1.1 --- a/patches/gcc/3.3.3/pr11162-1-test.patch Tue Aug 14 19:32:22 2007 +0000
1.2 +++ /dev/null Thu Jan 01 00:00:00 1970 +0000
1.3 @@ -1,25 +0,0 @@
1.4 ---- /dev/null Sat Dec 14 13:56:51 2002
1.5 -+++ gcc-3.3.1/gcc/testsuite/gcc.dg/pr11162-1.c Sun Sep 14 16:24:41 2003
1.6 -@@ -0,0 +1,22 @@
1.7 -+/*
1.8 -+ * PR optimization/11162
1.9 -+ * Reporter: Toshiyasu Morita <tm@kloo.net>
1.10 -+ * Summary: [3.4 Regression] [-fnew-ra] ICE compiling channel.i on sh
1.11 -+ * Keywords: ice-on-valid-code
1.12 -+ * Description:
1.13 -+ * [ICE] when channel.i from stress-1.17 is compiled with options:
1.14 -+ * -O2 -m4 -fnew-ra
1.15 -+ * ...
1.16 -+ * The CFG code gets confused by the addition of a USE insn after a sibcall.
1.17 -+ * Sibcalls are supposed to end the block (and the function!), so the fact
1.18 -+ * that we have an instruction following one (even if its fake) is confusing.
1.19 -+ *
1.20 -+ * testcase tweaked by dank@kegel.com
1.21 -+ */
1.22 -+/* { dg-do compile } */
1.23 -+/* { dg-options "-O2 -fnew-ra" } */
1.24 -+
1.25 -+int foo(char *p1)
1.26 -+{
1.27 -+ return bar(p1);
1.28 -+}