Thread (110 messages) 110 messages, 6 authors, 2024-08-21

Re: [PATCH v10 12/40] mm: Define VM_SHADOW_STACK for arm64 when we support GCS

From: Mark Brown <broonie@kernel.org>
Date: 2024-08-15 18:19:33
Also in: kvmarm, linux-arch, linux-arm-kernel, linux-fsdevel, linux-kselftest, linux-mm, linux-riscv, lkml

On Thu, Aug 15, 2024 at 05:53:19PM +0000, Edgecombe, Rick P wrote:
On Thu, 2024-08-15 at 17:39 +0100, Mark Brown wrote:
quoted
quoted
Oh, thanks for the heads up - I'd missed that.
quoted
Looking at this I think it makes sense to do as was done for x86 and
split this out into a separate series (part of why I'd missed it),
updating the generic implementation to do this by default.  That'll
touch a bunch of architectures and the series is already quite big,
it's not really an ABI impact.
The series is already upstream. You just need to add an arm version of that
linked patch. But up to you.
Your series modified the existing x86 custom arch_get_unmapped_area*()
functions, arm64 uses the generic implementation of those so I'd have to
either add custom implementations (which I can't imagine would be met
with great enthusiasm) or update the generic ones.  A generic
implementation seems reasonable and it looks like RISC-V would also end
up using it so while it's a bit invasive it does seem more sensible to
do the change there.

Attachments

Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help