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).

Notice: Only Gecko based browsers prior to FF4 support the multipart/mixed "server push" method used by this log reader to auto-update. Since you do not appear to use such a browser, this page will simply show the current log, and not automatically update.

#rockbox log for 2017-02-25

00:09:10 Quit ParkerR (Quit: ZNC -
00:09:58 Join ParkerR [0] (
00:09:58 Quit ParkerR (Changing host)
00:09:58 Join ParkerR [0] (ParkerR@unaffiliated/parkerr)
00:11:42 Join alexweissman [0] (
00:16:17 Quit alexweissman (Ping timeout: 268 seconds)
00:17:20 Quit fujisan (Quit: Textual IRC Client:
00:19:56 Join alexweissman [0] (
00:24:13 Quit lebellium (Quit: ChatZilla 0.9.93 [Firefox 51.0.1/20170125094131])
00:36:55 Quit pamaury_ (Ping timeout: 240 seconds)
00:40:51 Quit MrZeus2 (Ping timeout: 255 seconds)
00:45:39 Quit paulk-collins (Quit: Leaving)
01:04:15 Quit jhMikeS (Ping timeout: 255 seconds)
01:07:43***Saving seen data "./dancer.seen"
01:10:44 Quit dys (Ping timeout: 260 seconds)
01:12:17 Quit skapazzo (Quit: leaving)
01:27:17 Join krabador [0] (~krabador@unaffiliated/krabador)
01:31:13 Quit ender` (Quit: You have the right to remain silent. Anything you say will be misquoted, then used against you.)
01:53:42 Quit ZincAlloy (Quit: Leaving.)
02:00:00 Join gk_-1wm_- [0] (~su@
02:31:24 Quit gk_-1wm_- (Excess Flood)
02:34:02 Quit bastian__ (Quit: bastian__)
03:07:47***Saving seen data "./dancer.seen"
03:13:40 Quit alexweissman (Remote host closed the connection)
03:17:29 Quit krabador (Remote host closed the connection)
03:23:12 Join gk_-1wm [0] (~su@
03:23:13 Part gk_-1wm
05:07:48***Saving seen data "./dancer.seen"
05:28:13 Join jhMikeS [0] (
05:29:25 Quit Ruhan (Quit: Connection closed for inactivity)
06:12:01 Join shmibs_ [0] (
06:12:18 Join pixelma_ [0] (~pixelma@rockbox/staff/pixelma)
06:13:38 Join benedikt93_ [0] (~quassel@unaffiliated/benedikt93)
06:14:00 Join Strife1989 [0] (
06:14:09 Join kugel_ [0] (~kugel@rockbox/developer/kugel)
06:14:26 Join Marex_ [0] (~Marex@
06:14:28 Join mikroflops_ [0] (~yogurt@
06:14:31 Join rasher_ [0] (
06:14:38 Quit gevaerts (Disconnected by services)
06:14:42 Join mc2739_ [0] (~mc2739@rockbox/developer/mc2739)
06:14:45 Quit mc2739 (Disconnected by services)
06:14:46 Join gevaerts_ [0] (~fg@rockbox/developer/gevaerts)
06:14:47 Nick mc2739_ is now known as mc2739 (~mc2739@rockbox/developer/mc2739)
06:16:40 Quit fs-bluebot (Ping timeout: 260 seconds)
06:16:40 Quit munch (Ping timeout: 260 seconds)
06:16:40 Quit man_in_shack (Ping timeout: 260 seconds)
06:16:40 Quit Petri152 (Ping timeout: 260 seconds)
06:16:40 Quit Birch (Ping timeout: 260 seconds)
06:16:41 Quit kugel (Ping timeout: 260 seconds)
06:16:41 Quit rasher (Ping timeout: 260 seconds)
06:16:41 Quit Marex (Ping timeout: 260 seconds)
06:16:41 Quit Rower (Ping timeout: 260 seconds)
06:16:41 Quit pixelma (Ping timeout: 260 seconds)
06:16:42 Quit benedikt93 (Ping timeout: 260 seconds)
06:16:45 Quit mikroflops (Ping timeout: 240 seconds)
06:16:45 Quit Strife89 (Ping timeout: 240 seconds)
06:16:45 Quit gluytium_ (Ping timeout: 240 seconds)
06:16:46 Quit shmibs (Ping timeout: 240 seconds)
06:16:46 Quit tomflint (Ping timeout: 240 seconds)
06:16:46 Quit alexbobp (Ping timeout: 240 seconds)
06:16:51 Nick rasher_ is now known as rasher (
06:16:51 Nick shmibs_ is now known as shmibs (
06:17:12 Join alexbobp [0] (
06:17:28 Join fs-bluebot [0] (
06:17:34 Join Rower [0] (
06:17:35 Nick alexbobp is now known as Guest24161 (
06:17:37 Join Petri152 [0] (
06:17:40 Join man_in_shack [0] (~chat@unaffiliated/man-in-shack/x-4279753)
06:18:22 Join TorC [0] (~TorC@fsf/member/TorC)
06:19:01 Join tomflint [0] (~tomflint@unaffiliated/tomflint)
06:19:10 Join gluytium [0] (
06:21:01 Join munch` [0] (pls@gateway/shell/elitebnc/x-ldxtvsavwglssnyn)
06:36:49 Quit pixelma_ (Quit: No Ping reply in 120 seconds.)
06:38:28 Join pixelma [0] (~pixelma@rockbox/staff/pixelma)
06:41:15 Quit fs-bluebot (Ping timeout: 260 seconds)
06:41:15 Quit man_in_shack (Ping timeout: 260 seconds)
06:42:24 Join man_in_shack [0] (~chat@unaffiliated/man-in-shack/x-4279753)
06:42:26 Quit munch` (Ping timeout: 240 seconds)
06:42:27 Quit tomflint (Ping timeout: 240 seconds)
06:42:34 Join tomflint [0] (~tomflint@
06:42:34 Quit tomflint (Changing host)
06:42:34 Join tomflint [0] (~tomflint@unaffiliated/tomflint)
06:43:31 Join fs-bluebot [0] (
06:46:02 Join munch [0] (pls@gateway/shell/elitebnc/x-nldzxizmxfxcvytt)
06:48:21 Quit TheSeven (Ping timeout: 240 seconds)
06:49:27 Join TheSeven [0] (~quassel@rockbox/developer/TheSeven)
07:02:26 Quit furrywolf (Ping timeout: 240 seconds)
07:07:49***Saving seen data "./dancer.seen"
07:16:49 Quit pixelma (Quit: No Ping reply in 120 seconds.)
07:18:38 Join pixelma [0] (~pixelma@rockbox/staff/pixelma)
07:35:25 Join dys [0] (
08:10:00 Join alexweissman [0] (
08:19:26 Quit bluebrother (Ping timeout: 240 seconds)
08:19:41 Quit fs-bluebot (Ping timeout: 240 seconds)
08:21:10 Join bluebrother [0] (~dom@rockbox/developer/bluebrother)
08:33:37 Join fs-bluebot [0] (
08:35:48 Quit pixelma (Quit: .)
08:35:48 Quit amiconn (Quit: - Chat comfortably. Anywhere.)
08:35:59 Join pixelma [0] (~pixelma@rockbox/staff/pixelma)
08:36:01 Join amiconn [0] (~amiconn@rockbox/developer/amiconn)
08:53:02 Quit alexweissman (Remote host closed the connection)
08:56:33 Join alexweissman [0] (
08:58:54 Join paulk-collins [0] (
08:59:42 Quit alexweissman (Remote host closed the connection)
09:07:51***Saving seen data "./dancer.seen"
09:36:05 Join TheLemonMan [0] (~root@irssi/staff/TheLemonMan)
09:39:37 Join johnb2 [0] (
09:47:36 Join parchd [0] (~parchd@unaffiliated/parchd)
10:02:21 Quit johnb2 (Ping timeout: 240 seconds)
10:04:22 Join ken [0] (65e78a5c@gateway/web/freenode/ip.
10:04:45 Nick ken is now known as Guest71210 (65e78a5c@gateway/web/freenode/ip.
10:07:10 Quit Guest71210 (Client Quit)
10:09:01 Join johnb3 [0] (
10:09:21 Join ender` [0] (
10:09:31 Quit johnb3 (Client Quit)
10:10:14 Join alexweissman [0] (
10:13:22 Join lebellium [0] (
10:14:51 Quit alexweissman (Ping timeout: 260 seconds)
10:15:32 Quit Senji (Read error: Connection reset by peer)
10:15:49 Join Senji [0] (~Senji@
10:23:36 Join wodz [0] (
10:25:25wodzpamaury: ping
10:35:25 Quit amiconn (Quit: No Ping reply in 64 seconds.)
10:36:38 Join amiconn [0] (~amiconn@rockbox/developer/amiconn)
10:38:41 Quit parchd (Ping timeout: 240 seconds)
10:44:57lebelliumwodz: pamaury what's that "Wifi service"?!
11:04:34 Join pamaury_ [0] (~pamaury@rockbox/developer/pamaury)
11:07:53***Saving seen data "./dancer.seen"
11:10:54 Join parchd [0] (~parchd@unaffiliated/parchd)
11:11:39wodzlebellium: Afaik it is leftover from SDK
11:12:01wodzpamaury_: ping again
11:13:15pamaury_wodz: pong
11:14:36wodzpamaury_: I don't quite understand framebuffer configuration in nwz. nwzlinux.h claims pixel format 565 i.e 16bit is it really the case?
11:15:53pamaury_yes iirc
11:16:09pamaury_but on nwz, the framebuffer driver reports bullshit info anyway
11:16:25wodzpamaury_: also if framebuffer is mmap'ed do you really need to copy from rockbox's internal fb? Can't you simply access linux fb directly?
11:19:02pamaury_wodz: I could but then the screen would update without lcd_update which is wrong
11:19:14pamaury_because on NWZ, some LCDs refresh periodically
11:19:27wodzpamaury_: ah, ok.
11:20:45pamaury_my understanding of Sony's driver is that you can only select which buffer to use for redraw (and potentially set other info like alpha etc)
11:22:42wodzis there a way to not allocate rb internal framebuffer currently?
11:24:01wodzpamaury_: ^
11:26:44pamaury_I don't remember
11:28:07pamaury_wodz: but it's unclear to me because in lcd.h there is lcd_static_framebuffer which seems to be implemented by drivers/
11:28:33 Nick mikroflops_ is now known as mikroflops (~yogurt@
11:30:05pamaury_but in general I think you want to have two framebuffers: rockbox's static framebuffer and the framebuffer holding the data (being) sent to the lcd
11:30:22 Join skapazzo [0] (~skapazzo@
11:36:24pamaury_wodz: just so you know, most of the bluetooth services translate to calls to bt-adapter, so it's probably possible to not use sys_server at all
11:37:50wodzpamaury_: I know, but figuring out actual sequence may be problematic. I am wondering how usb insert/remove is signaled btw.
11:39:36 Join ZincAlloy [0] (~Adium@2a02:8108:8b80:1700:e49b:eb5a:835b:3c46)
11:41:15pamaury_I don't know, probably some dev/sys file that you need to poll. This does not seem to be handled by sys_server
11:49:56 Quit TheLemonMan (Remote host closed the connection)
11:50:19 Quit girafe (Read error: Connection reset by peer)
11:55:03pamaury_wodz: maybe look in /sys/devices/platform/ if you have a usb driver that exposes some info
11:55:20pamaury_possible in /proc/ as well
11:55:54wodzpamaury_: It seems it is somewhat modern system and most of the stuff is in /sys
11:55:55pamaury_I think linux has a hotplug detection mecamism in userspace, also there is inotify if you now that a particular file is create in /sys on usb plug
11:59:25wodzpamaury_: there is also dbus daemon running but don't know if it is used actually for anything
11:59:33 Quit pamaury (Remote host closed the connection)
12:01:03 Nick pamaury_ is now known as pamaury (~pamaury@rockbox/developer/pamaury)
12:20:33 Join TheLemonMan [0] (~root@irssi/staff/TheLemonMan)
12:34:15 Nick gevaerts_ is now known as gevaerts (~fg@rockbox/developer/gevaerts)
13:02:51 Join johnb2 [0] (
13:07:57***Saving seen data "./dancer.seen"
13:34:41 Quit amiconn (Quit: No Ping reply in 64 seconds.)
13:36:09 Join amiconn [0] (~amiconn@rockbox/developer/amiconn)
14:01:05 Join girafe [0] (
14:01:51 Join Ruhan [0] (uid76353@gateway/web/
14:10:04 Nick Marex_ is now known as Marex (~Marex@
14:33:45 Join Bilgus [0] (~Bilgus@gateway/tor-sasl/bilgus)
14:41:56BilgusjhMikeS, can you have a look at #g1562 when you get a chance
14:41:58fs-bluebot3Gerrit review #1562 at : 3Add DISABLE_O_CREAT define to allow removal of file creation codepath by William Wilgus
14:43:58Bilgusthough #g1572 should be universal I'm pretty sure the bootloader only uses the default codepages if someone can comfirm maybe we can just use it across all targets
14:43:59fs-bluebot3Gerrit review #1572 at : 3Add DISABLE_UNICODE_LOAD_CODEPAGES define to allow removal of load CodePage codepath by William Wilgus
14:47:51 Quit johnb2 (Ping timeout: 260 seconds)
14:53:39 Join johnb2 [0] (
15:01:09 Quit johnb2 (Quit: Nettalk6 -
15:03:31 Join Alexeyan [0] (5de5003c@gateway/web/freenode/ip.
15:07:59***Saving seen data "./dancer.seen"
15:15:51 Quit pixelma (Quit: No Ping reply in 120 seconds.)
15:17:38 Join pixelma [0] (~pixelma@rockbox/staff/pixelma)
15:34:21 Quit parchd (Ping timeout: 240 seconds)
15:50:04 Join Glitches [0] (
15:50:26GlitchesHello, I have a .rock file, how do I run it if I don't have a physicil device to setup rockbox?
15:50:53Glitchesor how do I analyze it, rather? are there debugging tools?
15:50:58GlitchesHow are these made?
15:55:40pamauryGlitches: just look at the code ?
15:56:06pamauryyou can't run them outside of rockbox, these are regular executable
15:56:06Glitchespamaury: this looks just like a binary file, can't make any sense of it
15:56:10pamaury*are not
15:56:22pamauryGlitches: the source code is in our repository of course
15:56:35Glitchespamaury: but what if you didn't write the application I'm analyzing?
15:57:06pamaurythat seems unlikely, I'm not aware of any out of tree plugin for rockbox
15:58:21Glitchespamaury: well, it's a hacking competition and they have apperantly made their own application for rockbox
15:59:27pamauryI see, well you can still look at our code to understand the format of the executable
15:59:30Glitchespamaury: but, so you're telling me that I need to setup rockbox on a physical device to be able to run this application?
15:59:31pamauryand then use a static analyzer
15:59:51Glitchespamaury: are there static analyzers for this out there?
16:00:02pixelmait may run in a simulator
16:00:05pamauryGlitches: to run it yes, unless it was complicated for a simulator build that runs on a PC
16:00:10Glitchespixelma: oh? is there one?
16:00:41Glitchespamaury: so you're saying that there is a simulator?
16:00:45pamaurywe have a simulator but it's not a simulator in the sense you imagine
16:00:59Glitchespamaury: well, I would like to try it out, where is it?
16:01:54pamaurythere isn't one simulator, each rockbox target can be turned into a simulator where all the kernel calls are essentially reimplemented on top on linux/windows
16:02:09Glitchesoh, I se
16:03:06pamaurybest way to get an emulator is to build it yourself on Linux. If you are running Windows, I think rasher provides some prebuilt images
16:04:42Glitchespamaury: how much time and knowledge about rockbox would that take?
16:05:16pamaurybuilding an emulator is easy on Linux if you are at least a bit familiar with command line
16:07:53Glitchespamaury: really?
16:08:24wodzuisimulator uses elf for plugins, no? Should be easy to check
16:08:38Glitchespamaury: sorry, my internet died, what did you say?
16:09:42 Join parchd [0] (~parchd@unaffiliated/parchd)
16:10:01Glitcheswodz: can you possibly run .rock binaries in that?
16:10:04pamaury explains how to build a simulator. But I am a bit confused about this hackking competition that you mentioned. Isn't it a bit strange to get an executable and no way to run it ?
16:10:38Glitchespamaury: that's kinda the point, you're just expected to figure shit out
16:10:48pamaurybut figure what exactly ?
16:11:01wodzGlitches: if this is for hacking competition, don't you think you should figure it out yourself?
16:11:09Glitchespamaury: how to run it, and if it's exploitable
16:11:28Glitcheswodz: yea, but I thought I would drop by here and ask if there was an emulator
16:11:50wodzmy bet is that it is plugin for simulator.
16:12:09wodzas I said afaik this should be elf file in this case
16:12:20Glitcheswodz: but I'm not given an elf file
16:13:03pamauryGlitches: you can easily check if a file is an ELF file by looking at the first bytes. If it contains the string ELF at the very beginning then it's an elf file
16:13:26Glitchespamaury: I have checked, this is raw binary data
16:14:28pamauryyes but ELF file start with the bytes 7f 45 4c 46 with correspond to the string ELF. You can check that with any decent hex editor
16:14:35wodzyeah, rocks for uisimulator are elfs
16:14:52wodzor simply file :-)
16:15:05wodzlamp.rock: ELF 64-bit LSB shared object, x86-64, version 1 (SYSV), dynamically linked, BuildID[sha1]=1d7ea01d3d986e7a743558cc98c9eaa981b5cd43, not stripped
16:16:04Glitcheswodz: well I was given that file as well, but...
16:16:10Glitches"lamp.rock: data"
16:17:01wodzGlitches: It means it is plugin to be run on embedded hardware
16:17:40Glitcheswodz: So I need to figure out what hardware, and then I need to get said hardware?
16:17:42wodzGlitches: you need to figure out what arch it is :-) Could be mipsel, arm, coldfire, sh
16:18:04wodzGlitches: lamp is small, I suppose you should figure out by static analysis only
16:19:17wodzanyway having hardware to run is not that helpful unless it is hacked to have some form of debug interface wired (as jtag or bdm)
16:19:37Glitcheswodz: right, so I'd need some emulator that runs that architecture?
16:20:11Glitcheswodz: but how do I statically analyze this? is there a tool or do I just need to side-by-side this with the manual fro a bunch of instruction sets?
16:20:34wodzGlitches: Have you heard about disassemblers?
16:20:49Glitcheswodz: I have, is there one for this?
16:21:15wodzgood god
16:21:28Glitchesright, so they can figure out that part?
16:23:11Glitcheswodz: but i'm still not following, does a disassembler not need to know what instuction set was used?
16:23:18wodzrock is header + raw binary, no relocations. When you figure out what arch is this you strip the header and feed it to disasembler. Then you have a fun to make sense of output you get.
16:23:51wodzGlitches: You are responsible of figuring out what arch it is but the choice is limited as I said
16:24:02pamauryyou have to try all architectures until you find the right one
16:24:10Glitchespamaury: ok, I see
16:24:20 Join smoke_fumus [0] (
16:24:36wodz+ arm can be both thumb or arm, still rather limited subset of arch on earth
16:25:08 Quit smoke_fumus (Client Quit)
16:26:41Glitcheswodz: yea, you're right
16:28:36wodzGlitches: can you provide the file? I am mildly interested what it actually is.
16:28:54Glitcheswodz: sure
16:31:53wodzGlitches: Ah you have full build not separated .rock? That makes things much easier. Hint rockbox.sansa file
16:32:06 Quit Moarc (Quit: i znowu NADMUCHAŁ BALONA)
16:32:20wodzor rockbox-info.txt
16:33:15Glitchesoh, I see
16:34:42wodzgood luck finding e200v2 :P
16:35:49Glitcheswodz: I highly doubt I actually need the hardware, there must be something I can gather from this
16:35:58 Join Moarc [0] (~chujko@
16:36:54wodzI think so too
16:37:52Glitcheswodz: lol, do you think this is usable?
16:38:28Glitchesit's not v2 tho
16:38:50wodzGlitches: it is very old and never fully working
16:39:02Glitchesyea, that's what I thought
16:39:49wodzGosh, should I hack the competition server to figure out what the hacking tasks are? Are there additional scores for knowing what you are supposed to do?
16:47:28 Join Bilgus_ [0] (~Bilgus@gateway/tor-sasl/bilgus)
16:48:05Glitcheswodz: ikr, this is so ambigous
16:49:42 Quit parchd (Ping timeout: 255 seconds)
16:50:18 Quit Bilgus (Ping timeout: 240 seconds)
16:50:24Alexeyanv2 has a different cpu so it shoudn't work..
16:50:54GlitchesAlexeyan: I was afraid of that
16:51:45AlexeyanThis challenge...
16:56:46 Join alexweissman [0] (
17:02:29 Join MrZeus1 [0] (~MrZeus@2a02:c7f:7018:6200:95c3:4473:850f:f22d)
17:02:43wodzpamaury: Do I think correctly that I need to reconfigure fb for 16 or 24 bpp for rockbox?
17:08:00***Saving seen data "./dancer.seen"
17:08:16jvoisin - Rockbox related
17:08:53jvoisinalso, hello Glitches ;)
17:10:01wodzIs there description of the task (without registering and stuff)?
17:27:09Glitcheswodz: yea, hang on
17:27:42Glitcheswodz: "Because I thought that you may be bored while listening to your music, I wrote a small application for your favorite jukebox, have fun!"
17:27:47Glitchesnot helpful at all...
17:32:01 Join furrywolf [0] (
17:40:27pamaurywodz: assuming the framebuffer supports it then yes
17:41:30pamaurymaking look at the code to see, or try to change the framebuffer settings, see what the ioctl returns
17:46:28 Quit pamaury (Remote host closed the connection)
17:46:53 Join pamaury [0] (~pamaury@rockbox/developer/pamaury)
17:49:48wodzpamaury: Just tried and fb only accepts 24 and 32 bpp BUT 24 bpp doesn't change mem layout so basically I am stuck at 32bpp
17:51:20pamaurywhat do you mean it doesn't change mem layout ?
17:51:29wodzah, but it looks like rockbox can be configured to support 32bpp
17:52:41wodzpamaury: it should change smem_length and line_length, no?
17:53:04 Quit ZincAlloy (Quit: Leaving.)
17:53:09pamaurywodz: don't necessarily trust those values
17:53:19pamauryon many framebuffer drivers they are wrong
17:53:37pamauryI would say just try to set 24-bit and see if it works as expected
17:53:54wodzah, and setting bits_per_pixel to 24 and rereading gives back 32
17:54:51pamaurythen maybe the driver only supports 32-bit
17:55:16wodzif rockbox can support 32bpp (and looking at lcd.h it should) it would be easier to not touch this
17:57:14pamauryyeah probably
17:58:00wodzapparently we don't have any single device with LCD_DEPTH == 32
17:58:51pamaurywodz: that means you may have to implement some functions
17:59:10CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
17:59:10*wodz grumbles
17:59:29pamaurybasically lcd-32bit.c, the same way we have lcd-24bit.c and lcd-16bit.c
18:00:08pamauryit's mostly copy-paste + fixes
18:00:41*pamaury suspects we have too much copy-paste in lcd drivers
18:02:18pamaurythe most important fixes will probably be in lcd.h to pack/unpack
18:03:33wodzor I can write lcd driver to perform extension 24->32. top 8 bits is alpha anyway
18:03:50wodzit will be slower then plain memcpy but easier
18:04:12pamauryyeah you can do that for starters, that's what I did for the ZEN until we had proper support for 24-bit framebuffer
18:04:40wodzThe screen is small, just 128x160.
18:04:41*pamaury finds it amusing that the Rocker has 32-bit per pixel given the terrible screen it has
18:05:49wodzOut of curiosity I must check if it really supports alpha
18:07:19pamauryit may actually support alpha if IPU is enabled so it can blend with video for example
18:11:47wodznop, driver itself just ignores top 8 bits
18:14:37lebelliumI just read the logs and didn't understand anything about this hacking challenge. That was a weird discussion :/
18:14:51 Join johnb2 [0] (
18:17:05wodzlebellium: I was contacted by author of this challenge and his description makes sense :-)
18:18:40wodzpamaury: I don't understand one thing - your test_fb operates on 32bpp. Do you reconfigure framebuffer on nwz?
18:20:13pamaurywodz: my test_fb operates on 24-bit
18:20:24lebelliumwodz: are the challengers given a .rock file? What are they supposed to do with it?
18:20:26pamaurythe code is simply doing 32-bit read/writes at the end
18:20:33pamaurybut each pixel is 24-bit
18:20:55wodzpamaury: but writes to mmap'ed fb memory are 32bit
18:21:17pamauryactually iirc my code can handle any pixel arrangement as reported by the kernel
18:21:35pamauryyes that's what I said, it does 32-bit read/write but masks everything properly
18:21:47pamauryit could support 17-bit per pixel even
18:22:07pamaurylook at put_pix and put_pix_mask
18:23:36*pamaury doesn't guarantee his code is correct in all cases
18:23:36wodzpamaury: I understand that part. But in rockbox you claim 24bpp and use lcd_copy_framebuffer_rect(). How it works to extend 24bit pixel into 32bit written to fb?
18:24:03pamauryno the framebuffer is 24-bit per pixel
18:24:53wodzI am lost. You write 32bits per pixel to framebuffer, no?
18:25:14pamauryno, I do 32-bit accesses to make the code simpler
18:25:38pamauryI didn't write test_fb for rockbox originally, it supports any pixel configuration
18:25:51pamauryand the easiest way to handle any configuration is to do 32-bit accesses
18:33:33pamauryso I am not getting any closer to fix audio on nwz but at least I know that noise cancelling is handled by the codec and seems simple enough to enable
18:34:09lebelliumNoise cancelling is the main feature
18:34:12lebelliumwe don't need sound
18:34:53pamauryit cancels everything, even sound :)
18:37:20lebelliumso after looking more deeply the issue, you still don't understand where it comes from or you actually didn't have time yet to look at it more deeply?
18:38:24pamauryI'm disassembling the codec driver to understand why acoustic mode somehow fixes the issue but makes sound too loud
18:39:23lebelliumah yes this strange acoustic mode
18:39:37lebelliumsounds like marketing bullshit
18:53:44 Quit amiconn (Quit: No Ping reply in 64 seconds.)
18:53:58 Join amiconn [0] (~amiconn@rockbox/developer/amiconn)
18:55:43 Join webguest044 [0] (
18:56:48 Quit webguest044 (Client Quit)
19:04:10 Quit johnb2 (Ping timeout: 268 seconds)
19:08:01***Saving seen data "./dancer.seen"
19:20:34 Quit wodz (Quit: Leaving)
19:48:12 Join polytux [0] (~oscar@2a02:7aa0:1619::cea:62cd) do I manage a 256 gb iPod 6th generation in Windows? it will not mount when connected, and transferring all my music everytime I need to sync it... :(
19:50:57 Join dipus [0] (
20:08:28 Quit Alexeyan (Ping timeout: 260 seconds)
20:11:55pamauryprobably prof_wolfff knows
20:13:26pamaurywodz (log): so looking at the audio table and confirming on target, I suspect that actually, in normal mode (ie acoustic=cuerev=0), the DAC runs at 0dB and then the volume controls HP volume. Thus it saturates.
20:14:03pamaurywhen acoustic mode is set, DAC volume is set to a fixed value (don't know exactly much), HP volume table is the same as in normal => no more saturation
20:14:35pamaurywhen cuerev is set, DAC volume is set the minimal possible value, and HP volume stays the same => even lower volume
20:14:45pamauryactually with cuerev, the volume is much more acceptablr
20:23:18 Join ZincAlloy [0] (~Adium@2a02:8108:8b80:1700:8caa:867d:cc24:627a)
20:27:05 Join johnb2 [0] (
20:43:18 Join wodz [0] (
20:44:14wodzpamaury: maybe I am missing something but if accustic mode attenuates DAC signal and volume tables controlling HP are the same how it is louder?
20:47:41 Quit johnb2 (Ping timeout: 240 seconds)
20:49:05pamaurywodz: I've check with the device, normal mode actually sounds louder, just completely distorted
20:49:53pamauryor not, I don't know it's obvious
20:50:16wodzpamaury: Didn't you say that accustic mode makes volume control useless as it is very loud?
20:50:16pamauryI mean it's hard to compare two signals when one of them is garbage
20:51:06pamaurybut in normal mode, it clearly sounds like it's clipping
20:51:29*pamaury double checks the tables
20:53:06wodzAnyway for me it is hard to imagine that DAC at 0dB causes clipping.
20:56:27pamauryok sorry I got it wrong, let's give some figures for lowest volume:
20:56:27pamaurynormal mode (dac/hp): 0x17 / 0xc3
20:56:27pamauryacoustic mode: 0x00 / 0xc9
20:56:27DBUGEnqueued KICK pamaury
20:56:27pamaurycuerev mode: 0xff / 0xc3
20:56:27pamauryso in fact in normal mode, the DAC is run at +x dB (whatever 0x17 means)
20:56:28***Alert Mode level 1
20:56:28pamauryin acoustic, it's 0dB
20:56:30pamauryin cuerev, it's the lowest (0xff is negative ?)
20:56:57pamauryacoustic may still sound louder if 0x17 + 0xc3 < 0x00 + 0xc9 (when adding the equivalent dB values)
20:57:13pamaurybut normal mode may clip while still being less loud
21:02:03 Quit dipus (Quit: WeeChat 1.6)
21:02:34pamauryhere are the raw numbers:
21:02:34pamaury[21:01:41][pamaury@lap-54-71:/home/pamaury/project/rockbox/sony_nwz]$ ./audio_table -f off -a off -c off -n off -o hp -i music -p dac,hpout
21:02:34pamauryfr=off in=music cr=off ac=off out=hp nc=off
21:02:34***Alert Mode level 2
21:02:34pamauryDAC: 33 17 17 17 17 17 17 17 17 17 17 17 17 17 17 17 17 17 17 17 17 17 17 17 17 17 17 17 17 18 18 18
21:02:34***Alert Mode level 3
21:02:34pamauryHPOUT: c3 c4 c7 ca cd d0 d3 d6 d9 dc df e2 e4 e6 e8 ea ec ee f0 f2 f3 f4 f5 f6 f7 f8 f9 fa fb fc fe fe
21:02:41pamaury(that's normal mode)
21:02:48pamaury[21:01:58][pamaury@lap-54-71:/home/pamaury/project/rockbox/sony_nwz]$ ./audio_table -f off -a on -c off -n off -o hp -i music -p dac,hpout
21:02:49pamauryfr=off in=music cr=off ac=on out=hp nc=off
21:02:49pamauryDAC: 33 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 04 08 0c 0e 10 12 14 16 18 18 18 18 18 18 18
21:02:49***Alert Mode level 4
21:02:49pamauryHPOUT: c9 cc d0 d6 dc e0 e3 e6 e9 ec ef f2 f4 f6 f8 fa fc fc fc fc fc fc fc fc fc fc fc fc fc fd fe fe
21:02:49***Alert Mode level 5
21:02:49pamaury(acoustic mode)
21:02:56pamaury[21:02:02][pamaury@lap-54-71:/home/pamaury/project/rockbox/sony_nwz]$ ./audio_table -f off -a off -c on -n off -o hp -i music -p dac,hpout
21:02:56pamauryfr=off in=music cr=on ac=off out=hp nc=off
21:02:56pamauryDAC: 33 ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff 00 00 00
21:02:58pamauryHPOUT: c3 c4 c7 ca cd d0 d3 d6 d9 dc df e2 e4 e6 e8 ea ec ee f0 f2 f3 f4 f5 f6 f7 f8 f9 fa fb fc fe fe
21:03:00pamaury(cuerev node)
21:06:42 Quit amiconn (Quit: No Ping reply in 64 seconds.)
21:07:23 Join amiconn [0] (~amiconn@rockbox/developer/amiconn)
21:08:05***Saving seen data "./dancer.seen"
21:08:56jvoisinnon-registered users are now allowed to see the challenges :)
21:09:18pamaurythis is the audio table codec, for those interested: pamaury/2ffc57d955e6bcf3702dfab79b690722">
21:09:53TheLemonManjvoisin, neet challenge :)
21:11:09polytuxpamaury: ok, thanks! prof_wolfff seems afk but I'm not in any rush
21:12:35jvoisinaw, TheLemonMan <3
21:12:50***Alert Mode OFF
21:13:46lebelliumjvoisin: it still asks me team name and password to see the challenges
21:14:20jvoisinit works here™
21:14:52lebelliumI checked in another browser to be sure it's not a cache issue
21:15:13jvoisinthis is weird.
21:17:02lebelliumoh it's OK now
21:27:48 Quit jhMikeS (Ping timeout: 255 seconds)
21:34:00 Part Glitches
21:39:46lebelliumNo more glitches pamaury. Clean sound
21:40:23pamaurylebellium: you mean you changed the code ?
21:40:35lebelliumI mean glitches left
21:40:44pamaurywith cuerev ?
21:40:47lebelliumsorry for the poor joke
21:40:53lebellium[21:34:03]<−−|Glitches has left #rockbox
21:41:25pamaurygood one, I completely missed it
21:44:17lebelliumotherwise it is very much unlikely I can improve any of your code
21:47:25pamauryI think I will switch the nwz code to set cue/rev since it seems to solve the problem, at least for now
21:47:39lebelliumdo you know what fr=off is?
21:47:46lebelliuma special setting for France?
21:50:44pamauryfr corresponds to volume regulation I believe (note that I did not include the fr1 table in the file because I'm lazy)
21:51:07pamauryreading sony's code over and over, I have been lead to believe that volume regulation was previously refered to as france setting
21:51:32lebelliumit's always ambiguous because I think that's a European regulation
21:51:57lebelliumbut maybe at first it was only in France
21:56:47pamaurywodz: lebellium: I have update g#1481 if you want to try it
21:56:49fs-bluebot3Gerrit review #1481 at : 3Initial commit for the Sony NWZ linux port (WIP) by Amaury Pouly
21:57:23pamaurythe new code now sets cue/rev to 1 by default, and there is a debug entry in System > Debug > Hardware Info > audio
21:57:42lebelliumI'll try on A10
21:57:46pamauryuse select to toggle, and up/down to switch between cue/rev and acoustic setting.
21:58:16pamauryAs far as I can tell, when both cue/rev and acoustic are set, it acts as if acoustic was 0 (ie cue/rev has priority)
22:10:03lebelliumpamaury: still the *Panic* control 'playback volume" on the A10
22:11:16pamaurylebellium: did you keep your old settings ?
22:11:36lebelliumpamaury: I overwrote the existing rockbox folder
22:11:44pamaurytry erasing it and do a fresh install
22:13:35lebelliumpamaury: doesn't help
22:14:00lebelliumthe panic comes directly when starting Rockbox
22:15:05pamauryah yes that's on the A10 ?
22:15:20pamauryI forgot the fix the volume thing were left and right are merge
22:15:39lebelliumotherwise I can try on another device
22:16:01pamaurytry it on the E580, in the mean time I will fix the code
22:25:43pamaurylebellium: I updated the task, hopefully it should fix A10
22:25:53lebelliumtrying on E580 for now
22:27:02lebelliumI agree
22:27:16lebelliumof cue=1, it doesn't matter if acoustic=0 or 1
22:28:12pamaurydoes it sound more reasonable with cuerev ? At the lowest volume (1), with cuerev sound is reasonably low. Still after level 5 or 6 it gets really loud but that's a clear improvement, I can actually listen to the device now
22:29:30lebelliumit's indeed better
22:29:34lebelliumnot much progressive volume
22:29:42lebelliumbut volume 1 is quiet enough
22:30:22lebelliumbut sound is still distorded
22:30:40pamaury? with cuerev ?
22:32:21lebellium6dB is reasonable volume with my IEM but sound is crappy
22:32:37pamauryit's not in dB :-p
22:32:46lebelliumless than with cue=0 but still very crappy
22:33:00pamauryon mine sound is not distorted, either in acoustic or cuerev mode.
22:33:34lebelliumI tried with Hello from Adele
22:33:43lebelliumon the quiet moment it's OK
22:33:46pamaurydid you change anything using the debug menu ?
22:33:50lebelliumbut on the main chorus it's terible
22:33:56lebelliumwith the debug menu
22:33:59lebelliumthat's how I test
22:34:03pamauryand if you don't change anything ?
22:34:09pamaurynow the default when it boots is cuerev=1
22:34:38lebelliumI just turned off and on again
22:34:58lebelliumsound is terrible at 6 dB (or whatever it is)
22:35:38pamauryon E580 ?
22:36:22lebelliumyes, let me try several songs
22:39:10pamauryand what about acoustic ?
22:39:22lebelliumconfirmed at least with FLAC files (Eminem and Adele sounding good with OF)
22:39:49lebelliumacoustic=1 is worse since it's louder
22:40:18lebelliumah no sorry, it's acoustic=0
22:40:24pamaurylouder doesn't mean more clipping though
22:40:43pamauryWe all agree that acoustic=0 and cuerev=0 is terrible because of clipping
22:40:51lebelliumwith acoustic=1 there is less clipping than with cuerev=1
22:41:18pamaurynow on my E580, acoustic=1 and cuerev=0, I have no clip but too loud sound
22:41:18pamaurywith acoustic=0 and cuerev=1, I have no clip and less loud sound
22:42:34lebelliumI think you do have clipping but you don't listen to the proper samples to hear it :P Do you want me to upload a file?
22:43:37pamauryit might depend on the headphones as well
22:43:52pamaurywhy not
22:44:27pamauryit may also depend on the file's own volume. Depending on whether on high the maximum volume in the file is, it may or may not clip
22:44:46pamaurybut you say you have clipping in acoustic mode ?
22:46:18lebelliumin acoustic mode it's louder but no clipping
22:47:55 Quit MrZeus1 (Read error: Connection reset by peer)
22:54:08 Join johnb2 [0] (
22:55:40lebelliumI confirm the sound is also crappy with mp3 files
22:57:02 Nick kugel_ is now known as kugel (~kugel@rockbox/developer/kugel)
22:58:33lebelliumacoustic is better and closer to the OF
22:59:03pamaurybut acoustic is too loud
22:59:55 Join MrZeus1 [0] (~MrZeus@2a02:c7f:7018:6200:5c0c:cd06:1ae4:bf3c)
23:00:00pamauryclearly the OF is not just using acoustic, either it uses something else or it preprocesses the file to reduce volume
23:00:53 Quit johnb2 (Ping timeout: 260 seconds)
23:00:56wodzpamaury: maybe you could run OF through ltrace/strace to figure out how it configures alsa
23:02:45pamauryI thought about it but I am also afraid that the amount of trace will be insane
23:03:25lebelliumpamaury: the OF has a much more progressive volume so it's not just about reducing the volume anyway
23:06:42pamauryolder Sony's allowed userspace to override the volume table but on E580 the driver does not seem to allow that
23:06:52CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
23:06:52*pamaury will try to trace the OF
23:08:09***Saving seen data "./dancer.seen"
23:12:25 Quit wodz (Quit: Leaving)
23:15:55prof_wolfffpolytux: if the storage media is not recognized by Apple firmware then probably it wont work on Rockbox, if it is compatible with Apple FW then it should work on Rockbox, an user with a similar problem:,51647.0.html , what is exactly your problem?
23:19:00pamauryoh wow, surprisingly enough, Sony put strace on the device...
23:21:05 Join initdctf [0] (60e76cfe@gateway/web/freenode/ip.
23:24:01pamauryhum, only on the E460 apparently
23:25:33 Quit initdctf (Client Quit)
23:27:38*pamaury now has a 2.5 MB trace log
23:34:34lebelliumpamaury: I guess this one is for you,51682.msg239213/topicseen.html#msg239213
23:35:49 Join smoke_fumus [0] (
23:57:11 Join jhMikeS [0] (

Previous day | Next day