02:00
sb0 has quit [Remote host closed the connection]
02:23
sb0 has joined #m-labs
02:24
<
sb0 >
rjo, set a watchdog and the master will terminate the worker
03:06
<
rjo >
then the keepalive stuff is only used to evict hung/crashed workers.
03:07
<
rjo >
(from the core device side)
03:15
<
sb0 >
disconnected network cables, OS crashes, etc.
04:04
<
sb0 >
the synthnv is remarkably buggy
04:05
<
sb0 >
connecting it via a USB hub makes its protocol timeout due to the increased USB latency
04:05
<
sb0 >
connecting it directly exceeds the USB power budget of some ports
04:10
<
sb0 >
the result of the protocol timeout is, of course, incorrect values being programmed for obscure reasons
04:15
<
rjo >
what is the usb-to-spi chip in there?
04:15
<
sb0 >
the protocol I'm talking about is what goes on top of serial
04:16
<
sb0 >
and delimitates each command
04:16
<
rjo >
sure. but do you know what the chip is?
04:18
<
sb0 >
I don't know... and is it really usb-to-spi? it's detected as a ACM device on Linux, and the commands are ASCII based
04:18
<
rjo >
re idle experiment: that falls apart if the master and a worker hang.
04:18
<
rjo >
the synthezier is spi iirc.
04:19
<
rjo >
i guess it is some µP.
04:19
<
rjo >
that does the protocol
04:19
<
rjo >
they could have just piped spi over a random ftdi chip.
04:20
<
sb0 >
and some custom firmware
04:20
<
sb0 >
"programmed with Windfreak Technologies proprietary firmware"
04:24
<
sb0 >
yes, if there is nothing to enforce the watchdogs, it'll break
04:26
<
rjo >
it will fix itself when we have the possibility to manage a little "kernel queue" on the core device. then the core device would just run idle experiments when the queue is empty.
04:27
<
sb0 >
you can still crash it e.g. if you RPC a tight loop
04:28
<
sb0 >
also, this kernel queue is incompatible with smooth handover
04:28
<
rjo >
if the master+worker hang, the rpcs would trigger the on-device watchdog, no?
04:29
<
sb0 >
(as the point of the latter is to execute host code between kernels, and that host code will determine what kernel is next)
04:29
<
rjo >
what is smooth handover? seamless handover?
04:29
<
rjo >
seamless as in "no idle exp" or as in "zero dead time"?
04:30
<
sb0 >
obviously you cannot run the idle exp between those, as it would insert additional rtio events
04:31
<
rjo >
ah. we would need something like the guard time/buffer in the fifos. only "commit" an experiment when now() is within guard time.
04:39
<
sb0 >
rjo, how trustworthy is the ad9914 hardware wrt keeping the magic smoke inside the kc705?
04:39
<
sb0 >
I still haven't powered it
04:41
<
sb0 >
12V and all that ...
04:43
anthony has joined #m-labs
04:43
anthony is now known as Guest80466
04:45
Guest80466 is now known as bentley`
04:55
rjo_ has joined #m-labs
04:57
<
rjo_ >
sb0: ha. we run that hardware on a few zedboards and i don't think there has been loss with those setups.
05:59
mumptai has joined #m-labs
06:58
fengling has quit [Ping timeout: 245 seconds]
07:05
fengling has joined #m-labs
07:17
cr1901_modern has quit [Read error: Connection reset by peer]
07:42
mumptai has quit [Remote host closed the connection]
08:29
rjo_ has quit [Ping timeout: 244 seconds]
10:18
bentley` has quit [Ping timeout: 256 seconds]
10:34
<
GitHub121 >
artiq/master e6cc041 Yann Sionneau: conda: give 10 more minutes to ISE to achieve routing
10:49
fengling has quit [Ping timeout: 245 seconds]
10:56
travis-ci has joined #m-labs
10:56
<
travis-ci >
m-labs/artiq#421 (master - e6cc041 : Yann Sionneau): The build has errored.
10:56
travis-ci has left #m-labs [#m-labs]
10:56
<
ysionneau >
ah crap
11:03
<
GitHub147 >
artiq/master c6bba4d Yann Sionneau: Revert "conda: give 10 more minutes to ISE to achieve routing"...
11:31
<
ysionneau >
oh, seems like someone is doing automated conda builds on windows for artiq?
11:31
<
ysionneau >
I'm seeing win32 builds being uploaded
11:36
travis-ci has joined #m-labs
11:36
<
travis-ci >
m-labs/artiq#422 (master - c6bba4d : Yann Sionneau): The build passed.
11:36
travis-ci has left #m-labs [#m-labs]
11:41
FabM has quit [Quit: ChatZilla 0.9.92 [Firefox 40.0.2/20150812163655]]
11:42
<
ysionneau >
or maybe it's just a date issue on anaconda.org website
11:42
FabM has joined #m-labs
13:02
<
mithro >
ysionneau: some time when I'm not about to give a lightning talk, I'd like to chat about conda
13:31
<
ysionneau >
sure :)
14:32
<
GitHub65 >
artiq/master df7489f Yann Sionneau: travis/conda: add automated builds for kc705 nist_qc2 subtarget
14:49
chiggs has quit [Quit: WeeChat 0.4.2]
15:41
bentley` has joined #m-labs
15:44
rjo_ has joined #m-labs
15:44
rjo_ has quit [Client Quit]
16:28
<
GitHub177 >
artiq/master 646721d Yann Sionneau: artiq_flash: add missing semicolons
17:07
travis-ci has joined #m-labs
17:07
<
travis-ci >
m-labs/artiq#424 (master - 646721d : Yann Sionneau): The build has errored.
17:07
travis-ci has left #m-labs [#m-labs]
17:19
cr1901_modern has joined #m-labs
17:21
mumptai has joined #m-labs
20:04
<
GitHub187 >
misoc/master 4acab79 Florent Kermarrec: sdram/module: cleanup indent
21:01
<
cr1901_modern >
Has anyone ever dealt with vvp infinite loops? Having significant trouble debugging why my simulation outright refuses to run to completion.
21:18
_whitelogger has joined #m-labs
22:16
mumptai has quit [Remote host closed the connection]
22:28
travis-ci has joined #m-labs
22:28
<
travis-ci >
m-labs/artiq#424 (master - 646721d : Yann Sionneau): The build passed.
22:28
travis-ci has left #m-labs [#m-labs]