Skip to content

Optimize sequenceLock writeAcquire/Release #1

@latentPrion

Description

@latentPrion

According to @tstamp 14:30 , (https://youtu.be/uiBMczcm69A?t=870&si=_aZukm9F-rl4QooR) we can reduce the number of RMW atomic instructions by reading once and then writing val+1 in writeAcquire and val+2 in writeRelease.

Investigate this and see if it's true, and impl in SequenceLock if true.

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions