--- Log for 09.03.122  Server: molybdenum.libera.chat  Channel: #rockbox
--- Nick: rb-logbot  Version: Dancer V4.16  Started: 1 month and 0 days ago

00.25.24 Quit    archpc (Ping timeout: 240 seconds)
00.25.43 Join    archpc [0] (~archpc@user/archpc)
00.40.47 Join    archpc5 [0] (~archpc@user/archpc)
00.41.31 Quit    archpc (Ping timeout: 272 seconds)
00.41.31 Nick    archpc5 is now known as archpc (~archpc@user/archpc)
00.54.15 ***     Saving seen data "./dancer.seen"
01.12.57 #       <rb-bluebot> Build Server message: 3New build round started. Revision 01e76548c3, 303 builds, 7 clients.
01.38.17 #       <spork> gerrit probably got overwhelmed by all of amachronic's patches
01.39.37 #       <rb-bluebot> Build Server message: 3Build round completed after 1599 seconds.
01.39.40 #       <rb-bluebot> Build Server message: 3Revision 01e76548c3 result: All green
02.01.40 Quit    braewoods (Quit: WeeChat 2.8)
02.05.46 Join    braewoods [0] (~braewoods@user/braewoods)
02.54.18 ***     Saving seen data "./dancer.seen"
03.18.36 #       <_bilgus_> __aeabi_idiv is related to the ARM  crash
03.30.03 Quit    Natch (Ping timeout: 256 seconds)
03.34.25 Join    Natch [0] (~natch@c-67bae255.014-297-73746f25.bbcust.telenor.se)
04.25.00 Quit    Malinux (Ping timeout: 240 seconds)
04.30.43 Join    Malinux [0] (~malin@2001:4641:4dfa::12c:c4a7)
04.32.57 Quit    Natch (Ping timeout: 240 seconds)
04.43.25 Quit    melmothX (Quit: WeeChat 3.0)
04.48.38 Join    Natch [0] (~natch@c-67bae255.014-297-73746f25.bbcust.telenor.se)
04.54.20 ***     Saving seen data "./dancer.seen"
05.06.13 Join    melmothX [0] (~marco@amusewiki/marco)
06.46.18 Quit    Romster (Ping timeout: 272 seconds)
06.48.18 Join    Romster [0] (~romster@user/romster)
06.54.24 ***     Saving seen data "./dancer.seen"
08.02.31 #       <braewoods> _bilgus_: division by zero?
08.02.44 #       <braewoods> that's the only thing i can think of
08.03.04 #       <braewoods> assuming it was division error and not another logic error
08.35.41 #       <_bilgus_> braewoods, its a hand tuned asm loop to generate reciprocal and use multiplication to divide
08.36.01 #       <_bilgus_> https://github.com/Rockbox/rockbox/blob/master/lib/arm_support/support-arm.S
08.37.55 #       <_bilgus_> dumping the the gcc version in there fixes it but even without the test for 0 its so much slower I'd prefer the crashing
08.40.35 #       <_bilgus_> i'm guessing there is some assumption that Os breaks for the code, and I was able to work around it by doing the division at compile time of usec/tick but now I think it might be the issue with fuze V2? V1?
08.54.28 ***     No seen item changed, no save performed.
08.57.40 Join    massiveH [0] (~massiveH@ool-4a5862ee.dyn.optonline.net)
09.25.15 #       <braewoods> _bilgus_: oh, that old trick for dividing by a constant integer? thought that should never trigger a division by zero.
09.26.06 #       <_bilgus_> there is no div by 0
09.28.08 #       <_bilgus_> its crashing on a data abort after Os
09.30.38 #       <_bilgus_> AFAIAA optimizations shouldn't affect ASM so its probably something with a bad (disallowed) register is current my guess i'll have to wade through the generated ASM
10.36.40 Quit    massiveH (Quit: Leaving)
10.44.54 #       <_bilgus_> forcing unsigned fixes it so I guess that gives me some hope
10.54.29 ***     Saving seen data "./dancer.seen"
12.54.32 ***     No seen item changed, no save performed.
12.57.19 Join    ZincAlloy [0] (~Adium@2a02:8108:943f:d824:d5c8:589a:7346:79f7)
13.02.48 #       <rb-bluebot> Build Server message: 3New build round started. Revision 57cffae369, 303 builds, 9 clients.
13.17.19 #       <rb-bluebot> Build Server message: 3Build round completed after 871 seconds.
13.17.23 #       <rb-bluebot> Build Server message: 3Revision 57cffae369 result: 1 errors 0 warnings
13.33.13 Join    lebellium [0] (~lebellium@2a01cb04012c090095e088e99ff73ea7.ipv6.abo.wanadoo.fr)
14.54.33 ***     No seen item changed, no save performed.
15.03.54 Quit    munkis (Remote host closed the connection)
15.04.18 Join    munkis [0] (~mendel_mu@ool-ae2cb229.dyn.optonline.net)
16.18.50 Quit    akaWolf (Read error: Connection reset by peer)
16.19.05 Join    akaWolf [0] (~akaWolf@akawolf.org)
16.19.10 Quit    ZincAlloy (Quit: Leaving.)
16.39.25 Quit    akaWolf (Ping timeout: 256 seconds)
16.54.00 Join    akaWolf [0] (~akaWolf@akawolf.org)
16.54.35 ***     Saving seen data "./dancer.seen"
16.59.29 #       <_bilgus_> hmm upon further experimenting it seems to be due to the interrupt, wonder if there is an op that doesn't like interrupts
17.01.33 #       <_bilgus_> reproducer: https://pastebin.com/qTacgFrw
17.02.29 #       <_bilgus_> I'm guessing this is all arm v6 not sure about the lesser versions
17.33.56 Quit    lebellium (Quit: Leaving)
18.54.38 ***     Saving seen data "./dancer.seen"
19.35.32 Join    archpc2 [0] (~archpc@user/archpc)
19.37.55 Quit    archpc (Ping timeout: 256 seconds)
19.37.55 Nick    archpc2 is now known as archpc (~archpc@user/archpc)
20.54.40 ***     Saving seen data "./dancer.seen"
22.02.15 #       <_bilgus_> Well I've figured it out I just don't know the 'WHY'
22.30.21 #       <_bilgus_> well still don't know the WHY but the equivalent instructions work great :) so fixed.
22.32.57 #       <_bilgus_> and spoke too soon :/
22.54.43 ***     No seen item changed, no save performed.