Messages in this thread | | | Subject | seq_file optimization? | From | Joe Perches <> | Date | Fri, 30 Jun 2017 19:48:17 -0700 |
| |
About this commit in mm:
--------------------------------------------------- commit c1aca87fb3f84a813fb3dda656bbee583c9c87a3 Author: Alexey Dobriyan <adobriyan@gmail.com> Date: Fri Jun 30 11:02:04 2017 +1000
fs/seq_file.c: delete small-value optimization num_to_str() optimizes printing small integers [0..9], so the same check higher in callchain is unnecessary. ---------------------------------------------------
Is removing this optimization really the same speed?
I don't think so.
The num_to_str call still does multiple moves of memory to the output, this is a single byte copy.
| |