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 2023-02-13

00:24:57 Quit akaWolf (Ping timeout: 252 seconds)
00:26:45 Quit S|h|a|w|n (Remote host closed the connection)
00:28:38***Saving seen data "./dancer.seen"
00:28:46 Join S|h|a|w|n [0] (~shawn156@user/shawn/x-4432647)
00:30:31 Quit S|h|a|w|n (Read error: Connection reset by peer)
00:49:27 Join S|h|a|w|n [0] (~shawn156@user/shawn/x-4432647)
01:00
01:08:05 Join akaWolf [0] (~akaWolf@akawolf.org)
01:14:03 Quit akaWolf (Ping timeout: 248 seconds)
01:15:14 Join akaWolf [0] (~akaWolf@188.243.245.20)
01:23:08 Quit massiveH (Quit: Leaving)
01:58:05 Quit akaWolf (Ping timeout: 252 seconds)
02:00
02:28:42***Saving seen data "./dancer.seen"
02:37:02 Join akaWolf [0] (~akaWolf@akawolf.org)
02:45:15 Quit S|h|a|w|n (Read error: Connection reset by peer)
02:58:17 Join S|h|a|w|n [0] (~shawn156@user/shawn/x-4432647)
03:00
03:20:51 Quit othello7 (Ping timeout: 260 seconds)
04:00
04:00:39 Quit akaWolf (Ping timeout: 260 seconds)
04:01:13 Quit m01 (*.net *.split)
04:01:13 Quit Galois (*.net *.split)
04:01:14 Quit Strife89 (*.net *.split)
04:01:14 Quit bleb (*.net *.split)
04:01:25 Join m01 [0] (~quassel@2001:41d0:304:200::d7f6)
04:01:27 Join Galois [0] (djao@efnet.math.uwaterloo.ca)
04:01:30 Join bleb [0] (~cm@user/bleb)
04:01:53 Join Strife89 [0] (sid399903@id-399903.lymington.irccloud.com)
04:08:24 Quit S|h|a|w|n (Read error: Connection reset by peer)
04:13:51 Join S|h|a|w|n [0] (~shawn156@user/shawn/x-4432647)
04:17:28 Quit S|h|a|w|n (Read error: Connection reset by peer)
04:26:25 Join S|h|a|w|n [0] (~shawn156@user/shawn/x-4432647)
04:28:45***Saving seen data "./dancer.seen"
04:43:14 Join akaWolf [0] (~akaWolf@akawolf.org)
04:50:49 Quit akaWolf (Ping timeout: 260 seconds)
04:52:09 Join akaWolf [0] (~akaWolf@akawolf.org)
04:57:00 Quit akaWolf (Ping timeout: 260 seconds)
04:58:02 Join akaWolf [0] (~akaWolf@akawolf.org)
05:00
05:05:21 Quit akaWolf (Ping timeout: 268 seconds)
05:12:46 Join akaWolf [0] (~akaWolf@akawolf.org)
05:28:23 Quit S|h|a|w|n (Read error: Connection reset by peer)
06:00
06:27:15 Join tchan1 [0] (~tchan@c-98-220-232-221.hsd1.il.comcast.net)
06:28:09 Join Retr0id5 [0] (~Retr0id@user/retr0id)
06:28:47***Saving seen data "./dancer.seen"
06:29:43 Join tertu2 [0] (~tertu@user/tertu)
06:31:50 Join gevaerts_ [0] (~fg@user/gevaerts)
06:32:01 Join larbob_ [0] (~larbob@159.65.42.191)
06:32:39 Join Skyrid3r [0] (~Skyrider@v220221267215211978.supersrv.de)
06:35:00 Quit akaWolf (Ping timeout: 260 seconds)
06:36:10 Join rasher_ [0] (~rasher@diti.rasher.dk)
06:36:28 Quit tchan (*.net *.split)
06:36:28 Quit tertu (*.net *.split)
06:36:28 Quit Retr0id (*.net *.split)
06:36:28 Quit GeekShadow (*.net *.split)
06:36:28 Quit rasher (*.net *.split)
06:36:28 Quit larbob (*.net *.split)
06:36:28 Quit Skyrider (*.net *.split)
06:36:28 Quit dys (*.net *.split)
06:36:28 Quit gevaerts (*.net *.split)
06:36:35 Nick Retr0id5 is now known as Retr0id (~Retr0id@user/retr0id)
06:36:35 Nick Skyrid3r is now known as Skyrider (~Skyrider@v220221267215211978.supersrv.de)
06:36:42 Join GeekShadow [0] (~antoine@82-64-164-139.subs.proxad.net)
07:00
07:32:30 Quit tchan1 (Quit: WeeChat 3.8)
07:34:49 Join tchan [0] (~tchan@98.220.232.221)
07:35:47 Nick gevaerts_ is now known as gevaerts (~fg@user/gevaerts)
08:00
08:28:51***Saving seen data "./dancer.seen"
08:44:44 Nick rasher_ is now known as rasher (~rasher@diti.rasher.dk)
08:44:52 Quit rasher (Changing host)
08:44:52 Join rasher [0] (~rasher@user/rasher)
09:00
09:59:35 Join othello7 [0] (~Thunderbi@pool-100-36-158-132.washdc.fios.verizon.net)
10:00
10:28:52***Saving seen data "./dancer.seen"
10:47:23speachyCarlo30: If you try the oldest daily build we still have on the site, does it solve anything?
12:00
12:28:54***No seen item changed, no save performed.
12:32:05 Join fourHZ [0] (~fourHZ@92-52-40-121.dynamic.orange.sk)
12:57:08 Join lebellium [0] (~lebellium@2a01cb040109a6008019ece5523b554e.ipv6.abo.wanadoo.fr)
13:00
13:19:30 Quit fourHZ (Quit: Client closed)
13:28:51 Join fourHZ [0] (~fourHZ@92-52-40-121.dynamic.orange.sk)
13:37:32 Join S|h|a|w|n [0] (~shawn156@user/shawn/x-4432647)
13:53:40 Quit jacobk (Ping timeout: 252 seconds)
14:00
14:28:55***Saving seen data "./dancer.seen"
14:30:23 Join jacobk [0] (~quassel@64.189.201.150)
15:00
15:29:09 Quit sam_d_ (Ping timeout: 252 seconds)
15:40:13 Join sam_d [0] (~sam@user/sam-d/x-8933526)
16:00
16:28:56***Saving seen data "./dancer.seen"
16:33:36 Quit prg_ (Quit: ZNC 1.8.2 - https://znc.in)
16:36:12 Join prg_ [0] (~prg@2001:19f0:5:9c1:1d2f:7937:5cb0:da6f)
16:50:25 Quit jacobk (Ping timeout: 260 seconds)
17:00
17:51:20 Quit lebellium (Quit: Leaving)
17:59:26 Join Carlo` [0] (~Carlo`@host-79-13-247-155.retail.telecomitalia.it)
17:59:53_bilgusCarlo` are you familar with git bisect?
18:00
18:00:32_bilgusi found this to help another user https://www.git-tower.com/learn/git/faq/git-bisect/
18:01:12_bilgusspecifically Git Bisect in Practice
18:02:35_bilgusstart where you last had a failure and whatever build you were coming from that works and it'll spit you out a few builds or hone it in a bit more and find another newer workin build up to you
18:03:23Carlo`I will do that, I was just trying to reduce the time window between the last known good version and the first bad one
18:03:34Carlo`I can confirm 20230106 has the issue
18:04:36Carlo`Clean install after make zip, changed the theme to Informative Beauty, repeat -> all and started cycling through a single directory
18:06:02_bilgustruly you should also do a make clean and typically I omit plugins in /tools/configure
18:06:07Carlo`https://git.rockbox.org/cgit/rockbox.git/commit/?id=5d0c382a59d84208be5ed719b510794f465ccd8b
18:06:12_bilgusthat way it speeds it up
18:06:21Carlo`This looks definitely suspicious
18:07:20_bilgusthe clip+ is monochrome
18:07:38_bilgusthats why git bisect is so helpful
18:08:04Carlo`Okay, will try bisect
18:11:23Carlo`rb@debian:~/rockbox/rockbox$ git bisect bad  5d0c382a59d84208be5ed719b510794f465ccd8b
18:11:23Carlo`rb@debian:~/rockbox/rockbox$ git bisect good 1773e56447785931968d46002b73307c1fb3a35a
18:11:24Carlo`Bisecting: 10 revisions left to test after this (roughly 4 steps)
18:11:38Carlo`bad is that commit I've linked, good is the last one from 20221230
18:12:19Carlo`Then i should simply build as usual?
18:13:07_bilgusyep then install and try
18:13:24Carlo`Roger that, doing it right now
18:13:26_bilgusif it works type good if it doesn't type bad
18:13:50_bilgusI should say 'git bisect good'
18:16:36_bilgusmake clean is a v. good idea but its probably close enough that its not necessary
18:16:45Carlo`Would copying my config.cfg from a previous build be of any harm?
18:17:01_bilguspossibly
18:17:38_bilgusI always start with a clean settings file but if you keep them the same between it should still work out fine
18:18:25_bilgusespecially with such a short window but if you need to do that try to keep it the same through the git bisect process wo sav hair pulling
18:18:26Carlo`Not a problem, I'm only changing theme and repeat, meanwhile I'm testing on another Clip+ with the cabbiev2 theme and the issues doesn't seem to be crop up with that
18:18:47_bilgusthere are some hardware differences with the clip's
18:19:15_bilgusand voltage issues between them with the CPU clocking
18:19:37_bilgusif it turns into a heisenbug it usually leads t that
18:20:01_bilguslater revisions handle lower CPU voltages it appears
18:20:08Carlo`I have seven Clip+, any way to tell them apart?
18:20:38_bilgusnot taht I've found I bet theres a serial number somewhere
18:20:50_bilgusbut hardware wise no
18:21:49Carlo`I haven't installed fonts this time, just the IB theme
18:23:05_bilgusas long as you can reproduce it then whatever you need but minimizing the amount of variables will save you a headache later
18:23:16Carlo`I'd hate to have a false negative mostly
18:23:48_bilgus4 steps is not much when its 50 it matters more
18:24:22_bilgusthats when I disable plugins or limit it to a subset of them
18:24:46Carlo`So, I'll reproduce the same condition as the last crash: clean install, repeat -> all, install fonts and IB
18:25:35Carlo`I'll test for another couple of minutes and then I'll tag this build as good and move on
18:25:51_bilgusonce we have a build I can then go look and we start adding or removing a variable at a time
18:26:23Carlo`Thank you so much for your time
18:27:04_bilgussome bugs jump right out others are subtle/insidious
18:27:15_bilgushey thank you for finding our bugs
18:27:49_bilgusblame will be applied later :p
18:27:50Carlo`Well, it's the least I can do, as I quite enjoy Rockbox on my devices
18:28:08Carlo`okay, this build is good, onto the next
18:29:00***Saving seen data "./dancer.seen"
18:29:10 Join jacobk [0] (~quassel@utdpat241055.utdallas.edu)
18:31:31Carlo`Your build with the cabbiev2 theme isn't affected by the bug
18:34:17 Quit jacobk (Ping timeout: 255 seconds)
18:36:33Carlo`I have two unusual characters in my tags, Ö and ø
18:37:54Carlo`however the bug appeared even while skipping files without strange characters in a different directory
18:38:59_bilgusI can try the theme what fonts are you using?
18:39:21Carlo`I've just installed the standard  font package with the rb utility
18:39:35Carlo`I haven't changed the default IB fonts
18:40:23Carlo`this build looks good too, let me compile the next one
18:43:21 Join massiveH [0] (~massiveH@2600:4040:a992:a300:5876:20dc:6e8b:80ec)
18:44:08_bilgusare you setting a particular font?
18:44:27Carlo`No
18:44:55_bilgusthen that shouldn't matter unless the theme is using one of em
18:44:56Carlo`The default Informative Beauty font is Rockfont 8px
18:48:01Carlo`hm.. three good builds in a row, I was hoping for a bad  one
18:50:52Carlo`Bisecting: 0 revisions left to test after this (roughly 1 step)
18:50:53Carlo`[b444ecfca2218068233a89c278cc86b9e9ef67d7] plugins: Properties: Add 'Last Modified' for audio files
18:52:19Carlo`this is going to be the final showdown
18:52:34Carlo`.. unless I have mis-tagged a build as good
18:58:50Carlo`Nope, this build doesn't seem to have the bug
18:59:31Carlo`Let me just try the next commit, the "Rewrite 16bit alpha bitmap blit" one
19:00
19:01:14Carlo`I'll just do a bisect reset to that commit
19:02:39Carlo`"plugins: Properties: Add 'Last Modified' for audio files" from 20230106 doesn't show any crashes
19:18:48_bilgus i tried the theme you referenced for a bit like 20 songs or so what about 15 20 mins and so far can't repro it i skipped fwd reverse fast forwarded rewound etc and so far no luck (unluck?)
19:20:03_bilgusif you try a few times and still no dice I can make you a build at max cpu voltages
19:20:03Carlo`This is truly a heisenbug, known bad builds aren't crashing anymore
19:21:09_bilguscan be your cfg file too I had one a week ago that we searched thru for a good hour and copying the settings file off and back on disappeared
19:22:14Carlo`Let me check one thing
19:23:16_bilgusbut likely you just need to find that bad variable thats why minimal reproducible case saves you a headache later
19:25:38Carlo`https://pastebin.com/pEwdeQx8
19:26:20Carlo`This may be just a coincidence, but with that config file 20230106 crashes, without it doesn't (or at least it doesn't seem to crash)
19:35:25_bilguswe did change the dsp stuff but it crashes with it the try it when you bisect
19:36:20_bilgushaas surround got a bug fixed if you look it the file itll say whats changed from default
19:36:35_bilgus(re settings)
19:37:32Carlo`It's definitely one of those settings, as the build you gave me earlier doesn't crash with a default config and the IB theme
19:39:27Carlo`"Haas Surround add flag so the settings can be read back from the config file"? That's from 20221212, but 20221230 works fine with my config
19:39:52Carlo`20221230 is the last known good configuration that I'm absolutely positive doesn't have the bug
19:40:01Carlo`*version I mean
19:41:25_bilguswell veify that it works through gitbisect and it might appear otherwise it could be a bad setting or a change to the settings format that is bad or even a bad spot in your disk which I my guess on my bug
19:41:39_bilgusis my guess*
19:42:01_bilgusi mean how else does copying and deleting and recopying fix a bug
19:43:04Carlo`I'll use another Clip+ for testing from now on
19:43:47Carlo`anyway, I did use my daily driver Clip+ for a couple of hours last week, updated to the 20230208 build, and the crashes started to appear immediately after that
19:44:38Carlo`Compiled 20221230 and it worked fine for another couple of hours, installed 20230115 and immediately got the crashes again
19:46:28_bilgusok so can't reproduce them now at all or is it still specific builds?
19:46:48Carlo`Your daily with the .map file you gave me earlier doesn't seem to crash with a stock config and the IB theme, but the crashes reappear as soon as I overwrite the stock config with the one I've pasted
19:47:27Carlo`Maybe I did tag as good a bad release just because the bug didn't trigger somehow
19:47:41_bilgusits always possible
19:47:42Carlo`I'll continue a bit more, then I'll get back at it tomorrow afternoon
19:48:14_bilgusill keep trying on this clip+ did you say you pasted the cfg file somewhere?
19:48:22Carlo`https://pastebin.com/pEwdeQx8
19:49:00_bilgusthe compressor is dealing with the dsp system
19:49:27_bilgusthere was a user having issues with it let me build you this build and try it with your current settings
19:50:44Carlo`Sure
19:50:56_bilgusoh nevermind it appears amachronic already merged it
19:51:45_bilgushttps://gerrit.rockbox.org/r/c/rockbox/+/4912 on 12-18
19:52:05_bilgusanyway ill apply those settings on mine
19:52:26Carlo`https://git.rockbox.org/cgit/rockbox.git/commit/?id=541960a11061213f682d67bb036f861b2605f9d3
19:53:39Carlo`Could that be the issue, finally?
19:55:17 Join jacobk [0] (~quassel@129.110.242.224)
19:55:51Carlo`"there is a race condition here that causes prefetch abort panics on the Clip+.", this sounds exactly like what I'm experiencing
19:56:58_bilgusmaybe it only looked fixed then
19:57:23_bilguswe were doing a dummy loop prior let me check what changed
19:57:53Carlo`It's triggered randomly, by repeteadly skipping songs, and it only happens as soon as a given file starts playing
19:59:40Carlo`I believed fast forwarding also triggered it, but I may simply have tapped forward instead of keeping it pressed
20:00
20:00:29_bilgusthis was done on Dec 18th https://gerrit.rockbox.org/r/c/rockbox/+/4915 didnt you say you were using 12/30 build fine?
20:00:50Carlo`Yes, I've used the 20221230 build for a couple of hours and the bug never showed once
20:02:53Carlo`However, given how elusive it is, I could keep using that build for a couple of days just to be sure it's bug free
20:03:51Carlo`It's quite random, however nothing crashed as fast as your 202302013 build you gave me earlier
20:07:10 Quit fourHZ (Quit: Client closed)
20:09:18_bilgusthats the build at HED
20:09:24_bilgusHEAD*
20:09:54Carlo`I know, it was just to say that the latest build crashes almost instantly
20:29:04***Saving seen data "./dancer.seen"
21:00
21:04:05 Quit Carlo` (Quit: Connection closed)
21:29:07 Quit massiveH (Quit: Leaving)
21:53:11 Quit jacobk (Ping timeout: 255 seconds)
22:00
22:07:36 Join jacobk [0] (~quassel@64.189.201.150)
22:13:16 Join amiconn_ [0] (jens@p200300ea870dd400305e95fffec66ff3.dip0.t-ipconnect.de)
22:13:16 Quit amiconn (Killed (calcium.libera.chat (Nickname regained by services)))
22:13:16 Nick amiconn_ is now known as amiconn (jens@p200300ea870dd400305e95fffec66ff3.dip0.t-ipconnect.de)
22:14:13 Join pixelma_ [0] (marianne@p200300ea870dd400305e95fffec66ff3.dip0.t-ipconnect.de)
22:14:13 Quit pixelma (Killed (calcium.libera.chat (Nickname regained by services)))
22:14:13 Nick pixelma_ is now known as pixelma (marianne@p200300ea870dd400305e95fffec66ff3.dip0.t-ipconnect.de)
22:29:06***Saving seen data "./dancer.seen"
23:00
23:37:11 Quit m01 (Quit: Konversation terminated.)
23:39:30 Join m01 [0] (~quassel@vps-b172b88b.vps.ovh.net)

Previous day | Next day