mirror of
https://github.com/c64scene-ar/llvm-6502.git
synced 2024-12-15 04:30:12 +00:00
198d8baafb
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
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@230786 91177308-0d34-0410-b5e6-96231b3b80d8
175 lines
4.8 KiB
LLVM
175 lines
4.8 KiB
LLVM
; Test 32-bit unsigned comparison in which the second operand is a variable.
|
|
;
|
|
; RUN: llc < %s -mtriple=s390x-linux-gnu | FileCheck %s
|
|
|
|
; Check register comparison.
|
|
define double @f1(double %a, double %b, i32 %i1, i32 %i2) {
|
|
; CHECK-LABEL: f1:
|
|
; CHECK: clrjl %r2, %r3
|
|
; CHECK: ldr %f0, %f2
|
|
; CHECK: br %r14
|
|
%cond = icmp ult i32 %i1, %i2
|
|
%res = select i1 %cond, double %a, double %b
|
|
ret double %res
|
|
}
|
|
|
|
; Check the low end of the CL range.
|
|
define double @f2(double %a, double %b, i32 %i1, i32 *%ptr) {
|
|
; CHECK-LABEL: f2:
|
|
; CHECK: cl %r2, 0(%r3)
|
|
; CHECK-NEXT: jl
|
|
; CHECK: ldr %f0, %f2
|
|
; CHECK: br %r14
|
|
%i2 = load i32 *%ptr
|
|
%cond = icmp ult i32 %i1, %i2
|
|
%res = select i1 %cond, double %a, double %b
|
|
ret double %res
|
|
}
|
|
|
|
; Check the high end of the aligned CL range.
|
|
define double @f3(double %a, double %b, i32 %i1, i32 *%base) {
|
|
; CHECK-LABEL: f3:
|
|
; CHECK: cl %r2, 4092(%r3)
|
|
; CHECK-NEXT: jl
|
|
; CHECK: ldr %f0, %f2
|
|
; CHECK: br %r14
|
|
%ptr = getelementptr i32, i32 *%base, i64 1023
|
|
%i2 = load i32 *%ptr
|
|
%cond = icmp ult i32 %i1, %i2
|
|
%res = select i1 %cond, double %a, double %b
|
|
ret double %res
|
|
}
|
|
|
|
; Check the next word up, which should use CLY instead of CL.
|
|
define double @f4(double %a, double %b, i32 %i1, i32 *%base) {
|
|
; CHECK-LABEL: f4:
|
|
; CHECK: cly %r2, 4096(%r3)
|
|
; CHECK-NEXT: jl
|
|
; CHECK: ldr %f0, %f2
|
|
; CHECK: br %r14
|
|
%ptr = getelementptr i32, i32 *%base, i64 1024
|
|
%i2 = load i32 *%ptr
|
|
%cond = icmp ult i32 %i1, %i2
|
|
%res = select i1 %cond, double %a, double %b
|
|
ret double %res
|
|
}
|
|
|
|
; Check the high end of the aligned CLY range.
|
|
define double @f5(double %a, double %b, i32 %i1, i32 *%base) {
|
|
; CHECK-LABEL: f5:
|
|
; CHECK: cly %r2, 524284(%r3)
|
|
; CHECK-NEXT: jl
|
|
; CHECK: ldr %f0, %f2
|
|
; CHECK: br %r14
|
|
%ptr = getelementptr i32, i32 *%base, i64 131071
|
|
%i2 = load i32 *%ptr
|
|
%cond = icmp ult i32 %i1, %i2
|
|
%res = select i1 %cond, double %a, double %b
|
|
ret double %res
|
|
}
|
|
|
|
; Check the next word up, which needs separate address logic.
|
|
; Other sequences besides this one would be OK.
|
|
define double @f6(double %a, double %b, i32 %i1, i32 *%base) {
|
|
; CHECK-LABEL: f6:
|
|
; CHECK: agfi %r3, 524288
|
|
; CHECK: cl %r2, 0(%r3)
|
|
; CHECK-NEXT: jl
|
|
; CHECK: ldr %f0, %f2
|
|
; CHECK: br %r14
|
|
%ptr = getelementptr i32, i32 *%base, i64 131072
|
|
%i2 = load i32 *%ptr
|
|
%cond = icmp ult i32 %i1, %i2
|
|
%res = select i1 %cond, double %a, double %b
|
|
ret double %res
|
|
}
|
|
|
|
; Check the high end of the negative aligned CLY range.
|
|
define double @f7(double %a, double %b, i32 %i1, i32 *%base) {
|
|
; CHECK-LABEL: f7:
|
|
; CHECK: cly %r2, -4(%r3)
|
|
; CHECK-NEXT: jl
|
|
; CHECK: ldr %f0, %f2
|
|
; CHECK: br %r14
|
|
%ptr = getelementptr i32, i32 *%base, i64 -1
|
|
%i2 = load i32 *%ptr
|
|
%cond = icmp ult i32 %i1, %i2
|
|
%res = select i1 %cond, double %a, double %b
|
|
ret double %res
|
|
}
|
|
|
|
; Check the low end of the CLY range.
|
|
define double @f8(double %a, double %b, i32 %i1, i32 *%base) {
|
|
; CHECK-LABEL: f8:
|
|
; CHECK: cly %r2, -524288(%r3)
|
|
; CHECK-NEXT: jl
|
|
; CHECK: ldr %f0, %f2
|
|
; CHECK: br %r14
|
|
%ptr = getelementptr i32, i32 *%base, i64 -131072
|
|
%i2 = load i32 *%ptr
|
|
%cond = icmp ult i32 %i1, %i2
|
|
%res = select i1 %cond, double %a, double %b
|
|
ret double %res
|
|
}
|
|
|
|
; Check the next word down, which needs separate address logic.
|
|
; Other sequences besides this one would be OK.
|
|
define double @f9(double %a, double %b, i32 %i1, i32 *%base) {
|
|
; CHECK-LABEL: f9:
|
|
; CHECK: agfi %r3, -524292
|
|
; CHECK: cl %r2, 0(%r3)
|
|
; CHECK-NEXT: jl
|
|
; CHECK: ldr %f0, %f2
|
|
; CHECK: br %r14
|
|
%ptr = getelementptr i32, i32 *%base, i64 -131073
|
|
%i2 = load i32 *%ptr
|
|
%cond = icmp ult i32 %i1, %i2
|
|
%res = select i1 %cond, double %a, double %b
|
|
ret double %res
|
|
}
|
|
|
|
; Check that CL allows an index.
|
|
define double @f10(double %a, double %b, i32 %i1, i64 %base, i64 %index) {
|
|
; CHECK-LABEL: f10:
|
|
; CHECK: cl %r2, 4092({{%r4,%r3|%r3,%r4}})
|
|
; CHECK-NEXT: jl
|
|
; CHECK: ldr %f0, %f2
|
|
; CHECK: br %r14
|
|
%add1 = add i64 %base, %index
|
|
%add2 = add i64 %add1, 4092
|
|
%ptr = inttoptr i64 %add2 to i32 *
|
|
%i2 = load i32 *%ptr
|
|
%cond = icmp ult i32 %i1, %i2
|
|
%res = select i1 %cond, double %a, double %b
|
|
ret double %res
|
|
}
|
|
|
|
; Check that CLY allows an index.
|
|
define double @f11(double %a, double %b, i32 %i1, i64 %base, i64 %index) {
|
|
; CHECK-LABEL: f11:
|
|
; CHECK: cly %r2, 4096({{%r4,%r3|%r3,%r4}})
|
|
; CHECK-NEXT: jl
|
|
; CHECK: ldr %f0, %f2
|
|
; CHECK: br %r14
|
|
%add1 = add i64 %base, %index
|
|
%add2 = add i64 %add1, 4096
|
|
%ptr = inttoptr i64 %add2 to i32 *
|
|
%i2 = load i32 *%ptr
|
|
%cond = icmp ult i32 %i1, %i2
|
|
%res = select i1 %cond, double %a, double %b
|
|
ret double %res
|
|
}
|
|
|
|
; Check the comparison can be reversed if that allows CL to be used.
|
|
define double @f12(double %a, double %b, i32 %i2, i32 *%ptr) {
|
|
; CHECK-LABEL: f12:
|
|
; CHECK: cl %r2, 0(%r3)
|
|
; CHECK-NEXT: jh {{\.L.*}}
|
|
; CHECK: ldr %f0, %f2
|
|
; CHECK: br %r14
|
|
%i1 = load i32 *%ptr
|
|
%cond = icmp ult i32 %i1, %i2
|
|
%res = select i1 %cond, double %a, double %b
|
|
ret double %res
|
|
}
|