From bc22c17e12c130dc929218a95aa347e0f3fd05dc Mon Sep 17 00:00:00 2001 From: Alain Knaff Date: Sun, 4 Jan 2009 22:46:16 +0100 Subject: bzip2/lzma: library support for gzip, bzip2 and lzma decompression Impact: Replaces inflate.c with a wrapper around zlib_inflate; new library code This is the first part of the bzip2/lzma patch The bzip patch is based on an idea by Christian Ludwig, includes support for compressing the kernel with bzip2 or lzma rather than gzip. Both compressors give smaller sizes than gzip. Lzma's decompresses faster than bzip2. It also supports ramdisks and initramfs' compressed using these two compressors. The functionality has been successfully used for a couple of years by the udpcast project This version applies to "tip" kernel 2.6.28 This part contains: - changed inflate.c to accomodate rest of patch - implementation of bzip2 compression (not used at this stage yet) - implementation of lzma compression (not used at this stage yet) - Makefile routines to support bzip2 and lzma kernel compression Signed-off-by: Alain Knaff Signed-off-by: H. Peter Anvin diff --git a/include/linux/decompress/bunzip2.h b/include/linux/decompress/bunzip2.h new file mode 100644 index 0000000..1152721 --- /dev/null +++ b/include/linux/decompress/bunzip2.h @@ -0,0 +1,10 @@ +#ifndef DECOMPRESS_BUNZIP2_H +#define DECOMPRESS_BUNZIP2_H + +int bunzip2(unsigned char *inbuf, int len, + int(*fill)(void*, unsigned int), + int(*flush)(void*, unsigned int), + unsigned char *output, + int *pos, + void(*error)(char *x)); +#endif diff --git a/include/linux/decompress/generic.h b/include/linux/decompress/generic.h new file mode 100644 index 0000000..f847f51 --- /dev/null +++ b/include/linux/decompress/generic.h @@ -0,0 +1,30 @@ +#ifndef DECOMPRESS_GENERIC_H +#define DECOMPRESS_GENERIC_H + +/* Minimal chunksize to be read. + *Bzip2 prefers at least 4096 + *Lzma prefers 0x10000 */ +#define COMPR_IOBUF_SIZE 4096 + +typedef int (*decompress_fn) (unsigned char *inbuf, int len, + int(*fill)(void*, unsigned int), + int(*writebb)(void*, unsigned int), + unsigned char *output, + int *posp, + void(*error)(char *x)); + +/* inbuf - input buffer + *len - len of pre-read data in inbuf + *fill - function to fill inbuf if empty + *writebb - function to write out outbug + *posp - if non-null, input position (number of bytes read) will be + * returned here + * + *If len != 0, the inbuf is initialized (with as much data), and fill + *should not be called + *If len = 0, the inbuf is allocated, but empty. Its size is IOBUF_SIZE + *fill should be called (repeatedly...) to read data, at most IOBUF_SIZE + */ + + +#endif diff --git a/include/linux/decompress/inflate.h b/include/linux/decompress/inflate.h new file mode 100644 index 0000000..f9b06cc --- /dev/null +++ b/include/linux/decompress/inflate.h @@ -0,0 +1,13 @@ +#ifndef INFLATE_H +#define INFLATE_H + +/* Other housekeeping constants */ +#define INBUFSIZ 4096 + +int gunzip(unsigned char *inbuf, int len, + int(*fill)(void*, unsigned int), + int(*flush)(void*, unsigned int), + unsigned char *output, + int *pos, + void(*error_fn)(char *x)); +#endif diff --git a/include/linux/decompress/mm.h b/include/linux/decompress/mm.h new file mode 100644 index 0000000..12ff8c3 --- /dev/null +++ b/include/linux/decompress/mm.h @@ -0,0 +1,87 @@ +/* + * linux/compr_mm.h + * + * Memory management for pre-boot and ramdisk uncompressors + * + * Authors: Alain Knaff + * + */ + +#ifndef DECOMPR_MM_H +#define DECOMPR_MM_H + +#ifdef STATIC + +/* Code active when included from pre-boot environment: */ + +/* A trivial malloc implementation, adapted from + * malloc by Hannu Savolainen 1993 and Matthias Urlichs 1994 + */ +static unsigned long malloc_ptr; +static int malloc_count; + +static void *malloc(int size) +{ + void *p; + + if (size < 0) + error("Malloc error"); + if (!malloc_ptr) + malloc_ptr = free_mem_ptr; + + malloc_ptr = (malloc_ptr + 3) & ~3; /* Align */ + + p = (void *)malloc_ptr; + malloc_ptr += size; + + if (free_mem_end_ptr && malloc_ptr >= free_mem_end_ptr) + error("Out of memory"); + + malloc_count++; + return p; +} + +static void free(void *where) +{ + malloc_count--; + if (!malloc_count) + malloc_ptr = free_mem_ptr; +} + +#define large_malloc(a) malloc(a) +#define large_free(a) free(a) + +#define set_error_fn(x) + +#define INIT + +#else /* STATIC */ + +/* Code active when compiled standalone for use when loading ramdisk: */ + +#include +#include +#include +#include + +/* Use defines rather than static inline in order to avoid spurious + * warnings when not needed (indeed large_malloc / large_free are not + * needed by inflate */ + +#define malloc(a) kmalloc(a, GFP_KERNEL) +#define free(a) kfree(a) + +#define large_malloc(a) vmalloc(a) +#define large_free(a) vfree(a) + +static void(*error)(char *m); +#define set_error_fn(x) error = x; + +#define INIT __init +#define STATIC + +#include + +#endif /* STATIC */ + +#endif /* DECOMPR_MM_H */ diff --git a/include/linux/decompress/unlzma.h b/include/linux/decompress/unlzma.h new file mode 100644 index 0000000..7796538 --- /dev/null +++ b/include/linux/decompress/unlzma.h @@ -0,0 +1,12 @@ +#ifndef DECOMPRESS_UNLZMA_H +#define DECOMPRESS_UNLZMA_H + +int unlzma(unsigned char *, int, + int(*fill)(void*, unsigned int), + int(*flush)(void*, unsigned int), + unsigned char *output, + int *posp, + void(*error)(char *x) + ); + +#endif diff --git a/lib/decompress_bunzip2.c b/lib/decompress_bunzip2.c new file mode 100644 index 0000000..5d3ddb5 --- /dev/null +++ b/lib/decompress_bunzip2.c @@ -0,0 +1,735 @@ +/* vi: set sw = 4 ts = 4: */ +/* Small bzip2 deflate implementation, by Rob Landley (rob@landley.net). + + Based on bzip2 decompression code by Julian R Seward (jseward@acm.org), + which also acknowledges contributions by Mike Burrows, David Wheeler, + Peter Fenwick, Alistair Moffat, Radford Neal, Ian H. Witten, + Robert Sedgewick, and Jon L. Bentley. + + This code is licensed under the LGPLv2: + LGPL (http://www.gnu.org/copyleft/lgpl.html +*/ + +/* + Size and speed optimizations by Manuel Novoa III (mjn3@codepoet.org). + + More efficient reading of Huffman codes, a streamlined read_bunzip() + function, and various other tweaks. In (limited) tests, approximately + 20% faster than bzcat on x86 and about 10% faster on arm. + + Note that about 2/3 of the time is spent in read_unzip() reversing + the Burrows-Wheeler transformation. Much of that time is delay + resulting from cache misses. + + I would ask that anyone benefiting from this work, especially those + using it in commercial products, consider making a donation to my local + non-profit hospice organization in the name of the woman I loved, who + passed away Feb. 12, 2003. + + In memory of Toni W. Hagan + + Hospice of Acadiana, Inc. + 2600 Johnston St., Suite 200 + Lafayette, LA 70503-3240 + + Phone (337) 232-1234 or 1-800-738-2226 + Fax (337) 232-1297 + + http://www.hospiceacadiana.com/ + + Manuel + */ + +/* + Made it fit for running in Linux Kernel by Alain Knaff (alain@knaff.lu) +*/ + + +#ifndef STATIC +#include +#endif /* !STATIC */ + +#include + +#ifndef INT_MAX +#define INT_MAX 0x7fffffff +#endif + +/* Constants for Huffman coding */ +#define MAX_GROUPS 6 +#define GROUP_SIZE 50 /* 64 would have been more efficient */ +#define MAX_HUFCODE_BITS 20 /* Longest Huffman code allowed */ +#define MAX_SYMBOLS 258 /* 256 literals + RUNA + RUNB */ +#define SYMBOL_RUNA 0 +#define SYMBOL_RUNB 1 + +/* Status return values */ +#define RETVAL_OK 0 +#define RETVAL_LAST_BLOCK (-1) +#define RETVAL_NOT_BZIP_DATA (-2) +#define RETVAL_UNEXPECTED_INPUT_EOF (-3) +#define RETVAL_UNEXPECTED_OUTPUT_EOF (-4) +#define RETVAL_DATA_ERROR (-5) +#define RETVAL_OUT_OF_MEMORY (-6) +#define RETVAL_OBSOLETE_INPUT (-7) + +/* Other housekeeping constants */ +#define BZIP2_IOBUF_SIZE 4096 + +/* This is what we know about each Huffman coding group */ +struct group_data { + /* We have an extra slot at the end of limit[] for a sentinal value. */ + int limit[MAX_HUFCODE_BITS+1]; + int base[MAX_HUFCODE_BITS]; + int permute[MAX_SYMBOLS]; + int minLen, maxLen; +}; + +/* Structure holding all the housekeeping data, including IO buffers and + memory that persists between calls to bunzip */ +struct bunzip_data { + /* State for interrupting output loop */ + int writeCopies, writePos, writeRunCountdown, writeCount, writeCurrent; + /* I/O tracking data (file handles, buffers, positions, etc.) */ + int (*fill)(void*, unsigned int); + int inbufCount, inbufPos /*, outbufPos*/; + unsigned char *inbuf /*,*outbuf*/; + unsigned int inbufBitCount, inbufBits; + /* The CRC values stored in the block header and calculated from the + data */ + unsigned int crc32Table[256], headerCRC, totalCRC, writeCRC; + /* Intermediate buffer and its size (in bytes) */ + unsigned int *dbuf, dbufSize; + /* These things are a bit too big to go on the stack */ + unsigned char selectors[32768]; /* nSelectors = 15 bits */ + struct group_data groups[MAX_GROUPS]; /* Huffman coding tables */ + int io_error; /* non-zero if we have IO error */ +}; + + +/* Return the next nnn bits of input. All reads from the compressed input + are done through this function. All reads are big endian */ +static unsigned int INIT get_bits(struct bunzip_data *bd, char bits_wanted) +{ + unsigned int bits = 0; + + /* If we need to get more data from the byte buffer, do so. + (Loop getting one byte at a time to enforce endianness and avoid + unaligned access.) */ + while (bd->inbufBitCount < bits_wanted) { + /* If we need to read more data from file into byte buffer, do + so */ + if (bd->inbufPos == bd->inbufCount) { + if (bd->io_error) + return 0; + bd->inbufCount = bd->fill(bd->inbuf, BZIP2_IOBUF_SIZE); + if (bd->inbufCount <= 0) { + bd->io_error = RETVAL_UNEXPECTED_INPUT_EOF; + return 0; + } + bd->inbufPos = 0; + } + /* Avoid 32-bit overflow (dump bit buffer to top of output) */ + if (bd->inbufBitCount >= 24) { + bits = bd->inbufBits&((1 << bd->inbufBitCount)-1); + bits_wanted -= bd->inbufBitCount; + bits <<= bits_wanted; + bd->inbufBitCount = 0; + } + /* Grab next 8 bits of input from buffer. */ + bd->inbufBits = (bd->inbufBits << 8)|bd->inbuf[bd->inbufPos++]; + bd->inbufBitCount += 8; + } + /* Calculate result */ + bd->inbufBitCount -= bits_wanted; + bits |= (bd->inbufBits >> bd->inbufBitCount)&((1 << bits_wanted)-1); + + return bits; +} + +/* Unpacks the next block and sets up for the inverse burrows-wheeler step. */ + +static int INIT get_next_block(struct bunzip_data *bd) +{ + struct group_data *hufGroup = NULL; + int *base = NULL; + int *limit = NULL; + int dbufCount, nextSym, dbufSize, groupCount, selector, + i, j, k, t, runPos, symCount, symTotal, nSelectors, + byteCount[256]; + unsigned char uc, symToByte[256], mtfSymbol[256], *selectors; + unsigned int *dbuf, origPtr; + + dbuf = bd->dbuf; + dbufSize = bd->dbufSize; + selectors = bd->selectors; + + /* Read in header signature and CRC, then validate signature. + (last block signature means CRC is for whole file, return now) */ + i = get_bits(bd, 24); + j = get_bits(bd, 24); + bd->headerCRC = get_bits(bd, 32); + if ((i == 0x177245) && (j == 0x385090)) + return RETVAL_LAST_BLOCK; + if ((i != 0x314159) || (j != 0x265359)) + return RETVAL_NOT_BZIP_DATA; + /* We can add support for blockRandomised if anybody complains. + There was some code for this in busybox 1.0.0-pre3, but nobody ever + noticed that it didn't actually work. */ + if (get_bits(bd, 1)) + return RETVAL_OBSOLETE_INPUT; + origPtr = get_bits(bd, 24); + if (origPtr > dbufSize) + return RETVAL_DATA_ERROR; + /* mapping table: if some byte values are never used (encoding things + like ascii text), the compression code removes the gaps to have fewer + symbols to deal with, and writes a sparse bitfield indicating which + values were present. We make a translation table to convert the + symbols back to the corresponding bytes. */ + t = get_bits(bd, 16); + symTotal = 0; + for (i = 0; i < 16; i++) { + if (t&(1 << (15-i))) { + k = get_bits(bd, 16); + for (j = 0; j < 16; j++) + if (k&(1 << (15-j))) + symToByte[symTotal++] = (16*i)+j; + } + } + /* How many different Huffman coding groups does this block use? */ + groupCount = get_bits(bd, 3); + if (groupCount < 2 || groupCount > MAX_GROUPS) + return RETVAL_DATA_ERROR; + /* nSelectors: Every GROUP_SIZE many symbols we select a new + Huffman coding group. Read in the group selector list, + which is stored as MTF encoded bit runs. (MTF = Move To + Front, as each value is used it's moved to the start of the + list.) */ + nSelectors = get_bits(bd, 15); + if (!nSelectors) + return RETVAL_DATA_ERROR; + for (i = 0; i < groupCount; i++) + mtfSymbol[i] = i; + for (i = 0; i < nSelectors; i++) { + /* Get next value */ + for (j = 0; get_bits(bd, 1); j++) + if (j >= groupCount) + return RETVAL_DATA_ERROR; + /* Decode MTF to get the next selector */ + uc = mtfSymbol[j]; + for (; j; j--) + mtfSymbol[j] = mtfSymbol[j-1]; + mtfSymbol[0] = selectors[i] = uc; + } + /* Read the Huffman coding tables for each group, which code + for symTotal literal symbols, plus two run symbols (RUNA, + RUNB) */ + symCount = symTotal+2; + for (j = 0; j < groupCount; j++) { + unsigned char length[MAX_SYMBOLS], temp[MAX_HUFCODE_BITS+1]; + int minLen, maxLen, pp; + /* Read Huffman code lengths for each symbol. They're + stored in a way similar to mtf; record a starting + value for the first symbol, and an offset from the + previous value for everys symbol after that. + (Subtracting 1 before the loop and then adding it + back at the end is an optimization that makes the + test inside the loop simpler: symbol length 0 + becomes negative, so an unsigned inequality catches + it.) */ + t = get_bits(bd, 5)-1; + for (i = 0; i < symCount; i++) { + for (;;) { + if (((unsigned)t) > (MAX_HUFCODE_BITS-1)) + return RETVAL_DATA_ERROR; + + /* If first bit is 0, stop. Else + second bit indicates whether to + increment or decrement the value. + Optimization: grab 2 bits and unget + the second if the first was 0. */ + + k = get_bits(bd, 2); + if (k < 2) { + bd->inbufBitCount++; + break; + } + /* Add one if second bit 1, else + * subtract 1. Avoids if/else */ + t += (((k+1)&2)-1); + } + /* Correct for the initial -1, to get the + * final symbol length */ + length[i] = t+1; + } + /* Find largest and smallest lengths in this group */ + minLen = maxLen = length[0]; + + for (i = 1; i < symCount; i++) { + if (length[i] > maxLen) + maxLen = length[i]; + else if (length[i] < minLen) + minLen = length[i]; + } + + /* Calculate permute[], base[], and limit[] tables from + * length[]. + * + * permute[] is the lookup table for converting + * Huffman coded symbols into decoded symbols. base[] + * is the amount to subtract from the value of a + * Huffman symbol of a given length when using + * permute[]. + * + * limit[] indicates the largest numerical value a + * symbol with a given number of bits can have. This + * is how the Huffman codes can vary in length: each + * code with a value > limit[length] needs another + * bit. + */ + hufGroup = bd->groups+j; + hufGroup->minLen = minLen; + hufGroup->maxLen = maxLen; + /* Note that minLen can't be smaller than 1, so we + adjust the base and limit array pointers so we're + not always wasting the first entry. We do this + again when using them (during symbol decoding).*/ + base = hufGroup->base-1; + limit = hufGroup->limit-1; + /* Calculate permute[]. Concurently, initialize + * temp[] and limit[]. */ + pp = 0; + for (i = minLen; i <= maxLen; i++) { + temp[i] = limit[i] = 0; + for (t = 0; t < symCount; t++) + if (length[t] == i) + hufGroup->permute[pp++] = t; + } + /* Count symbols coded for at each bit length */ + for (i = 0; i < symCount; i++) + temp[length[i]]++; + /* Calculate limit[] (the largest symbol-coding value + *at each bit length, which is (previous limit << + *1)+symbols at this level), and base[] (number of + *symbols to ignore at each bit length, which is limit + *minus the cumulative count of symbols coded for + *already). */ + pp = t = 0; + for (i = minLen; i < maxLen; i++) { + pp += temp[i]; + /* We read the largest possible symbol size + and then unget bits after determining how + many we need, and those extra bits could be + set to anything. (They're noise from + future symbols.) At each level we're + really only interested in the first few + bits, so here we set all the trailing + to-be-ignored bits to 1 so they don't + affect the value > limit[length] + comparison. */ + limit[i] = (pp << (maxLen - i)) - 1; + pp <<= 1; + base[i+1] = pp-(t += temp[i]); + } + limit[maxLen+1] = INT_MAX; /* Sentinal value for + * reading next sym. */ + limit[maxLen] = pp+temp[maxLen]-1; + base[minLen] = 0; + } + /* We've finished reading and digesting the block header. Now + read this block's Huffman coded symbols from the file and + undo the Huffman coding and run length encoding, saving the + result into dbuf[dbufCount++] = uc */ + + /* Initialize symbol occurrence counters and symbol Move To + * Front table */ + for (i = 0; i < 256; i++) { + byteCount[i] = 0; + mtfSymbol[i] = (unsigned char)i; + } + /* Loop through compressed symbols. */ + runPos = dbufCount = symCount = selector = 0; + for (;;) { + /* Determine which Huffman coding group to use. */ + if (!(symCount--)) { + symCount = GROUP_SIZE-1; + if (selector >= nSelectors) + return RETVAL_DATA_ERROR; + hufGroup = bd->groups+selectors[selector++]; + base = hufGroup->base-1; + limit = hufGroup->limit-1; + } + /* Read next Huffman-coded symbol. */ + /* Note: It is far cheaper to read maxLen bits and + back up than it is to read minLen bits and then an + additional bit at a time, testing as we go. + Because there is a trailing last block (with file + CRC), there is no danger of the overread causing an + unexpected EOF for a valid compressed file. As a + further optimization, we do the read inline + (falling back to a call to get_bits if the buffer + runs dry). The following (up to got_huff_bits:) is + equivalent to j = get_bits(bd, hufGroup->maxLen); + */ + while (bd->inbufBitCount < hufGroup->maxLen) { + if (bd->inbufPos == bd->inbufCount) { + j = get_bits(bd, hufGroup->maxLen); + goto got_huff_bits; + } + bd->inbufBits = + (bd->inbufBits << 8)|bd->inbuf[bd->inbufPos++]; + bd->inbufBitCount += 8; + }; + bd->inbufBitCount -= hufGroup->maxLen; + j = (bd->inbufBits >> bd->inbufBitCount)& + ((1 << hufGroup->maxLen)-1); +got_huff_bits: + /* Figure how how many bits are in next symbol and + * unget extras */ + i = hufGroup->minLen; + while (j > limit[i]) + ++i; + bd->inbufBitCount += (hufGroup->maxLen - i); + /* Huffman decode value to get nextSym (with bounds checking) */ + if ((i > hufGroup->maxLen) + || (((unsigned)(j = (j>>(hufGroup->maxLen-i))-base[i])) + >= MAX_SYMBOLS)) + return RETVAL_DATA_ERROR; + nextSym = hufGroup->permute[j]; + /* We have now decoded the symbol, which indicates + either a new literal byte, or a repeated run of the + most recent literal byte. First, check if nextSym + indicates a repeated run, and if so loop collecting + how many times to repeat the last literal. */ + if (((unsigned)nextSym) <= SYMBOL_RUNB) { /* RUNA or RUNB */ + /* If this is the start of a new run, zero out + * counter */ + if (!runPos) { + runPos = 1; + t = 0; + } + /* Neat trick that saves 1 symbol: instead of + or-ing 0 or 1 at each bit position, add 1 + or 2 instead. For example, 1011 is 1 << 0 + + 1 << 1 + 2 << 2. 1010 is 2 << 0 + 2 << 1 + + 1 << 2. You can make any bit pattern + that way using 1 less symbol than the basic + or 0/1 method (except all bits 0, which + would use no symbols, but a run of length 0 + doesn't mean anything in this context). + Thus space is saved. */ + t += (runPos << nextSym); + /* +runPos if RUNA; +2*runPos if RUNB */ + + runPos <<= 1; + continue; + } + /* When we hit the first non-run symbol after a run, + we now know how many times to repeat the last + literal, so append that many copies to our buffer + of decoded symbols (dbuf) now. (The last literal + used is the one at the head of the mtfSymbol + array.) */ + if (runPos) { + runPos = 0; + if (dbufCount+t >= dbufSize) + return RETVAL_DATA_ERROR; + + uc = symToByte[mtfSymbol[0]]; + byteCount[uc] += t; + while (t--) + dbuf[dbufCount++] = uc; + } + /* Is this the terminating symbol? */ + if (nextSym > symTotal) + break; + /* At this point, nextSym indicates a new literal + character. Subtract one to get the position in the + MTF array at which this literal is currently to be + found. (Note that the result can't be -1 or 0, + because 0 and 1 are RUNA and RUNB. But another + instance of the first symbol in the mtf array, + position 0, would have been handled as part of a + run above. Therefore 1 unused mtf position minus 2 + non-literal nextSym values equals -1.) */ + if (dbufCount >= dbufSize) + return RETVAL_DATA_ERROR; + i = nextSym - 1; + uc = mtfSymbol[i]; + /* Adjust the MTF array. Since we typically expect to + *move only a small number of symbols, and are bound + *by 256 in any case, using memmove here would + *typically be bigger and slower due to function call + *overhead and other assorted setup costs. */ + do { + mtfSymbol[i] = mtfSymbol[i-1]; + } while (--i); + mtfSymbol[0] = uc; + uc = symToByte[uc]; + /* We have our literal byte. Save it into dbuf. */ + byteCount[uc]++; + dbuf[dbufCount++] = (unsigned int)uc; + } + /* At this point, we've read all the Huffman-coded symbols + (and repeated runs) for this block from the input stream, + and decoded them into the intermediate buffer. There are + dbufCount many decoded bytes in dbuf[]. Now undo the + Burrows-Wheeler transform on dbuf. See + http://dogma.net/markn/articles/bwt/bwt.htm + */ + /* Turn byteCount into cumulative occurrence counts of 0 to n-1. */ + j = 0; + for (i = 0; i < 256; i++) { + k = j+byteCount[i]; + byteCount[i] = j; + j = k; + } + /* Figure out what order dbuf would be in if we sorted it. */ + for (i = 0; i < dbufCount; i++) { + uc = (unsigned char)(dbuf[i] & 0xff); + dbuf[byteCount[uc]] |= (i << 8); + byteCount[uc]++; + } + /* Decode first byte by hand to initialize "previous" byte. + Note that it doesn't get output, and if the first three + characters are identical it doesn't qualify as a run (hence + writeRunCountdown = 5). */ + if (dbufCount) { + if (origPtr >= dbufCount) + return RETVAL_DATA_ERROR; + bd->writePos = dbuf[origPtr]; + bd->writeCurrent = (unsigned char)(bd->writePos&0xff); + bd->writePos >>= 8; + bd->writeRunCountdown = 5; + } + bd->writeCount = dbufCount; + + return RETVAL_OK; +} + +/* Undo burrows-wheeler transform on intermediate buffer to produce output. + If start_bunzip was initialized with out_fd =-1, then up to len bytes of + data are written to outbuf. Return value is number of bytes written or + error (all errors are negative numbers). If out_fd!=-1, outbuf and len + are ignored, data is written to out_fd and return is RETVAL_OK or error. +*/ + +static int INIT read_bunzip(struct bunzip_data *bd, char *outbuf, int len) +{ + const unsigned int *dbuf; + int pos, xcurrent, previous, gotcount; + + /* If last read was short due to end of file, return last block now */ + if (bd->writeCount < 0) + return bd->writeCount; + + gotcount = 0; + dbuf = bd->dbuf; + pos = bd->writePos; + xcurrent = bd->writeCurrent; + + /* We will always have pending decoded data to write into the output + buffer unless this is the very first call (in which case we haven't + Huffman-decoded a block into the intermediate buffer yet). */ + + if (bd->writeCopies) { + /* Inside the loop, writeCopies means extra copies (beyond 1) */ + --bd->writeCopies; + /* Loop outputting bytes */ + for (;;) { + /* If the output buffer is full, snapshot + * state and return */ + if (gotcount >= len) { + bd->writePos = pos; + bd->writeCurrent = xcurrent; + bd->writeCopies++; + return len; + } + /* Write next byte into output buffer, updating CRC */ + outbuf[gotcount++] = xcurrent; + bd->writeCRC = (((bd->writeCRC) << 8) + ^bd->crc32Table[((bd->writeCRC) >> 24) + ^xcurrent]); + /* Loop now if we're outputting multiple + * copies of this byte */ + if (bd->writeCopies) { + --bd->writeCopies; + continue; + } +decode_next_byte: + if (!bd->writeCount--) + break; + /* Follow sequence vector to undo + * Burrows-Wheeler transform */ + previous = xcurrent; + pos = dbuf[pos]; + xcurrent = pos&0xff; + pos >>= 8; + /* After 3 consecutive copies of the same + byte, the 4th is a repeat count. We count + down from 4 instead *of counting up because + testing for non-zero is faster */ + if (--bd->writeRunCountdown) { + if (xcurrent != previous) + bd->writeRunCountdown = 4; + } else { + /* We have a repeated run, this byte + * indicates the count */ + bd->writeCopies = xcurrent; + xcurrent = previous; + bd->writeRunCountdown = 5; + /* Sometimes there are just 3 bytes + * (run length 0) */ + if (!bd->writeCopies) + goto decode_next_byte; + /* Subtract the 1 copy we'd output + * anyway to get extras */ + --bd->writeCopies; + } + } + /* Decompression of this block completed successfully */ + bd->writeCRC = ~bd->writeCRC; + bd->totalCRC = ((bd->totalCRC << 1) | + (bd->totalCRC >> 31)) ^ bd->writeCRC; + /* If this block had a CRC error, force file level CRC error. */ + if (bd->writeCRC != bd->headerCRC) { + bd->totalCRC = bd->headerCRC+1; + return RETVAL_LAST_BLOCK; + } + } + + /* Refill the intermediate buffer by Huffman-decoding next + * block of input */ + /* (previous is just a convenient unused temp variable here) */ + previous = get_next_block(bd); + if (previous) { + bd->writeCount = previous; + return (previous != RETVAL_LAST_BLOCK) ? previous : gotcount; + } + bd->writeCRC = 0xffffffffUL; + pos = bd->writePos; + xcurrent = bd->writeCurrent; + goto decode_next_byte; +} + +static int INIT nofill(void *buf, unsigned int len) +{ + return -1; +} + +/* Allocate the structure, read file header. If in_fd ==-1, inbuf must contain + a complete bunzip file (len bytes long). If in_fd!=-1, inbuf and len are + ignored, and data is read from file handle into temporary buffer. */ +static int INIT start_bunzip(struct bunzip_data **bdp, void *inbuf, int len, + int (*fill)(void*, unsigned int)) +{ + struct bunzip_data *bd; + unsigned int i, j, c; + const unsigned int BZh0 = + (((unsigned int)'B') << 24)+(((unsigned int)'Z') << 16) + +(((unsigned int)'h') << 8)+(unsigned int)'0'; + + /* Figure out how much data to allocate */ + i = sizeof(struct bunzip_data); + + /* Allocate bunzip_data. Most fields initialize to zero. */ + bd = *bdp = malloc(i); + memset(bd, 0, sizeof(struct bunzip_data)); + /* Setup input buffer */ + bd->inbuf = inbuf; + bd->inbufCount = len; + if (fill != NULL) + bd->fill = fill; + else + bd->fill = nofill; + + /* Init the CRC32 table (big endian) */ + for (i = 0; i < 256; i++) { + c = i << 24; + for (j = 8; j; j--) + c = c&0x80000000 ? (c << 1)^0x04c11db7 : (c << 1); + bd->crc32Table[i] = c; + } + + /* Ensure that file starts with "BZh['1'-'9']." */ + i = get_bits(bd, 32); + if (((unsigned int)(i-BZh0-1)) >= 9) + return RETVAL_NOT_BZIP_DATA; + + /* Fourth byte (ascii '1'-'9'), indicates block size in units of 100k of + uncompressed data. Allocate intermediate buffer for block. */ + bd->dbufSize = 100000*(i-BZh0); + + bd->dbuf = large_malloc(bd->dbufSize * sizeof(int)); + return RETVAL_OK; +} + +/* Example usage: decompress src_fd to dst_fd. (Stops at end of bzip2 data, + not end of file.) */ +STATIC int INIT bunzip2(unsigned char *buf, int len, + int(*fill)(void*, unsigned int), + int(*flush)(void*, unsigned int), + unsigned char *outbuf, + int *pos, + void(*error_fn)(char *x)) +{ + struct bunzip_data *bd; + int i = -1; + unsigned char *inbuf; + + set_error_fn(error_fn); + if (flush) + outbuf = malloc(BZIP2_IOBUF_SIZE); + else + len -= 4; /* Uncompressed size hack active in pre-boot + environment */ + if (!outbuf) { + error("Could not allocate output bufer"); + return -1; + } + if (buf) + inbuf = buf; + else + inbuf = malloc(BZIP2_IOBUF_SIZE); + if (!inbuf) { + error("Could not allocate input bufer"); + goto exit_0; + } + i = start_bunzip(&bd, inbuf, len, fill); + if (!i) { + for (;;) { + i = read_bunzip(bd, outbuf, BZIP2_IOBUF_SIZE); + if (i <= 0) + break; + if (!flush) + outbuf += i; + else + if (i != flush(outbuf, i)) { + i = RETVAL_UNEXPECTED_OUTPUT_EOF; + break; + } + } + } + /* Check CRC and release memory */ + if (i == RETVAL_LAST_BLOCK) { + if (bd->headerCRC != bd->totalCRC) + error("Data integrity error when decompressing."); + else + i = RETVAL_OK; + } else if (i == RETVAL_UNEXPECTED_OUTPUT_EOF) { + error("Compressed file ends unexpectedly"); + } + if (bd->dbuf) + large_free(bd->dbuf); + if (pos) + *pos = bd->inbufPos; + free(bd); + if (!buf) + free(inbuf); +exit_0: + if (flush) + free(outbuf); + return i; +} + +#define decompress bunzip2 diff --git a/lib/decompress_inflate.c b/lib/decompress_inflate.c new file mode 100644 index 0000000..163e66a --- /dev/null +++ b/lib/decompress_inflate.c @@ -0,0 +1,167 @@ +#ifdef STATIC +/* Pre-boot environment: included */ + +/* prevent inclusion of _LINUX_KERNEL_H in pre-boot environment: lots + * errors about console_printk etc... on ARM */ +#define _LINUX_KERNEL_H + +#include "zlib_inflate/inftrees.c" +#include "zlib_inflate/inffast.c" +#include "zlib_inflate/inflate.c" + +#else /* STATIC */ +/* initramfs et al: linked */ + +#include + +#include "zlib_inflate/inftrees.h" +#include "zlib_inflate/inffast.h" +#include "zlib_inflate/inflate.h" + +#include "zlib_inflate/infutil.h" + +#endif /* STATIC */ + +#include + +#define INBUF_LEN (16*1024) + +/* Included from initramfs et al code */ +STATIC int INIT gunzip(unsigned char *buf, int len, + int(*fill)(void*, unsigned int), + int(*flush)(void*, unsigned int), + unsigned char *out_buf, + int *pos, + void(*error_fn)(char *x)) { + u8 *zbuf; + struct z_stream_s *strm; + int rc; + size_t out_len; + + set_error_fn(error_fn); + rc = -1; + if (flush) { + out_len = 0x8100; /* 32 K */ + out_buf = malloc(out_len); + } else { + out_len = 0x7fffffff; /* no limit */ + } + if (!out_buf) { + error("Out of memory while allocating output buffer"); + goto gunzip_nomem1; + } + + if (buf) + zbuf = buf; + else { + zbuf = malloc(INBUF_LEN); + len = 0; + } + if (!zbuf) { + error("Out of memory while allocating input buffer"); + goto gunzip_nomem2; + } + + strm = malloc(sizeof(*strm)); + if (strm == NULL) { + error("Out of memory while allocating z_stream"); + goto gunzip_nomem3; + } + + strm->workspace = malloc(flush ? zlib_inflate_workspacesize() : + sizeof(struct inflate_state)); + if (strm->workspace == NULL) { + error("Out of memory while allocating workspace"); + goto gunzip_nomem4; + } + + if (len == 0) + len = fill(zbuf, INBUF_LEN); + + /* verify the gzip header */ + if (len < 10 || + zbuf[0] != 0x1f || zbuf[1] != 0x8b || zbuf[2] != 0x08) { + if (pos) + *pos = 0; + error("Not a gzip file"); + goto gunzip_5; + } + + /* skip over gzip header (1f,8b,08... 10 bytes total + + * possible asciz filename) + */ + strm->next_in = zbuf + 10; + /* skip over asciz filename */ + if (zbuf[3] & 0x8) { + while (strm->next_in[0]) + strm->next_in++; + strm->next_in++; + } + strm->avail_in = len - 10; + + strm->next_out = out_buf; + strm->avail_out = out_len; + + rc = zlib_inflateInit2(strm, -MAX_WBITS); + + if (!flush) { + WS(strm)->inflate_state.wsize = 0; + WS(strm)->inflate_state.window = NULL; + } + + while (rc == Z_OK) { + if (strm->avail_in == 0) { + /* TODO: handle case where both pos and fill are set */ + len = fill(zbuf, INBUF_LEN); + if (len < 0) { + rc = -1; + error("read error"); + break; + } + strm->next_in = zbuf; + strm->avail_in = len; + } + rc = zlib_inflate(strm, 0); + + /* Write any data generated */ + if (flush && strm->next_out > out_buf) { + int l = strm->next_out - out_buf; + if (l != flush(out_buf, l)) { + rc = -1; + error("write error"); + break; + } + strm->next_out = out_buf; + strm->avail_out = out_len; + } + + /* after Z_FINISH, only Z_STREAM_END is "we unpacked it all" */ + if (rc == Z_STREAM_END) { + rc = 0; + break; + } else if (rc != Z_OK) { + error("uncompression error"); + rc = -1; + } + } + + zlib_inflateEnd(strm); + if (pos) + /* add + 8 to skip over trailer */ + *pos = strm->next_in - zbuf+8; + +gunzip_5: + free(strm->workspace); +gunzip_nomem4: + free(strm); +gunzip_nomem3: + if (!buf) + free(zbuf); +gunzip_nomem2: + if (flush) + free(out_buf); +gunzip_nomem1: + return rc; /* returns Z_OK (0) if successful */ +} + +#define decompress gunzip diff --git a/lib/decompress_unlzma.c b/lib/decompress_unlzma.c new file mode 100644 index 0000000..546f2f4 --- /dev/null +++ b/lib/decompress_unlzma.c @@ -0,0 +1,647 @@ +/* Lzma decompressor for Linux kernel. Shamelessly snarfed + *from busybox 1.1.1 + * + *Linux kernel adaptation + *Copyright (C) 2006 Alain < alain@knaff.lu > + * + *Based on small lzma deflate implementation/Small range coder + *implementation for lzma. + *Copyright (C) 2006 Aurelien Jacobs < aurel@gnuage.org > + * + *Based on LzmaDecode.c from the LZMA SDK 4.22 (http://www.7-zip.org/) + *Copyright (C) 1999-2005 Igor Pavlov + * + *Copyrights of the parts, see headers below. + * + * + *This program is free software; you can redistribute it and/or + *modify it under the terms of the GNU Lesser General Public + *License as published by the Free Software Foundation; either + *version 2.1 of the License, or (at your option) any later version. + * + *This program is distributed in the hope that it will be useful, + *but WITHOUT ANY WARRANTY; without even the implied warranty of + *MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + *Lesser General Public License for more details. + * + *You should have received a copy of the GNU Lesser General Public + *License along with this library; if not, write to the Free Software + *Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA + */ + +#ifndef STATIC +#include +#endif /* STATIC */ + +#include + +#define MIN(a, b) (((a) < (b)) ? (a) : (b)) + +static long long INIT read_int(unsigned char *ptr, int size) +{ + int i; + long long ret = 0; + + for (i = 0; i < size; i++) + ret = (ret << 8) | ptr[size-i-1]; + return ret; +} + +#define ENDIAN_CONVERT(x) \ + x = (typeof(x))read_int((unsigned char *)&x, sizeof(x)) + + +/* Small range coder implementation for lzma. + *Copyright (C) 2006 Aurelien Jacobs < aurel@gnuage.org > + * + *Based on LzmaDecode.c from the LZMA SDK 4.22 (http://www.7-zip.org/) + *Copyright (c) 1999-2005 Igor Pavlov + */ + +#include + +#define LZMA_IOBUF_SIZE 0x10000 + +struct rc { + int (*fill)(void*, unsigned int); + uint8_t *ptr; + uint8_t *buffer; + uint8_t *buffer_end; + int buffer_size; + uint32_t code; + uint32_t range; + uint32_t bound; +}; + + +#define RC_TOP_BITS 24 +#define RC_MOVE_BITS 5 +#define RC_MODEL_TOTAL_BITS 11 + + +/* Called twice: once at startup and once in rc_normalize() */ +static void INIT rc_read(struct rc *rc) +{ + rc->buffer_size = rc->fill((char *)rc->buffer, LZMA_IOBUF_SIZE); + if (rc->buffer_size <= 0) + error("unexpected EOF"); + rc->ptr = rc->buffer; + rc->buffer_end = rc->buffer + rc->buffer_size; +} + +/* Called once */ +static inline void INIT rc_init(struct rc *rc, + int (*fill)(void*, unsigned int), + char *buffer, int buffer_size) +{ + rc->fill = fill; + rc->buffer = (uint8_t *)buffer; + rc->buffer_size = buffer_size; + rc->buffer_end = rc->buffer + rc->buffer_size; + rc->ptr = rc->buffer; + + rc->code = 0; + rc->range = 0xFFFFFFFF; +} + +static inline void INIT rc_init_code(struct rc *rc) +{ + int i; + + for (i = 0; i < 5; i++) { + if (rc->ptr >= rc->buffer_end) + rc_read(rc); + rc->code = (rc->code << 8) | *rc->ptr++; + } +} + + +/* Called once. TODO: bb_maybe_free() */ +static inline void INIT rc_free(struct rc *rc) +{ + free(rc->buffer); +} + +/* Called twice, but one callsite is in inline'd rc_is_bit_0_helper() */ +static void INIT rc_do_normalize(struct rc *rc) +{ + if (rc->ptr >= rc->buffer_end) + rc_read(rc); + rc->range <<= 8; + rc->code = (rc->code << 8) | *rc->ptr++; +} +static inline void INIT rc_normalize(struct rc *rc) +{ + if (rc->range < (1 << RC_TOP_BITS)) + rc_do_normalize(rc); +} + +/* Called 9 times */ +/* Why rc_is_bit_0_helper exists? + *Because we want to always expose (rc->code < rc->bound) to optimizer + */ +static inline uint32_t INIT rc_is_bit_0_helper(struct rc *rc, uint16_t *p) +{ + rc_normalize(rc); + rc->bound = *p * (rc->range >> RC_MODEL_TOTAL_BITS); + return rc->bound; +} +static inline int INIT rc_is_bit_0(struct rc *rc, uint16_t *p) +{ + uint32_t t = rc_is_bit_0_helper(rc, p); + return rc->code < t; +} + +/* Called ~10 times, but very small, thus inlined */ +static inline void INIT rc_update_bit_0(struct rc *rc, uint16_t *p) +{ + rc->range = rc->bound; + *p += ((1 << RC_MODEL_TOTAL_BITS) - *p) >> RC_MOVE_BITS; +} +static inline void rc_update_bit_1(struct rc *rc, uint16_t *p) +{ + rc->range -= rc->bound; + rc->code -= rc->bound; + *p -= *p >> RC_MOVE_BITS; +} + +/* Called 4 times in unlzma loop */ +static int INIT rc_get_bit(struct rc *rc, uint16_t *p, int *symbol) +{ + if (rc_is_bit_0(rc, p)) { + rc_update_bit_0(rc, p); + *symbol *= 2; + return 0; + } else { + rc_update_bit_1(rc, p); + *symbol = *symbol * 2 + 1; + return 1; + } +} + +/* Called once */ +static inline int INIT rc_direct_bit(struct rc *rc) +{ + rc_normalize(rc); + rc->range >>= 1; + if (rc->code >= rc->range) { + rc->code -= rc->range; + return 1; + } + return 0; +} + +/* Called twice */ +static inline void INIT +rc_bit_tree_decode(struct rc *rc, uint16_t *p, int num_levels, int *symbol) +{ + int i = num_levels; + + *symbol = 1; + while (i--) + rc_get_bit(rc, p + *symbol, symbol); + *symbol -= 1 << num_levels; +} + + +/* + * Small lzma deflate implementation. + * Copyright (C) 2006 Aurelien Jacobs < aurel@gnuage.org > + * + * Based on LzmaDecode.c from the LZMA SDK 4.22 (http://www.7-zip.org/) + * Copyright (C) 1999-2005 Igor Pavlov + */ + + +struct lzma_header { + uint8_t pos; + uint32_t dict_size; + uint64_t dst_size; +} __attribute__ ((packed)) ; + + +#define LZMA_BASE_SIZE 1846 +#define LZMA_LIT_SIZE 768 + +#define LZMA_NUM_POS_BITS_MAX 4 + +#define LZMA_LEN_NUM_LOW_BITS 3 +#define LZMA_LEN_NUM_MID_BITS 3 +#define LZMA_LEN_NUM_HIGH_BITS 8 + +#define LZMA_LEN_CHOICE 0 +#define LZMA_LEN_CHOICE_2 (LZMA_LEN_CHOICE + 1) +#define LZMA_LEN_LOW (LZMA_LEN_CHOICE_2 + 1) +#define LZMA_LEN_MID (LZMA_LEN_LOW \ + + (1 << (LZMA_NUM_POS_BITS_MAX + LZMA_LEN_NUM_LOW_BITS))) +#define LZMA_LEN_HIGH (LZMA_LEN_MID \ + +(1 << (LZMA_NUM_POS_BITS_MAX + LZMA_LEN_NUM_MID_BITS))) +#define LZMA_NUM_LEN_PROBS (LZMA_LEN_HIGH + (1 << LZMA_LEN_NUM_HIGH_BITS)) + +#define LZMA_NUM_STATES 12 +#define LZMA_NUM_LIT_STATES 7 + +#define LZMA_START_POS_MODEL_INDEX 4 +#define LZMA_END_POS_MODEL_INDEX 14 +#define LZMA_NUM_FULL_DISTANCES (1 << (LZMA_END_POS_MODEL_INDEX >> 1)) + +#define LZMA_NUM_POS_SLOT_BITS 6 +#define LZMA_NUM_LEN_TO_POS_STATES 4 + +#define LZMA_NUM_ALIGN_BITS 4 + +#define LZMA_MATCH_MIN_LEN 2 + +#define LZMA_IS_MATCH 0 +#define LZMA_IS_REP (LZMA_IS_MATCH + (LZMA_NUM_STATES << LZMA_NUM_POS_BITS_MAX)) +#define LZMA_IS_REP_G0 (LZMA_IS_REP + LZMA_NUM_STATES) +#define LZMA_IS_REP_G1 (LZMA_IS_REP_G0 + LZMA_NUM_STATES) +#define LZMA_IS_REP_G2 (LZMA_IS_REP_G1 + LZMA_NUM_STATES) +#define LZMA_IS_REP_0_LONG (LZMA_IS_REP_G2 + LZMA_NUM_STATES) +#define LZMA_POS_SLOT (LZMA_IS_REP_0_LONG \ + + (LZMA_NUM_STATES << LZMA_NUM_POS_BITS_MAX)) +#define LZMA_SPEC_POS (LZMA_POS_SLOT \ + +(LZMA_NUM_LEN_TO_POS_STATES << LZMA_NUM_POS_SLOT_BITS)) +#define LZMA_ALIGN (LZMA_SPEC_POS \ + + LZMA_NUM_FULL_DISTANCES - LZMA_END_POS_MODEL_INDEX) +#define LZMA_LEN_CODER (LZMA_ALIGN + (1 << LZMA_NUM_ALIGN_BITS)) +#define LZMA_REP_LEN_CODER (LZMA_LEN_CODER + LZMA_NUM_LEN_PROBS) +#define LZMA_LITERAL (LZMA_REP_LEN_CODER + LZMA_NUM_LEN_PROBS) + + +struct writer { + uint8_t *buffer; + uint8_t previous_byte; + size_t buffer_pos; + int bufsize; + size_t global_pos; + int(*flush)(void*, unsigned int); + struct lzma_header *header; +}; + +struct cstate { + int state; + uint32_t rep0, rep1, rep2, rep3; +}; + +static inline size_t INIT get_pos(struct writer *wr) +{ + return + wr->global_pos + wr->buffer_pos; +} + +static inline uint8_t INIT peek_old_byte(struct writer *wr, + uint32_t offs) +{ + if (!wr->flush) { + int32_t pos; + while (offs > wr->header->dict_size) + offs -= wr->header->dict_size; + pos = wr->buffer_pos - offs; + return wr->buffer[pos]; + } else { + uint32_t pos = wr->buffer_pos - offs; + while (pos >= wr->header->dict_size) + pos += wr->header->dict_size; + return wr->buffer[pos]; + } + +} + +static inline void INIT write_byte(struct writer *wr, uint8_t byte) +{ + wr->buffer[wr->buffer_pos++] = wr->previous_byte = byte; + if (wr->flush && wr->buffer_pos == wr->header->dict_size) { + wr->buffer_pos = 0; + wr->global_pos += wr->header->dict_size; + wr->flush((char *)wr->buffer, wr->header->dict_size); + } +} + + +static inline void INIT copy_byte(struct writer *wr, uint32_t offs) +{ + write_byte(wr, peek_old_byte(wr, offs)); +} + +static inline void INIT copy_bytes(struct writer *wr, + uint32_t rep0, int len) +{ + do { + copy_byte(wr, rep0); + len--; + } while (len != 0 && wr->buffer_pos < wr->header->dst_size); +} + +static inline void INIT process_bit0(struct writer *wr, struct rc *rc, + struct cstate *cst, uint16_t *p, + int pos_state, uint16_t *prob, + int lc, uint32_t literal_pos_mask) { + int mi = 1; + rc_update_bit_0(rc, prob); + prob = (p + LZMA_LITERAL + + (LZMA_LIT_SIZE + * (((get_pos(wr) & literal_pos_mask) << lc) + + (wr->previous_byte >> (8 - lc)))) + ); + + if (cst->state >= LZMA_NUM_LIT_STATES) { + int match_byte = peek_old_byte(wr, cst->rep0); + do { + int bit; + uint16_t *prob_lit; + + match_byte <<= 1; + bit = match_byte & 0x100; + prob_lit = prob + 0x100 + bit + mi; + if (rc_get_bit(rc, prob_lit, &mi)) { + if (!bit) + break; + } else { + if (bit) + break; + } + } while (mi < 0x100); + } + while (mi < 0x100) { + uint16_t *prob_lit = prob + mi; + rc_get_bit(rc, prob_lit, &mi); + } + write_byte(wr, mi); + if (cst->state < 4) + cst->state = 0; + else if (cst->state < 10) + cst->state -= 3; + else + cst->state -= 6; +} + +static inline void INIT process_bit1(struct writer *wr, struct rc *rc, + struct cstate *cst, uint16_t *p, + int pos_state, uint16_t *prob) { + int offset; + uint16_t *prob_len; + int num_bits; + int len; + + rc_update_bit_1(rc, prob); + prob = p + LZMA_IS_REP + cst->state; + if (rc_is_bit_0(rc, prob)) { + rc_update_bit_0(rc, prob); + cst->rep3 = cst->rep2; + cst->rep2 = cst->rep1; + cst->rep1 = cst->rep0; + cst->state = cst->state < LZMA_NUM_LIT_STATES ? 0 : 3; + prob = p + LZMA_LEN_CODER; + } else { + rc_update_bit_1(rc, prob); + prob = p + LZMA_IS_REP_G0 + cst->state; + if (rc_is_bit_0(rc, prob)) { + rc_update_bit_0(rc, prob); + prob = (p + LZMA_IS_REP_0_LONG + + (cst->state << + LZMA_NUM_POS_BITS_MAX) + + pos_state); + if (rc_is_bit_0(rc, prob)) { + rc_update_bit_0(rc, prob); + + cst->state = cst->state < LZMA_NUM_LIT_STATES ? + 9 : 11; + copy_byte(wr, cst->rep0); + return; + } else { + rc_update_bit_1(rc, prob); + } + } else { + uint32_t distance; + + rc_update_bit_1(rc, prob); + prob = p + LZMA_IS_REP_G1 + cst->state; + if (rc_is_bit_0(rc, prob)) { + rc_update_bit_0(rc, prob); + distance = cst->rep1; + } else { + rc_update_bit_1(rc, prob); + prob = p + LZMA_IS_REP_G2 + cst->state; + if (rc_is_bit_0(rc, prob)) { + rc_update_bit_0(rc, prob); + distance = cst->rep2; + } else { + rc_update_bit_1(rc, prob); + distance = cst->rep3; + cst->rep3 = cst->rep2; + } + cst->rep2 = cst->rep1; + } + cst->rep1 = cst->rep0; + cst->rep0 = distance; + } + cst->state = cst->state < LZMA_NUM_LIT_STATES ? 8 : 11; + prob = p + LZMA_REP_LEN_CODER; + } + + prob_len = prob + LZMA_LEN_CHOICE; + if (rc_is_bit_0(rc, prob_len)) { + rc_update_bit_0(rc, prob_len); + prob_len = (prob + LZMA_LEN_LOW + + (pos_state << + LZMA_LEN_NUM_LOW_BITS)); + offset = 0; + num_bits = LZMA_LEN_NUM_LOW_BITS; + } else { + rc_update_bit_1(rc, prob_len); + prob_len = prob + LZMA_LEN_CHOICE_2; + if (rc_is_bit_0(rc, prob_len)) { + rc_update_bit_0(rc, prob_len); + prob_len = (prob + LZMA_LEN_MID + + (pos_state << + LZMA_LEN_NUM_MID_BITS)); + offset = 1 << LZMA_LEN_NUM_LOW_BITS; + num_bits = LZMA_LEN_NUM_MID_BITS; + } else { + rc_update_bit_1(rc, prob_len); + prob_len = prob + LZMA_LEN_HIGH; + offset = ((1 << LZMA_LEN_NUM_LOW_BITS) + + (1 << LZMA_LEN_NUM_MID_BITS)); + num_bits = LZMA_LEN_NUM_HIGH_BITS; + } + } + + rc_bit_tree_decode(rc, prob_len, num_bits, &len); + len += offset; + + if (cst->state < 4) { + int pos_slot; + + cst->state += LZMA_NUM_LIT_STATES; + prob = + p + LZMA_POS_SLOT + + ((len < + LZMA_NUM_LEN_TO_POS_STATES ? len : + LZMA_NUM_LEN_TO_POS_STATES - 1) + << LZMA_NUM_POS_SLOT_BITS); + rc_bit_tree_decode(rc, prob, + LZMA_NUM_POS_SLOT_BITS, + &pos_slot); + if (pos_slot >= LZMA_START_POS_MODEL_INDEX) { + int i, mi; + num_bits = (pos_slot >> 1) - 1; + cst->rep0 = 2 | (pos_slot & 1); + if (pos_slot < LZMA_END_POS_MODEL_INDEX) { + cst->rep0 <<= num_bits; + prob = p + LZMA_SPEC_POS + + cst->rep0 - pos_slot - 1; + } else { + num_bits -= LZMA_NUM_ALIGN_BITS; + while (num_bits--) + cst->rep0 = (cst->rep0 << 1) | + rc_direct_bit(rc); + prob = p + LZMA_ALIGN; + cst->rep0 <<= LZMA_NUM_ALIGN_BITS; + num_bits = LZMA_NUM_ALIGN_BITS; + } + i = 1; + mi = 1; + while (num_bits--) { + if (rc_get_bit(rc, prob + mi, &mi)) + cst->rep0 |= i; + i <<= 1; + } + } else + cst->rep0 = pos_slot; + if (++(cst->rep0) == 0) + return; + } + + len += LZMA_MATCH_MIN_LEN; + + copy_bytes(wr, cst->rep0, len); +} + + + +STATIC inline int INIT unlzma(unsigned char *buf, int in_len, + int(*fill)(void*, unsigned int), + int(*flush)(void*, unsigned int), + unsigned char *output, + int *posp, + void(*error_fn)(char *x) + ) +{ + struct lzma_header header; + int lc, pb, lp; + uint32_t pos_state_mask; + uint32_t literal_pos_mask; + uint16_t *p; + int num_probs; + struct rc rc; + int i, mi; + struct writer wr; + struct cstate cst; + unsigned char *inbuf; + int ret = -1; + + set_error_fn(error_fn); + if (!flush) + in_len -= 4; /* Uncompressed size hack active in pre-boot + environment */ + if (buf) + inbuf = buf; + else + inbuf = malloc(LZMA_IOBUF_SIZE); + if (!inbuf) { + error("Could not allocate input bufer"); + goto exit_0; + } + + cst.state = 0; + cst.rep0 = cst.rep1 = cst.rep2 = cst.rep3 = 1; + + wr.header = &header; + wr.flush = flush; + wr.global_pos = 0; + wr.previous_byte = 0; + wr.buffer_pos = 0; + + rc_init(&rc, fill, inbuf, in_len); + + for (i = 0; i < sizeof(header); i++) { + if (rc.ptr >= rc.buffer_end) + rc_read(&rc); + ((unsigned char *)&header)[i] = *rc.ptr++; + } + + if (header.pos >= (9 * 5 * 5)) + error("bad header"); + + mi = 0; + lc = header.pos; + while (lc >= 9) { + mi++; + lc -= 9; + } + pb = 0; + lp = mi; + while (lp >= 5) { + pb++; + lp -= 5; + } + pos_state_mask = (1 << pb) - 1; + literal_pos_mask = (1 << lp) - 1; + + ENDIAN_CONVERT(header.dict_size); + ENDIAN_CONVERT(header.dst_size); + + if (header.dict_size == 0) + header.dict_size = 1; + + if (output) + wr.buffer = output; + else { + wr.bufsize = MIN(header.dst_size, header.dict_size); + wr.buffer = large_malloc(wr.bufsize); + } + if (wr.buffer == NULL) + goto exit_1; + + num_probs = LZMA_BASE_SIZE + (LZMA_LIT_SIZE << (lc + lp)); + p = (uint16_t *) large_malloc(num_probs * sizeof(*p)); + if (p == 0) + goto exit_2; + num_probs = LZMA_LITERAL + (LZMA_LIT_SIZE << (lc + lp)); + for (i = 0; i < num_probs; i++) + p[i] = (1 << RC_MODEL_TOTAL_BITS) >> 1; + + rc_init_code(&rc); + + while (get_pos(&wr) < header.dst_size) { + int pos_state = get_pos(&wr) & pos_state_mask; + uint16_t *prob = p + LZMA_IS_MATCH + + (cst.state << LZMA_NUM_POS_BITS_MAX) + pos_state; + if (rc_is_bit_0(&rc, prob)) + process_bit0(&wr, &rc, &cst, p, pos_state, prob, + lc, literal_pos_mask); + else { + process_bit1(&wr, &rc, &cst, p, pos_state, prob); + if (cst.rep0 == 0) + break; + } + } + + if (posp) + *posp = rc.ptr-rc.buffer; + if (wr.flush) + wr.flush(wr.buffer, wr.buffer_pos); + ret = 0; + large_free(p); +exit_2: + if (!output) + large_free(wr.buffer); +exit_1: + if (!buf) + free(inbuf); +exit_0: + return ret; +} + +#define decompress unlzma diff --git a/lib/zlib_inflate/inflate.h b/lib/zlib_inflate/inflate.h index df8a6c9..3d17b3d 100644 --- a/lib/zlib_inflate/inflate.h +++ b/lib/zlib_inflate/inflate.h @@ -1,3 +1,6 @@ +#ifndef INFLATE_H +#define INFLATE_H + /* inflate.h -- internal inflate state definition * Copyright (C) 1995-2004 Mark Adler * For conditions of distribution and use, see copyright notice in zlib.h @@ -105,3 +108,4 @@ struct inflate_state { unsigned short work[288]; /* work area for code table building */ code codes[ENOUGH]; /* space for code tables */ }; +#endif diff --git a/lib/zlib_inflate/inftrees.h b/lib/zlib_inflate/inftrees.h index 5f5219b..b70b473 100644 --- a/lib/zlib_inflate/inftrees.h +++ b/lib/zlib_inflate/inftrees.h @@ -1,3 +1,6 @@ +#ifndef INFTREES_H +#define INFTREES_H + /* inftrees.h -- header to use inftrees.c * Copyright (C) 1995-2005 Mark Adler * For conditions of distribution and use, see copyright notice in zlib.h @@ -53,3 +56,4 @@ typedef enum { extern int zlib_inflate_table (codetype type, unsigned short *lens, unsigned codes, code **table, unsigned *bits, unsigned short *work); +#endif diff --git a/scripts/Makefile.lib b/scripts/Makefile.lib index e0636577..70b4676 100644 --- a/scripts/Makefile.lib +++ b/scripts/Makefile.lib @@ -186,3 +186,17 @@ quiet_cmd_gzip = GZIP $@ cmd_gzip = gzip -f -9 < $< > $@ +# Bzip2 +# --------------------------------------------------------------------------- + +# Bzip2 does not include size in file... so we have to fake that +size_append=$(CONFIG_SHELL) $(srctree)/scripts/bin_size + +quiet_cmd_bzip2 = BZIP2 $@ +cmd_bzip2 = (bzip2 -9 < $< ; $(size_append) $<) > $@ || (rm -f $@ ; false) + +# Lzma +# --------------------------------------------------------------------------- + +quiet_cmd_lzma = LZMA $@ +cmd_lzma = (lzma -9 -c $< ; $(size_append) $<) >$@ || (rm -f $@ ; false) diff --git a/scripts/bin_size b/scripts/bin_size new file mode 100644 index 0000000..43e1b36 --- /dev/null +++ b/scripts/bin_size @@ -0,0 +1,10 @@ +#!/bin/sh + +if [ $# = 0 ] ; then + echo Usage: $0 file +fi + +size_dec=`stat -c "%s" $1` +size_hex_echo_string=`printf "%08x" $size_dec | + sed 's/\(..\)\(..\)\(..\)\(..\)/\\\\x\4\\\\x\3\\\\x\2\\\\x\1/g'` +/bin/echo -ne $size_hex_echo_string -- cgit v0.10.2 From 30d65dbfe3add7f010a075991dc0bfeaebb7d9e1 Mon Sep 17 00:00:00 2001 From: Alain Knaff Date: Sun, 4 Jan 2009 22:46:17 +0100 Subject: bzip2/lzma: config and initramfs support for bzip2/lzma decompression Impact: New code for initramfs decompression, new features This is the second part of the bzip2/lzma patch The bzip patch is based on an idea by Christian Ludwig, includes support for compressing the kernel with bzip2 or lzma rather than gzip. Both compressors give smaller sizes than gzip. Lzma's decompresses faster than bzip2. It also supports ramdisks and initramfs' compressed using these two compressors. The functionality has been successfully used for a couple of years by the udpcast project This version applies to "tip" kernel 2.6.28 This part contains: - support for new compressions (bzip2 and lzma) in initramfs and old-style ramdisk - config dialog for kernel compression (but new kernel compressions not yet supported) Signed-off-by: Alain Knaff Signed-off-by: H. Peter Anvin diff --git a/drivers/block/Kconfig b/drivers/block/Kconfig index 0344a8a..7955944 100644 --- a/drivers/block/Kconfig +++ b/drivers/block/Kconfig @@ -358,6 +358,31 @@ config BLK_DEV_XIP will prevent RAM block device backing store memory from being allocated from highmem (only a problem for highmem systems). +config RD_BZIP2 + bool "Initial ramdisk compressed using bzip2" + default n + depends on BLK_DEV_INITRD=y + help + Support loading of a bzip2 encoded initial ramdisk or cpio buffer + If unsure, say N. + +config RD_LZMA + bool "Initial ramdisk compressed using lzma" + default n + depends on BLK_DEV_INITRD=y + help + Support loading of a lzma encoded initial ramdisk or cpio buffer + If unsure, say N. + +config RD_GZIP + bool "Initial ramdisk compressed using gzip" + default y + depends on BLK_DEV_INITRD=y + select ZLIB_INFLATE + help + Support loading of a gzip encoded initial ramdisk or cpio buffer. + If unsure, say Y. + config CDROM_PKTCDVD tristate "Packet writing on CD/DVD media" depends on !UML diff --git a/init/Kconfig b/init/Kconfig index f628171..df84625 100644 --- a/init/Kconfig +++ b/init/Kconfig @@ -101,6 +101,56 @@ config LOCALVERSION_AUTO which is done within the script "scripts/setlocalversion".) +choice + prompt "Kernel compression mode" + default KERNEL_GZIP + help + The linux kernel is a kind of self-extracting executable. + Several compression algorithms are available, which differ + in efficiency, compression and decompression speed. + Compression speed is only relevant when building a kernel. + Decompression speed is relevant at each boot. + + If you have any problems with bzip2 or lzma compressed + kernels, mail me (Alain Knaff) . (An older + version of this functionality (bzip2 only), for 2.4, was + supplied by Christian Ludwig) + + High compression options are mostly useful for users, who + are low on disk space (embedded systems), but for whom ram + size matters less. + + If in doubt, select 'gzip' + +config KERNEL_GZIP + bool "Gzip" + help + The old and tried gzip compression. Its compression ratio is + the poorest among the 3 choices; however its speed (both + compression and decompression) is the fastest. + +config KERNEL_BZIP2 + bool "Bzip2" + help + Its compression ratio and speed is intermediate. + Decompression speed is slowest among the 3. + The kernel size is about 10 per cent smaller with bzip2, + in comparison to gzip. + Bzip2 uses a large amount of memory. For modern kernels + you will need at least 8MB RAM or more for booting. + +config KERNEL_LZMA + bool "LZMA" + help + The most recent compression algorithm. + Its ratio is best, decompression speed is between the other + 2. Compression is slowest. + The kernel size is about 33 per cent smaller with lzma, + in comparison to gzip. + +endchoice + + config SWAP bool "Support for paging of anonymous memory (swap)" depends on MMU && BLOCK diff --git a/init/do_mounts_rd.c b/init/do_mounts_rd.c index a7c748f..dcaeb1f 100644 --- a/init/do_mounts_rd.c +++ b/init/do_mounts_rd.c @@ -10,6 +10,12 @@ #include "do_mounts.h" +#include + +#include +#include +#include + int __initdata rd_prompt = 1;/* 1 = prompt for RAM disk, 0 = don't prompt */ static int __init prompt_ramdisk(char *str) @@ -28,7 +34,7 @@ static int __init ramdisk_start_setup(char *str) } __setup("ramdisk_start=", ramdisk_start_setup); -static int __init crd_load(int in_fd, int out_fd); +static int __init crd_load(int in_fd, int out_fd, decompress_fn deco); /* * This routine tries to find a RAM disk image to load, and returns the @@ -44,7 +50,7 @@ static int __init crd_load(int in_fd, int out_fd); * gzip */ static int __init -identify_ramdisk_image(int fd, int start_block) +identify_ramdisk_image(int fd, int start_block, decompress_fn *decompressor) { const int size = 512; struct minix_super_block *minixsb; @@ -70,6 +76,7 @@ identify_ramdisk_image(int fd, int start_block) sys_lseek(fd, start_block * BLOCK_SIZE, 0); sys_read(fd, buf, size); +#ifdef CONFIG_RD_GZIP /* * If it matches the gzip magic numbers, return 0 */ @@ -77,9 +84,39 @@ identify_ramdisk_image(int fd, int start_block) printk(KERN_NOTICE "RAMDISK: Compressed image found at block %d\n", start_block); + *decompressor = gunzip; + nblocks = 0; + goto done; + } +#endif + +#ifdef CONFIG_RD_BZIP2 + /* + * If it matches the bzip2 magic numbers, return -1 + */ + if (buf[0] == 0x42 && (buf[1] == 0x5a)) { + printk(KERN_NOTICE + "RAMDISK: Bzipped image found at block %d\n", + start_block); + *decompressor = bunzip2; + nblocks = 0; + goto done; + } +#endif + +#ifdef CONFIG_RD_LZMA + /* + * If it matches the lzma magic numbers, return -1 + */ + if (buf[0] == 0x5d && (buf[1] == 0x00)) { + printk(KERN_NOTICE + "RAMDISK: Lzma image found at block %d\n", + start_block); + *decompressor = unlzma; nblocks = 0; goto done; } +#endif /* romfs is at block zero too */ if (romfsb->word0 == ROMSB_WORD0 && @@ -143,6 +180,7 @@ int __init rd_load_image(char *from) int nblocks, i, disk; char *buf = NULL; unsigned short rotate = 0; + decompress_fn decompressor = NULL; #if !defined(CONFIG_S390) && !defined(CONFIG_PPC_ISERIES) char rotator[4] = { '|' , '/' , '-' , '\\' }; #endif @@ -155,12 +193,12 @@ int __init rd_load_image(char *from) if (in_fd < 0) goto noclose_input; - nblocks = identify_ramdisk_image(in_fd, rd_image_start); + nblocks = identify_ramdisk_image(in_fd, rd_image_start, &decompressor); if (nblocks < 0) goto done; if (nblocks == 0) { - if (crd_load(in_fd, out_fd) == 0) + if (crd_load(in_fd, out_fd, decompressor) == 0) goto successful_load; goto done; } @@ -259,138 +297,48 @@ int __init rd_load_disk(int n) return rd_load_image("/dev/root"); } -/* - * gzip declarations - */ - -#define OF(args) args - -#ifndef memzero -#define memzero(s, n) memset ((s), 0, (n)) -#endif - -typedef unsigned char uch; -typedef unsigned short ush; -typedef unsigned long ulg; - -#define INBUFSIZ 4096 -#define WSIZE 0x8000 /* window size--must be a power of two, and */ - /* at least 32K for zip's deflate method */ - -static uch *inbuf; -static uch *window; - -static unsigned insize; /* valid bytes in inbuf */ -static unsigned inptr; /* index of next byte to be processed in inbuf */ -static unsigned outcnt; /* bytes in output buffer */ static int exit_code; -static int unzip_error; -static long bytes_out; +static int decompress_error; static int crd_infd, crd_outfd; -#define get_byte() (inptr < insize ? inbuf[inptr++] : fill_inbuf()) - -/* Diagnostic functions (stubbed out) */ -#define Assert(cond,msg) -#define Trace(x) -#define Tracev(x) -#define Tracevv(x) -#define Tracec(c,x) -#define Tracecv(c,x) - -#define STATIC static -#define INIT __init - -static int __init fill_inbuf(void); -static void __init flush_window(void); -static void __init error(char *m); - -#define NO_INFLATE_MALLOC - -#include "../lib/inflate.c" - -/* =========================================================================== - * Fill the input buffer. This is called only when the buffer is empty - * and at least one byte is really needed. - * Returning -1 does not guarantee that gunzip() will ever return. - */ -static int __init fill_inbuf(void) +static int __init compr_fill(void *buf, unsigned int len) { - if (exit_code) return -1; - - insize = sys_read(crd_infd, inbuf, INBUFSIZ); - if (insize == 0) { - error("RAMDISK: ran out of compressed data"); - return -1; - } - - inptr = 1; - - return inbuf[0]; + int r = sys_read(crd_infd, buf, len); + if (r < 0) + printk(KERN_ERR "RAMDISK: error while reading compressed data"); + else if (r == 0) + printk(KERN_ERR "RAMDISK: EOF while reading compressed data"); + return r; } -/* =========================================================================== - * Write the output window window[0..outcnt-1] and update crc and bytes_out. - * (Used for the decompressed data only.) - */ -static void __init flush_window(void) +static int __init compr_flush(void *window, unsigned int outcnt) { - ulg c = crc; /* temporary variable */ - unsigned n, written; - uch *in, ch; - - written = sys_write(crd_outfd, window, outcnt); - if (written != outcnt && unzip_error == 0) { - printk(KERN_ERR "RAMDISK: incomplete write (%d != %d) %ld\n", - written, outcnt, bytes_out); - unzip_error = 1; - } - in = window; - for (n = 0; n < outcnt; n++) { - ch = *in++; - c = crc_32_tab[((int)c ^ ch) & 0xff] ^ (c >> 8); - } - crc = c; - bytes_out += (ulg)outcnt; - outcnt = 0; + int written = sys_write(crd_outfd, window, outcnt); + if (written != outcnt) { + if (decompress_error == 0) + printk(KERN_ERR + "RAMDISK: incomplete write (%d != %d)\n", + written, outcnt); + decompress_error = 1; + return -1; + } + return outcnt; } static void __init error(char *x) { printk(KERN_ERR "%s\n", x); exit_code = 1; - unzip_error = 1; + decompress_error = 1; } -static int __init crd_load(int in_fd, int out_fd) +static int __init crd_load(int in_fd, int out_fd, decompress_fn deco) { int result; - - insize = 0; /* valid bytes in inbuf */ - inptr = 0; /* index of next byte to be processed in inbuf */ - outcnt = 0; /* bytes in output buffer */ - exit_code = 0; - bytes_out = 0; - crc = (ulg)0xffffffffL; /* shift register contents */ - crd_infd = in_fd; crd_outfd = out_fd; - inbuf = kmalloc(INBUFSIZ, GFP_KERNEL); - if (!inbuf) { - printk(KERN_ERR "RAMDISK: Couldn't allocate gzip buffer\n"); - return -1; - } - window = kmalloc(WSIZE, GFP_KERNEL); - if (!window) { - printk(KERN_ERR "RAMDISK: Couldn't allocate gzip window\n"); - kfree(inbuf); - return -1; - } - makecrc(); - result = gunzip(); - if (unzip_error) + result = deco(NULL, 0, compr_fill, compr_flush, NULL, NULL, error); + if (decompress_error) result = 1; - kfree(inbuf); - kfree(window); return result; } diff --git a/init/initramfs.c b/init/initramfs.c index 4f5ba75..40bd4fb 100644 --- a/init/initramfs.c +++ b/init/initramfs.c @@ -389,11 +389,14 @@ static int __init write_buffer(char *buf, unsigned len) return len - count; } -static void __init flush_buffer(char *buf, unsigned len) + +static int __init flush_buffer(void *bufv, unsigned len) { + char *buf = (char *) bufv; int written; + int origLen = len; if (message) - return; + return -1; while ((written = write_buffer(buf, len)) < len && !message) { char c = buf[written]; if (c == '0') { @@ -407,73 +410,14 @@ static void __init flush_buffer(char *buf, unsigned len) } else error("junk in compressed archive"); } + return origLen; } -/* - * gzip declarations - */ - -#define OF(args) args - -#ifndef memzero -#define memzero(s, n) memset ((s), 0, (n)) -#endif - -typedef unsigned char uch; -typedef unsigned short ush; -typedef unsigned long ulg; - -#define WSIZE 0x8000 /* window size--must be a power of two, and */ - /* at least 32K for zip's deflate method */ - -static uch *inbuf; -static uch *window; - -static unsigned insize; /* valid bytes in inbuf */ -static unsigned inptr; /* index of next byte to be processed in inbuf */ -static unsigned outcnt; /* bytes in output buffer */ -static long bytes_out; - -#define get_byte() (inptr < insize ? inbuf[inptr++] : -1) - -/* Diagnostic functions (stubbed out) */ -#define Assert(cond,msg) -#define Trace(x) -#define Tracev(x) -#define Tracevv(x) -#define Tracec(c,x) -#define Tracecv(c,x) - -#define STATIC static -#define INIT __init - -static void __init flush_window(void); -static void __init error(char *m); - -#define NO_INFLATE_MALLOC +static unsigned my_inptr; /* index of next byte to be processed in inbuf */ -#include "../lib/inflate.c" - -/* =========================================================================== - * Write the output window window[0..outcnt-1] and update crc and bytes_out. - * (Used for the decompressed data only.) - */ -static void __init flush_window(void) -{ - ulg c = crc; /* temporary variable */ - unsigned n; - uch *in, ch; - - flush_buffer(window, outcnt); - in = window; - for (n = 0; n < outcnt; n++) { - ch = *in++; - c = crc_32_tab[((int)c ^ ch) & 0xff] ^ (c >> 8); - } - crc = c; - bytes_out += (ulg)outcnt; - outcnt = 0; -} +#include +#include +#include static char * __init unpack_to_rootfs(char *buf, unsigned len, int check_only) { @@ -482,9 +426,10 @@ static char * __init unpack_to_rootfs(char *buf, unsigned len, int check_only) header_buf = kmalloc(110, GFP_KERNEL); symlink_buf = kmalloc(PATH_MAX + N_ALIGN(PATH_MAX) + 1, GFP_KERNEL); name_buf = kmalloc(N_ALIGN(PATH_MAX), GFP_KERNEL); - window = kmalloc(WSIZE, GFP_KERNEL); - if (!window || !header_buf || !symlink_buf || !name_buf) + + if (!header_buf || !symlink_buf || !name_buf) panic("can't allocate buffers"); + state = Start; this_header = 0; message = NULL; @@ -504,22 +449,38 @@ static char * __init unpack_to_rootfs(char *buf, unsigned len, int check_only) continue; } this_header = 0; - insize = len; - inbuf = buf; - inptr = 0; - outcnt = 0; /* bytes in output buffer */ - bytes_out = 0; - crc = (ulg)0xffffffffL; /* shift register contents */ - makecrc(); - gunzip(); + if (!gunzip(buf, len, NULL, flush_buffer, NULL, + &my_inptr, error) && + message == NULL) + goto ok; + +#ifdef CONFIG_RD_BZIP2 + message = NULL; /* Zero out message, or else cpio will + think an error has already occured */ + if (!bunzip2(buf, len, NULL, flush_buffer, NULL, + &my_inptr, error) && + message == NULL) { + goto ok; + } +#endif + +#ifdef CONFIG_RD_LZMA + message = NULL; /* Zero out message, or else cpio will + think an error has already occured */ + if (!unlzma(buf, len, NULL, flush_buffer, NULL, + &my_inptr, error) && + message == NULL) { + goto ok; + } +#endif +ok: if (state != Reset) - error("junk in gzipped archive"); - this_header = saved_offset + inptr; - buf += inptr; - len -= inptr; + error("junk in compressed archive"); + this_header = saved_offset + my_inptr; + buf += my_inptr; + len -= my_inptr; } dir_utime(); - kfree(window); kfree(name_buf); kfree(symlink_buf); kfree(header_buf); diff --git a/lib/Makefile b/lib/Makefile index 32b0e64..e2a21d5 100644 --- a/lib/Makefile +++ b/lib/Makefile @@ -11,7 +11,8 @@ lib-y := ctype.o string.o vsprintf.o cmdline.o \ rbtree.o radix-tree.o dump_stack.o \ idr.o int_sqrt.o extable.o prio_tree.o \ sha1.o irq_regs.o reciprocal_div.o argv_split.o \ - proportions.o prio_heap.o ratelimit.o show_mem.o is_single_threaded.o + proportions.o prio_heap.o ratelimit.o show_mem.o is_single_threaded.o \ + decompress_inflate.o decompress_bunzip2.o decompress_unlzma.o lib-$(CONFIG_MMU) += ioremap.o lib-$(CONFIG_SMP) += cpumask.o -- cgit v0.10.2 From b172fd882d201cf8aa67cddd9c1a023421fd4956 Mon Sep 17 00:00:00 2001 From: "H. Peter Anvin" Date: Sun, 4 Jan 2009 14:42:52 -0800 Subject: bzip2/lzma: use a table to search for initramfs compression formats Impact: Code simplification Instead of open-coding testing for initramfs compression formats, use a table. Signed-off-by: H. Peter Anvin diff --git a/init/do_mounts_rd.c b/init/do_mounts_rd.c index dcaeb1f..9c9d7db 100644 --- a/init/do_mounts_rd.c +++ b/init/do_mounts_rd.c @@ -43,13 +43,31 @@ static int __init crd_load(int in_fd, int out_fd, decompress_fn deco); * numbers could not be found. * * We currently check for the following magic numbers: - * minix - * ext2 + * minix + * ext2 * romfs * cramfs - * gzip + * gzip */ -static int __init +static const struct compress_format { + unsigned char magic[2]; + const char *name; + decompress_fn decompressor; +} compressed_formats[] = { +#ifdef CONFIG_RD_GZIP + { {037, 0213}, "gzip", gunzip }, + { {037, 0236}, "gzip", gunzip }, +#endif +#ifdef CONFIG_RD_BZIP2 + { {0x42, 0x5a}, "bzip2", bunzip2 }, +#endif +#ifdef CONFIG_RD_LZMA + { {0x5d, 0x00}, "lzma", unlzma }, +#endif + { {0, 0}, NULL, NULL } +}; + +static int __init identify_ramdisk_image(int fd, int start_block, decompress_fn *decompressor) { const int size = 512; @@ -59,6 +77,7 @@ identify_ramdisk_image(int fd, int start_block, decompress_fn *decompressor) struct cramfs_super *cramfsb; int nblocks = -1; unsigned char *buf; + const struct compress_format *cf; buf = kmalloc(size, GFP_KERNEL); if (!buf) @@ -71,52 +90,21 @@ identify_ramdisk_image(int fd, int start_block, decompress_fn *decompressor) memset(buf, 0xe5, size); /* - * Read block 0 to test for gzipped kernel + * Read block 0 to test for compressed kernel */ sys_lseek(fd, start_block * BLOCK_SIZE, 0); sys_read(fd, buf, size); -#ifdef CONFIG_RD_GZIP - /* - * If it matches the gzip magic numbers, return 0 - */ - if (buf[0] == 037 && ((buf[1] == 0213) || (buf[1] == 0236))) { - printk(KERN_NOTICE - "RAMDISK: Compressed image found at block %d\n", - start_block); - *decompressor = gunzip; - nblocks = 0; - goto done; - } -#endif - -#ifdef CONFIG_RD_BZIP2 - /* - * If it matches the bzip2 magic numbers, return -1 - */ - if (buf[0] == 0x42 && (buf[1] == 0x5a)) { - printk(KERN_NOTICE - "RAMDISK: Bzipped image found at block %d\n", - start_block); - *decompressor = bunzip2; - nblocks = 0; - goto done; - } -#endif - -#ifdef CONFIG_RD_LZMA - /* - * If it matches the lzma magic numbers, return -1 - */ - if (buf[0] == 0x5d && (buf[1] == 0x00)) { - printk(KERN_NOTICE - "RAMDISK: Lzma image found at block %d\n", - start_block); - *decompressor = unlzma; - nblocks = 0; - goto done; + for (cf = compressed_formats; cf->decompressor; cf++) { + if (buf[0] == cf->magic[0] && buf[1] == cf->magic[1]) { + printk(KERN_NOTICE + "RAMDISK: %s image found at block %d\n", + cf->name, start_block); + *decompressor = cf->decompressor; + nblocks = 0; + goto done; + } } -#endif /* romfs is at block zero too */ if (romfsb->word0 == ROMSB_WORD0 && @@ -165,7 +153,7 @@ identify_ramdisk_image(int fd, int start_block, decompress_fn *decompressor) printk(KERN_NOTICE "RAMDISK: Couldn't find valid RAM disk image starting at %d.\n", start_block); - + done: sys_lseek(fd, start_block * BLOCK_SIZE, 0); kfree(buf); @@ -224,7 +212,7 @@ int __init rd_load_image(char *from) nblocks, rd_blocks); goto done; } - + /* * OK, time to copy in the data */ -- cgit v0.10.2 From ae03c49964af5033534e518eebe439c3b90f43a7 Mon Sep 17 00:00:00 2001 From: Alain Knaff Date: Sun, 4 Jan 2009 22:46:17 +0100 Subject: bzip2/lzma: x86 kernel compression support Impact: Replaces x86 kernel decompressor with new code This is the third part of the bzip2/lzma patch The bzip patch is based on an idea by Christian Ludwig, includes support for compressing the kernel with bzip2 or lzma rather than gzip. Both compressors give smaller sizes than gzip. Lzma's decompresses faster than bzip2. It also supports ramdisks and initramfs' compressed using these two compressors. The functionality has been successfully used for a couple of years by the udpcast project This version applies to "tip" kernel 2.6.28 This part contains: - support for new bzip2 and lzma kernel compression for x86 Signed-off-by: Alain Knaff Signed-off-by: H. Peter Anvin diff --git a/arch/x86/boot/compressed/Makefile b/arch/x86/boot/compressed/Makefile index 1771c80..3ca4c19 100644 --- a/arch/x86/boot/compressed/Makefile +++ b/arch/x86/boot/compressed/Makefile @@ -4,7 +4,7 @@ # create a compressed vmlinux image from the original vmlinux # -targets := vmlinux vmlinux.bin vmlinux.bin.gz head_$(BITS).o misc.o piggy.o +targets := vmlinux vmlinux.bin vmlinux.bin.gz vmlinux.bin.bz2 vmlinux.bin.lzma head_$(BITS).o misc.o piggy.o KBUILD_CFLAGS := -m$(BITS) -D__KERNEL__ $(LINUX_INCLUDE) -O2 KBUILD_CFLAGS += -fno-strict-aliasing -fPIC @@ -47,18 +47,35 @@ ifeq ($(CONFIG_X86_32),y) ifdef CONFIG_RELOCATABLE $(obj)/vmlinux.bin.gz: $(obj)/vmlinux.bin.all FORCE $(call if_changed,gzip) +$(obj)/vmlinux.bin.bz2: $(obj)/vmlinux.bin.all FORCE + $(call if_changed,bzip2) +$(obj)/vmlinux.bin.lzma: $(obj)/vmlinux.bin.all FORCE + $(call if_changed,lzma) else $(obj)/vmlinux.bin.gz: $(obj)/vmlinux.bin FORCE $(call if_changed,gzip) +$(obj)/vmlinux.bin.bz2: $(obj)/vmlinux.bin FORCE + $(call if_changed,bzip2) +$(obj)/vmlinux.bin.lzma: $(obj)/vmlinux.bin FORCE + $(call if_changed,lzma) endif LDFLAGS_piggy.o := -r --format binary --oformat elf32-i386 -T else + $(obj)/vmlinux.bin.gz: $(obj)/vmlinux.bin FORCE $(call if_changed,gzip) +$(obj)/vmlinux.bin.bz2: $(obj)/vmlinux.bin FORCE + $(call if_changed,bzip2) +$(obj)/vmlinux.bin.lzma: $(obj)/vmlinux.bin FORCE + $(call if_changed,lzma) LDFLAGS_piggy.o := -r --format binary --oformat elf64-x86-64 -T endif -$(obj)/piggy.o: $(obj)/vmlinux.scr $(obj)/vmlinux.bin.gz FORCE +suffix_$(CONFIG_KERNEL_GZIP) = gz +suffix_$(CONFIG_KERNEL_BZIP2) = bz2 +suffix_$(CONFIG_KERNEL_LZMA) = lzma + +$(obj)/piggy.o: $(obj)/vmlinux.scr $(obj)/vmlinux.bin.$(suffix_y) FORCE $(call if_changed,ld) diff --git a/arch/x86/boot/compressed/misc.c b/arch/x86/boot/compressed/misc.c index da06221..e45be73 100644 --- a/arch/x86/boot/compressed/misc.c +++ b/arch/x86/boot/compressed/misc.c @@ -116,71 +116,13 @@ /* * gzip declarations */ - -#define OF(args) args #define STATIC static #undef memset #undef memcpy #define memzero(s, n) memset((s), 0, (n)) -typedef unsigned char uch; -typedef unsigned short ush; -typedef unsigned long ulg; - -/* - * Window size must be at least 32k, and a power of two. - * We don't actually have a window just a huge output buffer, - * so we report a 2G window size, as that should always be - * larger than our output buffer: - */ -#define WSIZE 0x80000000 - -/* Input buffer: */ -static unsigned char *inbuf; - -/* Sliding window buffer (and final output buffer): */ -static unsigned char *window; - -/* Valid bytes in inbuf: */ -static unsigned insize; - -/* Index of next byte to be processed in inbuf: */ -static unsigned inptr; - -/* Bytes in output buffer: */ -static unsigned outcnt; - -/* gzip flag byte */ -#define ASCII_FLAG 0x01 /* bit 0 set: file probably ASCII text */ -#define CONTINUATION 0x02 /* bit 1 set: continuation of multi-part gz file */ -#define EXTRA_FIELD 0x04 /* bit 2 set: extra field present */ -#define ORIG_NAM 0x08 /* bit 3 set: original file name present */ -#define COMMENT 0x10 /* bit 4 set: file comment present */ -#define ENCRYPTED 0x20 /* bit 5 set: file is encrypted */ -#define RESERVED 0xC0 /* bit 6, 7: reserved */ - -#define get_byte() (inptr < insize ? inbuf[inptr++] : fill_inbuf()) - -/* Diagnostic functions */ -#ifdef DEBUG -# define Assert(cond, msg) do { if (!(cond)) error(msg); } while (0) -# define Trace(x) do { fprintf x; } while (0) -# define Tracev(x) do { if (verbose) fprintf x ; } while (0) -# define Tracevv(x) do { if (verbose > 1) fprintf x ; } while (0) -# define Tracec(c, x) do { if (verbose && (c)) fprintf x ; } while (0) -# define Tracecv(c, x) do { if (verbose > 1 && (c)) fprintf x ; } while (0) -#else -# define Assert(cond, msg) -# define Trace(x) -# define Tracev(x) -# define Tracevv(x) -# define Tracec(c, x) -# define Tracecv(c, x) -#endif -static int fill_inbuf(void); -static void flush_window(void); static void error(char *m); /* @@ -189,13 +131,8 @@ static void error(char *m); static struct boot_params *real_mode; /* Pointer to real-mode data */ static int quiet; -extern unsigned char input_data[]; -extern int input_len; - -static long bytes_out; - static void *memset(void *s, int c, unsigned n); -static void *memcpy(void *dest, const void *src, unsigned n); +void *memcpy(void *dest, const void *src, unsigned n); static void __putstr(int, const char *); #define putstr(__x) __putstr(0, __x) @@ -213,7 +150,17 @@ static char *vidmem; static int vidport; static int lines, cols; -#include "../../../../lib/inflate.c" +#ifdef CONFIG_KERNEL_GZIP +#include "../../../../lib/decompress_inflate.c" +#endif + +#ifdef CONFIG_KERNEL_BZIP2 +#include "../../../../lib/decompress_bunzip2.c" +#endif + +#ifdef CONFIG_KERNEL_LZMA +#include "../../../../lib/decompress_unlzma.c" +#endif static void scroll(void) { @@ -282,7 +229,7 @@ static void *memset(void *s, int c, unsigned n) return s; } -static void *memcpy(void *dest, const void *src, unsigned n) +void *memcpy(void *dest, const void *src, unsigned n) { int i; const char *s = src; @@ -293,38 +240,6 @@ static void *memcpy(void *dest, const void *src, unsigned n) return dest; } -/* =========================================================================== - * Fill the input buffer. This is called only when the buffer is empty - * and at least one byte is really needed. - */ -static int fill_inbuf(void) -{ - error("ran out of input data"); - return 0; -} - -/* =========================================================================== - * Write the output window window[0..outcnt-1] and update crc and bytes_out. - * (Used for the decompressed data only.) - */ -static void flush_window(void) -{ - /* With my window equal to my output buffer - * I only need to compute the crc here. - */ - unsigned long c = crc; /* temporary variable */ - unsigned n; - unsigned char *in, ch; - - in = window; - for (n = 0; n < outcnt; n++) { - ch = *in++; - c = crc_32_tab[((int)c ^ ch) & 0xff] ^ (c >> 8); - } - crc = c; - bytes_out += (unsigned long)outcnt; - outcnt = 0; -} static void error(char *x) { @@ -407,12 +322,8 @@ asmlinkage void decompress_kernel(void *rmode, memptr heap, lines = real_mode->screen_info.orig_video_lines; cols = real_mode->screen_info.orig_video_cols; - window = output; /* Output buffer (Normally at 1M) */ free_mem_ptr = heap; /* Heap */ free_mem_end_ptr = heap + BOOT_HEAP_SIZE; - inbuf = input_data; /* Input buffer */ - insize = input_len; - inptr = 0; #ifdef CONFIG_X86_64 if ((unsigned long)output & (__KERNEL_ALIGN - 1)) @@ -430,10 +341,9 @@ asmlinkage void decompress_kernel(void *rmode, memptr heap, #endif #endif - makecrc(); if (!quiet) putstr("\nDecompressing Linux... "); - gunzip(); + decompress(input_data, input_len, NULL, NULL, output, NULL, error); parse_elf(output); if (!quiet) putstr("done.\nBooting the kernel.\n"); diff --git a/arch/x86/include/asm/boot.h b/arch/x86/include/asm/boot.h index dd61616..c0e8e68 100644 --- a/arch/x86/include/asm/boot.h +++ b/arch/x86/include/asm/boot.h @@ -15,11 +15,21 @@ + (CONFIG_PHYSICAL_ALIGN - 1)) \ & ~(CONFIG_PHYSICAL_ALIGN - 1)) +#if (defined CONFIG_KERNEL_BZIP2) +#define BOOT_HEAP_SIZE 0x400000 +#else + #ifdef CONFIG_X86_64 #define BOOT_HEAP_SIZE 0x7000 -#define BOOT_STACK_SIZE 0x4000 #else #define BOOT_HEAP_SIZE 0x4000 +#endif + +#endif + +#ifdef CONFIG_X86_64 +#define BOOT_STACK_SIZE 0x4000 +#else #define BOOT_STACK_SIZE 0x1000 #endif -- cgit v0.10.2 From 0f5e2d2484ccd3062bb1f63083bafa37041bc868 Mon Sep 17 00:00:00 2001 From: "H. Peter Anvin" Date: Sun, 4 Jan 2009 15:10:40 -0800 Subject: bzip2/lzma: handle failures from bzip2 and lzma correctly Impact: Bug fix If bzip2 or lzma fails (for example, if they aren't installed on the system), we need to propagate the failure out to "make". However, they were masked by being followed by a semicolon. Signed-off-by: H. Peter Anvin diff --git a/scripts/Makefile.lib b/scripts/Makefile.lib index 70b4676..3b949a3 100644 --- a/scripts/Makefile.lib +++ b/scripts/Makefile.lib @@ -193,10 +193,10 @@ cmd_gzip = gzip -f -9 < $< > $@ size_append=$(CONFIG_SHELL) $(srctree)/scripts/bin_size quiet_cmd_bzip2 = BZIP2 $@ -cmd_bzip2 = (bzip2 -9 < $< ; $(size_append) $<) > $@ || (rm -f $@ ; false) +cmd_bzip2 = (bzip2 -9 < $< && $(size_append) $<) > $@ || (rm -f $@ ; false) # Lzma # --------------------------------------------------------------------------- quiet_cmd_lzma = LZMA $@ -cmd_lzma = (lzma -9 -c $< ; $(size_append) $<) >$@ || (rm -f $@ ; false) +cmd_lzma = (lzma -9 -c $< && $(size_append) $<) >$@ || (rm -f $@ ; false) -- cgit v0.10.2 From 2e9f3bddcbc711bb14d86c6f068a779bf3710247 Mon Sep 17 00:00:00 2001 From: "H. Peter Anvin" Date: Sun, 4 Jan 2009 15:41:25 -0800 Subject: bzip2/lzma: make config machinery an arch configurable Impact: Bug fix (we should not show this menu on irrelevant architectures) Make the config machinery to drive the gzip/bzip2/lzma selection dependent on the architecture advertising HAVE_KERNEL_* so that we don't display this for architectures where it doesn't matter. Signed-off-by: H. Peter Anvin diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig index 862adb9..7b66c34 100644 --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -39,6 +39,9 @@ config X86 select HAVE_GENERIC_DMA_COHERENT if X86_32 select HAVE_EFFICIENT_UNALIGNED_ACCESS select USER_STACKTRACE_SUPPORT + select HAVE_KERNEL_GZIP + select HAVE_KERNEL_BZIP2 + select HAVE_KERNEL_LZMA config ARCH_DEFCONFIG string diff --git a/init/Kconfig b/init/Kconfig index df84625..f9633c0 100644 --- a/init/Kconfig +++ b/init/Kconfig @@ -101,10 +101,20 @@ config LOCALVERSION_AUTO which is done within the script "scripts/setlocalversion".) +config HAVE_KERNEL_GZIP + bool + +config HAVE_KERNEL_BZIP2 + bool + +config HAVE_KERNEL_LZMA + bool + choice - prompt "Kernel compression mode" - default KERNEL_GZIP - help + prompt "Kernel compression mode" + default KERNEL_GZIP + depends on HAVE_KERNEL_GZIP || HAVE_KERNEL_BZIP2 || HAVE_KERNEL_LZMA + help The linux kernel is a kind of self-extracting executable. Several compression algorithms are available, which differ in efficiency, compression and decompression speed. @@ -123,34 +133,34 @@ choice If in doubt, select 'gzip' config KERNEL_GZIP - bool "Gzip" - help - The old and tried gzip compression. Its compression ratio is - the poorest among the 3 choices; however its speed (both - compression and decompression) is the fastest. + bool "Gzip" + depends on HAVE_KERNEL_GZIP + help + The old and tried gzip compression. Its compression ratio is + the poorest among the 3 choices; however its speed (both + compression and decompression) is the fastest. config KERNEL_BZIP2 bool "Bzip2" + depends on HAVE_KERNEL_BZIP2 help Its compression ratio and speed is intermediate. - Decompression speed is slowest among the 3. - The kernel size is about 10 per cent smaller with bzip2, - in comparison to gzip. - Bzip2 uses a large amount of memory. For modern kernels - you will need at least 8MB RAM or more for booting. + Decompression speed is slowest among the three. The kernel + size is about 10% smaller with bzip2, in comparison to gzip. + Bzip2 uses a large amount of memory. For modern kernels you + will need at least 8MB RAM or more for booting. config KERNEL_LZMA - bool "LZMA" - help - The most recent compression algorithm. - Its ratio is best, decompression speed is between the other - 2. Compression is slowest. - The kernel size is about 33 per cent smaller with lzma, - in comparison to gzip. + bool "LZMA" + depends on HAVE_KERNEL_LZMA + help + The most recent compression algorithm. + Its ratio is best, decompression speed is between the other + two. Compression is slowest. The kernel size is about 33% + smaller with LZMA in comparison to gzip. endchoice - config SWAP bool "Support for paging of anonymous memory (swap)" depends on MMU && BLOCK -- cgit v0.10.2 From c8531ab343dec88ed8005e403b1b304c710b7494 Mon Sep 17 00:00:00 2001 From: "H. Peter Anvin" Date: Mon, 5 Jan 2009 13:48:31 -0800 Subject: bzip2/lzma: proper Kconfig dependencies for the ramdisk options Impact: Partial resolution of build failure Make all the compression algorithms properly configurable, and make sure the ramdisk options pull in the proper compression algorithms, as they should. Signed-off-by: H. Peter Anvin diff --git a/drivers/block/Kconfig b/drivers/block/Kconfig index 7955944..cc9fa69 100644 --- a/drivers/block/Kconfig +++ b/drivers/block/Kconfig @@ -358,10 +358,20 @@ config BLK_DEV_XIP will prevent RAM block device backing store memory from being allocated from highmem (only a problem for highmem systems). +config RD_GZIP + bool "Initial ramdisk compressed using gzip" + default y + depends on BLK_DEV_INITRD=y + select DECOMPRESS_GZIP + help + Support loading of a gzip encoded initial ramdisk or cpio buffer. + If unsure, say Y. + config RD_BZIP2 bool "Initial ramdisk compressed using bzip2" default n depends on BLK_DEV_INITRD=y + select DECOMPRESS_BZIP2 help Support loading of a bzip2 encoded initial ramdisk or cpio buffer If unsure, say N. @@ -370,19 +380,11 @@ config RD_LZMA bool "Initial ramdisk compressed using lzma" default n depends on BLK_DEV_INITRD=y + select DECOMPRESS_LZMA help Support loading of a lzma encoded initial ramdisk or cpio buffer If unsure, say N. -config RD_GZIP - bool "Initial ramdisk compressed using gzip" - default y - depends on BLK_DEV_INITRD=y - select ZLIB_INFLATE - help - Support loading of a gzip encoded initial ramdisk or cpio buffer. - If unsure, say Y. - config CDROM_PKTCDVD tristate "Packet writing on CD/DVD media" depends on !UML diff --git a/lib/Kconfig b/lib/Kconfig index 03c2c24..e37f061 100644 --- a/lib/Kconfig +++ b/lib/Kconfig @@ -98,6 +98,19 @@ config LZO_DECOMPRESS tristate # +# These all provide a common interface (hence the apparent duplication with +# ZLIB_INFLATE; DECOMPRESS_GZIP is just a wrapper.) +# +config DECOMPRESS_GZIP + tristate + +config DECOMPRESS_BZIP2 + tristate + +config DECOMPRESS_LZMA + tristate + +# # Generic allocator support is selected if needed # config GENERIC_ALLOCATOR diff --git a/lib/Makefile b/lib/Makefile index e2a21d5..d9ac5a4 100644 --- a/lib/Makefile +++ b/lib/Makefile @@ -11,8 +11,7 @@ lib-y := ctype.o string.o vsprintf.o cmdline.o \ rbtree.o radix-tree.o dump_stack.o \ idr.o int_sqrt.o extable.o prio_tree.o \ sha1.o irq_regs.o reciprocal_div.o argv_split.o \ - proportions.o prio_heap.o ratelimit.o show_mem.o is_single_threaded.o \ - decompress_inflate.o decompress_bunzip2.o decompress_unlzma.o + proportions.o prio_heap.o ratelimit.o show_mem.o is_single_threaded.o lib-$(CONFIG_MMU) += ioremap.o lib-$(CONFIG_SMP) += cpumask.o @@ -66,6 +65,10 @@ obj-$(CONFIG_REED_SOLOMON) += reed_solomon/ obj-$(CONFIG_LZO_COMPRESS) += lzo/ obj-$(CONFIG_LZO_DECOMPRESS) += lzo/ +obj-$(CONFIG_DECOMPRESS_GZIP) += decompress_inflate.o +obj-$(CONFIG_DECOMPRESS_BZIP2) += decompress_bunzip2.o +obj-$(CONFIG_DECOMPRESS_LZMA) += decompress_unlzma.o + obj-$(CONFIG_TEXTSEARCH) += textsearch.o obj-$(CONFIG_TEXTSEARCH_KMP) += ts_kmp.o obj-$(CONFIG_TEXTSEARCH_BM) += ts_bm.o -- cgit v0.10.2 From 7856a16ea03ed9b17860d756ee6473c2e57882b2 Mon Sep 17 00:00:00 2001 From: "H. Peter Anvin" Date: Wed, 7 Jan 2009 00:01:43 -0800 Subject: bzip2/lzma: DECOMPRESS_GZIP should select ZLIB_INFLATE Impact: Partial resolution of build failure DECOMPRESS_GZIP is just a common-interface wrapper around the zlib_inflate code; it thus need to select it. Signed-off-by: H. Peter Anvin diff --git a/lib/Kconfig b/lib/Kconfig index e37f061..daa4818 100644 --- a/lib/Kconfig +++ b/lib/Kconfig @@ -102,6 +102,7 @@ config LZO_DECOMPRESS # ZLIB_INFLATE; DECOMPRESS_GZIP is just a wrapper.) # config DECOMPRESS_GZIP + select ZLIB_INFLATE tristate config DECOMPRESS_BZIP2 -- cgit v0.10.2 From fb9a4ca9820fd4d7c4906bd393004662451e273e Mon Sep 17 00:00:00 2001 From: "H. Peter Anvin" Date: Wed, 7 Jan 2009 00:03:49 -0800 Subject: bzip2/lzma: move initrd/ramfs options out of BLK_DEV Impact: Partial resolution of build failure Move the initrd/initramfs configuration options from drivers/block/Kconfig to usr/Kconfig, since they do not and should not depend on CONFIG_BLK_DEV. This fixes builds when CONFIG_BLK_DEV=n. Signed-off-by: H. Peter Anvin diff --git a/drivers/block/Kconfig b/drivers/block/Kconfig index cc9fa69..0344a8a 100644 --- a/drivers/block/Kconfig +++ b/drivers/block/Kconfig @@ -358,33 +358,6 @@ config BLK_DEV_XIP will prevent RAM block device backing store memory from being allocated from highmem (only a problem for highmem systems). -config RD_GZIP - bool "Initial ramdisk compressed using gzip" - default y - depends on BLK_DEV_INITRD=y - select DECOMPRESS_GZIP - help - Support loading of a gzip encoded initial ramdisk or cpio buffer. - If unsure, say Y. - -config RD_BZIP2 - bool "Initial ramdisk compressed using bzip2" - default n - depends on BLK_DEV_INITRD=y - select DECOMPRESS_BZIP2 - help - Support loading of a bzip2 encoded initial ramdisk or cpio buffer - If unsure, say N. - -config RD_LZMA - bool "Initial ramdisk compressed using lzma" - default n - depends on BLK_DEV_INITRD=y - select DECOMPRESS_LZMA - help - Support loading of a lzma encoded initial ramdisk or cpio buffer - If unsure, say N. - config CDROM_PKTCDVD tristate "Packet writing on CD/DVD media" depends on !UML diff --git a/usr/Kconfig b/usr/Kconfig index 86cecb5..a691a8f 100644 --- a/usr/Kconfig +++ b/usr/Kconfig @@ -44,3 +44,30 @@ config INITRAMFS_ROOT_GID owned by group root in the initial ramdisk image. If you are not sure, leave it set to "0". + +config RD_GZIP + bool "Initial ramdisk compressed using gzip" + default y + depends on BLK_DEV_INITRD=y + select DECOMPRESS_GZIP + help + Support loading of a gzip encoded initial ramdisk or cpio buffer. + If unsure, say Y. + +config RD_BZIP2 + bool "Initial ramdisk compressed using bzip2" + default n + depends on BLK_DEV_INITRD=y + select DECOMPRESS_BZIP2 + help + Support loading of a bzip2 encoded initial ramdisk or cpio buffer + If unsure, say N. + +config RD_LZMA + bool "Initial ramdisk compressed using lzma" + default n + depends on BLK_DEV_INITRD=y + select DECOMPRESS_LZMA + help + Support loading of a lzma encoded initial ramdisk or cpio buffer + If unsure, say N. -- cgit v0.10.2 From a26ee60f90daffe1de6be0d093af86e7279b3dfd Mon Sep 17 00:00:00 2001 From: Alain Knaff Date: Wed, 7 Jan 2009 00:10:27 -0800 Subject: bzip2/lzma: fix built-in initramfs vs CONFIG_RD_GZIP Impact: Resolves build failures in some configurations Makes it possible to disable CONFIG_RD_GZIP . In that case, the built-in initramfs will be compressed by whatever compressor is available (bzip2 or lzma) or left uncompressed if none is available. It also removes a couple of warnings which occur when no ramdisk compression at all is chosen. It also restores the select ZLIB_INFLATE in drivers/block/Kconfig which somehow came missing. This is needed to activate compilation of the stuff in zlib_deflate. Signed-off-by: Alain Knaff Signed-off-by: H. Peter Anvin diff --git a/init/initramfs.c b/init/initramfs.c index 40bd4fb..a3ba91c 100644 --- a/init/initramfs.c +++ b/init/initramfs.c @@ -389,7 +389,7 @@ static int __init write_buffer(char *buf, unsigned len) return len - count; } - +#if defined CONFIG_RD_GZIP || defined CONFIG_RD_BZIP2 || defined CONFIG_RD_LZMA static int __init flush_buffer(void *bufv, unsigned len) { char *buf = (char *) bufv; @@ -412,6 +412,7 @@ static int __init flush_buffer(void *bufv, unsigned len) } return origLen; } +#endif static unsigned my_inptr; /* index of next byte to be processed in inbuf */ @@ -449,10 +450,12 @@ static char * __init unpack_to_rootfs(char *buf, unsigned len, int check_only) continue; } this_header = 0; +#ifdef CONFIG_RD_GZIP if (!gunzip(buf, len, NULL, flush_buffer, NULL, &my_inptr, error) && message == NULL) goto ok; +#endif #ifdef CONFIG_RD_BZIP2 message = NULL; /* Zero out message, or else cpio will @@ -473,7 +476,9 @@ static char * __init unpack_to_rootfs(char *buf, unsigned len, int check_only) goto ok; } #endif +#if defined CONFIG_RD_GZIP || defined CONFIG_RD_BZIP2 || defined CONFIG_RD_LZMA ok: +#endif if (state != Reset) error("junk in compressed archive"); this_header = saved_offset + my_inptr; diff --git a/scripts/gen_initramfs_list.sh b/scripts/gen_initramfs_list.sh index 5f3415f..41041e4 100644 --- a/scripts/gen_initramfs_list.sh +++ b/scripts/gen_initramfs_list.sh @@ -5,7 +5,7 @@ # Released under the terms of the GNU GPL # # Generate a cpio packed initramfs. It uses gen_init_cpio to generate -# the cpio archive, and gzip to pack it. +# the cpio archive, and then compresses it. # The script may also be used to generate the inputfile used for gen_init_cpio # This script assumes that gen_init_cpio is located in usr/ directory @@ -16,8 +16,8 @@ usage() { cat << EOF Usage: $0 [-o ] [-u ] [-g ] {-d | } ... - -o Create gzipped initramfs file named using - gen_init_cpio and gzip + -o Create compressed initramfs file named using + gen_init_cpio and compressor depending on the extension -u User ID to map to user ID 0 (root). is only meaningful if is a directory. "squash" forces all files to uid 0. @@ -225,6 +225,7 @@ cpio_list= output="/dev/stdout" output_file="" is_cpio_compressed= +compr="gzip -9 -f" arg="$1" case "$arg" in @@ -233,11 +234,15 @@ case "$arg" in echo "deps_initramfs := \\" shift ;; - "-o") # generate gzipped cpio image named $1 + "-o") # generate compressed cpio image named $1 shift output_file="$1" cpio_list="$(mktemp ${TMPDIR:-/tmp}/cpiolist.XXXXXX)" output=${cpio_list} + echo "$output_file" | grep -q "\.gz$" && compr="gzip -9 -f" + echo "$output_file" | grep -q "\.bz2$" && compr="bzip2 -9 -f" + echo "$output_file" | grep -q "\.lzma$" && compr="lzma -9 -f" + echo "$output_file" | grep -q "\.cpio$" && compr="cat" shift ;; esac @@ -274,7 +279,7 @@ while [ $# -gt 0 ]; do esac done -# If output_file is set we will generate cpio archive and gzip it +# If output_file is set we will generate cpio archive and compress it # we are carefull to delete tmp files if [ ! -z ${output_file} ]; then if [ -z ${cpio_file} ]; then @@ -287,7 +292,7 @@ if [ ! -z ${output_file} ]; then if [ "${is_cpio_compressed}" = "compressed" ]; then cat ${cpio_tfile} > ${output_file} else - cat ${cpio_tfile} | gzip -f -9 - > ${output_file} + cat ${cpio_tfile} | ${compr} - > ${output_file} fi [ -z ${cpio_file} ] && rm ${cpio_tfile} fi diff --git a/usr/Makefile b/usr/Makefile index 201f27f..451cdff 100644 --- a/usr/Makefile +++ b/usr/Makefile @@ -5,14 +5,32 @@ klibcdirs:; PHONY += klibcdirs +# Find out "preferred" ramdisk compressor. Order of preference is +# 1. bzip2 efficient, and likely to be present +# 2. gzip former default +# 3. lzma +# 4. none + +# None of the above +suffix_y = + +# Lzma, but no gzip nor bzip2 +suffix_$(CONFIG_RD_LZMA) = .lzma + +# Gzip, but no bzip2 +suffix_$(CONFIG_RD_GZIP) = .gz + +# Bzip2 +suffix_$(CONFIG_RD_BZIP2) = .bz2 + # Generate builtin.o based on initramfs_data.o -obj-$(CONFIG_BLK_DEV_INITRD) := initramfs_data.o +obj-$(CONFIG_BLK_DEV_INITRD) := initramfs_data$(suffix_y).o -# initramfs_data.o contains the initramfs_data.cpio.gz image. +# initramfs_data.o contains the compressed initramfs_data.cpio image. # The image is included using .incbin, a dependency which is not # tracked automatically. -$(obj)/initramfs_data.o: $(obj)/initramfs_data.cpio.gz FORCE +$(obj)/initramfs_data$(suffix_y).o: $(obj)/initramfs_data.cpio$(suffix_y) FORCE ##### # Generate the initramfs cpio archive @@ -25,28 +43,28 @@ ramfs-args := \ $(if $(CONFIG_INITRAMFS_ROOT_UID), -u $(CONFIG_INITRAMFS_ROOT_UID)) \ $(if $(CONFIG_INITRAMFS_ROOT_GID), -g $(CONFIG_INITRAMFS_ROOT_GID)) -# .initramfs_data.cpio.gz.d is used to identify all files included +# .initramfs_data.cpio.d is used to identify all files included # in initramfs and to detect if any files are added/removed. # Removed files are identified by directory timestamp being updated # The dependency list is generated by gen_initramfs.sh -l -ifneq ($(wildcard $(obj)/.initramfs_data.cpio.gz.d),) - include $(obj)/.initramfs_data.cpio.gz.d +ifneq ($(wildcard $(obj)/.initramfs_data.cpio.d),) + include $(obj)/.initramfs_data.cpio.d endif quiet_cmd_initfs = GEN $@ cmd_initfs = $(initramfs) -o $@ $(ramfs-args) $(ramfs-input) -targets := initramfs_data.cpio.gz +targets := initramfs_data.cpio.gz initramfs_data.cpio.bz2 initramfs_data.cpio.lzma initramfs_data.cpio # do not try to update files included in initramfs $(deps_initramfs): ; $(deps_initramfs): klibcdirs -# We rebuild initramfs_data.cpio.gz if: -# 1) Any included file is newer then initramfs_data.cpio.gz +# We rebuild initramfs_data.cpio if: +# 1) Any included file is newer then initramfs_data.cpio # 2) There are changes in which files are included (added or deleted) -# 3) If gen_init_cpio are newer than initramfs_data.cpio.gz +# 3) If gen_init_cpio are newer than initramfs_data.cpio # 4) arguments to gen_initramfs.sh changes -$(obj)/initramfs_data.cpio.gz: $(obj)/gen_init_cpio $(deps_initramfs) klibcdirs - $(Q)$(initramfs) -l $(ramfs-input) > $(obj)/.initramfs_data.cpio.gz.d +$(obj)/initramfs_data.cpio$(suffix_y): $(obj)/gen_init_cpio $(deps_initramfs) klibcdirs + $(Q)$(initramfs) -l $(ramfs-input) > $(obj)/.initramfs_data.cpio.d $(call if_changed,initfs) diff --git a/usr/initramfs_data.S b/usr/initramfs_data.S index c2e1ad4..7c6973d 100644 --- a/usr/initramfs_data.S +++ b/usr/initramfs_data.S @@ -26,5 +26,5 @@ SECTIONS */ .section .init.ramfs,"a" -.incbin "usr/initramfs_data.cpio.gz" +.incbin "usr/initramfs_data.cpio" diff --git a/usr/initramfs_data.bz2.S b/usr/initramfs_data.bz2.S new file mode 100644 index 0000000..bc54d09 --- /dev/null +++ b/usr/initramfs_data.bz2.S @@ -0,0 +1,29 @@ +/* + initramfs_data includes the compressed binary that is the + filesystem used for early user space. + Note: Older versions of "as" (prior to binutils 2.11.90.0.23 + released on 2001-07-14) dit not support .incbin. + If you are forced to use older binutils than that then the + following trick can be applied to create the resulting binary: + + + ld -m elf_i386 --format binary --oformat elf32-i386 -r \ + -T initramfs_data.scr initramfs_data.cpio.gz -o initramfs_data.o + ld -m elf_i386 -r -o built-in.o initramfs_data.o + + initramfs_data.scr looks like this: +SECTIONS +{ + .init.ramfs : { *(.data) } +} + + The above example is for i386 - the parameters vary from architectures. + Eventually look up LDFLAGS_BLOB in an older version of the + arch/$(ARCH)/Makefile to see the flags used before .incbin was introduced. + + Using .incbin has the advantage over ld that the correct flags are set + in the ELF header, as required by certain architectures. +*/ + +.section .init.ramfs,"a" +.incbin "usr/initramfs_data.cpio.bz2" diff --git a/usr/initramfs_data.gz.S b/usr/initramfs_data.gz.S new file mode 100644 index 0000000..890c8dd --- /dev/null +++ b/usr/initramfs_data.gz.S @@ -0,0 +1,29 @@ +/* + initramfs_data includes the compressed binary that is the + filesystem used for early user space. + Note: Older versions of "as" (prior to binutils 2.11.90.0.23 + released on 2001-07-14) dit not support .incbin. + If you are forced to use older binutils than that then the + following trick can be applied to create the resulting binary: + + + ld -m elf_i386 --format binary --oformat elf32-i386 -r \ + -T initramfs_data.scr initramfs_data.cpio.gz -o initramfs_data.o + ld -m elf_i386 -r -o built-in.o initramfs_data.o + + initramfs_data.scr looks like this: +SECTIONS +{ + .init.ramfs : { *(.data) } +} + + The above example is for i386 - the parameters vary from architectures. + Eventually look up LDFLAGS_BLOB in an older version of the + arch/$(ARCH)/Makefile to see the flags used before .incbin was introduced. + + Using .incbin has the advantage over ld that the correct flags are set + in the ELF header, as required by certain architectures. +*/ + +.section .init.ramfs,"a" +.incbin "usr/initramfs_data.cpio.gz" diff --git a/usr/initramfs_data.lzma.S b/usr/initramfs_data.lzma.S new file mode 100644 index 0000000..e11469e --- /dev/null +++ b/usr/initramfs_data.lzma.S @@ -0,0 +1,29 @@ +/* + initramfs_data includes the compressed binary that is the + filesystem used for early user space. + Note: Older versions of "as" (prior to binutils 2.11.90.0.23 + released on 2001-07-14) dit not support .incbin. + If you are forced to use older binutils than that then the + following trick can be applied to create the resulting binary: + + + ld -m elf_i386 --format binary --oformat elf32-i386 -r \ + -T initramfs_data.scr initramfs_data.cpio.gz -o initramfs_data.o + ld -m elf_i386 -r -o built-in.o initramfs_data.o + + initramfs_data.scr looks like this: +SECTIONS +{ + .init.ramfs : { *(.data) } +} + + The above example is for i386 - the parameters vary from architectures. + Eventually look up LDFLAGS_BLOB in an older version of the + arch/$(ARCH)/Makefile to see the flags used before .incbin was introduced. + + Using .incbin has the advantage over ld that the correct flags are set + in the ELF header, as required by certain architectures. +*/ + +.section .init.ramfs,"a" +.incbin "usr/initramfs_data.cpio.lzma" -- cgit v0.10.2 From 5619448fc55650cb76cbb825e16e4bb5d1390399 Mon Sep 17 00:00:00 2001 From: "H. Peter Anvin" Date: Thu, 8 Jan 2009 15:09:12 -0800 Subject: bzip2/lzma: fix constant in decompress_inflate Impact: Cleanup Fix constant 0x8100 /* 32K */; according to Alain the value 0x8100 was left over test code to test misalignment, the correct value is indeed 0x8000 == 32K. Signed-off-by: H. Peter Anvin diff --git a/lib/decompress_inflate.c b/lib/decompress_inflate.c index 163e66a..d2e7c75 100644 --- a/lib/decompress_inflate.c +++ b/lib/decompress_inflate.c @@ -41,7 +41,7 @@ STATIC int INIT gunzip(unsigned char *buf, int len, set_error_fn(error_fn); rc = -1; if (flush) { - out_len = 0x8100; /* 32 K */ + out_len = 0x8000; /* 32 K */ out_buf = malloc(out_len); } else { out_len = 0x7fffffff; /* no limit */ -- cgit v0.10.2 From 6c11b12ac6f101732d43b5682f5b3ae4dda4205f Mon Sep 17 00:00:00 2001 From: Alain Knaff Date: Thu, 8 Jan 2009 15:10:19 -0800 Subject: bzip2/lzma: fix decompress_inflate.c vs multi-block-with-embedded-filename Impact: Bug fix Fix gunzip uncompression, so that it also works with files with embedded filenames that are larger than one block. Signed-off-by: Alain Knaff Signed-off-by: H. Peter Anvin diff --git a/lib/decompress_inflate.c b/lib/decompress_inflate.c index d2e7c75..839a329 100644 --- a/lib/decompress_inflate.c +++ b/lib/decompress_inflate.c @@ -97,7 +97,7 @@ STATIC int INIT gunzip(unsigned char *buf, int len, strm->next_in++; strm->next_in++; } - strm->avail_in = len - 10; + strm->avail_in = len - (strm->next_in - zbuf); strm->next_out = out_buf; strm->avail_out = out_len; -- cgit v0.10.2 From 889c92d21db40be0b7d22a59395060237895bb85 Mon Sep 17 00:00:00 2001 From: "H. Peter Anvin" Date: Thu, 8 Jan 2009 15:14:17 -0800 Subject: bzip2/lzma: centralize format detection Centralize the compression format detection to a common routine in the lib directory, and use it for both initramfs and initrd. Signed-off-by: H. Peter Anvin diff --git a/include/linux/decompress/generic.h b/include/linux/decompress/generic.h index f847f51..6dfb856 100644 --- a/include/linux/decompress/generic.h +++ b/include/linux/decompress/generic.h @@ -26,5 +26,8 @@ typedef int (*decompress_fn) (unsigned char *inbuf, int len, *fill should be called (repeatedly...) to read data, at most IOBUF_SIZE */ +/* Utility routine to detect the decompression method */ +decompress_fn decompress_method(const unsigned char *inbuf, int len, + const char **name); #endif diff --git a/init/do_mounts_rd.c b/init/do_mounts_rd.c index 9c9d7db..a06ed4f 100644 --- a/init/do_mounts_rd.c +++ b/init/do_mounts_rd.c @@ -12,9 +12,6 @@ #include -#include -#include -#include int __initdata rd_prompt = 1;/* 1 = prompt for RAM disk, 0 = don't prompt */ @@ -49,24 +46,6 @@ static int __init crd_load(int in_fd, int out_fd, decompress_fn deco); * cramfs * gzip */ -static const struct compress_format { - unsigned char magic[2]; - const char *name; - decompress_fn decompressor; -} compressed_formats[] = { -#ifdef CONFIG_RD_GZIP - { {037, 0213}, "gzip", gunzip }, - { {037, 0236}, "gzip", gunzip }, -#endif -#ifdef CONFIG_RD_BZIP2 - { {0x42, 0x5a}, "bzip2", bunzip2 }, -#endif -#ifdef CONFIG_RD_LZMA - { {0x5d, 0x00}, "lzma", unlzma }, -#endif - { {0, 0}, NULL, NULL } -}; - static int __init identify_ramdisk_image(int fd, int start_block, decompress_fn *decompressor) { @@ -77,7 +56,7 @@ identify_ramdisk_image(int fd, int start_block, decompress_fn *decompressor) struct cramfs_super *cramfsb; int nblocks = -1; unsigned char *buf; - const struct compress_format *cf; + const char *compress_name; buf = kmalloc(size, GFP_KERNEL); if (!buf) @@ -95,15 +74,12 @@ identify_ramdisk_image(int fd, int start_block, decompress_fn *decompressor) sys_lseek(fd, start_block * BLOCK_SIZE, 0); sys_read(fd, buf, size); - for (cf = compressed_formats; cf->decompressor; cf++) { - if (buf[0] == cf->magic[0] && buf[1] == cf->magic[1]) { - printk(KERN_NOTICE - "RAMDISK: %s image found at block %d\n", - cf->name, start_block); - *decompressor = cf->decompressor; - nblocks = 0; - goto done; - } + *decompressor = decompress_method(buf, size, &compress_name); + if (*decompressor) { + printk(KERN_NOTICE "RAMDISK: %s image found at block %d\n", + compress_name, start_block); + nblocks = 0; + goto done; } /* romfs is at block zero too */ diff --git a/init/initramfs.c b/init/initramfs.c index a3ba91c..2f42984 100644 --- a/init/initramfs.c +++ b/init/initramfs.c @@ -416,13 +416,13 @@ static int __init flush_buffer(void *bufv, unsigned len) static unsigned my_inptr; /* index of next byte to be processed in inbuf */ -#include -#include -#include +#include static char * __init unpack_to_rootfs(char *buf, unsigned len, int check_only) { int written; + decompress_fn decompress; + dry_run = check_only; header_buf = kmalloc(110, GFP_KERNEL); symlink_buf = kmalloc(PATH_MAX + N_ALIGN(PATH_MAX) + 1, GFP_KERNEL); @@ -450,35 +450,10 @@ static char * __init unpack_to_rootfs(char *buf, unsigned len, int check_only) continue; } this_header = 0; -#ifdef CONFIG_RD_GZIP - if (!gunzip(buf, len, NULL, flush_buffer, NULL, - &my_inptr, error) && - message == NULL) - goto ok; -#endif - -#ifdef CONFIG_RD_BZIP2 - message = NULL; /* Zero out message, or else cpio will - think an error has already occured */ - if (!bunzip2(buf, len, NULL, flush_buffer, NULL, - &my_inptr, error) && - message == NULL) { - goto ok; - } -#endif - -#ifdef CONFIG_RD_LZMA - message = NULL; /* Zero out message, or else cpio will - think an error has already occured */ - if (!unlzma(buf, len, NULL, flush_buffer, NULL, - &my_inptr, error) && - message == NULL) { - goto ok; - } -#endif -#if defined CONFIG_RD_GZIP || defined CONFIG_RD_BZIP2 || defined CONFIG_RD_LZMA -ok: -#endif + decompress = decompress_method(buf, len, NULL); + if (decompress) + decompress(buf, len, NULL, flush_buffer, NULL, + &my_inptr, error); if (state != Reset) error("junk in compressed archive"); this_header = saved_offset + my_inptr; diff --git a/lib/Makefile b/lib/Makefile index d9ac5a4..790de7c 100644 --- a/lib/Makefile +++ b/lib/Makefile @@ -11,7 +11,8 @@ lib-y := ctype.o string.o vsprintf.o cmdline.o \ rbtree.o radix-tree.o dump_stack.o \ idr.o int_sqrt.o extable.o prio_tree.o \ sha1.o irq_regs.o reciprocal_div.o argv_split.o \ - proportions.o prio_heap.o ratelimit.o show_mem.o is_single_threaded.o + proportions.o prio_heap.o ratelimit.o show_mem.o \ + is_single_threaded.o decompress.o lib-$(CONFIG_MMU) += ioremap.o lib-$(CONFIG_SMP) += cpumask.o @@ -65,9 +66,9 @@ obj-$(CONFIG_REED_SOLOMON) += reed_solomon/ obj-$(CONFIG_LZO_COMPRESS) += lzo/ obj-$(CONFIG_LZO_DECOMPRESS) += lzo/ -obj-$(CONFIG_DECOMPRESS_GZIP) += decompress_inflate.o -obj-$(CONFIG_DECOMPRESS_BZIP2) += decompress_bunzip2.o -obj-$(CONFIG_DECOMPRESS_LZMA) += decompress_unlzma.o +lib-$(CONFIG_DECOMPRESS_GZIP) += decompress_inflate.o +lib-$(CONFIG_DECOMPRESS_BZIP2) += decompress_bunzip2.o +lib-$(CONFIG_DECOMPRESS_LZMA) += decompress_unlzma.o obj-$(CONFIG_TEXTSEARCH) += textsearch.o obj-$(CONFIG_TEXTSEARCH_KMP) += ts_kmp.o diff --git a/lib/decompress.c b/lib/decompress.c new file mode 100644 index 0000000..edac55c --- /dev/null +++ b/lib/decompress.c @@ -0,0 +1,50 @@ +/* + * decompress.c + * + * Detect the decompression method based on magic number + */ + +#include + +#include +#include +#include + +#include +#include + +static const struct compress_format { + unsigned char magic[2]; + const char *name; + decompress_fn decompressor; +} compressed_formats[] = { +#ifdef CONFIG_DECOMPRESS_GZIP + { {037, 0213}, "gzip", gunzip }, + { {037, 0236}, "gzip", gunzip }, +#endif +#ifdef CONFIG_DECOMPRESS_BZIP2 + { {0x42, 0x5a}, "bzip2", bunzip2 }, +#endif +#ifdef CONFIG_DECOMPRESS_LZMA + { {0x5d, 0x00}, "lzma", unlzma }, +#endif + { {0, 0}, NULL, NULL } +}; + +decompress_fn decompress_method(const unsigned char *inbuf, int len, + const char **name) +{ + const struct compress_format *cf; + + if (len < 2) + return NULL; /* Need at least this much... */ + + for (cf = compressed_formats; cf->decompressor; cf++) { + if (!memcmp(inbuf, cf->magic, 2)) + break; + + } + if (name) + *name = cf->name; + return cf->decompressor; +} -- cgit v0.10.2 From 736f93236ce786d1bcf09ad4dcb38a360d35ea1b Mon Sep 17 00:00:00 2001 From: Ingo Molnar Date: Sat, 10 Jan 2009 12:06:19 +0100 Subject: bzip2/lzma: make flush_buffer() unconditional Impact: build fix flush_buffer() is used unconditionally: init/initramfs.c:456: error: 'flush_buffer' undeclared (first use in this function) init/initramfs.c:456: error: (Each undeclared identifier is reported only once init/initramfs.c:456: error: for each function it appears in.) So remove the decompressor #ifdefs from around it. Signed-off-by: Ingo Molnar diff --git a/init/initramfs.c b/init/initramfs.c index f8241e8..76f4a01 100644 --- a/init/initramfs.c +++ b/init/initramfs.c @@ -390,7 +390,6 @@ static int __init write_buffer(char *buf, unsigned len) return len - count; } -#if defined CONFIG_RD_GZIP || defined CONFIG_RD_BZIP2 || defined CONFIG_RD_LZMA static int __init flush_buffer(void *bufv, unsigned len) { char *buf = (char *) bufv; @@ -413,7 +412,6 @@ static int __init flush_buffer(void *bufv, unsigned len) } return origLen; } -#endif static unsigned my_inptr; /* index of next byte to be processed in inbuf */ -- cgit v0.10.2 From c9cbf3d3b35f198fab39e98d696312dd0b97a69a Mon Sep 17 00:00:00 2001 From: Dan Streetman Date: Sat, 10 Jan 2009 23:44:22 -0800 Subject: Input: usbtouchscreen - allow reporting calibrated data This patch adds a module parameter to report either the raw coordinate data or the hardware-calibrated coordinate data for MicroTouch/3M touchscreens. The default is set to the raw coordinates for backwards compatibilty. Signed-off-by: Dan Streetman Signed-off-by: Dmitry Torokhov diff --git a/drivers/input/touchscreen/usbtouchscreen.c b/drivers/input/touchscreen/usbtouchscreen.c index 5080b26..6d27a1d 100644 --- a/drivers/input/touchscreen/usbtouchscreen.c +++ b/drivers/input/touchscreen/usbtouchscreen.c @@ -60,6 +60,10 @@ static int swap_xy; module_param(swap_xy, bool, 0644); MODULE_PARM_DESC(swap_xy, "If set X and Y axes are swapped."); +static int hwcalib_xy; +module_param(hwcalib_xy, bool, 0644); +MODULE_PARM_DESC(hwcalib_xy, "If set hw-calibrated X/Y are used if available"); + /* device specifc data/functions */ struct usbtouch_usb; struct usbtouch_device_info { @@ -260,8 +264,13 @@ static int panjit_read_data(struct usbtouch_usb *dev, unsigned char *pkt) static int mtouch_read_data(struct usbtouch_usb *dev, unsigned char *pkt) { - dev->x = (pkt[8] << 8) | pkt[7]; - dev->y = (pkt[10] << 8) | pkt[9]; + if (hwcalib_xy) { + dev->x = (pkt[4] << 8) | pkt[3]; + dev->y = 0xffff - ((pkt[6] << 8) | pkt[5]); + } else { + dev->x = (pkt[8] << 8) | pkt[7]; + dev->y = (pkt[10] << 8) | pkt[9]; + } dev->touch = (pkt[2] & 0x40) ? 1 : 0; return 1; @@ -294,6 +303,12 @@ static int mtouch_init(struct usbtouch_usb *usbtouch) return ret; } + /* Default min/max xy are the raw values, override if using hw-calib */ + if (hwcalib_xy) { + input_set_abs_params(usbtouch->input, ABS_X, 0, 0xffff, 0, 0); + input_set_abs_params(usbtouch->input, ABS_Y, 0, 0xffff, 0, 0); + } + return 0; } #endif -- cgit v0.10.2 From 520abcdeb13de23e22b7d4367b1c3b136ef3b108 Mon Sep 17 00:00:00 2001 From: Daniel Mierswa Date: Sat, 10 Jan 2009 23:44:22 -0800 Subject: Input: atkbd - make forced_release_keys[] static Signed-off-by: Andrew Morton Signed-off-by: Dmitry Torokhov diff --git a/drivers/input/keyboard/atkbd.c b/drivers/input/keyboard/atkbd.c index f6e9f39..6df9ba1 100644 --- a/drivers/input/keyboard/atkbd.c +++ b/drivers/input/keyboard/atkbd.c @@ -839,7 +839,7 @@ static void atkbd_disconnect(struct serio *serio) */ static void atkbd_dell_laptop_keymap_fixup(struct atkbd *atkbd) { - const unsigned int forced_release_keys[] = { + static const unsigned int forced_release_keys[] = { 0x85, 0x86, 0x87, 0x88, 0x89, 0x8a, 0x8b, 0x8f, 0x93, }; int i; @@ -856,7 +856,7 @@ static void atkbd_dell_laptop_keymap_fixup(struct atkbd *atkbd) */ static void atkbd_hp_keymap_fixup(struct atkbd *atkbd) { - const unsigned int forced_release_keys[] = { + static const unsigned int forced_release_keys[] = { 0x94, }; int i; -- cgit v0.10.2 From 3c0340b774916caa4149892504169d290c8a720a Mon Sep 17 00:00:00 2001 From: Jean Delvare Date: Sat, 10 Jan 2009 23:44:22 -0800 Subject: Input: psmouse - make MOUSE_PS2_LIFEBOOK depend on X86 All Fujitsu-Siemens Lifebook systems are x86-based, so we might as well make MOUSE_PS2_LIFEBOOK depend on X86. This will avoid surprising things like: arch/arm/configs/s3c2410_defconfig:CONFIG_MOUSE_PS2_LIFEBOOK=y Signed-off-by: Jean Delvare Signed-off-by: Dmitry Torokhov diff --git a/drivers/input/mouse/Kconfig b/drivers/input/mouse/Kconfig index 093c8c1..9705f3a 100644 --- a/drivers/input/mouse/Kconfig +++ b/drivers/input/mouse/Kconfig @@ -70,7 +70,7 @@ config MOUSE_PS2_SYNAPTICS config MOUSE_PS2_LIFEBOOK bool "Fujitsu Lifebook PS/2 mouse protocol extension" if EMBEDDED default y - depends on MOUSE_PS2 + depends on MOUSE_PS2 && X86 help Say Y here if you have a Fujitsu B-series Lifebook PS/2 TouchScreen connected to your system. -- cgit v0.10.2 From 74ca11c2056d01d9ebb3615cd781a148450c3c82 Mon Sep 17 00:00:00 2001 From: Bastien Nocera Date: Sat, 10 Jan 2009 23:44:22 -0800 Subject: Input: uvc - the button on the camera is KEY_CAMERA Cameras should generate KEY_CAMERA, not BTN_0. Also call input_sync() on the device once the button has been pressed. Signed-off-by: Andrew Morton Signed-off-by: Dmitry Torokhov diff --git a/drivers/media/video/uvc/uvc_status.c b/drivers/media/video/uvc/uvc_status.c index 5d60b26..fdf7b41 100644 --- a/drivers/media/video/uvc/uvc_status.c +++ b/drivers/media/video/uvc/uvc_status.c @@ -47,8 +47,8 @@ static int uvc_input_init(struct uvc_device *dev) usb_to_input_id(udev, &input->id); input->dev.parent = &dev->intf->dev; - set_bit(EV_KEY, input->evbit); - set_bit(BTN_0, input->keybit); + __set_bit(EV_KEY, input->evbit); + __set_bit(KEY_CAMERA, input->keybit); if ((ret = input_register_device(input)) < 0) goto error; @@ -71,8 +71,10 @@ static void uvc_input_cleanup(struct uvc_device *dev) static void uvc_input_report_key(struct uvc_device *dev, unsigned int code, int value) { - if (dev->input) + if (dev->input) { input_report_key(dev->input, code, value); + input_sync(dev->input); + } } #else @@ -97,7 +99,7 @@ static void uvc_event_streaming(struct uvc_device *dev, __u8 *data, int len) return; uvc_trace(UVC_TRACE_STATUS, "Button (intf %u) %s len %d\n", data[1], data[3] ? "pressed" : "released", len); - uvc_input_report_key(dev, BTN_0, data[3]); + uvc_input_report_key(dev, KEY_CAMERA, data[3]); } else { uvc_trace(UVC_TRACE_STATUS, "Stream %u error event %02x %02x " "len %d.\n", data[1], data[2], data[3], len); -- cgit v0.10.2 From 57c1a24ee276a33190008243f986419e122c0dea Mon Sep 17 00:00:00 2001 From: Alexey Korolev Date: Tue, 6 Jan 2009 17:36:09 +0000 Subject: [MTD] [LPDDR] qinfo_probe depends on lpddr Signed-off-by: Alexey Korolev Signed-off-by: David Woodhouse diff --git a/drivers/mtd/lpddr/Kconfig b/drivers/mtd/lpddr/Kconfig index acd4ea9..5a401d8 100644 --- a/drivers/mtd/lpddr/Kconfig +++ b/drivers/mtd/lpddr/Kconfig @@ -12,6 +12,7 @@ config MTD_LPDDR DDR memories, intended for battery-operated systems. config MTD_QINFO_PROBE + depends on MTD_LPDDR tristate "Detect flash chips by QINFO probe" help Device Information for LPDDR chips is offered through the Overlay -- cgit v0.10.2 From 5f877607cdfe8b60bf96fb96e527e0ce2a21e68b Mon Sep 17 00:00:00 2001 From: Alan Cox Date: Sun, 11 Jan 2009 19:52:19 +0000 Subject: [MTD] map_rom has NULL erase pointer Which means if inftl or similar are loaded with it (which is a dumb thing to do admittedly) it may oops. Closes #8108 [dwmw2: change error to -EROFS to match write-protected flash] Signed-off-by: Alan Cox Signed-off-by: David Woodhouse diff --git a/drivers/mtd/chips/map_rom.c b/drivers/mtd/chips/map_rom.c index 821d0ed..c76d6e5 100644 --- a/drivers/mtd/chips/map_rom.c +++ b/drivers/mtd/chips/map_rom.c @@ -19,6 +19,7 @@ static int maprom_read (struct mtd_info *, loff_t, size_t, size_t *, u_char *); static int maprom_write (struct mtd_info *, loff_t, size_t, size_t *, const u_char *); static void maprom_nop (struct mtd_info *); static struct mtd_info *map_rom_probe(struct map_info *map); +static int maprom_erase (struct mtd_info *mtd, struct erase_info *info); static struct mtd_chip_driver maprom_chipdrv = { .probe = map_rom_probe, @@ -42,6 +43,7 @@ static struct mtd_info *map_rom_probe(struct map_info *map) mtd->read = maprom_read; mtd->write = maprom_write; mtd->sync = maprom_nop; + mtd->erase = maprom_erase; mtd->flags = MTD_CAP_ROM; mtd->erasesize = map->size; mtd->writesize = 1; @@ -71,6 +73,12 @@ static int maprom_write (struct mtd_info *mtd, loff_t to, size_t len, size_t *re return -EIO; } +static int maprom_erase (struct mtd_info *mtd, struct erase_info *info) +{ + /* We do our best 8) */ + return -EROFS; +} + static int __init map_rom_init(void) { register_mtd_chip_driver(&maprom_chipdrv); -- cgit v0.10.2 From 23a22d57a8962479ca630c9542e62d6f86fdf927 Mon Sep 17 00:00:00 2001 From: "H. Peter Anvin" Date: Mon, 12 Jan 2009 14:24:04 -0800 Subject: bzip2/lzma: comprehensible error messages for missing decompressor Instead of failing to identify a compressed image with a decompressor that we don't have compiled in, identify it and fail with a comprehensible panic message. Signed-off-by: H. Peter Anvin diff --git a/init/do_mounts_rd.c b/init/do_mounts_rd.c index a015e26..91d0cfc 100644 --- a/init/do_mounts_rd.c +++ b/init/do_mounts_rd.c @@ -79,9 +79,12 @@ identify_ramdisk_image(int fd, int start_block, decompress_fn *decompressor) sys_read(fd, buf, size); *decompressor = decompress_method(buf, size, &compress_name); - if (*decompressor) { + if (compress_name) { printk(KERN_NOTICE "RAMDISK: %s image found at block %d\n", compress_name, start_block); + if (!*decompressor) + printk(KERN_CRIT "RAMDISK: %s decompressor not configured!\n", + compress_name); nblocks = 0; goto done; } diff --git a/init/initramfs.c b/init/initramfs.c index 76f4a01..9a7290e 100644 --- a/init/initramfs.c +++ b/init/initramfs.c @@ -421,6 +421,8 @@ static char * __init unpack_to_rootfs(char *buf, unsigned len, int check_only) { int written; decompress_fn decompress; + const char *compress_name; + static __initdata char msg_buf[64]; dry_run = check_only; header_buf = kmalloc(110, GFP_KERNEL); @@ -449,10 +451,18 @@ static char * __init unpack_to_rootfs(char *buf, unsigned len, int check_only) continue; } this_header = 0; - decompress = decompress_method(buf, len, NULL); + decompress = decompress_method(buf, len, &compress_name); if (decompress) decompress(buf, len, NULL, flush_buffer, NULL, &my_inptr, error); + else if (compress_name) { + if (!message) { + snprintf(msg_buf, sizeof msg_buf, + "compression method %s not configured", + compress_name); + message = msg_buf; + } + } if (state != Reset) error("junk in compressed archive"); this_header = saved_offset + my_inptr; diff --git a/lib/decompress.c b/lib/decompress.c index edac55c..961f367 100644 --- a/lib/decompress.c +++ b/lib/decompress.c @@ -13,21 +13,25 @@ #include #include +#ifndef CONFIG_DECOMPRESS_GZIP +# define gunzip NULL +#endif +#ifndef CONFIG_DECOMPRESS_BZIP2 +# define bunzip2 NULL +#endif +#ifndef CONFIG_DECOMPRESS_LZMA +# define unlzma NULL +#endif + static const struct compress_format { unsigned char magic[2]; const char *name; decompress_fn decompressor; } compressed_formats[] = { -#ifdef CONFIG_DECOMPRESS_GZIP { {037, 0213}, "gzip", gunzip }, { {037, 0236}, "gzip", gunzip }, -#endif -#ifdef CONFIG_DECOMPRESS_BZIP2 { {0x42, 0x5a}, "bzip2", bunzip2 }, -#endif -#ifdef CONFIG_DECOMPRESS_LZMA { {0x5d, 0x00}, "lzma", unlzma }, -#endif { {0, 0}, NULL, NULL } }; -- cgit v0.10.2 From f2d8dc75a14479f8803a70cf637b5d79a3bb87f1 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Uwe=20Kleine-K=C3=B6nig?= Date: Mon, 12 Jan 2009 22:29:23 -0800 Subject: Input: corgikbd - mark probe function as __devinit MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit A pointer to corgikbd_probe is passed to the core via platform_driver_register and so the function must not disappear when the .init sections are discarded. Otherwise (if also having HOTPLUG=y) unbinding and binding a device to the driver via sysfs will result in an oops as does a device being registered late. An alternative to this patch is using platform_driver_probe instead of platform_driver_register plus removing the pointer to the probe function from the struct platform_driver. [dtor@mail.ru: fixed some more section markups] Signed-off-by: Uwe Kleine-König Signed-off-by: Dmitry Torokhov diff --git a/drivers/input/keyboard/corgikbd.c b/drivers/input/keyboard/corgikbd.c index c8ed065..abb04c8 100644 --- a/drivers/input/keyboard/corgikbd.c +++ b/drivers/input/keyboard/corgikbd.c @@ -288,7 +288,7 @@ static int corgikbd_resume(struct platform_device *dev) #define corgikbd_resume NULL #endif -static int __init corgikbd_probe(struct platform_device *pdev) +static int __devinit corgikbd_probe(struct platform_device *pdev) { struct corgikbd *corgikbd; struct input_dev *input_dev; @@ -368,7 +368,7 @@ static int __init corgikbd_probe(struct platform_device *pdev) return err; } -static int corgikbd_remove(struct platform_device *pdev) +static int __devexit corgikbd_remove(struct platform_device *pdev) { int i; struct corgikbd *corgikbd = platform_get_drvdata(pdev); @@ -388,7 +388,7 @@ static int corgikbd_remove(struct platform_device *pdev) static struct platform_driver corgikbd_driver = { .probe = corgikbd_probe, - .remove = corgikbd_remove, + .remove = __devexit_p(corgikbd_remove), .suspend = corgikbd_suspend, .resume = corgikbd_resume, .driver = { @@ -397,7 +397,7 @@ static struct platform_driver corgikbd_driver = { }, }; -static int __devinit corgikbd_init(void) +static int __init corgikbd_init(void) { return platform_driver_register(&corgikbd_driver); } -- cgit v0.10.2 From 840207edfa8b5e5b46e0d268bf33efe71fecea20 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Uwe=20Kleine-K=C3=B6nig?= Date: Mon, 12 Jan 2009 22:32:17 -0800 Subject: Input: corgi_ts - mark probe function as __devinit MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit A pointer to corgits_probe is passed to the core via platform_driver_register and so the function must not disappear when the .init sections are discarded. Otherwise (if also having HOTPLUG=y) unbinding and binding a device to the driver via sysfs will result in an oops as does a device being registered late. An alternative to this patch is using platform_driver_probe instead of platform_driver_register plus removing the pointer to the probe function from the struct platform_driver. [dtor@mail.ru: fixed some more section markups] Signed-off-by: Uwe Kleine-König Signed-off-by: Dmitry Torokhov diff --git a/drivers/input/touchscreen/corgi_ts.c b/drivers/input/touchscreen/corgi_ts.c index 65202c9..3fb51b5 100644 --- a/drivers/input/touchscreen/corgi_ts.c +++ b/drivers/input/touchscreen/corgi_ts.c @@ -268,7 +268,7 @@ static int corgits_resume(struct platform_device *dev) #define corgits_resume NULL #endif -static int __init corgits_probe(struct platform_device *pdev) +static int __devinit corgits_probe(struct platform_device *pdev) { struct corgi_ts *corgi_ts; struct input_dev *input_dev; @@ -343,7 +343,7 @@ static int __init corgits_probe(struct platform_device *pdev) return err; } -static int corgits_remove(struct platform_device *pdev) +static int __devexit corgits_remove(struct platform_device *pdev) { struct corgi_ts *corgi_ts = platform_get_drvdata(pdev); @@ -352,12 +352,13 @@ static int corgits_remove(struct platform_device *pdev) corgi_ts->machinfo->put_hsync(); input_unregister_device(corgi_ts->input); kfree(corgi_ts); + return 0; } static struct platform_driver corgits_driver = { .probe = corgits_probe, - .remove = corgits_remove, + .remove = __devexit_p(corgits_remove), .suspend = corgits_suspend, .resume = corgits_resume, .driver = { @@ -366,7 +367,7 @@ static struct platform_driver corgits_driver = { }, }; -static int __devinit corgits_init(void) +static int __init corgits_init(void) { return platform_driver_register(&corgits_driver); } -- cgit v0.10.2 From d63dab00e85641515a0a060e6de93e6b7f450665 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Uwe=20Kleine-K=C3=B6nig?= Date: Tue, 13 Jan 2009 18:20:20 -0800 Subject: Input: omap-keypad - mark probe function as __devinit MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit A pointer to omap_kp_probe is passed to the core via platform_driver_register and so the function must not disappear when the .init sections are discarded. Otherwise (if also having HOTPLUG=y) unbinding and binding a device to the driver via sysfs will result in an oops as does a device being registered late. An alternative to this patch is using platform_driver_probe instead of platform_driver_register plus removing the pointer to the probe function from the struct platform_driver. [dtor@mail.ru: fixed some more section markups] Signed-off-by: Uwe Kleine-König Signed-off-by: Dmitry Torokhov diff --git a/drivers/input/keyboard/omap-keypad.c b/drivers/input/keyboard/omap-keypad.c index ec0ebee..23b4fd1 100644 --- a/drivers/input/keyboard/omap-keypad.c +++ b/drivers/input/keyboard/omap-keypad.c @@ -279,7 +279,7 @@ static int omap_kp_resume(struct platform_device *dev) #define omap_kp_resume NULL #endif -static int __init omap_kp_probe(struct platform_device *pdev) +static int __devinit omap_kp_probe(struct platform_device *pdev) { struct omap_kp *omap_kp; struct input_dev *input_dev; @@ -422,7 +422,7 @@ err1: return -EINVAL; } -static int omap_kp_remove(struct platform_device *pdev) +static int __devexit omap_kp_remove(struct platform_device *pdev) { struct omap_kp *omap_kp = platform_get_drvdata(pdev); @@ -454,7 +454,7 @@ static int omap_kp_remove(struct platform_device *pdev) static struct platform_driver omap_kp_driver = { .probe = omap_kp_probe, - .remove = omap_kp_remove, + .remove = __devexit_p(omap_kp_remove), .suspend = omap_kp_suspend, .resume = omap_kp_resume, .driver = { @@ -463,7 +463,7 @@ static struct platform_driver omap_kp_driver = { }, }; -static int __devinit omap_kp_init(void) +static int __init omap_kp_init(void) { printk(KERN_INFO "OMAP Keypad Driver\n"); return platform_driver_register(&omap_kp_driver); -- cgit v0.10.2 From 27f23336bcc1b08512751c878b3e05ed6b815e86 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Uwe=20Kleine-K=C3=B6nig?= Date: Tue, 13 Jan 2009 18:20:31 -0800 Subject: Input: spitzkbd - mark probe function as __devinit MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit A pointer to spitzkbd_probe is passed to the core via platform_driver_register and so the function must not disappear when the .init sections are discarded. Otherwise (if also having HOTPLUG=y) unbinding and binding a device to the driver via sysfs will result in an oops as does a device being registered late. An alternative to this patch is using platform_driver_probe instead of platform_driver_register plus removing the pointer to the probe function from the struct platform_driver. [dtor@mail.ru: fixed some more section markups] Signed-off-by: Uwe Kleine-König Signed-off-by: Dmitry Torokhov diff --git a/drivers/input/keyboard/spitzkbd.c b/drivers/input/keyboard/spitzkbd.c index c48b76a..9d1781a 100644 --- a/drivers/input/keyboard/spitzkbd.c +++ b/drivers/input/keyboard/spitzkbd.c @@ -343,7 +343,7 @@ static int spitzkbd_resume(struct platform_device *dev) #define spitzkbd_resume NULL #endif -static int __init spitzkbd_probe(struct platform_device *dev) +static int __devinit spitzkbd_probe(struct platform_device *dev) { struct spitzkbd *spitzkbd; struct input_dev *input_dev; @@ -444,7 +444,7 @@ static int __init spitzkbd_probe(struct platform_device *dev) return err; } -static int spitzkbd_remove(struct platform_device *dev) +static int __devexit spitzkbd_remove(struct platform_device *dev) { int i; struct spitzkbd *spitzkbd = platform_get_drvdata(dev); @@ -470,7 +470,7 @@ static int spitzkbd_remove(struct platform_device *dev) static struct platform_driver spitzkbd_driver = { .probe = spitzkbd_probe, - .remove = spitzkbd_remove, + .remove = __devexit_p(spitzkbd_remove), .suspend = spitzkbd_suspend, .resume = spitzkbd_resume, .driver = { @@ -479,7 +479,7 @@ static struct platform_driver spitzkbd_driver = { }, }; -static int __devinit spitzkbd_init(void) +static int __init spitzkbd_init(void) { return platform_driver_register(&spitzkbd_driver); } -- cgit v0.10.2 From 73310a169aebe257efdd35a763cce1c7658f40c9 Mon Sep 17 00:00:00 2001 From: "H. Peter Anvin" Date: Wed, 14 Jan 2009 11:28:35 -0800 Subject: init: make initrd/initramfs decompression failure a KERN_EMERG event Impact: More consistent behaviour, avoid policy in the kernel Upgrade/downgrade initrd/initramfs decompression failure from inconsistently a panic or a KERN_ALERT message to a KERN_EMERG event. It is, however, possible do design a system which can recover from this (using the kernel builtin code and/or the internal initramfs), which means this is policy, not a technical necessity. A good way to handle this would be to have a panic-level=X option, to force a panic on a printk above a certain level. That is a separate patch, however. Signed-off-by: H. Peter Anvin diff --git a/init/do_mounts_rd.c b/init/do_mounts_rd.c index 91d0cfc..027a402 100644 --- a/init/do_mounts_rd.c +++ b/init/do_mounts_rd.c @@ -83,7 +83,8 @@ identify_ramdisk_image(int fd, int start_block, decompress_fn *decompressor) printk(KERN_NOTICE "RAMDISK: %s image found at block %d\n", compress_name, start_block); if (!*decompressor) - printk(KERN_CRIT "RAMDISK: %s decompressor not configured!\n", + printk(KERN_EMERG + "RAMDISK: %s decompressor not configured!\n", compress_name); nblocks = 0; goto done; diff --git a/init/initramfs.c b/init/initramfs.c index 9a7290e..7dcde7e 100644 --- a/init/initramfs.c +++ b/init/initramfs.c @@ -528,7 +528,7 @@ static int __init populate_rootfs(void) char *err = unpack_to_rootfs(__initramfs_start, __initramfs_end - __initramfs_start, 0); if (err) - panic(err); + panic(err); /* Failed to decompress INTERNAL initramfs */ if (initrd_start) { #ifdef CONFIG_BLK_DEV_RAM int fd; @@ -554,9 +554,12 @@ static int __init populate_rootfs(void) printk(KERN_INFO "Unpacking initramfs..."); err = unpack_to_rootfs((char *)initrd_start, initrd_end - initrd_start, 0); - if (err) - panic(err); - printk(" done\n"); + if (err) { + printk(" failed!\n"); + printk(KERN_EMERG "%s\n", err); + } else { + printk(" done\n"); + } free_initrd(); #endif } -- cgit v0.10.2 From 95c4bff0308eb0819436b730a836846d3e784657 Mon Sep 17 00:00:00 2001 From: Cyrill Gorcunov Date: Wed, 14 Jan 2009 23:37:46 +0300 Subject: x86: headers cleanup - boot.h Impact: cleanup 'make headers_check' warn us about leaking of kernel private (mostly compile time vars) data to userspace in headers. Fix it. Neither BOOT_HEAP_SIZE, BOOT_STACK_SIZE refs was found by searching thru net (ie in user-space area) so fence this all by __KERNEL__ guard. Signed-off-by: Cyrill Gorcunov Signed-off-by: H. Peter Anvin diff --git a/arch/x86/include/asm/boot.h b/arch/x86/include/asm/boot.h index c0e8e68..6526cf0 100644 --- a/arch/x86/include/asm/boot.h +++ b/arch/x86/include/asm/boot.h @@ -10,14 +10,16 @@ #define EXTENDED_VGA 0xfffe /* 80x50 mode */ #define ASK_VGA 0xfffd /* ask for it at bootup */ +#ifdef __KERNEL__ + /* Physical address where kernel should be loaded. */ #define LOAD_PHYSICAL_ADDR ((CONFIG_PHYSICAL_START \ + (CONFIG_PHYSICAL_ALIGN - 1)) \ & ~(CONFIG_PHYSICAL_ALIGN - 1)) -#if (defined CONFIG_KERNEL_BZIP2) +#ifdef CONFIG_KERNEL_BZIP2 #define BOOT_HEAP_SIZE 0x400000 -#else +#else /* !CONFIG_KERNEL_BZIP2 */ #ifdef CONFIG_X86_64 #define BOOT_HEAP_SIZE 0x7000 @@ -25,7 +27,7 @@ #define BOOT_HEAP_SIZE 0x4000 #endif -#endif +#endif /* !CONFIG_KERNEL_BZIP2 */ #ifdef CONFIG_X86_64 #define BOOT_STACK_SIZE 0x4000 @@ -33,4 +35,6 @@ #define BOOT_STACK_SIZE 0x1000 #endif +#endif /* __KERNEL__ */ + #endif /* _ASM_X86_BOOT_H */ -- cgit v0.10.2 From ee287587dafd77fd211e50637561224605c214b4 Mon Sep 17 00:00:00 2001 From: "H. Peter Anvin" Date: Wed, 14 Jan 2009 16:07:38 -0800 Subject: bzip2/lzma: update boot protocol specification Impact: documentation Update the boot protocol specification to include the currently supported file formats and their magic numbers. Signed-off-by: H. Peter Anvin diff --git a/Documentation/x86/boot.txt b/Documentation/x86/boot.txt index fcdc62b..9ac70ff 100644 --- a/Documentation/x86/boot.txt +++ b/Documentation/x86/boot.txt @@ -542,7 +542,10 @@ Protocol: 2.08+ The payload may be compressed. The format of both the compressed and uncompressed data should be determined using the standard magic - numbers. Currently only gzip compressed ELF is used. + numbers. The currently supported compression formats are gzip + (magic numbers 1F 8B or 1F 9E), bzip2 (magic number 42 5A) and LZMA + (magic number 5D 00). The uncompressed payload is currently always ELF + (magic number 7F 45 4C 46). Field name: payload_length Type: read -- cgit v0.10.2 From 3afd522de8d8ec446efe957b86e4f63e3dd8ce9d Mon Sep 17 00:00:00 2001 From: Roel Kluin Date: Mon, 19 Jan 2009 00:15:13 +0100 Subject: [MTD] slram: Handle negative devlength correctly A negative devlength won't get noticed and clean up: Signed-off-by: Roel Kluin Signed-off-by: David Woodhouse diff --git a/drivers/mtd/devices/slram.c b/drivers/mtd/devices/slram.c index a425d09..00248e8 100644 --- a/drivers/mtd/devices/slram.c +++ b/drivers/mtd/devices/slram.c @@ -267,22 +267,28 @@ static int parse_cmdline(char *devname, char *szstart, char *szlength) if (*(szlength) != '+') { devlength = simple_strtoul(szlength, &buffer, 0); devlength = handle_unit(devlength, buffer) - devstart; + if (devlength < devstart) + goto err_out; + + devlength -= devstart; } else { devlength = simple_strtoul(szlength + 1, &buffer, 0); devlength = handle_unit(devlength, buffer); } T("slram: devname=%s, devstart=0x%lx, devlength=0x%lx\n", devname, devstart, devlength); - if ((devstart < 0) || (devlength < 0) || (devlength % SLRAM_BLK_SZ != 0)) { - E("slram: Illegal start / length parameter.\n"); - return(-EINVAL); - } + if (devlength % SLRAM_BLK_SZ != 0) + goto err_out; if ((devstart = register_device(devname, devstart, devlength))){ unregister_devices(); return((int)devstart); } return(0); + +err_out: + E("slram: Illegal length parameter.\n"); + return(-EINVAL); } #ifndef MODULE -- cgit v0.10.2 From e8e0c02340f2d3e4f1ea75a136883d8797290929 Mon Sep 17 00:00:00 2001 From: Roel Kluin Date: Thu, 29 Jan 2009 22:56:08 -0800 Subject: Input: bf54x-keys - fix debounce time validation Signed-off-by: Roel Kluin Acked-by: Michael Hennerich Signed-off-by: Dmitry Torokhov diff --git a/drivers/input/keyboard/bf54x-keys.c b/drivers/input/keyboard/bf54x-keys.c index 1928401..ee855c5 100644 --- a/drivers/input/keyboard/bf54x-keys.c +++ b/drivers/input/keyboard/bf54x-keys.c @@ -209,8 +209,8 @@ static int __devinit bfin_kpad_probe(struct platform_device *pdev) goto out; } - if (!pdata->debounce_time || !pdata->debounce_time > MAX_MULT || - !pdata->coldrive_time || !pdata->coldrive_time > MAX_MULT) { + if (!pdata->debounce_time || pdata->debounce_time > MAX_MULT || + !pdata->coldrive_time || pdata->coldrive_time > MAX_MULT) { printk(KERN_ERR DRV_NAME ": Invalid Debounce/Columdrive Time from pdata\n"); bfin_write_KPAD_MSEL(0xFF0); /* Default MSEL */ -- cgit v0.10.2 From 4e8718a1f960db0c48427f4583f89f4cb62f2480 Mon Sep 17 00:00:00 2001 From: Kay Sievers Date: Thu, 29 Jan 2009 22:56:08 -0800 Subject: Input: struct device - replace bus_id with dev_name(), dev_set_name() Acked-by: Greg Kroah-Hartman Signed-off-by: Kay Sievers Signed-off-by: Dmitry Torokhov diff --git a/drivers/input/serio/ambakmi.c b/drivers/input/serio/ambakmi.c index b10ffae..af159c7 100644 --- a/drivers/input/serio/ambakmi.c +++ b/drivers/input/serio/ambakmi.c @@ -129,8 +129,8 @@ static int amba_kmi_probe(struct amba_device *dev, void *id) io->write = amba_kmi_write; io->open = amba_kmi_open; io->close = amba_kmi_close; - strlcpy(io->name, dev->dev.bus_id, sizeof(io->name)); - strlcpy(io->phys, dev->dev.bus_id, sizeof(io->phys)); + strlcpy(io->name, dev_name(&dev->dev), sizeof(io->name)); + strlcpy(io->phys, dev_name(&dev->dev), sizeof(io->phys)); io->port_data = kmi; io->dev.parent = &dev->dev; diff --git a/drivers/input/serio/gscps2.c b/drivers/input/serio/gscps2.c index adc3bd6..bd0f92d 100644 --- a/drivers/input/serio/gscps2.c +++ b/drivers/input/serio/gscps2.c @@ -359,7 +359,7 @@ static int __init gscps2_probe(struct parisc_device *dev) snprintf(serio->name, sizeof(serio->name), "GSC PS/2 %s", (ps2port->id == GSC_ID_KEYBOARD) ? "keyboard" : "mouse"); - strlcpy(serio->phys, dev->dev.bus_id, sizeof(serio->phys)); + strlcpy(serio->phys, dev_name(&dev->dev), sizeof(serio->phys)); serio->id.type = SERIO_8042; serio->write = gscps2_write; serio->open = gscps2_open; diff --git a/drivers/input/serio/sa1111ps2.c b/drivers/input/serio/sa1111ps2.c index 2ad8878..57953c0 100644 --- a/drivers/input/serio/sa1111ps2.c +++ b/drivers/input/serio/sa1111ps2.c @@ -246,8 +246,8 @@ static int __devinit ps2_probe(struct sa1111_dev *dev) serio->write = ps2_write; serio->open = ps2_open; serio->close = ps2_close; - strlcpy(serio->name, dev->dev.bus_id, sizeof(serio->name)); - strlcpy(serio->phys, dev->dev.bus_id, sizeof(serio->phys)); + strlcpy(serio->name, dev_name(&dev->dev), sizeof(serio->name)); + strlcpy(serio->phys, dev_name(&dev->dev), sizeof(serio->phys)); serio->port_data = ps2if; serio->dev.parent = &dev->dev; ps2if->io = serio; diff --git a/drivers/input/touchscreen/atmel_tsadcc.c b/drivers/input/touchscreen/atmel_tsadcc.c index a89a6a8..055969e 100644 --- a/drivers/input/touchscreen/atmel_tsadcc.c +++ b/drivers/input/touchscreen/atmel_tsadcc.c @@ -236,7 +236,7 @@ static int __devinit atmel_tsadcc_probe(struct platform_device *pdev) ts_dev->bufferedmeasure = 0; snprintf(ts_dev->phys, sizeof(ts_dev->phys), - "%s/input0", pdev->dev.bus_id); + "%s/input0", dev_name(&pdev->dev)); input_dev->name = "atmel touch screen controller"; input_dev->phys = ts_dev->phys; diff --git a/drivers/input/touchscreen/tsc2007.c b/drivers/input/touchscreen/tsc2007.c index b75dc29..4ab0702 100644 --- a/drivers/input/touchscreen/tsc2007.c +++ b/drivers/input/touchscreen/tsc2007.c @@ -289,7 +289,8 @@ static int tsc2007_probe(struct i2c_client *client, pdata->init_platform_hw(); - snprintf(ts->phys, sizeof(ts->phys), "%s/input0", client->dev.bus_id); + snprintf(ts->phys, sizeof(ts->phys), + "%s/input0", dev_name(&client->dev)); input_dev->name = "TSC2007 Touchscreen"; input_dev->phys = ts->phys; -- cgit v0.10.2 From bc34496d63ec0a669d6825ea42275fd6fcbe9969 Mon Sep 17 00:00:00 2001 From: Roel Kluin Date: Sun, 1 Feb 2009 16:54:19 -0800 Subject: Input: pxa930_trkball - fix write timeout handling With a postfix decrement i reaches -1 rather than 0, but after the loop it is tested whether it has become 0. Signed-off-by: Roel Kluin Signed-off-by: Dmitry Torokhov diff --git a/drivers/input/mouse/pxa930_trkball.c b/drivers/input/mouse/pxa930_trkball.c index a0f45c4..784be69 100644 --- a/drivers/input/mouse/pxa930_trkball.c +++ b/drivers/input/mouse/pxa930_trkball.c @@ -83,7 +83,7 @@ static int write_tbcr(struct pxa930_trkball *trkball, int v) __raw_writel(v, trkball->mmio_base + TBCR); - while (i--) { + while (--i) { if (__raw_readl(trkball->mmio_base + TBCR) == v) break; msleep(1); -- cgit v0.10.2 From 4ab73761faef832f6d378328f79d21e77c62af3b Mon Sep 17 00:00:00 2001 From: Roel Kluin Date: Sun, 1 Feb 2009 16:55:45 -0800 Subject: Input: ambakmi - fix timeout handling in amba_kmi_write() With a postfix decrement timeleft reaches -1 rather than 0, but after the loop it is tested to have become 0. Signed-off-by: Roel Kluin Signed-off-by: Dmitry Torokhov diff --git a/drivers/input/serio/ambakmi.c b/drivers/input/serio/ambakmi.c index af159c7..e29cdc1 100644 --- a/drivers/input/serio/ambakmi.c +++ b/drivers/input/serio/ambakmi.c @@ -57,7 +57,7 @@ static int amba_kmi_write(struct serio *io, unsigned char val) struct amba_kmi_port *kmi = io->port_data; unsigned int timeleft = 10000; /* timeout in 100ms */ - while ((readb(KMISTAT) & KMISTAT_TXEMPTY) == 0 && timeleft--) + while ((readb(KMISTAT) & KMISTAT_TXEMPTY) == 0 && --timeleft) udelay(10); if (timeleft) -- cgit v0.10.2 From 43f7392ba9e2585bf34f21399b1ed78692b5d437 Mon Sep 17 00:00:00 2001 From: Joerg Roedel Date: Sat, 3 Jan 2009 23:56:27 +0100 Subject: intel-iommu: fix build error with INTR_REMAP=y and DMAR=n This fix should be safe since iommu->agaw is only used in intel-iommu.c. And this file is only compiled with DMAR=y. Signed-off-by: Joerg Roedel Signed-off-by: David Woodhouse diff --git a/drivers/pci/dmar.c b/drivers/pci/dmar.c index f5a662a..2b4162d 100644 --- a/drivers/pci/dmar.c +++ b/drivers/pci/dmar.c @@ -491,7 +491,7 @@ int alloc_iommu(struct dmar_drhd_unit *drhd) int map_size; u32 ver; static int iommu_allocated = 0; - int agaw; + int agaw = 0; iommu = kzalloc(sizeof(*iommu), GFP_KERNEL); if (!iommu) @@ -507,6 +507,7 @@ int alloc_iommu(struct dmar_drhd_unit *drhd) iommu->cap = dmar_readq(iommu->reg + DMAR_CAP_REG); iommu->ecap = dmar_readq(iommu->reg + DMAR_ECAP_REG); +#ifdef CONFIG_DMAR agaw = iommu_calculate_agaw(iommu); if (agaw < 0) { printk(KERN_ERR @@ -514,6 +515,7 @@ int alloc_iommu(struct dmar_drhd_unit *drhd) iommu->seq_id); goto error; } +#endif iommu->agaw = agaw; /* the registers might be more than one page */ -- cgit v0.10.2 From 704126ad81b8cb7d3d70adb9ecb143f4d3fb38af Mon Sep 17 00:00:00 2001 From: Yu Zhao Date: Sun, 4 Jan 2009 16:28:52 +0800 Subject: VT-d: handle Invalidation Queue Error to avoid system hang When hardware detects any error with a descriptor from the invalidation queue, it stops fetching new descriptors from the queue until software clears the Invalidation Queue Error bit in the Fault Status register. Following fix handles the IQE so the kernel won't be trapped in an infinite loop. Signed-off-by: Yu Zhao Signed-off-by: David Woodhouse diff --git a/drivers/pci/dmar.c b/drivers/pci/dmar.c index 2b4162d..8d3e9c2 100644 --- a/drivers/pci/dmar.c +++ b/drivers/pci/dmar.c @@ -573,19 +573,49 @@ static inline void reclaim_free_desc(struct q_inval *qi) } } +static int qi_check_fault(struct intel_iommu *iommu, int index) +{ + u32 fault; + int head; + struct q_inval *qi = iommu->qi; + int wait_index = (index + 1) % QI_LENGTH; + + fault = readl(iommu->reg + DMAR_FSTS_REG); + + /* + * If IQE happens, the head points to the descriptor associated + * with the error. No new descriptors are fetched until the IQE + * is cleared. + */ + if (fault & DMA_FSTS_IQE) { + head = readl(iommu->reg + DMAR_IQH_REG); + if ((head >> 4) == index) { + memcpy(&qi->desc[index], &qi->desc[wait_index], + sizeof(struct qi_desc)); + __iommu_flush_cache(iommu, &qi->desc[index], + sizeof(struct qi_desc)); + writel(DMA_FSTS_IQE, iommu->reg + DMAR_FSTS_REG); + return -EINVAL; + } + } + + return 0; +} + /* * Submit the queued invalidation descriptor to the remapping * hardware unit and wait for its completion. */ -void qi_submit_sync(struct qi_desc *desc, struct intel_iommu *iommu) +int qi_submit_sync(struct qi_desc *desc, struct intel_iommu *iommu) { + int rc = 0; struct q_inval *qi = iommu->qi; struct qi_desc *hw, wait_desc; int wait_index, index; unsigned long flags; if (!qi) - return; + return 0; hw = qi->desc; @@ -603,7 +633,8 @@ void qi_submit_sync(struct qi_desc *desc, struct intel_iommu *iommu) hw[index] = *desc; - wait_desc.low = QI_IWD_STATUS_DATA(2) | QI_IWD_STATUS_WRITE | QI_IWD_TYPE; + wait_desc.low = QI_IWD_STATUS_DATA(QI_DONE) | + QI_IWD_STATUS_WRITE | QI_IWD_TYPE; wait_desc.high = virt_to_phys(&qi->desc_status[wait_index]); hw[wait_index] = wait_desc; @@ -614,13 +645,11 @@ void qi_submit_sync(struct qi_desc *desc, struct intel_iommu *iommu) qi->free_head = (qi->free_head + 2) % QI_LENGTH; qi->free_cnt -= 2; - spin_lock(&iommu->register_lock); /* * update the HW tail register indicating the presence of * new descriptors. */ writel(qi->free_head << 4, iommu->reg + DMAR_IQT_REG); - spin_unlock(&iommu->register_lock); while (qi->desc_status[wait_index] != QI_DONE) { /* @@ -630,15 +659,21 @@ void qi_submit_sync(struct qi_desc *desc, struct intel_iommu *iommu) * a deadlock where the interrupt context can wait indefinitely * for free slots in the queue. */ + rc = qi_check_fault(iommu, index); + if (rc) + goto out; + spin_unlock(&qi->q_lock); cpu_relax(); spin_lock(&qi->q_lock); } - - qi->desc_status[index] = QI_DONE; +out: + qi->desc_status[index] = qi->desc_status[wait_index] = QI_DONE; reclaim_free_desc(qi); spin_unlock_irqrestore(&qi->q_lock, flags); + + return rc; } /* @@ -651,13 +686,13 @@ void qi_global_iec(struct intel_iommu *iommu) desc.low = QI_IEC_TYPE; desc.high = 0; + /* should never fail */ qi_submit_sync(&desc, iommu); } int qi_flush_context(struct intel_iommu *iommu, u16 did, u16 sid, u8 fm, u64 type, int non_present_entry_flush) { - struct qi_desc desc; if (non_present_entry_flush) { @@ -671,10 +706,7 @@ int qi_flush_context(struct intel_iommu *iommu, u16 did, u16 sid, u8 fm, | QI_CC_GRAN(type) | QI_CC_TYPE; desc.high = 0; - qi_submit_sync(&desc, iommu); - - return 0; - + return qi_submit_sync(&desc, iommu); } int qi_flush_iotlb(struct intel_iommu *iommu, u16 did, u64 addr, @@ -704,10 +736,7 @@ int qi_flush_iotlb(struct intel_iommu *iommu, u16 did, u64 addr, desc.high = QI_IOTLB_ADDR(addr) | QI_IOTLB_IH(ih) | QI_IOTLB_AM(size_order); - qi_submit_sync(&desc, iommu); - - return 0; - + return qi_submit_sync(&desc, iommu); } /* diff --git a/drivers/pci/intr_remapping.c b/drivers/pci/intr_remapping.c index f78371b..45effc5 100644 --- a/drivers/pci/intr_remapping.c +++ b/drivers/pci/intr_remapping.c @@ -207,7 +207,7 @@ int alloc_irte(struct intel_iommu *iommu, int irq, u16 count) return index; } -static void qi_flush_iec(struct intel_iommu *iommu, int index, int mask) +static int qi_flush_iec(struct intel_iommu *iommu, int index, int mask) { struct qi_desc desc; @@ -215,7 +215,7 @@ static void qi_flush_iec(struct intel_iommu *iommu, int index, int mask) | QI_IEC_SELECTIVE; desc.high = 0; - qi_submit_sync(&desc, iommu); + return qi_submit_sync(&desc, iommu); } int map_irq_to_irte_handle(int irq, u16 *sub_handle) @@ -283,6 +283,7 @@ int clear_irte_irq(int irq, struct intel_iommu *iommu, u16 index) int modify_irte(int irq, struct irte *irte_modified) { + int rc; int index; struct irte *irte; struct intel_iommu *iommu; @@ -303,14 +304,15 @@ int modify_irte(int irq, struct irte *irte_modified) set_64bit((unsigned long *)irte, irte_modified->low | (1 << 1)); __iommu_flush_cache(iommu, irte, sizeof(*irte)); - qi_flush_iec(iommu, index, 0); - + rc = qi_flush_iec(iommu, index, 0); spin_unlock(&irq_2_ir_lock); - return 0; + + return rc; } int flush_irte(int irq) { + int rc; int index; struct intel_iommu *iommu; struct irq_2_iommu *irq_iommu; @@ -326,10 +328,10 @@ int flush_irte(int irq) index = irq_iommu->irte_index + irq_iommu->sub_handle; - qi_flush_iec(iommu, index, irq_iommu->irte_mask); + rc = qi_flush_iec(iommu, index, irq_iommu->irte_mask); spin_unlock(&irq_2_ir_lock); - return 0; + return rc; } struct intel_iommu *map_ioapic_to_ir(int apic) @@ -355,6 +357,7 @@ struct intel_iommu *map_dev_to_ir(struct pci_dev *dev) int free_irte(int irq) { + int rc = 0; int index, i; struct irte *irte; struct intel_iommu *iommu; @@ -375,7 +378,7 @@ int free_irte(int irq) if (!irq_iommu->sub_handle) { for (i = 0; i < (1 << irq_iommu->irte_mask); i++) set_64bit((unsigned long *)irte, 0); - qi_flush_iec(iommu, index, irq_iommu->irte_mask); + rc = qi_flush_iec(iommu, index, irq_iommu->irte_mask); } irq_iommu->iommu = NULL; @@ -385,7 +388,7 @@ int free_irte(int irq) spin_unlock(&irq_2_ir_lock); - return 0; + return rc; } static void iommu_set_intr_remapping(struct intel_iommu *iommu, int mode) diff --git a/include/linux/intel-iommu.h b/include/linux/intel-iommu.h index c4f6c10..d2e3cbf 100644 --- a/include/linux/intel-iommu.h +++ b/include/linux/intel-iommu.h @@ -194,6 +194,7 @@ static inline void dmar_writeq(void __iomem *addr, u64 val) /* FSTS_REG */ #define DMA_FSTS_PPF ((u32)2) #define DMA_FSTS_PFO ((u32)1) +#define DMA_FSTS_IQE (1 << 4) #define dma_fsts_fault_record_index(s) (((s) >> 8) & 0xff) /* FRCD_REG, 32 bits access */ @@ -328,7 +329,7 @@ extern int qi_flush_iotlb(struct intel_iommu *iommu, u16 did, u64 addr, unsigned int size_order, u64 type, int non_present_entry_flush); -extern void qi_submit_sync(struct qi_desc *desc, struct intel_iommu *iommu); +extern int qi_submit_sync(struct qi_desc *desc, struct intel_iommu *iommu); extern void *intel_alloc_coherent(struct device *, size_t, dma_addr_t *, gfp_t); extern void intel_free_coherent(struct device *, size_t, void *, dma_addr_t); -- cgit v0.10.2 From 1b0e235cc9bfae4bc0f5cd0cba929206fb0f6a64 Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Wed, 11 Feb 2009 00:54:07 -0800 Subject: sparc64: Fix crashes in jbusmc_print_dimm() Return was missing for the case where there is no dimm info match. Signed-off-by: David S. Miller diff --git a/arch/sparc/kernel/chmc.c b/arch/sparc/kernel/chmc.c index 3b9f4d6..e1a9598 100644 --- a/arch/sparc/kernel/chmc.c +++ b/arch/sparc/kernel/chmc.c @@ -306,6 +306,7 @@ static int jbusmc_print_dimm(int syndrome_code, buf[1] = '?'; buf[2] = '?'; buf[3] = '\0'; + return 0; } p = dp->controller; prop = &p->layout; -- cgit v0.10.2 From 270c5609e2540290d5d9b9f323b7b652cf0c5b75 Mon Sep 17 00:00:00 2001 From: Tobias Klauser Date: Mon, 9 Feb 2009 20:59:40 +0000 Subject: sh: Storage class should be before const qualifier The C99 specification states in section 6.11.5: The placement of a storage-class specifier other than at the beginning of the declaration specifiers in a declaration is an obsolescent feature. Signed-off-by: Tobias Klauser Signed-off-by: Paul Mundt diff --git a/arch/sh/kernel/cpu/sh2a/clock-sh7201.c b/arch/sh/kernel/cpu/sh2a/clock-sh7201.c index 020a96f..4a5e597 100644 --- a/arch/sh/kernel/cpu/sh2a/clock-sh7201.c +++ b/arch/sh/kernel/cpu/sh2a/clock-sh7201.c @@ -18,8 +18,8 @@ #include #include -const static int pll1rate[]={1,2,3,4,6,8}; -const static int pfc_divisors[]={1,2,3,4,6,8,12}; +static const int pll1rate[]={1,2,3,4,6,8}; +static const int pfc_divisors[]={1,2,3,4,6,8,12}; #define ifc_divisors pfc_divisors #if (CONFIG_SH_CLK_MD == 0) -- cgit v0.10.2 From 34aeb43e2d3800f4d8f96feb9f1b49cd506679d5 Mon Sep 17 00:00:00 2001 From: Kuninori Morimoto Date: Tue, 10 Feb 2009 09:04:00 +0000 Subject: serial: sh-sci: fix overrun error handling for SH7785 SCIF. There was a typo for the overrun bit definition, causing it not to be handled correctly on SH7785, fix it up. Signed-off-by: Kuninori Morimoto Signed-off-by: Paul Mundt diff --git a/drivers/serial/sh-sci.h b/drivers/serial/sh-sci.h index 3599828..022e89f 100644 --- a/drivers/serial/sh-sci.h +++ b/drivers/serial/sh-sci.h @@ -133,7 +133,7 @@ # define SCSPTR3 0xffed0024 /* 16 bit SCIF */ # define SCSPTR4 0xffee0024 /* 16 bit SCIF */ # define SCSPTR5 0xffef0024 /* 16 bit SCIF */ -# define SCIF_OPER 0x0001 /* Overrun error bit */ +# define SCIF_ORER 0x0001 /* Overrun error bit */ # define SCSCR_INIT(port) 0x3a /* TIE=0,RIE=0,TE=1,RE=1,REIE=1 */ #elif defined(CONFIG_CPU_SUBTYPE_SH7201) || \ defined(CONFIG_CPU_SUBTYPE_SH7203) || \ -- cgit v0.10.2 From e480814f138cd5d78a8efe397756ba6b6518fdb6 Mon Sep 17 00:00:00 2001 From: Atsushi Nemoto Date: Wed, 11 Feb 2009 13:12:17 -0800 Subject: [MTD] [MAPS] physmap: fix wrong free and del_mtd_{partition,device} commit 176bf2e0f10ecf1d20a97db3bd5bb2e6ba0b5668 ("physmap: fix leak of memory returned by parse_mtd_partitions") deals with a memory leak and frees the pointer array of mtd_partition after the call to add_mtd_partitions(). the problem is that mtd_table[x]->name still points to the freed memory. Aldo physmap_flash_remove() should call del_mtd_partitions() or del_mtd_device() only once. Signed-off-by: Atsushi Nemoto Reported-by: Matthias Kaehlcke Tested-by: Matthias Kaehlcke Signed-off-by: Andrew Morton Signed-off-by: David Woodhouse diff --git a/drivers/mtd/maps/physmap.c b/drivers/mtd/maps/physmap.c index 8774366..4b122e7 100644 --- a/drivers/mtd/maps/physmap.c +++ b/drivers/mtd/maps/physmap.c @@ -29,6 +29,7 @@ struct physmap_flash_info { struct map_info map[MAX_RESOURCES]; #ifdef CONFIG_MTD_PARTITIONS int nr_parts; + struct mtd_partition *parts; #endif }; @@ -45,25 +46,26 @@ static int physmap_flash_remove(struct platform_device *dev) physmap_data = dev->dev.platform_data; -#ifdef CONFIG_MTD_CONCAT - if (info->cmtd != info->mtd[0]) { +#ifdef CONFIG_MTD_PARTITIONS + if (info->nr_parts) { + del_mtd_partitions(info->cmtd); + kfree(info->parts); + } else if (physmap_data->nr_parts) + del_mtd_partitions(info->cmtd); + else del_mtd_device(info->cmtd); +#else + del_mtd_device(info->cmtd); +#endif + +#ifdef CONFIG_MTD_CONCAT + if (info->cmtd != info->mtd[0]) mtd_concat_destroy(info->cmtd); - } #endif for (i = 0; i < MAX_RESOURCES; i++) { - if (info->mtd[i] != NULL) { -#ifdef CONFIG_MTD_PARTITIONS - if (info->nr_parts || physmap_data->nr_parts) - del_mtd_partitions(info->mtd[i]); - else - del_mtd_device(info->mtd[i]); -#else - del_mtd_device(info->mtd[i]); -#endif + if (info->mtd[i] != NULL) map_destroy(info->mtd[i]); - } } return 0; } @@ -86,9 +88,6 @@ static int physmap_flash_probe(struct platform_device *dev) int err = 0; int i; int devices_found = 0; -#ifdef CONFIG_MTD_PARTITIONS - struct mtd_partition *parts; -#endif physmap_data = dev->dev.platform_data; if (physmap_data == NULL) @@ -167,10 +166,11 @@ static int physmap_flash_probe(struct platform_device *dev) goto err_out; #ifdef CONFIG_MTD_PARTITIONS - err = parse_mtd_partitions(info->cmtd, part_probe_types, &parts, 0); + err = parse_mtd_partitions(info->cmtd, part_probe_types, + &info->parts, 0); if (err > 0) { - add_mtd_partitions(info->cmtd, parts, err); - kfree(parts); + add_mtd_partitions(info->cmtd, info->parts, err); + info->nr_parts = err; return 0; } -- cgit v0.10.2 From 10715b8751dfc403aeb7fbc35166417fa1664eda Mon Sep 17 00:00:00 2001 From: Mike Frysinger Date: Wed, 11 Feb 2009 13:12:18 -0800 Subject: [MTD] [MAPS] blackfin: fix memory leak in error path Signed-off-by: Mike Frysinger Signed-off-by: Bryan Wu Signed-off-by: Andrew Morton Signed-off-by: David Woodhouse diff --git a/drivers/mtd/maps/bfin-async-flash.c b/drivers/mtd/maps/bfin-async-flash.c index 6fec86a..576611f 100644 --- a/drivers/mtd/maps/bfin-async-flash.c +++ b/drivers/mtd/maps/bfin-async-flash.c @@ -152,14 +152,18 @@ static int __devinit bfin_flash_probe(struct platform_device *pdev) if (gpio_request(state->enet_flash_pin, DRIVER_NAME)) { pr_devinit(KERN_ERR DRIVER_NAME ": Failed to request gpio %d\n", state->enet_flash_pin); + kfree(state); return -EBUSY; } gpio_direction_output(state->enet_flash_pin, 1); pr_devinit(KERN_NOTICE DRIVER_NAME ": probing %d-bit flash bus\n", state->map.bankwidth * 8); state->mtd = do_map_probe(memory->name, &state->map); - if (!state->mtd) + if (!state->mtd) { + gpio_free(state->enet_flash_pin); + kfree(state); return -ENXIO; + } #ifdef CONFIG_MTD_PARTITIONS ret = parse_mtd_partitions(state->mtd, part_probe_types, &pdata->parts, 0); -- cgit v0.10.2 From ab00d68276295a1b4da7ad924a35a3566e9c2698 Mon Sep 17 00:00:00 2001 From: Mike Frysinger Date: Wed, 11 Feb 2009 13:12:19 -0800 Subject: [MTD] [MAPS] blackfin async requires complex mappings Correct a build error. bfin-async uses complex mappings and so needs it. Signed-off-by: Mike Frysinger Signed-off-by: Bryan Wu Cc: David Woodhouse Signed-off-by: Andrew Morton Signed-off-by: David Woodhouse diff --git a/drivers/mtd/maps/Kconfig b/drivers/mtd/maps/Kconfig index 0225cbb..043d50f 100644 --- a/drivers/mtd/maps/Kconfig +++ b/drivers/mtd/maps/Kconfig @@ -491,7 +491,7 @@ config MTD_PCMCIA_ANONYMOUS config MTD_BFIN_ASYNC tristate "Blackfin BF533-STAMP Flash Chip Support" - depends on BFIN533_STAMP && MTD_CFI + depends on BFIN533_STAMP && MTD_CFI && MTD_COMPLEX_MAPPINGS select MTD_PARTITIONS default y help -- cgit v0.10.2 From 084eb960e81505680a9963665722d1bfd94af6a7 Mon Sep 17 00:00:00 2001 From: Tony Battersby Date: Wed, 11 Feb 2009 13:24:19 -0800 Subject: intel-iommu: fix endless "Unknown DMAR structure type" loop I have a SuperMicro C2SBX motherboard with BIOS revision 1.0b. With vt-d enabled in the BIOS, Linux gets into an endless loop printing "DMAR:Unknown DMAR structure type" when booting. Here is the DMAR ACPI table: DMAR @ 0x7fe86dec 0000: 44 4d 41 52 98 00 00 00 01 6f 49 6e 74 65 6c 20 DMAR.....oIntel 0010: 4f 45 4d 44 4d 41 52 20 00 00 04 06 4c 4f 48 52 OEMDMAR ....LOHR 0020: 01 00 00 00 23 00 00 00 00 00 00 00 00 00 00 00 ....#........... 0030: 01 00 58 00 00 00 00 00 00 a0 e8 7f 00 00 00 00 ..X............. 0040: ff ff ef 7f 00 00 00 00 01 08 00 00 00 00 1d 00 ................ 0050: 01 08 00 00 00 00 1d 01 01 08 00 00 00 00 1d 02 ................ 0060: 01 08 00 00 00 00 1d 07 01 08 00 00 00 00 1a 00 ................ 0070: 01 08 00 00 00 00 1a 01 01 08 00 00 00 00 1a 02 ................ 0080: 01 08 00 00 00 00 1a 07 01 08 00 00 00 00 1a 07 ................ 0090: c0 00 68 00 04 10 66 60 ..h...f` Here are the messages printed by the kernel: DMAR:Host address width 36 DMAR:RMRR base: 0x000000007fe8a000 end: 0x000000007fefffff DMAR:Unknown DMAR structure type DMAR:Unknown DMAR structure type DMAR:Unknown DMAR structure type ... Although I not very familiar with ACPI, to me it looks like struct acpi_dmar_header::length == 0x0058 is incorrect, causing parse_dmar_table() to look at an invalid offset on the next loop. This offset happens to have struct acpi_dmar_header::length == 0x0000, which prevents the loop from ever terminating. This patch checks for this condition and bails out instead of looping forever. Signed-off-by: Tony Battersby Signed-off-by: Andrew Morton Signed-off-by: David Woodhouse diff --git a/drivers/pci/dmar.c b/drivers/pci/dmar.c index 8d3e9c2..26c536b 100644 --- a/drivers/pci/dmar.c +++ b/drivers/pci/dmar.c @@ -330,6 +330,14 @@ parse_dmar_table(void) entry_header = (struct acpi_dmar_header *)(dmar + 1); while (((unsigned long)entry_header) < (((unsigned long)dmar) + dmar_tbl->length)) { + /* Avoid looping forever on bad ACPI tables */ + if (entry_header->length == 0) { + printk(KERN_WARNING PREFIX + "Invalid 0-length structure\n"); + ret = -EINVAL; + break; + } + dmar_table_print_dmar_entry(entry_header); switch (entry_header->type) { -- cgit v0.10.2 From efab0b5d3eed6aa71f8e3233e4e11774eedc04dc Mon Sep 17 00:00:00 2001 From: Andres Salomon Date: Wed, 11 Feb 2009 13:27:02 -0800 Subject: [JFFS2] force the jffs2 GC daemon to behave a bit better I've noticed some pretty poor behavior on OLPC machines after bootup, when gdm/X are starting. The GCD monopolizes the scheduler (which in turns means it gets to do more nand i/o), which results in processes taking much much longer than they should to start. As an example, on an OLPC machine going from OFW to a usable X (via auto-login gdm) takes 2m 30s. The majority of this time is consumed by the switch into graphical mode. With this patch, we cut a full 60s off of bootup time. After bootup, things are much snappier as well. Note that we have seen a CRC node error with this patch that causes the machine to fail to boot, but we've also seen that problem without this patch. Signed-off-by: Andres Salomon Signed-off-by: Andrew Morton Signed-off-by: David Woodhouse diff --git a/fs/jffs2/background.c b/fs/jffs2/background.c index 3cceef4..e958010 100644 --- a/fs/jffs2/background.c +++ b/fs/jffs2/background.c @@ -95,13 +95,17 @@ static int jffs2_garbage_collect_thread(void *_c) spin_unlock(&c->erase_completion_lock); - /* This thread is purely an optimisation. But if it runs when - other things could be running, it actually makes things a - lot worse. Use yield() and put it at the back of the runqueue - every time. Especially during boot, pulling an inode in - with read_inode() is much preferable to having the GC thread - get there first. */ - yield(); + /* Problem - immediately after bootup, the GCD spends a lot + * of time in places like jffs2_kill_fragtree(); so much so + * that userspace processes (like gdm and X) are starved + * despite plenty of cond_resched()s and renicing. Yield() + * doesn't help, either (presumably because userspace and GCD + * are generally competing for a higher latency resource - + * disk). + * This forces the GCD to slow the hell down. Pulling an + * inode in with read_inode() is much preferable to having + * the GC thread get there first. */ + schedule_timeout_interruptible(msecs_to_jiffies(50)); /* Put_super will send a SIGKILL and then wait on the sem. */ -- cgit v0.10.2 From 4aa3b2ee1945ed082430ae1fb988d60eef64ca07 Mon Sep 17 00:00:00 2001 From: Eric Leblond Date: Wed, 18 Feb 2009 15:28:46 +0100 Subject: netfilter: nf_conntrack_ipv6: fix nf_log_packet message in icmpv6 conntrack This patch fixes a trivial typo that was adding a new line at end of the nf_log_packet() prefix. It also make the logging conditionnal by adding a LOG_INVALID test. Signed-off-by: Eric Leblond Signed-off-by: Patrick McHardy diff --git a/net/ipv6/netfilter/nf_conntrack_proto_icmpv6.c b/net/ipv6/netfilter/nf_conntrack_proto_icmpv6.c index c323643..72dbb6d 100644 --- a/net/ipv6/netfilter/nf_conntrack_proto_icmpv6.c +++ b/net/ipv6/netfilter/nf_conntrack_proto_icmpv6.c @@ -201,8 +201,9 @@ icmpv6_error(struct net *net, struct sk_buff *skb, unsigned int dataoff, if (net->ct.sysctl_checksum && hooknum == NF_INET_PRE_ROUTING && nf_ip6_checksum(skb, hooknum, dataoff, IPPROTO_ICMPV6)) { - nf_log_packet(PF_INET6, 0, skb, NULL, NULL, NULL, - "nf_ct_icmpv6: ICMPv6 checksum failed\n"); + if (LOG_INVALID(net, IPPROTO_ICMPV6)) + nf_log_packet(PF_INET6, 0, skb, NULL, NULL, NULL, + "nf_ct_icmpv6: ICMPv6 checksum failed "); return -NF_ACCEPT; } -- cgit v0.10.2 From 5ca431f9ae8db8c6edb9c64bebe6d6521077afd6 Mon Sep 17 00:00:00 2001 From: Eric Leblond Date: Wed, 18 Feb 2009 15:29:23 +0100 Subject: netfilter: nfnetlink_log: fix per-rule qthreshold override In NFLOG the per-rule qthreshold should overrides per-instance only it is set. With current code, the per-rule qthreshold is 1 if not set and it overrides the per-instance qthreshold. This patch modifies the default xt_NFLOG threshold from 1 to 0. Thus a value of 0 means there is no per-rule setting and the instance parameter has to apply. Signed-off-by: Eric Leblond Signed-off-by: Patrick McHardy diff --git a/include/linux/netfilter/xt_NFLOG.h b/include/linux/netfilter/xt_NFLOG.h index cdcd0ed..4b36aeb 100644 --- a/include/linux/netfilter/xt_NFLOG.h +++ b/include/linux/netfilter/xt_NFLOG.h @@ -2,7 +2,7 @@ #define _XT_NFLOG_TARGET #define XT_NFLOG_DEFAULT_GROUP 0x1 -#define XT_NFLOG_DEFAULT_THRESHOLD 1 +#define XT_NFLOG_DEFAULT_THRESHOLD 0 #define XT_NFLOG_MASK 0x0 diff --git a/net/netfilter/nfnetlink_log.c b/net/netfilter/nfnetlink_log.c index fa49dc7..580b837 100644 --- a/net/netfilter/nfnetlink_log.c +++ b/net/netfilter/nfnetlink_log.c @@ -590,8 +590,10 @@ nfulnl_log_packet(u_int8_t pf, qthreshold = inst->qthreshold; /* per-rule qthreshold overrides per-instance */ - if (qthreshold > li->u.ulog.qthreshold) - qthreshold = li->u.ulog.qthreshold; + if (li->u.ulog.qthreshold) + if (qthreshold > li->u.ulog.qthreshold) + qthreshold = li->u.ulog.qthreshold; + switch (inst->copy_mode) { case NFULNL_COPY_META: -- cgit v0.10.2 From 2c6764b743f9d25dd0806a417f06920dcbd0f599 Mon Sep 17 00:00:00 2001 From: Eric Leblond Date: Wed, 18 Feb 2009 15:29:49 +0100 Subject: netfilter: nfnetlink_log: fix timeout handling NFLOG timeout was computed in timer by doing: flushtimeout*HZ/100 Default value of flushtimeout was HZ (for 1 second delay). This was wrong for non 100HZ computer. This patch modify the default delay by using 100 instead of HZ. Signed-off-by: Eric Leblond Signed-off-by: Patrick McHardy diff --git a/net/netfilter/nfnetlink_log.c b/net/netfilter/nfnetlink_log.c index 580b837..c712e9f 100644 --- a/net/netfilter/nfnetlink_log.c +++ b/net/netfilter/nfnetlink_log.c @@ -39,7 +39,7 @@ #endif #define NFULNL_NLBUFSIZ_DEFAULT NLMSG_GOODSIZE -#define NFULNL_TIMEOUT_DEFAULT HZ /* every second */ +#define NFULNL_TIMEOUT_DEFAULT 100 /* every second */ #define NFULNL_QTHRESH_DEFAULT 100 /* 100 packets */ #define NFULNL_COPY_RANGE_MAX 0xFFFF /* max packet size is limited by 16-bit struct nfattr nfa_len field */ -- cgit v0.10.2 From 5962fc6d5fff09c8e6fb8cadcb18327a0f4277f7 Mon Sep 17 00:00:00 2001 From: Patrick McHardy Date: Wed, 18 Feb 2009 15:30:34 +0100 Subject: netfilter: nf_conntrack: don't try to deliver events for untracked connections The untracked conntrack actually does usually have events marked for delivery as its not special-cased in that part of the code. Skip the actual delivery since it impacts performance noticeably. Signed-off-by: Patrick McHardy diff --git a/include/net/netfilter/nf_conntrack_core.h b/include/net/netfilter/nf_conntrack_core.h index e78afe7..c25068e 100644 --- a/include/net/netfilter/nf_conntrack_core.h +++ b/include/net/netfilter/nf_conntrack_core.h @@ -59,7 +59,7 @@ static inline int nf_conntrack_confirm(struct sk_buff *skb) struct nf_conn *ct = (struct nf_conn *)skb->nfct; int ret = NF_ACCEPT; - if (ct) { + if (ct && ct != &nf_conntrack_untracked) { if (!nf_ct_is_confirmed(ct) && !nf_ct_is_dying(ct)) ret = __nf_conntrack_confirm(skb); nf_ct_deliver_cached_events(ct); -- cgit v0.10.2 From eb132205ca2f7ad44d8c8c482815b6911200b6a0 Mon Sep 17 00:00:00 2001 From: Jan Engelhardt Date: Wed, 18 Feb 2009 16:42:19 +0100 Subject: netfilter: make proc/net/ip* print names from foreign NFPROTO When extensions were moved to the NFPROTO_UNSPEC wildcard in ab4f21e6fb1c09b13c4c3cb8357babe8223471bd, they disappeared from the procfs files. Signed-off-by: Jan Engelhardt Signed-off-by: Patrick McHardy diff --git a/net/netfilter/x_tables.c b/net/netfilter/x_tables.c index bfbf521..5baccfa 100644 --- a/net/netfilter/x_tables.c +++ b/net/netfilter/x_tables.c @@ -827,59 +827,143 @@ static const struct file_operations xt_table_ops = { .release = seq_release_net, }; -static void *xt_match_seq_start(struct seq_file *seq, loff_t *pos) +/* + * Traverse state for ip{,6}_{tables,matches} for helping crossing + * the multi-AF mutexes. + */ +struct nf_mttg_trav { + struct list_head *head, *curr; + uint8_t class, nfproto; +}; + +enum { + MTTG_TRAV_INIT, + MTTG_TRAV_NFP_UNSPEC, + MTTG_TRAV_NFP_SPEC, + MTTG_TRAV_DONE, +}; + +static void *xt_mttg_seq_next(struct seq_file *seq, void *v, loff_t *ppos, + bool is_target) { - struct proc_dir_entry *pde = (struct proc_dir_entry *)seq->private; - u_int16_t af = (unsigned long)pde->data; + static const uint8_t next_class[] = { + [MTTG_TRAV_NFP_UNSPEC] = MTTG_TRAV_NFP_SPEC, + [MTTG_TRAV_NFP_SPEC] = MTTG_TRAV_DONE, + }; + struct nf_mttg_trav *trav = seq->private; + + switch (trav->class) { + case MTTG_TRAV_INIT: + trav->class = MTTG_TRAV_NFP_UNSPEC; + mutex_lock(&xt[NFPROTO_UNSPEC].mutex); + trav->head = trav->curr = is_target ? + &xt[NFPROTO_UNSPEC].target : &xt[NFPROTO_UNSPEC].match; + break; + case MTTG_TRAV_NFP_UNSPEC: + trav->curr = trav->curr->next; + if (trav->curr != trav->head) + break; + mutex_unlock(&xt[NFPROTO_UNSPEC].mutex); + mutex_lock(&xt[trav->nfproto].mutex); + trav->head = trav->curr = is_target ? + &xt[trav->nfproto].target : &xt[trav->nfproto].match; + trav->class = next_class[trav->class]; + break; + case MTTG_TRAV_NFP_SPEC: + trav->curr = trav->curr->next; + if (trav->curr != trav->head) + break; + /* fallthru, _stop will unlock */ + default: + return NULL; + } - mutex_lock(&xt[af].mutex); - return seq_list_start(&xt[af].match, *pos); + if (ppos != NULL) + ++*ppos; + return trav; } -static void *xt_match_seq_next(struct seq_file *seq, void *v, loff_t *pos) +static void *xt_mttg_seq_start(struct seq_file *seq, loff_t *pos, + bool is_target) { - struct proc_dir_entry *pde = (struct proc_dir_entry *)seq->private; - u_int16_t af = (unsigned long)pde->data; + struct nf_mttg_trav *trav = seq->private; + unsigned int j; - return seq_list_next(v, &xt[af].match, pos); + trav->class = MTTG_TRAV_INIT; + for (j = 0; j < *pos; ++j) + if (xt_mttg_seq_next(seq, NULL, NULL, is_target) == NULL) + return NULL; + return trav; } -static void xt_match_seq_stop(struct seq_file *seq, void *v) +static void xt_mttg_seq_stop(struct seq_file *seq, void *v) { - struct proc_dir_entry *pde = seq->private; - u_int16_t af = (unsigned long)pde->data; + struct nf_mttg_trav *trav = seq->private; + + switch (trav->class) { + case MTTG_TRAV_NFP_UNSPEC: + mutex_unlock(&xt[NFPROTO_UNSPEC].mutex); + break; + case MTTG_TRAV_NFP_SPEC: + mutex_unlock(&xt[trav->nfproto].mutex); + break; + } +} - mutex_unlock(&xt[af].mutex); +static void *xt_match_seq_start(struct seq_file *seq, loff_t *pos) +{ + return xt_mttg_seq_start(seq, pos, false); } -static int xt_match_seq_show(struct seq_file *seq, void *v) +static void *xt_match_seq_next(struct seq_file *seq, void *v, loff_t *ppos) { - struct xt_match *match = list_entry(v, struct xt_match, list); + return xt_mttg_seq_next(seq, v, ppos, false); +} - if (strlen(match->name)) - return seq_printf(seq, "%s\n", match->name); - else - return 0; +static int xt_match_seq_show(struct seq_file *seq, void *v) +{ + const struct nf_mttg_trav *trav = seq->private; + const struct xt_match *match; + + switch (trav->class) { + case MTTG_TRAV_NFP_UNSPEC: + case MTTG_TRAV_NFP_SPEC: + if (trav->curr == trav->head) + return 0; + match = list_entry(trav->curr, struct xt_match, list); + return (*match->name == '\0') ? 0 : + seq_printf(seq, "%s\n", match->name); + } + return 0; } static const struct seq_operations xt_match_seq_ops = { .start = xt_match_seq_start, .next = xt_match_seq_next, - .stop = xt_match_seq_stop, + .stop = xt_mttg_seq_stop, .show = xt_match_seq_show, }; static int xt_match_open(struct inode *inode, struct file *file) { + struct seq_file *seq; + struct nf_mttg_trav *trav; int ret; - ret = seq_open(file, &xt_match_seq_ops); - if (!ret) { - struct seq_file *seq = file->private_data; + trav = kmalloc(sizeof(*trav), GFP_KERNEL); + if (trav == NULL) + return -ENOMEM; - seq->private = PDE(inode); + ret = seq_open(file, &xt_match_seq_ops); + if (ret < 0) { + kfree(trav); + return ret; } - return ret; + + seq = file->private_data; + seq->private = trav; + trav->nfproto = (unsigned long)PDE(inode)->data; + return 0; } static const struct file_operations xt_match_ops = { @@ -887,62 +971,63 @@ static const struct file_operations xt_match_ops = { .open = xt_match_open, .read = seq_read, .llseek = seq_lseek, - .release = seq_release, + .release = seq_release_private, }; static void *xt_target_seq_start(struct seq_file *seq, loff_t *pos) { - struct proc_dir_entry *pde = (struct proc_dir_entry *)seq->private; - u_int16_t af = (unsigned long)pde->data; - - mutex_lock(&xt[af].mutex); - return seq_list_start(&xt[af].target, *pos); + return xt_mttg_seq_start(seq, pos, true); } -static void *xt_target_seq_next(struct seq_file *seq, void *v, loff_t *pos) +static void *xt_target_seq_next(struct seq_file *seq, void *v, loff_t *ppos) { - struct proc_dir_entry *pde = (struct proc_dir_entry *)seq->private; - u_int16_t af = (unsigned long)pde->data; - - return seq_list_next(v, &xt[af].target, pos); -} - -static void xt_target_seq_stop(struct seq_file *seq, void *v) -{ - struct proc_dir_entry *pde = seq->private; - u_int16_t af = (unsigned long)pde->data; - - mutex_unlock(&xt[af].mutex); + return xt_mttg_seq_next(seq, v, ppos, true); } static int xt_target_seq_show(struct seq_file *seq, void *v) { - struct xt_target *target = list_entry(v, struct xt_target, list); - - if (strlen(target->name)) - return seq_printf(seq, "%s\n", target->name); - else - return 0; + const struct nf_mttg_trav *trav = seq->private; + const struct xt_target *target; + + switch (trav->class) { + case MTTG_TRAV_NFP_UNSPEC: + case MTTG_TRAV_NFP_SPEC: + if (trav->curr == trav->head) + return 0; + target = list_entry(trav->curr, struct xt_target, list); + return (*target->name == '\0') ? 0 : + seq_printf(seq, "%s\n", target->name); + } + return 0; } static const struct seq_operations xt_target_seq_ops = { .start = xt_target_seq_start, .next = xt_target_seq_next, - .stop = xt_target_seq_stop, + .stop = xt_mttg_seq_stop, .show = xt_target_seq_show, }; static int xt_target_open(struct inode *inode, struct file *file) { + struct seq_file *seq; + struct nf_mttg_trav *trav; int ret; - ret = seq_open(file, &xt_target_seq_ops); - if (!ret) { - struct seq_file *seq = file->private_data; + trav = kmalloc(sizeof(*trav), GFP_KERNEL); + if (trav == NULL) + return -ENOMEM; - seq->private = PDE(inode); + ret = seq_open(file, &xt_target_seq_ops); + if (ret < 0) { + kfree(trav); + return ret; } - return ret; + + seq = file->private_data; + seq->private = trav; + trav->nfproto = (unsigned long)PDE(inode)->data; + return 0; } static const struct file_operations xt_target_ops = { @@ -950,7 +1035,7 @@ static const struct file_operations xt_target_ops = { .open = xt_target_open, .read = seq_read, .llseek = seq_lseek, - .release = seq_release, + .release = seq_release_private, }; #define FORMAT_TABLES "_tables_names" -- cgit v0.10.2 From e0ae4f5503235ba4449ffb5bcb4189edcef4d584 Mon Sep 17 00:00:00 2001 From: Yinghai Lu Date: Tue, 17 Feb 2009 20:40:09 -0800 Subject: PCI quirk: enable MSI on 8132 David reported that LSI SAS doesn't work with MSI. It turns out that his BIOS doesn't enable it, but the HT MSI 8132 does support HT MSI. Add quirk to enable it Cc: stable@kernel.org Reported-by: David Lang Signed-off-by: Yinghai Lu Signed-off-by: Jesse Barnes diff --git a/drivers/pci/quirks.c b/drivers/pci/quirks.c index baad093..a21e1c2 100644 --- a/drivers/pci/quirks.c +++ b/drivers/pci/quirks.c @@ -1981,7 +1981,6 @@ static void __devinit quirk_msi_ht_cap(struct pci_dev *dev) DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_SERVERWORKS, PCI_DEVICE_ID_SERVERWORKS_HT2000_PCIE, quirk_msi_ht_cap); - /* The nVidia CK804 chipset may have 2 HT MSI mappings. * MSI are supported if the MSI capability set in any of these mappings. */ @@ -2032,6 +2031,9 @@ DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_SERVERWORKS, PCI_DEVICE_ID_SERVERWORKS_HT1000_PXB, ht_enable_msi_mapping); +DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_8132_BRIDGE, + ht_enable_msi_mapping); + /* The P5N32-SLI Premium motherboard from Asus has a problem with msi * for the MCP55 NIC. It is not yet determined whether the msi problem * also affects other devices. As for now, turn off msi for this device. -- cgit v0.10.2 From e4aa7ca5a2e6d44f07ceb87d9448113f5b48a334 Mon Sep 17 00:00:00 2001 From: Alain Knaff Date: Thu, 19 Feb 2009 13:36:55 -0800 Subject: bzip2/lzma: don't stop search at first unconfigured compression Impact: Bugfix, avoids kernels which build but panic on boot Fix a bug in decompress.c : only scanned until the first non-configured compressor (with disastrous result especially if that was gzip.) Signed-off-by: Alain Knaff Signed-off-by: H. Peter Anvin diff --git a/lib/decompress.c b/lib/decompress.c index 961f367..d2842f5 100644 --- a/lib/decompress.c +++ b/lib/decompress.c @@ -43,7 +43,7 @@ decompress_fn decompress_method(const unsigned char *inbuf, int len, if (len < 2) return NULL; /* Need at least this much... */ - for (cf = compressed_formats; cf->decompressor; cf++) { + for (cf = compressed_formats; cf->name; cf++) { if (!memcmp(inbuf, cf->magic, 2)) break; -- cgit v0.10.2 From ab59d3b70f7a47987f80e153d828363faec643fe Mon Sep 17 00:00:00 2001 From: Alain Knaff Date: Thu, 19 Feb 2009 13:39:21 -0800 Subject: bzip2/lzma: don't leave empty files around on failure Impact: Bugfix, silent build failures Fix a bug in gen_initramfs_list.sh: in case of failure, it left an empty output file behind, messing up the next make. Signed-off-by: Alain Knaff Signed-off-by: H. Peter Anvin diff --git a/scripts/gen_initramfs_list.sh b/scripts/gen_initramfs_list.sh index 41041e4..3eea8f1 100644 --- a/scripts/gen_initramfs_list.sh +++ b/scripts/gen_initramfs_list.sh @@ -292,7 +292,8 @@ if [ ! -z ${output_file} ]; then if [ "${is_cpio_compressed}" = "compressed" ]; then cat ${cpio_tfile} > ${output_file} else - cat ${cpio_tfile} | ${compr} - > ${output_file} + (cat ${cpio_tfile} | ${compr} - > ${output_file}) \ + || (rm -f ${output_file} ; false) fi [ -z ${cpio_file} ] && rm ${cpio_tfile} fi -- cgit v0.10.2 From ab76f3d771590d5c89faa3219559c5d3fc0ce0c2 Mon Sep 17 00:00:00 2001 From: Alain Knaff Date: Thu, 19 Feb 2009 13:43:51 -0800 Subject: bzip2/lzma: make internal initramfs compression configurable Impact: Avoids silent environment dependency Make builtin initramfs compression an explicit configurable. The previous version would pick a compression based on the binaries which were installed on the system, which could lead to unexpected results. It is now explicitly configured, and not having the appropriate binaries installed on the build host is simply an error. Signed-off-by: Alain Knaff Signed-off-by: H. Peter Anvin diff --git a/usr/Kconfig b/usr/Kconfig index a691a8f..43a3a0f 100644 --- a/usr/Kconfig +++ b/usr/Kconfig @@ -71,3 +71,65 @@ config RD_LZMA help Support loading of a lzma encoded initial ramdisk or cpio buffer If unsure, say N. + +choice + prompt "Built-in initramfs compression mode" + help + This setting is only meaningful if the INITRAMFS_SOURCE is + set. It decides by which algorithm the INITRAMFS_SOURCE will + be compressed. + Several compression algorithms are available, which differ + in efficiency, compression and decompression speed. + Compression speed is only relevant when building a kernel. + Decompression speed is relevant at each boot. + + If you have any problems with bzip2 or lzma compressed + initramfs, mail me (Alain Knaff) . + + High compression options are mostly useful for users who + are low on disk space (embedded systems), but for whom ram + size matters less. + + If in doubt, select 'gzip' + +config INITRAMFS_COMPRESSION_NONE + bool "None" + help + Do not compress the built-in initramfs at all. This may + sound wasteful in space, but, you should be aware that the + built-in initramfs will be compressed at a later stage + anyways along with the rest of the kernel, on those + architectures that support this. + However, not compressing the initramfs may lead to slightly + higher memory consumption during a short time at boot, while + both the cpio image and the unpacked filesystem image will + be present in memory simultaneously + +config INITRAMFS_COMPRESSION_GZIP + bool "Gzip" + depends on RD_GZIP + help + The old and tried gzip compression. Its compression ratio is + the poorest among the 3 choices; however its speed (both + compression and decompression) is the fastest. + +config INITRAMFS_COMPRESSION_BZIP2 + bool "Bzip2" + depends on RD_BZIP2 + help + Its compression ratio and speed is intermediate. + Decompression speed is slowest among the three. The initramfs + size is about 10% smaller with bzip2, in comparison to gzip. + Bzip2 uses a large amount of memory. For modern kernels you + will need at least 8MB RAM or more for booting. + +config INITRAMFS_COMPRESSION_LZMA + bool "LZMA" + depends on RD_LZMA + help + The most recent compression algorithm. + Its ratio is best, decompression speed is between the other + two. Compression is slowest. The initramfs size is about 33% + smaller with LZMA in comparison to gzip. + +endchoice diff --git a/usr/Makefile b/usr/Makefile index 451cdff..b84894b 100644 --- a/usr/Makefile +++ b/usr/Makefile @@ -5,24 +5,18 @@ klibcdirs:; PHONY += klibcdirs -# Find out "preferred" ramdisk compressor. Order of preference is -# 1. bzip2 efficient, and likely to be present -# 2. gzip former default -# 3. lzma -# 4. none -# None of the above -suffix_y = - -# Lzma, but no gzip nor bzip2 -suffix_$(CONFIG_RD_LZMA) = .lzma +# No compression +suffix_$(CONFIG_INITRAMFS_COMPRESSION_NONE) = # Gzip, but no bzip2 -suffix_$(CONFIG_RD_GZIP) = .gz +suffix_$(CONFIG_INITRAMFS_COMPRESSION_GZIP) = .gz # Bzip2 -suffix_$(CONFIG_RD_BZIP2) = .bz2 +suffix_$(CONFIG_INITRAMFS_COMPRESSION_BZIP2) = .bz2 +# Lzma +suffix_$(CONFIG_INITRAMFS_COMPRESSION_LZMA) = .lzma # Generate builtin.o based on initramfs_data.o obj-$(CONFIG_BLK_DEV_INITRD) := initramfs_data$(suffix_y).o -- cgit v0.10.2 From e32740d9786b8a6c54f6e3d670567d9ef57b3b8c Mon Sep 17 00:00:00 2001 From: Harvey Harrison Date: Thu, 19 Feb 2009 11:58:37 -0800 Subject: ALSA: pcxhr.h replace signed one-bit bitfields The usage and comments make it clear values of 1/0 were intended rather than -1/0 Noticed by sparse: sound/pci/pcxhr/pcxhr.h:100:20: error: dubious one-bit signed bitfield sound/pci/pcxhr/pcxhr.h:101:22: error: dubious one-bit signed bitfield sound/pci/pcxhr/pcxhr.h:102:24: error: dubious one-bit signed bitfield sound/pci/pcxhr/pcxhr.h:103:21: error: dubious one-bit signed bitfield sound/pci/pcxhr/pcxhr.h:104:25: error: dubious one-bit signed bitfield sound/pci/pcxhr/pcxhr.h:105:20: error: dubious one-bit signed bitfield Signed-off-by: Harvey Harrison Signed-off-by: Takashi Iwai diff --git a/sound/pci/pcxhr/pcxhr.h b/sound/pci/pcxhr/pcxhr.h index 84131a9..69d87de 100644 --- a/sound/pci/pcxhr/pcxhr.h +++ b/sound/pci/pcxhr/pcxhr.h @@ -97,12 +97,12 @@ struct pcxhr_mgr { int capture_chips; int fw_file_set; int firmware_num; - int is_hr_stereo:1; - int board_has_aes1:1; /* if 1 board has AES1 plug and SRC */ - int board_has_analog:1; /* if 0 the board is digital only */ - int board_has_mic:1; /* if 1 the board has microphone input */ - int board_aes_in_192k:1;/* if 1 the aes input plugs do support 192kHz */ - int mono_capture:1; /* if 1 the board does mono capture */ + unsigned int is_hr_stereo:1; + unsigned int board_has_aes1:1; /* if 1 board has AES1 plug and SRC */ + unsigned int board_has_analog:1; /* if 0 the board is digital only */ + unsigned int board_has_mic:1; /* if 1 the board has microphone input */ + unsigned int board_aes_in_192k:1;/* if 1 the aes input plugs do support 192kHz */ + unsigned int mono_capture:1; /* if 1 the board does mono capture */ struct snd_dma_buffer hostport; -- cgit v0.10.2 From 55290e1932102f57ea17e7cff895914c2dbdb4c4 Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Fri, 20 Feb 2009 15:59:01 +0100 Subject: ALSA: hda - Fix parse of init_verbs sysfs entry Fixed the parse of init_verbs hwdep sysfs entry. Simplieied using sscanf. Signed-off-by: Takashi Iwai diff --git a/sound/pci/hda/hda_hwdep.c b/sound/pci/hda/hda_hwdep.c index 482fb03..4ae51dc 100644 --- a/sound/pci/hda/hda_hwdep.c +++ b/sound/pci/hda/hda_hwdep.c @@ -277,18 +277,19 @@ static ssize_t init_verbs_store(struct device *dev, { struct snd_hwdep *hwdep = dev_get_drvdata(dev); struct hda_codec *codec = hwdep->private_data; - char *p; - struct hda_verb verb, *v; + struct hda_verb *v; + int nid, verb, param; - verb.nid = simple_strtoul(buf, &p, 0); - verb.verb = simple_strtoul(p, &p, 0); - verb.param = simple_strtoul(p, &p, 0); - if (!verb.nid || !verb.verb || !verb.param) + if (sscanf(buf, "%i %i %i", &nid, &verb, ¶m) != 3) + return -EINVAL; + if (!nid || !verb) return -EINVAL; v = snd_array_new(&codec->init_verbs); if (!v) return -ENOMEM; - *v = verb; + v->nid = nid; + v->verb = verb; + v->param = param; return count; } -- cgit v0.10.2 From 6a63209fc02d5483371f07e4913ee8abad608051 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Fri, 20 Feb 2009 11:00:09 -0500 Subject: Btrfs: add better -ENOSPC handling This is a step in the direction of better -ENOSPC handling. Instead of checking the global bytes counter we check the space_info bytes counters to make sure we have enough space. If we don't we go ahead and try to allocate a new chunk, and then if that fails we return -ENOSPC. This patch adds two counters to btrfs_space_info, bytes_delalloc and bytes_may_use. bytes_delalloc account for extents we've actually setup for delalloc and will be allocated at some point down the line. bytes_may_use is to keep track of how many bytes we may use for delalloc at some point. When we actually set the extent_bit for the delalloc bytes we subtract the reserved bytes from the bytes_may_use counter. This keeps us from not actually being able to allocate space for any delalloc bytes. Signed-off-by: Josef Bacik diff --git a/fs/btrfs/btrfs_inode.h b/fs/btrfs/btrfs_inode.h index a8c9693..72677ce 100644 --- a/fs/btrfs/btrfs_inode.h +++ b/fs/btrfs/btrfs_inode.h @@ -66,6 +66,9 @@ struct btrfs_inode { */ struct list_head delalloc_inodes; + /* the space_info for where this inode's data allocations are done */ + struct btrfs_space_info *space_info; + /* full 64 bit generation number, struct vfs_inode doesn't have a big * enough field for this. */ @@ -94,6 +97,11 @@ struct btrfs_inode { */ u64 delalloc_bytes; + /* total number of bytes that may be used for this inode for + * delalloc + */ + u64 reserved_bytes; + /* * the size of the file stored in the metadata on disk. data=ordered * means the in-memory i_size might be larger than the size on disk diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 766b31a..82491ba 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -596,13 +596,27 @@ struct btrfs_block_group_item { struct btrfs_space_info { u64 flags; - u64 total_bytes; - u64 bytes_used; - u64 bytes_pinned; - u64 bytes_reserved; - u64 bytes_readonly; - int full; - int force_alloc; + + u64 total_bytes; /* total bytes in the space */ + u64 bytes_used; /* total bytes used on disk */ + u64 bytes_pinned; /* total bytes pinned, will be freed when the + transaction finishes */ + u64 bytes_reserved; /* total bytes the allocator has reserved for + current allocations */ + u64 bytes_readonly; /* total bytes that are read only */ + + /* delalloc accounting */ + u64 bytes_delalloc; /* number of bytes reserved for allocation, + this space is not necessarily reserved yet + by the allocator */ + u64 bytes_may_use; /* number of bytes that may be used for + delalloc */ + + int full; /* indicates that we cannot allocate any more + chunks for this space */ + int force_alloc; /* set if we need to force a chunk alloc for + this space */ + struct list_head list; /* for block groups in our same type */ @@ -1782,6 +1796,16 @@ int btrfs_add_dead_reloc_root(struct btrfs_root *root); int btrfs_cleanup_reloc_trees(struct btrfs_root *root); int btrfs_reloc_clone_csums(struct inode *inode, u64 file_pos, u64 len); u64 btrfs_reduce_alloc_profile(struct btrfs_root *root, u64 flags); +void btrfs_set_inode_space_info(struct btrfs_root *root, struct inode *ionde); +int btrfs_check_metadata_free_space(struct btrfs_root *root); +int btrfs_check_data_free_space(struct btrfs_root *root, struct inode *inode, + u64 bytes); +void btrfs_free_reserved_data_space(struct btrfs_root *root, + struct inode *inode, u64 bytes); +void btrfs_delalloc_reserve_space(struct btrfs_root *root, struct inode *inode, + u64 bytes); +void btrfs_delalloc_free_space(struct btrfs_root *root, struct inode *inode, + u64 bytes); /* ctree.c */ int btrfs_previous_item(struct btrfs_root *root, struct btrfs_path *path, u64 min_objectid, @@ -2027,8 +2051,6 @@ int btrfs_merge_bio_hook(struct page *page, unsigned long offset, unsigned long btrfs_force_ra(struct address_space *mapping, struct file_ra_state *ra, struct file *file, pgoff_t offset, pgoff_t last_index); -int btrfs_check_free_space(struct btrfs_root *root, u64 num_required, - int for_del); int btrfs_page_mkwrite(struct vm_area_struct *vma, struct page *page); int btrfs_readpage(struct file *file, struct page *page); void btrfs_delete_inode(struct inode *inode); diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 0a5d796..e11875e 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -60,6 +60,10 @@ static int update_block_group(struct btrfs_trans_handle *trans, u64 bytenr, u64 num_bytes, int alloc, int mark_free); +static int do_chunk_alloc(struct btrfs_trans_handle *trans, + struct btrfs_root *extent_root, u64 alloc_bytes, + u64 flags, int force); + static int block_group_bits(struct btrfs_block_group_cache *cache, u64 bits) { return (cache->flags & bits) == bits; @@ -1909,6 +1913,7 @@ static int update_space_info(struct btrfs_fs_info *info, u64 flags, found->bytes_pinned = 0; found->bytes_reserved = 0; found->bytes_readonly = 0; + found->bytes_delalloc = 0; found->full = 0; found->force_alloc = 0; *space_info = found; @@ -1972,6 +1977,196 @@ u64 btrfs_reduce_alloc_profile(struct btrfs_root *root, u64 flags) return flags; } +static u64 btrfs_get_alloc_profile(struct btrfs_root *root, u64 data) +{ + struct btrfs_fs_info *info = root->fs_info; + u64 alloc_profile; + + if (data) { + alloc_profile = info->avail_data_alloc_bits & + info->data_alloc_profile; + data = BTRFS_BLOCK_GROUP_DATA | alloc_profile; + } else if (root == root->fs_info->chunk_root) { + alloc_profile = info->avail_system_alloc_bits & + info->system_alloc_profile; + data = BTRFS_BLOCK_GROUP_SYSTEM | alloc_profile; + } else { + alloc_profile = info->avail_metadata_alloc_bits & + info->metadata_alloc_profile; + data = BTRFS_BLOCK_GROUP_METADATA | alloc_profile; + } + + return btrfs_reduce_alloc_profile(root, data); +} + +void btrfs_set_inode_space_info(struct btrfs_root *root, struct inode *inode) +{ + u64 alloc_target; + + alloc_target = btrfs_get_alloc_profile(root, 1); + BTRFS_I(inode)->space_info = __find_space_info(root->fs_info, + alloc_target); +} + +/* + * for now this just makes sure we have at least 5% of our metadata space free + * for use. + */ +int btrfs_check_metadata_free_space(struct btrfs_root *root) +{ + struct btrfs_fs_info *info = root->fs_info; + struct btrfs_space_info *meta_sinfo; + u64 alloc_target, thresh; + + /* get the space info for where the metadata will live */ + alloc_target = btrfs_get_alloc_profile(root, 0); + meta_sinfo = __find_space_info(info, alloc_target); + + /* + * if the metadata area isn't maxed out then there is no sense in + * checking how much is used, since we can always allocate a new chunk + */ + if (!meta_sinfo->full) + return 0; + + spin_lock(&meta_sinfo->lock); + thresh = meta_sinfo->total_bytes * 95; + + do_div(thresh, 100); + + if (meta_sinfo->bytes_used + meta_sinfo->bytes_reserved + + meta_sinfo->bytes_pinned + meta_sinfo->bytes_readonly > thresh) { + spin_unlock(&meta_sinfo->lock); + return -ENOSPC; + } + spin_unlock(&meta_sinfo->lock); + + return 0; +} + +/* + * This will check the space that the inode allocates from to make sure we have + * enough space for bytes. + */ +int btrfs_check_data_free_space(struct btrfs_root *root, struct inode *inode, + u64 bytes) +{ + struct btrfs_space_info *data_sinfo; + int ret = 0; + + /* make sure bytes are sectorsize aligned */ + bytes = (bytes + root->sectorsize - 1) & ~((u64)root->sectorsize - 1); + + data_sinfo = BTRFS_I(inode)->space_info; +again: + /* make sure we have enough space to handle the data first */ + spin_lock(&data_sinfo->lock); + if (data_sinfo->total_bytes - data_sinfo->bytes_used - + data_sinfo->bytes_delalloc - data_sinfo->bytes_reserved - + data_sinfo->bytes_pinned - data_sinfo->bytes_readonly - + data_sinfo->bytes_may_use < bytes) { + /* + * if we don't have enough free bytes in this space then we need + * to alloc a new chunk. + */ + if (!data_sinfo->full) { + u64 alloc_target; + struct btrfs_trans_handle *trans; + + data_sinfo->force_alloc = 1; + spin_unlock(&data_sinfo->lock); + + alloc_target = btrfs_get_alloc_profile(root, 1); + trans = btrfs_start_transaction(root, 1); + if (!trans) + return -ENOMEM; + + ret = do_chunk_alloc(trans, root->fs_info->extent_root, + bytes + 2 * 1024 * 1024, + alloc_target, 0); + btrfs_end_transaction(trans, root); + if (ret) + return ret; + goto again; + } + spin_unlock(&data_sinfo->lock); + printk(KERN_ERR "no space left, need %llu, %llu delalloc bytes" + ", %llu bytes_used, %llu bytes_reserved, " + "%llu bytes_pinned, %llu bytes_readonly, %llu may use" + "%llu total\n", bytes, data_sinfo->bytes_delalloc, + data_sinfo->bytes_used, data_sinfo->bytes_reserved, + data_sinfo->bytes_pinned, data_sinfo->bytes_readonly, + data_sinfo->bytes_may_use, data_sinfo->total_bytes); + return -ENOSPC; + } + data_sinfo->bytes_may_use += bytes; + BTRFS_I(inode)->reserved_bytes += bytes; + spin_unlock(&data_sinfo->lock); + + return btrfs_check_metadata_free_space(root); +} + +/* + * if there was an error for whatever reason after calling + * btrfs_check_data_free_space, call this so we can cleanup the counters. + */ +void btrfs_free_reserved_data_space(struct btrfs_root *root, + struct inode *inode, u64 bytes) +{ + struct btrfs_space_info *data_sinfo; + + /* make sure bytes are sectorsize aligned */ + bytes = (bytes + root->sectorsize - 1) & ~((u64)root->sectorsize - 1); + + data_sinfo = BTRFS_I(inode)->space_info; + spin_lock(&data_sinfo->lock); + data_sinfo->bytes_may_use -= bytes; + BTRFS_I(inode)->reserved_bytes -= bytes; + spin_unlock(&data_sinfo->lock); +} + +/* called when we are adding a delalloc extent to the inode's io_tree */ +void btrfs_delalloc_reserve_space(struct btrfs_root *root, struct inode *inode, + u64 bytes) +{ + struct btrfs_space_info *data_sinfo; + + /* get the space info for where this inode will be storing its data */ + data_sinfo = BTRFS_I(inode)->space_info; + + /* make sure we have enough space to handle the data first */ + spin_lock(&data_sinfo->lock); + data_sinfo->bytes_delalloc += bytes; + + /* + * we are adding a delalloc extent without calling + * btrfs_check_data_free_space first. This happens on a weird + * writepage condition, but shouldn't hurt our accounting + */ + if (unlikely(bytes > BTRFS_I(inode)->reserved_bytes)) { + data_sinfo->bytes_may_use -= BTRFS_I(inode)->reserved_bytes; + BTRFS_I(inode)->reserved_bytes = 0; + } else { + data_sinfo->bytes_may_use -= bytes; + BTRFS_I(inode)->reserved_bytes -= bytes; + } + + spin_unlock(&data_sinfo->lock); +} + +/* called when we are clearing an delalloc extent from the inode's io_tree */ +void btrfs_delalloc_free_space(struct btrfs_root *root, struct inode *inode, + u64 bytes) +{ + struct btrfs_space_info *info; + + info = BTRFS_I(inode)->space_info; + + spin_lock(&info->lock); + info->bytes_delalloc -= bytes; + spin_unlock(&info->lock); +} + static int do_chunk_alloc(struct btrfs_trans_handle *trans, struct btrfs_root *extent_root, u64 alloc_bytes, u64 flags, int force) @@ -3105,6 +3300,10 @@ static void dump_space_info(struct btrfs_space_info *info, u64 bytes) (unsigned long long)(info->total_bytes - info->bytes_used - info->bytes_pinned - info->bytes_reserved), (info->full) ? "" : "not "); + printk(KERN_INFO "space_info total=%llu, pinned=%llu, delalloc=%llu," + " may_use=%llu, used=%llu\n", info->total_bytes, + info->bytes_pinned, info->bytes_delalloc, info->bytes_may_use, + info->bytes_used); down_read(&info->groups_sem); list_for_each_entry(cache, &info->block_groups, list) { @@ -3131,24 +3330,10 @@ static int __btrfs_reserve_extent(struct btrfs_trans_handle *trans, { int ret; u64 search_start = 0; - u64 alloc_profile; struct btrfs_fs_info *info = root->fs_info; - if (data) { - alloc_profile = info->avail_data_alloc_bits & - info->data_alloc_profile; - data = BTRFS_BLOCK_GROUP_DATA | alloc_profile; - } else if (root == root->fs_info->chunk_root) { - alloc_profile = info->avail_system_alloc_bits & - info->system_alloc_profile; - data = BTRFS_BLOCK_GROUP_SYSTEM | alloc_profile; - } else { - alloc_profile = info->avail_metadata_alloc_bits & - info->metadata_alloc_profile; - data = BTRFS_BLOCK_GROUP_METADATA | alloc_profile; - } + data = btrfs_get_alloc_profile(root, data); again: - data = btrfs_reduce_alloc_profile(root, data); /* * the only place that sets empty_size is btrfs_realloc_node, which * is not called recursively on allocations diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c index 872f104..dc78954 100644 --- a/fs/btrfs/file.c +++ b/fs/btrfs/file.c @@ -1091,19 +1091,24 @@ static ssize_t btrfs_file_write(struct file *file, const char __user *buf, WARN_ON(num_pages > nrptrs); memset(pages, 0, sizeof(struct page *) * nrptrs); - ret = btrfs_check_free_space(root, write_bytes, 0); + ret = btrfs_check_data_free_space(root, inode, write_bytes); if (ret) goto out; ret = prepare_pages(root, file, pages, num_pages, pos, first_index, last_index, write_bytes); - if (ret) + if (ret) { + btrfs_free_reserved_data_space(root, inode, + write_bytes); goto out; + } ret = btrfs_copy_from_user(pos, num_pages, write_bytes, pages, buf); if (ret) { + btrfs_free_reserved_data_space(root, inode, + write_bytes); btrfs_drop_pages(pages, num_pages); goto out; } @@ -1111,8 +1116,11 @@ static ssize_t btrfs_file_write(struct file *file, const char __user *buf, ret = dirty_and_release_pages(NULL, root, file, pages, num_pages, pos, write_bytes); btrfs_drop_pages(pages, num_pages); - if (ret) + if (ret) { + btrfs_free_reserved_data_space(root, inode, + write_bytes); goto out; + } if (will_write) { btrfs_fdatawrite_range(inode->i_mapping, pos, @@ -1136,6 +1144,8 @@ static ssize_t btrfs_file_write(struct file *file, const char __user *buf, } out: mutex_unlock(&inode->i_mutex); + if (ret) + err = ret; out_nolock: kfree(pages); diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 3cee77a..7d4f948 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -102,34 +102,6 @@ static int btrfs_init_inode_security(struct inode *inode, struct inode *dir) } /* - * a very lame attempt at stopping writes when the FS is 85% full. There - * are countless ways this is incorrect, but it is better than nothing. - */ -int btrfs_check_free_space(struct btrfs_root *root, u64 num_required, - int for_del) -{ - u64 total; - u64 used; - u64 thresh; - int ret = 0; - - spin_lock(&root->fs_info->delalloc_lock); - total = btrfs_super_total_bytes(&root->fs_info->super_copy); - used = btrfs_super_bytes_used(&root->fs_info->super_copy); - if (for_del) - thresh = total * 90; - else - thresh = total * 85; - - do_div(thresh, 100); - - if (used + root->fs_info->delalloc_bytes + num_required > thresh) - ret = -ENOSPC; - spin_unlock(&root->fs_info->delalloc_lock); - return ret; -} - -/* * this does all the hard work for inserting an inline extent into * the btree. The caller should have done a btrfs_drop_extents so that * no overlapping inline items exist in the btree @@ -1190,6 +1162,7 @@ static int btrfs_set_bit_hook(struct inode *inode, u64 start, u64 end, */ if (!(old & EXTENT_DELALLOC) && (bits & EXTENT_DELALLOC)) { struct btrfs_root *root = BTRFS_I(inode)->root; + btrfs_delalloc_reserve_space(root, inode, end - start + 1); spin_lock(&root->fs_info->delalloc_lock); BTRFS_I(inode)->delalloc_bytes += end - start + 1; root->fs_info->delalloc_bytes += end - start + 1; @@ -1223,9 +1196,12 @@ static int btrfs_clear_bit_hook(struct inode *inode, u64 start, u64 end, (unsigned long long)end - start + 1, (unsigned long long) root->fs_info->delalloc_bytes); + btrfs_delalloc_free_space(root, inode, (u64)-1); root->fs_info->delalloc_bytes = 0; BTRFS_I(inode)->delalloc_bytes = 0; } else { + btrfs_delalloc_free_space(root, inode, + end - start + 1); root->fs_info->delalloc_bytes -= end - start + 1; BTRFS_I(inode)->delalloc_bytes -= end - start + 1; } @@ -2245,10 +2221,6 @@ static int btrfs_unlink(struct inode *dir, struct dentry *dentry) root = BTRFS_I(dir)->root; - ret = btrfs_check_free_space(root, 1, 1); - if (ret) - goto fail; - trans = btrfs_start_transaction(root, 1); btrfs_set_trans_block_group(trans, dir); @@ -2261,7 +2233,6 @@ static int btrfs_unlink(struct inode *dir, struct dentry *dentry) nr = trans->blocks_used; btrfs_end_transaction_throttle(trans, root); -fail: btrfs_btree_balance_dirty(root, nr); return ret; } @@ -2284,10 +2255,6 @@ static int btrfs_rmdir(struct inode *dir, struct dentry *dentry) return -ENOTEMPTY; } - ret = btrfs_check_free_space(root, 1, 1); - if (ret) - goto fail; - trans = btrfs_start_transaction(root, 1); btrfs_set_trans_block_group(trans, dir); @@ -2304,7 +2271,6 @@ static int btrfs_rmdir(struct inode *dir, struct dentry *dentry) fail_trans: nr = trans->blocks_used; ret = btrfs_end_transaction_throttle(trans, root); -fail: btrfs_btree_balance_dirty(root, nr); if (ret && !err) @@ -2818,7 +2784,7 @@ int btrfs_cont_expand(struct inode *inode, loff_t size) if (size <= hole_start) return 0; - err = btrfs_check_free_space(root, 1, 0); + err = btrfs_check_metadata_free_space(root); if (err) return err; @@ -3014,6 +2980,7 @@ static noinline void init_btrfs_i(struct inode *inode) bi->last_trans = 0; bi->logged_trans = 0; bi->delalloc_bytes = 0; + bi->reserved_bytes = 0; bi->disk_i_size = 0; bi->flags = 0; bi->index_cnt = (u64)-1; @@ -3035,6 +3002,7 @@ static int btrfs_init_locked_inode(struct inode *inode, void *p) inode->i_ino = args->ino; init_btrfs_i(inode); BTRFS_I(inode)->root = args->root; + btrfs_set_inode_space_info(args->root, inode); return 0; } @@ -3455,6 +3423,7 @@ static struct inode *btrfs_new_inode(struct btrfs_trans_handle *trans, BTRFS_I(inode)->index_cnt = 2; BTRFS_I(inode)->root = root; BTRFS_I(inode)->generation = trans->transid; + btrfs_set_inode_space_info(root, inode); if (mode & S_IFDIR) owner = 0; @@ -3602,7 +3571,7 @@ static int btrfs_mknod(struct inode *dir, struct dentry *dentry, if (!new_valid_dev(rdev)) return -EINVAL; - err = btrfs_check_free_space(root, 1, 0); + err = btrfs_check_metadata_free_space(root); if (err) goto fail; @@ -3665,7 +3634,7 @@ static int btrfs_create(struct inode *dir, struct dentry *dentry, u64 objectid; u64 index = 0; - err = btrfs_check_free_space(root, 1, 0); + err = btrfs_check_metadata_free_space(root); if (err) goto fail; trans = btrfs_start_transaction(root, 1); @@ -3733,7 +3702,7 @@ static int btrfs_link(struct dentry *old_dentry, struct inode *dir, return -ENOENT; btrfs_inc_nlink(inode); - err = btrfs_check_free_space(root, 1, 0); + err = btrfs_check_metadata_free_space(root); if (err) goto fail; err = btrfs_set_inode_index(dir, &index); @@ -3779,7 +3748,7 @@ static int btrfs_mkdir(struct inode *dir, struct dentry *dentry, int mode) u64 index = 0; unsigned long nr = 1; - err = btrfs_check_free_space(root, 1, 0); + err = btrfs_check_metadata_free_space(root); if (err) goto out_unlock; @@ -4336,7 +4305,7 @@ int btrfs_page_mkwrite(struct vm_area_struct *vma, struct page *page) u64 page_start; u64 page_end; - ret = btrfs_check_free_space(root, PAGE_CACHE_SIZE, 0); + ret = btrfs_check_data_free_space(root, inode, PAGE_CACHE_SIZE); if (ret) goto out; @@ -4349,6 +4318,7 @@ again: if ((page->mapping != inode->i_mapping) || (page_start >= size)) { + btrfs_free_reserved_data_space(root, inode, PAGE_CACHE_SIZE); /* page got truncated out from underneath us */ goto out_unlock; } @@ -4631,7 +4601,7 @@ static int btrfs_rename(struct inode *old_dir, struct dentry *old_dentry, if (old_inode->i_ino == BTRFS_FIRST_FREE_OBJECTID) return -EXDEV; - ret = btrfs_check_free_space(root, 1, 0); + ret = btrfs_check_metadata_free_space(root); if (ret) goto out_unlock; @@ -4749,7 +4719,7 @@ static int btrfs_symlink(struct inode *dir, struct dentry *dentry, if (name_len > BTRFS_MAX_INLINE_DATA_SIZE(root)) return -ENAMETOOLONG; - err = btrfs_check_free_space(root, 1, 0); + err = btrfs_check_metadata_free_space(root); if (err) goto out_fail; diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index 988fdc8..bca729f 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -70,7 +70,7 @@ static noinline int create_subvol(struct btrfs_root *root, u64 index = 0; unsigned long nr = 1; - ret = btrfs_check_free_space(root, 1, 0); + ret = btrfs_check_metadata_free_space(root); if (ret) goto fail_commit; @@ -203,7 +203,7 @@ static int create_snapshot(struct btrfs_root *root, struct dentry *dentry, if (!root->ref_cows) return -EINVAL; - ret = btrfs_check_free_space(root, 1, 0); + ret = btrfs_check_metadata_free_space(root); if (ret) goto fail_unlock; @@ -374,7 +374,7 @@ static int btrfs_defrag_file(struct file *file) unsigned long i; int ret; - ret = btrfs_check_free_space(root, inode->i_size, 0); + ret = btrfs_check_data_free_space(root, inode, inode->i_size); if (ret) return -ENOSPC; -- cgit v0.10.2 From 4e06bdd6cbd5105376e7caf4e683ed131e777389 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Fri, 20 Feb 2009 10:59:53 -0500 Subject: Btrfs: try committing transaction before returning ENOSPC This fixes a problem where we could return -ENOSPC when we may actually have plenty of space, the space is just pinned. Instead of returning -ENOSPC immediately, commit the transaction first and then try and do the allocation again. This patch also does chunk allocation for metadata if we pass the 80% threshold for metadata space. This will help with stack usage since the chunk allocation will happen early on, instead of when the allocation is happening. Signed-off-by: Josef Bacik diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index e11875e..6b5966a 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -2017,26 +2017,49 @@ int btrfs_check_metadata_free_space(struct btrfs_root *root) struct btrfs_fs_info *info = root->fs_info; struct btrfs_space_info *meta_sinfo; u64 alloc_target, thresh; + int committed = 0, ret; /* get the space info for where the metadata will live */ alloc_target = btrfs_get_alloc_profile(root, 0); meta_sinfo = __find_space_info(info, alloc_target); - /* - * if the metadata area isn't maxed out then there is no sense in - * checking how much is used, since we can always allocate a new chunk - */ - if (!meta_sinfo->full) - return 0; - +again: spin_lock(&meta_sinfo->lock); - thresh = meta_sinfo->total_bytes * 95; + if (!meta_sinfo->full) + thresh = meta_sinfo->total_bytes * 80; + else + thresh = meta_sinfo->total_bytes * 95; do_div(thresh, 100); if (meta_sinfo->bytes_used + meta_sinfo->bytes_reserved + meta_sinfo->bytes_pinned + meta_sinfo->bytes_readonly > thresh) { + struct btrfs_trans_handle *trans; + if (!meta_sinfo->full) { + meta_sinfo->force_alloc = 1; + spin_unlock(&meta_sinfo->lock); + + trans = btrfs_start_transaction(root, 1); + if (!trans) + return -ENOMEM; + + ret = do_chunk_alloc(trans, root->fs_info->extent_root, + 2 * 1024 * 1024, alloc_target, 0); + btrfs_end_transaction(trans, root); + goto again; + } spin_unlock(&meta_sinfo->lock); + + if (!committed) { + committed = 1; + trans = btrfs_join_transaction(root, 1); + if (!trans) + return -ENOMEM; + ret = btrfs_commit_transaction(trans, root); + if (ret) + return ret; + goto again; + } return -ENOSPC; } spin_unlock(&meta_sinfo->lock); @@ -2052,7 +2075,7 @@ int btrfs_check_data_free_space(struct btrfs_root *root, struct inode *inode, u64 bytes) { struct btrfs_space_info *data_sinfo; - int ret = 0; + int ret = 0, committed = 0; /* make sure bytes are sectorsize aligned */ bytes = (bytes + root->sectorsize - 1) & ~((u64)root->sectorsize - 1); @@ -2065,13 +2088,14 @@ again: data_sinfo->bytes_delalloc - data_sinfo->bytes_reserved - data_sinfo->bytes_pinned - data_sinfo->bytes_readonly - data_sinfo->bytes_may_use < bytes) { + struct btrfs_trans_handle *trans; + /* * if we don't have enough free bytes in this space then we need * to alloc a new chunk. */ if (!data_sinfo->full) { u64 alloc_target; - struct btrfs_trans_handle *trans; data_sinfo->force_alloc = 1; spin_unlock(&data_sinfo->lock); @@ -2090,6 +2114,19 @@ again: goto again; } spin_unlock(&data_sinfo->lock); + + /* commit the current transaction and try again */ + if (!committed) { + committed = 1; + trans = btrfs_join_transaction(root, 1); + if (!trans) + return -ENOMEM; + ret = btrfs_commit_transaction(trans, root); + if (ret) + return ret; + goto again; + } + printk(KERN_ERR "no space left, need %llu, %llu delalloc bytes" ", %llu bytes_used, %llu bytes_reserved, " "%llu bytes_pinned, %llu bytes_readonly, %llu may use" -- cgit v0.10.2 From 4c41bd0ec953954158f92bed5d3062645062b98e Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Mon, 16 Feb 2009 21:29:31 +0100 Subject: [JFFS2] fix mount crash caused by removed nodes At scan time we observed following scenario: node A inserted node B inserted node C inserted -> sets overlapped flag on node B node A is removed due to CRC failure -> overlapped flag on node B remains while (tn->overlapped) tn = tn_prev(tn); ==> crash, when tn_prev(B) is referenced. When the ultimate node is removed at scan time and the overlapped flag is set on the penultimate node, then nothing updates the overlapped flag of that node. The overlapped iterators blindly expect that the ultimate node does not have the overlapped flag set, which causes the scan code to crash. It would be a huge overhead to go through the node chain on node removal and fix up the overlapped flags, so detecting such a case on the fly in the overlapped iterators is a simpler and reliable solution. Cc: stable@kernel.org Signed-off-by: Thomas Gleixner Signed-off-by: David Woodhouse diff --git a/fs/jffs2/readinode.c b/fs/jffs2/readinode.c index 6ca08ad..1fc1e92 100644 --- a/fs/jffs2/readinode.c +++ b/fs/jffs2/readinode.c @@ -220,7 +220,7 @@ static int jffs2_add_tn_to_tree(struct jffs2_sb_info *c, struct jffs2_tmp_dnode_info *tn) { uint32_t fn_end = tn->fn->ofs + tn->fn->size; - struct jffs2_tmp_dnode_info *this; + struct jffs2_tmp_dnode_info *this, *ptn; dbg_readinode("insert fragment %#04x-%#04x, ver %u at %08x\n", tn->fn->ofs, fn_end, tn->version, ref_offset(tn->fn->raw)); @@ -251,11 +251,18 @@ static int jffs2_add_tn_to_tree(struct jffs2_sb_info *c, if (this) { /* If the node is coincident with another at a lower address, back up until the other node is found. It may be relevant */ - while (this->overlapped) - this = tn_prev(this); - - /* First node should never be marked overlapped */ - BUG_ON(!this); + while (this->overlapped) { + ptn = tn_prev(this); + if (!ptn) { + /* + * We killed a node which set the overlapped + * flags during the scan. Fix it up. + */ + this->overlapped = 0; + break; + } + this = ptn; + } dbg_readinode("'this' found %#04x-%#04x (%s)\n", this->fn->ofs, this->fn->ofs + this->fn->size, this->fn ? "data" : "hole"); } @@ -360,7 +367,17 @@ static int jffs2_add_tn_to_tree(struct jffs2_sb_info *c, } if (!this->overlapped) break; - this = tn_prev(this); + + ptn = tn_prev(this); + if (!ptn) { + /* + * We killed a node which set the overlapped + * flags during the scan. Fix it up. + */ + this->overlapped = 0; + break; + } + this = ptn; } } @@ -456,8 +473,15 @@ static int jffs2_build_inode_fragtree(struct jffs2_sb_info *c, eat_last(&rii->tn_root, &last->rb); ver_insert(&ver_root, last); - if (unlikely(last->overlapped)) - continue; + if (unlikely(last->overlapped)) { + if (pen) + continue; + /* + * We killed a node which set the overlapped + * flags during the scan. Fix it up. + */ + last->overlapped = 0; + } /* Now we have a bunch of nodes in reverse version order, in the tree at ver_root. Most of the time, -- cgit v0.10.2 From b73a77494292b930642fbf87de3e3196593f7593 Mon Sep 17 00:00:00 2001 From: HighPoint Linux Team Date: Thu, 12 Feb 2009 11:28:31 +0800 Subject: [SCSI] hptiop: Add new PCI device ID Signed-off-by: HighPoint Linux Team Signed-off-by: James Bottomley diff --git a/drivers/scsi/hptiop.c b/drivers/scsi/hptiop.c index a48e499..34be88d 100644 --- a/drivers/scsi/hptiop.c +++ b/drivers/scsi/hptiop.c @@ -1251,6 +1251,7 @@ static struct pci_device_id hptiop_id_table[] = { { PCI_VDEVICE(TTI, 0x3530), (kernel_ulong_t)&hptiop_itl_ops }, { PCI_VDEVICE(TTI, 0x3560), (kernel_ulong_t)&hptiop_itl_ops }, { PCI_VDEVICE(TTI, 0x4322), (kernel_ulong_t)&hptiop_itl_ops }, + { PCI_VDEVICE(TTI, 0x4321), (kernel_ulong_t)&hptiop_itl_ops }, { PCI_VDEVICE(TTI, 0x4210), (kernel_ulong_t)&hptiop_itl_ops }, { PCI_VDEVICE(TTI, 0x4211), (kernel_ulong_t)&hptiop_itl_ops }, { PCI_VDEVICE(TTI, 0x4310), (kernel_ulong_t)&hptiop_itl_ops }, -- cgit v0.10.2 From 1648b11ea7cec5b95e5a71364ac1f40bfef702d0 Mon Sep 17 00:00:00 2001 From: Karen Xie Date: Fri, 13 Feb 2009 21:38:44 -0800 Subject: [SCSI] cxgb3i: transmit work-request fixes - resize the work-request credit array to be based on skb's MAX_SKB_FRAGS. - split the skb cb into tx and rx portion - increase the default transmit window to 128K. - stop queueing up the outgoing pdus if transmit window is full. Signed-off-by: Karen Xie Reviewed-by: Mike Christie Signed-off-by: James Bottomley diff --git a/drivers/scsi/cxgb3i/cxgb3i_offload.c b/drivers/scsi/cxgb3i/cxgb3i_offload.c index a865f1f..de3b3b6 100644 --- a/drivers/scsi/cxgb3i/cxgb3i_offload.c +++ b/drivers/scsi/cxgb3i/cxgb3i_offload.c @@ -23,19 +23,19 @@ #include "cxgb3i_ddp.h" #ifdef __DEBUG_C3CN_CONN__ -#define c3cn_conn_debug cxgb3i_log_info +#define c3cn_conn_debug cxgb3i_log_debug #else #define c3cn_conn_debug(fmt...) #endif #ifdef __DEBUG_C3CN_TX__ -#define c3cn_tx_debug cxgb3i_log_debug +#define c3cn_tx_debug cxgb3i_log_debug #else #define c3cn_tx_debug(fmt...) #endif #ifdef __DEBUG_C3CN_RX__ -#define c3cn_rx_debug cxgb3i_log_debug +#define c3cn_rx_debug cxgb3i_log_debug #else #define c3cn_rx_debug(fmt...) #endif @@ -47,9 +47,9 @@ static int cxgb3_rcv_win = 256 * 1024; module_param(cxgb3_rcv_win, int, 0644); MODULE_PARM_DESC(cxgb3_rcv_win, "TCP receive window in bytes (default=256KB)"); -static int cxgb3_snd_win = 64 * 1024; +static int cxgb3_snd_win = 128 * 1024; module_param(cxgb3_snd_win, int, 0644); -MODULE_PARM_DESC(cxgb3_snd_win, "TCP send window in bytes (default=64KB)"); +MODULE_PARM_DESC(cxgb3_snd_win, "TCP send window in bytes (default=128KB)"); static int cxgb3_rx_credit_thres = 10 * 1024; module_param(cxgb3_rx_credit_thres, int, 0644); @@ -301,8 +301,8 @@ static void act_open_req_arp_failure(struct t3cdev *dev, struct sk_buff *skb) static void skb_entail(struct s3_conn *c3cn, struct sk_buff *skb, int flags) { - CXGB3_SKB_CB(skb)->seq = c3cn->write_seq; - CXGB3_SKB_CB(skb)->flags = flags; + skb_tcp_seq(skb) = c3cn->write_seq; + skb_flags(skb) = flags; __skb_queue_tail(&c3cn->write_queue, skb); } @@ -457,12 +457,9 @@ static unsigned int wrlen __read_mostly; * The number of WRs needed for an skb depends on the number of fragments * in the skb and whether it has any payload in its main body. This maps the * length of the gather list represented by an skb into the # of necessary WRs. - * - * The max. length of an skb is controlled by the max pdu size which is ~16K. - * Also, assume the min. fragment length is the sector size (512), then add - * extra fragment counts for iscsi bhs and payload padding. + * The extra two fragments are for iscsi bhs and payload padding. */ -#define SKB_WR_LIST_SIZE (16384/512 + 3) +#define SKB_WR_LIST_SIZE (MAX_SKB_FRAGS + 2) static unsigned int skb_wrs[SKB_WR_LIST_SIZE] __read_mostly; static void s3_init_wr_tab(unsigned int wr_len) @@ -485,7 +482,7 @@ static void s3_init_wr_tab(unsigned int wr_len) static inline void reset_wr_list(struct s3_conn *c3cn) { - c3cn->wr_pending_head = NULL; + c3cn->wr_pending_head = c3cn->wr_pending_tail = NULL; } /* @@ -496,7 +493,7 @@ static inline void reset_wr_list(struct s3_conn *c3cn) static inline void enqueue_wr(struct s3_conn *c3cn, struct sk_buff *skb) { - skb_wr_data(skb) = NULL; + skb_tx_wr_next(skb) = NULL; /* * We want to take an extra reference since both us and the driver @@ -509,10 +506,22 @@ static inline void enqueue_wr(struct s3_conn *c3cn, if (!c3cn->wr_pending_head) c3cn->wr_pending_head = skb; else - skb_wr_data(skb) = skb; + skb_tx_wr_next(c3cn->wr_pending_tail) = skb; c3cn->wr_pending_tail = skb; } +static int count_pending_wrs(struct s3_conn *c3cn) +{ + int n = 0; + const struct sk_buff *skb = c3cn->wr_pending_head; + + while (skb) { + n += skb->csum; + skb = skb_tx_wr_next(skb); + } + return n; +} + static inline struct sk_buff *peek_wr(const struct s3_conn *c3cn) { return c3cn->wr_pending_head; @@ -529,8 +538,8 @@ static inline struct sk_buff *dequeue_wr(struct s3_conn *c3cn) if (likely(skb)) { /* Don't bother clearing the tail */ - c3cn->wr_pending_head = skb_wr_data(skb); - skb_wr_data(skb) = NULL; + c3cn->wr_pending_head = skb_tx_wr_next(skb); + skb_tx_wr_next(skb) = NULL; } return skb; } @@ -543,13 +552,14 @@ static void purge_wr_queue(struct s3_conn *c3cn) } static inline void make_tx_data_wr(struct s3_conn *c3cn, struct sk_buff *skb, - int len) + int len, int req_completion) { struct tx_data_wr *req; skb_reset_transport_header(skb); req = (struct tx_data_wr *)__skb_push(skb, sizeof(*req)); - req->wr_hi = htonl(V_WR_OP(FW_WROPCODE_OFLD_TX_DATA)); + req->wr_hi = htonl(V_WR_OP(FW_WROPCODE_OFLD_TX_DATA) | + (req_completion ? F_WR_COMPL : 0)); req->wr_lo = htonl(V_WR_TID(c3cn->tid)); req->sndseq = htonl(c3cn->snd_nxt); /* len includes the length of any HW ULP additions */ @@ -592,7 +602,7 @@ static int c3cn_push_tx_frames(struct s3_conn *c3cn, int req_completion) if (unlikely(c3cn->state == C3CN_STATE_CONNECTING || c3cn->state == C3CN_STATE_CLOSE_WAIT_1 || - c3cn->state == C3CN_STATE_ABORTING)) { + c3cn->state >= C3CN_STATE_ABORTING)) { c3cn_tx_debug("c3cn 0x%p, in closing state %u.\n", c3cn, c3cn->state); return 0; @@ -615,7 +625,7 @@ static int c3cn_push_tx_frames(struct s3_conn *c3cn, int req_completion) if (c3cn->wr_avail < wrs_needed) { c3cn_tx_debug("c3cn 0x%p, skb len %u/%u, frag %u, " "wr %d < %u.\n", - c3cn, skb->len, skb->datalen, frags, + c3cn, skb->len, skb->data_len, frags, wrs_needed, c3cn->wr_avail); break; } @@ -627,20 +637,24 @@ static int c3cn_push_tx_frames(struct s3_conn *c3cn, int req_completion) c3cn->wr_unacked += wrs_needed; enqueue_wr(c3cn, skb); - if (likely(CXGB3_SKB_CB(skb)->flags & C3CB_FLAG_NEED_HDR)) { - len += ulp_extra_len(skb); - make_tx_data_wr(c3cn, skb, len); - c3cn->snd_nxt += len; - if ((req_completion - && c3cn->wr_unacked == wrs_needed) - || (CXGB3_SKB_CB(skb)->flags & C3CB_FLAG_COMPL) - || c3cn->wr_unacked >= c3cn->wr_max / 2) { - struct work_request_hdr *wr = cplhdr(skb); + c3cn_tx_debug("c3cn 0x%p, enqueue, skb len %u/%u, frag %u, " + "wr %d, left %u, unack %u.\n", + c3cn, skb->len, skb->data_len, frags, + wrs_needed, c3cn->wr_avail, c3cn->wr_unacked); + - wr->wr_hi |= htonl(F_WR_COMPL); + if (likely(skb_flags(skb) & C3CB_FLAG_NEED_HDR)) { + if ((req_completion && + c3cn->wr_unacked == wrs_needed) || + (skb_flags(skb) & C3CB_FLAG_COMPL) || + c3cn->wr_unacked >= c3cn->wr_max / 2) { + req_completion = 1; c3cn->wr_unacked = 0; } - CXGB3_SKB_CB(skb)->flags &= ~C3CB_FLAG_NEED_HDR; + len += ulp_extra_len(skb); + make_tx_data_wr(c3cn, skb, len, req_completion); + c3cn->snd_nxt += len; + skb_flags(skb) &= ~C3CB_FLAG_NEED_HDR; } total_size += skb->truesize; @@ -735,8 +749,11 @@ static void process_act_establish(struct s3_conn *c3cn, struct sk_buff *skb) if (unlikely(c3cn_flag(c3cn, C3CN_ACTIVE_CLOSE_NEEDED))) /* upper layer has requested closing */ send_abort_req(c3cn); - else if (c3cn_push_tx_frames(c3cn, 1)) + else { + if (skb_queue_len(&c3cn->write_queue)) + c3cn_push_tx_frames(c3cn, 1); cxgb3i_conn_tx_open(c3cn); + } } static int do_act_establish(struct t3cdev *cdev, struct sk_buff *skb, @@ -1082,8 +1099,8 @@ static void process_rx_iscsi_hdr(struct s3_conn *c3cn, struct sk_buff *skb) return; } - CXGB3_SKB_CB(skb)->seq = ntohl(hdr_cpl->seq); - CXGB3_SKB_CB(skb)->flags = 0; + skb_tcp_seq(skb) = ntohl(hdr_cpl->seq); + skb_flags(skb) = 0; skb_reset_transport_header(skb); __skb_pull(skb, sizeof(struct cpl_iscsi_hdr)); @@ -1103,12 +1120,12 @@ static void process_rx_iscsi_hdr(struct s3_conn *c3cn, struct sk_buff *skb) goto abort_conn; skb_ulp_mode(skb) = ULP2_FLAG_DATA_READY; - skb_ulp_pdulen(skb) = ntohs(ddp_cpl.len); - skb_ulp_ddigest(skb) = ntohl(ddp_cpl.ulp_crc); + skb_rx_pdulen(skb) = ntohs(ddp_cpl.len); + skb_rx_ddigest(skb) = ntohl(ddp_cpl.ulp_crc); status = ntohl(ddp_cpl.ddp_status); c3cn_rx_debug("rx skb 0x%p, len %u, pdulen %u, ddp status 0x%x.\n", - skb, skb->len, skb_ulp_pdulen(skb), status); + skb, skb->len, skb_rx_pdulen(skb), status); if (status & (1 << RX_DDP_STATUS_HCRC_SHIFT)) skb_ulp_mode(skb) |= ULP2_FLAG_HCRC_ERROR; @@ -1126,7 +1143,7 @@ static void process_rx_iscsi_hdr(struct s3_conn *c3cn, struct sk_buff *skb) } else if (status & (1 << RX_DDP_STATUS_DDP_SHIFT)) skb_ulp_mode(skb) |= ULP2_FLAG_DATA_DDPED; - c3cn->rcv_nxt = ntohl(ddp_cpl.seq) + skb_ulp_pdulen(skb); + c3cn->rcv_nxt = ntohl(ddp_cpl.seq) + skb_rx_pdulen(skb); __pskb_trim(skb, len); __skb_queue_tail(&c3cn->receive_queue, skb); cxgb3i_conn_pdu_ready(c3cn); @@ -1151,12 +1168,27 @@ static int do_iscsi_hdr(struct t3cdev *t3dev, struct sk_buff *skb, void *ctx) * Process an acknowledgment of WR completion. Advance snd_una and send the * next batch of work requests from the write queue. */ +static void check_wr_invariants(struct s3_conn *c3cn) +{ + int pending = count_pending_wrs(c3cn); + + if (unlikely(c3cn->wr_avail + pending != c3cn->wr_max)) + cxgb3i_log_error("TID %u: credit imbalance: avail %u, " + "pending %u, total should be %u\n", + c3cn->tid, c3cn->wr_avail, pending, + c3cn->wr_max); +} + static void process_wr_ack(struct s3_conn *c3cn, struct sk_buff *skb) { struct cpl_wr_ack *hdr = cplhdr(skb); unsigned int credits = ntohs(hdr->credits); u32 snd_una = ntohl(hdr->snd_una); + c3cn_tx_debug("%u WR credits, avail %u, unack %u, TID %u, state %u.\n", + credits, c3cn->wr_avail, c3cn->wr_unacked, + c3cn->tid, c3cn->state); + c3cn->wr_avail += credits; if (c3cn->wr_unacked > c3cn->wr_max - c3cn->wr_avail) c3cn->wr_unacked = c3cn->wr_max - c3cn->wr_avail; @@ -1171,6 +1203,17 @@ static void process_wr_ack(struct s3_conn *c3cn, struct sk_buff *skb) break; } if (unlikely(credits < p->csum)) { + struct tx_data_wr *w = cplhdr(p); + cxgb3i_log_error("TID %u got %u WR credits need %u, " + "len %u, main body %u, frags %u, " + "seq # %u, ACK una %u, ACK nxt %u, " + "WR_AVAIL %u, WRs pending %u\n", + c3cn->tid, credits, p->csum, p->len, + p->len - p->data_len, + skb_shinfo(p)->nr_frags, + ntohl(w->sndseq), snd_una, + ntohl(hdr->snd_nxt), c3cn->wr_avail, + count_pending_wrs(c3cn) - credits); p->csum -= credits; break; } else { @@ -1180,15 +1223,24 @@ static void process_wr_ack(struct s3_conn *c3cn, struct sk_buff *skb) } } - if (unlikely(before(snd_una, c3cn->snd_una))) + check_wr_invariants(c3cn); + + if (unlikely(before(snd_una, c3cn->snd_una))) { + cxgb3i_log_error("TID %u, unexpected sequence # %u in WR_ACK " + "snd_una %u\n", + c3cn->tid, snd_una, c3cn->snd_una); goto out_free; + } if (c3cn->snd_una != snd_una) { c3cn->snd_una = snd_una; dst_confirm(c3cn->dst_cache); } - if (skb_queue_len(&c3cn->write_queue) && c3cn_push_tx_frames(c3cn, 0)) + if (skb_queue_len(&c3cn->write_queue)) { + if (c3cn_push_tx_frames(c3cn, 0)) + cxgb3i_conn_tx_open(c3cn); + } else cxgb3i_conn_tx_open(c3cn); out_free: __kfree_skb(skb); @@ -1452,7 +1504,7 @@ static void init_offload_conn(struct s3_conn *c3cn, struct dst_entry *dst) { BUG_ON(c3cn->cdev != cdev); - c3cn->wr_max = c3cn->wr_avail = T3C_DATA(cdev)->max_wrs; + c3cn->wr_max = c3cn->wr_avail = T3C_DATA(cdev)->max_wrs - 1; c3cn->wr_unacked = 0; c3cn->mss_idx = select_mss(c3cn, dst_mtu(dst)); @@ -1671,9 +1723,17 @@ int cxgb3i_c3cn_send_pdus(struct s3_conn *c3cn, struct sk_buff *skb) goto out_err; } - err = -EPIPE; if (c3cn->err) { c3cn_tx_debug("c3cn 0x%p, err %d.\n", c3cn, c3cn->err); + err = -EPIPE; + goto out_err; + } + + if (c3cn->write_seq - c3cn->snd_una >= cxgb3_snd_win) { + c3cn_tx_debug("c3cn 0x%p, snd %u - %u > %u.\n", + c3cn, c3cn->write_seq, c3cn->snd_una, + cxgb3_snd_win); + err = -EAGAIN; goto out_err; } diff --git a/drivers/scsi/cxgb3i/cxgb3i_offload.h b/drivers/scsi/cxgb3i/cxgb3i_offload.h index d231569..df1eae0 100644 --- a/drivers/scsi/cxgb3i/cxgb3i_offload.h +++ b/drivers/scsi/cxgb3i/cxgb3i_offload.h @@ -178,25 +178,33 @@ void cxgb3i_c3cn_release(struct s3_conn *); * @flag: see C3CB_FLAG_* below * @ulp_mode: ULP mode/submode of sk_buff * @seq: tcp sequence number - * @ddigest: pdu data digest - * @pdulen: recovered pdu length - * @wr_data: scratch area for tx wr */ +struct cxgb3_skb_rx_cb { + __u32 ddigest; /* data digest */ + __u32 pdulen; /* recovered pdu length */ +}; + +struct cxgb3_skb_tx_cb { + struct sk_buff *wr_next; /* next wr */ +}; + struct cxgb3_skb_cb { __u8 flags; __u8 ulp_mode; __u32 seq; - __u32 ddigest; - __u32 pdulen; - struct sk_buff *wr_data; + union { + struct cxgb3_skb_rx_cb rx; + struct cxgb3_skb_tx_cb tx; + }; }; #define CXGB3_SKB_CB(skb) ((struct cxgb3_skb_cb *)&((skb)->cb[0])) - +#define skb_flags(skb) (CXGB3_SKB_CB(skb)->flags) #define skb_ulp_mode(skb) (CXGB3_SKB_CB(skb)->ulp_mode) -#define skb_ulp_ddigest(skb) (CXGB3_SKB_CB(skb)->ddigest) -#define skb_ulp_pdulen(skb) (CXGB3_SKB_CB(skb)->pdulen) -#define skb_wr_data(skb) (CXGB3_SKB_CB(skb)->wr_data) +#define skb_tcp_seq(skb) (CXGB3_SKB_CB(skb)->seq) +#define skb_rx_ddigest(skb) (CXGB3_SKB_CB(skb)->rx.ddigest) +#define skb_rx_pdulen(skb) (CXGB3_SKB_CB(skb)->rx.pdulen) +#define skb_tx_wr_next(skb) (CXGB3_SKB_CB(skb)->tx.wr_next) enum c3cb_flags { C3CB_FLAG_NEED_HDR = 1 << 0, /* packet needs a TX_DATA_WR header */ -- cgit v0.10.2 From 949847d195d2bb86f61c289a57edb9207c4a3bbf Mon Sep 17 00:00:00 2001 From: Karen Xie Date: Fri, 13 Feb 2009 21:38:49 -0800 Subject: [SCSI] cxgb3i: added per-task data to track transmit progress added per-task struct cxgb3i_task_data to track the data transmiting progress and the state of the pdus to be transmitted. Signed-off-by: Karen Xie Reviewed-by: Mike Christie Signed-off-by: James Bottomley diff --git a/drivers/scsi/cxgb3i/cxgb3i.h b/drivers/scsi/cxgb3i/cxgb3i.h index fde6e4c..a7cf550 100644 --- a/drivers/scsi/cxgb3i/cxgb3i.h +++ b/drivers/scsi/cxgb3i/cxgb3i.h @@ -20,6 +20,7 @@ #include #include #include +#include #include /* from cxgb3 LLD */ @@ -113,6 +114,26 @@ struct cxgb3i_endpoint { struct cxgb3i_conn *cconn; }; +/** + * struct cxgb3i_task_data - private iscsi task data + * + * @nr_frags: # of coalesced page frags (from scsi sgl) + * @frags: coalesced page frags (from scsi sgl) + * @skb: tx pdu skb + * @offset: data offset for the next pdu + * @count: max. possible pdu payload + * @sgoffset: offset to the first sg entry for a given offset + */ +#define MAX_PDU_FRAGS ((ULP2_MAX_PDU_PAYLOAD + 512 - 1) / 512) +struct cxgb3i_task_data { + unsigned short nr_frags; + skb_frag_t frags[MAX_PDU_FRAGS]; + struct sk_buff *skb; + unsigned int offset; + unsigned int count; + unsigned int sgoffset; +}; + int cxgb3i_iscsi_init(void); void cxgb3i_iscsi_cleanup(void); diff --git a/drivers/scsi/cxgb3i/cxgb3i_iscsi.c b/drivers/scsi/cxgb3i/cxgb3i_iscsi.c index d83464b..f0434b7 100644 --- a/drivers/scsi/cxgb3i/cxgb3i_iscsi.c +++ b/drivers/scsi/cxgb3i/cxgb3i_iscsi.c @@ -364,7 +364,8 @@ cxgb3i_session_create(struct iscsi_endpoint *ep, u16 cmds_max, u16 qdepth, cls_session = iscsi_session_setup(&cxgb3i_iscsi_transport, shost, cmds_max, - sizeof(struct iscsi_tcp_task), + sizeof(struct iscsi_tcp_task) + + sizeof(struct cxgb3i_task_data), initial_cmdsn, ISCSI_MAX_TARGET); if (!cls_session) return NULL; @@ -844,7 +845,7 @@ static struct scsi_host_template cxgb3i_host_template = { .proc_name = "cxgb3i", .queuecommand = iscsi_queuecommand, .change_queue_depth = iscsi_change_queue_depth, - .can_queue = 128 * (ISCSI_DEF_XMIT_CMDS_MAX - 1), + .can_queue = CXGB3I_SCSI_QDEPTH_DFLT - 1, .sg_tablesize = SG_ALL, .max_sectors = 0xFFFF, .cmd_per_lun = ISCSI_DEF_CMD_PER_LUN, -- cgit v0.10.2 From f62d0896e67195d4407ef81c6f77a92f72a63e88 Mon Sep 17 00:00:00 2001 From: Karen Xie Date: Fri, 13 Feb 2009 21:38:54 -0800 Subject: [SCSI] cxgb3i: Outgoing pdus need to observe skb's MAX_SKB_FRAGS Need to make sure the outgoing pdu can fit into a single skb. When calulating the max. outgoing pdu payload size, take into consideration of - data can be held in the skb's fragment list, assume 512 bytes per fragment, and - data can be held in the headroom. Signed-off-by: Karen Xie Reviewed-by: Mike Christie Signed-off-by: James Bottomley diff --git a/drivers/scsi/cxgb3i/cxgb3i_ddp.c b/drivers/scsi/cxgb3i/cxgb3i_ddp.c index 08f3a09..a83d36e 100644 --- a/drivers/scsi/cxgb3i/cxgb3i_ddp.c +++ b/drivers/scsi/cxgb3i/cxgb3i_ddp.c @@ -639,10 +639,11 @@ static int ddp_init(struct t3cdev *tdev) write_unlock(&cxgb3i_ddp_rwlock); ddp_log_info("nppods %u (0x%x ~ 0x%x), bits %u, mask 0x%x,0x%x " - "pkt %u,%u.\n", + "pkt %u/%u, %u/%u.\n", ppmax, ddp->llimit, ddp->ulimit, ddp->idx_bits, ddp->idx_mask, ddp->rsvd_tag_mask, - ddp->max_txsz, ddp->max_rxsz); + ddp->max_txsz, uinfo.max_txsz, + ddp->max_rxsz, uinfo.max_rxsz); return 0; free_ddp_map: @@ -654,8 +655,8 @@ free_ddp_map: * cxgb3i_adapter_ddp_init - initialize the adapter's ddp resource * @tdev: t3cdev adapter * @tformat: tag format - * @txsz: max tx pkt size, filled in by this func. - * @rxsz: max rx pkt size, filled in by this func. + * @txsz: max tx pdu payload size, filled in by this func. + * @rxsz: max rx pdu payload size, filled in by this func. * initialize the ddp pagepod manager for a given adapter if needed and * setup the tag format for a given iscsi entity */ @@ -685,10 +686,12 @@ int cxgb3i_adapter_ddp_init(struct t3cdev *tdev, tformat->sw_bits, tformat->rsvd_bits, tformat->rsvd_shift, tformat->rsvd_mask); - *txsz = ddp->max_txsz; - *rxsz = ddp->max_rxsz; - ddp_log_info("ddp max pkt size: %u, %u.\n", - ddp->max_txsz, ddp->max_rxsz); + *txsz = min_t(unsigned int, ULP2_MAX_PDU_PAYLOAD, + ddp->max_txsz - ISCSI_PDU_NONPAYLOAD_LEN); + *rxsz = min_t(unsigned int, ULP2_MAX_PDU_PAYLOAD, + ddp->max_rxsz - ISCSI_PDU_NONPAYLOAD_LEN); + ddp_log_info("max payload size: %u/%u, %u/%u.\n", + *txsz, ddp->max_txsz, *rxsz, ddp->max_rxsz); return 0; } EXPORT_SYMBOL_GPL(cxgb3i_adapter_ddp_init); diff --git a/drivers/scsi/cxgb3i/cxgb3i_ddp.h b/drivers/scsi/cxgb3i/cxgb3i_ddp.h index 5c7c4d9..99d7014 100644 --- a/drivers/scsi/cxgb3i/cxgb3i_ddp.h +++ b/drivers/scsi/cxgb3i/cxgb3i_ddp.h @@ -85,8 +85,9 @@ struct cxgb3i_ddp_info { struct sk_buff **gl_skb; }; +#define ISCSI_PDU_NONPAYLOAD_LEN 312 /* bhs(48) + ahs(256) + digest(8) */ #define ULP2_MAX_PKT_SIZE 16224 -#define ULP2_MAX_PDU_PAYLOAD (ULP2_MAX_PKT_SIZE - ISCSI_PDU_NONPAYLOAD_MAX) +#define ULP2_MAX_PDU_PAYLOAD (ULP2_MAX_PKT_SIZE - ISCSI_PDU_NONPAYLOAD_LEN) #define PPOD_PAGES_MAX 4 #define PPOD_PAGES_SHIFT 2 /* 4 pages per pod */ diff --git a/drivers/scsi/cxgb3i/cxgb3i_iscsi.c b/drivers/scsi/cxgb3i/cxgb3i_iscsi.c index f0434b7..fa2a44f 100644 --- a/drivers/scsi/cxgb3i/cxgb3i_iscsi.c +++ b/drivers/scsi/cxgb3i/cxgb3i_iscsi.c @@ -403,17 +403,15 @@ static inline int cxgb3i_conn_max_xmit_dlength(struct iscsi_conn *conn) { struct iscsi_tcp_conn *tcp_conn = conn->dd_data; struct cxgb3i_conn *cconn = tcp_conn->dd_data; - unsigned int max = min_t(unsigned int, ULP2_MAX_PDU_PAYLOAD, - cconn->hba->snic->tx_max_size - - ISCSI_PDU_NONPAYLOAD_MAX); + unsigned int max = max(512 * MAX_SKB_FRAGS, SKB_TX_HEADROOM); + max = min(cconn->hba->snic->tx_max_size, max); if (conn->max_xmit_dlength) - conn->max_xmit_dlength = min_t(unsigned int, - conn->max_xmit_dlength, max); + conn->max_xmit_dlength = min(conn->max_xmit_dlength, max); else conn->max_xmit_dlength = max; align_pdu_size(conn->max_xmit_dlength); - cxgb3i_log_info("conn 0x%p, max xmit %u.\n", + cxgb3i_api_debug("conn 0x%p, max xmit %u.\n", conn, conn->max_xmit_dlength); return 0; } @@ -428,9 +426,7 @@ static inline int cxgb3i_conn_max_recv_dlength(struct iscsi_conn *conn) { struct iscsi_tcp_conn *tcp_conn = conn->dd_data; struct cxgb3i_conn *cconn = tcp_conn->dd_data; - unsigned int max = min_t(unsigned int, ULP2_MAX_PDU_PAYLOAD, - cconn->hba->snic->rx_max_size - - ISCSI_PDU_NONPAYLOAD_MAX); + unsigned int max = cconn->hba->snic->rx_max_size; align_pdu_size(max); if (conn->max_recv_dlength) { @@ -440,8 +436,7 @@ static inline int cxgb3i_conn_max_recv_dlength(struct iscsi_conn *conn) conn->max_recv_dlength, max); return -EINVAL; } - conn->max_recv_dlength = min_t(unsigned int, - conn->max_recv_dlength, max); + conn->max_recv_dlength = min(conn->max_recv_dlength, max); align_pdu_size(conn->max_recv_dlength); } else conn->max_recv_dlength = max; diff --git a/drivers/scsi/cxgb3i/cxgb3i_offload.h b/drivers/scsi/cxgb3i/cxgb3i_offload.h index df1eae0..6344b9e 100644 --- a/drivers/scsi/cxgb3i/cxgb3i_offload.h +++ b/drivers/scsi/cxgb3i/cxgb3i_offload.h @@ -225,6 +225,7 @@ struct sge_opaque_hdr { /* for TX: a skb must have a headroom of at least TX_HEADER_LEN bytes */ #define TX_HEADER_LEN \ (sizeof(struct tx_data_wr) + sizeof(struct sge_opaque_hdr)) +#define SKB_TX_HEADROOM SKB_MAX_HEAD(TX_HEADER_LEN) /* * get and set private ip for iscsi traffic diff --git a/drivers/scsi/cxgb3i/cxgb3i_pdu.c b/drivers/scsi/cxgb3i/cxgb3i_pdu.c index ce7ce8c..17115c2 100644 --- a/drivers/scsi/cxgb3i/cxgb3i_pdu.c +++ b/drivers/scsi/cxgb3i/cxgb3i_pdu.c @@ -32,6 +32,10 @@ #define cxgb3i_tx_debug(fmt...) #endif +/* always allocate rooms for AHS */ +#define SKB_TX_PDU_HEADER_LEN \ + (sizeof(struct iscsi_hdr) + ISCSI_MAX_AHS_SIZE) +static unsigned int skb_extra_headroom; static struct page *pad_page; /* @@ -146,12 +150,13 @@ static inline void tx_skb_setmode(struct sk_buff *skb, int hcrc, int dcrc) void cxgb3i_conn_cleanup_task(struct iscsi_task *task) { - struct iscsi_tcp_task *tcp_task = task->dd_data; + struct cxgb3i_task_data *tdata = task->dd_data + + sizeof(struct iscsi_tcp_task); /* never reached the xmit task callout */ - if (tcp_task->dd_data) - kfree_skb(tcp_task->dd_data); - tcp_task->dd_data = NULL; + if (tdata->skb) + __kfree_skb(tdata->skb); + memset(tdata, 0, sizeof(struct cxgb3i_task_data)); /* MNC - Do we need a check in case this is called but * cxgb3i_conn_alloc_pdu has never been called on the task */ @@ -159,28 +164,102 @@ void cxgb3i_conn_cleanup_task(struct iscsi_task *task) iscsi_tcp_cleanup_task(task); } -/* - * We do not support ahs yet - */ +static int sgl_seek_offset(struct scatterlist *sgl, unsigned int sgcnt, + unsigned int offset, unsigned int *off, + struct scatterlist **sgp) +{ + int i; + struct scatterlist *sg; + + for_each_sg(sgl, sg, sgcnt, i) { + if (offset < sg->length) { + *off = offset; + *sgp = sg; + return 0; + } + offset -= sg->length; + } + return -EFAULT; +} + +static int sgl_read_to_frags(struct scatterlist *sg, unsigned int sgoffset, + unsigned int dlen, skb_frag_t *frags, + int frag_max) +{ + unsigned int datalen = dlen; + unsigned int sglen = sg->length - sgoffset; + struct page *page = sg_page(sg); + int i; + + i = 0; + do { + unsigned int copy; + + if (!sglen) { + sg = sg_next(sg); + if (!sg) { + cxgb3i_log_error("%s, sg NULL, len %u/%u.\n", + __func__, datalen, dlen); + return -EINVAL; + } + sgoffset = 0; + sglen = sg->length; + page = sg_page(sg); + + } + copy = min(datalen, sglen); + if (i && page == frags[i - 1].page && + sgoffset + sg->offset == + frags[i - 1].page_offset + frags[i - 1].size) { + frags[i - 1].size += copy; + } else { + if (i >= frag_max) { + cxgb3i_log_error("%s, too many pages %u, " + "dlen %u.\n", __func__, + frag_max, dlen); + return -EINVAL; + } + + frags[i].page = page; + frags[i].page_offset = sg->offset + sgoffset; + frags[i].size = copy; + i++; + } + datalen -= copy; + sgoffset += copy; + sglen -= copy; + } while (datalen); + + return i; +} + int cxgb3i_conn_alloc_pdu(struct iscsi_task *task, u8 opcode) { + struct iscsi_conn *conn = task->conn; struct iscsi_tcp_task *tcp_task = task->dd_data; - struct sk_buff *skb; + struct cxgb3i_task_data *tdata = task->dd_data + sizeof(*tcp_task); + struct scsi_cmnd *sc = task->sc; + int headroom = SKB_TX_PDU_HEADER_LEN; + tcp_task->dd_data = tdata; task->hdr = NULL; - /* always allocate rooms for AHS */ - skb = alloc_skb(sizeof(struct iscsi_hdr) + ISCSI_MAX_AHS_SIZE + - TX_HEADER_LEN, GFP_ATOMIC); - if (!skb) + + /* write command, need to send data pdus */ + if (skb_extra_headroom && (opcode == ISCSI_OP_SCSI_DATA_OUT || + (opcode == ISCSI_OP_SCSI_CMD && + (scsi_bidi_cmnd(sc) || sc->sc_data_direction == DMA_TO_DEVICE)))) + headroom += min(skb_extra_headroom, conn->max_xmit_dlength); + + tdata->skb = alloc_skb(TX_HEADER_LEN + headroom, GFP_ATOMIC); + if (!tdata->skb) return -ENOMEM; + skb_reserve(tdata->skb, TX_HEADER_LEN); cxgb3i_tx_debug("task 0x%p, opcode 0x%x, skb 0x%p.\n", - task, opcode, skb); + task, opcode, tdata->skb); - tcp_task->dd_data = skb; - skb_reserve(skb, TX_HEADER_LEN); - task->hdr = (struct iscsi_hdr *)skb->data; - task->hdr_max = sizeof(struct iscsi_hdr); + task->hdr = (struct iscsi_hdr *)tdata->skb->data; + task->hdr_max = SKB_TX_PDU_HEADER_LEN; /* data_out uses scsi_cmd's itt */ if (opcode != ISCSI_OP_SCSI_DATA_OUT) @@ -192,13 +271,13 @@ int cxgb3i_conn_alloc_pdu(struct iscsi_task *task, u8 opcode) int cxgb3i_conn_init_pdu(struct iscsi_task *task, unsigned int offset, unsigned int count) { - struct iscsi_tcp_task *tcp_task = task->dd_data; - struct sk_buff *skb = tcp_task->dd_data; struct iscsi_conn *conn = task->conn; - struct page *pg; + struct iscsi_tcp_task *tcp_task = task->dd_data; + struct cxgb3i_task_data *tdata = tcp_task->dd_data; + struct sk_buff *skb = tdata->skb; unsigned int datalen = count; int i, padlen = iscsi_padding(count); - skb_frag_t *frag; + struct page *pg; cxgb3i_tx_debug("task 0x%p,0x%p, offset %u, count %u, skb 0x%p.\n", task, task->sc, offset, count, skb); @@ -209,90 +288,94 @@ int cxgb3i_conn_init_pdu(struct iscsi_task *task, unsigned int offset, return 0; if (task->sc) { - struct scatterlist *sg; - struct scsi_data_buffer *sdb; - unsigned int sgoffset = offset; - struct page *sgpg; - unsigned int sglen; - - sdb = scsi_out(task->sc); - sg = sdb->table.sgl; - - for_each_sg(sdb->table.sgl, sg, sdb->table.nents, i) { - cxgb3i_tx_debug("sg %d, page 0x%p, len %u offset %u\n", - i, sg_page(sg), sg->length, sg->offset); - - if (sgoffset < sg->length) - break; - sgoffset -= sg->length; + struct scsi_data_buffer *sdb = scsi_out(task->sc); + struct scatterlist *sg = NULL; + int err; + + tdata->offset = offset; + tdata->count = count; + err = sgl_seek_offset(sdb->table.sgl, sdb->table.nents, + tdata->offset, &tdata->sgoffset, &sg); + if (err < 0) { + cxgb3i_log_warn("tpdu, sgl %u, bad offset %u/%u.\n", + sdb->table.nents, tdata->offset, + sdb->length); + return err; } - sgpg = sg_page(sg); - sglen = sg->length - sgoffset; - - do { - int j = skb_shinfo(skb)->nr_frags; - unsigned int copy; - - if (!sglen) { - sg = sg_next(sg); - sgpg = sg_page(sg); - sgoffset = 0; - sglen = sg->length; - ++i; + err = sgl_read_to_frags(sg, tdata->sgoffset, tdata->count, + tdata->frags, MAX_PDU_FRAGS); + if (err < 0) { + cxgb3i_log_warn("tpdu, sgl %u, bad offset %u + %u.\n", + sdb->table.nents, tdata->offset, + tdata->count); + return err; + } + tdata->nr_frags = err; + + if (tdata->nr_frags > MAX_SKB_FRAGS || + (padlen && tdata->nr_frags == MAX_SKB_FRAGS)) { + char *dst = skb->data + task->hdr_len; + skb_frag_t *frag = tdata->frags; + + /* data fits in the skb's headroom */ + for (i = 0; i < tdata->nr_frags; i++, frag++) { + char *src = kmap_atomic(frag->page, + KM_SOFTIRQ0); + + memcpy(dst, src+frag->page_offset, frag->size); + dst += frag->size; + kunmap_atomic(src, KM_SOFTIRQ0); } - copy = min(sglen, datalen); - if (j && skb_can_coalesce(skb, j, sgpg, - sg->offset + sgoffset)) { - skb_shinfo(skb)->frags[j - 1].size += copy; - } else { - get_page(sgpg); - skb_fill_page_desc(skb, j, sgpg, - sg->offset + sgoffset, copy); + if (padlen) { + memset(dst, 0, padlen); + padlen = 0; } - sgoffset += copy; - sglen -= copy; - datalen -= copy; - } while (datalen); + skb_put(skb, count + padlen); + } else { + /* data fit into frag_list */ + for (i = 0; i < tdata->nr_frags; i++) + get_page(tdata->frags[i].page); + + memcpy(skb_shinfo(skb)->frags, tdata->frags, + sizeof(skb_frag_t) * tdata->nr_frags); + skb_shinfo(skb)->nr_frags = tdata->nr_frags; + skb->len += count; + skb->data_len += count; + skb->truesize += count; + } + } else { pg = virt_to_page(task->data); - while (datalen) { - i = skb_shinfo(skb)->nr_frags; - frag = &skb_shinfo(skb)->frags[i]; - - get_page(pg); - frag->page = pg; - frag->page_offset = 0; - frag->size = min((unsigned int)PAGE_SIZE, datalen); - - skb_shinfo(skb)->nr_frags++; - datalen -= frag->size; - pg++; - } + get_page(pg); + skb_fill_page_desc(skb, 0, pg, offset_in_page(task->data), + count); + skb->len += count; + skb->data_len += count; + skb->truesize += count; } if (padlen) { i = skb_shinfo(skb)->nr_frags; - frag = &skb_shinfo(skb)->frags[i]; - frag->page = pad_page; - frag->page_offset = 0; - frag->size = padlen; - skb_shinfo(skb)->nr_frags++; + get_page(pad_page); + skb_fill_page_desc(skb, skb_shinfo(skb)->nr_frags, pad_page, 0, + padlen); + + skb->data_len += padlen; + skb->truesize += padlen; + skb->len += padlen; } - datalen = count + padlen; - skb->data_len += datalen; - skb->truesize += datalen; - skb->len += datalen; return 0; } int cxgb3i_conn_xmit_pdu(struct iscsi_task *task) { - struct iscsi_tcp_task *tcp_task = task->dd_data; - struct sk_buff *skb = tcp_task->dd_data; struct iscsi_tcp_conn *tcp_conn = task->conn->dd_data; struct cxgb3i_conn *cconn = tcp_conn->dd_data; + struct iscsi_tcp_task *tcp_task = task->dd_data; + struct cxgb3i_task_data *tdata = tcp_task->dd_data; + struct sk_buff *skb = tdata->skb; unsigned int datalen; int err; @@ -300,13 +383,14 @@ int cxgb3i_conn_xmit_pdu(struct iscsi_task *task) return 0; datalen = skb->data_len; - tcp_task->dd_data = NULL; + tdata->skb = NULL; err = cxgb3i_c3cn_send_pdus(cconn->cep->c3cn, skb); - cxgb3i_tx_debug("task 0x%p, skb 0x%p, len %u/%u, rv %d.\n", - task, skb, skb->len, skb->data_len, err); if (err > 0) { int pdulen = err; + cxgb3i_tx_debug("task 0x%p, skb 0x%p, len %u/%u, rv %d.\n", + task, skb, skb->len, skb->data_len, err); + if (task->conn->hdrdgst_en) pdulen += ISCSI_DIGEST_SIZE; if (datalen && task->conn->datadgst_en) @@ -325,12 +409,14 @@ int cxgb3i_conn_xmit_pdu(struct iscsi_task *task) return err; } /* reset skb to send when we are called again */ - tcp_task->dd_data = skb; + tdata->skb = skb; return -EAGAIN; } int cxgb3i_pdu_init(void) { + if (SKB_TX_HEADROOM > (512 * MAX_SKB_FRAGS)) + skb_extra_headroom = SKB_TX_HEADROOM; pad_page = alloc_page(GFP_KERNEL); if (!pad_page) return -ENOMEM; @@ -366,7 +452,9 @@ void cxgb3i_conn_pdu_ready(struct s3_conn *c3cn) skb = skb_peek(&c3cn->receive_queue); while (!err && skb) { __skb_unlink(skb, &c3cn->receive_queue); - read += skb_ulp_pdulen(skb); + read += skb_rx_pdulen(skb); + cxgb3i_rx_debug("conn 0x%p, cn 0x%p, rx skb 0x%p, pdulen %u.\n", + conn, c3cn, skb, skb_rx_pdulen(skb)); err = cxgb3i_conn_read_pdu_skb(conn, skb); __kfree_skb(skb); skb = skb_peek(&c3cn->receive_queue); @@ -377,6 +465,11 @@ void cxgb3i_conn_pdu_ready(struct s3_conn *c3cn) cxgb3i_c3cn_rx_credits(c3cn, read); } conn->rxdata_octets += read; + + if (err) { + cxgb3i_log_info("conn 0x%p rx failed err %d.\n", conn, err); + iscsi_conn_failure(conn, ISCSI_ERR_CONN_FAILED); + } } void cxgb3i_conn_tx_open(struct s3_conn *c3cn) diff --git a/drivers/scsi/cxgb3i/cxgb3i_pdu.h b/drivers/scsi/cxgb3i/cxgb3i_pdu.h index a3f685c..0770b23 100644 --- a/drivers/scsi/cxgb3i/cxgb3i_pdu.h +++ b/drivers/scsi/cxgb3i/cxgb3i_pdu.h @@ -53,7 +53,7 @@ struct cpl_rx_data_ddp_norss { #define ULP2_FLAG_DCRC_ERROR 0x20 #define ULP2_FLAG_PAD_ERROR 0x40 -void cxgb3i_conn_closing(struct s3_conn *); +void cxgb3i_conn_closing(struct s3_conn *c3cn); void cxgb3i_conn_pdu_ready(struct s3_conn *c3cn); void cxgb3i_conn_tx_open(struct s3_conn *c3cn); #endif -- cgit v0.10.2 From 992040f54069c96a59343976950f174448f4a351 Mon Sep 17 00:00:00 2001 From: Karen Xie Date: Fri, 13 Feb 2009 21:38:59 -0800 Subject: [SCSI] cxgb3i: added missing include in cxgb3i_ddp.h Signed-off-by: Karen Xie Reviewed-by: Mike Christie Signed-off-by: James Bottomley diff --git a/drivers/scsi/cxgb3i/cxgb3i_ddp.h b/drivers/scsi/cxgb3i/cxgb3i_ddp.h index 99d7014..3faae78 100644 --- a/drivers/scsi/cxgb3i/cxgb3i_ddp.h +++ b/drivers/scsi/cxgb3i/cxgb3i_ddp.h @@ -13,6 +13,8 @@ #ifndef __CXGB3I_ULP2_DDP_H__ #define __CXGB3I_ULP2_DDP_H__ +#include + /** * struct cxgb3i_tag_format - cxgb3i ulp tag format for an iscsi entity * -- cgit v0.10.2 From 802d52734adf0f288c49c05ed433872d7559c932 Mon Sep 17 00:00:00 2001 From: Karen Xie Date: Fri, 13 Feb 2009 21:39:04 -0800 Subject: [SCSI] cxgb3i: Fix spelling errors in documentation Signed-off-by: Karen Xie Signed-off-by: James Bottomley diff --git a/Documentation/scsi/cxgb3i.txt b/Documentation/scsi/cxgb3i.txt index 8141fa0..7ac8032 100644 --- a/Documentation/scsi/cxgb3i.txt +++ b/Documentation/scsi/cxgb3i.txt @@ -4,7 +4,7 @@ Introduction ============ The Chelsio T3 ASIC based Adapters (S310, S320, S302, S304, Mezz cards, etc. -series of products) supports iSCSI acceleration and iSCSI Direct Data Placement +series of products) support iSCSI acceleration and iSCSI Direct Data Placement (DDP) where the hardware handles the expensive byte touching operations, such as CRC computation and verification, and direct DMA to the final host memory destination: @@ -31,9 +31,9 @@ destination: the TCP segments onto the wire. It handles TCP retransmission if needed. - On receving, S3 h/w recovers the iSCSI PDU by reassembling TCP + On receiving, S3 h/w recovers the iSCSI PDU by reassembling TCP segments, separating the header and data, calculating and verifying - the digests, then forwards the header to the host. The payload data, + the digests, then forwarding the header to the host. The payload data, if possible, will be directly placed into the pre-posted host DDP buffer. Otherwise, the payload data will be sent to the host too. @@ -68,9 +68,8 @@ The following steps need to be taken to accelerates the open-iscsi initiator: sure the ip address is unique in the network. 3. edit /etc/iscsi/iscsid.conf - The default setting for MaxRecvDataSegmentLength (131072) is too big, - replace "node.conn[0].iscsi.MaxRecvDataSegmentLength" to be a value no - bigger than 15360 (for example 8192): + The default setting for MaxRecvDataSegmentLength (131072) is too big; + replace with a value no bigger than 15360 (for example 8192): node.conn[0].iscsi.MaxRecvDataSegmentLength = 8192 -- cgit v0.10.2 From b7e7bd34465518f3527bf47a8055f35077d40c6c Mon Sep 17 00:00:00 2001 From: Karen Xie Date: Fri, 13 Feb 2009 21:39:09 -0800 Subject: [SCSI] cxgb3i: update the driver version to 1.0.1 Signed-off-by: Karen Xie Signed-off-by: James Bottomley diff --git a/drivers/scsi/cxgb3i/cxgb3i_init.c b/drivers/scsi/cxgb3i/cxgb3i_init.c index 091ecb4..1ce9f24 100644 --- a/drivers/scsi/cxgb3i/cxgb3i_init.c +++ b/drivers/scsi/cxgb3i/cxgb3i_init.c @@ -12,8 +12,8 @@ #include "cxgb3i.h" #define DRV_MODULE_NAME "cxgb3i" -#define DRV_MODULE_VERSION "1.0.0" -#define DRV_MODULE_RELDATE "Jun. 1, 2008" +#define DRV_MODULE_VERSION "1.0.1" +#define DRV_MODULE_RELDATE "Jan. 2009" static char version[] = "Chelsio S3xx iSCSI Driver " DRV_MODULE_NAME -- cgit v0.10.2 From 4034cc68157bfa0b6622efe368488d3d3e20f4e6 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Sat, 21 Feb 2009 11:04:45 +0900 Subject: [SCSI] sd: revive sd_index_lock Commit f27bac2761cab5a2e212dea602d22457a9aa6943 which converted sd to use ida instead of idr incorrectly removed sd_index_lock around id allocation and free. idr/ida do have internal locks but they protect their free object lists not the allocation itself. The caller is responsible for that. This missing synchronization led to the same id being assigned to multiple devices leading to oops. Reported and tracked down by Stuart Hayes of Dell. Signed-off-by: Tejun Heo Cc: Stable Tree Signed-off-by: James Bottomley diff --git a/drivers/scsi/sd.c b/drivers/scsi/sd.c index d57566b..55310db 100644 --- a/drivers/scsi/sd.c +++ b/drivers/scsi/sd.c @@ -107,6 +107,7 @@ static void scsi_disk_release(struct device *cdev); static void sd_print_sense_hdr(struct scsi_disk *, struct scsi_sense_hdr *); static void sd_print_result(struct scsi_disk *, int); +static DEFINE_SPINLOCK(sd_index_lock); static DEFINE_IDA(sd_index_ida); /* This semaphore is used to mediate the 0->1 reference get in the @@ -1914,7 +1915,9 @@ static int sd_probe(struct device *dev) if (!ida_pre_get(&sd_index_ida, GFP_KERNEL)) goto out_put; + spin_lock(&sd_index_lock); error = ida_get_new(&sd_index_ida, &index); + spin_unlock(&sd_index_lock); } while (error == -EAGAIN); if (error) @@ -1936,7 +1939,9 @@ static int sd_probe(struct device *dev) return 0; out_free_index: + spin_lock(&sd_index_lock); ida_remove(&sd_index_ida, index); + spin_unlock(&sd_index_lock); out_put: put_disk(gd); out_free: @@ -1986,7 +1991,9 @@ static void scsi_disk_release(struct device *dev) struct scsi_disk *sdkp = to_scsi_disk(dev); struct gendisk *disk = sdkp->disk; + spin_lock(&sd_index_lock); ida_remove(&sd_index_ida, sdkp->index); + spin_unlock(&sd_index_lock); disk->private_data = NULL; put_disk(disk); -- cgit v0.10.2 From 126c098296c8f96cf7f6ca0fdb47265ac7994f00 Mon Sep 17 00:00:00 2001 From: James Bottomley Date: Thu, 19 Feb 2009 21:48:54 +0000 Subject: [SCSI] fix ABORTED_COMMAND looping forever problem Instead of terminating after five retries, commands terminated by ABORTED_COMMAND sense are retrying forever. The problem was introduced by: commit b60af5b0adf0da24c673598c8d3fb4d4189a15ce Author: Alan Stern Date: Mon Nov 3 15:56:47 2008 -0500 [SCSI] simplify scsi_io_completion() Which introduced an error whereby ABORTED_COMMAND now gets erroneously retried in scsi_io_completion. Fix this by returning the behaviour back to the default no retry. Reported-by: Sitsofe Wheeler Tested-by: Sitsofe Wheeler Signed-off-by: James Bottomley diff --git a/drivers/scsi/scsi_lib.c b/drivers/scsi/scsi_lib.c index 940dc32..b82ffd9 100644 --- a/drivers/scsi/scsi_lib.c +++ b/drivers/scsi/scsi_lib.c @@ -1040,12 +1040,11 @@ void scsi_io_completion(struct scsi_cmnd *cmd, unsigned int good_bytes) action = ACTION_FAIL; break; case ABORTED_COMMAND: + action = ACTION_FAIL; if (sshdr.asc == 0x10) { /* DIF */ description = "Target Data Integrity Failure"; - action = ACTION_FAIL; error = -EILSEQ; - } else - action = ACTION_RETRY; + } break; case NOT_READY: /* If the device is in the process of becoming -- cgit v0.10.2 From 5ce7868e159a3ee4ddf95f8522643991fea97cf2 Mon Sep 17 00:00:00 2001 From: Yinghai Lu Date: Wed, 18 Feb 2009 11:25:32 -0800 Subject: [SCSI] mpt: fix disable lsi sas to use msi as default Impact: fix bug the third param in module_param(,,) is perm instead of default value. we still need to assign default at first. Also, the default is now zero not one, so fix the parameter text to reflect that. Signed-off-by: Yinghai Lu Acked-by: Kashyap Desai Signed-off-by: James Bottomley diff --git a/drivers/message/fusion/mptbase.c b/drivers/message/fusion/mptbase.c index 96ac883..ea3aafb 100644 --- a/drivers/message/fusion/mptbase.c +++ b/drivers/message/fusion/mptbase.c @@ -91,9 +91,9 @@ MODULE_PARM_DESC(mpt_msi_enable_fc, " Enable MSI Support for FC \ controllers (default=0)"); static int mpt_msi_enable_sas; -module_param(mpt_msi_enable_sas, int, 1); +module_param(mpt_msi_enable_sas, int, 0); MODULE_PARM_DESC(mpt_msi_enable_sas, " Enable MSI Support for SAS \ - controllers (default=1)"); + controllers (default=0)"); static int mpt_channel_mapping; -- cgit v0.10.2 From e8bf069c419c1dc0657e02636441fe1179a9db14 Mon Sep 17 00:00:00 2001 From: Anssi Hannula Date: Sun, 22 Feb 2009 14:42:54 +0200 Subject: ALSA: aw2: do not grab every saa7146 based device Audiowerk2 driver snd-aw2 is bound to any saa7146 device as it does not check subsystem ids. Many DVB devices are saa7146 based, so aw2 driver grabs them as well. According to http://lkml.org/lkml/2008/10/15/311 aw2 devices have the subsystem ids set to 0, the saa7146 default. Fix conflicts with DVB devices by checking for subsystem ids = 0 specifically. Signed-off-by: Anssi Hannula Signed-off-by: Takashi Iwai diff --git a/sound/pci/aw2/aw2-alsa.c b/sound/pci/aw2/aw2-alsa.c index 3f00ddf..c7c54e7 100644 --- a/sound/pci/aw2/aw2-alsa.c +++ b/sound/pci/aw2/aw2-alsa.c @@ -165,7 +165,7 @@ module_param_array(enable, bool, NULL, 0444); MODULE_PARM_DESC(enable, "Enable Audiowerk2 soundcard."); static struct pci_device_id snd_aw2_ids[] = { - {PCI_VENDOR_ID_SAA7146, PCI_DEVICE_ID_SAA7146, PCI_ANY_ID, PCI_ANY_ID, + {PCI_VENDOR_ID_SAA7146, PCI_DEVICE_ID_SAA7146, 0, 0, 0, 0, 0}, {0} }; -- cgit v0.10.2 From 5370d96f85962769ea3df3a81cc885f257c51589 Mon Sep 17 00:00:00 2001 From: Steve Chen Date: Sat, 21 Feb 2009 08:05:04 -0600 Subject: ALSA: fix excessive background noise introduced by OSS emulation rate shrink Incorrect variable was used to get the next sample which caused S2 to be stuck with the same value resulting in loud background noise. Signed-off-by: Steve Chen Cc: Signed-off-by: Takashi Iwai diff --git a/sound/core/oss/rate.c b/sound/core/oss/rate.c index a466443..2fa9299 100644 --- a/sound/core/oss/rate.c +++ b/sound/core/oss/rate.c @@ -157,7 +157,7 @@ static void resample_shrink(struct snd_pcm_plugin *plugin, while (dst_frames1 > 0) { S1 = S2; if (src_frames1-- > 0) { - S1 = *src; + S2 = *src; src += src_step; } if (pos & ~R_MASK) { -- cgit v0.10.2 From 2d4663816064fabb68935f920bbd7ccdc7f9392d Mon Sep 17 00:00:00 2001 From: Luke Yelavich Date: Mon, 23 Feb 2009 13:00:33 +1100 Subject: ALSA: hda - add another MacBook Pro 3,1 SSID Reference: Ubuntu bug #33245 https://bugs.launchpad.net/bugs/332456 Signed-off-by: Luke Yelavich Signed-off-by: Takashi Iwai diff --git a/sound/pci/hda/patch_realtek.c b/sound/pci/hda/patch_realtek.c index ed8fcbd..f657122 100644 --- a/sound/pci/hda/patch_realtek.c +++ b/sound/pci/hda/patch_realtek.c @@ -7017,6 +7017,7 @@ static int patch_alc882(struct hda_codec *codec) case 0x106b3e00: /* iMac 24 Aluminium */ board_config = ALC885_IMAC24; break; + case 0x106b00a0: /* MacBookPro3,1 - Another revision */ case 0x106b00a1: /* Macbook (might be wrong - PCI SSID?) */ case 0x106b00a4: /* MacbookPro4,1 */ case 0x106b2c00: /* Macbook Pro rev3 */ -- cgit v0.10.2 From cc374c477c9bf95f409fed16426856d86a97394f Mon Sep 17 00:00:00 2001 From: Juan Jesus Garcia de Soria Date: Mon, 23 Feb 2009 08:11:59 +0100 Subject: ALSA: hda - Quirk for Acer Aspire 6530G The Acer Aspire 6530G needs the 4930G "model" for the front mic to work properly. Signed-off-by: Takashi Iwai diff --git a/sound/pci/hda/patch_realtek.c b/sound/pci/hda/patch_realtek.c index f657122..a680be0 100644 --- a/sound/pci/hda/patch_realtek.c +++ b/sound/pci/hda/patch_realtek.c @@ -8470,6 +8470,8 @@ static struct snd_pci_quirk alc883_cfg_tbl[] = { ALC888_ACER_ASPIRE_4930G), SND_PCI_QUIRK(0x1025, 0x015e, "Acer Aspire 6930G", ALC888_ACER_ASPIRE_4930G), + SND_PCI_QUIRK(0x1025, 0x0166, "Acer Aspire 6530G", + ALC888_ACER_ASPIRE_4930G), SND_PCI_QUIRK(0x1025, 0, "Acer laptop", ALC883_ACER), /* default Acer */ SND_PCI_QUIRK(0x1028, 0x020d, "Dell Inspiron 530", ALC888_6ST_DELL), SND_PCI_QUIRK(0x103c, 0x2a3d, "HP Pavillion", ALC883_6ST_DIG), -- cgit v0.10.2 From 5c138dcee7d4a9e68cce546a45968bbf5dbfce80 Mon Sep 17 00:00:00 2001 From: Andrey Borzenkov Date: Sun, 15 Feb 2009 12:51:18 +0300 Subject: orinoco: do not resgister NULL pm_notifier function With DEBUG_NOTIFIERS it results in [11330.890966] WARNING: at /home/bor/src/linux-git/kernel/notifier.c:88 notifier_call_chain+0x91/0xa0() [11330.890977] Hardware name: PORTEGE 4000 [11330.890983] Invalid notifier called! ... Without DEBUG_NOTIFIERS it most likely crashes on NULL pointer. Signed-off-by: Andrey Borzenkov Acked-by: David Kilroy Signed-off-by: John W. Linville diff --git a/drivers/net/wireless/orinoco/orinoco.c b/drivers/net/wireless/orinoco/orinoco.c index 45a04fa..067d1a9 100644 --- a/drivers/net/wireless/orinoco/orinoco.c +++ b/drivers/net/wireless/orinoco/orinoco.c @@ -3157,8 +3157,20 @@ static int orinoco_pm_notifier(struct notifier_block *notifier, return NOTIFY_DONE; } + +static void orinoco_register_pm_notifier(struct orinoco_private *priv) +{ + priv->pm_notifier.notifier_call = orinoco_pm_notifier; + register_pm_notifier(&priv->pm_notifier); +} + +static void orinoco_unregister_pm_notifier(struct orinoco_private *priv) +{ + unregister_pm_notifier(&priv->pm_notifier); +} #else /* !PM_SLEEP || HERMES_CACHE_FW_ON_INIT */ -#define orinoco_pm_notifier NULL +#define orinoco_register_pm_notifier(priv) do { } while(0) +#define orinoco_unregister_pm_notifier(priv) do { } while(0) #endif /********************************************************************/ @@ -3648,8 +3660,7 @@ struct net_device priv->cached_fw = NULL; /* Register PM notifiers */ - priv->pm_notifier.notifier_call = orinoco_pm_notifier; - register_pm_notifier(&priv->pm_notifier); + orinoco_register_pm_notifier(priv); return dev; } @@ -3673,7 +3684,7 @@ void free_orinocodev(struct net_device *dev) kfree(rx_data); } - unregister_pm_notifier(&priv->pm_notifier); + orinoco_unregister_pm_notifier(priv); orinoco_uncache_fw(priv); priv->wpa_ie_len = 0; -- cgit v0.10.2 From 40b130a947672d0ca886b718bfc4bd7641b6b39d Mon Sep 17 00:00:00 2001 From: Vasanthakumar Thiagarajan Date: Mon, 16 Feb 2009 13:55:07 +0530 Subject: ath9k: Fix panic upon attach failure [246916.338046] [246916.338048] Pid: 29265, comm: insmod Not tainted (2.6.29-rc4-wl #64) 9461DUU [246916.338051] EIP: 0060:[] EFLAGS: 00010202 CPU: 0 [246916.338055] EIP is at rollback_registered+0x24/0x220 [246916.338057] EAX: 00000001 EBX: 00000000 ECX: 00000000 EDX: f122e8fc [246916.338059] ESI: 00000000 EDI: 00000000 EBP: f6595d30 ESP: f6595d1c [246916.338062] DS: 007b ES: 007b FS: 00d8 GS: 0033 SS: 0068 [246916.338064] Process insmod (pid: 29265, ti=f6594000 task=f7343fe0 task.ti=f6594000) [246916.338067] Stack: [246916.338068] c04a2920 22222222 f6595d48 00000000 f122f080 f6595d48 c02ca489 f122e8fc [246916.338076] f122e220 f122f080 f122e220 f6595d5c f8a03156 f122e220 f122f080 f122e220 [246916.338085] f6595d80 f87359af f122f080 00002000 f874e129 f122f150 f122f080 f6290000 [246916.338094] Call Trace: [246916.338096] [] ? unregister_netdevice+0x19/0x70 [246916.338100] [] ? ieee80211_unregister_hw+0x36/0xd0 [mac80211] [246916.338112] [] ? ath_detach+0xcf/0x250 [ath9k] [246916.338127] [] ? ath_attach+0x26c/0x740 [ath9k] [246916.338139] [] ? ath_pci_probe+0x13a/0x310 [ath9k] [246916.338151] [] ? _raw_spin_unlock+0x68/0x80 [246916.338158] [] ? local_pci_probe+0xe/0x10 [246916.338162] [] ? pci_device_probe+0x60/0x80 [246916.338169] [] ? driver_probe_device+0x82/0x1b0 [246916.338174] [] ? __driver_attach+0x89/0x90 [246916.338180] [] ? bus_for_each_dev+0x4b/0x70 [246916.338184] [] ? pci_device_remove+0x0/0x40 [246916.338190] [] ? driver_attach+0x19/0x20 [246916.338193] [] ? __driver_attach+0x0/0x90 [246916.338197] [] ? bus_add_driver+0x1b7/0x230 [246916.338203] [] ? pci_device_remove+0x0/0x40 [246916.338206] [] ? driver_register+0x69/0x140 [246916.338212] [] ? ath9k_init+0x0/0x54 [ath9k] [246916.338221] [] ? __pci_register_driver+0x4e/0x90 [246916.338225] [] ? ath9k_init+0x0/0x54 [ath9k] [246916.338232] [] ? ath_pci_init+0x17/0x19 [ath9k] [246916.338238] [] ? ath9k_init+0x17/0x54 [ath9k] [246916.338245] [] ? tracepoint_update_probe_range+0x7e/0xb0 [246916.338249] [] ? do_one_initcall+0x2a/0x170 [246916.338252] [] ? up_read+0x16/0x30 [246916.338256] [] ? __blocking_notifier_call_chain+0x4d/0x60 [246916.338265] [] ? sys_init_module+0x8a/0x1c0 [246916.338269] [] ? trace_hardirqs_on_thunk+0xc/0x10 [246916.338272] [] ? sysenter_do_call+0x12/0x43 [246916.338276] Code: 8d bc 27 00 00 00 00 55 89 e5 56 89 c6 53 83 ec 0c a1 74 27 4a c0 85 c0 0f 85 4b 01 00 00 e8 04 7d 00 00 85 c0 0f 84 c9 01 00 00 <8b> 86 18 03 00 00 85 c0 0f 84 86 01 00 00 83 e8 01 0f 85 71 01 [246916.338328] EIP: [] rollback_registered+0x24/0x220 SS:ESP 0068:f6595d1c [246916.338335] ---[ end trace 76357c56a75ea34e ]--- Signed-off-by: Vasanthakumar Thiagarajan Signed-off-by: John W. Linville diff --git a/drivers/net/wireless/ath9k/main.c b/drivers/net/wireless/ath9k/main.c index 727f067..0e80990 100644 --- a/drivers/net/wireless/ath9k/main.c +++ b/drivers/net/wireless/ath9k/main.c @@ -1538,6 +1538,7 @@ bad2: bad: if (ah) ath9k_hw_detach(ah); + ath9k_exit_debug(sc); return error; } @@ -1545,7 +1546,7 @@ bad: static int ath_attach(u16 devid, struct ath_softc *sc) { struct ieee80211_hw *hw = sc->hw; - int error = 0; + int error = 0, i; DPRINTF(sc, ATH_DBG_CONFIG, "Attach ATH hw\n"); @@ -1589,11 +1590,11 @@ static int ath_attach(u16 devid, struct ath_softc *sc) /* initialize tx/rx engine */ error = ath_tx_init(sc, ATH_TXBUF); if (error != 0) - goto detach; + goto error_attach; error = ath_rx_init(sc, ATH_RXBUF); if (error != 0) - goto detach; + goto error_attach; #if defined(CONFIG_RFKILL) || defined(CONFIG_RFKILL_MODULE) /* Initialze h/w Rfkill */ @@ -1601,8 +1602,9 @@ static int ath_attach(u16 devid, struct ath_softc *sc) INIT_DELAYED_WORK(&sc->rf_kill.rfkill_poll, ath_rfkill_poll); /* Initialize s/w rfkill */ - if (ath_init_sw_rfkill(sc)) - goto detach; + error = ath_init_sw_rfkill(sc); + if (error) + goto error_attach; #endif error = ieee80211_register_hw(hw); @@ -1611,8 +1613,16 @@ static int ath_attach(u16 devid, struct ath_softc *sc) ath_init_leds(sc); return 0; -detach: - ath_detach(sc); + +error_attach: + /* cleanup tx queues */ + for (i = 0; i < ATH9K_NUM_TX_QUEUES; i++) + if (ATH_TXQ_SETUP(sc, i)) + ath_tx_cleanupq(sc, &sc->tx.txq[i]); + + ath9k_hw_detach(sc->sc_ah); + ath9k_exit_debug(sc); + return error; } -- cgit v0.10.2 From 046ee5d26ac91316a8ac0a29c0b33139dc9da20d Mon Sep 17 00:00:00 2001 From: Larry Finger Date: Tue, 17 Feb 2009 14:31:12 -0600 Subject: rtl8187: New USB ID's for RTL8187L MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Add new USB ID codes. These come from two postings on forums and mailing lists, and four are derived from the .inf that accompanies the latest Realtek Windows driver for the RTL8187L. Thanks to Viktor IlijaÅ¡ić and Xose Vazquez Perez for reporting these new ID's. Signed-off-by: Larry Finger Signed-off-by: John W. Linville diff --git a/drivers/net/wireless/rtl818x/rtl8187_dev.c b/drivers/net/wireless/rtl818x/rtl8187_dev.c index 22bc07e..f4747a1 100644 --- a/drivers/net/wireless/rtl818x/rtl8187_dev.c +++ b/drivers/net/wireless/rtl818x/rtl8187_dev.c @@ -48,6 +48,10 @@ static struct usb_device_id rtl8187_table[] __devinitdata = { {USB_DEVICE(0x0bda, 0x8189), .driver_info = DEVICE_RTL8187B}, {USB_DEVICE(0x0bda, 0x8197), .driver_info = DEVICE_RTL8187B}, {USB_DEVICE(0x0bda, 0x8198), .driver_info = DEVICE_RTL8187B}, + /* Surecom */ + {USB_DEVICE(0x0769, 0x11F2), .driver_info = DEVICE_RTL8187}, + /* Logitech */ + {USB_DEVICE(0x0789, 0x010C), .driver_info = DEVICE_RTL8187}, /* Netgear */ {USB_DEVICE(0x0846, 0x6100), .driver_info = DEVICE_RTL8187}, {USB_DEVICE(0x0846, 0x6a00), .driver_info = DEVICE_RTL8187}, @@ -57,8 +61,16 @@ static struct usb_device_id rtl8187_table[] __devinitdata = { /* Sitecom */ {USB_DEVICE(0x0df6, 0x000d), .driver_info = DEVICE_RTL8187}, {USB_DEVICE(0x0df6, 0x0028), .driver_info = DEVICE_RTL8187B}, + /* Sphairon Access Systems GmbH */ + {USB_DEVICE(0x114B, 0x0150), .driver_info = DEVICE_RTL8187}, + /* Dick Smith Electronics */ + {USB_DEVICE(0x1371, 0x9401), .driver_info = DEVICE_RTL8187}, /* Abocom */ {USB_DEVICE(0x13d1, 0xabe6), .driver_info = DEVICE_RTL8187}, + /* Qcom */ + {USB_DEVICE(0x18E8, 0x6232), .driver_info = DEVICE_RTL8187}, + /* AirLive */ + {USB_DEVICE(0x1b75, 0x8187), .driver_info = DEVICE_RTL8187}, {} }; -- cgit v0.10.2 From 044fad0dbb4e814c061916fe5a36851af2fd1135 Mon Sep 17 00:00:00 2001 From: Dhananjay Phadke Date: Tue, 24 Feb 2009 03:42:59 -0800 Subject: netxen: fix physical port mapping The PCI function to physical port mapping is valid only for old firmware. New firmware (4.0.0+) abstracts this. So driver should never try to access phy using invalid mapping. The behavior is unpredictable when PCI functions 4-7 are enabled on the same NIC. Signed-off-by: Dhananjay Phadke Signed-off-by: David S. Miller diff --git a/drivers/net/netxen/netxen_nic_main.c b/drivers/net/netxen/netxen_nic_main.c index 9f33e44..f425811 100644 --- a/drivers/net/netxen/netxen_nic_main.c +++ b/drivers/net/netxen/netxen_nic_main.c @@ -795,9 +795,12 @@ netxen_nic_probe(struct pci_dev *pdev, const struct pci_device_id *ent) * See if the firmware gave us a virtual-physical port mapping. */ adapter->physical_port = adapter->portnum; - i = adapter->pci_read_normalize(adapter, CRB_V2P(adapter->portnum)); - if (i != 0x55555555) - adapter->physical_port = i; + if (adapter->fw_major < 4) { + i = adapter->pci_read_normalize(adapter, + CRB_V2P(adapter->portnum)); + if (i != 0x55555555) + adapter->physical_port = i; + } adapter->flags &= ~(NETXEN_NIC_MSI_ENABLED | NETXEN_NIC_MSIX_ENABLED); -- cgit v0.10.2 From 028e1415a78733fcd2cba4b4c001826cc37a373e Mon Sep 17 00:00:00 2001 From: Dhananjay Phadke Date: Tue, 24 Feb 2009 03:44:23 -0800 Subject: netxen: handle pci bar 0 mapping failure PCI bar 0 is used for memory mapped register access. If ioremap fails (returns NULL), register access results in crash. Use pci_ioremap_bar() instead of ioremap(), the latter fails on on 32 bit powerpc where pci resource address is > 32 bits. Signed-off-by: Dhananjay Phadke Signed-off-by: David S. Miller diff --git a/drivers/net/netxen/netxen_nic_main.c b/drivers/net/netxen/netxen_nic_main.c index f425811..1308778 100644 --- a/drivers/net/netxen/netxen_nic_main.c +++ b/drivers/net/netxen/netxen_nic_main.c @@ -588,7 +588,12 @@ netxen_nic_probe(struct pci_dev *pdev, const struct pci_device_id *ent) adapter->pci_mem_read = netxen_nic_pci_mem_read_2M; adapter->pci_mem_write = netxen_nic_pci_mem_write_2M; - mem_ptr0 = ioremap(mem_base, mem_len); + mem_ptr0 = pci_ioremap_bar(pdev, 0); + if (mem_ptr0 == NULL) { + dev_err(&pdev->dev, "failed to map PCI bar 0\n"); + return -EIO; + } + pci_len0 = mem_len; first_page_group_start = 0; first_page_group_end = 0; -- cgit v0.10.2 From 325fb5b4d26038cba665dd0d8ee09555321061f0 Mon Sep 17 00:00:00 2001 From: Josef Drexler Date: Tue, 24 Feb 2009 14:53:12 +0100 Subject: netfilter: xt_recent: fix proc-file addition/removal of IPv4 addresses Fix regression introduded by commit 079aa88 (netfilter: xt_recent: IPv6 support): From http://bugzilla.kernel.org/show_bug.cgi?id=12753: Problem Description: An uninitialized buffer causes IPv4 addresses added manually (via the +IP command to the proc interface) to never match any packets. Similarly, the -IP command fails to remove IPv4 addresses. Details: In the function recent_entry_lookup, the xt_recent module does comparisons of the entire nf_inet_addr union value, both for IPv4 and IPv6 addresses. For addresses initialized from actual packets the remaining 12 bytes not occupied by the IPv4 are zeroed so this works correctly. However when setting the nf_inet_addr addr variable in the recent_mt_proc_write function, only the IPv4 bytes are initialized and the remaining 12 bytes contain garbage. Hence addresses added in this way never match any packets, unless these uninitialized 12 bytes happened to be zero by coincidence. Similarly, addresses cannot consistently be removed using the proc interface due to mismatch of the garbage bytes (although it will sometimes work to remove an address that was added manually). Reading the /proc/net/xt_recent/ entries hides this problem because this only uses the first 4 bytes when displaying IPv4 addresses. Steps to reproduce: $ iptables -I INPUT -m recent --rcheck -j LOG $ echo +169.254.156.239 > /proc/net/xt_recent/DEFAULT $ cat /proc/net/xt_recent/DEFAULT src=169.254.156.239 ttl: 0 last_seen: 119910 oldest_pkt: 1 119910 [At this point no packets from 169.254.156.239 are being logged.] $ iptables -I INPUT -s 169.254.156.239 -m recent --set $ cat /proc/net/xt_recent/DEFAULT src=169.254.156.239 ttl: 0 last_seen: 119910 oldest_pkt: 1 119910 src=169.254.156.239 ttl: 255 last_seen: 126184 oldest_pkt: 4 125434, 125684, 125934, 126184 [At this point, adding the address via an iptables rule, packets are being logged correctly.] $ echo -169.254.156.239 > /proc/net/xt_recent/DEFAULT $ cat /proc/net/xt_recent/DEFAULT src=169.254.156.239 ttl: 0 last_seen: 119910 oldest_pkt: 1 119910 src=169.254.156.239 ttl: 255 last_seen: 126992 oldest_pkt: 10 125434, 125684, 125934, 126184, 126434, 126684, 126934, 126991, 126991, 126992 $ echo -169.254.156.239 > /proc/net/xt_recent/DEFAULT $ cat /proc/net/xt_recent/DEFAULT src=169.254.156.239 ttl: 0 last_seen: 119910 oldest_pkt: 1 119910 src=169.254.156.239 ttl: 255 last_seen: 126992 oldest_pkt: 10 125434, 125684, 125934, 126184, 126434, 126684, 126934, 126991, 126991, 126992 [Removing the address via /proc interface failed evidently.] Possible solutions: - initialize the addr variable in recent_mt_proc_write - compare only 4 bytes for IPv4 addresses in recent_entry_lookup Signed-off-by: Patrick McHardy diff --git a/net/netfilter/xt_recent.c b/net/netfilter/xt_recent.c index fe80b61..791e030 100644 --- a/net/netfilter/xt_recent.c +++ b/net/netfilter/xt_recent.c @@ -542,7 +542,7 @@ recent_mt_proc_write(struct file *file, const char __user *input, struct recent_entry *e; char buf[sizeof("+b335:1d35:1e55:dead:c0de:1715:5afe:c0de")]; const char *c = buf; - union nf_inet_addr addr; + union nf_inet_addr addr = {}; u_int16_t family; bool add, succ; -- cgit v0.10.2 From 1f9da5544073d38e05139f8ce9da24e78653c73e Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Tue, 24 Feb 2009 15:31:02 +0100 Subject: ALSA: emu10k1 - Fix digital/analog switch on audigy2 ZS Fix the inverted logic of shared spdif switch. Reference: Novell bnc#478496 https://bugzilla.novell.com/show_bug.cgi?id=478496 Signed-off-by: Takashi Iwai diff --git a/sound/pci/emu10k1/emu10k1_main.c b/sound/pci/emu10k1/emu10k1_main.c index 7958006..101a1c1 100644 --- a/sound/pci/emu10k1/emu10k1_main.c +++ b/sound/pci/emu10k1/emu10k1_main.c @@ -1528,6 +1528,7 @@ static struct snd_emu_chip_details emu_chip_details[] = { .ca0151_chip = 1, .spk71 = 1, .spdif_bug = 1, + .invert_shared_spdif = 1, /* digital/analog switch swapped */ .ac97_chip = 1} , {.vendor = 0x1102, .device = 0x0004, .subsystem = 0x10021102, .driver = "Audigy2", .name = "SB Audigy 2 Platinum [SB0240P]", -- cgit v0.10.2 From d22157b3d7a24d953ef5cce4515436be65b2e121 Mon Sep 17 00:00:00 2001 From: Chris Wright Date: Mon, 23 Feb 2009 21:50:35 -0800 Subject: PCI: add some sysfs ABI docs Add sysfs ABI docs for driver entries bind, unbind and new_id. These entries are pretty old, from 2.6.0 onwards AFAIK, so this documents current behaviour. Signed-off-by: Chris Wright Acked-by: Greg Kroah-Hartman Signed-off-by: Jesse Barnes diff --git a/Documentation/ABI/testing/sysfs-bus-pci b/Documentation/ABI/testing/sysfs-bus-pci index ceddcff..e638e15 100644 --- a/Documentation/ABI/testing/sysfs-bus-pci +++ b/Documentation/ABI/testing/sysfs-bus-pci @@ -1,3 +1,46 @@ +What: /sys/bus/pci/drivers/.../bind +Date: December 2003 +Contact: linux-pci@vger.kernel.org +Description: + Writing a device location to this file will cause + the driver to attempt to bind to the device found at + this location. This is useful for overriding default + bindings. The format for the location is: DDDD:BB:DD.F. + That is Domain:Bus:Device.Function and is the same as + found in /sys/bus/pci/devices/. For example: + # echo 0000:00:19.0 > /sys/bus/pci/drivers/foo/bind + (Note: kernels before 2.6.28 may require echo -n). + +What: /sys/bus/pci/drivers/.../unbind +Date: December 2003 +Contact: linux-pci@vger.kernel.org +Description: + Writing a device location to this file will cause the + driver to attempt to unbind from the device found at + this location. This may be useful when overriding default + bindings. The format for the location is: DDDD:BB:DD.F. + That is Domain:Bus:Device.Function and is the same as + found in /sys/bus/pci/devices/. For example: + # echo 0000:00:19.0 > /sys/bus/pci/drivers/foo/unbind + (Note: kernels before 2.6.28 may require echo -n). + +What: /sys/bus/pci/drivers/.../new_id +Date: December 2003 +Contact: linux-pci@vger.kernel.org +Description: + Writing a device ID to this file will attempt to + dynamically add a new device ID to a PCI device driver. + This may allow the driver to support more hardware than + was included in the driver's static device ID support + table at compile time. The format for the device ID is: + VVVV DDDD SVVV SDDD CCCC MMMM PPPP. That is Vendor ID, + Device ID, Subsystem Vendor ID, Subsystem Device ID, + Class, Class Mask, and Private Driver Data. The Vendor ID + and Device ID fields are required, the rest are optional. + Upon successfully adding an ID, the driver will probe + for the device and attempt to bind to it. For example: + # echo "8086 10f5" > /sys/bus/pci/drivers/foo/new_id + What: /sys/bus/pci/devices/.../vpd Date: February 2008 Contact: Ben Hutchings -- cgit v0.10.2 From 1dec6b054dd1fc780e18b815068bf5677409eb2d Mon Sep 17 00:00:00 2001 From: Yinghai Lu Date: Mon, 23 Feb 2009 11:51:59 -0800 Subject: PCI: don't enable too many HT MSI mappings Prakash reported that his c51-mcp51 ondie sound card doesn't work with MSI. But if he hacks out the HT-MSI quirk, MSI works fine. So this patch reworks the nv_msi_ht_cap_quirk(). It will now only enable ht_msi on own its root device, avoiding enabling it on devices following that root dev. Reported-by: Prakash Punnoor Tested-by: Prakash Punnoor Signed-off-by: Yinghai Lu Signed-off-by: Jesse Barnes diff --git a/drivers/pci/quirks.c b/drivers/pci/quirks.c index a21e1c2..a852329 100644 --- a/drivers/pci/quirks.c +++ b/drivers/pci/quirks.c @@ -2050,10 +2050,100 @@ DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_NVIDIA, PCI_DEVICE_ID_NVIDIA_NVENET_15, nvenet_msi_disable); -static void __devinit nv_msi_ht_cap_quirk(struct pci_dev *dev) +static void __devinit nv_ht_enable_msi_mapping(struct pci_dev *dev) { struct pci_dev *host_bridge; + int pos; + int i, dev_no; + int found = 0; + + dev_no = dev->devfn >> 3; + for (i = dev_no; i >= 0; i--) { + host_bridge = pci_get_slot(dev->bus, PCI_DEVFN(i, 0)); + if (!host_bridge) + continue; + + pos = pci_find_ht_capability(host_bridge, HT_CAPTYPE_SLAVE); + if (pos != 0) { + found = 1; + break; + } + pci_dev_put(host_bridge); + } + + if (!found) + return; + + /* root did that ! */ + if (msi_ht_cap_enabled(host_bridge)) + goto out; + + ht_enable_msi_mapping(dev); + +out: + pci_dev_put(host_bridge); +} + +static void __devinit ht_disable_msi_mapping(struct pci_dev *dev) +{ + int pos, ttl = 48; + + pos = pci_find_ht_capability(dev, HT_CAPTYPE_MSI_MAPPING); + while (pos && ttl--) { + u8 flags; + + if (pci_read_config_byte(dev, pos + HT_MSI_FLAGS, + &flags) == 0) { + dev_info(&dev->dev, "Enabling HT MSI Mapping\n"); + + pci_write_config_byte(dev, pos + HT_MSI_FLAGS, + flags & ~HT_MSI_FLAGS_ENABLE); + } + pos = pci_find_next_ht_capability(dev, pos, + HT_CAPTYPE_MSI_MAPPING); + } +} + +static int __devinit ht_check_msi_mapping(struct pci_dev *dev) +{ int pos, ttl = 48; + int found = 0; + + /* check if there is HT MSI cap or enabled on this device */ + pos = pci_find_ht_capability(dev, HT_CAPTYPE_MSI_MAPPING); + while (pos && ttl--) { + u8 flags; + + if (found < 1) + found = 1; + if (pci_read_config_byte(dev, pos + HT_MSI_FLAGS, + &flags) == 0) { + if (flags & HT_MSI_FLAGS_ENABLE) { + if (found < 2) { + found = 2; + break; + } + } + } + pos = pci_find_next_ht_capability(dev, pos, + HT_CAPTYPE_MSI_MAPPING); + } + + return found; +} + +static void __devinit nv_msi_ht_cap_quirk(struct pci_dev *dev) +{ + struct pci_dev *host_bridge; + int pos; + int found; + + /* check if there is HT MSI cap or enabled on this device */ + found = ht_check_msi_mapping(dev); + + /* no HT MSI CAP */ + if (found == 0) + return; /* * HT MSI mapping should be disabled on devices that are below @@ -2069,24 +2159,19 @@ static void __devinit nv_msi_ht_cap_quirk(struct pci_dev *dev) pos = pci_find_ht_capability(host_bridge, HT_CAPTYPE_SLAVE); if (pos != 0) { /* Host bridge is to HT */ - ht_enable_msi_mapping(dev); + if (found == 1) { + /* it is not enabled, try to enable it */ + nv_ht_enable_msi_mapping(dev); + } return; } - /* Host bridge is not to HT, disable HT MSI mapping on this device */ - pos = pci_find_ht_capability(dev, HT_CAPTYPE_MSI_MAPPING); - while (pos && ttl--) { - u8 flags; + /* HT MSI is not enabled */ + if (found == 1) + return; - if (pci_read_config_byte(dev, pos + HT_MSI_FLAGS, - &flags) == 0) { - dev_info(&dev->dev, "Disabling HT MSI mapping"); - pci_write_config_byte(dev, pos + HT_MSI_FLAGS, - flags & ~HT_MSI_FLAGS_ENABLE); - } - pos = pci_find_next_ht_capability(dev, pos, - HT_CAPTYPE_MSI_MAPPING); - } + /* Host bridge is not to HT, disable HT MSI mapping on this device */ + ht_disable_msi_mapping(dev); } DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_NVIDIA, PCI_ANY_ID, nv_msi_ht_cap_quirk); DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_AL, PCI_ANY_ID, nv_msi_ht_cap_quirk); -- cgit v0.10.2 From dbc7e1e567ef8cfc4b792ef6acb51d4ceb15746a Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Wed, 28 Jan 2009 19:31:18 -0800 Subject: PCI: pciehp: Handle interrupts that happen during initialization. Move the enabling of interrupts after all of the data structures are setup so that we can safely run the interrupt handler as soon as it is registered. Reviewed-by: Kenji Kaneshige Tested-by: Kenji Kaneshige Signed-off-by: Eric W. Biederman Signed-off-by: Jesse Barnes diff --git a/drivers/pci/hotplug/pciehp.h b/drivers/pci/hotplug/pciehp.h index db85284..39ae375 100644 --- a/drivers/pci/hotplug/pciehp.h +++ b/drivers/pci/hotplug/pciehp.h @@ -111,6 +111,7 @@ struct controller { int cmd_busy; unsigned int no_cmd_complete:1; unsigned int link_active_reporting:1; + unsigned int notification_enabled:1; }; #define INT_BUTTON_IGNORE 0 @@ -170,6 +171,7 @@ extern int pciehp_configure_device(struct slot *p_slot); extern int pciehp_unconfigure_device(struct slot *p_slot); extern void pciehp_queue_pushbutton_work(struct work_struct *work); struct controller *pcie_init(struct pcie_device *dev); +int pcie_init_notification(struct controller *ctrl); int pciehp_enable_slot(struct slot *p_slot); int pciehp_disable_slot(struct slot *p_slot); int pcie_enable_notification(struct controller *ctrl); diff --git a/drivers/pci/hotplug/pciehp_core.c b/drivers/pci/hotplug/pciehp_core.c index c248554..681e391 100644 --- a/drivers/pci/hotplug/pciehp_core.c +++ b/drivers/pci/hotplug/pciehp_core.c @@ -434,6 +434,13 @@ static int pciehp_probe(struct pcie_device *dev, const struct pcie_port_service_ goto err_out_release_ctlr; } + /* Enable events after we have setup the data structures */ + rc = pcie_init_notification(ctrl); + if (rc) { + ctrl_err(ctrl, "Notification initialization failed\n"); + goto err_out_release_ctlr; + } + /* Check if slot is occupied */ t_slot = pciehp_find_slot(ctrl, ctrl->slot_device_offset); t_slot->hpc_ops->get_adapter_status(t_slot, &value); diff --git a/drivers/pci/hotplug/pciehp_hpc.c b/drivers/pci/hotplug/pciehp_hpc.c index 71a8012..7a16c68 100644 --- a/drivers/pci/hotplug/pciehp_hpc.c +++ b/drivers/pci/hotplug/pciehp_hpc.c @@ -934,7 +934,7 @@ static void pcie_disable_notification(struct controller *ctrl) ctrl_warn(ctrl, "Cannot disable software notification\n"); } -static int pcie_init_notification(struct controller *ctrl) +int pcie_init_notification(struct controller *ctrl) { if (pciehp_request_irq(ctrl)) return -1; @@ -942,13 +942,17 @@ static int pcie_init_notification(struct controller *ctrl) pciehp_free_irq(ctrl); return -1; } + ctrl->notification_enabled = 1; return 0; } static void pcie_shutdown_notification(struct controller *ctrl) { - pcie_disable_notification(ctrl); - pciehp_free_irq(ctrl); + if (ctrl->notification_enabled) { + pcie_disable_notification(ctrl); + pciehp_free_irq(ctrl); + ctrl->notification_enabled = 0; + } } static int pcie_init_slot(struct controller *ctrl) @@ -1110,13 +1114,8 @@ struct controller *pcie_init(struct pcie_device *dev) if (pcie_init_slot(ctrl)) goto abort_ctrl; - if (pcie_init_notification(ctrl)) - goto abort_slot; - return ctrl; -abort_slot: - pcie_cleanup_slot(ctrl); abort_ctrl: kfree(ctrl); abort: -- cgit v0.10.2 From 1f9f13c8d59c1d8da1a602b71d1ab96d1d37d69e Mon Sep 17 00:00:00 2001 From: Andrew Patterson Date: Fri, 20 Feb 2009 16:04:59 -0700 Subject: PCI: Enable PCIe AER only after checking firmware support The PCIe port driver currently sets the PCIe AER error reporting bits for any root or switch port without first checking to see if firmware will grant control. This patch moves setting these bits to the AER service driver aer_enable_port routine. The bits are then set for the root port and any downstream switch ports after the check for firmware support (aer_osc_setup) is made. The patch also unsets the bits in a similar fashion when the AER service driver is unloaded. Reviewed-by: Alex Chiang Signed-off-by: Andrew Patterson Signed-off-by: Jesse Barnes diff --git a/drivers/pci/pcie/aer/aerdrv_core.c b/drivers/pci/pcie/aer/aerdrv_core.c index aac7006..d0c9736 100644 --- a/drivers/pci/pcie/aer/aerdrv_core.c +++ b/drivers/pci/pcie/aer/aerdrv_core.c @@ -108,6 +108,34 @@ int pci_cleanup_aer_correct_error_status(struct pci_dev *dev) } #endif /* 0 */ + +static void set_device_error_reporting(struct pci_dev *dev, void *data) +{ + bool enable = *((bool *)data); + + if (dev->pcie_type != PCIE_RC_PORT && + dev->pcie_type != PCIE_SW_UPSTREAM_PORT && + dev->pcie_type != PCIE_SW_DOWNSTREAM_PORT) + return; + + if (enable) + pci_enable_pcie_error_reporting(dev); + else + pci_disable_pcie_error_reporting(dev); +} + +/** + * set_downstream_devices_error_reporting - enable/disable the error reporting bits on the root port and its downstream ports. + * @dev: pointer to root port's pci_dev data structure + * @enable: true = enable error reporting, false = disable error reporting. + */ +static void set_downstream_devices_error_reporting(struct pci_dev *dev, + bool enable) +{ + set_device_error_reporting(dev, &enable); + pci_walk_bus(dev->subordinate, set_device_error_reporting, &enable); +} + static int find_device_iter(struct device *device, void *data) { struct pci_dev *dev; @@ -525,15 +553,11 @@ void aer_enable_rootport(struct aer_rpc *rpc) pci_read_config_dword(pdev, aer_pos + PCI_ERR_UNCOR_STATUS, ®32); pci_write_config_dword(pdev, aer_pos + PCI_ERR_UNCOR_STATUS, reg32); - /* Enable Root Port device reporting error itself */ - pci_read_config_word(pdev, pos+PCI_EXP_DEVCTL, ®16); - reg16 = reg16 | - PCI_EXP_DEVCTL_CERE | - PCI_EXP_DEVCTL_NFERE | - PCI_EXP_DEVCTL_FERE | - PCI_EXP_DEVCTL_URRE; - pci_write_config_word(pdev, pos+PCI_EXP_DEVCTL, - reg16); + /* + * Enable error reporting for the root port device and downstream port + * devices. + */ + set_downstream_devices_error_reporting(pdev, true); /* Enable Root Port's interrupt in response to error messages */ pci_write_config_dword(pdev, @@ -553,6 +577,12 @@ static void disable_root_aer(struct aer_rpc *rpc) u32 reg32; int pos; + /* + * Disable error reporting for the root port device and downstream port + * devices. + */ + set_downstream_devices_error_reporting(pdev, false); + pos = pci_find_ext_capability(pdev, PCI_EXT_CAP_ID_ERR); /* Disable Root's interrupt in response to error messages */ pci_write_config_dword(pdev, pos + PCI_ERR_ROOT_COMMAND, 0); diff --git a/drivers/pci/pcie/portdrv_pci.c b/drivers/pci/pcie/portdrv_pci.c index f9b874e..248b4db 100644 --- a/drivers/pci/pcie/portdrv_pci.c +++ b/drivers/pci/pcie/portdrv_pci.c @@ -97,8 +97,6 @@ static int __devinit pcie_portdrv_probe (struct pci_dev *dev, pcie_portdrv_save_config(dev); - pci_enable_pcie_error_reporting(dev); - return 0; } -- cgit v0.10.2 From a52b8bd338630f78a6bfe39fe17cb8469d2679ae Mon Sep 17 00:00:00 2001 From: Joe Perches Date: Tue, 24 Feb 2009 16:40:16 -0800 Subject: tcp_scalable: Update malformed & dead url Signed-off-by: Joe Perches Signed-off-by: David S. Miller diff --git a/net/ipv4/tcp_scalable.c b/net/ipv4/tcp_scalable.c index 2747ec7..4660b08 100644 --- a/net/ipv4/tcp_scalable.c +++ b/net/ipv4/tcp_scalable.c @@ -1,6 +1,6 @@ /* Tom Kelly's Scalable TCP * - * See htt://www-lce.eng.cam.ac.uk/~ctk21/scalable/ + * See http://www.deneholme.net/tom/scalable/ * * John Heffner */ -- cgit v0.10.2 From 09b4068a7fe442efc40e9dcbcf5ff37c3338ab15 Mon Sep 17 00:00:00 2001 From: NeilBrown Date: Wed, 25 Feb 2009 13:18:47 +1100 Subject: md/raid10: Don't skip more than 1 bitmap-chunk at a time during recovery. When doing recovery on a raid10 with a write-intent bitmap, we only need to recovery chunks that are flagged in the bitmap. However if we choose to skip a chunk as it isn't flag, the code currently skips the whole raid10-chunk, thus it might not recovery some blocks that need recovering. This patch fixes it. In case that is confusing, it might help to understand that there is a 'raid10 chunk size' which guides how data is distributed across the devices, and a 'bitmap chunk size' which says how much data corresponds to a single bit in the bitmap. This bug only affects cases where the bitmap chunk size is smaller than the raid10 chunk size. Cc: stable@kernel.org Signed-off-by: NeilBrown diff --git a/drivers/md/raid10.c b/drivers/md/raid10.c index 6736d6d..118f89e 100644 --- a/drivers/md/raid10.c +++ b/drivers/md/raid10.c @@ -2010,13 +2010,13 @@ static sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *skipped, i /* There is nowhere to write, so all non-sync * drives must be failed, so try the next chunk... */ - { - sector_t sec = max_sector - sector_nr; - sectors_skipped += sec; + if (sector_nr + max_sync < max_sector) + max_sector = sector_nr + max_sync; + + sectors_skipped += (max_sector - sector_nr); chunks_skipped ++; sector_nr = max_sector; goto skipped; - } } static int run(mddev_t *mddev) -- cgit v0.10.2 From 78200d45cde2a79c0d0ae0407883bb264caa3c18 Mon Sep 17 00:00:00 2001 From: NeilBrown Date: Wed, 25 Feb 2009 13:18:47 +1100 Subject: md/raid10: Don't call bitmap_cond_end_sync when we are doing recovery. For raid1/4/5/6, resync (fixing inconsistencies between devices) is very similar to recovery (rebuilding a failed device onto a spare). The both walk through the device addresses in order. For raid10 it can be quite different. resync follows the 'array' address, and makes sure all copies are the same. Recover walks through 'device' addresses and recreates each missing block. The 'bitmap_cond_end_sync' function allows the write-intent-bitmap (When present) to be updated to reflect a partially completed resync. It makes assumptions which mean that it does not work correctly for raid10 recovery at all. In particularly, it can cause bitmap-directed recovery of a raid10 to not recovery some of the blocks that need to be recovered. So move the call to bitmap_cond_end_sync into the resync path, rather than being in the common "resync or recovery" path. Cc: stable@kernel.org Signed-off-by: NeilBrown diff --git a/drivers/md/raid10.c b/drivers/md/raid10.c index 118f89e..e1feb87 100644 --- a/drivers/md/raid10.c +++ b/drivers/md/raid10.c @@ -1749,8 +1749,6 @@ static sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *skipped, i if (!go_faster && conf->nr_waiting) msleep_interruptible(1000); - bitmap_cond_end_sync(mddev->bitmap, sector_nr); - /* Again, very different code for resync and recovery. * Both must result in an r10bio with a list of bios that * have bi_end_io, bi_sector, bi_bdev set, @@ -1886,6 +1884,8 @@ static sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *skipped, i /* resync. Schedule a read for every block at this virt offset */ int count = 0; + bitmap_cond_end_sync(mddev->bitmap, sector_nr); + if (!bitmap_start_sync(mddev->bitmap, sector_nr, &sync_blocks, mddev->degraded) && !conf->fullsync && !test_bit(MD_RECOVERY_REQUESTED, &mddev->recovery)) { -- cgit v0.10.2 From 73d5c38a9536142e062c35997b044e89166e063b Mon Sep 17 00:00:00 2001 From: NeilBrown Date: Wed, 25 Feb 2009 13:18:47 +1100 Subject: md: avoid races when stopping resync. There has been a race in raid10 and raid1 for a long time which has only recently started showing up due to a scheduler changed. When a sync_read request finishes, as soon as reschedule_retry is called, another thread can mark the resync request as having completed, so md_do_sync can finish, ->stop can be called, and ->conf can be freed. So using conf after reschedule_retry is not safe. Similarly, when finishing a sync_write, calling md_done_sync must be the last thing we do, as it allows a chain of events which will free conf and other data structures. The first of these requires action in raid10.c The second requires action in raid1.c and raid10.c Cc: stable@kernel.org Signed-off-by: NeilBrown diff --git a/drivers/md/raid1.c b/drivers/md/raid1.c index 01e3cff..e246642 100644 --- a/drivers/md/raid1.c +++ b/drivers/md/raid1.c @@ -1237,8 +1237,9 @@ static void end_sync_write(struct bio *bio, int error) update_head_pos(mirror, r1_bio); if (atomic_dec_and_test(&r1_bio->remaining)) { - md_done_sync(mddev, r1_bio->sectors, uptodate); + sector_t s = r1_bio->sectors; put_buf(r1_bio); + md_done_sync(mddev, s, uptodate); } } diff --git a/drivers/md/raid10.c b/drivers/md/raid10.c index e1feb87..7301631 100644 --- a/drivers/md/raid10.c +++ b/drivers/md/raid10.c @@ -1236,6 +1236,7 @@ static void end_sync_read(struct bio *bio, int error) /* for reconstruct, we always reschedule after a read. * for resync, only after all reads */ + rdev_dec_pending(conf->mirrors[d].rdev, conf->mddev); if (test_bit(R10BIO_IsRecover, &r10_bio->state) || atomic_dec_and_test(&r10_bio->remaining)) { /* we have read all the blocks, @@ -1243,7 +1244,6 @@ static void end_sync_read(struct bio *bio, int error) */ reschedule_retry(r10_bio); } - rdev_dec_pending(conf->mirrors[d].rdev, conf->mddev); } static void end_sync_write(struct bio *bio, int error) @@ -1264,11 +1264,13 @@ static void end_sync_write(struct bio *bio, int error) update_head_pos(i, r10_bio); + rdev_dec_pending(conf->mirrors[d].rdev, mddev); while (atomic_dec_and_test(&r10_bio->remaining)) { if (r10_bio->master_bio == NULL) { /* the primary of several recovery bios */ - md_done_sync(mddev, r10_bio->sectors, 1); + sector_t s = r10_bio->sectors; put_buf(r10_bio); + md_done_sync(mddev, s, 1); break; } else { r10bio_t *r10_bio2 = (r10bio_t *)r10_bio->master_bio; @@ -1276,7 +1278,6 @@ static void end_sync_write(struct bio *bio, int error) r10_bio = r10_bio2; } } - rdev_dec_pending(conf->mirrors[d].rdev, mddev); } /* -- cgit v0.10.2 From 7c04d1d97a8d918b7ae2ef478229862b71a65f06 Mon Sep 17 00:00:00 2001 From: Jesse Barnes Date: Mon, 23 Feb 2009 15:36:40 -0800 Subject: drm/i915: remove PLL debugging messages These are normal; we walk through different values looking for the right one, so why flood the screen with messages? Signed-off-by: Jesse Barnes Reviewed-by: Eric Anholt Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 65b635c..a283427 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -217,7 +217,7 @@ bool intel_pipe_has_type (struct drm_crtc *crtc, int type) return false; } -#define INTELPllInvalid(s) do { DRM_DEBUG(s); return false; } while (0) +#define INTELPllInvalid(s) do { /* DRM_DEBUG(s); */ return false; } while (0) /** * Returns whether the given set of divisors are valid for a given refclk with * the given connectors. -- cgit v0.10.2 From 37df96736bfe6f5fd9a141d62946e1083d73e712 Mon Sep 17 00:00:00 2001 From: Jesse Barnes Date: Mon, 23 Feb 2009 15:36:42 -0800 Subject: drm/i915: handle bogus VBT panel timing We've seen cases in the wild where the VBT sync data is wrong, so add some code to fix it up in that case, taking care to make sure that the total is greater than the sync end. Signed-off-by: Jesse Barnes Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/i915/intel_bios.c b/drivers/gpu/drm/i915/intel_bios.c index 65be30d..fc28e2b 100644 --- a/drivers/gpu/drm/i915/intel_bios.c +++ b/drivers/gpu/drm/i915/intel_bios.c @@ -111,6 +111,12 @@ parse_panel_data(struct drm_i915_private *dev_priv, struct bdb_header *bdb) panel_fixed_mode->clock = dvo_timing->clock * 10; panel_fixed_mode->type = DRM_MODE_TYPE_PREFERRED; + /* Some VBTs have bogus h/vtotal values */ + if (panel_fixed_mode->hsync_end > panel_fixed_mode->htotal) + panel_fixed_mode->htotal = panel_fixed_mode->hsync_end + 1; + if (panel_fixed_mode->vsync_end > panel_fixed_mode->vtotal) + panel_fixed_mode->vtotal = panel_fixed_mode->vsync_end + 1; + drm_mode_set_name(panel_fixed_mode); dev_priv->vbt_mode = panel_fixed_mode; -- cgit v0.10.2 From c8766ac5933d6ee75e7ce379a1eb5ceb451fcb83 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Mon, 23 Feb 2009 08:44:33 -0800 Subject: drm: Fix shifts of EDID vsync offset/width fields. Signed-off-by: Linus Torvalds Reviewed-by: Eric Anholt Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/drm_edid.c b/drivers/gpu/drm/drm_edid.c index 5a4d324..e902b1c 100644 --- a/drivers/gpu/drm/drm_edid.c +++ b/drivers/gpu/drm/drm_edid.c @@ -320,10 +320,10 @@ static struct drm_display_mode *drm_mode_detailed(struct drm_device *dev, mode->htotal = mode->hdisplay + ((pt->hblank_hi << 8) | pt->hblank_lo); mode->vdisplay = (pt->vactive_hi << 8) | pt->vactive_lo; - mode->vsync_start = mode->vdisplay + ((pt->vsync_offset_hi << 8) | + mode->vsync_start = mode->vdisplay + ((pt->vsync_offset_hi << 4) | pt->vsync_offset_lo); mode->vsync_end = mode->vsync_start + - ((pt->vsync_pulse_width_hi << 8) | + ((pt->vsync_pulse_width_hi << 4) | pt->vsync_pulse_width_lo); mode->vtotal = mode->vdisplay + ((pt->vblank_hi << 8) | pt->vblank_lo); -- cgit v0.10.2 From fe56cf45f951b3810313584605c1d8a4f20b33a4 Mon Sep 17 00:00:00 2001 From: Jesse Barnes Date: Mon, 23 Feb 2009 15:36:41 -0800 Subject: drm: Fix ordering of bit fields in EDID structure leading huge vsync values. Signed-off-by: Jesse Barnes Reviewed-by: Eric Anholt Signed-off-by: Dave Airlie diff --git a/include/drm/drm_edid.h b/include/drm/drm_edid.h index c707c15..ff8d27a 100644 --- a/include/drm/drm_edid.h +++ b/include/drm/drm_edid.h @@ -58,10 +58,10 @@ struct detailed_pixel_timing { u8 hsync_pulse_width_lo; u8 vsync_pulse_width_lo:4; u8 vsync_offset_lo:4; - u8 hsync_pulse_width_hi:2; - u8 hsync_offset_hi:2; u8 vsync_pulse_width_hi:2; u8 vsync_offset_hi:2; + u8 hsync_pulse_width_hi:2; + u8 hsync_offset_hi:2; u8 width_mm_lo; u8 height_mm_lo; u8 height_mm_hi:4; -- cgit v0.10.2 From 7bec756c74b1a5079d5074144bb77a6b3e7d7783 Mon Sep 17 00:00:00 2001 From: Jesse Barnes Date: Mon, 23 Feb 2009 16:09:34 -0800 Subject: drm: disable encoders before re-routing them MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit In some cases we may receive a mode config that has a different CRTC<->encoder map that the current configuration. In that case, we need to disable any re-routed encoders before setting the mode, otherwise they may not pick up the new CRTC (if the output types are incompatible for example). Tested-by: Kristian Høgsberg Signed-off-by: Jesse Barnes Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/drm_crtc_helper.c b/drivers/gpu/drm/drm_crtc_helper.c index 733028b..1c3a8c5 100644 --- a/drivers/gpu/drm/drm_crtc_helper.c +++ b/drivers/gpu/drm/drm_crtc_helper.c @@ -452,6 +452,59 @@ static void drm_setup_crtcs(struct drm_device *dev) kfree(modes); kfree(enabled); } + +/** + * drm_encoder_crtc_ok - can a given crtc drive a given encoder? + * @encoder: encoder to test + * @crtc: crtc to test + * + * Return false if @encoder can't be driven by @crtc, true otherwise. + */ +static bool drm_encoder_crtc_ok(struct drm_encoder *encoder, + struct drm_crtc *crtc) +{ + struct drm_device *dev; + struct drm_crtc *tmp; + int crtc_mask = 1; + + WARN(!crtc, "checking null crtc?"); + + dev = crtc->dev; + + list_for_each_entry(tmp, &dev->mode_config.crtc_list, head) { + if (tmp == crtc) + break; + crtc_mask <<= 1; + } + + if (encoder->possible_crtcs & crtc_mask) + return true; + return false; +} + +/* + * Check the CRTC we're going to map each output to vs. its current + * CRTC. If they don't match, we have to disable the output and the CRTC + * since the driver will have to re-route things. + */ +static void +drm_crtc_prepare_encoders(struct drm_device *dev) +{ + struct drm_encoder_helper_funcs *encoder_funcs; + struct drm_encoder *encoder; + + list_for_each_entry(encoder, &dev->mode_config.encoder_list, head) { + encoder_funcs = encoder->helper_private; + /* Disable unused encoders */ + if (encoder->crtc == NULL) + (*encoder_funcs->dpms)(encoder, DRM_MODE_DPMS_OFF); + /* Disable encoders whose CRTC is about to change */ + if (encoder_funcs->get_crtc && + encoder->crtc != (*encoder_funcs->get_crtc)(encoder)) + (*encoder_funcs->dpms)(encoder, DRM_MODE_DPMS_OFF); + } +} + /** * drm_crtc_set_mode - set a mode * @crtc: CRTC to program @@ -547,6 +600,8 @@ bool drm_crtc_helper_set_mode(struct drm_crtc *crtc, encoder_funcs->prepare(encoder); } + drm_crtc_prepare_encoders(dev); + crtc_funcs->prepare(crtc); /* Set up the DPLL and any encoders state that needs to adjust or depend @@ -617,7 +672,7 @@ int drm_crtc_helper_set_config(struct drm_mode_set *set) struct drm_device *dev; struct drm_crtc **save_crtcs, *new_crtc; struct drm_encoder **save_encoders, *new_encoder; - struct drm_framebuffer *old_fb; + struct drm_framebuffer *old_fb = NULL; bool save_enabled; bool mode_changed = false; bool fb_changed = false; @@ -668,9 +723,10 @@ int drm_crtc_helper_set_config(struct drm_mode_set *set) * and then just flip_or_move it */ if (set->crtc->fb != set->fb) { /* If we have no fb then treat it as a full mode set */ - if (set->crtc->fb == NULL) + if (set->crtc->fb == NULL) { + DRM_DEBUG("crtc has no fb, full mode set\n"); mode_changed = true; - else if ((set->fb->bits_per_pixel != + } else if ((set->fb->bits_per_pixel != set->crtc->fb->bits_per_pixel) || set->fb->depth != set->crtc->fb->depth) fb_changed = true; @@ -682,7 +738,7 @@ int drm_crtc_helper_set_config(struct drm_mode_set *set) fb_changed = true; if (set->mode && !drm_mode_equal(set->mode, &set->crtc->mode)) { - DRM_DEBUG("modes are different\n"); + DRM_DEBUG("modes are different, full mode set\n"); drm_mode_debug_printmodeline(&set->crtc->mode); drm_mode_debug_printmodeline(set->mode); mode_changed = true; @@ -708,6 +764,7 @@ int drm_crtc_helper_set_config(struct drm_mode_set *set) } if (new_encoder != connector->encoder) { + DRM_DEBUG("encoder changed, full mode switch\n"); mode_changed = true; connector->encoder = new_encoder; } @@ -734,10 +791,20 @@ int drm_crtc_helper_set_config(struct drm_mode_set *set) if (set->connectors[ro] == connector) new_crtc = set->crtc; } + + /* Make sure the new CRTC will work with the encoder */ + if (new_crtc && + !drm_encoder_crtc_ok(connector->encoder, new_crtc)) { + ret = -EINVAL; + goto fail_set_mode; + } if (new_crtc != connector->encoder->crtc) { + DRM_DEBUG("crtc changed, full mode switch\n"); mode_changed = true; connector->encoder->crtc = new_crtc; } + DRM_DEBUG("setting connector %d crtc to %p\n", + connector->base.id, new_crtc); } /* mode_set_base is not a required function */ @@ -781,6 +848,7 @@ int drm_crtc_helper_set_config(struct drm_mode_set *set) fail_set_mode: set->crtc->enabled = save_enabled; + set->crtc->fb = old_fb; count = 0; list_for_each_entry(connector, &dev->mode_config.connector_list, head) { if (!connector->encoder) diff --git a/include/drm/drm_crtc_helper.h b/include/drm/drm_crtc_helper.h index 0b0d236..c7d4b2e 100644 --- a/include/drm/drm_crtc_helper.h +++ b/include/drm/drm_crtc_helper.h @@ -76,6 +76,7 @@ struct drm_encoder_helper_funcs { void (*mode_set)(struct drm_encoder *encoder, struct drm_display_mode *mode, struct drm_display_mode *adjusted_mode); + struct drm_crtc *(*get_crtc)(struct drm_encoder *encoder); /* detect for DAC style encoders */ enum drm_connector_status (*detect)(struct drm_encoder *encoder, struct drm_connector *connector); -- cgit v0.10.2 From b3f5e7329df1a508ac58ebe7509fb7a47b9eab6a Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Thu, 19 Feb 2009 14:48:22 +0000 Subject: drm: Correct unbalanced drm_vblank_put() during mode setting. The first time we install a mode, the vblank will be disabled for a pipe and so drm_vblank_get() in drm_vblank_pre_modeset() will fail. As we unconditionally call drm_vblank_put() afterwards, the vblank reference counter becomes unbalanced. Signed-off-by: Chris Wilson Acked-by: Jesse Barnes Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/drm_irq.c b/drivers/gpu/drm/drm_irq.c index 3795dbc..93e677a 100644 --- a/drivers/gpu/drm/drm_irq.c +++ b/drivers/gpu/drm/drm_irq.c @@ -435,6 +435,8 @@ EXPORT_SYMBOL(drm_vblank_get); */ void drm_vblank_put(struct drm_device *dev, int crtc) { + BUG_ON (atomic_read (&dev->vblank_refcount[crtc]) == 0); + /* Last user schedules interrupt disable */ if (atomic_dec_and_test(&dev->vblank_refcount[crtc])) mod_timer(&dev->vblank_disable_timer, jiffies + 5*DRM_HZ); @@ -460,8 +462,9 @@ void drm_vblank_pre_modeset(struct drm_device *dev, int crtc) * so that interrupts remain enabled in the interim. */ if (!dev->vblank_inmodeset[crtc]) { - dev->vblank_inmodeset[crtc] = 1; - drm_vblank_get(dev, crtc); + dev->vblank_inmodeset[crtc] = 0x1; + if (drm_vblank_get(dev, crtc) == 0) + dev->vblank_inmodeset[crtc] |= 0x2; } } EXPORT_SYMBOL(drm_vblank_pre_modeset); @@ -473,9 +476,12 @@ void drm_vblank_post_modeset(struct drm_device *dev, int crtc) if (dev->vblank_inmodeset[crtc]) { spin_lock_irqsave(&dev->vbl_lock, irqflags); dev->vblank_disable_allowed = 1; - dev->vblank_inmodeset[crtc] = 0; spin_unlock_irqrestore(&dev->vbl_lock, irqflags); - drm_vblank_put(dev, crtc); + + if (dev->vblank_inmodeset[crtc] & 0x2) + drm_vblank_put(dev, crtc); + + dev->vblank_inmodeset[crtc] = 0; } } EXPORT_SYMBOL(drm_vblank_post_modeset); -- cgit v0.10.2 From d61e7380b402a481ab1fa8027068a24918f701c8 Mon Sep 17 00:00:00 2001 From: Kyle McMartin Date: Tue, 24 Feb 2009 20:31:53 -0500 Subject: drm: edid revision 0 is valid edid->revision == 0 should be valid (at least, so the error message indicates. :) and wikipedia seems to indicate that EDID 1.0 existed. We can dump the entire check, since edid->revision is a u8, so it can't ever be less than 0. Marko reports in RH bz#476735 that his monitor claims to be EDID 1.0, and therefore hits the check and is stuck at 800x600 because of it. Reported-by: Marko Ristola Signed-off-by: Kyle McMartin Acked-by: Jesse Barnes Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/drm_edid.c b/drivers/gpu/drm/drm_edid.c index e902b1c..a839a28 100644 --- a/drivers/gpu/drm/drm_edid.c +++ b/drivers/gpu/drm/drm_edid.c @@ -125,7 +125,7 @@ static bool edid_is_valid(struct edid *edid) DRM_ERROR("EDID has major version %d, instead of 1\n", edid->version); goto bad; } - if (edid->revision <= 0 || edid->revision > 3) { + if (edid->revision > 3) { DRM_ERROR("EDID has minor version %d, which is not between 0-3\n", edid->revision); goto bad; } -- cgit v0.10.2 From dd0910b3c71b253c08111110f0399b924a8d5853 Mon Sep 17 00:00:00 2001 From: Dave Airlie Date: Wed, 25 Feb 2009 14:49:21 +1000 Subject: drm/i915: make hw page ioremap use ioremap_wc However we still have another issue with ioremap_wc not falling back properly or somehow doing something else stupid, this probably needs to be tracked down. Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/i915/i915_dma.c b/drivers/gpu/drm/i915/i915_dma.c index 2d797ff..cc46495 100644 --- a/drivers/gpu/drm/i915/i915_dma.c +++ b/drivers/gpu/drm/i915/i915_dma.c @@ -811,7 +811,7 @@ static int i915_set_status_page(struct drm_device *dev, void *data, dev_priv->hws_map.flags = 0; dev_priv->hws_map.mtrr = 0; - drm_core_ioremap(&dev_priv->hws_map, dev); + drm_core_ioremap_wc(&dev_priv->hws_map, dev); if (dev_priv->hws_map.handle == NULL) { i915_dma_cleanup(dev); dev_priv->status_gfx_addr = 0; -- cgit v0.10.2 From e08fb4f6d1dc95eff5b3fc1d0412bcb5afcae7f2 Mon Sep 17 00:00:00 2001 From: Dave Airlie Date: Wed, 25 Feb 2009 14:52:30 +1000 Subject: drm/i915: convert DRM_ERROR to DRM_DEBUG in phys object pwrite path This snuck in when I wrote phys object support. Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c index 28b726d..85685bf 100644 --- a/drivers/gpu/drm/i915/i915_gem.c +++ b/drivers/gpu/drm/i915/i915_gem.c @@ -3548,7 +3548,7 @@ i915_gem_phys_pwrite(struct drm_device *dev, struct drm_gem_object *obj, user_data = (char __user *) (uintptr_t) args->data_ptr; obj_addr = obj_priv->phys_obj->handle->vaddr + args->offset; - DRM_ERROR("obj_addr %p, %lld\n", obj_addr, args->size); + DRM_DEBUG("obj_addr %p, %lld\n", obj_addr, args->size); ret = copy_from_user(obj_addr, user_data, args->size); if (ret) return -EFAULT; -- cgit v0.10.2 From fef7cc0893146550b286b13c0e6e914556142730 Mon Sep 17 00:00:00 2001 From: Greg Kroah-Hartman Date: Tue, 24 Feb 2009 23:52:24 -0800 Subject: asix: new device ids This patch adds two new device ids to the asix driver. One comes directly from the asix driver on their web site, the other was reported by Armani Liao as needed for the MSI X320 to get the driver to work properly for it. Reported-by: Armani Liao Signed-off-by: Greg Kroah-Hartman Signed-off-by: David S. Miller diff --git a/drivers/net/usb/asix.c b/drivers/net/usb/asix.c index e009481..396f821 100644 --- a/drivers/net/usb/asix.c +++ b/drivers/net/usb/asix.c @@ -1451,6 +1451,14 @@ static const struct usb_device_id products [] = { // Cables-to-Go USB Ethernet Adapter USB_DEVICE(0x0b95, 0x772a), .driver_info = (unsigned long) &ax88772_info, +}, { + // ABOCOM for pci + USB_DEVICE(0x14ea, 0xab11), + .driver_info = (unsigned long) &ax88178_info, +}, { + // ASIX 88772a + USB_DEVICE(0x0db0, 0xa877), + .driver_info = (unsigned long) &ax88772_info, }, { }, // END }; -- cgit v0.10.2 From 6aa03ab06978e97b3e0720f83280d7841051916b Mon Sep 17 00:00:00 2001 From: Fenghua Yu Date: Wed, 25 Feb 2009 14:06:26 +0900 Subject: Fix iwlan DMA mapping direction When iwlan runs on IOMMU, IOMMU generates a lot of PTE write faults because PTE write bit is not set on some of PTE's. This is because iwlan driver calls DMA mapping with PCI_DMA_TODEVICE which is read only in mapping PTE. But iwlan device actually writes to the mapped page to update its contents. This issue is not exposed in swiotlb. But VT-d hardware can capture this fault and stop the fault transaction. The following patch fixes the issue. Signed-off-by: Fenghua Yu Reviewed-by: Bhavesh Davda Tested-by: Chris Wright Acked-by: Tomas Winkler Signed-off-by: David Woodhouse Signed-off-by: Linus Torvalds diff --git a/drivers/net/wireless/iwlwifi/iwl-tx.c b/drivers/net/wireless/iwlwifi/iwl-tx.c index b0ee86c..ab13ff2 100644 --- a/drivers/net/wireless/iwlwifi/iwl-tx.c +++ b/drivers/net/wireless/iwlwifi/iwl-tx.c @@ -148,7 +148,7 @@ static void iwl_hw_txq_free_tfd(struct iwl_priv *priv, struct iwl_tx_queue *txq) pci_unmap_single(dev, pci_unmap_addr(&txq->cmd[index]->meta, mapping), pci_unmap_len(&txq->cmd[index]->meta, len), - PCI_DMA_TODEVICE); + PCI_DMA_BIDIRECTIONAL); /* Unmap chunks, if any. */ for (i = 1; i < num_tbs; i++) { @@ -964,7 +964,7 @@ int iwl_tx_skb(struct iwl_priv *priv, struct sk_buff *skb) * within command buffer array. */ txcmd_phys = pci_map_single(priv->pci_dev, out_cmd, sizeof(struct iwl_cmd), - PCI_DMA_TODEVICE); + PCI_DMA_BIDIRECTIONAL); pci_unmap_addr_set(&out_cmd->meta, mapping, txcmd_phys); pci_unmap_len_set(&out_cmd->meta, len, sizeof(struct iwl_cmd)); /* Add buffer containing Tx command and MAC(!) header to TFD's @@ -1115,7 +1115,7 @@ int iwl_enqueue_hcmd(struct iwl_priv *priv, struct iwl_host_cmd *cmd) IWL_MAX_SCAN_SIZE : sizeof(struct iwl_cmd); phys_addr = pci_map_single(priv->pci_dev, out_cmd, - len, PCI_DMA_TODEVICE); + len, PCI_DMA_BIDIRECTIONAL); pci_unmap_addr_set(&out_cmd->meta, mapping, phys_addr); pci_unmap_len_set(&out_cmd->meta, len, len); phys_addr += offsetof(struct iwl_cmd, hdr); @@ -1212,7 +1212,7 @@ static void iwl_hcmd_queue_reclaim(struct iwl_priv *priv, int txq_id, pci_unmap_single(priv->pci_dev, pci_unmap_addr(&txq->cmd[cmd_idx]->meta, mapping), pci_unmap_len(&txq->cmd[cmd_idx]->meta, len), - PCI_DMA_TODEVICE); + PCI_DMA_BIDIRECTIONAL); for (idx = iwl_queue_inc_wrap(idx, q->n_bd); q->read_ptr != idx; q->read_ptr = iwl_queue_inc_wrap(q->read_ptr, q->n_bd)) { -- cgit v0.10.2 From 0af80c04e2f2e45ae09fceb17df8050f828a5c40 Mon Sep 17 00:00:00 2001 From: David Fries Date: Wed, 25 Feb 2009 20:28:21 +0100 Subject: ide: ide.c 'clear' fix, update "ide=nodma" documentation Documentation/kernel-parameters.txt - ide=nodma is no longer valid. drivers/ide/Kconfig - The module is ide-core.ko not ide. drivers/ide/ide.c - It took me a while to figure out what the arguments %d.%d:%d to nodma module parameter ment, so I added a comment to each. - Added a comment to each of the sscanf lines. - There is a bug, if j is 0 it would previously clear all the other bits except the current device, changed in three different places. mask &= (1 << i) should be mask &= ~(1 << i). Signed-off-by: David Fries [bart: s/disk/device/ in ide.c, beautify patch description] Signed-off-by: Bartlomiej Zolnierkiewicz diff --git a/Documentation/kernel-parameters.txt b/Documentation/kernel-parameters.txt index 319785b..0ed3234 100644 --- a/Documentation/kernel-parameters.txt +++ b/Documentation/kernel-parameters.txt @@ -868,8 +868,10 @@ and is between 256 and 4096 characters. It is defined in the file icn= [HW,ISDN] Format: [,[,[,]]] - ide= [HW] (E)IDE subsystem - Format: ide=nodma or ide=doubler + ide-core.nodma= [HW] (E)IDE subsystem + Format: =0.0 to prevent dma on hda, =0.1 hdb =1.0 hdc + .vlb_clock .pci_clock .noflush .noprobe .nowerr .cdrom + .chs .ignore_cable are additional options See Documentation/ide/ide.txt. idebus= [HW] (E)IDE subsystem - VLB/PCI bus speed diff --git a/drivers/ide/Kconfig b/drivers/ide/Kconfig index 3dad229..e072903 100644 --- a/drivers/ide/Kconfig +++ b/drivers/ide/Kconfig @@ -46,7 +46,7 @@ menuconfig IDE SMART parameters from disk drives. To compile this driver as a module, choose M here: the - module will be called ide. + module will be called ide-core.ko. For further information, please read . diff --git a/drivers/ide/ide.c b/drivers/ide/ide.c index 258805d..0920e3b 100644 --- a/drivers/ide/ide.c +++ b/drivers/ide/ide.c @@ -337,6 +337,7 @@ static int ide_set_dev_param_mask(const char *s, struct kernel_param *kp) int a, b, i, j = 1; unsigned int *dev_param_mask = (unsigned int *)kp->arg; + /* controller . device (0 or 1) [ : 1 (set) | 0 (clear) ] */ if (sscanf(s, "%d.%d:%d", &a, &b, &j) != 3 && sscanf(s, "%d.%d", &a, &b) != 2) return -EINVAL; @@ -349,7 +350,7 @@ static int ide_set_dev_param_mask(const char *s, struct kernel_param *kp) if (j) *dev_param_mask |= (1 << i); else - *dev_param_mask &= (1 << i); + *dev_param_mask &= ~(1 << i); return 0; } @@ -392,6 +393,8 @@ static int ide_set_disk_chs(const char *str, struct kernel_param *kp) { int a, b, c = 0, h = 0, s = 0, i, j = 1; + /* controller . device (0 or 1) : Cylinders , Heads , Sectors */ + /* controller . device (0 or 1) : 1 (use CHS) | 0 (ignore CHS) */ if (sscanf(str, "%d.%d:%d,%d,%d", &a, &b, &c, &h, &s) != 5 && sscanf(str, "%d.%d:%d", &a, &b, &j) != 3) return -EINVAL; @@ -407,7 +410,7 @@ static int ide_set_disk_chs(const char *str, struct kernel_param *kp) if (j) ide_disks |= (1 << i); else - ide_disks &= (1 << i); + ide_disks &= ~(1 << i); ide_disks_chs[i].cyl = c; ide_disks_chs[i].head = h; @@ -469,6 +472,8 @@ static int ide_set_ignore_cable(const char *s, struct kernel_param *kp) { int i, j = 1; + /* controller (ignore) */ + /* controller : 1 (ignore) | 0 (use) */ if (sscanf(s, "%d:%d", &i, &j) != 2 && sscanf(s, "%d", &i) != 1) return -EINVAL; @@ -478,7 +483,7 @@ static int ide_set_ignore_cable(const char *s, struct kernel_param *kp) if (j) ide_ignore_cable |= (1 << i); else - ide_ignore_cable &= (1 << i); + ide_ignore_cable &= ~(1 << i); return 0; } -- cgit v0.10.2 From 43a12216d3664a9fa6c8ceb398da6ef08fee7ff7 Mon Sep 17 00:00:00 2001 From: Roel Kluin Date: Wed, 25 Feb 2009 20:28:22 +0100 Subject: amd74xx: device/vendor confusion Device and vendor ids were confused Signed-off-by: Roel Kluin Cc: Andrew Morton Acked-by: Sergei Shtylyov Signed-off-by: Bartlomiej Zolnierkiewicz diff --git a/drivers/ide/amd74xx.c b/drivers/ide/amd74xx.c index 69660a4..77267c8 100644 --- a/drivers/ide/amd74xx.c +++ b/drivers/ide/amd74xx.c @@ -166,7 +166,7 @@ static unsigned int init_chipset_amd74xx(struct pci_dev *dev) * Check for broken FIFO support. */ if (dev->vendor == PCI_VENDOR_ID_AMD && - dev->vendor == PCI_DEVICE_ID_AMD_VIPER_7411) + dev->device == PCI_DEVICE_ID_AMD_VIPER_7411) t &= 0x0f; else t |= 0xf0; -- cgit v0.10.2 From f76bee16fc83f58d6c1b088977330f26ed7ae248 Mon Sep 17 00:00:00 2001 From: Roel Kluin Date: Wed, 25 Feb 2009 20:28:22 +0100 Subject: atiixp: fix missing parentheses Fix missing parentheses so PIO/DMA timings for master device on the second channel are programmed correctly (IOW "8 0 24 16" offset values should be used instead of the current "8 0 16 16"). [ The bug went unnoticed because after PIO/DMA timings get programmed incorrectly for the third device they are overwritten with timings for the fourth device and since BIOS should also program timings for the third device everything should work fine until suspend/resume cycle or user requested transfer mode changes. ] Signed-off-by: Roel Kluin Cc: Sergei Shtylyov Cc: Andrew Morton [bart: update patch description] Signed-off-by: Bartlomiej Zolnierkiewicz diff --git a/drivers/ide/atiixp.c b/drivers/ide/atiixp.c index b2735d2..ecd1e62 100644 --- a/drivers/ide/atiixp.c +++ b/drivers/ide/atiixp.c @@ -52,7 +52,7 @@ static void atiixp_set_pio_mode(ide_drive_t *drive, const u8 pio) { struct pci_dev *dev = to_pci_dev(drive->hwif->dev); unsigned long flags; - int timing_shift = (drive->dn & 2) ? 16 : 0 + (drive->dn & 1) ? 0 : 8; + int timing_shift = (drive->dn ^ 1) * 8; u32 pio_timing_data; u16 pio_mode_data; @@ -85,7 +85,7 @@ static void atiixp_set_dma_mode(ide_drive_t *drive, const u8 speed) { struct pci_dev *dev = to_pci_dev(drive->hwif->dev); unsigned long flags; - int timing_shift = (drive->dn & 2) ? 16 : 0 + (drive->dn & 1) ? 0 : 8; + int timing_shift = (drive->dn ^ 1) * 8; u32 tmp32; u16 tmp16; u16 udma_ctl = 0; -- cgit v0.10.2 From f38344b0a0898d2a8c13581ee61007719e16e1d7 Mon Sep 17 00:00:00 2001 From: Bartlomiej Zolnierkiewicz Date: Wed, 25 Feb 2009 20:28:22 +0100 Subject: it821x: remove dead URL Signed-off-by: Bartlomiej Zolnierkiewicz diff --git a/drivers/ide/it821x.c b/drivers/ide/it821x.c index e1c4f54..13b8153 100644 --- a/drivers/ide/it821x.c +++ b/drivers/ide/it821x.c @@ -5,9 +5,8 @@ * May be copied or modified under the terms of the GNU General Public License * Based in part on the ITE vendor provided SCSI driver. * - * Documentation available from - * http://www.ite.com.tw/pc/IT8212F_V04.pdf - * Some other documents are NDA. + * Documentation: + * Datasheet is freely available, some other documents under NDA. * * The ITE8212 isn't exactly a standard IDE controller. It has two * modes. In pass through mode then it is an IDE controller. In its smart -- cgit v0.10.2 From d3dd7107f4d843d0f01d0f77d49a7c5449130577 Mon Sep 17 00:00:00 2001 From: Bartlomiej Zolnierkiewicz Date: Wed, 25 Feb 2009 20:28:23 +0100 Subject: ide-cd: document capacity hack Just copy the comment from drivers/scsi/sr.c::sr_done() (from which the capacity hack has been originated). Cc: Borislav Petkov Acked-by: Sergei Shtylyov Signed-off-by: Bartlomiej Zolnierkiewicz diff --git a/drivers/ide/ide-cd.c b/drivers/ide/ide-cd.c index 0bfeb0c..690475b 100644 --- a/drivers/ide/ide-cd.c +++ b/drivers/ide/ide-cd.c @@ -194,6 +194,14 @@ static void cdrom_analyze_sense_data(ide_drive_t *drive, bio_sectors = max(bio_sectors(failed_command->bio), 4U); sector &= ~(bio_sectors - 1); + /* + * The SCSI specification allows for the value + * returned by READ CAPACITY to be up to 75 2K + * sectors past the last readable block. + * Therefore, if we hit a medium error within the + * last 75 2K sectors, we decrease the saved size + * value. + */ if (sector < get_capacity(info->disk) && drive->probed_capacity - sector < 4 * 75) set_capacity(info->disk, sector); -- cgit v0.10.2 From 8fed43684174b68f04d01d1210fd00536af790df Mon Sep 17 00:00:00 2001 From: Bartlomiej Zolnierkiewicz Date: Wed, 25 Feb 2009 20:28:24 +0100 Subject: ide: fix refcounting in device drivers During host driver module removal del_gendisk() results in a final put on drive->gendev and freeing the drive by drive_release_dev(). Convert device drivers from using struct kref to use struct device so device driver's object holds reference on ->gendev and prevents drive from prematurely going away. Also fix ->remove methods to not erroneously drop reference on a host driver by using only put_device() instead of ide*_put(). Reported-by: Stanislaw Gruszka Tested-by: Stanislaw Gruszka Signed-off-by: Bartlomiej Zolnierkiewicz diff --git a/drivers/ide/ide-cd.c b/drivers/ide/ide-cd.c index 690475b..ddfbea4 100644 --- a/drivers/ide/ide-cd.c +++ b/drivers/ide/ide-cd.c @@ -55,7 +55,7 @@ static DEFINE_MUTEX(idecd_ref_mutex); -static void ide_cd_release(struct kref *); +static void ide_cd_release(struct device *); static struct cdrom_info *ide_cd_get(struct gendisk *disk) { @@ -67,7 +67,7 @@ static struct cdrom_info *ide_cd_get(struct gendisk *disk) if (ide_device_get(cd->drive)) cd = NULL; else - kref_get(&cd->kref); + get_device(&cd->dev); } mutex_unlock(&idecd_ref_mutex); @@ -79,7 +79,7 @@ static void ide_cd_put(struct cdrom_info *cd) ide_drive_t *drive = cd->drive; mutex_lock(&idecd_ref_mutex); - kref_put(&cd->kref, ide_cd_release); + put_device(&cd->dev); ide_device_put(drive); mutex_unlock(&idecd_ref_mutex); } @@ -1798,15 +1798,17 @@ static void ide_cd_remove(ide_drive_t *drive) ide_debug_log(IDE_DBG_FUNC, "Call %s\n", __func__); ide_proc_unregister_driver(drive, info->driver); - + device_del(&info->dev); del_gendisk(info->disk); - ide_cd_put(info); + mutex_lock(&idecd_ref_mutex); + put_device(&info->dev); + mutex_unlock(&idecd_ref_mutex); } -static void ide_cd_release(struct kref *kref) +static void ide_cd_release(struct device *dev) { - struct cdrom_info *info = to_ide_drv(kref, cdrom_info); + struct cdrom_info *info = to_ide_drv(dev, cdrom_info); struct cdrom_device_info *devinfo = &info->devinfo; ide_drive_t *drive = info->drive; struct gendisk *g = info->disk; @@ -2005,7 +2007,12 @@ static int ide_cd_probe(ide_drive_t *drive) ide_init_disk(g, drive); - kref_init(&info->kref); + info->dev.parent = &drive->gendev; + info->dev.release = ide_cd_release; + dev_set_name(&info->dev, dev_name(&drive->gendev)); + + if (device_register(&info->dev)) + goto out_free_disk; info->drive = drive; info->driver = &ide_cdrom_driver; @@ -2019,7 +2026,7 @@ static int ide_cd_probe(ide_drive_t *drive) g->driverfs_dev = &drive->gendev; g->flags = GENHD_FL_CD | GENHD_FL_REMOVABLE; if (ide_cdrom_setup(drive)) { - ide_cd_release(&info->kref); + put_device(&info->dev); goto failed; } @@ -2029,6 +2036,8 @@ static int ide_cd_probe(ide_drive_t *drive) add_disk(g); return 0; +out_free_disk: + put_disk(g); out_free_cd: kfree(info); failed: diff --git a/drivers/ide/ide-cd.h b/drivers/ide/ide-cd.h index ac40d6c..c878bfc 100644 --- a/drivers/ide/ide-cd.h +++ b/drivers/ide/ide-cd.h @@ -80,7 +80,7 @@ struct cdrom_info { ide_drive_t *drive; struct ide_driver *driver; struct gendisk *disk; - struct kref kref; + struct device dev; /* Buffer for table of contents. NULL if we haven't allocated a TOC buffer for this device yet. */ diff --git a/drivers/ide/ide-gd.c b/drivers/ide/ide-gd.c index 7857b20..0471094 100644 --- a/drivers/ide/ide-gd.c +++ b/drivers/ide/ide-gd.c @@ -25,7 +25,7 @@ module_param(debug_mask, ulong, 0644); static DEFINE_MUTEX(ide_disk_ref_mutex); -static void ide_disk_release(struct kref *); +static void ide_disk_release(struct device *); static struct ide_disk_obj *ide_disk_get(struct gendisk *disk) { @@ -37,7 +37,7 @@ static struct ide_disk_obj *ide_disk_get(struct gendisk *disk) if (ide_device_get(idkp->drive)) idkp = NULL; else - kref_get(&idkp->kref); + get_device(&idkp->dev); } mutex_unlock(&ide_disk_ref_mutex); return idkp; @@ -48,7 +48,7 @@ static void ide_disk_put(struct ide_disk_obj *idkp) ide_drive_t *drive = idkp->drive; mutex_lock(&ide_disk_ref_mutex); - kref_put(&idkp->kref, ide_disk_release); + put_device(&idkp->dev); ide_device_put(drive); mutex_unlock(&ide_disk_ref_mutex); } @@ -66,17 +66,18 @@ static void ide_gd_remove(ide_drive_t *drive) struct gendisk *g = idkp->disk; ide_proc_unregister_driver(drive, idkp->driver); - + device_del(&idkp->dev); del_gendisk(g); - drive->disk_ops->flush(drive); - ide_disk_put(idkp); + mutex_lock(&ide_disk_ref_mutex); + put_device(&idkp->dev); + mutex_unlock(&ide_disk_ref_mutex); } -static void ide_disk_release(struct kref *kref) +static void ide_disk_release(struct device *dev) { - struct ide_disk_obj *idkp = to_ide_drv(kref, ide_disk_obj); + struct ide_disk_obj *idkp = to_ide_drv(dev, ide_disk_obj); ide_drive_t *drive = idkp->drive; struct gendisk *g = idkp->disk; @@ -348,7 +349,12 @@ static int ide_gd_probe(ide_drive_t *drive) ide_init_disk(g, drive); - kref_init(&idkp->kref); + idkp->dev.parent = &drive->gendev; + idkp->dev.release = ide_disk_release; + dev_set_name(&idkp->dev, dev_name(&drive->gendev)); + + if (device_register(&idkp->dev)) + goto out_free_disk; idkp->drive = drive; idkp->driver = &ide_gd_driver; @@ -373,6 +379,8 @@ static int ide_gd_probe(ide_drive_t *drive) add_disk(g); return 0; +out_free_disk: + put_disk(g); out_free_idkp: kfree(idkp); failed: diff --git a/drivers/ide/ide-gd.h b/drivers/ide/ide-gd.h index a86779f..b604bdd 100644 --- a/drivers/ide/ide-gd.h +++ b/drivers/ide/ide-gd.h @@ -17,7 +17,7 @@ struct ide_disk_obj { ide_drive_t *drive; struct ide_driver *driver; struct gendisk *disk; - struct kref kref; + struct device dev; unsigned int openers; /* protected by BKL for now */ /* Last failed packet command */ diff --git a/drivers/ide/ide-tape.c b/drivers/ide/ide-tape.c index d7ecd3c..bb450a7 100644 --- a/drivers/ide/ide-tape.c +++ b/drivers/ide/ide-tape.c @@ -169,7 +169,7 @@ typedef struct ide_tape_obj { ide_drive_t *drive; struct ide_driver *driver; struct gendisk *disk; - struct kref kref; + struct device dev; /* * failed_pc points to the last failed packet command, or contains @@ -267,7 +267,7 @@ static DEFINE_MUTEX(idetape_ref_mutex); static struct class *idetape_sysfs_class; -static void ide_tape_release(struct kref *); +static void ide_tape_release(struct device *); static struct ide_tape_obj *ide_tape_get(struct gendisk *disk) { @@ -279,7 +279,7 @@ static struct ide_tape_obj *ide_tape_get(struct gendisk *disk) if (ide_device_get(tape->drive)) tape = NULL; else - kref_get(&tape->kref); + get_device(&tape->dev); } mutex_unlock(&idetape_ref_mutex); return tape; @@ -290,7 +290,7 @@ static void ide_tape_put(struct ide_tape_obj *tape) ide_drive_t *drive = tape->drive; mutex_lock(&idetape_ref_mutex); - kref_put(&tape->kref, ide_tape_release); + put_device(&tape->dev); ide_device_put(drive); mutex_unlock(&idetape_ref_mutex); } @@ -308,7 +308,7 @@ static struct ide_tape_obj *ide_tape_chrdev_get(unsigned int i) mutex_lock(&idetape_ref_mutex); tape = idetape_devs[i]; if (tape) - kref_get(&tape->kref); + get_device(&tape->dev); mutex_unlock(&idetape_ref_mutex); return tape; } @@ -2256,15 +2256,17 @@ static void ide_tape_remove(ide_drive_t *drive) idetape_tape_t *tape = drive->driver_data; ide_proc_unregister_driver(drive, tape->driver); - + device_del(&tape->dev); ide_unregister_region(tape->disk); - ide_tape_put(tape); + mutex_lock(&idetape_ref_mutex); + put_device(&tape->dev); + mutex_unlock(&idetape_ref_mutex); } -static void ide_tape_release(struct kref *kref) +static void ide_tape_release(struct device *dev) { - struct ide_tape_obj *tape = to_ide_drv(kref, ide_tape_obj); + struct ide_tape_obj *tape = to_ide_drv(dev, ide_tape_obj); ide_drive_t *drive = tape->drive; struct gendisk *g = tape->disk; @@ -2407,7 +2409,12 @@ static int ide_tape_probe(ide_drive_t *drive) ide_init_disk(g, drive); - kref_init(&tape->kref); + tape->dev.parent = &drive->gendev; + tape->dev.release = ide_tape_release; + dev_set_name(&tape->dev, dev_name(&drive->gendev)); + + if (device_register(&tape->dev)) + goto out_free_disk; tape->drive = drive; tape->driver = &idetape_driver; @@ -2436,6 +2443,8 @@ static int ide_tape_probe(ide_drive_t *drive) return 0; +out_free_disk: + put_disk(g); out_free_tape: kfree(tape); failed: diff --git a/include/linux/ide.h b/include/linux/ide.h index 194da5a..fe235b6 100644 --- a/include/linux/ide.h +++ b/include/linux/ide.h @@ -663,7 +663,7 @@ typedef struct ide_drive_s ide_drive_t; #define to_ide_device(dev) container_of(dev, ide_drive_t, gendev) #define to_ide_drv(obj, cont_type) \ - container_of(obj, struct cont_type, kref) + container_of(obj, struct cont_type, dev) #define ide_drv_g(disk, cont_type) \ container_of((disk)->private_data, struct cont_type, driver) -- cgit v0.10.2 From 6b1ff036d4cde7834ef2f9dbea5747adaaac24e0 Mon Sep 17 00:00:00 2001 From: Kyle McMartin Date: Sat, 14 Feb 2009 04:11:29 -0500 Subject: [IA64] enable setting DMAR on by default The previous commit which introduced the DMAR_DEFAULT_ON setting in drivers/pci/dmar.c neglected to add the ability for ia64 to enable the IOMMU by default. Rectify that mistake, doh! Signed-off-by: Kyle McMartin Signed-off-by: Tony Luck diff --git a/arch/ia64/Kconfig b/arch/ia64/Kconfig index 4eb45c0..153e727 100644 --- a/arch/ia64/Kconfig +++ b/arch/ia64/Kconfig @@ -638,6 +638,17 @@ config DMAR and include PCI device scope covered by these DMA remapping devices. +config DMAR_DEFAULT_ON + def_bool y + prompt "Enable DMA Remapping Devices by default" + depends on DMAR + help + Selecting this option will enable a DMAR device at boot time if + one is found. If this option is not selected, DMAR support can + be enabled by passing intel_iommu=on to the kernel. It is + recommended you say N here while the DMAR code remains + experimental. + endmenu endif -- cgit v0.10.2 From aa2f63c95439a11dfac35c60d9160dcd0189aed3 Mon Sep 17 00:00:00 2001 From: Roel Kluin Date: Sun, 22 Feb 2009 02:33:28 +0100 Subject: [IA64] Do not go beyond ARRAY_SIZE of unw.hash static struct { ... :114 unsigned short hash[UNW_HASH_SIZE]; ... :2152 for (index = 0; index <= UNW_HASH_SIZE; ++index) { This is a bug, isn't it? s/<=/ Signed-off-by: Tony Luck diff --git a/arch/ia64/kernel/unwind.c b/arch/ia64/kernel/unwind.c index 67810b7..b6c0e63 100644 --- a/arch/ia64/kernel/unwind.c +++ b/arch/ia64/kernel/unwind.c @@ -2149,7 +2149,7 @@ unw_remove_unwind_table (void *handle) /* next, remove hash table entries for this table */ - for (index = 0; index <= UNW_HASH_SIZE; ++index) { + for (index = 0; index < UNW_HASH_SIZE; ++index) { tmp = unw.cache + unw.hash[index]; if (unw.hash[index] >= UNW_CACHE_SIZE || tmp->ip < table->start || tmp->ip >= table->end) -- cgit v0.10.2 From 5b5923975f07836fc7a5388f9fa5f459828ae4ee Mon Sep 17 00:00:00 2001 From: Roel Kluin Date: Sat, 21 Feb 2009 23:40:27 +0100 Subject: [IA64] Don't go beyond iosapic_intr_info's arraysize vi arch/ia64/kernel/iosapic.c +142 static struct iosapic_intr_info { ... } iosapic_intr_info[NR_IRQS]; But at line 510 we have: for (i = 0; i <= NR_IRQS; i++) { s/<=/ Signed-off-by: Tony Luck diff --git a/arch/ia64/kernel/iosapic.c b/arch/ia64/kernel/iosapic.c index 5cfd3d9..e131250 100644 --- a/arch/ia64/kernel/iosapic.c +++ b/arch/ia64/kernel/iosapic.c @@ -507,7 +507,7 @@ static int iosapic_find_sharable_irq(unsigned long trigger, unsigned long pol) if (trigger == IOSAPIC_EDGE) return -EINVAL; - for (i = 0; i <= NR_IRQS; i++) { + for (i = 0; i < NR_IRQS; i++) { info = &iosapic_intr_info[i]; if (info->trigger == trigger && info->polarity == pol && (info->dmode == IOSAPIC_FIXED || -- cgit v0.10.2 From 0b0a0806b0d8635e046bf533225a25903b1cddce Mon Sep 17 00:00:00 2001 From: Hugh Dickins Date: Tue, 24 Feb 2009 20:51:52 +0000 Subject: shmem: fix shared anonymous accounting Each time I exit Firefox, /proc/meminfo's Committed_AS goes down almost 400 kB: OVERCOMMIT_NEVER would be allowing overcommits it should prohibit. Commit fc8744adc870a8d4366908221508bb113d8b72ee "Stop playing silly games with the VM_ACCOUNT flag" changed shmem_file_setup() to set the shmem file's VM_ACCOUNT flag according to VM_NORESERVE not being set in the vma flags; but did so only _after_ the shmem_acct_size(flags, size) call which is expected to pre-account a shared anonymous object. It's all clearer if we switch shmem.c over to use VM_NORESERVE throughout in place of !VM_ACCOUNT. But I very nearly sent in a patch which mistakenly removed the accounting from tmpfs files: shmem_get_inode()'s memset was good for not setting VM_ACCOUNT, but now it needs to set VM_NORESERVE. Rather than setting that by default, then perhaps clearing it again in shmem_file_setup(), let's pass it as a flag to shmem_get_inode(): that allows us to remove the #ifdef CONFIG_SHMEM from shmem_file_setup(). Signed-off-by: Hugh Dickins Signed-off-by: Linus Torvalds diff --git a/mm/shmem.c b/mm/shmem.c index 19d566c..4103a23 100644 --- a/mm/shmem.c +++ b/mm/shmem.c @@ -169,13 +169,13 @@ static inline struct shmem_sb_info *SHMEM_SB(struct super_block *sb) */ static inline int shmem_acct_size(unsigned long flags, loff_t size) { - return (flags & VM_ACCOUNT) ? - security_vm_enough_memory_kern(VM_ACCT(size)) : 0; + return (flags & VM_NORESERVE) ? + 0 : security_vm_enough_memory_kern(VM_ACCT(size)); } static inline void shmem_unacct_size(unsigned long flags, loff_t size) { - if (flags & VM_ACCOUNT) + if (!(flags & VM_NORESERVE)) vm_unacct_memory(VM_ACCT(size)); } @@ -187,13 +187,13 @@ static inline void shmem_unacct_size(unsigned long flags, loff_t size) */ static inline int shmem_acct_block(unsigned long flags) { - return (flags & VM_ACCOUNT) ? - 0 : security_vm_enough_memory_kern(VM_ACCT(PAGE_CACHE_SIZE)); + return (flags & VM_NORESERVE) ? + security_vm_enough_memory_kern(VM_ACCT(PAGE_CACHE_SIZE)) : 0; } static inline void shmem_unacct_blocks(unsigned long flags, long pages) { - if (!(flags & VM_ACCOUNT)) + if (flags & VM_NORESERVE) vm_unacct_memory(pages * VM_ACCT(PAGE_CACHE_SIZE)); } @@ -1515,8 +1515,8 @@ static int shmem_mmap(struct file *file, struct vm_area_struct *vma) return 0; } -static struct inode * -shmem_get_inode(struct super_block *sb, int mode, dev_t dev) +static struct inode *shmem_get_inode(struct super_block *sb, int mode, + dev_t dev, unsigned long flags) { struct inode *inode; struct shmem_inode_info *info; @@ -1537,6 +1537,7 @@ shmem_get_inode(struct super_block *sb, int mode, dev_t dev) info = SHMEM_I(inode); memset(info, 0, (char *)inode - (char *)info); spin_lock_init(&info->lock); + info->flags = flags & VM_NORESERVE; INIT_LIST_HEAD(&info->swaplist); switch (mode & S_IFMT) { @@ -1779,9 +1780,10 @@ static int shmem_statfs(struct dentry *dentry, struct kstatfs *buf) static int shmem_mknod(struct inode *dir, struct dentry *dentry, int mode, dev_t dev) { - struct inode *inode = shmem_get_inode(dir->i_sb, mode, dev); + struct inode *inode; int error = -ENOSPC; + inode = shmem_get_inode(dir->i_sb, mode, dev, VM_NORESERVE); if (inode) { error = security_inode_init_security(inode, dir, NULL, NULL, NULL); @@ -1920,7 +1922,7 @@ static int shmem_symlink(struct inode *dir, struct dentry *dentry, const char *s if (len > PAGE_CACHE_SIZE) return -ENAMETOOLONG; - inode = shmem_get_inode(dir->i_sb, S_IFLNK|S_IRWXUGO, 0); + inode = shmem_get_inode(dir->i_sb, S_IFLNK|S_IRWXUGO, 0, VM_NORESERVE); if (!inode) return -ENOSPC; @@ -2332,7 +2334,7 @@ static int shmem_fill_super(struct super_block *sb, sb->s_flags |= MS_POSIXACL; #endif - inode = shmem_get_inode(sb, S_IFDIR | sbinfo->mode, 0); + inode = shmem_get_inode(sb, S_IFDIR | sbinfo->mode, 0, VM_NORESERVE); if (!inode) goto failed; inode->i_uid = sbinfo->uid; @@ -2574,12 +2576,12 @@ int shmem_unuse(swp_entry_t entry, struct page *page) return 0; } -#define shmem_file_operations ramfs_file_operations -#define shmem_vm_ops generic_file_vm_ops -#define shmem_get_inode ramfs_get_inode -#define shmem_acct_size(a, b) 0 -#define shmem_unacct_size(a, b) do {} while (0) -#define SHMEM_MAX_BYTES LLONG_MAX +#define shmem_vm_ops generic_file_vm_ops +#define shmem_file_operations ramfs_file_operations +#define shmem_get_inode(sb, mode, dev, flags) ramfs_get_inode(sb, mode, dev) +#define shmem_acct_size(flags, size) 0 +#define shmem_unacct_size(flags, size) do {} while (0) +#define SHMEM_MAX_BYTES LLONG_MAX #endif /* CONFIG_SHMEM */ @@ -2589,7 +2591,7 @@ int shmem_unuse(swp_entry_t entry, struct page *page) * shmem_file_setup - get an unlinked file living in tmpfs * @name: name for dentry (to be seen in /proc//maps * @size: size to be set for the file - * @flags: vm_flags + * @flags: VM_NORESERVE suppresses pre-accounting of the entire object size */ struct file *shmem_file_setup(char *name, loff_t size, unsigned long flags) { @@ -2623,13 +2625,10 @@ struct file *shmem_file_setup(char *name, loff_t size, unsigned long flags) goto put_dentry; error = -ENOSPC; - inode = shmem_get_inode(root->d_sb, S_IFREG | S_IRWXUGO, 0); + inode = shmem_get_inode(root->d_sb, S_IFREG | S_IRWXUGO, 0, flags); if (!inode) goto close_file; -#ifdef CONFIG_SHMEM - SHMEM_I(inode)->flags = (flags & VM_NORESERVE) ? 0 : VM_ACCOUNT; -#endif d_instantiate(dentry, inode); inode->i_size = size; inode->i_nlink = 0; /* It is unlinked */ -- cgit v0.10.2 From 7ba07d16bd62f931efec1fc8e63bf1aeebfe42a9 Mon Sep 17 00:00:00 2001 From: Ondrej Zary Date: Wed, 11 Feb 2009 13:08:43 -0800 Subject: pata_it821x: resume from hibernation fails with RAID volume Hibernation didn't work for me since I started to use IT8212 controller. I did some debugging (booting with no_console_suspend init=/bin/sh). Found that resume fails (2.6.28) with "serial number mismatch 'some garbage' != 'some other garbage'" and "revalidation failed" messages. That's because the controller firmware fills different serial number in the IDENTIFY every boot. The patch below fixes the resume simply clearing the serial number. The proper fix would be probably to fill in the serial number of the RAID volume instead. I assume that there must be something like that stored on the drives but I don't know where. Fix resume on pata_it821x RAID volume by clearing the serial number in IDENTIFY data, which is otherwise different on each boot. Signed-off-by: Ondrej Zary Acked-by: Alan Cox Signed-off-by: Andrew Morton Signed-off-by: Jeff Garzik diff --git a/drivers/ata/pata_it821x.c b/drivers/ata/pata_it821x.c index f1bb2f9..b05b86a 100644 --- a/drivers/ata/pata_it821x.c +++ b/drivers/ata/pata_it821x.c @@ -557,6 +557,9 @@ static unsigned int it821x_read_id(struct ata_device *adev, id[83] |= 0x4400; /* Word 83 is valid and LBA48 */ id[86] |= 0x0400; /* LBA48 on */ id[ATA_ID_MAJOR_VER] |= 0x1F; + /* Clear the serial number because it's different each boot + which breaks validation on resume */ + memset(&id[ATA_ID_SERNO], 0x20, ATA_ID_SERNO_LEN); } return err_mask; } -- cgit v0.10.2 From 6be96ac15e4d913e1f48299db083ada5321803b2 Mon Sep 17 00:00:00 2001 From: Mark Lord Date: Thu, 19 Feb 2009 10:38:04 -0500 Subject: sata_mv: fix SoC interrupt breakage For some reason, sata_mv doesn't clear interrupt status during init when it's running on an SoC host adapter. If the bootloader has touched the SATA controller before starting Linux, Linux can end up enabling the SATA interrupt with events pending, which will cause the interrupt to be marked as spurious and then be disabled, which then breaks all further accesses to the controller. This patch makes the SoC path clear interrupt status on init like in the non-SoC case. Signed-off-by: Lennert Buytenhek Signed-off-by: Mark Lord Signed-off-by: Jeff Garzik diff --git a/drivers/ata/sata_mv.c b/drivers/ata/sata_mv.c index 4ae1a41..7007edd 100644 --- a/drivers/ata/sata_mv.c +++ b/drivers/ata/sata_mv.c @@ -3114,19 +3114,17 @@ static int mv_init_host(struct ata_host *host, unsigned int board_idx) writelfl(0, hc_mmio + HC_IRQ_CAUSE_OFS); } - if (!IS_SOC(hpriv)) { - /* Clear any currently outstanding host interrupt conditions */ - writelfl(0, mmio + hpriv->irq_cause_ofs); + /* Clear any currently outstanding host interrupt conditions */ + writelfl(0, mmio + hpriv->irq_cause_ofs); - /* and unmask interrupt generation for host regs */ - writelfl(hpriv->unmask_all_irqs, mmio + hpriv->irq_mask_ofs); + /* and unmask interrupt generation for host regs */ + writelfl(hpriv->unmask_all_irqs, mmio + hpriv->irq_mask_ofs); - /* - * enable only global host interrupts for now. - * The per-port interrupts get done later as ports are set up. - */ - mv_set_main_irq_mask(host, 0, PCI_ERR); - } + /* + * enable only global host interrupts for now. + * The per-port interrupts get done later as ports are set up. + */ + mv_set_main_irq_mask(host, 0, PCI_ERR); done: return rc; } -- cgit v0.10.2 From c48052cc36e02fff6a9bb3cf83c4206b9127611f Mon Sep 17 00:00:00 2001 From: Alan Cox Date: Wed, 11 Feb 2009 13:08:41 -0800 Subject: [libata] pata_amd: program FIFO With 32bit PIO we can use the posted write buffers, but only for 32bit I/O cycles. This means we must disable the FIFO for ATAPI where a final 16bit cycle may occur. Rework the FIFO logic so that we disable the FIFO then selectively re-enable it when we set the timings on AMD devices. Also fix a case where we scribbled on PCI config 0x41 of Nvidia chips when we shouldn't. Signed-off-by: Alan Cox Signed-off-by: Andrew Morton Signed-off-by: Jeff Garzik diff --git a/drivers/ata/pata_amd.c b/drivers/ata/pata_amd.c index 63719ab..115b1cd 100644 --- a/drivers/ata/pata_amd.c +++ b/drivers/ata/pata_amd.c @@ -24,7 +24,7 @@ #include #define DRV_NAME "pata_amd" -#define DRV_VERSION "0.3.11" +#define DRV_VERSION "0.4.1" /** * timing_setup - shared timing computation and load @@ -145,6 +145,13 @@ static int amd_pre_reset(struct ata_link *link, unsigned long deadline) return ata_sff_prereset(link, deadline); } +/** + * amd_cable_detect - report cable type + * @ap: port + * + * AMD controller/BIOS setups record the cable type in word 0x42 + */ + static int amd_cable_detect(struct ata_port *ap) { static const u32 bitmask[2] = {0x03, 0x0C}; @@ -158,6 +165,40 @@ static int amd_cable_detect(struct ata_port *ap) } /** + * amd_fifo_setup - set the PIO FIFO for ATA/ATAPI + * @ap: ATA interface + * @adev: ATA device + * + * Set the PCI fifo for this device according to the devices present + * on the bus at this point in time. We need to turn the post write buffer + * off for ATAPI devices as we may need to issue a word sized write to the + * device as the final I/O + */ + +static void amd_fifo_setup(struct ata_port *ap) +{ + struct ata_device *adev; + struct pci_dev *pdev = to_pci_dev(ap->host->dev); + static const u8 fifobit[2] = { 0xC0, 0x30}; + u8 fifo = fifobit[ap->port_no]; + u8 r; + + + ata_for_each_dev(adev, &ap->link, ENABLED) { + if (adev->class == ATA_DEV_ATAPI) + fifo = 0; + } + if (pdev->device == PCI_DEVICE_ID_AMD_VIPER_7411) /* FIFO is broken */ + fifo = 0; + + /* On the later chips the read prefetch bits become no-op bits */ + pci_read_config_byte(pdev, 0x41, &r); + r &= ~fifobit[ap->port_no]; + r |= fifo; + pci_write_config_byte(pdev, 0x41, r); +} + +/** * amd33_set_piomode - set initial PIO mode data * @ap: ATA interface * @adev: ATA device @@ -167,21 +208,25 @@ static int amd_cable_detect(struct ata_port *ap) static void amd33_set_piomode(struct ata_port *ap, struct ata_device *adev) { + amd_fifo_setup(ap); timing_setup(ap, adev, 0x40, adev->pio_mode, 1); } static void amd66_set_piomode(struct ata_port *ap, struct ata_device *adev) { + amd_fifo_setup(ap); timing_setup(ap, adev, 0x40, adev->pio_mode, 2); } static void amd100_set_piomode(struct ata_port *ap, struct ata_device *adev) { + amd_fifo_setup(ap); timing_setup(ap, adev, 0x40, adev->pio_mode, 3); } static void amd133_set_piomode(struct ata_port *ap, struct ata_device *adev) { + amd_fifo_setup(ap); timing_setup(ap, adev, 0x40, adev->pio_mode, 4); } @@ -397,6 +442,16 @@ static struct ata_port_operations nv133_port_ops = { .set_dmamode = nv133_set_dmamode, }; +static void amd_clear_fifo(struct pci_dev *pdev) +{ + u8 fifo; + /* Disable the FIFO, the FIFO logic will re-enable it as + appropriate */ + pci_read_config_byte(pdev, 0x41, &fifo); + fifo &= 0x0F; + pci_write_config_byte(pdev, 0x41, fifo); +} + static int amd_init_one(struct pci_dev *pdev, const struct pci_device_id *id) { static const struct ata_port_info info[10] = { @@ -503,14 +558,8 @@ static int amd_init_one(struct pci_dev *pdev, const struct pci_device_id *id) if (type < 3) ata_pci_bmdma_clear_simplex(pdev); - - /* Check for AMD7411 */ - if (type == 3) - /* FIFO is broken */ - pci_write_config_byte(pdev, 0x41, fifo & 0x0F); - else - pci_write_config_byte(pdev, 0x41, fifo | 0xF0); - + if (pdev->vendor == PCI_VENDOR_ID_AMD) + amd_clear_fifo(pdev); /* Cable detection on Nvidia chips doesn't work too well, * cache BIOS programmed UDMA mode. */ @@ -536,18 +585,11 @@ static int amd_reinit_one(struct pci_dev *pdev) return rc; if (pdev->vendor == PCI_VENDOR_ID_AMD) { - u8 fifo; - pci_read_config_byte(pdev, 0x41, &fifo); - if (pdev->device == PCI_DEVICE_ID_AMD_VIPER_7411) - /* FIFO is broken */ - pci_write_config_byte(pdev, 0x41, fifo & 0x0F); - else - pci_write_config_byte(pdev, 0x41, fifo | 0xF0); + amd_clear_fifo(pdev); if (pdev->device == PCI_DEVICE_ID_AMD_VIPER_7409 || pdev->device == PCI_DEVICE_ID_AMD_COBRA_7401) ata_pci_bmdma_clear_simplex(pdev); } - ata_host_resume(host); return 0; } -- cgit v0.10.2 From c55af1f5abf606118b32e3ce9c3b1bbce5236e7e Mon Sep 17 00:00:00 2001 From: Alan Cox Date: Wed, 11 Feb 2009 13:08:42 -0800 Subject: [libata] pata_legacy: for VLB 32bit PIO don't try tricks with slop These devices are generally used with ATA anyway and it seems that some ATAPI will need us to issue the right number of words. Therefore as we can't switch mid burst on VLB devices we should only use 32bit I/O for suitable block sizes. Signed-off-by: Andrew Morton Signed-off-by: Jeff Garzik diff --git a/drivers/ata/pata_legacy.c b/drivers/ata/pata_legacy.c index 6c1d778..e3bc1b4 100644 --- a/drivers/ata/pata_legacy.c +++ b/drivers/ata/pata_legacy.c @@ -283,9 +283,10 @@ static void pdc20230_set_piomode(struct ata_port *ap, struct ata_device *adev) static unsigned int pdc_data_xfer_vlb(struct ata_device *dev, unsigned char *buf, unsigned int buflen, int rw) { - if (ata_id_has_dword_io(dev->id)) { + int slop = buflen & 3; + /* 32bit I/O capable *and* we need to write a whole number of dwords */ + if (ata_id_has_dword_io(dev->id) && (slop == 0 || slop == 3)) { struct ata_port *ap = dev->link->ap; - int slop = buflen & 3; unsigned long flags; local_irq_save(flags); @@ -735,7 +736,7 @@ static unsigned int vlb32_data_xfer(struct ata_device *adev, unsigned char *buf, struct ata_port *ap = adev->link->ap; int slop = buflen & 3; - if (ata_id_has_dword_io(adev->id)) { + if (ata_id_has_dword_io(adev->id) && (slop == 0 || slop == 3)) { if (rw == WRITE) iowrite32_rep(ap->ioaddr.data_addr, buf, buflen >> 2); else -- cgit v0.10.2 From 49f297f8df9adb797334155470ea9ca68bdb041e Mon Sep 17 00:00:00 2001 From: Michael Neuling Date: Thu, 19 Feb 2009 18:52:20 +0000 Subject: powerpc: Fix load/store float double alignment handler When we introduced VSX, we changed the way FPRs are stored in the thread_struct. Unfortunately we missed the load/store float double alignment handler code when updating how we access FPRs in the thread_struct. Below fixes this and merges the little/big endian case. Signed-off-by: Michael Neuling Signed-off-by: Benjamin Herrenschmidt diff --git a/arch/powerpc/kernel/align.c b/arch/powerpc/kernel/align.c index ada0692..73cb6a3 100644 --- a/arch/powerpc/kernel/align.c +++ b/arch/powerpc/kernel/align.c @@ -367,27 +367,24 @@ static int emulate_multiple(struct pt_regs *regs, unsigned char __user *addr, static int emulate_fp_pair(unsigned char __user *addr, unsigned int reg, unsigned int flags) { - char *ptr = (char *) ¤t->thread.TS_FPR(reg); - int i, ret; + char *ptr0 = (char *) ¤t->thread.TS_FPR(reg); + char *ptr1 = (char *) ¤t->thread.TS_FPR(reg+1); + int i, ret, sw = 0; if (!(flags & F)) return 0; if (reg & 1) return 0; /* invalid form: FRS/FRT must be even */ - if (!(flags & SW)) { - /* not byte-swapped - easy */ - if (!(flags & ST)) - ret = __copy_from_user(ptr, addr, 16); - else - ret = __copy_to_user(addr, ptr, 16); - } else { - /* each FPR value is byte-swapped separately */ - ret = 0; - for (i = 0; i < 16; ++i) { - if (!(flags & ST)) - ret |= __get_user(ptr[i^7], addr + i); - else - ret |= __put_user(ptr[i^7], addr + i); + if (flags & SW) + sw = 7; + ret = 0; + for (i = 0; i < 8; ++i) { + if (!(flags & ST)) { + ret |= __get_user(ptr0[i^sw], addr + i); + ret |= __get_user(ptr1[i^sw], addr + i + 8); + } else { + ret |= __put_user(ptr0[i^sw], addr + i); + ret |= __put_user(ptr1[i^sw], addr + i + 8); } } if (ret) -- cgit v0.10.2 From e423b9ecd6aa434ce9ba72a21fdc61079e620e0a Mon Sep 17 00:00:00 2001 From: Mark Nelson Date: Wed, 25 Feb 2009 13:26:48 +0000 Subject: powerpc: Fix 64bit memcpy() regression This fixes a regression introduced by commit 25d6e2d7c58ddc4a3b614fc5381591c0cfe66556 ("powerpc: Update 64bit memcpy() using CPU_FTR_UNALIGNED_LD_STD"). This commit allowed CPUs that have the CPU_FTR_UNALIGNED_LD_STD CPU feature bit present to do the memcpy() with unaligned load doubles. But, along with this came a bug where our final load double would read bytes beyond a page boundary and into the next (unmapped) page. This was caught by enabling CONFIG_DEBUG_PAGEALLOC, The fix was to read only the number of bytes that we need to store rather than reading a full 8-byte doubleword and storing only a portion of that. In order to minimise the amount of existing code touched we use the original do_tail for the src_unaligned case. Below is an example of the regression, as reported by Sachin Sant: Unable to handle kernel paging request for data at address 0xc00000003f380000 Faulting instruction address: 0xc000000000039574 cpu 0x1: Vector: 300 (Data Access) at [c00000003baf3020] pc: c000000000039574: .memcpy+0x74/0x244 lr: d00000000244916c: .ext3_xattr_get+0x288/0x2f4 [ext3] sp: c00000003baf32a0 msr: 8000000000009032 dar: c00000003f380000 dsisr: 40000000 current = 0xc00000003e54b010 paca = 0xc000000000a53680 pid = 1840, comm = readahead enter ? for help [link register ] d00000000244916c .ext3_xattr_get+0x288/0x2f4 [ext3] [c00000003baf32a0] d000000002449104 .ext3_xattr_get+0x220/0x2f4 [ext3] (unreliab le) [c00000003baf3390] d00000000244a6e8 .ext3_xattr_security_get+0x40/0x5c [ext3] [c00000003baf3400] c000000000148154 .generic_getxattr+0x74/0x9c [c00000003baf34a0] c000000000333400 .inode_doinit_with_dentry+0x1c4/0x678 [c00000003baf3560] c00000000032c6b0 .security_d_instantiate+0x50/0x68 [c00000003baf35e0] c00000000013c818 .d_instantiate+0x78/0x9c [c00000003baf3680] c00000000013ced0 .d_splice_alias+0xf0/0x120 [c00000003baf3720] d00000000243e05c .ext3_lookup+0xec/0x134 [ext3] [c00000003baf37c0] c000000000131e74 .do_lookup+0x110/0x260 [c00000003baf3880] c000000000134ed0 .__link_path_walk+0xa98/0x1010 [c00000003baf3970] c0000000001354a0 .path_walk+0x58/0xc4 [c00000003baf3a20] c000000000135720 .do_path_lookup+0x138/0x1e4 [c00000003baf3ad0] c00000000013645c .path_lookup_open+0x6c/0xc8 [c00000003baf3b70] c000000000136780 .do_filp_open+0xcc/0x874 [c00000003baf3d10] c0000000001251e0 .do_sys_open+0x80/0x140 [c00000003baf3dc0] c00000000016aaec .compat_sys_open+0x24/0x38 [c00000003baf3e30] c00000000000855c syscall_exit+0x0/0x40 Signed-off-by: Benjamin Herrenschmidt diff --git a/arch/powerpc/lib/memcpy_64.S b/arch/powerpc/lib/memcpy_64.S index fe2d34e..e178922 100644 --- a/arch/powerpc/lib/memcpy_64.S +++ b/arch/powerpc/lib/memcpy_64.S @@ -53,18 +53,19 @@ END_FTR_SECTION_IFCLR(CPU_FTR_UNALIGNED_LD_STD) 3: std r8,8(r3) beq 3f addi r3,r3,16 - ld r9,8(r4) .Ldo_tail: bf cr7*4+1,1f - rotldi r9,r9,32 + lwz r9,8(r4) + addi r4,r4,4 stw r9,0(r3) addi r3,r3,4 1: bf cr7*4+2,2f - rotldi r9,r9,16 + lhz r9,8(r4) + addi r4,r4,2 sth r9,0(r3) addi r3,r3,2 2: bf cr7*4+3,3f - rotldi r9,r9,8 + lbz r9,8(r4) stb r9,0(r3) 3: ld r3,48(r1) /* return dest pointer */ blr @@ -133,11 +134,24 @@ END_FTR_SECTION_IFCLR(CPU_FTR_UNALIGNED_LD_STD) cmpwi cr1,r5,8 addi r3,r3,32 sld r9,r9,r10 - ble cr1,.Ldo_tail + ble cr1,6f ld r0,8(r4) srd r7,r0,r11 or r9,r7,r9 - b .Ldo_tail +6: + bf cr7*4+1,1f + rotldi r9,r9,32 + stw r9,0(r3) + addi r3,r3,4 +1: bf cr7*4+2,2f + rotldi r9,r9,16 + sth r9,0(r3) + addi r3,r3,2 +2: bf cr7*4+3,3f + rotldi r9,r9,8 + stb r9,0(r3) +3: ld r3,48(r1) /* return dest pointer */ + blr .Ldst_unaligned: PPC_MTOCRF 0x01,r6 # put #bytes to 8B bdry into cr7 -- cgit v0.10.2 From f72b728bf100f276628e378e1fe6c6acd5d09401 Mon Sep 17 00:00:00 2001 From: Mark Nelson Date: Wed, 25 Feb 2009 13:46:24 +0000 Subject: powerpc: Fix 64bit __copy_tofrom_user() regression This fixes a regression introduced by commit a4e22f02f5b6518c1484faea1f88d81802b9feac ("powerpc: Update 64bit __copy_tofrom_user() using CPU_FTR_UNALIGNED_LD_STD"). The same bug that existed in the 64bit memcpy() also exists here so fix it here too. The fix is the same as that applied to memcpy() with the addition of fixes for the exception handling code required for __copy_tofrom_user(). This stops us reading beyond the end of the source region we were told to copy. Signed-off-by: Mark Nelson Signed-off-by: Benjamin Herrenschmidt diff --git a/arch/powerpc/lib/copyuser_64.S b/arch/powerpc/lib/copyuser_64.S index 70693a5..693b14a 100644 --- a/arch/powerpc/lib/copyuser_64.S +++ b/arch/powerpc/lib/copyuser_64.S @@ -62,18 +62,19 @@ END_FTR_SECTION_IFCLR(CPU_FTR_UNALIGNED_LD_STD) 72: std r8,8(r3) beq+ 3f addi r3,r3,16 -23: ld r9,8(r4) .Ldo_tail: bf cr7*4+1,1f - rotldi r9,r9,32 +23: lwz r9,8(r4) + addi r4,r4,4 73: stw r9,0(r3) addi r3,r3,4 1: bf cr7*4+2,2f - rotldi r9,r9,16 +44: lhz r9,8(r4) + addi r4,r4,2 74: sth r9,0(r3) addi r3,r3,2 2: bf cr7*4+3,3f - rotldi r9,r9,8 +45: lbz r9,8(r4) 75: stb r9,0(r3) 3: li r3,0 blr @@ -141,11 +142,24 @@ END_FTR_SECTION_IFCLR(CPU_FTR_UNALIGNED_LD_STD) 6: cmpwi cr1,r5,8 addi r3,r3,32 sld r9,r9,r10 - ble cr1,.Ldo_tail + ble cr1,7f 34: ld r0,8(r4) srd r7,r0,r11 or r9,r7,r9 - b .Ldo_tail +7: + bf cr7*4+1,1f + rotldi r9,r9,32 +94: stw r9,0(r3) + addi r3,r3,4 +1: bf cr7*4+2,2f + rotldi r9,r9,16 +95: sth r9,0(r3) + addi r3,r3,2 +2: bf cr7*4+3,3f + rotldi r9,r9,8 +96: stb r9,0(r3) +3: li r3,0 + blr .Ldst_unaligned: PPC_MTOCRF 0x01,r6 /* put #bytes to 8B bdry into cr7 */ @@ -218,7 +232,6 @@ END_FTR_SECTION_IFCLR(CPU_FTR_UNALIGNED_LD_STD) 121: 132: addi r3,r3,8 -123: 134: 135: 138: @@ -226,6 +239,9 @@ END_FTR_SECTION_IFCLR(CPU_FTR_UNALIGNED_LD_STD) 140: 141: 142: +123: +144: +145: /* * here we have had a fault on a load and r3 points to the first @@ -309,6 +325,9 @@ END_FTR_SECTION_IFCLR(CPU_FTR_UNALIGNED_LD_STD) 187: 188: 189: +194: +195: +196: 1: ld r6,-24(r1) ld r5,-8(r1) @@ -329,7 +348,9 @@ END_FTR_SECTION_IFCLR(CPU_FTR_UNALIGNED_LD_STD) .llong 72b,172b .llong 23b,123b .llong 73b,173b + .llong 44b,144b .llong 74b,174b + .llong 45b,145b .llong 75b,175b .llong 24b,124b .llong 25b,125b @@ -347,6 +368,9 @@ END_FTR_SECTION_IFCLR(CPU_FTR_UNALIGNED_LD_STD) .llong 79b,179b .llong 80b,180b .llong 34b,134b + .llong 94b,194b + .llong 95b,195b + .llong 96b,196b .llong 35b,135b .llong 81b,181b .llong 36b,136b -- cgit v0.10.2 From 8b1a8ff8b321a9384304aeea4dbdb9747daf7ee8 Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Sat, 28 Feb 2009 00:08:53 -0500 Subject: ext4: Remove duplicate call to ext4_commit_super() in ext4_freeze() Commit c4be0c1d added error checking to ext4_freeze() when calling ext4_commit_super(). Unfortunately the patch failed to remove the original call to ext4_commit_super(), with the net result that when freezing the filesystem, the superblock gets written twice, the first time without error checking. Signed-off-by: "Theodore Ts'o" diff --git a/fs/ext4/super.c b/fs/ext4/super.c index a5732c5..39d1993 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -3091,7 +3091,6 @@ static int ext4_freeze(struct super_block *sb) /* Journal blocked and flushed, clear needs_recovery flag. */ EXT4_CLEAR_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_RECOVER); - ext4_commit_super(sb, EXT4_SB(sb)->s_es, 1); error = ext4_commit_super(sb, EXT4_SB(sb)->s_es, 1); if (error) goto out; -- cgit v0.10.2 From d8ae4601a4b7ea1fa17fa395c3468c0e144d1275 Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Sat, 28 Feb 2009 09:50:01 -0500 Subject: ext4: Reorder fs/Makefile so that ext2 root fs's are mounted using ext2 In fs/Makefile, ext3 was placed before ext2 so that a root filesystem that possessed a journal, it would be mounted as ext3 instead of ext2. This was necessary because a cleanly unmounted ext3 filesystem was fully backwards compatible with ext2, and could be mounted by ext2 --- but it was desirable that it be mounted with ext3 so that the journaling would be enabled. The ext4 filesystem supports new incompatible features, so there is no danger of an ext4 filesystem being mistaken for an ext2 filesystem. At that point, the relative ordering of ext4 with respect to ext2 didn't matter until ext4 gained the ability to mount filesystems without a journal starting in 2.6.29-rc1. Now that this is the case, given that ext4 is before ext2, it means that root filesystems that were using the plain-jane ext2 format are getting mounted using the ext4 filesystem driver, which is a change in behavior which could be surprising to users. It's doubtful that there are that many ext2-only root filesystem users that would also have ext4 compiled into the kernel, but to adhere to the principle of least surprise, the correct ordering in fs/Makefile is ext3, followed by ext2, and finally ext4. Signed-off-by: "Theodore Ts'o" diff --git a/fs/Makefile b/fs/Makefile index 38bc735..dc20db3 100644 --- a/fs/Makefile +++ b/fs/Makefile @@ -69,10 +69,12 @@ obj-$(CONFIG_DLM) += dlm/ # Do not add any filesystems before this line obj-$(CONFIG_REISERFS_FS) += reiserfs/ obj-$(CONFIG_EXT3_FS) += ext3/ # Before ext2 so root fs can be ext3 -obj-$(CONFIG_EXT4_FS) += ext4/ # Before ext2 so root fs can be ext4 +obj-$(CONFIG_EXT2_FS) += ext2/ +# We place ext4 after ext2 so plain ext2 root fs's are mounted using ext2 +# unless explicitly requested by rootfstype +obj-$(CONFIG_EXT4_FS) += ext4/ obj-$(CONFIG_JBD) += jbd/ obj-$(CONFIG_JBD2) += jbd2/ -obj-$(CONFIG_EXT2_FS) += ext2/ obj-$(CONFIG_CRAMFS) += cramfs/ obj-$(CONFIG_SQUASHFS) += squashfs/ obj-y += ramfs/ -- cgit v0.10.2 From 8f64b32eb73fbfe9f38c4123121b63ee409278a7 Mon Sep 17 00:00:00 2001 From: Eric Sandeen Date: Thu, 26 Feb 2009 00:57:35 -0500 Subject: ext4: don't call jbd2_journal_force_commit_nested without journal Running without a journal, I oopsed when I ran out of space, because we called jbd2_journal_force_commit_nested() from ext4_should_retry_alloc() without a journal. This should take care of it, I think. Signed-off-by: Eric Sandeen Signed-off-by: "Theodore Ts'o" diff --git a/fs/ext4/balloc.c b/fs/ext4/balloc.c index 9a50b80..de9459b 100644 --- a/fs/ext4/balloc.c +++ b/fs/ext4/balloc.c @@ -609,7 +609,9 @@ int ext4_claim_free_blocks(struct ext4_sb_info *sbi, */ int ext4_should_retry_alloc(struct super_block *sb, int *retries) { - if (!ext4_has_free_blocks(EXT4_SB(sb), 1) || (*retries)++ > 3) + if (!ext4_has_free_blocks(EXT4_SB(sb), 1) || + (*retries)++ > 3 || + !EXT4_SB(sb)->s_journal) return 0; jbd_debug(1, "%s: retrying operation after ENOSPC\n", sb->s_id); diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 51cdd13..c7fed5b 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -2544,7 +2544,7 @@ retry: ext4_journal_stop(handle); - if (mpd.retval == -ENOSPC) { + if ((mpd.retval == -ENOSPC) && sbi->s_journal) { /* commit the transaction which would * free blocks released in the transaction * and try again -- cgit v0.10.2 From b2bf96833c5782befc3e7700f791fde754a47b01 Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Thu, 19 Feb 2009 08:50:26 +0100 Subject: block: fix bogus gcc warning for uninitialized var usage Newer gcc throw this warning: fs/bio.c: In function ?bio_alloc_bioset?: fs/bio.c:305: warning: ?p? may be used uninitialized in this function since it cannot figure out that 'p' is only ever used if 'bs' is non-NULL. Signed-off-by: Jens Axboe diff --git a/fs/bio.c b/fs/bio.c index 72ab251..124b95c 100644 --- a/fs/bio.c +++ b/fs/bio.c @@ -302,7 +302,7 @@ void bio_init(struct bio *bio) struct bio *bio_alloc_bioset(gfp_t gfp_mask, int nr_iovecs, struct bio_set *bs) { struct bio *bio = NULL; - void *p; + void *uninitialized_var(p); if (bs) { p = mempool_alloc(bs->bio_pool, gfp_mask); -- cgit v0.10.2 From 9e8c0bccdc944bd09361672d47660810c027bcaa Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?M=C3=A1rton=20N=C3=A9meth?= Date: Fri, 20 Feb 2009 08:12:51 +0100 Subject: block: add documentation for register_blkdev() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Add documentation for register_blkdev() function and for the parameters. Signed-off-by: Márton Németh Cc: Greg Kroah-Hartman Signed-off-by: Jens Axboe diff --git a/block/genhd.c b/block/genhd.c index e1eadcc..a9ec910 100644 --- a/block/genhd.c +++ b/block/genhd.c @@ -256,6 +256,22 @@ void blkdev_show(struct seq_file *seqf, off_t offset) } #endif /* CONFIG_PROC_FS */ +/** + * register_blkdev - register a new block device + * + * @major: the requested major device number [1..255]. If @major=0, try to + * allocate any unused major number. + * @name: the name of the new block device as a zero terminated string + * + * The @name must be unique within the system. + * + * The return value depends on the @major input parameter. + * - if a major device number was requested in range [1..255] then the + * function returns zero on success, or a negative error code + * - if any unused major number was requested with @major=0 parameter + * then the return value is the allocated major number in range + * [1..255] or a negative error code otherwise + */ int register_blkdev(unsigned int major, const char *name) { struct blk_major_name **n, *p; -- cgit v0.10.2 From 5e4c91c84b194b26cf592779e451f4b5be777cba Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Mon, 23 Feb 2009 08:53:35 +0100 Subject: cciss: shorten 30s timeout on controller reset If reset_devices is set for kexec, then cciss will delay 30 seconds since the old 5i controller _may_ need that long to recover. Replace the long sleep with incremental sleep and tests to reduce the 30 seconds to worst case for 5i, so that other controllers will proceed quickly. Reviewed-by: Mike Miller Signed-off-by: Jens Axboe diff --git a/drivers/block/cciss.c b/drivers/block/cciss.c index d2cb67b..b5a0611 100644 --- a/drivers/block/cciss.c +++ b/drivers/block/cciss.c @@ -3611,11 +3611,15 @@ static int __devinit cciss_init_one(struct pci_dev *pdev, schedule_timeout_uninterruptible(30*HZ); /* Now try to get the controller to respond to a no-op */ - for (i=0; i<12; i++) { + for (i=0; i<30; i++) { if (cciss_noop(pdev) == 0) break; - else - printk("cciss: no-op failed%s\n", (i < 11 ? "; re-trying" : "")); + + schedule_timeout_uninterruptible(HZ); + } + if (i == 30) { + printk(KERN_ERR "cciss: controller seems dead\n"); + return -EBUSY; } } -- cgit v0.10.2 From 1e42807918d17e8c93bf14fbb74be84b141334c1 Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Mon, 23 Feb 2009 09:03:10 +0100 Subject: block: reduce stack footprint of blk_recount_segments() blk_recalc_rq_segments() requires a request structure passed in, which we don't have from blk_recount_segments(). So the latter allocates one on the stack, using > 400 bytes of stack for that. This can cause us to spill over one page of stack from ext4 at least: 0) 4560 400 blk_recount_segments+0x43/0x62 1) 4160 32 bio_phys_segments+0x1c/0x24 2) 4128 32 blk_rq_bio_prep+0x2a/0xf9 3) 4096 32 init_request_from_bio+0xf9/0xfe 4) 4064 112 __make_request+0x33c/0x3f6 5) 3952 144 generic_make_request+0x2d1/0x321 6) 3808 64 submit_bio+0xb9/0xc3 7) 3744 48 submit_bh+0xea/0x10e 8) 3696 368 ext4_mb_init_cache+0x257/0xa6a [ext4] 9) 3328 288 ext4_mb_regular_allocator+0x421/0xcd9 [ext4] 10) 3040 160 ext4_mb_new_blocks+0x211/0x4b4 [ext4] 11) 2880 336 ext4_ext_get_blocks+0xb61/0xd45 [ext4] 12) 2544 96 ext4_get_blocks_wrap+0xf2/0x200 [ext4] 13) 2448 80 ext4_da_get_block_write+0x6e/0x16b [ext4] 14) 2368 352 mpage_da_map_blocks+0x7e/0x4b3 [ext4] 15) 2016 352 ext4_da_writepages+0x2ce/0x43c [ext4] 16) 1664 32 do_writepages+0x2d/0x3c 17) 1632 144 __writeback_single_inode+0x162/0x2cd 18) 1488 96 generic_sync_sb_inodes+0x1e3/0x32b 19) 1392 16 sync_sb_inodes+0xe/0x10 20) 1376 48 writeback_inodes+0x69/0xb3 21) 1328 208 balance_dirty_pages_ratelimited_nr+0x187/0x2f9 22) 1120 224 generic_file_buffered_write+0x1d4/0x2c4 23) 896 176 __generic_file_aio_write_nolock+0x35f/0x393 24) 720 80 generic_file_aio_write+0x6c/0xc8 25) 640 80 ext4_file_write+0xa9/0x137 [ext4] 26) 560 320 do_sync_write+0xf0/0x137 27) 240 48 vfs_write+0xb3/0x13c 28) 192 64 sys_write+0x4c/0x74 29) 128 128 system_call_fastpath+0x16/0x1b Split the segment counting out into a __blk_recalc_rq_segments() helper to avoid allocating an onstack request just for checking the physical segment count. Signed-off-by: Jens Axboe diff --git a/block/blk-merge.c b/block/blk-merge.c index b92f5b0..a104593 100644 --- a/block/blk-merge.c +++ b/block/blk-merge.c @@ -38,72 +38,84 @@ void blk_recalc_rq_sectors(struct request *rq, int nsect) } } -void blk_recalc_rq_segments(struct request *rq) +static unsigned int __blk_recalc_rq_segments(struct request_queue *q, + struct bio *bio, + unsigned int *seg_size_ptr) { - int nr_phys_segs; unsigned int phys_size; struct bio_vec *bv, *bvprv = NULL; - int seg_size; - int cluster; - struct req_iterator iter; - int high, highprv = 1; - struct request_queue *q = rq->q; + int cluster, i, high, highprv = 1; + unsigned int seg_size, nr_phys_segs; + struct bio *fbio; - if (!rq->bio) - return; + if (!bio) + return 0; + fbio = bio; cluster = test_bit(QUEUE_FLAG_CLUSTER, &q->queue_flags); seg_size = 0; phys_size = nr_phys_segs = 0; - rq_for_each_segment(bv, rq, iter) { - /* - * the trick here is making sure that a high page is never - * considered part of another segment, since that might - * change with the bounce page. - */ - high = page_to_pfn(bv->bv_page) > q->bounce_pfn; - if (high || highprv) - goto new_segment; - if (cluster) { - if (seg_size + bv->bv_len > q->max_segment_size) - goto new_segment; - if (!BIOVEC_PHYS_MERGEABLE(bvprv, bv)) - goto new_segment; - if (!BIOVEC_SEG_BOUNDARY(q, bvprv, bv)) + for_each_bio(bio) { + bio_for_each_segment(bv, bio, i) { + /* + * the trick here is making sure that a high page is + * never considered part of another segment, since that + * might change with the bounce page. + */ + high = page_to_pfn(bv->bv_page) > q->bounce_pfn; + if (high || highprv) goto new_segment; + if (cluster) { + if (seg_size + bv->bv_len > q->max_segment_size) + goto new_segment; + if (!BIOVEC_PHYS_MERGEABLE(bvprv, bv)) + goto new_segment; + if (!BIOVEC_SEG_BOUNDARY(q, bvprv, bv)) + goto new_segment; + + seg_size += bv->bv_len; + bvprv = bv; + continue; + } +new_segment: + if (nr_phys_segs == 1 && seg_size > + fbio->bi_seg_front_size) + fbio->bi_seg_front_size = seg_size; - seg_size += bv->bv_len; + nr_phys_segs++; bvprv = bv; - continue; + seg_size = bv->bv_len; + highprv = high; } -new_segment: - if (nr_phys_segs == 1 && seg_size > rq->bio->bi_seg_front_size) - rq->bio->bi_seg_front_size = seg_size; - - nr_phys_segs++; - bvprv = bv; - seg_size = bv->bv_len; - highprv = high; } - if (nr_phys_segs == 1 && seg_size > rq->bio->bi_seg_front_size) + if (seg_size_ptr) + *seg_size_ptr = seg_size; + + return nr_phys_segs; +} + +void blk_recalc_rq_segments(struct request *rq) +{ + unsigned int seg_size = 0, phys_segs; + + phys_segs = __blk_recalc_rq_segments(rq->q, rq->bio, &seg_size); + + if (phys_segs == 1 && seg_size > rq->bio->bi_seg_front_size) rq->bio->bi_seg_front_size = seg_size; if (seg_size > rq->biotail->bi_seg_back_size) rq->biotail->bi_seg_back_size = seg_size; - rq->nr_phys_segments = nr_phys_segs; + rq->nr_phys_segments = phys_segs; } void blk_recount_segments(struct request_queue *q, struct bio *bio) { - struct request rq; struct bio *nxt = bio->bi_next; - rq.q = q; - rq.bio = rq.biotail = bio; + bio->bi_next = NULL; - blk_recalc_rq_segments(&rq); + bio->bi_phys_segments = __blk_recalc_rq_segments(q, bio, NULL); bio->bi_next = nxt; - bio->bi_phys_segments = rq.nr_phys_segments; bio->bi_flags |= (1 << BIO_SEG_VALID); } EXPORT_SYMBOL(blk_recount_segments); diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h index dcaa0fd..465d6ba 100644 --- a/include/linux/blkdev.h +++ b/include/linux/blkdev.h @@ -708,6 +708,8 @@ struct req_iterator { }; /* This should not be used directly - use rq_for_each_segment */ +#define for_each_bio(_bio) \ + for (; _bio; _bio = _bio->bi_next) #define __rq_for_each_bio(_bio, rq) \ if ((rq->bio)) \ for (_bio = (rq)->bio; _bio; _bio = _bio->bi_next) -- cgit v0.10.2 From 9e973e64ac6dc504e6447d52193d4fff1a670156 Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Tue, 24 Feb 2009 08:10:09 +0100 Subject: xen/blkfront: use blk_rq_map_sg to generate ring entries On occasion, the request will apparently have more segments than we fit into the ring. Jens says: > The second problem is that the block layer then appears to create one > too many segments, but from the dump it has rq->nr_phys_segments == > BLKIF_MAX_SEGMENTS_PER_REQUEST. I suspect the latter is due to > xen-blkfront not handling the merging on its own. It should check that > the new page doesn't form part of the previous page. The > rq_for_each_segment() iterates all single bits in the request, not dma > segments. The "easiest" way to do this is to call blk_rq_map_sg() and > then iterate the mapped sg list. That will give you what you are > looking for. > Here's a test patch, compiles but otherwise untested. I spent more > time figuring out how to enable XEN than to code it up, so YMMV! > Probably the sg list wants to be put inside the ring and only > initialized on allocation, then you can get rid of the sg on stack and > sg_init_table() loop call in the function. I'll leave that, and the > testing, to you. [Moved sg array into info structure, and initialize once. -J] Signed-off-by: Jens Axboe Signed-off-by: Jeremy Fitzhardinge diff --git a/drivers/block/xen-blkfront.c b/drivers/block/xen-blkfront.c index 918ef725..b6c8ce2 100644 --- a/drivers/block/xen-blkfront.c +++ b/drivers/block/xen-blkfront.c @@ -40,6 +40,7 @@ #include #include #include +#include #include #include @@ -82,6 +83,7 @@ struct blkfront_info enum blkif_state connected; int ring_ref; struct blkif_front_ring ring; + struct scatterlist sg[BLKIF_MAX_SEGMENTS_PER_REQUEST]; unsigned int evtchn, irq; struct request_queue *rq; struct work_struct work; @@ -204,12 +206,11 @@ static int blkif_queue_request(struct request *req) struct blkfront_info *info = req->rq_disk->private_data; unsigned long buffer_mfn; struct blkif_request *ring_req; - struct req_iterator iter; - struct bio_vec *bvec; unsigned long id; unsigned int fsect, lsect; - int ref; + int i, ref; grant_ref_t gref_head; + struct scatterlist *sg; if (unlikely(info->connected != BLKIF_STATE_CONNECTED)) return 1; @@ -238,12 +239,13 @@ static int blkif_queue_request(struct request *req) if (blk_barrier_rq(req)) ring_req->operation = BLKIF_OP_WRITE_BARRIER; - ring_req->nr_segments = 0; - rq_for_each_segment(bvec, req, iter) { - BUG_ON(ring_req->nr_segments == BLKIF_MAX_SEGMENTS_PER_REQUEST); - buffer_mfn = pfn_to_mfn(page_to_pfn(bvec->bv_page)); - fsect = bvec->bv_offset >> 9; - lsect = fsect + (bvec->bv_len >> 9) - 1; + ring_req->nr_segments = blk_rq_map_sg(req->q, req, info->sg); + BUG_ON(ring_req->nr_segments > BLKIF_MAX_SEGMENTS_PER_REQUEST); + + for_each_sg(info->sg, sg, ring_req->nr_segments, i) { + buffer_mfn = pfn_to_mfn(page_to_pfn(sg_page(sg))); + fsect = sg->offset >> 9; + lsect = fsect + (sg->length >> 9) - 1; /* install a grant reference. */ ref = gnttab_claim_grant_reference(&gref_head); BUG_ON(ref == -ENOSPC); @@ -254,16 +256,12 @@ static int blkif_queue_request(struct request *req) buffer_mfn, rq_data_dir(req) ); - info->shadow[id].frame[ring_req->nr_segments] = - mfn_to_pfn(buffer_mfn); - - ring_req->seg[ring_req->nr_segments] = + info->shadow[id].frame[i] = mfn_to_pfn(buffer_mfn); + ring_req->seg[i] = (struct blkif_request_segment) { .gref = ref, .first_sect = fsect, .last_sect = lsect }; - - ring_req->nr_segments++; } info->ring.req_prod_pvt++; @@ -622,6 +620,8 @@ static int setup_blkring(struct xenbus_device *dev, SHARED_RING_INIT(sring); FRONT_RING_INIT(&info->ring, sring, PAGE_SIZE); + sg_init_table(info->sg, BLKIF_MAX_SEGMENTS_PER_REQUEST); + err = xenbus_grant_ring(dev, virt_to_mfn(info->ring.sring)); if (err < 0) { free_page((unsigned long)sring); -- cgit v0.10.2 From 3f53a38131a4e7a053c0aa060aba0411242fb6b9 Mon Sep 17 00:00:00 2001 From: Pavel Emelyanov Date: Thu, 26 Feb 2009 03:35:13 -0800 Subject: ipv6: don't use tw net when accounting for recycled tw We already have a valid net in that place, but this is not just a cleanup - the tw pointer can be NULL there sometimes, thus causing an oops in NET_NS=y case. The same place in ipv4 code already works correctly using existing net, rather than tw's one. The bug exists since 2.6.27. Signed-off-by: Pavel Emelyanov Signed-off-by: David S. Miller diff --git a/net/ipv6/inet6_hashtables.c b/net/ipv6/inet6_hashtables.c index 8fe267f..1bcc343 100644 --- a/net/ipv6/inet6_hashtables.c +++ b/net/ipv6/inet6_hashtables.c @@ -258,11 +258,11 @@ unique: if (twp != NULL) { *twp = tw; - NET_INC_STATS_BH(twsk_net(tw), LINUX_MIB_TIMEWAITRECYCLED); + NET_INC_STATS_BH(net, LINUX_MIB_TIMEWAITRECYCLED); } else if (tw != NULL) { /* Silly. Should hash-dance instead... */ inet_twsk_deschedule(tw, death_row); - NET_INC_STATS_BH(twsk_net(tw), LINUX_MIB_TIMEWAITRECYCLED); + NET_INC_STATS_BH(net, LINUX_MIB_TIMEWAITRECYCLED); inet_twsk_put(tw); } -- cgit v0.10.2 From 86883c2736e9697a38080a31c2794fa1316fd68f Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Thu, 26 Feb 2009 10:32:31 -0800 Subject: Make ieee1394_init a fs-initcall It needs to happen before any firewire driver actually registers itself, and that was previously handled by having the Makefile list the core ieee1394 files before the drivers. But now there are firewire drivers in drivers/media, and the Makefile games aren't enough. So just make ieee1394_init happen earlier in the init sequence, the way all other bus layers already do. Reported-and-tested-by: Ingo Molnar Cc: Stefan Richter Cc: Henrik Kurelid Cc: Mauro Carvalho Chehab Cc: Ben Backx Signed-off-by: Linus Torvalds diff --git a/drivers/ieee1394/ieee1394_core.c b/drivers/ieee1394/ieee1394_core.c index 1028e72..8723380 100644 --- a/drivers/ieee1394/ieee1394_core.c +++ b/drivers/ieee1394/ieee1394_core.c @@ -1275,7 +1275,7 @@ static void __exit ieee1394_cleanup(void) unregister_chrdev_region(IEEE1394_CORE_DEV, 256); } -module_init(ieee1394_init); +fs_initcall(ieee1394_init); module_exit(ieee1394_cleanup); /* Exported symbols */ -- cgit v0.10.2 From 47be12e4eec84c1846f29af64fe25a396b57a026 Mon Sep 17 00:00:00 2001 From: Tao Ma Date: Fri, 9 Jan 2009 07:32:48 +0800 Subject: ocfs2: Access and dirty the buffer_head in mark_written. In __ocfs2_mark_extent_written, when we meet with the situation of c_split_covers_rec, the old solution just replace the extent record and forget to access and dirty the buffer_head. This will cause a problem when the unwritten extent is in an extent block. So access and dirty it. Signed-off-by: Tao Ma Signed-off-by: Mark Fasheh diff --git a/fs/ocfs2/alloc.c b/fs/ocfs2/alloc.c index 60fe74035..3a9e5de 100644 --- a/fs/ocfs2/alloc.c +++ b/fs/ocfs2/alloc.c @@ -4796,6 +4796,29 @@ out: return ret; } +static int ocfs2_replace_extent_rec(struct inode *inode, + handle_t *handle, + struct ocfs2_path *path, + struct ocfs2_extent_list *el, + int split_index, + struct ocfs2_extent_rec *split_rec) +{ + int ret; + + ret = ocfs2_path_bh_journal_access(handle, inode, path, + path_num_items(path) - 1); + if (ret) { + mlog_errno(ret); + goto out; + } + + el->l_recs[split_index] = *split_rec; + + ocfs2_journal_dirty(handle, path_leaf_bh(path)); +out: + return ret; +} + /* * Mark part or all of the extent record at split_index in the leaf * pointed to by path as written. This removes the unwritten @@ -4885,7 +4908,9 @@ static int __ocfs2_mark_extent_written(struct inode *inode, if (ctxt.c_contig_type == CONTIG_NONE) { if (ctxt.c_split_covers_rec) - el->l_recs[split_index] = *split_rec; + ret = ocfs2_replace_extent_rec(inode, handle, + path, el, + split_index, split_rec); else ret = ocfs2_split_and_insert(inode, handle, path, et, &last_eb_bh, split_index, -- cgit v0.10.2 From 7dc102b737e9f49dac426161294cb2d326a97d8e Mon Sep 17 00:00:00 2001 From: Sunil Mushran Date: Tue, 3 Feb 2009 12:37:13 -0800 Subject: ocfs2/dlm: Retract fix for race between purge and migrate Mainline commit d4f7e650e55af6b235871126f747da88600e8040 attempts to delay the dlm_thread from sending the drop ref message if the lockres is being migrated. The problem is that we make the dlm_thread wait for the migration to complete. This causes a deadlock as dlm_thread also participates in the lockres migration process. A better fix for the original oss bugzilla#1012 is in testing. Signed-off-by: Sunil Mushran Acked-by: Joel Becker Signed-off-by: Mark Fasheh diff --git a/fs/ocfs2/dlm/dlmthread.c b/fs/ocfs2/dlm/dlmthread.c index d129520..4060bb3 100644 --- a/fs/ocfs2/dlm/dlmthread.c +++ b/fs/ocfs2/dlm/dlmthread.c @@ -181,8 +181,7 @@ static int dlm_purge_lockres(struct dlm_ctxt *dlm, spin_lock(&res->spinlock); /* This ensures that clear refmap is sent after the set */ - __dlm_wait_on_lockres_flags(res, (DLM_LOCK_RES_SETREF_INPROG | - DLM_LOCK_RES_MIGRATING)); + __dlm_wait_on_lockres_flags(res, DLM_LOCK_RES_SETREF_INPROG); spin_unlock(&res->spinlock); /* clear our bit from the master's refmap, ignore errors */ -- cgit v0.10.2 From c74ff8bb2235d848beb67fcfddae71ecbe3f92b1 Mon Sep 17 00:00:00 2001 From: Sunil Mushran Date: Tue, 3 Feb 2009 12:37:14 -0800 Subject: ocfs2: Cleanup the lockname print in dlmglue.c The dentry lock has a different format than other locks. This patch fixes ocfs2_log_dlm_error() macro to make it print the dentry lock correctly. Signed-off-by: Sunil Mushran Acked-by: Joel Becker Signed-off-by: Mark Fasheh diff --git a/fs/ocfs2/dlmglue.c b/fs/ocfs2/dlmglue.c index 206a237..7219a86 100644 --- a/fs/ocfs2/dlmglue.c +++ b/fs/ocfs2/dlmglue.c @@ -320,9 +320,14 @@ static void ocfs2_schedule_blocked_lock(struct ocfs2_super *osb, struct ocfs2_lock_res *lockres); static inline void ocfs2_recover_from_dlm_error(struct ocfs2_lock_res *lockres, int convert); -#define ocfs2_log_dlm_error(_func, _err, _lockres) do { \ - mlog(ML_ERROR, "DLM error %d while calling %s on resource %s\n", \ - _err, _func, _lockres->l_name); \ +#define ocfs2_log_dlm_error(_func, _err, _lockres) do { \ + if ((_lockres)->l_type != OCFS2_LOCK_TYPE_DENTRY) \ + mlog(ML_ERROR, "DLM error %d while calling %s on resource %s\n", \ + _err, _func, _lockres->l_name); \ + else \ + mlog(ML_ERROR, "DLM error %d while calling %s on resource %.*s%08x\n", \ + _err, _func, OCFS2_DENTRY_LOCK_INO_START - 1, (_lockres)->l_name, \ + (unsigned int)ocfs2_get_dentry_lock_ino(_lockres)); \ } while (0) static int ocfs2_downconvert_thread(void *arg); static void ocfs2_downconvert_on_unlock(struct ocfs2_super *osb, -- cgit v0.10.2 From dabc47de7a23f57522dc762d9d2ad875700d3497 Mon Sep 17 00:00:00 2001 From: Sunil Mushran Date: Tue, 3 Feb 2009 12:37:15 -0800 Subject: ocfs2/dlm: Use ast_lock to protect ast_list The code was using dlm->spinlock instead of dlm->ast_lock to protect the ast_list. This patch fixes the issue. Signed-off-by: Sunil Mushran Acked-by: Joel Becker Signed-off-by: Mark Fasheh diff --git a/fs/ocfs2/dlm/dlmunlock.c b/fs/ocfs2/dlm/dlmunlock.c index 86ca085..fcf879e 100644 --- a/fs/ocfs2/dlm/dlmunlock.c +++ b/fs/ocfs2/dlm/dlmunlock.c @@ -117,11 +117,11 @@ static enum dlm_status dlmunlock_common(struct dlm_ctxt *dlm, else BUG_ON(res->owner == dlm->node_num); - spin_lock(&dlm->spinlock); + spin_lock(&dlm->ast_lock); /* We want to be sure that we're not freeing a lock * that still has AST's pending... */ in_use = !list_empty(&lock->ast_list); - spin_unlock(&dlm->spinlock); + spin_unlock(&dlm->ast_lock); if (in_use) { mlog(ML_ERROR, "lockres %.*s: Someone is calling dlmunlock " "while waiting for an ast!", res->lockname.len, -- cgit v0.10.2 From 53ecd25e148615e0ed2a72635cc76f4773f97f90 Mon Sep 17 00:00:00 2001 From: Sunil Mushran Date: Tue, 3 Feb 2009 12:37:16 -0800 Subject: ocfs2/dlm: Make dlm_assert_master_handler() kill itself instead of the asserter In dlm_assert_master_handler(), if we get an incorrect assert master from a node that, we reply with EINVAL asking the asserter to die. The problem is that an assert is sent after so many hoops, it is invariably the node that thinks the asserter is wrong, is actually wrong. So instead of killing the asserter, this patch kills the assertee. This patch papers over a race that is still being addressed. Signed-off-by: Sunil Mushran Acked-by: Joel Becker Signed-off-by: Mark Fasheh diff --git a/fs/ocfs2/dlm/dlmmaster.c b/fs/ocfs2/dlm/dlmmaster.c index 54e182a..0a28139 100644 --- a/fs/ocfs2/dlm/dlmmaster.c +++ b/fs/ocfs2/dlm/dlmmaster.c @@ -1849,12 +1849,12 @@ int dlm_assert_master_handler(struct o2net_msg *msg, u32 len, void *data, if (!mle) { if (res->owner != DLM_LOCK_RES_OWNER_UNKNOWN && res->owner != assert->node_idx) { - mlog(ML_ERROR, "assert_master from " - "%u, but current owner is " - "%u! (%.*s)\n", - assert->node_idx, res->owner, - namelen, name); - goto kill; + mlog(ML_ERROR, "DIE! Mastery assert from %u, " + "but current owner is %u! (%.*s)\n", + assert->node_idx, res->owner, namelen, + name); + __dlm_print_one_lock_resource(res); + BUG(); } } else if (mle->type != DLM_MLE_MIGRATION) { if (res->owner != DLM_LOCK_RES_OWNER_UNKNOWN) { -- cgit v0.10.2 From 89a907afe073b8971a83d0ad54f391542b64d327 Mon Sep 17 00:00:00 2001 From: Tao Ma Date: Tue, 17 Feb 2009 04:39:28 +0800 Subject: ocfs2: Use the right access_* method in ctime update of xattr. In ctime updating of xattr, it use the wrong type of access for inode, so use ocfs2_journal_access_di instead. Reported-and-Tested-by: Tristan Ye Signed-off-by: Tao Ma Acked-by: Joel Becker Signed-off-by: Mark Fasheh diff --git a/fs/ocfs2/xattr.c b/fs/ocfs2/xattr.c index 915039f..e393315 100644 --- a/fs/ocfs2/xattr.c +++ b/fs/ocfs2/xattr.c @@ -2592,8 +2592,9 @@ static int __ocfs2_xattr_set_handle(struct inode *inode, if (!ret) { /* Update inode ctime. */ - ret = ocfs2_journal_access(ctxt->handle, inode, xis->inode_bh, - OCFS2_JOURNAL_ACCESS_WRITE); + ret = ocfs2_journal_access_di(ctxt->handle, inode, + xis->inode_bh, + OCFS2_JOURNAL_ACCESS_WRITE); if (ret) { mlog_errno(ret); goto out; -- cgit v0.10.2 From c8b9cf9a7cd25ba65166116d0a958f0bc709f0a7 Mon Sep 17 00:00:00 2001 From: Tao Ma Date: Tue, 24 Feb 2009 17:40:26 -0800 Subject: ocfs2: lock the metaecc process for xattr bucket For other metadata in ocfs2, metaecc is checked in ocfs2_read_blocks with io_mutex held. While for xattr bucket, it is calculated by the whole buckets. So we have to add a spin_lock to prevent multiple processes calculating metaecc. Signed-off-by: Tao Ma Tested-by: Tristan Ye Signed-off-by: Mark Fasheh diff --git a/fs/ocfs2/ocfs2.h b/fs/ocfs2/ocfs2.h index 0773841..946d3c3 100644 --- a/fs/ocfs2/ocfs2.h +++ b/fs/ocfs2/ocfs2.h @@ -341,6 +341,9 @@ struct ocfs2_super struct ocfs2_node_map osb_recovering_orphan_dirs; unsigned int *osb_orphan_wipes; wait_queue_head_t osb_wipe_event; + + /* used to protect metaecc calculation check of xattr. */ + spinlock_t osb_xattr_lock; }; #define OCFS2_SB(sb) ((struct ocfs2_super *)(sb)->s_fs_info) diff --git a/fs/ocfs2/super.c b/fs/ocfs2/super.c index b1cb38f..1c3acc4 100644 --- a/fs/ocfs2/super.c +++ b/fs/ocfs2/super.c @@ -1747,6 +1747,7 @@ static int ocfs2_initialize_super(struct super_block *sb, INIT_LIST_HEAD(&osb->blocked_lock_list); osb->blocked_lock_count = 0; spin_lock_init(&osb->osb_lock); + spin_lock_init(&osb->osb_xattr_lock); ocfs2_init_inode_steal_slot(osb); atomic_set(&osb->alloc_stats.moves, 0); diff --git a/fs/ocfs2/xattr.c b/fs/ocfs2/xattr.c index e393315..a7c1679 100644 --- a/fs/ocfs2/xattr.c +++ b/fs/ocfs2/xattr.c @@ -274,10 +274,12 @@ static int ocfs2_read_xattr_bucket(struct ocfs2_xattr_bucket *bucket, bucket->bu_blocks, bucket->bu_bhs, 0, NULL); if (!rc) { + spin_lock(&OCFS2_SB(bucket->bu_inode->i_sb)->osb_xattr_lock); rc = ocfs2_validate_meta_ecc_bhs(bucket->bu_inode->i_sb, bucket->bu_bhs, bucket->bu_blocks, &bucket_xh(bucket)->xh_check); + spin_unlock(&OCFS2_SB(bucket->bu_inode->i_sb)->osb_xattr_lock); if (rc) mlog_errno(rc); } @@ -310,9 +312,11 @@ static void ocfs2_xattr_bucket_journal_dirty(handle_t *handle, { int i; + spin_lock(&OCFS2_SB(bucket->bu_inode->i_sb)->osb_xattr_lock); ocfs2_compute_meta_ecc_bhs(bucket->bu_inode->i_sb, bucket->bu_bhs, bucket->bu_blocks, &bucket_xh(bucket)->xh_check); + spin_unlock(&OCFS2_SB(bucket->bu_inode->i_sb)->osb_xattr_lock); for (i = 0; i < bucket->bu_blocks; i++) ocfs2_journal_dirty(handle, bucket->bu_bhs[i]); -- cgit v0.10.2 From 4442f518269c6b3686fcbcadad22dc4475309b16 Mon Sep 17 00:00:00 2001 From: Tiger Yang Date: Fri, 20 Feb 2009 11:11:50 +0800 Subject: ocfs2: set gap to seperate entry and value when xattr in bucket This patch set a gap (4 bytes) between xattr entry and name/value when xattr in bucket. This gap use to seperate entry and name/value when a bucket is full. It had already been set when xattr in inode/block. Signed-off-by: Tiger Yang Signed-off-by: Mark Fasheh diff --git a/fs/ocfs2/xattr.c b/fs/ocfs2/xattr.c index a7c1679..4ddd788 100644 --- a/fs/ocfs2/xattr.c +++ b/fs/ocfs2/xattr.c @@ -82,13 +82,14 @@ struct ocfs2_xattr_set_ctxt { #define OCFS2_XATTR_ROOT_SIZE (sizeof(struct ocfs2_xattr_def_value_root)) #define OCFS2_XATTR_INLINE_SIZE 80 +#define OCFS2_XATTR_HEADER_GAP 4 #define OCFS2_XATTR_FREE_IN_IBODY (OCFS2_MIN_XATTR_INLINE_SIZE \ - sizeof(struct ocfs2_xattr_header) \ - - sizeof(__u32)) + - OCFS2_XATTR_HEADER_GAP) #define OCFS2_XATTR_FREE_IN_BLOCK(ptr) ((ptr)->i_sb->s_blocksize \ - sizeof(struct ocfs2_xattr_block) \ - sizeof(struct ocfs2_xattr_header) \ - - sizeof(__u32)) + - OCFS2_XATTR_HEADER_GAP) static struct ocfs2_xattr_def_value_root def_xv = { .xv.xr_list.l_count = cpu_to_le16(1), @@ -1511,7 +1512,7 @@ static int ocfs2_xattr_set_entry(struct inode *inode, last += 1; } - free = min_offs - ((void *)last - xs->base) - sizeof(__u32); + free = min_offs - ((void *)last - xs->base) - OCFS2_XATTR_HEADER_GAP; if (free < 0) return -EIO; @@ -2194,7 +2195,7 @@ static int ocfs2_xattr_can_be_in_inode(struct inode *inode, last += 1; } - free = min_offs - ((void *)last - xs->base) - sizeof(__u32); + free = min_offs - ((void *)last - xs->base) - OCFS2_XATTR_HEADER_GAP; if (free < 0) return 0; @@ -5065,8 +5066,8 @@ try_again: xh_free_start = le16_to_cpu(xh->xh_free_start); header_size = sizeof(struct ocfs2_xattr_header) + count * sizeof(struct ocfs2_xattr_entry); - max_free = OCFS2_XATTR_BUCKET_SIZE - - le16_to_cpu(xh->xh_name_value_len) - header_size; + max_free = OCFS2_XATTR_BUCKET_SIZE - header_size - + le16_to_cpu(xh->xh_name_value_len) - OCFS2_XATTR_HEADER_GAP; mlog_bug_on_msg(header_size > blocksize, "bucket %llu has header size " "of %u which exceed block size\n", @@ -5099,7 +5100,7 @@ try_again: need = 0; } - free = xh_free_start - header_size; + free = xh_free_start - header_size - OCFS2_XATTR_HEADER_GAP; /* * We need to make sure the new name/value pair * can exist in the same block. @@ -5132,7 +5133,8 @@ try_again: } xh_free_start = le16_to_cpu(xh->xh_free_start); - free = xh_free_start - header_size; + free = xh_free_start - header_size + - OCFS2_XATTR_HEADER_GAP; if (xh_free_start % blocksize < need) free -= xh_free_start % blocksize; -- cgit v0.10.2 From 28d57d437786eb3e44f1ca3f0f41e7cfe29c6dd4 Mon Sep 17 00:00:00 2001 From: wengang wang Date: Fri, 13 Feb 2009 10:11:47 +0800 Subject: ocfs2: add IO error check in ocfs2_get_sector() Check for IO error in ocfs2_get_sector(). Signed-off-by: Wengang Wang Signed-off-by: Mark Fasheh diff --git a/fs/ocfs2/super.c b/fs/ocfs2/super.c index 1c3acc4..7ac83a8 100644 --- a/fs/ocfs2/super.c +++ b/fs/ocfs2/super.c @@ -1537,6 +1537,13 @@ static int ocfs2_get_sector(struct super_block *sb, unlock_buffer(*bh); ll_rw_block(READ, 1, bh); wait_on_buffer(*bh); + if (!buffer_uptodate(*bh)) { + mlog_errno(-EIO); + brelse(*bh); + *bh = NULL; + return -EIO; + } + return 0; } -- cgit v0.10.2 From b50be33e42e2c87812b30aee1a2b2a5ac6cb3ffa Mon Sep 17 00:00:00 2001 From: David Woodhouse Date: Fri, 27 Feb 2009 04:51:33 +0900 Subject: [MTD] [MAPS] Remove MODULE_DEVICE_TABLE() from ck804rom driver. We really don't want the BIOS flash mapping hacks to get automatically loaded. No idea why it isn't using pci_register_driver() though -- that should be fine... and is even _present_ but disabled by #if 0. Signed-off-by: David Woodhouse diff --git a/drivers/mtd/maps/ck804xrom.c b/drivers/mtd/maps/ck804xrom.c index 5f7a245..424f17d 100644 --- a/drivers/mtd/maps/ck804xrom.c +++ b/drivers/mtd/maps/ck804xrom.c @@ -342,9 +342,9 @@ static struct pci_device_id ck804xrom_pci_tbl[] = { { 0, } }; +#if 0 MODULE_DEVICE_TABLE(pci, ck804xrom_pci_tbl); -#if 0 static struct pci_driver ck804xrom_driver = { .name = MOD_NAME, .id_table = ck804xrom_pci_tbl, -- cgit v0.10.2 From f6be37fdc62d0c0214bc49815d1180ebfbd716e2 Mon Sep 17 00:00:00 2001 From: Kyle McMartin Date: Thu, 26 Feb 2009 12:57:56 -0500 Subject: x86: enable DMAR by default Now that the obvious bugs have been worked out, specifically the iwlagn issue, and the write buffer errata, DMAR should be safe to turn back on by default. (We've had it on since those patches were first written a few weeks ago, without any noticeable bug reports (most have been due to the dma-api debug patchset.)) Signed-off-by: Kyle McMartin Acked-by: David Woodhouse Signed-off-by: Ingo Molnar diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig index 9c39095..bc2fbad 100644 --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -1803,7 +1803,7 @@ config DMAR remapping devices. config DMAR_DEFAULT_ON - def_bool n + def_bool y prompt "Enable DMA Remapping Devices by default" depends on DMAR help -- cgit v0.10.2 From ab65f649d38d910f48843a275f3f0596cdbf28bf Mon Sep 17 00:00:00 2001 From: Kiran Divekar Date: Thu, 19 Feb 2009 19:32:39 -0500 Subject: libertas: fix misuse of netdev_priv() and dev->ml_priv The mesh and radiotap interfaces need to use the same private data as the main wifi interface. If the main wifi interface uses netdev_priv(), but the other interfaces ->ml_priv, there's no way to figure out where the private data actually is in the WEXT handlers and netdevice callbacks. So make everything use ->ml_priv. Fixes botched netdev_priv() conversion introduced by "netdevice libertas: Fix directly reference of netdev->priv", though admittedly libertas' use of ->priv was somewhat "special". Signed-off-by: Kiran Divekar Acked-by: Dan Williams Tested-by: Chris Ball Signed-off-by: John W. Linville diff --git a/drivers/net/wireless/libertas/ethtool.c b/drivers/net/wireless/libertas/ethtool.c index 61d2f50..b118a35 100644 --- a/drivers/net/wireless/libertas/ethtool.c +++ b/drivers/net/wireless/libertas/ethtool.c @@ -23,7 +23,7 @@ static const char * mesh_stat_strings[]= { static void lbs_ethtool_get_drvinfo(struct net_device *dev, struct ethtool_drvinfo *info) { - struct lbs_private *priv = netdev_priv(dev); + struct lbs_private *priv = dev->ml_priv; snprintf(info->fw_version, 32, "%u.%u.%u.p%u", priv->fwrelease >> 24 & 0xff, @@ -47,7 +47,7 @@ static int lbs_ethtool_get_eeprom_len(struct net_device *dev) static int lbs_ethtool_get_eeprom(struct net_device *dev, struct ethtool_eeprom *eeprom, u8 * bytes) { - struct lbs_private *priv = netdev_priv(dev); + struct lbs_private *priv = dev->ml_priv; struct cmd_ds_802_11_eeprom_access cmd; int ret; @@ -76,7 +76,7 @@ out: static void lbs_ethtool_get_stats(struct net_device *dev, struct ethtool_stats *stats, uint64_t *data) { - struct lbs_private *priv = netdev_priv(dev); + struct lbs_private *priv = dev->ml_priv; struct cmd_ds_mesh_access mesh_access; int ret; @@ -113,7 +113,7 @@ static void lbs_ethtool_get_stats(struct net_device *dev, static int lbs_ethtool_get_sset_count(struct net_device *dev, int sset) { - struct lbs_private *priv = netdev_priv(dev); + struct lbs_private *priv = dev->ml_priv; if (sset == ETH_SS_STATS && dev == priv->mesh_dev) return MESH_STATS_NUM; @@ -143,7 +143,7 @@ static void lbs_ethtool_get_strings(struct net_device *dev, static void lbs_ethtool_get_wol(struct net_device *dev, struct ethtool_wolinfo *wol) { - struct lbs_private *priv = netdev_priv(dev); + struct lbs_private *priv = dev->ml_priv; if (priv->wol_criteria == 0xffffffff) { /* Interface driver didn't configure wake */ @@ -166,7 +166,7 @@ static void lbs_ethtool_get_wol(struct net_device *dev, static int lbs_ethtool_set_wol(struct net_device *dev, struct ethtool_wolinfo *wol) { - struct lbs_private *priv = netdev_priv(dev); + struct lbs_private *priv = dev->ml_priv; uint32_t criteria = 0; if (priv->wol_criteria == 0xffffffff && wol->wolopts) diff --git a/drivers/net/wireless/libertas/if_usb.c b/drivers/net/wireless/libertas/if_usb.c index 2fc637a..ea3dc038 100644 --- a/drivers/net/wireless/libertas/if_usb.c +++ b/drivers/net/wireless/libertas/if_usb.c @@ -59,7 +59,7 @@ static int if_usb_reset_device(struct if_usb_card *cardp); static ssize_t if_usb_firmware_set(struct device *dev, struct device_attribute *attr, const char *buf, size_t count) { - struct lbs_private *priv = netdev_priv(to_net_dev(dev)); + struct lbs_private *priv = to_net_dev(dev)->ml_priv; struct if_usb_card *cardp = priv->card; char fwname[FIRMWARE_NAME_MAX]; int ret; @@ -86,7 +86,7 @@ static DEVICE_ATTR(lbs_flash_fw, 0200, NULL, if_usb_firmware_set); static ssize_t if_usb_boot2_set(struct device *dev, struct device_attribute *attr, const char *buf, size_t count) { - struct lbs_private *priv = netdev_priv(to_net_dev(dev)); + struct lbs_private *priv = to_net_dev(dev)->ml_priv; struct if_usb_card *cardp = priv->card; char fwname[FIRMWARE_NAME_MAX]; int ret; diff --git a/drivers/net/wireless/libertas/main.c b/drivers/net/wireless/libertas/main.c index 4e0007d..f76623e 100644 --- a/drivers/net/wireless/libertas/main.c +++ b/drivers/net/wireless/libertas/main.c @@ -222,7 +222,7 @@ u8 lbs_data_rate_to_fw_index(u32 rate) static ssize_t lbs_anycast_get(struct device *dev, struct device_attribute *attr, char * buf) { - struct lbs_private *priv = netdev_priv(to_net_dev(dev)); + struct lbs_private *priv = to_net_dev(dev)->ml_priv; struct cmd_ds_mesh_access mesh_access; int ret; @@ -241,7 +241,7 @@ static ssize_t lbs_anycast_get(struct device *dev, static ssize_t lbs_anycast_set(struct device *dev, struct device_attribute *attr, const char * buf, size_t count) { - struct lbs_private *priv = netdev_priv(to_net_dev(dev)); + struct lbs_private *priv = to_net_dev(dev)->ml_priv; struct cmd_ds_mesh_access mesh_access; uint32_t datum; int ret; @@ -263,7 +263,7 @@ static ssize_t lbs_anycast_set(struct device *dev, static ssize_t lbs_prb_rsp_limit_get(struct device *dev, struct device_attribute *attr, char *buf) { - struct lbs_private *priv = netdev_priv(to_net_dev(dev)); + struct lbs_private *priv = to_net_dev(dev)->ml_priv; struct cmd_ds_mesh_access mesh_access; int ret; u32 retry_limit; @@ -286,7 +286,7 @@ static ssize_t lbs_prb_rsp_limit_get(struct device *dev, static ssize_t lbs_prb_rsp_limit_set(struct device *dev, struct device_attribute *attr, const char *buf, size_t count) { - struct lbs_private *priv = netdev_priv(to_net_dev(dev)); + struct lbs_private *priv = to_net_dev(dev)->ml_priv; struct cmd_ds_mesh_access mesh_access; int ret; unsigned long retry_limit; @@ -321,7 +321,7 @@ static void lbs_remove_mesh(struct lbs_private *priv); static ssize_t lbs_rtap_get(struct device *dev, struct device_attribute *attr, char * buf) { - struct lbs_private *priv = netdev_priv(to_net_dev(dev)); + struct lbs_private *priv = to_net_dev(dev)->ml_priv; return snprintf(buf, 5, "0x%X\n", priv->monitormode); } @@ -332,7 +332,7 @@ static ssize_t lbs_rtap_set(struct device *dev, struct device_attribute *attr, const char * buf, size_t count) { int monitor_mode; - struct lbs_private *priv = netdev_priv(to_net_dev(dev)); + struct lbs_private *priv = to_net_dev(dev)->ml_priv; sscanf(buf, "%x", &monitor_mode); if (monitor_mode) { @@ -383,7 +383,7 @@ static DEVICE_ATTR(lbs_rtap, 0644, lbs_rtap_get, lbs_rtap_set ); static ssize_t lbs_mesh_get(struct device *dev, struct device_attribute *attr, char * buf) { - struct lbs_private *priv = netdev_priv(to_net_dev(dev)); + struct lbs_private *priv = to_net_dev(dev)->ml_priv; return snprintf(buf, 5, "0x%X\n", !!priv->mesh_dev); } @@ -393,7 +393,7 @@ static ssize_t lbs_mesh_get(struct device *dev, static ssize_t lbs_mesh_set(struct device *dev, struct device_attribute *attr, const char * buf, size_t count) { - struct lbs_private *priv = netdev_priv(to_net_dev(dev)); + struct lbs_private *priv = to_net_dev(dev)->ml_priv; int enable; int ret, action = CMD_ACT_MESH_CONFIG_STOP; @@ -452,7 +452,7 @@ static struct attribute_group lbs_mesh_attr_group = { */ static int lbs_dev_open(struct net_device *dev) { - struct lbs_private *priv = netdev_priv(dev) ; + struct lbs_private *priv = dev->ml_priv; int ret = 0; lbs_deb_enter(LBS_DEB_NET); @@ -521,7 +521,7 @@ static int lbs_mesh_stop(struct net_device *dev) */ static int lbs_eth_stop(struct net_device *dev) { - struct lbs_private *priv = netdev_priv(dev); + struct lbs_private *priv = dev->ml_priv; lbs_deb_enter(LBS_DEB_NET); @@ -538,7 +538,7 @@ static int lbs_eth_stop(struct net_device *dev) static void lbs_tx_timeout(struct net_device *dev) { - struct lbs_private *priv = netdev_priv(dev); + struct lbs_private *priv = dev->ml_priv; lbs_deb_enter(LBS_DEB_TX); @@ -590,7 +590,7 @@ EXPORT_SYMBOL_GPL(lbs_host_to_card_done); */ static struct net_device_stats *lbs_get_stats(struct net_device *dev) { - struct lbs_private *priv = netdev_priv(dev); + struct lbs_private *priv = dev->ml_priv; lbs_deb_enter(LBS_DEB_NET); return &priv->stats; @@ -599,7 +599,7 @@ static struct net_device_stats *lbs_get_stats(struct net_device *dev) static int lbs_set_mac_address(struct net_device *dev, void *addr) { int ret = 0; - struct lbs_private *priv = netdev_priv(dev); + struct lbs_private *priv = dev->ml_priv; struct sockaddr *phwaddr = addr; struct cmd_ds_802_11_mac_address cmd; @@ -732,7 +732,7 @@ static void lbs_set_mcast_worker(struct work_struct *work) static void lbs_set_multicast_list(struct net_device *dev) { - struct lbs_private *priv = netdev_priv(dev); + struct lbs_private *priv = dev->ml_priv; schedule_work(&priv->mcast_work); } @@ -748,7 +748,7 @@ static void lbs_set_multicast_list(struct net_device *dev) static int lbs_thread(void *data) { struct net_device *dev = data; - struct lbs_private *priv = netdev_priv(dev); + struct lbs_private *priv = dev->ml_priv; wait_queue_t wait; lbs_deb_enter(LBS_DEB_THREAD); @@ -1184,6 +1184,7 @@ struct lbs_private *lbs_add_card(void *card, struct device *dmdev) goto done; } priv = netdev_priv(dev); + dev->ml_priv = priv; if (lbs_init_adapter(priv)) { lbs_pr_err("failed to initialize adapter structure.\n"); diff --git a/drivers/net/wireless/libertas/persistcfg.c b/drivers/net/wireless/libertas/persistcfg.c index d42b7a5..18fe29f 100644 --- a/drivers/net/wireless/libertas/persistcfg.c +++ b/drivers/net/wireless/libertas/persistcfg.c @@ -18,7 +18,7 @@ static int mesh_get_default_parameters(struct device *dev, struct mrvl_mesh_defaults *defs) { - struct lbs_private *priv = netdev_priv(to_net_dev(dev)); + struct lbs_private *priv = to_net_dev(dev)->ml_priv; struct cmd_ds_mesh_config cmd; int ret; @@ -57,7 +57,7 @@ static ssize_t bootflag_get(struct device *dev, static ssize_t bootflag_set(struct device *dev, struct device_attribute *attr, const char *buf, size_t count) { - struct lbs_private *priv = netdev_priv(to_net_dev(dev)); + struct lbs_private *priv = to_net_dev(dev)->ml_priv; struct cmd_ds_mesh_config cmd; uint32_t datum; int ret; @@ -100,7 +100,7 @@ static ssize_t boottime_get(struct device *dev, static ssize_t boottime_set(struct device *dev, struct device_attribute *attr, const char *buf, size_t count) { - struct lbs_private *priv = netdev_priv(to_net_dev(dev)); + struct lbs_private *priv = to_net_dev(dev)->ml_priv; struct cmd_ds_mesh_config cmd; uint32_t datum; int ret; @@ -152,7 +152,7 @@ static ssize_t channel_get(struct device *dev, static ssize_t channel_set(struct device *dev, struct device_attribute *attr, const char *buf, size_t count) { - struct lbs_private *priv = netdev_priv(to_net_dev(dev)); + struct lbs_private *priv = to_net_dev(dev)->ml_priv; struct cmd_ds_mesh_config cmd; uint32_t datum; int ret; @@ -210,7 +210,7 @@ static ssize_t mesh_id_set(struct device *dev, struct device_attribute *attr, struct cmd_ds_mesh_config cmd; struct mrvl_mesh_defaults defs; struct mrvl_meshie *ie; - struct lbs_private *priv = netdev_priv(to_net_dev(dev)); + struct lbs_private *priv = to_net_dev(dev)->ml_priv; int len; int ret; @@ -269,7 +269,7 @@ static ssize_t protocol_id_set(struct device *dev, struct cmd_ds_mesh_config cmd; struct mrvl_mesh_defaults defs; struct mrvl_meshie *ie; - struct lbs_private *priv = netdev_priv(to_net_dev(dev)); + struct lbs_private *priv = to_net_dev(dev)->ml_priv; uint32_t datum; int ret; @@ -323,7 +323,7 @@ static ssize_t metric_id_set(struct device *dev, struct device_attribute *attr, struct cmd_ds_mesh_config cmd; struct mrvl_mesh_defaults defs; struct mrvl_meshie *ie; - struct lbs_private *priv = netdev_priv(to_net_dev(dev)); + struct lbs_private *priv = to_net_dev(dev)->ml_priv; uint32_t datum; int ret; @@ -377,7 +377,7 @@ static ssize_t capability_set(struct device *dev, struct device_attribute *attr, struct cmd_ds_mesh_config cmd; struct mrvl_mesh_defaults defs; struct mrvl_meshie *ie; - struct lbs_private *priv = netdev_priv(to_net_dev(dev)); + struct lbs_private *priv = to_net_dev(dev)->ml_priv; uint32_t datum; int ret; diff --git a/drivers/net/wireless/libertas/scan.c b/drivers/net/wireless/libertas/scan.c index 57f6c12..9014950 100644 --- a/drivers/net/wireless/libertas/scan.c +++ b/drivers/net/wireless/libertas/scan.c @@ -945,7 +945,7 @@ int lbs_set_scan(struct net_device *dev, struct iw_request_info *info, union iwreq_data *wrqu, char *extra) { DECLARE_SSID_BUF(ssid); - struct lbs_private *priv = netdev_priv(dev); + struct lbs_private *priv = dev->ml_priv; int ret = 0; lbs_deb_enter(LBS_DEB_WEXT); @@ -1008,7 +1008,7 @@ int lbs_get_scan(struct net_device *dev, struct iw_request_info *info, struct iw_point *dwrq, char *extra) { #define SCAN_ITEM_SIZE 128 - struct lbs_private *priv = netdev_priv(dev); + struct lbs_private *priv = dev->ml_priv; int err = 0; char *ev = extra; char *stop = ev + dwrq->length; diff --git a/drivers/net/wireless/libertas/tx.c b/drivers/net/wireless/libertas/tx.c index dac4626..68bec31 100644 --- a/drivers/net/wireless/libertas/tx.c +++ b/drivers/net/wireless/libertas/tx.c @@ -60,7 +60,7 @@ static u32 convert_radiotap_rate_to_mv(u8 rate) int lbs_hard_start_xmit(struct sk_buff *skb, struct net_device *dev) { unsigned long flags; - struct lbs_private *priv = netdev_priv(dev); + struct lbs_private *priv = dev->ml_priv; struct txpd *txpd; char *p802x_hdr; uint16_t pkt_len; diff --git a/drivers/net/wireless/libertas/wext.c b/drivers/net/wireless/libertas/wext.c index c6102e0..f16d136 100644 --- a/drivers/net/wireless/libertas/wext.c +++ b/drivers/net/wireless/libertas/wext.c @@ -163,7 +163,7 @@ static int lbs_get_name(struct net_device *dev, struct iw_request_info *info, static int lbs_get_freq(struct net_device *dev, struct iw_request_info *info, struct iw_freq *fwrq, char *extra) { - struct lbs_private *priv = netdev_priv(dev); + struct lbs_private *priv = dev->ml_priv; struct chan_freq_power *cfp; lbs_deb_enter(LBS_DEB_WEXT); @@ -189,7 +189,7 @@ static int lbs_get_freq(struct net_device *dev, struct iw_request_info *info, static int lbs_get_wap(struct net_device *dev, struct iw_request_info *info, struct sockaddr *awrq, char *extra) { - struct lbs_private *priv = netdev_priv(dev); + struct lbs_private *priv = dev->ml_priv; lbs_deb_enter(LBS_DEB_WEXT); @@ -207,7 +207,7 @@ static int lbs_get_wap(struct net_device *dev, struct iw_request_info *info, static int lbs_set_nick(struct net_device *dev, struct iw_request_info *info, struct iw_point *dwrq, char *extra) { - struct lbs_private *priv = netdev_priv(dev); + struct lbs_private *priv = dev->ml_priv; lbs_deb_enter(LBS_DEB_WEXT); @@ -231,7 +231,7 @@ static int lbs_set_nick(struct net_device *dev, struct iw_request_info *info, static int lbs_get_nick(struct net_device *dev, struct iw_request_info *info, struct iw_point *dwrq, char *extra) { - struct lbs_private *priv = netdev_priv(dev); + struct lbs_private *priv = dev->ml_priv; lbs_deb_enter(LBS_DEB_WEXT); @@ -248,7 +248,7 @@ static int lbs_get_nick(struct net_device *dev, struct iw_request_info *info, static int mesh_get_nick(struct net_device *dev, struct iw_request_info *info, struct iw_point *dwrq, char *extra) { - struct lbs_private *priv = netdev_priv(dev); + struct lbs_private *priv = dev->ml_priv; lbs_deb_enter(LBS_DEB_WEXT); @@ -273,7 +273,7 @@ static int lbs_set_rts(struct net_device *dev, struct iw_request_info *info, struct iw_param *vwrq, char *extra) { int ret = 0; - struct lbs_private *priv = netdev_priv(dev); + struct lbs_private *priv = dev->ml_priv; u32 val = vwrq->value; lbs_deb_enter(LBS_DEB_WEXT); @@ -293,7 +293,7 @@ static int lbs_set_rts(struct net_device *dev, struct iw_request_info *info, static int lbs_get_rts(struct net_device *dev, struct iw_request_info *info, struct iw_param *vwrq, char *extra) { - struct lbs_private *priv = netdev_priv(dev); + struct lbs_private *priv = dev->ml_priv; int ret = 0; u16 val = 0; @@ -315,7 +315,7 @@ out: static int lbs_set_frag(struct net_device *dev, struct iw_request_info *info, struct iw_param *vwrq, char *extra) { - struct lbs_private *priv = netdev_priv(dev); + struct lbs_private *priv = dev->ml_priv; int ret = 0; u32 val = vwrq->value; @@ -336,7 +336,7 @@ static int lbs_set_frag(struct net_device *dev, struct iw_request_info *info, static int lbs_get_frag(struct net_device *dev, struct iw_request_info *info, struct iw_param *vwrq, char *extra) { - struct lbs_private *priv = netdev_priv(dev); + struct lbs_private *priv = dev->ml_priv; int ret = 0; u16 val = 0; @@ -359,7 +359,7 @@ out: static int lbs_get_mode(struct net_device *dev, struct iw_request_info *info, u32 * uwrq, char *extra) { - struct lbs_private *priv = netdev_priv(dev); + struct lbs_private *priv = dev->ml_priv; lbs_deb_enter(LBS_DEB_WEXT); @@ -385,7 +385,7 @@ static int lbs_get_txpow(struct net_device *dev, struct iw_request_info *info, struct iw_param *vwrq, char *extra) { - struct lbs_private *priv = netdev_priv(dev); + struct lbs_private *priv = dev->ml_priv; s16 curlevel = 0; int ret = 0; @@ -418,7 +418,7 @@ out: static int lbs_set_retry(struct net_device *dev, struct iw_request_info *info, struct iw_param *vwrq, char *extra) { - struct lbs_private *priv = netdev_priv(dev); + struct lbs_private *priv = dev->ml_priv; int ret = 0; u16 slimit = 0, llimit = 0; @@ -466,7 +466,7 @@ out: static int lbs_get_retry(struct net_device *dev, struct iw_request_info *info, struct iw_param *vwrq, char *extra) { - struct lbs_private *priv = netdev_priv(dev); + struct lbs_private *priv = dev->ml_priv; int ret = 0; u16 val = 0; @@ -542,7 +542,7 @@ static int lbs_get_range(struct net_device *dev, struct iw_request_info *info, struct iw_point *dwrq, char *extra) { int i, j; - struct lbs_private *priv = netdev_priv(dev); + struct lbs_private *priv = dev->ml_priv; struct iw_range *range = (struct iw_range *)extra; struct chan_freq_power *cfp; u8 rates[MAX_RATES + 1]; @@ -708,7 +708,7 @@ out: static int lbs_set_power(struct net_device *dev, struct iw_request_info *info, struct iw_param *vwrq, char *extra) { - struct lbs_private *priv = netdev_priv(dev); + struct lbs_private *priv = dev->ml_priv; lbs_deb_enter(LBS_DEB_WEXT); @@ -758,7 +758,7 @@ static int lbs_set_power(struct net_device *dev, struct iw_request_info *info, static int lbs_get_power(struct net_device *dev, struct iw_request_info *info, struct iw_param *vwrq, char *extra) { - struct lbs_private *priv = netdev_priv(dev); + struct lbs_private *priv = dev->ml_priv; lbs_deb_enter(LBS_DEB_WEXT); @@ -781,7 +781,7 @@ static struct iw_statistics *lbs_get_wireless_stats(struct net_device *dev) EXCELLENT = 95, PERFECT = 100 }; - struct lbs_private *priv = netdev_priv(dev); + struct lbs_private *priv = dev->ml_priv; u32 rssi_qual; u32 tx_qual; u32 quality = 0; @@ -886,7 +886,7 @@ static int lbs_set_freq(struct net_device *dev, struct iw_request_info *info, struct iw_freq *fwrq, char *extra) { int ret = -EINVAL; - struct lbs_private *priv = netdev_priv(dev); + struct lbs_private *priv = dev->ml_priv; struct chan_freq_power *cfp; struct assoc_request * assoc_req; @@ -943,7 +943,7 @@ static int lbs_mesh_set_freq(struct net_device *dev, struct iw_request_info *info, struct iw_freq *fwrq, char *extra) { - struct lbs_private *priv = netdev_priv(dev); + struct lbs_private *priv = dev->ml_priv; struct chan_freq_power *cfp; int ret = -EINVAL; @@ -994,7 +994,7 @@ out: static int lbs_set_rate(struct net_device *dev, struct iw_request_info *info, struct iw_param *vwrq, char *extra) { - struct lbs_private *priv = netdev_priv(dev); + struct lbs_private *priv = dev->ml_priv; u8 new_rate = 0; int ret = -EINVAL; u8 rates[MAX_RATES + 1]; @@ -1054,7 +1054,7 @@ out: static int lbs_get_rate(struct net_device *dev, struct iw_request_info *info, struct iw_param *vwrq, char *extra) { - struct lbs_private *priv = netdev_priv(dev); + struct lbs_private *priv = dev->ml_priv; lbs_deb_enter(LBS_DEB_WEXT); @@ -1079,7 +1079,7 @@ static int lbs_set_mode(struct net_device *dev, struct iw_request_info *info, u32 * uwrq, char *extra) { int ret = 0; - struct lbs_private *priv = netdev_priv(dev); + struct lbs_private *priv = dev->ml_priv; struct assoc_request * assoc_req; lbs_deb_enter(LBS_DEB_WEXT); @@ -1124,7 +1124,7 @@ static int lbs_get_encode(struct net_device *dev, struct iw_request_info *info, struct iw_point *dwrq, u8 * extra) { - struct lbs_private *priv = netdev_priv(dev); + struct lbs_private *priv = dev->ml_priv; int index = (dwrq->flags & IW_ENCODE_INDEX) - 1; lbs_deb_enter(LBS_DEB_WEXT); @@ -1319,7 +1319,7 @@ static int lbs_set_encode(struct net_device *dev, struct iw_point *dwrq, char *extra) { int ret = 0; - struct lbs_private *priv = netdev_priv(dev); + struct lbs_private *priv = dev->ml_priv; struct assoc_request * assoc_req; u16 is_default = 0, index = 0, set_tx_key = 0; @@ -1395,7 +1395,7 @@ static int lbs_get_encodeext(struct net_device *dev, char *extra) { int ret = -EINVAL; - struct lbs_private *priv = netdev_priv(dev); + struct lbs_private *priv = dev->ml_priv; struct iw_encode_ext *ext = (struct iw_encode_ext *)extra; int index, max_key_len; @@ -1501,7 +1501,7 @@ static int lbs_set_encodeext(struct net_device *dev, char *extra) { int ret = 0; - struct lbs_private *priv = netdev_priv(dev); + struct lbs_private *priv = dev->ml_priv; struct iw_encode_ext *ext = (struct iw_encode_ext *)extra; int alg = ext->alg; struct assoc_request * assoc_req; @@ -1639,7 +1639,7 @@ static int lbs_set_genie(struct net_device *dev, struct iw_point *dwrq, char *extra) { - struct lbs_private *priv = netdev_priv(dev); + struct lbs_private *priv = dev->ml_priv; int ret = 0; struct assoc_request * assoc_req; @@ -1685,7 +1685,7 @@ static int lbs_get_genie(struct net_device *dev, char *extra) { int ret = 0; - struct lbs_private *priv = netdev_priv(dev); + struct lbs_private *priv = dev->ml_priv; lbs_deb_enter(LBS_DEB_WEXT); @@ -1713,7 +1713,7 @@ static int lbs_set_auth(struct net_device *dev, struct iw_param *dwrq, char *extra) { - struct lbs_private *priv = netdev_priv(dev); + struct lbs_private *priv = dev->ml_priv; struct assoc_request * assoc_req; int ret = 0; int updated = 0; @@ -1816,7 +1816,7 @@ static int lbs_get_auth(struct net_device *dev, char *extra) { int ret = 0; - struct lbs_private *priv = netdev_priv(dev); + struct lbs_private *priv = dev->ml_priv; lbs_deb_enter(LBS_DEB_WEXT); @@ -1857,7 +1857,7 @@ static int lbs_set_txpow(struct net_device *dev, struct iw_request_info *info, struct iw_param *vwrq, char *extra) { int ret = 0; - struct lbs_private *priv = netdev_priv(dev); + struct lbs_private *priv = dev->ml_priv; s16 dbm = (s16) vwrq->value; lbs_deb_enter(LBS_DEB_WEXT); @@ -1936,7 +1936,7 @@ out: static int lbs_get_essid(struct net_device *dev, struct iw_request_info *info, struct iw_point *dwrq, char *extra) { - struct lbs_private *priv = netdev_priv(dev); + struct lbs_private *priv = dev->ml_priv; lbs_deb_enter(LBS_DEB_WEXT); @@ -1971,7 +1971,7 @@ static int lbs_get_essid(struct net_device *dev, struct iw_request_info *info, static int lbs_set_essid(struct net_device *dev, struct iw_request_info *info, struct iw_point *dwrq, char *extra) { - struct lbs_private *priv = netdev_priv(dev); + struct lbs_private *priv = dev->ml_priv; int ret = 0; u8 ssid[IW_ESSID_MAX_SIZE]; u8 ssid_len = 0; @@ -2040,7 +2040,7 @@ static int lbs_mesh_get_essid(struct net_device *dev, struct iw_request_info *info, struct iw_point *dwrq, char *extra) { - struct lbs_private *priv = netdev_priv(dev); + struct lbs_private *priv = dev->ml_priv; lbs_deb_enter(LBS_DEB_WEXT); @@ -2058,7 +2058,7 @@ static int lbs_mesh_set_essid(struct net_device *dev, struct iw_request_info *info, struct iw_point *dwrq, char *extra) { - struct lbs_private *priv = netdev_priv(dev); + struct lbs_private *priv = dev->ml_priv; int ret = 0; lbs_deb_enter(LBS_DEB_WEXT); @@ -2102,7 +2102,7 @@ static int lbs_mesh_set_essid(struct net_device *dev, static int lbs_set_wap(struct net_device *dev, struct iw_request_info *info, struct sockaddr *awrq, char *extra) { - struct lbs_private *priv = netdev_priv(dev); + struct lbs_private *priv = dev->ml_priv; struct assoc_request * assoc_req; int ret = 0; -- cgit v0.10.2 From bbe194433baeadc953f49e3795b41ffffc5486dd Mon Sep 17 00:00:00 2001 From: Stefan Assmann Date: Thu, 26 Feb 2009 10:46:48 -0800 Subject: PCI: AMD 813x B2 devices do not need boot interrupt quirk Turns out that the new AMD 813x devices do not need the quirk_disable_amd_813x_boot_interrupt quirk to be run on them. If it is, no interrupts are seen on the PCI-X adapter. From: Stefan Assmann Reported-by: Jamie Wellnitz Tested-by: Jamie Wellnitz Cc: stable Signed-off-by: Greg Kroah-Hartman Signed-off-by: Jesse Barnes diff --git a/drivers/pci/quirks.c b/drivers/pci/quirks.c index a852329..f20d553 100644 --- a/drivers/pci/quirks.c +++ b/drivers/pci/quirks.c @@ -1584,6 +1584,7 @@ DECLARE_PCI_FIXUP_RESUME(PCI_VENDOR_ID_SERVERWORKS, PCI_DEVICE_ID_SERVERWORKS_ */ #define AMD_813X_MISC 0x40 #define AMD_813X_NOIOAMODE (1<<0) +#define AMD_813X_REV_B2 0x13 static void quirk_disable_amd_813x_boot_interrupt(struct pci_dev *dev) { @@ -1591,6 +1592,8 @@ static void quirk_disable_amd_813x_boot_interrupt(struct pci_dev *dev) if (noioapicquirk) return; + if (dev->revision == AMD_813X_REV_B2) + return; pci_read_config_dword(dev, AMD_813X_MISC, &pci_config_dword); pci_config_dword &= ~AMD_813X_NOIOAMODE; -- cgit v0.10.2 From 1ac00cc21337b0b667493d9af79d88537de90aa3 Mon Sep 17 00:00:00 2001 From: Benjamin Herrenschmidt Date: Sun, 1 Feb 2009 14:24:18 +0000 Subject: powerpc/44x: Fix address decoding setup of PCI 2.x cells The PCI 2.x cells used on some 44x SoCs only let us configure the decode for the low 32-bit of the incoming PLB addresses. The top 4 bits (this is a 36-bit bus) are hard wired to different values depending on the specific SoC in use. Our code used to work "by accident" until I added support for the ISA memory holes and while at it added more validity checking of the addresses. This patch should bring it back to working condition. It still relies on the device-tree being correct but that's somewhat a pre-requisite for anything to work anyway. Signed-off-by: Benjamin Herrenschmidt Acked-by: Geert Uytterhoeven Acked-by: Josh Boyer diff --git a/arch/powerpc/sysdev/ppc4xx_pci.c b/arch/powerpc/sysdev/ppc4xx_pci.c index 77fae5f..5558d93 100644 --- a/arch/powerpc/sysdev/ppc4xx_pci.c +++ b/arch/powerpc/sysdev/ppc4xx_pci.c @@ -204,6 +204,23 @@ static int __init ppc4xx_setup_one_pci_PMM(struct pci_controller *hose, { u32 ma, pcila, pciha; + /* Hack warning ! The "old" PCI 2.x cell only let us configure the low + * 32-bit of incoming PLB addresses. The top 4 bits of the 36-bit + * address are actually hard wired to a value that appears to depend + * on the specific SoC. For example, it's 0 on 440EP and 1 on 440EPx. + * + * The trick here is we just crop those top bits and ignore them when + * programming the chip. That means the device-tree has to be right + * for the specific part used (we don't print a warning if it's wrong + * but on the other hand, you'll crash quickly enough), but at least + * this code should work whatever the hard coded value is + */ + plb_addr &= 0xffffffffull; + + /* Note: Due to the above hack, the test below doesn't actually test + * if you address is above 4G, but it tests that address and + * (address + size) are both contained in the same 4G + */ if ((plb_addr + size) > 0xffffffffull || !is_power_of_2(size) || size < 0x1000 || (plb_addr & (size - 1)) != 0) { printk(KERN_WARNING "%s: Resource out of range\n", -- cgit v0.10.2 From 7662b00c378fe638e84a853418cd833303fc050c Mon Sep 17 00:00:00 2001 From: Nicola Soranzo Date: Thu, 19 Feb 2009 13:41:56 -0300 Subject: V4L/DVB (10659): em28xx: register device to soundcard for sysfs As explained in "Writing an ALSA driver" (T. Iwai), audio drivers should set the struct device for the card before registering the card instance. This will add the correct /sys/class/sound/cardN/device symlink, so HAL can see the device and ConsoleKit sets its ACL permissions for the logged-in user. For em28xx audio capture cards found e.g. in Hauppauge WinTV-HVR-900 (R2), this patch fixes errors like: ALSA lib pcm_hw.c:1429:(_snd_pcm_hw_open) Invalid value for card Error opening audio: Permission denied when running mplayer as a normal user. Signed-off-by: Nicola Soranzo Signed-off-by: Mauro Carvalho Chehab diff --git a/drivers/media/video/em28xx/em28xx-audio.c b/drivers/media/video/em28xx/em28xx-audio.c index 5d882a4..2ac738f 100644 --- a/drivers/media/video/em28xx/em28xx-audio.c +++ b/drivers/media/video/em28xx/em28xx-audio.c @@ -463,6 +463,8 @@ static int em28xx_audio_init(struct em28xx *dev) pcm->info_flags = 0; pcm->private_data = dev; strcpy(pcm->name, "Empia 28xx Capture"); + + snd_card_set_dev(card, &dev->udev->dev); strcpy(card->driver, "Empia Em28xx Audio"); strcpy(card->shortname, "Em28xx Audio"); strcpy(card->longname, "Empia Em28xx Audio"); -- cgit v0.10.2 From 0ad675eb4533402fd7b03b25d1d4a0ab7a43ae6d Mon Sep 17 00:00:00 2001 From: Guennadi Liakhovetski Date: Mon, 23 Feb 2009 12:11:25 -0300 Subject: V4L/DVB (10663): soc-camera: fix S_CROP breakage on PXA and SuperH Recent format-negotiation patches caused S_CROP breakage in pxa_camera.c and sh_mobile_ceu_camera.c drivers, fix it. Signed-off-by: Guennadi Liakhovetski Signed-off-by: Mauro Carvalho Chehab diff --git a/drivers/media/video/pxa_camera.c b/drivers/media/video/pxa_camera.c index a1d6008..07c334f 100644 --- a/drivers/media/video/pxa_camera.c +++ b/drivers/media/video/pxa_camera.c @@ -1155,23 +1155,23 @@ static int pxa_camera_set_fmt(struct soc_camera_device *icd, { struct soc_camera_host *ici = to_soc_camera_host(icd->dev.parent); struct pxa_camera_dev *pcdev = ici->priv; - const struct soc_camera_data_format *host_fmt, *cam_fmt = NULL; - const struct soc_camera_format_xlate *xlate; + const struct soc_camera_data_format *cam_fmt = NULL; + const struct soc_camera_format_xlate *xlate = NULL; struct soc_camera_sense sense = { .master_clock = pcdev->mclk, .pixel_clock_max = pcdev->ciclk / 4, }; - int ret, buswidth; + int ret; - xlate = soc_camera_xlate_by_fourcc(icd, pixfmt); - if (!xlate) { - dev_warn(&ici->dev, "Format %x not found\n", pixfmt); - return -EINVAL; - } + if (pixfmt) { + xlate = soc_camera_xlate_by_fourcc(icd, pixfmt); + if (!xlate) { + dev_warn(&ici->dev, "Format %x not found\n", pixfmt); + return -EINVAL; + } - buswidth = xlate->buswidth; - host_fmt = xlate->host_fmt; - cam_fmt = xlate->cam_fmt; + cam_fmt = xlate->cam_fmt; + } /* If PCLK is used to latch data from the sensor, check sense */ if (pcdev->platform_flags & PXA_CAMERA_PCLK_EN) @@ -1201,8 +1201,8 @@ static int pxa_camera_set_fmt(struct soc_camera_device *icd, } if (pixfmt && !ret) { - icd->buswidth = buswidth; - icd->current_fmt = host_fmt; + icd->buswidth = xlate->buswidth; + icd->current_fmt = xlate->host_fmt; } return ret; diff --git a/drivers/media/video/sh_mobile_ceu_camera.c b/drivers/media/video/sh_mobile_ceu_camera.c index 9a2586b..ddcb81d 100644 --- a/drivers/media/video/sh_mobile_ceu_camera.c +++ b/drivers/media/video/sh_mobile_ceu_camera.c @@ -603,21 +603,18 @@ static int sh_mobile_ceu_set_fmt(struct soc_camera_device *icd, const struct soc_camera_format_xlate *xlate; int ret; + if (!pixfmt) + return icd->ops->set_fmt(icd, pixfmt, rect); + xlate = soc_camera_xlate_by_fourcc(icd, pixfmt); if (!xlate) { dev_warn(&ici->dev, "Format %x not found\n", pixfmt); return -EINVAL; } - switch (pixfmt) { - case 0: /* Only geometry change */ - ret = icd->ops->set_fmt(icd, pixfmt, rect); - break; - default: - ret = icd->ops->set_fmt(icd, xlate->cam_fmt->fourcc, rect); - } + ret = icd->ops->set_fmt(icd, xlate->cam_fmt->fourcc, rect); - if (pixfmt && !ret) { + if (!ret) { icd->buswidth = xlate->buswidth; icd->current_fmt = xlate->host_fmt; pcdev->camera_fmt = xlate->cam_fmt; -- cgit v0.10.2 From 382c5546d618f24dc7d6ae7ca33412083720efbf Mon Sep 17 00:00:00 2001 From: Patrick Boettcher Date: Mon, 23 Feb 2009 06:27:16 -0300 Subject: V4L/DVB (10694): [PATCH] software IRQ watchdog for Flexcop B2C2 DVB PCI cards With (some) Technisat cards you cannot run multiple DVB applications in parallel and switch the channel at the same time. There seems to be a problem on the interfaces or even inside the flexcop-device that can't handle interruption on the streaming interface. This patch adds a watchdog to check whether data is supposed to come in (streaming PIDs are requested) and if no data is seen within 400ms (default) it resets the streaming/pid-filtering hardware. This patch is urgently needed to support the rev 2.8 of the hardware and solves problem occassionally seen on older hardware. Signed-off-by: Uwe Bugla Signed-off-by: Patrick Boettcher Signed-off-by: Mauro Carvalho Chehab diff --git a/drivers/media/dvb/b2c2/flexcop-hw-filter.c b/drivers/media/dvb/b2c2/flexcop-hw-filter.c index b386cc6..451974b 100644 --- a/drivers/media/dvb/b2c2/flexcop-hw-filter.c +++ b/drivers/media/dvb/b2c2/flexcop-hw-filter.c @@ -192,6 +192,7 @@ int flexcop_pid_feed_control(struct flexcop_device *fc, struct dvb_demux_feed *d return 0; } +EXPORT_SYMBOL(flexcop_pid_feed_control); void flexcop_hw_filter_init(struct flexcop_device *fc) { diff --git a/drivers/media/dvb/b2c2/flexcop-pci.c b/drivers/media/dvb/b2c2/flexcop-pci.c index 5b30dfc..76e37fd 100644 --- a/drivers/media/dvb/b2c2/flexcop-pci.c +++ b/drivers/media/dvb/b2c2/flexcop-pci.c @@ -13,9 +13,9 @@ static int enable_pid_filtering = 1; module_param(enable_pid_filtering, int, 0444); MODULE_PARM_DESC(enable_pid_filtering, "enable hardware pid filtering: supported values: 0 (fullts), 1"); -static int irq_chk_intv; +static int irq_chk_intv = 100; module_param(irq_chk_intv, int, 0644); -MODULE_PARM_DESC(irq_chk_intv, "set the interval for IRQ watchdog (currently just debugging)."); +MODULE_PARM_DESC(irq_chk_intv, "set the interval for IRQ streaming watchdog."); #ifdef CONFIG_DVB_B2C2_FLEXCOP_DEBUG #define dprintk(level,args...) \ @@ -34,7 +34,9 @@ MODULE_PARM_DESC(irq_chk_intv, "set the interval for IRQ watchdog (currently jus static int debug; module_param(debug, int, 0644); -MODULE_PARM_DESC(debug, "set debug level (1=info,2=regs,4=TS,8=irqdma (|-able))." DEBSTATUS); +MODULE_PARM_DESC(debug, + "set debug level (1=info,2=regs,4=TS,8=irqdma,16=check (|-able))." + DEBSTATUS); #define DRIVER_VERSION "0.1" #define DRIVER_NAME "Technisat/B2C2 FlexCop II/IIb/III Digital TV PCI Driver" @@ -58,6 +60,8 @@ struct flexcop_pci { int active_dma1_addr; /* 0 = addr0 of dma1; 1 = addr1 of dma1 */ u32 last_dma1_cur_pos; /* position of the pointer last time the timer/packet irq occured */ int count; + int count_prev; + int stream_problem; spinlock_t irq_lock; @@ -103,18 +107,32 @@ static void flexcop_pci_irq_check_work(struct work_struct *work) container_of(work, struct flexcop_pci, irq_check_work.work); struct flexcop_device *fc = fc_pci->fc_dev; - flexcop_ibi_value v = fc->read_ibi_reg(fc,sram_dest_reg_714); - - flexcop_dump_reg(fc_pci->fc_dev,dma1_000,4); - - if (v.sram_dest_reg_714.net_ovflow_error) - deb_chk("sram net_ovflow_error\n"); - if (v.sram_dest_reg_714.media_ovflow_error) - deb_chk("sram media_ovflow_error\n"); - if (v.sram_dest_reg_714.cai_ovflow_error) - deb_chk("sram cai_ovflow_error\n"); - if (v.sram_dest_reg_714.cai_ovflow_error) - deb_chk("sram cai_ovflow_error\n"); + if (fc->feedcount) { + + if (fc_pci->count == fc_pci->count_prev) { + deb_chk("no IRQ since the last check\n"); + if (fc_pci->stream_problem++ == 3) { + struct dvb_demux_feed *feed; + + spin_lock_irq(&fc->demux.lock); + list_for_each_entry(feed, &fc->demux.feed_list, + list_head) { + flexcop_pid_feed_control(fc, feed, 0); + } + + list_for_each_entry(feed, &fc->demux.feed_list, + list_head) { + flexcop_pid_feed_control(fc, feed, 1); + } + spin_unlock_irq(&fc->demux.lock); + + fc_pci->stream_problem = 0; + } + } else { + fc_pci->stream_problem = 0; + fc_pci->count_prev = fc_pci->count; + } + } schedule_delayed_work(&fc_pci->irq_check_work, msecs_to_jiffies(irq_chk_intv < 100 ? 100 : irq_chk_intv)); @@ -216,16 +234,12 @@ static int flexcop_pci_stream_control(struct flexcop_device *fc, int onoff) flexcop_dma_control_timer_irq(fc,FC_DMA_1,1); deb_irq("IRQ enabled\n"); + fc_pci->count_prev = fc_pci->count; + // fc_pci->active_dma1_addr = 0; // flexcop_dma_control_size_irq(fc,FC_DMA_1,1); - if (irq_chk_intv > 0) - schedule_delayed_work(&fc_pci->irq_check_work, - msecs_to_jiffies(irq_chk_intv < 100 ? 100 : irq_chk_intv)); } else { - if (irq_chk_intv > 0) - cancel_delayed_work(&fc_pci->irq_check_work); - flexcop_dma_control_timer_irq(fc,FC_DMA_1,0); deb_irq("IRQ disabled\n"); @@ -299,8 +313,6 @@ static int flexcop_pci_init(struct flexcop_pci *fc_pci) IRQF_SHARED, DRIVER_NAME, fc_pci)) != 0) goto err_pci_iounmap; - - fc_pci->init_state |= FC_PCI_INIT; return ret; @@ -375,6 +387,10 @@ static int flexcop_pci_probe(struct pci_dev *pdev, const struct pci_device_id *e INIT_DELAYED_WORK(&fc_pci->irq_check_work, flexcop_pci_irq_check_work); + if (irq_chk_intv > 0) + schedule_delayed_work(&fc_pci->irq_check_work, + msecs_to_jiffies(irq_chk_intv < 100 ? 100 : irq_chk_intv)); + return ret; err_fc_exit: @@ -393,6 +409,9 @@ static void flexcop_pci_remove(struct pci_dev *pdev) { struct flexcop_pci *fc_pci = pci_get_drvdata(pdev); + if (irq_chk_intv > 0) + cancel_delayed_work(&fc_pci->irq_check_work); + flexcop_pci_dma_exit(fc_pci); flexcop_device_exit(fc_pci->fc_dev); flexcop_pci_exit(fc_pci); diff --git a/drivers/media/dvb/b2c2/flexcop.c b/drivers/media/dvb/b2c2/flexcop.c index 676413a..9106895 100644 --- a/drivers/media/dvb/b2c2/flexcop.c +++ b/drivers/media/dvb/b2c2/flexcop.c @@ -212,8 +212,7 @@ void flexcop_reset_block_300(struct flexcop_device *fc) v210.sw_reset_210.Block_reset_enable = 0xb2; fc->write_ibi_reg(fc,sw_reset_210,v210); - msleep(1); - + udelay(1000); fc->write_ibi_reg(fc,ctrl_208,v208_save); } -- cgit v0.10.2 From f5a3258335eef2baf534243da00bcd0e760dd2b5 Mon Sep 17 00:00:00 2001 From: Uwe Bugla Date: Mon, 23 Feb 2009 10:54:20 -0300 Subject: V4L/DVB (10695): Update Technisat card documentation Fixes for documentation of Technisat-based cards Signed-off-by: Uwe Bugla Signed-off-by: Patrick Boettcher Signed-off-by: Mauro Carvalho Chehab diff --git a/Documentation/dvb/technisat.txt b/Documentation/dvb/technisat.txt index cdf6ee4..3f435ff 100644 --- a/Documentation/dvb/technisat.txt +++ b/Documentation/dvb/technisat.txt @@ -1,5 +1,5 @@ -How to set up the Technisat devices -=================================== +How to set up the Technisat/B2C2 Flexcop devices +================================================ 1) Find out what device you have ================================ @@ -16,54 +16,60 @@ DVB: registering frontend 0 (Conexant CX24123/CX24109)... If the Technisat is the only TV device in your box get rid of unnecessary modules and check this one: "Multimedia devices" => "Customise analog and hybrid tuner modules to build" -In this directory uncheck every driver which is activated there. +In this directory uncheck every driver which is activated there (except "Simple tuner support" for case 9 only). Then please activate: 2a) Main module part: a.)"Multimedia devices" => "DVB/ATSC adapters" => "Technisat/B2C2 FlexcopII(b) and FlexCopIII adapters" -b.)"Multimedia devices" => "DVB/ATSC adapters" => "Technisat/B2C2 FlexcopII(b) and FlexCopIII adapters" => "Technisat/B2C2 Air/Sky/Cable2PC PCI" in case of a PCI card OR +b.)"Multimedia devices" => "DVB/ATSC adapters" => "Technisat/B2C2 FlexcopII(b) and FlexCopIII adapters" => "Technisat/B2C2 Air/Sky/Cable2PC PCI" in case of a PCI card +OR c.)"Multimedia devices" => "DVB/ATSC adapters" => "Technisat/B2C2 FlexcopII(b) and FlexCopIII adapters" => "Technisat/B2C2 Air/Sky/Cable2PC USB" in case of an USB 1.1 adapter d.)"Multimedia devices" => "DVB/ATSC adapters" => "Technisat/B2C2 FlexcopII(b) and FlexCopIII adapters" => "Enable debug for the B2C2 FlexCop drivers" Notice: d.) is helpful for troubleshooting 2b) Frontend module part: -1.) Revision 2.3: +1.) SkyStar DVB-S Revision 2.3: a.)"Multimedia devices" => "Customise DVB frontends" => "Customise the frontend modules to build" b.)"Multimedia devices" => "Customise DVB frontends" => "Zarlink VP310/MT312/ZL10313 based" -2.) Revision 2.6: +2.) SkyStar DVB-S Revision 2.6: a.)"Multimedia devices" => "Customise DVB frontends" => "Customise the frontend modules to build" b.)"Multimedia devices" => "Customise DVB frontends" => "ST STV0299 based" -3.) Revision 2.7: +3.) SkyStar DVB-S Revision 2.7: a.)"Multimedia devices" => "Customise DVB frontends" => "Customise the frontend modules to build" b.)"Multimedia devices" => "Customise DVB frontends" => "Samsung S5H1420 based" c.)"Multimedia devices" => "Customise DVB frontends" => "Integrant ITD1000 Zero IF tuner for DVB-S/DSS" d.)"Multimedia devices" => "Customise DVB frontends" => "ISL6421 SEC controller" -4.) Revision 2.8: +4.) SkyStar DVB-S Revision 2.8: a.)"Multimedia devices" => "Customise DVB frontends" => "Customise the frontend modules to build" b.)"Multimedia devices" => "Customise DVB frontends" => "Conexant CX24113/CX24128 tuner for DVB-S/DSS" c.)"Multimedia devices" => "Customise DVB frontends" => "Conexant CX24123 based" d.)"Multimedia devices" => "Customise DVB frontends" => "ISL6421 SEC controller" -5.) DVB-T card: +5.) AirStar DVB-T card: a.)"Multimedia devices" => "Customise DVB frontends" => "Customise the frontend modules to build" b.)"Multimedia devices" => "Customise DVB frontends" => "Zarlink MT352 based" -6.) DVB-C card: +6.) CableStar DVB-C card: a.)"Multimedia devices" => "Customise DVB frontends" => "Customise the frontend modules to build" b.)"Multimedia devices" => "Customise DVB frontends" => "ST STV0297 based" -7.) ATSC card 1st generation: +7.) AirStar ATSC card 1st generation: a.)"Multimedia devices" => "Customise DVB frontends" => "Customise the frontend modules to build" b.)"Multimedia devices" => "Customise DVB frontends" => "Broadcom BCM3510" -8.) ATSC card 2nd generation: +8.) AirStar ATSC card 2nd generation: a.)"Multimedia devices" => "Customise DVB frontends" => "Customise the frontend modules to build" b.)"Multimedia devices" => "Customise DVB frontends" => "NxtWave Communications NXT2002/NXT2004 based" -c.)"Multimedia devices" => "Customise DVB frontends" => "LG Electronics LGDT3302/LGDT3303 based" +c.)"Multimedia devices" => "Customise DVB frontends" => "Generic I2C PLL based tuners" -Author: Uwe Bugla December 2008 +9.) AirStar ATSC card 3rd generation: +a.)"Multimedia devices" => "Customise DVB frontends" => "Customise the frontend modules to build" +b.)"Multimedia devices" => "Customise DVB frontends" => "LG Electronics LGDT3302/LGDT3303 based" +c.)"Multimedia devices" => "Customise analog and hybrid tuner modules to build" => "Simple tuner support" + +Author: Uwe Bugla February 2009 -- cgit v0.10.2 From b391d0f08fb0858f7d2cc034ddcca6cd0545bd7d Mon Sep 17 00:00:00 2001 From: Uwe Bugla Date: Mon, 23 Feb 2009 10:56:11 -0300 Subject: V4L/DVB (10696): Remove outdated README for the flexcop-driver This patch removes an outdated README for the flexcop-driver. Signed-off-by: Uwe Bugla Signed-off-by: Patrick Boettcher Signed-off-by: Mauro Carvalho Chehab diff --git a/Documentation/dvb/README.flexcop b/Documentation/dvb/README.flexcop deleted file mode 100644 index 5515469..0000000 --- a/Documentation/dvb/README.flexcop +++ /dev/null @@ -1,205 +0,0 @@ -This README escorted the skystar2-driver rewriting procedure. It describes the -state of the new flexcop-driver set and some internals are written down here -too. - -This document hopefully describes things about the flexcop and its -device-offsprings. Goal was to write an easy-to-write and easy-to-read set of -drivers based on the skystar2.c and other information. - -Remark: flexcop-pci.c was a copy of skystar2.c, but every line has been -touched and rewritten. - -History & News -============== - 2005-04-01 - correct USB ISOC transfers (thanks to Vadim Catana) - - - - -General coding processing -========================= - -We should proceed as follows (as long as no one complains): - -0) Think before start writing code! - -1) rewriting the skystar2.c with the help of the flexcop register descriptions -and splitting up the files to a pci-bus-part and a flexcop-part. -The new driver will be called b2c2-flexcop-pci.ko/b2c2-flexcop-usb.ko for the -device-specific part and b2c2-flexcop.ko for the common flexcop-functions. - -2) Search for errors in the leftover of flexcop-pci.c (compare with pluto2.c -and other pci drivers) - -3) make some beautification (see 'Improvements when rewriting (refactoring) is -done') - -4) Testing the new driver and maybe substitute the skystar2.c with it, to reach -a wider tester audience. - -5) creating an usb-bus-part using the already written flexcop code for the pci -card. - -Idea: create a kernel-object for the flexcop and export all important -functions. This option saves kernel-memory, but maybe a lot of functions have -to be exported to kernel namespace. - - -Current situation -================= - -0) Done :) -1) Done (some minor issues left) -2) Done -3) Not ready yet, more information is necessary -4) next to be done (see the table below) -5) USB driver is working (yes, there are some minor issues) - -What seems to be ready? ------------------------ - -1) Rewriting -1a) i2c is cut off from the flexcop-pci.c and seems to work -1b) moved tuner and demod stuff from flexcop-pci.c to flexcop-tuner-fe.c -1c) moved lnb and diseqc stuff from flexcop-pci.c to flexcop-tuner-fe.c -1e) eeprom (reading MAC address) -1d) sram (no dynamic sll size detection (commented out) (using default as JJ told me)) -1f) misc. register accesses for reading parameters (e.g. resetting, revision) -1g) pid/mac filter (flexcop-hw-filter.c) -1i) dvb-stuff initialization in flexcop.c (done) -1h) dma stuff (now just using the size-irq, instead of all-together, to be done) -1j) remove flexcop initialization from flexcop-pci.c completely (done) -1l) use a well working dma IRQ method (done, see 'Known bugs and problems and TODO') -1k) cleanup flexcop-files (remove unused EXPORT_SYMBOLs, make static from -non-static where possible, moved code to proper places) - -2) Search for errors in the leftover of flexcop-pci.c (partially done) -5a) add MAC address reading -5c) feeding of ISOC data to the software demux (format of the isochronous data -and speed optimization, no real error) (thanks to Vadim Catana) - -What to do in the near future? --------------------------------------- -(no special order here) - -5) USB driver -5b) optimize isoc-transfer (submitting/killing isoc URBs when transfer is starting) - -Testing changes ---------------- - -O = item is working -P = item is partially working -X = item is not working -N = item does not apply here - = item need to be examined - - | PCI | USB -item | mt352 | nxt2002 | stv0299 | mt312 | mt352 | nxt2002 | stv0299 | mt312 --------+-------+---------+---------+-------+-------+---------+---------+------- -1a) | O | | | | N | N | N | N -1b) | O | | | | | | O | -1c) | N | N | | | N | N | O | -1d) | O | O -1e) | O | O -1f) | P -1g) | O -1h) | P | -1i) | O | N -1j) | O | N -1l) | O | N -2) | O | N -5a) | N | O -5b)* | N | -5c) | N | O - -* - not done yet - -Known bugs and problems and TODO --------------------------------- - -1g/h/l) when pid filtering is enabled on the pci card - -DMA usage currently: - The DMA is splitted in 2 equal-sized subbuffers. The Flexcop writes to first - address and triggers an IRQ when it's full and starts writing to the second - address. When the second address is full, the IRQ is triggered again, and - the flexcop writes to first address again, and so on. - The buffersize of each address is currently 640*188 bytes. - - Problem is, when using hw-pid-filtering and doing some low-bandwidth - operation (like scanning) the buffers won't be filled enough to trigger - the IRQ. That's why: - - When PID filtering is activated, the timer IRQ is used. Every 1.97 ms the IRQ - is triggered. Is the current write address of DMA1 different to the one - during the last IRQ, then the data is passed to the demuxer. - - There is an additional DMA-IRQ-method: packet count IRQ. This isn't - implemented correctly yet. - - The solution is to disable HW PID filtering, but I don't know how the DVB - API software demux behaves on slow systems with 45MBit/s TS. - -Solved bugs :) --------------- -1g) pid-filtering (somehow pid index 4 and 5 (EMM_PID and ECM_PID) aren't -working) -SOLUTION: also index 0 was affected, because net_translation is done for -these indexes by default - -5b) isochronous transfer does only work in the first attempt (for the Sky2PC -USB, Air2PC is working) SOLUTION: the flexcop was going asleep and never really -woke up again (don't know if this need fixes, see -flexcop-fe-tuner.c:flexcop_sleep) - -NEWS: when the driver is loaded and unloaded and loaded again (w/o doing -anything in the while the driver is loaded the first time), no transfers take -place anymore. - -Improvements when rewriting (refactoring) is done -================================================= - -- split sleeping of the flexcop (misc_204.ACPI3_sig = 1;) from lnb_control - (enable sleeping for other demods than dvb-s) -- add support for CableStar (stv0297 Microtune 203x/ALPS) (almost done, incompatibilities with the Nexus-CA) - -Debugging ---------- -- add verbose debugging to skystar2.c (dump the reg_dw_data) and compare it - with this flexcop, this is important, because i2c is now using the - flexcop_ibi_value union from flexcop-reg.h (do you have a better idea for - that, please tell us so). - -Everything which is identical in the following table, can be put into a common -flexcop-module. - - PCI USB -------------------------------------------------------------------------------- -Different: -Register access: accessing IO memory USB control message -I2C bus: I2C bus of the FC USB control message -Data transfer: DMA isochronous transfer -EEPROM transfer: through i2c bus not clear yet - -Identical: -Streaming: accessing registers -PID Filtering: accessing registers -Sram destinations: accessing registers -Tuner/Demod: I2C bus -DVB-stuff: can be written for common use - -Acknowledgements (just for the rewriting part) -================ - -Bjarne Steinsbo thought a lot in the first place of the pci part for this code -sharing idea. - -Andreas Oberritter for providing a recent PCI initialization template -(pluto2.c). - -Boleslaw Ciesielski for pointing out a problem with firmware loader. - -Vadim Catana for correcting the USB transfer. - -comments, critics and ideas to linux-dvb@linuxtv.org. -- cgit v0.10.2 From 9b58027bc23a73a036877f28422dad7a0a199f95 Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Thu, 26 Feb 2009 21:02:19 -0800 Subject: net: fix hp-plus build error hp-plus needs to call __alloc_eip_netdev() instead of __alloc_ei_netdev() since it is linked with 8390p.o. Fixes this build error: ERROR: "__alloc_ei_netdev" [drivers/net/hp-plus.ko] undefined! Signed-off-by: Randy Dunlap Signed-off-by: David S. Miller diff --git a/drivers/net/hp-plus.c b/drivers/net/hp-plus.c index 5e070f4..0486cbe 100644 --- a/drivers/net/hp-plus.c +++ b/drivers/net/hp-plus.c @@ -467,7 +467,7 @@ init_module(void) if (this_dev != 0) break; /* only autoprobe 1st one */ printk(KERN_NOTICE "hp-plus.c: Presently autoprobing (not recommended) for a single card.\n"); } - dev = alloc_ei_netdev(); + dev = alloc_eip_netdev(); if (!dev) break; dev->irq = irq[this_dev]; -- cgit v0.10.2 From f8af11af85fecbfa7b95fd79c043b16ae0ee0d55 Mon Sep 17 00:00:00 2001 From: Michael Buesch Date: Thu, 26 Feb 2009 22:33:00 -0800 Subject: b44: Unconditionally enable interrupt routing on reset Unconditionally setup the IRQ routing on chip reset. It's safe to call ssb_pcicore_dev_irqvecs_enable() unconditionally, because it has internal checks for redundant calls. This fixes problems where hardware will not come up properly due to quirks in the enable-bit hardware. Reported-by: Pantelis Koukousoulas Signed-off-by: Michael Buesch Signed-off-by: David S. Miller diff --git a/drivers/net/b44.c b/drivers/net/b44.c index c38512e..6b8c39f 100644 --- a/drivers/net/b44.c +++ b/drivers/net/b44.c @@ -1264,8 +1264,14 @@ static void b44_clear_stats(struct b44 *bp) static void b44_chip_reset(struct b44 *bp, int reset_kind) { struct ssb_device *sdev = bp->sdev; + bool was_enabled; - if (ssb_device_is_enabled(bp->sdev)) { + was_enabled = ssb_device_is_enabled(bp->sdev); + + ssb_device_enable(bp->sdev, 0); + ssb_pcicore_dev_irqvecs_enable(&sdev->bus->pcicore, sdev); + + if (was_enabled) { bw32(bp, B44_RCV_LAZY, 0); bw32(bp, B44_ENET_CTRL, ENET_CTRL_DISABLE); b44_wait_bit(bp, B44_ENET_CTRL, ENET_CTRL_DISABLE, 200, 1); @@ -1277,10 +1283,8 @@ static void b44_chip_reset(struct b44 *bp, int reset_kind) } bw32(bp, B44_DMARX_CTRL, 0); bp->rx_prod = bp->rx_cons = 0; - } else - ssb_pcicore_dev_irqvecs_enable(&sdev->bus->pcicore, sdev); + } - ssb_device_enable(bp->sdev, 0); b44_clear_stats(bp); /* -- cgit v0.10.2 From e92aa634a33739478958f4109d6bd35b36d13532 Mon Sep 17 00:00:00 2001 From: Michael Buesch Date: Thu, 26 Feb 2009 22:35:02 -0800 Subject: b44: Disable device on shutdown Disable the SSB core on device shutdown. This has two advantages: 1) A clean device shutdown is always desired here, because we disable the device's global crystal in the next statement. 2) This fixes a bug where the device will come up with the enable-bit set on the next initialization (without a reboot inbetween). This causes breakage on the second initialization due to code that checks this bit (ssb_device_is_enabled() checks). Reported-by: Pantelis Koukousoulas Signed-off-by: Michael Buesch Signed-off-by: David S. Miller diff --git a/drivers/net/b44.c b/drivers/net/b44.c index 6b8c39f..dc5f051 100644 --- a/drivers/net/b44.c +++ b/drivers/net/b44.c @@ -2240,6 +2240,7 @@ static void __devexit b44_remove_one(struct ssb_device *sdev) struct net_device *dev = ssb_get_drvdata(sdev); unregister_netdev(dev); + ssb_device_disable(sdev, 0); ssb_bus_may_powerdown(sdev->bus); free_netdev(dev); ssb_pcihost_set_power_state(sdev, PCI_D3hot); -- cgit v0.10.2 From 08c2f5b4d76f83213e379b12df504269d21c9e7c Mon Sep 17 00:00:00 2001 From: Paul Mundt Date: Fri, 27 Feb 2009 15:41:14 +0900 Subject: sh: ap325rxa: Revert ov772x support. This change depends on some v4l changes that have been pushed back to 2.6.30, so drop this and fall back on the old soc_camera code until then. Reported-by: Nobuhiro Iwamatsu Acked-by: Kuninori Morimoto Signed-off-by: Paul Mundt diff --git a/arch/sh/boards/board-ap325rxa.c b/arch/sh/boards/board-ap325rxa.c index 7c35787..72da416 100644 --- a/arch/sh/boards/board-ap325rxa.c +++ b/arch/sh/boards/board-ap325rxa.c @@ -22,7 +22,6 @@ #include #include #include -#include #include #include #include