| *** tpb has joined #tomu | 00:00 | |
| *** st-gourichon-f has joined #tomu | 00:03 | |
| *** st-gourichon-fid has quit IRC | 00:04 | |
| xobs | acathla: the wishbone-tool error looks like it's unable to open the device, or possibly it's not receiving the expected response (i.e. the target device doesn't have a debug bridge). | 02:22 |
|---|---|---|
| xobs | Blinking red usually means the CPU crashed, which means it didn't find the byte-order marker on the top of the flash image. | 02:22 |
| *** guan has quit IRC | 02:36 | |
| *** mithro has quit IRC | 02:37 | |
| *** mithro has joined #tomu | 02:38 | |
| *** guan has joined #tomu | 02:41 | |
| *** emeb has quit IRC | 03:59 | |
| CarlFK | xobs: I told someone about betrusted "verifiable hardware" , response: "does that mean we can have voting machines we can trust?" | 04:51 |
| xobs | Voting is hard :-/ | 04:52 |
| CarlFK | im sure you are up to it | 04:52 |
| CarlFK | once you betrusted is done, maybe make a larger thing that is a little more general purpose | 04:54 |
| Toba__ | voting should be done on paper | 05:31 |
| *** squelch has joined #tomu | 05:31 | |
| CarlFK | what do you think of a machine that prints out your ballet, and the human readable print out is later scanned? | 05:32 |
| Toba__ | i would have more thoughts if it wasn't half past midnight | 05:32 |
| Toba__ | im gonna go sleep | 05:32 |
| CarlFK | heh | 05:32 |
| CarlFK | see ya | 05:33 |
| *** st-gouri- has joined #tomu | 06:47 | |
| *** st-gourichon-f has quit IRC | 06:50 | |
| *** lucab has quit IRC | 08:38 | |
| *** leons has quit IRC | 08:38 | |
| *** promach3 has quit IRC | 08:38 | |
| *** jimt[m] has quit IRC | 08:38 | |
| *** nurelin has quit IRC | 08:38 | |
| *** david-sawatzke[m has quit IRC | 08:38 | |
| *** john_k[m] has quit IRC | 08:38 | |
| *** xobs has quit IRC | 08:38 | |
| *** john_k[m] has joined #tomu | 08:49 | |
| *** david-sawatzke[m has joined #tomu | 08:57 | |
| *** lucab has joined #tomu | 08:57 | |
| *** nurelin has joined #tomu | 08:57 | |
| *** xobs has joined #tomu | 08:57 | |
| *** jimt[m] has joined #tomu | 08:57 | |
| *** promach3 has joined #tomu | 08:57 | |
| *** leons has joined #tomu | 08:57 | |
| acathla | xobs, what's the command-line used to build the released foboot? --revision pvt and --with-debug usb only? | 09:20 |
| xobs | acathla: correct, adding --seed as necessary | 09:21 |
| acathla | ok, thank you. Still no way to do faster build using multiple cores? | 09:22 |
| *** im-tomu has quit IRC | 11:37 | |
| acathla | wishbone-tool seems to work now. xobs, any idea on how to debug the upload with dfu-util? How can I test the usb quality maybe? | 11:40 |
| xobs | acathla: `wishbone-tool -s random-test` will write random data to RAM and read it back. If the connection is flaky, data will come back with an error. | 11:41 |
| acathla | Oh, nice. Communication seems fine... | 11:44 |
| xobs | Are you uploading a riscv binary or an FPGA bitstream? | 11:48 |
| *** im-tomu has joined #tomu | 11:52 | |
| *** im-tomu has quit IRC | 11:56 | |
| *** im-tomu has joined #tomu | 12:24 | |
| acathla | xobs, I tried anything, it always end with a usb disconnection and red blinking LED | 12:30 |
| xobs | Even `riscv-blink`? https://github.com/im-tomu/fomu-workshop/blob/master/riscv-blink/src/main.c | 12:31 |
| tpb | Title: fomu-workshop/main.c at master · im-tomu/fomu-workshop · GitHub (at github.com) | 12:31 |
| xobs | (Note: That actually installs a USB driver that means it should still enumerate) | 12:31 |
| acathla | xobs, yes even riscv-blink fails | 12:35 |
| acathla | I had to expand the top-multiboot.bin to 2MB to program the flash with flashrom, so I added zeros, I hope it's ok. | 12:36 |
| xobs | Oh, so this isn't on Fomu, it's on something else? | 12:37 |
| acathla | Yes, sorry. It's a PCB with almost the same things as on fomu EVT | 12:40 |
| xobs | Are you sure the flash is getting written? You can poke around with `wishbone-tool`. I believe it should be at address 0x20040000 (but don't quote me on that) | 12:41 |
| acathla | Does dfu-util always write to the flash? FPGA bitstream and riscv binary? | 12:43 |
| xobs | `dfu-util` always writes to the flash. | 12:43 |
| xobs | It will write whatever you give it -- riscv binary or FPGA bitstream. | 12:44 |
| xobs | The bootloader determines which it is by looking for the bitstream sync bytes to determine if it's an FPGA bitstream, and assumes it's a binary if it can't find it. | 12:44 |
| acathla | Ok | 12:44 |
| xobs | Since it's blinking red, that means it hasn't determined it's a bitstream. | 12:44 |
| xobs | Which, if it can't write the flash, could be the source of your problems. | 12:44 |
| xobs | That might happen if, for example, you have quad mode disabled (which is the default) and the WP/ pin is... high? | 12:45 |
| acathla | Hum, IO2 and IO3 are defined in the platform, I guessed it's handled somewhere... | 12:48 |
| xobs | You definitely should see if data's actually getting written. | 12:49 |
| acathla | Value at 20040000: eeeeeeee | 12:49 |
| acathla | not good | 12:49 |
| xobs | I seem to recall the default SPI device actually gangs all four outputs together. | 12:49 |
| xobs | So one workaround would be to enable quad mode for your chip. | 12:50 |
| xobs | Fomu has that done in the factory as part of the test, so we don't need to check for it on the device. | 12:50 |
| acathla | inout wire [3:0] spiflash4x_dq in top.v, does it means quad_spi is enabled? | 12:52 |
| xobs | It means it's using a quad-enabled SPI block. | 12:52 |
| xobs | But your SPI flash chip needs to have quad mode enabled. | 12:52 |
| acathla | ah, in the fuses? | 12:53 |
| acathla | I read that somewhere | 12:53 |
| xobs | Yep | 12:53 |
| acathla | I'm still reading eeeeeee but it seems to work ! | 14:28 |
| *** emeb has joined #tomu | 15:48 | |
| *** CarlFK has quit IRC | 16:04 | |
| *** CarlFK has joined #tomu | 16:16 | |
| st-gouri- | Hi there! Have some difficuly sending "bulk" data (say, 100kb) through litex_server to a wishbone-based UART. | 16:31 |
| st-gouri- | To be clearer, PC communicates through a 3-wire UART to the design. We target registers that drive a second UART to some other device. | 16:32 |
| st-gouri- | We could not find clear documentation. First, we could not read any byte received by the second UART until we did self._wb.regs.uart_ev_pending.write(2) where 2 is a magic value found in litex/litex/soc/software/include/hw/flags.h | 16:34 |
| st-gouri- | We're accessing the wishbone bus of the design from a Python program on our PC. | 16:36 |
| xobs | You might get more help in #litex. | 16:37 |
| xobs | However, you may also want to generate the documentation for your design. | 16:37 |
| xobs | For example, the UART we use in Betrusted looks like https://ci.betrusted.io/betrusted-soc/doc/uart.html | 16:38 |
| tpb | Title: UART LiteX SoC Project documentation (at ci.betrusted.io) | 16:38 |
| st-gouri- | xobs, thanks for writing. Looking at links. | 16:38 |
| xobs | (Also, it's late here and I'm about ready to log off, sorry) | 16:38 |
| st-gouri- | okay, thanks all the more. | 16:39 |
| st-gouri- | That does shine some light. | 16:39 |
| st-gouri- | Confirms hypotheses. | 16:39 |
| xobs | Also also, if you're doing wishbone debugging stuff, you may want to look at wishbone-tool (https://github.com/litex-hub/wishbone-utils/releases/latest) which might give you more options on working with wishbone. | 16:39 |
| tpb | Title: Release v0.7.3 · litex-hub/wishbone-utils · GitHub (at github.com) | 16:39 |
| st-gouri- | We need to implement some protocol. Looks like Rust savvyness required? | 16:40 |
| st-gouri- | xobs, where on earth are you? Eastern Europe? | 16:41 |
| *** ppisati has quit IRC | 16:42 | |
| st-gouri- | Ahah! Using libeb-c from C may be a good option! Thanks xobs. | 16:44 |
| *** squelch has quit IRC | 17:21 | |
| *** wrtlprnft has quit IRC | 18:20 | |
| *** wrtlprnft has joined #tomu | 18:21 | |
| *** craigo has joined #tomu | 19:32 | |
| *** craigo has quit IRC | 20:01 | |
| *** ukembedded has quit IRC | 21:15 | |
| *** daveshah has quit IRC | 21:15 | |
| *** ukembedded has joined #tomu | 21:15 | |
| *** daveshah has joined #tomu | 21:15 | |
| *** js has quit IRC | 22:10 | |
| *** squelch has joined #tomu | 22:23 | |
| *** st-gouri- has quit IRC | 22:42 | |
| *** squelch has quit IRC | 23:00 | |
| *** squelch has joined #tomu | 23:01 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!