tmpfs_getattr fills > 100 bytes worth of data into struct vattr, which then has to be converted to struct stat format. The conversion happens to copy some of it as it is and branch on other stuff. tmpfs_stat elides the entire process.
On 8/8/20, Alan Somers <[email protected]> wrote: > On Fri, Aug 7, 2020 at 5:06 PM Mateusz Guzik <[email protected]> wrote: > >> Author: mjg >> Date: Fri Aug 7 23:06:40 2020 >> New Revision: 364044 >> URL: https://svnweb.freebsd.org/changeset/base/364044 >> >> Log: >> vfs: add VOP_STAT >> >> The current scheme of calling VOP_GETATTR adds avoidable overhead. >> >> An example with tmpfs doing fstat (ops/s): >> before: 7488958 >> after: 7913833 >> >> Reviewed by: kib (previous version) >> Differential Revision: https://reviews.freebsd.org/D25910 >> >> Modified: >> head/share/man/man9/Makefile >> head/share/man/man9/VOP_ATTRIB.9 >> head/sys/compat/linuxkpi/common/src/linux_compat.c >> head/sys/kern/vfs_default.c >> head/sys/kern/vfs_syscalls.c >> head/sys/kern/vfs_vnops.c >> head/sys/kern/vnode_if.src >> head/sys/security/audit/audit_arg.c >> head/sys/sys/vnode.h >> > > What avoidable overhead? The tmpfs_stat handler that you added in your next > commit looks pretty much the same as vop_getattr. I'm missing where the > performance improvement comes from. Could you please fill me in? > -Alan > -- Mateusz Guzik <mjguzik gmail.com> _______________________________________________ [email protected] mailing list https://lists.freebsd.org/mailman/listinfo/svn-src-head To unsubscribe, send any mail to "[email protected]"
