Date: Mon, 4 May 2015 18:49:25 +0000 (UTC) From: Gleb Smirnoff <glebius@FreeBSD.org> To: src-committers@freebsd.org, svn-src-all@freebsd.org, svn-src-head@freebsd.org Subject: svn commit: r282426 - head/sys/vm Message-ID: <201505041849.t44InPxC073254@svn.freebsd.org>
next in thread | raw e-mail | index | archive | help
Author: glebius Date: Mon May 4 18:49:25 2015 New Revision: 282426 URL: https://svnweb.freebsd.org/changeset/base/282426 Log: Fix arithmetical bug in vnode_pager_haspage(). The check against object size should be done not with the number of pages in the first block, but with the overall number of pages. While here, add KASSERT that makes sure that BMAP doesn't return completely irrelevant blocks. Reviewed by: kib Tested by: pho Sponsored by: Netflix Sponsored by: Nginx, Inc. Modified: head/sys/vm/vnode_pager.c Modified: head/sys/vm/vnode_pager.c ============================================================================== --- head/sys/vm/vnode_pager.c Mon May 4 18:20:31 2015 (r282425) +++ head/sys/vm/vnode_pager.c Mon May 4 18:49:25 2015 (r282426) @@ -340,16 +340,21 @@ vnode_pager_haspage(vm_object_t object, *before += poff; } if (after) { - int numafter; + /* + * The BMAP vop can report a partial block in the + * 'after', but must not count blocks after EOF. + * Assert the latter, and truncate 'after' in case + * of the former. + */ + KASSERT(reqblock + *after <= + object->size * pagesperblock, + ("%s: reqblock %jd after %d size %ju", __func__, + (intmax_t )reqblock, *after, + (uintmax_t )object->size)); *after *= pagesperblock; - numafter = pagesperblock - (poff + 1); - if (IDX_TO_OFF(pindex + numafter) > - object->un_pager.vnp.vnp_size) { - numafter = - OFF_TO_IDX(object->un_pager.vnp.vnp_size) - - pindex; - } - *after += numafter; + *after += pagesperblock - (poff + 1); + if (pindex + *after >= object->size) + *after = object->size - 1 - pindex; } } else { if (before) {
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?201505041849.t44InPxC073254>