linux-kernel - Re: [PATCH v7] tpm: Check size of response before accessing data

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:   Fri, 20 Jan 2017 18:21:58 +0200
From:   Jarkko Sakkinen <jarkko.sakkinen@...ux.intel.com>
To:     Stefan Berger <stefanb@...ux.vnet.ibm.com>
Cc:     tpmdd-devel@...ts.sourceforge.net,
        linux-security-module@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH v7] tpm: Check size of response before accessing data

On Fri, Jan 20, 2017 at 03:36:30PM +0200, Jarkko Sakkinen wrote:
> On Thu, Jan 19, 2017 at 07:19:12AM -0500, Stefan Berger wrote:
> > Make sure that we have not received less bytes than what is indicated
> > in the header of the TPM response. Also, check the number of bytes in
> > the response before accessing its data.
> > 
> > Signed-off-by: Stefan Berger <stefanb@...ux.vnet.ibm.com>
> 
> Reviewed-by: Jarkko Sakkinen <jarkko.sakkinen@...ux.intel.com>

Oops. I found some odd stuff after all so hold on for a moment.
I could do these updates myself probably...

> >  ssize_t tpm_transmit_cmd(struct tpm_chip *chip, const void *cmd,
> > -			 int len, unsigned int flags, const char *desc)
> > +			 size_t len, size_t min_rsp_body_length,
> > +			 unsigned int flags, const char *desc)

BTW, maybe the cmd_length would be actually a better idea because
it gets mixes witht local variable.

> >  {
> >  	const struct tpm_output_header *header;
> >  	int err;
> > +	ssize_t length;

Maybe it would make sense to name this as rsp_length.

> >  
> > -	len = tpm_transmit(chip, (const u8 *)cmd, len, flags);
> > -	if (len <  0)
> > -		return len;
> > -	else if (len < TPM_HEADER_SIZE)
> > +	length = tpm_transmit(chip, (const u8 *)cmd, len, flags);
> > +	if (length <  0)
> > +		return length;
> > +	else if (length < TPM_HEADER_SIZE)
> >  		return -EFAULT;
> >  
> >  	header = cmd;
> > +	if (length < be32_to_cpu(header->length))
> > +		return -EFAULT;

Why '<' and not '!='? In what legit case length would be larger?

> >  
> >  	err = be32_to_cpu(header->return_code);
> >  	if (err != 0 && desc)
> >  		dev_err(&chip->dev, "A TPM error (%d) occurred %s\n", err,
> >  			desc);
> > +	if (err)
> > +		return err;
> >  
> > -	return err;
> > +	if (be32_to_cpu(header->length) <
> > +	    min_rsp_body_length + TPM_HEADER_SIZE)
> > +		return -EFAULT;

Why couldn't you use 'length' here?

/Jarkko