[Intel-gfx] [PATCH] drm/i915: Fix VCS ring selection after uapi decoupling

[Tvrtko Ursulin]

On 27/01/16 13:00, Chris Wilson wrote:
> On Wed, Jan 27, 2016 at 12:38:43PM +0000, Tvrtko Ursulin wrote:
>> From: Tvrtko Ursulin <tvrtko.ursulin at intel.com>
>>
>> This got broken in:
>>
>>     commit de1add360522c876c25ef2bbbbab1c94bdb509ab
>>     Author: Tvrtko Ursulin <tvrtko.ursulin at intel.com>
>>     Date:   Fri Jan 15 15:12:50 2016 +0000
>>
>>         drm/i915: Decouple execbuf uAPI from internal implementation
>>
>> BSD ring flags need to be shifted before they can be considered
>> indices into the ring array.
>>
>> Reported by Zhipeng Gong.
>>
>
> This should be hit by gem_busy on appropriate hardware. I need to add
> the extra ring to gem_exec_nop and make those all basic test.

Right hardware is the key.

>> Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin at intel.com>
>> Cc: Daniel Vetter <daniel.vetter at ffwll.ch>
>> Cc: Chris Wilson <chris at chris-wilson.co.uk>
>> Cc: Zhipeng Gong <zhipeng.gong at intel.com>
>> ---
>>   drivers/gpu/drm/i915/i915_gem_execbuffer.c | 11 +++++++----
>>   include/uapi/drm/i915_drm.h                | 10 ++++++----
>>   2 files changed, 13 insertions(+), 8 deletions(-)
>>
>> diff --git a/drivers/gpu/drm/i915/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
>> index 2dc08ce1079a..6fc620af14fa 100644
>> --- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c
>> +++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
>> @@ -1395,12 +1395,15 @@ eb_select_ring(struct drm_i915_private *dev_priv,
>>   	}
>>
>>   	if (user_ring_id == I915_EXEC_BSD && HAS_BSD2(dev_priv)) {
>> -		unsigned int bsd_idx = args->flags & I915_EXEC_BSD_MASK;
>> +		unsigned int bsd_idx = (args->flags & I915_EXEC_BSD_MASK) >>
>> +					I915_EXEC_BSD_SHIFT;
>>
>> -		if (bsd_idx == I915_EXEC_BSD_DEFAULT) {
>> +		if (bsd_idx == (I915_EXEC_BSD_DEFAULT >> I915_EXEC_BSD_SHIFT)) {
>>   			bsd_idx = gen8_dispatch_bsd_ring(dev_priv, file);
>> -		} else if (bsd_idx >= I915_EXEC_BSD_RING1 &&
>> -			   bsd_idx <= I915_EXEC_BSD_RING2) {
>> +		} else if (bsd_idx >=
>> +			  (I915_EXEC_BSD_RING1 >> I915_EXEC_BSD_SHIFT) &&
>> +			  bsd_idx <=
>> +			  (I915_EXEC_BSD_RING2 >> I915_EXEC_BSD_SHIFT)) {
>
> Wouldn't inserting bsd_idx >>= I915_EXEC_BSD_SHIFT; here be simpler ?

Indeed it would, v2 is on the way.

Regards,

Tvrtko