microsoft / microsoft/terminal
Analyze PGO profiles during EIM
- Dominant language
- C++
- Stars
- 105k
- Forks
- 9.6k
- Avg merge
- 3d 17h
- Merged PRs (30d)
- 29
Description
@dmachaj noticed that a good chunk of our boot work is in .text$zz... so it's not in the PGO training.
When I put the initial PGO training tests in at https://github.com/microsoft/terminal/blob/main/src/cascadia/WindowsTerminal_UIATests/SmokeTests.cs, it was like the bare minimum to get us going on the PGO stuff.
EIM is probably a good time to revisit that and check how effective they're actually being and whether we need to make adjustments to the training set.
Contributor guide
Research direction
Start with src/cascadia/WindowsTerminal_UIATests/SmokeTests.cs and review the initial PGO training tests. Analyze the EIM profiles against the boot work in .text$zz... to assess whether the training set is effective. Done means identifying any gaps and making the necessary training-set adjustments.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- cpp
- Domain
- build-system, performance
- Issue type
- Refactor
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 25/100