mirror of
https://github.com/amd/blis.git
synced 2026-05-12 10:05:38 +00:00
09e34fd2bd
1. Two CGEMM function pointers are added for different storage schemes
1. bli_cgemmsup_rv_zen_asm_3x8m
2. bli_cgemmsup_rv_zen_asm_3x8n
2. In previous commit:
(Level-3 triangular routines now use different block sizes and kernels
Commit Id: 79e174ff0a)
1. bli_cntx_set_l3_sup_tri_kers cntx function was created
2. Function holds optimised function pointers for GEMMT/SYRK API's
3. It avoids over riding default block sizes which improves the
performance
4. This function did not include optimised CGEMM function pointers
leading to regression as reference kernels were invoked
3. With this commit, 2 optimized CGEMM function pointers are added in
bli_cntx_set_l3_sup_tri_kers
1. This fixes the regression as optimized CGEMM functions are invoked
AMD-Internal: [CPUPL-3831] [CPUPL-3830]
Change-Id: Ie8b41a5e62439de2a65e7df0b07d63ee2383e51e
For more information on sub-configurations and configuration families in BLIS, please read the Configuration Guide, which can be viewed in markdown-rendered form from the BLIS wiki page.
If you don't have time, or are impatient, take a look at the config_registry
file in the top-level directory of the BLIS distribution. It contains a
grammar-like mapping of configuration names, or families, to sub-configurations,
which may be other families. Keep in mind that the / notation:
<config>: <config>/<name>
means that the kernel set associated with <name> should be made available to
the configuration <config> if <config> is targeted at configure-time.
(Some configurations borrow kernels from other configurations, and this is how
we specify that requirement.)