[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <20171115084100.lkzww5e56asjtkyy@gmail.com>
Date: Wed, 15 Nov 2017 09:41:00 +0100
From: Ingo Molnar <mingo@...nel.org>
To: Ricardo Neri <ricardo.neri-calderon@...ux.intel.com>
Cc: Thomas Gleixner <tglx@...utronix.de>,
"H. Peter Anvin" <hpa@...or.com>, Borislav Petkov <bp@...e.de>,
Andy Lutomirski <luto@...nel.org>,
Tony Luck <tony.luck@...el.com>,
Paolo Bonzini <pbonzini@...hat.com>,
"Ravi V. Shankar" <ravi.v.shankar@...el.com>, x86@...nel.org,
ricardo.neri@...el.com, linux-kernel@...r.kernel.org
Subject: Re: [RESEND PATCH v2 4/4] x86/umip: Warn if UMIP-protected
instructions are used
* Ricardo Neri <ricardo.neri-calderon@...ux.intel.com> wrote:
> > > + snprintf(warn, sizeof(warn), "%s %s", umip_insns[umip_inst],
> > > + umip_warn_use);
> >
> > This is incredibly fragile against future buffer overflows, and warning about it
> > in comments does not make it less fragile!
>
> I need to concatenate the instruction mnemonic with the a string. Does something like
> this is more acceptable?
>
> unsigned char warn[50];
>
> ...
>
> strcpy(warn, umip_insns[umip_inst]);
> strcat(warn, " instruction cannot be used by applications.");
> umip_pr_warn(regs, warn, 0);
>
> In this manner I use the string literal directly but I still have a buffer that might
> overflow. Code looks more clear to me. I could #defines for the string lengths or
> set a maximum length.
This is still very fragile.
The right solution would be to make umip_pr_warn() a varargs helper function, so
that you can just use it to print things the usual way. I'd also use a
__attribute__((format(printf))) specification to get good build-time warnings.
Thanks,
Ingo
Powered by blists - more mailing lists