Re: [PATCH 1/7] fs: introduce kernel_pread_file* support

From: Luis Chamberlain
Date: Fri Aug 23 2019 - 17:29:37 EST


On Fri, Aug 23, 2019 at 12:55:30PM -0700, Scott Branden wrote:
> Hi Takashi
>
> On 2019-08-23 5:29 a.m., Takashi Iwai wrote:
> > On Thu, 22 Aug 2019 21:24:45 +0200,
> > Scott Branden wrote:
> > > Add kernel_pread_file* support to kernel to allow for partial read
> > > of files with an offset into the file. Existing kernel_read_file
> > > functions call new kernel_pread_file functions with offset=0 and
> > > flags=KERNEL_PREAD_FLAG_WHOLE.
> > Would this change passes the security check like ima?
> > I thought security_kernel_post_read_file() checks the whole content
> > for calculating the hash...
>
> It passes the fw_run_tests.sh.  How do you test the firmware loader passes
> this security check?

Its not a security check per code, its an audit of the code, to ensure
that no new cases are not covered and its why I had CC'd Mimi. The
question lies in *if* the approach exposes a new interface which cannot
be attested. Its unclear to me if we can attest currently through
security modules the fallback interface, as there are not APIs with a
respective callback yet.

Luis