00:41
fengling__ has joined #m-labs
00:46
fengling__ has quit [Ping timeout: 240 seconds]
01:18
fengling__ has joined #m-labs
01:28
sb0 has joined #m-labs
02:20
fengling__ is now known as fengling
02:22
<
sb0 >
whitequark, I have reflashed the kc705 with 1.1 and it pings again...
02:27
<
whitequark >
have you read the log above
02:27
<
whitequark >
there's a certain commit. the build before it works, any builds after it don't
02:27
<
whitequark >
by "the build before it" I mean any builds built on CI and uploaded to anaconda before that commit
02:28
<
sb0 >
yes, I have read the log, but I wanted to check that the failure was caused by a change in the code and not some hardware issue that popped up
02:28
<
whitequark >
I've tested the changes yesterday using 1.1
03:19
<
sb0 >
bb-m-labs, force build --revision=dd349b0701cc4168ebb5fa58e0dfd2e8508d755a artiq
03:19
<
bb-m-labs >
build forced [ETA 20m20s]
03:19
<
bb-m-labs >
I'll give a shout when the build finishes
03:32
mumptai has quit [Ping timeout: 253 seconds]
03:39
<
whitequark >
sb0: so it's completely deterministic
03:39
<
whitequark >
but I don't understand why reverting the next commit doesn't work...
03:43
mumptai has joined #m-labs
03:46
<
sb0 >
that commit worked before
03:46
<
sb0 >
that's a different failure
03:46
<
sb0 >
some windows breakage...
03:47
<
sb0 >
maybe since the python upgrade to 3.5.2?
03:48
<
sb0 >
bb-m-labs, force build --revision=4a6c270afe8e65f611d69ec5004cfca81a224672 artiq
03:48
<
bb-m-labs >
build forced [ETA 20m20s]
03:48
<
bb-m-labs >
I'll give a shout when the build finishes
03:49
<
sb0 >
did python fuck up asyncio.open_connection on windows in 3.5.2? that's going to be fun
03:55
<
whitequark >
sb0: I got your optics
04:08
<
sb0 >
bb-m-labs, force build --revision=71921de5bde602b7e65f4a79dffda6a60f3ffe84 artiq
04:08
<
bb-m-labs >
build forced [ETA 20m20s]
04:08
<
bb-m-labs >
I'll give a shout when the build finishes
04:27
<
sb0 >
vivado garbage
04:30
<
sb0 >
the vivado we have is pretty old, maybe they fixed that since then
04:31
<
cr1901_modern >
At the very least, their installation process has gotten far less fucked up, last I heard
04:32
<
cr1901_modern >
(Though maybe fixing two bugs is giving them too much credit :P)
04:32
<
sb0 >
can you install in command line now or do you still need X?
04:33
<
cr1901_modern >
Oh, AFAIK you still need X. The bug I was referring to is that the installer was hardcoded to check a specific interface to generate your product ID, and if you don't have an "eth0" you couldn't install it
04:33
<
sb0 >
my preferred way of deploying xilinx tools is to install it once, tarball it, decompress on every machine that needs it and use udev to set the MAC address to whatever their licensing shitware wants
04:35
<
cr1901_modern >
Doesn't cause problems on your network that two devices have same MAC address?
04:35
<
sb0 >
yes. many systems don't have a eth0, and since you need udev to create one, you might as well take care of the MAC as well
04:35
<
sb0 >
yes, that doesn't work if you connect two of those tweaked interfaces to the same net
04:36
<
sb0 >
you can also give it virtual tun/tap interface
04:37
<
cr1901_modern >
I don't understand how that solves the problem (b/c I don't actually know what a tun/tap interface does :P)
04:42
<
cr1901_modern >
I haven't actually installed Vivado yet b/c I need to nuke my Linux install and extend my Windoze partition. Only problem is that GRUB is right after my Windows partition and I'm unsure how to move it without have access to grubinst
04:42
<
cr1901_modern >
grub-install*
04:45
<
GitHub85 >
artiq/master 520b269 Sebastien Bourdeauducq: RELEASE_NOTES: artiq_client persist flag
04:45
<
GitHub85 >
artiq/master ea10a2a Sebastien Bourdeauducq: manual/faq: update for inter-experiment seamless handover
04:45
<
GitHub85 >
artiq/master d2d897a Sebastien Bourdeauducq: manual/getting_started: add examples location info, closes #501
04:49
<
sb0 >
well, I guess the problem is open_connection() will now raise OSError instead of ConnectionRefusedError when attempting to connect to a closed port on localhost
04:52
<
whitequark >
sb0: I need the kc705 now
04:52
<
whitequark >
going to use it for a few minutes
04:52
<
sb0 >
okay, you can stop that build
04:52
<
sb0 >
bb-m-labs, stop build artiq
04:52
<
bb-m-labs >
try 'stop build WHICH <REASON>'
04:52
<
sb0 >
bb-m-labs, stop build artiq kc705
04:52
<
bb-m-labs >
build 793 interrupted
05:41
FabM has joined #m-labs
05:43
<
GitHub129 >
artiq/release-1 d96d222 Sebastien Bourdeauducq: targets/kc705: fix import
05:45
<
whitequark >
why cannot you compare numpy arrays?
05:45
<
whitequark >
(numpy.array([42, 43]),) == (numpy.array([42, 43]),)
05:45
<
whitequark >
ValueError: The truth value of an array with more than one element is ambiguous. Use a.any() or a.all()
05:46
<
whitequark >
oh, it tries to do element-by-element comparison
06:06
<
sb0 >
bb-m-labs, force build artiq-kc705-nist_qc2
06:06
<
bb-m-labs >
build #172 forced
06:06
<
bb-m-labs >
I'll give a shout when the build finishes
06:07
<
sb0 >
and what causes it is migen commit 5630b7c81422184de272d49e0c9d9c29240278dd from June 11
06:08
<
sb0 >
hmm, does it take main or dev for installing migen?
06:14
<
sb0 >
and the SMA voltage situation is: the default is 2.5V indeed, but since NIST are connecting 3.3V DDSes directly on the same I/O bank and K7 FPGAs cannot take input voltages higher than VCCO (something that was found only after the boards were made, of course) the KC705 power controller is reprogrammed to set those to 3.3
06:34
<
GitHub56 >
artiq/master d90fd7d whitequark: compiler: implement numpy.full (#424).
06:34
<
GitHub56 >
artiq/master 7a671fb whitequark: embedding: treat numpy.{int32,int64,array} specially (#424)....
06:44
<
sb0 >
whitequark, done with the kc705?
06:58
<
GitHub34 >
artiq/master c065b58 Sebastien Bourdeauducq: require Python 3.5.2, remove monkey patches
06:58
<
GitHub34 >
artiq/master 8cb29fc Sebastien Bourdeauducq: targets/kc705: redefine user SMAs as 3.3V IO. Closes #502
06:58
<
GitHub153 >
artiq/release-1 adcf53f Sebastien Bourdeauducq: targets/kc705: redefine user SMAs as 3.3V IO. Closes #502
07:01
<
sb0 >
whitequark, if migen builds are going to trigger misoc and artiq builds, the latter should use the migen package just built ...
07:27
<
sb0 >
I'm downloading the new vivado, will try to upgrade the buildserver
07:32
<
whitequark >
sb0: I think that should work
07:33
<
whitequark >
I hope our vivado isn't exposed to the internet
07:34
<
sb0 >
it obviously doesn't: before those commits I had pushed, building from migen master results in the issue Raghu has reported, whereas the automatic build succeeds (see above)
07:34
<
sb0 >
we're using the node-locked license, fortunately, so it doesn't open ports
07:53
fengling has quit [Ping timeout: 240 seconds]
08:00
fengling has joined #m-labs
08:03
<
whitequark >
hey sb0 I just had an idea for a ghetto instrument
08:03
<
whitequark >
take any cheap LA
08:03
<
whitequark >
like those $15 chinese clones of saleae
08:03
<
whitequark >
put a voltage to frequency converter on every input, which you can possibly bypass
08:04
<
whitequark >
this allows you to capture analog channels synchronized with digital channels, on an arbitrary set of inputs, and with a natural choice between resolution and sample rate
08:33
<
GitHub24 >
artiq/master 373578b whitequark: embedding: fix location for diagnostics on quoted values....
08:34
mumptai has quit [Remote host closed the connection]
08:36
sb0 has quit [Ping timeout: 246 seconds]
08:40
fengling has quit [Ping timeout: 240 seconds]
08:49
sb0 has joined #m-labs
08:53
fengling has joined #m-labs
08:54
sb0 has quit [Ping timeout: 260 seconds]
08:57
sb0 has joined #m-labs
09:03
<
GitHub103 >
artiq/release-1 a801cde whitequark: embedding: fix location for diagnostics on quoted values....
09:05
<
whitequark >
uhm, that failed tests
09:05
<
whitequark >
I was about to push the fix
09:05
<
sb0 >
ok, please push it to release-1 as well
09:11
sb0 has quit [Ping timeout: 258 seconds]
09:17
rohitksingh has joined #m-labs
09:34
sb0 has joined #m-labs
09:44
sb0 has quit [Read error: Connection reset by peer]
09:46
sb0 has joined #m-labs
10:03
sb0 has quit [Ping timeout: 246 seconds]
10:11
sb0 has joined #m-labs
10:39
fengling has quit [Ping timeout: 240 seconds]
10:44
<
sb0 >
cr1901_modern, the installer did
*not* improve
10:44
<
sb0 >
I can't even run it
10:44
<
sb0 >
ERROR: This installation is not supported on 32 bit platforms.
10:44
<
sb0 >
even though it is running on 64
10:45
fengling has joined #m-labs
11:05
<
whitequark >
unpack it manually and fix the uname
11:05
<
whitequark >
change uname -i to uname -m in the xsetup.sh script, to be specififc
11:25
fengling has quit [Ping timeout: 240 seconds]
11:35
<
GitHub189 >
pythonparser/master 5cb107b whitequark: diagnostic.Engine: allow adding contextual notes.
11:35
<
GitHub189 >
pythonparser/master 03d87cb whitequark: Fix python default argument idiocy.
11:40
<
GitHub189 >
pythonparser/master c744600 whitequark: diagnostic.Engine: allow adding multiple notes at once.
11:49
<
GitHub65 >
artiq/master 5a79fcf whitequark: embedding: reimplement 373578bc properly....
11:50
<
whitequark >
bb-m-labs: force build --props=package=pythonparser conda-linux64
11:50
<
bb-m-labs >
no such builder 'conda-linux64'
11:50
<
whitequark >
bb-m-labs: force build --props=package=pythonparser conda-lin64
11:50
<
bb-m-labs >
build #163 forced
11:50
<
bb-m-labs >
I'll give a shout when the build finishes
11:50
<
GitHub32 >
conda-recipes/master aa093cc whitequark: pythonparser: bump.
11:53
<
whitequark >
bb-m-labs: force build artiq
11:53
<
bb-m-labs >
build forced [ETA 20m20s]
11:53
<
bb-m-labs >
I'll give a shout when the build finishes
11:54
rohitksingh has quit [Ping timeout: 246 seconds]
12:22
fengling has joined #m-labs
12:26
flaviusb has quit [Ping timeout: 250 seconds]
12:41
<
GitHub34 >
artiq/master 653eeb4 whitequark: runtime: fix serialization of object lists....
12:52
fengling has quit [Ping timeout: 240 seconds]
13:49
fengling has joined #m-labs
13:53
sb0 has quit [Read error: Connection reset by peer]
13:54
sb0 has joined #m-labs
13:54
fengling has quit [Ping timeout: 240 seconds]
14:43
<
sb0 >
whitequark, can you cherry-pick the other one? there are conflicts.
14:44
<
GitHub102 >
artiq/release-1 6a1706b whitequark: runtime: fix serialization of object lists....
14:51
fengling has joined #m-labs
14:56
fengling has quit [Ping timeout: 240 seconds]
14:56
<
GitHub33 >
conda-recipes/master 5803bd9 whitequark: libgit2: build as Release....
14:57
rohitksingh has joined #m-labs
14:59
<
whitequark >
bb-m-labs: force build --props=package=libgit2 conda-all
14:59
<
bb-m-labs >
build #61 forced
14:59
<
bb-m-labs >
I'll give a shout when the build finishes
15:05
<
sb0 >
the xilinx installer also cannot start into a Xvnc server
15:06
<
sb0 >
there's some java x11 garbage that crashes immediately
15:07
<
whitequark >
that's not java
15:07
<
whitequark >
Xvnc is complete trash
15:07
<
whitequark >
most likely what happens is it uses 8-bit color mode, which almost nothing supports these days
15:08
<
whitequark >
Xvnc is broken from both the X side and the VNC side, practically speaking. amazingly useless piece of software
15:08
<
sb0 >
sure. and if X worked correctly, we would not need Xvnc
15:08
<
whitequark >
try -depth 24
15:08
<
whitequark >
for Xvnc
15:12
<
sb0 >
hm, actually it starts, if I don't use sudo
15:24
<
sb0 >
installing ...
15:28
<
GitHub85 >
conda-recipes/master abe4f53 whitequark: libssh2: build as Release....
15:28
<
whitequark >
bb-m-labs: force build --props=package=libssh2 conda-all
15:28
<
bb-m-labs >
build forced [ETA 4m40s]
15:28
<
bb-m-labs >
I'll give a shout when the build finishes
15:35
<
whitequark >
*facedesk*
15:35
<
whitequark >
-DCMAKE_BUILD_TYPE doesn't do anything when building on msvc
15:38
<
whitequark >
bb-m-labs: force build --props=package=libssh2 conda-all
15:38
<
bb-m-labs >
build forced [ETA 3m28s]
15:38
<
bb-m-labs >
I'll give a shout when the build finishes
15:38
<
whitequark >
bb-m-labs: force build --props=package=libgit2 conda-all
15:38
<
bb-m-labs >
build forced [ETA 3m28s]
15:38
<
bb-m-labs >
I'll give a shout when the build finishes
15:38
<
GitHub111 >
conda-recipes/master 1b5c767 whitequark: libgit2, libssh2: actually build as Release on MSVC.
15:40
<
sb0 >
bb-m-labs, force build artiq
15:40
<
bb-m-labs >
build forced [ETA 20m20s]
15:40
<
bb-m-labs >
I'll give a shout when the build finishes
15:41
<
whitequark >
bb-m-labs: force build --props=package=libssh2 conda-all
15:41
<
bb-m-labs >
The build has been queued, I'll give a shout when it starts
15:41
<
GitHub25 >
conda-recipes/master 12ccd3b whitequark: libssh2: fix screwup.
15:45
<
bb-m-labs >
build forced [ETA 4m54s]
15:45
<
bb-m-labs >
I'll give a shout when the build finishes
15:47
sb0 has quit [Quit: Leaving]
15:52
fengling has joined #m-labs
15:57
fengling has quit [Ping timeout: 240 seconds]
16:01
<
GitHub52 >
artiq/release-1 2678bb0 whitequark: embedding: reimplement 373578bc properly....
16:47
rohitksingh has quit [Ping timeout: 264 seconds]
17:03
rohitksingh has joined #m-labs
17:20
rohitksingh has quit [Ping timeout: 272 seconds]
17:32
mumptai has joined #m-labs
17:34
rohitksingh has joined #m-labs
17:54
rohitksingh has quit [Ping timeout: 272 seconds]
17:55
fengling has joined #m-labs
18:00
fengling has quit [Ping timeout: 240 seconds]
18:04
FabM has quit [Quit: ChatZilla 0.9.92 [Firefox 47.0/20160604131506]]
18:06
rohitksingh has joined #m-labs
18:57
fengling has joined #m-labs
19:02
fengling has quit [Ping timeout: 240 seconds]
19:05
rohitksingh has quit [Quit: Leaving.]
21:00
fengling has joined #m-labs
21:00
sandeepkr has quit [Quit: Leaving]
21:05
fengling has quit [Ping timeout: 240 seconds]
21:10
sandeepkr has joined #m-labs
22:01
fengling has joined #m-labs
22:06
fengling has quit [Ping timeout: 240 seconds]
22:45
mumptai has quit [Quit: Verlassend]
23:03
fengling has joined #m-labs
23:08
fengling has quit [Ping timeout: 240 seconds]
23:29
<
GitHub69 >
artiq/master 2a5a1f3 Robert Jordens: browser, worker: feed experiments dummy devices, closes #454...
23:29
<
GitHub69 >
artiq/master b7cca38 Robert Jordens: browser: let the state manager handle the subcomponents