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 2011-11-20

00:01:31*[7] just flashed emcore to an ipod classic via scsi :)
00:01:53PoodlemastahBeastly. What kind of drugs do you take? :P
00:01:55PoodlemastahAwesome work.
00:02:02[7]gevaerts: we'll definitely have to write the firmware partition sequentially :/
00:03:06gevaerts[7]: no way to read back the image from the ipod?
00:03:13*gevaerts guesses not
00:03:44[7]none that we know of at least
00:04:36[7]there could possibly be one if we manage to find the code that's handling this magic 0xc6 scsi request
00:05:10 Join metaphys [0] (
00:05:24 Part metaphys
00:05:28[7]gevaerts: do you know of any magic numbers/strings that are likely to appear in scsi drivers?
00:05:36[7]the best thing I can think of right now is the CBW/CSW magic
00:05:59gevaertsYes, those at least
00:06:27gevaertsIt's operation code c6?
00:07:49*gevaerts wonders why they didn't use WRITE_BUFFER
00:07:57[7]followed by 31/90/91/92/94/96 bytes, which seem to be apple's internal command number
00:08:13[7]it's apple! they don't want others to figure out how this works!
00:08:31[7]I mean, why else would they hide away the firmware partition in the first place?
00:09:03gevaertsWell, they used WRITE_BUFFER to set the clock
00:09:39[7]that's nothing that needs to be hidden
00:09:45[7]why would anyone want to tamper with the clock?
00:09:58[7]oh, nice, there's exactly one occurrence of "USBC" in disk mode
00:10:22 Join Keripo [0] (~Keripo@
00:10:29[7]next to a "USBS" string :)
00:13:08*[7] loves ARM mnemonics :)
00:13:09[7]LDRLSB R0, [R5,#0xC]
00:13:09[7]BICLSS R0, R0, #0x80
00:13:19 Join TheLemonMan [0] (
00:23:39 Quit liar (Quit: hallowed are the ori!)
00:24:18 Quit n1s (Quit: Ex-Chat)
00:25:26 Quit TheLemonMan (Read error: Operation timed out)
00:27:52*[7] found the c6 command handler
00:34:27CIA-22New commit by fredwbauer (r31031): FS #12293 Global default glyph setting in System > Limits > Glyphs To Cache. Defaults to 250. This saves a lot of RAM while still allowing non-English ...
00:37:05CIA-22r31031 build result: 2 errors, 0 warnings (fredwbauer committed)
00:38:24 Quit freddybb (Quit: Leaving)
00:38:46 Join freddybb [0] (
00:43:51[7]command c6 94 is "repartition device", it accepts the new firmware partition size in sectors as its only argument
00:44:49[7]er, no, in kilobytes, not sectors
00:45:20[7]the firmware partition start sector is hardwired to 63, and the data partition immediately follows the firmware partition
00:45:56[7]itunes uses 201326592 bytes for the firmware partition
00:51:24 Quit saratoga (Quit: Page closed)
00:51:55 Join GeekShad1w [0] (
00:55:47 Quit GeekShadow (Ping timeout: 240 seconds)
00:58:31*[7] managed to resize the firmware partition to 131072 bytes using that command, and everything still works
00:58:38[7]so i'll take that as confirmed
01:00:40 Quit robin0800 (Quit: Leaving)
01:01:30***Saving seen data "./dancer.seen"
01:02:01 Quit ender` (Quit: Money is the root of all evil, and man needs roots.)
01:13:04 Quit Urne ()
01:16:03[7]command c6 90 seems to be completely useless. it does nothing besides zeroing some variables, and ignores the parameters that it gets from itunes (the type and size of the image to be uploaded)
01:16:32[7]leaving it out doesn't seem to have any effect
01:21:31[7]command c6 91 writes data to the firmware partition. gets a 16bit sector count argument.
01:22:53 Quit hilbert (Quit: Textual IRC Client:
01:26:05[7]command c6 92 is another one of the useless kind
01:33:48 Quit lebellium (Quit: ChatZilla 0.9.87 [Firefox 9.0/20111116091359])
01:38:14 Join domonoky1 [0] (
01:40:14 Quit domonoky (Ping timeout: 240 seconds)
01:40:36 Join robin0800 [0] (
01:45:52JdGordonflaco98: i have no specific plans, but I've already done a much nier gui for the database which would work nicely, so maybe during the week
01:54:29 Quit domonoky1 (Read error: Connection reset by peer)
01:56:15 Join advcomp2019 [0] (~advcomp20@unaffiliated/advcomp2019)
02:05:52 Quit advcomp2019 (Quit: Ex-Chat)
02:06:29 Join advcomp2019 [0] (
02:06:30 Quit advcomp2019 (Changing host)
02:06:30 Join advcomp2019 [0] (~advcomp20@unaffiliated/advcomp2019)
02:35:39 Quit robin0800 (Ping timeout: 244 seconds)
02:49:54 Quit bertrik (Read error: Operation timed out)
03:01:33***Saving seen data "./dancer.seen"
03:01:53 Join robin0800 [0] (~robin0800@
03:03:32 Quit nosa-j (Ping timeout: 252 seconds)
03:04:08 Join nosa-j [0] (
03:07:10 Quit robin0800 (Ping timeout: 244 seconds)
03:33:49 Quit freddybb (Quit: Leaving)
03:42:56 Quit tchan (Quit: WeeChat 0.3.6)
03:52:11 Join tchan [0] (~tchan@lunar-linux/developer/tchan)
04:09:28 Quit GeekShad1w (Ping timeout: 240 seconds)
04:18:22 Join robin0800 [0] (~robin0800@
04:27:50 Quit amiconn (Disconnected by services)
04:27:50 Join amiconn_ [0] (quassel@rockbox/developer/amiconn)
04:27:50 Quit pixelma (Disconnected by services)
04:27:52 Join pixelma_ [0] (quassel@rockbox/staff/pixelma)
04:27:54 Nick pixelma_ is now known as pixelma (quassel@rockbox/staff/pixelma)
04:28:12 Nick amiconn_ is now known as amiconn (quassel@rockbox/developer/amiconn)
04:38:45 Quit robin0800 (Ping timeout: 244 seconds)
04:48:52 Quit [7] (Disconnected by services)
04:49:08 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
04:57:59 Join robin0800 [0] (~robin0800@
05:01:35***Saving seen data "./dancer.seen"
05:06:26 Join Rob2222 [0] (
05:10:26 Quit Rob2223 (Ping timeout: 245 seconds)
05:21:59 Quit openstandards (Read error: Operation timed out)
05:23:54 Quit kkit (Remote host closed the connection)
05:25:00 Join kkit [0] (
05:26:06 Quit kkit (Remote host closed the connection)
05:28:48 Join JdGord [0] (~AndChat@
05:36:09 Join openstandards [0] (~openstand@
05:45:07 Join kkit [0] (
05:50:39 Quit robin0800 (Ping timeout: 240 seconds)
05:51:45 Join robin0800 [0] (~robin0800@
06:38:59 Join nosa [0] (
06:39:25 Quit Keripo (Quit: Leaving.)
06:39:53 Join Keripo [0] (~Keripo@
06:41:26 Quit nosa-j (Ping timeout: 252 seconds)
06:41:26 Nick nosa is now known as nosa-j (
06:41:49 Quit robin0800 (Quit: Leaving)
07:01:39***Saving seen data "./dancer.seen"
07:10:58 Quit ReimuHak_ (Remote host closed the connection)
07:31:45 Join H140 [0] (
07:33:03H140Hi. How do I set "Show Files" to "All" woth recent builds ?, there's no longer the option in the Quick menu.
07:33:12 Join nosa [0] (
07:35:41 Quit nosa-j (Ping timeout: 276 seconds)
07:35:42 Nick nosa is now known as nosa-j (
07:41:06 Join joey22 [0] (~Joe@
07:44:50 Part joey22
07:46:45H140aha, so that's where it resides. many thanks mc2739
07:48:05 Quit H140 (Quit: CGI:IRC (EOF))
07:50:12 Quit ReimuHakurei (Quit: Baibai.)
07:54:47 Join ReimuHakurei [0] (
08:10:25 Quit JdGord (Ping timeout: 260 seconds)
08:13:06 Join JdGord [0] (~AndChat@
08:16:08 Join ReimuHak_ [0] (
08:32:53 Join Horschti [0] (~Horscht@xbmc/user/horscht)
08:36:30 Quit Horscht (Ping timeout: 258 seconds)
09:01:43***Saving seen data "./dancer.seen"
09:44:51 Join jdgord_ [0] (~AndChat@
09:45:10 Quit JdGord (Read error: Connection reset by peer)
09:47:25 Join liar [0] (
09:51:31 Quit jdgord_ (Quit: Bye)
10:19:06 Join GeekShadow [0] (
10:23:01 Quit Keripo (Quit: Leaving.)
10:29:47 Join Highlander [0] (~Connor_Ma@2a01:e35:2ec2:dcd0:89ca:3aa8:6703:a45f)
10:34:52 Join ender` [0] (
10:39:22 Join TheLemonMan [0] (
10:44:48 Quit Highlander (Quit: ChatZilla 0.9.87 [Firefox 8.0/20111104165243])
10:48:19 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
10:52:33 Quit TheLemonMan (Remote host closed the connection)
11:01:46***Saving seen data "./dancer.seen"
11:02:56 Join TheLemonMan [0] (
11:02:58 Join y4n [0] (y4n@unaffiliated/y4ndexx)
11:09:06 Join Urne [0] (Webmaster@
11:11:40 Join Topy [0] (
11:14:39 Quit T44 (Ping timeout: 240 seconds)
11:25:00 Quit TheLemonMan (Read error: Connection reset by peer)
11:42:33desowinRTC on Sansa Connect seems to be connected somehow through the AVR. Any ideas how to deal with that? I think it's bad idea to put that in drivers/rtc. Would be ok to put the rtc_init, rtc_read_datetime and rtc_write_datetime in avr-sansaconnect.c?
12:22:10 Join pamaury [0] (
12:22:11 Quit pamaury (Changing host)
12:22:11 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
12:42:58 Quit GeekShadow (Remote host closed the connection)
12:43:09 Join GeekShadow [0] (
12:43:12 Join ranma_ [0] (ranma@2a01:4f8:62:2262::2)
12:49:15 Join benedikt93 [0] (~benedikt9@unaffiliated/benedikt93)
13:01:48***Saving seen data "./dancer.seen"
13:05:07 Quit Bagder (Quit: connection reset by beer)
13:08:50amiconnTheSeven: Why do you need to reformat in order to change endianess? Imo it can be done without losing data
13:08:51 Join Bagder [0] (
13:08:52 Quit Bagder (Changing host)
13:08:52 Join Bagder [241] (~daniel@rockbox/developer/bagder)
13:10:16 Quit ranma_ (Remote host closed the connection)
13:10:25TheSevenamiconn: byteswapping 160GB of data would probably take like 10 hours
13:11:36amiconnSure it would take a while, hence there should be a warning screen. It also might be a good idea to require the charger being connected
13:12:18TheSeventbh I think reformatting will be more convenient for the user
13:12:29*amiconn would expect this to be faster though
13:12:51amiconnAlso if the user backs up the data, then reformats and puts the data back, it will probably take longer
13:13:06TheSevenwhy would it? it's basically the same operation
13:13:19amiconnThis also includes reading all data and writing it back, and additionally transferring it twice through usb
13:13:53TheSevenmost users will probably have the data on their PCs anyway
13:14:01amiconnAnd if the disk isn't full, you don't have to byte-swap the whole disk, it's sufficient to swap the used parts
13:14:43amiconn(mbr, bpb, the two fats, and from the fats you know which sectors are used
13:15:21pamauryTheSeven: doesn't the ata controller have a byte swapping setting ? Or perhaps there is a memcpy/swap dma on this device ?
13:16:27amiconnEfficient in-memory byteswap should be easy on arm
13:16:43amiconnDon't know about dma, as that's SoC dependent
13:17:54TheSevenpamaury: I found a way to let the hardware do the byte swapping, yes
13:18:17pamaurythen there no issue anymore ?
13:18:20TheSeventhe bottleneck here is certainly the disk
13:18:45TheSevenpamaury: the issue is that there's a couple of currently rockboxed devices which are formatted with the wrong endian
13:19:10TheSevenso if we want to be compatible with the OF in the future (dualboot might come soon), we need to fix these somehow
13:19:36pamaurylet's say there were badly formattee for the sake of science :)
13:19:52user890104TheSeven: tell them to uninstall/reinstall?
13:20:46user890104they know that uninstalling reformats the hdd, so they have been warned
13:20:50pamauryI believe keeping compatibility with the OF is more important because it doesn't cost anything and is necessaryt o dualboot
13:31:27 Join fyre^OS [0] (
13:34:31 Quit fyrestorm (Ping timeout: 252 seconds)
13:43:49 Quit GeekShadow (Ping timeout: 244 seconds)
13:43:54 Join metaphys [0] (
13:51:51bertrikJdGordon, can it be that the "default:" case in the radio button handling is rarely called? It looks to me that the case ACTION_NONE is usually called, so the tuner driver isn't even asked for a radio event.
13:52:22bertrik(in apps/radio/radio.c around line 690)
13:54:56bertrikPerhaps the check for a radio event should be moved outside of the switch-case statement that checks for buttons
14:25:00 Quit bluebrother (Disconnected by services)
14:25:01 Join bluebrother^ [0] (~dom@rockbox/developer/bluebrother)
14:26:15 Quit fs-bluebot (Ping timeout: 240 seconds)
14:26:43 Part metaphys
14:27:56 Join fs-bluebot [0] (
14:32:50 Join fml [0] (
14:33:24 Quit fml (Client Quit)
14:36:19 Quit martii (*.net *.split)
14:36:19 Quit knittl (*.net *.split)
14:36:19 Quit Slasheri (*.net *.split)
14:36:22 Quit Zoiah (*.net *.split)
14:36:22 Quit JdGordon (*.net *.split)
14:36:22 Quit ender| (*.net *.split)
14:36:25 Join knittl [0] (
14:36:25 Quit knittl (Changing host)
14:36:25 Join knittl [0] (~knittl@unaffiliated/knittl)
14:36:28 Join Zoiah [0] (
14:36:30 Join Slasheri [0] (
14:36:31 Quit Slasheri (Changing host)
14:36:31 Join Slasheri [0] (miipekk@rockbox/developer/Slasheri)
14:36:35 Join JdGordon [0] (
14:36:37 Join martii [0] (
14:36:37 Quit JdGordon (Changing host)
14:36:38 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
14:37:01 Join ender| [0] (~ender1@2a01:260:4094:1:42:42:42:42)
14:39:13TheSevenoh well
14:39:23TheSevenlooks like apple has patched this exploit in some newer bootloader version
14:41:18 Join domonoky [0] (~Domonoky@rockbox/developer/domonoky)
14:42:12 Join mamarley [0] (~mamarley@2001:470:5:84d:960c:6dff:febc:4608)
14:42:56 Part mamarley (" - Chat comfortably. Anywhere.")
14:44:58TheSeventhis means that we'll definitely have to kick out apple's bootloader, and in that case a firmware partition installation doesn't make any sense
14:45:08TheSevenso we're back to DFU + emCORE :/
14:50:38TheSevengevaerts: do you know how exactly those WRITE BUFFER commands are constructed?
14:53:27gevaertsTheSeven: IIRC they're mostly vendor-specific
14:55:40TheSevengevaerts: I mean the apple ones
14:56:33gevaertsThe only thing I know in there is the clock setting. That's documented on (and in usb_storage.c)
15:00:17 Join n1s [0] (
15:00:18 Quit n1s (Changing host)
15:00:18 Join n1s [0] (~n1s@rockbox/developer/n1s)
15:00:41PoodlemastahThat's bad news seven. When was this patched? Versions?
15:01:50***Saving seen data "./dancer.seen"
15:02:46TheSevengevaerts: I'm asking because I've found the code that's processing these packets
15:02:48 Join MethoS [0] (~clemens@
15:03:14 Nick MethoS is now known as Guest20320 (~clemens@
15:03:19TheSevenPoodlemastah: I'm not quite sure, but it might be possible that the bug has never even existed on the 2g/3g classic
15:10:10 Quit zchs (Ping timeout: 252 seconds)
15:16:41TheSevengevaerts: while this seems to be the only WRITE BUFFER command, there's half a dozen READ BUFFER commands
15:18:16 Join hilbert [0] (
15:20:57pamaurybertrik: any idea why si4700_set_frequency would do an infinite loop
15:29:30sinthetekhey, i know this is slightly off-topic, but is it possible plugging devices into the mic port of my pc is causing it to crash?
15:30:05sinthetekseems like the last few times i went to play music through my good speakers that way, my system froze
15:31:37 Join zchs [0] (
15:50:29bertrikpamaury, you mean why the code has an infinite loop, or why the tuner never causes us to leave that loop?
15:50:48pamaurythe latter
15:51:05bertrikprobably because the tuner is powered down
15:52:10bertrikthe radio code asks the tuner to tune even when it has powered down the tuner before, so basically the calling protocol between the radio code and the tuner is unclear
15:52:37bertrikmaybe we should just use a finite number of tries in the tune loop
15:53:06pamauryhum, but then why does it work on other target ?!
15:53:42pamauryyou mean power down using tuner_power or just soft power down ?
15:54:13bertriksoft power down
15:55:45pamaurythey doesn't explain why it happen on the fuze+ and not on other targets
15:58:26bertrikI think we have a hack to not really power it down under some circumstances, maybe that hack is not in place for the fuze+
16:00:01bertrikor, perhaps the reference clock configuration is different
16:00:40bertrikthe beast supplies an external 32 kHz clock to the si4700 IIRC, the sansas uses the si4700 built-in oscillator
16:01:11bertriksee line 34 in firmware/drivers/tuner/si4700.c
16:01:30pamauryyeah, I'm trying to determine what is the setting on the fuze+...
16:02:22 Quit liar (Read error: Connection timed out)
16:09:03 Quit Guinness (Ping timeout: 248 seconds)
16:16:00 Join Guinness [0] (
16:17:26bertrikI'm in favour of removing the infinite loop and adding some reasonable tuning timeout
16:19:19 Quit Urne ()
16:19:38pamaurythe real would be that the code power up the tuner if ask to set the frequency no ?
16:19:40pamaury*real fix
16:20:13pamaurydamn, the OF code makes little sense, it's caches the register content like us but there is something strange
16:21:06 Quit sinthetek (Read error: Connection reset by peer)
16:22:51 Join sinthetek [0] (
16:22:51 Quit sinthetek (Changing host)
16:22:51 Join sinthetek [0] (~sinthetek@unaffiliated/sinthetek)
16:22:52bertrikpamaury, the access methods of the si4700 are strange anyways, reads start at a different register than writes and they both loop around
16:23:42pamauryyes but the writes starts at register 0x2 and the OF seems to start at register 0x8 but I can't be sure, the core is weird
16:27:41bertrikooh, RDS on the FMS now updates almost correctly
16:28:38bertrikscrolling of the long radio text message is reset is any valid RDS data comes in
16:29:17bertrikperhaps we should indicate new RDS data only if it is different from the current data
16:33:03pamauryhum, if I'm not mistake, the OF writes either 0 or 0x81 to the lower byte of TEST1 which doesn't make much sense
16:35:14bertrikIIRC bit 8 enables the internal oscillator
16:36:07pamauryyes but the 1 is reserved
16:36:21pamauryand enabling the internal oscillator doesn't make it work either
16:37:50pamauryhum in fact it does, so apparently it's use an internal oscillator
16:38:20bertrikI remember that getting the internal oscillator to work reliably for all si4700 types and firmware versions was quite tricky
16:38:36pamaurywhy ?
16:39:19bertrikthe different firmware versions react slightly differently and the init sequence for the oscillator is a bit vague in the datasheets and application notes
16:39:52bertrikand it's very hard to test, because I can't test it on all hardware
16:41:19CIA-22New commit by fredwbauer (r31032): Rename global_settings member glyphs to glyphs_to_cache and add descriptive comment.
16:42:37 Join freddybb [0] (
16:43:28 Join lebellium [0] (
16:43:43bertrikShouldn't we save settings before doing a rolo?
16:44:04CIA-22r31032 build result: 2 errors, 0 warnings (fredwbauer committed)
16:44:53bertrikFor example, I change the "Show files" property to properly show me the .sansa file I want to rolo, and then next time the Show Files settings is back at its original value
16:47:07 Join Stummi [0] (~Stummi@rockbox/developer/Stummi)
16:47:32n1sbertrik: yeah sounds like a bug
16:50:04bertrikmaybe not just save settings, but treat it a bit like a shutdown, to also save anything pending in ata idle callbacks
17:01:53***Saving seen data "./dancer.seen"
17:06:57 Quit n1s (Quit: Ex-Chat)
17:12:44 Join oy1r [0] (~Reggy@
17:13:09oy1rq: how do i make/play a playlist in rockbox ?
17:20:12gevaertsThere's quite a lot of information about playlists in the manual
17:20:42oy1ri bet there is.
17:21:28 Join GeekShadow [0] (
17:22:33bertrikpamaury, how is the tuner connected to the SoC in the fuze+, does it use a real i2c bus, or some bit-banged i2c?
17:25:34lebelliumbertrik: The FM UI is ready now: Just wait for you to implement the RDS feature^^
17:25:59 Quit GeekShadow (Ping timeout: 244 seconds)
17:26:35bertriklebellium, do you want an experimental .zip to test it? :)
17:26:43 Join funman [0] (~fun@rockbox/developer/funman)
17:27:05lebelliumI test whatever you want as long as it doesn't brick my device^^
17:30:26 Join herefornow [0] (~Joe@
17:31:49bertriklebellium, try this
17:32:35lebelliumok thx
17:33:03 Part herefornow
17:38:20oy1rq: does it matter where the audio resides ?
17:38:42oy1ri just made a folder named music up move my files to it.
17:41:05gevaertsrockbox doesn't care about how you want to organise your files
17:42:28lebelliumbertrik yeah the RDS works great hehe but I don't understand why it displays "2345678" in the left top corner. And strange things instead of my repeat mode tags. I don't think that comes from my theme
17:44:33lebelliumbertrik: ah understood! it displays the virtual keyboard in the top left corner Oo
17:50:15 Quit ej0rge_ (Read error: Operation timed out)
17:54:20 Join ej0rge [0] (~alhaz@
17:56:55lebelliumvery strange bug, any idea? But the RDS works perfectly here in Berlin, both the RDS name and text
17:58:34bertrikno, no idea, I get that 234567 thing too
17:59:23bertrikI sometimes get a "mixed message" from RDS, combining the previous RDS text with new test
18:00:32bertrikthe scrolling of the radio text resets when a new station name comes in, which is a bit annoying (and especially the commercial stations change their station name a lot to show all kinds of other info)
18:03:47 Join liar [0] (
18:04:42 Join dreamlayers [0] (
18:04:42 Quit dreamlayers (Changing host)
18:04:42 Join dreamlayers [0] (~dreamlaye@rockbox/developer/dreamlayers)
18:06:52dreamlayersAre there any known toolchain bugs related to CACHEALIGN_ATTR initialized data?
18:15:09lebelliumbertrik: you tried with my theme or another one?
18:16:06dreamlayersIf says some variable is at a certain address, then I'm supposed to find it at that file offset in rockbox.bin, right?
18:20:04funmanyou'll find it in rockbox.elf disassembly
18:20:43funmanbertrik: i'm considering asking the rockbox fund to sponsor me a Clip Zip
18:28:43 Join GeekShadow [0] (
18:38:45TheSevenhm, where can i get some scratch space in storage_init() to probe HDD data endianness?
18:38:50TheSevenI'd need 4KB of RAM for that
18:40:38 Join Urne [0] (
18:42:41funmanTheSeven: INIT_ATTR or something?
18:43:10TheSeveni had hoped to be able to just steal some (plugin?) buffer
18:44:23gevaertsCan't you get some RAM from bflib and give it back later?
18:44:39funmanhm storage_init() is called in usb mode as well
18:45:05funmanTheSeven: INIT_ATTR is in plugin or codec buffer (if defined for your target)
18:46:05TheSevenwhy would storage_init() be called more than once?
18:46:39funmanbecause storage is stopped completely
18:47:02funmani guess this only applies to model where USB storage mode is not controlled by software
18:47:25TheSevenso it doesn't apply here
18:47:42TheSevenand in storage_init I should be safe to steal fat_get_sector_buffer() :)
18:49:50CIA-22New commit by fredwbauer (r31033): Fix parameter order in font_load_in_memory(). FS #12395.
18:51:59CIA-22r31033 build result: 2 errors, 0 warnings (fredwbauer committed)
18:57:27CIA-22New commit by funman (r31034): Fix typo in comment
18:59:44CIA-22r31034 build result: 2 errors, 0 warnings (funman committed)
19:01:54***Saving seen data "./dancer.seen"
19:03:41CIA-22New commit by fredwbauer (r31035): Update filetree.c to use global_settings.glyphs_to_cache
19:05:56CIA-22r31035 build result: 2 errors, 0 warnings (fredwbauer committed)
19:06:59 Quit GeekShadow (Remote host closed the connection)
19:07:11 Join GeekShadow [0] (
19:07:24dreamlayersSo, objcopy -O binary outputs sections in the order they are in the .elf file, without caring about their addresses and alignment. Rockbox moves sections, but only a few targets move ".data". If a target doesn't move it, and it's more than word aligned, ".data" ends up at the wrong place.
19:08:59TheSevenmake: *** No rule to make target `/data/micha/dev/rockbox-trunk/apps/gui/skin_engine/skin_fonts.h', needed by `/data/micha/dev/rockbox-trunk/build/classic-app/apps/bookmark.o'. Stop.
19:10:10 Quit ReimuHak_ (Remote host closed the connection)
19:11:38 Join ReimuHak_ [0] (
19:12:17freddybbTheSeven: make reconf ; make clean ?
19:12:19dreamlayersMaybe that's an old dependency, and "make dep" will make the error go away?
19:12:39TheSeventrying that right now
19:13:24*TheSeven is used to the freemyipod project's makefiles which generate dependencies on the fly
19:24:58funmanadding INIT_ATTR to storage_init() would be an #ifdef hell
19:25:18funman!defined HAVE_USBSTACK && !defined(SIMULATOR) && !defined(USB_NONE) && defined(USB_FULL_INIT)
19:25:28funmanUSB_FULL_INIT being defined only in usb.c
19:27:01funman!defined HAVE_USBSTACK && !defined(SIMULATOR) && !defined(USB_NONE) && !defined(BOOTLOADER) looks better though
19:30:18funman#if defined HAVE_USBSTACK || defined(USB_NONE)\n INIT_ATTR\n #endif
19:40:19 Quit ender` (Quit: Saying time travel is impossible is a really safe bet. If it turns out time travel is possible, they can always go back in time and change their statement. But since they haven't changed their statement then that means time travel is impossible.)
19:41:48 Quit hilbert (Ping timeout: 240 seconds)
19:44:41CIA-22New commit by funman (r31036): Add INIT_ATTR to storage_init() when possible
19:45:20 Join saratoga [0] (9803c31c@gateway/web/freenode/ip.
19:46:52CIA-22r31036 build result: 2 errors, 0 warnings (funman committed)
19:47:31 Quit openstandards (Ping timeout: 245 seconds)
19:47:38funmanbertrik: on which parts of clip zip port could i help except for USB?
19:48:55funman(also yay, some green!)
19:49:02 Join hilbert [0] (
19:49:42bertrikplugins (about 8 not working yet IIRC), graphics, help to finish RDS (nearly done), and USB :)
19:49:54funmanalso video playback
19:50:13funmanBagder: do you handle rockbox fund too or is it only Zagor?
19:50:21bertrikMaybe if you get one with the wisechip display, you could look closer at the problems we have with that display (horizontal position in lcd_update_rect seems incorrect)
19:50:32Bagderfunman: zagor only
19:50:55saratogafunman: maybe voltage/frequency scaling ?
19:51:19bertriksaratoga, yes, but that's something that applies to the entire AMSv2 range I think
19:51:31funmansaratoga: hm yes that'd be nice to compare clip zip to other models we had
19:51:54bertrikI haven't done a battery benchmark on the clip zip yet, but it feels similar to the clip+
19:51:55 Quit dreamlayers (Quit: dreamlayers)
19:52:10saratogapeople have compared mp3 to ogg (which is faster) and showed very little difference in battery life, so if we could get frequency scaling working that would probably help
19:52:18bertrikThe clip zip currently uses the battery curve of the clip+
19:54:00bertrikI'm a bit confused regarding the whole clocking/voltage thing on AMSv1 and AMSv2 to be honest
19:54:21funmanbertrik: what's confusing on AMSv1?
19:54:24bertriksaratoga, I declare you the expert on that now :)
19:54:42saratogayeah i don't really understand that either
19:54:52saratogaPCLK seems to have a HUGE impact on battery life, but FCLK only a small one
19:54:55saratogawhich makes no sense
19:55:07bertrikwell, I was confusing the voltage settings from AMSv1 with AMSv2
19:55:36bertriksaratoga, and FCLK has also relatively small impact on performance, right?
19:55:57saratogabertrik: FCLK is the ARM processor, so its fairly important :)
19:56:09saratogabut PCLK also matter a lot because of IRAM if thats what you mean
19:56:49bertrikthat also confuses me, the as3525 talks of 1T-RAM, is that what you call IRAM?
19:57:32saratogaInternal RAM
19:58:10saratogadepending on how well cache works for codecs, PCLK either matters a lot or very little
19:59:06 Quit freddybb (Quit: Leaving)
19:59:37saratogai tested low FCLK and slightly higher PCLK and it actually reduced battery life a little
19:59:58saratogai guess the main CPU is pretty good at saving idle power, whereas whatever uses the PCLK less so
20:00:07saratogai still want to try lower cpu voltage too though
20:00:32 Join openstandards [0] (~openstand@
20:01:02bertrikI think I still have a patch somewhere to show the extended ADC values (which includes the processor clock values CVDD1 and CVDD2)
20:01:10saratogaoh neat
20:01:27saratogai wonder if its something stupid like the IRAM is run off the cpu voltage or something and that makes it use a lot of power
20:02:02saratogai went through the clock enable registers and as far as I can tell we have everything we don't need disabled
20:02:14bertrikI'm pretty sure now that we're using the same voltages as the OF (not claiming the OF does it in the absolutely optimal way)
20:02:16saratogaalthough we do leave the display stuff clocked when the screen is off AFAICT
20:02:28saratogawell we beat the OF on the Clip+
20:02:59saratogai was thinking that maybe the display controller uses more power with a faster PCLK, as far as I know its driven off of that
20:12:28 Join nosa [0] (
20:12:47 Quit nosa-j (Ping timeout: 248 seconds)
20:12:47 Nick nosa is now known as nosa-j (
20:13:04funmansaratoga: display clock didn't matter for clip*
20:13:22saratogait uses the SSP orwhatever?
20:13:24 Join TomColler [0] (
20:13:52funmanonly clip+
20:14:00funmanand zip probably
20:15:16bertrikzip uses SSP too indeed
20:15:39saratogaIIRC it was just initialized at startup and then left running when the screen was off
20:15:41saratogano idea if that matters
20:16:14 Join ender` [0] (
20:16:46bertrikI can imagine that the SSP turns off its external clock to the display when there's nothing to send/receive
20:17:52bertrik(if it's a standard ARM peripheral we should be able to look that up)
20:18:18saratogai just wondered if it was using some extra power when we increased PCLK
20:28:27 Join robin0800 [0] (~robin0800@
20:31:09*GodEater seems to have a broken tool chain =/
20:31:15GodEatercan't build android client
20:32:16GodEaterhas anyone successfully built with the latest ndk ?
20:32:59GodEaterI'm getting lots of : undefined reference to `__aeabi_ldivmod'
20:33:42 Quit bzed (Ping timeout: 258 seconds)
20:33:46 Join GermanMushroom [0] (
20:34:48pamaurybertrik: i2c using bit banging, the real i2c bus is used for the touchpad, is this important ?
20:34:50funmansaratoga: #11330
20:34:55funmansaratoga: FS #11330
20:34:56fs-bluebot Clip*: disable peripherals in lcd_enable() (patches, closed)
20:35:24saratogafunman: thanks
20:35:37saratogai should try that at higher PCLK and see if it helps at all
20:36:03 Join ChickeNES-laptop [0] (~ChickeNES@
20:46:21 Quit GeekShadow (Ping timeout: 244 seconds)
20:46:39 Quit y4n (Quit: PÆNTS ØLF!)
20:48:54 Quit robin0800 (Quit: Leaving)
20:50:37 Join robin0800 [0] (
20:52:04 Join TheLemonMan [0] (
21:01:58***Saving seen data "./dancer.seen"
21:09:50 Join bluefoxx [0] (
21:10:16 Quit ChickeNES-laptop (Quit: Computer has gone to sleep.)
21:10:23 Join GeekShadow [0] (
21:11:17 Quit bluefoxx (Read error: Connection reset by peer)
21:14:36 Join bluefoxx [0] (
21:15:17 Join bzed [0] (
21:15:20 Quit robin0800 (Read error: Connection timed out)
21:18:59 Join robin0800 [0] (
21:34:40 Part TomColler
21:36:34 Quit robin0800 (Ping timeout: 244 seconds)
21:36:37 Quit GermanMushroom (Quit: Ik ga weg)
21:36:53 Quit Stummi (Quit: Bye!)
21:39:00 Join NoPinky [0] (
21:39:10NoPinkyhello people
21:41:20 Quit benedikt93 (Quit: Bye ;))
21:48:36 Join GeekShad1w [0] (
21:52:29 Quit GeekShadow (Ping timeout: 244 seconds)
21:53:32 Join robin0800 [0] (
22:07:11soapmy bad
22:19:10 Quit NoPinky (Remote host closed the connection)
22:19:20 Join NoPinky [0] (
22:37:58 Part domonoky
22:42:30lebelliumas the Zip port is considered as "unstable" what does it keem from getting a theme page?
22:43:03lebelliumthere are already 2
22:43:09lebelliumdfkt's and mine
22:43:26funmanhmm then nothing i guess
22:44:10lebelliumIt's a new target (no other rboxed player with the same screen resolution) so I can't add my theme for another target
22:46:08funmansvn:// probably needs update
22:48:13lebelliumyeah probably but I mean how does it work usually? Users have to submit a request to an admin to update the target database or we just have to wait until an admin thinks about this topic?
22:51:15funmanlebellium: if you send me a patch i can apply it to the website
22:51:34funmandunno if it's updated automatically frm the svn repo though
22:52:49lebelliumI don't compile anything about RB, just making some themes as a newbie so I can't really send you a patch, I don't know what you need :S
22:54:15funmanwell i mean someone needs to patch the website, whose repo URL i just gave
22:55:27lebelliumYeah I understood that but actually I never know who I have to talk to for this kind of things. Is it written anywhere on the RB website who needs to be contacted for that or that subject?
22:55:47saratogayou probably have to ping one of the swedes so that they cp it over from svn to the webserver
22:56:05funmananyway #rockbox is always your first point of contact :)
22:59:59lebelliumIt's very unclear to me how the RB team works. Who decides what and when. Who can modify that or that page on the website. funman and saratoga, you are both in the Team as developers but you can't modify the page for instance?
23:00:50saratogalebellium: developers have SVN access, but most don't have terminal access to things like the physical web servers
23:01:45saratogaBagder, Zagor do the web server stuff, scorche the forums
23:02:01***Saving seen data "./dancer.seen"
23:02:04saratogaso if we update one of those in SVN someone has to copy the changes to the relevant machines
23:03:05scorcheforums.r.o themes.r.o and translate.r.o are all on my server
23:03:26saratogaah ok
23:03:29 Quit NoPinky (Remote host closed the connection)
23:03:37 Join NoPinky [0] (
23:08:25*pamaury begins to grumbles when he sees a check for si4702 vs si4703 in OF source code
23:08:39lebelliumok thank you for the info saratoga
23:13:10saratogabut yes if you ask on IRC theres usually someone who can fix things
23:16:35 Quit TheLemonMan (Remote host closed the connection)
23:16:42 Quit icarusfactor (Ping timeout: 244 seconds)
23:21:10 Quit ender` (Quit: Puritanism - the haunting fear that someone, somewhere, may be happy. -- Henry Mencken)
23:23:52 Join Keripo [0] (
23:31:51 Join Scromple [0] (
23:33:45 Nick funman is now known as noone (~fun@rockbox/developer/funman)
23:34:16 Nick noone is now known as Guest14160 (~fun@rockbox/developer/funman)
23:35:33 Nick Guest14160 is now known as funman (~fun@rockbox/developer/funman)
23:41:41bertriklebellium, it's not always clear for me too, especially who decides what :)
23:42:09bertrikpamaury, is there a big difference, in rockbox I think we only have to check for RDS capability
23:43:09pamaurybertrik: it only checks for chip id, if it's si4703 then it enables RDS
23:43:36pamauryit also enables RDS interrupt so I guess it's using a pin as irq, I think I found it but the code is so intricate that it's hard to tell
23:43:53pamaurybut anyway the tuner doesn't work for now :)
23:45:04bertrikoh, I thought you finally did
23:47:18pamauryno, I was reverse engineering to find out and found nothing paticular except that it uses the internal oscillator and doesn't even wait for 500ms before power up
23:50:48lebelliumbertrik: I'm still wondering how it is possible RB displays the virtual keyboard instead of my numeric volume and repeat mode tags. What's the link with the RDS feature you enabled Oo
23:54:50bertrikI have no idea
23:57:34pamaurybertrik: isn't the logic of si4700_init reversed ? the tuner need to power down to set internal oscillator
23:58:01bertrikhm, perhaps, but it does seem to work now
23:58:10pamauryyes and it doesn't solve my problem either

Previous day | Next day