Previous day | Jump to hour: 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 | Next day

Seconds: Show Hide | Joins: Show Hide | View raw
Font: Serif Sans-Serif Monospace | Size: Small Medium Large

Click in the nick column to highlight everything a person has said.
The Logo icon identifies that the person is a core developer (has commit access).

#rockbox log for 2012-01-07

00:01:34 Quit pamaury (Remote host closed the connection)
00:03:27 Quit Topy (Ping timeout: 240 seconds)
00:05:01 Quit dfkt (Ping timeout: 252 seconds)
00:09:33 Join nosa [0] (
00:11:48 Quit nosa-j (Ping timeout: 248 seconds)
00:11:49 Nick nosa is now known as nosa-j (
00:14:54 Quit benedikt93 (Quit: Bye ;))
00:15:48 Quit lovasoa (Remote host closed the connection)
00:27:04 Join [Saint_] [0] (~Saint]@unaffiliated/saint/x-8516940)
00:38:28 Quit bertrik (Read error: Operation timed out)
00:54:36 Quit n1s (Quit: Ex-Chat)
00:55:15 Join anewuser [0] (~anewuser@
00:55:15 Quit anewuser (Changing host)
00:55:15 Join anewuser [0] (~anewuser@unaffiliated/anewuser)
00:57:34 Quit factor (Ping timeout: 240 seconds)
00:59:36***Saving seen data "./dancer.seen"
01:15:54 Join domonoky1 [0] (
01:17:07 Quit domonoky (Ping timeout: 240 seconds)
01:34:59 Quit lebellium (Quit: ChatZilla 0.9.88 [Firefox 10.0/20111228055358])
01:37:50 Join factor [0] (~factor@
01:42:06 Join Horscht [0] (
01:42:06 Quit Horscht (Changing host)
01:42:06 Join Horscht [0] (~Horscht@xbmc/user/horscht)
02:03:05Tuggerssaratoga_: I'm interested in doing audio codec optimizations in asm
02:05:19saratoga_why that project?
02:06:34Tuggersbecause I'm interested in embedded systems and speed optimization
02:07:17TuggersI do have MIPS experience, but AFAIK most music players use ARM
02:07:39Tuggersand from what I've seen they're fairly similar
02:07:57saratoga_somewhat, although learning arm isn't very hard
02:08:16saratoga_codecs mostly just do load > mul > store, so you need about 10 instructions
02:08:42saratoga_do you have any supported target?
02:08:55TuggersI don't
02:09:04Tuggersif my proposal is accepted I'll buy one
02:09:25saratoga_do you have a phone or mp3 player capable of running 3rd party software?
02:10:28TuggersMy dad has an android phone, and I'm sure I could get an android phone on our verizon plan if I asked
02:10:58saratoga_where are you in the US?
02:11:02Tuggersbut on the front page it says that android is "unusable"
02:11:18saratoga_well if you just want to do development it doesn't really matter how polished the software is
02:12:02saratoga_you can probably buy a sansa clip for ~ $20 if you keep your eyes open, might be a good idea
02:12:38Tuggersthe advantage of running it on android, I assume, is that you get access to all of the arm instructions up to v7
02:14:02saratoga_yeah, although for optimization purposes that may be less interesting
02:14:14saratoga_its not clear to me how useful optimizing things that run so quickly is on 1+GHz processors
02:14:32saratoga_although i suppose adding things like NEON might be interesting
02:15:05saratoga_usually though its arm7/9/11 devices that benefit the most
02:15:30saratoga_i don't know what the situation with battery life is even like on Android at this point
02:19:29Tuggerswhat would be a good way to learn about armv4/v5/v6 programming?
02:20:01Tuggersany guides? or should I just look at the ARM manuals and the Rockbox source?
02:20:02saratoga_Tuggers: best way is to learn by doing with the arm instruction reference
02:20:29saratoga_the instruction reference card linked there is really handy
02:21:59saratoga_have you done anything with codecs or dsp before?
02:22:36Tuggersbut I have programmed for the NES and its 6502-based processor
02:22:59saratoga_so i guess you've done a little fixed point stuff before?
02:23:55Tuggersin my computer architecture class I used fixed point to store a value in dollars and cents
02:24:38saratoga_which school are you at?
02:24:44Tuggersuniversity of texas at dallas
02:25:19saratoga_ah ok, i know some people at UTA
02:27:14saratoga_anyway, since thats a pretty open ended project, it wouldn't be a bad idea to look at some of the codecs and come up with a good plan
02:27:18 Quit Zarggg (Quit: Rebooting client...)
02:27:40saratoga_we get a few people apply for that project every year and most of them don't convince anyone that they can do anything useful
02:28:21Tuggersok, so what I can do now is read the arm reference and the Rockbox codec source
02:28:23Tuggersthank you
02:28:32saratoga_i think the codec source is probably more useful
02:29:34saratoga_libmad, libfaad, libalac are probably good places to start
02:31:03saratoga_ liba52 doesn't need optimization, but its such a simple codec that it also might not be a bad place to start, and i'm sure theres low hanging fruit there
02:31:30saratoga_its pretty much a textbook codec, so pretty easy to understand
02:32:10saratoga_libwmavoice is currently not converted to fixed point, so theres that as well
02:33:09 Join Zarggg [0] (
02:39:43 Quit TheLemonMan (Quit: WeeChat 0.3.6)
02:48:43 Quit domonoky1 (Read error: Connection reset by peer)
02:59:35 Join Rob2223 [0] (
02:59:37***Saving seen data "./dancer.seen"
03:01:00 Quit Rob2222 (Ping timeout: 276 seconds)
03:26:14 Join blind_tester [0] (
03:29:10blind_testerIs there a way to reset a Sansa Clip+ with a completely frozen rockbox? No keypress works. I tried the imageviewer and i saw a flickering image but after i went back to the list of entries in the directory all got frozen.
03:41:34 Join Brijesh__ [0] (Brijesh@2002:3ba1:8841::3ba1:8841)
03:42:33 Nick [Saint_] is now known as [Saint] (~Saint]@unaffiliated/saint/x-8516940)
03:45:49 Join CaptainKewl [0] (
03:46:28saratogablind_tester: power for 30 seconds is the hardware power off
03:54:05 Quit [Saint] (Remote host closed the connection)
04:02:41 Quit mystica555_ (Remote host closed the connection)
04:06:53 Quit curtism (Quit: Live Long and Prosper)
04:10:25 Join curtism [0] (
04:15:15 Quit pixelma (Disconnected by services)
04:15:17 Join pixelma_ [0] (quassel@rockbox/staff/pixelma)
04:15:19 Nick pixelma_ is now known as pixelma (quassel@rockbox/staff/pixelma)
04:16:34 Quit amiconn (Disconnected by services)
04:16:35 Join amiconn_ [0] (quassel@rockbox/developer/amiconn)
04:16:57 Nick amiconn_ is now known as amiconn (quassel@rockbox/developer/amiconn)
04:19:55 Quit CaptainKewl (Quit: ( :: NoNameScript 4.22 :: ))
04:21:08 Join [Saint] [0] (~Saint]@unaffiliated/saint/x-8516940)
04:22:39 Quit Brijesh__ (Quit: Leaving)
04:24:29 Join dys` [0] (
04:26:16 Quit dys (Ping timeout: 252 seconds)
04:27:44 Quit TheSeven (Disconnected by services)
04:27:56 Join [7] [0] (~TheSeven@rockbox/developer/TheSeven)
04:59:40***Saving seen data "./dancer.seen"
05:37:06 Join [Saint_] [0] (~Saint]@unaffiliated/saint/x-8516940)
05:39:23 Quit [Saint] (Ping timeout: 240 seconds)
05:40:31 Nick [Saint_] is now known as [Saint] (~Saint]@unaffiliated/saint/x-8516940)
05:42:19 Join [Saint_] [0] (~Saint]@unaffiliated/saint/x-8516940)
05:43:09 Nick [Saint] is now known as Guest11989 (~Saint]@unaffiliated/saint/x-8516940)
05:44:55 Quit Guest11989 (Ping timeout: 252 seconds)
05:45:29 Nick [Saint_] is now known as [Saint] (~Saint]@unaffiliated/saint/x-8516940)
05:55:40 Join Rob2222 [0] (
05:58:05 Join RusselKubes [0] (
05:59:19 Quit Rob2223 (Ping timeout: 240 seconds)
06:00:23 Join [Saint_] [0] (~Saint]@unaffiliated/saint/x-8516940)
06:01:37 Quit [Saint] (Read error: Connection reset by peer)
06:02:05 Join [Saint] [0] (~Saint]@unaffiliated/saint/x-8516940)
06:05:01 Quit [Saint_] (Ping timeout: 255 seconds)
06:06:06 Nick [Saint] is now known as [Saint_] (~Saint]@unaffiliated/saint/x-8516940)
06:06:12 Nick [Saint_] is now known as [Saint__] (~Saint]@unaffiliated/saint/x-8516940)
06:06:19 Nick [Saint__] is now known as [Saint___] (~Saint]@unaffiliated/saint/x-8516940)
06:06:24 Nick [Saint___] is now known as [Saint] (~Saint]@unaffiliated/saint/x-8516940)
06:06:24DBUGEnqueued KICK [Saint]
06:19:04 Quit [Saint] (Remote host closed the connection)
06:20:05 Join [Saint] [0] (~Saint]@unaffiliated/saint/x-8516940)
06:25:42 Quit Tuggers (Quit: Leaving)
06:26:52 Join Tuggers [0] (
06:27:35 Quit Tuggers (Client Quit)
06:28:03 Join Tuggers [0] (
06:46:48 Quit RusselKubes (Quit: Ex-Chat)
06:59:43***Saving seen data "./dancer.seen"
07:05:09 Join [Saint_] [0] (~Saint]@unaffiliated/saint/x-8516940)
07:08:17 Quit [Saint] (Ping timeout: 268 seconds)
07:13:13 Quit pjm0616 (Ping timeout: 276 seconds)
07:24:37 Join pjm0616 [0] (~user@
07:25:06 Join [Saint] [0] (~Saint]@unaffiliated/saint/x-8516940)
07:28:43 Quit [Saint_] (Ping timeout: 252 seconds)
07:54:42 Quit perrikwp (Read error: Connection reset by peer)
07:55:53 Join perrikwp [0] (
08:15:09Mirblind_tester: you could reinstall rockbox fresh
08:15:17Mirdeleteing your settings
08:15:27Miri had that issue with my H10
08:15:39Mirchanged the theme and all shit broke loose
08:19:58 Join robin0800 [0] (~robin0800@
08:22:29 Quit saratoga_ (Quit: Page closed)
08:23:51 Quit T44 (Read error: Connection reset by peer)
08:26:23[Saint] Mir: reinstalling Rockbox to reset settings is rather similar in my opinion to cutting off your finger to remove a splinter :)
08:29:53[Saint]Settings - Manage Settings - Reset Settings
08:30:50[Saint]Individual settings can be reset by bringing up the context menu on the setting you want to reset also.
08:30:56 Join Topy44 [0] (
08:35:16[Saint]I didn't see the issue that blind_tester had initially, but if its a configuration issue, the above will be plenty of info to resolve it.
08:37:08 Quit bieber (Ping timeout: 240 seconds)
08:37:35 Join bieber [0] (
08:51:58 Quit robin0800 (Ping timeout: 252 seconds)
08:52:10[Saint]Oh...I forgot about the logs. Yes, yes I did see blind_tester's issue. And configuration settings wasn't the issue or the fix in this case, it just needed to be reset (the method of which differs between players).
08:54:29 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
08:54:58 Join robin0800 [0] (~robin0800@
08:59:18blind_testerSaint: I waited until battery has gone. Now it works again. It was no configuration problem.
08:59:21blind_testerMir: I waited until battery has gone. Now it works again. It was no configuration problem.
08:59:45***Saving seen data "./dancer.seen"
09:01:37[Saint]If "long power" didn't work, then I've no idea what kind of state the player must have gotten itself into.
09:02:16[Saint]AFAIK "long power" should always work.
09:03:16[Saint]'should' being the operative word.
09:06:02blind_testerI pressed the power button for several seconds. Don't know how long "long power" is :-) ... I tried maybe 5 seconds.
09:06:26blind_testerSaint: I pressed the power button for several seconds. Don't know how long "long power" is :-) ... I tried maybe 5 seconds.
09:06:46[Saint]~30 seconds
09:07:21blind_testerah... ok thank you. I didn't find this information in the manual - but i searched only for reset! :-(
09:09:27[Saint]It should be in the manual, but I'm not sure how hard/easy it is to dig up via the search function. Especially if you're not sure which exact keyword you need to be searching.
09:10:17[Saint]Even I have trouble finding things I know exist in the manuals sometimes, human error on my part I'm sure. :)
09:11:12blind_testerSaint: There i can read: Long Power -> Shutdown Rockbox. It seems to be the normal shutdown sequence... normally i must press it ~ 2-3 seconds. This did not work.
09:13:17blind_testerSaint: I think the device was frozen. No way to get out of it. I will avoid copying images to the device in the future. This will help.
09:19:14CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
09:19:14*[Saint] isn't so sure about that.
09:20:11[Saint]The type of digital media you chose to store on the device shouldn't have anything to do with it crashing or not.
09:24:49blind_testerSaint: It crashed after the viewing of a jpg. The image flickered... after coming back to the directory list all was frozen. I think this was the problem. Ok i go now... Bye.
09:31:05 Join pamaury [0] (
09:31:05 Quit pamaury (Changing host)
09:31:05 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
09:39:32 Join n1s [0] (
09:39:32 Quit n1s (Changing host)
09:39:32 Join n1s [0] (~n1s@rockbox/developer/n1s)
09:55:39 Join RusselKubes [0] (
10:04:22 Quit perrikwp (Read error: Connection reset by peer)
10:07:48 Join perrikwp [0] (
10:10:30 Join ender` [0] (
10:14:44 Join Horschti [0] (~Horscht@xbmc/user/horscht)
10:17:48 Quit Horscht (Ping timeout: 240 seconds)
10:39:08 Quit RusselKubes (Quit: Ex-Chat)
10:40:25 Join RusselKubes [0] (
10:41:33 Quit RusselKubes (Client Quit)
10:42:02 Quit blind_tester (Quit: Verlassend)
10:42:29 Join RusselKubes [0] (
10:45:04 Quit perrikwp (Read error: Connection reset by peer)
10:46:18 Join perrikwp [0] (
10:53:01 Join stoffel [0] (
10:59:48***Saving seen data "./dancer.seen"
11:17:46 Join y4n [0] (y4n@unaffiliated/y4ndexx)
11:24:10 Quit robin0800 (Ping timeout: 252 seconds)
11:42:30 Join lebellium [0] (
12:07:37 Join robin0800 [0] (~robin0800@
12:27:48 Join dfkt [0] (dfkt@unaffiliated/dfkt)
12:30:30 Quit qnm (Ping timeout: 252 seconds)
12:32:47 Join TheLemonMan [0] (
12:33:43 Join qnm [0] (~qnm@2001:44b8:3110:f300:208:9bff:fec0:179a)
12:50:32 Quit qnm (Ping timeout: 268 seconds)
12:50:34 Quit Zambezi (Read error: Connection reset by peer)
12:52:12 Join qnm [0] (~qnm@2001:44b8:3110:f300:208:9bff:fec0:179a)
12:55:01 Quit stoffel (Remote host closed the connection)
12:55:36 Join stoffel [0] (
12:55:44 Quit factor (Read error: Connection reset by peer)
12:59:49 Join Zambezi [0] (
12:59:50***Saving seen data "./dancer.seen"
13:04:23 Quit dv_ (Read error: Operation timed out)
13:06:07 Join {phoenix} [0] (
13:09:24bluebrotheris there a way to tell sapi_voice.vbs to output PCM? I'm getting ADPCM which is a bit inconvenient.
13:13:08 Join factor [0] (~factor@
13:15:26kugelhm, embedded quesheet, nice
13:15:49kugelnow the "browse chesheet" and playlist viewer need to be merged
13:21:53bluebrotherhmm, from reading the web the output should be PCM, not ADPCM :/
13:26:53 Join martinst_ [0] (
13:42:30 Join TomColler [0] (
13:53:40bluebrotheryay, now it works on Windows :)
13:54:18bluebrotherstupid Windows putting in a different fmt header
13:57:15 Join liar [0] (
14:05:48 Nick Jack87 is now known as Jack87|Away (Jack87@nasadmin/admin/jack87)
14:08:40pixelmadfkt: hmm, not sure if a new bootloader has been released:
14:08:53pixelmait talks about being "optional"
14:09:06dfktpixelma, i checked svn, seems you have to uncomment a line for dualboot
14:09:30dfktso i guess the target status page is still somewhat correct
14:09:31pixelmawell yes, same result here :)
14:10:42pixelmaok, then I misremembered because my M5 was also the test target for the bootloader and thus having dual boot
14:10:52bluebrotherhmm, that means we need to update Rockbox Utility as well ...
14:11:12bluebrothersince I assume we can't distribute a pre-patched bootloader
14:11:23*bluebrother remembers
14:13:12pixelmabluebrother: you mean the Utility should offer both?
14:13:38bluebrotherpixelma: yes, or at least offer the possibility to patch a firmware with the dual-boot bootloader
14:13:59bluebrotherso you can either provide an OF and get dual-boot, or not provide an OF file and get single boot
14:14:24pixelmasounds fair
14:14:27bluebrotherwe could of course keep dual-boot for those targets as advanced only that is only supported if you install manually yourself
14:15:33bluebrotherthough for now I'm trying to get libmp3lame support going. Way simpler setup than using lame as executable
14:18:15KohlrabiIs there a page which lists the current build number of bootloaders?
14:18:55KohlrabiSomeone told me USB support was fixed for Clip+, and I recently upgraded Rockbox on my device.
14:19:20KohlrabiBut I don't want to run another installation just to find out nothing has changed :)
14:20:40dfktKohlrabi -
14:20:50bluebrotheryou can check the timestamp of the bootloader on the server and compare to the timestamp Rockbox Utility stored
14:21:43Kohlrabiah, OK, thanks
14:21:54pixelmadfkt: but has a new bootloader been released? ;)
14:22:10bluebrothersince we can't extract the version number from the binary we store the server timestamp. Which should be unique enough
14:22:25pixelmaor does it need to be compiled on your own
14:22:29dfktpixelma, i do not know these things, always build them myself :)
14:23:17dfktfunman released a "v4.0" bootloader for the ams sansas, which should be the same as last svn
14:26:28 Quit TheLemonMan (Ping timeout: 240 seconds)
14:27:47bluebrotherinteresting −− using espeak via SAPI I get errors in the created files.
14:28:06bluebrothermaybe that's related to the long-standing voicefile bug?
14:36:37 Quit user890104 (Remote host closed the connection)
14:37:47 Join user890104 [0] (
14:49:08 Quit stoffel (Ping timeout: 248 seconds)
14:49:37 Quit martinst_ (Ping timeout: 240 seconds)
14:59:52***Saving seen data "./dancer.seen"
15:02:18 Quit {phoenix} (Remote host closed the connection)
15:04:53 Join rarog [0] (
15:21:18 Quit robin0800 (Quit: Leaving)
15:22:04 Join robin0800 [0] (~robin0800@
15:25:05 Quit ender` (*.net *.split)
15:25:05 Quit kadath (*.net *.split)
15:25:08 Quit maraz (*.net *.split)
15:25:39 Join ender` [0] (
15:25:39 Join kadath [0] (
15:25:39 Join maraz [0] (
15:26:22 Join Ave [0] (
15:26:40Avehi, I have a sansa device that always autocreates few directories when its in usb mode
15:26:59Aveis there a way to remove them somehow automatically too, like scripting
15:28:29kugela lua script can do it probably, but you need to run it manually
15:32:44dfktthe disktidy plugin could remove them as well
15:32:46Aveno "autoexec"
15:33:41Avedisktidy sounds good
15:34:38[Saint]Autorock dicktidy
15:35:49Aveis autorock enabled in normao builds
15:35:56[Saint]I can't remember the exact naming scheme for autorock, but, someone will.
15:35:58kugelno it isnt
15:36:31[Saint]Its not? Hmmm.
15:38:14 Join Xerion [0] (
15:38:33[Saint]Another alternative is making them hidden folders with a windows system, using the "supported" file view mode, and just forgetting about them.
15:38:44[Saint]They do no harm.
15:39:18AveI tried attrib +h
15:39:31Avewell kinda, since I'm on linux ..
15:39:42[Saint]Yeah...that doesn't work.
15:39:59 Join perrikwp_ [0] (
15:40:07Aveso what do to you then, if not +h
15:40:36[Saint]Make them hidden files with a windows system, as I said.
15:40:44Aveisnt that what attrib +h does :D
15:40:52AveI mean I know of other way
15:41:00Avedont know
15:41:44Aveyeah they do no harm, but its annoying having stuff like "audible" and "audiobooks" in there that you dont use
15:42:14 Part TomColler
15:42:36dfktif you only use rockbox usb mode with a current bootloader/revision, those folders shouldn't be recreated, afaik
15:42:46 Quit perrikwp (Ping timeout: 252 seconds)
15:43:08dfktat least on my clip+ and zip they stay gone for good
15:43:14Avewell thats the thing, whenever I connect usb cable, the device boots into some "native" usb mode
15:43:17Aveits not rockbox usb
15:43:26dfktthen maybe upgrade your bootloader
15:43:55Aveok lets do that
15:47:53 Join lorenzo92 [0] (
15:48:45AveI have clip+ as well
15:49:32dfktyou can use my prebuilt bootloader if you want -
15:50:11dfktoops, wrong link
15:51:20[Saint]Just remember "third party stuff, if it blows up its your support..blah blah disclaimer".
15:52:08dfkti used make && make bulletproof ;)
15:55:29 Join domonoky [0] (~Domonoky@rockbox/developer/domonoky)
15:57:04Aveok, now everything is updated, it doesnt reboot anymore, but I dont see usb mode either
16:08:11AveI see the usb symbol and its charging, but no mass storage
16:11:15 Quit prof_wolfff (Quit: CGI:IRC (EOF))
16:14:00 Quit lorenzo92 (Remote host closed the connection)
16:17:24 Join TheLemonMan [0] (
16:19:30bluebrotherdfkt: we have a "make bulletproof" target? ;-)
16:31:20 Join martinst_ [0] (
16:34:05 Join mortalis [0] (~mortalis@
16:34:14bluebrotherhmm, do the voice scripts handle english-us separately these days?
16:34:59 Quit Zambezi (Ping timeout: 268 seconds)
16:35:05 Quit pamaury (Remote host closed the connection)
16:38:14 Join Highlander [0] (~Connor_Ma@2a01:e35:2ec2:dcd0:3544:c31:5bce:bd36)
16:48:18 Join Zambezi [0] (
16:48:43 Join pamaury [0] (
16:48:43 Quit pamaury (Changing host)
16:48:43 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
16:49:18 Quit TheLemonMan (Quit: WeeChat 0.3.6)
16:52:26 Join stoffel [0] (
16:59:54***Saving seen data "./dancer.seen"
16:59:59 Join Brijesh222 [0] (Brijesh@
17:00:13 Quit rarog (Quit: ChatZilla 0.9.88 [SeaMonkey 2.4.1/20111011102430])
17:06:21 Quit robin0800 (Ping timeout: 260 seconds)
17:12:37 Join benedikt93 [0] (~benedikt9@unaffiliated/benedikt93)
17:18:00 Join robin0800 [0] (~robin0800@
17:25:18 Join Keripo [0] (
17:26:44 Quit Highlander (Quit: ChatZilla 0.9.88 [Firefox 9.0.1/20111220165912])
17:27:56mortalisWhat's the difference between HAVE_LCD_SLEEP, HAVE_LCD_SHUTDOWN and HAVE_LCD_ENABLE?
17:29:21CIA-88New commit by pamaury (r31596): fuze+: #define TARGET_EXTRA_THREADS because of the extra touchpad thread. Should fix mpeg player not working
17:30:27pamaurymortalis: not sure but: HAVE_LCD_SLEEP if you implement lcd_sleep (put lcd in sleep state but still active I guess). HAVE_LCD_ENABLE if you implement lcd_enable (can turn lcd on and off). Don't know of HAVE_LCD_SHUTDOWN
17:31:26CIA-88r31596 build result: All green
17:32:55kugelmortalis: LCD_SLEEP is deep sleep that might take long to wake up from(it comes with a setting to turn off), LCD_ENABLE is the same but no setting waking up is fast (you can also use both if you have some light sleep and deep sleep). LCD_SHUTDOWN is entirely different, there's a shutdown function that needs to be called when the player shuts down (so the opposite to init())
17:33:30 Nick amithkk is now known as rolandixor (u4289@2buntu/writers/amithkk)
17:34:00 Nick rolandixor is now known as Guest19592 (u4289@2buntu/writers/amithkk)
17:34:11 Nick Guest19592 is now known as amithkk (u4289@2buntu/writers/amithkk)
17:34:23 Nick amithkk is now known as jrgifford_ (u4289@2buntu/writers/amithkk)
17:34:32 Nick jrgifford_ is now known as amithkk (u4289@2buntu/writers/amithkk)
17:34:54kugelwhat "long time to wakeup" means is up to you :)
17:35:42 Quit Keripo (Quit: Leaving.)
17:36:30mortalisnow I understand, thanks
17:37:49jhMikeSin this day and age do we still need TARGET_TREE?
17:49:20kugeljhMikeS: did it ever have a point?
17:50:42funmangit grep says on lcd-remote.h use it, once
17:51:44jhMikeSkugel: when it was introduced probably
17:52:19funmanso no it's not needed anymore
17:52:32funmanJdGordon: ping
17:52:36jhMikeSprobably not. I think it must have been 7 years ago or so.
17:53:14funmanjhMikeS: you can remove it
17:53:26funmanall config.h probably need some cleanup
17:54:43funmanlcd-remote-target.h can also be reduced
17:55:11jhMikeSI noticed the hosted stuff doesn't have it defined in the configs
17:55:35funmanjhMikeS: i can remove it if you want
17:56:46jhMikeSI get 73 hits in export/config but have 81 headers
17:57:55jhMikeSno archos defines it, but clearly they're using it
17:59:20 Quit martinst_ (Ping timeout: 252 seconds)
18:00 doesn't appear to be doing anything anymore
18:05:25jhMikeSfunman: go ahead. I was right in the middle of something else anyway but I just thought to ask out of nowhere
18:06:07pamauryhas anyone seen low_light recently ?
18:10:15funmanpamaury: i think logbot has a 'seen' command (/msg logbot seen low_light)
18:11:04 Quit liar (Read error: Connection reset by peer)
18:11:13 Nick amithkk is now known as rolandixor (u4289@2buntu/writers/amithkk)
18:11:34 Nick rolandixor is now known as amithkk (u4289@2buntu/writers/amithkk)
18:12:03 Join liar [0] (
18:13:38 Quit stoffel (Ping timeout: 252 seconds)
18:14:42CIA-88New commit by funman (r31597): TARGET_TREE is not needed anymore ...
18:16:28CIA-88r31597 build result: All green
18:17:59 Quit RusselKubes (Remote host closed the connection)
18:23:20jhMikeS /msg logbot seen jhMikeS => Yeah, I see you. Boy, you're ugly ;) @@
18:24:38 Join captainkwel [0] (
18:36:02 Join jlbiasini [0] (
18:37:08jlbiasinipamaury: I saw you are committing again, could you also commit my plugin keymaps patch and the little power-volume button bug fix?
18:37:13bertrikIs anybody still hearing a "zip" sound on AMSv2 when pausing?
18:39:13 Join remlap1 [0] (
18:39:54jlbiasinipamaury: the keymaps patch for plugin is here :
18:41:08 Quit remlap (Ping timeout: 252 seconds)
18:43:40jlbiasiniand the fix for power and volume keys is just above except you wanted to change the comment fallback to fallthrough
18:46:06 Quit robin0800 (Quit: Leaving)
18:51:16 Join AlexP_mob [0] (~AlexP@rockbox/staff/AlexP)
18:51:33 Quit AlexP_mob (Client Quit)
18:51:39jhMikeSbertrik: because of fading? fading a sine wav would create that
18:51:56pamauryjlbiasini: ah yeah, forgot that. Give me two minutes :)
18:52:04 Join AlexP_mob [0] (~AlexP@rockbox/staff/AlexP)
18:52:51bertrikjhMikeS, I remember it being quite conspicuous but it haven't noticed it anymore for a while now
18:53:46jhMikeSI don't hear one with normal music currently
18:55:13jhMikeSmight have been the old volume control fading method
18:59:18bluebrotherdo HWCODEC targets automatically reload the voice file after USB (dis-)connect?
18:59:41 Join stoffel [0] (
18:59:56***Saving seen data "./dancer.seen"
19:00:58CIA-88New commit by pamaury (r31598): fuze+: Update plugin keymaps (FS #12405). Thanks to Jean-Louis Biasini
19:01:03CIA-88New commit by pamaury (r31599): fuze+: workaround the power/volume+ multiplexing that would trigger a false power button detection on heavy usage (FS #12405). Thanks to Jean-Louis ...
19:02:02 Quit remlap1 (Quit: Leaving.)
19:02:13 Quit AlexP_mob (Read error: Connection reset by peer)
19:02:50CIA-88r31598 build result: All green
19:04:25jlbiasiniI'm cleaning my manual patch, should be ready soon
19:04:30CIA-88r31599 build result: All green
19:06:50 Join remlap [0] (
19:07:04 Join AlexP_mob [0] (~AlexP@rockbox/staff/AlexP)
19:07:34 Quit AlexP_mob (Client Quit)
19:07:56 Join AlexP_mob [0] (~AlexP@rockbox/staff/AlexP)
19:08:15AlexP_mobI nearly give up :)
19:08:31CIA-88New commit by pamaury (r31600): usb: fix typo in DEBUGF
19:09:25CIA-88New commit by pamaury (r31601): usb: fix another typo
19:09:54kugelheh, /me has some make file magic running for selecting asm files over c files in an automatic manner
19:10:09CIA-88r31600 build result: All green
19:10:41kugeli.e. if arm/foo.S exists it'll be compiled, if not generic/foo.c will be picked up automatically
19:10:55 Quit AlexP_mob (Client Quit)
19:11:44CIA-88r31601 build result: All green
19:15:34CIA-88New commit by funman (r31602): lcd-remote-target.h: remove duplicate declarations
19:17:32CIA-88r31602 build result: All green
19:22:16CIA-88New commit by funman (r31603): lcd-remote-target.h: remove more duplicates
19:22:21bertrikjhMikeS, we currently still have a kind of autofade feature enabled in the analog codec, which changes the gain in 2ms steps towards the configured gain. In the past, disabling this fixed the "zip" sound for me, but turned it into a more pronounced click. Can't seem to reproduce it now though ... :)
19:23:30 Join TheLemonMan [0] (
19:24:04CIA-88r31603 build result: All green
19:24:05bertrikpossibly this fixed itself with the analog voltage changes
19:25:19bertriklebellium, can you test a patch on your clip zip with wisechip display?
19:26:06lebelliumbertrik; I go to the supermarket now but you send it to me, I test it in 40min
19:26:14lebelliumyou can send*
19:26:46bertrikpatch is here: it is supposed to improve the reddish tint of the display at low brightness
19:30:06lebelliumOk I let you know but I don't find my display to be particularly reddish at low brightness hum
19:31:44 Join TomColler [0] (
19:31:55lebelliumoh but I can only test compiled things. I don't know what to do with your patch :S
19:33:20jlbiasinibluebrother: regarding "OF" occurence in my IMX233_install.tex, it is starting with: An original firmware (OF) from \playerman's website" so the abrieviation is explained. If this is still not acceptable then we also have to correct the sansaAMS_install.tex file which i
19:33:41jlbiasini*is using the very same abreviation
19:34:12CIA-88New commit by funman (r31604): move lcd remote common declarations to lcd-remote.h
19:34:30funmanisn't there a "OF tag" ?
19:34:39 Join dv_ [0] (
19:34:45jlbiasiniI basicly made a nearly perfect copy of this sansaAMS
19:34:58jlbiasininot in the sansaAMS_install.tex
19:35:48funmanjlbiasini: nothing can be factored then?
19:36:18funmani'm trying to remove some of those 'copy everything from another target and change only one bit there and there' :/
19:36:20CIA-88r31604 build result: All green
19:36:30jlbiasiniand no: there is a bootloader one
19:36:40jlbiasinibut no OF tag
19:37:49jlbiasiniI offered to edit the ams file and so ask for a renaming of it (IMX233 is not ams and not reserved to sansa) but bluebrother told me to create a different file for a start
19:38:45funmanOTOH this install procedure is sansa specific, not imx233-specific
19:38:55funmanspecific to *some* sansa at least
19:38:57jlbiasinifunman: I think that for a start bluebrother is right: first we add a new target we will mix everything later
19:39:11funmanyeah if it's not complete perhaps it's better
19:39:28jlbiasiniwell for the moment imx233 is sansa fuze plus specific
19:39:45jlbiasinibut some creative device could follow (pamaury is on it
19:39:55 Quit bluebrother (Disconnected by services)
19:39:56 Join bluebrother^ [0] (~dom@rockbox/developer/bluebrother)
19:43:07 Quit fs-bluebot (Ping timeout: 240 seconds)
19:44:41 Join fs-bluebot [0] (
19:47:25 Part TomColler
19:49:37funmanah ok, that's why he talked about it
19:49:39pamauryafaik the creative uses the same update procedure
19:49:58pamauryin fact, all imx233/stmp3700 sdk based use this one
19:50:14pamauryplus they have a recovery mode than *could* be handled by rbutil for example
19:53:24 Join krabador [0] (
19:56:25 Quit stoffel (Remote host closed the connection)
20:06:12kugeljhMikeS: ping
20:07:38CIA-88New commit by funman (r31605): remote_detect(): move to lcd-remote.h ...
20:09:34CIA-88r31605 build result: 0 errors, 66 warnings (funman committed)
20:14:26funmando we leave that yellow to remind us of branching hwcodec ?
20:15:36gevaerts(a) no, and (b) I don't see the link between those yellows and hwcodec
20:16:59jlbiasinibluebrother: I updated my patch following your directions
20:18:07funmangevaerts: yellow wouldn't have happened if player build was not made from this branch
20:18:28CIA-88New commit by funman (r31606): Only need lcd-remote.h if HAVE_REMOTE_LCD is defined
20:20:22CIA-88r31606 build result: 0 errors, 60 warnings (funman committed)
20:22:19funmanhum i gues i'll need sh compiler
20:32:04 Join chkktri [0] (chikakitaa@unaffiliated/chkktri)
20:32:25chkktriare there any known issues with TTA playback?
20:32:53CIA-88New commit by funman (r31607): move _remote_backlight_* to common header ...
20:33:00bluebrother^leave yellow to get reminded of branching? Has anything about branching been decided *at all*?
20:33:12bluebrother^a couple of people are in favour of branching, ok.
20:33:30bluebrother^but branching doesn't make sense unless the transition to git is complete IMO
20:33:45jhMikeSkugel: pong
20:33:50bluebrother^and why on earth does Qt fail to load that lib on OS X?
20:34:50funmanbluebrother^: indeed
20:34:51kugeljhMikeS: are there C implementations of lcd_write_yuv420_lines_odither for lcd-memframe?
20:34:52CIA-88r31607 build result: 0 errors, 60 warnings (funman committed)
20:35:01jhMikeSkugel: nope
20:35:47CIA-88New commit by funman (r31608): apps: lcd-remote.h is not needed if HAVE_REMOTE_LCD is not defined ...
20:36:42jhMikeSthere wasn't really a motive for it at the time
20:36:42gevaertschkktri: I don't think so
20:37:43CIA-88r31608 build result: 0 errors, 21 warnings (funman committed)
20:38:45jhMikeSkugel: there also wasn't an lcd-memframe at the time !
20:39:04kugeljhMikeS: can the sdl code be adapted?
20:39:34kugelits yuv code does the algo and then calls lcd update, so I guess it's similar to lcd-memframe
20:40:26 Join evilnick [0] (5186c1c8@rockbox/staff/evilnick)
20:42:53bluebrother^urgh? Encoding using liblame seems to be _slower_ on OS X :o
20:43:21 Join chkktri_ [0] (
20:44:17 Quit chkktri (Read error: Connection reset by peer)
20:44:17 Quit factor (Read error: Connection reset by peer)
20:44:31CIA-88New commit by funman (r31609): mr500: move remote button reading code to buttom-mr500.c
20:45:11CIA-88New commit by funman (r31610): rolo: include lcd-remote.h only when needed
20:46:34CIA-88r31609 build result: 0 errors, 21 warnings (funman committed)
20:47:47jhMikeSkugel: erm, duh, no, I don't think that will work. the asm functions in lcd-as-memframe.S write TWO lines of YUV, so that any 4:2:0 chroma value is handled once for each block of 4 Y pixels
20:48:34CIA-88r31610 build result: 0 errors, 15 warnings (funman committed)
20:48:51CIA-88New commit by funman (r31611): lcd_remote_powersave: unused
20:49:37CIA-88New commit by funman (r31612): powermgmt.c : include lcd-remote.h only when needed
20:49:51funmanyellow should be fixed now
20:50:00bluebrother^wow, it's really that byte order code that causes the slowdown :(
20:50:28kugeljhMikeS: hm. I have some local stuff to select asm over C automatically, but that requires a C fallback for everything
20:50:38CIA-88r31611 build result: 0 errors, 15 warnings (funman committed)
20:52:26CIA-88r31612 build result: 0 errors, 6 warnings (funman committed)
20:52:37funmangah :/
20:53:04Mirat least we know you are commited
20:53:17Mirpa dum dumm
20:54:46 Part evilnick
20:54:48CIA-88New commit by funman (r31613): Only use lcd-remote.h #ifdef HAVE_REMOTE_LCD
20:55:44jhMikeSkugel: it should be safe since only ARM is using afaik, or will it just not work if it doesn't see C?
20:56:43CIA-88r31613 build result: All green
20:56:49 Quit TheLemonMan (Quit: WeeChat 0.3.6)
20:57:41jlbiasinipamaury there is still a problem with mpeg-plugin... It's starting but then hang on some video (like the 320x240 from wiki) or play sound displaying black screen with sometime partial glitching of the screen showing the upper right part of the video
20:59:01jlbiasiniI know this is nothing like a priority, just to let you know
20:59:59***Saving seen data "./dancer.seen"
21:01:30 Join factor [0] (~factor@
21:02:08kugeljhMikeS: it will work but defeats the purpose of my work
21:02:35kugelI also find it less than ideal that we have asm code which we have no C equivalent for
21:03:27pamauryjlbiasini: I have never used the mepg plugin so I don't know if it's supposed to work well or not, perhaps it's not related to the fuze+
21:04:18jlbiasiniyeah as is it is not usuable
21:04:28jlbiasinibut I don't know either
21:05:15jhMikeSkugel: to what extent should we concern ourselves with that? every single thing?
21:05:17jlbiasiniI tried with a lower resolution (128x128 file from wiki) and I get half of the screen
21:05:45CIA-88New commit by funman (r31614): lcd-remote-target.h: Use only in firmware/target/
21:06:09jhMikeSkugel: there is quite a bit of lcd code that has no C code, not just that either
21:06:19jlbiasinibut this is all very slow
21:06:28jlbiasininever mind we'll see later
21:06:46kugeljhMikeS: which bits other than arch specifics which cannot really be expressed in C?
21:07:39CIA-88r31614 build result: All green
21:08:16bluebrother^interesting. Creating a voice file on Windows: 0:10 / 0:40 / 1:00 (libmp3lame, libmp3lame w/ byte order code, running lame directly)
21:08:26bluebrother^so it's 6x faster with libmp3lame.
21:08:53bluebrother^on Linux it's almost no increase in speed: 0:05 / 0:22 / 0:07
21:09:25bluebrother^and OS X ... seems to be worse using libmp3lame, which is strange.
21:09:35jhMikeSkugel: I'm sure alot could be, but why bother? do we need C code for the x5 lcd driver? the whole memframe idea was the presence of optimized assembly already available to any device using a memory framebuffer
21:09:46*bluebrother^ goes trying that again
21:10:59jhMikeSso, if the memframe c-code weren't shared (as it wasn't previously), would this be a concern at all?
21:11:41 Quit mortalis (Quit: KVIrc 4.1.1 Equilibrium
21:11:49*bluebrother^ just got the encoding error on OS X that seems to be Windows only
21:14:35kugeljhMikeS: I wouldnt have noticed it at least :)
21:14:57kugelI want to use it on the ypr0 which doesnt have ASM enabled yet
21:15:03bluebrother^hmm. 0:07 / 0:35 / 0:07 on OS X.
21:15:16bluebrother^so running the executable isn't really slower
21:15:23jhMikeSkugel: well then, I guess it's there's now a reason
21:15:33bluebrother^OTOH that's all stopwatch-timed, so not extremely exact.
21:15:56kugeljhMikeS: well, yes and no, it'll eventually enable asm :)
21:16:08chkktri_looks like this happens only with big files
21:16:27chkktri_should i open an issue on flyspray?
21:16:28jhMikeSwhat arch is it? that code is well proven by now for ARM
21:16:41chkktri_(that's all about TTA playback problem)
21:17:08*bluebrother^ considers keeping the old code on OS X −− getting the correct dylib is (unfortunately) a bit of a pain on OS X as well
21:17:17kugeljhMikeS: it is arm
21:22:48Mirbluebrother^: why mp3
21:22:57Mirwhy no FLAC 8D
21:23:02bluebrother^Mir: ?
21:23:03Mir<3 flac
21:23:37bluebrother^I'm working on voicefile generation. For HWCODEC devices we need to use mp3. For SWCODEC speex.
21:23:45bluebrother^flac would be too huge
21:23:59bluebrother^since it's only about voice clips size is more important than quality
21:24:27bluebrother^the problem is that setting up lame is a bit ... well, ugly in Rockbox Utility right now.
21:24:50Mirand ogg isnt satisfactory?
21:24:58Miri hate ogg btw
21:25:00bluebrother^so I tried using libmp3lame instead of calling the lame binary. Which is much more elegant
21:25:07*bluebrother^ uses vorbis all the time
21:25:17Miri rather use flac than ogg
21:25:26bluebrother^how am I supposed to play a vorbis file on a device that has a hardware mp3 chip?
21:25:30Miri will use wma before ogg
21:25:41Mirgood point
21:25:45bluebrother^maybe, but I don't care about that :)
21:25:53Mirwhat a sharp point ;)
21:25:56 Part jlbiasini
21:26:04bluebrother^If I want lossy I use vorbis. If I want lossless ... well, flac maybe.
21:26:32bluebrother^either case, that doesn't make a difference for the actual problem I'm looking at.
21:27:25bluebrother^the annoying thing on OS X is that I have to handle endianess issues (since ppc is BE) and the available libmp3lame.dylib binaries are either placed in the wrong location or use the wrong architecture :(
21:27:59Mirwait are you on a G4/G5?
21:28:17bluebrother^but Rockbox Utility still supports OS X 10.4
21:28:21bluebrother^i.e. ppc
21:28:23Mirthen you are on a intel mac
21:28:33bluebrother^well, yes.
21:28:35Mirsome PPC
21:28:51Mirfor example 10.4 WONT install onmy G4 tower
21:28:51*bluebrother^ used Rosetta to debug some PPC code on an intel Mac some time ago
21:29:17bluebrother^well, Rockbox Utility is built for 10.4+. And 10.4 also supports ppc.
21:29:26bluebrother^so I need to handle ppc as well.
21:29:29Mirand to get it to install will require 800$ worth ofa single upgrade
21:29:36bluebrother^though maybe we should drop support for ppc ...
21:30:05Mirfirst find out how many PPC users there are before hapazardly dropping support
21:30:28bluebrother^that's the problematic part :)
21:30:35Mirwhat is
21:30:42bluebrother^since I assume that Apple's numbers cannot be trusted on that.
21:30:58Mirmaking a poll to find out how many PPC users there are
21:31:06bluebrother^but dropping ppc support for Rockbox Utility wouldn't be _that_ much of a deal-breaker.
21:31:13Miror tracking the download # for the PPC version?
21:31:18bluebrother^people can always use a different machine or a different OS
21:31:43Mirand te only reason i wold ask fr a PPC rockbox is lnnux ppc
21:31:44bluebrother^and the ppc machines aren't being sold since a couple of years, as well as not being supported by recent OS X versions
21:32:05Mir10.5.4 was the last
21:32:12bluebrother^though as long as it's not soo much of a problem I
21:32:17bluebrother^'d like to keep ppc support
21:32:26Mirrequred a 1.6GHZ G4/G5 or faste
21:32:29bluebrother^the annoying thing is this dylib issue
21:36:14bluebrother^hmm, I could also hardcode the path the Audacity dylib uses since that's not in the search path.
21:36:21bluebrother^but that's pretty ugly too
21:36:44*bluebrother^ decides to let this issue wait until tomorrow.
21:51:40 Join goa [0] (
21:52:37 Quit goa (Client Quit)
21:55:28gevaertschkktri_: if you have an actual bug, reporting it is always best, yes
21:56:13chkktri_but i shoul give actual file somehow, right?
22:00:46Mir*O* i so jizzed my pants
22:03:20 Nick Jack87|Away is now known as Jack87 (Jack87@nasadmin/admin/jack87)
22:05:42 Part Ave
22:10:00funmanbluebrother^: can't you use dual-arch static libs now?
22:19:42 Join Lonely_Dragon|2 [0] (
22:22:47 Quit Lonely_Dragon (Ping timeout: 268 seconds)
22:27:26kugeljhMikeS: lcd_update() works with a simple memcpy loop, as expected, but I don't think I'm smart enough for the yuv part
22:28:26funmankugel: there's a C imple;entation already no?
22:29:31kugelfunman: not for lcd-as-memframe.S
22:29:54funmanbut you can move out the yuv code and reuse it?
22:30:30funmanthere's a loop which updates 2 rows at a time
22:30:54funmandirectly in the framebuffer
22:31:40funmanhum i see lcd_blit_yuv() in lcd-memframe.c
22:31:43jhMikeSwhy not just enable the use of lcd-as-memframe.S?
22:31:57kugeljhMikeS: it seems the sdl code also does 2 lines at once?
22:32:50jhMikeSdoes it?
22:33:25saratogathe imx233 does hardware YUV conversion right?
22:34:59kugeljhMikeS: the one I'm looking at yes
22:36:33saratogaah yes, manual says YUV, rescaling and alpha bending in hardware, neat
22:38:28jhMikeSis it the same IPU as the beast?
22:38:42saratogai'm not sure what hte beast uses
22:38:46kugelthe imx37 in this ypr0 also has this
22:39:01saratogaprobably not though since its originally from sigmatel, unless they licensed it from freescale
22:39:10jhMikeSI would like an LCD driver that can use alternate framebuffers so I can use overlays
22:39:51jhMikeSit might be possible to render fonts with it too
22:40:20saratogait apparently can do hardware overlays
22:40:44saratogaThe PXP supports up to eight overlays that can be used to merge graphic data with video (or other graphic data). Each overlay consists of a rectangular area that is a multiple of eight pixels in both the vertical and horizontal directions
22:41:00kugelfunman, jhMikeS: appears to work in test_fps. need to check a real movie
22:41:01jhMikeSwhich sectin is it in?
22:42:33saratogajhMikeS: 17.2.5
22:42:48kugeljhMikeS: JdGordon was working on that IIRC
22:43:12pamaurysaratoga: iirc, it doesn't have a YUV conversion that is usable, for which purpose would you like it ?
22:43:38jhMikeSsaratoga: looks nothing like programming the i.MX31's IPU :)
22:43:45saratogapamaury: RGB > YUV in mpegplayer would be pretty useful
22:43:51saratogaerr otherway :)
22:44:00pamauryiirc, it doesn't have such a thing but let me check
22:44:20saratogathe manual says it does, although its possible that sandisk removed it
22:44:21pamauryand yes indeed, it does
22:44:27pamauryusing the PXP
22:44:39saratoga17.2.4 Colorspace Conversion
22:44:43jhMikeSwouldn't be such a bad thing to draw and decode in parallel
22:45:02saratogainterestingly, it can also do bilinear filtering and colorspace conversion on a frame
22:45:27funmanit would be cool if it had mpeg2video decoding
22:45:32saratogawhich would be extremely useful if we ever got support for things like .vob files or DVD isos in mpegplayer :)
22:45:32funmanso we could #define HWCODEC
22:45:35pamaurywould it require a hack to use it in mpegplayer ?
22:45:54funmansaratoga: vob should work i think
22:45:59jhMikeSif something can do all lcd stuff with hardware acceleration, that might be a good idea to work that out
22:46:17 Quit y4n (Quit: PANTS OFF!)
22:46:41saratogai guess the smart way would be to make the YUV blit operations use the hardware, although i'm not sure API-wise how easy that is
22:47:23funmandepends what the HW interface is
22:48:23jhMikeSperhaps the lcd api needs to mature to accomodate such things
22:48:53pamauryI don't remember how the pxp works, if it interacts with the lcdif or just blits to a buffer
22:49:18 Join liuzhoou [0] (~user@
22:49:26 Join Strife89 [0] (
22:50:10saratogafunman: from their example code, you give it a pointer to Y, U, and V tables, and an empty buffer
22:50:23saratogait then fills the buffer with the decoded data
22:51:00saratogayou can also specify the rescaling ratio while it does this
22:51:24funmanlooks simple enough to not need new API
22:52:14saratogayeah actually their example code is remarkably simple
22:52:25saratogaAMS could learn from these guys
22:52:28funmanlcd_blit_yuv converts and write to lcd in one shot, HW would convert the video and then call lcd_update(_rect_ though
22:52:53jhMikeSrendering the OSD as an overlay would be nice, rather than cropping the YUV
22:53:00*funman looks at other -target.h files
22:53:46pamauryI'll see what I can do to speed up lcd_yuv_blit if mpegpolayer uses that
22:53:47funmanthe publicspi.h only has a "#include "spi-target.h"
22:54:27pamauryI also saw interesting undocumented registers in the DCP that look like colorspace conversion, might be interesting
22:54:39saratogapamaury: yuv_blit tends to be quite slow, so if the hardware can do it, then we should use it
22:57:02kugeljhMikeS: lcd-memframe deals with orientation for yuv already right?
22:57:34kugelI still seem to need the corresponding #ifdef using the sdl code
22:57:54jhMikeSkugel: yeah
22:58:35 Quit liuzhoou (Ping timeout: 252 seconds)
23:00:03***Saving seen data "./dancer.seen"
23:00:49kugelah the sdl code is the same as what's in lcd-16bit-common.c
23:04:17CIA-88New commit by funman (r31615): Clean spi.h and spi-target.h ...
23:04:38funmankugel: hm should the sdl code be removed then?
23:05:44kugelfunman: android as well
23:05:57 Quit saratoga (Ping timeout: 258 seconds)
23:06:10kugelI assume the code in lcd-16bit-common.c was pulled from sdl?
23:06:20funmanmost likely
23:06:20CIA-88r31615 build result: All green
23:11:01 Quit remlap (Read error: Connection reset by peer)
23:11:25CIA-88New commit by funman (r31616): ingenic_jz: remove ata-nand-target.h
23:13:16CIA-88r31616 build result: All green
23:15:12 Quit [Saint] (Read error: Connection reset by peer)
23:20:40 Join remlap [0] (
23:27:51 Join tr|nn| [0] (
23:29:15 Join saratoga [0] (9803c31c@gateway/web/freenode/ip.
23:32:54CIA-88New commit by funman (r31617): powermgmt-target.h: move prototypes to powermgmt.h ...
23:34:52CIA-88r31617 build result: 0 errors, 1 warnings (funman committed)
23:34:56funmanwe have power-target.h , powermgmt-target.h, and pmu-target.h
23:39:22CIA-88New commit by funman (r31618): cowond2's power-target.h -> rename in pmu-target.h ...
23:40:43funmanif CIA-88 could tell us in which targets/features the red/yellow happen that'd be awesome
23:41:00funman174 warnings (in LCD_CHARCELL, CONFIG_TUNER)
23:41:21CIA-88r31618 build result: 0 errors, 1 warnings (funman committed)
23:43:54CIA-88New commit by funman (r31619): charging_algorithm_close() declaration has moved
23:45:45CIA-88r31619 build result: All green
23:47:38kugelhow do you put a lookup table into cf assembly?
23:48:00funmanwrite it in C, look at gcc -S
23:49:07kugelgood idea
23:49:14funmankugel: what's the difference between android and ypr0 dir-target.h ?
23:49:41kugelalmost none
23:49:45n1si think you just write .long .byte or .word and then your data, not sure of the exact syntax
23:49:46kugelI'll merge them later
23:50:53n1skugel: any reason to not just put the table in a c file and just link into the assembly?
23:51:03 Quit bertrik (Ping timeout: 255 seconds)
23:54:38n1sffs does that as an example
23:54:44kugelhehe :)
23:54:50kugelI'm undoing exactly that
23:55:03jhMikeSdon't undo it
23:55:56jhMikeSthen an optimization isn't possible with table location exploit
23:56:07kugeljhMikeS: what?
23:56:23jhMikeSyou mean ffs-arm.S right?
23:56:32 Quit benedikt93 (Quit: Bye ;))
23:56:47jhMikeSsee the ldrb instruction
23:57:06jhMikeSyes, that has a separate table
23:57:09kugelI assume the same trick can possibly be used on cf?
23:57:16*jhMikeS freaked out too much then
23:57:24jhMikeSno, it can't use pc that way
23:58:15kugeln1s: my new framework doesn't allow that nicely
23:58:22jhMikeSsorry, I thought you said you were moving to a C table :P
23:58:46kugelwould need to put the .c for the table into a different SOURCES

Previous day | Next day