Re: [dev] st: Use after free

From: <moosotc_AT_gmail.com>
Date: Mon, 23 Jan 2017 04:43:49 +0300

Hiltjo Posthuma <hiltjo_AT_codemadness.org> writes:

> On Sun, Jan 22, 2017 at 11:00:28PM +0300, moosotc_AT_gmail.com wrote:
>> Martin K├╝hne <mysatyre_AT_gmail.com> writes:
>>
>> > On Sun, Jan 22, 2017 at 5:17 PM, <moosotc_AT_gmail.com> wrote:
>> >>
>> >> `valgrind st -f mono-2 cat full-bmp.txt' [1]
>> >>
>> >> Yields quite a few invalid reads from freed blocks, the issue is related
>> >> to cache management. In the real world those dangling pointer issues
>> >> lead to segfaults or X11 errors (eventually)
>> >>
>> >
>> >
>> > I think your patch might have been lost on the way?
>> >
>>
>> Nope - there was no patch since I don't really understand what's going
>> on, I only have symptoms and reproducer.
>>
>> --
>> mailto:moosotc_AT_gmail.com
>>
>
> Hi,
>
> Not to be mean, but please learn more about Valgrind (and it's caveats)
> before posting to the ML. It is expected to provide a patch fix or at the
> very least give detailed information.

`st cat ~/x/txt/full-bmp.txt -
zsh: segmentation fault (core dumped) st cat ~/x/txt/full-bmp.txt -`

That's after 4-5 runs of st, not the absence of valgrind in the above
command line, with valgrind the splats are predictable and stable,
always pointing to the same stack trace at least.

I would love to provide a patch, but as mentioned before I don't really
understand the code in question.

-- 
mailto:moosotc_AT_gmail.com
Received on Mon Jan 23 2017 - 02:43:49 CET

This archive was generated by hypermail 2.3.0 : Mon Jan 23 2017 - 02:48:13 CET