Date: Mon, 8 Dec 2003 16:04:19 -0600 From: Dan Nelson <dnelson@allantgroup.com> To: Kris Kennaway <kris@obsecurity.org> Cc: Martin <nakal@web.de> Subject: Re: Is "vm_fault" a kernel bug here? Message-ID: <20031208220419.GB2435@dan.emsphone.com> In-Reply-To: <20031208214056.GA52416@xor.obsecurity.org> References: <1070918549.702.22.camel@klotz.local> <20031208214056.GA52416@xor.obsecurity.org>
next in thread | previous in thread | raw e-mail | index | archive | help
In the last episode (Dec 08), Kris Kennaway said: > On Mon, Dec 08, 2003 at 10:22:30PM +0100, Martin wrote: > > Here is the result of a read access on a file, which is obviously > > unreadable: > > > > acd0: FAILURE - READ BIG status=51<READY,DSC,ERROR> sensekey=MEDIUM > > ERROR error=4<ABORTED> > > vm_fault: pager read error, pid 606 (cp) > > cp: filename: Bad address > > > > I want to report it, because of the "vm_fault", it sounds "scary". > > Is this a correct behaviour or a bug in the kernel? > > That's just what happens when an error is detected by the disk > driver. This can be caused by hardware failure or incompatibility, or > a driver bug. More specifically, you get a vm_fault if the kernel is trying to page in a block of memory for a process but can't. The reason you get this instead of a regular read error is because cp uses mmap() on files smaller than 8 MB. -- Dan Nelson dnelson@allantgroup.com
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20031208220419.GB2435>