On Fri, Sep 30, 2022 at 06:01:16PM +0200, Sebastian Andrzej Siewior wrote: > On 2022-09-30 17:56:17 [+0200], Daniel Wagner wrote: > > On Fri, Sep 30, 2022 at 03:13:30PM +0200, Sebastian Andrzej Siewior wrote: > > > On 2022-09-30 10:37:19 [+0200], Daniel Wagner wrote: > > > > I guess I should also pick your version of the local-lock fixup for > > > > v4.19-rt? > > > > > > If it is missing, yes please. > > > > I've came up with a very similar patch a few releases ago. But to keep > > things consistent, I think it makes sense to use your version. > > I just sent Mark a small update but if v4.19 has local_lock_t and then > things may be smooth. IIRC, local_lock_t is just part of the -rt set. It looks like I need the same fix as Mark for !rt builds. The 'struct local_irq_lock' type is unknown 2022-10-05 20:50:36 - /home/wagi/rt/v4.19-rt/drivers/char/random.c:233:31: error: field ‘lock’ has incomplete type 2022-10-05 20:50:36 - 233 | struct local_irq_lock lock; Shall I wrap it in a #ifdef section? > In general I would prefer to have the same patches > across each kernels because if something breaks then you can have > at least the same set of patches in each kernel. Sure it's a good idea. I'll pick all your patches from https://lore.kernel.org/all/20220819092446.980320-1-bigeasy@xxxxxxxxxxxxx/ too.
