search for: ptx_float_3op

Displaying 5 results from an estimated 5 matches for "ptx_float_3op".

2011 Oct 07
6
[LLVMdev] Enhancing TableGen
...definitions have no value, and I > would welcome a for-loop construct that can create them. I think this > was the primary motivation behind Che-Liang's proposal. Ok. We agree here. > I do NOT want to factor out redundancy from instruction definitions like this: > > multiclass PTX_FLOAT_3OP<string opcstr> { > def rr32 : InstPTX<(outs RegF32:$d), > (ins RndMode:$r, RegF32:$a, RegF32:$b), > !strconcat(opcstr, "$r.f32\t$d, $a, $b"), []>; > def ri32 : InstPTX<(outs RegF32:$d), > (ins...
2011 Oct 07
0
[LLVMdev] Enhancing TableGen
...; Such 'embarrassingly sequential' definitions have no value, and I would welcome a for-loop construct that can create them. I think this was the primary motivation behind Che-Liang's proposal. I do NOT want to factor out redundancy from instruction definitions like this: multiclass PTX_FLOAT_3OP<string opcstr> { def rr32 : InstPTX<(outs RegF32:$d), (ins RndMode:$r, RegF32:$a, RegF32:$b), !strconcat(opcstr, "$r.f32\t$d, $a, $b"), []>; def ri32 : InstPTX<(outs RegF32:$d), (ins RndMode:$r, RegF32:$a, f...
2011 Oct 07
0
[LLVMdev] Enhancing TableGen
...swer '...and where they came from'. > This could be mitigated somewhat by doing something like this: > > class binary_pattern<string opcstr, string type> { > string pattern = !strconcat(opcstr, "$r."#type#"\t$d, $a, $b"); > } > > multiclass PTX_FLOAT_3OP<string opcstr> { > def rr32 : InstPTX<(outs RegF32:$d), > (ins RndMode:$r, RegF32:$a, RegF32:$b), > binary_pattern<opcstrm "f32">.pattern, []>; > def ri32 : InstPTX<(outs RegF32:$d), > (ins R...
2011 Oct 07
6
[LLVMdev] Enhancing TableGen
Evan Cheng <evan.cheng at apple.com> writes: > David, we cannot accept the 'multidef' keyword. Please revert it. Working on it now. > We appreciate you thinking ahead about MIC, but we are against the > massive refactoring and complicated abstraction scheme. We'll never > accept those patches. How about a less massive and complicated scheme? I think we can make
2011 Oct 08
3
[LLVMdev] Enhancing TableGen
...me from'. > >> This could be mitigated somewhat by doing something like this: >> >> class binary_pattern<string opcstr, string type> { >>  string pattern = !strconcat(opcstr, "$r."#type#"\t$d, $a, $b"); >> } >> >> multiclass PTX_FLOAT_3OP<string opcstr> { >>  def rr32 : InstPTX<(outs RegF32:$d), >>                     (ins RndMode:$r, RegF32:$a, RegF32:$b), >>                     binary_pattern<opcstrm "f32">.pattern, []>; >>  def ri32 : InstPTX<(outs RegF32:$d), >>      ...