On 5/14/21 12:35 AM, Martin Sebor wrote:
> On 5/13/21 11:36 AM, Martin Sebor wrote:
>> On 5/13/21 11:20 AM, Bernd Edlinger wrote:
>>> On 5/13/21 3:55 AM, Martin Sebor via Gcc-patches wrote:
>>>> A logic bug in the handling of PHI arguments in compute_objsize
>>>> that are all null pointers lets an incompletely populated struct
>>>> be used in a way that triggers an assertion causing an ICE.
>>>>
>>>> The attached patch corrects that by having compute_objsize fail
>>>> when the struct isn't fully populated (when all os the PHI's
>>>> arguments are null).
>>>>
>>>> Martin
>>>
>>> Martin,
>>>
>>> I'm getting test failures with your patch here:
>>>
>>> Running target unix/-m32
>>> FAIL: g++.dg/pr100574.C -std=gnu++14 (test for excess errors)
>>> FAIL: g++.dg/pr100574.C -std=gnu++17 (test for excess errors)
>>> FAIL: g++.dg/pr100574.C -std=gnu++2a (test for excess errors)
>>>
>>> /home/ed/gnu/gcc-trunk/gcc/testsuite/g++.dg/pr100574.C:6:7: error:
>>> 'operator new' takes type 'size_t' ('unsigned int') as first parameter
>>> [-fpermissive]^M
>>> compiler exited with status 1
>>
>> Thanks, I've just fixed it.
>
> I hadn't checked in the patch yet. I'm only now about to do it and
> see I inadvertently committed the test in response to your email
> about the failures. I didn't realize you were testing the patch
> I had posted for review, before I committed it.
>
Oh, well.
I just wanted to help, since you didn't tell how you tested the patch.
Bernd.
> Martin
>
>>
>> Martin
>>
>>>
>>>
>>> Bernd.
>>>
>>
>
