01:06 <sb0> _florent_, any progress?

01:08 <sb0> _florent_, and why is this bug causing delays now? didn't you order a kcu105 in January (!) to deal with this sort of problem?

01:27 <FelixVi2> hi there, I am trying to get around the problem that python 3.6 is not supported

01:28 <FelixVi2> cygwin has 3.4.5-1 as well, is that one known to be working?

01:28 <FelixVi2> or is there another suggested strategy to use migen under cygwin?

01:31 <FelixVi2> sorry, the error actually occurs in misoc

01:43 <sb0> FelixVi2, instead of working around the problem, maybe you can look into cleaning up the patch that was posted in the issues?

02:02 <FelixVi2> sb0, I agree. I was just wondering if there's a known solution as of now - seems like the issue has popped up in various places thus far

02:04 <FelixVi2> sb0, is there a patch? I'm having trouble finding it

02:04 <sb0> FelixVi2, https://github.com/m-labs/migen/issues/62

02:05 <FelixVi2> sb0, thanks! I'll have a look there

02:24 <FelixVi2> sb0, on the first glance the patch does work - now I'm stuck on the issue that lm32-elf-gcc isn't found under cygwin

02:24 <FelixVi2> sb0, I'll report back once progress is made - but I may have to switch to linux instead

02:27 <sb0> you probably have to compile lm32-gcc yourself

02:27 <sb0> if that helps, or1k is also supported via clang/llvm

09:15 <mithro> FelixVi2: if you use the HDMI2USB-litex-firmware you can easily get Python 3.5, gcc, etc

09:26 <GitHub169> [smoltcp] whitequark commented on issue #68: @dhslichter I plan to have it fixed in 2 days. https://github.com/m-labs/smoltcp/issues/68#issuecomment-344534551

09:27 <GitHub72> [smoltcp] phil-opp commented on issue #74: I didn't mean that `get(2..1)` is a bug itself. In the DHCP protocol an option with length 4 has 6 octets (4 + code + length fields) and I thought that TCP was identical. But it seems like in TCP the length field includes the whole option, i.e. length 4 means 4 octets. So no bug here, just a strange difference between the TCP and DHCP option format. https://github.com/m-labs/smoltcp/issues/74#issuecomm

10:26 <_florent_> sb0: I'll try to find a solution today. (castrated if needed) I was only able to test that on kcu105 with loopback, which was ok.

11:52 <GitHub159> [smoltcp] phil-opp commented on pull request #75 e9e9ce3: That's useful, thanks! https://github.com/m-labs/smoltcp/pull/75#discussion_r151105235

13:07 <GitHub46> [smoltcp] phil-opp commented on issue #75: I implemented basic parsing and emitting of packets and DHCP options. I'll add a few more essential options, but I don't think that we need the rest right now. https://github.com/m-labs/smoltcp/pull/75#issuecomment-344586858

14:50 <sb0> _florent_, what fails exactly? is it the delay determination?

14:51 <sb0> _florent_, what was the problem with the phase detector

14:51 <sb0> ?

15:22 <sb0> _florent_, if you do half-rate, you can use half the SERDES outputs for data and one in the other half for phase detection. I did that on milkymist for a HDMI-over-TTL hack (single ended, so phase detector not supported)

15:22 <sb0> that worked surprisingly well

15:25 <cr1901_modern> single-ended HDMI?

15:27 <sb0> convert to LVDS first with a resistor/capacitor hack, then into a LVDS-to-TTL chip

16:21 <bb-m-labs> build #277 of misoc is complete: Failure [failed conda_build] Build details are at http://buildbot.m-labs.hk/builders/misoc/builds/277 blamelist: Sebastien Bourdeauducq <sb@m-labs.hk>

16:33 <sb0> bb-m-labs, force build --branch=master misoc

16:33 <GitHub20> [misoc] sbourdeauducq pushed 1 new commit to master: https://git.io/vFX1o

16:33 <GitHub20> misoc/master 641bfce Sebastien Bourdeauducq: liteeth/1000basex: add full PCS and link management (untested)

16:33 <bb-m-labs> I'll give a shout when the build finishes

16:33 <bb-m-labs> build #279 forced

16:34 <GitHub61> [smoltcp] dhslichter commented on issue #68: Thanks! https://github.com/m-labs/smoltcp/issues/68#issuecomment-344649746

16:35 <bb-m-labs> build #278 of misoc is complete: Failure [failed conda_build] Build details are at http://buildbot.m-labs.hk/builders/misoc/builds/278 blamelist: Sebastien Bourdeauducq <sb@m-labs.hk>

16:35 <bb-m-labs> build #279 of misoc is complete: Failure [failed conda_build] Build details are at http://buildbot.m-labs.hk/builders/misoc/builds/279

21:47 <sb0> _florent_, ping