Date: Fri, 29 Aug 2003 17:26:31 -0700 From: Tim Kientzle <kientzle@acm.org> To: Pav Lucistnik <pav@oook.cz> Cc: hackers@freebsd.org Subject: Re: gcc weirdness Message-ID: <3F4FEF37.7070303@acm.org> In-Reply-To: <1062201750.641.21.camel@hood.oook.cz> References: <1062195438.641.18.camel@hood.oook.cz> <3F4FE60B.1050002@acm.org> <1062201750.641.21.camel@hood.oook.cz>
next in thread | previous in thread | raw e-mail | index | archive | help
Pav Lucistnik wrote:
> V so, 30. 08. 2003 v 01:47, Tim Kientzle píše:
>
>
>>>What am I missing here?
>>>
>>>Let we have a function called popup(). Inside this function there is
>>>this code:
>>>
>>>list = get_children(ds, x, TYPE_ELEMENT);
>>>
>>>gdb shows these values:
>>>
>>>(gdb) print ds
>>>$46 = (Type *) 0x1
>>>
>>>(gdb) print x
>>>$47 = 0x86adb80
>>>
>>>So far, everything is fine. Now, let step down one frame in gdb:
>>>
>>>(gdb) down
>>>#5 0x0808b807 in get_children(ds=0x86adb80, node=0x1, tag_type=TYPE_E...
>>>
>>>In the source, there is a
>>>
>>>GList* get_children (Type* ds, PtrNode node, enum Types tag_type) {
>>>
>>>Do you see ds and node values swapped? The program crashes because of
>>>this swapping. What is wrong? I use gcc-3.3.1 on today's -CURRENT.
>
>
>>First, try putting printf() statements at the beginning of
>>get_children and just before the corresponding call to see
>>if that's really what's going on.
>
>
> Yes, printf gives out same (correct) values. And, now it crashes in
> different part of get_children.
>
>
>>I presume you're seeing this while debugging a core dump after
>>a crash? The crash may have mangled the stack, in which case,
>>gdb may not be correctly identifying the function arguments.
>
>
> So it is a memory corruption? Is there a good way how to debug such bugs?
>
> Thanks for help so far.
>
You have an error in your program.
Run the program under gdb and step carefully through get_children
to try to figure out what is wrong.
Tim
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?3F4FEF37.7070303>