Re: [PATCH] munmap sealed memory cause memory to split (bug)

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

 



On Thu, Oct 17, 2024 at 10:46:10AM +0100, Lorenzo Stoakes wrote:
[snip]

> Anyway, let's try to decode (please take this as input as to how you should
> try to communicate these things):
>
>
> So we start with a VMA like this:
>
> 012345678901
> xxxxxxxxxxxx
>
> We then seal the middle, starting at offset 4:
>
> 012345678901
> xxxx****xxxx
>
> This sets the VM_SEALED flag in the middle and splits VMAs resulting in 3
> VMAs.
>
> We then attempt to unmap 4 pages from offset 2, but this fails, as
> expected.
>
> 012345678901
> xxxx****xxxx
>   |--| fail
>
> We then attempt to unmap 4 pages from offset 6, but this fails, as
> expected.
>
> 012345678901
> xxxx****xxxx
>       |--| fail
>
> At each stage we should observe 4 VMAs.

CORRECTION: 3 VMAs.

[snip]




[Index of Archives]     [Linux ARM Kernel]     [Linux ARM]     [Linux Omap]     [Fedora ARM]     [IETF Annouce]     [Bugtraq]     [Linux OMAP]     [Linux MIPS]     [eCos]     [Asterisk Internet PBX]     [Linux API]

  Powered by Linux