Hi all,
I'd like to commit this instead of the OK'd patch 05/14 from the
unpacked FP series [1], as I later realised that:
* I can use the existing unpacked_fcm_1.c file to test these
splitting patterns, since the stores to out[i] are conditional
on the comparison result and the loop mask.
* I missed an existing optimisation that we have for BIC, where
we prefer NOT and do-away with a PTRUE.
Bootstrapped & regtested on aarch64-linux-gnu. OK for master?
Thanks,
Spencer
[1]
https://inbox.sourceware.org/gcc-patches/20250602100707.2659-6-spencer.ab...@arm.com/
Spencer Abson (1):
aarch64: AND/BIC combines for unpacked SVE FP comparisons
gcc/config/aarch64/aarch64-sve.md | 26 +++++++-------
.../aarch64/sve/unpacked_fcm_combines_1.c | 17 +++++++++
.../aarch64/sve/unpacked_fcm_combines_2.c | 35 +++++++++++++++++++
3 files changed, 65 insertions(+), 13 deletions(-)
create mode 100644
gcc/testsuite/gcc.target/aarch64/sve/unpacked_fcm_combines_1.c
create mode 100644
gcc/testsuite/gcc.target/aarch64/sve/unpacked_fcm_combines_2.c
--
2.34.1
