| *** tpb has joined #tomu | 00:00 | |
| xobs | mithro: you probably could use litex-buildenv by commenting out "import lxbuildenv" and then providing your own copies of the relevant support libraries. | 00:11 |
|---|---|---|
| *** emeb has quit IRC | 00:44 | |
| *** AmosSam has left #tomu | 00:52 | |
| *** AmosSam has joined #tomu | 00:52 | |
| *** awe00 has quit IRC | 01:38 | |
| TomKeddie | mithro: I kinda do this on my arty hack, everything comes from litex-buildenv except valentyusb which I've hacked in as "sys.path.append(os.path.dirname(os.path.realpath(__file__)) + "/../foboot/hw/deps/valentyusb")" - see https://github.com/TomKeddie/prj-litex/tree/master/arty/usb/hw | 03:38 |
| tpb | Title: prj-litex/arty/usb/hw at master · TomKeddie/prj-litex · GitHub (at github.com) | 03:38 |
| TomKeddie | xobs: fyi we upstreamed your rgb driver today, https://github.com/timvideos/litex-buildenv/pull/164 | 03:42 |
| tpb | Title: platform/icebreaker: port rgb led driver from fomu by TomKeddie · Pull Request #164 · timvideos/litex-buildenv · GitHub (at github.com) | 03:42 |
| mithro | TomKeddie: Still need to fix -> https://github.com/m-labs/nmigen-boards/issues/22 | 03:45 |
| tpb | Title: Support silkscreen references (and other information) in platform definitions · Issue #22 · m-labs/nmigen-boards · GitHub (at github.com) | 03:45 |
| TomKeddie | Nice idea, I can't touch this though. BTW would like to see the descriptions as comments in the constraints file, very helpful for more advanced players. | 03:48 |
| mithro | TomKeddie: That was not the issue I meant to link | 03:48 |
| TomKeddie | (I wondered) | 03:48 |
| mithro | TomKeddie: https://github.com/im-tomu/foboot/issues/20 | 03:48 |
| tpb | Title: Rework code in SBLED so LED information is in the platform · Issue #20 · im-tomu/foboot · GitHub (at github.com) | 03:48 |
| TomKeddie | Oh yes, I left the current draws parameters hard coded, will fix this. I didn't want to interrupt your merging today. I was thinking a vector argument to init (like pads). | 03:50 |
| mithro | TomKeddie: It's more the color mapping issue | 03:51 |
| TomKeddie | mithro: oh, I see, the upstreamed platforms need updating to capture this information? | 03:52 |
| mithro | TomKeddie: I think what needs to be done is that the upstream platforms have "user_ledr_n", "user_ledg_n" and "user_ledb_n" and names on the right pins and then SBLED compares the mapping between colors and pins to generate the lookup matrix that xobs hardcoded... | 03:54 |
| TomKeddie | mithro: I'm not sure it's possible to map out everything, am I right that the there is no pin mux for the hardened block, to map the registers to pins? | 03:56 |
| mithro | TomKeddie: the map needs to be between the SB_LEDDA_IP PWM outputs and the SB_RGBA_DRV inputs if I understand correctly | 03:57 |
| TomKeddie | mithro: ah, that makes sense. ok, will look at this. | 03:58 |
| mithro | First thing is to decode where the actual colors are connected on each platform | 03:59 |
| TomKeddie | mithro: the rgb block is kinda cemented in the fomu/foboot platform. It occurred to me it would probably be easier to write an emulator for non-ice40 platforms then all the fomu material would just work on 7series and ecp5. | 04:00 |
| mithro | TomKeddie: would be a pretty easy block to write I think? | 04:02 |
| TomKeddie | mithro: that's what I'm thinking, might be easier to emulate than futz around with all the platform stuff. I've already emulated "touch" on my arty port so that part of the (current) foboot s/w just works. | 04:03 |
| TomKeddie | The part of foboot that is going to be hard is the pointers to the s/w. Each chip has a very different image size so it's difficult to hard code them. I was thinking that a table/metadata in the last page of the flash might give us what we want. | 04:06 |
| xobs | TomKeddie: wouldn't it make sense to have them as platform variables? | 04:17 |
| TomKeddie | xobs: but then the same code wouldn't run on each platform. | 04:23 |
| TomKeddie | I guess we could move the variables to registers but that burns precious fpga resources. | 04:23 |
| TomKeddie | xobs: or does the code contain all the variables (or syscalls) for all the platforms (which bloats the code). | 04:26 |
| TomKeddie | xobs: it seems to be s/w compatibility is desirable (from a maintenance pov) we need separate fpga images but running the same copy of micropython on all platforms would be cool. | 04:27 |
| xobs | I think Foboot 2.0 will need to incorporate a lot of feedback. Things such as syscalls to get those addresses. And a better USB stack. | 04:32 |
| TomKeddie | Cool, we need to let you ship 1.0 first. | 04:36 |
| xobs | Up to 1.8.6 now! | 04:38 |
| mithro | xobs - was the csr / SRAM fix enough to unblock the spiflash stuff? | 04:40 |
| TomKeddie | mithro: has there been any thought given to building active low inversion into the signal definitions? I'm think whitequark probably has this in nmigen? I hate the _n stuff we should bury at the edge and mostly hide it from the logic (imho). | 04:48 |
| mithro | Ask whitequark is probably the best way to get an answer :-) | 04:49 |
| TomKeddie | Do we want to add it here or are we just biding our time until nmigen takes over the world? | 04:52 |
| *** rohitksingh_work has joined #tomu | 04:53 | |
| xobs | mithro: mostly unblocked, yes. | 05:17 |
| xobs | I'm using foboot-2.0 to test foboot-1.8.6. I'm developing the spi stuff on a pvt board. | 05:18 |
| *** AmosSam has left #tomu | 09:13 | |
| *** AmosSam has joined #tomu | 09:14 | |
| *** AmosSam has left #tomu | 09:30 | |
| *** AmosSam has joined #tomu | 09:30 | |
| *** awe00 has joined #tomu | 09:45 | |
| *** TomKeddie has quit IRC | 10:16 | |
| *** TomKeddie has joined #tomu | 10:18 | |
| *** flammit has quit IRC | 12:04 | |
| *** jacobrosenthal has quit IRC | 12:04 | |
| *** ukembedded has quit IRC | 12:04 | |
| *** mithro has quit IRC | 12:04 | |
| *** Kitlith has quit IRC | 12:07 | |
| *** flammit has joined #tomu | 12:10 | |
| *** jacobrosenthal has joined #tomu | 12:10 | |
| *** ukembedded has joined #tomu | 12:10 | |
| *** mithro has joined #tomu | 12:10 | |
| *** Kitlith has joined #tomu | 12:10 | |
| *** rohitksingh_work has quit IRC | 13:11 | |
| *** AmosSam has left #tomu | 14:07 | |
| *** AmosSam has joined #tomu | 14:08 | |
| *** AmosSam has left #tomu | 14:10 | |
| *** AmosSam has joined #tomu | 14:10 | |
| *** awe00 has quit IRC | 14:51 | |
| *** ghanashyamprabhu has joined #tomu | 15:16 | |
| *** ghanashyamprabhu has quit IRC | 15:17 | |
| *** awe00 has joined #tomu | 15:17 | |
| *** rohitksingh has joined #tomu | 15:18 | |
| *** xkapastel has joined #tomu | 15:22 | |
| *** earthnative has quit IRC | 15:25 | |
| *** awe00 has quit IRC | 15:25 | |
| *** awe00 has joined #tomu | 15:46 | |
| *** earthnative has joined #tomu | 16:29 | |
| *** earthnative has quit IRC | 16:50 | |
| *** earthnative has joined #tomu | 17:30 | |
| *** xkapastel has quit IRC | 17:32 | |
| *** awe00 has quit IRC | 17:51 | |
| *** AmosSam has left #tomu | 17:52 | |
| *** AmosSam has joined #tomu | 17:52 | |
| *** AmosSam has left #tomu | 17:58 | |
| *** AmosSam has joined #tomu | 17:59 | |
| *** AmosSam has left #tomu | 18:10 | |
| *** AmosSam has joined #tomu | 18:10 | |
| *** rohitksingh has quit IRC | 18:37 | |
| *** AmosSam has left #tomu | 19:07 | |
| *** AmosSam has joined #tomu | 19:08 | |
| *** AmosSam has left #tomu | 19:18 | |
| *** AmosSam has joined #tomu | 19:18 | |
| *** awe00 has joined #tomu | 19:27 | |
| *** ovf has quit IRC | 19:58 | |
| *** ovf has joined #tomu | 19:59 | |
| *** awe00 has quit IRC | 20:07 | |
| *** awe00 has joined #tomu | 20:15 | |
| *** awe00 has quit IRC | 20:26 | |
| *** awe00 has joined #tomu | 20:35 | |
| mithro | xobs: https://github.com/elmsfu/py_ice40 | 20:47 |
| tpb | Title: GitHub - elmsfu/py_ice40: python code to configure iCE40 FPGAs (at github.com) | 20:47 |
| *** acathla has quit IRC | 21:57 | |
| *** acathla has joined #tomu | 21:57 | |
| *** AmosSam has left #tomu | 22:37 | |
| *** AmosSam has joined #tomu | 22:37 | |
| *** lathiat has quit IRC | 22:44 | |
| *** earthnative has quit IRC | 22:55 | |
| *** earthnative has joined #tomu | 23:15 | |
| *** awe00 has quit IRC | 23:53 | |
Generated by irclog2html.py 2.13.1 by Marius Gedminas - find it at mg.pov.lt!