Details
-
Improvement
-
Resolution: Fixed
-
Minor
-
None
-
None
-
3
-
9223372036854775807
Description
When copying one page to another in kernel memory (not to/from userspace, so this can't be used for the general purpose IO stuff), the kernel has an optimized copy_page which can be used instead of memcpy().
This is relevant in the lnet loopback subsystem, which copies from the kiov from the client to that on the server. (Using the same page is nasty for a lot of reasons, so copying is best.) The most common case is full page to full page, so add a check for that.
On my little tiny VM system, this improves maximum write performance (with the fake write fail_loc enabled) by about 20%, from 4.4 GiB/s to 5.7 GiB/s.
Attachments
Issue Links
- is related to
-
LU-10191 FLR2: Server Local Client (SLC)
- Reopened