Re: [PATCH] tpm: Do not dereference NULL pointer if acpi_os_map_memory() fails.

[Kent Yoder <key@linux.vnet.ibm.com>]

On Thu, Aug 16, 2012 at 12:16:33AM +0200, Jesper Juhl wrote:
> In drivers/char/tpm/tpm_acpi.c::read_log() we call
> acpi_os_map_memory(). That call may fail for a number of reasons
> (invalid address, out of memory etc). If the call fails it returns
> NULL and we just pass that to memcpy() unconditionally, which will go
> bad when it tries to dereference the pointer.
> 
> Unfortunately we just get NULL back, so we can't really tell the user
> exactely what went wrong, but we can at least avoid crashing and
> return an error (-EIO seemed more generic and more suitable here than
> -ENOMEM or something else, so I picked that).

  Thanks Jesper, applied here:

git://github.com/shpedoikal/linux.git v3.6-rc1-tpmdd-staging

Kent

> 
> Signed-off-by: Jesper Juhl <jj@chaosbits.net>
> ---
>  drivers/char/tpm/tpm_acpi.c | 5 +++++
>  1 file changed, 5 insertions(+)
> 
>  note: this patch is against git://github.com/shpedoikal/linux.git v3.6-rc1-tpmdd-staging
> 
> diff --git a/drivers/char/tpm/tpm_acpi.c b/drivers/char/tpm/tpm_acpi.c
> index a1bb5a18..fe3fa94 100644
> --- a/drivers/char/tpm/tpm_acpi.c
> +++ b/drivers/char/tpm/tpm_acpi.c
> @@ -96,6 +96,11 @@ int read_log(struct tpm_bios_log *log)
>  	log->bios_event_log_end = log->bios_event_log + len;
> 
>  	virt = acpi_os_map_memory(start, len);
> +	if (!virt) {
> +		kfree(log->bios_event_log);
> +		printk("%s: ERROR - Unable to map memory\n", __func__);
> +		return -EIO;
> +	}
> 
>  	memcpy(log->bios_event_log, virt, len);
> 
> -- 
> 1.7.11.4
> 
> 
> -- 
> Jesper Juhl <jj@chaosbits.net>       http://www.chaosbits.net/
> Don't top-post http://www.catb.org/jargon/html/T/top-post.html
> Plain text mails only, please.
>