seanm wrote:
@bwendling thanks for your reply. No idea how representative it is of other
projects, but 3 of the 6 flexible arrays in
[libusb](https://github.com/libusb/libusb) structures have these kinds of
non-trivial counts.
The GCC folks seem to have
[considered](https://gcc.gnu.org/bugzi
bwendling wrote:
> @bwendling is there any plan / possibility for simple expressions (with no
> side effects)? Like:
>
> ```c
> struct libusb_bos_dev_capability_descriptor {
> uint8_t bLength;
> uint8_t bDescriptorType;
> uint8_t bDevCapabilityType;
> uint8_t dev_cap
seanm wrote:
@bwendling is there any plan / possibility for simple expressions (with no side
effects)? Like:
```c
struct libusb_bos_dev_capability_descriptor {
uint8_t bLength;
uint8_t bDescriptorType;
uint8_t bDevCapabilityType;
uint8_t dev_capability_data[
bwendling wrote:
Resubmitted with fix. Sorry about the failures:
```
To https://github.com/llvm/llvm-project.git
e8790027b169..164f85db876e main -> main
```
https://github.com/llvm/llvm-project/pull/76348
___
cfe-commits mailing list
cfe-commits@l
nico wrote:
Reverted in 2dce77201c0c6b541a53aa7a09ec06e7561e8f74 for now.
https://github.com/llvm/llvm-project/pull/76348
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
dyung wrote:
@bwendling the test you added `CodeGen/attr-counted-by.c` appears to be failing
on quite a few bots, can you take a look?
Some failing bots:
- https://lab.llvm.org/buildbot/#/builders/164/builds/48660
- https://lab.llvm.org/buildbot/#/builders/188/builds/40261
- https://lab.llvm.or
https://github.com/bwendling closed
https://github.com/llvm/llvm-project/pull/76348
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
bwendling wrote:
Thanks everyone for their insightful feedback!!
https://github.com/llvm/llvm-project/pull/76348
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/rapidsna edited
https://github.com/llvm/llvm-project/pull/76348
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/rapidsna approved this pull request.
> Do you have anything to add?
LGTM
https://github.com/llvm/llvm-project/pull/76348
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-comm
bwendling wrote:
@rapidsna & @nickdesaulniers: Do you have anything to add?
https://github.com/llvm/llvm-project/pull/76348
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/kees approved this pull request.
I can't speak to the LLVM code changes, but behaviorally, this passes all the
torture tests I'd expect it to (kernel-tools/fortify/array-bounds.c and full
kernel builds with its hundreds of `counted_by` annotations).
https://github.com/llvm/l
kees wrote:
That's it! Everything works. :) Ship it!
https://github.com/llvm/llvm-project/pull/76348
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
bwendling wrote:
Never mind. I think I found the issue anyway...
https://github.com/llvm/llvm-project/pull/76348
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
bwendling wrote:
@kees Could you supply a pre-processed file and command line? (It should have
been emitted when you got the backtrace.)
https://github.com/llvm/llvm-project/pull/76348
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://li
kees wrote:
Thanks! It's fixed for me now. I continue to be highly unlucky, though, and
keep finding weird stuff. It seems to segfault just parsing libc headers under
`-D_FORTIFY_SOURCE=3` (but not `=2` or lower):
```
0. Program arguments:
/srv/built-compilers/llvm/counted_by/install/bin
github-actions[bot] wrote:
:warning: C/C++ code formatter, clang-format found issues in your code.
:warning:
You can test this locally with the following command:
``bash
git-clang-format --diff 061e4f24b24a3b59d73a94dc6f2f0d21a2b7beac
557604301ab98d89b5ddf6d857d85326d7d3da3d --
bwendling wrote:
Like I said, I can fix the `mi->ints` issue. It's when we're trying to ask for
the size of the whole struct with more than one `__counted_by` attribute that
we don't be able to handle.
https://github.com/llvm/llvm-project/pull/76348
kees wrote:
Accessing `mi->ints` is unambiguous (it would use the declared `count_ints`)
but I'm fine to wait and fix that later. The flex array union is gloriously
rare in the kernel. As for whole object, I say pick smallest from all available
under bdos(x, 1) and largest for bdos(x, 0), or j
bwendling wrote:
> > @kees, the issue should be fixed with the newest push.
>
> Nice! This is so extremely close. It fixed all but 1 instance in the torture
> testing. The remaining seems to be union order sensitive. O_o This
> arrangement still reports non-zero for the `int count_bytes` one,
kees wrote:
> @kees, the issue should be fixed with the newest push.
Nice! This is so extremely close. It fixed all but 1 instance in the torture
testing. The remaining seems to be union order sensitive. O_o This arrangement
still reports non-zero for the `int count_bytes` one, unless I move i
bwendling wrote:
@kees, the issue should be fixed with the newest push.
https://github.com/llvm/llvm-project/pull/76348
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
kees wrote:
> > but the value is nonsense, so we must return 0 so that anything checking
> > lengths will not write anything to the array.
>
> @kees Oh, I see. I did not know such the convention but it makes sense. Is it
> documented somewhere?
This is new territory (having a multiplier for f
rapidsna wrote:
> The return value of -1 means "I don't know how large this is".
@kees Oh, I see. I did not know such the convention. Is it documented somewhere?
https://github.com/llvm/llvm-project/pull/76348
___
cfe-commits mailing list
cfe-commits
kees wrote:
> > This prints a wrapped calculation instead of the expected "0".
>
> Shouldn't `getDefaultBuiltinObjectSizeResult` return `-1`? Have you tried
> `-2` to see if it's returning the negative count value or it happens to be
> equal to the default value?
Well, maybe it's not wrapped,
rapidsna wrote:
> This prints a wrapped calculation instead of the expected "0".
Shouldn't `getDefaultBuiltinObjectSizeResult` return `-1`? Have you tried `-2`
to see if it's returning the negative count value or it happens to be equal to
the default value?
https://github.com/llvm/llvm-projec
kees wrote:
Thanks! The update fixes the anon struct issue I hit. I've found one more
issue, though this appears to be a miscalculation with a pathological `count`
value (i.e. `count` is signed type and contains a negative value):
```
struct annotated {
unsigned long flags;
int count;
@@ -944,22 +951,259 @@ static llvm::Value
*getArrayIndexingBound(CodeGenFunction &CGF,
return nullptr;
}
+namespace {
+
+/// \p StructAccessBase returns the base \p Expr of a field access. It returns
+/// either a \p DeclRefExpr, representing the base pointer to the struct,
@@ -944,22 +951,259 @@ static llvm::Value
*getArrayIndexingBound(CodeGenFunction &CGF,
return nullptr;
}
+namespace {
+
+/// \p StructAccessBase returns the base \p Expr of a field access. It returns
+/// either a \p DeclRefExpr, representing the base pointer to the struct,
@@ -818,6 +819,189 @@ CodeGenFunction::evaluateOrEmitBuiltinObjectSize(const
Expr *E, unsigned Type,
return ConstantInt::get(ResType, ObjectSize, /*isSigned=*/true);
}
+const FieldDecl *CodeGenFunction::FindFlexibleArrayMemberField(
+ASTContext &Ctx, const RecordDecl *R
@@ -944,22 +951,259 @@ static llvm::Value
*getArrayIndexingBound(CodeGenFunction &CGF,
return nullptr;
}
+namespace {
+
+/// \p StructAccessBase returns the base \p Expr of a field access. It returns
+/// either a \p DeclRefExpr, representing the base pointer to the struct,
bwendling wrote:
@kees Should be fixed now.
https://github.com/llvm/llvm-project/pull/76348
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
@@ -944,22 +951,262 @@ static llvm::Value
*getArrayIndexingBound(CodeGenFunction &CGF,
return nullptr;
}
+namespace {
+
+/// \p StructAccessBase returns the base \p Expr of a field access. It returns
+/// either a \p DeclRefExpr, representing the base pointer to the struct,
@@ -818,6 +819,189 @@ CodeGenFunction::evaluateOrEmitBuiltinObjectSize(const
Expr *E, unsigned Type,
return ConstantInt::get(ResType, ObjectSize, /*isSigned=*/true);
}
+const FieldDecl *CodeGenFunction::FindFlexibleArrayMemberField(
+ASTContext &Ctx, const RecordDecl *R
@@ -818,6 +819,189 @@ CodeGenFunction::evaluateOrEmitBuiltinObjectSize(const
Expr *E, unsigned Type,
return ConstantInt::get(ResType, ObjectSize, /*isSigned=*/true);
}
+const FieldDecl *CodeGenFunction::FindFlexibleArrayMemberField(
+ASTContext &Ctx, const RecordDecl *R
@@ -944,22 +951,259 @@ static llvm::Value
*getArrayIndexingBound(CodeGenFunction &CGF,
return nullptr;
}
+namespace {
+
+/// \p StructAccessBase returns the base \p Expr of a field access. It returns
+/// either a \p DeclRefExpr, representing the base pointer to the struct,
kees wrote:
Possibly due to bug #72032 , I can get this tree to crash using the latest
`array-bounds.c` test from
https://github.com/kees/kernel-tools/tree/trunk/fortify
Specifically:
```
struct anon_struct {
unsigned long flags;
long count;
int array[] __counted_by(cou
@@ -818,6 +819,189 @@ CodeGenFunction::evaluateOrEmitBuiltinObjectSize(const
Expr *E, unsigned Type,
return ConstantInt::get(ResType, ObjectSize, /*isSigned=*/true);
}
+const FieldDecl *CodeGenFunction::FindFlexibleArrayMemberField(
+ASTContext &Ctx, const RecordDecl *R
@@ -818,6 +819,189 @@ CodeGenFunction::evaluateOrEmitBuiltinObjectSize(const
Expr *E, unsigned Type,
return ConstantInt::get(ResType, ObjectSize, /*isSigned=*/true);
}
+const FieldDecl *CodeGenFunction::FindFlexibleArrayMemberField(
+ASTContext &Ctx, const RecordDecl *R
@@ -944,22 +951,259 @@ static llvm::Value
*getArrayIndexingBound(CodeGenFunction &CGF,
return nullptr;
}
+namespace {
+
+/// \p StructAccessBase returns the base \p Expr of a field access. It returns
+/// either a \p DeclRefExpr, representing the base pointer to the struct,
@@ -944,22 +951,262 @@ static llvm::Value
*getArrayIndexingBound(CodeGenFunction &CGF,
return nullptr;
}
+namespace {
+
+/// \p StructAccessBase returns the base \p Expr of a field access. It returns
+/// either a \p DeclRefExpr, representing the base pointer to the struct,
@@ -944,22 +951,259 @@ static llvm::Value
*getArrayIndexingBound(CodeGenFunction &CGF,
return nullptr;
}
+namespace {
+
+/// \p StructAccessBase returns the base \p Expr of a field access. It returns
+/// either a \p DeclRefExpr, representing the base pointer to the struct,
@@ -818,6 +819,189 @@ CodeGenFunction::evaluateOrEmitBuiltinObjectSize(const
Expr *E, unsigned Type,
return ConstantInt::get(ResType, ObjectSize, /*isSigned=*/true);
}
+const FieldDecl *CodeGenFunction::FindFlexibleArrayMemberField(
+ASTContext &Ctx, const RecordDecl *R
@@ -818,6 +819,189 @@ CodeGenFunction::evaluateOrEmitBuiltinObjectSize(const
Expr *E, unsigned Type,
return ConstantInt::get(ResType, ObjectSize, /*isSigned=*/true);
}
+const FieldDecl *CodeGenFunction::FindFlexibleArrayMemberField(
+ASTContext &Ctx, const RecordDecl *R
@@ -1155,15 +1159,14 @@ const FieldDecl
*CodeGenFunction::FindCountedByField(const FieldDecl *FD) {
return nullptr;
auto GetNonAnonStructOrUnion = [](const RecordDecl *RD) {
-while (RD && !RD->getDeclName())
- if (const auto *R = dyn_cast(RD->getDeclContext()))
@@ -944,22 +951,259 @@ static llvm::Value
*getArrayIndexingBound(CodeGenFunction &CGF,
return nullptr;
}
+namespace {
+
+/// \p StructAccessBase returns the base \p Expr of a field access. It returns
+/// either a \p DeclRefExpr, representing the base pointer to the struct,
https://github.com/rapidsna commented:
I left some comments. Looking good overall but `LocalDeclMap` doesn't seem to
serve the purpose for this PR. And I'm not sure the code block using it would
be actually necessary. Please see my inlined review.
https://github.com/llvm/llvm-project/pull/7634
@@ -944,22 +951,262 @@ static llvm::Value
*getArrayIndexingBound(CodeGenFunction &CGF,
return nullptr;
}
+namespace {
+
+/// \p StructAccessBase returns the base \p Expr of a field access. It returns
+/// either a \p DeclRefExpr, representing the base pointer to the struct,
@@ -944,22 +951,259 @@ static llvm::Value
*getArrayIndexingBound(CodeGenFunction &CGF,
return nullptr;
}
+namespace {
+
+/// \p StructAccessBase returns the base \p Expr of a field access. It returns
+/// either a \p DeclRefExpr, representing the base pointer to the struct,
@@ -818,6 +819,189 @@ CodeGenFunction::evaluateOrEmitBuiltinObjectSize(const
Expr *E, unsigned Type,
return ConstantInt::get(ResType, ObjectSize, /*isSigned=*/true);
}
+const FieldDecl *CodeGenFunction::FindFlexibleArrayMemberField(
+ASTContext &Ctx, const RecordDecl *R
@@ -818,6 +819,189 @@ CodeGenFunction::evaluateOrEmitBuiltinObjectSize(const
Expr *E, unsigned Type,
return ConstantInt::get(ResType, ObjectSize, /*isSigned=*/true);
}
+const FieldDecl *CodeGenFunction::FindFlexibleArrayMemberField(
+ASTContext &Ctx, const RecordDecl *R
@@ -944,22 +951,259 @@ static llvm::Value
*getArrayIndexingBound(CodeGenFunction &CGF,
return nullptr;
}
+namespace {
+
+/// \p StructAccessBase returns the base \p Expr of a field access. It returns
+/// either a \p DeclRefExpr, representing the base pointer to the struct,
@@ -944,22 +951,259 @@ static llvm::Value
*getArrayIndexingBound(CodeGenFunction &CGF,
return nullptr;
}
+namespace {
+
+/// \p StructAccessBase returns the base \p Expr of a field access. It returns
+/// either a \p DeclRefExpr, representing the base pointer to the struct,
@@ -818,6 +819,189 @@ CodeGenFunction::evaluateOrEmitBuiltinObjectSize(const
Expr *E, unsigned Type,
return ConstantInt::get(ResType, ObjectSize, /*isSigned=*/true);
}
+const FieldDecl *CodeGenFunction::FindFlexibleArrayMemberField(
+ASTContext &Ctx, const RecordDecl *R
@@ -1155,15 +1159,14 @@ const FieldDecl
*CodeGenFunction::FindCountedByField(const FieldDecl *FD) {
return nullptr;
auto GetNonAnonStructOrUnion = [](const RecordDecl *RD) {
-while (RD && !RD->getDeclName())
- if (const auto *R = dyn_cast(RD->getDeclContext()))
https://github.com/rapidsna edited
https://github.com/llvm/llvm-project/pull/76348
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
@@ -850,6 +1034,13 @@ CodeGenFunction::emitBuiltinObjectSize(const Expr *E,
unsigned Type,
}
}
+ if (IsDynamic) {
+// Emit special code for a flexible array member with the "counted_by"
+// attribute.
+if (Value *V = emitFlexibleArrayMemberSize(E, Type, Res
@@ -365,6 +365,10 @@ class DefaultFilterCCC final : public
CorrectionCandidateCallback {
template
class DeclFilterCCC final : public CorrectionCandidateCallback {
public:
+ explicit DeclFilterCCC(const IdentifierInfo *Typo = nullptr,
+ NestedNameSpeci
@@ -818,6 +819,189 @@ CodeGenFunction::evaluateOrEmitBuiltinObjectSize(const
Expr *E, unsigned Type,
return ConstantInt::get(ResType, ObjectSize, /*isSigned=*/true);
}
+const FieldDecl *CodeGenFunction::FindFlexibleArrayMemberField(
+ASTContext &Ctx, const RecordDecl *R
@@ -850,6 +1034,13 @@ CodeGenFunction::emitBuiltinObjectSize(const Expr *E,
unsigned Type,
}
}
+ if (IsDynamic) {
+// Emit special code for a flexible array member with the "counted_by"
+// attribute.
+if (Value *V = emitFlexibleArrayMemberSize(E, Type, Res
@@ -818,6 +819,189 @@ CodeGenFunction::evaluateOrEmitBuiltinObjectSize(const
Expr *E, unsigned Type,
return ConstantInt::get(ResType, ObjectSize, /*isSigned=*/true);
}
+const FieldDecl *CodeGenFunction::FindFlexibleArrayMemberField(
+ASTContext &Ctx, const RecordDecl *R
@@ -365,6 +365,10 @@ class DefaultFilterCCC final : public
CorrectionCandidateCallback {
template
class DeclFilterCCC final : public CorrectionCandidateCallback {
public:
+ explicit DeclFilterCCC(const IdentifierInfo *Typo = nullptr,
+ NestedNameSpeci
bwendling wrote:
Pinging for after-holidays visibility. :-)
https://github.com/llvm/llvm-project/pull/76348
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
bwendling wrote:
Some (hopefully) helpful comments on this admittedly very large patch.
General Comment:
I pulled back on allowing the `count` to be in a non-anonymous struct that
doesn't contain the flexible array. So this is no longer allowed:
```
struct count_not_in_substruct {
unsigned
https://github.com/bwendling edited
https://github.com/llvm/llvm-project/pull/76348
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
llvmbot wrote:
@llvm/pr-subscribers-clang
Author: Bill Wendling (bwendling)
Changes
The 'counted_by' attribute is used on flexible array members. The
argument for the attribute is the name of the field member holding the
count of elements in the flexible array. This information is used to
66 matches
Mail list logo
