linux-kernel - Re: [PATCH 4/4] uprobe: only do FOLL_SPLIT

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: <20191016121031.GA31585@redhat.com>
Date:   Wed, 16 Oct 2019 14:10:31 +0200
From:   Oleg Nesterov <oleg@...hat.com>
To:     Song Liu <songliubraving@...com>
Cc:     linux-kernel@...r.kernel.org, linux-mm@...ck.org,
        akpm@...ux-foundation.org, matthew.wilcox@...cle.com,
        kernel-team@...com, william.kucharski@...cle.com,
        "Kirill A . Shutemov" <kirill.shutemov@...ux.intel.com>,
        Srikar Dronamraju <srikar@...ux.vnet.ibm.com>
Subject: Re: [PATCH 4/4] uprobe: only do FOLL_SPLIT_PMD for uprobe register

On 10/16, Song Liu wrote:
>
> --- a/kernel/events/uprobes.c
> +++ b/kernel/events/uprobes.c
> @@ -474,14 +474,17 @@ int uprobe_write_opcode(struct arch_uprobe *auprobe, struct mm_struct *mm,
>  	struct vm_area_struct *vma;
>  	int ret, is_register, ref_ctr_updated = 0;
>  	bool orig_page_huge = false;
> +	unsigned int gup_flags = FOLL_FORCE;
>  
>  	is_register = is_swbp_insn(&opcode);
>  	uprobe = container_of(auprobe, struct uprobe, arch);
>  
>  retry:
> +	if (is_register)
> +		gup_flags |= FOLL_SPLIT_PMD;
>  	/* Read the page with vaddr into memory */
> -	ret = get_user_pages_remote(NULL, mm, vaddr, 1,
> -			FOLL_FORCE | FOLL_SPLIT_PMD, &old_page, &vma, NULL);
> +	ret = get_user_pages_remote(NULL, mm, vaddr, 1, gup_flags,
> +				    &old_page, &vma, NULL);
>  	if (ret <= 0)
>  		return ret;
>  
> @@ -489,6 +492,9 @@ int uprobe_write_opcode(struct arch_uprobe *auprobe, struct mm_struct *mm,
>  	if (ret <= 0)
>  		goto put_old;
>  
> +	WARN(!is_register && PageCompound(old_page),
> +	     "uprobe unregister should never work on compound page\n");

But this can happen with the change above. You can't know if *vaddr was
previously changed by install_breakpoint() or not.

If not, verify_opcode() should likely save us, but we can't rely on it.
Say, someone can write "int3" into vm_file at uprobe->offset.

And I am not sure it is safe to continue in this case, I'd suggest to
return -EWHATEVER to avoid the possible crash.

Oleg.