[Mesa-dev] [PATCH] Revert "compiler/glsl: handle case where we have multiple users for types"

[Timothy Arceri]

On 18/4/19 4:38 pm, Tapani Pälli wrote:
> On 4/18/19 8:50 AM, Timothy Arceri wrote:
>> On 18/4/19 3:08 pm, Tapani Pälli wrote:
>>>
>>>
>>> On 4/18/19 8:05 AM, Timothy Arceri wrote:
>>>> This reverts commit 624789e3708c87ea2a4c8d2266266b489b421cba.
>>>>
>>>> It caused 400+ piglit tests to crash on radeonsi, I haven't been able
>>>> to track down the reason yet.
>>>
>>> Could you share the backtrace?
>>
>> It's more complicated than that. There is some kind of race condition, 
>> we end up crashing in nir validation as it seems the types are 
>> released before validation finishes.
> 
> I do wonder how that can happen :/ I'm ok with revert;
> 
> Acked-by: Tapani Pälli <tapani.palli at intel.com>

I'll hold off pushing for now and try to take a closer look tomorrow. 
Your patch seems correct to me but so did the threading in radeonsi so 
I'm surprised its falling over.

> 
> 
>> e.g. for ./bin/getuniform-03 -auto -fbo
>>
>> ...
>>      vec1 32 ssa_404 = deref_array &(*ssa_403)[0] (uniform vec4) /* 
>> &gl_TextureEnvColor[0] */
>> error: glsl_type_is_array(parent->type) || 
>> glsl_type_is_matrix(parent->type) 
>> (../src/compiler/nir/nir_validate.c:453)
> 
> // Tapani