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 2009-11-04

00:03:01 Join webguest11 [0] (
00:06:17 Nick Ypsy is now known as YPSY (n=ypsy@
00:10:13kugelBob_C: I see a "/* Does this get optimised out ? */" comment in sd-s3c2440.c
00:11:35kugelI'm not sure how it works for dedicated functions, but those loops within code are potentially optimized away. attaching a "asm volatile("nop\n");" is the safer way. you can always check the generated assembly of course
00:13:53Tornekugel: you shouldn't ened the actual nop
00:13:57 Quit webguest11 ("CGI:IRC")
00:13:59kugelassuming you added that question, and the code you code you copied from :p
00:14:00Tornejust an empty asm volatile block should do
00:17:40kugelinteresting, the mini2440 keeps loading rockbox when you remove the sd card during transfer
00:17:54kugeli.e. during loading the rockbox binary
00:19:50Bob_C_kugel: at the level of optimisation we use, I don't think empty loops are removed. I'm surprised more aggressive optimisation is not used
00:21:24TorneBob_C_: Lots of -O2 level optimisations increase code size, it's an awkward tradeoff on small devices
00:21:29Torneespecially ones with no cache
00:22:29kugelthe mini2440 doesn't even build with eabi
00:22:31kugelassembler error
00:22:32Bob_C_-Os seems to lots of opt on AVRs anyway
00:22:44TorneBob_C_: Yes, many space optimisations are *also* time optimisations
00:23:08Torneit's also possible taht not everything actually *works* with -O2, of course :)
00:23:15Bob_C_yeah, so why not -Os?
00:23:20Tornegcc can be a hilarious minefield sometime
00:23:41TorneBob_C_: -Os is used on some platforms, -O on others
00:24:04TorneIt's whatever seemed to be the est tradeoff between speed and size (and actually worked) last time soemone was motivated to test it, i expect
00:26:04Torneif you'd like to paintakinly test various combinations of gcc flags across multiple targets, i doubt anyone would stop you :)
00:26:50Bob_C_Ha, who wants to cross the minefield first?
00:27:26Torneatm there are some people working on getting rockbox to work with an EABI ARM toolchain (the new standard ARM binary spec)
00:27:32Tornewhich is related, I guess.
00:28:26Torneeabi has some nice tricks for handling long calls and the like which could help us, but not all targets work commpiled that way atm :)
00:28:33kugelsome == me and Unhelpful
00:28:59kugel"/home/kugel/rbdev/rockbox-git/firmware/target/arm/s3c2440/crt0.S:558: Error: symbol vectors is in a different section" :'(
00:29:51Tornehooray for mysterious compile and link problems
00:30:25kugelwtf is "adr lr, vectors"
00:30:39kugelthat's the problematic line
00:30:47Torneadr is an assembler pseudoop
00:31:06Torneit turns into an add lr, pc, #offset
00:31:12Tornepc-relative addressing
00:31:43Torneso it will indeed not work if the symbol referred to is in a different section, since it can't do pc-relative addresses for symbols it doesn't know where they are going.
00:31:55Torneldr lr, =vectors would work
00:31:57*kugel wonders how it works in the current setup then
00:32:03Tornebut is slower.
00:32:17TorneNo idea :)
00:32:36Tornemaybe it makes hilarious assumptions
00:32:49kugelthat's 1 instruction in crt0.S, speed *hardly* matters
00:33:06Bob_C_I noticed that the s3c crt0 seems to rely on position independent code
00:33:27Tornewell yes, entry point code often is PIC
00:33:37Torneinc ase you are running XIP from flash and are going to relocate to ram or similar
00:33:48Tornepic is not very difficult on ARM so it's not much of a hardship
00:33:59Torneif vectors is a fixed address then just ldr it from a pool
00:34:09Tornethe compiler will generate a reloc, linker will fix it, done
00:34:30Torneif it's supposed to be pic then you'll have to do something else..
00:34:42Bob_C_I don't just mean it happens to use it, it *relies* on it
00:35:44 Quit evilnick_B ("Page closed")
00:35:52Torneyah, but i've not looked at the code in question so i'm not sure what vectors is :)
00:36:46Bob_C_although in case vectors is surely always 0, and it set lr to the reset vector in case main() ever returns
00:37:19 Quit HellDragon (Read error: 104 (Connection reset by peer))
00:37:24 Join HellDragon [0] (
00:37:39Tornewell if it's really a reference to the ARM low reset vector then just hardcode #0
00:38:00Bob_C_everything done after call to enable_mmu is done with VA, so the linker addresses are valid
00:38:18TorneThe reset vector is always 0 or ffff0000 so hey
00:38:18*kugel just saw rockbox rebooting on the mini2440 eabi build
00:38:27Tornejust hardcode it :)
00:38:40kugelduring main(), is it possible that it indeed hit that case and branched to the vectors?
00:39:12kugelthe e200 would just freeze at that point
00:39:46Tornei would hope that if you are turning the MMU on that you are moving to high vectors, though
00:39:57Torneand leaving the page at 0 unmapped, to catch null pointer derefs. No? :)
00:40:19Bob_C_Torne: That's the other way of doing ;)
00:40:31Bob_C_.. it, but we dont't
00:40:38TorneWell, you should :)
00:40:47Torneleaving 0 unmapped is a really, really handy debugging tool...
00:40:54Tornecatches loads of issues
00:41:18Bob_C_Its a lot more fun to overwrite the vectors though ;)
00:42:03Bob_C_we seem to allocate VA in 1MB sections ?
00:42:03Torneat least map it readonly if you aren't gonna blank it :)
00:42:22Torneyeah, saves having pagetables
00:42:51Tornejust leave 0-1mb unmapped unless you have a really good reaso to need that range
00:42:55 Join Stephen [0] (
00:42:58Bob_C_It's funny, I worked on the kernel team for this SoC while at Samsung, ah such memories
00:43:07kugelBob_C_: after undeffing SD_DEBUG the reboot turned into a "no partition found. insert USB cable and fix it"
00:43:12 Nick Stephen is now known as Stephen__ (
00:43:14Torneoh, but then mapping the high vectors is a pain.
00:43:15 Quit hillshum ("Leaving")
00:43:34Tornewell, if there is some *reasonble* way, do it :0
00:43:39Bob_C_I wrote the MMU code on that project
00:44:34kugelUnhelpful: the mini2440 survives setttings_reset
00:44:49Bob_C_There was a great bug where someone wrote to NULL pointer and put the FLash in program mode
00:45:39Tornehah. yeah :)
00:47:37Bob_C_Shame to waste 1MB...
00:47:59Torneyou have 4GB, there's not that much to map.. ;)
00:48:17 Part froggyman
00:48:32saratogawhats the correct way to watch for dock connection so that we know to enable line out on AMS?
00:48:38saratogajust check in the button code?
00:48:43saratogaor the audio driver?
00:49:09kugelUnhelpful: the mini2440 gets even farther. most of the e200 eabi issues seem dual core related indeed
00:49:14 Part toffe82_
00:49:26kugelcheck gpio I'd say
00:49:57Bob_C_kugel: sounds like sd_read fails to read MBR
00:50:19saratogakugel: fwiw its probably on DBOP
00:50:41kugelsaratoga: I wouldn't be sure about that
00:51:44saratogathough we have yet to confirm it
00:51:49kugelTorne: the crt0.S might not even be eabi related, but instead due to my newer version of binutils
00:51:59Tornecould be..
00:52:27TheSevenbtw, WTH does dbop mean? Division of Biologic Oncology Products? Design for the Base of the Pyramid?
00:52:47saratogaits an arm component for updating LCDs
00:53:19saratogabasically a 4MHz bus that can read/write control signals to an LCD, but sandisk uses it for button reading too since AMS devices have hardly any GPIO
00:53:50saratogaotherwise its basically a GPIO with a FIFO and limited ability to trigger interrupts
00:54:11TheSevenok, so basically some kind of external I/O externder chip?
00:54:34saratogaits actually built into the SOC though, and shares a lot of physical pins with GPIO
00:55:08TheSevenhuh? why don't they just use the GPIO unit to access these pins then?
00:55:23saratogasome of them can be used either way, some are only one or the other
00:55:35TheSevenweird thing...
00:55:43saratogathe DBOP is much faster for some thigns since the pins have a FIFO and don't need to be serviced as often
00:55:49saratogayes its quite stupid
00:56:06saratogaparticuarly since it means you can't use interrupts to read out the wheel on the AMS devices
00:56:26saratogaand since some physical pins are connected to both the screen AND wheel
00:57:44kugeldbop is a ARM thing?
00:58:01kugelTheSeven: it means data output pin IIRC
00:58:25kugelhm, the b means something too :p
00:58:30 Quit TheSeven (Nick collision from services.)
00:58:46 Join The_Seven [0] (n=theseven@rockbox/developer/TheSeven)
00:58:58 Nick The_Seven is now known as TheSeven (n=theseven@rockbox/developer/TheSeven)
00:58:58saratogamaybe not
00:59:12TorneARM are mysterious like that
01:01:28 Join lazaruso [0] (
01:01:33JdGordon|kugel: is there a bug somewhere with the title position in lists?;topic=22971.0;attach=3898;image has the title aligned with the list?
01:01:47 Quit Bob_C_ ("ChatZilla 0.9.85 [Firefox 3.5.4/20091016092926]")
01:01:57kugelJdGordon|: that's since configurable scrollbar
01:02:09lazarusowill rockbox install on the ipod touch 2g on 3.1.2 fw
01:02:16JdGordon|if its on the right it does that?
01:02:16kugelif you put it on the right and don't chose icon line selector it looks like this
01:02:40JdGordon|you know what we need here.....?
01:02:53JdGordon|a seperate user defineable viewport for the title!
01:03:17JdGordon|... or skinability in the list control!
01:03:27*JdGordon| cackles like an evil witch...
01:03:30JdGordon|viewports everywhere!!!!
01:04:45LambdaCalculus37lazaruso: No.
01:05:11kugelBob_C: Bad boot sector signature
01:05:57kugelsd_transfer_sectors() fails apparently
01:06:16kugelmaybe readding SD_DEBUG helps :/
01:07:10 Quit TheSeven (Nick collision from services.)
01:07:17kugelthat sd debug is so bloody useless :p
01:07:26 Join The_Seven [0] (n=theseven@rockbox/developer/TheSeven)
01:07:28kugelexcept for keeping code working that is ;)
01:07:38 Nick The_Seven is now known as TheSeven (n=theseven@rockbox/developer/TheSeven)
01:08:21kugelthe eabi build boots!
01:08:33kugeland even plays music
01:08:59kugelstrange, the codec api should be incompatible, and I didn't compile eabi-codecs yet
01:09:28***Saving seen data "./dancer.seen"
01:10:50 Quit TheSeven (Nick collision from services.)
01:10:54 Join AndyIL [0] (n=pasha_in@
01:11:05 Join The_Seven [0] (n=theseven@rockbox/developer/TheSeven)
01:11:17 Nick The_Seven is now known as TheSeven (n=theseven@rockbox/developer/TheSeven)
01:11:44 Quit AEnima1577 ("Leaving.")
01:12:18kugelif the mini2440 runs it with so little modifications, F/X will make it even better
01:12:30kugelI think we should add a arm eabi option to
01:12:50kugelof course tagged with "under construction and may not work at all"
01:13:16kugelinterestingly, the fuze eabi build suffers from sd failure too
01:15:48 Join phanboy4 [0] (
01:15:51 Quit TheSeven (Nick collision from services.)
01:16:07 Join The_Seven [0] (n=theseven@rockbox/developer/TheSeven)
01:16:18 Nick The_Seven is now known as TheSeven (n=theseven@rockbox/developer/TheSeven)
01:16:26 Quit JdGordon| ("Miranda IM! Smaller, Faster, Easier.")
01:16:39kugeloh, the eabi build doesn't cope with the enum statusbar_values global_settings member
01:16:41 Quit antil33t (Read error: 104 (Connection reset by peer))
01:18:04kugelsaratoga: the dbop pin is an interesting find
01:19:21 Quit TheSeven (Nick collision from services.)
01:19:39 Join The_Seven [0] (n=theseven@rockbox/developer/TheSeven)
01:19:51 Nick The_Seven is now known as TheSeven (n=theseven@rockbox/developer/TheSeven)
01:22:03 Quit AndyI (Read error: 110 (Connection timed out))
01:25:42 Quit ender` (" Never say "Oooops" ... always say "Ahhh, interesting..."")
01:26:59 Join Sajber^ [0] (
01:27:59kugelBob_C: I haven't looked at the data sheet, but it sd insertion is on GPHDAT bit 8, can we use that?
01:30:31 Quit TheSeven (Nick collision from services.)
01:30:47 Join The_Seven [0] (n=theseven@rockbox/developer/TheSeven)
01:30:59 Nick The_Seven is now known as TheSeven (n=theseven@rockbox/developer/TheSeven)
01:32:25 Join [TiZ] [0] (
01:33:09[TiZ]Hi. I'm working on a custom statusbar. Is there any way to draw text over an image?
01:33:10Bob_Csure, isn't there a function already?
01:34:11kugelnot really
01:35:40kugelweeh, fs corruption
01:35:57kugelBob_C: I mean interrupt to detect ejection and insertion
01:36:26Bob_Cthe sd_thread is supposed to do that i think
01:36:43 Quit Thundercloud (Remote closed the connection)
01:38:27Bob_CH8 is write protect switch
01:39:56Bob_CThe SD card does not do write protect, the host is supposed to read the swicth
01:40:24kugelso, that doesn't indicate inserted at all?
01:40:56Bob_CCD_SD is the card detect SD switch, on G8
01:41:21kugelah I see, I only saw H8 and didn't look at others
01:41:39 Quit DerPapst ("Leaving.")
01:41:53Bob_CIn the original code the sd_thread monitors the card detect, but I didn't bother with most of the Hotswap code
01:42:01kugelbut well, we're currently not masking interrupts it seems, and the isr is #if 0'd
01:42:26kugelyea, I'm basically talking about hotswap ;)
01:42:40kugelit's a pitty that rockbox kept loading while I ejected the card :p
01:42:47Bob_Cthere is no CD ISR?
01:43:45 Join AEnima1577 [0] (
01:44:11Bob_CThe architecture doesn't allow a single drive and hotswap
01:44:15kugelthere's so many #if 0 code in that file ...
01:44:39Bob_Cyeah, but the comments explain everything ;)
01:44:58kugelBob_C: well, we could hack and don't actually unmount. but at least remount on reinsertion (and possibly detect that rockbox.mini2440 changed)
01:45:03Bob_Cer, apart from the ones still in my head...
01:45:23Bob_CWhy are you taking the card anyway? That is bad.
01:45:48kugelI'm used to do that...all my other targets have hotswap
01:46:16Bob_CNormally rockbox firmware would be on the fixed (internal) drive?
01:46:30*kugel wonders why sd_init_card() is no sd-generic function
01:47:51Bob_CA lot of the functions are nearly generic, but with target bits sprinkled through
01:49:48Bob_Ci think sd_init_card has one target specific line to set the clock rate
01:50:54kugelthat could be solved
01:50:59 Quit [TiZ] ("Leaving")
01:51:43Bob_Cthe sd.c driver doesn't really do much in the way of sd driving
01:52:02 Join brn2dth [0] (n=chatzill@
01:52:59Bob_CI think it could be usefully refactored
01:53:19*kugel could imagine rockbox enters some mode similar to ums usb mode
01:53:44kugelbasically wait until the storage is available again
01:54:52Bob_CLike DOS? "Please insert SYS disk" or whatever
01:55:02 Quit brn2dth (Client Quit)
01:55:55kugelI've never seen that, but probably
01:56:57Bob_CAh, the good old days, IBM PC with twin floppy and no HDD
01:57:15kugelI'm too young for that stuff :p
01:58:08Bob_CIt was only 1980 ;)
01:59:04Bob_COr was it 85...
02:00:17kugelany idea why transfers are so slow ?
02:01:27Bob_CYes, the question is why doesn't it work without delays. We need to sort that one.
02:02:12Bob_CDomonoky was looking at it recently, but failed to solve it as well
02:02:23kugeltiming issues are bad
02:02:43kugelwe should get FlynDice a mini2440, he's grown to an sd expert :)
02:04:42Bob_CI could maybe use a spare
02:05:38Bob_CI like to solve the SD issue, but I'd also like to work on a HDD interface PCB
02:06:36 Quit LambdaCalculus37 ("Leaving")
02:07:00kugelis there some fifo ?
02:07:19Bob_Cthere are lots of fifos - for the SD?
02:07:36Bob_CI think its 64 bytes on the SD
02:08:02kugellots of issues on the samsas were solved after we figured that waiting for the dma to finish isn't enough, the fifo must also be empty
02:08:53Bob_Cthe sd driver waits for the SD complete int, but not for DMA complete, I wondered if that was correct
02:09:52Bob_CIf we are unlucky its a bug in the SoC. I know of at least one in the SD controller
02:13:03Bob_CAnyway, X-files is on now, back tomorrow.
02:13:10kugelsee you
02:13:13 Quit kugel (Remote closed the connection)
02:13:47 Quit Topy ("Leaving")
02:13:54 Quit amiconn (Nick collision from services.)
02:13:56 Join amiconn_ [0] (i=quassel@rockbox/developer/amiconn)
02:14:16 Nick amiconn_ is now known as amiconn (i=quassel@rockbox/developer/amiconn)
02:16:00 Quit pixelma (Nick collision from services.)
02:16:00 Join pixelma_ [0] (i=quassel@rockbox/staff/pixelma)
02:16:20 Nick pixelma_ is now known as pixelma (i=quassel@rockbox/staff/pixelma)
02:19:41Unhelpfulkugel: any particular setting can specify that it's reset via a custom callback, right? perhaps a platform-specific function called by one of these is the problem?
02:25:49 Nick fxb is now known as fxb__ (n=felixbru@
02:26:47 Join dmb [0] (n=Dmb@unaffiliated/dmb)
02:27:57 Quit lazaruso ("Leaving")
02:28:53 Join BHSPitMonkey [0] (n=stephen@unaffiliated/bhspitmonkey)
02:34:04 Quit Stephen__ ("Leaving")
02:39:03 Join AEnima15771 [0] (
02:41:20 Quit dmb (Connection reset by peer)
02:45:45 Quit AEnima1577 (Read error: 60 (Operation timed out))
02:46:10 Quit lifeless_ (Remote closed the connection)
02:46:28 Join lifeless_ [0] (n=lifeless@
02:46:34 Quit MethoS- (Read error: 104 (Connection reset by peer))
02:54:00 Join aaron424 [0] (
02:56:56MG_ManAwww damn, I think I bricked the iriver
02:57:17MG_ManI got the EU 1.30 FW, the bootloader-h300.bin, and ran fwpatcher.exe on it
02:57:21MG_ManThen put it on the iriver root
02:57:32MG_ManDid the firmware upgrade, now it wont even turn on
03:06:41aaron424im here, but i cant help with iriver stuff, sorry
03:08:06 Join Casainho [0] (
03:08:12 Join MG_Man_ [0] (i=6165d00b@gateway/web/freenode/x-a754fde5bf69c27e)
03:08:30MG_Man_So, uh, like I said, I think I bricked the h320
03:09:04MG_Man_I got the EU 1.30 firmware, the bootloader-h300.bin, ran fwpatcher.exe, put it in root, then did firmware update
03:09:07MG_Man_now it won't turn on
03:09:14 Part MG_Man ("FLEEING...")
03:09:32***Saving seen data "./dancer.seen"
03:10:58evilnick_Were there any error messages when you upgraded the f/w?
03:11:07evilnick_Or when patching the f/w
03:11:11MG_Man_I didn't see any
03:12:07MG_Man_If there were error messages when upgrading, they went away fast because now the screen is blank, and I did not touch it
03:12:22MG_Man_so is that it then? it's toast?
03:12:38evilnick_Try pressing the reset button and then turn it on again. I'm guessing that there's no light on the display right now
03:13:07MG_Man_Wait what, it turned on, but earlier I did everything I could...odd
03:13:15MG_Man_I had to touble click on or something
03:13:32MG_Man_anyway, it sais 1.29 EU even though it was 1.30
03:13:56MG_Man_Also, there's no rockbox bootloader first, it goes right into iriver
03:14:20evilnick_Sounds like it didn't flash successfully then.
03:14:57MG_Man_oh well, sure am glad it isn't bricked, yet
03:16:37MG_Man_so according to the manual, the proper procedure is: get firmware (Non USA), get correct bootloader, run fwpatcher on the h300.hex that you recently got?
03:16:48MG_Man_Or does it have to be the current version the player is on?
03:18:07evilnick_the h300.hex file *is* the firmware, so it's not important which version it is.
03:18:30evilnick_Although I'd guess that going to a more recent f/w is a useful tactic
03:18:35rasherMG_Man_: You should use Rockbox Utility
03:19:24MG_Man_RB-Util patches the firmware to a compatible version too?
03:19:29rasherThere's nothing gained in using fwpatcher except more chances to mess up
03:19:45MG_Man_Because before hand it was 1.08 USA
03:20:11MG_Man_Okay then, I'll upgrade firmware to unpatched 1.30 EU then instead of a fwpatcher one
03:20:16MG_Man_then use rbutil
03:20:37MG_Man_Maybe I'll stick with 1.29, since thats what the iriver screen says it is anyway
03:21:27MG_Man_oh, or its because I DID end up grabbing 1.29, irregardless, I doubt I'll use the IFW anyway
03:21:53MG_Man_so doing this shouldn't likely mess it up, should it?
03:22:26rasherRockbox doesn't care which firmware version you use
03:23:05rasher(as long as it's a reasonably "new" one)
03:23:20MG_Man_I meant, is there a significant chance that I could ruin the iriver if I upgrade from an fwpatcher.exe patched firmware to an unpatched one before using rbutil?
03:23:20 Quit aaron424 (Read error: 60 (Operation timed out))
03:23:52MG_Man_Okay then
03:24:15MG_Man_Alright, it's doing it's thing
03:24:58MG_Man_okay, success, rbutil time
03:26:56MG_Man_I'll have to tackle that tomorrow, so see you guys later
03:27:14 Part MG_Man_
03:27:54 Quit JdGordon ("Leaving.")
03:29:53 Join JdGordon [0] (
03:40:00 Quit Casainho (Read error: 113 (No route to host))
03:44:17 Join CaptainKwel [0] (
03:50:04 Quit AaronM ("Emo Time In My Corner... //_-")
03:53:45 Quit JdGordon (Read error: 110 (Connection timed out))
03:58:20 Join FOAD_ [0] (
04:08:13 Join GeekShadow [0] (n=Antoine@reactos/tester/GeekShadow)
04:08:16 Quit goffa (Read error: 54 (Connection reset by peer))
04:08:48 Join goffa [0] (n=goffa@
04:09:02 Quit GeekShad__ (Read error: 104 (Connection reset by peer))
04:09:09 Quit Rondom (Nick collision from services.)
04:09:19 Join Rondom [0] (
04:14:30 Quit FOAD (Read error: 110 (Connection timed out))
04:14:31 Nick FOAD_ is now known as FOAD (
04:14:40 Quit adiroiban (Read error: 60 (Operation timed out))
04:15:00 Quit TheSeven (Nick collision from services.)
04:15:18 Join The_Seven [0] (n=theseven@rockbox/developer/TheSeven)
04:15:30 Nick The_Seven is now known as TheSeven (n=theseven@rockbox/developer/TheSeven)
04:25:13 Join Blue_Dude [0] (n=chatzill@rockbox/developer/Blue-Dude)
04:30:56 Join Rob2223 [0] (n=Miranda@
04:31:58 Join Horschti [0] (n=Horscht2@xbmc/user/horscht)
04:38:02 Quit Rob2222 (Read error: 60 (Operation timed out))
04:38:15 Quit Horscht (Read error: 60 (Operation timed out))
04:40:42 Part AB3JU
04:49:00 Quit amiconn (Read error: 104 (Connection reset by peer))
04:49:16 Join amiconn [0] (i=quassel@rockbox/developer/amiconn)
04:49:38 Join Horscht [0] (n=Horscht2@xbmc/user/horscht)
04:49:59 Join Rob2222 [0] (n=Miranda@
04:50:39 Quit Kopfgeldjaeger (Read error: 104 (Connection reset by peer))
04:50:43 Quit panni_ ("( :: NoNameScript 3.81 :: )")
04:51:19 Join Kopfgeldjaeger [0] (
04:56:16 Quit Rob2223 (Read error: 60 (Operation timed out))
04:56:27Blue_DudeNote to log: I am committing a change to pcmbuf and playback that duplicates current code during track transitions, but written more clearly and efficiently to show what's happening. I plan to give this more attention later because I'm not fully satisfied with the way track transitions are handled now, but at least this way I'm not introducing functional changes or new bugs.
04:57:33 Quit GeekShadow ("The cake is a lie !")
04:58:35CIA-8New commit by blue_dude (r23506): Clarify track transition code in pcmbuf and playback. No functional changes yet.
05:05:13 Quit Horschti (Read error: 113 (No route to host))
05:07:28 Join Horschti [0] (n=Horscht2@xbmc/user/horscht)
05:07:46 Quit Blue_Dude ("ChatZilla 0.9.85 [Firefox 3.5.4/20091016092926]")
05:08:59 Quit Horschti (Read error: 104 (Connection reset by peer))
05:09:25 Join Horschti [0] (n=Horscht2@xbmc/user/horscht)
05:09:35***Saving seen data "./dancer.seen"
05:23:24 Join FOAD_ [0] (
05:23:41 Join JdGordon [0] (n=jonno@rockbox/developer/JdGordon)
05:24:50 Join Tech-Guy [0] (
05:25:05 Quit Horscht (Read error: 113 (No route to host))
05:28:56 Quit FOAD (Read error: 60 (Operation timed out))
05:28:56 Nick FOAD_ is now known as FOAD (
05:31:11 Quit r0b- (Read error: 110 (Connection timed out))
05:33:24 Quit lifeless_ (Remote closed the connection)
05:33:43 Join lifeless_ [0] (n=lifeless@
05:35:19JdGordonam I going to end up pissing off everyone if the current statusbar is removed without supplying a skinned one which exactly mimics it?
05:38:46 Quit Tech-Guy (Read error: 104 (Connection reset by peer))
05:41:24 Quit Horschti ("Verlassend")
06:10:55CIA-8New commit by jdgordon (r23507): Fix viewport "UI area" to work based on what was agreed on last week in the mailing lists and IRC: ...
06:16:53*JdGordon doesnt think many people use the statusbar at the bottom of the display and have a wps with the %we tag...
06:24:31CIA-8New commit by jdgordon (r23508): Fix the WPS to work properly with the statusbar for the previous change.. Side effect is it fixes FS #10721 where %pb didnt work
06:27:39CIA-8New commit by jdgordon (r23509): fix charcell yellow
06:35:04 Join brn2dth_ [0] (n=chatzill@
06:35:15 Quit fyrestorm ("lamers envy me like they envy bill g -- main boot xp, just the way it should be!")
06:35:24 Quit brn2dth_ (Remote closed the connection)
06:36:23*JdGordon curses at checkwps
06:37:12 Quit BHSPitMonkey (Success)
06:40:11CIA-8New commit by jdgordon (r23510): Fix checkwps. Reimplement viewport_set_fullscreen() and viewport_set_defaults() to always return the "full display". This doesn't change anything ...
06:43:26 Join BHSPitMonkey [0] (n=stephen@unaffiliated/bhspitmonkey)
06:44:31 Quit Hadaka (
06:44:31 Quit jon-kha (
06:44:31 Quit shodanX (
06:44:31 Quit YPSY (
06:44:31 Quit Slasheri (
06:44:51NJoinHadaka [0] (n=naked@
06:44:51NJoinjon-kha [0] (i=jon-kha@
06:44:51NJoinshodanX [0] (
06:44:51NJoinYPSY [0] (n=ypsy@
06:44:51NJoinSlasheri [0] (i=miipekk@rockbox/developer/Slasheri)
06:55:27 Join Hilikus [0] (n=hilikus@unaffiliated/hilikus)
06:55:39Hilikushey guys
06:55:44Hilikusis there any podcast agregator for rockbox?
06:58:45JdGordonrockbox just plays your mp3/whatever files... so you dont need one... use any that will just copy the files to your DAP
07:02:33Hilikusi see
07:04:03 Quit CaptainKwel (Remote closed the connection)
07:09:38***Saving seen data "./dancer.seen"
07:14:40 Join dmb [0] (n=Dmb@unaffiliated/dmb)
07:28:22 Quit Hilikus ("Suerrrrrrte")
07:30:50 Quit bluebrother (Nick collision from services.)
07:30:53 Join bluebroth3r [0] (n=dom@rockbox/developer/bluebrother)
07:34:00 Join fejfighter [0] (
07:38:49 Join kugel [0] (n=kugel@rockbox/developer/kugel)
07:39:12kugelJdGordon: why don't you just move it into the __PCTOOL__ part of checkwps?
07:39:38JdGordonthought you had to go to uni? :)
07:39:58kugelI'm there now ;)
07:40:40JdGordonthis was the simpler change
07:40:53JdGordoncheckwps doesnt care about the crap with the tatusbars and stuff
07:41:13kugelUnhelpful: I think the problem is that enum doesn't have a very defined size with eabi anymore. settings reset seems to simply cast it to int
07:44:39kugelJdGordon: I find the comment in viewport.c confusing (the one that talks about overlapping)
07:45:14JdGordonwhich part?
07:45:31 Quit liar (Read error: 113 (No route to host))
07:45:59kugelthe whole comment
07:53:08JdGordonI'm not sure how to unconfuse you...
07:54:06 Join alexbobp [0] (n=alex@
07:54:45 Join ruckus1 [0] (n=Earworm@
07:54:57ruckus1hey ey everyonbe
07:55:32pixelmaJdGordon: just a quick idea about the quality vs. bitrate settings for the rec screen status bar - would a "does setting x exist" tag be possible?
07:55:48pixelmamaybe this could also help in other situations
07:55:49ruckus1i've been reading around the internets, and on the site itself, and i am not clear only on one thing. is it possible to install 160gb drive in ipod 5.5 ?
07:55:58JdGordonpixelma: sure it would be possible
07:56:06CIA-8New commit by kugel (r23511): Reorder a bit in viewport.c so that we don't need stubs for checkwps.
07:56:45pixelmaJdGordon: I mean - without much overhead and easy to use later?
07:58:26JdGordonwe could possibly change %St to have different meaning if the setting doesnt exist
08:00:11ruckus1anyone knows?!
08:00:12pixelmasounds a bit confusing at first glance
08:02:21JdGordonsure is :)
08:07:57ruckus1hey JdGordon, you weren't talking to me earlier, right? just making sure...
08:12:47 Quit kugel (Read error: 110 (Connection timed out))
08:13:05JdGordonkugel: you got red!
08:14:05CIA-8New commit by kugel (r23512): Remove wps_data.remote_wps. It's redundant with the new global curr_screen.
08:16:47 Quit Dhraakellian (Read error: 104 (Connection reset by peer))
08:17:30 Join KBH [0] (
08:20:45CIA-8New commit by kugel (r23513): Fix player yellows and reds. I'm not sure why viewport.c was never compiled for charcell.
08:21:04JdGordonmy fix was cleaner :)
08:21:41 Join kugel [0] (n=kugel@
08:22:17JdGordonremoving ->remote_wps may be a bad idea...
08:24:05 Join HBK- [0] (
08:26:32 Join JdGordon| [0] (
08:26:57kugelwell, currently it was rather a mess and created a lot of #ifdefs. much cleaner code with curr_screen
08:27:17 Quit KBH (Read error: 60 (Operation timed out))
08:27:25JdGordonthats not exactly pretty either though
08:28:15CIA-8New commit by kugel (r23514): Forgot to remove wps_data.remove_wps in checkwps too.
08:29:55 Quit JdGordon| (Client Quit)
08:31:10JdGordonpixelma: actually.... %?Se|setting|<%?St|setting|<some|value>> I tihnk looks worse than %?St|setting|<doesnt exist|value|next|value>
08:31:13 Part ruckus1 ("Leaving.")
08:32:20kugelpixelma: now albumart on remotes should work out of the box
08:32:23 Join ender` [0] (
08:35:15 Quit HBK (Read error: 110 (Connection timed out))
08:35:19 Nick HBK- is now known as HBK (
08:35:20 Join Casainho [0] (n=chatzill@
08:35:23kugelJdGordon: we wouldn't need curr_screen if simply gui_wps instead of wps_data was passed to the tag parsers
08:35:47JdGordonyeah, the api could be cleaned up a bit...
08:35:51JdGordonbut later
08:37:12pixelmaJdGordon: the thing with the latter is that it's not obvious for the themer with only one target and not being aware that the theme is usable on other targets where the setting doesn't exist
08:38:05JdGordonI'm not sure how the first one helps either... or really what the problem is?
08:39:11pixelmaoh well, your way might work if the "doesn't exist" part is mandatory
08:39:49JdGordonwhich will break all current themes
08:39:58kugel%St is hardly used
08:39:58JdGordonalthough I doubt many use this currently
08:40:19kugelwe really need statistics on the theme site about tag usage
08:42:56pixelmathe problem I thougt this would help solving was the difference between the hwcodec and swcodec rec screen status bar (because on hwcodec you can chose recording qualiity not the bitrate (files are vbr and the MAS is responsible for the average bitrate based on the quality setting) and the bitrate setting that's in place for it on swcodec targets)
08:43:36pixelmathought it can help in other places too
08:45:05JdGordonI tihnk I've slightly changed my plans for that....
08:45:42JdGordonI'm tihnking seriously about actually getting skins working properly there, then hardcode a replacement bar into that screen so hwcodec/swcodec isnt an issue anymore
08:46:17pixelmachange one hardcoded thing to another?
08:46:41 Join Utchybann [0] (
08:47:00pixelmaby the way, the rec screen once used the standard status bar until someone decided to move recording specific things from the rest of the screen to the bar
08:52:23 Quit HBK (Read error: 104 (Connection reset by peer))
08:54:17 Join JdGordon| [0] (
08:54:19 Join einhirn [0] (
08:55:04 Quit einhirn (Read error: 104 (Connection reset by peer))
08:55:05 Join maruk [0] (
08:56:11 Join Zagor [242] (n=bjorn@rockbox/developer/Zagor)
08:58:16 Join HBK [0] (
08:59:05 Join Rob2223 [0] (
09:02:14 Quit JdGordon| ("Miranda IM! Smaller, Faster, Easier.")
09:03:55JdGordonpixelma: well yes, but im trying to get rid of the statusbar, replacing that with a hardcoded skin is cool
09:06:03*kugel likes the idea of %St|setting|<does not exist|1|2|3>
09:06:43JdGordonthat will fail the parser :)
09:07:01kugelthe current one yes
09:07:11JdGordonno... even with the change....
09:07:14JdGordonyou're missing the ?
09:07:37JdGordonthe problem with that is most settings are universal
09:07:53kugelhow's that a problem?
09:08:13 Join B4gder [0] (
09:08:16JdGordon|<|item|> looks wierd
09:08:30kugelthat's not a real problem...
09:08:57JdGordonnot a real one... no...
09:09:40***Saving seen data "./dancer.seen"
09:11:10 Join flydutch [0] (
09:16:12 Quit Rob2222 (Read error: 110 (Connection timed out))
09:18:33 Join einhirn [0] (
09:20:03 Quit Casainho (Read error: 113 (No route to host))
09:25:03 Join Thundercloud [0] (
09:29:53 Join petur [50] (n=petur@rockbox/developer/petur)
09:33:08pixelma%?St already exists (but doesn't work properly with negative values) such as bass and treble
09:40:30JdGordonthats another issue...
09:47:05 Join antil33t [0] (
09:48:03 Quit Thundercloud (Remote closed the connection)
09:49:32 Quit kugel (Read error: 110 (Connection timed out))
09:50:23 Join kugel [0] (n=kugel@rockbox/developer/kugel)
09:52:48 Quit kugel (Remote closed the connection)
09:57:24 Quit swilde (Read error: 104 (Connection reset by peer))
09:59:19 Join kugel [0] (i=kugel@rockbox/developer/kugel)
10:00:35 Quit faemir ("Leaving")
10:01:46 Join MethoS- [0] (n=clemens@
10:04:21 Quit MethoS- (Remote closed the connection)
10:11:51 Quit reid04 (Read error: 110 (Connection timed out))
10:12:23 Join reid04 [0] (
10:13:15 Join antil33t1 [0] (
10:13:22 Quit antil33t (Nick collision from services.)
10:23:40 Join seani_ [0] (n=seani@
10:24:45 Join DerPapst [0] (
10:42:34 Quit BHSPitMonkey (Read error: 113 (No route to host))
10:53:04 Quit seani_ ("Leaving")
10:53:14 Quit seani ("Leaving")
10:53:43 Join seani [0] (n=seani@
10:58:24 Quit Unhelpful (
10:58:24 Quit Tristan (
10:58:24 Quit pjm0616 (
10:58:24 Quit rphillips (
10:58:24 Quit Kohlrabi (
10:58:49NJoinUnhelpful [0] (n=quassel@rockbox/developer/Unhelpful)
10:58:49NJoinTristan [0] (i=tristan@
10:58:49NJoinpjm0616 [0] (n=user@
10:58:49NJoinrphillips [0] (
10:58:49NJoinKohlrabi [0] (
11:01:57 Quit Kohlrabi (
11:01:57 Quit rphillips (
11:01:57 Quit Unhelpful (
11:01:57 Quit Tristan (
11:01:57 Quit pjm0616 (
11:02:21NJoinUnhelpful [0] (n=quassel@rockbox/developer/Unhelpful)
11:02:21NJoinTristan [0] (i=tristan@
11:02:21NJoinpjm0616 [0] (n=user@
11:02:21NJoinrphillips [0] (
11:02:21NJoinKohlrabi [0] (
11:03:47 Quit Kohlrabi (
11:03:47 Quit rphillips (
11:03:47 Quit Unhelpful (
11:03:47 Quit Tristan (
11:03:47 Quit pjm0616 (
11:04:11NJoinUnhelpful [0] (n=quassel@rockbox/developer/Unhelpful)
11:04:11NJoinTristan [0] (i=tristan@
11:04:11NJoinpjm0616 [0] (n=user@
11:04:11NJoinrphillips [0] (
11:04:11NJoinKohlrabi [0] (
11:05:37 Quit Kohlrabi (
11:05:37 Quit rphillips (
11:05:37 Quit Unhelpful (
11:05:37 Quit Tristan (
11:05:37 Quit pjm0616 (
11:06:02NJoinUnhelpful [0] (n=quassel@rockbox/developer/Unhelpful)
11:06:02NJoinTristan [0] (i=tristan@
11:06:02NJoinpjm0616 [0] (n=user@
11:06:02NJoinrphillips [0] (
11:06:02NJoinKohlrabi [0] (
11:08:59 Quit Kohlrabi (
11:08:59 Quit rphillips (
11:08:59 Quit Unhelpful (
11:08:59 Quit Tristan (
11:08:59 Quit pjm0616 (
11:09:24NJoinUnhelpful [0] (n=quassel@rockbox/developer/Unhelpful)
11:09:24NJoinTristan [0] (i=tristan@
11:09:24NJoinpjm0616 [0] (n=user@
11:09:24NJoinrphillips [0] (
11:09:24NJoinKohlrabi [0] (
11:09:43***Saving seen data "./dancer.seen"
11:10:49 Quit Kohlrabi (
11:10:49 Quit rphillips (
11:10:49 Quit Unhelpful (
11:10:49 Quit Tristan (
11:10:49 Quit pjm0616 (
11:11:13NJoinUnhelpful [0] (n=quassel@rockbox/developer/Unhelpful)
11:11:13NJoinTristan [0] (i=tristan@
11:11:13NJoinpjm0616 [0] (n=user@
11:11:13NJoinrphillips [0] (
11:11:13NJoinKohlrabi [0] (
11:12:39 Quit Kohlrabi (
11:12:39 Quit rphillips (
11:12:39 Quit Unhelpful (
11:12:39 Quit Tristan (
11:12:39 Quit pjm0616 (
11:13:04NJoinUnhelpful [0] (n=quassel@rockbox/developer/Unhelpful)
11:13:04NJoinTristan [0] (i=tristan@
11:13:04NJoinpjm0616 [0] (n=user@
11:13:04NJoinrphillips [0] (
11:13:04NJoinKohlrabi [0] (
11:14:51 Quit kugel (Read error: 104 (Connection reset by peer))
11:16:14 Quit Kohlrabi (
11:16:14 Quit rphillips (
11:16:14 Quit Unhelpful (
11:16:14 Quit Tristan (
11:16:14 Quit pjm0616 (
11:16:38NJoinUnhelpful [0] (n=quassel@rockbox/developer/Unhelpful)
11:16:38NJoinTristan [0] (i=tristan@
11:16:38NJoinpjm0616 [0] (n=user@
11:16:38NJoinrphillips [0] (
11:16:38NJoinKohlrabi [0] (
11:18:04 Quit Kohlrabi (
11:18:04 Quit rphillips (
11:18:04 Quit Unhelpful (
11:18:04 Quit Tristan (
11:18:04 Quit pjm0616 (
11:18:27NJoinUnhelpful [0] (n=quassel@rockbox/developer/Unhelpful)
11:18:27NJoinTristan [0] (i=tristan@
11:18:27NJoinpjm0616 [0] (n=user@
11:18:27NJoinrphillips [0] (
11:18:27NJoinKohlrabi [0] (
11:20:11 Join Casainho [0] (
11:21:39 Quit Kohlrabi (
11:21:39 Quit rphillips (
11:21:39 Quit Unhelpful (
11:21:39 Quit Tristan (
11:21:39 Quit pjm0616 (
11:22:04NJoinUnhelpful [0] (n=quassel@rockbox/developer/Unhelpful)
11:22:04NJoinTristan [0] (i=tristan@
11:22:04NJoinpjm0616 [0] (n=user@
11:22:04NJoinrphillips [0] (
11:22:04NJoinKohlrabi [0] (
11:23:29 Quit Kohlrabi (
11:23:29 Quit rphillips (
11:23:29 Quit Unhelpful (
11:23:29 Quit Tristan (
11:23:29 Quit pjm0616 (
11:23:53NJoinUnhelpful [0] (n=quassel@rockbox/developer/Unhelpful)
11:23:53NJoinTristan [0] (i=tristan@
11:23:53NJoinpjm0616 [0] (n=user@
11:23:53NJoinrphillips [0] (
11:23:53NJoinKohlrabi [0] (
11:25:19 Quit Kohlrabi (
11:25:19 Quit rphillips (
11:25:19 Quit Unhelpful (
11:25:19 Quit Tristan (
11:25:19 Quit pjm0616 (
11:25:43NJoinUnhelpful [0] (n=quassel@rockbox/developer/Unhelpful)
11:25:43NJoinTristan [0] (i=tristan@
11:25:43NJoinpjm0616 [0] (n=user@
11:25:43NJoinrphillips [0] (
11:25:43NJoinKohlrabi [0] (
11:27:09 Quit Kohlrabi (
11:27:09 Quit rphillips (
11:27:09 Quit Unhelpful (
11:27:09 Quit Tristan (
11:27:09 Quit pjm0616 (
11:27:33NJoinUnhelpful [0] (n=quassel@rockbox/developer/Unhelpful)
11:27:33NJoinTristan [0] (i=tristan@
11:27:33NJoinpjm0616 [0] (n=user@
11:27:33NJoinrphillips [0] (
11:27:33NJoinKohlrabi [0] (
11:28:59 Quit Kohlrabi (
11:28:59 Quit rphillips (
11:28:59 Quit Unhelpful (
11:28:59 Quit Tristan (
11:28:59 Quit pjm0616 (
11:29:05 Join LayA [0] (
11:29:24NJoinUnhelpful [0] (n=quassel@rockbox/developer/Unhelpful)
11:29:24NJoinTristan [0] (i=tristan@
11:29:24NJoinpjm0616 [0] (n=user@
11:29:24NJoinrphillips [0] (
11:29:24NJoinKohlrabi [0] (
11:29:31LayAi managed too install rockbox on my nano 2g right,,,
11:29:57LayAis there any chance that i can have the original ipod theme back?
11:30:09LayAhas someone made one?
11:30:30LayAand btw getting rockbox too load on ipod nano 2g was a sinch
11:30:49 Quit Kohlrabi (
11:30:49 Quit rphillips (
11:30:49 Quit Unhelpful (
11:30:49 Quit Tristan (
11:30:49 Quit pjm0616 (
11:31:14NJoinUnhelpful [0] (n=quassel@rockbox/developer/Unhelpful)
11:31:14NJoinTristan [0] (i=tristan@
11:31:14NJoinpjm0616 [0] (n=user@
11:31:14NJoinrphillips [0] (
11:31:14NJoinKohlrabi [0] (
11:31:31LayAthnx topik ive alredy looked there :(
11:32:41 Quit Kohlrabi (
11:32:41 Quit rphillips (
11:32:41 Quit Unhelpful (
11:32:41 Quit Tristan (
11:32:41 Quit pjm0616 (
11:32:45topikthen your question was rhetorical
11:33:06NJoinUnhelpful [0] (n=quassel@rockbox/developer/Unhelpful)
11:33:06NJoinTristan [0] (i=tristan@
11:33:06NJoinpjm0616 [0] (n=user@
11:33:06NJoinrphillips [0] (
11:33:06NJoinKohlrabi [0] (
11:33:16 Quit fejfighter (Read error: 110 (Connection timed out))
11:33:39LayAi bet no 1 else hasnt hacked a 2g nano before...
11:35:39B4gderLayA: right, you must be the first
11:36:24topikthe reason for no lookalike themes could be the use of copyrighted images
11:36:41B4gderwhy would anyone want an "original ipod" theme?
11:36:43 Quit Kohlrabi (
11:36:43 Quit rphillips (
11:36:43 Quit Unhelpful (
11:36:43 Quit Tristan (
11:36:43 Quit pjm0616 (
11:37:08NJoinUnhelpful [0] (n=quassel@rockbox/developer/Unhelpful)
11:37:08NJoinTristan [0] (i=tristan@
11:37:08NJoinpjm0616 [0] (n=user@
11:37:08NJoinrphillips [0] (
11:37:08NJoinKohlrabi [0] (
11:37:16B4gderand it's easy to make a theme yourself
11:40:22 Quit Kohlrabi (
11:40:22 Quit rphillips (
11:40:22 Quit Unhelpful (
11:40:22 Quit Tristan (
11:40:22 Quit pjm0616 (
11:40:51NJoinUnhelpful [0] (n=quassel@rockbox/developer/Unhelpful)
11:40:51NJoinTristan [0] (i=tristan@
11:40:51NJoinpjm0616 [0] (n=user@
11:40:51NJoinrphillips [0] (
11:40:51NJoinKohlrabi [0] (
11:42:08LayAtrue ill actually give u that 1
11:44:12 Quit Kohlrabi (
11:44:12 Quit rphillips (
11:44:12 Quit Unhelpful (
11:44:12 Quit Tristan (
11:44:12 Quit pjm0616 (
11:44:40NJoinUnhelpful [0] (n=quassel@rockbox/developer/Unhelpful)
11:44:40NJoinTristan [0] (i=tristan@
11:44:40NJoinpjm0616 [0] (n=user@
11:44:40NJoinrphillips [0] (
11:44:40NJoinKohlrabi [0] (
11:46:53LayAgot a psp? goto
11:46:55 Quit LayA ("CGI:IRC")
11:48:00 Quit Kohlrabi (
11:48:00 Quit rphillips (
11:48:00 Quit Unhelpful (
11:48:00 Quit Tristan (
11:48:00 Quit pjm0616 (
11:48:24NJoinUnhelpful [0] (n=quassel@rockbox/developer/Unhelpful)
11:48:24NJoinTristan [0] (i=tristan@
11:48:24NJoinpjm0616 [0] (n=user@
11:48:24NJoinrphillips [0] (
11:48:24NJoinKohlrabi [0] (
11:49:50 Quit Kohlrabi (
11:49:50 Quit rphillips (
11:49:50 Quit Unhelpful (
11:49:50 Quit Tristan (
11:49:50 Quit pjm0616 (
11:50:14NJoinUnhelpful [0] (n=quassel@rockbox/developer/Unhelpful)
11:50:14NJoinTristan [0] (i=tristan@
11:50:14NJoinpjm0616 [0] (n=user@
11:50:14NJoinrphillips [0] (
11:50:14NJoinKohlrabi [0] (
11:51:14 Quit Kohlrabi (
11:51:14 Quit rphillips (
11:51:14 Quit Unhelpful (
11:51:14 Quit Tristan (
11:51:14 Quit pjm0616 (
11:51:18 Join Unhelpful_ [0] (
11:51:21NJoinKohlrabi [0] (
11:51:23 Join rphillip1 [0] (
11:51:28 Join Tristan [0] (
11:51:36NJoinpjm0616 [0] (n=user@
11:58:22 Quit fxb__ ("changing servers")
11:58:36 Join fxb__ [0] (
11:58:48 Join cas [0] (n=chatzill@
11:59:18 Quit Casainho ("ChatZilla 0.9.85 [Firefox 3.5.4/20091028153816]")
11:59:44 Nick cas is now known as Casainho (n=chatzill@
12:01:45 Join kugel [0] (i=kugel@
12:26:50 Quit phanboy4 ("Leaving")
12:39:17 Join GeekShadow [0] (n=Antoine@reactos/tester/GeekShadow)
13:09:45***Saving seen data "./dancer.seen"
13:16:39 Quit kugel (Read error: 104 (Connection reset by peer))
13:18:14 Join Lynx_ [0] (
13:23:51Unhelpful_kugel: ick. shouldn't make assumptions about enum size. it's tricky, though, as this leaves no way to set values via pointer, unless it's with a custom function per enum type. :/
13:26:08 Join aaron424 [0] (
13:36:23 Join teru [0] (
13:40:52 Nick fxb__ is now known as fxb (
13:43:25 Join dfkt [0] (i=dfkt@unaffiliated/dfkt)
13:46:33CIA-8New commit by teru (r23515): Fix volume change in wps. %mv tag keeped showing volume is changed if volume is changed in wps.
13:47:46Unhelpful_...or you can use ints with enum values.
13:48:26 Quit Casainho (Remote closed the connection)
13:49:08 Join Casainho [0] (
13:54:47 Join kugel [0] (n=kugel@rockbox/developer/kugel)
13:55:09kugelUnhelpful_: there's a gcc switch that linux uses to keep int size
13:55:21kugelor the statusbar could be just changed to int
13:55:52kugelI don't know whether it's desirable to have enum sizes dynamic
13:59:16Unhelpful_kugel: -fshort-enums to *enable* enum shrinking, manual does not mention -fno-short-enums or similar :/
14:00:13 Join Titleist026 [0] (
14:01:45TorneUnhelpful_: EABI should default to short enums, but it's fine to change it
14:01:51Titleist026hey all, just installed rockbox on my ipod and wanted to stop by and say how much i love it, keep up the good work!
14:02:09Tornethe AAPCS says that short enums are normal but platforms are allowed to define enums as being at-least-as-big-as-int if they document it that way
14:02:28Unhelpful_Torne: i don't see an option to change it... i could change it while building the compiler, i suppose...
14:02:53kugelcan we use __attribute__((align)) on enums?
14:02:55Torneisn't every -f option reversible with no- ?
14:03:25Unhelpful_is it? *usually* such options are documented, aren't they?
14:03:33Tornetry it? :)
14:03:44CIA-8New commit by teru (r23516): FS #10649: calendar: make it selectable what day is first day of week.
14:05:46 Join adiroiban [0] (
14:12:23 Quit Titleist026 ()
14:12:26 Join Blue_Dude [0] (n=chatzill@rockbox/developer/Blue-Dude)
14:15:40kugelUnhelpful_: did you find anything further about the boot failure in create_thread?
14:17:09kugelthe mini2440 boots an eabi build almost without changes needed
14:18:06kugelalso, can you post your gcc patch again that disables exceptions?
14:18:36Tornewait, do you mean the generation of the eh info section?
14:18:59kugelI mean pulling in libunwind exception handling
14:19:45TorneThat shouldn't happen.. :)
14:20:04Tornesorry, i thought you meant a different bug
14:20:22kugelbut it does happen
14:20:26Torne(gcc 4.4 on arm-eabi genreates some exception sections even when it shouldn't because it mistakenly thinks they are needed for debugging)
14:20:40Tornewhich gcc?
14:21:26Tornehm, that's the toolchain I use for everything
14:21:33Tornewhat actually happens?
14:21:43kugeldo you also compile rockbox with it?
14:22:14kugeltry it :)
14:22:39Tornei just mean, i build for a bare metal target with custom link scritps etc using it, and don't get it pulling in libunwind :)
14:22:59kugeltry it :)
14:24:30Tornehow do i get our makefiles to run arm-eabi-gcc? :)
14:25:24kugelor edit configure
14:25:50B4gderor make symlinks in a new dir to prefix the path with
14:29:10kugelTorne: you might not experience it if you have an abort() function or stub
14:30:57TorneI just tried mini2440
14:31:07Tornedies with no memory region specified for loadable section .ARM.extab
14:31:30Tornewhich is, iirc, the exact problem that I had with pycorn, which is the aforementioned gcc bug and is fixed by -fno-dwarf2-cfi-asm
14:31:36Tornei may be wrong though :)
14:31:40kugelyea, that's already part of the bug/problem I believe
14:31:44Torneit was a while ago
14:31:57kugelanyway, if you have those region (even discarded) you'll get a lot of undefined references
14:32:16TorneThey only get generated in the first place by accident
14:32:24Torne-fno-dwarf2-cfi-asm stops them existing at all
14:32:29Torneso it's not discarding
14:32:39Tornei can't find the gcc bug number but i had it here somewhere
14:32:43kugelI mean if you manually insert those in the lds
14:32:55Tornei'm trying hacking that in to see :)
14:33:08kugelbut it's nice that you know how to fix the problem we were fighting with over the last weeks :)
14:33:21Tornewell it might not be the same thing
14:33:24Tornebut it seems similar ;)
14:34:15TorneOh, no, i don't htink it is
14:34:20Tornethe cfi thing creates .eh_frame
14:35:50Tornehm, which binary has .ARM.extab?
14:37:37*Torne forgets how to make the linker be more useful
14:37:52 Quit Sajber^ (Read error: 104 (Connection reset by peer))
14:38:34*Unhelpful_ can't find where he pastebinned that
14:38:53Torne is the gcc bug I had on pycorn
14:38:55 Nick Unhelpful_ is now known as Unhelpful (n=quassel@rockbox/developer/Unhelpful)
14:39:00Tornewhich may or may not be relevant :)
14:40:00Unhelpfulfound it:
14:40:22UnhelpfulTorne: the issue is that libgcc objects are being built with -fexceptions :/, then why doesn't mine go wrong? :)
14:41:00Tornei use libgcc :)
14:41:06Tornefor mine read "things other than rockbox"
14:41:14Torneit doe sindeed go wrong if i build rockbox
14:41:28*Torne looks in his libgcc
14:41:33UnhelpfulTorne: i don't know :/
14:42:24 Join Sajber^ [0] (
14:42:40TorneHm, yes, my libgcc does indeed have lots of things with unwinding info in.
14:42:44TorneThat's.. interesting :)
14:42:56Tornesurely python must use division, somewhere.
14:43:39Unhelpfulif you're not discarding all of that, i think things will link properly, though?
14:44:05Tornei should be discarding it :)
14:44:24Tornethe final elf has no unwindings
14:44:33Torneand doesn't link to anything i didn't explicitly tell it to
14:45:29Tornemy linker script doesn't put .ARM.ex{tab,idx} anywhere, and earlier in development i was accidentally generating them in various places, which did indeed break linking in exactly this way
14:47:42 Quit Sajber^ (Read error: 104 (Connection reset by peer))
14:48:18kugel-fno-dwarf2-cfi-asm helps it?
14:48:30 Quit togetic (Read error: 110 (Connection timed out))
14:48:36TorneNo, that's not related
14:48:41TorneI tried that on rockbox, makes no difference
14:48:51 Join Sajber^ [0] (
14:48:58Tornethat's for
14:49:01kugelbut you just said you had the problems too with pycorn?
14:49:06Tornemakes -g generate bad cfi's
14:49:10TorneI thought it was the same problem bu tit's not
14:49:23Tornethe cfi problem makes spurious .eh_frame sections
14:49:27Tornein *every* object
14:49:35Torneeven helloworld.c type things
14:49:44Torne(if you compile with -g)
14:50:06Tornei had problems with .ARM.ex{idx,tab} appearing but i fixed those without changing gcc or libgcc
14:50:20Tornethey were appearing in my own binaries because i was doing stuff wrong
14:50:24Tornerockbox isn't doing that :)
14:50:32kugeldo you have abort()?
14:50:40Torneit looks like pycorn just literally doesn't link to the objects inside libgcc.a which have unwinding info
14:50:49kugelbecause that finally magically "fixes" everything
14:50:58Tornee.g. it doesn't use _divdi3
14:51:02 Join jgarvey [0] (
14:51:14Torneyes, I have an abort()
14:51:21Tornepycorn is a full hosted C implementation, i have a complete C library
14:51:53kugelthat might be the reason you aren't getting the undefined reference also
14:52:14kugeladd "char abort[]" in main.c and if compiles
14:52:53Torneit just does look like i don't need any of the functions in libgcc that have unwinding :)
14:53:05kugelI mean for compiling rockbox
14:54:11TorneThat makes no difference, it still has the extab section
14:54:32*Torne shrugs, sorry, maybe I can't help after all :)
14:54:54kugelabort fixes the lot of undefined references
14:55:02Tornei don't have any undefined references
14:55:13Torneit does seem unlikely that the entire C library and Python interpreter just coincidentally manage not to hit any of the libgcc functions which have unwinding tables, but maybe that really i sjust a coincidence
14:56:00Tornei thought i had poked it enough that there was no exception handling naywhere in my toolchain but that's clearly not true
14:56:05Torneit's just not in the bits I use ;)
14:56:31kugelI'm talking about rockbox...
14:56:53Tornei'm talking about multiple things simultaneously, i'm afraid :)
14:56:56kugelyou get a lot of undefined references if you add the sections to the lds files
14:57:35Torneyes, if you keep extab then it's going to generate references to all the functions in libunwind and also to abort.
14:58:07TorneHeh, you already have a /DISCARD/ for .eh_frame which may be why the cfi bug doesn't matter :)
14:58:11kugeland adding something that's named abort makes them go away
15:01:09Torneif i just discard those sections i get just the undefined references to the exidx linker symbols from unwind-arm.o
15:02:12Tornei forget how you get ld to tell you why it included each object
15:03:40 Join Horscht [0] (n=Horscht2@xbmc/user/horscht)
15:03:45Torneah there we go
15:05:25 Quit jgarvey ("Leaving")
15:08:04 Join evilnick_B [0] (i=0c140464@rockbox/staff/evilnick)
15:09:14kugelTorne: we are failry sure it's _udivd3 and _divd3
15:09:18 Quit HBK (Read error: 60 (Operation timed out))
15:09:49***Saving seen data "./dancer.seen"
15:10:24kugelUnhelpful: I can't find your patch either, did you by chance post the link in #rockbox-community ?
15:11:15Torneyes, it is
15:11:21Tornei meant the bits in between that, though
15:11:47Tornegrargh, crazy gcc stuff.
15:11:54 Part evilnick_B
15:12:00 Join evilnick_B [0] (i=0c140464@rockbox/staff/evilnick)
15:12:01kugelTorne: the patch I'm asking Unhelpful about IIRC patched gcc at these very bits
15:14:13 Join panni_ [0] (
15:14:30 Quit lifeless_ (Remote closed the connection)
15:15:05 Join GeekShado_ [0] (
15:15:27 Quit antil33t1 ()
15:16:27Tornekugel: yes, that's in the pastebin above, just building libgcc with -fno-exceptions
15:16:34Tornethat seems reasonable, i'm just puzzled as to why i don't have to dot hat ;)
15:17:22 Join pamaury [0] (n=pamaury@
15:18:11Tornewell, to be more specific i'm puzzled as to why division on long longs needs to call something that can unwind the stack but division of ints doesn't
15:18:27Torne_divdi3 has unwinding data, _divsi3 does not
15:19:37Torneoh, hm, that's not even longlong..
15:19:56*Torne boggles and wonders what these things even are
15:21:10pamaurygevaerts: ping
15:21:22*Torne reads libgcc source
15:23:44gevaertspamaury: gnip
15:25:35TorneAh, right
15:25:38Tornenow i get it :)
15:25:42pamaurygevaerts: hello, I have a question about MTP/MSC. If MTP were to be included in the code, there should be a menu or something where to choose between them. In which menu should I include it ?
15:26:02Torne32-bit division is all implemented in ARM asm
15:26:14Torne64-bit division is implemented in C
15:26:35Torneand python doesn't use long long.
15:27:02Torneso yes, pycorn works by coincidence because it just happens that it never uses the runtime functions which are written in C and thus have exception info
15:28:10Torneso, yah, keep doing what you're doing, sorry :)
15:29:10gevaertspamaury: I'd say near the hid enable/disable setting (added in r23322). Also, while I agree that MSC/MTP should probably be presented to the user as a choice, I think the actual internal settings should probably be two booleans (similar to the curent HID setting)
15:29:17*gevaerts isn't entirely sure though
15:29:54gevaertspamaury: can you remind us again of where your working tree is?
15:29:57 Join aaron424_ [0] (
15:30:06*gevaerts always forgets
15:30:14 Quit Sajber^ ("Leaving.")
15:30:15 Quit kugel (Read error: 110 (Connection timed out))
15:30:31pamaurymtp branch
15:31:03 Quit aaron424 (Read error: 104 (Connection reset by peer))
15:31:15 Join funman [0] (n=fun@rockbox/developer/funman)
15:32:05 Join Sajber^ [0] (
15:32:10pamauryFor know it seems to work on my linux. I've not implemented copy because it's not critical. I've also not implemented move because of the comments I put in the code.
15:33:02 Quit GeekShadow (Read error: 110 (Connection timed out))
15:33:05pamauryIt works for some thinks with Windows but it's not recognized as a MTP device, rather as a PTP device which could explain the thing. I'm not sure why.
15:33:33pamauryIt doesn't work at all under MacOSX (tested on once) but I don't have a Mac so I couldn't debug it
15:33:39pamaury*only once
15:34:37gevaertsI don't think it has to work perfectly everywhere before it goes in. It can always be disabled in the build by default
15:34:50*gevaerts will have a proper look tonight
15:35:29pamauryThere are still lots of thing to improve in the code. Of course.
15:35:30 Quit funman (Read error: 113 (No route to host))
15:37:55 Quit Sajber^ ("Leaving.")
15:39:06pamauryI have to leave, see you
15:39:11 Join Sajber^ [0] (
15:39:12 Quit pamaury ("exit(*(int *)0 / 0);")
15:45:00 Join kugel [0] (n=kugel@rockbox/developer/kugel)
15:47:23 Join funman [0] (n=fun@rockbox/developer/funman)
15:53:14 Join HBK [0] (
15:58:16 Quit AEnima15771 ("Leaving.")
16:05:11Unhelpfulkugel: <Unhelpful> found it:
16:05:19kugelyea, I saw
16:06:04Unhelpfulwe should also patch out short enums, as it *builds* with -fno-short-enum but has link warnings for libgcc
16:06:26 Join fyrestorm [0] (
16:07:55 Quit funman ("free(random());")
16:10:00 Join mcuelenaere [0] (
16:12:43 Quit fyrestorm (Read error: 54 (Connection reset by peer))
16:14:16 Join fyrestorm [0] (
16:15:13 Quit teru ("Quit")
16:17:41 Nick rphillip1 is now known as rphillips (
16:21:44 Join jgarvey [0] (
16:23:58 Join toffe82 [0] (n=chatzill@
16:28:31 Join Lss [0] (
16:28:35 Join Ayshara [0] (
16:29:01 Quit Lss (Client Quit)
16:29:15 Join Lss [0] (
16:30:24AysharaMy Iriver 20gb won't detect rockbox and i've tried both the automated client and installing it manually, nothing works
16:30:59evilnick_BAyshara: What model iriver do you have?
16:32:15*Ayshara is very frustraited
16:34:32Aysharaanything else you need to know?
16:35:10evilnick_BDid you do the UMS trick with it?
16:35:45evilnick_BThe manual (that I linked to) should have thorough instructions for installing Rockbox, so let me know which part you're having problems with.
16:36:30AysharaYes and it seems to install correctly but when I go to boot it afterwards, it doesn't detect rockbox at all
16:37:09evilnick_BWhen you switch it on, do you get a black screen with white writing on it? (which is the bootloader)
16:38:22evilnick_BSo the install didn't work. What messages did rbutil give you when you tried to install Rockbox?
16:38:57Aysharano messages at all, it looked like the install went flawlessly but....
16:40:13Aysharajust a second
16:41:25kugelUnhelpful: what about -mabi=aapcs-linux ?
16:41:37Aysharaok, back
16:42:04kugelUnhelpful: I don't think it's worth patching gcc just for that single global_settings member
16:42:05evilnick_BAyshara: You might need to run rbutil with sudo (or as an admin) - which OS are you running?
16:42:27AysharaUbuntu 9.10
16:43:22AysharaIt says the bootloader installed correctly along with everything else
16:44:28 Quit Horscht ("Verlassend")
16:45:33AysharaIt's still giving me that %^& message about the system files being missing
16:46:14evilnick_BWhat precisely does it say? It's hard to give help without seeing what's happening onscreen
16:47:28Aysharasystem files missing, to restore the system files please connect H10 to recovery tool
16:50:22gevaertsAyshara: what files are there now in the system/ directory on the player?
16:50:52Ayshara.rockbox and .Trash-1000
16:51:02gevaertsthat's not in the system directiry
16:51:50AysharaI'm looking at the player itself right now, that's all that's there
16:52:08gevaertsthere's no system directory?
16:52:57AysharaNot that I can tell
16:54:30 Join Strife89 [0] (n=michael@
16:55:02 Join funman [0] (n=fun@rockbox/developer/funman)
16:55:36gevaertsok. Try making such a directory, download and put it there. If I understand things correctly, that should make the original firmware work again. After that we can look into getting rockbox running
16:58:20CIA-8New commit by mc2739 (r23517): Fix FS #10751 - Fuze freezes when using radio while charging
16:59:30 Nick YPSY is now known as Ypsy (n=ypsy@
17:00:17AysharaIt's hanging
17:00:32 Quit Casainho ("ChatZilla 0.9.85 [Firefox 3.0.15/2009102814]")
17:01:25AysharaThe original firmware is hanging on me -_-
17:02:16 Quit Zagor ("Don't panic")
17:02:47gevaertsok, let's go for rockbox first then
17:02:58gevaertsReset the player and connect to usb again
17:03:31gevaertsthen, replace that H10_20GC.mi4 file with (the rockbox bootloader), and see if that helps
17:06:42 Join AEnima1577 [0] (
17:06:47funmanmc2739: we don't know how to detect usb on m200v4 ?
17:07:16Aysharai'll brb I have to reboot
17:07:49 Quit Ayshara ("CGI:IRC")
17:09:06 Join Horscht [0] (n=Horscht2@xbmc/user/horscht)
17:09:52***Saving seen data "./dancer.seen"
17:11:19 Join Ayshara [0] (
17:11:39Unhelpfulkugel: you could try that. the reason i suggest patching libgcc is that simply changing the short size, or the abi, will result in link warnings or errors when linking libgcc
17:12:36AysharaThe system is now giving me an i/o error when I try to copy the file given
17:13:24 Join liar [0] (n=liar@
17:14:00gevaertsin that case you either have a corrupted filesystem or a bad disk. Try fsck.vfat first
17:14:19Unhelpfulindeed, aapcs-linux does build, but produces the same link warnings. better just to make sure libgcc is built with -fno-short-enum, as well. it's not like we're not already telling it how to build libgcc
17:15:30 Quit bubsy ()
17:18:19kugelisn't it just multilib?
17:18:32kugelthat's different, it doesn't change generated code
17:18:52kugelwhat are the warnings btw?
17:23:06 Quit TopyMobile (Remote closed the connection)
17:29:06 Join Aero [0] (n=hehe@
17:35:28 Part Aero
17:39:41 Join bertrik [0] (
17:40:40 Join stoffel [0] (n=quassel@
17:42:23 Quit beta2k (Read error: 60 (Operation timed out))
17:51:22 Part B4gder
17:54:25 Quit petur ("work->home")
17:59:07saratogaso assuming we have the right register to check for a dock connection on the fuze, what is the right way to monitor it for an insertion? just have the button driver check it and set a flag when theres a dock?
18:00:09kugelbasically yes
18:03:38bertrikif it's on the DBOP then yes, the button driver looks like the right place
18:03:48 Join MethoS- [0] (n=clemens@
18:05:30mc2739funman: not that I know of, nothing is listed on SansaAMSHardwareMappings
18:06:43 Quit Strife89 ("End of class; lunch.")
18:17:14funmanIs there a page which gives information on the mtp driver being written ? (I didn't find anything on the wiki/flyspray but I remember some people working on it on irc)
18:19:33 Join BdN3504 [0] (
18:20:13BdN3504is there any kind of documentation on how bmark files are created in rb?
18:20:50 Join beta2k [0] (
18:22:06mc2739funman: I am not finding any documentation on MTP. pamaury was/is working on it.
18:22:24funman18:22 < mc2739> funman: I am not finding any documentation on MTP. pamaury was/is working on it.
18:22:25 Join bubsy [0] (n=bubsy@
18:22:29BdN3504i have a bmark file and wanted to use it as a reference for certain track positions
18:23:32BdN3504but the numbers in there are rather cryptic to me
18:23:53BdN3504what does this mean for example
18:25:28gevaertsfunman: I only know the git repository
18:25:43funmanwhere is it ?
18:26:23gevaertsthe mtp branch
18:32:12 Quit funman ("free(random());")
18:34:12mc2739BdN3504: The bookmark file will contain N number of bookmarks in the following format: resume_index;resume_offset;resume_seed;resume_first_index;resume_file;milliseconds;MP3 Title;
18:34:19 Nick Ypsy is now known as YPSY (n=ypsy@
18:37:35 Join phanboy4 [0] (
18:38:12 Nick ch4os is now known as chaos (n=ch4os@gentoo/user/ch4os)
18:39:43 Quit phanboy4 (Client Quit)
18:40:51 Quit maruk ("Leaving.")
18:41:57 Join JdGordon| [0] (
19:02:06bertrikwow, meizu sure uses a varied mix of hardware, the M6 firmware contains initialisation code for both SDR and DDR SDRAM
19:03:05 Quit kugel (Read error: 110 (Connection timed out))
19:03:38 Quit jgarvey ("Leaving")
19:06:59 Quit beta2k (Read error: 110 (Connection timed out))
19:07:17 Quit JdGordon| ("Miranda IM! Smaller, Faster, Easier.")
19:09:54***Saving seen data "./dancer.seen"
19:14:37CIA-8New commit by blue_dude (r23518): Compressor: simplify makeup gain setting, expand release range, finally provide documention in the manual!
19:17:53 Join Grahack [0] (
19:19:25 Join JdGordon| [0] (n=Miranda@nat/microsoft/x-c0886bd22075bd5c)
19:22:22 Quit Blue_Dude ("ChatZilla 0.9.85 [Firefox 3.5.4/20091016092926]")
19:27:02 Quit BdN3504 ("CGI:IRC (Ping timeout)")
19:29:24 Join antil33t [0] (
19:30:30TheSevenbertrik: I would guess they just use some generic library or startup code, which they use on multiple devices
19:30:47TheSevenor maybe it's a remnant from dev/prototyping boards
19:31:03TheSeveni don't think there are m6 players with different ram speeds...
19:31:13*gitster pings on FS #10680 and FS #10681
19:33:08 Nick evilnick_B is now known as evilnick_\0 (i=0c140464@rockbox/staff/evilnick)
19:33:24 Nick evilnick_\0 is now known as evilnick_B (i=0c140464@rockbox/staff/evilnick)
19:33:53amiconnUnhelpful: I'd rather keep short enums
19:35:10 Join binarywise [0] (n=binarywi@
19:36:45 Part binarywise
19:38:09CIA-8New commit by learman (r23519): Updated Swedish translation and some minor fixes.
19:41:02DerPapsti just updated my beasts bootloader and now i'm stuck in a "no partition found" error and it doesn't connect to usb. any chance to backup some data of the hdd without buying a zif to ide converter?
19:41:23gevaertsDerPapst: do you have an ipod video?
19:41:44DerPapsti guess that's the answer :P
19:41:46 Join Strife89 [0] (n=michael@
19:41:56gevaertsif you feel like opening that one as well :)
19:42:05gevaertsDerPapst: doesn't the bootloader do usb?
19:42:15gevaertsah wait, you said...
19:42:34DerPapstit restarts after a while
19:42:49DerPapst"it" being the rockbox bootloader
19:44:28DerPapstbefore i updated the bootloader and rockbox everything worked. i didn't mess with the partitions or filesystem in any way. i even "safely disconnected" the beast first
19:45:15 Quit Strife89 (Client Quit)
19:46:22DerPapstdumb question but current svn bootloader is supposed to work on the S?
19:46:56gevaertsI'm not sure if anyone tested it recently, but I can't think of anything that could have broken it
19:47:37DerPapstto bad. otherwise i could have yelled at someone ;)
19:48:15gevaertsjust bisect the thing, and then yell :)
19:48:29 Quit einhirn ("Miranda IM! Smaller, Faster, Easier.")
19:48:56*DerPapst is away for a while and happy for more ideas if anybody has some.
19:49:30bertrikTheSeven, could be. They seem to configure it for SDR first and test if the SDRAM works, then configure it for DDR and test again.
19:50:02bertrikThey do use dual drivers (and various hardware) for the audio codecs and LCDs
19:50:51 Join domonoky [0] (n=Domonoky@rockbox/developer/domonoky)
19:51:18 Quit stoffel (Remote closed the connection)
19:51:23bertrikI don't have any other indications that they actually really use DDR in any of the meizu m6 devices
19:52:52*GodEater yawns
19:55:57 Join BdN3504 [0] (
19:56:06 Join peace [0] (
19:57:44BdN3504mc2739: hm mine look different: 25;682784;0;0;47125;1;0;folder;file
19:57:45peaceHi, I need a litle hand please. An old version of rockbox corrupted the memory of my e280v2. I want to reformat it and put fresh box, but windows canot read the drive. Any thought please?
19:57:58 Join Strife89 [0] (n=michael@
19:58:26peacenow it wants. nice
19:59:13peacehow big clusters have to be for rockbox?
20:04:36 Quit peace ("CGI:IRC (Ping timeout)")
20:07:56BdN3504is there no documentation on bmark files?
20:10:20 Quit domonoky (Read error: 104 (Connection reset by peer))
20:11:57 Join domonoky [0] (n=Domonoky@rockbox/developer/domonoky)
20:12:56 Quit BdN3504 ("CGI:IRC")
20:16:54 Quit Strife89 ("Switching OSes.")
20:17:04 Join kugel [0] (n=kugel@rockbox/developer/kugel)
20:21:17 Join Strife89 [0] (n=michael@
20:22:48JdGordon|who knows the database tagnavi.config stuff?
20:23:14 Join BOBdotEXE [0] (
20:27:15 Join Stephen [0] (
20:27:36 Nick Stephen is now known as Stephen__ (
20:30:24 Join phanboy4 [0] (
20:35:24bluebroth3rdomonoky: reworked mkamsboot Makefile −− comments?
20:35:43 Nick bluebroth3r is now known as bluebrother (n=dom@rockbox/developer/bluebrother)
20:36:51JdGordon|kugel: is that comment still confusing you?
20:39:39kugelhas it changed?
20:40:09domonokybluebrother: the way you create LIBOBJS looks a bit strange... but if it works for all uses, i think that fine :-)
20:41:05bluebrotherdomonoky: well, I'm just trying to avoid maintaining two lists with source files. Btw, any idea why the main function isn't in a separate main.c like for ipodpatcher?
20:41:25kugelJdGordon|: then yes
20:41:28bluebrotherthat would make it easier :)
20:42:10domonokylazyness ? :-)
20:42:11JdGordon|its supposed to say that if the viewports dont overlap, then its the users stupity and we cant do anything about it, so give the full display
20:42:41bluebrotherwell, and it compiles fine as far as I tested. Couldn't try the OS X fat binary build though.
20:42:46 Join n1s [0] (n=n1s@rockbox/developer/n1s)
20:42:58*bluebrother considers changing that main thing later
20:43:29kugelJdGordon|: but they shoudn't overlap
20:43:40kugeloverlapping is what's causing redraw problems
20:43:55JdGordon|the user setting and %Vi need to overlap
20:44:05DerPapstthe hdd in the gigabeat s is zif 40?
20:44:05JdGordon|if they dont then we are in an error condition
20:44:44kugelwait a second, you're right
20:44:51JdGordon|course I'm rigth!
20:45:07JdGordon|:) right
20:45:28kugelI was thinking of the statusbar viewport(s) not %Vi
20:45:32evilnick_BDerPapst: It's the same interface as the iPod Video, so yes.
20:45:54kugelbut, with overlap, you mean that the ui vp should be within %Vi entirely?
20:46:21DerPapstthanks. i'd rather not try to open my video because it's full of those ribbon cables.
20:47:19JdGordon|kugel: no, they just need to overlap somewhere... even 1x1 is enough
20:48:00kugelif ui vp is not completely within %vi it might overlap with statusbar viewports
20:48:20JdGordon|which statusbar are you talking about?
20:49:20JdGordon|rememebr we always set the viewport to the intersection (i.e overlap) of the setting and %Vi
20:49:27kugelthe idea of intersection was to crop the ui vp so that it's completely within %Vi, wasn't it?
20:49:56gevaertsbasically, yes
20:50:02JdGordon|yes, thats what it does
20:50:19kugelyou said the ui vp doesn't have to be within %Vi
20:50:36JdGordon|no i didnt... I said if that happens its an error state
20:51:30kugelyou said they just need to overlap, the error is when they don't overlap at all
20:52:17JdGordon|how can they overlap but not being within eachother?
20:52:27gevaertssaratoga: I heard from a usually reliable source that ipods can indeed use some sort of accessory protocol over USB. I personally suspect that it's a similar mechanism to how they do clock setting. The unnamed person who told me now has the Official rockbox usb tracer to investigate
20:52:41DerPapstjust to make sure i don't buy crap... this claims to work with ipod hdds
20:53:02*gevaerts wonders if JdGordon| has had time for this yet :)
20:53:22JdGordon|unfortunatly my video has kicked th ebucket apparnelty... which is very annoying
20:53:36JdGordon|that is high on my todo list though
20:54:57n1s... on #rockbox, see JdGordon| trying to save his video, will he be successfull or will it end in tears, stay tuned!
20:54:58kugelthe XX area is where they overlap, but neither vp is within the other
20:55:52JdGordon|the XX area is whwat the viewport will be set to then
20:56:09JdGordon|thats at least what I understood the agreement to meaning
20:56:35kugelthe commends reads as if the whole if and the block would "just make sure they overlap"
20:57:59JdGordon|the if does make sure they overlap
20:58:07JdGordon|if they dont overlap, the if will fail
20:58:12kugelthe next thing, it says "if they dont its user error, so use the full display and live with redraw problems" but it isn't doing any of that sort. if the if fails the vp is left unset entirely
20:58:47JdGordon|woops.. thats because I moved the set_fullscreen() call to the end which iirc was your suggestion :)
20:58:53 Join Casainho [0] (n=chatzill@
20:58:57JdGordon|can you fix that?
20:59:22kugelI can
20:59:46kugelcan i change the comment too?
20:59:54JdGordon|to what?
21:00:13kugelmention that the if block does intersction for example
21:01:17kugelbtw, why fullscreen if they overlap, and not %Vi or ui vp?
21:01:41kugelif they don't overlap*
21:01:57JdGordon|because none of those three will guarentee any sort of behaviour
21:01:57evilnick_BDerPapst: What are you connecting it to?
21:02:58kugelui vp respects the user setting, %Vi won't give any redraw problems. fullscreen is worst
21:03:01DerPapstthe beats hdd and a usb ide connector (for 2,5'', 3,5'' HDDs)
21:03:52JdGordon|ui respects the setting, but guarentees redraw if there was no overlap, %Vi actually makes the most sense
21:04:01JdGordon|yeah, change it to %Vi
21:04:57*kugel doesn't think this edge case has any importance anyway
21:05:16evilnick_BDerPapst: I'd be a little wary of that in case the OF decides to randomly format it
21:05:37JdGordon|kugel: having fullscreen makes it more obvious what the problem was...
21:05:42evilnick_BUnless there's a way to boot RB on the beast without the OF existing, but that'd be news to me
21:06:24BOBdotEXEis it possible to make a rockbox plugIn without recompileing the entire firmware?
21:06:45JdGordon|umm... actually
21:06:48 Quit thegeek (Read error: 113 (No route to host))
21:06:50JdGordon|make plugins might do it
21:07:13BOBdotEXEOn the documents page, it said I have to get the source code
21:07:29 Quit toens (Read error: 110 (Connection timed out))
21:07:35DerPapstevilnick_B: the rockbox bootloader doesn't go into usb mode. i started the OF once and it said something about having to recover. then i pulled the charger cable (my beast doesn't have a battery anymore since yesterday).
21:07:54kugelI'm thinking it should be done differently now
21:08:17kugelor maybe not
21:08:39domonokyBOBdotEXE: was there a question in you last sentence ? :-)
21:08:47DerPapstBOBdotEXE: actually it should only compile the module if nothing else changes (and relink everything), thanks to the dependency system.
21:09:47kugelthat you can't pass a viewport to option_screen is a problem
21:09:56***Saving seen data "./dancer.seen"
21:10:01kugelor rather to set_int and the like
21:10:22BOBdotEXEok, thanks! :)
21:10:50JdGordon|kugel: a comepltly seperate problem
21:11:41kugelnot completely. the option screen currently relies on viewport_get_current_vp() to get the viewport that the list before used
21:11:51kugelthat returns the ui vp which isn't fixed up
21:12:14JdGordon|viewport_get_current_vp() was a bad idea (sorry.. it was)... it should never have got in
21:12:33JdGordon|change option screen to using viewport_set_defaults() and all is good
21:13:10kugelI think I tried that, but that gave other problems
21:13:30JdGordon|such as?
21:13:36kugelcan't remember
21:14:44 Join thegeek [0] (
21:15:59kugelsuppose you choose an entirely different viewport for a menu in a plugin
21:16:18kugelyou pass that to do_menu(), it won't come through any set_int though
21:17:24 Quit GeekShado_ ("The cake is a lie !")
21:17:47Casainhoanyone knows if 140MHz coldfire would be equivalent to 140MHz ARM9? ::
21:17:53kugelsuch situations are in the core too I think
21:19:22JdGordon|kugel: ok, thats a seperate issue... set_int() should accept a viewport
21:20:29kugelI think you disagreed with that when I had the problem for the first time
21:20:48*JdGordon| doesnt remember
21:21:25saratogaCasainho: depends which arm9
21:22:05JdGordon|kugel: I probably disagreed for a different reason... *all* "controls" should accept a viewport
21:22:22JdGordon|the set_*(), option_select(), lists, etc
21:22:22CasainhoARM Cortex M3? −− because I was thinking if Phillips LPC Cortex M3 running at 100MHz would work for a audio player device and run Rockbox...
21:22:38JdGordon|Casainho: one board at a time would kill you?
21:23:36CasainhoJdGordon|: why not? There are small boards using cheap ARM Cortex M3, they even aren't BGA...
21:23:57Casainhothat MPIOHD200 is a cheap device for sure...
21:27:58Bob_Cmy rule of thumb is around 50 MIPS for a sw codec, but the real way to find out is get a dev board and try it
21:31:47Bob_Cwhich reminds, is there a way to report CPU usage in Rockbox?
21:34:54kugelyou can log boosting, and watch some debug screens for the current frequency
21:35:05kugelthe stack debug menu might be interesting too
21:35:26domonokykugel: mini2440 doesnt do boosting at moment :-)
21:36:35Bob_CI can implement that pretty easily
21:37:36domonokywe often use boost ratio to measure codec performance, i think at moment there is no other way to measure cpu usage.
21:38:16Bob_CIs there an "idle" task?
21:41:49 Join bmbl [0] (n=Miranda@unaffiliated/bmbl)
21:42:47JdGordon|one for each core apparently
21:44:16Bob_Cok, so if I create a CPU meter is there a way to show it on the wps?
21:45:30domonokyJdGordon|: really ? i thought we didnt have a idle thread, but maybe that changed with the ipods (dual.core)
21:45:52JdGordon|my mini2g has 2 idle threads in the OS stacks listing
21:46:11JdGordon|Bob_C: sure
21:46:27JdGordon|fix it up to return a percentage and we can do anything with it
21:46:43n1show long is a cpu meter anyway? ;)
21:46:51*domonoky cant see a idle thread on his m200v4, lets check other targets :-)
21:46:52 Quit Sajber^ (Read error: 104 (Connection reset by peer))
21:47:16JdGordon|it might only be COP-ed targets then?
21:47:18Bob_Cwel, 1ns = 30cm, if that helps
21:47:33JdGordon|it is?
21:48:01Bob_Cfor a photon, electrons are a bit slower
21:48:07 Join froggyman [0] (
21:48:26saratogaCasainho: i've never heard of the M3 before but it looks like a very good choice for an MP3 player
21:48:40saratogabasically an ARM7TDMI with a fast multiplier
21:48:52saratogaso its probably comparable to coldfire
21:49:23Bob_Capparently the latest gcc has support for some of these Cortex chips
21:49:57Casainhosaratoga: and look here a board using that ARM Cortex M3 at 100MHz:
21:50:50Casainhosaratoga: but that one do not have external access to external SDRAM, I must find one that have.
21:50:50bertrikIIRC, the cortex m3 does thumb only
21:50:56 Join AaronM [0] (
21:51:11Casainhobertrik: means that is slow?
21:52:03saratogaoh thumb only is odd
21:52:11domonokyJdGordon|: looks like only dual-core targets have idle threads. single core targets just sleep the core if there is nothing shedule.
21:53:03 Join Computer [0] (n=Computer@unaffiliated/computer)
21:53:20saratogai don't know what thumb2 is
21:53:34bertrikCasainho, not sure, I think thumb instructions are not as powerful as 32-bit arm, but you can fit more of them in the cache
21:54:17Bob_Cthe thumb architecture is basically a 16 bit instruction set
21:55:08saratogathumb2 looks like a 32 bit instruction set with variable sized instructions
21:55:51saratogabasically the reference card has all the thumb1 stuff as 16 bit ops, and a lot of the armv5e stuff as 32 bit opps
21:55:52 Quit phanboy4 ("Leaving")
21:56:01 Join Sajber^ [0] (
21:56:17Bob_Cactually I think it is 16 bit plus some essential 32 bit instructions (because you cant switch into full 32 bit ARM mode)
21:56:33Bob_Cbut I could be wrong ;)
21:57:05Computeri have an ipod classic oh god what do i do
21:57:30saratogagive up
21:57:49saratogaor i guess get to work on a port
21:58:37saratogaif thumb2 has gcc support and access tot he 32x32=64 bit mul instructions (MULL, MULAL, etc) then i think it would be plenty fast
21:58:42CIA-8New commit by bluebrother (r23520): Clean up mkamsboot building. No functional changes. ...
21:58:46saratogabut i don't know much about it
22:00:13bluebrotherComputer: no idea what you do with your Ipod. I usually use it for listening to music.
22:01:17Computeris there any way to play ogg and/or flac on this thing?
22:02:52Computeraaaaahh i would have bought a different player if i had known
22:02:56Computershould i have looked it up
22:03:09domonokyComputer: yes, exchange it to another ipod where rockbox runs on :-)
22:03:27Computeri am so upset :<
22:03:38Computeri need some cake
22:04:56*Torne gives Computer a slice of cake: <)
22:07:07*bluebrother wonders when computers started eating *g*
22:08:22CIA-8New commit by kugel (r23521): Fix a possiblity where viewport_set_default left the viewport unchanged. Improve comments a bit.
22:27:42 Join LambdaCalculus37 [0] (n=LambdaCa@rockbox/staff/LambdaCalculus37)
22:29:53 Quit AaronM ("Emo Time In My Corner... //_-")
22:30:37 Quit Grahack (Client Quit)
22:31:15 Nick aaron424_ is now known as aaron424 (
22:31:32bluebrotherdomonoky: any idea why telechips.c is built from instead of being put into libmktccboot.a?
22:33:14domonokyno idea
22:33:40*bluebrother considers changing that
22:34:50bluebrotherbtw, what do you think about mktccboot not only using a source from outside (which is nothing special) but also compiling it outside its folder?
22:35:06bluebrotherkinda breaks out-of-tree building
22:36:19domonokydoes it really build outside ?
22:37:01domonokythe telechips.o rule puts the .o file into the $OUT dir. but the clean rule is wrong.
22:37:53bluebrotherhmm right, it's the clean rule.
22:38:19gitsterAnybody familiar with firmware/drivers/rtc/rtc_as3514.c and has access to both AMS and non AMS Sansa?
22:38:25domonokybut linking all into one .a file would make sense.
22:38:28bluebrotherthough I'm about to change that Makefile similar to the changes to the mkamsboot one
22:38:34Tornetalking of building, why does the stuff in tools get built in tree?
22:38:41*bluebrother doing that right now
22:38:51domonokyand it also uses this -DLIB mechanism, so avoid having a seperate main.c :-)
22:39:30gitsterOn my Fuze, it is Nov 4th 01:40 am with the original firmware but rockbox shows 13:40 am (i.e. 12 hours ahead).
22:39:41gitsterI am wondering which one of the following is true:
22:39:50gitster(1) SECS_ADJUST for CONFIG_CPU==AS3525 in rtc_as3514.c is wrong to assume that AMS Sansa counts from Jan 1st 1970 but it needs 12 more hours of adjustment.
22:40:09gitster(2) SECS_ADJUST in rtc_as3514.c is wrong for both AMS and non AMS Sansas and it needs 12 more hours of adjustment.
22:40:23gitster(3) SECS_ADJUST in rtc_as3514.c needs conditional not just on AMS-ness but Fuze is different from other AMS Sansa and it needs 12 more hours of adjustment.
22:40:31bluebrotherdomonoky: well, that will go the same time :)
22:40:38domonokyTorne: for the rockbox build, its probably because the tools are not target dependend, so can be reused to build for different targets in one tree.
22:41:28bluebrotherbtw, I'm considering to move ipodpatcher / sansapatcher to a lib as well. Any reasons against doing so?
22:41:42 Join RalpH_himself [0] (
22:41:47Tornedomonoky: well yes, but it's not like they are big or take very long
22:42:14domonokybluebrother: no, that would be good. make sure to add those mac trickery for them aswell then.
22:43:00kugelgitster: it's correct here. are you sure you set 12h/24h display correctly?
22:45:38RalpH_himselfhey guys
22:45:59gitsterkugel: the same symptom either way with 12h/24h settings.
22:46:12bluebrotherdomonoky: ideally everything outside of the rbutil folder should be a lib IMO.
22:46:41domonokybluebrother: yes, that is a good goal, but there are still many things left..
22:48:02bluebrothersure. But someone needs to start, and I'm currently at doing some Makefile stuff, so ... ;-)
22:48:08AysharaI got the bootloader file moved to a system folder
22:48:18domonokysure, if you are in the mood for it.. do it !
22:49:45RalpH_himselffeature request: Maybe you have already heard of a "symmetric signal", which is used in PA quite often. It consists of twice the same mono signal, but one of which is inverted. Then, the difference of these two sigals is taken, which elliminates all distortion created by the cable. Many pro devices have such inputs. Would it be possible to have another "Channel configuration"? Wouldn't be too much work, just taking the mono signal a
22:49:45RalpH_himselfon one channel...
22:51:21RalpH_himselfI've done some sound programming in assembler, if nobody else does it, i will somewhen do it. But right now, I'm simply missing the time :(
22:52:40domonokyRalpH_himself: as you are perhaps the only one wanting this feature, you probably have to code it yourself.. but we are happy to help if have questions :-)
22:52:53 Quit Strife89 ("My number of files is OVER 9000!")
22:53:12domonokythings only get done, if someone wants something bad enough, so he codes it himself :-)
22:53:16kugelhrm, viewvc breaks lines that are less than 80 chars wide
22:53:35RalpH_himselfdomonoky: it just popped up right now, when i was considering the construction of an active party speaker. and the amp would have such an input.
22:54:51AysharaYAY! I got rockbox working, thanks everyone
22:55:35AysharaThat's one problem off my list
22:55:49 Quit Ayshara ("CGI:IRC")
22:56:43CIA-8New commit by bluebrother (r23522): Clean up mktccboot building. No functional changes. ...
22:57:30CIA-8New commit by bluebrother (r23523): Remove unused variable.
22:58:18 Part toffe82
23:02:56 Quit AlexP ("Please insert girder")
23:05:13 Join AlexP [0] (n=alex@rockbox/staff/AlexP)
23:09:57***Saving seen data "./dancer.seen"
23:19:36 Join toffe82 [0] (n=chatzill@
23:23:07CIA-8New commit by bluebrother (r23524): mktccboot doesn't use any version string passed by the Makefile, remove unneeded call to
23:24:00 Part RalpH_himself
23:24:48CIA-8New commit by bluebrother (r23525): Use mingw32-make when deploy-building on Windows / MinGW as make might not be the MinGW version one.
23:26:07 Join stripwax [0] (
23:28:28 Quit zu (Read error: 60 (Operation timed out))
23:28:35 Quit bzed (Read error: 60 (Operation timed out))
23:29:05 Join zu_ [0] (
23:29:33 Join bzed_ [0] (
23:29:33 Nick bzed_ is now known as bzed (
23:29:47JdGordon|kugel: you watching the "customisable statusbar" forum thread? interesting issue there (although its user error...) _set_defaults() needs to be changed to always set the user font (I think)...
23:30:43kugelit seems to me that the guy doesn't get the font thing
23:30:53kugel"It's intended, that all viewports in the sbs file are set to font 0." is wrong
23:31:26JdGordon|I tinhk he means there that the sbs should have its viewports with font 0... but the menu should be 1
23:31:40kugelbut right, set_default should probably set user font and global_settings colors
23:31:41JdGordon|the problem is when the %Vi says 0 and setting says 1...
23:31:53JdGordon|yeah, I tihnk so too
23:32:03kugeldidn't it do that before?
23:32:37JdGordon|it may have which is why that commit changed it
23:32:47JdGordon|I dont think I took those into account at all
23:33:04kugelhm, no it didn't
23:33:32kugelbut it didn't return a skin viewport (where it makes sense to not use "-" for these fields)
23:34:34kugelalthough CustomWPS says that %Vi is used if no ui vp is available
23:35:13kugelwhich sort of implies that font and colors could be taken from it too . for %Vi, anything other than "-" for font and colors don't make sense at all either
23:35:22JdGordon|right here we have a problem where the font and colours probably shouldnt be a seperate setting anymore...
23:36:15bertrikyay, I think I got sdram working on the meizu m6 sp \o/
23:37:06kugelwe can fix that or not. I don't really care. Altough I'd rather convince people to use "-" instead of filling it with dumb values that are not used anyway
23:37:18kugel(I don't consider it as a bug really)
23:38:22kugelone could actually want a colored list combined with his sbs
23:38:27JdGordon|%Vi especially should be -'s
23:38:38kugelalthough %Vi isn't designed for that
23:38:53JdGordon|so now the colours settings are redundant...
23:39:10JdGordon|if we can force the user setting to always be there
23:39:16kugelhow? "-" means use settings
23:42:18 Quit BOBdotEXE (Read error: 60 (Operation timed out))
23:43:29 Quit bertrik ("De groeten")
23:44:17 Quit LambdaCalculus37 ("off I go!")
23:46:30kugelbut viewport_set_defaults doesn't set colors if both %Vi and UI vp are set
23:47:15 Join GeekShadow [0] (n=Antoine@reactos/tester/GeekShadow)
23:54:31 Join Lynx0 [0] (
23:55:04CIA-8New commit by kugel (r23526): Get font and colors from UI viewport before intersecting, instead of leaving them empty.
23:57:34mc2739gitster: try adjusting the clock in your OF and see if that helps
23:58:01CIA-8New commit by lenzone10 (r23527): Updated italian translation and correct a little mistake in English.
23:58:50JdGordon|kugel: wait.. ui viewport? is that the setting or Vi?

Previous day | Next day