03:50 <Sprite_tm> tnt: Spent yesterday fighting Yosys, seems I had one inferred latch that messed up routing that I missed over and over again.

03:51 <Sprite_tm> Got the hardware hooked up, I hope. Need to write the actual software... I assume I should be able to port ice40-playground/projects/riscv_usb/fw over and get some results out of it?

05:24 <Sprite_tm> \o/

05:24 <Sprite_tm> [147397.277158] usb 1-2: New USB device found, idVendor=1d50, idProduct=6147, bcdDevice= 0.02

05:24 <Sprite_tm> [147397.277172] usb 1-2: New USB device strings: Mfr=2, Product=3, SerialNumber=1

05:24 <Sprite_tm> [147397.277179] usb 1-2: Product: iCE40 USB Demo

05:24 <Sprite_tm> [147397.277185] usb 1-2: Manufacturer: osmocom

05:24 <Sprite_tm> [147397.277190] usb 1-2: SerialNumber: 0000000000000000

05:24 <Sprite_tm> [147397.354982] cdc_acm 1-2:1.0: ttyACM0: USB ACM device

05:25 <Sprite_tm> Now to take a stab at porting TinyUSB... with a working PoC, at least I don't have to worry about the hardware being OK.

06:24 <tnt> Sprite_tm: oh wow, that was fast ! Nice !

06:41 <Sprite_tm> tnt: Errm... does your USB hardware have a name?

06:47 <Sprite_tm> Gonna call it tntusb for now :P

06:55 <tnt> Sprite_tm: oh yeah, didn't think about that :p

06:55 <Sprite_tm> Hehe, I can imagine, naming my thing is also something that doesn't always happen when I write stuff.

06:58 <Sprite_tm> Are you ok with tntusb? If not, I can still rename it.

06:58 <tnt> Sprite_tm: no, it's as good a name as any and I can't immediately think of something else.

07:06 <Sprite_tm> tnt: Cool :)

07:42 <Sprite_tm> tnt: Hey, erm, where do you have the code to initialize eps other than ep0?

07:42 <Sprite_tm> Perhaps I'm overlooking them, but your cdc code doesn't seem to open an ep and dfu doesn't need it.

07:44 <tnt> Sprite_tm: I just pushed some code in the icepick branch.

07:45 <tnt> https://github.com/smunaut/ice40-playground/commit/97684983243550a93cf05e3e0fd18e8af2ccfeb2#diff-4b7d0c8dbfdeb8e67ba7660fe46b3afcR256

07:45 <tnt> for instance

07:46 <tnt> I gotta go now but I'll be back in a couple of hours. The 'e1' branch also has some code showing off dual buffered isochronous endpoints with 512 bytes packets.

07:47 <Sprite_tm> Nice, thanks!

07:47 <Sprite_tm> I don't quite need iso support yet; I'll hack in the requirements for msc first as I need that.

09:47 <tnt> Sprite_tm: yeah getting it to enumerate is probably the first step :p

12:44 <Sprite_tm> tnt : Whee, TinyUSB does something as well.

12:44 <Sprite_tm> [173877.367164] usb 1-2: New USB device found, idVendor=cafe, idProduct=4002, bcdDevice= 1.00

12:44 <Sprite_tm> [173877.367175] usb 1-2: New USB device strings: Mfr=1, Product=2, SerialNumber=3

12:44 <Sprite_tm> [173877.367180] usb 1-2: Product: TinyUSB Device

12:44 <Sprite_tm> [173877.367184] usb 1-2: Manufacturer: TinyUSB

12:44 <Sprite_tm> [173877.367188] usb 1-2: SerialNumber: 123456

12:44 <Sprite_tm> [173877.478115] usb-storage 1-2:1.0: USB Mass Storage device detected

12:44 <Sprite_tm> [173877.478424] scsi host1: usb-storage 1-2:1.0

12:44 <Sprite_tm> Everything but EP0 is borked, and error handling is entirely absent, but there ya go :)

12:50 <tnt> Sprite_tm: oh wow, that was pretty quick :p

12:51 <tnt> Sprite_tm: do you have the code up somewhere ?

12:54 <Sprite_tm> Sec, I'll push.

12:56 <Sprite_tm> https://github.com/Spritetm/hadbadge2019_fpgasoc/tree/master/soc/app <- most of it is in dcd_tntusb.

12:57 <Sprite_tm> Apologies in advance if you find some infuriating hack that I did you your otherwise pretty nice code.

13:01 <tnt> Sprite_tm: hehe.

13:02 <tnt> Sprite_tm: for the stall and clear_stall, I think maybe setting the 'HALT' bit in the EP type would match the API better. Setting one BD to STALL will only send a single STALL packet. Setting the HALT bit makes it respond with STALL until it's un-halted.

14:46 <Sprite_tm> tnt: Ah gotcha. That's what ctrl-f 'stall' nets ya :P

16:20 <tnt> Sprite_tm: oh, something I'm just thinking of now ... the "length" that you read off a receive buffer descriptor when it's done includes the CRC ... (so it's 2 byte added to it basically).

16:22 <tnt> (there is no need to account for it in the 'size' you reserve when allocating or prepapring it though so you can safely set it up with a max length of 64 bytes and it will happily receive 64 bytes packets, but the length written when xfer is done is 66 ...)