@User I did some manual fixes to the error

Min area of metal1 holes > 0.14um^2 (met1.7)

in my

Ghazi_SoC.gds

, it showed 0 DRC and everything looked good. However when I did

file

-->

write gds

it printed this message.

Completed 50%

35799 problems occured. See feedback entries.

1. I can load the resultant

.gds

file in Klayout and I can see some layers missing 2. The file size of the DRC clean

.gds

is greater than that of the old one 3. Does the above message mean half the design is discarded while writing the gds? 4. Where do I look for feedback entries, does it create any log file somewhere?

No, you are not missing 50% of the design; that was just a progress marker. The command "select top cell ; feedback why" will tell you what the errors are. You shouldn't get errors on a standard cell layout, so it would be helpful to know what the error messages are. What layers are missing in the klayout view? That would also be a clue.

pwell.pin -122/16

nwell.pin -64/16

metal1.pin -68/16

These three layers

That's not a serious issue, because the pin layers should be coincident with drawing layers. There is definitely a different in the way magic handles labels and pins that I need to look into, but it should not be a problem in the layout. It still remains to figure out whether the GDS write issues are important or not.

@Tim Edwards Do I need to be concerned about this parent child disagreement? [That's a funny terminology by the way :)]

@Wajeh ul hasan: Sometimes this is an issue and sometimes not. It is related to the way that magic automatically generates implant layers; first each cell has GDS generated and output individually, and then magic checks how the GDS in cells interacts with other cells, and attempts to fix any problems it finds. If the generation rules act differently on the full flattened layout than they do for the individual cells, magic will output this error message. It can be very hard to tell what the problem is, or if there is a problem. The best thing to do is to do "feedback find" to get the location of some of the errors, and then look at them in klayout. If you have a standard cell design, then the best thing to do is to use "cif *hier write disable" before doing "gds write", because the standard cells are designed to be placed without causing bad interactions, so the hierarchical checks can be skipped. If part of your layout is analog, then that's not really a good option to use.