Meghana 28bb28b79f Modified Function definition for BLAS and CBLAS interfaces of DOTV and SWAPV APIs ...

Details:
-Kernel is called directly from API call to avoid framework
 overhead in case of single and double precisions.
-Currently these changes are applicable only for zen2 configuration.
 They will be enabled for zen family processors in future.
-These changes improve performance of BLAS and CBLAS interfaces of API.
 They do not affect BLIS-specific APIs.

Change-Id: I1eb7ca470ced82c3cfa8b22f2b53000d42fef96c
Signed-off-by: Meghana Vankadari <Meghana.Vankadari@amd.com>
AMD-Internal: [CPUPL-847,CPUPL-816]

2020-05-04 15:06:07 +05:30

..

0

Use void_fp for function pointers instead of void*.

2019-08-23 14:18:08 +05:30

1

Use void_fp for function pointers instead of void*.

2019-08-23 14:18:08 +05:30

1d

Use void_fp for function pointers instead of void*.

2019-08-23 14:18:08 +05:30

1f

Use void_fp for function pointers instead of void*.

2019-08-23 14:18:08 +05:30

1m

Merge branch 'master' into amd

2019-10-11 11:53:51 -05:00

2

Use void_fp for function pointers instead of void*.

2019-08-23 14:18:08 +05:30

3

Made some critical changes to small_gemm kernels

2020-03-19 16:30:04 +05:30

base

JIRA: CPUPL-853: Fix for the redefinition of _unsigned int __get_cpuid_max(unsigned int, unsigned int*)_. http://ontrack-internal.amd.com/browse/CPUPL-853 https://github.com/flame/blis/issues/393

2020-04-23 17:14:24 +05:30

compat

Modified Function definition for BLAS and CBLAS interfaces of DOTV and SWAPV APIs

2020-05-04 15:06:07 +05:30

include

Modified Function definition for BLAS and CBLAS interfaces of DOTV and SWAPV APIs

2020-05-04 15:06:07 +05:30

ind

"Merge Selective Packing code from amd branch flame/blis"

2019-12-16 14:48:53 +05:30

thread

Added missing return to bli_thread_partition_2x2().

2020-03-16 12:28:25 +05:30

util

Code review + tweaks to AMD's AOCL 2.0 PR (#349).

2019-10-11 10:24:24 -05:00