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 2011-12-14

00:00:53funmanTheSeven: can you etst r31220 ?
00:01:04kugellorenzo92: why that?
00:01:17lorenzo92to be able to run rockbox...
00:01:25kugelI'm sure you can script the generation of it based on a stock rom?
00:04:46lorenzo92of course, but the stock rom has no scriptable starter or what from the user memory. And the modded firware has the safe mode, that's very important since the OF has a terrible bug. Without it there are chances to brick device
00:05:48saratogaare their proprietary bits in the stock rom though? those might be tricky to redistribute
00:06:08saratogaalthough i'm not sure how that works for linux distros
00:07:08kugellorenzo92: we can add all that to svn+plus a script to create the rom, just not the rom itself
00:09:38lorenzo92kugel: uhm okay could be fine, but I remind you that if a rom fails, device is bricked ;)
00:10:11lorenzo92kugel: anyway we will discuss that better
00:10:29lorenzo92now I check rockbox some minutes then I close ;)
00:12:23CIA-1New commit by funman (r31235): Sansa Clip: simplify matrix key scan ...
00:14:03CIA-1r31235 build result: All green
00:14:22kugellorenzo92: that risk is a given :P
00:15:00lorenzo92but if you want to test some particular roms, just ask me ;)
00:16:56 Quit lorenzo92 (Quit: ChatZilla 0.9.87 [Firefox 8.0/20111115183813])
00:20:00funmanhm my fuzev2 shows the drive but they have no medium inside
00:20:40CIA-1New commit by bluebrother (r31236): Remove duplicate entry in russian translation.
00:21:53funmanit's dying though, not sure i should rely on it
00:22:04CIA-1r31236 build result: All green
00:22:39 Join Thra11 [0] (~thrall@
00:25:02 Join Rob2222 [0] (
00:26:37 Quit ChickeNES (Quit: Computer has gone to sleep.)
00:35:33 Join Topy44 [0] (
00:35:51 Quit liar (Ping timeout: 245 seconds)
00:54:26 Quit bertrik (Ping timeout: 244 seconds)
00:56:52 Quit lebellium (Quit: ChatZilla 0.9.87 [Firefox 9.0/20111206234556])
00:59:25 Quit kadoban_ (Ping timeout: 255 seconds)
01:16:18 Join liar [0] (
01:23:21 Join Zarggg [0] (
01:26:17***Saving seen data "./dancer.seen"
01:26:21 Quit rarog (Quit: ChatZilla 0.9.87 [SeaMonkey 2.4.1/20111011102430])
01:27:25 Quit Rob2222 (Quit: Rob2222)
01:29:31 Join nosa [0] (
01:29:52 Quit nosa-j (Ping timeout: 268 seconds)
01:30:17CIA-1New commit by funman (r31237): usb-s3c6400: fix more index / endpoint number confusion
01:30:20 Join kadoban_ [0] (
01:30:20funmanTheSeven: current build might be fixed for nano2g
01:30:20 Nick nosa is now known as nosa-j (
01:31:59CIA-1r31237 build result: All green
01:33:05 Quit captainkwel (Read error: Connection reset by peer)
01:34:46 Join Rob2222 [0] (
01:37:49 Quit liar (Remote host closed the connection)
01:37:56funmanJdGordon: did you see FS #12370 ?
01:37:57fs-bluebot RDS support for Si4701/Si4703 (Sansa Clip Zip proof-of-concept) (patches, new)
01:43:30funmanbertrik: anything against making Zip stable? I use it everyday without problems
01:45:38 Join CaptainKewl [0] (
01:57:41 Quit kadoban_ (Read error: Operation timed out)
02:01:05 Quit keyb_gr (Ping timeout: 244 seconds)
02:07:35JdGordonfunman: no, is there anything I should look at?
02:07:39JdGordonremind me tonight :)
02:14:02 Join captainkwel [0] (
02:14:37 Quit CaptainKewl (Ping timeout: 252 seconds)
02:16:21 Join lieutenantmeow [0] (
02:17:15lieutenantmeowhello, i have just installed rockbox on my sansa + but the device has not loaded up to its proper theme, what would i do from this point on?
02:27:59 Quit Thra11 (Ping timeout: 276 seconds)
02:32:10 Quit maal (Quit: CGI:IRC)
02:57:12funmanJdGordon: yes the last messages talk about skins probklems preventing inclusion
02:58:37*JdGordon lols at the last comment :)
02:59:41JdGordonI dont understand why hat is a blocker?
02:59:46JdGordoncommit and open a bug?
03:00:07funmanJdGordon: no clue
03:00:17funmani pinged you because of last comment :)
03:01:05JdGordonask zagor to buy me a clip zip and I'll fix it on arrival :)
03:01:32funmanJdGordon: you can ask him i think
03:01:51JdGordon(or fix the rds simulator code)
03:03:43funmanusb in bootloader is broken in amsv1 too
03:07:08JdGordonfunman: wanna try a quick fix?
03:07:11JdGordonif you've got the target
03:07:26 Quit Strife89 (Quit: Vamoose!)
03:10:55 Quit lieutenantmeow (Quit: Lost terminal)
03:12:02JdGordonremove %s from line 57,58 in apps/radio/radio_skin.c
03:15:40funmando i need the patch applied?
03:16:05JdGordondont' you have a clip zip?
03:16:14JdGordonor was that just wondering why it was blocked?
03:17:28funmanyes i have a clip zip
03:17:50JdGordonso yeah, apply the rds patch and that 2 line fix and see if it works :)
03:17:51funmani didn't try this patch though, not sure what's the status of rds in current build.
03:18:18JdGordonit almost certainly will work, but it will make those lines not scroll
03:18:35JdGordonsomeone needs to look into making scrolling lines be able ot have their text changes
03:25:28funmani'm not sure where i should be able to see the RDS stuff
03:25:52JdGordonhow many lines fit on the radio screen?
03:26:01funmani see 5
03:26:06funman1/ station name
03:26:07funman2/ freq
03:26:10funman3/ mono/stereo
03:26:15funman4/ mode
03:26:21***Saving seen data "./dancer.seen"
03:26:21funman5/ signal strength? (it's cut)
03:26:24funman6/ search bar
03:26:40JdGordonshould be under that
03:26:44funmanunder the bar?
03:26:58JdGordonwith a blank line between the bar and rds info
03:27:11JdGordonis HAVE_RDS_CAPS defined?
03:27:55funmanyes in zip config.h
03:28:08funmanperhaps my radios just have no rds data
03:28:18funmanah i see it!
03:28:36funmana single word "PLANETE", the name of the radio, where you said it should be
03:28:51funmanit appears a bit after tuning
03:28:54JdGordonand of course it wont scroll
03:29:05JdGordontry to find a station which updates the info
03:29:07funmanprobably but then there are other non scrolling lines
03:33:15funmani only see "RADIO X" "PLANETE" and "LE FM 98", the title for 3 channels, but they don't update
03:35:05 Quit Topy44 (Ping timeout: 240 seconds)
03:38:25 Join Topy44 [0] (
03:46:30 Join nosa [0] (
03:49:52 Quit nosa-j (Ping timeout: 248 seconds)
03:49:52 Nick nosa is now known as nosa-j (
03:51:50 Quit Zarggg (Ping timeout: 276 seconds)
04:09:16 Quit captainkwel (Quit: ( :: NoNameScript 4.22 :: ))
04:09:38funmanJdGordon: suggested changes for settings_parseline(): add \ to escape #
04:09:54funmani want to add ##MUSIC# and ##PORT# to disktidy.config but they got ignored
04:10:05funmanso i suggest \##MUSIC# \##PORT#
04:12:26funmanand \\ as first character would give '\'
04:12:26funmani could do that n disktidy.c but it's better to be in generic code
04:13:21 Join Keripo [0] (
04:17:34 Join JdGord [0] (
04:17:58JdGordFunman: ?
04:18:00JdGordWhen is that needed?
04:18:41JdGord# is only checked at the beggining of the line isn't it?
04:19:52funmannope, after whitespace
04:20:22funmani just added if (*str == '\') str++; in disktidy.c
04:29:11 Quit TheSeven (Disconnected by services)
04:29:30 Join [7] [0] (~TheSeven@rockbox/developer/TheSeven)
04:35:13 Quit amiconn (Disconnected by services)
04:35:14 Join amiconn_ [0] (quassel@rockbox/developer/amiconn)
04:35:14 Quit pixelma (Disconnected by services)
04:35:17 Join pixelma_ [0] (quassel@rockbox/staff/pixelma)
04:35:19 Nick pixelma_ is now known as pixelma (quassel@rockbox/staff/pixelma)
04:35:36 Nick amiconn_ is now known as amiconn (quassel@rockbox/developer/amiconn)
04:40:29JdGordonfunman: remind me when im not busy at work :) I'm not underastanidng why its needed and cant look into it atm
04:40:50JdGordonis the leading # in the value or name return val?
04:42:46 Quit Topy44 (Ping timeout: 240 seconds)
04:43:35funmanlike this: " ##MUSIC#: no"
04:43:38 Join Topy44 [0] (
04:43:55JdGordonyeah, that shold be fine to add
04:44:04JdGordonI thought you wanted it in value which causes problems
04:44:14funmanyeah in value it should work already
04:44:31JdGordon# isnt checked for in value i tihnk
04:49:02funmanhm i made disktidy fully erase my disk :/
04:49:14JdGordonhaha well done!
04:49:30jhMikeStidiness win!
04:50:51JdGordonjhMikeS: hey,
04:50:58funmanwhat changes could have broken usb in ams bootloader?
04:51:03JdGordonthe stereo/mono line doesnt scroll so it is no issue to update it every redraw
04:51:21jhMikeSJdGordon/funman: hello
04:51:27JdGordonthe lcd_scroll function cant handle the scrolling text changing its text, so we only update every err... not often
04:51:32JdGordonotherwise the line could never scroll
04:51:39JdGordonso, how would you like to fix that? :D
04:51:51jhMikeSfunman: I'm trying to remember. I remember being concerned about it. :\
04:52:23JdGordonsee gui/skin_engine/skin_render.c around line 740
04:52:25jhMikeSJdGordon: fix scrollers to not reset if the text is the same? maybe just keep scrolling anyway :)
04:52:58JdGordonideally we should be able to change the text and have the scroller continue scrolling from where it was
04:53:09JdGordonas long as the new string is at least as long as the last one
04:54:13jhMikeSfunman: probably whenever I changed the bootloader, for whatever that was for. I just remember looking and going "I hope it still works."
04:54:33jhMikeSJdGordon: I suppose just hit the limit and reverse it in that case
04:57:00jhMikeSno, I suppose haven't altered the bootloader afterall
04:57:54funmanjhMikeS: i don't see your commits in bootloader/sansa_as3525.c
04:59:43jhMikeSright, didn't commit there but something just had me thinking about its welfare, but I don't remember exacty what project that was
05:00:43 Join kadoban_ [0] (
05:02:01funmanUSB_DETECT_BY_DRV ?.
05:02:23 Join Buschel [0] (
05:03:11 Quit Keripo (Read error: Connection reset by peer)
05:03:37 Quit mgue (Ping timeout: 240 seconds)
05:05:42 Join mgue [0] (
05:06:02jhMikeSfunman: quite possibly that one :)
05:09:33funmandoes usb bootloader still work on sa9200 ?
05:11:28 Join Keripo [0] (
05:11:38funmanjhMikeS: ah there is a HAVE_BOOTLOADER_USB_MODE
05:14:21funmanjhMikeS: probably when you changed PP bootloaders
05:19:21 Quit Keripo (Read error: Connection reset by peer)
05:21:39 Quit Buschel (Quit: ChatZilla 0.9.87 [Firefox 8.0/20111104165243])
05:21:57 Join Rob2223 [0] (
05:25:52 Quit Rob2222 (Ping timeout: 252 seconds)
05:26:24***Saving seen data "./dancer.seen"
05:30:03jhMikeSfunman: no idea about sa9200. I could check it.
05:33:11 Join Zarggg [0] (
05:33:34 Quit ps-auxw (Ping timeout: 252 seconds)
05:35:24 Join ps-auxw [0] (~arneb@2001:470:c807:0:1532:4e5f:2ad3:4123)
05:36:49jhMikeSyeah, bl usb mode works
05:38:08funmanok i'll try on clipv1
05:38:32CIA-1New commit by funman (r31238): disktidy: simplify ...
05:38:46CIA-1New commit by funman (r31239): disktidy.config: add a preset for Sansa AMS OF files
05:39:25mc2739funman: see FS #12384
05:39:27fs-bluebot AMSv1 bootloader USB mode does not attach since r30548 (bugs, new)
05:40:13CIA-1r31238 build result: 166 errors, 0 warnings (funman committed)
05:40:19funmanah thanks
05:41:02 Join nosa [0] (
05:41:40CIA-1r31239 build result: 166 errors, 0 warnings (funman committed)
05:42:04CIA-1New commit by funman (r31240): disktidy: don't use "abort" for variable name as it might conflict
05:42:10 Quit nosa-j (Ping timeout: 252 seconds)
05:42:10 Nick nosa is now known as nosa-j (
05:43:26 Quit Topy44 (Ping timeout: 240 seconds)
05:43:34CIA-1r31240 build result: All green
05:43:37 Join Keripo [0] (
05:45:41 Join Topy44 [0] (
05:52:05 Quit Keripo (Ping timeout: 248 seconds)
05:54:20funmanjhMikeS: what does HAVE_BOOTLOADER_USB_MOD do?
05:55:03jhMikeSit was something to deal with some complications on PP originally but it also creates a full usb.c
05:55:21funmanyeah but we do that already on AMS
05:56:23jhMikeSbut somehow it makes the bootloader bloat?
05:57:04jhMikeSit causes "current_tick" to actually exist
05:57:21funmanisn't that PP specific anyway?
05:57:38jhMikeSyeah, so I don't get it
05:57:56funman#define HAVE_PRIORITY_SCHEDULING
05:57:56funman#define USB_STATUS_BY_EVENT
05:57:56funman#define USB_DETECT_BY_CORE
05:58:00funmani think that's all it does
05:58:20funmanwell for ams: only HAVE_PRIORITY_SCHEDULING
05:58:21jhMikeSah you got to the .h files alread :0
05:58:32jhMikeSthat will add major code size
05:59:15funman(what is "that")
05:59:25jhMikeSthe priority scheduler
05:59:38funmanwell we don't need it i think, do we?
05:59:54funmanperhaps USB_DETECT_BY_CORE over USB_DETECT_BY_DRV adds more code
05:59:56jhMikeSprobably not in there but other targets yes
06:00:28jhMikeSI wouldn't expect that to be too expensive
06:00:54jhMikeSacutally it might be smaller
06:00:55funmansansa_as3525.c:usb_mode() looks ok?
06:02:28jhMikeSI think so
06:02:37funmanadding usb_enable(true/false) fix it
06:03:47funman(on clipv1, trying clip zip)
06:06:01jhMikeSshouldn't that have been done in the usb thread?
06:06:44funmani think we don't use the usb thread
06:07:01funmanlook at gigabeat.c loop
06:07:13CIA-1New commit by funman (r31241): Sansa AMS: fix usb in bootloader ...
06:07:24funmanclip zip works but: [11758.128551] usb 1-2: config 1 has no interfaces?
06:07:35jhMikeSI wonder if there's enough time for usb_status to get updated?
06:08:21jhMikeSno? who runs the driver?
06:08:34 Join Torino10 [0] (
06:08:53CIA-1r31241 build result: All green
06:09:33Torino10I'm having trouble getting my sansa Clip+ to boot to the original firmware
06:09:36funmanthat's equivalent to usb_core_init()
06:09:43funmanTorino10: press left key
06:10:10Torino10It starts to boot but it gets stuck
06:10:23funmanjhMikeS: it is probably interrupt driven?
06:10:55funmanTorino10: try to format it, anyway the OF is not on topic here, you should ask #rockbox-community or sandisk
06:13:59jhMikeSfunman: sure it is, but the thread is created for it
06:14:16funmanwhat does the thread do?
06:14:34 Join saratoga_ [0] (9803c31c@gateway/web/freenode/ip.
06:16:05funmanah i see
06:16:12funmanUSB_POWERED event
06:16:54funmanhm USB_STATUS_BY_EVENT is defined
06:18:24funmanascodec should be enabled and generate interrupts though
06:21:50Torino10Rockbox is working, but device does not show on usb, can I replace the firmware by copying the renamed patched.bin file from sd card to root?
06:21:51 Quit psychohistory (Read error: Connection reset by peer)
06:22:18saratoga_Torino10: you can, but you'd need to be able to boot at least a little ways into the sandisk software for it to take
06:23:24saratoga_what did you actually do to it?
06:23:45 Join BHSPitMonkey_ [0] (
06:25:48Torino10I think the patched bootloader is corrupt, it starts to boot to OF but repeats then gets stuck on graphics screen
06:27:02funmansaratoga_: i start seeing data aborts and panics on my zip in usb mode
06:27:17jhMikeSfunman: why would a force usb_enable be needed? I'm not seeing why it should be.
06:28:17funmanjhMikeS: not sure ..
06:28:27jhMikeSoh, I think the acking doesn't happen if no messages were broadcast
06:28:28funmani copied gigabeat code
06:28:45jhMikeSgigabeat f? It's a bridge IC
06:29:04funmanIIUC the problem was that connection isn't detected
06:29:09funmanso not related to bridge
06:29:56funmanwhere does tha tcome from ? the code uses usb_detect()
06:30:04jhMikeSif no queue exists for SYS_USB_CONNECTED, num_acks_to_expect stays 0, thus never reaching USB_SLAVE_MODE
06:30:50funmandunno what you're talking about :)
06:31:09jhMikeSin usb.c
06:31:23jhMikeSif(num_acks_to_expect > 0 && −−num_acks_to_expect == 0)
06:32:18funman num_acks_to_expect += queue_broadcast(SYS_USB_CONNECTED, 0) - 1;
06:33:03jhMikeSyeah, if there's no queue besides the USB queue, num_acks_to_expect will be 0
06:34:40jhMikeShrm, I guess the backlight queue is registered
06:35:49jhMikeSor is it? I can't tell yet - depends on BACKLIGHT_FULL_INIT
06:36:05 Quit robin0800 (Ping timeout: 252 seconds)
06:36:50funman#define HAVE_PRIORITY_SCHEDULING
06:36:56funman#define HAVE_PRIORITY_SCHEDULING should only be enabled for PP ?
06:38:32jhMikeSIt was preferred for the beast but for PP probably isn't the critical in the bootloaders
06:39:24jhMikeSnope, no backlight queue for an AMS bootloader
06:48:46 Quit Topy44 (Ping timeout: 240 seconds)
06:54:22funmansomeone with a nano2g can test usb in current build?
06:54:26funmanor ipod classic
06:54:43 Join Topy44 [0] (
07:04:46 Quit factor (Ping timeout: 240 seconds)
07:14:13 Quit othniel (Ping timeout: 248 seconds)
07:18:34 Join factor [0] (~factor@
07:19:01 Nick kugel is now known as kugelp (~kugel@rockbox/developer/kugel)
07:21:14 Quit BHSPitMonkey_ (Remote host closed the connection)
07:26:27***Saving seen data "./dancer.seen"
07:30:19CIA-1New commit by funman (r31242): usb-s3c6400: make isr more clear ...
07:30:39CIA-1New commit by funman (r31243): usb-drv-as3525v2: simplify usb_reset isr
07:31:59CIA-1r31242 build result: All green
07:32:31 Quit jordan` (Ping timeout: 252 seconds)
07:33:43CIA-1r31243 build result: All green
07:34:04CIA-1New commit by funman (r31244): usb-drv-as3525v2: cosmetics
07:35:46CIA-1r31244 build result: All green
07:39:54 Join jordan` [0] (
07:43:50 Nick kugelp is now known as kugel (~kugel@rockbox/developer/kugel)
07:49:07 Quit Topy44 (Ping timeout: 268 seconds)
07:52:48 Quit Torino10 (Quit: CGI:IRC)
08:04:42 Join Zagor [242] (~bjst@rockbox/developer/Zagor)
08:12:54 Quit saratoga_ (Quit: Page closed)
08:13:06 Join Rob2222 [0] (
08:13:48[7]funman: still broken with r31244 :(
08:19:02 Quit Rob2223 (Ping timeout: 245 seconds)
08:22:08 Join GodEater_ [0] (93722cc9@rockbox/staff/GodEater)
08:23:49 Join Keripo [0] (
08:26:13[7]however it doesn't seem to lock up completely any more, it just doesn't work
08:27:06*[7] might do some bus sniffing when he arrives at his office
08:39:04 Quit Keripo (Read error: Connection reset by peer)
08:41:40 Join bertrik [0] (
08:41:46 Quit bertrik (Changing host)
08:41:46 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
08:41:59 Quit Zoiah (Remote host closed the connection)
08:42:05 Join Zoiah [0] (
08:43:41 Join jdgord_ [0] (~AndChat@
08:47:32 Quit JdGord (Ping timeout: 245 seconds)
08:50:32 Join wodz [0] (
08:50:35 Quit jdgord_ (Ping timeout: 255 seconds)
08:52:20 Join paradoxgo [0] (
08:52:32 Part paradoxgo
08:54:38 Quit kadoban_ (Quit: bye)
08:55:07 Join kadoban [0] (
08:58:12 Join ender` [0] (
08:59:48 Join Keripo [0] (
09:00:34 Quit Keripo (Client Quit)
09:17:22jhMikeSfunman: I found what the real problem is. There's a missing usb_enable(true) in usb.c when USB_DETECT_BY_CORE is not defined and the USB stack is used
09:19:58 Quit perrikwp (Read error: Connection reset by peer)
09:21:06 Join perrikwp [0] (
09:21:26 Join othniel [0] (~othniel@
09:21:26 Quit thegeek (Read error: Connection reset by peer)
09:21:36 Join thegeek [0] (
09:26:31***Saving seen data "./dancer.seen"
09:36:09 Quit factor (Read error: Connection reset by peer)
09:54:09 Join factor [0] (~factor@
10:02:50 Join pamaury [0] (~pamaury@rockbox/developer/pamaury)
10:04:11 Quit Farthen (Ping timeout: 268 seconds)
10:05:48 Join Farthen [0] (~Farthen@2a01:4f8:101:2a4:0:bc28:b2e1:9)
10:10:36pamauryfunman: any success with usb ?
10:14:46 Quit fs-bluebot (Read error: Connection reset by peer)
10:16:45 Quit kadoban (Read error: Operation timed out)
10:18:20 Join fs-bluebot [0] (
10:19:39 Join einhirn [0] (
10:19:51CIA-1New commit by jethead71 (r31245): Fix the real issue with AMS bootloader USB mode. A call to usb_enable was missing in usb.c when using the USB stack and USB_DETECT_BY_CORE was not ...
10:21:46CIA-1r31245 build result: All green
10:22:50wodzpamaury: considering the lack of success with ld have you tried two pass link approach?
10:23:56 Join wilsonsamm [0] (
10:24:18wilsonsammhey. I've just installed rockbox onto my sansa clip
10:24:32wilsonsammit's great, but I have a problem running lua scripts
10:25:10wilsonsammwhen I choose one from the file menu, I get an error saying "couldnt open viewer: .rockbox/rocks/viewers/lua."
10:25:34wilsonsammthat's right, that file doesn't exist. Where can I get it?
10:25:53wilsonsammI've already looked over
10:26:31pamaurywodz: not yet, I didn't have time. I don't really like the two pass approach but I must admit I don't have another idea. There are still questions about how to do it: either build a map from, then use bin2c to produce de c file which is compiled in, or directly include the binary using the linker or some objcopy. In any case the problem is to make sure no address changes between two links
10:27:15JdGordonbertrik: pin
10:34:40 Join rarog [0] (
10:37:47 Join nick-p [0] (
10:42:53JdGordonnick-p: I'm gonna commit that wps fix if you think its good to go :)
10:43:08JdGordonbut i got another wishlist if you want osmethin to have a crack at? :)
10:43:16 Join gwlegion [0] (~gwlegion@unaffiliated/gwlegion)
10:44:03nick-pJdGordon: While I didn't test the right align stuff,nothing crashed with it - I reckon it can't be worse.
10:44:11JdGordongevaerts: whats the usual story with bugfixes after a branch? do we try backporting anything and release in a couple weeks again?
10:44:11gwlegionwhat about ipod nano 4g dev ?
10:44:25nick-pJdGordon: I'm a bit busy at the moment to do any testing though
10:46:19gevaertsJdGordon: basically the policy is "use your own judgement" :)
10:47:04gevaertsIt's a tradeoff. The branch will get little or no testing in practice at this point, so if there's a risk of breaking things for a minor bug, don't do it
10:47:07JdGordonok, so then I tihnk this commit shold go through, but I don't have (and cant get :D )a 3.10 checkout
10:47:36JdGordon... after a week or so to make sure it doesnt cause any issues :)(
10:47:36*gevaerts has some doubts about "can't get" :)
10:47:50JdGordonon my connection and data allowance it will take too long :)
10:47:52gevaertsping me when it's ready then
10:48:55CIA-1New commit by jdgordon (r31246): Fix FS #12237 (caused by r30302) where centered skin text isnt eqactly centered
10:48:57 Join LinusN [0] (
10:49:04JdGordonI'll do it from melbourne next week if i remember
10:51:32CIA-1r31246 build result: All green
10:51:46JdGordonoh nice... git svn rebase is still working after that commit :)
10:57:35JdGordongevaerts: will the build system and install stuff all DIAF if plugins had spaces in the names?
11:00:24gevaertsno idea
11:01:28JdGordonI'm toying with the idea of cleaning up the plugin file listsing (and openwith menu) to replace _'s with spaces and Title Case the names on the fly
11:01:38 Join lorenzo92 [0] (
11:01:41JdGordonkugel thinks it should be done with a hardcoded lookup instead
11:01:45JdGordonor we just fix the filenames
11:04:31 Join dfkt [0] (dfkt@unaffiliated/dfkt)
11:05:19lorenzo92kugel: you did a good job with keymappings, but I need to fix something. Context menu is done by pressing menu key, not keeping pressed central key ;)
11:06:50JdGordoncontext menu is done with the "center/ok" button on every target
11:07:02JdGordonthe keymap wars have begun :p
11:07:07lorenzo92yeah :D
11:07:16lorenzo92but we have a dedicated key...
11:09:44kugelmany targets have a menu key
11:10:09kugelit's used to go to the main menu on all. and all targets use long select for the context menu
11:11:28kugelJdGordon: I would favor fixing the filename
11:12:01kugelthe conversion can be automated when making the .rock file if you like
11:12:11kugelno need to add code and special cases to rockbox for this
11:12:16JdGordonthats true
11:13:37 Quit preglow (Ping timeout: 240 seconds)
11:14:33rarogpamaury: I thought about the sd and boot problem, our problem compared to OF is that rockbox is in the file system, while the firmware is in a separate space. what if the external card is detected, but mounted only after the boot?
11:15:17kugelJdGordon: also I'm unsure if it should be done at all :)
11:15:36JdGordonrandom_folder_advance_config looks like crap
11:15:39pamauryit's not a problem, we do it on other targets as well; the problem is that for some reason our code tries to read/write from/to the sd card before it's initialized and my code somehow doesn't handle that properly
11:16:38rarogah, so this is something like the dining philosophers problem?
11:18:09rarogdo you have your current local implementation somewhere like pastebin, so I could have a look on it?
11:19:21 Join Barahir [0] (
11:20:06pamauryyes, wait a minute
11:20:33lorenzo92kugel: okay for the key but still there is a double action_std_context I guess...
11:22:13 Join preglow [0] (
11:26:34***Saving seen data "./dancer.seen"
11:26:40pamauryrarog:, this my raw diff so it's not exactly clean and include the backtracer
11:27:07pamauryI've put a panic where the dramatic sequence of events starts basically
11:28:30*Zagor moves the elfs and maps release files into a "debug" subdir
11:28:32rarogok, so as long as I boot with build with diff, I'll have a forced debug?
11:28:50pamauryno, only if you boot with sd inserted
11:28:56pamaury(or rolo)
11:29:12pamauryand perhaps on usb unplug :)
11:29:41pamauryyou can probably remove the panic and then it will lockup only on boot with sd inserted
11:30:32rarogwell, it's ok, as long as I can cicrumvent the lockup if I need to :D this is the situation here. :D
11:34:38 Quit lorenzo92 (Quit: ChatZilla 0.9.87 [Firefox 8.0/20111115183813])
11:48:40 Join markun [0] (~markun@rockbox/developer/markun)
11:55:21rarogah, I know why it can't apply. a git diff. >.>
12:48:42rarogpamaury: do you also have the diff for unwarminder? or is this somewhere in the flyspray?
12:54:31rarogok, can't convert it... I'll just use git instead of svn now...
12:56:56Tornerarog: we are aiming to switch to git for all development within a week or so, so it won't hurt ;P
12:58:21 Quit Guinness (Read error: Connection reset by peer)
13:02:26rarogyeah. got it now here. :D
13:03:08rarognow I'm looking into the unwarminder patch... as far as I understand, i should just apply the latest version. (v3) but it gives me an error, if I check the patch.
13:03:33rarogerror: cannot apply binary patch to 'firmware/target/arm/unwarminder/.get_sp.c.swp' without full index line
13:04:05 Join MethoS- [0] (~clemens@
13:09:05JdGordonmaking text which changes dynamically scroll is damn simple!
13:09:08 Join Guinness [0] (
13:09:13JdGordonnow, why the heck wasnt this dne years aog?
13:14:11rarogok, I had to use a different patch version it seems.
13:14:14Tornerarog: that looks like the patch incorrectly includes some vim tempfiles
13:18:09pamauryrarog: it's on FS bu
13:18:15pamauryt it's included in my diff
13:18:30pamaurywithout modification except for client.h and which doesn't do anything
13:19:30JdGordondoes anyone have any idea why this should cause a line to not scroll? "if (current_vp->width - x * 8 >= w) return;"?
13:19:37JdGordoni mean, why that check might be there...
13:19:45JdGordonamiconn: kugel?
13:19:58JdGordonin lcd_bitmap-common.c around line 400ish
13:21:08gevaertsJdGordon: looks like a test to see if the string fits
13:21:15rarogpamaury: not really, at least not if I download your diff.
13:21:26rarogthe files are referenced, but not created.
13:21:27JdGordongevaerts: fits "how" though?
13:21:32JdGordonthat seems very arbitrary?
13:21:54JdGordonunless im missing something obvious?
13:21:54pamauryrarog: what do you mean ?
13:22:14pamauryoh you're right
13:22:17gevaertsJdGordon: I'm not sure about the *8, but (width - x) looks like the remaining size in the viewport, and w is the width of the string. Is x in character coordinates there?
13:22:43JdGordonchar cell, not pixel
13:22:45pamauryrarog: FS #12302, sorry :)
13:22:47fs-bluebot Add backtrace to the panic screen on ARM (patches, new)
13:22:54JdGordonso *8 makes it a "pixel"
13:23:38rarogpamaury: so I have to apply v3 patch and after it your diff?
13:23:40kugelJdGordon: I guess it checks whether the string is long enough for scrolling
13:23:42gevaertsright, then assuming a character is 8 pixels wide (which I don't know), width - x*8 is the available space for the string
13:23:48kugelit seems poor, though
13:24:27pamauryrarog: yes, but there will be a conflict on some file, wait a minute I will upload a complete diff
13:25:03rarogpamaury: yes, found the conflict now while checking the sd.diff. :D
13:25:40JdGordonoh fuck i hate our lcd_puts* api
13:26:14pamauryrarog: I've update the gist, you can redownload it
13:26:35***Saving seen data "./dancer.seen"
13:27:45kugelJdGordon: ask Zagor about his branch again :)
13:27:47JdGordonthat check *must* be to see if it needs to scroll
13:27:58JdGordonin which case it is very wrong
13:28:07 Quit nick-p (Quit: Leaving)
13:33:29rarogpamaury: now it applies and compiles like a charm. :)
13:36:10 Quit soap (Changing host)
13:36:10 Join soap [0] (~soap@rockbox/staff/soap)
13:38:39JdGordon <- any comments?
13:38:59JdGordoncurrently lines have to explicitly say they want to not restart the scroll for the line
13:39:05JdGordonI'm not sure that needs to be done though
13:40:20JdGordonand of course make sure the scrolling line is in the current_vp :)
13:41:53 Part gwlegion ("Quitte")
13:46:11JdGordonyeah, that extra style flag doesnt appear to be needed
13:53:24CIA-1New commit by jdgordon (r31247): Allow scrolling lines to have their content changed without restarting the scroll line. This means skin lines with dynamic tags can be updated in ...
13:54:30JdGordonIf that ends up making things look crap I'm open to reverting
13:55:32CIA-1r31247 build result: All green
13:58:35 Quit kugel (Remote host closed the connection)
14:01:10 Join Misanthropos [0] (
14:03:42 Join kugel [0] (~kugel@rockbox/developer/kugel)
14:05:29rarogpaumary: I don't know why, but when I uncomment the panic line, everything starts nice and clean. Or should I again make also the firmware to flash?
14:10:00 Join jlbiasini [0] (
14:11:10jlbiasinihello, I just made a new battery bench... but I forget I had the backtrace patch... Can it have any effect on the result?
14:15:10gevaertsjlbiasini: do you want a correct answer or a useful answer? :)
14:17:24jlbiasini?? not sure to understand?
14:17:41gevaertsWell, the correct answer is "yes"
14:18:00gevaertsThe effect will not be measurable though
14:18:09soapI'd run a battery bench to answer the question. ;)
14:18:19gevaertsWell, it most likely won't be measurable
14:18:20jlbiasiniok I did understood then! :D
14:19:13jlbiasiniIt would be fun to find out that it increase playback time :D
14:19:42gevaertsIt won't. The effect is "less RAM available for buffering"
14:20:11rarogjlbiasini: btw, you can update the bootloader and wiki for fuze+. :)
14:20:14jlbiasiniyeah I know but this is the theory !
14:20:43jlbiasiniyeah I'm on the baterry time's one
14:23:08raroghm... is the git not yet in sync with svn?
14:23:44Tornethe new git repository is being mirrored manually
14:23:46Torneby me :)
14:23:52TorneGive me a moment :p
14:24:09rarogok. :P
14:24:46kugelis the new repo accessible already?
14:24:51Tornekugel: it has been for weeks
14:25:00kugeloh nice
14:25:27Tornestill not automatically mirrored. but there's no point now
14:25:41Tornei'm gonna finish up the last things we need to do Real Soon and then get zagor to shut off write access to svn
14:25:43JdGordonwhy not?
14:25:45kugelfor weeks? wasnt it just a few days ago when the old mirror was moved away?
14:25:58Tornekugel: It was already accessible through gerrit (http or ssh)
14:26:04Tornethe git daemon pointing at it is recent
14:26:25TorneJdGordon: because i'm hoping we can switch to git for real by the weekend :)
14:26:31JdGordonfucking aye!
14:26:36JdGordonis gerrit actually usable yet?
14:26:52Tornei'm not turning on code reviews just yet
14:26:57 Join keyb_gr [0] (
14:27:06JdGordonso thats a no? :)
14:27:07Torneone thing at a time; we need to move main infrastrucure first
14:27:18Torneyou can create account, get permissions, read code
14:27:26kugelhow can you read code?
14:27:28Torneonce we shut svn off i will turn on push permissions for committers
14:27:42JdGordonsure, though it owuld be nice to get a feel of it with an actual project, even if we can use it directly to push things
14:27:44Tornethen once we deal with all the buildbots and so on i'll get around to allowing code review
14:27:51JdGordonbut yeah, move to git and ignore everything else :)
14:27:52Tornethe sandbox repo is still there; you can use it
14:27:57Tornei know it's not "real"
14:28:06kugelI mean the new git repo
14:28:16Tornekugel: git clone
14:28:22Tornewe don't have a web based browser up right now
14:28:31Torneif you'd like to suggest one, feel free to propose it to zagor ;)
14:28:43Tornegerrit supports integrating with, i believe, anything you can specify a url pattern for
14:28:50Torneit's meant to be used with gitweb but other stuff probably work
14:28:54kugelah I thought we did when you said it's available through gerrit
14:28:58Tornewe don't need access control so it doesn't have to be implemented tightly
14:29:05Torneit's available through gerrit, yes. as a repository you can clone
14:29:10Torneover ssh or http
14:29:22Tornegerrit doesn't provide a browser
14:31:49Torneanyway, i'm probably not going to get anywhere today. currently i'm at my last day of work this year so i need to finish up pending stuff; then going to work xmas party for free booze, then will be at home and drunk. ;)
14:31:58Tornebut tomorrow, i'll deal with more of it
14:32:18JdGordonwith the hangover or git?
14:32:34Tornei don't get hung over
14:32:40Tornei can absorb infinite alcohol
14:33:44rarogTorne: please post your diff to body.c
14:34:16raroginfinite alcohol is a nice feature.
14:34:31Tornewell okay probably not infinite
14:34:41Tornebut as much as i comfortably care to consume in the first place
14:36:54jlbiasinirarog: is the modification for the second lcd issue already in the tree?
14:37:09jlbiasiniI meant svn
14:37:17rarogjlbiasini: yes. and working like a charm!
14:38:14jlbiasiniwiki updated then
14:41:10rarogjlbiasini: do you need the updated or will you compile it yourself?
14:44:05 Join abloomfield [0] (
14:44:14jlbiasiniI've got the first kind. It was just to know if I could write in the wiki that both lcd are handled
14:47:40rarogyes sure, but the non-compiling people will need an updated firmware to see the bug go away.
14:48:02Misanthroposhey there - i built a rockbox with the current svn tree - and listening to .flac files i noticed the right channel channel dies on low frequencies.... first i had my earplug blaimed but the OF does not show that behaviour - anybody had similar experiences?
14:48:05 Join webguest731 [0] (
14:48:24Misanthroposi heard its related to FS12371
14:48:44 Quit webguest731 (Client Quit)
14:49:12MisanthroposFS #12371
14:49:13fs-bluebot FLAC 32 bits 96khz surround 5.1 file not playing (bugs, unconfirmed)
14:49:25jlbiasiniyou mean that is a bootloader upgrade??
14:50:51 Quit abloomfield (Quit: CGI:IRC)
14:54:01rarogthe screen must be initialized via firmware so it's working directly at start and not at rolo.
14:54:10rarogso it's time for v2.
14:54:15Misanthroposso - I reverse patched with flac_v5 and my flacs play nicely again
14:55:40rarogand by the way, do you use an micro-sd card? I can't reproduce any lock paumary has. his patch works like a charm for me, no locks, recognition of cd card on boot time and later.
14:57:27 Join Topy44 [0] (
14:57:28rarogMisanthropos: And now you might experience some hanging while playing AFAIR.
14:58:27Misanthroposrarog, funny - but i remember a hang in once occasion - where the player started looping .5 secs or so / thought it was a problem accessing the sd
14:58:53Misanthroposbeside that - but i think i dont have 5.1 flacs
14:59:20jlbiasiniWell I do have a sd but a 32GB's one... Not if I want to test stuff with that....
15:00:24rarogwhy? risking data loss?
15:00:27Misanthroposi use mine with a clip+ ... usb support works with some glitches.... especially on connect, it can crash rockbox
15:01:31jlbiasinithis is still a bit expensive for me , I would want to have it dead because of it. I'd rather buy a 2GB to run test...
15:02:17rarogfor paumary it just doesn't let rockbox boot. for me it does read like a charm.
15:06:11rarogok, very, very strange....
15:06:56rarogIn 80% of time everything starts normally with sd patch, and once i had a hanging at boot-time and once a white panic screen with backtrace.
15:06:56jlbiasiniafter ROLO it doesn't hang? ;)
15:06:58pamauryrarog: perhaps I made a mistake and upload the wrong version to my device when I tested it, it was late and I was very tired
15:07:19pamaurybut I think it's timing related so it's a bit randon
15:07:26pamaury*random :)
15:07:32rarogbut removing the card, starting rockbox and shutting it down and starting again solved it.
15:07:39rarogit really seems to be random.
15:08:18rarogif hanging at boot time I didn't see anything besides "CAFEBABE" line.
15:10:29rarogBut I also experienced some glitches with OF sometimes, where everything hanged and I had to reboot it with the power button holding for 14 seconds. though this could be completely unrelated.
15:11:09 Join TheLemonMan [0] (
15:12:00pamauryrarog: no I think it's really rockbox really
15:12:34rarogok. but indeed I had hard time to reproduce this.
15:12:47rarogand in most cases everything works.
15:13:14rarogat least at my side.
15:18:21rarogah, and if I have the sd card inserted, rockbox doesn't offer me neither of the drives for my os.
15:19:26rarogwith sd card out, pluggin in usb offers me the internal storage again.
15:20:02jlbiasinibootloader's link updated on the wiki
15:26:37***Saving seen data "./dancer.seen"
15:33:56 Join lorenzo92 [0] (
15:39:04 Quit lorenzo92 (Quit: ChatZilla 0.9.87 [Firefox 8.0/20111115183813])
15:47:43 Join WalkGood [0] (~4@unaffiliated/walkgood)
15:49:18pamaurywodz: I hacked the Makefiles to produce a symbol map, I'll finish that during my coffee break and tell you if it works
15:52:13jlbiasinipamaury: if you have a look at the batterry bench files i updated on the wiki, you'll see that the player is starting stating there is 15 hours left although it last only 10. This is also a minimal use condition because I just let the beast working in a drawer which meams no lcd consumption at all...
15:53:01pamauryjlbiasini: the battery has not been calibrated, the estimation cannot be right at the moment
15:54:11 Join Buschel [0] (
15:54:33rarogpamaury: the thing devices not showing up if the sd card is inserted is just an enumeration bug or is it just a not yet implemented feature?
15:54:41BuschelMisanthropos: hi, you reported some details about the flac issue with current svn
15:54:52BuschelMisanthropos: can you please upload such file?
15:55:19pamauryrarog: it's probably the same bug as the lock up
15:55:49raroghm... so it's not clean enough yet or has some race conditions.
15:56:39pamauryI told you, there is a bug left, the very fact that it lock up randomly is the proof :)
15:57:27rarogindeed. but yesterday it sounded as if it's reproduced instantly. and not as rare as I see it. :D
15:57:38MisanthroposBuschel, any flac - i tried several
15:57:55rarogand I hate the random bugs, they are so hard to track down. :/
15:57:58pamauryon my device it's 99% of the time a lock up but since it's timing related, it depends on lots of factors
15:58:19Buschel? I compared two files via test_codec, the output was binary identical before and after my change
15:58:25pamaurybut I know exactly the sequence of events which lead to it, but not why it freezes though
15:59:35rarogand you hope to track it better down when you have the better debug/symbol output, right?
16:00:02pamauryfor one part, but it won't solve everything magically
16:00:56rarogprobably not
16:01:31wodzrarog: it is rather that pamaury expressed interest in symbol tables some time ago :-)
16:02:28pamaurywell, I also need the backtrace to debug it and the symbol table will make me win some time (actually not because I need to write it, but then it will :D)
16:03:28rarogpraeoptimization. :D
16:05:29funmanpamaury: usb seems to work much better now
16:06:01funmanpamaury: i still have locks at insertion though
16:06:12funmanbut at a much lower rate at least
16:08:15wodzdo nano-2g works after this changes?
16:08:29wodzdoes * work
16:08:49 Join y4n [0] (y4n@unaffiliated/y4ndexx)
16:11:52BuschelMisanthropos: I am helpless here. The official test files flac_5 and flac_8 decode to identical output. please provide a test sample to allow me to fix your issue
16:13:24 Quit mudd1 (Read error: Connection reset by peer)
16:14:32funmanwodz: apparently not < [7]> funman: still broken with r31244
16:15:26funmanpotential problem is r31229
16:16:28wodzthats rather unfortunate this change was not tested on actual target
16:16:59pamaurywodz: symbol table works ! \o/
16:17:37wodzshow the screenshot :P
16:17:56funmanwodz: indeed, i thought the change were simple but i managed to let errors slip through
16:19:27pamaurywodz: I don't have a camera unfortunately
16:19:41 Quit Llorean (Read error: Connection reset by peer)
16:19:46wodzwhat a pitty
16:20:23pamaurythe bad point is that it requires to modify all linker files as I'm not sure we cannot include the symbol in a special section. And it requires some modifications in the makefiles
16:20:41pamauryso it will need a cleanup to be an optional
16:21:07pamauryI'll upload the patch tonigh, except if you want some raw ugly diff now
16:22:17wodzI can't play with this now anyway
16:24:39 Part LinusN
16:26:28 Part jlbiasini
16:30:51 Quit TheLemonMan (Quit: WeeChat 0.3.6)
16:31:29CIA-1New commit by funman (r31248): usb-s3c6400: fix confusion between USB_DIR_IN and DIR_IN ...
16:32:28funman[7]: r31248 should help a lot
16:34:14CIA-1r31248 build result: All green
16:35:23funmanpamaury: usb doesn't work in zip bootloader btw, i see a config with 0 interfaces
16:36:23pamaurywithout any others errors ?
16:36:23pamauryperhaps you need to enable some usb drivers ?
16:37:11funmanit should work already, it's enabled for amsv1
16:37:38funmanhmm no you're right it's missing
16:42:05CIA-1New commit by funman (r31249): Sansa AMSv2: enable usb storage in bootloader
16:42:12funmanworks fine except the messages aren't displayed on the zip screen
16:44:17CIA-1r31249 build result: All green
16:51:24funmanpamaury: only ep0 is actually bidirectional
16:51:32pamauryfunman: yes
16:51:41pamaury(on this controller)
16:51:44funmanand input and output eps are separated by step of 2
16:52:25pamaurythis is not mandatory but both devices have ep like this
16:52:49pamauryin full generality, you can read some register which tells you the ep configuration
16:52:56funmanhm yeah HWCFG1
16:53:10funmanhm i can't use ep1 as input though
16:53:11 Quit scorche (Ping timeout: 252 seconds)
16:53:21pamauryI need to leave, good luck
16:53:24 Part pamaury
16:54:31funmankernel panic:
16:54:43 Join scorche [0] (~scorche@rockbox/administrator/scorche)
16:56:19 Part Zagor
17:04:52 Quit antil33t (Ping timeout: 252 seconds)
17:08:48 Join dreamlayers [0] (
17:08:48 Quit dreamlayers (Changing host)
17:08:48 Join dreamlayers [0] (~bgjenero@rockbox/developer/dreamlayers)
17:11:01 Join Misan [0] (
17:13:22 Join antil33t [0] (~Ahurhurr@
17:18:36CIA-1New commit by funman (r31250): disktidy: clean more Sansa AMS OF files
17:20:38CIA-1r31250 build result: All green
17:26:39funmanhow could i test fullspeed on linux? ehci is not a module anymore on ubuntu kernel
17:26:41***Saving seen data "./dancer.seen"
17:26:57wodzfunman: you can disable ehci at runtime
17:28:08wodzfunman: /sys/bus/pci/drivers/ehci_hcd/ and you have to echo 'unbind' to appropriate file
17:29:12wodzit should be a link in form similar to 0000:00:1a.7
17:31:42funmanseems to work fine in full speed
17:31:55 Quit einhirn (Quit: Miranda IM! Smaller, Faster, Easier.
17:33:57funmanstill can't use EP1 though :o
17:35:34wodzhmm, this weird jumps in rk27xx ftl init code do some binary search
17:37:19 Quit mystica555_ (Quit: Ekkusu Chatto)
17:38:49 Join mystica555_ [0] (
17:42:05 Quit mystica555_ (Remote host closed the connection)
17:44:01 Join mystica555_ [0] (
17:45:55 Join pamaury [0] (
17:45:55 Quit pamaury (Changing host)
17:45:55 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
17:47:37pamauryfunman: why can't you use EP1 ?
17:48:18 Quit bluefoxx (Quit: bluefoxx)
17:52:13funmandon't know. if i replace the list by 0, 1, 3 or 0, 1, 3, 5 usb doesn't work
17:52:31funmani have another commit incoming
17:52:53pamauryperhaps that's why I didn't put ep1 in the list ?
17:54:32pamaurybut I really can't remember doing that, which is a bit weird :)
17:56:58keyb_grSeems like the 3.10 manual is outdated, not reflecting changes in the main menu. Oops.
17:57:07keyb_grHow'd that happen?
17:57:33 Join bluefoxx [0] (
18:02:08 Join kadoban [0] (
18:05:54funmanpamaury: it works
18:06:08funman0, 1, 3
18:06:21pamaurywhat was wrong ?
18:07:01funmani change stuff so it looks more like usb-s3c6400.c
18:08:15 Join GeekShad1w [0] (
18:08:47 Quit GeekShadow (Ping timeout: 252 seconds)
18:12:54CIA-1New commit by funman (r31251): usb-drv-as3525v2: make it look more like usb-s3c6400 ...
18:14:57CIA-1r31251 build result: All green
18:15:06funmanbertrik: mpegplayer fps is ~23
18:17:08funmanpamaury: is there something wrong with using 4 in endpoints and 3 out endpoints (including ep0)
18:18:15gevaertsThat's a lot of endpoints. What are you running?
18:19:20gevaertsWell yes
18:19:42gevaertsI mean, UMS uses one in and one out, HID (in our setup) uses one in
18:20:10funmanah so that's why i only see 2 and 3 in lsusb
18:20:18funmanHID doesn't work (no isochronous)
18:20:28gevaertsHID doesn't use isochronous
18:21:17funmanah we don't use isochronous at all?
18:21:30gevaertsno, only pamaury's audio branch has isochronous
18:21:45pamauryand it requires a driver rewrite more or less anyway :)
18:22:06funmanwhat's needed for HID then?
18:22:32 Quit factor (Ping timeout: 245 seconds)
18:22:57gevaertsOn the wire that's really the same as bulk. IIRC in the ARC driver we treat those exactly the same
18:24:10funmani remember that it was impossible to do for the nano2g and amsv2 use the same controller
18:24:31gevaertsThere were bugs on nano2g. That
18:24:38gevaertss not really the same as "impossible" :)
18:24:42pamauryyeah there is absolutely no difference between bulk and interrupt
18:24:47pamauryit's only a host difference
18:25:24funmanpacket size maybe?
18:25:30funmanthe thing i just removed
18:25:37gevaertsI'm pretty sure the nano2g bugs are either driver bugs involving simultaneous MSC and HID, or our HID code not being too good
18:26:03funmaninterrupt packetsize is twice bigger in high speed
18:26:24pamauryfunman: the packet size constraints are a bit difference but again, there is no hardware different, except for the fifo size perhaps
18:26:34gevaertsHID packet size should never go near that I'd say
18:27:35funmangevaerts: indeed, 512 ought to be enough for everyone
18:28:01gevaertsfunman: I'd be surprised at HID packets for keyboard or mouse emulation larger than a few tens of bytes :)
18:32:07MisanBuschel, i meant any non 5.1 flac
18:33:02BuschelMisan: that was understood. another user already attached a file to the bug report. with this I already can reproduce this issue
18:39:21funmanthere's some lang strings corruption when i enabled HID
18:42:30dreamlayersfunman: do you just find that the USB screen text is displayed incorrectly, or does other stuff get corrupted?
18:42:53funmanusb screen text i think
18:43:12funmanand usb settings
18:43:36dreamlayersDuring the USB screen, the normally used font is unloaded and the built-in font is used.
18:44:19funmanmake clean makes it better (but hid deoesn' twork)
18:45:42BuschelMisan: got it, baaad mistake that could easily been overseen :/
18:47:20CIA-1New commit by buschel (r31252): Correct decorrelation within flac decoding. Finally fixes FS #12371.
18:49:08MisanBuschel, i can update to current svn then?
18:49:15CIA-1r31252 build result: All green
18:49:28dreamlayersfunman: so I guess you just needed to rebuild the language files after enabling HID. I was wondering if you were just seeing problems with USB screen text due to FS #12436.
18:49:29fs-bluebot USB keypad mode text unreadable in Russian regardless of selected font (bugs, unconfirmed)
18:50:03funmangevaerts: we always use 512 bytes on amsv1 and it works fine
18:50:09gevaertsdreamlayers: isn't that one just sysfont not having anything non-ascii?
18:50:48gevaertsfunman: yes, but that's just the maximum. You're allowed to send smaller packets
18:50:48dreamlayersgevaerts: yeah.
18:51:54funmangevaerts: yeah i just meant that the problem with HID is not in packet size
18:51:57pamauryI have a toy question for C experts: let's say I want to build a symbol tables knowing the names of the symbols. I would like to do something like struct symbol_t symbols[] = { {.name = "usb_enable", .addr = (uint32_t)&usb_enable;} ... }; problem is: usb_enable would be undeclared. Of course I could define it like void usb_enable(); but this wouldn't work if some function has a prototype imported via a header
18:52:36funmanpamaury: you need to declare it of course
18:52:57*gevaerts nods
18:53:21pamaurybut I don't know the right type of the function and I don't want any type mismatch
18:53:55funmanif you have a mismatch you knwo the type already, right?
18:54:09pamauryI want to autogenerate it !
18:55:02pamaury(there are around 1600 symbols)
18:55:13gevaertsCan't you generate them from the elf file and then insert them in the binary in some way?
18:55:55*gevaerts isn't convinced that something that doesn't handle static functions is *that* useful
18:56:31funmanpamaury: then don't include any header which could declare those functions?
18:56:59pamauryyes, that's the only solution I found :-/ Anyway I don't do that, that's why it's a toy question
18:57:15dreamlayersWhat about gcc -mpoke-function-name ?
18:59:31pamaurythat's interesting
19:01:34 Quit wilsonsamm (Quit: Leaving)
19:02:58CIA-1New commit by buschel (r31253): FS #12443: Implement downmixing to stereo for multichannel flac.
19:05:02CIA-1r31253 build result: All green
19:06:39funmani receive eopframe interrupt, not sure what it means
19:06:40pamaurythere might be a little problem though: the arm unwinder will not give us the pc of the beginning of the function
19:06:50 Join factor [0] (~factor@
19:06:58pamauryfunman: that's for isochronous, you should mask it
19:07:02funmanpamaury: fp?
19:07:29funmanHID with storage disabled doesn't work either
19:11:22funman[ 7866.945691] generic-usb: probe of 0003:0781:74E5.0008 failed with error -22
19:12:45MisanBuschel, thx a bundle :-) Non 5.1 flacs sound nice again
19:13:34pamauryfunman: the frame pointer is not given by the uwninder but I guess that can be changed
19:13:46funmanwhen stotrage is enabled i see the hid screen, else i just see the normal menu
19:13:59BuschelMisa: thanks for the feedback
19:14:11funmanin dmesg (with storage enabled): [ 7999.588180] usb 1-1: new high-speed USB device number 71 using ehci_hcd
19:14:14funman[ 7999.806748] scsi83 : usb-storage 1-1:1.0
19:21:08pamaurygevaerts: I disagree, non-static functions are already of a great help
19:21:31funmangevaerts: lsusb reports max packet size as 8 bytes for hid
19:26:38 Join lebellium [0] (
19:26:45***Saving seen data "./dancer.seen"
19:32:01 Join Keripo [0] (
19:32:14 Join Strife89 [0] (
19:37:34[7]funman: what are you doing with the usb drivers btw? are you trying to merge them? if yes, did you already do that? i.e. is the clip+ or whatever currently using the s3c6400x driver? if yes, does it work there?
19:37:53CIA-1New commit by funman (r31254): Sansa AMSv2: don't panic if microsd init failed
19:37:58funman[7]: first i'm trying to have it work on amsv2 and it's going well
19:38:09funmani only have a clip zip and a dying fuzev2, no clip+ or clipv2 though
19:38:34funmanno i didn't merge them yet because i can't test on nano2g and my small modifications are enough to break it
19:38:55funmanand no i didn't use (successfully) usb-s3c6400.c on AMSv2 (yet)
19:39:13funmanusb-drv-as3525v2.c has a few more features though
19:39:29[7]so what are you doing with the s3c6400x one in the first place? i.e. what did you modify and why?
19:39:38CIA-1r31254 build result: All green
19:39:50funmani am using the same header (register defines) for both
19:40:03funmani changed DIEPCTL/DOEPCTL to DEPCTL(direction)
19:40:05[7]ah, that makes sense :)
19:40:20funmanand factored identical code for both directions (made several mistakes in the process as you saw)
19:40:57funmandoes r31248 finally works again?
19:43:33[7]i'll test that in a minute
19:47:30 Join Horscht [0] (~Horscht@xbmc/user/horscht)
19:47:39 Quit Strife89 (Quit: leaving)
19:48:04 Join Strife89 [0] (
19:48:51*[7] needs a faster build machine
19:49:35funman[7]: :)
19:57:21ukleinekon a rockboxed sansa Fuze, can I just rename or remove /.rockbox without bricking?
19:59:01 Join liar [0] (
19:59:22funmandon't expect rockbox to work though
19:59:59ukleinekfunman: obviously, yes. My wife doesn't like rockbox compared with the OF, so this is the desired outcome :-(
20:00:12funmanukleinek: upgrade the OF to remove the bootloader
20:00:56ukleinekfunman: but it doesn't hurt to have it, does it? This way I can more easily borrow it for quick tests.
20:01:42funmanthe problem is that rockbox is booted by default and you can't change that
20:02:19ukleinekfunman: ah, so I still have to press 'left' to boot the OF
20:05:27 Quit Keripo (Quit: Leaving.)
20:07:17ukleinekalso there are some .bin and .sys files in the / of the player, I think they originate from the OF. Probably I can remove these, too?
20:07:29funmanyeah but they'll be recreated by of
20:08:36gevaertsAlso, I would imagine they didn't put those there for fun. If the player will be used with the OF most of the time, I'd leave them
20:11:51[7]funman: :(
20:14:42ukleinekan opt-in additionally to the opt-out bootloader would be great
20:15:52funman[7]: no luck with current build? :/
20:16:10[7]sadly yes
20:16:21[7]windows fails to enumerate it
20:16:22 Quit tholin ()
20:16:37 Join Don_Giles [0] (
20:16:52Don_GilesHow can I help Funman
20:17:30funmanDon_Giles: i will send you a build
20:17:57funmanDon_Giles: hold and power work?
20:18:09Don_Gilespower works
20:21:11 Join T44 [0] (
20:23:38 Quit Don_Giles (Quit: CGI:IRC (Ping timeout))
20:24:28 Quit Topy44 (Ping timeout: 248 seconds)
20:25:46 Quit funman (Ping timeout: 268 seconds)
20:31:05 Join fun [0] (
20:36:48 Quit dreamlayers (Quit: back later)
20:38:04funhm he left
20:39:41fun : should fix buttons on clipv2
20:39:51ukleinekfun: it seems irc doesn't work :-)
20:43:15 Nick fun is now known as namnuf (
20:44:09 Join maal [0] (
20:46:24Buschelmaal: thanks for supporting the bughunting with the latest flac-issue
20:47:21maalBuschel: thanks for fix, i just wanted to confirm it works
20:47:41desowinhmm, something else is draining these milliamps... removing wifi firmware and modules from OF on Connect didn't affect the runtime at all
20:49:07 Quit maal (Quit: CGI:IRC)
20:50:59 Join Topy [0] (
20:52:16desowinBuschel: could you give me any hints about implementing something like this in Rockbox ?
20:54:11 Quit T44 (Ping timeout: 245 seconds)
20:55:36wodzdesowin: Our implementation is much simpler. We choose 2 freqs sets and switch between them.
20:55:55desowinwodz: yes, I know that
20:56:23 Join funman [0] (~fun@rockbox/developer/funman)
20:57:21 Quit namnuf (Quit: leaving)
20:58:11wodzdesowin: look at firmware/target/arm/s5l8700/system-s5l8700.c set_cpu_frequency() for reference
20:59:28wodzdesowin: basicaly you need to know what is the sequence of freq change on your platform and what are constraints. Ability to change core voltage is a nice bonus if available
20:59:39desowinyeah, it's switching between fast/slow for all periphs
21:00:14desowindm320 has very limited implementation of that, just switching AHB and ARM clocks
21:01:11wodzIf I were you I would first write two versions of init code - one running at high speed and second running at low speed. When this two setups are working you will need only minor glue code and you are done
21:02:04desowinby low speed you mean lowest clock setup required to boot?
21:03:02desowinI wished I had m:robe 500 to test such changes btw.
21:03:14 Join Don_Giles [0] (
21:03:44wodzIt is a trade off. Usually we try to take low speed such that least demanding codecs doesn't boost.
21:03:49funmanDon_Giles: re, can you try funman/rockbox.sansa"> ?
21:04:22Don_GilesI pasted the file on to Clip V2 rockbox folder, no help, buttons still unresponsive
21:04:43desowindefine:east demanding codecs
21:05:14desowinso I guess BYPASSying all clocks doesn't really go into considerations for low speed
21:05:44funmanDon_Giles: after rebooting? what is the rockbox version showed on screen?
21:06:22wodzdesowin: that depends what this bypassed clocks are and how fast you can recover plls/divs
21:07:48 Join benedikt93 [0] (~benedikt9@unaffiliated/benedikt93)
21:07:53wodzdesowin: other limiting factors when choosing clocks might be the stability and slow lcd updates.
21:08:49 Quit WalkGood (Quit: me fui)
21:09:13wodzdesowin: usually it is best to change only dividers as this avoids busy-waiting for pll lock
21:10:40wodzdesowin: regarding mrobe500 - ask on the list maybe some developer may lend you one.
21:12:11funmanDon_Giles: ah i think i know what's going on
21:12:39saratogadesowin: for arm9e 30mHz is a good low speed, although if you have a slow screen you may want faster just so the UI is responsive
21:13:30desowinscreen updates could probably be handled by DMA
21:13:30funmanDon_Giles: can you download frm the same url again?
21:13:49 Quit Don_Giles (Quit: CGI:IRC (Ping timeout))
21:14:28saratogadesowin: probably worth trying a battery bench with the clock forced lower (say half of what you run it at now) to check if its actually CPU or some peripherals that are wasting power
21:15:09saratogaa lot of ARM CPUs are pretty good at idling when unused so the savings from frequency scaling can be masked by a peripheral thats using a lot of power
21:15:36 Join Don_Giles [0] (
21:15:50Don_Gilesfunman.. I am getting 404 file not found
21:16:06 Quit y4n (Quit: only amiga makes it possible)
21:17:24funmanDon_Giles: funman/rockbox.sansa"> ?
21:17:56Don_GilesNot Found
21:17:56Don_GilesThe requested URL /~<span class= was not found on this server.
21:18:15funmanweird. try
21:19:20 Join robin0800 [0] (
21:20:52Don_GilesThat link worked and yes the buttons are working again! Thanks!
21:21:13funmanthanks for your help, i'll commit a proper fix soon
21:21:19funmansorry for the bug in the first place
21:21:34funmanit should give me an exmaple to not commit code untested on target :s
21:22:04Don_Gilesno problem, that you for all your hard work on rockbox
21:22:05 Quit Xerion (Ping timeout: 240 seconds)
21:22:19Don_Gilesnow to get line out to work on Sansa Fuze V2
21:22:47 Join T44 [0] (
21:23:48dfktis it normal that the clip+ still uses OF usb when plugged in while powered off? (the zip always uses rockbox usb now)
21:24:03saratogadfkt: we haven't updated the bootloaders yet
21:24:11saratogaor rather mkamsboot I guess
21:24:14dfktah, i just built a new bootloader for both
21:24:21saratogayeah sorry its not the bootloader
21:24:30saratogaits the actual mkamsboot binary that needs updating
21:24:34dfktso i guess i was too soon :)
21:24:54saratogai think funman wanted to wait a bit longer to make sure USB was stable before pulling that
21:25:00dfktok, thanks
21:25:33 Join dreamlayers [0] (~bgjenero@rockbox/developer/dreamlayers)
21:25:47 Quit Topy (Ping timeout: 248 seconds)
21:26:11saratogai guess you also want to make sure the bootloader USB mode is stable before releasing a clipzip bootloader?
21:26:17 Quit Don_Giles (Quit: CGI:IRC)
21:26:46***Saving seen data "./dancer.seen"
21:27:01 Quit robin0800 (Ping timeout: 255 seconds)
21:27:06 Quit Buschel (Quit: ChatZilla 0.9.87 [Firefox 8.0/20111104165243])
21:28:23funmanhm yes but we could still do another at the same time than other amsv2
21:31:06 Quit Sundiver (Ping timeout: 252 seconds)
21:32:38 Join Sundiver [0] (
21:39:43*bluebrother^ finds the tone on the website linked in FS #12451 pretty inappropriate
21:39:44fs-bluebot Some metadata characters are not being displayed properly (bugs, unconfirmed)
21:40:37*dreamlayers agrees with bluebrother
21:41:10saratogai'm just amazed people continue to report problems with a specific file and not think its important to link the file
21:41:45saratogaalthough i think thats the guy who posted in the forums asking how to find the patch tracker, so maybe its not so suprising
21:42:38 Join robin0800 [0] (~robin0800@
21:44:13bluebrother^it looks like "bug! bug" screaming to me without doing ones due diligence (i.e checking if the problem is actually a bug or if there is some setting that affects this) first.
21:44:58[7]well, he claims to have played with fonts at least
21:46:28funman[7]: did r31224 work on nano2g ?
21:47:01[7]you mean r31254?
21:48:25*[7] builds r31224 just in case
21:48:53[7]anyway, r31254 didn't work
21:49:10funman224 added the direction parameter to DEPCTL macro
21:49:11[7]so let's try with r31224...
21:49:15 Join Thra11 [0] (
21:49:15 Quit robin0800 (Ping timeout: 252 seconds)
21:49:57 Quit dfkt (Quit: -= SysReset 2.55=- Sic gorgiamus allos subjectatos nunc.)
21:50:32bluebrother^does ID3V2.3 include information about the encoding used? Or is that v2.4 only?
21:53:37[7]funman: r31224 seems to be even worse
21:53:44[7](enumeration timeout)
21:54:18lebelliumbluebrother: id3v2.3 may include that, at least dbpoweramp is able to tell me exact encoding settings with id3v2.3
21:54:33funman[7]: hm it helps us since it's closer to the last working rev
21:55:24[7]this seems to be a timeout, while r31254 seems more like it responded the wrong way, judging from windows' behavior
21:55:42funman31223 is ok ?
21:56:52funmani just rechecked that it should have no functional changes
21:59:14 Join schanulleke [0] (
22:00:29bluebrother^lebellium: sounds a bit like it's an optional value :o
22:02:13saratoga[7]: does the Nano2G not have car adapter mode or is the manual wrong?
22:02:16 Join Keripo [0] (
22:02:31[7]the last time i looked into that menu it did have it
22:03:00funman[7]: i fear i have f*cked up whe nconverting values to bitfields in r31222
22:04:05bluebrother^hmm, according to Wikipedia only v2.4 allows storing data as utf-8
22:04:07CIA-1New commit by funman (r31255): Clipv2: fix buttons broken by me in r31235 ...
22:04:28bluebrother^which brings me to the conclusion that most likely the encoding is set wrong.
22:04:39 Quit Misan (Remote host closed the connection)
22:04:48funmansetup a 10 pages wordpress site with pictures to prove your point
22:05:33 Join Rob2223 [0] (
22:06:04CIA-1r31255 build result: All green
22:06:41lebelliumbluebrother^ : sorry but what's the link between utf-8 and encoding settings?
22:06:44rarogbluebrother^: is v2.3 UTF16 only?
22:06:50 Quit factor (Ping timeout: 240 seconds)
22:07:34lebelliumrarog: 2.3 may also be ISO-8859
22:07:40bluebrother^rarog: as far as I understand Wikipedia it's local encoding plus optionally UTF-16
22:07:56bluebrother^Rockbox defaults to utf-8, right?
22:08:08 Quit benedikt93 (Quit: Bye ;))
22:08:35 Quit schanulleke (Quit: Lost terminal)
22:08:53 Quit Rob2222 (Ping timeout: 252 seconds)
22:09:13funmanbluebrother^: rbutil built from svn tells me about a new version available
22:10:29dreamlayersunicode_munge() in apps/metadata/id3tags.c decides how to decode tag text, based on the encoding byte at the start
22:11:02bluebrother^funman: yes, I know :)
22:11:44bluebrother^that's because svn thinks the most recent is 1.2.11 while the server has a 1.2.11-1. Only affects the Linux versions.
22:12:50 Join TheLemonMan [0] (
22:20:47 Nick GeekShad1w is now known as GeekShadow (
22:21:21 Quit GeekShadow (Changing host)
22:21:21 Join GeekShadow [0] (~antoine@reactos/tester/GeekShadow)
22:26:44 Join lorenzo92 [0] (~chatzilla@
22:27:18 Quit lorenzo92 (Client Quit)
22:28:10 Quit wodz (Quit: Leaving)
22:28:54froggymanAnyone here have a current SVN build of RaaA for 540x960 (qHD)?
22:31:43 Quit TheLemonMan (Quit: WeeChat 0.3.6)
22:35:26*bluebrother^ prepares pushing the first part of the Rockbox Utility tools' Makefile rework
22:35:36 Quit ThomasAH (Ping timeout: 252 seconds)
22:36:18 Join ThomasAH [0] (
22:36:32 Join factor [0] (~factor@
22:40:54funmanbertrik: what is blocking RDS patch ?
22:43:33funmanwhy can't i see my last reply on,28709.msg186988.html#msg186988 ? i see it in "recent posts"
22:44:36gevaertsfunman: because The System decided it looked like spam
22:45:08 Quit Thra11 (Ping timeout: 252 seconds)
22:45:27CIA-1New commit by dreamlayers (r31256): FS #12378 : Remove various unused code, and comment out some unused code and data for reference or future use.
22:48:03CIA-1r31256 build result: All green
22:48:24 Join Thra11 [0] (
22:51:31funmangevaerts: how can i have it back as non spam (with the 3 links in my message)?
22:51:34funmanah it's back
22:51:44funmani didn't see it in the 'removed posts' forum
22:52:02gevaertsIt gets put in a "special" place
22:52:54*dreamlayers is not sure if he should work on adding some conditionals identified here
22:53:39dreamlayersDo we intend to rely on -ffunction-sections -Wl,−−gc-sections soon?
22:53:57funmani hope so
22:54:04funmanwe could rely on it to reduce the ifdef hell
22:55:13 Join robin0800 [0] (
22:55:58funmansaratoga: any progress on hwcodec fork?
22:56:33funmana few people seem to be running rockbox 2.5.1
22:57:37 Join bzed_ [0] (
22:58:17saratogafunman: no i don't think theres been much discussion
22:59:03funmanbasically it boils down to having amiconn's opinion
22:59:17saratogai was going to send an email summarizing the discussions we've had about it over the years, but haven't gotten around to it
22:59:21*bluebrother^ fires
22:59:39CIA-1New commit by bluebrother (r31257): Initial common Makefile for Rockbox Utility tools / libs. ...
22:59:48 Join lovasoa [0] (~olojkine@
22:59:50bluebrother^hmpf, what's up with gnome-keyring on that box? :(
22:59:53CIA-1New commit by bluebrother (r31258): libucl: allow cross compiling via CROSS as well. ...
23:00:09CIA-1New commit by bluebrother (r31259): mkmpioboot: use common Makefile. ...
23:00:10 Quit bzed (Ping timeout: 252 seconds)
23:00:10 Nick bzed_ is now known as bzed (
23:00:26CIA-1New commit by bluebrother (r31260): Rockbox Utility: pass BUILD_DIR to libs Makefiles. ...
23:00:43funmansaratoga: ok i'll wait for your mail
23:00:50CIA-1New commit by bluebrother (r31261): mkimxboot: use common Makefile. ...
23:00:53saratogadid you want to mail the list?
23:01:55CIA-1r31257 build result: All green
23:02:50 Quit robin0800 (Ping timeout: 240 seconds)
23:03:17JdGordonbertrik: did you try RDS with latest svn?
23:03:49 Quit dreamlayers (Quit: back later)
23:04:13CIA-1r31261 build result: All green
23:04:59 Quit Zarggg (Ping timeout: 252 seconds)
23:11:54 Quit Thra11 (Read error: Operation timed out)
23:24:43CIA-1New commit by bluebrother (r31262): Pass version when building libmktccboot for Rockbox Utility. ...
23:26:41CIA-1r31262 build result: All green
23:26:48***Saving seen data "./dancer.seen"
23:27:08 Join Thra11 [0] (
23:28:04bluebrother^wtf? Why is libmkamsboot not getting built correctly anymore on OS X? I haven't changed that Makefile yet ...
23:30:17 Quit liar (Remote host closed the connection)
23:32:15 Join robin0800 [0] (
23:35:05 Join Strife1989 [0] (~Strife89@
23:36:41 Join jlbiasini [0] (
23:38:34 Quit robin0800 (Ping timeout: 240 seconds)
23:38:43 Quit Strife89 (Ping timeout: 252 seconds)
23:38:55 Join liar [0] (
23:40:09lovasoawhen compiling for android, I get
23:40:11lovasoa/home/olojkine/rockbox/firmware/common/rbpaths.c:78:21: error: macro "paths_init" passed 1 arguments, but takes just 0
23:40:13lovasoa/home/olojkine/rockbox/firmware/common/rbpaths.c:79: error: expected '=', ',', ';', 'asm' or '__attribute__' before '{' token
23:40:15lovasoamake: *** [/home/olojkine/rockbox/android-build/firmware/common/rbpaths.o] Erreur 1
23:40:17lovasoaAny idea why ?
23:45:12 Quit pamaury (Remote host closed the connection)
23:48:21 Part jlbiasini
23:48:58 Quit rarog (Quit: ChatZilla 0.9.87 [SeaMonkey 2.4.1/20111011102430])

Previous day | Next day