projects / platform / upstream / openblas.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 69edc5b) | patch
Fix inline assembly constraints in Bulldozer TRSM kernels
authorMartin Kroeker <martin@ruby.chemie.uni-freiburg.de>
Sat, 16 Feb 2019 19:06:48 +0000 (20:06 +0100)
committerGitHub <noreply@github.com>
Sat, 16 Feb 2019 19:06:48 +0000 (20:06 +0100)
commitf9bb76d29af48f448a8ab2bdfffc962d9623a3df
treec02800ca64b53784f44dbe504fa36cb0d442c0e5tree | snapshot
parent69edc5bbe79af88710666aa909e7b39c89558b9ccommit | diff
Fix inline assembly constraints in Bulldozer TRSM kernels

rework indices to allow marking i,as and bs as both input and output (marked operand n1 as well for simplicity). For #2009
kernel/x86_64/dtrsm_kernel_RT_bulldozer.c diff | blob | history
kernel/x86_64/strsm_kernel_LN_bulldozer.c diff | blob | history
kernel/x86_64/strsm_kernel_LT_bulldozer.c diff | blob | history
kernel/x86_64/strsm_kernel_RN_bulldozer.c diff | blob | history
kernel/x86_64/strsm_kernel_RT_bulldozer.c diff | blob | history
Domain: Machine Learning / ML Framework;