C++ PATCH for c++/44282 (ia32 calling convention attributes and mangling)

Uros Bizjak ubizjak@gmail.com
Tue Jun 2 11:52:00 GMT 2015


On Tue, Jun 2, 2015 at 1:03 PM, Uros Bizjak <ubizjak@gmail.com> wrote:
> On Tue, Jun 2, 2015 at 9:26 AM, Uros Bizjak <ubizjak@gmail.com> wrote:
>> Hello!
>>
>>>    PR c++/44282
>>>    gcc/cp/
>>>    * mangle.c (attr_strcmp): New.
>>>    (write_CV_qualifiers_for_type): Also write out attributes that
>>>    affect type identity.
>>>    (write_type): Strip all attributes after writing qualifiers.
>>>    libiberty/
>>>    * cp-demangle.c (cplus_demangle_type): Handle arguments to vendor
>>>    extended qualifier.
>>
>> +++ b/gcc/testsuite/g++.dg/abi/mangle-regparm.C
>> @@ -0,0 +1,29 @@
>> +// { dg-do run { target i?86-*-* } }
>>
>> This should read:
>>
>> +// { dg-do run { target { { i?86-*-* x86_64-*-* } && ia32 } } }
>>
>> The test wasn't actually run on x86_64-linux target. I'll commit the
>> following patch after regtest:
>
> Unfortunately, something is wrong with the testcase itself:
>
> FAIL: g++.dg/abi/mangle-regparm.C  -std=c++98 (test for excess errors)
> WARNING: g++.dg/abi/mangle-regparm.C  -std=c++98 compilation failed to
> produce executable
> FAIL: g++.dg/abi/mangle-regparm.C  -std=c++11 (test for excess errors)
> WARNING: g++.dg/abi/mangle-regparm.C  -std=c++11 compilation failed to
> produce executable
> FAIL: g++.dg/abi/mangle-regparm.C  -std=c++14 (test for excess errors)
> WARNING: g++.dg/abi/mangle-regparm.C  -std=c++14 compilation failed to
> produce executable
>
> with the following error:
>
> FAIL: g++.dg/abi/mangle-regparm.C  -std=c++98 (test for excess errors)
> Excess errors:
> /usr/bin/ld: /tmp/ccU8LttY.o: bad reloc symbol index (0x5b550 >= 0x12)
> for offset 0x6c in section `.text'
> /tmp/ccU8LttY.o: could not read symbols: Bad value
>
> WARNING: g++.dg/abi/mangle-regparm.C  -std=c++98 compilation failed to
> produce executable
> g++.dg/abi/mangle-regparm.C  -std=c++98 : output file does not exist
> UNRESOLVED: g++.dg/abi/mangle-regparm.C  -std=c++98  scan-assembler
> _Z18IndirectExternCallIPU7stdcallU7regparmILi3EEFviiEiEvT_T0_S3_

FYI, everything links and runs OK if .set is removed from the
following part of the asm:

.LFE4:
       .size   _Z18IndirectExternCallIPU7stdcallU7regparmILi3EEFviiEiEvT_T0_S3_,
.-_Z18IndirectExternCallIPU7stdcallU7regparmILi3EEFviiEiEvT_T0_S3_
       .weak   _Z18IndirectExternCallIPFviiEiEvT_T0_S3_
       .set    _Z18IndirectExternCallIPFviiEiEvT_T0_S3_,_Z18IndirectExternCallIPU7stdcallU7regparmILi3EEFviiEiEvT_T0_S3_
       .section
.text._Z18IndirectExternCallIPFviiEiEvT_T0_S3_,"axG",@progbits,_Z18IndirectExternCallIPFviiEiEvT_T0_S3_,comdat
       .weak   _Z18IndirectExternCallIPFviiEiEvT_T0_S3_
       .type   _Z18IndirectExternCallIPFviiEiEvT_T0_S3_, @function
_Z18IndirectExternCallIPFviiEiEvT_T0_S3_:

Uros.



More information about the Gcc-patches mailing list