From: Uros Bizjak <ubizjak@gmail.com>
To: "gcc-patches@gcc.gnu.org" <gcc-patches@gcc.gnu.org>
Cc: Richard Sandiford <richard.sandiford@arm.com>,
Richard Biener <rguenther@suse.de>
Subject: [PATCH] simplify-rtx: Fix VOIDmode operand handling in simplify_subreg [PR108805]
Date: Thu, 16 Feb 2023 18:39:16 +0100 [thread overview]
Message-ID: <CAFULd4b_k7Az5ycESE-_7nqJF4y8s4KLbdYx3Y=LbCVP+q_xfA@mail.gmail.com> (raw)
[-- Attachment #1: Type: text/plain, Size: 758 bytes --]
simplify_subreg can return VOIDmode const_int operand and will
cause ICE in simplify_gen_subreg when this operand is passed to it.
The patch prevents VOIDmode temporary from entering simplify_gen_subreg.
We can't process const_int operand any further, since outermode
is not an integer mode here.
2023-02-16 Uroš Bizjak <ubizjak@gmail.com>
gcc/ChangeLog:
PR target/108805
* simplify_rtx.cc (simplify_context::simplify_subreg): Prevent
VOIDmode const_int result from simplify_subreg from entering
simplify_gen_subreg.
gcc/testsuite/ChangeLog:
PR target/108805
* gcc.dg/pr108805.c: New test.
Bootstrapped and regression tested on x86_64-linux-gnu {,-m32}.
OK for master and release branches?
Uros.
[-- Attachment #2: p.diff.txt --]
[-- Type: text/plain, Size: 1174 bytes --]
diff --git a/gcc/simplify-rtx.cc b/gcc/simplify-rtx.cc
index 0a1dd88b0a8..281bc418df0 100644
--- a/gcc/simplify-rtx.cc
+++ b/gcc/simplify-rtx.cc
@@ -7664,7 +7664,7 @@ simplify_context::simplify_subreg (machine_mode outermode, rtx op,
0).exists (&int_outermode))
{
rtx tem = simplify_subreg (int_outermode, op, innermode, byte);
- if (tem)
+ if (tem && GET_MODE (tem) != VOIDmode)
return simplify_gen_subreg (outermode, tem, GET_MODE (tem), 0);
}
diff --git a/gcc/testsuite/gcc.dg/pr108805.c b/gcc/testsuite/gcc.dg/pr108805.c
new file mode 100644
index 00000000000..280d3f5c377
--- /dev/null
+++ b/gcc/testsuite/gcc.dg/pr108805.c
@@ -0,0 +1,20 @@
+/* { dg-do compile { target longlong64 } } */
+/* { dg-options "-O" } */
+/* { dg-additional-options "-msse2" { target x86_64-*-* i?86-*-* } } */
+
+typedef __INT8_TYPE__ __attribute__((__vector_size__ (4))) U;
+typedef __INT32_TYPE__ __attribute__((__vector_size__ (4))) V;
+typedef __UINT64_TYPE__ __attribute__((__vector_size__ (8))) W;
+
+int i;
+U h;
+W g;
+
+U
+foo (void)
+{
+ W w = i != g;
+ V v = __builtin_convertvector (i | w >> 2, V);
+ U u = (U) v[0] + h;
+ return u;
+}
next reply other threads:[~2023-02-16 17:39 UTC|newest]
Thread overview: 9+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-02-16 17:39 Uros Bizjak [this message]
2023-02-17 7:38 ` Richard Biener
2023-02-17 8:51 ` Uros Bizjak
2023-02-17 11:31 ` Richard Biener
2023-02-17 15:05 ` Uros Bizjak
2023-02-27 15:46 ` Richard Sandiford
2023-03-02 10:08 ` Andre Vieira (lists)
2023-03-02 10:13 ` Richard Sandiford
2023-03-02 14:15 ` Andre Vieira (lists)
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to='CAFULd4b_k7Az5ycESE-_7nqJF4y8s4KLbdYx3Y=LbCVP+q_xfA@mail.gmail.com' \
--to=ubizjak@gmail.com \
--cc=gcc-patches@gcc.gnu.org \
--cc=rguenther@suse.de \
--cc=richard.sandiford@arm.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
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).