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 2019-07-18

00:28:47__builtincould I get a quick code review? assembly: original C:
00:29:47__builtinthe C code compiled to 31 instructions/loop with -O3, and the assembly version uses just 13
00:35:21***Saving seen data "./dancer.seen"
00:39:59 Join krabador [0] (~krabador@unaffiliated/krabador)
00:49:46 Quit rasher (Quit: buster)
00:57:21 Join rasher [0] (~rasher@rockbox/developer/rasher)
01:09:34 Quit rasher (Quit: buster 2: electric boogaloo)
01:22:50 Join rasher [0] (~rasher@rockbox/developer/rasher)
01:30:46 Quit ZincAlloy (Quit: Leaving.)
01:55:10 Join blubbblubb [0] (
01:57:11 Quit blubbblubb (Remote host closed the connection)
01:57:30 Quit foolsh (Ping timeout: 245 seconds)
02:05:43 Quit MrZeus (Ping timeout: 252 seconds)
02:35:24***Saving seen data "./dancer.seen"
02:42:39 Join tchan1 [0] (
02:42:50 Quit tchan1 (Client Quit)
02:43:12 Join tchan1 [0] (
02:43:24 Quit tchan1 (Client Quit)
02:43:42 Join tchan1 [0] (
02:43:54 Quit tchan1 (Client Quit)
02:44:59 Quit tchan (Ping timeout: 245 seconds)
02:47:57 Join tchan [0] (~tchan@lunar-linux/developer/tchan)
02:49:00 Quit krabador (Remote host closed the connection)
04:21:18__builtindoes someone have a gigabeast to test quake?
04:35:27***Saving seen data "./dancer.seen"
04:44:14 Nick IncognitoMan_ is now known as IncognitoMan (~Incognito@unaffiliated/incognitoman)
05:05:27 Quit lagfra (Ping timeout: 246 seconds)
05:07:33 Join lagfra [0] (
05:15:13 Quit TheSeven (Disconnected by services)
05:15:21 Join [7] [0] (~quassel@rockbox/developer/TheSeven)
06:35:31***Saving seen data "./dancer.seen"
07:10:58 Quit JanC (Read error: Connection reset by peer)
07:10:59 Join JanC_ [0] (~janc@lugwv/member/JanC)
07:12:33 Nick JanC_ is now known as JanC (~janc@lugwv/member/JanC)
07:20:51 Quit dys (Ping timeout: 272 seconds)
08:35:34***Saving seen data "./dancer.seen"
08:46:15 Quit [7] (Ping timeout: 264 seconds)
08:46:36 Join ZincAlloy [0] (~Adium@2a02:8108:9440:dfc:719a:509b:a51:16c8)
08:47:39 Join TheSeven [0] (~quassel@rockbox/developer/TheSeven)
09:11:31 Join TheSphinX_ [0] (
09:13:51 Quit TheSphinX^ (Ping timeout: 264 seconds)
09:13:51 Nick TheSphinX_ is now known as TheSphinX^ (
09:30:40 Quit JanC (Remote host closed the connection)
09:31:01 Join JanC [0] (~janc@lugwv/member/JanC)
10:26:06 Quit yosafbridge (Quit: Leaving)
10:30:34 Quit Huntereb (Ping timeout: 258 seconds)
10:30:48 Join yosafbridge [0] (
10:35:35***Saving seen data "./dancer.seen"
10:54:32 Join Huntereb [0] (
11:01:41 Join dys [0] (~dys@
11:13:42 Join deevious [0] (~Thunderbi@
12:35:36***No seen item changed, no save performed.
12:37:18 Join vmx [0] (
13:36:32 Join dandels [0] (~dandels@unaffiliated/dandels)
13:50:52 Quit dandels (Ping timeout: 245 seconds)
13:51:20 Join krabador [0] (~krabador@unaffiliated/krabador)
14:19:08 Join foolsh [0] (~quassel@
14:19:22 Join massiveH [0] (
14:35:38***Saving seen data "./dancer.seen"
14:56:02 Join pamaury [0] (~pamaury@rockbox/developer/pamaury)
14:57:59 Quit tchan (Ping timeout: 246 seconds)
14:59:02 Join tchan [0] (~tchan@lunar-linux/developer/tchan)
15:05:06 Join tchan1 [0] (
15:05:16 Quit tchan1 (Client Quit)
15:07:12 Quit tchan (Ping timeout: 248 seconds)
15:34:54 Quit massiveH (Quit: Leaving)
15:41:15 Join speachy [0] (40eebded@
15:41:38fs-bluebotBuild Server message: New build round started. Revision c9aeb42, 280 builds, 9 clients.
15:56:01 Quit dys (Ping timeout: 258 seconds)
15:58:10speachyhmm. we ought to update the main web page to include the AGPTek Rocker/Benjie T6 and the Xduoo X3 as unstable targets.
16:01:30speachyhuh. apparently I did that a while back. or at least I committed it.
16:02:42speachyI guess there's another step needed to publish the changes
16:03:39 Quit ender| (Ping timeout: 276 seconds)
16:05:31 Join tchan [0] (
16:05:31 Quit tchan (Changing host)
16:05:31 Join tchan [0] (~tchan@lunar-linux/developer/tchan)
16:09:10 Quit michaelni (Ping timeout: 245 seconds)
16:16:20 Quit krabador (Remote host closed the connection)
16:16:53 Join ender| [0] (~ender1@2a01:260:4094:1:225:90ff:fe2a:83dc)
16:21:32 Quit ender| (Ping timeout: 252 seconds)
16:23:02 Join michaelni [0] (
16:25:25 Quit vmx (Ping timeout: 245 seconds)
16:31:31 Quit foolsh (Read error: Connection reset by peer)
16:34:38 Join foolsh [0] (~quassel@
16:35:42***Saving seen data "./dancer.seen"
16:40:27 Join krabador [0] (~krabador@unaffiliated/krabador)
16:50:07 Quit krabador (Remote host closed the connection)
16:57:16__builtinspeachy: yeah, you'll need to contact one of the site admins
17:19:48 Join ender| [0] (~ender1@2a01:260:4094:1:225:90ff:fe2a:83dc)
17:35:18 Quit ender| (Ping timeout: 276 seconds)
17:47:55fs-bluebotBuild Server message: New build round started. Revision d5908f5, 280 builds, 9 clients.
17:48:33 Join ender| [0] (~ender1@2a01:260:4094:1:225:90ff:fe2a:83dc)
18:35:44***Saving seen data "./dancer.seen"
19:14:23 Join vmx [0] (
19:28:53 Quit pamaury (Quit: Konversation terminated!)
19:36:06 Quit beencubed (Quit: Leaving)
19:50:35foolshI can't find anything in the git logs, but was something changed in the configure script, I get a lot of warnings now. Mostly about __builtin's sdl library and games wolf3d and duke3d
19:51:16foolshit was compiling fine without warnings this morning until I did a git pull
19:57:09 Join dys [0] (
20:03:44 Join beencubed [0] (~beencubed@
20:04:16__builtinI may have fudged the flags
20:05:26__builtinfoolsh: what exactly are the warnings?
20:10:32foolshHang on I'll pastebin the output
20:11:04foolshrunning a clean build to make doubly sure it's not just my env
20:25:55speachynone of my targets build the SDL stuff, but one has to go back a ways before anything changed in the configure script or build system.
20:35:46***Saving seen data "./dancer.seen"
20:39:07foolshYeah I think was just old cruft, I swear I cleaned ccache before I built last time. It was doing it on my 64bit machine no matter what, thought my build env got boogered up, rebuilt the arm crosscompiler and now everythings peachy
20:40:10foolsh__builtin: no worries its all working now
20:44:23foolshgreat now it's doning it it again, after running "ccache -c" and rm'ing my build directory, WTF
20:45:16__builtinfoolsh: can you pastebin the output?
20:45:45 Join ulmutul-guest [0] (
20:48:32ulmutul-guest__builtin: not an ARM expert at all, but didn't you miss a "mov r5, r5, asl #16" after your "mul r5, r4, r1"? (Line 31)
20:49:06__builtinulmutul-guest: nice catch, but I already just fixed that :)
20:49:39__builtinlet me show you my latest version
20:49:59__builtinit appears to be causing sporadic crashes, but I don't know why
20:50:17__builtinit works fine usually, but very rarely it causes an abort in unrelated code
20:50:32__builtinI suspect I'm messing up the calling convention somehow
20:53:18__builtin(by the way, the whole shift-by-16 nonsense is to work with QADD to provide saturating addition)
20:53:38__builtinQADD only handles 32-bit values, so I shift the 16-bit samples to the high half and add them that way
20:54:08ulmutul-guestIs there any reliable calling convention if the code is inlined?
20:54:30__builtinI don't think it's inlining though
20:54:54foolsh__builtin: I purged ccahe something fishy going on with it, I had it plugged in to icecc for distrubuted builds, something some where went wonky
20:56:12foolshI'll dive into wolf3d a little bit, and duke3d fails at writing the config file somewhere on target clip zip
20:57:18*foolsh wonders if his ssd is failing and causing errors
21:05:01 Join Rower [0] (
21:13:29 Quit speachy (Remote host closed the connection)
21:37:12 Quit Rower (Quit: Hmmm...)
21:52:40fs-bluebotBuild Server message: New build round started. Revision 8bd992c, 280 builds, 9 clients.
22:00:21 Join lebellium [0] (
22:06:56 Quit pR0Ps (Ping timeout: 248 seconds)
22:07:03 Join MrZeus [0] (~MrZeus@2a02:c7f:70d4:cd00:89a3:eff3:40d7:2991)
22:11:51 Quit vmx (Remote host closed the connection)
22:23:40foolsh__builtin: wolf3d freezes at "rockbox/apps/plugins/sdl/src/file/SDL_rwops.c" on line "SDL_RWread(src, &value, (sizeof value), 1);" in function "Uint32 SDL_ReadLE32 (SDL_RWops *src)"
22:24:39__builtinthis is clip zip?
22:26:50__builtinI can almost guarantee that it's reading from an unaligned address
22:27:57 Join pR0Ps [0] (
22:28:35 Part ulmutul-guest
22:30:05__builtinwell, that wouldn't make sense...
22:31:29__builtinfoolsh: could you try a build without wolf3d's SOUND_ENABLE?
22:31:35__builtinuncomment the #define in wl_def.h
22:31:43__builtiner.. comment it out
22:35:48***Saving seen data "./dancer.seen"
22:39:07__builtinhmm... I tried out my ipod6.0G and the lcd flashes are completely gone...
22:44:00 Quit lebellium (Quit: Leaving)
22:47:54foolsh__builtin: Yeah wolf3d runs beautifully with sound disabled
22:49:05__builtinthat's really odd... could you trace the crash further down the call chain??
22:50:56__builtinI took a quick look, and it seems that the memory accesses are through SDL_memcpy
22:51:19__builtinwhich only deals with char* pointers
22:55:16__builtinugh, it looks like it's optimized into a mess by gcc
22:59:20foolshTo sum it up: Use following directives for GCC “-mshort-load-bytes” and “-mno-short-load-words”
22:59:58foolshI'll pop some flags into the Makefile
23:00:20__builtinyou'll probably want SDLFLAGS sdl.make
23:09:25__builtinulmutul (logs): I think I should mention that assembly runs in an IRQ context as well
23:32:05foolsh__builtin: Geez no of these are present in our compiler "-mword-relocations" "-mno-alignment-traps" "-mno-short-load-words" "-mno-short-load-bytes"
23:32:46foolshI tried "-mword-relocations" not know exactly what it does, but it still freezes
23:42:46__builtinI'm not surprised. 4.4.4 is ancient
23:57:54 Quit ZincAlloy (Quit: Leaving.)

Previous day | Next day