| Message ID | 20100625021739.GA24512@bromo.med.uc.edu |
|---|---|
| State | New |
| Headers | show |
On Jun 24, 2010, at 7:17 PM, Jack Howarth wrote: > The attached patch backports r159700, r158929, r159837 and > r159924 from trunk to gcc 4.5 branch which eliminate several > lto and alias related testsuite failures on darwin. Tested on > x86_64-apple-darwin10. Okay for gcc 4.5.1? Ok.
On 25 Jun 2010, at 06:14, Mike Stump wrote: > On Jun 24, 2010, at 7:17 PM, Jack Howarth wrote: >> The attached patch backports r159700, r158929, r159837 and >> r159924 from trunk to gcc 4.5 branch which eliminate several >> lto and alias related testsuite failures on darwin. Tested on >> x86_64-apple-darwin10. Okay for gcc 4.5.1? > > Ok. done, r162054 Iain
Index: gcc/testsuite/gcc.dg/lto/20091216-1_0.c =================================================================== --- gcc/testsuite/gcc.dg/lto/20091216-1_0.c (revision 161355) +++ gcc/testsuite/gcc.dg/lto/20091216-1_0.c (working copy) @@ -1,6 +1,6 @@ /* { dg-lto-do run } */ -asm (".globl start; start: nop"); +asm (".globl start_; start_: nop"); int main () Index: gcc/testsuite/gcc.dg/lto/20081222_0.c =================================================================== --- gcc/testsuite/gcc.dg/lto/20081222_0.c (revision 161355) +++ gcc/testsuite/gcc.dg/lto/20081222_0.c (working copy) @@ -1,3 +1,4 @@ +/* { dg-require-alias "" } */ #include "20081222_0.h" extern void abort (void); Index: gcc/testsuite/gcc.dg/lto/20090914-2_0.c =================================================================== --- gcc/testsuite/gcc.dg/lto/20090914-2_0.c (revision 161355) +++ gcc/testsuite/gcc.dg/lto/20090914-2_0.c (working copy) @@ -1,5 +1,6 @@ /* { dg-lto-do run } */ /* { dg-skip-if "x86 only" { ! { x86_64-*-* i?86-*-* } } { "*" } { "" } } */ +/* { dg-skip-if "no .type" { *-*-darwin* } { "*" } { "" } } */ /* Doesn't work without this dummy function with -fwhopr. */ int foo(void) { } Index: gcc/testsuite/g++.dg/lto/20100302_0.C =================================================================== --- gcc/testsuite/g++.dg/lto/20100302_0.C (revision 161355) +++ gcc/testsuite/g++.dg/lto/20100302_0.C (working copy) @@ -1,5 +1,5 @@ // Test for ABI forward-compatibility aliases with LTO. -// { dg-skip-if "" { ! { i?86-*-* x86_64-*-* } } { "*" } { "" } } +// { dg-skip-if "" { { ! { i?86-*-* x86_64-*-* } } || { *-*-darwin* } } { "*" } { "" } } // { dg-lto-options {"-flto -fabi-version=2"} } #include "20100302.h" Index: gcc/testsuite/g++.dg/abi/mangle40.C =================================================================== --- gcc/testsuite/g++.dg/abi/mangle40.C (revision 161355) +++ gcc/testsuite/g++.dg/abi/mangle40.C (working copy) @@ -1,7 +1,7 @@ // PR c++/12909 // { dg-do compile { target i?86-*-* x86_64-*-* } } // { dg-require-weak } -// { dg-require-alias } +// { dg-require-alias "" } // { dg-options "-mavx -Wabi -fabi-version=2" } // { dg-final { scan-assembler "weak\[^\n\]*_Z1fIDv4_fEvT_" } } // { dg-final { scan-assembler "weak\[^\n\]*_Z1fIU8__vectorfEvT_" } }
The attached patch backports r159700, r158929, r159837 and r159924 from trunk to gcc 4.5 branch which eliminate several lto and alias related testsuite failures on darwin. Tested on x86_64-apple-darwin10. Okay for gcc 4.5.1? Jack 2010-06-24 Jack Howarth <howarth@bromo.med.uc.edu> Backport from mainline 2010-05-27 Jack Howarth <howarth@bromo.med.uc.edu> * g++.dg/lto/20100302_0.C: Skip on darwin. 2010-06-24 Jack Howarth <howarth@bromo.med.uc.edu> Backport from mainline 2009-04-30 Steven Bosscher <steven@gcc.gnu.org> * gcc.dg/lto/20091216-1_0.c: Adjust test case to avoid conflict with darwin crt1.o 'start' symbol. 2010-06-24 Jack Howarth <howarth@bromo.med.uc.edu> Backport from mainline 2010-05-22 Iain Sandoe <iains@gcc.gnu.org> PR lto/44238 * gcc.dg/lto/20090914-2_0.c: XFAIL *-*-darwin*. 2010-06-24 Jack Howarth <howarth@bromo.med.uc.edu> Backport from mainline 2010-05-25 Jack Howarth <howarth@bromo.med.uc.edu> Iain Sandoe <iains@gcc.gnu.org> * g++.dg/abi/mangle40.C: Adjust dg-require-alias syntax. * gcc.dg/lto/20081222_0.c: Require alias support.