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 2008-11-10

00:00:12funmani only suggested it was a 'bank switching' because it was mentioned in the e200v1 driver
00:00:21kugelfunman: also, I think bertrik noticed some csd issues
00:00:36 Quit Bensawsome ("The awsome is gone :(")
00:00:40 Quit n1s ()
00:00:42funmankugel: what do you mean ?
00:02:47 Join Bensawsome [0] (n=Bensawso@unaffiliated/bensawsome)
00:03:37funmani noticed the same thing
00:03:52funmani just thought "bah they didn't care about the specification they wrote themselves"
00:04:05kugeli doubt that
00:04:19 Nick mcuelenaere_ is now known as mcuelenaere (n=mcuelena@rockbox/developer/mcuelenaere)
00:04:45kugelif they want the zillions of cards they sell to be accepted by card readers, they should stick to their specs
00:05:13funmanit's not "cards" here, not in the removable sense
00:06:12 Quit domonoky (Read error: 104 (Connection reset by peer))
00:06:48rasherI'd be amazed if they bothered produce separate chips for the memory too. Just grab some of those they're already producing millions of
00:07:10 Join vac33 [0] (
00:07:57vac33Hey anyone know where I can get themes for sansa e200
00:08:20kugelfunman: I didn't mean the embedded sd card only
00:08:33 Quit bmbl ("Woah!")
00:08:40 Quit faemir (Remote closed the connection)
00:09:07funmankugel: nobody is gonna desolder the storage inside the Sansav2 , so they are free to violate the specification as they want
00:10:44 Quit vladanian ("CGI:IRC")
00:10:45 Quit BHSPitLappy (Remote closed the connection)
00:11:14kugelfunman: anyway. I've looked at ata_sd_as3525.c a bit. the first sdhc detection looks a bit weird to me
00:11:45 Join t0mas [0] (n=tomas@rockbox/developer/t0mas)
00:12:15funmankugel: feel free to fix it then, without a SD slot or a SDHC card it's a bit hard for me to do so
00:13:06kugelfunman: well, I could fix it. The specs say hosts should look for ocr bit 30. but your sd driver needs sdhc before retrieving that bit
00:13:17funmanI used the SD specification version 2.00
00:14:10funmanand this is also *exactly* what the OF does (the argument for the SEND_OCR command is different for SDHC and not HC cards)
00:14:40kugelwell, i have doubts anyway that this fixes the issue
00:14:48kugeljust looked weird
00:15:42 Quit Acksaw (Connection timed out)
00:15:59 Quit massiveH ("Leaving")
00:16:13 Join massiveH [0] (
00:16:20 Join dany_21a [0] (
00:18:35 Join MethoS- [0] (
00:19:05 Quit vac33 ("ChatZilla 0.9.83 [Firefox 3.0.1/2008070208]")
00:20:39 Quit bluebrother ("leaving")
00:21:48 Join kid_digit [0] (
00:25:46 Quit massiveH ("Leaving")
00:26:46 Quit MethoS (Read error: 60 (Operation timed out))
00:27:06 Join massiveH [0] (
00:30:20 Part dany_21a
00:30:25 Join Thundercloud [0] (
00:31:51 Quit massiveH (Client Quit)
00:38:19 Quit funman ("leaving")
00:44:25gevaertsamiconn: I'm not sure about that #if. It probably is unlikely that we ever get another hardware USB hotswap target, but shouldn't those #ifs be placed even then, for completeness and correctness?
00:45:39gevaertsI agree that this isn't the correct fix for this specific compile failure though. I missed the missing HAVE_USBSTACK bit
00:46:13*gevaerts thinks that if you don't have a usb stack implementation yet, you just shouldn't compile usb.c in
00:50:39kid_digithey guys pretty newb question but what file can/do I use for Last.Fm scrobber? (running on a 60gb Ipod Vid)
00:51:19 Quit aarcane ("Leaving")
00:53:08 Quit _lifeless (Read error: 104 (Connection reset by peer))
00:53:28 Join _lifeless [0] (n=lifeless@
00:59:28 Join funman [0] (
01:00:21 Join massiveH [0] (
01:03:44 Quit ender` (" There are two major products that come out of Berkeley: LSD and UNIX. We don't believe this to be a coincidence. -- Jeremy")
01:04:17funmanbertrik: also the name of the SD card is not 100% ASCII
01:06:14 Quit funman ("leaving")
01:09:46 Join Calcipher [0] (
01:10:41CalcipherI have a very specific situation here
01:11:47CalcipherI have a Sansa E280R, that will crash when in its original firmware, after trying to refresh DB, and therefore will not allow a USB sync to occur with a computer
01:12:20Calcipheryet same player continues to work almost flawlesly in RockBox mode
01:12:37gevaertsHave you checked the filesystem?
01:13:05CalcipherI figure, I should be able to do some repair in Restore mode right?
01:13:16CalcipherIt won't usb sync
01:13:23Calcipherfreezes before that point
01:14:11Calcipherin restore mode I should be able to reformat it If I remember right
01:14:13gevaertsI'd suggest using the sansa.fmt trick to reformat, but I don't know if that works on e200r
01:15:22CalcipherI just want to make sure I can get rb working on it again, Its my friends player, and hes totally blind and loves RB
01:15:34gevaerts tells all about it
01:15:48 Join linuxstb_ [0] (n=linuxstb@rockbox/developer/linuxstb)
01:15:57Calciphernice, thanks
01:16:22gevaertsUnless the hardware is broken, I see no reason why you shouldn't be able to reinstall rockbox afterwards
01:16:50CalcipherI knew this was the place to come for answers, all the googling didn't even mention restore mode, I just remembered that from when I set up RB on it last year
01:16:58gevaertsIf you need to keep what's currently on the player, I'd try copying it to a microsd card
01:18:11CalcipherI'll see if I can grab stuff off in restore mode, cause I don't think I can copy and paste directly through the player
01:18:48gevaertsRecovery mode doesn't expose the flash.
01:19:22gevaertsYou can copy and paste in rockbox
01:20:32CalcipherNiiice, atleast I can see my songs in Recovery mode
01:20:39Calciphersome priceless tracks in there
01:21:01 Part kid_digit
01:21:23CalcipherHonestly I think I should be able to just back up my rb build, and songs, then reformat
01:21:34gevaertsThat doesn't sound like recovery mode then
01:22:06Calcipherset hold switch to ON, hold down record, and menu when powering on
01:22:35Calcipherand it lets you see player as mass storage without loading
01:22:37 Quit MethoS- (Remote closed the connection)
01:22:55gevaertsIs that something e200r specific then?
01:23:19gevaertsNormal recovery mode shows a 16MB "disk" containing one file, version.txt
01:23:26 Quit reacocard (Read error: 110 (Connection timed out))
01:23:31CalcipherI jumped the gun
01:23:35Calcipherby alot
01:23:57Calcipheryour right
01:24:09Calcipherthats what I got, 16mb
01:24:25gevaertsWhatever you do, _don't_ format that
01:24:59Calcipherso, let me check that link you gave me thoroughly haha
01:29:46 Quit kugel ("ChatZilla 0.9.83 [Firefox 3.0.3/2008101315]")
01:30:56 Quit knnylftn (Remote closed the connection)
01:31:16 Join knnylftn [0] (
01:31:44mcuelenaeregevaerts: have CONFIG_STORAGE & STORAGE_RAMDISK and the USB stack been tested?
01:31:57mcuelenaereah you're asleep
01:33:23gevaertsmcuelenaere: not totally asleep yet :). Yes, I've run an ipod mini build from ramdisk
01:34:01mcuelenaereI'm getting linker errors because of missing mmc_detect() and mmc_enable_monitoring() functions
01:34:15mcuelenaeresounds familiar?
01:34:32mcuelenaere(they're in disk_mount_all() )
01:34:51 Quit linuxstb (Read error: 110 (Connection timed out))
01:35:21gevaertsNo. I haven't seen those
01:35:52mcuelenaereah I see the problem
01:36:03mcuelenaereit's in hotswap.h:51
01:36:14mcuelenaereshouldn't that be #elseif CONFIG_STORAGE & STORAGE_MMC ?
01:36:27gevaertsThe only "issue" I had is that if rockbox goes to usb mode because it can't mount a disk, it reboots after disconnect, which is reasonable for a real disk but not helpful with ramdisks
01:36:40*gevaerts looks
01:37:16gevaertsYou're running with both hotswap and ramdisk?
01:37:34mcuelenaereperhaps a misconfiguration
01:37:56mcuelenaereit's probably due to that I'm hacking a kind of multi volume support
01:38:30mcuelenaerebut that's unrelated to the USB ramdisk
01:38:57mcuelenaereok, commenting out have_hotswap fixed it
01:38:59gevaertsI'd say you're right about the elif, but it's a bit late for me to think clearly
01:39:20gevaertsAnd that alone won't make multidriver play nice with hotswap anyway I guess
01:39:55*gevaerts now really goes to sleep :)
01:40:02mcuelenaeregood night then :)
01:41:28Calcipherhmm, I'm looking at the instructions on recovering originall fw through recovery mode using a MI4 file
01:42:01Calcipherbut, there are 3 different versions of the fw on the linked page, for sansa e200 american version
01:42:19 Quit t0mas ("good night!")
01:42:34***Saving seen data "./dancer.seen"
01:42:39Calciphertheres no explanation as to what BL file, is which is different for some of the mi4 files listed
01:43:09CalcipherI just want to make sure I can use RB on this after I'm done
01:45:23Calcipherthats the link provided on the rb wiki for unbricking sansa
01:49:57 Join massiveH_ [0] (
01:50:38 Quit moos ("Rockbox rules the DAP world")
01:51:48Calcipherare all the MI4 files for the sansa e200? the section says so yet there are Iriver and samsung labeled MI4 files
01:54:12CalcipherI think one more sentence right around that section could clarify this a hell of a lot
01:55:09Calcipherdon't know what the boardsupportSD.rom files are for and why they're listed for each MI4
01:56:08CalcipherBL_SD_BoardSupportSD.rom, is the exact file name
02:00:14 Quit DerDome ("Leaving.")
02:01:36 Quit Thundercloud (Remote closed the connection)
02:03:30 Quit mcuelenaere ()
02:07:05 Join BHSPitMonkey [0] (n=stephen@unaffiliated/bhspitmonkey)
02:07:10 Quit culture (Read error: 110 (Connection timed out))
02:07:30 Quit massiveH (Read error: 110 (Connection timed out))
02:08:46 Join reacocard [0] (n=reacocar@WL-112.CINE.HMC.Edu)
02:15:17 Nick massiveH_ is now known as massiveH (
02:22:40CalcipherOK, I copied the 1.02.15A.mi4 to the 16MB drive when sansa was in recovery mode
02:23:05Calcipherreceuved a load failure message on player restart that instructed me to switch to recovery mode
02:25:49 Quit pixelma (" -")
02:26:12 Join pixelma2 [0] (n=marianne@rockbox/staff/pixelma)
02:30:55Calcipherset it to recovery mode and now the mi4 file i placed in there is gone
02:31:30CalcipherI sort of remember reading something about renaming the mi4 to some other filename.mi4
02:32:16Calciphersansa.mi4? is that it? I really don't want to just experiment at this point
02:55:04 Quit Rob2223 (Read error: 104 (Connection reset by peer))
02:55:41 Join blkhawk- [0] (n=blkhawk@
02:55:45 Join Rob2222 [0] (
03:06:26 Join Horschti [0] (
03:07:10 Quit Horscht (Nick collision from services.)
03:08:46 Nick fxb is now known as fxb__ (
03:08:50 Join Darksair [0] (n=user@
03:10:37 Join Nico_P [50] (n=nicolas@rockbox/developer/NicoP)
03:12:16 Quit blkhawk (Read error: 113 (No route to host))
03:23:21 Quit Bensawsome ("The awsome is gone :(")
03:24:35 Quit jhulst (Remote closed the connection)
03:26:12 Join AndyIL [0] (i=AndyI@
03:28:23 Quit Nico_P (Remote closed the connection)
03:30:35 Join jhulst [0] (n=jhulst@unaffiliated/jhulst)
03:32:11 Quit |AhIoRoS| ("Abandonando, see you")
03:33:41 Quit XavierGr ()
03:39:51 Quit AndyI (Read error: 110 (Connection timed out))
03:39:59 Join |AhIoRoS| [0] (n=ahioros@
03:42:05 Quit PaulJam (".")
03:42:39***Saving seen data "./dancer.seen"
03:51:21 Join homielowe [0] (
03:52:08Calcipherwell. the error message I get on the sansa is "Main image load failure, switch to recovery mode"
03:53:23Calcipherto be exact it says "Load main image failed, switch to recovery mode"
03:53:51Calcipherbasically, now I'm worse off then when I started
03:54:30 Join john [0] (
03:55:02 Nick john is now known as Guest36047 (
03:58:10 Join webguest29 [0] (n=45ea90f2@gateway/web/cgi-irc/
03:59:10 Quit webguest29 (Client Quit)
04:03:12 Quit HBK ()
04:03:18mc2739Calcipher: still here?
04:03:44mc2739which firmware file did you use?
04:04:41Calcipherfor an e280R, which could not complete db refresh under its own fw, but could still load rb and play perferctly
04:04:50Calcipherbefore i tried fixing it ofcoursew
04:04:54mc2739that firmware is not for an e200r
04:05:42mc2739you need to look for 01.02.031
04:06:08Calcipheris that available on daniels page?
04:06:21CalcipherI really appreciate the help by the way
04:07:00Calcipherthe unbrick wiki was vague enough where I got lost in my situation
04:09:02 Join HBK [0] (
04:10:33mc2739you're welcome
04:10:34CalcipherI'm guessing I have to re do the bootloader and rb set up after this
04:10:49JdGordonsoap: whats your view on commiting the ipod accesory patch?
04:10:52 Quit Guest94694 (Read error: 110 (Connection timed out))
04:11:18mc2739yes, but you may have other steps for recovery yet
04:11:41Calcipherooh, lets see, I'll do it now so I can let you know
04:12:01 Quit HBK (Client Quit)
04:12:03soapI, personally, have no accessories to test it on. That being said I have seen no reports of it causing problems. I only do the Unsupported Build because I feel it is one of those patches which benefits from lots of users.
04:12:37 Nick blkhawk- is now known as blkhawk (n=blkhawk@
04:12:59JdGordonas long as it doesnt cause problems for ipods which dont need it, I tihnk its time it went in
04:13:07JdGordonunless you like keeping that build going? :)
04:15:26 Join HBK [0] (
04:15:33Calciphermc2739, it came back into original fw, then attempted to refresh db, and crashed at a blank blue screen again
04:16:12Calcipheras it was doing before, except this time it didn't try to load rb automatically
04:16:42Calciphersound like i may need to do that .fmt file thing
04:16:59mc2739ok, now go back to recovery mode and add an empty file named sansa.fmt to the 16MB drive
04:17:43soapIf I say a word now in support of it - the problems will come rolling in. Murphy's law. I shall remain mum ;)
04:18:38JdGordonok, so thats a yes then? ;p
04:18:55JdGordonpatch seems fine to me, apart from the lang part of the diff
04:19:19JdGordonthe fun will start when people start complainig about devices which arn't handled
04:20:54 Quit jhulst (Remote closed the connection)
04:21:02soapI'll try to get the time to make a more modern IpodAccessories wiki page and dump all the new reports in it.
04:22:35Calcipherits back to life!
04:22:46Calciphernow I treat it as a new rb installation
04:23:00mc2739cool, now to reload rockbox, follow this wiki -
04:23:08 Join miepchen^schlaf_ [0] (
04:23:09CalcipherI have no idea what my friend did to get it to crash like that in the first place
04:23:28mc2739probably a bad file
04:24:11Calciphernow the only scary thing is I have to remember how it was I was making builds with the sexy chick tts voices
04:24:33Calcipherthat took a while to get for someone who never messed with builds and svn
04:24:38mc2739sorry, can't help you with that one
04:25:10Calciphernah don't worry, all the work folders are still in tact
04:25:39Calcipherjust haven't updated my rb since feb
04:26:26mc2739thats old for rockbox, lot of fixes since then
04:26:28Calcipherdb is still not being voiced right?
04:26:47Calcipheroh man, thats gonna be a pain
04:27:01mc2739I believe that is correct, I don't use the database so I am not sure
04:27:21CalcipherI'm just curious
04:27:31Calcipherhow do you play your music if you don't use db
04:28:07CalcipherI just find the db useful since you can select all tracks by a certain artist, instead of having to organize folders differently
04:28:16mc2739I use playlists built on the pc
04:29:20Calcipherthanks for the info, i'll be back in a few after reinstalling rb
04:29:31mc2739ok - good luck
04:31:21 Quit miepchen^schlaf (Read error: 145 (Connection timed out))
04:36:49 Join jhulst [0] (n=jhulst@unaffiliated/jhulst)
04:51:29 Join blkhawk- [0] (
04:56:43 Join ajonat [0] (n=ajonat@
04:57:33Calcipherooh, pretty, new default rb has taxi skin
04:57:45CalcipherI think thats what it was called
04:58:23Calcipherthank you all so much, especially mc2739
04:59:20mc2739no problem, glad to help
05:06:07CalcipherI was always clammering about landscape mode selection and db being voiced, don't think any of those ever came to happen
05:07:09mc2739no, I don't think so
05:07:47 Quit blkhawk (Read error: 110 (Connection timed out))
05:08:26 Nick blkhawk- is now known as blkhawk (
05:09:51Calcipherhopefully not too much changed since my last build, so maybe voice config steps are still working
05:09:54Calcipherwe'll see
05:12:07Calcipherupping from svn now
05:22:40Calcipheroh damn..., I don't remember what this syntax was for voice build options...
05:25:48 Quit Llorean (Read error: 113 (No route to host))
05:31:27Calcipherlooks like things have changed, less tts options, looks like only sapi is available
05:38:22 Join __lifeless [0] (n=lifeless@
05:39:38 Quit _lifeless (Read error: 113 (No route to host))
05:42:28 Quit Horschti ("")
05:42:43***Saving seen data "./dancer.seen"
05:56:21 Quit mc2739 ()
05:58:45 Quit massiveH ("Leaving")
05:59:58 Join massiveH [0] (
06:06:54Calcipherhey jd
06:11:02Calcipherwell I was going to ask about some patches in FS tracker
06:11:19Calcipherbut i just searched for keywords and didn't see them in there
06:11:33Calcipherlet me run through my old emails maybe i can find the numbers
06:15:45Calcipherwell I just answered my own question, nothing has been updated on the items I was interested in
06:22:07CalcipherI was looking at 8083 and 6637
06:41:34 Quit massiveH ("Leaving")
06:52:00 Quit miepchen^schlaf_ ()
07:06:17CalcipherI'm still awake, and simply amazed at how well rockboxutility is working, even allowing proper selection of voice engine and talk file creation
07:07:43 Nick __lifeless is now known as _lifeless (n=lifeless@
07:07:51Calcipherthings have improved so much, I had no idea
07:08:09_lifelesshello, is somebody have pp5022 datasheet/pinout ?
07:09:00Calciphernot me, sorry
07:09:56 Quit ajonat ()
07:13:08 Quit reacocard (Read error: 110 (Connection timed out))
07:18:30_lifelesstrying to enable touchpad
07:21:20advcomp2019there is no datasheet for the pp5022 as far as i know
07:24:26 Quit Zarggg ()
07:25:43 Part Calcipher
07:30:56 Join Zarggg [0] (
07:31:58 Join J-23 [0] (
07:42:47***Saving seen data "./dancer.seen"
07:47:27 Join bmbl [0] (n=Miranda@unaffiliated/bmbl)
07:48:30 Join LinusN [0] (n=linus@rockbox/developer/LinusN)
07:51:46 Quit bmbl (Client Quit)
07:52:37 Join Thundercloud [0] (
07:54:22 Join avis [0] (n=ident@pdpc/supporter/student/avis)
08:06:51 Join ender` [0] (
08:14:51 Join GodEater [0] (i=c2cbc962@gateway/web/ajax/
08:21:01 Quit fyrestorm (Read error: 54 (Connection reset by peer))
08:24:28 Join fyrestorm [0] (
08:26:55 Nick Jabone_ is now known as Jabone (
08:27:43 Join lasser [0] (
08:28:21 Join Makuseru [0] (
08:30:31 Join Bagderr [241] (n=daniel@rockbox/developer/bagder)
08:30:56 Nick Bagderr is now known as B4gder (n=daniel@rockbox/developer/bagder)
08:31:30 Join Rob2223 [0] (
08:38:33 Nick linuxstb_ is now known as linuxstb (n=linuxstb@rockbox/developer/linuxstb)
08:48:23 Quit Rob2222 (Read error: 110 (Connection timed out))
08:51:57 Quit Darksair ("Everything that has a beginning has an end.")
09:03:04 Join BHSPitLappy [0] (n=BHSPitLa@unaffiliated/bhspitmonkey)
09:06:32 Quit amiconn (Nick collision from services.)
09:06:39 Join amiconn [50] (n=jens@rockbox/developer/amiconn)
09:06:56 Join pixelma [0] (i=pixelma@
09:14:04JdGordonLinusN: hey, I dont know how much the ipod accessory patch has changed since you opened it, but feel like giving a quick look to see if its ok to commit? looks fine to me....
09:14:12JdGordonFS #8624
09:14:22LinusNi was about to do just that
09:14:45 Quit BigBambi (Read error: 104 (Connection reset by peer))
09:18:29 Quit Guest36047 ("Leaving.")
09:19:18LinusNleaves a little to wish for stylewise
09:20:02LinusNbut nothing serious
09:22:24 Quit pixelma2 (Read error: 110 (Connection timed out))
09:28:54MakuseruDoes anyone have experiance converting files so that they work with Mpegplayer?
09:29:20B4gdersome must have since there's a nice wiki page about it...
09:29:42Makuseruthat doesnt mean they are in here at this moment
09:29:53B4gderno, and you didn't ask for that...
09:30:15Makuseruwell im asking in here, at this moment, am i not?
09:30:42JdGordonto be precise... no your not
09:30:43*B4gder shuts up and just looks at Makuseru
09:30:59B4gdercan you instead just ask the question?
09:32:54MakuseruIve got a video file, an MKV, and it has different subtitle tracks in it, ive converted other filetypes using the string for mencoder at the bottom of the wiki on the mpegplayer plugin, but never MKV's, and i just cant figure out how to make this subtitle track appear in the final mpeg
09:34:47JdGordonarn't subtitles comepltyl seperate from the video? i.e displayed by the viewer?
09:35:28linuxstbThey can be, but mpegplayer doesn't, so Makuseru needs to render them in the video. I'm sure mencoder can do that, but I've no idea how.
09:35:52MakuseruJdGordon: im not sure i completly understand what your saying, but if you mean arent they a seperate sub file, than no
09:41:37B4gderMakuseru: aren't you using a specific subtitle option for mplayer to watch the movie? isn't that then the same you should use when transcoding?
09:42:05 Quit MrDuck (Read error: 113 (No route to host))
09:42:12Makuserui dont know the subtitle option, ive only ever used VLC to play MKVs
09:42:31Makuseruif i knew that option, would that work to add it to the mpeg?
09:42:41B4gderI'd assume so
09:42:49***Saving seen data "./dancer.seen"
09:43:08 Quit esthar ("KVIrc 3.4.0 Virgo")
09:43:47 Quit avis ("Ex-Chat")
09:44:08linuxstbMakuseru: You would get better support asking the mplayer/mencoder people, or google - mpegplayer just plays standard mpeg-1/mpeg-2 video files, so this isn't anything specific to Rockbox.
09:45:17Makuseruive been to a few forums and the maktrosa irc room, and so far you guys have been the most help
09:45:51 Join Zagor [0] (n=bjorn@rockbox/developer/Zagor)
09:51:38 Join Darksair [0] (n=user@
09:51:40 Join funman [0] (
09:53:02 Join petur [50] (n=petur@rockbox/developer/petur)
09:54:20funmanBagder: you are right, sansav2 bootloaders should be added to the build table, so even if they don't work they at least build :-)
09:54:48B4gderyes that's my point
09:55:23B4gderwe have several builds in there that don't actually work yet
09:56:29linuxstbB4gder: Can you delete the "Fuze" wiki page?
09:57:13funmanlinuxstb: hi! what is the capacity of your Clip?
09:57:20B4gderlinuxstb: will do!
09:57:27linuxstbfunman: 1GB I think
09:58:47funmanthe SD specification mentions that 'all SD cards should send their relative addresses (RCA)' maybe there's more than 1 card in the high capacity models; but I'm not sure how to select each cards one by one, perhaps with MCISelect register ? Someone on the forum thread proposed to have a look on his e200
10:02:51B4gderfunman: what bootloader(s) do you think would be suitable to start with?
10:03:52funmanClip/Fuze/e200v2/m200v2 should all build - I can't objectively choose one or two of these 4
10:03:54 Quit perrikwp (Remote closed the connection)
10:03:54 Quit GodEater (Read error: 54 (Connection reset by peer))
10:04:05 Join esthar [0] (
10:04:32B4gderit probably makes sense to add them all anyway
10:07:50linuxstbWhat do people think about renaming the m200v2 to m200v4 ? That seems to be the name used by Sandisk on their forums (in the firmware download threads)
10:08:35 Join einhirn [0] (
10:08:42B4gderI'm fine with that
10:09:20*linuxstb doesn't have time to do it now though...
10:09:28ZagorB4gder: you can add the lillebror build server again
10:09:53B4gderit's there already, it "should just work"
10:10:24B4gderbut if m200 is v4 is the e200 still v2?
10:11:20funmanafaiu the differents m200 (v1 v2 and v3) have different hardware each ?
10:11:30linuxstbB4gder: Yes, I think so - see the stickies here -
10:12:09linuxstbfunman: Yes, I think they have different NAND flash controllers. Same CPU though (TCC770)
10:14:14 Quit J-23 (Read error: 113 (No route to host))
10:15:44B4gderwhat is the output file names from the v2 sansa bootloader build process?
10:16:19B4gderbootloader-*.sansa ?
10:16:41funmanyes seems that's it - I have a bootloader-clip.sansa
10:17:50 Quit esthar ("KVIrc 3.4.0 Virgo")
10:20:32B4gdernow we just need a commit...
10:21:00B4gderhm I did my own a bit too early I think ;-)
10:21:36 Join fyre^OS [0] (
10:22:37 Join esthar [0] (
10:24:52 Quit Darksair ("People who are zhuangbility want to show their niubility but only reflect their shability.")
10:25:25 Quit Kopfgeldjaeger (Read error: 60 (Operation timed out))
10:25:35B4gdermaybe we should now stop referring to them as "sansa v2" players...
10:25:45B4gderand call them sansa ams or something
10:26:10 Quit BHSPitMonkey (Read error: 104 (Connection reset by peer))
10:26:39 Quit linuxstb (Read error: 104 (Connection reset by peer))
10:26:47qwmscorche: offtopic!
10:27:05 Join Kopfgeldjaeger [0] (
10:28:14 Quit BHSPitLappy (Remote closed the connection)
10:29:05 Quit fyrestorm (Read error: 145 (Connection timed out))
10:30:31*JdGordon too slow...
10:30:44JdGordonwe should go to *-ams and *-tcc instead of numbers
10:33:20JdGordonalthough if sansa is calling them v2 it probably does make sense for us to call them that also
10:33:32B4gder"them" is not v2, the e200 is
10:33:42B4gderand possibly the c200
10:33:50B4gderbut the m200 is v4 and clip and fuze are without v
10:34:25B4gderI don't think sandisk has a name for these
10:35:16 Join GodEater [0] (i=c2cbc962@gateway/web/ajax/
10:36:19funmanthe different Clips are named "hardware revision " v1.0 and v2.0 - the one we work on is the v1.0
10:36:59B4gderah right, even more confusing ;-)
10:37:42 Join bodymind [0] (n=bodymind@
10:38:16B4gderand we have 86 builds in the table!
10:38:56JdGordonok, enoughs enough.... 1 more build and it wont fit on my 1680x1050 lcd....
10:39:12bodymindi have a sansa e260v2... and i would like to help out on developing the new port... :)
10:39:27bodymindi program in c, and have some small experience in assembler...
10:40:34funmanbodymind: hello, your help is welcome, a few people work on the specific e200v2 model
10:41:14bodymindi've read the thread in the foruns... but i didn't understand in what state the port is...
10:41:15B4gderyellow samsa ams bootloader builds...
10:41:58funmanbodymind: since yesterday we have a SD driver but it doesn't work the high capacity models (only work for 1GB currently)
10:42:15funmanthe LCD screen, the i2c driver, and some buttons work on the e200v2
10:42:36funmanB4gder: I don't know/can't test rolo but I think adding the AS3525 cpu case in rolo.c would be enough
10:43:28B4gderwell, the yellow is about an unused variable and that can be nuked no matter what...
10:43:46funmanB4gder: imo nuking warnings is not good, they are here for a reason
10:44:17bodymindso what's next? :o
10:44:27B4gderfunman: I disagree with that actually
10:44:28funman'Someone' should either fix the real issue, or leave the warning here for the future, even if the table is not perfectly green
10:44:54B4gderthe warning will not work as that, it will just prevent people from noticign when "real" warnings appear
10:44:57funmanbodymind: next is fix the SD driver so that it supports >1GB models, I can give directions where to look for that
10:45:59funmanif someone who knows rolo.c can look at line 233 and see if the fix is to add the as3525 case, it'll be fixed easily :-)
10:46:23bodymindok, i will compile svn for now... for some testing
10:46:44B4gderthe m200v4 build seems broken, or did I really cause that failure?
10:47:09funmanB4gder: I don't know what you used for tools/configure : m200v2 or m200v4 ?
10:47:19JdGordonfunman: looking now.. yeah my guess is adding the AS3525 case should work correctly
10:47:20B4gder"Platform set to m200v4"
10:47:26B4gderthere is no v2 anymore...
10:48:32B4gderah my fault it is
10:48:55funmansansa-m200v2 -> sansa-m200v4 ?
10:49:00*B4gder didn't do a case insensitive grep... :-/
10:49:27B4gderfunman: exactly
10:52:17JdGordonB4gder: why is the warning in the clip output ok but garbeled a bit in the fuze one? 'i' vs ‘i’
10:53:25JdGordonboth pages are western iso-8859-1 accoridng to ff?
10:53:45B4gderyes but not the output from the build
10:53:53B4gderat the server's place
10:53:57 Join reacocard [0] (n=reacocar@WL-112.CINE.HMC.Edu)
10:53:57JdGordonah right
10:54:18bodymindok.. i've downloaded the svn and compile rbutil.. but it isn't yet installing in e200v2.. how do i install by hand?
10:54:35JdGordonthe e200v2 is only a bootloader so far
10:55:03bodymindhow do i compile and test on my player? :o
10:55:13JdGordonalthough if you manually do a regular build you can start stubbing missing stuff
10:55:23funmanuse tools/configure and select a bootloader build
10:55:24 Join MrDuck [0] (
10:55:43funmanthen patch a SanDisk firmware with this file, with the help of rbutil/mkamsboot/ utility
10:59:07bodymindwhere do i get arm-elf-gcc? 8-)
10:59:13bodymindubuntu here :p
10:59:19funmanuse tools/ to build everything you need
11:04:13bodymindROCKBOXDEV: patch is required for this script to work.
11:04:14bodymindROCKBOXDEV: Please install patch and re-run the script.
11:04:21funmansudo apt-get install build-essential
11:06:39GodEaterdid we get the patch to working on Ubuntu 8.10 yet btw?
11:06:51GodEateror is he going to have to specify that -U_FORTIFY_SOURCE thing manually still ?
11:07:28funmanit should work out of the box
11:08:22*GodEater hasn't tried it since B4gder's first attempt at the patch
11:11:25bodymindnice work :)
11:14:33 Join jeffdame` [0] (
11:15:28 Quit nuonguy ("This computer has gone to sleep")
11:22:44B4gderI still get a link problem with m200v4
11:23:14B4gderI'll get back to that later if nobody else does, I need to do some real work for a while now
11:23:52funmani'll have a look
11:26:06 Join culture [0] (
11:34:23bodymindok.. i've run make.. what patch do i need?
11:36:05 Quit esthar ("KVIrc 3.4.0 Virgo")
11:36:26funmanbodymind: no patch needed, everything is in the repository. once you have built the bootloader, use rbutil/mkamsboot/mkamsboot (run make in that directory) to write the bootloader into a e200 firmware, and copy this patched firmware as e200pa.bin on your e200
11:42:30 Join esthar [0] (
11:42:52***Saving seen data "./dancer.seen"
11:44:03 Quit culture (Connection timed out)
11:52:12 Quit martian67 (Excess Flood)
11:52:31 Join martian67 [0] (i=lol3izer@about/linux/regular/martian67)
11:57:30 Nick fxb__ is now known as fxb (
11:59:19 Join Nico_P [50] (n=nicolas@rockbox/developer/NicoP)
11:59:39 Join PaulJam [0] (
11:59:58 Join kugel [0] (n=chatzill@unaffiliated/kugel)
12:01:40kugelfunman: if I #ifdef out the csd v1.0 case (and assume always csd v2.0) for the fuze, I get ~4GB reported
12:02:02kugela bit less even
12:03:12 Quit martian67 (Excess Flood)
12:03:22 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
12:03:31 Join martian67 [0] (i=lol3izer@about/linux/regular/martian67)
12:03:52bodyminddualboot.S:138:10: error: #error No target-specific key check defined!
12:03:59bodymindon compiling mkamsboot.. :S
12:04:02funmanbodymind: run svn up
12:04:06funmankugel: hm ..
12:04:58 Quit jeffdame` (Read error: 113 (No route to host))
12:05:03bodymindwhere do i have to put the e200 firmware?
12:05:08funmankugel: but the csd version is set to 1.0 (bits 31:30 == 0x0) ?
12:05:28funmanbodymind: just copy it on your player, as "e200pa.bin", then unmount and disconnect it
12:05:29bodymindi found it.. 8-)
12:06:01kugelfunman: yep (i.e. if I don't #ifdef it out, it'll branch the 1.0 case and report 1GB)
12:06:29bodymindUsage: mkamsboot <firmware file> <boot file> <output file> what is the bootfile? :x
12:06:37kugelfunman: the capacity reported is ~76MB higher than windows reports
12:06:40funmanbodymind: bootloader-e200v2.sansa
12:07:05funmankugel: take the exact number, divide it by 512 to get the number of sectors in your SD card
12:07:14funmanthen substract from it the number of sectors reported by dmesg
12:07:24kugelwill do
12:07:26funmanand use this value instead of 20480, and see if it can read the disk
12:08:55bodymindaw right :o
12:10:29funmankugel: maybe you can force the use of CSD2 for drive 0, but is it also the case for Fuze/e200 1GB ?
12:11:05kugelafaik there's neither fuze nor e200 in a 1GB version
12:11:58bodymindaw right :o PANIC, disk init failed :p
12:12:10bodymindwhich key is used to choose the other firmware?
12:12:37funmanbodymind: don't remember, it's written in rbutil/mkamsboot/dualboot.S
12:13:56bodymindok ;)
12:14:56funmankugel: the SD card doesn't respond to CMD8 (send_if_cond) , right ?
12:15:42linuxstbfunman: What's the disk layout on the Clip? Do I understand correctly that there's no partition table at sector 0, then there is the OF, then the main partition (10MB from the start)?
12:17:09linuxstbThe 10MB is just hard-coded currently?
12:18:20bodymindok :)
12:18:29bodymindit's up and runing !
12:18:46bodymindwhy are you doing SD before disk init?
12:18:54 Join tyfoo [0] (
12:19:57GodEaterPresumably so that a Rockbox firmware can be loaded from something disposable (the SD card) before trying to have modify the internal storage, and possibly damage the player ?
12:20:35 Join funman_ [0] (
12:20:41 Quit funman (Read error: 113 (No route to host))
12:20:52funman_yes it's hardcoded for the Clip (and maybe valid for m200v4 and c200v2)
12:21:16bodymindgodEater ok!
12:21:38 Join J-23 [0] (
12:21:50bodymindi don't have no SDcard.. i can't help on that.. :s
12:22:04GodEaterbodymind: that's a complete guess - you might want to ask funman again - I guess he probably missed it due to dropping from IRC.
12:22:08linuxstbbodymind: The internal flash also uses an SD interface
12:22:17GodEaterah see - I was wrong ;)
12:22:19funman_linuxstb will know, I have to go :)
12:22:20 Quit funman_ (Client Quit)
12:22:43linuxstbThe "sd" driver is the low-level stuff, "disk" is the higher level stuff, dealing with partition tables etc.
12:24:03bodymindah ah :o
12:26:13bodymindwhere i can see that sd code?
12:26:27bodymindor where can i read about the rockbox tree?
12:27:29kugelGodEater, linuxstb: the 10MB is the OF offset (which is hidden on computers). it's no real partition, but a kind of
12:32:30kugelfunman: I replaced the 20480 with my number (157696), I still get disk_init failed
12:33:58kugelwith latest svn
12:34:31kugellinuxstb: if you like to you can commit my fuze sim (that'd get my tree clean)
12:35:21kugelbut I'd test it before, it's not a "clean" svn diff (should work though)
12:35:55 Join perrikwp [0] (i=4aa794a0@gateway/web/ajax/
12:36:34 Join robin0800 [0] (
12:38:55bodymindkugel: i have a e260v2 i can do testing... 8) i know a bit of c.. but no so low level.. :$
12:39:55 Quit Thundercloud (Remote closed the connection)
12:39:56kugelwell, what you can do, is hassle with ata_sd_as3525.c and try to get your capacity reported correctly
12:41:06 Quit kugel (Read error: 131 (Connection reset by peer))
12:43:26JdGordonhmm... shouldnt alot of that ata_sd_ code be generic and shared with the pp sansas?
12:45:29linuxstbkugel: Lots of other devices have such a partition (Sansa v1, ipods etc) - it might be cleaner to fake a partition table, so the firmware partition is treated as partition 1, and the main part of the disk as partition 2.
12:46:07 Quit Nico_P (Remote closed the connection)
12:47:32bodymindcant compile last svn.. :s
12:47:45bodymindIn function `INT_TIMER2':
12:47:45bodymindkernel-as3525.c:(.text.INT_TIMER2+0x4c): undefined reference to `current_tick'
12:48:37JdGordonbodymind: the main build isnt ready yet... you either have to fix those yourself or stub them....
12:50:36bodyminderr, am i in the wrong path? i only want to build the bootloader
12:50:43bodymind30min ago was building ok!
12:51:18JdGordonlinuxstb: fake it where? in common/disk.c?
12:52:29linuxstbMaybe, or at least, at that level (i.e. higher than the SD driver)
12:58:20bodymindi can't mout my sansa...
12:59:06linuxstbHave you started the original firmware?
12:59:26 Join tvelocity [0] (n=tony@
12:59:29*linuxstb shrugs then
12:59:29bodymind[ 9324.000582] exo-mount[20516]: segfault at 00000008 eip 0804ad21 esp bfd76fe0 error 4
12:59:29bodymind[ 9327.736680] exo-eject[20521]: segfault at 00000008 eip 0804ad21 esp bf98dc00 error 4
12:59:29bodymind[ 9330.396844] exo-mount[20526]: segfault at 00000008 eip 0804ad21 esp bf9743e0 error 4
12:59:32bodymindon dmesg
13:00:00bodymindi didn't write any code.. just compile the svn..
13:00:13Zagorwth is exo-mount? what's wrong with "mount"?
13:02:08bodymindthere is no partition on fdisk...
13:02:35linuxstbCorrect - it uses "superfloppy" mode (IIRC)
13:03:00B4gderso many of the new players are superfloppy
13:03:20bodymindis it bricked? FUCK!
13:03:36linuxstbJust try "mount /dev/sdX /mnt"
13:03:42Zagorhow loevely of exo-mount to barf just because it doesn't recognize the disk...
13:03:48B4gderbodymind: nobody said it is
13:04:14JdGordonoh, ruddy hell.. is that why I could never get the bloody thing to work with linux?
13:04:16bodymindok, sorry. :s
13:04:31bodymindis it because of ubuntu with exo-mount etc?
13:04:45bodymindi can't mount because there is no /dev/sdX...
13:04:56linuxstbThen how did you use fdisk?
13:05:00bodymindfdisk -l only give me the partition of my HD
13:05:05bodymindsudo fdisk -l
13:07:17bodymindis it because of exo?
13:08:25linuxstbHow many internal disks do you have in your computer?
13:08:46Zagorfrom fdisk man page: -l List the partition tables for the specified devices and then exit. If no devices are given, those mentioned in /proc/partitions (if that exists) are used.
13:09:05linuxstbAnd this worked (doing exactly what you're doing now) before you installed Rockbox?
13:09:36 Quit Zarggg ()
13:11:30linuxstbDoes your Sansa have a setting (in the original firmware) to switch between MTP and MSC USB modes?
13:11:31 Join fredddy [0] (
13:17:19bodymindi've tried both modes.. :s
13:17:39bodymindi will restart my system... maybe it's a hal problem or something.. :s
13:17:41 Quit bodymind ("Ex-Chat")
13:20:48 Quit leachim6 (Read error: 110 (Connection timed out))
13:21:17 Quit intrados (Read error: 110 (Connection timed out))
13:23:11 Join BigBambi [0] (n=Alex@rockbox/staff/BigBambi)
13:24:33 Join Bensawsome [0] (n=Bensawso@unaffiliated/bensawsome)
13:26:28 Quit reacocard (".")
13:26:42 Join reacocard [0] (n=reacocar@WL-112.CINE.HMC.Edu)
13:27:57 Join bodymind [0] (n=bodymind@
13:28:02bodymindok it's a linux problem :o
13:28:36bodymindwhen i mounted the OF (updated yesterday) with would to to the special mode (of daniel site)
13:28:56bodymindnow it just won't mount in linux
13:29:02bodymindbut in windows mounted well ;)
13:29:22gevaertsbodymind: try to mount it manually
13:29:25bodymindbut i can't understand why.. :s
13:30:32bodymindi got this error in dmesg:
13:30:33bodymind 295.332904] usb 7-6: USB disconnect, address 3
13:30:33bodymind[ 301.623713] usb 7-6: new high speed USB device using ehci_hcd and address 4
13:30:33DBUGEnqueued KICK bodymind
13:30:33bodymind[ 303.091287] ehci_hcd 0000:00:1d.7: port 6 reset error -110
13:30:33bodymind[ 303.091301] hub 7-0:1.0: hub_port_status failed (err = -32)
13:32:58bodymindbut the OF without rockbox worked
13:33:20bodymindso you maybe missing a block that activates the special mode in linux?
13:33:57gevaertsspecial mode?
13:34:41bodymindi'm going for lunch... bbl
13:35:19linuxstbYou don't need a special mode to use it in Linux - just MSC.
13:36:36bodymindi'm on ubuntu.. maybe it's something with hal.. ? because the first time i mounted the sansa in linux there were different directories from windows
13:36:54 Join Thundercloud [0] (
13:36:55bodymindand it seemed just like special mode that daniel writes about...
13:37:17bodymindi will install the OF withou RB later to try it ouy again... cya later
13:37:26bodymind*ouy = out
13:39:47 Join Darksair [0] (n=user@
13:40:11 Join linuxstb_ [0] (n=linuxstb@rockbox/developer/linuxstb)
13:42:55 Quit linuxstb (Nick collision from services.)
13:42:56***Saving seen data "./dancer.seen"
13:42:59 Nick linuxstb_ is now known as linuxstb (n=linuxstb@rockbox/developer/linuxstb)
13:43:29JdGordonbodymind: rmmod ehci_hcd seems to work for me
13:44:34GodEatergevaerts: that would be because my Mum looks after the paper ;)
13:50:55 Join Schmogel [0] (
13:51:26 Join kugel [0] (n=chatzill@unaffiliated/kugel)
13:56:21 Join intrados [0] (
14:02:35 Quit gevaerts (Nick collision from services.)
14:02:44 Join gevaerts [0] (n=fg@rockbox/developer/gevaerts)
14:08:49 Join nplus [0] (n=nplus@141.25.Globcom.Net)
14:10:10 Nick JdGordon is now known as JdGordon|zzz (n=jonno@rockbox/developer/JdGordon)
14:17:01 Quit kugel (Read error: 113 (No route to host))
14:18:26 Join n1s [0] (n=nils@rockbox/developer/n1s)
14:27:02 Quit tvelocity (Read error: 104 (Connection reset by peer))
14:33:24Zagorurgh. plugin.h is filled with #include of irrelevant files. :-O
14:33:33 Quit J-23 ("Flying cow pressed ^D on my keyboard.")
14:34:22 Quit robin0800 (Read error: 104 (Connection reset by peer))
14:34:28 Join J-23 [0] (
14:44:57 Join LambdaCalculus37 [0] (i=44a04303@gateway/web/ajax/
14:52:50 Join kugel [0] (n=chatzill@unaffiliated/kugel)
14:58:58 Join tvelocity [0] (n=tony@
15:10:40bodymindJdGordon|zzz: worked for me 2!
15:12:49 Quit J-23 (Remote closed the connection)
15:16:39 Join J-23 [0] (
15:17:46 Join Llorean [0] (
15:20:19 Quit kugel (Remote closed the connection)
15:24:34bodymindok.. now it's the normal build... is it dangerous for a RBn00b to mess up with it?
15:26:17 Join meven [0] (
15:29:22 Join Nico_P [50] (n=nicolas@rockbox/developer/NicoP)
15:29:59 Quit bodymind ("Ex-Chat")
15:31:38 Join moos [0] (
15:36:24B4gderargh, lostlogic's server is now killing m68k builds...
15:37:02B4gderI'm now disabling it completely
15:37:28amiconnZagor: plugin.h *has* to include many other .h files, because plugins themselves *must not* include core stuff other than plugin.h
15:37:37 Join XavierGr [0] (n=xavier@rockbox/staff/XavierGr)
15:38:19Zagoramiconn: well plugin.h is also abused to include code for plugin.c
15:38:41Zagorinclude headers
15:40:34B4gderpower outage in the main rockbox server room on nov 22, 18:00 - 20:00 CET
15:40:57CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
15:40:57*B4gder is incorrect
15:41:05*B4gder shuts up and goes to the corner
15:41:18*B4gder is now whistling in the corner
15:41:21LambdaCalculus37B4gder: Remember that for when the 22nd rolls around. ;)
15:41:36*gevaerts planned to do his controversial commits then :(
15:41:45*LambdaCalculus37 thinks B4gder is predicting the future
15:42:10B4gderbtw, I have a crystal ball up for sale here
15:42:27*LambdaCalculus37 already has a nice crystal ball :P
15:42:57***Saving seen data "./dancer.seen"
15:43:03*B4gder switched to rolling dice
15:44:48*LambdaCalculus37 rolls some D20s and predicts that we'll have sound on the Sansa v2's by Christmas ;)
15:45:41B4gderno we won't
15:45:53B4gderon the samsa amses on the other hand... :-O
15:46:01Zagoramiconn: ...which has resulted in several other files also relying on plugin.h to get the function declarations they need.
15:54:46ZagorOh dear, it's a mess. I'll leave it for now.
15:56:38Zagormany plugins depend on preprocess symbols outside of plugin.h.
15:56:48 Quit Schmogel ("Miranda IM! Smaller, Faster, Easier.")
15:57:09Zagorwhich means our plugin api is not defined in plugin.h. it's spread all over the place.
15:57:57Zagorbut simply duplicating all those symbols is hardly optimal either
15:59:00LambdaCalculus37B4gder: +1 vote for "AMS Sansa" as the reference name for the newer Sansas.
16:03:32linuxstbUntil we have "AMS v2" (e.g. the revision 2 Clip, which IIUC has a different AMS SoC)...
16:04:14LambdaCalculus37Curse SanDisk and their constant changing of what's under the hood. :/
16:04:29B4gderlinuxstb: uuuh, let's not port to any such device! ;-)
16:04:54B4gderthe 3525sansas!
16:05:01linuxstbThat won't help though - we still need to call it something to say we don't support it... :(
16:05:07LloreanWell. these days we're less referring to them as a whole, and more referring to them on a device-by-device basis anyway, right?
16:05:12 Join robin0800 [0] (
16:05:47B4gderwe need a commit to test drive and celebrate lillebror's return to the build system
16:06:06n1sZagor: the plugin API aslo depends on structs like the user settings etc in the core
16:06:40Zagorn1s: yes but those are at least divulged in the plugin.h file
16:08:50 Quit J-23 (Remote closed the connection)
16:10:24 Join J-23 [0] (
16:15:24 Quit tvelocity (Read error: 110 (Connection timed out))
16:20:07 Quit MrDuck (Read error: 110 (Connection timed out))
16:20:41 Quit perrikwp (" ajax IRC Client")
16:23:55 Quit Darksair ("People who are zhuangbility want to show their niubility but only reflect their shability.")
16:24:54 Join tvelocity [0] (
16:25:17Zagorwe obviously haven't been very strict about plugins not #including core files
16:26:32 Quit robin0800 (Remote closed the connection)
16:26:50Zagorcalculator.c:#include "math.h"
16:27:32Zagorcalendar.c:#include <timefuncs.h>
16:27:33linuxstbWe even have math.h?
16:30:16linuxstbAnd also in libmusepack/math.h...
16:31:32Zagoryeah, but that's not what he wanted :-)
16:31:52 Join {phoenix} [0] (
16:37:31Zagoris plugins/midi2wav.c dead? it is never built and hasn't been touched for 3 years, other than generic code updates that accidentally touched it
16:39:31LambdaCalculus37Zagor: 3 years without an update nor building? I think that qualifies as dead.
16:40:02mooswasn't it replaced?
16:40:04Zagorfirst commit comment (3 years, 3 months ago) says "Will turn into real codec later, when plugins support the codec api."
16:40:23Zagormoos: yeah we have another midi plugin. but I have never used either...
16:40:53moosis stevenm still around time to time?
16:42:27Zagorlogbot says he was here 17 days ago
16:43:18 Quit meven ("Konversation terminated!")
16:44:34linuxstbZagor: I would say "svn rm" it. It will always be in the svn history for anyone that wants it.
16:47:07 Join Nibbler [0] (n=Nibbler@
16:50:21amiconnZagor: I would assume midi2wav is for testing some details of the midi synthesis
16:51:11Zagorperhaps. but it's very stale by now.
16:51:30 Join jgarvey [0] (
16:51:31 Join robin0800 [0] (
16:52:46LambdaCalculus37I also think that we should just "svn rm" it.
16:52:59LambdaCalculus37rasher: Ping
16:53:30linuxstbI don't think it's used for testing midi playback - looking at the logs, the midi player plugin has seen many changes which haven't also been made to midi2wav.
16:53:51moosmidi2wav was deprecated IIRC
16:54:07LambdaCalculus37It's basically rotting code in SVN. Might as well take it out before it stinks up the place.
16:54:29*linuxstb doesn't feel that strongly ;)
16:54:51Zagorgoing, going, gone
16:55:43*LambdaCalculus37 watches midi2wav fly away into the sunset
16:55:45 Quit suom1 (Connection reset by peer)
16:57:31LambdaCalculus37Zagor: I was thinking... rather than just dump the code away completely, should we just post it to Flyspray with a note saying "This code is no longer part of the official SVN. If you'd like to work on it for re-inclusion, feel free to"?
16:58:14linuxstbDo we really want a midi2wav plugin?
16:58:44ZagorLambdaCalculus37: I don't think we should have stale code in flyspray either
16:59:12LloreanWe're working at closing more "stale" tasks these days. Patches that are being maintained in current condition, rather than being improved toward commit.
17:00:08LambdaCalculus37Zagor: That's understandable. What's to say that it'll ever even see any work done, whether on Flyspray or SVN or anywhere?
17:00:44 Join suom1 [0] (
17:03:13moospitty that we swear off the idea proposed by MarcGay(?) about concentrating on few patches, to make them commitable....
17:04:01LambdaCalculus37moos: We shouldn't swear that off. There are some patches that show a lot of promise.
17:04:32 Quit suom1 (Client Quit)
17:05:08moosbut they are rotting for lo,g for few of them :(
17:05:33 Join Schmogel [0] (
17:05:39 Quit Zagor ("Client exiting")
17:07:45 Join suom1 [0] (
17:14:57 Join TheSphinX^ [0] (
17:17:03 Quit _lifeless (Remote closed the connection)
17:17:06 Join perrikwp [0] (i=d1a8d351@gateway/web/ajax/
17:21:22 Join _lifeless [0] (n=lifeless@
17:22:21 Join kugel [0] (n=chatzill@unaffiliated/kugel)
17:24:23 Quit moos ("Rockbox rules the DAP world")
17:25:05 Quit Nico_P (Remote closed the connection)
17:27:26 Quit wurscht (Remote closed the connection)
17:27:33 Join funman [0] (
17:31:01funmankugel: maybe you can do #ifdef SANSA_FUZE || defined(SANSA_E200V2) force the csd version to be 2.0, if these models don't have a 1GB version (if i understand correctly there is non-SDHC 2GB cards as well ?)
17:33:45rasherLambdaCalculus37: Yes?
17:34:14kugelfunman: that's what I did
17:34:20funmankugel: only if drive == 0 however, we don't want to force that for cards in the SD slot
17:34:28funmancan you access the disk now ?
17:34:46kugelfunman: I tried JdGordon's "magic" value, I got to the "loading firmware" with both csd 1.0 and 2.0
17:35:31funmanif the data accessed is within the 1st GB then it's normal
17:35:31kugelalso, my value which I got with csd 2.0 was obviously wrong as well
17:35:52funmanthen the answer is somewhere else :)
17:36:34funmantry to guess what is the meaning of the capacity reported by the card, now that you have the number of bytes reserved for the OF, and the value reported by dmesg
17:37:08funmanreal_value = dmesg + magic_offset, perhaps == csd_value*8 ?
17:37:20LambdaCalculus37rasher: I was going through the list of lang files in SVN and on your site, and was wondering about some of the translations that are "red" (i.e. less than 50% of their strings translated).
17:37:51kugelfunman: I strongly assume JdGordon's e200v2 value is also valid for the fuze
17:37:55LambdaCalculus37rasher: Are there any plans to do anything about the ones that are in really bad shape? (e.g. Afrikaans, Turkish...)
17:38:09funmankugel: sure, but that doesn't tell us how we can access the whole 8GB
17:38:28kugelI have 4GB, but there's a 8GB variant,yes
17:39:03rasherLambdaCalculus37: It's been discussed a million times, and we usually end up with a status quo, because there aren't any really compelling arguments to remove or keep.
17:39:58kugelLambdaCalculus37: nice to see you, I'd like you to test my latest backlight patch :)
17:40:28LambdaCalculus37kugel: Is it up on Flyspray?
17:40:58kugelfunman: I'm trying to build a fuze build, I'm getting many multiple and undefined references :(
17:41:04kugelLambdaCalculus37: not yet
17:41:10 Join smacnay [0] (
17:42:36funmankugel: normal, nothing is implemented :-)
17:42:59***Saving seen data "./dancer.seen"
17:43:29kugelfunman: multiple reference shouldn't happen though
17:43:40kugelmultiple definition*
17:43:48LambdaCalculus37rasher: Afrikaans was first committed five years ago, and since then, nothing's been done to it. It's already got 700 missing strings (only 10.83% translated). IMHO I would consider it "depreciated" and just remove it, but that's just my opinion.
17:44:21rasherLambdaCalculus37: But what if those 10% are just enough that someone might use it, or better yet, might be annoyed enough to improve it?
17:44:26funmankugel: you mean redefinitions ?
17:44:39LambdaCalculus37rasher: Point.
17:44:51kugelfunman: most likely
17:45:14rasherLambdaCalculus37: And they don't really do any harm. The worst that can happen if that people just don't use them
17:45:21 Quit BigBambi (Read error: 104 (Connection reset by peer))
17:45:27 Join linuxstb_ [0] (n=linuxstb@rockbox/developer/linuxstb)
17:45:29LambdaCalculus37rasher: Also point.
17:45:32 Quit linuxstb (Read error: 131 (Connection reset by peer))
17:45:38 Nick linuxstb_ is now known as linuxstb (n=linuxstb@rockbox/developer/linuxstb)
17:47:45funmankugel: in comment of lcd_enable: /* I'm guessing this function is lcd_enable, but it may not be... */ perhaps you can test that ?
17:49:14funmanwe need to know also what we can reuse directly of as3514 code, i expect a lot
17:50:25 Join miepchen^schlaf [0] (
17:50:58funmanlinuxstb: what's the interest of faking a partition table ?
17:52:55 Join bodymind [0] (n=bodymind@
17:53:21linuxstbfunman: I thought it may simplify things, but I'm not sure I still think that. We probably don't ever want to expose that partition via USB - it's too dangerous.
17:54:16linuxstbfunman: Did you try looking at the structure of the OF in the firmware part of the disk, to see if there's a header indicating the size?
17:54:42funmanat offset 0, there is the file we copy on the player
17:55:13 Quit lastebil_ (Remote closed the connection)
17:55:16 Join lastebil [0] (
17:55:52smacnayIf you were to buy a player today, which would you buy for supported features? I am looking at the Buyer's Guide but am unsure of which features are most important.
17:56:13funmanit always have the same size across different versions (and there is even only 2 possible sizes across different models)
17:56:32 Join jhMikeS [50] (n=jethead7@rockbox/developer/jhMikeS)
17:57:04 Join domonoky [0] (n=Domonoky@rockbox/developer/domonoky)
17:57:07gevaertssmacnay: how should we know?
17:57:20gevaertsImportance of features is a very personal thing
17:57:43 Join meven [0] (
18:00:49 Part B4gder
18:01:16kugelfunman: your audiohw_enable_output-stub in audio-as3525.c was kinda bad, since that is already declared in as3514.c (which is still assumed to share alot of code so it's compiled)
18:03:01funmankugel: i noticed, but this was not built for the Clip target
18:03:25 Quit miepchen^schlaf ()
18:03:50kugelah ok
18:04:03funmanif you intend to fix that, don't break the Clip !! :)
18:04:51*kugel can't get behind "loading firmware"
18:04:57bodymindhei.. my sansa e260v2... is up and running.. how can i confirm that the size of SD is correct?
18:05:54kugelfunman: btw, I've noticed the ocr is the same if sdhc is true or false
18:06:31 Join fragilematter [0] (n=fragilem@
18:07:33funmankugel: yes .. because the card isn't sdhc
18:07:59funmankugel: you can't get behind "loading firmware" if the FAT driver needs access to data beyond 1GB - sounds quite normal
18:08:43gevaertskugel: depending on how adventurous you feel you could try formatting it with a smaller filesystem
18:08:46funmanthe SD driver is also very unstable in the loaded firmware - i think i can get 1 time out of 10 to the menu, but the firmware is always loaded correctly from the FAT32 partition however, so it's more stable in the bootloader
18:09:24gevaertsfunman: does the bootloader use threads? If not, my money is on timing issues
18:09:25funmanbodymind: add a printf(); while(1); in ata_sd_as3525.c after the capacity is calculated (~line 260)
18:09:26kugelfunman: I mean, if I forced sdhc = true, the ocr answer was still the same
18:09:41funmanthe bootloader use threads
18:09:52 Join Llorean1 [0] (
18:10:06 Quit Llorean (Nick collision from services.)
18:10:09 Nick Llorean1 is now known as Llorean (
18:10:10kugeli.e. I removed the check for sdhc and replaced it with sdhc = true;
18:10:13funmankugel: did you read what the 'sdhc' variable does exactly ?
18:10:39funmanit sets (or not) the SDHC bit in the argument sent to the card, read the SD specification for the detailed process
18:10:48kugelit alters the argument of the SD_APP_OP_COND comamnd
18:11:15funmansdhc cards will not answer if this bit isn't set as far as i understand. they don't say non-sdhc cards will answer differently
18:12:28funmangevaerts: what do you think of timeout = 0x80000 in ata-sd-pp.c line 249 ? if the ticks is increased by 100 each second, that means 1h30
18:14:18 Quit petur ("work->home")
18:14:24gevaertsfunman: that's not ticks
18:16:19funmanoh right, because the PP don't have ticks in the bootloader
18:16:38 Join BigBambi [0] (n=Alex@rockbox/staff/BigBambi)
18:16:59funmanthen I don't understand this code ..
18:17:37kugelI might just try to run rockbox from my microsd
18:17:44gevaertsUSEC_TIMER is a microsecond timer
18:18:17funmanthat's 0.5s then
18:18:35funmanI've used 10s timeouts I think ..
18:19:07 Join MethoS [0] (
18:19:50fragilematterhey guys
18:20:28*amiconn doesn't like those hardcoded ad-hoc timeouts in the PP SD driver
18:20:30bodymindfunman: it printed a strange string
18:20:36fragilematterI got rockbox very close too booting on the e200v2
18:20:41bodymind card_info[drive].numblocks = c_size << 10;
18:20:41bodymind card_info[drive].capacity = card_info[drive].numblocks * card_info[drive].block_size;
18:20:41bodymind }
18:20:41DBUGEnqueued KICK bodymind
18:20:41***Alert Mode level 1
18:21:09funmanbodymind: can you code C?
18:21:13bodymindi've also tried: printf("%s",card_info[drive].capacity); :o
18:21:27bodymindsame thing.. |:
18:21:32funmansince it's an int, %d would be more suited
18:21:38kugelbodymind: capacity is an integer obviously, an unsigned even
18:21:42kugelso, use %u
18:21:55amiconnI assume that SD cards specify their timeouts in a similar way as MMC, and imho rockbox should use those timeouts
18:22:05bodymindhm, i don't know so low level c... :\ they keep insisting in java.. :|
18:22:19CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
18:22:19*domonoky tries to build a m200v4 build.. but its missing things...
18:23:08kugeldomonoky: heh, we all are trying to build builds for our ams sansas :)
18:24:43fragilematterfunman: when you booted rb on the clip did you have only the binary onto it or some other stuff?
18:24:53funmanfragilematter: I used make bin zip
18:25:19fragilematterand then extracted the zip on the root of the drive...
18:25:59funmanfragilematter: i thought it was clear the SD driver is not working on >1GB models
18:26:03fragilematterI'm getting a white screen after it loads rockbox.sansa, so I guess I need to get the zip to build
18:26:26fragilematterI moved everything to the start of the drive and it works for now :D
18:26:53funmanok ;)
18:29:53funmanamiconn: the SD spec mentions which timeout to use, for example for read it is the lowest of 100ms and 100 times the usual read access time
18:30:25funmanthe access times being reported by the card (in the CSD register)
18:30:42***Alert Mode OFF
18:31:25amiconnYes, that's what I mean. But iirc the PP SD driver uses fixed timeouts
18:32:45 Join bertrik [0] (
18:32:48fragilematterany fixed values for TIMER_FREQ, or should I use a dummy one just to get it to build?
18:33:04funmanhello bertrik, I was expecting you :-)
18:34:24*domonoky got it to build.. bootloader works, but it shows the rockbox logo, then a short flash of text and then the logo again ..
18:34:26funmanI was wondering what is the difference between i2c-pp.c and as3525-codec.c
18:34:47funmandomonoky: the SD driver is utterly broken in the non-bootloader environment
18:35:01 Quit bodymind ("Ex-Chat")
18:35:27funmandomonoky: is it the same logo ? on the Clip I can tell because the bootloader writes the version at the top of the screen (in the yellow part)
18:35:58amiconnfunman: Btw, I wonder why you disabled the SD driver init for the main build
18:36:21amiconnSo far the rockbox philosophy was that the main build should be as independent as possible
18:37:00funmanamiconn: binsize is an argument
18:37:27funmanit seemed also to improve stability when I tried it, but in fact it's not true
18:37:29LambdaCalculus37jhMikeS: Regarding FS #9312... I don't mind having USB charging only for now on the beast. It's working quite well here, and I understand that getting charging without USB is a pain right now.
18:37:49domonokyfunman: i dont know if its the same logo or not, but it looks like rockbox starts..
18:37:54jhMikeSLambdaCalculus37: I was thining of adapter charging, USB off for the moment.
18:38:20*jhMikeS wonders where he said "usb only"
18:38:21LloreanjhMikeS: Can it just have an option (in debug perhaps) to change the two?
18:38:26 Quit flux (Read error: 104 (Connection reset by peer))
18:38:28LloreanIf it's possible to do either one?
18:38:31linuxstbfunman: What's the status of SVN for the Clip? I've just tried compiling SVN bootloader and main build, and it boots, but freezes on the main Rockbox logo. Is that expected?
18:38:33*Llorean just stepped in here.
18:38:39*LambdaCalculus37 misread the comment... sorry
18:39:22bertrikI think the main clip rockbox hangs after mounting the disk, I tried putting a panic after the mount and I didn't go off most of the time
18:39:26jhMikeSLlorean: It's just a matter of working out how to properly power the unit with USB only _after_ the battery is charged.
18:39:32kugelwhat do I need to change BOOTDIR to in order to boot rockbox from my microsd?
18:39:50LambdaCalculus37jhMikeS: Sorry, that's where I misread.
18:39:50kugelLambdaCalculus37: did you already run a backlight test?
18:40:11 Quit robin0800 (Connection timed out)
18:40:11LambdaCalculus37kugel: Yep. Works nicely. Want to post a new patch?
18:40:41jhMikeSI mean, it will power it but it cycles over just a few minutes between charging cycle and draining the battery with no AC.
18:41:00bertrikfunman, i2c-pp is an i2c driver specific for the portalplayer chips
18:41:38*domonoky fixes SVN so m200v4 builds....
18:42:35kugelLambdaCalculus37: No, I'd like you to test if the old code is still working. To do so, comment #define HAVE_BACKLIGHT_THREAD_FADING out in your config-gigabeat-s.h
18:42:57linuxstbDoes the m200v4 have charging? I thought it just took replacable AA(A) batteries?
18:43:01bertrikare the mutexes in rockbox re-entrant? I mean will mutex_lock block or not when calling it twice in a row from the same thread?
18:43:02LambdaCalculus37kugel: Let me finish eating and I'll roll another build.
18:43:28kugelbertrik: do you have an idea?
18:44:14bertrikkugel, about what?
18:44:19*kugel would like to printf a ls of the root dir in the bootloader
18:44:34kugelbertrik: what to change BOOTDIR into in order to boot from the microsd
18:45:12domonokylinuxstb: yes, m200v2 uses 1 standard AAA Battery, but i dont know if it can recharge it...
18:45:27linuxstbdomonoky: I was referring to funman's last commit
18:45:42 Quit tvelocity ("Αποχώρησε")
18:46:25 Quit einhirn ("Miranda IM! Smaller, Faster, Easier.")
18:46:35bertrikkugel, no idea
18:49:11 Join nuonguy [0] (
18:49:47 Quit TheSphinX^ ("XChat@Linux")
18:50:14funmanlinuxstb: it's the same here, dunno what's happenning except that it's the SD driver's fault
18:50:35 Join mcflow [0] (
18:50:43linuxstbfunman: OK - so your photo of Rockbox working is a fake? ;)
18:50:50funmanbertrik: is as3525-codec.c an i2c driver for as3525 chips ? i.e. the equivalent of i2c-pp.c ?
18:50:55linuxstbOr was it bertrik's photo?
18:50:56 Quit {phoenix} ("Konversation terminated!")
18:51:00funmanlinuxstb: i made no photo, but i can get to the menu 1 time out of 10
18:51:10 Join webguest47 [0] (n=4491b125@gateway/web/cgi-irc/
18:52:01Lloreanfunman: So, what's with the "use simple charging" for m200v4 commit?
18:52:51*domonoky just got another result then the logo on m200v4 : partition not found.. :-)
18:52:51bertrikfunman, yes it basically an i2c driver specifically tailored for the codec part of the as3525
18:53:28 Quit webguest47 (Client Quit)
18:53:50funmanLlorean: warning killing
18:53:51 Join bmbl [0] (n=Miranda@unaffiliated/bmbl)
18:54:17linuxstbfunman: But afaik, the m200 doesn't charge - so that #define shouldn't be there.
18:54:43LloreanYeah, killing warnings that actually reference something that does need fixed is a bad practice, I think.
18:54:58linuxstbOr rather, it was the wrong fix to the warning...
18:55:33 Join {phoenix} [0] (
18:55:34linuxstbfunman: Was this a bootloader warning, or main build?
18:56:10*linuxstb sees it was a bootloader warning, and investigates
18:56:44 Part fragilematter
18:56:47LambdaCalculus37kugel: Rolling a build now with your patch applied, and #define HAVE_BACKLIGHT_THREAD_FADING commented.
18:59:24kugelfunman: I get disk_init failed on my microsd, even without having start +=
18:59:44LambdaCalculus37Oops... error in build due to FS #7331.
18:59:49*LambdaCalculus37 reverts that patch
18:59:54LambdaCalculus37Let's try this again...
18:59:55funmanLlorean: yes very right. I thought it would be correct since the other 'ams sansa' used this config already
18:59:57funmanthe exact problem is that usb_detect() use charger_inserted() (which is implemented), and charger_inserted was not defined
19:00:11linuxstbfunman: I'm about to fix (m200v4 charging)
19:00:12funmankugel: patch welcome
19:00:14*LambdaCalculus37 runs make clean and builds a second time
19:00:30funmanlinuxstb: then do it for every other 'sansa ams' targets
19:00:36amiconnIf the m200 has no charging, it shouldn't define it at all (same as the Ondio then)
19:00:40linuxstbfunman: ?
19:00:50funmancharger_inserted is implemented, and depends on CONFIG_CHARGING (in export/power.h)
19:01:07LambdaCalculus37amiconn: Neither should the c100.
19:01:12funmandoesn't the m200 charge batteries over USB ?
19:01:23LambdaCalculus37funman: AFAIK, no.
19:01:29Lloreanfunman: The m200 uses non-rechargeable AAs.
19:01:29amiconnWell, the m200 runs from AAA, doesn't it?
19:01:34LloreanEr AAA
19:01:39n1sjhMikeS: I have been messing a bit with the Gigabeasts' radio and can read and write to its i2c regs, so i was looking at the wmcodec driver to try to figure out how to enable the fm input and AFAICT setting the BOOSTENL and BOOSTENR bits in Power management 2 should be enough, do you think i need to do anything else?
19:01:46amiconnTrying to charge those would be dangerous...
19:01:49kugelfunman: you don't need to tell me "patch welcome" again and again. I know this by myself. I just thought it's interesting
19:01:56funmanok, then charger_inserted should be conditional on !M200V4
19:02:14LambdaCalculus37amiconn: I put a rechargable NiMH in my m200 with no ill effects.
19:02:28amiconnThat's not what I mean...
19:02:37amiconnI'm using rechargeables with my Ondio too
19:02:39funmanlinuxstb: in power-as3525.c
19:02:53 Join MrDuck [0] (
19:03:07funmanbertrik: what is the 'codec part' of as3525 ? I thought it was a synonym for i2c
19:03:11pixelmaLambdaCalculus37: but you didn't try to charge an alkaline in your m200, right?
19:03:46LambdaCalculus37pixelma: No, I never have.
19:03:58funmankugel: i'm not going to code or fix things for targets I don't own
19:04:02LambdaCalculus37I don't buy alkalines, anyway.
19:04:20 Quit FOAD (Remote closed the connection)
19:05:43bertrikwith codec part I mean the thing that controls audio, power, rtc, adc, unique id, it's basically an as3514. I2c is just the type of bus used to connect it to the rest of the as3525 and can be used to connect other things too (like an fmchip)
19:05:54linuxstbfunman: That seems an odd way to detect USB in usb-as3525.c - charger != USB
19:06:22kugelfunman: I didn't mean to ask you to fix that. I just told you because you might find that interesting
19:06:24 Join MethoS- [0] (
19:06:37 Join miepchen^schlaf [0] (
19:06:54 Join BigBambi_ [0] (n=Alex@rockbox/staff/BigBambi)
19:06:57funmanlinuxstb: do you mean a USB charger without an actual USB connection (like to a computer) ? I think I saw that in another target
19:07:21 Join FOAD [0] (
19:07:39linuxstbYes, you can buy usb chargers
19:08:05funmanit's in tcc77x
19:08:48linuxstbIn fact, I've just tried a USB charger on my Clip, and the OF correctly starts charging, but doesn't enter USB mode.
19:08:50funmanon the Clip there is a GPI pin which changes when USB is present, not sure if that's true when only a charger is there
19:09:14 Join bodymind [0] (
19:09:45jhMikeSn1s: The i2c driver as-is did the trick? it had never been tested for reads.
19:10:00funmanbertrik: ok, I was confused by the fact that i2c-pp.c use the ascodec_ api
19:10:07n1sjhMikeS: it seems so :)
19:10:36jhMikeSn1s: Also, I'd suggest looking at the block diagram for the WM codec. I usually draw paths on those with a pencil to make things easier.
19:10:47linuxstbfunman: Any objection to removing that check in usb_detect() for charger_inserted (usb-as3525.c)
19:10:59n1sjhMikeS: also i'm not sure if the clock from the pmic to the fm chip is enabled, do you know anything about that?
19:11:22jhMikeSn1s: the block diagram should that it is connected to 32.768kHz
19:11:30kugelLambdaCalculus37: any updates from the backlight fron?
19:11:47LambdaCalculus37kugel: Still building.
19:11:54n1syes, but would it need to be enabled in the pmic somehow?
19:11:58funmanlinuxstb: no, we don't have a better solution at the moment, but we don't need it either
19:12:31 Quit Slasheri (Read error: 145 (Connection timed out))
19:12:34linuxstbfunman: No, in fact it's probably better to not detect USB at the moment anyway.
19:12:37funmanby the way the as3514 RTC works fine , but it's not powered while the Clip is off
19:12:45 Join BigBambi__ [0] (
19:12:58 Quit BigBambi__ (Read error: 104 (Connection reset by peer))
19:13:05jhMikeSn1s: I'm guessing no but I'm not positive. It has CLK32KMCU and CLK32K. You don't have the datasheet?
19:13:20 Quit MethoS (Read error: 60 (Operation timed out))
19:15:13bertrikfunman, too bad about the clip RTC. I hope we can find a way to keep it enabled somehow
19:15:41n1sjhMikeS: yes i do, but they usually confuse me so i though if you knew off the top of your head :)
19:21:27bodymindhei... my sansa e260v2 only detects 1027342336 bytes of memory.. it should have 4Gb.. :x any workaround?
19:22:22jhMikeSn1s: Power Control 0 bit 6
19:22:43funmanbodymind: in fact I had understood that you would look for the real problem
19:23:00jhMikeSSays it's on by default. I have a cold boot bit list somewhere to check that.
19:23:07bodymindonly testing here :$
19:23:15bodymindi'm not familiar with rockbox code.. :s
19:23:31kugelOk, what do we need BOOTDIR for, if it's not used in load_firmware()?
19:23:53jhMikeSyeah, they have it off by default
19:24:27 Quit BigBambi (Read error: 110 (Connection timed out))
19:24:36 Quit meven ("Konversation terminated!")
19:24:57kugelhe found rockbox.sansa on my microsd
19:25:00kugelbad checksum though
19:25:21bertrikn1s, I stumbled on an app note for the si4700/01 chip
19:25:38 Join BigBambi [0] (n=Alex@rockbox/staff/BigBambi)
19:26:24n1sbertrik: i've got a couple of different docs from jhMikeS, which did you find?
19:26:24jhMikeSbertrik: a new one? I have a whole collection of docs for it.
19:26:33bertrikit's a programming guide that explains how to set it up for various basic operations
19:26:44n1sthe "AN230" ?
19:26:46bertrikit's called AN230Rev0_3
19:27:08n1sI have that, but thanks anyway :)
19:29:36 Quit BigBambi_ (Read error: 110 (Connection timed out))
19:29:59jhMikeSn1s: I guess I had a couple duplicate docs in that set :)
19:30:35kugelanyone know how to fix the checksum?
19:30:46n1syes, but no worries, I like the one with screenshots of how to install some software for a dev board in windows though :P
19:33:18bertrikn1s, this radio chip basically tunes itself, it shouldn't be that hard to get some sound of it I think
19:33:31kugelfunman: I succeed with my attempt to load from microsd. only checksum is wrong now, any idea (I remember I read that lately so someone might have had the probelm as well)
19:33:58n1sbertrik: it turns out that supplying the clock to it is probably a good idea :)
19:34:19LambdaCalculus37kugel: Your code for the backlight is still working.
19:34:21bertrikn1s, but you could already communicate with the chip without the clock?
19:34:47kugelLambdaCalculus37: so you have no backlight fading? or the half-baked one which is present on the beast?
19:34:48jhMikeSthe 32K clock doesn't control the serial then
19:35:07funmankugel: FIRMWARE_OFFSET_FILE_* (in config-clip.h) were not defined correctly, but it looks correct for the fuze. The calculated and read checksums should be printed on the screen, by how much do they differ: a few only ?
19:35:21n1sbertrik: yes, the datasheet even states that you can at least set some bits in i2c regs before supplying the clock
19:35:21pixelmaeh? Now also produced a weird failed build... maybe some timing issues after the changes for "make" lately?
19:35:37kugelfunman: no, it's quite a bit difference
19:35:40bodymindfunman: the problem is that the MULTIVOLUME is not activated.. :o
19:35:47funmankugel: then the data read is likely wrong
19:35:48LambdaCalculus37kugel: I currently have no backlight fading on the beast.
19:35:51kugelbodymind: you can activate it
19:36:00bodymindyes.. just to know :P
19:36:19jhMikeSn1s: so you're positive the correct values are being read and written with i2c? I just want to make perfectly sure in case that needs some debugging.
19:36:21bodymindwith multivolume activate is shows 4Gb on sansa260v2
19:36:46funmanbodymind: 4GB for your internal storage ? is it correct ?
19:36:58kugelfunman: checksum 318cc9e, sum: 1db9f01
19:37:17bertrikan incorrect checksum probably means that reads from the flash are corrupted
19:37:38n1sjhMikeS: not _positive_ but i get the correct chip ID and manufacturer ID 's a chip revision that makes sense and if i change bits and read them back they have changed as expected
19:39:07 Join aarcane [0] (
19:39:13jhMikeSn1s: well, if that's correct then it sounds alright
19:39:35kugelfunman, bertrik: the len is quite right though. it reads the whole file
19:40:06funmankugel: print the data byte per byte and check if they are correct
19:41:00funmanZagor: " SP - Normal" < some dependancy problem with make -j ?
19:41:00jhMikeSn1s: I'm just surprised it worked since I just guessed using the flowchart in the imx datasheet :p
19:41:08 Join nanok [0] (n=nanok@
19:41:33bertriksometimes, just sometimes, things work they way they are supposed to :P
19:41:41nanok q: are there rockbox targets which have a visible display without the backlight? (i assume the monochrome lcd ones do, so i guess i mean the color ones)
19:42:15nanokbertrik: for that, there's always some big and blue code around the corner to fix everything..
19:42:34n1sjhMikeS: if i read with subaddresses it doesn't seem to work but a raw read of the whole regset does
19:43:04***Saving seen data "./dancer.seen"
19:43:42 Join Horscht [0] (n=Horscht@xbmc/user/horscht)
19:43:53Bagdernanok: I believe one or a few of the ipods have such screens
19:43:57LambdaCalculus37nanok: The iPod color has a visible display with the backlight off.
19:43:58funmanare you hacking the GigaBeat FM chip ?
19:43:59jhMikeSn1s: well, then there is a small bug or it needs a tweak for that chip
19:45:11jhMikeSn1s: you mean if you set reg to -1 and read the whole chip?
19:45:18pixelmananok: the Nano's and the Ipod Video's displays are quite readable without backlight (the 4th gen Ipod Photo too but the one that I have seen was not as good as the former)
19:45:37 Quit amiconn (" rb")
19:46:06nanokhow about the iaudio x5?
19:46:28nanok(thanks btw, that's very usefull already)
19:47:02Bagderthe x5 is not readable without backlight iirc
19:47:05pixelmablack without backlight and as it's not readable anyways Rockbox switches the display off
19:47:27Bagderright, but it does so because it is unreadable
19:47:56nanokhowever, if i manage to get an x5, i think i;ll have to live with it anyway. i am much inclined towards the iaudio, for some reason..
19:48:02nanokthanks a lot
19:48:06pixelmathat's what I meant
19:48:22 Join amiconn [50] (n=jens@rockbox/developer/amiconn)
19:48:59nanokpixelma: well, rockbox is known to squeze every possible minute of battery lifetime from the x5, so i guess it does right whatever it does
19:49:09nanokdoes it also swith the "video" chip off?
19:49:17nanokor am i talking rubish
19:49:46bodymindfunman: can i comment out card_detect_target() on line 533 of ata_sd_etc..
19:50:10bodymindit's undeclared... and it's only called if HAVE_MULTIVOLUME is not defined.. :s
19:50:12nanokoh, one more question: i found some 2 and 4g nanos, but not sure how to tell if they are first generation (if i am not mistaking it's the only nano which is rockbox-worthy)
19:52:43kugelundefined instruction at 3000024
19:54:05LambdaCalculus37nanok: Black or white front and chrome back is the best way to tell.
19:54:43jhMikeSn1s: Does subaddressing work with writes?
19:54:48nanokaham, so the chrome back is a dead giveaway
19:55:07nanokone i have seen, with a chrome back, had also printed on it a manufacturing date of about 2005
19:55:18nanokso i guess that makes sense
19:55:24nanokLambdaCalculus37: thanks a lot
19:56:30n1sjhMikeS: is that implemented? i just do i2c_write(&si4700_i2c_node, bar, 8);
19:56:34 Join tyfoo2 [0] (
19:56:48jhMikeSn1s: subaddressing is implemented for r/w in the driver
19:56:54n1swhere 'bar' is an array 8 unsigned chars
19:57:18jhMikeSI'm not sure if the chip supports it actually. It's not clear.
19:57:47jhMikeSthe philips IC doesn't so it wouldn't be so odd
19:57:48n1sI may of course have messed something up so don't take this with too much credibility
19:58:00 Join petur [50] (n=petur@rockbox/developer/petur)
19:58:02bertriksubaddressing is just an extra byte before the other data, and as far I've read the datasheet, the fm chip does not use sub-addressing
19:58:22pixelmaBagder: did you see the failed OndioSP build on Looks similar weird as the coldfire build which failed on lostlogic's server earlier, so maybe there's another thing going on there?
19:58:25jhMikeSbertrik: right. it looks like it may not to me as well.
19:59:03Bagderpixelma: yes indeed.... I think I'll re-enable lostlogic's server again and see if we get this in a way that makes us understand it
20:00:23bodymindwhat is: ATA error: -2 ?
20:00:51bodymindi've enabled MULTIVOLUME, and now i have this error... instead of: firmware not found
20:01:47bodymindand comment the the function call: card_detect_target() line 533 ata_sd_as3525.c
20:02:11n1sBagder: Iäm not sure how rombox works but that failed build tried to link it before a lot of the core was compiled...
20:02:14 Quit BigBambi (Read error: 113 (No route to host))
20:02:15nanokarghh, just talked to a potential iaudio x5, battery is quite dead (5-6h), so the natural question arrises: is it possible to replace that battery? where to get such battery?
20:03:06jhMikeSnanok: sure is. google, google
20:03:22nanokjhMikeS: thanks, will do
20:04:30bodymindanyone on e200 ?
20:05:10LambdaCalculus37nanok: And eBay, eBay.
20:07:51nanokLambdaCalculus37: yup, readily available on ebay, radioshack
20:08:11kugelfunman: do we even need FIRMWARE_OFFSET_FILE_DATA?
20:08:37nanoki am thinking of driving 250km tomorrow evening just to get the unit from the owner :)
20:08:50kugelthe comment says it's only needed for the old rockbox.e200 format (which happans to be a special format, with e.g. 512byte padding)
20:10:43n1sjhMikeS: enabling the clock made some more things appear in the i2c regs, like the firmware version which was previously "00" is now "0E" etc, still no sound out but i guess i'll save that for tomorrow
20:11:24funmankugel: see bootloaader/common.c
20:13:09kugelfunman: I see, the comment is missleading
20:13:15 Quit tyfoo (Connection timed out)
20:13:28nanokso, iaudio x5: battery sorted out (sort off). what elese should i be careful of: that pesky port replicator (useless without it), and besides that? any common faults i should check before buying?
20:14:06nanok(note: please feel free to shut me up if i am disturbing/being offtopic ;) )
20:14:10bodymindfunman: what is ATA error: -2 ?
20:14:57kugelno card found iirc
20:15:39bodymindhm, have_multivolume is to activate the microSD?
20:16:14kugelbasically, yes
20:16:19 Join Slasheri [0] (i=miipekk@rockbox/developer/Slasheri)
20:18:36bertrikn1s, nice to hear you have it almost working. I should have a good look at how it is connected in the clip soon
20:21:10n1sbertrik: i2c seems to be common for this kind of thing
20:23:26n1sand the beast has 3 gpios connected to it too
20:23:42n1si guess you need something similar at least
20:24:29 Join Lear [0] (i=chatzill@rockbox/developer/lear)
20:26:22 Join Strife89 [0] (n=michael@
20:26:26bodymindfunman: i found out that the calculation method used is wrong... it should use /* CSD version 2.0 */ instead of version 1.0
20:26:27 Quit Thundercloud (Read error: 104 (Connection reset by peer))
20:26:59funmanbodymind: this was suggested by kugel as well, but he later changed his mind, I don't know why
20:27:20bodymindi've tested right now
20:27:46bodymindthe first version shows up 1Gb the second 2.0 shows the right size.. 4Gb
20:28:17kugelfunman: I changed my mind, because the reported size is obviously wrong too (although it's much nearer to 4GB)
20:29:07kugelthe of offset I get with that size and dmesg is bigger than jdgordons value, which leads to "loading firmware"
20:29:30bertrikn1s, the i2c pins are shared with the GPIOs on the as3525 which already seem to be overpopulated
20:30:59funmankugel: what is the difference between the correct size and the size reported if you use CSD2.0 ?
20:33:56bertrikfunman, I wrote a little csd parser, want to have it?
20:34:22funmanhum the PP ascodec-target.h is in target/arm .. then we can't use a ascodec-target.h in as3525/
20:34:33jhMikeSn1s: and the power handling? HAVE_TUNER_POWER_CTRL?
20:35:02funmanbertrik: well I won't use it, I only have a 1GB card which works fine, better propose it to kugel. Or post it on the forum?
20:35:26 Join culture [0] (
20:36:02funmanbertrik: or maybe we should use this header for both PP and as3525, what do you think?
20:36:16kugelfunman: my value was 157696, jdgordons is 61440 (number of blocks)
20:36:41n1sjhMikeS: i have not done any integration yet, everything lives in a debug screed :) but i imagine we could just disable the chip and the clock (which IIUC should keep all the reg settings intact) when it's not used
20:36:45funmanbertrik: hm it might be useful when messing with the reported timeouts.. just put it somewhere ;)
20:37:51n1sdisable the chip == setting the DISABLE bit, but not pulling RST low
20:38:05funmanthe difference is 47*2048 or 188*512
20:38:19jhMikeSn1s: it comes to mind quite a bit needs integrating, like the audio muxing stuff. yes, disable the chip and clock and i2c interface in tuner_power.
20:38:44kugelfunman: according to previous reports, the firmware should load on >1GB devices too, if rockbox.sansa is within the first GB, right?
20:39:28kugelfunman: Ok, well. I get exact the same bad checksum when loading from internal sd
20:39:47jhMikeSthe only thing on i2c2 is the FM but the codec and TV enc share i2c1
20:39:59n1sjhMikeS: yes, i saw that no source selection seems to be implemented in the w8978 driver, but i prefer to get a horrible working mess first and cleaning up that than doing everything "right" from the start...
20:40:03kugelif I ignore the checksum and load it anyway, I get undefined instruction
20:40:25n1sjhMikeS: ah, yes
20:40:57*jhMikeS is usually somewhat of a neat freak from the start :) :P
20:41:15funmankugel: what is the reported block length ? (I think it's fixed in the driver to 512)
20:41:17kugelfunman: does the difference help you? I say it again, my value didn't work at all (disk_init error)
20:41:40kugelfunman: it's fixed. 512bytes
20:41:50funmankugel: i'm just trying to give you suggestions, i have no clue
20:42:01kugeldmesg also reports 512bytes if you mean that
20:42:21funmankugel: bits 83:80 of CSD reports the maximal block len for reads
20:42:37funmanwhat dmesg reports is not linked to the SD card
20:43:25kugelyou mean card_info.block_size
20:43:36 Join Thundercloud [0] (
20:44:01bodymindin dmesg it reports: [ 5093.583245] sd 15:0:0:0: [sdb] 7964672 512-byte hardware sectors (4078 MB
20:45:36kugelfunman: I read it by printf'ing "card_info[drive].csd & (0xff << 80)" right?
20:46:15 Join Vorador [0] (
20:46:24funmanno, csd is a int[4]. also be careful because the 4 ints are stored in a weird order
20:46:48*jhMikeS still wants a nice suggestion to incorporate HW EQ along with software.
20:47:22kugelfunman: ok, let me read the specs again
20:49:21n1sjhMikeS: all i can think of is a separate hw EQ screen (which would be weird and confusing) or a limited "Bass" and "Treble" thing similar to other targets (which would not use the full potential of the chip)
20:49:46bodymindkugel: i'm printing only the capacity...
20:50:14kugelbodymind: I'm already beyond that :)
20:50:21jhMikeSn1s: It's puzzling me why we get to a certain point and then want to limit the use of hardware features when the device has "too much".
20:50:24kugelfunman: card_info[drive].csd[2] & (0xf << (80-64)) should work now
20:51:38Unhelpfulan ugly hack might be to have one set of EQ settings, that are translated to hardware settings when possible, and applied via DSP when not... but that's just as confusing as two different EQs :/
20:52:21n1sassuming that it cannot do everything the SW eq can but works similarly maybe we can have some kind of seamless transition or only apply parts of the SW and do what's possible with HW?
20:52:49jhMikeSThey would make louse bass/treble controls. The switchover wouldn't integrate well either...different filters.
20:52:54kugelfunman: does 0x90000 make any sense?
20:53:21funmanyou mean 9
20:53:28Unhelpfuln1s: the thing that i suspect would confuse people there is 1) filters might sound different 2) why does battery life decrease when i change this bar in the EQ?
20:53:34jhMikeSIt would be improbable anything SW would match HW
20:53:42funmanif it's 2^9 then it's the hardcoded 512 we use
20:53:56funmanbut I just suppose the unit is powers of two
20:54:00amiconnjhMikeS: What target does have real hw eq?
20:54:07jhMikeSamiconn: the beast
20:54:41kugelfunman: 9 could be right, damn endianess
20:54:41kugelprintf gives 0x90000
20:54:41n1sjhMikeS, Unhelpful: then I'd say go with a separate screen and put a nice description of the differences in the manual.
20:54:46Unhelpfulexposed as presets in the OF, i believe. is the configurability *similar* to rockbox EQ?
20:54:53*amiconn didn't know
20:55:14funmankugel: 80-64 = 16; so 9 << 16 = 0x90000
20:55:23jhMikeSthe audio chip is rather elaborate
20:56:12amiconnIf a target has real hw eq, I'd think we should offer it and drop sw eq instead
20:56:18nanokjhMikeS: but surely the hw eq is not as flexible as the rockbox parametric eq (?)
20:56:34*amiconn only uses simple treble/bass anyway
20:56:43jhMikeSit's not, that's why the SW still has appeal
20:57:18jhMikeSunless the bands could be "joined" in such a way as to imitate bass/treble controls
20:57:21nanokjhMikeS: so the question is how to implement it from the point of view of the user experience, do i understand correctly?
20:57:29Unhelpfulcompile-time option, or a screen for each, and perhaps a setting to decide which is the "default" eq?
20:57:51jhMikeSnanok: if we use both to allow full HW use and not lose the SW advantages, yes.
20:58:01kugelfunman: ok. I assume my build is broken, that explains why it fails on both internal and microsd for the exact same reason
20:58:35n1sThe more i think about it, two eq screens seems like the best way to go, but maybe it's time for one of those fun mailing list discussions?
20:59:02*amiconn thinks there are a bunch of more important problems to solve on the beast
20:59:11jhMikeSn1s: hehe. well, it could be indeed. If somehow it's clear enough, I can't imagine too much confusion.
20:59:44nanokjhMikeS: Unhelpful how about a config option (themes or such?) which can either display both in the menu, disable one altogether or the other? that should be clean enough, no fuss for those who don't want it, no problem for those who do want to choose/have both
21:00:01jhMikeSamiconn: it came up, just because. I think DMA and scaling should be high up.
21:00:07bodymindis there anymore testing or anything i can do for e200v2port?? :'|
21:00:30Unhelpfulmy "which is default" idea is along the lines of displaying EQ and HW EQ, or EQ and SW EQ, as separate menu items.
21:00:32n1snanok: IMHO the problem is not people that don't _want_ it but people who don't understand the difference...
21:00:32amiconnjhMikeS: What annoys me most atm is that pcm (i.e.) voice)
21:00:32jhMikeSnanok: we don't do configurable menus (that can be worse)
21:00:50amiconn...doesn't work after boot before playing music
21:00:54funmanhow can I use target/arm/sandisk/adc-target.h for 'ams sansa' as well? I was thinking put a adc-as3514-target.h in target/arm/ and include it in */adc-target.h
21:01:30*n1s wonders how many of our users know what Q means in an eq anyway...
21:01:32jhMikeSamiconn: ??
21:01:58 Quit nuonguy ("This computer has gone to sleep")
21:02:08nanokn1s: i get it. so the idea would be to burry one of them somewhere deep enough to not disturb the "innocent", but accessible enough to be usable by the "initiated" :)
21:02:25amiconnjhMikeS: Enable voice (you also need a voice file, obviously). After boot voice doesn't work. If you play some music, voice starts working. It then continues to work until shutdown
21:02:46amiconnOne of the reasons why I don't use the beast for real
21:02:53funmanwell or put it in as3514.h
21:03:22*amiconn leaves for a while
21:03:24jhMikeSamiconn: not intentional. something must just be overlooked like an init but I can't imagine what.
21:03:26nanokn1s: in all honesty, i first found out what a parametric eq actually is while reading the rockbox manual, and i am gratefull for that (no kidding). so i guess you could look at it that way too
21:04:12nanokn1s: the next thing i did was go and buy a sansa (the most accessible rockbox target at the moment), thouhg i had a music player i was quite happy with
21:05:02n1snanok: heh, then i guess you are in the category that will not be confused by "Hardware EQ" and "Software EQ" ;)
21:05:59kugelfunman: I have similar problems. I can'*t get the "implicit declaration of function ‘ascodec_read’" warning away
21:06:05nanokand now, to put things into perspective, i am wondering why a parametric eq is not "standard feature" for most computer music player (i guess it would be so easy, so much less limitations than when coding it for rockbox targets)
21:06:11funmankugel: svn up ?
21:06:41kugelfunman: haha ok
21:07:12kugeldidn't notice the commit. THat happens if you don't sit 24/7 pressing F5 on the rockbox page
21:08:06nanokn1s: i remember what a proeminent member of the oss community once said (while rather angry, granted): somehting along the lines "if you will assume your users are dumb, they will become [be naturaly sleected to be] dumb" i'd dare to add "if you assume they are smart, they might take the time to smarten up" ;)
21:08:49kugelhe's wrong
21:08:56n1sthat might very well be true :)
21:09:34nanokkugel: are you the guy always seeing the holes in the cheese? :)
21:09:51funmana dose of tolerance is also needed :/
21:10:00kugelnanok: I'm the guy who EATS the holes in the cheese :p
21:10:14nanokkugel: let me dream, at least. i have enough reality every day at the office :)
21:10:23kugelok :D
21:10:26nanokkugel: oh damn. join the club..
21:11:49nanokkugel: well, this is not "the office". our users are smarter than most users.
21:12:11nanoksorry, prefix that with "independent studies clearly show"
21:13:08kugeloh, I already thought you lie :p
21:13:20kugelbut that studies impress me ;)
21:14:28nanokkugel: let me rephrase that: 82% of our users
21:14:35*domonoky moves the offtopic chatters to #rockbox-community.. :-)
21:14:49nanok(yeah, the 80-20 bullshit rule, with a 2% twist for credibility)
21:14:56nanokdomonoky: aye sir
21:15:42 Quit atrus ()
21:18:23 Quit bodymind ("Ex-Chat")
21:19:48 Quit nplus (Remote closed the connection)
21:19:49 Join faemir [0] (
21:21:57J-23hmm, why isn't as3525-codec.h in the SVN repo?
21:22:19kugelfunman: you forget to edit sansa_as3525.c to include ascodec-target.h
21:22:21funmanJ-23: check the last commit message
21:22:49funmanoops I didn't test the bootloader
21:23:41 Quit intrados ("WeeChat 0.2.6")
21:23:44kugelfunman: haha, tell you you haven't checked the build table :)
21:24:46funmanI didn't :)
21:25:10linuxstbfunman: ascodec-target.h was intended to be a PortalPlayer specific file - i.e. there should also be an as3525/ascodec-target.h
21:25:46n1sWow, that latest ondio SP build failure is pretty weird...
21:26:39funmanlinuxstb: it looks like a as3514 specific file, not portal player specific, and can be shared with the as3525
21:27:14linuxstbAnything to be shared should be in ascodec.h, not ascodec-target.h
21:27:55funmani'm a bit lost with this system
21:28:12funmancan you do it the correct way (tm) ?
21:28:27kugelwell, that's probably because "mr.someone" didn't clean the target tree system
21:28:39linuxstbThe problem is that there isn't (currently) a portalplayer target directory - so lots of files in target/arm/ should really be in target/arm/pp/ The comment in ascodec-target.h that you removed said that.
21:28:52funmanI noticed it
21:28:57kugelthe old manufactor/ (not cpu/)is messing a bit
21:29:00funmanBut I made this file not Portal Player specific
21:29:26linuxstbBut it is specific - it has portalplayer code in it.
21:29:37linuxstb(with #ifdef...)
21:30:16kugelwell, I'd probably be mr. someone for that clean up, but as long as svn mv fails with svn diff, I don't really feel like
21:30:21 Join __lifeless [0] (n=lifeless@
21:30:33jhMikeSamiconn: (for logs) does the volume setting get applied before voice is played?
21:31:06funmanlinuxstb: I was about to modify it again, do you want to move the functions into a specific folder, so we keep that files for the declarations ?
21:31:26funmans/do you want to/can you please/
21:31:34linuxstbfunman: It also depends if the implementation for as3525 are going to be "static inline", or if they are going to be in a .c file.
21:34:08kugelfunman: so, I'm able to load rockbox.sansa from either external or internal memory (depending on whether the card is inserted). I can't get either to load, both times check sum fails
21:34:54kugelwhich I don't really understand since the filesize is read correctly, and the sum calculation is in a for loop which loops len(=filesize) times
21:36:57bertrikhmm, I think the as3525-codec.c should have locking
21:37:03kugelfunman: if I try to load it anyway, I get a undefined instruction at 0x3000024 which happens to be crt0.s area
21:37:30linuxstbfunman: I can't do it for another 10 minutes or so (I'm in the middle of something else), but I think you simply need to revert your change to arm/ascodec-target.h, and move the as3525 specific bits of it to target/arm/as3525/ascodec-target.h
21:37:57 Join stripwax [0] (
21:39:41kugellinuxstb: do you know how the checksum is calculated (the one in the first 4bytes of the rockbox binary?
21:39:45 Quit Nibbler ("Ex-Chat")
21:41:48 Join krazykit [0] (
21:42:20funmanlinuxstb: but won't arm/ascodec-target.h be included before arm/*/ascoded-target.h ?
21:42:54funmanthat is the reason why I have put the as3525 code in it
21:43:07***Saving seen data "./dancer.seen"
21:43:28funmankugel: see bootloader/common.c:189
21:43:37 Quit J-23 (Remote closed the connection)
21:43:58kugelfunman: I mean the checksum written in the header
21:44:19funmankugel: well it's calculated in the same way of course ..
21:44:27linuxstbfunman: No, the more specific files (i.e. deeper in the tree) should be included - that's my understanding anyway.
21:45:18funmanlinuxstb: ah right (TARGET_INC in root Makefile)
21:46:09 Quit _lifeless (Read error: 110 (Connection timed out))
21:48:38 Join J-23 [0] (
21:49:40 Join Nibbler [0] (n=Nibbler@
21:49:45 Quit Strife89 ("Stuff to do elsewhere......")
21:50:04kugelanyone has an idea why the BOOTDIR #define isn't used in load_firmware? it's hardcoded to /.rockbox
21:50:04 Join flux [0] (
21:51:56funmanlinuxstb: ok
21:56:56 Join BigBambi [0] (n=Alex@rockbox/staff/BigBambi)
21:57:50linuxstbfunman: BTW, I think the bootloader should just use ascodec.h, rather than ascodec-target.h.
21:58:24funmanoh ?
21:58:40 Quit LambdaCalculus37 (" ajax IRC Client")
21:58:53funmanthis ascodec header looks pretty empty ..
21:59:26funmannow I think I'm done with the commit frenzy ..
22:01:53amiconnjhMikeS: It should, during settings_load()
22:03:25jhMikeSamiconn: ok. Starting music shouldn't do anything special over voice, so atm I'm stumped.
22:04:21funman.. or not
22:04:33amiconnfunman: Loadsa red
22:05:02jhMikeSamiconn: is there anything peculiar about the way it doesn't work? (something other than just being silent, for example)
22:05:16funmanamiconn: yup it's already fixed (for the sansa ams bootloader)
22:05:24amiconnjhMikeS: It's silent
22:05:26funmanhmm more red
22:05:44jhMikeSamiconn: but eats PCM or not?
22:05:57amiconnHow would I know?
22:06:41*amiconn didn't try to track down the cause so far
22:06:43jhMikeSamiconn: I don't know. I thought you could have investigated some stuff/
22:07:13amiconnIt's just a reason why I don't use the beast. /me has lots of other choices
22:07:45jhMikeScould you email me a voice file if that isn't much trouble?
22:08:09amiconnAnother annoying thing (but probably unfixable for a looong time) is the excessive boot time
22:08:15 Quit {phoenix} (Remote closed the connection)
22:08:59jhMikeSunless we flash it it's pretty much stuck just like the F/X
22:09:15 Quit J-23 ("Flying cow pressed ^D on my keyboard.")
22:09:30jhMikeSat least if you need retailos. without that it's not so bad.
22:09:33XavierGrdidn't kkbjurn had some plans about Gigabeat F flashing?
22:09:45markunXavierGr: he's still working on it
22:09:51XavierGrmarkun: good to hear
22:10:01markunin fact it's working find on his player, including wakup alarm
22:10:16markunfine even :)
22:10:33XavierGrand what is the obstacle for further usage?
22:10:54 Join MethoS-- [0] (
22:10:59amiconnIf I start my archos recorder and my beast at the same time, the recorder already resumed playback when the beast is about 2/3 through the windows mobile progress bar...
22:11:10funmanamiconn: r19079 should fix it - do the build servers build each revision, or the most recent revision when the build has completed ?
22:11:21XavierGrmarkun: I mean what's holding it back for official support?
22:13:36amiconnfunman: Most recent. If the master is idle when a commit happens, it also waits 60 seconds before starting the build, and if another commit happens during this delay, it's also included
22:13:43jhMikeSamiconn: the native loader loads the entire firmware image first. with just a pure rb bootloader it's hardly seen.
22:14:31amiconnAlso, builds don't run back-to-back. The master does some post processing (binsize calculation...) before it's ready to pick up the next changes
22:14:50jhMikeSamiconn: having things setup to support a sleep mode (like you were going to do) would also help :)
22:15:12BigBambiThe beast boot is pretty quick without dual boot
22:15:21BigBambiWith dual boot, it does indeed take ages
22:15:22amiconnI wasn't going to do sleep mode on anything but PP5002, and even that's on hold atm
22:15:44 Join saLOUt [0] (
22:16:29*amiconn still didn't gather enough motivation to wade through powermgmt and sys messaiging
22:16:44 Quit petur ("let's try to install a new video driver")
22:16:55amiconnasm is more fun... and less complex
22:17:27 Quit fredddy (Remote closed the connection)
22:18:06amiconnBigBambi: How do you charge without dual boot??
22:18:14BigBambiWith the patch
22:18:29BigBambiWhich works very nicely
22:19:53 Quit aarcane ("reboot timez!!")
22:20:29amiconnI know there is a patch... but I think it's too complex to safely apply it to my main tree
22:20:40BigBambiWell, that is how I charge it.
22:20:54BigBambiAnd without the OF, the boot is quick
22:21:26funmandbg_ports is displayed in the menu conditionally, but is built unconditionally (in debug_menu.c)
22:21:28jhMikeSI don't see it being too much simpler other than to rework chargning code with it in place
22:21:50amiconnDoes the patch also fix battery status? In SVN it seems like battery status uses rand()...
22:21:51 Quit Nibbler (Read error: 113 (No route to host))
22:22:00 Join bluebrother [0] (n=dom@rockbox/staff/bluebrother)
22:22:11n1sanother thing that is bad about the current dualboot is that it's pretty easy to accidentaly boot the OF and then ypu're stuck if you don't have a paperclip
22:22:22BigBambiI'd love to see the patch committed, but I can't comment on the technical side :)
22:22:23jhMikeSI did cut some cruft in my own tree though
22:22:55 Join aarcane [0] (
22:25:26aarcanehaha, new kernel and terminal emulator
22:26:14jhMikeSamiconn: the problem mentioned earlier?
22:26:17 Quit Lear ("ChatZilla 0.9.83 [Firefox 3.1b2pre/20081110034911]")
22:26:18BigBambiaarcane: ?
22:26:29aarcanesorry, I meant that for the channel below this
22:26:58 Quit MethoS- (Read error: 110 (Connection timed out))
22:27:50amiconnjhMikeS: ?
22:28:47jhMikeSamiconn: Unhelpful had a problem with status updates under certain conditions that should have been addressed but nothing should seem like rand()
22:29:09 Join Chronon [0] (
22:30:00amiconnWhat I mean is that it happened several times that I shut down the beast with rockbox still showing >50% battery. But when trying to boot a few hours later, the OF bootloader complains about low bat and refuses to boot
22:30:39BigBambiI get that too
22:30:57jhMikeSwell, the curves have never been made for that target.
22:31:18BigBambiSo some battery benchmarking is in order?
22:31:51jhMikeSBigBambi: you bet. original battery prerferably.
22:32:11BigBambiOK, I'll have a look for it - I have a 1000 mAh one in at the moment :)
22:32:43jhMikeSdo you have a final reading of voltage after charging it?
22:33:07BigBambiI'll have a look, I can't remember exactly off the top of my head
22:33:53funmanthe builds are a bit long to complete, some server lag ?
22:34:21jhMikeSIt must be less than 4.242 volts (1% tolerance)
22:34:47 Quit esthar ("KVIrc 3.4.0 Virgo")
22:34:47funmankugel: are you sure that the Fuze SDRAM can be accessed correctly ?
22:35:07kugelfunman: no
22:35:17amiconnfunman: Hmm? Buikds were done a few minutes ago... and there's still red
22:35:21BigBambijhMikeS: I'll let you know - I bought it as a specific beast replacement (Cameron Sino as well), so I hope it is OK
22:35:28kugelfunman: I made a post in the v2/sansa ams thread, it sums up my findings
22:35:53kugelfunman: I'm using the same code as you for the clip and others for their ams sansas
22:35:55amiconnWhy do the other PP targets (i.e. all except the v1 Sansas) even include as3514.h? This can't be right
22:36:23funmanno I think it isn't right
22:36:49funmanthe shared i2c-pp.c needs it for some targets (c200 & e200 I think)
22:37:28amiconnYeah, but why does it include it for those where it's not necessary?
22:37:58 Join petur [50] (n=petur@rockbox/developer/petur)
22:38:34linuxstbfunman: I moved the definition of AS3514_I2C_ADDR from as3514.h into ascodec-target.h a couple of weeks ago, and added that #error...
22:39:16funmanamiconn: fs#8023 / r15297
22:39:43funmanlinuxstb: oh .. sorry to undo your work (not intentionally) but what's the point ?
22:40:18*kugel thinks funman should probably browse svn history a bit before reverting others work :)
22:40:37 Join Nibbler [0] (
22:41:15linuxstbThe point is that it's target-specific, so IMO belongs in the target files - it has to be public in ascodec-target.h for PP because it's used by static inline functions, but can just be private in ascodec-as3525.c
22:41:20funmanI just thought that things in a specific area are here because they are only used by one target, and should be moved if more than one target use it
22:41:20amiconn#ifdef HAVE_AS3514 / #include "as3514.h" / #endif
22:41:46funmanlinuxstb: well it seems it's not target-specific in fact
22:43:19linuxstbIt may differ for different targets though, and makes as3514.h nice and generic
22:43:46funmanit is still generic and doesn't differ for any target.
22:44:35linuxstbIf it has target-specific #ifs, then it's not generic.
22:45:08funmanthen just remove the #ifs (and #include "config.h") - I wondered why they were here in the first place
22:45:19 Join esthar [0] (
22:46:01funmanthis value just happens to be shared by all targets using as3514, imo it should just be generic to as3514 and not specific to each model/SoC
22:47:49 Join MichaelRoss [0] (
22:47:58 Join webguest83 [0] (n=562d6a4d@gateway/web/cgi-irc/
22:48:57 Quit Thundercloud (Read error: 104 (Connection reset by peer))
22:49:04MichaelRossmay i have write permission for the Wiki so i can upload my themes?
22:49:21MichaelRossthis is the nickname i have registered there
22:49:27BigBambisure, one mo
22:50:03*funman likes green
22:50:03MichaelRossthey are themes for the Sansa e200 if that matters
22:50:36BigBambiNot really, just promise no spamming now :)
22:50:45MichaelRossi promise
22:50:51BigBambiOK, all done
22:50:54 Join nbg [0] (i=574464b2@gateway/web/ajax/
22:50:58MichaelRossi don't even know what i would spam!
22:51:07BigBambino problem
22:51:16kugelfunman: if you have time look at my forum post please. I'm not familiar with asm, but that looks not quite correct to me
22:52:05MichaelRossi see the link redirects to the WpsGallery now
22:52:14funmankugel: I just answered - you are looking in the wrong way, if the checksum is invalid, the data read from SD card into SDRAM is incorrect, and doesn't represent valid code (which is why the checksum test was here in the first place)
22:53:03funmani'm off for a bit, see you next time :)
22:53:17 Quit funman ("leaving")
22:56:40 Quit mcflow ("Leaving.")
22:56:44MichaelRossi haven't uploaded anything yet, but it works... thanks
22:56:54MichaelRossbye for now :D
22:57:11 Quit MichaelRoss ("Java user signed off")
23:05:27 Quit nbg (" ajax IRC Client")
23:06:54BigBambijhMikeS: 4.192 V
23:06:56 Quit webguest83 ("CGI:IRC (EOF)")
23:07:13 Join mirak [0] (
23:08:36jhMikeSBigBambi: That's a good number :)
23:08:43BigBambiCool :)
23:08:51BigBambiI'll do some benches
23:09:25jhMikeSyou set the batt capacity to 1000mAh?
23:13:59 Join Zagor [242] (
23:14:23 Quit mirak (Remote closed the connection)
23:14:36amiconnlinuxstb: i2c addresses are usually hardcoded - so the as3514 should always have that address. Some chips have partially variable i2c addresses though (so you can connect more than one of them to the same i2c bus)
23:17:16 Join kushal_12_27_200 [0] (n=kushal@
23:23:03 Quit BigBambi (Read error: 104 (Connection reset by peer))
23:23:19 Join BigBambi [0] (n=Alex@rockbox/staff/BigBambi)
23:24:52 Join MethoS_ [0] (
23:25:39 Quit bertrik ("Leaving")
23:29:46 Quit aarcane ("Leaving")
23:33:00 Quit MethoS-- (Read error: 145 (Connection timed out))
23:35:13 Nick JdGordon|zzz is now known as JdGordon (n=jonno@rockbox/developer/JdGordon)
23:36:32 Quit bmbl ("Woah!")
23:37:00 Quit kushal_12_27_200 ("Leaving")
23:37:03 Quit bluebrother ("leaving")
23:39:49 Join LambdaCalculus37 [0] (
23:41:45 Quit Schmogel (Read error: 104 (Connection reset by peer))
23:43:11 Quit domonoky (Read error: 104 (Connection reset by peer))
23:43:12***Saving seen data "./dancer.seen"
23:44:19 Quit petur ("Zzzzz")
23:46:20 Quit saLOUt ("Kopete 0.12.7 :")
23:49:05 Quit scorche (Nick collision from services.)
23:50:20 Join scorche [50] (n=scorche@rockbox/administrator/scorche)
23:50:37 Quit n1s ()
23:51:49 Join tvelocity [0] (n=tony@
23:56:51 Join MethoS [0] (
23:57:14 Join Zarggg [0] (

Previous day | Next day