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

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

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

#rockbox log for 2010-12-29

00:01:08thomasjfoxNice
00:01:13thomasjfoxWell, have a good night
00:01:19selectohhyou too
00:01:30 Quit thomasjfox (Remote host closed the connection)
00:01:36 Quit kevku (Ping timeout: 272 seconds)
00:02:57 Quit benedikt93 (Quit: Bye ;))
00:06:13 Quit ender` (Quit: Trying to establish voice contact ... please yell into keyboard.)
00:18:14 Quit Keripo (Quit: Leaving.)
00:21:45 Join feisar- [0] (jljhook@irkki.fi)
00:41:26 Quit dfkt_ (Quit: -= SysReset 2.53=- Sic gorgiamus allos subjectatos nunc.)
00:42:01 Quit kadoban (Ping timeout: 276 seconds)
00:43:30 Join kadoban [0] (~kadoban@ip98-165-177-158.ph.ph.cox.net)
00:52:14 Quit kadoban (Ping timeout: 240 seconds)
00:52:29 Join kugel [0] (~kugel@46.114.13.136)
00:52:30 Quit kugel (Changing host)
00:52:30 Join kugel [0] (~kugel@rockbox/developer/kugel)
00:53:12 Quit pamaury (Remote host closed the connection)
00:54:00 Quit TheSeven|Mobile (Ping timeout: 260 seconds)
01:00
01:05:50 Join TheSeven|Mobile [0] (~theseven@rockbox/developer/TheSeven)
01:07:17 Quit merbanan (Ping timeout: 265 seconds)
01:08:50 Quit kugel (Read error: Connection reset by peer)
01:08:59 Join tempname [0] (generic@server1.unitedservers.de)
01:09:13 Join telliott [0] (~Tim@d27-96-170-180.evv.wideopenwest.com)
01:09:51 Quit TheSeven|Mobile (Read error: Connection reset by peer)
01:11:19 Join kugel [0] (~kugel@e178062044.adsl.alicedsl.de)
01:11:20 Quit kugel (Changing host)
01:11:21 Join kugel [0] (~kugel@rockbox/developer/kugel)
01:11:45telliottI need help with playlists on the clip+
01:13:03 Join TheSeven|Mobile [0] (~theseven@rockbox/developer/TheSeven)
01:13:06telliottI make playlists in MediaMonkey for music on the microsd but Rockbox needs "/<microSD1>" in the path
01:13:31 Quit TheSeven|Mobile (Client Quit)
01:13:58 Join TheSeven|Mobile [0] (~theseven@rockbox/developer/TheSeven)
01:14:12 Quit TheSeven|Mobile (Client Quit)
01:14:46 Join TheSeven|Mobile [0] (~theseven@rockbox/developer/TheSeven)
01:19:24 Join merbanan [0] (~banan@c-94-255-218-107.cust.bredband2.com)
01:20:51 Quit bertrik (Quit: :tiuQ)
01:24:21 Quit kugel (Quit: Yaaic - Yet another Android IRC client - http://www.yaaic.org)
01:24:28 Join kugel_ [0] (~kugel@e178062044.adsl.alicedsl.de)
01:25:13the_Kyletelliott: Is the music already on the microSD card? I have found that the easiest way to make a playlist for Rockbox on a microSD card is to use Rockbox itself to create the playlist.
01:25:54 Join Strife1989 [0] (~Strife89@adsl-80-154-224.mcn.bellsouth.net)
01:26:51 Quit kugel_ (Read error: Connection reset by peer)
01:27:05 Join kugel_ [0] (~kugel@e178062044.adsl.alicedsl.de)
01:27:46 Quit Strife89 (Ping timeout: 246 seconds)
01:27:50 Quit kugel_ (Read error: Connection reset by peer)
01:28:13 Quit Strife89TX (Ping timeout: 240 seconds)
01:28:33 Join kugel [0] (~kugel@e178062044.adsl.alicedsl.de)
01:28:34 Quit kugel (Changing host)
01:28:35 Join kugel [0] (~kugel@rockbox/developer/kugel)
01:32:11 Nick Strife1989 is now known as Strife89 (~Strife89@adsl-80-154-224.mcn.bellsouth.net)
01:37:09telliottthe_Kyle, yes the music is on the card. I love dynamic playlists bit I have a particular playlist I want to transfer to the card.
01:38:12the_KyleDoes the playlist not work as is if it resides on the card?
01:38:22telliottMaybe I just need an ecitor that will let me insert a string at the front of every line
01:39:06telliottNo
01:39:29telliottIt just hangs and gives up
01:40:51the_KyleIn that case, the only thing to do would be to insert the <microSD1> string at the beginning of each line. The easiest way I can think to do this is to insert it at the beginning of the first line and then copy and paste it on each subsequent line.
01:42:56telliottYeah. I tried to figure out a way to use search and replace
01:44:07the_KyleIf the files aren't in a folder, you could replace / with <microSD1>/, but if the files are in a folder, that definitely won't work.
01:44:42telliottTrue
01:45:06telliottI am going to try mocing the playlist to the root instead of /playlist
01:45:20telliottor /playlists
01:46:07the_KyleYou will still need <microSD1> at the beginning, unless you putthe playlist in the same folder as the files.
01:46:36 Join eRivas [0] (~je@190.150.0.99)
01:46:56eRivashello, is there a theming guidelines document or something?
01:48:24eRivasI want to create a theme, but have some trouble understanding .cfg files
01:50:22***Saving seen data "./dancer.seen"
01:51:03telliottThere must be some way to get MediaMonkey to do it. Plaulists on the main internal storage work fine
01:52:14the_KyleI'm not familiar with Media Monkey. It will work if you have some kind of prefix option.
01:54:02 Join saratoga [0] (9803c6dd@gateway/web/freenode/ip.152.3.198.221)
01:54:39 Quit GeekShadow (Quit: The cake is a lie !)
01:54:42saratogaselectohh: the clipv1 has 2MB of DRAM, and 320KB of IRAM, and thats all
01:55:42pixelmaeRivas: if you mean what cfg files can contain, there is a chapter in the manual about it. Or do you have another problem?
01:55:42saratogaif plugin_get_audio_buffer returned 6MB, you probably have a Clip V2, not a Clip V1
01:55:57saratogahe probably means the WPS files given the phrasing of his question
01:56:46eRivaspixelma, saratoga: I want to customize not only the wps, but also the other screens
01:56:57saratogayou can't really do that
01:57:05saratogaother then changing the fonts at least
01:57:22eRivashow about .sbs files?
01:57:34pixelmayou can create a custom status bar and the radio screen is themeable too
01:58:38eRivasso its an universal backdrop and a wps backdrop, right?
01:58:40pixelmaeRivas: I'd recommend the CustomWPS page on our wiki and the manual too
01:58:50eRivasthanks, will check now
01:59:29pixelmaand looking at other themes
02:00
02:04:35eRivaswow guys, what a great manual, cheers to everyone working on it
02:05:42pixelmanice to hear, thanks :)
02:12:55soapam I missing something? I "rbconf" and pick "21" for ipod nano, "a" for advanced options, "t" for test plugins, and it just sits there.
02:13:14pixelmaanother enter
02:13:30soapgah!
02:13:35soapthank you pixelma
02:13:57saratogai've done that before
02:14:21pixelmait's made this way to make input of more than one advanced option possible - as you even used
02:16:10pixelmanot very intuitive, I agree though
02:18:30 Join The_Pwny [0] (~IceChat7@27-33-16-78.tpgi.com.au)
02:18:47saratogayeah i just can't be bothered to add another print line after the first option
02:23:18 Join bmbl [0] (~bmbl@unaffiliated/bmbl)
02:23:29saratogaI've been testing boosting on the clip+ lately by flying around the US with it
02:23:37saratogait definitely behaves oddly
02:23:52saratogai had it boot up once in such a way that everything worked correctly except playback, just won't resume
02:24:01saratogaalso got some odd data aborts
02:24:06saratogaso something is screwy
02:24:33saratogaby just "just wouldn't resume" I mean selecting anything just resulted in silence like the audio driver wasn't feeding data to the DAC
02:31:30 Join kadoban [0] (~kadoban@ip98-165-177-158.ph.ph.cox.net)
02:34:57 Quit TeruFSX (Ping timeout: 264 seconds)
02:35:48 Quit froggyman (Remote host closed the connection)
02:42:18 Join selectahh [0] (~irc@69.17.254.184)
02:45:25 Quit selectohh (Ping timeout: 265 seconds)
02:49:01soapam I really reading LcdFrameRate wiki correct? The gigabeat S can pump > 80MB/s to the screen? Faster than my bloody LAN?
02:51:10 Join JdGord [0] (~jonno@211.31.73.76)
02:52:00soapBuschel, FS task re:Nano updated, LcdFrameRate wiki updated.
02:54:58saratogawell its only sending those bytes a couple mm, where as your lan can send them 100 meters
02:56:32soapit was a bad choice of comparisons, yes, but my LAN beats all other RB DAPs.
02:56:44saratogathe bandwidth of really short interconnects can be astronomical
03:00
03:05:01 Quit JdGord (Quit: Bye)
03:06:12 Quit bmbl (Quit: Verlassend)
03:08:43 Join mikroflops_ [0] (~yogurt@h-34-71.A238.priv.bahnhof.se)
03:12:18 Quit mikroflops (Ping timeout: 260 seconds)
03:23:35 Join madalu [0] (~user@unaffiliated/madalu)
03:33:16 Join hart [0] (~6c07e938@giant.haxx.se)
03:34:45 Join hLVEHEART [0] (~Jeremy@pool-108-7-233-56.bstnma.fios.verizon.net)
03:35:02hLVEHEARTHey guys, stupid question. What would happen if I tried installing rb on my Nano video?
03:35:04hLVEHEART:3
03:35:15hLVEHEARTWould I brick it?
03:38:16 Quit hart (Quit: CGI:IRC (Ping timeout))
03:38:18*hLVEHEART slaps himself around with Windows Me
03:41:01hLVEHEARTer, better worded... nano 3g
03:46:35[Saint]hLVEHEART: And what build do you think you will be installing on your Nano 3G?
03:46:58[Saint]Rockbox doesn't "just work", it's designed to run on each platform it runs on.
03:47:30 Quit fyrestorm (Read error: Connection reset by peer)
03:50:17 Join TeruFSX [0] (~TeruFSX@174-30-253-16.mpls.qwest.net)
03:50:23***Saving seen data "./dancer.seen"
03:50:55saratogahLVEHEART: you aren't able to try and do that, theres no way to install it on your player
03:53:39 Join Bonkers [0] (~4b014beb@giant.haxx.se)
03:53:47 Quit Bonkers (Client Quit)
03:54:02[Saint]well, you're able to try it. It won;t work though.
03:54:28 Join Bonkers [0] (~4b014beb@giant.haxx.se)
03:54:47 Join bonkers_ [0] (~4b014beb@giant.haxx.se)
03:54:47 Quit Bonkers (Client Quit)
03:55:00bonkers_Hello?
03:55:10bonkers_I need help with my iPod video 30g
03:55:21bonkers_I was redirected here by a mod
03:55:25[Saint]best to just ask your question bonkers_, no need to "ask to ask".
03:55:33bonkers_alrighty
03:56:03bonkers_0) in the top left corner
03:56:07bonkers_oops
03:56:46bonkers_I'm getting a screen that says "Data abort at 0002D444
03:57:20[Saint]does this reliably happen, or has it just happened and you want to know what to do about it?
03:57:29bonkers_just happened right now
03:57:36bonkers_and the iPod won't restart
03:57:43[Saint]press menu+select to reset the device.
03:57:47hLVEHEART:3
03:57:58bonkers_oh
03:58:02bonkers_i was pressing the bottom two
03:58:03hLVEHEARTSaint, are there any alts to rockbox available for it yet?
03:58:04bonkers_thanks
03:58:12[Saint]hLVEHEART: Nope.
03:58:19hLVEHEART=[
03:58:24hLVEHEARTthat's a shame.
03:58:27saratogaalts?
03:58:36[Saint]If there were, it wouldn't be a secret ;)
03:58:48hLVEHEART;p
03:58:55hLVEHEARTsaratoga - alternatives
03:58:57hLVEHEARTI was too lazy to type it
03:58:59saratogaoh
03:59:04saratogabad use english so i can understand you
03:59:06saratogabah
03:59:17hLVEHEART:3
04:00
04:01:41bonkers_Thank you very much
04:01:44bonkers_My ipod works now
04:02:10[Saint]No worries, just remember not to use menu+select reset excessively.
04:02:18bonkers_okay
04:02:22[Saint]It's the same as pulling the power plug on a PC really.
04:02:27bonkers_ah
04:02:38bonkers_I was having some other problems earlier
04:02:40[Saint]only use it if there's no way to shut down normally.
04:02:53bonkers_I played a song and it froze
04:02:56[Saint]what problems?
04:03:08 Quit TeruFSX (Disconnected by services)
04:03:10bonkers_i had to reconnect it several imes to fi it
04:03:16bonkers_times
04:03:24[Saint]ah...I could only speculate about the reason for that.
04:03:47 Quit telliott (Ping timeout: 246 seconds)
04:03:49saratogaif you have a song that freezes everytime you try and play it, then upload it to the bug tracker so someone can look into fixing whatever it crashes
04:04:23bonkers_Well, It was a song that I had deleted
04:04:28bonkers_but it was still listed
04:04:35bonkers_so i pressed it to see what would happen
04:04:48bonkers_it had no title, only something in <>
04:04:51bonkers_like
04:04:57bonkers_<something>
04:05:54bonkers_well im going to watch the spurs game
04:05:57bonkers_we are kicking ass :D
04:06:02bonkers_bye, thanks again
04:06:28 Quit bonkers_ (Quit: CGI:IRC (EOF))
04:07:13*[Saint] wonders if this was a "Manual Read Fail" or if "What to do when all else fails" really needs to be covered better in the manual.
04:10:52 Quit hLVEHEART (Quit: ( www.nnscript.com :: NoNameScript 4.22 :: www.esnation.com ))
04:12:11 Join pppp1 [0] (~3a60516e@giant.haxx.se)
04:13:32pppp1how do you get your ipod to not use the rockbox once its installed?
04:14:00[Saint]a: by reading the manual, b: alternatively, by holding menu or hold at boot.
04:14:29pppp1okay, thanks
04:21:32 Quit pppp1 (Quit: CGI:IRC)
04:22:41 Join kugel_ [0] (~kugel@e178059234.adsl.alicedsl.de)
04:23:03 Quit kugel (Disconnected by services)
04:24:17 Join Barahir_ [0] (~jonathan@frnk-4d0094ae.pool.mediaWays.net)
04:27:15 Quit Barahir (Ping timeout: 246 seconds)
04:36:13 Join Strife1989 [0] (~Strife89@adsl-80-181-57.mcn.bellsouth.net)
04:38:31 Quit Strife89 (Ping timeout: 260 seconds)
04:52:11 Nick Strife1989 is now known as Strife89 (~Strife89@adsl-80-181-57.mcn.bellsouth.net)
04:54:55 Quit pixelma (Disconnected by services)
04:54:57 Join pixelma_ [0] (quassel@rockbox/staff/pixelma)
04:54:59 Nick pixelma_ is now known as pixelma (quassel@rockbox/staff/pixelma)
04:56:10 Quit amiconn (Disconnected by services)
04:56:12 Join amiconn_ [0] (quassel@rockbox/developer/amiconn)
04:56:29 Nick amiconn_ is now known as amiconn (quassel@rockbox/developer/amiconn)
05:00
05:01:14 Quit Dreamxtreme (Read error: Connection reset by peer)
05:01:58 Join Dreamxtreme [0] (~Dre@92.30.156.129)
05:21:25 Quit Strife89 (Quit: Reboot.)
05:29:56 Quit madalu (Ping timeout: 250 seconds)
05:30:51 Quit Horscht (Quit: Verlassend)
05:30:56 Join Strife89 [0] (~Strife89@adsl-80-181-57.mcn.bellsouth.net)
05:34:16[Saint]Is it correct that USB is supposed to work on Fuze V1?
05:35:35[Saint]I seem to just get an endless boot cycle (I can only tell by the wheel light, it has no LCD) with RB release and SVN head, and have to mount using the OF>
05:36:17saratogayes USB works on the fuze v1
05:36:37saratogaassuming you're already in rockbox, it will not reboot
05:36:45[Saint]Do I have to do some weird magic or trickery with the bootloader?
05:37:16[Saint]I let RBUtil handle that, I haven't tried an SVN bootloader yet as I don't have a build environment set up.
05:37:22saratogaif you're in plugging in with the player off, you'll need to have patched your firmware in the last couple months
05:37:37saratogaotherwise, no, the bootloader doesn't matter once rockbox is running
05:37:41[Saint]neither seems to work.
05:37:59saratogawhat happens?
05:38:09[Saint]if I plug it in when it's on or off, boot cycle. I have to specifically boot the OF apparently.
05:38:37saratogawhat revision are you on?
05:38:47[Saint]one sec.
05:40:20[Saint].rockbox=r28913
05:40:32[Saint]can't seem to find the bootloader info.
05:41:14saratogabootloader doesn't matter
05:41:30saratogayou're sure it reboots and doesn't just freeze?
05:42:01[Saint]the wheel light comes in in a regular pattern, so yeah...I'm pretty sure it's a boot cycle.
05:42:15[Saint]without the CLD though, it makes it a little hard.
05:42:32saratogai'm not sure thats rebooting
05:42:47[Saint]well, it sertainly isn't mounting.
05:42:58[Saint]ew, that was bad. *certainly
05:43:36saratogai guess it could be freezing in someway that just reboots without showing any message
05:44:20[Saint]fingers crossed, I'll be pillaging an LCD from Strife89's old Fuze V1...so, time will tell what it's doing.
05:44:33saratogaalthough if the screen never comes on, it sounds like the OF bootloader never runs
05:44:56[Saint]well, I can't tell if the screen fires or not ;)
05:45:04[Saint]it's not actually in the DAP.
05:45:17saratogathis is one of those things thats helpful to mention
05:47:42 Join kam_ [0] (~chatzilla@user-0c90t1m.cable.mindspring.com)
05:48:26[Saint]neither the OF nor RB seems to care a bit about the missing LCD, t's easy to tell when you're in rockbox due to the voice...I simply can't get it to mount under RB though. I plug it on, either on or off and the wheel light just comes on every ~20 seconds or so.
05:49:28saratogamy guess would be you're actually running a much older build then you think you are then
05:49:35saratogaif not, i have no idea
05:50:19[Saint]well, if it's running an ancient build then RBUtil is seriously messed up and rockboxinfo lies.
05:50:27***Saving seen data "./dancer.seen"
05:50:32*[Saint] shrugs
05:54:59 Quit parafin (Ping timeout: 276 seconds)
06:00
06:02:52saratogahow many samples at a time does the DSP engine typically operate on?
06:07:49 Quit sasquatch (Ping timeout: 260 seconds)
06:12:06pikytcus_Hey, so I've got the bootloader and rb installed, and noticed when using the rockbox bootloader, if it's plugged in, the clickwheel still works
06:12:28pikytcus_It actually sends something to the system it's connected to
06:13:04 Quit eRivas (Quit: Saliendo)
06:13:17saratogayes
06:13:33saratogaalthough thats not when you're using the bootloader, but rather USB mode
06:14:06pikytcus_Yeah that
06:14:56 Join Judas_PhD [0] (~kevin@misterfluffy.dsl.xmission.com)
06:15:14[Saint]If that is undesirable, you can turn it off.
06:15:24[Saint]Or change the mode.
06:16:52pikytcus_Ah how do you do that?
06:17:17[Saint]Settings - General Settings - USB HID (iirc)
06:17:41pikytcus_Ah, wondered what that was for. Thanks
06:17:42[Saint]or, something similar.
06:17:52*pikytcus_ needs to go read up on the new system settings
06:18:30pikytcus_my gigabeatf pretty much died, and it never had that. So it's new to me anyway
06:19:58 Quit The_Pwny (Ping timeout: 276 seconds)
06:22:11 Quit kadoban (Ping timeout: 240 seconds)
06:23:11 Join sasquatch [0] (~username@2.209.19.72)
06:24:16pikytcus_Also, I've noticed if you turn the bass up, and crossfeed on, everything sounds all clippy, even with the volume really low. The gigabeat didn't do that, or at least not nearly that much. Is there something I can do to fix it?
06:25:40Lloreanpikytcus_: Bass is often handled by the hardware, so may perform differently on different players.
06:26:00LloreanOn some players though it may be handled in software (I don't know which)
06:26:10pikytcus_Ah
06:26:41LloreanYou could try using the EQ to add precut and see if that helps
06:26:56LloreanIf you've got a player with software bass, I think EQ precut or replaygain could both help
06:28:07pikytcus_Cool, will try that
06:30:13 Part Llorean
06:30:39 Join Llorean [0] (~DarkkOne@99-68-45-56.lightspeed.hstntx.sbcglobal.net)
06:30:42 Quit Llorean (Changing host)
06:30:42 Join Llorean [0] (~DarkkOne@rockbox/user/Llorean)
06:37:11 Join JdGordon| [0] (~jonno@211.31.73.76)
06:37:11 Quit JdGordon| (Changing host)
06:37:11 Join JdGordon| [0] (~jonno@rockbox/developer/JdGordon)
06:44:06 Join kadoban [0] (~kadoban@ip98-165-177-158.ph.ph.cox.net)
06:48:54 Quit kam_ (Quit: ChatZilla 0.9.86 [Firefox 3.6.13/20101206122652])
06:53:52 Quit niekie (Ping timeout: 260 seconds)
06:58:09 Join niekie [0] (~niek@CAcert/Assurer/niekie)
07:00
07:02:11 Quit Judas_PhD (Quit: This is a quitting message)
07:06:11 Join Judas_PhD [0] (~kevin@misterfluffy.dsl.xmission.com)
07:10:47 Join JesusFreak316_ [0] (~JesusFrea@pool-173-65-105-252.tampfl.fios.verizon.net)
07:16:20 Nick JesusFreak316_ is now known as JesusFreak316 (~JesusFrea@pool-173-65-105-252.tampfl.fios.verizon.net)
07:32:43 Join nowt [0] (~opera@ppp121-45-210-61.lns20.cbr1.internode.on.net)
07:34:37nowtI've found what may be a (ridiculously obscure) bug in Jewels in 3.4, I haven't found anything related to it on flyspray so I wonder whether I should report it anyway?
07:35:08nowtEssentially, if you finish a level with your last possible move, the game ends once it's completed the new level animation, even if new moves have been made possible
07:35:30nowtis this worth reporting at all?
07:36:11[Saint]is that even a bug?
07:36:27[Saint]seems as though it could be intended.
07:36:46nowtthat's what I wonder, hence "may be"
07:37:00nowthowever if your last remaining move results in more moves being possible without changing level, the game continues
07:37:19 Quit anewuser ()
07:37:24nowtmy guess is that it isn't intended, but I don't presume to know
07:37:57nowtI found it an interesting curiosity, and it's certainly very unlikely to occur
07:38:49 Quit feisar- (Ping timeout: 260 seconds)
07:39:15LloreanIt doesn't sound like expected behaviour (in the sense that the expectation is generally "if my move results in more possible moves, then I haven't lost") so I'd say report it. Though 3.4 is awfully old. Jewels may have changed since then, I'm not sure. I doubt that bug has been fixed, of course, but 3.4 is quite old.
07:39:40LloreanAt the very least, the manual should describe the game far enough to let you know what "lose conditions" exist.
07:40:01nowtYeah, I'd test it on 3.7.1 but by the time it happened again there would probably be an even more recent release!
07:41:59nowtI'm still on 3.4 because of some strange behaviour in anything more recent on my H340, but that's something I haven't looked up yet
07:44:13nowtI don't have the capability to set up the game in that state to "debug" it
07:48:00nowtAlright, submitted. Let's see where it goes
07:48:07nowtthanks for your help
07:48:24 Part nowt
07:50:30***Saving seen data "./dancer.seen"
07:57:19 Quit sinthetek (Ping timeout: 276 seconds)
07:59:26 Quit Strife89 (Quit: Sleep is but a mattress away!)
08:00
08:03:49 Quit JdGordon| (Ping timeout: 276 seconds)
08:10:59 Quit kugel_ (Ping timeout: 240 seconds)
08:13:14 Join sinthetek [0] (~sinthetek@unaffiliated/sinthetek)
08:19:50 Quit Unhelpful (Read error: Connection reset by peer)
08:20:10 Join Unhelpful [0] (~quassel@rockbox/developer/Unhelpful)
08:20:36 Quit elcan (Read error: Connection reset by peer)
08:23:59 Quit JesusFreak316 (Remote host closed the connection)
08:24:10 Part Amec
08:25:09 Quit Judas_PhD (Quit: This is a quitting message)
08:28:36 Join elcan [0] (user36@2610:150:c229::2)
08:29:35 Join Buschel [0] (~chatzilla@p54B67F52.dip.t-dialin.net)
08:35:15 Join feisar- [0] (jljhook@irkki.fi)
08:39:25Buschelsoap: thanks for testing and updating Wiki. I also think it is a good idea to let FS #11830 settle a bit and keep testing logner term.
08:39:30 Quit feisar- (Ping timeout: 240 seconds)
08:46:34Buschel@all users with PP502x target and ATA_DMA enabled: can you please test the v07 patch in FS #11830 ? I do not expect any issues, but I want to have the patch verified on as many targets as possible.
08:48:08 Quit Unhelpful (Read error: Connection reset by peer)
08:49:15 Join Unhelpful [0] (~quassel@rockbox/developer/Unhelpful)
08:56:12 Join JdGordon [0] (d31f494c@gateway/web/freenode/ip.211.31.73.76)
08:58:54 Join stoffel [0] (~quassel@p57B4DD32.dip.t-dialin.net)
08:59:17 Join ender` [0] (krneki@foo.eternallybored.org)
09:00
09:10:59 Quit JdGordon (Ping timeout: 265 seconds)
09:16:19 Join Judas_PhD [0] (~kevin@misterfluffy.dsl.xmission.com)
09:20:47 Quit elcan (Ping timeout: 272 seconds)
09:25:23 Join elcan [0] (user36@pr0.us)
09:28:16Buschelsoap: I again need to ask you for support :) I have provided a patch to speed up the RGB/YUV on the nano1G a bit. Can you take a look at FS #11843 ? Thanks in advance!
09:28:20 Quit Buschel (Quit: ChatZilla 0.9.86 [Firefox 3.6.13/20101203075014])
09:36:27 Join feisar- [0] (jljhook@irkki.fi)
09:40:13 Join LinusN [0] (~linus@rockbox/developer/LinusN)
09:41:06 Quit feisar- (Ping timeout: 255 seconds)
09:45:11 Join TheSphinX^ [0] (~cold@p54A5B450.dip.t-dialin.net)
09:45:11 Quit TheSphinX^ (Client Quit)
09:45:18 Join yawny [0] (user36@pr0.us)
09:45:36 Quit elcan (Read error: Connection reset by peer)
09:46:58 Quit TheSeven (Ping timeout: 255 seconds)
09:47:53 Quit liar (Ping timeout: 255 seconds)
09:47:53 Quit krazykit (Ping timeout: 255 seconds)
09:47:54 Quit crwl (Ping timeout: 255 seconds)
09:48:37 Join krazykit [0] (~krazykit@ppp-70-236-46-241.dsl.ipltin.ameritech.net)
09:50:32***Saving seen data "./dancer.seen"
09:51:21 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
10:00
10:00:04 Join liar [0] (~liar@clnet-p09-185.ikbnet.co.at)
10:01:18 Join crwl [0] (~crwlll@dsl-jklbrasgw1-fe8edf00-29.dhcp.inet.fi)
10:10:39 Join JdGordon [0] (d31f494c@gateway/web/freenode/ip.211.31.73.76)
10:13:27 Join chickens [0] (~3a60516e@giant.haxx.se)
10:13:50chickenshow do you use the remote control app?
10:14:00chickensfor the ipod nano 1st gen
10:17:42 Quit JdGordon (Ping timeout: 265 seconds)
10:18:35chickenscan anyone help me?
10:18:39chickenscause it looks cool
10:19:55pixelmathe plugin is not maintained anymore, but the fuctionality is in the core (unless you turned HID support off). Just connect your player with USB and use it, provided you run a recent enough build
10:20:58chickensso i cannot use the features like escape, f10 etc. anymore?
10:23:12 Quit chickens (Quit: CGI:IRC)
10:23:19 Join chickens [0] (~3a60516e@giant.haxx.se)
10:28:53 Quit chickens (Quit: CGI:IRC)
10:30:34 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
10:38:46 Join feisar- [0] (jljhook@irkki.fi)
10:50:06 Quit feisar- (Ping timeout: 240 seconds)
10:52:31 Join JdGordon [0] (d31f494c@gateway/web/freenode/ip.211.31.73.76)
11:00
11:00:38 Quit kadoban (Ping timeout: 276 seconds)
11:01:26 Join parafin [0] (parafin@paraf.in)
11:10:36 Join kugel [0] (~kugel@rockbox/developer/kugel)
11:11:59 Join feisar- [0] (jljhook@irkki.fi)
11:15:13 Quit jae (Remote host closed the connection)
11:16:40 Quit JdGordon (Ping timeout: 265 seconds)
11:17:08 Join n1s [0] (~n1s@90-230-78-242-no134.tbcn.telia.com)
11:17:08 Quit n1s (Changing host)
11:17:08 Join n1s [0] (~n1s@rockbox/developer/n1s)
11:17:26 Join DerPapst [0] (~Alexander@91-66-226-46-dynip.superkabel.de)
11:19:33 Join benedikt93 [0] (~benedikt9@unaffiliated/benedikt93)
11:35:04 Quit sinthetek (Ping timeout: 272 seconds)
11:38:43 Join sinthetek [0] (~sinthetek@cpe-174-111-239-037.triad.res.rr.com)
11:38:43 Quit sinthetek (Changing host)
11:38:43 Join sinthetek [0] (~sinthetek@unaffiliated/sinthetek)
11:39:19 Join fml [0] (~chatzilla@p5DF3E855.dip.t-dialin.net)
11:41:04fmlNote to gevaerts, kugel and whoever: if HAVE_PLUGIN_CHECK_OPEN_CLOSE is not defined in plugin.c, sim build, then calling rb->creat in a plugin does not lead to calling sim_creat. If it's defined then sim_creat is called.
11:41:38fmlBut what does get called? The native creat function? But then it should fail since the sim root dir is not inserted.
11:41:51fmlI'll file a bug report.
11:44:45 Quit DerPapst (Quit: Leaving.)
11:50:35***Saving seen data "./dancer.seen"
11:57:42 Quit benedikt93 (Quit: Bye ;))
12:00
12:02:06 Join benedikt93 [0] (~benedikt9@unaffiliated/benedikt93)
12:12:57 Quit TheSeven|Mobile (Read error: Connection reset by peer)
12:20:39 Join Kitar|st [0] (~Kitarist@89.142.76.239)
12:32:36 Quit Topy44 (Read error: Connection reset by peer)
12:35:08 Join Topy44 [0] (~Topy44@89.207.248.177)
12:37:04 Quit user890104 ()
12:41:32 Quit sasquatch (Ping timeout: 276 seconds)
12:45:05 Join sasquatch [0] (~username@2.209.147.85)
12:48:42 Join bmbl [0] (~bmbl@dsl-217-162-40.pool.bitel.net)
12:48:42 Quit bmbl (Changing host)
12:48:42 Join bmbl [0] (~bmbl@unaffiliated/bmbl)
12:50:49 Quit fml (Quit: ChatZilla 0.9.86 [Firefox 3.6.13/20101206122310])
12:54:31 Join {phoenix} [0] (~dirk@p57AA7620.dip.t-dialin.net)
12:56:58selectahhthanks for your response saratoga (re: 10 hours ago)
12:59:52 Quit {phoenix} (Read error: Connection reset by peer)
13:00
13:22:54 Quit crwl (Ping timeout: 264 seconds)
13:27:57 Join crwl [0] (~crwlll@dsl-jklbrasgw1-fe8edf00-29.dhcp.inet.fi)
13:50:19 Part LinusN
13:50:37***Saving seen data "./dancer.seen"
13:50:51 Join LinusN [0] (~linus@rockbox/developer/LinusN)
14:00
14:01:57 Join mirak [0] (~mirak@85-170-106-220.rev.numericable.fr)
14:05:56 Part LinusN
14:20:10 Quit stoffel (Ping timeout: 246 seconds)
14:29:25 Quit Topy44 (Ping timeout: 240 seconds)
14:45:46 Join dfkt [0] (dfkt@unaffiliated/dfkt)
14:49:01 Join user890104 [0] (Venci@venci-notebook-lan.ipv6.6bez10.info)
14:51:17TheSevenlinuxstb: ping
14:53:33CIA-7New commit by jethead71 (r28923): Just a technical matter even though all is well anyway: struct blocker::thread should be volatile for > 1 core and the change should make no ...
14:55:34 Quit [Saint] (Disconnected by services)
14:55:36 Join S_a_i_n_t [0] (S_a_i_n_t@203.184.0.89)
14:56:05CIA-7r28923 build result: All green
14:58:35 Join LinusN [0] (~linus@rockbox/developer/LinusN)
15:00
15:00:41LloreanSpeex has a decoder feature called "perceptual enhancement." Does anybody know about this and whether or not we use it?
15:03:02S_a_i_n_ttheir narrowband mode?
15:03:25S_a_i_n_tI've looked into it before for something, no idea if RB uses it or no.
15:04:21LloreanThe manual doesn't suggest it has anything to do with your choice of narrowband or not.
15:04:45LloreanJust says it's a feature of the decoder (which arguably means it shouldn't matter if you've chosen n/w/uw)
15:09:27 Join komputes [0] (~komputes@ubuntu/member/komputes)
15:10:02CIA-7New commit by jethead71 (r28924): interrupt_level should've been volatile as well when I changed this file. Obtuseness FTW.
15:12:01CIA-7r28924 build result: All green
15:13:38 Join Buschel [0] (~chatzilla@p54A3A07A.dip.t-dialin.net)
15:16:45 Quit user890104 (Ping timeout: 272 seconds)
15:17:37 Quit yawny (Read error: Connection reset by peer)
15:17:43 Join elcan [0] (user36@pr0.us)
15:19:20 Join freddyb [0] (~chatzilla@216.8.239.112.etczone.com)
15:27:27 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
15:36:42soapBuschel, regarding the Nano LCD patches, is there a reason to test v01 (w/o YUV) if v02 works ok?
15:36:56Buschelno
15:37:11Buscheldoes this mean it works ok? ;-)
15:37:43soapabout to unmount and see
15:46:10soap31% speedup in YUV, no change in LCD
15:47:47Buschelnice :)
15:48:12BuschelRGB might have reached the maximum possible frame rate
15:48:36S_a_i_n_tGotta meet the Nano2G LCD speed-up, or I won't recognize the achievement :P
15:48:45Buschelif it the serial LCD IF is clocked at 80MHz (PLL) 80/8 = 10 MB/s is the max
15:49:54soapmovies appear to be playing fine. Is there reason to suspect more covert issues with the YUV changes?
15:50:31Buschelsoap: could you try the following and retest RGB full screen? -> r+=4 in line 338 and copying line 342 to call this twice?
15:50:38***Saving seen data "./dancer.seen"
15:50:58Buschelsoap: no. if movies play fine and the test screen loks fine there will be no futher issue
15:51:06soapof lcd-color_nano.c?
15:51:10Buschelyes
15:51:19*Buschel forgot to mention the file to change...
15:51:37 Join bertrik [0] (~bertrik@ip117-49-211-87.adsl2.static.versatel.nl)
15:51:38 Quit bertrik (Changing host)
15:51:38 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
15:52:35soapso r+=2 changed to r+=4
15:52:40Buschelyes
15:53:21soapand just LCD2_BLOCK_DATA = *addr++;
15:53:25soapliterally twice in a row?
15:53:28Buschelyes
15:54:15Buschelyou'll only need to test 80 MHz as this is more critical
15:56:09soapgarbled boot logo
15:58:28soapgoing to the gym, BBIAB
15:58:46Buschelok, thanks so far :)
16:00
16:00:02 Quit S_a_i_n_t (Ping timeout: 250 seconds)
16:00:21 Quit mirak (Quit: Ex-Chat)
16:00:30 Join kevku [0] (~kevku@2001:7d0:0:f000::135d)
16:03:07 Join eWill [0] (~chatzilla@adsl-76-235-39-45.dsl.dytnoh.sbcglobal.net)
16:03:14eWillWhere in the source code is the setting for how many entries are kept in the "Most Recent Bookmarks" list?
16:05:21 Join [Saint] [0] (S_a_i_n_t@203.184.0.89)
16:07:24freddybShould I revert FS #11765 (r28834)? I'm stumped on why it broke uSD for some.
16:09:20Buschelis a use with an iPod color present?
16:09:25Buscheluser
16:09:42[Saint]indeedy, though I can't build presently.
16:10:00[Saint]roll me a build, and I can test it for you.
16:10:05 Part LinusN
16:10:12Buschelgive me some minutes
16:10:42[Saint]No problem. I'm doing an all-nighter for work...uurgh :/
16:11:48Buschelgreat, then you can do multiple tests with a single build (1. ATA DMA change, 2. RGB, 3. YUV)
16:12:18[Saint]Sure.
16:13:40bertrikfreddyb, yes I think it should be reverted to get uSD working for everyone. My guess it's probably some issue in the sd controller driver though. There's probably something very marginal there, we've seen problems earlier when slightly increasing the CPU clock from 240 MHz to 248 MHz.
16:14:13Buschel[Saint]: build started, will take max. 10min
16:14:28*Buschel has an old CPU
16:14:52*[Saint] wonders what it is...
16:15:33[Saint]I can beat 10min with an AMD 2100+ on CygWin, so I'm expecting *old*
16:15:56 Join anewuser [0] (anewuser@unaffiliated/anewuser)
16:17:39Buschelsome 1.7 GHz Pentium... do not know the precise naming
16:17:46Buschelon cygwin
16:20:26Buschel[Saint]: can you do a short test_fps test run on your color? to have a reference
16:20:37[Saint]yep.
16:21:01[Saint]Oh, ummm...no. No test_fps *facepalm*.
16:21:12Buschelouch
16:21:24 Quit krazykit (Quit: awe yeeeeeee)
16:21:31Buscheldo you use a current build?
16:22:02[Saint]Yes, well...I can install one if the rev is off.
16:22:33[Saint]Not having a build environment set up is rather inconvenient.
16:24:09Buschellet's just await the test of the build i am preparing
16:24:13 Quit AlexP_ (Remote host closed the connection)
16:24:53 Join AlexP [0] (~alex@rockbox/staff/AlexP)
16:31:25[Saint]Just installed 28924 so any test_* plugins should be transferrable also.
16:31:58Buschelbuild is uploaded in a minute
16:34:31Buschelhttp://www.sendspace.com/file/b6vicb
16:34:42[Saint]thanks.
16:53:44Buscheldoes it start and buffer ok?
16:59:59 Quit sasquatch (Ping timeout: 260 seconds)
17:00
17:03:33 Quit eWill (Quit: ChatZilla 0.9.86 [Firefox 3.6.13/20101203075014])
17:07:19CIA-7New commit by fredwbauer (r28925): revert r28834 because it causes problems with uSD
17:08:03 Join GigabeatUser [0] (~c869dcf4@giant.haxx.se)
17:08:17[Saint]LCD 1/1: 19.4, 1/4: 76.0 ; LCD YUV 1/1: 19.1, 1/4:75.5 @ 30MHZ
17:08:55GigabeatUserhi there
17:09:30CIA-7r28925 build result: All green
17:09:59[Saint]LCD 1/1: 51.7, 1/4: 202.5 ; LCD YUV 1/1: 51.0, 1/4: 202.0 @ 80MHZ
17:10:08[Saint]Buschel: ^
17:10:25[Saint]seems to buffer ok, boots fine. anything else?
17:10:51 Quit TheSeven (Ping timeout: 240 seconds)
17:11:25GigabeatUserthere is someone here that can give me a hand with rockbox?
17:11:48[Saint]GigabeatUser: There's many people, all you need to do is ask ;)
17:11:56[Saint]We're not mind readers ;p
17:11:57Buschel[Saint]: is test_fps ok as well?
17:12:18Buscheland what numbers do you reach for 80 MHz and 30 MHz?
17:12:19[Saint]Buschel: yes, there's some results above.
17:12:28Buschelshit, overseen those
17:12:33GigabeatUserthanks, I just didnt knew if one of all the people around here was something like an admin
17:12:57GigabeatUserI have an old toshiba gigabeat F series 20GB
17:13:04GigabeatUserhad rockbox before
17:13:21GigabeatUserthen I lost it, and when i found it again, I formated the whole player
17:13:32GigabeatUsernow I want to install rockbox again
17:13:55GigabeatUserautomatically)
17:14:05GigabeatUserwith the program, it gives me the following error
17:14:08Buschel[Saint]: any glitches when playing with YUV with boosted CPU (e.g. playing a movie)?
17:14:08 Join sasquatch [0] (~username@92.116.109.239)
17:14:43[Saint]Buschel: One second, I'll just transfer one onto the disc.
17:14:45 Quit Judas_PhD (Quit: This is a quitting message)
17:14:54 Join user890104 [0] (~Venci@6bez10.info)
17:15:04GigabeatUsercache used)- Installing rockbox bootloader- *error accessing output folder*
17:15:36[Saint]GigabeatUser: What OS is this? Do you have write permissions?
17:15:57GigabeatUserits windows 7 and I'm admin
17:15:59 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
17:16:09GigabeatUserI runned the program with admin permisions
17:16:16gevaertsGigabeatUser: do you have ☕ GBSYSTEM directory on the player? I assume not, if you just formatted it...
17:16:26gevaertsHave a look at http://www.rockbox.org/wiki/GigabeatFXPort#Improved_Boot_Time_40Optional_41
17:16:30GigabeatUseryes, I formated, i
17:16:32 Quit freddyb (Ping timeout: 255 seconds)
17:17:07GigabeatUserwill check it now
17:17:17gevaertsDownload the zip file from there and unpack it on your player. I think that will do the trick
17:17:57gevaertsYou'll only have rockbox though. If you want the original firmware too, you'll have to restore that somehow (I can't help there)
17:18:38GigabeatUsernah, why would I want the OF if I have rockbox
17:18:55GigabeatUserOF only plays songs and views pictures, rockbox is games, videos, its amasing man
17:19:00GigabeatUserbi <3 rockbox
17:19:04GigabeatUserbig*
17:19:18GigabeatUserthe other file *.DAT will have to put that too?
17:19:32GigabeatUseror its ok with the unzip on the / of the player
17:19:37[Saint]Buschel: Seems fine.
17:19:37 Quit AndyI (Ping timeout: 272 seconds)
17:21:08gevaertsThe other .DAT file is the rockbox bootloader. You can put it in GBSYSTEM too. If you do that, you don't have to install a bootloader using rockbox utility any moire
17:23:22 Join Strife89TX [0] (~cstrife89@207.144.201.128)
17:24:40 Quit GigabeatUser (Quit: CGI:IRC (Ping timeout))
17:25:29 Join Keripo [0] (~Keripo@CPE0022b0d4bdb7-CM001a6680d4fe.cpe.net.cable.rogers.com)
17:26:07 Join AndyI [0] (~pasha_int@212.14.211.251)
17:29:25 Quit kugel (Remote host closed the connection)
17:32:51Buschel[Saint]: can you retest YUV/RGB again with this version? -> http://www.sendspace.com/file/h3wcq8
17:33:10Buschel[Saint]: this is the safe variant which polls for LCD IF ready
17:34:05[Saint]yep, gimme a while. I'll post when I have the results.
17:34:11Buschelperfect
17:35:20Buschelwill be back later
17:37:33 Join Buschel_ [0] (~chatzilla@p54A3A07A.dip.t-dialin.net)
17:40:19 Quit Buschel (Ping timeout: 264 seconds)
17:40:30 Nick Buschel_ is now known as Buschel (~chatzilla@p54A3A07A.dip.t-dialin.net)
17:43:44soapBuschel, I'm back if there is anything else I can test for you.
17:44:29Buschelsoap: could you also test v03 in FS #11843 ? speed as well as if there are any flaws in YUV blitting or movie playing?
17:44:37soapcan do
17:44:43Buschel(in boosted state)
17:45:44soapI _ass_ume pluginmpegplayer is boosting? My prior report regarding v02 and no visible glitching was playing Elephant's Dream w/o manually setting the boost count high.
17:46:21[Saint]iiuc, pluginmpegplayer is always boosting
17:46:46Buschelyes, it is boosting on its own. the v03 patch does not wait for FIFO ready. so, it might be showing flaws (which i do not expect as the conversion should take longer than a needed wait)
17:47:07 Quit Gareth (Read error: Operation timed out)
17:47:11 Join Gareth [0] (~gareth@www.wiked.org)
17:47:35 Join WonTu [0] (~WonTu@p57B56CC9.dip.t-dialin.net)
17:47:50 Part WonTu
17:49:03*Buschel has a real life interrupt now
17:50:41***Saving seen data "./dancer.seen"
17:51:47soapBuschel, /home/soap/rockbox/firmware/target/arm/ipod/lcd-as-color-nano.S:201: Error: symbol `lcd_yuv_write_inner_loop' is already defined
17:51:47soapmake: *** [/home/soap/rockbox/buildnano/firmware/target/arm/ipod/lcd-as-color-nano.o] Error 1
17:53:39 Quit Buschel (Ping timeout: 260 seconds)
18:00
18:03:30 Join Buschel [0] (~chatzilla@p54A3A07A.dip.t-dialin.net)
18:14:05 Quit AndyI (Ping timeout: 272 seconds)
18:19:07 Join kugel [0] (~kugel@df01ppp216.eplus-online.de)
18:19:08 Quit kugel (Changing host)
18:19:08 Join kugel [0] (~kugel@rockbox/developer/kugel)
18:19:09 Join AndyI [0] (~pasha_int@212.14.211.84)
18:29:31 Join froggyman [0] (~seth@unaffiliated/froggyman)
18:37:17 Quit feisar- (Remote host closed the connection)
18:39:25 Join LambdaCalculus37 [0] (~3f74f70d@rockbox/staff/LambdaCalculus37)
18:45:53 Quit parafin (Quit: So long and thanks for all the fish)
18:45:57 Join GigabeatUser [0] (~cdf1265f@giant.haxx.se)
18:46:14 Nick benedikt93 is now known as benedikt93|AFK (~benedikt9@unaffiliated/benedikt93)
18:47:32 Join eRivas [0] (~je@190.150.0.99)
18:49:39eRivashi guys, some theming questions here
18:50:01[Saint]shoot.
18:50:02eRivasmy volume icon is not working correctly
18:50:14eRivasit has only four stages
18:50:26eRivasmute, min, med, and max, to say something
18:50:27[Saint]define "not working correctly", and pastebin your code.
18:51:41eRivasI would expect rockbox to show each one for equal intervals
18:52:20eRivasbut rather one of the icons is really long, and the last three almost change on every step I do on volume
18:52:27[Saint]wait...are you talking about your own theme, or a theme shipped with rockbox?
18:52:42eRivasoops, sorry, my own theme, on sansa fuze v2
18:52:52[Saint]pastebin your code please.
18:52:59[Saint]I think I know what you're doing wrong.
18:53:00eRivasall the wps?
18:53:09[Saint]why not.
18:53:11eRivasok
18:55:15[Saint]iirc, the volume is actually <mute|.|.|...|line out|greater than line out>
18:55:25 Quit GigabeatUser (Quit: CGI:IRC (Ping timeout))
18:55:37[Saint]it soean't magically adjust itself like some other things which are similar do.
18:55:39eRivasfyi, right now its just a mod of the darkness theme, so it has commentarys and file name for it, will change later
18:55:46[Saint]*doesn't
18:55:48eRivashttp://pastebin.com/FsZNj3Uq
18:58:31[Saint]right, yeah...it is as I thought.
18:58:49[Saint]the %pv tag segments are actually "the volume is actually <mute|.|.|...|line out|greater than line out>"
18:59:29[Saint]so, you just need to adjust the amount of segments to fit your percieved idea of "15%" increments.
18:59:45[Saint]*25% rather
19:00
19:00:25eRivasI'm sorry, could you provide the code, I did this thing for the first time last night
19:01:52 Join Horscht [0] (~Horschti@p5DD57F6D.dip.t-dialin.net)
19:01:53 Quit Horscht (Changing host)
19:01:53 Join Horscht [0] (~Horschti@xbmc/user/horscht)
19:02:53[Saint]http://pastebin.com/RfGgu0nt should look a bit better, assuming that a == mute and d == max in the sub-images.
19:03:06eRivasa is the top?
19:03:39[Saint]My code hasn't touched positioning...so, if it was before, then yes.
19:03:48eRivasok
19:04:16eRivasthanks, btw, I've also seen the battery icon dissapear when the battery is very low
19:07:27[Saint]eRivas: http://pastebin.com/rMmYXBQ9 fixes the battery issue
19:07:30[Saint]or, should.
19:08:01*LambdaCalculus37 is trying to figure out the last signed/unsigned warning he's getting from FS #7505
19:08:26Buschelsoap: do you still have the build issue?
19:08:51 Nick benedikt93|AFK is now known as benedikt93 (~benedikt9@unaffiliated/benedikt93)
19:11:26LambdaCalculus37This is a bit more tricky for me...
19:11:28Buschelsoap: I do not understand your build problem, the v03 patch is nearly identical to the v02 −− except an additional "#if'defing".
19:16:54eRivas[Saint]: hmm, the theme broke, it fell back to the stock theme
19:17:32[Saint]eRivas: One second. I'll have another look.
19:18:21eRivas[Saint]: I have notice that the old battery code has "||%xd(ia)". Are those two bars necessary
19:19:17[Saint]no, it skips out he condition for battery state unknown, and the first step of the battery state that is known.
19:19:21[Saint]*the
19:20:34CIA-7New commit by jethead71 (r28926): Gigabeat S: Do simple direct keypad scanning rather than triggering a separate scan interrupt and it all does the same thing but with 1 tick less ...
19:21:14[Saint]eRivas: http://pastebin.com/RfNAB2FR
19:21:24[Saint]The code parses as free from error here.
19:22:06[Saint]I reduced the size rather considerably, also.
19:22:32CIA-7r28926 build result: All green
19:23:20[Saint]...which may or may not bring up some alignment issues if newlines were used for alignment (I *hate* that...) now I think of it, but I'll cross that bridge if it happens.
19:23:57eRivasok, will try
19:24:19 Join parafin [0] (parafin@paraf.in)
19:24:25eRivason another topic, how do theme makers take screenshots?
19:24:48LambdaCalculus37eRivas: The UI sim is a good way of doing so.
19:25:46[Saint]you can also enable screendump on the player through the debug screen.
19:26:12[Saint]when in screendump mode, as dump of the current screen state is made when USB is plugged.
19:26:54[Saint]I'm unsure if the manual describes screendump or not, as it's in the debug menu.
19:27:24eRivaswill read about it
19:30:01 Quit LambdaCalculus37 (Quit: back to constructing those pylons)
19:31:53eRivas[Saint]: your code works fine
19:32:03[Saint]good to know.
19:33:18 Quit froggyman (Quit: Ex-Chat)
19:50:38 Quit liar (Ping timeout: 240 seconds)
19:50:43***Saving seen data "./dancer.seen"
19:52:10 Join liar [0] (~liar@clnet-p09-185.ikbnet.co.at)
19:55:18pikytcus_Hey, I think rockbox has wrongly detected the features of my IPod. It says I have line-in and line-out, and no fm radio. But according to the manual for the IPod video, it has an fm radio. Not sure about line-in and line-out
19:55:37 Join GuySoft [0] (guy@bzq-79-183-3-236.red.bezeqint.net)
19:56:01GuySofthi all, might anyone know where i can get the sansa fuze firmware so i can install rockbox?
19:56:15eRivasGuySoft: the forum thread at sansa
19:56:46GuySofteRivas, thanks :
19:56:48GuySoft:)
19:57:27pikytcus_I don't see any of those on the player though. All I can find on the player (besides the clickwheel) are a headphone jack and a place for the sync cable
19:58:09pikytcus_What am I missing?
19:58:28saratogapikytcus_: the ipod video doesn't have an FM radio, although you can add one using an external ipod FM accessory
19:59:23pikytcus_saratoga: Ah k. What about the line-in and out though? I don't see those.
19:59:30saratogayes it has those
19:59:38pikytcus_Oh hmm
19:59:47saratogain the future, if you have questions about what your device has, the apple website, wikipedia or even google are a great way to find out
19:59:50 Join krazykit [0] (~krazykit@99-126-205-52.lightspeed.cicril.sbcglobal.net)
19:59:57*pikytcus_ nods
20:00
20:00:38pikytcus_So, if this is not an IPod video as claimed, wonder what it is?
20:02:56[Saint]can you post a picture?
20:03:10benedikt93pikytcus_, http://support.apple.com/kb/HT1353 may help you to find out
20:05:30 Join JesusFreak316 [0] (~JesusFrea@pool-173-65-105-252.tampfl.fios.verizon.net)
20:06:14 Join fml [0] (~chatzilla@manz-590ee82f.pool.mediaWays.net)
20:06:47pikytcus_It seems to be a fifth generation IPod
20:07:00fmlgevaerts: hello. Could you please help me to sort out the problem described in FS #11844 ?
20:08:15pikytcus_if I could look at them I could tell. Grr
20:08:17fmlI think it's not disktidy specific, but I noticed it there and hence written so in the bug report. jhMikes has reported to have similar problems in mpeg player
20:09:35pikytcus_or, it could be something else... Several of them seem to have a click wheel and 60gb
20:10:24[Saint]pikytcus_: There will be a serial number on the back, printed with the capacity (assuming it is an actual iPod), google it.
20:10:45[Saint]this will give you a hit about the device type for surre.
20:10:51[Saint]*sure
20:11:00 Join kadoban [0] (~kadoban@ip98-165-177-158.ph.ph.cox.net)
20:12:21fmlkugel: the same request as for gevaerts (FS #11844)
20:18:25 Quit kugel (Ping timeout: 255 seconds)
20:18:58 Quit benedikt93 (Quit: Bye ;))
20:23:38eRivaswhich files are necessary to create radio screen themes?
20:23:48[Saint].fms
20:23:54eRivasok
20:24:29[Saint]an .sbs may also be present in a wps/fms also
20:25:22 Quit B4gder (Ping timeout: 260 seconds)
20:25:37 Join Keripo1 [0] (~Keripo@CPE0022b0d4bdb7-CM001a6680d4fe.cpe.net.cable.rogers.com)
20:26:36pikytcus_Well blah. That didn't help much. GOt the serial number, but Apple's website didn't help
20:27:01pikytcus_The last three characters don't match any of the ones on their site
20:27:19[Saint]just punch it into google.
20:28:06pikytcus_No results for it
20:28:23saratogaeither you're wrong or Apple is wrong then . . .
20:28:28saratogahttp://samples.mplayerhq.hu/A-codecs/
20:28:39saratogai wonder what happens if i try to build a database in rockbox with all those samples ...
20:29:13 Quit Keripo (Ping timeout: 255 seconds)
20:29:25gevaertspikytcus_: if rockbox utility says it's an ipod video, it *is* an ipod video
20:30:19pikytcus_It's all their fault
20:30:31saratogawait you already have rockbox on it?
20:30:44pikytcus_Ok. Apparently, it is an older fifth generation IPod from 2005
20:31:15saratogahow do i get wget to grab that link without going one level below and grabbing the video codecs too?
20:31:45gevaertssaratoga: -np
20:31:53gevaertsoh, wait
20:31:58pikytcus_Yeah I do. Didn't use the utility though
20:32:10saratogawell just look at which rockbox build you're running
20:32:14GuySofteRivas, i cant find it in the thread, its all about sansa fuze V2. i have V1
20:32:29GuySoftwhere can i find the sansa fuze firmware so i can install rockbox?
20:32:30gevaertssaratoga: the -l option (depth)
20:32:39eRivasGuySoft: they both are on the same thread, give me a min
20:32:51saratogaso -l 1?
20:32:53gevaertsGuySoft: rockbox utility should show a working link
20:33:05eRivasGuySoft: http://forums.sandisk.com/t5/Fuze/Sansa-Fuze-Firmware-Update-01-02-31-amp-02-03-33/td-p/139175
20:33:06saratogai still want it to recurse, just only recurse down the tree, and not up to the root
20:33:10gevaertssaratoga: I think so. I tend to think off-by-one though :)
20:33:26gevaertssaratoga: -np is no parent (so only down), and -l says how deep
20:33:36eRivasGuySoft: direct link: http://mp3support.sandisk.com/firmware/fuze/fuze01.02.31.zip
20:33:50saratogaah great
20:34:04 Join B4gder [0] (~daniel@rockbox/developer/bagder)
20:34:18pikytcus_idk, I built it from source, and don't configure it was an ipodvideo (target 22 in the list)
20:34:33pikytcus_urg, typing failure. Ah well
20:34:36saratogadoes the build run on your ipod?
20:34:52GuySofteRivas, the version i have on it now is 01.01.11A , is it ok to use another?
20:35:03gevaertspikytcus_: then it's an ipod video...
20:35:06saratogayes it doesn't matter, just give it a firmware file
20:35:53*pikytcus_ nods
20:36:22gevaertsfml: I got "Could not create file /.rockbox/rocks/apps/disktidy_custom.config" on the console output. I assume that's also what you get?
20:36:54pikytcus_are the line-in and line-out separate from the actual device, then? Guessing an accessory you plugin to the dock port?
20:37:13gevaertspikytcus_: they're probably some pins on the dock connector
20:37:26gevaertsLine out definitely is, not sure about line in
20:37:44pikytcus_Ah k. Thanks
20:38:06 Quit Strife89TX (Quit: Heading out.)
20:38:23saratogacrap 1.5GB downloaded so far
20:38:24saratogathey're going to fill my sansa
20:39:14eRivaswhere do I have to put playlists so that they appear in the catalog?
20:39:33GuySofteRivas, thanks for the help, installing now
20:39:40eRivasGuySoft: np
20:40:06CIA-7New commit by gevaerts (r28927): Properly prefix the creat function for sim builds, otherwise the system creat() is called, which doesn't do proper root-mangling to make the file end ...
20:42:03CIA-7r28927 build result: All green
20:45:47eRivascould someone give and example of a working path from /playlist to the microSD card?
20:49:01fmlgevaerts: yes, that was the message I got. And thanks for the fix! Shouldn't "remove" and "rename" also be prefixed?
20:49:07 Quit mortalscan (Remote host closed the connection)
20:49:22 Join mortalscan [0] (~mortalsca@109.169.55.155)
20:49:41*gevaerts checks
20:50:29gevaertshm, probably
20:52:18gevaertsor maybe not
20:52:29fmlgevaerts: ...as well as mkdir and rmdir
20:53:14gevaertsThose do work
20:53:47gevaertshm, no
20:53:58fmlgevaerts: I don't understand the principle. What funcs should be prefixed and which not? The all exist as sim_xxx.
20:54:03*gevaerts didn't test from a plugin, so he doesn't actually know...
20:54:09fmls/The/They
20:54:40gevaertsI guess a lot of them need sim_...
20:55:04fmlgevaerts: that's what PREFIX does IIUC
20:55:11gevaertsyes, exactly
20:57:34 Quit Buschel (Ping timeout: 255 seconds)
21:00
21:03:39 Quit parafin (Quit: So long and thanks for all the fish)
21:03:50CIA-7New commit by jethead71 (r28928): Return thread ID as a string for name when a thread isn't named or name cannot be used, as comment says, not slot pointer.
21:04:50 Join thomasjfox [0] (~thomasjfo@dslb-088-066-094-042.pools.arcor-ip.net)
21:05:55CIA-7r28928 build result: All green
21:07:57GuySoftis there a way to get the sansa fuze to multiboot?
21:08:42gevaertsYes. The manual should tell you how
21:09:05fmlgevaerts: if those functions need PREFIX then I don't understand how and why "open" works, because it's included with the wrapper, and in the wrapper, PREFIX is not used.
21:09:46GuySoftgevaerts, thanks ill ook
21:09:51GuySoftlook*
21:10:13CIA-7New commit by gevaerts (r28929): Also PREFIX() other filesystem functions. I hope I got them all now...
21:10:20gevaertsfml: I'm not sure what you mean
21:11:03 Quit thomasjfox (Read error: Connection reset by peer)
21:11:32 Join thomasjfox [0] (~thomasjfo@dslb-088-066-094-042.pools.arcor-ip.net)
21:12:24CIA-7r28929 build result: All green
21:12:43GuySofthmm, is there a way to change the boot priority
21:12:45GuySoft?
21:13:05 Join stoffel [0] (~quassel@p57B4CB51.dip.t-dialin.net)
21:13:39 Join parafin [0] (parafin@paraf.in)
21:20:10saratogaGuySoft: no
21:21:06 Quit komputes (Remote host closed the connection)
21:24:22 Join CaptainKewl [0] (~captainke@207-38-215-126.c3-0.nyr-ubr1.nyr.ny.cable.rcn.com)
21:28:05GuySofthmm, what about hebrew fonts? its listed here in feature requests: http://forums.rockbox.org/index.php?topic=23161.0
21:28:49[Saint]rockbox has great support for Hebrew
21:29:48gevaertsThat post just suggests to have *more* fonts with Hebrew support
21:30:11saratogareal audio lossless files seem to crash the database parser
21:31:22[Saint]GuySoft: rasher.dk/rockbox/fontstats/">http://rasher.dk/rockbox/fontstats/
21:31:51[Saint]here, you can check the availability of support for a particular lang vs. a particular font
21:32:38GuySoft[Saint], yes just figured it now
21:34:24GuySoftyay done
21:34:56saratogathat metadata log is really handy
21:37:10saratogai guess the next step is to write a script that randomly exchanges the file extensions and see what happens!
21:43:10fmlgevaerts: I mean that the open function is included into the plugin API via its wrapper, open_wrapper. In that wrapper function, no PREFIX is used. There is a comment there explaining things but I don't understand it.
21:43:43 Join komputes [0] (~komputes@ubuntu/member/komputes)
21:44:34gevaertsfml: "open()" is already a macro that expands to open_uncached on target. I assume it expands to something on the sim too
21:45:37fmlgevaerts: ah... and where is that "open" macro defined?
21:45:47gevaertsfile.h I guess
21:46:58 Quit jhMikeS (Read error: Connection reset by peer)
21:47:00fmlgevaerts: yes. But there, all other file funcs are defined as macros too. :-/
21:47:03 Join jhMikeS [0] (~jethead71@rockbox/developer/jhMikeS)
21:48:16 Join froggyman [0] (~seth@unaffiliated/froggyman)
21:48:22fmlgevaerts: what confuses me is the fact that the file I/O functions are treated differently in the plugin API.
21:49:05gevaertsAre they?
21:49:05 Quit jhMikeS (Read error: Connection reset by peer)
21:49:11 Join jhMikeS [0] (~jethead71@rockbox/developer/jhMikeS)
21:50:45***Saving seen data "./dancer.seen"
21:50:59thomasjfoxhey jhMikeS
21:51:14thomasjfoxI've seen you are currently working on making rockbox "thread safe"
21:52:37 Quit jhMikeS (Read error: Connection reset by peer)
21:52:58fmlgevaerts: yes, a little bit. In file.h, sim_open and sim_creat are declared but not the other functions, which are only defined as macros.
21:53:56fmlgevaerts: I mean, declared as "extern int sim_open ..."
21:54:06fmlWhy is it needed?
21:54:11gevaertsno idea
21:54:16*gevaerts didn't write most of this!
21:55:24thomasjfoxgevaerts: hey, I've seen kugel created a n900-demo branch in his repository. How did that work out?
21:55:52 Quit stoffel (Ping timeout: 276 seconds)
21:56:20gevaertsthomasjfox: that's a merge of your tree, his ucontext branch and his cabbie branch. It runs :)
21:56:37gevaertsThe usual "pulseaudio uses lots of CPU" things, but we can't help that
21:57:29gevaertsHalf of that disappears if you plug in headphones, and another 2.5% or so vanishes if you tell gstreamer that rockbox outputs at 48kHz (which of course doesn't sound entirely right)
21:57:38thomasjfoxTrue that. So it didn't crash on you yet :)
21:57:56gevaertsThe built-in media player has the same pulseaudio behaviour
21:57:56thomasjfoxhuh, the CPU load goes down if you plug in headphones?
21:58:20gevaertsIt crashed once yesterday. I can't remember who was playing with it at the time though
21:58:40gevaertsYes. Apparently they need to do some heavy eq to make the speakers sound right at higher volume
21:58:42 Join jhMikeS [0] (~jethead71@adsl-99-26-177-70.dsl.sfldmi.sbcglobal.net)
21:58:42 Quit jhMikeS (Changing host)
21:58:42 Join jhMikeS [0] (~jethead71@rockbox/developer/jhMikeS)
21:58:52thomasjfoxThe stack size is too small. With the cabbie theme it crashed on my while pressing pause/play
21:59:17gevaertsAh, maybe
21:59:27thomasjfoxI just increased the stack size and now it doesn't crash anymore
21:59:28gevaertsOh, speaking of pause/play...
21:59:50gevaertsDid you notice that if you pause (with default settings) the fade on stop takes down the system volume?
22:00
22:00:11gevaertsMaybe linking the volume directly isn't ideal
22:00:13thomasjfoxRight now that is "by design". I admit this is not ideal
22:00:17 Quit fml (Quit: ChatZilla 0.9.86 [Firefox 3.6.13/20101203075014])
22:00:34thomasjfoxOk, so we should keep the volumes separate
22:00:52gevaertsyes, but with linked settings maybe
22:01:10thomasjfoxHow would that linking look like?
22:01:19gevaertsi.e. link the system volume to the rockbox volume *setting*, not the active one
22:01:29*gevaerts doesn't really know that area...
22:01:57thomasjfoxI'll ask kugel how he implemented it on android. Better yet, I'll look at the code :)
22:02:17thomasjfoxI've found another issue I can't make my mind up yet:
22:02:26gevaertsMaybe we should start thinking about a .desktop file, so mere mortals can run the thing :)
22:02:33thomasjfoxI already have one :)
22:02:39thomasjfox.deb package creation is next
22:02:51gevaertsThat should be pretty simple I think
22:03:07*thomasjfox is not a debian guy
22:03:22*gevaerts has done a bit of debian packaging in his dark and obscure past
22:03:55thomasjfoxif you are bored and could give it a quick shot
22:03:58thomasjfox...
22:03:59thomasjfox:)
22:04:21thomasjfoxRegarind the issue I found:
22:04:23gevaertsI will, at some point.
22:04:31gevaertsThat might be next week though
22:04:49thomasjfoxWhen I switch from .mp3 codec to .wav codec, the .wav codec doesn't run. Just outputs "codec failure"
22:04:59thomasjfoxFirst I thought it's also the stack size
22:05:20thomasjfoxThough I increased it from 0x1000 to 0x6000 and it still doesn't work (codec usage is at 57%)
22:05:46thomasjfoxIf I go from WPS back to the main menu an re-enter, the codec works. Strange heh?
22:06:21thomasjfoxIt is also more likely to happen if I really stuff up the playlist (=add all files on the device)
22:06:37gevaertshm
22:06:47gevaertsmaybe kugel knows
22:07:08*gevaerts can do some work on this, but he's at about 600km from his scratchbox installation right now...
22:07:33thomasjfoxno worries. Just wanted to ask, maybe you have an idea what might be going on
22:08:01 Join Bagder [0] (~daniel@81-235-141-52-no63.tbcn.telia.com)
22:08:01 Quit Bagder (Changing host)
22:08:01 Join Bagder [0] (~daniel@rockbox/developer/bagder)
22:08:37thomasjfoxjhMikeS: ping
22:09:47jhMikeSyes?
22:09:58thomasjfoxhi
22:10:14thomasjfoxI've seen you are commting changes to make rockbox more multi core safe
22:10:32thomasjfoxEspecially adding "volatile" keywords
22:11:12GuySofthm, can i get the lastfm log to be located on an external memory card? i dont mind compiling if that is required
22:11:12thomasjfoxThis help with threading / multi core but doesn't fully solve the problem
22:11:18 Quit liar (Ping timeout: 255 seconds)
22:11:26saratogathomasjfox: when you load a codec on a lot of CPUs, you need to flush some of the CPU caches
22:11:35jhMikeSauditing things really and just fixing things that for now just silently work but aren't guaranteed to
22:11:45saratogathe codec loading code should handle that, but maybe its not implemented on your CPU?
22:11:50thomasjfoxvolatile won't really help you with this
22:12:00 Join liar [0] (~liar@clnet-p09-185.ikbnet.co.at)
22:12:03thomasjfoxIt just makes sure the compiler doesn't cache the var in a register
22:12:08saratogaalthough i'd expect that on teh A8 they'd have finally got around to fixing all those weird cache coherence issues on arm
22:12:21jhMikeSI know what volatile does and some thing need rereading every time they're compared
22:12:40thomasjfoxVolatile won't help you with hardware reordering
22:12:45jhMikeStechnically the scheduler itself is multicore safe, or PP wouldn't work
22:12:47thomasjfoxAtleast on Intel CPUs. Dunno about the details of ARM
22:13:03jhMikeSARM is no concern really
22:13:05thomasjfoxYou can get away with a lot of stuff if you just add the volatile keyword
22:13:07jhMikeSnot yet
22:13:21saratogaassuming you mean out of order execution, the CPU ensures that everything finishes as if it happened in order
22:13:34jhMikeSwhat am I getting away with? it's technically correct c code or the compiler may decide to cache it
22:13:50thomasjfoxYes
22:14:03thomasjfoxJust think about a multi core system. The CPU might decide to cache it
22:14:52jhMikeSit might, id hope so though I wonder, what is handled by the compiler there?
22:14:54thomasjfoxThis is a good starting point: http://stackoverflow.com/questions/2484980/why-is-volatile-not-considered-useful-in-multithreaded-c-or-c-programming
22:15:22thomasjfoxI just burned my fingers on the boost mailinglist recently
22:15:36thomasjfoxLet me look up the thread, it was pretty educating
22:16:09*jhMikeS wonders to what extent this is another "phear threading and cores" take...will find out
22:17:01thomasjfoxor what about atomic increments if two cores try f.e. to increment the "interrupt level" at the same time
22:19:03thomasjfoxjhMikes: Found it: http://groups.google.com/group/boost-list/browse_thread/thread/c6f07525f939967c
22:19:30thomasjfoxNot directly related to rockbox but it easily shows what can happen on Intel hardware
22:20:40jhMikeSthis probably doesn't apply to the sim because the mutexing prevents simultaneous access anyway
22:21:11thomasjfoxI'm more thinking about rockbox as application :)
22:21:47gevaertsWe're pretty single-threaded there as far as the CPU is concerned
22:22:19thomasjfoxThen volatile is enough
22:22:35thomasjfoxOh no scrap that
22:22:51thomasjfoxF.e. the gstreamer backend of the maemo port runs as separate OS thread
22:23:04thomasjfoxThis might run on a different core in a multi-core system
22:23:31gevaertsYes, but it should do locking, and locking should provide barriers
22:23:55thomasjfoxTrue. Let me just find the interrupt_level example I've recently seen
22:24:21jhMikeSthat's serialized with a mutex and condition
22:24:37thomasjfoxTrue. So that's fine, too
22:25:06thomasjfoxI just checked gcc's arm implemtation for atomic operations. They use compare and swap loops to do it
22:26:09jhMikeSswp(b) is deprecated on armv6 (of which no dual core exist right now) and doesn't really work properly on PP targets
22:26:55jhMikeSpeterson's algo was the final choice for PP
22:27:31thomasjfoxPP is the dual core arm also used f.e. in the Rio Karma
22:27:45*thomasjfox is new to these player CPUs
22:29:31thomasjfoxHow is "current_tick" updated?
22:29:34gevaertsIn the rockbox context, the big portalplayer users are older ipods and sansas
22:29:46jhMikeSarm11 has some IMB concerns but not too much
22:30:14gevaertsThere are a few more, but I suspect they're less popular
22:31:12 Quit selectahh ()
22:35:14jhMikeSthimosjfox: in the timer tick interrupt
22:35:26jhMikeS*thomasjfox
22:36:16thomasjfoxGuessed so. Just searching where it is :)
22:43:07*thomasjfox is reading the "ARM11 MPCore Processor Technical Reference Manual"
22:43:42 Quit bmbl (Quit: Verlassend)
22:45:58*jhMikeS gets annoyed at people that apply overly general rules to things about programming that when applied generally aren't really correct
22:47:03thomasjfox;)
22:48:07jhMikeSwho would init things like in that example? it looks like a total race condition anyway
22:49:27thomasjfoxThe code is fine in its current state and adding volatile definately helps against compiler caching
22:50:04thomasjfoxI don't think the compiler is smart enough to detect "thread" context switch performed by rockbox
22:50:36thomasjfoxSo it's a good thing what you are doing. I was just thinking about real multi-core systems like the Cortex-A9.
22:50:57 Join kugel [0] (~kugel@2.210.168.189)
22:50:58 Quit kugel (Changing host)
22:50:58 Join kugel [0] (~kugel@rockbox/developer/kugel)
22:54:36kugelthomasjfox: on android the volume separated
22:56:10 Join fml [0] (~chatzilla@manz-590ee82f.pool.mediaWays.net)
22:56:31jhMikeSthomasjfox: really, we avoid to much cache handling and there's no SMP so shared data tends to be in uncached, ordered ram and properly aligned
22:57:22thomasjfoxYes, this is true (for now)
22:57:37jhMikeScooperative threading really doesn't scale well to multiprocessor
22:57:37fmlkugel: hello. Do you know how all the file I/O functions are supposed to work in different environments (target/sim, core/plugin)? See my conversation with gevaerts a couple of hours ago.
22:57:53thomasjfoxWhen rockbox as an application is running on future multi-core CPUs, the fun might start
22:58:04 Quit dfkt (Quit: -= SysReset 2.53=- Sic gorgiamus allos subjectatos nunc.)
22:58:06thomasjfoxbtw: found a nice read: http://blogs.arm.com/software-enablement/169-locks-swps-and-two-smoking-barriers/
22:58:15thomasjfoxThere's also a part two to this article
22:58:34kugelfml: It's completely black magic
22:58:34jhMikeSthomasjfox: to my knowledge, the sim has been running on them for awhile now
22:58:37thomasjfoxAt least for ARM newbies like me.
22:58:44saratogai don't see why that would cause a problem
22:59:03saratogawe'll just lock all the threads to one cpu like we do now
22:59:16thomasjfoxsaratoga: That is fine if we can control it
22:59:29gevaertsthomasjfox: why wouldn't we?
22:59:38thomasjfoxA future meego port might not be able to do this or f.e. don't use gstreamer as a backend
22:59:39jhMikeSsaratoga: in the sim, we don't do that
22:59:42fmlSeems so. In order to understand it, one has to know what "application" means, how all the thing are compiled etc. I, for instance, don't know whether a sim build is an "application"
22:59:50thomasjfoxgstreamer uses OS threads to fetch PCM data
22:59:51saratogajhMikeS: ?
22:59:56fmlkugel: ^
23:00
23:00:09gevaertsthomasjfox: yes, but those are well isolated and easy to get right
23:00:11thomasjfoxkugel: I'll look into the volume issue.
23:00:16kugelthe sim can run on multiple cpus, it's using OS threads
23:00:17jhMikeSsaratoga: it's serialized but who know what cores those threads run on, could be any available
23:00:26saratogawell ok, same thing
23:01:11thomasjfoxThe current sim is fine as it locks -every- thread with a global mutex
23:01:19thomasjfoxSo this automatically issues a memory barrier
23:01:26thomasjfox(and is very expensive)
23:01:35saratogawhy is it expensive?
23:01:38kugelthomasjfox: I think we only might run into trouble if the rockbox threads are shared between cpus
23:02:03jhMikeSI sort of lie, the disk IO can leave the serialization and read/write
23:02:03thomasjfoxkugel: Exactly that could happen on a multi-core meego device
23:02:13kugelfor the hardware bits we consider them ad interrupt-like
23:02:39thomasjfoxkugel: The gstreamer "need-data" callback f.e. is running in a different OS thread
23:04:02 Join Buschel [0] (~chatzilla@p54A3B9CF.dip.t-dialin.net)
23:04:07thomasjfoxsaratoga: Because many threads are waiting for a single global lock
23:04:10eRivascould someone help me debug this radio screen theme: http://pastebin.com/L3JrpYdK
23:04:23saratogathomasjfox: isn't that the idea though?
23:04:32saratogathe global lock is to make sure each only runs one at a time right?
23:04:35thomasjfoxsaratoga: For the sim that's ok. When I use native threads instead of this approach, we save around 15% CPU
23:04:49thomasjfoxon the n900 f.e.
23:04:52kugelthomasjfox: no, the rockbox threads always run on the same core (if we use our scheduler)
23:04:53kugelthe gstreamer cb is no rockbox thread
23:04:53kugelfml: the io functions go through several layers of macros and wrapping
23:05:28thomasjfoxkugel: What about the thread that handles the SDL input events?
23:05:44thomasjfoxit also uses the queue_* functions
23:06:57kugelIt's complicated but not really doable in another way
23:07:33 Quit fml (Ping timeout: 255 seconds)
23:07:52thomasjfoxThe SDL input thread might run another core (like the gstreamer threads)
23:08:22thomasjfoxIn linxu there's a sched_setaffinity() function to bind a thread to a CPU
23:09:06thomasjfoxOr to be precise, it binds a process to a CPU
23:09:15kugelthe input event is also not a rockbox thread
23:09:50 Join JdGordon [0] (d31f494c@gateway/web/freenode/ip.211.31.73.76)
23:10:00thomasjfoxkugel: Yes, but the input event calls into rockbox
23:10:22kugelthe whole many core thing wouldn't work if the host couldn't handle it for us
23:10:46jhMikeSthomasjfox: maybe it does but they do go throught the "interrupt" serialization first
23:11:00 Quit thomasjfox (Read error: Connection reset by peer)
23:11:01kugelwe only do context switches on threads which are all on the same CPU
23:11:15gevaertsthomasjfox: basically the things that run in external threads in the application run in an interrupt handler on real hardware, so the mechanisms they use for communication are already well protected
23:11:38kugelright
23:13:02kugelthey can only use kernel primitives which do work as expected
23:13:45eRivascould someone help me debug this radio screen theme: http://pastebin.com/L3JrpYdK
23:14:01jhMikeSfunny thing is, none of it was protected at all for a long time...lol
23:14:06kugelthose which are interrupt-safe on real hardware
23:17:34 Join thomasjfox [0] (~thomasjfo@dslb-088-066-094-042.pools.arcor-ip.net)
23:18:08*kugel hopes thomasjfox knows the irc log page ;-)
23:18:22*thomasjfox is already on the irc log page :)
23:18:46*thomasjfox hates it when his first-generation core i7 board crashes the X server on a daily basis
23:21:33thomasjfoxkugel: Ahh, the sim_enter_interrupt_handler() does the trick
23:21:33 Join stoffel [0] (~quassel@p57B4CB51.dip.t-dialin.net)
23:21:44thomasjfoxkugel: for the SDL event thread
23:22:07n1ssoap: re FS #11843 svn revert does not delete new files so i think your problem is that the contents of the new .S file are in there 3 times (from aech applying of the patch)
23:22:48n1ssince the patch only adds lines to the file, the patch applies fine regardless of the file's contents
23:22:51soapn1s, thank you very much!
23:23:15soapis there a SVN command to delete non-svn files?
23:23:16thomasjfoxkugel: What about gstreamer calling pcm_play_get_more_callback()?
23:23:23soapwait, I don't want that anyway!
23:23:40n1ssoap: svn status should show them with a ?
23:24:05soapyea, it does.
23:24:18[Saint]svn revert -R . && svn st | awk '/^?/ {print substr($0,8)}' | (while read a ; do rm -vr "$a" ; done)
23:24:34soapI was just curious if there was a "svn cleanhouse" command, but then I thought to myself "I don't want that!"
23:24:41[Saint]shit, - the revert in front.
23:24:46gevaertssaratoga: "print $2" will work just as well :)
23:25:52thomasjfoxkugel: The SDL audio callback also runs in a separate OS thread (see SDL_OpenAudio manual)
23:26:31kugelpcm_get_more is safe I'd say
23:27:14kugelaudio stuff usually runs on a separate thread, on android too
23:27:55kugelon dma interrupt on real hardware
23:28:41Buschelsoap: I just see you're still fighting with the *.S file ?
23:31:27Buschel[Saint]: could you already give the second build a chance?
23:31:41 Quit parafin (Quit: So long and thanks for all the fish)
23:33:14Buschel[Saint]: interesting to see that with the v03 patch YUV and RGB have nearly the same speed. this means the LCD IF is the bottle neck. right now there is no register known to control the clock dividers for color displays, but only for monochrome.
23:36:28thomasjfoxkugel: I guess the get_more() function most somehow lock as it might occur on any time
23:36:30 Quit kugel (Quit: Yaaic - Yet another Android IRC client - http://www.yaaic.org)
23:36:37 Join kugel_ [0] (~kugel@e178059234.adsl.alicedsl.de)
23:36:51thomasjfoxkugel_: I guess the get_more() function most somehow lock as it might occur on any time
23:37:43[Saint]LCD 1/1: 19.4, 1/4: 76.0 LCD YUV 1/1: 17.0, 1/4: 67.1 @ 30MHz ; LCD 1/1: 51.7, 1/4: 202.5 LCD YUV 1/1: 45.5, 1/4: 180.0 @ 80MHz
23:37:47[Saint]Buschel: ^
23:39:00 Join jgarvey [0] (~jgarvey@cpe-065-190-066-089.nc.res.rr.com)
23:39:04 Quit kugel_ (Read error: Connection reset by peer)
23:39:10 Quit jgarvey (Client Quit)
23:39:22 Join kugel [0] (~kugel@rockbox/developer/kugel)
23:39:48Buschel[Saint]: thanks :)
23:40:05[Saint]Buschel: as expected?
23:40:33Buschelkind of. faster than svn, slower than the last test
23:41:05 Join kugel_ [0] (~kugel@g231108137.adsl.alicedsl.de)
23:41:09Buschelthis is the safe version which still polls for fifo ready before writing to FIFO
23:41:14 Quit kugel (Disconnected by services)
23:41:32[Saint]Any more tests you'd like to throw at me?
23:41:51[Saint]Is there some form of "Who has what target?" page?
23:42:02Buschela pitty you cannot build, i will prepare one or two binaries for you...
23:42:13 Nick kugel_ is now known as kugel (~kugel@g231108137.adsl.alicedsl.de)
23:42:17 Quit kugel (Changing host)
23:42:17 Join kugel [0] (~kugel@rockbox/developer/kugel)
23:42:20n1s[Saint]: RockboxTesting IIRC
23:43:17n1syes, that's the one, doesn't seem people are ever taken off the list though
23:44:03gevaertsWell, people are supposed to edit this for themselves
23:44:43[Saint]Buschel: What is your connection like? If you could deflate/compact/compress the eabi toolchain (or just usr/local) I'd be able to, rockboxdev.sh is giving me an error I can't figure out.
23:44:49 Join kugel_ [0] (~kugel@e178059025.adsl.alicedsl.de)
23:45:09 Quit kugel (Disconnected by services)
23:45:18Buschel[Saint]: my upload is poor
23:45:19[Saint]I'm unsure *exactly* wjhat files are needed, I've been meaning to organise making an eabi cygwin package for ages.
23:45:46n1s[Saint]: wouldnt' it be bette to fix your error?
23:46:14[Saint]n1s: Almost vertainly, but probably not quicker.
23:46:19[Saint]*certainly.
23:48:17Buschel[Saint]: can you please check 'View I/O ports' with this -> http://www.sendspace.com/file/j72rit
23:49:15[Saint]yes, what am I expecting to see there?
23:49:21Buschel[Saint]: I am interested in MLCD_SCLK_DIV and the three items below
23:49:33[Saint]Ok.
23:49:38Buschel[Saint]: there will be some register contents displayed
23:50:07 Quit Llorean (Quit: Leaving.)
23:50:14[Saint]is this just the rockbox.ipod file?
23:50:24Buschelyep
23:50:50***Saving seen data "./dancer.seen"
23:54:51thomasjfoxkugel_: Is this working? :)
23:55:02 Quit kugel_ (Ping timeout: 240 seconds)
23:55:09Buschel[Saint]: btw, I have added a copy-paste error to the code. it will display the same register on the left side for all three items below MLCD_SLCK_DIV
23:55:36Buschel[Saint]: the register contents are correct though
23:55:42[Saint]inl (0x60006024): 00000000
23:55:46[Saint]for all three
23:55:50thomasjfoxjhMikeS: Can I ask you one more question on the DMA <-> pcm data transfer?
23:56:00Buscheland MLCD... ?
23:56:25[Saint]00000000
23:56:38 Join froggyman_ [0] (~seth@98.115.0.7)
23:56:43 Quit froggyman (Read error: Connection reset by peer)
23:57:22[Saint]MLCD_SCLK_DIV: 00000000
23:57:24[Saint]inl (0x60006024): 00000000
23:57:24[Saint]inl (0x60006024): 00000000
23:57:25DBUGEnqueued KICK [Saint]
23:57:25[Saint]inl (0x60006024): 00000000
23:58:07 Join kugel [0] (~kugel@rockbox/developer/kugel)
23:58:27[Saint]Buschel: is that what you wanted?
23:58:58kugelthomasjfox: sorry, having net problems currently. I don't know the details about get_more() but it is interrupt-safe

Previous day | Next day