linux-kernel - Re: [RFC][PATCH] x86: ptrace: Add function argument access API

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]
Date:   Mon, 15 Oct 2018 23:56:46 +0900
From:   Masami Hiramatsu <mhiramat@...nel.org>
To:     Steven Rostedt <rostedt@...dmis.org>
Cc:     Andy Lutomirski <luto@...capital.net>,
        LKML <linux-kernel@...r.kernel.org>,
        Thomas Gleixner <tglx@...utronix.de>,
        Peter Zijlstra <peterz@...radead.org>,
        Ingo Molnar <mingo@...nel.org>,
        "H. Peter Anvin" <hpa@...or.com>,
        Masami Hiramatsu <mhiramat@...nel.org>,
        Josh Poimboeuf <jpoimboe@...hat.com>
Subject: Re: [RFC][PATCH] x86: ptrace: Add function argument access API

On Fri, 12 Oct 2018 15:54:39 -0400
Steven Rostedt <rostedt@...dmis.org> wrote:

> On Fri, 12 Oct 2018 11:21:28 -0700
> Andy Lutomirski <luto@...capital.net> wrote:
> 
> > On Fri, Oct 12, 2018 at 9:26 AM Steven Rostedt <rostedt@...dmis.org> wrote:
> > >
> > >
> > > Anyone have any issues with this patch?
> > >  
> > 
> > I'm conceptually okay with it.  That being said,
> > regs_within_kernel_stack(), which you're indirectly using, is
> > off-by-a-few.  And updating it to use probe_kernel_read() might be
> > nice for robustness.
> > 
> 
> Something like this?
> 
> -- Steve
> 
> From: "Steven Rostedt (VMware)" <rostedt@...dmis.org>
> Date: Fri, 12 Oct 2018 15:44:20 -0400
> Subject: [PATCH] x86: ptrace.h: Add regs_get_kernel_stack_nth_safe() function
> 
> Andy had some concerns about using regs_get_kernel_stack_nth() in a new
> function regs_get_kernel_argument() as if there's any error in the stack
> code, it could cause a bad memory access. Instead, add a new function called
> regs_get_kernel_stack_nth_safe() that does a probe_kernel_read() on the
> stack address to be extra careful in accessing the memory. To share the
> code, regs_get_kernel_stack_nth_addr() was added to just return the stack
> address (or NULL if not on the stack), that both regs_get_kernel_stack_nth()
> and the _safe() version can use.

This patch looks good to me.
But if the concern is real, all regs_get_kernel_stack_nth() user must move
onto _safe() version, at least all tracers code.

Reviewed-by: Masami Hiramatsu <mhiramat@...nel.org>

Thanks,

> 
> Link: http://lkml.kernel.org/r/CALCETrXn9zKTb9i1LP3qoFcpqZHF34BdkuZ5D3N0uCmRr+VnbA@mail.gmail.com
> Requested-by: Andy Lutomirski <luto@...capital.net>
> Signed-off-by: Steven Rostedt (VMware) <rostedt@...dmis.org>
> ---
>  arch/x86/include/asm/ptrace.h | 57 ++++++++++++++++++++++++++++++++++++++++---
>  1 file changed, 54 insertions(+), 3 deletions(-)
> 
> diff --git a/arch/x86/include/asm/ptrace.h b/arch/x86/include/asm/ptrace.h
> index c2304b25e2fd..8df7ab6a17c5 100644
> --- a/arch/x86/include/asm/ptrace.h
> +++ b/arch/x86/include/asm/ptrace.h
> @@ -237,6 +237,27 @@ static inline int regs_within_kernel_stack(struct pt_regs *regs,
>  }
>  
>  /**
> + * regs_get_kernel_stack_nth_addr() - get the address of the Nth entry on stack
> + * @regs:	pt_regs which contains kernel stack pointer.
> + * @n:		stack entry number.
> + *
> + * regs_get_kernel_stack_nth() returns the address of the @n th entry of the
> + * kernel stack which is specified by @regs. If the @n th entry is NOT in
> + * the kernel stack, this returns NULL.
> + */
> +static inline unsigned long *regs_get_kernel_stack_nth_addr(struct pt_regs *regs,
> +							    unsigned int n)
> +{
> +	unsigned long *addr = (unsigned long *)kernel_stack_pointer(regs);
> +
> +	addr += n;
> +	if (regs_within_kernel_stack(regs, (unsigned long)addr))
> +		return addr;
> +	else
> +		return NULL;
> +}
> +
> +/**
>   * regs_get_kernel_stack_nth() - get Nth entry of the stack
>   * @regs:	pt_regs which contains kernel stack pointer.
>   * @n:		stack entry number.
> @@ -248,14 +269,44 @@ static inline int regs_within_kernel_stack(struct pt_regs *regs,
>  static inline unsigned long regs_get_kernel_stack_nth(struct pt_regs *regs,
>  						      unsigned int n)
>  {
> -	unsigned long *addr = (unsigned long *)kernel_stack_pointer(regs);
> -	addr += n;
> -	if (regs_within_kernel_stack(regs, (unsigned long)addr))
> +	unsigned long *addr;
> +
> +	addr = regs_get_kernel_stack_nth_addr(regs, n);
> +	if (addr)
>  		return *addr;
>  	else
>  		return 0;
>  }
>  
> +/* To avoid include hell, we can't include uaccess.h */
> +extern long probe_kernel_read(void *dst, const void *src, size_t size);
> +
> +/**
> + * regs_get_kernel_stack_nth_safe() - safely get Nth entry of the stack
> + * @regs:	pt_regs which contains kernel stack pointer.
> + * @n:		stack entry number.
> + *
> + * Same as regs_get_kernel_stack_nth(), but references the stack value
> + * with a probe_kernel_read() in case there's a bad stack pointer, it
> + * will not cause a bad memory access. If the @n is not on the stack,
> + * or a bad memory access happened, it returns zero.
> + */
> +static inline unsigned long regs_get_kernel_stack_nth_safe(struct pt_regs *regs,
> +							   unsigned int n)
> +{
> +	unsigned long *addr;
> +	unsigned long val;
> +	long ret;
> +
> +	addr = regs_get_kernel_stack_nth_addr(regs, n);
> +	if (addr) {
> +		ret = probe_kernel_read(&val, addr, sizeof(val));
> +		if (!ret)
> +			return val;
> +	}
> +	return 0;
> +}
> +
>  /**
>   * regs_get_kernel_argument() - get Nth function argument in kernel
>   * @regs:	pt_regs of that context
> -- 
> 2.13.6
> 


-- 
Masami Hiramatsu <mhiramat@...nel.org>