open-source-silicon.dev

image.png

Hi, I have run into a strange issue with the XOR check.  I directly integrated the empty_analog_project_wrapper.gds directly into my design and connected the internal pins to it.  The XOR check fails, and the resulting GDS appears to show a shrink and offset of the system.  Any guidance on how to debug this?  I've looked through a number of answers and don't see any obvious steps.

(I manually added the ruler for scale)

`Reading file /prj/sky_130/taurus/gds/user_analog_project_wrapper.gds for cell user_analog_project_wrapper`
`dbu=0.001`
`cell user_analog_project_wrapper dbu-bbox(ll;ur)=(-11710,86275;2916290,3614275)`
`cell user_analog_project_wrapper dbu-bbox(left,bottom,right,top)=(-11710,86275,2916290,3614275)`
`cell user_analog_project_wrapper dbu-size(width,height)=(2928000,3528000)`
`cell user_analog_project_wrapper micron-bbox(left,bottom,right,top)=(-11.71,86.275,2916.29,3614.275)`
`cell user_analog_project_wrapper micron-size(width,height)=(2928.0,3528.0)`
`Done.`

`Magic 8.3 revision 420 - Compiled on Fri Sep  8 02:05:00 UTC 2023.`
`Starting magic under Tcl interpreter`
`Using the terminal as the console.`
`Using NULL graphics device.`
`Processing system .magicrc file`
`Sourcing design .magicrc for technology sky130A ...`
`2 Magic internal units = 1 Lambda`
`Input style sky130(): scaleFactor=2, multiplier=2`
`The following types are not handled by extraction and will be treated as non-electrical types:`
    `ubm`
`Scaled tech values by 2 / 1 to match internal grid scaling`
`Loading sky130A Device Generator Menu ...`
`Loading "/home/vlad/mpw_precheck/checks/xor_check/erase_box.tcl" from command line.`
`Input style sky130(vendor): scaleFactor=2, multiplier=2`
`CIF input style is now "sky130(vendor)"`
`Warning: Calma reading is not undoable!  I hope that's OK.`
`Library written using GDS-II Release 6.0`
`Library name: LIB`
`Reading "nmos_CDNS_6942952479347".`
`CIF file read warning: CIF style sky130(vendor): units rescaled by factor of 5 / 1`

standard.PNG

offset.PNG

Figured out the issue.  Despite the strange area ID, this was simply an issue with cell location.  Centering the empty_analog_project_wrapper cell at (0,0) was the solution.

I figured this out by looking at the XOR pins, and noticed that my version was erased 7.71 further right than desired.  Sure enough, that was the xoffset of the cell.