public inbox for gcc-patches@gcc.gnu.org
 help / color / mirror / Atom feed
* [committed] i386: Fix up *avx2_eq<mode>3 constraints [PR114783]
@ 2024-04-19 22:22 Jakub Jelinek
  0 siblings, 0 replies; only message in thread
From: Jakub Jelinek @ 2024-04-19 22:22 UTC (permalink / raw)
  To: Hongtao Liu, Kong Lingling; +Cc: gcc-patches

Hi!

The r14-4456 change (part of APX EGPR support) seems to have mistakenly
changed in the
@@ -16831,7 +16831,7 @@ (define_insn "*avx2_eq<mode>3"
   [(set (match_operand:VI_256 0 "register_operand" "=x")
        (eq:VI_256
          (match_operand:VI_256 1 "nonimmediate_operand" "%x")
-         (match_operand:VI_256 2 "nonimmediate_operand" "xm")))]
+         (match_operand:VI_256 2 "nonimmediate_operand" "jm")))]
   "TARGET_AVX2 && !(MEM_P (operands[1]) && MEM_P (operands[2]))"
   "vpcmpeq<ssemodesuffix>\t{%2, %1, %0|%0, %1, %2}"
   [(set_attr "type" "ssecmp")
hunk the xm constraint to jm, while in many other spots it changed correctly
xm to xjm.  The instruction doesn't require the last operand to be in
memory, it can handle 3 256-bit registers just fine, just it is a VEX only
encoded instruction and so can't allow APX EGPR regs in the memory operand.

The following patch fixes it, so that we don't force one of the == operands
into memory all the time.

Bootstrapped/regtested on x86_64-linux and i686-linux, committed to trunk as
obvious.

2024-04-20  Jakub Jelinek  <jakub@redhat.com>

	PR target/114783
	* config/i386/sse.md (*avx2_eq<mode>3): Change last operand's
	constraint from "jm" to "xjm".

	* gcc.target/i386/avx2-pr114783.c: New test.

--- gcc/config/i386/sse.md.jj	2024-04-10 09:55:05.849877708 +0200
+++ gcc/config/i386/sse.md	2024-04-19 19:37:34.110320387 +0200
@@ -17029,7 +17029,7 @@ (define_insn "*avx2_eq<mode>3"
   [(set (match_operand:VI_256 0 "register_operand" "=x")
 	(eq:VI_256
 	  (match_operand:VI_256 1 "nonimmediate_operand" "%x")
-	  (match_operand:VI_256 2 "nonimmediate_operand" "jm")))]
+	  (match_operand:VI_256 2 "nonimmediate_operand" "xjm")))]
   "TARGET_AVX2 && !(MEM_P (operands[1]) && MEM_P (operands[2]))"
   "vpcmpeq<ssemodesuffix>\t{%2, %1, %0|%0, %1, %2}"
   [(set_attr "type" "ssecmp")
--- gcc/testsuite/gcc.target/i386/avx2-pr114783.c.jj	2024-04-19 19:28:50.216644029 +0200
+++ gcc/testsuite/gcc.target/i386/avx2-pr114783.c	2024-04-19 19:45:02.943054262 +0200
@@ -0,0 +1,12 @@
+/* PR target/114783 */
+/* { dg-do compile } */
+/* { dg-options "-O2 -mavx2 -mno-avx512f -masm=att" } */
+/* { dg-final { scan-assembler "vpcmpeqd\[ \\t\]+%ymm\[01\], %ymm\[01\], %ymm0" } } */
+
+typedef int V __attribute__((vector_size (32)));
+
+V
+foo (V x, V y)
+{
+  return x == y;
+}

	Jakub


^ permalink raw reply	[flat|nested] only message in thread

only message in thread, other threads:[~2024-04-19 22:22 UTC | newest]

Thread overview: (only message) (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2024-04-19 22:22 [committed] i386: Fix up *avx2_eq<mode>3 constraints [PR114783] Jakub Jelinek

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).