llvm-project/llvm/test/tools/llvm-mca
David Green e9adcbde31 [AArch64] Model Cortex-A55 Q register NEON instructions
Cortex-A55 has 2 64bit NEON vector units, meaning a 128bit instruction
requires taking both units (and can only be issued as the first
instruction in a dual issue pair). This patch models that by splitting
the WriteV SchedWrite into two - the WriteVd that reads/writes only
64bit operands, and the WriteVq that read/writes 128bit registers. The
A55 schedule then uses this distinction to model the WriteVq as taking
both resource units, and starting a Schedule Group and WriteVd as taking
one as before.

I believe this is more correct, even if it does not lead to much better
performance.

Differential Revision: https://reviews.llvm.org/D108766
2021-09-29 16:55:31 +01:00
..
AArch64 [AArch64] Model Cortex-A55 Q register NEON instructions 2021-09-29 16:55:31 +01:00
AMDGPU [MCA] Adding an AMDGPUCustomBehaviour implementation. 2021-08-24 13:33:58 -07:00
ARM [MCA][InstrBuilder] Check for the presence of flag VariadicOpsAreDefs. 2021-06-15 09:52:38 +01:00
JSON/X86 [llvm-mca][JSON] Store extra information about driver flags used for the simulation 2021-07-16 09:18:40 +02:00
SystemZ
X86 [X86][SLM] Fix ADDQ/SUBQ/CMPEQQ throughput to account for running on either port. 2021-09-24 10:06:14 +01:00
invalid_input_file_name.test [test] Use host platform specific error message substitution in lit tests 2021-01-29 07:16:30 -05:00
lit.local.cfg