Re: [PATCH] drm/i915: relax debug BUG_ON() for closed context in hw_id pin

From: Zhenyu Wang <zhenyuw@linux.intel.com>
To: Chris Wilson <chris@chris-wilson.co.uk>
Cc: intel-gfx@lists.freedesktop.org, intel-gvt-dev@lists.freedesktop.org
Subject: Re: [PATCH] drm/i915: relax debug BUG_ON() for closed context in hw_id pin
Date: Sat, 9 Mar 2019 11:39:36 +0800	[thread overview]
Message-ID: <20190309033936.GA27119@zhen-hp.sh.intel.com> (raw)
In-Reply-To: <155203390621.27405.3964566776159217086@skylake-alporthouse-com>

[-- Attachment #1.1: Type: text/plain, Size: 1028 bytes --]

On 2019.03.08 08:31:51 +0000, Chris Wilson wrote:
> Quoting Zhenyu Wang (2019-03-08 07:52:37)
> > Current GVT created context is marked closed as not to be used for
> > host user. But its hw_id should still be used. So this is to relax
> > debug BUG_ON() in __i915_gem_context_pin_hw_id() for GVT contexts
> > which can use force single submission flag to identify.
> 
> The alternative strategy would be to always pin the id for GVT. How many
> gvt contexts? One per host or one per client? Or we don't mark them as
> closed (not so keen on that as it does provide some protection).
>

Currently one per VM guest, always pin the id would also be good.

> I think I'd rather delete the GEM_BUG_ON() if it's not invariant -- we
> only escape it firing for kernel contexts because they pin their id.

I think anyway we'd better pin id for gvt context, as it's good to keep
it for one VM cycle.

thanks

-- 
Open Source Technology Center, Intel ltd.

$gpg --keyserver wwwkeys.pgp.net --recv-keys 4D781827

[-- Attachment #1.2: signature.asc --]
[-- Type: application/pgp-signature, Size: 195 bytes --]

[-- Attachment #2: Type: text/plain, Size: 159 bytes --]

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx