Re: [PATCH] mm: fix possible cause of a page_mapped BUG

["Robert Święcki" <robert@swiecki.net>]

On Sat, Apr 2, 2011 at 6:01 AM, Hui Zhu <teawater@gmail.com> wrote:
> On Sat, Apr 2, 2011 at 00:35, Linus Torvalds
> <torvalds@linux-foundation.org> wrote:
>> On Fri, Apr 1, 2011 at 9:21 AM, Robert Święcki <robert@swiecki.net> wrote:
>>>
>>> Is it possible to turn it off via config flags? Looking into
>>> arch/x86/include/asm/bug.h it seems it's unconditional (as in "it
>>> always manifests itself somehow") and I have
>>> CONFIG_DEBUG_BUGVERBOSE=y.
>>
>> Ok, if you have CONFIG_DEBUG_BUGVERBOSE then, you do have the bug-table.
>>
>> Maybe it's just kdb that is broken, and doesn't print it. I wouldn't
>> be surprised. It's not the first time I've seen debugging features
>> that just make debugging a mess.
>>
>>> Anything that could help you debugging this? Uploading kernel image
>>> (unfortunately I've overwritten this one), dumping more kgdb data?
>>
>> So in this case kgdb just dropped the most important data on the floor.
>>
>> But if you have kdb active next time, print out the vma/old contents
>> in that function that has the BUG() in it.
>>
>>> I must admit I'm not up-to-date with current linux kernel debugging
>>> techniques. The kernel config is here:
>>> http://alt.swiecki.net/linux_kernel/ise-test-2.6.38-kernel-config.txt
>>>
>>> For now I'll compile with -O0 -fno-inline (are you sure you'd like -Os?)
>
> Hi Robert,
>
> I am not sure you can success with build trunk with  -O0 -fno-inline.
> I suggest you try the patch in
> http://code.google.com/p/kgtp/downloads/detail?name=co.patch.
> It add a option in "Kernel hacking" called "Compile with almost no
> optimization". It will make kernel be built without -O2. It support
> x86_32, x86_64 and arm.

HI,

Yeah.. -O0 doesn't build smoothly, it seems that building with -O0 is
not required right now, but I'll keep your patch in mind in case it
becomes necessary. Thanks for the tip.

> PS, maybe you can try kgtp (https://code.google.com/p/kgtp/)  debug your kernel.
>
>>
>> Oh, don't do that. -O0 makes the code totally unreadable (the compiler
>> just does _stupid_ things, making the asm code look so horrible that
>> you can't match it up against anything sane), and -fno-inline isn't
>> worth the pain either.
>>
>> -Os is much better than those.
>>
>> But in this case, just getting the filename and line number would have
>> made the thing moot anyway - without kdb it _should_ have said
>> something clear like
>>
>>   kernel BUG at %s:%u!
>>
>> where %s:%u is the filename and line number.
>>
>>                          Linus
>> --
>> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
>> the body of a message to majordomo@vger.kernel.org
>> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>> Please read the FAQ at  http://www.tux.org/lkml/
>>
>



-- 
Robert Święcki

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>