Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]


Groups > linux.kernel > #1321981 > unrolled thread

Re: [PATCH v4 00/22] arm64: implement support for KASLR

Started byCatalin Marinas <catalin.marinas@arm.com>
First post2016-01-29 19:30 +0100
Last post2016-01-29 19:50 +0100
Articles 2 — 2 participants

Back to article view | Back to linux.kernel

This discussion starts older than the indexed window; earlier articles aren't shown. The article labeled Started by below is the oldest one visible, not the original post.


Contents

  Re: [PATCH v4 00/22] arm64: implement support for KASLR Catalin Marinas <catalin.marinas@arm.com> - 2016-01-29 19:30 +0100
    Re: [PATCH v4 00/22] arm64: implement support for KASLR Ard Biesheuvel <ard.biesheuvel@linaro.org> - 2016-01-29 19:50 +0100

#1321981 — Re: [PATCH v4 00/22] arm64: implement support for KASLR

FromCatalin Marinas <catalin.marinas@arm.com>
Date2016-01-29 19:30 +0100
SubjectRe: [PATCH v4 00/22] arm64: implement support for KASLR
Message-ID<qWlXZ-79P-49@gated-at.bofh.it>
Hi Ard,

On Tue, Jan 26, 2016 at 06:10:27PM +0100, Ard Biesheuvel wrote:
> Code can be found here:
> git://git.linaro.org/people/ard.biesheuvel/linux-arm.git arm64-kaslr-v4a
> https://git.linaro.org/people/ard.biesheuvel/linux-arm.git/shortlog/refs/heads/arm64-kaslr-v4a

The overall series looks fine but I'd like more time to review the KASLR
part together with the module PLT stuff.

So could you please split this series in 2-3 parts for easy merging
(possibly without the KASLR part, it depends on how the review goes)? It
looks like a mix of features in random order like huge-vmap, relative
extable, kernel memory layout changes, kernel load address, PIE and
KASLR. So something like:

1. relative extable
2. huge-vmap
3. kernel memory layout changes (moving kernel to the base of vmalloc
range)
4. allow kernel loading at different phys offsets
5. module PLTs, relocations, PIE, relative kallsyms, KASLR
6. efi_get_random_bytes

1-4 can be in the same branch but I currently find it hard to
cherry-pick the non-PIE/non-KASLR patches without conflicts.

Thanks.

-- 
Catalin

[toc] | [next] | [standalone]


#1321997

FromArd Biesheuvel <ard.biesheuvel@linaro.org>
Date2016-01-29 19:50 +0100
Message-ID<qWmhj-7kP-5@gated-at.bofh.it>
In reply to#1321981
giOn 29 January 2016 at 19:26, Catalin Marinas <catalin.marinas@arm.com> wrote:
> Hi Ard,
>
> On Tue, Jan 26, 2016 at 06:10:27PM +0100, Ard Biesheuvel wrote:
>> Code can be found here:
>> git://git.linaro.org/people/ard.biesheuvel/linux-arm.git arm64-kaslr-v4a
>> https://git.linaro.org/people/ard.biesheuvel/linux-arm.git/shortlog/refs/heads/arm64-kaslr-v4a
>
> The overall series looks fine but I'd like more time to review the KASLR
> part together with the module PLT stuff.
>
> So could you please split this series in 2-3 parts for easy merging
> (possibly without the KASLR part, it depends on how the review goes)?

Sure

> It
> looks like a mix of features in random order like huge-vmap, relative
> extable, kernel memory layout changes, kernel load address, PIE and
> KASLR. So something like:
>
> 1. relative extable

This has been picked up by akpm in the mean time

> 2. huge-vmap

This is a single patch which is completely independent. I don't expect
it to conflict when applied in isolation.

> 3. kernel memory layout changes (moving kernel to the base of vmalloc
> range)
> 4. allow kernel loading at different phys offsets
> 5. module PLTs, relocations, PIE, relative kallsyms, KASLR

Relative kallsyms has also been picked up by akpm

> 6. efi_get_random_bytes
>
> 1-4 can be in the same branch but I currently find it hard to
> cherry-pick the non-PIE/non-KASLR patches without conflicts.
>

I think 3 logically coherent series that apply in sequence is
feasible. I will look into this on Monday

-- 
Ard.

[toc] | [prev] | [standalone]


Back to top | Article view | linux.kernel


csiph-web