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 2010-12-09

00:01:35 Quit pamaury (Remote host closed the connection)
00:03:32 Join AlexP_mob [0] (~ap@rockbox/staff/AlexP)
00:09:43***Saving seen data "./dancer.seen"
00:10:45 Quit bertrik (Quit: :tiuQ)
00:12:48 Quit mortalscan (Ping timeout: 245 seconds)
00:14:15 Quit casainho (Remote host closed the connection)
00:15:24 Quit komputes (Remote host closed the connection)
00:15:44 Quit wodz (Quit: Leaving)
00:17:14 Quit jgarvey (Quit: Leaving)
00:20:48 Quit mc2739 (Quit: leaving)
00:23:51 Quit timccc (Quit: Leaving.)
00:26:07n1ssaratoga: ii wouldn't mind if you deleted the "simple guide to compiling" especially since it's longer and more involved than any of the regular guides (and as you say it doesn't even work)
00:27:36n1simo, since noone seems interested in keeping the directions for cygwin up to date we should drop them as they just confuse people more than they help
00:28:04 Join leavittx [0] (~leavittx@
00:31:51 Join timccc [0] (~timccc@
00:32:14 Quit The_Pwny (Quit: Light travels faster then sound, which is why some people appear bright, until you hear them speak)
00:34:10 Quit AlexP_mob (Quit: Cheers)
00:39:26 Quit n1s (Quit: Lämnar)
00:41:38 Quit markun (Read error: Connection reset by peer)
00:50:38 Quit user890104 (Ping timeout: 272 seconds)
01:00:43 Quit dfkt (Quit: -= SysReset 2.53=- Sic gorgiamus allos subjectatos nunc.)
01:02:44 Join markun [0] (~markun@rockbox/developer/markun)
01:21:29 Join krabador [0] (
01:21:50 Join mc2739 [0] (~mc2739@rockbox/developer/mc2739)
01:24:09 Part toffe82
01:33:10 Join fyrestorm [0] (
01:37:54 Quit fyrestorm (Client Quit)
01:39:41 Join fyrestorm [0] (
01:43:37rasherJdGordon|: so you think the metadata log patch is fine to go in?
01:45:39JdGordon|well you ignored the one code comment i had
01:48:33rasher Mostly out of laziness, admittedly
01:48:55 Quit GeekShadow (Quit: The cake is a lie !)
01:49:04 Quit Llorean (Read error: Connection reset by peer)
01:49:30JdGordon|not reaslly the best or arguments....
01:49:34 Quit hebz0rl (Quit: Leaving)
01:50:08rasherWell is there actually any good reason to bother with the snprintf?
01:50:19rasherOther than "that's how we usually do it"
01:50:55JdGordon|1 write is better than 2
01:51:10rasherBut is 1 write and a snprintf better than 2 writes?
01:51:20rasherAnd in what way is it better?
01:51:53JdGordon|if you asked someone else they might give you athe answer you want to hear
01:52:08rasherI am actually asking what the argument is
01:53:39 Join Llorean [0] (~DarkkOne@rockbox/user/Llorean)
01:53:45JdGordon|it is cleaner the other way, clean is subjective
01:54:15 Quit tchan (Read error: Connection reset by peer)
01:54:25rasherI see
01:54:41JdGordon|the point is you asked my opinion and i gave it
01:55:05 Join tchan [0] (~tchan@lunar-linux/developer/tchan)
01:55:30 Quit Rob2222 (Quit: Rob2222)
01:55:43 Quit efyx (Remote host closed the connection)
01:57:11rasherSure, just curious to know what you were basing it on
02:09:46***Saving seen data "./dancer.seen"
02:16:11 Quit JdGordon (Ping timeout: 255 seconds)
02:19:12 Quit robin0800 (Remote host closed the connection)
02:23:51 Join madalu [0] (~user@unaffiliated/madalu)
02:33:47 Join BHSPitMonkey [0] (~stephen@unaffiliated/bhspitmonkey)
02:33:49 Quit kadoban (Remote host closed the connection)
02:34:02 Join kadoban [0] (
02:41:06 Quit [Saint] (Ping timeout: 265 seconds)
02:53:19 Join coquito222 [0] (
02:53:59 Quit Keripo (Ping timeout: 240 seconds)
02:54:17 Quit Jamo (Quit: Leaving)
02:55:47 Join mortalscan [0] (~mortalsca@
02:58:06 Join Keripo [0] (
03:00:15 Quit coquito222 (Quit: CGI:IRC (Ping timeout))
03:12:59the_KyleI have the real Clip+ now and have installed Rockbox. The problem is that the context menu in the file browser is not speaking correctly. In the simulator, if I held select, I heard "playlist" and in that menu there several functions including "insert," "insert shuffled," etc. On the real Clip+ instead of "playlist" I hear "insert next." Then if I go into the menu, I get "insert last" and "queue next."
03:16:05the_KyleI installed the prebuilt release on the real Clip+ instead of building from source, if that helps.
03:17:17 Join JesusFreak316 [0] (
03:17:30*the_Kyle wonders if he should have built the Clip+ firmware from source if using a built voice file. The source I have is from the 3.7 release.
03:21:17rasherthe_Kyle: are you using the same voice file?
03:22:00rasherThat won't always work. Using a voice file of the "wrong" version will get shifts like that, I believe
03:23:11the_KyleI've tried two different files now, both a file I built using eSpeak and the voice file I built from my patched version that uses Pico. I believe both versions are from the 3.7 source I downloaded. I can try the original voice file with the 3.7.1 release I installed, but I don't like the Festival speech.
03:23:32*the_Kyle tries the 3.7.1 voice file.
03:25:21 Quit mortalscan (Ping timeout: 264 seconds)
03:27:40the_KyleThat worked. So what sources do I need to rebuild my Pico and eSpeak voice files for 3.7.1?
03:31:09 Quit MethoS- (Read error: Connection reset by peer)
03:31:15rasherThe 3.7.1 sources, of course :)
03:31:21rashersvn co svn:// should get them for you
03:31:56the_KyleSorry, I meant to ask for the link rather than the version, which I already knew.
03:32:03*the_Kyle runs away and hides.
03:32:30rasherNono, it's fine
03:32:34 Join [Saint] [0] (cbb802e9@gateway/web/freenode/ip.
03:34:14the_KyleHas anything changed in the voice tool and the configure tool from 3.7 to 3.7.1? I mean can I copy my patched versions of the tools from 3.7 so I can use Pico, or should I patch them in the 3.7.1 source?
03:35:57rasherI can say with high confidence that nothing has changed there
03:42:10 Join T44 [0] (
03:42:34mordocaiIs there anyone who is currently actively working on the sandisk sansa familty of players?
03:43:14rashercould you be more specific? That's a pretty broad question
03:44:11mordocaiWell, mainly this bug is annoying me and I was wondering who to ask about it:
03:45:19 Quit Topy44 (Ping timeout: 245 seconds)
03:46:46 Quit Keripo (Quit: Leaving.)
03:46:59 Join eWill [0] (
03:47:41 Join Keripo [0] (
03:48:05eWill(building from Cygwin) Does it matter what dir you run '' from? I'm asking because is downloading 'binutils-2.20.1.tar.bz2', but I already installed a later version through Cygwin setup....?
03:48:33[Saint]that's to be expected.
03:48:46 Quit Keripo (Read error: Connection reset by peer)
03:48:47[Saint]it requires a very specific set of tools to build the toolchain.
03:49:29[Saint]eWill: ^
03:49:50eWillthanks. −−−− checking for correct version of CLooG... no
03:50:13eWillchecking for gnatbind... no checking for gnatmake... no
03:50:49[Saint]that's's good to have those things...but it won't fail if it doesn't.
03:51:00 Join Keripo [0] (
03:51:09[Saint]cloog must have missed a dependancy during the install.
03:51:53krabadori have a theme issue with fuzev1: when i load a theme, i load a second, and i come back to the first, the first isn't properly loaded, it have some setting of the last theme
03:52:31[Saint]that's a problem with the theme not setting all elements correctly.
03:53:00[Saint]you could try figure out what isn't getting set, and add it to the theme.cfg, or attempt to contact the theme author.
03:53:05[Saint]krabador: ^
03:55:27eWill[Saint]: will the results of building the eabi toolchain be the same with or without cloog?
03:57:29 Quit krabador (Ping timeout: 272 seconds)
03:58:44 Join krabador [0] (
03:59:51krabadorthe bad thing is that the various theme will never load their settings correctly
04:00:49krabadora RB reinstallation would needed to use the theme
04:00:54 Quit Strife89 (Quit: Reboot)
04:01:10[Saint]krabador: a lot of themes don't bother to unset things that may have been set by other themes.
04:01:32[Saint]you could maybe get around it by reverting to the fallback theme inbetween.
04:02:03*[Saint] needs to wander off for an hour or so...
04:02:16 Quit designate72 (Ping timeout: 265 seconds)
04:02:32krabador[Saint], good time :)
04:03:31 Join Strife89 [0] (
04:03:59 Quit Keripo (Quit: Leaving.)
04:04:43 Join Keripo [0] (
04:07:14 Quit Keripo (Client Quit)
04:07:24 Join Topy [0] (
04:07:59 Join Keripo [0] (
04:09:02 Quit sasquatch (Quit: WeeChat 0.3.2)
04:09:27 Join sasquatch [0] (
04:09:47***Saving seen data "./dancer.seen"
04:10:31 Quit InsDel (Read error: Connection reset by peer)
04:10:40 Quit T44 (Ping timeout: 245 seconds)
04:11:49the_KyleIs it safe to remove clppa.bin from my player now? What about all those extra sys files and the like?
04:12:19 Join Barahir [0] (
04:13:35the_KyleI can backup those files if they are important for the original firmware, but at this point, I have no plans of running the original firmware.
04:13:48the_KyleUnless it's needed for USB.
04:15:44 Quit Barahir_ (Ping timeout: 260 seconds)
04:16:00 Join Rob2222 [0] (
04:16:17 Join DerPapst1 [0] (
04:16:44mordocaithe_Kyle: Yeah, the original firmware is currently still used for USB.
04:18:07the_KyleSo do I need all the extra sys files and the sdk something file, or can those be removed? I wouldn't think those would be needed for USB, since USB is needed before those files can be accessed.
04:18:09 Quit DerPapst (Ping timeout: 240 seconds)
04:18:57mordocaithe_Kyle: Truthfully, I don't know (hopefully someone else does?). I personally just stayed on the safe side and left them all. (I have a clip+ as well)
04:19:42the_KyleAnd is clppa.bin still needed, or can I remove that, since it was just used to patch the OF with the bootloader?
04:20:48 Join fdinel [0] (
04:21:47Lloreanthe_Kyle: I didn't even have a clppa.bin on my Clip+ after installation.
04:22:07*the_Kyle looks. Maybe it gets deleted automatically.
04:22:50the_KyleSeems it does. It's not on mine now either.
04:23:27LloreanI can't speak for the Clip+ but many of the Sandisk players (e200/c200 for example) didn't require any files for the OF to work, everything the OF required was in a different partition that wasn't accessible in a file browser.
04:24:36*Llorean just got his Clip+ earlier today and didn't even look at the other files on it.
04:25:15the_KyleI have something called DID.BIN, MTABLE.SYSRES_INFO.SYS, SYS_CONF.SYS and VERSION.SDK.
04:25:44 Join T44 [0] (
04:27:10the_KyleMost of these files are very small, but MTABLE.SYS is almost 2 MB.
04:27:16LloreanMy *guess* is that they're all unnecessary, but I wouldn't bother deleting them anyway. They're probably the database and such of the original firmware, and I'd bet that if they were unnecessary they'd still get re-created each time the OF is booted.
04:27:35LloreanI believe mtables.sys is specifically the original firmware's database of tracks on the device.
04:28:11the_KyleIn that case, it's huge, especially since there were no tracks on the device before I installed Rockbox.
04:28:21LloreanIf you're using a build with Rockbox's USB mode enabled, you might be able to remove them (or rather, you almost certainly can) but until then, there's not much point.
04:28:35 Quit Topy (Ping timeout: 245 seconds)
04:28:57the_KyleWhen was USB enabled for this device? I'd like to be able to get rid of this stuff.
04:29:29LloreanI don't believe it has been yet, but you could enable it in a custom build. There's a patch on flyspray you'd want to install first, as USB isn't reliable
04:29:43LloreanWIth the patch it's more reliable (possibly 100%) but has some problems under Windows still (slow speeds)
04:30:35the_KyleIf it's good on Linux, I may try it.
04:30:37 Quit krabador (Quit: Sto andando via)
04:30:47the_KyleI won't be putting this on Windows.
04:31:21LloreanI don't know what the flyspray task is but it's mentioned on the SansaAMS page on the wiki I believe
04:32:16the_KyleSounds good. Do you know if it patches against 3.7.1, or if I need latest svn?
04:34:48mordocaiLlorean: the_Kyle: By the way,(not really related to the usb things) I would appreciate it if both of you would check if you have this issue I had it with both 3.7.1 as well as the lastest build.
04:35:29 Quit amiconn (Disconnected by services)
04:35:30 Quit pixelma (Disconnected by services)
04:35:30 Join amiconn_ [0] (quassel@rockbox/developer/amiconn)
04:35:31 Join pixelma_ [0] (quassel@rockbox/staff/pixelma)
04:35:46 Nick pixelma_ is now known as pixelma (quassel@rockbox/staff/pixelma)
04:35:50 Nick amiconn_ is now known as amiconn (quassel@rockbox/developer/amiconn)
04:35:51Lloreanthe_Kyle: Probably the latest SVN. It's rare for patches to be posted against a release build
04:36:39the_KyleLiving on the edge. But probably a good thing, since I really want to test and contribute what I can, input if nothing else.
04:36:45 Quit fdinel (Quit: Miranda IM! Smaller, Faster, Easier.
04:36:55Lloreanthe_Kyle: 99% of the time SVN is better to use than the latest release anyway. :)
04:37:23Lloreanmordocai: I can't really test that right now, I only have a single audio file on my Clip+ (an audiobook), as I don't intend for it to be my primary player, but rather something for when I'm exercising or whatnot. I'll try to remember to test it later.
04:37:23the_KyleSounds good if I can be assured that no unrecoverable breakage will occur.
04:39:14the_KyleUnlike Llorean, my clip+ is now my primary player.
04:39:51 Quit TheSeven (Ping timeout: 272 seconds)
04:39:52the_KyleIf I can be sure that I can fix something if it breaks, I'll certainly update to latest svn.
04:42:06mordocaithe_Kyle: I got my clip+ recently, but it is my only player. I've used svn for weeks and never had a problem... I'd say even if something breaks, you could always roll back to the release version or an earlier commit.
04:42:44the_Kylemordocai: Sounds good. Well, consider me a Rockbox tester on clip+.
04:43:04 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
04:46:41the_KylePardon my slight ignorance of how svn workd, but if I have a release, do I need to checkout the latest source, or can I update directly from my release source?
04:48:21 Join Topy [0] (
04:49:39 Quit DerPapst1 (Quit: Leaving.)
04:50:40 Quit T44 (Ping timeout: 245 seconds)
04:53:48eWillthe_Kyle: I had a folder on my Fuze v2 called "##PORT#" that was ~200MB. I deleted it, then filled the device so that there is less than 177MB free space (the size of the biggest file in that dir). Now no more "Refreshing Your Database". Though I can't boot the OF (says I must free up space for the database), I can still charge and transfer files in the OF.
04:55:16the_KyleSounds safe to me.
04:55:43eWillthe_Kyle: Oh yeah −− you can't see that dir unless you are in Rockbox, or have a build that allows RB to handle USB, and connect through RB.
04:56:28the_KyleI don't plan on using the OF for anymore than is absolutely necessary anyway. I need the voice, and my clip+ is essentially a brick if it doesn't talk.
05:00:46the_KyleIf I apply a patch to svn, can I roll back to before tht patch if necessary? I know it patches usually survive commits.
05:08:34the_Kylemordocai: Does the bug you link to have any effect on what you hear? Unfortunately, I'm unable to see the error message, but in the simulator I noticed that if I insert two albums the music stops. I am able to resume playback, but that's all I can confirm at this point. I haven't tried it on the player yet.
05:09:19mordocaithe_Kyle: Well, the playback will certainly stop. But I have to turn off the player and turn it back on to get playback to resume.
05:10:23the_Kylemordocai: I was able to resume playback by hitting what seems to be the simulated power button a couple of times. I can't confirm on the player yet though. Will try as soon as I get it updated to trunk.
05:16:00mordocaieWill: Thanks for the ##PORT# tip. Let me get a few more songs on the main player instead of the SD card.
05:16:26 Join T44 [0] (
05:18:16 Join user890104 [0] (
05:19:00 Quit Topy (Ping timeout: 245 seconds)
05:25:27 Quit user890104 (Ping timeout: 272 seconds)
05:38:35 Quit shai (Ping timeout: 245 seconds)
05:42:35the_KyleWhat is the compiler I need to build Rockbox for Clip+? It says I need arm-elf-eabi-gcc, but I have no package by that name. gcc-arm-linux-eabi-gnu doesn't work.
05:43:38kadobanthe_Kyle: use the script to generate it for you (on linux, right?) it's in tools/
05:47:12the_KyleGot it. Thanks.
06:03:00the_KyleIs fs11664 the patch I need to make USB mode work on Sansa?
06:04:25 Quit JesusFreak316 (Ping timeout: 245 seconds)
06:09:29 Quit MagusG (Ping timeout: 260 seconds)
06:09:40 Quit linuxguy3 (Ping timeout: 240 seconds)
06:09:49***Saving seen data "./dancer.seen"
06:11:50 Join linuxguy3 [0] (
06:28:26the_KyleI'm getting a "no PREFIX given" error when trying to build my install. The sim didn't do this. Should I go ahead and set it to .rockbox?
06:37:47 Join Zarggg_ [0] (
06:38:36 Join shai [0] (
06:40:11 Quit Zarggg (Ping timeout: 255 seconds)
06:45:13eWillfuuuuu. I just spent 3 hours with running, i get an error, and it quit. Anyone willing to take a look at the last few lines?
06:45:55 Join mortalscan [0] (~mortalsca@
06:47:41eWillrunning in Cygwin, btw.
06:48:05 Quit madalu (Ping timeout: 264 seconds)
06:54:52the_KyleI'm not getting a .rockbox directory anywhere. I even set my prefix to the directory where the Makefile is, but .rockbox isn't being built.
06:56:18[Saint]the_Kyle: what are the steps you are taking in trying to build?
06:57:55[Saint]eWill: I can't really say what xgcc was unavailable at the time, but the only real option is to start again. You *must* delete tmp/rbdev-build/ to run the script again.
06:58:11[Saint]s/say what/say why/
06:58:48eWillcould high cpu usage cause that maybe?
06:58:57eWilli was playing a game
06:58:59the_KyleMy source is in ~/src/rockbox. I'm trying to build in ~/cliprock. cd ~/cliprock, ~/src/rockbox/tools/configure, selected target 62 normal build, make, make fullinstall.
06:59:31the_KyleMaybe I need to build in my source directory?
06:59:47the_KyleBut the sim builds fine in an alternate directory.
07:01:46eWill[Saint]: But I couldn't have gotten that far if xgcc was installed-incorrectly/not-installed −− right?
07:03:10the_KyleI also should mention that I'm trying to build trunk with the USB patches applied.
07:09:07[Saint]the_Kyle: yes, you need to build inside the source tree
07:09:35[Saint]make "~/src/rockbox/build/" or similar, and build in there.
07:11:25[Saint]eWill: it's possible that high CPU usage was a factor...It's really too hard to say why the particular resource was unavailable at that point in time.
07:11:35*the_Kyle tries to build in ~/src/rockbox/build. Standing by for build.
07:11:51[Saint]And yes, it is possible that the script got that far before encountering an error.
07:17:34the_KyleI'm still getting a "No PREFIX given" error.
07:17:57the_KyleI do, however, have a .rockbox directory this time.
07:22:33the_KyleLooks like I'm missing nvram.bin in my build, unless that's created on the player. The nvram.bin file is on my player but not in my new build. It seems to be the only file that's missing in the top-level .rockbox directory.
07:24:17the_KyleStrang thin though is that the .rockbox directory on my player is 8.6 MB, but my build is 21 MB.
07:27:54 Quit JdGordon| (Quit: leaving)
07:30:28 Join JdGord [0] (~jd@
07:33:17eWillis it ok to install libiconv AND libiconv2?
07:34:33the_KyleWell, apparently I don't have a problem. My build seems to work on the player. WOE! Does this wakeup stuff really work?
07:49:44the_KyleLooks like the USB stuff is working as well. I have the ##PORT# directory and the ##MUSIC# directory, which I assume can be removed safely now that I got it working. Starts much faster too.
07:52:40 Quit simonrvn (Ping timeout: 240 seconds)
07:54:22eWilldoes '' have a 'clean' or similar argument?
07:55:37eWillI asking because I deleted everything an the Cygwin dir except the 'home' dir, reinstalled Cygwin, and now fails with: /tmp/rbdev-dl/binutils-2.20.1.tar.bz2: No such file or directory
07:56:31[Saint]eWill: No, all you need to do is delete <CygWin>/tmp/redev-build/
07:57:30eWill[Saint]: ROCKBOXDEV: Downloading using wget
07:57:49eWill..../tmp/rbdev-dl/binutils-2.20.1.tar.bz2: No such file or directory
08:00:04 Join simonrvn [0] (simon@
08:00:10[Saint]did you delete rbdev-dl instead?
08:01:08eWilldeleted both
08:01:25[Saint]It definitely shouldn't complain about rbdev-dl existing...I always place mine manually, as opposed to letting the script download it each time, I just backed it up.
08:01:26eWillEVERYTHING outside of /home
08:01:50eWillI didn't mean to delete the dl dir
08:02:05[Saint]well, with rbdev-dl and rbdev-build deleted, there shouldn't be any remnants of the failed toolchain build left.
08:02:43eWilli'm gonna download the file manually
08:04:22[Saint]well, I said it doesn't care *how* the files get there, as long as they're there. It just attempts to download them using wget (or curl if it can't fing wget) if it doesn't find the files it needs in the rbdev-dl dir
08:05:47eWillROCKBOXDEV: Downloading using wget
08:05:49eWill/tmp/rbdev-dl/binutils-2.20.1-ld-thumb-interwork-long-call.diff: No such file or directory
08:06:38eWillmanually downloading now. wtf?
08:07:43[Saint]are you making sure the files are in that dir *before* running the script?
08:08:33eWillyeah, after I manually dl them and put them there, it is working, but −−−− checking for version 0.10 of PPL... no checking for correct version of CLooG... no −−−− how do I install v0.1 of ppl? and what is correct version of Cloog?
08:09:14[Saint]you may not have installed the devel versions of cloog and ppl
08:09:42[Saint]just use the search feature of the cygwin setup and search for cloog and ppl and select all those packages.
08:09:46[Saint]they aren't large.
08:09:52***Saving seen data "./dancer.seen"
08:09:55eWillthe_Kyle: it's in ''
08:10:04[Saint]However, not having them isn't going to make the build fail.
08:12:16eWilldo you mean to install 'ppl-devel'?
08:12:38[Saint]the_Kyle: you should be doing "cd <build_dir> && ./../tools/configure/ && make && make fullzip"
08:12:46[Saint]not "make fullinstall"
08:13:21the_Kyle[Saint]: That would explain things. Thanks.
08:14:42[Saint]you'll only need to do "make fullzip" once unless you change the fonts...after doing "make fullzip" if you build again in the same build dir after that you'll only need to do "make zip" from then on.
08:15:24the_KyleIt worked. So now I can just unzip into my player, just like the release?
08:17:56eWill[Saint]: do I need to install the '*-devel' versions of any of these: gmp, mpfr, curl, wget, cloog-ppl, flex, expect, bison, ppl, ccache, libiconv?
08:18:21 Quit CaptainKewl (Ping timeout: 240 seconds)
08:18:42 Join MagusG [0] (
08:21:49[Saint]eWill: won't wurt to.
08:21:54the_KyleI still don't see an nvram.bin, but the zip built without errors, and not having this file at install time doesn't seem to be a problem.
08:23:33[Saint]nvram.bin isn't built at this time.
08:23:56[Saint]it is a file that Rockbox uses, it isn't built at compile time.
08:24:10eWillwell rockboxdev is running again. I'll see you in 3 hours (maybe).
08:24:18the_KyleSo it builds on the player?
08:26:14eWill[Saint]: Can you check your current version of ppl and cloog?
08:26:14 Quit markun (Read error: Connection reset by peer)
08:30:18*scorche adds another field to the forum registration page in a further effort to combat spam (might be successful, but who knows) - I can always change it to something else if it causes problems too...
08:33:07 Join Zagor [0] (~bjst@rockbox/developer/Zagor)
08:33:38scorcheif we are still having spam problems, i can see about trying this mod:
08:37:00the_KyleIs there any way to use tools/ to create a in the top-level without having to recurse through every directory on my player yet again
08:37:17the_KyleI just want to add one more folder that isn't .talked yet.
08:38:27 Join wodz [0] (
08:38:29the_KyleI know I can create the .talk file inside that directory that I added, but it doesn't do that way except for subdirectories.
08:38:33wodz[Saint]: ping
08:39:14 Quit amiconn (Remote host closed the connection)
08:39:15 Quit pixelma (Remote host closed the connection)
08:40:50[Saint]wodz: pong
08:41:22 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
08:46:48 Quit JdGord (Quit: Bye)
08:47:24wodz[Saint]: you seem to be art guy, would you mind looking at svg included in tarball on FS #11796 and give it more realistic look? As a reference
08:47:45 Quit mortalscan (Ping timeout: 245 seconds)
08:48:11Zagorscorche: is akismet an option? it's free for smallish sites.
08:48:28scorcheZagor: yes, it is
08:49:13Bagderakismet is quite good for wordpress at least
08:49:18Zagoryes I saw that mod. I meant rather "is there a reason I'm unaware of not to use it"
08:49:31scorchei just havent looked into it
08:49:49 Quit Keripo (Quit: Leaving.)
08:51:32wodzhow one get preprocessed source file? in functions.make there is a rule but I can't find how to use it
08:52:29 Join pixelma [0] (quassel@rockbox/staff/pixelma)
08:52:29 Join amiconn [0] (quassel@rockbox/developer/amiconn)
08:52:46Bagderuse make V=1, then copy the compile command line and add -E
08:58:49 Join ender` [0] (
09:00:14 Part the_Kyle
09:01:37 Join the_Kyle [0] (~kyle@
09:04:21[Saint]wodz: I may not have time to look at it tonight, but, yes...I can look into it for you.
09:05:49wodz[Saint]: thx
09:08:24 Quit eWill (Quit: ChatZilla 0.9.86 [Firefox 3.6.12/20101026210630])
09:10:20 Quit Rob2222 (Ping timeout: 240 seconds)
09:13:08 Quit BHSPitMonkey (Remote host closed the connection)
09:20:53pixelma<[Saint]> the_Kyle: yes, you need to build inside the source tree <−− that's wrong, I build outside the source tree all the time. This is of course has an influence on the path to configure (more typing) but I like it better because my source tree stays "clean" and "make reconf" helps reducing the typing in most cases...
09:24:52pixelmathe_Kyle: about your voice file being a bit off at the beginning - in most cases you can't use voice files made for the sim on the player as the sim may have different features (e.g. in the Clip+ case there is no microSD simulation and this comes at least with one extra language string). Just saying as I can't tell from the logs if you got that, if you did please ignore me :)
09:27:53 Join markun [0] (~markun@rockbox/developer/markun)
09:30:45 Quit markun (Read error: Connection reset by peer)
09:33:43the_Kylepixelma: Thanks for the info. I did end up building inside the source tree, but I can easily move my build out, since I created it as a subdirectory. As for the voice files, they seem to work both on the sim and on the real player. The problem seems to have been that I was trying to use a 3.7 voice file on 3.7.1, which I have since corrected, a positive sideaffect of updating to trunk to apply the USB patches.
09:36:19pixelmayes, the voice files will work as in "sound correctly" but the spoken string doesn't match the actual highlighted menu item (that's what I meant with being "off")
09:36:36pixelmaso you can't really use it
09:37:27the_KyleRight. But I don't think it was the sim so much as different Rockbox versions.
09:37:57the_KyleI have it matched up now, and it works better than ever so far.
09:38:35pixelmaI wpould have thought that voice files for 3.7 should work with 3.7.1 as the latter is just 3.7 with backported bug fixes. I don't know for sure though if one or two new language strings slipped in
09:39:00 Join pamaury [0] (
09:39:00 Quit pamaury (Changing host)
09:39:00 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
09:42:32the_KyleI'm not sure either. The sim seems to like the 3.7.1 foice file and the 3.7 file I built, but only 3.7.1 was able to run correctly on the player.
09:43:40the_KyleI built the 3.7 voice file using the 3.7 source tree configured for sansaclipplus. No reason why that shouldn't have worked.
09:44:24the_KyleUnless, of course, there was a string or two that snuck in somewhere.
09:44:56the_KyleI know that current svn has several strings that aren't in 3.7.1.
09:50:07the_KyleRegarding the data abort mentioned earlier by mordocai, I'm not getting it here. I just added three albums to the playlist without any errors. So that's either good news, because the error has been corrected in the latest build, or bad news, because it can't be reproduced easily.
09:51:14the_KyleRunning svn trunk at the moment, and it's virtually error free over the last 20 minutes or so. No noticeable bugs yet.
09:52:03wodz20 minutes is not a good measure
09:52:45the_KyleTrue that.
09:59:07wodz <- maybe some one will see what is the major difference in this sequences which render USB bridge malfunctioning in rockbox?
10:06:37 Join markun [0] (
10:06:37 Quit markun (Changing host)
10:06:37 Join markun [0] (~markun@rockbox/developer/markun)
10:09:53***Saving seen data "./dancer.seen"
10:10:00wodzcommenting out disk_unmount_all() doesn't solve the problem
10:15:38 Quit markun (Read error: Connection reset by peer)
10:15:57pamauryperhaps some power management, the ide_power_enable is only is the working sequence
10:16:19 Join efyx [0] (
10:17:34 Join eWill [0] (
10:18:56eWill[Saint]: I just failed again, but on a different file. I HOPE I figured it out −− I have a defragger running on auto. I'll shut it down and try again.
10:19:53 Quit Judas_PhD (Quit: This is a quitting message)
10:21:55[Saint]ah,'s quite possible it attempted to defrag something cygwin was using.
10:24:44 Quit factor (Read error: Connection reset by peer)
10:29:48 Join factor [0] (
10:30:49 Join MethoS- [0] (~clemens@
10:31:33 Quit eWill (Quit: ChatZilla 0.9.86 [Firefox 3.6.12/20101026210630])
10:31:55 Join markun [0] (
10:31:55 Quit markun (Changing host)
10:31:55 Join markun [0] (~markun@rockbox/developer/markun)
10:40:58 Quit markun (Read error: Connection reset by peer)
10:45:36 Join einhirn [0] (
10:58:20pixelmaAlexP: cr0fghfd
10:59:05pixelmaor any other forum admin
11:00:37 Join user890104 [0] (
11:03:39 Quit TheSeven (Ping timeout: 250 seconds)
11:18:53 Join Judas_PhD [0] (
11:18:55 Join dfkt [0] (dfkt@unaffiliated/dfkt)
11:20:33 Quit linuxstb (Ping timeout: 265 seconds)
11:26:52 Quit Judas_PhD (Quit: This is a quitting message)
11:30:57 Join markun [0] (~markun@rockbox/developer/markun)
11:32:47 Join linuxstb [0] (~linuxstb@rockbox/developer/linuxstb)
11:35:56 Join casainho [0] (
11:40:45 Join Rob2222 [0] (
11:47:57 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
11:49:32pixelmaAlexP: another spammer 9skalux50 :\
11:56:43Zagorscorche: if feasible, automatically adding rel="nofollow" to links in posts would make spamming less attractive
11:58:14 Join teru [0] (
11:58:17pixelmaZagor: wouldn't it make linking to the manual etc. harder too?
11:59:50pixelmaand while you are around, maybe you could take care of the above mentioned spammer? ;)
12:03:38 Quit TheSeven (Ping timeout: 276 seconds)
12:05:13 Quit user890104 ()
12:06:23amee2kis it bad if the make script throws a message " line 22: svnversion: command not found" ?
12:06:53 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
12:08:21 Quit markun (Read error: Connection reset by peer)
12:09:55***Saving seen data "./dancer.seen"
12:11:34wodzI am out of ideas why USB bridge works in bootloader and not in rb
12:13:56 Quit MethoS- (Remote host closed the connection)
12:14:38 Join InsDel [0] (
12:19:29 Join GeekShadow [0] (
12:19:30 Quit GeekShadow (Changing host)
12:19:30 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
12:24:29 Join markun [0] (
12:24:30 Quit markun (Changing host)
12:24:30 Join markun [0] (~markun@rockbox/developer/markun)
12:31:11CIA-7New commit by wodz (r28780): MPIO HD200, HD300: Make USB bridge handling more correct (this doesn't solve problems with USB inside rockbox on HD300 unfortunately).
12:33:25CIA-7r28780 build result: All green
12:40:14 Quit Rob2222 (Quit: Rob2222)
12:41:33 Quit wodz (Quit: Leaving)
12:43:06 Join eWill [0] (
12:44:02Zagorpixelma: nofollow just tells search engines to not follow or count the link. in the case of google, it means the spammer doesn't get pagerank credit for it.
12:44:13Zagorit's still a normal link for web browsers > eabi toolchain > Cygwin > 3600MHz dual-core AMD Athon II >>>> 2 hours 22 minutes. [Saint]
12:48:40 Quit GeekShadow (Read error: Connection reset by peer)
12:52:06 Quit InsDel (Ping timeout: 260 seconds)
12:54:16 Nick dionoea_ is now known as dionoea (
12:54:22 Quit dionoea (Changing host)
12:54:22 Join dionoea [0] (~dionoea@videolan/developer/dionoea)
12:59:11eWillbuilding with Cygwin −− ccache is installed. When I run make, does the 'CC' preceding each line mean ccache is running automatically? (I didn't precede make with 'ccache')
13:08:27 Quit markun (Read error: Connection reset by peer)
13:13:07 Join markun [0] (
13:13:07 Quit markun (Changing host)
13:13:07 Join markun [0] (~markun@rockbox/developer/markun)
13:16:33 Join b0hoon [0] (~quassel@
13:22:01 Quit markun (Read error: Connection reset by peer)
13:32:29 Join markun [0] (
13:32:29 Quit markun (Changing host)
13:32:29 Join markun [0] (~markun@rockbox/developer/markun)
13:34:23 Quit markun (Read error: Connection reset by peer)
13:35:46 Quit bluebrother (Disconnected by services)
13:35:47 Join bluebroth3r [0] (~dom@rockbox/developer/bluebrother)
13:44:23 Join simonrvn_ [0] (
13:47:02 Quit simonrvn (Ping timeout: 240 seconds)
13:47:02 Nick simonrvn_ is now known as simonrvn (
13:49:49 Join kugel [0] (~kugel@rockbox/developer/kugel)
13:50:19kugeleWill: yes
13:53:24 Quit fyrestorm (Quit: Ur skills' fireproof like a wooden panel -- U got feds talking leet on your IRC channel!)
13:55:23amee2khm. this is odd.
13:55:50amee2ki just installed debian on the laptop i want to mod into a jukebox, compiled the rockbox application port and ran it
13:56:28amee2kthe interface is rather slugish and the CPU idles at ~60..70% load
13:56:46amee2kbut htop for some reason doesn't show any process using it >_<
13:56:59Zagoramee2k: that's winter mode. designed to keep you warm :-)
13:57:29Tornewhat're you looking at in top, the CPU % of each process
13:57:34Tornelook at time instead
13:57:51Torneif something is using 40% of the cpu then somebody's scheduled time must be going up by four seconds in every ten
13:58:26Torneit's possible to entirely miss a process that's sleeping in short frequent bursts, if the sample rate lines up unfortunately ;)
13:59:28amee2kall the rockbox processes are listed as zero time
13:59:43amee2ki just started playing something and CPU load is constantly at 100% now
14:00:10*Torne wonders, out of curiosity, how loaded the machine is when you 8exit* rockbox :)
14:01:38 Join DerPapst [0] (
14:03:11Tornethis is a veyr slow machine, then :)
14:04:07Zagorstill, clearly the rockbox application should not spend time when idle
14:04:08Torneanyway, i don't really know a lot about the app port, but I think it's safe to assume that it's not *intended* to consume significant cpu time when doing nothing
14:04:12Torneassuming it really is
14:04:33Tornewe use SDL threading at the moment, right?
14:04:39Torneis this a pre-pentiumpro machine?
14:05:41amee2kP2 266
14:05:57amee2kcompaq armada 1700
14:06:15amee2kbut despite the internet insisting on 32M RAM it seems to have 64
14:06:48TornePentium II? that's new enough that everything under the hood should be working, at least
14:07:02Torne486/586 have some peculiarities that i'd rather not get into ;)
14:07:03amee2kthat what i thought too
14:07:11kugelTorne: yes, sdl threading, but I have a patch to replace it with *context() (as far as the rockbox threads go, we need at least 2 sdl threads still)
14:07:22Tornewell i don't mean in terms of performance, i mean in terms of SDL threading being reliable and sensible, and locking not being crazy broken
14:07:49amee2kkugel: does that improve performance?
14:08:18kugelit does reduce cpu usage on my machine
14:08:29amee2kcan i download it somewhere to try?
14:08:52amee2ki mean, worst case i still have the same problem
14:09:21Torneworth a try
14:09:35Tornebut it sounds like maybe we have some real issue there as well
14:09:47Tornewhich is maybe just not particularly visible on a faster machine
14:09:58***Saving seen data "./dancer.seen"
14:11:03Tornethe speed of the machine shouldn't make much difference when it's *idle* :)
14:11:23amee2kfrom the channel activity here at least i take it other than kugel and possibly me very little people are having a developing interrest in the app port
14:12:12Torneprobably not, no
14:12:18Tornewell, not the SDL app port
14:12:25TorneVarious people are interested in the Android app port
14:12:37Tornewhich does share code, but not everything
14:12:58Torneactually, at some point I will be interested in it
14:13:01amee2kto get a second data point i'm installing moc and mp3blaster to test playback performance
14:13:11Tornewhen my pandora finally arrives i'll probably want rockbox on it
14:13:18Torneand I'm pondering trying to port it both ways and compare
14:13:24Torne(have been meaning to do an OMAP port for a while)
14:13:41amee2kbut that has worked on a 166MHz Pentium (1) without significant impact
14:14:24amee2kwell, the vague idea i have in my head is an x86 port that uses an embedded linux system as bootloader and HAL
14:14:49Torneyes, that's what we discussed before
14:15:02Tornethe Pandora is an ARM-based handheld mostly intended for games; it runs linux
14:15:04amee2kwhich will probably double the specs, but if its for stationary use anyway then power is not an issue
14:15:18TorneI was going to port Rockbox to its hardware directly, but also get it running as a linux app with the SDL port
14:15:22Torneand see how it compares.
14:15:25Torneespecially for battery life
14:15:36Torneso, if i do that I will be interested in the SDL app port too.
14:15:40Tornebut right now I don't have the hardware yet
14:16:43kugelI'd be happy if someone would make a linux port not involving sdl, but I doubt it's worth it
14:17:05 Join robin0800 [0] (
14:17:12Torneyou mean drawing direct to the framebuffer and playing direct to /dev/dsp or similar?
14:17:36Tornedrawing and playback are probably not too bad for that; non-keyboard input is slightly trickier
14:18:02Tornethe benefit might indeed not be worth it, tho
14:19:48Tornei want to compare native vs linux on the pandora because someone once worked out from the hardware specs that it should be able to manage 100+ hours of battery life playing music (with screen off)
14:19:52kugelthe threading sdl implements is quite heavy (I think) because the semaphore locking leads to a huge number of syscalls (two per context switch)
14:20:03Tornebut currently it doesn't have anything like that using a regular linux mp3 player
14:20:16Tornethe battery on that thing is huge
14:21:25Torne4200mAh for a system whose SoC draws max 2W with every single peripheral active :)
14:28:25 Join mortalscan [0] (~mortalsca@
14:30:16 Quit teru (Quit: Quit)
14:31:23amee2khm... i seem to have some permissions issues left to work out, but this is a test system anyway
14:31:55 Quit antil33t (Read error: Connection reset by peer)
14:32:05 Join antil33t [0] (
14:32:06amee2kdecoding 320kbps, 44kHz 2ch MP3 results in a steady state CPU load of ~25% +/- 2
14:32:18amee2kso the proc should definitely be up to it
14:34:05amee2kokay, so to apply the patch i go to the top source directory ("rockbox") and run "patch </path/to/patch", right?
14:35:33kugelwith -p1
14:37:06amee2koh, which reminds me of an earlier question. do the docs have a vaguely complete list of build prerequisites somewhere?
14:37:29kugelfor the sdl app you only need a compiler, libc and libsdl
14:37:41kugel(plus the development headers of the libs)
14:38:23amee2ki'm not sure if subversion is a requirement too. when running make it repeatedly complained about not finding "svnversion" in tools/
14:39:30kugelthe version is not critical so you should be fine, assuming it doesn't cancel compilation
14:40:31amee2kno, it conginued running anyway
14:43:52amee2kokay, rebuilding it now
14:48:04 Join MethoS- [0] (~clemens@
14:54:28 Quit DerPapst (Read error: Connection timed out)
14:55:06kugelamee2k: still building? :)
14:55:44 Join casainho_ [0] (
14:56:43 Join DerPapst [0] (
15:01:46amee2kbut its already in apps/codecs
15:02:09amee2ki took the time to refine my serial terminal configuration
15:05:45 Join CaptainKewl [0] (
15:14:40amee2kkugel: that didn't really help much
15:15:08amee2know theres only 4 rockbox processes instead of 7 or 8
15:15:22amee2kbut idle load is still like 65%
15:15:58amee2kits also quite annoying that i can't find out whats sinking all this CPU time
15:16:07 Join komputes [0] (~komputes@ubuntu/member/komputes)
15:16:07kugelstrange, it went from ~10% to 4-5% on my laptop
15:16:26kugelwhat does top say? powertop output could also be interesting
15:17:22kugelamee2k: htop isn't showing processes but also user-space threads, which makes it hard to get a picture of the whole
15:17:37kugelrockbox is of course a single process
15:17:45amee2kinstalling powertop now
15:18:11kugel(you can also configure htop to not show threads seperately)
15:18:14 Join panni_ [0] (
15:18:16amee2kusing plain old top has always been a bit weee for me. hence why i prefer htop. it looks like norton commander >_>
15:18:40 Quit komputes (Remote host closed the connection)
15:18:41 Join hebz0rl [0] (
15:18:41amee2ki enabled the tree mode so it shows thread processes indented under processes
15:19:39amee2kthere also seem to be some odd issues with the mouse. when i move the cursor around, it sometimes "drops" a cursor symbol
15:19:49amee2klike its not properly ereasing the old one while redrawing
15:20:03amee2kbut the artifacts disappear after a second or two on their own
15:22:11amee2kkugel: << this is the powertop result
15:25:23 Join komputes [0] (~komputes@ubuntu/member/komputes)
15:25:27[Saint]eWill: Good news, and yes...if ccache is found it will be used automatically.
15:26:53amee2ki tried its suggestion, increasing the writeback timer, but not much difference if any
15:26:57amee2k 85.1% (1061.9) rockbox : do_nanosleep (hrtimer_wakeup)
15:27:40eWill[Saint]: Is there some folder I can backup now, so I don't have to rebuild the tool chain if I re-install?
15:28:40[Saint]you can just backup the entire C://CygWin folder
15:28:49kugelamee2k: that's 1061 wakeups per second, quite a lot. is that with my patch?
15:28:55[Saint]only info about setup.exe is stored in the registry.
15:28:58kugelI would be interested in the number without the patch
15:29:11amee2kkugel: yeah
15:29:14Torne[Saint]: the mount points are also stored in the registry
15:29:47amee2klet me try
15:29:49Tornethe toolchain is all installed to /usr/local so you can probably just back up that
15:29:54[Saint]I'm not sure that stil;l holds true.
15:30:07amee2kis there a quick way to revert the patch?
15:30:20kugelpatch -p1 -R < /path/to/patch
15:30:28amee2koh, cool :)
15:30:37Torne[Saint]: which? installs with −−prefix=/usr/local
15:30:45kugelalthough the 146% of the time in C2 state is very much questionable
15:30:48 Quit CaptainKewl (Ping timeout: 240 seconds)
15:30:53Torneso everything installed by the toolchain setup is under there.
15:31:01TorneYou can't *just* back up /usr/local/arm-eabi or similar
15:31:03[Saint]that mountpoints are stored in the registry.
15:31:10Torne[Saint]: i assure you they are
15:31:22amee2kkugel: in what way? what exactly does that number show?
15:31:28Torne[Saint]: i can see them right now :0
15:31:31Torneand they aren't stored anywhere else, that I know of
15:31:33amee2k(note: gtg in 5, bb in ~30)
15:31:54TorneHKLM\Software\Cygnus Solutions\Cygwin\mounts v2\*
15:32:18kugelamee2k: it shows how much % of the time the CPU was in which "sleep level"
15:32:25[Saint]Hmmm...I don't have that string.
15:32:31kugelC2 is "do nothing"
15:32:56amee2kum... how can it be doing nothing more than 100% of the time? its a single core system
15:33:07 Quit cjcopi (Ping timeout: 264 seconds)
15:33:21Torne[Saint]: Ther'es one under HKCU as well
15:34:06[Saint]Ah...right, there we go.
15:34:21 Join CaptainKewl [0] (~captainke@
15:34:32kugelamee2k: exactly :)
15:34:39Torneit *also* looks in /etc/fstab, it seems, but it doesn't store the default ones there.
15:34:49amee2kokay, compiling. i gtg pick something up now while make is running. bb in 30
15:35:13[Saint] & learn. Surely it's stored somewhere else, I've been moving the same CygWin install around for ages now.
15:35:25TorneIt infers where the root directory is using magic
15:35:41Torneso it doesn't actually need a mount for / to work
15:35:46Tornebut the others are not magic
15:35:54Torneso the binds for /usr/bin and /usr/lib break if you move it
15:36:27[Saint]yeah..that's why I'm thinking it must be stored somewhere else.
15:36:41TorneNo, it probably just regenerates them if you don't have a mount table in your registry
15:36:46Torneusing the same method it uses to decide what they were on install
15:36:57Torneso i fyo uhaven't actually *changed any mounts* it might work
15:37:02Tornebut if you did the ni twon't remember
15:38:00[Saint]that's probably how I've been getting away with it for so long then.
15:38:20[Saint]my setup rarely changes, even if it moves to different machiens.
15:41:46 Join Rob2222 [0] (
15:42:03 Quit antil33t (Read error: Connection reset by peer)
15:42:12 Join antil33t [0] (
15:43:43 Part b0hoon ("Back to work.")
15:45:13 Join Keripo [0] (
15:46:29eWillso.... can I just backup /usr/local?
15:47:17[Saint] theory.
15:47:52 Quit CaptainKewl (Ping timeout: 255 seconds)
15:48:04eWillI'm gonna just do the whole thing. Thanks for all the help lately.
15:54:48 Quit Keripo (Quit: Leaving.)
16:03:59 Quit robin0800 (Remote host closed the connection)
16:10:00***Saving seen data "./dancer.seen"
16:13:48 Join ReimuHakurei_ [0] (~reimu@
16:13:49 Quit ReimuHakurei (Read error: Connection reset by peer)
16:14:12 Join Keripo [0] (
16:16:28 Join jgarvey [0] (
16:18:44 Part Zagor
16:27:38 Quit kadoban (Remote host closed the connection)
16:32:36 Join n1s [0] (
16:32:37 Quit n1s (Changing host)
16:32:37 Join n1s [0] (~n1s@rockbox/developer/n1s)
16:33:21 Quit kugel (Remote host closed the connection)
16:41:53 Quit TheSeven (Ping timeout: 272 seconds)
16:48:37 Quit liar (Ping timeout: 255 seconds)
16:49:10 Join toffe82 [0] (
16:56:33 Quit evilnick_B (Quit: Page closed)
16:59:53 Quit panni_ (Read error: Connection reset by peer)
17:09:02 Join Strife89TX [0] (~cstrife89@
17:09:43 Join evilnick_B [0] (0c140464@rockbox/staff/evilnick)
17:12:00 Join Judas_PhD [0] (
17:16:59 Join stoffel [0] (
17:20:01 Join krabador [0] (
17:20:52 Quit evilnick_B (Quit: Page closed)
17:23:18 Join evilnick_B [0] (0c140464@rockbox/staff/evilnick)
17:23:28 Quit antil33t (Read error: Connection reset by peer)
17:23:38 Join antil33t [0] (
17:27:48 Quit pamaury (Read error: Operation timed out)
17:30:54 Quit evilnick_B (Ping timeout: 265 seconds)
17:32:19 Join evilnick_B [0] (0c140464@rockbox/staff/evilnick)
17:34:43 Join pamaury [0] (~quassel@
17:34:43 Quit pamaury (Changing host)
17:34:43 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
17:37:39 Quit MethoS- (Remote host closed the connection)
17:38:34 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
17:42:27 Join maltin [0] (~maltin@unaffiliated/maltin)
17:42:30maltinhi guys
17:43:36 Join kevku [0] (~kevku@2001:7d0:0:f000::135d)
17:43:59 Quit evilnick_B (Quit: Page closed)
17:48:56amee2kmmmh, damn
17:49:01amee2kthe kugel is gone :(
17:49:56 Quit casainho_ (Quit: ChatZilla 0.9.86 [Firefox 3.6.12/20101027124101])
17:53:03 Join evilnick_B [0] (0c140464@rockbox/staff/evilnick)
17:55:53saratogaAlexP: uggboots9
18:08:49 Quit eWill (Quit: ChatZilla 0.9.86 [Firefox 3.6.12/20101026210630])
18:10:02***Saving seen data "./dancer.seen"
18:11:51 Quit mystica555_ (Read error: Operation timed out)
18:19:55 Join domonoky [0] (~Domonoky@rockbox/developer/domonoky)
18:26:09 Quit Strife89TX (Quit: brb)
18:28:04 Join Strife89TX [0] (~cstrife89@
18:31:10 Join TheLemonMan [0] (
18:32:38 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
18:36:02 Join Maggux [0] (
18:42:20 Join kugel [0] (~kugel@
18:42:22 Quit kugel (Changing host)
18:42:22 Join kugel [0] (~kugel@rockbox/developer/kugel)
18:42:46kugelamee2k: pong
18:43:37 Quit Maggux (Remote host closed the connection)
18:43:46amee2khey kugel :)
18:44:08amee2kgot a bunch of pastes for you
18:44:41amee2k << with your patch, while idling in the main menu. i re-ran it another time to be sure that the 280% shit isn't a false positive
18:45:18amee2k << without your patch (un-patched with the command you recommended and built in a new build directory), idling in the main menu
18:45:33amee2k << same unpatched build, trying to play something back
18:46:32gevaertsamee2k: I'd try a newer powertop. Those numbers don't make sense...
18:46:57amee2kas a side note, after a while powertop suggested that i enable the HPET in my bios and load appropriate kernel extensions
18:47:18amee2kwhich is entirely academic since the system is certainly too old to have a HPET
18:47:30amee2kgevaerts: how much newer?
18:48:01gevaertswell, "latest" seems appropriate
18:48:39amee2kwhat i'm using as dev system is debian stable, 2.6.26 and fully updated
18:49:24gevaertsWell, that's 1.10. 1.13 is available
18:49:46amee2kits debian stable. that means whatever it is, its 2 years old ;)
18:50:02 Join domonoky1 [0] (
18:50:08gevaertsAnd any tool that claims 42610 wakeups per second, 99% of which are the 1824 wakeups caused by rockbox, is *broken*
18:50:18 Quit domonoky (Ping timeout: 245 seconds)
18:51:03gevaertsIf you want to actually use those numbers, feel free, but I'd really try powertop 1.13 first
18:51:26amee2kcompiling it right now
18:53:00 Quit krabador (Ping timeout: 272 seconds)
18:53:09kugelI was about to believe that unpatched causes 4x more wakeups, but not anymore when the same built causes 10x less with music playing
18:54:01amee2ki think the 10x less is only because the processor is overloaded
18:54:17kugelupgrading to squeeze might be worth it. they in freeze so it should be about as stable as lenny
18:54:18amee2kits ramping straight into 100% and stays there when i hit play
18:55:34kugelhmm, you have a point there. I guess it does less wakeups because the CPU can't sleep (?)
18:56:32amee2kor its simply stalling. i can't hear anything and the playback is maybe 20..30% realtime at best
18:56:59amee2kmocp plays fine so alsa is working correctly
18:57:40kugelperhaps the GUI is too demanding
18:57:41amee2kalso, not sure if the lack of a HPET forces some awkwardly inefficient workaround to be employed
18:57:55amee2khmmm maybe
18:58:15amee2kbut that shouldn't explain the excess idle load, unless it is forcing needless redraws
18:58:35kugelonly 1 per second
18:59:14n1sdoes the sdl app have core_sleep ?
18:59:15amee2kis there a way to gauge the load with subsecond precision?
18:59:54amee2ktop and other tools are integrating over at least a second or so to to assess cpu load
19:00:11amee2kthat may be throwing it off, but it would also mean that the redraw is REALLY inefficient
19:01:01kugelsomething is definitely wrong with 60%+ idle load
19:01:51 Join JesusFreak316 [0] (
19:01:52amee2ki'm wondering what though
19:02:04amee2kif its just a scheduler bug, or the graphics
19:02:26kugelyou can try commenting the redraw code out (lcd_update() and _rect()) but It's
19:02:39kugelprobably not the problem
19:02:58amee2khm... i'll try the newer powertop first. maybe that clears some things up
19:06:08kugelamee2k: is the load caused by a single thread? (htop shows this)
19:06:58amee2khtop oddly reports 0% CPU and zero TIME+ for all rockbox processes
19:07:31amee2keven though a good 50% of the load are user processes, and around another 15% are system load
19:07:44kugelmake sure you sort by CPU usage
19:08:04amee2ki've sorted by command name so all the rockbox processes end up together
19:08:35amee2kbut resorting by cpu% doesn't show any perpetrators either
19:08:57kugelso what pegs the CPU according to htop?
19:09:04amee2knothing :P
19:09:14amee2k << current powertop, still the same unpatched build, idling in main menu
19:09:30amee2kkugel: at first i thought htop bug, but regular top shows the same picture
19:10:10amee2k40..50%us, ~15%sy, but no processes creating the load
19:11:43amee2kthe suggestions in the new powertop are entirely useless. only tells me to enable various kernel options to allow power-managing devices i don't have anyway >_<
19:12:47amee2kwant me to recompile with your patch on again?
19:13:07n1sfor me the rockbox app causes 50-60% of the wakeups when idling (in do_nanosleep)
19:13:47amee2kwell, the percentage won't be as important. this box isn't running anything else
19:14:11 Join Horscht [0] (
19:14:11 Quit Horscht (Changing host)
19:14:11 Join Horscht [0] (~Horschti@xbmc/user/horscht)
19:14:17amee2kno desktop environment and fancy devices to create lots of wakeups
19:15:19n1sof course, i just wanted to point out that it causes a lot of wakeups here too and that powertop identifies the "do_nanosleep" function as the cause (which i assume is an sdl thing)
19:15:42amee2kah, okay. i misunderstood then :)
19:16:31amee2kwhat kind of system are you running it on, if i may ask?
19:16:47amee2kespecially, is your system using a HPET?
19:16:50n1sdesktop core i5
19:17:05n1sshould be
19:17:31n1syes, i have a number of hpet lines in powertop
19:17:48amee2kthe 1.10 powertop has recommended i enable it in the bios/kernel but i can't test it due to lack of the actual hardware
19:18:11 Quit Horscht (Client Quit)
19:18:16kugelamee2k: gtg sorry. that needs investigation, but I tend to assume that sdl is the problem.
19:18:25 Quit kugel (Quit: Yaaic - Yet another Android IRC client -
19:19:46 Quit Judas_PhD (Quit: This is a quitting message)
19:20:32 Join Horscht [0] (
19:20:32 Quit Horscht (Changing host)
19:20:32 Join Horscht [0] (~Horschti@xbmc/user/horscht)
19:21:31*gevaerts no blames amee2k's kernel
19:22:15 Quit maltin (Read error: Operation timed out)
19:22:52amee2know* ?
19:23:03saratogaAlexP: forums getting owned
19:23:09amee2ktoo old or too new?
19:23:37gevaertsyes, now :)
19:24:06 Join Judas_PhD [0] (
19:24:18amee2k# uname -a
19:24:19amee2kLinux rocklap 2.6.26-2-686 #1 SMP Thu Nov 25 01:53:57 UTC 2010 i686 GNU/Linux
19:24:57gevaertsNo idea, but if most of the numbers you get are meaningless, regardless of the tool used, there must be a reason
19:25:16 Join maltin [0] (
19:25:16amee2kyeah, there metter be one >_>
19:26:29gevaertsPossibly some linux-related channel can provide more help though
19:26:46n1sour sdl thread code uses SDL_SemWaitTimeout about which the sdl docs say "This function is to be avoided if possible since on some platforms it is implemented by polling the semaphore every millisecond in a busy loop."
19:26:57n1sso this sounds like a likely candidate to me
19:27:29amee2k1k wakeups per second = ~1ms interval
19:27:45gevaertsBut didn't you see the same with that patch?
19:28:03amee2kyeah, i did
19:28:12n1soh, i missed the patch
19:28:45amee2klet me reapply the patch and try again with the new htop version
19:29:09 Quit einhirn (Quit: Miranda IM! Smaller, Faster, Easier.
19:29:14amee2kerr... powertop
19:29:21n1swhere is this patch?
19:30:38amee2kapplied with cd rockbox; patch -p1 </path/to/patch
19:31:21alexbobPsomebody should make a rockbox plugin that lets you charge the player by leaving it in the sun
19:31:43amee2kUSB solar charger
19:31:59alexbobPbut buying new hardware costs money!
19:32:17amee2khm... that would be a cool hardware mod for my ipod... solar cells on the back
19:32:30amee2kbut not much room left in the case to fit a boost converter :/
19:34:23amee2kdoing make fullinstall will overwrite the entire previous installation, right?
19:35:46n1si get much fewer wakeups with that patch
19:36:31n1sabout 100 now, which seems right
19:36:54n1scpu usage dropped from 3% to 2% too
19:37:38 Quit stoffel (Remote host closed the connection)
19:39:29 Quit jgarvey (Ping timeout: 240 seconds)
19:53:04 Join jgarvey [0] (
19:55:33 Quit Strife89TX (Quit: Leaving)
20:02:03 Nick YPSY is now known as Ypsy (
20:10:05***Saving seen data "./dancer.seen"
20:17:46pixelmaoops, sorry
20:29:12 Quit Judas_PhD (Quit: This is a quitting message)
20:30:42 Quit maltin (Quit: Lost terminal)
20:35:53 Quit factor (Ping timeout: 255 seconds)
20:45:12 Join mystica555_ [0] (
20:46:06 Join factor [0] (
20:47:56 Join slooopy [0] (
20:51:28 Quit mystica555_ (Ping timeout: 276 seconds)
20:51:58*scorche installed akismet for the forums
20:55:22 Quit n1s (Quit: Lämnar)
20:58:18 Quit evilnick_B (Quit: Page closed)
20:58:21 Join InsDel [0] (
20:59:11 Join Kupop [0] (
21:00:04amee2kthis day somehow is going out of its way to not let me get anything done >_<
21:09:38amee2k << freshly patched and rebuilt, updated powertop version, idling in main menu
21:10:00amee2kthe 200+% idle time thing is back, apparently
21:10:06amee2kthis time, even more gross than before
21:20:26the_KyleDo the recording gain settings not work on clip+? All settings sound the same. It's really quiet.
21:22:20 Join ej0rge [0] (
21:22:51ej0rgehey. I have a weird issue with my beast.
21:23:20ej0rgeS60 upgraded to 120gb, 111gb of music on it, if i attempt to build the database it gets to 1103 and then hangs.
21:23:40ej0rgeI presume there is a corrupt file or something, but i'm at a loss with regard to how to identify it
21:24:21ej0rgeRan checkdisk in winxp, it cranked for 3 hours, got to 100%, told me it was unable to complete
21:24:26ej0rgeno difference in behavior
21:25:02ej0rgeis there a way i can find out the last file added to the database before it got stuck?
21:26:27ej0rgethe harddrive sounds like it's performing the same seek and reset over and over, so it may even be a bad sector. At this point i think that short of identifying the file and adding it to the ignore list, i may have to find a way to hook the drive up directly to a PC and run better diagnostics on it than i can do at the other end of a USB cable
21:27:12saratogaej0rge: check this out:
21:27:40saratogathat should tell you which file it dies on
21:27:55ej0rgeyeah, that looks like it should help a lot. Thanks
21:28:05ej0rgefwiw you may know me as ericj, but someone else is using that as a nick these days
21:28:35ej0rgejob/life kept asserting too many non-maskable interrupts so i stopped playing with DAPs so much.
21:30:33pixelmaif you think it's a bad sector, I guess running a file system check first would be good
21:31:29ej0rgepixelma: yeah, but windows scandisk utterly failed to do it. I suppose i could boot up my laptop and see if dosfsck can do any good over usb. (at work, win-centric job)
21:32:13saratogathen its probably either a bad file or a broken disk
21:32:55ej0rgebad file wouldn't surprise me in the least. I need to go through my music archives with mediamonkey or something. Round tuit not yet acquired.
21:33:18saratogaif you can find a file that crashes our database please file a bug report with a link to it so that we can fix the bug
21:33:59ej0rgewill do
21:35:09 Join wodz [0] (
21:35:20CIA-7New commit by wodz (r28781): fix rbutil.ini entries for MPIOs as pointed out by Michael Leslie
21:37:06CIA-7r28781 build result: All green
21:37:55amee2ki'm so tempted to wipe this dev box and install testing
21:38:52 Quit pamaury (Remote host closed the connection)
21:42:21wodzI am rather surprised how small feedback gives my mails to ML about i2c unifiaction and doxygen.
21:51:22 Quit Battousai (Read error: Operation timed out)
21:51:49the_KyleI figured out the recording gain. Looks like is just more fine-grained than I expected. Works really nicely now.
21:53:21 Join Battousai [0] (~bryan@gentoo/developer/battousai)
21:54:11 Join pamaury [0] (
21:54:12 Quit pamaury (Changing host)
21:54:12 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
22:09:05bertrikthe_Kyle, I vaguely remember that we used to have a problem with recording gain, there there was always some kind of auto-gain control active. Not sure if that was ever fixed.
22:10:06***Saving seen data "./dancer.seen"
22:10:27TheLemonManfreesale kindly rejected my request about bootloader usb protocol infos
22:10:33 Join fml [0] (
22:11:48the_Kylebertrik: Seems that still happens. I'm still testing things. That could get annoying though. I noticed there's no configuration setting for the auto gain control.
22:12:11 Quit Keripo (Read error: Connection reset by peer)
22:12:15fmlwodz: I think no reply means no opinion, or, positively formulated, "I don't care". For example, a very well written mail from sideral about a feature that would be assumably used by some got no replies. Nobody cares.
22:12:23 Quit factor (Ping timeout: 245 seconds)
22:13:19 Join factor [0] (
22:16:18 Join liar [0] (
22:17:43 Join Keripo [0] (
22:19:48bertrikwodz, looks like a nice prototype, but there are some weird i2c devices out there (that want a repeated start for example) and I don't know for sure that rockbox only used "normal" i2c devices.
22:20:52bertrikhm, or maybe I should only worry about that when we actually encounter something weird and think of a solution then
22:22:53bertrikthe_Kyle, it's a hardware thing AFAIK
22:23:30 Quit bertrik (Quit: :tiuQ)
22:23:56the_KyleWell, not much can be done about that I guess but to work around it when recording.
22:24:12 Quit wodz (Quit: Leaving)
22:24:45fmlI remember some i2c related errors were ironed out by try-and-error
22:24:53 Quit slooopy (Ping timeout: 260 seconds)
22:36:33 Quit fml (Quit: ChatZilla 0.9.86 [Firefox 3.6.12/20101026210630])
22:37:55 Join slooopy [0] (
22:39:36 Nick Ypsy is now known as YPSY (
22:50:20ej0rgean hour later, dosfsck is all the way to /Da*
22:50:28ej0rgethis is looking like something i should leave running overnight
22:50:46ej0rgeit has found some issues though
22:57:20 Quit TheLemonMan (Quit: free(me))
22:57:21 Join MethoS- [0] (~clemens@
22:57:57 Join ze_ [0] (~ze@
22:59:30 Quit Keripo (Read error: Connection reset by peer)
23:00:20 Join evilnick_B [0] (0c140464@rockbox/staff/evilnick)
23:06:44 Quit slooopy (Ping timeout: 264 seconds)
23:10:49 Quit hebz0rl (Quit: Leaving)
23:11:37 Quit kevku (Quit: KVIrc 4.0.2 Insomnia
23:13:44 Join JdGord [0] (~jd@
23:15:03JdGordA DB auto-update should cause playback to get sstarved right?
23:19:58 Join slooopy [0] (
23:20:58 Quit domonoky1 (Read error: Connection reset by peer)
23:26:15 Quit casainho (Read error: Connection timed out)
23:27:11 Join casainho [0] (
23:28:21saratogashouldn't i think
23:30:34 Join grndslm [0] (
23:30:52 Quit komputes (Remote host closed the connection)
23:33:08JdGordMaybe my disk is crap then. I disabalsd autoupdate and its all good after rebooting
23:33:19JdGordBefore buffering was conpletly starved
23:36:56 Join kugel [0] (~kugel@rockbox/developer/kugel)
23:37:18kugelamee2k: it's sad that none of these tools give helpful numbers
23:38:17kugelpowertop is useless if it reports 50k wakeups (with rockbox causing 99.5% of them with 1k wakeups) and being 430% of the time in C2 mode
23:38:53kugel(h)top is useless if they can't tell what's causing the cpu load but still do report high load
23:40:17kugelyou could do 2 things: a) upgrade to squeeze or even unstable and hope that the bugs are fixed, or b) downgrade a lot until yuo reach the point where P2 was reasonably supported
23:41:15kugelat least n1s confirms what I expected to see from my patch
23:43:46 Quit slooopy (Ping timeout: 250 seconds)
23:50:36amee2kive already downloaded the squeeze netinst
23:51:01amee2kwhile you were afk i've tried your patch + the new powertop version again
23:51:42amee2kkugel: how the fsck did you get that message after you left?!
23:52:03kugelwe have irc logs
23:52:14amee2k(in entorely unrelated news, my flatmate has a damn fine selection fo whiskys in his stask)
23:52:34kugel :)
23:52:59kugelthe topic also mentions it
23:53:49amee2kright. i susually read the topic on join, and if it conetains no relevant info to my cause i forget about it >_>
23:54:01kugelamee2k: I think I need to meed your flatmate, but that's stuff for #rockbox-community
23:54:16amee2kwow, i'm drunk i almost threw my terminal off the desk >_>
23:54:26amee2klol okay
23:54:40 Join JdGordon| [0] (~jonno@rockbox/developer/JdGordon)
23:54:58amee2kjust don't be as tight assed as abrotman about it please
23:55:15amee2ki don't want to put you on ignore ;)
23:55:48 Join slooopy [0] (
23:55:48amee2koh, thats a dude from #debian in case you don't know him
23:56:52kugelno worries, but I suggest you to put #rockbox-community onto your favorites, then you can talk about whatever you want while being drunk (or not). this channel is for development (hence it's logged). -community is for chit chat of all sorts
23:57:14amee2khehe, okay
23:58:47amee2kanyway, i've downloaded the testing netinst. since i'm onyly geting bullshit from stable and gevaerts (i think? cbf to check the log...) suggested its a kernel issue

Previous day | Next day