Re: Separating stdout and stderr

On 30 March 2016 at 10:47, Ben Elliston <[email protected]> wrote:
> On Wed, Mar 30, 2016 at 10:15:46AM +0200, Christophe Lyon wrote:
>
>> Well, I re-ran it, and I still have no evidence of using standard_wait.
>> Are you running native validations?
>
> Yes.
>
>> I remind you that we are using a cross-testing configuration, so we might be
>> using a different code path than the one you are looking at.
>
> Oh, I see. Sorry.
>
>> As I said earlier, I'm looking at local_exec in remote.exp and it
>> does contain a code fragment similar to the one you've just
>> patched. Does it need the same change?
>
> Yes, probably. Can you give it a try?
>

Rob cherry-picked this patch into our branch, and I manage to
re-run the tests with it. I'm still seeing "random" results
for *asan tests.
For instance:
https://ci.linaro.org/job/tcwg-compare-results/66/artifact/artifacts/logs/diff-x86_64.aarch64-linux-gnu.txt


> Cheers, Ben