**** BEGIN LOGGING AT Wed Dec 09 02:59:57 2009
Dec 09 03:06:34 <zhenhua1> holtmann: do you know how to create a pseudo tty device in user space?
Dec 09 03:15:55 <denkenz> man openpty
Dec 09 04:11:16 <denkenz> zhenhua1: I don't see how release_specific is supposed to work
Dec 09 04:11:41 <denkenz> is there a poll somewhere?
Dec 09 04:19:03 <zhenhua1> denkenz: i tested on iphone, and we will get chld update when you release a call in three way calls
Dec 09 04:19:33 <denkenz> chld update?
Dec 09 04:19:41 <denkenz> you mean callheld switches to 0?
Dec 09 04:19:50 <zhenhua1> right
Dec 09 04:20:03 <zhenhua1> can u give an example that doesn't work
Dec 09 04:20:20 <zhenhua1> we only use release_specific when call num > 1
Dec 09 04:20:37 <denkenz> No I'm just wondering what triggers the poll
Dec 09 04:20:43 <denkenz> So today its triggered by callheld=0
Dec 09 04:20:52 <zhenhua1> also callheld=1
Dec 09 04:20:52 <denkenz> which polls clcc
Dec 09 04:21:30 <denkenz> yep ok
Dec 09 04:22:41 <zhenhua1> so...
Dec 09 04:22:45 <zhenhua1> is that ok?
Dec 09 04:24:48 <zhenhua1> in callheld=2, all calls are held. so no call has been released before.
Dec 09 04:25:40 <denkenz> yep its fine
Dec 09 04:25:44 <denkenz> The 3rd patch...
Dec 09 04:26:05 <zhenhua1> yes
Dec 09 04:26:05 <denkenz> in poll_clcc you don't reset clcc_source to 0
Dec 09 04:26:21 <zhenhua1> ops. good catch
Dec 09 04:26:38 <zhenhua1> you're right.
Dec 09 04:26:58 <denkenz> also
Dec 09 04:27:10 <denkenz> +               if (nc) {
Dec 09 04:27:12 <denkenz> +                       if (nc->status == CALL_STATUS_ACTIVE)
Dec 09 04:27:13 <denkenz> +                               num_active++;
Dec 09 04:27:15 <denkenz> +
Dec 09 04:27:16 <denkenz> +                       if (nc->status == CALL_STATUS_HELD)
Dec 09 04:27:18 <denkenz> +                               num_held++;
Dec 09 04:27:19 <denkenz> +               }
Dec 09 04:27:21 <denkenz> +
Dec 09 04:27:23 <denkenz> I would probably prefer if (nc && nc->status...)
Dec 09 04:27:29 <denkenz> than a nested if
Dec 09 04:27:31 <zhenhua1> no probs
Dec 09 04:29:04 <zhenhua1> any others
Dec 09 04:29:08 <zhenhua1> how about patch 4
Dec 09 04:29:56 <denkenz> I actually think that is useless
Dec 09 04:30:01 <denkenz> But it doesn't hurt
Dec 09 04:30:37 <denkenz> Actually..
Dec 09 04:30:46 <denkenz> The problem with CLIP is the broken devices
Dec 09 04:30:53 <zhenhua1> in case the phone doesn't have CLIP
Dec 09 04:30:55 <denkenz> I'd rather not poll those
Dec 09 04:31:14 <denkenz> Especially your weird LG device that messes with call IDs
Dec 09 04:31:19 <zhenhua1> but you cannot receive incoming update 
Dec 09 04:31:21 <zhenhua1> heh
Dec 09 04:31:58 <zhenhua1> anyway, you can decide accept or not. i am fine.
Dec 09 04:31:59 <denkenz> we don't, we track callsetup
Dec 09 04:32:11 <denkenz> We don't poll from RING
Dec 09 04:33:01 <denkenz> However, we must notify the call in case the CLIP never arrives
Dec 09 04:33:16 <denkenz> So modify patch 4 to do that
Dec 09 04:34:48 <zhenhua1> you poll from callsetup=1 ?
Dec 09 04:35:26 <zhenhua1> what's difference
Dec 09 04:36:05 <denkenz> nope
Dec 09 04:36:08 <denkenz> callsetup=1 is ignored
Dec 09 04:37:08 <zhenhua1> we notify call in CLIP. so how do you notify call without CLIP
Dec 09 04:44:54 <denkenz> do the same CLIP_TIMEOUT thingy
Dec 09 04:45:04 <denkenz> just don't poll, but instead notify the call as is
Dec 09 04:45:11 <denkenz> its created with unknown caller
Dec 09 04:45:44 <denkenz> simply find incoming call and notify
Dec 09 04:47:12 <denkenz> might want to make the timeout just a tad bigger, like 500
Dec 09 04:50:37 <zhenhua1> ok. i understand your point
Dec 09 04:51:02 <zhenhua1> for the patch I just send out. the patch number should be 3/4 instead of 3/3 in attachment.
Dec 09 04:52:57 <denkenz> drivers/hfpmodem/voicecall.c: In function ?hfp_release_specific?:
Dec 09 04:52:58 <denkenz> drivers/hfpmodem/voicecall.c:532: error: ?req? may be used uninitialized in this function
Dec 09 04:53:16 <zhenhua1> let me check
Dec 09 04:54:11 <zhenhua1> ops. will fix that.
Dec 09 04:57:02 <denkenz> Ok, resubmit #1, get me a CLIP patch as discussed and I think we got the protocol part done
Dec 09 04:57:31 <denkenz> Now just need audio
Dec 09 04:58:53 <zhenhua1> that's great
Dec 09 05:37:25 <denkenz> zhenhua1: Patch 4, don't resuse clcc_source for clip, just pick another variable
Dec 09 05:40:49 <denkenz> and make sure to remove the source in _remove
Dec 09 05:48:24 <zhenhua1> denkenz: if so, i have to add a varialbe in voicecall_data again
Dec 09 05:49:07 <zhenhua1> something like clip_source
Dec 09 05:49:12 <zhenhua1> is that ok?
Dec 09 05:51:18 <denkenz> yep
Dec 09 05:51:32 <zhenhua1> again. why not reuse clcc_source? 
Dec 09 05:51:58 <zhenhua1> if it's incoming call, we are safe because only one incoming call allowed
Dec 09 05:53:36 <zhenhua1> maybe we can rename it to something like 'timer' or others
Dec 09 05:54:52 <denkenz> Remember, code should be clear and readable
Dec 09 05:55:10 <denkenz> wasting extra 4 bytes is a small price to pay 
Dec 09 05:55:40 <zhenhua1> ...ok.
Dec 09 05:57:36 <denkenz> And it is actually safer
Dec 09 05:57:47 <denkenz> Think of the end all case where a call goes from waiting to ringing
Dec 09 06:01:26 <zhenhua1> i have resend the patch. even if waiting to ringing, i don't see the problem.
Dec 09 06:05:15 <denkenz> I know the variables don't overlap, but in 6 months you'll forget, make the code readable from the very beginning
Dec 09 06:31:30 <martin_lab> holtmann, denkenz, we have a new different modem desc on my f3607 modem on Samos platform. That make the modem can not be detected, just like my f3507 card issue. 
Dec 09 06:31:52 <denkenz> so submit the id
Dec 09 06:32:29 <denkenz> This is getting silly though
Dec 09 06:32:33 <denkenz> How many IDs do they have :P
Dec 09 06:35:21 <martin_lab> denkenz, I will submitted the patch soon. 
Dec 09 06:40:16 <martin_lab> denkenz, I think we need to add string table to hold these IDs, so it is convenient for us the add new ID. 
Dec 09 06:40:28 <denkenz> that is probably a good idea
Dec 09 07:18:03 <martin_lab> denkenz, forget the set git_config on my Samos platform. ;-), So the author of the patch is Root. Will resend it. sorry. 
Dec 09 09:44:11 <dharman> 'giorno giovani!!!!
Dec 09 09:44:18 <dharman> come va la nuova esperienza domestica???
Dec 09 09:44:19 <dharman> :D
Dec 09 09:44:32 <dharman> ooops!
Dec 09 09:45:34 <dharman> sorry
**** ENDING LOGGING AT Mon Dec 14 23:56:40 2009