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 2020-09-30

00:04:22 Quit akaWolf (Ping timeout: 246 seconds)
00:29:40 Join akaWolf [0] (~akaWolf@akawolf.org)
00:31:01***Saving seen data "./dancer.seen"
01:00
01:00:55 Quit massiveH (Quit: Leaving)
01:27:44 Join ZincAlloy [0] (~Adium@ip5f5acf9f.dynamic.kabel-deutschland.de)
01:28:56 Quit ac_laptop (Ping timeout: 256 seconds)
01:31:47 Quit ZincAlloy (Ping timeout: 240 seconds)
01:32:45braewoodshm.
01:38:02braewoodsspeachy: what does it take to register for the wiki? I was told registration is disabled.
01:38:24braewoodsI was trying to correct some outdated links...
01:40:46 Join ZincAlloy [0] (~Adium@2a02:8108:943f:d824:6de1:ea53:aa02:1121)
01:45:51 Quit ZincAlloy (Ping timeout: 272 seconds)
02:00
02:19:46 Quit efqw (Read error: Connection reset by peer)
02:20:00 Join efqw [0] (uid412670@gateway/web/irccloud.com/x-pzlbsqthyhjvsueh)
02:29:15 Quit amiconn (Quit: No Ping reply in 64 seconds.)
02:29:27 Join amiconn [0] (jens@rockbox/developer/amiconn)
02:31:05***Saving seen data "./dancer.seen"
02:46:30 Join petur [0] (~petur@rockbox/developer/petur)
03:00
03:03:32 Join pamaury [0] (~pamaury@rockbox/developer/pamaury)
03:17:10 Quit michaelni (Ping timeout: 272 seconds)
03:18:13 Join michaelni [0] (~michael@213-47-68-29.cable.dynamic.surfer.at)
04:00
04:31:09***Saving seen data "./dancer.seen"
06:00
06:15:47 Quit pamaury (Ping timeout: 240 seconds)
06:16:05 Join Telehubis [0] (594cb4e3@89-76-180-227.dynamic.chello.pl)
06:16:40TelehubisHi, any other than Paypal donations possibilities?
06:16:59 Quit Telehubis (Remote host closed the connection)
06:19:13braewoodsOw. We got hit by another hit and run greeter.
06:19:34braewoodsI mean... 20 seconds?
06:19:37braewoodsthat's not long to wait
06:20:42 Join pamaury [0] (~pamaury@rockbox/developer/pamaury)
06:31:11***Saving seen data "./dancer.seen"
06:42:32speachybraewoods: PM me your desired email and full name, I'll create the account.
06:42:43speachybots are heavily slamming the account creation page to this day.
06:45:37speachyTelehubis, we have nothing else set up. What did you have in mind?
08:00
08:15:11 Join Telehubis [0] (594cb4e3@89-76-180-227.dynamic.chello.pl)
08:15:49TelehubisOk then. Hi :- )
08:17:03Telehubisas for paypal I was using it but they have a fairly high additional fee (which is indirectly added) that is why I though maybe something else...?
08:20:01speachyif there's one thing I've learned, the money people will _alway_ find a way to take their cut
08:21:08Telehubiswell then they are not charity
08:21:47speachyare you in the US?
08:21:51Telehubisnope
08:21:56TelehubisEurope
08:22:20speachysince international borders are involved I think there are no non-sucky options.
08:23:35Telehubisok, paypal it is
08:31:12***No seen item changed, no save performed.
08:36:24 Quit Telehubis (Remote host closed the connection)
10:00
10:25:53genevinook filling a 256gb sd card is a bit harder than i thought
10:31:13***Saving seen data "./dancer.seen"
11:00
11:05:43 Quit Huntereb (Ping timeout: 260 seconds)
11:28:30_bilgus__INDEED
11:30:21_bilgus__speachy, not sure I have any reason to want to be involved in that, not a form factor I'd ever want in my pocket
11:31:10speachythat's what pine64 came back with; my emails are saying that's not what "we" want
11:31:32speachy(where "we" is really "I" so far..)
11:34:23_bilgus__the touch screen Is where it all really goes to shit i'd make a way to shove a player in my pocket but I break touch screens I count lifetimes in weeks
11:34:53efqwwhat I really want is pretty much an iPod Classic clone, with USB-C and a decent IPS LCD
11:35:36 Join johnb2 [0] (~johnb2@p5b3af332.dip0.t-ipconnect.de)
11:35:46efqwimo some of those linux based sony players are quite nice but the proprietary connector is a total deal breaker for me
11:36:26efqwI have no interest in anything that doesn't use MicroUSB or USB-C.
11:37:15 Join MrZeus_ [0] (~MrZeus@2a02:c7f:70d0:6a00:9d0d:8a12:6a9b:2ed)
11:37:55_bilgus__I feel like USB-c is pretty inevitable at this point I mean unless you like fucking your customers by making them buy overpriced custom dongles
11:39:02_bilgus__well I guess thats possible even with USB-C if they really want headphones
11:39:28 Quit akaWolf (Ping timeout: 246 seconds)
11:42:30efqwspeachy: please also clone the m3k kernel source from github just in case fiio decides to remove it in the future (albeit highly unlikely)
11:43:19 Join lonoxmont [0] (~lonoxmont@024-180-058-254.res.spectrum.com)
12:00
12:07:04 Join ZincAlloy [0] (~Adium@ip5f5acf9f.dynamic.kabel-deutschland.de)
12:11:23 Quit ZincAlloy (Ping timeout: 240 seconds)
12:12:37 Quit petur (Quit: Connection reset by beer)
12:19:59 Join ZincAlloy [0] (~Adium@2a02:8108:943f:d824:5049:3abb:4a19:2040)
12:21:34 Join sakax [0] (~r0b0t@unaffiliated/r0b0t)
12:24:31 Quit ZincAlloy (Ping timeout: 244 seconds)
12:30:33 Join akaWolf [0] (~akaWolf@188.243.183.39)
12:31:16***Saving seen data "./dancer.seen"
12:33:40fs-bluebotBuild Server message: New build round started. Revision 1c0648c, 282 builds, 9 clients.
12:35:57 Join ZincAlloy [0] (~Adium@2a02:8108:943f:d824:3065:daf3:1401:3dd7)
12:50:34fs-bluebotBuild Server message: Build round completed after 1015 seconds.
12:50:39fs-bluebotBuild Server message: Revision 1c0648c result: All green
12:51:22 Join kugel [0] (~kugel@ip5b40d8e6.dynamic.kabel-deutschland.de)
12:51:22 Quit kugel (Changing host)
12:51:22 Join kugel [0] (~kugel@rockbox/developer/kugel)
12:52:20 Quit kugel_ (Ping timeout: 256 seconds)
13:00
13:03:47 Quit johnb2 (Ping timeout: 240 seconds)
13:07:27 Quit kugel (Ping timeout: 240 seconds)
13:10:36genevino_bilgus__: unfortunately, a lot of manufacturers still use micro-usb jacks.
13:11:27_bilgus__still an open and standard connector..
13:14:47genevinotrue, but one that breaks every now and then...
13:15:37genevinousb-c is like MUCH more robust.
13:15:43 Quit pamaury (Ping timeout: 246 seconds)
13:18:43lonoxmontalso usb-c isnt 4-dimensional
13:18:51lonoxmont>try to insert cable
13:18:56lonoxmont>doesnt fit, flip over
13:19:02lonoxmont>still doesnt fit, flip over
13:19:04lonoxmont>fits
13:19:09lonoxmontpikachu.jpg
13:19:36genevinohehe
13:22:13 Join tor_ [0] (~tor@185.9.19.107)
13:41:01_bilgus__if robust was the key criteria a well supported usb mini was fine
13:50:12speachywhat's the least robust are the original B connectors, which wear the socket rather than the cable. wtf.
13:59:00lonoxmonti think i like the micro over the mini
13:59:08lonoxmontseems to stay in more tightly
13:59:16lonoxmontat least the ones ive used
14:00
14:01:24braewoodsspeachy: the usb type primarily used in printers?
14:01:49speachyprinters do seem to be the last holdout of B sockets, for reasons that I do not comprehend
14:03:01speachy1500 insertion cycles, and it's the socket that wears out. mini-USB upped that to 5K, and micro-usb over 10K.
14:03:06genevinoso from the library view, i can just long-tap the middle button on my agptek rocker and "insert into current playlist" or even "insert into current playlist (shuffled)" which is exactly what i need to play all songs on my device shuffled. but where can i a) view this "current playlist"? also how the hell do i even clear the current playlist?
14:03:36genevinooh, the latter question obviously was b)
14:03:41speachycontext menu while playing.
14:03:58speachyI forget how to bring that up on the rocker though
14:04:39genevinowell when long-tapping the middle button, i get the database context menu, which is not what you mean, or?
14:05:01 Quit tor_ (Quit: tor_)
14:05:21braewoodsspeachy: i was reading a more in-depth analysis of IDE UDMA modes... apparently the higher speed ones aren't even an option due to low # of pins in most ATA based players
14:05:47braewoodsthey need the full 80 pins to use higher modes or so but then how does CF do UDMA 7 then?
14:05:48speachybraewoods: it's not the pin count; PATA drives only had 40 connectors, after all.
14:06:00lonoxmontspeachy: some other stuff uses the B sockets, not sure if they can carry more current or if its jsut cheap
14:06:02braewoodsAh.
14:06:07speachybut the _cables_ had 80 wires, pairing each wire with a ground.
14:06:15braewoodsOh.
14:06:22lonoxmonti have some HAM radio SDR stuff thats powered off of a b socket
14:06:29braewoodssounds like overkill...
14:06:35braewoodswhy do you need that many separate grounds?
14:06:36speachyso a direct-attached drive should JustWork(tm) in "80-wire" mode.
14:06:42genevinospeachy: so do you have any further pointer how to simply clear the current playlist?
14:06:53speachysignal integrity; minimizing crosstalk between the data lines
14:07:02lonoxmontbraewoods: because parallel
14:07:12genevinospeachy: going to "playlist catalogue" just tells me that /Playlists doesn't exist btw.
14:07:13lonoxmontalso what speachy says
14:07:38lonoxmontevery other wire in the cable is a ground
14:07:43braewoodsI see.
14:07:59lonoxmontdata ground data ground
14:08:18braewoodsand i'm guessing sata took over due to the limitations of PATA
14:08:21lonoxmontyep
14:08:36lonoxmontturns out its easier/cheaper to use serial instead
14:08:38braewoodsUDMA 7 being the last speed mode ever added
14:08:59lonoxmontthen you have less wires to worry about for thigns like timing skew and so forth
14:09:33braewoodsthough we still see parallel interfaces
14:09:42braewoodse.g., old school character lcds
14:10:06lonoxmontyeah
14:10:21lonoxmontbut even then afaik they are usually less wires than the old pata stuff
14:10:53braewoodshow does serial work anyway? does it use shift registers or something to transfer bits?
14:11:02lonoxmontthen again nowadays its usually those flexible plastic ribbon cables
14:11:07lonoxmonti think so yeah
14:11:08braewoodsi assumed due to low pin count it has to pulse the pins it does have
14:11:12lonoxmontjust shift registeres for days
14:12:10speachygenevino: I don't think there is an explicit "clear playlist" since you can replace the runing playlist with something else at any time
14:12:36speachygranted you can manually remove each item one by one
14:13:19lonoxmonteasy/lazy workaround would be replace it with a single song then delete that song
14:13:20lonoxmont:B
14:13:28speachyor ... just stop playing. :)
14:13:32lonoxmontthat too
14:14:54 Quit craftyguy (Quit: WeeChat 2.9)
14:18:12genevinohmmm, it keeps complaining that /mnt/sd_0/Playlists doesn't exist when i try to select the playlist catalogue from the main menu. o.O
14:18:34speachyI take it you have that directory?
14:18:53speachy(ie "Playlists" in the top-level of the SD card?)
14:19:02genevinoi tried creating it but that didn't change something, yes.
14:20:01braewoodsgenevino: did you try regenerating the database under settings?
14:20:14braewoodsmight be related or not
14:20:17genevinobraewoods: not yet, should i?
14:20:31braewoodsyou normally need a database of your music to add to playlists
14:20:42braewoodsthough i guess it does support manual insertions by browsing
14:21:08genevinoah wait a second, inserting means actually "clear, then insert", as apposed to "appending"
14:21:14genevinooookay
14:21:22genevinowell that solves all the problems basically
14:25:00genevinoso yea lonoxmont was right, you could just "insert a single song"
14:25:19genevinobut i still don't know what's up with that /Playlists thing
14:25:48speachythe error is odd; it's suppsoed to create the directory if it doesn't exist.
14:27:03braewoodsspeachy: i get an error about updating databases when i disconnect RB from usb. any idea why that happens?
14:27:10braewoodsseems to happen on every RB thing i've ever owned
14:27:39braewoodsusually before I've installed anything to play
14:27:55speachyI think that'/mnt/sd_0/' shouldn't be there though.
14:28:15braewoodstoo bad a shell can't be opened
14:28:30braewoodswouldn't that be nice though? shell over usb like adb has.
14:28:40speachy"patches welcome" :D
14:28:40braewoodsfor Linux targets
14:28:51braewoodsyou could then directly inspect what's going on.
14:28:52speachywell, we _do_ have adb on the rocker
14:29:10speachyincluding strace
14:29:13braewoodsi see.
14:31:18***Saving seen data "./dancer.seen"
14:31:44genevinowell at least none of the issues i discovered on the rocker are actual dealbreakers that make using the device on a daily basis impossible.
14:31:55genevinoso there's that. ;)
14:33:48braewoodsi've been working with an iriver h120
14:34:13braewoodsit's possible to replace the OF entirely and install RB entirely to ROM save for its extras that must still be loaded from disk
14:34:31braewoodssince the device has 2MB of EEPROM
14:35:12braewoodsbut you need to do some hacks to get it to work since this isn't the default mode of operation for RB
14:35:22braewoodslike flashing a newer bootloader
14:35:31speachyok, the playlist catalog thing is broken on hosted targets and simulator builds.
14:36:05braewoodsspeachy: how many native targets are there?
14:36:07braewoodserr
14:36:09braewoodshosted
14:36:52speachysamxung yp-r0/r1, xduoo x3ii/x20, rocker, ibasso dx50/90, and most of the sonys.
14:37:54braewoodsah... so pretty much all the "modern" ones
14:38:34braewoodson the bright side, at least the system call ABI for Linux is totally exposed so you can write ASM if you have to...
14:38:42braewoodsyou can bypass libc and such.
14:39:13braewoodslibc is mainly there for portability between architectures
14:41:30speachyyeah, but why write our own libc when we don't need to? :)
14:41:43speachythis is a case of GIGO
14:43:52braewoodsso you link to their libc?
14:43:58braewoodswhatever they use since linux can use any
14:44:26braewoodsnot sure how you would though without the libc headers for that system
14:44:40braewoodsit would probably be easier to statically link in your own choice of libc
14:45:19bluebrotherspeachy: re that rocker install the other day: so Rockbox Utility should have a hint for the user to perform an update from the OF? I.e. the rocker needs to get added to BootloaderInstallHelper::postinstallHints()?
14:45:51speachybluebrother: yeah, that's a good idea. I could have sworn I did something like that already though
14:46:02speachythe rocker, x3, x3ii, and x20 all need that same blurb
14:46:25bluebrotherthat existing part for h100 should fit, right?
14:48:05 Join tor_ [0] (~tor@185.9.19.107)
14:55:11speachyokay, g#2771 will fix the playlist (and recording) paths on hosted targets.
14:56:06speachyoh, simulator too
14:56:22braewoodsspeachy: it even simulates the bugs!
14:56:23braewoodslol
14:56:50speachybraewoods: that's actually a pretty important quality
14:56:57fs-bluebotBuild Server message: New build round started. Revision ff408fd, 282 builds, 9 clients.
15:00
15:09:21genevinospeachy: oh nice!
15:11:10 Join kugel [0] (~kugel@ip5b40d8e6.dynamic.kabel-deutschland.de)
15:11:10 Quit kugel (Changing host)
15:11:10 Join kugel [0] (~kugel@rockbox/developer/kugel)
15:11:48fs-bluebotBuild Server message: Build round completed after 892 seconds.
15:11:50fs-bluebotBuild Server message: Revision ff408fd result: All green
15:13:25fs-bluebotBuild Server message: New build round started. Revision 728299e, 282 builds, 9 clients.
15:19:37braewoodsspeachy: apparently there's been bootloader fixes in "SVN" that isn't in the latest build for h100 series
15:19:49braewoodsfrom 10 years ago...
15:19:55braewoodswhy no new bootloaders?
15:19:57braewoodsI'm puzzled.
15:20:22speachy...nobody's bothered to test it out properly and upload it?
15:20:31braewoodsI see.
15:21:08braewoodshow can I build it myself?
15:21:21braewoodsor is that not advised?
15:22:19braewoodseither way... guess it's time to get a wiki account first
15:27:26bluebrotherbraewoods: the problem with the bootloader on h100 is that if the flash goes wrong you need special hardware to revive the player. That only left this work to a few people back then. Unless you're feeling confident enough to (not) brick your player :)
15:27:47braewoodsi see.
15:29:37fs-bluebotBuild Server message: Build round completed after 972 seconds.
15:29:41fs-bluebotBuild Server message: Revision 728299e result: All green
15:32:45lonoxmontbluebrother: need a jtag setup to unbrick the player i take it?
15:33:08lonoxmontor is it something more janky?
15:35:41bluebrotherjtag afaik.
15:36:09lonoxmontthats not as janky as it could be, at least its ostensibly an industry standard
15:36:20lonoxmontbut you are right that most people dont have jtag hardware lying about
15:36:35bluebrotherhttps://www.rockbox.org/wiki/IriverBDM
15:38:05bluebrothernot sure what you'd need to do to fix a failed flash. Probably loading rockbox to RAM, executing it and using iriver_flash. Never did jtag stuff on that player.
15:38:23bluebrotheranother problem with the jtag is that you have to open the player and solder around :)
15:38:29lonoxmonttrue
15:39:02braewoodsbluebrother: so how would one get or build a newer build for the bootloader?
15:39:12lonoxmontvery carefully
15:39:14lonoxmont( ≖‿≖)
15:39:14braewoodsit's probably time someone released an updated one that has more of the fixes
15:39:44lonoxmontwhat fixes could the bootloader even have?
15:39:48speachywe shoudl also probably do anoher ipod bootloader build with the iflash fixes
15:40:00speachyto answer your question indirectly..
15:40:02lonoxmontlike as long as it loads the os im not sure what else there is to fix
15:40:22lonoxmontunless it hangs around and acts like a bios or something and handles low level storage calls and whatnot
15:42:27bluebrotherthere's a v7pre4 bootloader in the wiki. That's the one I'm using (with Rockbox in flash)
15:43:57bluebrotherthough that's also 10+ years old :)
15:44:10bluebrotherand yeah, for h100 the bootloader isn't doing much afaik.
15:45:19braewoodsbluebrother: there's fixes with CF mod related to usb mode
15:45:33braewoodsfixed in SVN but not 7pre4
15:45:40bluebrotherwhat's svn? ;-)
15:45:52braewoodsbasically means "fixed in git but not released"
15:46:12bluebrotherok. I wouldn't mind an updated bootloader. Though someone has to do it.
15:46:21braewoodsindeed.
15:46:34braewoodswhich is why i was trying to find out how to build an updated bootloader only
15:46:44*bluebrother wants to update the bootloader on his clip+ for sd booting
15:46:55bluebrotherthat hasn't been released either, right? As well as e200?
15:47:13braewoodssansa clip+, isn't that stable?
15:47:35braewoodsyep
15:48:40braewoodshttps://www.rockbox.org/wiki/CFModGuide
15:48:45braewoodsNote: On iRiver H1x0 targets, if a CFMod has been performed, returning from USB disk mode is not fully supported from the bootloader. Disk mode works to and from Rockbox (after RB has booted) but to Rockbox from bootloader USB disk mode. In other words, if a user plugs in the USB cable with the player turned off, then boots the player they will get into a working USB disk mode but will be unable
15:48:47braewoodsto get out of this mode without seeing the dreaded "ATA error: -80". A reset is then required.
15:49:13braewoodshttp://www.rockbox.org/tracker/task/9642
15:49:16braewoodsis the linked issue
15:49:30braewoodssays svn bootloader reported to work
15:49:38braewoodsbut that's from 2010 which is far newer than 7pre4
15:49:42braewoodsso
15:49:47bluebrotherthe clip+ is stable, but iirc support to boot from sd has been added some time ago but no new release of the bootloader been done.
15:49:49braewoodsit probably would require an updated bootloader
15:50:06braewoodsindeed.
15:51:02braewoods" Be careful! Do NOT attempt to build your own bootloader from SVN unless you know for sure what you're doing. "
15:51:10braewoodsSo... who does?
15:51:18 Join lebellium [0] (~lebellium@89-92-69-66.hfc.dyn.abo.bbox.fr)
15:51:44braewoodsi've been able to build uboot before and flash that but...
15:51:54braewoodsthat was well documented
15:52:40bluebrotherthe problem simply is that while the build server ensures the bootloader builds there's no way to tell if it's working. And as said before, if you flash a bad bootloader you'll brick your player.
15:52:52speachy../tools/configure −−target=iriverh100 −−type=b && make
15:52:59*bluebrother ran into this with flashing coreboot to a laptop
15:53:08braewoodsindeed.
15:53:16braewoodsbut there's only one way to find out i suppose
15:53:19speachyand you um, break it, you're stuck with it. :)
15:53:42bluebrotherbraewoods: get a bdm setup up and running? ;-)
15:55:05braewoodsbdm?
15:55:45braewoodsspeachy: thanks. any recommended build host?
15:55:47braewoodsubuntu LTS or so?
15:56:10braewoodsor would it be better to try the build service's bootloader?
15:56:16genevinospeachy: since i'm just looking at your latest commit in git: i see both /sdcard and /mnt/sd_0 somewhere in the code, is /mnt/sd_0 just "wrong"?
15:57:24bluebrotherbdm is the jtag of the coldfire chip
15:57:39bluebrotherso to say
15:59:39braewoodshow hard is jtag to setup?
15:59:49braewoodsi've heard you usually need to...
15:59:52bluebrotherthe adapter mentioned on the wiki seems to connect to the parallel port. That could become hard these days.
15:59:56braewoodssolder pins
16:00
16:00:14braewoodsbluebrother: adapter?
16:00:44braewoodsbluebrother: which url?
16:00:53bluebrothertypically: take the adapter and plug it in. If it's not development hardware you'd need to solder the pins (assuming you know about those and can access them)
16:01:20bluebrotherand of course jtag functionality could be locked. But back in those days you still had a chance to get to those things :)
16:01:36bluebrotherthe IriverBDM wiki page I linked earlier
16:03:14braewoodsthe alternative would be desoldering the EEPROM and direct reprogramming?
16:05:57 Join johnb5 [0] (~johnb2@p5b3af332.dip0.t-ipconnect.de)
16:06:09speachyyou could use a clip-on connector on the eeprom
16:06:45braewoodsinteresting.
16:06:46 Join pamaury [0] (~pamaury@rockbox/developer/pamaury)
16:06:59braewoodsthat was how libreboot suggested first installs
16:07:03johnb5bluebrother : I am pretty confident the SD bootloader for clip+ is already active.
16:07:09speachybut you'd presumably need a full eeprom dump from a working unit.
16:07:27johnb5It's not for sansa e200.
16:07:28bluebrotherpossibly. If you get it desoldered. Looking at the pictures on the wiki it seems that one is bga
16:07:35braewoodsOh.
16:07:39bluebrotherjohnb5: good to know, I'll give that a try
16:07:52speachythere's a AT24 I2C-attached eeprom on U23
16:08:58bluebrotherhttps://www.rockbox.org/wiki/IriverHardwareComponents −− scroll to 39VF160
16:09:12bluebrotherI _think_ that's the one.
16:09:22genevinoso can i read somewhere how the exact process i need to walk through to build some custom firmware image from the git repository checkout will look like?
16:09:48braewoodsbluebrother: you're right. no legs or otherwise to grab onto.
16:09:59braewoodsso JTAG is probably only option
16:10:02genevinooh, looks like it's all in the README
16:10:04speachyah, ok, the tiny one's not actually used int he boot
16:10:54bluebrotherafaik the EEPROM isn't used at all by Rockbox but only used by the OF.
16:11:17braewoodsit is if you load the firmware to ram
16:11:21braewoodsor rom
16:11:21bluebrotherI also have no idea about internal flash (and its use) of the cpu itself.
16:12:08bluebrotherwell, if you have a working jtag you should be able to load Rockbox to RAM and then execute it. And then it should be able to use iriver_flash.
16:13:08braewoodsSeeing what it would even mount to first
16:15:17braewoodsi see one unpopulated 5 pin thing
16:15:28braewoodsprobably serial
16:15:33braewoodsbut that's not jtag
16:16:09bluebrotherserial. Check the pictures in the wiki.
16:16:31speachyBTW, there are a ton of usb-attached jtag adapters to be had. openocd supports quite a few
16:16:33braewoodsbluebrother: i was.
16:16:36bluebrotherstart with IriverPort, check the subpages.
16:16:49braewoodsjust not seeing where you'd attach the BDM
16:17:04braewoodsOh.
16:19:09braewoodsinteresting.
16:19:20braewoodsiriver_flash has hardcoded checksums so
16:19:27braewoodsyou'd need to mod it to add a new bootloader
16:19:59braewoodsgo figure
16:20:14braewoodsso it would need a custom build regardless to develop a new bootloader release
16:22:50bluebrotherspeachy: the only problem is that those usually target jtag for arm. For CF we'd need bdm. AFAIU that's different to those cheap jtag adapters.
16:24:41braewoodsbluebrother: would you suggest building from latest development or latest stable if i was trying to build a new bootloader?
16:24:51braewoodseither is far newer than current one
16:25:18braewoodsi have a server I can use for this... ECC RAM and all.
16:26:09bluebrotherbuilding the bootloader is not a problem. That's rather quick on even non-modern hardware.
16:26:39braewoodsok.
16:26:57braewoodsi just thought for something this sensitive probably should build with Debian or something.
16:27:04braewoodsand use a proper server that I have
16:27:25bluebrothertook 10 seconds on my laptop :P
16:27:28braewoodsi'll need to build a custom RB in either case
16:27:37braewoodsto patch the flasher
16:27:40speachythere's nothing special about compiling the bootloader versus any other rockbox code
16:28:22bluebrotherhttps://www.rockbox.org/wiki/pub/Main/DataSheets/COLDFIRE2UM.pdf −− it only talks about bdm, not jtag. So I guess we need bdm. And according to wikipedia that's a 1 wire interface, so not standard jtag.
16:30:02bluebrotherthough it shows the typical clock, di, do lines.
16:31:19***Saving seen data "./dancer.seen"
16:31:57bluebrotherso it might be possible to get that working with those cheap wigglers.
16:32:13*bluebrother wonders why wikipedia calls it a 1 wire interface then
16:32:40bluebrotheranyway, off for today. And I won't have the time to look into that, even though it's kinda interesting :)
16:34:17 Quit sakax (Remote host closed the connection)
16:34:17speachythere are patches for bdm and openocd, but it doesn't look like they ever landed.
16:35:45braewoodsspeachy: i guess i'll try to get a cheap throwaway h120 and experiment from there.
16:35:55braewoodsi would like to try it.
16:36:20braewoodsan updated bootloader would be nice.
16:36:28 Quit johnb5 (Quit: Nettalk6 - www.ntalk.de)
16:39:25braewoodsspeachy: would you build from master or stable?
16:39:33braewoodsnot sure myself
16:39:43braewoodssince i don't know what the existing ones were compiled form
16:41:03speachymaster is where I'd start, because if it doesn't work, at least you're at a point worth debugging.
16:42:15speachythere is very little "pure bootloader" code in the tree; it's really cut-down rockbox binary that still relies on the common code/drivers/etc.
16:43:45speachy(obviously cutting out everyhting that's not essential in the bootup..)
16:44:47 Quit tor_ (Ping timeout: 240 seconds)
16:57:35braewoodsok.
17:00
17:24:54 Quit tchan (Quit: WeeChat 2.8)
17:45:20 Join tchan [0] (~tchan@c-98-220-238-152.hsd1.il.comcast.net)
17:45:20 Quit tchan (Changing host)
17:45:20 Join tchan [0] (~tchan@lunar-linux/developer/tchan)
17:54:07 Quit pamaury (Ping timeout: 240 seconds)
17:57:04 Join ac_laptop [0] (~ac_laptop@186.2.247.129)
17:58:49 Quit lebellium (Quit: Leaving)
18:00
18:08:39speachyone of the changes in the m3k dump is enabling >96KHz audio and a kinda broken implementation of automatic freq selection.
18:10:45speachyI've been working on cleaning up that part.. but most targets seem to #define SAMPR_CAP_ALL which clearly is wrong, as that incorrectly enables 96KHz, and possibly even 48KHz.
18:31:21***Saving seen data "./dancer.seen"
18:35:22 Quit ZincAlloy (Quit: Leaving.)
18:42:13genevinocurrently trying to find out how to build a custom firmware image - can i find a ./tools/configure somewhere that someone used to successfully build an image for the agptek rocker?
19:00
19:01:26speachywhat's committed is what's used to build it.
19:01:43speachyif it's failing, the error will be informative.
19:05:58genevinoallright, thank you.
19:09:26 Join kevin__ [0] (~kevin@2601:648:8681:4d10::a347)
19:11:18kevin__i got a used clip zip. it starts up okay, but it's never recognized as a usb device on windows or linux. Tried all the obvious stuff other than opening it up. now, plugging it into usb causes a shutdown into bootloop.
19:12:01kevin__it charges on usb, but can't seem to connect to any computer without borking itself
19:13:42braewoodskevin__: anything weird like buttons behaving strangely? i had a philips gogear hdd1630 that i got used that had a stuck power button. had to open it to repair the issue.
19:15:15kevin__buttons all seem to work, music plays fine etc. it reboots and gets stuck on the flower screen whenever i plug it into a comp
19:15:49kevin__at this point i'm suspecting a faulty microusb
19:15:50kevin__port
19:16:30kevin__i cant copy stuff to flash over a new firmware... :(. if only i could install firmware from microsd.
19:19:38kevin__hmmm, so if i plug it into my sleeping laptop, it charges. and only once i open the laptop, it reboots. it's some issue with the mounting? idk
19:20:37braewoodskevin__: did you try seeing if it boots while disconnected?
19:21:21speachyany kernel messages on the linux side before it reboots?
19:24:32braewoodsyes that would also be helpful
19:24:43braewoodsclear your kernel ring buffer and then see what it spits out
19:24:53braewoodsdmesg -C
19:24:55braewoodsas root
19:27:18kevin__it's not normally in a bootloop. it only bootloop once its usb is plugged in
19:27:23kevin__[Wed Sep 30 15:05:06 2020] usb usb1-port4: attempt power cycle
19:27:24kevin__[Wed Sep 30 15:05:07 2020] usb 1-4: new high-speed USB device number 17 using xhci_hcd
19:27:24kevin__[Wed Sep 30 15:05:07 2020] usb 1-4: Device not responding to setup address.
19:27:24DBUGEnqueued KICK kevin__
19:27:24kevin__[Wed Sep 30 15:05:07 2020] usb 1-4: Device not responding to setup address.
19:27:26kevin__[Wed Sep 30 15:05:07 2020] usb 1-4: device not accepting address 17, error -71
19:27:29kevin__[Wed Sep 30 15:05:07 2020] usb 1-4: new high-speed USB device number 18 using xhci_hcd
19:27:34kevin__[Wed Sep 30 15:05:07 2020] usb 1-4: Device not responding to setup address.
19:27:37kevin__[Wed Sep 30 15:05:08 2020] usb 1-4: Device not responding to setup address.
19:27:40kevin__[Wed Sep 30 15:05:08 2020] usb 1-4: device not accepting address 18, error -71
19:27:43kevin__[Wed Sep 30 15:05:08 2020] usb usb1-port4: unable to enumerate USB device
19:27:46kevin__[Wed Sep 30 16:17:56 2020] usb 1-9: reset full-speed USB device number 4 using xhci_hcd
19:27:49kevin__[Wed Sep 30 16:17:56 2020] usb 1-8: reset high-speed USB device number 3 using xhci_hcd
19:27:52kevin__[Wed Sep 30 16:18:26 2020] usb 1-9: reset full-speed USB device number 4 using xhci_hcd
19:27:55kevin__[Wed Sep 30 16:18:26 2020] usb 1-8: reset high-speed USB device number 3 using xhci_hcd
19:29:17kevin__oops sry for spam. ive tried diff cables diff OSs, same issue so far. it was spotted once on windows (said device unrecognized), but then it started this whole bootloop on usb thing.
19:33:09 Quit MrZeus_ (Read error: Connection reset by peer)
19:33:11lonoxmontid say try and do a hard reset or power cycle on it
19:33:18kevin__yeah, so sudo dmesg | grep -i usb isnt picking up anything if i try again idk if those messages above were even related, even though it only reboots on computer plugins
19:33:32kevin__yeah i'm gonna let the battery drain fully and try some more tommorow
19:33:39lonoxmontworst case crack it open and figure out how to unhook the battery and force it to completely power off
19:34:16kevin__yeah i see this page https://www.rockbox.org/wiki/SansaAMSUnbrick (im clip zip)
19:34:35 Join MrZeus_ [0] (~MrZeus@90.203.212.4)
19:36:19kevin__thanks yall
19:39:14 Join fs-bluebot_ [0] (~fs-bluebo@55d4659c.access.ecotel.net)
19:39:26 Quit bluebrother (Disconnected by services)
19:39:31 Join bluebrother^ [0] (~dom@rockbox/developer/bluebrother)
19:41:25 Quit fs-bluebot (Ping timeout: 246 seconds)
20:00
20:06:24braewoodsi'll take the risk later
20:06:40braewoodsspeachy: i mean, i'll try to make a new bootloader later on. right now got other plans.
20:06:56braewoodsneed to wait for my CF parts to get here, etc
20:07:10braewoodsso i can actually test for the problem i'm trying to repair
20:13:30speachythat's wise.
20:15:08braewoodsspeachy: is there an official guide for testing bootloaders?
20:15:18braewoodsi saw the 7pre4 zip had a readme with something like that
20:15:41speachynot that I'm aware of, beyond "bricking is a distinct possibility"
20:15:51braewoodsok.
20:15:59braewoodsi expect v6 to remain the default for now
20:16:05braewoodsindefiniteyl for that matter
20:16:13braewoodsv7pre5 is what i'll label what i'm trying to do
20:16:32braewoodsi wonder what happened to the one who uploaded the previous one?
20:16:39braewoodsMiikaPekkarinen
20:16:41braewoodsor whoever
20:17:28speachythat way predates my involvement (as a developer, anyway)
20:17:36braewoodsah.
20:17:51speachyI used to have an ihp-120 but it was stolen out of my car during tropical storm Fey
20:18:57braewoodsouch.
20:19:06braewoodsi'm currently watching an auction for one of those
20:19:20braewoodshttps://www.ebay.com/itm/iRiver-iHP-120-Multi-Codex-Jukebox-20G-Black/133529400598
20:20:04braewoodsi was going to try to get it for testing
20:20:20braewoodsbut if you'd rather
20:21:42 Join massiveH [0] (~massiveH@ool-18e4e82f.dyn.optonline.net)
20:22:44speachy_bilgus__: comments on g#2773? Discovered it while trying to hunt down an unrelated issue. Seems like an oversight..
20:22:55fs-bluebot_Build Server message: New build round started. Revision 91197dd, 282 builds, 9 clients.
20:23:09braewoodsspeachy: i'm surprised anyone would steal an ancient mp3 player
20:23:11braewoodso.O
20:23:41braewoodsthey're not something many people want afaik
20:23:54speachyand g#2774 is what I spent most of the afternoon working on, finally builds on everything I have access to.
20:24:50speachybraewoods: TS Fay hit in August 2008. They also stole a pair of prescription sunglasses too.
20:26:07braewoodsOh. That long ago.
20:26:26braewoodsyou said Fey originally so I thought it was this year
20:26:41braewoodsOh.
20:26:45braewoodswth wikipedia
20:26:48braewoodshttps://en.wikipedia.org/wiki/Tropical_Storm_Fay_(2020)
20:27:07speachythey recycle names every what, 9 years?
20:27:15braewoodslol i see that
20:28:05speachyFay dropped something like 24" (~60cm) of rain on us..
20:31:24***Saving seen data "./dancer.seen"
20:32:01_bilgus__I wish the sim was even more tightly linked than it is currently
20:32:48braewoodsdoes the simulator work for bootloader testing?
20:32:50speachy_bilgus__: as a general rule, the sansa ams targets support 96KHz audio, right?
20:33:29speachythe simulator is only a UI simulator. doesn't run any hw-specific code.
20:33:37speachyand doesn't emulate anything at all
20:34:08braewoodsok
20:36:23speachyI'm trying to figure out how many targets that used to claim SAMPR_CAP_ALL actually supported >48KHz.
20:36:32_bilgus__braewoods, bluebrother^ the clip+ bootloader is up to date tested and used in rbutil since last? year
20:37:31_bilgus__amongst other players supported by that patch that we had on hand to test clip zip, fuze+ fuzev2 and maybe e200
20:39:00braewoods_bilgus__: interesting. i'm looking to push out a new bootloader for the iriver h120 at least.
20:39:03fs-bluebot_Build Server message: Build round completed after 968 seconds.
20:39:04fs-bluebot_Build Server message: Revision 91197dd result: All green
20:39:16braewoodsbut a bit later. mostly for fixes that were never made available in it.
20:40:21braewoodsfor sensitive builds i usually use my LXD containers on my server
20:40:29braewoodssince it has ECC RAM and such at its disposal
20:42:39genevinoso what's a recommendable environment for building? i noticed sh-elf-gcc is only available from the AUR on arch/artix and it's not exactly trivial to use that.
20:43:46braewoodsmost things prefer debian or ubuntu
20:43:50braewoodsespecially cross-compiling
20:43:56genevinogotcha.
20:44:08braewoodse.g., openwrt, uboot, and lineage os
20:44:10braewoodsfor starters
20:47:54speachygenevino, https://www.rockbox.org/wiki/CrossCompiler
20:48:08speachystep 2a works quite well.
20:49:42genevinooh wow, thank you.
20:50:20braewoodsor
20:50:22braewoodshttps://www.rockbox.org/wiki/Main/LinuxSimpleGuideToCompiling
20:51:01speachythe whole page is out of date except for 2a. heh.
20:51:13braewoodsi updated one of them to include wget
20:51:19braewoodssince my container was pretty bare
20:53:04genevino"Instructions here might be obsolete. as nobody does it manually really."
20:53:04genevinohehe
20:53:14_bilgus__speachy I'd just say its an oversight was before me but I do wonder HTH you found it
20:53:22_bilgus__(2773)
20:54:44speachyI put an #error check in config.h if none of the target config files got #included
20:56:09speachyokay, this is odd. my clip+'s screen is _very_ dim and washed out. only when the USB is plugged in is it visible too.
21:00
21:00:40speachy_bilgus__: if you have a clip+ (or any sansa ams) player handy with the test plugins installed, can you tell me if it has 96KHz as an audio rate, and if it actually works?
21:01:19fs-bluebot_Build Server message: New build round started. Revision 127862c, 282 builds, 9 clients.
21:09:45braewoodsspeachy: is that something you can view from the debug options from the standard stable build?
21:10:01braewoodsi have a sansa clip+ with RB 3.15 already installed
21:10:32speachyNo, I need to see if it actually works, as on 3.15 it claims to support it.
21:11:06braewoodshow would I know? i don't have anything that supports 96khz afaik
21:11:07speachythough.. with a dev build you could try to dial the mikmod plugin up to 96KHz
21:12:01speachyI think there are a lot of targets that used to have SAMPR_CAP_ALL when that was limited to 48KHz, and when the upper cap got bumpted to 96, those weren't properly audited/updated.
21:13:34speachy_bilgus__: I think thre's going to be some red in this build. :/ Just discovered ipod nano broke with the lua change.
21:15:21fs-bluebot_Build Server message: Build round completed after 843 seconds.
21:15:22fs-bluebot_Build Server message: Revision 127862c result: 149 errors 0 warnings
21:24:40fs-bluebot_Build Server message: New build round started. Revision c2c5945, 282 builds, 9 clients.
21:25:01speachyso, looks like lua will gain the ability to interoperate with the remote control buttons on the ipods.
21:26:57speachycool, gave it a whack and the screen started working again
21:30:25speachyokay, clip+ does _work_ with the higher rates but they're off a bit; only 88K sounds correct.
21:38:12fs-bluebot_Build Server message: Build round completed after 812 seconds.
21:38:14fs-bluebot_Build Server message: Revision c2c5945 result: All green
21:38:26speachyanyone here have an imx233-based device? (fuze+, most creative zens)
21:39:11speachy(or the sony e360/e370, and samsung yp-z5)
21:43:00speachyand an ipod6g/classic (since the nano2g works with 96K, the same-guts 6g/classic should too..)
21:43:58fs-bluebot_Build Server message: New build round started. Revision 01650b8, 282 builds, 9 clients.
21:44:44_bilgus__sorry AFK
21:48:23_bilgus__speachy I have a fuze+ @ home I can get it this week sometime
21:50:49speachyno worries. of the imx233 players, only the sonys were seemed to be advertised with 96K support, so I left the rest at 48.
21:52:40_bilgus__I think we already tried this on the clip+ or zip at one point
21:53:06speachyI also have a codec here that's floating-point. which pretty much limits its use to the newer hosted linux targets (ie armhf or mips) with FPUs.
21:53:28_bilgus__ended up settling for decent 48k and good 44
21:53:58speachy88K sounds good, the others above 44 are a bit off.
21:54:15_bilgus__yeah the int fp stuff is black magic as it is
21:54:39_bilgus__88 doesn't seem like a common format though
21:55:02speachy88 and 172 are common DSD->PCM downsampling targets.
21:55:11_bilgus__not that 96k is either :p
21:56:05speachyalso there's a ticket asking about supporting various DSD-ish file formats; I found three different libraries to do DSD->PCM conversion but they all use FP too.
21:56:29speachywhich is probably fine, as the only targets that have enough ooomph to do that crap (and support high bitrates) all hve hardfloat anyway.
21:57:00speachythough I can't imagine the jz4760's threading code is sane with respect to the FP state.
21:57:25speachy(threading & interrupt)
21:58:27speachyand doing native DSD playback on the HW that supports it would mean bypassing our mixer layer altogether. Might be feasible in a PoC plugin though.
22:00
22:06:03speachyStill have a few more chunks to pull/rework out of the m3k dump before I can properly tackle the m3k-specific code.
22:12:27fs-bluebot_Build Server message: Build round completed after 1708 seconds.
22:12:28fs-bluebot_Build Server message: Revision 01650b8 result: All green
22:12:29fs-bluebot_Build Server message: New build round started. Revision 1cd004f, 282 builds, 9 clients.
22:31:27***No seen item changed, no save performed.
22:34:12fs-bluebot_Build Server message: Build round completed after 1304 seconds.
22:34:13fs-bluebot_Build Server message: Revision 1cd004f result: All green
23:00
23:11:03fs-bluebot_Build Server message: New build round started. Revision cb9b5d3, 282 builds, 9 clients.
23:27:33 Quit ac_laptop (Ping timeout: 258 seconds)
23:28:35 Quit MrZeus_ (Ping timeout: 240 seconds)
23:31:47fs-bluebot_Build Server message: Build round completed after 1244 seconds.
23:31:48fs-bluebot_Build Server message: Revision cb9b5d3 result: All green
23:32:31 Quit massiveH (Quit: Leaving)
23:40:55 Quit TheSeven (Ping timeout: 240 seconds)
23:41:33 Join TheSeven [0] (~quassel@rockbox/developer/TheSeven)

Previous day | Next day