Re: [dpdk-dev] aesni mb on Centos KO for three weeks

[Thomas Monjalon <thomas@monjalon.net>]

+Cc more people

05/06/2019 14:05, David Marchand:
> I noticed build errors on patches I sent yesterday:
> http://mails.dpdk.org/archives/test-report/2019-June/084633.html

I tried to reproduce this error with make on my system
but I don't manage to see such error.
It may be specific to a GCC version.

> I went and looked at the test-report mailing list.
> 
> The failures on master builds started on May 14th:
> http://mails.dpdk.org/archives/test-report/2019-May/082432.html
> But still, the test was reported as successfull.
> 
> The status got properly reported starting May 28th:
> http://mails.dpdk.org/archives/test-report/2019-May/083421.html
> Looking at patchwork this day, I can see that it changed between
> http://patchwork.dpdk.org/patch/53721/ and
> http://patchwork.dpdk.org/patch/53722/
> 
> So I suppose someone noticed that the meson build failures were not
> properly reported and fixed it, but the error still remained.
> 
> So, we have this error for 3 weeks, is there someone looking at fixing it
> please?
> I did not manage to build the ipsec_mb thing on my rhel.
> Not sure what is wrong, but I don't want to waste more time on it.
> 
> 
> As a sidenote, we don't have Intel build reports on the patches since this
> morning, around
> http://mails.dpdk.org/archives/test-report/2019-June/084647.html
> Has it been disabled intentionnally?

I agree with David that there is a serious issue
with compilation testing for a long time.
Please can we fix it quickly?