[Mesa-dev] [PATCH v2 2/3] i965/fs: Change SEL and MOV types as needed to propagate source modifiers

[Jason Ekstrand]

On Mon, Apr 6, 2015 at 6:13 PM, Jason Ekstrand <jason at jlekstrand.net> wrote:
> On Mon, Apr 6, 2015 at 4:55 PM, Matt Turner <mattst88 at gmail.com> wrote:
>> On Fri, Apr 3, 2015 at 2:06 PM, Jason Ekstrand <jason at jlekstrand.net> wrote:
>>> SEL and MOV instructions, as long as they don't have source modifiers, are
>>> just copying bits around.  This commit adds support to copy propagation to
>>> switch the type of a SEL or MOV instruction as needed so that it can
>>> propagate source modifiers.  This is needed because NIR generates integer
>>> SEL and MOV instructions whenver it doesn't know what else to generate.
>>>
>>> shader-db results with NIR:
>>> total FS instructions in shared programs: 4360910 -> 4360186 (-0.02%)
>>> FS instructions in affected programs:     59094 -> 58370 (-1.23%)
>>> helped:                                   341
>>> HURT:                                     0
>>> GAINED:                                   2
>>> LOST:                                     0
>>> ---
>>>  .../drivers/dri/i965/brw_fs_copy_propagation.cpp   | 33 +++++++++++++++++++---
>>>  1 file changed, 29 insertions(+), 4 deletions(-)
>>>
>>> diff --git a/src/mesa/drivers/dri/i965/brw_fs_copy_propagation.cpp b/src/mesa/drivers/dri/i965/brw_fs_copy_propagation.cpp
>>> index e8d092c..d321509 100644
>>> --- a/src/mesa/drivers/dri/i965/brw_fs_copy_propagation.cpp
>>> +++ b/src/mesa/drivers/dri/i965/brw_fs_copy_propagation.cpp
>>> @@ -275,6 +275,16 @@ is_logic_op(enum opcode opcode)
>>>             opcode == BRW_OPCODE_NOT);
>>>  }
>>>
>>> +static bool
>>> +can_change_source_types(fs_inst *inst)
>>> +{
>>> +   return !inst->src[0].abs && !inst->src[0].negate &&
>>> +          (inst->opcode == BRW_OPCODE_MOV ||
>>> +           (inst->opcode == BRW_OPCODE_SEL &&
>>> +            inst->conditional_mod == BRW_CONDITIONAL_NONE &&
>>
>> I understand what this means, but I'd change this line to
>> inst->predicate (!= BRW_PREDICATE_NONE).
>
> Seems reasonable.  I'll do that.
>
>>> +            !inst->src[1].abs && !inst->src[1].negate));
>>> +}
>>> +
>>>  bool
>>>  fs_visitor::try_copy_propagate(fs_inst *inst, int arg, acp_entry *entry)
>>>  {
>>> @@ -346,7 +356,9 @@ fs_visitor::try_copy_propagate(fs_inst *inst, int arg, acp_entry *entry)
>>>          type_sz(inst->src[arg].type)) % type_sz(entry->src.type) != 0)
>>>        return false;
>>>
>>> -   if (has_source_modifiers && entry->dst.type != inst->src[arg].type)
>>> +   if (has_source_modifiers &&
>>> +       entry->dst.type != inst->src[arg].type &&
>>> +       !can_change_source_types(inst))
>>>        return false;
>>>
>>>     if (brw->gen >= 8 && (entry->src.negate || entry->src.abs) &&
>>> @@ -412,9 +424,22 @@ fs_visitor::try_copy_propagate(fs_inst *inst, int arg, acp_entry *entry)
>>>        break;
>>>     }
>>>
>>> -   if (!inst->src[arg].abs) {
>>> -      inst->src[arg].abs = entry->src.abs;
>>> -      inst->src[arg].negate ^= entry->src.negate;
>>> +   if (has_source_modifiers) {
>>> +      if (entry->dst.type != inst->src[arg].type) {
>>> +         /* We are propagating source modifiers from a MOV with a different
>>
>> Like Anuj mentioned, MOV or SEL.
>
> You can only propagate *from* a MOV
>
>>> +          * type.  If we got here, then we can just change the source and
>>> +          * destination types of the instruction and keep going.
>>> +          */
>>> +         assert(can_change_source_types(inst));
>>> +         for (int i = 0; i < inst->sources; i++)
>>> +            inst->src[i].type = entry->dst.type;
>>
>> Probably put braces araound this.
>
> Sure.
>
>>> +         inst->dst.type = entry->dst.type;
>>> +      }
>>> +
>>> +      if (!inst->src[arg].abs) {
>>> +         inst->src[arg].abs = entry->src.abs;
>>> +         inst->src[arg].negate ^= entry->src.negate;
>>> +      }
>>>     }
>>>
>>>     return true;
>>> --
>>> 2.3.4
>>
>> Seems totally believable. I'm a little surprised the shader-db results
>> aren't bigger. I'd like to play with it and confirm it handles what I
>> was seeing with SEL instructions.

Ping