https://open-source-silicon.dev logo
Join Slack
Powered by
A little confused by the net order inside this par...
# caravan
c
A little confused by the net order inside this particular cell... It seems A is corresponded to Y?
t
The order in which pins are written to the output is not meaningful. I did add some code to sort them, which seems to not be running or working in this case, but the underlying results in netgen are just unsorted. I have not seen the "match locally but not globally" message for a long time, since some changes I made about half a year ago. What is the version of netgen (per Weston's post above yours, I'm also looking at an issue that was caused by the most recent netgen update, which I will be investigating today).
c
Thanks for the reply, I am using the Netgen 1.5.159. @Tim Edwards
t
That version is half a year and over 25 commits out of date. Which corresponds to my statement above about not having seen that particular error for over half a year.
c
I see
Thanks!
t
I believe that if you update, it will start printing the side-by-side pins in matching order A->A, Y->Y (which you will probably not see because the error will go away).
c
I have updated it to 1.5.186 on server and tried again. It is wired that the filler cell and tap cell is counted as mismatch devices.. Although I have passed the netgen_setup.tcl to netgen in the command as I did previously.
Here is the screenshot of the error and lvs_results.log @Tim Edwards
Copy code
Subcircuit summary:
Circuit 1: fine_freq_track                 |Circuit 2: fine_freq_track                 
-------------------------------------------|-------------------------------------------
sky130_fd_sc_hs__fill_2 (1)                |(no matching element)                      
sky130_fd_sc_hs__fill_1 (1)                |(no matching element)                      
sky130_fd_sc_hs__fill_8 (1)                |(no matching element)                      
sky130_fd_sc_hs__inv_4 (67)                |sky130_fd_sc_hs__inv_4 (67)                
sky130_fd_sc_hs__fill_4 (1)                |(no matching element)                      
sky130_fd_sc_hs__tapvpwrvgnd_1 (1)         |(no matching element)                      
sky130_fd_sc_hs__fa_2 (12)                 |sky130_fd_sc_hs__fa_2 (12)                 
sky130_fd_sc_hs__dfrtp_4 (46)              |sky130_fd_sc_hs__dfrtp_4 (46)              
sky130_fd_sc_hs__nor2b_1 (24)              |sky130_fd_sc_hs__nor2b_1 (24)              
sky130_fd_sc_hs__dfxtp_4 (6)               |sky130_fd_sc_hs__dfxtp_4 (6)               
sky130_fd_sc_hs__dfrbp_1 (24)              |sky130_fd_sc_hs__dfrbp_1 (24)              
sky130_fd_sc_hs__a32oi_1 (5)               |sky130_fd_sc_hs__a32oi_1 (5)               
sky130_fd_sc_hs__a22o_1 (15)               |sky130_fd_sc_hs__a22o_1 (15)               
sky130_fd_sc_hs__nor2_1 (60)               |sky130_fd_sc_hs__nor2_1 (60)               
sky130_fd_sc_hs__nand4_1 (3)               |sky130_fd_sc_hs__nand4_1 (3)               
sky130_fd_sc_hs__o211ai_1 (5)              |sky130_fd_sc_hs__o211ai_1 (5)              
sky130_fd_sc_hs__a21oi_1 (61)              |sky130_fd_sc_hs__a21oi_1 (61)              
sky130_fd_sc_hs__nand2_1 (59)              |sky130_fd_sc_hs__nand2_1 (59)              
sky130_fd_sc_hs__nand2_4 (8)               |sky130_fd_sc_hs__nand2_4 (8)               
sky130_fd_sc_hs__o21a_1 (39)               |sky130_fd_sc_hs__o21a_1 (39)               
sky130_fd_sc_hs__xnor2_1 (7)               |sky130_fd_sc_hs__xnor2_1 (7)               
sky130_fd_sc_hs__inv_2 (4)                 |sky130_fd_sc_hs__inv_2 (4)                 
sky130_fd_sc_hs__nand2_2 (9)               |sky130_fd_sc_hs__nand2_2 (9)               
sky130_fd_sc_hs__and2_2 (2)                |sky130_fd_sc_hs__and2_2 (2)                
sky130_fd_sc_hs__a22oi_1 (13)              |sky130_fd_sc_hs__a22oi_1 (13)              
sky130_fd_sc_hs__nor3_1 (7)                |sky130_fd_sc_hs__nor3_1 (7)                
sky130_fd_sc_hs__a31oi_2 (1)               |sky130_fd_sc_hs__a31oi_2 (1)               
sky130_fd_sc_hs__o31ai_1 (4)               |sky130_fd_sc_hs__o31ai_1 (4)               
sky130_fd_sc_hs__o21ai_1 (5)               |sky130_fd_sc_hs__o21ai_1 (5)               
sky130_fd_sc_hs__nand2b_1 (4)              |sky130_fd_sc_hs__nand2b_1 (4)              
sky130_fd_sc_hs__a211oi_1 (3)              |sky130_fd_sc_hs__a211oi_1 (3)              
sky130_fd_sc_hs__dfxtp_1 (3)               |sky130_fd_sc_hs__dfxtp_1 (3)               
sky130_fd_sc_hs__nand3_2 (2)               |sky130_fd_sc_hs__nand3_2 (2)               
sky130_fd_sc_hs__o22ai_1 (6)               |sky130_fd_sc_hs__o22ai_1 (6)               
sky130_fd_sc_hs__dfxtp_2 (2)               |sky130_fd_sc_hs__dfxtp_2 (2)               
sky130_fd_sc_hs__o2bb2ai_1 (1)             |sky130_fd_sc_hs__o2bb2ai_1 (1)             
sky130_fd_sc_hs__inv_1 (1)                 |sky130_fd_sc_hs__inv_1 (1)                 
sky130_fd_sc_hs__dfrtn_1 (25)              |sky130_fd_sc_hs__dfrtn_1 (25)              
sky130_fd_sc_hs__a222o_1 (1)               |sky130_fd_sc_hs__a222o_1 (1)               
sky130_fd_sc_hs__dfstp_2 (4)               |sky130_fd_sc_hs__dfstp_2 (4)               
sky130_fd_sc_hs__maj3_1 (2)                |sky130_fd_sc_hs__maj3_1 (2)                
sky130_fd_sc_hs__o21ai_2 (2)               |sky130_fd_sc_hs__o21ai_2 (2)               
sky130_fd_sc_hs__nor2b_4 (2)               |sky130_fd_sc_hs__nor2b_4 (2)               
sky130_fd_sc_hs__dfsbp_2 (1)               |sky130_fd_sc_hs__dfsbp_2 (1)               
sky130_fd_sc_hs__nor2_4 (2)                |sky130_fd_sc_hs__nor2_4 (2)                
sky130_fd_sc_hs__nor4_1 (2)                |sky130_fd_sc_hs__nor4_1 (2)                
sky130_fd_sc_hs__nor2_2 (2)                |sky130_fd_sc_hs__nor2_2 (2)                
sky130_fd_sc_hs__clkbuf_2 (2)              |sky130_fd_sc_hs__clkbuf_2 (2)              
sky130_fd_sc_hs__sdlclkp_1 (1)             |sky130_fd_sc_hs__sdlclkp_1 (1)             
sky130_fd_sc_hs__o2bb2ai_2 (1)
t
@Can Wang: Because of something the OpenLane developers added, it is necessary to set the environment variable "MAGIC_EXT_USE_GDS" before running netgen to get it to ignore the tap and fill cells. As for the segmentation fault, I don't know why it segfaults, but I consider that an error that needs attention, so if you can post the two netlists that caused it, I'd appreciate it.
c
Gotcha! @Tim Edwards Thanks for taking the time to help with us!
design_extracted.spicedesign_lvs.spice
Oh after 
EXPORT MAGIC_EXT_USE_GDS=1
the lvs passed. The seg fault also disappears. @Tim Edwards
t
Good to know!
3 Views
Open in Slack
PreviousNext
Powered by