| Age | Commit message (Collapse) | Author |
|---|
|
With this error output becomes almost readable. The line numbers are
still totally bonghits, but that's a lot harder to pull out of
kerneldoc. We'd essentially have to insert some special markers in the
kernel-doc output, split the output along these markers and then
insert each block separately using
state_machine.insert_input(block, source, first_line)
Cc: Jani Nikula <jani.nikula@intel.com>
Cc: linux-doc@vger.kernel.org
Cc: Jonathan Corbet <corbet@lwn.net>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
|
|
Reconcile differences between python2 and python3 on dealing with
stdout, stderr from Popen. This fixes "name 'unicode' is not defined"
errors on python3. We'll need to try to keep the extension working on
both python-sphinx and python3-sphinx so we don't need two copies.
Reported-and-tested-by: Marius Vlad <marius.c.vlad@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
|
|
If the documentation comment does not have params or sections, the
section heading may leak from the previous documentation comment.
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
|
|
If there are multiple sections with the same section name, the current
implementation results in several sections by the same heading, with the
content duplicated from the last section to all. Even if there's the
error message, a more graceful approach is to combine all the
identically named sections into one, with concatenated contents.
With the supported sections already limited to select few, there are
massively fewer collisions than there used to be, but this is still
useful for e.g. when function parameters are documented in the middle of
a documentation comment, with description spread out above and
below. (This is not a recommended documentation style, but used in the
kernel nonetheless.)
We can now also demote the error to a warning.
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
|
|
kernel-doc currently identifies anything matching "section header:"
(specifically a string of word characters and spaces followed by a
colon) as a new section in the documentation comment, and renders the
section header accordingly.
Unfortunately, this turns all uses of colon into sections, mostly
unintentionally. Considering the output, erroneously creating sections
when not intended is always worse than erroneously not creating sections
when intended. For example, a line with "http://example.com" turns into
a "http" heading followed by "//example.com" in normal text style, which
is quite ugly. OTOH, "WARNING: Beware of the Leopard" is just fine even
if "WARNING" does not turn into a heading.
It is virtually impossible to change all the kernel-doc comments, either
way. The compromise is to pick the most commonly used and depended on
section headers (with variants) and accept them as section headers.
The accepted section headers are, case insensitive:
* description:
* context:
* return:
* returns:
Additionally, case sensitive:
* @return:
All of the above are commonly used in the kernel-doc comments, and will
result in worse output if not identified as section headers. Also,
kernel-doc already has some special handling for all of them, so there's
nothing particularly controversial in adding more special treatment for
them.
While at it, improve the whitespace handling surrounding section
names. Do not consider the whitespace as part of the name.
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
|
|
Yes, for our purposes the type should contain typedef.
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
|
|
The latter isn't special to rst.
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
|
|
If a param description spans multiple lines, check any leading
whitespace in the first continuation line, and remove same amount of
whitespace from following lines.
This allows indentation in the multi-line parameter descriptions for
aesthetical reasons while not causing accidentally significant
indentation in the rst output.
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
|
|
Handle whitespace on the first line of param text as if it was the empty
string. There is no need to add the newline in this case. This improves
the rst output in particular, where blank lines may be problematic in
parameter lists.
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
|
|
Move away from field lists, and simply use **strong emphasis** for
section headings on lines of their own. Do not use rst section headings,
because their nesting depth depends on the surrounding context, which
kernel-doc has no knowledge of. Also, they do not need to end up in any
table of contexts or indexes.
There are two related immediate benefits. Field lists are typically
rendered in two columns, while the new style uses the horizontal width
better. With no extra indent on the left, there's no need to be as fussy
about it. Field lists are more susceptible to indentation problems than
the new style.
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
|
|
The inline member markup allows whitespace lines before the actual
documentation starts. Strip the leading blank lines. This improves the
rst output.
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
|
|
Current approach leads to two blank lines, while one is enough.
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
|
|
No functional changes.
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
|
|
The use of these is confusing in the script, and per this grep, they're
not used anywhere anyway:
$ git grep " \* [%$&][a-zA-Z0-9_]*:" -- *.[ch] | grep -v "\$\(Id\|Revision\|Date\)"
While at it, throw out the constants array, nothing is ever put there
again.
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
|
|
Let the user use @foo, &bar, %baz, etc. in the first kernel-doc purpose
line too.
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
|
|
This bit is already done by xml_unescape() above.
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
|
|
Link "&foo->bar", "&foo->bar()", "&foo.bar", and "&foo.bar()" to the
struct/union/enum foo definition. The members themselves do not
currently have anchors to link to, but this is better than nothing, and
promotes a universal notation.
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
|
|
Let the user use "&union foo" and "&typedef foo" to reference foo. The
difference to using "union &foo", "typedef &foo", or just "&foo" (which
are valid too) is that "union" and "typedef" become part of the link
text.
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
|
|
It's possible to use &foo to reference structs, enums, typedefs, etc. in
the Sphinx C domain. Thus do not prefix the links with "struct".
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
|
|
The Sphinx C domain spec says function references should include the
parens ().
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
|
|
If the user requests a specific DOC: section by name, do not output its
section title. In these cases, the surrounding context already has a
heading, and the DOC: section title is only used as an identifier and a
heading for clarity in the source file.
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
|
|
Make the output selection a bit more readable by adding constants for
the various types of output selection. While at it, actually call the
variable for choosing what to output $output_selection.
No functional changes.
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
|
|
Make the state machine a bit more readable by adding constants for
parser states and inline member documentation parser substates. While at
it, rename the "split" documentation to "inline" documentation.
No functional changes.
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
|
|
Add "struct" in the label of the reference.
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
|
|
Function references should include the parens (), struct references
should not include "struct".
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
|
|
This script uses pandoc to convert existing DocBook template files to RST
templates. A couple of sed scripts are need to massage things both before
and after the conversion, but the result is then usable with no hand
editing.
[Jani: Change usage to tmplcvt <in> <out>. Fix escaping for docproc
directives. Add support the new kernel-doc extension.]
Signed-off-by: Jonathan Corbet <corbet@lwn.net>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
|
|
Read the version and release from the top level Makefile (for use when
Sphinx is invoked directly, by e.g. Read the Docs), but override them
via Sphinx command line arguments in a normal documentation build.
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
|
|
Tell Sphinx where to find the extension, and pass on the kernel src tree
and kernel-doc paths to the extension.
With this, any .rst files under Documentation may contain the kernel-doc
rst directive to include kernel-doc documentation from any source file.
While building, it may be handy to pass kernel-doc extension
configuration on the command line. For example, 'make SPHINXOPTS="-D
kerneldoc_verbosity=0" htmldocs' silences all stderr output from
kernel-doc when the kernel-doc exit code is 0. (The stderr will be
logged unconditionally when the exit code is non-zero.)
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
|
|
Add an extension to handle kernel-doc directives, to call kernel-doc
according to the arguments and parameters given to the reStructuredText
directive.
The syntax for the kernel-doc directive is:
.. kernel-doc:: FILENAME
:export:
:internal:
:functions: FUNCTION [FUNCTION ...]
:doc: SECTION TITLE
Of the directive options export, internal, functions, and doc, currently
only one option may be given at a time.
The FILENAME is relative from the kernel source tree root.
The extension notifies Sphinx about the document dependency on FILENAME,
causing the document to be rebuilt when the file has been changed.
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
|
|
The Sphinx output directory is generated.
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
|
|
Add basic configuration and makefile to build documentation from any
.rst files under Documentation using Sphinx. For starters, there's just
the placeholder index.rst.
At the top level Makefile, hook Sphinx documentation targets alongside
(but independent of) the DocBook toolchain, having both be run on the
various 'make *docs' targets.
All Sphinx processing is placed into Documentation/Makefile.sphinx. Both
that and the Documentation/DocBook/Makefile are now expected to handle
all the documentation targets, explicitly ignoring them if they're not
relevant for that particular toolchain. The changes to the existing
DocBook Makefile are kept minimal.
There is graceful handling of missing Sphinx and rst2pdf (which is
needed for pdf output) by checking for the tool and python module,
respectively, with informative messages to the user.
If the Read the Docs theme (sphinx_rtd_theme) is available, use it, but
otherwise gracefully fall back to the Sphinx default theme, with an
informative message to the user, and slightly less pretty HTML output.
Sphinx can now handle htmldocs, pdfdocs (if rst2pdf is available),
epubdocs and xmldocs targets. The output documents are written into per
output type subdirectories under Documentation/output.
Finally, you can pass options to sphinx-build using the SPHINXBUILD make
variable. For example, 'make SPHINXOPTS=-v htmldocs' for more verbose
output from Sphinx.
This is based on the original work by Jonathan Corbet, but he probably
wouldn't recognize this as his own anymore.
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
|
|
Currently we use docproc to figure out which symbols are exported, and
then docproc calls kernel-doc on specific functions, to get
documentation on exported functions. According to git blame and docproc
comments, this is due to historical reasons, as functions and their
corresponding EXPORT_SYMBOL* may have been in different files. However
for more than ten years the recommendation in CodingStyle has been to
place the EXPORT_SYMBOL* immediately after the closing function brace
line.
Additionally, the kernel-doc comments for functions are generally placed
above the function definition in the .c files (i.e. where the
EXPORT_SYMBOL* is) rather than above the declaration in the .h
files. There are some exceptions to this, but AFAICT none of these are
included in DocBook documentation using the "!E" docproc directive.
Therefore, assuming the EXPORT_SYMBOL* and kernel-doc are with the
function definition, kernel-doc can extract the exported vs. not
information by making two passes on the input file. Add support for that
via the new -export and -internal parameters.
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
|
|
I'm not quite sure why the errors below are happening, but this fixes
them.
Use of uninitialized value in string ne at ./scripts/kernel-doc line 1819, <IN> line 6494.
Use of uninitialized value $_[0] in join or string at ./scripts/kernel-doc line 1759, <IN> line 6494.
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
|
|
|
|
The self-test was updated to cover zero-length strings; the function
needs to be updated, too.
Reported-by: Geert Uytterhoeven <geert@linux-m68k.org>
Signed-off-by: George Spelvin <linux@sciencehorizons.net>
Fixes: fcfd2fbf22d2 ("fs/namei.c: Add hashlen_string() function")
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
|
|
The original name was simply hash_string(), but that conflicted with a
function with that name in drivers/base/power/trace.c, and I decided
that calling it "hashlen_" was better anyway.
But you have to do it in two places.
[ This caused build errors for architectures that don't define
CONFIG_DCACHE_WORD_ACCESS - Linus ]
Signed-off-by: George Spelvin <linux@sciencehorizons.net>
Reported-by: Guenter Roeck <linux@roeck-us.net>
Fixes: fcfd2fbf22d2 ("fs/namei.c: Add hashlen_string() function")
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
|
|
The HPFS filesystem used generic_show_options to produce string that is
displayed in /proc/mounts. However, there is a problem that the options
may disappear after remount. If we mount the filesystem with option1
and then remount it with option2, /proc/mounts should show both option1
and option2, however it only shows option2 because the whole option
string is replaced with replace_mount_options in hpfs_remount_fs.
To fix this bug, implement the hpfs_show_options function that prints
options that are currently selected.
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Cc: stable@vger.kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
|
|
Commit c8f33d0bec99 ("affs: kstrdup() memory handling") checks if the
kstrdup function returns NULL due to out-of-memory condition.
However, if we are remounting a filesystem with no change to
filesystem-specific options, the parameter data is NULL. In this case,
kstrdup returns NULL (because it was passed NULL parameter), although no
out of memory condition exists. The mount syscall then fails with
ENOMEM.
This patch fixes the bug. We fail with ENOMEM only if data is non-NULL.
The patch also changes the call to replace_mount_options - if we didn't
pass any filesystem-specific options, we don't call
replace_mount_options (thus we don't erase existing reported options).
Fixes: c8f33d0bec99 ("affs: kstrdup() memory handling")
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Cc: stable@vger.kernel.org # v4.1+
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
|
|
Commit ce657611baf9 ("hpfs: kstrdup() out of memory handling") checks if
the kstrdup function returns NULL due to out-of-memory condition.
However, if we are remounting a filesystem with no change to
filesystem-specific options, the parameter data is NULL. In this case,
kstrdup returns NULL (because it was passed NULL parameter), although no
out of memory condition exists. The mount syscall then fails with
ENOMEM.
This patch fixes the bug. We fail with ENOMEM only if data is non-NULL.
The patch also changes the call to replace_mount_options - if we didn't
pass any filesystem-specific options, we don't call
replace_mount_options (thus we don't erase existing reported options).
Fixes: ce657611baf9 ("hpfs: kstrdup() out of memory handling")
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Cc: stable@vger.kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
|
|
Pull more MIPS updates from Ralf Baechle:
"This is the secondnd batch of MIPS patches for 4.7. Summary:
CPS:
- Copy EVA configuration when starting secondary VPs.
EIC:
- Clear Status IPL.
Lasat:
- Fix a few off by one bugs.
lib:
- Mark intrinsics notrace. Not only are the intrinsics
uninteresting, it would cause infinite recursion.
MAINTAINERS:
- Add file patterns for MIPS BRCM device tree bindings.
- Add file patterns for mips device tree bindings.
MT7628:
- Fix MT7628 pinmux typos.
- wled_an pinmux gpio.
- EPHY LEDs pinmux support.
Pistachio:
- Enable KASLR
VDSO:
- Build microMIPS VDSO for microMIPS kernels.
- Fix aliasing warning by building with `-fno-strict-aliasing' for
debugging but also tracing them might result in recursion.
Misc:
- Add missing FROZEN hotplug notifier transitions.
- Fix clk binding example for varioius PIC32 devices.
- Fix cpu interrupt controller node-names in the DT files.
- Fix XPA CPU feature separation.
- Fix write_gc0_* macros when writing zero.
- Add inline asm encoding helpers.
- Add missing VZ accessor microMIPS encodings.
- Fix little endian microMIPS MSA encodings.
- Add 64-bit HTW fields and fix its configuration.
- Fix sigreturn via VDSO on microMIPS kernel.
- Lots of typo fixes.
- Add definitions of SegCtl registers and use them"
* 'upstream' of git://git.linux-mips.org/pub/scm/ralf/upstream-linus: (49 commits)
MIPS: Add missing FROZEN hotplug notifier transitions
MIPS: Build microMIPS VDSO for microMIPS kernels
MIPS: Fix sigreturn via VDSO on microMIPS kernel
MIPS: devicetree: fix cpu interrupt controller node-names
MIPS: VDSO: Build with `-fno-strict-aliasing'
MIPS: Pistachio: Enable KASLR
MIPS: lib: Mark intrinsics notrace
MIPS: Fix 64-bit HTW configuration
MIPS: Add 64-bit HTW fields
MAINTAINERS: Add file patterns for mips device tree bindings
MAINTAINERS: Add file patterns for mips brcm device tree bindings
MIPS: Simplify DSP instruction encoding macros
MIPS: Add missing tlbinvf/XPA microMIPS encodings
MIPS: Fix little endian microMIPS MSA encodings
MIPS: Add missing VZ accessor microMIPS encodings
MIPS: Add inline asm encoding helpers
MIPS: Spelling fix lets -> let's
MIPS: VR41xx: Fix typo
MIPS: oprofile: Fix typo
MIPS: math-emu: Fix typo
...
|
|
Various builds (such as i386:allmodconfig) fail with
fs/binfmt_aout.c:133:2: error: expected identifier or '(' before 'return'
fs/binfmt_aout.c:134:1: error: expected identifier or '(' before '}' token
[ Oops. My bad, I had stupidly thought that "allmodconfig" covered this
on x86-64 too, but it obviously doesn't. Egg on my face. - Linus ]
Fixes: 5d22fc25d4fc ("mm: remove more IS_ERR_VALUE abuses")
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
|
|
Pull string hash improvements from George Spelvin:
"This series does several related things:
- Makes the dcache hash (fs/namei.c) useful for general kernel use.
(Thanks to Bruce for noticing the zero-length corner case)
- Converts the string hashes in <linux/sunrpc/svcauth.h> to use the
above.
- Avoids 64-bit multiplies in hash_64() on 32-bit platforms. Two
32-bit multiplies will do well enough.
- Rids the world of the bad hash multipliers in hash_32.
This finishes the job started in commit 689de1d6ca95 ("Minimal
fix-up of bad hashing behavior of hash_64()")
The vast majority of Linux architectures have hardware support for
32x32-bit multiply and so derive no benefit from "simplified"
multipliers.
The few processors that do not (68000, h8/300 and some models of
Microblaze) have arch-specific implementations added. Those
patches are last in the series.
- Overhauls the dcache hash mixing.
The patch in commit 0fed3ac866ea ("namei: Improve hash mixing if
CONFIG_DCACHE_WORD_ACCESS") was an off-the-cuff suggestion.
Replaced with a much more careful design that's simultaneously
faster and better. (My own invention, as there was noting suitable
in the literature I could find. Comments welcome!)
- Modify the hash_name() loop to skip the initial HASH_MIX(). This
would let us salt the hash if we ever wanted to.
- Sort out partial_name_hash().
The hash function is declared as using a long state, even though
it's truncated to 32 bits at the end and the extra internal state
contributes nothing to the result. And some callers do odd things:
- fs/hfs/string.c only allocates 32 bits of state
- fs/hfsplus/unicode.c uses it to hash 16-bit unicode symbols not bytes
- Modify bytemask_from_count to handle inputs of 1..sizeof(long)
rather than 0..sizeof(long)-1. This would simplify users other
than full_name_hash"
Special thanks to Bruce Fields for testing and finding bugs in v1. (I
learned some humbling lessons about "obviously correct" code.)
On the arch-specific front, the m68k assembly has been tested in a
standalone test harness, I've been in contact with the Microblaze
maintainers who mostly don't care, as the hardware multiplier is never
omitted in real-world applications, and I haven't heard anything from
the H8/300 world"
* 'hash' of git://ftp.sciencehorizons.net/linux:
h8300: Add <asm/hash.h>
microblaze: Add <asm/hash.h>
m68k: Add <asm/hash.h>
<linux/hash.h>: Add support for architecture-specific functions
fs/namei.c: Improve dcache hash function
Eliminate bad hash multipliers from hash_32() and hash_64()
Change hash_64() return value to 32 bits
<linux/sunrpc/svcauth.h>: Define hash_str() in terms of hashlen_string()
fs/namei.c: Add hashlen_string() function
Pull out string hash to <linux/stringhash.h>
|
|
This will improve the performance of hash_32() and hash_64(), but due
to complete lack of multi-bit shift instructions on H8, performance will
still be bad in surrounding code.
Designing H8-specific hash algorithms to work around that is a separate
project. (But if the maintainers would like to get in touch...)
Signed-off-by: George Spelvin <linux@sciencehorizons.net>
Cc: Yoshinori Sato <ysato@users.sourceforge.jp>
Cc: uclinux-h8-devel@lists.sourceforge.jp
|
|
Microblaze is an FPGA soft core that can be configured various ways.
If it is configured without a multiplier, the standard __hash_32()
will require a call to __mulsi3, which is a slow software loop.
Instead, use a shift-and-add sequence for the constant multiply.
GCC knows how to do this, but it's not as clever as some.
Signed-off-by: George Spelvin <linux@sciencehorizons.net>
Cc: Alistair Francis <alistair.francis@xilinx.com>
Cc: Michal Simek <michal.simek@xilinx.com>
|
|
This provides a multiply by constant GOLDEN_RATIO_32 = 0x61C88647
for the original mc68000, which lacks a 32x32-bit multiply instruction.
Yes, the amount of optimization effort put in is excessive. :-)
Shift-add chain found by Yevgen Voronenko's Hcub algorithm at
http://spiral.ece.cmu.edu/mcm/gen.html
Signed-off-by: George Spelvin <linux@sciencehorizons.net>
Cc: Geert Uytterhoeven <geert@linux-m68k.org>
Cc: Greg Ungerer <gerg@linux-m68k.org>
Cc: Andreas Schwab <schwab@linux-m68k.org>
Cc: Philippe De Muyter <phdm@macq.eu>
Cc: linux-m68k@lists.linux-m68k.org
|
|
This is just the infrastructure; there are no users yet.
This is modelled on CONFIG_ARCH_RANDOM; a CONFIG_ symbol declares
the existence of <asm/hash.h>.
That file may define its own versions of various functions, and define
HAVE_* symbols (no CONFIG_ prefix!) to suppress the generic ones.
Included is a self-test (in lib/test_hash.c) that verifies the basics.
It is NOT in general required that the arch-specific functions compute
the same thing as the generic, but if a HAVE_* symbol is defined with
the value 1, then equality is tested.
Signed-off-by: George Spelvin <linux@sciencehorizons.net>
Cc: Geert Uytterhoeven <geert@linux-m68k.org>
Cc: Greg Ungerer <gerg@linux-m68k.org>
Cc: Andreas Schwab <schwab@linux-m68k.org>
Cc: Philippe De Muyter <phdm@macq.eu>
Cc: linux-m68k@lists.linux-m68k.org
Cc: Alistair Francis <alistai@xilinx.com>
Cc: Michal Simek <michal.simek@xilinx.com>
Cc: Yoshinori Sato <ysato@users.sourceforge.jp>
Cc: uclinux-h8-devel@lists.sourceforge.jp
|
|
Patch 0fed3ac866 improved the hash mixing, but the function is slower
than necessary; there's a 7-instruction dependency chain (10 on x86)
each loop iteration.
Word-at-a-time access is a very tight loop (which is good, because
link_path_walk() is one of the hottest code paths in the entire kernel),
and the hash mixing function must not have a longer latency to avoid
slowing it down.
There do not appear to be any published fast hash functions that:
1) Operate on the input a word at a time, and
2) Don't need to know the length of the input beforehand, and
3) Have a single iterated mixing function, not needing conditional
branches or unrolling to distinguish different loop iterations.
One of the algorithms which comes closest is Yann Collet's xxHash, but
that's two dependent multiplies per word, which is too much.
The key insights in this design are:
1) Barring expensive ops like multiplies, to diffuse one input bit
across 64 bits of hash state takes at least log2(64) = 6 sequentially
dependent instructions. That is more cycles than we'd like.
2) An operation like "hash ^= hash << 13" requires a second temporary
register anyway, and on a 2-operand machine like x86, it's three
instructions.
3) A better use of a second register is to hold a two-word hash state.
With careful design, no temporaries are needed at all, so it doesn't
increase register pressure. And this gets rid of register copying
on 2-operand machines, so the code is smaller and faster.
4) Using two words of state weakens the requirement for one-round mixing;
we now have two rounds of mixing before cancellation is possible.
5) A two-word hash state also allows operations on both halves to be
done in parallel, so on a superscalar processor we get more mixing
in fewer cycles.
I ended up using a mixing function inspired by the ChaCha and Speck
round functions. It is 6 simple instructions and 3 cycles per iteration
(assuming multiply by 9 can be done by an "lea" instruction):
x ^= *input++;
y ^= x; x = ROL(x, K1);
x += y; y = ROL(y, K2);
y *= 9;
Not only is this reversible, two consecutive rounds are reversible:
if you are given the initial and final states, but not the intermediate
state, it is possible to compute both input words. This means that at
least 3 words of input are required to create a collision.
(It also has the property, used by hash_name() to avoid a branch, that
it hashes all-zero to all-zero.)
The rotate constants K1 and K2 were found by experiment. The search took
a sample of random initial states (I used 1023) and considered the effect
of flipping each of the 64 input bits on each of the 128 output bits two
rounds later. Each of the 8192 pairs can be considered a biased coin, and
adding up the Shannon entropy of all of them produces a score.
The best-scoring shifts also did well in other tests (flipping bits in y,
trying 3 or 4 rounds of mixing, flipping all 64*63/2 pairs of input bits),
so the choice was made with the additional constraint that the sum of the
shifts is odd and not too close to the word size.
The final state is then folded into a 32-bit hash value by a less carefully
optimized multiply-based scheme. This also has to be fast, as pathname
components tend to be short (the most common case is one iteration!), but
there's some room for latency, as there is a fair bit of intervening logic
before the hash value is used for anything.
(Performance verified with "bonnie++ -s 0 -n 1536:-2" on tmpfs. I need
a better benchmark; the numbers seem to show a slight dip in performance
between 4.6.0 and this patch, but they're too noisy to quote.)
Special thanks to Bruce fields for diligent testing which uncovered a
nasty fencepost error in an earlier version of this patch.
[checkpatch.pl formatting complaints noted and respectfully disagreed with.]
Signed-off-by: George Spelvin <linux@sciencehorizons.net>
Tested-by: J. Bruce Fields <bfields@redhat.com>
|
|
The "simplified" prime multipliers made very bad hash functions, so get rid
of them. This completes the work of 689de1d6ca.
To avoid the inefficiency which was the motivation for the "simplified"
multipliers, hash_64() on 32-bit systems is changed to use a different
algorithm. It makes two calls to hash_32() instead.
drivers/media/usb/dvb-usb-v2/af9015.c uses the old GOLDEN_RATIO_PRIME_32
for some horrible reason, so it inherits a copy of the old definition.
Signed-off-by: George Spelvin <linux@sciencehorizons.net>
Cc: Antti Palosaari <crope@iki.fi>
Cc: Mauro Carvalho Chehab <m.chehab@samsung.com>
|
|
That's all that's ever asked for, and it makes the return
type of hash_long() consistent.
It also allows (upcoming patch) an optimized implementation
of hash_64 on 32-bit machines.
I tried adding a BUILD_BUG_ON to ensure the number of bits requested
was never more than 32 (most callers use a compile-time constant), but
adding <linux/bug.h> to <linux/hash.h> breaks the tools/perf compiler
unless tools/perf/MANIFEST is updated, and understanding that code base
well enough to update it is too much trouble. I did the rest of an
allyesconfig build with such a check, and nothing tripped.
Signed-off-by: George Spelvin <linux@sciencehorizons.net>
|
|
Finally, the first use of previous two patches: eliminate the
separate ad-hoc string hash functions in the sunrpc code.
Now hash_str() is a wrapper around hash_string(), and hash_mem() is
likewise a wrapper around full_name_hash().
Note that sunrpc code *does* call hash_mem() with a zero length, which
is why the previous patch needed to handle that in full_name_hash().
(Thanks, Bruce, for finding that!)
This also eliminates the only caller of hash_long which asks for
more than 32 bits of output.
The comment about the quality of hashlen_string() and full_name_hash()
is jumping the gun by a few patches; they aren't very impressive now,
but will be improved greatly later in the series.
Signed-off-by: George Spelvin <linux@sciencehorizons.net>
Tested-by: J. Bruce Fields <bfields@redhat.com>
Acked-by: J. Bruce Fields <bfields@redhat.com>
Cc: Jeff Layton <jlayton@poochiereds.net>
Cc: linux-nfs@vger.kernel.org
|
