tmlind | freemangordon: maybe you want to use a completion? | 04:26 |
---|---|---|
freemangordon | tmlind: it uses wait queue behind the scenes, so the same question is still valid. I guess I'll just implement it like that | 07:38 |
tmlind | freemangordon: ok yeah, no idea about the latencies here | 07:41 |
freemangordon | tmlind: for some reason, on n900 usleep_range(728, 728) sleeps for 10955 us :) | 08:05 |
freemangordon | seems there is something wrong with hrtimers on omap3 | 08:05 |
tmlind | hmm sounds like there's some issue somewhere | 08:08 |
freemangordon | or, when asked for 307us sleeps for 2166748ns. Unless I am doing something stupid when logging the time https://pastebin.com/tMuYC1Cy | 08:11 |
freemangordon | the result https://pastebin.com/wDaFYHGX | 08:12 |
freemangordon | no wonder IR TX does not work :) | 08:12 |
tmlind | maybe run timer tests like cyclictest to check which timers are broken | 08:13 |
freemangordon | tmlind: ktime_t is timestamp in nanoseconds, right? | 08:14 |
tmlind | cyclictest and others are in the rt-tests package | 08:14 |
freemangordon | in the repos? | 08:14 |
tmlind | might be | 08:14 |
freemangordon | lemme try to find | 08:14 |
* tmlind bbl | 08:15 | |
freemangordon | tmlind: no idea how to read the results: | 08:20 |
freemangordon | T: 0 ( 5683) P: 0 I:1000 C: 10000 Min: 172 Act: 383 Avg: 446 Max: 4979 | 08:20 |
freemangordon | WARN: measured clock resolution approximately: 30517 nsec | 08:31 |
tmlind | freemangordon: the max value is way too long it seems | 09:16 |
tmlind | hmm well it's high but it's not with rt patches | 09:22 |
freemangordon | tmlind: well, average 446 us does not look good to me | 09:50 |
freemangordon | given that timer resolution is 30us | 09:50 |
tmlind | is this with the mainline v6.5 or something else? | 09:51 |
freemangordon | 6.1.48 | 09:51 |
freemangordon | this https://github.com/maemo-leste/droid4-linux/tree/maemo-6.1.y | 09:52 |
tmlind | ok | 09:52 |
tmlind | well CONFIG_NO_HZ_IDLE=y is needed for pm, but also causes the clockevent to be often reprogrammed for the next timer so there's some extra overhead | 09:53 |
tmlind | sched_clock() is slow because there is no localtimer | 09:54 |
freemangordon | have to run, ttyl | 09:59 |
tmlind | later | 10:07 |
Wizzup | uvos: do we have a way to enter dtmf in sphone? | 15:40 |
bencoh | Wizzup: I don't see any either at least | 16:26 |
bencoh | which reminds me ... does anyone see odd cases of missed calls / sphone not making the phone vibrate / interrupted vibration? | 16:28 |
Wizzup | no, but I have had repeated smses still some times | 16:33 |
bencoh | Wizzup: I've had repeated SMSes for a day (and counting) | 17:42 |
bencoh | (I switched yesterday to d4 as a main phone) | 17:43 |
bencoh | is there a workaround to that btw? | 17:43 |
Wizzup | I don't know why it happens, but I think it has something to do with setting the display on/off (by that I mean what mce writes to the modem to silence the signal strength notifications on display on/off) | 17:51 |
Wizzup | because I get the repeated smses the moment I turn on the display | 17:51 |
Wizzup | check @ main phone, I finally have some peace so I'm going to revive the sphone tp part | 17:52 |
Wizzup | then we will hopefully finally also have sip/xmpp calls | 17:52 |
bencoh | how does the cellular part work nowadays? sphone takes care of everything (ie ofono, pulseaudio routing, vibration, ...)? | 18:00 |
bencoh | I noticed that I often get SMSes (either repeating, or just regular SMSes that didn't show up thus far) whenever I call or get a call | 18:01 |
bencoh | 41 | 18:02 |
bencoh | (woops) | 18:02 |
tmlind | Wizzup, uvos: dtmf should be trivial to implement, there's the dtmf shell script at https://github.com/tmlind/droid4-sms-tools | 18:52 |
Wizzup | thanks | 19:08 |
Wizzup | sicelo: arno11: I have the n900 set up with serial again, lmk if I need to test anything else | 21:49 |
freemangordon | tmlind: any idea where to look for why usleep_range() misbehaves? | 23:27 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!