I saw this tweet claiming this flag makes libdeflate run 20% faster on
WebAssembly: https://twitter.com/Algunenano/status/1317098341377900550.
Indeed, when tried even in a complex PNG compression benchmark I've
observed 10-15% improvement when this flag is enabled.
Even though WebAssembly might be running on top of a variety of
underlying platforms, the spec requires it to support unaligned access,
and on majority of platforms it will translate to a faster code.
Hence, I think it makes sense to enable this flag by default.
This is needed to avoid the following error when using
-fsanitize=undefined with gcc:
lib/x86/adler32_impl.h:214:2: runtime error: signed integer overflow:
1951294680 + 1956941400 cannot be represented in type 'int'
Note that this isn't seen when using -fsanitize=undefined with clang.
Old compilers don't have unsigned vector types, so work around that.
If support for CRC32 instructions is detected, set
ARM_CPU_FEATURE_CRC32. Also define
COMPILER_SUPPORTS_CRC32_TARGET_INTRINSICS when appropriate, and update
run_tests.sh to toggle the crc32 feature for testing.
gcc 10 is miscompiling libdeflate on x86_64 at -O3 due to a regression
in how packed structs are handled
(https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94994).
Work around this by just always using memcpy() for unaligned accesses.
It's unclear that the "packed struct" approach is worthwhile to maintain
anymore. Currently I'm only aware that it's useful with old gcc's on
arm32. Hopefully, compilers are good enough now that we can simply use
memcpy() everywhere.
Update https://github.com/ebiggers/libdeflate/issues/64
NEON intrinsics cannot be used when compiling for an ARM CPU without
hardware floating point support, e.g. the Debian armel port. In this
case arm_neon.h cannot even be included as it causes an #error.
[Based on a patch by Adrian Bunk <bunk@debian.org>, but changed to check
for __ARM_FP instead of !__SOFTFP__ to be consistent with arm_neon.h,
and added a comment.]
All 64bit PowerPC CPUs handle unaligned accesses reasonably fast, so
set UNALIGNED_ACCESS_IS_FAST.
Decompression of the snappy html test case is almost 50% faster on
POWER9 with this patch applied.
Now that we detect CPU features on 32-bit x86, allow the SSE2
implementation of Adler-32 to be selected at runtime based on the
presence of the SSE2 feature.
Replace COMPILER_SUPPORTS_TARGET_INTRINSICS with macros for the
individual features, since COMPILER_SUPPORTS_TARGET_INTRINSICS was
x86-specific and would cause confusion when we try to use intrinsics in
'target' functions for other architectures.
With clang 3.9:
warning: macro expansion producing 'defined' has undefined
behavior [-Wexpansion-to-defined]
Just eliminate the tests for clang and icc; they shouldn't be necessary.
* Bring in common headers and program code from xpack project
* Move program code to programs/
* Move library code to lib/
* GNU89 and MSVC2010 compatibility
* Other changes
