[Bug testsuite/50085] New: [4.7 Regression] FAIL: g++.dg/opt/life1.C scan-assembler GLOBAL_OFFSET_TABLE on x86_64-apple-darwin10

[Bug testsuite/50085] New: [4.7 Regression] FAIL: g++.dg/opt/life1.C scan-assembler GLOBAL_OFFSET_TABLE on x86_64-apple-darwin10

[dominiq at lps dot ens.fr]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50085

             Bug #: 50085
           Summary: [4.7 Regression] FAIL: g++.dg/opt/life1.C
                    scan-assembler GLOBAL_OFFSET_TABLE on
                    x86_64-apple-darwin10
    Classification: Unclassified
           Product: gcc
           Version: 4.7.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: testsuite
        AssignedTo: unassigned@gcc.gnu.org
        ReportedBy: dominiq@lps.ens.fr
                CC: hjl@gcc.gnu.org
              Host: x86_64-apple-darwin10
            Target: x86_64-apple-darwin10
             Build: x86_64-apple-darwin10


The change at revision 176090 changed g++.dg/opt/life1.C from "unsupported",
i.e., with 

// { dg-do compile { target i?86-*-linux* x86_64-*-linux* } }
// { dg-require-effective-target ilp32 }

to "unexpected failures", i.e., with

// { dg-do compile { target { { i?86-*-* x86_64-*-* } && ia32 } } }

[Bug testsuite/50085] [4.7 Regression] FAIL: g++.dg/opt/life1.C scan-assembler GLOBAL_OFFSET_TABLE on x86_64-apple-darwin10

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50085

H.J. Lu <hjl.tools at gmail dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |hjl.tools at gmail dot com

--- Comment #1 from H.J. Lu <hjl.tools at gmail dot com> 2011-08-15 13:29:46 UTC ---
Can you try this

---
diff --git a/gcc/testsuite/g++.dg/opt/life1.C
b/gcc/testsuite/g++.dg/opt/life1.C
index 42fd744..cacd365 100644
--- a/gcc/testsuite/g++.dg/opt/life1.C
+++ b/gcc/testsuite/g++.dg/opt/life1.C
@@ -1,7 +1,6 @@
 // This testcase did not set up the pic register on IA-32 due
 // to bug in calculate_global_regs_live EH edge handling.
-// { dg-do compile { target { { i?86-*-* x86_64-*-* } && ia32 } } }
-// { dg-require-effective-target fpic }
+// { dg-do compile { target { { i?86-*-* x86_64-*-* } && ia32 && fpic} } }
 // { dg-options "-O2 -fPIC" }

 struct A { };
---

[Bug testsuite/50085] [4.7 Regression] FAIL: g++.dg/opt/life1.C scan-assembler GLOBAL_OFFSET_TABLE on x86_64-apple-darwin10

[dominiq at lps dot ens.fr]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50085

--- Comment #2 from Dominique d'Humieres <dominiq at lps dot ens.fr> 2011-08-15 13:38:15 UTC ---
> Can you try this ...

With the change I get:

ERROR: g++.dg/opt/life1-c.C: syntax error in target selector "target  i?86-*-*
x86_64-*-*  && ia32 && fpic" for " dg-do 3 compile { target { { i?86-*-*
x86_64-*-* } && ia32  && fpic } } "

If I use

// { dg-do compile { target { { i?86-*-* x86_64-*-* } && { ia32  && fpic } } }
}

The test fails with -m32 (unsupported with -m64).

[Bug testsuite/50085] [4.7 Regression] FAIL: g++.dg/opt/life1.C scan-assembler GLOBAL_OFFSET_TABLE on x86_64-apple-darwin10

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50085

--- Comment #3 from H.J. Lu <hjl.tools at gmail dot com> 2011-08-15 13:42:59 UTC ---
(In reply to comment #2)
> > Can you try this ...
> 
> With the change I get:
> 
> ERROR: g++.dg/opt/life1-c.C: syntax error in target selector "target  i?86-*-*
> x86_64-*-*  && ia32 && fpic" for " dg-do 3 compile { target { { i?86-*-*
> x86_64-*-* } && ia32  && fpic } } "
> 
> If I use
> 
> // { dg-do compile { target { { i?86-*-* x86_64-*-* } && { ia32  && fpic } } }
> }
> 
> The test fails with -m32 (unsupported with -m64).

Did this test work with -m32 before?

[Bug testsuite/50085] [4.7 Regression] FAIL: g++.dg/opt/life1.C scan-assembler GLOBAL_OFFSET_TABLE on x86_64-apple-darwin10

[dominiq at lps dot ens.fr]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50085

--- Comment #4 from Dominique d'Humieres <dominiq at lps dot ens.fr> 2011-08-15 14:55:20 UTC ---
> Did this test work with -m32 before?

It was unsupported.

[Bug testsuite/50085] [4.7 Regression] FAIL: g++.dg/opt/life1.C scan-assembler GLOBAL_OFFSET_TABLE on x86_64-apple-darwin10

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50085

--- Comment #5 from H.J. Lu <hjl.tools at gmail dot com> 2011-08-15 15:01:51 UTC ---
(In reply to comment #4)
> > Did this test work with -m32 before?
> 
> It was unsupported.

It required PIC and ILP32 before.  Why wasn't supported?

[Bug testsuite/50085] [4.7 Regression] FAIL: g++.dg/opt/life1.C scan-assembler GLOBAL_OFFSET_TABLE on x86_64-apple-darwin10

[dominiq at lps dot ens.fr]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50085

--- Comment #6 from Dominique d'Humieres <dominiq at lps dot ens.fr> 2011-08-15 15:09:10 UTC ---
> It required PIC and ILP32 before.  Why wasn't supported?

Good question;-) but I don't know the answer!-(

[Bug testsuite/50085] [4.7 Regression] FAIL: g++.dg/opt/life1.C scan-assembler GLOBAL_OFFSET_TABLE on x86_64-apple-darwin10

[dominiq at lps dot ens.fr]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50085

Dominique d'Humieres <dominiq at lps dot ens.fr> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |iains at gcc dot gnu.org,
                   |                            |mrs at gcc dot gnu.org

--- Comment #7 from Dominique d'Humieres <dominiq at lps dot ens.fr> 2011-08-15 21:48:18 UTC ---
> > It required PIC and ILP32 before.  Why wasn't supported?
>
> Good question;-) but I don't know the answer!-(

And the answer is

-// { dg-do compile { target i?86-*-linux* x86_64-*-linux* } }
-// { dg-require-effective-target ilp32 }
+// { dg-do compile { target { { i?86-*-* x86_64-*-* } && ia32 } } }

Before the patch, the test was run on linux only. After it it is run on all the
{ i?86-*-* x86_64-*-* } platforms.

An alternative is

// { dg-final { scan-assembler "GLOBAL_OFFSET_TABLE" { xfail *-*-darwin* } } }

I don't know what is supposed to replace "GLOBAL_OFFSET_TABLE" on darwin.

[Bug testsuite/50085] [4.7 Regression] FAIL: g++.dg/opt/life1.C scan-assembler GLOBAL_OFFSET_TABLE on x86_64-apple-darwin10

[hjl at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50085

--- Comment #8 from hjl at gcc dot gnu.org <hjl at gcc dot gnu.org> 2011-08-15 22:04:09 UTC ---
Author: hjl
Date: Mon Aug 15 22:04:04 2011
New Revision: 177772

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=177772
Log:
Only run g++.dg/opt/life1.C on Linux.

2011-08-15  H.J. Lu  <hongjiu.lu@intel.com>

    PR testsuite/50085
    * g++.dg/opt/life1.C: Only run on Linux.

Modified:
    trunk/gcc/testsuite/ChangeLog
    trunk/gcc/testsuite/g++.dg/opt/life1.C

[Bug testsuite/50085] [4.7 Regression] FAIL: g++.dg/opt/life1.C scan-assembler GLOBAL_OFFSET_TABLE on x86_64-apple-darwin10

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50085

H.J. Lu <hjl.tools at gmail dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |RESOLVED
         Resolution|                            |FIXED
   Target Milestone|---                         |4.7.0

--- Comment #9 from H.J. Lu <hjl.tools at gmail dot com> 2011-08-15 22:06:54 UTC ---
Fixed.