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

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

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

#rockbox log for 2012-06-11

00:07:49 Quit Guest2615 (Read error: Connection reset by peer)
00:09:01 Quit GermanMushroom (Quit: Ik ga weg)
00:09:19 Join GermanMushroom [0] (
00:09:22 Quit GermanMushroom (Read error: Connection reset by peer)
00:11:52 Quit bertrik (Quit: And That, My Liege, Is How We Know the Earth to Be Banana Shaped)
00:13:01 Quit lebellium (Quit: ChatZilla [Firefox 13.0/20120531155942])
00:22:50 Quit enthdegree (Read error: Connection reset by peer)
00:23:49 Join enthdegree [0] (~enthdegre@wikimedia/enthdegree)
00:25:10 Quit ender` (Quit: The better I get to know men, the more I find myself loving dogs. -- Charles de Gaulle)
00:31:11 Join WilliamC [0] (~William@
00:31:50WilliamCI'm having a problem my flash music player has been filled up with junk, how do I purge it, it won't boot into default boot it says there isn't enough free space
00:38:48 Quit [Saint] (Ping timeout: 244 seconds)
00:41:56 Join [Saint] [0] (~Saint]@unaffiliated/saint/x-8516940)
00:43:35WilliamCMy Sansa Clip Zip has been filled up with junk and it won't boot into default firmware
00:51:06 Quit pamaury (Remote host closed the connection)
00:51:21WilliamCI want to throw it into the wall
00:54:57RFroehningWilliamC, whats the default firmware for you? rockbox or sansa?
00:55:21WilliamCRockbox by default
00:55:44RFroehninghave you tried to boot the sansa firmware yet?
00:56:07WilliamCIt says Not enough space for MusicDB, Please free ..."
00:56:14WilliamCBut then it shuts off
00:56:33RFroehningand rockbox says the same?
00:56:50RFroehningi mean it cant boot because of that?
00:57:02WilliamCNo, it says no firmware
00:57:23RFroehningoh ... now i understand youi better
00:59:07RFroehningconnecting it directly to usb in sansa mode works?
01:00:12RFroehningeven if trying to press home or someting while connecting?
01:00:30RFroehningto make sure the sansa firmware boots
01:00:38WilliamCIt boot sbut shuts off
01:01:09RFroehningwhat boots? the sansa firmware?
01:01:54RFroehningtoo bad.. that would have been my way to fix the mess...
01:02:04WilliamCShould I take it apart?
01:02:39WilliamCIt gets recognized in Rockbox mode but I cannot access the internet memory
01:04:14RFroehningwhat do you mean by it gets recognized? its in usb mode and you see a drive?
01:05:33RFroehningcan you read anything from that drive?
01:05:53WilliamCIt says that there is nothing in the drive
01:06:29RFroehningcan you create a file there?
01:06:48RFroehningbut nothing... hmmm
01:07:40 Quit thegeek_ (Read error: Operation timed out)
01:07:45RFroehningso rockbox firmware seems to boot?
01:08:00RFroehningbut if you try to boot into rockbox it says no firmware?
01:08:16RFroehningi am lost :-D
01:09:02WilliamCWhat should I do?
01:09:09WilliamCShould I take it apart?
01:09:26RFroehningyou have made sure you connect it while turned off and press a key to force sansa mode?
01:09:33RFroehningand it fails?
01:10:04RFroehningif you took it apart... what do you want to achieve? you would need to get access to the inernal memory
01:11:16WilliamCWhat should I do?
01:11:27WilliamCIt doesn't seem like installing Rockbox should do this, but it did
01:15:03 Quit perrikwp (Read error: Connection reset by peer)
01:15:10RFroehningmaybe the filesystem got corrupted...
01:15:34RFroehningi cant help you here..... but if the firmware starts and says you need to make space... the usb mode should work too
01:15:52***Saving seen data "./dancer.seen"
01:19:00WilliamCBut it doesn't
01:22:47WilliamCRemind me not to use Rockbox again
01:29:10WilliamCI got it to work
01:29:15WilliamCI switched USB ports
01:34:13RFroehningyou want to remove some stuff and reinstall rockbox... or alternatively: put a firmware update from sansa on the root directory and reboot
01:34:28RFroehningso the new firmware gets installed and removes the rockbox boot loader
01:36:45WilliamCI like Rockbox but it's unstable and it causes me headache
01:41:55 Quit WilliamC (Read error: Connection reset by peer)
01:43:05 Join fastreload [0] (~fastreloa@
01:46:37 Quit fastreload (Read error: Connection reset by peer)
01:46:46 Join perrikwp [0] (
01:50:04 Join user485763 [0] (~aoeu@
02:09:50 Quit [Saint] (Remote host closed the connection)
02:15:17 Quit nomada (Ping timeout: 245 seconds)
02:16:50 Quit amithkk (Max SendQ exceeded)
02:17:29 Join amithkk [0] (u4289@2buntu/writers/amithkk)
02:18:18 Join nomada [0] (
02:21:37 Join thegeek [0] (
02:34:37 Join WilliamC [0] (~William@
02:34:44WilliamCNow I have another yet similar issue
02:39:55 Quit rasher (Ping timeout: 248 seconds)
02:40:59 Quit prof_wolfff (Ping timeout: 260 seconds)
02:47:24 Join rasher [0] (~rasher@rockbox/developer/rasher)
02:52:22 Quit enthdegree (Quit: HydraIRC -> <- Wibbly Wobbly IRC)
02:54:58WilliamCI should just install a jtag
02:55:53 Join BHSPitMonkey [0] (~stephen@unaffiliated/bhspitmonkey)
03:11:52WilliamCHow would I go about installing jtag?
03:15:53***Saving seen data "./dancer.seen"
03:16:37 Quit Syconaut (Read error: Connection reset by peer)
03:16:42WilliamCWhy does using Rockbox do this?
03:17:05 Join Mir [0] (
03:21:05WilliamCAll I did was update the firmware.:(
03:25:47 Quit thegeek (Ping timeout: 244 seconds)
03:31:15Mirwow >>
03:32:24WilliamCI feel like throwing the thing out the window
03:33:15 Join thegeek [0] (
03:34:32WilliamCIt boots into original firmware, it says that the FAT is corrupted, connect to computer, I connect to computer, it gets recognized but the internal memory is inaccessible.
03:35:26Miri had a similar problem with a clip i flashed
03:35:39Mirthen agian it turns out the clip got fried
03:35:56WilliamCWhat happened?
03:39:14WilliamCLife sucks
03:40:00Mirt died after installing rockbox
03:40:07Mirjust would not power on
03:40:09Mirat all
03:40:13WilliamCIt powers on
03:40:23Mirsent it to a friend who tried debricking it
03:40:30Mirmine didnt :)
03:40:39WilliamCRight now I hate Rockbox
03:46:08WilliamCNo one is ever on this channel
03:50:05WilliamCI just got it 6 months ago
03:57:03 Join factor [0] (
03:57:44 Quit XavierGr ()
04:25:33 Quit factor (Read error: Connection reset by peer)
04:36:56 Join enthdegree [0] (~enthdegre@wikimedia/enthdegree)
04:43:20 Join factor [0] (
04:43:27 Join TheSphinX^ [0] (
04:44:58 Quit [7] (Disconnected by services)
04:45:05 Join TheSeven [0] (~quassel@rockbox/developer/TheSeven)
04:46:49 Quit TheSphinX_ (Ping timeout: 245 seconds)
04:48:08serhi. it seems internal drive of my sansa fuze is broken, when i connect usb, it works unpredictable, sometimes freezes, etc. what should i do?
04:48:54seris any type of low-level format possible and re-install of rockbox?
04:56:32 Join amiconn_ [0] (amiconn@rockbox/developer/amiconn)
04:56:32 Quit amiconn (Disconnected by services)
04:56:35 Nick amiconn_ is now known as amiconn (amiconn@rockbox/developer/amiconn)
04:57:31 Quit pixelma (Disconnected by services)
04:57:31 Join pixelma_ [0] (pixelma@rockbox/staff/pixelma)
04:57:34 Nick pixelma_ is now known as pixelma (pixelma@rockbox/staff/pixelma)
05:15:56***Saving seen data "./dancer.seen"
05:21:15funmanser: fuze v2 ?
05:22:11serfunman: yes, but it seems i am on the way to recover, i formated internal drive with mkfs.vfat and now i am installing rockbox from scratch
05:22:27funmanusb is a bit flaky on amsv2
05:22:55serit ssems i will use sdcard adapter to fill it up next time, thanks a lot
05:28:11funmanthat, or use the OF
05:28:32serwhat is OF?
05:40:14funmanOriginal Firmware
05:40:47serOK, thanks for decypting :)
05:52:58 Quit enthdegree (Quit: HydraIRC -> <- Go on, try it!)
06:15:14WilliamCI just don't think that Rockbox is meant to run with AMSv2
06:18:08 Quit CaptainKewl (Ping timeout: 240 seconds)
06:26:23Mode"#rockbox +o funman" by ChanServ (ChanServ@services.)
06:26:45Kick(#rockbox WilliamC :go spread your hate and lies somewhere else) by funman!~fun@rockbox/developer/funman
06:26:45 Join WilliamC [0] (~William@
06:27:11 Join [Saint] [0] (~Saint]@
06:27:11 Quit [Saint] (Changing host)
06:27:11 Join [Saint] [0] (~Saint]@unaffiliated/saint/x-8516940)
06:30:49 Quit Rower85 (Quit: Hmmm...)
06:33:34 Quit anewuser (Read error: Connection reset by peer)
06:46:31 Quit user485763 (Ping timeout: 246 seconds)
06:47:53 Quit WilliamC (Read error: Connection reset by peer)
06:57:11 Quit amee2k (Ping timeout: 252 seconds)
06:58:07 Join bluebrother [0] (~dom@rockbox/developer/bluebrother)
06:58:14 Join amee2k [0] (
06:59:44 Quit fs-bluebot (Ping timeout: 245 seconds)
06:59:52 Quit bluebrother^ (Ping timeout: 265 seconds)
07:00:51 Join fs-bluebot [0] (
07:01:05 Quit KiwiCam (Quit: Leaving)
07:12:56*[Saint] wonders how on eart to debug this touchscreen bug
07:13:04[Saint]*earth, too
07:14:06[Saint]I appear to have discovered a recipe to make the main menu totally stop responding to scroll/tap events.
07:15:58***Saving seen data "./dancer.seen"
07:18:01 Join WilliamC [0] (~William@
07:48:29 Quit CIA-4 (Ping timeout: 245 seconds)
07:50:14 Join CIA-4 [0] (
08:03:11 Quit CIA-4 (Ping timeout: 265 seconds)
08:04:22 Join CIA-47 [0] (
08:18:51 Join harch [0] (1f196a41@gateway/web/freenode/ip.
08:19:11harchhello to you all
08:22:14 Quit BHSPitMonkey (Remote host closed the connection)
08:23:21 Quit harch (Ping timeout: 245 seconds)
08:46:07RFroehningJdGordon, ping?
08:51:08 Join ender` [0] (
08:57:26 Join wodz [0] (
09:00:36 Join Zagor [242] (~bjst@rockbox/developer/Zagor)
09:02:44wodzgevaerts: Calling me to fullfill DevCon MoM I guess you are referring to 143d2b5baca7ae7fbc9dc42935a4c7813f784fe8 and c029c77796e88ee991af09ffe094309e4df528e6 :-)
09:04:18 Join bertrik [0] (
09:04:19 Quit bertrik (Changing host)
09:04:19 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
09:05:02 Join ks [0] (
09:10:24 Join LinusN [0] (
09:13:31 Quit kaner (Quit: leaving)
09:16:01***Saving seen data "./dancer.seen"
09:34:07 Nick untrack|ZzZpoke is now known as untrack (
09:36:04 Quit bertrik (Read error: Operation timed out)
09:45:46[Saint]kugel: any chance you can reproduce this weird touchscreen bug: "enter WPS, use the "quick action panel" (filebrowser/database, quickscreen, context menu) to get to the filebrowser or database, use HW Back (or titlebar, if you're using my theme) to get back to the main menu, lists should now have stopped responding to scroll/tap events"?
09:46:50[Saint]Stopping playback returns the lists to normal (touch events work) behaviour.
09:49:08 Join TheLemonMan [0] (
09:49:46gevaertswodz: apparently ;)
09:52:42 Quit wodz (Quit: Leaving)
10:00:50*gevaerts wonders if anyone has ever tried to get RaaAoA working on MIPS
10:04:50 Quit pjm0616 (Remote host closed the connection)
10:08:49[Saint]Hmmmmm....this touchscreen bug is annoying me.
10:09:14[Saint]I can't for the life of me see what is causing it. I feel someone with a clue needs to have a look.
10:11:01 Quit RFroehning (Ping timeout: 248 seconds)
10:12:44[Saint]Can someone else try to recreate this for me? I'd like to know someone else can do it before I add it as a bug officially on flyspray. My Android devices have a habit of showing some rather non-normal behaviour when compared to those of others.
10:13:08[Saint]"enter WPS, use the "quick action panel" (filebrowser/database, quickscreen, context menu) to get to the filebrowser or database, use HW Back (or titlebar, if you're using my theme) to get back to the main menu, lists should now have stopped responding to scroll/tap events"
10:14:11[Saint]I'm not 100% sure its possible to recreate if you don't have a hardware back button and aren't using my theme, unless you're in grid-mode.
10:15:14 Join evilnick| [0] (~evilnick@rockbox/staff/evilnick)
10:15:19 Quit evilnick (Read error: Connection reset by peer)
10:19:53 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
10:23:37 Join RFroehning [0] (
10:54:34 Quit kevku (Quit: KVIrc 4.0.4 Insomnia
10:54:43 Join kevku [0] (
11:12:39 Join einhirn [0] (
11:16:04***Saving seen data "./dancer.seen"
11:29:18 Join MethoS- [0] (~clemens@
11:34:27 Quit TheLemonMan (Quit: WeeChat 0.3.8)
11:35:21 Join TheLemonMan [0] (
12:09:41 Join swilde [0] (
12:12:47 Join Topy [0] (
12:16:44 Quit T44 (Ping timeout: 260 seconds)
12:38:38 Quit amithkk (Ping timeout: 244 seconds)
12:42:39 Quit evilnick| (Read error: Connection reset by peer)
12:55:08 Join XavierGr [0] (~xavier@rockbox/staff/XavierGr)
13:10:22 Join TomColler [0] (
13:10:31 Part TomColler
13:16:06***Saving seen data "./dancer.seen"
13:55:23 Join anewuser [0] (~anewuser@
13:55:23 Quit anewuser (Changing host)
13:55:23 Join anewuser [0] (~anewuser@unaffiliated/anewuser)
13:59:07 Quit Bagder (Ping timeout: 244 seconds)
14:01:00 Join Bagder [0] (
14:01:01 Quit Bagder (Changing host)
14:01:01 Join Bagder [241] (~daniel@rockbox/developer/bagder)
14:01:04 Quit swilde (Read error: Connection reset by peer)
14:02:13 Quit sinthetek (Ping timeout: 244 seconds)
14:03:20 Join sinthetek [0] (
14:03:20 Quit sinthetek (Changing host)
14:03:20 Join sinthetek [0] (~sinthetek@unaffiliated/sinthetek)
14:04:59 Join dfkt [0] (dfkt@unaffiliated/dfkt)
14:12:11 Quit [Saint] (Quit: Quit)
14:12:33 Join [Saint] [0] (~Saint]@
14:12:34 Quit [Saint] (Changing host)
14:12:34 Join [Saint] [0] (~Saint]@unaffiliated/saint/x-8516940)
14:12:53 Quit [Saint] (Read error: Connection reset by peer)
14:13:18 Join [Saint] [0] (~Saint]@
14:13:19 Quit [Saint] (Changing host)
14:13:19 Join [Saint] [0] (~Saint]@unaffiliated/saint/x-8516940)
14:38:40 Quit WilliamC (Read error: Connection reset by peer)
14:39:48kugelgevaerts: on which mips-based android device?
14:40:01gevaertsainol novo 7 paladin
14:46:03kugelgevaerts: I don't own that one, so no :)
14:46:22gevaertskugel: I have one
14:50:28kugelgevaerts: so why are you asking if someone else did some work :)
14:50:41gevaertsBecause it would be a good start1
14:50:44kugelyou should already have done it :p
14:52:58 Join WilliamC [0] (~William@
15:13:05 Join user485763 [0] (~aoeu@
15:15:08 Quit MethoS- (Ping timeout: 248 seconds)
15:16:08***Saving seen data "./dancer.seen"
15:18:35 Join Rower85 [0] (
15:33:59 Part Zoiah
15:34:12 Quit factor (Ping timeout: 245 seconds)
15:52:44 Join factor [0] (
16:01:15 Quit Zagor (Quit: Clint excited)
16:07:46 Quit factor (Read error: Connection reset by peer)
16:25:30 Join factor [0] (
16:25:42 Quit perrikwp (Ping timeout: 245 seconds)
16:30:57 Join WalkGood [0] (~4@unaffiliated/walkgood)
16:43:23 Join [Saint_] [0] (~Saint]@unaffiliated/saint/x-8516940)
16:43:28 Quit [Saint_] (Remote host closed the connection)
17:06:46 Join Keripo [0] (
17:08:10 Quit Rower85 (Quit: Hmmm...)
17:13:38 Quit guymann (Ping timeout: 246 seconds)
17:15:40 Join guymann [0] (
17:16:10***Saving seen data "./dancer.seen"
17:16:38 Join Rower85 [0] (
17:17:40 Quit user485763 (Ping timeout: 248 seconds)
17:24:35 Quit einhirn (Quit: Miranda IM! Smaller, Faster, Easier.
17:57:01 Join lebellium_gs2 [0] (
18:03:24 Join domonoky [0] (~Domonoky@rockbox/developer/domonoky)
18:17:05 Join prof_wolfff [0] (
18:17:10 Quit lebellium_gs2 (Quit: Bye)
18:23:16 Quit pamaury (Ping timeout: 246 seconds)
18:23:20 Quit factor (Quit: Leaving)
18:23:52 Join evilnick [0] (~evilnick@rockbox/staff/evilnick)
18:25:35 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
18:34:50 Quit TheLemonMan (Quit: WeeChat 0.3.8)
18:40:15 Join amithkk [0] (u4289@2buntu/writers/amithkk)
18:43:43 Join Strife89 [0] (~Strife89@
18:45:00bertrikhm, odd, could be a bug
18:49:16gevaertsbertrik: maybe. It could also be that cards that big with a smaller cluster size have another field that's technically out of spec, in which case we might reject them correctly
18:50:11gevaertsNumbers need to be worked out
18:52:38WalkGoodi wouldn't have a clue what it is but my version 3.10 runs fine with a 32gb microsd
18:52:57WalkGoodbut i don't have the clip+, mines the zip
18:54:24gevaertsWalkGood: could you go to the debug menu, pick "disk info", press right (I think...) to get to the sd card, scroll down, and look at the cluster size?
18:54:50gevaerts32GB with 32K clusters should be fine...
18:55:06gevaerts(my gigabeat has 32K clusters on 60GB)
18:55:55gevaertsbut maybe 'default' means 16K for the people who have issues?
18:57:16WalkGoodi don't see the clusters there
18:58:14gevaertsHm, maybe only ATA targets show that
18:59:02WalkGoodhas disk info but no cluster size there
19:00:07*gevaerts looks at the code
19:00:14gevaertsRight, that's not shown for SD
19:00:35gevaertsThat's a bug I'd say, even though it's just debug info
19:01:49evilnickIf a theme is using the same background image as cabbiev2, is it okay to package the exact same cabbiev2.bmp image in the theme's zip?
19:02:03WalkGoodi formatted mine w/sdassociations formatter, i'm assuming it's cluster size is 64k but don't know for sure
19:02:27gevaertsevilnick: I'd say no, but I can't speak for everyone
19:02:47TorneWalkGood: unlikely; default allocation size is pretty universallt 4kb
19:02:57gevaertsTorne: on 32GB FAT32?
19:03:22Torneer, okay, not on 32GB
19:03:26Tornebut then it'll just pick 8kb
19:03:30Torne64kb is not techncially even standard
19:03:35Torneold FAT implementations choke on it
19:03:39Torne32KB was the DOS limit
19:03:57WalkGoodonly said that because that post said it fixed it, but mine always worked from the start
19:04:16Tornea larger allocation size is generally a perfectly reasonable thing to do on a card that's just full of large files like mp3s :p
19:04:17WalkGoodlike i said, don't know for sure
19:04:19Torneand will indeed make stuff faster
19:05:02TornePossibly the poster there had filesystem errors
19:05:09Tornewhich will have been fixed by formatting with a different cluster size :P
19:06:01WalkGoodsince mine always works fine from the sd formatter, no need to ever check for me
19:07:00WalkGoodi did two different cards like that and both worked fine, i've had vertually no truble other than in very early days of v1 but no sd card back then
19:07:51WalkGoodi was a little surprised to read that, that's the only reason i posted
19:08:28 Quit guymann (Ping timeout: 245 seconds)
19:10:30 Join guymann [0] (
19:14:58bertrikgevaerts, want me to look at adding the cluster size to the debug info?
19:16:14***Saving seen data "./dancer.seen"
19:18:54bertrikcurrently, the disk info debug menu shows purely sd-related fields, and the partition info debug menu shows partition stuff, I don't think we have a fat info debug menu, but we can try to add it to disk info
19:20:39 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
19:25:10 Quit pamaury (Remote host closed the connection)
19:25:16bertrikhow about this: ?
19:25:48bertrikNot completely sure if cardnum correctly maps to the fat_get_cluster_size() argument, but it works for me
19:26:49gevaertsbertrik: I'd say we should get the FAT info out of the disk screen, and have a separate filesystem debug screen
19:28:05bertrikok, I guess there is other interesting stuff to show about a file system
19:28:35gevaertsFor one thing there might be more than one filesystem on a disk
19:30:36 Join pamaury [0] (
19:30:36 Quit pamaury (Changing host)
19:30:36 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
19:37:56saratogadidn't google open source some text to speech engine recently? i'm not seeing one as i search
19:42:32 Join TheLemonMan [0] (
19:51:28 Join n1s [0] (
19:51:28 Quit n1s (Changing host)
19:51:28 Join n1s [0] (~n1s@rockbox/developer/n1s)
19:56:35bertriksaratoga, I don't know, but I found eyes-free not sure yet if this is a TTS in itself, or a framework around a TTS
19:57:09saratogai think thats just using espeak
19:57:31saratogafor some reason i thought google bought some TTS engine recently and open sourced it but i'm finding nothing
19:58:43 Quit guymann (Ping timeout: 244 seconds)
20:00:45 Join guymann [0] (
20:14:53 Quit n1s (Read error: Connection timed out)
20:16:45 Quit ks (Quit: Nettalk6 -
20:23:27 Nick untrack is now known as untrack|stdy (
20:24:21bertrikI pinpointed the "Error accessing playlist control file" problem after USB connect. The problem is that playlist_shutdown is called when USB is connected, leaving the playlist sub-system in a half-deinitialised state. The problem is triggered when the playlist thread tries to access certain fields again.
20:25:55kugelgevaerts: fwiw, I managed RaaAoA on the x86 emulator with some build system changes
20:28:57gevaertskugel: I did end up getting it to build. There's one bug in firmware/SOURCES (I'll push a fix later), then libmad refuses to build with mips asm for me so I had to disable that, and then I had to tell android.make that this is mips. After that it builds and installs, but it crashes immediately.
20:30:20 Join enthdegree [0] (~enthdegre@wikimedia/enthdegree)
20:34:53 Quit enthdegree (Ping timeout: 244 seconds)
20:37:06 Join enthdegree [0] (~enthdegre@wikimedia/enthdegree)
20:43:41 Nick untrack|stdy is now known as utrack|ZzZpoke (
20:48:17 Quit mikroflops (Quit: <(^_^)>)
20:49:57kugelgevaerts: i went with sigaltstack threads for x86 (because no asm for native ones exist), perhaps it's worth a try?
20:50:05kugelalso logcat output is always helpful
20:53:12gevaertskugel: how do i enable/get sigalstack threads?
20:53:17 Quit FOAD (Quit: I'll be back)
20:54:50 Join Wardo [0] (
20:56:31*gevaerts finds things
21:04:06 Quit WalkGood ()
21:12:51 Join lebellium [0] (
21:15:03gevaertskugel: it works!
21:16:18***Saving seen data "./dancer.seen"
21:19:04 Quit Wardo (Quit: Blarglarg)
21:23:15 Quit enthdegree (Ping timeout: 256 seconds)
21:25:49gevaertsHow do I tell git to only push a single change, even if I have other changes after it and changes in other branches?
21:27:36bertrikgevaerts, I'm no git expert, but I think you should move all of your changes to a branch, then cherry-pick the change and push that
21:28:11gevaertsbertrik: last time I pushed something in this sort of situation, it pushed from all branches I had
21:32:41AlexPFor e.g. android, do we have anything specific to different arm versions?
21:34:56funmani think we assume armv6 ?
21:35:15gevaertsI think so, yes
21:35:16Mode"#rockbox -o funman" by funman (~fun@rockbox/developer/funman)
21:35:23AlexPOK, cheers
21:35:32gevaertsI don't know if we have any armv7 asm, actually
21:35:52AlexPI was just wondering if we say assume v6 but some v7 stuff can be turned on if you select it
21:35:54funmanwe'd need cpu detection code and/or packaging changes
21:44:49bluebrothergevaerts: git push origin <hash>:master
21:44:56gevaertsAh, thanks!
21:45:04bluebrotherpretty much the same as the git push origin HEAD:/ref/for/master :)
21:45:14bluebrothererr, HEAD:refs/for/master
21:45:22bluebrotherexcept that you replace HEAD with the hash
21:45:28bluebrotherworks with gerrit as well :)
21:46:11bluebrotheramiconn: interested in giving g#272 a look?
21:46:12fs-bluebotGerrit review #272 at : Drop size entries from voicefiles. by Dominik Riebeling (changes/72/272/1)
22:02:05 Join enthdegree [0] (~enthdegre@wikimedia/enthdegree)
22:02:19amiconngevaerts: We do have some armv7 asm. Guess where...
22:02:38gevaertsape? :)
22:02:39amiconnIt's used on N900
22:03:13amiconnFor android we'd need some multilib mechanism, I think. We don't have that yet
22:03:56amiconnNeon helps a lot in the ape filters. >2x speedup...
22:04:27amiconn-c5000 is no problem on the N900 (less than 50% CPU load with stock kernel)
22:05:06AlexPWould it be worthwhile in the future being able to select at compile time (if there were ever more)?
22:07:55AlexPwe could just do a firefox/flash etc. and drop v6 :)
22:08:41funmanyou can package v6 and v7 code in one .apk
22:08:48funmanincluding C code built for both
22:09:12AlexPMuch increase in size?
22:09:26AlexPAs I would guess older v6 devices are also those that struggle for space
22:10:15funmanof course yeah
22:10:25amiconnIt's only really interesting for a few code paths
22:10:51funmanhaving only a few specialized functions means we need CPU detection code
22:11:10 Join FOAD [0] (~foad@
22:11:10 Quit FOAD (Changing host)
22:11:10 Join FOAD [0] (~foad@unaffiliated/foad)
22:11:21amiconnCan't we just ask the OS?
22:11:33amiconnI mean, multilib is only really interesting for hosted rockbox
22:11:42funmanhow do you ask the OS?
22:12:05funmanCPU detection can use existing OS functions of course
22:12:09amiconnI don't know, but I would expect the OS to know about cpu features
22:14:11gevaertsany MIPS people around?
22:16:03 Join n1s [0] (~n1s@rockbox/developer/n1s)
22:23:22 Join liar [0] (
22:26:11gevaertsHow can I reorder local commits (or get rid of one in between)?
22:26:33 Join RAThomas [0] (~chatzilla@
22:28:03gevaertsah, rebase -i
22:32:43CIA-47Commit 268b725 in rockbox by Frank Gevaerts: Rearrange #ifdefs a bit, so SIGALTSTACK overrides CPU-specific threading.
22:32:43CIA-47Commit d1560e5 in rockbox by Frank Gevaerts: Add the Android/MIPS target
22:35:58CIA-47d1560e5 build result: 0 errors, 8 warnings (Frank Gevaerts committed)
22:36:41gevaertsThat was not really intended
22:38:05 Join mikroflops [0] (
22:39:11gevaertsBuilding for mips/android still gives unless I disable asm for libmad
22:39:27 Quit [Saint] (Ping timeout: 245 seconds)
22:40:06kugelgevaerts: was this the only change required?
22:40:21gevaertskugel: I only tried mp3, so who knows?
22:40:34funmangevaerts: all in mad_f_mul ?
22:40:44gevaertsfunman: yes
22:41:05gevaertsMaybe other asm also breaks in more subtle ways
22:41:19*gevaerts first wants to fix the ypr0 build
22:41:56 Quit RAThomas (Quit: ChatZilla [Firefox 12.0/20120420145725])
22:42:41kugelgevaerts: is the asm perhaps targetting a other/newer mips version than the android toolchain?
22:43:12 Join saratoga_ [0] (98032941@gateway/web/freenode/ip.
22:43:18gevaertsI doubt it. I basically copied the mips CFLAGS from the onda port
22:43:25saratoga_FS #12013
22:43:25fs-bluebot Build libmad using mips-sde-elf-gcc 4.5.1:impossible constraint in 'asm' (patches, unconfirmed)
22:43:47kugelgevaerts: btw, I was about to suggest to switch to sigaltstack for all android builds
22:44:09gevaertssaratoga_: that's probably it!
22:44:32kugelgevaerts: ah okay, I copied the gcc cmdline from $NDK/ndk-build back then
22:44:50gevaertsI didn't think of looking there :)
22:45:19funmanwhich gcc version does android NDK use?
22:45:37funmanah the report is for 4.5.1 but says "since 4.4"
22:45:50bertrikhas anyone tried gcc 4.7 yet?
22:46:01funmani did on arm
22:46:15funmandont remember the results except -flto still doesn't work with -mthumb
22:47:33bertrikwe don't use -flto at all yet, do we?
22:47:38gevaertsDoes anyone object to me committing FS #12013 later today?
22:47:39fs-bluebot Build libmad using mips-sde-elf-gcc 4.5.1:impossible constraint in 'asm' (patches, unconfirmed)
22:48:26funmanbertrik: no of course we don't, it appeared in 4.6 i think
22:48:38funmanand is mostly only tested on x86
22:48:46funmanalso reported to be crap
22:48:48saratoga_i don't think that bug affects anything but MIPS, so go for it
22:49:01funmanbertrik: i'm interested into trying what makes the code smaller
22:49:15funmanso whole program optimization is the logical next step after thumb ;)
22:49:21kugelgevaerts: I dont think anyone is actively working on a mips target or even using one
22:49:35funmangevaerts is!
22:50:21 Quit remlap (Ping timeout: 244 seconds)
22:53:10n1sfunman: i think flto is supposed to be much better in 4.7 than it was in 4.6
22:53:22 Quit Strife89 (Quit: Heading out)
22:53:50funmann1s: in any case it doesn't work properly when mixing thumb and arm
22:54:35n1sfunman: hmm
22:55:01n1sdreamlayers posted a patch to build for sh with lto
22:56:41funmanand (more important)
22:56:44n1si guess it might be worthwhile to test if lto gives a bigger space benefit than thumb but i doubt it
22:57:24funmanit didnt
22:58:02n1sthe other disadvantage is that all code is effectively reoptimized by the compiler for each source change so build times get much longer than a traditional build when only doing small changes
22:58:33funmanthat's unavoidable but the same thing applies for thumb
22:58:50funmanwe try to build with -mthumb and if that fails (likely because it uses arm asm) we build again
22:58:52n1si'd be interested to try lto for codecs to see if there are any speed benefits
22:59:47n1sfunman: yeah but only for modified files, with lto all the bytecode is run through (most of the) compile each build
22:59:54bertrikHow does lto mainly reduce size, by removing unused functions?
23:00:06n1sbertrik: also cross module inlining
23:00:29n1sand other cross module optimixzations i gues
23:00:52 Join [Saint] [0] (~Saint]@unaffiliated/saint/x-8516940)
23:01:22n1sso the benefits are depending on the structure of the program
23:02:05amiconngevaerts: Regarding fs#12013, there is no need to roll your own gcc version detection
23:02:05fs-bluebot Build libmad using mips-sde-elf-gcc 4.5.1:impossible constraint in 'asm' (patches, unconfirmed)
23:02:43amiconnJust use GCCNUM (exported from the main makefile)
23:03:25amiconn#if GCCNUM >= 404
23:04:12*gevaerts nods
23:06:24CIA-47Commit 10b2143 in rockbox by Frank Gevaerts: The threading model should be set from configure, not config.h.
23:09:18CIA-4710b2143 build result: 0 errors, 8 warnings (Frank Gevaerts committed)
23:11:09gevaertskugel: I think there's something fishy there. YPR0 defines sigaltstack threads from configure, but due to the way ifdefs were ordered previously that wasn't actually used I think
23:11:59 Quit n1s (Quit: Ex-Chat)
23:16:20***Saving seen data "./dancer.seen"
23:17:33kugelthat's possible :)
23:18:13kugelit's not intended to use sigaltstack threads iirc
23:18:33kugelbut I can remember that I've seen those warnings locally too in the past
23:19:14gevaertsOK, I'll fix it by removing sigaltstack for ypr0 in configure then :)
23:19:34CIA-47Commit fbe33e0 in rockbox by Frank Gevaerts: (Author: Huan Zhang) Fix libmad compilation on MIPS with newer gcc. FS #12013
23:22:24CIA-47fbe33e0 build result: 42 errors, 8 warnings (Huan Zhang committed)
23:23:35 Part LinusN
23:23:44CIA-47Commit 20261dd in rockbox by Frank Gevaerts: Fix YP-R0 build.
23:24:26 Join knob [0] (~cela@
23:24:58 Join dfkt_ [0] (dfkt@unaffiliated/dfkt)
23:25:09 Part knob
23:25:49 Quit dfkt (Ping timeout: 260 seconds)
23:26:13CIA-4720261dd build result: 42 errors, 0 warnings (Frank Gevaerts committed)
23:26:20CIA-47Commit bf6abf7 in rockbox by Frank Gevaerts: Define GCCNUM properly. Case matters!
23:26:54gevaertsI suspect maemo and pandora might have the same issue as YP-R0
23:27:06gevaertsI don't have toolchains for those though
23:28:45 Join [Saint_] [0] (~Saint]@
23:28:46 Quit [Saint_] (Changing host)
23:28:47 Join [Saint_] [0] (~Saint]@unaffiliated/saint/x-8516940)
23:28:53CIA-47bf6abf7 build result: All green
23:28:54 Join [Saint__] [0] (~Saint]@unaffiliated/saint/x-8516940)
23:28:55amiconnBuilding for maemo is special. I have no idea how we'd do that with our own toolchain
23:29:41amiconnYou use scratchbox, which partially emulates arm, to make it appear to be native maemo
23:30:25amiconnAfaiu this is done because many programs' build systems don't handle cross compilation well (if at all)
23:31:49gevaertsI know. I used to have that installed, but I don't have it anymore
23:31:56 Quit [Saint] (Ping timeout: 252 seconds)
23:32:21WilliamCIt took me an hour to do a full format of my Clip Zip
23:32:24amiconnMaybe we could run our own toolchain (isn't ypr0 also linux?)
23:32:35WilliamCBut it works now
23:35:22 Join remlap [0] (
23:37:04WilliamCI'm not sure I'll trust it to run
23:38:02 Quit dfkt_ (Quit: -= SysReset 2.55=- Sic gorgiamus allos subjectatos nunc.)
23:42:38kugelgevaerts: maemo and Pandora always intentionally used sigaltstack, but I might have accidentally enabled native threads for them
23:43:08 Quit enthdegree (Ping timeout: 246 seconds)
23:43:27 Join Syconaut [0] (
23:53:01gevaertskugel: so possibly they're now fixed? Let's leave this to the port maintainers...
23:53:42kugelgood idea
23:54:17 Quit [Saint_] (Quit: Quit)
23:54:33 Join [Saint] [0] (~Saint]@
23:54:33 Quit [Saint] (Changing host)
23:54:33 Join [Saint] [0] (~Saint]@unaffiliated/saint/x-8516940)
23:55:11 Quit [Saint] (Client Quit)
23:55:28 Join [Saint] [0] (~Saint]@
23:55:28 Quit [Saint] (Changing host)
23:55:28 Join [Saint] [0] (~Saint]@unaffiliated/saint/x-8516940)
23:55:28 Quit [Saint__] (Read error: Connection reset by peer)
23:55:53 Join enthdegree [0] (~enthdegre@wikimedia/enthdegree)
23:56:52 Join bitcraft [0] (~bitcraft@

Previous day | Next day