public inbox for gcc-patches@gcc.gnu.org
 help / color / mirror / Atom feed
From: Siddhesh Poyarekar <siddhesh@gotplt.org>
To: gcc-patches@gcc.gnu.org
Cc: jakub@redhat.com
Subject: [PATCH] tree-object-size: Always set computed bit for bdos [PR113012]
Date: Mon, 18 Dec 2023 11:42:52 -0500	[thread overview]
Message-ID: <20231218164252.1963249-1-siddhesh@gotplt.org> (raw)

It is always safe to set the computed bit for dynamic object sizes at
the end of collect_object_sizes_for because even in case of a dependency
loop encountered in nested calls, we have an SSA temporary to actually
finish the object size expression.  The reexamine pass for dynamic
object sizes is only for propagation of unknowns and gimplification of
the size expressions, not for loop resolution as in the case of static
object sizes.

gcc/ChangeLog:

	PR tree-optimization/113012
	* gcc.dg/ubsan/pr113012.c: New test case.

gcc/testsuite/ChangeLog:

	PR tree-optimization/113012
	* tree-object-size.cc (compute_builtin_object_size): Expand
	comment for dynamic object sizes.
	(collect_object_sizes_for): Always set COMPUTED bitmap for
	dynamic object sizes.

Signed-off-by: Siddhesh Poyarekar <siddhesh@gotplt.org>
---
Testing:

- Bootstrapped x86_64 and config=ubsan
- Tested i686

OK for trunk and backport to gcc 13 branch?

 gcc/testsuite/gcc.dg/ubsan/pr113012.c | 17 +++++++++++++++++
 gcc/tree-object-size.cc               | 17 ++++++++++++-----
 2 files changed, 29 insertions(+), 5 deletions(-)
 create mode 100644 gcc/testsuite/gcc.dg/ubsan/pr113012.c

diff --git a/gcc/testsuite/gcc.dg/ubsan/pr113012.c b/gcc/testsuite/gcc.dg/ubsan/pr113012.c
new file mode 100644
index 00000000000..4fc38cd1171
--- /dev/null
+++ b/gcc/testsuite/gcc.dg/ubsan/pr113012.c
@@ -0,0 +1,17 @@
+/* { dg-do compile } */
+/* { dg-options "-fsanitize=undefined" } */
+
+int *
+foo (int x, int y, int z, int w)
+{
+  int *p = __builtin_malloc (z * sizeof (int));
+  int *q = p - 1;
+  while (--x > 0)
+    {
+      if (w + 1 > y)
+	q = p - 1;
+      ++*q;
+      ++q;
+    }
+  return p;
+}
diff --git a/gcc/tree-object-size.cc b/gcc/tree-object-size.cc
index 28f27adf9ca..434b2fc0bf5 100644
--- a/gcc/tree-object-size.cc
+++ b/gcc/tree-object-size.cc
@@ -1185,10 +1185,12 @@ compute_builtin_object_size (tree ptr, int object_size_type,
 	  osi.tos = NULL;
 	}
 
-      /* First pass: walk UD chains, compute object sizes that
-	 can be computed.  osi.reexamine bitmap at the end will
-	 contain what variables were found in dependency cycles
-	 and therefore need to be reexamined.  */
+      /* First pass: walk UD chains, compute object sizes that can be computed.
+	 osi.reexamine bitmap at the end will contain what variables that need
+	 to be reexamined.  For both static and dynamic size computation,
+	 reexamination is for propagation across dependency loops. The dynamic
+	 case has the additional use case where the computed expression needs
+	 to be gimplified.  */
       osi.pass = 0;
       osi.changed = false;
       collect_object_sizes_for (&osi, ptr);
@@ -1823,11 +1825,16 @@ collect_object_sizes_for (struct object_size_info *osi, tree var)
       gcc_unreachable ();
     }
 
-  if (! reexamine || object_sizes_unknown_p (object_size_type, varno))
+  /* Dynamic sizes use placeholder temps to return an answer, so it is always
+   * safe to set COMPUTED for them.  */
+  if ((object_size_type & OST_DYNAMIC)
+      || !reexamine || object_sizes_unknown_p (object_size_type, varno))
     {
       bitmap_set_bit (computed[object_size_type], varno);
       if (!(object_size_type & OST_DYNAMIC))
 	bitmap_clear_bit (osi->reexamine, varno);
+      else if (reexamine)
+	bitmap_set_bit (osi->reexamine, varno);
     }
   else
     {
-- 
2.43.0


             reply	other threads:[~2023-12-18 16:42 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-12-18 16:42 Siddhesh Poyarekar [this message]
2023-12-19 22:57 ` Jakub Jelinek
2023-12-19 23:23   ` Siddhesh Poyarekar

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=20231218164252.1963249-1-siddhesh@gotplt.org \
    --to=siddhesh@gotplt.org \
    --cc=gcc-patches@gcc.gnu.org \
    --cc=jakub@redhat.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).