public inbox for gcc-patches@gcc.gnu.org
 help / color / mirror / Atom feed
From: Richard Biener <rguenther@suse.de>
To: gcc-patches@gcc.gnu.org
Subject: [PATCH] tree-optimization/107935 - fixup equivalence handling in PHI VN
Date: Thu, 1 Dec 2022 10:18:22 +0100 (CET)	[thread overview]
Message-ID: <20221201091823.3D45F13B4A@imap2.suse-dmz.suse.de> (raw)

The following makes sure to honor the backedge processing logic
that forces VARYING there.

Bootstrapped and tested on x86_64-unknown-linux-gnu, pushed.

	PR tree-optimization/107935
	* tree-ssa-sccvn.cc (visit_phi): Honor forced VARYING on
	backedges.

	* gcc.dg/torture/pr107935.c: New testcase.
---
 gcc/testsuite/gcc.dg/torture/pr107935.c | 18 ++++++++++++++++++
 gcc/tree-ssa-sccvn.cc                   |  7 ++++++-
 2 files changed, 24 insertions(+), 1 deletion(-)
 create mode 100644 gcc/testsuite/gcc.dg/torture/pr107935.c

diff --git a/gcc/testsuite/gcc.dg/torture/pr107935.c b/gcc/testsuite/gcc.dg/torture/pr107935.c
new file mode 100644
index 00000000000..78175100f80
--- /dev/null
+++ b/gcc/testsuite/gcc.dg/torture/pr107935.c
@@ -0,0 +1,18 @@
+/* { dg-do compile } */
+
+int *a, **b;
+int main() {
+  int d = 0, *e = &d;
+ L:
+  *e = d;
+  if (a) {
+    int *g = e = *b;
+    if (!e)
+      __builtin_abort();
+    if (**b)
+      return 0;
+    *g = 1;
+    goto L;
+  }
+  return 0;
+}
diff --git a/gcc/tree-ssa-sccvn.cc b/gcc/tree-ssa-sccvn.cc
index 6895ae84d13..fa2f65df159 100644
--- a/gcc/tree-ssa-sccvn.cc
+++ b/gcc/tree-ssa-sccvn.cc
@@ -5861,7 +5861,12 @@ visit_phi (gimple *phi, bool *inserted, bool backedges_varying_p)
 		  continue;
 	      }
 	    /* There's also the possibility to use equivalences.  */
-	    if (!FLOAT_TYPE_P (TREE_TYPE (def)))
+	    if (!FLOAT_TYPE_P (TREE_TYPE (def))
+		/* But only do this if we didn't force any of sameval or
+		   val to VARYING because of backedge processing rules.  */
+		&& (TREE_CODE (sameval) != SSA_NAME
+		    || SSA_VAL (sameval) == sameval)
+		&& (TREE_CODE (def) != SSA_NAME || SSA_VAL (def) == def))
 	      {
 		vn_nary_op_t vnresult;
 		tree ops[2];
-- 
2.35.3

                 reply	other threads:[~2022-12-01  9:18 UTC|newest]

Thread overview: [no followups] expand[flat|nested]  mbox.gz  Atom feed

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=20221201091823.3D45F13B4A@imap2.suse-dmz.suse.de \
    --to=rguenther@suse.de \
    --cc=gcc-patches@gcc.gnu.org \
    /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).