Typed property assign should be optimized more
Nobody has claimed this yet.
- Dominant language
- C
- Stars
- 40.4k
- Forks
- 8.1k
- Avg merge
- 2d 13h
- Merged PRs (30d)
- 96
Description
Description
Classes are very often used as a simple data holding structure which is created and destroyed millions times per script lifecycle/run.
In https://3v4l.org/CMktE/rfc#vgit.master I have discovered a "class like struct" is slower than an array. This is something not fully surprising on the first sight.
However, https://3v4l.org/hSSFJ/rfc#vgit.master, same demo with propoted properties only, with empty constructor, is not (much if any) faster.
Initially I thought the interpreted code in the constructor will contribute the most to the total run time. This seems to be not true, it seems the most of the time is spent in object allocation and release. If an object has no cycles, it should be fast. I hope there is a room to optimize the object allocation/release more and make this core part of php faster.
let's focus on typed property assign performance solely here
repro: https://3v4l.org/FGeEf
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start with the 3v4l.org/FGeEf reproduction and profile the typed-property assignment case against the current PHP interpreter. Determine which allocation or assignment path dominates before proposing an optimization; done means a measured improvement without changing semantics, supported by a focused regression or performance check if the project provides one.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- c, php
- Domain
- backend, performance
- Issue type
- Refactor
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Stale
- Clarity
- Mostly clear
- Newbie friendliness
- 25/100