-
Notifications
You must be signed in to change notification settings - Fork 13.4k
[libc++] Add _LIBCPP_NO_UNIQUE_ADDRESS to flat_{,multi}map::value_compare #137594
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Conversation
@llvm/pr-subscribers-libcxx Author: Nikolas Klauser (philnik777) ChangesThis breaks the ABI of Full diff: https://github.com/llvm/llvm-project/pull/137594.diff 2 Files Affected:
diff --git a/libcxx/include/__flat_map/flat_map.h b/libcxx/include/__flat_map/flat_map.h
index f5abfd0985280..f5e9756ff6a60 100644
--- a/libcxx/include/__flat_map/flat_map.h
+++ b/libcxx/include/__flat_map/flat_map.h
@@ -114,7 +114,7 @@ class flat_map {
class value_compare {
private:
- key_compare __comp_;
+ _LIBCPP_NO_UNIQUE_ADDRESS key_compare __comp_;
_LIBCPP_HIDE_FROM_ABI value_compare(key_compare __c) : __comp_(__c) {}
friend flat_map;
diff --git a/libcxx/include/__flat_map/flat_multimap.h b/libcxx/include/__flat_map/flat_multimap.h
index ea77fb5d79bd2..15fcd7995ad0a 100644
--- a/libcxx/include/__flat_map/flat_multimap.h
+++ b/libcxx/include/__flat_map/flat_multimap.h
@@ -115,7 +115,7 @@ class flat_multimap {
class value_compare {
private:
- key_compare __comp_;
+ _LIBCPP_NO_UNIQUE_ADDRESS key_compare __comp_;
_LIBCPP_HIDE_FROM_ABI value_compare(key_compare __c) : __comp_(__c) {}
friend flat_multimap;
|
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I agree this is very unlikely to break anything, because value_compare
is not stored in flat_map
. The only situation this can break anything is when user do m.value_compare()
and then store the results in a struct.
Could we add a unittest in libcxx folder?
Otherwise LGTM
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I think this is very low risk if we do it right now and backport it. I also don't see what test we'd add for this in order to really add value. LGTM as is.
/cherry-pick ed0aa99 |
/pull-request #138880 |
This breaks the ABI of
flat_{,multi}map::value_compare
, but this type has only been introduced in LLVM 20, so it should be very unlikely that we break anybody if we back-port this now.