Hello! I am getting this error for fastroute in op...
# openlane
p
Hello! I am getting this error for fastroute in openlane, what is possibly the issue? Is it because the routing is too complex to finish?
m
you need to find out why the process is getting killed. (3rd line). You can ignore everything after because those errors are all due to the process getting killed unexpectedly. Most often the reason why this happens is running out memory. So watch memory usage with a program like top, and after the process gets killed, check the kernel log (/var/log/kern.log) for messages about oomkiller (that's the process that will kill processes that are using too much memory);.
p
Thank you for your reply! I am not able to find the /var/log/kern.log, probably because I am using docker. But I check the last log file in the process. this is 11-fastroute.log and I don't see error, it is like the process suddenly stop at iteration 24
m
Use top to track memory as it crashes
p
@Matt Venn @mehdi Hello! I hope you are having a wonderful weekend. I am writing to ask some question about the drc error on sky130_sram_4kbyte_1rw1r_32x1024_8. I am encountering drc error on it. The drc error is about met2 and met3, which is in the first and second figures. I pick up a drc error on met3 in the third figure. This error is not in the bit cell of the sram but in the peripheral circuit of the sram. I remember you once told me that this sram is drc clean. So I want to ask can I ignore this error or am I doing something wrong? The fourth figure is the git log of the sky130_sram_macros
m
@Matthew Guthaus
m
You have old version.
I am out today but can check if it's a mistake and the new one isn't in the repo. You can look at our test chip for the latest ones
m
Thanks @Matthew Guthaus. @jeffdi @mkk we would need to update to the latest one
p
@Matthew Guthaus Hello, I pull from sky130_sram_macros yesterday so I am sure I am using the latest version. Could you please share with me the new sram that is not in the repo? Thank you!
m
@Peijun Houmeanwhile take a look at: https://github.com/VLSIDA/openram_testchip
p
Yes, I am looking at it.
👍 1
It seems that the sky130_sram_4kbyte_1rw1r_32x1024_8.gds.gz in this github also does not work for this met3 error also appears in other version @mehdi @jeffdi @Matthew Guthaus
m
I'm have to look tomorrow. I remember specifically fixing that
m
@Matthew Guthaus That would be really helpful! We also have some density issues inside the memory. @Peijun Hou maybe you could show that on here?
p
Yes, I suspect in the lower left of the sram the diffusion area and tap is not enough so there is a FOM error of drc. the link shows the error https://skywater-pdk.slack.com/files/U016HSAA3RQ/F02BBHC4XLJ/klayout_drc_fom.log
m
That should be reported to Tim to fix in the Magic DRC deck then...
@Peijun Hou You are using an old version. I can confirm that it is fixed:
I am using this commit in the dev branch: commit b32280e1ae2226a26c18fb035aaeda6fa17cff29 (HEAD -> dev, origin/dev) Author: mrg <mrg@ucsc.edu> Date: Wed Jul 28 083251 2021 -0700 Fix MSB/LSB bug in lib files
You said that you were too though. So I'm not sure what is going on...
Which repo are you using? git@github.com:efabless/sky130_sram_macros.git ?
m
@Matthew Guthaus it has the same issue on @Peijun Hou's screenshot. @Peijun Hou please double check.
m
@mehdi His screenshot doesn't correspond to the version in the repo
Or even the one in the openram_testchip repo...
m
Thanks @Matthew Guthaus. @Peijun Hou please double check. I am checking as well
p
Yes, it is the version problem I think
Thanks for your help Matthew Guthanus, I am making a mistake for the sram.
m
No problem. I think you may still run into that density error. Nothing was changed about that, but also they didn't complain during the tape in