On Thursday 27 August 2009 00:44:13 Guennadi Liakhovetski wrote: > On Thu, 27 Aug 2009, Hans Verkuil wrote: > > > On Wednesday 26 August 2009 23:55:29 Guennadi Liakhovetski wrote: > > > On Wed, 26 Aug 2009, Mauro Carvalho Chehab wrote: > > > > > > > Em Tue, 25 Aug 2009 17:11:26 +0200 (CEST) > > > > Guennadi Liakhovetski <g.liakhovetski@xxxxxx> escreveu: > > > > > > > > > Hi Mauro > > > > > > > > > > Here they come... The first one is a kernel-sync, I skipped two more > > > > > similar patches for other ARM platforms, because they are not in hg at > > > > > all. The rest are marked "Priority: low" as you requested to stress, that > > > > > they should go in after those three ARM (PXA) patches. I still have a > > > > > couple more patches, which I will also try to prepare and push for 2.6.32, > > > > > but let's get these ones first in. Some patches do not pass checkpatch > > > > > cleanly, but if you check the cumulative diff, it should be clean apart > > > > > from a couple false positives. The last patch in the series does the clean > > > > > up. > > > > > > > > > > Please pull from http://linuxtv.org/hg/~gliakhovetski/v4l-dvb > > > > > > > > > > for the following 35 changesets: > > > > > > [snip] > > > > > > > There are some compilation troubles now on x86_64 compilation: > > > > > > > > /home/v4l/master/v4l/soc_camera_platform.c: In function 'soc_camera_platform_probe': > > > > /home/v4l/master/v4l/soc_camera_platform.c:140: warning: cast from pointer to integer of different size > > > > /home/v4l/master/v4l/soc_camera.c: In function 'soc_camera_init_i2c': > > > > /home/v4l/master/v4l/soc_camera.c:860: warning: cast from pointer to integer of different size > > > > > > > > I'll take a look later and try to fix. > > > > > > Hm, these are lines like > > > > > > subdev->grp_id = (__u32)icd; > > > > > > I think, grp_id should become unsigned long, Hans? > > > > Where and how is grp_id used? This seems to me to be an abuse of this field. > > It is really not meant to contain a pointer. > > > > I can't find any actual use of this field in this driver, so either I'm > > missing something or this is simply not needed at all. > > We discussed this before, here's an excerpt from your earlier reply to my > mail: > > <quote> > > In fact, what I actually need is to call a specific method, if it is > > implemented, from one specific subdevice, and get its error code - not > > from all and not until the first error. I am currently abusing your > > grp_id for this, but it might eventually be better to add such a wrapper. > > That's actually what grp_id is intended for (or one of the intended uses > at least). > </quote> I'm pretty sure I didn't realize at the time that you would attempt to store a pointer in grp_id. > > But, I think, since then I switched to using v4l2_subdev_call() > _everywhere_, so, it might well be, that this grp_id can go completely. > Not sure off the top of my head though. If you only use v4l2_subdev_call then you do not need to set grp_id at all. So these lines can be removed. Regards, Hans -- Hans Verkuil - video4linux developer - sponsored by TANDBERG Telecom -- To unsubscribe from this list: send the line "unsubscribe linux-media" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html