<@U016ULGAUNM> (<@U017E7L119N>): Some of this und...
# openroad
t
@User (@User): Some of this undoubtedly comes from magic's reading in the original LEF marked as version (something around) 5.5 and then writing out version 5.7, and not paying attention to items that might not be compatible with 5.7 (yet. . . why on Earth would they arbitrarily make some random statement not backwardly compatible?). I have corrected magic so that it does not do this any more. I'm not sure whether other scripts are recasting any of the original LEF files as version 5.6+ without scrubbing any "SOURCE" statements. I'll leave it to you to check that. Otherwise, if you update magic, then it should not generate SOURCE statements in the LEF output any more.
u
Thanks, since the pdk contains pre-made *.magic.lef view, I should probably file a ticket right? Regarding updating locally, will re building open_pdk fix it under latest magic?
t
@20Mhz: The last commit fixes the "SOURCE" statements; there is an ongoing fix regarding the "CLASS" definition for the cells that did not have a LEF file given to us by SkyWater, and there was another issue regarding the use of the "NAMESCASESENSITIVE" keyword in LEF, which like "SOURCE" was deprecated in LEF version 5.6. I committed a fix for that one today.
🙌 1
u
thanks