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 2013-11-16

00:01:20pamauryhum, there is definitely something fishy, maybe mobile ddr is not set to low-power mode by the bootloader ? Not sure if that could explain such a difference
00:01:44pamauryor maybe it is related to audio settings
00:02:49pamauryI need to plug hwstub into this OF to see
00:08:30 Join mudlord [0] (
00:08:51pamauryI guess we could even lower the power but I'm not sure we could decode mp3 with this
00:09:46[Saint]There's clearly something awry if the OF has the potential to beat us by ~75%
00:10:02[Saint]...what, however, is anyone's guess.
00:10:43[Saint]Of course there's always the possibility that the OF's battery bar is *really* poorly calibrated.
00:10:46pamaurywhat is even stranger is that we beat the Fuze+ OF by large with the same code
00:11:25pamaurythat's why i'm thinking about some ram related setting maybe
00:12:01[Saint]you think some chip not running in a low power state could eat *so* much power?
00:12:07 Quit Narod ()
00:12:26[Saint]Currently, assuming the F's battery bar is anywhere near accurate - that's a massive difference.
00:13:32mudlordHi I have a general question regarding embedded C since I want to tinker with rockbox on my Sansa Clip Zip which is being shipped.
00:13:40mudlordI already have experience in C
00:13:49mudlordI am wondering is things like malloc() allowed?
00:14:31lebelliumI don't believe the battery indicator is accurate but I expect it to last at least 20hrs like pamaury's E370
00:14:33mudlordthank you
00:14:47pamaurydammit, [Saint] was faster
00:15:09mudlordso strcmp and things are still allowed?
00:16:03pamaurywe do have limited forms of memory allocation but they have great restrictions due to the lack of virtual memory.
00:16:03pamauryYes of course
00:16:21mudlordOkay, so mainly the limits are for memory size, etc
00:17:15[Saint]Some targets have ~2MB or less to play with.
00:21:13[Saint]If you're only counting "real DAPs", then I think targets can have anywhere between 1 and 64MB of RAM available.
00:23:04mudlordmy intended target for the tinkering is the Sansa Clip Zip, 8GB model
00:23:19[Saint]A fine wee beast.
00:23:20mudlordnot sure on the amount of ram it has
00:23:44[Saint]8MB for th
00:23:49[Saint] beat me this time.
00:25:22[Saint]For such tiny wee things, the Clip(s) are all really nice targets, really.
00:25:58[Saint]I prefer the big, actual spinning platter disc, brick-type players myself - but there's nothing wrong with the wee flash guys.
00:26:18pamauryyeah only USB is problematic with the clips
00:26:50[Saint]Do you think its worth returning to booting the OF on connect?
00:27:16[Saint]It might piss a few people off, but, it would alter the "the crap doesn't work" perception that seems to be building.
00:27:47mudlordI guess then I will have plenty of questions then when my shipment arrives
00:28:01mudlordatm I downloaded vbox and the rockbox virtualbox package
00:28:03[Saint]looking forward to it. :)
00:28:33[Saint]Making sure you can compile the project is definitely a great first step.
00:28:46[Saint]Jump around the source a bit, have a look at how its structured.
00:29:24pamaury[Saint]: no opinion on this, but that would definitely spare us the "usb doesn't work" and us responding "yes we know, please read the manual/wiki before talking"
00:29:47[Saint]pamaury: indeed.
00:30:06[Saint]Frankly, it looks bad.
00:30:23[Saint]"USB is broken", "yes, we know", "...but, it says stable"
00:30:45gevaertsThe problem is having to release different bootloaders for this
00:31:01[Saint]Ah yes.
00:31:16[Saint]I think its worth considering, nonetheless.
00:31:43[Saint]But I have no desire to get into another quasi-bricking bootloader experience either.
00:34:00[Saint]SOmeone wanted to do a release soonish - perhaps that is the time to do so?
00:34:42gevaertsThe tricky thing to watch out for is a setup where rockbox reboots for USB and the bootloader thinks rockbox wants to handle it
00:35:18[Saint]Man, that was ugly.
00:37:08gevaertsThe other direction just means that booting with USB goes to the OF, which may be a bit confusing but without actually sending people into a panic
00:37:38[Saint]I know my opinion alone means squat - and this needs some debate - but I think we passed the point a while back where this fails for more people than it works for.
00:38:12[Saint] /something/ needs to be done, and I know some very smart people have looked at fixing USB - if that is currently not an option, we need to consider going back to the OF>
00:38:35[Saint]I'm more thinking about potential users than existing users that have already been forced to figure this out.
00:39:28[Saint]I don;t think we can call these targets "stable" in good faith knowing that USB almost certainly won't work.
00:39:48gevaertsI don't disagree
00:40:15gevaertsOh, how does rbutil handle these bootloaders? Don't we also need a properly coordinated release of that?
00:40:50[Saint]Yes, I believe so.
00:41:14[Saint]This is definitely something that needs a few points of coordination should it happen.
00:41:33funmanrockbox USB is definitely better than OF USB
00:41:55[Saint]funman: IFF it works...which is basically never.
00:42:05funmanworks fine for me
00:42:26lebelliumif it only works fine for you, I don't see how you could claim it is better then OF USB :)
00:42:27[Saint]Right, but the people whom it works for are a tiny percentile.
00:42:40[Saint] /tiny/
00:42:49funman[Saint]: do you have numbers?
00:42:55*gevaerts points out that we don't have any idea about the numbers
00:43:13funmanah well, i can't test on my broken clip+ obviously :/
00:43:28gevaertsI know we have the STALL recovery issue
00:44:03gevaertsAnd that one is going to show up for first time users on windows
00:44:41gevaertsI don't know what happens after that point, but I think it's not unlikely to be a one time issue
00:44:57[Saint]Ok - yes, I'll admit. There's no real data I have on numbers of success vs. fail - that was improper. Based purely on support requests alone there's no denying that there's a problem, though.
00:45:12[Saint]And shrugging and saying "works for me" isn;t really an option.
00:45:36[Saint]Well - of course its an option - but its a crap one.
00:45:54funman[Saint]: I am giving results, not options
00:46:20 Quit Guest17648 (Quit: Hmmm...)
00:46:42[Saint]I realize that but saying it works for you is rather meaningless when we know it doesn't work for so many others.
00:46:44funmanI obviously know USB has a problem but it works most of the time for me (1 plug or unplug out of 10 will give me hard lock up)
00:47:04funman[Saint]: how many
00:47:18pamauryif like gevaerts suggest this is all (or even only mostly) related to the STALL recovery issue (and so to MTP detection on first connect on Windows), then there is a trivial solution to the problem
00:47:37funmani doubt it is a Windows problem, I have the same results on Linux / Windows and OSX
00:47:49gevaertsfunman: this particular one is
00:47:50funmanthere could be more than one issue though
00:48:01gevaertspamaury: provide an OS descriptor?
00:48:12pamauryyes, or provide "something"
00:48:31pamaurywe are not infrindging anyone IP by sending an empty string for this particular value
00:48:41gevaertsAh, indeed
00:48:49gevaertsWell, it better be a valid string then :)
00:49:34gevaertsfunman: if you run the python script I attached to FS #12910, I promise you it will kill the USB connection
00:49:35pamauryyes probably :) Not sure how Windows will interpret it though, needs some experiment.
00:49:35fs-bluebot Sansa zip clip fails to enumerate over USB under Windows. (bugs, unconfirmed)
00:49:46gevaertsAnd that's what windows does on first connect with new devices
00:49:59funmangevaerts: will look, updating to latest git first
00:50:04pamauryBut I don't see any disadvantage here: we know for sure that if you plug it under Windows, it will 99% crash the usb core because of this bug
00:51:05pamaurygevaerts: did you try with the OF ? Ask for some other invalid string ? Does it recover from STALL ?
00:51:06gevaertspamaury: do you know if windows rememembers not to try again if it crashed the device? IIRC I read somewhere they do remember not to ask again if you don't have the string
00:51:25[Saint]Bah - there's no non-trivial way to scrape the forums for specific complaints.
00:51:42pamaurygevaerts: I'm mixed on this, I personally experienced cases where Windows would never remember
00:52:01gevaertspamaury: I tried the OF with 0xEE, yes, and it works. My e200v1 also works with rockbox, so it's not a general issue
00:52:19gevaerts[Saint]: thing is, you won't find any numbers for those where it works
00:52:32pamauryBecause I still have my work in progress disassembly of the Clip+ (or V2) bootrom, I went pretty far in USB
00:52:39gevaertspamaury: I didn't try any other string
00:52:42pamauryit must know how to recover
00:54:07funman make zip
00:54:07funmancan't open database.ignore at /media/dev/rockbox/tools/ line 374.
00:54:09funmanhmm :/
00:56:51 Quit ender` (Quit: There are only two hard things in Computer Science: cache invalidation and naming things.)
00:57:07pamaurymemory settings are fine on the nwz370 so the battery life difference doesn't come from here
00:57:40funmanmy clip+ died again
00:59:47funmangevaerts: AttributeError: 'module' object has no attribute 'busses'
01:00:04gevaertsfunman: bah. pyusb 0.4 vs 1.0 I bet
01:00:37gevaertsI'm on 0.4, and the thing is that I know nothing about it, really...
01:00:49funmanhmm it wasn't installed at all
01:00:59funmanI should not have saved this script as
01:01:16funmandescriptor: (4, 3, 9, 4)
01:01:17funman0xEE failed as expected
01:01:35funmanfuzev2 is not responding now
01:01:59gevaertsWhat it does is get a normal string descriptor, then 0xEE which should fail, then a normal again which should work
01:03:40funmanapparently Ubuntu Saucy's perl doesn't like
01:13:50 Quit bertrik (Ping timeout: 260 seconds)
01:17:31funmangevaerts: the fuzev2 still works fine on unplug though
01:17:42funmanno hard lock up
01:18:05gevaertsI've had it both lock up and work
01:18:12gevaertsWell, not at the same time :)
01:18:25gevaertsBut the lock up could be the "other" issue
01:24:10funman[Saint]: did you mention a jtagged amsv2 the other day?
01:35:17***Saving seen data "./dancer.seen"
01:44:08 Quit krabador (Quit: Sto andando via)
02:02:34 Quit lebellium (Quit: ChatZilla [Firefox 26.0/20131111154639])
02:10:32 Quit mudlord (Disconnected by services)
02:10:38 Join mudlord [0] (
02:45:03 Quit foolsh (Quit: foolsh)
03:19:58 Quit mudlord (Disconnected by services)
03:20:03 Join mudlord [0] (
03:35:18***Saving seen data "./dancer.seen"
03:40:38[Saint]funman: TheSeven was trying to get rid of one a while ago.
03:41:06[Saint]Prod him I guess, not sure if it found a home yet.
03:41:13 Part mudlord
03:45:37TheSevenit's still sitting on my desk, with apparently broken internal flash
03:58:00 Quit [Saint] (Quit: Quit.)
04:39:08 Quit pixelma (Disconnected by services)
04:39:08 Join pixelma_ [0] (pixelma@rockbox/staff/pixelma)
04:39:09 Join amiconn_ [0] (amiconn@rockbox/developer/amiconn)
04:39:09 Quit amiconn (Disconnected by services)
04:39:10 Nick pixelma_ is now known as pixelma (pixelma@rockbox/staff/pixelma)
04:39:13 Nick amiconn_ is now known as amiconn (amiconn@rockbox/developer/amiconn)
05:06:58 Join Strife89 [0] (
05:14:09 Quit TheSeven (Disconnected by services)
05:14:22 Join [7] [0] (~quassel@rockbox/developer/TheSeven)
05:35:20***Saving seen data "./dancer.seen"
06:32:36 Join DexterLB [0] (
06:32:59 Quit dfkt_ (Quit: -= SysReset 2.55=- Sic gorgiamus allos subjectatos nunc.)
07:35:21***Saving seen data "./dancer.seen"
07:57:41 Quit nosa-j (Ping timeout: 248 seconds)
07:58:49 Quit amiconn (Remote host closed the connection)
07:58:49 Quit pixelma (Remote host closed the connection)
07:59:42 Join pixelma [0] (pixelma@rockbox/staff/pixelma)
07:59:43 Join amiconn [0] (amiconn@rockbox/developer/amiconn)
08:04:03lonoxmontandypotter: i think either i did something wrong or it didnt compile right or something, because there werent any commands visible in the 'view log file' in the debug menu
08:04:05 Join nosa-j [0] (~m00k@
08:04:17lonoxmontand there wasnt a log file in the .rockbox folder either
08:04:25lonoxmontit may be iap over usb after all :/
08:08:53lonoxmontrather not reencode all my shit to the various apple formats and use the fisher-price firmware, but may have to do so :/
08:10:02lonoxmontthough i suspect it may also somewhat come down to the cable, theres two versions of it for my car, one that supports tagging tracks off of the radio in itunes and one that does not
08:10:19lonoxmontmaybe the tagging one is iap over usb while the other one isnt
08:17:34 Join kevku [0] (~kevku@2001:470:27:773:0:feed:c0f:fee)
08:18:13 Join n1s [0] (~n1s@rockbox/developer/n1s)
08:20:10 Join pretty_function [0] (~sigBART@
08:23:40 Quit pretty_function (Client Quit)
08:32:21 Quit nosa-j (Ping timeout: 248 seconds)
08:36:38 Join nosa-j [0] (~m00k@
08:59:08 Quit ender| (Ping timeout: 256 seconds)
09:00:08 Join ender| [0] (krneki@2a01:260:4094:1:42:42:42:42)
09:08:02 Quit nosa-j (Ping timeout: 264 seconds)
09:14:35 Nick SuperBrainAk is now known as DormantBrain (~andy@
09:22:46 Join nosa-j [0] (~m00k@
09:31:44 Join lorenzo92 [0] (
09:35:22***Saving seen data "./dancer.seen"
09:38:33 Quit nosa-j (Ping timeout: 272 seconds)
09:51:12lorenzo92pamaury: ping
09:57:45 Join nosa-j [0] (~m00k@
10:07:11 Quit nosa-j (Ping timeout: 240 seconds)
10:12:05 Join [Saint] [0] (~saint@rockbox/user/saint)
10:15:45 Join nosa-j [0] (~m00k@
10:16:34lorenzo92pamaury: now z5 booted again, but removing the skin loading stuff line (very last one in main), memory corruption otherwise
10:23:36 Quit nosa-j (Ping timeout: 272 seconds)
10:25:46 Join nosa-j [0] (~m00k@
10:52:03[Saint]The sustained transfer speed on the CEATA CLassic is awful.
10:52:56[Saint]Filling the disk takes approximately seventeen trillion years.
10:53:24copperI think you're exagerating.
10:53:51 Join lebellium [0] (
10:54:01[Saint]Just a tiny bit.
10:54:07copperEver so slightly.
10:55:38[Saint]But, seriously though - it takes like 8 hours.
10:55:52[Saint]WHich is nuts. Luckily I do it overnight.
10:55:54 Quit pamaury (Ping timeout: 272 seconds)
10:55:55copperAlso, I think that USB / database fuck ups tend to bugger the Classic and make it run slow
10:56:21 Quit nosa-j (Ping timeout: 248 seconds)
10:56:28copperevery time I wipe it clean and reinstall Rockbox, copy my files and build the database, it runs much smoother and faster
10:56:59copperI have no idea how to reliably reproduce it and / or debug it
10:57:31coppereven with my large lossyFLACs
10:57:45copper(400+ kbps)
10:59:02[Saint]I can't say I have noticed that myself.
10:59:27copperi think maybe one reason is filesystem problems due to Rockbox crashing while doing USB transfers… then the database tries to auto-update with fucked up data, or something, and possibly, I turn off the DAP before giving it a chance to let it finish
10:59:28[Saint]USing the device is fine, its just that USB transfer for bulk transfers is really, really, really slow.
11:00:13[Saint]The Database should, in theory, be able to recover itself.
11:00:23copperI went into the debug menu last time before wiping the HDD clean, to see how fast the database was updating: it was going at like one song per second
11:00:42copperafter the wipe, it went a hell of a lot faster (many songs per second)
11:01:05 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
11:01:09[Saint]that can be a tricky-trickson, that screen.
11:01:22[Saint]User input can affect the update rate of the information displayed.
11:01:44copperI wasn't doing anything while watching progress
11:01:52coppernot touching anything
11:02:33[Saint]I can't think of a reason for that.
11:02:35copperbesides, overall time of completion (i.e. when the HDD activity indicator stopped) was consistent with that assessment
11:02:49 Join nosa-j [0] (~m00k@
11:05:22copper[Saint]: about your speed problem: does changing block size when writing with dd improve speed?
11:05:46n1slistening to music while it updates will make it slower too
11:06:35n1salso if the filesystem is corrupten, maybe a stab at it with fsck.vfat will help
11:09:59 Quit nosa-j (Ping timeout: 252 seconds)
11:11:47 Join kugel [0] (~kugel@rockbox/developer/kugel)
11:15:50 Join nosa-j [0] (~m00k@
11:26:02 Quit nosa-j (Ping timeout: 264 seconds)
11:28:25 Join bertrik [0] (
11:28:25 Quit bertrik (Changing host)
11:28:25 Join bertrik [0] (~quassel@rockbox/developer/bertrik)
11:30:56 Join mortalis [0] (~mortalis@
11:35:26***Saving seen data "./dancer.seen"
11:47:33 Quit kugel (Ping timeout: 248 seconds)
11:48:27 Join nosa-j [0] (~m00k@
11:58:04 Quit Ketturi (Quit: WeeChat 0.4.0)
12:01:23 Join Guest17648 [0] (
12:06:45 Quit nosa-j (Ping timeout: 272 seconds)
12:10:26 Join nosa-j [0] (~m00k@
12:22:06 Join benedikt93 [0] (~benedikt9@unaffiliated/benedikt93)
12:24:23 Quit lorenzo92 (Remote host closed the connection)
13:04:48 Join krabador [0] (~krabador_@unaffiliated/krabador)
13:07:04 Quit krabador (Client Quit)
13:07:05funman[7]: is your jtagged player amsv1 or amsv2 ?
13:07:39 Join shai [0] (
13:11:46 Quit shai (Client Quit)
13:15:57 Quit benedikt93 (Ping timeout: 245 seconds)
13:19:17 Quit Scall (Ping timeout: 248 seconds)
13:27:06 Join Narod [0] (
13:31:35 Join Narod- [0] (
13:31:35 Quit Narod (Read error: Connection reset by peer)
13:32:38 Quit Narod- (Client Quit)
13:35:28***Saving seen data "./dancer.seen"
13:38:39 Join Scall [0] (~chat@unaffiliated/scall)
13:41:38 Join DuperMan [0] (
13:42:17DuperManpamaury: anyplace I can download plugins for zen? xD
13:43:26DuperManah. so will need to open the wiki, see which arch to aim at and build like I have half a clue what I'm doing
13:43:34DuperManso much work! :)
13:43:57pamaurythey don't compile for now, we don't have keymap for them
13:44:16DuperManw00t. isn't there a hal?
13:45:06DuperManlike, io's handled differently for plugins than rockbox or charsets? neat either way. guessing latter
13:45:31pamaurythis is not the matter, the matter is you need to teach each plugin/game/app what key to use for what purpose
13:45:44pamauryit's not like a computer, there is a no standard layout of keys
13:45:49DuperManwhoa. considered that but thought nowai
13:46:06DuperMandos-like. awesome.
13:46:43DuperManI honestly dig the implied simplistic nature of rockbox this conveys
13:47:22pamauryit's not of being simplistic, it's about being usable
13:47:46DuperMannot mutually exclusive. wasn't using simplistic as a derogative
13:48:24DuperMansimilar approach to diskette era games, 5'12 floppies remember?
13:48:56DuperManmakes me think mem extenders but that would totally negate the point lol
13:49:24 Quit Cosmoe (Quit: ZNC -
13:49:48DuperMan(forgot rockbox isn't an os but more of a load to mem app which handles native processor calls as it's os, forcing the wrong paradigm)
13:50:03pamauryfor the simplest ones we have a standard set of keys but the more complex one, you have to come up with a a reasonable assignement of the keys
13:50:06DuperMan(that is, I forced the wrong paradigm on the description to understand)
13:50:47DuperManstill not worth putting in a dll like extension when some devices have as little as 32k mem presumably
13:52:19DuperManhow much work, in theory, will it take to put some buffer lib to remap keys from a device with the same architecture?
13:53:35pamaurywe already have a such a lib, but there is no way to make it work for all plugins
13:53:56DuperManoh. I weep for doom :)
13:54:17n1sDuperMan: if two devices share the same physical button layout, using the same keymap is rather easy
13:54:24pamauryI assure you, people have tried and have failed providing a general framework, it's just not possible, the players are too different
13:54:34n1swith the exception of combo limitations
13:54:56DuperManI understand, sorry for redefining counter-productivity but I'm asking to understand better... :)
13:55:26pamauryunfortunately the ZEN has quite unique physical layout, fortunately though it has a lot of keys
13:56:13DuperMansure redefines a lot.. must be easiest on devices that can take a remote. more places to hook and stuff
13:56:29 Join Narod [0] (
13:56:36 Quit nosa-j (Ping timeout: 248 seconds)
13:58:07lebelliumpamaury: 25.5hrs and still running
13:58:10pamaurydo you want to help with this matter ?
13:58:48pamauryDuperMan: ^
13:59:13DuperManif I could.. using the software to learn here, still. :/ shouldv'e learned proper coding way earlier
13:59:18pamaurylebellium: ok, i've begun to compare the register map of E370 vs Rockbox, couldn't spot anything to relevant currently
13:59:35 Join nosa-j [0] (~m00k@
13:59:52pamauryDuperMan: have you ever programmed before or not ?
13:59:59pamauryBecause it can be done in the simulator
14:01:07DuperManpascal. I'll try set up the emulator - I doubt the zen is a known target yet :)
14:01:51pamauryyes it is, the simulator is working, just the plugins are disabled but it can be fixed easily
14:02:02pamauryare you using Windows or Linux ?
14:03:16DuperManah. yeah, you answered how to do that yesterday iirc so it should be in the irc logs. linux, a bit hesitant to setup the env with it's required compilers chain cause I blindly build cyanogenmod as well on the same box but I'll take the "risk"
14:04:30pamauryYou don't need any cross compiler for this, except if you want to test it on a real target, but in my opinion it's just faster to test it in the simulator
14:04:33DuperManhuess I can't say anything productive before doing that, so wiki re-reading time for me:) thanks
14:04:57pamauryIf you choose the simulator, read this:
14:05:26DuperManoh. figures, just need a simulated io system - ergo simulator, not emulator. thanks again ^_^
14:05:38 Quit Guest17648 (Ping timeout: 264 seconds)
14:06:14 Quit nosa-j (Ping timeout: 245 seconds)
14:07:35 Join Guest17648 [0] (
14:08:36 Join nosa-j [0] (~m00k@
14:09:05DuperMangood old target 98. herp derp
14:09:31 Quit thegeek (Ping timeout: 260 seconds)
14:10:39DuperMandrats. Xming should suffice as a gui right?
14:11:02DuperManmy setup remotes all the things through windows
14:11:55pamauryI guess, trying is still the best way to know, right ?
14:12:56DuperManshall be done momentarily :D
14:14:31DuperManheh. crashed Xming. will try in xfce oO
14:17:13 Quit Guest17648 (Ping timeout: 272 seconds)
14:17:25 Join krabador [0] (~krabador_@unaffiliated/krabador)
14:17:32DuperMannow updates to many of the things. will report laters
14:18:07 Join Guest17648 [0] (
14:19:37 Join rela [0] (~x@pdpc/supporter/active/rela)
14:20:50 Quit krabador (Read error: Connection reset by peer)
14:25:15DuperManfunny how framework dependant the simulator is in contrast to rockbox itself :) laters
14:25:32 Quit nosa-j (Ping timeout: 272 seconds)
14:27:39 Join nosa-j [0] (~m00k@
14:32:54 Join benedikt93 [0] (~benedikt9@unaffiliated/benedikt93)
14:35:12 Quit JdGordon (Ping timeout: 250 seconds)
14:40:06 Quit nosa-j (Ping timeout: 272 seconds)
14:41:39 Join nosa-j [0] (~m00k@
14:42:29 Quit n1s (Ping timeout: 245 seconds)
14:48:57 Quit benedikt93 (Quit: Bye ;))
14:50:55 Join n1s [0] (~n1s@rockbox/developer/n1s)
15:13:38 Quit rela (Read error: Connection reset by peer)
15:14:59 Quit lebellium (Ping timeout: 252 seconds)
15:15:33 Join lebellium [0] (
15:19:47 Join y4n [0] (~y4n@unaffiliated/y4ndexx)
15:26:36 Quit mortalis (Quit: KVIrc 4.3.1 Aria
15:28:53 Quit [Saint] (Remote host closed the connection)
15:31:03 Join [Saint] [0] (~saint@rockbox/user/saint)
15:35:29***Saving seen data "./dancer.seen"
15:42:44 Quit nosa-j (Ping timeout: 248 seconds)
15:45:15 Join nosa-j [0] (~m00k@
15:47:24DuperManpamaury: hooray. when I load xfce beforehand the simulator loads dandy
15:47:46DuperManno audio output at the moment but I doubt that part merits more tweaking :)
15:50:29DuperMan(meaning my workstation isn't set up for audio - I have no reason to assume it works nor to test that:)
15:55:28 Join foolsh [0] (
15:56:11 Join krabador [0] (~krabador_@unaffiliated/krabador)
15:58:51 Quit ungali_ (Quit: Lost terminal)
15:59:43lebelliumpamaury: have to leave for the weekend now, so I stop the benchmark and will resume it on Sunday evening. 27.5hrs so far!!
16:01:16DuperManruntime lebellium?!?!?!?!
16:03:06 Quit lebellium (Quit: ChatZilla [Firefox 26.0/20131114085019])
16:12:54 Quit nosa-j (Ping timeout: 245 seconds)
16:14:25 Join nosa-j [0] (~m00k@
16:30:13 Quit nosa-j (Ping timeout: 265 seconds)
16:31:48 Join nosa-j [0] (~m00k@
16:41:24[7]funman: clip+
16:47:38funman[7]: did you look at the USB issues with it?
16:48:05funmanif you don't want it I might take a look but I don't know how much time I want to invest
16:48:17funman(specifically on USB)
17:12:37 Join thegeek [0] (
17:16:20 Join kugel [0] (
17:16:20 Quit kugel (Changing host)
17:16:20 Join kugel [0] (~kugel@rockbox/developer/kugel)
17:17:42 Quit thegeek (Ping timeout: 272 seconds)
17:22:32 Join Ketturi [0] (
17:25:47 Join thegeek [0] (
17:34:04 Quit Strife89 (Quit: Vamoose!)
17:35:32***Saving seen data "./dancer.seen"
17:36:17[7]no, I didn't work on USB with that one
17:36:52[7]that was mostly an attempt to figure out a way how to unbrick those things without jtag, but it turned out that it probably has irrecoverable damage
17:42:27 Quit Narod ()
17:49:19 Quit krabador (Quit: Sto andando via)
18:00:08 Join mortalis [0] (~mortalis@
18:10:15funmancan you still load custom binary with jtag? I guess I could use the uSD
18:14:03 Quit thegeek (Read error: No route to host)
18:14:15 Join thegeek [0] (
18:20:52 Quit n1s (Quit: Ex-Chat)
18:29:29 Quit shamus (Read error: Connection reset by peer)
18:29:35 Join shamus [0] (
18:48:02 Join sakax [0] (~sakax@unaffiliated/sakax)
19:11:41 Quit shamus (Read error: Connection reset by peer)
19:12:06 Join ender` [0] (
19:14:03pamaury[7]: don't remember, what is the datasheet you suggest for the usb core desc ? I'm still disassembling the clip+ ROM to see how it handles this STALL
19:29:00 Join mirak [0] (
19:30:09pamaurythis code is seriously brain dead, I've already found 300 usb functions/methods and there is more
19:30:36funmanpamaury: i have S3C6400X_UserManual_rev1-0_2008-02_661558um.pdf
19:31:28pamauryI'm looking at STM32F4xx
19:31:35pamauryit has a lot of details
19:33:34[7]funman: gevaerts managed to make a build that seemed to mostly work on that jtaghed clip+
19:33:46gevaertsI did?
19:33:49[7]pamaury: yes, I suggested the stm32 one
19:34:00[7]hm, I think it was you... months ago
19:34:12[7]or possibly funman? not sure
19:34:18gevaertsI'm not saying it wasn't, but I don't remember anything about that
19:34:23[7]too long ago
19:34:30gevaertsSomeone did :)
19:35:34***Saving seen data "./dancer.seen"
19:36:19funmaniirc it was me
19:36:23[7]pamaury: we could look at ST's reference implementation as well... it is equally overcomplicated with OTG and everything, but open source
19:37:04[7]it's kinda funny that nano2g and classic behave differently wrt. that stall bug, with the very same code
19:37:08pamaury[7]: do you have the link to the code ?
19:38:04pamaurybut at least the ROM code works for sure ! It's just a f*** nightmare, I haven't found anything special about stall by the way
19:38:53[7]I think it's
19:41:06[7] hm, or could be that's for F1
19:41:51pamaurythe explaination about the fifo allocation are interesting too, it's clearly more complicated than I expected
19:46:11[7]various usb implementations can be found there
19:49:08lonoxmont[Saint]: any reason that the log file that is supposed to capture iap commands isnt capturing any?
19:50:45 Join shamus [0] (
19:56:13 Quit bertrik (Remote host closed the connection)
20:09:40gevaertslonoxmont: one reason could be that there aren't any commands to log. Do you happen to have access to any ipod accessory that's not your car so you can try there?
20:11:06 Quit DexterLB (Ping timeout: 265 seconds)
20:13:56 Join DexterLB [0] (~dex@
20:28:31lonoxmontgevaerts: ill look around but im not sure that i have anything smart enough to send iap, just some docks and things
20:37:24 Quit mortalis (Quit: KVIrc 4.3.1 Aria
21:01:10 Quit shamus (Ping timeout: 240 seconds)
21:06:01 Join n1s [0] (~n1s@rockbox/developer/n1s)
21:23:48 Join wodz [0] (
21:33:35 Quit sakax (Quit: Leaving)
21:33:56 Join sakax [0] (~sakax@unaffiliated/sakax)
21:34:43 Quit sakax (Client Quit)
21:35:04 Join sakax [0] (~sakax@unaffiliated/sakax)
21:35:36***Saving seen data "./dancer.seen"
21:36:13andypotterlonoxmont:Most ipod hifi docks should use iap if they allow control from the hifi. Anything that can control the ipod should do.
21:36:14andypotterOn my ipod video, I get 3 lines in the log file regarding codecs even if the ipod is not connected to a dock. Do you get any info in the log?
21:36:55lonoxmontandypotter: nothing at all in the file, it doenst even create one as far as i can tell
21:37:03lonoxmontdo i have to make the file manually and then it will log to it?
21:37:36wodzlonoxmont: no
21:37:47lonoxmontmaybe i didnt compile it right
21:38:00wodzlonoxmont: I would double check that you enabled logging in the build
21:38:16lonoxmonti think i forgot to do a make clean the last time i tried to compile
21:38:23lonoxmontill try doing that first then building again
21:39:29wodzmake clean && configure && make to be sure
21:40:05 Quit thegeek (Read error: Connection reset by peer)
21:41:46lonoxmontalso in the options in configure, under advanced build do i need to make sure (D)EBUG is selected as well or just (L)ogf?
21:42:08wodzlogf should be enough
21:42:59wodzBUT debug doesn't harm in this case. I don't know iap code so I can't say if something is not protected with DEBUG define
21:44:20 Quit kugel (Read error: Operation timed out)
21:47:12lonoxmonthmm, when i went to make zip it spit out another line of compiler stuff
21:47:25lonoxmontlonox@patchouli:~/rockbox/build$ make zip
21:47:25lonoxmontLD rockboy.rock
21:47:57lonoxmonthasnt done that previously
21:52:58wodzits about plugin - irrelevant here
21:54:06andypotterShould not need debug, I just use (A)dvanced and (L)ogf and it creates the log. When finished, the rockbox.ipod binary should be about 750784 bytes in size. The normal build version should be about 747504 bytes (based on current git).
21:58:00lonoxmontsize matches your number
21:58:06lonoxmontill grab the zip now
21:58:56sakaxhello rockbox lovers, how is ipod nano 2g progressing? especially the usb issue :)
21:59:57 Join bluebrother [0] (~dom@rockbox/developer/bluebrother)
22:00:18lonoxmontdo i need to delete the .rockbox folder on the ipod before unzipping this onto the ipod? was thinking of maybe renaming the current .rockbox so we know for sure its loading the debug version i just made
22:01:25 Quit fs-bluebot (Ping timeout: 264 seconds)
22:02:52 Join fs-bluebot [0] (
22:03:17 Join saratoga [0] (123e1c4b@gateway/web/freenode/ip.
22:03:42 Quit bluebrother^ (Ping timeout: 272 seconds)
22:04:00lonoxmontalright, i moved my current .rockbox to .rockbox-old, so it shuold for sure load the dev version now
22:04:07lonoxmont(after unzipping)
22:04:24andypottershouldn't need to delete the .rockbox folder. I just copy in the rockbox.ipod binary, leaving everything else alone and it works OK.
22:05:34lonoxmontstill nothing in debug > show log file
22:05:47lonoxmontnot even lines about codecs
22:05:56lonoxmontill try recompinling with debug set as well
22:06:27saratogathis is with logf? usually you need to compile a debug build AND enable logf in the file you want to log. i assume you did both of those?
22:07:14lonoxmontive done what andypotter said to do, made a build with logf set
22:07:42lonoxmonti dont know what to do past that
22:07:44saratogadid you enable it for whatever files you want to log?
22:07:53lonoxmonthow do i do that?
22:08:00saratogaI think you have to add a #define and maybe uncomment logf.h
22:08:53lonoxmontwhere would i do that?
22:09:27saratogaeach file that calls logf needs to have it enabled in that specific file
22:09:40saratogaotherwise your log would fill up with thousands of unrelated debug messages very quickly
22:10:02andypotterThe IAP code will automatically output to logf.txt if an (A)dvanced build with (L)ogf is selected. Shouldn't need anything more than that, mine doesn't.
22:10:38saratogaah it shouldn't do that, but I guess its ok since you'll only get output if using an accessory which is unlikely to conflict with anything
22:13:28 Join thegeek [0] (
22:37:49 Join kugel [0] (
22:37:49 Quit kugel (Changing host)
22:37:49 Join kugel [0] (~kugel@rockbox/developer/kugel)
22:48:51 Join Narod [0] (
22:55:39lonoxmontlooks like its logging correctly when connecting to an ihome dock alarm thing my sister had, ipod didnt do much besides that though
22:56:17lonoxmontill go try it on my car one more time
22:56:33lonoxmontunless you wanted that logfile for whatever reason
23:04:22lonoxmontunfortunately it logged nothing when connecting to my car, meaning its probably using iap over usb, and thus undebuggable
23:04:58wodzunsupported rather
23:05:15lonoxmontim guessing permanently unsupported due to lisencing issues
23:05:23andypotterAnyone want to write an IAP over USB patch? ;)
23:05:45wodzlonoxmont: we would happily accept patch for this :-)
23:06:54lonoxmontwell ive got a possible lead on the whitepaper describing iap over usb, dunno if itl pan out or not, wont find out for several months
23:07:23saratogatheres detailed specifications for IAP over USB on google
23:07:52lonoxmontso its an untractable problem or something?
23:08:05saratogabut it uses public key encryption to authenticate devices, so unless theres a flaw in their crypto, or your vendor was lazy about verification, you won't be able to talk to it
23:08:35lonoxmontah, unless apples private key gets leaked...
23:08:55saratogai wouldn't be surprised if a lot of devices are lazy about authentication though, the point of the PKA was to prevent unauthorized ipod accessories, probably not to prevent unauthorized ipods
23:09:14andypotterIt's on my list of things to do but it's nowhere near the top. Plus I've very little info to go on other than that in the Apple Ipod Accessory Protocol Docs.
23:09:33saratogaandypotter: which ones do you have?
23:10:38 Quit saratoga (Changing host)
23:10:38 Join saratoga [0] (123e1c4b@rockbox/developer/saratoga)
23:12:40wodzpamaury: I made another attempt to run hwstub on rk27xx and failed. My observations are as follows: 1) At least on my box it takes considerable time for device to get setup request. The one which passes blocking wait in usb_drv_recv_setup() is just before usb bus reset. 2) The device simply doesn't respond to this request. usb_drv_send() is entered but wireshark doesn't see any data :-/
23:13:47pamauryyeah, same observation here. Last I time I tried it was similar iirc
23:14:33andypottersaratoga: Ive got Apple Accessory Protocol Interface Spec R38 ( a lot of screenshots from chinese site), R35 as a PDF and Ipod Extended Interface Spec R21.
23:15:00saratogaok thats probably all the ones i've seen floating aorund
23:15:55lonoxmontandypotter: if you had access to more could oyu do more?
23:16:13wodzpamaury: I spotted a couple of possible errors but still I can't make it going
23:17:45andypotterI could try. I know it wont be easy but I could do with it for the wife's Renault head unit which is IAP over USB.
23:19:51saratogai believe the USB protocol uses some variant of the USB audio spec
23:19:56saratogaso supporting that would be required too
23:20:26saratogathis could be useful on other devices as well
23:21:02wodzpamaury: Next time I'll try interrupt version of usb driver.
23:23:27pamauryon the rknanob I used interrupts and I have a similar problem: I never get the setup interrupt for some reason
23:23:38andypotterI'll look through what I have but I don't have a lot of time available and there's three things I want to do first, 1G Mini IAP Support, 2G Nano IAP Support and 6G Classic IAP.
23:24:23andypotterThe first one I have working, the others will take a lot longer.
23:26:42saratogayeah getting IAP on non-PP devices would make people happy
23:28:05wodzpamaury: Can you pastebin somewhere irq version of the stub?
23:29:28 Quit Scall (Ping timeout: 272 seconds)
23:31:09wodzpamaury: What is the sequence of events from the device POV? Reset, Setup ?
23:31:39pamaurywith rknanob ?
23:31:43 Quit n1s (Quit: Ex-Chat)
23:32:55wodzpamaury: with every device, no?
23:33:20pamauryI was referring to your first statement
23:33:40 Join Guest26236 [0] (
23:33:41pamauryirq are handled quite differently on rknano since it's cortex-m3
23:34:42wodzpamaury: I am interested in changes which makes stub asynchronous rather then fully synchronous.
23:35:08pamaurythere are none, only the driver in asynchronous, and it's not working
23:35:27wodzpamaury: I guess you setup some flag or something which is cleared in irq.
23:35:37***Saving seen data "./dancer.seen"
23:36:17pamauryYeah but I don't see the point, I can post the code if you want but I mean the rknano driver has been using irq since the beginning
23:38:07pamauryand it's quite incomplete too since I can't even get a setuo

Previous day | Next day