Nallani Bhaskar 1d31386c02 Fixed few out of bound memory reads in sgemmsup kernels ...

Details:
 Fixed memory access bugs in the bli_sgemmsup_rd_zen_asm_s1x16()
  kernel. The bugs were caused by loading four
  single-precision elements of C, via instructions such as:

	vfmadd231ps(mem(rcx, 0*32), ymm3, ymm4)

        or

        vfmadd231ps(mem(rcx, 0*32), xmm3, xmm4)

  in situations where only two elements are guaranteed to exist. (These
  bugs may not have manifested in earlier tests due to the leading
  dimension alignment that BLIS employs by default.) The issue was fixed
  by replacing lines like the one above with:

	vmovsd(mem(rcx), xmm0)
	vfmadd231ps(xmm0, xmm3, xmm4)

  Thus, we use vmovsd to explicitly load only two elements of C into
  registers, and then operate on those values using register addressing.

  AMD_CPUPLID: CPUPL-2279

Change-Id: Ic39290d651f5218b2e548351a87ac5e4b5b79c68

2022-07-29 09:09:12 -04:00

..

1

Added support for AVX512 for Windows and AMAVX

2022-06-08 11:09:48 +05:30

1f

Removed Arch specific code from BLIS framework.

2022-05-17 20:35:40 +05:30

2

Fixed CMake files for HER

2022-06-01 12:25:43 +05:30

3

Fixed few out of bound memory reads in sgemmsup kernels

2022-07-29 09:09:12 -04:00

util

DGEMMT : Tuning SUP threshold to improve ST and MT performance.

2022-05-17 18:09:22 +05:30

bli_kernels_zen.h

ZGEMM kernel performance improvement for k=1 sizes:

2022-06-30 07:19:52 -04:00

CMakeLists.txt

Removed packm kernels of zen

2021-11-12 08:58:51 +05:30