Date: Sat, 4 Jun 2016 14:59:01 -0400 From: Michael Butler <imb@protected-networks.net> To: Konstantin Belousov <kostikbel@gmail.com>, Matthew Macy <mmacy@nextbsd.org> Cc: "freebsd-current@freebsd.org" <freebsd-current@freebsd.org>, alc@freebsd.org Subject: Re: repeatable panic on pageout with 945GM Message-ID: <88ad4228-2583-8a91-1751-d16f7a51de91@protected-networks.net> In-Reply-To: <20160604174745.GB38613@kib.kiev.ua> References: <2490f1c7-8153-ece3-49ed-4b3886564fd7@protected-networks.net> <da19738b-6bf1-10a3-4428-43b6095ec35a@protected-networks.net> <205d4423-b834-9a21-785f-fa15d44c78ec@protected-networks.net> <CAHM0Q_PR5Aoak6A7f=tsRy0DJFCmLDVfRGpceZ0mXU3P%2BxO8DA@mail.gmail.com> <1551419a1db.12929035f45012.326107747932338888@nextbsd.org> <939f9d2b-e925-e8e0-0ff3-8d90623728c6@protected-networks.net> <1551c5dbd86.c68532b5123717.566503881838650848@nextbsd.org> <20160604174745.GB38613@kib.kiev.ua>
next in thread | previous in thread | raw e-mail | index | archive | help
On 06/04/16 13:47, Konstantin Belousov wrote: [ .. snip .. ] > I believe that this is a bug in amd64 pmap. Fictitious pages are not > promoted, in particular, the pv_table array does not span over the > dynamically registered fictitious ranges. As result, pa_to_pvh() returns > garbage and pvh must not be accessed in the case of 'small_mappings' in > several pmap functions. It is typically not accessed, except in case > when we have to drop and reacquire pv lock, to avoid LOR with pmap. > > i386 does not have the issue, due to pvh_global_lock. > > Below is the supposed fix (not tested). [ .. snip .. ] Is this something I should test and, should it not introduce any other issues, might get committed? imb
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?88ad4228-2583-8a91-1751-d16f7a51de91>