[llvm] r209155 - Fix testers by removing dubious testcase for r209154.

Rafael EspĂ­ndola rafael.espindola at gmail.com
Wed May 21 14:02:28 PDT 2014


Can't you just require the host to be 86 on the test?

On 19 May 2014 16:53, Lang Hames <lhames at gmail.com> wrote:
> The RuntimeDyld layer doesn't have great testing infrastructure yet. I could
> write an MCJIT test that produced the right relocation on x86-64, but that
> doesn't really help much: It would run on all testers, but only test the
> desired behavior on x86-64. Moreover if there were any changes to CodeGen,
> MCJIT might end producing different relocations and so quietly cease testing
> what we wanted.
>
> The ideal way to test this would be to teach llvm-rtdyld how to dump the
> relocated sections. Then we'd need a utility to hoover up the interesting
> bytes (representing GOT entries) from a byte stream, reconstruct addresses,
> then run some expressions on them to verify that things are laid out as
> expected. That would be awesome, but it would (will?) also take some work.
>
> Despite the lack of tests, this change should stay in tree: the previous
> behavior was wrong, and reverting to known broken behavior isn't the answer
> to missing tests.
>
> - Lang.
>
>
>
> On Mon, May 19, 2014 at 12:51 PM, David Blaikie <dblaikie at gmail.com> wrote:
>>
>> On Mon, May 19, 2014 at 12:38 PM, Lang Hames <lhames at gmail.com> wrote:
>> > Author: lhames
>> > Date: Mon May 19 14:38:48 2014
>> > New Revision: 209155
>> >
>> > URL: http://llvm.org/viewvc/llvm-project?rev=209155&view=rev
>> > Log:
>> > Fix testers by removing dubious testcase for r209154.
>> >
>> > It turns out that not all the world is x86-64. Who knew?
>> >
>> > I'll get to work on a more appropriate test case for this patch.
>>
>> Reverting/removing test cases while leaving functionality untested
>> in-tree isn't usually ideal.
>>
>> I /think/ we usually just revert the whole patch, fix the tests, then
>> recommit. (also makes it a bit easier to understand when spelunking
>> back through the history later - the test goes with the code that it
>> verifies)
>>
>> >
>> >
>> > Removed:
>> >
>> > llvm/trunk/test/ExecutionEngine/RuntimeDyld/Inputs/x86_64_macho_pic_globals_GOT.o
>> > Modified:
>> >     llvm/trunk/test/ExecutionEngine/RuntimeDyld/macho_relocations.test
>> >
>> > Removed:
>> > llvm/trunk/test/ExecutionEngine/RuntimeDyld/Inputs/x86_64_macho_pic_globals_GOT.o
>> > URL:
>> > http://llvm.org/viewvc/llvm-project/llvm/trunk/test/ExecutionEngine/RuntimeDyld/Inputs/x86_64_macho_pic_globals_GOT.o?rev=209154&view=auto
>> >
>> > ==============================================================================
>> > Binary file - no diff available.
>> >
>> > Modified:
>> > llvm/trunk/test/ExecutionEngine/RuntimeDyld/macho_relocations.test
>> > URL:
>> > http://llvm.org/viewvc/llvm-project/llvm/trunk/test/ExecutionEngine/RuntimeDyld/macho_relocations.test?rev=209155&r1=209154&r2=209155&view=diff
>> >
>> > ==============================================================================
>> > --- llvm/trunk/test/ExecutionEngine/RuntimeDyld/macho_relocations.test
>> > (original)
>> > +++ llvm/trunk/test/ExecutionEngine/RuntimeDyld/macho_relocations.test
>> > Mon May 19 14:38:48 2014
>> > @@ -1,2 +1 @@
>> >  RUN: llvm-rtdyld -printline %p/Inputs/arm_secdiff_reloc.o
>> > -RUN: llvm-rtdyld %p/Inputs/x86_64_macho_pic_globals_GOT.o
>> >
>> >
>> > _______________________________________________
>> > llvm-commits mailing list
>> > llvm-commits at cs.uiuc.edu
>> > http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits
>
>
>
> _______________________________________________
> llvm-commits mailing list
> llvm-commits at cs.uiuc.edu
> http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits
>



More information about the llvm-commits mailing list