hi , I am not sure how the obstruction layers and ...
# magic
r
hi , I am not sure how the obstruction layers and metal layers are getting generated in LEF file ....I see my pins going inside the OBS layer , which will make the routing access to the pins not feasible
t
There are several solutions to this. The easiest one is just to make sure that all your pins are on the cell boundary. For other solutions, see the documentation on magic's
lef write
command.
r
yes the pins are on the boundary... Let me try the distance option -- Option -hide may take an additional value argument distance which will hide the central part of a cell starting a distance distance from the edge
@Tim Edwards - not getting to what to do to make the pins accessible.. I used below option , and now the pins seems clipped / shape shifted...
t
Can you please post the layout? If the pins are really on the cell boundary, then there should not be a metal obstruction layer blocking them.
r
@Tim Edwards PFA
It is taking the max and min boundaries while calculating the obs
t
The pins at the bottom left extend to X = -5.87um, but the pins at the top extend to only X = -3.26um. So the simplest solution is just to move the top pins out to -5.87um or move the bottom pins in to -3.26um so that they are all aligned with the same edge. You could also find the obstruction rectangle in the LEF file that is blocking the pins and remove it manually. A third solution is not to use the
-hide
option, since that automatically assumes the whole layout is a rectangle. You'll get a more complicated LEF view, but since it's a small analog block, it shouldn't be a problem.
r
If i move the pins , still the obs* layers are covering...I am removing manually now , seems to work ! Thanks