cvs commit: src/share/man/man9 Makefile sx.9 src/sys/conf NOTES
options src/sys/dev/acpica acpi_ec.c src/sys/dev/mxge
if_mxge.c src/sys/dev/usb if_aue.c if_axe.c
src/sys/gnu/fs/xfs/FreeBSD/support mrlock.c mrlock.h ...
attilio at freebsd.org
Sun Apr 1 00:36:18 UTC 2007
2007/4/1, Kris Kennaway <kris at obsecurity.org>:
> eOn Sat, Mar 31, 2007 at 11:23:42PM +0000, John Baldwin wrote:
> > jhb 2007-03-31 23:23:42 UTC
> > FreeBSD src repository
> > Modified files:
> > share/man/man9 Makefile sx.9
> > sys/conf NOTES options
> > sys/dev/acpica acpi_ec.c
> > sys/dev/mxge if_mxge.c
> > sys/dev/usb if_aue.c if_axe.c
> > sys/gnu/fs/xfs/FreeBSD/support mrlock.c mrlock.h
> > sys/i386/acpica acpi_machdep.c
> > sys/kern kern_sx.c
> > sys/netinet6 in6_src.c
> > sys/sys sleepqueue.h sx.h
> > Added files:
> > sys/sys _sx.h
> > Log:
> > Optimize sx locks to use simple atomic operations for the common cases of
> > obtaining and releasing shared and exclusive locks. The algorithms for
> > manipulating the lock cookie are very similar to that rwlocks. This patch
> > also adds support for exclusive locks using the same algorithm as mutexes.
> Thanks to Attilio for doing this work and to John for committing it.
> This is a significant step forward for 7.0 and will be the basis for
> some major performance optimizations to be committed in the near
> future (e.g. filedesc locking from rwatson, which gives even better
> mysql performance than the "tophalf" mutexes Jeff and I recently
Thanks a lot to you for the biggest effort you did in testing and
benchmarking the patch and to pho@ who did first stability tests on
the first revision of the patch.
These credits should be however shared completely with John who did an
excellent job of revision on the code and in particulary for the
challenging discussions we had during the time of sx rewriting.
I think I learned a lot by his incredible way of coding.
Peace can only be achieved by understanding - A. Einstein
More information about the freebsd-current