Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

[msan] Unpoison indirect outputs for userspace using memset for large operands #79924

Merged
merged 3 commits into from
Jan 30, 2024
Merged

[msan] Unpoison indirect outputs for userspace using memset for large operands #79924

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
07043d2
[𝘀𝗽𝗿] initial version
MaskRay Jan 30, 2024
d0bd88e
use StoreInst when size<=32, similar to -ftrivial-auto-init=
MaskRay Jan 30, 2024
d4d4c73
improve test and comment
MaskRay Jan 30, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
12 changes: 9 additions & 3 deletions llvm/lib/Transforms/Instrumentation/MemorySanitizer.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4552,16 +4552,22 @@ struct MemorySanitizerVisitor : public InstVisitor<MemorySanitizerVisitor> {
}
if (!ElemTy->isSized())
return;
Value *SizeVal =
IRB.CreateTypeSize(MS.IntptrTy, DL.getTypeStoreSize(ElemTy));
auto Size = DL.getTypeStoreSize(ElemTy);
Value *SizeVal = IRB.CreateTypeSize(MS.IntptrTy, Size);
if (MS.CompileKernel) {
IRB.CreateCall(MS.MsanInstrumentAsmStoreFn, {Operand, SizeVal});
} else {
// ElemTy, derived from elementtype(), does not encode the alignment of
// the pointer. Conservatively assume that the shadow memory is unaligned.
// When Size is large, avoid StoreInst as it would expand to many
// instructions.
auto [ShadowPtr, _] =
getShadowOriginPtrUserspace(Operand, IRB, IRB.getInt8Ty(), Align(1));
IRB.CreateAlignedStore(getCleanShadow(ElemTy), ShadowPtr, Align(1));
if (Size <= 32)
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Maybe some comment wouldn't hurt here to clarify this constant.

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This is a bit arbitrary. Perhaps

// The size threshold matches shouldUseBZeroPlusStoresToInitialize for -ftrivial-auto-var-init=zero?

Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Memset will go into interceptor and do a lot of check, see __msan_memset, so it's more expensive than a regular memset. I can't tell what can be optimal constant here, but I would expect something larger. But I doubt it will make a meaningful difference.

Copy link
Member Author

@MaskRay MaskRay Feb 3, 2024 •
edited
Loading

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thanks for your previous comment about the interceptor. The committed patch does contain this description:
"The intrinsic, if lowered to libcall, will use the msan interceptor."

Inline asm isn't commonly used:) This patch is for =m in extended asm, which I believe is more uncommon, and if used, usually with small objects. I guess 32 and 64 won't make a difference.

IRB.CreateAlignedStore(getCleanShadow(ElemTy), ShadowPtr, Align(1));
else
IRB.CreateMemSet(ShadowPtr, ConstantInt::getNullValue(IRB.getInt8Ty()),
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

it will hit interceptor

Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Never mind, we do this all the time here.

SizeVal, Align(1));
}
}

Expand Down
14 changes: 13 additions & 1 deletion llvm/test/Instrumentation/MemorySanitizer/msan_asm_conservative.ll
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -14,6 +14,7 @@ target datalayout = "e-m:e-i64:64-f80:128-n8:16:32:64-S128"
target triple = "x86_64-unknown-linux-gnu"

%struct.pair = type { i32, i32 }
%struct.large = type { [33 x i8] }

@id1 = common dso_local global i32 0, align 4
@is1 = common dso_local global i32 0, align 4
Expand All @@ -28,6 +29,7 @@ target triple = "x86_64-unknown-linux-gnu"
@memcpy_d2 = common dso_local global ptr null, align 8
@memcpy_s1 = common dso_local global ptr null, align 8
@memcpy_s2 = common dso_local global ptr null, align 8
@large = common dso_local global %struct.pair zeroinitializer, align 4

; The functions below were generated from a C source that contains declarations like follows:
; void f1() {
Expand Down Expand Up @@ -201,7 +203,6 @@ entry:
; CHECK: call void @__msan_warning
; CHECK: call i32 asm "", "=r,=*m,r,*m,~{dirflag},~{fpsr},~{flags}"(ptr elementtype(i32) @id1, i32 [[IS1_F7]], ptr elementtype(i32) @is1)


; Three outputs, first and last returned via regs, second via mem:
; asm("" : "=r" (id1), "=m"(id2), "=r" (id3):);
define dso_local void @f_3o_reg_mem_reg() sanitize_memory {
Expand Down Expand Up @@ -265,6 +266,17 @@ entry:
; CHECK-CONS: call void @__msan_instrument_asm_store({{.*}}@memcpy_d1{{.*}}, i64 8)
; CHECK: call void asm "", "=*m,=*m,=*m,*m,*m,*m,~{dirflag},~{fpsr},~{flags}"(ptr elementtype(%struct.pair) @pair2, ptr elementtype(i8) @c2, ptr elementtype(ptr) @memcpy_d1, ptr elementtype(%struct.pair) @pair1, ptr elementtype(i8) @c1, ptr elementtype(ptr) @memcpy_s1)

; Use memset when the size is larger.
define dso_local void @f_1i_1o_mem_large() sanitize_memory {
entry:
%0 = call i32 asm "", "=r,=*m,*m,~{dirflag},~{fpsr},~{flags}"(ptr elementtype(%struct.large) @large, ptr elementtype(%struct.large) @large)
store i32 %0, ptr @id1, align 4
ret void
}
; CHECK-LABEL: @f_1i_1o_mem_large(
; USER-CONS: call void @llvm.memset.p0.i64(ptr align 1 inttoptr (i64 xor (i64 ptrtoint (ptr @large to i64), i64 87960930222080) to ptr), i8 0, i64 33, i1 false)
; CHECK-CONS: call void @__msan_instrument_asm_store({{.*}}@large{{.*}}, i64 33)
; CHECK: call i32 asm "", "=r,=*m,*m,~{dirflag},~{fpsr},~{flags}"(ptr elementtype(%struct.large) @large, ptr elementtype(%struct.large) @large)

; A simple asm goto construct to check that callbr is handled correctly:
; int asm_goto(int n) {
Expand Down
Loading