#milkymist IRC log for Friday, 2011-11-11

wpwrak(unwind=0x6983a0) i'm kinda curious about that myself ;-)00:43
wpwrak(30 min) sorry, took a bit longer00:43
wpwrak(sentinel) of course, 0x82d231c isn't all that much better a value. it's a host address anyway01:09
wpwrakhmm, seems that gdb has a debug option for the DWARF parser, but not for the evaluator. and even the first only operates at the global level. well, i'm quite confident that there's a lot of debug data in that flickernoise executable ...06:06
wpwrakmaybe it's time to unearth umlsim :)06:06
wpwrakgrmbl. the handling of optimized-out expressions is annoying.  p *info -> Cannot access memory at address 0x0 ("ah, it's a NULL pointer"). but then p info -> $1 = <value optimized out>06:32
wpwrakas a small mercy, we don't seem to have to get into execute_cfa_program. that would be fun to debug. (DWARFS has its own "virtual machine" with instruction set. it then has code snippets to calculate all sorts of things from the registers and memory of the debugged program. last time i checked, the GNU tools, gcc, gdb, etc., only used about 1% of the true power that lies there ;-)06:37
wpwrakjust sent further encouragement to faderfox :)08:17
wolfspraulyes seems that exchange is moving now08:28
wpwraklet's hope it goes through before the next set of customs restrictions here. the ones that shut down everything on monday have since been lifted, but they threaten to bring that back08:31
wpwrak"they" = the tax authority. they're working on slapping new controls on customs, basically a triplicate to existing redundant controls. they did something similar to the currency exchange a two weeks ago. see how well it worked: http://fx.sauder.ubc.ca/cgi/fxplot?b=USD&c=ARS&y=daily08:34
wpwrakmwalle: the problem seems to be in the { frame_unwind_register_value (frame=0,regnum=28(sp),...) -> computed bytes=[408d8108] }08:45
wpwrakat get_frame_id (fi=1)08:46
lekernelwpwrak, and as you can see, my MIDI sliders are working correctly (the software still had all the bugs you fixed)09:05
wpwrakhmm, i see people setting up an m1. and when i clock to jump forward, the screen goes black.09:21
wpwrakbut yes, i believe you that things worked for you ;-)09:21
wpwrakhmm, mixing of patches ? doesn't that need a lot more mmeory bandwidth ?09:42
lekernelno no.... alternate on frames09:45
wpwrakaah !09:45
lekernelthat will actually produce funnier results than just mixing the final results09:45
lekernelyeah that flash player is a bit crappy... i've dumped the video and reuploading atm09:46
wpwrakhave you tried it already ? (alternating patches)09:46
lekernelnot precisely, I just sent quick patch switch midi messages09:47
lekerneland the result was interesting09:48
wpwrakhmm, two more parameters: period length and duty cycle ;-)10:01
wpwrakand yes, i can imagine that this should be fun, considering that most patches are highly iterative10:01
lekernelgnome 3 slowness with fedora 16 is amazing. even this IRC text input box has 6 seconds of latency10:01
lekerneli moved away from kde 4 because it had a similar BUG, and now it's coming to gnome too ... :(10:03
wpwraktry fvwm ;-)10:03
lekernelbetter video link: http://vimeo.com/3194846310:58
Fallenouhehe nice sound recording quality, we hear very well what you say12:01
wpwrakaw_: would you see any problems with using a SOT-323 for the AND gate (for the reset circuit) ? it's a little big but availability is great12:02
aw_wpwrak, the SOT-323 is okay though12:02
aw_wpwrak, i remembered you drawed one page with two AND gate sch. Do you have that link?12:03
wpwrak(two) lemme see ..12:04
wpwrakbut that's already obsolete because we now know that we don't need to drive INIT_B12:04
wpwrakthe final set of reworks on rc3 is ... http://en.qi-hardware.com/wiki/File:M1_rc3_hw_fix2.png12:06
aw_yup..yes. no for INIT_B12:06
wpwrak... minus the wire+diode from INIT_B, correct ? or were there any other changes ?12:06
wpwrakin particular, are the resistors as indicated ?12:06
aw_yes. the added diode in red color is not needed. no more other changes. and the values of resistors are all correct in that link12:08
Fallenoutoo bad audio isn't synced with video12:08
Fallenou(in the recording)12:08
wpwrakaw_: perfect, thanks !12:10
aw_wpwrak, so will you change that rst.ps again? I know their wirings that's for rc4.12:12
wpwraki'll make a new file ;-) also with the reset chip update12:13
aw_m...thanks !12:14
wpwrak... during this weekend. it'll be too late for you do order parts today anyway :)12:15
wolfspraullekernel: do you have an original download link for the new video?12:22
lekerneluse the .flv link on vimeo... that's just the justin.tv file cut with mencoder (keeping the original codec and data)12:25
wpwraklekernel: hmm, is there a human-readable description of the lm32 instruction set somewhere ? all i can find is the cryptic http://sourceware.org/cgen/gen-doc/lm32-insn.html17:22
lekernelin lm32_archman.pdf ?17:23
wpwrakah, perfect. thanks ! you're better organized than lattice are :)17:24
kristianpaullekernel: milkymist conbus is ver b3 right?18:45
lekernelwhat do you want to know exactly?19:41
sh4rm4grr why cant stupid vimeo support html5 videos ?20:00
sh4rm4lol at the guy in the audience who means C must be C++20:02
sh4rm4and the other one who mumbles bullshit about classes at the end20:03
kristianpaulwhat types of bus cycles it support?, also, for example when i do a mr 0xaddr and mr 0xaddr what cycle is used in both escenarios?20:06
Fallenoumr and mr ?20:25
lekernelburst or single transfers, depending on whether the address is L1-cacheable or not20:29
kristianpaulArtyom: well come !20:42
kristianpaulFallenou: bios commands20:42
wpwrakmwalle: hmm. something just clicked. you're on 32 bit. i'm on 64 bit.20:58
lars_still talking about gdb backtraces?21:00
wpwrakyes. but i think i just fixed it :)21:01
wpwrakat least in 7.121:02
wpwraklet's see if the same nonsense exists in 7.3.1 ...21:02
wpwraki just wish people would actually take the time to read K&R21:02
lars_they never worked for me either and i'm on 64bit too21:02
wpwrakworks :)21:05
wpwrakposting patch ...21:06
kristianpaullekernel: btw you aware or any free routed on PCB gclk pin with BUFIO2 that can be taken from current M1 board? :-)21:10
kristianpauli reviewed seems, not but asking just in case21:10
kristianpaullekernel: i know M1 is not a devel board, but you know how i'm missing that external clock input those days :)21:11
mwallewpwrak: nice ;)22:52
wpwrakthanks :) does it also work on your 32 bit system ?22:58
mwallelemme test23:21
mwallewpwrak: still works23:27
wpwrakkewl, thanks !23:29
wpwrakah ... and are you my upstream for lm32-related gdb patches ? :)23:29
mwalledunno if its moderated :)23:30
wpwrakokay, mailed23:38
wpwraksweet dreams, untroubled by corrupt frames ! :)23:38
--- Sat Nov 12 201100:00

Generated by irclog2html.py 2.9.2 by Marius Gedminas - find it at mg.pov.lt!