[Intel-gfx] [igt-dev] [PATCH i-g-t 24/25] gem_wsim: Discover engines
Tvrtko Ursulin
tvrtko.ursulin at linux.intel.com
Fri May 17 12:19:18 UTC 2019
On 17/05/2019 13:10, Andi Shyti wrote:
> On Fri, May 17, 2019 at 12:25:25PM +0100, Tvrtko Ursulin wrote:
>> From: Tvrtko Ursulin <tvrtko.ursulin at intel.com>
>>
>> Instead of hardcoding the VCS balancing engines, discover, both with the
>> new engines query, or with the legacy get_param in the fallback case, so
>> class based addressing always works.
>>
>> Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin at intel.com>
>> ---
>> benchmarks/gem_wsim.c | 180 ++++++++++++++++++++++++++++++++++++++++--
>> 1 file changed, 173 insertions(+), 7 deletions(-)
>>
>> diff --git a/benchmarks/gem_wsim.c b/benchmarks/gem_wsim.c
>> index d43e7c767801..539de243f6e8 100644
>> --- a/benchmarks/gem_wsim.c
>> +++ b/benchmarks/gem_wsim.c
>> @@ -365,34 +365,198 @@ static int str_to_engine(const char *str)
>> return -1;
>> }
>>
>> +static bool __engines_queried;
>> +static unsigned int __num_engines;
>> +static struct i915_engine_class_instance *__engines;
>> +
>> +static int
>> +__i915_query(int i915, struct drm_i915_query *q)
>> +{
>> + if (igt_ioctl(i915, DRM_IOCTL_I915_QUERY, q))
>> + return -errno;
>> + return 0;
>> +}
>> +
>> +static int
>> +__i915_query_items(int i915, struct drm_i915_query_item *items, uint32_t n_items)
>> +{
>> + struct drm_i915_query q = {
>> + .num_items = n_items,
>> + .items_ptr = to_user_pointer(items),
>> + };
>> + return __i915_query(i915, &q);
>> +}
>> +
>> +static void
>> +i915_query_items(int i915, struct drm_i915_query_item *items, uint32_t n_items)
>> +{
>> + igt_assert_eq(__i915_query_items(i915, items, n_items), 0);
>> +}
>> +
>> +static bool has_query(int i915)
>> +{
>> + struct drm_i915_query query = {};
>> +
>> + return __i915_query(i915, &query) == 0;
>> +}
>> +
>> +static bool has_engine_query(int i915)
>> +{
>> + struct drm_i915_query_item item = {
>> + .query_id = DRM_I915_QUERY_ENGINE_INFO,
>> + };
>> +
>> + return __i915_query_items(i915, &item, 1) == 0 && item.length > 0;
>> +}
>> +
>> +static void query_engines(void)
>> +{
>> + struct i915_engine_class_instance *engines;
>> + unsigned int num;
>> +
>> + if (__engines_queried)
>> + return;
>> +
>> + __engines_queried = true;
>> +
>> + if (!has_query(fd) || !has_engine_query(fd)) {
>
> One question, still. What is the real use of this check and
> 'has_query' that is used only here.
>
> I mean... here you want to check whether the "ioctl is not
> implemented" or "ioctl is not implemented and length is 0".
>
> Wouldn't in this case just '!has_engine_query()' be enough? or
> have I missed any case?
You haven't missed anything. I have been pointlessly verbose and a bit
lazy by copy-pasting a lot.
has_engine_query is a superset of has_query for the purpose of ioctl
detection.
>
>> + unsigned int num_bsd = gem_has_bsd(fd) + gem_has_bsd2(fd);
>> + unsigned int i = 0;
>> +
>> + igt_assert(num);
>> +
>> + num = 1 + num_bsd;
>
> did you mean the above two lines swapped?
No, I want to avoid running on platforms with no vcs engines since no
one ever tested gem_wsim there.
>> +
>> + if (gem_has_blt(fd))
>> + num++;
>> +
>> + if (gem_has_vebox(fd))
>> + num++;
>> +
>> + engines = calloc(num,
>> + sizeof(struct i915_engine_class_instance));
>> + igt_assert(engines);
>> +
>> + engines[i].engine_class = I915_ENGINE_CLASS_RENDER;
>> + engines[i].engine_instance = 0;
>> + i++;
>> +
>> + if (gem_has_blt(fd)) {
>> + engines[i].engine_class = I915_ENGINE_CLASS_COPY;
>> + engines[i].engine_instance = 0;
>> + i++;
>> + }
>> +
>> + if (gem_has_bsd(fd)) {
>> + engines[i].engine_class = I915_ENGINE_CLASS_VIDEO;
>> + engines[i].engine_instance = 0;
>> + i++;
>> + }
>> +
>> + if (gem_has_bsd2(fd)) {
>> + engines[i].engine_class = I915_ENGINE_CLASS_VIDEO;
>> + engines[i].engine_instance = 1;
>> + i++;
>> + }
>> +
>> + if (gem_has_vebox(fd)) {
>> + engines[i].engine_class =
>> + I915_ENGINE_CLASS_VIDEO_ENHANCE;
>> + engines[i].engine_instance = 0;
>> + i++;
>> + }
>
> mmhhh... isn't this the intel_execution_engine2[]? Yet another
> way for having engine list... in the long run, updating here (as
> well) won't be easy to remember.
Not here, gem_wsim uses some of the IGT libraries, but should keep it at
minimum. So I think we don't want to pull in the engine array etc.
Regards,
Tvrtko
More information about the Intel-gfx
mailing list