| commit | 7ac3ee2d5980ad6ba00a7529b307f5f0e57e63ed | [log] [tgz] |
|---|---|---|
| author | Luis Chamberlain <mcgrof@kernel.org> | Mon Dec 02 16:20:40 2024 -0800 |
| committer | Luis Chamberlain <mcgrof@kernel.org> | Mon Dec 02 16:20:40 2024 -0800 |
| tree | 0d4db9e8dcb9e8adf22170940c6b776401971d95 | |
| parent | 40384c840ea1944d7c5a392e8975ed088ecf0b37 [diff] |
x86: make execmem_alloc() cache optional This let's us evaluate the gains with minimal kernel changes since the only information we had before this implementation was a "AB testing" for eBPF claim for about 0.5% gains whatever that means... and then a student who saw "10% performance improvement" but with no details spelled out at all. Signed-off-by: Luis Chamberlain <mcgrof@kernel.org>