Page MenuHomeFreeBSD

per-cpu free page caching
ClosedPublic

Authored by jeff on Mar 30 2018, 3:47 AM.
Tags
None
Referenced Files
Unknown Object (File)
Jan 21 2025, 11:48 PM
Unknown Object (File)
Jan 21 2025, 12:44 AM
Unknown Object (File)
Jan 8 2025, 11:56 PM
Unknown Object (File)
Jan 8 2025, 11:26 PM
Unknown Object (File)
Dec 22 2024, 10:16 PM
Unknown Object (File)
Dec 21 2024, 8:07 PM
Unknown Object (File)
Nov 21 2024, 6:43 AM
Unknown Object (File)
Nov 18 2024, 8:03 AM
Subscribers

Details

Summary

This implements per-cpu free page caching only for FREEPOOL_DEFAULT. I haven't seen enough contention yet from FREEPOOL_DIRECT that it is an issue. This also brings in vm_phys_alloc_npages() which gives you as many as 'count' pages. This is an optimization to reduce round-trips to the backend phys allocator. Contrary to other instances of this patch, this is now completely compatible with reservations.

It might be more elegant to push this into vm_phys.c along with all of the vm_domain_free_lock() calls. However, pushing it there would mean the cached pages were reflected in the free count. I believe this will make us more prone to out of memory deadlocks.

Diff Detail

Repository
rS FreeBSD src repository - subversion
Lint
Lint Skipped
Unit
Tests Skipped

Event Timeline

vm/vm_page.c
224

Should it be sizeof(struct vm_page *) ?

vm/vm_page.c
224

actually in this case uma uses it as a cost function to limit the bucket size. But we override with MAXBUCKET. So it is correct to intent but unused.

This revision is now accepted and ready to land.Apr 1 2018, 4:20 AM
This revision was automatically updated to reflect the committed changes.