2014-05-30 18:09:59 +08:00
|
|
|
; RUN: llc -O3 -mtriple=thumb-eabi -mcpu=cortex-a8 %s -o - -arm-atomic-cfg-tidy=0 | FileCheck %s
|
2012-06-16 06:46:31 +08:00
|
|
|
;
|
|
|
|
|
; LSR should only check for valid address modes when the IV user is a
|
|
|
|
|
; memory address.
|
|
|
|
|
; svn r158536, rdar://11635990
|
|
|
|
|
;
|
|
|
|
|
; Note that we still don't produce the best code here because we fail
|
|
|
|
|
; to coalesce the IV. See <rdar://problem/11680670> [coalescer] IVs
|
|
|
|
|
; need to be scheduled to expose coalescing.
|
|
|
|
|
|
|
|
|
|
; LSR before the fix:
|
|
|
|
|
;The chosen solution requires 4 regs, with addrec cost 1, plus 3 base adds, plus 2 setup cost:
|
|
|
|
|
; LSR Use: Kind=Special, Offsets={0}, all-fixups-outside-loop, widest fixup type: i32
|
|
|
|
|
; reg(%v3) + reg({0,+,-1}<%while.cond.i.i>) + imm(1)
|
|
|
|
|
; LSR Use: Kind=ICmpZero, Offsets={0}, widest fixup type: i32
|
|
|
|
|
; reg(%v3) + reg({0,+,-1}<%while.cond.i.i>)
|
|
|
|
|
; LSR Use: Kind=Address of i32, Offsets={0}, widest fixup type: i32*
|
|
|
|
|
; reg((-4 + (4 * %v3) + %v1)) + 4*reg({0,+,-1}<%while.cond.i.i>)
|
|
|
|
|
; LSR Use: Kind=Address of i32, Offsets={0}, widest fixup type: i32*
|
|
|
|
|
; reg((-4 + (4 * %v3) + %v4)) + 4*reg({0,+,-1}<%while.cond.i.i>)
|
|
|
|
|
; LSR Use: Kind=Special, Offsets={0}, all-fixups-outside-loop, widest fixup type: i32
|
|
|
|
|
; reg(%v3)
|
|
|
|
|
;
|
|
|
|
|
; LSR after the fix:
|
|
|
|
|
;The chosen solution requires 4 regs, with addrec cost 1, plus 1 base add, plus 2 setup cost:
|
|
|
|
|
; LSR Use: Kind=Special, Offsets={0}, all-fixups-outside-loop, widest fixup type: i32
|
|
|
|
|
; reg({%v3,+,-1}<nsw><%while.cond.i.i>) + imm(1)
|
|
|
|
|
; LSR Use: Kind=ICmpZero, Offsets={0}, widest fixup type: i32
|
|
|
|
|
; reg({%v3,+,-1}<nsw><%while.cond.i.i>)
|
|
|
|
|
; LSR Use: Kind=Address of i32, Offsets={0}, widest fixup type: i32*
|
|
|
|
|
; reg((-4 + %v1)) + 4*reg({%v3,+,-1}<nsw><%while.cond.i.i>)
|
|
|
|
|
; LSR Use: Kind=Address of i32, Offsets={0}, widest fixup type: i32*
|
|
|
|
|
; reg((-4 + %v4)) + 4*reg({%v3,+,-1}<nsw><%while.cond.i.i>)
|
|
|
|
|
; LSR Use: Kind=Special, Offsets={0}, all-fixups-outside-loop, widest fixup type: i32
|
|
|
|
|
; reg(%v3)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
%s = type { i32* }
|
|
|
|
|
|
|
|
|
|
@ncol = external global i32, align 4
|
|
|
|
|
|
|
|
|
|
declare i32* @getptr() nounwind
|
|
|
|
|
declare %s* @getstruct() nounwind
|
|
|
|
|
|
|
|
|
|
; CHECK: @main
|
|
|
|
|
; Check that the loop preheader contains no address computation.
|
Generalize MergeBlockIntoPredecessor. Replace uses of MergeBasicBlockIntoOnlyPred.
Summary:
Two utils methods have essentially the same functionality. This is an attempt to merge them into one.
1. lib/Transforms/Utils/Local.cpp : MergeBasicBlockIntoOnlyPred
2. lib/Transforms/Utils/BasicBlockUtils.cpp : MergeBlockIntoPredecessor
Prior to the patch:
1. MergeBasicBlockIntoOnlyPred
Updates either DomTree or DeferredDominance
Moves all instructions from Pred to BB, deletes Pred
Asserts BB has single predecessor
If address was taken, replace the block address with constant 1 (?)
2. MergeBlockIntoPredecessor
Updates DomTree, LoopInfo and MemoryDependenceResults
Moves all instruction from BB to Pred, deletes BB
Returns if doesn't have a single predecessor
Returns if BB's address was taken
After the patch:
Method 2. MergeBlockIntoPredecessor is attempting to become the new default:
Updates DomTree or DeferredDominance, and LoopInfo and MemoryDependenceResults
Moves all instruction from BB to Pred, deletes BB
Returns if doesn't have a single predecessor
Returns if BB's address was taken
Uses of MergeBasicBlockIntoOnlyPred that need to be replaced:
1. lib/Transforms/Scalar/LoopSimplifyCFG.cpp
Updated in this patch. No challenges.
2. lib/CodeGen/CodeGenPrepare.cpp
Updated in this patch.
i. eliminateFallThrough is straightforward, but I added using a temporary array to avoid the iterator invalidation.
ii. eliminateMostlyEmptyBlock(s) methods also now use a temporary array for blocks
Some interesting aspects:
- Since Pred is not deleted (BB is), the entry block does not need updating.
- The entry block was being updated with the deleted block in eliminateMostlyEmptyBlock. Added assert to make obvious that BB=SinglePred.
- isMergingEmptyBlockProfitable assumes BB is the one to be deleted.
- eliminateMostlyEmptyBlock(BB) does not delete BB on one path, it deletes its unique predecessor instead.
- adding some test owner as subscribers for the interesting tests modified:
test/CodeGen/X86/avx-cmp.ll
test/CodeGen/AMDGPU/nested-loop-conditions.ll
test/CodeGen/AMDGPU/si-annotate-cf.ll
test/CodeGen/X86/hoist-spill.ll
test/CodeGen/X86/2006-11-17-IllegalMove.ll
3. lib/Transforms/Scalar/JumpThreading.cpp
Not covered in this patch. It is the only use case using the DeferredDominance.
I would defer to Brian Rzycki to make this replacement.
Reviewers: chandlerc, spatel, davide, brzycki, bkramer, javed.absar
Subscribers: qcolombet, sanjoy, nemanjai, nhaehnle, jlebar, tpr, kbarton, RKSimon, wmi, arsenm, llvm-commits
Differential Revision: https://reviews.llvm.org/D48202
llvm-svn: 335183
2018-06-21 06:01:04 +08:00
|
|
|
; CHECK: %while.cond.i.i
|
2012-06-16 06:46:31 +08:00
|
|
|
; CHECK-NOT: add{{.*}}lsl
|
|
|
|
|
; CHECK: ldr{{.*}}lsl #2
|
|
|
|
|
; CHECK: ldr{{.*}}lsl #2
|
|
|
|
|
define i32 @main() nounwind ssp {
|
|
|
|
|
entry:
|
2015-02-28 05:17:42 +08:00
|
|
|
%v0 = load i32, i32* @ncol, align 4
|
2012-06-16 06:46:31 +08:00
|
|
|
%v1 = tail call i32* @getptr() nounwind
|
|
|
|
|
%cmp10.i = icmp eq i32 %v0, 0
|
|
|
|
|
br label %while.cond.outer
|
|
|
|
|
|
|
|
|
|
while.cond.outer:
|
|
|
|
|
%call18 = tail call %s* @getstruct() nounwind
|
|
|
|
|
br label %while.cond
|
|
|
|
|
|
|
|
|
|
while.cond:
|
|
|
|
|
%cmp20 = icmp eq i32* %v1, null
|
|
|
|
|
br label %while.body
|
|
|
|
|
|
|
|
|
|
while.body:
|
2015-02-28 05:17:42 +08:00
|
|
|
%v3 = load i32, i32* @ncol, align 4
|
2012-08-14 13:19:07 +08:00
|
|
|
br label %end_of_chain
|
2012-06-16 06:46:31 +08:00
|
|
|
|
2012-08-14 13:19:07 +08:00
|
|
|
end_of_chain:
|
[opaque pointer type] Add textual IR support for explicit type parameter to getelementptr instruction
One of several parallel first steps to remove the target type of pointers,
replacing them with a single opaque pointer type.
This adds an explicit type parameter to the gep instruction so that when the
first parameter becomes an opaque pointer type, the type to gep through is
still available to the instructions.
* This doesn't modify gep operators, only instructions (operators will be
handled separately)
* Textual IR changes only. Bitcode (including upgrade) and changing the
in-memory representation will be in separate changes.
* geps of vectors are transformed as:
getelementptr <4 x float*> %x, ...
->getelementptr float, <4 x float*> %x, ...
Then, once the opaque pointer type is introduced, this will ultimately look
like:
getelementptr float, <4 x ptr> %x
with the unambiguous interpretation that it is a vector of pointers to float.
* address spaces remain on the pointer, not the type:
getelementptr float addrspace(1)* %x
->getelementptr float, float addrspace(1)* %x
Then, eventually:
getelementptr float, ptr addrspace(1) %x
Importantly, the massive amount of test case churn has been automated by
same crappy python code. I had to manually update a few test cases that
wouldn't fit the script's model (r228970,r229196,r229197,r229198). The
python script just massages stdin and writes the result to stdout, I
then wrapped that in a shell script to handle replacing files, then
using the usual find+xargs to migrate all the files.
update.py:
import fileinput
import sys
import re
ibrep = re.compile(r"(^.*?[^%\w]getelementptr inbounds )(((?:<\d* x )?)(.*?)(| addrspace\(\d\)) *\*(|>)(?:$| *(?:%|@|null|undef|blockaddress|getelementptr|addrspacecast|bitcast|inttoptr|\[\[[a-zA-Z]|\{\{).*$))")
normrep = re.compile( r"(^.*?[^%\w]getelementptr )(((?:<\d* x )?)(.*?)(| addrspace\(\d\)) *\*(|>)(?:$| *(?:%|@|null|undef|blockaddress|getelementptr|addrspacecast|bitcast|inttoptr|\[\[[a-zA-Z]|\{\{).*$))")
def conv(match, line):
if not match:
return line
line = match.groups()[0]
if len(match.groups()[5]) == 0:
line += match.groups()[2]
line += match.groups()[3]
line += ", "
line += match.groups()[1]
line += "\n"
return line
for line in sys.stdin:
if line.find("getelementptr ") == line.find("getelementptr inbounds"):
if line.find("getelementptr inbounds") != line.find("getelementptr inbounds ("):
line = conv(re.match(ibrep, line), line)
elif line.find("getelementptr ") != line.find("getelementptr ("):
line = conv(re.match(normrep, line), line)
sys.stdout.write(line)
apply.sh:
for name in "$@"
do
python3 `dirname "$0"`/update.py < "$name" > "$name.tmp" && mv "$name.tmp" "$name"
rm -f "$name.tmp"
done
The actual commands:
From llvm/src:
find test/ -name *.ll | xargs ./apply.sh
From llvm/src/tools/clang:
find test/ -name *.mm -o -name *.m -o -name *.cpp -o -name *.c | xargs -I '{}' ../../apply.sh "{}"
From llvm/src/tools/polly:
find test/ -name *.ll | xargs ./apply.sh
After that, check-all (with llvm, clang, clang-tools-extra, lld,
compiler-rt, and polly all checked out).
The extra 'rm' in the apply.sh script is due to a few files in clang's test
suite using interesting unicode stuff that my python script was throwing
exceptions on. None of those files needed to be migrated, so it seemed
sufficient to ignore those cases.
Reviewers: rafael, dexonsmith, grosser
Differential Revision: http://reviews.llvm.org/D7636
llvm-svn: 230786
2015-02-28 03:29:02 +08:00
|
|
|
%state.i = getelementptr inbounds %s, %s* %call18, i32 0, i32 0
|
2015-02-28 05:17:42 +08:00
|
|
|
%v4 = load i32*, i32** %state.i, align 4
|
2012-06-16 06:46:31 +08:00
|
|
|
br label %while.cond.i.i
|
|
|
|
|
|
|
|
|
|
while.cond.i.i:
|
2012-08-14 13:19:07 +08:00
|
|
|
%counter.0.i.i = phi i32 [ %v3, %end_of_chain ], [ %dec.i.i, %land.rhs.i.i ]
|
2012-06-16 06:46:31 +08:00
|
|
|
%dec.i.i = add nsw i32 %counter.0.i.i, -1
|
|
|
|
|
%tobool.i.i = icmp eq i32 %counter.0.i.i, 0
|
|
|
|
|
br i1 %tobool.i.i, label %where.exit, label %land.rhs.i.i
|
|
|
|
|
|
|
|
|
|
land.rhs.i.i:
|
[opaque pointer type] Add textual IR support for explicit type parameter to getelementptr instruction
One of several parallel first steps to remove the target type of pointers,
replacing them with a single opaque pointer type.
This adds an explicit type parameter to the gep instruction so that when the
first parameter becomes an opaque pointer type, the type to gep through is
still available to the instructions.
* This doesn't modify gep operators, only instructions (operators will be
handled separately)
* Textual IR changes only. Bitcode (including upgrade) and changing the
in-memory representation will be in separate changes.
* geps of vectors are transformed as:
getelementptr <4 x float*> %x, ...
->getelementptr float, <4 x float*> %x, ...
Then, once the opaque pointer type is introduced, this will ultimately look
like:
getelementptr float, <4 x ptr> %x
with the unambiguous interpretation that it is a vector of pointers to float.
* address spaces remain on the pointer, not the type:
getelementptr float addrspace(1)* %x
->getelementptr float, float addrspace(1)* %x
Then, eventually:
getelementptr float, ptr addrspace(1) %x
Importantly, the massive amount of test case churn has been automated by
same crappy python code. I had to manually update a few test cases that
wouldn't fit the script's model (r228970,r229196,r229197,r229198). The
python script just massages stdin and writes the result to stdout, I
then wrapped that in a shell script to handle replacing files, then
using the usual find+xargs to migrate all the files.
update.py:
import fileinput
import sys
import re
ibrep = re.compile(r"(^.*?[^%\w]getelementptr inbounds )(((?:<\d* x )?)(.*?)(| addrspace\(\d\)) *\*(|>)(?:$| *(?:%|@|null|undef|blockaddress|getelementptr|addrspacecast|bitcast|inttoptr|\[\[[a-zA-Z]|\{\{).*$))")
normrep = re.compile( r"(^.*?[^%\w]getelementptr )(((?:<\d* x )?)(.*?)(| addrspace\(\d\)) *\*(|>)(?:$| *(?:%|@|null|undef|blockaddress|getelementptr|addrspacecast|bitcast|inttoptr|\[\[[a-zA-Z]|\{\{).*$))")
def conv(match, line):
if not match:
return line
line = match.groups()[0]
if len(match.groups()[5]) == 0:
line += match.groups()[2]
line += match.groups()[3]
line += ", "
line += match.groups()[1]
line += "\n"
return line
for line in sys.stdin:
if line.find("getelementptr ") == line.find("getelementptr inbounds"):
if line.find("getelementptr inbounds") != line.find("getelementptr inbounds ("):
line = conv(re.match(ibrep, line), line)
elif line.find("getelementptr ") != line.find("getelementptr ("):
line = conv(re.match(normrep, line), line)
sys.stdout.write(line)
apply.sh:
for name in "$@"
do
python3 `dirname "$0"`/update.py < "$name" > "$name.tmp" && mv "$name.tmp" "$name"
rm -f "$name.tmp"
done
The actual commands:
From llvm/src:
find test/ -name *.ll | xargs ./apply.sh
From llvm/src/tools/clang:
find test/ -name *.mm -o -name *.m -o -name *.cpp -o -name *.c | xargs -I '{}' ../../apply.sh "{}"
From llvm/src/tools/polly:
find test/ -name *.ll | xargs ./apply.sh
After that, check-all (with llvm, clang, clang-tools-extra, lld,
compiler-rt, and polly all checked out).
The extra 'rm' in the apply.sh script is due to a few files in clang's test
suite using interesting unicode stuff that my python script was throwing
exceptions on. None of those files needed to be migrated, so it seemed
sufficient to ignore those cases.
Reviewers: rafael, dexonsmith, grosser
Differential Revision: http://reviews.llvm.org/D7636
llvm-svn: 230786
2015-02-28 03:29:02 +08:00
|
|
|
%arrayidx.i.i = getelementptr inbounds i32, i32* %v4, i32 %dec.i.i
|
2015-02-28 05:17:42 +08:00
|
|
|
%v5 = load i32, i32* %arrayidx.i.i, align 4
|
[opaque pointer type] Add textual IR support for explicit type parameter to getelementptr instruction
One of several parallel first steps to remove the target type of pointers,
replacing them with a single opaque pointer type.
This adds an explicit type parameter to the gep instruction so that when the
first parameter becomes an opaque pointer type, the type to gep through is
still available to the instructions.
* This doesn't modify gep operators, only instructions (operators will be
handled separately)
* Textual IR changes only. Bitcode (including upgrade) and changing the
in-memory representation will be in separate changes.
* geps of vectors are transformed as:
getelementptr <4 x float*> %x, ...
->getelementptr float, <4 x float*> %x, ...
Then, once the opaque pointer type is introduced, this will ultimately look
like:
getelementptr float, <4 x ptr> %x
with the unambiguous interpretation that it is a vector of pointers to float.
* address spaces remain on the pointer, not the type:
getelementptr float addrspace(1)* %x
->getelementptr float, float addrspace(1)* %x
Then, eventually:
getelementptr float, ptr addrspace(1) %x
Importantly, the massive amount of test case churn has been automated by
same crappy python code. I had to manually update a few test cases that
wouldn't fit the script's model (r228970,r229196,r229197,r229198). The
python script just massages stdin and writes the result to stdout, I
then wrapped that in a shell script to handle replacing files, then
using the usual find+xargs to migrate all the files.
update.py:
import fileinput
import sys
import re
ibrep = re.compile(r"(^.*?[^%\w]getelementptr inbounds )(((?:<\d* x )?)(.*?)(| addrspace\(\d\)) *\*(|>)(?:$| *(?:%|@|null|undef|blockaddress|getelementptr|addrspacecast|bitcast|inttoptr|\[\[[a-zA-Z]|\{\{).*$))")
normrep = re.compile( r"(^.*?[^%\w]getelementptr )(((?:<\d* x )?)(.*?)(| addrspace\(\d\)) *\*(|>)(?:$| *(?:%|@|null|undef|blockaddress|getelementptr|addrspacecast|bitcast|inttoptr|\[\[[a-zA-Z]|\{\{).*$))")
def conv(match, line):
if not match:
return line
line = match.groups()[0]
if len(match.groups()[5]) == 0:
line += match.groups()[2]
line += match.groups()[3]
line += ", "
line += match.groups()[1]
line += "\n"
return line
for line in sys.stdin:
if line.find("getelementptr ") == line.find("getelementptr inbounds"):
if line.find("getelementptr inbounds") != line.find("getelementptr inbounds ("):
line = conv(re.match(ibrep, line), line)
elif line.find("getelementptr ") != line.find("getelementptr ("):
line = conv(re.match(normrep, line), line)
sys.stdout.write(line)
apply.sh:
for name in "$@"
do
python3 `dirname "$0"`/update.py < "$name" > "$name.tmp" && mv "$name.tmp" "$name"
rm -f "$name.tmp"
done
The actual commands:
From llvm/src:
find test/ -name *.ll | xargs ./apply.sh
From llvm/src/tools/clang:
find test/ -name *.mm -o -name *.m -o -name *.cpp -o -name *.c | xargs -I '{}' ../../apply.sh "{}"
From llvm/src/tools/polly:
find test/ -name *.ll | xargs ./apply.sh
After that, check-all (with llvm, clang, clang-tools-extra, lld,
compiler-rt, and polly all checked out).
The extra 'rm' in the apply.sh script is due to a few files in clang's test
suite using interesting unicode stuff that my python script was throwing
exceptions on. None of those files needed to be migrated, so it seemed
sufficient to ignore those cases.
Reviewers: rafael, dexonsmith, grosser
Differential Revision: http://reviews.llvm.org/D7636
llvm-svn: 230786
2015-02-28 03:29:02 +08:00
|
|
|
%arrayidx1.i.i = getelementptr inbounds i32, i32* %v1, i32 %dec.i.i
|
2015-02-28 05:17:42 +08:00
|
|
|
%v6 = load i32, i32* %arrayidx1.i.i, align 4
|
2012-06-16 06:46:31 +08:00
|
|
|
%cmp.i.i = icmp eq i32 %v5, %v6
|
|
|
|
|
br i1 %cmp.i.i, label %while.cond.i.i, label %equal_data.exit.i
|
|
|
|
|
|
|
|
|
|
equal_data.exit.i:
|
|
|
|
|
ret i32 %counter.0.i.i
|
|
|
|
|
|
|
|
|
|
where.exit:
|
|
|
|
|
br label %while.end.i
|
|
|
|
|
|
|
|
|
|
while.end.i:
|
|
|
|
|
ret i32 %v3
|
|
|
|
|
}
|