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 2010-02-05

00:16:48 Quit maraz (Ping timeout: 248 seconds)
00:16:51 Join maraz [0] (
00:16:52 Quit bertrik (Quit: De groeten)
00:18:55 Quit ender` (Quit: Marriage is not a word, it's a sentence. A life sentence...)
00:19:28 Quit PaulJam_ (Ping timeout: 248 seconds)
00:21:09*Unhelpful is curious to see what happens with the reciprocal table on nano2g... i understand iram is slower than cache there, although still not as slow as dram...
00:30:03CIA-8New commit by stripwax (r24514): XPROD31_R, XNPROD31_R no longer require/expect the first argument to be a read/write register (e.g. could be a constant)
00:30:42 Quit arbingordon (Read error: Connection reset by peer)
00:31:04 Join arbingordon [0] (
00:32:25 Part froggyman
00:34:09 Quit JdGordon_ (Quit: Page closed)
00:34:13CIA-8New commit by stripwax (r24515): restore previous fft-ffmpeg.c - no idea what I did to break it so..
00:36:15 Quit Rob2223 (Quit: Rob2223)
00:42:16CIA-8New commit by stripwax (r24516): Make same change to XPROD32, XPROD31, XNPROD31 , to avoid requiring that arg0 is read/write. Might save register reloads/etc in calling code and ...
00:42:48stripwax^all mdctexp branch
00:59:41CIA-8New commit by stripwax (r24517): Didn't mean to make XPROD31_R and XNPROD31_R asm 'volatile' as part of the committed version. It performs about 0.5MHz faster without the volatile ...
01:03:13 Quit saratoga (Ping timeout: 248 seconds)
01:03:13 Quit evilnick_B (Ping timeout: 248 seconds)
01:03:36 Quit Farthen (Disconnected by services)
01:03:58 Join Farthen_ [0] (
01:04:00 Nick Farthen_ is now known as Farthen (
01:06:22 Quit dfkt (Quit: -= SysReset 2.53=- Ph'nglui mglw'nafh Cthulhu R'lyeh wgah'nagl fhtagn.)
01:08:56CIA-8New commit by stripwax (r24518): Correct me if I'm wrong, but %r doesn't mean anything for inline asm register inputs list..
01:11:44stripwaxI learn a new thing every day. I *am* wrong. fab!
01:12:07Unhelpfulwhat *does* % mean on a constraint? i can't recall having seen it before...
01:17:13stripwaxYeah. In this case though I still don't want to use % modifier, as I'm (hopefully) passing in the args in the most efficient order for early termination of the SMULL
01:21:43 Quit fyrestorm (Read error: Connection reset by peer)
01:23:55 Join froggyman [0] (
01:24:23 Quit flydutch (Quit: /* empty */)
01:24:46 Part froggyman
01:27:10 Quit anewuser (Quit: =o)
01:27:20CIA-8New commit by stripwax (r24519): You obviously cannot have an early-terminating MUL if cPI2_8 is the second arg, so make it the first arg
01:35:46***Saving seen data "./dancer.seen"
01:42:02 Join panni__ [0] (
01:42:18 Quit panni__ (Read error: Connection reset by peer)
01:42:22 Quit panni_ (Read error: Connection reset by peer)
01:42:33 Join panni_ [0] (
01:46:28 Quit stripwax (Quit:
01:58:48 Part toffe82_
01:59:35 Quit martian67 (Read error: Operation timed out)
02:00:58 Join martian67 [0] (~martian67@about/linux/regular/martian67)
02:02:42 Quit Farthen (Ping timeout: 258 seconds)
02:03:56 Join Farthen [0] (
02:23:32 Quit DerPapst (Quit: Leaving.)
02:25:59 Join froggyman [0] (
02:27:52 Join Tomis2 [0] (~Tomis@
02:29:50 Quit Tomis (Ping timeout: 240 seconds)
02:29:51 Nick Tomis2 is now known as Tomis (~Tomis@
02:58:33 Join BHSPitMonkey [0] (~stephen@unaffiliated/bhspitmonkey)
02:59:15 Join CaptainKewl [0] (
03:07:15 Join eliasp [0] (
03:07:59eliaspaccording to the changelog ( → 1.2.4) there's now support for the Cowon D2 in rbutil, but it is 'currently disabled'... is there some CLI switch to re-enable it?
03:14:43 Join Tomis2 [0] (~Tomis@
03:14:52 Quit beta2k (Ping timeout: 276 seconds)
03:16:05 Join beta2k [0] (
03:16:40 Quit Tomis (Ping timeout: 240 seconds)
03:16:40 Nick Tomis2 is now known as Tomis (~Tomis@
03:28:35 Quit mc2739 (Ping timeout: 260 seconds)
03:30:35 Join mc2739 [0] (~mc2739@rockbox/developer/mc2739)
03:32:00 Quit komputes (Ping timeout: 248 seconds)
03:35:50***Saving seen data "./dancer.seen"
03:36:52 Join gigatheme [0] (
03:37:30gigathemeany way to edit a theme i uploaded? put wrong file in on the shuffle and repeat images
03:42:18 Quit Sajber^ (Read error: Connection reset by peer)
03:44:19krazykitgigatheme, please have patience in waiting for a response. those that know about the theme site may not be around right now, but they'll answer when they see the question
03:47:44 Join flyback [0] (
03:52:12gigathemewhat is the new playlist viewer in the WPS on the 3.5 release
03:55:00 Quit gigatheme (Quit: CGI:IRC)
03:57:22 Join Strife89 [0] (~michael@
03:59:31 Quit phanboy4 (Quit: Leaving)
04:02:08 Quit MethoS- (Remote host closed the connection)
04:05:09eliasphmm, is it possible installing .rockbox onto a ext2/3 volume or will it only work on a vfat volume?
04:10:52 Part froggyman
04:11:06Unhelpfuleliasp: only fat32 is supported
04:11:32 Join Casainho [0] (
04:12:02eliaspUnhelpful: ok, thx, so i'll use vfat for the boot volume... and what about secondary storage volumes? does rockbox support reading files from ext2/3?
04:12:53eliasphmm... then i'll have to find a script which overrides these crazy vfat filename restrictions by renaming the files.. ;-(
04:13:42Unhelpfuleliasp: utilities for generating FS-safe filenames from tags are extremely common
04:14:28eliaspUnhelpful: i'm actually syncing my files using rsync... so let's see how i can integrate such a script into this proces...
04:15:41Unhelpfuleliasp: well, the easiest way would be to sync them after giving them names that are safe...
04:16:16eliaspUnhelpful: i sync them directly from my mediaserver, so this would involve copying them first to a temporary location and then onto the device...
04:16:25eliaspok, what i could do... create safe named symlinks
04:16:41eliaspthat's probably the best way
04:16:46Unhelpful...or rename the files with safe names on your server?
04:16:50eliaspsymlinks with safe names in a temporary location...
04:17:11eliaspUnhelpful: nope... just because of this crappy vfat i will not rename all my files on my server
04:18:36Unhelpfulit's by far the lowest-maintenance solution. and "crappy" though it may be vfat is the most widely-supported filesystem, unless you count feature-incomplete ext2 support on windows as making ext3 "widely supported"
04:19:15 Quit Tomis (Quit: Tomis)
04:19:29Unhelpfulmy suggestion is that you do what works. it's not as if filenames are the most important thing for music. :)
04:19:36eliaspUnhelpful: i don't have any windows computers at all.... so i don't want to restrict my self to such a common denominator just because of a mobile player...
04:19:58eliaspbut i think i'm fine with the symlink+script solution
04:20:12Unhelpfuleliasp: your choice then, have fun. i don't know anything that will help you with your symlinks. :)
04:20:34eliaspsome script-foo will do... /me starts the ruby-engines ;-)
04:20:45eliaspUnhelpful: thx anyways for your help
04:21:13 Join Tomis [0] (~Tomis@
04:28:17 Quit TheSeven (Disconnected by services)
04:28:29 Join The_Seven [0] (~theseven@rockbox/developer/TheSeven)
04:28:40 Nick The_Seven is now known as TheSeven (~theseven@rockbox/developer/TheSeven)
04:32:27 Join phanboy4 [0] (
04:37:39 Join fyrestorm [0] (
04:44:53 Quit antil33t ()
04:52:55 Quit Barahir (Ping timeout: 246 seconds)
04:54:51 Join Barahir [0] (
05:06:28 Quit Casainho (Quit: ChatZilla 0.9.86 [Firefox 3.6/20100115133306])
05:09:20 Join komputes [0] (~komputes@ubuntu/member/komputes)
05:09:47 Quit kaniini (Quit: E.T. go home.)
05:15:13 Join kaniini [0] (
05:17:35 Quit komputes (Ping timeout: 256 seconds)
05:27:43 Quit BHSPitMonkey (Remote host closed the connection)
05:28:24 Quit moos (Quit: ChatZilla 0.9.86 [Firefox 3.6/20100115144158])
05:31:11 Quit CaptainKewl (Read error: Connection reset by peer)
05:33:38 Quit Horscht (Quit: Verlassend)
05:33:49 Join wind [0] (
05:35:54***Saving seen data "./dancer.seen"
05:39:35 Join Rob2222 [0] (
05:48:03 Join CaptainKewl [0] (
05:49:34Unhelpfulhrm, to what degree is a previously-used codec "reused"? can i be certain that the codec's static data will be freshly copied from the buffer or the disk before the start of each file? if so, i think i see a way to get best-possible performance out of ape for each profile without greatly complicating its startup...
05:51:33 Join komputes [0] (~komputes@ubuntu/member/komputes)
05:56:00Unhelpfulinstead of having a real malloc for iram, do something like this: fill available iram with table, but calculate a table size for each profile based on the space that would remain after statically allocating that profile's buffers from the end of iram. make a table with the compare instructions for each table size. on codec init, copy the appropriate instructions from the table to the start of udiv32_arm (which can easily be ordered to put
05:56:00Unhelpful this compare first), and allocate filter buffers starting from the end of iram, aliasing the reciprocal table. if we make sure the end-of-iram address is available at link time, these can even be declared static and initialized in much the same way as they are now.
05:57:31 Quit panni_ (Quit: ( :: NoNameScript 3.81 :: ))
06:01:38 Quit komputes (Quit: I haven't slept for ten days, because that would be too long.)
06:02:22flybackGOD DAMMIT
06:02:43*flyback continues surgery on his h10
06:03:29flybackhow do I find out how many mod formats the mod routines handle and how I can mabye help
06:03:53 Join komputes [0] (~komputes@ubuntu/member/komputes)
06:05:18*flyback used to be friends with mikmak
06:08:51 Quit wind (Quit: CGI:IRC (EOF))
06:13:05 Join Rob2223 [0] (
06:16:31 Quit Rob2222 (Ping timeout: 256 seconds)
06:17:54 Join TopyMobile_ [0] (
06:19:43 Quit komputes (Ping timeout: 246 seconds)
06:21:44 Quit TopyMobile (Ping timeout: 252 seconds)
06:32:11 Join TopyMobile__ [0] (
06:36:21 Quit TopyMobile_ (Ping timeout: 256 seconds)
06:37:48 Join komputes [0] (~komputes@ubuntu/member/komputes)
06:42:55flybackIT LIVES
06:46:26 Join TopyMobile_ [0] (
06:47:41 Quit Slasheri (Ping timeout: 256 seconds)
06:49:16 Join Slasheri [0] (miipekk@rockbox/developer/Slasheri)
06:49:57 Quit TopyMobile__ (Ping timeout: 256 seconds)
06:55:20 Join wind [0] (
07:00:44 Join TopyMobile__ [0] (
07:03:52windWhy can‘t I rotate a picture with Rockbox?
07:04:41 Quit TopyMobile_ (Ping timeout: 256 seconds)
07:05:26 Quit TopyMobile__ (Client Quit)
07:12:44JdGordonrockbox is no photoshop/gimp?
07:14:23 Quit nimak (Changing host)
07:14:23 Join nimak [0] (~nima@unaffiliated/nima)
07:24:47 Quit grndslm (Read error: Operation timed out)
07:25:54 Part davidb_csh
07:27:49 Quit Farthen (Read error: Connection reset by peer)
07:28:34 Join grndslm [0] (
07:31:50 Join Farthen [0] (
07:32:06 Quit Farthen (Read error: Connection reset by peer)
07:33:07 Join matsl [0] (
07:33:50 Join Farthen [0] (
07:34:11 Quit Farthen (Read error: Connection reset by peer)
07:35:55***Saving seen data "./dancer.seen"
07:36:17 Join Farthen [0] (
07:37:08 Quit Farthen (Read error: Connection reset by peer)
07:38:13 Join Farthen [0] (
07:38:47 Quit Farthen (Read error: Connection reset by peer)
07:39:37 Join Farthen [0] (
07:40:26 Quit grndslm (Remote host closed the connection)
07:40:35 Quit Farthen (Read error: Connection reset by peer)
07:41:22 Quit rhodan (Remote host closed the connection)
07:41:40 Join Farthen [0] (
07:41:48 Quit Farthen (Read error: Connection reset by peer)
07:43:58 Join Farthen [0] (
07:44:03 Quit Farthen (Read error: Connection reset by peer)
07:44:54 Join Farthen [0] (
07:44:54 Quit Farthen (Read error: Connection reset by peer)
07:45:56 Join Farthen [0] (
07:46:03 Quit Farthen (Read error: Connection reset by peer)
07:46:56 Join Farthen [0] (
07:46:58 Quit Farthen (Read error: Connection reset by peer)
07:48:25 Quit Slasheri (Ping timeout: 264 seconds)
07:50:12 Join Farthen [0] (
07:50:25 Quit Farthen (Read error: Connection reset by peer)
07:52:34Unhelpfulhrm, wonder how much codec ram ape typically has left... i could just put a copy of the reciprocal table in dram and copy "enough" of it to iram on codec init
07:52:40 Join Farthen [0] (
07:53:08 Quit Farthen (Read error: Connection reset by peer)
07:53:48 Quit komputes (Ping timeout: 265 seconds)
07:53:58 Join Farthen [0] (
07:54:50 Quit Farthen (Read error: Connection reset by peer)
07:55:57 Join stoffel [0] (
07:56:25 Join Farthen [0] (
07:56:43 Quit Farthen (Read error: Connection reset by peer)
07:58:58Mode"#rockbox +o TheSeven" by ChanServ (ChanServ@services.)
07:59:08Mode"#rockbox +b *!*" by TheSeven (~theseven@rockbox/developer/TheSeven)
07:59:12Mode"#rockbox -o TheSeven" by ChanServ (ChanServ@services.)
08:02:10 Join Farthen [0] (
08:06:08Mode"#rockbox +o TheSeven" by ChanServ (ChanServ@services.)
08:06:10Mode"#rockbox -b *!*" by TheSeven (~theseven@rockbox/developer/TheSeven)
08:06:15Mode"#rockbox -o TheSeven" by ChanServ (ChanServ@services.)
08:07:12 Join komputes [0] (~komputes@ubuntu/member/komputes)
08:15:05 Join Zagor [0] (
08:15:05 Quit Zagor (Changing host)
08:15:05 Join Zagor [0] (~bjst@rockbox/developer/Zagor)
08:15:29 Join stoffel_ [0] (
08:17:19 Quit stoffel (Ping timeout: 246 seconds)
08:20:03 Quit CaptainKewl (Remote host closed the connection)
08:22:44 Quit arbingordon (Ping timeout: 252 seconds)
08:25:17 Join ender` [0] (
08:28:05 Quit komputes (Ping timeout: 245 seconds)
08:40:16 Join komputes [0] (~komputes@ubuntu/member/komputes)
08:48:39amiconnUnhelpful: Reloading from disk would defeat the point of buffering codecs in the first place. And since buffering it multiple times would be a waste, reuse was implemented. Nothing gets reloaded, the codec asks the core whether there is another track to decode
08:49:42Unhelpfulamiconn: ok... so it will need to be able to re-initialize the reciprocal table. and udiv32 and its table will need to be the very last thing in iram... maybe add a section for that purpose?
08:49:45amiconnI think that filling the reciprocal table programmatically (doing actual divisions) wouldn't take long
08:50:32 Join bertrik [0] (
08:50:34amiconnBut probably it's not even worth the hassle
08:50:51amiconnDid you compare the results for PP5002 and PP5022?
08:51:13 Join flydutch [0] (
08:53:42 Quit TheSeven (Quit: ChatZilla 0.9.86 [Firefox 3.5.7/20091221164558])
08:54:06amiconnPP5022 is only 1.8% faster than PP5002 at -c1000, although the table is 3.6 times as large
08:54:13 Quit komputes (Ping timeout: 265 seconds)
08:55:49Unhelpfulright... the bulk of the value of the table is up front, and the only reason to maximize it is to get whatever small gain is possible out of the available space. :)
08:56:42Unhelpfulthinking that way, it's probably not worth doing dynamic table allocation... just have enough static buffer in iram for modes that might possibly decode in realtime, and use the rest for the table
09:00:20Unhelpfuli note that the speex codec has some saturate operations in it also. is it important for the codec that saturate be to -32767/+32767, rather than full 16-bit range?
09:00:54 Join petur [0] (~petur@rockbox/developer/petur)
09:07:17 Join komputes [0] (~komputes@ubuntu/member/komputes)
09:07:48 Join Tomis2 [0] (~Tomis@
09:09:25 Quit Tomis (Ping timeout: 264 seconds)
09:09:25 Nick Tomis2 is now known as Tomis (~Tomis@
09:11:40 Join Buschel [0] (
09:12:49 Join Bagder [0] (~daniel@rockbox/developer/bagder)
09:12:52Buschelamiconn: r24512 does not make use of the fact that iram size on PP5022/5024 is 128KB instead of 96KB (all other PPs)?
09:15:13amiconnHuh? Of course it does
09:15:20 Join TheSeven|Mobile [0] (~theseven@rockbox/developer/TheSeven)
09:15:56UnhelpfulBuschel: the space not used by filter buffers is being used for exact reciprocals - that's handled at link time, targets with more iram get a bigger table
09:16:29 Quit Tomis (Read error: Connection reset by peer)
09:17:04 Join Tomis [0] (~Tomis@
09:17:51 Quit TheSeven|Mobile (Read error: Connection reset by peer)
09:18:01*amiconn is puzzled by the weird behaviour of both his PP5022 targets which means he cannot get results for -c5000
09:18:15 Join TheSeven|Mobile [0] (~theseven@rockbox/developer/TheSeven)
09:18:56amiconnBoth the mini G2 and the c240 shut down due to low-bat at a point not even halfway through the file
09:19:22amiconnNo other target does this, even though PP5020 and PP5002 are slower
09:19:41Unhelpfulamiconn: can't you plug them in while testing?
09:19:53amiconnAnd after powering back up, the battery is nowhere near empty. It even happens when keeping the charger connected
09:20:47amiconnI suspect that somehow the low yield rate affects the powermanagement thread, but that would be really strange
09:20:49Unhelpfuloh :/ there's not a watchdog on them, is there? that's maybe not being serviced if the decoder is too long between yields on very very slow files?
09:21:03amiconnNo watchdog...
09:21:40*amiconn plans to disable low-bat shutdown on those targets for a test tonight
09:21:58 Quit Strife89 (Quit: See ya.)
09:22:49 Quit TheSeven|Mobile (Ping timeout: 258 seconds)
09:24:53 Join utchybann [0] (
09:26:01amiconnUnhelpful: You have an e200?
09:26:21 Quit bertrik (Ping timeout: 256 seconds)
09:26:43Unhelpfulamiconn: yes. i didn't have any trouble of this sort with the normal/high testfiles.
09:27:39BuschelUnhelpful: ok, didn't take into account link time stuff. so everything's fine :o)
09:28:18 Quit Buschel ()
09:28:49 Quit GodEater (Ping timeout: 248 seconds)
09:28:50amiconnUnhelpful: I'm not talking about normal/high. It *only* happens for c5000 here
09:29:11amiconnCould you perhaps do that test on your e200 using the official testfile?
09:29:24Unhelpfulamiconn: i've never run any tests with a c5000 sample... did i see earlier that you added ones for every level to the testfiles set?
09:29:36amiconnIt may take 1.5..2 hours ...
09:29:58Unhelpfulunless it mysteriously turns itself off :)
09:31:10*amiconn should do that test twice
09:31:28Unhelpfulhrm... it seems the same logic used in the ape codec saturate code could be used to get the -32767..32767 range used by speex with the addition of an orrne result, result, #1...
09:31:40amiconnOnce with just low-bat shutdown disabled so that the test finishes, and another time logging the measured batt voltage in parallel
09:33:35Unhelpfuli know speex is a voice codec... are there some standard testfiles for it?
09:35:56***Saving seen data "./dancer.seen"
09:38:14 Join Strife89DS [0] (~nds@
09:42:49 Quit Tomis (Quit: Tomis)
09:43:52 Quit komputes (Read error: Operation timed out)
09:46:53 Join PaulJam_ [0] (
09:52:05 Quit thegeek (Read error: No route to host)
09:53:43Unhelpfulit's definitely starving other threads a bit, it took a while for the backlight to go off and it pauses visibly between steps whil fading
09:59:16 Join komputes [0] (~komputes@ubuntu/member/komputes)
10:10:57 Quit AndyI (Ping timeout: 248 seconds)
10:11:25 Quit phanboy4 (Ping timeout: 245 seconds)
10:15:01 Quit Strife89DS (Quit: ClIRC - IRC client for Nintendo DS)
10:17:14 Join AndyI [0] (~pasha_int@
10:22:17 Quit Tuplanolla (Ping timeout: 240 seconds)
10:28:57 Quit komputes (Ping timeout: 265 seconds)
10:33:04 Join GodEater [0] (
10:33:04 Quit GodEater (Changing host)
10:33:04 Join GodEater [0] (~bibble@rockbox/staff/GodEater)
10:41:57 Join komputes [0] (~komputes@ubuntu/member/komputes)
10:50:14 Quit Frampis (Ping timeout: 258 seconds)
10:52:08Unhelpfulamiconn: ok, at some point my e200 shut off :/
10:54:20 Quit komputes (Ping timeout: 245 seconds)
10:58:10 Join Frampis [0] (
11:01:14 Quit utchybann (Quit: WeeChat 0.3.0)
11:02:41 Quit Farthen (Ping timeout: 256 seconds)
11:06:49 Join komputes [0] (~komputes@ubuntu/member/komputes)
11:10:18 Quit jhulst (Read error: Operation timed out)
11:10:37 Quit PaulJam_ (Ping timeout: 256 seconds)
11:13:37 Join jhulst [0] (
11:25:52 Quit sinuc (Read error: Operation timed out)
11:27:06 Join sinuc [0] (
11:30:19 Join petur2 [0] (~petur@rockbox/developer/petur)
11:30:31 Join n1s [0] (~n1s@rockbox/developer/n1s)
11:31:00 Quit komputes (Ping timeout: 264 seconds)
11:31:29 Quit petur (Ping timeout: 240 seconds)
11:33:58 Nick petur2 is now known as petur (~petur@rockbox/developer/petur)
11:36:01***Saving seen data "./dancer.seen"
11:39:36 Join dfkt [0] (dfkt@unaffiliated/dfkt)
11:43:13 Join komputes [0] (~komputes@ubuntu/member/komputes)
11:57:01 Join mt [0] (~chatzilla@rockbox/developer/mt)
11:57:35 Join LinusN [0] (~linus@rockbox/developer/LinusN)
12:01:14 Join Sajber^ [0] (
12:03:06 Quit komputes (Read error: Operation timed out)
12:04:28 Join Tuplanol1a [0] (
12:18:36 Join komputes [0] (~komputes@ubuntu/member/komputes)
12:19:19 Join einhirn [0] (
12:38:55 Join bmbl [0] (~Miranda@unaffiliated/bmbl)
12:46:41 Nick Tuplanol1a is now known as Tuplis (
12:46:50 Quit Tuplis (Changing host)
12:46:50 Join Tuplis [0] (~jani@unaffiliated/tuplanolla)
13:00:58 Join thegeek [0] (
13:01:10 Quit komputes (Ping timeout: 246 seconds)
13:14:31 Join komputes [0] (~komputes@ubuntu/member/komputes)
13:16:59 Join PaulJam_ [0] (
13:19:17 Join DataGhost [0] (dataghost@unaffiliated/dataghost)
13:24:36 Join bimbel [0] (~Miranda@unaffiliated/bmbl)
13:25:10 Quit bmbl (Disconnected by services)
13:25:18 Nick bimbel is now known as bmbl (~Miranda@unaffiliated/bmbl)
13:28:16 Quit n1s (Ping timeout: 265 seconds)
13:36:04***Saving seen data "./dancer.seen"
13:40:28CIA-8New commit by funman (r24520): Make Clip+ bootloader build ...
13:48:41CIA-8New commit by funman (r24521): Sansa ASM: clock-target.h needs to know the CPU
13:55:09 Join teru [0] (
14:07:11 Join vegtoruci [0] (~vegtoruci@
14:08:26 Join Xerion_ [0] (
14:10:29 Quit Xerion (Ping timeout: 240 seconds)
14:10:29 Nick Xerion_ is now known as Xerion (
14:37:09 Quit Zagor (Quit: Clint excited)
14:37:26 Join froggymana [0] (
14:45:22 Quit Bagder (Quit: It is time to say moo)
14:51:19 Join MethoS- [0] (~clemens@
15:00:12 Quit fyrestorm (Ping timeout: 260 seconds)
15:06:52 Join Casainho [0] (~chatzilla@
15:09:04 Join OnlyByGrace [0] (~4f387f5a@gateway/web/freenode/x-laogyoesrhljsdwi)
15:13:22 Quit OnlyByGrace (Ping timeout: 248 seconds)
15:14:25 Join fyrestorm [0] (
15:14:27 Part LinusN
15:26:10 Quit Zambezi (Ping timeout: 256 seconds)
15:27:43 Join TheSeven [0] (~theseven@rockbox/developer/TheSeven)
15:29:24 Join Zambezi [0] (Zulu@
15:33:12 Join Farthen [0] (
15:34:36 Quit komputes (Ping timeout: 264 seconds)
15:36:08***Saving seen data "./dancer.seen"
15:39:25 Join evilnick_B [0] (~0c140464@rockbox/staff/evilnick)
15:47:16 Join komputes [0] (~komputes@ubuntu/member/komputes)
15:52:04 Quit froggymana (Quit: CGI:IRC)
15:55:23 Quit matsl (Ping timeout: 252 seconds)
15:56:01Torneis there a reason we have seperate drivers for pcf50605 and pcf50606 when they are almost identical and described by the same datasheet
15:56:16Torneespecially since the ipods which use the pcf50605 actually use the ADC, which is only present in the 06 anyway
15:56:18 Join DerPapst [0] (
15:56:25Torne(suggesting that they're actually all 06)
15:56:45 Quit Zambezi (Ping timeout: 272 seconds)
16:00:07linuxstbTorne: I doubt there's a good reason. If you think they can be merged, go ahead.
16:01:29 Quit komputes (Ping timeout: 240 seconds)
16:01:38Tornethe drivers aren't entirely the same, because they incorporate a few target specific things they needn't
16:01:51Tornebut it should be doable, yah
16:02:24CIA-8New commit by teru (r24522): superdom: check range of input in get_number(). stop scrolling text when leave the screen. ...
16:07:10 Quit HBK ()
16:07:49 Join WalkGood [0] (~TubeRider@unaffiliated/walkgood)
16:12:10CIA-8New commit by funman (r24523): Untested attempt at a Clip+ LCD driver using SSP (PrimeCell PL022)
16:14:29 Join swilde [0] (
16:14:41 Join komputes [0] (~komputes@ubuntu/member/komputes)
16:20:36Tornegevaerts: if I ask "how does USB work" am i going to like the answer? :)
16:21:24gevaertsTorne: it entirely depends on what exactly you mean :)
16:21:25Tornegevaerts: i'm wondering about adding a new USB state for "USB charger attached"
16:22:14TorneI think the best way to handle the charging stuff is to drive it from the usb stack, and only communicate simple "charge / don't charge" to the target code
16:22:26Tornebut then the target needs some way to signal to the usb stack that it's detected an actual charger.
16:22:42 Quit Casainho (Quit: ChatZilla 0.9.86 [Firefox 3.6/20100115133306])
16:23:55Tornethis code is all rather excessively ifdef'ed thoguh :)
16:24:12 Join Zambezi [0] (Zulu@
16:24:12 Quit wind (Quit: CGI:IRC (EOF))
16:25:17 Join domonoky [0] (~Domonoky@rockbox/developer/domonoky)
16:25:24gevaertsTorne: I'll answer later. I have to go now
16:25:32Tornei'll keep poking it
16:33:21 Quit komputes (Ping timeout: 248 seconds)
16:38:04 Quit Res1 (Read error: Operation timed out)
16:40:37 Join toffe82 [0] (~chatzilla@
16:40:56 Join perfectdrug [0] (
16:41:54perfectdrugso I'm quite boored at the moment, what can I do for rockbox?
16:43:02linuxstb234 ideas...
16:43:02TorneFix FS #10910 :)
16:43:47 Quit DataGhost (Ping timeout: 252 seconds)
16:44:58perfectdrugnot a developer you know, I would like to do some svgs or something, checking bugreports may also be an option yes
16:45:22Tornechecking bug reports is pretty useful. look for old bugs on a player you have and see if you can reproduce them.
16:48:38perfectdrugbtw is there an option to sort flyspray tasks by last commented, so that you can see which ones are active at the moment
16:48:57pixelmaTorne: I believe there was a small discussions about it just recently and this is still true on the Archos targets (with alarm functionality) and has to stay this way due to hardware initialisation or something
16:49:26pixelmaguess it's not true for other targets then
16:50:20TorneYes, possibly
16:50:25Tornethat's why I didn't just change the manual myself
16:50:32Tornebecause i wasn't sure what to change it to :)
16:50:38Tornei raised the bug instead so someone else might :)
16:55:10 Quit teru (Quit: Quit)
16:56:47gevaertsTorne: isn't the charger presence pretty much independent from the actual usb connection state? I'm not sure uif just adding a state can really work
16:57:24Tornegevaerts: how so? if a usb charger is connected then nothing else can be connected
16:59:18gevaertsTorne: oh, you don't want to combine it with "usb connected,500mA available"? Ok, then it can indeed go in the same state machine
16:59:29Tornei mean as an alternative to USB_POWERED
16:59:56Torneso instead of going from POWERED to INSERTED when it detects a host, it would go POWERED to CHARGER when it detects a charger (e.g. by D+/D- short)
17:00:16Tornebecause POWERED doesn't tell you anything about whether it's safe to draw power
17:00:32gevaertsthat should work
17:00:35 Quit Zambezi (Ping timeout: 245 seconds)
17:00:56Tornewhat does USB_DETECT_BY_DRV imply?
17:01:02Tornethe ifdefs are really confusing :)
17:01:25*gevaerts was afraid that Torne would ask that
17:01:46Tornewell that affects the state machine somewhat
17:02:32 Join Tomis [0] (~Tomis@
17:03:02gevaertsIt allows drivers to detect a connection with other means than polling, e.g. with an insert interrupt
17:03:40gevaertsI didn't write that code, and I must admit that I don't actually know how it works either
17:05:00Tornethe state machine with USB_DETECT_BY_DRV defined seems quite comprehensible
17:05:25Tornewithout it looks weird
17:05:34 Join Tomis2 [0] (~Tomis@
17:07:00 Quit Tomis (Ping timeout: 264 seconds)
17:07:00 Nick Tomis2 is now known as Tomis (~Tomis@
17:07:02gevaertsusb.c could use some love indeed :)
17:07:31gevaertsJust don't break early usb!
17:08:09Tornei do like the way that the status gets ampped between INSERTED and POWERED quite so often
17:08:41Tornehm, so USB_DETECT_BY_DRV is for imx31, pp50xx and jz4740, and everything else isn't? i think?
17:09:08gevaertsI'm not entirely sure
17:09:15Torneit's definately defined for imx31
17:09:21Torneand those are the usb drivers which call usb_status_event..
17:09:46Torneoh wait, USB_STATUS_BY_EVENT is *seperate* from USB_DETECT_BY_DRV
17:09:52*Torne 's head explodes
17:09:52 Join komputes [0] (~komputes@ubuntu/member/komputes)
17:10:07Torneok i get what "by event" vs not means :)
17:10:25 Join Zambezi [0] (Zulu@
17:11:41Torneah, hm. they are currently defined in exactly the same circumstances.
17:11:58Tornelegacy of an older approach maybe?
17:12:09gevaertsnot sure
17:13:25*gevaerts 's main concern back when he worked on usb.c was to not touch the hardware usb logic. That probably explains some of the confusion
17:18:41 Quit petur (Quit: work->bar->home->fosdem)
17:19:23 Quit toffe82 (Read error: Connection reset by peer)
17:24:42 Join casainho [0] (~chatzilla@
17:30:31 Join HBK [0] (
17:32:48TorneAnyway, the charger thing can probably wait for now since we don't have any charger detection code ;)
17:32:58Torneso even if i added a new state we'd never go into it
17:33:41 Join toffe82 [0] (~chatzilla@
17:34:19 Quit HBK (Client Quit)
17:35:27gevaertstrue enough
17:35:44 Join _zic [0] (
17:36:10***Saving seen data "./dancer.seen"
17:36:15TorneTheSeven: Am I right to assume that the USB charging enable setting on nano2g just switches the ipod between being limited to 100mA and limited to 500mA?
17:36:26Torne(as with FS #8802 for the older ipods)
17:36:40Tornei dunno who to ask about h300, though :)
17:37:50Torneanyone got a h300 and a multimeter?
17:37:57gevaertsIIRC on h300 it means just that. It mirrors an OF setting, that IIRC is seriously confusing
17:37:59Torneand, er, a usb cable they can cut in half?
17:39:08*gevaerts has a h300, a multimeter, *and* spare cables, but he won't be near them before monday
17:39:18TorneWell, ther'es no rush
17:39:33Tornei just need to fit the existing h300/nano2g code into what i'm working out here at some point.
17:39:41Tornewell, some point befor ei commit it ;)
17:40:04Torneand it doesn't really say in the source *what* the gpio that usb_charging_enable() pokes actually does
17:40:37gevaertsdoesn't the wiki say more?
17:40:38TheSevenTorne: yes
17:40:48TorneTheSeven: ok, np.
17:40:54TorneTheSeven: i'm probably going to reinvent this interface :)
17:41:09Tornedo you know if there are any other optoins possible on nano2g?
17:41:13Tornee.g. can you stop it drawin gpower at all?
17:41:21 Join HBK [0] (
17:41:24Torneand is there any way to do charger detection? (usb standard or apple)
17:41:45TheSevenTorne: I can set 0/100/500mA
17:41:57Torneooh, you can do 0 as well?
17:42:02TheSevenand i can detect if there is voltage present on the bus
17:42:14TheSevennot sure about detecting chargers (in terms of the usb charging spec)
17:42:27Torneto detect a standard USB charger you would need to know if D+ and D- were connected
17:42:32TheSevenbut I think we can assume that things that don't want us to enumerate are chargers
17:42:39Torneto detect an Apple one you have to use ADCs on D+/D-
17:42:52TorneTheSeven: no, that's not a safe assumption
17:43:19TorneI am intending to rework all this such that the user can choose whether to be safe or not, by changing the usb charging setting to yes/no/force
17:43:33 Join panni_ [0] (
17:43:41Tornefor "force" i will be assuming that it's a charger if we don't get a bus reset
17:43:49Tornebut for "yes" I want to follow specs where possible
17:44:33TheSevenTorne: the hardware default on the nano (e.g. during boot) is 100mA
17:44:43TheSevenso going to 0 doesn't really make that much sense
17:44:43Torneyes, that's normal
17:44:47TorneSure it does
17:45:04Tornenot for the current setting, maybe
17:45:09Tornebut once i've reworked it, i would use 0
17:45:12Torne(for "no")
17:45:41Tornethe use case for setting usb charging to no is "i have my dap attached to my laptop to use it as a disk/whatever, and i don't want to drain the laptop's battery just to charge the dap's"
17:45:54Torneso if you have the ability to stop it drawing current entirely then that's desirable
17:46:05TheSevenTorne: I would be careful about that
17:46:32TheSevenwould it mean rockbox shutting down immediately if the battery is empty, before you have a chance to allow charging?
17:46:33Tornegevaerts: not very more helpful, but thanks :)
17:46:49TorneTheSeven: Hm, that's a good point
17:46:52TorneI hadn't thought of that one
17:46:58Tornebut that can probably be handled.
17:47:02TheSevenmaybe only go to 0mA if the batter is >10%
17:47:05gevaertsTorne: in that case, look at the schematics :)
17:47:31TorneTheSeven: Yeah, I'll make a note to consider that
17:47:49Tornefor now i'm trying to work out what the UI should be and what the usbstack <-> target interface should be
17:49:13Tornegevaerts: the schematics w ehave appear to have been written by linus as well
17:49:19Torneso they only say the same thing :)
17:50:34Torneaha, the datasheet for the battery charger will probably say
17:53:55Torneyeah it's switching VID1 on the lx2201 I think, so it should be selecting 0 vs 500, actually
17:54:40Torne..or maybe not.
17:55:01gevaertshm, 0 could mean different things depending on how things are connected
17:56:07gevaertsIf the thing can be powered without going through the battery, the 0 is only about charging and not about power usage
17:59:45 Join Horscht [0] (~Horscht2@xbmc/user/horscht)
18:01:32 Join phanboy4 [0] (
18:04:02 Quit stoffel_ (Remote host closed the connection)
18:04:53 Quit bmbl (Quit: Bye!)
18:05:12 Join Adub- [0] (
18:06:38pixelmaI thought there was some charger detection code for the PP Ipods back when we didn't have Rockbox USB though. It was there to make it stay in Rockbox and not reboot to the Apple OF but it was a bit buggy and caused Rockbox to hang in the USB screen on normal USB insertion
18:07:02pixelmawith some builds
18:08:10gevaertspixelma: that was actually more an abseense of host detection than a proper charger detection
18:08:28 Join wojtek111 [0] (
18:08:41gevaertsas far as I know anyway
18:08:56 Quit wojtek111 (Client Quit)
18:10:34pixelmapossibly. Was it completely removed though?
18:10:44Torneno, we still have that facility
18:10:51Tornethta's the POWERED state
18:11:14Torne"5V is available but we have not received a bus reset yet"
18:11:48Torneit is not spec compliant to assume that just because you haven't got a bus reset command that the thing you are attached to is a charger :)
18:11:54 Quit Farthen (Quit: ChatZilla 0.9.86 [Firefox 3.6/20100115144158])
18:12:41TheSevenTorne: actually, it is true in 99% of the cases though
18:12:54TheSevenand there is no other way to detect a charger
18:13:04TheSeven(a cheap, non-spec-compliant one, like most are)
18:14:06TorneThat doesn't mean it's not desirable to implement charger detection
18:14:43TorneMy intention is to change the setting to a tristate, off/on/force, where "force" will make exactly that assumption
18:14:50 Join wojtek111 [0] (
18:14:54Tornebut "on" will not, and will follow the spec
18:15:12Torne(so it will charge correctly from an active USB host, but not from a dumb USB charger that can't be positively identified)
18:15:20Tornethen "on" can be the default, beacuse it will be safe. even "on" could lead to "bricked" because of empty battery conditions?
18:15:52TorneTheSeven: er, how?
18:16:05wojtek111I have question but not about rb
18:16:12TheSevenbecause it will shut down before it has a chance to catch the reset if the battery is empty?
18:16:21TorneTheSeven: why would it shut down?
18:16:43Torneyou can draw 100mA as soon as you get +5V on Vcc
18:17:00TheSevenTorne: no, you can't neccessarily
18:17:11TheSevenonly if there is bus traffic
18:17:21TheSeven(if you really want to be spec-compliant)
18:17:38wojtek111does anybody know osx86 at insanelymac?
18:17:47TheSevenso you'd say that "on" will always draw 100mA, even if it can't detect the charger?
18:18:00TheSeventhat's roughly what the OFW is doing
18:18:02Torneaccording to USB-IF compliance stuff you can draw 100mA
18:18:20TheSevenTorne: not if the bus is asleep
18:18:21wojtek111whats is withe their irc channel?
18:18:39TorneTheSeven: nope
18:18:49TorneBattery powered peripherals are still allowed to draw 100mA while the bus is asleep
18:19:11TheSevenoh, you mean they changed that when they introduced that charging spec?
18:19:13wojtek111do you know how update ideneb 10.5.6 to 10.6?
18:19:14 Join rhodan [0] (~quassel@
18:19:24Torneno, this is just usb 2.0 afaik
18:19:39Tornewojtek111: this is off topic, ask somewhere else
18:20:32wojtek111i asked because can't find help anywhere, sorry
18:20:33Horschthm... Torne, is the DMA stuff in current builds yet?
18:20:38TorneHorscht: yes
18:20:46Tornealmost all buffering reads should be using DMA on pp502x
18:20:51Horschtsince when?
18:20:55Tornea week or so?
18:21:06Tornerevisino number should be posted on FS #9708
18:21:47TheSevenTorne: I think i've seen something similar in the usb charging spec, but the usb 2.0 spec clearly says that you may not draw more than a few µA if there is no bus activity
18:22:04TheSeven(which will obviously be the case if you are connected to a dumb charger)
18:22:11TorneAh, yes
18:22:13Tornethat is new
18:22:20Horschtok..., then DMA doesn't help. I am using r24442 and I still get my buffering pause :/
18:22:32Tornehrm, actually it only applies to dead batteries i think :(
18:22:51Torneanyway, i am not that bothered about exactly following the letter of the law
18:23:07 Part wojtek111
18:23:15TheSevenso you'd go for always 100mA, and 500mA if enumerated properly?
18:24:28Torneor if D+/D- are detected as shorted
18:24:34Torneor if the magical apple ADC values appear on D+/D_
18:24:45Tornewhich is not really part of the USB spec but is very unlikely to happen by accident
18:25:03TheSevenTorne: and it's also followed by some other vendors, like for example creative
18:25:10Tornethen the "force" mode would, in addition to that, also switch to 500mA if there was no bus activity within some timeout
18:25:29TorneNot all targets can actually implement that charger detection, though
18:25:40TheSevenand "brute force" would be always 500mA, even if enumerated at 100mA? ;-)
18:25:48TorneWe're not doing that :)
18:26:20TheSevenwhen designing that, you should also consider devices that can draw 1A, like the iPod Nano 4G
18:26:39TheSeven(even though I doubt it will ever draw more than 500mA, even if allowed to)
18:26:41TorneYour point about 0mA needs considering but even if you rule that out (most hardware can't do 0mA anyway), then "no" still has a bonus feature: you can make it not request bus power at enumeration time
18:26:59Torneif usb charging is disabled when plugged in you can only ask for 100mA
18:27:08Tornewhich makes the player work when connected to a bus powered hub (currently impossible)
18:27:27TorneSo yah, i think this plan delivers quite a few potential features for targets which are adapted to support it
18:27:42Torneand just changing the bool to a tristate doesn't make it any more difficult for the user to understand what the hell is going on, i don't think ;)
18:27:56Tornedoes the 4g actually draw more than 500mA?
18:28:04TheSeveni never measured it
18:28:05Tornethe iphone/touch do
18:28:14Tornethey charge at 1A
18:28:21TheSeveni guess it's actually a relic from the itouch...
18:28:33Torneafaik *all* apple chargers are rated at 1A
18:28:38Torneeven ones made long before
18:28:46Torneand afaik all the PP based iPods *can't* do more than 500
18:28:52Torneexcept via firewire.
18:29:09TheSevento my knowledge even most of the non-apple chargers are rated 1A
18:29:18Tornethe USB charging spec allows 1.5A
18:29:25Tornewhich is a weird number but what the hell
18:29:31Torneand the chinese almost-the-same charging spec is 1.2A
18:29:37TorneHooray for standards.
18:29:41Tornethere are so many to choose from.
18:30:06Tornebut anyway. if we had suitable detection then yes, i'd let it draw 1A if possible
18:30:10Tornebut *only* if it was detected
18:30:15Torne"force" would still only do 500mA
18:30:36Tornei'm going for minimal probability of hardware damage even if the user *does* pick "force"
18:31:04 Join mt_ [0] (~mtee@
18:31:10 Quit swilde (Quit: ERC Version 5.3 (IRC client for Emacs))
18:32:37 Quit casainho (Quit: ChatZilla 0.9.86 [Firefox 3.0.16/2009121602])
18:32:41Torneanyway i'm putting a patch for it together, and will show it around before commit, so hey :)
18:38:32 Join DataGhost [0] (
18:38:32 Quit DataGhost (Changing host)
18:38:32 Join DataGhost [0] (dataghost@unaffiliated/dataghost)
18:38:56 Quit perfectdrug (Read error: Connection reset by peer)
18:40:34 Join bertrik [0] (
18:55:55amiconnTorne: Keep in mind that not all targets utilizing the PCF5060x have software usb
18:57:34amiconnAlso there are targets which can charge from USB or from external charger, and the latter can be present in parallel to an usb connection (e.g. H300)
19:02:50Torneamiconn: yes, i know
19:03:03Tornethe pcf5060x comment was unrelated
19:04:33Tornei noticed it a while ago and seeing the h300 charge code just reminded me
19:06:10 Join stoffel [0] (
19:12:41 Nick fxb__ is now known as fxb (
19:16:07 Quit Kitar|st ()
19:23:05 Join Kitar|st [0] (Kitr88@
19:23:50 Quit vegtoruci (Ping timeout: 252 seconds)
19:35:45 Quit pixelma (Ping timeout: 248 seconds)
19:36:12***Saving seen data "./dancer.seen"
19:36:31 Quit amiconn (Ping timeout: 272 seconds)
19:37:36 Quit TheSeven (Quit: ChatZilla 0.9.86 [Firefox 3.5.7/20091221164558])
19:39:35 Join amiconn[h] [0] (
19:48:18 Quit gevaerts (Disconnected by services)
19:48:27 Join gevaerts [0] (~fg@rockbox/developer/gevaerts)
19:55:11 Quit FlynDice (Remote host closed the connection)
19:58:49 Join FlynDice [0] (
20:04:39 Join jgarvey [0] (
20:07:05 Join TheSeven|Mobile [0] (~theseven@rockbox/developer/TheSeven)
20:08:34 Quit TheSeven|Mobile (Remote host closed the connection)
20:08:54 Join TheSeven|Mobile [0] (~theseven@rockbox/developer/TheSeven)
20:09:24 Join moos [0] (moos@rockbox/staff/moos)
20:09:45 Quit TheSeven|Mobile (Read error: Connection reset by peer)
20:11:21 Join TheSeven|Mobile [0] (~theseven@rockbox/developer/TheSeven)
20:12:02 Quit TheSeven|Mobile (Read error: Connection reset by peer)
20:12:20 Join TheSeven|Mobile [0] (~theseven@rockbox/developer/TheSeven)
20:14:16 Quit moos (Ping timeout: 265 seconds)
20:21:06 Quit MethoS- (Remote host closed the connection)
20:33:15 Quit flydutch (Quit: /* empty */)
20:34:03 Quit stoffel (Remote host closed the connection)
20:35:04 Join pixelma [0] (~pixelma@rockbox/staff/pixelma)
20:35:45 Join amiconn[h]_ [0] (
20:38:24 Quit amiconn[h] (Ping timeout: 256 seconds)
20:38:24 Nick amiconn[h]_ is now known as amiconn[h] (
20:39:25 Quit TheSeven|Mobile (Quit: TheSeven|Mobile)
20:47:21 Quit einhirn (Quit: Miranda IM! Smaller, Faster, Easier.
20:50:10bertrikFlynDice, how the clip+ work coming along?
20:50:51FlynDiceI'v'e found most of the buttons, all except pwr & USB and have dual boot working
20:51:42FlynDicefunman got the clip+ bootloader building but when I tried to boot into rockbox I got no response...
20:52:54FlynDiceI'm trying to see if USB is on a pin that starts high on reset and is perhaps pulled down when USB is inserted
20:53:37 Quit DataGhost (Ping timeout: 248 seconds)
20:54:16 Quit CIA-8 (Ping timeout: 256 seconds)
20:54:27 Join CIA-88 [0] (~CIA@
20:54:41FlynDiceall the other buttons simply set the pin they're associated with to high when pressed so that part was pretty simple
20:55:02bertrikI guess it would help to look at the OF
20:56:45 Join TheSeven [0] (~theseven@rockbox/developer/TheSeven)
20:57:26FlynDiceThat's next if these high pins don't work out, unfortunately brute force is going to be quicker for me... ;-)
20:58:12bertrikok, maybe I'll have a go at it this weekend
20:59:12FlynDiceThat would be great!
21:00:24 Quit PaulJam_ (Ping timeout: 264 seconds)
21:05:27JdGordondid you need to do anything special to not brick it?
21:07:02 Join anewuser [0] (anewuser@unaffiliated/anewuser)
21:07:22 Join gigatheme [0] (
21:08:38 Quit gigatheme (Client Quit)
21:09:01FlynDiceJdGordon: Not really, funman gave me some code that was careful to reset any of the registers that were changed
21:09:04 Quit pixelma (Disconnected by services)
21:09:04 Join pixelma_ [0] (~pixelma@rockbox/staff/pixelma)
21:09:12 Join amiconn [0] (~jens@rockbox/developer/amiconn)
21:09:24 Nick pixelma_ is now known as pixelma (~pixelma@rockbox/staff/pixelma)
21:09:38 Join pixelma_ [0] (~pixelma@rockbox/staff/pixelma)
21:10:41FlynDicebut right now I'm just relying on the fact that pins come from reset in an input mode and buttons set the pin high. Simple enough even for me to understand... ;-0
21:13:56bertrikwith the buttons (almost) working, and once the display works, the rest becomes a lot easier to figure out
21:16:58 Quit pixelma (Disconnected by services)
21:16:58 Nick pixelma_ is now known as pixelma (~pixelma@rockbox/staff/pixelma)
21:17:00 Join pixelma2 [0] (quassel@rockbox/staff/pixelma)
21:17:32 Quit amiconn (Disconnected by services)
21:17:35 Join amiconn_ [0] (quassel@rockbox/developer/amiconn)
21:17:57 Nick amiconn_ is now known as amiconn (quassel@rockbox/developer/amiconn)
21:18:04 Quit amiconn[h] (Ping timeout: 256 seconds)
21:22:26 Join moos [0] (moos@rockbox/staff/moos)
21:31:36bertrikfunman's clipsplit doesn't seem to work on the clip+ :/
21:31:54 Join vegtoruci [0] (~vegtoruci@
21:36:13***Saving seen data "./dancer.seen"
21:36:35 Join waran [0] (
21:36:36bertrikhm, the clip+ firmware file mentions AS3525, which is the same SoC as used in the clip v1
21:36:54waranForwarding my question from #~community:
21:36:55waran[21:35] waran: I don't know if this is an appropriate question but... how "stable" is the iPod nano 2G port?
21:36:55waran[21:35] waran: Usable (on the edge) or certainly "not recommended"?
21:37:02AlexPIt is not bad - your FTL might get corrupted from time to time
21:37:13AlexPWhich means keep a back up of your music
21:38:21waranI just wonder which MP3-Player to buy (since my last one, iPod nano 1G, went missing (and stolen, I suppose))
21:38:33AlexPDepends what you want
21:38:38waranActually I thought 1G would be nice... but those are very hard to find nowadays.
21:38:45AlexPI wouldn't get an ipod personally, but that is just me
21:39:13waranApple isn't "my company" as well, but I like the comfortable control via the touch wheel :)
21:39:23AlexPNothing to do with it being Apple
21:39:30waranThe firmware/player is pretty much unusable, thats why I use rockbox after all ;)
21:39:32AlexPIt just isn't very good hardware IMO
21:39:41AlexPAnd touch wheels are awful
21:39:45AlexPagain IMO
21:39:54*evilnick_B seconds AlexP's O
21:40:04waranSo which player would you recommend?
21:40:19AlexPDepends what you want - capacity, recording, radio, ....
21:40:41evilnick_BIt depends what you want to use it for, but if you're after a flash player then probably the Sansa E2x0 series
21:40:53waranjust decent playback quality, moderatly high uptime (5-7h) and around 4GB, preferrably flash, storage.
21:41:24AlexPa sansa then
21:41:25*evilnick_B thinks that the v2 versions are probably better than the v1 nowadays (in terms of RB usage)
21:41:37AlexP(one of the supported ones obviously)
21:41:54AlexPThe ipods are fine, just not my choice
21:42:18waranWell, I'll just check the sansas... thanks :)
21:42:26 Join matsl [0] (
21:42:35AlexPthere is a bit of an issue with the e200 (v1 only?)
21:43:14AlexPWith the right combination of earphones, some people can hear some noice when it is buffering (especially (only?) when from the microsd)
21:43:34AlexPI can't hear it, but again, that's just me
21:43:35JdGordonisnt it only the internal storage?
21:43:56AlexPJdGordon: dionoea tested it yesterday and said only sd for him
21:44:03AlexPSo again, maybe it varies :))
21:44:04JdGordonand touch wheels arn't all that bad
21:44:10AlexPPersonal preference
21:44:14AlexPI dislike them
21:44:35evilnick_BThe v1s have the electrical... oh
21:45:25waranIndeed, personal preference. Hower from what I heard the iPods audio hardware isn't actually high quality (at least thats for the 1G nano).
21:45:39evilnick_BBut that noise is only noticable (in my experience) with decent IEM earphones
21:45:58AlexPwaran: There is nothing inherently wrong with the ipods, I'm just stating my preference :)
21:46:16waranAlexP, they get stolen easily ;)
21:46:23AlexPHowever, it is certain the the nano 2nd gen is less stable with Rockbox currently than some other targets
21:47:17waranI'll bet thats something you (the dev team?) will manage some day using one dirty hack after another :>
21:48:33waranWhat apple could do is to release documentation on "their" hardware, IF they choose to make their default firmware THAT awful.
21:48:44AlexPyeah, good luck
21:48:49AlexPAnd it isn't just apple
21:48:52waranBut thats as likely as Microsoft releasing their next windows open source :)
21:49:53CIA-88New commit by Domonoky (r24524): Add recaptcha to the report function to prevent spam.
21:51:23*bertrik summons funman
21:54:35 Join liar [0] (
21:56:24 Quit pixelma (Quit: .)
21:57:42 Nick pixelma2 is now known as pixelma (quassel@rockbox/staff/pixelma)
21:57:45 Quit anewuser (Ping timeout: 272 seconds)
22:06:13FlynDicebertrik: Does the clip have a buttonlight? Does the pad somehow light up?
22:07:14bertrikit's on GPIO D7 IIRC
22:07:27FlynDiceBummer, not on clip+....
22:08:09bertrikI assume the clip+ has an OLED display too, so no backlight you can turn on, right?
22:08:21ThomasAHFlynDice: one of the + features for me :)
22:08:34 Quit Zambezi (Read error: Connection reset by peer)
22:08:35ThomasAHbertrik: correct
22:09:09 Join Zambezi [0] (Zulu@
22:09:29 Quit kaniini (Ping timeout: 240 seconds)
22:11:29 Join froggyman [0] (
22:12:02 Nick fxb is now known as fxb__ (
22:16:19 Join funman [0] (
22:16:47 Nick funman is now known as Guest65061 (
22:17:05 Nick Guest65061 is now known as namnuf (
22:17:17namnufFlynDice: ah there is no button light on Clip+ ? :(
22:18:12 Join Assembler [0] (
22:18:22FlynDiceNope, just some nice white paint that I'm sure will last forever.....
22:18:39 Join killan_ [0] (
22:19:10namnufI'll try to look if I can find what powers the LCD on, I'm not sure of the SSP setup
22:19:36Assemblerhiya if the fuze has 25 MHz Proc. speed and ~1 Mb of RAM, How about sansa e200?
22:19:46Assembler250 MHz*
22:19:49namnufOnly the lcd_write_register function is correct, I didn't find the clock setup and data write (I supposed 16 bits transfer was ok)
22:20:47FlynDicenamnuf: This is you version of dark sunglasses I suppose..;-)
22:21:40krazykitAssembler, which one? the v1 is a portalplayer device and the v2 is AMS like the fuze
22:21:52Assembleryaaaaaaay! ROCKBOX IS STABLE ON FUZE V1!!!!!!!!!
22:22:01*krazykit bets it's documented on the wiki
22:22:01FlynDiceStill no luck with USB but I've had a bunch of interruptions
22:22:09 Quit killan (Ping timeout: 248 seconds)
22:23:06domonokyAssembler: v2 is like the fuze but has 8mb ram. v1 is pp (dual core 60mhz?) with 32mb ram.
22:23:13namnufYou mean you didn't try all the pins ?
22:23:34namnufbertrik: I could extract the Clip+ firmware with utils/AMS/hacking/amsinfo.c
22:23:46domonokyÄh, fuze also has 8mb ram. only clip and m200 have 2mb.
22:23:59namnufHowever some code was not extracted (the string "Batt." contained in the test code was not dumped to a file)
22:24:40namnufthe 'library' detection code looked like hazardous, perhaps it can be improved
22:24:46Assemblerwait wait im confuzed.
22:24:50FlynDiceNo I tried all the pins but some were steady high, I'm trying to see if USB will pull one of them low. After that I move on to your GPIOB write
22:25:03Assemblerwhat is the processor speed and RAM o fuze v1?
22:25:40namnufAssembler: you said the processor speed yourself, domonoky just told you the RAM amount
22:26:09 Quit _zic (Remote host closed the connection)
22:28:19 Join BHSPitLappy [0] (~BHSPitLap@unaffiliated/bhspitmonkey)
22:33:15Assemblerbut when i got to rockbox info it says: "Skin RAM usage; 28.0KB/9.3MB" doesent it means the fuze has 1MB of ram?
22:36:17AlexPe200v1 is dual 80 MHz, no?
22:37:21Assemblerim using the fuze
22:37:37AlexPYou also asked about the e200
22:37:42AlexP<Assembler> hiya if the fuze has 25 MHz Proc. speed and ~1 Mb of RAM, How about sansa e200?
22:38:00 Quit ender` (Quit: sex is like math: - the clothes, + the bed, / the legs, ...and pray to god you don't multiply.)
22:38:48Assembleri know
22:39:25Assembleroh god i have visited too many posts...
22:40:02AssemblerTMI one said he lemited the RAM to 0.5 MB rockbox says 1 MB im really confused
22:40:21AlexPI am confused by that sentence
22:41:08AssemblerToo Much Infornation. Someone posted in a forum he limited the RAM to 0.5 MB
22:41:33Assemblerrockbot says Skin RAM usage 28\93 MB
22:41:43AlexPI'm sure it doesn't
22:41:49AlexPThat doesn't make any sense
22:41:59TheSevenwish me luck, guys...
22:42:03Assembleri have just checked it on my fuze
22:42:10*TheSeven is about to try flashing the NOR flash on his ipod
22:42:11Assemblergood luck?
22:42:18pixelmaalso, Skin RAM usage doesn't mean much
22:42:19AlexPTheSeven: Good luck :)
22:42:26Assemblerwhat? whats NOR?
22:42:57TheSevenlet's first sign some recovery images, in case i need them...
22:45:04Assemblerhere System>Rockbox Info>Skin RAM usage 28.0/93.0KB
22:45:15Assemblerit was in KB :P
22:45:25AlexPThank you
22:45:36 Join ender` [0] (
22:45:56 Join kaniini [0] (
22:45:57Assemblerwait so it means the e200 has more RAM?
22:46:52AlexPAssembler: Please read the responses you had a few minutes ago last time you asked that
22:48:41Assemblerhay i just got a crazy idea
22:48:46AlexPoh good
22:49:16Assemblerhow about taking a 2GB fuze and adding it a 16 GB microSD and using the 2GB as RAM?
22:49:28AlexPdon't be silly
22:50:31rasherAssembler: flash memory is not fast enough to be used in that way.
22:50:59Assemblerrasher: and a microSD?
22:51:07rasherSame thing
22:51:41Assembleroh brother....
22:51:47Assemblerthe RAM is bothering me
22:51:56Assemblercan i add it RAM is some way?
22:52:19Assemblermaybe deassemble it and reaplace its chip?
22:52:39Assemblerlet it be the intire "system on a chip" how they call it
22:53:47Assemblerjust a random question
22:53:59Assemblerdo you guys ave windows 3.1 source code?
22:54:13AlexPNo, and stay on topic
22:55:08AlexPnot successful?
22:55:20TheSevennot at all
22:56:18AlexPnot dead I hope?
22:56:52TheSeventhe flasher failed - it didn't write anything
22:57:11bertrikFlynDice, there seems to be something on gpio A2
22:58:06FlynDicebertrik: as far as USB?
22:58:08bertrikon other sansas that's the micro sd detect
22:58:33waranthis fuze v1 vs. v2 is mean
22:58:34FlynDiceA2 is one of the pins that was high on reset
22:58:49waranthey look identical O_o
22:58:52bertrikFlynDice, I don't know yet, but it's checked during startup, there even seems to be an interrupt on that pin
22:59:04AlexPwaran: OF version
22:59:25TheSevenoriginal firmware
22:59:43FlynDiceUSB did not pull it to low for me
23:00:33namnufbertrik: you summoned me ? :)
23:03:57bertrikfunman, yes I wanted to know how to adapt clipsplit, but I worked around that
23:03:58FlynDiceThe pins I found high on reset were: A2 B6 B7 C0 C6 D0 D1 D2 D3 D4 D7
23:04:08bertrikFlynDice, I think I found the button read routine
23:05:36 Quit evilnick_B (Quit: Page closed)
23:07:05namnufbertrik: I don't remember what "clipslit" is, you might want to use amsinfo
23:09:15 Quit Assembler (Quit: CGI:IRC)
23:11:43*TheSeven wonders if his code has a bug, or if there is some kind of write protection that he overlooked
23:11:45TheSevenmy code:
23:12:53 Quit Zarggg (Read error: Connection reset by peer)
23:13:00 Join Zarggg_ [0] (~zarggg@2001:0:4137:9e74:0:fbfc:beb1:ba3d)
23:15:15FlynDiceWell USB does not pull any of those pins low, time for a break and then I'll try the write to gpiob
23:15:51TheSevenany obvious errors in there?
23:16:12 Quit shaggy-h (Ping timeout: 240 seconds)
23:19:09TheSevenok, it seems to have something to do with the delay loops
23:19:34bertrikFlynDice, in the button read routine I'm seeing a read of GPIO D6, but before that one is read, GPIO B0 is made high, then a little delay, then the GPIO D6 read, then GPIO B0 is made low
23:20:05namnufTheSeven: looks ok
23:20:30TheSevenif i just wait some thousand iterations, it'll work
23:20:55TheSevenso i probably missed something regarding the protocol (status detection during a write operation)
23:25:00 Quit waran (Quit: Miranda IM! Smaller, Faster, Easier.
23:26:06bertrikFlynDice, I can confirm all of the other GPIO's you already found being read in the button read code
23:27:27 Quit mt (Ping timeout: 240 seconds)
23:28:05 Quit mt_ (Ping timeout: 245 seconds)
23:34:08 Quit jgarvey (Quit: Leaving)
23:34:53bertrikI'm now also seeing lcd stuff being written through SSP, but funman figured that out already
23:36:14***Saving seen data "./dancer.seen"
23:37:43 Quit Llorean (Read error: Connection reset by peer)
23:39:10namnufbertrik: perhaps power button is D6 ?
23:41:30bertrikyes maybe, there's also another pin being made temporarily high while reading D6
23:45:46 Join mt [0] (~chatzilla@
23:47:35TheSevenany nor flash expert around?
23:55:09 Join anewuser [0] (anewuser@unaffiliated/anewuser)
23:57:10 Quit domonoky (Read error: Connection reset by peer)
23:57:28 Join Strife89 [0] (
23:57:34FlynDiceDomestic duties in the way right now ;-) Back to work a bit later!

Previous day | Next day