public inbox for gcc-patches@gcc.gnu.org
 help / color / mirror / Atom feed
* [PATCH] Fix middle-end/103645: empty struct store not removed when using compound literal
@ 2022-08-08  3:37 apinski
  2022-08-08  7:59 ` Richard Biener
  0 siblings, 1 reply; 2+ messages in thread
From: apinski @ 2022-08-08  3:37 UTC (permalink / raw)
  To: gcc-patches; +Cc: Andrew Pinski

From: Andrew Pinski <apinski@marvell.com>

For compound literals empty struct stores are not removed as they go down a
different path of the gimplifier; trying to optimize the init constructor.
This fixes the problem by not adding the gimple assignment at the end
of gimplify_init_constructor if it was an empty type.

Note this updates gcc.dg/pr87052.c where we had:
const char d[0] = { };
And was expecting a store to d but after this, there is no store
as the decl's type is zero in size.

OK? Bootstrapped and tested on x86_64-linux-gnu with no regressions.

gcc/ChangeLog:

	PR middle-end/103645
	* gimplify.c (gimplify_init_constructor): Don't build/add
	gimple assignment of an empty type.

testsuite/ChangeLog:
	* gcc.dg/pr87052.c: Update d var to expect nothing.
---
 gcc/gimplify.cc                | 7 +++++--
 gcc/testsuite/gcc.dg/pr87052.c | 6 +++---
 2 files changed, 8 insertions(+), 5 deletions(-)

diff --git a/gcc/gimplify.cc b/gcc/gimplify.cc
index 2ac7ca0855e..f0fbdb48012 100644
--- a/gcc/gimplify.cc
+++ b/gcc/gimplify.cc
@@ -5488,8 +5488,11 @@ gimplify_init_constructor (tree *expr_p, gimple_seq *pre_p, gimple_seq *post_p,
   if (ret == GS_ERROR)
     return GS_ERROR;
   /* If we have gimplified both sides of the initializer but have
-     not emitted an assignment, do so now.  */
-  if (*expr_p)
+     not emitted an assignment, do so now.   */
+  if (*expr_p
+      /* If the type is an empty type, we don't need to emit the
+	 assignment. */
+      && !is_empty_type (TREE_TYPE (TREE_OPERAND (*expr_p, 0))))
     {
       tree lhs = TREE_OPERAND (*expr_p, 0);
       tree rhs = TREE_OPERAND (*expr_p, 1);
diff --git a/gcc/testsuite/gcc.dg/pr87052.c b/gcc/testsuite/gcc.dg/pr87052.c
index 18e092c4674..796fe6440c1 100644
--- a/gcc/testsuite/gcc.dg/pr87052.c
+++ b/gcc/testsuite/gcc.dg/pr87052.c
@@ -23,8 +23,7 @@ void test (void)
 
   const char d[0] = { };
 
-  /* Expect the following:
-     d = ""; */
+  /* Expect nothing.  */
 
   const char e[0] = "";
 
@@ -36,6 +35,7 @@ void test (void)
 /* { dg-final { scan-tree-dump-times "a = \"\\\\x00ab\";" 1 "gimple" } }
    { dg-final { scan-tree-dump-times "b = \"a\\\\x00bc\";"  1 "gimple" } }
    { dg-final { scan-tree-dump-times "c = \"\";"  1 "gimple" } }
-   { dg-final { scan-tree-dump-times "d = { *};"  1 "gimple" } }
+   { dg-final { scan-tree-dump-times "d = "  1 "gimple" } }
+   { dg-final { scan-tree-dump-times "d = {CLOBBER\\(eol\\)}"  1 "gimple" } }
    { dg-final { scan-tree-dump-times "e = "  1 "gimple" } }
    { dg-final { scan-tree-dump-times "e = {CLOBBER\\(eol\\)}"  1 "gimple" } }  */
-- 
2.27.0


^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2022-08-08  7:59 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2022-08-08  3:37 [PATCH] Fix middle-end/103645: empty struct store not removed when using compound literal apinski
2022-08-08  7:59 ` Richard Biener

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).