Hello, I'm just testing the LVS with a circuit that I know has a correct LVS. If I send this message, you can imagine the LVS didn't pass. When I check the log, I have no problem with vias and nets. The final result says: 'LVS result: Final result: Circuits match uniquely.' In checking ext.log, I saw this: _Reading "sky130_ef_sc_hd__decap_12$1"._ _Error while reading cell "$$$CONTEXT_INFO$$$" (byte position 1718): Cell "sky130_ef_sc_hd__decap_12$1" was already defined in this file._ _Error while reading cell "$$$CONTEXT_INFO$$$" (byte position 1718): Ignoring duplicate definition_ _Reading "sky130_fd_sc_hd__decap_8$1"._ _Error while reading cell "$$$CONTEXT_INFO$$$" (byte position 5514): Cell "sky130_fd_sc_hd__decap_8$1" was already defined in this file._ _Error while reading cell "$$$CONTEXT_INFO$$$" (byte position 5514): Ignoring duplicate definition_ _Reading "sky130_fd_sc_hd__decap_3$1"._ _Error while reading cell "$$$CONTEXT_INFO$$$" (byte position 9054): Cell "sky130_fd_sc_hd__decap_3$1" was already defined in this file._ _Error while reading cell "$$$CONTEXT_INFO$$$" (byte position 9054): Ignoring duplicate definition_ _Reading "sky130_fd_sc_hd__fill_1$1"._ _Error while reading cell "$$$CONTEXT_INFO$$$" (byte position 11730): Cell "sky130_fd_sc_hd__fill_1$1" was already defined in this file._ _Error while reading cell "$$$CONTEXT_INFO$$$" (byte position 11730): Ignoring duplicate definition_ _Reading "sky130_fd_sc_hd__tapvpwrvgnd_1$1"._ _Error while reading cell "$$$CONTEXT_INFO$$$" (byte position 13046): Cell "sky130_fd_sc_hd__tapvpwrvgnd_1$1" was already defined in this file._ _Error while reading cell "$$$CONTEXT_INFO$$$" (byte position 13046): Ignoring duplicate definition_ _Reading "sky130_fd_sc_hd__decap_6$1"._ _Error while reading cell "$$$CONTEXT_INFO$$$" (byte position 14724): Cell "sky130_fd_sc_hd__decap_6$1" was already defined in this file._ _Error while reading cell "$$$CONTEXT_INFO$$$" (byte position 14724): Ignoring duplicate definition_ _Reading "sky130_fd_sc_hd__fill_2$1"._ _Error while reading cell "$$$CONTEXT_INFO$$$" (byte position 17880): Cell "sky130_fd_sc_hd__fill_2$1" was already defined in this file._ _Error while reading cell "$$$CONTEXT_INFO$$$" (byte position 17880): Ignoring duplicate definition_ _Reading "sky130_fd_sc_hd__clkbuf_1$1"._ _Error while reading cell "$$$CONTEXT_INFO$$$" (byte position 19320): Cell "sky130_fd_sc_hd__clkbuf_1$1" was already defined in this file._ _Error while reading cell "$$$CONTEXT_INFO$$$" (byte position 19320): Ignoring duplicate definition_ _Reading "sky130_fd_sc_hd__dfxtp_1$1"._ _Error while reading cell "$$$CONTEXT_INFO$$$" (byte position 23140): Cell "sky130_fd_sc_hd__dfxtp_1$1" was already defined in this file._ _Error while reading cell "$$$CONTEXT_INFO$$$" (byte position 23140): Ignoring duplicate definition_ _Reading "sky130_fd_sc_hd__decap_4$1"._ _Error while reading cell "$$$CONTEXT_INFO$$$" (byte position 35272): Cell "sky130_fd_sc_hd__decap_4$1" was already defined in this file._ _Error while reading cell "$$$CONTEXT_INFO$$$" (byte position 35272): Ignoring duplicate definition_ _Reading "sky130_fd_sc_hd__buf_2$1"._ _Error while reading cell "$$$CONTEXT_INFO$$$" (byte position 38074): Cell "sky130_fd_sc_hd__buf_2$1" was already defined in this file._ _Error while reading cell "$$$CONTEXT_INFO$$$" (byte position 38074): Ignoring duplicate definition_ _Reading "vias_gen"._ _Error while reading cell "$$$CONTEXT_INFO$$$" (byte position 42562): Cell "vias_gen" was already defined in this file._ _Error while reading cell "$$$CONTEXT_INFO$$$" (byte position 42562): Ignoring duplicate definition_ Reading "Dff$1". _Error while reading cell "$$$CONTEXT_INFO$$$" (byte position 42988): Cell "Dff$1" was already defined in this file._ _Error while reading cell "$$$CONTEXT_INFO$$$" (byte position 42988): Ignoring duplicate definition_ _Reading "user_analog_project_wrapper"._ _Processing sky130_ef_sc_hd__decap_12$1_ Processing Dff$1 Do you think this is the cause of the LVS failing in the percheck process?

@Charly Meyer Looks like the

$$$CONTEXT_INFO$$$

cell has duplicate definitions of some cells that are being ignored. If you’re getting a

'LVS result: Final result: Circuits match uniquely.'

message, where is LVS failing? Could it be failing because there are missing cells in the source? Can you share your

LVS_check.log

file?

The error is mentioned in the precheck: 'ERROR LVS FAILED'. This is the result placed here: precheck_results/02_JUL_2024___14_24_17 If the LVS precheck failed, it is for a reason (not necessarily a good reason). It is important to understand why. If it is only a warning, how do I skip it?

@Charly Meyer See this message.

LVS Done.
Warning: device level LVS may be incomplete because 2 cell(s) was/were ignored: see /opt/input/myproject/test_lvs_d/precheck_results/02_JUL_2024___14_24_17/tmp/ignore.glob
LVS problem: check the following files
/opt/input/myproject/test_lvs_d/precheck_results/02_JUL_2024___14_24_17/logs/ext.log
/opt/input/myproject/test_lvs_d/precheck_results/02_JUL_2024___14_24_17/logs/lvs.log
/opt/input/myproject/test_lvs_d/precheck_results/02_JUL_2024___14_24_17/outputs/reports/lvs.report

LVS ends with a non-zero error code if there are cells that ignored. What is in

ignore.glob

?

There is something I didn't mention before. The extracted netlist gives

sky130_fd_pr__special_nfet_01v8

instead of

sky130_fd_pr__nfet_01v8

because the size of the transistors is too small. I ignored the subcircuit containing

sky130_fd_pr__special_nfet_01v8

, and the LVS showed no errors in the log, but it failed in the precheck. Then I decided to change the name of the transistor to

sky130_fd_pr__special_nfet_01v8

(in another SPICE file), and the same thing happened. The log said the circuit matched, but the LVS in the precheck failed.

_`$$$CONTEXT_INFO$$$`_ is a metadata cell generated by, I think, gdsfactory. I think there is an option when you write out GDS from klayout to not write metadata; that will help avoid confusing tools downstream (but you'll lose the metadata, so that's a one-way street).

@Charly Meyer

I ignored the subcircuit containing

sky130_fd_pr__special_nfet_01v8

, and the LVS showed no errors in the log, but it failed in the precheck.

precheck LVS finishes with no errors if the layout and the schematic match. When you use ignore cells, LVS might finish with no errors, but precheck does not consider this to be a true match (because cells were ignored).

The log says, “The following cells had property errors: {Dff$1}”

but I don’t know if this is important. Above all, I don’t know how or where to change this property.

Search the LVS report for the

Dff$1

. It should list the devices that have mismatched properties.

@Tim Edwards Thanks for the advice. I reopened and rewrote the GDS file with Magic, and now I have no errors _$$$CONTEXT_INFO$$$_

If the layout was done in klayout, then I would advise against piping through magic to get GDS. Magic may not parse the hierarchy properly and can alter the layout in ways that break the circuit. In this case, I really think it is just a matter of selecting an option in klayout to remove the metadata when writing final GDS.