Previous day | Jump to hour: 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 | Next day

Seconds: Show Hide | Joins: Show Hide | View raw
Font: Serif Sans-Serif Monospace | Size: Small Medium Large

Click in the nick column to highlight everything a person has said.
The Logo icon identifies that the person is a core developer (has commit access).

Notice: Only Gecko based browsers prior to FF4 support the multipart/mixed "server push" method used by this log reader to auto-update. Since you do not appear to use such a browser, this page will simply show the current log, and not automatically update.

#rockbox log for 2011-01-13

00:00:39thomasjfoxAccording to one dev it contained a good map of register, timers etc.
00:05:00 Quit panni_ (Read error: Connection reset by peer)
00:06:32 Quit dfkt (Read error: Connection reset by peer)
00:09:36 Join panni_ [0] (
00:11:20 Join JdGordon| [0] (
00:11:20 Quit JdGordon| (Changing host)
00:11:20 Join JdGordon| [0] (~jonno@rockbox/developer/JdGordon)
00:13:31*B4gder kicked his client to become m68k-gcc452
00:15:31 Quit ender` (Quit: There are two ways of constructing a software design. One way is to make it so simple that there are obviously no deficiencies. And the other way is to make it so complicated that there are no obvious deficiencies.-- C.A.R. Hoare)
00:15:56n1sB4gder: great
00:20:22jhMikeSthomasjfox: I see
00:28:59***Saving seen data "./dancer.seen"
00:31:32 Join kugel [0] (~kugel@rockbox/developer/kugel)
00:33:52 Quit thomasjfox (Remote host closed the connection)
00:36:28 Join thomasjfox [0] (
00:37:13 Join mudd1_ [0] (
00:44:08 Quit Torne (Remote host closed the connection)
00:45:05 Quit utanapischti (Quit: WeeChat 0.3.2)
00:45:29 Join sasquatch [0] (
00:51:47 Quit pamaury (Remote host closed the connection)
00:54:28 Quit domonoky (Read error: Connection reset by peer)
01:02:09 Quit n1s (Quit: Lämnar)
01:03:01 Part Zagor
01:06:30 Quit thomasjfox (Remote host closed the connection)
01:13:12 Quit sideral (Quit: Leaving.)
01:19:38 Quit kugel (Remote host closed the connection)
01:20:30 Quit mystica555 (Ping timeout: 240 seconds)
01:32:43 Quit timccc (Ping timeout: 276 seconds)
01:33:15 Join mystica555 [0] (
01:34:00 Join madalu [0] (~user@unaffiliated/madalu)
01:35:36 Quit GeekShadow (Quit: The cake is a lie !)
01:39:13 Quit Kitar|st (Ping timeout: 276 seconds)
01:42:12 Join Torne [0] (~torne@rockbox/developer/Torne)
01:44:31 Join Kitar|st [0] (
01:49:42 Quit mudd1_ (Quit: Ex-Chat)
01:55:27 Quit factor (Ping timeout: 246 seconds)
01:55:28 Quit esperegu (Read error: Connection reset by peer)
02:10:51 Join factor [0] (~factor@
02:21:31 Quit leavittx (Ping timeout: 255 seconds)
02:21:34 Quit DerPapst1 (Quit: Leaving.)
02:21:35 Quit DerPapst (Quit: Leaving.)
02:28:20 Quit pjm0616 (Ping timeout: 264 seconds)
02:29:02***Saving seen data "./dancer.seen"
02:43:11 Quit Judas_PhD (Quit: This is a quitting message)
02:43:17 Join pjm0616 [0] (~user@
02:59:44 Join LambdaCalculus37 [0] (
02:59:44 Quit LambdaCalculus37 (Changing host)
02:59:44 Join LambdaCalculus37 [0] (~rmenes@rockbox/staff/LambdaCalculus37)
03:01:46 Quit LambdaCalculus37 (Client Quit)
03:18:22 Quit Kitar|st ()
03:36:04 Join Saij [0] (
03:36:40 Quit krabador (Remote host closed the connection)
03:42:18 Join Keripo [0] (
03:52:25 Quit Keripo (Quit: Leaving.)
03:53:15 Join Judas_PhD [0] (
03:56:57 Join Keripo [0] (
04:05:26 Quit Judas_PhD (Ping timeout: 240 seconds)
04:08:52 Join Barahir_ [0] (
04:12:12 Quit Barahir (Ping timeout: 260 seconds)
04:15:11S_a_i_n_tJdGordon|: touch areas *are* viewport relative...correct?
04:15:32JdGordon|not sure
04:16:13S_a_i_n_tHmmm...I assumed so myself, and you can definitely fire a touch area in a conditional viewport right?
04:16:34S_a_i_n_tthe theme editor seems to not do either of those things.
04:17:05S_a_i_n_t(touch areas are always true...and screen relative, not VP relative)
04:17:19S_a_i_n_tin the theme editor^
04:19:52 Join Judas_PhD [0] (
04:29:05***Saving seen data "./dancer.seen"
04:36:06 Quit amiconn (Disconnected by services)
04:36:07 Join amiconn_ [0] (quassel@rockbox/developer/amiconn)
04:36:22 Quit pixelma (Disconnected by services)
04:36:23 Join pixelma_ [0] (quassel@rockbox/staff/pixelma)
04:36:25 Nick amiconn_ is now known as amiconn (quassel@rockbox/developer/amiconn)
04:36:39 Nick pixelma_ is now known as pixelma (quassel@rockbox/staff/pixelma)
04:50:24 Quit TheSeven (Ping timeout: 276 seconds)
04:53:27 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
05:07:40S_a_i_n_tIs there some way I'm missing that you would be able to make a touch area for "Mute"?
05:08:17S_a_i_n_tI'd like the volume to mute when I hold the volume icon for <X>, is that possible and I'm just missing it?
05:11:25 Quit Rob2222 (Ping timeout: 250 seconds)
05:16:43JdGordon|should be dead simple to implement
05:16:52JdGordon|actually.... thats something we really should add
05:16:57JdGordon|or, just press pause
05:18:07 Quit parafin (Read error: Operation timed out)
05:19:32 Quit simonrvn (Quit: ZNC -
05:22:28 Join simonrvn [0] (
05:26:35JdGordon|S_a_i_n_t: I can see logic in adding mute for the fm screen, it should be really simple to do if you want to try
05:26:58JdGordon|just need to add the touch region code and 3 lines to handle it
05:29:16 Quit fdinel (Quit: Miranda IM! Smaller, Faster, Easier.
05:35:28JdGordon|S_a_i_n_t: untested and WPS only (needs adding for use in sbs/fms)
05:42:06 Quit Horscht (Quit: Verlassend)
05:53:11 Join timccc [0] (~timccc@
05:54:39 Quit Judas_PhD (Quit: This is a quitting message)
05:58:22 Quit madalu (Ping timeout: 260 seconds)
06:03:40 Quit fyre^OS (Quit: lamers envy me like they envy bill g -- main boot xp, just the way it should be!)
06:04:32 Quit Keripo (Quit: Leaving.)
06:23:36 Quit Saij (Read error: Connection reset by peer)
06:29:08***Saving seen data "./dancer.seen"
06:56:41 Join phizz [0] (~phizz@unaffiliated/phizz)
07:04:10 Quit JdGordon| (Quit: leaving)
07:23:44 Join esperegu [0] (~quassel@
07:29:44 Join Judas_PhD [0] (
07:36:26 Join JdGordon| [0] (
07:36:26 Quit JdGordon| (Changing host)
07:36:26 Join JdGordon| [0] (~jonno@rockbox/developer/JdGordon)
07:37:03JdGordon|S_a_i_n_t: actually... yeah maybe it doesnt really make sense for a mute button
07:37:15JdGordon|because you need to turn the screen on first and pulling the headphones is faster
07:40:35 Join PurlingNayuki [0] (~PurlingNa@
07:40:54PurlingNayukiHi everyone.
07:41:20PurlingNayukiWhat files should I edit if I want to add a item in WPS context menu?
07:41:43JdGordon|apps/onplay.c i tihnk
07:44:27PurlingNayukiSo you see, some one have added 'picture flow' in it.
07:44:48 Quit panni_ (Quit: ( :: NoNameScript 3.81 :: ))
07:45:17PurlingNayukiHow should I add a item named lryplayer in it, and when I press it I can start lryplayer?
07:45:40JdGordon|find the commi that added the pictureflow code and copy it
07:46:47PurlingNayukiI can't find which commit it's in...
07:47:03PurlingNayukiI only get a patch, but it seems to be out-of-date.
07:48:57CIA-7New commit by 03jdgordon (r29043): FS #11867 - Add 2 new tags to allow skins to display themed peakmeters ...
07:52:22CIA-7r29043 build result: 4 errors, 0 warnings (jdgordon committed)
07:52:48PurlingNayukiOh, errors.
07:53:06 Join Buschel [0] (
07:53:24 Quit Judas_PhD (Ping timeout: 265 seconds)
07:54:37CIA-7New commit by 03jdgordon (r29044): remove unecessary (duplicated even) header
07:56:21 Join ruckus [0] (
07:56:39ruckushey Buschel
07:57:11BuschelI saw you were searching for me
07:57:17ruckusso i am back with some (perhaps) frightening results
07:57:31Buschellet's see :)
07:58:11ruckusok. so, i experienced no freezes of the kind i had before, until tonight, which is pretty good since they used to happen lot more often
07:58:27CIA-7r29044 build result: 4 errors, 0 warnings (jdgordon committed)
07:58:33 Join einhirn [0] (
07:58:58ruckusa different problem appeared though. my iPod wouldn't turn on when off. twice i had to force restart so it would turn on
07:59:25ruckusthe first happened yesterday, and the other time tonight
07:59:50ruckusalthough, once i restarted it tonight i got an icon
08:00:03Buschelhmm, the build I gave you is pure svn with one change reverted. this reverted change is not connected to startup...
08:00:25 Join GeekShadow [0] (
08:00:26 Quit GeekShadow (Changing host)
08:00:26 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
08:00:26ruckusan exclamation next to an ipod, with a link to apple's support page about ipods
08:01:00CIA-7New commit by 03jdgordon (r29045): here too
08:01:02ruckusi haven't done anything to it, aside of running the diagnostics and trying to restart it several times
08:01:09Buschel? battery empty ?
08:01:21ruckusnope. it's pretty full.
08:01:54Buschelnever saw such behaviour on my iPod
08:02:02ruckusi've never seen this before as well
08:02:26PurlingNayukiSo did you try to re-flash your iPod?
08:02:43ruckusi haven't tried anything. i thought it's best to ask you guys first
08:03:38PurlingNayukiFlash your iPod may solve this. Remember to backup your data.
08:03:43ruckusthis is strange
08:03:53ruckusi am confident that i charged my ipod today fully
08:04:31ruckusi just plugged it in the usb drive, and i again got that exclamation business
08:04:45Buschelcan you enter the emergeny disk mode?
08:04:55CIA-7r29045 build result: All green
08:04:57ruckusi moved it at a different angle, and i can see the screen saying that the battery is extremely low
08:05:18ruckusnope, back to the error screen
08:05:40PurlingNayukiI don't think your battery is truly low.
08:05:49PurlingNayukiIt must be software error.
08:06:08 Join Judas_PhD [0] (
08:06:08ruckusgood thing i backed up everything today
08:07:10PurlingNayukiSo you should get into the OF first, I guess.
08:07:39ruckusOF ?
08:07:54Buschelwould you mind using another build? this one does not have the suspicious change reverted. but it has a special PANIC debug output, if a thread deadlock is detected
08:08:14ruckusbut that's what i'm saying
08:08:27ruckusi mean, the ipod won't mount at all now
08:08:48PurlingNayukiLet's start afresh.
08:09:00PurlingNayukiWhich build of Rockbox you're using?
08:09:19ruckusBuschel should know that
08:09:39Buschelit is a buold based on r28992 w/o r2887
08:09:44Buschelw/o r28877
08:10:47BuschelI am using it as well and it does not show such behaviour by default
08:11:15ruckusi am fearing that the restarts damaged the ipod's os somehow
08:12:21ruckusbecause both rockbox and the ipod's os refuse to boot, whether i plug the usb cable, or the adapter
08:13:25Buschelagain: can you enter the OF's emergeny disk mode?
08:13:53ruckusBuschel: i am just looking up the instructions how to do that
08:13:54 Join Bagder [0] (
08:13:58ruckusi haven't tried that yet
08:14:57Buschelruckus: start the ipod via pressing OK. when you see the apple logo on the screen press both MENU and PLAY/PAUSE until you enter the emergency disk mode
08:17:28ruckusno luck so far
08:17:46Buschelno logo visible?
08:20:31ruckusi see the error
08:20:58kisakwhy is the "status summary of unstable and unusable ports." link on the main page of incomplete?
08:21:23kisakthe # tag is incomplete
08:21:32Buschelwhen you reset the ipod (MENU + OK several senconds), do you then see the apple logo for some seconds? if so, press OK+PLAY/PAUSE for several seconds.
08:21:50ruckusi see the logo
08:22:04kisakruckus: it should take 10-15 seconds
08:22:46ruckusthis time i got it
08:23:06Buschelgood, if it is emergency disk mode, plug USB to your PC
08:23:30Buschelok, does it show that it is charging?
08:23:39ruckusyes it does
08:23:51Buschelcan you access the disk via USB?
08:24:14ruckusholy crap
08:24:16Buschelwhat does the content look like?
08:24:26ruckusi see whole bunch of garbled junk in one of my folders
08:24:57ruckusthat is, on more than one of my folders
08:25:12Buschelyou should run chkdsk (or something similar)
08:25:26Buscheland reinstall
08:26:47PurlingNayukiFilesystem damaged.
08:26:56ruckuslooks like it
08:27:40ruckusUnrecoverable error in folder \sgf
08:27:40ruckusConvert folder to file (Y/N)?
08:27:48Buschelhmm, weren't there some such reports in the past weeks? I thought they were connected to the clip/fuze though
08:28:14PurlingNayukiIf you chose Y, you will get a lot of junk.
08:28:22ruckusi see
08:28:48ruckusoh crap
08:28:55ruckusit's asking me for every folder
08:29:11***Saving seen data "./dancer.seen"
08:29:39 Join T44 [0] (
08:29:39ruckusok, not every
08:29:42ruckusbut many
08:30:19 Quit elcan (Ping timeout: 272 seconds)
08:30:43 Join elcan [0] (
08:31:23 Join sideral [0] (~sideral@unaffiliated/sideral)
08:31:34PurlingNayukiSo is it have a quiet mode?
08:32:33ruckusno idea, i'm using chkdsk
08:32:48 Quit Topy44 (Ping timeout: 240 seconds)
08:32:49PurlingNayukiI don't know much about it.
08:32:56ruckusfuuuuuuuu, bad clusters
08:33:11PurlingNayukiMaybe chkdsk [Device] /f /q ?
08:33:12Buschelruckus: I need to leave for work now. so, when you reinstall rockbox, can you please use the following build: it is a plain r28992 build + an additional patch to show panic screen when detecting deadlock situation. hopefully this is the reason for the lockups you experienced each day.
08:33:41PurlingNayukiWait, which patch do you include?
08:33:45Buschelruckus: do you have an iPod with 30 or 60/80 GB HDD?
08:34:03ruckusit's 30gig version, but i put a 160gb drive
08:34:08PurlingNayukiBuschel: What patch(es) do you include?
08:34:46BuschelPurlingNayuki: the patch jhMikeS attached to FS #11863.
08:35:12PurlingNayukiBuschel: Thanks. I will see it.
08:35:34BuschelPurlingNayuki: the build which was used by ruckus so far was plain r28992 (with only r28877 reverted).
08:36:12 Quit Llorean (Read error: Connection reset by peer)
08:36:17PurlingNayukiThank you.
08:37:02ruckusBuschel: i'll see what i can do. i have to see what's going on with this drive here
08:37:24 Quit ruckus (Read error: Connection reset by peer)
08:37:31 Join Llorean [0] (
08:37:44Buschelruckus: yes, of course. good luck!
08:37:50Buschelsee you later
08:37:53 Join ruckus [0] (
08:37:54 Quit Buschel (Quit: ChatZilla 0.9.86 [Firefox 3.6.13/20101203075014])
08:38:24 Nick ruckus is now known as ruckus_ (
08:38:58ruckus_did anyone say anything after my last comment? i got disconnected for a second
08:39:09 Quit Xerion (Read error: Connection reset by peer)
08:39:55 Join Xerion [0] (
08:40:40amiconnTorne: The H300 exists with two different LCD panels which have different subpixel orientation, without a way for the firmware to know which panel type is connected. The LCD controller is the same
08:41:37sideralruckus: Yes:
08:42:43ruckus_i didn't notice the link in the topic... argh. time to sleep, i've been awake for nearly 24 hrs
08:46:15ruckus_goodnight folks
08:46:17 Join ender` [0] (
08:46:36 Join wodz [0] (
08:47:05JdGordon|wodz: i figured out the problem :)
08:47:28wodzTheSeven: my problems with ipods are coming out from ehci_hcd driver in linux definitely.
08:47:34wodzJdGordon|: nice
08:47:54wodzJdGordon|: what was the problem?
08:48:20JdGordon|fiddly nonesense with how peakmeters are enabled in the wps
08:49:18wodzdo you need to enable it in some special way or you fixed this and you simply use %pL/pR as advertised?
08:49:19JdGordon|my wiki account is stuffed so untill zagor can fix it i cant edit CustomWPS to add the new tags thoguh
08:49:26JdGordon|the latter
08:49:44wodzis it in svn or you posted to FS?
08:49:44ruckus_hey PurlingNayuki. would it be easier just to format the disk? i mean, i have everything backed up anyway
08:50:05PurlingNayukiYes, of course.
08:50:15JdGordon|wodz: svn as of an hour ago
08:50:31wodzok, I'll test it on the target later today
08:52:41wodzJdGordon|: BTW. I spend over an hour digging in and my conclusion is that it is hardly extensible hack rather than framework solution
08:53:56JdGordon|wodz: I dont know who wrote it origionally, and yes I agree.... we sort of discussed ages ago about removing all skins from svn except cabbie which would mean a much simpler
08:54:13PurlingNayukiRockbox now use system built-in font in EQ screen. How can I change it in to user font?
08:55:03wodzJdGordon|: it is not the matter of different skins in svn - it is a matter of horrible inconsistent parsing perform.
08:55:15 Join LinusN [0] (~linus@rockbox/developer/LinusN)
08:57:15ruckus_PurlingNayuki: should i format the drive with itunes, or the command line?
08:57:36PurlingNayukiJust format it into FAT32, I guess.
08:57:51ruckus_because i got win7
08:57:55PurlingNayukiRockbox only support FAT32.
08:58:13ruckus_i am not too confident formating anything with the command line, and the gui won't let me pick fat32
08:58:45ruckus_the formatting gui that comes with win7
08:58:57ruckus_so does anyone know what command should i use ?
08:58:59PurlingNayukiSo you can format it in explorer.
08:59:23ruckus_i can, but not in fat32. it's only ntfs and exfat
08:59:49PurlingNayukiTry to use some tools like Partition Manager.
09:00:08ruckus_do you mean to say that itunes won't do the job ?
09:00:38PurlingNayukiBut I never use itunes.
09:00:51Bagderuse whatever you want, but select fat32 =)
09:00:54 Quit bluebrother (Disconnected by services)
09:00:56 Join bluebroth3r [0] (~dom@rockbox/developer/bluebrother)
09:01:06PurlingNayukiThat's true.
09:01:47 Join Zagor [0] (~bjst@rockbox/developer/Zagor)
09:02:40PurlingNayukiIs there anyone know how to change font in EQ screen?
09:04:51 Join petur [0] (d408b802@rockbox/developer/petur)
09:08:04S_a_i_n_tEQ Screen can only use Userfont
09:08:12S_a_i_n_tor sysfont.
09:08:52phizzanyone know if the new house came on tonight?
09:09:38PurlingNayukiS_a_i_n_t: Then how to make it use userfont?
09:10:08S_a_i_n_tDon't load a sysfont.
09:10:23S_a_i_n_tI mean, "don;t laod a userfont".
09:10:41S_a_i_n_t*"don't load a userfont"
09:11:10PurlingNayukiMust I change the source code?
09:11:34S_a_i_n_tThere's no tage to differentiate the EQ screen from any other, so a custom font can't be loaded conditionally for this screen.
09:12:23S_a_i_n_tIf a user font is loaded, the EQ screen will use this...if not, sysfont will be used.
09:14:11PurlingNayukiNo, I don't mean this.
09:15:42PurlingNayukiI mean, though I select my own font, when I try to use graphical EQ settings the font is still the built-in one.
09:15:51PurlingNayukiHow to change it?
09:20:10S_a_i_n_tIt should be using the userfont if one is selcted...if it isn't, I have no idea why.
09:22:03TheSevenwodz: how can you tell that?
09:22:22PurlingNayukiI saw this in source code: LANG_SYSFONT_EQUALIZER_MODE
09:22:24PurlingNayukiCan I change it into LANG_UIFONT_EQUALIZER_MODE
09:22:27TheSevenif it works on uhci that can also have ipod-side timing or error handling problems as a cause
09:24:28wodzTheSeven: Unbinding ehci_hcd which effectively means falling back to uhci makes ipod work. And the problem showed up after upgrading my laptop from previous LTS ubutnu to current one
09:24:49TheSevenok, so you say older ehci versions work as well?
09:25:00TheSevennow why doesn't this happen for all those thumbdrives then?
09:25:58 Join efyx [0] (
09:27:25wodzTheSeven: it happens for some although the issue is not that widespread
09:27:46ruckus_okay guys
09:27:55ruckus_i found a way to format my drive in fat32
09:28:16ruckus_does anyone know what should be my allocation unit size ?
09:28:24TheSevenwodz: is there a kernel bug item for that?
09:28:36ruckus_my drive is 120gb btw
09:29:14JdGordon|Zagor: hey, when you get a few min can you look at my wiki account please? I cant login with my password and cant reset because no email or soemthing
09:29:33wodzTheSeven: I couldn't find one. There are a few entries in bugtrackers of major distributions but this are left unanswered mostly
09:29:51PurlingNayukiruckus_: Just use the default size.
09:30:04ruckus_it's not giving me that option
09:31:31sideralSaint, gevaerts, Torne: With the unmount-sync patch, my experience is much more stable now. But last night I had a strange case of FS corruption again, so the solution may not be complete :(
09:31:50TheSevenruckus_: you can basically use anything, and what the optimum value is depends on what you plan to do with the file system
09:31:56S_a_i_n_tGah...same here.
09:32:00S_a_i_n_tsideral: ^
09:32:19S_a_i_n_tAll seemed to be going well, then the FS all of a sudden TRASHED itself on an un-mount.
09:32:20TheSevenfile sizes will basically be rounded up to the next allocation size boundary, and thus low allocation sizes waste some space on the disk
09:32:48TheSevenhowever, each allocation unit needs 4 bytes in the FAT (and another 4 in the backup FAT if present)
09:32:52S_a_i_n_tthe solution we're running definitely helps, but, doesn't seem to be quite complete.
09:33:02 Join kevku [0] (~kevku@2001:7d0:0:f9af:feed:feed:feed:feed)
09:33:21TheSevenso too small allocation sizes waste space as well, and increase the time needed to mount the volume, and slightly decrease the sequential access speed
09:33:30S_a_i_n_tthe Color got a goot 8 hours playback today, with multiple mounts/ *was* fine, until it suddenly wasn't.
09:33:37ruckus_TheSeven: ok, and what about too large ?
09:34:15TheSevenso if you have a rather low number of files, and most of them are rather large, use a large allocation size, while if you have several thousands of tiny files, use small allocation sizes
09:34:32TheSevenalso there is a limit of IIRC 256M allocation units
09:34:52ruckus_ok, what do you mean by tiny and large
09:35:09TheSevenruckus_: "too" large means faster accesses/mounts, but wasted space for small files
09:35:10sideralSaint: The strangeness I experienced was that after umounting the FS, fsck saw lots of errors (FATs don't match, "cluster chain length is 0 bytes. Truncating file to 0 bytes." and "FOO and BAR share clusters."), but after a disconnect / reconnect the errors were mostly gone (only the free-block summary info was wrong)
09:35:31TheSevenif your allocation size is 32KB (the maximum supported by fat32), each and every a-few-bytes file will need 32KB on the disk
09:36:02siderallooks like Rockbox managed to flush some write buffers only after the reconnect
09:36:10S_a_i_n_tsideral: My FS was just.....trashed, for lack of a batter word.
09:36:12TheSeventhat's the major drawback of large allocation units
09:36:34ruckus_well most of my library is V0 mp3s
09:36:36S_a_i_n_ta few thousand corrupt files immediately after unmounting.
09:36:42TheSevenon the other hand my ipod classic 80GB with 4KB allocation units takes more than 10 seconds to mount on windows even though there is no backup fat present
09:37:14TheSeven(because it reads the whole 80MB-sized FAT to RAM)
09:38:10ruckus_TheSeven: ok. so what would you use for for a large V0 mp3 library ?
09:38:41TheSevenprobably the maximum possible (32K)
09:38:45sideralSaint: To verify my theory, next time you see a trashed FS, can you try to disconnect and reconnect the device without fixing the FS and without rebooting Rockbox?
09:38:58ruckus_TheSeven: 32768 ?
09:39:14ruckus_there's one more after that over here
09:39:57TheSeveni think the fat spec states that (while it's technically possible) cluster sizes bigger than 32K are not allowed
09:40:08TheSevenso while windows can probably handle it nevertheless, there might be tools that can't
09:40:17ruckus_ok. fair enough
09:41:07ruckus_TheSeven: would a quickformat do the job? or should i do a full format ?
09:41:48S_a_i_n_tsideral: Yes, I cam try that.
09:42:13TheSevenruckus_: if you expect your drive to have bad sectors, do a full format
09:42:19S_a_i_n_tsideral: Do you have any further idea why this may be happening? (I'm a themer, more than a coder ;))
09:42:24TheSevenotherwise a quick format should be sufficient
09:42:33TheSeven(a full format would probably take several hours)
09:43:00ruckus_that's right. i'm not sure what to expect, so i'll leave it as it is
09:43:02sideralSaint: Not yet :(
09:43:07ruckus_full format
09:43:32ruckus_i wish i was more awake so i can actually comprehend what you were explaining
09:43:34ruckus_anyway, thank you
09:44:09S_a_i_n_tsideral: Is it possible that file handles are leaking?
09:44:52sideralSaint: I don't think so. The file handles are reset hard in disk_unmount_all
09:45:27S_a_i_n_t(we'll, supposed to be ;)) Just a quess...I'm pretty in the dark here in this area of the code.
09:46:02*wodz enjoys music from his rockboxed nano2g after working around stupid bug with linux ehci_hcd
09:46:37*S_a_i_n_t was not aware wodz *had* a Nano2G
09:46:44wodznow I have to craft smart udev rule
09:46:51S_a_i_n_tI would have hassled you more to test iLike Nano if I did ;)
09:47:14wodzS_a_i_n_t: You have memory leak - you offered me lcd replacement a few days ago
09:47:31S_a_i_n_tHahah!...yes, seems about right.
09:49:27wodzit's nice player the only thing which I dislike is how hard is to use hold switch
09:49:50S_a_i_n_tthis is correct.
09:50:03 Part ruckus_
09:50:04S_a_i_n_tespecially if the player has been opened, it never is quite the same afterward.
09:50:08S_a_i_n_tit's cheap plastic.
09:50:38wodzit behaves about the same as was before I opened it
09:51:50*wodz needs some nice box for his rockboxed players - the box of rockbox :P
09:52:46*JdGordon| cant figure out the theme editor code :/
09:53:26wodzJdGordon|: there is some description on the wiki how themeeditor is designed
09:54:19 Quit wodz (Quit: Leaving)
09:54:37PurlingNayukiHow can you do this: * wodz needs some nice box for his rockboxed players - the box of rockbox
09:55:01BagderPurlingNayuki: start the line you write with /me
09:55:10Bagder /me says so
09:55:19S_a_i_n_tFAIL! :)
09:55:42Bagderno, that line started with /me ! =)
09:55:46S_a_i_n_t...oh, right...demonstration...aha.
09:55:49*PurlingNayuki understand what Bagder said and try to use it.
09:55:51Bagder(notice the extra space :-)
09:57:50*PurlingNayuki is confused with adding a item in WPS context menu.
10:00:35S_a_i_n_tif you grep the code for MAKE_MENU you'll easily be able to see how the menu structure is defined.
10:01:17S_a_i_n_tI can't think of what you would possibly want to add to the context menu that isn't there already though.
10:01:35PurlingNayukiSo? I'll try. Thanks.
10:09:25PurlingNayukiS_a_i_n_t: I want to add a menu, when I press it, it starts snclrc.
10:09:59S_a_i_n_tyour best bet is looking at how the WPS Hotkey is defined, and adding another option to it.
10:10:25S_a_i_n_tthe Hotkey code has been made so it ias reasonably painless to add additional fields.
10:12:41PurlingNayukiNow we can run PictureFlow directly from WPS context menu, can't we?
10:12:51S_a_i_n_tjust replicate the code for WPS Pictureflow Hotkey, but change the relevant info to point to lrcviewer instead.
10:13:09PurlingNayukiI'd like to do something like that.
10:13:20S_a_i_n_tjust replicate the code for WPS Pictureflow Hotkey, but change the relevant info to point to lrcviewer instead.
10:15:50PurlingNayukiSo, when I try to do this, I'll get 'GO_TO_LYRICS undeclared'.
10:16:48JdGordon|bluebroth3r: ping?
10:17:12PurlingNayukiBut however I do, I can't find where the GO_TO_PICTUREFLOW is defined.
10:17:37JdGordon|root_menu.h I tihnk
10:19:16JdGordon|anyone awake who knows qt?
10:19:31 Quit S_a_i_n_t (Quit: I'm only going to Heaven if it feels like Hell, I'm only going to Heaven if it tastes like caramel...)
10:23:37JdGordon|why the hell does the theme editor create the timer and device options screens before a project is opened?
10:23:52 Quit phizz (Quit: phizz)
10:24:47 Join [Saint] [0] (S_a_i_n_t@
10:28:20 Quit [Saint] (Disconnected by services)
10:28:22 Join S_a_i_n_t [0] (S_a_i_n_t@
10:28:28 Nick S_a_i_n_t is now known as [Saint] (S_a_i_n_t@
10:29:14***Saving seen data "./dancer.seen"
10:32:31 Part LinusN
10:36:14 Quit chattr (Ping timeout: 240 seconds)
10:37:48 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
10:40:01 Join chattr [0] (
10:42:22 Join DerPapst [0] (
10:42:31 Join DerPapst1 [0] (
10:46:50 Join Rob2222 [0] (
10:54:48sideralI made a typo in the Summary line of FS #11873. Can anyone with edit powers fix it, please? "hutdown" -> "shutdown"
10:54:49 Join n1s [0] (~n1s@rockbox/developer/n1s)
10:55:57PurlingNayukiCan't you edit it?
10:56:21[Saint]Obviously not.
10:57:20sideralNo, I'm a simple user without edit dignity
10:58:19[Saint]sideral: Just something you noticed, or possibly relevant to our FS issues?
10:58:39sideralThanks Torne!
10:59:11sideralSaint: I just noticed while hunting that bug, but I don't think it's relevant to it
11:00:55 Quit GeekShadow (Ping timeout: 276 seconds)
11:00:56sideralI still don't understand what the shutdown code does to clean up the filesystem. Apparently nothing beyond sending a disk spindown event, politely asking all subsystems to commit their dirty data?
11:01:41[Saint]I'm not sure how politely it asks ;)
11:02:12TheSevenapparently too politely...
11:03:50Torneyou are aware that the shutdown is kinda done in two almost-entirely-unconnected places, yes?
11:03:53sideralTheSeven: Possibly politely enough: The FS seems to be clean after shutdown, but not when entering USB slave mode
11:04:06TorneThe app/ui level shutdown code is pretty disconnected from the actual firmware level poweroff
11:04:20Torneand the latter is queued on a timeout at the start of the shutdown process
11:04:33Torneso if the app code takes more than, er, 40 seconds i think? then the device gets hard powered off
11:04:59*TheSeven occasionally sees his ipods hit that timeout
11:05:09sideralTorne: I saw that, yes. But I think the stuff done after the timeout is not concerned with any state flushing, it just switches off the device hard, right?
11:05:44TorneNot entirely hard
11:05:55TorneIt's not going to be flushing the fat buffers ro anything, no
11:06:07Tornebut it does still turn the *hardware* off in a fairly controlled and clean way
11:06:13Tornerather than just cutting the power
11:06:24Torneit should still shut down the lcd and so on
11:06:52TheSevenhm, i did some fixes some months ago that makes sure that all file handles are closed when entering usb mode
11:07:18TheSevenso the problem is that there are still dirty fat sectors in the cache?
11:07:27TheSevensounds like a pretty trivial fix to me...
11:07:49 Join LinusN [0] (~linus@rockbox/developer/LinusN)
11:08:01TheSevenbut i think disk_unmount should be called in that case, and that should make sure the cache is flushed
11:08:51sideralTheSeven: Saint and I are testing a patch that does the FAT flushing. It helps, but doesn't seem to be complete
11:09:13JdGordon|[Saint]: did you see the mute patch earlier?
11:09:36[Saint]JdGordon|: No, I didn't...any chance of committing it? Or is it already?
11:09:45[Saint]I don't think it's entirely useless in SVN
11:09:53JdGordon|It needs a teency bit of discussion i think
11:10:17JdGordon|and testing...
11:10:22sideralTheSeven: It's this patch:
11:10:44[Saint]Well, I can't test it without an .apk which I can't build for at least 3 days or so.
11:10:48[Saint]JdGordon|: ^
11:10:57[Saint]But, I'll be happy to then.
11:11:07[Saint]As for discussion...I'm easy.
11:11:36sideralTheSeven: It calls fat_unmount(flush=true) for each volume before entering USB mode
11:12:02sideralTheSeven: The problem is we still see occasional FS inconsistency
11:13:17[Saint]JdGordon|: What is it about said patch that needs discussion?
11:13:27sideralBTW, there was a call to disk_unmount, but as disk_unmount doesn't flush buffers, this patch replaces it.
11:13:28JdGordon|if its wanted mostly
11:13:31gevaertssideral: you're using resume points I assume?
11:14:05[Saint]JdGordon|: Well...on a device like a smartphone...I'd like to be able to mute it from the wps/sbs without pulling the phones...personally.
11:14:12[Saint]I can't see why anyone wouldn't.
11:14:48JdGordon|wouldnt pause make more sense though?
11:14:56sideralgevaerts: Yes, I do. And with playback stats. So there's likely a fair amount of disk traffic before entering USB mode.
11:15:26[Saint]JdGordon|: I think both have their merits.
11:15:29TheSevensideral: er, shouldn't that code maybe go into disk_unmount(volume) instead?
11:16:36sideraldisk_unmount currently only is for removing a volume hard after is has been ejected, so it assumes it cannot write to the volume any longer. But yes, we can refactor the code once we got it to work
11:17:22gevaertsTheSeven: the problem is that callers of disk_unmount(volume) want fat_unmount(flush=false), because they respond to card eject, so flush won't work
11:17:39TheSevenso the old code was actually assuming that the media wasn't accessible any more?
11:18:47 Quit PurlingNayuki (Ping timeout: 240 seconds)
11:20:19sideralNow I'm wondering about two things: What is it that we don't flush before entering USB mode, and why is it flushed when powering off?
11:20:29 Join user890104 [0] (~Venci@2001:0:5ef5:79fd:28b2:318:2b16:10ec)
11:21:43TheSevenso on shutdown unmounting was previously handled like it is handled now for usb?
11:22:14sideralI can't seem to find any code related to unmounting in the shutdown code
11:22:42TheSevensideral: sounds like you want to do a breakpoint on usb connect, get a memory dump and look at the FAT buffers :)
11:22:44sideralLooks like it purely relies on all recipients of the disk-spindown event to flush their stuff
11:23:42sideralYeah. I was fantasizing about adding gcc-remote-debugging support via serial over USB...
11:25:07JdGordon|mute works.... unmute doesnt
11:25:54[Saint]Aaaaaawwww crap :/
11:27:23TheSevensideral: the nano2g can actually do that
11:27:50TheSevenjust call post_moretem_stub() somewhere, and it will freeze and allow you to download a RAM dump via USB
11:28:33sideral=-Owow. Not exactly gcc remote debugging, but sounds useful...
11:29:11TheSevenFarthen is working on a gdb stub for the embios kernel, but it isn't as easy as expected :/
11:31:07*TheSeven runs off...
11:31:13TheSeveni'll be back in an hour or so
11:33:36JdGordon|hehe im an idiot
11:35:23 Quit TheSeven (Ping timeout: 240 seconds)
11:37:32 Join TheSeven|Mobile [0] (~theseven@rockbox/developer/TheSeven)
11:39:51TheSeven|Mobileto download the dump you'll need i have no idea whether this was committed to svn or only attached to the flyspray task
11:43:43 Join TheLemonMan [0] (
11:44:45 Join GeekShadow [0] (
11:44:45 Quit GeekShadow (Changing host)
11:44:45 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
11:46:36 Join leavittx [0] (~lev@
11:48:06 Join wodz [0] (
11:49:14wodzgevaerts: I am cleaning my desk now and I found my BDM pod - are you interested in it?
11:49:34wodzor you decided to buy another?
11:49:57 Join dfkt [0] (dfkt@unaffiliated/dfkt)
11:51:05gevaertsLinusN: what sort of hardware skills are required to use BDM things on irivers?
11:52:00n1sgevaerts: maybe you can borrow any player with the bdm connector soldered that LinusN has (if he still has any)
11:52:01 Part LinusN
11:52:11n1sah, bad timing
11:52:32 Join LinusN [0] (~linus@rockbox/developer/LinusN)
11:52:48*gevaerts accuses LinusN of trying to hide :)
11:53:47LinusNgevaerts: mostly soldering of tiny SMD:s
11:53:47Bagderbut failing so!
11:53:52 Part LinusN
11:54:02gevaertswodz: I'm going to say I'm interested, and then I'll figure out how to get the soldering done
11:54:13wodzLinusN: but is there an unpopulated connector or you have to solder into test pads ?
11:54:17 Join LinusN [0] (~linus@rockbox/developer/LinusN)
11:54:21LinusNand the ability to swear and curse when you try to build the bdm gdb tools :-)
11:54:34gevaertsOh, but that's software! That's *easy*!
11:54:53BagderI believe the correct term is "just software" :-)
11:55:04gevaertsAh, yes :)
11:55:30wodzI build software for bdm quite recently and it was rather easy
11:55:45JdGordon|AlexP: pixelma: ping?
11:56:15LinusNi remember it as fairly painless, but it was a while
11:56:42CIA-7New commit by 03jdgordon (r29046): New Touchscreen region type... 'mute' which un/mutes volume without pausing playback
11:57:06LinusNi have planned a nice bdm revival session this weekend
11:57:22[Saint]JdGordon|: is it just mute/&mute ?
11:57:36LinusNso i guess i'll have a go at documenting it in the wiki
11:57:54wodzLinusN: I believe you worked before m68k-bdm server era
11:58:03JdGordon|[Saint]: yes
11:58:06LinusNwodz: that's correct
11:58:16JdGordon|you are now required to add it to all touch cabbies :D
11:58:18[Saint]and, thanks heaps's not often you pounce so quickly on my suggestions. Just want you to know I appreciate it when you do. ;)
11:58:25[Saint]JdGordon|: ^
11:58:55LinusNthe problem nowadays is finding a computer with a parallel port :-(
11:59:01 Quit The_Pwny (Quit: Pull the pin and count to what?)
11:59:09wodzLinusN: I have usb bdm pod
11:59:28LinusNwodz: i never managed to make my usb pod work :-(
11:59:29[Saint]JdGordon|: I said, I'm *hoping* to be able to do this one well enough to get it into's 240x320...I have to have the liberty of coding for a target I own first :P
11:59:44LinusNwodz: which pod is it?
11:59:52wodzLinusN: tblcf
11:59:59LinusNwodz: oh... :-)
12:00:19*sideral notices Rockbox already has a USB serial controller mode
12:00:21 Quit TheSeven|Mobile (Remote host closed the connection)
12:00:39LinusNwodz: perhaps i should have another go at it...
12:00:55CIA-7r29046 build result: All green
12:01:25wodzLinusN: did you managed to flash firmware onto MCU of the pod? This was the hardest part as linux tools are broken for that
12:01:43LinusNwodz: i believe i did
12:01:57wodzso the rest is simple :-)
12:03:05LinusNas far as i can remember, it broke somehow, possibly because i failed to connect the pod correctly to the target
12:03:49wodzI patched m68k-bdm to bypass check of MCU type and specify it from commandline as I wasn't able to craft rule for autodiscovery for 5249
12:04:56AlexPJdGordon|: Yo
12:05:09LinusNwodz: hmmm, it's coming back to me now
12:05:39LinusNi seem to remember that i had problem woth just that... maybe it wasn't broken after all
12:05:50LinusNcare to share the patch?
12:05:57JdGordon|AlexP: is OK? I just noticed these regions are missing
12:06:05wodzLinusN: yes, I'll have to dig a bit
12:06:45*wodz hates cleaning his desk
12:08:09JdGordon|AlexP: and the line "\config{\%T(x,y,width,\tabnlindentheight,action)}"... why is the \tabnlindent needed?
12:08:15AlexPJdGordon|: The patch looks fine - I'm not sure I understand the description, but that may just be as I'm not familar with touchscreen tags
12:08:20JdGordon|the manual has []'s around width and height which is wrong
12:09:08JdGordon|AlexP: %T(0,0,10,10,setting_inc,volume) will create a region which increments the volume when it is pressed
12:09:35AlexPJdGordon|: I think the tabnlindent bit is needed to make the line break but remain in the same table cell
12:10:03JdGordon|will it work like that or does it need a space?
12:10:15JdGordon|or shall i just FS this and leave it for osmeone else? :D
12:10:29AlexPI'm not sure, I'd have to test
12:10:57AlexPI'm happy to play with it :)
12:11:09AlexPI didn't get round to the other one last night, but tonight!
12:12:29 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
12:12:37AlexPJdGordon|: OK, I see what that means now - I didn't immediately twig that setting_inc referred to volume
12:12:58AlexPJdGordon|: Oh, and the example for setting8dec has a typo
12:13:06AlexPJdGordon|: It should be setting8dec
12:13:42AlexPJdGordon|: Maybe just add something like "e.g \config{%T(0,0, setting_inc, volume)} will increase the volume by one step)
12:14:17AlexP*setting_dec of course :)
12:14:17JdGordon|I'm putting it on FS :)
12:14:21AlexPOK :)
12:15:39JdGordon|FS #11874 :)
12:16:02AlexPI'll definatley try to attempt to maybe look at it tonight :)
12:16:40JdGordon|doesnt worry me... I never read the manual :D
12:19:26preglowbooltoader \o/
12:21:47 Quit leavittx (Ping timeout: 255 seconds)
12:29:17***Saving seen data "./dancer.seen"
12:29:34wodzpreglow: rather cryptic hurray :-)
12:42:10wodzLinusN: <- here is the patch for m68k-bdm
12:43:50LinusNwodz: thanks a bunch!
12:43:50 Join newClipUser [0] (
12:44:55wodzLinusN: I used such gdb script to debug bootloader on MPIO:
12:45:27newClipUserI have kind of bricked my sansa clip+ due to a power failure. now the FAT is corrupted. Any suggestion how to recover my data?
12:46:00wodzLinusN: and finally udev rule for tblcf:
12:47:18wodzLinusN: I used gdb 7.0.1 + m68k-bdm-1.4-pre4 patched
12:47:38LinusNwodz: thanks a lot
12:47:46wodzhope this helps
12:48:39wodzthe patch is 'alpha' quality as I failed to craft some smart autodiscovery for the part - that's why I introduced bypass method
12:49:07LinusNmaybe i can have a go at it
12:49:10newClipUseranyone here has experience in data recovering from corrupted FAT file system?
12:50:13LinusNnewClipUser: try pc inspector (
12:50:36*Torne has had good results with getdataback from runtime software but it's not free
12:50:49LinusNTorne: me too, i have a license
12:51:23newClipUser@LinusN: thanks, would it only help me to format or also recover my lost data?
12:52:00LinusNnewClipUser: pure recovery
12:52:08newClipUseroh, its german, are you german?
12:52:13LinusNnope, swedish
12:52:27LinusNyou can select language in the top right corner
12:52:41newClipUserno problem, because I do speak german :)
12:52:47LinusNoh :-)
12:53:52newClipUserand you are the starter of the rockbox project, right?
12:54:00newClipUserI really like it, thanks :)
12:54:18LinusNnewClipUser: yes, i am one of the project founders
12:54:56LinusNalthough i am not in the loop nowadays
12:56:57CIA-7New commit by 03rmenes (r29047): SA9200: Modify pacbox keymap a slight bit to avoid bunching ...
12:57:25 Join LambdaCalculus37 [0] (
12:57:25 Quit LambdaCalculus37 (Changing host)
12:57:25 Join LambdaCalculus37 [0] (~rmenes@rockbox/staff/LambdaCalculus37)
12:58:09 Quit LambdaCalculus37 (Client Quit)
12:59:32 Quit wodz (Quit: Leaving)
13:00:27CIA-7r29047 build result: All green
13:10:27[Saint]In the loop or no, you still deserve thanks where it's due LinusN ;)
13:10:43[Saint]Many, many people should be pretty bloody proud of themselves.
13:13:53 Join Saij [0] (
13:14:44 Part LinusN
13:15:28 Quit Saij (Read error: Connection reset by peer)
13:15:31 Part Zagor
13:20:36 Join LinusN [0] (~linus@rockbox/developer/LinusN)
13:21:57 Join Zagor [0] (~bjst@rockbox/developer/Zagor)
13:24:44Tornedo we have any docs anywhere on how to actually use bloat-o-meter/objdiff from utils?
13:27:22n1sTorne: i was just about to try that bloat o meter thingy, iirc i used it once before, i think you just pass it two elf files
13:38:45n1syes, just bloat-o-meter old.elf new.elf and it prints size per function and diff for the whole elf
13:39:36n1sa lot of diffs in inlining between toolchains so not terribly usefull for me
13:41:24n1si mean the 8 functions that grew the most according to this don't exist at all in the old elf
13:46:52n1sand similarly a bunch of functions present in the old elf aren't in the new one
13:53:23Zagoryou mean we actually changed things? ;-)
13:55:53 Quit JdGordon| (Quit: sleeeeeeeeeeeep)
13:57:33n1sfor the x5 the mayor size diff seems to cone from lcd_enable which probably means it just inlines less there since that function is just calling a few other functions
14:00:23n1sperhaps i should double check some of this inlining
14:02:30 Quit MethoS- (Read error: Connection reset by peer)
14:04:26n1soh, yes not much is inlined anymore
14:04:26 Quit antil33t (Read error: Connection reset by peer)
14:04:35 Join antil33t [0] (
14:04:45n1swarning: inlining failed in call to ‘lcd_write_reg’: optimizing for size and code size would grow
14:04:55 Join MethoS- [0] (~clemens@
14:05:37n1sshould we stop using Os (optmizing for size) or should we start sprinkling some always_inline attributes?
14:05:43 Join kugel [0] (~kugel@rockbox/developer/kugel)
14:06:01[Saint]lulz...I've introduced the GF to Abduction on the Galaxy 5
14:06:15[Saint]oops, wrong channel.
14:06:24[Saint]damn, that was embarrasing. :/
14:08:38 Join mystica555_ [0] (
14:15:26 Join WonTu [0] (
14:15:40 Part WonTu
14:20:13 Quit factor (Read error: Connection reset by peer)
14:29:18***Saving seen data "./dancer.seen"
14:32:34 Join robin0800 [0] (
14:36:11 Join factor [0] (~factor@
14:45:11 Quit GeekShadow (Read error: Connection reset by peer)
14:45:23 Join GeekShad0w [0] (
14:55:06 Join wodz [0] (
14:56:27wodzn1s: what's the delta when we build with -O2 instead of -Os?
14:56:31 Join Strife89 [0] (~Strife89@
15:00:27 Quit robin0800 (Remote host closed the connection)
15:00:42 Join robin0800 [0] (
15:05:55 Quit Llorean (Changing host)
15:05:56 Join Llorean [0] (~DarkkOne@rockbox/user/Llorean)
15:13:55LloreanWhy do we need a "mute without pausing" touchregion? Does any player even have that as a button?
15:14:17 Quit newClipUser (Ping timeout: 276 seconds)
15:20:38 Quit Strife89 (Quit: To class.)
15:28:36Stummiis it normal that building the database on android (htc desire hd) seems to need a lot of time?
15:29:29Torneyes, it probably traverses the entire FS
15:29:52 Join leavittx [0] (~lev@
15:30:41 Join slooopy [0] (
15:35:15Stummihm, can i tell rockbox to not search on the entire FS at buildung database but on a specific folder?
15:35:26 Join komputes [0] (~komputes@ubuntu/member/komputes)
15:36:09Tornei don't think we've sorted that kind of thing out int eh app build yet
15:36:58LloreanBut databas.ignore in the root, then database.unignore in the folder?
15:37:05LloreanOr did .unignore never happen?
15:37:29 Quit Judas_PhD (Quit: This is a quitting message)
15:37:45Tornethat doesn't stop it actually traversing the entire FS
15:37:58Tornethat just stops it getting metadata/inspecting the files it finds
15:38:40Torneand you can't put a database.ignore in the root unless you have rooted your device since / is neither user writable nor even mounted writable at all by default
15:39:00LloreanAh, it probably shouldn't traverse anywhere the user can't have written to, at the very least.
15:39:27*Torne is only repeating what he's heard other people say, btw; don't know for sure :)
15:39:36Tornewell, about our lack of ability to control it, that is
15:39:44Tornenot being able to write to the root on android is right :)
15:40:21 Join Judas_PhD [0] (
15:40:51Stummimaybe there could added an option, on which folder the indexing should start for RaaA
15:45:50 Join BaDeD [0] (
15:47:40BaDeDI need help
15:48:10BaDeDI have sansa clip and i need voice in spanish
15:48:35BaDeDanybody help me?
15:48:42Tornethe installer will install any voice file you want
15:49:25 Join evilnick__ [0] (
15:49:41BaDeDI have not file in spanish
15:50:20Tornethe installer can generate them
15:51:29Tornewe only have english voice files for download unfortunately as generating every possible language would be too large
15:51:37Tornebut you can generate it yourself using the rockbox utility
15:52:37BaDeDmanual can generate?
15:53:03 Quit evilnick_ (Ping timeout: 240 seconds)
15:53:45BaDeDi have error in TTS
15:54:47StummiBaDeD, here, just for you ;)">
15:56:38Stummi(hope that works
15:57:20BaDeDthank you stummi
15:59:08BaDeDi have not installate stummi
15:59:15BaDeDyou help me?
15:59:29Stummihm? What do you mean?
15:59:48 Quit factor (Ping timeout: 240 seconds)
16:00:29Tornejust copy the file onto your player in the folder .rockbox\langs
16:00:39Tornenext to espanol.lang
16:02:44 Join Strife89 [0] (~Strife89@
16:13:16BaDeDi have not listen the menu
16:14:19pixelmaStummi: did you make the voice for a Clip?
16:14:50StummiBaDeD, do you have a clip or clip+?
16:16:01Stummiok, thats may the problem. I made it for the clip ;)
16:16:17 Quit TheLemonMan (Quit: free(me))
16:16:37BaDeDclip+ no voice
16:17:12Stummifor each language and device there is a voice-file
16:17:20BaDeDahhh ok
16:20:49 Join parafin [0] (
16:21:32sideralSaint, TheSeven, gevaerts: Re FS corruption: I've verified there there are still at least 2 files and 2 dirs open when entering USB mode during playback. But even when closing all files before fat_unmount(flush=true), I can still reproduce FS corruption. The strange thing is that often disconnecting/reconnecting the player makes (most of) the corruption go away. Which buffer may I have forgotten to flush? dircache and dir_uncached don't seem to cac
16:21:43 Quit mystica555_ (Read error: Connection reset by peer)
16:22:22gevaertssideral: maybe one of the file operations fails to check for file open?
16:23:45 Quit Strife89 (Quit: Leaving)
16:23:50sideralgevaerts: Good idea... But that would explain corruption after returning from USB mode, not before entering it, no?
16:24:40gevaertsI don't know
16:24:55sideralftruncate is one such candidate
16:25:26gevaertsIf we have code that doesn't properly close files on USB connect, it might also continue doing things while USB is connected, or after, and do lots of weird things
16:25:33n1swodz: -o is about 30k larger than -Os i can try -O2 too
16:26:16gevaertssideral: maybe time to add some logf() debugging to file.c, and disk.c, and check what actually happens?
16:27:16n1swodz: -O2 is about 78k larger than -Os
16:27:23sideralgevaerts: yeah. I suspect the playlist-control file is one of the files that remain open, as I've seen it corrupted several times
16:27:57 Join Facefoxdotcom [0] (
16:29:19***Saving seen data "./dancer.seen"
16:30:39 Quit leavittx (Ping timeout: 255 seconds)
16:31:01 Quit kugel (Remote host closed the connection)
16:31:42wodzn1s: hmm that is about 20% larger so we should probably stick to -Os and force inlines for what we need
16:34:06 Quit BaDeD (Quit: CGI:IRC (Ping timeout))
16:34:22n1syeah, very little code in the core is speed critical so it makes sense but is ugly :/
16:34:47wodzn1s: does from svn builds new gcc for CF?
16:35:51n1s-Winline is very useful for this if you want to look at it
16:36:16n1sit prints a warning whenever a call to an inline function isn't inlined and why
16:36:23wodzbut does it substitute the previous one or you can build both (as for arm)?
16:37:42n1sif you want to have both available it's probably best to use a different prefix so it gets its own dir but they should be able to coexist, the old binutils will be overwritten though
16:38:55 Join JesusFreak316 [0] (
16:39:03n1sand the default m68k-elf-gcc but there's a m68k-elf-gcc-3.4.6 still there with all its multilibs and stuff still available
16:39:31Zagoryou can safely compile the new over the old
16:39:45Zagorthat's what I have
16:39:57n1sbut does the old one still work?
16:40:29Zagoror, I haven't done much testing. but it runs at least.
16:40:49gevaertsMaking the build system actually use it might be fun though
16:41:37n1sis anyone against defining a macro such as: #define INLINE inline __attribute__((always_inline)) ?
16:41:50n1sand sprinkling that liberally?
16:42:04Bagderif it helps, I'm in
16:42:40sideralgevaerts: The funny thing is the disappearing corruption when rebooting to the OF. It's as if RB's storage layer returns outdated data that isn't actually on the disk
16:43:04gevaertsoh, that's interesting too...
16:43:38gevaertsNot that I could guess how that would happen
16:44:40sideralI know that the Clip's OF fiddles around with the disk before exposing it to hide the ##MUSIC and ##PORT directories, but I doubt that it would go as far as repairing the FS during bootup :)
16:44:53Tornen1s: only if you call it PRETTY_PLEASE_INLINE
16:46:01 Join leavittx [0] (~lev@
16:46:12n1sTorne: should there be a smiley after that or are you serious?
16:46:31n1soh, nevermind
16:46:49n1si asked if people were against it
16:48:40n1si'll cook something up
16:49:22TorneOkay, to be really serious: yes, add such a thing, but I would prefer it wasn't called INLINE
16:49:38Tornebecause it's fairly common for INLINE to expand to regular inline or nothing, according to compiler/config/etc
16:49:41Zagormaybe FORCE_INLINE
16:49:51TorneYeah, FORCE_INLINE i like
16:51:57 Join benedikt93 [0] (~benedikt9@unaffiliated/benedikt93)
16:54:24 Join delt [0] (
16:54:28deltHello everyone
16:55:16deltquestion about sandisk e200: what device is the microSD slot on the pc?
16:55:48 Join BaDeD [0] (
16:56:02n1sdelt: depends on your OS etc
16:56:35deltkernel sees both sda1 and sda2 even when there's no card.... can mount sda1 (internal 4gb) but sda2 can't find valid vfat (seems to be formatted)
16:57:03n1sthat's the firmware partition
16:57:07deltand i don't want to risk scrapping the original firmware by doing "mkdosfs /dev/sda2"
16:57:09gevaertsThose are both partitions on the internal flash
16:57:10Tornesda is the internal drive
16:57:18deltoh, tha's what i thought
16:57:19n1sthe sd card is a different device
16:57:20Tornethe slot will be sdb, probably
16:57:31Torneif the internal storage is sda :)
16:57:54deltmount: special device /dev/sdb1 does not exist
16:58:05gevaertsis it partitioned?
16:58:10gevaertsDoes /dev/sdb exist?
16:58:15n1sis there a card inserted?
16:58:18deltor is the filesystem written directly without a partition table?
16:58:27deltmount: special device /dev/sdb does not exist
16:58:43gevaertsAre you running a self-compiled kernel?
16:58:50delt[pts/9][user@deimos]:~$ ls /dev/sd*
16:58:50delt/dev/sda /dev/sda1 /dev/sda2
16:59:09n1sdelt: you *do* have a card in the slot, right?
16:59:10deltwell, gcc compiled it, not me =)
16:59:12gevaertsYou probably have CONFIG_SCSI_MULTI_LUN=n
16:59:15gevaertsn1s: doesn't matter
16:59:49n1sgevaerts: oh
16:59:57gevaertsOr set the max_report_luns option for sd_mod
17:00:10delt[pts/9][user@deimos]:~$ zcat /proc/config.gz | grep CONFIG_SCSI_MULTI_LUN
17:00:10delt# CONFIG_SCSI_MULTI_LUN is not set
17:00:37gevaertsn1s: it's not a removable device, it's a removable *media* device. Don't believe what 99.9% of the people are telling you!
17:01:07 Quit BaDeD (Quit: CGI:IRC (Ping timeout))
17:01:07deltso what does CONFIG_SCSI_MULTI_LUN do?
17:01:11n1sright, these usb people really knew how to give things clear descriptive names ;)
17:01:49gevaertsIt enables scanning multiple LUNS, which you need to use devices with multiple LUNS, such as most card readers, or sansas :)
17:02:06deltie. how would it be called in the kernel's make *config ?
17:02:29delt*checks LUN on wikipedia* ...
17:02:43gevaertsno idea, but I think searching for CONFIG_SCSI_MULTI_LUN (using "/" as the search shortcut key) works in make menuconfig
17:03:13deltyeah it would
17:03:27 Join ehntoo [0] (
17:04:10gevaertsdelt: possibly playing with max_luns and max_report_luns in /sys/module/scsi_mod/parameters is enough
17:04:11Torneyou don't need the CONFIG_
17:04:13deltlet's see.... ah Prompt: Probe all LUNs on each SCSI device
17:05:01 Quit T44 (Quit: Leaving)
17:05:17delt[pts/2][root@deimos]:~# cat /sys/module/scsi_mod/parameters/max_luns
17:05:26deltmax_report_luns is 511 however
17:05:36gevaertsok. Set max_luns to 2 or more
17:05:37deltso "echo ....." > the file(s)?
17:05:57 Quit Bagder (Remote host closed the connection)
17:06:21deltok, now it says 2 when i cat the file back to the term
17:06:31deltlet's remove and replug the player...
17:07:20gevaertsright. In recent-ish kernels (probably years and years...) CONFIG_SCSI_MULTI_LUN only sets a different default for max_luns
17:07:46deltah, now it sees sdb as 7.94gb (8gb card)
17:08:20 Quit jepler (Quit: .)
17:08:30deltbut weird, kernel says sda: (nothing - usually lists sda1 and sda2 here)
17:08:38 Join jepler [0] (~jepler@emc/developer/pdpc.professional.jepler)
17:08:49delt/dev/sda1 and sda2 exist tho
17:09:33deltcan mount it, seems to work fine etc etc... now let's try the msd card.... i created a directory "Test" on it from the player
17:09:40gevaertsSometimes the sda1 and sda2 appear on another line. The kernel isn't single-threaded any more
17:10:08deltremoved the card and put it back in, "Test/" was stil there, so looks like it's formatted
17:10:21delthere's my "Test" ....
17:11:14 Join BaDeD [0] (
17:11:54deltseems to be working fine, thanks :D
17:12:08 Quit krazykit (Quit: awe yeeeeeee)
17:12:15 Quit BaDeD (Client Quit)
17:12:32deltso i guess i'll just add echo "2" > /sys/module/scsi_mod/parameters/max_luns to my startup scripts
17:12:43 Part Zagor
17:13:10gevaertsif scsi_mod is a module, you should add it to the usual place for module parameters
17:13:22gevaertsBut that's not really on-topic here...
17:15:56deltoh yeah, it auto-mounted modules sg and usbhid
17:16:27 Quit robin0800 (Remote host closed the connection)
17:16:42deltbah, i have a whole section in my rc.local that's "echo stuff > /proc or /sys/whatever"
17:17:57deltanyway thanks a lot for the help!
17:17:59 Part delt
17:19:07 Join delt [0] (
17:19:43deltoh just wondering 1 more thing.... if i create a partition table (and multiple partitions) on the SD card, can rockbox still use (probably the first one)?
17:20:29 Join domonoky [0] (~Domonoky@rockbox/developer/domonoky)
17:21:12Torneif you partition it rockbox will probably just use the first one
17:21:24*gevaerts checks
17:21:27Torneyou can build rockbox with multi partition support but i suspect that isn't
17:21:30 Quit leavittx (Ping timeout: 255 seconds)
17:22:41gevaertsWell, "the first one" being "the first usable one", i.e. the first FAT partition
17:23:01deltso to be consistent with most other devices, i can create a partition table and just one big fat32 that takes up the whole card.... that would work?
17:23:41 Join BaDeD [0] (
17:23:50delta filesystem directly on sdb seems to me kind of like the first chapter of a book starting right on the cover
17:24:13deltbut that's just me... thanks again
17:24:14 Part delt
17:25:31 Quit kevku (Ping timeout: 272 seconds)
17:26:04 Join Topy44 [0] (
17:27:31 Join kevku [0] (~kevku@2001:7d0:0:f9af:feed:feed:feed:feed)
17:27:44BaDeDmy sansa clip+ not read my lcd
17:28:05sideralgevaerts: The two remaining open files when entering USB slave mode are the mp3 file being played back and /.rockbox/fonts/08-Rockfont.fnt −− both should be read only, and I see no flushing going on for them. This explains why closing the files does not help with syncing the disk :)
17:30:45gevaertssideral: the font file being open is interesting
17:31:04Tornethe font cache doesn't ever close the UI font, no?
17:31:17Tornewe just stop using the UI font, rather than doing font_load(NULL)
17:31:23Torneunless i forget
17:31:36gevaertsI think so
17:32:16gevaertsBut if the file is still open, things may well do weird things later on
17:32:53sideralthere's lots of other strange stuff going on before and after the USB connection, but nothing looks particularly suspicious. For example: what is /.rockbox/wps/ ?
17:33:03 Quit BaDeD (Quit: CGI:IRC (Ping timeout))
17:33:32 Join t0rc [0] (~t0rc@
17:33:34 Quit t0rc (Changing host)
17:33:34 Join t0rc [0] (~t0rc@unaffiliated/t0rc/x-5233201)
17:33:37 Join leavittx [0] (~lev@
17:33:51sideraland /.rockbox/icons/-.icons ?
17:34:23 Quit t0rc (Read error: Connection reset by peer)
17:34:29sideraland /.rockbox/wps/.rockbox/wps/rockbox_failsafe.fms ?
17:34:29gevaertssideral: '-' is a common default in settings. Thoseprobably just mean that no sbs or icons file was specified
17:35:30 Quit kevku (Ping timeout: 260 seconds)
17:36:24 Quit petur (Quit: Page closed)
17:37:14 Quit STeeF (Remote host closed the connection)
17:37:24sideralthere's also lots of lseek activity and single-byte writes (which hopefully are cached in the FAT layer? −− I'll look into those later)
17:37:37 Quit einhirn (Quit: Miranda IM! Smaller, Faster, Easier.
17:38:32 Quit TheSeven (Ping timeout: 250 seconds)
17:46:04 Join factor [0] (~factor@
17:46:37sideralNeed to run now −− I'll check later whether you guys have had other interesting ideas. Bye
17:46:55 Join stoffel [0] (
17:47:59 Quit sideral (Quit: Leaving.)
17:48:50 Quit GeekShad0w (Ping timeout: 246 seconds)
17:52:33 Quit leavittx (Ping timeout: 255 seconds)
17:57:50 Join panni_ [0] (
18:18:32 Quit swilde (Remote host closed the connection)
18:29:22***Saving seen data "./dancer.seen"
18:35:33 Join kevku [0] (~kevku@2001:7d0:0:f9af:215:c5ff:fe7e:794d)
18:36:04 Join sideral [0] (~sideral@unaffiliated/sideral)
18:36:36 Quit [fred] (Ping timeout: 276 seconds)
18:37:58 Join leavittx [0] (~lev@
18:39:31 Join newClipUser [0] (
18:40:13 Quit JesusFreak316 (Read error: Connection reset by peer)
18:41:01 Join [fred] [0] (
18:43:01 Quit sideral (Remote host closed the connection)
18:43:38 Join sideral [0] (~sideral@unaffiliated/sideral)
18:46:21 Join JesusFreak316 [0] (
18:52:44 Join TheLemonMan [0] (
18:53:16 Quit sideral (Remote host closed the connection)
18:53:27 Part LinusN
18:53:31 Join sideral [0] (~sideral@
18:53:34 Quit sideral (Changing host)
18:53:34 Join sideral [0] (~sideral@unaffiliated/sideral)
19:02:54 Quit markun (Read error: Connection reset by peer)
19:03:31 Quit sideral (Remote host closed the connection)
19:04:16 Join sideral [0] (~sideral@unaffiliated/sideral)
19:06:07 Quit newClipUser (Ping timeout: 264 seconds)
19:06:33 Quit simonrvn (Quit: ZNC -
19:06:57 Join markun [0] (
19:06:57 Quit markun (Changing host)
19:06:57 Join markun [0] (~markun@rockbox/developer/markun)
19:07:22 Join simonrvn [0] (
19:09:42CIA-7New commit by 03alex (r29048): Manual: Fix part 1 of FS #11874 - touchscreen region in the manual isn't 100% correct.
19:09:44CIA-7New commit by 03alex (r29049): Manual: Fix part 2 of FS #11874 - setting_inc and setting_dec missing from actions list.
19:09:52CIA-7New commit by 03alex (r29050): Manual: Fix FS #11872 - %pL/pR aren't in the manual.
19:11:57 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
19:14:26 Quit JesusFreak316 (Remote host closed the connection)
19:14:45CIA-7r29048 build result: All green
19:15:59pixelmaAlexP: doesn't completely have to do with the commit but I noticed something in the diff - above the %pm etc. there is an \nopt{player} - I'd think \opt{lcd_bitmap} is more suitable
19:16:53AlexPpixelma: Could well be - to be honest I have no idea what any of the archos players can and can't do :)
19:17:27pixelmateh Player is the only charcell device
19:18:27pixelmathe others have monochrome displays, so are "only" limited by that when looking at the skinning abilities
19:18:37 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
19:19:07CIA-7r29050 build result: All green
19:19:52*wodz is reminded of putting rockbox on his archos studio 20
19:20:08amiconnpixelma: Regarding peakmeters - the player not having it has nothing to do with bitmap vs. charcell
19:20:37amiconnIt's indeed Player specific - the MAS3507D doesn't have a way to read out peak data
19:21:02pixelmathere's also a MASD opt from features.txt
19:21:52pixelmaor a "masd" feature
19:22:26 Quit DerPapst1 (Quit: Leaving.)
19:22:27 Quit DerPapst (Quit: Leaving.)
19:22:29amiconnThe peakmeters should probably be opted for swcodec and masf
19:23:21amiconnOther potential (unlikely, but still...) hwcodec targets would need special adjustment too (if possible to implement)
19:24:36pixelmathat's also possible
19:24:44pixelmato do in the manual
19:28:09 Join newClipUser [0] (
19:32:33 Join GeekShadow [0] (
19:32:33 Quit GeekShadow (Changing host)
19:32:33 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
19:32:50*pixelma is reminded of finishing the target rename in the manual (including general \\opt cleanup) :/
19:34:52 Join Horscht [0] (~Horscht@xbmc/user/horscht)
19:40:33 Quit leavittx (Ping timeout: 255 seconds)
19:43:23 Quit stoffel (Remote host closed the connection)
19:44:34 Join BaDeD [0] (
19:44:34 Quit BaDeD (Client Quit)
19:44:45 Join BaDeD [0] (
19:45:08 Join Keripo [0] (~Keripo@
19:45:27 Join JesusFreak316 [0] (
19:48:19 Join solexx [0] (
19:51:43 Quit solexx_ (Ping timeout: 264 seconds)
19:52:06 Quit BaDeD (Quit: CGI:IRC (Ping timeout))
19:53:09 Join 20QAAE1S1 [0] (
19:53:09 Join 52AAACRYL [0] (
19:55:58 Join Buschel [0] (
19:58:05Buscheln1s: do you have any codec comparisons of old and new gcc on CF?
20:00:13Buschel(ore better: pre-r29042 vs. post-r29042)
20:00:14 Join evilnick [0] (
20:01:15 Quit evilnick__ (Ping timeout: 240 seconds)
20:08:45pixelmaBuschel: dev-ml ;)
20:11:45 Join leavittx [0] (~lev@
20:13:54 Quit newClipUser (Ping timeout: 265 seconds)
20:14:10 Join kugel [0] (~kugel@rockbox/developer/kugel)
20:23:58 Join newClipUser [0] (~newClipUs@
20:29:26***Saving seen data "./dancer.seen"
20:33:39sideralgevaerts: Re FS corruption vanishing when player is reconnected or rebooted to OF: What's the chance of this being a DMA/caching problem, where the right data is committed to disk but masked by invalid CPU-cache contents when reread?
20:44:00 Quit newClipUser (Ping timeout: 255 seconds)
20:44:48 Join bmbl [0] (~bmbl@unaffiliated/bmbl)
20:46:51 Quit bmbl (Client Quit)
20:50:26 Quit Keripo (Quit: Leaving.)
20:52:26gevaertssideral: sounds unlikely to me
20:56:15 Quit chattr (Quit: gone)
20:56:22 Join newClipUser [0] (
20:59:30 Join YouKay [0] (~Aryan@unaffiliated/rocket16)
20:59:39YouKayHello gentlemen
21:00:07YouKayI have a Philips GoGear Muse, 8 GB. Can rockbox be installed on it?
21:01:47YouKaygevaerts: I see. Thanks. But can't it be customized in any way to replace the original firmware? (just curious(
21:02:29 Join LambdaCalculus37 [0] (~3f74f70d@rockbox/staff/LambdaCalculus37)
21:03:01LambdaCalculus37jhMikeS: Ping
21:03:23 Quit user890104 (Ping timeout: 272 seconds)
21:04:26gevaertsYouKay: I'm not aware of anything similar to rockbox out there. There might be some mods specifically for that player around, but nobody here is likely to know about those. You could port rockbox, but that's a lot of work (see
21:06:07YouKaygevaerts: I see. Many thanks for the link, I guess I will give it a shot (although I am not sure how much I can perform).
21:07:20LambdaCalculus37YouKay: The first thing to do is find out what kind of hardware is inside this GoGear model.
21:09:49*LambdaCalculus37 is having a little trouble getting rbutil to build on OS X
21:12:39CIA-7New commit by 03bluebrother (r29051): Make m68k-elf 4.5.2 toolchain build on OS X 10.6. ...
21:13:24YouKayI see. Thanks LambdaCalculus37 and gevaerts
21:13:49 Part YouKay ("We shall overcome!")
21:15:07 Quit LambdaCalculus37 (Quit: back to work again)
21:16:24 Quit leavittx (Ping timeout: 255 seconds)
21:18:20CIA-7r29051 build result: All green
21:19:12 Quit esperegu (Remote host closed the connection)
21:19:32n1sbluebroth3r: ah, cool
21:20:28n1sdidn't realize we needed that workaround, is the −−disable-nls part not needed too?
21:20:47bluebroth3rn1s: no, figured that it's not necessary.
21:21:01bluebroth3rso we could enable nls for arm-eabi as well
21:21:10 Join leavittx [0] (~lev@
21:21:33n1smissing libiconv or something else?
21:21:48bluebroth3rthough I'm not sure if nls is of too much use for a compiler. I'm always annoyed when looking at localized compiler output (it's worse to understand even if it's my native language!)
21:21:59n1si agree
21:22:13gevaertsIf it's for the build system, I'd say we explicitely *don't* want it
21:22:37n1sbut since our generated makefiles export LANGUAGE=english i always get English messages anyway
21:23:04n1swhich is a leftover from when you could select the default lang but still nice :)
21:23:58bluebroth3rgevaerts: build system as in automated build system or in build system for Rockbox?
21:24:27gevaertsWell, that LANGUAGE=english thing fixes it anyway
21:24:40 Join user890104 [0] (~Venci@2001:0:5ef5:79fd:28b2:318:2b16:10ec)
21:25:30 Quit preglow (Ping timeout: 246 seconds)
21:25:38bluebroth3rnow let's see what's wrong with zezayer
21:26:31 Quit aevin (Ping timeout: 260 seconds)
21:26:39bluebroth3rbtw, what do people think about using png images for the sim? When using SDL_image this is pretty easy and makes the background images noticably smaller
21:26:49 Join chattr [0] (
21:26:55bluebroth3rthough downloading the bmp files might not be that much of an issue since bmp usually compresses quite nice
21:27:51bluebroth3rso it might not be worth the additional dependency
21:29:30 Join aevin [0] (eivindsy@unaffiliated/aevin)
21:35:42pixelmabluebroth3r: what about zezayer?
21:35:46bluebroth3rhmm, looks like a filesystem problem on my player that broke zezayer :/
21:35:57bluebroth3rat least it works fine in the sim. Time for fsck
21:36:47pixelmathe amount of filesystem problems reported lately is a bit "scary"
21:38:36pixelmabluebroth3r: is this on your Mini or H100?
21:38:45bluebroth3rwell, in this case I don
21:38:53bluebroth3r#t consider it a Rockbox issue :)
21:39:19bluebroth3rpixelma: h100, and for some strange reason it disconnected itself from the Mac after I copied the new build onto it
21:40:12bluebroth3rprobably just a somewhat broken cable in combination with the keyboard hub I've used (had the impression that this hub is a bit flaky in the past anyway)
21:42:09pixelmainteresting that it works at all on it, when I tried (on an older OSX probably) the Mac didn't even recognise my c200 when connected to the keyboard hub
21:44:46 Quit newClipUser (Ping timeout: 265 seconds)
21:48:22 Quit user890104 (Ping timeout: 272 seconds)
21:48:42 Quit Buschel (Quit: ChatZilla 0.9.86 [Firefox 3.6.13/20101203075014])
21:48:55gevaertsI'd expect most keyboard hubs to only allow 100mA devices
21:49:54bluebroth3rshouldn't be a problem with the h100 since it cannot charge from USB anyway :)
21:50:50 Quit chattr (Ping timeout: 240 seconds)
21:51:09gevaertsAh, that would probably be the reason for the difference then
21:52:34 Join Highlander [0] (
21:52:38bluebroth3rseems I can crash the UI reproducably with that build :o
21:53:44n1swhat UI?
21:54:08bluebroth3rh100. Music keeps playing, buttons toggle backlight, but I can't navigate in the file browser anymore
21:55:48n1show do you do it?
21:56:04bluebroth3rlet's see if that still happens when I flash rombox
21:56:39 Quit captainkewll (Quit: Page closed)
21:56:41bluebroth3rjust use the Stop button
21:56:46*bluebroth3r tries again
21:57:59bluebroth3rcrash again
21:59:24 Quit JesusFreak316 (Ping timeout: 240 seconds)
21:59:37bluebroth3rmaybe it's related to the fact that my h100 is rtc modded?
22:02:01n1sstopping playback works fine on my h300 at least
22:02:31n1sbluebroth3r: are you certain it was introduced at the same time as the compiler switch?
22:02:57bluebroth3rn1s: no, I haven't updated the h100 since like half a year or so
22:04:57bluebroth3rnow crash when trying to load a file while another one is playing
22:05:42bluebroth3rI guess I need to build with the old gcc and check that
22:05:45n1scould you try with a build made with the old compiler? And did you fix the disk corruption?
22:06:09n1sdailies from before yesterday were made with the old gcc
22:06:30bluebroth3rnice, I still have the old m68k-elf around on this machine (and no updated one)
22:10:21 Quit benedikt93 (Quit: Bye ;))
22:19:55bluebroth3rn1s: same problem with r29039
22:20:09 Join bmbl [0] (
22:20:09 Quit bmbl (Changing host)
22:20:09 Join bmbl [0] (~bmbl@unaffiliated/bmbl)
22:20:44n1sbluebroth3r: ah, so not cause by the new compiler, then
22:21:05bluebroth3rskipping doesn't work, trying to play another track while music is playing or hitting stop results in music continuing to play, but display showing the last active screen. If it's the WPS it's static
22:21:41bluebroth3rmaybe I should try with rockbox_default as theme
22:21:48 Join JesusFreak316 [0] (
22:21:59 Join chattr [0] (
22:22:00 Join petteyg [0] (
22:22:11n1sbluebroth3r: very odd behaviour
22:23:04bluebroth3rhmm, using rockbox_failsafe causes the same issue.
22:23:12bluebroth3rso it doesn't look like being theme related
22:24:10 Part chattr
22:24:19bluebroth3rlet's try a build without RTC mod
22:26:50 Quit Barahir_ (Ping timeout: 265 seconds)
22:28:07 Join Barahir [0] (
22:29:04wodzn1s: how should I build new compiler to coexist with old one?
22:29:30***Saving seen data "./dancer.seen"
22:30:36n1swodz: either use a different prefix and adjust the PATH or just use the same and hack the makefile slightly to select the old one when you need it
22:31:20wodzso the binary itself will have different name like m68k-elf-gcc-xxx ?
22:33:06n1syes, you'll have a m68k-elf-gcc-3.4.6 and m68k-elf-gcc-4.5.2 but the plain m68k-elf-gcc will be the one that is installed last (if you use the same prefix) or the one that's first in PATH if you use different prefixes
22:36:10 Quit n1s (Quit: Lämnar)
22:36:58 Quit GeekShadow (Ping timeout: 255 seconds)
22:47:20 Quit Highlander (Quit: Quitte)
22:48:28wodzTheSeven: for the record - one of the distro tracker entry about ehci_hcd problems
22:50:23 Join user890104 [0] (~Venci@2001:0:4137:9e76:c7f:3a54:2b16:10ec)
22:52:24bluebroth3rhmm, 3.7.1 _also_ shows the issue? Now it's getting really weird.
22:56:30 Quit JesusFreak316 (Ping timeout: 240 seconds)
22:56:55bluebroth3rn1s: 3.6 works, 3.7 works, 3.7.1 does show the problem.
22:57:16TheSeventhat limits it to a very small number of commits at least
22:57:22bluebroth3rso it's "only" around 1 month of commits to bisect
22:57:31wodzbluebroth3r: 3.7 -> 3.7.1 should be fairly easy to bisect
22:57:43bluebroth3rwodz: true, but it's still work ;-)
22:57:53 Quit bertrik (Ping timeout: 240 seconds)
22:58:10wodzbluebroth3r: It's your fault - you found the bug :-)
22:58:27bluebroth3rargh crap :)
22:59:04*bluebroth3r needs to remove that rubber case from the player −− using the paperclip debug tool is somewhat ugly with that
22:59:31 Quit factor (Read error: Connection reset by peer)
22:59:47 Join factor [0] (~factor@
23:01:30wodzTheSeven: Problems with my ipods are caused by ehci_hcd in recentish linux kernels definitely. Various people report problems with different players/thumb drives/and some other usb2.0 stuff. The only workaround is to disable ehci temporary for now.
23:14:40 Quit kugel (Remote host closed the connection)
23:15:20 Quit factor (Read error: Connection reset by peer)
23:20:30 Join JdGordon| [0] (
23:20:30 Quit JdGordon| (Changing host)
23:20:30 Join JdGordon| [0] (~jonno@rockbox/developer/JdGordon)
23:20:51 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
23:20:56 Quit TheLemonMan (Quit: free(me))
23:22:55 Join factor [0] (~factor@
23:24:11 Quit mystica555 (Read error: Connection reset by peer)
23:27:46 Join newClipUser [0] (
23:30:33 Join Buschel [0] (
23:30:33 Quit bmbl (Quit: Verlassend)
23:30:42 Quit domonoky (Read error: Connection reset by peer)
23:31:30Buschelbluebroth3r: as you are bisecting your playback issue -> r28652 (3.7 branch) or r28608 (trunk) look suspicous to my eyes
23:32:05Buschelbluebroth3r: this may also be another hint/idea regarding FS #11863 (playback lockup's or lockup's during startup)
23:32:25 Quit newClipUser (Ping timeout: 260 seconds)
23:32:48Buscheljust wanted to write this to the logs before my notebook battery dies (in a few minutes)
23:35:23 Quit Buschel (Client Quit)
23:36:00 Join newClipUser [0] (
23:43:03 Join fdinel [0] (
23:43:43 Quit wodz (Quit: Leaving)
23:44:55 Quit Rob2222 (Read error: Connection reset by peer)
23:49:11 Join Rob2222 [0] (
23:51:13 Quit newClipUser (Ping timeout: 276 seconds)
23:51:29 Join mystica555 [0] (
23:55:31 Join mystica555_ [0] (
23:55:34 Join kugel [0] (~kugel@rockbox/developer/kugel)
23:56:27 Quit mystica555 (Ping timeout: 240 seconds)

Previous day | Next day