hi, now I'm getting this error in my project in op...
# openlanei
Ivan Rodriguez
12/21/2021, 11:07 AMhi, now I'm getting this error in my project in open lane "[ERROR]: There are hold violations in the design at the typical corner", this error point me to a file with the timings. What I need to change to fix the hold violation error?, increment target density, bigger chip, etc ..
v
Vijayan Krishnan
12/21/2021, 11:12 AMpost CTS timing logs check it. How much hold violation exist?
Try
Vijayan Krishnan
12/21/2021, 11:12 AM Copy code
repair_timing -holdVijayan Krishnan
12/21/2021, 11:13 AMWhich Openlane version currently using? Try with latest version once
i
Ivan Rodriguez
12/21/2021, 11:14 AM6.76 data required time
-4.87 data arrival time
-1.90 slack (VIOLATED)
Ivan Rodriguez
12/21/2021, 11:14 AMfor the openlane version I'm using 2021.11.23_01.42.34
Ivan Rodriguez
12/21/2021, 11:15 AMhow I use the ''' repair_timing -hold'''
👀 1
Ivan Rodriguez
12/21/2021, 11:15 AM?
v
Vijayan Krishnan
12/21/2021, 11:30 AM2021.12.21_01.44.20 latest openlane tag available
Vijayan Krishnan
12/21/2021, 11:32 AMi
Ivan Rodriguez
12/21/2021, 11:39 AMok, I will try with the new openlane tag and see. Also I move some of the GPIO to be closer together, becuase the timing viloation was in the outer most GPIO pins
Ivan Rodriguez
12/21/2021, 12:27 PMwith the latest's version of open lane that you suggested I get an assertion error in yosys
Ivan Rodriguez
12/21/2021, 12:27 PMERROR: Assert `p != NULL' failed in kernel/yosys.cc:453.
v
Vijayan Krishnan
12/21/2021, 12:29 PMis this caravel based project?
i
Ivan Rodriguez
12/21/2021, 12:33 PMyes
v
Vijayan Krishnan
12/21/2021, 12:40 PMYou might try increasing PL_RESIZER_HOLD_SLACK_MARGIN
✅ 1
i
Ivan Rodriguez
12/21/2021, 12:57 PMyes, I will try that, for now I'm stuck with the "ERROR: Assert `p != NULL' failed in kernel/yosys.cc:453."
Ivan Rodriguez
12/21/2021, 12:58 PMin the ABC pass
v
Vijayan Krishnan
12/21/2021, 1:15 PM2021.11.23_01.42.34 try with this openlane tag itself
i
Ivan Rodriguez
12/21/2021, 1:47 PMyes, I re-download that tag but the error is same. I don't know why
Ivan Rodriguez
12/21/2021, 2:03 PMdoing the make test in openlane I get the same error. So I don't know what to do now.
v
Vijayan Krishnan
12/21/2021, 4:03 PMtry to install in new directory and try it.
i
Ivan Rodriguez
12/21/2021, 4:05 PMI just found the issue, xD. No space left because of the multiple images of doker. I just manage to return to the correct openlane tag
v
Vijayan Krishnan
12/21/2021, 4:05 PMyes
Vijayan Krishnan
12/23/2021, 12:35 PMwhich openlane tag resolved your hold violation issue?
i
Ivan Rodriguez
12/23/2021, 5:46 PMstill working on that, now I'm having issues to finish the routing in less than 64 steps. For now I want to try again with the new output configuration ( Not using the outer most IO ports) and if still happens I will start testing different tags to solve the issue. I will keep you posted.
Ivan Rodriguez
12/25/2021, 6:55 PMHi @User, I'm testing the setting that you mention. For now I just have a finish openlane with slew violations and hold violation. The slew were low (0.1). But for the Hold violation, I have and error popping but in the report was only one that have 0.00 as a result and that was mark as a violation. Now I'm trying again with the margings a little bit up hopping to have it now without errors.
v
Vijayan Krishnan
12/27/2021, 8:25 AMyou can ignore the message if 0.00
i
Ivan Rodriguez
12/27/2021, 9:50 AMok, thanks. I'm think I'm close to dile in the parameters.
62 Views