The ifunc method avoids indirection via the function pointer
crc64_func. This works on GNU/Linux and probably on FreeBSD too.
The previous __attribute((__constructor__)) method is kept for
compatibility with ELF platforms which do support ifunc.
The ifunc method has some limitations, for example, building
liblzma with -fsanitize=address will result in segfaults.
The configure option --disable-ifunc must be used for such builds.
Thanks to Hans Jansen for the original patch.
Closes: https://github.com/tukaani-project/xz/pull/53
It gives C4146 here since unary minus with unsigned integer
is still unsigned (which is the intention here). Doing it
with substraction makes it clearer and avoids the warning.
Thanks to Nathan Moinvaziri for reporting this.
This affects only 32-bit x86 builds. x86-64 is OK as is.
I still cannot easily test this myself. The reporter has tested
this and it passes the tests included in the CMake build and
performance is good: raw CRC64 is 2-3 times faster than the
C version of the slice-by-four method. (Note that liblzma doesn't
include a MSVC-compatible version of the 32-bit x86 assembly code
for the slice-by-four method.)
Thanks to Iouri Kharon for figuring out a fix, testing, and
benchmarking.
This reverts commit 36edc65ab4.
It was reported that it wasn't a good enough fix and MSVC
still produced (different kind of) bad code when building
for 32-bit x86 if optimizations are enabled.
Thanks to Iouri Kharon.
I haven't tested with MSVC myself and there doesn't seem to be
information about the problem online, so I'm relying on the bug report.
Thanks to Iouri Kharon for the bug report and the patch.
It also works on E2K as it supports these intrinsics.
On x86-64 runtime detection is used so the code keeps working on
older processors too. A CLMUL-only build can be done by using
-msse4.1 -mpclmul in CFLAGS and this will reduce the library
size since the generic implementation and its 8 KiB lookup table
will be omitted.
On 32-bit x86 this isn't used by default for now because by default
on 32-bit x86 the separate assembly file crc64_x86.S is used.
If --disable-assembler is used then this new CLMUL code is used
the same way as on 64-bit x86. However, a CLMUL-only build
(-msse4.1 -mpclmul) won't omit the 8 KiB lookup table on
32-bit x86 due to a currently-missing check for disabled
assembler usage.
The configure.ac check should be such that the code won't be
built if something in the toolchain doesn't support it but
--disable-clmul-crc option can be used to unconditionally
disable this feature.
CLMUL speeds up decompression of files that have compressed very
well (assuming CRC64 is used as a check type). It is know that
the CLMUL code is significantly slower than the generic code for
tiny inputs (especially 1-8 bytes but up to 16 bytes). If that
is a real-world problem then there is already a commented-out
variant that uses the generic version for small inputs.
Thanks to Ilya Kurdyukov for the original patch which was
derived from a white paper from Intel [1] (published in 2009)
and public domain code from [2] (released in 2016).
[1] https://www.intel.com/content/dam/www/public/us/en/documents/white-papers/fast-crc-computation-generic-polynomials-pclmulqdq-paper.pdf
[2] https://github.com/rawrunprotected/crc
This uses it for CRC table initializations when using --disable-small.
It avoids mythread_once() overhead. It also means that then
--disable-small --disable-threads is thread-safe if this attribute
is supported.
When Intel CET is enabled, we need to include <cet.h> in assembly codes
to mark Intel CET support and add _CET_ENDBR to indirect jump targets.
Tested on Intel Tiger Lake under CET enabled Linux.
Using the aligned methods requires more care to ensure that
the address really is aligned, so it's nicer if the aligned
methods are prefixed. The next commit will remove the unaligned_
prefix from the unaligned methods which in liblzma are used in
more places than the aligned ones.
This is the sane thing to do. The conflict with OpenSSL
on some OSes and especially that the OS-provided versions
can be significantly slower makes it clear that it was
a mistake to have the external SHA-256 support enabled by
default.
Those who want it can now pass --enable-external-sha256 to
configure. INSTALL was updated with notes about OSes where
this can be a bad idea.
The SHA-256 detection code in configure.ac had some bugs that
could lead to a build failure in some situations. These were
fixed, although it doesn't matter that much now that the
external SHA-256 is disabled by default.
MINIX >= 3.2.0 uses NetBSD's libc and thus has SHA256_Init
in libc instead of libutil. Support for the libutil version
was removed.
The Maj macro is used where multiple things are added
together, so making Maj a sum of two expressions allows
some extra freedom for the compiler to schedule the
instructions.
I learned this trick from
<http://www.hackersdelight.org/corres.txt>.
This looks weird because the rotations become sequential,
but it helps quite a bit on both 32-bit and 64-bit x86:
- It requires fewer instructions on two-operand
instruction sets like x86.
- It requires one register less which matters especially
on 32-bit x86.
I hope this doesn't hurt other archs.
I didn't invent this idea myself, but I don't remember where
I saw it first.
This way a branch isn't needed for each operation
to choose between blk0 and blk2, and still the code
doesn't grow as much as it would with full unrolling.
If the operating system libc or other base libraries
provide SHA-256, use that instead of our own copy.
Note that this doesn't use OpenSSL or libgcrypt or
such libraries to avoid creating dependencies to
other packages.
This supports at least FreeBSD, NetBSD, OpenBSD, Solaris,
MINIX, and Darwin. They all provide similar but not
identical SHA-256 APIs; everyone is a little different.
Thanks to Wim Lewis for the original patch, improvements,
and testing.
This replaces bswap.h and integer.h.
The tuklib module uses <byteswap.h> on GNU,
<sys/endian.h> on *BSDs and <sys/byteorder.h>
on Solaris, which may contain optimized code
like inline assembly.
Seems that it is a problem in some cases if the same
version of XZ Utils produces different output on different
endiannesses, so this commit fixes that problem. The output
will still vary between different XZ Utils versions, but I
cannot avoid that for now.
This commit bloatens the code on big endian systems by 1 KiB,
which should be OK since liblzma is bloated already. ;-)
and use a fix that works on all systems using
GNU assembler.
Maybe the assembler code is used e.g. on Solaris x86
but let's worry about it if this doesn't work on it.
Don't use libtool convenience libraries to avoid recently
discovered long-standing subtle but somewhat severe bugs
in libtool (at least 1.5.22 and 2.2.6 are affected). It
was found when porting XZ Utils to Windows
<http://lists.gnu.org/archive/html/libtool/2009-06/msg00070.html>
but the problem is significant also e.g. on GNU/Linux.
Unless --disable-shared is passed to configure, static
library built from a set of convenience libraries will
contain PIC objects. That is, while libtool builds non-PIC
objects too, only PIC objects will be used from the
convenience libraries. On 32-bit x86 (tested on mobile XP2400+),
using PIC instead of non-PIC makes the decompressor 10 % slower
with the default CFLAGS.
So while xz was linked against static liblzma by default,
it got the slower PIC objects unless --disable-shared was
used. I tend develop and benchmark with --disable-shared
due to faster build time, so I hadn't noticed the problem
in benchmarks earlier.
This commit also adds support for building Windows resources
into liblzma and executables.
- Use call/ret pair to get instruction pointer for PIC.
- Use PIC only if PIC or __PIC__ is #defined.
- The code should work on MinGW and Darwin in addition
to GNU/Linux and Solaris.
Half of developers were already forgetting to use these
functions, which could have caused total breakage in some future
liblzma version or even now if --enable-small was used. Now
liblzma uses pthread_once() to do the initializations unless
it has been built with --disable-threads which make these
initializations thread-unsafe.
When --enable-small isn't used, liblzma currently gets needlessly
linked against libpthread (on systems that have it). While it is
stupid for now, liblzma will need threads in future anyway, so
this stupidity will be temporary only.
When --enable-small is used, different code CRC32 and CRC64 is
now used than without --enable-small. This made the resulting
binary slightly smaller, but the main reason was to clean it up
and to handle the lack of lzma_init_check().
The pkg-config file lzma.pc was renamed to liblzma.pc. I'm not
sure if it works correctly and portably for static linking
(Libs.private includes -pthread or other operating system
specific flags). Hopefully someone complains if it is bad.
lzma_rc_prices[] is now included as a precomputed array even
with --enable-small. It's just 128 bytes now that it uses uint8_t
instead of uint32_t. Smaller array seemed to be at least as fast
as the more bloated uint32_t array on x86; hopefully it's not bad
on other architectures.
broken. API has changed a lot and it will still change a
little more here and there. The command line tool doesn't
have all the required changes to reflect the API changes, so
it's easy to get "internal error" or trigger assertions.
specification. Simplify things by removing most of the
support for known uncompressed size in most places.
There are some miscellaneous changes here and there too.
The API of liblzma has got many changes and still some
more will be done soon. While most of the code has been
updated, some things are not fixed (the command line tool
will choke with invalid filter chain, if nothing else).
Subblock filter is somewhat broken for now. It will be
updated once the encoded format of the Subblock filter
has been decided.
or no inttypes.h. This is useful when the compiler has
good enough support for C99, but libc headers don't.
Changed liblzma API so that sys/types.h and inttypes.h
have to be #included before #including lzma.h. On systems
that don't have C99 inttypes.h, it's the problem of the
applications to provide the required types and macros
before #including lzma.h.
If lzma.h defined the missing types and macros, it could
conflict with third-party applications whose configure
has detected that the types are missing and defined them
in config.h already. An alternative would have been
introducing lzma_uint32 and similar types, but that would
just be an extra pain on modern systems.
so that the preprocessor removes the /* */ style comments,
which are not supported by some non-GNU assemblers (Solaris)
that otherwise work with this code.
Lasse Collin
Ed Maste
H.J. Lu
Jim Meyering