[llvm-commits] Fix floating point numbers in tests on MinGW

[Dale Johannesen]

I'd rather see this fixed permanently by changing the code not to differ.  IIRC there's a PR with a suggested (rather ugly but functional) patch in it, maybe we could get that checked in?  I can't find the number right now unfortunately.

On Dec 8, 2010, at 11:42 AMPST, Danil Malyshev wrote:

> Hello All,
>  
> Some LLVM tests expect exact matching on printed out floating point numbers, and they fail (at least on Windows) because of that.
> For example, the test CodeGen/ARM/fpconsts.ll expects "-1.300000e+01" but the actual output could be "-1.300000e+001" (extra 0 in magnitude).
>  
> Please find attached the patch to fix such tests by ignoring insignificant zeroes.
> The following tests are affected:
>  
> * CodeGen/ARM/fpconsts.ll
> * Transforms/GVN/rle.ll
> * Transforms/InstCombine/bitcast.ll
> * Transforms/InstCombine/fsub.ll
> * Transforms/InstCombine/phi.ll
> * Transforms/InstCombine/shufflevec-constant.ll
>  
>  
> Thank you,
> Danil
> _______________________________________________
> llvm-commits mailing list
> llvm-commits at cs.uiuc.edu
> http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20101208/0e56d27e/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: fix-scientific-01.diff
Type: application/octet-stream
Size: 3876 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20101208/0e56d27e/attachment.obj>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20101208/0e56d27e/attachment-0001.html>