IRC log of #maemo-ssu for Thursday, 2012-03-15

*** NIN101 has quit IRC00:12
*** Pali has quit IRC00:37
*** BCMM has joined #maemo-ssu00:52
*** LaoLang_cool has joined #maemo-ssu01:00
DocScrutinizerkp in its recent form won't go into cssu01:03
freemangordon_what is the recent form of kp?01:05
*** M4rtinK has quit IRC01:07
freemangordon_DocScrutinizer, will you be more specific, as I am not aware of any problems in KP50, while your statement is too definitive01:13
DocScrutinizerpowertop doesn't work.01:14
DocScrutinizerjust for one point01:14
freemangordon_what?01:14
*** M4rtinK has joined #maemo-ssu01:15
freemangordon_DocScrutinizer, who said powertop does not work with KP? It works for me.01:17
DocScrutinizeruhuh01:17
DocScrutinizerpastebin it01:18
DocScrutinizerI bet first 3 lines look like:01:20
DocScrutinizerC#      | Ratio  | Avg/dura | Frequency | Ratio01:20
DocScrutinizer--------+--------+----------+-----------+--------+01:20
DocScrutinizer     C0 |   1.5% |          |  1150 MHz |   nan% |01:20
freemangordon_so? is it different with omap1? is there a reported bug for that?01:22
DocScrutinizerC#      | Ratio  | Avg/dura | Frequency | Ratio01:23
DocScrutinizer--------+--------+----------+-----------+--------+01:23
DocScrutinizer     C0 |   0.4% |          |   600 MHz |   0.0% |01:23
DocScrutinizer     C1 |   0.0% |    0.2ms |   550 MHz |   0.0% |01:23
DocScrutinizerNFC if anybody reported a bug for that, I don't care if anybody did01:23
freemangordon_and that is better? in what extact way?01:24
DocScrutinizerc'mon, either you want to make me angry, or you are actually too tired01:24
freemangordon_the information for C0 is exactly the same, hold on to paste mine01:25
freemangordon_   C0 |  10.5% |          |  1150 MHz |   nan% |01:25
DocScrutinizerpfff, you probably should finish reading the lines, not stop at first 30%01:25
freemangordon_     C1 |   0.0% |    0.4ms |01:25
DocScrutinizeractually your way to argue is in line with quality of PK01:26
freemangordon_yeah, sure01:26
*** LinuxCode has quit IRC01:27
DocScrutinizeron a first cursory glance it looks just like it works same way as stock01:28
DocScrutinizerooh, it doesn't report *any* duratins for different clockspeeds? nah, who cares, it's PK, it's cool01:29
DocScrutinizer(just to premept your next objection: I'm well aware the problem is mainly in powertop itself here)01:30
DocScrutinizerbut on CSSU we have to live with legacy01:30
DocScrutinizerit's meant to stay compatible to stock - this includes powertop as well as any other things that show weird strange behaviour on PK01:31
merlin1991wtf why is there a tmp in my documents01:31
merlin1991must have been on of those days where I just name my folders tmp :D01:32
freemangordonDocScrutinizer, it is the first time I hear of that problem. And as I am not the kind of men who just says "it is KP so it is cool", the correct way to deal of such things it to report them.01:33
DocScrutinizerit's been known since PK0.000.0101:33
freemangordonNot to make statements like "No, this piece of crap will not make it into CSSU"01:33
freemangordonWho knows that you?01:34
freemangordonbecause I am not aware ot the problem and it is not reported on the bigtracker01:34
DocScrutinizeragain, I don't ccare who does01:34
DocScrutinizerand actually there's no proper way to fix this bug01:35
DocScrutinizerozher than breaking PK's extended clock range01:36
DocScrutinizersince powertop is closed01:36
freemangordonWell, at least there is proper way to report it. And once it is reported in a proper way I am sure it can be workarounded somehow. And even it cannot be fixed/workarounded, do you really think this is a showstopper bug?01:37
*** trbs has quit IRC01:38
*** ekze-nyan has joined #maemo-ssu01:43
*** ekze has quit IRC01:45
*** FireFly has quit IRC01:45
*** LaoLang_cool has quit IRC01:45
DocScrutinizerthat'S exactly why I was reluctant to answer at all first instance - now you again try to drive me into a quantity competition, while I clearly said it's just an example01:46
DocScrutinizerfact is that PK is not about bug fixing. Actually it introduced more bugs than it fixed01:47
DocScrutinizerand01:47
DocScrutinizer**obviously it's not needed for CSSU - IOW there's not a single dependency to PK in CSSU**01:48
DocScrutinizerso it won't go into CSSU base/core package, same way OMP won't and nikocam never should have01:48
DocScrutinizerusers are free to install PK independently of CSSU any time they like01:49
DocScrutinizernot the faintest reason to include it01:49
*** FireFly has joined #maemo-ssu01:49
DocScrutinizerif there was a baseline PK that had *only* the few real bugfixes, then you could maybe convince me that it should get shipped with CSSU01:51
freemangordonDocScrutinizer, you are correct that there is no dependency. But if omap1 is to make it into CSSU, with only bugfixes, who will maintain it? So far the only ones who still keep some development/backporting on the kernel front is me and Pali. NFC if Pali wants to maintain/support two kernels, but for dure I don't.01:51
freemangordonHave in mind that I don't feel myself like "kernel developer"01:51
DocScrutinizerso just keep PK a separate entity01:52
DocScrutinizerand everybody's happy01:52
freemangordonWhith resources we have right now, merging KP and CSSU is the best option AIUI01:52
DocScrutinizernonsense01:52
freemangordonAnd what about bug reports against omap1?01:52
DocScrutinizerwhat about it?01:53
DocScrutinizerpester Nokia to fix it?01:53
freemangordonAre you kidding me?01:53
DocScrutinizeror suggest users that want to do sth about it that they shall install PK01:54
freemangordonDid you check KP50 build result? As it is the same as KP49, it does not appear in maemo extras web interface01:55
DocScrutinizerbut WTH shall somebody that wants CSSU rootfs be forced to also install PK, and WHY for the love of god shall all the OC fanboys be forced to install CSSU rootfs and then report bugs against CSSU that are caused by their idiocy01:55
DocScrutinizernow that's for sure an invincible argument for including it to CSSU, as an integral component01:56
freemangordonAnd noone knows how long will it take until KP50 can be promoted to -testing01:56
amiconnKP does maintain time_in_state statistics01:57
DocScrutinizernite01:57
DocScrutinizeraniI know that01:57
DocScrutinizeramiconn: ^^^01:57
amiconnSo if powertop doesn't handle those, it's a powertop bug imo, not a KP bug01:57
DocScrutinizeramiconn: but it does in a way incompatible to powertop01:58
DocScrutinizeramiconn: please read backscroll - I hoped to stop this argument already01:58
DocScrutinizernayway, got some other duties in a few hours, so bbl01:59
freemangordonDocScrutinizer, nite01:59
*** M4rtinK has quit IRC02:44
*** freemangordon_ has left #maemo-ssu03:14
*** LaoLang_cool has joined #maemo-ssu04:12
*** wmarone has quit IRC04:16
*** wmarone_ has joined #maemo-ssu04:16
*** wmarone_ has quit IRC04:26
*** wmarone__ has joined #maemo-ssu04:26
*** LaoLang_cool has quit IRC04:30
*** LaoLang_cool has joined #maemo-ssu04:30
*** amiconn has quit IRC05:10
*** amiconn_ has joined #maemo-ssu05:10
*** amiconn_ is now known as amiconn05:10
*** LaoLang_cool has quit IRC05:18
*** LaoLang_cool has joined #maemo-ssu07:33
*** M4rtinK has joined #maemo-ssu08:46
*** M4rtinK has quit IRC08:53
*** bindi__ has joined #maemo-ssu09:16
*** peetah_ has joined #maemo-ssu09:16
*** bindi has quit IRC09:21
*** Raimu has quit IRC09:21
*** peetah has quit IRC09:21
*** LaoLang_cool has quit IRC09:39
*** freemangordon_ has joined #maemo-ssu09:51
*** freemangordon_ has left #maemo-ssu10:01
*** dafox has joined #maemo-ssu10:05
*** freemangordon has quit IRC10:08
*** psycho_oreos has joined #maemo-ssu10:53
*** trumee_afk has quit IRC10:58
*** trumee_afk has joined #maemo-ssu10:59
*** trumee_afk is now known as trumee10:59
*** Pali has joined #maemo-ssu11:00
*** lizardo has joined #maemo-ssu12:31
*** DocScrutinizer has quit IRC12:34
*** DocScrutinizer has joined #maemo-ssu12:34
*** andre__ has joined #maemo-ssu12:41
*** ekze-nyan has quit IRC12:49
*** ekze-nyan has joined #maemo-ssu12:49
*** ekze_nyan has joined #maemo-ssu12:52
*** ekze-nyan has quit IRC12:55
*** ekze_nyan has quit IRC13:00
*** ekze has joined #maemo-ssu13:00
*** LaoLang_cool has joined #maemo-ssu13:15
*** LaoLang_cool has quit IRC13:16
*** lartza_ has joined #maemo-ssu13:23
*** dafox has quit IRC13:35
*** dafox has joined #maemo-ssu13:59
*** Raimu has joined #maemo-ssu14:23
*** Pali has quit IRC14:32
*** M4rtinK has joined #maemo-ssu14:33
*** Pali has joined #maemo-ssu14:33
*** Pali has quit IRC14:36
*** Pali has joined #maemo-ssu14:37
*** Pali has quit IRC15:00
*** Pali has joined #maemo-ssu15:00
*** LinuxCode has joined #maemo-ssu15:18
*** mase76 has joined #maemo-ssu16:06
*** M4rtinK has quit IRC16:11
*** M4rtinK has joined #maemo-ssu16:15
*** M4rtinK2 has joined #maemo-ssu16:17
*** M4rtinK2 has quit IRC16:18
*** andre__ has quit IRC16:20
*** mase76 has quit IRC16:29
*** Pali has quit IRC16:32
*** mase76 has joined #maemo-ssu16:33
*** peetah_ is now known as peetah16:48
*** M4rtinK has quit IRC16:48
*** Pali has joined #maemo-ssu16:55
*** dafox has quit IRC17:19
*** andre__ has joined #maemo-ssu17:26
*** andre__ has joined #maemo-ssu17:26
*** M4rtinK has joined #maemo-ssu18:16
*** andre__ has quit IRC18:29
*** mase_76 has joined #maemo-ssu18:36
*** mase76 has quit IRC18:39
*** trbs has joined #maemo-ssu18:46
Paliping merlin199118:52
merlin1991pong18:52
Palimerlin1991, I have program which write kernel version which is flashed into mtd318:53
Palihttp://atrey.karlin.mff.cuni.cz/~pali/kernel-version.c18:53
Paliit will scan for piggy.gz in zImage format, try to unpack piggy.gz and try to find version string18:54
Paliworking with mtd or block device or with regular file18:54
Palimerlin1991, can you test it?18:54
merlin1991sure18:55
PaliI think this is way how to determinate if user has flashed U-Boot, nokia stock kernel or kernel power18:55
merlin1991though can you make me an arm build of it?18:55
Palihttp://atrey.karlin.mff.cuni.cz/~pali/kernel-version.armel18:55
Palirun as root: ./kernel-version.armel /dev/mtd3ro18:56
merlin1991Version string not found18:56
Palican you create dump of your mtd3ro?18:57
*** NIN101 has joined #maemo-ssu18:58
merlin1991is it safe to just dd if=/dev/mtd3 of=somefile?18:58
Paliyes18:58
Palibetter mtd3ro18:59
Palimerlin1991, which kernel version are you using?18:59
merlin1991power46 from the bleeding edge package19:00
merlin1991btw are you going to include the "new" driver from lpx1 into powerkernel?19:00
Palimaybe in future in separate package19:00
Palinow problem is that kernel-power v50 is builded but not uploaded from autobuilder to extras-devel :-(19:01
merlin1991hm it fails aswell with the stock pr 1.3 kernel19:01
merlin1991ah wait19:02
merlin1991I don't have the stock kernel here but power46-wl1 again19:02
merlin1991trying on yet another n900 :D19:02
Palican you send me dump?19:02
merlin1991yeah I'm only trying to figure out where I dropped my ssh key :D19:03
Paliit should show version of each zImage kernels19:03
Paliok19:03
merlin1991http://188.40.39.11/~christian/maemo/mtd3dump_power46-wl1.bin19:04
PaliVersion: 2.6.28.10power46-wl119:05
Pali(on x86_64)19:05
merlin1991I ran it on the device19:05
Paliok19:05
merlin1991I'll try to run it against the dump19:05
merlin1991(on the device again)19:05
PaliVersion: 2.6.28.10power46-wl119:06
Paliworking too19:06
Palican you try to run: ./kernel-version.armel mtd3dump_power46-wl1.bin19:06
merlin1991that works19:06
merlin1991trying again agains /dev/mtd3ro19:06
Palimake sure you are root19:07
merlin1991I sshed in as root :D19:07
merlin1991Nokia-N900:~# ./kernel-version.armel /dev/mtd3ro19:07
merlin1991Version string not found19:07
merlin1991but19:07
merlin1991Nokia-N900:~# ./kernel-version.armel mtd3dump_power46-wl1.bin19:07
merlin1991Version: 2.6.28.10power46-wl119:07
PaliI updated armel binary on server19:08
Palinow it show size of file/mtd device19:08
Palicheck if size output is same19:08
merlin1991Size: 209715219:09
merlin1991but still no version info19:09
Palisize is ok19:09
merlin1991dd also reads same size19:10
*** M4rtinK has quit IRC19:10
merlin1991ls19:10
PaliI will try to investigate where is problem19:10
Paliprogram mapping file to memory19:10
Paliso maybe mmap is not working correctly19:10
merlin1991odd thing is it works against the dd dump19:10
merlin1991maybe you can't mmap the /dev device?19:11
PaliI will try to find out some more info19:16
Paliat least uncompressing and finding correct version string working19:17
*** freemangordon has joined #maemo-ssu19:26
Palimerlin1991, I found problem. mmap returning (void *)-1 if failed (not NULL)19:28
Paliand it failed with errno ENODEV19:28
Paliso maybe really mtd dev does not support mmap19:29
*** mase_76 has quit IRC19:43
freemangordonPali, why do you do mmap? Isn't it possible to just read()?19:44
Palimmap is easy way19:45
Palibut I rewrite it to use read19:45
freemangordonok :)19:45
Paliprogram needs to uncompress piggy.gz19:45
Paliin zImage19:45
Paliand for unpacking is possible to use only zllib.h19:45
freemangordonaah, I see, you pass the mmaped memory to library functions.19:46
Paliso I need to use other read buffer19:47
*** infobot has quit IRC19:47
Paliand create new while loop for reading19:47
freemangordonyeah. did you try to mmap /dev/mtd3 instead of /dev/mtd3ro?19:47
Paliyes19:48
Paliseems that kernel does not support mapping mtd devices19:48
freemangordonwell, that is reasonable, mmaping flash devices is a very very bad idea, because of the writing19:49
*** xmlich02 has quit IRC19:55
*** xmlich02 has joined #maemo-ssu19:57
*** Pali has quit IRC20:09
*** NIN102 has joined #maemo-ssu20:20
*** NIN101 has quit IRC20:22
*** psycho_oreos has quit IRC20:26
*** Free-MG has joined #maemo-ssu20:42
*** wmarone__ has quit IRC20:47
*** arcean has joined #maemo-ssu20:48
*** wmarone__ has joined #maemo-ssu20:48
*** Free-MG has quit IRC21:55
*** bindi__ is now known as bindi22:45
*** lizardo has quit IRC23:00
*** LinuxCode has quit IRC23:50

Generated by irclog2html.py 2.15.1 by Marius Gedminas - find it at mg.pov.lt!