Skip to content

[Docs] Document persistent metadata printing #220967

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.

Sign up for GitHub

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

Merged
yxsamliu merged 5 commits into from
Sep 8, 2026
Merged

[Docs] Document persistent metadata printing #220967

Changes from 1 commit
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
e89a277
[Docs] Document persistent metadata printing
yxsamliu Sep 3, 2026
3d90d05
[Docs] Clarify persistent metadata renumbering
yxsamliu Sep 4, 2026
ddafe0e
[Docs] Clarify direct module printing changes
yxsamliu Sep 4, 2026
acea702
[Docs] Explain metadata numbering choices
yxsamliu Sep 7, 2026
f8f991d
[Docs] Mention the metadata printing speedup
yxsamliu Sep 7, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
8 changes: 8 additions & 0 deletions llvm/docs/ReleaseNotes.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -52,6 +52,14 @@ Makes programs 10x faster by doing Special New Thing.

### Changes to the LLVM IR

* LLVM now assigns persistent print IDs to metadata nodes. This keeps metadata
IDs stable across repeated debug and pass output, but can change their
numbering and ordering compared with earlier releases. Standalone metadata
printing now uses numbered definitions such as `!1 = !DIFile(...)` instead
of pointer-based forms such as `<0x...> = !DIFile(...)`. Tools and tests that
compare such output may need updating. Final assembly output is still
renumbered into canonical order.
Comment thread
yxsamliu marked this conversation as resolved.
Resolved

* Added `llvm.vector.reduce.fmaximumnum` and `llvm.vector.reduce.fminimumnum`
intrinsics, the reduction variants of `llvm.maximumnum` and
`llvm.minimumnum`.
Expand Down
Loading