| commit | c7c492fea07982e0142d0e8840ec9c774e47395c | |
| author | Douglas Katzman <dougk@google.com> | |
| Fri, 2 Dec 2016 22:32:58 +0000 (2 17:32 -0500) | ||
| committer | Douglas Katzman <dougk@google.com> | |
| Fri, 2 Dec 2016 22:55:31 +0000 (2 17:55 -0500) | ||
| tree | 5864b6a986c0de0646ca34f79e9fda5077cd4b2f | treesnapshot (tar.gz zip) |
| parent | 7cb250af6bd99c2ee2d3b6007d25b7611e2042f8 | commitdiff |
Deobfuscate pointer preservation in gencgc a tiny bit.
Too many names expressed nearly the same idea, and some weren't memorable:
- valid_lisp_pointer_p (this one remains as is).
- valid_conservative_root_p becomes "conservative_root_p"
because I don't know what an "invalid" conservative root is.
- looks_like_valid_lisp_pointer_p becomes "properly_tagged_descriptor_p"
whereas "improperly" tagged ones have wrong lowtag bits for a widetag.
- possibly_valid_dynamic_space_pointer_s is gone,
being wholly absorbed absorbed into conservative_root_p.
Too many names expressed nearly the same idea, and some weren't memorable:
- valid_lisp_pointer_p (this one remains as is).
- valid_conservative_root_p becomes "conservative_root_p"
because I don't know what an "invalid" conservative root is.
- looks_like_valid_lisp_pointer_p becomes "properly_tagged_descriptor_p"
whereas "improperly" tagged ones have wrong lowtag bits for a widetag.
- possibly_valid_dynamic_space_pointer_s is gone,
being wholly absorbed absorbed into conservative_root_p.
| src/runtime/gc-common.c | diffblobblamehistory | |
| src/runtime/gc-internal.h | diffblobblamehistory | |
| src/runtime/gencgc.c | diffblobblamehistory |