**** BEGIN LOGGING AT Sat Dec 13 02:59:58 2014 Dec 13 16:05:47 lumag_, hi Dec 13 16:05:54 this is collie, same deferral Dec 13 16:05:56 http://pastebin.com/s77YzMHC Dec 13 16:07:35 Hello Dec 13 16:07:45 there is that input scan to remove Dec 13 16:07:57 Yes. And increase scan frequency. Dec 13 16:08:12 And drop the extra "power" handling. Dec 13 16:08:19 And handle home key. Dec 13 16:08:36 btw: there is a deferral, but the lcd works, doesn't it? Dec 13 16:08:46 is ok Dec 13 16:09:19 I'll take a look if changing the order of devices will help with that deferral. Dec 13 16:09:24 on boot and on resume Dec 13 16:09:40 so collie doesn't stay suspended Dec 13 16:10:15 but lcd is fine on resume Dec 13 16:11:52 PM: suspend of devices complete after 839.405 msecs Dec 13 16:11:52 PM: late suspend of devices complete after 5.509 msecs Dec 13 16:11:52 PM: noirq suspend of devices complete after 4.639 msecs Dec 13 16:11:52 PM: noirq resume of devices complete after 725.576 msecs Dec 13 16:11:52 PM: early resume of devices complete after 4.069 msecs Dec 13 16:12:04 the irq changes ? Dec 13 16:14:32 halt/poweroff is ok on collie Dec 13 16:14:48 and it restarts (kexecboot 0.5.9 atm) Dec 13 17:30:40 ant__, I got tired of loosing various pointers, ideas, patches and hacks. I've started to put information here: https://linuxpda.github.io/devices/Sharp/Zaurus/sl5500.html Dec 13 17:31:12 very nice Dec 13 17:31:18 you should have full access once you accept an invitation. Dec 13 17:31:30 at github. Dec 13 17:32:24 If you think of person that also should have access, please invite them. Dec 13 17:34:15 btw, sa1100 patches also got merged. Dec 13 17:47:00 yes, Igot your mail Dec 13 17:47:05 thanks Dec 13 17:47:10 I'll register later Dec 13 17:47:37 btw now waiting the last locomo touches Dec 13 17:47:47 I'll spend some time in userspace Dec 13 17:47:55 i.e. modules autoloading Dec 13 17:48:15 about soc-snd, do you think we should compile it in kernel? Dec 13 17:48:33 I don't know. I'm used to having huge all-in kernels. Dec 13 17:48:43 iirc there was an issue OSS/ALSA Dec 13 17:48:45 But that is my custom. Dec 13 17:48:52 for OPIE is using oss iirc Dec 13 17:48:59 me too I prefer Dec 13 17:49:11 There is an oss emulation layer for alsa. Dec 13 18:08:47 hmm Dec 13 18:09:32 # echo devices > /sys/power/pm_test Dec 13 18:09:52 # echo mem > /sys/power/state Dec 13 18:10:19 suspend debug: Waiting for 5 seconds. Dec 13 18:10:29 it really waits 5 secs suspended Dec 13 18:10:42 root@collie:~# cat /sys/kernel/debug/suspend_stats Dec 13 18:10:42 success: 3 Dec 13 18:10:42 fail: 0 Dec 13 18:10:42 failed_freeze: 0 Dec 13 18:10:42 failed_prepare: 0 Dec 13 18:10:43 failed_suspend: 0 Dec 13 18:10:45 failed_suspend_late: 0 Dec 13 18:10:47 failed_suspend_noirq: 0 Dec 13 18:10:49 failed_resume: 0 Dec 13 18:10:51 failed_resume_early: 0 Dec 13 18:10:53 failed_resume_noirq: 0 Dec 13 18:10:55 failures: Dec 13 18:10:57 last_failed_dev: Dec 13 18:10:59 Dec 13 18:11:01 last_failed_errno: 0 Dec 13 18:11:03 0 Dec 13 18:11:05 last_failed_step: Dec 13 18:11:09 Dec 13 18:11:11 root@collie:~# Dec 13 18:11:13 hm... Dec 13 18:11:21 success: 4 Dec 13 18:12:19 Good Dec 13 18:12:44 https://www.kernel.org/doc/Documentation/power/basic-pm-debugging.txt Dec 13 18:12:47 these steps Dec 13 18:15:59 # echo platform > /sys/power/pm_test Dec 13 18:16:04 same good results Dec 13 18:16:12 success: 8 Dec 13 18:16:38 but once I try normal suspend it wakes up Dec 13 18:19:18 root@collie:~# cat /sys/power/pm_test Dec 13 18:19:18 [none] core processors platform devices freezer Dec 13 18:22:03 success: 13 Dec 13 18:22:07 tried all tests... Dec 13 18:22:19 (even freezer is ok ;) Dec 13 20:46:26 ant__, got disconnected a while ago. Great that it works. Dec 13 20:47:01 * lumag_ likes the simplicity of strongarm - nothing fancy, just simple code (most of the time). Dec 13 22:01:36 it only remains to test suspend without locomo-new patches Dec 13 22:41:42 lumag_: yes, collie stays suspended Dec 13 22:44:15 must be one of the new locomo patches Dec 13 22:44:33 ant__, and nothing in dmesg? Dec 13 22:45:08 What about PM_DEBUG? It should tell, which driver doesn't let it to suspend. Dec 13 22:45:53 nothing Dec 13 22:46:04 I'm s Dec 13 22:46:44 eeing strange things on suspend/resume but surely the defconfig with the new locomo drivers doen't match if the patches are reverted... Dec 13 22:47:43 wrt PM_DEBUG I was hoping to see Dec 13 22:47:45 failures: Dec 13 22:47:46 last_failed_dev: alarm Dec 13 22:47:53 or alike but no errors Dec 13 22:49:03 And if you try to suspend it and it resumes, was there anything in dmesg? Dec 13 22:49:12 I hoped for PM_DEBUG to do the trick. Dec 13 22:52:22 https://www.kernel.org/doc/Documentation/power/suspend-and-interrupts.txt Dec 13 22:52:28 reading taht now... Dec 13 22:52:58 PM: noirq resume of devices complete after 726.478 msecs Dec 13 22:52:58 PM: early resume of devices complete after 4.003 msecs Dec 13 22:53:32 noirq resume Dec 13 23:04:18 hm.. Dec 13 23:04:19 [PATCH] i2c: exynos5: Properly use the "noirq" variants of suspend/resume Dec 13 23:05:21 interesting but only you know if it applies Dec 13 23:05:24 http://lists.infradead.org/pipermail/linux-arm-kernel/2014-June/265044.html Dec 13 23:24:34 zZZZ Dec 13 23:24:35 gn **** ENDING LOGGING AT Sun Dec 14 02:59:58 2014