From mboxrd@z Thu Jan 1 00:00:00 1970 From: Pavel Machek Date: Thu, 09 Dec 2004 10:57:53 +0000 Subject: Re: Anticipatory prefaulting in the page fault handler V1 Message-Id: <20041209105753.GB1131@elf.ucw.cz> List-Id: References: <41AEB44D.2040805@pobox.com> <20041201223441.3820fbc0.akpm@osdl.org> <41AEBAB9.3050705@pobox.com> <20041201230217.1d2071a8.akpm@osdl.org> <179540000.1101972418@[10.10.2.4]> <41AEC4D7.4060507@pobox.com> <20041202101029.7fe8b303.cliffw@osdl.org> In-Reply-To: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: Christoph Lameter Cc: nickpiggin@yahoo.com.au, Jeff Garzik , torvalds@osdl.org, hugh@veritas.com, benh@kernel.crashing.org, linux-mm@kvack.org, linux-ia64@vger.kernel.org, linux-kernel@vger.kernel.org Hi! > Standard Kernel on a 512 Cpu machine allocating 32GB with an increasing > number of threads (and thus increasing parallellism of page faults): > > Gb Rep Threads User System Wall flt/cpu/s fault/wsec > 32 3 1 1.416s 138.165s 139.050s 45073.831 45097.498 ... > Patched kernel: > > Gb Rep Threads User System Wall flt/cpu/s fault/wsec > 32 3 1 1.098s 138.544s 139.063s 45053.657 45057.920 ... > These number are roughly equal to what can be accomplished with the > page fault scalability patches. > > Kernel patches with both the page fault scalability patches and > prefaulting: > > Gb Rep Threads User System Wall flt/cpu/s fault/wsec > 32 10 1 4.103s 456.384s 460.046s 45541.992 45544.369 ... > > The fault rate doubles when both patches are applied. ... > We are getting into an almost linear scalability in the high end with > both patches and end up with a fault rate > 3 mio faults per second. Well, with both patches you also slow single-threaded case more than twice. What are the effects of this patch on UP system? Pavel -- People were complaining that M$ turns users into beta-testers... ...jr ghea gurz vagb qrirybcref, naq gurl frrz gb yvxr vg gung jnl!