[llvm] r338507 - [X86] Improved sched models for X86 BT*rr instructions.
Andrew V. Tischenko via llvm-commits
llvm-commits at lists.llvm.org
Wed Aug 1 03:24:27 PDT 2018
Author: avt77
Date: Wed Aug 1 03:24:27 2018
New Revision: 338507
URL: http://llvm.org/viewvc/llvm-project?rev=338507&view=rev
Log:
[X86] Improved sched models for X86 BT*rr instructions.
Differential Revision: https://reviews.llvm.org/D49243
Modified:
llvm/trunk/lib/Target/X86/X86InstrInfo.td
llvm/trunk/lib/Target/X86/X86SchedBroadwell.td
llvm/trunk/lib/Target/X86/X86SchedHaswell.td
llvm/trunk/lib/Target/X86/X86SchedSandyBridge.td
llvm/trunk/lib/Target/X86/X86SchedSkylakeClient.td
llvm/trunk/lib/Target/X86/X86SchedSkylakeServer.td
llvm/trunk/lib/Target/X86/X86Schedule.td
llvm/trunk/lib/Target/X86/X86ScheduleAtom.td
llvm/trunk/lib/Target/X86/X86ScheduleBtVer2.td
llvm/trunk/lib/Target/X86/X86ScheduleSLM.td
llvm/trunk/lib/Target/X86/X86ScheduleZnver1.td
Modified: llvm/trunk/lib/Target/X86/X86InstrInfo.td
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Target/X86/X86InstrInfo.td?rev=338507&r1=338506&r2=338507&view=diff
==============================================================================
--- llvm/trunk/lib/Target/X86/X86InstrInfo.td (original)
+++ llvm/trunk/lib/Target/X86/X86InstrInfo.td Wed Aug 1 03:24:27 2018
@@ -1750,7 +1750,7 @@ def LAHF : I<0x9F, RawFrm, (outs),
// Bit tests instructions: BT, BTS, BTR, BTC.
let Defs = [EFLAGS] in {
-let SchedRW = [WriteALU] in {
+let SchedRW = [WriteBitTest] in {
def BT16rr : I<0xA3, MRMDestReg, (outs), (ins GR16:$src1, GR16:$src2),
"bt{w}\t{$src2, $src1|$src1, $src2}",
[(set EFLAGS, (X86bt GR16:$src1, GR16:$src2))]>,
@@ -1783,7 +1783,7 @@ let mayLoad = 1, hasSideEffects = 0, Sch
[]>, TB, NotMemoryFoldable;
}
-let SchedRW = [WriteALU] in {
+let SchedRW = [WriteBitTest] in {
def BT16ri8 : Ii8<0xBA, MRM4r, (outs), (ins GR16:$src1, i16i8imm:$src2),
"bt{w}\t{$src2, $src1|$src1, $src2}",
[(set EFLAGS, (X86bt GR16:$src1, i16immSExt8:$src2))]>,
@@ -1818,7 +1818,7 @@ def BT64mi8 : RIi8<0xBA, MRM4m, (outs),
} // SchedRW
let hasSideEffects = 0 in {
-let SchedRW = [WriteALU], Constraints = "$src1 = $dst" in {
+let SchedRW = [WriteBitTest], Constraints = "$src1 = $dst" in {
def BTC16rr : I<0xBB, MRMDestReg, (outs GR16:$dst), (ins GR16:$src1, GR16:$src2),
"btc{w}\t{$src2, $src1|$src1, $src2}", []>,
OpSize16, TB, NotMemoryFoldable;
@@ -1842,7 +1842,7 @@ def BTC64mr : RI<0xBB, MRMDestMem, (outs
NotMemoryFoldable;
}
-let SchedRW = [WriteALU], Constraints = "$src1 = $dst" in {
+let SchedRW = [WriteBitTest], Constraints = "$src1 = $dst" in {
def BTC16ri8 : Ii8<0xBA, MRM7r, (outs GR16:$dst), (ins GR16:$src1, i16i8imm:$src2),
"btc{w}\t{$src2, $src1|$src1, $src2}", []>, OpSize16, TB;
def BTC32ri8 : Ii8<0xBA, MRM7r, (outs GR32:$dst), (ins GR32:$src1, i32i8imm:$src2),
@@ -1861,7 +1861,7 @@ def BTC64mi8 : RIi8<0xBA, MRM7m, (outs),
Requires<[In64BitMode]>;
}
-let SchedRW = [WriteALU], Constraints = "$src1 = $dst" in {
+let SchedRW = [WriteBitTest], Constraints = "$src1 = $dst" in {
def BTR16rr : I<0xB3, MRMDestReg, (outs GR16:$dst), (ins GR16:$src1, GR16:$src2),
"btr{w}\t{$src2, $src1|$src1, $src2}", []>,
OpSize16, TB, NotMemoryFoldable;
@@ -1885,7 +1885,7 @@ def BTR64mr : RI<0xB3, MRMDestMem, (outs
NotMemoryFoldable;
}
-let SchedRW = [WriteALU], Constraints = "$src1 = $dst" in {
+let SchedRW = [WriteBitTest], Constraints = "$src1 = $dst" in {
def BTR16ri8 : Ii8<0xBA, MRM6r, (outs GR16:$dst), (ins GR16:$src1, i16i8imm:$src2),
"btr{w}\t{$src2, $src1|$src1, $src2}", []>,
OpSize16, TB;
@@ -1908,7 +1908,7 @@ def BTR64mi8 : RIi8<0xBA, MRM6m, (outs),
Requires<[In64BitMode]>;
}
-let SchedRW = [WriteALU], Constraints = "$src1 = $dst" in {
+let SchedRW = [WriteBitTest], Constraints = "$src1 = $dst" in {
def BTS16rr : I<0xAB, MRMDestReg, (outs GR16:$dst), (ins GR16:$src1, GR16:$src2),
"bts{w}\t{$src2, $src1|$src1, $src2}", []>,
OpSize16, TB, NotMemoryFoldable;
@@ -1932,7 +1932,7 @@ def BTS64mr : RI<0xAB, MRMDestMem, (outs
NotMemoryFoldable;
}
-let SchedRW = [WriteALU], Constraints = "$src1 = $dst" in {
+let SchedRW = [WriteBitTest], Constraints = "$src1 = $dst" in {
def BTS16ri8 : Ii8<0xBA, MRM5r, (outs GR16:$dst), (ins GR16:$src1, i16i8imm:$src2),
"bts{w}\t{$src2, $src1|$src1, $src2}", []>, OpSize16, TB;
def BTS32ri8 : Ii8<0xBA, MRM5r, (outs GR32:$dst), (ins GR32:$src1, i32i8imm:$src2),
Modified: llvm/trunk/lib/Target/X86/X86SchedBroadwell.td
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Target/X86/X86SchedBroadwell.td?rev=338507&r1=338506&r2=338507&view=diff
==============================================================================
--- llvm/trunk/lib/Target/X86/X86SchedBroadwell.td (original)
+++ llvm/trunk/lib/Target/X86/X86SchedBroadwell.td Wed Aug 1 03:24:27 2018
@@ -137,6 +137,7 @@ def : WriteRes<WriteSETCCStore, [BWPort
let NumMicroOps = 3;
}
def : WriteRes<WriteLAHFSAHF, [BWPort06]>;
+def : WriteRes<WriteBitTest,[BWPort06]>; // Bit Test instrs
// Bit counts.
defm : BWWriteResPair<WriteBSF, [BWPort1], 3>;
@@ -603,14 +604,6 @@ def BWWriteResGroup6 : SchedWriteRes<[BW
let ResourceCycles = [1];
}
def: InstRW<[BWWriteResGroup6], (instrs CDQ, CQO)>;
-def: InstRW<[BWWriteResGroup6], (instregex "BT(16|32|64)ri8",
- "BT(16|32|64)rr",
- "BTC(16|32|64)ri8",
- "BTC(16|32|64)rr",
- "BTR(16|32|64)ri8",
- "BTR(16|32|64)rr",
- "BTS(16|32|64)ri8",
- "BTS(16|32|64)rr")>;
def BWWriteResGroup7 : SchedWriteRes<[BWPort15]> {
let Latency = 1;
Modified: llvm/trunk/lib/Target/X86/X86SchedHaswell.td
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Target/X86/X86SchedHaswell.td?rev=338507&r1=338506&r2=338507&view=diff
==============================================================================
--- llvm/trunk/lib/Target/X86/X86SchedHaswell.td (original)
+++ llvm/trunk/lib/Target/X86/X86SchedHaswell.td Wed Aug 1 03:24:27 2018
@@ -150,6 +150,7 @@ def : WriteRes<WriteSETCCStore, [HWPort
let NumMicroOps = 3;
}
def : WriteRes<WriteLAHFSAHF, [HWPort06]>;
+def : WriteRes<WriteBitTest,[HWPort06]>;
// This is for simple LEAs with one or two input operands.
// The complex ones can only execute on port 1, and they require two cycles on
@@ -895,14 +896,6 @@ def HWWriteResGroup7 : SchedWriteRes<[HW
let ResourceCycles = [1];
}
def: InstRW<[HWWriteResGroup7], (instrs CDQ, CQO)>;
-def: InstRW<[HWWriteResGroup7], (instregex "BT(16|32|64)ri8",
- "BT(16|32|64)rr",
- "BTC(16|32|64)ri8",
- "BTC(16|32|64)rr",
- "BTR(16|32|64)ri8",
- "BTR(16|32|64)rr",
- "BTS(16|32|64)ri8",
- "BTS(16|32|64)rr")>;
def HWWriteResGroup8 : SchedWriteRes<[HWPort15]> {
let Latency = 1;
Modified: llvm/trunk/lib/Target/X86/X86SchedSandyBridge.td
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Target/X86/X86SchedSandyBridge.td?rev=338507&r1=338506&r2=338507&view=diff
==============================================================================
--- llvm/trunk/lib/Target/X86/X86SchedSandyBridge.td (original)
+++ llvm/trunk/lib/Target/X86/X86SchedSandyBridge.td Wed Aug 1 03:24:27 2018
@@ -145,6 +145,7 @@ def : WriteRes<WriteSETCCStore, [SBPort
let NumMicroOps = 3;
}
def : WriteRes<WriteLAHFSAHF, [SBPort05]>;
+def : WriteRes<WriteBitTest,[SBPort05]>;
// This is for simple LEAs with one or two input operands.
// The complex ones can only execute on port 1, and they require two cycles on
@@ -570,14 +571,6 @@ def SBWriteResGroup4 : SchedWriteRes<[SB
let ResourceCycles = [1];
}
def: InstRW<[SBWriteResGroup4], (instrs CDQ, CQO)>;
-def: InstRW<[SBWriteResGroup4], (instregex "BT(16|32|64)ri8",
- "BT(16|32|64)rr",
- "BTC(16|32|64)ri8",
- "BTC(16|32|64)rr",
- "BTR(16|32|64)ri8",
- "BTR(16|32|64)rr",
- "BTS(16|32|64)ri8",
- "BTS(16|32|64)rr")>;
def SBWriteResGroup5 : SchedWriteRes<[SBPort15]> {
let Latency = 1;
Modified: llvm/trunk/lib/Target/X86/X86SchedSkylakeClient.td
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Target/X86/X86SchedSkylakeClient.td?rev=338507&r1=338506&r2=338507&view=diff
==============================================================================
--- llvm/trunk/lib/Target/X86/X86SchedSkylakeClient.td (original)
+++ llvm/trunk/lib/Target/X86/X86SchedSkylakeClient.td Wed Aug 1 03:24:27 2018
@@ -136,6 +136,7 @@ def : WriteRes<WriteSETCCStore, [SKLPor
let NumMicroOps = 3;
}
def : WriteRes<WriteLAHFSAHF, [SKLPort06]>;
+def : WriteRes<WriteBitTest,[SKLPort06]>; //
// Bit counts.
defm : SKLWriteResPair<WriteBSF, [SKLPort1], 3>;
@@ -605,14 +606,6 @@ def SKLWriteResGroup7 : SchedWriteRes<[S
let ResourceCycles = [1];
}
def: InstRW<[SKLWriteResGroup7], (instrs CDQ, CQO, CLAC, STAC)>;
-def: InstRW<[SKLWriteResGroup7], (instregex "BT(16|32|64)ri8",
- "BT(16|32|64)rr",
- "BTC(16|32|64)ri8",
- "BTC(16|32|64)rr",
- "BTR(16|32|64)ri8",
- "BTR(16|32|64)rr",
- "BTS(16|32|64)ri8",
- "BTS(16|32|64)rr")>;
def SKLWriteResGroup8 : SchedWriteRes<[SKLPort15]> {
let Latency = 1;
Modified: llvm/trunk/lib/Target/X86/X86SchedSkylakeServer.td
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Target/X86/X86SchedSkylakeServer.td?rev=338507&r1=338506&r2=338507&view=diff
==============================================================================
--- llvm/trunk/lib/Target/X86/X86SchedSkylakeServer.td (original)
+++ llvm/trunk/lib/Target/X86/X86SchedSkylakeServer.td Wed Aug 1 03:24:27 2018
@@ -136,6 +136,7 @@ def : WriteRes<WriteSETCCStore, [SKXPor
let NumMicroOps = 3;
}
def : WriteRes<WriteLAHFSAHF, [SKXPort06]>;
+def : WriteRes<WriteBitTest,[SKXPort06]>; //
// Integer shifts and rotates.
defm : SKXWriteResPair<WriteShift, [SKXPort06], 1>;
@@ -618,14 +619,6 @@ def SKXWriteResGroup7 : SchedWriteRes<[S
let ResourceCycles = [1];
}
def: InstRW<[SKXWriteResGroup7], (instrs CDQ, CQO, CLAC, STAC)>;
-def: InstRW<[SKXWriteResGroup7], (instregex "BT(16|32|64)ri8",
- "BT(16|32|64)rr",
- "BTC(16|32|64)ri8",
- "BTC(16|32|64)rr",
- "BTR(16|32|64)ri8",
- "BTR(16|32|64)rr",
- "BTS(16|32|64)ri8",
- "BTS(16|32|64)rr")>;
def SKXWriteResGroup8 : SchedWriteRes<[SKXPort15]> {
let Latency = 1;
Modified: llvm/trunk/lib/Target/X86/X86Schedule.td
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Target/X86/X86Schedule.td?rev=338507&r1=338506&r2=338507&view=diff
==============================================================================
--- llvm/trunk/lib/Target/X86/X86Schedule.td (original)
+++ llvm/trunk/lib/Target/X86/X86Schedule.td Wed Aug 1 03:24:27 2018
@@ -142,6 +142,7 @@ def WriteFCMOV : SchedWrite; // X87 con
def WriteSETCC : SchedWrite; // Set register based on condition code.
def WriteSETCCStore : SchedWrite;
def WriteLAHFSAHF : SchedWrite; // Load/Store flags in AH.
+def WriteBitTest : SchedWrite; // Bit Test - TODO add memory folding support
// Integer shifts and rotates.
defm WriteShift : X86SchedWritePair;
Modified: llvm/trunk/lib/Target/X86/X86ScheduleAtom.td
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Target/X86/X86ScheduleAtom.td?rev=338507&r1=338506&r2=338507&view=diff
==============================================================================
--- llvm/trunk/lib/Target/X86/X86ScheduleAtom.td (original)
+++ llvm/trunk/lib/Target/X86/X86ScheduleAtom.td Wed Aug 1 03:24:27 2018
@@ -108,6 +108,7 @@ def : WriteRes<WriteLAHFSAHF, [AtomPort
let Latency = 2;
let ResourceCycles = [2];
}
+def : WriteRes<WriteBitTest,[AtomPort01]>;
defm : X86WriteResUnsupported<WriteIMulH>;
Modified: llvm/trunk/lib/Target/X86/X86ScheduleBtVer2.td
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Target/X86/X86ScheduleBtVer2.td?rev=338507&r1=338506&r2=338507&view=diff
==============================================================================
--- llvm/trunk/lib/Target/X86/X86ScheduleBtVer2.td (original)
+++ llvm/trunk/lib/Target/X86/X86ScheduleBtVer2.td Wed Aug 1 03:24:27 2018
@@ -188,6 +188,7 @@ defm : X86WriteRes<WriteFCMOV, [JFPU0, J
def : WriteRes<WriteSETCC, [JALU01]>; // Setcc.
def : WriteRes<WriteSETCCStore, [JALU01,JSAGU]>;
def : WriteRes<WriteLAHFSAHF, [JALU01]>;
+def : WriteRes<WriteBitTest,[JALU01]>;
// This is for simple LEAs with one or two input operands.
def : WriteRes<WriteLEA, [JALU01]>;
Modified: llvm/trunk/lib/Target/X86/X86ScheduleSLM.td
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Target/X86/X86ScheduleSLM.td?rev=338507&r1=338506&r2=338507&view=diff
==============================================================================
--- llvm/trunk/lib/Target/X86/X86ScheduleSLM.td (original)
+++ llvm/trunk/lib/Target/X86/X86ScheduleSLM.td Wed Aug 1 03:24:27 2018
@@ -120,6 +120,7 @@ def : WriteRes<WriteSETCCStore, [SLM_IE
let ResourceCycles = [2,1];
}
def : WriteRes<WriteLAHFSAHF, [SLM_IEC_RSV01]>;
+def : WriteRes<WriteBitTest,[SLM_IEC_RSV01]>;
// This is for simple LEAs with one or two input operands.
// The complex ones can only execute on port 1, and they require two cycles on
Modified: llvm/trunk/lib/Target/X86/X86ScheduleZnver1.td
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Target/X86/X86ScheduleZnver1.td?rev=338507&r1=338506&r2=338507&view=diff
==============================================================================
--- llvm/trunk/lib/Target/X86/X86ScheduleZnver1.td (original)
+++ llvm/trunk/lib/Target/X86/X86ScheduleZnver1.td Wed Aug 1 03:24:27 2018
@@ -198,6 +198,7 @@ defm : ZnWriteResPair<WriteCMOV2, [ZnAL
def : WriteRes<WriteSETCC, [ZnALU]>;
def : WriteRes<WriteSETCCStore, [ZnALU, ZnAGU]>;
defm : X86WriteRes<WriteLAHFSAHF, [ZnALU], 2, [1], 2>;
+def : WriteRes<WriteBitTest,[ZnALU]>;
// Bit counts.
defm : ZnWriteResPair<WriteBSF, [ZnALU], 3>;
More information about the llvm-commits
mailing list