Re: [PATCH 1/6] firmware: permit LSMs and IMA to fail firmware sysfs fallback loading

[Mimi Zohar <zohar@linux.vnet.ibm.com>]

On Fri, 2018-05-04 at 00:02 +0000, Luis R. Rodriguez wrote:
> If you can add Andres Rodriguez <andresx7@gmail.com>, and Greg to your Cc list
> in the future patches that'd be appreciated.
> 
> On Tue, May 01, 2018 at 09:48:18AM -0400, Mimi Zohar wrote:
> > Add an LSM hook prior to allowing firmware sysfs fallback loading.
> > 
> > Signed-off-by: Mimi Zohar <zohar@linux.vnet.ibm.com>
> > Cc: Luis R. Rodriguez <mcgrof@suse.com>
> > Cc: David Howells <dhowells@redhat.com>
> > Cc: Kees Cook <keescook@chromium.org>
> > Cc: Matthew Garrett <mjg59@google.com>
> > ---
> >  drivers/base/firmware_loader/fallback.c | 7 +++++++
> >  include/linux/fs.h                      | 1 +
> >  2 files changed, 8 insertions(+)
> > 
> > diff --git a/drivers/base/firmware_loader/fallback.c b/drivers/base/firmware_loader/fallback.c
> > index 31b5015b59fe..23d2af30474e 100644
> > --- a/drivers/base/firmware_loader/fallback.c
> > +++ b/drivers/base/firmware_loader/fallback.c
> > @@ -651,6 +651,8 @@ static bool fw_force_sysfs_fallback(unsigned int opt_flags)
> >  
> >  static bool fw_run_sysfs_fallback(unsigned int opt_flags)
> >  {
> > +	int ret;
> > +
> >  	if (fw_fallback_config.ignore_sysfs_fallback) {
> >  		pr_info_once("Ignoring firmware sysfs fallback due to sysctl knob\n");
> >  		return false;
> > @@ -659,6 +661,11 @@ static bool fw_run_sysfs_fallback(unsigned int opt_flags)
> >  	if ((opt_flags & FW_OPT_NOFALLBACK))
> >  		return false;
> >  
> > +	/* Also permit LSMs and IMA to fail firmware sysfs fallback */
> > +	ret = security_kernel_read_file(NULL, READING_FIRMWARE_FALLBACK);
> > +	if (ret < 0)
> > +		return ret;
> > +
> >  	return fw_force_sysfs_fallback(opt_flags);
> >  }
> >  
> > diff --git a/include/linux/fs.h b/include/linux/fs.h
> > index 760d8da1b6c7..dc16a73c3d38 100644
> > --- a/include/linux/fs.h
> > +++ b/include/linux/fs.h
> > @@ -2810,6 +2810,7 @@ extern int do_pipe_flags(int *, int);
> >  	id(UNKNOWN, unknown)		\
> >  	id(FIRMWARE, firmware)		\
> >  	id(FIRMWARE_PREALLOC_BUFFER, firmware)	\
> > +	id(FIRMWARE_FALLBACK, firmware)	\
> 
> If you're going to add this perhaps FIRMWARE_FALLBACK_SYSFS as we may later
> get FIRMWARE_FALLBACK_EFI.

>From an IMA signature verification perspective, both are buffer based.
 The file signature is stored as a security xattr.  Without a file
descriptor, the kernel cannot verify the firmware signature.

I don't have a problem with defining another enumeration.  Perhaps
other LSMs will want to be able to differentiate between sysfs and EFI
fallback methods.

Mimi