Only retain no_mangle static symbols across LTO #29676
Conversation
|
Hm ideally this would have the same logic as the function that actually decides the symbol name which not only handles |
|
Oh! For some reason I thought |
arcnmx
force-pushed
the
static-extern
branch
from
cc8d548
to
70b49d8
Compare
|
@alexcrichton I made changes to account for |
| @@ -312,6 +312,15 @@ pub fn find_export_name_attr(diag: &SpanHandler, attrs: &[Attribute]) -> Option< | |||
| }) | |||
| } | |||
|
|
|||
| pub fn contains_extern_indicator(attrs: &[Attribute]) -> bool { | |||
| contains_name(attrs, "no_mangle") || | |||
| contains_name(attrs, "export_name") || | |||
There was a problem hiding this comment.
This may want to use find_export_name_attr instead to consolidate the logic for the detection of "export_name".
There was a problem hiding this comment.
The only thing is that method takes a SpanHandler since it tries to actually parse the inner value. Should I add that to this method signature?
There was a problem hiding this comment.
Yeah that's fine to just add another argument.
|
☔ The latest upstream changes (presumably #30043) made this pull request unmergeable. Please resolve the merge conflicts. |
|
What's the status of this PR? |
|
I've been busy, but it's on my todo list once I get some space to actually be able to compile things again. |
|
Closing due to inactivity for now, but feel free to resubmit with a rebase! |
|
o I was going to work on it today >_> |
|
Oh no worries! I'll take a look at #30830 |
Follow-up to #29656
r? @alexcrichton