Fix for PR14824: Optimization arm_ldst_opt inserts newly generated instruction vldmia at incorrect position

[Stepan Dyatkovskiy]

Oh.. sorry, guys. I had pretty old test-suite version, the modified x86 
one. Renato, thanks, and sorry again.
-Stepan

Renato Golin wrote:
> Hi Stepan,
>
> Your patch broke a lot of tests on the ARM test-suite:
>
> http://lab.llvm.org:8011/builders/clang-native-arm-lnt/builds/442
>
> And the fix didn't fix all of them:
>
> http://lab.llvm.org:8011/builders/clang-native-arm-lnt/builds/443
>
> CoyoteBench was passing before you committed:
>
> http://lab.llvm.org:8011/builders/clang-native-arm-lnt/builds/441
>
> I'm surprised that you managed to run the test-suite both times and not
> see those failures...
>
> If you can't fix the broken test, I suggest we revert your patch and
> make sure you can run the test-suite and see the same errors as the one
> in the build, and make sure all the tests pass before committing. I
> followed the instructions from the website and it behaves exactly like
> the buildbot:
>
> http://llvm.org/docs/lnt/quickstart.html
>
> cheers,
> --renato
>
>
>
>
> On 5 April 2013 07:47, Stepan Dyatkovskiy <stpworld at narod.ru
> <mailto:stpworld at narod.ru>> wrote:
>
>     Got failure on clang-native-arm-cortex-a9. Fixing...
>     -Stepan
>
>     Stepan Dyatkovskiy wrote:
>
>         Committed as r178851. The only regression was added. Difficult
>         to guess
>         code that reproduces this bug.
>         -Stepan
>
>         Stepan Dyatkovskiy wrote:
>
>             OK. I'll do. Currently I got some little trouble with
>             test/CodeGen/ARM/stm.ll. Hope fix it soon.
>
>             -Stepan
>
>             Renato Golin wrote:
>
>                 On 3 April 2013 06:29, Stepan Dyatkovskiy
>                 <stpworld at narod.ru <mailto:stpworld at narod.ru>
>                 <mailto:stpworld at narod.ru <mailto:stpworld at narod.ru>>>
>                 wrote:
>
>                      Hello Renato,
>                      Thanks. I'll run test-suite today, if everything
>                 would fine, I
>                      commit it then.
>
>
>                 Can you also add a few tests, just to make sure it won't
>                 regress in the
>                 future?
>
>                 Thanks!
>                 --renato
>
>
>             _________________________________________________
>             llvm-commits mailing list
>             llvm-commits at cs.uiuc.edu <mailto:llvm-commits at cs.uiuc.edu>
>             http://lists.cs.uiuc.edu/__mailman/listinfo/llvm-commits
>             <http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits>
>
>
>         _________________________________________________
>         llvm-commits mailing list
>         llvm-commits at cs.uiuc.edu <mailto:llvm-commits at cs.uiuc.edu>
>         http://lists.cs.uiuc.edu/__mailman/listinfo/llvm-commits
>         <http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits>
>
>
>