shylie/llvm-project
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
llvm-project/llvm/test/Transforms/MemCpyOpt/preserve-memssa.ll
Nikita Popov 92c55a315e
[IR] Only allow lifetime.start/end on allocas (#149310) 
lifetime.start and lifetime.end are primarily intended for use on
allocas, to enable stack coloring and other liveness optimizations. This
is necessary because all (static) allocas are hoisted into the entry
block, so lifetime markers are the only way to convey the actual
lifetimes.

However, lifetime.start and lifetime.end are currently *allowed* to be
used on non-alloca pointers. We don't actually do this in practice, but
just the mere fact that this is possible breaks the core purpose of the
lifetime markers, which is stack coloring of allocas. Stack coloring can
only work correctly if all lifetime markers for an alloca are
analyzable.

* If a lifetime marker may operate on multiple allocas via a select/phi,
we don't know which lifetime actually starts/ends and handle it
incorrectly (https://github.com/llvm/llvm-project/issues/104776).
* Stack coloring operates on the assumption that all lifetime markers
are visible, and not, for example, hidden behind a function call or
escaped pointer. It's not possible to change this, as part of the
purpose of lifetime markers is that they work even in the presence of
escaped pointers, where simple use analysis is insufficient.

I don't think there is any way to have coherent semantics for lifetime
markers on allocas, while also permitting them on arbitrary pointer
values.

This PR restricts lifetimes to operate on allocas only. As a followup, I
will also drop the size argument, which is superfluous if we always
operate on an alloca. (This change also renders various code handling
lifetime markers on non-alloca dead. I plan to clean up that kind of
code after dropping the size argument as well.)

In practice, I've only found a few places that currently produce
lifetimes on non-allocas:

* CoroEarly replaces the promise alloca with the result of an intrinsic,
which will later be replaced back with an alloca. I think this is the
only place where there is some legitimate loss of functionality, but I
don't think this is particularly important (I don't think we'd expect
the promise in a coroutine to admit useful lifetime optimization.)
* SafeStack moves unsafe allocas onto a separate frame. We can safely
drop lifetimes here, as SafeStack performs its own stack coloring.
* Similar for AddressSanitizer, it also moves allocas into separate
memory.
* LSR sometimes replaces the lifetime argument with a GEP chain of the
alloca (where the offsets ultimately cancel out). This is just
unnecessary. (Fixed separately in
https://github.com/llvm/llvm-project/pull/149492.)
* InferAddrSpaces sometimes makes lifetimes operate on an addrspacecast
of an alloca. I don't think this is necessary.
2025-07-21 15:04:50 +02:00

147 lines
5.6 KiB
LLVM
Raw Blame History

; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
; RUN: opt -aa-pipeline=basic-aa -passes='require<memoryssa>,memcpyopt' -verify-memoryssa -S %s | FileCheck %s
; REQUIRES: asserts
target datalayout = "e-m:o-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128"
target triple = "x86_64-apple-macosx10.15.0"
%t = type <{ ptr, [4 x i8], ptr, ptr, i32, [8192 x i8] }>
define i32 @test1(ptr %ptr) {
; CHECK-LABEL: @test1(
; CHECK-NEXT: invoke.cont6:
; CHECK-NEXT: [[P_2:%.*]] = getelementptr inbounds [[T:%.*]], ptr [[PTR:%.*]], i64 0, i32 4
; CHECK-NEXT: [[P_3:%.*]] = getelementptr inbounds [[T]], ptr [[PTR]], i64 0, i32 5, i64 0
; CHECK-NEXT: call void @llvm.memset.p0.i64(ptr align 8 [[PTR]], i8 0, i64 20, i1 false)
; CHECK-NEXT: call void @llvm.memset.p0.i64(ptr align 8 [[P_2]], i8 0, i64 8195, i1 false)
; CHECK-NEXT: ret i32 0
;
invoke.cont6:
call void @llvm.memset.p0.i64(ptr %ptr, i8 0, i64 20, i1 false)
store ptr null, ptr %ptr, align 8
%p.2 = getelementptr inbounds %t, ptr %ptr, i64 0, i32 4
store i32 0, ptr %p.2, align 8
%p.3 = getelementptr inbounds %t, ptr %ptr, i64 0, i32 5, i64 0
call void @llvm.memset.p0.i64(ptr %p.3, i8 0, i64 8191, i1 false)
ret i32 0
}
declare ptr @get_ptr()
define void @test2(ptr noalias %in) {
; CHECK-LABEL: @test2(
; CHECK-NEXT: entry:
; CHECK-NEXT: [[CALL_I1_I:%.*]] = tail call ptr @get_ptr()
; CHECK-NEXT: tail call void @llvm.memcpy.p0.p0.i64(ptr [[CALL_I1_I]], ptr [[IN:%.*]], i64 10, i1 false)
; CHECK-NEXT: ret void
;
entry:
%call.i1.i = tail call ptr @get_ptr()
tail call void @llvm.memset.p0.i64(ptr %call.i1.i, i8 0, i64 10, i1 false)
tail call void @llvm.memcpy.p0.p0.i64(ptr %call.i1.i, ptr %in, i64 10, i1 false)
ret void
}
declare ptr @malloc(i64)
define i32 @test3(ptr noalias %in) {
; CHECK-LABEL: @test3(
; CHECK-NEXT: [[CALL_I_I_I:%.*]] = tail call ptr @malloc(i64 20)
; CHECK-NEXT: tail call void @llvm.memcpy.p0.p0.i64(ptr [[CALL_I_I_I]], ptr [[IN:%.*]], i64 20, i1 false)
; CHECK-NEXT: ret i32 10
;
%call.i.i.i = tail call ptr @malloc(i64 20)
tail call void @llvm.memmove.p0.p0.i64(ptr %call.i.i.i, ptr %in, i64 20, i1 false)
ret i32 10
}
define void @test4(i32 %n, ptr noalias %ptr.0, ptr noalias %ptr.1, ptr %ptr.2) unnamed_addr {
; CHECK-LABEL: @test4(
; CHECK-NEXT: [[ELEM_I:%.*]] = getelementptr i8, ptr [[PTR_0:%.*]], i64 8
; CHECK-NEXT: store i32 [[N:%.*]], ptr [[PTR_2:%.*]], align 8
; CHECK-NEXT: call void @llvm.memcpy.p0.p0.i64(ptr [[ELEM_I]], ptr [[PTR_1:%.*]], i64 10, i1 false)
; CHECK-NEXT: ret void
;
%elem.i = getelementptr i8, ptr %ptr.0, i64 8
call void @llvm.memset.p0.i64(ptr %elem.i, i8 0, i64 10, i1 false)
store i32 %n, ptr %ptr.2, align 8
call void @llvm.memcpy.p0.p0.i64(ptr %elem.i, ptr %ptr.1, i64 10, i1 false)
ret void
}
declare void @decompose(ptr nocapture)
define void @test5(ptr %ptr) {
; CHECK-LABEL: @test5(
; CHECK-NEXT: entry:
; CHECK-NEXT: [[EARLY_DATA:%.*]] = alloca [128 x i8], align 8
; CHECK-NEXT: [[TMP:%.*]] = alloca [[T:%.*]], align 8
; CHECK-NEXT: call void @llvm.lifetime.start.p0(i64 32, ptr [[EARLY_DATA]])
; CHECK-NEXT: [[TMP0:%.*]] = load i32, ptr [[PTR:%.*]], align 8
; CHECK-NEXT: call fastcc void @decompose(ptr [[TMP]])
; CHECK-NEXT: call void @llvm.memcpy.p0.p0.i64(ptr [[EARLY_DATA]], ptr [[TMP]], i64 32, i1 false)
; CHECK-NEXT: ret void
;
entry:
%early_data = alloca [128 x i8], align 8
%tmp = alloca %t, align 8
call void @llvm.lifetime.start.p0(i64 32, ptr %early_data)
%0 = load i32, ptr %ptr, align 8
call fastcc void @decompose(ptr %tmp)
call void @llvm.memcpy.p0.p0.i64(ptr %early_data, ptr %tmp, i64 32, i1 false)
ret void
}
define void @test7(ptr %ptr) {
; CHECK-LABEL: @test7(
; CHECK-NEXT: entry:
; CHECK-NEXT: [[TMP0:%.*]] = getelementptr inbounds [4 x i32], ptr [[PTR:%.*]], i64 0, i32 1
; CHECK-NEXT: [[TMP1:%.*]] = getelementptr inbounds [4 x i32], ptr [[PTR]], i64 0, i32 2
; CHECK-NEXT: [[TMP2:%.*]] = getelementptr inbounds [4 x i32], ptr [[PTR]], i64 0, i32 3
; CHECK-NEXT: call void @llvm.memset.p0.i64(ptr align 1 [[PTR]], i8 0, i64 16, i1 false)
; CHECK-NEXT: call void @clobber()
; CHECK-NEXT: ret void
;
entry:
store i32 0, ptr %ptr, align 1
%0 = getelementptr inbounds [4 x i32], ptr %ptr, i64 0, i32 1
store i32 0, ptr %0, align 1
%1 = getelementptr inbounds [4 x i32], ptr %ptr, i64 0, i32 2
store i32 0, ptr %1, align 1
%2 = getelementptr inbounds [4 x i32], ptr %ptr, i64 0, i32 3
store i32 0, ptr %2, align 1
call void @clobber()
ret void
}
define void @test8(ptr noalias %src, ptr %dst) {
; CHECK-LABEL: @test8(
; CHECK-NEXT: call void @llvm.memcpy.p0.p0.i64(ptr align 1 [[DST:%.*]], ptr align 1 [[SRC:%.*]], i64 8224, i1 false)
; CHECK-NEXT: call void @llvm.memset.p0.i64(ptr align 1 [[SRC]], i8 0, i64 8224, i1 false)
; CHECK-NEXT: ret void
;
%1 = load %t, ptr %src
store %t zeroinitializer, ptr %src
store %t %1, ptr %dst
ret void
}
declare void @clobber()
; Function Attrs: argmemonly nounwind willreturn
declare void @llvm.lifetime.start.p0(i64 immarg, ptr nocapture) #0
; Function Attrs: argmemonly nounwind willreturn
declare void @llvm.memcpy.p0.p0.i64(ptr noalias nocapture writeonly, ptr noalias nocapture readonly, i64, i1 immarg) #0
; Function Attrs: argmemonly nounwind willreturn writeonly
declare void @llvm.memset.p0.i64(ptr nocapture writeonly, i8, i64, i1 immarg) #1
; Function Attrs: argmemonly nounwind willreturn
declare void @llvm.memmove.p0.p0.i64(ptr nocapture, ptr nocapture readonly, i64, i1 immarg) #0
attributes #0 = { argmemonly nounwind willreturn }
attributes #1 = { argmemonly nounwind willreturn writeonly }
Reference in New Issue View Git Blame Copy Permalink