09:11 <larsc> hm, does anybody know whether it is possible to just configure a pin in a ucf file without connecting a signal to it? E.g. I want to configure one pin to have a pullup

09:18 <lekernel> use the PULLUP primitive

09:21 <larsc> yea, but if I use "NET FOOBAR LOC=123 | PULLUP | S" it complains that it can't find the FOOBAR net

09:32 <_florent_> in ngdbuild you can enable "Allow Unmatched LOC Constraints", it won't complains anymore about FOOBAR net, but I don't know if the PULLUP constraint will be ignored or not...

09:35 <larsc> ok, thanks

09:52 <lekernel> I'm not talking of the PULLUP constraint, but the primitive

09:52 <lekernel> something like: output some_signal; ... PULLUP(.S(some_signal));

09:52 <lekernel> that's done in the HDL, not UCF though - but it could solve your problem still

10:10 <larsc> ok, thanks

10:49 <wpwrak> wheee ! shipment reached BUE. now let's see how long it takes to get to/through customs. CPI is already good. customs dwell there, too.

10:49 <larsc> it doesn't have a power plug, so you should be fine ;)

10:51 <wpwrak> yeah. and that's the not so picky customs. the one where all the regular mail goes through.

10:53 <wpwrak> also the paperwork is considerably simpler there. something like two signatures and you're done. there are still massive waits between things, though. ah, and they have a very high probability of requiring your presence. something like 80% or more.

10:53 <wpwrak> so it's a high risk of a small annoyance while the other has a small risk of a major annoyance.

14:37 <lekernel> _florent_, have you tried migen on windows already?

15:02 <_florent_> yes I'm using it on windows

15:03 <_florent_> I don't have issues

15:08 <_florent_> the only issues I had was with mibuild and the Xilinx module, I just added a --no-source option and it's ok

15:09 <_florent_> http://git.io/OumLOQ

15:16 <lekernel> great, i'll merge that patch (and detect windows and make --no-source the default)

15:16 <lekernel> thanks!