e3c72e1075
We currently generate code like this on x86 for a jump table with 5 elements, assuming the call target is in rbx: lea global_addr(%rip), %rax # initialize temporary rax with base address mov %rbx, %rcx # initialize another temporary rcx for index (rbx will be used for the call, so it is still live) sub %rax, %rcx # compute `address - base` ror $0x3, %rcx # compute `(address - base) ror 3` i.e. index cmp $0x4, %rcx # check index <= 4 ja .Ltrap [...] .Ltrap: ud1 A more efficient instruction sequence, that only needs one temporary register and one fewer instruction, is possible by subtracting the address we are testing from the fixed address instead of vice versa: lea (global_addr + 4*8)(%rip), %rax # initialize temporary rax with address of last element sub %rbx, %rax # compute `last element - address` ror $0x3, %rax # compute `(last element - address) ror 3` i.e. 4 - index cmp $0x4, %rax # check 4 - index <= 4 (same as above) ja .Ltrap [...] .Ltrap: ud1 Change LowerTypeTests to generate that sequence. As a consequence, the order of bits in the bitsets is reversed. Because it doesn't matter how we do the subtraction on other architectures (to the best of my knowledge), do so unconditionally. Reviewers: fmayer, vitalybuka Reviewed By: fmayer Pull Request: https://github.com/llvm/llvm-project/pull/142887
210 lines
9.5 KiB
LLVM
210 lines
9.5 KiB
LLVM
; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --include-generated-funcs --version 2
|
|
;; Check the cases involving internal CFI instrumented functions where we do not expect functions to be merged.
|
|
; RUN: opt -S -passes=mergefunc < %s | FileCheck %s
|
|
; RUN: opt -S -passes=mergefunc,lowertypetests < %s | FileCheck --check-prefix=LOWERTYPETESTS %s
|
|
|
|
target datalayout = "e-m:e-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128"
|
|
target triple = "x86_64-unknown-linux"
|
|
|
|
@0 = private unnamed_addr constant { i16, i16, [12 x i8] } { i16 -1, i16 0, [12 x i8] c"'int (int)'\00" }
|
|
|
|
; Function Attrs: noinline nounwind optnone
|
|
define dso_local i32 @f(i32 noundef %arg) #0 !type !3 !type !4 {
|
|
entry:
|
|
%arg.addr = alloca i32, align 4
|
|
%a = alloca i32, align 4
|
|
%b = alloca i32, align 4
|
|
store i32 %arg, ptr %arg.addr, align 4
|
|
store i32 0, ptr %b, align 4
|
|
%0 = load i32, ptr %arg.addr, align 4
|
|
%cmp = icmp sgt i32 %0, 0
|
|
br i1 %cmp, label %if.then, label %if.end
|
|
|
|
if.then: ; preds = %entry
|
|
store i32 1, ptr %a, align 4
|
|
br label %if.end
|
|
|
|
if.end: ; preds = %if.then, %entry
|
|
%1 = load i32, ptr %a, align 4
|
|
%2 = load i32, ptr %b, align 4
|
|
%add = add nsw i32 %1, %2
|
|
ret i32 %add
|
|
}
|
|
|
|
; Function Attrs: noinline nounwind optnone
|
|
define dso_local i32 @f_thunk(i32 noundef %arg) #0 !type !3 !type !4 {
|
|
entry:
|
|
%arg.addr = alloca i32, align 4
|
|
%a = alloca i32, align 4
|
|
%b = alloca i32, align 4
|
|
store i32 %arg, ptr %arg.addr, align 4
|
|
store i32 0, ptr %b, align 4
|
|
%0 = load i32, ptr %arg.addr, align 4
|
|
%cmp = icmp sgt i32 %0, 0
|
|
br i1 %cmp, label %if.then, label %if.end
|
|
|
|
if.then: ; preds = %entry
|
|
store i32 1, ptr %a, align 4
|
|
br label %if.end
|
|
|
|
if.end: ; preds = %if.then, %entry
|
|
%1 = load i32, ptr %a, align 4
|
|
%2 = load i32, ptr %b, align 4
|
|
%add = add nsw i32 %1, %2
|
|
ret i32 %add
|
|
}
|
|
|
|
; Function Attrs: noinline nounwind optnone
|
|
define dso_local i32 @g(i32 noundef %b) #0 !type !3 !type !4 {
|
|
entry:
|
|
%b.addr = alloca i32, align 4
|
|
%fp = alloca ptr, align 8
|
|
store i32 %b, ptr %b.addr, align 4
|
|
%0 = load i32, ptr %b.addr, align 4
|
|
%tobool = icmp ne i32 %0, 0
|
|
%1 = zext i1 %tobool to i64
|
|
%cond = select i1 %tobool, ptr @f, ptr @f_thunk
|
|
store ptr %cond, ptr %fp, align 8
|
|
%2 = load ptr, ptr %fp, align 8
|
|
%3 = call i1 @llvm.type.test(ptr %2, metadata !"_ZTSFiiE"), !nosanitize !5
|
|
br i1 %3, label %cont, label %trap, !nosanitize !5
|
|
|
|
trap: ; preds = %entry
|
|
call void @llvm.ubsantrap(i8 2) #3, !nosanitize !5
|
|
unreachable, !nosanitize !5
|
|
|
|
cont: ; preds = %entry
|
|
%4 = load i32, ptr %b.addr, align 4
|
|
%call = call i32 %2(i32 noundef %4)
|
|
ret i32 %call
|
|
}
|
|
|
|
; Function Attrs: nocallback nofree nosync nounwind speculatable willreturn memory(none)
|
|
declare i1 @llvm.type.test(ptr, metadata) #1
|
|
|
|
; Function Attrs: cold noreturn nounwind
|
|
declare void @llvm.ubsantrap(i8 immarg) #2
|
|
|
|
attributes #0 = { noinline nounwind optnone "min-legal-vector-width"="0" "no-trapping-math"="true" "stack-protector-buffer-size"="8" "target-features"="+cx8,+mmx,+sse,+sse2,+x87" }
|
|
attributes #1 = { nocallback nofree nosync nounwind speculatable willreturn memory(none) }
|
|
attributes #2 = { cold noreturn nounwind }
|
|
attributes #3 = { noreturn nounwind }
|
|
|
|
!llvm.module.flags = !{!0, !1}
|
|
|
|
!0 = !{i32 1, !"wchar_size", i32 4}
|
|
!1 = !{i32 4, !"CFI Canonical Jump Tables", i32 0}
|
|
!3 = !{i64 0, !"_ZTSFiiE"}
|
|
!4 = !{i64 0, !"_ZTSFiiE.generalized"}
|
|
!5 = !{}
|
|
; CHECK-LABEL: define dso_local i32 @f
|
|
; CHECK-SAME: (i32 noundef [[ARG:%.*]]) #[[ATTR0:[0-9]+]] !type [[META2:![0-9]+]] !type [[META3:![0-9]+]] {
|
|
; CHECK-NEXT: entry:
|
|
; CHECK-NEXT: [[ARG_ADDR:%.*]] = alloca i32, align 4
|
|
; CHECK-NEXT: [[A:%.*]] = alloca i32, align 4
|
|
; CHECK-NEXT: [[B:%.*]] = alloca i32, align 4
|
|
; CHECK-NEXT: store i32 [[ARG]], ptr [[ARG_ADDR]], align 4
|
|
; CHECK-NEXT: store i32 0, ptr [[B]], align 4
|
|
; CHECK-NEXT: [[TMP0:%.*]] = load i32, ptr [[ARG_ADDR]], align 4
|
|
; CHECK-NEXT: [[CMP:%.*]] = icmp sgt i32 [[TMP0]], 0
|
|
; CHECK-NEXT: br i1 [[CMP]], label [[IF_THEN:%.*]], label [[IF_END:%.*]]
|
|
; CHECK: if.then:
|
|
; CHECK-NEXT: store i32 1, ptr [[A]], align 4
|
|
; CHECK-NEXT: br label [[IF_END]]
|
|
; CHECK: if.end:
|
|
; CHECK-NEXT: [[TMP1:%.*]] = load i32, ptr [[A]], align 4
|
|
; CHECK-NEXT: [[TMP2:%.*]] = load i32, ptr [[B]], align 4
|
|
; CHECK-NEXT: [[ADD:%.*]] = add nsw i32 [[TMP1]], [[TMP2]]
|
|
; CHECK-NEXT: ret i32 [[ADD]]
|
|
;
|
|
;
|
|
; CHECK-LABEL: define dso_local i32 @g
|
|
; CHECK-SAME: (i32 noundef [[B:%.*]]) #[[ATTR0]] !type [[META2]] !type [[META3]] {
|
|
; CHECK-NEXT: entry:
|
|
; CHECK-NEXT: [[B_ADDR:%.*]] = alloca i32, align 4
|
|
; CHECK-NEXT: [[FP:%.*]] = alloca ptr, align 8
|
|
; CHECK-NEXT: store i32 [[B]], ptr [[B_ADDR]], align 4
|
|
; CHECK-NEXT: [[TMP0:%.*]] = load i32, ptr [[B_ADDR]], align 4
|
|
; CHECK-NEXT: [[TOBOOL:%.*]] = icmp ne i32 [[TMP0]], 0
|
|
; CHECK-NEXT: [[TMP1:%.*]] = zext i1 [[TOBOOL]] to i64
|
|
; CHECK-NEXT: [[COND:%.*]] = select i1 [[TOBOOL]], ptr @f, ptr @f_thunk
|
|
; CHECK-NEXT: store ptr [[COND]], ptr [[FP]], align 8
|
|
; CHECK-NEXT: [[TMP2:%.*]] = load ptr, ptr [[FP]], align 8
|
|
; CHECK-NEXT: [[TMP3:%.*]] = call i1 @llvm.type.test(ptr [[TMP2]], metadata !"_ZTSFiiE"), !nosanitize [[META4:![0-9]+]]
|
|
; CHECK-NEXT: br i1 [[TMP3]], label [[CONT:%.*]], label [[TRAP:%.*]], !nosanitize [[META4]]
|
|
; CHECK: trap:
|
|
; CHECK-NEXT: call void @llvm.ubsantrap(i8 2) #[[ATTR3:[0-9]+]], !nosanitize [[META4]]
|
|
; CHECK-NEXT: unreachable, !nosanitize [[META4]]
|
|
; CHECK: cont:
|
|
; CHECK-NEXT: [[TMP4:%.*]] = load i32, ptr [[B_ADDR]], align 4
|
|
; CHECK-NEXT: [[CALL:%.*]] = call i32 [[TMP2]](i32 noundef [[TMP4]])
|
|
; CHECK-NEXT: ret i32 [[CALL]]
|
|
;
|
|
;
|
|
; CHECK-LABEL: define dso_local i32 @f_thunk
|
|
; CHECK-SAME: (i32 noundef [[TMP0:%.*]]) #[[ATTR0]] !type [[META2]] !type [[META3]] {
|
|
; CHECK-NEXT: [[TMP2:%.*]] = tail call i32 @f(i32 noundef [[TMP0]]) #[[ATTR0]]
|
|
; CHECK-NEXT: ret i32 [[TMP2]]
|
|
;
|
|
;
|
|
; LOWERTYPETESTS-LABEL: define dso_local i32 @f
|
|
; LOWERTYPETESTS-SAME: (i32 noundef [[ARG:%.*]]) #[[ATTR0:[0-9]+]] !type [[META2:![0-9]+]] !type [[META3:![0-9]+]] {
|
|
; LOWERTYPETESTS-NEXT: entry:
|
|
; LOWERTYPETESTS-NEXT: [[ARG_ADDR:%.*]] = alloca i32, align 4
|
|
; LOWERTYPETESTS-NEXT: [[A:%.*]] = alloca i32, align 4
|
|
; LOWERTYPETESTS-NEXT: [[B:%.*]] = alloca i32, align 4
|
|
; LOWERTYPETESTS-NEXT: store i32 [[ARG]], ptr [[ARG_ADDR]], align 4
|
|
; LOWERTYPETESTS-NEXT: store i32 0, ptr [[B]], align 4
|
|
; LOWERTYPETESTS-NEXT: [[TMP0:%.*]] = load i32, ptr [[ARG_ADDR]], align 4
|
|
; LOWERTYPETESTS-NEXT: [[CMP:%.*]] = icmp sgt i32 [[TMP0]], 0
|
|
; LOWERTYPETESTS-NEXT: br i1 [[CMP]], label [[IF_THEN:%.*]], label [[IF_END:%.*]]
|
|
; LOWERTYPETESTS: if.then:
|
|
; LOWERTYPETESTS-NEXT: store i32 1, ptr [[A]], align 4
|
|
; LOWERTYPETESTS-NEXT: br label [[IF_END]]
|
|
; LOWERTYPETESTS: if.end:
|
|
; LOWERTYPETESTS-NEXT: [[TMP1:%.*]] = load i32, ptr [[A]], align 4
|
|
; LOWERTYPETESTS-NEXT: [[TMP2:%.*]] = load i32, ptr [[B]], align 4
|
|
; LOWERTYPETESTS-NEXT: [[ADD:%.*]] = add nsw i32 [[TMP1]], [[TMP2]]
|
|
; LOWERTYPETESTS-NEXT: ret i32 [[ADD]]
|
|
;
|
|
;
|
|
; LOWERTYPETESTS-LABEL: define dso_local i32 @g
|
|
; LOWERTYPETESTS-SAME: (i32 noundef [[B:%.*]]) #[[ATTR0]] !type [[META2]] !type [[META3]] {
|
|
; LOWERTYPETESTS-NEXT: entry:
|
|
; LOWERTYPETESTS-NEXT: [[B_ADDR:%.*]] = alloca i32, align 4
|
|
; LOWERTYPETESTS-NEXT: [[FP:%.*]] = alloca ptr, align 8
|
|
; LOWERTYPETESTS-NEXT: store i32 [[B]], ptr [[B_ADDR]], align 4
|
|
; LOWERTYPETESTS-NEXT: [[TMP0:%.*]] = load i32, ptr [[B_ADDR]], align 4
|
|
; LOWERTYPETESTS-NEXT: [[TOBOOL:%.*]] = icmp ne i32 [[TMP0]], 0
|
|
; LOWERTYPETESTS-NEXT: [[TMP1:%.*]] = zext i1 [[TOBOOL]] to i64
|
|
; LOWERTYPETESTS-NEXT: [[COND:%.*]] = select i1 [[TOBOOL]], ptr @.cfi.jumptable, ptr getelementptr inbounds ([2 x [8 x i8]], ptr @.cfi.jumptable, i64 0, i64 1)
|
|
; LOWERTYPETESTS-NEXT: store ptr [[COND]], ptr [[FP]], align 8
|
|
; LOWERTYPETESTS-NEXT: [[TMP2:%.*]] = load ptr, ptr [[FP]], align 8
|
|
; LOWERTYPETESTS-NEXT: [[TMP3:%.*]] = ptrtoint ptr [[TMP2]] to i64
|
|
; LOWERTYPETESTS-NEXT: [[TMP4:%.*]] = sub i64 ptrtoint (ptr getelementptr (i8, ptr @.cfi.jumptable, i64 8) to i64), [[TMP3]]
|
|
; LOWERTYPETESTS-NEXT: [[TMP5:%.*]] = call i64 @llvm.fshr.i64(i64 [[TMP4]], i64 [[TMP4]], i64 3)
|
|
; LOWERTYPETESTS-NEXT: [[TMP6:%.*]] = icmp ule i64 [[TMP5]], 1
|
|
; LOWERTYPETESTS-NEXT: br i1 [[TMP6]], label [[CONT:%.*]], label [[TRAP:%.*]], !nosanitize [[META4:![0-9]+]]
|
|
; LOWERTYPETESTS: trap:
|
|
; LOWERTYPETESTS-NEXT: call void @llvm.ubsantrap(i8 2) #[[ATTR4:[0-9]+]], !nosanitize [[META4]]
|
|
; LOWERTYPETESTS-NEXT: unreachable, !nosanitize [[META4]]
|
|
; LOWERTYPETESTS: cont:
|
|
; LOWERTYPETESTS-NEXT: [[TMP7:%.*]] = load i32, ptr [[B_ADDR]], align 4
|
|
; LOWERTYPETESTS-NEXT: [[CALL:%.*]] = call i32 [[TMP2]](i32 noundef [[TMP7]])
|
|
; LOWERTYPETESTS-NEXT: ret i32 [[CALL]]
|
|
;
|
|
;
|
|
; LOWERTYPETESTS-LABEL: define dso_local i32 @f_thunk
|
|
; LOWERTYPETESTS-SAME: (i32 noundef [[TMP0:%.*]]) #[[ATTR0]] !type [[META2]] !type [[META3]] {
|
|
; LOWERTYPETESTS-NEXT: [[TMP2:%.*]] = tail call i32 @f(i32 noundef [[TMP0]]) #[[ATTR0]]
|
|
; LOWERTYPETESTS-NEXT: ret i32 [[TMP2]]
|
|
;
|
|
;
|
|
; LOWERTYPETESTS-LABEL: define private void @.cfi.jumptable
|
|
; LOWERTYPETESTS-SAME: () #[[ATTR3:[0-9]+]] align 8 {
|
|
; LOWERTYPETESTS-NEXT: entry:
|
|
; LOWERTYPETESTS-NEXT: call void asm sideeffect "jmp ${0:c}@plt\0Aint3\0Aint3\0Aint3\0A", "s"(ptr @f)
|
|
; LOWERTYPETESTS-NEXT: call void asm sideeffect "jmp ${0:c}@plt\0Aint3\0Aint3\0Aint3\0A", "s"(ptr @f_thunk)
|
|
; LOWERTYPETESTS-NEXT: unreachable
|
|
;
|