summaryrefslogtreecommitdiff
AgeCommit message (Collapse)Author
2016-09-23ext4: Use correct descriptor size when reading the block group descriptorStefan Brüns
The correct descriptor size must be used when calculating offsets, and also to read the correct amount of data. Signed-off-by: Stefan Brüns <stefan.bruens@rwth-aachen.de>
2016-09-23ext4: Add helper functions for block group descriptor field accessStefan Brüns
The helper functions encapsulate access of the block group descriptors, independent of group descriptor size. The helpers also deal with the endianess of the fields, and with split fields like free_blocks/ free_blocks_high. Signed-off-by: Stefan Brüns <stefan.bruens@rwth-aachen.de>
2016-09-23ext4: determine group descriptor size for 64bit featureStefan Brüns
If EXT4_FEATURE_INCOMPAT_64BIT is set, the descriptor can be read from the superblocks, otherwise it defaults to 32. Signed-off-by: Stefan Brüns <stefan.bruens@rwth-aachen.de>
2016-09-23ext4: Update ext2/3/4 superblock, group descriptor and inode structuresStefan Brüns
Most importantly, the superblock provides the used group descriptor size, which is required for the EXT4_FEATURE_INCOMPAT_64BIT. Signed-off-by: Stefan Brüns <stefan.bruens@rwth-aachen.de>
2016-09-23ext4: Fix memory leak of journal buffer if block is updated multiple timesStefan Brüns
If the same block is updated multiple times in a row during a single file system operation, gd_index is decremented to use the same journal entry again. Avoid loosing the already allocated buffer. Signed-off-by: Stefan Brüns <stefan.bruens@rwth-aachen.de>
2016-09-23ext4: Correct block number handling, empty block vs. error codeStefan Brüns
read_allocated block may return block number 0, which is just an indicator a chunk of the file is not backed by a block, i.e. it is sparse. During file deletions, just continue with the next logical block, for other operations treat blocknumber <= 0 as an error. For writes, blocknumber 0 should never happen, as U-Boot always allocates blocks for the whole file. Reading already handles this correctly, i.e. the read buffer is 0-fillled. Not treating block 0 as sparse block leads to FS corruption, e.g. ./sandbox/u-boot -c 'host bind 0 ./sandbox/test/fs/3GB.ext4.img ; ext4write host 0 0 /2.5GB.file 1 ' The 2.5GB.file from the fs test is actually a sparse file. Signed-off-by: Stefan Brüns <stefan.bruens@rwth-aachen.de>
2016-09-23ext4: remove duplicated block release code for extentsStefan Brüns
The data blocks are identical for files using traditional direct/indirect block allocation scheme and extent trees, thus this code part can be common. Only the code to deallocate the indirect blocks to record the used blocks has to be seperate, respectively the code to release extent tree index blocks. Actually the code to release the extent tree index blocks is still missing, but at least add a FIXME at the appropriate place. Signed-off-by: Stefan Brüns <stefan.bruens@rwth-aachen.de>
2016-09-23ext4: initialize full inode for inodes bigger than 128 bytesStefan Brüns
Make sure the the extra_isize field (offset 128) is initialized to 0, to mark any extra data as invalid. Signed-off-by: Stefan Brüns <stefan.bruens@rwth-aachen.de> Reviewed-by: Lukasz Majewski <l.majewski@samsung.com>
2016-09-23ext4: Use correct value for inode size even on revision 0 filesystemsStefan Brüns
fs->inodesz is already correctly (i.e. dependent on fs revision) initialized in ext4fs_mount. Signed-off-by: Stefan Brüns <stefan.bruens@rwth-aachen.de> Reviewed-by: Lukasz Majewski <l.majewski@samsung.com>
2016-09-23ext4: Fix memory leak in case of failureStefan Brüns
temp_ptr should always be freed, even if the function is left via goto fail. Signed-off-by: Stefan Brüns <stefan.bruens@rwth-aachen.de> Reviewed-by: Lukasz Majewski <l.majewski@samsung.com>
2016-09-23ext4: Avoid out-of-bounds access of block bitmapStefan Brüns
If the blocksize is 1024, count is initialized with 1. Incrementing count by 8 will never match (count == fs->blksz * 8), and ptr may be incremented beyond the buffer end if the bitmap is filled. Add the startblock offset after the loop. Remove the second loop, as only the first iteration will be done. Signed-off-by: Stefan Brüns <stefan.bruens@rwth-aachen.de> Reviewed-by: Lukasz Majewski <l.majewski@samsung.com>
2016-09-23ext4: After completely filled group, scan next group from the beginningStefan Brüns
The last free block of a block group may be in its middle. After it has been allocated, the next block group should be scanned from its beginning. The following command triggers the bad behaviour (on a blocksize 1024 fs): ./sandbox/u-boot -c 'i=0; host bind 0 ./disk.raw ; while test $i -lt 260 ; do echo $i; setexpr i $i + 1; ext4write host 0:2 0 /X${i} 0x1450; done ; ext4write host 0:2 0 /X240 0x2000 ; ' When 'X240' is extended from 5200 byte to 8192 byte, the new blocks should start from the first free block (8811), but it uses the blocks 8098-8103 and 16296-16297 -- 8103 + 1 + 8192 = 16296. This can be shown with debugfs, commands 'ffb' and 'stat X240'. Signed-off-by: Stefan Brüns <stefan.bruens@rwth-aachen.de> Reviewed-by: Lukasz Majewski <l.majewski@samsung.com>
2016-09-23ext4: Do not clear zalloc'ed buffers a second timeStefan Brüns
zero_buffer is never written, thus clearing it is pointless. journal_buffer is completely initialized by ext4fs_devread (or in case of failure, not used). Signed-off-by: Stefan Brüns <stefan.bruens@rwth-aachen.de> Reviewed-by: Lukasz Majewski <l.majewski@samsung.com>
2016-09-23ext4: Only update number of of unused inodes if GDT_CSUM feature is setStefan Brüns
e2fsck warns about "Group descriptor 0 marked uninitialized without feature set." The bg_itable_unused field is only defined if FEATURE_RO_COMPAT_GDT_CSUM is set, and should be set (kept) zero otherwise. Signed-off-by: Stefan Brüns <stefan.bruens@rwth-aachen.de> Reviewed-by: Lukasz Majewski <l.majewski@samsung.com>
2016-09-23ext4: Scan all directory blocks when looking up an entryStefan Brüns
Scanning only the direct blocks of the directory file may falsely report an existing file as nonexisting, and worse can also lead to creation of a duplicate entry on file creation. Signed-off-by: Stefan Brüns <stefan.bruens@rwth-aachen.de> Reviewed-by: Lukasz Majewski <l.majewski@samsung.com>
2016-09-23ext4: Avoid corruption of directories with hash tree indexesStefan Brüns
While directories can be read using the old linear scan method, adding a new file would require updating the index tree (alternatively, the whole tree could be removed). Signed-off-by: Stefan Brüns <stefan.bruens@rwth-aachen.de> Reviewed-by: Lukasz Majewski <l.majewski@samsung.com>
2016-09-23ext4: Scan all directory blocks for space when inserting a new entryStefan Brüns
Previously, only the last directory block was scanned for available space. Instead, scan all blocks back to front, and if no sufficient space is found, eventually append a new block. Blocks are only appended if the directory does not use extents or the new block would require insertion of indirect blocks, as the old code does. Signed-off-by: Stefan Brüns <stefan.bruens@rwth-aachen.de> Reviewed-by: Lukasz Majewski <l.majewski@samsung.com>
2016-09-23ext4: Do not crash when trying to grow a directory using extentsStefan Brüns
The following command crashes u-boot: ./sandbox/u-boot -c 'i=0; host bind 0 ./sandbox/test/fs/3GB.ext4.img ; while test $i -lt 200 ; do echo $i; setexpr i $i + 1; ext4write host 0 0 /foobar${i} 0; done' Previously, the code updated the direct_block even for extents, and fortunately crashed before pushing garbage to the disk. Signed-off-by: Stefan Brüns <stefan.bruens@rwth-aachen.de> Reviewed-by: Lukasz Majewski <l.majewski@samsung.com>
2016-09-23ext4: propagate error if creation of directory entry failsStefan Brüns
In case the dir entry creation failed, ext4fs_write would later overwrite a random inode, as inodeno was never initialized. Signed-off-by: Stefan Brüns <stefan.bruens@rwth-aachen.de> Reviewed-by: Lukasz Majewski <l.majewski@samsung.com>
2016-09-23ext4: fix possible crash on directory traversal, ignore deleted entriesStefan Brüns
The following command triggers a segfault in search_dir: ./sandbox/u-boot -c 'host bind 0 ./sandbox/test/fs/3GB.ext4.img ; ext4write host 0 0 /./foo 0x10' The following command triggers a segfault in check_filename: ./sandbox/u-boot -c 'host bind 0 ./sandbox/test/fs/3GB.ext4.img ; ext4write host 0 0 /. 0x10' "." is the first entry in the directory, thus previous_dir is NULL. The whole previous_dir block in search_dir seems to be a bad copy from check_filename(...). As the changed data is not written to disk, the statement is mostly harmless, save the possible NULL-ptr reference. Typically a file is unlinked by extending the direntlen of the previous entry. If the entry is the first entry in the directory block, it is invalidated by setting inode=0. The inode==0 case is hard to trigger without crafted filesystems. It only hits if the first entry in a directory block is deleted and later a lookup for the entry (by name) is done. Signed-off-by: Stefan Brüns <stefan.bruens@rwth-aachen.de> Reviewed-by: Lukasz Majewski <l.majewski@samsung.com>
2016-09-23ext4: fix wrong usage of le32_to_cpu()Michael Walle
le32_to_cpu() must only convert the revision_level and not the boolean result. Signed-off-by: Michael Walle <michael@walle.cc>
2016-09-23ext4: fix endianess problems in ext4 write supportMichael Walle
All fields were accessed directly instead of using the proper byte swap functions. Thus, ext4 write support was only usable on little-endian architectures. Fix this. Signed-off-by: Michael Walle <michael@walle.cc>
2016-09-23ext4: use kernel names for byte swapsMichael Walle
Instead of __{be,le}{16,32}_to_cpu use {be,le}{16,32}_to_cpu. Signed-off-by: Michael Walle <michael@walle.cc>
2016-09-23ext4: change structure fields to __le/__be typesMichael Walle
Change all the types of ext2/4 fields to little endian types and all the JBD fields to big endian types. Now we can use sparse (make C=1) to check for statements where we need byteswaps. Signed-off-by: Michael Walle <michael@walle.cc>
2016-09-23test/fs: Check writes using "." (same dir) relative pathStefan Brüns
<path>/<fname> and <path>/./<fname> should reference the same file. Signed-off-by: Stefan Brüns <stefan.bruens@rwth-aachen.de>
2016-09-23test/fs: Check ext4 behaviour if dirent is first entry in directory blockStefan Brüns
This is a regression test for a crash happening if the first dirent in the block matches. Code tried to access a predecessor entry which does not exist. The crash happened for any block, but "." is always the first entry in the first directory block and thus easy to check for. Signed-off-by: Stefan Brüns <stefan.bruens@rwth-aachen.de>
2016-09-23test/fs: strip noise from filesystem code prior to checking resultsStefan Brüns
ext4 and fat code emit some diagnostic messages during command execution. These additional lines force a match window size which strictly is not necessary. Signed-off-by: Stefan Brüns <stefan.bruens@rwth-aachen.de>
2016-09-23test/fs: remove use of undefined WRITE_FILE variableStefan Brüns
The write file is created from $SMALL_FILE by appending ".w" on all other occurences in the code. Signed-off-by: Stefan Brüns <stefan.bruens@rwth-aachen.de>
2016-09-23test/fs: Restructure file path specification to allow some flexibilityStefan Brüns
Instead of providing the full path, specify directory and filename separately. This allows to specify intermediate directories, required for some additional tests. Signed-off-by: Stefan Brüns <stefan.bruens@rwth-aachen.de>
2016-09-23cmd/fat: Do not crash on write when <bytes> is not specifiedStefan Brüns
argc is checked, but is off by one. In case <bytes> is not specified, create an empty file, which is identical to the ext4write behaviour. Signed-off-by: Stefan Brüns <stefan.bruens@rwth-aachen.de> Reviewed-by: Simon Glass <sjg@chromium.org>
2016-09-23fs/fat: Correct description of determine_fatent functionStefan Brüns
Current description does not match the function behaviour. Signed-off-by: Stefan Brüns <stefan.bruens@rwth-aachen.de> Acked-by: Lukasz Majewski <l.majewski@samsung.com>
2016-09-23fs/fat: Do not write unmodified fat entries to diskStefan Brüns
The code caches 6 sectors of the FAT. On FAT traversal, the old contents needs to be flushed to disk, but only if any FAT entries had been modified. Explicitly flag the buffer on modification. Currently, creating a new file traverses the whole FAT up to the first free cluster and rewrites the on-disk blocks. Signed-off-by: Stefan Brüns <stefan.bruens@rwth-aachen.de> Reviewed-by: Lukasz Majewski <l.majewski@samsung.com>
2016-09-23fs/fat: Remove two statements without effectStefan Brüns
fatlength is a local variable which is no more used after the assignment. s_name is not used in the function, save the strncpy. Signed-off-by: Stefan Brüns <stefan.bruens@rwth-aachen.de> Acked-by: Lukasz Majewski <l.majewski@samsung.com> Acked-by: Stephen Warren <swarren@nvidia.com> Reviewed-by: Benoît Thébaudeau <benoit.thebaudeau.dev@gmail.com>
2016-09-22Merge git://git.denx.de/u-boot-rockchipTom Rini
2016-09-22ns16650: Make sure we have CONFIG_CLK set before using infrastructureTom Rini
We cannot call on the CONFIG_CLK based clk_get_rate function unless CONFIG_CLK is set. Signed-off-by: Tom Rini <trini@konsulko.com>
2016-09-22Merge branch 'master' of git://git.denx.de/u-boot-uniphierTom Rini
2016-09-22ARM: dts: uniphier: sync clock/reset controller nodes with LinuxMasahiro Yamada
Sync device trees with Linux for easier DT life. Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
2016-09-22clk: uniphier: allow to have clock node under syscon nodeMasahiro Yamada
To sync the DT binding with Linux, the register base must be taken from the parent syscon node. Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
2016-09-22clk: uniphier: move U_BOOT_DRIVER entry to core codeMasahiro Yamada
Move U_BOOT_DRIVER() entry from the data file (clk-uniphier-mio.c) to the core support file (clk-uniphier-core.c) because I do not want to repeat the driver boilerplate when I add more clock data. Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
2016-09-22clk: uniphier: constify clock data arrays/structuresMasahiro Yamada
Clarify these clock data are constant. Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
2016-09-22ARM: uniphier: add PLL init code for LD11 SoCMasahiro Yamada
- Initialize PLLs (SPL initializes only DPLL to save the precious SPL memory footprint) - Adjust CPLL/MPLL to the final tape-out frequency - Set the Cortex-A53 clock to the maximum frequency since it is running at 500MHz (SPLL/4) on startup Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
2016-09-22ARM: uniphier: move CONFIG_SPL_* to defconfig or selectMasahiro Yamada
As I repeated in the ML, I am unhappy with config entries with bare defaults. Kick them out of arch/arm/mach-uniphier/Kconfig. Currently, CONFIG_SPL_SERIAL_SUPPORT is not user-configurable (build fails without it), but it should be fixed later anyway, so I am moving CONFIG_SPL_SERIAL_SUPPORT to defconfigs. Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com> Reviewed-by: Tom Rini <trini@konsulko.com>
2016-09-22Merge branch 'master' of git://git.denx.de/u-boot-spiTom Rini
2016-09-22Merge branch 'master' of git://www.denx.de/git/u-boot-microblazeTom Rini
2016-09-22Makefile: rm u-boot.cfg dependencies are missingStephen Warren
Prior to the previous patch, a freshly created .u-boot.cfg.cmd may not correctly represent all dependencies for u-boot.cfg. The previous change only solved this issue for fresh builds; when performing an incremental build, the deficient .u-boot.cfg.cmd is already present, so u-boot.cfg is not rebuilt, and hence .u-boot.cfg.cmd is not rebuilt with the correct content. Solve this by explicitly detecting when the dependency file .u-boot.cfg.d has not been integrated into .u-boot.cfg.cmd, and force u-boot.cfg to be rebuilt in this case by deleting it first. This is possible since if_changed_dep will always delete .u-boot.cfg.d when it executes successfully, so its presence means either that the previous build was made by a source tree that contained a Makefile that didn't include the previous patch, or that the build failed part way through executing if_changed_dep for u-boot.cfg. Forcing a rebuild of u-boot.cfg is required in the former case, and will cause no additional work in the latter case, since the file would be rebuilt anyway for the same reason it was being rebuilt by the previous build. Signed-off-by: Stephen Warren <swarren@nvidia.com>
2016-09-22Makefile: use if_change_dep for u-boot.cfgStephen Warren
cmd_cpp_cfg generates a dependency output, but because it's invoked using if_changed rather than if_changed_dep, that dependency file is ignored. This results in Kbuild not knowing about which files u-boot.cfg depends on, so it may not be rebuilt when required. A practical result of this is that u-boot.cfg may continue to reference CONFIG_ options that no longer exist in the source tree, and this can cause the adhoc config options check to fail. This change modifies Makefile to use if_changed_dep, which in turn causes all dependencies to be known to the next make invocation. Signed-off-by: Stephen Warren <swarren@nvidia.com>
2016-09-22test/py/tests/test_vboot.py: Add check that we boot the imageTom Rini
Make sure that when we're telling bootm to boot an image, and we expect the image to boot we get the output from sandbox that we attempted to run Linux and that U-Boot completed its job. Cc: Simon Glass <sjg@chromium.org> Cc: Stephen Warren <swarren@nvidia.com> Signed-off-by: Tom Rini <trini@konsulko.com> Reviewed-by: Simon Glass <sjg@chromium.org> Acked-by: Stephen Warren <swarren@nvidia.com>
2016-09-22image-fit: Fix fit_get_node_from_config semanticsPaul Burton
Commit bac17b78dace ("image-fit: switch ENOLINK to ENOENT") changed fit_get_node_from_config to return -ENOENT when a property doesn't exist, but didn't change any of its callers which check return values. Notably it didn't change boot_get_ramdisk, which leads to U-Boot failing to boot FIT images which don't include ramdisks with the following message: Ramdisk image is corrupt or invalid It also didn't take into account that by returning -ENOENT to denote the lack of a property we lost the ability to determine from the return value of fit_get_node_from_config whether it was the property or the configuration node that was missing, which may potentially lead callers to accept invalid FIT images. Fix this by having fit_get_node_from_config return -EINVAL when the configuration node isn't found and -ENOENT when the property isn't found, which seems to make semantic sense. Callers that previously checked for -ENOLINK are adjusted to check for -ENOENT, which fixes the breakage introduced by commit bac17b78dace ("image-fit: switch ENOLINK to ENOENT"). The only other user of the return fit_get_node_from_config return value, indirectly, is bootm_find_os which already checked for -ENOENT. From a read-through of the code I suspect it ought to have been checking for -ENOLINK prior to bac17b78dace ("image-fit: switch ENOLINK to ENOENT") anyway, which would make it right after this patch, but this would be good to get verified by someone who knows this x86 code or is able to test it. Signed-off-by: Paul Burton <paul.burton@imgtec.com> Cc: Jonathan Gray <jsg@jsg.id.au> Cc: Marek Vasut <marex@denx.de> Acked-by: Marek Vasut <marex@denx.de> Acked-by: Stefan Roese <sr@denx.de> Acked-by: George McCollister <george.mccollister@gmail.com> Tested-by: George McCollister <george.mccollister@gmail.com>
2016-09-22clk: rk3288: add PWM clock get rateKever Yang
This patch add clk_get_rate for PWM device. Signed-off-by: Kever Yang <kever.yang@rock-chips.com> Acked-by: Simon Glass <sjg@chromium.org>
2016-09-22clk: rk3399: add pmucru controller supportKever Yang
pmucru is a module like cru which is a clock controller manage some PLL and module clocks. Signed-off-by: Kever Yang <kever.yang@rock-chips.com> Acked-by: Simon Glass <sjg@chromium.org>