RPCS3/llvm-mirror
1
0
Fork 0
mirror of https://github.com/RPCS3/llvm-mirror.git synced 2024-11-01 00:12:50 +01:00
Code Issues Projects Releases Wiki Activity
f92cb857aa
llvm-mirror/test/CodeGen/X86/masked-iv-safe.ll

245 lines
6.5 KiB
LLVM
Raw Normal View History

Eliminate more uses of llvm-as and llvm-dis. llvm-svn: 81290
2009-09-09 01:54:48 +02:00
; RUN: llc < %s -march=x86-64 > %t
Teach getZeroExtendExpr and getSignExtendExpr to use trip-count information to simplify [sz]ext({a,+,b}) to {zext(a),+,[zs]ext(b)}, as appropriate. These functions and the trip count code each call into the other, so this requires careful handling to avoid infinite recursion. During the initial trip count computation, conservative SCEVs are used, which are subsequently discarded once the trip count is actually known. Among other benefits, this change lets LSR automatically eliminate some unnecessary zext-inreg and sext-inreg operation where the operand is an induction variable. llvm-svn: 70241
2009-04-27 22:16:15 +02:00
; RUN: not grep and %t
; RUN: not grep movz %t
; RUN: not grep sar %t
; RUN: not grep shl %t
Factor the code for collecting IV users out of LSR into an IVUsers class, and generalize it so that it can be used by IndVarSimplify. Implement the base IndVarSimplify transformation code using IVUsers. This removes TestOrigIVForWrap and associated code, as ScalarEvolution now has enough builtin overflow detection and folding logic to handle all the same cases, and more. Run "opt -iv-users -analyze -disable-output" on your favorite loop for an example of what IVUsers does. This lets IndVarSimplify eliminate IV casts and compute trip counts in more cases. Also, this happens to finally fix the remaining testcases in PR1301. Now that IndVarSimplify is being more aggressive, it occasionally runs into the problem where ScalarEvolutionExpander's code for avoiding duplicate expansions makes it difficult to ensure that all expanded instructions dominate all the instructions that will use them. As a temporary measure, IndVarSimplify now uses a FixUsesBeforeDefs function to fix up instructions inserted by SCEVExpander. Fortunately, this code is contained, and can be easily removed once a more comprehensive solution is available. llvm-svn: 71535
2009-05-12 04:17:14 +02:00
; RUN: grep add %t | count 2
Revert LoopStrengthReduce.cpp to pre-r94061 for now. llvm-svn: 94123
2010-01-22 01:46:49 +01:00
; RUN: grep inc %t | count 4
Factor the code for collecting IV users out of LSR into an IVUsers class, and generalize it so that it can be used by IndVarSimplify. Implement the base IndVarSimplify transformation code using IVUsers. This removes TestOrigIVForWrap and associated code, as ScalarEvolution now has enough builtin overflow detection and folding logic to handle all the same cases, and more. Run "opt -iv-users -analyze -disable-output" on your favorite loop for an example of what IVUsers does. This lets IndVarSimplify eliminate IV casts and compute trip counts in more cases. Also, this happens to finally fix the remaining testcases in PR1301. Now that IndVarSimplify is being more aggressive, it occasionally runs into the problem where ScalarEvolutionExpander's code for avoiding duplicate expansions makes it difficult to ensure that all expanded instructions dominate all the instructions that will use them. As a temporary measure, IndVarSimplify now uses a FixUsesBeforeDefs function to fix up instructions inserted by SCEVExpander. Fortunately, this code is contained, and can be easily removed once a more comprehensive solution is available. llvm-svn: 71535
2009-05-12 04:17:14 +02:00
; RUN: grep dec %t | count 2
Revert LoopStrengthReduce.cpp to pre-r94061 for now. llvm-svn: 94123
2010-01-22 01:46:49 +01:00
; RUN: grep lea %t | count 2
Teach getZeroExtendExpr and getSignExtendExpr to use trip-count information to simplify [sz]ext({a,+,b}) to {zext(a),+,[zs]ext(b)}, as appropriate. These functions and the trip count code each call into the other, so this requires careful handling to avoid infinite recursion. During the initial trip count computation, conservative SCEVs are used, which are subsequently discarded once the trip count is actually known. Among other benefits, this change lets LSR automatically eliminate some unnecessary zext-inreg and sext-inreg operation where the operand is an induction variable. llvm-svn: 70241
2009-04-27 22:16:15 +02:00
; Optimize away zext-inreg and sext-inreg on the loop induction
; variable using trip-count information.
define void @count_up(double* %d, i64 %n) nounwind {
entry:
br label %loop
loop:
%indvar = phi i64 [ 0, %entry ], [ %indvar.next, %loop ]
%indvar.i8 = and i64 %indvar, 255
%t0 = getelementptr double* %d, i64 %indvar.i8
%t1 = load double* %t0
Split the Add, Sub, and Mul instruction opcodes into separate integer and floating-point opcodes, introducing FAdd, FSub, and FMul. For now, the AsmParser, BitcodeReader, and IRBuilder all preserve backwards compatability, and the Core LLVM APIs preserve backwards compatibility for IR producers. Most front-ends won't need to change immediately. This implements the first step of the plan outlined here: http://nondot.org/sabre/LLVMNotes/IntegerOverflow.txt llvm-svn: 72897
2009-06-05 00:49:04 +02:00
%t2 = fmul double %t1, 0.1
Teach getZeroExtendExpr and getSignExtendExpr to use trip-count information to simplify [sz]ext({a,+,b}) to {zext(a),+,[zs]ext(b)}, as appropriate. These functions and the trip count code each call into the other, so this requires careful handling to avoid infinite recursion. During the initial trip count computation, conservative SCEVs are used, which are subsequently discarded once the trip count is actually known. Among other benefits, this change lets LSR automatically eliminate some unnecessary zext-inreg and sext-inreg operation where the operand is an induction variable. llvm-svn: 70241
2009-04-27 22:16:15 +02:00
store double %t2, double* %t0
%indvar.i24 = and i64 %indvar, 16777215
%t3 = getelementptr double* %d, i64 %indvar.i24
%t4 = load double* %t3
Split the Add, Sub, and Mul instruction opcodes into separate integer and floating-point opcodes, introducing FAdd, FSub, and FMul. For now, the AsmParser, BitcodeReader, and IRBuilder all preserve backwards compatability, and the Core LLVM APIs preserve backwards compatibility for IR producers. Most front-ends won't need to change immediately. This implements the first step of the plan outlined here: http://nondot.org/sabre/LLVMNotes/IntegerOverflow.txt llvm-svn: 72897
2009-06-05 00:49:04 +02:00
%t5 = fmul double %t4, 2.3
Teach getZeroExtendExpr and getSignExtendExpr to use trip-count information to simplify [sz]ext({a,+,b}) to {zext(a),+,[zs]ext(b)}, as appropriate. These functions and the trip count code each call into the other, so this requires careful handling to avoid infinite recursion. During the initial trip count computation, conservative SCEVs are used, which are subsequently discarded once the trip count is actually known. Among other benefits, this change lets LSR automatically eliminate some unnecessary zext-inreg and sext-inreg operation where the operand is an induction variable. llvm-svn: 70241
2009-04-27 22:16:15 +02:00
store double %t5, double* %t3
%t6 = getelementptr double* %d, i64 %indvar
%t7 = load double* %t6
Split the Add, Sub, and Mul instruction opcodes into separate integer and floating-point opcodes, introducing FAdd, FSub, and FMul. For now, the AsmParser, BitcodeReader, and IRBuilder all preserve backwards compatability, and the Core LLVM APIs preserve backwards compatibility for IR producers. Most front-ends won't need to change immediately. This implements the first step of the plan outlined here: http://nondot.org/sabre/LLVMNotes/IntegerOverflow.txt llvm-svn: 72897
2009-06-05 00:49:04 +02:00
%t8 = fmul double %t7, 4.5
Teach getZeroExtendExpr and getSignExtendExpr to use trip-count information to simplify [sz]ext({a,+,b}) to {zext(a),+,[zs]ext(b)}, as appropriate. These functions and the trip count code each call into the other, so this requires careful handling to avoid infinite recursion. During the initial trip count computation, conservative SCEVs are used, which are subsequently discarded once the trip count is actually known. Among other benefits, this change lets LSR automatically eliminate some unnecessary zext-inreg and sext-inreg operation where the operand is an induction variable. llvm-svn: 70241
2009-04-27 22:16:15 +02:00
store double %t8, double* %t6
%indvar.next = add i64 %indvar, 1
%exitcond = icmp eq i64 %indvar.next, 10
br i1 %exitcond, label %return, label %loop
return:
ret void
}
define void @count_down(double* %d, i64 %n) nounwind {
entry:
br label %loop
loop:
%indvar = phi i64 [ 10, %entry ], [ %indvar.next, %loop ]
%indvar.i8 = and i64 %indvar, 255
%t0 = getelementptr double* %d, i64 %indvar.i8
%t1 = load double* %t0
Split the Add, Sub, and Mul instruction opcodes into separate integer and floating-point opcodes, introducing FAdd, FSub, and FMul. For now, the AsmParser, BitcodeReader, and IRBuilder all preserve backwards compatability, and the Core LLVM APIs preserve backwards compatibility for IR producers. Most front-ends won't need to change immediately. This implements the first step of the plan outlined here: http://nondot.org/sabre/LLVMNotes/IntegerOverflow.txt llvm-svn: 72897
2009-06-05 00:49:04 +02:00
%t2 = fmul double %t1, 0.1
Teach getZeroExtendExpr and getSignExtendExpr to use trip-count information to simplify [sz]ext({a,+,b}) to {zext(a),+,[zs]ext(b)}, as appropriate. These functions and the trip count code each call into the other, so this requires careful handling to avoid infinite recursion. During the initial trip count computation, conservative SCEVs are used, which are subsequently discarded once the trip count is actually known. Among other benefits, this change lets LSR automatically eliminate some unnecessary zext-inreg and sext-inreg operation where the operand is an induction variable. llvm-svn: 70241
2009-04-27 22:16:15 +02:00
store double %t2, double* %t0
%indvar.i24 = and i64 %indvar, 16777215
%t3 = getelementptr double* %d, i64 %indvar.i24
%t4 = load double* %t3
Split the Add, Sub, and Mul instruction opcodes into separate integer and floating-point opcodes, introducing FAdd, FSub, and FMul. For now, the AsmParser, BitcodeReader, and IRBuilder all preserve backwards compatability, and the Core LLVM APIs preserve backwards compatibility for IR producers. Most front-ends won't need to change immediately. This implements the first step of the plan outlined here: http://nondot.org/sabre/LLVMNotes/IntegerOverflow.txt llvm-svn: 72897
2009-06-05 00:49:04 +02:00
%t5 = fmul double %t4, 2.3
Teach getZeroExtendExpr and getSignExtendExpr to use trip-count information to simplify [sz]ext({a,+,b}) to {zext(a),+,[zs]ext(b)}, as appropriate. These functions and the trip count code each call into the other, so this requires careful handling to avoid infinite recursion. During the initial trip count computation, conservative SCEVs are used, which are subsequently discarded once the trip count is actually known. Among other benefits, this change lets LSR automatically eliminate some unnecessary zext-inreg and sext-inreg operation where the operand is an induction variable. llvm-svn: 70241
2009-04-27 22:16:15 +02:00
store double %t5, double* %t3
%t6 = getelementptr double* %d, i64 %indvar
%t7 = load double* %t6
Split the Add, Sub, and Mul instruction opcodes into separate integer and floating-point opcodes, introducing FAdd, FSub, and FMul. For now, the AsmParser, BitcodeReader, and IRBuilder all preserve backwards compatability, and the Core LLVM APIs preserve backwards compatibility for IR producers. Most front-ends won't need to change immediately. This implements the first step of the plan outlined here: http://nondot.org/sabre/LLVMNotes/IntegerOverflow.txt llvm-svn: 72897
2009-06-05 00:49:04 +02:00
%t8 = fmul double %t7, 4.5
Teach getZeroExtendExpr and getSignExtendExpr to use trip-count information to simplify [sz]ext({a,+,b}) to {zext(a),+,[zs]ext(b)}, as appropriate. These functions and the trip count code each call into the other, so this requires careful handling to avoid infinite recursion. During the initial trip count computation, conservative SCEVs are used, which are subsequently discarded once the trip count is actually known. Among other benefits, this change lets LSR automatically eliminate some unnecessary zext-inreg and sext-inreg operation where the operand is an induction variable. llvm-svn: 70241
2009-04-27 22:16:15 +02:00
store double %t8, double* %t6
%indvar.next = sub i64 %indvar, 1
%exitcond = icmp eq i64 %indvar.next, 0
br i1 %exitcond, label %return, label %loop
return:
ret void
}
define void @count_up_signed(double* %d, i64 %n) nounwind {
entry:
br label %loop
loop:
%indvar = phi i64 [ 0, %entry ], [ %indvar.next, %loop ]
%s0 = shl i64 %indvar, 8
%indvar.i8 = ashr i64 %s0, 8
%t0 = getelementptr double* %d, i64 %indvar.i8
%t1 = load double* %t0
Split the Add, Sub, and Mul instruction opcodes into separate integer and floating-point opcodes, introducing FAdd, FSub, and FMul. For now, the AsmParser, BitcodeReader, and IRBuilder all preserve backwards compatability, and the Core LLVM APIs preserve backwards compatibility for IR producers. Most front-ends won't need to change immediately. This implements the first step of the plan outlined here: http://nondot.org/sabre/LLVMNotes/IntegerOverflow.txt llvm-svn: 72897
2009-06-05 00:49:04 +02:00
%t2 = fmul double %t1, 0.1
Teach getZeroExtendExpr and getSignExtendExpr to use trip-count information to simplify [sz]ext({a,+,b}) to {zext(a),+,[zs]ext(b)}, as appropriate. These functions and the trip count code each call into the other, so this requires careful handling to avoid infinite recursion. During the initial trip count computation, conservative SCEVs are used, which are subsequently discarded once the trip count is actually known. Among other benefits, this change lets LSR automatically eliminate some unnecessary zext-inreg and sext-inreg operation where the operand is an induction variable. llvm-svn: 70241
2009-04-27 22:16:15 +02:00
store double %t2, double* %t0
%s1 = shl i64 %indvar, 24
%indvar.i24 = ashr i64 %s1, 24
%t3 = getelementptr double* %d, i64 %indvar.i24
%t4 = load double* %t3
Split the Add, Sub, and Mul instruction opcodes into separate integer and floating-point opcodes, introducing FAdd, FSub, and FMul. For now, the AsmParser, BitcodeReader, and IRBuilder all preserve backwards compatability, and the Core LLVM APIs preserve backwards compatibility for IR producers. Most front-ends won't need to change immediately. This implements the first step of the plan outlined here: http://nondot.org/sabre/LLVMNotes/IntegerOverflow.txt llvm-svn: 72897
2009-06-05 00:49:04 +02:00
%t5 = fmul double %t4, 2.3
Teach getZeroExtendExpr and getSignExtendExpr to use trip-count information to simplify [sz]ext({a,+,b}) to {zext(a),+,[zs]ext(b)}, as appropriate. These functions and the trip count code each call into the other, so this requires careful handling to avoid infinite recursion. During the initial trip count computation, conservative SCEVs are used, which are subsequently discarded once the trip count is actually known. Among other benefits, this change lets LSR automatically eliminate some unnecessary zext-inreg and sext-inreg operation where the operand is an induction variable. llvm-svn: 70241
2009-04-27 22:16:15 +02:00
store double %t5, double* %t3
%t6 = getelementptr double* %d, i64 %indvar
%t7 = load double* %t6
Split the Add, Sub, and Mul instruction opcodes into separate integer and floating-point opcodes, introducing FAdd, FSub, and FMul. For now, the AsmParser, BitcodeReader, and IRBuilder all preserve backwards compatability, and the Core LLVM APIs preserve backwards compatibility for IR producers. Most front-ends won't need to change immediately. This implements the first step of the plan outlined here: http://nondot.org/sabre/LLVMNotes/IntegerOverflow.txt llvm-svn: 72897
2009-06-05 00:49:04 +02:00
%t8 = fmul double %t7, 4.5
Teach getZeroExtendExpr and getSignExtendExpr to use trip-count information to simplify [sz]ext({a,+,b}) to {zext(a),+,[zs]ext(b)}, as appropriate. These functions and the trip count code each call into the other, so this requires careful handling to avoid infinite recursion. During the initial trip count computation, conservative SCEVs are used, which are subsequently discarded once the trip count is actually known. Among other benefits, this change lets LSR automatically eliminate some unnecessary zext-inreg and sext-inreg operation where the operand is an induction variable. llvm-svn: 70241
2009-04-27 22:16:15 +02:00
store double %t8, double* %t6
%indvar.next = add i64 %indvar, 1
%exitcond = icmp eq i64 %indvar.next, 10
br i1 %exitcond, label %return, label %loop
return:
ret void
}
define void @count_down_signed(double* %d, i64 %n) nounwind {
entry:
br label %loop
loop:
%indvar = phi i64 [ 10, %entry ], [ %indvar.next, %loop ]
%s0 = shl i64 %indvar, 8
%indvar.i8 = ashr i64 %s0, 8
%t0 = getelementptr double* %d, i64 %indvar.i8
%t1 = load double* %t0
Split the Add, Sub, and Mul instruction opcodes into separate integer and floating-point opcodes, introducing FAdd, FSub, and FMul. For now, the AsmParser, BitcodeReader, and IRBuilder all preserve backwards compatability, and the Core LLVM APIs preserve backwards compatibility for IR producers. Most front-ends won't need to change immediately. This implements the first step of the plan outlined here: http://nondot.org/sabre/LLVMNotes/IntegerOverflow.txt llvm-svn: 72897
2009-06-05 00:49:04 +02:00
%t2 = fmul double %t1, 0.1
Teach getZeroExtendExpr and getSignExtendExpr to use trip-count information to simplify [sz]ext({a,+,b}) to {zext(a),+,[zs]ext(b)}, as appropriate. These functions and the trip count code each call into the other, so this requires careful handling to avoid infinite recursion. During the initial trip count computation, conservative SCEVs are used, which are subsequently discarded once the trip count is actually known. Among other benefits, this change lets LSR automatically eliminate some unnecessary zext-inreg and sext-inreg operation where the operand is an induction variable. llvm-svn: 70241
2009-04-27 22:16:15 +02:00
store double %t2, double* %t0
%s1 = shl i64 %indvar, 24
%indvar.i24 = ashr i64 %s1, 24
%t3 = getelementptr double* %d, i64 %indvar.i24
%t4 = load double* %t3
Split the Add, Sub, and Mul instruction opcodes into separate integer and floating-point opcodes, introducing FAdd, FSub, and FMul. For now, the AsmParser, BitcodeReader, and IRBuilder all preserve backwards compatability, and the Core LLVM APIs preserve backwards compatibility for IR producers. Most front-ends won't need to change immediately. This implements the first step of the plan outlined here: http://nondot.org/sabre/LLVMNotes/IntegerOverflow.txt llvm-svn: 72897
2009-06-05 00:49:04 +02:00
%t5 = fmul double %t4, 2.3
Teach getZeroExtendExpr and getSignExtendExpr to use trip-count information to simplify [sz]ext({a,+,b}) to {zext(a),+,[zs]ext(b)}, as appropriate. These functions and the trip count code each call into the other, so this requires careful handling to avoid infinite recursion. During the initial trip count computation, conservative SCEVs are used, which are subsequently discarded once the trip count is actually known. Among other benefits, this change lets LSR automatically eliminate some unnecessary zext-inreg and sext-inreg operation where the operand is an induction variable. llvm-svn: 70241
2009-04-27 22:16:15 +02:00
store double %t5, double* %t3
%t6 = getelementptr double* %d, i64 %indvar
%t7 = load double* %t6
Split the Add, Sub, and Mul instruction opcodes into separate integer and floating-point opcodes, introducing FAdd, FSub, and FMul. For now, the AsmParser, BitcodeReader, and IRBuilder all preserve backwards compatability, and the Core LLVM APIs preserve backwards compatibility for IR producers. Most front-ends won't need to change immediately. This implements the first step of the plan outlined here: http://nondot.org/sabre/LLVMNotes/IntegerOverflow.txt llvm-svn: 72897
2009-06-05 00:49:04 +02:00
%t8 = fmul double %t7, 4.5
Teach getZeroExtendExpr and getSignExtendExpr to use trip-count information to simplify [sz]ext({a,+,b}) to {zext(a),+,[zs]ext(b)}, as appropriate. These functions and the trip count code each call into the other, so this requires careful handling to avoid infinite recursion. During the initial trip count computation, conservative SCEVs are used, which are subsequently discarded once the trip count is actually known. Among other benefits, this change lets LSR automatically eliminate some unnecessary zext-inreg and sext-inreg operation where the operand is an induction variable. llvm-svn: 70241
2009-04-27 22:16:15 +02:00
store double %t8, double* %t6
%indvar.next = sub i64 %indvar, 1
%exitcond = icmp eq i64 %indvar.next, 0
br i1 %exitcond, label %return, label %loop
return:
ret void
}
define void @another_count_up(double* %d, i64 %n) nounwind {
entry:
br label %loop
loop:
%indvar = phi i64 [ 18446744073709551615, %entry ], [ %indvar.next, %loop ]
%indvar.i8 = and i64 %indvar, 255
%t0 = getelementptr double* %d, i64 %indvar.i8
%t1 = load double* %t0
Split the Add, Sub, and Mul instruction opcodes into separate integer and floating-point opcodes, introducing FAdd, FSub, and FMul. For now, the AsmParser, BitcodeReader, and IRBuilder all preserve backwards compatability, and the Core LLVM APIs preserve backwards compatibility for IR producers. Most front-ends won't need to change immediately. This implements the first step of the plan outlined here: http://nondot.org/sabre/LLVMNotes/IntegerOverflow.txt llvm-svn: 72897
2009-06-05 00:49:04 +02:00
%t2 = fmul double %t1, 0.1
Teach getZeroExtendExpr and getSignExtendExpr to use trip-count information to simplify [sz]ext({a,+,b}) to {zext(a),+,[zs]ext(b)}, as appropriate. These functions and the trip count code each call into the other, so this requires careful handling to avoid infinite recursion. During the initial trip count computation, conservative SCEVs are used, which are subsequently discarded once the trip count is actually known. Among other benefits, this change lets LSR automatically eliminate some unnecessary zext-inreg and sext-inreg operation where the operand is an induction variable. llvm-svn: 70241
2009-04-27 22:16:15 +02:00
store double %t2, double* %t0
%indvar.i24 = and i64 %indvar, 16777215
%t3 = getelementptr double* %d, i64 %indvar.i24
%t4 = load double* %t3
Split the Add, Sub, and Mul instruction opcodes into separate integer and floating-point opcodes, introducing FAdd, FSub, and FMul. For now, the AsmParser, BitcodeReader, and IRBuilder all preserve backwards compatability, and the Core LLVM APIs preserve backwards compatibility for IR producers. Most front-ends won't need to change immediately. This implements the first step of the plan outlined here: http://nondot.org/sabre/LLVMNotes/IntegerOverflow.txt llvm-svn: 72897
2009-06-05 00:49:04 +02:00
%t5 = fmul double %t4, 2.3
Teach getZeroExtendExpr and getSignExtendExpr to use trip-count information to simplify [sz]ext({a,+,b}) to {zext(a),+,[zs]ext(b)}, as appropriate. These functions and the trip count code each call into the other, so this requires careful handling to avoid infinite recursion. During the initial trip count computation, conservative SCEVs are used, which are subsequently discarded once the trip count is actually known. Among other benefits, this change lets LSR automatically eliminate some unnecessary zext-inreg and sext-inreg operation where the operand is an induction variable. llvm-svn: 70241
2009-04-27 22:16:15 +02:00
store double %t5, double* %t3
%t6 = getelementptr double* %d, i64 %indvar
%t7 = load double* %t6
Split the Add, Sub, and Mul instruction opcodes into separate integer and floating-point opcodes, introducing FAdd, FSub, and FMul. For now, the AsmParser, BitcodeReader, and IRBuilder all preserve backwards compatability, and the Core LLVM APIs preserve backwards compatibility for IR producers. Most front-ends won't need to change immediately. This implements the first step of the plan outlined here: http://nondot.org/sabre/LLVMNotes/IntegerOverflow.txt llvm-svn: 72897
2009-06-05 00:49:04 +02:00
%t8 = fmul double %t7, 4.5
Teach getZeroExtendExpr and getSignExtendExpr to use trip-count information to simplify [sz]ext({a,+,b}) to {zext(a),+,[zs]ext(b)}, as appropriate. These functions and the trip count code each call into the other, so this requires careful handling to avoid infinite recursion. During the initial trip count computation, conservative SCEVs are used, which are subsequently discarded once the trip count is actually known. Among other benefits, this change lets LSR automatically eliminate some unnecessary zext-inreg and sext-inreg operation where the operand is an induction variable. llvm-svn: 70241
2009-04-27 22:16:15 +02:00
store double %t8, double* %t6
%indvar.next = add i64 %indvar, 1
%exitcond = icmp eq i64 %indvar.next, 0
br i1 %exitcond, label %return, label %loop
return:
ret void
}
define void @another_count_down(double* %d, i64 %n) nounwind {
entry:
br label %loop
loop:
%indvar = phi i64 [ 0, %entry ], [ %indvar.next, %loop ]
%indvar.i8 = and i64 %indvar, 255
%t0 = getelementptr double* %d, i64 %indvar.i8
%t1 = load double* %t0
Split the Add, Sub, and Mul instruction opcodes into separate integer and floating-point opcodes, introducing FAdd, FSub, and FMul. For now, the AsmParser, BitcodeReader, and IRBuilder all preserve backwards compatability, and the Core LLVM APIs preserve backwards compatibility for IR producers. Most front-ends won't need to change immediately. This implements the first step of the plan outlined here: http://nondot.org/sabre/LLVMNotes/IntegerOverflow.txt llvm-svn: 72897
2009-06-05 00:49:04 +02:00
%t2 = fmul double %t1, 0.1
Teach getZeroExtendExpr and getSignExtendExpr to use trip-count information to simplify [sz]ext({a,+,b}) to {zext(a),+,[zs]ext(b)}, as appropriate. These functions and the trip count code each call into the other, so this requires careful handling to avoid infinite recursion. During the initial trip count computation, conservative SCEVs are used, which are subsequently discarded once the trip count is actually known. Among other benefits, this change lets LSR automatically eliminate some unnecessary zext-inreg and sext-inreg operation where the operand is an induction variable. llvm-svn: 70241
2009-04-27 22:16:15 +02:00
store double %t2, double* %t0
%indvar.i24 = and i64 %indvar, 16777215
%t3 = getelementptr double* %d, i64 %indvar.i24
%t4 = load double* %t3
Reapply the new LoopStrengthReduction code, with compile time and bug fixes, and with improved heuristics for analyzing foreign-loop addrecs. This change also flattens IVUsers, eliminating the stride-oriented groupings, which makes it easier to work with. llvm-svn: 95975
2010-02-12 11:34:29 +01:00
%t5 = fdiv double %t4, 2.3
Teach getZeroExtendExpr and getSignExtendExpr to use trip-count information to simplify [sz]ext({a,+,b}) to {zext(a),+,[zs]ext(b)}, as appropriate. These functions and the trip count code each call into the other, so this requires careful handling to avoid infinite recursion. During the initial trip count computation, conservative SCEVs are used, which are subsequently discarded once the trip count is actually known. Among other benefits, this change lets LSR automatically eliminate some unnecessary zext-inreg and sext-inreg operation where the operand is an induction variable. llvm-svn: 70241
2009-04-27 22:16:15 +02:00
store double %t5, double* %t3
%t6 = getelementptr double* %d, i64 %indvar
%t7 = load double* %t6
Split the Add, Sub, and Mul instruction opcodes into separate integer and floating-point opcodes, introducing FAdd, FSub, and FMul. For now, the AsmParser, BitcodeReader, and IRBuilder all preserve backwards compatability, and the Core LLVM APIs preserve backwards compatibility for IR producers. Most front-ends won't need to change immediately. This implements the first step of the plan outlined here: http://nondot.org/sabre/LLVMNotes/IntegerOverflow.txt llvm-svn: 72897
2009-06-05 00:49:04 +02:00
%t8 = fmul double %t7, 4.5
Teach getZeroExtendExpr and getSignExtendExpr to use trip-count information to simplify [sz]ext({a,+,b}) to {zext(a),+,[zs]ext(b)}, as appropriate. These functions and the trip count code each call into the other, so this requires careful handling to avoid infinite recursion. During the initial trip count computation, conservative SCEVs are used, which are subsequently discarded once the trip count is actually known. Among other benefits, this change lets LSR automatically eliminate some unnecessary zext-inreg and sext-inreg operation where the operand is an induction variable. llvm-svn: 70241
2009-04-27 22:16:15 +02:00
store double %t8, double* %t6
%indvar.next = sub i64 %indvar, 1
%exitcond = icmp eq i64 %indvar.next, 18446744073709551615
br i1 %exitcond, label %return, label %loop
return:
ret void
}
define void @another_count_up_signed(double* %d, i64 %n) nounwind {
entry:
br label %loop
loop:
%indvar = phi i64 [ 18446744073709551615, %entry ], [ %indvar.next, %loop ]
%s0 = shl i64 %indvar, 8
%indvar.i8 = ashr i64 %s0, 8
%t0 = getelementptr double* %d, i64 %indvar.i8
%t1 = load double* %t0
Split the Add, Sub, and Mul instruction opcodes into separate integer and floating-point opcodes, introducing FAdd, FSub, and FMul. For now, the AsmParser, BitcodeReader, and IRBuilder all preserve backwards compatability, and the Core LLVM APIs preserve backwards compatibility for IR producers. Most front-ends won't need to change immediately. This implements the first step of the plan outlined here: http://nondot.org/sabre/LLVMNotes/IntegerOverflow.txt llvm-svn: 72897
2009-06-05 00:49:04 +02:00
%t2 = fmul double %t1, 0.1
Teach getZeroExtendExpr and getSignExtendExpr to use trip-count information to simplify [sz]ext({a,+,b}) to {zext(a),+,[zs]ext(b)}, as appropriate. These functions and the trip count code each call into the other, so this requires careful handling to avoid infinite recursion. During the initial trip count computation, conservative SCEVs are used, which are subsequently discarded once the trip count is actually known. Among other benefits, this change lets LSR automatically eliminate some unnecessary zext-inreg and sext-inreg operation where the operand is an induction variable. llvm-svn: 70241
2009-04-27 22:16:15 +02:00
store double %t2, double* %t0
%s1 = shl i64 %indvar, 24
%indvar.i24 = ashr i64 %s1, 24
%t3 = getelementptr double* %d, i64 %indvar.i24
%t4 = load double* %t3
Reapply the new LoopStrengthReduction code, with compile time and bug fixes, and with improved heuristics for analyzing foreign-loop addrecs. This change also flattens IVUsers, eliminating the stride-oriented groupings, which makes it easier to work with. llvm-svn: 95975
2010-02-12 11:34:29 +01:00
%t5 = fdiv double %t4, 2.3
Teach getZeroExtendExpr and getSignExtendExpr to use trip-count information to simplify [sz]ext({a,+,b}) to {zext(a),+,[zs]ext(b)}, as appropriate. These functions and the trip count code each call into the other, so this requires careful handling to avoid infinite recursion. During the initial trip count computation, conservative SCEVs are used, which are subsequently discarded once the trip count is actually known. Among other benefits, this change lets LSR automatically eliminate some unnecessary zext-inreg and sext-inreg operation where the operand is an induction variable. llvm-svn: 70241
2009-04-27 22:16:15 +02:00
store double %t5, double* %t3
%t6 = getelementptr double* %d, i64 %indvar
%t7 = load double* %t6
Split the Add, Sub, and Mul instruction opcodes into separate integer and floating-point opcodes, introducing FAdd, FSub, and FMul. For now, the AsmParser, BitcodeReader, and IRBuilder all preserve backwards compatability, and the Core LLVM APIs preserve backwards compatibility for IR producers. Most front-ends won't need to change immediately. This implements the first step of the plan outlined here: http://nondot.org/sabre/LLVMNotes/IntegerOverflow.txt llvm-svn: 72897
2009-06-05 00:49:04 +02:00
%t8 = fmul double %t7, 4.5
Teach getZeroExtendExpr and getSignExtendExpr to use trip-count information to simplify [sz]ext({a,+,b}) to {zext(a),+,[zs]ext(b)}, as appropriate. These functions and the trip count code each call into the other, so this requires careful handling to avoid infinite recursion. During the initial trip count computation, conservative SCEVs are used, which are subsequently discarded once the trip count is actually known. Among other benefits, this change lets LSR automatically eliminate some unnecessary zext-inreg and sext-inreg operation where the operand is an induction variable. llvm-svn: 70241
2009-04-27 22:16:15 +02:00
store double %t8, double* %t6
%indvar.next = add i64 %indvar, 1
%exitcond = icmp eq i64 %indvar.next, 0
br i1 %exitcond, label %return, label %loop
return:
ret void
}
define void @another_count_down_signed(double* %d, i64 %n) nounwind {
entry:
br label %loop
loop:
%indvar = phi i64 [ 0, %entry ], [ %indvar.next, %loop ]
%s0 = shl i64 %indvar, 8
%indvar.i8 = ashr i64 %s0, 8
%t0 = getelementptr double* %d, i64 %indvar.i8
%t1 = load double* %t0
Split the Add, Sub, and Mul instruction opcodes into separate integer and floating-point opcodes, introducing FAdd, FSub, and FMul. For now, the AsmParser, BitcodeReader, and IRBuilder all preserve backwards compatability, and the Core LLVM APIs preserve backwards compatibility for IR producers. Most front-ends won't need to change immediately. This implements the first step of the plan outlined here: http://nondot.org/sabre/LLVMNotes/IntegerOverflow.txt llvm-svn: 72897
2009-06-05 00:49:04 +02:00
%t2 = fmul double %t1, 0.1
Teach getZeroExtendExpr and getSignExtendExpr to use trip-count information to simplify [sz]ext({a,+,b}) to {zext(a),+,[zs]ext(b)}, as appropriate. These functions and the trip count code each call into the other, so this requires careful handling to avoid infinite recursion. During the initial trip count computation, conservative SCEVs are used, which are subsequently discarded once the trip count is actually known. Among other benefits, this change lets LSR automatically eliminate some unnecessary zext-inreg and sext-inreg operation where the operand is an induction variable. llvm-svn: 70241
2009-04-27 22:16:15 +02:00
store double %t2, double* %t0
%s1 = shl i64 %indvar, 24
%indvar.i24 = ashr i64 %s1, 24
%t3 = getelementptr double* %d, i64 %indvar.i24
%t4 = load double* %t3
Reapply the new LoopStrengthReduction code, with compile time and bug fixes, and with improved heuristics for analyzing foreign-loop addrecs. This change also flattens IVUsers, eliminating the stride-oriented groupings, which makes it easier to work with. llvm-svn: 95975
2010-02-12 11:34:29 +01:00
%t5 = fdiv double %t4, 2.3
Teach getZeroExtendExpr and getSignExtendExpr to use trip-count information to simplify [sz]ext({a,+,b}) to {zext(a),+,[zs]ext(b)}, as appropriate. These functions and the trip count code each call into the other, so this requires careful handling to avoid infinite recursion. During the initial trip count computation, conservative SCEVs are used, which are subsequently discarded once the trip count is actually known. Among other benefits, this change lets LSR automatically eliminate some unnecessary zext-inreg and sext-inreg operation where the operand is an induction variable. llvm-svn: 70241
2009-04-27 22:16:15 +02:00
store double %t5, double* %t3
%t6 = getelementptr double* %d, i64 %indvar
%t7 = load double* %t6
Split the Add, Sub, and Mul instruction opcodes into separate integer and floating-point opcodes, introducing FAdd, FSub, and FMul. For now, the AsmParser, BitcodeReader, and IRBuilder all preserve backwards compatability, and the Core LLVM APIs preserve backwards compatibility for IR producers. Most front-ends won't need to change immediately. This implements the first step of the plan outlined here: http://nondot.org/sabre/LLVMNotes/IntegerOverflow.txt llvm-svn: 72897
2009-06-05 00:49:04 +02:00
%t8 = fmul double %t7, 4.5
Teach getZeroExtendExpr and getSignExtendExpr to use trip-count information to simplify [sz]ext({a,+,b}) to {zext(a),+,[zs]ext(b)}, as appropriate. These functions and the trip count code each call into the other, so this requires careful handling to avoid infinite recursion. During the initial trip count computation, conservative SCEVs are used, which are subsequently discarded once the trip count is actually known. Among other benefits, this change lets LSR automatically eliminate some unnecessary zext-inreg and sext-inreg operation where the operand is an induction variable. llvm-svn: 70241
2009-04-27 22:16:15 +02:00
store double %t8, double* %t6
%indvar.next = sub i64 %indvar, 1
%exitcond = icmp eq i64 %indvar.next, 18446744073709551615
br i1 %exitcond, label %return, label %loop
return:
ret void
}
Reference in New Issue Copy Permalink