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 2012-12-14

00:05:55 Quit Wardo (Read error: Connection reset by peer)
00:07:38***Saving seen data "./dancer.seen"
00:08:57 Join [Saint_] [0] (~quassel@rockbox/user/saint)
00:09:08 Quit lebellium (Quit: ChatZilla 0.9.89 [Firefox 18.0/20121205060959])
00:13:59 Quit Rower85 (Quit: Hmmm...)
00:19:09 Quit bertrik (Remote host closed the connection)
00:26:34pamaurylorenzo92: I already found these codes long ago, don't remember where actually. Useful indeed
00:27:51pamaurylorenzo92: thanks for testing sbloader
00:31:30 Quit melmothX (Quit: bau)
00:32:10 Quit ender (Quit: Do not believe any statistic you didn't falsify yourself.)
00:40:09 Quit [Saint_] (Remote host closed the connection)
00:41:42 Join [Saint_] [0] (~quassel@rockbox/user/saint)
00:44:42 Nick dongs_ is now known as dongs (
00:52:23 Quit mc2739 (Ping timeout: 264 seconds)
00:54:16 Join mc2739 [0] (~mc2739@rockbox/developer/mc2739)
00:57:25 Quit yuriks (Ping timeout: 244 seconds)
00:58:56 Quit [Saint_] (Remote host closed the connection)
01:01:04 Join [Saint_] [0] (~quassel@rockbox/user/saint)
01:03:38 Quit pamaury (Ping timeout: 260 seconds)
01:04:59 Quit kevku (Ping timeout: 264 seconds)
01:07:14 Quit einhirn (Quit: Miranda IM! Smaller, Faster, Easier.
01:07:57 Quit [Saint_] (Remote host closed the connection)
01:09:41 Join [Saint_] [0] (~quassel@rockbox/user/saint)
01:10:40 Quit [Saint_] (Remote host closed the connection)
01:16:13 Quit Horschti (Quit: Verlassend)
01:20:27 Quit dfkt (Quit: -= SysReset 2.55=- Sic gorgiamus allos subjectatos nunc.)
01:21:43 Quit prof_wolfff (Ping timeout: 260 seconds)
01:29:01 Quit AlexP (Quit: No Ping reply in 180 seconds.)
01:42:53 Join AlexP [0] (~alex@rockbox/staff/AlexP)
01:43:36 Join Topy44|2 [0] (
01:45:38 Quit Topy44 (Ping timeout: 260 seconds)
02:07:39***Saving seen data "./dancer.seen"
02:16:13 Join Topy44 [0] (
02:17:08 Quit Topy44|2 (Ping timeout: 260 seconds)
02:21:27 Quit kermie (Quit: WeeChat 0.3.7)
02:34:37 Join foo|sh [0] (
02:52:10 Quit Chronon (Read error: Connection reset by peer)
03:22:28 Quit AlexP (Ping timeout: 260 seconds)
03:32:56 Quit mc2739 (Quit: leaving)
03:34:14 Join mc2739 [0] (~mc2739@rockbox/developer/mc2739)
04:03:36 Quit pixelma (Disconnected by services)
04:03:36 Join pixelma_ [0] (pixelma@rockbox/staff/pixelma)
04:03:37 Nick pixelma_ is now known as pixelma (pixelma@rockbox/staff/pixelma)
04:03:45 Quit amiconn (Disconnected by services)
04:03:45 Join amiconn_ [0] (quassel@rockbox/developer/amiconn)
04:03:47 Nick amiconn_ is now known as amiconn (quassel@rockbox/developer/amiconn)
04:05:06 Quit foo|sh (Remote host closed the connection)
04:07:43***Saving seen data "./dancer.seen"
04:14:15 Quit thegeek_ (Read error: Connection reset by peer)
04:14:27 Join thegeek [0] (
04:15:14 Join SirRamsay [0] (bd8ac3c3@gateway/web/freenode/ip.
04:33:26 Quit [Saint] (Remote host closed the connection)
04:34:26 Join [Saint] [0] (~saint@rockbox/user/saint)
04:40:02 Join TheSphinX^ [0] (
04:41:08 Quit SirRamsay (Quit: Page closed)
04:43:32 Nick dongs is now known as _dongs (
04:44:00 Quit TheSphinX_ (Ping timeout: 264 seconds)
05:16:22 Join Rower85 [0] (
05:17:59 Quit TheSeven (Disconnected by services)
05:18:11 Join [7] [0] (~quassel@rockbox/developer/TheSeven)
05:32:07 Quit SuperBrainAK (Quit: pbly going to sleep /_\)
05:39:04 Quit nosa-j (Quit: k)
05:55:15 Join nosa-j [0] (~m00k@
05:55:49 Quit XavierGr (Ping timeout: 256 seconds)
06:02:33 Quit TheSphinX^ (Read error: Operation timed out)
06:07:47***Saving seen data "./dancer.seen"
06:09:37 Join TheSphinX^ [0] (
06:13:22 Quit anewuser ()
06:34:38 Join anewuser [0] (~anewuser@
06:34:38 Quit anewuser (Changing host)
06:34:38 Join anewuser [0] (~anewuser@unaffiliated/anewuser)
07:00:09 Join mortalis [0] (
07:18:07 Join theli_ua [0] (
07:25:41[Saint]Is there a way to know the storage type in Rockbox? Other than internal/external?
07:26:08[Saint]As in, flash vs. an actual spinning platter, for instance.
07:27:31 Join yuriks [0] (~yuriks@opentyrian/developer/yuriks)
07:30:55JdGordonerr, you mean on devices which assume spinning but not?
07:32:08[Saint]Well, yeah, I was thinking about devices which are supposed to be spinning platter, but are frequently modified by the user with a CF or SSD, yet the settings still assume a spinning platter.
07:33:08[Saint]And then they spin out about still having spindown and anti-skip buffer et al.
07:33:54_dongspretty sure last media player with a hard disk in it has bene osboleted more than a decade ago
07:34:07_dongslunix removed support for 386, time to dump support for dead tech here too
07:34:19_dongsmaybe while you're at it, you can fix triggered recording to actually perform as the manual describes
07:34:40[Saint]maybe you could stop being a douche?
07:35:16JdGordon[Saint]: the watermark stuff works off the disc access time (real, not reported) so that shouldnt matter
07:36:21[Saint]I just think it would be nice if the build could adapt to the storage type being changed, and not present those settings.
07:37:16_dongs64gbyte emmc is like $60 in single qtys from places like arrow/digikey
07:37:25_dongswhy would anyone want a spinning disk on a modern media player
07:37:41_dongs(or have more than 64gb of audio media)
07:37:52_dongsyou'd spend more time copying shit to it than you'd ever spend listening to it
07:38:55[Saint]I guess it is fairly safe to consider that if the user has modified their DAP to include an SSD or CF card that they could compile their own build and make the trivial modifications needed to dis-include the needless settings though...
07:39:11Galoisyou don't have to sit there waiting for it to copy
07:39:14_dongsdoes rockbox still require gcc to build?
07:39:19JdGordon[Saint]: hmm, well, if you make a list of the ones to remove we can talk about it
07:39:41_dongsI remember bringing up trying to compile it with a real compiler and seeing immediate performance gains, but I guess noone cared.
07:39:47_dongsthis was back in like 2006
07:56:59 Quit sciopath (Read error: Connection reset by peer)
08:00:12 Join sciopath [0] (
08:05:13 Join Topy44|2 [0] (
08:05:21 Quit Topy44 (Read error: Connection reset by peer)
08:07:41 Join akaWolf [0] (~akaWolf@unaffiliated/akawolf)
08:07:48***Saving seen data "./dancer.seen"
08:26:15 Quit theli_ua (Quit: Залишаю)
08:31:27 Quit brkfstsqd (Remote host closed the connection)
08:41:27 Join wodz [0] (
08:43:16 Join LinusN [0] (
08:44:55 Quit [Saint] (Remote host closed the connection)
08:45:59 Join [Saint] [0] (~saint@rockbox/user/saint)
08:48:34wodz_dongs: Please disclose to us whats the name of the 'real compiler' which will bring us 'immediate performance gains'. Seriously, if you think rockbox is broken and developers are so lazy and stupid please go away. Nothing prevents you from forking and showing the rest of the world how dull current rb team was. Your comments bring nothing but irritating people. Please STOP.
08:49:13 Quit [Saint] (Read error: Connection reset by peer)
08:50:24 Join [Saint] [0] (~saint@rockbox/user/saint)
08:54:17 Join ender [0] (
08:54:44 Join kevku [0] (x@2001:470:dcc6:0:69:69:69:69)
08:56:25 Join melmothX [0] (~melmoth@unaffiliated/melmothx)
09:02:02 Quit wodz (Ping timeout: 276 seconds)
09:21:44 Quit Topy44|2 (Read error: Connection reset by peer)
09:21:50 Quit [Saint] (Remote host closed the connection)
09:24:03 Join [Saint] [0] (~saint@rockbox/user/saint)
09:47:17 Join lebellium [0] (
09:47:34 Join Topy44 [0] (
09:48:16 Join einhirn [0] (~Miranda@2001:638:605:4:8c3a:63a1:8e70:487)
10:02:17 Join dfkt [0] (dfkt@unaffiliated/dfkt)
10:07:51***Saving seen data "./dancer.seen"
10:10:44 Quit Topy44 (Ping timeout: 265 seconds)
10:11:31 Join petur [0] (~petur@rockbox/developer/petur)
10:13:05 Quit yuriks (Ping timeout: 256 seconds)
10:27:07 Join wodz [0] (
10:30:00 Join Topy44 [0] (
10:33:56 Nick funman_ is now known as funman (~fun@rockbox/developer/funman)
10:36:26 Quit Staphylo` (Quit: Bye Bye =))
10:38:24 Join Staphylo [0] (
10:41:46 Join lorenzo92 [0] (~chatzilla@
10:42:31 Quit dfkt (Quit: -= SysReset 2.55=- Sic gorgiamus allos subjectatos nunc.)
10:46:45lorenzo92 pamaury (logs): you're welcome ;)
10:49:38 Quit Barahir_ (Read error: Operation timed out)
10:49:47 Join Barahir [0] (
10:56:24 Join Zagor [0] (
10:56:24 Quit Zagor (Changing host)
10:56:24 Join Zagor [242] (~bjst@rockbox/developer/Zagor)
11:16:50 Join pamaury [0] (
11:16:50 Quit pamaury (Changing host)
11:16:50 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
11:18:37 Quit jhMikeS (Ping timeout: 255 seconds)
11:27:17 Join bluebrother [0] (~dom@rockbox/developer/bluebrother)
11:27:53 Quit TheJacquerie (Ping timeout: 255 seconds)
11:29:42 Quit fs-bluebot (Ping timeout: 245 seconds)
11:30:07 Quit bluebrother^ (Ping timeout: 245 seconds)
11:30:57 Join fs-bluebot [0] (
11:30:58 Quit lorenzo92 (Ping timeout: 260 seconds)
11:36:33 Join TheJacquerie [0] (
11:42:48wodzpamaury: Maybe I found out why bulk doesn't work. bit7 of RXnCON is described as 'Set as one to enable the INTERRUPT OUTPUT generation initiated by the assertion of RXACK'. Both OF and stage2 loader of rk27load set this bit.
11:43:48pamaurycool, will you have time to check this theory today ?
11:44:43wodzI am not sure
11:46:08wodzah, similar situation is for TXnCON.
11:53:11 Join lorenzo92 [0] (~chatzilla@
11:53:57lorenzo92pamaury: okay so I think we have all the tools to start the stmp36xx platform, well, I have lots of ideas in my mind but not sure I know what to do next :D
11:54:39pamaurythe missing tool is elftosb, since it doesn't create sb file for the stmp36xx right now. There is the windows tool but we don't want to use that.
11:55:00pamauryI'll write it soon, we the existing code it should be really fast
11:55:58lorenzo92ah right indeed I was missing this detail :)
11:57:39pamauryIf I were you, my first step would be to write a small tool that you can send to RAM and which has working usb. For example you could modify imxtools/hxemul to run on the stmp36xx (should be easy, the usb core doesn't modifications). When you're done, I send you other tools I wrote to have the screen working (using usb as a logger, very handy). When you have usb and the screen working, it will be much easier
12:00:00pamaurywodz: I'll try to fix usb on rk27xx now, I have a bit of time. Your code is on gerrit ?
12:01:10lorenzo92pamaury: indeed that was my as I probably understood your hwemul doesn't use any other rockbox code right? then yeah I can test this :)
12:01:41pamaurylorenzo92: yes, it's really minimal. It already runs on stmp37xx so that should give you an idea of what needs to be modified
12:02:23pamaurymost probably: pin registers and clock control
12:02:50pamaurybut I don't remember if they are used ^^
12:03:56 Join TheSphinX_ [0] (
12:04:08pamaurylorenzo92: and remove the pwm stuff, it shouldn't be there, that's a mistake of mine
12:05:45lorenzo92ok :)
12:06:07lorenzo92and so for the moment i will use the elftosb 4 windows
12:06:51 Quit TheSphinX^ (Ping timeout: 252 seconds)
12:07:53pamaurylorenzo92: if the tool crashes, tell me, I'll give you some command to run to strip the guilty sections of the executable
12:07:55***Saving seen data "./dancer.seen"
12:07:58wodzpamaury: yes the code is on gerrit
12:09:31wodzare you going to do native port or RaaA?
12:10:21pamaurywodz: do you know if RXACKINTEN needs to be set once or once per transfer ?
12:11:05lorenzo92wodz: native, I hope to be able to help, at least, hehe
12:11:35wodzpamaury: I think once
12:12:54wodzpamaury: utils/rk27utils/rk27load/stage2/*.S has working usb out transfers
12:13:02wodz*bulk out
12:13:41pamaurywodz: by the way, last time I tried, when I plug usb and unplug, the device doesn't detect the unplug
12:14:00wodzyes I know about this
12:14:20wodzI am not quite sure yet how to fix this
12:18:22Raptorspamaury, btw, how did you figure out how to fix the annoying ass static sound on boot?
12:18:53pamauryRaptors: it seems to me that the current HEAD is much better in this regard, have you tried it ?
12:19:12Raptorscurrent head?
12:19:30RaptorsI am running the firmware from the 11th
12:19:36Raptorsand there is no static sound
12:20:07pamauryah, sorry, you question is about how I found out ?
12:20:30Raptorsyou weren't sure how to fix it before
12:21:04RaptorsJust looked at the original firmware?
12:21:22pamauryfirst I did RMAA tests to understand what was wrong. We figured out that some bias was wrong. And then I found out to charge the capacitor without a pop. Many trial and errors, looking at the OF and I finally found a solution without it
12:21:53pamauryI must say I don't completely understand why you have to do that but hey
12:22:52pamaurythere still is a problem with the frequency response if you looked at the RMAA test on the forum but that's less of an issue
12:23:53lorenzo92pamaury: I don't have the device right with me, but seems that pin addresses are pretty the same
12:25:32pamaurywodz: with this bit set, the interrupt is generated for the first bulk out. Then it doesn't seem to work but that's a first step
12:27:29pamauryI'm not sure I understand your code. Why if(ep->cnt > 0) blk_read(...) else usb_core_transfer_complete() ?
12:28:42pamaurywodz: say you do a bulk out and after the first transfer you're done. Then ep->cnt=0 but you won't get an interrupt anymore
12:30:31 Quit lorenzo92 (Ping timeout: 255 seconds)
12:30:52wodzpamaury: I don't have the code to look but it is quite likely I screwed this up
12:32:10 Join philhug_ [0] (
12:37:18_dongssansa clip+ usb was much better in old version
12:37:25_dongscurrent -stable crashes half the time when USB is pluged in.
12:37:37_dongsand for some reason on my system, also kills the hub its connected to.
12:37:56_dongsso i have to pull the hub's power and reconnect.
12:39:24 Join Topy44|2 [0] (
12:41:21wodzpamaury: ok, so in interrupt one should check how much was transfered in this transaction actually, you are right.
12:42:41 Quit Topy44 (Ping timeout: 244 seconds)
12:51:03 Join lorenzo92 [0] (
13:05:10 Nick _dongs is now known as dongs (
13:07:56 Join scrts [0] (~quassel@unaffiliated/scrts)
13:15:33 Quit petur (Ping timeout: 256 seconds)
13:37:48 Quit lorenzo92 (Quit: ChatZilla 0.9.89 [Firefox 17.0.1/20121129165506])
13:49:42 Nick philhug_ is now known as philhug (
13:49:55 Part philhug
14:01:49pamaurywodz: I'm a bit afraid by the dma code of this driver because it doesn't have a way to limit the amount of incoming data. For example if you usb_drv_recv 31 bytes but the host sends 32, it will write one byte past the end of buffer :-/
14:05:18wodzpamaury: true, but I can't see a way to prevent this
14:05:49wodzpamaury: there is no counter in dma regs
14:06:16pamauryI know. By the way I fixed bulk in, it now passes more or less inquiry. Some transfer still fail though
14:06:34wodzgreat. Please put it on gerrit
14:07:59***Saving seen data "./dancer.seen"
14:08:08pamauryI don't quite understand why the host resets the device at some point, strange
14:10:25wodzthat was bothering me too, but I know too little about usb
14:10:52pamaurywodz: do you know if I can push the change on the same gerrit task ?
14:11:19pamauryor I can pastebin the diff, it's very small
14:11:23wodzyes you can. You only need to preserve this ID line
14:11:37wodzpamaury: whatever you like
14:12:40pamauryok, I'll pastebin the diff. I would need to have a look at this with the usb analyzer but my time is up. I think it would be simpler to disable ums and only debug HID, it's a simpler protocol at the beginning. And iirc UMS uses weird stall in his protocol, perhaps the driver doesn't handle it
14:13:30wodzhid doesn't use bulk transfers, no?
14:13:47pamauryno interrupt, but it's exactly as bulk exactly the max packet size
14:14:16pamaury(exactly for the programming point of view, it's the job of the host to ensure at interrupt transfer get the proper bandwidth)
14:14:52pamaurywodz: if you have a doubt about usb, I highly recommand USB in a Nutshell
14:15:29wodzpamaury: I know this document but there is huge difference between 'have idea' and 'deeply understand'
14:15:40wodz*have an idea
14:16:59wodzpamaury: but rk27xx have only interrupt IN endpoints so only half of the chain can be debugged this way
14:17:07pamauryI admit the rk27xx controller is really tricky because the doc is vague
14:17:22pamaurythat's true
14:17:40pamaurythen debug using logf over usb or even a custom driver
14:17:51wodzand misleading - DMA start bit is described as bit0 in one place and as bit1 in other
14:19:45 Quit wodz (Quit: Leaving)
14:27:09 Quit FOAD (Changing host)
14:27:09 Join FOAD [0] (~foad@unaffiliated/foad)
14:43:03 Quit hype (Quit: ["Textual IRC Client:"])
14:56:18 Part LinusN
14:58:09 Quit lebellium (Ping timeout: 244 seconds)
15:01:35 Join nateloaf [0] (
15:03:23 Quit mortalis (Quit: Leaving)
15:07:32 Join the-grue [0] (~gru@
15:08:38 Join lebellium [0] (
15:09:25the-grueHi, I'd like to give opus a try. Which version do I have to install on my sansa clip+? I suppose bugreports go to, right?
15:11:02gevaertsA current development build, and yes
15:20:29the-grueThank you :)
15:20:39 Part the-grue
15:31:27 Join Horscht [0] (
15:31:27 Quit Horscht (Changing host)
15:31:27 Join Horscht [0] (~Horscht@xbmc/user/horscht)
15:34:16 Join jhMikeS [0] (~jethead71@
15:34:16 Quit jhMikeS (Changing host)
15:34:16 Join jhMikeS [0] (~jethead71@rockbox/developer/jhMikeS)
15:44:55 Quit mc2739 (Quit: leaving)
15:45:16 Join XavierGr [0] (XavierGr@rockbox/staff/XavierGr)
15:49:15 Join mc2739 [0] (~mc2739@rockbox/developer/mc2739)
15:58:35 Join y4n [0] (~y4n@unaffiliated/y4ndexx)
16:08:02***Saving seen data "./dancer.seen"
16:08:30pamaurywodz (logs): with my code (slightly modified), it fails because at some point the received CBW (for UMS) has bad signature. I check with the usb analyzer and the signature is correct so there might a problem in the dma setup or something
16:09:56pamaurywodz (logs): by the way, I've found another request handled by the hardware: clear feature (for example, clear stall)
16:10:46 Quit shamus (Quit: chaos reigns within reflect repent and reboot order shall return)
16:11:21pamaurywodz (logs): though it's not clear if it *actually* clear the stall or just notify it
16:18:50pamauryhum, thinking about it now, there might be a problem with the cache if you didn't set the proper defines
16:22:06 Quit einhirn (Ping timeout: 245 seconds)
16:25:06 Join prof_wolfff [0] (
16:25:46 Join shamus [0] (
16:32:17 Nick knitt1 is now known as knittl (~knittl@unaffiliated/knittl)
16:34:17 Quit prof_wolfff (Ping timeout: 264 seconds)
16:37:55pamaurywodz (logs): is there a way to discard a cache entry on the rk27xx ? (and not simply commit and discard)
16:48:59pamaurywodz (logs): funny, if I cache aligned things and commit_discard_idcache() it seems to work much better, but if I only commit_discard_dcache_range(), it doesn't work
16:57:27 Quit the-kyle (Read error: Connection reset by peer)
16:59:29 Quit Zagor (Quit: Clint excited)
17:04:41pamaurywodz (logs): why don't you check for disconnection in CONN_INTR ? it explains why you can't detect unplug !
17:10:14 Join Robin0800 [0] (
17:11:10 Quit jhMikeS (Ping timeout: 260 seconds)
17:13:24 Join the-kyle [0] (~kyle@
17:14:15pamaurywodz (logs): I don't know why but ums reports two devices: sd card and *something* but the something has incredibly high storage size so the hosts send READ CAPACITY (16) which we don't handle and that kills the whole thing
17:15:31 Join Robin0800_ [0] (
17:31:19 Quit Robin0800_ (Remote host closed the connection)
17:43:08 Join theli_ua [0] (
17:46:22 Join n1s [0] (~n1s@rockbox/developer/n1s)
17:47:00 Quit sakax (Ping timeout: 244 seconds)
17:49:10 Quit lebellium (Ping timeout: 244 seconds)
17:50:50 Quit zoktar (Ping timeout: 260 seconds)
17:52:37pamaurygevaerts: in usb_storage.c, in the case of an unhandled command (search "unknown cmd"), shouldn't we stall the bulk endpoints too ?
17:53:12pamauryI checked and linux seems to wait for a response even though the cbw failed
18:00:13 Join sakax [0] (
18:00:26 Join zoktar [0] (~zoktar@unaffiliated/zoktar)
18:08:04***Saving seen data "./dancer.seen"
18:12:14 Join Ward [0] (
18:12:38 Nick Ward is now known as Guest81569 (
18:20:41 Join pretty_function [0] (~sigBART@
18:21:56 Join lebellium [0] (
18:47:31gevaertspamaury: maybe. I seem to remember changing that a few times...
18:47:42pamauryi'm trying to see
18:49:15pamaurythis rk27xx controller is really weird. Handling clear feature in hardware but *not* actually clear stall...
18:49:29pamauryit generates an interrupt but that's just...weird !
18:55:19 Quit lebellium (Ping timeout: 252 seconds)
18:57:03 Join bertrik [0] (~quassel@rockbox/developer/bertrik)
19:01:13 Quit the-kyle (Ping timeout: 256 seconds)
19:13:10 Join the-kyle [0] (~kyle@
19:13:42pamaurywodz (logs): current state:
19:14:16pamauryI can enumerate, report drives but all read/write fail/timeout. I must leave so I don't have any time to investigate the issue but we are getting closer !!
19:27:36 Quit Horscht (Quit: Verlassend)
19:34:37 Quit pretty_function (Remote host closed the connection)
19:42:30 Join lebellium [0] (
19:42:39 Quit lebellium (Client Quit)
19:44:49 Join amayer_ [0] (
19:48:18 Quit guymann (Ping timeout: 265 seconds)
19:50:00 Join guymann [0] (~c@unaffiliated/guymann)
20:08:08***Saving seen data "./dancer.seen"
20:21:51 Join lorenzo92 [0] (
20:22:57lorenzo92pamaury: I wanted to do a quick try compiling your hwemul without many changes, okay got the elf, but then why elftosb crashes with an unhandled exception? can you please guide me I don't really know how to use it
20:29:35 Join SuperBrainAK [0] (
20:41:15 Join wodz [0] (
20:43:48wodzpamaury: Great job! I reacall comment somewhere that dma buffer for udc needs to be double word aligned. Regarding the cache - you cannot simply discard. Only commit+discard is supported. range discard work in lcd driver so I don't know why it doesn't work for you.
20:45:11wodzpamaury: Could it be that 'big' other drive is nand which is not reported correctly? I mean all nand functions are stubs so maybe some uninitialized var?
20:46:41wodzpamaury: Checking (DEV_INFO & VBUS_STS) simply didn't work for me. If it works thats great.
20:47:43 Join solarcloud_3scrn [0] (~solarclou@
21:08:05 Quit theli_ua (Ping timeout: 276 seconds)
21:12:38 Quit ParkerR_ (Ping timeout: 276 seconds)
21:30:55 Join amiconn_ [0] (amiconn@rockbox/developer/amiconn)
21:30:55 Quit amiconn (Disconnected by services)
21:30:58 Nick amiconn_ is now known as amiconn (amiconn@rockbox/developer/amiconn)
21:31:02 Join ParkerR [0] (ParkerR@unaffiliated/parkerr)
21:31:36 Quit wodz (Quit: Leaving)
21:31:50 Quit pixelma (Disconnected by services)
21:31:51 Join pixelma_ [0] (pixelma@rockbox/staff/pixelma)
21:31:53 Nick pixelma_ is now known as pixelma (pixelma@rockbox/staff/pixelma)
21:35:28 Quit solarcloud (Ping timeout: 265 seconds)
21:35:28 Quit lorenzo92 (Quit: ChatZilla 0.9.89 [Firefox 17.0.1/20121129165506])
21:36:08 Join solarcloud [0] (~solarclou@
21:49:22 Quit Guest81569 (*.net *.split)
21:49:22 Quit Topy44|2 (*.net *.split)
21:49:22 Quit Barahir (*.net *.split)
21:49:22 Quit jm (*.net *.split)
22:08:10***Saving seen data "./dancer.seen"
22:17:40 Join AlexP [0] (~alex@rockbox/staff/AlexP)
22:45:10 Quit y4n (Quit: 6,000,000 ways to die — choose one.)
22:50:26 Join Guest81569 [0] (
22:50:26 Join Topy44|2 [0] (
22:50:26 Join Barahir [0] (
22:50:26 Join jm [0] (
22:52:51 Quit akaWolf (Ping timeout: 248 seconds)
23:03:39 Join akaWolf [0] (~akaWolf@unaffiliated/akawolf)
23:05:47pamaurywodz: yeah, sounds like uninitialized var
23:07:21pamaurylorenzo92: you must discard the .ARM.attributes section: arm-elf-eabi-objcopy -R .ARM.attributes in.elf out.elf
23:14:19 Quit melmothX (Quit: bau)
23:19:48 Quit kevku (Quit: KVIrc 4.3.1 Aria
23:31:23pamaurywodz (logs): I tried to disable the nand storage and now it works almost ! For some reason the driver reads the partition table then the FAT and it fails on the first real read. It's peculiar since it's big (64kio) contrary to all the previous ones (1024 bytes). I currently don't have an explaination though because the usb analyzer reports the transfer as successful but the CBW is never sent it seems
23:33:04pamauryhum interesting, actually the host request for more than 64kio but the device only sends 64kio. It seems there is a bug somewhere
23:34:46 Join brkfstsqd [0] (
23:34:59 Join |akaWolf| [0] (~akaWolf@
23:37:53 Quit akaWolf (Ping timeout: 264 seconds)
23:39:02 Nick scorche|1h is now known as scorche|sh (
23:39:10 Quit scorche|sh (Changing host)
23:39:10 Join scorche|sh [0] (~scorche@rockbox/administrator/scorche)
23:41:25pamaurywodz (logs): yeah, my theory is right, it sends a ZLP after 64kio instead of continuing
23:42:54pamauryok, that's because usb_storage split the transfer into two parts and the driver sends a ZLP after the first whereas it probably shouldn't
23:46:05 Join akaWolf [0] (~akaWolf@unaffiliated/akawolf)
23:46:06 Quit |akaWolf| (Read error: Connection reset by peer)
23:48:10 Part amayer_
23:48:59pamaurywodz: it works \o/
23:49:42pamaurywodz: working version:
23:50:09pamauryneeds lot of cleanup. The fix was easy: don't send a ZLP in bulk transfer except if the transfer length is 0 (that was a change from >0 to >=0)
23:50:27pamauryI haven't checked HID, perhaps interrupt transfers need the same fixes as bulk in
23:50:43 Quit akaWolf (Ping timeout: 244 seconds)
23:51:33 Quit n1s (Quit: Ex-Chat)

Previous day | Next day