public inbox for gcc-patches@gcc.gnu.org
 help / color / mirror / Atom feed
From: Martin Jambor <mjambor@suse.cz>
To: GCC Patches <gcc-patches@gcc.gnu.org>
Cc: Jan Hubicka <jh@suse.cz>, Jan Hubicka <hubicka@ucw.cz>
Subject: [PATCH] ipa-sra: Consider the first parameter of methods safe to dereference
Date: Wed, 14 Dec 2022 16:18:58 +0100	[thread overview]
Message-ID: <ri6o7s6rp25.fsf@suse.cz> (raw)

Hi,

Honza requested this after reviewing the patch that taught IPA-SRA
that REFERENCE_TYPEs are always non-NULL that the pass also handles
the first parameters of methods, this pointers, in the same way.  So
this patch does that.

The patch is undergoing bootstrap and testing on an x86_64-linux right
now.  OK if it passes?

Thanks,

Martin


gcc/ChangeLog:

2022-12-14  Martin Jambor  <mjambor@suse.cz>

	* ipa-sra.cc (create_parameter_descriptors): Consider the first
	parameter of a method safe to dereference.

gcc/testsuite/ChangeLog:

2022-12-14  Martin Jambor  <mjambor@suse.cz>

	* g++.dg/ipa/ipa-sra-6.C: New test.
---
 gcc/ipa-sra.cc                       |  7 +++-
 gcc/testsuite/g++.dg/ipa/ipa-sra-6.C | 62 ++++++++++++++++++++++++++++
 2 files changed, 68 insertions(+), 1 deletion(-)
 create mode 100644 gcc/testsuite/g++.dg/ipa/ipa-sra-6.C

diff --git a/gcc/ipa-sra.cc b/gcc/ipa-sra.cc
index bcabdedfc6c..6fe336eeb19 100644
--- a/gcc/ipa-sra.cc
+++ b/gcc/ipa-sra.cc
@@ -1206,7 +1206,12 @@ create_parameter_descriptors (cgraph_node *node,
       if (POINTER_TYPE_P (type))
 	{
 	  desc->by_ref = true;
-	  desc->safe_ref = (TREE_CODE (type) == REFERENCE_TYPE);
+	  if (TREE_CODE (type) == REFERENCE_TYPE
+	      || (num == 0
+		  && TREE_CODE (TREE_TYPE (node->decl)) == METHOD_TYPE))
+	    desc->safe_ref = true;
+	  else
+	    desc->safe_ref = false;
 	  type = TREE_TYPE (type);
 
 	  if (TREE_CODE (type) == FUNCTION_TYPE
diff --git a/gcc/testsuite/g++.dg/ipa/ipa-sra-6.C b/gcc/testsuite/g++.dg/ipa/ipa-sra-6.C
new file mode 100644
index 00000000000..d6b7822533f
--- /dev/null
+++ b/gcc/testsuite/g++.dg/ipa/ipa-sra-6.C
@@ -0,0 +1,62 @@
+/* { dg-do compile } */
+/* { dg-options "-O2 -fdump-ipa-sra"  } */
+
+namespace {
+
+class C
+{
+
+  int mi;
+
+public:
+  C (int i)
+    : mi(i)
+  {}
+
+  void foo (int c);
+};
+
+volatile int vi;
+
+
+void __attribute__((noinline))
+C::foo (int cond)
+{
+  int i;
+  if (cond)
+    i = mi;
+  else
+    i = 0;
+  vi = i;
+}
+
+static C c_instance(1);
+}
+
+void __attribute__((noinline))
+bar (C *p, int cond)
+{
+  p->foo (cond);
+}
+
+
+class C *gp;
+
+void something(void);
+
+void
+baz (int cond)
+{
+  C c(vi);
+  gp = &c;
+  something ();
+  bar (gp, cond);
+}
+
+void
+hoo(void)
+{
+  gp = &c_instance;
+}
+
+/* { dg-final { scan-ipa-dump "Will split parameter" "sra" } } */
-- 
2.38.1


             reply	other threads:[~2022-12-14 15:19 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-12-14 15:18 Martin Jambor [this message]
2022-12-14 15:20 ` Jan Hubicka
2022-12-15  8:20   ` Richard Biener
2022-12-15 17:52     ` Jan Hubicka

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=ri6o7s6rp25.fsf@suse.cz \
    --to=mjambor@suse.cz \
    --cc=gcc-patches@gcc.gnu.org \
    --cc=hubicka@ucw.cz \
    --cc=jh@suse.cz \
    /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).