pan-users
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Pan-users] use-after-free in latest pan.git [FIXED]


From: walt
Subject: Re: [Pan-users] use-after-free in latest pan.git [FIXED]
Date: Sat, 10 May 2014 13:21:16 -0700
User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:24.0) Gecko/20100101 Thunderbird/24.5.0

On 05/10/2014 12:56 AM, Heinrich Müller wrote:
> Am 10.05.2014 00:16, schrieb walt:
>> On 05/09/2014 11:34 AM, Heinrich Müller wrote:
>>> Am 07.05.2014 00:54, schrieb walt:
>>>> On 05/06/2014 11:42 AM, walt wrote:
>>>>> Hi all.  This is a bug I reported some time ago, but I just learned about
>>>>> the gcc -fsanitize=address option and now I have some fresh info, which I
>>>>> don't know enough to interpret. (see attached)
>>>>
>>>> Using addr2line (from binutils) I added some source code annotations at the
>>>> end of each line of debugging info. (see attached)
>>>>
>>
>>> I think I fixed this. (please pull)
>>> Could you try to watch if the error persists or has gone away?
>>
>> Thanks Heinrich :)  I wonder if you forgot to commit some of your changes?
>>
>> I'm now running this version:
>>
>> commit ecb67b1f18dccbe53c95569bc21b80e4daa9dd50
>> Author: Heinrich Müller <address@hidden>
>> Date:   Fri May 9 20:33:17 2014 +0200
>>
>>      fix use-after-free
>>
>> pan: ../../../pan/tasks/nntp.cc:100: virtual bool 
>> pan::NNTP::on_socket_response(pan::Socket*, const pan::StringView&): 
>> Assertion `_listener != 0' failed.
>>
>> Program received signal SIGABRT, Aborted.
>> 0x00007ffff3bc4899 in __GI_raise (address@hidden) at 
>> ../nptl/sysdeps/unix/sysv/linux/raise.c:56
>> 56      ../nptl/sysdeps/unix/sysv/linux/raise.c: No such file or directory.
>> (gdb) bt
>> #0  0x00007ffff3bc4899 in __GI_raise (address@hidden) at 
>> ../nptl/sysdeps/unix/sysv/linux/raise.c:56
>> #1  0x00007ffff3bc5e27 in __GI_abort () at abort.c:89
>> #2  0x00007ffff3bbd846 in __assert_fail_base (fmt=0x7ffff3cf6cd0 "%s%s%s:%u: 
>> %s%sAssertion `%s' failed.\n%n", address@hidden "_listener != 0",
>>      address@hidden "../../../pan/tasks/nntp.cc", address@hidden,
>>      address@hidden <pan::NNTP::on_socket_response(pan::Socket*, 
>> pan::StringView const&)::__PRETTY_FUNCTION__> "virtual bool 
>> pan::NNTP::on_socket_response(pan::Socket*, const pan::StringView&)") at 
>> assert.c:92
>> #3  0x00007ffff3bbd902 in __GI___assert_fail (assertion=0x63b8f7 "_listener 
>> != 0", file=0x63b898 "../../../pan/tasks/nntp.cc", line=100,
>>      function=0x63bc00 <pan::NNTP::on_socket_response(pan::Socket*, 
>> pan::StringView const&)::__PRETTY_FUNCTION__> "virtual bool 
>> pan::NNTP::on_socket_response(pan::Socket*, const pan::StringView&)") at 
>> assert.c:101
>> #4  0x0000000000544883 in pan::NNTP::on_socket_response (this=0xb344a0, 
>> sock=<optimized out>, line_in=...) at ../../../pan/tasks/nntp.cc:100
>> #5  0x000000000055b803 in pan::GIOChannelSocket::do_read 
>> (this=0x7fffd40010c0) at ../../../pan/tasks/socket-impl-gio.cc:346
>> #6  0x000000000055c69d in pan::GIOChannelSocket::gio_func 
>> (this=0x7fffd40010c0, channel=<optimized out>, cond=<optimized out>)
>>      at ../../../pan/tasks/socket-impl-gio.cc:457
>> #7  0x00007ffff536ea3f in g_main_context_dispatch () from 
>> /usr/lib64/libglib-2.0.so.0
>> #8  0x00007ffff536edc8 in g_main_context_iterate.isra () from 
>> /usr/lib64/libglib-2.0.so.0
>> #9  0x00007ffff536f232 in g_main_loop_run () from /usr/lib64/libglib-2.0.so.0
>> #10 0x00007ffff76c4d6f in gtk_main () from /usr/lib64/libgtk-x11-2.0.so.0
>> #11 0x0000000000461694 in run_pan_in_window (group_prefs=..., 
>> window=0x9260f0, prefs=..., queue=..., data=..., _gui=0x925a80) at 
>> ../../../pan/gui/pan.cc:553
>> #12 main (argc=1, argv=0x7fffffffdd08) at ../../../pan/gui/pan.cc:1107
>>
> Seems like I need to investigate this further... not fixed atm, reverted this 
> commit

I pulled again this morning:
 
commit 4c6d3bd7fc0dbd1ce1ad7e40690948b39764a89f
Author: Heinrich Müller <address@hidden>
Date:   Sat May 10 09:55:24 2014 +0200

    fix use-after-free

which I've been running for many hours without a single problem, not even one of
the frequent non-fatal assertions I've been seeing for many months.

Absolutely perfect, Heinrich, thank you! :)





reply via email to

[Prev in Thread] Current Thread [Next in Thread]