00:57 <GitHub-m-labs> [artiq] jordens commented on issue #979: Are you trying this with artiq_run or artiq_master? https://github.com/m-labs/artiq/issues/979#issuecomment-383346036

01:01 <GitHub-m-labs> [artiq] whitequark commented on issue #979: artiq_master, like you said. https://github.com/m-labs/artiq/issues/979#issuecomment-383346314

01:08 <GitHub7> [smoltcp] whitequark commented on pull request #196 662cc02: What exactly is not being done appropriately here? https://github.com/m-labs/smoltcp/pull/196#discussion_r183222988

01:08 <GitHub199> [smoltcp] whitequark commented on pull request #196 662cc02: This function doesn't panic, as written. https://github.com/m-labs/smoltcp/pull/196#discussion_r183223003

01:08 <GitHub138> [smoltcp] whitequark commented on pull request #196 662cc02: Which packet types we don't handle, but should? https://github.com/m-labs/smoltcp/pull/196#discussion_r183222991

01:08 <GitHub84> [smoltcp] whitequark commented on pull request #196 662cc02: Why? https://github.com/m-labs/smoltcp/pull/196#discussion_r183222989

01:17 <GitHub69> [smoltcp] whitequark commented on pull request #178 0eb0467: No, I mean an index. Like an `usize`. https://github.com/m-labs/smoltcp/pull/178#discussion_r183223084

01:17 <GitHub44> [smoltcp] whitequark commented on pull request #178 0eb0467: `has_multicast_group` for consistency with other functions. https://github.com/m-labs/smoltcp/pull/178#discussion_r183223094

01:17 <GitHub114> [smoltcp] whitequark commented on pull request #178 0eb0467: I don't think this code processes an IGMP packet. https://github.com/m-labs/smoltcp/pull/178#discussion_r183223057

01:17 <GitHub156> [smoltcp] whitequark commented on pull request #178 0eb0467: The other examples use `expect`, so this one should too. https://github.com/m-labs/smoltcp/pull/178#discussion_r183223064

01:17 <GitHub107> [smoltcp] whitequark commented on issue #178: @podhrmic See comments above. https://github.com/m-labs/smoltcp/pull/178#issuecomment-383347053

01:21 <GitHub175> [smoltcp] whitequark commented on pull request #190 9f89675: > I am happy to explain in more detail, but in short, the problem was that to access fragments from self I needed a mutable borrow of self, but the following processing (such as process_udp) uses a non-mutable borrow.... https://github.com/m-labs/smoltcp/pull/190#discussion_r183223161

01:26 <GitHub61> [smoltcp] whitequark commented on pull request #190 9f89675: This needs an added test. https://github.com/m-labs/smoltcp/pull/190#discussion_r183223209

01:26 <GitHub181> [smoltcp] whitequark commented on pull request #190 9f89675: I would like this to use a ManagedMap as I have described previously. https://github.com/m-labs/smoltcp/pull/190#discussion_r183223213

01:27 <GitHub99> [smoltcp] dlrobertson commented on pull request #196 662cc02: I'll remove this https://github.com/m-labs/smoltcp/pull/196#discussion_r183223225

01:31 <GitHub31> [smoltcp] dlrobertson commented on pull request #196 662cc02: RouterAdvertisements, RouterSolicitations, and Redirect messages aren't currently handled. https://github.com/m-labs/smoltcp/pull/196#discussion_r183223257

01:32 <GitHub129> [smoltcp] whitequark commented on pull request #196 662cc02: Can you add the current status of NDISC support to README in the usual format instead of this TODO? https://github.com/m-labs/smoltcp/pull/196#discussion_r183223268

02:01 <GitHub41> [smoltcp] dlrobertson commented on pull request #196 662cc02: We don't consider the Override flag. Essentially we assume that Override is always set. https://github.com/m-labs/smoltcp/pull/196#discussion_r183223522

02:01 <GitHub27> [smoltcp] dlrobertson commented on pull request #196 662cc02: Good catch. https://github.com/m-labs/smoltcp/pull/196#discussion_r183223523

02:03 <GitHub> [conda-recipes] whitequark pushed 1 new commit to master: https://github.com/m-labs/conda-recipes/commit/69da604b4841cc5313fab28c0a0f898c9e97d73f

02:03 <GitHub> conda-recipes/master 69da604 whitequark: rustc: run Windows build using `start /wait`.

02:04 <GitHub194> [smoltcp] whitequark commented on pull request #196 662cc02: What are the implications of this? https://github.com/m-labs/smoltcp/pull/196#discussion_r183223544

02:04 <whitequark> bb-m-labs: force build --props=package=rustc conda-all

02:04 <bb-m-labs> build forced [ETA 32m11s]

02:04 <bb-m-labs> I'll give a shout when the build finishes

02:09 <GitHub88> [smoltcp] whitequark commented on pull request #187 4297051: Why are you manipulating the buffer directly in this test and not via the socket API? This defeats the purpose of testing the socket. https://github.com/m-labs/smoltcp/pull/187#discussion_r183223577

02:10 <GitHub131> [smoltcp] whitequark commented on issue #193: So, the underlying cause is that PacketBuffer may use one entry in the metadata ring buffer for the padding, which causes buffer exhaustion when the actual metadata entry is inserted.... https://github.com/m-labs/smoltcp/pull/193#issuecomment-383348894

02:40 <GitHub100> [smoltcp] dlrobertson commented on pull request #196 662cc02: Honestly shouldn't be too hard to implement. I'll just implement support for the override flag. https://github.com/m-labs/smoltcp/pull/196#discussion_r183223909

02:47 <bb-m-labs> build #246 of conda-win64 is complete: Failure [failed conda_build] Build details are at http://buildbot.m-labs.hk/builders/conda-win64/builds/246

02:52 <GitHub39> [smoltcp] dlrobertson commented on pull request #196 402a7bf: This should panic as described in the documentation. https://github.com/m-labs/smoltcp/pull/196#discussion_r183224035

02:53 <GitHub35> [smoltcp] dlrobertson commented on pull request #196 402a7bf: Added some sections to the README https://github.com/m-labs/smoltcp/pull/196#discussion_r183224047

03:24 <bb-m-labs> build #397 of conda-lin64 is complete: Failure [failed anaconda_upload] Build details are at http://buildbot.m-labs.hk/builders/conda-lin64/builds/397

03:24 <bb-m-labs> build #120 of conda-all is complete: Failure [failed] Build details are at http://buildbot.m-labs.hk/builders/conda-all/builds/120

06:40 <GitHub-m-labs> [artiq] jordens commented on issue #979: And with the dashboard running as well? https://github.com/m-labs/artiq/issues/979#issuecomment-383359337

06:41 <sb0> whitequark, maybe artiq_coreboot, artiq_coreconfig and artiq_corelog can be merged into one artiq_coremgmt?

06:41 <sb0> whitequark, and artiq_coredebug

06:48 <sb0> whitequark, fyi llvmlite has a hidden dependency on libtinfo.so.5

07:03 <GitHub-m-labs> [artiq] sbourdeauducq pushed 3 new commits to master: https://github.com/m-labs/artiq/compare/3b054855ecb7...4fe09fddd501

07:03 <GitHub-m-labs> artiq/master 4fe09fd Sebastien Bourdeauducq: examples/kc705_nist_clock: update to new ad53xx driver

07:03 <GitHub-m-labs> artiq/master c47d3ec Sebastien Bourdeauducq: zotino: use None as default for ldac and clr...

07:03 <GitHub-m-labs> artiq/master d2d02ae Sebastien Bourdeauducq: artiq_coredebug: fix permissions

07:41 <bb-m-labs> build #1441 of artiq-board is complete: Success [build successful] Build details are at http://buildbot.m-labs.hk/builders/artiq-board/builds/1441

07:45 <bb-m-labs> build #2269 of artiq is complete: Failure [failed python_unittest_2] Build details are at http://buildbot.m-labs.hk/builders/artiq/builds/2269 blamelist: Sebastien Bourdeauducq <sb@m-labs.hk>

14:18 <GitHub-m-labs> [artiq] whitequark commented on issue #979: @jordens ... https://github.com/m-labs/artiq/issues/979#issuecomment-383384831

14:24 <whitequark> sb0: why did test_clock_generator_loopback start to fail?

14:24 <whitequark> there were no compiler changes

14:32 <GitHub-m-labs> [artiq] jordens commented on issue #979: @whitequark Ack. I'll see whether I can isolate it. https://github.com/m-labs/artiq/issues/979#issuecomment-383385695

16:44 <rjo> whitequark: did the repro that cjbe posted not work for you?

16:47 <whitequark> rjo: oh sorry, I missed the fact that there is a repro

16:47 <whitequark> going to check now

18:12 <rjo> whitequark: thanks. if that doesn't do it, i'll reconstruct what i saw.

18:14 <rjo> whitequark: also, ping RTT seems to be ~70ms when those kind of experiments are running (a bit of log traffic, monitoring traffic, and rpcs). is that expected?

18:15 <rjo> and one other thing that i couldn't figure out: the scheduler.pause() RPC is ~500ms (c.f. the dummy scheduler.pause() call with artiq_run being <10ms)

18:16 <whitequark> rjo: re ping: the smoltcp poll code currently isn't fair

18:16 <whitequark> if the NIC can send packets faster than it forms them it's going to spit out the entire buffer however long it takes

18:16 <whitequark> it might make sense to make transmit buffers smaller, or to add a deadline to poll

18:17 <whitequark> probably the latter

18:18 <whitequark> rjo: the scheduler.pause() issue is very odd

18:19 <whitequark> for both of these, file issues if you want me to look at them and I'll do that

18:45 <rjo> whitequark: ack. will do.