[Mesa-dev] [PATCH] i965: possible typo in ir_unop_f2b case.

[Dave Airlie]

On 4 September 2014 05:02, Matt Turner <mattst88 at gmail.com> wrote:
> On Mon, Sep 1, 2014 at 4:15 PM, Dave Airlie <airlied at gmail.com> wrote:
>> From: Dave Airlie <airlied at redhat.com>
>>
>> Coverity reported this, I'm not sure this patch is correct, but I'm sure
>> someone who knows can fix this or push my fix.
>>
>> Signed-off-by: Dave Airlie <airlied at redhat.com>
>> ---
>>  src/mesa/drivers/dri/i965/brw_fs_visitor.cpp | 2 +-
>>  1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/src/mesa/drivers/dri/i965/brw_fs_visitor.cpp b/src/mesa/drivers/dri/i965/brw_fs_visitor.cpp
>> index 2fa90a4..5d10eb1 100644
>> --- a/src/mesa/drivers/dri/i965/brw_fs_visitor.cpp
>> +++ b/src/mesa/drivers/dri/i965/brw_fs_visitor.cpp
>> @@ -2300,7 +2300,7 @@ fs_visitor::emit_bool_to_cond_code(ir_rvalue *ir)
>>
>>     case ir_unop_f2b:
>>        if (brw->gen >= 6) {
>> -         emit(CMP(reg_null_d, op[0], fs_reg(0.0f), BRW_CONDITIONAL_NZ));
>> +         emit(CMP(reg_null_f, op[0], fs_reg(0.0f), BRW_CONDITIONAL_NZ));
>>        } else {
>>           inst = emit(MOV(reg_null_f, op[0]));
>>           inst->conditional_mod = BRW_CONDITIONAL_NZ;
>> --
>
> I don't think this should matter. I'm mostly curious about what
> warning Coverity gave for this? It seems like it would have to a
> pretty deep understanding of the code to warn about this
> (alternatively, no understanding at all).

It thought it was a cut-n-paste error, since the i2b case uses
reg_null_d in both sides of the if, but the f2b uses reg_null_d and
reg_null_f.

Dave.