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


Groups > linux.kernel > #1730399 > unrolled thread

[PATCH v3 0/3] arm64: Add CONFIG_KUSER_HELPERS

Started byMark Salyzyn <salyzyn@android.com>
First post2017-09-11 17:30 +0200
Last post2017-09-20 11:20 +0200
Articles 3 — 3 participants

Back to article view | Back to linux.kernel


Contents

  [PATCH v3 0/3] arm64: Add CONFIG_KUSER_HELPERS Mark Salyzyn <salyzyn@android.com> - 2017-09-11 17:30 +0200
    Re: [PATCH v3 0/3] arm64: Add CONFIG_KUSER_HELPERS Kees Cook <keescook@chromium.org> - 2017-09-20 00:10 +0200
      Re: [PATCH v3 0/3] arm64: Add CONFIG_KUSER_HELPERS Will Deacon <will.deacon@arm.com> - 2017-09-20 11:20 +0200

#1730399 — [PATCH v3 0/3] arm64: Add CONFIG_KUSER_HELPERS

FromMark Salyzyn <salyzyn@android.com>
Date2017-09-11 17:30 +0200
Subject[PATCH v3 0/3] arm64: Add CONFIG_KUSER_HELPERS
Message-ID<uoyOS-5b8-11@gated-at.bofh.it>
Make it possible to disable the kuser helpers by adding a KUSER_HELPERS
config option (enabled by default). When disabled, all kuser
helpers-related code is removed from the kernel and no mapping is done
at the fixed high address (0xffff0000); any attempt to use a kuser
helper from a 32-bit process will result in a segfault.

Signed-off-by: Mark Salyzyn <salyzyn@android.com>

v2:
- vectors_page and compat_vdso_spec as array of 2
- free sigpage if vectors allocation failed
- adjust makefile so one line for each of the assembler source modules
- split off assembler changes to a new previous patch in series to reduce churn
- modify slightly the feature documentation to reduce its reach
- modify slightly the feature documentation to rationalize the yes default.
- There are more ifdefs as a result of the rebase.

v3:
- rebase (minor conflicts)

[toc] | [next] | [standalone]


#1735291

FromKees Cook <keescook@chromium.org>
Date2017-09-20 00:10 +0200
Message-ID<urySl-2Ga-7@gated-at.bofh.it>
In reply to#1730399
On Mon, Sep 11, 2017 at 8:26 AM, Mark Salyzyn <salyzyn@android.com> wrote:
> Make it possible to disable the kuser helpers by adding a KUSER_HELPERS
> config option (enabled by default). When disabled, all kuser
> helpers-related code is removed from the kernel and no mapping is done
> at the fixed high address (0xffff0000); any attempt to use a kuser
> helper from a 32-bit process will result in a segfault.
>
> Signed-off-by: Mark Salyzyn <salyzyn@android.com>
>
> v2:
> - vectors_page and compat_vdso_spec as array of 2
> - free sigpage if vectors allocation failed
> - adjust makefile so one line for each of the assembler source modules
> - split off assembler changes to a new previous patch in series to reduce churn
> - modify slightly the feature documentation to reduce its reach
> - modify slightly the feature documentation to rationalize the yes default.
> - There are more ifdefs as a result of the rebase.
>
> v3:
> - rebase (minor conflicts)

This looks great to me. I'd love to see kuser helpers gone. :)

Catalin, Will, does this need anything else?

-Kees

-- 
Kees Cook
Pixel Security

[toc] | [prev] | [next] | [standalone]


#1735661

FromWill Deacon <will.deacon@arm.com>
Date2017-09-20 11:20 +0200
Message-ID<urJkK-11L-17@gated-at.bofh.it>
In reply to#1735291
On Tue, Sep 19, 2017 at 03:02:22PM -0700, Kees Cook wrote:
> On Mon, Sep 11, 2017 at 8:26 AM, Mark Salyzyn <salyzyn@android.com> wrote:
> > Make it possible to disable the kuser helpers by adding a KUSER_HELPERS
> > config option (enabled by default). When disabled, all kuser
> > helpers-related code is removed from the kernel and no mapping is done
> > at the fixed high address (0xffff0000); any attempt to use a kuser
> > helper from a 32-bit process will result in a segfault.
> >
> > Signed-off-by: Mark Salyzyn <salyzyn@android.com>
> >
> > v2:
> > - vectors_page and compat_vdso_spec as array of 2
> > - free sigpage if vectors allocation failed
> > - adjust makefile so one line for each of the assembler source modules
> > - split off assembler changes to a new previous patch in series to reduce churn
> > - modify slightly the feature documentation to reduce its reach
> > - modify slightly the feature documentation to rationalize the yes default.
> > - There are more ifdefs as a result of the rebase.
> >
> > v3:
> > - rebase (minor conflicts)
> 
> This looks great to me. I'd love to see kuser helpers gone. :)
> 
> Catalin, Will, does this need anything else?

I haven't had chance to look at the patches yet (buried in other stuff), but
I aim to get to it in the next couple of weeks.

Will

[toc] | [prev] | [standalone]


Back to top | Article view | linux.kernel


csiph-web