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).

#rockbox log for 2011-10-02

00:05:33 Join ChickeNES [0] (
00:06:10 Quit pamaury (Ping timeout: 244 seconds)
00:12:16 Join pamaury [0] (
00:12:16 Quit pamaury (Changing host)
00:12:16 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
00:12:19 Quit ender` (Quit: facebook is like a jail: you sit around, waste time, have a profile picture, write on walls and get poked by guys you don't really know.)
00:20:44 Quit keyb_gr (Ping timeout: 260 seconds)
00:28:47pamaurywodz: did you know there were rk27xx available ?
00:32:37 Join powell14ski_ [0] (
00:36:00 Quit robin0800 (Quit: Leaving)
00:36:54 Quit nrg_ (Quit: leaving)
00:38:10 Join robin0800 [0] (~robin0800@
01:01:37 Quit ReimuHakurei (Remote host closed the connection)
01:03:17 Quit pamaury (Remote host closed the connection)
01:05:39 Join ReimuHakurei [0] (
01:07:57 Quit bertrik (Quit: And That, My Liege, Is How We Know the Earth to Be Banana Shaped)
01:21:26***Saving seen data "./dancer.seen"
01:26:21 Part lovasoa
01:27:11 Quit mshathlonxp ()
01:31:52 Quit ReimuHakurei (Remote host closed the connection)
01:35:33 Join ReimuHakurei [0] (
01:47:10 Quit AlexP (Remote host closed the connection)
01:58:42 Join Jak_o_Shadows [0] (
02:15:22 Quit KiwiCam (Remote host closed the connection)
02:16:52 Quit robin0800 (Ping timeout: 260 seconds)
02:29:21 Join robin0800 [0] (
02:30:08 Quit n1s (Remote host closed the connection)
02:43:29 Join T44 [0] (
02:46:38 Quit Topy (Ping timeout: 248 seconds)
03:06:28JdGordonZagor: hope you read logs... so i dont forget, skin viewports can be made screen size independant using negative numbers in the skin file
03:06:55JdGordonany viewports setup in code already automagically scale to the screen size
03:07:28JdGordonso now the issue is really the background image (which needs it to be the correct size) and what to do aobut the native statusbar
03:08:52JdGordonFS #11615 would unify all the raaa screen sizes to 1 build which would make sense if the usable lcd size changes anyway
03:08:53fs-bluebot Dynamic screen size (patches, new)
03:09:55JdGordonIMO the best way to deal with the native statusbar is to push the rockbox framebuffer down so 0,0 is still the top left of the usaeble frambebuffer
03:10:11JdGordonthen if skins are written correctly it will all just work
03:10:13 Join Rob2222 [0] (
03:14:07 Quit Rob2223 (Ping timeout: 258 seconds)
03:18:50 Quit ChickeNES (Quit: Computer has gone to sleep.)
03:18:52 Join Jak_o_Shadows1 [0] (
03:21:21 Join Topy [0] (
03:21:30***Saving seen data "./dancer.seen"
03:21:34 Quit Jak_o_Shadows (Ping timeout: 260 seconds)
03:25:02 Quit T44 (Ping timeout: 248 seconds)
03:38:46 Quit robin0800 (Quit: Leaving)
03:42:57 Join robin0800 [0] (
04:04:42 Quit robin0800 (Quit: Leaving)
04:05:26 Join robin0800 [0] (
04:09:49 Quit CaptainKewl (Quit: ( :: NoNameScript 4.22 :: ))
04:18:35 Quit mystica555 (Ping timeout: 258 seconds)
04:34:24 Quit amiconn (Disconnected by services)
04:34:24 Join amiconn_ [0] (quassel@rockbox/developer/amiconn)
04:34:46 Nick amiconn_ is now known as amiconn (quassel@rockbox/developer/amiconn)
04:35:24 Quit pixelma (Disconnected by services)
04:35:26 Join pixelma_ [0] (quassel@rockbox/staff/pixelma)
04:35:28 Nick pixelma_ is now known as pixelma (quassel@rockbox/staff/pixelma)
04:37:48 Quit TheSeven (Disconnected by services)
04:37:59 Join [7] [0] (~TheSeven@rockbox/developer/TheSeven)
05:10:16 Quit Rob2222 (Ping timeout: 255 seconds)
05:16:16 Join Rob2222 [0] (
05:16:29 Quit froggyman (Read error: Connection reset by peer)
05:16:33 Join froggyman_ [0] (~seth@
05:17:01 Quit froggyman_ (Client Quit)
05:17:35 Join froggyman [0] (~seth@unaffiliated/froggyman)
05:21:33***Saving seen data "./dancer.seen"
05:26:27 Quit ReimuHakurei (Remote host closed the connection)
05:30:38 Join ReimuHakurei [0] (
05:33:16 Quit hilbert (Quit: Computer has gone to sleep.)
05:38:49 Join hilbert [0] (~hilbert@
05:43:13 Join [Saint_AndChat] [0] (~Saint]@
05:52:12 Join FoH [0] (
06:22:15 Quit [Saint_AndChat] (Remote host closed the connection)
06:32:00 Quit robin0800 (Quit: Leaving)
06:44:04 Quit hilbert (Quit: Textual IRC Client:
06:51:46 Join [Saint_AndChat] [0] (~Saint]@
06:52:11[Saint_AndChat]Finally got the hang of fake pich-to-zoom.
06:53:10[Saint_AndChat]I'd really prefer the "fake" finger was indeed a real finger, and not a fixed position in the middle of the screen, though.
06:53:42[Saint_AndChat]Dammit...this isn't the channel I wanted ;)
06:53:57*[Saint_AndChat] skulks away...
07:21:36***Saving seen data "./dancer.seen"
07:24:00 Quit [Saint_AndChat] (Read error: Connection reset by peer)
07:58:34 Join T44 [0] (
08:01:51 Quit Topy (Ping timeout: 248 seconds)
08:07:14 Join Keripo [0] (
08:10:02 Quit powell14ski_ (Quit: powell14ski_)
08:29:12 Quit FoH (Quit: ¡ooʇ ‘ǝןdoǝd ǝɹɐ sʇɐq)
08:42:23 Quit Bagder (Ping timeout: 248 seconds)
09:14:20 Join Topy44 [0] (
09:17:03 Quit T44 (Ping timeout: 248 seconds)
09:21:40***Saving seen data "./dancer.seen"
09:31:50 Join dfkt|n [0] (
09:31:51 Quit dfkt|n (Changing host)
09:31:51 Join dfkt|n [0] (~dfkt@unaffiliated/dfkt)
09:36:09 Join stoffel [0] (
09:53:39 Join T44 [0] (
09:57:35 Quit Topy44 (Ping timeout: 248 seconds)
10:08:11 Join n1s [0] (~quassel@rockbox/developer/n1s)
10:14:07 Quit T44 (Ping timeout: 248 seconds)
10:14:59 Join Topy44 [0] (
10:16:26 Join fiesh [0] (
10:17:21fiesha question that might have come up before, but if so I didn't find the answer −− has it ever been considered to have the Bauer stereophonic-to-binaural DSP as a plugin for rockbox? (
10:19:59 Quit Topy44 (Ping timeout: 248 seconds)
10:22:13 Join Jak_o_Shadows [0] (
10:23:30 Quit Jak_o_Shadows1 (Ping timeout: 258 seconds)
10:24:35 Join keyb_gr [0] (
10:24:56 Quit keyb_gr (Client Quit)
10:38:45 Quit TBFOOL (Quit: sigh)
10:43:11 Join Bagder [0] (
10:43:12 Quit Bagder (Changing host)
10:43:12 Join Bagder [241] (~daniel@rockbox/developer/bagder)
10:44:23 Join Topy44 [0] (
10:46:53 Join bertrik [0] (
10:46:54 Quit bertrik (Changing host)
10:46:54 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
10:52:53 Join TBCOOL [0] (
10:57:12CIA-14New commit by bluebrother (r30631): Rockbox Utility: add some more icon sizes to the window icon on Windows.
10:59:32CIA-14r30631 build result: All green
11:06:03 Quit dfkt|n (Ping timeout: 258 seconds)
11:09:01 Join ender` [0] (
11:13:08 Join T44 [0] (
11:15:00 Quit Topy44 (Ping timeout: 240 seconds)
11:21:42***Saving seen data "./dancer.seen"
11:27:20 Join pamaury [0] (
11:27:20 Quit pamaury (Changing host)
11:27:20 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
11:31:36 Join Horschti [0] (~Horscht@xbmc/user/horscht)
11:35:01 Quit Horscht (Ping timeout: 255 seconds)
11:36:49 Join bluebrother [0] (
11:36:50 Quit bluebrother (Changing host)
11:36:50 Join bluebrother [0] (~dom@rockbox/developer/bluebrother)
11:36:58bertrikDid anything result from the FOR_NB_SCREEN(i) discussion yesterday? I had to leave early before any conclusion.
11:38:55 Quit fs-bluebot (Ping timeout: 248 seconds)
11:40:03 Join fs-bluebot [0] (
11:40:10 Quit bluebrother^ (Ping timeout: 258 seconds)
11:52:27wodzbertrik: not really
11:53:26bertrikI was confused thinking that i was used as an argument in viewportmanager_theme_enable, but that function is an empty define for HWCODEC
11:54:07bertrikand I remember amiconn saying that the problem is actually in the viewportmanager_... code
11:54:25bertrikI see now that these functions take an enum screen_type instead of an int
11:54:50amiconnIt's not empty for hwcodec, it's empty for charcell
11:54:56amiconnThat's the problem/ bug
11:56:43n1swe could just #ifndec the whole loops for charcell i guess
11:58:02amiconnThat would be a stopgap fix for the warning, yes
11:59:09bertrikWhat about using a real but empty viewportmanager_theme_enable function for charcell? or would that waste binsize?
11:59:37amiconnThe proper fix would be enabling the viewport stuff for charcell
12:00:05wodzrk27xx lcdif block drives me mad :/
12:00:19n1sis that ever going to happen?
12:00:21amiconnThat's something I might even try to do, provided I get svn rockbox boot on hwcodec again
12:00:50wodzamiconn: this will have considerable impact on binsize, no?
12:01:23amiconnYes, but that's no problem
12:01:52amiconnThe Player has the smallest binary of all archoses anyway, even when viewports would be enabled
12:04:40amiconnIn fact having viewports on the Player would be rather useful, because of the small display
12:14:39 Join Jerom [0] (~jerome@
12:15:35[7]er, was that a phone typing in a pocket? :)
12:22:28bertrikanyone here with a clipv2?
12:22:51 Join Jak_o_Shadows1 [0] (
12:24:38 Quit Jak_o_Shadows (Ping timeout: 258 seconds)
13:03:28 Join JdGord [0] (~AndChat@
13:10:47 Quit Jak_o_Shadows1 (Remote host closed the connection)
13:21:46***Saving seen data "./dancer.seen"
13:22:00bertrikTorne, can you have a look at FS #12294 ?
13:22:01fs-bluebot No SD card in current build (bugs, unconfirmed)
13:33:30Tornei don't see how that can happen
13:34:41Tornewe don't have any check for plausibility on the MBR, s oit'll just interpret random crap as a partition table
13:34:46Tornebut fat_mount should just fail
13:34:55Torneand if it fails to mount any of the four partitions then it tries to mount the whole disk
13:35:24TorneRelying on partition type checks to make that work is a pretty terrible idea, that's still not validating that it's really an MBR :)
13:36:16 Join TomColler [0] (
13:37:30Tornethe only way i can see that happening is if one of the garbage "partitions" *does* succeed in fat_mount :)
13:39:32CIA-14New commit by roolku (r30632): revert r30566 and r30567 to fix FS #12294 (superfloppy not mounting)
13:39:58Tornenot a fix!
13:41:11CIA-14r30632 build result: All green
13:45:00 Quit TomColler (Ping timeout: 240 seconds)
13:47:51bertrikoh pity, this needs further investigation, not just a revert
13:49:06Tornefat_mount may not be bein careful enough about checking if the fs is sane
13:49:32Tornei've reopened the task and commented
13:49:52Tornewe should probably go see what partition_msdos in linux does to decide if an MBR is plausible :)
13:53:06Tornei don't have time to do it today, probably, but i'll try and remember to look soon
14:03:55 Join hilbert [0] (~hilbert@
14:20:01 Quit Jerom (Quit: Leaving.)
14:46:31 Nick [Sanyi] is now known as sanyi (thg@gateway/shell/
14:47:37 Nick sanyi is now known as [Sanyi] (thg@gateway/shell/
14:49:26 Join chkktri_ [0] (
14:58:02 Join GermanMushroom [0] (
14:59:37 Join robin0800 [0] (
15:01:56 Quit linuxstb (Ping timeout: 256 seconds)
15:06:36 Join liar [0] (
15:07:18 Join DMJC [0] (
15:07:28DMJChey, has anyone here played with the android build?
15:08:19 Quit GermanMushroom (Ping timeout: 255 seconds)
15:08:25DMJCI used to run Rockbox on my iRiver H140, but I'm trying to get it going under android atm. I got it built/installed/executing, but the font size is tiny
15:08:29pamauryhow often is button_read_device() called ?
15:09:03DMJCdo I need .rockbox on an android device or is the rockbox folder used instead?
15:11:35 Join stripwax [0] (
15:12:59 Quit stripwax (Client Quit)
15:14:12 Join linuxstb [0] (~linuxstb@rockbox/developer/linuxstb)
15:19:02DMJCwhere are fonts saved?
15:21:28Lalufupamaury: once every tick as far as I can make out.
15:21:48***Saving seen data "./dancer.seen"
15:22:59wodzpamaury: it is called from tick task
15:24:08 Join antil33t [0] (
15:24:54DMJCmy fonts are too small
15:25:00DMJCwhere should fonts be stored?
15:25:02 Quit antil33t| (Ping timeout: 252 seconds)
15:25:58Lalufupamaury: button_tick()->button_read()->button_read_device() (all in firmware/drivers/button.c)
15:27:42pamauryHum, writing a good touchpad driver for the fuze+ it definitely not trivial
15:28:11bluebrotherDMJC: the default cabbiev2 is quite usable. You can also try the one in FS #12254
15:28:12fs-bluebot RaaA cabbiev2 [Full Touch Application Port] (patches, new)
15:28:51bluebrotherRockbox needs to be able to access the SD card for it to work properly btw −− that might be the reason why your font is "tiny"
15:31:44 Quit hilbert (Ping timeout: 252 seconds)
15:33:57 Quit Keripo (Quit: Leaving.)
15:40:31 Join RansomTime [0] (~RT@wikia/vstf/Randomtime)
15:41:07RansomTimeWould it be possible to change the location log to the root directory of the microSD card on SanDisk players?
15:41:11RansomTimeof the*
15:45:39 Join Topy [0] (
15:49:03 Quit T44 (Ping timeout: 248 seconds)
15:50:33DMJCI have 10gb internal storage and 16gb SDcard
15:50:42DMJCshould the files be on the SD in .rockbox?
15:52:19 Join hilbert [0] (
15:57:20 Join benedikt93 [0] (~benedikt9@unaffiliated/benedikt93)
15:57:33bluebrotherDMJC: what files?
15:59:54 Join dfkt|n [0] (
16:00:00 Quit dfkt|n (Changing host)
16:00:00 Join dfkt|n [0] (~dfkt@unaffiliated/dfkt)
16:08:25pamauryHum, this means I can't yield in a button_read_device()
16:11:12 Quit hilbert (Quit: Computer has gone to sleep.)
16:11:24DMJCfonts/themes etc
16:11:31DMJCcabbiev2 looks very good
16:12:29bluebrotherDMJC: you store data on the sd card. Rockbox itself has its internally stored defaults (like cabbiev2) that is stored internally (unless you use some custom build that has this changed)
16:13:32DMJCall I've done is checkout svn and done tools/configure 201 960x540 and then make, make zip, make apk, adb install rockbox.apk
16:15:05pamauryDoes any see how I can avoid creating a thread to handle the f+ touchpad since I need to do some i2c requests and I can't to it in an interrupt handler :-/
16:15:08bluebrotherwell, since there is no theme available for that resolution the sizes are likely to not match at all.
16:15:11pamaury*does anyone
16:15:33DMJCso I should tell it to use 800x400?
16:15:44DMJC480x800 sorry
16:18:36bluebrothernot necessarily.
16:19:19DMJCthat worked perfectly
16:19:21bluebrotherbut IIRC the not matching cabbiev2 for 480x800 won't be put in the apk, so you only get the fallback theme −− which uses sysfont, which in turn _is_ tiny on a display of that resolution
16:20:10bluebrotheryou can use the correct resolution and put a somewhat matching version of cabbie on the sd card yourself
16:20:12DMJC480x800 fixes the theme display issue
16:20:18DMJCyeah I will do that in time
16:20:34bluebrotheror use another theme of a similar resolution. You can use any theme that has a smaller resolution.
16:21:59DMJCso how do I access my sdcard?
16:22:35bluebrotherby connecting the phone to the PC and setting it up to share the card as drive?
16:22:50 Join powell14ski_ [0] (
16:22:53bertrikpamaury, yup, it runs in interrupt context even, you can hardly do anything ...
16:22:57bluebrotheror use adb
16:23:01DMJCmy phone already mounts the sdcard
16:23:40DMJCnot quite sure where/how but it does it because the phone has built in media software that accesses it
16:24:28DMJChmm /mnt/sdcard-ext
16:25:47DMJCis there a way to set the default path?
16:26:37bluebrothersdcard-ext is a different path (i.e. different partition on the card)
16:26:46DMJCyeah I know
16:26:57DMJCis there a way to set sdcard-ext as the default path?
16:27:08bluebrotherAndroid can export the sd card (i.e. /sdcard) as storage device, which (obviously) makes it unavailable to the phone when connected
16:27:16 Join antil33t| [0] (
16:27:23bluebrotheryou can change the Rockbox code to look in sdcard-ext
16:27:36DMJCah, so there's no configure setting for it
16:27:41bluebrotherbut apps usually assume the sdcard to be /sdcard
16:27:52bluebrothersince that's the paths of the sdcard on Android.
16:28:00bluebrothersdcard-ext is something different
16:28:12 Quit antil33t (Ping timeout: 252 seconds)
16:28:47DMJCI think they've done my phone differently
16:28:51DMJCit's a Motorola Atrix
16:29:07DMJClooks like /sdcard is internal phone storage and /sdcard-ext is the external 16gb card
16:29:41bluebrotherno idea about that phone
16:30:07CIA-14New commit by bluebrother (r30633): Rockbox Utility: listen to translation change events. ...
16:30:14DMJCyeah I just proved it
16:31:37DMJChmm I can see what I want to change in rockbox, but it's a brilliant start
16:31:43 Quit zchs (Ping timeout: 248 seconds)
16:31:43CIA-14r30633 build result: All green
16:32:14 Quit powell14ski_ (Quit: powell14ski_)
16:36:39CIA-14New commit by bluebrother (r30634): Update cache information after translation change. ...
16:37:41 Join antil33t [0] (
16:38:31CIA-14r30634 build result: All green
16:38:50 Quit antil33t| (Ping timeout: 256 seconds)
16:41:51 Join GermanMushroom [0] (
16:43:38 Quit JdGord (Quit: Bye)
16:44:47CIA-14New commit by pamaury (r30635): fuze+: squeeze keymap to bare minimum
16:44:57CIA-14New commit by pamaury (r30636): fuze+: set backlight brightness to default only when there is not backlight infrastructure
16:45:02CIA-14New commit by pamaury (r30637): fuze+: implement basic touchpad driver
16:46:55CIA-14r30635 build result: All green
16:47:37CIA-14r30637 build result: All green
16:49:44pamauryhuh ?! I get a "ld: error: no memory region specified for loadable section `.icode'" but I have defined a region for it !
16:50:04bluebrotherpress save in the editor?
16:50:34pamauryhum, I guess codec use a custom linker script ?
16:50:43bluebrothercompiling old code but watching at new one in the editor happened to me this weekend. Gives funny experiences ;-)
16:51:12wodzpamaury: yes codecs use its own script
16:51:22pamaury:) but no I haven't changed the linker script in a while and it only happen for codecs, where is it ?
16:51:55pamaurywodz: where is it ?
16:52:05n1spamaury: apps/plugins/ i think
16:52:38pamauryurg, one for all targets :-/
16:54:30n1syeah, a nice ifdef mess
16:55:09pamaurywhat is DRAMORIG supposed to be ? the base of the DRAM and it does so magic to put it in the correct place in memory ?
16:55:09n1sSomeone should clean that up
16:56:02n1syeah, dram orig is the base of dram dunno if there's any magic going on
16:57:46 Quit DMJC (Quit: This computer has gone to sleep)
16:58:01pamauryhum it doesn't seem to be happy and complain about undefined symbols like iramstart :-/
16:59:03 Join hilbert [0] (
17:00:45*bluebrother likes the new language change feature in Rockbox Utility :)
17:00:53bluebrothershould have finished that much earlier ...
17:01:59wodzThis lcdif block on rk27xx is so f****** complicated with highly unclear documentation. The dma engine supposed to work in pair with lcdif is even more complicated. And because this all mess with display the work on this is hit-or-miss
17:02:23pamaurybluebrother: what is change about ?
17:02:39wodzAfter 1+ hour of reading manual I am still unclear how to enable interrupt from this dma to see if the transfer actually happen
17:03:32pamaurywodz: did you look at the rk27xx code you have ?
17:04:33wodzpamaury: sure - took me two days to produce call graph of display related functions
17:04:55 Join zchs [0] (
17:05:10bluebrotherpamaury: no need to restart Rockbox Utility anymore after changing its language
17:10:26pamauryI don't get it, if I want to use some iram for codecs, I need to know how much the firmware uses ?
17:11:52wodzisn't there static split between core and codecs?
17:12:39 Join Buschel [0] (
17:12:54pamauryanyway something is broken, codec_crt0.S complains about missing symbol, another ifdef hell consequence
17:13:30pamauryand I need to determine how much iram I give to the core :-/
17:14:06bertrikgood luck
17:14:38wodzwow - first time I see *some* pixels from what I am transferring
17:14:50BuscheljhMikeS: is my last patch in FS #12303 the obvious and correct fix? or is there still something unclear?
17:14:51fs-bluebot iPod Video USB broken since r30546 (bugs, new)
17:16:02pamauryproblem is: the core uses all the iram (31 out of 32 Kb), and if I set the plugin iram size to 0 then I get some error :(
17:17:26wodzpamaury: look how it is done for rk27xx - I claim no iram at all
17:17:43pamaurybut it doesn't link here
17:19:21pamauryhum ok I found out, I will add more to the USE_IRAM #ifdef hell
17:21:51***Saving seen data "./dancer.seen"
17:22:12wodzwe should move it to the target tree probably
17:24:43pamaurycodec_crt0.S is suspicious: #if USE_IRAM ends with #endif /* PLUGIN_USE_IRAM */ ??
17:25:51pamauryit still doesn't work, don't know why
17:27:53CIA-14New commit by bluebrother (r30638): Rockbox Utility: handle layout direction on language changes as well.
17:28:07wodzpamaury: what do you had in mind asking this "<pamaury> wodz: did you know there were rk27xx available ?"
17:28:12 Quit user829385 (Ping timeout: 260 seconds)
17:28:38pamauryrk27xx emulators sorry
17:29:23CIA-14r30638 build result: All green
17:30:24wodzpamaury: it is mostly the same thing as our emulator
17:30:37wodzit doesn't emulate lowlevel stuff
17:31:05wodzit is crap by the way :-)
17:34:37 Quit Buschel (Quit: ChatZilla 0.9.87 [Firefox 7.0.1/20110928134238])
17:39:21CIA-14New commit by pamaury (r30639): fuze+: add chessclock keymap
17:39:26CIA-14New commit by pamaury (r30640): fuze+: add pluginlib keymap
17:39:32CIA-14New commit by pamaury (r30641): imx233/fuze+: don't use iram in plugins and in codecs
17:39:35CIA-14New commit by pamaury (r30642): fuze+: add cube keymap
17:39:39CIA-14New commit by pamaury (r30643): fuze+: add logo keymap
17:39:45CIA-14New commit by pamaury (r30644): fuze+: add mosaique keymap
17:39:48CIA-14New commit by pamaury (r30645): fuze+: add rockblox keymap
17:41:01CIA-14r30639 build result: 0 errors, 4 warnings (pamaury committed)
17:41:04pamaurymore to come :)
17:41:33wodzand that was you who complained about my rk27xx related commit flood :P
17:41:34pamauryhum, isn't the buildserver doing bad things ?
17:42:31n1swhat bad things?
17:42:42wodzI observed that with flood of commits it takes the latest one to rebuild
17:42:52CIA-14r30645 build result: 0 errors, 4 warnings (pamaury committed)
17:43:12pamauryand no, again the set but not used one :)
17:43:15n1syes, the cliebnts just svn up and build
17:43:33 Join keyb_gr [0] (
17:43:38n1sso it'll build the first and last rev of a set like this
17:44:16rasherIs that new? It used to wait for x seconds before starting a build round
17:44:47n1sno idea
17:44:57pamauryarg, I don't have enough buttons for some plugins !
17:46:19 Quit robin0800 (Quit: Leaving)
17:46:42 Join robin0800 [0] (
17:50:43 Quit robin0800 (Client Quit)
17:50:48CIA-14New commit by pamaury (r30646): fuze+: add snow keymap
17:50:52CIA-14New commit by pamaury (r30647): fuze+: add stopwatch keymap
17:50:56CIA-14New commit by pamaury (r30648): fuze+: add stat keymap
17:51:00CIA-14New commit by pamaury (r30649): fuze+: add flipit keymap
17:51:05CIA-14New commit by pamaury (r30650): fuze+: add lamp keymap
17:51:08 Join robin0800 [0] (
17:51:08CIA-14New commit by pamaury (r30651): fuze+: add mp3_encoder keymap
17:51:12CIA-14New commit by pamaury (r30652): fuze+: add battery_bench keymap
17:51:15CIA-14New commit by pamaury (r30653): fuze+: add rockblox1d keymap
17:52:35CIA-14r30646 build result: 0 errors, 4 warnings (pamaury committed)
17:54:31CIA-14r30653 build result: All green
17:54:50 Quit GermanMushroom (Read error: Connection reset by peer)
17:55:47keyb_grbertrik: Currently thinking about which settings to include in an "official" "audio-only" PMU patch for all AMSv2. PVDD1 and AVDD17 of course, DC/DC also makes sense. Not sure 'bout oscillator drive current.
17:57:09pamaurywhat does that control exactly ?
17:57:16pamaurywhat is the impact ?
17:59:05bertrikI think the drive strength and DC/DC speed probably have no impact on audio
17:59:39bertrikpamaury, these are various power management related settings that the OF does, but we don't all do in rockbox yet
18:01:07keyb_grpamaury: The preliminary patch as attached to FS #11664 changes several things at once: Codec PLL setup (we can leave that out), DAC and headphone amp voltages are up to 2.5 V (vs. 1.65 V), USB detection threshold, DC/DC frequency, 32k and 24M oscillator drive currents.
18:01:08fs-bluebot Workaround for usb random failures on amsv2. (patches, new)
18:01:53keyb_grI'm not sure of the effect of reduced drive current, /might/ increase jitter. Will whip out RMAA and check.
18:01:59pamauryyes I know but my question about the last two (DC/DC and oscillator driver). If they are unrelated to the audio they should be part of another patch no ?
18:02:29 Join T44 [0] (
18:03:35bertrikthe patch is basically already applied, except for the voltage stuff which is now under an #if 0
18:04:43CIA-14New commit by pamaury (r30654): fuze+: add brickmania keymap
18:04:46CIA-14New commit by pamaury (r30655): fuze+: add calendar keymap
18:04:52CIA-14New commit by pamaury (r30656): fuze+: add blackjack keymap (unsure about resume)
18:04:56CIA-14New commit by pamaury (r30657): fuze+: add spacerocks keymap
18:04:58CIA-14New commit by pamaury (r30658): fuze+: add xobox keymap
18:05:05bertrikthe two things that might have some impact on USB is 1) lowering of the USB detection threshold (applied already) and possibly 2) increase of the internal voltages
18:05:06CIA-14New commit by pamaury (r30659): fuze+: add bounce keymap
18:05:10pamaurybertrik: I would say that if the OF does it it's pretty safe to do it no ?
18:05:11CIA-14New commit by pamaury (r30660): fuze+: add calculator keymap
18:05:14CIA-14New commit by pamaury (r30661): fuze+: fix calculator keymap
18:05:17CIA-14New commit by pamaury (r30662): fuze+: add chip8 keymap
18:05:18bertrikpamaury, yes
18:05:18keyb_grIMO, DC/DC setup is "nice to have and doesn't do any harm". It works fine at 1 MHz vs. the default of 2, and should reduce power consumption a bit.
18:05:21CIA-14New commit by pamaury (r30663): fuze+: add chopper keymap
18:05:24CIA-14New commit by pamaury (r30664): fuze+: add jewels keymap
18:05:37 Quit Topy (Ping timeout: 260 seconds)
18:06:04wodzkeyb_gr: 1MHz of what? you mean converter frequency?
18:06:22keyb_grwodz: yep
18:06:25CIA-14r30654 build result: All green
18:06:59wodzkeyb_gr: lowering freq may have impact on converter efficiency
18:07:27 Join Mineo [0] (~wh@2001:638:904:ffca:4261:86ff:fe87:5544)
18:08:00keyb_grYeah, usually losses increase with higher freq. Lower limit is usually set by how much filtering you can afford.
18:08:15CIA-14r30664 build result: All green
18:08:23wodzkeyb_gr: but lower freq means you need bigger coil
18:08:44pamaurydamn, some many plugin use similar keymap, can't they use a standard one ??
18:08:48pamaury*so many
18:09:19bertrikpamaury, yes, it seems to be silly that some plugins have more code for the keymaps than for the actual plugin itself
18:10:16keyb_grwodz: I'll trust the OF developers if they say 1 MHz should be fine.
18:10:21pamauryand isn't there some standard keymap in pluginlib that they can use
18:10:49wodzthere is
18:10:55 Quit zchs (Ping timeout: 248 seconds)
18:11:07 Quit liar (Remote host closed the connection)
18:11:11keyb_grPlugin keymaps are annoying indeed. Seemingly almost always nonstandard.
18:11:30wodzkeyb_gr: ah yes I missed that bit - if they designed converter for 1MHz we should use this
18:11:31 Quit robin0800 (Quit: Leaving)
18:11:59wodzI would love to be able to build ONLY test_* plugins
18:12:23pamauryI mean lots of keymap have standard left,right,up,down,volume up,volume down,select; that would save *lots* of define if they use some standard one
18:12:29n1spamaury: it was tried using a standard keymap for simple plugins, turns out it didn't work well
18:12:56bertrikperhaps it wasn't tried hard enough :)
18:13:43 Join zchs [0] (
18:14:18bertrikonce I get the sansa clip zip working on a basic level, I think I'll leave the keymaps to someone else :P
18:14:43bertrik(though clip zip can basically steal/reuse the clipv1 keymaps)
18:17:13bertrikI'm still not completely sure what to do about the internal voltage for AMSv2. On one hand I'd like to commit it on the basis that it's what the OF does (and it apparently solves clicks/pops), on the other hand we might reconsider that if the runtime impact is too big (but what is too big?)
18:18:10bertrik< 5%?
18:19:01wodzbertrik: and what IS runtime impact?
18:20:04bertrikwodz, I don't know, too lazy to test, I use my clip+ daily and don't want to wait for a battery bench
18:20:48CIA-14New commit by pamaury (r30665): fuze+: add oscilloscope keymap
18:20:51CIA-14New commit by pamaury (r30666): fuze+: add pegbox keymap
18:20:55CIA-14New commit by pamaury (r30667): fuze+: add minesweeper keymap
18:20:58CIA-14New commit by pamaury (r30668): fuze+: add pong keymap
18:21:01CIA-14New commit by pamaury (r30669): fuze+: add snake keymap
18:21:06CIA-14New commit by pamaury (r30670): fuze+: add snake2 keymap
18:21:09CIA-14New commit by pamaury (r30671): fuze+: add solitaire keymap (unsure about it)
18:21:11pamaurybertrik: I can battery bench if you want
18:21:12CIA-14New commit by pamaury (r30672): fuze+: add sliding_puzzle keymap
18:21:16CIA-14New commit by pamaury (r30673): fuze+: fix solitaire keymap
18:22:30CIA-14r30665 build result: All green
18:23:14bertrikpamaury, please do, the thing to change is line 284 in firmware/target/arm/as3525/system-as3525.c
18:23:34pamaurybertrik: shouldn't I first battery bench with svn and then with the change ?
18:23:54bertrikwhatever order you prefer
18:24:14CIA-14r30673 build result: All green
18:24:16pamauryis the bootloader important ? I don't think I have the latest version
18:25:13bertrikpamaury, no, the bootloader is not important, this stuff is re-inited in the main rockbox image too
18:30:29pamaurybench started
18:38:02bertrikyou know a single bench might take up to 17h? :P
18:38:46bertrikand of course you need to make sure that the battery is completely full before starting a bench
18:40:14bertrikIf there is a really big difference between runs, you might be able to see it on a shorter bench, just by looking at the slope of the voltage
18:42:38 Join Topy [0] (
18:43:22wodzI am out of ideas how rk27xx maps lcdif sram I know which registers change this but I can't figure out how it is done
18:44:39pamaurybertrik: yes, I hardly use my clip+ these times and it happens that I fully charged it two days ago so all conditions are met. I doubt the difference will be sufficient enough to see it on the slope :)
18:45:03 Quit T44 (Ping timeout: 248 seconds)
18:51:18CIA-14New commit by pamaury (r30674): fuze+: add star keymap
18:51:22CIA-14New commit by pamaury (r30675): fuze+: fix solitaire keymap again
18:51:25CIA-14New commit by pamaury (r30676): fuze+: add starfield jeymap
18:51:29CIA-14New commit by pamaury (r30677): fuze+: add sokoban keymap
18:51:34CIA-14New commit by pamaury (r30678): fuze+: add wormlet keymap
18:51:37CIA-14New commit by pamaury (r30679): fuze+: add vu_meter keymap
18:53:01CIA-14r30674 build result: 4 errors, 4 warnings (pamaury committed)
18:54:01pamauryhum, very strange error
18:54:10pamaurygcc: › : No such file or directory
18:54:23wodzclient problem?
18:54:44CIA-14r30679 build result: 0 errors, 4 warnings (pamaury committed)
18:59:46 Quit benedikt93 (Quit: Bye ;))
19:07:29pamaurywodz: how is your display code working right now ?
19:08:51wodzpamaury: I switch to bypass mode which means any write into lcdif sram address ends up on lcd internal bus
19:09:16wodzthe downside is that I have to convert pixels to native format which is truly weird
19:10:50pamaurythe display controller can do the conversion ?
19:11:14pamauryand what is your problem when trying to use dma mode ?
19:11:38wodzpamaury: It doesn't work as expected :-)
19:12:20pamauryis the data sent ?
19:12:49wodzat least partially since I can see some pixels on the screen
19:15:52 Join robin0800 [0] (
19:16:16wodzdma engine uses linked list to setup transfer so I thought it would update whole screen at one shot but I get a few lines at best
19:17:09pamaurythe lcd controller probably only has a limited buffer and doesn't sustain the dma burst
19:17:37bertrikthat would be one sorry lcd controller then
19:18:25wodzbut it is mem2peri transfer so I guess lcdif can hold transfer if needed
19:18:27pamauryhey that's a rockchip one !
19:18:53bertrikI had a lot of trouble with lcd controller in the s5l8700 (in the samsung yp-s3), I could enable it and actually get a rockbox logo on the screen, but the CPU would hang when the lcd controller was enabled
19:19:02bertrikI gave up on that one
19:19:52wodzbertrik: and how do you drive it?
19:21:03bertriksetup the usual display stuff then setup DMA and start DMA
19:21:54***Saving seen data "./dancer.seen"
19:22:18wodzI mean how you drive it now
19:29:51 Join FoH [0] (
19:30:47 Quit robin0800 (Ping timeout: 245 seconds)
19:31:52 Join robin0800 [0] (
19:34:08CIA-14New commit by pamaury (r30680): fuze+: add clix keymap
19:34:14CIA-14New commit by pamaury (r30681): fuze+: add rockpaint keymap
19:34:15pamaury\o/ finished keymaps
19:34:17CIA-14New commit by pamaury (r30682): fuze+: add fireworks keymap
19:34:21CIA-14New commit by pamaury (r30683): fuze+: add matrix keymap
19:34:24CIA-14New commit by pamaury (r30684): fuze+: add invadrox keymap
19:34:33CIA-14New commit by pamaury (r30685): fuze+: add superdom keymap
19:34:37CIA-14New commit by pamaury (r30686): fuze+: add text viewer keymap
19:34:41CIA-14New commit by pamaury (r30687): fuze+: add rockboy keymap
19:34:45CIA-14New commit by pamaury (r30688): fuze+: add chessbox keymap
19:34:49CIA-14New commit by pamaury (r30689): fuze+: add fft keymap
19:34:53CIA-14New commit by pamaury (r30690): fuze+: add fractal keymap
19:34:56CIA-14New commit by pamaury (r30691): fuze+: add imageviewer keymap
19:35:02CIA-14New commit by pamaury (r30692): fuze+: add sudoky keymap
19:35:05CIA-14New commit by pamaury (r30693): fuze+: add reverso keymap
19:35:12CIA-14New commit by pamaury (r30694): fuze+: add goban keymap
19:35:15CIA-14New commit by pamaury (r30695): fuze+: add zxbox keymap
19:35:19CIA-14New commit by pamaury (r30696): fuze+: add doom and pacbox keymap
19:35:22CIA-14New commit by pamaury (r30697): fuze+: add midiplay keymap
19:35:26CIA-14New commit by pamaury (r30698): fuze+°: add mpeg keymap
19:35:31bertrikwodz, I don't, I abandoned the effort
19:35:54CIA-14r30680 build result: 0 errors, 4 warnings (pamaury committed)
19:36:59 Join [Saint_AndChat] [0] (~Saint]@
19:37:39CIA-14r30698 build result: All green
19:38:19 Join AHadley [0] (
19:38:53wodzpamaury: this rknano looks like stripped down version of rk27xx
19:39:58 Quit AHadley (Client Quit)
19:40:40pamauryyes, it basically has the bare minimum amount of controllers, a low freq cpu and hardware decoding I think
19:41:05wodzdecoding of what?
19:41:34wodzI don't think so
19:42:13wodzits impossible - I've never seen ogg decoding in hardware
19:42:31pamauryit doesn't support ogg
19:43:01pamauryhum, you might be right
19:43:15 Join T44 [0] (
19:43:30pamauryanyway, the question is whether is has the same registers as, say, the rk27xx or a completely different one
19:43:38 Quit Topy (Read error: Connection reset by peer)
19:43:40wodzI bet this is stripped down ARM7ejs based SoC
19:43:56pamauryactually I was amazed by how simple the hardware is, did you look at the photo ?
19:45:27wodzPeripherials are different - lcdif, pmu (rk27xx doesn't have something which one may call pmu)
19:46:12wodzfor decoding audio armv5te @ 120MHz is a lot
19:46:32pamauryUnfortunately I have no datasheet, no sdk, nothing. I found a schematic diagram and a set of windows tool; one should be able to read from the flash but it doesn't work; I have a firmware upgrade for another rknano based player
19:47:46wodzyou may try to short nand address pins during startup - this should produce a fault and through you into rom code
19:48:07wodzif you are lucky they didn't changed routines
19:48:50pamaurywhat is rom code ?
19:49:31wodzall rk chips I am aware of (which excludes rknano) have masked rom.
19:50:02wodzIn this rom are routines to boot from nand and failsafe mode (sort of dfu)
19:50:49pamauryI already found the dfu mode but I don't know what to do with it
19:51:17 Quit amiconn (Quit: No Ping reply in 64 seconds.)
19:51:26 Join amiconn_ [0] (quassel@rockbox/developer/amiconn)
19:51:45wodzYou mean firmware upgrade from menu right? Read answer from mortalis in rk27xx thread on the forum
19:51:48 Nick amiconn_ is now known as amiconn (quassel@rockbox/developer/amiconn)
19:52:14wodzthis may be two different things
19:52:55pamaurydfu from menu and from booting with "M" hold; it's the same here
19:53:10pamaury(I think)
19:54:13wodztry flash fault method - this is the only way to be sure you are not messing with higher level code
19:55:12wodzMy device for example had 'dfu' mode serviced by nand bootloader which was slightly different then rom one.
19:55:16 Quit factor (Read error: Connection reset by peer)
19:55:29wodzboth modes presented the same VID/PID
19:56:26pamaurywhich pin should I short ?
19:57:11wodzsome adress lines (preferably to GND)
19:58:20 Join Topy [0] (
19:59:19pamauryI'll try and then what can I do in this mode ?
20:01:17wodzif it works as in rk27xx you can load arbitrary code to the device
20:01:19 Quit T44 (Ping timeout: 248 seconds)
20:02:07pamaurybut since the registers probably don't match :-/
20:03:26wodzin which form is this firmware upgrade you mentioned?
20:04:19pamaurya binary file but it's not for my player :)
20:04:41wodzthe hardware is indeed super simple I wonder whats on the second side of the PCB
20:05:01wodzpamaury: binary as .bin or binary as unknown format?
20:11:31 Quit stoffel (Ping timeout: 260 seconds)
20:11:55wodzok this sot-25 IC near battery terminals is most probably EMC5754 lithium-ion battery charger
20:13:03 Join factor [0] (~factor@
20:13:59pamaurywodz: it's a bin file, I don't know the format but I have a tool which can read it so it's a structured one
20:14:10pamauryI can send it to you it you want
20:15:53wodzunfortunately I'll be extremely short of time in incoming weeks
20:29:41pamauryyeah, I you'll better spend your time on rk27xx rather than this :)
20:32:22wodzGood god - mortali's DAP is They used 20+ old tda1543 as DAC (at least that is what this site states).
20:33:21wodzthis would mean one have to write fake codec driver and process everything through our dsp.
20:34:27keyb_grbertrik: RMAA on Clip+ done. Result = boring. With PMU patch (no PLL), it's about 3 dB louder and has a bit more distortion as a result. And that's it.
20:36:10 Join Zagor [242] (~bjst@rockbox/developer/Zagor)
20:45:09 Quit Bagder (Ping timeout: 244 seconds)
20:45:27 Quit robin0800 (Ping timeout: 260 seconds)
20:45:28 Quit dfkt|n ()
20:47:27 Join Bagder [241] (~daniel@rockbox/developer/bagder)
20:52:54 Join antil33t| [0] (
20:53:27 Join T44 [0] (
20:54:05 Quit antil33t (Ping timeout: 258 seconds)
20:56:41pamauryhum, when do plugin print "incompatible model" when tried run ?
20:56:47 Quit Topy (Ping timeout: 248 seconds)
21:02:52bertrikI guess it's based on the plugin api version
21:03:13pamauryBut I've recompiled everything
21:03:28bertrikIt happens to me sometimes when I try to run an old test plugin
21:04:22bertrikMaybe it was still on the player and isn't normally compiled
21:04:35pamauryno that's the first time I compile the plugins for the f+
21:05:21bertrikI'd look in the source code for the "incompatible model" text and reasons for showing it
21:05:38pamauryI found it and it doesn't make any sense that it fail /-
21:11:19pamauryI give up for today, I can't find the problem
21:21:56***Saving seen data "./dancer.seen"
21:22:14 Join Topy [0] (
21:25:03 Quit T44 (Ping timeout: 248 seconds)
21:28:35 Quit ReimuHakurei (Remote host closed the connection)
21:30:15 Join ReimuHakurei [0] (
21:32:10 Quit keyb_gr (Quit: ChatZilla 0.9.87 [SeaMonkey 2.4.1/20110928161145])
21:44:44 Join robin0800 [0] (~quassel@
21:49:59 Quit FoH (Ping timeout: 260 seconds)
21:50:39 Quit n1s (Ping timeout: 248 seconds)
21:50:43 Join T44 [0] (
21:53:51 Quit Topy (Ping timeout: 248 seconds)
21:58:02 Join thomasjfox [0] (~thomasjfo@rockbox/developer/thomasjfox)
22:04:46CIA-14New commit by thomasjfox (r30699): Fix file descriptor double close(): ...
22:06:24CIA-14r30699 build result: All green
22:14:07 Quit thomasjfox (Remote host closed the connection)
22:18:23 Join Strife89 [0] (~Strife89@
22:21:34 Quit robin0800 (Ping timeout: 276 seconds)
22:22:17 Join n1s [0] (~quassel@rockbox/developer/n1s)
22:22:47n1spamaury: it's when the model id in the plugin header doesn't match with what the loader expects
22:23:48pamauryn1s: no, I found it out, it was because of a loading address mismatch
22:24:05pamauryare there some numbers of test_disk to compare targets ?
22:24:18n1syeah, the wiki has some
22:26:15 Join robin0800 [0] (~robin0800@
22:27:53pamaurydoes test_disk writes the output to a file ?
22:29:07pamaurythe wiki is not quite up to date...
22:31:27n1syeah, it creates a dir in which it does everything
22:32:14 Join Topy [0] (
22:32:18n1sin fact no, it creates /__TEST__ and /test_disk.tmp
22:33:16CIA-14New commit by pamaury (r30700): imx233/fuze+: enable plugins
22:34:00 Nick soap is now known as soap_ (~soap@rockbox/staff/soap)
22:35:03CIA-14r30700 build result: All green
22:35:07 Quit T44 (Ping timeout: 260 seconds)
22:35:45pamaurythe fuze+ port is hopefully getting closer to a completely functional DAP
22:36:08n1spamaury: congratulations
22:36:27pamauryonly the audio driver missing plus some quirks here and there
22:37:09[Saint_AndChat]So its just a DP for now? No A? ;)
22:37:12n1swho cares about audio?
22:37:37n1sdoom works, right?
22:37:39[Saint_AndChat]Doom is all that's needed, clearly.
22:37:49[Saint_AndChat]Heh, snap.
22:38:01pamauryit's a MDAP (Mute DAP)
22:44:44 Quit n1s (Remote host closed the connection)
23:01:17 Quit robin0800 (Remote host closed the connection)
23:04:15 Quit ReimuHakurei (Remote host closed the connection)
23:12:33 Part Zagor
23:12:39 Quit crwll (Ping timeout: 256 seconds)
23:13:19 Join crwll [0] (
23:14:52 Join robin0800 [0] (
23:21:59***Saving seen data "./dancer.seen"
23:26:02 Quit froggyman (Ping timeout: 255 seconds)
23:27:54 Quit wodz (Quit: Leaving)
23:29:58 Join ReimuHakurei [0] (
23:32:22 Quit bertrik (Quit: And That, My Liege, Is How We Know the Earth to Be Banana Shaped)
23:41:38 Join dfkt [0] (~dfkt@unaffiliated/dfkt)
23:43:41 Quit dfkt (Read error: Connection reset by peer)
23:47:26 Join dfkt [0] (~dfkt@unaffiliated/dfkt)
23:50:37 Quit dfkt (Client Quit)

Previous day | Next day