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 2022-09-19

00:37:10 Quit m01 (Quit: Konversation terminated.)
00:39:05 Join m01 [0] (
00:44:40***Saving seen data "./dancer.seen"
01:29:04 Join smithjd [0] (
02:04:41 Quit smithjd (Ping timeout: 250 seconds)
02:33:49 Join smithjd [0] (
02:44:44***Saving seen data "./dancer.seen"
02:45:15 Quit smithjd (Ping timeout: 268 seconds)
03:15:13 Quit _bilgus (Ping timeout: 252 seconds)
04:44:47***Saving seen data "./dancer.seen"
06:44:51***No seen item changed, no save performed.
08:44:52***No seen item changed, no save performed.
09:53:41 Join massiveH [0] (
10:36:51 Quit akaWolf (Ping timeout: 265 seconds)
10:44:54***Saving seen data "./dancer.seen"
10:53:44 Join nihilazo [0] (
10:54:18nihilazoare sansas still the best cheaper players to run rockbox? On ebay they seem to have been going up in price since my last one died and I'm wondering if there are better options
10:56:16nihilazoI previously also had an x3 but the whole player died when one button failed which was frustrating
11:14:12 Quit massiveH (Quit: Leaving)
11:20:06 Join akaWolf [0] (
12:12:02 Join buZz [0] (
12:12:07buZzooooo \o hello
12:12:48buZzi was trying to build rockbox-as-a-application on newish debian on ARM , but many issues regarding inline ASM not specifying if it was thumb or ARM asm
12:12:57buZzis this a known issue?
12:34:06 Quit michaelni (Ping timeout: 260 seconds)
12:44:56***Saving seen data "./dancer.seen"
13:22:05 Join michaelni [0] (
13:43:40 Quit munkis (Remote host closed the connection)
13:44:01 Join munkis [0] (
13:45:15 Join lebellium [0] (
14:00:02paulcarrotynihilazo, probably yes if you're looking for sdcard and/or radio support.
14:20:50 Join amachronic [0] (~amachroni@user/amachronic)
14:33:25amachronicbuZz: if you are hitting inline asm when building the sdl app on ARM i'd say something has gone wrong
14:34:27buZzi agree :P
14:34:39amachroniccould you post build logs?
14:34:45buZzi -think- this was on gcc8 , i'll collect more logs soonish when i redo
14:35:03buZzwait, maybe i can do it right now
14:37:52rb-bluebotBuild Server message: New build round started. Revision 6beebd75e7, 303 builds, 9 clients.
14:44:59***Saving seen data "./dancer.seen"
14:46:51buZzamachronic: the first i hit is
14:48:51buZzadding -Wa,-mimplicit-it=thumb to gcc options gets past that , but later its still 'implicitly thumb' , yet its arm asm (iirc in libmad)
14:50:01amachronicokay, I guess the problem is from filters_arm4.h maybe if you disable that include in libspeex/filters.c it'll get past it?
14:50:17amachronicat least until the next error :)
14:50:33buZz^_^ lets see
14:50:49buZzi assume if i dont add implicit thumb , it would just assume the arm is arm
14:50:53popcorn9499[m]the hunt for the next error continues with Hobbes on his journey!
14:52:39buZzok, continueing make, it gets past filters.c
14:52:46buZznow lsp.c :P same thing i assume
14:52:58popcorn9499[m]sorry for going off topic
14:53:20buZzhmm, not the same thing
14:53:36amachronici don't see any arm asm there
14:54:57amachronicok, lemme see how you get the build system to show compiler flags
14:55:26amachronicmake V=1
14:55:57amachroniccould you try that and let's see what flags it's using
14:57:51buZzi guess i could compile it by hand with -Wa,-mimplicit-it=thumb and then let make continue?
14:58:36amachronicnothing looks funny to me in the cflags, but I'm not too familiar with ARM assembler
14:58:38buZzdoes make accept it? :P yes
14:59:07buZzhmhm, arm has two different assembly languages, arm and thumb
14:59:18buZzthumb is smaller, but has way less instructions
14:59:23amachronicyeah that much I know
15:00:01buZzit seems gcc8 is just confused about it all
15:00:09amachronicI think we're battling with some GCC default in debian vs. what the RB native ARM toolchains do by default
15:00:12rb-bluebotBuild Server message: Build round completed after 1341 seconds.
15:00:14rb-bluebotBuild Server message: Revision 6beebd75e7 result: All green
15:00:22buZzare you building with custom toolchains?
15:00:26buZzohh,, RB native
15:00:50amachronicwhich are gcc 4.9.4 to boot
15:01:05buZzhehe, well its a old project :P
15:01:19amachronici seem to remember debian supports a couple different ARM flavors
15:01:37buZzok, after compiling lsp and ltp.c by hand with -Wa,-mimplicit-it=thumb , it seems to continue
15:01:40amachronicmaybe a newer ARM subtarget than what rockbox usually deals with?
15:02:05buZzhmhm, well , maybe just a wiki page with 'wanna compile on recentish arm debian? you might need these tiny mods'
15:03:33buZzlibtremor/floor0.c , next one :P
15:04:41amachronicwhat if you use -Wa,-mimplicit-it=always?
15:04:51buZzthen it crashes later on on libmad
15:05:08buZzwhich is assuming implicit-it=arm , i guess
15:05:45amachronicbut =always is supposed to include the functionality of both as far as I can tell...
15:06:58buZzlets see
15:09:50buZz(its not a really fast armv7, but which one is ;) )
15:09:55rb-bluebotBuild Server message: New build round started. Revision f47aa584a8, 303 builds, 9 clients.
15:14:40buZzit just got past libspeex \o
15:15:13buZzand libtremor
15:15:20amachronichaha great let's see if it can pass libmad :)
15:16:23 Join smithjd [0] (
15:18:50buZz-o /home/user/code/rockbox/buildalways/lib/rbcodec/codecs/libmad/layer3.o
15:18:50buZz{standard input}: Assembler messages:
15:18:50buZz{standard input}:1764: Error: shift by register not allowed in thumb mode −− `rsb r0,r0,r0,lsl r3'
15:19:09buZzmaybe i'll try that one with implicit-it=arm
15:19:44buZzeh , same error , *scratches head*
15:20:29 Part nihilazo (Leaving)
15:22:52amachronicthat one comes from MASK() on layer3.c:926
15:23:03amachronicI'm guessing none of the inline asm was designed to be used on thumb
15:23:28buZzi'll remove the arm stuff there, have it just be be the MASK() below it
15:23:37buZz# if defined(CPU_HARM)
15:24:01amachronicwhat you could try is disabling thumb code generation entirely
15:24:13buZzthe upstream libmad might have thumb support though
15:24:53amachronicprobably; the RB libraries have heavily diverged in places
15:25:27amachronici'm pretty skeptical of inline asm in general though
15:25:40buZzyeah agreed
15:27:13 Quit smithjd (Ping timeout: 250 seconds)
15:29:59buZz{standard input}:296: Error: selected processor does not support `rsc r10,r10,#0' in Thumb mode
15:30:19buZzpff, from libmad/synth.c , but not seeing any inline there
15:31:05buZzoh, guess its the FPM_ARM stuff
15:33:26buZzdang, synth.c seems a hard cookie to crack
15:34:09amachronicwell, good luck, I have to step away for now
15:34:20buZz\o thanks for the pointers anyway :)
15:34:22 Quit amachronic (Quit: amachronic)
15:34:32buZzand the hint about gcc4 used by rockbox :P
15:43:17rb-bluebotBuild Server message: Build round completed after 2002 seconds.
15:43:20rb-bluebotBuild Server message: Revision f47aa584a8 result: All green
16:28:33 Join smithjd [0] (
16:39:41 Quit smithjd (Ping timeout: 260 seconds)
16:45:01***Saving seen data "./dancer.seen"
16:50:58 Join smithjd [0] (
17:07:45 Quit smithjd (Ping timeout: 250 seconds)
17:31:26 Quit lebellium (Quit: Leaving)
17:33:25 Join Romster [0] (~romster@
18:45:03***Saving seen data "./dancer.seen"
18:49:30 Quit paulk (Ping timeout: 244 seconds)
18:51:32 Join paulk [0] (
20:09:26 Join smithjd [0] (
20:38:15 Quit smithjd (Ping timeout: 268 seconds)
20:45:07***Saving seen data "./dancer.seen"
21:32:39 Quit akaWolf (Ping timeout: 252 seconds)
21:38:30 Join akaWolf [0] (
22:02:03 Join smithjd [0] (
22:37:57 Quit smithjd (Ping timeout: 250 seconds)
22:45:11***Saving seen data "./dancer.seen"
23:01:06 Quit skipwich (Quit: DISCONNECT)

Previous day | Next day