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 2013-02-26

00:00:12 Quit lebellium (Quit: ChatZilla 0.9.90 [Firefox 20.0/20130220104816])
00:05:35 Quit Rower (Quit: Hmmm...)
00:08:09 Quit Rondom_ (Quit: No Ping reply in 180 seconds.)
00:08:45 Join Rondom [0] (~rondom@2a01:488:66:1000:b24d:4f2f:0:1)
00:10:00 Quit n1s (Quit: Ex-Chat)
00:12:08***Saving seen data "./dancer.seen"
00:37:46 Join lebellium [0] (
00:38:11 Join sakax [0] (~sakax@
00:43:37lebelliumokay as expected, with my theme, USB connection fails on C200v1. No time to dig further now but I'm so fed up with being obliged to use cabbiev2 on all my players to get USB working :(
00:44:27 Quit kadoban (Quit: bye)
00:45:05 Join kadoban [0] (~kadoban@
00:46:04 Quit lebellium (Quit: ChatZilla 0.9.90 [Firefox 20.0/20130220104816])
00:48:33 Quit dfkt (Ping timeout: 252 seconds)
00:48:51 Join dfkt [0] (dfkt@unaffiliated/dfkt)
00:50:36 Quit kugel (Quit: leaving)
00:50:45 Join kugel [0] (~kugel@
00:50:45 Quit kugel (Changing host)
00:50:45 Join kugel [0] (~kugel@rockbox/developer/kugel)
00:52:07 Quit XavierGr ()
00:56:46 Quit kadoban (Ping timeout: 245 seconds)
00:57:58 Quit prof_wolfff (Ping timeout: 252 seconds)
01:09:21 Quit pxb|honeycrisp (Quit: leaving)
01:09:55 Join XavierGr [0] (~XavierGr@rockbox/staff/XavierGr)
01:16:39 Join saratoga [0] (123e0c92@gateway/web/freenode/ip.
01:16:44saratogashould the 10 band EQ really have been enabled on PP devices?
01:16:49saratogais it actually useable?
01:17:24 Quit AlexP_ (Ping timeout: 252 seconds)
01:18:29soapwasn't it you who said it was only a few mults per band so around 2-3mhz on PP per band?
01:18:40soap(in the context of battery consumption)
01:18:53saratogasoap: thats on ARMv5
01:19:12saratogai bet its a lot slower on older ARM
01:19:43saratogai remember people complaining about skipping on audio on PP with the 5 band EQ and MP3 before we got it running on the second core
01:19:48soapOne of us is mistaken as to your original context. For I was telling [Saint] at the time what a power drag the EQ was on PP.
01:20:22saratogamultiplies are much slower on ARMv4, and loads and stores take longer
01:20:27saratogaso running filters is harder
01:20:40soapand that evening (our time) when [Saint] and I were talking is when you stepped in with the 2-3Mhz number. So either you didn't realize we were talking PP or you explicitly said ARMv5 and I missed it.
01:20:55saratogait should be tested at least
01:21:21 Join jhMikeS [0] (~jethead71@
01:21:21 Quit jhMikeS (Changing host)
01:21:21 Join jhMikeS [0] (~jethead71@rockbox/developer/jhMikeS)
01:21:41saratogapreferably before 3.13
01:22:54soaphow w/o adding some more code? Filters can't be part of the codec performance plugin, eh?
01:23:18soapor do you mean just brute force and see how many bands enabled before the UI goes to shit?
01:23:39saratogano the test_codec plugin can test DSP too
01:23:57saratogajust benchmark a wav file with DSP enabled (and the EQ bands set in config)
01:24:13saratogai think saint did this, but on newer players
01:24:28soapwhy WAV out of curiosity?
01:25:29saratogatheres nothing to decode
01:25:42saratogayou can use MP3 too, but you'll have to test with and without DSP and subtract
01:25:56saratogamaybe not a bad idea with WAV just in case
01:29:09saratogasince 5 bands used to lag the UI on PP with a 40 mhz mp3 decoder, its probably like 6-8MHz per band would be my random guess
01:30:59 Quit ender| (Ping timeout: 246 seconds)
01:43:55 Join ender| [0] (krneki@2a01:260:4094:1:42:42:42:42)
01:51:03 Join krabador [0] (~krabador_@
01:57:00 Quit dv_ (Ping timeout: 255 seconds)
02:02:24 Quit saratoga (Quit: Page closed)
02:12:12***Saving seen data "./dancer.seen"
02:16:06 Join dv_ [0] (
03:32:14*[Saint] tested the 10 band EQ on a PP target, and found playback worked fine for vorbis@500, mp3@320, and FLAC8
03:32:30[Saint]soap: saratoga: ^
03:33:29[Saint]Yes, admittedly, those are the only codecs I tested...but, I figured that would be "Good Enough".
03:35:58JdGordonpeople can always disable bands they don't need if its lagy right?
03:39:51 Quit bluebrother (Disconnected by services)
03:39:56 Join bluebrother [0] (~dom@rockbox/developer/bluebrother)
03:43:27 Quit fs-bluebot (Ping timeout: 260 seconds)
03:45:11 Join fs-bluebot [0] (~fs-bluebo@
03:49:07 Join amayer [0] (
03:55:41 Join kadoban [0] (
04:00:49 Quit ender| (Ping timeout: 246 seconds)
04:06:23 Quit krabador (Remote host closed the connection)
04:06:29 Quit Scromple (Quit: Leaving)
04:10:36 Join Scromple [0] (~Simon@
04:12:15***Saving seen data "./dancer.seen"
04:13:13 Join ender| [0] (krneki@2a01:260:4094:1:42:42:42:42)
04:23:56 Quit amayer (Quit: Leaving)
04:42:19 Join amiconn_ [0] (amiconn@rockbox/developer/amiconn)
04:42:19 Join pixelma_ [0] (pixelma@rockbox/staff/pixelma)
04:42:19 Quit amiconn (Disconnected by services)
04:42:19 Quit pixelma (Disconnected by services)
04:42:21 Nick pixelma_ is now known as pixelma (pixelma@rockbox/staff/pixelma)
04:42:22 Nick amiconn_ is now known as amiconn (amiconn@rockbox/developer/amiconn)
04:53:19 Quit ender| (Ping timeout: 246 seconds)
05:06:18 Join ender| [0] (krneki@2a01:260:4094:1:42:42:42:42)
05:38:54 Quit [7] (Disconnected by services)
05:39:02 Join TheSeven [0] (~quassel@rockbox/developer/TheSeven)
05:58:10 Quit SuperBrainAK (Quit: pbly going to sleep /_\)
06:04:58 Join akaWolf [0] (~akaWolf@unaffiliated/akawolf)
06:08:32 Join webguest25 [0] (
06:09:28 Quit webguest25 (Client Quit)
06:12:16***Saving seen data "./dancer.seen"
06:20:41 Quit zamboni ()
07:00:02 Join mortalis [0] (~kvirc@
07:01:11 Quit [Saint] (Remote host closed the connection)
07:03:44 Join [Saint] [0] (~quassel@rockbox/user/saint)
07:05:50 Join mrkiko [0] (
07:14:20 Quit kiwicam (Remote host closed the connection)
07:28:17 Join esperegu [0] (
07:42:04 Quit djp_ (Ping timeout: 260 seconds)
07:42:35 Join djp_ [0] (
07:42:36 Quit djp_ (Changing host)
07:42:36 Join djp_ [0] (~djp@fsf/member/djp-)
07:47:16 Quit djp_ (Ping timeout: 248 seconds)
07:47:28 Join djp_ [0] (djp@fsf/member/djp-)
07:49:42 Join melmothX [0] (~melmoth@unaffiliated/melmothx)
08:10:17 Quit amiconn (Remote host closed the connection)
08:10:17 Quit pixelma (Remote host closed the connection)
08:11:50 Join pixelma [0] (pixelma@rockbox/staff/pixelma)
08:11:50 Join amiconn [0] (amiconn@rockbox/developer/amiconn)
08:12:15 Join Rower [0] (
08:12:19***Saving seen data "./dancer.seen"
08:16:05 Quit froggyman (Ping timeout: 256 seconds)
08:26:35 Quit kadoban (Ping timeout: 264 seconds)
08:35:19 Join ender` [0] (~ender@
08:37:40 Join olspookishmagus [0] (
08:39:16 Quit zchs (*.net *.split)
08:39:16 Quit nosa-j (*.net *.split)
08:39:16 Quit ruskie (*.net *.split)
08:39:16 Quit user890104 (*.net *.split)
08:39:16 Quit rudi_s (*.net *.split)
08:43:26 Join nosa-j [0] (~m00k@
08:46:26 Join LinusN [0] (
08:51:07 Join zchs [0] (
08:51:07 Join ruskie [0] (ruskie@sourcemage/mage/ruskie)
08:51:07 Join user890104 [0] (Venci@unaffiliated/user890104)
08:51:07 Join rudi_s [0] (
08:53:53 Join petur [0] (~petur@rockbox/developer/petur)
09:02:58 Join einhirn [0] (
09:09:51 Join sarg [0] (~sarg@
09:10:55 Quit ender| (Ping timeout: 246 seconds)
09:19:52 Join kevku [0] (~kevku@2a01:d0:ffff:34a::8:3)
09:23:38 Join ender| [0] (krneki@2a01:260:4094:1:42:42:42:42)
09:36:05 Quit Guinness (Ping timeout: 244 seconds)
09:46:34 Quit simabeis (Remote host closed the connection)
09:48:23 Join wodz [0] (
09:53:14wodzamiconn: ping
10:05:52 Quit melmothX (Quit: bau)
10:10:07 Quit einhirn (Ping timeout: 252 seconds)
10:12:21***Saving seen data "./dancer.seen"
10:13:03 Join melmothX [0] (~melmoth@unaffiliated/melmothx)
10:29:41JdGordonbahaha fuze+ != fuzev2
10:30:00JdGordonespecially the screen size! and the skin engine doesnt cope too well with the wrong one trying to be loaded
10:37:40 Join pxb|honeycrisp [0] (
10:39:06JdGordon[Saint]: pingaroony
10:39:43 Quit [Saint] (Remote host closed the connection)
10:42:13 Join [Saint] [0] (~quassel@rockbox/user/saint)
10:51:09JdGordonso, the %Cl and %Cd tags are stupid
10:51:56sargwodz: dev/mem idea has failed, but I got another. Seems it is possible to inject some code in current loader, in FlashBoot.bin precisely. Code would copy bootrom to user accessible place.
10:56:14sargAll I need, is some safe offset in memory which will survive linux kernel loading.
10:56:35sargAnd later I access this offset through /dev/mem to obtain copied data.
10:56:57 Quit TBCOOL (Quit: .)
10:56:58sargwodz: can you suggest me what this offset can be ?
10:57:09 Join y4n [0] (~y4n@unaffiliated/y4ndexx)
10:57:23wodznot really, I don't know linux kernel internals
10:58:10wodzhow much code can you inject?
10:58:43sarg500 bytes min
10:59:12sargFlashBoot.bin in first place copying new vector table to offset 0x0, overwriting bootrom
11:00:01sargbut it has about 500 bytes array in beginning, initially filled with zeros, so I may use this space to place my code
11:00:22 Join TBCOOL [0] (
11:00:37wodzbootrom can't be overwritten by definition, you know
11:01:01sargbootrom is copied to RAM, offset 0x0 ?
11:01:08gevaertswodz, LinusN: I'm still not entirely clear on how to use this bdm thing. I can connect using gdb (although I often get a "packet too short" warning), and I can upload an elf file (provided it's not too big, I have to keep those pins shorted), but I haven't managed yet to get it to stay on without the pins, and I understand from some parts of the wiki that I have to get the cpu into halt mode first, which is also not clear to me
11:01:18wodzsarg: usualy not
11:01:42sarghm, ok, but first instruction at 0x0 should be jump to real bootrom location
11:01:44gevaertsAlso, if you happen to have reflashing code, I wouldn't mind getting a copy, writing my own seems like a waste of time if it already exists :)
11:01:59sargso if I retrieve vector table from 0x0, I can find where bootrom is really placed
11:02:07gevaerts(still on h300. I hope to find a volunteer to fix the h120 connector today)
11:02:48wodzgevaerts: bdm pod should put processor in halted state AFAIK.
11:03:07gevaertsah, ok. So I can ignore that part, probably
11:04:00wodzgevaerts: I also had "packet too short" thing until I placed ferrite bed over bdm flex cable.
11:04:24gevaertsOh, ok. There are ferrite things around it, but maybe I have to move them a bit then
11:05:11wodzgevaerts: One of the very first thing crt0.S is doing is to activate pull-up/pull-down on KEEP_ACT line so pins shorting should not be needed when uploaded executable started\
11:06:27gevaertsThat's what I thought...
11:06:39gevaertsOK, so I'll investigate those ferrite things
11:06:40copperwas the theme bad, or was Rockbox bad?
11:06:56wodzI recall I used gdb init script to configure pullups/pulldowns on startup so I didn't have to short anything after gdb startup
11:06:59JdGordoncopper: the theme, but i've fixed the code so it doesnt matter
11:07:42kugelJdGordon: what's happening now if the AA is larger than the vp?
11:07:53kugelpreviously it was scaled down IIUC?
11:09:28gevaertswodz: is that a fixed GPIO for all coldfire-based devices, or does it vary?
11:10:34JdGordonkugel: bitmap_part shold crop it
11:10:57wodzit does vary, see firmare/target/coldfire/crt0.S starting from line 315
11:11:05JdGordonand no, previously it was resized down to the viewport size, but that doesnt really make sense because its not necessarily the vp it is being displayed in
11:11:40 Join lebellium [0] (
11:13:01pixelmahmm, I thought it was cropped before too. That's at least what I remember my c200 sbs was doing
11:13:28kugelJdGordon: what I said :)
11:13:59pixelmahadn't updated the build there for quite some time, I mostly use my M5 or my phone currently
11:14:48gevaertswodz: what's %a1 there? I assume the base address for the GPIO in question
11:15:07lebelliumpixelma: did you make a theme for c200? I'm looking for SBS which are compatible with USB connection :D
11:17:12pixelmayes I did, but I haven't published it anywhere (as it also relies on my personal tagging style)
11:17:36wodzgevaerts: its MBAR2 base address, gpios are in MBAR2 mem region
11:17:59JdGordonkugel: actually i lie, previously (i.e before the change i just undid) it checked against the SCREEN size which made even less sense
11:18:00pixelmaand as I said, the build is quite old
11:19:27lebelliumpixelma: a build from before all these USB connection and memory corruption issues?
11:21:23pixelmaJdGordon: is it now possible to have AA size filling the complete screen even if it is rectangular - so scaled to fit the longest dimension of the screen and crop the rest? That wasn't possible before (it only scaled up to the shortest dimension of the screen). That would be neat
11:21:28gevaertswodz: ok. Sorry if I appear stupid, but how do I get at that in gdb? Can I assume a fixed value?
11:22:03pixelmalebellium: probably
11:23:50JdGordonpixelma: I don't know.
11:24:11copperJdGordon: thx, btw
11:24:30lebelliumpixelma: would you mind trying an early beta version of my theme on your c200 and check if USB is working?
11:25:42wodzgevaerts: IIRC you setup the base address for MBAR2. line 77-78 seems to confirm this. I believe we use the same mbar2 base for all cf targets
11:26:31wodzgevaerts: wait, you have unpatched bdm-server, right? Then gdb doesn't know about mbar2 register
11:27:01gevaertsI have the one LinusN sent me
11:27:15*gevaerts will grep in there
11:27:41gevaertsIt knows about mbar2
11:27:58gevaertsSo I run set %mbar2 = <something>?
11:28:05wodzshould do
11:29:33gevaertsOK, I have enough to experiment with further tonight then :)
11:30:41wodzI may check the backup I found yesterday, maybe there are some notes and scripts from my work on hd200
11:32:11gevaertsThat could be useful
11:32:14pixelmalebellium: if you mean with the old build, then I guess the theme won't work and I'll only get the failsafe one
11:32:23 Join einhirn [0] (
11:32:57lebelliumyes I meant with the old build. Is it too old for current thme tags?
11:36:46lebelliumbtw, here it is"> it's a beta version so the code is still dirty and not cleaned but it *should not* make the USB connection fail though! That's killing me :(
11:37:02pixelmadepends on which tags exactly
11:37:30wodzlebellium: not correct, that is killing usb connection :-)
11:43:22JdGordonlebellium: what target?
11:44:45lebelliumUSB is only working everytime with cabbiev2. With all other themes it randomly works and with mine it actually never works properly
11:45:15JdGordondoes that still happen if only the wps is loaded?
11:46:28lebelliumIf I use cabbiev2 theme but with my WPS USB is working fine.
11:47:31lebelliumthe only difference I see at first is that I made my own status bar (not coded as a status bar) while cabbiev2 uses the default status bar
11:54:18 Join kiwicam [0] (~quassel@
12:02:56copperThis uses an icon most likely copyrighted to Apple:
12:03:16copperI used a UTF-8 icon here with the DejaVu Sans font:
12:03:38copperwould the second one be a copyright problem?
12:03:59copperi.e. could I use that on my theme and upload it on Rockbox's website?
12:04:40copper[Saint]: ^^
12:05:19coppers/UTF-8 icon/UTF-8 character/
12:06:47wodzfor me this two are sufficiently different
12:09:44copperwhat about DejaVu copyright
12:10:13copperoh, they have an IRC channel
12:10:16copperI could ask there
12:12:25***Saving seen data "./dancer.seen"
12:12:38 Join kadoban [0] (
12:16:48 Quit wodz (Read error: Connection reset by peer)
12:16:56 Join wodz [0] (~wodz@
12:27:57 Quit preglow (Quit: leaving)
12:28:49 Join preglow [0] (
12:29:09 Quit preglow (Client Quit)
12:29:47 Join preglow [0] (
12:29:47 Quit preglow (Changing host)
12:29:47 Join preglow [0] (thomj@rockbox/developer/preglow)
12:45:01sargwodz: please take a look">
12:45:35sargit is stage2 on rk30
12:46:44sarginterrupt vectors seems valid, but reset vector pointing to ANDEQ operation, wtf?
12:47:53wodzit doesn't look sane
12:48:46sargrk30 loader file has 4 parts in it, 30_LPDDR2_300MHz_DD.bin FlashBoot.bin FlashData.bin rk30usbplug.bin
12:49:17wodzok, first is ram controller setup the same as in rk27xx
12:49:30sargi think, boot sequence is bootrom -> DDR_300Mhz -> rk30usbplug -> FlashBoot
12:50:06sargrk30usbplug and FlashBoot overwriting vector table at first
12:50:10wodzI guess it is reverse actually, rk30usbplug is failsafe image probably in case flash boot failed
12:50:27sargnice idea
12:50:35wodzthats how it is in rk27xx
12:51:00sargthis would be great, as I can't find enough space in rk30usbplug to place injected code
12:51:52wodzfirst try to inject delay loop. this will give you the answer which images are executed
12:53:28wodzthats interesting they kept 0x60000000 as dram base address
12:54:03sargi am perfectly sure that this is valid address
12:54:42wodzthe same base is in rk27xx
12:55:05sargchinese developers )
12:56:18wodzthe vectors are taken from 0x6000C898, right?
12:56:57sargand 6000c898+0x7c is right at the end of file
12:58:13wodzwhat are hex values of 0x74 and 0x78? are this any special?
12:59:00sargwhere did you get those values?
12:59:42wodzI mean the values interpreted as andeq and mulvs
13:00:51wodzirq/fiq handler looks interesting
13:03:53sargflashboot vector table seems to be correct">
13:06:07 Quit petur (Quit: *plop*)
13:08:08sargok, enough for today, I will continue this tomorrow
13:09:38 Quit sarg (Quit: see you later)
13:11:12 Join |fly| [0] (~fly@
13:52:59 Quit mrkiko (Ping timeout: 264 seconds)
13:59:32 Join mrkiko [0] (
14:08:13 Join amayer_ [0] (
14:10:38 Quit mrkiko (Ping timeout: 260 seconds)
14:12:29***Saving seen data "./dancer.seen"
14:12:51 Quit wodz (Quit: Leaving)
14:27:27 Quit olspookishmagus (Quit: All for nothing)
14:39:13 Quit scorche (Disconnected by services)
14:39:16 Join scorche` [0] (~scorche@rockbox/administrator/scorche)
14:40:35 Quit rudi_s (Quit: leaving)
14:47:25 Part amayer_
14:51:01ukleinekamiconn: he usually reads the logs, so just address him.
14:51:30amiconnHe only pinged, so I need to know first what he wanted
14:54:59 Join amayer_ [0] (
14:56:38 Part amayer_
14:58:24 Quit kadoban (Ping timeout: 240 seconds)
15:02:08 Join amayer [0] (
15:05:51 Quit jhMikeS (Ping timeout: 256 seconds)
15:05:55 Join wodz [0] (
15:06:29 Quit amayer (Client Quit)
15:07:21 Join amayer [0] (
15:08:21wodzamiconn: I wanted to ask how ucl self unpack image for recorder works. My understanding is that decompressor is coupled with binary image, decompressor is moved to iram, ucl image is moved to the end of dram, then unpack function is called with dst at the begining of dram and finally control is passed to unpacked content. Is this correct?
15:27:30 Quit amayer (Quit: Leaving)
15:28:42 Join amayer [0] (
15:51:43 Join cac2s [0] (
15:55:51 Quit cac2s (Client Quit)
15:59:32 Quit [Saint] (Remote host closed the connection)
16:01:39 Join [Saint] [0] (~quassel@rockbox/user/saint)
16:09:41 Join mrkiko [0] (
16:10:23 Quit kevku (Ping timeout: 264 seconds)
16:12:33***Saving seen data "./dancer.seen"
16:17:48 Join Guinness [0] (Slayer@
16:25:33 Quit |fly| (Quit: Leaving)
16:28:54wodzIs there a way to build bootloader in different format then main binary? configure seems to use the same command for both
16:36:11 Quit wodz (Ping timeout: 264 seconds)
16:48:02 Join kevku [0] (~kevku@2001:470:27:773:0:feed:c0f:fee)
16:51:17 Join scorche [0] (~scorche@rockbox/administrator/scorche)
16:54:13 Quit scorche` (Ping timeout: 244 seconds)
16:57:47 Join sakax_ [0] (
16:58:43 Quit sakax (Quit: Leaving)
16:59:41 Join krabador [0] (~krabador_@
17:08:24 Quit Guinness (Read error: Connection reset by peer)
17:10:56 Join Guinness [0] (~Slayer@
17:16:51 Join n1s [0] (
17:16:51 Quit n1s (Changing host)
17:16:51 Join n1s [0] (~n1s@rockbox/developer/n1s)
17:27:30 Join thegeek [0] (
17:28:50 Quit thegeek_ (Ping timeout: 244 seconds)
17:32:06 Join wodz [0] (
17:42:17 Join froggyman [0] (~me@unaffiliated/froggyman)
17:42:31 Join kaputnik [0] (
17:43:18 Quit thegeek (Ping timeout: 246 seconds)
17:47:47 Quit mortalis (Quit: KVIrc 4.1.3 Equilibrium
17:48:39 Quit Scall (Quit: Bye bye)
17:49:05 Join thegeek [0] (
17:51:05 Join Scall [0] (~chat@unaffiliated/scall)
17:52:11 Quit wodz (Quit: Leaving)
18:02:27 Join prof_wolfff [0] (
18:03:50 Join bertrik [0] (~quassel@rockbox/developer/bertrik)
18:04:51 Join zamboni [0] (~bottledwa@unaffiliated/zamboni)
18:07:59 Quit pxb|honeycrisp (Quit: leaving)
18:10:19 Quit thegeek (Read error: Connection reset by peer)
18:10:38 Join thegeek [0] (
18:12:21 Join pxb [0] (pxb@2001:1b40:5600:e00::5079:3edb)
18:12:36***Saving seen data "./dancer.seen"
18:21:08 Join pxb^ [0] (
18:35:54 Join kadoban [0] (
18:44:14 Quit sakax_ (Remote host closed the connection)
18:55:07 Quit [Saint] (Ping timeout: 244 seconds)
18:58:04 Quit Scall (Quit: Bye bye)
18:58:50 Quit pxb (Quit: leaving)
19:00:55 Quit pxb^ (Remote host closed the connection)
19:01:07 Join pxb [0] (pxb@2001:1b40:5600:e00::5079:3edb)
19:02:25 Join bertrik_ [0] (
19:02:26 Quit bertrik_ (Changing host)
19:02:26 Join bertrik_ [0] (~quassel@rockbox/developer/bertrik)
19:09:31 Join Scall [0] (~chat@unaffiliated/scall)
19:10:25 Join [Saint] [0] (~quassel@rockbox/user/saint)
19:14:05lebelliumI'm trying to identify a potential part of my code or a particular tag causing the USB issues but no luck: after removing some stuff from the status bar the USB connection seemed to be more stable but soon it failed again. I hate those random things :(
19:14:34 Join AlexP [0] (~alex@rockbox/staff/AlexP)
19:37:06 Quit shamus (Read error: Connection reset by peer)
19:37:24 Join shamus [0] (
19:40:06 Quit kaputnik (Ping timeout: 260 seconds)
19:45:55 Quit einhirn (Quit: Miranda IM! Smaller, Faster, Easier.
19:50:50copperbtw, I haven't had a single lockup with my Fuze+ during USB transfers
19:51:08lebelliumeven with my theme?
19:52:31lebelliumThere is no USB in Rockbox on the R0 and I don't have a Fuze+ so I haven't had the chance to try it out but I'm sure it won't work. USB just can't work with my themes... lol
19:59:41gevaertsLinusN: I'm usually getting results like
20:00:13*gevaerts is building a newer gdb
20:01:59gevaertsThat might actually help...
20:12:37***Saving seen data "./dancer.seen"
20:17:57gevaertsWell, except that x/x doesn't show the data I uploaded, unless I'm doing things wrong
20:18:21LinusNdid you upload it to dram?
20:18:54gevaertsI used "load rockbox.elf"
20:19:22LinusNif you did it directly from reset, the dram might not be initialized
20:19:38gevaertsAh, right. That will probably be the case here...
20:20:16LinusNor rather, the dram *controller*, but i guess you understood that
20:20:21*gevaerts nods
20:20:39gevaertsSo what I need is probably our crt0.S linked to iram?
20:21:53LinusNor you could initialize the dram controller with a gdb script
20:22:12 Join ender1 [0] (
20:22:36LinusNi remember doing that back in the day
20:23:54*gevaerts looks at crt0.S to see what he has to do
20:24:06*gevaerts pretends to understand it all
20:24:42 Quit ender` (Ping timeout: 245 seconds)
20:25:00 Quit ender| (Ping timeout: 246 seconds)
20:25:01*LinusN finds a gdb script from back in the days
20:25:07 Join wodz [0] (
20:26:07n1sperhaps such scripts could be put in the repo for when someone else comes along and wnats to fiddle with it in another 6 years or so :)
20:26:26wodzgevaerts: Back in days I simply assembled crt0.S into separate small bin run from iram and executed it before actual rb loading.
20:27:52 Join simabeis [0] (
20:29:01 Quit ml| ()
20:29:21 Join ml| [0] (~ml@unaffiliated/ml/x-3958674)
20:34:51 Quit Scall (Quit: Bye bye)
20:35:46 Quit bertrik (Remote host closed the connection)
20:37:19 Join bertrik [0] (~quassel@rockbox/developer/bertrik)
20:37:24wodzgevaerts: unfortunately I don't have my gdb scripts to fiddle with 5249
20:37:49gevaertswodz: I'm now tweaking the lkinker file to load crt0 to iram properly
20:37:52 Join ender| [0] (krneki@2a01:260:4094:1:42:42:42:42)
20:39:35gevaertsI probably need to make that code halt :)
20:40:32wodzput inifinite loop instead of branching to main()
20:40:53wodzor set breakpoint earlier :-)
20:42:05 Join Scall [0] (~chat@unaffiliated/scall)
20:48:31 Join kaputnik [0] (
20:51:49 Quit Gallomimia (Ping timeout: 255 seconds)
20:52:56 Join Gallomimia [0] (
20:53:39 Join kaputnik_ [0] (
20:56:34 Quit kaputnik (Read error: Operation timed out)
20:59:16 Join Strife89 [0] (~Strife89@2602:306:250d:8809:225:d3ff:fed6:15a)
21:04:34 Quit amayer (Quit: Leaving)
21:05:12 Join toffe82 [0] (
21:09:49 Join amayer [0] (
21:13:03gevaertshmm, I should have disabled bdm debug output before uploading this full rockbox image
21:14:05gevaertsWoo! Backlight!
21:16:44 Quit shamus (Read error: Connection reset by peer)
21:17:00 Join shamus [0] (
21:18:03 Join pxb^ [0] (
21:18:59 Quit akaWolf (Ping timeout: 248 seconds)
21:19:25gevaertsWoo! Rockbox!
21:19:42gevaertsWell, the spash screen anyway
21:19:48 Join stripwax [0] (~Miranda@rockbox/developer/stripwax)
21:19:53*gevaerts assumes the bootloader normally sets up a bit more
21:20:33wodzmissing 'l' in splash screen is a bit worrying :-)
21:22:02gevaertsAnyway, I think I'll first clean up the stuff I have so I can recreate this binary when I need it
21:22:16wodzdo you upload main binary or bootloader?
21:23:15gevaertsI'm assuming the bootloader is linked to flash addresses
21:23:38wodzyes, but nothing prevents you from linking it at iram address
21:23:50gevaertsAh, true. I should try that
21:23:58wodzthat way you will know if bootloader actually works
21:24:30gevaertsAs in, don't even try flashing it if it already fails there? Yes...
21:25:15gevaertsGood idea
21:25:30 Quit stripwax (Quit: Miranda IM! Smaller, Faster, Easier.
21:28:53kugelgevaerts: trying to repair the h300 bootloader?
21:29:03wodzOk, I have POC loader which supports ucl packed binary payload in rkw firmware format. Now I need to test if it makes any noticeable difference in boot time on rk27xx
21:29:08gevaertskugel: that's the idea, yes
21:37:26 Quit wodz (Quit: Leaving)
21:39:53 Join tjb0607 [0] (~tjb0607@
21:49:07 Quit LinusN (Quit: disconnecting from stoned server.)
21:50:08 Join LinusN [0] (
21:51:08 Quit tjb0607 (Ping timeout: 256 seconds)
21:57:01 Quit LinusN (Quit: disconnecting from stoned server.)
21:57:16 Join LinusN [0] (
21:59:43gevaertsOK, so does anyone remember when the bootloader worked?
22:03:37 Quit y4n (Disconnected by services)
22:03:44 Join y4n [0] (~y4n@unaffiliated/y4ndexx)
22:05:03 Join tjb0607 [0] (~tjb0607@
22:06:25 Part LinusN
22:09:10 Quit melmothX (Quit: bau)
22:09:19n1sgevaerts:,17202.0.html is the best clue i can think of
22:11:03n1siirc, one theory was that the breakage had something to do with the patching into the OF or the flashing of the patched hex
22:11:18n1salso note that the toolchain has changed
22:11:30gevaertsI know. I'm hoping for now that the toolchain is not related
22:11:51n1sit was definitely broken before the toolchain upgrade at least
22:12:38***Saving seen data "./dancer.seen"
22:22:11AlexPIs it known that the H100 bootloader is also broken, or are we just assuming it is? And how similar are they? Just wondering if the H100 v7pre4 might give any clues
22:26:24gevaertsI don't know
22:26:46gevaertsAnd I broke the bdm connector on the h120, so that needs to be fixed before I find out
22:35:31 Join Ward [0] (
22:35:55 Nick Ward is now known as Guest72310 (
22:43:00bluebrothergevaerts: didn't LinusN work on the h100 bootloader at DevCon 2 years ago (i.e. in London)?
22:43:22*gevaerts isn't sure
22:44:29bluebrotherI seem to remember him even borrowing a debug tool (paperclip :)
22:46:17 Quit mrkiko (Ping timeout: 276 seconds)
22:46:46bertrikentire player families have been ported to rockbox since iriver bootloaders last worked!
22:48:00bluebrotherentire families?
22:50:48bertrikallow me to exaggerate a bit please :)
22:51:11 Quit y4n (Quit: PANTS OFF!)
22:51:36derfFathers, daughters, nephews, third cousins...
23:02:26 Join mrkiko [0] (
23:02:50 Quit amayer (Ping timeout: 276 seconds)
23:03:14 Quit n1s (Quit: Ex-Chat)
23:08:38 Quit kevku (Quit: What you must always keep in mind as a developer? That's of course- cake.)
23:09:09 Quit Rower (Quit: Hmmm...)
23:24:17soap[Saint], saratoga, iPod Nano 1st. WAV: No EQ=1.81Mhz. Low Shelf=7.77Mhz total. 1 Peaking=7.78 total. So 6Mhz per band. So we'd expect 61.41Mhz needed for wav and all ten bands and the test says: 44.13
23:24:31soapdoesn't smell like all the bands are actually applied.
23:24:47soapanyone actually RMAA this with narrow values for Q?
23:26:40 Quit pxb (Ping timeout: 245 seconds)
23:27:47 Join pxb [0] (pxb@2001:1b40:5600:e00::5079:3edb)
23:37:09soapor am I blanking on something? 5 bands on + wav = 24.13. 4 bands on + wav=19.98. 3 bands on + wav=15.86, etc. So 1.8 for decoding wav, 4.2 per band and another 1.8 fixed cost for turning on the EQ?
23:37:23 Join stripwax [0] (~Miranda@rockbox/developer/stripwax)
23:37:43 Quit stripwax (Client Quit)
23:45:10 Join rudi_s [0] (

Previous day | Next day