#rockbox 2010-11-12

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 2010-11-12

00:05:23		Quit Strife89 (Quit: Heading home - night all! o/)
00:14:39		Quit bertrik (Quit: :tiuQ)
00:15:05		Quit ender` (Quit: I have the heart of a small boy. I keep it in a jar on my desk.)
00:15:38		Quit Super_Dork (Quit: Leaving)
00:20:30		Join JdGord [0] (~jd@pa58-109-129-213.pa.nsw.optusnet.com.au)
00:25:57		Join user890104 [0] (Venci@Venci-Notebook-LAN.ipv6.6bez10.info)
00:29:25		Quit hebz0rl (Quit: Leaving)
00:30:23		Join anewuser [0] (anewuser@unaffiliated/anewuser)
00:31:30		Join Llorean1 [0] (~DarkkOne@adsl-69-153-207-220.dsl.hstntx.swbell.net)
00:34:25		Join maraz [0] (maraz@kapsi.fi)
00:35:18	CIA-7	New commit by moos (r28560): Update the french translation regarding the previous revision.
00:36:16		Join The3_14ed\|r [0] (~ntryon@cpe-67-240-248-41.rochester.res.rr.com)
00:37:19	CIA-7	r28560 build result: All green
00:40:30		Join JdGordon\| [0] (~jonno@vl10.gw.ok-labs.com)
00:40:30		Quit JdGordon\| (Changing host)
00:40:30		Join JdGordon\| [0] (~jonno@rockbox/developer/JdGordon)
00:41:08		Quit Llorean (.net .split)
00:41:08		Quit iq (.net .split)
00:41:08		Quit maraz_ (.net .split)
00:41:08		Quit Dhraakellian (.net .split)
00:41:08		Quit simonrvn (.net .split)
00:41:08		Quit literal (.net .split)
00:41:21	***	Saving seen data "./dancer.seen"
00:47:45		Join iq [0] (~iq@unaffiliated/iq)
00:48:28		Quit jgarvey (Quit: Leaving)
00:50:46		Nick The3_14ed\|r is now known as Dhraakellian (~ntryon@cpe-67-240-248-41.rochester.res.rr.com)
00:54:56		Join CaptainKewl [0] (~jason@207-38-215-126.c3-0.nyr-ubr1.nyr.ny.cable.rcn.com)
00:57:41		Join simonrvn [0] (simon@209.165-ppp.3menatwork.com)
00:57:52		Join literal [0] (hinrik@v.nix.is)
00:58:22		Nick simonrvn is now known as Guest33011 (simon@209.165-ppp.3menatwork.com)
01:00
01:01:23		Quit moos (Quit: ChatZilla 0.9.86 [Firefox 3.6.12/20101026210630])
01:03:27		Quit GeekShadow (Quit: The cake is a lie !)
01:09:51		Quit DerPapst (Quit: Leaving.)
01:24:34		Join Horschti [0] (~Horscht@xbmc/user/horscht)
01:29:21		Quit Horscht (Ping timeout: 276 seconds)
01:29:54		Join fdinel [0] (~Miranda@modemcable235.127-131-66.mc.videotron.ca)
01:32:30		Join clone4crw [0] (~Calvin@97-86-227-168.dhcp.roch.mn.charter.com)
01:32:54		Part toffe82
01:47:09		Join T44 [0] (~Topy44@f049099199.adsl.alicedsl.de)
01:49:39		Join Bacon [0] (~43a47747@giant.haxx.se)
01:49:44	Bacon	hello?
01:50:19	Bacon	can someone help me with my ipod?
01:50:36		Quit Topy44 (Ping timeout: 245 seconds)
01:50:59	JdGordon\|	not untill you ask a question
01:51:09	Bacon	huh?
01:51:12	Bacon	oh, haha
01:51:29	Bacon	yeah, i have a 5th gen video with rockbox on it
01:51:44	Bacon	i want to know how to reboot it
01:52:34	xxcv	press play
01:53:16	Bacon	play?
01:53:54	xxcv	don't release
01:53:55	Bacon	theres an icon on the main menu that says resume playback?
01:54:16	Bacon	...
01:54:23	Bacon	im a little confused..
01:54:48	xxcv	oh my you need some sleep..
01:55:17	xxcv	press ok + menu hold for 10 seconds
01:56:23	Bacon	ok
01:56:35	Bacon	cool
01:56:39	JdGordon\|	xxcv: press means press... press+release is "hold", use the right terminology so people dont get confused
01:56:50	JdGordon\|	fuck, press+hold is "hold"
01:57:10	Bacon	jd, did you write rockbox
01:57:23	JdGordon\|	not all of it, but a fair bit
01:57:33	Bacon	did you use C++
01:58:16	*	xxcv remains calm.
01:58:35	JdGordon\|	hell no!
01:58:45	Bacon	lol...?
01:58:58	Bacon	whats wrong with c++
01:59:25	JdGordon\|	everything, and this isnt the channel to discuss... #rockbox-community is for non rockbox talk
02:00
02:00:49	Bacon	ah, ok
02:00:55	Bacon	i have another question
02:01:22	Bacon	how to you change the skins
02:01:26	Bacon	or shell
02:01:40	JdGordon\|	http://themes.rockbox.org/
02:03:15	Bacon	im talking about on the ipod
02:04:30	pixelma	we have a nice manual, I think it answers questions like that and was written for that purpose (not having to answer the same things time and again) ;)
02:05:47	Bacon	ook, thanx
02:05:49		Quit Bacon (Quit: CGI:IRC)
02:24:42		Join dfkt [0] (dfkt@unaffiliated/dfkt)
02:29:19		Quit Farthen (Ping timeout: 260 seconds)
02:41:23	***	Saving seen data "./dancer.seen"
03:00
03:04:27		Join Farthen [0] (~Farthen@static.225.178.40.188.clients.your-server.de)
03:05:29		Quit clone4crw (Read error: Connection reset by peer)
03:05:50		Join clone4crw [0] (~Calvin@97-86-227-168.dhcp.roch.mn.charter.com)
03:06:22		Quit Llorean1 (Read error: Connection reset by peer)
03:06:53		Join Llorean [0] (~DarkkOne@adsl-69-153-207-195.dsl.hstntx.swbell.net)
03:14:08		Quit evilnick (Ping timeout: 240 seconds)
03:14:21		Join antil33t [0] (antil33t@124-197-51-80.callplus.net.nz)
03:20:40		Join BHSPitMonkey [0] (~stephen@unaffiliated/bhspitmonkey)
03:22:35		Quit MethoS- (Remote host closed the connection)
03:23:53		Quit JdGord (Quit: Bye)
03:35:47		Nick Guest33011 is now known as simonrvn (simon@209.165-ppp.3menatwork.com)
03:59:01		Quit bluebroth3r (Read error: Connection reset by peer)
03:59:14		Join bluebrother [0] (~dom@f053154068.adsl.alicedsl.de)
03:59:18		Quit bluebrother (Changing host)
03:59:18		Join bluebrother [0] (~dom@rockbox/developer/bluebrother)
04:00
04:08:59		Quit InsDel (Read error: Connection reset by peer)
04:11:11		Quit amiconn (Disconnected by services)
04:11:12		Join amiconn_ [0] (quassel@rockbox/developer/amiconn)
04:11:32		Nick amiconn_ is now known as amiconn (quassel@rockbox/developer/amiconn)
04:12:12		Quit pixelma (Disconnected by services)
04:12:14		Join pixelma_ [0] (quassel@rockbox/staff/pixelma)
04:12:16		Nick pixelma_ is now known as pixelma (quassel@rockbox/staff/pixelma)
04:14:17		Quit anewuser ()
04:15:57		Quit Sundiver (Quit: Leaving)
04:16:47		Quit froggyman (Quit: Bye)
04:17:08		Join froggyman [0] (~seth@unaffiliated/froggyman)
04:18:55		Quit Llorean (Changing host)
04:18:55		Join Llorean [0] (~DarkkOne@rockbox/user/Llorean)
04:22:24		Join anewuser [0] (anewuser@unaffiliated/anewuser)
04:25:07		Join Barahir_ [0] (~jonathan@frnk-590f51a4.pool.mediaWays.net)
04:28:50		Quit Barahir (Ping timeout: 260 seconds)
04:34:15	Llorean	soap: Has mpegplayer even changed between 3.6 and 3.7?
04:34:28	soap	Llorean, I'm being polite
04:35:20	soap	Accepting the possibility of fault while kindly asking the user to investigate their end. Flies / honey / vinegar / all that.
04:35:46	Llorean	Also, I saw in another thread you mentioned IAP not working in cars probably being them checking the filesystem. Someone mentioned to me that our IAP supports is via serial, don't most cars just use a USB cable to hook up (and thus, we wouldn't support IAP on them at all yet?)
04:36:03	soap	No
04:36:23	Llorean	To which part?
04:36:27	soap	Some cars see the Apple USB ID and try to read the apple database
04:36:36	Llorean	Interesting
04:36:50	Llorean	My car gives me an error, but I have to plug it in with an iPod USB cable
04:37:10	Llorean	But it's an "iPod not recognized" or something style error
04:37:27	soap	prior to 5th gen ipod?
04:37:59	Llorean	iPod Nano G1
04:38:24	soap	Other cars apparently have UMS capability, but if see Apple ID refuse to even try.
04:38:33	soap	(but that's not a IAP issue).
04:38:36	Llorean	I don't believe mine has UMS capability
04:38:44	Llorean	Though I guess I haven't checked.
04:40:28	Llorean	So the first test would be to sync a few things using iTunes, then boot in Rockbox and see if it then recognizes it?
04:40:55		Quit T44 (Read error: No route to host)
04:40:58	soap	one sec
04:41:03		Join T44 [0] (~Topy44@f049099199.adsl.alicedsl.de)
04:41:26	***	Saving seen data "./dancer.seen"
04:42:38	soap	There is also the iHome IH36 - I'm trying to get lixer to reply back, in email he strongly _implied_ that he only got the device working with itunes loaded music, but he's been hard to pin down.
04:44:00		Quit dfkt (Quit: -= SysReset 2.53=- Sic gorgiamus allos subjectatos nunc.)
04:44:29	froggyman	so the USB stack on fuze v2 works fine on linux and only has problems underwindows?
04:48:20		Join FIReun [0] (~fireun@unaffiliated/fireun)
04:49:47		Quit panni_ (Read error: Connection reset by peer)
04:56:08		Quit dys (Ping timeout: 276 seconds)
04:56:26		Join dys [0] (~andreas@krlh-5f72fea3.pool.mediaWays.net)
04:58:25		Quit TheSeven (Ping timeout: 250 seconds)
04:59:35		Quit FIReun (Quit: Leaving.)
05:00
05:01:35	[Saint]	xxcv: fwiw, when you're instructed to reboot an iPod with rockbox on it you should always press and hold play to shut down the player, then any key to start it again. You should never use Menu+Select unless Rockbox is in a freeze and won't shut down manually.
05:02:19		Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
05:04:52	soap	ie reboot, not _reset_
05:14:01		Quit anewuser ()
05:30:08		Quit JdGordon\| (Quit: leaving)
05:32:43		Quit Judas_PhD (Quit: This is a quitting message)
05:32:51		Quit Horschti (Quit: Verlassend)
05:34:59		Quit ps-auxw (Ping timeout: 276 seconds)
05:35:47		Quit xxcv ()
05:37:53		Quit clone4crw (Ping timeout: 255 seconds)
05:45:42		Join ps-auxw [0] (~arneb@p4FF7EF75.dip.t-dialin.net)
05:46:20		Join binaryhermit [0] (~binaryher@adsl-75-49-220-78.dsl.emhril.sbcglobal.net)
05:51:20		Join Super_Dork [0] (~Super_Dor@ip68-103-180-128.ks.ok.cox.net)
05:51:41		Join xxcv [0] (~IMNOTADEV@c211-30-174-99.carlnfd1.nsw.optusnet.com.au)
05:57:54	Super_Dork	I subscribed to the dev mailing list, how do I ask the devs a question?
06:00
06:02:17		Join JdGordon\| [0] (~jonno@rockbox/developer/JdGordon)
06:05:32		Join saratoga [0] (9803c6dd@gateway/web/freenode/ip.152.3.198.221)
06:06:31	Super_Dork	I hate to ask it again, but nobody has answered it yet. I subscribed to the dev mailing list, how do I ask the devs a question?
06:07:31	saratoga	what do you mean?
06:12:09		Quit binaryhermit (Quit: Leaving)
06:16:58	Super_Dork	I need to ask one of the USA devs if I could send them my Fuze V2 that won't get past the splash screen, to see if that would help solve the issue. There are a bunch of us that have that problem, but non that do are devs. Thus I am offering to send a dev my player to have the dev look at it and see if there is a solution. I hope that makes sense.
06:17:56	Super_Dork	I just need to know how to ask a question to the dev mailing list.
06:18:06	saratoga	oh
06:18:17	saratoga	you just send an email to the list and it goes to everyone
06:18:43	saratoga	see: http://cool.haxx.se/mailman/listinfo/rockbox-dev
06:23:03		Quit JesusFreak316 (Ping timeout: 245 seconds)
06:31:00		Quit fdinel (Quit: Miranda IM! Smaller, Faster, Easier. http://miranda-im.org)
06:32:04		Join Judas_PhD [0] (~kevin@misterfluffy.dsl.xmission.com)
06:32:34		Quit JdGordon\| (Ping timeout: 265 seconds)
06:36:33		Quit factor (Read error: Connection reset by peer)
06:39:23		Quit xxcv ()
06:41:29	***	Saving seen data "./dancer.seen"
06:45:19	Super_Dork	heh. "when all else fails, read the instructions" I guess. Sorry for the noob question. Thanks for your help saratoga.
06:45:30	saratoga	no problem
06:46:52	Super_Dork	I just want to make double sure I do everything right, as this is my first time installing Rockbox on a music player. I had Rockbox on an e260 before, but it was my brother that put it onto it for me.
06:47:08	Super_Dork	And now this isssue.
06:53:32		Join factor [0] (~factor@r74-195-220-23.msk1cmtc02.mskgok.ok.dh.suddenlink.net)
06:54:04		Quit Super_Dork (Quit: Leaving)
07:00
07:08:33		Join xxcv [0] (~IMNOTADEV@c211-30-174-99.carlnfd1.nsw.optusnet.com.au)
07:17:23		Join stoffel [0] (~quassel@p57B4DF1A.dip.t-dialin.net)
07:38:09		Join Buschel [0] (~chatzilla@p54A3B019.dip.t-dialin.net)
07:44:15	[Saint]	Anyone have any suggestions on a creative way I can do the ">" chars that the Apple OF has in its menu structure on the iLike theme?
07:46:53		Quit xxcv (Ping timeout: 255 seconds)
07:49:04	[Saint]	I thought about trying to do it by just drawing the chars on the menu backdrop but that would look stupid and obvious because there's not a gauranteed ammount of items in each menu so there could easily be more ">"s than items in the menu.
08:00
08:02:43		Join einhirn [0] (~Miranda@bsod.rz.tu-clausthal.de)
08:03:27	Llorean	Manually add them to an unused character in the font you choose to use?
08:04:09	CIA-7	New commit by Buschel (r28561): Set alignment for major data arrays in the mpc codec. This avoids performance regressions on S5L870x CPUs when changing code and allows further ...
08:06:03	CIA-7	r28561 build result: All green
08:07:59	CIA-7	New commit by Buschel (r28562): Re-submit ARM asm optimizations in mpc synthesis as the performance regressions on S5L870x have been solved with r28561.
08:09:11	CIA-7	New commit by Buschel (r28563): Minor speed up of mpc requantization.
08:09:52	CIA-7	r28562 build result: All green
08:11:37	CIA-7	r28563 build result: 3 errors, 0 warnings (Buschel committed)
08:12:36	*	Buschel assumes this failure has not been caused by r28563
08:17:15		Quit BHSPitMonkey (Remote host closed the connection)
08:23:38		Join kugel [0] (~kugel@rockbox/developer/kugel)
08:24:57	[Saint]	Llorean: I'm not sure you're understanding what it is I'm trying to do. I have that char in my font sets, but I'm stuck on how to make this happen: http://imgur.com/GlZQk.png
08:25:01	Stummi	good morning #rockbox
08:25:34	[Saint]	As you can see, it's more complicated than "every menu entry gets a ">""...it's only menu entries with submenus.
08:25:54	[Saint]	And I have NO idea how to pull it off. :(
08:26:44		Quit Buschel (Ping timeout: 240 seconds)
08:37:23		Join Zagor [0] (~bjst@rockbox/developer/Zagor)
08:39:59		Quit stoffel (Ping timeout: 255 seconds)
08:40:55		Quit kugel (Remote host closed the connection)
08:41:03	*	TheSeven wonders if n02-roolku's memory is dying :)
08:41:32	***	Saving seen data "./dancer.seen"
08:44:30		Join DerPapst [0] (~Alexander@p5DE5B6E4.dip.t-dialin.net)
08:48:38		Join ender` [0] (krneki@foo.eternallybored.org)
08:54:28	Stummi	I have a general c-question: http://pastebin.de/11620 is there a difference between the declaration of test1 (without arraysize) and test2 (with arraysize)?
08:56:41		Join n1s [0] (~n1s@90-230-78-242-no134.tbcn.telia.com)
08:56:42		Quit n1s (Changing host)
08:56:42		Join n1s [0] (~n1s@rockbox/developer/n1s)
08:59:09		Join TheSeven\|Mobile [0] (~theseven@rockbox/developer/TheSeven)
09:00
09:03:48		Quit TheSeven (Ping timeout: 276 seconds)
09:07:36		Quit avacore (Ping timeout: 245 seconds)
09:10:07	kadoban	Stummi: no, no difference
09:13:16		Join petur [0] (~petur@rockbox/developer/petur)
09:13:57		Quit simonrvn (Ping timeout: 240 seconds)
09:15:17		Join xxcv [0] (~IMNOTADEV@c211-30-174-99.carlnfd1.nsw.optusnet.com.au)
09:21:20		Join JdGord [0] (~jd@114.75.116.22)
09:21:40		Quit petur (Quit: plop)
09:22:16		Join simonrvn [0] (simon@209.165-ppp.3menatwork.com)
09:30:14		Join Buschel [0] (~chatzilla@p54B66DB5.dip.t-dialin.net)
09:31:06		Quit JdGord (Ping timeout: 276 seconds)
09:32:48		Join JdGord [0] (~jd@114.75.116.22)
09:35:14	AlexP	bertrik: r28559 contains skin tokens right (tr, tl, th)? Do they just return a number? I ask as they need adding to the manual
09:36:49		Join Rob2222 [0] (~Miranda@p4FFF2435.dip.t-dialin.net)
09:37:05		Quit DerPapst (Quit: Leaving.)
09:40:51		Quit Rob2223 (Ping timeout: 276 seconds)
09:40:53		Join swilde [0] (~wilde@aktaia.intevation.org)
09:42:33		Quit Judas_PhD (Quit: This is a quitting message)
09:43:56		Quit JdGord (Quit: Bye)
09:49:49		Quit Buschel (Ping timeout: 272 seconds)
10:00
10:00:10		Join LinusN [0] (~linus@rockbox/developer/LinusN)
10:00:51		Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
10:02:03		Quit utanapischti (Quit: WeeChat 0.3.2)
10:02:29		Join sasquatch [0] (~username@p4FF2D44C.dip.t-dialin.net)
10:02:30		Quit TheSeven\|Mobile (Quit: TheSeven\|Mobile)
10:03:35		Join kugel [0] (~kugel@rockbox/developer/kugel)
10:11:27		Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
10:12:32		Quit TheSeven (Remote host closed the connection)
10:22:45		Quit user890104 (Ping timeout: 272 seconds)
10:33:48		Join user890104 [0] (Venci@Venci-Notebook-WLAN.ipv6.6bez10.info)
10:41:36	***	Saving seen data "./dancer.seen"
10:42:11		Join Judas_PhD [0] (~kevin@misterfluffy.dsl.xmission.com)
10:45:33		Join DerPapst [0] (~Alexander@p5DE5B6E4.dip.t-dialin.net)
10:47:13		Join earcar [0] (~carmine@93-39-229-194.ip78.fastwebnet.it)
10:51:39		Quit alexbobP (Ping timeout: 264 seconds)
10:53:00		Quit earcar (Quit: bye)
10:53:23		Join alexbobP [0] (~alex@adsl-75-63-1-71.dsl.austtx.sbcglobal.net)
10:56:15		Join mystica555_ [0] (~mike@c-75-70-179-25.hsd1.co.comcast.net)
10:59:32		Quit togetic (Ping timeout: 240 seconds)
11:00
11:02:25		Quit DerPapst (Quit: Leaving.)
11:02:46		Quit mystica555_ (Ping timeout: 265 seconds)
11:02:58	Stummi	snprintf(buf[i], 10, "%i Max", i); .. on sim-build this works (I get "7 Max" for example), on normal build i get "i Max". Somebody an idea, why?
11:03:10	Stummi	*has somebody..
11:04:05	Stummi	context: http://pastebin.de/11622
11:05:30	ranmachan	Try %d instead
11:05:45	ranmachan	The sim-build probably uses the system libc for snprintf I'd guess
11:06:54		Join InsDel [0] (~haqr.net@c-98-231-87-43.hsd1.fl.comcast.net)
11:07:18	Stummi	ranmachan, and for the normal-build there is an other implementation of snprintf?
11:08:19	Stummi	ok, with %d it works. thanks
11:10:03	Zagor	Stummi: yes, rockbox has its' own libc
11:11:56	Stummi	ok
11:12:14		Join bmbl [0] (~Miranda@vpn224-55.fh-bielefeld.de)
11:12:14		Quit bmbl (Changing host)
11:12:14		Join bmbl [0] (~Miranda@unaffiliated/bmbl)
11:16:09		Join mystica555_ [0] (~mike@m4c2636d0.tmodns.net)
11:16:10		Quit mystica555_ (Excess Flood)
11:16:40		Join mystica555_ [0] (~mike@m3c2636d0.tmodns.net)
11:18:51		Quit kugel (Ping timeout: 276 seconds)
11:23:12		Join earcar [0] (~carmine@93-39-229-194.ip78.fastwebnet.it)
11:23:30		Join evilnick [0] (~evilnick@bb-87-82-1-42.ukonline.co.uk)
11:29:37		Quit mystica555_ (Read error: Connection reset by peer)
11:30:42		Quit [Saint] (Ping timeout: 276 seconds)
11:32:05		Join S_a_i_n_t [0] (S_a_i_n_t@203.184.0.187)
11:38:00		Join mystica555_ [0] (~mike@c-75-70-179-25.hsd1.co.comcast.net)
11:43:44		Join Jerom [0] (~jerome@95.171.142.251)
11:47:50		Quit bmbl (Quit: Bye!)
11:56:26		Quit Jerom (Ping timeout: 240 seconds)
12:00
12:04:25		Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
12:06:31		Join stoffel [0] (~quassel@p57B4DF1A.dip.t-dialin.net)
12:19:06		Join Jerom [0] (~jerome@79.132.60.222)
12:30:15		Quit xxcv (Ping timeout: 265 seconds)
12:39:32		Join teru [0] (~teru@KD059133111160.ppp.dion.ne.jp)
12:41:37	***	Saving seen data "./dancer.seen"
12:46:42		Quit Jerom (Quit: Leaving.)
12:47:32		Join MethoS- [0] (~clemens@134.102.106.250)
12:52:58		Quit InsDel (Read error: Connection reset by peer)
13:00
13:02:40	TheSeven	any opinions on how the keyclick settings menu should look like?
13:02:54		Quit earcar (Quit: bye)
13:03:20	TheSeven	i proposed two possibilities at FS #5111, please comment :)
13:13:36	n1s	as long as the two clicks can be set independently i
13:13:44	n1s	m fine with either
13:14:58		Join anewuser [0] (anewuser@unaffiliated/anewuser)
13:32:32		Quit anewuser ()
13:34:54	Stummi	a little "survey": atm I am thinking about saving a "snapshot"-savegame automatically on exiting rockboy which is also automatically loaded, when the rom is loaded the next time. Would you say, something like this would be reasonable?
13:42:03		Join bzed_ [0] (~bzed@devel.recluse.de)
13:42:27		Nick bzed_ is now known as Guest39244 (~bzed@devel.recluse.de)
13:42:46		Quit Farthen (Ping timeout: 240 seconds)
13:43:01		Quit simabeis_ (Ping timeout: 255 seconds)
13:43:08		Join simabeis [0] (~simabeis@lobmenschen.de)
13:43:29		Quit bzed (Ping timeout: 245 seconds)
13:43:34		Nick Guest39244 is now known as bzed (~bzed@devel.recluse.de)
13:44:02		Join Farthen [0] (~Farthen@static.225.178.40.188.clients.your-server.de)
13:45:23		Quit stoffel (Ping timeout: 250 seconds)
13:58:36	Stummi	teru, you seem to work often with rockboy. What would you say to this?
13:59:26	AlexP	Stummi: I'd like an option for the auto bit, but sounds nice"
14:00
14:00:25	Stummi	hm, yes. An option for that will be there of course
14:01:55		Join wodz [0] (~wodz@chello087206240131.chello.pl)
14:03:11	wodz	AlexP: r28534 is possible candidate for backport
14:04:24	AlexP	wodz: That is just for replacement batteries?
14:05:19	wodz	yes, but minis are quite old so more and more people will replace bateries
14:06:16	AlexP	yeah
14:06:20	AlexP	I'm not sure
14:06:30	AlexP	It isn't really a bug bug
14:21:01		Join Jerom [0] (~jerome@79.132.60.222)
14:24:01		Join fdinel [0] (~Miranda@modemcable235.127-131-66.mc.videotron.ca)
14:25:03	wodz	AlexP: It's so simple change that I think it won't hurt
14:41:41	***	Saving seen data "./dancer.seen"
14:52:13		Join LambdaCalculus37 [0] (~3f74f70d@rockbox/staff/LambdaCalculus37)
14:52:45		Quit LambdaCalculus37 (Client Quit)
14:54:06		Join LambdaCalculus37 [0] (~3f74f70d@rockbox/staff/LambdaCalculus37)
14:59:09		Quit n1s (Quit: L�mnar)
15:00
15:02:31	LambdaCalculus37	Hmmm, the touchpad on the HDD6330 is a little flaky in Rockbox.
15:02:36		Join n1s [0] (~n1s@90-230-78-242-no134.tbcn.telia.com)
15:02:36		Quit n1s (Changing host)
15:02:36		Join n1s [0] (~n1s@rockbox/developer/n1s)
15:04:40	LambdaCalculus37	b0hoon, lowlight: (for the logs) There's a little flakiness with the touchpad on the HDD6330. Pressing View in the lower right corner to bring up the context menu causes the rest of the touchpad to behave erratically.
15:08:10		Join komputes [0] (~komputes@ubuntu/member/komputes)
15:08:31		Quit teru (Quit: Quit)
15:15:36		Quit LambdaCalculus37 (Quit: CGI:IRC (EOF))
15:18:39		Quit S_a_i_n_t (Disconnected by services)
15:18:42		Join S_a_i_n_t [0] (S_a_i_n_t@203.184.0.144)
15:22:10		Join LambdaCalculus37 [0] (~3f74f70d@rockbox/staff/LambdaCalculus37)
15:25:29		Join benedikt93 [0] (~benedikt9@unaffiliated/benedikt93)
15:28:27		Quit TheSeven (Ping timeout: 245 seconds)
15:30:33		Quit krazykit (Quit: brb)
15:30:41		Join krazykit [0] (krazykit@silenceisdefeat.com)
15:33:13		Quit LambdaCalculus37 (Quit: CGI:IRC (EOF))
15:33:42		Join LambdaCalculus37 [0] (~3f74f70d@rockbox/staff/LambdaCalculus37)
15:38:10		Part Zagor
15:39:44		Quit CaptainKewl (Ping timeout: 265 seconds)
15:41:20		Quit GeekShadow (Ping timeout: 272 seconds)
15:47:54		Join GeekShadow [0] (~Antoine@44.181.204-77.rev.gaoland.net)
15:47:54		Quit GeekShadow (Changing host)
15:47:54		Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
15:54:37		Part LinusN
15:56:16		Join panni_ [0] (hannes@ip-178-203-81-220.unitymediagroup.de)
16:00
16:07:42	wodz	is it possible that CF cards do not support correctly 'standby' ata command 0xE2 (0x96) but support 'sleep' ata command 0xE6 (0x99) ?
16:08:32	LambdaCalculus37	wodz: There are a few cards that don't support those commands, I believe.
16:09:12		Quit user890104 (Ping timeout: 272 seconds)
16:10:10	LambdaCalculus37	wodz: I had an A-Data 4GB "Speedy" that exhibited that behavior in my iPod video.
16:10:24		Quit GeekShadow (Read error: No route to host)
16:10:26	wodz	rockbox uses 'sleep' command and works with my CFed MPIO, OF uses 'standby' and freezes after 'spinning down'
16:10:49	LambdaCalculus37	wodz: What brand of CF card?
16:11:05	wodz	Sandisk Ultra II
16:11:52	LambdaCalculus37	wodz: Wanna post your findings here? http://www.rockbox.org/wiki/CFModGuide
16:12:07	wodz	LambdaCalculus37: first I have to confirm this. I
16:12:15	wodz	I'll patch OF and see
16:12:37	LambdaCalculus37	wodz: Okay.
16:14:49	wodz	heh reading technical documentation to some CF card I see that they do not conform to ATA standard in this area. It looks like they swap standby and sleep. It is said that any command issued after sleep will recover normal operation. This exactly contrary to what ATA spec sais
16:17:02		Join dfkt [0] (dfkt@unaffiliated/dfkt)
16:17:25		Join GeekShadow [0] (~Antoine@44.181.204-77.rev.gaoland.net)
16:17:25		Quit GeekShadow (Changing host)
16:17:25		Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
16:18:05		Join captainkewll [0] (2669ecc2@gateway/web/freenode/ip.38.105.236.194)
16:18:12	n1s	wodz: yeah, i think quite a few pwople have had trouble with non conforming cf cards
16:20:03	LambdaCalculus37	n1s: IIRC the manufacturers that conform to ATA standards are A-Data, Transcend, and Kingston.
16:20:32		Quit GeekShadow (Read error: No route to host)
16:20:34	wodz	I always wonder why the hell people write standards. It seems that standards are made only for producers to not conform to them
16:21:09	wodz	4GB Transcendend card had the same problem in my MPIO
16:21:44		Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
16:21:57		Join Horscht [0] (~Horscht@xbmc/user/horscht)
16:22:05		Join Alchimysta [0] (~a4844b20@giant.haxx.se)
16:22:54	Alchimysta	There are italian in this chat?? I have a big problem...!!
16:28:06		Quit Alchimysta (Quit: CGI:IRC (EOF))
16:29:28		Join Topy [0] (~Topy44@f048113253.adsl.alicedsl.de)
16:31:50		Quit T44 (Ping timeout: 245 seconds)
16:32:54	Stummi	lol
16:35:31	n1s	LambdaCalculus37: ah, good to know
16:36:08		Join user890104__ [0] (~Venci@213.226.63.34)
16:36:23		Nick user890104__ is now known as user890104 (~Venci@213.226.63.34)
16:36:42	n1s	i have given up on trying to find a working mini to cf mod that isn't pink for a decent price though so i'll probably buy a 32GB �SDHC for my fuze to get a flash player with space
16:36:55		Join jgarvey [0] (~jgarvey@cpe-065-190-066-089.nc.res.rr.com)
16:39:50		Quit Topy (Read error: Connection reset by peer)
16:41:04		Quit user890104 (Ping timeout: 255 seconds)
16:41:43	***	Saving seen data "./dancer.seen"
16:43:30		Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
16:44:51		Quit n1s (Quit: L�mnar)
16:48:22		Join toffe82 [0] (~chatzilla@maf.wirelesstcp.net)
16:55:29		Quit Horscht (Quit: Verlassend)
16:55:52		Join Horscht [0] (~Horschti@xbmc/user/horscht)
16:56:56		Quit benedikt93 (Quit: Bye ;))
17:00
17:06:59		Quit einhirn (Quit: Miranda IM! Smaller, Faster, Easier. http://miranda-im.org)
17:34:07		Join Buschel [0] (~chatzilla@p54B677FD.dip.t-dialin.net)
17:36:13	saratoga	Buschel: ping
17:43:39	saratoga	hmm nevermind
17:53:45	Llorean	S_a_i_n_t: Wouldn't you pull it off by adjusting the strings of those menu entries to include that character? Another option would be adding support for "icons on the right" and then making that the icon for "menus with submenus"
17:58:18	S_a_i_n_t	Llorean: Y'know...that last idea there might just be crazy enogh to work.
17:58:21		Join Super_Dork [0] (~Super_Dor@ip68-103-180-128.ks.ok.cox.net)
17:58:35	S_a_i_n_t	*enough
17:59:20	S_a_i_n_t	I knew it was a good idea to ask "the cloud" what I should/could do to achieve this.
18:00
18:00:44		Quit antil33t (Read error: Connection reset by peer)
18:00:54		Join antil33t [0] (~Mudkips@124-197-51-80.callplus.net.nz)
18:00:59	S_a_i_n_t	I'd thought about your first idea, but the font is not fixed width, so the results wouldn't be terrible accuratem especially having to rely on using spaces for alignment.
18:01:11		Quit MethoS- (Remote host closed the connection)
18:01:16	S_a_i_n_t	*terribly
18:02:54	Super_Dork	Once I post a message to the dev mailing list, about how long before I should expect a response?
18:03:20	Torne	minutes/days/week or two
18:03:28	S_a_i_n_t	Super_Dork: You may never get one, you may get one immediately.
18:03:43		Join TheLemonMan [0] (~lemonboy@ppp-58-213.32-151.iol.it)
18:03:44	Super_Dork	That's what I was afraid of.
18:03:51	S_a_i_n_t	It depends if some one knows the answer or not.
18:04:05	Super_Dork	And from my research, nobody does.
18:04:19	S_a_i_n_t	If you'd prefer a quick answer that might not be correct at all, I can give you one now ;)
18:05:10	Super_Dork	:) Unfortunately, I have a problem that can't be solved with words alone.:)
18:05:35	Super_Dork	I have a V2 Fuze that won't go past the splash screen.
18:06:16	Super_Dork	So I asked the USA devs if I could send one of them my player to look at.
18:06:31	Super_Dork	Maybe help solve the problem.
18:07:13		Quit saratoga (Quit: Page closed)
18:08:17		Join T44 [0] (~Topy44@f048113253.adsl.alicedsl.de)
18:10:12	LambdaCalculus37	Super_Dork: You can send it to one if you want to, but IIRC it was mentioned before that the USA developers had little to do with the port. (Someone correct me if I'm wrong!)
18:10:58	Super_Dork	Yeah, I know that but I asked just in case one wanted to take an interest.
18:11:54	LambdaCalculus37	Super_Dork: I haven't seen an email from you appear on the list yet, but what exactly is wrong with the Fuze v2 you have, aside from not progressing past the splash screen?
18:12:34	Super_Dork	If I could send it to somebody that did work on that port it would be better, for sure. However, I live in the US and so the shipping to out of country is expensive. btw it would be from dana.laberge@gmail.com
18:12:45		Quit Judas_PhD (Ping timeout: 250 seconds)
18:12:57	Super_Dork	Just that. Nothing I do gets it past the splash screen.
18:13:10	Super_Dork	Except booting into the OF.
18:13:21		Join Judas_PhD [0] (~kevin@misterfluffy.dsl.xmission.com)
18:13:36	LambdaCalculus37	Super_Dork: Posting an email in a logged channel is a bit of a bad idea... just letting ya know. ;)
18:13:46	Super_Dork	I know.
18:13:56	LambdaCalculus37	Super_Dork: You could email it to a staff or dev in the USA and we could pass it along.
18:15:15	Super_Dork	Maybe.
18:16:29	Super_Dork	I'll have to think about it.
18:16:47	LambdaCalculus37	No problem.
18:17:03		Quit Jerom (Quit: Leaving.)
18:17:21	Super_Dork	I looked at the dev map and theres one in Colorado.
18:17:37	Super_Dork	The shipping there wouldn't be too bad.
18:17:47	Super_Dork	I live one state away.
18:18:25	S_a_i_n_t	you really need to send it to a dev that's interested, not a dev that happens to live nearby
18:19:02	LambdaCalculus37	Super_Dork: I can take it and pass it along if you wish; I live in New Jersey so shipping to Europe would be a bit cheaper from the East Coast.
18:19:19	LambdaCalculus37	Super_Dork: And yes, I do count as a developer/staff. ;)
18:19:29	Super_Dork	Cool.
18:21:02		Join Alchimysta [0] (~a4841eda@giant.haxx.se)
18:21:50	Alchimysta	Ci sei?
18:22:04	LambdaCalculus37	Alchimysta: English, please.
18:22:08	Alchimysta	There are a italian in this chat??
18:22:12	Alchimysta	Help me!!!
18:22:24	Alchimysta	I have a problem..
18:23:08	LambdaCalculus37	Alchimysta: You've been saying that every single time you come in. Just state what's wrong and someone will try to help you.
18:23:34	Super_Dork	Isn't that kind of covered in the rules?
18:23:54	LambdaCalculus37	Super_Dork: Yup.
18:24:01	Super_Dork	Thought so.
18:28:17		Quit toffe82 (Read error: Connection reset by peer)
18:28:21	Alchimysta	What do I do?
18:28:30	Super_Dork	Well, I'm off.
18:28:33		Quit Super_Dork (Quit: Leaving)
18:29:08		Quit Alchimysta (Quit: CGI:IRC (EOF))
18:29:49		Join Strife89TX [0] (~cstrife89@207.144.201.128)
18:31:05		Quit TheLemonMan (Quit: Help me, i got shot! DIES)
18:33:44		Quit Evious (Disconnected by services)
18:34:15		Join Zeiris_ [0] (~Zeiris@66.183.73.102)
18:36:00		Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
18:39:20		Quit swilde (Remote host closed the connection)
18:39:37		Quit Kohlrabi (Quit: No Ping reply in 180 seconds.)
18:40:44		Join Kohlrabi [0] (~kohlrabi@kohlio.de)
18:41:46	***	Saving seen data "./dancer.seen"
18:55:36		Quit panni_ (Read error: Connection reset by peer)
19:00
19:03:47		Join Alchimysta [0] (~a4843f4f@giant.haxx.se)
19:04:29		Quit Alchimysta (Client Quit)
19:11:23		Join DSStrife89 [0] (~Strife89@207.144.201.128)
19:16:05		Join togetic [0] (~togetic@unaffiliated/ibuffy)
19:19:23		Quit CIA-7 (Ping timeout: 276 seconds)
19:28:07		Join kevku [0] (~kevku@arch.tunnel.ipv6.estpak.ee)
19:28:07		Quit markun (Read error: Connection reset by peer)
19:29:27		Join markun [0] (~markun@5ED33C2C.cm-7-4a.dynamic.ziggo.nl)
19:29:27		Quit markun (Changing host)
19:29:27		Join markun [0] (~markun@rockbox/developer/markun)
19:29:33		Join CIA-7 [0] (~CIA@208.69.182.149)
19:34:08		Quit DSStrife89 (Quit: ClIRC v0.05 (homebrew IRC client for Nintendo DS))
19:43:28		Quit Strife89TX (Quit: Vamoose!)
19:43:52	Buschel	saratoga: pong
19:50:36		Join kugel [0] (~kugel@rockbox/developer/kugel)
19:50:54	kugel	I've found the cause of Rockbox idle cpu usage in the android port
19:51:25	S_a_i_n_t	oh?
19:51:28	S_a_i_n_t	nice.
19:51:53	kugel	6-7% on my phone, but apparently up to 15% on a galaxy s
19:52:01	kugel	with the fix it's down to zero
19:52:36		Quit dfkt (Quit: -= SysReset 2.53=- Sic gorgiamus allos subjectatos nunc.)
19:52:46		Nick YPSY is now known as Ypsy (~ypsy@geekpadawan.de)
19:57:01		Join panni_ [0] (hannes@ip-178-203-81-220.unitymediagroup.de)
20:00
20:00:26		Quit mystica555_ (Ping timeout: 255 seconds)
20:02:40		Join MethoS- [0] (~clemens@134.102.106.250)
20:03:25		Join saratoga [0] (9803c6dd@gateway/web/freenode/ip.152.3.198.221)
20:04:06	saratoga	Buschel: I looked at the dct32 used in mpc and mp3, and think it could probably be speed up a good bit on arm if the asm version was rewritten
20:04:42	saratoga	most of the trig accesses can be arranged so that they're sequential, and could probably use 16 bit packed values like the atrac changes
20:07:24	saratoga	also are the two dcts really identical? if so, we should put them in the codec lib and just use one for both
20:08:29	Buschel	saratoga: they are not 100% identical. the output is sorted differently and mpc uses 64 output samples instead of 32 −− to be able to use further optimization in the dewindowing
20:09:24		Join JesusFreak316 [0] (~JesusFrea@pool-173-65-109-252.tampfl.fios.verizon.net)
20:09:26		Quit riotz (Read error: Operation timed out)
20:09:46		Join riotz [0] (riot33r@gateway/shell/shellium.org/x-swytrihvklwklpby)
20:09:56	Buschel	saratoga: 16 bit for the costab values will be too few precision. i experimented on this for a while... i guess this is caused by the butterfly architecture which lowers precision in each stage
20:10:12	saratoga	thats good to know
20:10:37	saratoga	the mpc's dct order actually seemed like a disadvantage to me
20:11:02	saratoga	since it requires you to make very largely spaced loads
20:11:26	saratoga	thats fine on arm7 and coldfire, but quite painful on arm9 and arm11
20:12:22		Join InsDel [0] (~haqr.net@c-98-231-87-43.hsd1.fl.comcast.net)
20:13:57	Buschel	saratoga: the mpc dewndowing asm shows what is the advantage, you only need to load ~50% of the coefs compared to the reference implementation −− as the are symmetries. at the cost of 32 additional writes per dct and a doubled effort for the memmove.
20:14:08	Buschel	saratoga: expensive on arm9, true
20:14:42	Buschel	btw, I made some profiling on nano2g for mp3 and mpc, interested?
20:14:46	saratoga	yes very
20:14:58	Buschel	wait a few seconds
20:15:16	saratoga	btw, i'm now convinced that the libmad synthesis filter can be rewritten so that at very least all loads can be done two at a time, and possibly 4 at a time, which should lead to a very large increase in speed on arm11
20:15:23		Quit fdinel (Quit: Miranda IM! Smaller, Faster, Easier. http://miranda-im.org)
20:16:23	Buschel	http://pastie.org/1293417
20:17:25	Buschel	saratoga: bitstream = bit buffer handling / huffman decoding
20:18:25	saratoga	interesting
20:19:10	Buschel	you see that mpc's synthesis is still a little faster, but the mp3 synthesis isn't fully optimized
20:19:19		Nick Ypsy is now known as YPSY (~ypsy@geekpadawan.de)
20:19:53	Buschel	i am convinced that the mp3 synth will overcome mpc on arm9 with a bit more teaking −− mp3 implementation is less memory intense
20:20:19		Join stoffel [0] (~quassel@p57B4D122.dip.t-dialin.net)
20:21:09		Quit MethoS- (Remote host closed the connection)
20:21:17		Quit factor (Read error: Connection reset by peer)
20:21:29	saratoga	yes the current synth filter is quite bad on arm9
20:21:37	saratoga	it was not scheduled correctly
20:21:38	Buschel	btw, you can read the results? left side = over all needed MHz, right side = MHz for this single step
20:21:43	saratoga	yeah
20:23:25	saratoga	right now in the mp3 version, every single load has 1 cycle stall due to bad scheduling on arm9, and a 3 cycle one on arm11
20:23:47	saratoga	that can be trivially corrected, but getting larger speed ups will require tinkering with the memory accesses and precision
20:24:07		Join Kupop [0] (~Kupo@cpc2-bsfd7-2-0-cust220.5-3.cable.virginmedia.com)
20:24:29		Join factor [0] (~factor@r74-195-220-23.msk1cmtc02.mskgok.ok.dh.suddenlink.net)
20:24:43	Buschel	mp3 = needs ~50% of mpc's memory / rotates the filterbanck coefs, so that the are no symmetries exploitable versus mpc = doubled memory usage / reduced loads due to exploit of symmetries
20:24:58	Buschel	mpc should be less effective on CPU's with slow ram
20:25:13	Buschel	a pitty that the nano 2g cpu is one of those...
20:26:42	saratoga	for targets without IRAM, its worth wasting ram if it allows for better caching
20:26:54	*	Buschel should check his mpc resort patch again after his alignment change
20:27:22	saratoga	I am thinking about rewriting the mdctlib eventually to work differently on arm9e/arm11 such that the trig tables are 50% duplicated, but are arranged sequentially for the most common block size
20:27:38	saratoga	i expect this to give a very large speed up on arm11
20:28:24	Buschel	would this be same for arm9?
20:32:38	saratoga	its not so important there
20:37:57	saratoga	on arm9 you have a 1 cycle delay before a loaded register can be used or else you stall
20:38:19	saratoga	on arm11 you have a 3 cycle delay (IIRC) and furthermore, the load/store path is 64 bits wide so you can do two at once
20:39:04	saratoga	so in theory ldm with proper pipeling can do 2 words a clock, whereas the worst case ldr without pipelining can do 0.25 words per clock
20:41:49	***	Saving seen data "./dancer.seen"
20:46:08	saratoga	we should really start a wiki page with all this info
20:46:15	saratoga	its so hard to keep track which arm core needs what
20:46:20	saratoga	and i don't even bothered with coldfire or mips
20:47:22	Buschel	you are volunteering to start with it ? ;)
20:47:50		Join xxcv [0] (~IMNOTADEV@c211-30-174-99.carlnfd1.nsw.optusnet.com.au)
20:48:15	saratoga	if no one beats me to it, i'll try to find time later
20:49:55		Quit panni_ (Read error: Connection reset by peer)
20:50:56		Join panni_ [0] (hannes@ip-178-203-81-220.unitymediagroup.de)
20:53:14	saratoga	whats a good page name, ProcessorOptimization?
20:54:24	Buschel	CPUspecificOptimization (the camel case writing looks strange though..)
20:54:27	Buschel	?
20:55:45	Buschel	HowToOptimizePerformance
20:56:10	saratoga	TargetSpecificOptimization
20:56:16	Buschel	good!
20:56:21	Buschel	i like that
20:58:54	kugel	Buschel, saratoga: why didn't you use CACHEALIGN_ATTR in your commits?
20:59:14	kugel	(and fix the attr to be 32 bytes on armv5+
20:59:14	saratoga	laziness
20:59:15	kugel	)
21:00
21:01:40		Join anewuser [0] (anewuser@unaffiliated/anewuser)
21:06:04	saratoga	Buschel: http://www.rockbox.org/wiki/TargetSpecificOptimization
21:06:10	saratoga	will fill out with actual info when I get time
21:06:53		Quit Kupop (Ping timeout: 272 seconds)
21:07:41		Join toffe82 [0] (~chatzilla@maf.wirelesstcp.net)
21:08:13	Buschel	kugel: CACHEALIGN_ATTR did not work for me on nano 2g. when checking system.h in the buld process it is defined to ""
21:08:37	kugel	strange
21:08:43	Buschel	PROC_NEEDS_CACHEALIGN is not defined
21:09:12	kugel	IIRC mmu-arm.S depends on it being set
21:10:14		Join Kupop [0] (~Kupo@cpc2-bsfd7-2-0-cust220.5-3.cable.virginmedia.com)
21:10:22	kugel	it should be always defined IMO
21:10:38	Buschel	it is define at one single place: system-target.h
21:12:11	kugel	system.h #includes system-target.h
21:17:32		Join Buschel_ [0] (~chatzilla@p54B677FD.dip.t-dialin.net)
21:18:03		Quit Buschel (Ping timeout: 245 seconds)
21:18:13		Nick Buschel_ is now known as Buschel (~chatzilla@p54B677FD.dip.t-dialin.net)
21:18:46		Quit stoffel (Ping timeout: 240 seconds)
21:21:34		Join benedikt93 [0] (~benedikt9@unaffiliated/benedikt93)
21:32:13		Join freddyb [0] (~chatzilla@216.8.239.112.etczone.com)
21:33:31	saratoga	http://www.rockbox.org/wiki/TargetSpecificOptimization
21:33:38	saratoga	just dumped a mountain of info on that wiki
21:33:42	saratoga	got to get back to work now
21:34:40	freddyb	Could someone smart take a look at my patch for FS #11608? I switched from sync to async bus and it fixed my playback but I don't if it might cause other problems.
21:35:28	saratoga	i think async is supposed to be slower
21:35:47	saratoga	you should probably ask funman, he discussed the clocking modes in great detail before
21:36:17	freddyb	I was hoping he'd be online, but no luck.
21:36:55	saratoga	i sent you his email
21:37:10	freddyb	Thanks.
21:39:10		Join bmbl [0] (~bmbl@dsl16-200.pool.bitel.net)
21:39:17		Quit bmbl (Changing host)
21:39:17		Join bmbl [0] (~bmbl@unaffiliated/bmbl)
21:39:59		Quit freddyb (Quit: ChatZilla 0.9.86 [Firefox 3.6.8/20100723085406])
21:50:44		Join fdinel [0] (~Miranda@modemcable235.127-131-66.mc.videotron.ca)
21:58:28		Quit LambdaCalculus37 (Quit: All aboard for funtime!)
22:00
22:03:13		Join Kitr88 [0] (~Kitarist@BSN-182-143-183.dial-up.dsl.siol.net)
22:04:29	CIA-7	New commit by kugel (r28564): Android: Change how detecting call state (introduced in r27746) works, from polling to event based. ...
22:05:10	saratoga	TheSeven: would you mind adding the Nano2G's processor here: http://www.rockbox.org/wiki/TargetSpecificOptimization
22:06:35		Quit Kitar\|st (Ping timeout: 255 seconds)
22:06:37	CIA-7	r28564 build result: All green
22:07:20		Quit Kitr88 (Ping timeout: 240 seconds)
22:07:48		Quit riotz (Changing host)
22:07:48		Join riotz [0] (riot33r@unaffiliated/riotz)
22:07:48		Quit riotz (Changing host)
22:07:48		Join riotz [0] (riot33r@gateway/shell/shellium.org/x-swytrihvklwklpby)
22:09:28		Join robin0800 [0] (~robin0800@genld-224-239.t-mobile.co.uk)
22:10:06		Join Kitar\|st [0] (Kitarist@BSN-182-78-196.dial-up.dsl.siol.net)
22:15:01		Quit toffe82 (Read error: Connection reset by peer)
22:15:11		Quit robin0800 (Ping timeout: 245 seconds)
22:21:22	CIA-7	New commit by kugel (r28565): Android: Fix small error in install&run script.
22:21:39	CIA-7	New commit by kugel (r28566): Android: Fix compilation failure
22:23:06	CIA-7	r28565 build result: All green
22:24:52	CIA-7	r28566 build result: All green
22:25:38		Quit kevku (Read error: Operation timed out)
22:27:33		Join stoffel [0] (~quassel@p57B4D122.dip.t-dialin.net)
22:32:18	TheSeven	saratoga: I can add some details about the S5L8701 SoC, but honestly I don't know much about the ARM940T core in it, or the memory clocking
22:32:36		Quit GeekShadow (Read error: No route to host)
22:34:19		Quit benedikt93 (Quit: Bye ;))
22:37:07		Join Trixx [0] (~chatzilla@cpc2-slam1-0-0-cust926.2-4.cable.virginmedia.com)
22:40:27	Buschel	kugel: #define PROC_NEEDS_CACHEALIGN is only set in arm/target-config.h, but not i the CPU specific ones like arm/s5l8700/target-config.h
22:40:50	kugel	what is target-config.h?
22:40:59		Quit stoffel (Remote host closed the connection)
22:41:19	Buschel	sorry, system-target.h
22:41:52	***	Saving seen data "./dancer.seen"
22:42:04		Join freddyb [0] (~chatzilla@pool-71-171-240-252.chi01.dsl-w.verizon.net)
22:42:05	kugel	ah, meh, that arm/system-target.h is only meant for PP, apparently
22:46:14	Buschel	yes
22:48:06	kugel	I suspect cache align attrs were forgotten for newer targets
22:49:19		Quit freddyb (Quit: ChatZilla 0.9.86 [Firefox 3.6.8/20100723085406])
22:49:49	Buschel	it seems like it isn't set for any target except PP
22:51:36		Join kevku [0] (~kevku@arch.tunnel.ipv6.estpak.ee)
22:52:38	*	TheSeven thinks he fixed that some time ago for s5l870x
22:52:52	TheSeven	CACHEALIGN_ATTR appears to be defined at least
22:53:13	TheSeven	(to __attribute__((aligned(16))))
22:53:16	Buschel	i am just preparing a patch
22:53:30	CIA-7	New commit by wodz (r28567): MPIO: move files a bit in preparation for HD300 port
22:53:36	Buschel	CACHEALIGN_ATTR is defined to ""
22:53:40	Buschel	empty
22:53:41		Part Trixx
22:53:45	*	TheSeven doubts that
22:54:00	TheSeven	at least it used to work some time ago
22:54:26	TheSeven	if someone broke it, this might explain all the weird trouble we've been experiencing recently
22:54:35		Join KiwiCam [0] (~Kiwicam@ip-118-90-86-119.xdsl.xnet.co.nz)
22:55:07	CIA-7	r28567 build result: All green
22:57:55	Buschel	for S5L870x: http://pastie.org/1293801
22:58:06	TheSeven	it must be pure luck that USB is working at all under these circumstances
22:58:07		Quit bmbl (Quit: Verlassend)
22:58:34	TheSeven	while it will just cause weird cache coherency trouble for other DMA, the USB core's AHB master seems to be very picky about alignment
22:58:47	Buschel	all other CPUs need such change in their respective system-target.h as well
22:58:53		Quit kevku (Quit: KVIrc 4.0.2 Insomnia http://www.kvirc.net/)
22:59:04	TheSeven	are you sure that this isn't defined in config/ipodnano2g.h?
23:00
23:00:05	Buschel	yes
23:00:12	*	Buschel just checked
23:00:58	CIA-7	New commit by kugel (r28568): Android: Use adb install -r to reinstall the app, as it doesn't remove icons from the homescreen.
23:01:31	Buschel	I will recompile now and make a short test
23:01:34	kugel	Buschel: perhaps it should always default to 32, leaving the possibility to have smaller aligns if wanted
23:02:14	kugel	alignment doesn't hurt but potentially increases performance drastically, so I'd go for "opt-out" rather than "opt-in"
23:02:35	CIA-7	r28568 build result: All green
23:02:36	*	TheSeven blames funman
23:03:16	TheSeven	r25633 was the culprit
23:03:39	Buschel	kugel: it may default to 32, if not set. but right now the default is that there is no alignment at all
23:05:02	TheSeven	Buschel: apparently funman moved it to export/s5l8700.h
23:05:12	TheSeven	and apparently that isn't included early enough
23:05:24	TheSeven	so what's the proper way to do this?
23:05:43	Buschel	see my patch I have linked above
23:06:15	TheSeven	IIRC we don't need PROC_NEEDS_CACHEALIGN
23:06:27	TheSeven	this seems to be a multicore thing
23:08:08	TheSeven	and funman intentionally removed the CACHEALIGN_SIZE define in r25633, probably for a reason
23:09:04	Buschel	CACHEALIGN_SIZE = (1<<CACHEALIGN_BITS), just a matter of taste what to define
23:09:32	Buschel	but to me it seems like PROC_NEEDS_CACHEALIGN needs to be defined to have CACHEALIGN_ATTR defined correctly
23:09:32	CIA-7	New commit by kugel (r28569): Android: When the backlight goes off, the RockboxFramebuffer/View isn't made invisible. Force that in order to disable screen updates while backlight ...
23:11:10	CIA-7	r28569 build result: All green
23:11:40	Buschel	TheSeven: see r25337
23:12:18	Buschel	ahem, r25339 (the diff to 25337)
23:13:19	TheSeven	yeah, he also broke STORAGE_WANTS_ALIGN... (which i fixed in the meantime)
23:16:23	TheSeven	what I would do now is remove the definition from s5l8700.h (which doesn't seem to belong there), and add #define CACHEALIGN_BITS 4 to firmware/target/s5l8700/system-target.h
23:17:01	Buschel	and define PROC_NEEDS_CACHEALIGN to define macros
23:17:01		Join GeekShadow [0] (~Antoine@44.181.204-77.rev.gaoland.net)
23:17:02		Quit GeekShadow (Changing host)
23:17:02		Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
23:18:39	TheSeven	Buschel: http://www.rockbox.org/irc/log-20100620#19:19:14
23:22:24	Buschel	if you do not define it, the define for CACHEALIGN_ATTR is empty and not defined to "__attribute__ ((aligned(16)))""
23:22:49	TheSeven	which makes me wonder if CACHEALIGN_ATTR is actually what we want to use for DMA
23:23:55	Buschel	and also: if we do this change for S5L870, what do we do for all other ARM cpu's?
23:24:18	TheSeven	how are the other CPUs using CACHEALIGN_ATTR?
23:24:56	TheSeven	it might be possible that I abused this when writing the S5L870x drivers without knowing what it actually means, and that nobody else uses it for that purpose
23:25:23	Buschel	to me it seems like the former implementation (pre 25339) defined the macros, if CACHEALIGN_BITS was defined. is it possible to implement it this way again? CACHEALIGN_BITS for other cpu's is also defined in the export\cpu.h
23:25:31	TheSeven	how is the alignment that's neccessary for PCM buffers being determined?
23:25:39	Buschel	cpu = S5L8700 as3525 etc.
23:26:07	TheSeven	...which doesn't seem to be included globally for s5l8700.h, so it won't work
23:26:32	TheSeven	or did I miss where it is included from?
23:27:46		Join Luca_S [0] (~5d2b980e@giant.haxx.se)
23:31:21		Quit komputes (Quit: I haven't slept for ten days, because that would be too long.)
23:38:22		Quit Luca_S (Quit: CGI:IRC)
23:41:19	TheSeven	Buschel: I just grepped my working copy, and can only find 6 places where CACHEALIGN_ATTR is used at all
23:42:01	Buschel	yet, the idea is to use it for target specific codec optimization to avoid all those __attribute__ (()) stuff
23:42:21	Buschel	also, the other macros are used at some places
23:43:13	Buschel	the easiest way seems to be to just use "#if 1" instead of "#ifdef PROC_NEEDS_CACHEALIGN" in export/system.h
23:43:43	Buschel	this way those defines are available and the cpu specific CACHEALIGN_BITS is used
23:43:45	TheSeven	there are also STORAGE_ALIGN_ATTR and USB_DEVBSS_ATTR
23:49:25	TheSeven	grepping case-insensitively for ((STORAGE_ALIGN\|USB_DEVBSS\|CACHEALIGN\|ALIGNED)_ATTR\|attribute.*aligned\() reveals quite some mess
23:51:42	Buschel	how are coldfire CPUs performing with aligned buffers?
23:51:56	Buschel	cuold they use the same define?
23:52:00	Buschel	*could
23:52:08		Quit jgarvey (Quit: Leaving)
23:53:29		Join Topy [0] (~Topy44@f049101069.adsl.alicedsl.de)
23:56:09		Quit ender` (Quit: Of course, this is not a very secure program... But, we weren't trying to write a security system, just an example for page 9 in this book. -- Learning Perl, 2nd Edition)
23:56:51		Quit T44 (Ping timeout: 245 seconds)

Previous day | Next day