Date: Fri, 15 Feb 2013 15:12:00 +0100 From: Gary Jennejohn <gljennjohn@googlemail.com> To: Dag-Erling =?ISO-8859-1?Q?Sm=F8rgrav?= <des@des.no> Cc: Konstantin Belousov <kostikbel@gmail.com>, arch@freebsd.org Subject: Re: Unmapped I/O Message-ID: <20130215151200.69cc8554@ernst.jennejohn.org> In-Reply-To: <86621trc50.fsf@ds4.des.no> References: <20121219135451.GU71906@kib.kiev.ua> <20130214194731.GK2522@kib.kiev.ua> <86621trc50.fsf@ds4.des.no>
next in thread | previous in thread | raw e-mail | index | archive | help
On Fri, 15 Feb 2013 13:32:11 +0100 Dag-Erling Smørgrav <des@des.no> wrote: > Konstantin Belousov <kostikbel@gmail.com> writes: > > I consider the current patch ready to be committed into the HEAD. > > Some elements of the design were discussed with Jeff Roberson. The > > patch was tested by Peter Holm, a backport to stable/9 got load > > testing by Scott Long. I see an ~30% reduction in the system time on > > reading large files over UFS/ahci on the 4-core HTT machine. > > That sounds very nice indeed. Have you observed a similar reduction in > buildworld times? > I applied the patch and am currently running with it. Had to fix a fatal error in subr_bus_dma.c - "error may be used unitialized" or something to that effect. I don't have the patched source any longer, but I think it was at line 131. I started a build world with -j6 but unfortunately it broke off with an error which is no longer visible in the output. -- Gary Jennejohn
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20130215151200.69cc8554>