Thursday, 2019-01-10

*** tpb has joined #symbiflow00:00
*** citypw has quit IRC02:31
*** citypw has joined #symbiflow03:52
*** citypw has quit IRC04:23
digshadownats`: have some more time now. Let me know if you still have any qs09:12
digshadowlitghost is a better contact for 07x series, but I can try to answer qs09:12
*** citypw has joined #symbiflow10:11
*** citypw has quit IRC10:21
nats`I'll write the splitted version this evening to test it I think it could work correctly and I'll also try with 2018 to see if the leak is still here10:40
digshadownats`: FYI we are locked to 2017.2 unless something really critical comes pu10:48
*** storm_ has quit IRC10:53
nats`yep remember that :)10:53
nats`but I wanted to see if I can run it on 201810:54
nats`just you know to know :D10:54
*** sxpert has quit IRC12:17
*** sxpert has joined #symbiflow12:57
*** igorangst has joined #symbiflow13:28
digshadowlitghost: I'm fixing / reviewing IOB_SIGN issue I saw a while back13:39
digshadowtentatively, looks like this will result in a *negative tilegrid offset*13:40
digshadowdo you have any fundamental objections to this?13:40
digshadowthe issue is that only half of the normal tile is present, both as displayed in the GUi and in the bitstream13:40
digshadowas the normal tile is 4 words high, but we only have the Y1 case13:41
digshadowI fully expect this to at least break some of our tools on prints13:41
*** igorangst has quit IRC14:01
*** igorangst has joined #symbiflow14:22
*** efg has joined #symbiflow14:53
*** igorangst has quit IRC15:38
litghost> tentatively, looks like this will result in a *negative tilegrid offset* -> Why not base the tile at the lower address, and shift everybody else up?15:56
litghostUltimately I don't actually care about a negative offset, as long as does the right thing, and xc7patch accepts the input15:57
litghostIf you can round trip (bit -> fasm -> bit -> fasm) with no errors, than I'm good15:58
digshadowlitghost: its the word offset, not the tile address16:17
digshadowwe're out of things to go back to16:18
litghostA negative word offset just causes the frame address to drop below the base address, so shifting the base address down would resolve that.17:13
digshadowlitghost: hmm not sure I agree with that17:24
digshadowit would at the very least introduce a lot more special cases than it would solve17:24
digshadowin any case, I just tested with segprint and the iob fuzzer17:24
digshadowand both seemed happy with it17:25
digshadowI'll put the details in the PR for you to review17:25
digshadowpush back if you think it will be problematic17:25
digshadowcreated PR:
tpbTitle: Iob sign by mcmasterg · Pull Request #488 · SymbiFlow/prjxray · GitHub (at
*** swetland has joined #symbiflow19:13
*** Rahix has joined #symbiflow20:15
nats`litghost digshadow I'm starting testing my parallel 072 fuzzer20:26
nats`it seems I have something working20:39
nats`I'll run it long time to test if someone is motivated I can make a zip of that dirty code and you can try20:39
*** igorangst has joined #symbiflow20:43
nats` <= screencap of fuzzer running :)20:44
nats` <= change of block seems to help !20:45
nats`maybe it's just a feeling but it seems to go faster too20:50
nats`seems 1/3rd in 10 minutes oO20:51
nats`there is a problem I think20:51
*** Rahix has quit IRC21:02
nats`my fuzzer ended work in 30 minutes....21:23
nats`I'm not sure it's good :p21:23
nats`I get 2.7GB downhill file and a 3.6GB uphill file21:30
nats`is it ok ?21:30
nats`anyway if it's good it means we can generate fuzzer 072 in 30 minutes with not eating more than 8GB21:31
nats`and it can be even more optimized with more parallel block if more ram is available21:31
litghostnats: Sounds great21:35
litghostThat size looks about right21:35
nats`litghost, how could we check diff between results file ?21:35
nats`do you have them ?21:36
nats`I could download them pretty fast I have FFTH21:36
litghostAssuming the files were emitted in order, lets start with sha1sum21:36
nats`ahhhhhh you know you're smart ? :D21:36
litghostIf we are lucky, they'll just match21:36
* nats` dumb21:36
nats`anyway where can I get original one because I can't run 072 fuzzer on my computer21:37
litghostI was planning to just run it on my machine21:37
nats`someone must have them no ?21:38
nats`2ade66e56811d9753e48d51aa0fdcc1692e5d1ef  uphill_wires.txt21:38
nats`I concated all file together21:38
litghost072 gets consume in 074 and only tileconn.json is output21:39
litghost072 generates input data21:39
nats`ce520ef646faf050651189290e1128818c19d426  downhill_wires.txt21:39
litghostIt doesn't directly output to the database21:39
nats`yep I saw that21:39
nats`I was thinking someone would have kept them21:39
nats`should I make a clean patch and put a PR or zip it for you or someone else to test it ?21:40
nats`uhhmmm wait21:41
nats`maybe I can feed them to 07421:41
nats`we could compare output21:41
nats` <= you can have the 3 needed fuzzer file here21:51
tpbTitle: Index of /~nats/symbi/ (at
nats`if you want to run it put it in a folder because it'll generate a huge mess21:51
nats`and you need to be in the prjxray env21:51
*** Ultrasauce has quit IRC22:30
*** Ultrasauce has joined #symbiflow22:37
nats`oky I adapted it in fuzzer 072 with good build dir etc22:44
nats`I just started it one last time to see if it "works"22:44
nats`I'll make the PR after that and someone will have to compare output file22:44
litghostIt's worth mentioning that if your version of 072 completes in 30 minutes, that is actually faster than the old version on my machine22:44
litghostit still hasn't completed, and has been running for more than 1 hr22:45
nats`can you get my fuzzer and run it too ?22:46
nats`by the way I have a server with a core i3 on it if needed to deploy a build system22:46
nats`maybe as the leak becomes huge the vivado process slows down22:47
nats`it would be really possibl22:47
nats`already at 20% !22:50
nats`in about 8 minutes22:50
nats`I launched it with time command like you recommended22:50
nats`don't forget I run it on 4 processor too :D22:53
nats`it may explains why it goes faster22:53
*** igorangst has left #symbiflow22:55
nats`about 4 minutes per 10% :)22:59
*** vup2 has quit IRC23:04
*** vup2 has joined #symbiflow23:06
nats`Work done !23:15
nats`touch build/specimen_001/OK23:15
nats`real    31m49,181s23:15
nats`user    104m14,978s23:15
nats`sys     17m32,685s23:16
nats`not bad23:16
nats`and it generates the two file in the end23:16
nats`I push on my fork and PR but it'll be a PR to be checked by someone23:16
*** vup2 has quit IRC23:22
*** vup2 has joined #symbiflow23:22
tpbTitle: Fixing Bug #171 and #465 by natsfr · Pull Request #489 · SymbiFlow/prjxray · GitHub (at
nats`pull request made23:31
nats`I'll write the bug explanation in bed I'm sleepy :)23:32
nats`good night folks23:32
*** vup2 has quit IRC23:52
*** vup2 has joined #symbiflow23:55

Generated by 2.13.1 by Marius Gedminas - find it at!