On Tue, Aug 16, 2022 at 04:45:14PM +0200, Ard Biesheuvel wrote:
On Sun, 14 Aug 2022 at 17:30, Sasha Levin <sashal@xxxxxxxxxx> wrote:
From: Lecopzer Chen <lecopzer.chen@xxxxxxxxxxxx>
[ Upstream commit 565cbaad83d83e288927b96565211109bc984007 ]
Simply make shadow of vmalloc area mapped on demand.
Since the virtual address of vmalloc for Arm is also between
MODULE_VADDR and 0x100000000 (ZONE_HIGHMEM), which means the shadow
address has already included between KASAN_SHADOW_START and
KASAN_SHADOW_END.
Thus we need to change nothing for memory map of Arm.
This can fix ARM_MODULE_PLTS with KASan, support KASan for higmem
and support CONFIG_VMAP_STACK with KASan.
Signed-off-by: Lecopzer Chen <lecopzer.chen@xxxxxxxxxxxx>
Tested-by: Linus Walleij <linus.walleij@xxxxxxxxxx>
Reviewed-by: Linus Walleij <linus.walleij@xxxxxxxxxx>
Signed-off-by: Russell King (Oracle) <rmk+kernel@xxxxxxxxxxxxxxx>
Signed-off-by: Sasha Levin <sashal@xxxxxxxxxx>
This patch does not belong in -stable. It has no fixes: or cc:stable
tags, and the contents are completely inappropriate for backporting
anywhere. In general, I think that no patch that touches arch/arm
(with the exception of DTS updates, perhaps) should ever be backported
unless proposed or acked by the maintainer.
I'll drop it.
I know I shouldn't ask, but how were these patches build/boot tested?
KAsan is very tricky to get right, especially on 32-bit ARM ...
They were only build tested at this stage. They go through
boot/functional test only after they are actually queued up for the
various trees.
--
Thanks,
Sasha