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 2020-10-25

00:08:10_bilgus__looks like there isn't a direct ioctl call but maybe writing a new brightness level for the device however that works
00:08:41speachythe only direct ioctl is on vs off, which might not do anything to the backlight anyway.
00:09:16speachydon't think there's any way to avoid polling for power/usb/battery level though.
00:09:41speachywell, usb and power yes, but battery level no.
00:11:06speachyif we were using native threads this wouldn't be as big of a deal
00:17:48_bilgus__same as anything going through a translation layer I suppose
00:21:54 Quit TheSeven (Disconnected by services)
00:22:04 Join [7] [0] (~quassel@rockbox/developer/TheSeven)
00:27:20speachyno.. we appear to the host os as a single thread, so any I/O blocks everything else.
00:34:05_bilgus__why can't we run a helper app to poll for us then?
00:34:37speachywe could, but that info has to get back to us. via a signal or shared memory
00:34:47speachyor a pipe I suppose too..
00:35:51speachyanyway. I did find two bugs in the ALSA driver but they're related to recording.
00:37:46_bilgus__I think it sounds quite plausible that the IO is hanging it but why it doesn't happen on the other targets ya got me there
00:39:00speachybecause tthe other targets are using native threading. :)
00:40:25speachy(as in, our bare-metal kernel)
00:40:44speachyit's not fully preemptible but it does well enough
00:42:47_bilgus__got it sorry I'm not really up speed on the hosted side
00:44:13speachyneither was I until relatively recently. But I have the advantage of being pretty familiar with embedded linux hackery and my earliest F/OSS contributions were sound-related. :)
00:44:28***Saving seen data "./dancer.seen"
00:45:49fs-bluebotBuild Server message: New build round started. Revision 4afd1f9, 293 builds, 9 clients.
00:46:35speachylet's see how this box performs now. :D
00:49:09speachyanyway, I'm going to pack it in. if you do get a logf dump when that playback hang is triggered, please send it over
00:49:36speachy(as soon as it hangs, dump it immediately)
00:52:36_bilgus__will do I have it built I was just off fixing other stuff sorry didn't realize you were waiting on it
01:00:13fs-bluebotBuild Server message: Build round completed after 864 seconds.
01:00:22fs-bluebotBuild Server message: Revision 4afd1f9 result: 10 errors 0 warnings
01:02:36 Quit massiveH (Quit: Leaving)
02:32:31 Quit yosafbridge (Quit: Leaving)
02:44:32***Saving seen data "./dancer.seen"
02:53:07 Join yosafbridge [0] (
03:03:17 Quit Rower (Ping timeout: 260 seconds)
03:08:51 Quit logbot (Ping timeout: 265 seconds)
03:08:51***ERROR: (Closing Link: (Ping timeout: 265 seconds)) from
03:08:51***Saving seen data "./dancer.seen"
03:08:52***Started Dancer V4.16
03:08:52DEBUGgethostbyname(2) failed for (No such device or address) (snapshot: netstuff.c line 99)
03:08:52DEBUGgethostbyname(2) failed for (No such device or address) (snapshot: netstuff.c line 99)
03:08:52***Connected to on port 6667
03:08:52***Logfile for #rockbox started
03:09:21Mode"logbot :+i" by logbot
03:09:21CtcpVersion from freenode-connect!frigg@freenode/utility-bot/frigg
03:09:27***Server message 501: 'logbot :Unknown MODE flag'
03:09:28 Join logbot [0] (
03:09:28 Join yosafbridge [0] (
03:09:28 Join [7] [0] (~quassel@rockbox/developer/TheSeven)
03:09:28 Join koniu [0] (~koniu@gateway/tor-sasl/koniu)
03:09:28 Join S|h|a|w|n [0] (~shawn156@unaffiliated/shawn156)
03:09:28 Join J_Darnley [0] (
03:09:28 Join fs-bluebot [0] (
03:09:28 Join bluebrother [0] (~dom@rockbox/developer/bluebrother)
03:09:28 Join speachy [0] (~speachy@
03:09:28 Join vvrng [0] (
03:09:28 Join St3ak [0] (
03:09:28 Join Moarc [0] (
03:09:28 Join Tsesarevich [0] (Tsesarevic@fluxbuntu/founder/joejaxx)
03:09:28 Join ps-auxw [0] (
03:09:28 Join Slasherii [0] (
03:09:28 Join rudi_s [0] (
03:09:28 Join xcin [0] (~x@
03:09:28 Join blackyus- [0] (
03:09:28 Join shrizza [0] (
03:09:28 Join trfl [0] (
03:09:28 Join @ChanServ [0] (ChanServ@services.)
03:09:28 Join plum [0] (~plum@unaffiliated/plum)
03:09:28 Join copper [0] (~copper@unaffiliated/copper)
03:09:28 Join rasher [0] (~rasher@rockbox/developer/rasher)
03:09:28 Join gevaerts [0] (~fg@rockbox/developer/gevaerts)
03:09:28 Join user890104 [0] (
03:09:28 Join Galois [0] (
03:09:28 Join igitoor [0] (igitur@unaffiliated/contempt)
03:09:28 Join aevin [0] (~brootvors@unaffiliated/aevin)
03:09:28 Join bleb [0] (~cm@unaffiliated/bleb)
03:09:28 Join WakiMiko [0] (~WakiMiko@unaffiliated/wakimiko)
03:09:28 Join prg318 [0] (~prg@deadcodersociety/prg318)
03:09:28 Join amdj [0] (~aaron@freenode/staff/atheme.amdj)
03:09:28 Join mutax [0] (
03:09:28 Join ParkerR [0] (~ParkerR@unaffiliated/parkerr)
03:09:28 Join benedikt93 [0] (~quassel@unaffiliated/benedikt93)
03:09:28 Join Topy44 [0] (
03:09:28 Join mikroflops [0] (
03:09:28 Join Strife89 [0] (sid399903@gateway/web/
03:09:28 Join kirvesAxe [0] (
03:09:28 Join lonoxmont [0] (
03:09:28 Join Rondom [0] (
03:09:28 Join __builtin [0] (~quassel@rockbox/developer/builtin)
03:09:28 Join TorC [0] (~Tor@fsf/member/TorC)
03:09:28 Join asaba [0] (~asaba@
03:09:28 Join ecs [0] (
03:09:28 Join sh4 [0] (shapeless@unaffiliated/sh4)
03:09:28 Join CommunistWitchDr [0] (
03:09:28 Join ArsenArsen [0] (~Arsen@kshare/developer/ArsenArsen)
03:09:28 Join jerwin [0] (~flippy-fl@unaffiliated/flippy)
03:09:28 Join klock [0] (~freeklock@unaffiliated/klock)
03:09:28 Join skrzyp [0] (
03:09:28 Join XDjackieXD [0] (
03:09:28 Join akaWolf [0] (
03:09:28 Join funman [0] (~fun@rockbox/developer/funman)
03:09:28 Join tchan [0] (~tchan@lunar-linux/developer/tchan)
03:09:28 Join Xeha [0] (~Xeha@unaffiliated/xeha)
03:09:28 Join alexbobp [0] (
03:09:28 Join vup [0] (~~~~@
03:09:28 Join Soap [0] (~Soap@rockbox/staff/soap)
03:09:28 Join dys [0] (
03:09:28 Join mendel_munkis [0] (
03:09:28 Join emacsomancer [0] (
03:09:28 Join _bilgus__ [0] (~bilgus@
03:09:28 Join atsampson [0] (
03:09:28 Join JanC [0] (~janc@lugwv/member/JanC)
03:09:28 Join Ckat [0] (~Ckat@xn--z7x.xn--6frz82g)
03:09:28 Join TheSphinX^ [0] (
03:09:28 Join michaelni [0] (
03:09:28 Join Barlow [0] (~barlow@unaffiliated/barlow)
03:09:28 Join genevino [0] (
03:09:28 Join braewoods [0] (~braewoods@learnprogramming/staff/braewoods)
03:09:28 Join lemon_jesus [0] (
03:09:28 Join advcomp2019__ [0] (~advcomp20@unaffiliated/advcomp2019)
03:09:28 Join wodz [0] (
03:09:28 Join sakax [0] (~r0b0t@unaffiliated/r0b0t)
03:09:28 Join paulk-leonov [0] (
03:09:28 Join APLU [0] (~mulx@2a03:7220:8081:2900::1)
03:09:28 Join t0mato [0] (~t0mato@
03:09:28 Join amiconn [0] (jens@rockbox/developer/amiconn)
03:09:28 Join pixelma [0] (marianne@rockbox/staff/pixelma)
03:09:28 Join toruvinn [0] (
03:09:28 Join fauweh [0] (
03:09:28 Join efqw [0] (uid412670@gateway/web/
03:09:28 Join blbro[m] [0] (blbrostrat@gateway/shell/
03:09:28 Join danielp3344 [0] (danielp334@gateway/shell/
03:09:28 Join bzed [0] (
03:09:28 Join heredoc [0] (heredoc@2a01:7e01::f03c:91ff:fec1:de1d)
03:09:28 Join ender| [0] (~ender1@2a01:260:4094:1:6045:6fff:fedd:cbf3)
03:09:28 Join kadoban [0] (kadobanmat@gateway/shell/
03:09:28 Join SammysHP [0] (
03:09:28 Join kugel_ [0] (~kugel@rockbox/developer/kugel)
03:09:28 Join pookie [0] (
03:09:28 Join Acou_Bass [0] (
03:09:28 Join _3dsv [0] (
03:09:28 Join rogeliodh [0] (
03:09:28 Join beencubed [0] (~beencubed@
03:09:28 Join mixfix41 [0] (~popebook2@unaffiliated/mixfix41)
03:09:28 Join galaxy_knuckles [0] (~gknux@unaffiliated/galaxy-knuckles/x-1756549)
03:16:07_bilgus__speachy logf from the rocker
03:32:48 Quit vvrng (Remote host closed the connection)
04:23:01 Join ufdm [0] (
04:23:06 Join prof_wolfff [0] (
04:28:09 Quit S|h|a|w|n (Quit: Leaving)
04:30:05 Quit prof_wolfff (Ping timeout: 260 seconds)
05:07:48 Join johnb3 [0] (
05:08:55***Saving seen data "./dancer.seen"
05:18:35 Join lebellium [0] (
05:25:30lebellium_bilgus__: indeed I think you broke the icons 8x8 in May 2020. The 1st offsets were 42 4d 7a like for the other sizes and now the 1st offsets are 42 4d 38. I assume we lost transparency somewhere here
05:25:57lebellium6 months without noticing it :(
05:26:49lebelliumA few years earlier the Clip Zip was the most popular target, it would have been noticed immediately
05:36:22lebelliumPlus some icons like the 'Rockbox' icon don't match with the other sizes
05:37:40lebelliumFrom my point of view the icons should be the same across all sizes to get the same name "tango"
05:44:58 Quit SammysHP (Quit: *wuff*)
05:45:19 Join SammysHP [0] (
05:49:39 Quit SammysHP (Client Quit)
05:50:08 Join SammysHP [0] (
07:08:56***Saving seen data "./dancer.seen"
08:08:33 Join MrZeus_ [0] (~MrZeus@2a02:c7f:70d0:6a00:20d5:a6ab:6b85:44)
08:28:04 Quit johnb3 (Ping timeout: 240 seconds)
08:38:44fs-bluebotBuild Server message: New build round started. Revision d4a80ce, 293 builds, 9 clients.
08:42:25speachy_bilgus__: thank you. unfortunately it's perfectly normal-looking.
08:51:43fs-bluebotBuild Server message: Build round completed after 778 seconds.
08:51:46fs-bluebotBuild Server message: Revision d4a80ce result: All green
09:08:57***Saving seen data "./dancer.seen"
09:12:45 Join prof_wolfff [0] (
10:02:21 Quit prof_wolfff (Ping timeout: 260 seconds)
11:08:14 Join prof_wolfff [0] (
11:09:00***Saving seen data "./dancer.seen"
11:09:18 Join johnb3 [0] (
11:16:28 Quit Strife89 (Read error: Connection reset by peer)
11:16:45 Join Strife89 [0] (sid399903@gateway/web/
11:18:25 Quit prof_wolfff (Ping timeout: 260 seconds)
11:25:02 Quit Moarc (Read error: Connection reset by peer)
11:26:09 Join Moarc [0] (
11:34:18_bilgus__lebellium, I've been using the clipzip daily since then
11:37:03 Quit Moarc (Read error: Connection reset by peer)
11:38:25 Join Moarc [0] (
11:38:30_bilgus__the user that submitted them did rightfully argue that the tiny icons were very blurry
11:40:18_bilgus__but honestly IDC I spend maybe 30 mins a week actually looking at the screen
11:46:46lebelliumwell, if you use a theme with dark background, you would see the issue everytime you turn on the device so it's quite annoying for those
11:47:02lebelliumI'm using my own theme with a white background so the transparency issue isn't seen
11:48:10lebelliumcan you ask the user to add the necessary alpha channel?
11:50:29_bilgus__oh I'm sure he is nowhere to be found but ill put it on the list..
11:52:59lebelliumotherwise revert back to the older icons. I think the transparency is more important than debatable blurriness
12:00:03_bilgus__sounds good lets revert maybe that'll bring em backhuh so only one target even defines = LCD_FBSTRIDE(w, h)
12:00:28_bilgus__sorry second half different train of thought
12:01:17_bilgus__inly the mrobe 500 even defines LCD_STRIDEFORMAT
12:01:48_bilgus__I'm thinking this might the be the final wonkyness in 2 bit horizontal
12:11:48 Join TheLemonMan [0] (~lemonboy@irssi/staff/TheLemonMan)
12:15:02 Quit emacsomancer (Ping timeout: 265 seconds)
12:22:25fs-bluebotBuild Server message: New build round started. Revision 3fbca49, 293 builds, 8 clients.
12:23:12_bilgus__ok lebellium Its been reverted If you can test it after its done and make sure everything came out right
12:23:44lebelliummeanwhile I'm trying to add transparency myself with Photoshop Elements but it's a fail so far :D
12:23:47_bilgus__If he comes back I'll ask him to update otherwise oh well
12:24:30_bilgus__lebellium, I tried and failed too unfortunately its been dithered and it doesn't look right
12:25:06_bilgus__probably why they look less blurry than the old ones
12:29:42_bilgus__hmm so I failed to see that the LCD STRIDE vs PIXEL STRIDEare not related
12:30:06ArsenArsenhow does rockbox pick what partition to mount?
12:30:23_bilgus__that means I have 30 lines or so of conditionals that just will never ever get hit
12:30:26ArsenArsenbecause right now it's saying No partition found when there is a partition with a .rockbox directory on a mkfs.fat -F 32'd fat32 fs
12:30:56_bilgus__ArsenArsen, it just grabs the first valid one it finds IIRC
12:31:05ArsenArsen_bilgus__: define valid
12:31:24ArsenArsen"/dev/sdd1 * 16510 33554429 33537920 128G c W95 FAT32 (LBA)" -> definitely there
12:31:47speachyArsenArsen: what target, what version?
12:32:00_bilgus__but just because you think you formatted properly it means nothing unless it is perfect doing a reformat over exfat sometimes leaves dirty bits laying around
12:32:07ArsenArsenipod6g, dfae5d881d (I believe)
12:32:16ArsenArsenit's a fairly recent build
12:32:23speachywith an iflash card...
12:32:35ArsenArsenyes, but this worked before
12:33:02ArsenArsenonly started breaking down after I started trying to extend the rockbox partition to full size
12:33:05speachya while back (ie several months) the filesystem mount code was made more robust in the face of wonky partitioning/formatting
12:33:20_bilgus__as in this card worked last time you booted and today it doesn't?
12:34:02ArsenArsenno, it worked fine, now I removed the partition after sdd1, extended sdd1 and reformatted it with mkfs.fat, and unzipped my (so that there is a .rockbox directory), and now it doesn't find a partition
12:34:09_bilgus__ah no linux is smart enoght to figure that out we are far less robust
12:34:34ArsenArsenlinux probably does have more robust routines yah but I haven't read eithers
12:35:32_bilgus__I can't tell you exactly where but I know its an issue
12:35:54fs-bluebotBuild Server message: Build round completed after 809 seconds.
12:35:56fs-bluebotBuild Server message: Revision 3fbca49 result: All green
12:36:21ArsenArsenI mean, it could be some code breaking due to unusual sector sizes or something
12:36:39ArsenArsenwait, does it check the boot flag?
12:37:13_bilgus__doubtful but its been a while since i've been in the basement
12:37:30 Part johnb3
12:38:01 Join johnb3 [0] (
12:38:32ArsenArsenwhat does the panic screens how?
12:38:45ArsenArsenmount: 0 looks interesting
12:39:54ArsenArsenah, it's from panicf
12:40:28ArsenArsenit didn't mount any partitions afaict, so it just panics
12:41:09_bilgus__now thats odd really should fail gracefully
12:42:05ArsenArsenit happens after unplugging the usb cord in bootloader usb mode
12:42:09ArsenArsenI think that's what should happen
12:43:30_bilgus__oh hmm I get that in the rocker too it must be new
12:43:47 Join amiconn_ [0] (jens@rockbox/developer/amiconn)
12:43:47 Nick amiconn is now known as Guest22936 (jens@rockbox/developer/amiconn)
12:43:47 Quit Guest22936 (Killed ( (Nickname regained by services)))
12:43:47 Nick amiconn_ is now known as amiconn (jens@rockbox/developer/amiconn)
12:44:53 Join pixelma_ [0] (marianne@rockbox/staff/pixelma)
12:44:53 Quit pixelma (Killed ( (Nickname regained by services)))
12:44:53 Nick pixelma_ is now known as pixelma (marianne@rockbox/staff/pixelma)
12:52:53ArsenArsenCONFIG_DEFAULT_PARTNUM sounds interesting
12:54:16ArsenArsenlooks like it's zero
12:54:30 Join emacsomancer [0] (
12:55:46 Join edhelas [0] (
12:56:59 Quit johnb3 (Quit: Nettalk6 -
12:57:10ArsenArsenwhat does DEBUGF do?
12:57:17ArsenArsenI'm in the basement now lol
13:03:12ArsenArsentime to break out the ol hexdump
13:03:59ArsenArsenaccidential caps lock?
13:04:05_bilgus__sorry caps lock
13:04:09ArsenArsenI hope I didn't cause such frustration
13:04:15_bilgus__was typing defines
13:04:28ArsenArsenyeah, I assume the build I have (from rbutil) doesn't have them
13:04:51ArsenArsenoh :^) it would appear I overwrote the MBR with a FAT partition
13:05:14ArsenArsenoh, nope, I just can't read
13:08:15ArsenArsenthe MBR entry for this part looks fine
13:09:03***Saving seen data "./dancer.seen"
13:09:27ArsenArsenI think it's down to fat_mount failing
13:12:01_bilgus__copy off thew old mbr do a good wipe and try again, I don't remember what helps in linux but I've been pushing windows people towards LowLevelFormatter and rufus? or someother 3rd party formatter
13:12:21ArsenArsenthe old MBR is as good as useless
13:12:39_bilgus__keep that mbr and compare to the new one, Ive yet to get a copy of a bad mbr
13:12:45ArsenArsenI'm suspecting a missing BPB because I don't understand what's meant to create it
13:13:05ArsenArsenor what it does, actually, the BPB is genuinely a mistery to me
13:15:09_bilgus__it carries all the setup for the disk I think
13:16:10_bilgus__here ya go
13:28:29 Join _bilgus_ [0] (~bilgus@
13:28:53_bilgus_and a very comprehensive reference
13:29:20 Quit _bilgus__ (Ping timeout: 256 seconds)
13:32:01_bilgus_I've spent more time fixing OSD than I have doing the whole rest of the FrameBuffer rewrite
13:53:13ArsenArsenI'd love to see where this fails
13:53:19ArsenArsenI might just have to build a debug bootloader
13:58:21 Join cockroach [0] (~blattodea@pdpc/supporter/active/cockroach)
14:29:31ArsenArsenso how do I flash on a dev bootloader?
14:32:34_bilgus_same way as a regular one? the wiki should have steps
14:37:01ArsenArsenhm, I failed to build it with debug
14:38:06ArsenArseni wonder if the simulator could be helpful here
14:43:58ArsenArsenit's unfortunate that disk_mount discards the error from fat_mount
14:44:24 Join _bilgus__ [0] (~bilgus@
14:45:28 Quit _bilgus_ (Ping timeout: 260 seconds)
14:45:33ArsenArsenlooks like the biggest requirement is that it's fat32
14:53:19_bilgus__indeed we only support F32
14:53:47ArsenArsenwait, what's bootloader/fat32format.c
15:01:23lebellium_bilgus__: the best I could achieve
15:01:36lebelliumpixel by pixel in gimp
15:02:41lebelliumBasically I added transparency based on the previous user work. I made some slight adjustments to some icons and reused the settings icon from the older iconset
15:03:30lebelliumregarding the rockbox icon, I left his new icon as the G key is almost impossible to reproduce with 8x8
15:04:35lebelliumI put both the old icons set and this new icons set on my 2 Clip Zip side by side and I'm not sure which one looks best :D
15:05:37 Quit _bilgus__ (Ping timeout: 260 seconds)
15:09:07***Saving seen data "./dancer.seen"
15:09:09lebelliumI like smaller size of the old icons, so that they do not touch each other but indeed there was blurriness. With this new set the blurriness is gone but the icons are bigger and touch each other
15:16:04 Join petur [0] (~petur@rockbox/developer/petur)
15:19:27 Join _bilgus [0] (~bilgus@
15:19:57_bilgussorry bad internet today apparently, its not wired into the bootloaders
15:21:41_bilgusugh @ ArsenArsen
15:22:26_bilguslebellium, if we are putting them back I'll re-add him to the credits when I do the commit
15:26:44_bilgusok this is about as much 'change this break that change that break everything' I can can handle for today
15:28:56lebelliumany other Clip Zip user around here wanting to compare the old and new icons?
15:32:55 Join prof_wolfff [0] (
15:33:02 Join kugel [0] (
15:33:02 Quit kugel (Changing host)
15:33:02 Join kugel [0] (~kugel@rockbox/developer/kugel)
15:35:05 Quit kugel_ (Ping timeout: 260 seconds)
15:38:44ArsenArsen gparted appears to be killing itself consisetntly
15:39:17 Join pamaury [0] (~pamaury@rockbox/developer/pamaury)
15:58:57ArsenArsencan I force the bootloader into usb mode?
16:00:21 Join _bilgus_ [0] (~bilgus@
16:01:37 Quit _bilgus (Ping timeout: 260 seconds)
16:08:06ArsenArsenI can't find a way to resize a FAT filesystem
16:08:19ArsenArsenI even tried windows but it just grays out the expand in the disk manager
16:09:02efqwHow well does rb handle very large databases?
16:09:09 Quit _bilgus_ (Ping timeout: 260 seconds)
16:09:47 Join _bilgus [0] (~bilgus@
16:09:53efqwSuppose I have two 128GB MicroSD in an xDuoo X3 filled with music, will this become a problem?
16:21:04ArsenArsen_bilgus: got it booting again
16:21:11ArsenArsenby restoring it
16:21:18ArsenArsennow I can't resize the partition again
16:22:40 Quit mixfix41 (Remote host closed the connection)
16:38:01braewoodsspeachy: can you offer any insight into what firmware/ does? it seems involved when ROM images are built and it does special stuff for the H100 series but i'm not following it too well atm.
16:38:15braewoodsall i know is it's a linker script.
16:58:42 Join ungali [0] (~ungali@unaffiliated/ungali)
17:09:09***Saving seen data "./dancer.seen"
17:24:59 Quit cockroach (Quit: leaving)
17:31:24ArsenArsensurely writing code to rewrite a fat isn't hard, right?
17:33:48edhelasefqw it takes already a few minutes to index the 60gb of vorbis that I have on my iPod, I can't imagine with 256Gb…
17:40:16speachybraewoods: booting out of ROM requires putting things in different places.
17:40:28speachydifferent things in different places, I should say
17:41:50braewoodsspeachy: ok. i'd like some help there later but for now it doesn't matter.
17:42:08braewoodsi'm reviewing how iriver_flash works and comparing the docs on how the flash chips work from source side
17:42:12braewoodsappears to be identical
17:42:23braewoodsonly real difference is probably timing
17:46:15braewoodsanyway i can't find any issues with the waiting code...
17:46:16speachyArsenArsen: so, what's unknown at the moment is (1) if it's the bootloader or rockbox complaining about the partition info, (2) if it's rockbox, is it the partitioning or the filesystem
17:46:24braewoodsit seems longer than the actual datasheet
17:46:38speachyArsenArsen: if you rename .rockbox to something else, is there a different error?
17:47:42ArsenArsenspeachy: no, it boots, I just need to extend the fs and part
17:47:45ArsenArsenand gparted segfaults
17:49:22 Quit ungali (Quit: ungali)
17:51:18speachythere's always parted on the cmdline instead.
17:52:19speachybut if it works pre-resize, but fails post-resize, it follows that the resize is breaking something. But my original question still stands, is it the bootloader or the rockbox binary that's generatnig that error on the possibly-b0rked drive?
17:58:10ArsenArsenspeachy: no, I simply don't have a tool that can resize a FAT filesystem
17:58:22ArsenArsenI was screwing with the MBR when it failed
17:58:44speachyah, okay.
17:59:22speachythere's always gparted live; put it on a CD or usb stick, boot off it, and do it from there
17:59:38ArsenArsenI'm not sure that would change anything
17:59:44ArsenArsenafter all, this filesystem isn't mounted or anything
18:00:53speachyI mean, that will give you a working resize tool
18:07:24 Quit pamaury (Ping timeout: 240 seconds)
18:07:44ArsenArsenhonestly, I think I'm just facing a parted bug here
18:09:38ArsenArsenreportedly the sector size (in gparted) is 4096
18:09:44ArsenArsenparted seems to not like anything but 512
18:09:49ArsenArsenbut rockbox believes its 512
18:10:50ArsenArsenI should probably spend some time reopening this to take out the card and do it by hand
18:21:12 Quit TheLemonMan (Quit: "It's now safe to turn off your computer.")
18:22:45ArsenArsenspeachy: got it to work
18:22:47ArsenArsengod bless mtools
18:23:05 Quit lebellium (Quit: Leaving)
18:23:38ArsenArsenI just used minfo or whatever to get the right mformat command and changed the total sector count to size in bytes (as reported by lsblk -b) / 4096, and it worked like a charm after I unzipped my rb backup
18:24:43ArsenArsenbloody fat, I swear
18:24:52ArsenArsenmight have to send in a patch for an ext2 driver :)
18:31:39 Join Airwave [0] (
18:32:48AirwaveI run Rockbox development build on Agptek Rocker. After updating a few weeks ago the volume suddenly only goes down to -127dB, which is way too loud. It used to go all the way down to -400dB.
18:47:20speachy-400dB is a volume level that is vastly exceeded in deep space.
18:47:42ArsenArsen^ sounds kinda weird
18:47:57ArsenArsenisn't that below the hearing floor?
18:49:51speachyeven -127dB is below our hearing floor, requiring an anechoic chamber
18:50:34 Quit petur (Remote host closed the connection)
18:51:35AirwaveThe volume levels on my player now go from -127 dB to 0 dB
18:51:44speachybut back to the actual complaint −− I did re-scale the Rocker's volume, because it presented a value from -1020 -> 0, with a step size of 5. I got rid of the scaling.
18:52:18AirwaveI used to have it at around -225 dB before the update
18:52:36AirwaveNow -127 dB is the quietest and it's ear-bustingly loud
18:53:15speachygimme a few, brb
18:54:29AirwaveOh and if it matters it's actually a Benjie T6
18:54:58gevaertsThat sounds wrong... If -127dB is too loud, 0dB would mean the end of the world... I assume that's not that case :)
18:55:43AirwaveI could try 0 dB, but I'm afraid it would blow out my headphones
18:56:36AirwaveWhen I'm in the car I set it to 0 dB and use the car's volume adjuster instead.
18:57:19ArsenArsenI'm not 100% confident in this thing I did with mformat to be honest
18:58:37ArsenArsen255 heads, 60915712 sectors, 63 SPT, 3791 cylinders
18:58:50ArsenArsensounds very wrong, but I guess it's just emulation or something
18:59:15gevaertsIf 60915712 seems plausible. That's all that matters
18:59:27speachywith LBA, C/H/S is irrelevant
18:59:45gevaertsThat's 30GB if I'm not mistaken
18:59:53ArsenArsennot with 4096B sectors :)
19:00:07ArsenArsenspeachy: and yeah, I know, but FAT still contains that info
19:00:27gevaertsAh, then indeed not :) 240GB?
19:00:37ArsenArsenye that's about right
19:00:46ArsenArsena weird sector size, though
19:00:50ArsenArsenI do wonder what's up with that
19:00:54braewoodsspeachy: i noticed the coldfire code appears to search for ROM/RAM rockbox images regardless of platform at hard-coded locations... the H300 has more ROM space than these other ones do... should I just ignore the extra space? whatever I choose we'll be stuck with but 2MB has been enough up till now for the H100 series, which is the only thing that uses this
19:01:40speachythere's no need to ignore the extra space on the H300, indeed we'll need it due to larger resource requirements (ie color bitmaps)
19:02:10braewoodsas it stands the H100 effectively splits the ROM space in 2
19:02:18braewoodsnot exactly but close enough
19:02:40braewoodsH300 has twice the ROM space available so I was considering putting the ROM start location higher to allow for more expansion later
19:02:53braewoodsbut pondering what i'd need to do for the bootloader to support it
19:03:05braewoodsi'll poke around a bit more
19:03:58speachyAirwave: I think I missed part of the de-scaling. try this build:
19:04:20Airwavespeachy: Okay gimme a second
19:05:01speachyignore the "number" it shows, the important bit is the actual relative volume.
19:06:59braewoodsspeachy: see firmware/target/coldfire/crt0.S
19:07:05braewoodsi suspect it was intended to pull from
19:07:38speachybraewoods: no, generally speaking you can't move the start location, that's hardwired into the SoC.
19:07:54braewoodsbraewoods: i don't think you understand what i'm getting at
19:07:54speachyI mean, the CPU is going to start looking at a specific location regardless.
19:08:13braewoodsspeachy: FLASH_RAMIMAGE_ENTRY / FLASH_ROMIMAGE_ENTRY are where we store rockbox in ROM on the H100 series
19:08:30braewoodswhich are 0x1000 and 0x1000000 respectfuly
19:08:34braewoodsthese are hard-coded in that file
19:08:51braewoodsfollowed by comments around it saying ROM / RAM rockbox image
19:09:12***Saving seen data "./dancer.seen"
19:09:19braewoodse.g., line 153
19:09:22braewoodsthen 160
19:09:32braewoodsthey're hard-coded. i think they were intended to use macros
19:09:50braewoodsto leverage the extra H300 space they need to be configurable
19:10:47speachyugh, there's leftover sh cruft in the
19:11:32braewoodsso, do you think i can make the base address use a macro or so?
19:11:51braewoodsin case someone decides to try expanding this feature to non-iriver targets
19:12:00braewoodselse I can use #if defined for the iriver h300
19:12:07Airwavespeachy: Volume still goes from -127 dB to 0 dB, but I now noticed something new. If I go up, -122 is all of a sudden the lowest and it goes higher all the way up till -98, then at -97 it's super quiet again, then it resets again at -71 dB.
19:12:19braewoodsprobably best if i do that I think
19:12:34AirwaveNot sure if this was the case wit the build I was using before, I never tried increasing the volume above -127 dB then
19:13:49braewoodseither way i think i'll be committing the first part of my work
19:13:52AirwaveNext reset is -46 dB
19:13:54braewoodslet's see
19:14:20AirwaveThen -23
19:15:18AirwaveThat's the last reset. Well at least now I have a workaround, I can get it low again by a little higher
19:15:49speachyAirwave: in other words, the hardware lies about the volume scale.
19:16:39AirwaveYeah guess so
19:17:41speachy(I wonder if we're better off just forcing the HW mixer to max and using SW mixer volume operation..)
19:17:54AirwaveOh and there's also this new bug recently where sometimes when I pause it, unpausing doesn't work (it just stays stuck at the same play position) and I have to reboot to get it playing again
19:19:17speachyyeah, that's just been reported yesterday
19:19:21speachythe reason is unclear
19:19:33braewoodsone small change that should be simple to integrate
19:20:50AirwaveSeems to happen if I leave it paused until the display turns off before I unpause it again. If I play/pause while the display is still on it works fine.
19:23:59speachythe built I sent you has extra logging enabled, when it gets stuck, can you go into the system/debug/logf dispaly
19:24:06speachyonly interested in the last line
19:25:11Airwavespeachy: Oh and I just tested the volume thing with the latest development build vs. the build you sent. With latest dev it "resets" at -115 and then doesn't reset again. With your build it resets at -122, -97, -71, -46, and -23.
19:26:11Airwavespeachy: Sure, one moment
19:27:02Airwavespeachy: Do you want me to trigger the bug before I go in there, or just go there straight away?
19:27:12speachytrigger it first
19:29:06speachyif I set the volume on the cmdline properly, it works. something in our ALSA code is ... very wrong.
19:29:31Airwavespeachy: THere's no "logf display" here. Only thing with "log" in the name is "Metadata log"
19:30:18speachyAirwave: well, lovely. guess I didn't turn it on in that build tree
19:31:48AirwaveIf you build it again I'll try again
19:37:26speachyuploaded, same filename.
19:38:29fs-bluebotBuild Server message: New build round started. Revision 1419070, 293 builds, 9 clients.
19:41:57Airwavespeachy: "Show Log File"?
19:42:22speachyprobably the last line is "pcm state X"
19:42:42AirwaveWhat denotes a newline?
19:45:01speachyor just do the "dump log" and upload it somewhere
19:46:46AirwaveThe last several lines
19:48:55speachyok, that's unfortunately all nominal.
19:51:23AirwaveLet me know if I can help debug anymore
19:54:27fs-bluebotBuild Server message: Build round completed after 958 seconds.
19:54:28fs-bluebotBuild Server message: Revision 1419070 result: All green
20:04:19speachyAirwave: I'll have anotehr build for you in a few.
20:06:09Airwavespeachy: Nice
20:10:16speachynew file, same location.
20:11:36speachyvolume should be sane.
20:12:08 Quit bluebrother (Disconnected by services)
20:12:13 Join bluebrother^ [0] (~dom@rockbox/developer/bluebrother)
20:12:24speachy(and volume works differently from all of the other settings. bah)
20:13:09 Join fs-bluebot_ [0] (
20:15:19 Quit fs-bluebot (Ping timeout: 246 seconds)
20:28:02fs-bluebot_Build Server message: Build round completed after 907 seconds.
20:28:03fs-bluebot_Build Server message: Revision 39a0d8b result: All green
20:30:02speachyAirwave: with regards to the playback not resuming, in the settings, disable the backlight fading, and see if that mitigates things.
20:30:09Airwavespeachy: Sorry, I was away for a bit. Testing now.
20:33:05Airwavespeachy: Well first thing I noticed is that in settings, volume goes from -127 dB to 0 dB in .5 increments, but in the top bar it shows 10x that value.
20:33:43AirwaveSo if I e.g. set it to -124.5 in settings, the top bar says -1245
20:35:20AirwaveIt's still weird. Very loud at "minimum", then resets at at -115.0 dB.
20:35:36Airwave-115.0 dB is the only reset though
20:36:28speachyso there's still some overflow going on in the hw side of things. -115.0 is the true minoimum point?
20:37:29Airwavespeachy: So turn off both "Backlight Fade In" and "Backlight Fade Out" to test the play/pause issue?
20:37:52speachyit's the fade in that I'm interested in
20:38:36speachythe build that just finished sets the minimum to -102.0
20:39:32Airwavespeachy: Okay with Backlight Fade In set to off I can't reproduce the play/pause issue
20:41:27speachyone more to try, it's building now.
20:41:37speachyplease retore the fade settings to default
20:42:05speachyI think I know what's going on now, but if correct, it points at a really stupid kernel bug.
20:42:09AirwaveWill do, I'll just remove that setting from config.cfg
20:42:40speachynew file, same place
20:43:00AirwaveOkay I'll test volume and play/pause, right?
20:43:26speachyvolume's unchanged, though it's capped at -115 now
20:43:34AirwaveOkay one minute
20:45:38speachyuploading another one.
20:46:01speachythe other one would have resulted in no audio output at all
20:46:14AirwaveI was just about to type that!
20:46:57AirwaveOkay installing the latest
20:48:57Airwavespeachy: Volume is good with this one at least, no resets at all
20:49:31AirwavePlay/pause is good as well!
20:50:19AirwaveNice work, both issues fixed!
20:52:01AirwaveWell only kinda odd thing now is that the top bar shows 10x the dB value that the sound setting does, but that's no biggie
20:53:18speachywhat's actualy happening is that the command to mute/unmute the codec is colliding with the backlight control.
20:53:30 Join S|h|a|w|n [0] (~shawn156@unaffiliated/shawn156)
20:53:51speachymy guess is they're both attached to i2c, and there's improper locking in the kernel drivers.
20:55:08AirwaveIs it the hardware that's to blame?
20:55:16speachywith the volume dialed up, is there a pop at initial startup?
20:55:34speachynot the hardware, the OEM Linux kernel.
20:56:30Airwavespeachy: I tried booting it now with volume set to max. I didn't hear a pop.
20:59:08braewoodsspeachy: next minor patch
21:00:20speachyAirwave: I'll have one more for you, and if that's good, I'll commit this. The wonky scale-on-top bug is purely a display problem, and should be easy to fix once I find where the offending bit of code is.
21:00:22braewoodsthere. i finished my patching to coldfire bootloader common code. it took awhile to figure out what it was doing.
21:00:40Airwavespeachy: Okay, let me know when it's ready
21:00:48speachyAirwave: is ready.
21:01:40Airwavespeachy: Installing now. What do you want me to test?
21:02:12speachymake sure playback works, pause, etc.
21:02:15 Quit MrZeus_ (Ping timeout: 272 seconds)
21:03:50Airwavespeachy: Yep everything working fine
21:04:14braewoodsspeachy: please review my patch sets when you get a chance. some of it was book-keeping to make it easier to work on H300.
21:04:43braewoods3 of them are pending
21:05:40speachylooks ok, but does that cover all coldfires?
21:06:04Airwavespeachy: Thanks for the great work on this. Loving Rockbox on this player!
21:06:07braewoodsspeachy: the last one sure does. why do you think i had to add extra macros?
21:06:23braewoodsspeachy: i modified all the coldfire targets to have the necessary macros
21:06:28braewoodsonly the iriver ones already had them
21:07:15fs-bluebot_Build Server message: New build round started. Revision bde72ef, 293 builds, 8 clients.
21:08:16braewoodsthough i did notice some of the bootloaders were failing to build due to missing LCD functions
21:08:20braewoodsnot sure what's going on there
21:08:25braewoodsbut not due to my changes
21:08:31speachyum.. so we're going to get some red, then.
21:08:41braewoodsyea.. i assume so.
21:08:55braewoodsbut i didn't change anything LCD related
21:09:07braewoodsiaudiom3 was one failure
21:09:14***Saving seen data "./dancer.seen"
21:09:17speachylet's see how it shakes out, and fix properly.
21:09:40braewoodsi guess it was due to stale build
21:10:45braewoodsreally weird
21:10:57braewoodsi guess i should make a fresh build dir when i reconfigure
21:11:05braewoodsit's not happening now
21:11:26AirwaveOh by the way, is there any way to set USB to charge-only? When I use it in the car the power goes off and on when I turn the ignition, so it goes into USB disk mode unless I hold a button. Kinda annoying to have to remember to hold a button every time.
21:12:30AirwaveOr to get it to detect that it's plugged into a USB with power only/no data
21:12:53braewoodsanyway, it's time to start testing out the new iriver_flash
21:14:49AirwaveRight now this little patch is my best solution to the issue:
21:20:34fs-bluebot_Build Server message: Build round completed after 798 seconds.
21:20:35fs-bluebot_Build Server message: Revision bde72ef result: All green
21:20:36fs-bluebot_Build Server message: New build round started. Revision 28b9c8b, 293 builds, 8 clients.
21:36:51fs-bluebot_Build Server message: Build round completed after 975 seconds.
21:36:52fs-bluebot_Build Server message: Revision 28b9c8b result: All green
21:38:56braewoodsspeachy: when you get a moment,
21:39:13braewoodsi expect that to be the last of it before I get around to testing iriver_flash and the bootloader
21:39:32braewoodseverything else seems to be in place
21:39:47braewoodsnow i need to generate a test bootloader
22:24:03braewoodswow. good thing i reviewed the source code prior to flashing.
22:25:05braewoodsgiven the existing code it would fail to erase the bootloader on the H300
22:27:38 Quit Moarc (Read error: Connection reset by peer)
22:29:32 Join Moarc [0] (
23:04:56braewoodsspeachy: well.. iriver_flash crashed rockbox... but it didn't make it to the flash tampering phase so nothing was harmed...
23:05:02braewoodsspeachy: any idea what this means?
23:05:17braewoodsspeachy: I27:Trap7 at 32f800ee
23:09:15***Saving seen data "./dancer.seen"
23:17:19_bilgussure does look like it trapped error 7 at SP or hopefully PC mem address 32f800ee
23:17:54_bilgushave a look at the mapfile generated and search for 32f8 ish
23:37:16braewoods_bilgus: hm... i don't think i need to do that
23:37:31braewoodsiriver_flash disables the ... memory protection
23:37:50braewoods0x32f800ee is beyond the DRAM origin point
23:37:54braewoodsso it's not the flash region
23:38:15braewoodssubtracting DRAMINIT and i get...
23:38:39braewoodswhich is more RAM than this thing ha
23:38:58braewoodsso my best theory is its an error from attempting to go outside the confines of RAM
23:39:32 Join Stanley00 [0] (~stanley00@unaffiliated/stanley00)
23:39:34braewoodsor... hm
23:39:36braewoodsnot quite
23:39:49braewoods_bilgus: where would the map file be?
23:41:11braewoods_bilgus: also, can you approve these?
23:44:57braewoodsi'm going to see if memcpy is working or not
23:45:05braewoodsthat's right before the first erase attempt
23:55:47_bilgusbraewoods, you need to alter the make file to generate a mapfile if it doesn't do it
23:56:35braewoods_bilgus: i've never done that before.
23:57:02_bilguseasiest to see what is going on is probably pictureflow in plugins

Previous day | Next day