[Intel-gfx] [PATCH 2/2] drm/i915/tgl: Restrict availables engines to rcs0 by default

[Summers, Stuart]

On Tue, 2019-10-01 at 14:54 +0100, Chris Wilson wrote:
> CI is still unstable whenever we enable more than one engine, and we
> have not yet found a better hack than restricting it to using just
> rcs0.
> 
> However, to allow testing to continue on the other engines by
> developers, we allow the available set of engines to be overridden on
> the command line with just the default set limited to [rcs0].
> 
> Signed-off-by: Chris Wilson <chris at chris-wilson.co.uk>
> Cc: Andi Shyti <andi.shyti at intel.com>
> Cc: Mika Kuoppala <mika.kuoppala at linux.intel.com>
> ---
>  drivers/gpu/drm/i915/gt/intel_engine_cs.c | 4 ++++
>  1 file changed, 4 insertions(+)
> 
> diff --git a/drivers/gpu/drm/i915/gt/intel_engine_cs.c
> b/drivers/gpu/drm/i915/gt/intel_engine_cs.c
> index 690da64ec256..9c8c7c8af394 100644
> --- a/drivers/gpu/drm/i915/gt/intel_engine_cs.c
> +++ b/drivers/gpu/drm/i915/gt/intel_engine_cs.c
> @@ -406,6 +406,10 @@ static bool engine_available(struct
> drm_i915_private *i915, int id)
>  	if (!HAS_ENGINE(i915, id))
>  		return false;
>  
> +	/* XXX reduced by default for CI stability XXX */
> +	if (IS_TIGERLAKE(i915) && i915_modparams.engines == -1u)
> +		return id == RCS0;
> +

So I'm not completely against this, and I generally (from a debug
perspective) like the idea of being able to tweak this kind of thing.
But given our CI, is this really needed on top of just reducing the
engines in i915_pci.c?

Thanks,
Stuart

>  	if (!(i915_modparams.engines & param_bit[id]))
>  		return false;
>  
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/x-pkcs7-signature
Size: 3270 bytes
Desc: not available
URL: <https://lists.freedesktop.org/archives/intel-gfx/attachments/20191008/0a8de8ee/attachment-0001.bin>