RedDocMD added a comment.
In D98726#2721228 <https://reviews.llvm.org/D98726#2721228>, @NoQ wrote:
>> when the visitor encounters an `ExplodedNode`
>
> Weird. `finalizeVisitor()` accepts not any node but the error node. Your
> screenshot suggests that the error node is not in the standard library but in
> user code. Might it be that there are multiple error nodes and you're looking
> at the wrong one? As usual, you can set conditional breakpoints by node IDs.
Okay no you are right. I was looking at the wrong error node. There are two.
Lets take the following code:
#include <memory>
int foo() {
std::unique_ptr<int> P(new int(10));
int *raw = new int(13);
int b = *raw;
return b;
}
There is clearly a leak but that is getting squelched. Now for the two error
nodes:
- This is the error node we expect, it points to our source code.
F16513445: Screenshot 2021-04-28 at 19-14-34 Screenshot.png
<https://reviews.llvm.org/F16513445>
- This is another error, somewhere in the `unique_ptr.h` file.
F16513472: Screenshot 2021-04-28 at 19-15-04 Screenshot.png
<https://reviews.llvm.org/F16513472>
The second one is the one that causes the first one to get squelched. I suppose
how this point was reached was due to the `StaticAnalyzer` trying to reason
about the destructor of `unique_ptr`, which is implicitly called in at the end
of the function.
For reference, here <https://termbin.com/q1v0> is the dot file for the exploded
graph dump.
Repository:
rG LLVM Github Monorepo
CHANGES SINCE LAST ACTION
https://reviews.llvm.org/D98726/new/
https://reviews.llvm.org/D98726
_______________________________________________
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
