[igt-dev] [PATCH] [PATCH i-g-t][V2]tests/i915/gem_exec_nop:-Fixed Crash issue seen on few platform

[Chris Wilson]

Quoting Arjun Melkaveri (2020-03-27 05:04:35)
> Crash issue was seen w.r.t engines on different platforms.
> Fixed this by initializing engine and engine name with
> maximum supported engines value.
> 
> V2:
> 
> Made changes for other tests using engine array.
> 
> Reviewed-by: Antonio Argenziano <antonio.argenziano at intel.com>
> Signed-off-by: Arjun Melkaveri <arjun.melkaveri at intel.com>
> ---
>  tests/i915/gem_exec_nop.c | 18 ++++++++++--------
>  1 file changed, 10 insertions(+), 8 deletions(-)
> 
> diff --git a/tests/i915/gem_exec_nop.c b/tests/i915/gem_exec_nop.c
> index 6da6234f..823788f8 100644
> --- a/tests/i915/gem_exec_nop.c
> +++ b/tests/i915/gem_exec_nop.c
> @@ -55,6 +55,7 @@
>  
>  #define MAX_PRIO LOCAL_I915_CONTEXT_MAX_USER_PRIORITY
>  #define MIN_PRIO LOCAL_I915_CONTEXT_MIN_USER_PRIORITY
> +#define GEM_MAX_ENGINES I915_EXEC_RING_MASK + 1

I'm not that keen on GEM_ since it makes it seem like it's a fundamental
limit, but in reality it's just the execbuf2 ABI.

Rather than squeeze in EB_ABI_MAX_ENGINES, I'd just use MAX_ENGINES for
the local define.

#define I915_EXEC_MAX_ENGINES (I915_EXEC_RING_MASK + 1)

although we will argue with ourselves for using I915_EXEC in our own
gem_engine_topology, so maybe EXECBUF_MAX_ENGINES is good enough.

As for iterating multiple times, I was hoping that having derived the
list of engines in gem_engine_topology, we would be able to reuse it!
-Chris