Messages in this thread | | | From | David Laight <> | Subject | RE: Performance Regression in Linux Kernel 5.19 | Date | Fri, 9 Sep 2022 21:22:12 +0000 |
| |
From: Manikandan Jagatheesan > Sent: 09 September 2022 12:46 > > As part of VMware's performance regression testing for Linux > Kernel upstream releases, we have evaluated the performance > of Linux kernel 5.19 against the 5.18 release and we have > noticed performance regressions in Linux VMs on ESXi as shown > below. > - Compute(up to -70%) > - Networking(up to -30%) > - Storage(up to -13%) > > After performing the bisect between kernel 5.18 and 5.19, we > identified the root cause to be the enablement of IBRS mitigation > for spectre_v2 vulnerability by commit 6ad0ad2bf8a6 ("x86/bugs: > Report Intel retbleed vulnerability").
As a matter of interest how much faster does it go if you boot with all mitigations disabled and compile without retpolines and without page table separation?
There are plenty of semi-embedded systems (even running on x86) where there are a limited set of binaries, it is difficult to add new binaries, and everything basically runs as root.
David
- Registered Address Lakeside, Bramley Road, Mount Farm, Milton Keynes, MK1 1PT, UK Registration No: 1397386 (Wales)
| |