https://gcc.gnu.org/g:f0e56a8159d60e2f9ceb50f4f918f0e93e38c420
commit f0e56a8159d60e2f9ceb50f4f918f0e93e38c420
Author: Michael Meissner <meiss...@linux.ibm.com>
Date: Wed Jul 9 01:52:03 2025 -0400
PR target/117251: Improve vector and to vector nor fusion
See the following post for a complete explanation of what the patches
for PR target/117251:
* https://gcc.gnu.org/pipermail/gcc-patches/2025-June/686474.html
This is patch #34 of 45 to generate the 'XXEVAL' instruction on power10
and power11 instead of using the Altivec 'VAND' instruction feeding
into 'VNOR'. The 'XXEVAL' instruction can use all 64 vector registers,
instead of the 32 registers that traditional Altivec vector
instructions use. By allowing all of the vector registers to be used,
it reduces the amount of spilling that a large benchmark generated.
Currently the following code:
vector int a, b, c, d;
a = ~ ((c & d) | b);
Generates:
vand t,c,d
vnor a,t,b
Now in addition with this patch, if the arguments or result is
allocated to a traditional FPR register, the GCC compiler will now
generate the following code instead of adding vector move instructions:
xxeval a,b,c,224
Since fusion using 2 Altivec instructions is slightly faster than using
the 'XXEVAL' instruction we prefer to generate the Altivec instructions
if we can. In addition, because 'XXEVAL' is a prefixed instruction, it
possibly might generate an extra NOP instruction to align the 'XXEVAL'
instruction.
I have tested these patches on both big endian and little endian
PowerPC servers, with no regressions. Can I check these patchs into
the trunk?
2025-07-09 Michael Meissner <meiss...@linux.ibm.com>
gcc/
PR target/117251
* config/rs6000/fusion.md: Regenerate.
* config/rs6000/genfusion.pl (gen_logical_addsubf): Add support
to generate vector and => nor fusion if XXEVAL is supported.
Diff:
---
gcc/config/rs6000/fusion.md | 15 +++++++++------
gcc/config/rs6000/genfusion.pl | 1 +
2 files changed, 10 insertions(+), 6 deletions(-)
diff --git a/gcc/config/rs6000/fusion.md b/gcc/config/rs6000/fusion.md
index e3d9f7376a8d..68b52d4f5893 100644
--- a/gcc/config/rs6000/fusion.md
+++ b/gcc/config/rs6000/fusion.md
@@ -2480,20 +2480,23 @@
;; logical-logical fusion pattern generated by gen_logical_addsubf
;; vector vand -> vnor
(define_insn "*fuse_vand_vnor"
- [(set (match_operand:VM 3 "altivec_register_operand" "=&0,&1,&v,v")
- (and:VM (not:VM (and:VM (match_operand:VM 0 "altivec_register_operand"
"v,v,v,v")
- (match_operand:VM 1 "altivec_register_operand"
"v,v,v,v")))
- (not:VM (match_operand:VM 2 "altivec_register_operand"
"v,v,v,v"))))
- (clobber (match_scratch:VM 4 "=X,X,X,&v"))]
+ [(set (match_operand:VM 3 "vector_fusion_operand" "=&0,&1,&v,wa,v")
+ (and:VM (not:VM (and:VM (match_operand:VM 0 "vector_fusion_operand"
"v,v,v,wa,v")
+ (match_operand:VM 1 "vector_fusion_operand"
"v,v,v,wa,v")))
+ (not:VM (match_operand:VM 2 "vector_fusion_operand"
"v,v,v,wa,v"))))
+ (clobber (match_scratch:VM 4 "=X,X,X,X,&v"))]
"(TARGET_P10_FUSION)"
"@
vand %3,%1,%0\;vnor %3,%3,%2
vand %3,%1,%0\;vnor %3,%3,%2
vand %3,%1,%0\;vnor %3,%3,%2
+ xxeval %x3,%x2,%x1,%x0,224
vand %4,%1,%0\;vnor %3,%4,%2"
[(set_attr "type" "fused_vector")
(set_attr "cost" "6")
- (set_attr "length" "8")])
+ (set_attr "length" "8")
+ (set_attr "prefixed" "*,*,*,yes,*")
+ (set_attr "isa" "*,*,*,xxeval,*")])
;; logical-logical fusion pattern generated by gen_logical_addsubf
;; vector vandc -> vnor
diff --git a/gcc/config/rs6000/genfusion.pl b/gcc/config/rs6000/genfusion.pl
index 3a603eb09675..56e5d96ec5f3 100755
--- a/gcc/config/rs6000/genfusion.pl
+++ b/gcc/config/rs6000/genfusion.pl
@@ -248,6 +248,7 @@ sub gen_logical_addsubf
"vorc_vor" => 191,
"vandc_vnor" => 208,
"vandc_veqv" => 210,
+ "vand_vnor" => 224,
);
KIND: foreach $kind ('scalar','vector') {
