[Bug libffi/64799] New: [5 regression] libffi.special/unwindtest.cc FAILs on Solaris/SPARC

[Bug libffi/64799] New: [5 regression] libffi.special/unwindtest.cc FAILs on Solaris/SPARC

[ro at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64799

            Bug ID: 64799
           Summary: [5 regression] libffi.special/unwindtest.cc FAILs on
                    Solaris/SPARC
           Product: gcc
           Version: 5.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: libffi
          Assignee: unassigned at gcc dot gnu.org
          Reporter: ro at gcc dot gnu.org
                CC: rth at gcc dot gnu.org
              Host: *-*-solaris2.1[01]
            Target: *-*-solaris2.1[01]
             Build: *-*-solaris2.1[01]

Since the libffi merge, two testcases FAIL on Solaris/SPARC when using Sun as,
both for 32 and 64-bit:

FAIL: libffi.special/unwindtest.cc  -shared-libgcc -lstdc++ execution test
FAIL: libffi.special/unwindtest_ffi_call.cc  -shared-libgcc -lstdc++ execution
test

I'm also seeing a couple of libjava regressions that are almost certainly
related:

FAIL: noclass execution - gij test
FAIL: pr11951 run
FAIL: throwit execution - gij test
FAIL: pr29812 execution - gij test
FAIL: ExtraClassLoader execution - source compiled test
FAIL: ExtraClassLoader -findirect-dispatch execution - source compiled test
FAIL: ExtraClassLoader -O3 execution - source compiled test
FAIL: ExtraClassLoader -O3 -findirect-dispatch execution - source compiled test
FAIL: invokethrow execution - source compiled test
FAIL: invokethrow -findirect-dispatch execution - source compiled test
FAIL: invokethrow -O3 execution - source compiled test
FAIL: invokethrow -O3 -findirect-dispatch execution - source compiled test

I'm pretty sure that this happens because the merge lost the handcrafted
.eh_frame sections, relying on .cfi_* directives that the native Solaris/SPARC
assembler does not support.

  Rainer

[Bug libffi/64799] [5 regression] libffi.special/unwindtest.cc FAILs on Solaris/SPARC

[ro at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64799

Rainer Orth <ro at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|---                         |5.0

[Bug libffi/64799] [5 regression] libffi.special/unwindtest.cc FAILs on Solaris/SPARC

[ro at CeBiTec dot Uni-Bielefeld.DE]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64799

--- Comment #2 from ro at CeBiTec dot Uni-Bielefeld.DE <ro at CeBiTec dot Uni-Bielefeld.DE> ---
> --- Comment #1 from Richard Henderson <rth at gcc dot gnu.org> ---
> Created attachment 34583
>   --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=34583&action=edit
> proposed patch
>
> That does seem likely.  I tested this on linux; hopefully
> I got all of the macroization right for solaris.

Looks good so far: I've applied the patch, rebuilt libffi and run the
libffi testsuite (both multilibs).  I'll run a full bootstrap (both as
and gas configurations) for good measure and check if the libjava
failures are gone, too.

Thanks.
    Rainer

[Bug libffi/64799] [5 regression] libffi.special/unwindtest.cc FAILs on Solaris/SPARC

[ro at CeBiTec dot Uni-Bielefeld.DE]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64799

--- Comment #3 from ro at CeBiTec dot Uni-Bielefeld.DE <ro at CeBiTec dot Uni-Bielefeld.DE> ---
> --- Comment #2 from ro at CeBiTec dot Uni-Bielefeld.DE <ro at CeBiTec dot
> Uni-Bielefeld.DE> ---
[...]
> Looks good so far: I've applied the patch, rebuilt libffi and run the
> libffi testsuite (both multilibs).  I'll run a full bootstrap (both as
> and gas configurations) for good measure and check if the libjava
> failures are gone, too.

Those bootstraps have completed now.  The Solaris 11 and 32-bit Solaris
10 results look good, both libffi and libjava, but 64-bit Solaris 10
still FAILs:

FAIL: libffi.special/unwindtest.cc  -shared-libgcc -lstdc++ execution test
terminate called after throwing an instance of 'int'

Comparing sparcv9 fficonfig.h between S10 and S11, I see only one
difference:

--- fficonfig.h.s10     2015-01-28 00:45:48.714103199 +0100
+++ fficonfig.h.s11     2015-01-27 18:27:05.122136400 +0100
@@ -54,7 +54,7 @@

 /* Define if your assembler and linker support unaligned PC relative relocs.
    */
-/* #undef HAVE_AS_SPARC_UA_PCREL */
+#define HAVE_AS_SPARC_UA_PCREL 1

And checking the GCC 4.9 v9.S, I found that the attached change also
fixes those FAILures.  Tested for both 32 and 64-bit on S10 and S11.

    Rainer

[Bug libffi/64799] [5 regression] libffi.special/unwindtest.cc FAILs on Solaris/SPARC

[ro at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64799

--- Comment #4 from Rainer Orth <ro at gcc dot gnu.org> ---
Created attachment 34600
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=34600&action=edit
fix for 64-bit Solaris 10 failures

[Bug libffi/64799] [5 regression] libffi.special/unwindtest.cc FAILs on Solaris/SPARC

[rth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64799

--- Comment #5 from Richard Henderson <rth at gcc dot gnu.org> ---
Duh, that should have been obvious.
Thanks for the fix.  I'll push them
upstream at the same time.

[Bug libffi/64799] [5 regression] libffi.special/unwindtest.cc FAILs on Solaris/SPARC

[rth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64799

Richard Henderson <rth at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|ASSIGNED                    |RESOLVED
         Resolution|---                         |FIXED

--- Comment #6 from Richard Henderson <rth at gcc dot gnu.org> ---
Fixed in r220222.