maxjhandsome
/
llvm-project
forked from OSchip/llvm-project
1
0
Fork 0
Code Issues Pull Requests Packages Releases Wiki Activity
llvm-project/llvm/test/Transforms/TailCallElim/basic.ll

201 lines
5.1 KiB
LLVM
Raw Normal View History

[TRE] Merged several tests into the the test basic.ll. llvm-svn: 185723
2013-07-06 04:45:13 +08:00
; RUN: opt < %s -tailcallelim -S | FileCheck %s
declare void @noarg()
declare void @use(i32*)
Teach TailRecursionElimination to handle certain cases of nocapture escaping allocas. Without the changes introduced into this patch, if TRE saw any allocas at all, TRE would not perform TRE *or* mark callsites with the tail marker. Because TRE runs after mem2reg, this inadequacy is not a death sentence. But given a callsite A without escaping alloca argument, A may not be able to have the tail marker placed on it due to a separate callsite B having a write-back parameter passed in via an argument with the nocapture attribute. Assume that B is the only other callsite besides A and B only has nocapture escaping alloca arguments (*NOTE* B may have other arguments that are not passed allocas). In this case not marking A with the tail marker is unnecessarily conservative since: 1. By assumption A has no escaping alloca arguments itself so it can not access the caller's stack via its arguments. 2. Since all of B's escaping alloca arguments are passed as parameters with the nocapture attribute, we know that B does not stash said escaping allocas in a manner that outlives B itself and thus could be accessed indirectly by A. With the changes introduced by this patch: 1. If we see any escaping allocas passed as a capturing argument, we do nothing and bail early. 2. If we do not see any escaping allocas passed as captured arguments but we do see escaping allocas passed as nocapture arguments: i. We do not perform TRE to avoid PR962 since the code generator produces significantly worse code for the dynamic allocas that would be created by the TRE algorithm. ii. If we do not return twice, mark call sites without escaping allocas with the tail marker. *NOTE* This excludes functions with escaping nocapture allocas. 3. If we do not see any escaping allocas at all (whether captured or not): i. If we do not have usage of setjmp, mark all callsites with the tail marker. ii. If there are no dynamic/variable sized allocas in the function, attempt to perform TRE on all callsites in the function. Based off of a patch by Nick Lewycky. rdar://14324281. llvm-svn: 186057
2013-07-11 12:40:01 +08:00
declare void @use_nocapture(i32* nocapture)
declare void @use2_nocapture(i32* nocapture, i32* nocapture)
[TRE] Merged several tests into the the test basic.ll. llvm-svn: 185723
2013-07-06 04:45:13 +08:00
; Trivial case. Mark @noarg with tail call.
define void @test0() {
; CHECK: tail call void @noarg()
call void @noarg()
ret void
}
; PR615. Make sure that we do not move the alloca so that it interferes with the tail call.
define i32 @test1() {
; CHECK: i32 @test1()
; CHECK-NEXT: alloca
%A = alloca i32 ; <i32*> [#uses=2]
store i32 5, i32* %A
call void @use(i32* %A)
[TRE] Combined another test into basic.ll llvm-svn: 185729
2013-07-06 06:24:06 +08:00
; CHECK: tail call i32 @test1
[TRE] Merged several tests into the the test basic.ll. llvm-svn: 185723
2013-07-06 04:45:13 +08:00
%X = tail call i32 @test1() ; <i32> [#uses=1]
ret i32 %X
}
; This function contains intervening instructions which should be moved out of the way
define i32 @test2(i32 %X) {
; CHECK: i32 @test2
; CHECK-NOT: call
; CHECK: ret i32
entry:
%tmp.1 = icmp eq i32 %X, 0 ; <i1> [#uses=1]
br i1 %tmp.1, label %then.0, label %endif.0
then.0: ; preds = %entry
%tmp.4 = add i32 %X, 1 ; <i32> [#uses=1]
ret i32 %tmp.4
endif.0: ; preds = %entry
%tmp.10 = add i32 %X, -1 ; <i32> [#uses=1]
%tmp.8 = call i32 @test2(i32 %tmp.10) ; <i32> [#uses=1]
%DUMMY = add i32 %X, 1 ; <i32> [#uses=0]
ret i32 %tmp.8
}
; Though this case seems to be fairly unlikely to occur in the wild, someone
; plunked it into the demo script, so maybe they care about it.
define i32 @test3(i32 %c) {
; CHECK: i32 @test3
; CHECK-NOT: call
; CHECK: ret i32 0
entry:
%tmp.1 = icmp eq i32 %c, 0 ; <i1> [#uses=1]
br i1 %tmp.1, label %return, label %else
else: ; preds = %entry
%tmp.5 = add i32 %c, -1 ; <i32> [#uses=1]
%tmp.3 = call i32 @test3(i32 %tmp.5) ; <i32> [#uses=0]
ret i32 0
return: ; preds = %entry
ret i32 0
}
Teach TailRecursionElimination to handle certain cases of nocapture escaping allocas. Without the changes introduced into this patch, if TRE saw any allocas at all, TRE would not perform TRE *or* mark callsites with the tail marker. Because TRE runs after mem2reg, this inadequacy is not a death sentence. But given a callsite A without escaping alloca argument, A may not be able to have the tail marker placed on it due to a separate callsite B having a write-back parameter passed in via an argument with the nocapture attribute. Assume that B is the only other callsite besides A and B only has nocapture escaping alloca arguments (*NOTE* B may have other arguments that are not passed allocas). In this case not marking A with the tail marker is unnecessarily conservative since: 1. By assumption A has no escaping alloca arguments itself so it can not access the caller's stack via its arguments. 2. Since all of B's escaping alloca arguments are passed as parameters with the nocapture attribute, we know that B does not stash said escaping allocas in a manner that outlives B itself and thus could be accessed indirectly by A. With the changes introduced by this patch: 1. If we see any escaping allocas passed as a capturing argument, we do nothing and bail early. 2. If we do not see any escaping allocas passed as captured arguments but we do see escaping allocas passed as nocapture arguments: i. We do not perform TRE to avoid PR962 since the code generator produces significantly worse code for the dynamic allocas that would be created by the TRE algorithm. ii. If we do not return twice, mark call sites without escaping allocas with the tail marker. *NOTE* This excludes functions with escaping nocapture allocas. 3. If we do not see any escaping allocas at all (whether captured or not): i. If we do not have usage of setjmp, mark all callsites with the tail marker. ii. If there are no dynamic/variable sized allocas in the function, attempt to perform TRE on all callsites in the function. Based off of a patch by Nick Lewycky. rdar://14324281. llvm-svn: 186057
2013-07-11 12:40:01 +08:00
; Make sure that a nocapture pointer does not stop adding a tail call marker to
; an unrelated call and additionally that we do not mark the nocapture call with
; a tail call.
;
; rdar://14324281
define void @test4() {
; CHECK: void @test4
; CHECK-NOT: tail call void @use_nocapture
; CHECK: tail call void @noarg()
; CHECK: ret void
%a = alloca i32
call void @use_nocapture(i32* %a)
call void @noarg()
ret void
}
; Make sure that we do not perform TRE even with a nocapture use. This is due to
; bad codegen caused by PR962.
;
; rdar://14324281.
define i32* @test5(i32* nocapture %A, i1 %cond) {
; CHECK: i32* @test5
; CHECK-NOT: tailrecurse:
; CHECK: ret i32* null
%B = alloca i32
br i1 %cond, label %cond_true, label %cond_false
cond_true:
call i32* @test5(i32* %B, i1 false)
ret i32* null
cond_false:
call void @use2_nocapture(i32* %A, i32* %B)
call void @noarg()
ret i32* null
}
; PR14143: Make sure that we do not mark functions with nocapture allocas with tail.
;
; rdar://14324281.
define void @test6(i32* %a, i32* %b) {
Update Transforms tests to use CHECK-LABEL for easier debugging. No functionality change. This update was done with the following bash script: find test/Transforms -name "*.ll" | \ while read NAME; do echo "$NAME" if ! grep -q "^; *RUN: *llc" $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 mv $TEMP $NAME fi done llvm-svn: 186268
2013-07-14 09:42:54 +08:00
; CHECK-LABEL: @test6(
Teach TailRecursionElimination to handle certain cases of nocapture escaping allocas. Without the changes introduced into this patch, if TRE saw any allocas at all, TRE would not perform TRE *or* mark callsites with the tail marker. Because TRE runs after mem2reg, this inadequacy is not a death sentence. But given a callsite A without escaping alloca argument, A may not be able to have the tail marker placed on it due to a separate callsite B having a write-back parameter passed in via an argument with the nocapture attribute. Assume that B is the only other callsite besides A and B only has nocapture escaping alloca arguments (*NOTE* B may have other arguments that are not passed allocas). In this case not marking A with the tail marker is unnecessarily conservative since: 1. By assumption A has no escaping alloca arguments itself so it can not access the caller's stack via its arguments. 2. Since all of B's escaping alloca arguments are passed as parameters with the nocapture attribute, we know that B does not stash said escaping allocas in a manner that outlives B itself and thus could be accessed indirectly by A. With the changes introduced by this patch: 1. If we see any escaping allocas passed as a capturing argument, we do nothing and bail early. 2. If we do not see any escaping allocas passed as captured arguments but we do see escaping allocas passed as nocapture arguments: i. We do not perform TRE to avoid PR962 since the code generator produces significantly worse code for the dynamic allocas that would be created by the TRE algorithm. ii. If we do not return twice, mark call sites without escaping allocas with the tail marker. *NOTE* This excludes functions with escaping nocapture allocas. 3. If we do not see any escaping allocas at all (whether captured or not): i. If we do not have usage of setjmp, mark all callsites with the tail marker. ii. If there are no dynamic/variable sized allocas in the function, attempt to perform TRE on all callsites in the function. Based off of a patch by Nick Lewycky. rdar://14324281. llvm-svn: 186057
2013-07-11 12:40:01 +08:00
; CHECK-NOT: tail call
; CHECK: ret void
%c = alloca [100 x i8], align 16
%tmp = bitcast [100 x i8]* %c to i32*
call void @use2_nocapture(i32* %b, i32* %tmp)
ret void
}
; PR14143: Make sure that we do not mark functions with nocapture allocas with tail.
;
; rdar://14324281
define void @test7(i32* %a, i32* %b) nounwind uwtable {
entry:
Update Transforms tests to use CHECK-LABEL for easier debugging. No functionality change. This update was done with the following bash script: find test/Transforms -name "*.ll" | \ while read NAME; do echo "$NAME" if ! grep -q "^; *RUN: *llc" $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 mv $TEMP $NAME fi done llvm-svn: 186268
2013-07-14 09:42:54 +08:00
; CHECK-LABEL: @test7(
Teach TailRecursionElimination to handle certain cases of nocapture escaping allocas. Without the changes introduced into this patch, if TRE saw any allocas at all, TRE would not perform TRE *or* mark callsites with the tail marker. Because TRE runs after mem2reg, this inadequacy is not a death sentence. But given a callsite A without escaping alloca argument, A may not be able to have the tail marker placed on it due to a separate callsite B having a write-back parameter passed in via an argument with the nocapture attribute. Assume that B is the only other callsite besides A and B only has nocapture escaping alloca arguments (*NOTE* B may have other arguments that are not passed allocas). In this case not marking A with the tail marker is unnecessarily conservative since: 1. By assumption A has no escaping alloca arguments itself so it can not access the caller's stack via its arguments. 2. Since all of B's escaping alloca arguments are passed as parameters with the nocapture attribute, we know that B does not stash said escaping allocas in a manner that outlives B itself and thus could be accessed indirectly by A. With the changes introduced by this patch: 1. If we see any escaping allocas passed as a capturing argument, we do nothing and bail early. 2. If we do not see any escaping allocas passed as captured arguments but we do see escaping allocas passed as nocapture arguments: i. We do not perform TRE to avoid PR962 since the code generator produces significantly worse code for the dynamic allocas that would be created by the TRE algorithm. ii. If we do not return twice, mark call sites without escaping allocas with the tail marker. *NOTE* This excludes functions with escaping nocapture allocas. 3. If we do not see any escaping allocas at all (whether captured or not): i. If we do not have usage of setjmp, mark all callsites with the tail marker. ii. If there are no dynamic/variable sized allocas in the function, attempt to perform TRE on all callsites in the function. Based off of a patch by Nick Lewycky. rdar://14324281. llvm-svn: 186057
2013-07-11 12:40:01 +08:00
; CHECK-NOT: tail call
; CHECK: ret void
%c = alloca [100 x i8], align 16
%0 = bitcast [100 x i8]* %c to i32*
call void @use2_nocapture(i32* %0, i32* %a)
call void @use2_nocapture(i32* %b, i32* %0)
ret void
}
; If we have a mix of escaping captured/non-captured allocas, ensure that we do
; not do anything including marking callsites with the tail call marker.
;
; rdar://14324281.
define i32* @test8(i32* nocapture %A, i1 %cond) {
; CHECK: i32* @test8
; CHECK-NOT: tailrecurse:
; CHECK-NOT: tail call
; CHECK: ret i32* null
%B = alloca i32
%B2 = alloca i32
br i1 %cond, label %cond_true, label %cond_false
cond_true:
call void @use(i32* %B2)
call i32* @test8(i32* %B, i1 false)
ret i32* null
cond_false:
call void @use2_nocapture(i32* %A, i32* %B)
call void @noarg()
ret i32* null
}
Fix PR7272 in -tailcallelim instead of the inliner The -tailcallelim pass should be checking if byval or inalloca args can be captured before marking calls as tail calls. This was the real root cause of PR7272. With a better fix in place, revert the inliner change from r105255. The test case it introduced still passes and has been moved to test/Transforms/Inline/byval-tail-call.ll. Reviewers: chandlerc Differential Revision: http://reviews.llvm.org/D3403 llvm-svn: 206789
2014-04-22 04:48:47 +08:00
; Don't tail call if a byval arg is captured.
define void @test9(i32* byval %a) {
; CHECK-LABEL: define void @test9(
Revert "Transforms: reapply SVN r219899" This reverts commit r220811 and r220839. It made an incorrect change to musttail handling. llvm-svn: 221226
2014-11-04 10:02:14 +08:00
; CHECK: {{^ *}}call void @use(
Fix PR7272 in -tailcallelim instead of the inliner The -tailcallelim pass should be checking if byval or inalloca args can be captured before marking calls as tail calls. This was the real root cause of PR7272. With a better fix in place, revert the inliner change from r105255. The test case it introduced still passes and has been moved to test/Transforms/Inline/byval-tail-call.ll. Reviewers: chandlerc Differential Revision: http://reviews.llvm.org/D3403 llvm-svn: 206789
2014-04-22 04:48:47 +08:00
call void @use(i32* %a)
ret void
}
Improve 'tail' call marking in TRE. A bootstrap of clang goes from 375k calls marked tail in the IR to 470k, however this improvement does not carry into an improvement of the call/jmp ratio on x86. The most common pattern is a tail call + br to a block with nothing but a 'ret'. The number of tail call to loop conversions remains the same (1618 by my count). The new algorithm does a local scan over the use-def chains to identify local "alloca-derived" values, as well as points where the alloca could escape. Then, a visit over the CFG marks blocks as being before or after the allocas have escaped, and annotates the calls accordingly. llvm-svn: 208017
2014-05-06 07:59:03 +08:00
%struct.X = type { i8* }
declare void @ctor(%struct.X*)
define void @test10(%struct.X* noalias sret %agg.result, i1 zeroext %b) {
Fix a bunch of trivial cases of 'CHECK[^:]*$' in the tests. NFCI I looked into adding a warning / error for this to FileCheck, but there doesn't seem to be a good way to avoid it triggering on the instances of it in RUN lines. llvm-svn: 244481
2015-08-11 03:01:27 +08:00
; CHECK-LABEL: @test10
Improve 'tail' call marking in TRE. A bootstrap of clang goes from 375k calls marked tail in the IR to 470k, however this improvement does not carry into an improvement of the call/jmp ratio on x86. The most common pattern is a tail call + br to a block with nothing but a 'ret'. The number of tail call to loop conversions remains the same (1618 by my count). The new algorithm does a local scan over the use-def chains to identify local "alloca-derived" values, as well as points where the alloca could escape. Then, a visit over the CFG marks blocks as being before or after the allocas have escaped, and annotates the calls accordingly. llvm-svn: 208017
2014-05-06 07:59:03 +08:00
entry:
%x = alloca %struct.X, align 8
br i1 %b, label %if.then, label %if.end
if.then: ; preds = %entry
call void @ctor(%struct.X* %agg.result)
; CHECK: tail call void @ctor
br label %return
if.end:
call void @ctor(%struct.X* %x)
; CHECK: call void @ctor
br label %return
return:
ret void
}
We may visit a call that uses an alloca multiple times in callUsesLocalStack, sometimes with IsNocapture true and sometimes with IsNocapture false. We accidentally skipped work we needed to do in the IsNocapture=false case if we were called with IsNocapture=true the first time. Fixes PR20405! llvm-svn: 213726
2014-07-23 14:24:49 +08:00
declare void @test11_helper1(i8** nocapture, i8*)
declare void @test11_helper2(i8*)
define void @test11() {
; CHECK-LABEL: @test11
; CHECK-NOT: tail
%a = alloca i8*
%b = alloca i8
call void @test11_helper1(i8** %a, i8* %b) ; a = &b
[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
%c = load i8*, i8** %a
We may visit a call that uses an alloca multiple times in callUsesLocalStack, sometimes with IsNocapture true and sometimes with IsNocapture false. We accidentally skipped work we needed to do in the IsNocapture=false case if we were called with IsNocapture=true the first time. Fixes PR20405! llvm-svn: 213726
2014-07-23 14:24:49 +08:00
call void @test11_helper2(i8* %c)
; CHECK: call void @test11_helper2
ret void
}
[OperandBundles] Have TailCallElim play nice with operand bundles A call site's use of a Value might not correspond to an argument operand but to a bundle operand. This fixes PR25928. llvm-svn: 256328
2015-12-23 17:58:43 +08:00
; PR25928
define void @test12() {
entry:
; CHECK-LABEL: @test12
; CHECK: {{^ *}} call void undef(i8* undef) [ "foo"(i8* %e) ]
%e = alloca i8
call void undef(i8* undef) [ "foo"(i8* %e) ]
unreachable
}