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).

Notice: Only Gecko based browsers prior to FF4 support the multipart/mixed "server push" method used by this log reader to auto-update. Since you do not appear to use such a browser, this page will simply show the current log, and not automatically update.

#rockbox log for 2014-08-04

00:00:33 Quit IgoRCL (Read error: Connection reset by peer)
00:10:10 Quit lebellium (Quit: ChatZilla [Firefox 32.0/20140731191115])
00:19:38 Quit ZincAlloy (Ping timeout: 244 seconds)
00:25:12 Quit bertrik (Read error: Connection reset by peer)
00:26:11 Quit Xerion (Read error: Connection reset by peer)
00:27:57 Join Xerion [0] (
00:28:06 Part Xerion
00:28:14 Join Xerion [0] (
00:29:38 Quit Xyem (Ping timeout: 260 seconds)
00:32:29 Join Xyem [0] (
00:35:32 Join ikeboy [0] (
00:37:59 Quit Xyem (Remote host closed the connection)
00:40:19 Join Xyem [0] (
00:44:44 Quit ikeboy (Quit: Leaving)
00:46:46 Join franklin [0] (
00:48:53 Quit Rower (Quit: Hmmm...)
00:49:06 Quit the-kyle (Remote host closed the connection)
00:49:15 Join the-kyle [0] (
00:56:39 Quit bp0 (Quit: Leaving)
01:21:38 Nick franklin is now known as franklinbot (
01:22:03 Nick franklinbot is now known as franklin (
01:23:00 Nick franklin is now known as franklinbot (
01:23:34 Nick franklinbot is now known as franklin (
01:27:00 Quit franklin (Quit: Leaving)
01:32:28[Saint]semi-related hardware-ish topic:
01:32:51[Saint]Is there any reason I'm overlooking for driving a pre-amp high when it offers a low power mode?
01:33:07[Saint]This device is doing some weird shit with throwing the juice around.
01:36:48[Saint]In this device there's a pre-amp driven high, which then feeds into a secondary amp. Setting the first stage amp to low power mode makes no appreciable difference in quality or performance and extends battery life dramatically.
01:37:08[Saint]I'm wondering if Samsung is really this naive, or if there's something I'm overlooking.
01:42:01 Join franklin [0] (
01:46:38***Saving seen data "./dancer.seen"
01:48:30 Quit franklin (Ping timeout: 244 seconds)
01:53:29 Quit benedikt93 (Quit: Bye ;))
02:01:55 Join franklin [0] (
02:09:24 Quit franklin (Quit: Leaving)
02:15:54 Join sakax [0] (~sakax@unaffiliated/sakax)
02:27:52 Quit sakax (Quit: Leaving)
02:42:58 Join sakax [0] (~sakax@unaffiliated/sakax)
02:54:48 Quit sakax (Quit: Leaving)
02:55:24 Join ygrek [0] (~user@
02:59:59 Quit AlexP (Remote host closed the connection)
03:12:11[Saint]slider is really quite fucked up.
03:12:56[Saint]oh. no. hmmmmm.
03:13:04[Saint]its its touch area being weird.
03:13:41[Saint]wellp. I can live with that. I'll nuke the bar's touch area and overlay my own larger one.
03:21:45[Saint]copper, pixelma, kugel, other people who may care...^
03:22:23[Saint](the clashing yellow album art was a pure coincidence)
03:23:51[Saint]I'm still not sure about the order and content of the overflow menu.
03:44:55JdGordon_[Saint]: looks nice
03:45:06JdGordon_is that progress slider big enough to actually use thouhg?
03:45:55[Saint]the bar itself passes the notouch param and there's a MUCH larger viewport behind it handling the scrubber touch region.
03:46:06[Saint]old trick I learned.
03:46:23JdGordon_oh, cool
03:46:28[Saint]so you can go way off the reservation when seeking.
03:46:40***Saving seen data "./dancer.seen"
04:00:53 Join HeadWhacker [0] (84934c34@gateway/web/freenode/ip.
04:01:24JdGordon_[Saint]: seriously... whats it going to cost me to get a 1080*1920 version?
04:02:52[Saint]now I'm reasonably happy with the layout, despite the drawing artifacts, and got positioning cleaned up a bit, ports to alternate resolutions will be fairly trivial.
04:04:17JdGordon_so I'll have one for my drive home in 3h? :)
04:04:24[Saint]I suspect people will gripe about the overflow menu and the lack of immediately observable status icons and state display.
04:04:32HeadWhackerI am hunting down what is causing some random crackling noise while music is playing on rockbox.
04:05:28[Saint]HeadWhacker: have you watched the debug menu to see if it corresponds with either boosting or buffer refill?
04:05:40[Saint]those are two likely candidates.
04:09:21 Quit ygrek (Ping timeout: 256 seconds)
04:11:40HeadWhackerI enable logf
04:12:29[Saint]no I mean there's a debug menu view that allows you to view the buffer state.
04:12:34HeadWhackerit looks like the noise happen when the buffer handle is close and a fill is started
04:12:45HeadWhackerah no not yet
04:13:12[Saint]Ah. No debug menu on this target?
04:13:41HeadWhackerthere is
04:14:28HeadWhackerbut I have to compile with debug options before I can see debug messages right?
04:17:02[Saint]not for what I was talking about.
04:17:16[Saint]settings - system - debug - view buffering thread
04:17:53[Saint]easiest place to watch the current buffer and cpu state from.
04:18:18 Quit cmhobbs (Remote host closed the connection)
04:18:57HeadWhackerOk I have that in the debug menu
04:19:51*JdGordon_ sets up a bot to prod [Saint] every 5 min till he gets his theme on the phone!
04:24:07 Quit JoeyJoeJo (Remote host closed the connection)
04:24:36 Join ygrek [0] (~user@
04:54:09 Join amiconn_ [0] (amiconn@rockbox/developer/amiconn)
04:54:09 Quit pixelma (Disconnected by services)
04:54:09 Quit amiconn (Disconnected by services)
04:54:10 Join pixelma_ [0] (pixelma@rockbox/staff/pixelma)
04:54:12 Nick amiconn_ is now known as amiconn (amiconn@rockbox/developer/amiconn)
04:54:12 Nick pixelma_ is now known as pixelma (pixelma@rockbox/staff/pixelma)
05:03:53 Join JdGordon [0] (
05:03:53 Quit JdGordon (Changing host)
05:03:53 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
05:05:58 Quit JdGordon_ (Ping timeout: 255 seconds)
05:08:44HeadWhackerSaint: I looked at the bufferin thread and I can see the noise appear while the buffer is getting filled.
05:09:45JdGordonHeadWhacker: on a sansa clip or similar dap?
05:10:13HeadWhackerNo on DX90
05:22:40 Quit TheSeven (Ping timeout: 250 seconds)
05:23:50 Join TheSeven [0] (~quassel@rockbox/developer/TheSeven)
05:26:20HeadWhackerI think I may have found a workaround
05:26:58HeadWhackerLooks like the noise appear when rockbox loads 2 or more tracks in the buffer
05:29:10HeadWhackerI reduce the Memory size to force RB to load only one track of 16bit/44k file
05:29:35HeadWhackerI don't seems hear noise anymore
05:34:49 Quit jhMikeS (Ping timeout: 244 seconds)
05:44:05 Join steffengy [0] (
05:46:44***Saving seen data "./dancer.seen"
05:47:14 Quit steffengy1 (Ping timeout: 260 seconds)
06:08:20HeadWhackerbtw it looks like Rockbox use all available memory for audio buffer. Is it possible to force it to use a fix buffer size for audio?
06:15:58 Quit tchan (Quit: WeeChat 0.4.3)
06:37:21 Join tchan [0] (~tchan@lunar-linux/developer/tchan)
07:13:59[Saint] Yes that's possible.
07:14:09[Saint]Let me get to a workstation.
07:17:50ParkerR[Saint], I have made a huge mistake...
07:17:56ParkerR(Using rockbox on android lol)
07:18:14ParkerRMy current dpi made the default freaking small
07:18:46[Saint]ParkerR: target resolution?
07:19:09ParkerRNo extra fonts
07:19:11ParkerROr themes
07:19:18ParkerRSo Im going to have to grab those
07:19:30ParkerRthen it will be readable :D
07:19:34[Saint]no themes exist for this resolution.
07:19:44ParkerR[Saint], sure they do
07:19:44[Saint]I have a font pack in the extras section, though.
07:19:50ParkerRRockboxed for the Video :P
07:20:02[Saint]That won't work.
07:20:11ParkerRRockboxed works pretty much everywhere...
07:20:17ParkerRIts worked before
07:20:27[Saint]You must ve using an ancient build.
07:20:45ParkerRUnless it changed. It worked when I tried a while back
07:21:00[Saint]Rockbox on Android will flip its shit if the compiled binary doesn't match the framebuffer.
07:21:09[Saint]this regression started in May of this year.
07:21:15ParkerRNo I dont mean the binary
07:21:20ParkerRI just mean the wps theme
07:21:20[Saint]the theme now MUST match the framebuffer.
07:21:48[Saint]Doesn't matter.
07:22:25[Saint]Oh, actually, you might get away with it if there's no backdrop loaded by the theme.
07:22:33[Saint]But, I'm fairly sure there is.
07:22:39ParkerRYeah its justa background color
07:23:28[Saint](I finally have it stuck together)
07:23:39[Saint]Now I have to port it to a fucktonne of additional targets.
07:23:42ParkerR:O wat is this
07:23:45ParkerRA native theme...
07:23:57ParkerROr just an android-ish theme
07:24:15[Saint]Ooooohhhhh - tricked 'ya. I must be on the right track.
07:24:28[Saint]Its the Rb theme engine /trying/ to be native. :)
07:24:29ParkerRLooks good
07:24:52ParkerR[Saint], +1
07:25:43ParkerR[Saint], I dont think I ever looked this up. So does the android .rockbox reside in /sdcard/rockbox or in the folder
07:26:41ParkerROr does wps just go in /sdcard/rockbox/
07:26:57[Saint]"/.rockbox" is /storage/sdcard0/rockbox or whatever that path equates to on your system.
07:27:09ParkerRAhh ok
07:27:14[Saint]usually additionally symliked as /sdcard/rockbox
07:28:03[Saint]There's a great and varied number of paths that can end up pointing to the external or emulated storage.
07:28:15[Saint]its very difficult to just say "look here".
07:28:41ParkerR/sdcard/ Ive found is usally linked correctly on single emulated/micro sd phones
07:28:50ParkerRIts when there is both that it get weirder
07:29:49ParkerR[Saint], is there an android specific font pack or just choose randomly?
07:29:58[Saint]There's usually a bunch of symlinks that point to the real or emulated storage to cope with braindead apps expecting a particular path.
07:30:04ParkerROh nvm
07:30:12ParkerR woo
07:30:18[Saint]ParkerR: I put up an antialiased font pack a few years ago.
07:30:31ParkerROoh I see now
07:30:42 Join mortalis [0] (~kvirc@
07:30:58[Saint]Oh. Wow. Weird.
07:31:03[Saint]Its got my real name up there.
07:31:16*[Saint] doesn't recall writing that, maybe he didn't.
07:31:42ParkerRHayden Pearce ?
07:32:08[Saint]The one and only.
07:32:30[Saint](actually, there's several of us, even one that has my exact same middle initials and birthdate)
07:32:35ParkerRWell it is a wiki
07:32:41ParkerRSee who made the edit lol
07:33:01[Saint]Me, apparently. Heh.
07:33:13ParkerR3rd person mmmmm
07:33:42[Saint]I've been told that some archivers fricken HATE that zip.
07:33:55[Saint]If you have issues, lemme know and I'll repackage it.
07:33:57ParkerR[Saint], wth kind of compression did you use
07:33:58ParkerR skipping: 10 Ubuntu [Bold].fnt need PK compat. v6.3 (can do v4.6)
07:34:17ParkerRAhh ok
07:34:34ParkerR[parker@localhost Downloads]$ 7z e
07:34:36ParkerRThat worked
07:34:59[Saint]I crushed the almighty hell out of that zip iirc.
07:35:00ParkerRYeah I tried in ES on the phone and it just wouldnt extract
07:35:46[Saint]AH, yeah, ~130MB of fonts in ~14MB of zip.
07:35:59*[Saint] seems to recall that taking _quite_ a while to compress.
07:37:28[Saint]There's GNU Unifont if you need extensive locale support, and three variants of Ubuntu if you want a simple, pretty font.
07:38:01ParkerRRight now Im struggling to get adb to accept wildcards :P
07:38:08[Saint](the GNU unifont ones have like 20K glyphs each or something, hench the ~30MB filesize on the larger ones.
07:41:05HeadWhackerI can see the default Android target only has 8MB of allocated memory. Is there a reason/benefit to go more than that size?
07:42:30[Saint]larger buffer means that you'll hit the disk a lot less, obviously, but this isn't really a concern on flash based targets.
07:42:43ParkerR[Saint], Much better :)
07:42:50[Saint]Having a larger allocation also makes it more likely you'll be the one that gets killed in an OOM situation.
07:44:58[Saint]Rockboxed *is* loading a backdrop.
07:45:07[Saint]How the fuck are you getting away with tis?
07:45:24[Saint]ParkerR: what's the version of this binary?
07:45:40ParkerRAndroid? I just pulled from rasher
07:46:45***Saving seen data "./dancer.seen"
07:47:22[Saint]I don't get it. The backdrop *must* match the binary resolution, and the binary resolution *must* match the framebuffer.
07:47:30[Saint]By all rights, this should fall over.
07:47:39ParkerRI used the iPod Video Rockboxed
07:47:46[Saint]can you screencap the wps for me please?
07:48:00[Saint]right, I'm looking at that wps now.
07:48:01ParkerRNow playing?
07:48:15ParkerRJust need to get an mp3 lol Hadnt actually put any msuic on it yet
07:48:49[Saint]something almost certainly shat some media in /sdcard/media*
07:49:03[Saint]Hangouts, Facebook, Twitter, etc.
07:49:13ParkerRI had sonic drowning in Downloads heh
07:49:44ParkerRYeah WPS is funky but the touch regions work
07:50:12[Saint]Applications can dump system notification in /sdcard/media/{ringtones/alarms/notifications} or just sdcard/{Alarms/Notifications/Ringtones}.
07:50:43[Saint]That's very...odd.
07:50:47[Saint]It should fall over.
07:51:23ParkerR[Saint], published that theme your working on at all?
07:51:40ParkerRId be happy to test
07:52:04[Saint]No, not as yet. I finally finished it this afternoon. Well, some resemblance of finished.
07:52:13[Saint]Just working on locale support and cleanup now.
07:52:39ParkerRWell Im us english so dont have to worry about me :P
07:52:53[Saint]Then I'll put it up and with any luck hopefully someone'll step in and help port it to the numerous obscure locations.
07:53:06[Saint]By design its largely resolution agnostic though.
07:53:26[Saint]Its capable of fitting to multiple resolutions, within reason.
07:53:59[Saint]No backdrop, very little fixed width/placment definitions, and a LOT of negative offset placement.
07:54:01ParkerRI know it would mean more coding but I wonder if a native Ui could be done for rockbox. Like just use the codecs (I may be talking out of my ass)
07:54:23[Saint]No, you're absolutely on the right track. And, its very possible.
07:54:41[Saint]We already have a minimally functional stanalone core playback library.
07:55:02[Saint]Many people never heard of the project at all.
07:55:14[Saint]Its called Warble, and it was a Google SoC project.
07:55:15ParkerRFor archaic format support I usually use deadbeef (Can play those video game music files)
07:55:48[Saint]Rockbox plays .mod - what more do you need? ;)
07:56:25ParkerRWell it was the type where it was literaly the music ripped from the rom
07:56:30ParkerRSo it actual emualted it to some extent
07:56:50[Saint]JdGordon: I am correct in thinking that a theme _absolutely_ should fall over if it attempts to load a backdrop that isn't the exact size of the framebuffer, yes?
07:57:20[Saint]the above situation is confusing the shit out of me.
07:57:38ParkerR[Saint], Extracting 01 - Opening Theme.vgm
07:57:52ParkerRStuff like that :)
07:58:26[Saint]I'm just going through now and making sure that everything works the way it should with RTL text and viewport flipping.
07:58:51[Saint]And translation where applicable.
07:59:38[Saint]ParkerR: is it immediately obvious where my theme inspiration came from do you think?
07:59:50[Saint]I wanted to modernize Rockboxed.
08:00:00ParkerRYeah Android pretty much
08:00:34[Saint]I stole the scrubbers from Android, blatantly.
08:00:43[Saint]And the switch toggles.
08:00:48*[Saint] shrugs
08:00:58[Saint]Great artists steal, etc.
08:01:14ParkerRProgrammers too :P
08:02:08ParkerRDat music
08:03:31[Saint]HeadWhacker: another downside of increasing the buffer is obvious latency with DSP effects.
08:03:48[Saint](as you'll need to wait for the useful buffer to run out before the effect is applied)
08:03:56[Saint]more buffer, more wait time.
08:05:44ParkerR[Saint], Just got a new phone and was playing around with some apps. I was like heh almost forgot about rockbox
08:06:06[Saint]Ah, yes. Moto G, IIRC>
08:06:22[Saint]Welcome to the world of vaguely current Android.
08:06:32ParkerRIm on 4.4.4...
08:06:35[Saint]You must've been fairly shocked.
08:06:36ParkerRPaaoid Android
08:06:55ParkerR[Saint], Naah Ive had the Nexus 7 2012 for newer stuff :P
08:07:06ParkerRI am shocked at good IO speeds though...
08:07:19ParkerRMoto G blows the nexus 7 away at IO
08:07:24ParkerRAt least the 2012 model
08:12:50HeadWhackerSaint: thanks, I think 16MB is the right balance for DX90.
08:13:01HeadWhackerI have been listening in the past 30 minutes and don't the noise yet and I watch the buffer emptied and refill many times over.
08:13:23HeadWhacker I don't know why mem was set to 64MB from the DX50 source/port.
08:13:28[Saint]Where is it you're increasing the buffer?
08:13:45[Saint](just want to make sure you're not irrelvantly increasing the plugin buffer)
08:15:29JdGordon[Saint]: with the %X tag yes
08:15:55[Saint]then the above makes no sense.
08:16:24HeadWhackerIt's not the plug in buffer, it's the MEMORYSIZE from the Makefile
08:16:53HeadWhackerI can see Rockbox use a larger audio buffer with large memory available
08:18:29HeadWhackerI am looking at the code yet where to limit the audio buffer to a fixed size regardless of the available mem
08:18:35[Saint]That's a bit of an ugly way of doing it.
08:18:42[Saint]You should probably be looking at /android/src/org/rockbox/
08:18:55[Saint] Math.max(32<<10, 2*getMinBufferSize(samplerate, channels, encoding));
08:19:31HeadWhackerok I'll try that thanks
08:19:36[Saint]perhaps try 4*
08:19:56[Saint]though I can't honestly think of a logical reason for this.
08:20:36 Quit ygrek (Ping timeout: 255 seconds)
08:35:29 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
08:44:37 Join Zagor [242] (~bjst@rockbox/developer/Zagor)
08:51:47 Quit Jinx (Ping timeout: 260 seconds)
08:52:06JdGordon[Saint]: remind me tomorow or wednesday, going afk for the rest of the day
08:57:52 Join Jinx [0] (Dojo@unaffiliated/jinx)
08:59:28 Quit Zagor (Ping timeout: 256 seconds)
08:59:44JdGordon[Saint]: looks like not much error checking happens there
08:59:48 Join Zagor [0] (~bjst@
08:59:59 Quit Zagor (Changing host)
08:59:59 Join Zagor [242] (~bjst@rockbox/developer/Zagor)
09:00:00[Saint]It used to spit.
09:00:35[Saint]if the backdrop was even 1px out in any direction it would fall over on parse.
09:01:07JdGordonyeah, probably before i rejigged the backdrop sharing magic
09:01:20JdGordonfile a bug, and/or email me/dev-ml to remind me to do something about that
09:01:27[Saint]will do.
09:01:31[Saint]and thanks.
09:21:27 Join ygrek [0] (~user@
09:22:46 Quit pamaury (Ping timeout: 264 seconds)
09:45:50 Quit Zagor (Remote host closed the connection)
09:46:23 Join Zagor [242] (~bjst@rockbox/developer/Zagor)
09:46:48***Saving seen data "./dancer.seen"
09:52:10 Quit HeadWhacker (Ping timeout: 246 seconds)
09:56:15 Join petur [0] (5bb7304d@rockbox/developer/petur)
09:57:28 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
10:22:05 Quit pystar89 (Ping timeout: 250 seconds)
10:26:18 Join pystar89 [0] (
10:28:12 Join AlexP [0] (~alex@rockbox/staff/AlexP)
11:34:31 Join kugel [0] (
11:34:31 Quit kugel (Changing host)
11:34:31 Join kugel [0] (~kugel@rockbox/developer/kugel)
11:46:52***Saving seen data "./dancer.seen"
12:31:45 Join ikeboy [0] (
12:39:17 Join kuldeepdhaka [0] (~kuldeepdh@unaffiliated/kuldeepdhaka)
12:55:24kugelgevaerts: have you seen FS #12990 ?
12:55:25fs-bluebot 3PANIC in usb_storage_init_connection on iPod Classic (bugs, unconfirmed)
12:59:13*[Saint] is reminded of TheSeven's USB driver rework.
12:59:48[Saint]I was supposed to it.
12:59:56[Saint]Ah. Bitfields.
13:00:07 Quit ikeboy (Quit: Leaving)
13:11:06 Join HeadWhacker [0] (84934c34@gateway/web/freenode/ip.
13:16:26 Quit HeadWhacker (Quit: Page closed)
13:29:24 Join Ra_ [0] (75c2e80c@gateway/web/freenode/ip.
13:30:50Ra_is there any way to use rockbox on iPod classic 6G without emCORE?
13:32:22coppernot yet
13:36:47 Part Ra_
13:46:55***Saving seen data "./dancer.seen"
13:57:41 Quit ygrek (Ping timeout: 245 seconds)
14:00:03 Quit kuldeepdhaka (Ping timeout: 250 seconds)
14:00:30 Join kuldeepdhaka_ [0] (~kuldeepdh@unaffiliated/kuldeepdhaka)
14:10:08 Join scorche [0] (~scorche@rockbox/administrator/scorche)
14:13:06 Quit kuldeepdhaka_ (Ping timeout: 245 seconds)
14:15:01 Join sLite_ [0] (
14:15:09 Join copper_ [0] (~copper@unaffiliated/copper)
14:15:44 Join charlie_ [0] (~c@
14:16:08 Nick charlie_ is now known as Guest76678 (~c@
14:18:41 Join Xyem_ [0] (
14:19:07 Quit Xyem (*.net *.split)
14:19:07 Quit scorche` (*.net *.split)
14:19:09 Quit copper (*.net *.split)
14:19:11 Quit sLite (*.net *.split)
14:19:14 Quit charlie (*.net *.split)
14:19:23 Nick copper_ is now known as copper (~copper@unaffiliated/copper)
14:56:17 Join kuldeepdhaka [0] (~kuldeepdh@unaffiliated/kuldeepdhaka)
14:57:15 Join amayer [0] (
15:15:12 Join maruk1 [0] (
15:31:26 Quit alucryd (Ping timeout: 240 seconds)
15:32:12 Join alucryd [0] (~alucryd@archlinux/trusteduser/alucryd)
15:35:50 Quit Strife89 (Ping timeout: 255 seconds)
15:46:59***Saving seen data "./dancer.seen"
16:00:20 Join lleeloo [0] (b27d9335@gateway/web/freenode/ip.
16:02:40lleelooanyone alive here?
16:10:30[Saint]Just ask your question.
16:10:52[Saint]If someone can assist they will.
16:14:52lleelooI made Cortex-M4 port of rockbox(dummy codec, SD LCD320x240, SRAM 16M). But my code a little ugly in file structure. Someoune who intersting in this, I need adviсe. (Sorry for my English)
16:24:19 Join wodz [0] (
16:25:02wodzlleeloo: What kind of advice?
16:25:39lleeloofile naming and file tree structure
16:26:17lleeloohow insert V7M specific code to common parts of kernel&
16:27:07lleeloohow to separate V7M asm from common ARM asm files?
16:27:26lleelooconfigure script advices=(
16:30:02wodzIts a bit complicated. The starting point is firmware/export/config.h where different macros are defined. You are interested in ARM_ARCH I think
16:30:16wodzand then usual #ifdefs in kernel code
16:31:43wodzthe ARCH_VERSION is set in tools/configure individually for every supported target
16:32:40lleelooARCH_VERSION is 7 and is cortex-m4 is v7m
16:34:27lleeloowhen configure finished, ARCH_VERSION = 7, but 7 is cortex-r(a) too/ and thats why i define in my config.h cpu=STM32F4
16:35:14 Quit mortalis (Ping timeout: 255 seconds)
16:35:51lleeloomy working day finished. I'l come back little later when i came home
16:35:57 Quit lleeloo (Quit: Page closed)
16:36:02 Quit kuldeepdhaka (Ping timeout: 240 seconds)
16:50:32 Quit Scall (Ping timeout: 255 seconds)
16:52:52 Join ygrek [0] (~user@
16:53:39wodzHmm, actually this is interesting problem. And actually this is v7e-m architecture
16:57:21 Quit Zagor (Quit: Clint excited)
16:57:35 Join Scall [0] (~chat@unaffiliated/scall)
17:03:37wodzdefining ARM_PROFILE would be the most general
17:04:30dongsi tried rockbox on assdroid teh other day, and all it did was blank screen
17:08:46pamauryIndeed for ARM v7+ we would need the profile
17:09:28pamauryI wonder what device this is
17:13:45wodzI guess dev board or something
17:13:59wodzM4 is not very popular in consumer electronic market
17:16:14dongswat it m4
17:18:58 Join benedikt93 [0] (~benedikt9@unaffiliated/benedikt93)
17:26:32 Quit kugel (Ping timeout: 240 seconds)
17:43:20 Quit SolarAquarion (Read error: Connection reset by peer)
17:45:00 Quit petur (Ping timeout: 246 seconds)
17:46:40 Join kuldeepdhaka [0] (~kuldeepdh@unaffiliated/kuldeepdhaka)
17:47:00***Saving seen data "./dancer.seen"
17:56:40 Quit pamaury (Ping timeout: 256 seconds)
18:06:24 Quit benedikt93 (Quit: Bye ;))
18:20:42 Quit maruk1 (Quit: Leaving.)
18:28:45 Join ivanf [0] (~ivan@unaffiliated/kferdous)
18:28:50ivanfHiya all
18:30:44 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
18:34:01wodzpamaury: I fully decoded CP3 upgrade file :-)
18:34:35pamaurywodz: cool :) so how does it work ?
18:37:18wodzpamaury: The .ROM file is actually RKW but a 'complete one'. It has various parts: 1) RKRS section which is sort of script what upgrade routine is supposed to do with this file. 2) RKST section which is kind of custom archive. 3) Additional resources where bootloaders and helper recovery binaries are stored.
18:40:06wodzThe upgrade routine follows the directives from RKRS and 1) check the version 2) check nand bootloader version and upgrade if needed 3) format system partition 4) copy files onto system partition 5) unpack Rock27Boot.bin file 6) reset
18:41:54wodzThe interesting thing is that Rock27Boot.bin contains the same bootloaders (s1 and s2) as stored in 'extended' resources in the .ROM file itself
18:47:34pamaurynot sure I understand everything ^^ So it's a bit different from other rk27xx devices ?
18:50:35wodzThis seems to be standard upgrade procedure in SDK but Hifiman for example doesn't follow this scheme
18:51:17 Join ikeboy [0] (
18:55:43 Join rela [0] (~x@pdpc/supporter/active/rela)
18:57:04pamauryI'm still working on the Sansa View, the code is a nightmare to RE, so much C++ :(
18:59:02 Quit the-kyle (Ping timeout: 240 seconds)
19:00:47 Join the-kyle [0] (
19:05:38 Join jhMikeS [0] (
19:05:38 Quit jhMikeS (Changing host)
19:05:38 Join jhMikeS [0] (~jethead71@rockbox/developer/jhMikeS)
19:06:17 Quit wodz (Remote host closed the connection)
19:13:37 Join bertrik [0] (~quassel@rockbox/developer/bertrik)
19:15:44 Join lebellium [0] (
19:25:02 Quit ygrek (Ping timeout: 240 seconds)
19:39:41 Quit AlexP (Remote host closed the connection)
19:39:59 Quit bluebrother (Disconnected by services)
19:40:04 Join bluebrother^ [0] (~dom@rockbox/developer/bluebrother)
19:40:55 Join AlexP [0] (~alex@rockbox/staff/AlexP)
19:42:33 Quit fs-bluebot (Ping timeout: 255 seconds)
19:43:47 Quit kuldeepdhaka (Ping timeout: 255 seconds)
19:45:53 Join fs-bluebot [0] (
19:47:01***Saving seen data "./dancer.seen"
19:53:13 Join SolarAquarion [0] (~SolarAqua@unaffiliated/solaraquarion)
20:14:50 Quit ikeboy (Ping timeout: 255 seconds)
20:43:50 Join lleeloo [0] (25d7a744@gateway/web/freenode/ip.
20:46:47lleelooSo about cortex-m4: what is the best way to identify v7e-m in export/config.h?
20:47:22lleeloothis define will help me to use #ifdef in code and source include file list
20:50:13lleelooand at that moment and edit tools/configure and add code for my devboard with target ID=90. Is this ID free or not?
20:51:53pamaurylleeloo: I think you need to make sure the configure set ARCH_VERSION/ARM_ARCH to 7, but you should also create some #define for the profile, maybe something like ARM_PROFILE and define ARM_PROFILE_M as a possible value. In configure you also need to create your own function, just like arm1176jzscc() for example
20:52:31lleelooi add code to scrambler tool and write bootloader for demoboard. What should I read (or know) to add this files to rockbox tree
20:53:01pamaurythen you can create a target which uses this function and see
20:53:09lleelooconfigure already work! i can use this script to configure Rockbox sources for my board
20:54:40lleelooat this moment i use configure to build bootloader and main rockbox firmware. But CPU identification done by defining cpu=stm32f4 =(
20:55:09pamaurythat's normal
20:55:24pamauryoh you mean the cpu given to the compiler ?
20:56:49lleeloono? i mean that in config.h I do next: #define CONFIG_CPU STM32F4
20:57:29lleelooand lateer in source code i use this define to identify cortex-m4 code
20:58:26lleelooin configure i use armcortexM4Fcc () { findarmgcc GCCOPTS="$CCOPTS -mthumb -march=armv7e-m -mtune=cortex-m4" GCCOPTIMIZE="-fomit-frame-pointer" endian="little" }
20:58:38pamaurythat's normal
20:59:01pamauryCONFIG_CPU should be read the soc really, it's more than the CPU architecture
20:59:16pamauryand for some reason we do the cpu architecture in configure and the soc define in config
21:00:48lleelooso configure set ARCH_VERSION/ARM_ARCH to 7. so i try to define ARM_PROFILE_M in my config.h. is it a good way?
21:01:34lleelooso if i use STM32F417 what must be defined as CONFIG_CPU?
21:01:40pamauryI think it would be better if the profile is set by configure script
21:01:51pamauryCONFIG_CPU should be STFM32F417
21:02:14pamauryor perhaps a slightly more general name, like STFM32F4xx, depending on how general your code is
21:02:15lleeloo"<pamaury> CONFIG_CPU should be STFM32F417" OK I'll do that
21:02:48pamaurylike me just check I'm not saying any mistake ;)
21:03:05lleelooWhat should i do to make configure set ARM_PROFILE_M?
21:04:26pamaurygood question, let me check
21:04:31lleelooi mean how configure make define like that
21:05:42pamauryit's a bit tricky
21:05:53pamauryconfigure generates the makefile, which contains some exported variables
21:06:13pamauryiirc, thos variables are used by the makefile to generate config.h which contains a subset of those variables
21:07:36pamaurysorry autoconf.h
21:08:01pamauryoh actually autoconf.h is generated by configure so it's much easier
21:09:08pamauryYou should add a line below /* Optionally define architecture version */ ${Darch_version} with your custom define I guess
21:10:07pamauryI think it would be better to have some define like ARM_PROFILE which be ARM_M or ARM_A for example, where ARM_M and ARM_A are predefined in autoconfig.h
21:11:53lleeloook, i'm not so good in makefiles=( I'm like beginner. May you explain me more details? Where and what I should write
21:13:02pamauryok, give a minute, I'll write something
21:14:37lleeloook, thank you for you patience!
21:15:06lleelooyour patience
21:15:30pamaurylleeloo: by the way, what does the FCC stands for after the cortexM4 ?
21:18:13lleelooactually it's M4F cross compiller (M4FCC in short). I can rena it if it is mistake! My English is not so good as i want=(
21:19:25pamauryI think cortexm4 would be more correct, it's the name of the cpu, not the name of the toolchain in theory
21:20:03lleeloook, I'll FIX this!
21:25:35pamauryI didn't try it, but it should give you the idea
21:28:58lleelooi think i understand what did you mean! I will try this way. I one more question. How my code will sync with main rockbox sources? If new version of RB will release, my local version will not be correct=(
21:31:04pamaurybest is to follow and push your patch to gerrit ( This way anyone can see your changes, comment on them, and your code is not lost. And if the code is fine, we could consider it for inclusion in our repository
21:32:52 Join smokesig [0] (
21:32:57 Part smokesig ("Konversation terminated!")
21:34:30lleelooI will read about UsingGit. I never use a patches =( Now it's time=) OK i will try
21:35:29lleelooThank you very much ащк нщг рудз
21:35:39lleeloofor you help
21:47:04***Saving seen data "./dancer.seen"
22:06:06 Quit lleeloo (Ping timeout: 246 seconds)
22:10:58pamauryyeah, finally found how to select between internal storage and sd card on the Sansa View \o/
22:23:24 Join lleeloo [0] (25d7b2c4@gateway/web/freenode/ip.
22:24:09 Join einhirn [0] (
22:25:26 Join einhirn_ [0] (
22:28:31 Quit einhirn (Ping timeout: 245 seconds)
22:32:16 Quit pamaury (Ping timeout: 245 seconds)
22:37:57 Quit lleeloo (Ping timeout: 246 seconds)
22:41:42 Join sakax [0] (~sakax@unaffiliated/sakax)
22:56:50 Join wodz [0] (
23:07:47 Quit rela (Quit: Leaving)
23:13:21 Join prof_wolfff [0] (
23:17:11 Quit amayer (Quit: Leaving)
23:32:40 Join petur [0] (~petur@rockbox/developer/petur)
23:32:47 Quit petur (Client Quit)
23:39:08 Quit wodz (Quit: Leaving)
23:45:08 Quit einhirn_ (Quit: Miranda IM! Smaller, Faster, Easier.
23:47:06***Saving seen data "./dancer.seen"
23:49:11 Quit lebellium (Quit: ChatZilla [Firefox 32.0/20140731191115])

Previous day | Next day