#milkymist IRC log for Monday, 2011-01-31

lekernelhttp://www.linux-magazin.de/NEWS/VJ-System-Milkymist-als-Entwicklerboard-erhaeltlich04:33
aerisadam a re├žu le bidule ?04:34
lekernelah, je ne sais pas, je demande04:41
lekerneltu vas au fosdem au fait?04:41
aerisUi04:41
rohhackabledevices down?04:47
wolfspraullekernel: nice, congrats for the article!05:00
wolfspraulha, Anika Kehrer again - she's cool!05:01
rohhm. found a bug. it was 27c3 not 26c3 ;)05:58
lekernelno, i did present it at 26c305:59
kristianpauland booted linux at that time ;-)06:00
rejonlekernel where you staying at for fosdem?06:00
lekernelat a hotel06:01
rejonwhich?06:02
rejonlooking for pointers06:02
rejonb4 i get ripped off06:02
rohlekernel: ah.. ok. so you have 2 congresses on it now ;)06:07
rohgood things need time06:08
lekernelwell, only 26c3 was successful. they really fucked up the workshops at 2706:08
lekernelrejon: hotel bloom, rue royale06:13
wolfspraullekernel: btw Adam is working on the power-up bug06:21
wolfspraulhe went to xilinx last week and spent 2.5h there, discussing the thing. he has some new ideas he is trying now (just in case you don't already know all this)06:21
lekernelhe found it, didn't he?06:22
lekernelfrom my understanding, it's a race condition between fpga and flash startup, and for some reason (Adam didn't tell me why) the capacitor I had planned to solve this potential problem didn't work as expected06:23
rejonlekernel cool06:27
wolfspraullekernel: I don't know, I am waiting.06:28
wolfspraulI hate to cause stress on people.06:29
wolfspraullet him do his proper work until he is happy :-)06:29
wolfspraulif he found it - great!06:29
kristianpaulhttp://www.eetimes.com/electronics-news/4212668/Xilinx-buys-high-level-synthesis-EDA-vendor09:30
mwallehi15:48
kristianpaulhola16:02
mumptaihi16:40
mumptaithe fpga is the only device on the jtag chain on the mm1?17:00
mwallemumptai: yes17:00
kristianpaulyes17:00
mumptaiokay, then i got to read more, thx17:01
mumptaidid the urjtag systax hange lately?17:10
mumptaii'm getting a syntax error with that fjmem batch17:11
mwallei dont think so17:11
mwallewhat error17:11
mwalle?17:11
mumptaithis one: "Error: part.c:355 urj_part_instruction_define() invalid parameter: invalid instruction length"17:11
mumptaiwhich is the response to: "instruction CFG_OUT  000100 BYPASS"17:13
mwalledid the detect return successfully?17:13
mumptaiit detects the prom an the spartan17:14
mwallePROM?17:14
mwallewhat board17:14
mumptaiand i select the later one with "part 1"17:14
mumptainexys217:15
mwalleunfortunately, i dont have a working urjtag right now, but i tried it with my spartan 3e eval board17:20
mwalleare you sure you've chosen the right part?17:20
mwallethere is a command which should display the instruction length17:21
mumptaidetect does17:22
mumptaiIR length is 1417:22
mwallethe complete chain?17:26
mumptaiyes17:28
mwalletry print17:29
mumptaiit says "non" multiple times17:30
mumptais/non/none/17:30
mwallewhere?17:30
mwallecould you paste it somewhre17:31
kristianpaulmumptai: did you : instruction CFG_OUT  000100 BYPASS17:31
kristianpaulinstruction CFG_IN   000101 BYPASS17:31
kristianpaulbefore pld load right?17:31
mumptaithere is one row17:31
mumptailabled as No. 117:31
mumptaino mfg., no part17:32
mumptaiand Stepping instr. and register are "(none)"17:32
mumptaikristianpaul, yes17:33
mwalleand detect detect, detects a xilinx device?17:33
mumptaiand says that is can't find info about the stepping17:34
kristianpaulpaste i tsomwhere17:34
mumptaibut it decodes the IDs correctly17:34
mwalleah ok, so you have to add the stepping in data/xilinx/<part>/STEPPINGS17:35
mwalledid it ever work?17:35
mumptaiyes17:36
mumptaiwait17:36
mumptaibut that wasn't in the interactive mode17:36
mwallewith exactly that fpga?17:38
kristianpaulmumptai: what jtag cable are you using btw?17:38
mumptaiaww17:39
mwallei guess you are using the onboard 'cable' of the nexys2?17:41
mumptaiworks better with proper scan files ;)17:41
mwallebut why have it worked before?17:42
kristianpaulmumptai: [OT] are you aware if there is support for Bus pirate from urjtag project?17:42
mumptaikristianpaul, no wasn't. but i can actually load a bit-file into the fpga using urjtag, so it can't be that broken ;)17:43
kristianpaul:)17:43
mwallekristianpaul: have you tried the bus pirate with urjtag?17:44
mwallelast time i tried it was very slow17:44
mumptainow i get this: http://codepad.org/pdO5778417:45
kristianpaulmwalle: but, but i willing to do so with my avnet board17:45
kristianpauls/but/no17:45
kristianpaultry  frequency 12000000 ??17:46
mumptaithe tck rate is actually hard-coded into the cable's controller code17:48
kristianpaulhmm17:48
mumptaiits a cypress fx2 running some altera usb-blaster emulation17:50
mumptaimaybe i messed up the ucf, or the flash isn't yet supported17:52
mumptaiwill double check that when i find another spare evening17:53
mwallemumptai: i guess your are too fast for the flash detection17:54
mumptaimhmmm17:54
mwalleyou get the same error with the FT2232H if you dont set the frequency to 6MHz17:54
mumptaiits not really running at 12MHz, probably not even above 2MHz17:55
kristianpauldid flash memory mind about it? it seems..17:55
mumptaithat is a 8051 core doing some bit-banging ...17:55
kristianpaulis not that worst? :) (bit bang?.. above 2Mhz..)17:56
kristianpaulmumptai: can you confirm with flash memory data sheet?17:56
mumptaii can, but i got to get up in less than 7 hours17:57
mumptaiso, good night17:59
mumptaiand thanks for you help17:59
kristianpaulsure18:00
kristianpaulhey 7 hrs still un fair !18:00
kristianpaulsleep from 5 to 6 hrs a day is posible :-)18:01
mwalleim going to bed too18:01
mwallegn8 :)18:01
--- Tue Feb 1 201100:00

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