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 2012-03-23

00:06:26digineton those that have them, does Rockbox have any support for DSP? If not, is the design of RB conducive to adding support?
00:07:07 Quit LambdaCalculus37 (Quit: This computer has gone to sleep)
00:08:10wodzdiginet: could you be more specific?
00:08:56diginetwell, for example some of the SoCS used in various players have DSPs, like TMS320 and DSP56k, does Rockbox have any support for utilizing those?
00:09:26diginetand if not, is there a non-hackish way to add support? (not asking for someone to do it, just curious)
00:09:59gevaertsThe way I understand it, one major problem with hardware DSP is compiler support
00:10:12wodzdiginet: we use DSP in TMS320 in some way. Don't know the details though
00:11:06gevaertsAs far as I know we only use DSPs if things like audio or video are wired through them so we have no choice
00:11:37 Quit lebellium (Quit: ChatZilla [Firefox 12.0/20120314195616])
00:12:59wodzyeah, free compiler are main blockers anyway
00:14:28saratogagenerally speaking, rockbox doesn't need to "support" a DSP
00:14:37saratogaif you can program it, you can run whatever you want on it
00:14:46saratogaregardless of what we support
00:15:25saratogaa programmable DSP is just a CPU, so you can do whatever you want on it
00:17:22wodzHeh, a2a dma engine on rk27xx is sooo inefficient. That partially explains poor sd performance. With dumb 4x ldr + stmia I get over 4 times better throughput for aligned access.
00:22:49 Quit Scromple (Ping timeout: 246 seconds)
00:22:56 Join Scromple [0] (~Simon@
00:23:20saratogahave you tried a battery bench with the new memory timings?
00:23:56wodzsaratoga: I have desoldered battery.
00:24:18wodzbut test_mem shows only tiny difference anyway
00:26:52 Join Llorean1 [0] (
00:26:52 Quit leavittx_ (Remote host closed the connection)
00:28:26 Quit Llorean (Ping timeout: 260 seconds)
00:31:37 Quit wodz (Quit: Leaving)
00:32:55***Saving seen data "./dancer.seen"
00:36:55 Join Llorean [0] (
00:37:31 Quit Llorean1 (Ping timeout: 246 seconds)
00:44:52 Quit bertrik (Ping timeout: 246 seconds)
00:46:33 Join factor [0] (~factor@
00:47:29 Quit TomColler (Remote host closed the connection)
01:04:32 Join bertrik [0] (
01:04:32 Quit bertrik (Changing host)
01:04:32 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
01:04:36 Quit z180 (Ping timeout: 260 seconds)
01:05:31 Quit bertrik (Client Quit)
01:07:20 Quit Llorean (Changing host)
01:07:20 Join Llorean [0] (~DarkkOne@rockbox/user/Llorean)
01:33:21 Quit dfkt (Quit: -= SysReset 2.55=- Sic gorgiamus allos subjectatos nunc.)
01:36:01 Join bitcraft_ [0] (~bitcraft@
01:36:01 Quit bitcraft (Read error: Connection reset by peer)
01:36:01 Quit bitcraft_ (Remote host closed the connection)
01:36:26 Join bitcraft [0] (~bitcraft@
01:48:13 Quit prof_wolfff (Ping timeout: 246 seconds)
02:31:57 Join [Saint_] [0] (~Saint]@unaffiliated/saint/x-8516940)
02:32:59***Saving seen data "./dancer.seen"
02:36:39 Join TBCOOL [0] (
02:48:46 Quit enthdegree (Ping timeout: 245 seconds)
02:54:11 Join [Saint__] [0] (~Saint]@unaffiliated/saint/x-8516940)
02:57:46 Quit [Saint_] (Ping timeout: 260 seconds)
02:59:41 Quit [Saint__] (Remote host closed the connection)
02:59:45 Quit Thra11 (Remote host closed the connection)
03:01:58 Join enthdegree [0] (~BitchX@unaffiliated/enthdegree)
03:11:34 Quit MethoS- (Ping timeout: 265 seconds)
03:33:13 Join bitcraft_ [0] (~bitcraft@
03:33:18 Quit bitcraft_ (Remote host closed the connection)
03:33:20 Quit bitcraft (Read error: Connection reset by peer)
03:33:43 Join bitcraft [0] (~bitcraft@
04:18:27 Join chattr [0] (~mike@
04:18:39 Part chattr ("gone")
04:22:29 Join amiconn_ [0] (amiconn@rockbox/developer/amiconn)
04:22:30 Quit amiconn (Disconnected by services)
04:22:51 Nick amiconn_ is now known as amiconn (amiconn@rockbox/developer/amiconn)
04:22:57 Quit TheSeven (Disconnected by services)
04:23:02 Join [7] [0] (~quassel@rockbox/developer/TheSeven)
04:23:20 Quit pixelma (Disconnected by services)
04:23:22 Join pixelma_ [0] (pixelma@rockbox/staff/pixelma)
04:23:24 Nick pixelma_ is now known as pixelma (pixelma@rockbox/staff/pixelma)
04:26:09 Join LambdaCalculus37 [0] (
04:26:09 Quit LambdaCalculus37 (Changing host)
04:26:09 Join LambdaCalculus37 [0] (~rmenes@rockbox/staff/LambdaCalculus37)
04:33:00***Saving seen data "./dancer.seen"
05:00:12 Quit anewuser ()
05:02:50 Join Rasdere [0] (724d6a5a@gateway/web/freenode/ip.
05:04:55RasdereHey guys, I've recently installed Rockbox on a 2GB clip+ and put a 32GB SDHC card in it; how do I access the SDHC card via usb?
05:07:46 Quit LambdaCalculus37 (Quit: Fwump)
05:29:53 Join bitcraft_ [0] (~bitcraft@
05:29:53 Quit bitcraft (Read error: Connection reset by peer)
05:34:58 Quit ps-auxw (Ping timeout: 260 seconds)
05:41:51 Join ps-auxw [0] (
05:43:24 Quit curtism (Quit: Live Long and Prosper)
05:45:10 Join curtism [0] (~curtis@unaffiliated/programble)
05:47:12 Quit enthdegree (Read error: Connection reset by peer)
05:49:07 Quit curtism (Client Quit)
05:52:12 Join Rob2222 [0] (
05:53:03 Join curtism [0] (
05:53:04 Quit curtism (Changing host)
05:53:04 Join curtism [0] (~curtis@unaffiliated/programble)
05:56:13 Quit Rob2223 (Ping timeout: 246 seconds)
05:56:52 Quit ps-auxw (Quit: leaving)
05:57:56 Join ps-auxw [0] (
05:58:13 Quit curtism (Quit: Live Long and Prosper)
06:01:32 Join curtism [0] (
06:01:33 Quit curtism (Changing host)
06:01:33 Join curtism [0] (~curtis@unaffiliated/programble)
06:03:25 Join Keripo [0] (
06:04:54 Quit nosa-j (Read error: Connection reset by peer)
06:05:41 Join nosa-j [0] (
06:06:46 Quit curtism (Quit: Live Long and Prosper)
06:07:56 Quit nosa-j (Excess Flood)
06:11:05 Join nosa-j [0] (
06:16:38 Join perrikwp [0] (
06:21:14 Quit bitcraft_ (Remote host closed the connection)
06:23:13 Join curtism [0] (
06:23:14 Quit curtism (Changing host)
06:23:14 Join curtism [0] (~curtis@unaffiliated/programble)
06:25:10 Quit curtism (Client Quit)
06:25:38 Join curtism [0] (
06:25:38 Quit curtism (Changing host)
06:25:38 Join curtism [0] (~curtis@unaffiliated/programble)
06:29:55 Quit curtism (Client Quit)
06:30:24 Join curtism [0] (
06:30:24 Quit curtism (Changing host)
06:30:24 Join curtism [0] (~curtis@unaffiliated/programble)
06:33:01***Saving seen data "./dancer.seen"
06:39:56 Quit curtism (Quit: Live Long and Prosper)
06:45:50 Join curtism [0] (
06:45:51 Quit curtism (Changing host)
06:45:51 Join curtism [0] (~curtis@unaffiliated/programble)
07:07:39 Quit Rower85 (Read error: Connection reset by peer)
07:12:07 Join PaulJam [0] (
07:25:54 Quit PaulJam (Ping timeout: 264 seconds)
07:36:15 Quit nosa-j (Excess Flood)
07:37:34 Join nosa-j [0] (
07:40:25funmanRasdere: use a current build
07:46:39RasdereI'm using 3.10, from what I can tell, that is the newest build
07:55:47 Join mortalis [0] (~mortalis@
07:58:38funmannope, current. not last stable
08:21:30 Join bitcraft [0] (~bitcraft@
08:25:59 Quit bitcraft (Ping timeout: 245 seconds)
08:31:08 Quit Scromple (Quit: Leaving)
08:33:03***Saving seen data "./dancer.seen"
08:47:08 Join XavierGr [0] (~xavier@rockbox/staff/XavierGr)
08:47:33 Join ender` [0] (
09:04:37 Join LinusN [0] (
09:20:26 Quit AlexP (Ping timeout: 246 seconds)
09:25:22 Join AlexP [0] (~alex@rockbox/staff/AlexP)
09:28:48 Quit [Saint] (Quit: Quit)
09:29:09 Join [Saint] [0] (~Saint]@
09:29:10 Quit [Saint] (Changing host)
09:29:10 Join [Saint] [0] (~Saint]@unaffiliated/saint/x-8516940)
09:51:34 Part LinusN
09:54:45 Join einhirn [0] (
09:59:04 Join LinusN [0] (
10:06:34 Quit kadoban (Ping timeout: 260 seconds)
10:07:26 Join kadoban [0] (
10:15:18 Join wodz [0] (
10:16:45 Join davo [0] (
10:18:49davotried googling and searching the rockbox forums, but only found info for making plugins, when trying to build the simulator, i'm make gives me: 'rockbox/build-dir/apps/bitmaps/native/rockboxlogo.220x68x16.o: could not read symbols: File in wrong format' collect2: ld returned 1 exit status
10:24:49 Join swilde [0] (
10:24:56wodzdavo: never saw such error
10:25:25davoi'm finding some stuff on other opensource software build logs, maybe it'll help
10:26:31swildehi *, IIRC it is possible to have more than one version of rockbox installed on a device in different directories. How does one boot one of the non default installations?
10:27:49wodzswilde: you can rolo()
10:28:18wodzswilde: I mean while in rockbox select in filebrowser firmware file
10:28:36swildewodz: Ah, thank you!
10:28:40wodzthe resource will be used from .rockbox always though
10:28:51davoit seems it might have something to do with the 'export ARCH=unknown' in the Makefile, if the compiler doesn't know what arch my machine is :\
10:29:07swildewodz: Good to know...
10:29:10 Quit [Saint] (Read error: Connection reset by peer)
10:33:04***Saving seen data "./dancer.seen"
10:50:54 Join lebellium [0] (
10:52:24wodzmortalis: ping
10:53:07wodzmortalis: I think I found the reason of slow sd transfers (at least one of the reason)
10:53:19wodzmortalis: a2a dma engine is dead slow
10:53:59wodzmortalis: dumb 4x ldr + stmia gives ~4 times performance boost for aligned access.
10:54:45wodzat least on my DAP
10:56:54davofound out what the problem with building the simulator was, the Makefile just needed the location of arm-elf-eabi-gcc
10:56:58davoall works now
10:57:04davothanks btw
10:57:20wodzdavo: you should have it in your PATH don't you :-)
10:57:42 Quit factor (Read error: Connection reset by peer)
10:58:21mortaliswodz: what is used in OF for sd transfers?
10:58:23davono, i think manually installed it, but it never got added to my PATH
10:58:31davobut good point, it should be there :)
10:59:17gevaertsdavo: that's definitely not it. You don't need arm-elf-eabi-gcc for the simulator
10:59:56gevaertsThat error looks like you tried to use a directory for building a simulator that had a non-sim build in it without cleaning first though
11:00:24davowell i dunno what else it could be, without it in the Makefile i was getting that build error
11:00:26wodzmortalis: simple loop but I didn't checked if ms compiler do something smart with it.
11:01:01davobut i just watned to state for the record that i resolved it in case anyone had seen my post
11:01:32lebelliumwhat do you think guys, should the FMS display the current preset name even in scan mode?
11:03:28wodzdavo, our dev script gives you a message after successful install that you should add it to your PATH
11:04:19davoah, then that was clearly my mistake not to notice that message
11:04:28davomy bad, and thanks for pointing that out
11:05:40 Quit nosa-j (Read error: Connection reset by peer)
11:07:34 Join nosa-j [0] (
11:09:31 Quit Torne (Ping timeout: 252 seconds)
11:15:41 Join factor [0] (~factor@
11:26:07 Join Torne [0] (~torne@rockbox/developer/Torne)
11:44:08 Join [Saint] [0] (~Saint]@
11:44:08 Quit [Saint] (Changing host)
11:44:08 Join [Saint] [0] (~Saint]@unaffiliated/saint/x-8516940)
11:47:35 Join TomColler [0] (
12:25:06 Join anewuser [0] (~anewuser@
12:25:06 Quit anewuser (Changing host)
12:25:06 Join anewuser [0] (~anewuser@unaffiliated/anewuser)
12:33:07***Saving seen data "./dancer.seen"
12:35:24 Join Topy44 [0] (
12:37:50 Quit Topy (Ping timeout: 246 seconds)
12:41:35 Join Llorean1 [0] (
12:44:38 Quit Llorean (Ping timeout: 244 seconds)
12:49:49 Join T44 [0] (
12:51:00 Quit Topy44 (Read error: Connection reset by peer)
12:51:48wodzTorne: Could you take a look if looks correct? This snippets are going to transfer one sector of data from fixed sd module address to the memory.
12:56:15 Quit T44 (Ping timeout: 276 seconds)
12:56:46 Join Topy44 [0] (
12:58:41Tornenot entirely sure what that's intended to be
12:58:49Tornea fixed-source memcpy?
13:02:08Torneand you want to support efficient writes to unaligned destination buffers
13:05:23Tornei'd honestly suggest not bothering
13:05:36Torneand just aligning all the targets
13:06:22Torneif you're always reading a sector then requiring word alignment is not exactly onerous
13:07:27wodzhow is alignment handled in our storage system then?
13:07:45Tornebadly, generally
13:08:13Tornei would strongly support removing support for non-word-aligned storage accesses in *all* the storage drivers
13:08:19Torneand fixing all the code that breaks :)
13:08:32wodzeek, go for it :-)
13:08:55Tornein fact, in general we want cache line aligned storage buffers where feasible
13:09:03Torneso that DMA is easier
13:09:28Torneanyway, i don't desperately want to sit down and poke at your asm right now to validate it
13:09:39wodzok no problem
13:09:39Torneespecially since i'm supposedly at work :)
13:09:49Torneif you think it's worth having i may look at it later
13:09:55Tornebut i would generally vote that it's not :p
13:09:57Tornesorry :p
13:35:35 Quit nosa-j (Excess Flood)
13:36:34 Join nosa-j [0] (
13:57:13 Quit nosa-j (Ping timeout: 260 seconds)
13:57:34 Join PaulJam [0] (
13:59:04 Join nosa-j [0] (
13:59:57 Nick Llorean1 is now known as Llorean (
14:08:01 Quit perrikwp (Read error: Connection reset by peer)
14:09:15 Join perrikwp [0] (
14:13:32 Join dfkt [0] (~dfkt@unaffiliated/dfkt)
14:23:10 Quit [Saint] (Read error: Connection reset by peer)
14:24:14 Join [Saint] [0] (~Saint]@unaffiliated/saint/x-8516940)
14:24:38 Join Rower85 [0] (
14:33:08***Saving seen data "./dancer.seen"
14:46:34 Quit PaulJam (Ping timeout: 244 seconds)
15:06:18 Join TheLemonMan [0] (
15:07:44diginetthis might be a silly question, but how is the digital audio bitstream generally represented in DAPs? That is, at the elctronic level, how do they "get" the PCM stream to the DAC? Is it I2S?
15:10:28wodzusually i2s but we have targets with memmapped interfaces
15:11:40 Join enthdegree [0] (~BitchX@unaffiliated/enthdegree)
15:11:56diginetdoes Rockbox have to deal with the protocol, or is that handled by hardware?
15:12:06wodzwhat you mean?
15:13:16diginethmm, I guess I'm not sure :P
15:14:00diginetI've been trying to find a suitable SoC for this DIY DAP project, but none of the otherwise suitable ones have I2C
15:15:04diginetThe only one I could find was that SFC5200 chip in the Hx0, but I feel it's kind of old, I'm scared Freescale might discontinue it or something
15:15:23digineterrr, I2S not I2C, sorry
15:16:31wodzHx0 uses scf5249/5250 this is quite different from other members of 5200
15:16:52wodzThere are lots of nice SoCs.
15:17:08diginetsorry, I meant 5250
15:17:34wodzI thought 5249/50 are discontinued
15:17:43diginetI'd kind of prefer to use Coldfire, because I know more about 68k/Coldfire than ARM
15:17:48diginetnope, they're still around
15:18:25diginetI wish they had faster Coldfire+ MCUs, those would be otherwise perfect, but they're too slow, only 50mhz
15:18:28wodzlast time I asked for sample they answered me this is discontinued line of products
15:19:17wodz5249 goes up to ~130MHz (we use 124 in boost mode)
15:20:15wodzIf you want to try something new, there are a few Cortex-M3 chips with external mem bus.
15:20:45diginetthere's also those, although they're BGA. I can do most SMD stuff, not BGA though
15:21:01diginetI know, I'm just not too familiar with ARM
15:21:15wodzarm is quite easy
15:21:39diginetare the ARM and Coldfire branches of rockbox about equal in optimization?
15:23:27diginetI guess I should probably just go with ARM
15:23:40 Join AlexP_ [0] (~alex@rockbox/staff/AlexP)
15:23:52diginetas much as I do like Coldfire
15:23:59diginetI can always use CF for another project
15:24:19 Quit anewuser (Ping timeout: 265 seconds)
15:24:26 Quit AlexP (Ping timeout: 246 seconds)
15:24:36wodzIf you don't wan't to develop everything from the ground up look for SoC already supported in rb.
15:24:52diginetah okay, any suggestions?
15:25:12wodzimx233 is quite recent
15:25:56wodzs3c2440 should be possible to get also
15:28:12wodzWasn't it you to go with softcore?
15:28:20diginetyeah, that was me
15:28:31diginetI decided it was too hard, and impractical
15:28:40diginetI didn't realize how power-hungry FPGAs are
15:29:59wodzwhat a pitty, we will not have SPARC port :/
15:30:33 Quit wodz (Quit: Leaving)
15:30:49gevaertsWhat will we ever do without a SPARC port?
15:31:02diginethahah, not the most practical thing I suppose
15:31:17diginetthe i.MX233 looks good
15:32:03 Join PaulJam [0] (
15:51:10 Part LinusN
15:51:18 Join perrikwp_ [0] (
15:54:15 Quit perrikwp (Ping timeout: 244 seconds)
15:59:45 Join tjb0607_ [0] (c6ed31da@gateway/web/freenode/ip.
15:59:51 Join WalkGood [0] (~4@unaffiliated/walkgood)
16:00:03tjb0607_wow, the game boy emulator's sound engine causes tons of lag
16:01:50 Join maffe [0] (
16:01:54gevaertsemulation isn't exactly a CPU-friendly thing to do, and DAPs tend not to have massive CPUs
16:02:20Tornemost targets are several times too slow to run a gameboy properly under all circumstances
16:02:41tjb0607_the sound just makes it die
16:03:00Tornewell, sound and graphics are what make it die (emulating the cpu is pretty trivial)
16:03:09Tornebut you can't generally turn the graphics off in a game :p
16:03:18Torneso saving that part is not really an option :P
16:11:56tjb0607_I find it kind of funny that Doom runs faster with sound than Pokemon
16:12:17Tornedoom isn't emulated.
16:12:35Torneeven if it were, the graphics hardware in a basic PC of that era is *much* cheaper to emulate than the gameboy's
16:13:17Tornebecause classic pc graphics hardware is a dumb 2d framebuffer (like ours!) - the gameboy's is a hardware sprite and tilemap engine with affine transforms
16:17:09tjb0607_yeah, makes sense
16:17:14 Join y4n [0] (~y4n@unaffiliated/y4ndexx)
16:28:16 Quit enthdegree (Ping timeout: 252 seconds)
16:29:35 Quit Keripo (Quit: Leaving.)
16:29:52 Quit TheLemonMan (Quit: WeeChat 0.3.7)
16:33:10***Saving seen data "./dancer.seen"
16:34:53 Join bitcraft [0] (~bitcraft@
16:37:55 Quit tjb0607_ (Ping timeout: 245 seconds)
16:38:11 Join enthdegree [0] (~BitchX@unaffiliated/enthdegree)
16:42:24 Join vsrinivas [0] (
16:43:44vsrinivashi folks! just started using Rockbox on a sansa clip, am really impressed, particularly w/ Crossfeed!
16:44:25vsrinivasam looking at bringing rockbox's crossfeed to my desktop in-fact :); the basic implementation looks fairly straightforward.
16:44:56vsrinivashas anyone tinkered with more sophisticated (fft-based?) ways of doing it?
16:45:28vsrinivasand for some other things, is there an fft in the rockbox tree?
16:46:34 Join saratoga_ [0] (980329e4@gateway/web/freenode/ip.
16:47:56saratoga_vsrinivas: yes there are FFts available in rockbox
16:48:13saratoga_i don't know how you would use an fft for crossfeed, I don't think it would be useful
16:49:20vsrinivasiiuc you'd want to mix some frequency bands more than others?
16:49:29 Join benedikt93 [0] (~benedikt9@unaffiliated/benedikt93)
16:49:42saratoga_vsrinivas: a filter is a more efficient way to do that
16:49:58saratoga_i believe this is what the current implementation does
16:50:07vsrinivasit is;
16:53:41vsrinivasre: filter −− the filter would definitely be more efficient. but if i want a little equalizer-like-thingy to modify gains on each band, seemed more straightforward w/ fft?;
16:56:32vsrinivasit is surprising how many computrons these little mp3 players have too!
16:57:11saratoga_EQ is generally done with filters as well
16:57:39saratoga_FFTs are problematic because the bands are uniformly spaced and of equal size, which is usually not what you want
16:58:56saratoga_for simple things like adjusting a band's gain, its often easier to just design a simple bandpass, apply the gain, and then mix it back in
16:59:13saratoga_you don't require a sharp edge, so very basic filters will work
17:03:25vsrinivasfair, i was referring to modifying crossfeed gains, but that makes sense for EQ.
17:06:16 Quit maffe (Quit: IRC ist obsolet!)
17:38:45 Quit saratoga_ (Ping timeout: 245 seconds)
17:39:10 Quit Rasdere (Ping timeout: 245 seconds)
17:40:07 Quit perrikwp_ (Read error: Connection reset by peer)
17:41:22 Join perrikwp [0] (
17:45:07 Quit perrikwp (Read error: Connection reset by peer)
17:46:19 Join perrikwp [0] (
17:49:52 Join domonoky [0] (~Domonoky@rockbox/developer/domonoky)
17:49:56 Quit enthdegree (Ping timeout: 248 seconds)
17:53:42 Join dhrasmus [0] (
17:57:35 Quit perrikwp (Read error: Connection reset by peer)
17:58:50 Join perrikwp [0] (
18:19:08 Join enthdegree [0] (~BitchX@unaffiliated/enthdegree)
18:20:49 Quit fs-bluebot (Read error: Connection reset by peer)
18:20:50 Quit bluebrother (Read error: Connection reset by peer)
18:21:00 Join bluebrother [0] (~dom@rockbox/developer/bluebrother)
18:21:09 Quit einhirn (Quit: Miranda IM! Smaller, Faster, Easier.
18:24:02 Join fs-bluebot [0] (
18:24:50 Join anewuser [0] (~anewuser@
18:24:50 Quit anewuser (Changing host)
18:24:50 Join anewuser [0] (~anewuser@unaffiliated/anewuser)
18:33:02diginetdo you guys think that might be a good candidate for a rockbox port?
18:33:11***Saving seen data "./dancer.seen"
18:39:51 Join stoffel [0] (
18:56:18saratogado we have new targets for 3.11? if so, please add them to the release notes
18:56:37saratogai think the plan was to add the c200v2 for instance
18:56:49saratogasince several users have confirmed that the port is working well
18:57:02saratogaand the known issues have been fixed
19:03:50saratogadiginet: we don't currently have any thumb2 processors supported by rockbox, so porting to one might be a lot of work
19:11:34funmanwith the raaa changes it should be possible to disable inline asm though
19:21:03 Quit stoffel (Ping timeout: 272 seconds)
19:24:00diginetsaratoga, thanks! It's so bewildering trying to figure what chip is the best ot use for a given application, there's so many!
19:24:15saratogai recommend one we already support
19:24:17diginetwhat was the last iteration of ARM that wasn't thumb-2?
19:24:45saratogaarmv4/5/6 targets are likely to be easiest, i'm not sure how much changed with thumb2
19:24:48diginetwell, I originally wanted to use the Coldfire in the Hx0, but I think its deprecated
19:25:51diginetI wish Freescale offerred faster versions of their Coldfire+ MCUs, they're perfect save for maxing out at 50mhz
19:26:07saratogai have no idea what the new cortex processors are like in terms of support,, although maybe someone here could tell you
19:26:14saratogathey're so far too new to be in mp3 players
19:26:32diginetsomeone earlier suggested the i.MX233
19:26:43diginetis that in the Mini2440 thing, as mentioned in the Wiki?
19:26:59saratogamaybe, it is used in the fuze+ at least
19:28:53diginetwhat's the newest ARM SoC that Rockbox supports with no more than minor issues?
19:34:19 Join bertrik [0] (
19:34:20 Quit bertrik (Changing host)
19:34:20 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
19:34:46bertrikbluebrother, RockboxUtility will be able to install a clip zip on the time of the release of 3.11, right?
19:36:55saratogaare we releasing for the Zip? I think some stuff is still missing like a theme and such
19:38:20bertrikZip is stable enough for daily use and people can always install a theme themselves
19:38:57saratogayeah but we should at least have a default theme and such
19:39:03funmanmanual too
19:39:28bertrikshould we hold off support in RockboxUtility for that?
19:39:32saratogaif its working but stuff is missing we should make it a stable target but not release IMO
19:40:36 Quit swilde (Remote host closed the connection)
19:41:27funmansoon we'll have one target status per target
19:41:58saratogaalthough wasn't cabbie about ready to go?
19:42:15saratogamaybe we could just commit it before the release
19:43:13bertrikthe cabbiev2 theme doesn't appear to get applied completely if it's used as a builtin theme, it works fine (for me) when selected manually
19:43:35saratogathat was because the font wasn't set in settings_list.c right?
19:44:33bertrikand I think some other problem too, the little icons in front of menu items IIRC
19:45:03saratogathe font at least is dead simple to fix, just add an entry here for 96x96:;a=blob;f=apps/settings_list.c;h=5acebef2a5a64581525d20eb011e75277f581d6a;hb=HEAD#l212
19:49:17 Join TheLemonMan [0] (
19:50:25saratogaapparently i posted a patch for the font change on teh tracker and then forgot about it
19:50:26funmandid we spend like 10 times more time to discuss this font than what we need to fix it ?
19:50:52saratogai think i did fix it actually, although i didn't try the fix
19:51:04funmando you have a zip ?
19:51:13saratogano, but i should have compiled the sim
19:51:20saratogajust busy
20:01:44bertrikI pushed my WIP of FS #12390 to g197 and g198
20:01:45fs-bluebot Sansa Clip Zip Cabbiev2 Port (patches, new)
20:01:46fs-bluebotGerrit review #197 at,197 : FS #12390 - Sansa Clip Zip Cabbiev2 Port by Stephen Carroll by Bertrik Sikken (changes/97/197/1)
20:02:36 Quit TBCOOL (Quit: .)
20:07:58gevaertssaratoga: is the c200v2 installable with rockbox utility?
20:08:12 Quit benedikt93 (Quit: Bye ;))
20:08:32gevaertsfunman: what's the status on the nano2g?
20:10:14funmanusb doesn't work
20:10:46funmanwe might as well just disable USB for 3.11 release
20:11:24 Quit anewuser ()
20:12:37gevaertsCan it reboot to OF disk mode like the PP ipods?
20:12:42gevaerts(...used to do)
20:13:00 Join anewuser [0] (~anewuser@
20:13:01 Quit anewuser (Changing host)
20:13:01 Join anewuser [0] (~anewuser@unaffiliated/anewuser)
20:13:34funmanprobably, emCore has a disk mode entry
20:14:45gevaerts[7]: any ideas about that?
20:16:38saratogagevaerts: hardware-wise its basically identical to the clip but with a color screen, so i assume it can be, although maybe we have to toggle something in the code to enable it
20:17:02saratogawe didn't release it before because we had some issues with the playback engine on low memory, color targets, but those were fixed a while back
20:17:47saratogarbutil does have a clipv2 option in the current release
20:17:51saratogasorry c200v2 option
20:18:00gevaertsAh, that sounds good then
20:18:44gevaertsAlexP_: ping
20:23:23 Quit PaulJam (Ping timeout: 260 seconds)
20:23:28 Join Strife89 [0] (~Strife89@
20:33:15***Saving seen data "./dancer.seen"
20:38:46 Join stoffel [0] (
20:42:53 Quit Strife89 (Quit: Quick reboot)
20:45:00[7]gevaerts, funman: it can, and I think it is actually implemented if USB is disabled
20:45:30gevaertsOK, so if someone can test that and commit it to the 3.11 branch...
20:45:32[7]"diskmodehotstuff\x01\0\0\0" somewhere at the end of IRAM is the signature IIRC
20:46:08 Join prof_wolfff [0] (
20:46:42[7]what's going on here btw? i didn't have time to follow this the last couple of months, and I probably won't have that time for the next month either...
20:46:49[7]did nano2g usb ever work with the new driver?
20:46:54[7]why not just use the old one?
20:46:58[7]what exactly did break here?
20:47:58gevaertsIf the old driver is an option, I'd definitely prefer that
20:50:01funmani don't know if the new driver works, because the old driver was broken by usb.c changes
20:52:06funman[7]: first bad commit is bfd69f2aa19
20:52:33funmana (conflicting) revert didn't work, and i don't have the patch i tried (for the revert)
20:54:09[7]gevaerts: can you give me an executive summary on the usb core changes?
20:56:20funmanand yeah the old driver is a perfect option because the changes didn't add new features (and potentially regressions)
20:58:34 Join Strife89 [0] (~Strife89@
20:59:51funman[7]: usb-s3c6400x.c and usb-s3c6400x.h - both working with bfd69f2aa19^
21:01:15gevaerts[7]: nearly nothing, really. If you have USB_DETECT_BY_REQUEST (which was called USB_DETECT_BY_CORE before) now usb.c decides that the host is there when it sees a transfer completion, instead of usb_core.c when it saw a control transfer. If you don't have USB_DETECT_BY_REQUEST (which IIRC is the case for the nano2g) I don't think anything changed in practice at all
21:01:44gevaertsI'd like to see the nano2g (and classic) move to USB_DETECT_BY_REQUEST, by the way
21:04:24funmani can try a patch for that, i have no idea how to implement it
21:04:43funmanwhen other drivers were converted, I only saw code removal inside usb drivers
21:07:52gevaertsI *think* it's just a matter of changing config.h
21:10:09[7]gevaerts: I tried to move the nano2g to that when it was introduced, and it didn't want to work at the first try
21:10:20[7]and, as usual, I didn't have much time to investigage
21:10:36gevaerts3f709ead is the change for amsv1, and 82259b7a is amsv2
21:10:42[7]i probably just didn't understand correctly how this needs to be implemented
21:11:17[7]so you say bfd69f2aa19 broke it?
21:11:35[7]where can I get a diff of that?
21:11:42*[7] doesn't even have a git account yet
21:11:43funmangit show bfd69f2aa19
21:11:54[7]let alone a clone :)
21:15:22[7]how does that one fail?
21:17:04[7]looks like what has changed is basically when the driver is being started/stopped
21:17:43[7]AFAIK the old nano2g driver reports a spurious bus reset while starting up the usb core
21:18:23[7]and the whole insert/extract magic is controlled by a thread that's polling vbus
21:18:35funmanwhere's that thread?
21:18:54[7]in usb-s3c6400x.c, somewhere at the bottom
21:19:08[7]I suspect that it just isn't started any more because of some change
21:19:26funmani don't see specific mention of a thread
21:20:44 Quit Strife89 (Ping timeout: 272 seconds)
21:20:54[7]oh, wait, I mixed that up
21:21:13[7]in rockbox that thread is somewhere in the usb core, and is calling usb_detect()
21:21:19[7]in emcore, it's in the driver itself
21:22:41 Quit perrikwp (Read error: Connection reset by peer)
21:23:25funman -> charge only, no USB screen
21:23:53 Join perrikwp [0] (
21:24:28funmanah, AMS signals usb plug/unplug from isr
21:26:51funmanUSB_STATUS_BY_EVENT is defined -> usb_tick should poll usb_detect every tick
21:29:10[7]that might help on nano2g then
21:29:24 Join perrikwp_ [0] (
21:30:30 Quit perrikwp (Read error: Operation timed out)
21:30:35funmanbut no USB screen tho
21:32:18funmanusb_enable(1) ain't called
21:32:19 Quit y4n (Quit: Assumption is the mother of all fuckups)
21:35:23funmanusb_detect is called at boot but not anymore
21:37:00 Quit mortalis (Quit: KVIrc 4.1.1 Equilibrium
21:37:54funmanah i need to remove USB_STATUS_BY_EVENT if detection is made by polling
21:42:11funmani still get the same infinite loop with GRSTCTL & 1 == 1
21:48:29 Join kakashi_ [0] (~kakashi_@nltk/kakashi)
21:52:54 Quit domonoky (Quit: Leaving.)
21:54:21 Join Strife89 [0] (~Strife89@
21:55:35kakashi_I have a sansa clip
21:55:43kakashi_the battery seems not to work
21:55:50kakashi_it just suddenly stopped working
21:56:03 Join domonoky [0] (~Domonoky@rockbox/developer/domonoky)
21:59:03 Join perrikwp [0] (
22:01:13 Quit perrikwp_ (Ping timeout: 252 seconds)
22:02:47 Quit perrikwp (Read error: Connection reset by peer)
22:04:16 Join perrikwp [0] (
22:05:17 Part parafin ("If you don't like rock'n'roll | It's too late now")
22:06:14funmanif i ignore the infinite loop, i still see charging only
22:06:29funmaniiuc that means no transfers took place?
22:13:01 Quit perrikwp (Read error: Connection reset by peer)
22:13:17[7]funman: if it gets stuck at that location that definitely means that something is wrong
22:13:22[7]most likely a clock not running
22:14:16 Join perrikwp [0] (
22:14:39funmanso what do you suggest
22:16:01 Quit kakashi_ (Ping timeout: 246 seconds)
22:16:20 Join perrikwp_ [0] (
22:16:50 Join kakashi_ [0] (~kakashi_@nltk/kakashi)
22:17:01[7]short term: revert to the old driver and figure out why that core change broke it
22:17:32[7]long term: double-check everything and bisect between a working driver and the current state
22:18:28[7]i.e. do dumps of all OTG, PHY and SYSCON registers at the point where the lockup happens both with the broken and a working driver, and look for differences
22:18:30 Quit perrikwp (Read error: Operation timed out)
22:19:33funman[7]: i mean please give a diff for 'figure out why that core change broke it
22:19:56funmanor list of registers to watch
22:20:57[7]funman: is usb_detect being called regularly in bfd69f2aa19? if not, why?
22:21:59funmani don't know
22:22:09[7]add some logf() then :)
22:22:36funmani mean in my current code usb_detect works so it's not very interesting
22:23:05funmanusb_reset() is called - and here the core never acks reset
22:23:36 Part WalkGood
22:26:40 Quit dfkt (Quit: -= SysReset 2.55=- Sic gorgiamus allos subjectatos nunc.)
22:30:02[7]funman: I think we might be facing different bugs with those two drivers
22:30:12[7]and I figure getting the old one to work again will be easier
22:30:23funmanthat's what i'm doing ..
22:30:35funmanall your commits that 'fix USB' since last december didn't work for me
22:30:39[7]we should fix the new one as well at some point, but if we aren't even sure whether that did ever work, we don't really have time for that now
22:30:50funmanno, i don't care about the new one either
22:30:55funmani'll do that later
22:31:08[7]so IIUC bfd69f2aa19 did break the old one
22:31:18funmani pasted the old driver i'm using and i gave the exact revision in the irc log i mentioned
22:31:35[7]my next step would be trying to figure out what exactly is going on with it, i.e. *why* it actually broke, because I can't see an obvious reason
22:32:03[7]so I'd try building bfd69f2aa19 with a logf in usb_detect, to check if that one is even called. if not, that might explain the whole trouble. if it is, we'll need to dig deeper
22:33:16***Saving seen data "./dancer.seen"
22:33:58funmanlogf? i use panicf
22:34:23[7]that wouldn't help us if it's called just once
22:34:41[7]for that particular one I'd use logf and check if the log is being spammed with lots of calls to it
22:35:01*funman needs to see how to use logf
22:35:54[7]if nothing has changed since I used it for the last time: #define LOGF_ENABLE #include "logf.h"
22:36:08[7]select advanced, logf in configure
22:36:23[7]and add logf("whatever") calls in the code
22:37:38funmanyes, it's called frequently
22:38:47[7]what's the USB failure behavior with that revision? just not detected at all?
22:38:55[7]or do you at least get a USB screen?
22:41:50funmaninfinite loop
22:42:01funmanin usb_reset(); , so before USB screen
22:42:45 Quit dhrasmus (Quit: Leaving)
22:44:30[7]funman: er, wait... you're trying to tell me that this particular commit caused GRSTCTL & 1 to not be cleared any more?
22:45:27funmanyes, that's what i'm telling
22:46:40[7]am I correct that usb_stack_enable ends up calling usb_drv_init/usb_drv_exit
22:47:29 Quit stoffel (Remote host closed the connection)
22:50:01[7]funman: it never gets past this?
22:50:01[7]while (GRSTCTL & GRSTCTL_csftrst); /* Wait for OTG to ack reset */
22:51:17funmanyes, that's what i'm telling
22:52:25 Quit XavierGr (Ping timeout: 260 seconds)
22:54:07[7]funman: can you dump the following regs immediately before asserting CSFTRST? PWRCON, PWRCONEXT, PCGCCTL, GRSTCTL, INTMSK
22:54:38[7]after that, can you check if the usb irq handler is ever called?
22:58:41[7]and the new driver shows the same behavior?
23:02:37 Quit perrikwp_ (Read error: Connection reset by peer)
23:03:51 Join perrikwp [0] (
23:09:29 Join z180 [0] (
23:10:00 Quit enthdegree (Ping timeout: 252 seconds)
23:11:11 Join LinusN [0] (
23:22:34 Join perrikwp_ [0] (
23:24:41 Join perrikwp__ [0] (
23:24:58 Quit perrikwp (Ping timeout: 246 seconds)
23:26:27 Quit perrikwp_ (Read error: Operation timed out)
23:27:50 Join perrikwp [0] (
23:29:56 Quit Strife89 (Ping timeout: 272 seconds)
23:30:21 Quit perrikwp__ (Ping timeout: 244 seconds)
23:32:12 Quit TheLemonMan (Quit: WeeChat 0.3.7)
23:32:31 Join perrikwp_ [0] (
23:35:18 Quit perrikwp (Ping timeout: 252 seconds)
23:35:34 Join perrikwp [0] (
23:36:22 Quit perrikwp (Read error: Connection reset by peer)
23:37:37 Join perrikwp [0] (
23:38:48 Quit perrikwp_ (Ping timeout: 272 seconds)
23:40:38 Part LinusN
23:44:24bluebrother bertrik: yes, release status is determined by the server, not Rockbox Utility
23:46:22 Quit perrikwp (Read error: Connection reset by peer)
23:47:33 Join perrikwp [0] (
23:48:52 Quit z180 (Ping timeout: 250 seconds)
23:49:34 Join perrikwp_ [0] (
23:52:19 Join cnutor [0] (~none@
23:52:36 Quit perrikwp (Ping timeout: 248 seconds)
23:52:45 Quit cnutor (Client Quit)
23:54:29 Part TomColler
23:56:00 Quit nosa-j (Ping timeout: 260 seconds)
23:57:04 Join nosa-j [0] (

Previous day | Next day