Friday, 2021-04-09

*** tpb has joined #litex00:00
*** thorns514 has joined #litex00:47
*** Degi has quit IRC01:01
*** Degi has joined #litex01:03
*** FFY00_ has quit IRC01:45
*** FFY00_ has joined #litex01:45
*** Bleepshop|2 is now known as Bleepshop02:02
*** thorns514 has quit IRC02:07
*** Bertl_oO is now known as Bertl_zZ05:07
*** kgugala has joined #litex06:24
*** kgugala_ has quit IRC06:24
*** kgugala has quit IRC06:38
*** kgugala has joined #litex06:39
*** peeps[zen] has quit IRC08:02
*** davidlattimore has quit IRC08:02
*** apolkosnik[m] has quit IRC08:02
*** guan has quit IRC08:02
*** SpaceCoaster has quit IRC08:02
*** feldim2425 has quit IRC08:02
*** dormito has quit IRC08:02
*** CarlFK has quit IRC08:02
*** st-gourichon-fid has quit IRC08:02
*** pepijndevos has quit IRC08:02
*** carlomaragno has quit IRC08:02
*** shoragan has quit IRC08:02
*** Emantor has quit IRC08:02
*** tucanae47 has quit IRC08:02
*** [Matt] has quit IRC08:02
*** cr1901_modern has quit IRC08:02
*** felix_ has quit IRC08:02
*** shenki has quit IRC08:02
*** xobs has quit IRC08:02
*** promach3 has quit IRC08:02
*** leons has quit IRC08:02
*** _whitelogger has quit IRC08:02
*** shoragan[m] has quit IRC08:02
*** tcal has quit IRC08:02
*** midnight has quit IRC08:02
*** vup has quit IRC08:02
*** joseng has quit IRC08:02
*** benh has quit IRC08:02
*** pdp7 has quit IRC08:02
*** Xesxen has quit IRC08:02
*** _florent_ has quit IRC08:02
*** lf has quit IRC08:02
*** Claude has quit IRC08:02
*** sorear has quit IRC08:02
*** acathla has quit IRC08:02
*** esden has quit IRC08:02
*** the_jinx has quit IRC08:02
*** xfxf has quit IRC08:02
*** y2kbugger has quit IRC08:02
*** gruetzkopf has quit IRC08:02
*** disasm[m] has quit IRC08:02
*** JJJollyjim1 has quit IRC08:02
*** Bertl_zZ has quit IRC08:02
*** awygle has quit IRC08:02
*** flammit has quit IRC08:02
*** kgugala has quit IRC08:02
*** Bleepshop has quit IRC08:02
*** _franck_ has quit IRC08:02
*** captain_morgan has quit IRC08:02
*** awordnot has quit IRC08:02
*** tmbinc has quit IRC08:02
*** ric96 has quit IRC08:02
*** tannewt has quit IRC08:02
*** levi has quit IRC08:02
*** rektide has quit IRC08:02
*** miek has quit IRC08:02
*** Stary has quit IRC08:02
*** gatecat has quit IRC08:02
*** mithro has quit IRC08:02
*** alanvgreen has quit IRC08:02
*** x56_ has quit IRC08:02
*** key2 has quit IRC08:02
*** rohitksingh has quit IRC08:02
*** palmer has quit IRC08:02
*** somlo has quit IRC08:02
*** RaYmAn has quit IRC08:02
*** mntmn has quit IRC08:02
*** Degi has quit IRC08:02
*** zyp has quit IRC08:02
*** TMM has quit IRC08:02
*** proteusguy has quit IRC08:02
*** keesj has quit IRC08:02
*** kbeckmann has quit IRC08:02
*** anuejn has quit IRC08:02
*** geertu has quit IRC08:02
*** trabucayre has quit IRC08:02
*** sajattack[m] has quit IRC08:02
*** jryans has quit IRC08:02
*** tumbleweed has quit IRC08:02
*** bubble_buster has quit IRC08:02
*** simeonm has quit IRC08:02
*** j4cbo has quit IRC08:02
*** pftbest_ has quit IRC08:02
*** indy has quit IRC08:02
*** lkcl has quit IRC08:02
*** FFY00_ has quit IRC08:02
*** nickoe has quit IRC08:02
*** Finde has quit IRC08:02
*** bastelmann has quit IRC08:02
*** nats` has quit IRC08:02
*** disasm[m] has joined #litex08:49
*** [Matt] has joined #litex08:49
*** midnight has joined #litex08:49
*** vup has joined #litex08:49
*** joseng has joined #litex08:49
*** pdp7 has joined #litex08:49
*** benh has joined #litex08:49
*** st-gourichon-fid has joined #litex08:49
*** pepijndevos has joined #litex08:49
*** carlomaragno has joined #litex08:49
*** shoragan has joined #litex08:49
*** Emantor has joined #litex08:49
*** tucanae47 has joined #litex08:49
*** _florent_ has joined #litex08:49
*** FFY00_ has joined #litex08:49
*** nickoe has joined #litex08:49
*** Finde has joined #litex08:49
*** bastelmann has joined #litex08:49
*** nats` has joined #litex08:49
*** Xesxen has joined #litex08:49
*** pftbest_ has joined #litex08:49
*** lkcl has joined #litex08:49
*** proteusguy has joined #litex08:49
*** keesj has joined #litex08:49
*** kbeckmann has joined #litex08:49
*** anuejn has joined #litex08:49
*** geertu has joined #litex08:49
*** trabucayre has joined #litex08:49
*** peeps[zen] has joined #litex08:49
*** dormito has joined #litex08:49
*** davidlattimore has joined #litex08:49
*** guan has joined #litex08:49
*** SpaceCoaster has joined #litex08:49
*** feldim2425 has joined #litex08:49
*** tumbleweed has joined #litex08:49
*** j4cbo has joined #litex08:49
*** bubble_buster has joined #litex08:49
*** simeonm has joined #litex08:49
*** flammit has joined #litex08:49
*** awygle has joined #litex08:49
*** Bertl_zZ has joined #litex08:49
*** gruetzkopf has joined #litex08:49
*** y2kbugger has joined #litex08:49
*** xfxf has joined #litex08:49
*** the_jinx has joined #litex08:49
*** acathla has joined #litex08:49
*** esden has joined #litex08:49
*** sorear has joined #litex08:49
*** Claude has joined #litex08:49
*** lf has joined #litex08:49
*** TMM has joined #litex08:49
*** shenki has joined #litex08:49
*** felix_ has joined #litex08:49
*** cr1901_modern has joined #litex08:49
*** Degi has joined #litex08:49
*** zyp has joined #litex08:49
*** kgugala has joined #litex08:49
*** Bleepshop has joined #litex08:49
*** _franck_ has joined #litex08:49
*** captain_morgan has joined #litex08:49
*** awordnot has joined #litex08:49
*** tmbinc has joined #litex08:49
*** tannewt has joined #litex08:49
*** levi has joined #litex08:49
*** miek has joined #litex08:49
*** rektide has joined #litex08:49
*** Stary has joined #litex08:49
*** gatecat has joined #litex08:49
*** x56_ has joined #litex08:49
*** alanvgreen has joined #litex08:49
*** rohitksingh has joined #litex08:49
*** key2 has joined #litex08:49
*** palmer has joined #litex08:49
*** somlo has joined #litex08:49
*** RaYmAn has joined #litex08:49
*** mntmn has joined #litex08:49
*** _whitelogger has joined #litex08:49
*** tcal has joined #litex08:49
*** indy has joined #litex08:49
*** ric96 has joined #litex08:49
*** mithro has joined #litex08:49
*** leons has joined #litex08:57
*** JJJollyjim has joined #litex08:57
*** CarlFK has joined #litex08:57
*** sajattack[m] has joined #litex08:57
*** jryans has joined #litex08:57
*** shoragan[m] has joined #litex08:57
*** promach3 has joined #litex08:57
*** xobs has joined #litex08:57
*** apolkosnik[m] has joined #litex08:57
*** Bertl_zZ is now known as Bertl11:47
*** kgugala_ has joined #litex13:25
*** thorns514 has joined #litex13:27
*** kgugala has quit IRC13:28
*** thorns51_ has joined #litex13:40
*** thorns514 has quit IRC13:42
*** thorns514 has joined #litex13:43
*** thorns51_ has quit IRC13:44
*** thorns51_ has joined #litex13:58
*** thorns514 has quit IRC13:59
*** thorns51_ has quit IRC14:09
*** thorns514 has joined #litex17:46
*** kgugala_ has quit IRC18:05
*** kgugala has joined #litex18:05
*** rj has joined #litex18:07
*** pftbest_ has quit IRC18:11
*** pftbest has joined #litex18:13
*** m4ssi has joined #litex18:39
*** thorns514 has quit IRC19:01
leons_florent_: thanks so much for integrating my LiteEth timestamping PR!19:02
leonsI have one question though (I feel like I'm missing something). You changed the code to make the SRAM Writer only have a result status FIFO when timestamping is enabled19:03
leonsI was under the impression by introducing the result status FIFO I was implicitly solving another issue: how does a LiteEth user know how many packets have been sent when not having a result FIFO?19:04
leonsSure, they can count interrupts. Imagine however the OS is pretty busy and it will take a while to do anything reasonable with the interrupt. LiteEth might have sent multiple packets, right?19:05
*** rj has quit IRC19:30
*** pftbest has quit IRC19:35
_florent_leons: The reasoning is fine and I agree with it, I did this change mostly because it is changing the software API (with CSRs moving) and so would break the current drivers using this API, so I just want to be sure we do it at the right time (and maybe after doing https://github.com/enjoy-digital/litex/issues/548 that could avoid breaking current drivers with the change).19:39
*** pftbest has joined #litex19:50
*** pftbest has quit IRC19:54
*** pftbest has joined #litex20:11
*** thorns514 has joined #litex20:14
leonsAh, thanks for the explanation, this makes sense. I already prepared to get my hands dirty in the Linux LiteEth driver to adjust it, but it makes sense that there are many clients relying on it20:31
leonsI think it's very important nonetheless, as my driver (without timestamps retrofitted) currently relies on one interrupt routed to the driver per packet, and I'm reasonably sure that caused issues at least a few times already20:33
leonsI'm a huge fan of having more static, well defined CSRs as proposed in https://github.com/enjoy-digital/litex/issues/54820:34
leonsIn case one would, in the future, run into the case where the CSR interface is deemed sufficiently broken to need to redesign it backwards-incompatible, wouldn't it make sense to version the interface while we're at it? I'd love to fail fast in my driver if the interface is changed in a backwards-incompatible way. If that might make sense, I'll be sure to add a comment to #54820:36
zypwhat you might want to do is add a version register20:37
leonszyp: yup! ideally, always at the start of the CSR bank for each peripheral. An 8-bit number should probably suffice20:38
zypI've got an upcoming usecase for this too20:39
zypI've designed an ecp5 based board for the orbuculum project that'll act as a swd/jtag/trace probe: https://bin.jvnv.net/file/e4wot.png20:40
leonszyp: hah, my usecase is "I have a bunch of drivers, and I don't know if updating LiteX will break my drivers in a way that manifests itself in a weird edge-case"20:40
leonsI suppose that's a pretty common "usecase" haha20:40
zypand I'm planning to eventually port the blackmagic probe firmware to run on a vexriscv in the fpga20:41
zypso it would be nice to be able to present a consistent register interface towards the firmware, since the firmware would be fairly decoupled from the gateware platform20:41
leonszyp: That looks pretty awesome, I didn't know orbuculum yet!20:42
leonsSpecifically what makes backwards-incompatible changes in CSRs tricky for me is that the OS I'm maintaining the LiteX target for is written in Rust, so I can't integrate it with the usual generated headers20:44
leonsAnd auto-generating the driver's register abstraction from JSON/SVD isn't really a great option either20:44
*** kgugala_ has joined #litex21:59
*** kgugala has quit IRC22:02
*** kgugala has joined #litex22:03
*** kgugala_ has quit IRC22:06
*** lf has quit IRC23:34
*** lf has joined #litex23:34

Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!