Date: Thu, 22 Mar 2012 11:44:35 -0500 From: Alan Cox <alc@rice.edu> To: Sergey Kandaurov <pluknet@freebsd.org> Cc: Alan Cox <alc@freebsd.org>, svn-src-head@freebsd.org, svn-src-all@freebsd.org, src-committers@freebsd.org Subject: Re: svn commit: r233290 - head/sys/amd64/amd64 Message-ID: <4F6B56F3.6030209@rice.edu> In-Reply-To: <CAE-mSO%2BFe8_xDU9%2BPnQBxWR2EF=P8wB2REyhnHgeuz9ojndQwQ@mail.gmail.com> References: <201203220440.q2M4eMSl007771@svn.freebsd.org> <CAE-mSO%2BFe8_xDU9%2BPnQBxWR2EF=P8wB2REyhnHgeuz9ojndQwQ@mail.gmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On 03/22/2012 03:12, Sergey Kandaurov wrote: > On 22 March 2012 08:40, Alan Cox<alc@freebsd.org> wrote: >> Author: alc >> Date: Thu Mar 22 04:40:22 2012 >> New Revision: 233290 >> URL: http://svn.freebsd.org/changeset/base/233290 >> >> Log: >> Change pv_entry_count to a long. During the lifetime of FreeBSD 10.x, >> physical memory sizes at the high-end will likely reach a point that >> the number of pv entries could overflow an int. >> >> Submitted by: kib >> >> Modified: >> head/sys/amd64/amd64/pmap.c >> > While there, maybe also do something with these semi-debug-ish > vm.pmap.pc_chunk_(allocs|frees) like change them to [u_]long too, move > under debug or just drop them? While they can be useful for someone, > pc_chunk_(allocs|frees) have also an "int" and quite quickly overflow. > Though e.g. vm.pmap.pv_entry_(allocs|frees) have a long. > Yes, in the next month or two I'm going to be making some changes to the way that pmap_collect() works. I will be eliminating a couple of the counters as part of those changes. I will deal with the rest of the counters at the same time. Alan
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?4F6B56F3.6030209>