Clang should be built with ThinLTO
Categories
(Firefox Build System :: Toolchains, enhancement)
Tracking
(firefox111 fixed)
Tracking | Status | |
---|---|---|
firefox111 | --- | fixed |
People
(Reporter: sergesanspaille, Assigned: sergesanspaille)
Details
Attachments
(1 file)
If we build our core compiler with LTO in addition to the existing PGO step, it should run faster, which in turn should make fx builds faster.
Assignee | ||
Comment 1•3 years ago
|
||
Comment 2•3 years ago
|
||
The Bugbug bot thinks this bug should belong to the 'Firefox Build System::Toolchains' component, and is moving the bug to that component. Please correct in case you think the bot is wrong.
Comment 3•3 years ago
|
||
There's a r+ patch which didn't land and no activity in this bug for 2 weeks.
:sergesanspaille, could you have a look please?
If you still have some work to do, you can add an action "Plan Changes" in Phabricator.
For more information, please visit auto_nag documentation.
Assignee | ||
Comment 4•3 years ago
|
||
This should be ok to land now that lld usage has reached a decent state.
Updated•3 years ago
|
Updated•3 years ago
|
Assignee | ||
Comment 5•3 years ago
|
||
edit: there was some warnings in the build logs that hints at using llvm-ar instead of ar when linking with lto, /me checks this.
Comment 6•3 years ago
|
||
(In reply to [:sergesanspaille] from comment #5)
edit: there was some warnings in the build logs that hints at using llvm-ar instead of ar when linking with lto, /me checks this.
where did that lead?
Assignee | ||
Comment 7•3 years ago
|
||
I ended up forcing usage of llvm-ar, llvm-ranlib (from teh same toolchain as clang), and that worked nice.
Comment 9•3 years ago
|
||
bugherder |
Description
•