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 2020-10-17

00:13:47efqwspeachy: something else for the wishlist: comprehensive Bluetooth AVRCP support
00:13:55efqw(if BT is a thing of course)
00:14:27efqwIf you connect the device to a car stereo, it'd be cool to have the cover art and metadata available.
00:16:53efqwThe minimal profile set we need to support would be AVRCP + A2DP, add HFP if you want recording over BT, and possibly mSBC (wideband speech) to improve the audio quality.
00:17:20efqw(mSBC is the optional codec, not a profile)
00:23:33 Quit emacsomancer (Read error: Connection reset by peer)
00:29:08 Quit massiveH (Quit: Leaving)
00:30:55_bilgus__atm only hosted targets have bt support?
00:34:52efqwyea, this is mostly just ideas for the wishlist
00:35:24 Join emacsomancer [0] (
00:39:33***Saving seen data "./dancer.seen"
00:43:31_bilgus__I mean I hope we someday get these things bare metal butTBH it'd be enough to make a plugin as long as the backend existed to control it
00:50:05 Quit emacsomancer (Read error: Connection reset by peer)
00:59:28 Quit Stanley|00 (Remote host closed the connection)
01:02:37 Join emacsomancer [0] (
01:07:13efqwmy personal interest is mostly on the hosted side but I currently still don't own a hosted target that has BT
01:07:51efqwthe fiio m5 fits the bill but taking it apart is a pain (I want console access for a dev machine)
01:08:18efqwmight look into the eros stuff sometime
01:13:58efqwbut the eros devices only have 32MB of ram, and I'd rather have 64MB just to be safe
01:18:35_bilgus__this osd code is baffling me
01:20:14_bilgus__If i flip it to view the back buffer it works almost properly but it clearly has a good copy in it yet if I try to display the actual osd buffer shows garbled junk when it loads from the back buffer
01:40:49 Join Stanley00 [0] (~stanley00@unaffiliated/stanley00)
01:42:49Stanley00efqw: fiio m5 is cool, if adb shell access is enough for you, I can help you with that
02:19:13efqwknowing myself I'd probably try to pull it apart and rip the screen fpc, lol
02:21:01Stanley00in that case, I think follow the guide of Shanling M0 could help
02:21:25Stanley00it's strange that no one ever teardown the M5 but M0
02:22:05speachy__builtin: yep, they need to generally work, be installable via rbutil, and have a manual
02:23:12efqwspeachy: anything else you need to test for the m3k?
02:23:13speachyefqw: BT-wise, on these hosted platforms we might be stuck with whatever hiby implemented, which isn't much more than basic audio.
02:24:15speachywe need to get discovery/pairing/etc working first, then figure out how to hook up the audio side of things.
02:24:45speachyand then figure out if more advanced stuff is feasible with what we have.
02:25:03speachy(ie without recompiling anything in the base platform)
02:25:14efqwwhich hiby based thing would you recommend getting?
02:25:23efqwI want x1000e if possible
02:25:37speachythe m3k is hte only one I'm aware of with the -e
02:25:58Stanley00looking at kernel source code for M5, they control most the thing via fiio_ctrl driver
02:26:11speachythere may be other touchscreen based ones but I didn't bother looking at any of those.
02:26:21speachybecause I want real buttons, damnit.
02:26:47Stanley00actually, just 3 buttons is not bad, you know
02:27:04efqwm3 pro has x1000e as well but it doesn't have bt
02:27:07speachythe rocker is reasonably solid, and the eros q/k are also good, albeit much larger physically.
02:27:14Stanley00play/pause, vol+/-, and ff/rw, next/prev should be fine with just 3 buttons
02:27:52Stanley00efqw: please, m5 is also x1000e and with bt, and touchscreen :3
02:28:28efqwyeah, it's on my list but I think it's a bit *too* small if you get what I mean
02:28:45efqwthey made it so tiny and the battery is really small
02:29:06efqwa 1000mAh+ battery would be nicer
02:29:41speachyif I were to pick a target for a native port it would be the eros.
02:29:42Stanley00I see, but it can keep solid 10+ hours playing, so I think it's not that bad
02:31:13speachyanyway. if the pine64 thing bears fruit in the end, it's not goign to be an x1000[e].
02:31:39Stanley00I was playing around with M5, and got the music play, and some games work too. But having another developer working on it should be great
02:32:45speachyStanley00: patches (in gerrit) welcome!
02:33:20Stanley00speachy: sure, when I feel confident enough, currently, the code is just a mess, and it still use m3k build script
02:33:42_bilgus__we jump in where we find bugs or something interesting or in the case of speachy everywhere :p
02:34:17speachyI'm not terribly motivated to work on hardware that I wouldn't personally want to use
02:34:31Stanley00also, I can't seems to find a way to install the zip for M5, I'm using adb push for now :(
02:35:45speachyanyway. it's 0235 here, and I _Really_ need to get some sleep. g'nite y'all.
02:35:53_bilgus__thats kinda the whole RB ethos anyway :)
02:36:24Stanley00night speachy
02:39:37***Saving seen data "./dancer.seen"
03:02:22 Join johnb7 [0] (
03:40:07 Join S|h|a|w|n [0] (~shawn156@unaffiliated/shawn156)
03:48:05 Join bonfire [0] (
04:17:46 Quit Stanley00 ()
04:37:13 Join sakax [0] (~r0b0t@unaffiliated/r0b0t)
04:39:39***Saving seen data "./dancer.seen"
04:44:19 Join johnb3 [0] (
04:49:18 Quit Acou_Bass (Ping timeout: 256 seconds)
04:52:50 Join Acou_Bass [0] (
04:58:01 Quit S|h|a|w|n (Read error: Connection reset by peer)
04:58:08 Join lebellium [0] (
04:59:05 Quit Acou_Bass (Ping timeout: 240 seconds)
05:02:38 Join Acou_Bass [0] (
05:04:02lebelliumspeachy: Rocker fixed. Thank you!
06:09:34 Quit Oksana (Read error: Connection reset by peer)
06:11:38 Join pamaury [0] (~pamaury@rockbox/developer/pamaury)
06:19:31lebelliumspeachy: YP-R0 broken (fresh install latest build). Rockbox doesn't even start, immediate crash. My previous build (20200915) is working
06:34:11lebellium4f8736909a-201011 OK
06:34:25lebellium135b3f6416-201011 NOK
06:38:12lebelliumcould be db6f21e295: hosted: sanitize handling of HOME_DIR ?
06:39:43***Saving seen data "./dancer.seen"
07:21:14 Quit johnb7 (Ping timeout: 272 seconds)
07:31:27efqwupon further inspection, the allwinner v3 doesn't seem to have I2S at all?
07:31:37efqwto me this is kind of a deal breaker
07:56:32 Join MrZeus [0] (~MrZeus@2a02:c7f:70d0:6a00:f0dd:4f09:1639:8b72)
08:02:01 Quit pamaury (Ping timeout: 264 seconds)
08:06:43 Quit johnb3 (Quit: Nettalk6 -
08:20:41lebelliumAgptek H3 battery benchmark
08:39:46***Saving seen data "./dancer.seen"
08:49:16 Join MrZeus_ [0] (~MrZeus@2a02:c7f:70d0:6a00:fd44:d60b:489:2f5e)
08:53:35 Quit MrZeus (Ping timeout: 272 seconds)
09:32:15 Join Rower [0] (
09:32:22 Quit Rower (Client Quit)
09:48:11 Quit TheSphinX^ (Ping timeout: 240 seconds)
09:50:20 Join cockroach [0] (~blattodea@pdpc/supporter/active/cockroach)
10:23:51speachyefqw: it has bi-directional i2s.
10:39:50***Saving seen data "./dancer.seen"
10:47:09speachyefqw: section of the S3's datasheet. (8-32bit, 8-192KHz)
10:47:33speachy(Sochip S3 == Allwinner V3, S3L = V3s)
10:48:51speachylebellium: glad the rocker works now; and your H3 lasted a few minutes longer than mine, so I'm going to leave the curve as-is
10:49:00 Join pamaury [0] (~pamaury@rockbox/developer/pamaury)
10:51:23speachyas for the YP-R0, can you confirm a few things for me? sd card is /mnt/media0, and all rockbox files live under there?
10:53:46lebelliumNo. media0 is internal memory. media1 is hidden internal memory, mmc is SD card
10:54:00lebelliumrockbox files are on media0
10:55:22speachyok, then in that case rockbox should at least _start_. How much low-level access do you have? Any possibility of manually starting the rockbox binary with strace? (if it's even present?)
10:56:18 Join MrZeus__ [0] (~MrZeus@2a02:c7f:70d0:6a00:fd44:d60b:489:2f5e)
10:59:37 Quit MrZeus_ (Ping timeout: 272 seconds)
11:03:44lebelliumI'm afraid your questions are too technical for me :/ the developer is lorenzo92 but no longer active unfortunately
11:04:30speachyok. :)
11:05:00lebelliumI only know that in the hidden memory media1 we put
11:05:13lebelliumthis file defines how to launch Rockbox
11:05:57speachyok. so ypr0 uses MULTIDRIVE, which is incompatible with PIVOT_ROOT. Need to figure out how to fix that properly.
11:06:13*speachy sighs.
11:12:03speachylebellium: do you happen to have a DX50 or DX90?
11:12:15speachyor any of the Sony NWZ models?
11:13:07lebelliumI don't have any Ibasso but have several NWZ, including E580, A10 and A20
11:16:40 Join TheSphinX^ [0] (~briehl@2a01:4f8:c17:714f::3)
11:21:49speachyI probably bnroke them in similar manner
11:26:30lebelliumshould I check?
11:27:05speachyif you don't mind.
11:32:13lebelliumI thought pamaury updated his NWZ every day to make sure they always work :)
11:35:01 Join johnb3 [0] (
11:35:19lebelliumcontrary to YP-R0, at least here I have a *panic*
11:35:23lebelliumpanic cannot read /dev/input directory
11:35:29lebelliumfor NWZ-A10
11:37:45 Quit pamaury (Ping timeout: 240 seconds)
11:43:38 Join ac_laptop [0] (~ac_laptop@
12:00:19 Quit ac_laptop (Quit: WeeChat 2.9)
12:39:51***Saving seen data "./dancer.seen"
12:46:37 Join johnb7 [0] (
13:19:08speachylebellium: ok... that's unexpected. probaly related, but will have to check.
13:19:48speachybut.. yard work comes first.
13:20:45 Quit johnb3 (Quit: Nettalk6 -
13:26:03 Join MrZeus_ [0] (~MrZeus@
13:30:21 Quit MrZeus__ (Ping timeout: 272 seconds)
13:42:30 Quit johnb7 (Ping timeout: 272 seconds)
13:57:28 Join S|h|a|w|n [0] (~shawn156@unaffiliated/shawn156)
14:18:28 Join pamaury [0] (~pamaury@rockbox/developer/pamaury)
14:22:53genevinospeachy: just read your messages on the away-log - thx for getting back, will be able to test in 1-2hrs. :)
14:25:10fs-bluebotBuild Server message: New build round started. Revision 0e667ff, 291 builds, 10 clients.
14:25:22speachylebellium: the build going now should fix the Sony NWZ panic you ran into.
14:26:21speachythe YP-R0/R1 is a more challenging beast, due to the internal/external storage split.
14:27:30lebelliumYP-R1 has no external storage
14:27:43lebelliumonly the two internal partitions media0 and media1
14:28:03speachyah, then the odds are decent that the YP-R1 will JustWork(tm).
14:29:26speachyno visible reason for the -R0 not working? Any display output at all?
14:30:54lebelliumno, black screen
14:31:09lebelliumno panic or anything
14:32:24lebelliumbut I don't get the difference between YP-R0 and NWZ-A10, they both have internal and external storage
14:33:37speachyfound it
14:34:42speachythe low-level code was, in a few places, using the internal/cooked open() etc instead of the proper OS open()
14:35:01speachy(in the LCD driver and PMU code)
14:35:25speachystill doing a build, when it's done I'll make sure I caught everything
14:35:49speachynow external storage will be inaccessible.
14:35:49 Quit _bilgus__ (Remote host closed the connection)
14:37:45 Join _bilgus__ [0] (~bilgus@
14:38:20speachyokay, just committed the fixes for the R0/R1.
14:39:53***Saving seen data "./dancer.seen"
14:40:35fs-bluebotBuild Server message: Build round completed after 926 seconds.
14:40:40fs-bluebotBuild Server message: Revision 0e667ff result: All green
14:40:42fs-bluebotBuild Server message: New build round started. Revision 13f5e12, 291 builds, 9 clients.
14:42:15speachylebellium: rev 13f5e12 has the YP fixes
14:43:33speachylebellium: the differnce in behavior betwen the R0 and A10 is their different low-level driver code.
14:43:47speachyon both we haven't got to the point where we can access rockbox proper.
14:45:15speachyI haven't yet found anything ont he NWZ about where the sd card gets mounted
14:45:26speachy(versus the internal storage)
14:53:44fs-bluebotBuild Server message: Build round completed after 781 seconds.
14:53:49fs-bluebotBuild Server message: Revision 13f5e12 result: All green
14:53:55speachyI think I have a fix for the R0's external storage too.
14:55:36lebelliumI confirm YP-R0 is now working
14:55:48speachyinternal storage only, correct?
14:57:08lebelliumwell, I can only see media0 now, not even media1 (and not mmc either)
14:57:28speachymedia1 is read-only for OS-stuff only, correct?
14:59:11lebelliumnot read only.
14:59:26lebelliumwe put the rockbox loader there
14:59:33lebelliumbut it's an hidden partition with OF
14:59:46speachyie not intended to be touched by runtime stuffs
15:00:37lebelliumwe're not supposed to touch it but we should be able to see the content in the rockbox browser
15:01:18 Join johnb7 [0] (
15:01:24lebelliumOn NWZ-A10 the panic is gone but I can't see the sd card either
15:04:58lebelliumbasically, I think Rockbox should display all partitions in the file browser. Then it's up to the user to set "start browsing here" at the directory he wants
15:06:12speachyI've been undoing a lot of special cases trying to work around that ability
15:07:23speachysee if that allows you to see the sd card via the browser
15:08:31speachy(it's supposed to present itself as a special "link" on the top level of the browser)
15:10:28speachybasically the goal is to mimic the behavior of native targets. and simulator builds too. only show rockbox's "internal" and "external" storage.
15:12:38lebelliumstill only displays media0
15:17:53 Quit johnb7 (Ping timeout: 260 seconds)
15:36:06 Quit pamaury (Ping timeout: 258 seconds)
15:40:08speachyheh, aren't even any of tehse things listed on ebay
15:42:32lebelliumYP-R0 regularly shows up on ebay as it was quite popular in the UK, Germany and Italy
15:44:13lebelliumit was not sold in America and Oceania though
15:44:38lebelliumSony NWZ should be easy to find
15:49:31speachysee if there's an entry in the "top" level file browser that points at the SD card. in theory it should be called whatever the underlying volume name happens to be.
15:50:49speachywhoops, missed something. gimme a momen
15:52:06lebelliumI now have <Volume1> inside media0
15:52:15speachydoes it point at the SD card?
15:52:47speachyand do they play back properly? :)
15:52:52lebelliumbut obviously it shouldn't be here
15:53:13speachyThis is how things are _supposed_ to work btw, in keeping with all of the other rockbox targets (and simulator)
15:53:16lebelliumyes playback is fine
15:53:47speachytop-level is the volume where .rockbox lies, with <whatever> pointing at the external storage
15:54:11speachyno visibility to the rest of the running system (if there even is any)
15:54:49lebelliumHum I just checked my Clip Zip and indeed <microSD1> is inside the internal memory browser
15:54:55lebelliumDidn't remember
15:55:07lebelliumI prefer how it was before. media0, media1 and mmc :)
15:55:17speachy(damnit, this means I need to redo how the ingenic hosted targets mount external USB storage)
15:56:04speachy(...since I fixed the underlying problem preventing them from working how I want)
15:56:42speachy(and this also means respinning the bootloader images for everything. @~#&%^$#!!!)
15:57:31lebelliumCan you rename <Volume1>?
15:57:43speachyyes, it should be called <microSD>
15:59:05lebelliumWon't be able to open with the embedded rockbox text viewer anymore :(
16:00:03speachywhen wodz did the original Rocker port he got about 3/4ths of the way towards making hosted storage work the same way as everything else.
16:00:29speachytbh the hosted targets were all rather "special" and not quite entirely working when it came to paths
16:00:42speachylots of hacks in the code for special cases; most of those are gone now
16:01:36speachyif you could confirm the <microSD> name in that -c image, I can move on to fixing the rest of the mess I made. Do you know where the NWZ targets put the external SD storage?
16:02:12lebelliummaybe but giving the opportunity to see the system files was fine and fun. I think it would be better to leave this opportunity (even if hidden by default) than trying to match the usual targets
16:03:38speachy(on them, '/storage' is the only path mentioned in the rb source)
16:05:39lebelliumit's written <microSD1> as on Clip Zip
16:06:26lebelliumNWZ have emmc, media and sd
16:06:28speachyI mean, should report <microSD> instead of <Volume>
16:06:59lebelliumYes, I tell you, it's written <microSD1> (not <microSD>), as on Clip Zip
16:07:30lebelliumand I couldn't know that it's emmc, media and sd on NWZ if I couldn't see the system files :)
16:08:09speachyah, okay.
16:08:15speachysorry, mis-understood you there
16:08:59 Join johnb7 [0] (
16:10:53lebelliummedia is internal storage. sd and emmc are empty although I have a sd card inserted
16:11:17speachyI didn't do anything that would affect that though.
16:13:14lebelliumso mnt\emmc and mnt\sd are empty. But I also have allcontents\int and allcontents\ext
16:13:22lebelliumand here ext I see the SD
16:16:24speachyok. sigh.
16:18:01speachywhere is .rockbox located?
16:18:17speachyjust trying to figure out how sony mapped everything..
16:19:51lebelliumhum my mistake. As I have the same directories on the internal storage and my SD, it's quite confusing. I'll put a text file to make the difference
16:23:40 Quit johnb7 (Quit: Nettalk6 -
16:24:13lebelliumcontents = allcontents\int = internal memory (where .rockbox is). allcontents\ext = mnt\media = sdcard
16:26:05lebellium.rockbox is also visible directly at the file browser root (so even before getting to contents or allcontents\int)
16:32:10lebelliumhum I notice the device automatically restarts instead of shutting down
16:34:36lebelliumwell, that's with the 2017 build. Will see with a fresh one
16:39:55***Saving seen data "./dancer.seen"
17:11:23fs-bluebotBuild Server message: New build round started. Revision 012908e, 291 builds, 10 clients.
17:26:38fs-bluebotBuild Server message: Build round completed after 916 seconds.
17:26:40fs-bluebotBuild Server message: Revision 012908e result: All green
17:54:20_bilgus__Sometime I really really hate casts
17:55:12_bilgus__I spent a day trying to find a bug that was directly the result of (void *) covering up my improper dereference of a pointer
17:55:28_bilgus__on the plus side I really cleaned some stuff up trying to find it
18:03:15speachyI'm down another three rabbit holes myself
18:03:44speachybut yeah, I'm not a fan of explicit casting to/from void*
18:06:39_bilgus__I got rid of some but unfortunately some are still reqd
18:07:43 Join pamaury [0] (~pamaury@rockbox/developer/pamaury)
18:07:44speachyI'm still unwinding the layers and layers of hacks in the hosted filesytem code
18:10:59pamauryspeachy: you mean the rockbox hosted FS code is a hack? I am shocked :P
18:12:06speachypamaury: oh, it worked well enough for the first target(s), the problem was that each new platform layered new hacks on top of the old ones, and they don't like to share. :)
18:14:07_bilgus__speachy remember that file system redirect patch of JhMikeS' that I totally failed to rebase?
18:14:27speachy...not really
18:14:34_bilgus__let me find it
18:16:43_bilgus__ g#1613
18:16:45fs-bluebotGerrit review #1613 at : Multiboot Firmware Root Redirect by William Wilgus
18:18:03_bilgus__sorry that isn't it it is the parent of that
18:18:20speachyJust after I published new bootloaders for the hiby targets.
18:18:26speachyI need to respin them all. $@#%@%
18:19:30speachybut the good news is that I now have proper hotplug going via the rockbox-native multidrive framework
18:20:04_bilgus__ g#1556
18:20:05fs-bluebotGerrit review #1556 at : Allow mounting of any directory as the root directory. by William Wilgus
18:21:28speachyyou merged that ~2mo ago...
18:22:09speachyunfortunately the hosted stuff sorta bolts onto the side of everything and is a petulant misbehaving child
18:25:12_bilgus__and I reverted it
18:25:23speachyooohyeah, that's right
18:25:46speachysome of the craziness you ended up with might be mitigated with the changes I've been making to the hosted code
18:26:08_bilgus__I'll eventually revisit but I'm pretty sure it would make what you want to do pretty seamless
18:26:45speachymost of what I'm doing involves unwinding hardcoded assumptions about paths. and mixing os-native and rockbox-cooked filesystem calls
18:28:02_bilgus__what that does is allows you to set sym links essentially
18:37:34lebelliumspeachy: a user complained that the sleep timer doesn't work on YP-R0. I could fix this issue with adding #define HAVE_POWEROFF_WHILE_CHARGING to samsungypr0.h. I don't know if lorenzo92 forgot it back to the time or if there was a specific reason not to have it. It's relevant, maybe it could be added to trunk?
18:37:51lebelliumIf it's relevant*
18:37:54speachyif it does the trick, I dont' see why not
18:39:58***No seen item changed, no save performed.
18:40:22fs-bluebotBuild Server message: New build round started. Revision 0a4a920, 291 builds, 10 clients.
18:44:54lebelliumcould you also add YP-R1 to the theme page? I have an old 2013 touch theme for it I should rather share instead of keeping on my computer
18:52:32fs-bluebotBuild Server message: Build round completed after 730 seconds.
18:52:37fs-bluebotBuild Server message: Revision 0a4a920 result: All green
18:52:38fs-bluebotBuild Server message: New build round started. Revision 6f9a157, 291 builds, 10 clients.
19:20:31fs-bluebotBuild Server message: Build round completed after 1673 seconds.
19:20:32fs-bluebotBuild Server message: Revision 6f9a157 result: All green
19:29:34 Quit pamaury (Ping timeout: 272 seconds)
19:44:35 Quit Ckat (Excess Flood)
19:44:59 Join Ckat [0] (~Ckat@xn--z7x.xn--6frz82g)
19:45:40 Quit bluebrother (Disconnected by services)
19:45:45 Join bluebrother^ [0] (~dom@rockbox/developer/bluebrother)
19:45:50 Join fs-bluebot_ [0] (
19:48:33 Quit fs-bluebot (Ping timeout: 260 seconds)
19:53:07 Quit lebellium (Quit: Leaving)
20:40:00***Saving seen data "./dancer.seen"
20:48:55 Quit MrZeus_ (Ping timeout: 260 seconds)
20:50:47speachylebellium: Added to the themesite. If any compatible themes are present they will get accounted for in the next theme refresh (15 minutes after the next build finishes)
20:54:36 Join massiveH [0] (
21:05:38fs-bluebot_Build Server message: New build round started. Revision b1e21bb, 291 builds, 10 clients.
21:06:46speachylebellium, also one more question −− on the yp-r0, what is the device node for the sd card? (eg /dev/mmcblk0 ?)
21:25:22fs-bluebot_Build Server message: Build round completed after 1184 seconds.
21:25:23fs-bluebot_Build Server message: Revision b1e21bb result: 19 errors 3 warnings
21:41:36fs-bluebot_Build Server message: New build round started. Revision edbdefa, 291 builds, 10 clients.
21:43:02speachyand with this, it's time to rebuild the bootloader binaries and patches. At least I have that scripted now.
21:47:51_bilgus__well everything was going well with my char * void * fb_data* fb_remote_data * union till I discovered that my current arm toolchain doesn't like anonymous unions
21:51:00speachystill on 444 I guess?
21:51:12 Quit beencubed (Quit: Leaving)
21:53:20_bilgus__yeah and I truly hesitate to start locking stuff to compiler versions
21:54:02_bilgus__at least named unions are forward compatible I guess
21:54:18speachywell, we do officially use 494 everywhere now. :)
21:54:41fs-bluebot_Build Server message: Build round completed after 785 seconds.
21:54:43fs-bluebot_Build Server message: Revision edbdefa result: All green
21:55:33_bilgus__hmm ok but only because its so damn ugly using named unions :p
21:59:11speachyin all seriousness the nice thing about a more "modern" baseline compiler is so we dont' have to put in workarounds for older crap
21:59:47_bilgus__yeah I was drifting towards some kind of macro hack but this sounds better
22:00:02_bilgus__I'll just have to remeber to update at home next time I'm there
22:01:27_bilgus__so OSD finally works RIGHT
22:01:51_bilgus__still chasing some weird issues on 1 bit screens (clip+)
22:02:21 Quit cockroach (Quit: leaving)
22:02:26speachyI'm regenerating a bootloader set for the hiby targets, AGAIN.
22:02:34speachyat least it's scripted now
22:04:53_bilgus__the one with the upgrade in the bootloader?
22:05:45_bilgus__I noticed on the rocker at least that you can't shutdown till you boot into a fw
22:05:57_bilgus__mildly annoying
22:08:43efqwbetter than the m3k I suppose?
22:09:02efqwyou can't shut down at all unless you use the stock fw
22:10:03speachy_bilgus__: the loader has a shutdown option in the tools menu
22:11:11speachythis new iteration has proper hotplug of both SD and external USB, identically integrated to how our native targets handle it.
22:11:43speachy(had to change the mountpoint and fix a couple of bugs in the integration scripts)
22:15:47 Join Stanley00 [0] (~stanley00@unaffiliated/stanley00)
22:16:27speachythe m3k shutdown problem might be fixed now.
22:18:44_bilgus__oh nice I missed it
22:23:44*speachy waits for 450MB of images to upload.
22:25:02speachygive me another ~15 min and you'll be able to use rbutil to install the newly-updated stuff
22:26:09Stanley00hmm... I was trying to make a separate build for M5, but the whole plugin things is scary...
22:26:30Stanley00can we have a way to disable all plugins by default for the initial port?
22:26:56speachyedit apps/plugins/SOURCES and add #if 0 / #endif at the top/bottom
22:27:02speachyditto for apps/plugins/SUBDIRS
22:27:25Stanley00ah, I see, thanks
22:28:46Stanley00I saw that we have `plugins="yes"` in tools/configure, let me see if we can use that config somehow
22:29:50efqwthe ridiculous part is that the m3k's power controls are broken, lol
22:30:11efqwas far as I can tell from the logs, the player binary tells axp192 to cut the power
22:30:14efqwnot the kernel
22:30:35speachyefqw: oh that's awful.
22:33:09efqwthis is probably one of the hackiest linux device I've ever seen
22:36:13Stanley00speachy: found it, <= we need to check for this ifeq (yes,$(ENABLEDPLUGINS)) too
22:40:05***Saving seen data "./dancer.seen"
22:45:21_bilgus__theres actually a lot of 'hidden' functionality in the config stuff
22:45:40_bilgus__thumb stuff even but I think that may be enabled now?
22:45:53_bilgus__arm thumb..
22:46:14_bilgus__I don't know if that made it clearer or not
22:47:48_bilgus__annnnd 1 bit seems to be fixed now after setting it to the right union member
22:49:40speachythe thumb stuff is still going through that hacky wrapper script
22:52:33speachyIIRC we only turn it on when we need the smallest possbie code size.
23:04:01efqwfound a function like this: `open64("/dev/axp173",2);`
23:04:40efqwcan this get any worse
23:06:29Stanley00wait, that looks familiar to me too =]]
23:09:19efqwso apparently there are no proper drivers for the dac and the pmu
23:09:29efqwit's all ioctl() calls from the player binary
23:42:16Stanley00speachy: <= not sure if this is fixed, but fiio is still using -march=mips32 only, bump to mips32r2 make it crash on my m5
23:49:55 Quit [7] (Ping timeout: 240 seconds)
23:50:31 Join TheSeven [0] (~quassel@rockbox/developer/TheSeven)

Previous day | Next day