[cfe-dev] choosing the gcc toolchain

[Daniel Sanders]

If I understand correctly, patching manually using the raw diff results in the failure Reed mentions but this test would pass if the patch is applied using 'arc patch D2911'. Is that right?
If so, the patch has already been reviewed and LGTM'd so I think you're ok to commit.

> -----Original Message-----
> From: Reed Kotler [mailto:rkotler at mips.com]
> Sent: 04 March 2014 17:14
> To: Simon Atanasyan; Daniel Sanders
> Cc: cfe-dev at cs.uiuc.edu
> Subject: Re: choosing the gcc toolchain
> 
> I'm running another build to just check some things but otherwise; I think
> that if it's okay with Daniel that we should check in your patch today.
> 
> 
> On 03/03/2014 08:33 PM, Simon Atanasyan wrote:
> > On Tue, Mar 4, 2014 at 4:39 AM, Reed Kotler <rkotler-
> 8NJIiSa5LzA at public.gmane.org> wrote:
> >> On the big endian machine; which is the one we have as the build
> >> slave to the official llvm build bot, there is only one failure with
> >> your patch and it's some mips specific test in the driver.
> >
> > [...]
> >
> >> FAIL: Clang :: Driver/mips-reduced-toolchain.cpp (2882 of 16898)
> >
> > It's ok. The patch from the Phabricator does not contain new input
> > test directory for the driver debian_reduced_mips_tree. The revision
> > page shows these new files though. I do not know the reason.
> >
> > Thanks for the checking.
> >