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

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

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

#rockbox log for 2011-02-18

00:00:32 Join JdGordon1 [0] (~jonno@vl10.gw.ok-labs.com)
00:01:22 Quit bluebroth3r (Ping timeout: 255 seconds)
00:04:31 Quit GeekShad0w (Ping timeout: 240 seconds)
00:09:13 Join Guinness` [0] (Slayer@c-68-55-111-159.hsd1.va.comcast.net)
00:11:31 Quit Guinness (Ping timeout: 240 seconds)
00:16:01 Quit ender` (Quit: All power corrupts, but we need electricity.)
00:18:54 Join Keripo [0] (~Keripo@eng173.wireless-resnet.upenn.edu)
00:32:22 Quit kevku (Ping timeout: 246 seconds)
00:42:34 Quit pamaury (Remote host closed the connection)
00:44:49*TheSeven is running out of ideas
00:45:28TheSeveni've even done captures of the GPIO and MMC controller state in diagmode, but can't spot any obvious problem...
00:47:04sideraljhMikeS: Thanks for uploading your readings!
00:47:38sideralI'll summarize my findings in that tracker item as soon as I find the time.
00:47:44TheSevensideral: you can of course manually boost/unboost, if that is of any use
00:48:17sideraljhMikeS: Let me know if you manage to reproduce the ClipV1 noise...
00:48:34sideralTheSeven: How so? Or do you mean by changing the code?
00:48:43TheSeventhe "CPU frequency" debug screen
00:48:55TheSevenyou can increase/decrease the boost counter using some buttons
00:49:05sideraloh?
00:49:09TheSevenon the ipods it's turning the wheel, no idea about other targets
00:50:47sideralprobably no key is mapped for the Clip
00:51:00sideralI'll investigate tomorrow −− thanks for the hint!
00:51:02*TheSeven doubts that
00:52:19TheSevensideral: it's ACTION_STD_PREV and ACTION_STD_NEXT, so the buttons you usually use to scroll in a list
00:52:47[Saint]I'm quite certain it's the same for all targets.
00:53:19TheSeven[Saint]: in that it is PREV/NEXT, yes. but that's of course not the same button on all targets :)
00:53:41sideralthose would be the up and down buttons on the Clip. On the HW info screen "Clock Frequencies", Up does nothing, and Down switches to the next screen.
00:53:51[Saint]yes, well...at least you knew what I meant ;)
00:54:04TheSevensideral: that's the wrong screen
00:54:26TheSeventhere is (or at least should be) a dedicated "cpu frequency" entry in the debug menu
00:54:44TheSeven(which shows nothing but the current frequency and boost counter
00:55:11sideralcan't find it. perhaps my target cannot boost the CPU?
00:55:26[Saint]errrrrm.
00:56:08TheSevenapparently, as it's #ifdef HAVE_ADJUSTABLE_CPU_FREQ
00:56:32TheSevensideral: which exact target is it?
00:56:42[Saint]system - debug - CPU frequency (if it's not there, it can't)
00:56:45sideralClipV2
00:57:01sideraldoesn't have HAVE_ADJUSTABLE_CPU_FREQ
00:57:16TheSevenyep
00:57:30TheSevenso that one probably always runs at the full clock currently
00:58:07sideralApparently. I think the SoC has some "dynamic power management", though
00:58:24 Nick krazykit is now known as kkit|sh (~kkit@li135-248.members.linode.com)
00:59:14*TheSeven wonders what to do about that CE-ATA drive not working, and decides to go to sleep and continue poking at it tomorrow
01:00
01:00:49sideralAnyway, need to log off now. Thanks for your assistance!
01:01:43 Join krazykit [0] (~krazykit@99-126-205-52.lightspeed.cicril.sbcglobal.net)
01:03:47 Quit sideral (Quit: Leaving.)
01:20:23Suit_Of_Sablesdoes anyone in here use EAC under wine?
01:20:35krazykitwhat does that have to do with rockbox?
01:39:45***Saving seen data "./dancer.seen"
01:41:07Strife89Suit_Of_Sables: /join #winehg
01:41:16Strife89* #winehq
01:44:23Suit_Of_SablesStrife89: Yeah, wasn't getting any help in there today :*( I just thought there would be some linux users who are into ripping CDs in this channel. sorry for that
01:45:57Strife89Suit_Of_Sables: Sorry to hear that.
01:47:35Strife89Suit_Of_Sables: You might consider joining the channel that supports your distro, or joining #winehq again to see if there are different people in there.
01:52:26Strife89Suit_Of_Sables: See PM
02:00
02:04:36 Quit liar (Quit: Leaving)
02:11:55 Quit panni_ (Read error: Connection reset by peer)
02:12:37 Join panni_ [0] (hannes@ip-178-203-73-7.unitymediagroup.de)
02:44:39 Quit MethoS- (Remote host closed the connection)
02:46:27 Quit dfkt (Quit: -= SysReset 2.53=- Sic gorgiamus allos subjectatos nunc.)
03:00
03:03:28 Quit leavittx_ (Ping timeout: 272 seconds)
03:08:44 Join Dreamxtreme [0] (Dre@92.30.106.22)
03:09:52 Nick scorche` is now known as scorche (~scorche@rockbox/administrator/scorche)
03:10:50 Quit mudd1 (Ping timeout: 255 seconds)
03:11:01 Quit afk (Ping timeout: 276 seconds)
03:11:44 Join afk [0] (Dre@92.30.243.114)
03:13:05 Quit Dreamxtreme (Ping timeout: 246 seconds)
03:13:33 Join Dreamxtreme [0] (Dre@92.30.107.33)
03:17:39 Quit afk (Ping timeout: 276 seconds)
03:18:40 Quit Dreamxtreme (Ping timeout: 272 seconds)
03:39:47***Saving seen data "./dancer.seen"
03:56:20 Quit krazykit (Quit: awe yeeeeeee)
03:58:59 Join krazykit [0] (~krazykit@99-126-205-52.lightspeed.cicril.sbcglobal.net)
04:00
04:00:18 Quit Suit_Of_Sables (Remote host closed the connection)
04:06:24 Quit DerPapst1 (Quit: Leaving.)
04:15:54 Nick Rondom_ is now known as Rondom (~rondom@lvps178-77-79-47.dedicated.hosteurope.de)
04:27:13 Quit Battousai (Ping timeout: 260 seconds)
04:37:48 Quit amiconn (Disconnected by services)
04:37:48 Join amiconn_ [0] (quassel@rockbox/developer/amiconn)
04:37:51 Quit pixelma (Disconnected by services)
04:37:51 Join pixelma_ [0] (quassel@rockbox/staff/pixelma)
04:38:06 Nick amiconn_ is now known as amiconn (quassel@rockbox/developer/amiconn)
04:38:07 Nick pixelma_ is now known as pixelma (quassel@rockbox/staff/pixelma)
04:39:57 Join Battousai [0] (~bryan@gentoo/developer/battousai)
04:45:27 Quit factor (Read error: Connection reset by peer)
04:45:51 Join factor [0] (~factor@75.108.68.114)
04:48:25 Quit factor (Client Quit)
04:48:28JdGordon1Buschel: going over last nights logs, my mini2g also occasionally does some really bad lockups, including playing static when it does
04:48:44JdGordon1but I always assumed it was the CF card or filesystem stuffing up
04:49:00 Join factor [0] (~factor@75.108.68.114)
04:58:27 Quit TheSeven (Ping timeout: 240 seconds)
04:59:13 Quit Barahir_ (Read error: Operation timed out)
05:00
05:02:01 Join Barahir [0] (~jonathan@frnk-590f7c59.pool.mediaWays.net)
05:02:35 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
05:13:03 Quit factor (Ping timeout: 240 seconds)
05:13:31 Join factor [0] (~factor@75.108.68.114)
05:14:55 Join Dreamxtreme [0] (Dre@92.30.245.150)
05:29:07 Join t0rc [0] (~t0rc@unaffiliated/t0rc/x-5233201)
05:29:18 Join Zarggg [0] (~zarggg@24.229.139.169.res-cmts.sm.ptd.net)
05:30:08 Join Rob2222 [0] (~Miranda@p4FFF2A10.dip.t-dialin.net)
05:34:43 Quit Rob2223 (Ping timeout: 276 seconds)
05:39:51***Saving seen data "./dancer.seen"
05:39:56 Quit Horscht (Quit: Verlassend)
05:48:57 Quit antil33t ()
05:59:13 Join Judas_PhD [0] (~kevin@misterfluffy.dsl.xmission.com)
06:00
06:14:50 Quit elcan (Ping timeout: 240 seconds)
06:15:39 Join elcan [0] (user36@pr0.us)
06:31:26 Quit Battousai (Quit: Leaving)
06:32:26 Join Battousai [0] (~bryan@gentoo/developer/battousai)
06:35:32 Quit Zarggg (Quit: Zarggg)
06:35:32 Quit milk (Quit: baaaiiii)
06:48:47 Quit panni_ (Quit: ( www.nnscript.de :: NoNameScript 3.81 :: www.XLhost.de ))
06:51:03 Join Izzy__ [0] (~Izzy@50.35.228.110)
06:53:33 Join BHSPitMonkey [0] (~stephen@unaffiliated/bhspitmonkey)
06:54:33 Quit Izzy_ (Ping timeout: 276 seconds)
06:56:50 Join Unhelpful_ [0] (~quassel@rockbox/developer/Unhelpful)
07:00
07:00:50 Quit Unhelpful (Ping timeout: 240 seconds)
07:08:33 Quit JdGordon1 (Quit: leaving)
07:09:39 Quit mystica555 (Remote host closed the connection)
07:16:51 Quit factor (Read error: Connection reset by peer)
07:17:58 Join factor [0] (~factor@75.108.68.114)
07:26:19 Join mystica555 [0] (~Mike@m3e2636d0.tmodns.net)
07:35:44 Join Buschel [0] (~chatzilla@p54A39D29.dip.t-dialin.net)
07:38:37BuschelJdGordon|: could try to disable HAVE_ATA_DMA for your build and see if the lockups still occur?
07:39:53***Saving seen data "./dancer.seen"
07:42:08JdGordon|Buschel: well the problem is there is no reliable repro (for me anyway) and recently i've noticed it alot less (i think)... but sure
07:42:19JdGordon|if i remember tomorow night, unless you want to build me one quickly? :)
07:43:15Buschelwell, to have reasonable results from such build it should be used for a while. so, no hurry
07:46:30JdGordon|i can leave it going for the next 25h and see if it makes it the whole?
07:50:35Buschelwhy not, if this is normally sufficient to get it stuck.
07:54:51 Quit t0rc (Remote host closed the connection)
07:57:39 Join eGen_ [0] (generat0r@gate.mmdecin.cz)
08:00
08:04:28 Join esperegu [0] (~quassel@145.116.15.244)
08:15:07 Join GeekShadow [0] (~Antoine@ree79-1-78-237-225-34.fbx.proxad.net)
08:15:07 Quit GeekShadow (Changing host)
08:15:07 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
08:19:53 Quit Buschel (Quit: ChatZilla 0.9.86 [Firefox 3.6.13/20101203075014])
08:22:38 Join Loto [0] (~nfs@S01060012171a84e3.no.shawcable.net)
08:22:38 Quit Loto (Changing host)
08:22:39 Join Loto [0] (~nfs@xbmc/user/Loto)
08:27:19 Join FoH [0] (~foh@adsl-240-126-222.bhm.bellsouth.net)
08:29:22 Quit JdGordon| (Ping timeout: 255 seconds)
08:34:23 Join JdGordon| [0] (~jonno@124-168-132-222.dyn.iinet.net.au)
08:34:23 Quit JdGordon| (Changing host)
08:34:23 Join JdGordon| [0] (~jonno@rockbox/developer/JdGordon)
08:38:14 Join B4gder [0] (~danielx@2a00:1a28:1200:9::2)
08:38:14 Quit B4gder (Changing host)
08:38:15 Join B4gder [0] (~danielx@rockbox/developer/bagder)
08:41:04 Quit Keripo (Quit: Leaving.)
08:45:39 Join Keripo [0] (~Keripo@eng173.wireless-resnet.upenn.edu)
08:47:00 Join ender` [0] (krneki@foo.eternallybored.org)
08:54:08JdGordon|Buschel: ok, I've just loaded svn+!HAVE_ATA_DMA, If its still playing this time tomorrow night I'll do another test tomorow night without the change
08:54:34JdGordon|I do have a second mini2g although its not really an equivilant test (that one doesnt have CF and no idea if it shows the problem)
08:54:49 Quit eGen_ (Quit: ... gettin' screew my wife ....)
08:54:57 Quit mystica555 (Ping timeout: 240 seconds)
08:58:27 Join mystica555 [0] (~Mike@m3e2636d0.tmodns.net)
09:00
09:05:32 Join mudd1 [0] (~cmertes@ip-78-94-203-49.unitymediagroup.de)
09:05:39 Join Zagor [0] (bjst@rockbox/developer/Zagor)
09:06:35 Join petur [0] (~petur@rockbox/developer/petur)
09:14:18 Join n1s [0] (~n1s@sb-fw.bmc.uu.se)
09:14:19 Quit n1s (Changing host)
09:14:19 Join n1s [0] (~n1s@rockbox/developer/n1s)
09:15:42 Join kevku [0] (~kevku@2001:7d0:0:f9af:feed:feed:feed:feed)
09:28:28 Quit kevku (Ping timeout: 272 seconds)
09:30:02 Join kevku [0] (~kevku@2001:7d0:0:f9af:feed:feed:feed:feed)
09:37:04 Join leavittx_ [0] (~lev@89.221.199.187)
09:38:36 Quit user890104 (Ping timeout: 272 seconds)
09:39:57***Saving seen data "./dancer.seen"
09:44:40 Join shai [0] (~Shai@l192-117-110-233.cable.actcom.net.il)
09:59:02 Quit sasquatch (Quit: WeeChat 0.3.2)
09:59:26 Join sasquatch [0] (~username@p4FF2D980.dip.t-dialin.net)
10:00
10:03:48 Quit BHSPitMonkey (Remote host closed the connection)
10:07:29 Join eGen_ [0] (generat0r@gate.mmdecin.cz)
10:07:51 Join sideral [0] (~sideral@213.165.85.248)
10:07:51 Quit sideral (Changing host)
10:07:51 Join sideral [0] (~sideral@rockbox/developer/sideral)
10:19:06sideralliteral: Have you had a chance to try whether the background noise is also audible during regular playback (when there's no disk access, no UI interaction)?
10:19:08 Quit leavittx_ (Ping timeout: 272 seconds)
10:32:11 Join LinusN [0] (~linus@rockbox/developer/LinusN)
10:35:35 Join DerPapst [0] (~Alexander@dslb-188-107-174-146.pools.arcor-ip.net)
10:42:57 Join sideral1 [0] (~sideral@213.165.85.248)
10:42:57 Quit sideral (Disconnected by services)
10:42:57 Nick sideral1 is now known as sideral (~sideral@213.165.85.248)
10:42:58 Quit sideral (Changing host)
10:42:58 Join sideral [0] (~sideral@rockbox/developer/sideral)
10:45:12 Join casainho [0] (~chatzilla@pal-213-228-181-14.netvisao.pt)
10:45:13 Quit sideral (Remote host closed the connection)
10:46:16 Join sideral [0] (~sideral@213.165.85.248)
10:46:16 Quit sideral (Changing host)
10:46:16 Join sideral [0] (~sideral@rockbox/developer/sideral)
10:55:29 Quit sideral (Remote host closed the connection)
10:56:18 Join sideral [0] (~sideral@213.165.85.248)
10:56:18 Quit sideral (Changing host)
10:56:18 Join sideral [0] (~sideral@rockbox/developer/sideral)
11:00
11:00:26 Quit casainho (Remote host closed the connection)
11:04:42 Quit GeekShadow (Ping timeout: 250 seconds)
11:05:43 Quit sideral (Remote host closed the connection)
11:05:59 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
11:06:16 Join sideral [0] (~sideral@213.165.85.248)
11:06:16 Quit sideral (Changing host)
11:06:16 Join sideral [0] (~sideral@rockbox/developer/sideral)
11:19:54 Join dfkt [0] (dfkt@unaffiliated/dfkt)
11:39:58***Saving seen data "./dancer.seen"
11:50:25 Join Kitr88 [0] (~Kitarist@BSN-182-104-42.dial-up.dsl.siol.net)
11:52:31Zagorhmm, I want to move draw_album_art() out of albumart.c. but where to?
11:54:05 Quit Kitar|st (Ping timeout: 260 seconds)
11:54:05Zagor(because it isn't really album art code. it is wps code.)
11:54:29Zagorand therefore does not belong in the playback library
11:55:35Zagorsomewhere in skin_engine?
11:55:39 Quit Kitr88 (Ping timeout: 276 seconds)
11:59:53 Join Kitar|st [0] (Kitarist@BSN-182-66-53.dial-up.dsl.siol.net)
12:00
12:00:25 Join user890104 [0] (Venci@venci-notebook-lan.ipv6.6bez10.info)
12:38:21 Join robin0800 [0] (~robin0800@cpc2-brig8-0-0-cust964.3-3.cable.virginmedia.com)
12:39:42[Saint]Zagor: seems appropriate to me...though I imagine JdGordon| is probably one to pester about this.
13:00
13:05:37[Saint]is " export ANDROID_NDK_PATH="$HOME/android-dev/android-ndk-r5b-linux-x86/" " correct syntax to add thos environment variable to my ~/.profile ?
13:05:55[Saint]it doesn't seem to be finding the toolchain :-/
13:08:49 Join evilnick [0] (18bcf602@rockbox/staff/evilnick)
13:19:41bluefoxxOkay, I have a Philips HDD6330 here that my friend has managed to fisk the headphone jack on, is there a takeapart on it someplace? Lacks the conventional screws
13:20:28 Join MethoS- [0] (~clemens@134.102.106.250)
13:22:09 Quit Unhelpful_ (Read error: Connection reset by peer)
13:22:49 Join Unhelpful [0] (~quassel@rockbox/developer/Unhelpful)
13:30:44 Quit Keripo (Quit: Leaving.)
13:40:03***Saving seen data "./dancer.seen"
13:57:09 Quit krazykit (Quit: baby bye bye bye)
14:00
14:05:41 Quit GeekShadow (Read error: Connection reset by peer)
14:06:06 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
14:08:03 Join dfkt|x [0] (~dfkt@chello062178002170.1.11.univie.teleweb.at)
14:08:04 Quit dfkt|x (Changing host)
14:08:04 Join dfkt|x [0] (~dfkt@unaffiliated/dfkt)
14:16:06 Join einhirn [0] (~Miranda@bsod.rz.tu-clausthal.de)
14:18:07sideralIs there a reason why Flyspray doesn't offer all (or at least all supported) platforms or targets for the Player Type field? I'm missing the Sansa AMSv2 platform or targets.
14:19:10sideralI think it would be best if that field captured the platform, so that all bugs related to a particular family of players could be queried
14:20:36Zagorsideral: added
14:21:18 Join Highlander [0] (~Highlande@mek33-4-82-236-45-205.fbx.proxad.net)
14:21:35sideralthat was quick, thanks Zagor!
14:25:29ZagorI'm here to help :)
14:26:11dfkt|xis there an explanation why the biggest screen resolution raaa build has a smaller file size than the other ones? - http://pastie.org/1578657
14:37:31sideralZagor: Looks like I can't search Flyspray by Player Type... Is it possible to add that field to the query mask?
14:38:16Zagorumm, I guess. "possible" is a wide word :)
14:38:38sideralZagor: do it now!
14:38:43sideralbetter? ;)
14:39:34ZagorI'd have to muck about in the php code, which I try to avoid
14:40:21sideralToo bad. I'd have wagered there's a list of fields to present in the search mask...
14:41:11Zagornot that I can find
14:43:32 Quit shai (Quit: Leaving)
14:44:02 Quit dfkt|x (Remote host closed the connection)
14:45:26dfkt(i'm still here, if anyone knows the answer to the above ;)
14:47:52n1sdfkt: maybe some resolution dependant files have not been adapted, like backdrops
14:48:17n1syou know, you could check which files are included in each build
14:49:05dfktn1s, thanks, i will check
14:50:51 Join TheLemonMan [0] (~lem0n@151.62.146.31)
14:51:29 Quit user890104 (Ping timeout: 272 seconds)
14:54:02sideralZagor: Should we file a bug with the Flyspray developers?
14:56:36n1ssideral: unfortunately flyspray development seems to have stagnated AFAICT
14:56:39ZagorI don't know if they are alive anymore. last release was nearly two years ago.
14:57:04n1slast commit was in November so that's a bad sing too
14:57:25sideralHmm... probably because Flyspray already is nearly perfect −− save for this one last feature ;)
14:57:45Zagorthere was a "is the project still alive" question on the mailing list in september, without any dev response...
14:57:58B4gderouch
14:59:04 Join MrChimp [0] (~Miranda@81.168.114.190)
14:59:26sideralSo, back to the php hacking plan...
15:00
15:00:39sideralRight now it's impossible to search or filter by any of the extended fields
15:01:57sideralAh no, Player Type is the only one :)
15:02:50Zagorplayer type was be called "operating system" originally
15:03:44Zagorin case you want to experiment with url construction
15:03:54Zagorvariable name "os", I think
15:04:06sideralI'm doing that right now :)
15:06:34MrChimphi everyone...I have an iPod Nano 4G. Am I right in saying installing rockbox will leave the ipod OS unaffected?
15:07:00TorneInstalling rockbox won't work, as we don't support the nano 4g
15:07:07Torneon models we do support, it leaves the ipod OS unaffected
15:08:21MrChimpAh, I mis-read. iPod 4th gen, not Nano 4th Gen. Ah well.
15:09:23 Quit robin0800 (Quit: Leaving)
15:10:08sideralZagor: Nope, doesn't seem to work. (Tried "os[]=34" and "operating_system[]=34" as URL components. The 34 is the ID of "AMSv2" in the New Task dialog.)
15:10:33 Part MrChimp ("I'm a happy Miranda IM user! Get it here: http://miranda-im.org")
15:16:55sideralZagor: Does Flyspray have a web API to export its DB as XML or JSON so that I could run my own queries on it?
15:21:35 Join leavittx [0] (~lev@nat172-248-205-109.tvoe.tv)
15:23:35 Join komputes [0] (~komputes@ubuntu/member/komputes)
15:26:20 Quit Highlander (Quit: Quitte)
15:34:39 Quit n1s (Quit: Lämnar)
15:35:17 Part LinusN
15:37:06 Quit factor (Ping timeout: 246 seconds)
15:37:19 Quit [Saint] (Quit: I'm only going to Heaven if it feels like Hell, I'm only going to Heaven if it tastes like caramel...)
15:40:04***Saving seen data "./dancer.seen"
15:41:32Zagorsideral: I don't know
15:42:02ZagorI suppose we could expose some of the mysql tables read-only
15:42:18 Join [Saint] [0] (S_a_i_n_t@203.184.0.13)
15:43:42 Join enthdegree [0] (~enthdegre@vpn2.waketech.edu)
15:48:14sideralZagor: Sounds like it's much more trouble that it's worth. Thanks for now!
15:48:36Zagorok
15:49:46 Quit B4gder (Remote host closed the connection)
15:51:19 Quit [Saint] (Ping timeout: 240 seconds)
15:56:47 Join [Saint] [0] (S_a_i_n_t@203.184.0.13)
16:00
16:04:15 Join user890104 [0] (Venci@venci-notebook-lan.ipv6.6bez10.info)
16:05:43 Quit Topy44 (Ping timeout: 240 seconds)
16:06:35 Join Topy44 [0] (~Topy44@89.204.137.146)
16:09:45 Quit user890104 (Read error: Connection reset by peer)
16:09:53 Join user890104 [0] (Venci@venci-notebook-lan.ipv6.6bez10.info)
16:12:53 Join user890104_ [0] (~Venci@6bez10.info)
16:13:05 Quit user890104 (Read error: Connection reset by peer)
16:15:49Zagorskin_display.c looks like a reasonable place to put it
16:16:48 Quit tintamarre (Remote host closed the connection)
16:17:38 Quit GeekShadow (Quit: The cake is a lie !)
16:30:52 Quit enthdegree (Quit: leaving)
16:46:14 Quit Judas_PhD (Quit: This is a quitting message)
16:48:42 Join Judas_PhD [0] (~kevin@misterfluffy.dsl.xmission.com)
16:49:47 Quit evilnick_B (Ping timeout: 245 seconds)
16:51:21 Quit Judas_PhD (Client Quit)
16:58:54 Part Zagor
17:00
17:03:59 Quit TheLemonMan (Quit: free(me))
17:08:14 Quit petur (Quit: *plop*)
17:18:39 Quit bluefoxx (Quit: Can we, should we, will we?)
17:19:25 Join bluefoxx [0] (fuzzylomba@S0106485b3917092d.vs.shawcable.net)
17:23:22 Join Horscht [0] (~Horscht@p5DD56DE5.dip.t-dialin.net)
17:23:22 Quit Horscht (Changing host)
17:23:22 Join Horscht [0] (~Horscht@xbmc/user/horscht)
17:37:16 Join Keripo [0] (~Keripo@SEAS315.wlan.seas.upenn.edu)
17:40:08***Saving seen data "./dancer.seen"
17:46:09 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
17:46:23 Join vnl [0] (~slayer@cpc5-king10-2-0-cust73.perr.cable.virginmedia.com)
17:46:57 Join liar [0] (~liar@clnet-p09-185.ikbnet.co.at)
17:56:32 Quit Keripo (Quit: Leaving.)
18:00
18:02:55 Quit powell14ski_ (Ping timeout: 276 seconds)
18:04:24 Join stoffel [0] (~quassel@p57B4CA6B.dip.t-dialin.net)
18:09:07 Join Izzy_ [0] (~Izzy@50.35.228.110)
18:12:32 Quit Izzy__ (Ping timeout: 240 seconds)
18:21:58 Join panni_ [0] (hannes@ip-178-203-73-7.unitymediagroup.de)
18:33:59 Join Jerom [0] (~jerome@79.132.59.245)
18:40:26 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
18:42:22 Quit komputes (Quit: I haven't slept for ten days, because that would be too long.)
18:48:51 Join toffe82 [0] (~chatzilla@maf.wirelesstcp.net)
18:54:00 Part miceh
18:56:00 Join robin0800 [0] (~robin0800@cpc2-brig8-0-0-cust964.3-3.cable.virginmedia.com)
19:00
19:02:04 Quit sideral (Ping timeout: 276 seconds)
19:06:40 Join lixxus [0] (~Mehdi@94-193-44-17.zone7.bethere.co.uk)
19:10:49 Join sweetandy [0] (~sweetandy@unaffiliated/sweetandy)
19:12:28 Quit robin0800 (Remote host closed the connection)
19:19:26 Join LambdaCalculus37 [0] (www-data@rockbox/staff/LambdaCalculus37)
19:20:01LambdaCalculus37TheSeven: Ping
19:21:27sweetandyHmm, my screen will only light up if (and only if) it is connected to my computer. Otherwise it's as if the battery is completely dead, even though it is fully charged. Huh.
19:22:03sweetandyI'm a genius. I hit it opposite to the way I dropped it and it works now
19:22:19TheSevenLambdaCalculus37: gniP
19:22:19 Part sweetandy ("Ex-Chat")
19:23:05 Join komputes [0] (~komputes@ubuntu/member/komputes)
19:24:28 Quit DerPapst (Quit: Leaving.)
19:25:07LambdaCalculus37TheSeven: Found out something interesting about the 6th gen nano. It is indeed running iOS.
19:25:59LambdaCalculus37Albeit a very, very scaled down version meant for only audio playback, picture display, and the clock, and adding outside apps disabled.
19:26:20TheSevenLambdaCalculus37: how do you know that it's ios?
19:26:40LambdaCalculus37TheSeven: Tore apart a firmware update.
19:27:34LambdaCalculus37Plus got into an interesting conversation with an iOS dev at an Apple Store about the subject.
19:29:43*LambdaCalculus37 has to go back to work
19:29:47 Quit LambdaCalculus37 (Quit: back to work)
19:36:38 Quit einhirn (Quit: Miranda IM! Smaller, Faster, Easier. http://miranda-im.org)
19:40:11***Saving seen data "./dancer.seen"
19:41:56 Quit lixxus (Quit: Leaving.)
19:42:10 Join saratoga [0] (9803c6dd@gateway/web/freenode/ip.152.3.198.221)
19:43:05 Join krazykit [0] (~krazykit@99-126-205-52.lightspeed.cicril.sbcglobal.net)
19:43:49 Quit komputes (Quit: I haven't slept for ten days, because that would be too long.)
19:44:36 Join lixxus [0] (~Mehdi@94-193-44-17.zone7.bethere.co.uk)
19:52:29dionoeahello
19:53:17dionoeakugel: in case you read the logs: shouldn't FS #11766 be closed ?
19:55:20 Part vnl
19:55:26 Quit bluefoxx (Quit: Can we, should we, will we?)
19:55:33 Join bluefoxx [0] (fuzzylomba@S0106485b3917092d.vs.shawcable.net)
19:59:52 Quit bluefoxx (Client Quit)
20:00
20:00:00 Join bluefoxx [0] (fuzzylomba@S0106485b3917092d.vs.shawcable.net)
20:04:49 Join sideral [0] (~sideral@rockbox/developer/sideral)
20:06:56 Quit user890104_ ()
20:10:10 Quit Bagder (Quit: connection reset by beer)
20:16:30 Join Jerom1 [0] (~jerome@79.132.59.245)
20:16:30 Quit Jerom (Read error: Connection reset by peer)
20:17:08 Join Stummi [0] (~Stummi@rockbox/developer/Stummi)
20:18:01 Join factor_ [0] (~factor@75.108.68.114)
20:25:20 Join Buschel [0] (~chatzilla@p54B66667.dip.t-dialin.net)
20:26:35 Quit bluefoxx (Quit: Can we, should we, will we?)
20:26:40 Join bluefoxx [0] (fuzzylomba@S0106485b3917092d.vs.shawcable.net)
20:35:29 Join {phoenix} [0] (~dirk@p57AA5B77.dip.t-dialin.net)
20:37:37 Quit leavittx (Ping timeout: 276 seconds)
20:44:07 Join bertrik [0] (~bertrik@ip117-49-211-87.adsl2.static.versatel.nl)
20:44:07 Quit bertrik (Changing host)
20:44:07 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
20:46:52 Join shai [0] (~Shai@l192-117-110-233.cable.actcom.net.il)
20:52:22 Quit dionoea (Ping timeout: 255 seconds)
20:53:09 Join dionoea [0] (~dionoea@yop.chewa.net)
20:53:32 Quit dionoea (Changing host)
20:53:32 Join dionoea [0] (~dionoea@videolan/developer/dionoea)
20:59:43 Quit Buschel (Quit: ChatZilla 0.9.86 [Firefox 3.6.13/20101203075014])
21:00
21:03:18 Quit factor_ (Read error: Connection reset by peer)
21:07:43 Quit evilnick (Ping timeout: 244 seconds)
21:09:03 Join factor_ [0] (~factor@75.108.68.114)
21:22:35 Join user890104 [0] (~Venci@6bez10.info)
21:27:58 Quit [Saint] (Disconnected by services)
21:28:00 Join S_a_i_n_t [0] (S_a_i_n_t@203.184.1.244)
21:30:17 Join TheLemonMan [0] (~lem0n@151.62.146.31)
21:33:03 Quit S_a_i_n_t (Ping timeout: 272 seconds)
21:36:19 Join [Saint] [0] (S_a_i_n_t@203.184.1.244)
21:40:14***Saving seen data "./dancer.seen"
21:43:04 Part lixxus
21:43:14 Join Bagder [0] (~daniel@1-1-5-26a.hud.sth.bostream.se)
21:43:14 Quit Bagder (Changing host)
21:43:14 Join Bagder [0] (~daniel@rockbox/developer/bagder)
21:43:51 Join kugel [0] (~kugel@e178058159.adsl.alicedsl.de)
21:43:53 Quit kugel (Changing host)
21:43:53 Join kugel [0] (~kugel@rockbox/developer/kugel)
21:46:29 Join leavittx [0] (~lev@89.221.199.187)
21:50:00 Join thomasjfox [0] (~thomasjfo@dslb-088-067-037-140.pools.arcor-ip.net)
21:53:31 Join fml [0] (~chatzilla@manz-590ee198.pool.mediaWays.net)
21:55:33fmlHello. I'm trying to investigate the error in disktidy plugin described in http://forums.rockbox.org/index.php/topic,27247.0.html. I start the H120 simulator, launch the plugin and get many errors "SDL_WaitEvent() error", and after that a Segmentation fault. What could be the reason? Where should I start to look?
21:55:44 Quit Stummi (Quit: Bye!)
21:56:32fmlThe error seems to have nothing to do with the particular folder name (##MUSIC#). It also occurs if I just try to clean e.g. Linux trash files.
21:59:18 Nick thomasjfox is now known as thomasjfox_ (~thomasjfo@dslb-088-067-037-140.pools.arcor-ip.net)
22:00
22:01:51 Quit thomasjfox_ (Quit: Konversation terminated!)
22:02:12 Join thomasjfox [0] (~thomasjfo@dslb-088-067-037-140.pools.arcor-ip.net)
22:02:45fmlI couldn't find any occurrences of "SDL_WaitEvent() error" in the source code
22:03:11 Quit thomasjfox (Remote host closed the connection)
22:03:22 Join thomasjfox [0] (~thomasjfo@dslb-088-067-037-140.pools.arcor-ip.net)
22:04:16 Quit mystica555 (Read error: Connection reset by peer)
22:04:17 Join Buschel [0] (~chatzilla@p54A3A5FA.dip.t-dialin.net)
22:04:41Buschelfml: firmware/target/hosted/sdl/button-sdl.c, line 283
22:05:04fmlBuschel: yes, sorry, now I've found that too
22:05:51kugeldionoea: hey, I tried changing the intent mechanism for the widgets (KeyEvent in the intent's extra data)
22:06:00kugelbut for some reason it doesn't work at all
22:06:22thomasjfoxBuschel: Thanks for the uninitialized vars fixes!
22:06:40thomasjfoxBuschel: Any comment about FS #11947? I'm about to commit the off-by-one-fix
22:06:41kugelthe keyevent is wrong when I read it in the service
22:06:56Buschelthomasjfox: no problem. it remined me of my good old PCLint times ;)
22:07:15 Quit stoffel (Remote host closed the connection)
22:07:31thomasjfoxBuschel: Did you try PCLint on rockbox? That's something I wanted to do -sometime-
22:07:44Buschelthomasjfox: no, I never tried
22:07:50kugelthomasjfox: it should go in. it seems like an obvious and needed fix for me
22:08:04thomasjfoxkugel: What puzzles me is the comment
22:08:11thomasjfoxkugel: Is state "Zero terminate the string"
22:08:16thomasjfoxIt just search for \r or \n
22:08:24thomasjfoxMaybe it should also search for \0?
22:09:52Buschelwill compilers always create code that first checks for "(i<max)"? I would expect that it might depend on the compiler
22:10:04Buschel<- FS #11947
22:10:31 Quit Xerion (Quit: )
22:10:40kugelthomasjfox: your patch doesnt change the logic (whatever the logic is), does it?
22:10:40thomasjfoxBuschel: Yes, I'm pretty sure about that. && and || will be execute in order.
22:11:09thomasjfoxkugel: I'll just fix the buffer access. Logic fixes might come in a separate commit
22:11:26 Quit fml (Quit: ChatZilla 0.9.86 [Firefox 3.6.13/20101206122310])
22:11:45kugelBuschel: the compiler doesn't (and must not) reorder the conditions
22:11:59Buschelthomasjfox: any details about your rew/ff obervations?
22:12:35 Quit thomasjfox (Changing host)
22:12:35 Join thomasjfox [0] (~thomasjfo@rockbox/developer/thomasjfox)
22:13:18Buschelif so the change should go in. it will not harm anything
22:13:29*scorche salutes thomasjfox
22:15:42thomasjfoxBuschel: That's what I'm about to investigate now. Got delayed by my neighbours car battery...
22:16:01 Join maxsin [0] (~maxsin@188-115-150-221.broadband.tenet.odessa.UA)
22:16:33*thomasjfox is now officially rockbox cloaked. thanks scorche ;)
22:21:46 Quit Buschel (Quit: ChatZilla 0.9.86 [Firefox 3.6.13/20101203075014])
22:21:54CIA-70New commit by thomasjfox (r29324): Fix off-by-one buffer read access in format_track_path(). Part of #11947 ...
22:23:26saratogathat sounds fairly serious
22:23:37saratogadid that off by 1 error cause any problems?
22:24:49thomasjfoxNot that I detected any. Though it depends on the memory location. If it reads past the memory allocated by rockbox, it will segfault.
22:25:49CIA-70r29324 build result: All green
22:28:24 Join {-phoenix-} [0] (~dirk@p57AA7389.dip.t-dialin.net)
22:29:51dionoeakugel: hum ... it's hard to guess what could be wrong without seeing the code :)
22:30:39kugeldionoea: http://pastie.org/1580309
22:32:33 Quit {phoenix} (Ping timeout: 276 seconds)
22:35:44 Quit GeekShadow (Ping timeout: 246 seconds)
22:36:36 Quit leavittx (Ping timeout: 250 seconds)
22:40:01CIA-70New commit by thomasjfox (r29325): Don't underflow the buffer in format_track_path() if it consists of whitespaces
22:42:03 Join Xerion [0] (~xerion@5419A4D7.cm-5-2c.dynamic.ziggo.nl)
22:42:30dionoeakugel: hum ... the code looks fine to me :/ (Why do you run putExtra() in getPendingIntent() instead of in the constructor?)
22:43:35kugeldionoea: a lame attempt to make it work :)
22:43:51CIA-70r29325 build result: All green
22:43:53dionoeaah ok. And is intent.keycode ok? (If you cast the intent in the service)
22:43:58 Join ej0rge [0] (~alhaz@alhaz.fttp.xmission.com)
22:44:24kugelI also tried without this wrapper class without success
22:44:46kugelno, the keycode is wrong (always stop button)
22:44:46ej0rgeDo i presume correctly that the patch in task 8894 is more or less rejected and abandoned since mid-2009?
22:45:52gevaertsej0rge: huh?
22:45:52kugelthomasjfox: about to commit the thread rework, what was the issue you found again?
22:46:00AlexPej0rge: Errr, it says accepted
22:46:13thomasjfoxkugel: SDL threads didn't compile anymore
22:46:29ej0rgewhoops, sorry, i fail at reading comprehension
22:46:32thomasjfoxkugel: Some variable was not defined
22:51:20dionoeakugel: and if you always use MEDIA_NEXT will it also send MEDIA_STOP? (did you try a clean build directory?)
22:51:47dionoea(i had weird build problems once ... not due to changes in the java code though but you never know)
22:51:51kugelit'll only send stop. yes I tried a clean build
22:53:44dionoeahum ... I don't have the slighest clue
22:55:26dionoeamaybe something due to the media button receiver? (I doubt it since the intent is targeted specifically at the service)
22:56:28dionoeaI'll give it a try tomorrow
22:56:51CIA-70New commit by thomasjfox (r29326): Make sure we don't read past the end of a C-string in format_track_path. Second part of FS #11947
23:00
23:00:46CIA-70r29326 build result: All green
23:05:25thomasjfoxIs this normal rockbox behavior: I queue up about 150 tracks in the playlist and press ff really fast. Suddenly the complete playlist changes to something else like a full album not even in that folder??
23:05:35thomasjfoxOr is there some kind of ff to next folder feature?
23:05:50 Quit esperegu (Read error: Connection reset by peer)
23:09:09kugelthomasjfox: that's a feature yes
23:09:33kugelshort ff followed by long ff within 1s or so will advance to the next folder
23:10:18thomasjfoxkugel: Thanks. Now I can trigger it ;) Rockbox is really a feature monster...
23:12:10 Join kkurbjun [0] (~kkurbjun@c-24-9-122-9.hsd1.co.comcast.net)
23:12:33 Quit kkurbjun (Changing host)
23:12:33 Join kkurbjun [0] (~kkurbjun@rockbox/developer/kkurbjun)
23:13:00kkurbjunDoes anyone know if any of the bootloaders work with software USB?
23:13:17kkurbjunI am trying to get it working on the M:Robe, but it doesn't seem to be active
23:13:18 Join casainho [0] (~chatzilla@bl20-146-26.dsl.telepac.pt)
23:13:27kkurbjunworks fine with the main build
23:13:32gevaertskkurbjun: gigabeat S, and possibly some of the PP ones these days
23:14:06kkurbjungevaerts - do you know what needs to be done? I tried emulating what the S was doing, but it still didn't work
23:14:26kkurbjunalso, I was looking at the GSOC - I saw the idea for relocatable plugins
23:14:34gevaertskkurbjun: not really. jhMikeS did that
23:14:43kkurbjunI think you could do it without elf if you use -fpic
23:15:07kkurbjunyou would just need to follow the abi for position independent code for each target
23:15:24gevaertskkurbjun: you need relocation handling, whether that is elf or something else. Only -fpic is not sufficient
23:15:37thomasjfoxBuschel (logs): With your recent initialization fixes, the rew/ff issue seems gone in valgrind
23:15:51kkurbjunCouldn't you just load a binary blob to an address and setup the offset table?
23:16:28gevaertsIt's been a while since I looked at it...
23:16:40*gevaerts waits for Torne to comment. he knows better :)
23:17:23kkurbjunwell no worries, it was just a thought - I believe that u-boot does it without a full elf loader
23:17:32kkurbjunbut that might just be relocation
23:18:10gevaertsOh, you definitely don't need full elf
23:18:40kkurbjunwell, yes, I should have said with a binary blob
23:19:03kkurbjunin anycase I think it's a cool idea for a project
23:19:52gevaertsYes. I think we really need such a thing
23:23:36*kugel ponders :)
23:28:10thomasjfoxkugel: Do you sometimes get stuck keyboard keys on android? Like you pressed the up key only once and rockbox keeps scrolling?
23:28:38kugelkeyboard?
23:28:50Tornekkurbjun: -fpic is basically useless :)
23:28:58kkurbjunwhy is that?
23:29:09thomasjfoxkugel: Okay, so your phone doesn't have that :)
23:29:18Tornebecause you still need ld.so
23:29:55Torneand because you can't output as binary, so you end up needing, er, ELF :)
23:30:04kkurbjunhmm
23:30:10kkurbjunI've seen it done on microblaze
23:30:20kkurbjunbut I havn't messed around with it on arm
23:30:39kkurbjunwithout a special loader - the abi on that only used one register to calculate the offsets though
23:30:47TorneYes, -fpic doesn't work like that
23:31:05kkurbjunyou just had to set r12 to the load address and the program was compiled with offsets that are calculated based on that register
23:31:15Torneyes. gcc doesn't do that. :)
23:31:21Torneexcept for MIPS
23:31:29kkurbjunthat was gcc on microblaze
23:31:47kkurbjunis there documentation on fpic or fPIC?
23:31:54Tornemicroblaze is MIPS, no?
23:31:56kkurbjunI don't know if fPIC is supported on ARM though
23:32:06Torne-fpic and -fPIC are almost identical
23:32:16kkurbjunmicroblaze is the soft processor Xilinx makes
23:32:20Torneyes, but it's a MIPS
23:32:23Tornebecaue MIPS is free
23:32:33Torneunless i'm misremembering
23:32:44kkurbjunits not mips to my knowledge
23:32:50Torneanyway, the register based pic data implementation is implemented in the arch backend
23:32:54TorneARM doesn't have one
23:32:58Tornelike most other arches :)
23:33:44TornePIC otherwise uses a PLT and a function to get the base address over and over.
23:34:18Torneit's purely a mechanism to reduce the number of unique pages you need on a system, not to avoid the need for a full dynamic loader
23:34:43 Quit TheLemonMan (Quit: free(me))
23:35:03kkurbjundo you have any documentation you could point to on the pic implementation for ARM?
23:35:11Tornethere is no pic implementatoin for ARM
23:35:22Tornethat's the whole point
23:35:32Bagdermicroblaze is not mips, no
23:35:33Tornethe pic implementation is for linux, not any arch
23:35:43Torneand thus assumes ld-linux.so
23:36:17kkurbjunI think they are using it in U-boot now
23:36:23kkurbjunon arm
23:36:46Tornefor what?
23:37:00kkurbjunso that uboot doesn't need a static address when compiling
23:37:09kkurbjunso that it can run with one binary on a number of boards
23:37:21kkurbjunwith different memory sizes (they load at the end of memory)
23:39:41TorneThat's not position independant at all
23:39:44TorneIt just relocates itself
23:39:49Tornehttp://git.denx.de/cgi-bin/gitweb.cgi?p=u-boot.git;a=blob;f=doc/README.arm-relocation;h=c0957c210ceca37efd1c9c236768df57b8456c6e;hb=HEAD
23:40:16***Saving seen data "./dancer.seen"
23:40:38Tornethe startup code is written to be position independant by hand, either by being in asm or by being extremely careful in C, and then it applies a relocation table generated by the linker to itself
23:40:47Tornewhich is exactly what we're proposing to do for rockbox
23:41:50Tornewriting PIC code in ARM is easy until you need to refer to something outside the .text section ;)
23:41:53kkurbjungotcha, but the u-boot image is a binary blob
23:42:02kkurbjunso I am missing something on the need for elf support
23:42:25Torneyes, and the linker script throws the ELF relocation records onto the end of the blob
23:42:32Torneso the relocation code needs to be able to parse ELF relocs
23:42:54Torneand it has to know where the relocatoins are, which it does by having the linker script just emit symbols for them
23:42:59Tornewhich works because u-boot is a single binary
23:43:17TorneFor us to load plugins we would at minimum need some kind of header for the location of that info
23:43:47Tornei suspect it might be a better idea to use an existing lightweight binary format rather than invent our own wrapper around a blob with ELF relocs appended
23:43:57Tornesince we're not linking everything into one binary.
23:44:50kkurbjunhmm, in terms of the callbacks to the main firmware you mean?
23:44:52Tornethe u-boot relocatoin loop depends on the linker only outputting type 23 and type 2 relocations since it can't parse anything else, and that's kinda fragile
23:45:19Tornesince future ARM instruction set changes or whatever might need different reloc types and then the compiler will generate something that its loader can' thandle
23:45:51TorneThe calls back to the rockbox binary aren't really relevant
23:46:07TorneTo begin with it would be easiest to just keep doing them exactly as we do now
23:46:07CIA-70New commit by kugel (r29327): Implement cooperative threads on hosted platforms using C code. ...
23:46:12Tornevia the rb structure.
23:46:15kkurbjunwhat else is being linked seperately?
23:46:24Tornehm? the plugins ar elinked seperately to the main binary
23:46:43Torneso the relocating loader won't have access to symbols in the plugins
23:46:58Tornebecause that's in a different file :)
23:48:37kugelthomasjfox: you can now adapt maemo
23:48:58thomasjfoxkugel: Yeah!
23:49:18thomasjfoxkugel: Currently tracking down the stuck buttons in the SDL app build...
23:49:25kkurbjun.. so plugins only communicate/use code through the rb structure to the main binary - that part I understand - but I don't get how the plugins are significantly different from a u-boot binary
23:49:41Tornebecause the u-boot binary contains both the relocatoin code and the code that needs to be relocated
23:49:41kkurbjunI mean besides the use of that structure
23:49:48kkurbjunahh
23:49:49Torneso it already knows where the relocations are, because it's linked to them
23:49:57kkurbjunI see so you need the header to find them
23:50:00kkurbjunmakes sense
23:50:20Torneyes. and rather than keep inventing more ad hoc data structures for doing htat, i think it might be more sensible to use FLT or some other known binary format
23:50:35Tornewhich is lighter and simpler than ELF but already has existing tools to create/manipulate it
23:50:38Tornethat's a minor detail, though
23:50:48kkurbjungotcha - I'm not familiar with FLT - I was just thinking that ELF might be overkill
23:50:51Torneanyway, the main point is yes, we want to do it in broadly the way that u-boot does it.
23:51:01Tornebut u-boot doesn't use -fpic or anything similar
23:51:03Tornebecause that sucks
23:51:04Torne:)
23:51:21Tornethat would generate crazy relocs that it'd never be able to apply without understanding way more of ELF
23:51:25kkurbjunso they use pie instead correct?
23:51:38CIA-70r29327 build result: 987 errors, 121 warnings (kugel committed)
23:51:56Torneyes, but we needn't.
23:51:58Tornei don't think.
23:52:13Torneplugins don't need to have a final link done on them, they're not really executables
23:52:22Tornethe FLT way doesn't require any such thing
23:52:37Torneyou just build the code and link it with -r so the relocations are retained and symbols are allowed to just be undefined
23:52:47Tornethen elf2flt converts it to FLT from that.
23:52:51thomasjfoxkugel: new highscore?
23:53:04kugelno :)
23:53:06kkurbjuninteresting
23:53:25Torneonce regular relocation works we could also throw out the rb structure and just allow plugins to link directly to functions in core
23:53:36gevaertsthomasjfox: that's not anywhere near a highscore
23:53:44Tornethose would then be undefined at link time, and the flt image would tell you where to look them up
23:54:01Tornethat's fiddlier, though; ABI compatibility and the like
23:54:03kkurbjunhow "heavy" is flt?
23:54:07Torneand it's a second step, unrelated
23:54:09Torneminiscule
23:54:18kkurbjunI mean would it have a significant impact on the binary size of rockbox?
23:54:31Tornenope, the relocation loop is probably eight lines or so :)
23:54:31 Quit [Saint] (Disconnected by services)
23:54:32 Join S_a_i_n_t [0] (S_a_i_n_t@203.184.1.146)
23:54:44kkurbjunoh nice
23:54:45kkurbjunhaha
23:54:54Torneit literally is just a header, text section, data section, and a table of relocations which are all the same format/type
23:55:16kugelweeh, that warning in thread.c is suspicious
23:55:18kkurbjuncool, sounds like a nice solution then
23:55:28Torneyou just need to look at each offset, deduce from its locatoin whether it's a data or text reloc, and then patch the target site by the base address of the relevant section
23:55:30gevaertsThe actual work is going to be in the build system I think
23:55:41Tornethe problem is that it doesn't support arbitrary sections so plugins that use iram may be an issue
23:55:47gevaertsAnd of course making use of the new possibilities
23:55:56TorneIf we only allow one thing at a time to be using iram then that's okay
23:56:04Tornesince you can just link those as fixed addresses
23:56:32Tornebut yeah, elf2flt is a readily available tool and it's used/maintained by uClinux so it's an active thing
23:56:37Tornethough sadly it's not super well documented ;)
23:57:06Torneso yes, the work would be working out what to change in the build system and link scripts
23:57:13Torneuntil it produces the right binaries
23:57:19Tornethe actual relocating loader shouldn't be super hard :)
23:57:39Torneand if it turns out FLT isn't really suitable you could do something else.
23:57:47Tornei just think inventing our own format should be a last resort
23:58:12Torne(if nothing else because it means someone *might* have to write a tool that can read ELF binaries and that's Not Fun At All) :)
23:58:38 Quit bluebrother (Disconnected by services)
23:58:40 Join bluebroth3r [0] (~dom@rockbox/developer/bluebrother)

Previous day | Next day