Date: Mon, 06 Mar 1995 07:13:45 -0800 From: David Greenman <davidg@Root.COM> To: starkhome!gene@sbstark.cs.sunysb.edu (Gene Stark) Cc: current@FreeBSD.org, dyson@Root.COM Subject: Re: Page fault panics during make world in -current Message-ID: <199503061513.HAA00557@corbin.Root.COM> In-Reply-To: Your message of "Mon, 06 Mar 95 08:28:22 EST." <199503061328.IAA03874@starkhome.cs.sunysb.edu>
next in thread | previous in thread | raw e-mail | index | archive | help
>I notice that control has arrived at line 1005 in vfs_bio.c with "obj" >set to a VM object with ref_count = 0, even though the object was obtained >from vp->v_vmdata. It seems wrong to have a pointer to a VM object >stored in a vnode without a positive reference count on the VM object. Cached objects have reference counts of 0. A reference to the vnode is held as long as the pager exists (which is for the life of the object). When the pager is destroyed, vp->v_vmdata is set to NULL. >The size field of the VM object also seems bogus: 0xf0668da4, though this >could perhaps be a consequence of the object's already having been freed >and the memory used for something else. That is strange. The object hasn't been freed. This is starting to look like the memory corruption problem(s) that I've been trying to find. -DG
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199503061513.HAA00557>