Re: Add MODULE_ALIAS with cra_driver_name

2006-05-23 Thread Herbert Xu
On Wed, May 24, 2006 at 12:19:21PM +1200, Michal Ludvig wrote: > > Have you got any pointers to the source? Which parts of kernel do this, > where can I learn about it? Or at least some keywords that I can grep > for ;-) Checkout include/asm-i386/uaccess.h and arch/i386/lib/getuser.S. Cheers, --

Re: Add MODULE_ALIAS with cra_driver_name

2006-05-23 Thread Michal Ludvig
Herbert Xu wrote: > On Wed, May 24, 2006 at 11:31:41AM +1200, Michal Ludvig wrote: >> Can we safely handle these pagefaults in the atomic context? Or is it a >> nonsense? > > It should work. You need to setup an exception table to trap the > fault. Have you got any pointers to the source? Which

Re: Add MODULE_ALIAS with cra_driver_name

2006-05-23 Thread Herbert Xu
On Wed, May 24, 2006 at 11:31:41AM +1200, Michal Ludvig wrote: > > Lots of troubles with the fallback thing, isn't it? It is pretty ugly. > Is it possible to implement in the kernel the approach with arbitrary > page faults that I did in phe_sum > (http://www.logix.cz/michal/devel/padlock/phe_su

Re: Add MODULE_ALIAS with cra_driver_name

2006-05-23 Thread Michal Ludvig
Herbert Xu wrote: > On Mon, May 22, 2006 at 03:38:36PM +1200, Michal Ludvig wrote: >> +static void padlock_sha_init(struct crypto_tfm *tfm) >> +{ >> +CTX(tfm)->used = 0; >> +CTX(tfm)->bypass = 0; >> +CTX(tfm)->data = (char*)__get_free_page(GFP_KERNEL); >> +if (!CTX(tfm)->data) >> +

Re: Add MODULE_ALIAS with cra_driver_name

2006-05-23 Thread Herbert Xu
On Mon, May 22, 2006 at 03:38:36PM +1200, Michal Ludvig wrote: > > +static void padlock_sha_init(struct crypto_tfm *tfm) > +{ > + CTX(tfm)->used = 0; > + CTX(tfm)->bypass = 0; > + CTX(tfm)->data = (char*)__get_free_page(GFP_KERNEL); > + if (!CTX(tfm)->data) > + padlock_s

Re: Add MODULE_ALIAS with cra_driver_name

2006-05-21 Thread Herbert Xu
On Mon, May 22, 2006 at 04:19:28PM +1200, Michal Ludvig wrote: > > Then we could add a compile-time option (in Kconfig) for the default and > set it to sha1-i586 once that is available. Or have a list of default > fallbacks like sha1_fallback=sha1-i586,sha1-x86-64,sha1-generic. Neither > is a big

Re: Add MODULE_ALIAS with cra_driver_name

2006-05-21 Thread Michal Ludvig
Herbert Xu wrote: > In general we want to minimise the number of people who need to tweak > configuration parameters. I have nothing against a module parameter > that selected a specific fallback. However, if a large proportion of > people end up having to set the parameter once we get an assemb

Re: Add MODULE_ALIAS with cra_driver_name

2006-05-21 Thread Herbert Xu
On Mon, May 22, 2006 at 03:38:36PM +1200, Michal Ludvig wrote: > > That's configurable with module parameters. Defaults to sha1-generic but > you can easily add sha1_fallback=sha1-i586 to your modprobe command. To > use padlock or sha1-i586 by default you'll have to edit > /etc/modprobe.conf anywa

Re: Add MODULE_ALIAS with cra_driver_name

2006-05-21 Thread Michal Ludvig
Herbert Xu wrote: > Hi Michal: > > On Mon, May 22, 2006 at 02:15:07PM +1200, Michal Ludvig wrote: >> we should be able to autoload modules via their cra_driver_name, not >> only by the module name. I added a few MODULE_ALIAS()es, for now only >> for aes.ko which sets its driver-name and for SHA1/2

Re: Add MODULE_ALIAS with cra_driver_name

2006-05-21 Thread Herbert Xu
Hi Michal: On Mon, May 22, 2006 at 02:15:07PM +1200, Michal Ludvig wrote: > > we should be able to autoload modules via their cra_driver_name, not > only by the module name. I added a few MODULE_ALIAS()es, for now only > for aes.ko which sets its driver-name and for SHA1/256 modules that I > need

Add MODULE_ALIAS with cra_driver_name

2006-05-21 Thread Michal Ludvig
Hi Herbert, we should be able to autoload modules via their cra_driver_name, not only by the module name. I added a few MODULE_ALIAS()es, for now only for aes.ko which sets its driver-name and for SHA1/256 modules that I need as fallbacks for padlock-sha. Should I resend this patch with all soft-