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 2015-09-22

00:02:25 Quit ender` (Quit: 99% of lawyers give the rest a bad name.)
00:06:36 Quit Markmaster (Quit: My Mac has gone to sleep. ZZZzzz…)
00:07:31[Saint]prof_wolfff: Well done, by the way. That's awesome.
00:14:33 Quit xorly (Ping timeout: 264 seconds)
00:21:21 Quit girafe (Read error: Connection reset by peer)
00:22:11 Join [Franklin] [0] (~franklin@unaffiliated/franklin)
00:24:32 Quit amayer (Quit: Leaving)
00:30:18***Saving seen data "./dancer.seen"
00:38:38 Quit TheLemonMan (Quit: "It's now safe to turn off your computer.")
01:03:56 Quit dys (Ping timeout: 240 seconds)
01:10:12 Join dys [0] (
01:12:04 Join chrisb [0] (
01:15:26 Quit ZincAlloy (Quit: Leaving.)
01:49:10 Quit [Saint] (Remote host closed the connection)
01:49:59 Join [Saint] [0] (~hayden@rockbox/staff/saint)
02:10:00 Quit scorche (Disconnected by services)
02:10:04 Join scorche` [0] (~scorche@rockbox/administrator/scorche)
02:15:33 Quit ParkerR (Ping timeout: 252 seconds)
02:18:21 Quit emphaser ()
02:21:15 Join ParkerR [0] (~ParkerR@unaffiliated/parkerr)
02:22:40 Quit alexbobp (Ping timeout: 244 seconds)
02:23:42 Join alexbobp [0] (
02:24:04 Nick alexbobp is now known as Guest93303 (
02:30:20***Saving seen data "./dancer.seen"
02:32:18 Join tpr888 [0] (32a517de@gateway/web/freenode/ip.
02:42:21 Join Strife89 [0] (
02:43:52 Quit tpr888 (Quit: Page closed)
02:46:20 Quit chrisb (Ping timeout: 240 seconds)
02:51:48 Quit JdGordon_ (Ping timeout: 240 seconds)
02:59:22 Join JdGordon [0] (
02:59:22 Quit JdGordon (Changing host)
02:59:22 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
03:02:25jtdesigns01guys, why am i getting "No rule to make target `/home/jonathan/rockbox/builds/sansa/apps/plugins/dark/typedef.o', needed by `/home/jonathan/rockbox/builds/sansa/apps/plugins/dark/gob_viewer.rock'. Stop."
03:02:54jtdesigns01 when i use this makefile i made
03:03:35jtdesigns01(gob_viewer.c is just a simple hello world plugin)
03:07:11[Franklin]is there a file called typedef.c?
03:09:15jtdesigns01no, not in my folder, just gob_viewer.c and the makefile
03:09:18[Franklin]post your code please
03:10:20[Franklin]ok, so it's nothing special
03:10:55jtdesigns01i added the "dark" folder to SUBDIRS
03:12:00jtdesigns01and "gob_viewer,demos" to CATEGORIES
03:13:20[Franklin]something with the makefile
03:13:35[Franklin]try as a template instead
03:14:26jtdesigns01one thing i want is to have multiple rocks coming from that folder
03:14:33jtdesigns01like "shortcuts"
03:14:48jtdesigns01(the plugin)
03:16:14[Franklin]I think that should be ok
03:16:32[Franklin]just write a new make target for each rock
03:16:46jtdesigns01make target?
03:16:57[Franklin]a target in the makefile
03:17:10[Franklin]i.e. TARGET_NAME: DEPENDENCIES
03:17:52jtdesigns01hmm ok
03:18:06[Franklin]in your makefile, you have a target for gob_viewer.rock
03:19:39jtdesigns01can you see anything wrong with mine? i hand-copied it from sudoku and changed and renamed parts of it
03:19:52[Franklin]try this:
03:21:38jtdesigns01still getting the error
03:21:50[Franklin]can you pastebin the lines before the error?
03:21:52[Franklin]just a few
03:23:19jtdesigns01i`ll try a clean recompile in that folder
03:23:19jtdesigns01*build folder
03:23:23[Franklin]for some reason it wants to build typedef.o
03:24:40jtdesigns01on a clean recompile i get
03:25:08[Franklin]have you changed the makefile?
03:25:15jtdesigns01i`m using yours
03:25:45jtdesigns01but i get the same thing with mine
03:26:01[Franklin]latest HEAD?
03:26:35jtdesigns01i`m pretty sure its up to date if thats what you mean, updated last night
03:27:08[Franklin]I can't seem to figure it out
03:27:26[Franklin]Anyone else have some ideas? I gotta run now
03:27:44[Franklin]the sysfont.h error really puzzles me
03:27:52[Franklin]have you changed anything else in the source tree?
03:28:13jtdesigns01nope, but i will try rebasing
03:28:35[Franklin][Saint]: can you help here? I gotta run
03:28:46[Franklin]good night
03:28:55jtdesigns01thanks for the time
03:29:02 Quit [Franklin] (Quit: leaving)
03:52:51 Join [Saint_] [0] (~hayden@rockbox/staff/saint)
03:54:13 Quit [Saint] (Read error: Connection reset by peer)
03:58:04jtdesigns01ok guys i found out my problem!
03:58:30jtdesigns01my problem was here: "DARK_SRC := $(call preprocess, $(DARKSRCDIR)/gob_viewer.c)"
03:59:13jtdesigns01where it says gob_viewer,c
03:59:27jtdesigns01it used to say SOURCES
04:00:13jtdesigns01therefore, it looks in the file given for a list of sources, it doesnt use it as a source
04:00:33jtdesigns01so, it was looking in my source for a list of sources and getting all messed up.
04:00:45jtdesigns01hooray, problem solved!
04:02:29jtdesigns01(now, time to go to bed so tomorrow i can write a paper for school )
04:02:34 Quit jtdesigns01 (Quit: Leaving)
04:17:20 Join krabador [0] (~krabador@unaffiliated/krabador)
04:30:21***Saving seen data "./dancer.seen"
04:40:20 Quit kvieta- (Excess Flood)
04:40:55 Join kvieta [0] (~kvieta@
04:44:24 Quit kvieta (Excess Flood)
04:44:55 Join kvieta [0] (~kvieta@
04:45:43 Nick Guest93303 is now known as alexbobp (
04:46:13 Nick alexbobp is now known as Guest38107 (
04:46:37 Nick Guest38107 is now known as alexbobp (
04:48:20 Quit krabador (Remote host closed the connection)
04:54:06 Quit JanC (Ping timeout: 248 seconds)
05:08:16 Join JanC [0] (~janc@lugwv/member/JanC)
05:31:21 Join chrisb [0] (
05:55:38 Quit [7] (Ping timeout: 240 seconds)
05:57:30 Join TheSeven [0] (~quassel@rockbox/developer/TheSeven)
05:58:06 Join Guest66888 [0] (
06:00:13 Quit Guest18189 (Ping timeout: 244 seconds)
06:11:05 Quit chrisb (Ping timeout: 240 seconds)
06:30:24***Saving seen data "./dancer.seen"
06:31:42 Quit Strife89 (Ping timeout: 260 seconds)
07:44:50 Nick [Saint_] is now known as [Saint] (~hayden@rockbox/staff/saint)
08:04:34 Join wodz [0] (
08:08:29 Quit djukon (Ping timeout: 240 seconds)
08:10:18 Join djukon [0] (transitor@gateway/shell/insomnia247/x-pwwcbodtqsgntkio)
08:20:49 Join ender` [0] (
08:30:14 Quit amiconn (Remote host closed the connection)
08:30:14 Quit pixelma (Remote host closed the connection)
08:30:25 Join pixelma [0] (~pixelma@rockbox/staff/pixelma)
08:30:27 Join amiconn [0] (~amiconn@rockbox/developer/amiconn)
08:30:29***Saving seen data "./dancer.seen"
08:51:23 Quit [Saint] (Remote host closed the connection)
08:52:13 Join [Saint] [0] (~hayden@rockbox/staff/saint)
08:53:09 Join petur [0] (~petur@
08:53:09 Quit petur (Changing host)
08:53:09 Join petur [0] (~petur@rockbox/developer/petur)
09:41:21 Quit ruhannn (K-Lined)
09:41:23 Quit Ivoah2 (K-Lined)
10:30:31***Saving seen data "./dancer.seen"
10:34:32 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
10:38:03 Join maruk [0] (
10:44:52 Quit pamaury (Remote host closed the connection)
10:45:22 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
10:56:25 Quit yosafbridge (Ping timeout: 240 seconds)
11:01:43 Join yosafbridge [0] (~yosafbrid@2607:5300:100:200::160d)
11:25:35 Join einhirn [0] (
11:51:20 Join Ivoah2 [0] (uid49352@gateway/web/
12:02:12kugelwodz: ping
12:05:46 Join ruhannn [0] (uid76353@gateway/web/
12:17:01wodzkugel: where is firmware/kernel/pthread/* used?
12:17:14kugelnowhere, i checked it in by accident
12:17:16wodzkugel: I was surprised it is not used on android
12:17:56kugelit was part of my playbacklib effort, i don't know how i managed that
12:19:56wodzkugel: Other question - there is 'malloc backend for buflib' patch on gerrit. What's holding this? I looks pretty interesting.
12:20:26kugeli was never happy with it
12:20:34kugelwell not completely happy*
12:21:15wodzwhat was the problem?
12:21:17kugelyou can't shrink a buffer without making a new allocation (which a) fail and b) changes the pointer)
12:22:06kugeli changed some code to cope with shrinking changing pointers but it never felt great to me
12:23:56kugelalso the original buflib feature (IMO) that mem usage is capped is not implemented
12:26:36kugelthen i realized that this feature contradicts with the main motivation of making it malloc based (to allow continuing playback while other code makes allocs)
12:29:05wodzyes, definitely this should work like this on hosted targets
12:30:23kugelso i kind of got in the moot it's not actually worth it, instead the playback engine should be able to deal with shrink buffers in a glitch-free manner
12:30:30kugelbut I never got it to work properly
12:30:34***Saving seen data "./dancer.seen"
12:30:39kugelbecause buffering is so complicated to my brain
12:42:47wodzkugel: Probably I terribly miss something but why buflib_shrink can't simply call realloc()?
12:43:32kugelthat's what I said, it's (potentially) a new allocation
12:44:21wodzThe realloc() function changes the size of the memory block pointed to by ptr to size bytes. The contents will be unchanged in the range from the start of the region up to the minimum of the old and new sizes.
12:44:43wodzI read it as realloc with shrinking does not make new allocation
12:44:53kugelalso, realloc always keeps the memory at the start of the original alloc. buflib allows to shrink from the front
12:45:35 Join emphaser [0] (
12:45:50wodzIsn't it one of MMUless peculiarities of buflib?
12:46:00kugelthe latter can be worked around
12:46:45kugelanyway even if realloc was a perfect fit, you still have to stop playback for a brief moment
12:47:39kugelinitially I thought we simply don't shrink, but then i began to realize that capped mem usage is actually a feature and wanted to maintain that
12:47:39wodzwhy exactly? (I am trying to build up bigger image of the problem)
12:48:15kugelthe buffering engine needs to update its management
12:48:47wodzkugel: I'd say capped mem is the solution for a) low mem targets b) MMUless
12:49:09kugelone way to to do it race free is to stop playback, there are other methods but I haven't got them to work
12:49:09wodzkugel: As long as we don't leak mem in the long run
12:50:04kugelwodz: I'd love to be able to configure a potential playbacklib to use max. N MB memory
12:51:30wodzkugel: Thats nice sure, but honestly considering a) typical mem usage b) ram size in current mobiles is it really important?
12:52:14kugelyou are free to pick it up and do what you want with it, I'm not into rockbox at the moment
12:53:18kugelwhat I didn't like most is that we have 2 or 3 places that do a "give me everything you got, I'll work with it and give back what I don't need"
12:53:58kugelyou have to limit those so some maximum regardless
12:55:14kugelso you have to put (artificial) limits and still don't get the "buflib use max. N MB as a whole". kind of the worst of both worlds
12:55:44kugelanyway i didnt say it didn't work (i think the patch mostly works), just that I was never completely happy with it#
12:57:27kugeli was also unsure about if the move callback should be called if shrinking leads to a new allocation or not, if the general intention was to about needing move and shrink
12:59:41kugelthings would be easier if there was a reliable promise that realloc never makes a new allocation when shrinking
13:02:15wodzkugel: I guess the only way to know is to look at realloc() implementation
13:02:28kugelwhich one :-)
13:04:13kugel(we could also ship our own like using our tlsf copy)
13:05:23wodzkugel: one way would be to realloc() into tmp pointer and compare tmp and original and panicf() when block is moved. That way we should pretty fast find whether shink realloc() does not move holds or not.
13:09:23wodzkugel: But in general realloc() indeed does not guarantee that block is not moved. So you were right
13:11:30kugeladditionally to that, realloc is not really suitable for shrinking from the front
13:11:54kugelwhich is the main use case for shrinking
13:12:34kugelthough that can probably be solved elsewhere
13:13:43kugelshrinking from the front is really only needed in a flat memory layout like on native targets
13:13:57wodzthats what I said
13:17:21kugeli'd love if you'd get it to work
13:21:38wodzLooks like new allocation is the only reliable way.
13:40:23 Join chrisb [0] (
13:47:27 Quit ruhannn (Quit: Updating details, brb)
13:47:37 Join ruhan [0] (uid76353@gateway/web/
14:04:40 Quit chrisb (Ping timeout: 256 seconds)
14:09:01 Join krabador [0] (~krabador@unaffiliated/krabador)
14:15:23 Quit krabador (Quit: Take The Time)
14:27:58 Quit yosafbridge (Ping timeout: 240 seconds)
14:30:37***Saving seen data "./dancer.seen"
14:33:20 Quit soap_ (Read error: Connection reset by peer)
14:33:42 Join yosafbridge [0] (~yosafbrid@2607:5300:100:200::160d)
14:35:31 Join soap [0] (~soap@rockbox/staff/soap)
14:40:56 Join chrisb [0] (
14:45:11 Join amayer [0] (
14:45:18 Quit chrisb (Ping timeout: 250 seconds)
14:46:44 Quit soap (Read error: Connection reset by peer)
14:48:19 Join soap [0] (~soap@rockbox/staff/soap)
15:07:44 Quit soap (Read error: Connection reset by peer)
15:09:02 Join soap [0] (~soap@rockbox/staff/soap)
15:17:29 Join chrisb [0] (
15:20:25 Quit wodz (Remote host closed the connection)
15:26:25 Quit chrisb (Ping timeout: 240 seconds)
15:33:45 Join chrisb [0] (
15:40:17 Quit emphaser ()
15:47:09 Quit chrisb (Ping timeout: 244 seconds)
15:51:58 Join ribblle [0] (
15:53:44ribbllei'm trying to install rockbox on a speedo mmp3 player. It was made by iriver but i can't tell what drivers they've used. I've also formatted it. Is there a way of upgrading it with your lovely kit?
15:57:55ribbllehope that made any sense
15:57:57gevaertsNo. Rockbox runs only on the exact devices listed on the front page
15:58:36ribbllealright. Is there any other way for me to hack it?
15:59:38ribblleand would i have any luck in the unsupported builds part of the forum?
15:59:44gevaertsWell, you can start a port, but that's not exactly simple or quick. Have a look at
16:01:12gevaertsIt's not a matter of "supported". Unless there's specific code to run on a particular device, it won't work
16:04:49ribbllewill the work already done for iriver players help much?
16:05:20ribblleor can you recommend me a different way to scrobble music on my mp3. That's all i want
16:05:26gevaertsThat entirely depends on how similar the thing is
16:05:55gevaertsBut I suspect the answer is "no"
16:09:40 Quit pamaury (Read error: No route to host)
16:12:28 Quit yosafbridge (Ping timeout: 246 seconds)
16:17:48 Join yosafbridge [0] (~yosafbrid@2607:5300:100:200::160d)
16:18:05 Part ribblle
16:22:48 Join chrisb [0] (
16:30:40***Saving seen data "./dancer.seen"
16:55:07 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
17:30:00 Join ZincAlloy [0] (
17:31:20 Quit petur (Quit: Leaving)
18:19:27 Quit Moarc (Ping timeout: 272 seconds)
18:26:15 Quit maruk (Quit: Leaving.)
18:30:41***Saving seen data "./dancer.seen"
18:40:55 Nick ruhan is now known as ruhans (uid76353@gateway/web/
18:58:30 Join Moarc [0] (
19:04:18 Quit ender` (Read error: Connection reset by peer)
19:04:42 Join ender` [0] (
19:23:12 Join lebellium [0] (
19:32:30 Join girafe [0] (
19:44:39 Join TheLemonMan [0] (~lemonboy@unaffiliated/thelemonman)
19:57:56 Join petur [0] (~petur@rockbox/developer/petur)
20:26:07 Quit Bray90820 (Read error: Connection reset by peer)
20:30:42***Saving seen data "./dancer.seen"
20:36:44 Quit chrisb (Ping timeout: 240 seconds)
20:46:10 Quit pamaury (Ping timeout: 246 seconds)
20:57:49 Quit fs-bluebot_ (Ping timeout: 240 seconds)
20:58:57 Quit bluebrother^ (Ping timeout: 240 seconds)
21:01:13 Join bluebrother [0] (~dom@rockbox/developer/bluebrother)
21:01:14 Join fs-bluebot [0] (
21:36:17 Quit shamus (Ping timeout: 256 seconds)
21:36:51 Join shamus [0] (
21:59:09 Join wodz [0] (
22:01:25wodzkugel: tlsf does not move block when shrinking with realloc(). But this doesn't help much as IMO the whole point is to relay on host malloc() and friends. Otherwise you would need to create tlsf pool with host malloc/sbrk/mmap and operate like on static buffer.
22:03:02kugelI dont know the tlsf internals but I'd think it manages the pool so that it gives back unallocated pages
22:05:30kugeland then it depends on whether you want to keep the property of max. mem usage
22:06:31 Join Markmast_ [0] (
22:07:02wodzkugel: I just looked at source out of curiosity
22:08:53kugeli'd say it doesn't have to be the *host* malloc but one that can reclaim unused memory and deal with fragmentation
22:11:28wodzkugel: But where do you see the benefit? buflib can coalescent mem
22:12:12kugelbut only by actually memmoving allocs
22:13:14wodzkugel: As I understand the main problem is the static pool not memmoving
22:13:21kugelas i said earlier, the initial motiation was to avoid having to move or shrink other allocs when a new one is made, thereby getting around the playback stop
22:14:40kugelthis goal is in and of itself with the max. mem usage feature, but i didn't think dropping that feature is the end of the world
22:18:10wodzI still can't see how tlsf helps here
22:19:46wodzunless you do tlsf allocation from other pool then buflib pool
22:22:30 Join Bray90820 [0] (~Bray90820@2604:2d80:800a:83e2:1c6b:a421:dde0:d215)
22:28:26 Join Markmas__ [0] (
22:30:46***Saving seen data "./dancer.seen"
22:30:46 Quit Markmast_ (Ping timeout: 240 seconds)
22:48:21 Quit wodz (Ping timeout: 268 seconds)
22:51:53 Join Bray9082_ [0] (
22:54:47 Quit Bray90820 (Ping timeout: 240 seconds)
23:05:15 Quit lebellium (Quit: ChatZilla 0.9.92 [Firefox 41.0/20150917150946])
23:09:36 Quit einhirn (Quit: Miranda IM! Smaller, Faster, Easier.
23:13:34 Quit krnlyng (Ping timeout: 264 seconds)
23:15:05 Nick Bray9082_ is now known as Bray90820 (
23:17:25 Quit girafe (Quit: Leaving)
23:18:05 Quit ender` (Quit: There are three deaths. The first is when the body ceases to function. The second is when the body is consigned to the grave. The third is that moment, sometime in the future, when your name is spoken for the last time. -- David Eagleman)
23:21:21 Quit Bray90820 (Read error: Connection reset by peer)
23:21:47 Join Bray90820 [0] (~Bray90820@2604:2d80:800a:83e2:5c77:fbe7:fa33:47a7)
23:32:11 Join Bray9082_ [0] (
23:34:46 Quit Bray90820 (Ping timeout: 240 seconds)

Previous day | Next day