RPCS3/llvm-mirror
1
0
Fork 0
mirror of https://github.com/RPCS3/llvm-mirror.git synced 2025-02-01 05:01:59 +01:00
Code Issues Projects Releases Wiki Activity
llvm-mirror/test/CodeGen/PowerPC/fast-isel-load-store.ll

221 lines
5.0 KiB
LLVM
Raw Normal View History

Use PIC relocation model as default for PowerPC64 ELF. Most of the PowerPC64 code generation for the ELF ABI is already PIC. There are four main exceptions: (1) Constant pointer arrays etc. should in writeable sections. (2) The TOC restoration NOP after a call is needed for all global symbols. While GNU ld has a workaround for questionable GCC self-calls, we trigger the checks for calls from COMDAT sections as they cross input sections and are therefore not considered self-calls. The current decision is questionable and suboptimal, but outside the scope of the change. (3) TLS access can not use the initial-exec model. (4) Jump tables should use relative addresses. Note that the current encoding doesn't work for the large code model, but it is more compact than the default for any non-trivial jump table. Improving this is again beyond the scope of this change. At least (1) and (3) are assumptions made in target-independent code and introducing additional hooks is a bit messy. Testing with clang shows that a -fPIC binary is 600KB smaller than the corresponding -fno-pic build. Separate testing from improved jump table encodings would explain only about 100KB or so. The rest is expected to be a result of more aggressive immediate forming for -fno-pic, where the -fPIC binary just uses TOC entries. This change brings the LLVM output in line with the GCC output, other PPC64 compilers like XLC on AIX are known to produce PIC by default as well. The relocation model can still be provided explicitly, i.e. when using MCJIT. One test case for case (1) is included, other test cases with relocation mode sensitive behavior are wired to static for now. They will be reviewed and adjusted separately. Differential Revision: https://reviews.llvm.org/D26566 llvm-svn: 289743
2016-12-15 00:01:53 +00:00
; RUN: llc -relocation-model=static < %s -O0 -verify-machineinstrs -fast-isel -fast-isel-abort=1 -mattr=-vsx -mtriple=powerpc64-unknown-linux-gnu -mcpu=pwr7 | FileCheck %s --check-prefix=ELF64
[NFC][PowerPC] Fast-isel VSX support test We have fixed most of the VSX limitation in Fast-isel, so we can remove the -mattr=-vsx for most testcases now. llvm-svn: 372345
2019-09-19 18:18:18 +00:00
; RUN: llc -relocation-model=static < %s -O0 -verify-machineinstrs -fast-isel -fast-isel-abort=1 -mattr=+vsx -mtriple=powerpc64-unknown-linux-gnu -mcpu=pwr7 | FileCheck %s --check-prefix=VSX
Introduce codegen for the Signal Processing Engine Summary: The Signal Processing Engine (SPE) is found on NXP/Freescale e500v1, e500v2, and several e200 cores. This adds support targeting the e500v2, as this is more common than the e500v1, and is in SoCs still on the market. This patch is very intrusive because the SPE is binary incompatible with the traditional FPU. After discussing with others, the cleanest solution was to make both SPE and FPU features on top of a base PowerPC subset, so all FPU instructions are now wrapped with HasFPU predicates. Supported by this are: * Code generation following the SPE ABI at the LLVM IR level (calling conventions) * Single- and Double-precision math at the level supported by the APU. Still to do: * Vector operations * SPE intrinsics As this changes the Callee-saved register list order, one test, which tests the precise generated code, was updated to account for the new register order. Reviewed by: nemanjai Differential Revision: https://reviews.llvm.org/D44830 llvm-svn: 337347
2018-07-18 04:25:10 +00:00
; RUN: llc -relocation-model=static < %s -O0 -verify-machineinstrs -fast-isel -fast-isel-abort=1 -mattr=spe -mtriple=powerpc-unknown-linux-gnu -mcpu=e500 | FileCheck %s --check-prefix=SPE
[PowerPC] Add loads, stores, and related things to fast-isel. This is the next big chunk of fast-isel code. The primary purpose is to implement selection of loads and stores, but there is a lot of drag-along to support this. The common code to analyze addresses for both loads and stores is substantial. It's also necessary to add the materialization code for global values. Related to load-store processing is the code to fold loads into integer extends, since otherwise we generate lots of redundant instructions. We also need to add some overrides to some FastEmit routines to ensure we don't assign GPR 0 to a virtual register when this would change the meaning of an instruction. I added handling selection of a few binary arithmetic instructions, to enable committing some test cases I wrote a while back. Finally, ap couple of miscellaneous changes: * I cleaned up some poor style from a previous patch in PPCISelLowering.cpp, pointed out by David Blaikie. * I enlarged the Addr.Offset field to avoid sign problems with 32-bit offsets. llvm-svn: 189636
2013-08-30 02:29:45 +00:00
; This test verifies that load/store instructions are properly generated,
; and that they pass MI verification.
[PowerPC][test] Add explicit dso_local to definitions in ELF static relocation model tests TargetMachine::shouldAssumeDSOLocal currently implies dso_local for such definitions. Adding explicit dso_local makes these tests align with the clang -fpic behavior and allow the removal of the TargetMachine::shouldAssumeDSOLocal special case. Rewrite preemption.ll to dsolocal-static.ll and dsolocal-pic.ll, and add "PIC Level" metadata.
2020-12-30 10:32:34 -08:00
@a = dso_local global i8 1, align 1
@b = dso_local global i16 2, align 2
@c = dso_local global i32 4, align 4
@d = dso_local global i64 8, align 8
@e = dso_local global float 1.25, align 4
@f = dso_local global double 3.5, align 8
[PowerPC] Add loads, stores, and related things to fast-isel. This is the next big chunk of fast-isel code. The primary purpose is to implement selection of loads and stores, but there is a lot of drag-along to support this. The common code to analyze addresses for both loads and stores is substantial. It's also necessary to add the materialization code for global values. Related to load-store processing is the code to fold loads into integer extends, since otherwise we generate lots of redundant instructions. We also need to add some overrides to some FastEmit routines to ensure we don't assign GPR 0 to a virtual register when this would change the meaning of an instruction. I added handling selection of a few binary arithmetic instructions, to enable committing some test cases I wrote a while back. Finally, ap couple of miscellaneous changes: * I cleaned up some poor style from a previous patch in PPCISelLowering.cpp, pointed out by David Blaikie. * I enlarged the Addr.Offset field to avoid sign problems with 32-bit offsets. llvm-svn: 189636
2013-08-30 02:29:45 +00:00
%struct.s = type<{ i8, i32 }>
%struct.t = type<{ i8, i64 }>
[PowerPC][test] Add explicit dso_local to definitions in ELF static relocation model tests TargetMachine::shouldAssumeDSOLocal currently implies dso_local for such definitions. Adding explicit dso_local makes these tests align with the clang -fpic behavior and allow the removal of the TargetMachine::shouldAssumeDSOLocal special case. Rewrite preemption.ll to dsolocal-static.ll and dsolocal-pic.ll, and add "PIC Level" metadata.
2020-12-30 10:32:34 -08:00
@g = dso_local global %struct.s <{ i8 1, i32 2 }>, align 1
@h = dso_local global %struct.t <{ i8 1, i64 2 }>, align 1
[PowerPC] Add loads, stores, and related things to fast-isel. This is the next big chunk of fast-isel code. The primary purpose is to implement selection of loads and stores, but there is a lot of drag-along to support this. The common code to analyze addresses for both loads and stores is substantial. It's also necessary to add the materialization code for global values. Related to load-store processing is the code to fold loads into integer extends, since otherwise we generate lots of redundant instructions. We also need to add some overrides to some FastEmit routines to ensure we don't assign GPR 0 to a virtual register when this would change the meaning of an instruction. I added handling selection of a few binary arithmetic instructions, to enable committing some test cases I wrote a while back. Finally, ap couple of miscellaneous changes: * I cleaned up some poor style from a previous patch in PPCISelLowering.cpp, pointed out by David Blaikie. * I enlarged the Addr.Offset field to avoid sign problems with 32-bit offsets. llvm-svn: 189636
2013-08-30 02:29:45 +00:00
[PowerPC][test] Add explicit dso_local to definitions in ELF static relocation model tests TargetMachine::shouldAssumeDSOLocal currently implies dso_local for such definitions. Adding explicit dso_local makes these tests align with the clang -fpic behavior and allow the removal of the TargetMachine::shouldAssumeDSOLocal special case. Rewrite preemption.ll to dsolocal-static.ll and dsolocal-pic.ll, and add "PIC Level" metadata.
2020-12-30 10:32:34 -08:00
@i = common dso_local global [8192 x i64] zeroinitializer, align 8
[PowerPC] Add loads, stores, and related things to fast-isel. This is the next big chunk of fast-isel code. The primary purpose is to implement selection of loads and stores, but there is a lot of drag-along to support this. The common code to analyze addresses for both loads and stores is substantial. It's also necessary to add the materialization code for global values. Related to load-store processing is the code to fold loads into integer extends, since otherwise we generate lots of redundant instructions. We also need to add some overrides to some FastEmit routines to ensure we don't assign GPR 0 to a virtual register when this would change the meaning of an instruction. I added handling selection of a few binary arithmetic instructions, to enable committing some test cases I wrote a while back. Finally, ap couple of miscellaneous changes: * I cleaned up some poor style from a previous patch in PPCISelLowering.cpp, pointed out by David Blaikie. * I enlarged the Addr.Offset field to avoid sign problems with 32-bit offsets. llvm-svn: 189636
2013-08-30 02:29:45 +00:00
; load
Clean up function attributes on PPC fast-isel tests. llvm-svn: 243079
2015-07-24 01:07:50 +00:00
define i8 @t1() nounwind {
NFC: Use -LABEL more There were a number of tests needing updates for D91734, and I added a bunch of LABEL directives to help track down where those had to go. These directives are an improvement independent of the functional patch, so I'm committing them as their own separate patch.
2021-01-11 07:03:08 -08:00
; ELF64-LABEL: t1:
[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 21:17:42 +00:00
%1 = load i8, i8* @a, align 1
[PowerPC] Add loads, stores, and related things to fast-isel. This is the next big chunk of fast-isel code. The primary purpose is to implement selection of loads and stores, but there is a lot of drag-along to support this. The common code to analyze addresses for both loads and stores is substantial. It's also necessary to add the materialization code for global values. Related to load-store processing is the code to fold loads into integer extends, since otherwise we generate lots of redundant instructions. We also need to add some overrides to some FastEmit routines to ensure we don't assign GPR 0 to a virtual register when this would change the meaning of an instruction. I added handling selection of a few binary arithmetic instructions, to enable committing some test cases I wrote a while back. Finally, ap couple of miscellaneous changes: * I cleaned up some poor style from a previous patch in PPCISelLowering.cpp, pointed out by David Blaikie. * I enlarged the Addr.Offset field to avoid sign problems with 32-bit offsets. llvm-svn: 189636
2013-08-30 02:29:45 +00:00
; ELF64: lbz
%2 = add nsw i8 %1, 1
; ELF64: addi
ret i8 %2
}
Clean up function attributes on PPC fast-isel tests. llvm-svn: 243079
2015-07-24 01:07:50 +00:00
define i16 @t2() nounwind {
NFC: Use -LABEL more There were a number of tests needing updates for D91734, and I added a bunch of LABEL directives to help track down where those had to go. These directives are an improvement independent of the functional patch, so I'm committing them as their own separate patch.
2021-01-11 07:03:08 -08:00
; ELF64-LABEL: t2:
[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 21:17:42 +00:00
%1 = load i16, i16* @b, align 2
[PowerPC] Add loads, stores, and related things to fast-isel. This is the next big chunk of fast-isel code. The primary purpose is to implement selection of loads and stores, but there is a lot of drag-along to support this. The common code to analyze addresses for both loads and stores is substantial. It's also necessary to add the materialization code for global values. Related to load-store processing is the code to fold loads into integer extends, since otherwise we generate lots of redundant instructions. We also need to add some overrides to some FastEmit routines to ensure we don't assign GPR 0 to a virtual register when this would change the meaning of an instruction. I added handling selection of a few binary arithmetic instructions, to enable committing some test cases I wrote a while back. Finally, ap couple of miscellaneous changes: * I cleaned up some poor style from a previous patch in PPCISelLowering.cpp, pointed out by David Blaikie. * I enlarged the Addr.Offset field to avoid sign problems with 32-bit offsets. llvm-svn: 189636
2013-08-30 02:29:45 +00:00
; ELF64: lhz
%2 = add nsw i16 %1, 1
; ELF64: addi
ret i16 %2
}
[PowerPC][test] Add explicit dso_local to definitions in ELF static relocation model tests TargetMachine::shouldAssumeDSOLocal currently implies dso_local for such definitions. Adding explicit dso_local makes these tests align with the clang -fpic behavior and allow the removal of the TargetMachine::shouldAssumeDSOLocal special case. Rewrite preemption.ll to dsolocal-static.ll and dsolocal-pic.ll, and add "PIC Level" metadata.
2020-12-30 10:32:34 -08:00
define dso_local i32 @t3() nounwind {
NFC: Use -LABEL more There were a number of tests needing updates for D91734, and I added a bunch of LABEL directives to help track down where those had to go. These directives are an improvement independent of the functional patch, so I'm committing them as their own separate patch.
2021-01-11 07:03:08 -08:00
; ELF64-LABEL: t3:
[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 21:17:42 +00:00
%1 = load i32, i32* @c, align 4
[PowerPC] Add loads, stores, and related things to fast-isel. This is the next big chunk of fast-isel code. The primary purpose is to implement selection of loads and stores, but there is a lot of drag-along to support this. The common code to analyze addresses for both loads and stores is substantial. It's also necessary to add the materialization code for global values. Related to load-store processing is the code to fold loads into integer extends, since otherwise we generate lots of redundant instructions. We also need to add some overrides to some FastEmit routines to ensure we don't assign GPR 0 to a virtual register when this would change the meaning of an instruction. I added handling selection of a few binary arithmetic instructions, to enable committing some test cases I wrote a while back. Finally, ap couple of miscellaneous changes: * I cleaned up some poor style from a previous patch in PPCISelLowering.cpp, pointed out by David Blaikie. * I enlarged the Addr.Offset field to avoid sign problems with 32-bit offsets. llvm-svn: 189636
2013-08-30 02:29:45 +00:00
; ELF64: lwz
%2 = add nsw i32 %1, 1
; ELF64: addi
ret i32 %2
}
Clean up function attributes on PPC fast-isel tests. llvm-svn: 243079
2015-07-24 01:07:50 +00:00
define i64 @t4() nounwind {
NFC: Use -LABEL more There were a number of tests needing updates for D91734, and I added a bunch of LABEL directives to help track down where those had to go. These directives are an improvement independent of the functional patch, so I'm committing them as their own separate patch.
2021-01-11 07:03:08 -08:00
; ELF64-LABEL: t4:
[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 21:17:42 +00:00
%1 = load i64, i64* @d, align 4
[PowerPC] Add loads, stores, and related things to fast-isel. This is the next big chunk of fast-isel code. The primary purpose is to implement selection of loads and stores, but there is a lot of drag-along to support this. The common code to analyze addresses for both loads and stores is substantial. It's also necessary to add the materialization code for global values. Related to load-store processing is the code to fold loads into integer extends, since otherwise we generate lots of redundant instructions. We also need to add some overrides to some FastEmit routines to ensure we don't assign GPR 0 to a virtual register when this would change the meaning of an instruction. I added handling selection of a few binary arithmetic instructions, to enable committing some test cases I wrote a while back. Finally, ap couple of miscellaneous changes: * I cleaned up some poor style from a previous patch in PPCISelLowering.cpp, pointed out by David Blaikie. * I enlarged the Addr.Offset field to avoid sign problems with 32-bit offsets. llvm-svn: 189636
2013-08-30 02:29:45 +00:00
; ELF64: ld
%2 = add nsw i64 %1, 1
; ELF64: addi
ret i64 %2
}
[PowerPC][test] Add explicit dso_local to definitions in ELF static relocation model tests TargetMachine::shouldAssumeDSOLocal currently implies dso_local for such definitions. Adding explicit dso_local makes these tests align with the clang -fpic behavior and allow the removal of the TargetMachine::shouldAssumeDSOLocal special case. Rewrite preemption.ll to dsolocal-static.ll and dsolocal-pic.ll, and add "PIC Level" metadata.
2020-12-30 10:32:34 -08:00
define dso_local float @t5() nounwind {
NFC: Use -LABEL more There were a number of tests needing updates for D91734, and I added a bunch of LABEL directives to help track down where those had to go. These directives are an improvement independent of the functional patch, so I'm committing them as their own separate patch.
2021-01-11 07:03:08 -08:00
; ELF64-LABEL: t5:
; SPE-LABEL: t5:
[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 21:17:42 +00:00
%1 = load float, float* @e, align 4
[PowerPC] Add loads, stores, and related things to fast-isel. This is the next big chunk of fast-isel code. The primary purpose is to implement selection of loads and stores, but there is a lot of drag-along to support this. The common code to analyze addresses for both loads and stores is substantial. It's also necessary to add the materialization code for global values. Related to load-store processing is the code to fold loads into integer extends, since otherwise we generate lots of redundant instructions. We also need to add some overrides to some FastEmit routines to ensure we don't assign GPR 0 to a virtual register when this would change the meaning of an instruction. I added handling selection of a few binary arithmetic instructions, to enable committing some test cases I wrote a while back. Finally, ap couple of miscellaneous changes: * I cleaned up some poor style from a previous patch in PPCISelLowering.cpp, pointed out by David Blaikie. * I enlarged the Addr.Offset field to avoid sign problems with 32-bit offsets. llvm-svn: 189636
2013-08-30 02:29:45 +00:00
; ELF64: lfs
Introduce codegen for the Signal Processing Engine Summary: The Signal Processing Engine (SPE) is found on NXP/Freescale e500v1, e500v2, and several e200 cores. This adds support targeting the e500v2, as this is more common than the e500v1, and is in SoCs still on the market. This patch is very intrusive because the SPE is binary incompatible with the traditional FPU. After discussing with others, the cleanest solution was to make both SPE and FPU features on top of a base PowerPC subset, so all FPU instructions are now wrapped with HasFPU predicates. Supported by this are: * Code generation following the SPE ABI at the LLVM IR level (calling conventions) * Single- and Double-precision math at the level supported by the APU. Still to do: * Vector operations * SPE intrinsics As this changes the Callee-saved register list order, one test, which tests the precise generated code, was updated to account for the new register order. Reviewed by: nemanjai Differential Revision: https://reviews.llvm.org/D44830 llvm-svn: 337347
2018-07-18 04:25:10 +00:00
; SPE: lwz
[PowerPC] Add loads, stores, and related things to fast-isel. This is the next big chunk of fast-isel code. The primary purpose is to implement selection of loads and stores, but there is a lot of drag-along to support this. The common code to analyze addresses for both loads and stores is substantial. It's also necessary to add the materialization code for global values. Related to load-store processing is the code to fold loads into integer extends, since otherwise we generate lots of redundant instructions. We also need to add some overrides to some FastEmit routines to ensure we don't assign GPR 0 to a virtual register when this would change the meaning of an instruction. I added handling selection of a few binary arithmetic instructions, to enable committing some test cases I wrote a while back. Finally, ap couple of miscellaneous changes: * I cleaned up some poor style from a previous patch in PPCISelLowering.cpp, pointed out by David Blaikie. * I enlarged the Addr.Offset field to avoid sign problems with 32-bit offsets. llvm-svn: 189636
2013-08-30 02:29:45 +00:00
%2 = fadd float %1, 1.0
; ELF64: fadds
Introduce codegen for the Signal Processing Engine Summary: The Signal Processing Engine (SPE) is found on NXP/Freescale e500v1, e500v2, and several e200 cores. This adds support targeting the e500v2, as this is more common than the e500v1, and is in SoCs still on the market. This patch is very intrusive because the SPE is binary incompatible with the traditional FPU. After discussing with others, the cleanest solution was to make both SPE and FPU features on top of a base PowerPC subset, so all FPU instructions are now wrapped with HasFPU predicates. Supported by this are: * Code generation following the SPE ABI at the LLVM IR level (calling conventions) * Single- and Double-precision math at the level supported by the APU. Still to do: * Vector operations * SPE intrinsics As this changes the Callee-saved register list order, one test, which tests the precise generated code, was updated to account for the new register order. Reviewed by: nemanjai Differential Revision: https://reviews.llvm.org/D44830 llvm-svn: 337347
2018-07-18 04:25:10 +00:00
; SPE: efsadd
[PowerPC] Add loads, stores, and related things to fast-isel. This is the next big chunk of fast-isel code. The primary purpose is to implement selection of loads and stores, but there is a lot of drag-along to support this. The common code to analyze addresses for both loads and stores is substantial. It's also necessary to add the materialization code for global values. Related to load-store processing is the code to fold loads into integer extends, since otherwise we generate lots of redundant instructions. We also need to add some overrides to some FastEmit routines to ensure we don't assign GPR 0 to a virtual register when this would change the meaning of an instruction. I added handling selection of a few binary arithmetic instructions, to enable committing some test cases I wrote a while back. Finally, ap couple of miscellaneous changes: * I cleaned up some poor style from a previous patch in PPCISelLowering.cpp, pointed out by David Blaikie. * I enlarged the Addr.Offset field to avoid sign problems with 32-bit offsets. llvm-svn: 189636
2013-08-30 02:29:45 +00:00
ret float %2
}
[PowerPC][test] Add explicit dso_local to definitions in ELF static relocation model tests TargetMachine::shouldAssumeDSOLocal currently implies dso_local for such definitions. Adding explicit dso_local makes these tests align with the clang -fpic behavior and allow the removal of the TargetMachine::shouldAssumeDSOLocal special case. Rewrite preemption.ll to dsolocal-static.ll and dsolocal-pic.ll, and add "PIC Level" metadata.
2020-12-30 10:32:34 -08:00
define dso_local double @t6() nounwind {
NFC: Use -LABEL more There were a number of tests needing updates for D91734, and I added a bunch of LABEL directives to help track down where those had to go. These directives are an improvement independent of the functional patch, so I'm committing them as their own separate patch.
2021-01-11 07:03:08 -08:00
; ELF64-LABEL: t6:
; SPE-LABEL: t6:
[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 21:17:42 +00:00
%1 = load double, double* @f, align 8
[PowerPC] Add loads, stores, and related things to fast-isel. This is the next big chunk of fast-isel code. The primary purpose is to implement selection of loads and stores, but there is a lot of drag-along to support this. The common code to analyze addresses for both loads and stores is substantial. It's also necessary to add the materialization code for global values. Related to load-store processing is the code to fold loads into integer extends, since otherwise we generate lots of redundant instructions. We also need to add some overrides to some FastEmit routines to ensure we don't assign GPR 0 to a virtual register when this would change the meaning of an instruction. I added handling selection of a few binary arithmetic instructions, to enable committing some test cases I wrote a while back. Finally, ap couple of miscellaneous changes: * I cleaned up some poor style from a previous patch in PPCISelLowering.cpp, pointed out by David Blaikie. * I enlarged the Addr.Offset field to avoid sign problems with 32-bit offsets. llvm-svn: 189636
2013-08-30 02:29:45 +00:00
; ELF64: lfd
[NFC][PowerPC] Fast-isel VSX support test We have fixed most of the VSX limitation in Fast-isel, so we can remove the -mattr=-vsx for most testcases now. llvm-svn: 372345
2019-09-19 18:18:18 +00:00
; VSX: lxsdx
Introduce codegen for the Signal Processing Engine Summary: The Signal Processing Engine (SPE) is found on NXP/Freescale e500v1, e500v2, and several e200 cores. This adds support targeting the e500v2, as this is more common than the e500v1, and is in SoCs still on the market. This patch is very intrusive because the SPE is binary incompatible with the traditional FPU. After discussing with others, the cleanest solution was to make both SPE and FPU features on top of a base PowerPC subset, so all FPU instructions are now wrapped with HasFPU predicates. Supported by this are: * Code generation following the SPE ABI at the LLVM IR level (calling conventions) * Single- and Double-precision math at the level supported by the APU. Still to do: * Vector operations * SPE intrinsics As this changes the Callee-saved register list order, one test, which tests the precise generated code, was updated to account for the new register order. Reviewed by: nemanjai Differential Revision: https://reviews.llvm.org/D44830 llvm-svn: 337347
2018-07-18 04:25:10 +00:00
; SPE: evldd
[PowerPC] Add loads, stores, and related things to fast-isel. This is the next big chunk of fast-isel code. The primary purpose is to implement selection of loads and stores, but there is a lot of drag-along to support this. The common code to analyze addresses for both loads and stores is substantial. It's also necessary to add the materialization code for global values. Related to load-store processing is the code to fold loads into integer extends, since otherwise we generate lots of redundant instructions. We also need to add some overrides to some FastEmit routines to ensure we don't assign GPR 0 to a virtual register when this would change the meaning of an instruction. I added handling selection of a few binary arithmetic instructions, to enable committing some test cases I wrote a while back. Finally, ap couple of miscellaneous changes: * I cleaned up some poor style from a previous patch in PPCISelLowering.cpp, pointed out by David Blaikie. * I enlarged the Addr.Offset field to avoid sign problems with 32-bit offsets. llvm-svn: 189636
2013-08-30 02:29:45 +00:00
%2 = fadd double %1, 1.0
; ELF64: fadd
[NFC][PowerPC] Fast-isel VSX support test We have fixed most of the VSX limitation in Fast-isel, so we can remove the -mattr=-vsx for most testcases now. llvm-svn: 372345
2019-09-19 18:18:18 +00:00
; VSX: xsadddp
Introduce codegen for the Signal Processing Engine Summary: The Signal Processing Engine (SPE) is found on NXP/Freescale e500v1, e500v2, and several e200 cores. This adds support targeting the e500v2, as this is more common than the e500v1, and is in SoCs still on the market. This patch is very intrusive because the SPE is binary incompatible with the traditional FPU. After discussing with others, the cleanest solution was to make both SPE and FPU features on top of a base PowerPC subset, so all FPU instructions are now wrapped with HasFPU predicates. Supported by this are: * Code generation following the SPE ABI at the LLVM IR level (calling conventions) * Single- and Double-precision math at the level supported by the APU. Still to do: * Vector operations * SPE intrinsics As this changes the Callee-saved register list order, one test, which tests the precise generated code, was updated to account for the new register order. Reviewed by: nemanjai Differential Revision: https://reviews.llvm.org/D44830 llvm-svn: 337347
2018-07-18 04:25:10 +00:00
; SPE: efdadd
[PowerPC] Add loads, stores, and related things to fast-isel. This is the next big chunk of fast-isel code. The primary purpose is to implement selection of loads and stores, but there is a lot of drag-along to support this. The common code to analyze addresses for both loads and stores is substantial. It's also necessary to add the materialization code for global values. Related to load-store processing is the code to fold loads into integer extends, since otherwise we generate lots of redundant instructions. We also need to add some overrides to some FastEmit routines to ensure we don't assign GPR 0 to a virtual register when this would change the meaning of an instruction. I added handling selection of a few binary arithmetic instructions, to enable committing some test cases I wrote a while back. Finally, ap couple of miscellaneous changes: * I cleaned up some poor style from a previous patch in PPCISelLowering.cpp, pointed out by David Blaikie. * I enlarged the Addr.Offset field to avoid sign problems with 32-bit offsets. llvm-svn: 189636
2013-08-30 02:29:45 +00:00
ret double %2
}
; store
[PowerPC][test] Add explicit dso_local to definitions in ELF static relocation model tests TargetMachine::shouldAssumeDSOLocal currently implies dso_local for such definitions. Adding explicit dso_local makes these tests align with the clang -fpic behavior and allow the removal of the TargetMachine::shouldAssumeDSOLocal special case. Rewrite preemption.ll to dsolocal-static.ll and dsolocal-pic.ll, and add "PIC Level" metadata.
2020-12-30 10:32:34 -08:00
define dso_local void @t7(i8 %v) nounwind {
NFC: Use -LABEL more There were a number of tests needing updates for D91734, and I added a bunch of LABEL directives to help track down where those had to go. These directives are an improvement independent of the functional patch, so I'm committing them as their own separate patch.
2021-01-11 07:03:08 -08:00
; ELF64-LABEL: t7:
[PowerPC] Add loads, stores, and related things to fast-isel. This is the next big chunk of fast-isel code. The primary purpose is to implement selection of loads and stores, but there is a lot of drag-along to support this. The common code to analyze addresses for both loads and stores is substantial. It's also necessary to add the materialization code for global values. Related to load-store processing is the code to fold loads into integer extends, since otherwise we generate lots of redundant instructions. We also need to add some overrides to some FastEmit routines to ensure we don't assign GPR 0 to a virtual register when this would change the meaning of an instruction. I added handling selection of a few binary arithmetic instructions, to enable committing some test cases I wrote a while back. Finally, ap couple of miscellaneous changes: * I cleaned up some poor style from a previous patch in PPCISelLowering.cpp, pointed out by David Blaikie. * I enlarged the Addr.Offset field to avoid sign problems with 32-bit offsets. llvm-svn: 189636
2013-08-30 02:29:45 +00:00
%1 = add nsw i8 %v, 1
store i8 %1, i8* @a, align 1
; ELF64: addi
[FastISel] Flush local value map on every instruction Local values are constants or addresses that can't be folded into the instruction that uses them. FastISel materializes these in a "local value" area that always dominates the current insertion point, to try to avoid materializing these values more than once (per block). https://reviews.llvm.org/D43093 added code to sink these local value instructions to their first use, which has two beneficial effects. One, it is likely to avoid some unnecessary spills and reloads; two, it allows us to attach the debug location of the user to the local value instruction. The latter effect can improve the debugging experience for debuggers with a "set next statement" feature, such as the Visual Studio debugger and PS4 debugger, because instructions to set up constants for a given statement will be associated with the appropriate source line. There are also some constants (primarily addresses) that could be produced by no-op casts or GEP instructions; the main difference from "local value" instructions is that these are values from separate IR instructions, and therefore could have multiple users across multiple basic blocks. D43093 avoided sinking these, even though they were emitted to the same "local value" area as the other instructions. The patch comment for D43093 states: Local values may also be used by no-op casts, which adds the register to the RegFixups table. Without reversing the RegFixups map direction, we don't have enough information to sink these instructions. This patch undoes most of D43093, and instead flushes the local value map after(*) every IR instruction, using that instruction's debug location. This avoids sometimes incorrect locations used previously, and emits instructions in a more natural order. In addition, constants materialized due to PHI instructions are not assigned a debug location immediately; instead, when the local value map is flushed, if the first local value instruction has no debug location, it is given the same location as the first non-local-value-map instruction. This prevents PHIs from introducing unattributed instructions, which would either be implicitly attributed to the location for the preceding IR instruction, or given line 0 if they are at the beginning of a machine basic block. Neither of those consequences is good for debugging. This does mean materialized values are not re-used across IR instruction boundaries; however, only about 5% of those values were reused in an experimental self-build of clang. (*) Actually, just prior to the next instruction. It seems like it would be cleaner the other way, but I was having trouble getting that to work. This reapplies commits cf1c774d and dc35368c, and adds the modification to PHI handling, which should avoid problems with debugging under gdb. Differential Revision: https://reviews.llvm.org/D91734
2021-01-11 08:32:36 -08:00
; ELF64: addis
Revert "[FastISel] Flush local value map on ever instruction" and dependent patches This reverts commit cf1c774d6ace59c5adc9ab71b31e762c1be695b1. This change caused several regressions in the gdb test suite - at least a sample of which was due to line zero instructions making breakpoints un-lined. I think they're worth investigating/understanding more (& possibly addressing) before moving forward with this change. Revert "[FastISel] NFC: Clean up unnecessary bookkeeping" This reverts commit 3fd39d3694d32efa44242c099e923a7f4d982095. Revert "[FastISel] NFC: Remove obsolete -fast-isel-sink-local-values option" This reverts commit a474657e30edccd9e175d92bddeefcfa544751b2. Revert "Remove static function unused after cf1c774." This reverts commit dc35368ccf17a7dca0874ace7490cc3836fb063f. Revert "[lldb] Fix TestThreadStepOut.py after "Flush local value map on every instruction"" This reverts commit 53a14a47ee89dadb8798ca8ed19848f33f4551d5.
2020-12-01 13:23:30 -08:00
; ELF64: addi
[PowerPC] Add loads, stores, and related things to fast-isel. This is the next big chunk of fast-isel code. The primary purpose is to implement selection of loads and stores, but there is a lot of drag-along to support this. The common code to analyze addresses for both loads and stores is substantial. It's also necessary to add the materialization code for global values. Related to load-store processing is the code to fold loads into integer extends, since otherwise we generate lots of redundant instructions. We also need to add some overrides to some FastEmit routines to ensure we don't assign GPR 0 to a virtual register when this would change the meaning of an instruction. I added handling selection of a few binary arithmetic instructions, to enable committing some test cases I wrote a while back. Finally, ap couple of miscellaneous changes: * I cleaned up some poor style from a previous patch in PPCISelLowering.cpp, pointed out by David Blaikie. * I enlarged the Addr.Offset field to avoid sign problems with 32-bit offsets. llvm-svn: 189636
2013-08-30 02:29:45 +00:00
; ELF64: stb
ret void
}
[PowerPC][test] Add explicit dso_local to definitions in ELF static relocation model tests TargetMachine::shouldAssumeDSOLocal currently implies dso_local for such definitions. Adding explicit dso_local makes these tests align with the clang -fpic behavior and allow the removal of the TargetMachine::shouldAssumeDSOLocal special case. Rewrite preemption.ll to dsolocal-static.ll and dsolocal-pic.ll, and add "PIC Level" metadata.
2020-12-30 10:32:34 -08:00
define dso_local void @t8(i16 %v) nounwind {
NFC: Use -LABEL more There were a number of tests needing updates for D91734, and I added a bunch of LABEL directives to help track down where those had to go. These directives are an improvement independent of the functional patch, so I'm committing them as their own separate patch.
2021-01-11 07:03:08 -08:00
; ELF64-LABEL: t8:
[PowerPC] Add loads, stores, and related things to fast-isel. This is the next big chunk of fast-isel code. The primary purpose is to implement selection of loads and stores, but there is a lot of drag-along to support this. The common code to analyze addresses for both loads and stores is substantial. It's also necessary to add the materialization code for global values. Related to load-store processing is the code to fold loads into integer extends, since otherwise we generate lots of redundant instructions. We also need to add some overrides to some FastEmit routines to ensure we don't assign GPR 0 to a virtual register when this would change the meaning of an instruction. I added handling selection of a few binary arithmetic instructions, to enable committing some test cases I wrote a while back. Finally, ap couple of miscellaneous changes: * I cleaned up some poor style from a previous patch in PPCISelLowering.cpp, pointed out by David Blaikie. * I enlarged the Addr.Offset field to avoid sign problems with 32-bit offsets. llvm-svn: 189636
2013-08-30 02:29:45 +00:00
%1 = add nsw i16 %v, 1
store i16 %1, i16* @b, align 2
; ELF64: addi
[FastISel] Flush local value map on every instruction Local values are constants or addresses that can't be folded into the instruction that uses them. FastISel materializes these in a "local value" area that always dominates the current insertion point, to try to avoid materializing these values more than once (per block). https://reviews.llvm.org/D43093 added code to sink these local value instructions to their first use, which has two beneficial effects. One, it is likely to avoid some unnecessary spills and reloads; two, it allows us to attach the debug location of the user to the local value instruction. The latter effect can improve the debugging experience for debuggers with a "set next statement" feature, such as the Visual Studio debugger and PS4 debugger, because instructions to set up constants for a given statement will be associated with the appropriate source line. There are also some constants (primarily addresses) that could be produced by no-op casts or GEP instructions; the main difference from "local value" instructions is that these are values from separate IR instructions, and therefore could have multiple users across multiple basic blocks. D43093 avoided sinking these, even though they were emitted to the same "local value" area as the other instructions. The patch comment for D43093 states: Local values may also be used by no-op casts, which adds the register to the RegFixups table. Without reversing the RegFixups map direction, we don't have enough information to sink these instructions. This patch undoes most of D43093, and instead flushes the local value map after(*) every IR instruction, using that instruction's debug location. This avoids sometimes incorrect locations used previously, and emits instructions in a more natural order. In addition, constants materialized due to PHI instructions are not assigned a debug location immediately; instead, when the local value map is flushed, if the first local value instruction has no debug location, it is given the same location as the first non-local-value-map instruction. This prevents PHIs from introducing unattributed instructions, which would either be implicitly attributed to the location for the preceding IR instruction, or given line 0 if they are at the beginning of a machine basic block. Neither of those consequences is good for debugging. This does mean materialized values are not re-used across IR instruction boundaries; however, only about 5% of those values were reused in an experimental self-build of clang. (*) Actually, just prior to the next instruction. It seems like it would be cleaner the other way, but I was having trouble getting that to work. This reapplies commits cf1c774d and dc35368c, and adds the modification to PHI handling, which should avoid problems with debugging under gdb. Differential Revision: https://reviews.llvm.org/D91734
2021-01-11 08:32:36 -08:00
; ELF64: addis
Revert "[FastISel] Flush local value map on ever instruction" and dependent patches This reverts commit cf1c774d6ace59c5adc9ab71b31e762c1be695b1. This change caused several regressions in the gdb test suite - at least a sample of which was due to line zero instructions making breakpoints un-lined. I think they're worth investigating/understanding more (& possibly addressing) before moving forward with this change. Revert "[FastISel] NFC: Clean up unnecessary bookkeeping" This reverts commit 3fd39d3694d32efa44242c099e923a7f4d982095. Revert "[FastISel] NFC: Remove obsolete -fast-isel-sink-local-values option" This reverts commit a474657e30edccd9e175d92bddeefcfa544751b2. Revert "Remove static function unused after cf1c774." This reverts commit dc35368ccf17a7dca0874ace7490cc3836fb063f. Revert "[lldb] Fix TestThreadStepOut.py after "Flush local value map on every instruction"" This reverts commit 53a14a47ee89dadb8798ca8ed19848f33f4551d5.
2020-12-01 13:23:30 -08:00
; ELF64: addi
[PowerPC] Add loads, stores, and related things to fast-isel. This is the next big chunk of fast-isel code. The primary purpose is to implement selection of loads and stores, but there is a lot of drag-along to support this. The common code to analyze addresses for both loads and stores is substantial. It's also necessary to add the materialization code for global values. Related to load-store processing is the code to fold loads into integer extends, since otherwise we generate lots of redundant instructions. We also need to add some overrides to some FastEmit routines to ensure we don't assign GPR 0 to a virtual register when this would change the meaning of an instruction. I added handling selection of a few binary arithmetic instructions, to enable committing some test cases I wrote a while back. Finally, ap couple of miscellaneous changes: * I cleaned up some poor style from a previous patch in PPCISelLowering.cpp, pointed out by David Blaikie. * I enlarged the Addr.Offset field to avoid sign problems with 32-bit offsets. llvm-svn: 189636
2013-08-30 02:29:45 +00:00
; ELF64: sth
ret void
}
[PowerPC][test] Add explicit dso_local to definitions in ELF static relocation model tests TargetMachine::shouldAssumeDSOLocal currently implies dso_local for such definitions. Adding explicit dso_local makes these tests align with the clang -fpic behavior and allow the removal of the TargetMachine::shouldAssumeDSOLocal special case. Rewrite preemption.ll to dsolocal-static.ll and dsolocal-pic.ll, and add "PIC Level" metadata.
2020-12-30 10:32:34 -08:00
define dso_local void @t9(i32 %v) nounwind {
NFC: Use -LABEL more There were a number of tests needing updates for D91734, and I added a bunch of LABEL directives to help track down where those had to go. These directives are an improvement independent of the functional patch, so I'm committing them as their own separate patch.
2021-01-11 07:03:08 -08:00
; ELF64-LABEL: t9:
[PowerPC] Add loads, stores, and related things to fast-isel. This is the next big chunk of fast-isel code. The primary purpose is to implement selection of loads and stores, but there is a lot of drag-along to support this. The common code to analyze addresses for both loads and stores is substantial. It's also necessary to add the materialization code for global values. Related to load-store processing is the code to fold loads into integer extends, since otherwise we generate lots of redundant instructions. We also need to add some overrides to some FastEmit routines to ensure we don't assign GPR 0 to a virtual register when this would change the meaning of an instruction. I added handling selection of a few binary arithmetic instructions, to enable committing some test cases I wrote a while back. Finally, ap couple of miscellaneous changes: * I cleaned up some poor style from a previous patch in PPCISelLowering.cpp, pointed out by David Blaikie. * I enlarged the Addr.Offset field to avoid sign problems with 32-bit offsets. llvm-svn: 189636
2013-08-30 02:29:45 +00:00
%1 = add nsw i32 %v, 1
store i32 %1, i32* @c, align 4
; ELF64: addi
[FastISel] Flush local value map on every instruction Local values are constants or addresses that can't be folded into the instruction that uses them. FastISel materializes these in a "local value" area that always dominates the current insertion point, to try to avoid materializing these values more than once (per block). https://reviews.llvm.org/D43093 added code to sink these local value instructions to their first use, which has two beneficial effects. One, it is likely to avoid some unnecessary spills and reloads; two, it allows us to attach the debug location of the user to the local value instruction. The latter effect can improve the debugging experience for debuggers with a "set next statement" feature, such as the Visual Studio debugger and PS4 debugger, because instructions to set up constants for a given statement will be associated with the appropriate source line. There are also some constants (primarily addresses) that could be produced by no-op casts or GEP instructions; the main difference from "local value" instructions is that these are values from separate IR instructions, and therefore could have multiple users across multiple basic blocks. D43093 avoided sinking these, even though they were emitted to the same "local value" area as the other instructions. The patch comment for D43093 states: Local values may also be used by no-op casts, which adds the register to the RegFixups table. Without reversing the RegFixups map direction, we don't have enough information to sink these instructions. This patch undoes most of D43093, and instead flushes the local value map after(*) every IR instruction, using that instruction's debug location. This avoids sometimes incorrect locations used previously, and emits instructions in a more natural order. In addition, constants materialized due to PHI instructions are not assigned a debug location immediately; instead, when the local value map is flushed, if the first local value instruction has no debug location, it is given the same location as the first non-local-value-map instruction. This prevents PHIs from introducing unattributed instructions, which would either be implicitly attributed to the location for the preceding IR instruction, or given line 0 if they are at the beginning of a machine basic block. Neither of those consequences is good for debugging. This does mean materialized values are not re-used across IR instruction boundaries; however, only about 5% of those values were reused in an experimental self-build of clang. (*) Actually, just prior to the next instruction. It seems like it would be cleaner the other way, but I was having trouble getting that to work. This reapplies commits cf1c774d and dc35368c, and adds the modification to PHI handling, which should avoid problems with debugging under gdb. Differential Revision: https://reviews.llvm.org/D91734
2021-01-11 08:32:36 -08:00
; ELF64: addis
Revert "[FastISel] Flush local value map on ever instruction" and dependent patches This reverts commit cf1c774d6ace59c5adc9ab71b31e762c1be695b1. This change caused several regressions in the gdb test suite - at least a sample of which was due to line zero instructions making breakpoints un-lined. I think they're worth investigating/understanding more (& possibly addressing) before moving forward with this change. Revert "[FastISel] NFC: Clean up unnecessary bookkeeping" This reverts commit 3fd39d3694d32efa44242c099e923a7f4d982095. Revert "[FastISel] NFC: Remove obsolete -fast-isel-sink-local-values option" This reverts commit a474657e30edccd9e175d92bddeefcfa544751b2. Revert "Remove static function unused after cf1c774." This reverts commit dc35368ccf17a7dca0874ace7490cc3836fb063f. Revert "[lldb] Fix TestThreadStepOut.py after "Flush local value map on every instruction"" This reverts commit 53a14a47ee89dadb8798ca8ed19848f33f4551d5.
2020-12-01 13:23:30 -08:00
; ELF64: addi
[PowerPC] Add loads, stores, and related things to fast-isel. This is the next big chunk of fast-isel code. The primary purpose is to implement selection of loads and stores, but there is a lot of drag-along to support this. The common code to analyze addresses for both loads and stores is substantial. It's also necessary to add the materialization code for global values. Related to load-store processing is the code to fold loads into integer extends, since otherwise we generate lots of redundant instructions. We also need to add some overrides to some FastEmit routines to ensure we don't assign GPR 0 to a virtual register when this would change the meaning of an instruction. I added handling selection of a few binary arithmetic instructions, to enable committing some test cases I wrote a while back. Finally, ap couple of miscellaneous changes: * I cleaned up some poor style from a previous patch in PPCISelLowering.cpp, pointed out by David Blaikie. * I enlarged the Addr.Offset field to avoid sign problems with 32-bit offsets. llvm-svn: 189636
2013-08-30 02:29:45 +00:00
; ELF64: stw
ret void
}
[PowerPC][test] Add explicit dso_local to definitions in ELF static relocation model tests TargetMachine::shouldAssumeDSOLocal currently implies dso_local for such definitions. Adding explicit dso_local makes these tests align with the clang -fpic behavior and allow the removal of the TargetMachine::shouldAssumeDSOLocal special case. Rewrite preemption.ll to dsolocal-static.ll and dsolocal-pic.ll, and add "PIC Level" metadata.
2020-12-30 10:32:34 -08:00
define dso_local void @t10(i64 %v) nounwind {
NFC: Use -LABEL more There were a number of tests needing updates for D91734, and I added a bunch of LABEL directives to help track down where those had to go. These directives are an improvement independent of the functional patch, so I'm committing them as their own separate patch.
2021-01-11 07:03:08 -08:00
; ELF64-LABEL: t10:
[PowerPC] Add loads, stores, and related things to fast-isel. This is the next big chunk of fast-isel code. The primary purpose is to implement selection of loads and stores, but there is a lot of drag-along to support this. The common code to analyze addresses for both loads and stores is substantial. It's also necessary to add the materialization code for global values. Related to load-store processing is the code to fold loads into integer extends, since otherwise we generate lots of redundant instructions. We also need to add some overrides to some FastEmit routines to ensure we don't assign GPR 0 to a virtual register when this would change the meaning of an instruction. I added handling selection of a few binary arithmetic instructions, to enable committing some test cases I wrote a while back. Finally, ap couple of miscellaneous changes: * I cleaned up some poor style from a previous patch in PPCISelLowering.cpp, pointed out by David Blaikie. * I enlarged the Addr.Offset field to avoid sign problems with 32-bit offsets. llvm-svn: 189636
2013-08-30 02:29:45 +00:00
%1 = add nsw i64 %v, 1
store i64 %1, i64* @d, align 4
; ELF64: addi
[FastISel] Flush local value map on every instruction Local values are constants or addresses that can't be folded into the instruction that uses them. FastISel materializes these in a "local value" area that always dominates the current insertion point, to try to avoid materializing these values more than once (per block). https://reviews.llvm.org/D43093 added code to sink these local value instructions to their first use, which has two beneficial effects. One, it is likely to avoid some unnecessary spills and reloads; two, it allows us to attach the debug location of the user to the local value instruction. The latter effect can improve the debugging experience for debuggers with a "set next statement" feature, such as the Visual Studio debugger and PS4 debugger, because instructions to set up constants for a given statement will be associated with the appropriate source line. There are also some constants (primarily addresses) that could be produced by no-op casts or GEP instructions; the main difference from "local value" instructions is that these are values from separate IR instructions, and therefore could have multiple users across multiple basic blocks. D43093 avoided sinking these, even though they were emitted to the same "local value" area as the other instructions. The patch comment for D43093 states: Local values may also be used by no-op casts, which adds the register to the RegFixups table. Without reversing the RegFixups map direction, we don't have enough information to sink these instructions. This patch undoes most of D43093, and instead flushes the local value map after(*) every IR instruction, using that instruction's debug location. This avoids sometimes incorrect locations used previously, and emits instructions in a more natural order. In addition, constants materialized due to PHI instructions are not assigned a debug location immediately; instead, when the local value map is flushed, if the first local value instruction has no debug location, it is given the same location as the first non-local-value-map instruction. This prevents PHIs from introducing unattributed instructions, which would either be implicitly attributed to the location for the preceding IR instruction, or given line 0 if they are at the beginning of a machine basic block. Neither of those consequences is good for debugging. This does mean materialized values are not re-used across IR instruction boundaries; however, only about 5% of those values were reused in an experimental self-build of clang. (*) Actually, just prior to the next instruction. It seems like it would be cleaner the other way, but I was having trouble getting that to work. This reapplies commits cf1c774d and dc35368c, and adds the modification to PHI handling, which should avoid problems with debugging under gdb. Differential Revision: https://reviews.llvm.org/D91734
2021-01-11 08:32:36 -08:00
; ELF64: addis
Revert "[FastISel] Flush local value map on ever instruction" and dependent patches This reverts commit cf1c774d6ace59c5adc9ab71b31e762c1be695b1. This change caused several regressions in the gdb test suite - at least a sample of which was due to line zero instructions making breakpoints un-lined. I think they're worth investigating/understanding more (& possibly addressing) before moving forward with this change. Revert "[FastISel] NFC: Clean up unnecessary bookkeeping" This reverts commit 3fd39d3694d32efa44242c099e923a7f4d982095. Revert "[FastISel] NFC: Remove obsolete -fast-isel-sink-local-values option" This reverts commit a474657e30edccd9e175d92bddeefcfa544751b2. Revert "Remove static function unused after cf1c774." This reverts commit dc35368ccf17a7dca0874ace7490cc3836fb063f. Revert "[lldb] Fix TestThreadStepOut.py after "Flush local value map on every instruction"" This reverts commit 53a14a47ee89dadb8798ca8ed19848f33f4551d5.
2020-12-01 13:23:30 -08:00
; ELF64: addi
[PowerPC] Add loads, stores, and related things to fast-isel. This is the next big chunk of fast-isel code. The primary purpose is to implement selection of loads and stores, but there is a lot of drag-along to support this. The common code to analyze addresses for both loads and stores is substantial. It's also necessary to add the materialization code for global values. Related to load-store processing is the code to fold loads into integer extends, since otherwise we generate lots of redundant instructions. We also need to add some overrides to some FastEmit routines to ensure we don't assign GPR 0 to a virtual register when this would change the meaning of an instruction. I added handling selection of a few binary arithmetic instructions, to enable committing some test cases I wrote a while back. Finally, ap couple of miscellaneous changes: * I cleaned up some poor style from a previous patch in PPCISelLowering.cpp, pointed out by David Blaikie. * I enlarged the Addr.Offset field to avoid sign problems with 32-bit offsets. llvm-svn: 189636
2013-08-30 02:29:45 +00:00
; ELF64: std
ret void
}
[PowerPC][test] Add explicit dso_local to definitions in ELF static relocation model tests TargetMachine::shouldAssumeDSOLocal currently implies dso_local for such definitions. Adding explicit dso_local makes these tests align with the clang -fpic behavior and allow the removal of the TargetMachine::shouldAssumeDSOLocal special case. Rewrite preemption.ll to dsolocal-static.ll and dsolocal-pic.ll, and add "PIC Level" metadata.
2020-12-30 10:32:34 -08:00
define dso_local void @t11(float %v) nounwind {
NFC: Use -LABEL more There were a number of tests needing updates for D91734, and I added a bunch of LABEL directives to help track down where those had to go. These directives are an improvement independent of the functional patch, so I'm committing them as their own separate patch.
2021-01-11 07:03:08 -08:00
; ELF64-LABEL: t11:
; SPE-LABEL: t11:
[PowerPC] Add loads, stores, and related things to fast-isel. This is the next big chunk of fast-isel code. The primary purpose is to implement selection of loads and stores, but there is a lot of drag-along to support this. The common code to analyze addresses for both loads and stores is substantial. It's also necessary to add the materialization code for global values. Related to load-store processing is the code to fold loads into integer extends, since otherwise we generate lots of redundant instructions. We also need to add some overrides to some FastEmit routines to ensure we don't assign GPR 0 to a virtual register when this would change the meaning of an instruction. I added handling selection of a few binary arithmetic instructions, to enable committing some test cases I wrote a while back. Finally, ap couple of miscellaneous changes: * I cleaned up some poor style from a previous patch in PPCISelLowering.cpp, pointed out by David Blaikie. * I enlarged the Addr.Offset field to avoid sign problems with 32-bit offsets. llvm-svn: 189636
2013-08-30 02:29:45 +00:00
%1 = fadd float %v, 1.0
store float %1, float* @e, align 4
; ELF64: fadds
; ELF64: stfs
Introduce codegen for the Signal Processing Engine Summary: The Signal Processing Engine (SPE) is found on NXP/Freescale e500v1, e500v2, and several e200 cores. This adds support targeting the e500v2, as this is more common than the e500v1, and is in SoCs still on the market. This patch is very intrusive because the SPE is binary incompatible with the traditional FPU. After discussing with others, the cleanest solution was to make both SPE and FPU features on top of a base PowerPC subset, so all FPU instructions are now wrapped with HasFPU predicates. Supported by this are: * Code generation following the SPE ABI at the LLVM IR level (calling conventions) * Single- and Double-precision math at the level supported by the APU. Still to do: * Vector operations * SPE intrinsics As this changes the Callee-saved register list order, one test, which tests the precise generated code, was updated to account for the new register order. Reviewed by: nemanjai Differential Revision: https://reviews.llvm.org/D44830 llvm-svn: 337347
2018-07-18 04:25:10 +00:00
; SPE: efsadd
; SPE: stw
[PowerPC] Add loads, stores, and related things to fast-isel. This is the next big chunk of fast-isel code. The primary purpose is to implement selection of loads and stores, but there is a lot of drag-along to support this. The common code to analyze addresses for both loads and stores is substantial. It's also necessary to add the materialization code for global values. Related to load-store processing is the code to fold loads into integer extends, since otherwise we generate lots of redundant instructions. We also need to add some overrides to some FastEmit routines to ensure we don't assign GPR 0 to a virtual register when this would change the meaning of an instruction. I added handling selection of a few binary arithmetic instructions, to enable committing some test cases I wrote a while back. Finally, ap couple of miscellaneous changes: * I cleaned up some poor style from a previous patch in PPCISelLowering.cpp, pointed out by David Blaikie. * I enlarged the Addr.Offset field to avoid sign problems with 32-bit offsets. llvm-svn: 189636
2013-08-30 02:29:45 +00:00
ret void
}
[PowerPC][test] Add explicit dso_local to definitions in ELF static relocation model tests TargetMachine::shouldAssumeDSOLocal currently implies dso_local for such definitions. Adding explicit dso_local makes these tests align with the clang -fpic behavior and allow the removal of the TargetMachine::shouldAssumeDSOLocal special case. Rewrite preemption.ll to dsolocal-static.ll and dsolocal-pic.ll, and add "PIC Level" metadata.
2020-12-30 10:32:34 -08:00
define dso_local void @t12(double %v) nounwind {
NFC: Use -LABEL more There were a number of tests needing updates for D91734, and I added a bunch of LABEL directives to help track down where those had to go. These directives are an improvement independent of the functional patch, so I'm committing them as their own separate patch.
2021-01-11 07:03:08 -08:00
; ELF64-LABEL: t12:
; SPE-LABEL: t12:
[PowerPC] Add loads, stores, and related things to fast-isel. This is the next big chunk of fast-isel code. The primary purpose is to implement selection of loads and stores, but there is a lot of drag-along to support this. The common code to analyze addresses for both loads and stores is substantial. It's also necessary to add the materialization code for global values. Related to load-store processing is the code to fold loads into integer extends, since otherwise we generate lots of redundant instructions. We also need to add some overrides to some FastEmit routines to ensure we don't assign GPR 0 to a virtual register when this would change the meaning of an instruction. I added handling selection of a few binary arithmetic instructions, to enable committing some test cases I wrote a while back. Finally, ap couple of miscellaneous changes: * I cleaned up some poor style from a previous patch in PPCISelLowering.cpp, pointed out by David Blaikie. * I enlarged the Addr.Offset field to avoid sign problems with 32-bit offsets. llvm-svn: 189636
2013-08-30 02:29:45 +00:00
%1 = fadd double %v, 1.0
store double %1, double* @f, align 8
; ELF64: fadd
; ELF64: stfd
[NFC][PowerPC] Fast-isel VSX support test We have fixed most of the VSX limitation in Fast-isel, so we can remove the -mattr=-vsx for most testcases now. llvm-svn: 372345
2019-09-19 18:18:18 +00:00
; VSX: xsadddp
; VSX: stxsdx
Introduce codegen for the Signal Processing Engine Summary: The Signal Processing Engine (SPE) is found on NXP/Freescale e500v1, e500v2, and several e200 cores. This adds support targeting the e500v2, as this is more common than the e500v1, and is in SoCs still on the market. This patch is very intrusive because the SPE is binary incompatible with the traditional FPU. After discussing with others, the cleanest solution was to make both SPE and FPU features on top of a base PowerPC subset, so all FPU instructions are now wrapped with HasFPU predicates. Supported by this are: * Code generation following the SPE ABI at the LLVM IR level (calling conventions) * Single- and Double-precision math at the level supported by the APU. Still to do: * Vector operations * SPE intrinsics As this changes the Callee-saved register list order, one test, which tests the precise generated code, was updated to account for the new register order. Reviewed by: nemanjai Differential Revision: https://reviews.llvm.org/D44830 llvm-svn: 337347
2018-07-18 04:25:10 +00:00
; SPE: efdadd
; SPE: evstdd
[PowerPC] Add loads, stores, and related things to fast-isel. This is the next big chunk of fast-isel code. The primary purpose is to implement selection of loads and stores, but there is a lot of drag-along to support this. The common code to analyze addresses for both loads and stores is substantial. It's also necessary to add the materialization code for global values. Related to load-store processing is the code to fold loads into integer extends, since otherwise we generate lots of redundant instructions. We also need to add some overrides to some FastEmit routines to ensure we don't assign GPR 0 to a virtual register when this would change the meaning of an instruction. I added handling selection of a few binary arithmetic instructions, to enable committing some test cases I wrote a while back. Finally, ap couple of miscellaneous changes: * I cleaned up some poor style from a previous patch in PPCISelLowering.cpp, pointed out by David Blaikie. * I enlarged the Addr.Offset field to avoid sign problems with 32-bit offsets. llvm-svn: 189636
2013-08-30 02:29:45 +00:00
ret void
}
;; lwa requires an offset divisible by 4, so we need lwax here.
Clean up function attributes on PPC fast-isel tests. llvm-svn: 243079
2015-07-24 01:07:50 +00:00
define i64 @t13() nounwind {
NFC: Use -LABEL more There were a number of tests needing updates for D91734, and I added a bunch of LABEL directives to help track down where those had to go. These directives are an improvement independent of the functional patch, so I'm committing them as their own separate patch.
2021-01-11 07:03:08 -08:00
; ELF64-LABEL: t13:
[opaque pointer type] Add textual IR support for explicit type parameter to gep operator Similar to gep (r230786) and load (r230794) changes. Similar migration script can be used to update test cases, which successfully migrated all of LLVM and Polly, but about 4 test cases needed manually changes in Clang. (this script will read the contents of stdin and massage it into stdout - wrap it in the 'apply.sh' script shown in previous commits + xargs to apply it over a large set of test cases) import fileinput import sys import re rep = re.compile(r"(getelementptr(?:\s+inbounds)?\s*\()((<\d*\s+x\s+)?([^@]*?)(|\s*addrspace\(\d+\))\s*\*(?(3)>)\s*)(?=$|%|@|null|undef|blockaddress|getelementptr|addrspacecast|bitcast|inttoptr|zeroinitializer|<|\[\[[a-zA-Z]|\{\{)", re.MULTILINE | re.DOTALL) def conv(match): line = match.group(1) line += match.group(4) line += ", " line += match.group(2) return line line = sys.stdin.read() off = 0 for match in re.finditer(rep, line): sys.stdout.write(line[off:match.start()]) sys.stdout.write(conv(match)) off = match.end() sys.stdout.write(line[off:]) llvm-svn: 232184
2015-03-13 18:20:45 +00:00
%1 = load i32, i32* getelementptr inbounds (%struct.s, %struct.s* @g, i32 0, i32 1), align 1
[PowerPC] Add loads, stores, and related things to fast-isel. This is the next big chunk of fast-isel code. The primary purpose is to implement selection of loads and stores, but there is a lot of drag-along to support this. The common code to analyze addresses for both loads and stores is substantial. It's also necessary to add the materialization code for global values. Related to load-store processing is the code to fold loads into integer extends, since otherwise we generate lots of redundant instructions. We also need to add some overrides to some FastEmit routines to ensure we don't assign GPR 0 to a virtual register when this would change the meaning of an instruction. I added handling selection of a few binary arithmetic instructions, to enable committing some test cases I wrote a while back. Finally, ap couple of miscellaneous changes: * I cleaned up some poor style from a previous patch in PPCISelLowering.cpp, pointed out by David Blaikie. * I enlarged the Addr.Offset field to avoid sign problems with 32-bit offsets. llvm-svn: 189636
2013-08-30 02:29:45 +00:00
%2 = sext i32 %1 to i64
; ELF64: li
; ELF64: lwax
%3 = add nsw i64 %2, 1
; ELF64: addi
ret i64 %3
}
;; ld requires an offset divisible by 4, so we need ldx here.
Clean up function attributes on PPC fast-isel tests. llvm-svn: 243079
2015-07-24 01:07:50 +00:00
define i64 @t14() nounwind {
NFC: Use -LABEL more There were a number of tests needing updates for D91734, and I added a bunch of LABEL directives to help track down where those had to go. These directives are an improvement independent of the functional patch, so I'm committing them as their own separate patch.
2021-01-11 07:03:08 -08:00
; ELF64-LABEL: t14:
[opaque pointer type] Add textual IR support for explicit type parameter to gep operator Similar to gep (r230786) and load (r230794) changes. Similar migration script can be used to update test cases, which successfully migrated all of LLVM and Polly, but about 4 test cases needed manually changes in Clang. (this script will read the contents of stdin and massage it into stdout - wrap it in the 'apply.sh' script shown in previous commits + xargs to apply it over a large set of test cases) import fileinput import sys import re rep = re.compile(r"(getelementptr(?:\s+inbounds)?\s*\()((<\d*\s+x\s+)?([^@]*?)(|\s*addrspace\(\d+\))\s*\*(?(3)>)\s*)(?=$|%|@|null|undef|blockaddress|getelementptr|addrspacecast|bitcast|inttoptr|zeroinitializer|<|\[\[[a-zA-Z]|\{\{)", re.MULTILINE | re.DOTALL) def conv(match): line = match.group(1) line += match.group(4) line += ", " line += match.group(2) return line line = sys.stdin.read() off = 0 for match in re.finditer(rep, line): sys.stdout.write(line[off:match.start()]) sys.stdout.write(conv(match)) off = match.end() sys.stdout.write(line[off:]) llvm-svn: 232184
2015-03-13 18:20:45 +00:00
%1 = load i64, i64* getelementptr inbounds (%struct.t, %struct.t* @h, i32 0, i32 1), align 1
[PowerPC] Add loads, stores, and related things to fast-isel. This is the next big chunk of fast-isel code. The primary purpose is to implement selection of loads and stores, but there is a lot of drag-along to support this. The common code to analyze addresses for both loads and stores is substantial. It's also necessary to add the materialization code for global values. Related to load-store processing is the code to fold loads into integer extends, since otherwise we generate lots of redundant instructions. We also need to add some overrides to some FastEmit routines to ensure we don't assign GPR 0 to a virtual register when this would change the meaning of an instruction. I added handling selection of a few binary arithmetic instructions, to enable committing some test cases I wrote a while back. Finally, ap couple of miscellaneous changes: * I cleaned up some poor style from a previous patch in PPCISelLowering.cpp, pointed out by David Blaikie. * I enlarged the Addr.Offset field to avoid sign problems with 32-bit offsets. llvm-svn: 189636
2013-08-30 02:29:45 +00:00
; ELF64: li
; ELF64: ldx
%2 = add nsw i64 %1, 1
; ELF64: addi
ret i64 %2
}
;; std requires an offset divisible by 4, so we need stdx here.
[PowerPC][test] Add explicit dso_local to definitions in ELF static relocation model tests TargetMachine::shouldAssumeDSOLocal currently implies dso_local for such definitions. Adding explicit dso_local makes these tests align with the clang -fpic behavior and allow the removal of the TargetMachine::shouldAssumeDSOLocal special case. Rewrite preemption.ll to dsolocal-static.ll and dsolocal-pic.ll, and add "PIC Level" metadata.
2020-12-30 10:32:34 -08:00
define dso_local void @t15(i64 %v) nounwind {
NFC: Use -LABEL more There were a number of tests needing updates for D91734, and I added a bunch of LABEL directives to help track down where those had to go. These directives are an improvement independent of the functional patch, so I'm committing them as their own separate patch.
2021-01-11 07:03:08 -08:00
; ELF64-LABEL: t15:
[PowerPC] Add loads, stores, and related things to fast-isel. This is the next big chunk of fast-isel code. The primary purpose is to implement selection of loads and stores, but there is a lot of drag-along to support this. The common code to analyze addresses for both loads and stores is substantial. It's also necessary to add the materialization code for global values. Related to load-store processing is the code to fold loads into integer extends, since otherwise we generate lots of redundant instructions. We also need to add some overrides to some FastEmit routines to ensure we don't assign GPR 0 to a virtual register when this would change the meaning of an instruction. I added handling selection of a few binary arithmetic instructions, to enable committing some test cases I wrote a while back. Finally, ap couple of miscellaneous changes: * I cleaned up some poor style from a previous patch in PPCISelLowering.cpp, pointed out by David Blaikie. * I enlarged the Addr.Offset field to avoid sign problems with 32-bit offsets. llvm-svn: 189636
2013-08-30 02:29:45 +00:00
%1 = add nsw i64 %v, 1
[opaque pointer type] Add textual IR support for explicit type parameter to gep operator Similar to gep (r230786) and load (r230794) changes. Similar migration script can be used to update test cases, which successfully migrated all of LLVM and Polly, but about 4 test cases needed manually changes in Clang. (this script will read the contents of stdin and massage it into stdout - wrap it in the 'apply.sh' script shown in previous commits + xargs to apply it over a large set of test cases) import fileinput import sys import re rep = re.compile(r"(getelementptr(?:\s+inbounds)?\s*\()((<\d*\s+x\s+)?([^@]*?)(|\s*addrspace\(\d+\))\s*\*(?(3)>)\s*)(?=$|%|@|null|undef|blockaddress|getelementptr|addrspacecast|bitcast|inttoptr|zeroinitializer|<|\[\[[a-zA-Z]|\{\{)", re.MULTILINE | re.DOTALL) def conv(match): line = match.group(1) line += match.group(4) line += ", " line += match.group(2) return line line = sys.stdin.read() off = 0 for match in re.finditer(rep, line): sys.stdout.write(line[off:match.start()]) sys.stdout.write(conv(match)) off = match.end() sys.stdout.write(line[off:]) llvm-svn: 232184
2015-03-13 18:20:45 +00:00
store i64 %1, i64* getelementptr inbounds (%struct.t, %struct.t* @h, i32 0, i32 1), align 1
[PowerPC] Add loads, stores, and related things to fast-isel. This is the next big chunk of fast-isel code. The primary purpose is to implement selection of loads and stores, but there is a lot of drag-along to support this. The common code to analyze addresses for both loads and stores is substantial. It's also necessary to add the materialization code for global values. Related to load-store processing is the code to fold loads into integer extends, since otherwise we generate lots of redundant instructions. We also need to add some overrides to some FastEmit routines to ensure we don't assign GPR 0 to a virtual register when this would change the meaning of an instruction. I added handling selection of a few binary arithmetic instructions, to enable committing some test cases I wrote a while back. Finally, ap couple of miscellaneous changes: * I cleaned up some poor style from a previous patch in PPCISelLowering.cpp, pointed out by David Blaikie. * I enlarged the Addr.Offset field to avoid sign problems with 32-bit offsets. llvm-svn: 189636
2013-08-30 02:29:45 +00:00
; ELF64: addi
[FastISel] Flush local value map on every instruction Local values are constants or addresses that can't be folded into the instruction that uses them. FastISel materializes these in a "local value" area that always dominates the current insertion point, to try to avoid materializing these values more than once (per block). https://reviews.llvm.org/D43093 added code to sink these local value instructions to their first use, which has two beneficial effects. One, it is likely to avoid some unnecessary spills and reloads; two, it allows us to attach the debug location of the user to the local value instruction. The latter effect can improve the debugging experience for debuggers with a "set next statement" feature, such as the Visual Studio debugger and PS4 debugger, because instructions to set up constants for a given statement will be associated with the appropriate source line. There are also some constants (primarily addresses) that could be produced by no-op casts or GEP instructions; the main difference from "local value" instructions is that these are values from separate IR instructions, and therefore could have multiple users across multiple basic blocks. D43093 avoided sinking these, even though they were emitted to the same "local value" area as the other instructions. The patch comment for D43093 states: Local values may also be used by no-op casts, which adds the register to the RegFixups table. Without reversing the RegFixups map direction, we don't have enough information to sink these instructions. This patch undoes most of D43093, and instead flushes the local value map after(*) every IR instruction, using that instruction's debug location. This avoids sometimes incorrect locations used previously, and emits instructions in a more natural order. In addition, constants materialized due to PHI instructions are not assigned a debug location immediately; instead, when the local value map is flushed, if the first local value instruction has no debug location, it is given the same location as the first non-local-value-map instruction. This prevents PHIs from introducing unattributed instructions, which would either be implicitly attributed to the location for the preceding IR instruction, or given line 0 if they are at the beginning of a machine basic block. Neither of those consequences is good for debugging. This does mean materialized values are not re-used across IR instruction boundaries; however, only about 5% of those values were reused in an experimental self-build of clang. (*) Actually, just prior to the next instruction. It seems like it would be cleaner the other way, but I was having trouble getting that to work. This reapplies commits cf1c774d and dc35368c, and adds the modification to PHI handling, which should avoid problems with debugging under gdb. Differential Revision: https://reviews.llvm.org/D91734
2021-01-11 08:32:36 -08:00
; ELF64: addis
Revert "[FastISel] Flush local value map on ever instruction" and dependent patches This reverts commit cf1c774d6ace59c5adc9ab71b31e762c1be695b1. This change caused several regressions in the gdb test suite - at least a sample of which was due to line zero instructions making breakpoints un-lined. I think they're worth investigating/understanding more (& possibly addressing) before moving forward with this change. Revert "[FastISel] NFC: Clean up unnecessary bookkeeping" This reverts commit 3fd39d3694d32efa44242c099e923a7f4d982095. Revert "[FastISel] NFC: Remove obsolete -fast-isel-sink-local-values option" This reverts commit a474657e30edccd9e175d92bddeefcfa544751b2. Revert "Remove static function unused after cf1c774." This reverts commit dc35368ccf17a7dca0874ace7490cc3836fb063f. Revert "[lldb] Fix TestThreadStepOut.py after "Flush local value map on every instruction"" This reverts commit 53a14a47ee89dadb8798ca8ed19848f33f4551d5.
2020-12-01 13:23:30 -08:00
; ELF64: addi
[PowerPC] Add loads, stores, and related things to fast-isel. This is the next big chunk of fast-isel code. The primary purpose is to implement selection of loads and stores, but there is a lot of drag-along to support this. The common code to analyze addresses for both loads and stores is substantial. It's also necessary to add the materialization code for global values. Related to load-store processing is the code to fold loads into integer extends, since otherwise we generate lots of redundant instructions. We also need to add some overrides to some FastEmit routines to ensure we don't assign GPR 0 to a virtual register when this would change the meaning of an instruction. I added handling selection of a few binary arithmetic instructions, to enable committing some test cases I wrote a while back. Finally, ap couple of miscellaneous changes: * I cleaned up some poor style from a previous patch in PPCISelLowering.cpp, pointed out by David Blaikie. * I enlarged the Addr.Offset field to avoid sign problems with 32-bit offsets. llvm-svn: 189636
2013-08-30 02:29:45 +00:00
; ELF64: li
; ELF64: stdx
ret void
}
;; ld requires an offset that fits in 16 bits, so we need ldx here.
Clean up function attributes on PPC fast-isel tests. llvm-svn: 243079
2015-07-24 01:07:50 +00:00
define i64 @t16() nounwind {
NFC: Use -LABEL more There were a number of tests needing updates for D91734, and I added a bunch of LABEL directives to help track down where those had to go. These directives are an improvement independent of the functional patch, so I'm committing them as their own separate patch.
2021-01-11 07:03:08 -08:00
; ELF64-LABEL: t16:
[opaque pointer type] Add textual IR support for explicit type parameter to gep operator Similar to gep (r230786) and load (r230794) changes. Similar migration script can be used to update test cases, which successfully migrated all of LLVM and Polly, but about 4 test cases needed manually changes in Clang. (this script will read the contents of stdin and massage it into stdout - wrap it in the 'apply.sh' script shown in previous commits + xargs to apply it over a large set of test cases) import fileinput import sys import re rep = re.compile(r"(getelementptr(?:\s+inbounds)?\s*\()((<\d*\s+x\s+)?([^@]*?)(|\s*addrspace\(\d+\))\s*\*(?(3)>)\s*)(?=$|%|@|null|undef|blockaddress|getelementptr|addrspacecast|bitcast|inttoptr|zeroinitializer|<|\[\[[a-zA-Z]|\{\{)", re.MULTILINE | re.DOTALL) def conv(match): line = match.group(1) line += match.group(4) line += ", " line += match.group(2) return line line = sys.stdin.read() off = 0 for match in re.finditer(rep, line): sys.stdout.write(line[off:match.start()]) sys.stdout.write(conv(match)) off = match.end() sys.stdout.write(line[off:]) llvm-svn: 232184
2015-03-13 18:20:45 +00:00
%1 = load i64, i64* getelementptr inbounds ([8192 x i64], [8192 x i64]* @i, i32 0, i64 5000), align 8
[PowerPC] Add loads, stores, and related things to fast-isel. This is the next big chunk of fast-isel code. The primary purpose is to implement selection of loads and stores, but there is a lot of drag-along to support this. The common code to analyze addresses for both loads and stores is substantial. It's also necessary to add the materialization code for global values. Related to load-store processing is the code to fold loads into integer extends, since otherwise we generate lots of redundant instructions. We also need to add some overrides to some FastEmit routines to ensure we don't assign GPR 0 to a virtual register when this would change the meaning of an instruction. I added handling selection of a few binary arithmetic instructions, to enable committing some test cases I wrote a while back. Finally, ap couple of miscellaneous changes: * I cleaned up some poor style from a previous patch in PPCISelLowering.cpp, pointed out by David Blaikie. * I enlarged the Addr.Offset field to avoid sign problems with 32-bit offsets. llvm-svn: 189636
2013-08-30 02:29:45 +00:00
; ELF64: lis
; ELF64: ori
; ELF64: ldx
%2 = add nsw i64 %1, 1
; ELF64: addi
ret i64 %2
}
;; std requires an offset that fits in 16 bits, so we need stdx here.
[PowerPC][test] Add explicit dso_local to definitions in ELF static relocation model tests TargetMachine::shouldAssumeDSOLocal currently implies dso_local for such definitions. Adding explicit dso_local makes these tests align with the clang -fpic behavior and allow the removal of the TargetMachine::shouldAssumeDSOLocal special case. Rewrite preemption.ll to dsolocal-static.ll and dsolocal-pic.ll, and add "PIC Level" metadata.
2020-12-30 10:32:34 -08:00
define dso_local void @t17(i64 %v) nounwind {
NFC: Use -LABEL more There were a number of tests needing updates for D91734, and I added a bunch of LABEL directives to help track down where those had to go. These directives are an improvement independent of the functional patch, so I'm committing them as their own separate patch.
2021-01-11 07:03:08 -08:00
; ELF64-LABEL: t17:
[PowerPC] Add loads, stores, and related things to fast-isel. This is the next big chunk of fast-isel code. The primary purpose is to implement selection of loads and stores, but there is a lot of drag-along to support this. The common code to analyze addresses for both loads and stores is substantial. It's also necessary to add the materialization code for global values. Related to load-store processing is the code to fold loads into integer extends, since otherwise we generate lots of redundant instructions. We also need to add some overrides to some FastEmit routines to ensure we don't assign GPR 0 to a virtual register when this would change the meaning of an instruction. I added handling selection of a few binary arithmetic instructions, to enable committing some test cases I wrote a while back. Finally, ap couple of miscellaneous changes: * I cleaned up some poor style from a previous patch in PPCISelLowering.cpp, pointed out by David Blaikie. * I enlarged the Addr.Offset field to avoid sign problems with 32-bit offsets. llvm-svn: 189636
2013-08-30 02:29:45 +00:00
%1 = add nsw i64 %v, 1
[opaque pointer type] Add textual IR support for explicit type parameter to gep operator Similar to gep (r230786) and load (r230794) changes. Similar migration script can be used to update test cases, which successfully migrated all of LLVM and Polly, but about 4 test cases needed manually changes in Clang. (this script will read the contents of stdin and massage it into stdout - wrap it in the 'apply.sh' script shown in previous commits + xargs to apply it over a large set of test cases) import fileinput import sys import re rep = re.compile(r"(getelementptr(?:\s+inbounds)?\s*\()((<\d*\s+x\s+)?([^@]*?)(|\s*addrspace\(\d+\))\s*\*(?(3)>)\s*)(?=$|%|@|null|undef|blockaddress|getelementptr|addrspacecast|bitcast|inttoptr|zeroinitializer|<|\[\[[a-zA-Z]|\{\{)", re.MULTILINE | re.DOTALL) def conv(match): line = match.group(1) line += match.group(4) line += ", " line += match.group(2) return line line = sys.stdin.read() off = 0 for match in re.finditer(rep, line): sys.stdout.write(line[off:match.start()]) sys.stdout.write(conv(match)) off = match.end() sys.stdout.write(line[off:]) llvm-svn: 232184
2015-03-13 18:20:45 +00:00
store i64 %1, i64* getelementptr inbounds ([8192 x i64], [8192 x i64]* @i, i32 0, i64 5000), align 8
[PowerPC] Add loads, stores, and related things to fast-isel. This is the next big chunk of fast-isel code. The primary purpose is to implement selection of loads and stores, but there is a lot of drag-along to support this. The common code to analyze addresses for both loads and stores is substantial. It's also necessary to add the materialization code for global values. Related to load-store processing is the code to fold loads into integer extends, since otherwise we generate lots of redundant instructions. We also need to add some overrides to some FastEmit routines to ensure we don't assign GPR 0 to a virtual register when this would change the meaning of an instruction. I added handling selection of a few binary arithmetic instructions, to enable committing some test cases I wrote a while back. Finally, ap couple of miscellaneous changes: * I cleaned up some poor style from a previous patch in PPCISelLowering.cpp, pointed out by David Blaikie. * I enlarged the Addr.Offset field to avoid sign problems with 32-bit offsets. llvm-svn: 189636
2013-08-30 02:29:45 +00:00
; ELF64: addi
[FastISel] Flush local value map on every instruction Local values are constants or addresses that can't be folded into the instruction that uses them. FastISel materializes these in a "local value" area that always dominates the current insertion point, to try to avoid materializing these values more than once (per block). https://reviews.llvm.org/D43093 added code to sink these local value instructions to their first use, which has two beneficial effects. One, it is likely to avoid some unnecessary spills and reloads; two, it allows us to attach the debug location of the user to the local value instruction. The latter effect can improve the debugging experience for debuggers with a "set next statement" feature, such as the Visual Studio debugger and PS4 debugger, because instructions to set up constants for a given statement will be associated with the appropriate source line. There are also some constants (primarily addresses) that could be produced by no-op casts or GEP instructions; the main difference from "local value" instructions is that these are values from separate IR instructions, and therefore could have multiple users across multiple basic blocks. D43093 avoided sinking these, even though they were emitted to the same "local value" area as the other instructions. The patch comment for D43093 states: Local values may also be used by no-op casts, which adds the register to the RegFixups table. Without reversing the RegFixups map direction, we don't have enough information to sink these instructions. This patch undoes most of D43093, and instead flushes the local value map after(*) every IR instruction, using that instruction's debug location. This avoids sometimes incorrect locations used previously, and emits instructions in a more natural order. In addition, constants materialized due to PHI instructions are not assigned a debug location immediately; instead, when the local value map is flushed, if the first local value instruction has no debug location, it is given the same location as the first non-local-value-map instruction. This prevents PHIs from introducing unattributed instructions, which would either be implicitly attributed to the location for the preceding IR instruction, or given line 0 if they are at the beginning of a machine basic block. Neither of those consequences is good for debugging. This does mean materialized values are not re-used across IR instruction boundaries; however, only about 5% of those values were reused in an experimental self-build of clang. (*) Actually, just prior to the next instruction. It seems like it would be cleaner the other way, but I was having trouble getting that to work. This reapplies commits cf1c774d and dc35368c, and adds the modification to PHI handling, which should avoid problems with debugging under gdb. Differential Revision: https://reviews.llvm.org/D91734
2021-01-11 08:32:36 -08:00
; ELF64: addis
Revert "[FastISel] Flush local value map on ever instruction" and dependent patches This reverts commit cf1c774d6ace59c5adc9ab71b31e762c1be695b1. This change caused several regressions in the gdb test suite - at least a sample of which was due to line zero instructions making breakpoints un-lined. I think they're worth investigating/understanding more (& possibly addressing) before moving forward with this change. Revert "[FastISel] NFC: Clean up unnecessary bookkeeping" This reverts commit 3fd39d3694d32efa44242c099e923a7f4d982095. Revert "[FastISel] NFC: Remove obsolete -fast-isel-sink-local-values option" This reverts commit a474657e30edccd9e175d92bddeefcfa544751b2. Revert "Remove static function unused after cf1c774." This reverts commit dc35368ccf17a7dca0874ace7490cc3836fb063f. Revert "[lldb] Fix TestThreadStepOut.py after "Flush local value map on every instruction"" This reverts commit 53a14a47ee89dadb8798ca8ed19848f33f4551d5.
2020-12-01 13:23:30 -08:00
; ELF64: addi
[PowerPC] Add loads, stores, and related things to fast-isel. This is the next big chunk of fast-isel code. The primary purpose is to implement selection of loads and stores, but there is a lot of drag-along to support this. The common code to analyze addresses for both loads and stores is substantial. It's also necessary to add the materialization code for global values. Related to load-store processing is the code to fold loads into integer extends, since otherwise we generate lots of redundant instructions. We also need to add some overrides to some FastEmit routines to ensure we don't assign GPR 0 to a virtual register when this would change the meaning of an instruction. I added handling selection of a few binary arithmetic instructions, to enable committing some test cases I wrote a while back. Finally, ap couple of miscellaneous changes: * I cleaned up some poor style from a previous patch in PPCISelLowering.cpp, pointed out by David Blaikie. * I enlarged the Addr.Offset field to avoid sign problems with 32-bit offsets. llvm-svn: 189636
2013-08-30 02:29:45 +00:00
; ELF64: lis
; ELF64: ori
; ELF64: stdx
ret void
}
Reference in New Issue Copy Permalink