On 12/22/16 at 10:33P, Gleb Smirnoff wrote: > Hi! > > On Wed, Dec 21, 2016 at 04:19:28PM -0800, hiren panchasara wrote: > h> On 12/09/16 at 05:58P, Gleb Smirnoff wrote: > h> > Author: glebius > h> > Date: Fri Dec 9 17:58:34 2016 > h> > New Revision: 309745 > h> > URL: https://svnweb.freebsd.org/changeset/base/309745 > h> > > h> > Log: > h> > Provide counter_ratecheck(), a MP-friendly substitution to > ppsratecheck(). > h> > When rated event happens at a very quick rate, the ppsratecheck() is > not > h> > only racy, but also becomes a performance bottleneck. > h> > > h> > Together with: rrs, jtl > h> > h> Is it possible for you to MFC this and the following fix to it to 11? > > Have you tried it and found useful? If yes, you are welcome to merge it.
As you stated yourself in the commit message, this is a clear bug and an opportunity for performance enhancement. Do you feel this bug won't be in 11 and systems running on 11 won't have similar performance gains as those running on -head? Specially when 11 is the latest released version that many (like myself) would be running in the production. > > My current work involves only FreeBSD head, I don't have any STABLE production > boxes, that's why I am very conservative with my MFCs: I do only obvious > conservative bugfixes, not performance improvements. The obvious reason we encourage original submitter to do MFCs is that (s)he is intimately familiar with the changes. I can understand your point of not having any 11 based systems to test. I'll test this on our $work prod and do the MFC myself. Thanks for the original fix. Cheers, Hiren
pgpb1Ed4Di3I2.pgp
Description: PGP signature
