ppc64 clang LNT buildbot / reference_output / test failures

[Renato Golin]

On 25 April 2013 20:33, Daniel Dunbar <daniel at zuster.org> wrote:

> I believe that almost all of the work to make reference outputs portable
> is done. I believe the current status is that the reference outputs largely
> work across Darwin x {x86, ARM, PPC} and Linux x { x86?, ARM }, which makes
> it increasingly unlikely that the reference outputs are incorrect for your
> platform.
>

Linux ARM is good, Linux x86 still has one issue (Olden/voroni) left.


There isn't currently a way from the LNT command line to disable the use of
> reference outputs, and I personally would prefer this serve as a forcing
> function to fix them in the few remaining places they don't work.
>

I agree with Daniel here.

The main idea is to create a reference output that doesn't depend on which
platform, OS or library version the tests are running, and this is achieved
by implementing the library routines that are not portable (RNGs, Sorts,
etc) locally, so that the output is stable and consistent.

While making the ARM test-suite green I haven't come across a single
example where this was impossible or undesirable, and the remaining x86
case is also simple, so I believe the same is true for PPC.

cheers,
--renato
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20130426/d8e1d529/attachment.html>