maxjhandsome
/
llvm-project
forked from OSchip/llvm-project
1
0
Fork 0
Code Issues Pull Requests Packages Releases Wiki Activity
llvm-project/llvm/test/CodeGen/X86/pr32345.ll

171 lines
6.0 KiB
LLVM
Raw Normal View History

[SDAG] Avoid deleted SDNodes PromoteIntBinOp Reorder work in PromoteIntBinOp to prevent stale (deleted) nodes from being used. Fixes PR32340 and PR32345. Reviewers: hfinkel, dbabokin Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D31148 llvm-svn: 298923
2017-03-28 23:41:12 +08:00
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
[FastISel] Disable local value sinking by default This is causing compilation timeouts on code with long sequences of local values and calls (i.e. foo(1); foo(2); foo(3); ...). It turns out that code coverage instrumentation is a great way to create sequences like this, which how our users ran into the issue in practice. Intel has a tool that detects these kinds of non-linear compile time issues, and Andy Kaylor reported it as PR37010. The current sinking code scans the whole basic block once per local value sink, which happens before emitting each call. In theory, local values should only be introduced to be used by instructions between the current flush point and the last flush point, so we should only need to scan those instructions. llvm-svn: 329822
2018-04-12 00:03:07 +08:00
; RUN: llc -fast-isel-sink-local-values -O0 -mtriple=x86_64-unknown-linux-gnu -o - %s | FileCheck %s -check-prefix=X640
; RUN: llc -fast-isel-sink-local-values -O0 -mtriple=i686-unknown -o - %s | FileCheck %s -check-prefix=6860
; RUN: llc -fast-isel-sink-local-values -mtriple=x86_64-unknown-linux-gnu -o - %s | FileCheck %s -check-prefix=X64
; RUN: llc -fast-isel-sink-local-values -mtriple=i686-unknown -o - %s | FileCheck %s -check-prefix=686
[SDAG] Avoid deleted SDNodes PromoteIntBinOp Reorder work in PromoteIntBinOp to prevent stale (deleted) nodes from being used. Fixes PR32340 and PR32345. Reviewers: hfinkel, dbabokin Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D31148 llvm-svn: 298923
2017-03-28 23:41:12 +08:00
@var_22 = external global i16, align 2
@var_27 = external global i16, align 2
define void @foo() {
; X640-LABEL: foo:
[CodeGen] Unify MBB reference format in both MIR and debug output As part of the unification of the debug format and the MIR format, print MBB references as '%bb.5'. The MIR printer prints the IR name of a MBB only for block definitions. * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)->getNumber\(\)/" << printMBBReference(*\1)/g' * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)\.getNumber\(\)/" << printMBBReference(\1)/g' * find . \( -name "*.txt" -o -name "*.s" -o -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#([0-9]+)/%bb.\1/g' * grep -nr 'BB#' and fix Differential Revision: https://reviews.llvm.org/D40422 llvm-svn: 319665
2017-12-05 01:18:51 +08:00
; X640: # %bb.0: # %bb
[FastISel] Sink local value materializations to first use Summary: Local values are constants, global addresses, and stack addresses that can't be folded into the instruction that uses them. For example, when storing the address of a global variable into memory, we need to materialize that address into a register. FastISel doesn't want to materialize any given local value more than once, so it generates all local value materialization code at EmitStartPt, which always dominates the current insertion point. This allows it to maintain a map of local value registers, and it knows that the local value area will always dominate the current insertion point. The downside is that local value instructions are always emitted without a source location. This is done to prevent jumpy line tables, but it means that the local value area will be considered part of the previous statement. Consider this C code: call1(); // line 1 ++global; // line 2 ++global; // line 3 call2(&global, &local); // line 4 Today we end up with assembly and line tables like this: .loc 1 1 callq call1 leaq global(%rip), %rdi leaq local(%rsp), %rsi .loc 1 2 addq $1, global(%rip) .loc 1 3 addq $1, global(%rip) .loc 1 4 callq call2 The LEA instructions in the local value area have no source location and are treated as being on line 1. Stepping through the code in a debugger and correlating it with the assembly won't make much sense, because these materializations are only required for line 4. This is actually problematic for the VS debugger "set next statement" feature, which effectively assumes that there are no registers live across statement boundaries. By sinking the local value code into the statement and fixing up the source location, we can make that feature work. This was filed as https://bugs.llvm.org/show_bug.cgi?id=35975 and https://crbug.com/793819. This change is obviously not enough to make this feature work reliably in all cases, but I felt that it was worth doing anyway because it usually generates smaller, more comprehensible -O0 code. I measured a 0.12% regression in code generation time with LLC on the sqlite3 amalgamation, so I think this is worth doing. There are some special cases worth calling out in the commit message: 1. local values materialized for phis 2. local values used by no-op casts 3. dead local value code Local values can be materialized for phis, and this does not show up as a vreg use in MachineRegisterInfo. In this case, if there are no other uses, this patch sinks the value to the first terminator, EH label, or the end of the BB if nothing else exists. 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. Lastly, if the local value register has no other uses, we can delete it. This comes up when fastisel tries two instruction selection approaches and the first materializes the value but fails and the second succeeds without using the local value. Reviewers: aprantl, dblaikie, qcolombet, MatzeB, vsk, echristo Subscribers: dotdash, chandlerc, hans, sdardis, amccarth, javed.absar, zturner, llvm-commits, hiraditya Differential Revision: https://reviews.llvm.org/D43093 llvm-svn: 327581
2018-03-15 05:54:21 +08:00
; X640-NEXT: movzwl var_22, %eax
[SDAG] Avoid deleted SDNodes PromoteIntBinOp Reorder work in PromoteIntBinOp to prevent stale (deleted) nodes from being used. Fixes PR32340 and PR32345. Reviewers: hfinkel, dbabokin Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D31148 llvm-svn: 298923
2017-03-28 23:41:12 +08:00
; X640-NEXT: movzwl var_27, %ecx
[FastISel] Sink local value materializations to first use Summary: Local values are constants, global addresses, and stack addresses that can't be folded into the instruction that uses them. For example, when storing the address of a global variable into memory, we need to materialize that address into a register. FastISel doesn't want to materialize any given local value more than once, so it generates all local value materialization code at EmitStartPt, which always dominates the current insertion point. This allows it to maintain a map of local value registers, and it knows that the local value area will always dominate the current insertion point. The downside is that local value instructions are always emitted without a source location. This is done to prevent jumpy line tables, but it means that the local value area will be considered part of the previous statement. Consider this C code: call1(); // line 1 ++global; // line 2 ++global; // line 3 call2(&global, &local); // line 4 Today we end up with assembly and line tables like this: .loc 1 1 callq call1 leaq global(%rip), %rdi leaq local(%rsp), %rsi .loc 1 2 addq $1, global(%rip) .loc 1 3 addq $1, global(%rip) .loc 1 4 callq call2 The LEA instructions in the local value area have no source location and are treated as being on line 1. Stepping through the code in a debugger and correlating it with the assembly won't make much sense, because these materializations are only required for line 4. This is actually problematic for the VS debugger "set next statement" feature, which effectively assumes that there are no registers live across statement boundaries. By sinking the local value code into the statement and fixing up the source location, we can make that feature work. This was filed as https://bugs.llvm.org/show_bug.cgi?id=35975 and https://crbug.com/793819. This change is obviously not enough to make this feature work reliably in all cases, but I felt that it was worth doing anyway because it usually generates smaller, more comprehensible -O0 code. I measured a 0.12% regression in code generation time with LLC on the sqlite3 amalgamation, so I think this is worth doing. There are some special cases worth calling out in the commit message: 1. local values materialized for phis 2. local values used by no-op casts 3. dead local value code Local values can be materialized for phis, and this does not show up as a vreg use in MachineRegisterInfo. In this case, if there are no other uses, this patch sinks the value to the first terminator, EH label, or the end of the BB if nothing else exists. 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. Lastly, if the local value register has no other uses, we can delete it. This comes up when fastisel tries two instruction selection approaches and the first materializes the value but fails and the second succeeds without using the local value. Reviewers: aprantl, dblaikie, qcolombet, MatzeB, vsk, echristo Subscribers: dotdash, chandlerc, hans, sdardis, amccarth, javed.absar, zturner, llvm-commits, hiraditya Differential Revision: https://reviews.llvm.org/D43093 llvm-svn: 327581
2018-03-15 05:54:21 +08:00
; X640-NEXT: xorl %ecx, %eax
; X640-NEXT: movzwl var_27, %ecx
; X640-NEXT: xorl %ecx, %eax
; X640-NEXT: movslq %eax, %rdx
; X640-NEXT: movq %rdx, -{{[0-9]+}}(%rsp)
; X640-NEXT: movzwl var_22, %eax
; X640-NEXT: movzwl var_27, %ecx
; X640-NEXT: xorl %ecx, %eax
; X640-NEXT: movzwl var_27, %ecx
; X640-NEXT: xorl %ecx, %eax
; X640-NEXT: movslq %eax, %rdx
; X640-NEXT: movzwl var_27, %eax
; X640-NEXT: subl $16610, %eax # imm = 0x40E2
; X640-NEXT: movl %eax, %eax
; X640-NEXT: movl %eax, %ecx
Followup on Proposal to move MIR physical register namespace to '$' sigil. Discussed here: http://lists.llvm.org/pipermail/llvm-dev/2018-January/120320.html In preparation for adding support for named vregs we are changing the sigil for physical registers in MIR to '$' from '%'. This will prevent name clashes of named physical register with named vregs. llvm-svn: 323922
2018-02-01 06:04:26 +08:00
; X640-NEXT: # kill: def $cl killed $rcx
[FastISel] Sink local value materializations to first use Summary: Local values are constants, global addresses, and stack addresses that can't be folded into the instruction that uses them. For example, when storing the address of a global variable into memory, we need to materialize that address into a register. FastISel doesn't want to materialize any given local value more than once, so it generates all local value materialization code at EmitStartPt, which always dominates the current insertion point. This allows it to maintain a map of local value registers, and it knows that the local value area will always dominate the current insertion point. The downside is that local value instructions are always emitted without a source location. This is done to prevent jumpy line tables, but it means that the local value area will be considered part of the previous statement. Consider this C code: call1(); // line 1 ++global; // line 2 ++global; // line 3 call2(&global, &local); // line 4 Today we end up with assembly and line tables like this: .loc 1 1 callq call1 leaq global(%rip), %rdi leaq local(%rsp), %rsi .loc 1 2 addq $1, global(%rip) .loc 1 3 addq $1, global(%rip) .loc 1 4 callq call2 The LEA instructions in the local value area have no source location and are treated as being on line 1. Stepping through the code in a debugger and correlating it with the assembly won't make much sense, because these materializations are only required for line 4. This is actually problematic for the VS debugger "set next statement" feature, which effectively assumes that there are no registers live across statement boundaries. By sinking the local value code into the statement and fixing up the source location, we can make that feature work. This was filed as https://bugs.llvm.org/show_bug.cgi?id=35975 and https://crbug.com/793819. This change is obviously not enough to make this feature work reliably in all cases, but I felt that it was worth doing anyway because it usually generates smaller, more comprehensible -O0 code. I measured a 0.12% regression in code generation time with LLC on the sqlite3 amalgamation, so I think this is worth doing. There are some special cases worth calling out in the commit message: 1. local values materialized for phis 2. local values used by no-op casts 3. dead local value code Local values can be materialized for phis, and this does not show up as a vreg use in MachineRegisterInfo. In this case, if there are no other uses, this patch sinks the value to the first terminator, EH label, or the end of the BB if nothing else exists. 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. Lastly, if the local value register has no other uses, we can delete it. This comes up when fastisel tries two instruction selection approaches and the first materializes the value but fails and the second succeeds without using the local value. Reviewers: aprantl, dblaikie, qcolombet, MatzeB, vsk, echristo Subscribers: dotdash, chandlerc, hans, sdardis, amccarth, javed.absar, zturner, llvm-commits, hiraditya Differential Revision: https://reviews.llvm.org/D43093 llvm-svn: 327581
2018-03-15 05:54:21 +08:00
; X640-NEXT: sarq %cl, %rdx
RegAllocFast: Improve hinting heuristic Trace through multiple COPYs when looking for a physreg source. Add hinting for vregs that will be copied into physregs (we only hinted for vregs getting copied to a physreg previously). Give hinted a register a bonus when deciding which value to spill. This is part of my rewrite regallocfast series. In fact this one doesn't even have an effect unless you also flip the allocation to happen from back to front of a basic block. Nonetheless it helps to split this up to ease review of D52010 Patch by Matthias Braun llvm-svn: 360887
2019-05-16 20:50:39 +08:00
; X640-NEXT: # kill: def $dl killed $dl killed $rdx
; X640-NEXT: # implicit-def: $rsi
; X640-NEXT: movb %dl, (%rsi)
[SDAG] Avoid deleted SDNodes PromoteIntBinOp Reorder work in PromoteIntBinOp to prevent stale (deleted) nodes from being used. Fixes PR32340 and PR32345. Reviewers: hfinkel, dbabokin Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D31148 llvm-svn: 298923
2017-03-28 23:41:12 +08:00
; X640-NEXT: retq
;
; 6860-LABEL: foo:
[CodeGen] Unify MBB reference format in both MIR and debug output As part of the unification of the debug format and the MIR format, print MBB references as '%bb.5'. The MIR printer prints the IR name of a MBB only for block definitions. * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)->getNumber\(\)/" << printMBBReference(*\1)/g' * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)\.getNumber\(\)/" << printMBBReference(\1)/g' * find . \( -name "*.txt" -o -name "*.s" -o -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#([0-9]+)/%bb.\1/g' * grep -nr 'BB#' and fix Differential Revision: https://reviews.llvm.org/D40422 llvm-svn: 319665
2017-12-05 01:18:51 +08:00
; 6860: # %bb.0: # %bb
[SDAG] Avoid deleted SDNodes PromoteIntBinOp Reorder work in PromoteIntBinOp to prevent stale (deleted) nodes from being used. Fixes PR32340 and PR32345. Reviewers: hfinkel, dbabokin Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D31148 llvm-svn: 298923
2017-03-28 23:41:12 +08:00
; 6860-NEXT: pushl %ebp
; 6860-NEXT: .cfi_def_cfa_offset 8
; 6860-NEXT: .cfi_offset %ebp, -8
; 6860-NEXT: movl %esp, %ebp
; 6860-NEXT: .cfi_def_cfa_register %ebp
; 6860-NEXT: pushl %ebx
; 6860-NEXT: pushl %edi
; 6860-NEXT: pushl %esi
; 6860-NEXT: andl $-8, %esp
; 6860-NEXT: subl $32, %esp
; 6860-NEXT: .cfi_offset %esi, -20
; 6860-NEXT: .cfi_offset %edi, -16
; 6860-NEXT: .cfi_offset %ebx, -12
[FastISel] Sink local value materializations to first use Summary: Local values are constants, global addresses, and stack addresses that can't be folded into the instruction that uses them. For example, when storing the address of a global variable into memory, we need to materialize that address into a register. FastISel doesn't want to materialize any given local value more than once, so it generates all local value materialization code at EmitStartPt, which always dominates the current insertion point. This allows it to maintain a map of local value registers, and it knows that the local value area will always dominate the current insertion point. The downside is that local value instructions are always emitted without a source location. This is done to prevent jumpy line tables, but it means that the local value area will be considered part of the previous statement. Consider this C code: call1(); // line 1 ++global; // line 2 ++global; // line 3 call2(&global, &local); // line 4 Today we end up with assembly and line tables like this: .loc 1 1 callq call1 leaq global(%rip), %rdi leaq local(%rsp), %rsi .loc 1 2 addq $1, global(%rip) .loc 1 3 addq $1, global(%rip) .loc 1 4 callq call2 The LEA instructions in the local value area have no source location and are treated as being on line 1. Stepping through the code in a debugger and correlating it with the assembly won't make much sense, because these materializations are only required for line 4. This is actually problematic for the VS debugger "set next statement" feature, which effectively assumes that there are no registers live across statement boundaries. By sinking the local value code into the statement and fixing up the source location, we can make that feature work. This was filed as https://bugs.llvm.org/show_bug.cgi?id=35975 and https://crbug.com/793819. This change is obviously not enough to make this feature work reliably in all cases, but I felt that it was worth doing anyway because it usually generates smaller, more comprehensible -O0 code. I measured a 0.12% regression in code generation time with LLC on the sqlite3 amalgamation, so I think this is worth doing. There are some special cases worth calling out in the commit message: 1. local values materialized for phis 2. local values used by no-op casts 3. dead local value code Local values can be materialized for phis, and this does not show up as a vreg use in MachineRegisterInfo. In this case, if there are no other uses, this patch sinks the value to the first terminator, EH label, or the end of the BB if nothing else exists. 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. Lastly, if the local value register has no other uses, we can delete it. This comes up when fastisel tries two instruction selection approaches and the first materializes the value but fails and the second succeeds without using the local value. Reviewers: aprantl, dblaikie, qcolombet, MatzeB, vsk, echristo Subscribers: dotdash, chandlerc, hans, sdardis, amccarth, javed.absar, zturner, llvm-commits, hiraditya Differential Revision: https://reviews.llvm.org/D43093 llvm-svn: 327581
2018-03-15 05:54:21 +08:00
; 6860-NEXT: movw var_22, %ax
; 6860-NEXT: movzwl var_27, %ecx
; 6860-NEXT: movw %cx, %dx
; 6860-NEXT: xorw %dx, %ax
; 6860-NEXT: # implicit-def: $esi
; 6860-NEXT: movw %ax, %si
; 6860-NEXT: xorl %ecx, %esi
RegAllocFast: Improve hinting heuristic Trace through multiple COPYs when looking for a physreg source. Add hinting for vregs that will be copied into physregs (we only hinted for vregs getting copied to a physreg previously). Give hinted a register a bonus when deciding which value to spill. This is part of my rewrite regallocfast series. In fact this one doesn't even have an effect unless you also flip the allocation to happen from back to front of a basic block. Nonetheless it helps to split this up to ease review of D52010 Patch by Matthias Braun llvm-svn: 360887
2019-05-16 20:50:39 +08:00
; 6860-NEXT: # kill: def $si killed $si killed $esi
; 6860-NEXT: movzwl %si, %ecx
[FastISel] Sink local value materializations to first use Summary: Local values are constants, global addresses, and stack addresses that can't be folded into the instruction that uses them. For example, when storing the address of a global variable into memory, we need to materialize that address into a register. FastISel doesn't want to materialize any given local value more than once, so it generates all local value materialization code at EmitStartPt, which always dominates the current insertion point. This allows it to maintain a map of local value registers, and it knows that the local value area will always dominate the current insertion point. The downside is that local value instructions are always emitted without a source location. This is done to prevent jumpy line tables, but it means that the local value area will be considered part of the previous statement. Consider this C code: call1(); // line 1 ++global; // line 2 ++global; // line 3 call2(&global, &local); // line 4 Today we end up with assembly and line tables like this: .loc 1 1 callq call1 leaq global(%rip), %rdi leaq local(%rsp), %rsi .loc 1 2 addq $1, global(%rip) .loc 1 3 addq $1, global(%rip) .loc 1 4 callq call2 The LEA instructions in the local value area have no source location and are treated as being on line 1. Stepping through the code in a debugger and correlating it with the assembly won't make much sense, because these materializations are only required for line 4. This is actually problematic for the VS debugger "set next statement" feature, which effectively assumes that there are no registers live across statement boundaries. By sinking the local value code into the statement and fixing up the source location, we can make that feature work. This was filed as https://bugs.llvm.org/show_bug.cgi?id=35975 and https://crbug.com/793819. This change is obviously not enough to make this feature work reliably in all cases, but I felt that it was worth doing anyway because it usually generates smaller, more comprehensible -O0 code. I measured a 0.12% regression in code generation time with LLC on the sqlite3 amalgamation, so I think this is worth doing. There are some special cases worth calling out in the commit message: 1. local values materialized for phis 2. local values used by no-op casts 3. dead local value code Local values can be materialized for phis, and this does not show up as a vreg use in MachineRegisterInfo. In this case, if there are no other uses, this patch sinks the value to the first terminator, EH label, or the end of the BB if nothing else exists. 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. Lastly, if the local value register has no other uses, we can delete it. This comes up when fastisel tries two instruction selection approaches and the first materializes the value but fails and the second succeeds without using the local value. Reviewers: aprantl, dblaikie, qcolombet, MatzeB, vsk, echristo Subscribers: dotdash, chandlerc, hans, sdardis, amccarth, javed.absar, zturner, llvm-commits, hiraditya Differential Revision: https://reviews.llvm.org/D43093 llvm-svn: 327581
2018-03-15 05:54:21 +08:00
; 6860-NEXT: movl %ecx, {{[0-9]+}}(%esp)
[SDAG] Avoid deleted SDNodes PromoteIntBinOp Reorder work in PromoteIntBinOp to prevent stale (deleted) nodes from being used. Fixes PR32340 and PR32345. Reviewers: hfinkel, dbabokin Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D31148 llvm-svn: 298923
2017-03-28 23:41:12 +08:00
; 6860-NEXT: movl $0, {{[0-9]+}}(%esp)
[FastISel] Sink local value materializations to first use Summary: Local values are constants, global addresses, and stack addresses that can't be folded into the instruction that uses them. For example, when storing the address of a global variable into memory, we need to materialize that address into a register. FastISel doesn't want to materialize any given local value more than once, so it generates all local value materialization code at EmitStartPt, which always dominates the current insertion point. This allows it to maintain a map of local value registers, and it knows that the local value area will always dominate the current insertion point. The downside is that local value instructions are always emitted without a source location. This is done to prevent jumpy line tables, but it means that the local value area will be considered part of the previous statement. Consider this C code: call1(); // line 1 ++global; // line 2 ++global; // line 3 call2(&global, &local); // line 4 Today we end up with assembly and line tables like this: .loc 1 1 callq call1 leaq global(%rip), %rdi leaq local(%rsp), %rsi .loc 1 2 addq $1, global(%rip) .loc 1 3 addq $1, global(%rip) .loc 1 4 callq call2 The LEA instructions in the local value area have no source location and are treated as being on line 1. Stepping through the code in a debugger and correlating it with the assembly won't make much sense, because these materializations are only required for line 4. This is actually problematic for the VS debugger "set next statement" feature, which effectively assumes that there are no registers live across statement boundaries. By sinking the local value code into the statement and fixing up the source location, we can make that feature work. This was filed as https://bugs.llvm.org/show_bug.cgi?id=35975 and https://crbug.com/793819. This change is obviously not enough to make this feature work reliably in all cases, but I felt that it was worth doing anyway because it usually generates smaller, more comprehensible -O0 code. I measured a 0.12% regression in code generation time with LLC on the sqlite3 amalgamation, so I think this is worth doing. There are some special cases worth calling out in the commit message: 1. local values materialized for phis 2. local values used by no-op casts 3. dead local value code Local values can be materialized for phis, and this does not show up as a vreg use in MachineRegisterInfo. In this case, if there are no other uses, this patch sinks the value to the first terminator, EH label, or the end of the BB if nothing else exists. 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. Lastly, if the local value register has no other uses, we can delete it. This comes up when fastisel tries two instruction selection approaches and the first materializes the value but fails and the second succeeds without using the local value. Reviewers: aprantl, dblaikie, qcolombet, MatzeB, vsk, echristo Subscribers: dotdash, chandlerc, hans, sdardis, amccarth, javed.absar, zturner, llvm-commits, hiraditya Differential Revision: https://reviews.llvm.org/D43093 llvm-svn: 327581
2018-03-15 05:54:21 +08:00
; 6860-NEXT: movw var_22, %ax
; 6860-NEXT: movzwl var_27, %ecx
; 6860-NEXT: movw %cx, %dx
; 6860-NEXT: xorw %dx, %ax
RegAllocFast: Improve hinting heuristic Trace through multiple COPYs when looking for a physreg source. Add hinting for vregs that will be copied into physregs (we only hinted for vregs getting copied to a physreg previously). Give hinted a register a bonus when deciding which value to spill. This is part of my rewrite regallocfast series. In fact this one doesn't even have an effect unless you also flip the allocation to happen from back to front of a basic block. Nonetheless it helps to split this up to ease review of D52010 Patch by Matthias Braun llvm-svn: 360887
2019-05-16 20:50:39 +08:00
; 6860-NEXT: # implicit-def: $edi
; 6860-NEXT: movw %ax, %di
; 6860-NEXT: xorl %ecx, %edi
; 6860-NEXT: # kill: def $di killed $di killed $edi
; 6860-NEXT: movzwl %di, %ebx
; 6860-NEXT: # kill: def $cl killed $cl killed $ecx
; 6860-NEXT: addb $30, %cl
; 6860-NEXT: xorl %eax, %eax
; 6860-NEXT: movb %cl, {{[-0-9]+}}(%e{{[sb]}}p) # 1-byte Spill
; 6860-NEXT: shrdl %cl, %eax, %ebx
; 6860-NEXT: movb {{[-0-9]+}}(%e{{[sb]}}p), %cl # 1-byte Reload
; 6860-NEXT: testb $32, %cl
; 6860-NEXT: movl %ebx, {{[-0-9]+}}(%e{{[sb]}}p) # 4-byte Spill
; 6860-NEXT: movl %eax, {{[-0-9]+}}(%e{{[sb]}}p) # 4-byte Spill
[SDAG] Avoid deleted SDNodes PromoteIntBinOp Reorder work in PromoteIntBinOp to prevent stale (deleted) nodes from being used. Fixes PR32340 and PR32345. Reviewers: hfinkel, dbabokin Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D31148 llvm-svn: 298923
2017-03-28 23:41:12 +08:00
; 6860-NEXT: jne .LBB0_2
[CodeGen] Unify MBB reference format in both MIR and debug output As part of the unification of the debug format and the MIR format, print MBB references as '%bb.5'. The MIR printer prints the IR name of a MBB only for block definitions. * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)->getNumber\(\)/" << printMBBReference(*\1)/g' * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)\.getNumber\(\)/" << printMBBReference(\1)/g' * find . \( -name "*.txt" -o -name "*.s" -o -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#([0-9]+)/%bb.\1/g' * grep -nr 'BB#' and fix Differential Revision: https://reviews.llvm.org/D40422 llvm-svn: 319665
2017-12-05 01:18:51 +08:00
; 6860-NEXT: # %bb.1: # %bb
[DAGCombiner][X86] When promoting loads don't use ZEXTLOAD even its legal We were previously prefering ZEXTLOAD over EXTLOAD if it is legal. This triggers during X86's promotion of i16->i32. Not sure about other targets. Using ZEXTLOAD can prevent folding it to SEXTLOAD later if we were to promote a sign extended operand like we would need for SRA. However, X86 doesn't currently promote i16 SRA. I was looking into doing that which is how I found this issue. This is also blocking our ability to fold 4 byte aligned EXTLOADs with "loadi32". This is what caused most of the test changes here. Differential Revision: https://reviews.llvm.org/D45585#inline-402825 llvm-svn: 330781
2018-04-25 06:35:27 +08:00
; 6860-NEXT: movl {{[-0-9]+}}(%e{{[sb]}}p), %eax # 4-byte Reload
; 6860-NEXT: movl %eax, {{[-0-9]+}}(%e{{[sb]}}p) # 4-byte Spill
[SDAG] Avoid deleted SDNodes PromoteIntBinOp Reorder work in PromoteIntBinOp to prevent stale (deleted) nodes from being used. Fixes PR32340 and PR32345. Reviewers: hfinkel, dbabokin Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D31148 llvm-svn: 298923
2017-03-28 23:41:12 +08:00
; 6860-NEXT: .LBB0_2: # %bb
[DAGCombiner][X86] When promoting loads don't use ZEXTLOAD even its legal We were previously prefering ZEXTLOAD over EXTLOAD if it is legal. This triggers during X86's promotion of i16->i32. Not sure about other targets. Using ZEXTLOAD can prevent folding it to SEXTLOAD later if we were to promote a sign extended operand like we would need for SRA. However, X86 doesn't currently promote i16 SRA. I was looking into doing that which is how I found this issue. This is also blocking our ability to fold 4 byte aligned EXTLOADs with "loadi32". This is what caused most of the test changes here. Differential Revision: https://reviews.llvm.org/D45585#inline-402825 llvm-svn: 330781
2018-04-25 06:35:27 +08:00
; 6860-NEXT: movl {{[-0-9]+}}(%e{{[sb]}}p), %eax # 4-byte Reload
RegAllocFast: Improve hinting heuristic Trace through multiple COPYs when looking for a physreg source. Add hinting for vregs that will be copied into physregs (we only hinted for vregs getting copied to a physreg previously). Give hinted a register a bonus when deciding which value to spill. This is part of my rewrite regallocfast series. In fact this one doesn't even have an effect unless you also flip the allocation to happen from back to front of a basic block. Nonetheless it helps to split this up to ease review of D52010 Patch by Matthias Braun llvm-svn: 360887
2019-05-16 20:50:39 +08:00
; 6860-NEXT: # kill: def $al killed $al killed $eax
; 6860-NEXT: # implicit-def: $ecx
; 6860-NEXT: movb %al, (%ecx)
[SDAG] Avoid deleted SDNodes PromoteIntBinOp Reorder work in PromoteIntBinOp to prevent stale (deleted) nodes from being used. Fixes PR32340 and PR32345. Reviewers: hfinkel, dbabokin Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D31148 llvm-svn: 298923
2017-03-28 23:41:12 +08:00
; 6860-NEXT: leal -12(%ebp), %esp
; 6860-NEXT: popl %esi
; 6860-NEXT: popl %edi
; 6860-NEXT: popl %ebx
; 6860-NEXT: popl %ebp
Correct dwarf unwind information in function epilogue This patch aims to provide correct dwarf unwind information in function epilogue for X86. It consists of two parts. The first part inserts CFI instructions that set appropriate cfa offset and cfa register in emitEpilogue() in X86FrameLowering. This part is X86 specific. The second part is platform independent and ensures that: * CFI instructions do not affect code generation (they are not counted as instructions when tail duplicating or tail merging) * Unwind information remains correct when a function is modified by different passes. This is done in a late pass by analyzing information about cfa offset and cfa register in BBs and inserting additional CFI directives where necessary. Added CFIInstrInserter pass: * analyzes each basic block to determine cfa offset and register are valid at its entry and exit * verifies that outgoing cfa offset and register of predecessor blocks match incoming values of their successors * inserts additional CFI directives at basic block beginning to correct the rule for calculating CFA Having CFI instructions in function epilogue can cause incorrect CFA calculation rule for some basic blocks. This can happen if, due to basic block reordering, or the existence of multiple epilogue blocks, some of the blocks have wrong cfa offset and register values set by the epilogue block above them. CFIInstrInserter is currently run only on X86, but can be used by any target that implements support for adding CFI instructions in epilogue. Patch by Violeta Vukobrat. Differential Revision: https://reviews.llvm.org/D42848 llvm-svn: 330706
2018-04-24 18:32:08 +08:00
; 6860-NEXT: .cfi_def_cfa %esp, 4
[SDAG] Avoid deleted SDNodes PromoteIntBinOp Reorder work in PromoteIntBinOp to prevent stale (deleted) nodes from being used. Fixes PR32340 and PR32345. Reviewers: hfinkel, dbabokin Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D31148 llvm-svn: 298923
2017-03-28 23:41:12 +08:00
; 6860-NEXT: retl
;
; X64-LABEL: foo:
[CodeGen] Unify MBB reference format in both MIR and debug output As part of the unification of the debug format and the MIR format, print MBB references as '%bb.5'. The MIR printer prints the IR name of a MBB only for block definitions. * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)->getNumber\(\)/" << printMBBReference(*\1)/g' * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)\.getNumber\(\)/" << printMBBReference(\1)/g' * find . \( -name "*.txt" -o -name "*.s" -o -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#([0-9]+)/%bb.\1/g' * grep -nr 'BB#' and fix Differential Revision: https://reviews.llvm.org/D40422 llvm-svn: 319665
2017-12-05 01:18:51 +08:00
; X64: # %bb.0: # %bb
[DAGCombiner][X86] When promoting loads don't use ZEXTLOAD even its legal We were previously prefering ZEXTLOAD over EXTLOAD if it is legal. This triggers during X86's promotion of i16->i32. Not sure about other targets. Using ZEXTLOAD can prevent folding it to SEXTLOAD later if we were to promote a sign extended operand like we would need for SRA. However, X86 doesn't currently promote i16 SRA. I was looking into doing that which is how I found this issue. This is also blocking our ability to fold 4 byte aligned EXTLOADs with "loadi32". This is what caused most of the test changes here. Differential Revision: https://reviews.llvm.org/D45585#inline-402825 llvm-svn: 330781
2018-04-25 06:35:27 +08:00
; X64-NEXT: movzwl {{.*}}(%rip), %ecx
[DAGCombiner] allow narrowing of add followed by truncate trunc (add X, C ) --> add (trunc X), C' If we're throwing away the top bits of an 'add' instruction, do it in the narrow destination type. This makes the truncate-able opcode list identical to the sibling transform done in IR (in instcombine). This change used to show regressions for x86, but those are gone after D55494. This gets us closer to deleting the x86 custom function (combineTruncatedArithmetic) that does almost the same thing. Differential Revision: https://reviews.llvm.org/D55866 llvm-svn: 350006
2018-12-23 01:10:31 +08:00
; X64-NEXT: movzwl {{.*}}(%rip), %eax
; X64-NEXT: xorw %cx, %ax
; X64-NEXT: xorl %ecx, %eax
; X64-NEXT: movzwl %ax, %eax
[SDAG] Avoid deleted SDNodes PromoteIntBinOp Reorder work in PromoteIntBinOp to prevent stale (deleted) nodes from being used. Fixes PR32340 and PR32345. Reviewers: hfinkel, dbabokin Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D31148 llvm-svn: 298923
2017-03-28 23:41:12 +08:00
; X64-NEXT: movq %rax, -{{[0-9]+}}(%rsp)
[DAGCombiner] allow narrowing of add followed by truncate trunc (add X, C ) --> add (trunc X), C' If we're throwing away the top bits of an 'add' instruction, do it in the narrow destination type. This makes the truncate-able opcode list identical to the sibling transform done in IR (in instcombine). This change used to show regressions for x86, but those are gone after D55494. This gets us closer to deleting the x86 custom function (combineTruncatedArithmetic) that does almost the same thing. Differential Revision: https://reviews.llvm.org/D55866 llvm-svn: 350006
2018-12-23 01:10:31 +08:00
; X64-NEXT: addb $30, %cl
Followup on Proposal to move MIR physical register namespace to '$' sigil. Discussed here: http://lists.llvm.org/pipermail/llvm-dev/2018-January/120320.html In preparation for adding support for named vregs we are changing the sigil for physical registers in MIR to '$' from '%'. This will prevent name clashes of named physical register with named vregs. llvm-svn: 323922
2018-02-01 06:04:26 +08:00
; X64-NEXT: # kill: def $cl killed $cl killed $ecx
[SDAG] Avoid deleted SDNodes PromoteIntBinOp Reorder work in PromoteIntBinOp to prevent stale (deleted) nodes from being used. Fixes PR32340 and PR32345. Reviewers: hfinkel, dbabokin Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D31148 llvm-svn: 298923
2017-03-28 23:41:12 +08:00
; X64-NEXT: shrq %cl, %rax
; X64-NEXT: movb %al, (%rax)
; X64-NEXT: retq
;
; 686-LABEL: foo:
[CodeGen] Unify MBB reference format in both MIR and debug output As part of the unification of the debug format and the MIR format, print MBB references as '%bb.5'. The MIR printer prints the IR name of a MBB only for block definitions. * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)->getNumber\(\)/" << printMBBReference(*\1)/g' * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)\.getNumber\(\)/" << printMBBReference(\1)/g' * find . \( -name "*.txt" -o -name "*.s" -o -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#([0-9]+)/%bb.\1/g' * grep -nr 'BB#' and fix Differential Revision: https://reviews.llvm.org/D40422 llvm-svn: 319665
2017-12-05 01:18:51 +08:00
; 686: # %bb.0: # %bb
[SDAG] Avoid deleted SDNodes PromoteIntBinOp Reorder work in PromoteIntBinOp to prevent stale (deleted) nodes from being used. Fixes PR32340 and PR32345. Reviewers: hfinkel, dbabokin Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D31148 llvm-svn: 298923
2017-03-28 23:41:12 +08:00
; 686-NEXT: pushl %ebp
; 686-NEXT: .cfi_def_cfa_offset 8
; 686-NEXT: .cfi_offset %ebp, -8
; 686-NEXT: movl %esp, %ebp
; 686-NEXT: .cfi_def_cfa_register %ebp
; 686-NEXT: andl $-8, %esp
; 686-NEXT: subl $8, %esp
[DAGCombiner][X86] When promoting loads don't use ZEXTLOAD even its legal We were previously prefering ZEXTLOAD over EXTLOAD if it is legal. This triggers during X86's promotion of i16->i32. Not sure about other targets. Using ZEXTLOAD can prevent folding it to SEXTLOAD later if we were to promote a sign extended operand like we would need for SRA. However, X86 doesn't currently promote i16 SRA. I was looking into doing that which is how I found this issue. This is also blocking our ability to fold 4 byte aligned EXTLOADs with "loadi32". This is what caused most of the test changes here. Differential Revision: https://reviews.llvm.org/D45585#inline-402825 llvm-svn: 330781
2018-04-25 06:35:27 +08:00
; 686-NEXT: movzwl var_27, %ecx
[DAGCombiner] allow narrowing of add followed by truncate trunc (add X, C ) --> add (trunc X), C' If we're throwing away the top bits of an 'add' instruction, do it in the narrow destination type. This makes the truncate-able opcode list identical to the sibling transform done in IR (in instcombine). This change used to show regressions for x86, but those are gone after D55494. This gets us closer to deleting the x86 custom function (combineTruncatedArithmetic) that does almost the same thing. Differential Revision: https://reviews.llvm.org/D55866 llvm-svn: 350006
2018-12-23 01:10:31 +08:00
; 686-NEXT: movzwl var_22, %eax
; 686-NEXT: xorw %cx, %ax
; 686-NEXT: xorl %ecx, %eax
; 686-NEXT: movzwl %ax, %eax
[SDAG] Avoid deleted SDNodes PromoteIntBinOp Reorder work in PromoteIntBinOp to prevent stale (deleted) nodes from being used. Fixes PR32340 and PR32345. Reviewers: hfinkel, dbabokin Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D31148 llvm-svn: 298923
2017-03-28 23:41:12 +08:00
; 686-NEXT: movl %eax, (%esp)
; 686-NEXT: movl $0, {{[0-9]+}}(%esp)
[DAGCombiner] allow narrowing of add followed by truncate trunc (add X, C ) --> add (trunc X), C' If we're throwing away the top bits of an 'add' instruction, do it in the narrow destination type. This makes the truncate-able opcode list identical to the sibling transform done in IR (in instcombine). This change used to show regressions for x86, but those are gone after D55494. This gets us closer to deleting the x86 custom function (combineTruncatedArithmetic) that does almost the same thing. Differential Revision: https://reviews.llvm.org/D55866 llvm-svn: 350006
2018-12-23 01:10:31 +08:00
; 686-NEXT: addb $30, %cl
[SDAG] Avoid deleted SDNodes PromoteIntBinOp Reorder work in PromoteIntBinOp to prevent stale (deleted) nodes from being used. Fixes PR32340 and PR32345. Reviewers: hfinkel, dbabokin Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D31148 llvm-svn: 298923
2017-03-28 23:41:12 +08:00
; 686-NEXT: xorl %edx, %edx
; 686-NEXT: shrdl %cl, %edx, %eax
; 686-NEXT: testb $32, %cl
; 686-NEXT: jne .LBB0_2
[CodeGen] Unify MBB reference format in both MIR and debug output As part of the unification of the debug format and the MIR format, print MBB references as '%bb.5'. The MIR printer prints the IR name of a MBB only for block definitions. * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)->getNumber\(\)/" << printMBBReference(*\1)/g' * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)\.getNumber\(\)/" << printMBBReference(\1)/g' * find . \( -name "*.txt" -o -name "*.s" -o -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#([0-9]+)/%bb.\1/g' * grep -nr 'BB#' and fix Differential Revision: https://reviews.llvm.org/D40422 llvm-svn: 319665
2017-12-05 01:18:51 +08:00
; 686-NEXT: # %bb.1: # %bb
[SDAG] Avoid deleted SDNodes PromoteIntBinOp Reorder work in PromoteIntBinOp to prevent stale (deleted) nodes from being used. Fixes PR32340 and PR32345. Reviewers: hfinkel, dbabokin Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D31148 llvm-svn: 298923
2017-03-28 23:41:12 +08:00
; 686-NEXT: movl %eax, %edx
; 686-NEXT: .LBB0_2: # %bb
; 686-NEXT: movb %dl, (%eax)
; 686-NEXT: movl %ebp, %esp
; 686-NEXT: popl %ebp
Correct dwarf unwind information in function epilogue This patch aims to provide correct dwarf unwind information in function epilogue for X86. It consists of two parts. The first part inserts CFI instructions that set appropriate cfa offset and cfa register in emitEpilogue() in X86FrameLowering. This part is X86 specific. The second part is platform independent and ensures that: * CFI instructions do not affect code generation (they are not counted as instructions when tail duplicating or tail merging) * Unwind information remains correct when a function is modified by different passes. This is done in a late pass by analyzing information about cfa offset and cfa register in BBs and inserting additional CFI directives where necessary. Added CFIInstrInserter pass: * analyzes each basic block to determine cfa offset and register are valid at its entry and exit * verifies that outgoing cfa offset and register of predecessor blocks match incoming values of their successors * inserts additional CFI directives at basic block beginning to correct the rule for calculating CFA Having CFI instructions in function epilogue can cause incorrect CFA calculation rule for some basic blocks. This can happen if, due to basic block reordering, or the existence of multiple epilogue blocks, some of the blocks have wrong cfa offset and register values set by the epilogue block above them. CFIInstrInserter is currently run only on X86, but can be used by any target that implements support for adding CFI instructions in epilogue. Patch by Violeta Vukobrat. Differential Revision: https://reviews.llvm.org/D42848 llvm-svn: 330706
2018-04-24 18:32:08 +08:00
; 686-NEXT: .cfi_def_cfa %esp, 4
[SDAG] Avoid deleted SDNodes PromoteIntBinOp Reorder work in PromoteIntBinOp to prevent stale (deleted) nodes from being used. Fixes PR32340 and PR32345. Reviewers: hfinkel, dbabokin Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D31148 llvm-svn: 298923
2017-03-28 23:41:12 +08:00
; 686-NEXT: retl
bb:
%tmp = alloca i64, align 8
%tmp1 = load i16, i16* @var_22, align 2
%tmp2 = zext i16 %tmp1 to i32
%tmp3 = load i16, i16* @var_27, align 2
%tmp4 = zext i16 %tmp3 to i32
%tmp5 = xor i32 %tmp2, %tmp4
%tmp6 = load i16, i16* @var_27, align 2
%tmp7 = zext i16 %tmp6 to i32
%tmp8 = xor i32 %tmp5, %tmp7
%tmp9 = sext i32 %tmp8 to i64
store i64 %tmp9, i64* %tmp, align 8
%tmp10 = load i16, i16* @var_22, align 2
%tmp11 = zext i16 %tmp10 to i32
%tmp12 = load i16, i16* @var_27, align 2
%tmp13 = zext i16 %tmp12 to i32
%tmp14 = xor i32 %tmp11, %tmp13
%tmp15 = load i16, i16* @var_27, align 2
%tmp16 = zext i16 %tmp15 to i32
%tmp17 = xor i32 %tmp14, %tmp16
%tmp18 = sext i32 %tmp17 to i64
%tmp19 = load i16, i16* @var_27, align 2
%tmp20 = zext i16 %tmp19 to i32
%tmp21 = sub nsw i32 %tmp20, 16610
%tmp22 = zext i32 %tmp21 to i64
%tmp23 = ashr i64 %tmp18, %tmp22
%tmp24 = trunc i64 %tmp23 to i8
store i8 %tmp24, i8* undef, align 1
ret void
}