Re: [PATCH 1/3] module: Invalidate signatures on force-loaded modules

2016-04-27 Thread Rusty Russell
Ben Hutchings writes: > On Tue, 2016-04-26 at 20:07 +0930, Rusty Russell wrote: >> Ben Hutchings writes: >> > - if (info->len > markerlen && >> > + /* >> > +  * Require flags == 0, as a module with version information >> > +  * removed is no longer the module that was signed >> > +  */ >> >

Re: [PATCH 1/3] module: Invalidate signatures on force-loaded modules

2016-04-26 Thread Ben Hutchings
On Tue, 2016-04-26 at 20:07 +0930, Rusty Russell wrote: > Ben Hutchings writes: > > > > Signing a module should only make it trusted by the specific kernel it > > was built for, not anything else.  Loading a signed module meant for a > > kernel with a different ABI could have interesting effects.

Re: [PATCH 1/3] module: Invalidate signatures on force-loaded modules

2016-04-26 Thread Rusty Russell
Ben Hutchings writes: > Signing a module should only make it trusted by the specific kernel it > was built for, not anything else. Loading a signed module meant for a > kernel with a different ABI could have interesting effects. > Therefore, treat all signatures as invalid when a module is > forc

[PATCH 1/3] module: Invalidate signatures on force-loaded modules

2016-04-23 Thread Ben Hutchings
Signing a module should only make it trusted by the specific kernel it was built for, not anything else. Loading a signed module meant for a kernel with a different ABI could have interesting effects. Therefore, treat all signatures as invalid when a module is force-loaded. Signed-off-by: Ben Hut