linux-kernel - Re: [PATCH v11 1/6] arm64: mte: Sync tags for pages where PTE is untagged

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20210427174357.GA17872@arm.com>
Date:   Tue, 27 Apr 2021 18:43:58 +0100
From:   Catalin Marinas <catalin.marinas@....com>
To:     Steven Price <steven.price@....com>
Cc:     Marc Zyngier <maz@...nel.org>, Will Deacon <will@...nel.org>,
        James Morse <james.morse@....com>,
        Julien Thierry <julien.thierry.kdev@...il.com>,
        Suzuki K Poulose <suzuki.poulose@....com>,
        kvmarm@...ts.cs.columbia.edu, linux-arm-kernel@...ts.infradead.org,
        linux-kernel@...r.kernel.org, Dave Martin <Dave.Martin@....com>,
        Mark Rutland <mark.rutland@....com>,
        Thomas Gleixner <tglx@...utronix.de>, qemu-devel@...gnu.org,
        Juan Quintela <quintela@...hat.com>,
        "Dr. David Alan Gilbert" <dgilbert@...hat.com>,
        Richard Henderson <richard.henderson@...aro.org>,
        Peter Maydell <peter.maydell@...aro.org>,
        Haibo Xu <Haibo.Xu@....com>, Andrew Jones <drjones@...hat.com>
Subject: Re: [PATCH v11 1/6] arm64: mte: Sync tags for pages where PTE is
 untagged

On Fri, Apr 16, 2021 at 04:43:04PM +0100, Steven Price wrote:
> A KVM guest could store tags in a page even if the VMM hasn't mapped
> the page with PROT_MTE. So when restoring pages from swap we will
> need to check to see if there are any saved tags even if !pte_tagged().
> 
> However don't check pages which are !pte_valid_user() as these will
> not have been swapped out.

You should remove the pte_valid_user() mention from the commit log as
well.

> diff --git a/arch/arm64/include/asm/pgtable.h b/arch/arm64/include/asm/pgtable.h
> index e17b96d0e4b5..cf4b52a33b3c 100644
> --- a/arch/arm64/include/asm/pgtable.h
> +++ b/arch/arm64/include/asm/pgtable.h
> @@ -312,7 +312,7 @@ static inline void set_pte_at(struct mm_struct *mm, unsigned long addr,
>  		__sync_icache_dcache(pte);
>  
>  	if (system_supports_mte() &&
> -	    pte_present(pte) && pte_tagged(pte) && !pte_special(pte))
> +	    pte_present(pte) && (pte_val(pte) & PTE_USER) && !pte_special(pte))

I would add a pte_user() macro here or, if we restore the tags only when
the page is readable, use pte_access_permitted(pte, false). Also add a
comment why we do this.

There's also the pte_user_exec() case which may not have the PTE_USER
set (exec-only permission) but I don't think it matters. We don't do tag
checking on instruction fetches, so if the user adds a PROT_READ to it,
it would go through set_pte_at() again. I'm not sure KVM does anything
special with exec-only mappings at stage 2, I suspect they won't be
accessible by the guest (but needs checking).

>  		mte_sync_tags(ptep, pte);
>  
>  	__check_racy_pte_update(mm, ptep, pte);
> diff --git a/arch/arm64/kernel/mte.c b/arch/arm64/kernel/mte.c
> index b3c70a612c7a..e016ab57ea36 100644
> --- a/arch/arm64/kernel/mte.c
> +++ b/arch/arm64/kernel/mte.c
> @@ -26,17 +26,23 @@ u64 gcr_kernel_excl __ro_after_init;
>  
>  static bool report_fault_once = true;
>  
> -static void mte_sync_page_tags(struct page *page, pte_t *ptep, bool check_swap)
> +static void mte_sync_page_tags(struct page *page, pte_t *ptep, bool check_swap,
> +			       bool pte_is_tagged)
>  {
>  	pte_t old_pte = READ_ONCE(*ptep);
>  
>  	if (check_swap && is_swap_pte(old_pte)) {
>  		swp_entry_t entry = pte_to_swp_entry(old_pte);
>  
> -		if (!non_swap_entry(entry) && mte_restore_tags(entry, page))
> +		if (!non_swap_entry(entry) && mte_restore_tags(entry, page)) {
> +			set_bit(PG_mte_tagged, &page->flags);
>  			return;
> +		}
>  	}
>  
> +	if (!pte_is_tagged || test_and_set_bit(PG_mte_tagged, &page->flags))
> +		return;

I don't think we need another test_bit() here, it was done in the
caller (bar potential races which need more thought).

> +
>  	page_kasan_tag_reset(page);
>  	/*
>  	 * We need smp_wmb() in between setting the flags and clearing the
> @@ -54,11 +60,13 @@ void mte_sync_tags(pte_t *ptep, pte_t pte)
>  	struct page *page = pte_page(pte);
>  	long i, nr_pages = compound_nr(page);
>  	bool check_swap = nr_pages == 1;
> +	bool pte_is_tagged = pte_tagged(pte);
>  
>  	/* if PG_mte_tagged is set, tags have already been initialised */
>  	for (i = 0; i < nr_pages; i++, page++) {
> -		if (!test_and_set_bit(PG_mte_tagged, &page->flags))
> -			mte_sync_page_tags(page, ptep, check_swap);
> +		if (!test_bit(PG_mte_tagged, &page->flags))
> +			mte_sync_page_tags(page, ptep, check_swap,
> +					   pte_is_tagged);
>  	}
>  }

You were right in the previous thread that if we have a race, it's
already there even without your patches KVM patches.

If it's the same pte in a multithreaded app, we should be ok as the core
code holds the ptl (the arch code also holds the mmap_lock during
exception handling but only as a reader, so you can have multiple
holders).

If there are multiple ptes to the same page, for example mapped with
MAP_ANONYMOUS | MAP_SHARED, metadata recovery is done via
arch_swap_restore() before we even set the pte and with the page locked.
So calling lock_page() again in mte_restore_tags() would deadlock.

I can see that do_swap_page() also holds the page lock around
set_pte_at(), so I think we are covered.

Any other scenario I may have missed? My understanding is that if the
pte is the same, we have the ptl. Otherwise we have the page lock for
shared pages.

-- 
Catalin