Script 'mail_helper' called by obssrc Hello community, here is the log from the commit of package primesieve for openSUSE:Factory checked in at 2026-11-08 19:22:15 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ Comparing /work/SRC/openSUSE:Factory/primesieve (Old) and /work/SRC/openSUSE:Factory/.primesieve.new.3759290 (New) ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Package is "primesieve" Sun Nov 8 19:22:15 2026 rev:27 rq:1396127 version:12.17 Changes: -------- --- /work/SRC/openSUSE:Factory/primesieve/primesieve.changes 2026-09-29 17:55:27.457396896 +0200 +++ /work/SRC/openSUSE:Factory/.primesieve.new.3759290/primesieve.changes 2026-11-08 19:22:41.471631881 +0100 @@ -1,0 +2,34 @@ +Sat Nov 07 15:44:00 UTC 2026 - Kim Walisch <[email protected]> + +- Update to primesieve-12.17 +- PreSieve_arm_sve.hpp: Fix ambiguous svwhilelt_b8() call on macOS +- ci.yml: Compile Linux x64 tests with -march=x86-64, Ubuntu 26.04 enables AVX2 +- primesieveConfig.cmake.in: Avoid overwriting the LOWER_COMPONENTS variable of consumers +- CMake: Prefix internal cache variables with PRIMESIEVE_ to avoid name collisions +- PreSieveTables.hpp: Simplify the lookup table generator program +- CMake: Require C++14 for the executable and tests +- Vector.hpp: Make Array member functions constexpr +- PreSieveTables.hpp: Fix MSVC initializing preSieveTables at startup +- scripts: Remove unused build_mingw64_x64.sh and build_mingw64_arm64.sh +- Add AGENTS.md file for AI agents +- ci.yml: Migrate Windows GCC CI from MINGW64 to UCRT64 +- ci.yml: Check that SIMD runtime dispatch probes succeed +- CMake: Reuse cpu_arch_macros.hpp in SIMD runtime dispatch probes +- Implement SIMD runtime dispatching for MSVC compiler +- main.cpp: Add -march=native support to --cpu-info option +- popcnt.hpp: Add ARM NEON popcnt64() implementation for MSVC compiler +- popcnt.hpp: Get rid of popcnt64_native(), use _mm_popcnt_u64() instead +- cpu_arch_macros.hpp: New header to simplify SIMD preprocessor logic +- PrimeGenerator_arm_neon.hpp: Convert 1-bits from sieve array to primes using ARM NEON +- PrimeGenerator_x86_avx512.hpp: New AVX512 algorithm with fewer branch mispredictions +- PrimeGenerator_x86_avx2.hpp: Convert 1-bits from sieve array to primes using AVX2 +- PrimeGenerator.hpp: Update SIMD runtime dispatching +- PreSieve.cpp: Update SIMD runtime dispatching +- PreSieve.cpp: Use ARM NEON only on 64-bit ARM CPUs +- PreSieve.cpp: Enable SSE2 algorithm for MSVC on x64 +- PreSieve_x86_avx2.hpp: Add AVX2 pre-sieving algorithm +- CMakeLists.txt: Building now requires C++14 (linking still works using C++11) +- macros.hpp: Add NO_UNROLL_LOOP and NO_VECTORIZE_LOOP macros +- CpuInfo.cpp: Fix uninitialized read of 32-bit hw.logicalcpu sysctl on macOS + +------------------------------------------------------------------- Old: ---- primesieve-12.16.tar.gz New: ---- primesieve-12.17.tar.gz ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ Other differences: ------------------ ++++++ primesieve.spec ++++++ --- /var/tmp/diff_new_pack.qXrH11/_old 2026-11-08 19:22:42.276665183 +0100 +++ /var/tmp/diff_new_pack.qXrH11/_new 2026-11-08 19:22:42.278665266 +0100 @@ -17,7 +17,7 @@ Name: primesieve -Version: 12.16 +Version: 12.17 Release: 0 Summary: A prime number generator License: BSD-2-Clause ++++++ primesieve-12.16.tar.gz -> primesieve-12.17.tar.gz ++++++ ++++ 16589 lines of diff (skipped)
