[Mesa-dev] [PATCH 14/24] i965: fix atomic buffer index for bindings other than 0

[Ilia Mirkin]

On Sat, Sep 26, 2015 at 5:17 PM, Timothy Arceri <t_arceri at yahoo.com.au> wrote:
>
>
> On 27 September 2015 6:23:42 am AEST, Ilia Mirkin <imirkin at alum.mit.edu> wrote:
>>On Thu, Sep 17, 2015 at 3:02 AM, Timothy Arceri <t_arceri at yahoo.com.au>
>>wrote:
>>> Since commit c0cd5b var->data.binding was being used as a replacement
>>> for atomic buffer index, but they don't have to be the same value
>>they
>>> just happen to end up the same when binding is 0.
>>>
>>> Now that we store the atomic uniform location in var->data.location
>>> we can use this to lookup the atomic buffer index in uniform storage.
>>
>>FWIW I ran into a similar problem when trying to implement atomic
>>counters in gallium. I assume that this logic depends on some other
>>bit from this patch series, since I don't see that we store the value
>>in var->data.location in the upstream code?
>
> It wad added recently [1] as part of my struct indirect indexing fix.
>
> [1] http://cgit.freedesktop.org/mesa/mesa/tree/src/glsl/link_uniforms.cpp#n750

Oh, that's sneaky. It wasn't where I expected :)

  -ilia