ACK: [PATCH 0/1][SRU][D] PAN is broken for execute-only user mappings on ARMv8

[Connor Kuehl]

On 1/8/20 9:21 AM, Tyler Hicks wrote:
> BugLink: https://launchpad.net/bugs/1858815
> 
> Clean cherry-pick that is only build tested at this time. I'm relying on
> our SRU tests for regression testing.
> 
> [Impact]
> 
> It was discovered that upstream kernel commit cab15ce604e5 ("arm64:
> Introduce execute-only page access permissions"), which introduced
> execute-only user mappings, subverted the Privileged Access Never
> protections.
> 
> The fix is to effectively revert commit cab15ce604e5. This is done in
> upstream kernel commit 24cecc377463 ("arm64: Revert support for
> execute-only user mappings").
> 
> [Test Case]
> 
> I'm not aware of any PAN test cases. Booting our arm64 kernels on an
> ARMv8 device and running through our typical regression tests is
> probably the best we can do at this time.
> 
> [Regression Potential]
> 
> Touching the page handling code always carries significant risk.
> However, the fix is simply reverting the change that added the
> execute-only user mappings feature in v4.9.
> 
> Tyler
> 
> Catalin Marinas (1):
>    arm64: Revert support for execute-only user mappings
> 
>   arch/arm64/include/asm/pgtable-prot.h |  5 ++---
>   arch/arm64/include/asm/pgtable.h      | 10 +++-------
>   arch/arm64/mm/fault.c                 |  2 +-
>   mm/mmap.c                             |  6 ------
>   4 files changed, 6 insertions(+), 17 deletions(-)
> 


Acked-by: Connor Kuehl <connor.kuehl at canonical.com>