[PATCH] D26559: Insert a type check before reading vtable.

Ivan Krasin via cfe-commits cfe-commits at lists.llvm.org
Tue Nov 15 18:31:56 PST 2016


Okay, I will keep it as a separate file then. It's very likely that I will
soon send a follow up, as there's another very similar case known to not
work properly (based on the experience with UBSan Vptr bot in Chromium)

On Tue, Nov 15, 2016 at 6:27 PM, Richard Smith <richard at metafoo.co.uk>
wrote:

> On Tue, Nov 15, 2016 at 6:20 PM, Ivan Krasin <krasin at google.com> wrote:
>
>> Thank you, Richard.
>>
>> Shall I merge the newly introduced test/CodeGenCXX/ubsan-vtable-checks.cpp
>> into catch-undef-behavior.cpp or it's more clear when it's standalone?
>>
>
> Up to you. catch-undef-behavior.cpp is getting unwieldy, so a separate
> test file doesn't seem like a bad thing.
>
>
>> On Tue, Nov 15, 2016 at 5:51 PM, Richard Smith <richard at metafoo.co.uk>
>> wrote:
>>
>>> On Fri, Nov 11, 2016 at 3:02 PM, Ivan Krasin via cfe-commits <
>>> cfe-commits at lists.llvm.org> wrote:
>>>
>>>> krasin added a comment.
>>>>
>>>> Small correction: all UBSan type checks tests live in compiler-rt.
>>>
>>>
>>> Actually, most of the UBSan tests live in test/CodeGenCXX/catch-undef-behavior.cpp
>>> in Clang; the compiler-rt tests are merely aiming to test that the runtime
>>> produces the correct diagnostics. There are a few more test files testing
>>> UBSan besides that one; you can find the tests by grepping for "fsanitize="
>>> in Clang's test/ directory.
>>>
>>>
>>>> There is a test for UBsan + devirtualization inside tools/clang. My
>>>> point still stands.
>>>>
>>>>
>>>> https://reviews.llvm.org/D26559
>>>>
>>>>
>>>>
>>>> _______________________________________________
>>>> cfe-commits mailing list
>>>> cfe-commits at lists.llvm.org
>>>> http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
>>>>
>>>
>>>
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/cfe-commits/attachments/20161115/1ab3fc39/attachment-0001.html>


More information about the cfe-commits mailing list