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 2015-03-24

00:00:40 Quit Elfish (Ping timeout: 272 seconds)
00:03:31 Join Elfish [0] (amba@2001:1608:12:1:13:3:3:7)
00:07:47 Join JdGordon_ [0] (~jonno@rockbox/developer/JdGordon)
00:08:35 Quit JdGordon (Ping timeout: 265 seconds)
00:12:46 Join JdGordon [0] (
00:12:46 Quit JdGordon (Changing host)
00:12:46 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
00:12:57 Quit JdGordon_ (Ping timeout: 264 seconds)
00:29:45 Quit [Franklin] (Ping timeout: 264 seconds)
00:30:21 Quit AlexP (Read error: Connection reset by peer)
00:31:27 Join [Franklin] [0] (
00:44:09 Quit JanC (Ping timeout: 264 seconds)
00:57:36 Join JanC [0] (~janc@lugwv/member/JanC)
01:01:38 Quit ParkerR (Read error: Connection reset by peer)
01:02:25 Quit Scromple (Quit: Leaving)
01:12:22 Join Scromple [0] (~Simon@
01:17:41 Join JdGordon_ [0] (~jonno@rockbox/developer/JdGordon)
01:19:53 Quit JdGordon (Ping timeout: 255 seconds)
01:20:31***Saving seen data "./dancer.seen"
01:22:13 Quit JdGordon_ (Ping timeout: 252 seconds)
01:22:43 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
01:36:58 Quit dfkt (Read error: Connection reset by peer)
01:40:08 Quit user890104 (Ping timeout: 255 seconds)
01:43:22 Join user890104 [0] (Venci@unaffiliated/user890104)
01:52:05 Quit GeekShadow (Ping timeout: 252 seconds)
01:53:11 Quit JdGordon (Ping timeout: 252 seconds)
01:53:41 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
02:19:00 Join GeekShadow [0] (~antoine@reactos/tester/GeekShadow)
02:22:21 Join AlexP [0] (~alex@rockbox/staff/AlexP)
02:40:31 Join JdGordon_ [0] (~jonno@rockbox/developer/JdGordon)
02:42:21 Quit JdGordon (Ping timeout: 264 seconds)
02:48:35 Join Strife89 [0] (
02:53:18 Quit AlexP (Remote host closed the connection)
03:03:07 Quit ZincAlloy (Quit: Leaving.)
03:06:21 Quit RiD (Quit: A good plan today is better than a perfect plan tomorrow.)
03:20:33***Saving seen data "./dancer.seen"
03:45:42 Quit krabador (Quit: Sto andando via)
03:56:21 Quit [Franklin] (Quit: Lost terminal)
04:19:35 Quit ungali (Quit: ungali)
04:20:16 Join ungali [0] (~ungali@unaffiliated/ungali)
04:29:09 Join CrashBash-Kun [0] (
05:20:35***Saving seen data "./dancer.seen"
05:36:08 Quit TheSeven (Disconnected by services)
05:36:22 Join [7] [0] (~quassel@rockbox/developer/TheSeven)
05:36:59 Quit CrashBash-Kun (Ping timeout: 272 seconds)
06:04:34 Join CrashBash-Kun [0] (
06:19:28 Quit Strife89 (Ping timeout: 256 seconds)
06:29:58 Quit CrashBash-Kun (Ping timeout: 255 seconds)
06:37:59 Join CrashBash-Kun [0] (
06:40:26 Quit CrashBash-Kun (Client Quit)
06:41:52 Join AlexP [0] (~alex@rockbox/staff/AlexP)
07:05:17 Quit kiwicam (Remote host closed the connection)
07:15:14 Quit AlexP (Remote host closed the connection)
07:18:36 Join AlexP [0] (~alex@rockbox/staff/AlexP)
07:20:37***Saving seen data "./dancer.seen"
07:27:25 Join mortalis [0] (~kvirc@
07:52:07 Join xorly [0] (
08:01:22 Quit AlexP (Remote host closed the connection)
08:05:48 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
08:16:46 Quit xorly (Ping timeout: 256 seconds)
08:26:55 Join kiwicam [0] (
08:36:45 Join wodz [0] (
08:38:19wodzpamaury: looking at descrambled x1 sys.bin I think descrambling is not totally correct. I might be wrong of course but some instruction sequences look just wrong.
08:42:38pamaurywodz: do you have precise examples ? in particular at which offset ?
08:43:18pamauryI need to go, I'll read the logs to see the answer
08:45:48wodzpamaury: I have to go too. Particularly I am looking at context saves and restores. This are long sequence of lw/sw with grabbing some stuff from coprocessor0 and from coprocessor1. There are blocks of undefined instructions in between which looks just wrong. Additionally functions prologues are strange here and there (which might be related to compiler but that would be very strange as they use gcc)
08:46:32wodzpamaury: what is the cipher block size?
08:48:24 Quit pamaury (Ping timeout: 264 seconds)
08:58:16 Quit wodz (Quit: Leaving)
09:04:41 Join einhirn [0] (
09:15:47 Quit shamus (Ping timeout: 244 seconds)
09:16:11 Join shamus [0] (
09:20:38***Saving seen data "./dancer.seen"
09:23:52 Join ender` [0] (
09:33:13 Join pamaury [0] (
09:33:30 Quit pamaury (Changing host)
09:33:30 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
09:34:08pamaurywodz (logs): cipher size is 344
09:41:48 Quit shamus (Read error: Connection reset by peer)
09:41:53pamauryI'll try to have a quick look at the problem
09:42:19 Join shamus [0] (
09:46:09 Join pamaury_ [0] (~quassel@rockbox/developer/pamaury)
09:51:29 Join wodz [0] (
09:51:47 Join Markmaster [0] (
09:53:50pamaurywodz: I had a look and I'm a bit puzzled
09:54:33pamauryassuming the cipher is a rotated XOR of size 344, I'm almost sure it's correct, just looking at the strings and ther remarkable features of the binary, any error would be quite obvious
09:55:10pamauryon the other hand, some pieces of code have undefined instruction when loaded in IDA indeed. However, may those undefined instruction be XBurst specific ones ?
09:55:22pamaurywhat did you use to disassemble it ?
09:55:37wodzpamaury: r2 and objdump
09:56:18wodzpamaury: It could be xburst instructions yes BUT as far as I know xmu extension is about simd which are rather unlikely to be used in context save
09:56:59wodz344 is block size?
09:57:17 Quit pamaury_ (Ping timeout: 246 seconds)
09:57:26pamauryyes that's my best guess
09:57:58wodzis it coding byte-by-byte?
09:59:04pamauryI don't know, I don't have any code to reverse engineer, that's just a guess. I think it's byte by byte xor yes
09:59:49pamauryI posted the code on gerrit g#1165 if you want to see it by yourself
09:59:54fs-bluebotGerrit review #1165 at : Add jz4760 tool by Amaury Pouly
10:04:13wodzpamaury: looking block-by-block decoding looks sane so this looks like this particular binary weirdness
10:05:24 Join pamaury_ [0] (~quassel@rockbox/developer/pamaury)
10:05:29pamauryone way to be sure would be to find the decoding sequence in the firmware, assuming it's not done by the bootrom
10:06:19wodzkey should be easy to find out unless it isn't generated dynamically on startup
10:06:59pamauryyes but this binary looks a bit weird, like if there were many binaries in it, not so easy to find the code if it references some data somewhere
10:07:47pamaury_I have found a "IHFS" string in the binary
10:07:56pamaury_but I couldn't find the xor key
10:08:33wodzyes this binary pretty much looks like concatenation of a few images
10:09:37pamaurybut the binaries seem to share the data sections and so on, so maybe some linker magic
10:12:38pamaury_haha, loading the file at 0x80001000 in ida seems to give good results
10:14:05pamaury_I'm wondering which OS is that
10:14:39wodzuCOS-II mos probably
10:14:55pamaury_is that open source ?
10:14:56wodzI'd say it should be loaded at 0x80001000
10:15:01wodzsort of
10:15:44pamaury_this link may be of interest:
10:16:26wodzprogramming manual says that cache is memmapped @ 0x80000000 and is 16kB each (icache and dcache)
10:17:05wodzthis is weird however as I never heard of MIPS with mem mapped cache
10:17:17wodzmaybe they mean iram/TCM or something
10:17:32wodzthe SoC is dualcore BTW
10:17:41 Quit Markmaster (Quit: Markmaster)
10:18:07pamaury_yes but apparently the second code is for VPU only, I don't know what that means exactly, I guess it doesn't have a MMU
10:19:04pamaury_does ucos uses a bootloader like u-boot ?
10:19:04wodzit is referenced as VPU but it looks like plain MIPS core with c0, c1 and mmu stripped
10:19:34wodzat least iriver use custom loading code
10:21:47pamauryI need to work, I'll see this later. Would be helpful to check it's indeed uCOS-II, that would be really helpful to RE
10:22:30wodzBut in general I started to think if x1 is really good target for rb. The selling point is support for hires files which rb does not support (you know what I mean - we decode this but downsample and cast to 16bits)
10:23:13wodzIt is unlikely someone will stepup and add support for this to rb. I am excluding the bit about real need for something like this in DAP
10:25:48 Quit pamaury_ (Ping timeout: 252 seconds)
10:26:25pamauryyeah, that would definitely be a port just for fun, and saratoga said we are underpar on mips wrt to codec optimisation too
10:27:24wodzThis I see as challenge opportunity personally :-)
10:29:12pamauryand working on mips sounds like fun, I'm really thinking we should do it ^^
10:29:49pamauryNot to mention people will probably adore us for porting rockbox to a hifi player ;) We just have to pretend rockbox sound better right ? :-p
10:33:14wodzmore neutral, without colorization, with sweet and crystal clear tweets, deep and precise middle, broad and detailed stage and, of course, strong bass base but nowhere thumping :P
10:36:40pamauryI wouldn't say it better
10:37:21 Quit prof_wolfff (Remote host closed the connection)
10:37:49wodzI would but at least 2 beers are needed :P
10:44:35pixelmaeveryone knows that rockbox sounds like a bag of shit
10:49:29 Join tittom [0] (6dbe8e7b@gateway/web/freenode/ip.
11:11:23 Quit einhirn (Ping timeout: 272 seconds)
11:19:57 Join Markmaster [0] (
11:20:40***Saving seen data "./dancer.seen"
11:27:41 Quit tittom (Quit: Page closed)
11:40:08 Join Markmaster__ [0] (
11:40:34 Quit Markmaster (Ping timeout: 264 seconds)
11:40:34 Nick Markmaster__ is now known as Markmaster (
11:44:17 Join Markmaster_ [0] (
11:45:00 Quit Markmaster (Ping timeout: 252 seconds)
11:45:00 Nick Markmaster_ is now known as Markmaster (
11:47:35 Quit Markmaster (Read error: Connection reset by peer)
11:56:51 Join prof_wolfff [0] (
12:11:54 Join ZincAlloy [0] (
12:29:05 Join pamaury_ [0] (~quassel@rockbox/developer/pamaury)
12:42:13 Join AlexP [0] (~alex@rockbox/staff/AlexP)
12:44:02 Join Markmaster [0] (
12:58:12 Join xorly [0] (
13:03:18 Join dfkt [0] (~dfkt@unaffiliated/dfkt)
13:13:21 Quit AlexP (Remote host closed the connection)
13:20:44***Saving seen data "./dancer.seen"
13:26:22 Quit pamaury_ (Ping timeout: 244 seconds)
13:32:23 Quit Markmaster (Ping timeout: 265 seconds)
13:34:14 Join Markmaster [0] (
13:37:26 Quit xorly (Ping timeout: 246 seconds)
13:38:53 Quit Markmaster (Client Quit)
13:49:25 Join ii_ [0] (
13:59:13 Join AlexP [0] (~alex@rockbox/staff/AlexP)
14:07:54 Join amayer [0] (
14:24:06 Quit prof_wolfff (Ping timeout: 265 seconds)
15:01:13 Quit wodz (Quit: Leaving)
15:16:36saratoga24 bit should more or less work in rockbox if you had a 24 bit driver and made some minor changes to the PCM code
15:19:20pamaurybut is there any quality advantage to using 24-bit appart from saying you do ?
15:19:41saratogaif the DAC has more than 16 effective bits resolution, you would increase SNR somewhat
15:20:30pamauryyou are suggesting to upsample everything from 16-bit to 24-bit and not touch any codec right ?
15:20:41saratogai doubt it matters much in practice, but if the hardware is very good there is no reason not use that extra capability
15:20:47***Saving seen data "./dancer.seen"
15:20:59saratogathe output of the codecs is 32 bit
15:21:10saratogawe just convert to 16 bit right before the driver
15:21:20pamauryoh wow
15:21:24saratogasince we have 32 bit GPRs on our cpus
15:21:24pamauryI didn't know that
15:21:54saratogadoing math on 16 bit values is actually hard, they tend to overflow
15:22:14saratogaso we try to avoid it
15:22:37saratogafor lossless formats the output will have zeros in the extra bits of course
15:22:52saratogaassuming no EQ at least
15:24:56pamauryso we should definitely port rockbox to the fiio x1 :)
15:26:47saratogathe ibasso player can do 17 or 18 effective bits, so i had suggested getting 24 bit output to work on it
15:27:06saratogatheres no cost aside from some tiny bit more DMA traffic
15:27:59ZincAlloyIt might not be the only player to benefit from it. As far as I understand, signals below the noise floor can be audible..
15:28:41saratogadepends on what you mean by that statement
15:29:19ZincAlloydynamic range is usually defined between noise floor and clipping. but apparently some signals may be audible below the noise floor.
15:29:44saratogaaudibility of one signal in the presence of another is usually determined by things like masking, not absolute intensity
15:29:57saratogabut in this case we're not talking about audibility
15:29:59ZincAlloynot that it would make any difference at reguar listening levels..
15:30:08saratogarather the effective resolution of piece of hardware
15:31:27saratogaa simple test would be to turn on noise shaping in the sound settings and then run RMAA and see if the noise floor falls
15:33:12ZincAlloyin any case, audiophiles are gonna love 24bit playback. it might create a lot of interest in those circles..
15:35:48saratogaas i understand it (and I should ask jhMikeS since this is his code), this is where we go from 32 to 16 bit at the end of the DSP engine
15:39:13saratogalooking at pcmbuf.c, crossfading assumes 16 bit, so thats annoying
15:45:23 Quit AlexP (Remote host closed the connection)
16:10:39 Join pamaury_ [0] (~quassel@rockbox/developer/pamaury)
16:22:55 Quit pamaury_ (Ping timeout: 252 seconds)
16:35:59 Join krabador [0] (~krabador@unaffiliated/krabador)
16:42:00 Quit mortalis (Ping timeout: 250 seconds)
17:07:49 Join advcomp2019__ [0] (~advcomp20@unaffiliated/advcomp2019)
17:11:33 Quit advcomp2019_ (Ping timeout: 256 seconds)
17:20:50***Saving seen data "./dancer.seen"
17:21:26 Quit ii_ (Remote host closed the connection)
17:58:11 Join AlexP [0] (~alex@rockbox/staff/AlexP)
18:03:08 Join pamaury_ [0] (~quassel@rockbox/developer/pamaury)
18:11:38 Join wodz [0] (
18:12:57wodzsaratoga: If we can do 24bits without much hassle we should offer such possibility on hardware which supports this. This goona be strong selling point.
18:15:08 Quit pixelma (Ping timeout: 265 seconds)
18:15:08 Quit amiconn (Ping timeout: 265 seconds)
18:19:15 Join pixelma [0] (~pixelma@rockbox/staff/pixelma)
18:19:18 Join amiconn [0] (~amiconn@rockbox/developer/amiconn)
18:20:08 Join bluebrother [0] (~dom@rockbox/developer/bluebrother)
18:22:52 Quit fs-bluebot (Ping timeout: 265 seconds)
18:23:01 Quit bluebrother^ (Ping timeout: 246 seconds)
18:23:14 Join fs-bluebot [0] (
18:33:54 Join CaptainKewll [0] (4081bb72@gateway/web/freenode/ip.
18:35:38 Quit AlexP (Ping timeout: 245 seconds)
18:37:51 Join AlexP [0] (~alex@rockbox/staff/AlexP)
18:38:33 Quit pamaury_ (Ping timeout: 245 seconds)
18:41:17 Join ii_ [0] (
18:41:50 Quit CaptainKewll (Quit: Page closed)
18:42:22 Join pamaury_ [0] (~quassel@rockbox/developer/pamaury)
18:47:40pamaury_wodz: I had a quick look at the firmware and I think I found one of the main entry point (ie not just the OS), at least with IDA this triggers the creation of many functions
18:59:13 Quit wodz (Ping timeout: 256 seconds)
19:07:33 Quit pamaury (Remote host closed the connection)
19:12:24 Quit pamaury_ (Ping timeout: 256 seconds)
19:12:27 Join RiD [0] (
19:20:52***Saving seen data "./dancer.seen"
19:23:45 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
19:24:47 Quit JdGordon_ (Ping timeout: 252 seconds)
19:29:22 Join lebellium [0] (
19:30:19saratogawodz: sure, but right now i think we only have one device that could even use it
19:43:23 Join TheLemonMan [0] (~lemonboy@unaffiliated/thelemonman)
19:46:02 Join xorly [0] (
19:49:00 Quit djukon (Ping timeout: 264 seconds)
19:50:20 Join djukon [0] (~djukon@gateway/shell/insomnia247/x-pckmpgmufucfwjmn)
20:04:18 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
20:14:51 Join JdGordon_ [0] (~jonno@rockbox/developer/JdGordon)
20:16:41 Quit JdGordon (Ping timeout: 272 seconds)
20:21:54 Quit RiD (Quit: A good plan today is better than a perfect plan tomorrow.)
20:23:56 Join prof_wolfff [0] (
20:38:12 Quit ZincAlloy (Ping timeout: 265 seconds)
20:40:01 Join wodz [0] (
20:42:29wodzpamaury, what offset?
20:44:35pamaurywodz: wait a sec
20:47:22wodzbinary loaded @ 0x80001000 looks pretty sane.
20:48:10pamauryyeah I think that's the right address
20:48:26pamauryI found strings which prove it's uCOS-II as you suggested
20:49:04pamauryI don't really know how this OS works but there are strings which suggest many libraries are included in this thing, I just don't know if there an actual notion of program/lib/userspace and so on
20:50:19 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
20:51:05wodzin general uCOS-II can load elf modules but I don't see anything like this. Other then this it is just regular RTOS
20:51:15 Quit JdGordon_ (Ping timeout: 255 seconds)
20:57:32pamaurymaybe it load non-elf modules ? anyway, I prefer to work on headergen first
21:02:27wodzgenerated code is much more optimized then in iriver e1xx firmware
21:05:15pamaurythe delay slot jump on mips are kind of mind stretching when you are not used to read mips code
21:12:12 Join eliudnir [0] (
21:16:35eliudnirHi, I am trying to use RockBoxUtility in arch linux on this ipod 5th gen (mounted) but the utility configuration GUI won't show the mountpoint - any suggestions?
21:17:00wodzpamaury: yeah.
21:17:17 Quit wodz (Quit: Leaving)
21:18:36 Join bertrik [0] (~quassel@rockbox/developer/bertrik)
21:20:52pamauryeliudnir: I think the released version of RBUtil doesn't always detect mountpoints correctly. I think you can try to enter it manually
21:20:54***Saving seen data "./dancer.seen"
21:21:43eliudnirpamaury: ah, do you happen to know how I can direct it to the right mountpoint? it's my first time using this tool
21:23:56pamauryeliudnir: I would need to check, I hardly use the tool since I'm a dev, I do things by hand ^^ Let me check
21:25:26 Join RiD [0] (
21:28:10pamauryhum, running version 1.4 indeed the mount point list is in a combobox and I can't enter a mountpoint by hand
21:28:35gevaertsThe diagnostics log might help
21:29:24pamauryeliudnir: can you paste ( the output of Help > Debug > System trace (something like this)
21:29:25eliudniryeah - the Help - Troubleshoot - System Info dialogue from within RockBoxUtility shows the player as mounted and np - but it is not showing in the dialogue for the configuration so I cannot select it :(
21:29:34eliudnirpamaury: yeah one sec
21:29:46pamauryand also the system information
21:30:34eliudnirsystem info:
21:31:38gevaerts"VID: 05ac PID: 1261, Apple Inc. iPod" is a classic, not a video
21:31:40pamauryeliudnir: where is you device supposed to be mounted ?
21:31:56gevaertsi.e. 6th gen
21:31:59eliudnirgevaerts: sorry, you are correct of course :)
21:32:09eliudnirpamaury: /mnt/ipod
21:32:53gevaertsRocbox Utility doesn't support those, but if you're slightly adventurous you can have a look at to get rockbox on it
21:32:59pamaury[configure.cpp:672 WARNING] mountpoint not writable, skipping: "/mnt/ipod"
21:33:35gevaertsWell, rockbox utility might support installing the main build, but not a bootloader
21:33:35eliudnirgevaerts: ah no easy sailing then I guess :)
21:33:55gevaertseliudnir: I don't know. Some people find it fairly easy, some get stuck...
21:34:12*gevaerts doesn't have a 6th gen ipod so he doesn't have personal experience with it
21:34:24eliudnirI've done a lot of similar firmware swaps and cyanogenmod etc. on phones and game consoles - but never tried on my ipod
21:34:29eliudnirrockbox is very tempting tho
21:34:46gevaertsRight, that should be well within your reach then :)
21:35:22eliudnirI hope so - i hate how it's tied to itunes atm :(
21:44:24 Quit bertrik (Remote host closed the connection)
21:46:43 Join [Franklin] [0] (~franklin@unaffiliated/franklin)
21:50:13 Quit JdGordon (Ping timeout: 265 seconds)
21:53:49 Join JdGordon [0] (
21:53:49 Quit JdGordon (Changing host)
21:53:49 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
21:54:28 Join bertrik [0] (~quassel@rockbox/developer/bertrik)
22:10:22 Join einhirn [0] (
22:11:58 Quit [Franklin] (Ping timeout: 265 seconds)
22:12:11 Join [Franklin] [0] (
22:12:33 Join ZincAlloy [0] (
22:15:22 Quit einhirn (Ping timeout: 264 seconds)
22:23:44eliudnirpamaury gevaerts : thanks to both of ya - all done and installed now :)
22:26:29[Franklin]eliudnir: congrats :)
22:31:26 Quit ii_ ()
22:33:27eliudnir[Franklin]: thanks! this is liberating :)
22:33:33eliudnirit's more than a paperweight! :)
22:33:49[Franklin]have fun ;)
22:48:23 Quit Marex (*.net *.split)
22:48:23 Quit ranmachan (*.net *.split)
22:49:50 Join Marex [0] (~Marex@
22:49:50 Join ranmachan [0] (
22:56:05 Join Markmaster [0] (
22:58:08 Quit eliudnir (Ping timeout: 250 seconds)
23:01:27 Quit [Saint] (Remote host closed the connection)
23:02:12 Quit pamaury (Ping timeout: 255 seconds)
23:04:54 Join [Saint] [0] (~saint@rockbox/staff/saint)
23:15:43 Quit ender` (Quit: drug, n: A substance that, injected into a rat, produces a scientific paper.)
23:20:57***Saving seen data "./dancer.seen"
23:26:08 Quit amayer (Quit: Leaving)
23:34:01 Quit lebellium (Quit: ChatZilla [Firefox 37.0/20150319212106])
23:40:26 Quit TheLemonMan (Quit: leaving)

Previous day | Next day