On Mit, 2013-02-20 at 18:46 +0100, Christian König wrote:
>
> diff --git a/lib/Target/R600/SIInstructions.td
> b/lib/Target/R600/SIInstructions.td
> index 700b8f8..866c7cb 100644
> --- a/lib/Target/R600/SIInstructions.td
> +++ b/lib/Target/R600/SIInstructions.td
> @@ -620,7 +620,7 @@ def V_INTERP_P1_F32 : VINTRP <
> 0x00000000,
> (outs VReg_32:$dst),
> (ins VReg_32:$i, i32imm:$attr_chan, i32imm:$attr, M0Reg:$m0),
> - "V_INTERP_P1_F32",
> + "V_INTERP_P1_F32 $dst, $i, $attr_chan, $attr, $m0",
I didn't put $m0 in the V_INTERP_MOV_F32 asm string because it's not an
explicit operand but used implicitly. I can see the value in printing it
anyway, but it should probably be distinguished from the explicit
operands somehow, e.g. "[$m0]" or something like that.
> let DisableEncoding = "$exec" in {
> def S_CBRANCH_EXECZ : SOPP <
> 0x00000008, (ins brtarget:$target, EXECReg:$exec),
> - "S_CBRANCH_EXECZ",
> + "S_CBRANCH_EXECZ $target",
> []
> >;
> def S_CBRANCH_EXECNZ : SOPP <
> 0x00000009, (ins brtarget:$target, EXECReg:$exec),
> - "S_CBRANCH_EXECNZ",
> + "S_CBRANCH_EXECNZ $target",
> []
> >;
> } // End DisableEncoding = "$exec"
> @@ -722,16 +722,19 @@ def S_WAITCNT : SOPP <0x0000000c, (ins i32imm:$simm16),
> "S_WAITCNT $simm16",
> //def S_TTRACEDATA : SOPP_ <0x00000016, "S_TTRACEDATA", []>;
>
> def V_CNDMASK_B32_e32 : VOP2 <0x00000000, (outs VReg_32:$dst),
> - (ins VSrc_32:$src0, VReg_32:$src1, VCCReg:$vcc), "V_CNDMASK_B32_e32",
> + (ins VSrc_32:$src0, VReg_32:$src1, VCCReg:$vcc),
> + "V_CNDMASK_B32_e32 $dst, $src0, $src1, $vcc",
> []
> >{
> let DisableEncoding = "$vcc";
These asm strings should be made consistent with whichever way we decide
to handle implicit operands as well.
Other than these nits, the series looks good to me. Be sure to run make
check, in case the lit tests need to be updated for the asm string
changes.
--
Earthling Michel Dänzer | http://www.amd.com
Libre software enthusiast | Debian, X and DRI developer
_______________________________________________
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/mesa-dev
