[PATCH RESEND] mm, kasan: don't poison boot memory

[Catalin Marinas]

On Wed, Feb 17, 2021 at 09:59:24PM +0100, Andrey Konovalov wrote:
> During boot, all non-reserved memblock memory is exposed to the buddy
> allocator. Poisoning all that memory with KASAN lengthens boot time,
> especially on systems with large amount of RAM. This patch makes
> page_alloc to not call kasan_free_pages() on all new memory.
> 
> __free_pages_core() is used when exposing fresh memory during system
> boot and when onlining memory during hotplug. This patch adds a new
> FPI_SKIP_KASAN_POISON flag and passes it to __free_pages_ok() through
> free_pages_prepare() from __free_pages_core().
> 
> This has little impact on KASAN memory tracking.
> 
> Assuming that there are no references to newly exposed pages before they
> are ever allocated, there won't be any intended (but buggy) accesses to
> that memory that KASAN would normally detect.
> 
> However, with this patch, KASAN stops detecting wild and large
> out-of-bounds accesses that happen to land on a fresh memory page that
> was never allocated. This is taken as an acceptable trade-off.
> 
> All memory allocated normally when the boot is over keeps getting
> poisoned as usual.
> 
> Signed-off-by: Andrey Konovalov <andreyknvl at google.com>

The approach looks fine to me. If you don't like the trade-off, I think
you could still leave the kasan poisoning in if CONFIG_DEBUG_KERNEL.

Reviewed-by: Catalin Marinas <catalin.marinas at arm.com>

Just curious, have you noticed any issue booting a KASAN_SW_TAGS-enabled
kernel on a system with sufficiently large RAM? Is the boot slow-down
significant?

For MTE, we could look at optimising the poisoning code for page size to
use STGM or DC GZVA but I don't think we can make it unnoticeable for
large systems (especially with DC GZVA, that's like zeroing the whole
RAM at boot).

-- 
Catalin