*** fiferboy has quit IRC | 00:05 | |
*** tbf has quit IRC | 00:21 | |
*** rcampbell has joined #maemo-devel | 00:24 | |
*** silbo___ has quit IRC | 00:29 | |
*** SpeedEvil has quit IRC | 00:34 | |
*** DocScrutinizer51 has quit IRC | 00:36 | |
*** SpeedEvil has joined #maemo-devel | 00:36 | |
*** tbf has joined #maemo-devel | 00:36 | |
*** Docscrutemp has joined #maemo-devel | 00:36 | |
*** Docscrutemp is now known as DocScrutinizer51 | 00:36 | |
*** lbt has quit IRC | 00:41 | |
*** DocScrutinizer51 has quit IRC | 00:45 | |
*** Docscrutemp has joined #maemo-devel | 00:45 | |
*** Docscrutemp is now known as DocScrutinizer51 | 00:45 | |
*** tbf has quit IRC | 00:46 | |
*** JPohlmann has quit IRC | 00:50 | |
*** Ian-- has quit IRC | 01:02 | |
*** Ian-- has joined #maemo-devel | 01:03 | |
*** Acedip has quit IRC | 01:13 | |
*** achipa has quit IRC | 01:16 | |
*** Mobsan has quit IRC | 01:22 | |
*** edisson has joined #maemo-devel | 01:23 | |
*** User_ has joined #maemo-devel | 01:26 | |
*** SpeedEvil has quit IRC | 01:27 | |
*** shinkamui has joined #maemo-devel | 01:27 | |
*** User_ has quit IRC | 01:29 | |
*** T7g has quit IRC | 01:36 | |
*** T7g has joined #maemo-devel | 01:41 | |
*** rsalveti has quit IRC | 01:45 | |
*** rsalveti has joined #maemo-devel | 01:49 | |
*** DocScrutinizer51 has quit IRC | 01:54 | |
*** Docscrutemp has joined #maemo-devel | 01:54 | |
*** Docscrutemp is now known as DocScrutinizer51 | 01:55 | |
*** DocScrutinizer51 has quit IRC | 02:05 | |
*** Docscrutemp has joined #maemo-devel | 02:05 | |
*** Docscrutemp is now known as DocScrutinizer51 | 02:06 | |
*** SpeedEvil has joined #maemo-devel | 02:32 | |
*** edisson has quit IRC | 02:58 | |
*** edisson has joined #maemo-devel | 02:58 | |
*** lcuk2 has joined #maemo-devel | 03:08 | |
*** swc|666 has quit IRC | 03:10 | |
*** ptl_demands_PR13 has quit IRC | 03:10 | |
*** ptl_demands_PR13 has joined #maemo-devel | 03:10 | |
*** lcuk has quit IRC | 03:12 | |
*** lcuk2 is now known as lcuk | 03:14 | |
*** lcuk has joined #maemo-devel | 03:14 | |
*** njsf_ has joined #maemo-devel | 03:19 | |
*** edisson has quit IRC | 04:40 | |
*** SpeedEvil has quit IRC | 04:42 | |
*** SpeedEvil has joined #maemo-devel | 04:43 | |
*** edisson has joined #maemo-devel | 04:51 | |
*** GeneralAntilles has quit IRC | 05:12 | |
*** GeneralAntilles has joined #maemo-devel | 05:16 | |
*** swc|666 has joined #maemo-devel | 05:33 | |
*** njsf_ has quit IRC | 05:33 | |
*** swc|666 has quit IRC | 06:03 | |
*** edisson has quit IRC | 06:33 | |
*** njsf_ has joined #maemo-devel | 06:56 | |
*** swc|666 has joined #maemo-devel | 06:59 | |
*** t7g_ has joined #maemo-devel | 07:13 | |
*** T7g has quit IRC | 07:16 | |
*** DocScrutinizer51 has quit IRC | 07:33 | |
*** DocScrutinizer has quit IRC | 07:33 | |
*** DocScrutinizer has joined #maemo-devel | 07:33 | |
*** Docscrutemp has joined #maemo-devel | 07:34 | |
*** Docscrutemp is now known as DocScrutinizer51 | 07:34 | |
*** microlith has quit IRC | 08:15 | |
*** microlith has joined #maemo-devel | 08:16 | |
*** njsf_ has quit IRC | 08:22 | |
*** swc|666 has quit IRC | 08:40 | |
*** rsalveti has quit IRC | 09:06 | |
*** rcampbell has quit IRC | 09:12 | |
*** ppenz has joined #maemo-devel | 09:51 | |
*** rsalveti has joined #maemo-devel | 09:53 | |
*** csaavedra has joined #maemo-devel | 10:00 | |
*** lbt has joined #maemo-devel | 10:14 | |
*** DocScrutinizer51 has quit IRC | 10:14 | |
*** Docscrutemp has joined #maemo-devel | 10:15 | |
*** Docscrutemp is now known as DocScrutinizer51 | 10:15 | |
*** tbf has joined #maemo-devel | 10:22 | |
*** DocScrutinizer51 has quit IRC | 10:31 | |
*** Docscrutemp has joined #maemo-devel | 10:32 | |
*** Docscrutemp is now known as DocScrutinizer51 | 10:32 | |
*** Mobsan has joined #maemo-devel | 10:41 | |
*** kamui__ has joined #maemo-devel | 11:16 | |
*** shinkamui has quit IRC | 11:21 | |
*** DocScrutinizer51 has quit IRC | 11:41 | |
*** Docscrutemp has joined #maemo-devel | 11:43 | |
*** Docscrutemp is now known as DocScrutinizer51 | 11:43 | |
*** dazo_afk is now known as dazo | 11:46 | |
scoopr | since pr1.2, in my app exit_app_view (task switcher) started going directly in to dashboard, instead of the task switcher. I found the culprit to be that my window is in portrait mode. If I unset portrait property before taskswitching, and then reset it on EnterNotify, it works as a workaround | 11:52 |
---|---|---|
scoopr | but if I do ctrl-backspace for taskswitcher, then it again goes to dashboard, because I won't know to unset the portrait mode property | 11:53 |
scoopr | this used to work in pr1.1, so I would call this a regression | 11:53 |
scoopr | is there something magical that needs to change or should I just file a bug? | 11:53 |
*** achipa has joined #maemo-devel | 11:58 | |
*** DocScrutinizer51 has quit IRC | 12:21 | |
*** Docscrutemp has joined #maemo-devel | 12:22 | |
*** Docscrutemp is now known as DocScrutinizer51 | 12:22 | |
*** DocScrutinizer51 has quit IRC | 12:32 | |
*** Docscrutemp has joined #maemo-devel | 12:33 | |
*** Docscrutemp is now known as DocScrutinizer51 | 12:33 | |
*** rsalveti has quit IRC | 13:47 | |
*** rsalveti has joined #maemo-devel | 14:26 | |
*** DocScrutinizer51 has quit IRC | 14:33 | |
*** dazo has quit IRC | 14:50 | |
*** dazo_ has joined #maemo-devel | 14:51 | |
*** dazo_ is now known as dazo | 14:52 | |
*** StippenG has joined #maemo-devel | 14:57 | |
*** JPohlmann has joined #maemo-devel | 15:01 | |
*** Docscrutemp has joined #maemo-devel | 15:04 | |
*** Docscrutemp is now known as DocScrutinizer51 | 15:05 | |
StippenG | Hello! Is there a good way of counting cpu cycles between two points in a program on fremantle? I suppose it's possible using inline assembly at least, but are there some C functions for this too? | 15:05 |
*** edisson has joined #maemo-devel | 15:05 | |
SpeedEvil | yes | 15:10 |
SpeedEvil | actually - what do you mean by cycles | 15:10 |
SpeedEvil | do you mean one or two, or a few million? | 15:10 |
SpeedEvil | /proc/nnn/stat may be useful for the latter | 15:11 |
StippenG | I would like to time some opengl commands mostly, and I figure reading some cpu cycle count would be a good way. It doesn't have to be super accurate, but shouldn't have a high overhead either. I'll look at /proc/nnn/stat and see if that might be enough, but some function that just reads a clock value directly seems like a better idea if that exists | 15:14 |
Wolfie | i'm assuming, you want to count the cpu clocks for your current process? | 15:15 |
Wolfie | and not the whole system... multitasking might skew that number a bit ;) | 15:15 |
StippenG | If that is possible yes, otherwise I figure I will have to run the function multiple times and get an average time to account for other processes running in between | 15:17 |
*** tbf has quit IRC | 15:17 | |
StippenG | I'm a bit worried that multitasking could skew the results a bit yes :) since there are quite a few processes running on the N900 | 15:17 |
StippenG | Should still give me a nice idea of memory transfer times to the GPU for example I figure | 15:18 |
SpeedEvil | gettimeofday before and after? | 15:20 |
*** dazo has quit IRC | 15:23 | |
SpeedEvil | /proc/timer_list has the time at the top. There is certainly a better way to call this thoushg. Wander around in the kernel docs | 15:24 |
StippenG | Ah, that would be a good backup plan if nothing better comes around. It seems to have microsecond precision at least ;) I just heard that for some systems you can quite easily recover a clock count since power on. | 15:24 |
*** dazo has joined #maemo-devel | 15:24 | |
zokier | clock_gettime can have up to nanosecond precision iirc | 15:27 |
StippenG | Ah, thanks! clock_gettime with CLOCK_PROCESS_CPUTIME_ID ("High-resolution per-process timer from the CPU.") seems like it would be worth trying | 15:30 |
*** JPohlmann has quit IRC | 15:32 | |
StippenG | Trying to do some video decoding on the N900 GPU, and need to find out the cost of uploading textures and coordinates to GPU memory and such to decide if it's worth it :) | 15:33 |
zokier | im not sure if cpu time is what you want if you want to account for time used by the gpu | 15:35 |
StippenG | hmm, I guess it depends. If the cpu is blocked during a texture upload to the GPU, the CPU time will tell me how much time the upload took pretty much, right? If the CPU is not blocked, but only sets up some DMA transfer, I suppose I will have to use system time | 15:44 |
StippenG | and figure out wether the transfer is actually completed when the function returns | 15:45 |
*** fiferboy has joined #maemo-devel | 15:51 | |
*** fiferboy has quit IRC | 15:51 | |
*** fiferboy has joined #maemo-devel | 15:51 | |
*** fiferboy has quit IRC | 15:52 | |
*** fiferboy has joined #maemo-devel | 15:54 | |
*** fiferboy has quit IRC | 15:54 | |
*** fiferboy has joined #maemo-devel | 15:54 | |
*** fiferboy has quit IRC | 16:01 | |
*** fiferboy has joined #maemo-devel | 16:05 | |
*** fiferboy has quit IRC | 16:05 | |
*** fiferboy has joined #maemo-devel | 16:05 | |
SpeedEvil | has anyone else see anything touching /home deadlock? | 16:14 |
*** achipa has quit IRC | 16:35 | |
*** achipa has joined #maemo-devel | 16:44 | |
*** ppenz has quit IRC | 17:05 | |
*** aboyer has quit IRC | 17:19 | |
*** JPohlmann has joined #maemo-devel | 17:21 | |
*** tbf has joined #maemo-devel | 17:49 | |
*** zhuangyao has joined #maemo-devel | 17:54 | |
*** zhuangyao has quit IRC | 18:12 | |
*** silbo_ has joined #maemo-devel | 18:13 | |
*** VDVsx has joined #maemo-devel | 18:31 | |
*** silbo_ has quit IRC | 18:38 | |
*** rsalveti has quit IRC | 20:02 | |
*** dazo is now known as dazo_afk | 20:07 | |
*** rcampbell has joined #maemo-devel | 20:09 | |
*** Aranel_ has joined #maemo-devel | 20:27 | |
*** Aranel_ has quit IRC | 20:27 | |
*** Aranel has quit IRC | 20:28 | |
*** dazo_afk is now known as dazo | 20:29 | |
*** VDVsx has quit IRC | 20:33 | |
*** lcuk2 has joined #maemo-devel | 20:35 | |
*** lcuk has quit IRC | 20:37 | |
*** lcuk2 is now known as lcuk | 20:37 | |
*** lcuk has quit IRC | 20:37 | |
*** lcuk has joined #maemo-devel | 20:37 | |
*** csaavedra has quit IRC | 20:38 | |
*** JPohlmann has quit IRC | 20:42 | |
*** rsalveti has joined #maemo-devel | 20:52 | |
*** lostinmirkwood has joined #maemo-devel | 21:02 | |
*** achipa has quit IRC | 21:08 | |
*** swc|666 has joined #maemo-devel | 21:21 | |
*** fiferboy_ has joined #maemo-devel | 22:13 | |
*** fiferboy_ has quit IRC | 22:13 | |
*** fiferboy_ has joined #maemo-devel | 22:13 | |
*** fiferboy has quit IRC | 22:14 | |
*** lostinmirkwood has quit IRC | 22:26 | |
*** csaavedra has joined #maemo-devel | 22:32 | |
*** csaavedra has quit IRC | 22:38 | |
*** apoi has quit IRC | 23:08 | |
*** dazo is now known as dazo_afk | 23:09 | |
*** apoi has joined #maemo-devel | 23:13 | |
*** shinkamui has joined #maemo-devel | 23:16 | |
*** kamui__ has quit IRC | 23:20 | |
*** tbf has quit IRC | 23:36 | |
*** edisson has quit IRC | 23:40 | |
*** silbo_ has joined #maemo-devel | 23:56 |
Generated by irclog2html.py 4.0.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!