RPCS3/llvm-mirror
1
0
Fork 0
mirror of https://github.com/RPCS3/llvm-mirror.git synced 2024-10-29 23:12:55 +01:00
Code Issues Projects Releases Wiki Activity
64b8621170
llvm-mirror/test/CodeGen/X86/switch-bt.ll

102 lines
2.6 KiB
LLVM
Raw Normal View History

Optimize the "bit test" code path for switch lowering in the case where the bit mask has exactly one bit. llvm-svn: 106716
2010-06-24 04:06:24 +02:00
; RUN: llc -march=x86-64 -asm-verbose=false < %s | FileCheck %s
; This switch should use bit tests, and the third bit test case is just
; testing for one possible value, so it doesn't need a bt.
; CHECK: movabsq $2305843009482129440, %r
; CHECK-NEXT: btq %rax, %r
Flip the new block-placement pass to be on by default. This is mostly to test the waters. I'd like to get results from FNT build bots and other bots running on non-x86 platforms. This feature has been pretty heavily tested over the last few months by me, and it fixes several of the execution time regressions caused by the inlining work by preventing inlining decisions from radically impacting block layout. I've seen very large improvements in yacr2 and ackermann benchmarks, along with the expected noise across all of the benchmark suite whenever code layout changes. I've analyzed all of the regressions and fixed them, or found them to be impossible to fix. See my email to llvmdev for more details. I'd like for this to be in 3.1 as it complements the inliner changes, but if any failures are showing up or anyone has concerns, it is just a flag flip and so can be easily turned off. I'm switching it on tonight to try and get at least one run through various folks' performance suites in case SPEC or something else has serious issues with it. I'll watch bots and revert if anything shows up. llvm-svn: 154816
2012-04-16 15:49:17 +02:00
; CHECK-NEXT: jae
; CHECK: movl $671088640, %e
Optimize the "bit test" code path for switch lowering in the case where the bit mask has exactly one bit. llvm-svn: 106716
2010-06-24 04:06:24 +02:00
; CHECK-NEXT: btq %rax, %r
Flip the new block-placement pass to be on by default. This is mostly to test the waters. I'd like to get results from FNT build bots and other bots running on non-x86 platforms. This feature has been pretty heavily tested over the last few months by me, and it fixes several of the execution time regressions caused by the inlining work by preventing inlining decisions from radically impacting block layout. I've seen very large improvements in yacr2 and ackermann benchmarks, along with the expected noise across all of the benchmark suite whenever code layout changes. I've analyzed all of the regressions and fixed them, or found them to be impossible to fix. See my email to llvmdev for more details. I'd like for this to be in 3.1 as it complements the inliner changes, but if any failures are showing up or anyone has concerns, it is just a flag flip and so can be easily turned off. I'm switching it on tonight to try and get at least one run through various folks' performance suites in case SPEC or something else has serious issues with it. I'll watch bots and revert if anything shows up. llvm-svn: 154816
2012-04-16 15:49:17 +02:00
; CHECK-NEXT: jae
; CHECK: testq %rax, %r
Optimize the "bit test" code path for switch lowering in the case where the bit mask has exactly one bit. llvm-svn: 106716
2010-06-24 04:06:24 +02:00
; CHECK-NEXT: j
define void @test(i8* %l) nounwind {
entry:
%l.addr = alloca i8*, align 8 ; <i8**> [#uses=2]
store i8* %l, i8** %l.addr
%tmp = load i8** %l.addr ; <i8*> [#uses=1]
%tmp1 = load i8* %tmp ; <i8> [#uses=1]
%conv = sext i8 %tmp1 to i32 ; <i32> [#uses=1]
switch i32 %conv, label %sw.default [
i32 62, label %sw.bb
i32 60, label %sw.bb
i32 38, label %sw.bb2
i32 94, label %sw.bb2
i32 61, label %sw.bb2
i32 33, label %sw.bb4
]
sw.bb: ; preds = %entry, %entry
call void @foo(i32 0)
br label %sw.epilog
sw.bb2: ; preds = %entry, %entry, %entry
call void @foo(i32 1)
br label %sw.epilog
sw.bb4: ; preds = %entry
call void @foo(i32 3)
br label %sw.epilog
sw.default: ; preds = %entry
call void @foo(i32 97)
br label %sw.epilog
sw.epilog: ; preds = %sw.default, %sw.bb4, %sw.bb2, %sw.bb
ret void
}
declare void @foo(i32)
Avoid zero extend bit test operands to pointer type if all the masks fit in the original type of the switch statement key. rdar://8781238 llvm-svn: 122935
2011-01-06 02:02:44 +01:00
; Don't zero extend the test operands to pointer type if it can be avoided.
; rdar://8781238
define void @test2(i32 %x) nounwind ssp {
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
2013-07-13 22:38:47 +02:00
; CHECK-LABEL: test2:
Avoid zero extend bit test operands to pointer type if all the masks fit in the original type of the switch statement key. rdar://8781238 llvm-svn: 122935
2011-01-06 02:02:44 +01:00
; CHECK: cmpl $6
; CHECK: ja
; CHECK-NEXT: movl $91
; CHECK-NOT: movl
; CHECK-NEXT: btl
Flip the new block-placement pass to be on by default. This is mostly to test the waters. I'd like to get results from FNT build bots and other bots running on non-x86 platforms. This feature has been pretty heavily tested over the last few months by me, and it fixes several of the execution time regressions caused by the inlining work by preventing inlining decisions from radically impacting block layout. I've seen very large improvements in yacr2 and ackermann benchmarks, along with the expected noise across all of the benchmark suite whenever code layout changes. I've analyzed all of the regressions and fixed them, or found them to be impossible to fix. See my email to llvmdev for more details. I'd like for this to be in 3.1 as it complements the inliner changes, but if any failures are showing up or anyone has concerns, it is just a flag flip and so can be easily turned off. I'm switching it on tonight to try and get at least one run through various folks' performance suites in case SPEC or something else has serious issues with it. I'll watch bots and revert if anything shows up. llvm-svn: 154816
2012-04-16 15:49:17 +02:00
; CHECK-NEXT: jae
Avoid zero extend bit test operands to pointer type if all the masks fit in the original type of the switch statement key. rdar://8781238 llvm-svn: 122935
2011-01-06 02:02:44 +01:00
entry:
switch i32 %x, label %if.end [
i32 6, label %if.then
i32 4, label %if.then
i32 3, label %if.then
i32 1, label %if.then
i32 0, label %if.then
]
if.then: ; preds = %entry, %entry, %entry, %entry, %entry
tail call void @bar() nounwind
ret void
if.end: ; preds = %entry
ret void
}
declare void @bar()
Don't emit a bit test if there is only one case the test can yield false. A simple SETNE is sufficient. llvm-svn: 135126
2011-07-14 03:38:42 +02:00
define void @test3(i32 %x) nounwind {
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
2013-07-13 22:38:47 +02:00
; CHECK-LABEL: test3:
Don't emit a bit test if there is only one case the test can yield false. A simple SETNE is sufficient. llvm-svn: 135126
2011-07-14 03:38:42 +02:00
; CHECK: cmpl $5
; CHECK: ja
; CHECK: cmpl $4
Flip the new block-placement pass to be on by default. This is mostly to test the waters. I'd like to get results from FNT build bots and other bots running on non-x86 platforms. This feature has been pretty heavily tested over the last few months by me, and it fixes several of the execution time regressions caused by the inlining work by preventing inlining decisions from radically impacting block layout. I've seen very large improvements in yacr2 and ackermann benchmarks, along with the expected noise across all of the benchmark suite whenever code layout changes. I've analyzed all of the regressions and fixed them, or found them to be impossible to fix. See my email to llvmdev for more details. I'd like for this to be in 3.1 as it complements the inliner changes, but if any failures are showing up or anyone has concerns, it is just a flag flip and so can be easily turned off. I'm switching it on tonight to try and get at least one run through various folks' performance suites in case SPEC or something else has serious issues with it. I'll watch bots and revert if anything shows up. llvm-svn: 154816
2012-04-16 15:49:17 +02:00
; CHECK: je
Don't emit a bit test if there is only one case the test can yield false. A simple SETNE is sufficient. llvm-svn: 135126
2011-07-14 03:38:42 +02:00
switch i32 %x, label %if.end [
i32 0, label %if.then
i32 1, label %if.then
i32 2, label %if.then
i32 3, label %if.then
i32 5, label %if.then
]
if.then:
tail call void @bar() nounwind
ret void
if.end:
ret void
}
Reference in New Issue Copy Permalink