@Tim Edwards I am trying to open an sram macro from here: https://github.com/google/globalfoundries-pdk-ip-gf180mcu_fd_ip_sram/tree/main/cells/gf180mcu_fd_ip_sram__sram128x8m8wm1 But I am getting the following error on magic:

Error while reading cell "saout_wm1_x4_128x8m81" (byte position 1681042): Warning:  Cell saout_wm1_x4_128x8m81 boundary was redefined.
Error while reading cell "saout_wm1_x4_128x8m81" (byte position 1681042): Warning:  Cell saout_wm1_x4_128x8m81 boundary was redefined.
Error while reading cell "saout_wm1_x4_128x8m81" (byte position 1681042): Warning:  Cell saout_wm1_x4_128x8m81 boundary was redefined.

Magic version:

8.3.326

Open_pdk commit:

41e2da101560e

gf180mcu_fd_ip_sram commit:

9c411928870ce15

I am still working out what constitutes a "boundary layer" in this process. GF uses (0, 0), (63, 63), and (63, 0) (the last of which is used in vendor GDS but not documented anywhere). In general, you can ignore that error, but eventually I will figure out and correct how magic interprets those layers. Tentatively, after talking with Matt Guthaus about the same SRAM, I've determined that the (0, 0) layer surrounds the extent of geometry, while the (63, 63) layer (63, 0) (?) is the abutment box (PR boundary). It was only last week that I was looking at it, and I haven't put that understanding into practice in the tech file yet.

Alright thanks