fix: Do not replace items annotated with builtin attrs with the attr input #9943
Conversation
Veykril
force-pushed
the
attr-repl
branch
2 times, most recently
from
e4929bd to
45d7f30
Compare
|
Interestingly this does not break the |
Veykril
changed the title
|
|
Nevermind I think we can 🎉 |
|
bors r+ |
9965: minor: Don't ask for the builtin attribute input twice r=Veykril a=Veykril `tt` and `item` here were the same, I misunderstood what the main input for attributes was in #9943 bors r+ Co-authored-by: Lukas Wirth <[email protected]>
|
This increased the memory usage on |
|
This is to be expected, as we now expand builtin attributes back to the item so basically duplicate builtin-attr annotated items into a macrofile. Might make sense to investigate whether we actually need this expansion for builtin attributes? |
10044: minor: Ignore text and bench attributes again r=Veykril a=Veykril Reverts part of #9943 Co-authored-by: Lukas Wirth <[email protected]>
This causes runnables to work for paths that actually resolve to the
testattribute now.Prior to this we replaced items annotated with builtin attributes by the attribute input instead of the item in our dummy expansion which is why the fully written path didn't work as we actually discarded the item while
testwas just ignored.Fixes #9935