maxjhandsome
/
llvm-project
forked from OSchip/llvm-project
1
0
Fork 0
Code Issues Pull Requests Packages Releases Wiki Activity
llvm-project/llvm/test/CodeGen/ARM/atomic-64bit.ll

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

421 lines
14 KiB
LLVM
Raw Normal View History

ARM big endian function argument passing llvm-svn: 208316
2014-05-08 22:06:24 +08:00
; RUN: llc < %s -mtriple=armv7-apple-ios | FileCheck %s --check-prefix=CHECK --check-prefix=CHECK-LE
; RUN: llc < %s -mtriple=thumbv7-none-linux-gnueabihf | FileCheck %s --check-prefix=CHECK-THUMB --check-prefix=CHECK-THUMB-LE
Add a new string member to the TargetOptions struct for the name of the abi we should be using. For targets that don't use the option there's no change, otherwise this allows external users to set the ABI via string and avoid some of the -backend-option pain in clang. Use this option to move the ABI for the ARM port from the Subtarget to the TargetMachine and update the testcases accordingly since it's no longer valid to set via -mattr. llvm-svn: 224492
2014-12-18 10:20:58 +08:00
; RUN: llc < %s -mtriple=armebv7 -target-abi apcs | FileCheck %s --check-prefix=CHECK --check-prefix=CHECK-BE
ARM big endian function argument passing llvm-svn: 208316
2014-05-08 22:06:24 +08:00
; RUN: llc < %s -mtriple=thumbebv7-none-linux-gnueabihf | FileCheck %s --check-prefix=CHECK-THUMB --check-prefix=CHECK-THUMB-BE
Some 64-bit atomic operations on ARM. 64-bit cmpxchg coming next. llvm-svn: 138845
2011-08-31 08:31:29 +08:00
define i64 @test1(i64* %ptr, i64 %val) {
Convert CodeGen/*/*.ll tests to use the new CHECK-LABEL for easier debugging. No functionality change and all tests pass after conversion. This was done with the following sed invocation to catch label lines demarking function boundaries: sed -i '' "s/^;\( *\)\([A-Z0-9_]*\):\( *\)test\([A-Za-z0-9_-]*\):\( *\)$/;\1\2-LABEL:\3test\4:\5/g" test/CodeGen/*/*.ll which was written conservatively to avoid false positives rather than false negatives. I scanned through all the changes and everything looks correct. llvm-svn: 186258
2013-07-14 04:38:47 +08:00
; CHECK-LABEL: test1:
ARM: relax the atomic release barrier to "dmb ishst" on Swift Swift cores implement store barriers that are stronger than the ARM specification but weaker than general barriers. They are, in fact, just about enough to provide the ordering needed for atomic operations with release semantics. This patch makes use of that quirk. llvm-svn: 185527
2013-07-03 17:20:36 +08:00
; CHECK: dmb {{ish$}}
Remove hard coded registers in ARM ldrexd and strexd instructions This patch replaces the hard coded GPR pair [R0, R1] of Intrinsic:arm_ldrexd and [R2, R3] of Intrinsic:arm_strexd with even/odd GPRPair reg class. Similar to the lowering of atomic_64 operation. llvm-svn: 168207
2012-11-17 05:55:34 +08:00
; CHECK: ldrexd [[REG1:(r[0-9]?[02468])]], [[REG2:(r[0-9]?[13579])]]
ARM big endian function argument passing llvm-svn: 208316
2014-05-08 22:06:24 +08:00
; CHECK-LE: adds [[REG3:(r[0-9]?[02468])]], [[REG1]]
; CHECK-LE: adc [[REG4:(r[0-9]?[13579])]], [[REG2]]
; CHECK-BE: adds [[REG4:(r[0-9]?[13579])]], [[REG2]]
; CHECK-BE: adc [[REG3:(r[0-9]?[02468])]], [[REG1]]
Remove hard coded registers in ARM ldrexd and strexd instructions This patch replaces the hard coded GPR pair [R0, R1] of Intrinsic:arm_ldrexd and [R2, R3] of Intrinsic:arm_strexd with even/odd GPRPair reg class. Similar to the lowering of atomic_64 operation. llvm-svn: 168207
2012-11-17 05:55:34 +08:00
; CHECK: strexd {{[a-z0-9]+}}, [[REG3]], [[REG4]]
Some 64-bit atomic operations on ARM. 64-bit cmpxchg coming next. llvm-svn: 138845
2011-08-31 08:31:29 +08:00
; CHECK: cmp
; CHECK: bne
ARM: relax the atomic release barrier to "dmb ishst" on Swift Swift cores implement store barriers that are stronger than the ARM specification but weaker than general barriers. They are, in fact, just about enough to provide the ordering needed for atomic operations with release semantics. This patch makes use of that quirk. llvm-svn: 185527
2013-07-03 17:20:36 +08:00
; CHECK: dmb {{ish$}}
Fix 64-bit atomic operations in Thumb mode. The ARM and Thumb variants of LDREXD and STREXD have different constraints and take different operands. Previously the code expanding atomic operations didn't take this into account and asserted in Thumb mode. llvm-svn: 173780
2013-01-29 17:06:13 +08:00
Mass update to CodeGen tests to use CHECK-LABEL for labels corresponding to function definitions for more informative error messages. No functionality change and all updated tests passed locally. This update was done with the following bash script: find test/CodeGen -name "*.ll" | \ while read NAME; do echo "$NAME" if ! grep -q "^; *RUN: *llc.*debug" $NAME; then TEMP=`mktemp -t temp` cp $NAME $TEMP sed -n "s/^define [^@]*@\([A-Za-z0-9_]*\)(.*$/\1/p" < $NAME | \ while read FUNC; do sed -i '' "s/;\(.*\)\([A-Za-z0-9_-]*\):\( *\)$FUNC: *\$/;\1\2-LABEL:\3$FUNC:/g" $TEMP done sed -i '' "s/;\(.*\)-LABEL-LABEL:/;\1-LABEL:/" $TEMP sed -i '' "s/;\(.*\)-NEXT-LABEL:/;\1-NEXT:/" $TEMP sed -i '' "s/;\(.*\)-NOT-LABEL:/;\1-NOT:/" $TEMP sed -i '' "s/;\(.*\)-DAG-LABEL:/;\1-DAG:/" $TEMP mv $TEMP $NAME fi done llvm-svn: 186280
2013-07-14 14:24:09 +08:00
; CHECK-THUMB-LABEL: test1:
ARM: relax the atomic release barrier to "dmb ishst" on Swift Swift cores implement store barriers that are stronger than the ARM specification but weaker than general barriers. They are, in fact, just about enough to provide the ordering needed for atomic operations with release semantics. This patch makes use of that quirk. llvm-svn: 185527
2013-07-03 17:20:36 +08:00
; CHECK-THUMB: dmb {{ish$}}
Fix 64-bit atomic operations in Thumb mode. The ARM and Thumb variants of LDREXD and STREXD have different constraints and take different operands. Previously the code expanding atomic operations didn't take this into account and asserted in Thumb mode. llvm-svn: 173780
2013-01-29 17:06:13 +08:00
; CHECK-THUMB: ldrexd [[REG1:[a-z0-9]+]], [[REG2:[a-z0-9]+]]
ARM big endian function argument passing llvm-svn: 208316
2014-05-08 22:06:24 +08:00
; CHECK-THUMB-LE: adds.w [[REG3:[a-z0-9]+]], [[REG1]]
; CHECK-THUMB-LE: adc.w [[REG4:[a-z0-9]+]], [[REG2]]
; CHECK-THUMB-BE: adds.w [[REG4:[a-z0-9]+]], [[REG2]]
; CHECK-THUMB-BE: adc.w [[REG3:[a-z0-9]+]], [[REG1]]
Fix 64-bit atomic operations in Thumb mode. The ARM and Thumb variants of LDREXD and STREXD have different constraints and take different operands. Previously the code expanding atomic operations didn't take this into account and asserted in Thumb mode. llvm-svn: 173780
2013-01-29 17:06:13 +08:00
; CHECK-THUMB: strexd {{[a-z0-9]+}}, [[REG3]], [[REG4]]
; CHECK-THUMB: cmp
; CHECK-THUMB: bne
ARM: relax the atomic release barrier to "dmb ishst" on Swift Swift cores implement store barriers that are stronger than the ARM specification but weaker than general barriers. They are, in fact, just about enough to provide the ordering needed for atomic operations with release semantics. This patch makes use of that quirk. llvm-svn: 185527
2013-07-03 17:20:36 +08:00
; CHECK-THUMB: dmb {{ish$}}
Fix 64-bit atomic operations in Thumb mode. The ARM and Thumb variants of LDREXD and STREXD have different constraints and take different operands. Previously the code expanding atomic operations didn't take this into account and asserted in Thumb mode. llvm-svn: 173780
2013-01-29 17:06:13 +08:00
Some 64-bit atomic operations on ARM. 64-bit cmpxchg coming next. llvm-svn: 138845
2011-08-31 08:31:29 +08:00
%r = atomicrmw add i64* %ptr, i64 %val seq_cst
ret i64 %r
}
define i64 @test2(i64* %ptr, i64 %val) {
Convert CodeGen/*/*.ll tests to use the new CHECK-LABEL for easier debugging. No functionality change and all tests pass after conversion. This was done with the following sed invocation to catch label lines demarking function boundaries: sed -i '' "s/^;\( *\)\([A-Z0-9_]*\):\( *\)test\([A-Za-z0-9_-]*\):\( *\)$/;\1\2-LABEL:\3test\4:\5/g" test/CodeGen/*/*.ll which was written conservatively to avoid false positives rather than false negatives. I scanned through all the changes and everything looks correct. llvm-svn: 186258
2013-07-14 04:38:47 +08:00
; CHECK-LABEL: test2:
ARM: relax the atomic release barrier to "dmb ishst" on Swift Swift cores implement store barriers that are stronger than the ARM specification but weaker than general barriers. They are, in fact, just about enough to provide the ordering needed for atomic operations with release semantics. This patch makes use of that quirk. llvm-svn: 185527
2013-07-03 17:20:36 +08:00
; CHECK: dmb {{ish$}}
Remove hard coded registers in ARM ldrexd and strexd instructions This patch replaces the hard coded GPR pair [R0, R1] of Intrinsic:arm_ldrexd and [R2, R3] of Intrinsic:arm_strexd with even/odd GPRPair reg class. Similar to the lowering of atomic_64 operation. llvm-svn: 168207
2012-11-17 05:55:34 +08:00
; CHECK: ldrexd [[REG1:(r[0-9]?[02468])]], [[REG2:(r[0-9]?[13579])]]
ARM big endian function argument passing llvm-svn: 208316
2014-05-08 22:06:24 +08:00
; CHECK-LE: subs [[REG3:(r[0-9]?[02468])]], [[REG1]]
; CHECK-LE: sbc [[REG4:(r[0-9]?[13579])]], [[REG2]]
; CHECK-BE: subs [[REG4:(r[0-9]?[13579])]], [[REG2]]
; CHECK-BE: sbc [[REG3:(r[0-9]?[02468])]], [[REG1]]
Remove hard coded registers in ARM ldrexd and strexd instructions This patch replaces the hard coded GPR pair [R0, R1] of Intrinsic:arm_ldrexd and [R2, R3] of Intrinsic:arm_strexd with even/odd GPRPair reg class. Similar to the lowering of atomic_64 operation. llvm-svn: 168207
2012-11-17 05:55:34 +08:00
; CHECK: strexd {{[a-z0-9]+}}, [[REG3]], [[REG4]]
Some 64-bit atomic operations on ARM. 64-bit cmpxchg coming next. llvm-svn: 138845
2011-08-31 08:31:29 +08:00
; CHECK: cmp
; CHECK: bne
ARM: relax the atomic release barrier to "dmb ishst" on Swift Swift cores implement store barriers that are stronger than the ARM specification but weaker than general barriers. They are, in fact, just about enough to provide the ordering needed for atomic operations with release semantics. This patch makes use of that quirk. llvm-svn: 185527
2013-07-03 17:20:36 +08:00
; CHECK: dmb {{ish$}}
Fix 64-bit atomic operations in Thumb mode. The ARM and Thumb variants of LDREXD and STREXD have different constraints and take different operands. Previously the code expanding atomic operations didn't take this into account and asserted in Thumb mode. llvm-svn: 173780
2013-01-29 17:06:13 +08:00
Mass update to CodeGen tests to use CHECK-LABEL for labels corresponding to function definitions for more informative error messages. No functionality change and all updated tests passed locally. This update was done with the following bash script: find test/CodeGen -name "*.ll" | \ while read NAME; do echo "$NAME" if ! grep -q "^; *RUN: *llc.*debug" $NAME; then TEMP=`mktemp -t temp` cp $NAME $TEMP sed -n "s/^define [^@]*@\([A-Za-z0-9_]*\)(.*$/\1/p" < $NAME | \ while read FUNC; do sed -i '' "s/;\(.*\)\([A-Za-z0-9_-]*\):\( *\)$FUNC: *\$/;\1\2-LABEL:\3$FUNC:/g" $TEMP done sed -i '' "s/;\(.*\)-LABEL-LABEL:/;\1-LABEL:/" $TEMP sed -i '' "s/;\(.*\)-NEXT-LABEL:/;\1-NEXT:/" $TEMP sed -i '' "s/;\(.*\)-NOT-LABEL:/;\1-NOT:/" $TEMP sed -i '' "s/;\(.*\)-DAG-LABEL:/;\1-DAG:/" $TEMP mv $TEMP $NAME fi done llvm-svn: 186280
2013-07-14 14:24:09 +08:00
; CHECK-THUMB-LABEL: test2:
ARM: relax the atomic release barrier to "dmb ishst" on Swift Swift cores implement store barriers that are stronger than the ARM specification but weaker than general barriers. They are, in fact, just about enough to provide the ordering needed for atomic operations with release semantics. This patch makes use of that quirk. llvm-svn: 185527
2013-07-03 17:20:36 +08:00
; CHECK-THUMB: dmb {{ish$}}
Fix 64-bit atomic operations in Thumb mode. The ARM and Thumb variants of LDREXD and STREXD have different constraints and take different operands. Previously the code expanding atomic operations didn't take this into account and asserted in Thumb mode. llvm-svn: 173780
2013-01-29 17:06:13 +08:00
; CHECK-THUMB: ldrexd [[REG1:[a-z0-9]+]], [[REG2:[a-z0-9]+]]
ARM big endian function argument passing llvm-svn: 208316
2014-05-08 22:06:24 +08:00
; CHECK-THUMB-LE: subs.w [[REG3:[a-z0-9]+]], [[REG1]]
; CHECK-THUMB-LE: sbc.w [[REG4:[a-z0-9]+]], [[REG2]]
; CHECK-THUMB-BE: subs.w [[REG4:[a-z0-9]+]], [[REG2]]
; CHECK-THUMB-BE: sbc.w [[REG3:[a-z0-9]+]], [[REG1]]
Fix 64-bit atomic operations in Thumb mode. The ARM and Thumb variants of LDREXD and STREXD have different constraints and take different operands. Previously the code expanding atomic operations didn't take this into account and asserted in Thumb mode. llvm-svn: 173780
2013-01-29 17:06:13 +08:00
; CHECK-THUMB: strexd {{[a-z0-9]+}}, [[REG3]], [[REG4]]
; CHECK-THUMB: cmp
; CHECK-THUMB: bne
ARM: relax the atomic release barrier to "dmb ishst" on Swift Swift cores implement store barriers that are stronger than the ARM specification but weaker than general barriers. They are, in fact, just about enough to provide the ordering needed for atomic operations with release semantics. This patch makes use of that quirk. llvm-svn: 185527
2013-07-03 17:20:36 +08:00
; CHECK-THUMB: dmb {{ish$}}
Fix 64-bit atomic operations in Thumb mode. The ARM and Thumb variants of LDREXD and STREXD have different constraints and take different operands. Previously the code expanding atomic operations didn't take this into account and asserted in Thumb mode. llvm-svn: 173780
2013-01-29 17:06:13 +08:00
Some 64-bit atomic operations on ARM. 64-bit cmpxchg coming next. llvm-svn: 138845
2011-08-31 08:31:29 +08:00
%r = atomicrmw sub i64* %ptr, i64 %val seq_cst
ret i64 %r
}
define i64 @test3(i64* %ptr, i64 %val) {
Convert CodeGen/*/*.ll tests to use the new CHECK-LABEL for easier debugging. No functionality change and all tests pass after conversion. This was done with the following sed invocation to catch label lines demarking function boundaries: sed -i '' "s/^;\( *\)\([A-Z0-9_]*\):\( *\)test\([A-Za-z0-9_-]*\):\( *\)$/;\1\2-LABEL:\3test\4:\5/g" test/CodeGen/*/*.ll which was written conservatively to avoid false positives rather than false negatives. I scanned through all the changes and everything looks correct. llvm-svn: 186258
2013-07-14 04:38:47 +08:00
; CHECK-LABEL: test3:
ARM: relax the atomic release barrier to "dmb ishst" on Swift Swift cores implement store barriers that are stronger than the ARM specification but weaker than general barriers. They are, in fact, just about enough to provide the ordering needed for atomic operations with release semantics. This patch makes use of that quirk. llvm-svn: 185527
2013-07-03 17:20:36 +08:00
; CHECK: dmb {{ish$}}
Remove hard coded registers in ARM ldrexd and strexd instructions This patch replaces the hard coded GPR pair [R0, R1] of Intrinsic:arm_ldrexd and [R2, R3] of Intrinsic:arm_strexd with even/odd GPRPair reg class. Similar to the lowering of atomic_64 operation. llvm-svn: 168207
2012-11-17 05:55:34 +08:00
; CHECK: ldrexd [[REG1:(r[0-9]?[02468])]], [[REG2:(r[0-9]?[13579])]]
ARM big endian function argument passing llvm-svn: 208316
2014-05-08 22:06:24 +08:00
; CHECK-LE-DAG: and [[REG3:(r[0-9]?[02468])]], [[REG1]]
; CHECK-LE-DAG: and [[REG4:(r[0-9]?[13579])]], [[REG2]]
; CHECK-BE-DAG: and [[REG4:(r[0-9]?[13579])]], [[REG2]]
; CHECK-BE-DAG: and [[REG3:(r[0-9]?[02468])]], [[REG1]]
Remove hard coded registers in ARM ldrexd and strexd instructions This patch replaces the hard coded GPR pair [R0, R1] of Intrinsic:arm_ldrexd and [R2, R3] of Intrinsic:arm_strexd with even/odd GPRPair reg class. Similar to the lowering of atomic_64 operation. llvm-svn: 168207
2012-11-17 05:55:34 +08:00
; CHECK: strexd {{[a-z0-9]+}}, [[REG3]], [[REG4]]
Some 64-bit atomic operations on ARM. 64-bit cmpxchg coming next. llvm-svn: 138845
2011-08-31 08:31:29 +08:00
; CHECK: cmp
; CHECK: bne
ARM: relax the atomic release barrier to "dmb ishst" on Swift Swift cores implement store barriers that are stronger than the ARM specification but weaker than general barriers. They are, in fact, just about enough to provide the ordering needed for atomic operations with release semantics. This patch makes use of that quirk. llvm-svn: 185527
2013-07-03 17:20:36 +08:00
; CHECK: dmb {{ish$}}
Fix 64-bit atomic operations in Thumb mode. The ARM and Thumb variants of LDREXD and STREXD have different constraints and take different operands. Previously the code expanding atomic operations didn't take this into account and asserted in Thumb mode. llvm-svn: 173780
2013-01-29 17:06:13 +08:00
Mass update to CodeGen tests to use CHECK-LABEL for labels corresponding to function definitions for more informative error messages. No functionality change and all updated tests passed locally. This update was done with the following bash script: find test/CodeGen -name "*.ll" | \ while read NAME; do echo "$NAME" if ! grep -q "^; *RUN: *llc.*debug" $NAME; then TEMP=`mktemp -t temp` cp $NAME $TEMP sed -n "s/^define [^@]*@\([A-Za-z0-9_]*\)(.*$/\1/p" < $NAME | \ while read FUNC; do sed -i '' "s/;\(.*\)\([A-Za-z0-9_-]*\):\( *\)$FUNC: *\$/;\1\2-LABEL:\3$FUNC:/g" $TEMP done sed -i '' "s/;\(.*\)-LABEL-LABEL:/;\1-LABEL:/" $TEMP sed -i '' "s/;\(.*\)-NEXT-LABEL:/;\1-NEXT:/" $TEMP sed -i '' "s/;\(.*\)-NOT-LABEL:/;\1-NOT:/" $TEMP sed -i '' "s/;\(.*\)-DAG-LABEL:/;\1-DAG:/" $TEMP mv $TEMP $NAME fi done llvm-svn: 186280
2013-07-14 14:24:09 +08:00
; CHECK-THUMB-LABEL: test3:
ARM: relax the atomic release barrier to "dmb ishst" on Swift Swift cores implement store barriers that are stronger than the ARM specification but weaker than general barriers. They are, in fact, just about enough to provide the ordering needed for atomic operations with release semantics. This patch makes use of that quirk. llvm-svn: 185527
2013-07-03 17:20:36 +08:00
; CHECK-THUMB: dmb {{ish$}}
Fix 64-bit atomic operations in Thumb mode. The ARM and Thumb variants of LDREXD and STREXD have different constraints and take different operands. Previously the code expanding atomic operations didn't take this into account and asserted in Thumb mode. llvm-svn: 173780
2013-01-29 17:06:13 +08:00
; CHECK-THUMB: ldrexd [[REG1:[a-z0-9]+]], [[REG2:[a-z0-9]+]]
ARM big endian function argument passing llvm-svn: 208316
2014-05-08 22:06:24 +08:00
; CHECK-THUMB-LE-DAG: and.w [[REG3:[a-z0-9]+]], [[REG1]]
; CHECK-THUMB-LE-DAG: and.w [[REG4:[a-z0-9]+]], [[REG2]]
; CHECK-THUMB-BE-DAG: and.w [[REG4:[a-z0-9]+]], [[REG2]]
; CHECK-THUMB-BE-DAG: and.w [[REG3:[a-z0-9]+]], [[REG1]]
Fix 64-bit atomic operations in Thumb mode. The ARM and Thumb variants of LDREXD and STREXD have different constraints and take different operands. Previously the code expanding atomic operations didn't take this into account and asserted in Thumb mode. llvm-svn: 173780
2013-01-29 17:06:13 +08:00
; CHECK-THUMB: strexd {{[a-z0-9]+}}, [[REG3]], [[REG4]]
; CHECK-THUMB: cmp
; CHECK-THUMB: bne
ARM: relax the atomic release barrier to "dmb ishst" on Swift Swift cores implement store barriers that are stronger than the ARM specification but weaker than general barriers. They are, in fact, just about enough to provide the ordering needed for atomic operations with release semantics. This patch makes use of that quirk. llvm-svn: 185527
2013-07-03 17:20:36 +08:00
; CHECK-THUMB: dmb {{ish$}}
Fix 64-bit atomic operations in Thumb mode. The ARM and Thumb variants of LDREXD and STREXD have different constraints and take different operands. Previously the code expanding atomic operations didn't take this into account and asserted in Thumb mode. llvm-svn: 173780
2013-01-29 17:06:13 +08:00
Some 64-bit atomic operations on ARM. 64-bit cmpxchg coming next. llvm-svn: 138845
2011-08-31 08:31:29 +08:00
%r = atomicrmw and i64* %ptr, i64 %val seq_cst
ret i64 %r
}
define i64 @test4(i64* %ptr, i64 %val) {
Convert CodeGen/*/*.ll tests to use the new CHECK-LABEL for easier debugging. No functionality change and all tests pass after conversion. This was done with the following sed invocation to catch label lines demarking function boundaries: sed -i '' "s/^;\( *\)\([A-Z0-9_]*\):\( *\)test\([A-Za-z0-9_-]*\):\( *\)$/;\1\2-LABEL:\3test\4:\5/g" test/CodeGen/*/*.ll which was written conservatively to avoid false positives rather than false negatives. I scanned through all the changes and everything looks correct. llvm-svn: 186258
2013-07-14 04:38:47 +08:00
; CHECK-LABEL: test4:
ARM: relax the atomic release barrier to "dmb ishst" on Swift Swift cores implement store barriers that are stronger than the ARM specification but weaker than general barriers. They are, in fact, just about enough to provide the ordering needed for atomic operations with release semantics. This patch makes use of that quirk. llvm-svn: 185527
2013-07-03 17:20:36 +08:00
; CHECK: dmb {{ish$}}
Remove hard coded registers in ARM ldrexd and strexd instructions This patch replaces the hard coded GPR pair [R0, R1] of Intrinsic:arm_ldrexd and [R2, R3] of Intrinsic:arm_strexd with even/odd GPRPair reg class. Similar to the lowering of atomic_64 operation. llvm-svn: 168207
2012-11-17 05:55:34 +08:00
; CHECK: ldrexd [[REG1:(r[0-9]?[02468])]], [[REG2:(r[0-9]?[13579])]]
ARM big endian function argument passing llvm-svn: 208316
2014-05-08 22:06:24 +08:00
; CHECK-LE-DAG: orr [[REG3:(r[0-9]?[02468])]], [[REG1]]
; CHECK-LE-DAG: orr [[REG4:(r[0-9]?[13579])]], [[REG2]]
; CHECK-BE-DAG: orr [[REG4:(r[0-9]?[13579])]], [[REG2]]
; CHECK-BE-DAG: orr [[REG3:(r[0-9]?[02468])]], [[REG1]]
Remove hard coded registers in ARM ldrexd and strexd instructions This patch replaces the hard coded GPR pair [R0, R1] of Intrinsic:arm_ldrexd and [R2, R3] of Intrinsic:arm_strexd with even/odd GPRPair reg class. Similar to the lowering of atomic_64 operation. llvm-svn: 168207
2012-11-17 05:55:34 +08:00
; CHECK: strexd {{[a-z0-9]+}}, [[REG3]], [[REG4]]
Some 64-bit atomic operations on ARM. 64-bit cmpxchg coming next. llvm-svn: 138845
2011-08-31 08:31:29 +08:00
; CHECK: cmp
; CHECK: bne
ARM: relax the atomic release barrier to "dmb ishst" on Swift Swift cores implement store barriers that are stronger than the ARM specification but weaker than general barriers. They are, in fact, just about enough to provide the ordering needed for atomic operations with release semantics. This patch makes use of that quirk. llvm-svn: 185527
2013-07-03 17:20:36 +08:00
; CHECK: dmb {{ish$}}
Fix 64-bit atomic operations in Thumb mode. The ARM and Thumb variants of LDREXD and STREXD have different constraints and take different operands. Previously the code expanding atomic operations didn't take this into account and asserted in Thumb mode. llvm-svn: 173780
2013-01-29 17:06:13 +08:00
Mass update to CodeGen tests to use CHECK-LABEL for labels corresponding to function definitions for more informative error messages. No functionality change and all updated tests passed locally. This update was done with the following bash script: find test/CodeGen -name "*.ll" | \ while read NAME; do echo "$NAME" if ! grep -q "^; *RUN: *llc.*debug" $NAME; then TEMP=`mktemp -t temp` cp $NAME $TEMP sed -n "s/^define [^@]*@\([A-Za-z0-9_]*\)(.*$/\1/p" < $NAME | \ while read FUNC; do sed -i '' "s/;\(.*\)\([A-Za-z0-9_-]*\):\( *\)$FUNC: *\$/;\1\2-LABEL:\3$FUNC:/g" $TEMP done sed -i '' "s/;\(.*\)-LABEL-LABEL:/;\1-LABEL:/" $TEMP sed -i '' "s/;\(.*\)-NEXT-LABEL:/;\1-NEXT:/" $TEMP sed -i '' "s/;\(.*\)-NOT-LABEL:/;\1-NOT:/" $TEMP sed -i '' "s/;\(.*\)-DAG-LABEL:/;\1-DAG:/" $TEMP mv $TEMP $NAME fi done llvm-svn: 186280
2013-07-14 14:24:09 +08:00
; CHECK-THUMB-LABEL: test4:
ARM: relax the atomic release barrier to "dmb ishst" on Swift Swift cores implement store barriers that are stronger than the ARM specification but weaker than general barriers. They are, in fact, just about enough to provide the ordering needed for atomic operations with release semantics. This patch makes use of that quirk. llvm-svn: 185527
2013-07-03 17:20:36 +08:00
; CHECK-THUMB: dmb {{ish$}}
Fix 64-bit atomic operations in Thumb mode. The ARM and Thumb variants of LDREXD and STREXD have different constraints and take different operands. Previously the code expanding atomic operations didn't take this into account and asserted in Thumb mode. llvm-svn: 173780
2013-01-29 17:06:13 +08:00
; CHECK-THUMB: ldrexd [[REG1:[a-z0-9]+]], [[REG2:[a-z0-9]+]]
ARM big endian function argument passing llvm-svn: 208316
2014-05-08 22:06:24 +08:00
; CHECK-THUMB-LE-DAG: orr.w [[REG3:[a-z0-9]+]], [[REG1]]
; CHECK-THUMB-LE-DAG: orr.w [[REG4:[a-z0-9]+]], [[REG2]]
; CHECK-THUMB-BE-DAG: orr.w [[REG4:[a-z0-9]+]], [[REG2]]
; CHECK-THUMB-BE-DAG: orr.w [[REG3:[a-z0-9]+]], [[REG1]]
Fix 64-bit atomic operations in Thumb mode. The ARM and Thumb variants of LDREXD and STREXD have different constraints and take different operands. Previously the code expanding atomic operations didn't take this into account and asserted in Thumb mode. llvm-svn: 173780
2013-01-29 17:06:13 +08:00
; CHECK-THUMB: strexd {{[a-z0-9]+}}, [[REG3]], [[REG4]]
; CHECK-THUMB: cmp
; CHECK-THUMB: bne
ARM: relax the atomic release barrier to "dmb ishst" on Swift Swift cores implement store barriers that are stronger than the ARM specification but weaker than general barriers. They are, in fact, just about enough to provide the ordering needed for atomic operations with release semantics. This patch makes use of that quirk. llvm-svn: 185527
2013-07-03 17:20:36 +08:00
; CHECK-THUMB: dmb {{ish$}}
Fix 64-bit atomic operations in Thumb mode. The ARM and Thumb variants of LDREXD and STREXD have different constraints and take different operands. Previously the code expanding atomic operations didn't take this into account and asserted in Thumb mode. llvm-svn: 173780
2013-01-29 17:06:13 +08:00
Some 64-bit atomic operations on ARM. 64-bit cmpxchg coming next. llvm-svn: 138845
2011-08-31 08:31:29 +08:00
%r = atomicrmw or i64* %ptr, i64 %val seq_cst
ret i64 %r
}
define i64 @test5(i64* %ptr, i64 %val) {
Convert CodeGen/*/*.ll tests to use the new CHECK-LABEL for easier debugging. No functionality change and all tests pass after conversion. This was done with the following sed invocation to catch label lines demarking function boundaries: sed -i '' "s/^;\( *\)\([A-Z0-9_]*\):\( *\)test\([A-Za-z0-9_-]*\):\( *\)$/;\1\2-LABEL:\3test\4:\5/g" test/CodeGen/*/*.ll which was written conservatively to avoid false positives rather than false negatives. I scanned through all the changes and everything looks correct. llvm-svn: 186258
2013-07-14 04:38:47 +08:00
; CHECK-LABEL: test5:
ARM: relax the atomic release barrier to "dmb ishst" on Swift Swift cores implement store barriers that are stronger than the ARM specification but weaker than general barriers. They are, in fact, just about enough to provide the ordering needed for atomic operations with release semantics. This patch makes use of that quirk. llvm-svn: 185527
2013-07-03 17:20:36 +08:00
; CHECK: dmb {{ish$}}
Remove hard coded registers in ARM ldrexd and strexd instructions This patch replaces the hard coded GPR pair [R0, R1] of Intrinsic:arm_ldrexd and [R2, R3] of Intrinsic:arm_strexd with even/odd GPRPair reg class. Similar to the lowering of atomic_64 operation. llvm-svn: 168207
2012-11-17 05:55:34 +08:00
; CHECK: ldrexd [[REG1:(r[0-9]?[02468])]], [[REG2:(r[0-9]?[13579])]]
ARM big endian function argument passing llvm-svn: 208316
2014-05-08 22:06:24 +08:00
; CHECK-LE-DAG: eor [[REG3:(r[0-9]?[02468])]], [[REG1]]
; CHECK-LE-DAG: eor [[REG4:(r[0-9]?[13579])]], [[REG2]]
; CHECK-BE-DAG: eor [[REG4:(r[0-9]?[13579])]], [[REG2]]
; CHECK-BE-DAG: eor [[REG3:(r[0-9]?[02468])]], [[REG1]]
Remove hard coded registers in ARM ldrexd and strexd instructions This patch replaces the hard coded GPR pair [R0, R1] of Intrinsic:arm_ldrexd and [R2, R3] of Intrinsic:arm_strexd with even/odd GPRPair reg class. Similar to the lowering of atomic_64 operation. llvm-svn: 168207
2012-11-17 05:55:34 +08:00
; CHECK: strexd {{[a-z0-9]+}}, [[REG3]], [[REG4]]
Some 64-bit atomic operations on ARM. 64-bit cmpxchg coming next. llvm-svn: 138845
2011-08-31 08:31:29 +08:00
; CHECK: cmp
; CHECK: bne
ARM: relax the atomic release barrier to "dmb ishst" on Swift Swift cores implement store barriers that are stronger than the ARM specification but weaker than general barriers. They are, in fact, just about enough to provide the ordering needed for atomic operations with release semantics. This patch makes use of that quirk. llvm-svn: 185527
2013-07-03 17:20:36 +08:00
; CHECK: dmb {{ish$}}
Fix 64-bit atomic operations in Thumb mode. The ARM and Thumb variants of LDREXD and STREXD have different constraints and take different operands. Previously the code expanding atomic operations didn't take this into account and asserted in Thumb mode. llvm-svn: 173780
2013-01-29 17:06:13 +08:00
Mass update to CodeGen tests to use CHECK-LABEL for labels corresponding to function definitions for more informative error messages. No functionality change and all updated tests passed locally. This update was done with the following bash script: find test/CodeGen -name "*.ll" | \ while read NAME; do echo "$NAME" if ! grep -q "^; *RUN: *llc.*debug" $NAME; then TEMP=`mktemp -t temp` cp $NAME $TEMP sed -n "s/^define [^@]*@\([A-Za-z0-9_]*\)(.*$/\1/p" < $NAME | \ while read FUNC; do sed -i '' "s/;\(.*\)\([A-Za-z0-9_-]*\):\( *\)$FUNC: *\$/;\1\2-LABEL:\3$FUNC:/g" $TEMP done sed -i '' "s/;\(.*\)-LABEL-LABEL:/;\1-LABEL:/" $TEMP sed -i '' "s/;\(.*\)-NEXT-LABEL:/;\1-NEXT:/" $TEMP sed -i '' "s/;\(.*\)-NOT-LABEL:/;\1-NOT:/" $TEMP sed -i '' "s/;\(.*\)-DAG-LABEL:/;\1-DAG:/" $TEMP mv $TEMP $NAME fi done llvm-svn: 186280
2013-07-14 14:24:09 +08:00
; CHECK-THUMB-LABEL: test5:
ARM: relax the atomic release barrier to "dmb ishst" on Swift Swift cores implement store barriers that are stronger than the ARM specification but weaker than general barriers. They are, in fact, just about enough to provide the ordering needed for atomic operations with release semantics. This patch makes use of that quirk. llvm-svn: 185527
2013-07-03 17:20:36 +08:00
; CHECK-THUMB: dmb {{ish$}}
Fix 64-bit atomic operations in Thumb mode. The ARM and Thumb variants of LDREXD and STREXD have different constraints and take different operands. Previously the code expanding atomic operations didn't take this into account and asserted in Thumb mode. llvm-svn: 173780
2013-01-29 17:06:13 +08:00
; CHECK-THUMB: ldrexd [[REG1:[a-z0-9]+]], [[REG2:[a-z0-9]+]]
ARM big endian function argument passing llvm-svn: 208316
2014-05-08 22:06:24 +08:00
; CHECK-THUMB-LE-DAG: eor.w [[REG3:[a-z0-9]+]], [[REG1]]
; CHECK-THUMB-LE-DAG: eor.w [[REG4:[a-z0-9]+]], [[REG2]]
; CHECK-THUMB-BE-DAG: eor.w [[REG4:[a-z0-9]+]], [[REG2]]
; CHECK-THUMB-BE-DAG: eor.w [[REG3:[a-z0-9]+]], [[REG1]]
Fix 64-bit atomic operations in Thumb mode. The ARM and Thumb variants of LDREXD and STREXD have different constraints and take different operands. Previously the code expanding atomic operations didn't take this into account and asserted in Thumb mode. llvm-svn: 173780
2013-01-29 17:06:13 +08:00
; CHECK-THUMB: strexd {{[a-z0-9]+}}, [[REG3]], [[REG4]]
; CHECK-THUMB: cmp
; CHECK-THUMB: bne
ARM: relax the atomic release barrier to "dmb ishst" on Swift Swift cores implement store barriers that are stronger than the ARM specification but weaker than general barriers. They are, in fact, just about enough to provide the ordering needed for atomic operations with release semantics. This patch makes use of that quirk. llvm-svn: 185527
2013-07-03 17:20:36 +08:00
; CHECK-THUMB: dmb {{ish$}}
Fix 64-bit atomic operations in Thumb mode. The ARM and Thumb variants of LDREXD and STREXD have different constraints and take different operands. Previously the code expanding atomic operations didn't take this into account and asserted in Thumb mode. llvm-svn: 173780
2013-01-29 17:06:13 +08:00
Some 64-bit atomic operations on ARM. 64-bit cmpxchg coming next. llvm-svn: 138845
2011-08-31 08:31:29 +08:00
%r = atomicrmw xor i64* %ptr, i64 %val seq_cst
ret i64 %r
}
define i64 @test6(i64* %ptr, i64 %val) {
Convert CodeGen/*/*.ll tests to use the new CHECK-LABEL for easier debugging. No functionality change and all tests pass after conversion. This was done with the following sed invocation to catch label lines demarking function boundaries: sed -i '' "s/^;\( *\)\([A-Z0-9_]*\):\( *\)test\([A-Za-z0-9_-]*\):\( *\)$/;\1\2-LABEL:\3test\4:\5/g" test/CodeGen/*/*.ll which was written conservatively to avoid false positives rather than false negatives. I scanned through all the changes and everything looks correct. llvm-svn: 186258
2013-07-14 04:38:47 +08:00
; CHECK-LABEL: test6:
ARM: relax the atomic release barrier to "dmb ishst" on Swift Swift cores implement store barriers that are stronger than the ARM specification but weaker than general barriers. They are, in fact, just about enough to provide the ordering needed for atomic operations with release semantics. This patch makes use of that quirk. llvm-svn: 185527
2013-07-03 17:20:36 +08:00
; CHECK: dmb {{ish$}}
Remove hard coded registers in ARM ldrexd and strexd instructions This patch replaces the hard coded GPR pair [R0, R1] of Intrinsic:arm_ldrexd and [R2, R3] of Intrinsic:arm_strexd with even/odd GPRPair reg class. Similar to the lowering of atomic_64 operation. llvm-svn: 168207
2012-11-17 05:55:34 +08:00
; CHECK: ldrexd [[REG1:(r[0-9]?[02468])]], [[REG2:(r[0-9]?[13579])]]
; CHECK: strexd {{[a-z0-9]+}}, {{r[0-9]?[02468]}}, {{r[0-9]?[13579]}}
Some 64-bit atomic operations on ARM. 64-bit cmpxchg coming next. llvm-svn: 138845
2011-08-31 08:31:29 +08:00
; CHECK: cmp
; CHECK: bne
ARM: relax the atomic release barrier to "dmb ishst" on Swift Swift cores implement store barriers that are stronger than the ARM specification but weaker than general barriers. They are, in fact, just about enough to provide the ordering needed for atomic operations with release semantics. This patch makes use of that quirk. llvm-svn: 185527
2013-07-03 17:20:36 +08:00
; CHECK: dmb {{ish$}}
Fix 64-bit atomic operations in Thumb mode. The ARM and Thumb variants of LDREXD and STREXD have different constraints and take different operands. Previously the code expanding atomic operations didn't take this into account and asserted in Thumb mode. llvm-svn: 173780
2013-01-29 17:06:13 +08:00
Mass update to CodeGen tests to use CHECK-LABEL for labels corresponding to function definitions for more informative error messages. No functionality change and all updated tests passed locally. This update was done with the following bash script: find test/CodeGen -name "*.ll" | \ while read NAME; do echo "$NAME" if ! grep -q "^; *RUN: *llc.*debug" $NAME; then TEMP=`mktemp -t temp` cp $NAME $TEMP sed -n "s/^define [^@]*@\([A-Za-z0-9_]*\)(.*$/\1/p" < $NAME | \ while read FUNC; do sed -i '' "s/;\(.*\)\([A-Za-z0-9_-]*\):\( *\)$FUNC: *\$/;\1\2-LABEL:\3$FUNC:/g" $TEMP done sed -i '' "s/;\(.*\)-LABEL-LABEL:/;\1-LABEL:/" $TEMP sed -i '' "s/;\(.*\)-NEXT-LABEL:/;\1-NEXT:/" $TEMP sed -i '' "s/;\(.*\)-NOT-LABEL:/;\1-NOT:/" $TEMP sed -i '' "s/;\(.*\)-DAG-LABEL:/;\1-DAG:/" $TEMP mv $TEMP $NAME fi done llvm-svn: 186280
2013-07-14 14:24:09 +08:00
; CHECK-THUMB-LABEL: test6:
ARM: relax the atomic release barrier to "dmb ishst" on Swift Swift cores implement store barriers that are stronger than the ARM specification but weaker than general barriers. They are, in fact, just about enough to provide the ordering needed for atomic operations with release semantics. This patch makes use of that quirk. llvm-svn: 185527
2013-07-03 17:20:36 +08:00
; CHECK-THUMB: dmb {{ish$}}
Fix 64-bit atomic operations in Thumb mode. The ARM and Thumb variants of LDREXD and STREXD have different constraints and take different operands. Previously the code expanding atomic operations didn't take this into account and asserted in Thumb mode. llvm-svn: 173780
2013-01-29 17:06:13 +08:00
; CHECK-THUMB: ldrexd [[REG1:[a-z0-9]+]], [[REG2:[a-z0-9]+]]
; CHECK-THUMB: strexd {{[a-z0-9]+}}, {{[a-z0-9]+}}, {{[a-z0-9]+}}
; CHECK-THUMB: cmp
; CHECK-THUMB: bne
ARM: relax the atomic release barrier to "dmb ishst" on Swift Swift cores implement store barriers that are stronger than the ARM specification but weaker than general barriers. They are, in fact, just about enough to provide the ordering needed for atomic operations with release semantics. This patch makes use of that quirk. llvm-svn: 185527
2013-07-03 17:20:36 +08:00
; CHECK-THUMB: dmb {{ish$}}
Fix 64-bit atomic operations in Thumb mode. The ARM and Thumb variants of LDREXD and STREXD have different constraints and take different operands. Previously the code expanding atomic operations didn't take this into account and asserted in Thumb mode. llvm-svn: 173780
2013-01-29 17:06:13 +08:00
Some 64-bit atomic operations on ARM. 64-bit cmpxchg coming next. llvm-svn: 138845
2011-08-31 08:31:29 +08:00
%r = atomicrmw xchg i64* %ptr, i64 %val seq_cst
ret i64 %r
Some minor cleanups for r138845. llvm-svn: 138846
2011-08-31 08:41:05 +08:00
}
64-bit atomic cmpxchg for ARM. llvm-svn: 138868
2011-09-01 01:52:22 +08:00
define i64 @test7(i64* %ptr, i64 %val1, i64 %val2) {
Convert CodeGen/*/*.ll tests to use the new CHECK-LABEL for easier debugging. No functionality change and all tests pass after conversion. This was done with the following sed invocation to catch label lines demarking function boundaries: sed -i '' "s/^;\( *\)\([A-Z0-9_]*\):\( *\)test\([A-Za-z0-9_-]*\):\( *\)$/;\1\2-LABEL:\3test\4:\5/g" test/CodeGen/*/*.ll which was written conservatively to avoid false positives rather than false negatives. I scanned through all the changes and everything looks correct. llvm-svn: 186258
2013-07-14 04:38:47 +08:00
; CHECK-LABEL: test7:
Atomics: make use of the "cmpxchg weak" instruction. This also simplifies the IR we create slightly: instead of working out where success & failure should go manually, it turns out we can just always jump to a success/failure block created for the purpose. Later phases will sort out the mess without much difficulty. llvm-svn: 210917
2014-06-14 00:45:52 +08:00
; CHECK-DAG: mov [[VAL1LO:r[0-9]+]], r1
Remove hard coded registers in ARM ldrexd and strexd instructions This patch replaces the hard coded GPR pair [R0, R1] of Intrinsic:arm_ldrexd and [R2, R3] of Intrinsic:arm_strexd with even/odd GPRPair reg class. Similar to the lowering of atomic_64 operation. llvm-svn: 168207
2012-11-17 05:55:34 +08:00
; CHECK: ldrexd [[REG1:(r[0-9]?[02468])]], [[REG2:(r[0-9]?[13579])]]
ARM: sink atomic release barrier as far as possible into cmpxchg. DMB instructions can be expensive, so it's best to avoid them if possible. In atomicrmw operations there will always be an attempted store so a release barrier is always needed, but in the cmpxchg case we can delay the DMB until we know we'll definitely try to perform a store (and so need release semantics). In the strong cmpxchg case this isn't quite free: we must duplicate the LDREX instructions to skip the barrier on subsequent iterations. The basic outline becomes: ldrex rOld, [rAddr] cmp rOld, rDesired bne Ldone dmb Lloop: strex rRes, rNew, [rAddr] cbz rRes Ldone ldrex rOld, [rAddr] cmp rOld, rDesired beq Lloop Ldone: So we'll skip this version for strong operations in "minsize" functions. llvm-svn: 261568
2016-02-23 04:55:50 +08:00
; CHECK-LE-DAG: eor [[MISMATCH_LO:.*]], [[REG1]], [[VAL1LO]]
; CHECK-LE-DAG: eor [[MISMATCH_HI:.*]], [[REG2]], r2
; CHECK-BE-DAG: eor [[MISMATCH_LO:.*]], [[REG2]], r2
; CHECK-BE-DAG: eor [[MISMATCH_HI:.*]], [[REG1]], r1
ARM: expand atomic ldrex/strex loops in IR The previous situation where ATOMIC_LOAD_WHATEVER nodes were expanded at MachineInstr emission time had grown to be extremely large and involved, to account for the subtly different code needed for the various flavours (8/16/32/64 bit, cmpxchg/add/minmax). Moving this transformation into the IR clears up the code substantially, and makes future optimisations much easier: 1. an atomicrmw followed by using the *new* value can be more efficient. As an IR pass, simple CSE could handle this efficiently. 2. Making use of cmpxchg success/failure orderings only has to be done in one (simpler) place. 3. The common "cmpxchg; did we store?" idiom can be exposed to optimisation. I intend to gradually improve this situation within the ARM backend and make sure there are no hidden issues before moving the code out into CodeGen to be shared with (at least ARM64/AArch64, though I think PPC & Mips could benefit too). llvm-svn: 205525
2014-04-03 19:44:58 +08:00
; CHECK: orrs {{r[0-9]+}}, [[MISMATCH_LO]], [[MISMATCH_HI]]
64-bit atomic cmpxchg for ARM. llvm-svn: 138868
2011-09-01 01:52:22 +08:00
; CHECK: bne
ARM: sink atomic release barrier as far as possible into cmpxchg. DMB instructions can be expensive, so it's best to avoid them if possible. In atomicrmw operations there will always be an attempted store so a release barrier is always needed, but in the cmpxchg case we can delay the DMB until we know we'll definitely try to perform a store (and so need release semantics). In the strong cmpxchg case this isn't quite free: we must duplicate the LDREX instructions to skip the barrier on subsequent iterations. The basic outline becomes: ldrex rOld, [rAddr] cmp rOld, rDesired bne Ldone dmb Lloop: strex rRes, rNew, [rAddr] cbz rRes Ldone ldrex rOld, [rAddr] cmp rOld, rDesired beq Lloop Ldone: So we'll skip this version for strong operations in "minsize" functions. llvm-svn: 261568
2016-02-23 04:55:50 +08:00
; CHECK-DAG: dmb {{ish$}}
Remove hard coded registers in ARM ldrexd and strexd instructions This patch replaces the hard coded GPR pair [R0, R1] of Intrinsic:arm_ldrexd and [R2, R3] of Intrinsic:arm_strexd with even/odd GPRPair reg class. Similar to the lowering of atomic_64 operation. llvm-svn: 168207
2012-11-17 05:55:34 +08:00
; CHECK: strexd {{[a-z0-9]+}}, {{r[0-9]?[02468]}}, {{r[0-9]?[13579]}}
64-bit atomic cmpxchg for ARM. llvm-svn: 138868
2011-09-01 01:52:22 +08:00
; CHECK: cmp
ARM: sink atomic release barrier as far as possible into cmpxchg. DMB instructions can be expensive, so it's best to avoid them if possible. In atomicrmw operations there will always be an attempted store so a release barrier is always needed, but in the cmpxchg case we can delay the DMB until we know we'll definitely try to perform a store (and so need release semantics). In the strong cmpxchg case this isn't quite free: we must duplicate the LDREX instructions to skip the barrier on subsequent iterations. The basic outline becomes: ldrex rOld, [rAddr] cmp rOld, rDesired bne Ldone dmb Lloop: strex rRes, rNew, [rAddr] cbz rRes Ldone ldrex rOld, [rAddr] cmp rOld, rDesired beq Lloop Ldone: So we'll skip this version for strong operations in "minsize" functions. llvm-svn: 261568
2016-02-23 04:55:50 +08:00
; CHECK: beq
ARM: relax the atomic release barrier to "dmb ishst" on Swift Swift cores implement store barriers that are stronger than the ARM specification but weaker than general barriers. They are, in fact, just about enough to provide the ordering needed for atomic operations with release semantics. This patch makes use of that quirk. llvm-svn: 185527
2013-07-03 17:20:36 +08:00
; CHECK: dmb {{ish$}}
Fix 64-bit atomic operations in Thumb mode. The ARM and Thumb variants of LDREXD and STREXD have different constraints and take different operands. Previously the code expanding atomic operations didn't take this into account and asserted in Thumb mode. llvm-svn: 173780
2013-01-29 17:06:13 +08:00
Mass update to CodeGen tests to use CHECK-LABEL for labels corresponding to function definitions for more informative error messages. No functionality change and all updated tests passed locally. This update was done with the following bash script: find test/CodeGen -name "*.ll" | \ while read NAME; do echo "$NAME" if ! grep -q "^; *RUN: *llc.*debug" $NAME; then TEMP=`mktemp -t temp` cp $NAME $TEMP sed -n "s/^define [^@]*@\([A-Za-z0-9_]*\)(.*$/\1/p" < $NAME | \ while read FUNC; do sed -i '' "s/;\(.*\)\([A-Za-z0-9_-]*\):\( *\)$FUNC: *\$/;\1\2-LABEL:\3$FUNC:/g" $TEMP done sed -i '' "s/;\(.*\)-LABEL-LABEL:/;\1-LABEL:/" $TEMP sed -i '' "s/;\(.*\)-NEXT-LABEL:/;\1-NEXT:/" $TEMP sed -i '' "s/;\(.*\)-NOT-LABEL:/;\1-NOT:/" $TEMP sed -i '' "s/;\(.*\)-DAG-LABEL:/;\1-DAG:/" $TEMP mv $TEMP $NAME fi done llvm-svn: 186280
2013-07-14 14:24:09 +08:00
; CHECK-THUMB-LABEL: test7:
Fix 64-bit atomic operations in Thumb mode. The ARM and Thumb variants of LDREXD and STREXD have different constraints and take different operands. Previously the code expanding atomic operations didn't take this into account and asserted in Thumb mode. llvm-svn: 173780
2013-01-29 17:06:13 +08:00
; CHECK-THUMB: ldrexd [[REG1:[a-z0-9]+]], [[REG2:[a-z0-9]+]]
ARM big endian function argument passing llvm-svn: 208316
2014-05-08 22:06:24 +08:00
; CHECK-THUMB-LE-DAG: eor.w [[MISMATCH_LO:[a-z0-9]+]], [[REG1]], r2
; CHECK-THUMB-LE-DAG: eor.w [[MISMATCH_HI:[a-z0-9]+]], [[REG2]], r3
ARM & AArch64: make use of common cmpxchg idioms after expansion The C and C++ semantics for compare_exchange require it to return a bool indicating success. This gets mapped to LLVM IR which follows each cmpxchg with an icmp of the value loaded against the desired value. When lowered to ldxr/stxr loops, this extra comparison is redundant: its results are implicit in the control-flow of the function. This commit makes two changes: it replaces that icmp with appropriate PHI nodes, and then makes sure earlyCSE is called after expansion to actually make use of the opportunities revealed. I've also added -{arm,aarch64}-enable-atomic-tidy options, so that existing fragile tests aren't perturbed too much by the change. Many of them either rely on undef/unreachable too pervasively to be restored to something well-defined (particularly while making sure they test the same obscure assert from many years ago), or depend on a particular CFG shape, which is disrupted by SimplifyCFG. rdar://problem/16227836 llvm-svn: 209883
2014-05-30 18:09:59 +08:00
; CHECK-THUMB-BE-DAG: eor.w [[MISMATCH_HI:[a-z0-9]+]], [[REG1]], r2
; CHECK-THUMB-BE-DAG: eor.w [[MISMATCH_LO:[a-z0-9]+]], [[REG2]], r3
ARM: sink atomic release barrier as far as possible into cmpxchg. DMB instructions can be expensive, so it's best to avoid them if possible. In atomicrmw operations there will always be an attempted store so a release barrier is always needed, but in the cmpxchg case we can delay the DMB until we know we'll definitely try to perform a store (and so need release semantics). In the strong cmpxchg case this isn't quite free: we must duplicate the LDREX instructions to skip the barrier on subsequent iterations. The basic outline becomes: ldrex rOld, [rAddr] cmp rOld, rDesired bne Ldone dmb Lloop: strex rRes, rNew, [rAddr] cbz rRes Ldone ldrex rOld, [rAddr] cmp rOld, rDesired beq Lloop Ldone: So we'll skip this version for strong operations in "minsize" functions. llvm-svn: 261568
2016-02-23 04:55:50 +08:00
; CHECK-THUMB-LE: orrs.w {{.*}}, [[MISMATCH_LO]], [[MISMATCH_HI]]
Fix 64-bit atomic operations in Thumb mode. The ARM and Thumb variants of LDREXD and STREXD have different constraints and take different operands. Previously the code expanding atomic operations didn't take this into account and asserted in Thumb mode. llvm-svn: 173780
2013-01-29 17:06:13 +08:00
; CHECK-THUMB: bne
ARM: sink atomic release barrier as far as possible into cmpxchg. DMB instructions can be expensive, so it's best to avoid them if possible. In atomicrmw operations there will always be an attempted store so a release barrier is always needed, but in the cmpxchg case we can delay the DMB until we know we'll definitely try to perform a store (and so need release semantics). In the strong cmpxchg case this isn't quite free: we must duplicate the LDREX instructions to skip the barrier on subsequent iterations. The basic outline becomes: ldrex rOld, [rAddr] cmp rOld, rDesired bne Ldone dmb Lloop: strex rRes, rNew, [rAddr] cbz rRes Ldone ldrex rOld, [rAddr] cmp rOld, rDesired beq Lloop Ldone: So we'll skip this version for strong operations in "minsize" functions. llvm-svn: 261568
2016-02-23 04:55:50 +08:00
; CHECK-THUMB: dmb {{ish$}}
Fix 64-bit atomic operations in Thumb mode. The ARM and Thumb variants of LDREXD and STREXD have different constraints and take different operands. Previously the code expanding atomic operations didn't take this into account and asserted in Thumb mode. llvm-svn: 173780
2013-01-29 17:06:13 +08:00
; CHECK-THUMB: strexd {{[a-z0-9]+}}, {{[a-z0-9]+}}, {{[a-z0-9]+}}
; CHECK-THUMB: cmp
ARM: sink atomic release barrier as far as possible into cmpxchg. DMB instructions can be expensive, so it's best to avoid them if possible. In atomicrmw operations there will always be an attempted store so a release barrier is always needed, but in the cmpxchg case we can delay the DMB until we know we'll definitely try to perform a store (and so need release semantics). In the strong cmpxchg case this isn't quite free: we must duplicate the LDREX instructions to skip the barrier on subsequent iterations. The basic outline becomes: ldrex rOld, [rAddr] cmp rOld, rDesired bne Ldone dmb Lloop: strex rRes, rNew, [rAddr] cbz rRes Ldone ldrex rOld, [rAddr] cmp rOld, rDesired beq Lloop Ldone: So we'll skip this version for strong operations in "minsize" functions. llvm-svn: 261568
2016-02-23 04:55:50 +08:00
; CHECK-THUMB: beq
ARM: relax the atomic release barrier to "dmb ishst" on Swift Swift cores implement store barriers that are stronger than the ARM specification but weaker than general barriers. They are, in fact, just about enough to provide the ordering needed for atomic operations with release semantics. This patch makes use of that quirk. llvm-svn: 185527
2013-07-03 17:20:36 +08:00
; CHECK-THUMB: dmb {{ish$}}
Fix 64-bit atomic operations in Thumb mode. The ARM and Thumb variants of LDREXD and STREXD have different constraints and take different operands. Previously the code expanding atomic operations didn't take this into account and asserted in Thumb mode. llvm-svn: 173780
2013-01-29 17:06:13 +08:00
IR: add "cmpxchg weak" variant to support permitted failure. This commit adds a weak variant of the cmpxchg operation, as described in C++11. A cmpxchg instruction with this modifier is permitted to fail to store, even if the comparison indicated it should. As a result, cmpxchg instructions must return a flag indicating success in addition to their original iN value loaded. Thus, for uniformity *all* cmpxchg instructions now return "{ iN, i1 }". The second flag is 1 when the store succeeded. At the DAG level, a new ATOMIC_CMP_SWAP_WITH_SUCCESS node has been added as the natural representation for the new cmpxchg instructions. It is a strong cmpxchg. By default this gets Expanded to the existing ATOMIC_CMP_SWAP during Legalization, so existing backends should see no change in behaviour. If they wish to deal with the enhanced node instead, they can call setOperationAction on it. Beware: as a node with 2 results, it cannot be selected from TableGen. Currently, no use is made of the extra information provided in this patch. Test updates are almost entirely adapting the input IR to the new scheme. Summary for out of tree users: ------------------------------ + Legacy Bitcode files are upgraded during read. + Legacy assembly IR files will be invalid. + Front-ends must adapt to different type for "cmpxchg". + Backends should be unaffected by default. llvm-svn: 210903
2014-06-13 22:24:07 +08:00
%pair = cmpxchg i64* %ptr, i64 %val1, i64 %val2 seq_cst seq_cst
%r = extractvalue { i64, i1 } %pair, 0
64-bit atomic cmpxchg for ARM. llvm-svn: 138868
2011-09-01 01:52:22 +08:00
ret i64 %r
}
Generic expansion for atomic load/store into cmpxchg/atomicrmw xchg; implements 64-bit atomic load/store for ARM. llvm-svn: 138872
2011-09-01 02:26:09 +08:00
[ARM] Use the load-acquire/store-release instructions optimally in AArch32. Patch by Artyom Skrobov. llvm-svn: 191428
2013-09-26 20:22:36 +08:00
; Compiles down to a single ldrexd
Generic expansion for atomic load/store into cmpxchg/atomicrmw xchg; implements 64-bit atomic load/store for ARM. llvm-svn: 138872
2011-09-01 02:26:09 +08:00
define i64 @test8(i64* %ptr) {
Convert CodeGen/*/*.ll tests to use the new CHECK-LABEL for easier debugging. No functionality change and all tests pass after conversion. This was done with the following sed invocation to catch label lines demarking function boundaries: sed -i '' "s/^;\( *\)\([A-Z0-9_]*\):\( *\)test\([A-Za-z0-9_-]*\):\( *\)$/;\1\2-LABEL:\3test\4:\5/g" test/CodeGen/*/*.ll which was written conservatively to avoid false positives rather than false negatives. I scanned through all the changes and everything looks correct. llvm-svn: 186258
2013-07-14 04:38:47 +08:00
; CHECK-LABEL: test8:
Remove hard coded registers in ARM ldrexd and strexd instructions This patch replaces the hard coded GPR pair [R0, R1] of Intrinsic:arm_ldrexd and [R2, R3] of Intrinsic:arm_strexd with even/odd GPRPair reg class. Similar to the lowering of atomic_64 operation. llvm-svn: 168207
2012-11-17 05:55:34 +08:00
; CHECK: ldrexd [[REG1:(r[0-9]?[02468])]], [[REG2:(r[0-9]?[13579])]]
AArch64: use ldxp/stxp pair to implement 128-bit atomic loads. The ARM ARM is clear that 128-bit loads are only guaranteed to have been atomic if there has been a corresponding successful stxp. It's less clear for AArch32, so I'm leaving that alone for now. llvm-svn: 254524
2015-12-03 02:12:57 +08:00
; CHECK-NOT: strexd
; CHECK: clrex
; CHECK-NOT: strexd
ARM: relax the atomic release barrier to "dmb ishst" on Swift Swift cores implement store barriers that are stronger than the ARM specification but weaker than general barriers. They are, in fact, just about enough to provide the ordering needed for atomic operations with release semantics. This patch makes use of that quirk. llvm-svn: 185527
2013-07-03 17:20:36 +08:00
; CHECK: dmb {{ish$}}
Fix 64-bit atomic operations in Thumb mode. The ARM and Thumb variants of LDREXD and STREXD have different constraints and take different operands. Previously the code expanding atomic operations didn't take this into account and asserted in Thumb mode. llvm-svn: 173780
2013-01-29 17:06:13 +08:00
Mass update to CodeGen tests to use CHECK-LABEL for labels corresponding to function definitions for more informative error messages. No functionality change and all updated tests passed locally. This update was done with the following bash script: find test/CodeGen -name "*.ll" | \ while read NAME; do echo "$NAME" if ! grep -q "^; *RUN: *llc.*debug" $NAME; then TEMP=`mktemp -t temp` cp $NAME $TEMP sed -n "s/^define [^@]*@\([A-Za-z0-9_]*\)(.*$/\1/p" < $NAME | \ while read FUNC; do sed -i '' "s/;\(.*\)\([A-Za-z0-9_-]*\):\( *\)$FUNC: *\$/;\1\2-LABEL:\3$FUNC:/g" $TEMP done sed -i '' "s/;\(.*\)-LABEL-LABEL:/;\1-LABEL:/" $TEMP sed -i '' "s/;\(.*\)-NEXT-LABEL:/;\1-NEXT:/" $TEMP sed -i '' "s/;\(.*\)-NOT-LABEL:/;\1-NOT:/" $TEMP sed -i '' "s/;\(.*\)-DAG-LABEL:/;\1-DAG:/" $TEMP mv $TEMP $NAME fi done llvm-svn: 186280
2013-07-14 14:24:09 +08:00
; CHECK-THUMB-LABEL: test8:
Fix 64-bit atomic operations in Thumb mode. The ARM and Thumb variants of LDREXD and STREXD have different constraints and take different operands. Previously the code expanding atomic operations didn't take this into account and asserted in Thumb mode. llvm-svn: 173780
2013-01-29 17:06:13 +08:00
; CHECK-THUMB: ldrexd [[REG1:[a-z0-9]+]], [[REG2:[a-z0-9]+]]
AArch64: use ldxp/stxp pair to implement 128-bit atomic loads. The ARM ARM is clear that 128-bit loads are only guaranteed to have been atomic if there has been a corresponding successful stxp. It's less clear for AArch32, so I'm leaving that alone for now. llvm-svn: 254524
2015-12-03 02:12:57 +08:00
; CHECK-THUMB-NOT: strexd
; CHECK-THUMB: clrex
; CHECK-THUMB-NOT: strexd
ARM: relax the atomic release barrier to "dmb ishst" on Swift Swift cores implement store barriers that are stronger than the ARM specification but weaker than general barriers. They are, in fact, just about enough to provide the ordering needed for atomic operations with release semantics. This patch makes use of that quirk. llvm-svn: 185527
2013-07-03 17:20:36 +08:00
; CHECK-THUMB: dmb {{ish$}}
Fix 64-bit atomic operations in Thumb mode. The ARM and Thumb variants of LDREXD and STREXD have different constraints and take different operands. Previously the code expanding atomic operations didn't take this into account and asserted in Thumb mode. llvm-svn: 173780
2013-01-29 17:06:13 +08:00
[opaque pointer type] Add textual IR support for explicit type parameter to load instruction Essentially the same as the GEP change in r230786. A similar migration script can be used to update test cases, though a few more test case improvements/changes were required this time around: (r229269-r229278) import fileinput import sys import re pat = re.compile(r"((?:=|:|^)\s*load (?:atomic )?(?:volatile )?(.*?))(| addrspace\(\d+\) *)\*($| *(?:%|@|null|undef|blockaddress|getelementptr|addrspacecast|bitcast|inttoptr|\[\[[a-zA-Z]|\{\{).*$)") for line in sys.stdin: sys.stdout.write(re.sub(pat, r"\1, \2\3*\4", line)) Reviewers: rafael, dexonsmith, grosser Differential Revision: http://reviews.llvm.org/D7649 llvm-svn: 230794
2015-02-28 05:17:42 +08:00
%r = load atomic i64, i64* %ptr seq_cst, align 8
Generic expansion for atomic load/store into cmpxchg/atomicrmw xchg; implements 64-bit atomic load/store for ARM. llvm-svn: 138872
2011-09-01 02:26:09 +08:00
ret i64 %r
}
; Compiles down to atomicrmw xchg; there really isn't any more efficient
; way to write it.
define void @test9(i64* %ptr, i64 %val) {
Convert CodeGen/*/*.ll tests to use the new CHECK-LABEL for easier debugging. No functionality change and all tests pass after conversion. This was done with the following sed invocation to catch label lines demarking function boundaries: sed -i '' "s/^;\( *\)\([A-Z0-9_]*\):\( *\)test\([A-Za-z0-9_-]*\):\( *\)$/;\1\2-LABEL:\3test\4:\5/g" test/CodeGen/*/*.ll which was written conservatively to avoid false positives rather than false negatives. I scanned through all the changes and everything looks correct. llvm-svn: 186258
2013-07-14 04:38:47 +08:00
; CHECK-LABEL: test9:
ARM: relax the atomic release barrier to "dmb ishst" on Swift Swift cores implement store barriers that are stronger than the ARM specification but weaker than general barriers. They are, in fact, just about enough to provide the ordering needed for atomic operations with release semantics. This patch makes use of that quirk. llvm-svn: 185527
2013-07-03 17:20:36 +08:00
; CHECK: dmb {{ish$}}
Remove hard coded registers in ARM ldrexd and strexd instructions This patch replaces the hard coded GPR pair [R0, R1] of Intrinsic:arm_ldrexd and [R2, R3] of Intrinsic:arm_strexd with even/odd GPRPair reg class. Similar to the lowering of atomic_64 operation. llvm-svn: 168207
2012-11-17 05:55:34 +08:00
; CHECK: ldrexd [[REG1:(r[0-9]?[02468])]], [[REG2:(r[0-9]?[13579])]]
; CHECK: strexd {{[a-z0-9]+}}, {{r[0-9]?[02468]}}, {{r[0-9]?[13579]}}
Generic expansion for atomic load/store into cmpxchg/atomicrmw xchg; implements 64-bit atomic load/store for ARM. llvm-svn: 138872
2011-09-01 02:26:09 +08:00
; CHECK: cmp
; CHECK: bne
ARM: relax the atomic release barrier to "dmb ishst" on Swift Swift cores implement store barriers that are stronger than the ARM specification but weaker than general barriers. They are, in fact, just about enough to provide the ordering needed for atomic operations with release semantics. This patch makes use of that quirk. llvm-svn: 185527
2013-07-03 17:20:36 +08:00
; CHECK: dmb {{ish$}}
Fix 64-bit atomic operations in Thumb mode. The ARM and Thumb variants of LDREXD and STREXD have different constraints and take different operands. Previously the code expanding atomic operations didn't take this into account and asserted in Thumb mode. llvm-svn: 173780
2013-01-29 17:06:13 +08:00
Mass update to CodeGen tests to use CHECK-LABEL for labels corresponding to function definitions for more informative error messages. No functionality change and all updated tests passed locally. This update was done with the following bash script: find test/CodeGen -name "*.ll" | \ while read NAME; do echo "$NAME" if ! grep -q "^; *RUN: *llc.*debug" $NAME; then TEMP=`mktemp -t temp` cp $NAME $TEMP sed -n "s/^define [^@]*@\([A-Za-z0-9_]*\)(.*$/\1/p" < $NAME | \ while read FUNC; do sed -i '' "s/;\(.*\)\([A-Za-z0-9_-]*\):\( *\)$FUNC: *\$/;\1\2-LABEL:\3$FUNC:/g" $TEMP done sed -i '' "s/;\(.*\)-LABEL-LABEL:/;\1-LABEL:/" $TEMP sed -i '' "s/;\(.*\)-NEXT-LABEL:/;\1-NEXT:/" $TEMP sed -i '' "s/;\(.*\)-NOT-LABEL:/;\1-NOT:/" $TEMP sed -i '' "s/;\(.*\)-DAG-LABEL:/;\1-DAG:/" $TEMP mv $TEMP $NAME fi done llvm-svn: 186280
2013-07-14 14:24:09 +08:00
; CHECK-THUMB-LABEL: test9:
ARM: relax the atomic release barrier to "dmb ishst" on Swift Swift cores implement store barriers that are stronger than the ARM specification but weaker than general barriers. They are, in fact, just about enough to provide the ordering needed for atomic operations with release semantics. This patch makes use of that quirk. llvm-svn: 185527
2013-07-03 17:20:36 +08:00
; CHECK-THUMB: dmb {{ish$}}
Fix 64-bit atomic operations in Thumb mode. The ARM and Thumb variants of LDREXD and STREXD have different constraints and take different operands. Previously the code expanding atomic operations didn't take this into account and asserted in Thumb mode. llvm-svn: 173780
2013-01-29 17:06:13 +08:00
; CHECK-THUMB: ldrexd [[REG1:[a-z0-9]+]], [[REG2:[a-z0-9]+]]
; CHECK-THUMB: strexd {{[a-z0-9]+}}, {{[a-z0-9]+}}, {{[a-z0-9]+}}
; CHECK-THUMB: cmp
; CHECK-THUMB: bne
ARM: relax the atomic release barrier to "dmb ishst" on Swift Swift cores implement store barriers that are stronger than the ARM specification but weaker than general barriers. They are, in fact, just about enough to provide the ordering needed for atomic operations with release semantics. This patch makes use of that quirk. llvm-svn: 185527
2013-07-03 17:20:36 +08:00
; CHECK-THUMB: dmb {{ish$}}
Fix 64-bit atomic operations in Thumb mode. The ARM and Thumb variants of LDREXD and STREXD have different constraints and take different operands. Previously the code expanding atomic operations didn't take this into account and asserted in Thumb mode. llvm-svn: 173780
2013-01-29 17:06:13 +08:00
Generic expansion for atomic load/store into cmpxchg/atomicrmw xchg; implements 64-bit atomic load/store for ARM. llvm-svn: 138872
2011-09-01 02:26:09 +08:00
store atomic i64 %val, i64* %ptr seq_cst, align 8
ret void
}
Added atomic 64 min/max/umin/umax instrinsics support in the ARM backend. llvm-svn: 168886
2012-11-29 22:41:25 +08:00
define i64 @test10(i64* %ptr, i64 %val) {
Convert CodeGen/*/*.ll tests to use the new CHECK-LABEL for easier debugging. No functionality change and all tests pass after conversion. This was done with the following sed invocation to catch label lines demarking function boundaries: sed -i '' "s/^;\( *\)\([A-Z0-9_]*\):\( *\)test\([A-Za-z0-9_-]*\):\( *\)$/;\1\2-LABEL:\3test\4:\5/g" test/CodeGen/*/*.ll which was written conservatively to avoid false positives rather than false negatives. I scanned through all the changes and everything looks correct. llvm-svn: 186258
2013-07-14 04:38:47 +08:00
; CHECK-LABEL: test10:
ARM: relax the atomic release barrier to "dmb ishst" on Swift Swift cores implement store barriers that are stronger than the ARM specification but weaker than general barriers. They are, in fact, just about enough to provide the ordering needed for atomic operations with release semantics. This patch makes use of that quirk. llvm-svn: 185527
2013-07-03 17:20:36 +08:00
; CHECK: dmb {{ish$}}
Added atomic 64 min/max/umin/umax instrinsics support in the ARM backend. llvm-svn: 168886
2012-11-29 22:41:25 +08:00
; CHECK: ldrexd [[REG1:(r[0-9]?[02468])]], [[REG2:(r[0-9]?[13579])]]
ARM: expand atomic ldrex/strex loops in IR The previous situation where ATOMIC_LOAD_WHATEVER nodes were expanded at MachineInstr emission time had grown to be extremely large and involved, to account for the subtly different code needed for the various flavours (8/16/32/64 bit, cmpxchg/add/minmax). Moving this transformation into the IR clears up the code substantially, and makes future optimisations much easier: 1. an atomicrmw followed by using the *new* value can be more efficient. As an IR pass, simple CSE could handle this efficiently. 2. Making use of cmpxchg success/failure orderings only has to be done in one (simpler) place. 3. The common "cmpxchg; did we store?" idiom can be exposed to optimisation. I intend to gradually improve this situation within the ARM backend and make sure there are no hidden issues before moving the code out into CodeGen to be shared with (at least ARM64/AArch64, though I think PPC & Mips could benefit too). llvm-svn: 205525
2014-04-03 19:44:58 +08:00
; CHECK: mov [[OUT_HI:[a-z0-9]+]], r2
ARM: Better codegen for 64-bit compares. This introduces a custom lowering for ISD::SETCCE (introduced in r253572) that allows us to emit a short code sequence for 64-bit compares. Before: push {r7, lr} cmp r0, r2 mov.w r0, #0 mov.w r12, #0 it hs movhs r0, #1 cmp r1, r3 it ge movge.w r12, #1 it eq moveq r12, r0 cmp.w r12, #0 bne .LBB1_2 @ BB#1: @ %bb1 bl f pop {r7, pc} .LBB1_2: @ %bb2 bl g pop {r7, pc} After: push {r7, lr} subs r0, r0, r2 sbcs.w r0, r1, r3 bge .LBB1_2 @ BB#1: @ %bb1 bl f pop {r7, pc} .LBB1_2: @ %bb2 bl g pop {r7, pc} Saves around 80KB in Chromium's libchrome.so. Some notes on this patch: - I don't much like the ARMISD::BRCOND and ARMISD::CMOV combines I introduced (nothing else needs them). However, they are necessary in order to avoid poor codegen, and they seem similar to existing combines in other backends (e.g. X86 combines (brcond (cmp (setcc Compare))) to (brcond Compare)). - No support for Thumb-1. This is in principle possible, but we'd need to implement ARMISD::SUBE for Thumb-1. Differential Revision: http://reviews.llvm.org/D15256 llvm-svn: 263962
2016-03-22 02:00:02 +08:00
; CHECK-LE: subs {{[^,]+}}, r1, [[REG1]]
; CHECK-BE: subs {{[^,]+}}, r2, [[REG2]]
; CHECK-LE: sbcs {{[^,]+}}, r2, [[REG2]]
; CHECK-BE: sbcs {{[^,]+}}, r1, [[REG1]]
; CHECK: mov [[CMP:[a-z0-9]+]], #0
; CHECK: movwge [[CMP]], #1
; CHECK: cmp [[CMP]], #0
ARM: expand atomic ldrex/strex loops in IR The previous situation where ATOMIC_LOAD_WHATEVER nodes were expanded at MachineInstr emission time had grown to be extremely large and involved, to account for the subtly different code needed for the various flavours (8/16/32/64 bit, cmpxchg/add/minmax). Moving this transformation into the IR clears up the code substantially, and makes future optimisations much easier: 1. an atomicrmw followed by using the *new* value can be more efficient. As an IR pass, simple CSE could handle this efficiently. 2. Making use of cmpxchg success/failure orderings only has to be done in one (simpler) place. 3. The common "cmpxchg; did we store?" idiom can be exposed to optimisation. I intend to gradually improve this situation within the ARM backend and make sure there are no hidden issues before moving the code out into CodeGen to be shared with (at least ARM64/AArch64, though I think PPC & Mips could benefit too). llvm-svn: 205525
2014-04-03 19:44:58 +08:00
; CHECK: movne [[OUT_HI]], [[REG2]]
; CHECK: mov [[OUT_LO:[a-z0-9]+]], r1
; CHECK: movne [[OUT_LO]], [[REG1]]
ARM: Better codegen for 64-bit compares. This introduces a custom lowering for ISD::SETCCE (introduced in r253572) that allows us to emit a short code sequence for 64-bit compares. Before: push {r7, lr} cmp r0, r2 mov.w r0, #0 mov.w r12, #0 it hs movhs r0, #1 cmp r1, r3 it ge movge.w r12, #1 it eq moveq r12, r0 cmp.w r12, #0 bne .LBB1_2 @ BB#1: @ %bb1 bl f pop {r7, pc} .LBB1_2: @ %bb2 bl g pop {r7, pc} After: push {r7, lr} subs r0, r0, r2 sbcs.w r0, r1, r3 bge .LBB1_2 @ BB#1: @ %bb1 bl f pop {r7, pc} .LBB1_2: @ %bb2 bl g pop {r7, pc} Saves around 80KB in Chromium's libchrome.so. Some notes on this patch: - I don't much like the ARMISD::BRCOND and ARMISD::CMOV combines I introduced (nothing else needs them). However, they are necessary in order to avoid poor codegen, and they seem similar to existing combines in other backends (e.g. X86 combines (brcond (cmp (setcc Compare))) to (brcond Compare)). - No support for Thumb-1. This is in principle possible, but we'd need to implement ARMISD::SUBE for Thumb-1. Differential Revision: http://reviews.llvm.org/D15256 llvm-svn: 263962
2016-03-22 02:00:02 +08:00
; CHECK: strexd {{[a-z0-9]+}}, [[OUT_LO]], [[OUT_HI]]
Added atomic 64 min/max/umin/umax instrinsics support in the ARM backend. llvm-svn: 168886
2012-11-29 22:41:25 +08:00
; CHECK: cmp
; CHECK: bne
ARM: relax the atomic release barrier to "dmb ishst" on Swift Swift cores implement store barriers that are stronger than the ARM specification but weaker than general barriers. They are, in fact, just about enough to provide the ordering needed for atomic operations with release semantics. This patch makes use of that quirk. llvm-svn: 185527
2013-07-03 17:20:36 +08:00
; CHECK: dmb {{ish$}}
Fix 64-bit atomic operations in Thumb mode. The ARM and Thumb variants of LDREXD and STREXD have different constraints and take different operands. Previously the code expanding atomic operations didn't take this into account and asserted in Thumb mode. llvm-svn: 173780
2013-01-29 17:06:13 +08:00
Mass update to CodeGen tests to use CHECK-LABEL for labels corresponding to function definitions for more informative error messages. No functionality change and all updated tests passed locally. This update was done with the following bash script: find test/CodeGen -name "*.ll" | \ while read NAME; do echo "$NAME" if ! grep -q "^; *RUN: *llc.*debug" $NAME; then TEMP=`mktemp -t temp` cp $NAME $TEMP sed -n "s/^define [^@]*@\([A-Za-z0-9_]*\)(.*$/\1/p" < $NAME | \ while read FUNC; do sed -i '' "s/;\(.*\)\([A-Za-z0-9_-]*\):\( *\)$FUNC: *\$/;\1\2-LABEL:\3$FUNC:/g" $TEMP done sed -i '' "s/;\(.*\)-LABEL-LABEL:/;\1-LABEL:/" $TEMP sed -i '' "s/;\(.*\)-NEXT-LABEL:/;\1-NEXT:/" $TEMP sed -i '' "s/;\(.*\)-NOT-LABEL:/;\1-NOT:/" $TEMP sed -i '' "s/;\(.*\)-DAG-LABEL:/;\1-DAG:/" $TEMP mv $TEMP $NAME fi done llvm-svn: 186280
2013-07-14 14:24:09 +08:00
; CHECK-THUMB-LABEL: test10:
ARM: relax the atomic release barrier to "dmb ishst" on Swift Swift cores implement store barriers that are stronger than the ARM specification but weaker than general barriers. They are, in fact, just about enough to provide the ordering needed for atomic operations with release semantics. This patch makes use of that quirk. llvm-svn: 185527
2013-07-03 17:20:36 +08:00
; CHECK-THUMB: dmb {{ish$}}
Fix 64-bit atomic operations in Thumb mode. The ARM and Thumb variants of LDREXD and STREXD have different constraints and take different operands. Previously the code expanding atomic operations didn't take this into account and asserted in Thumb mode. llvm-svn: 173780
2013-01-29 17:06:13 +08:00
; CHECK-THUMB: ldrexd [[REG1:[a-z0-9]+]], [[REG2:[a-z0-9]+]]
ARM: Better codegen for 64-bit compares. This introduces a custom lowering for ISD::SETCCE (introduced in r253572) that allows us to emit a short code sequence for 64-bit compares. Before: push {r7, lr} cmp r0, r2 mov.w r0, #0 mov.w r12, #0 it hs movhs r0, #1 cmp r1, r3 it ge movge.w r12, #1 it eq moveq r12, r0 cmp.w r12, #0 bne .LBB1_2 @ BB#1: @ %bb1 bl f pop {r7, pc} .LBB1_2: @ %bb2 bl g pop {r7, pc} After: push {r7, lr} subs r0, r0, r2 sbcs.w r0, r1, r3 bge .LBB1_2 @ BB#1: @ %bb1 bl f pop {r7, pc} .LBB1_2: @ %bb2 bl g pop {r7, pc} Saves around 80KB in Chromium's libchrome.so. Some notes on this patch: - I don't much like the ARMISD::BRCOND and ARMISD::CMOV combines I introduced (nothing else needs them). However, they are necessary in order to avoid poor codegen, and they seem similar to existing combines in other backends (e.g. X86 combines (brcond (cmp (setcc Compare))) to (brcond Compare)). - No support for Thumb-1. This is in principle possible, but we'd need to implement ARMISD::SUBE for Thumb-1. Differential Revision: http://reviews.llvm.org/D15256 llvm-svn: 263962
2016-03-22 02:00:02 +08:00
; CHECK-THUMB: mov [[OUT_LO:[a-z0-9]+]], r2
; CHECK-THUMB-LE: subs.w {{[^,]+}}, r2, [[REG1]]
; CHECK-THUMB-BE: subs.w {{[^,]+}}, r3, [[REG2]]
; CHECK-THUMB-LE: sbcs.w {{[^,]+}}, r3, [[REG2]]
; CHECK-THUMB-BE: sbcs.w {{[^,]+}}, r2, [[REG1]]
; CHECK-THUMB: mov.w [[CMP:[a-z0-9]+]], #0
; CHECK-THUMB: movge.w [[CMP]], #1
; CHECK-THUMB: cmp.w [[CMP]], #0
; CHECK-THUMB: mov [[OUT_HI:[a-z0-9]+]], r3
ARM: expand atomic ldrex/strex loops in IR The previous situation where ATOMIC_LOAD_WHATEVER nodes were expanded at MachineInstr emission time had grown to be extremely large and involved, to account for the subtly different code needed for the various flavours (8/16/32/64 bit, cmpxchg/add/minmax). Moving this transformation into the IR clears up the code substantially, and makes future optimisations much easier: 1. an atomicrmw followed by using the *new* value can be more efficient. As an IR pass, simple CSE could handle this efficiently. 2. Making use of cmpxchg success/failure orderings only has to be done in one (simpler) place. 3. The common "cmpxchg; did we store?" idiom can be exposed to optimisation. I intend to gradually improve this situation within the ARM backend and make sure there are no hidden issues before moving the code out into CodeGen to be shared with (at least ARM64/AArch64, though I think PPC & Mips could benefit too). llvm-svn: 205525
2014-04-03 19:44:58 +08:00
; CHECK-THUMB: movne [[OUT_HI]], [[REG2]]
; CHECK-THUMB: movne [[OUT_LO]], [[REG1]]
ARM: Better codegen for 64-bit compares. This introduces a custom lowering for ISD::SETCCE (introduced in r253572) that allows us to emit a short code sequence for 64-bit compares. Before: push {r7, lr} cmp r0, r2 mov.w r0, #0 mov.w r12, #0 it hs movhs r0, #1 cmp r1, r3 it ge movge.w r12, #1 it eq moveq r12, r0 cmp.w r12, #0 bne .LBB1_2 @ BB#1: @ %bb1 bl f pop {r7, pc} .LBB1_2: @ %bb2 bl g pop {r7, pc} After: push {r7, lr} subs r0, r0, r2 sbcs.w r0, r1, r3 bge .LBB1_2 @ BB#1: @ %bb1 bl f pop {r7, pc} .LBB1_2: @ %bb2 bl g pop {r7, pc} Saves around 80KB in Chromium's libchrome.so. Some notes on this patch: - I don't much like the ARMISD::BRCOND and ARMISD::CMOV combines I introduced (nothing else needs them). However, they are necessary in order to avoid poor codegen, and they seem similar to existing combines in other backends (e.g. X86 combines (brcond (cmp (setcc Compare))) to (brcond Compare)). - No support for Thumb-1. This is in principle possible, but we'd need to implement ARMISD::SUBE for Thumb-1. Differential Revision: http://reviews.llvm.org/D15256 llvm-svn: 263962
2016-03-22 02:00:02 +08:00
; CHECK-THUMB: strexd {{[a-z0-9]+}}, [[OUT_LO]], [[OUT_HI]]
Fix 64-bit atomic operations in Thumb mode. The ARM and Thumb variants of LDREXD and STREXD have different constraints and take different operands. Previously the code expanding atomic operations didn't take this into account and asserted in Thumb mode. llvm-svn: 173780
2013-01-29 17:06:13 +08:00
; CHECK-THUMB: cmp
; CHECK-THUMB: bne
ARM: relax the atomic release barrier to "dmb ishst" on Swift Swift cores implement store barriers that are stronger than the ARM specification but weaker than general barriers. They are, in fact, just about enough to provide the ordering needed for atomic operations with release semantics. This patch makes use of that quirk. llvm-svn: 185527
2013-07-03 17:20:36 +08:00
; CHECK-THUMB: dmb {{ish$}}
Fix 64-bit atomic operations in Thumb mode. The ARM and Thumb variants of LDREXD and STREXD have different constraints and take different operands. Previously the code expanding atomic operations didn't take this into account and asserted in Thumb mode. llvm-svn: 173780
2013-01-29 17:06:13 +08:00
Added atomic 64 min/max/umin/umax instrinsics support in the ARM backend. llvm-svn: 168886
2012-11-29 22:41:25 +08:00
%r = atomicrmw min i64* %ptr, i64 %val seq_cst
ret i64 %r
}
define i64 @test11(i64* %ptr, i64 %val) {
Convert CodeGen/*/*.ll tests to use the new CHECK-LABEL for easier debugging. No functionality change and all tests pass after conversion. This was done with the following sed invocation to catch label lines demarking function boundaries: sed -i '' "s/^;\( *\)\([A-Z0-9_]*\):\( *\)test\([A-Za-z0-9_-]*\):\( *\)$/;\1\2-LABEL:\3test\4:\5/g" test/CodeGen/*/*.ll which was written conservatively to avoid false positives rather than false negatives. I scanned through all the changes and everything looks correct. llvm-svn: 186258
2013-07-14 04:38:47 +08:00
; CHECK-LABEL: test11:
ARM: relax the atomic release barrier to "dmb ishst" on Swift Swift cores implement store barriers that are stronger than the ARM specification but weaker than general barriers. They are, in fact, just about enough to provide the ordering needed for atomic operations with release semantics. This patch makes use of that quirk. llvm-svn: 185527
2013-07-03 17:20:36 +08:00
; CHECK: dmb {{ish$}}
Added atomic 64 min/max/umin/umax instrinsics support in the ARM backend. llvm-svn: 168886
2012-11-29 22:41:25 +08:00
; CHECK: ldrexd [[REG1:(r[0-9]?[02468])]], [[REG2:(r[0-9]?[13579])]]
ARM: expand atomic ldrex/strex loops in IR The previous situation where ATOMIC_LOAD_WHATEVER nodes were expanded at MachineInstr emission time had grown to be extremely large and involved, to account for the subtly different code needed for the various flavours (8/16/32/64 bit, cmpxchg/add/minmax). Moving this transformation into the IR clears up the code substantially, and makes future optimisations much easier: 1. an atomicrmw followed by using the *new* value can be more efficient. As an IR pass, simple CSE could handle this efficiently. 2. Making use of cmpxchg success/failure orderings only has to be done in one (simpler) place. 3. The common "cmpxchg; did we store?" idiom can be exposed to optimisation. I intend to gradually improve this situation within the ARM backend and make sure there are no hidden issues before moving the code out into CodeGen to be shared with (at least ARM64/AArch64, though I think PPC & Mips could benefit too). llvm-svn: 205525
2014-04-03 19:44:58 +08:00
; CHECK: mov [[OUT_HI:[a-z0-9]+]], r2
ARM: Better codegen for 64-bit compares. This introduces a custom lowering for ISD::SETCCE (introduced in r253572) that allows us to emit a short code sequence for 64-bit compares. Before: push {r7, lr} cmp r0, r2 mov.w r0, #0 mov.w r12, #0 it hs movhs r0, #1 cmp r1, r3 it ge movge.w r12, #1 it eq moveq r12, r0 cmp.w r12, #0 bne .LBB1_2 @ BB#1: @ %bb1 bl f pop {r7, pc} .LBB1_2: @ %bb2 bl g pop {r7, pc} After: push {r7, lr} subs r0, r0, r2 sbcs.w r0, r1, r3 bge .LBB1_2 @ BB#1: @ %bb1 bl f pop {r7, pc} .LBB1_2: @ %bb2 bl g pop {r7, pc} Saves around 80KB in Chromium's libchrome.so. Some notes on this patch: - I don't much like the ARMISD::BRCOND and ARMISD::CMOV combines I introduced (nothing else needs them). However, they are necessary in order to avoid poor codegen, and they seem similar to existing combines in other backends (e.g. X86 combines (brcond (cmp (setcc Compare))) to (brcond Compare)). - No support for Thumb-1. This is in principle possible, but we'd need to implement ARMISD::SUBE for Thumb-1. Differential Revision: http://reviews.llvm.org/D15256 llvm-svn: 263962
2016-03-22 02:00:02 +08:00
; CHECK-LE: subs {{[^,]+}}, r1, [[REG1]]
; CHECK-BE: subs {{[^,]+}}, r2, [[REG2]]
; CHECK-LE: sbcs {{[^,]+}}, r2, [[REG2]]
; CHECK-BE: sbcs {{[^,]+}}, r1, [[REG1]]
; CHECK: mov [[CMP:[a-z0-9]+]], #0
; CHECK: movwhs [[CMP]], #1
; CHECK: cmp [[CMP]], #0
ARM: expand atomic ldrex/strex loops in IR The previous situation where ATOMIC_LOAD_WHATEVER nodes were expanded at MachineInstr emission time had grown to be extremely large and involved, to account for the subtly different code needed for the various flavours (8/16/32/64 bit, cmpxchg/add/minmax). Moving this transformation into the IR clears up the code substantially, and makes future optimisations much easier: 1. an atomicrmw followed by using the *new* value can be more efficient. As an IR pass, simple CSE could handle this efficiently. 2. Making use of cmpxchg success/failure orderings only has to be done in one (simpler) place. 3. The common "cmpxchg; did we store?" idiom can be exposed to optimisation. I intend to gradually improve this situation within the ARM backend and make sure there are no hidden issues before moving the code out into CodeGen to be shared with (at least ARM64/AArch64, though I think PPC & Mips could benefit too). llvm-svn: 205525
2014-04-03 19:44:58 +08:00
; CHECK: movne [[OUT_HI]], [[REG2]]
; CHECK: mov [[OUT_LO:[a-z0-9]+]], r1
; CHECK: movne [[OUT_LO]], [[REG1]]
ARM: Better codegen for 64-bit compares. This introduces a custom lowering for ISD::SETCCE (introduced in r253572) that allows us to emit a short code sequence for 64-bit compares. Before: push {r7, lr} cmp r0, r2 mov.w r0, #0 mov.w r12, #0 it hs movhs r0, #1 cmp r1, r3 it ge movge.w r12, #1 it eq moveq r12, r0 cmp.w r12, #0 bne .LBB1_2 @ BB#1: @ %bb1 bl f pop {r7, pc} .LBB1_2: @ %bb2 bl g pop {r7, pc} After: push {r7, lr} subs r0, r0, r2 sbcs.w r0, r1, r3 bge .LBB1_2 @ BB#1: @ %bb1 bl f pop {r7, pc} .LBB1_2: @ %bb2 bl g pop {r7, pc} Saves around 80KB in Chromium's libchrome.so. Some notes on this patch: - I don't much like the ARMISD::BRCOND and ARMISD::CMOV combines I introduced (nothing else needs them). However, they are necessary in order to avoid poor codegen, and they seem similar to existing combines in other backends (e.g. X86 combines (brcond (cmp (setcc Compare))) to (brcond Compare)). - No support for Thumb-1. This is in principle possible, but we'd need to implement ARMISD::SUBE for Thumb-1. Differential Revision: http://reviews.llvm.org/D15256 llvm-svn: 263962
2016-03-22 02:00:02 +08:00
; CHECK: strexd {{[a-z0-9]+}}, [[OUT_LO]], [[OUT_HI]]
Added atomic 64 min/max/umin/umax instrinsics support in the ARM backend. llvm-svn: 168886
2012-11-29 22:41:25 +08:00
; CHECK: cmp
; CHECK: bne
ARM: relax the atomic release barrier to "dmb ishst" on Swift Swift cores implement store barriers that are stronger than the ARM specification but weaker than general barriers. They are, in fact, just about enough to provide the ordering needed for atomic operations with release semantics. This patch makes use of that quirk. llvm-svn: 185527
2013-07-03 17:20:36 +08:00
; CHECK: dmb {{ish$}}
Fix 64-bit atomic operations in Thumb mode. The ARM and Thumb variants of LDREXD and STREXD have different constraints and take different operands. Previously the code expanding atomic operations didn't take this into account and asserted in Thumb mode. llvm-svn: 173780
2013-01-29 17:06:13 +08:00
Mass update to CodeGen tests to use CHECK-LABEL for labels corresponding to function definitions for more informative error messages. No functionality change and all updated tests passed locally. This update was done with the following bash script: find test/CodeGen -name "*.ll" | \ while read NAME; do echo "$NAME" if ! grep -q "^; *RUN: *llc.*debug" $NAME; then TEMP=`mktemp -t temp` cp $NAME $TEMP sed -n "s/^define [^@]*@\([A-Za-z0-9_]*\)(.*$/\1/p" < $NAME | \ while read FUNC; do sed -i '' "s/;\(.*\)\([A-Za-z0-9_-]*\):\( *\)$FUNC: *\$/;\1\2-LABEL:\3$FUNC:/g" $TEMP done sed -i '' "s/;\(.*\)-LABEL-LABEL:/;\1-LABEL:/" $TEMP sed -i '' "s/;\(.*\)-NEXT-LABEL:/;\1-NEXT:/" $TEMP sed -i '' "s/;\(.*\)-NOT-LABEL:/;\1-NOT:/" $TEMP sed -i '' "s/;\(.*\)-DAG-LABEL:/;\1-DAG:/" $TEMP mv $TEMP $NAME fi done llvm-svn: 186280
2013-07-14 14:24:09 +08:00
; CHECK-THUMB-LABEL: test11:
ARM: relax the atomic release barrier to "dmb ishst" on Swift Swift cores implement store barriers that are stronger than the ARM specification but weaker than general barriers. They are, in fact, just about enough to provide the ordering needed for atomic operations with release semantics. This patch makes use of that quirk. llvm-svn: 185527
2013-07-03 17:20:36 +08:00
; CHECK-THUMB: dmb {{ish$}}
Fix 64-bit atomic operations in Thumb mode. The ARM and Thumb variants of LDREXD and STREXD have different constraints and take different operands. Previously the code expanding atomic operations didn't take this into account and asserted in Thumb mode. llvm-svn: 173780
2013-01-29 17:06:13 +08:00
; CHECK-THUMB: ldrexd [[REG1:[a-z0-9]+]], [[REG2:[a-z0-9]+]]
ARM: Better codegen for 64-bit compares. This introduces a custom lowering for ISD::SETCCE (introduced in r253572) that allows us to emit a short code sequence for 64-bit compares. Before: push {r7, lr} cmp r0, r2 mov.w r0, #0 mov.w r12, #0 it hs movhs r0, #1 cmp r1, r3 it ge movge.w r12, #1 it eq moveq r12, r0 cmp.w r12, #0 bne .LBB1_2 @ BB#1: @ %bb1 bl f pop {r7, pc} .LBB1_2: @ %bb2 bl g pop {r7, pc} After: push {r7, lr} subs r0, r0, r2 sbcs.w r0, r1, r3 bge .LBB1_2 @ BB#1: @ %bb1 bl f pop {r7, pc} .LBB1_2: @ %bb2 bl g pop {r7, pc} Saves around 80KB in Chromium's libchrome.so. Some notes on this patch: - I don't much like the ARMISD::BRCOND and ARMISD::CMOV combines I introduced (nothing else needs them). However, they are necessary in order to avoid poor codegen, and they seem similar to existing combines in other backends (e.g. X86 combines (brcond (cmp (setcc Compare))) to (brcond Compare)). - No support for Thumb-1. This is in principle possible, but we'd need to implement ARMISD::SUBE for Thumb-1. Differential Revision: http://reviews.llvm.org/D15256 llvm-svn: 263962
2016-03-22 02:00:02 +08:00
; CHECK-THUMB: mov [[OUT_LO:[a-z0-9]+]], r2
; CHECK-THUMB-LE: subs.w {{[^,]+}}, r2, [[REG1]]
; CHECK-THUMB-BE: subs.w {{[^,]+}}, r3, [[REG2]]
; CHECK-THUMB-LE: sbcs.w {{[^,]+}}, r3, [[REG2]]
; CHECK-THUMB-BE: sbcs.w {{[^,]+}}, r2, [[REG1]]
; CHECK-THUMB: mov.w [[CMP:[a-z0-9]+]], #0
; CHECK-THUMB: movhs.w [[CMP]], #1
; CHECK-THUMB: cmp.w [[CMP]], #0
; CHECK-THUMB: mov [[OUT_HI:[a-z0-9]+]], r3
ARM: expand atomic ldrex/strex loops in IR The previous situation where ATOMIC_LOAD_WHATEVER nodes were expanded at MachineInstr emission time had grown to be extremely large and involved, to account for the subtly different code needed for the various flavours (8/16/32/64 bit, cmpxchg/add/minmax). Moving this transformation into the IR clears up the code substantially, and makes future optimisations much easier: 1. an atomicrmw followed by using the *new* value can be more efficient. As an IR pass, simple CSE could handle this efficiently. 2. Making use of cmpxchg success/failure orderings only has to be done in one (simpler) place. 3. The common "cmpxchg; did we store?" idiom can be exposed to optimisation. I intend to gradually improve this situation within the ARM backend and make sure there are no hidden issues before moving the code out into CodeGen to be shared with (at least ARM64/AArch64, though I think PPC & Mips could benefit too). llvm-svn: 205525
2014-04-03 19:44:58 +08:00
; CHECK-THUMB: movne [[OUT_HI]], [[REG2]]
; CHECK-THUMB: movne [[OUT_LO]], [[REG1]]
ARM: Better codegen for 64-bit compares. This introduces a custom lowering for ISD::SETCCE (introduced in r253572) that allows us to emit a short code sequence for 64-bit compares. Before: push {r7, lr} cmp r0, r2 mov.w r0, #0 mov.w r12, #0 it hs movhs r0, #1 cmp r1, r3 it ge movge.w r12, #1 it eq moveq r12, r0 cmp.w r12, #0 bne .LBB1_2 @ BB#1: @ %bb1 bl f pop {r7, pc} .LBB1_2: @ %bb2 bl g pop {r7, pc} After: push {r7, lr} subs r0, r0, r2 sbcs.w r0, r1, r3 bge .LBB1_2 @ BB#1: @ %bb1 bl f pop {r7, pc} .LBB1_2: @ %bb2 bl g pop {r7, pc} Saves around 80KB in Chromium's libchrome.so. Some notes on this patch: - I don't much like the ARMISD::BRCOND and ARMISD::CMOV combines I introduced (nothing else needs them). However, they are necessary in order to avoid poor codegen, and they seem similar to existing combines in other backends (e.g. X86 combines (brcond (cmp (setcc Compare))) to (brcond Compare)). - No support for Thumb-1. This is in principle possible, but we'd need to implement ARMISD::SUBE for Thumb-1. Differential Revision: http://reviews.llvm.org/D15256 llvm-svn: 263962
2016-03-22 02:00:02 +08:00
; CHECK-THUMB: strexd {{[a-z0-9]+}}, [[OUT_LO]], [[OUT_HI]]
Fix 64-bit atomic operations in Thumb mode. The ARM and Thumb variants of LDREXD and STREXD have different constraints and take different operands. Previously the code expanding atomic operations didn't take this into account and asserted in Thumb mode. llvm-svn: 173780
2013-01-29 17:06:13 +08:00
; CHECK-THUMB: cmp
; CHECK-THUMB: bne
ARM: relax the atomic release barrier to "dmb ishst" on Swift Swift cores implement store barriers that are stronger than the ARM specification but weaker than general barriers. They are, in fact, just about enough to provide the ordering needed for atomic operations with release semantics. This patch makes use of that quirk. llvm-svn: 185527
2013-07-03 17:20:36 +08:00
; CHECK-THUMB: dmb {{ish$}}
Fix 64-bit atomic operations in Thumb mode. The ARM and Thumb variants of LDREXD and STREXD have different constraints and take different operands. Previously the code expanding atomic operations didn't take this into account and asserted in Thumb mode. llvm-svn: 173780
2013-01-29 17:06:13 +08:00
Added atomic 64 min/max/umin/umax instrinsics support in the ARM backend. llvm-svn: 168886
2012-11-29 22:41:25 +08:00
%r = atomicrmw umin i64* %ptr, i64 %val seq_cst
ret i64 %r
}
define i64 @test12(i64* %ptr, i64 %val) {
Convert CodeGen/*/*.ll tests to use the new CHECK-LABEL for easier debugging. No functionality change and all tests pass after conversion. This was done with the following sed invocation to catch label lines demarking function boundaries: sed -i '' "s/^;\( *\)\([A-Z0-9_]*\):\( *\)test\([A-Za-z0-9_-]*\):\( *\)$/;\1\2-LABEL:\3test\4:\5/g" test/CodeGen/*/*.ll which was written conservatively to avoid false positives rather than false negatives. I scanned through all the changes and everything looks correct. llvm-svn: 186258
2013-07-14 04:38:47 +08:00
; CHECK-LABEL: test12:
ARM: relax the atomic release barrier to "dmb ishst" on Swift Swift cores implement store barriers that are stronger than the ARM specification but weaker than general barriers. They are, in fact, just about enough to provide the ordering needed for atomic operations with release semantics. This patch makes use of that quirk. llvm-svn: 185527
2013-07-03 17:20:36 +08:00
; CHECK: dmb {{ish$}}
Added atomic 64 min/max/umin/umax instrinsics support in the ARM backend. llvm-svn: 168886
2012-11-29 22:41:25 +08:00
; CHECK: ldrexd [[REG1:(r[0-9]?[02468])]], [[REG2:(r[0-9]?[13579])]]
ARM: expand atomic ldrex/strex loops in IR The previous situation where ATOMIC_LOAD_WHATEVER nodes were expanded at MachineInstr emission time had grown to be extremely large and involved, to account for the subtly different code needed for the various flavours (8/16/32/64 bit, cmpxchg/add/minmax). Moving this transformation into the IR clears up the code substantially, and makes future optimisations much easier: 1. an atomicrmw followed by using the *new* value can be more efficient. As an IR pass, simple CSE could handle this efficiently. 2. Making use of cmpxchg success/failure orderings only has to be done in one (simpler) place. 3. The common "cmpxchg; did we store?" idiom can be exposed to optimisation. I intend to gradually improve this situation within the ARM backend and make sure there are no hidden issues before moving the code out into CodeGen to be shared with (at least ARM64/AArch64, though I think PPC & Mips could benefit too). llvm-svn: 205525
2014-04-03 19:44:58 +08:00
; CHECK: mov [[OUT_HI:[a-z0-9]+]], r2
ARM: Better codegen for 64-bit compares. This introduces a custom lowering for ISD::SETCCE (introduced in r253572) that allows us to emit a short code sequence for 64-bit compares. Before: push {r7, lr} cmp r0, r2 mov.w r0, #0 mov.w r12, #0 it hs movhs r0, #1 cmp r1, r3 it ge movge.w r12, #1 it eq moveq r12, r0 cmp.w r12, #0 bne .LBB1_2 @ BB#1: @ %bb1 bl f pop {r7, pc} .LBB1_2: @ %bb2 bl g pop {r7, pc} After: push {r7, lr} subs r0, r0, r2 sbcs.w r0, r1, r3 bge .LBB1_2 @ BB#1: @ %bb1 bl f pop {r7, pc} .LBB1_2: @ %bb2 bl g pop {r7, pc} Saves around 80KB in Chromium's libchrome.so. Some notes on this patch: - I don't much like the ARMISD::BRCOND and ARMISD::CMOV combines I introduced (nothing else needs them). However, they are necessary in order to avoid poor codegen, and they seem similar to existing combines in other backends (e.g. X86 combines (brcond (cmp (setcc Compare))) to (brcond Compare)). - No support for Thumb-1. This is in principle possible, but we'd need to implement ARMISD::SUBE for Thumb-1. Differential Revision: http://reviews.llvm.org/D15256 llvm-svn: 263962
2016-03-22 02:00:02 +08:00
; CHECK-LE: subs {{[^,]+}}, r1, [[REG1]]
; CHECK-BE: subs {{[^,]+}}, r2, [[REG2]]
; CHECK-LE: sbcs {{[^,]+}}, r2, [[REG2]]
; CHECK-BE: sbcs {{[^,]+}}, r1, [[REG1]]
; CHECK: mov [[CMP:[a-z0-9]+]], #0
; CHECK: movwlt [[CMP]], #1
; CHECK: cmp [[CMP]], #0
ARM: expand atomic ldrex/strex loops in IR The previous situation where ATOMIC_LOAD_WHATEVER nodes were expanded at MachineInstr emission time had grown to be extremely large and involved, to account for the subtly different code needed for the various flavours (8/16/32/64 bit, cmpxchg/add/minmax). Moving this transformation into the IR clears up the code substantially, and makes future optimisations much easier: 1. an atomicrmw followed by using the *new* value can be more efficient. As an IR pass, simple CSE could handle this efficiently. 2. Making use of cmpxchg success/failure orderings only has to be done in one (simpler) place. 3. The common "cmpxchg; did we store?" idiom can be exposed to optimisation. I intend to gradually improve this situation within the ARM backend and make sure there are no hidden issues before moving the code out into CodeGen to be shared with (at least ARM64/AArch64, though I think PPC & Mips could benefit too). llvm-svn: 205525
2014-04-03 19:44:58 +08:00
; CHECK: movne [[OUT_HI]], [[REG2]]
; CHECK: mov [[OUT_LO:[a-z0-9]+]], r1
; CHECK: movne [[OUT_LO]], [[REG1]]
ARM: Better codegen for 64-bit compares. This introduces a custom lowering for ISD::SETCCE (introduced in r253572) that allows us to emit a short code sequence for 64-bit compares. Before: push {r7, lr} cmp r0, r2 mov.w r0, #0 mov.w r12, #0 it hs movhs r0, #1 cmp r1, r3 it ge movge.w r12, #1 it eq moveq r12, r0 cmp.w r12, #0 bne .LBB1_2 @ BB#1: @ %bb1 bl f pop {r7, pc} .LBB1_2: @ %bb2 bl g pop {r7, pc} After: push {r7, lr} subs r0, r0, r2 sbcs.w r0, r1, r3 bge .LBB1_2 @ BB#1: @ %bb1 bl f pop {r7, pc} .LBB1_2: @ %bb2 bl g pop {r7, pc} Saves around 80KB in Chromium's libchrome.so. Some notes on this patch: - I don't much like the ARMISD::BRCOND and ARMISD::CMOV combines I introduced (nothing else needs them). However, they are necessary in order to avoid poor codegen, and they seem similar to existing combines in other backends (e.g. X86 combines (brcond (cmp (setcc Compare))) to (brcond Compare)). - No support for Thumb-1. This is in principle possible, but we'd need to implement ARMISD::SUBE for Thumb-1. Differential Revision: http://reviews.llvm.org/D15256 llvm-svn: 263962
2016-03-22 02:00:02 +08:00
; CHECK: strexd {{[a-z0-9]+}}, [[OUT_LO]], [[OUT_HI]]
Added atomic 64 min/max/umin/umax instrinsics support in the ARM backend. llvm-svn: 168886
2012-11-29 22:41:25 +08:00
; CHECK: cmp
; CHECK: bne
ARM: relax the atomic release barrier to "dmb ishst" on Swift Swift cores implement store barriers that are stronger than the ARM specification but weaker than general barriers. They are, in fact, just about enough to provide the ordering needed for atomic operations with release semantics. This patch makes use of that quirk. llvm-svn: 185527
2013-07-03 17:20:36 +08:00
; CHECK: dmb {{ish$}}
Fix 64-bit atomic operations in Thumb mode. The ARM and Thumb variants of LDREXD and STREXD have different constraints and take different operands. Previously the code expanding atomic operations didn't take this into account and asserted in Thumb mode. llvm-svn: 173780
2013-01-29 17:06:13 +08:00
Mass update to CodeGen tests to use CHECK-LABEL for labels corresponding to function definitions for more informative error messages. No functionality change and all updated tests passed locally. This update was done with the following bash script: find test/CodeGen -name "*.ll" | \ while read NAME; do echo "$NAME" if ! grep -q "^; *RUN: *llc.*debug" $NAME; then TEMP=`mktemp -t temp` cp $NAME $TEMP sed -n "s/^define [^@]*@\([A-Za-z0-9_]*\)(.*$/\1/p" < $NAME | \ while read FUNC; do sed -i '' "s/;\(.*\)\([A-Za-z0-9_-]*\):\( *\)$FUNC: *\$/;\1\2-LABEL:\3$FUNC:/g" $TEMP done sed -i '' "s/;\(.*\)-LABEL-LABEL:/;\1-LABEL:/" $TEMP sed -i '' "s/;\(.*\)-NEXT-LABEL:/;\1-NEXT:/" $TEMP sed -i '' "s/;\(.*\)-NOT-LABEL:/;\1-NOT:/" $TEMP sed -i '' "s/;\(.*\)-DAG-LABEL:/;\1-DAG:/" $TEMP mv $TEMP $NAME fi done llvm-svn: 186280
2013-07-14 14:24:09 +08:00
; CHECK-THUMB-LABEL: test12:
ARM: relax the atomic release barrier to "dmb ishst" on Swift Swift cores implement store barriers that are stronger than the ARM specification but weaker than general barriers. They are, in fact, just about enough to provide the ordering needed for atomic operations with release semantics. This patch makes use of that quirk. llvm-svn: 185527
2013-07-03 17:20:36 +08:00
; CHECK-THUMB: dmb {{ish$}}
Fix 64-bit atomic operations in Thumb mode. The ARM and Thumb variants of LDREXD and STREXD have different constraints and take different operands. Previously the code expanding atomic operations didn't take this into account and asserted in Thumb mode. llvm-svn: 173780
2013-01-29 17:06:13 +08:00
; CHECK-THUMB: ldrexd [[REG1:[a-z0-9]+]], [[REG2:[a-z0-9]+]]
ARM: Better codegen for 64-bit compares. This introduces a custom lowering for ISD::SETCCE (introduced in r253572) that allows us to emit a short code sequence for 64-bit compares. Before: push {r7, lr} cmp r0, r2 mov.w r0, #0 mov.w r12, #0 it hs movhs r0, #1 cmp r1, r3 it ge movge.w r12, #1 it eq moveq r12, r0 cmp.w r12, #0 bne .LBB1_2 @ BB#1: @ %bb1 bl f pop {r7, pc} .LBB1_2: @ %bb2 bl g pop {r7, pc} After: push {r7, lr} subs r0, r0, r2 sbcs.w r0, r1, r3 bge .LBB1_2 @ BB#1: @ %bb1 bl f pop {r7, pc} .LBB1_2: @ %bb2 bl g pop {r7, pc} Saves around 80KB in Chromium's libchrome.so. Some notes on this patch: - I don't much like the ARMISD::BRCOND and ARMISD::CMOV combines I introduced (nothing else needs them). However, they are necessary in order to avoid poor codegen, and they seem similar to existing combines in other backends (e.g. X86 combines (brcond (cmp (setcc Compare))) to (brcond Compare)). - No support for Thumb-1. This is in principle possible, but we'd need to implement ARMISD::SUBE for Thumb-1. Differential Revision: http://reviews.llvm.org/D15256 llvm-svn: 263962
2016-03-22 02:00:02 +08:00
; CHECK-THUMB: mov [[OUT_LO:[a-z0-9]+]], r2
; CHECK-THUMB-LE: subs.w {{[^,]+}}, r2, [[REG1]]
; CHECK-THUMB-BE: subs.w {{[^,]+}}, r3, [[REG2]]
; CHECK-THUMB-LE: sbcs.w {{[^,]+}}, r3, [[REG2]]
; CHECK-THUMB-BE: sbcs.w {{[^,]+}}, r2, [[REG1]]
; CHECK-THUMB: mov.w [[CMP:[a-z0-9]+]], #0
; CHECK-THUMB: movlt.w [[CMP]], #1
; CHECK-THUMB: cmp.w [[CMP]], #0
; CHECK-THUMB: mov [[OUT_HI:[a-z0-9]+]], r3
ARM: expand atomic ldrex/strex loops in IR The previous situation where ATOMIC_LOAD_WHATEVER nodes were expanded at MachineInstr emission time had grown to be extremely large and involved, to account for the subtly different code needed for the various flavours (8/16/32/64 bit, cmpxchg/add/minmax). Moving this transformation into the IR clears up the code substantially, and makes future optimisations much easier: 1. an atomicrmw followed by using the *new* value can be more efficient. As an IR pass, simple CSE could handle this efficiently. 2. Making use of cmpxchg success/failure orderings only has to be done in one (simpler) place. 3. The common "cmpxchg; did we store?" idiom can be exposed to optimisation. I intend to gradually improve this situation within the ARM backend and make sure there are no hidden issues before moving the code out into CodeGen to be shared with (at least ARM64/AArch64, though I think PPC & Mips could benefit too). llvm-svn: 205525
2014-04-03 19:44:58 +08:00
; CHECK-THUMB: movne [[OUT_HI]], [[REG2]]
; CHECK-THUMB: movne [[OUT_LO]], [[REG1]]
ARM: Better codegen for 64-bit compares. This introduces a custom lowering for ISD::SETCCE (introduced in r253572) that allows us to emit a short code sequence for 64-bit compares. Before: push {r7, lr} cmp r0, r2 mov.w r0, #0 mov.w r12, #0 it hs movhs r0, #1 cmp r1, r3 it ge movge.w r12, #1 it eq moveq r12, r0 cmp.w r12, #0 bne .LBB1_2 @ BB#1: @ %bb1 bl f pop {r7, pc} .LBB1_2: @ %bb2 bl g pop {r7, pc} After: push {r7, lr} subs r0, r0, r2 sbcs.w r0, r1, r3 bge .LBB1_2 @ BB#1: @ %bb1 bl f pop {r7, pc} .LBB1_2: @ %bb2 bl g pop {r7, pc} Saves around 80KB in Chromium's libchrome.so. Some notes on this patch: - I don't much like the ARMISD::BRCOND and ARMISD::CMOV combines I introduced (nothing else needs them). However, they are necessary in order to avoid poor codegen, and they seem similar to existing combines in other backends (e.g. X86 combines (brcond (cmp (setcc Compare))) to (brcond Compare)). - No support for Thumb-1. This is in principle possible, but we'd need to implement ARMISD::SUBE for Thumb-1. Differential Revision: http://reviews.llvm.org/D15256 llvm-svn: 263962
2016-03-22 02:00:02 +08:00
; CHECK-THUMB: strexd {{[a-z0-9]+}}, [[OUT_LO]], [[OUT_HI]]
Fix 64-bit atomic operations in Thumb mode. The ARM and Thumb variants of LDREXD and STREXD have different constraints and take different operands. Previously the code expanding atomic operations didn't take this into account and asserted in Thumb mode. llvm-svn: 173780
2013-01-29 17:06:13 +08:00
; CHECK-THUMB: cmp
; CHECK-THUMB: bne
ARM: relax the atomic release barrier to "dmb ishst" on Swift Swift cores implement store barriers that are stronger than the ARM specification but weaker than general barriers. They are, in fact, just about enough to provide the ordering needed for atomic operations with release semantics. This patch makes use of that quirk. llvm-svn: 185527
2013-07-03 17:20:36 +08:00
; CHECK-THUMB: dmb {{ish$}}
Fix 64-bit atomic operations in Thumb mode. The ARM and Thumb variants of LDREXD and STREXD have different constraints and take different operands. Previously the code expanding atomic operations didn't take this into account and asserted in Thumb mode. llvm-svn: 173780
2013-01-29 17:06:13 +08:00
Added atomic 64 min/max/umin/umax instrinsics support in the ARM backend. llvm-svn: 168886
2012-11-29 22:41:25 +08:00
%r = atomicrmw max i64* %ptr, i64 %val seq_cst
ret i64 %r
}
define i64 @test13(i64* %ptr, i64 %val) {
Convert CodeGen/*/*.ll tests to use the new CHECK-LABEL for easier debugging. No functionality change and all tests pass after conversion. This was done with the following sed invocation to catch label lines demarking function boundaries: sed -i '' "s/^;\( *\)\([A-Z0-9_]*\):\( *\)test\([A-Za-z0-9_-]*\):\( *\)$/;\1\2-LABEL:\3test\4:\5/g" test/CodeGen/*/*.ll which was written conservatively to avoid false positives rather than false negatives. I scanned through all the changes and everything looks correct. llvm-svn: 186258
2013-07-14 04:38:47 +08:00
; CHECK-LABEL: test13:
ARM: relax the atomic release barrier to "dmb ishst" on Swift Swift cores implement store barriers that are stronger than the ARM specification but weaker than general barriers. They are, in fact, just about enough to provide the ordering needed for atomic operations with release semantics. This patch makes use of that quirk. llvm-svn: 185527
2013-07-03 17:20:36 +08:00
; CHECK: dmb {{ish$}}
Added atomic 64 min/max/umin/umax instrinsics support in the ARM backend. llvm-svn: 168886
2012-11-29 22:41:25 +08:00
; CHECK: ldrexd [[REG1:(r[0-9]?[02468])]], [[REG2:(r[0-9]?[13579])]]
ARM: expand atomic ldrex/strex loops in IR The previous situation where ATOMIC_LOAD_WHATEVER nodes were expanded at MachineInstr emission time had grown to be extremely large and involved, to account for the subtly different code needed for the various flavours (8/16/32/64 bit, cmpxchg/add/minmax). Moving this transformation into the IR clears up the code substantially, and makes future optimisations much easier: 1. an atomicrmw followed by using the *new* value can be more efficient. As an IR pass, simple CSE could handle this efficiently. 2. Making use of cmpxchg success/failure orderings only has to be done in one (simpler) place. 3. The common "cmpxchg; did we store?" idiom can be exposed to optimisation. I intend to gradually improve this situation within the ARM backend and make sure there are no hidden issues before moving the code out into CodeGen to be shared with (at least ARM64/AArch64, though I think PPC & Mips could benefit too). llvm-svn: 205525
2014-04-03 19:44:58 +08:00
; CHECK: mov [[OUT_HI:[a-z0-9]+]], r2
ARM: Better codegen for 64-bit compares. This introduces a custom lowering for ISD::SETCCE (introduced in r253572) that allows us to emit a short code sequence for 64-bit compares. Before: push {r7, lr} cmp r0, r2 mov.w r0, #0 mov.w r12, #0 it hs movhs r0, #1 cmp r1, r3 it ge movge.w r12, #1 it eq moveq r12, r0 cmp.w r12, #0 bne .LBB1_2 @ BB#1: @ %bb1 bl f pop {r7, pc} .LBB1_2: @ %bb2 bl g pop {r7, pc} After: push {r7, lr} subs r0, r0, r2 sbcs.w r0, r1, r3 bge .LBB1_2 @ BB#1: @ %bb1 bl f pop {r7, pc} .LBB1_2: @ %bb2 bl g pop {r7, pc} Saves around 80KB in Chromium's libchrome.so. Some notes on this patch: - I don't much like the ARMISD::BRCOND and ARMISD::CMOV combines I introduced (nothing else needs them). However, they are necessary in order to avoid poor codegen, and they seem similar to existing combines in other backends (e.g. X86 combines (brcond (cmp (setcc Compare))) to (brcond Compare)). - No support for Thumb-1. This is in principle possible, but we'd need to implement ARMISD::SUBE for Thumb-1. Differential Revision: http://reviews.llvm.org/D15256 llvm-svn: 263962
2016-03-22 02:00:02 +08:00
; CHECK-LE: subs {{[^,]+}}, r1, [[REG1]]
; CHECK-BE: subs {{[^,]+}}, r2, [[REG2]]
; CHECK-LE: sbcs {{[^,]+}}, r2, [[REG2]]
; CHECK-BE: sbcs {{[^,]+}}, r1, [[REG1]]
; CHECK: mov [[CMP:[a-z0-9]+]], #0
; CHECK: movwlo [[CMP]], #1
; CHECK: cmp [[CMP]], #0
ARM: expand atomic ldrex/strex loops in IR The previous situation where ATOMIC_LOAD_WHATEVER nodes were expanded at MachineInstr emission time had grown to be extremely large and involved, to account for the subtly different code needed for the various flavours (8/16/32/64 bit, cmpxchg/add/minmax). Moving this transformation into the IR clears up the code substantially, and makes future optimisations much easier: 1. an atomicrmw followed by using the *new* value can be more efficient. As an IR pass, simple CSE could handle this efficiently. 2. Making use of cmpxchg success/failure orderings only has to be done in one (simpler) place. 3. The common "cmpxchg; did we store?" idiom can be exposed to optimisation. I intend to gradually improve this situation within the ARM backend and make sure there are no hidden issues before moving the code out into CodeGen to be shared with (at least ARM64/AArch64, though I think PPC & Mips could benefit too). llvm-svn: 205525
2014-04-03 19:44:58 +08:00
; CHECK: movne [[OUT_HI]], [[REG2]]
; CHECK: mov [[OUT_LO:[a-z0-9]+]], r1
; CHECK: movne [[OUT_LO]], [[REG1]]
ARM: Better codegen for 64-bit compares. This introduces a custom lowering for ISD::SETCCE (introduced in r253572) that allows us to emit a short code sequence for 64-bit compares. Before: push {r7, lr} cmp r0, r2 mov.w r0, #0 mov.w r12, #0 it hs movhs r0, #1 cmp r1, r3 it ge movge.w r12, #1 it eq moveq r12, r0 cmp.w r12, #0 bne .LBB1_2 @ BB#1: @ %bb1 bl f pop {r7, pc} .LBB1_2: @ %bb2 bl g pop {r7, pc} After: push {r7, lr} subs r0, r0, r2 sbcs.w r0, r1, r3 bge .LBB1_2 @ BB#1: @ %bb1 bl f pop {r7, pc} .LBB1_2: @ %bb2 bl g pop {r7, pc} Saves around 80KB in Chromium's libchrome.so. Some notes on this patch: - I don't much like the ARMISD::BRCOND and ARMISD::CMOV combines I introduced (nothing else needs them). However, they are necessary in order to avoid poor codegen, and they seem similar to existing combines in other backends (e.g. X86 combines (brcond (cmp (setcc Compare))) to (brcond Compare)). - No support for Thumb-1. This is in principle possible, but we'd need to implement ARMISD::SUBE for Thumb-1. Differential Revision: http://reviews.llvm.org/D15256 llvm-svn: 263962
2016-03-22 02:00:02 +08:00
; CHECK: strexd {{[a-z0-9]+}}, [[OUT_LO]], [[OUT_HI]]
Added atomic 64 min/max/umin/umax instrinsics support in the ARM backend. llvm-svn: 168886
2012-11-29 22:41:25 +08:00
; CHECK: cmp
; CHECK: bne
ARM: relax the atomic release barrier to "dmb ishst" on Swift Swift cores implement store barriers that are stronger than the ARM specification but weaker than general barriers. They are, in fact, just about enough to provide the ordering needed for atomic operations with release semantics. This patch makes use of that quirk. llvm-svn: 185527
2013-07-03 17:20:36 +08:00
; CHECK: dmb {{ish$}}
Fix 64-bit atomic operations in Thumb mode. The ARM and Thumb variants of LDREXD and STREXD have different constraints and take different operands. Previously the code expanding atomic operations didn't take this into account and asserted in Thumb mode. llvm-svn: 173780
2013-01-29 17:06:13 +08:00
Mass update to CodeGen tests to use CHECK-LABEL for labels corresponding to function definitions for more informative error messages. No functionality change and all updated tests passed locally. This update was done with the following bash script: find test/CodeGen -name "*.ll" | \ while read NAME; do echo "$NAME" if ! grep -q "^; *RUN: *llc.*debug" $NAME; then TEMP=`mktemp -t temp` cp $NAME $TEMP sed -n "s/^define [^@]*@\([A-Za-z0-9_]*\)(.*$/\1/p" < $NAME | \ while read FUNC; do sed -i '' "s/;\(.*\)\([A-Za-z0-9_-]*\):\( *\)$FUNC: *\$/;\1\2-LABEL:\3$FUNC:/g" $TEMP done sed -i '' "s/;\(.*\)-LABEL-LABEL:/;\1-LABEL:/" $TEMP sed -i '' "s/;\(.*\)-NEXT-LABEL:/;\1-NEXT:/" $TEMP sed -i '' "s/;\(.*\)-NOT-LABEL:/;\1-NOT:/" $TEMP sed -i '' "s/;\(.*\)-DAG-LABEL:/;\1-DAG:/" $TEMP mv $TEMP $NAME fi done llvm-svn: 186280
2013-07-14 14:24:09 +08:00
; CHECK-THUMB-LABEL: test13:
ARM: relax the atomic release barrier to "dmb ishst" on Swift Swift cores implement store barriers that are stronger than the ARM specification but weaker than general barriers. They are, in fact, just about enough to provide the ordering needed for atomic operations with release semantics. This patch makes use of that quirk. llvm-svn: 185527
2013-07-03 17:20:36 +08:00
; CHECK-THUMB: dmb {{ish$}}
Fix 64-bit atomic operations in Thumb mode. The ARM and Thumb variants of LDREXD and STREXD have different constraints and take different operands. Previously the code expanding atomic operations didn't take this into account and asserted in Thumb mode. llvm-svn: 173780
2013-01-29 17:06:13 +08:00
; CHECK-THUMB: ldrexd [[REG1:[a-z0-9]+]], [[REG2:[a-z0-9]+]]
ARM: Better codegen for 64-bit compares. This introduces a custom lowering for ISD::SETCCE (introduced in r253572) that allows us to emit a short code sequence for 64-bit compares. Before: push {r7, lr} cmp r0, r2 mov.w r0, #0 mov.w r12, #0 it hs movhs r0, #1 cmp r1, r3 it ge movge.w r12, #1 it eq moveq r12, r0 cmp.w r12, #0 bne .LBB1_2 @ BB#1: @ %bb1 bl f pop {r7, pc} .LBB1_2: @ %bb2 bl g pop {r7, pc} After: push {r7, lr} subs r0, r0, r2 sbcs.w r0, r1, r3 bge .LBB1_2 @ BB#1: @ %bb1 bl f pop {r7, pc} .LBB1_2: @ %bb2 bl g pop {r7, pc} Saves around 80KB in Chromium's libchrome.so. Some notes on this patch: - I don't much like the ARMISD::BRCOND and ARMISD::CMOV combines I introduced (nothing else needs them). However, they are necessary in order to avoid poor codegen, and they seem similar to existing combines in other backends (e.g. X86 combines (brcond (cmp (setcc Compare))) to (brcond Compare)). - No support for Thumb-1. This is in principle possible, but we'd need to implement ARMISD::SUBE for Thumb-1. Differential Revision: http://reviews.llvm.org/D15256 llvm-svn: 263962
2016-03-22 02:00:02 +08:00
; CHECK-THUMB: mov [[OUT_LO:[a-z0-9]+]], r2
; CHECK-THUMB-LE: subs.w {{[^,]+}}, r2, [[REG1]]
; CHECK-THUMB-BE: subs.w {{[^,]+}}, r3, [[REG2]]
; CHECK-THUMB-LE: sbcs.w {{[^,]+}}, r3, [[REG2]]
; CHECK-THUMB-BE: sbcs.w {{[^,]+}}, r2, [[REG1]]
; CHECK-THUMB: mov.w [[CMP:[a-z0-9]+]], #0
; CHECK-THUMB: movlo.w [[CMP]], #1
; CHECK-THUMB: cmp.w [[CMP]], #0
; CHECK-THUMB: mov [[OUT_HI:[a-z0-9]+]], r3
ARM: expand atomic ldrex/strex loops in IR The previous situation where ATOMIC_LOAD_WHATEVER nodes were expanded at MachineInstr emission time had grown to be extremely large and involved, to account for the subtly different code needed for the various flavours (8/16/32/64 bit, cmpxchg/add/minmax). Moving this transformation into the IR clears up the code substantially, and makes future optimisations much easier: 1. an atomicrmw followed by using the *new* value can be more efficient. As an IR pass, simple CSE could handle this efficiently. 2. Making use of cmpxchg success/failure orderings only has to be done in one (simpler) place. 3. The common "cmpxchg; did we store?" idiom can be exposed to optimisation. I intend to gradually improve this situation within the ARM backend and make sure there are no hidden issues before moving the code out into CodeGen to be shared with (at least ARM64/AArch64, though I think PPC & Mips could benefit too). llvm-svn: 205525
2014-04-03 19:44:58 +08:00
; CHECK-THUMB: movne [[OUT_HI]], [[REG2]]
; CHECK-THUMB: movne [[OUT_LO]], [[REG1]]
ARM: Better codegen for 64-bit compares. This introduces a custom lowering for ISD::SETCCE (introduced in r253572) that allows us to emit a short code sequence for 64-bit compares. Before: push {r7, lr} cmp r0, r2 mov.w r0, #0 mov.w r12, #0 it hs movhs r0, #1 cmp r1, r3 it ge movge.w r12, #1 it eq moveq r12, r0 cmp.w r12, #0 bne .LBB1_2 @ BB#1: @ %bb1 bl f pop {r7, pc} .LBB1_2: @ %bb2 bl g pop {r7, pc} After: push {r7, lr} subs r0, r0, r2 sbcs.w r0, r1, r3 bge .LBB1_2 @ BB#1: @ %bb1 bl f pop {r7, pc} .LBB1_2: @ %bb2 bl g pop {r7, pc} Saves around 80KB in Chromium's libchrome.so. Some notes on this patch: - I don't much like the ARMISD::BRCOND and ARMISD::CMOV combines I introduced (nothing else needs them). However, they are necessary in order to avoid poor codegen, and they seem similar to existing combines in other backends (e.g. X86 combines (brcond (cmp (setcc Compare))) to (brcond Compare)). - No support for Thumb-1. This is in principle possible, but we'd need to implement ARMISD::SUBE for Thumb-1. Differential Revision: http://reviews.llvm.org/D15256 llvm-svn: 263962
2016-03-22 02:00:02 +08:00
; CHECK-THUMB: strexd {{[a-z0-9]+}}, [[OUT_LO]], [[OUT_HI]]
Fix 64-bit atomic operations in Thumb mode. The ARM and Thumb variants of LDREXD and STREXD have different constraints and take different operands. Previously the code expanding atomic operations didn't take this into account and asserted in Thumb mode. llvm-svn: 173780
2013-01-29 17:06:13 +08:00
; CHECK-THUMB: cmp
; CHECK-THUMB: bne
ARM: relax the atomic release barrier to "dmb ishst" on Swift Swift cores implement store barriers that are stronger than the ARM specification but weaker than general barriers. They are, in fact, just about enough to provide the ordering needed for atomic operations with release semantics. This patch makes use of that quirk. llvm-svn: 185527
2013-07-03 17:20:36 +08:00
; CHECK-THUMB: dmb {{ish$}}
Added atomic 64 min/max/umin/umax instrinsics support in the ARM backend. llvm-svn: 168886
2012-11-29 22:41:25 +08:00
%r = atomicrmw umax i64* %ptr, i64 %val seq_cst
ret i64 %r
}