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

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

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

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

#rockbox log for 2011-12-18

00:00:56 Quit wodz (Quit: Leaving)
00:06:44 Quit deute (Quit: CGI:IRC (Ping timeout))
00:12:39 Join rarog [0] (
00:13:25 Quit Xerion (Ping timeout: 240 seconds)
00:14:11ack /win 2
00:14:57rarogI managed to register in the wiki, could someone please give me the write permission? Username is "AndrejSinicyn"
00:24:11 Join dhrasmus [0] (
00:25:11gevaertsrarog: done
00:28:15 Quit dhrasmus (Client Quit)
00:33:14 Quit saratoga_ (Quit: Page closed)
00:37:37raroggevaerts: thx
00:38:20 Join liar [0] (
00:40:27 Quit ender` (Quit: A computer program will always do what you tell it to, and seldom what you want it to.)
00:41:05 Quit domonoky (Read error: Connection reset by peer)
00:41:28rarogworks like a charm. :)
00:49:22scanfUSB doesnt work at all on the sansa clip zip
00:49:25scanfseems to be the verdict
00:53:31scanfalso, it never comes back after hooking it to usb
00:53:35scanfso there is no way to charge
00:53:45 Quit froggyman (Ping timeout: 240 seconds)
00:54:36scanfhow can i revert this
00:59:34 Join nick-p [0] (
01:00:36 Quit bertrik (Ping timeout: 240 seconds)
01:00:44CIA-1New commit by 03nick_p (r31348): Hard code the cuesheet offset for double byte encoding, rather than pointless calculation.
01:02:53 Quit rarog (Quit: ChatZilla 0.9.87 [SeaMonkey 2.4.1/20111011102430])
01:03:05CIA-1r31348 build result: All green
01:03:44JdGordonlebellium: its not finished
01:10:13 Quit Misan (Quit: Verlassend)
01:20:47 Quit nick-p (Quit: Leaving)
01:23:17 Quit metaphysis (Ping timeout: 276 seconds)
01:28:22***Saving seen data "./dancer.seen"
01:45:54 Quit keyb_gr (Ping timeout: 244 seconds)
02:03:53*Torne has the new git repo mirroring from svn by the super high tech method of a while loop with a "sleep 5m" in it on his machine.
02:04:10Torneso, yeah. it should be up to date within a short lag
02:04:16Tornefrom now until we switch :)
02:05:38Tornealso wikified the commit policy
02:05:40JdGordonwhy not a svn post-commit hook?
02:06:08Tornebecause that requires dicking about doing it on the svn server
02:06:28Tornethe transitional repo i have here which pulls from svn and pushes to gerrit is all weird and crazy
02:06:49Tornesince it's pulling/pushing from multipe sources and rewriting branches enroute
02:07:36 Quit Strife89 (Ping timeout: 244 seconds)
02:15:15 Quit dfkt (Quit: -= SysReset 2.55=- Sic gorgiamus allos subjectatos nunc.)
02:15:25JdGordonTorne: are you mirroring all the svn release branches? or just trunk?
02:15:48Torneall of them
02:15:57Tornethe new git has everything in it
02:16:09Torne(for a value of everything that i don't want to have to actually explain)
02:16:21JdGordondanm that caveat
02:16:24Tornesome of them are not in refs/heads though so they won't come down by default in a clone
02:16:34Torneyou can change your fetchspec to get them if you want
02:16:46Tornei didn't want to clutter up the autocomplete namespace too much
02:17:19Torneanyway yeah i painstakingly remapped everything to logical sensible git branches and tags
02:17:24 Join nick-p [0] (
02:17:44*JdGordon has a ton of local branches to move across :/
02:19:40TorneYeah, i'm gonna write up a doc on that soon
02:19:52Tornehave you been merging onto them?
02:20:01Tornethere are a number of Entertaining(tm) options for how to "fix" things
02:20:07Tornei need to experiment a bit more
02:20:16Torneif you have a nice repo with some good example cases tha tmight help
02:20:23Tornewhere by nice i mean complicated
02:20:30Tornesimple things are easy, just 3-arg rebase
02:20:43Tornegit rebase −−onto=newgit/master oldgit/master mybranch
02:21:00JdGordonna, pretty sure most of my local branches are dead anyway, so its just trying to figure out what can be nuked and what can't
02:21:09Torneah, ok.
02:22:06JdGordonis the plan still to turn off svn this weekend?
02:22:11JdGordonor did i misread? :)
02:22:20Tornethe time window was kinda vague
02:22:30Tornei have only one or two things left on my checklist
02:22:40Tornethen it's down to availability of zagor probably
02:22:56Torneso yeah, maybe not tomorrow as it turns out there are six people in my house
02:23:11Tornebut ASAP
02:24:26Torneif you want to help go proofread/test/suggest improvements for the wiki pages ;)
02:25:08Torneor yaknow just improve, it is a wiki
02:29:15 Quit kadoban_ (Ping timeout: 252 seconds)
02:29:29 Quit pamaury (Remote host closed the connection)
02:36:34 Join RTN [0] (
02:38:38RTNHey guys
02:40:32RTNAre have question, are 'sound settings' EQ etc disabled for FM radio by default? My device is Fuze v2
02:41:03Tornefm radio plays directly to the output generally
02:41:07Tornethere's not much chance to do DSP to it
02:41:27Tornedepending what device its possible the tone controls/similar might work
02:41:32Tornebut rockbox's software equaliser is not going to
02:41:40Tornewe'd ahve to be recording from radio and then playing back in realtime
02:41:57Tornewe don't support that (and most devices' hardware can't do it anyway)
02:42:34RTNOkay. Thanks, Torne.
02:43:58CIA-1New commit by 03nick_p (r31349): Type 1 text encoding uses BOM to decide byte order, type 2 has no BOM and is BE.
02:45:58CIA-1r31349 build result: All green
03:08:12CIA-1New commit by 03nick_p (r31350): I forgot to flip the true response from memcmp
03:09:58CIA-1r31350 build result: All green
03:10:25nick-pSorry about all the commits
03:16:54 Quit nick-p (Quit: Leaving)
03:28:23***Saving seen data "./dancer.seen"
03:29:56RTNMy Fuzev2 has full duplex support under Rockbox. Is this enough to enable EQ for radio, in theory? Sorry I looked at the ASM SoC documents but they were over my head.
03:30:25RTNnot the end of the world though I could always record a show
03:30:39RTNand listen with eq later
03:31:47 Quit lebellium (Quit: ChatZilla 0.9.87 [Firefox 9.0/20111212185108])
03:32:27 Quit robin0800 (Ping timeout: 240 seconds)
03:35:33 Join robin0800 [0] (~robin0800@
04:04:27 Join Vector850 [0] (
04:04:58 Join Strife1989 [0] (
04:07:08 Quit liar (Ping timeout: 245 seconds)
04:17:52 Join deute [0] (
04:18:18 Quit Strife1989 (Quit: Heading home, memtest overnight.)
04:20:02deutehey guys, are there any plugin sourcecodes available?
04:23:57 Quit TheSeven (Disconnected by services)
04:24:34 Join [7] [0] (~TheSeven@rockbox/developer/TheSeven)
04:26:05 Join nick-p [0] (
04:29:06 Quit nick-p (Client Quit)
04:31:23 Quit Vector850 (Quit: Miranda IM! Smaller, Faster, Easier.
04:31:23 Quit amiconn (Disconnected by services)
04:31:23 Join amiconn_ [0] (quassel@rockbox/developer/amiconn)
04:31:45 Nick amiconn_ is now known as amiconn (quassel@rockbox/developer/amiconn)
04:32:25 Quit pixelma (Disconnected by services)
04:32:27 Join pixelma_ [0] (quassel@rockbox/staff/pixelma)
04:32:29 Nick pixelma_ is now known as pixelma (quassel@rockbox/staff/pixelma)
04:43:08 Quit dys (Ping timeout: 252 seconds)
04:44:04 Join dys [0] (
04:46:41 Quit robin0800 (Ping timeout: 255 seconds)
04:53:44 Join robin0800 [0] (~robin0800@
04:57:57 Quit CaptainKewl (Quit: ( :: NoNameScript 4.22 :: ))
05:03:37 Quit mgue (Ping timeout: 240 seconds)
05:03:53 Quit Az|Out (Quit: 5a 65 75 73 73 2d 73 6f 64 64 69 74 2d 66 75 63 6b 2c 20 49 27 6c 6c 20 62 65 20 62 61 63 6b 2e 2e 2e)
05:03:54 Join erson [0] (~erson@
05:03:54 Quit deute (Quit: CGI:IRC (Ping timeout))
05:04:05 Join bluefoxx [0] (
05:04:35ersonHi, I have a big problem with my cowon j3.anyone can help me to solve the issue?
05:05:18 Join mgue [0] (
05:06:07ersonAnyone here?
05:07:11ersonI am trying to use TCCtool
05:07:45ersonI could use it to upload J3 firmware 2.26 to device
05:07:59ersonBut it didn't react anyway
05:08:51ersonanyone hearing me?
05:09:07ersonplease someone help me
05:17:11 Join Rob2222 [0] (
05:18:31 Quit dv_ (Read error: Connection reset by peer)
05:20:52 Quit Rob2223 (Ping timeout: 252 seconds)
05:21:16 Quit erson (Quit: Leaving)
05:22:19 Join dv_ [0] (
05:28:26***Saving seen data "./dancer.seen"
05:32:26 Quit Topy (Ping timeout: 240 seconds)
05:34:16 Quit ps-auxw (Ping timeout: 252 seconds)
05:34:32 Join Topy [0] (
05:35:59 Join ps-auxw [0] (~arneb@2001:470:c807:0:1532:4e5f:2ad3:4123)
05:46:41 Quit Topy (Ping timeout: 244 seconds)
05:57:24 Join Topy [0] (
06:01:34 Join Zarggg [0] (
06:03:15 Join dreamlayers [0] (
06:03:15 Quit dreamlayers (Changing host)
06:03:15 Join dreamlayers [0] (~bgjenero@rockbox/developer/dreamlayers)
06:24:57 Quit robin0800 (Quit: Leaving)
06:30:54 Join T44 [0] (
06:33:06 Quit Topy (Ping timeout: 240 seconds)
07:28:29***Saving seen data "./dancer.seen"
07:33:26 Quit T44 (Ping timeout: 240 seconds)
07:36:34 Join froggyman [0] (~froggyman@unaffiliated/froggyman)
07:40:22 Join T44 [0] (
07:41:21 Quit froggyman (Quit: People who think they know everything are really annoying to those of us who actually do.)
07:43:29 Join froggyman [0] (~froggyman@
07:43:29 Quit froggyman (Changing host)
07:43:29 Join froggyman [0] (~froggyman@unaffiliated/froggyman)
07:54:11 Join Prodicus [0] (
07:57:16ProdicusFew weeks ago I installed Rockbox on a relative's Clip+, worked like a charm. Tried installing on my own Clip v2−− bricked :( Given the current price of a new clip+ and my previous disappointing experience clumsily trying to solder teeny parts, maybe I just have to write this one off :(
08:00:05ProdicusThe Clip page does say there's always a risk of bricking and "The risk seems particularly high for certain hardware variants of the Clip V2." but it's easy to miss, and the problem isn't mentioned anywhere in the port status, hardware status tables, etc.
08:11:58CIA-1New commit by 03dreamlayers (r31352): Add linker script wildcards for -ffunction-sections and -fdata-sections. ...
08:12:27CIA-1r31352 build result: All green
08:32:18 Join Topy [0] (
08:33:04desowinenabling imagebuffer clock on DM320 just for being able to let DSP interrupt ARM increased the runtime by 40 minutes
08:34:54desowinI should check if there are any side effects, but unfortunately mrobe 500 uses that to speed up LCD transfers (actually on connect my tests showed over 6 times faster screen update when using imagebuffer for that purpose)
08:35:23desowinI need to think of some good synchronization method for this clock between the devices
08:35:28 Quit T44 (Ping timeout: 245 seconds)
08:35:43desowinor maybe some other way for DSP->ARM interrupt
08:37:37 Quit dreamlayers (Quit: goodnight)
08:43:46desowinusing imagebuffer for lcd updates is handy as it can transfer up to 2048 pixels in one dma go (and it handles copying not full lines as well)
09:03:23 Join Horscht [0] (
09:03:23 Quit Horscht (Changing host)
09:03:23 Join Horscht [0] (~Horscht@xbmc/user/horscht)
09:04:43desowincan we have DM320-targets in flyspray as well as Sansa Connect?
09:05:28desowinI have reported my findings as FS #12462 and it affects all DM320 devices
09:05:28fs-bluebot 3TMS320DM320 runtime improvement by disabling ImageBuffer clock (patches, new)
09:05:53 Quit Horschti (Ping timeout: 268 seconds)
09:08:26 Join remlap1 [0] (
09:10:09 Join jordan`` [0] (~gromit@2001:660:3302:2826:225:90ff:fe20:d9a8)
09:10:27 Join TorneWuff [0] (
09:13:49 Join guymann [0] (
09:15:05 Join kevku [0] (x@2001:470:28:773::)
09:15:59 Join nosa [0] (
09:16:14 Quit Torne (Ping timeout: 240 seconds)
09:16:14 Quit remlap (Ping timeout: 240 seconds)
09:16:14 Quit jordan` (Ping timeout: 240 seconds)
09:16:14 Quit pjm0616 (Ping timeout: 240 seconds)
09:16:14 Quit guymann_ (Ping timeout: 240 seconds)
09:16:14 Quit nosa-j (Ping timeout: 240 seconds)
09:16:16 Nick TorneWuff is now known as Torne (
09:16:16 Quit Torne (Changing host)
09:16:16 Join Torne [0] (~torne@rockbox/developer/Torne)
09:16:18 Nick nosa is now known as nosa-j (
09:16:31 Join pjm0616 [0] (~user@
09:18:05 Join Buschel [0] (
09:20:31desowingevaerts: could you please test the runtime impact of FS #12462 on m:robe 500?
09:20:31fs-bluebot 3TMS320DM320 runtime improvement by disabling ImageBuffer clock (patches, new)
09:28:10 Join liar [0] (
09:28:32***Saving seen data "./dancer.seen"
09:36:06 Quit Topy (Ping timeout: 240 seconds)
09:39:10 Join Topy [0] (
09:48:18desowinwait, no, hopefully after reading datasheet it turns out to not be prone to race conditions as the clock enables in IO_CLK_MOD1 and CP_CLKC are ORed!
10:00:30Buschelanyone awake who could test a patch on a MCF5249 an/or MCF5250?
10:04:21desowinI will just wait for someone to confirm I didn't break m:robe 500 playback before commiting the changes
10:04:49 Join ender` [0] (
10:05:26 Quit RTN (Quit: CGI:IRC)
10:07:57desowinBuschel: does iAudio x5 qualify?
10:08:08Buscheldesowin: absolutely
10:08:19desowinso, I can test it
10:09:21Buschelgreat! Can you please run test_codec on with svn and the following patch:
10:09:45BuschelI want to compare the codec speed of pure svn against the patch
10:10:32desowinmy x5 doesn't turn on :( I hope it's just empty battery and it'll turn up soon
10:18:12 Join nick-p [0] (
10:18:26 Quit fs-bluebot (Ping timeout: 240 seconds)
10:18:41 Join James_Hunt [0] (2ea48860@gateway/web/freenode/ip.
10:18:51 Quit bluebrother (Read error: Operation timed out)
10:20:00 Join fs-bluebot [0] (
10:20:12nick-pCould NickPeskett be added to WikiUsersGroup please?
10:21:04 Join bluebrother [0] (
10:21:04 Quit bluebrother (Changing host)
10:21:04 Join bluebrother [0] (~dom@rockbox/developer/bluebrother)
10:30:12 Quit James_Hunt (Ping timeout: 258 seconds)
10:36:21desowincrap, battery measures 0v :( guess it's fried
10:39:05Buschelbad news
10:39:08 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
10:39:26 Quit Topy (Ping timeout: 240 seconds)
10:39:34Buschelanyway, I have created FS #12463. hopefully somebody else can measure the impact
10:39:34fs-bluebot 3Speed up FLAC multichannel (patches, assigned)
10:39:52JdGordonnick-p: done
10:42:15 Join Parsi1 [0] (~Maysam@
10:46:26 Quit Parsi1 (Ping timeout: 240 seconds)
10:48:07desowinyeah, battery warranty ended literally week ago
11:00:49 Join Stummi [0] (~Stummi@rockbox/developer/Stummi)
11:18:41 Quit Prodicus (Quit: ChatZilla 0.9.87 [Firefox 8.0/20111104165243])
11:19:52 Join y4n [0] (y4n@unaffiliated/y4ndexx)
11:28:36***Saving seen data "./dancer.seen"
11:41:12 Join TheLemonMan [0] (
11:44:54 Join n1s [0] (~n1s@rockbox/developer/n1s)
11:58:57 Join lebellium [0] (
12:00:22 Nick Stummi is now known as Stummi|afk (~Stummi@rockbox/developer/Stummi)
12:04:13gevaertsdesowin: I'll try to have a go somewhere today. Probably not in the first several hours though
12:04:40gevaertsI guess you need the usual "does it still work" + "how much longer does it run"?
12:05:04 Nick Stummi|afk is now known as Stummi (~Stummi@rockbox/developer/Stummi)
12:05:06desowingevaerts: yes
12:05:32desowinbut the "does it still work" is more important
12:07:08gevaerts"It now runs three times as long, at the slight cost of having no audio or screen" :)
12:10:34 Join Topy44 [0] (
12:19:04 Quit saratoga (Ping timeout: 258 seconds)
12:28:12 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
12:31:57 Join ruskie [0] (ruskie@sourcemage/mage/ruskie)
12:32:25TheLemonManpamaury: ping
12:32:37 Join dfkt [0] (dfkt@unaffiliated/dfkt)
12:34:45TheLemonMandoes the imx233 memory layout supports booting from recovery mode ?
12:37:30pamauryyou mean the hard memory layout in recovery mode ?
12:38:03pamauryor the soft memory layout used by rockbox ?
12:39:08TheLemonManif the latter works on the former, thats the question :)
12:39:44gevaertsdesowin: it still boots and plays (I did a very short test). The runtime tests will have to wait for the battery to charge
12:40:02pamauryyes, the recovery mode is no different than any other boot mode so rockbox should boot out of recovery mode
12:41:38pamauryTheLemonMan: however, a bit of warning, if you're working on stmp3770, the physical memory layout is different so you'll need to change a few things in the linker script
12:42:20TheLemonMandang, then i think ive found why its just resetting when i upload the bootloader
12:42:41pamaurydoes the stmp3770 have a mmu ?
12:42:47TheLemonMan(i'm using your old elf2sb tool anyway, too lazy to modify mkimxboot atm, but that shouldnt matter
12:43:18TheLemonMani think it does, cant be 100% sure as there are no datasheets around
12:43:55pamaurywait a minute, I can upload a modified linker script if you want, no guaranty though :)
12:44:29TheLemonMancool, ill search for some docs in the meanwhile
12:44:46 Quit Topy44 (Ping timeout: 240 seconds)
12:45:56TheLemonManand yes, it has a MMU
12:46:53pamaurydo you know how much ram it has ?
12:48:27TheLemonManhalf of the fuze plus, 32mb (again, im not 100% sure, i found that value by searching into the kernel tree)
12:48:57pamaurydo you know where is it mapped ?
12:49:10pamauryit must be somwhere in the kernel tree
12:52:15TheLemonManits hard to tell heh, its such a big codebase
12:52:47 Join kadoban_ [0] (
12:53:25pamaurytry to find the linker script, or have a look at the headers of the platform. By the way, do you still have the script and the diff between the registers ?
12:54:32TheLemonMansure, want me to send them again ?
12:55:27pamauryyes, thanks
12:58:17TheLemonMan and
13:00:10TheLemonManthe kernel crt0 also disables the watchdog, might be worth a try too
13:01:41pamauryTheLemonMan: do you have a sb file from samsung ? (
13:02:07pamauryif you run sbtoelf to it, you can have a look at the adresses and see where is the memory
13:04:14 Join Topy44 [0] (
13:04:27TheLemonManthe nand bootloader is loaded in the ocram mapped at 0x0
13:09:25pamaurythat doesn't help :(
13:09:54pamauryyou should really find the memory setup in the kernel I'm afraid
13:21:32pamauryTheLemonMan: did you open the device ? what are the chips inside ?
13:21:52 Join wodz [0] (
13:22:32TheLemonManpamaury: i didnt as it my only player and i need it heh, but lebellium did, you can ask him for part numbers
13:22:48TheLemonMando have anyway the lcd/nand/fm tuner models
13:24:34lebelliumhere is my disassembly guide
13:28:03wodz16Mx16 ddr sram
13:28:38***Saving seen data "./dancer.seen"
13:31:31 Join fyre^OS [0] (
13:33:07wodzand standard 8bit wide nand chip
13:34:16 Quit fyrestorm (Ping timeout: 244 seconds)
13:42:16 Join lorenzo92 [0] (
13:43:21 Join T44 [0] (
13:43:46 Quit Topy44 (Ping timeout: 240 seconds)
13:45:04 Quit lorenzo92 (Client Quit)
13:45:07pamauryTheLemonMan: the sdram is at the same address as in the imx233 then, only the size is different, and the iram is bigger
13:45:23CIA-1r31354 build result: All green
13:45:32 Join honschu [0] (
13:45:53 Join domonoky [0] (~Domonoky@rockbox/developer/domonoky)
13:47:21 Quit kevku (Quit: KVIrc 4.0.4 Insomnia
13:48:55 Quit honschu_ (Ping timeout: 268 seconds)
13:51:43 Quit nick-p (Quit: Leaving)
13:52:57pamauryTheLemonMan: if you want to have the bootloader working, you should change the memory size, that might explain the reset too
13:54:46wodzpamaury: don't you have to init mem controller first when using recovery mode?
13:55:47pamauryin any mode you have to, for the fuze+ I just include in the sb file some boot sections of the OF which init the memory
13:56:03wodzmaybe that is the problem also?
13:57:14pamaurysurely if you write to non-init ram it will reset too :)
14:11:06 Quit Buschel (Ping timeout: 240 seconds)
14:49:14 Quit T44 (Ping timeout: 252 seconds)
14:49:43 Join T44 [0] (
14:51:43 Join benedikt93 [0] (
14:51:51 Quit benedikt93 (Changing host)
14:51:51 Join benedikt93 [0] (~benedikt9@unaffiliated/benedikt93)
14:52:05 Join Buschel [0] (
15:10:46 Quit wodz (Quit: Leaving)
15:12:24 Join keyb_gr [0] (
15:27:46SynrGdo .fmr files allow comment lines?
15:28:11SynrGI would like to add a comment header like:
15:28:41***Saving seen data "./dancer.seen"
15:29:00SynrG# Halifax.fmr by Ben Armstrong <> 2011/12/18
15:29:15lebelliumyes they do
15:29:23SynrG# From:
15:29:31SynrGand maybe also:
15:30:05lebelliumah no sorry I thought you spoke about fms files
15:30:07SynrG# Callsign Branding (Region) -(format)
15:30:09lebelliumfmr I don't know
15:30:20dfktjust try it ;)
15:30:24SynrGya. i will :)
15:30:40SynrGbut if it works 'by accident' ... it might break something later on
15:31:03SynrGso i really need to know if it is formally supported.
15:36:53n1slook at the parser code?
15:43:01TheLemonManpamaury: i fiddled a bit with ram values but got no luck, what could be wrong ?
15:43:23pamaurylots of things :)
15:44:19pamaurythe mmu setup, the stack, some register, ...
15:44:40TheLemonMangonna love blind troubleshooting
15:45:08pamauryif the bootloader is small enough, I would advise that you tweak the linker file to put everything in iram (since it's rather big)
15:45:23pamauryand that you disable the mmu
15:45:53pamauryunfortunately that will require to modify crt0 I think
15:47:07 Quit T44 (Ping timeout: 240 seconds)
15:47:11pamaurydid you already succeed in executing something on the device ? Or is it when you try to use the rockbox code that it fails ?
15:47:49TheLemonMani already had stuff running on it and some drivers too
15:48:10bluebrothersome kind of radio presets website similar to would be nice.
15:48:11scanfhow can I revert to OF
15:48:18bluebrotherby reverting
15:48:33 Join T44 [0] (
15:48:36scanfi know how to do it on my ipod 5.5, but im not sure how to on my clip zip
15:48:49*scanf is thinking about exchanging it for a clip+
15:48:49pamauryTheLemonMan: and then, you put everything in iram ?
15:48:52bluebrotherso you're talking about dualboot on the clip zip or what?
15:49:48scanfno, rockbox on the clip zip just doesnt really seem usable, so i want to revert somehow and exchange it for a wellsupported clip+
15:50:30TheLemonManyep i had everything in ocram and the bss in sdram
15:50:34lebelliumrockbox on clip zip not usable?!
15:50:41pamauryTheLemonMan: so the sdram was working ?
15:51:19scanflebellium: maybe if you only charge via USB and transfer files to SDHC using an external reader
15:51:25bluebrotheraccording to the fine manual you can update with an unmodified Sandisk firmware.
15:51:28scanfusb crashes it and it cant read files
15:51:47scanfit wont even connect to my PC without freezing
15:51:49*bluebrother considers reading a book :P
15:51:54TheLemonManit was just the bss that was there and didnt check it thorougly
15:52:55lebelliumscanf: I own both Zip and Clip+ and the clip+ doesn't seem more stable. There are the same USB connection issues, it's not perfectly implemented yet
15:53:15pamauryok, then first try to put some code there, to make sure it works. Then the best thing you can do is probably to disable the mmu in rockbox. In fact what you should do is something like modifying crt0.S to hang at some places and see where it resets (I know this is painful)
15:54:47TheLemonMangood, ill use the keypad leds to ease this pain a bit
15:55:19pamauryalso disable caches (this is do at the beginning of crt0, just don't enable it back, so wipe out the calls to memory_init if you're using imx233/crt0.S, disable the dramcopy too
15:55:39pamauryoh, a major point: how do you include rockbox in the sb file ? do you include rockbox.elf directly ?
15:56:34TheLemonMani reused my old config file and the elf2sb tool to include a part of the samsung firmware and bootloader.elf from the build directory
15:56:48pamauryok, are you using the imx233.h from rockbox ?
15:57:51pamaurythen replace CACHED_DRAM_ADDR and BUFFERED_DRAM_ADDR by 0x40000000, since you don't want to use the mmu first
15:58:27pamaurydon't forget to put the right value for MEMORYSIZE in tools/configure
16:00:13pamaurystill you are lucky: you have a led to debug ! ;)
16:01:13TheLemonMani fear it wont work heh, luck isnt by my side lately
16:01:30TheLemonManalso, no news from low_light ?
16:03:09SynrGhumm. so i've confirmed the comment lines are ignored, but ...
16:03:28SynrGi can't actually include as much information as I have. yes, the long lines are scrolled, but ...
16:03:38SynrGthey are truncated at a certain # of characters. can this be expanded?
16:04:48SynrGunless i miscounted, it's 27. that's an unusual length.
16:04:57SynrG(i'm talking about the .fmr format)
16:07:47 Quit remlap1 (Quit: Leaving.)
16:07:48scanflebellium: shucks, for some reason i thought the sansas were perfectly supported
16:07:57 Join remlap [0] (
16:08:41lebelliumscanf: well you can still use the OF USB connection, personally I don't think it's a big deal. Everything else works fine in RB :)
16:10:12scanfhow do you do that
16:13:08scanfis there a button you hold or something
16:13:40 Quit Buschel (Ping timeout: 240 seconds)
16:14:08 Join Buschel [0] (
16:20:27 Join stoffel [0] (
16:25:15 Quit keyb_gr (Ping timeout: 252 seconds)
16:26:36 Join keyb_gr [0] (
16:29:09lebelliumscanf: you can press and hold volume down to boot the OF. Use that for the USB connection
16:31:33 Quit stoffel (Remote host closed the connection)
16:40:12*pamaury grumbles against the imx233 sdk once more, it's just crazy to reverse engineer
16:49:34 Join Topy [0] (
16:49:43scanflebellium: cool thanks
16:49:58scanfhmm even in that i get a million errors when trying to mount the sd
16:50:17scanfit only shows up as /dev/sde it doesnt show up as an actal partititon like /dev/sde1
16:50:45scanfand when i remove the card and put it in an MMC reader it shows p as /dev/mmcblk0p1 and seems to work fine
16:50:56scanf(formatted fat32)
16:52:48 Quit T44 (Ping timeout: 276 seconds)
16:56:48lebelliumscanf : I don't use Linux so I can't help you :/ but if that happens with the OF, that's Sandisk's fault, not RB fault.
17:03:02pamauryomg, how can one write so complicated code to do something so simple :(
17:28:42***Saving seen data "./dancer.seen"
17:30:45 Quit kugel (Ping timeout: 248 seconds)
17:33:34 Quit Buschel (Quit: ChatZilla 0.9.87 [Firefox 8.0/20111104165243])
17:34:56 Quit kadoban_ (Ping timeout: 252 seconds)
17:35:22 Join dreamlayers [0] (
17:35:22 Quit dreamlayers (Changing host)
17:35:22 Join dreamlayers [0] (~bgjenero@rockbox/developer/dreamlayers)
17:41:36TheLemonManpamaury: disabling the mmu made the trick :D
17:41:50pamaurynice :)
17:42:04pamaurythen perhaps it doesn't have a mmu, which stmp37xx is it ?
17:43:07 Join TomColler [0] (
17:43:35TheLemonMan3750, it should have it tho
17:44:43pamauryyeah, then something else was
17:45:38pamaurythat's also probabyl because of the way the binary is packed, there is a great different between loading bootloader.elf and bootloader.bin, crt0.S makes some assumptions
17:49:41 Join T44 [0] (
17:49:46 Quit Topy (Ping timeout: 240 seconds)
17:51:19 Join Topy [0] (
17:54:32 Part TomColler
17:55:07 Quit T44 (Ping timeout: 252 seconds)
17:56:25pamauryI have the screen of the creative zen x-fi2 nearly working
17:57:17gevaertsdesowin: the battery_bench keymap seems to be broken on touchscreen targets...
17:59:12TheLemonManso you already started work on the x-fi2? great
18:00:19desowingevaerts: just open apps/plugins/battery_bench.c and change line 519 to button = BATTERY_ON
18:00:28desowinthat's a quick hack I can think of
18:00:32SynrGsomeone authorize me for the wiki please, so I can submit my Halifax.fmr?
18:00:35pamauryjust tried to make the lcd work, since it's the same SoC, based on the same sdk, I can reuse my code which I used for the fuze+ (which is not rockbox)
18:01:05SynrGi just filled out the registration, then was told that the change was suspicious, then was denied and taken to a page to ask here
18:02:41SynrGmy wiki username would be BenArmstrong
18:04:04SynrGincidentally, here's a copy of the file in question:
18:05:27desowinSynrG: you should be able to edit wiki now
18:08:23gevaertsdesowin: right. I'll do that later though. Cooking dinner now :)
18:09:50TheLemonManok, the code seems to hang at kernel_init
18:11:23 Join chkktri [0] (~user@unaffiliated/chkktri)
18:11:45TheLemonManthat i just discovered it just bootstraps the thread system
18:14:18SynrGdesowin: it's not obvious to me how to update an attached file
18:15:03SynrGclicking on a file takes me directly to a page to view it, not to edit it
18:15:15SynrGthere are attachments at the bottom of this page, but only one is listed:
18:15:38dreamlayersbuflib_compact() is still broken, but the fix for this seems simple (FS #12441). It's good that FS #12409 didn't end up in 3.10.
18:15:39fs-bluebot 3Sansa Clip+: a segmentation fault occurs with sim at r31101 (bugs, unconfirmed)
18:15:39fs-bluebot 3problems with buflib_compact() (bugs, closed)
18:16:38desowinSynrG: click the manage button and you can change the file
18:17:15SynrGthere is no manage button for my city ...
18:17:23desowinSynrG: you can find it for every file after clicking Attach
18:17:24SynrGi can reverse-engineer the correct URL from caracas.fmr ...
18:17:41SynrGi mean, there is already an existing Halifax.fmr ...
18:17:50SynrGso i would have expected to find it attached at the bottom
18:17:52SynrGit is not
18:17:58desowinSynrG: yes
18:18:08desowinSynrG: it centainly is there
18:18:17SynrGah, got it
18:18:23SynrGgot to click "Attach" first. i see
18:18:31SynrGwhy is caracas different? it's misleading
18:18:40desowinbecause it's not hidden
18:19:08SynrGgot it
18:19:11desowinSynrG: I've just changed its properties
18:23:27SynrGdesowin: when viewing this file, the accents look strange. is that my browser just applying an incorrect default encoding, or was it corrupted upon upload?
18:24:24SynrGmy guess is the former. looks like the original file was Unicode
18:24:56*SynrG tries to convince Chromium this is Unicode ..
18:25:26SynrGthere. looks better (was defaulting to ISO Latin-1)
18:26:28SynrGone of the consequences of the 27 character limit is that any line with accents has one less character space per 2-character unicode character
18:29:21SynrGhah, my ruler is broken
18:30:53*desowin has new suspection about what can help improve runtime on Connect
18:34:20 Join mortalis [0] (~mortalis@
18:45:53 Join lorenzo92 [0] (
18:46:11 Quit lorenzo92 (Client Quit)
18:53:29 Quit Topy (Ping timeout: 252 seconds)
18:54:32 Join Topy [0] (
18:57:32 Join limescout [0] (
18:58:08 Part limescout
18:58:12 Join kugel [0] (~kugel@rockbox/developer/kugel)
18:59:32pamauryI have a working creative zen x-fi2 screen:
19:00:20pamaurythat was nearly too easy ;)
19:04:44 Join kadoban_ [0] (
19:05:36 Quit Stummi (Read error: Connection reset by peer)
19:08:54kugelpamaury: good luck with the touchscreen interface :P
19:09:22 Join Stummi [0] (~Stummi@rockbox/developer/Stummi)
19:09:25pamaurykugel: please avoid these kind of subjects, can you ? ;)
19:11:55pamauryso what is the state of the touchscreen interface ?
19:13:52kugeldepends if there's a cabbie for your resolution
19:14:05kugelit works good generally
19:14:38kugelwith lots of quirks
19:14:48pamauryno there is not :( it's 400x240
19:14:48 Join TomColler [0] (
19:15:09 Quit Horscht (Quit: Verlassend)
19:17:37kugelpamaury: then it'll work not as good because the fallback wps uses the grid mode
19:18:10kugelbut grid mode is probably default anyway for new targets (can't remember if the selection uses a black or whitelist=
19:19:31 Quit dreamlayers (Quit: Leaving)
19:19:57SynrGok, still after upgrading my son's fuze v2 to 3.10, no FM ...
19:20:23SynrGhowever, i can't reproduce the "select FM radio and the system locks up" issue because FM does not appear in the menu (which is an old, known issue apparently)
19:20:51SynrGi don't even know what happened to make it decide to show FM in the menu one day. we've tried everything to make it come back, to no avail :p
19:21:26 Quit Topy (Read error: Connection reset by peer)
19:22:50mc2739SynrG: fuze v2 is known to have problems with the radio chip going bad
19:24:00SynrGit seems to still work on OF though
19:24:04SynrGso what's up with that?
19:26:27mc2739There are two different fm radio chips available. Maybe there is a problem in the detection of the chip. Can you try installing an older version of Rockbox to see if it is some recent change causing the problem?
19:27:07 Quit mgue (Quit: leaving)
19:27:26 Join metaphysis [0] (
19:27:46SynrGhmm. i'm back into the OF and although i get an FM display, i am tuned to a station i know should be there and get no sound :(
19:27:49 Quit mortalis (Read error: Operation timed out)
19:28:31SynrGjust a tiny ticking sound as i spin the wheel to turn the volume up and down
19:28:44***Saving seen data "./dancer.seen"
19:29:53metaphysispamaury: is the VERSION.SDK file in the fuze+'s root typical off imx? (like sb file are) because if so we could use it for autodetection in rockbox utility!
19:30:28 Quit metaphysis (Client Quit)
19:30:44 Join jlbiasini [0] (
19:30:48pamauryno idea
19:31:03pamauryI don't think so
19:31:47jlbiasiniI tought there where people working on other imx player?
19:31:55 Join Topy [0] (
19:32:00pamauryfor example the creative zen x-fi2 doesn't have it, neither the creative zen style m100/m300
19:32:47jlbiasiniwell at least I could use it for the fuze then, assuming people won't delete it...
19:32:48mc2739jlbiasini: other sansa players have that file, although the contents may be different
19:32:49SynrGmc2739: well, based on this, i think it's not worth trying the downgrade. sounds like the chip has indeed gone bad ...
19:33:02SynrGmy son indicates the unit has been dropped a lot. maybe a trace broken in there or something
19:33:07mc2739SynrG: yes, I would agree
19:33:41pamaurylet me see if there are other files
19:33:46SynrGso far, i have found the clip+ much more robust than the fuze
19:33:53jlbiasinimc27739: this is the content that is interessant: it give the model's name and firmware version!
19:34:07jlbiasiniideal for autodetection...
19:34:10SynrGi've had fuze display connections go flaky ... wheel become unresponsive ...
19:34:14SynrGand now this FM chip thing
19:34:45SynrGthe household has had three fuzes (two v1, one v2). only one of the v1 units is still operational
19:34:52SynrGand the v2 now has dead FM
19:35:55mc2739jlbiasini: I don't agree that an OF file would be good for autodetection. That file can be deleted while in Rockbox and will not be recreated unless the OF is used.
19:35:59pamauryjlbiasini: at least the fuze+ version.sdk is reliable because there is "product: fuzep" in it
19:38:23jlbiasinipamaury: for now I've got something working like that: it takes bootloader-fuzeplus.sansa in the folder it is launched from, ask for OF, patch it and copy the result to the target directory. It do so with general info and nothing specific to fuzeplus so it should work on other imx target to. I still need to implement error descritption and a check that the firmware file given is consequent with the player
19:38:39jlbiasiniyes this is why I was asking
19:38:53jlbiasinifor autodetection it is perfect
19:39:48pamauryjlbiasini: I can modify mkimxboot to force a check of the firmware file, so you can specify that for example you are patching a fuze+ image
19:40:44jlbiasiniI saw that imxboot is already saying what version the file is
19:40:58jlbiasiniI guess I could do it
19:41:39jlbiasinianyway I have to confirm that normal usb mode is totally broken with sd on my fuze+ :(
19:43:23pamauryonly with sd inserted ? it works flawlessly on mine :-/
19:43:59 Join Horscht [0] (~Horscht@xbmc/user/horscht)
19:44:32jlbiasiniyes only with sd in, else no problem
19:45:19pamauryeven the internal storage breaks or just the sd don't show up ?
19:46:11jlbiasiniwait i've have to check it briefly again
19:47:23jlbiasinidata abort :(
19:47:36jlbiasinibut not all the time
19:47:47TheLemonMani fear some sections of code gets messed somehow as it locks here and there
19:48:58jlbiasinithe strange thing is that the screen get corrupted always the same part of it
19:49:14pamaurythat's with svn head ?
19:50:10jlbiasiniand then it's data abord or usb hid mode work but stay after deconnection (and no storage acces anyway
19:50:19jlbiasinisvn head?
19:50:41pamaurylatest svn revision ? when did it start ?
19:51:13jlbiasinias you implemented sd
19:51:24pamauryin bootloader or in main firmware or both ?
19:51:52jlbiasiniyes I did rebuild the bootloader also
19:52:29pamauryok, I'll update to check, I'm not running the latest bootloader
19:52:30jlbiasiniyou really don't get anything similar? that's strange
19:54:51pamaurythere have been recent changes by other people, perhaps they broke something
19:57:09jlbiasiniI'm also reinstalling everything just in case...
19:57:45 Quit Topy (Ping timeout: 240 seconds)
19:58:11 Quit nosa-j (Read error: Connection reset by peer)
19:59:59funmananythingbutipod forum is clear: amsv2 usb is still quite buggy
20:00:07 Join nosa-j [0] (
20:00:13 Join Topy [0] (
20:02:36pamauryjlbiasini: I reinstalled and it works
20:03:49jlbiasiniI'm reinstalling and I forget I've still have the unwarminder patch... could it be related?
20:04:25pamaurydon't think so
20:04:52jlbiasiniI take it out anyway
20:07:30*bluebrother is almost finished in figuring the Visual Studio stuff for Rockbox Utility :)
20:07:56bluebrotherMinGW can be quite annoying if you call Makefiles that assume a Linux-like system
20:07:59TheLemonManpamaury: i tracked down the crt0 problem in the dram copy loop
20:08:14pamauryTheLemonMan: you don't need the dram copy loop
20:08:59TheLemonManim trying to use the mmu too as i fear theres some memory corruption ongoing
20:09:07 Quit Zarggg (Ping timeout: 252 seconds)
20:11:07 Join Zarggg [0] (
20:15:16pamauryTheLemonMan: if you can wait a bit, I'll explain you what the different steps of crt0 are useful for and the assumptions
20:16:07TheLemonMansure, ill st
20:16:18TheLemonMan*stay here for another couple of hours
20:16:49TheLemonManhad to study but this port took all the time >.<
20:22:40 Part jlbiasini
20:23:00 Join wodz [0] (
20:23:36wodzJdGordon: Any idea what can cause FS #12464?
20:23:37fs-bluebot 3can't load user theme in MPIO HD300 (bugs, unconfirmed)
20:23:56wodzI can confirm the issue
20:24:54 Join jlbiasini [0] (
20:26:14jlbiasinipamaury: I'm sorry this is working now! I must have forget something...
20:28:02wodzhmm whole statusbar is f* up - battery indicator doesn't update, disk icon is corrupted, time is truncated
20:29:25pamauryTheLemonMan: ok, so basically you have three things: 1) bootloader.bin vs bootloader.elf 2) dram relocation, 3) mmu
20:29:25pamaury1) the bootloader/firmware is made of several pieces: some in iram, some in dram but for simplicity, we want to load everything at once; to do so, we pack everything in one binary file: bootloader.bin, so then crt0.S has to put back the right pieces at the right places: that's the job of iramcopy
20:29:25pamaury2) because the binary can be either rolo'ed or loaded by the bootloader, we can't make any assumption on the address at which the binary is loaded; yet the binary was linked to run at some address so we if it was not loaded at the linking address, the binary copy itself at the right location: that's the job of dramcopy
20:29:25DBUGEnqueued KICK pamaury
20:29:25pamaury3) the mmu brings further complexity because the binary is normally running at a virtual, cached address, but the part which is enabling the virtual memory must have the same address before and after, and its stack too. Thus I use the physical address of the stack (stackend_phys), but after that, we jump back to the real address, that's why there is the strange jump to remap
20:29:51kugelwodz: perhaps FS #12441?
20:29:52fs-bluebot 3Sansa Clip+: a segmentation fault occurs with sim at r31101 (bugs, unconfirmed)
20:30:34pamaurysince you are building the sb file with rockbox.elf, you don't need the iramcopy, and you also don't need the dram copy, if you disable the mmu, you also don't need the memory_init and remap thing
20:31:00pamauryrunning without the mmu is simpler and I'm confident that the vast majority of the code can work with and without
20:31:36pamauryyou also don't need the bss init since the sb loader does it for you
20:32:04wodzkugel: which one of the linked patches are supposed to be proper fix?
20:32:35pamauryand I also advise that you disable HAVE_INIT_ATTR for a start, it can only make things worse
20:32:43TheLemonMani wouldnt mind running w/o mmu but seems some code gets messed somehwere
20:33:22pamaurythat's unlikely
20:33:41TheLemonManthen cant explain why kernel_init never returns
20:33:57pamaurycan you pastebin you current crt0.S ?
20:34:11pamauryand the linker file if you modified it
20:34:18TheLemonMani traced what kernel_init does and it just calls thread_init, and in that theres a comment that says there might be problems in case the .bss wasnt cleared
20:34:53pamauryhum, actually, you might need the iram copy thing, depending on the linker script
20:34:59pamaurysame for bss
20:35:44TheLemonManthe linkerscript is the imx233 one, wasnt unsure about how to modify the config.h for the cpu to reflect the layout differences
20:36:18pamauryok, then if you kept the linker script, you need to do the iram copy and bss init
20:37:18pamaurywhen you produce the sb file, it should produce some debug output (or perhaps you need to enable it don't remember) of the commands, can you pastebin it ?
20:39:15pamauryI know this is a bit tricky :) and that part is not really documented
20:39:22TheLemonMan here it is
20:40:14TheLemonManmy crt0 were never more complex than some stack setup and a jump to the main function
20:40:28TheLemonManso im a bit lost here heh
20:40:41 Quit jlbiasini (Remote host closed the connection)
20:41:13pamauryhmm, seems elftosb is keeping the sections in iram, that's unexpected, I don't really know what will happen :-/
20:41:53pamauryI guess that's because I made it use the virtual address and not the physical one
20:42:06 Join jlbiasini [0] (
20:42:41pamaurydid you modify imx233.h as I told you ?
20:42:53pamaurythe addresses look really suspicious
20:43:06pamauryexcept if you enabled the mmu
20:43:21pamaurybut even then
20:43:54pamauryarf, I understand, I fear you'll have to stop building using bootloader.elf and use bootloader.bin
20:45:54TheLemonManawesome, time to hack mkimxboot \o/
20:45:57jlbiasinipamaury: WTF! this is very strange: the but I had is theme related! It happens only with lebellium Samsung-like!!
20:46:21TheLemonManalso, how should i modify the memory config ?
20:46:38jlbiasinivery interesting to read you both!
20:46:49pamauryit will be just simpler if you do it the same way as we do, so revert everything I told you, keep everything in crt0.S and it you elftosb command, instead of doing "load rockbox_elf" you do load "rockbox_bin > 0x40000000;" you don't need to modify mkimxboot know
20:47:05pamauryjlbiasini: that's weird indeed
20:47:50pamauryTheLemonMan: also revert imx233.h to the svn value, since you are using the virtual memory, you need to keep the different address for the cached, uncached and buffered segments, otherwise the world might collapse ;)
20:49:31TheLemonMani suppose i also need a "jump 0x40000000" instead of "call recovery"
20:50:19TheLemonMani hope it wont jump in a black hole
20:52:21pamauryI hope too :) Hopefully it will just work and the time I spent once to get it right will be useful for you
20:55:31TheLemonManaww no luck, got reset again
20:56:12 Quit y4n (Quit: 6,000,000 ways to die — choose one.)
20:56:21pamaurycan you tell where ?
20:56:45 Quit Topy (Ping timeout: 240 seconds)
20:57:16pamauryin the elftosb script, you replace rockbox.elf by rockbox.bin in the source section ?
20:57:35TheLemonMans/rockbox/bootloader/ yep
20:57:58 Join Topy [0] (
20:58:11pamauryhmm, to sum up, you are usually the svn imx233.h, linker script and crt0.S ?
20:59:28TheLemonManhrm basically
21:00:35pamaurythen if you changed the memory size to the right value, I can't see what could cause a reset :-/ Can you pastebin the output of elftosb ? otherwise try to loop at various places of crt0.S to see where it crashes
21:03:24TheLemonManhrm, if you mean the memory size in configure then i did
21:05:48jlbiasinilebellium: you were right your theme isn't fabulous! ;)
21:06:06lebelliumjlbiasini : ?
21:06:12jlbiasiniit is causing bug
21:06:23jlbiasiniat least on the fuze+
21:06:29lebelliumI guess you see it the wrong way
21:06:42lebelliummy theme respects all the WPS tags
21:06:45 Quit bertrik (Ping timeout: 240 seconds)
21:06:46jlbiasinithis is very weird
21:06:57jlbiasiniit bug in usb mode
21:07:08jlbiasiniwith other theme no problem
21:07:12pamauryI'm not sure if it's related to the fuze+ then, it would be interested to test it on other, I think the gigabeat has the same screen size right ?
21:07:37lebelliumOn the Samsung R0 there is no USB mode yet so I can't try
21:07:53lebelliumbut definitely the bug comes from RB or your Fuze build
21:08:25lebelliumI mean if it respects all the WPS tags, you can't blame my theme :p
21:08:37pamaurythen we blame JdGordon :)
21:09:51TheLemonManpamaury: hah gotcha, even lcd is working :D
21:09:57jlbiasinithe interessant part is there is some screen corruption of the size of the hide_background.bmp and then it just hang or go into data abort
21:09:59TheLemonManguess what was the error!
21:10:12pamauryyou uploaded the wrong file ?
21:10:19pamaurygood work anyway \o/
21:10:53wodzTheLemonMan, congrats - I guess new port approaches :-)
21:11:03TheLemonManas it had no info about the entrypoint it was always executing keypad_led_on instead of main
21:11:11jlbiasiniyes nice job!
21:11:17CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
21:11:17*[7] wonders whether we should add SMART support to our USB interface
21:11:22[7]apparently there is a way to access SMART data through some USB bridges...
21:11:38pamaury[7]: using scsi passthrough ?
21:11:43kugelwodz: the last one
21:11:52TheLemonManthanks for the help and support \o/
21:11:58 Quit Horscht (Quit: Verlassend)
21:12:05wodzkugel: I can try if this fixes the issue
21:12:07[7]hm, no idea how that works
21:12:09pamauryTheLemonMan: I can't wait to see your patch :)
21:12:22[7]but SMART isn't SCSI, right?
21:12:25pamaury[7]: how do you know some bridge do then ?
21:12:40pamauryI don't remember but I don't think so
21:12:47[7]it's something that only ATA has, so I have no idea what that interface looks like
21:13:04[7]pamaury: I have one sitting here on my desk
21:13:19[7]and hdtune just accesses the smart data of that drive somehow
21:13:29pamauryhave a look at the code :)
21:13:53[7]and that data is consistent with the values that I get when hooking that drive up using SATA
21:14:04[7]so it seems to actually work and not just read some garbage
21:14:38[7]would be nice if we could diagnose iPod HDDs that way .)
21:15:48pamaury[7]: then you know what do to ;)
21:16:09pamauryit should be pretty easy once you figured out the interface used
21:16:17jlbiasinilebellium: I confirm! If I delete hide_background.bmp from the theme it doesn't bug anymore! And it is awfull :)
21:16:27lebelliumlol strange bug
21:16:30[7]it even works with a stoneage 12GB PATA laptop drive hooked up to it
21:16:30*[7] fires up the usb analyzer
21:16:59jlbiasiniperhraps related to the format of the file??
21:17:06n1s[7]: i think Torne was looking at SMART possibly creating a plugin that would read it
21:17:06lebelliumit's probably because I use conditional main menu viewport and for the USB mode it doesn't like that
21:18:01lebelliumor maybe because it's a 32bit pic
21:18:04jlbiasiniwell if you have an idea to correct it I can test it
21:18:06lebelliumconvert it into 24bit
21:18:21 Join Horscht [0] (
21:18:21 Quit Horscht (Changing host)
21:18:21 Join Horscht [0] (~Horscht@xbmc/user/horscht)
21:18:29jlbiasinieuh... how do i do that?
21:18:37jlbiasinigimp and then?
21:18:50Tornen1s: yeah, I have some very basic smart code
21:18:52jlbiasiniah! I know then : google
21:18:54Torneit needs ot be more clever to be of any use
21:18:57Tornewill get around to it :)
21:19:00pamaury[7]: isn't hdtune open source ?
21:19:22wodzkugel: The patch seems to improve the behaviour but I still get malfunctioning statusbar just after reboot. Loading themes does work though and fixes statusbar
21:19:44[7]don't think so, they're providing an evaluation version, so probably net
21:19:53n1sTorne: cool, just hoping you and [7] don't duplicate effort
21:20:02lebelliumjlbiasini : lebellium/R0/Rockbox/hide_background.bmp">
21:20:06jlbiasinilebellium: anyway why would it work on other menu as as 32 bits and not in usb??
21:20:10scanfso if i have rockbox working and i delete .rockbox it should auto load the iRiver BS right?
21:20:15TorneJust supporting the relevant requests over usb and forwarding the request/response to/from teh drive woul dbe easy
21:20:19Torneand kinda independant
21:20:21scanfcurrentl it says 'firmware not found' and says 'plug usb cable'
21:20:29TorneThat would let smartmontools or whatever on the host deal with smart interpretation :)
21:20:33Tornewhich is easier
21:20:40n1sah, ok
21:20:42Torne[7]: if you feel like doing that, go for it
21:20:44wodzscanf: it behaves as expected.
21:21:08scanfi thought it was supposed to fallback to the default factory firmware
21:21:11[7]Torne: smart code in terms of accessing the drive's smart data or the usb interface?
21:21:15n1sso interpreting the data is the tricky bit?
21:21:18[7]AFAIK smartmontools currently doesn't support that particular USB bridge's interface... but that could be added
21:21:23Tornen1s: yes, interpreting the data is basically insanity
21:21:33Tornegetting the data is trivial
21:21:53Tornesmartmontools has a huge load of code to do it, with a giant database of weird special cases :)
21:22:11n1sfun :)
21:22:39Tornei know roughly how to get some kind of useful info out of it, but haven't had time to sit down and write codew
21:23:04*[7] just hopes that the bridge doesn't do translation between SMART and the SCSI variant of it, but instead just provides the raw data somehow
21:23:20Tornei'm pretty sure it just sends the same data
21:23:39[7]Torne: do you happen to have a comprehensive list of scsi commands around?
21:24:15scanfhow can i get it to default bootup the OF?
21:24:19pamaury[7]: there is a scsi spec ! :)
21:24:34jlbiasinilebellium: now that i remember I also experienced hang while selecting another theme in theme list menu after having yours selected
21:24:51topikscanf reflash the original firmware ?
21:24:55funmanscanf: did you check the manual?
21:25:05scanffunman: i cant find relevant info
21:25:05[7]pamaury: no, there's like a dozen different scsi specs, each containing a subset of each other
21:25:13scanfon how to reflash
21:25:27jlbiasinilebellium: no improvement, this is still bugging!
21:25:39pamaury[7]: I know I know, that's why I'm smiling
21:26:27jlbiasinilebellium: your theme seems to be the only one with space in the name and files name... are you sure this is safe?
21:26:59[7]it's done through a SCSI "drive status page" command
21:27:36jlbiasinipamaury: the bmp causing trouble is only a part of the screen: could it be related to the fact we are miissing some partial update for the lcd?
21:28:20 Join saratoga [0] (9803ec71@gateway/web/freenode/ip.
21:28:45***Saving seen data "./dancer.seen"
21:28:53scanfokay yeah its what i was saying
21:28:56scanfIf you wish to clean up your disk, you may also wish to delete the .rockbox directory and its contents. Turn the Sansa off. Turn the player back on and the original Sansa software will load.
21:29:00scanf^ that does not function in that way
21:29:09pamauryjlbiasini: currently any partial update will redraw the whole screen, but if there is a corruption, it means the framebuffer was corrupted which is bad
21:29:25lebelliumjlbiasini : then I don't know why there is this bug. Maybe because USB mode is sth particular. I mean, my theme displays a mini-player or this problematic picture according to the playmode but I don't know how it works for USB mode
21:29:44desowin[7]: what usb analyzer do you use?
21:29:46pamaury[7]: is this standard ?
21:30:09wodzscanf: It does. You simply didn't update the player with plain OF.
21:30:13scanfwith patched firmware, deleting the .rockbox directory makes it sit on 'firmware not found' and 'plug usb cable', it does *not* boot into the original sansa software without holding a left arrow
21:30:57scanfoh it wants the unpatched firmware?
21:31:09wodzscanf: The manual says "Copy an *unmodified* original firmware to your player and update it."
21:31:10scanfthere isnt a method to use rockbox's firmware to boot into sansa's stuff
21:31:12jlbiasinipamaury: you mean worse than the data abort that follow? ;)
21:31:12[7]desowin: I'm playing around with multiple of them, depending on what I'm trying to do
21:31:25[7]if it can be done on linux, wireshark is just great
21:31:41desowin[7]: so no hardware gear?
21:31:46topiknot as a default scanf
21:31:52[7]but if you're stuck with windows, there's snoopypro, sysnucleus usbtrace and usblyzer
21:32:12pamaurydesowin: a software analyzer is much better if you can, my experience with the rockbox hardware usb analyzer is pretty terrible
21:32:36desowinpamaury: what is rockbox hardware usb analyzer/
21:32:49[7]no, i don't have a hardware analyzer, but wireshark will usually do equally well for anything that's not at the electrical level
21:32:54desowinwell, I wish the OpenVizsla came out ;-)
21:33:24pamauryit's a hardware analyzer that someone (gevaerts: who ?) bought for rockbox, and that I currently have
21:33:27pamaurywireshark is quite good, even though it could probably be made much better for usb
21:33:48[7]desowin: rockbox owns some hardware gear, no idea where it is currently though
21:33:48[7]apparently pamaury did have it at some point
21:33:52saratogafunman: is USB on AMSv2 working in current SVN for you?
21:34:06saratogai tried 6 times in a row and it hard locked the player every time, with or without sd card
21:34:09funmansometimes it works sometimes i need to reste the player
21:34:12[7]eeek, IRC is laggy today
21:34:32saratogawith the previous build I had 100% success rate after a dozen tries on the same pc
21:35:16 Quit jlbiasini (Remote host closed the connection)
21:35:19wodzsaratoga: moon phase have changed :-)
21:35:47saratogawell there were a dozen or so commits touching USB as well :)
21:35:51 Join jlbiasini [0] (
21:35:58funmansaratoga: do you remember which rev?
21:36:26saratogafunman: no but i can probably figure it out based on the svn logs
21:37:09funmanand btw the other driver still works on nano2g (but not fully on classic it seems)
21:38:18saratogafunman: pretty sure it was r31231 that I was testing
21:38:26 Quit benedikt93 (Quit: Bye ;))
21:40:39 Quit n1s (Quit: Ex-Chat)
21:52:21TheLemonManthe charging pump works too
21:56:38 Quit quem (Excess Flood)
21:56:52 Join quem [0] (
21:59:33pamauryTheLemonMan: is it hardware controlled or you enabled it ?
22:00:06 Quit Topy (Ping timeout: 240 seconds)
22:00:54[7]ah, smartmontools does support it if you pass -d usbjmicron
22:01:18TheLemonManit has to be enabled and managed by hand, had to write some code to wrap all those high/low toggling
22:01:59pamauryhigh/low toggling ?
22:02:53funmansaratoga: i get lock up with r31237 (daily build)
22:03:50funmansaratoga: so r31232 mgiht cause problems
22:05:31[7]ok, so that particular bridge has an ATA passthrough SCSI command (0xdf)
22:05:37 Join Topy [0] (
22:06:12pamaury[7]: so you don't have to translate valyes ?
22:06:31[7]but there's also a standard version of that command
22:07:05 Quit Torne (Ping timeout: 255 seconds)
22:07:08[7](command 0x85/0xa1)
22:08:06[7]we might even implement both interfaces to be compatible with more analysis tools
22:08:10[7]this is trivial from the scsi side of things
22:09:20[7]we just need a way to send raw ATA commands to the drive from our SCSI layer
22:10:36[7] documents the standard way to do it, and the JMicron way seems to be very similar (just a different command number and attributes)
22:22:06 Join Torne [0] (~torne@rockbox/developer/Torne)
22:28:34funmanwhen SD writes error our: [13680.504130] sd 14:0:0:0: [sdb] Unhandled error code
22:28:46funman(hadn't noticed that SD write was disabled for AMSv2 bootlaoder)
22:29:17 Join webguest02 [0] (
22:29:51 Quit webguest02 (Client Quit)
22:31:28 Quit Stummi (Quit: Bye!)
22:32:19funmansaratoga: that diff doesn't seem to fix the lock outs
22:32:40saratogafunman: ok was just about to try it
22:33:45funmanbootloader usb works, but not rockbox.sansa yet (after 3 tries)
22:34:20funmanworked at 4th
22:35:30 Quit remlap (Read error: Connection reset by peer)
22:36:47 Part funman
22:36:47 Join funman [0] (~fun@rockbox/developer/funman)
22:39:08 Quit wodz (Quit: Leaving)
22:43:32 Quit TheLemonMan (Quit: WeeChat 0.3.6)
22:44:06 Join dreamlayers [0] (~bgjenero@rockbox/developer/dreamlayers)
22:52:38 Quit kadoban_ (Read error: Operation timed out)
22:57:50 Join T44 [0] (
23:00:22saratogafunman: r31231 works much better for me
23:00:56funmanno locks ever?
23:01:12saratogajust mounted it a few times without trouble, before it wouldn't mount at all
23:01:18 Quit Topy (Ping timeout: 248 seconds)
23:02:20saratogajust did 8 plugin, unplug cycles and it mounted and exited USB mode normally for each
23:02:33saratogashould I try 31232 and see if that broke it?
23:03:36 Quit JdGordon (Ping timeout: 240 seconds)
23:08:18saratogar31232 didn't crash over 10 mount/disconnect cycles either
23:10:02funmandaily build 31238 should be indentical for the zip and i still had problems
23:15:09saratogar31254 crashes for me
23:15:47 Join JdGordon| [0] (
23:15:47 Quit JdGordon| (Changing host)
23:15:47 Join JdGordon| [0] (~jonno@rockbox/developer/JdGordon)
23:16:18 Nick JdGordon| is now known as JdGordon (~jonno@rockbox/developer/JdGordon)
23:19:14 Part jlbiasini
23:21:15saratogar31250 is no good either
23:26:04 Join Topy [0] (
23:28:17 Quit dreamlayers (Quit: Leaving)
23:28:48***Saving seen data "./dancer.seen"
23:29:12 Quit domonoky (Read error: Connection reset by peer)
23:29:37 Quit T44 (Ping timeout: 240 seconds)
23:36:35 Quit JdGordon (Remote host closed the connection)
23:38:12 Join Scromple [0] (~Simon@
23:52:32 Join remlap [0] (
23:57:00 Join rarog [0] (
23:58:09 Part TomColler

Previous day | Next day