RPCS3/llvm-mirror
1
0
Fork 0
mirror of https://github.com/RPCS3/llvm-mirror.git synced 2024-10-25 22:12:57 +02:00
Code Issues Projects Releases Wiki Activity
f3adc07abf
llvm-mirror/test/CodeGen/X86/SwizzleShuff.ll

70 lines
2.1 KiB
LLVM
Raw Normal View History

[x86] Make vector legalization of extloads work more like the "normal" vector operation legalization with support for custom target lowering and fallback to expand when it fails, and use this to implement sext and anyext load lowering for x86 in a more principled way. Previously, the x86 backend relied on a target DAG combine to "combine away" sextload and extload nodes prior to legalization, or would expand them during legalization with terrible code. This is particularly problematic because the DAG combine relies on running over non-canonical DAG nodes at just the right time to match several common and important patterns. It used a combine rather than lowering because we didn't have good lowering support, and to expose some tricks being employed to more combine phases. With this change it becomes a proper lowering operation, the backend marks that it can lower these nodes, and I've added support for handling the canonical forms that don't have direct legal representations such as sextload of a v4i8 -> v4i64 on AVX1. With this change, our test cases for this behavior continue to pass even after the DAG combiner beigns running more systematically over every node. There is some noise caused by this in the test suite where we actually use vector extends instead of subregister extraction. This doesn't really seem like the right thing to do, but is unlikely to be a critical regression. We do regress in one case where by lowering to the target-specific patterns early we were able to combine away extraneous legal math nodes. However, this regression is completely addressed by switching to a widening based legalization which is what I'm working toward anyways, so I've just switched the test to that mode. Differential Revision: http://reviews.llvm.org/D4654 llvm-svn: 213897
2014-07-25 00:09:56 +02:00
; RUN: llc < %s -mtriple=x86_64-apple-darwin -mcpu=corei7-avx -mattr=+avx -x86-experimental-vector-widening-legalization | FileCheck %s
This commit contains a few changes that had to go in together. 1. Simplify xor/and/or (bitcast(A), bitcast(B)) -> bitcast(op (A,B)) (and also scalar_to_vector). 2. Xor/and/or are indifferent to the swizzle operation (shuffle of one src). Simplify xor/and/or (shuff(A), shuff(B)) -> shuff(op (A, B)) 3. Optimize swizzles of shuffles: shuff(shuff(x, y), undef) -> shuff(x, y). 4. Fix an X86ISelLowering optimization which was very bitcast-sensitive. Code which was previously compiled to this: movd (%rsi), %xmm0 movdqa .LCPI0_0(%rip), %xmm2 pshufb %xmm2, %xmm0 movd (%rdi), %xmm1 pshufb %xmm2, %xmm1 pxor %xmm0, %xmm1 pshufb .LCPI0_1(%rip), %xmm1 movd %xmm1, (%rdi) ret Now compiles to this: movl (%rsi), %eax xorl %eax, (%rdi) ret llvm-svn: 153848
2012-04-01 21:31:22 +02:00
; Check that we perform a scalar XOR on i32.
; CHECK: pull_bitcast
; CHECK: xorl
; CHECK: ret
define void @pull_bitcast (<4 x i8>* %pA, <4 x i8>* %pB) {
[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
2015-02-27 22:17:42 +01:00
%A = load <4 x i8>, <4 x i8>* %pA
%B = load <4 x i8>, <4 x i8>* %pB
This commit contains a few changes that had to go in together. 1. Simplify xor/and/or (bitcast(A), bitcast(B)) -> bitcast(op (A,B)) (and also scalar_to_vector). 2. Xor/and/or are indifferent to the swizzle operation (shuffle of one src). Simplify xor/and/or (shuff(A), shuff(B)) -> shuff(op (A, B)) 3. Optimize swizzles of shuffles: shuff(shuff(x, y), undef) -> shuff(x, y). 4. Fix an X86ISelLowering optimization which was very bitcast-sensitive. Code which was previously compiled to this: movd (%rsi), %xmm0 movdqa .LCPI0_0(%rip), %xmm2 pshufb %xmm2, %xmm0 movd (%rdi), %xmm1 pshufb %xmm2, %xmm1 pxor %xmm0, %xmm1 pshufb .LCPI0_1(%rip), %xmm1 movd %xmm1, (%rdi) ret Now compiles to this: movl (%rsi), %eax xorl %eax, (%rdi) ret llvm-svn: 153848
2012-04-01 21:31:22 +02:00
%C = xor <4 x i8> %A, %B
store <4 x i8> %C, <4 x i8>* %pA
ret void
}
Optimizing swizzles of complex shuffles may generate additional complex shuffles. Do not try to optimize swizzles of shuffles if the source shuffle has more than a single user, except when the source shuffle is also a swizzle. llvm-svn: 153864
2012-04-02 09:11:12 +02:00
; CHECK: multi_use_swizzle
[x86] Teach the 128-bit vector shuffle lowering routines to take advantage of the existence of a reasonable blend instruction. The 256-bit vector shuffle lowering has leveraged the general technique of decomposed shuffles and blends for quite some time, but this never made it back into the 128-bit code, and there are a large number of patterns where this is substantially better. For example, this removes almost all domain crossing in vector shuffles that involve some blend and some permutation with SSE4.1 and later. See the massive reduction in 'shufps' for integer test cases in this commit. This isn't perfect yet for a few reasons: 1) The v8i16 shuffle lowering continues to plague me. We don't always form an unpack-based blend when that would be better. But the wins pretty drastically outstrip the losses here. 2) The v16i8 shuffle lowering is just a disaster here. I never went and implemented blend support here for some terrible reason. I'll do that next probably. I've not updated it for now. More variations on this technique are coming as well -- we don't shuffle-into-unpack or shuffle-into-palignr, both of which would also be profitable. Note that some test cases grow significantly in the number of instructions, but I expect to actually be faster. We use pshufd+pshufd+blendw instead of a single shufps, but the pshufd's are very likely to pipeline well (two ports on most modern intel chips) and the blend is a *very* fast instruction. The domain switch penalty will essentially always be more than a blend instruction, which is the only increase in tree height. llvm-svn: 229350
2015-02-16 02:52:02 +01:00
; CHECK: pshufd
; CHECK-NEXT: pshufd
; CHECK-NEXT: pblendw
; CHECK-NEXT: pshufd
; CHECK-NEXT: pshufd
; CHECK-NEXT: pxor
Optimizing swizzles of complex shuffles may generate additional complex shuffles. Do not try to optimize swizzles of shuffles if the source shuffle has more than a single user, except when the source shuffle is also a swizzle. llvm-svn: 153864
2012-04-02 09:11:12 +02:00
; CHECK-NEXT: ret
define <4 x i32> @multi_use_swizzle (<4 x i32>* %pA, <4 x i32>* %pB) {
[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
2015-02-27 22:17:42 +01:00
%A = load <4 x i32>, <4 x i32>* %pA
%B = load <4 x i32>, <4 x i32>* %pB
Optimizing swizzles of complex shuffles may generate additional complex shuffles. Do not try to optimize swizzles of shuffles if the source shuffle has more than a single user, except when the source shuffle is also a swizzle. llvm-svn: 153864
2012-04-02 09:11:12 +02:00
%S = shufflevector <4 x i32> %A, <4 x i32> %B, <4 x i32> <i32 1, i32 1, i32 5, i32 6>
%S1 = shufflevector <4 x i32> %S, <4 x i32> undef, <4 x i32> <i32 1, i32 3, i32 2, i32 2>
%S2 = shufflevector <4 x i32> %S, <4 x i32> undef, <4 x i32> <i32 2, i32 1, i32 0, i32 2>
%R = xor <4 x i32> %S1, %S2
ret <4 x i32> %R
}
Add an additional testcase which checks ops with multiple users. llvm-svn: 153939
2012-04-03 09:39:36 +02:00
; CHECK: pull_bitcast2
; CHECK: xorl
; CHECK: ret
define <4 x i8> @pull_bitcast2 (<4 x i8>* %pA, <4 x i8>* %pB, <4 x i8>* %pC) {
[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
2015-02-27 22:17:42 +01:00
%A = load <4 x i8>, <4 x i8>* %pA
Add an additional testcase which checks ops with multiple users. llvm-svn: 153939
2012-04-03 09:39:36 +02:00
store <4 x i8> %A, <4 x i8>* %pC
[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
2015-02-27 22:17:42 +01:00
%B = load <4 x i8>, <4 x i8>* %pB
Add an additional testcase which checks ops with multiple users. llvm-svn: 153939
2012-04-03 09:39:36 +02:00
%C = xor <4 x i8> %A, %B
store <4 x i8> %C, <4 x i8>* %pA
ret <4 x i8> %C
}
1. Remove the part of r153848 which optimizes shuffle-of-shuffle into a new shuffle node because it could introduce new shuffle nodes that were not supported efficiently by the target. 2. Add a more restrictive shuffle-of-shuffle optimization for cases where the second shuffle reverses the transformation of the first shuffle. llvm-svn: 154266
2012-04-07 23:19:08 +02:00
; CHECK: reverse_1
[x86] Teach the 128-bit vector shuffle lowering routines to take advantage of the existence of a reasonable blend instruction. The 256-bit vector shuffle lowering has leveraged the general technique of decomposed shuffles and blends for quite some time, but this never made it back into the 128-bit code, and there are a large number of patterns where this is substantially better. For example, this removes almost all domain crossing in vector shuffles that involve some blend and some permutation with SSE4.1 and later. See the massive reduction in 'shufps' for integer test cases in this commit. This isn't perfect yet for a few reasons: 1) The v8i16 shuffle lowering continues to plague me. We don't always form an unpack-based blend when that would be better. But the wins pretty drastically outstrip the losses here. 2) The v16i8 shuffle lowering is just a disaster here. I never went and implemented blend support here for some terrible reason. I'll do that next probably. I've not updated it for now. More variations on this technique are coming as well -- we don't shuffle-into-unpack or shuffle-into-palignr, both of which would also be profitable. Note that some test cases grow significantly in the number of instructions, but I expect to actually be faster. We use pshufd+pshufd+blendw instead of a single shufps, but the pshufd's are very likely to pipeline well (two ports on most modern intel chips) and the blend is a *very* fast instruction. The domain switch penalty will essentially always be more than a blend instruction, which is the only increase in tree height. llvm-svn: 229350
2015-02-16 02:52:02 +01:00
; CHECK-NOT: pshufd
1. Remove the part of r153848 which optimizes shuffle-of-shuffle into a new shuffle node because it could introduce new shuffle nodes that were not supported efficiently by the target. 2. Add a more restrictive shuffle-of-shuffle optimization for cases where the second shuffle reverses the transformation of the first shuffle. llvm-svn: 154266
2012-04-07 23:19:08 +02:00
; CHECK: ret
define <4 x i32> @reverse_1 (<4 x i32>* %pA, <4 x i32>* %pB) {
[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
2015-02-27 22:17:42 +01:00
%A = load <4 x i32>, <4 x i32>* %pA
%B = load <4 x i32>, <4 x i32>* %pB
1. Remove the part of r153848 which optimizes shuffle-of-shuffle into a new shuffle node because it could introduce new shuffle nodes that were not supported efficiently by the target. 2. Add a more restrictive shuffle-of-shuffle optimization for cases where the second shuffle reverses the transformation of the first shuffle. llvm-svn: 154266
2012-04-07 23:19:08 +02:00
%S = shufflevector <4 x i32> %A, <4 x i32> %B, <4 x i32> <i32 1, i32 0, i32 3, i32 2>
%S1 = shufflevector <4 x i32> %S, <4 x i32> undef, <4 x i32> <i32 1, i32 0, i32 3, i32 2>
ret <4 x i32> %S1
}
; CHECK: no_reverse_shuff
[x86] Teach the 128-bit vector shuffle lowering routines to take advantage of the existence of a reasonable blend instruction. The 256-bit vector shuffle lowering has leveraged the general technique of decomposed shuffles and blends for quite some time, but this never made it back into the 128-bit code, and there are a large number of patterns where this is substantially better. For example, this removes almost all domain crossing in vector shuffles that involve some blend and some permutation with SSE4.1 and later. See the massive reduction in 'shufps' for integer test cases in this commit. This isn't perfect yet for a few reasons: 1) The v8i16 shuffle lowering continues to plague me. We don't always form an unpack-based blend when that would be better. But the wins pretty drastically outstrip the losses here. 2) The v16i8 shuffle lowering is just a disaster here. I never went and implemented blend support here for some terrible reason. I'll do that next probably. I've not updated it for now. More variations on this technique are coming as well -- we don't shuffle-into-unpack or shuffle-into-palignr, both of which would also be profitable. Note that some test cases grow significantly in the number of instructions, but I expect to actually be faster. We use pshufd+pshufd+blendw instead of a single shufps, but the pshufd's are very likely to pipeline well (two ports on most modern intel chips) and the blend is a *very* fast instruction. The domain switch penalty will essentially always be more than a blend instruction, which is the only increase in tree height. llvm-svn: 229350
2015-02-16 02:52:02 +01:00
; CHECK: pshufd
1. Remove the part of r153848 which optimizes shuffle-of-shuffle into a new shuffle node because it could introduce new shuffle nodes that were not supported efficiently by the target. 2. Add a more restrictive shuffle-of-shuffle optimization for cases where the second shuffle reverses the transformation of the first shuffle. llvm-svn: 154266
2012-04-07 23:19:08 +02:00
; CHECK: ret
define <4 x i32> @no_reverse_shuff (<4 x i32>* %pA, <4 x i32>* %pB) {
[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
2015-02-27 22:17:42 +01:00
%A = load <4 x i32>, <4 x i32>* %pA
%B = load <4 x i32>, <4 x i32>* %pB
1. Remove the part of r153848 which optimizes shuffle-of-shuffle into a new shuffle node because it could introduce new shuffle nodes that were not supported efficiently by the target. 2. Add a more restrictive shuffle-of-shuffle optimization for cases where the second shuffle reverses the transformation of the first shuffle. llvm-svn: 154266
2012-04-07 23:19:08 +02:00
%S = shufflevector <4 x i32> %A, <4 x i32> %B, <4 x i32> <i32 1, i32 0, i32 3, i32 2>
%S1 = shufflevector <4 x i32> %S, <4 x i32> undef, <4 x i32> <i32 3, i32 2, i32 3, i32 2>
ret <4 x i32> %S1
}
Reference in New Issue Copy Permalink