Script 'mail_helper' called by obssrc Hello community, here is the log from the commit of package htscodecs for openSUSE:Factory checked in at 2022-09-11 21:57:28 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ Comparing /work/SRC/openSUSE:Factory/htscodecs (Old) and /work/SRC/openSUSE:Factory/.htscodecs.new.2083 (New) ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Package is "htscodecs" Sun Sep 11 21:57:28 2022 rev:2 rq:1002486 version:1.3.0 Changes: -------- --- /work/SRC/openSUSE:Factory/htscodecs/htscodecs.changes 2022-05-04 15:10:57.480157813 +0200 +++ /work/SRC/openSUSE:Factory/.htscodecs.new.2083/htscodecs.changes 2022-09-11 21:57:30.953656204 +0200 @@ -1,0 +2,23 @@ +Thu Sep 8 21:52:47 UTC 2022 - Stefan Br??ns <stefan.bru...@rwth-aachen.de> + +- Update to 1.3.0: + * The primary change in this release is a new SIMD enabled + rANS codec. + There is a 32-way unrolled rANS implementation. This is accessed + using the existing rans 4x16 API with the RANS_ORDER_X32 bit set. + * Improved memory allocation via a new htscodecs_tls_alloc function. + * Some external functions have been renamed, with the old ones + still existing in a deprecated fashion. + * Improved test framework with an "entropy" tool that iterates + over all entropy encoders. + * Reworked fuzzing infrastructure. + * Small speed improvements to various rANS encoders and decoders. + * Substantial memory reduction to the name tokeniser (tok3). + * Fixed undefined behaviour in our use of _builtin_clz(). + * Fixed a few redundant #includes. + * Work around strict aliasing bugs, uncovered with gcc -O2. + * Fixed an issue with encoding data blocks close to 2GB in size. + * Fix encode error with large blocks using RANS_ORDER_STRIPE. +- Add fix_ix86_build.patch + +------------------------------------------------------------------- Old: ---- htscodecs-1.2.2.tar.gz New: ---- fix_ix86_build.patch htscodecs-1.3.0.tar.gz ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ Other differences: ------------------ ++++++ htscodecs.spec ++++++ --- /var/tmp/diff_new_pack.ARk3Sa/_old 2022-09-11 21:57:31.473657668 +0200 +++ /var/tmp/diff_new_pack.ARk3Sa/_new 2022-09-11 21:57:31.477657679 +0200 @@ -18,7 +18,7 @@ %define sonum 2 Name: htscodecs -Version: 1.2.2 +Version: 1.3.0 Release: 0 Summary: C library for custom compression for CRAM and other formats License: MIT @@ -26,6 +26,9 @@ URL: https://github.com/samtools/htscodecs Source0: https://github.com/samtools/htscodecs/releases/download/v%{version}/htscodecs-%{version}.tar.gz Source100: baselibs.conf +# PATCH-FIX-UPSTREAM +Patch0: https://github.com/samtools/htscodecs/commit/843d4f63b1c64905881b4648916a4d027baa1a1c.patch#/fix_ix86_build.patch +BuildRequires: autoconf BuildRequires: gcc-c++ BuildRequires: pkgconfig BuildRequires: pkgconfig(bzip2) @@ -58,9 +61,11 @@ Header files and libraries of the samtools project for compiling against %{name}. %prep -%setup -q +%autosetup -p1 %build +# Rebuild configure script after Patch0 +autoconf %configure --disable-static %make_build @@ -76,7 +81,7 @@ %{_libdir}/lib%{name}.so.* %files devel -%doc README.md NEWS +%doc README.md NEWS.md %{_includedir}/htscodecs %{_libdir}/lib%{name}.so ++++++ fix_ix86_build.patch ++++++ >From 843d4f63b1c64905881b4648916a4d027baa1a1c Mon Sep 17 00:00:00 2001 From: James Bonfield <j...@sanger.ac.uk> Date: Wed, 17 Aug 2022 15:27:04 +0100 Subject: [PATCH] Improvements for intel -m32 builds. On this platform _mm256_extract_epi64 isn't defined, but the rest of AVX2 is. It needs to fail auto-detection. Also we get unaligned accesses in the SSE4 code with tbuf due to differing data alignment caused by 32-bit pointers instead of 64-bit. This exposes an underlying problem of using aligned SIMD writes on tbuf without explicitly asking for alignment. (The new code is also sometimes a little faster.) See also samtools/htslib#1500 --- configure.ac | 3 ++- htscodecs/rANS_static32x16pr_avx2.c | 2 +- htscodecs/rANS_static32x16pr_avx512.c | 2 +- htscodecs/rANS_static32x16pr_sse4.c | 2 +- m4/ax_check_compile_flag.m4 | 6 +++++- 5 files changed, 10 insertions(+), 5 deletions(-) diff --git a/configure.ac b/configure.ac index 4c8ad54..eb2f44e 100644 --- a/configure.ac +++ b/configure.ac @@ -165,7 +165,8 @@ AX_CHECK_COMPILE_FLAG([-mavx2], [ ]],[[ __m256i a = _mm256_set_epi32(1, 2, 3, 4, 5, 6, 7, 8); __m256i b = _mm256_add_epi32(a, a); - return *((char *) &b); + long long c = _mm256_extract_epi64(b, 0); + return (int) c; ]]) ]) AM_CONDITIONAL([RANS_32x16_AVX2],[test "x$MAVX2" != "x"]) diff --git a/htscodecs/rANS_static32x16pr_avx2.c b/htscodecs/rANS_static32x16pr_avx2.c index 2dd14f1..c36f23c 100644 --- a/htscodecs/rANS_static32x16pr_avx2.c +++ b/htscodecs/rANS_static32x16pr_avx2.c @@ -1123,7 +1123,7 @@ unsigned char *rans_uncompress_O1_32x16_avx2(unsigned char *in, union { unsigned char tbuf[32][32]; uint64_t tbuf64[32][4]; - } u; + } u __attribute__((aligned(32))); unsigned int tidx = 0; if (0) { diff --git a/htscodecs/rANS_static32x16pr_avx512.c b/htscodecs/rANS_static32x16pr_avx512.c index 0d1456f..e9cce46 100644 --- a/htscodecs/rANS_static32x16pr_avx512.c +++ b/htscodecs/rANS_static32x16pr_avx512.c @@ -735,7 +735,7 @@ unsigned char *rans_uncompress_O1_32x16_avx512(unsigned char *in, union { unsigned char tbuf[32][32]; uint64_t tbuf64[32][4]; - } u; + } u __attribute__((aligned(32))); #else uint32_t tbuf[32][32]; #endif diff --git a/htscodecs/rANS_static32x16pr_sse4.c b/htscodecs/rANS_static32x16pr_sse4.c index fe0345f..88bcbe0 100644 --- a/htscodecs/rANS_static32x16pr_sse4.c +++ b/htscodecs/rANS_static32x16pr_sse4.c @@ -1423,7 +1423,7 @@ unsigned char *rans_uncompress_O1_32x16_sse4(unsigned char *in, uint16_t *sp = (uint16_t *)ptr; const uint32_t mask = ((1u << TF_SHIFT_O1_FAST)-1); __m128i maskv = _mm_set1_epi32(mask); // set mask in all lanes - uint8_t tbuf[32][32]; + uint8_t tbuf[32][32] __attribute__((aligned(32))); int tidx = 0; LOAD128(Rv, R); LOAD128(Lv, l); diff --git a/m4/ax_check_compile_flag.m4 b/m4/ax_check_compile_flag.m4 index bd753b3..c2bd2c6 100644 --- a/m4/ax_check_compile_flag.m4 +++ b/m4/ax_check_compile_flag.m4 @@ -36,13 +36,17 @@ #serial 6 +# LOCAL modification; change AC_COMPILE_IFELSE to AC_LINK_IFELSE so +# _mm256_extract_epi64 tests on a x86_64 running under -m32 still +# fails. (Otherwise it compiles, but fails to link.) + AC_DEFUN([AX_CHECK_COMPILE_FLAG], [AC_PREREQ(2.64)dnl for _AC_LANG_PREFIX and AS_VAR_IF AS_VAR_PUSHDEF([CACHEVAR],[ax_cv_check_[]_AC_LANG_ABBREV[]flags_$4_$1])dnl AC_CACHE_CHECK([whether _AC_LANG compiler accepts $1], CACHEVAR, [ ax_check_save_flags=$[]_AC_LANG_PREFIX[]FLAGS _AC_LANG_PREFIX[]FLAGS="$[]_AC_LANG_PREFIX[]FLAGS $4 $1" - AC_COMPILE_IFELSE([m4_default([$5],[AC_LANG_PROGRAM()])], + AC_LINK_IFELSE([m4_default([$5],[AC_LANG_PROGRAM()])], [AS_VAR_SET(CACHEVAR,[yes])], [AS_VAR_SET(CACHEVAR,[no])]) _AC_LANG_PREFIX[]FLAGS=$ax_check_save_flags]) ++++++ htscodecs-1.2.2.tar.gz -> htscodecs-1.3.0.tar.gz ++++++ ++++ 26090 lines of diff (skipped)
