|
1 | 1 | OpenBLAS ChangeLog
|
| 2 | +==================================================================== |
| 3 | +Version 0.3.15 |
| 4 | + 2-May-2021 |
| 5 | + |
| 6 | +common: |
| 7 | + - imported improvements and bugfixes from Reference-LAPACK 3.9.1 |
| 8 | + - imported LAPACKE interface fixes from Reference-LAPACK PRs 534 + 537 |
| 9 | + - fixed a problem in the cpu detection of 0.3.14 that prevented cross-compilation |
| 10 | + - fixed a sequence problem in the generation of softlinks to the library in GMAKE |
| 11 | + |
| 12 | +RISC V: |
| 13 | + - fixed compilation on RISCV (missing entry in getarch) |
| 14 | + - fixed a potential division by zero in CROTG and ZROTG |
| 15 | + |
| 16 | +POWER: |
| 17 | + - fixed LAPACK testsuite failures seen with the NVIDIA HPC compiler |
| 18 | + - improved CGEMM, DGEMM and ZGEMM performance on POWER10 |
| 19 | + - added an optimized ZGEMV kernel for POWER10 |
| 20 | + - fixed a potential division by zero in CROTG and ZROTG |
| 21 | + |
| 22 | +x86_64: |
| 23 | + - added support for Intel Control-flow Enforcement Technology (CET) |
| 24 | + - reverted the DOMATCOPY_RT code to the generic C version |
| 25 | + - fixed a bug in the AVX512 SGEMM kernel introduced in 0.3.14 |
| 26 | + - fixed misapplication of -msse flag to non-SSE cpus in DYNAMIC_ARCH |
| 27 | + - added support for compilation of the benchmarks on older OSX versions |
| 28 | + - fix propagation of the NO_AVX512 option in CMAKE builds |
| 29 | + - fix compilation of the AVX512 SGEMM kernel with clang-cl on Windows |
| 30 | + - fixed compilation of the CTESTs with INTERFACE64=1 (random faults on OSX) |
| 31 | + - corrected the Haswell DROT kernel to require AVX2/FMA3 rather than AVX512 |
| 32 | + |
| 33 | +ARM: |
| 34 | + - fixed a potential division by zero in CROTG and ZROTG |
| 35 | + - fixed a potential overflow in IMATCOPY/ZIMATCOPY and the CTESTs |
| 36 | + |
| 37 | +ARM64: |
| 38 | + - fixed spurious reads outside the array in the SGEMM tcopy macro |
| 39 | + - fixed a potential division by zero in CROTG and ZROTG |
| 40 | + - fixed a segmentation fault in DYNAMIC_ARCH builds (reappeared in 0.3.14) |
| 41 | + |
| 42 | +MIPS |
| 43 | + - fixed a potential division by zero in CROTG and ZROTG |
| 44 | + - fixed a potential overflow in IMATCOPY/ZIMATCOPY and the CTESTs |
| 45 | + |
| 46 | +MIPS64: |
| 47 | + - fixed a potential division by zero in CROTG and ZROTG |
| 48 | + |
| 49 | +SPARC: |
| 50 | + - fixed a potential division by zero in CROTG and ZROTG |
| 51 | + |
2 | 52 | ====================================================================
|
3 | 53 | Version 0.3.14
|
4 | 54 | 17-Mar-2021
|
|
0 commit comments