[LSF/MM TOPIC] The kernel address sanitizer

[LSF/MM TOPIC] The kernel address sanitizer

[Andrey Ryabinin]

Hello

I'm working on the kernel address sanitizer (KASan) -
http://thread.gmane.org/gmane.linux.kernel.mm/120041
KASan is runtime memory debugger designed to find use-after-free and
out-of-bounds accesses.

Seems we've come to agreement that KASan is useful and deserves to be
in mainline, yet the feedback on patches is poor.
It seems like they are stalled, so I would like to discuss the future
of it. I hope this will help in pushing it forward.

Besides we have ideas for further improvements, like:

 * Detecting reads of uninitialized memory.
 * Quarantine - delaying reallocation of freed memory to increase
chance of catching use after free bugs.
                    In combination with DEBUG_PAGEALLOC or slab
poisoning it's useful even without KASan.
 * and some more...

Perhaps it's worth to discuss them as well. I'll be able to come up
with some prototype until summit if needed.

--
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/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

Re: [LSF/MM TOPIC] The kernel address sanitizer

[Dmitry Vyukov]

+some more people

On Thu, Dec 25, 2014 at 3:01 PM, Andrey Ryabinin <ryabinin.a.a@gmail.com> wrote:
> Hello
>
> I'm working on the kernel address sanitizer (KASan) -
> http://thread.gmane.org/gmane.linux.kernel.mm/120041
> KASan is runtime memory debugger designed to find use-after-free and
> out-of-bounds accesses.
>
> Seems we've come to agreement that KASan is useful and deserves to be
> in mainline, yet the feedback on patches is poor.
> It seems like they are stalled, so I would like to discuss the future
> of it. I hope this will help in pushing it forward.
>
> Besides we have ideas for further improvements, like:
>
>  * Detecting reads of uninitialized memory.
>  * Quarantine - delaying reallocation of freed memory to increase
> chance of catching use after free bugs.
>                     In combination with DEBUG_PAGEALLOC or slab
> poisoning it's useful even without KASan.
>  * and some more...
>
> Perhaps it's worth to discuss them as well. I'll be able to come up
> with some prototype until summit if needed.

--
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/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

Re: [LSF/MM TOPIC] The kernel address sanitizer

[Dave Hansen]

On 12/25/2014 04:01 AM, Andrey Ryabinin wrote:
> Seems we've come to agreement that KASan is useful and deserves to be
> in mainline, yet the feedback on patches is poor.
> It seems like they are stalled, so I would like to discuss the future
> of it. I hope this will help in pushing it forward.

I think this should more broadly be a talk about our memory-related
debugging options.  This is an especially good audience for seeing what
gets used and if we need to start culling any of them.

--
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/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

Re: [LSF/MM TOPIC] The kernel address sanitizer

[Andrey Ryabinin]

2015-01-14 1:00 GMT+03:00 Dave Hansen <dave.hansen@intel.com>:
> On 12/25/2014 04:01 AM, Andrey Ryabinin wrote:
>> Seems we've come to agreement that KASan is useful and deserves to be
>> in mainline, yet the feedback on patches is poor.
>> It seems like they are stalled, so I would like to discuss the future
>> of it. I hope this will help in pushing it forward.
>
> I think this should more broadly be a talk about our memory-related
> debugging options.  This is an especially good audience for seeing what
> gets used and if we need to start culling any of them.
>

No objections, I tend to agree with you.

--
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/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>