Re: Questions about mm

[Chintan Pandya <cpandya@codeaurora.org>]

Hi Sanidhya,

Could you explain your use-case or debug method ? Why do you need to 
keep a page even after its original process have got killed ?

thanks,

On 12/08/2014 09:54 AM, Sanidhya Kashyap wrote:
> Hello everyone,
>
> I have some questions about page allocation and locking.
>
> - Suppose that a process is about to be killed and before that happens, I want
> to keep the content of the page intact in the memory, i.e. the page should
> neither be zeored or allocated to some other process unless required. In order
> to achieve this, what can be the most optimal approach in which the internals of
> the kernel is not changed besides adding a syscall or something.
>
> - Another is what happens if I increase the count of mm_users and mm_count
> before and later that process gets killed. Assuming that the mm was linked only
> to the killed process. What will happen in this case?
>
> - Last question that I wanted to know is what will happen if I change the flags
> of the pages to be reserved and unevictable?
> Is it possible for the pages to be set pinned as well?
> Can this approach help me soling the first issue or I might get a BUG by some
> other component in kernel?
>
> Since, I have just started playing with the kernel, so there is a possibility
> that I might have asked something very silly/horrific. Please bear with me.
>
> Thanks,
> Sanidhya
>
> --
> 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>


-- 
Chintan Pandya

QUALCOMM INDIA, on behalf of Qualcomm Innovation Center, Inc. is a
member of the Code Aurora Forum, hosted by The Linux Foundation

--
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>