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 2011-12-09

00:00:38alexbobpdfkt: yeah, it's a great device
00:00:50alexbobpjust the bare essentials. well... most of them. (hold switch)
00:07:53 Quit TheLemonMan (Quit: WeeChat 0.3.6)
00:20:08 Part lovasoa
00:24:14 Quit robin0800 (Read error: Connection timed out)
00:24:54 Join robin0800 [0] (~robin0800@
00:29:24 Quit bertrik (Ping timeout: 248 seconds)
00:32:46 Part low_light
00:35:06 Nick kugelp is now known as kugel (~kugel@rockbox/developer/kugel)
00:39:35 Quit dfkt (Quit: -= SysReset 2.55=- Sic gorgiamus allos subjectatos nunc.)
00:49:24***Saving seen data "./dancer.seen"
00:51:12 Join saratoga [0] (9803c31c@gateway/web/freenode/ip.
01:01:44JdGordondoes anyone know how the filebrowser voice clips work?
01:02:12JdGordoni.e can a similar thing be easily done to add voiceing to the shortcuts menu?
01:02:21 Quit lebellium (Quit: ChatZilla 0.9.87 [Firefox 9.0/20111130065942])
01:02:52 Join rarog [0] (
01:03:20 Quit rarog (Client Quit)
01:13:04 Quit TomColler (Remote host closed the connection)
01:16:17 Join Strife89 [0] (~Strife89@
01:30:29JdGordonwe are on lwn
01:44:29 Join dreamlayers [0] (
01:44:29 Quit dreamlayers (Changing host)
01:44:29 Join dreamlayers [0] (~dreamlaye@rockbox/developer/dreamlayers)
01:49:17 Part Poodlemastah
01:50:25 Join Poodlemastah [0] (
01:50:56 Quit keyb_gr (Ping timeout: 244 seconds)
01:54:06 Join CaptainKewl [0] (
02:12:54 Quit saratoga (Quit: Page closed)
02:22:26 Quit robin0800 (Read error: Connection timed out)
02:27:11 Quit Thra11_ (Ping timeout: 252 seconds)
02:30:11 Join robin0800 [0] (~robin0800@
02:33:11 Quit Strife89 (Ping timeout: 268 seconds)
02:34:52 Join Strife89 [0] (~Strife89@
02:49:27***Saving seen data "./dancer.seen"
03:07:28CIA-10New commit by dreamlayers (r31186): Add STORAGE_INIT_ATTR to some target-specific ATA functions.
03:08:36funmanrasher: are your sim builds broken atm?
03:09:55CIA-10r31186 build result: All green
03:14:12 Quit mgue (Ping timeout: 248 seconds)
03:23:47dreamlayersfunman: yes, see how the dates and versions differ at">
03:26:06 Join mgue [0] (
03:26:58funmandreamlayers: i'm asking help on #mingw-w64 channel (on OFTC)
03:37:12 Quit dreamlayers (Quit: back later)
03:53:01 Join brulap [0] (
03:53:04brulapHey !
03:53:37brulapCan someone add me to the wikiuser, i want to add the radio preset for Montreal
03:58:38JdGordonwhats your username?
04:01:53funmanis that your first name and last name?
04:02:56brulapBruno Lapointe
04:03:20 Quit scorche (Read error: Connection reset by peer)
04:04:02funmanplease use BrunoLapointe then
04:07:00funmantell me when you registered so i can add you to the wiki group
04:08:10 Quit Llorean (Read error: Connection reset by peer)
04:08:25funmandreamlayers: re 31186, ata_reset could be simpler
04:09:13 Join Llorean [0] (
04:09:25 Quit robin0800 (Read error: Connection timed out)
04:09:31brulapim registered
04:11:03 Quit Llorean (Changing host)
04:11:03 Join Llorean [0] (~DarkkOne@rockbox/user/Llorean)
04:11:33 Join robin0800 [0] (~robin0800@
04:13:09funmanbrulap: done!
04:14:08brulapTy !
04:17:31 Join scorche [0] (~scorche@rockbox/administrator/scorche)
04:18:52 Quit brulap (Quit: CGI:IRC (EOF))
04:28:37 Quit dys (Ping timeout: 252 seconds)
04:29:04 Join dys [0] (
04:34:45 Quit [7] (Disconnected by services)
04:34:55 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
04:38:25 Join brulap [0] (
04:40:28 Quit brulap (Client Quit)
04:41:27 Quit amiconn (Disconnected by services)
04:41:28 Join amiconn_ [0] (quassel@rockbox/developer/amiconn)
04:41:29 Quit pixelma (Disconnected by services)
04:41:31 Join pixelma_ [0] (quassel@rockbox/staff/pixelma)
04:41:34 Nick pixelma_ is now known as pixelma (quassel@rockbox/staff/pixelma)
04:41:50 Nick amiconn_ is now known as amiconn (quassel@rockbox/developer/amiconn)
04:49:28***Saving seen data "./dancer.seen"
04:51:05 Join dreamlayers [0] (
04:51:05 Quit dreamlayers (Changing host)
04:51:05 Join dreamlayers [0] (~dreamlaye@rockbox/developer/dreamlayers)
04:53:23dreamlayersfunman: what simplification do you suggest re. what r31186 does with ata_reset()?
04:56:17dreamlayersIs just putting the attribute in part of the prototype in the #ifdef better? (instead of the whole prototype)
05:28:05 Join Rob2222 [0] (
05:29:43 Quit zchs (Quit: Leaving)
05:32:01 Quit Rob2223 (Ping timeout: 268 seconds)
05:39:48 Quit Horscht (Quit: Verlassend)
05:59:25funmandreamlayers: exactly
05:59:46funmannot sure what you prefer though
06:04:55dreamlayersfunman: I thought a single prototype split over several lines with an #ifndef in the middle is harder to read.
06:11:52dreamlayersfunman: any progress re. weak and windows?
06:13:39funmani think the guys are sleeping right now
06:14:07funmanbut as i said i'm not sure if my method is reliable
06:14:25funmani hope the definition which has the weak attribute would be the 'weakest' when linking but i'm not sure
06:17:26dreamlayersfirmware/target/sh/system-sh.c uses weak for interrupt vectors, making interrupts go to the undefined interrupt handler unless a non-weak handler is defined in another file. The weak symbols and references to them are in asm, and the non-weak ones are in C.
06:20:36dreamlayersfirmware/target/coldfire/system-coldfire.c does the same thing but in C.
06:21:27funmanyep but they use alias though
06:22:23funmanperhaps we could e.g. define lcd_blit_yuv_c
06:22:36funmanand have lcd_blit_yuv be a weak alias to lcd_blit_yuv_c
06:22:54funmanthen lcd_blit_yuv_c could be eliminated if it's not used (not sure about this)
06:23:28dreamlayersIt's worth a try.
06:24:53funman__attribute__((weak, alias("f"))) void f_c(void) {}
06:25:01funmana.o:a.c:(.text.startup+0xc): undefined reference to `_f'
06:25:14funmannope, there's definitely something windows specific here
06:35:22funmanRAM usage: 0
06:35:32funmanwhen building mrobe500 binary
06:36:22dreamlayerswith some added switches?
06:37:26funmanit's generated by
06:38:06 Quit robin0800 (Quit: Leaving)
06:42:03dreamlayersyeah, it seems just isn't figuring out the size properly
06:43:03funmanworks for other targets (zip, ipod color) though
06:45:06dreamlayersthe lacks the _loadaddress and _end symbols needed for this.
06:46:14funmanwe could do with a linker template
06:49:16funman__attribute__((weak, alias("lcd_blit_yuv"))) void lcd_blit_yuv_c => the 2 functions end up in rockbox.elf
06:49:29***Saving seen data "./dancer.seen"
06:51:12funmanalias seem to have no effect on win32
06:52:53dreamlayersit's late. i'll look at this tomorrow.
06:52:57 Quit dreamlayers (Quit: goodnight)
06:53:57funman__attribute__((weakref, alias("f_c"))) void f(void);
06:54:04funmanb.c:1:45: erreur: weakref ‘f’ must have static linkage
06:54:18funmangood night!
07:03:53funmanwe should try weak symbols in asm : , perhaps that'll help us pinpoint what's going on
07:05:57 Join BHSPitMonkey [0] (~stephen@unaffiliated/bhspitmonkey)
07:09:34funmandreamlayers: IIUC it just doesn't work, but support should be easy to add in binutils
07:10:07funmanso if we need win32 sims then let's just agree to not use weak symbols (at least in code shared with win32)
07:13:14 Nick kugel is now known as kugelp (~kugel@rockbox/developer/kugel)
07:16:00 Nick kugelp is now known as kugel (~kugel@rockbox/developer/kugel)
07:18:12 Join PaulJam [0] (
07:21:58 Join n1s [0] (
07:21:59 Quit n1s (Changing host)
07:21:59 Join n1s [0] (~n1s@rockbox/developer/n1s)
07:27:32 Join fyrestorm [0] (
07:32:03 Quit Strife89 (Quit: Every time you make a typo, the errorists win.)
07:34:07 Quit PaulJam (Ping timeout: 268 seconds)
07:43:22 Quit BHSPitMonkey (Remote host closed the connection)
07:57:40 Join Keripo [0] (
08:01:56 Join Zagor [242] (~bjst@rockbox/developer/Zagor)
08:02:44 Quit CaptainKewl (Ping timeout: 248 seconds)
08:08:34 Nick ntrly_ow is now known as ntrly_owl (
08:27:39 Quit Scromple (Quit: Leaving)
08:48:51 Join ender` [0] (
08:49:33***Saving seen data "./dancer.seen"
08:51:53 Join chkktri_ [0] (
08:52:40 Join wodz [0] (
08:53:27 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
08:53:45wodzspeaking about compilers - someone with the passion should try to build rb with clang targeting ARM. There are some instructions floating around how to use clang in android sdk/ndk so this should be a good starting point
09:08:31 Join lorenzo92 [0] (
09:18:48 Join TheLemonMan [0] (
09:21:17 Join mortalis [0] (
09:33:07 Quit bertrik (Ping timeout: 240 seconds)
09:38:30 Join einhirn [0] (
09:51:01 Join keyb_gr [0] (
09:54:01 Join dys` [0] (
09:57:03 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
09:57:31 Quit dys (Ping timeout: 252 seconds)
10:06:42 Quit TheLemonMan (Quit: WeeChat 0.3.6)
10:08:32 Quit mortalis (Quit: CGI:IRC)
10:08:35 Join mortalis [0] (
10:10:14 Quit factor (Read error: Connection reset by peer)
10:11:53 Join TheLemonMan [0] (
10:17:11 Join keyb_gr_ [0] (
10:17:19 Nick dys` is now known as dys (
10:17:39 Quit Llorean (Quit: Leaving.)
10:17:54 Quit keyb_gr (Ping timeout: 244 seconds)
10:18:10 Nick keyb_gr_ is now known as keyb_gr (
10:24:31 Quit keyb_gr (Ping timeout: 244 seconds)
10:26:01 Join keyb_gr [0] (
10:27:23 Join factor [0] (~factor@
10:43:14 Quit kadoban (Ping timeout: 240 seconds)
10:44:41 Quit Keripo (Quit: Leaving.)
10:49:35***Saving seen data "./dancer.seen"
10:55:07lorenzo92kugel: I'll upload the result of your screen fix on my device before the crash...
10:57:34lorenzo92so before that, the logo comes up
10:57:43 Join swilde [0] (~wilde@2001:470:9c7c:abc:8c0d:78ff:fe95:cab1)
10:57:52kugelthe logo is fine, and then the garbage appears?
10:58:06kugelor is the garbage there from the beginning?
10:58:14lorenzo92exactly. logo with revision number etc is fine, then garbage
10:58:32kugelI notice some glitches on my device too, but not that bad
10:58:58lorenzo92uhm maybe I'm using lebellium's theme...
10:59:15kugelah interesting to know. will try that too
10:59:31lorenzo92yeah in the meanwhile I try with cabbie
11:02:09lorenzo92oh mah lol
11:02:14lorenzo92with cabbie works uhm
11:02:44kugelthe shutdown splash doesnt appear
11:02:58kugelquickscreen also shows glitches at the top I think
11:03:04kugelor pitchscreen?
11:03:35lorenzo92yeah indeed several issues
11:03:52lorenzo92pitch screen works after some cicles of redraw I guess
11:04:19lorenzo92and yes
11:04:27lorenzo92loading lebelliums theme crashes everything
11:05:43lorenzo92the thing is: how redrawing? if full size, okay, the array is copied fully to the framebuffer. but then if only some points? maybe we could simplify it redrawing only lines
11:06:02lorenzo92no well guess in this way we have other glitches...
11:06:19kugelupdating more than needed can be done
11:06:41kugeldont expect glitches. update_rect() is only an optimization means
11:06:51lorenzo92ah really? okay then
11:07:27lorenzo92but still don't understand why it crashes...
11:07:38kugelyea that's strange
11:07:55 Quit aevin (Ping timeout: 252 seconds)
11:10:14lorenzo92I could try to increase virtual resolution (now it's 240, 640)
11:12:19kugelvirtual resolution?
11:15:36lorenzo92no maybe I said something horrible...anyway is the virtual res of the framebuffer but actually i don't know really what's the meaning :9
11:16:53lorenzo92maybe it goes out of bounds somewhere
11:17:00lorenzo92the array...
11:17:12kugelthe framebuffer is 240x320
11:18:09lorenzo922 byte sized pixel, right?
11:19:29kugeldont worry too much. I'll try 4byte (x and width must be even) aligned addresses on the weeked, if all fails we stick to what you implemented
11:19:50TheLemonManmaybe it's meant to be used with panning to have double buffering
11:20:22lorenzo92kugel: okay ;)
11:20:29kugellorenzo92: what is 240x640 what you were refering to?
11:21:27lorenzo92I read that in the virtual_size property of the framebuffer...
11:24:06lorenzo92kugel: I suspect that alsa is still using too much cpu :(
11:24:26kugelwhat "virtual_size property"?
11:25:02 Join dfkt [0] (dfkt@unaffiliated/dfkt)
11:26:30lorenzo92goto: /sys/class/graphics/fb0/virtual_size
11:26:36lorenzo92open with text editor
11:26:40lorenzo92and you can see that
11:30:43kugeldont know what that is :)
11:30:48kugelperhaps TheLemonMan is right
11:31:06lorenzo92yes could be
11:33:28lorenzo92and to change topic....battery level :D any idea how could we implement the already written ascodec "driver"? because I tried in the past but I ended up with mess only hehe
11:43:42 Quit TheLemonMan (Quit: WeeChat 0.3.6)
11:43:59 Join TheLemonMan [0] (
11:51:21 Quit TheLemonMan (Read error: Connection reset by peer)
11:51:42 Join TheLemonMan [0] (
11:51:45 Quit mortalis (Quit: CGI:IRC)
12:06:01 Quit TheLemonMan (Ping timeout: 252 seconds)
12:09:47 Quit swilde (Quit: ERC Version 5.3 (IRC client for Emacs))
12:11:20 Join TheLemonMan [0] (
12:16:29 Quit TheLemonMan (Quit: WeeChat 0.3.6)
12:16:44 Join TheLemonMan [0] (
12:19:17 Join gss_zh [0] (~star@
12:25:29 Quit einhirn (Quit: Miranda IM! Smaller, Faster, Easier.
12:27:14 Quit lorenzo92 (Quit: ChatZilla 0.9.87 [Firefox 8.0/20111115183813])
12:29:51 Nick kugel is now known as kugelp (~kugel@rockbox/developer/kugel)
12:45:42 Quit maraz (Read error: Operation timed out)
12:45:49 Join maraz [0] (
12:47:23 Quit TheLemonMan (Quit: WeeChat 0.3.6)
12:47:38 Join TheLemonMan [0] (
12:49:36***Saving seen data "./dancer.seen"
12:56:25 Quit maraz (Ping timeout: 240 seconds)
13:00:21 Join maraz [0] (
13:03:39 Quit maraz (Read error: Operation timed out)
13:09:45 Join maraz [0] (
13:30:15 Quit evilnick (Ping timeout: 240 seconds)
13:32:12 Join evilnick [0] (~evilnick@rockbox/staff/evilnick)
13:32:44 Join fml [0] (
13:33:14fmlIs it correct that the sleep time is only available for devices with the option "rtc"?
13:33:25fmler... timer
13:35:10fmlThe menu entry "Time and date −−> Sleep timer" is only present in the manual for rtc devices. But I think it's not correct.
13:44:52CIA-10New commit by alle (r31187): Manual: Some minor corrections to the 'Shortcuts' section
13:47:22CIA-10r31187 build result: All green
13:53:05 Nick kugelp is now known as kugel (~kugel@rockbox/developer/kugel)
14:05:15pamauryWow, I *finally* found the recovery mode of the Creative Zen Style M300 \o:
14:06:35amiconnfml: Sleep timer is available on all targets (not sure about raaa)
14:08:19wodzpamaury: found as in binary or found as finger breaking combo?
14:11:15 Nick kugel is now known as kugelp (~kugel@rockbox/developer/kugel)
14:11:51fmlamiconn: yes, that's what I think too (judging by the code). But in the manual, sleep timer is only present for rtc targets (the whole Time and Date section is included only then).
14:12:20 Quit factor (Ping timeout: 248 seconds)
14:12:25amiconnI don't have a "Time & Date" option on the Ondio, the sleep timer resides directly under "Settings"
14:14:44 Quit MethoS- (Quit: Konversation terminated!)
14:21:14 Quit keyb_gr (Ping timeout: 240 seconds)
14:21:35 Join nick-p [0] (
14:24:52nick-pA forum posting I made seems to have been deleted, any idea why? I was hoping to document theme changes there:,29497.0.html
14:25:24fmlamiconn: Then the structure of the menu in the manual is quite wrong. There, Sleep time sits in "Date and Date".
14:25:29 Quit fml (Quit: ChatZilla 0.9.87 [Firefox 3.6.24/20111107173218])
14:30:38pamaurywodz: finger breaking combo :) Actually you "just" need to plug usb while hold power but if you plug usb before power or power before plugging usb it will not work :)
14:31:24nick-pfml: Sleep Timer only sits in Time & Date if you have a Real Time Clock, otherwise it's in the Settings menu. There is no Time & Date menu if you don't have a RTC.
14:34:16 Nick kugelp is now known as kugel (~kugel@rockbox/developer/kugel)
14:35:07wodzhow to use __attribute__ to disable optimization for particular function?
14:35:59 Join gmanol [0] (
14:37:31pamaurywodz: __attribute((optimize(0))) ?
14:38:48wodzdoesn't work
14:38:56wodzit throws error
14:40:00gmanolhi, I'm trying to submit a patch for a new codec i created (zx spectrum tap files). The question is against which svn revision should i do it ? what do you recommend ?
14:40:04pamauryI think there are gcc specific pragmas: #pragma GCC optimize(push/pop) and #pragma GCC optimize(...)
14:40:14pamaurywhere ... can be 0 I guess
14:41:00wodzaccording to docs __attribute__ ((optimize("0"))) and #pragma GCC optimize 0 are equivalent
14:41:05 Quit evilnick (Ping timeout: 240 seconds)
14:42:03 Quit kugel (Remote host closed the connection)
14:42:12pamaurynot exactly, __attribute__ ((optimize("0"))) is not the same as __attribute__ ((optimize(0)))
14:42:59pamaurybut __attribute__ ((optimize("O0"))) should be the same as __attribute__ ((optimize("0")))
14:43:17nick-pgmanol: you post a patch against the current trunk, then keep it up to date for as long as the developers argue about/ ignore it ;)
14:43:26wodzok resolved this
14:44:43gmanolnick-p: thank you. But the current trunk (rev 31186) causes a segmentation fault when executed. I builded the simulator for sansa clip+ without any modifications from my side.
14:45:32 Join rivia [0] (
14:46:03wodzmeh - compiler ignores this attribute apparently and optimizes out variable I want to look at
14:46:07wodzscrew you GCC
14:46:16riviahi! will be supported in future iRiver T6 ?
14:47:25wodzrivia: if you do the port why not
14:48:22nick-pgmanol: might be best to start a flyspray entry with the latest revision you can compile/ run against. There's a good chance the patch can be applied to the latest revisionfor others anyway.
14:48:25nick-pYou can keep posting updates on the FS page you create.
14:49:40***Saving seen data "./dancer.seen"
14:49:48wodzmeh, stupid me __attribute__ works - I just run wrong binary
14:51:28gmanolnick-p: great. One last question. In flyspray when i'm submitting a new task there is a field "Reported Version" with has only four options. "Release 3.9", "Daily build", "Rbutil 1.2.7" & "Rbutil SVN". Which one should i select. How can i declare the revision my patch works against ?
14:52:01 Quit pamaury (Read error: No route to host)
14:52:09nick-pgmanol: I think that's mainly for bugs, I never bothered with it for patches
14:52:13 Join pamaury [0] (
14:52:13 Quit pamaury (Changing host)
14:52:13 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
14:53:21gmanolnick-p : should i write in the details field the revision for my patch ?
14:54:11nick-pWouldn't hurt, though if you made the patch with a svn diff it'll be in there anyway
14:54:42gmanolaaa.. you're right! thank you nick :)
15:07:09 Join mystica555_ [0] (
15:14:37 Quit gevaerts (Ping timeout: 252 seconds)
15:14:38 Quit wodz (Quit: Leaving)
15:24:13 Part rivia
15:25:38 Quit chkktri_ (Quit: Leaving)
15:26:09 Join evilnick [0] (~evilnick@rockbox/staff/evilnick)
15:31:03 Join keyb_gr [0] (
15:33:01 Join benedikt93 [0] (~benedikt9@unaffiliated/benedikt93)
15:35:33 Part gss_zh
15:37:16 Quit mystica555_ (Remote host closed the connection)
15:38:19 Join mystica555_ [0] (
15:42:08 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
15:48:34 Join WalkGood [0] (~4@unaffiliated/walkgood)
15:57:58 Join dfkt|n [0] (dfktn@unaffiliated/dfkt)
16:10:12 Join y4n [0] (y4n@unaffiliated/y4ndexx)
16:15:51 Join dreamlayers [0] (
16:15:51 Quit dreamlayers (Changing host)
16:15:52 Join dreamlayers [0] (~dreamlaye@rockbox/developer/dreamlayers)
16:17:03 Quit gmanol ()
16:28:02TheLemonManassuming main is the first function called by rockbox i can safely say that it doesnt even reach that
16:28:21TheLemonManso, what the hell happens during the load ?
16:29:59 Join gevaerts [0] (~fg@rockbox/developer/gevaerts)
16:34:00CIA-10New commit by dreamlayers (r31188): Introduce USED_ATTR wrapper for __attribute__((used)).
16:34:07 Join factor [0] (~factor@
16:34:21Tornecrt0 runs first
16:34:28Tornemain is not called until the basic C environment has been set up
16:35:15 Join wodz [0] (
16:35:31wodzTorne: he talks about RaaA I think
16:36:06CIA-10r31188 build result: All green
16:36:10Tornewell, same answer, no? :)
16:36:21Tornedo we have anything called at init time?
16:36:27wodzsort of
16:36:48wodzI guess RaaA crt0.S is vastly different than our hand crafted ones
16:37:11Torneright, but it still (potentially) does more than just call main()
16:37:25TorneTheLemonMan: what's actually happening that you can observe?
16:37:28Tornealso, strace/etc
16:37:37wodzobjdump should tell
16:38:05TheLemonMani should be able to get a strace in a few mins
16:38:48 Join saratoga [0] (9803c57f@gateway/web/freenode/ip.
16:39:09Tornewodz: possibly not trivially, though, thanks to exciting platform abi variation
16:40:38wodzwhat do you mean? ctr0.S should be in binary, no?
16:42:10Torneyes, but if you know what the ABI says you can deduce what it will do without actually reading the assembly
16:42:22Tornee.g. on EABI you can just look in .init_array
16:43:10wodzbut if we suspect something odd is happening before main() we shouldn't relay on this
16:43:18TheLemonManstrace doesnt seem to work
16:43:27TheLemonManat least, the output is just blank
16:43:41 Quit pamaury (Remote host closed the connection)
16:43:42wodzTheLemonMan: for rb or in general?
16:44:01wodzTheLemonMan: can you run 'hello world' program correctly?
16:44:01Tornewhat does it actually do when you run it?
16:44:02TheLemonManin general, its really a pain to debug
16:44:30TheLemonMansure i do, i already had ported several apps
16:44:56TheLemonManit doesnt do anything, its like it returns as soon as its launched
16:45:27wodzok, so toolchain is probably ok and crt0.S should be as well
16:46:07saratoganick-p: where as that post originally?
16:46:46nick-psaratoga: the forum post?
16:47:48wodzTheLemonMan: can you do serial-over-usb?
16:47:52nick-pit disappeared, I posted it again, but I think it needs to be accepted:,29499.0.html
16:48:22wodzTheLemonMan: or plain oldfashioned uart is supported?
16:48:47TheLemonMancant use either, this is the point
16:49:04wodzTheLemonMan: lack of driver or hw limitation?
16:49:07nick-psaratoga: the original which I quoted in the theme which got wasted:,29497.0.html
16:49:35TheLemonManwodz: no soldering skills :(
16:49:42***Saving seen data "./dancer.seen"
16:49:51saratoganick-p: it looks like the spam software is blocking your posts
16:49:58saratogaone of the forums admins can fix that i think
16:50:07saratogawhere as the post originally made? the wps forum?
16:50:14wodzwell for serial-over-usb you don't need soldering skills - only driver is needed
16:50:36nick-psaratoga: yes, the theming forum section
16:50:40saratogado we have software FP support for codecs?
16:50:47saratogalet me see if i can put it back
16:50:56saratogasomeone else will have to whitelist you for the spam software
16:51:12saratogafor whatever reason it sometimes locks onto accounts, it was blocking me for a while
16:51:14TheLemonManwodz: no kernel-side support for that
16:51:36nick-psaratoga: thanks
16:51:46saratogawell i moved it back but now it shows up as blocked :)
16:51:54saratogascorche: can you fix that?
16:52:17nick-pno problem, it's for a e200 theme, not exactly high traffic ;)
16:52:31wodzTheLemonMan: in this situation you don't have much options
16:55:25TheLemonManits already the 2nd time i stop porting rockbox here
16:55:39nick-psaratoga: thanks, I'm seeing 29497 now - any chance 29499 could be deleted?
16:56:42n1ssaratoga: i'm pretty sure gcc includes the necessary emulation routines if you use fp
16:57:24saratogaso do we care if codecs use FP in non-performance sensitive areas?
16:57:55Tornehm, possibly not, but it's doing stuff like "generating sin()"
16:58:00Tornewhich i'm pretty sure we already have code for, no?
16:58:15saratogayes we have fixed point versions of those functions
16:58:39Tornethe codec is pretty simple (it's just a multitone generator, really)
16:58:46Torneso it doesn't seem like it should be super hard to do it "right"
16:58:46n1si think some codecs do or did that
16:58:50Torneeven if doing it with floating point works
16:59:52 Part Zagor
17:00:22 Join gmanol [0] (
17:05:17 Join PaulJam [0] (
17:06:27bertrikdo we have a plugin to enforce a data abort? :)
17:06:54Lalufuthat should not be too difficult.
17:09:45 Quit keyb_gr (Ping timeout: 252 seconds)
17:10:38bertrikI think the current intention is a reboot on keypress when showing an abort, but on the clip it just hangs. This could be related to us stopping the ROM clock to save power.
17:10:39 Join Zagor [242] (~bjst@rockbox/developer/Zagor)
17:11:34 Quit mystica555 (Quit: WeeChat 0.3.7-dev)
17:12:30 Quit nick-p (Quit: Leaving)
17:12:35 Join mystica [0] (
17:13:05CIA-10New commit by dreamlayers (r31189): Convert remaining __attribute__((noreturn)) to NORETURN_ATTR.
17:13:21 Nick mystica is now known as mystica555 (
17:15:11CIA-10r31189 build result: All green
17:15:40saratogabertrik: turning the ROM back on is fine if it causes problems
17:15:51 Join toffe82 [0] (
17:15:53bertriksaratoga, yes, I'll investigate that
17:16:18saratogai switched it off because I was under the impression it would be restarted as soon as the device rebooted and so we didn't need it
17:17:43bertrikhm, I'll also check how we do the reboot exactly, maybe not rebooting "hard" enough
17:18:00 Quit dreamlayers (Quit: back later)
17:18:30bertrikwe use the watchdog, which I expect to do a pretty hard reset
17:19:26 Join keyb_gr [0] (
17:23:10 Quit gmanol ()
17:23:12 Join T44 [0] (
17:24:09 Quit B4gder (Ping timeout: 244 seconds)
17:26:25 Quit Topy44 (Ping timeout: 240 seconds)
17:27:32 Quit Sundiver (Ping timeout: 248 seconds)
17:34:28n1sah, we're on lwn
17:37:09 Join kadoban [0] (
17:39:04pixelmaI wonder a bit about the .tap codec as there is a tap viewer (ZX Box)
17:39:30pixelmahaven't looked into the tracker entry yet though
17:42:55 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
17:48:28 Part Zagor
17:55:06 Join Sundiver [0] (~angel@
18:00:42 Join dreamlayers [0] (
18:00:42 Quit dreamlayers (Changing host)
18:00:42 Join dreamlayers [0] (~bgjenero@rockbox/developer/dreamlayers)
18:15:09 Join lovasoa [0] (~olojkine@
18:18:53 Join aevin [0] (eivindsy@unaffiliated/aevin)
18:24:11 Join Llorean [0] (~DarkkOne@rockbox/user/Llorean)
18:27:36 Join Horscht [0] (
18:27:36 Quit Horscht (Changing host)
18:27:36 Join Horscht [0] (~Horscht@xbmc/user/horscht)
18:49:43***Saving seen data "./dancer.seen"
18:52:42 Quit saratoga (Ping timeout: 265 seconds)
18:59:03 Join Buschel [0] (
19:10:13 Join thomasjfox [0] (~thomasjfo@rockbox/developer/thomasjfox)
19:11:32 Quit evilnick (Read error: Connection reset by peer)
19:17:37 Part lovasoa
19:24:26pamauryhow can the sd init work ? most drivers grab a mutex on sd card init, then call disk_mount with the mutex hold and the read/write functions need to grab the mutex too :-/
19:26:47funmandreamlayers: i think we could just skip the weak attribute if building for windows
19:27:07funmandreamlayers: we won't use a better/optimized function when building windows sims
19:29:18dreamlayersfunman: yeah, that's an easy solution in this case
19:29:24bertrikpamaury, I think one thread can get the mutex again without blocking itself, that's the difference between a 1-count semaphore and a mutex
19:29:26 Join Bagder [0] (
19:29:26 Quit Bagder (Changing host)
19:29:26 Join Bagder [241] (~daniel@rockbox/developer/bagder)
19:42:54jhMikeSindeed the mutex is implemented to be recursive
19:51:28 Quit Buschel (Quit: ChatZilla 0.9.87 [Firefox 8.0/20111104165243])
20:02:21 Quit mystica555 (Quit: WeeChat 0.3.7-dev)
20:02:53 Join mystica555 [0] (
20:04:11 Join Strife89 [0] (~Strife89@
20:04:20 Quit WalkGood (Ping timeout: 248 seconds)
20:08:22 Join WalkGood [0] (
20:12:05 Quit WalkGood (Changing host)
20:12:05 Join WalkGood [0] (~4@unaffiliated/walkgood)
20:22:26 Quit dfkt|n ()
20:22:45 Quit WalkGood (Quit: ♪ ♫ ♪ ♫ ♪ ♫ ♪ ♫)
20:26:48 Quit Strife89 (Ping timeout: 268 seconds)
20:36:57 Join Strife89 [0] (
20:39:51 Quit PaulJam (Ping timeout: 252 seconds)
20:46:04CIA-10New commit by thomasjfox (r31190): Nokia N900: Implement pcm_play_lock() / pcm_play_unlock() ...
20:47:08 Quit y4n (Quit: We're fucking 3LN!)
20:48:02CIA-10r31190 build result: All green
20:48:33 Quit Strife89 (Ping timeout: 252 seconds)
20:49:46***Saving seen data "./dancer.seen"
21:04:19 Join Strife89 [0] (~Strife89@
21:04:43CIA-10New commit by thomasjfox (r31191): Nokia N900: Increase the number of samples for the software mixer ...
21:07:17CIA-10r31191 build result: All green
21:18:05 Join fml [0] (
21:18:40fmlwhy is "sleeptimer on startup" only available for non-rtc targets?
21:19:41amiconnIt's not
21:19:55amiconnIt's available on non-rtc too
21:20:43fmlsorry, I looked at an old version
21:21:54fmlbut the timer on non-rtc works differently. the values are in decreasing order.
21:27:08 Quit TheLemonMan (Quit: WeeChat 0.3.6)
21:27:29 Join Stummi [0] (~Stummi@rockbox/developer/Stummi)
21:28:56dreamlayersamiconn: do you have any files which reliably reproduce the sh gcc jump table bug that's addressed by the gcc patch?
21:32:35pamauryjhMikeS: there is something strange with the sd mutex: if I do "return -1;" rockbox boots, and if I do "mutex_lock;mutex_unlock;return -1;" it doesn't boot !
21:36:12 Quit keyb_gr (Ping timeout: 252 seconds)
21:39:59bertrikpamaury, is the mutex properly initialised?
21:40:52funmancan you inspect its memory before and after each step?
21:41:31 Join keyb_gr [0] (
21:42:59pamauryI guess I can panic to do that
21:45:07pamaurywhat should I print ?
21:45:13 Quit fml (Quit: ChatZilla 0.9.87 [Firefox 8.0/20111104165243])
21:45:15CIA-10New commit by dreamlayers (r31192): Add labels for calculating RAM usage on tms320dm320 targets. Thanks to Rafaël Carré for pointing this out.
21:45:19pamaurymutex.recursion ?
21:47:07CIA-10r31192 build result: All green
21:47:23 Quit dfkt (Quit: -= SysReset 2.55=- Sic gorgiamus allos subjectatos nunc.)
21:51:47pamaurythat's strange, it's called with recursion=0 so it might get called *before* the init !
21:52:09pamauryor not, depends on how recursion is updated
21:55:20pamauryI really don't understand why it freezes
21:56:41CIA-10New commit by funman (r31193): lcd_blit_yuv(): don't use weak attribute when building for windows ...
21:56:49bertrikcan we take a look at your code, pamaury
21:57:15funmanwindows sims should be back
21:58:33pamaurysd_init_card succeeds, returns 0, then disk_mount gets called and the player lock up. In transfer_sectors if I remove the mutex_lock/unlock pair before the return -1 it boots
21:59:02CIA-10r31193 build result: All green
22:01:29 Join thegeek [0] (
22:07:36 Quit fs-bluebot (Read error: Connection reset by peer)
22:09:21bertrikpamaury, I don't see anything obviously wrong with the mutex locking
22:10:46 Join fs-bluebot [0] (
22:11:41pamauryit's the same code as as3525 basically for most part
22:14:52 Quit dreamlayers (Quit: brb)
22:16:42 Join dreamlayers [0] (
22:16:46 Quit dreamlayers (Changing host)
22:16:46 Join dreamlayers [0] (~dreamlaye@rockbox/developer/dreamlayers)
22:19:12bertrikfunman, you tested your clip zip dualboot code, right? I see there's a new official clip zip firmware now.
22:26:38 Quit factor (Read error: Connection reset by peer)
22:30:49 Join Keripo [0] (
22:36:00n1shmm the clang static analyzer might have actually spotted a real bg
22:36:11Bagderthat's an awesome tool
22:36:17n1sor at least some bogus code
22:36:23n1syes it's very nice
22:37:03n1scould anyone that knows dircache check out dircache.c:dircache_resume ?
22:38:56 Join robin0800 [0] (~robin0800@
22:39:11 Join KBH [0] (
22:43:57 Join factor [0] (~factor@
22:44:45 Quit robin0800 (Ping timeout: 252 seconds)
22:44:53dreamlayersfunman, thanks for the windows sim fix. It works when building in cygwin with i686-pc-mingw32-gcc or i686-w64-mingw32-gcc.
22:45:49 Quit dreamlayers (Quit: bye)
22:46:59 Quit toffe82 (Quit: ChatZilla 0.9.87 [Firefox 8.0.1/20111120135848])
22:49:47***Saving seen data "./dancer.seen"
22:53:29jhMikeSpamaury: maybe it's not that one but perhaps another one elsewhere causing trouble. also, recursion should be 0 unless a thread locks twice or more
22:58:47 Join freddybb [0] (
23:00:18 Quit Stummi (Quit: Bye!)
23:03:18 Quit thomasjfox (Remote host closed the connection)
23:13:30 Quit wodz (Quit: Leaving)
23:15:29 Quit benedikt93 (Quit: Bye ;))
23:22:09 Join Ericc [0] (
23:24:36EriccHi, Im not affiliated with but they have a good deal on Sansa clip zips. Buy 8GB one and get a 16GB card with it and an extra 16GB for £1.99
23:26:48EriccHappy Birthday ROCKBOX
23:28:15 Join robin0800 [0] (~robin0800@
23:30:24 Quit Strife89 (Ping timeout: 252 seconds)
23:30:51keyb_grGuys - is it intentional that the 3.10 release notes link on the download page is still pointing to 3.9.1 notes?
23:36:27 Quit Ericc (Quit: CGI:IRC)
23:37:35n1skeyb_gr: i'd say no
23:38:01pamauryweird, the sd init only lock up at boot, but if I insert the sd later it works. It might be due to sd/mmc ordering, the sd is considered the first partition...
23:38:28CIA-10New commit by gevaerts (r31194): Fix release notes link
23:38:58gevaertsBagder: can you do the necessary server magic to use the new
23:40:21CIA-10r31194 build result: All green
23:41:08funmanbertrik: yes i tested it
23:41:31funmanbertrik: we could make a mkamsboot / zip bootloader release
23:46:37 Join PaulJam [0] (
23:47:29 Quit freddybb (Quit: ChatZilla 0.9.87 [Firefox 8.0/20111115183813])

Previous day | Next day