[llvm] r295765 - Don't modify archive members unless really needed.

[Rui Ueyama via llvm-commits]

No worries, I already fixed it.

On Wed, Feb 22, 2017 at 10:36 AM, Rafael Avila de Espindola <
rafael.espindola at gmail.com> wrote:

> Oops. I forgot to put the unreachable after the switch. I will fix that
> when I get back from lunch.
>
> Cheers,
> Rafael
>
>
> On February 22, 2017 12:53:10 PM EST, Rui Ueyama <ruiu at google.com> wrote:
>>
>> Huh, that's right. Maybe it was an MSVC thing that it cannot defer that
>> the end of function is unreachable.
>>
>> On Wed, Feb 22, 2017 at 6:17 AM, Rafael Avila de Espindola <
>> rafael.espindola at gmail.com> wrote:
>>
>>> Rui Ueyama <ruiu at google.com> writes:
>>>
>>> > On Tue, Feb 21, 2017 at 12:40 PM, Rafael Espindola via llvm-commits <
>>> > llvm-commits at lists.llvm.org> wrote:
>>> >
>>> >> Author: rafael
>>> >> Date: Tue Feb 21 14:40:54 2017
>>> >> New Revision: 295765
>>> >>
>>> >> URL: http://llvm.org/viewvc/llvm-project?rev=295765&view=rev
>>> >> Log:
>>> >> Don't modify archive members unless really needed.
>>> >>
>>> >> For whatever reason ld64 requires that member headers (not the member
>>> >> themselves) should be aligned. The only way to do that is to edit the
>>> >> previous member so that it ends at an aligned boundary.
>>> >>
>>> >> Since modifying data put in an archive is an undesirable property,
>>> >> llvm-ar should only do it when it is absolutely necessary.
>>> >>
>>> >> Added:
>>> >>     llvm/trunk/test/Object/archive-pad.test
>>> >> Modified:
>>> >>     llvm/trunk/include/llvm/Object/Archive.h
>>> >>     llvm/trunk/lib/Object/ArchiveWriter.cpp
>>> >>     llvm/trunk/test/Object/archive-extract.test
>>> >>     llvm/trunk/test/Object/archive-format.test
>>> >>     llvm/trunk/tools/llvm-ar/llvm-ar.cpp
>>> >>
>>> >> Modified: llvm/trunk/include/llvm/Object/Archive.h
>>> >> URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/include/
>>> >> llvm/Object/Archive.h?rev=295765&r1=295764&r2=295765&view=diff
>>> >> ============================================================
>>> >> ==================
>>> >> --- llvm/trunk/include/llvm/Object/Archive.h (original)
>>> >> +++ llvm/trunk/include/llvm/Object/Archive.h Tue Feb 21 14:40:54 2017
>>> >> @@ -212,6 +212,7 @@ public:
>>> >>      K_GNU,
>>> >>      K_MIPS64,
>>> >>      K_BSD,
>>> >> +    K_DARWIN,
>>> >>      K_DARWIN64,
>>> >>      K_COFF
>>> >>    };
>>> >>
>>> >> Modified: llvm/trunk/lib/Object/ArchiveWriter.cpp
>>> >> URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Object/
>>> >> ArchiveWriter.cpp?rev=295765&r1=295764&r2=295765&view=diff
>>> >> ============================================================
>>> >> ==================
>>> >> --- llvm/trunk/lib/Object/ArchiveWriter.cpp (original)
>>> >> +++ llvm/trunk/lib/Object/ArchiveWriter.cpp Tue Feb 21 14:40:54 2017
>>> >> @@ -122,12 +122,26 @@ static void printWithSpacePadding(raw_fd
>>> >>    }
>>> >>  }
>>> >>
>>> >> +static bool isBSDLike(object::Archive::Kind Kind) {
>>> >> +  switch (Kind) {
>>> >> +  case object::Archive::K_GNU:
>>> >> +    return false;
>>> >> +  case object::Archive::K_BSD:
>>> >> +  case object::Archive::K_DARWIN:
>>> >> +    return true;
>>> >> +  case object::Archive::K_MIPS64:
>>> >> +  case object::Archive::K_DARWIN64:
>>> >> +  case object::Archive::K_COFF:
>>> >> +    llvm_unreachable("not supported for writting");
>>> >> +  }
>>> >> +}
>>> >>
>>> >
>>> > This `switch` doesn't cover all Kind enums, which is bad because this
>>> > function should return a value.
>>>
>>> I am pretty sure it covers all the enum values. Which one do you think
>>> is missing?
>>>
>>> Cheers,
>>> Rafael
>>>
>>
>>
> --
> Sent from my Android device with K-9 Mail. Please excuse my brevity.
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20170222/6a610b59/attachment-0001.html>