-
Notifications
You must be signed in to change notification settings - Fork 15.3k
One more fix for P3144R2 implementation #149406
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.
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
base: main
Are you sure you want to change the base?
Changes from 3 commits
94e652b
a2123f9
69a65b2
2f751bc
6c21226
eff9ccb
268265f
d9449b2
File filter
Filter by extension
Conversations
Jump to
Diff view
Diff view
There are no files selected for viewing
| Original file line number | Diff line number | Diff line change |
|---|---|---|
|
|
@@ -224,11 +224,11 @@ void bad_deletes() | |
| delete 0; // expected-error {{cannot delete expression of type 'int'}} | ||
| delete [0] (int*)0; // expected-error {{expected variable name or 'this' in lambda capture list}} | ||
| delete (void*)0; | ||
| // cxx98-23-warning@-1 {{cannot delete expression with pointer-to-'void' type 'void *'}} | ||
| // since-cxx26-error@-2 {{cannot delete pointer to incomplete type 'void'}} | ||
| // cxx98-23-warning@-1 {{cannot delete pointer to 'void'}} | ||
| // since-cxx26-error@-2 {{cannot delete pointer to 'void'}} | ||
| delete (T*)0; | ||
| // cxx98-23-warning@-1 {{deleting pointer to incomplete type}} | ||
| // since-cxx26-error@-2 {{cannot delete pointer to incomplete type 'T'}} | ||
| // cxx98-23-warning@-1 {{deleting pointer to incomplete struct 'T'}} | ||
| // since-cxx26-error@-2 {{cannot delete pointer to incomplete struct 'T'}} | ||
| ::S::delete (int*)0; // expected-error {{expected unqualified-id}} | ||
| } | ||
|
|
||
|
|
@@ -570,8 +570,8 @@ namespace DeleteIncompleteClassPointerError { | |
| struct A; // expected-note {{forward declaration}} | ||
| void f(A *x) { 1+delete x; } | ||
| // expected-error@-1 {{invalid operands to binary expression}} | ||
| // cxx98-23-warning@-2 {{deleting pointer to incomplete type}} | ||
| // since-cxx26-error@-3 {{cannot delete pointer to incomplete type 'A'}} | ||
| // cxx98-23-warning@-2 {{deleting pointer to incomplete struct 'A'}} | ||
| // since-cxx26-error@-3 {{cannot delete pointer to incomplete struct 'A'}} | ||
| } | ||
|
|
||
| namespace PR10504 { | ||
|
|
@@ -595,6 +595,10 @@ struct GH99278_2 { | |
| } f; | ||
| }; | ||
| GH99278_2<void> e; | ||
| void GH99278_3(int(*p)[]) { | ||
| delete p; | ||
| // expected-warning@-1 {{'delete' applied to a pointer-to-array type 'int (*)[]' treated as 'delete[]'}} | ||
| }; | ||
|
Collaborator
There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. Maybe also test deleting an incomplete array of a complete class type?
Contributor
Author
There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. Hmm, that made me think, what about a pointer to an array of incomplete class type? If that's done through Comparing to what GCC does, GCC doesn't accept this example (even with The example does not appear to violate any rule (at least when the pointer is a null pointer) and clearly was not intended to be banned by P3144R2, yet at the same time I am getting the feeling it never should have been allowed.
Contributor
Author
There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. What EDG does (thanks Compiler Explorer for having that available) is allow it for arrays of complete class type, and disallow it for arrays of incomplete class type. That seems like a sensible way of going about it, it effectively interprets "If the object being deleted has incomplete class type at the point of deletion, the program is ill-formed" as applying also to subobjects of the object being deleted. Which is not what the standard says, but necessary for what P3144R2 aimed to accomplish, so I'll see if I can implement that.
Contributor
Author
There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. That turned out to be a trivial one-line change, so I've done it and added tests for it.
Collaborator
There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. The general issue here is that there's a hole in the standard: a non-array new-expression can't return a pointer to an array, so [expr.delete]p2 effectively says it's always undefined behavior to delete a pointer to an array (unless it's null). Existing compilers treat this as if you wrote Probably we should ask the committee to address this. That said, given the current state of things, this patch seems fine. On a sort of related note, the following currently crashes in codegen:
Contributor
Author
There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more.
True for the tests that I added with struct S { ~S(); };
void del(S(*p)[]) { delete[] p; }
void test() { del(new S[4][4]); }The checks that I implemented in this PR handle this, they allow it in that test, but reject it if The codegen crash is worrying, the test I'm showing here shows that this can occur in legitimate code. I'll have a look at that when I have some extra time.
Contributor
Author
There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. I've changed the array tests to use
Contributor
Author
There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. For the codegen crash, I have opened #150359 |
||
| #endif | ||
|
|
||
| struct PlacementArg {}; | ||
|
|
||
Uh oh!
There was an error while loading. Please reload this page.