Search
2:01:24
Bike
oh, i think i recognize this problem though
2:01:29
Bike
maybe i screwed up the merge
2:02:13
Bike
well, the problem i hit before was that the analyzer-generated code uses a define GC_STAMP where the scraper-generated code uses GC_ENUM
2:02:34
Bike
so i changed everything to use GC_ENUM uniformly, but if you don't do that it's missing all the STAMPWTAG_whatever definitions
2:02:57
Bike
wait, are you just using the clasp_gc_cando.cc that's in the branch? i didn't regenerate any of those files
2:03:16
Bike
(well, i might have done clasp_gc.cc, but it would be an accident)
2:03:32
Bike
i changed the gctools to use GC_ENUM instead of GC_STAMP, but clasp_gc_cando.cc still has GC_STAMP
2:11:45
drmeister
Ah - I haven't integrated anything yet.
2:12:50
drmeister
Does the scraper generate the file at this point?
2:13:07
drmeister
"not called by anything"
2:28:01
Bike
Yeah, like I said, I've been calling it and writing it into clasp_gc.cc myself
2:28:33
drmeister
Will all the info we need be in clasp_gc.cc for build_cboehm ?
2:28:55
Bike
i thought for cboehm we just wanted to use the scraper like we've been doing?
2:29:31
drmeister
So all we are doing is moving the location of clasp_gc_xxx.cc and it's always called clasp_gc.cc
3:05:18
beach
Good morning everyone!