[Beignet] 9e8874c breaking (more) Haswell systems?

[Zhigang Gong]

I will raise this issue to Linux i915 kernel team. And hope we can get a
clean fix in Linux kernel side finally.

Thanks for pointing this out.

> -----Original Message-----
> From: Beignet [mailto:beignet-bounces at lists.freedesktop.org] On Behalf Of
> Zhenyu Wang
> Sent: Thursday, April 23, 2015 2:14 AM
> To: Rebecca N. Palmer; Zhigang Gong
> Cc: beignet at lists.freedesktop.org
> Subject: Re: [Beignet] 9e8874c breaking (more) Haswell systems?
> 
> On 2015.04.18 22:16:07 +0100, Rebecca N. Palmer wrote:
> > http://cgit.freedesktop.org/beignet/commit/?h=Release_v1.0&id=9e8874c7
> > a16b029d5b857a11ff9b23102752b4db was intended to catch the "silently
> > does nothing on large arrays" bug, and does so (with the message
> > "drm_intel_gem_bo_context_exec() failed: No space left on device").
> >
> > However, a user reports that it also breaks a previously
> > working-without-workarounds Haswell system, with
> > "drm_intel_gem_bo_context_exec() failed: Invalid argument":
> > https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=781875#46
> >
> > The usual i915_cmd_parser workaround did work, so it might be a good
> > idea to suggest it in this error message.  (The Debian kernel does
> > have c9224fa...,
> > https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=767148 , but as
> > noted earlier this isn't enough for all Haswell systems.)
> >
> 
> Zhigang, please double check this, mostly for
> ef7127c03bd533277afc443b335c37a69927250a.
> 
> If this kernel workaround is not required on some HW stepping, then please
> send a kernel patch to fix it. If that might fail for some context, or you
need to
> set it in your context, then they might be in kernel white list.
> 
> Otherwise current beignet just failed on HSW with cmd parser and we're
very
> close to really enable cmd parser for HSW in normal kernel config.
> 
> --
> Open Source Technology Center, Intel ltd.
> 
> $gpg --keyserver wwwkeys.pgp.net --recv-keys 4D781827