| *** tpb has joined #yosys | 00:00 | |
| *** kristianpaul has quit IRC | 00:01 | |
| *** kristianpaul has joined #yosys | 00:02 | |
| *** lf has quit IRC | 00:13 | |
| *** lf has joined #yosys | 00:13 | |
| *** FL4SHK has quit IRC | 02:15 | |
| *** FL4SHK has joined #yosys | 02:16 | |
| *** FFY00 has quit IRC | 03:32 | |
| *** FFY00 has joined #yosys | 03:32 | |
| *** Degi has quit IRC | 04:19 | |
| *** Degi has joined #yosys | 04:19 | |
| *** jfcaron has quit IRC | 05:19 | |
| *** FL4SHK has quit IRC | 06:52 | |
| *** emeb_mac has quit IRC | 06:54 | |
| *** FL4SHK has joined #yosys | 06:55 | |
| *** stzsch has quit IRC | 07:00 | |
| *** kraiskil has joined #yosys | 07:15 | |
| *** srk has quit IRC | 07:36 | |
| *** elGamal has joined #yosys | 07:49 | |
| *** srk has joined #yosys | 08:02 | |
| *** FL4SHK has quit IRC | 08:18 | |
| *** FL4SHK has joined #yosys | 08:18 | |
| *** citypw has joined #yosys | 08:43 | |
| *** kraiskil has quit IRC | 08:57 | |
| *** kraiskil has joined #yosys | 08:59 | |
| *** kraiskil has quit IRC | 10:11 | |
| *** gmc has quit IRC | 10:18 | |
| *** srk has quit IRC | 10:34 | |
| *** srk has joined #yosys | 10:35 | |
| *** citypw has quit IRC | 11:06 | |
| *** citypw has joined #yosys | 11:09 | |
| *** citypw has quit IRC | 11:16 | |
| *** citypw has joined #yosys | 11:19 | |
| *** vidbina_ has joined #yosys | 11:28 | |
| *** citypw has quit IRC | 11:46 | |
| *** citypw has joined #yosys | 11:52 | |
| *** s_frit_ has joined #yosys | 12:01 | |
| *** s_frit has quit IRC | 12:02 | |
| *** kraiskil has joined #yosys | 12:17 | |
| *** kraiskil has quit IRC | 13:00 | |
| *** show1 has quit IRC | 13:49 | |
| *** citypw has quit IRC | 14:45 | |
| *** kraiskil has joined #yosys | 14:50 | |
| *** lansiir has quit IRC | 15:15 | |
| *** s_frit has joined #yosys | 15:16 | |
| *** lansiir has joined #yosys | 15:16 | |
| *** vidbina_ has quit IRC | 15:17 | |
| *** s_frit_ has quit IRC | 15:18 | |
| *** vidbina_ has joined #yosys | 15:43 | |
| *** kraiskil has quit IRC | 16:14 | |
| *** FFY00 has quit IRC | 16:49 | |
| *** FFY00 has joined #yosys | 16:50 | |
| *** srk has quit IRC | 16:52 | |
| *** emeb has joined #yosys | 16:53 | |
| *** show has joined #yosys | 17:04 | |
| *** kraiskil has joined #yosys | 17:11 | |
| *** srk has joined #yosys | 17:20 | |
| *** moony has quit IRC | 17:41 | |
| *** moony has joined #yosys | 17:41 | |
| *** jfcaron has joined #yosys | 18:16 | |
| *** FL4SHK has quit IRC | 18:39 | |
| *** FL4SHK has joined #yosys | 18:39 | |
| *** vidbina_ has quit IRC | 19:30 | |
| *** peepsalot has quit IRC | 19:54 | |
| *** peepsalot has joined #yosys | 20:17 | |
| *** emeb_mac has joined #yosys | 20:43 | |
| *** peepsalot has quit IRC | 20:51 | |
| *** peepsalot has joined #yosys | 20:53 | |
| *** kraiskil has quit IRC | 21:00 | |
| awygle | what kind of error does `ERROR: Failed to pack flipflop 'pin_ddr2_0__dq.U$$32' with 'syn_useioff' set into IOLOGIC.` from nextpnr-ecp5 indicate? | 21:39 |
|---|---|---|
| awygle | like obviously there's something wrong with the pin mapping/constraints but an idea of what to look at would be very helpful if somebody has one | 21:39 |
| tnt | You asked for a FF to be packed in the IO block and it's unable to do that. | 21:43 |
| awygle | is there a way i can get more information about why it was unable to do so? | 21:47 |
| tnt | Not that I know. | 21:52 |
| tnt | Basically open the .json file, find that FF, and how it's configured and try to figure out why it wouldn't be packable. | 21:53 |
| tnt | Where is that design from ? | 21:54 |
| *** peeps[zen] has joined #yosys | 21:55 | |
| awygle | i wrote it | 21:55 |
| awygle | it's a 4x DDR output | 21:56 |
| awygle | so not hugely surprising that i configur3ed it wrong | 21:56 |
| awygle | but i wrote it in nmigen so there's a fair bit of stuff between me and waht's actually happening | 21:56 |
| awygle | which is why i was hoping to get a few more pointers from nextpnr et al | 21:56 |
| *** peeps[zen] has quit IRC | 21:57 | |
| *** peepsalot has quit IRC | 21:57 | |
| *** peepsalot has joined #yosys | 21:58 | |
| daveshah | Tristate registers and 4:1 aren't supported at all | 22:06 |
| awygle | oh. well... that's probably why then | 22:06 |
| daveshah | Diamomd doesn't support bidirectional 4:1 at all. nextpnr supports it based on empirical testing on a best effort basis | 22:06 |
| awygle | diamond has those weird primitives for memory specifically, can i use those? | 22:07 |
| daveshah | Yes, you have to use those | 22:07 |
| tnt | but you need all the DQS stuff for those. | 22:07 |
| awygle | and based on looking at the code nmigen doesn't automagic them | 22:07 |
| awygle | which makes sense i spose | 22:07 |
| tnt | it's not a generic 4:1 | 22:07 |
| daveshah | The choice is basically use 2:1 IO, use 4:1 IO with nextpnr only or use 4:1 with the DQS stuff for both tools | 22:08 |
| daveshah | I don't know why Diamond doesn't support 4:1 bidirectional, it is possible that the hardware is subtly broken | 22:08 |
| awygle | i mean i am only using nextpnr | 22:08 |
| daveshah | I know it works well enough in nextpnr for Greg to get a HyperRAM core working though | 22:08 |
| daveshah | Yeah but sometimes it's useful to test in Diamond too | 22:09 |
| awygle | it does seem pretty dumb that diamond doesn't support generic 4:1 bidir but :shrug: i guess | 22:09 |
| tnt | And in nextpnr, how is the tristate controlled when using 4:1 ? is it packed ? | 22:09 |
| daveshah | No, it isn't | 22:10 |
| daveshah | From what I know of the hardware I doubt a tristate reg with 4:1 IO but I haven't tested, right now it just errors | 22:10 |
| awygle | so i need to do dir='-' in nmigen and manually glue the Instances of the memory primitives together, then? | 22:11 |
| daveshah | Afraid so | 22:11 |
| tnt | yeah, I guess it'd be like DDR when you only control it at clk_1x, but for DDR you can still pack it, so at least it toggles "in-sync". | 22:11 |
| awygle | mk, that's fine | 22:11 |
| awygle | thanks for letting me know | 22:11 |
| *** modwizcode has quit IRC | 22:37 | |
| *** emeb has left #yosys | 22:38 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!