From owner-freebsd-fs@freebsd.org Sun Oct 29 10:36:46 2017 Return-Path: Delivered-To: freebsd-fs@mailman.ysv.freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:1900:2254:206a::19:1]) by mailman.ysv.freebsd.org (Postfix) with ESMTP id E35E7E5FD18 for ; Sun, 29 Oct 2017 10:36:46 +0000 (UTC) (envelope-from se@freebsd.org) Received: from mailout10.t-online.de (mailout10.t-online.de [194.25.134.21]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (Client CN "mailout00.t-online.de", Issuer "TeleSec ServerPass DE-2" (verified OK)) by mx1.freebsd.org (Postfix) with ESMTPS id 8CC1574BF6; Sun, 29 Oct 2017 10:36:46 +0000 (UTC) (envelope-from se@freebsd.org) Received: from fwd02.aul.t-online.de (fwd02.aul.t-online.de [172.20.26.148]) by mailout10.t-online.de (Postfix) with SMTP id AFC5441F702C; Sun, 29 Oct 2017 11:36:37 +0100 (CET) Received: from Stefans-MBP-2.fritz.box (rCC2fZZE8hMI4h5HwjrVX0-gS-qqS4SPz7mMM1Pe+sc8p6U1lTyDZREEdvhQH50wk2@[84.154.121.171]) by fwd02.t-online.de with (TLSv1.2:ECDHE-RSA-AES256-GCM-SHA384 encrypted) esmtp id 1e8kxE-2qdRmi0; Sun, 29 Oct 2017 11:36:32 +0100 Subject: Re: softdep as a mount(8) option To: Kirk McKusick , Mark Johnston Cc: freebsd-fs@FreeBSD.org References: <201710282225.v9SMPDCZ074228@chez.mckusick.com> From: Stefan Esser Message-ID: Date: Sun, 29 Oct 2017 11:36:32 +0100 User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.13; rv:52.0) Gecko/20100101 Thunderbird/52.4.0 MIME-Version: 1.0 In-Reply-To: <201710282225.v9SMPDCZ074228@chez.mckusick.com> Content-Type: multipart/mixed; boundary="------------24E4D2D2320EE5A929E27EF5" Content-Language: en-GB X-ID: rCC2fZZE8hMI4h5HwjrVX0-gS-qqS4SPz7mMM1Pe+sc8p6U1lTyDZREEdvhQH50wk2 X-TOI-MSGID: 0e32c1f2-a3a8-415e-b6b7-c539b65a095e X-BeenThere: freebsd-fs@freebsd.org X-Mailman-Version: 2.1.23 Precedence: list List-Id: Filesystems List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sun, 29 Oct 2017 10:36:47 -0000 This is a multi-part message in MIME format. --------------24E4D2D2320EE5A929E27EF5 Content-Type: text/plain; charset=windows-1252 Content-Transfer-Encoding: 7bit Am 29.10.17 um 00:25 schrieb Kirk McKusick: >> Date: Fri, 27 Oct 2017 11:39:00 -0400 >> From: Mark Johnston >> To: freebsd-fs@FreeBSD.org >> Subject: softdep as a mount(8) option >> >> Hi, >> >> I'd like to finally enable the use of SU (not SU+J) on some small UFS >> filesystems. The fact that SU is enabled using a flag in the superblock >> poses a problem for me, however: the systems containing these >> filesystems may at any time be repurposed to run a kernel that supports >> SU but contains bugs[*] that render it unusable. I therefore can't >> persistently enable SU in these systems. >> >> I'm wondering if it would be possible to enable SU using a mount >> option rather than with a persistent flag. fsck_ffs conditionalizes some >> of its logic on whether SU is configured - is this necessary for >> correctness? That is, if I run fsck on an unclean filesystem that had >> been mounted with SU, and fsck runs as though SU hadn't been configured, >> what problems might arise? >> >> [*] These bugs are a result of local modifications and aren't in >> FreeBSD. > > While it is safe and possible to add soft-updates (but not journalled > soft updates) as a mount option, it means that fsck will not know that > soft updates were in use, so it will always run in full (slow) mode at > boot time. This is why I have not added it as an option. Hi Kirk, I had implemented this feature more than 10 years ago and even got your reply, that this was a better solution than the mount option, when I discussed the implementation with you at that time. (That was shortly after soft-updates made it into FreeBSD - a long time ago ...) But despite your agreement that I should commit that change (for the next major FreeBSD release), it was later considered a violation of POLA and I kept (and maintained it) on my system only. I had answered the same question regarding FSCK more than 10 years ago in the following archived message: https://lists.freebsd.org/pipermail/freebsd-hackers/2005-December/014865.html My solution was very simple: The mount option sets the soft-updetes flag in the super-block, if a file system is mounted RW with soft-updates enabled. If it is not cleanly unmounted, then FSCK will correctly check it even on a system that does not support soft-updates in the kernel (but in FSCK, where it's not optional). When such file system is mounted on a system without my patch, it is indistinguishable from a drive that had the soft-updates flag set via tunefs. In fact, my patch made the kernel do the equivalent of tunefs -j on each mount, depending on the mount option. The only special case is that mount -u must not change the soft-updates state on a file-system that is mounted RW (but tunefs cannot do this, too), since I did not want to implement the required flushing of all buffers before switching the mode. I have stopped using UFS (have switched to ZFS) a few years ago, but have maintained my changes in my local version of -CURRENT to this date. (For that reason I had no interest in SU+J and there might be a need to adapt the code to correctly work with SU+J ...) Regards, STefan PS: I have attached my patch to this mail, will be stripped by the mail-list, but may be of interest to the direct recipients ... --------------24E4D2D2320EE5A929E27EF5 Content-Type: text/plain; charset=UTF-8; x-mac-type="0"; x-mac-creator="0"; name="soft-updates-option.diff" Content-Transfer-Encoding: 7bit Content-Disposition: attachment; filename="soft-updates-option.diff" Index: sys/ufs/ffs/ffs_vfsops.c =================================================================== --- sys/ufs/ffs/ffs_vfsops.c (Revision 325051) +++ sys/ufs/ffs/ffs_vfsops.c (Arbeitskopie) @@ -138,7 +138,7 @@ */ static const char *ffs_opts[] = { "acls", "async", "noatime", "noclusterr", "noclusterw", "noexec", "export", "force", "from", "groupquota", - "multilabel", "nfsv4acls", "fsckpid", "snapshot", "nosuid", "suiddir", + "multilabel", "nfsv4acls", "fsckpid", "snapshot", "softdep", "nosuid", "suiddir", "nosymfollow", "sync", "union", "userquota", NULL }; static int @@ -237,17 +237,26 @@ ump = VFSTOUFS(mp); fs = ump->um_fs; devvp = ump->um_devvp; if (fsckpid == -1 && ump->um_fsckpid > 0) { if ((error = ffs_flushfiles(mp, WRITECLOSE, td)) != 0 || (error = ffs_sbupdate(ump, MNT_WAIT, 0)) != 0) return (error); g_topology_lock(); + if (fs->fs_ronly == 0) { /* - * Return to normal read-only mode. + * Preserve current softdep status, + * unless mounted read-only. */ + MNT_ILOCK(mp); + if (fs->fs_flags & FS_DOSOFTDEP) { + mp->mnt_flag |= MNT_SOFTDEP; + } else { + mp->mnt_flag &= ~MNT_SOFTDEP; + } + MNT_IUNLOCK(mp); error = g_access(ump->um_cp, 0, -1, 0); g_topology_unlock(); ump->um_fsckpid = 0; } if (fs->fs_ronly == 0 && vfs_flagopt(mp->mnt_optnew, "ro", NULL, 0)) { @@ -266,7 +265,8 @@ flags = WRITECLOSE; if (mp->mnt_flag & MNT_FORCE) flags |= FORCECLOSE; - if (MOUNTEDSOFTDEP(mp)) { +// if (MOUNTEDSOFTDEP(mp)) { + if (fs->fs_flags & FS_DOSOFTDEP) { error = softdep_flushfiles(mp, flags, td); } else { error = ffs_flushfiles(mp, flags, td); @@ -346,6 +346,15 @@ (fs->fs_flags & FS_DOSOFTDEP))) { printf("WARNING: %s was not properly " "dismounted\n", fs->fs_fsmnt); + /* + * Preserve softdep state if unclean + */ + MNT_ILOCK(mp); + if ((fs->fs_flags & FS_DOSOFTDEP) == 0) + mp->mnt_flag &= ~MNT_SOFTDEP; + else + mp->mnt_flag |= MNT_SOFTDEP; + MNT_IUNLOCK(mp); } else { vfs_mount_error(mp, "R/W mount of %s denied. %s.%s", @@ -373,7 +380,7 @@ MNT_IUNLOCK(mp); fs->fs_mtime = time_second; /* check to see if we need to start softdep */ - if ((fs->fs_flags & FS_DOSOFTDEP) && + if ((mp->mnt_flag & MNT_SOFTDEP) && (error = softdep_mount(devvp, mp, fs, td->td_ucred))){ vn_finished_write(mp); return (error); @@ -851,6 +858,15 @@ (fs->fs_flags & FS_DOSOFTDEP))) { printf("WARNING: %s was not properly dismounted\n", fs->fs_fsmnt); + /* + * Preserve softdep state if unclean. + */ + MNT_ILOCK(mp); + if ((fs->fs_flags & FS_DOSOFTDEP) == 0) + mp->mnt_flag &= ~MNT_SOFTDEP; + else + mp->mnt_flag |= MNT_SOFTDEP; + MNT_IUNLOCK(mp); } else { vfs_mount_error(mp, "R/W mount of %s denied. %s%s", fs->fs_fsmnt, "Filesystem is not clean - run fsck.", @@ -875,6 +889,11 @@ fs->fs_pendingblocks = 0; fs->fs_pendinginodes = 0; } + /* Clear softdep flag in superblock, if not a softdep mount. */ + MNT_ILOCK(mp); + if ((mp->mnt_flag & MNT_SOFTDEP) == 0) + fs->fs_flags &= ~FS_DOSOFTDEP; + MNT_IUNLOCK(mp); if ((fs->fs_flags & FS_GJOURNAL) != 0) { #ifdef UFS_GJOURNAL /* @@ -1071,7 +1088,7 @@ if (ronly == 0) { fs->fs_mtime = time_second; - if ((fs->fs_flags & FS_DOSOFTDEP) && + if ((mp->mnt_flag & MNT_SOFTDEP) && (error = softdep_mount(devvp, mp, fs, cred)) != 0) { free(fs->fs_csp, M_UFSMNT); ffs_flushfiles(mp, FORCECLOSE, td); Index: sbin/mount/mntopts.h =================================================================== --- sbin/mount/mntopts.h (Revision 325058) +++ sbin/mount/mntopts.h (Arbeitskopie) @@ -52,6 +52,7 @@ #define MOPT_NOCLUSTERW { "clusterw", 1, MNT_NOCLUSTERW, 0 } #define MOPT_SUIDDIR { "suiddir", 0, MNT_SUIDDIR, 0 } #define MOPT_SNAPSHOT { "snapshot", 0, MNT_SNAPSHOT, 0 } +#define MOPT_SOFTDEP { "softdep", 0, MNT_SOFTDEP, 0 } #define MOPT_MULTILABEL { "multilabel", 0, MNT_MULTILABEL, 0 } #define MOPT_ACLS { "acls", 0, MNT_ACLS, 0 } #define MOPT_NFS4ACLS { "nfsv4acls", 0, MNT_NFS4ACLS, 0 } Index: sbin/mount/mount.c =================================================================== --- sbin/mount/mount.c (Revision 325058) +++ sbin/mount/mount.c (Arbeitskopie) @@ -958,6 +958,7 @@ if (flags & MNT_NOCLUSTERW) res = catopt(res, "noclusterw"); if (flags & MNT_NOSYMFOLLOW) res = catopt(res, "nosymfollow"); if (flags & MNT_SUIDDIR) res = catopt(res, "suiddir"); + if (flags & MNT_SOFTDEP) res = catopt(res, "softdep"); if (flags & MNT_MULTILABEL) res = catopt(res, "multilabel"); if (flags & MNT_ACLS) res = catopt(res, "acls"); if (flags & MNT_NFS4ACLS) res = catopt(res, "nfsv4acls"); --------------24E4D2D2320EE5A929E27EF5-- From owner-freebsd-fs@freebsd.org Sun Oct 29 10:57:10 2017 Return-Path: Delivered-To: freebsd-fs@mailman.ysv.freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:1900:2254:206a::19:1]) by mailman.ysv.freebsd.org (Postfix) with ESMTP id B3CFDE60251 for ; Sun, 29 Oct 2017 10:57:10 +0000 (UTC) (envelope-from se@freebsd.org) Received: from mailout02.t-online.de (mailout02.t-online.de [194.25.134.17]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (Client CN "mailout00.t-online.de", Issuer "TeleSec ServerPass DE-2" (verified OK)) by mx1.freebsd.org (Postfix) with ESMTPS id 5F29E7548F; Sun, 29 Oct 2017 10:57:10 +0000 (UTC) (envelope-from se@freebsd.org) Received: from fwd33.aul.t-online.de (fwd33.aul.t-online.de [172.20.27.144]) by mailout02.t-online.de (Postfix) with SMTP id E3F7241AE1A8; Sun, 29 Oct 2017 11:57:01 +0100 (CET) Received: from Stefans-MBP-2.fritz.box (r1-YDvZE8htzYMhDyrh7hpKNSEfP0y7R2lUtXcbZ-JnbELUm21D1oLyNvqlMVJRQJm@[84.154.121.171]) by fwd33.t-online.de with (TLSv1.2:ECDHE-RSA-AES256-GCM-SHA384 encrypted) esmtp id 1e8lGz-0MXSXw0; Sun, 29 Oct 2017 11:56:57 +0100 Subject: [PATCH] Re: softdep as a mount(8) option From: Stefan Esser To: Kirk McKusick , Mark Johnston Cc: freebsd-fs@FreeBSD.org References: <201710282225.v9SMPDCZ074228@chez.mckusick.com> Message-ID: <0bc1fba0-cda6-064a-add5-02f702e533f7@freebsd.org> Date: Sun, 29 Oct 2017 11:56:57 +0100 User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.13; rv:52.0) Gecko/20100101 Thunderbird/52.4.0 MIME-Version: 1.0 In-Reply-To: Content-Type: multipart/mixed; boundary="------------772410BE6C4FC62BABB50287" Content-Language: de-DE X-ID: r1-YDvZE8htzYMhDyrh7hpKNSEfP0y7R2lUtXcbZ-JnbELUm21D1oLyNvqlMVJRQJm X-TOI-MSGID: d1ed271d-aba3-480b-bac7-0b6fc3ca9afc X-BeenThere: freebsd-fs@freebsd.org X-Mailman-Version: 2.1.23 Precedence: list List-Id: Filesystems List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sun, 29 Oct 2017 10:57:10 -0000 This is a multi-part message in MIME format. --------------772410BE6C4FC62BABB50287 Content-Type: text/plain; charset=windows-1252 Content-Transfer-Encoding: 7bit Am 29.10.17 um 11:36 schrieb Stefan Esser: > Hi Kirk, > > I had implemented this feature more than 10 years ago and even got your > reply, that this was a better solution than the mount option, when I > discussed the implementation with you at that time. (That was shortly > after soft-updates made it into FreeBSD - a long time ago ...) > > But despite your agreement that I should commit that change (for the > next major FreeBSD release), it was later considered a violation of POLA > and I kept (and maintained it) on my system only. > > > I had answered the same question regarding FSCK more than 10 years ago > in the following archived message: > > https://lists.freebsd.org/pipermail/freebsd-hackers/2005-December/014865.html > > > My solution was very simple: The mount option sets the soft-updetes flag > in the super-block, if a file system is mounted RW with soft-updates > enabled. If it is not cleanly unmounted, then FSCK will correctly check > it even on a system that does not support soft-updates in the kernel > (but in FSCK, where it's not optional). > > When such file system is mounted on a system without my patch, it is > indistinguishable from a drive that had the soft-updates flag set via > tunefs. > > In fact, my patch made the kernel do the equivalent of tunefs -j on > each mount, depending on the mount option. The only special case is that > mount -u must not change the soft-updates state on a file-system that > is mounted RW (but tunefs cannot do this, too), since I did not want to > implement the required flushing of all buffers before switching the > mode. > > I have stopped using UFS (have switched to ZFS) a few years ago, but > have maintained my changes in my local version of -CURRENT to this date. > > (For that reason I had no interest in SU+J and there might be a need to > adapt the code to correctly work with SU+J ...) > > Regards, STefan > > PS: I have attached my patch to this mail, will be stripped by the > mail-list, but may be of interest to the direct recipients ... I noticed that patch missed one non-critical part, which allows to update the soft-updates state with mount -u on a R/O mounted file- system (e.g. / when booting in single-user mode). Seems that the patch made it to the list, so I'm attaching the full diff (now with the part missed in the previous version) again. Maybe it is of use to somebody ... Regards, STefan --------------772410BE6C4FC62BABB50287 Content-Type: text/plain; charset=UTF-8; x-mac-type="0"; x-mac-creator="0"; name="soft-updates-option.diff" Content-Transfer-Encoding: 7bit Content-Disposition: attachment; filename="soft-updates-option.diff" Index: sys/ufs/ffs/ffs_vfsops.c =================================================================== --- sys/ufs/ffs/ffs_vfsops.c (Revision 325051) +++ sys/ufs/ffs/ffs_vfsops.c (Arbeitskopie) @@ -138,7 +138,7 @@ */ static const char *ffs_opts[] = { "acls", "async", "noatime", "noclusterr", "noclusterw", "noexec", "export", "force", "from", "groupquota", - "multilabel", "nfsv4acls", "fsckpid", "snapshot", "nosuid", "suiddir", + "multilabel", "nfsv4acls", "fsckpid", "snapshot", "softdep", "nosuid", "suiddir", "nosymfollow", "sync", "union", "userquota", NULL }; static int @@ -237,17 +237,26 @@ ump = VFSTOUFS(mp); fs = ump->um_fs; devvp = ump->um_devvp; if (fsckpid == -1 && ump->um_fsckpid > 0) { if ((error = ffs_flushfiles(mp, WRITECLOSE, td)) != 0 || (error = ffs_sbupdate(ump, MNT_WAIT, 0)) != 0) return (error); g_topology_lock(); + if (fs->fs_ronly == 0) { /* - * Return to normal read-only mode. + * Preserve current softdep status, + * unless mounted read-only. */ + MNT_ILOCK(mp); + if (fs->fs_flags & FS_DOSOFTDEP) { + mp->mnt_flag |= MNT_SOFTDEP; + } else { + mp->mnt_flag &= ~MNT_SOFTDEP; + } + MNT_IUNLOCK(mp); error = g_access(ump->um_cp, 0, -1, 0); g_topology_unlock(); ump->um_fsckpid = 0; } if (fs->fs_ronly == 0 && vfs_flagopt(mp->mnt_optnew, "ro", NULL, 0)) { @@ -266,7 +265,8 @@ flags = WRITECLOSE; if (mp->mnt_flag & MNT_FORCE) flags |= FORCECLOSE; - if (MOUNTEDSOFTDEP(mp)) { +// if (MOUNTEDSOFTDEP(mp)) { + if (fs->fs_flags & FS_DOSOFTDEP) { error = softdep_flushfiles(mp, flags, td); } else { error = ffs_flushfiles(mp, flags, td); @@ -346,6 +346,15 @@ (fs->fs_flags & FS_DOSOFTDEP))) { printf("WARNING: %s was not properly " "dismounted\n", fs->fs_fsmnt); + /* + * Preserve softdep state if unclean + */ + MNT_ILOCK(mp); + if ((fs->fs_flags & FS_DOSOFTDEP) == 0) + mp->mnt_flag &= ~MNT_SOFTDEP; + else + mp->mnt_flag |= MNT_SOFTDEP; + MNT_IUNLOCK(mp); } else { vfs_mount_error(mp, "R/W mount of %s denied. %s.%s", @@ -373,7 +380,7 @@ MNT_IUNLOCK(mp); fs->fs_mtime = time_second; /* check to see if we need to start softdep */ - if ((fs->fs_flags & FS_DOSOFTDEP) && + if ((mp->mnt_flag & MNT_SOFTDEP) && (error = softdep_mount(devvp, mp, fs, td->td_ucred))){ vn_finished_write(mp); return (error); @@ -851,6 +858,15 @@ (fs->fs_flags & FS_DOSOFTDEP))) { printf("WARNING: %s was not properly dismounted\n", fs->fs_fsmnt); + /* + * Preserve softdep state if unclean. + */ + MNT_ILOCK(mp); + if ((fs->fs_flags & FS_DOSOFTDEP) == 0) + mp->mnt_flag &= ~MNT_SOFTDEP; + else + mp->mnt_flag |= MNT_SOFTDEP; + MNT_IUNLOCK(mp); } else { vfs_mount_error(mp, "R/W mount of %s denied. %s%s", fs->fs_fsmnt, "Filesystem is not clean - run fsck.", @@ -875,6 +889,11 @@ fs->fs_pendingblocks = 0; fs->fs_pendinginodes = 0; } + /* Clear softdep flag in superblock, if not a softdep mount. */ + MNT_ILOCK(mp); + if ((mp->mnt_flag & MNT_SOFTDEP) == 0) + fs->fs_flags &= ~FS_DOSOFTDEP; + MNT_IUNLOCK(mp); if ((fs->fs_flags & FS_GJOURNAL) != 0) { #ifdef UFS_GJOURNAL /* @@ -1071,7 +1088,7 @@ if (ronly == 0) { fs->fs_mtime = time_second; - if ((fs->fs_flags & FS_DOSOFTDEP) && + if ((mp->mnt_flag & MNT_SOFTDEP) && (error = softdep_mount(devvp, mp, fs, cred)) != 0) { free(fs->fs_csp, M_UFSMNT); ffs_flushfiles(mp, FORCECLOSE, td); Index: sbin/mount/mntopts.h =================================================================== --- sbin/mount/mntopts.h (Revision 325058) +++ sbin/mount/mntopts.h (Arbeitskopie) @@ -52,6 +52,7 @@ #define MOPT_NOCLUSTERW { "clusterw", 1, MNT_NOCLUSTERW, 0 } #define MOPT_SUIDDIR { "suiddir", 0, MNT_SUIDDIR, 0 } #define MOPT_SNAPSHOT { "snapshot", 0, MNT_SNAPSHOT, 0 } +#define MOPT_SOFTDEP { "softdep", 0, MNT_SOFTDEP, 0 } #define MOPT_MULTILABEL { "multilabel", 0, MNT_MULTILABEL, 0 } #define MOPT_ACLS { "acls", 0, MNT_ACLS, 0 } #define MOPT_NFS4ACLS { "nfsv4acls", 0, MNT_NFS4ACLS, 0 } Index: sbin/mount/mount.c =================================================================== --- sbin/mount/mount.c (Revision 325058) +++ sbin/mount/mount.c (Arbeitskopie) @@ -958,6 +958,7 @@ if (flags & MNT_NOCLUSTERW) res = catopt(res, "noclusterw"); if (flags & MNT_NOSYMFOLLOW) res = catopt(res, "nosymfollow"); if (flags & MNT_SUIDDIR) res = catopt(res, "suiddir"); + if (flags & MNT_SOFTDEP) res = catopt(res, "softdep"); if (flags & MNT_MULTILABEL) res = catopt(res, "multilabel"); if (flags & MNT_ACLS) res = catopt(res, "acls"); if (flags & MNT_NFS4ACLS) res = catopt(res, "nfsv4acls"); Index: sys/sys/mount.h =================================================================== --- sys/sys/mount.h (Revision 325058) +++ sys/sys/mount.h (Arbeitskopie) @@ -337,6 +337,7 @@ MNT_SYNCHRONOUS | MNT_UNION | MNT_ASYNC | \ MNT_NOATIME | \ MNT_NOSYMFOLLOW | MNT_IGNORE | \ + MNT_SOFTDEP | \ MNT_NOCLUSTERR | MNT_NOCLUSTERW | MNT_SUIDDIR | \ MNT_ACLS | MNT_USER | MNT_NFS4ACLS | \ MNT_AUTOMOUNTED) --------------772410BE6C4FC62BABB50287-- From owner-freebsd-fs@freebsd.org Sun Oct 29 11:25:42 2017 Return-Path: Delivered-To: freebsd-fs@mailman.ysv.freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:1900:2254:206a::19:1]) by mailman.ysv.freebsd.org (Postfix) with ESMTP id 25FBAE60DA9 for ; Sun, 29 Oct 2017 11:25:42 +0000 (UTC) (envelope-from se@freebsd.org) Received: from mailout09.t-online.de (mailout09.t-online.de [194.25.134.84]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (Client CN "mailout00.t-online.de", Issuer "TeleSec ServerPass DE-2" (verified OK)) by mx1.freebsd.org (Postfix) with ESMTPS id DDCDD7617C; Sun, 29 Oct 2017 11:25:41 +0000 (UTC) (envelope-from se@freebsd.org) Received: from fwd23.aul.t-online.de (fwd23.aul.t-online.de [172.20.26.128]) by mailout09.t-online.de (Postfix) with SMTP id E3A9E42756C3; Sun, 29 Oct 2017 12:25:38 +0100 (CET) Received: from Stefans-MBP-2.fritz.box (Tl8UDeZLZhJZL9qtz0Pg5UelmE0NCMwW5qt0qkJpgBZ1eePsHbihDwGImqto2uggD3@[84.154.121.171]) by fwd23.t-online.de with (TLSv1.2:ECDHE-RSA-AES256-GCM-SHA384 encrypted) esmtp id 1e8lij-28eAim0; Sun, 29 Oct 2017 12:25:37 +0100 Subject: Re: softdep as a mount(8) option To: Kirk McKusick , Mark Johnston Cc: freebsd-fs@FreeBSD.org References: <201710282225.v9SMPDCZ074228@chez.mckusick.com> From: Stefan Esser Message-ID: <659abe72-69f9-8e89-24ba-36a9fdfb3fd5@freebsd.org> Date: Sun, 29 Oct 2017 12:25:37 +0100 User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.13; rv:52.0) Gecko/20100101 Thunderbird/52.4.0 MIME-Version: 1.0 In-Reply-To: <201710282225.v9SMPDCZ074228@chez.mckusick.com> Content-Type: text/plain; charset=windows-1252 Content-Language: en-GB Content-Transfer-Encoding: 7bit X-ID: Tl8UDeZLZhJZL9qtz0Pg5UelmE0NCMwW5qt0qkJpgBZ1eePsHbihDwGImqto2uggD3 X-TOI-MSGID: 2e3bb460-f905-4a04-a334-b2271149dcd6 X-BeenThere: freebsd-fs@freebsd.org X-Mailman-Version: 2.1.23 Precedence: list List-Id: Filesystems List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sun, 29 Oct 2017 11:25:42 -0000 Am 29.10.17 um 00:25 schrieb Kirk McKusick: >> Date: Fri, 27 Oct 2017 11:39:00 -0400 >> From: Mark Johnston >> To: freebsd-fs@FreeBSD.org >> Subject: softdep as a mount(8) option >> >> Hi, >> >> I'd like to finally enable the use of SU (not SU+J) on some small UFS >> filesystems. The fact that SU is enabled using a flag in the superblock >> poses a problem for me, however: the systems containing these >> filesystems may at any time be repurposed to run a kernel that supports >> SU but contains bugs[*] that render it unusable. I therefore can't >> persistently enable SU in these systems. >> >> I'm wondering if it would be possible to enable SU using a mount >> option rather than with a persistent flag. fsck_ffs conditionalizes some >> of its logic on whether SU is configured - is this necessary for >> correctness? That is, if I run fsck on an unclean filesystem that had >> been mounted with SU, and fsck runs as though SU hadn't been configured, >> what problems might arise? >> >> [*] These bugs are a result of local modifications and aren't in >> FreeBSD. > > While it is safe and possible to add soft-updates (but not journalled > soft updates) as a mount option, it means that fsck will not know that > soft updates were in use, so it will always run in full (slow) mode at > boot time. This is why I have not added it as an option. Hi Kirk, this is a link to the message where you agreed to make SU a mount option: http://docs.freebsd.org/cgi/mid.cgi?200006282051.NAA05776 Seems you were (rightly) not too worried about FSCK, back then ;-) Regards, STefan From owner-freebsd-fs@freebsd.org Sun Oct 29 18:28:25 2017 Return-Path: Delivered-To: freebsd-fs@mailman.ysv.freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:1900:2254:206a::19:1]) by mailman.ysv.freebsd.org (Postfix) with ESMTP id AFEC1E49496 for ; Sun, 29 Oct 2017 18:28:25 +0000 (UTC) (envelope-from mckusick@mckusick.com) Received: from chez.mckusick.com (chez.mckusick.com [70.36.157.235]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (Client did not present a certificate) by mx1.freebsd.org (Postfix) with ESMTPS id 967113077; Sun, 29 Oct 2017 18:28:25 +0000 (UTC) (envelope-from mckusick@mckusick.com) Received: from chez.mckusick.com (localhost [IPv6:::1]) by chez.mckusick.com (8.15.2/8.15.2) with ESMTP id v9TIPjRE095791; Sun, 29 Oct 2017 11:25:45 -0700 (PDT) (envelope-from mckusick@chez.mckusick.com) Message-Id: <201710291825.v9TIPjRE095791@chez.mckusick.com> From: Kirk McKusick To: Stefan Esser Subject: Re: softdep as a mount(8) option cc: Mark Johnston , freebsd-fs@FreeBSD.org In-reply-to: <659abe72-69f9-8e89-24ba-36a9fdfb3fd5@freebsd.org> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-ID: <95789.1509301545.1@chez.mckusick.com> Date: Sun, 29 Oct 2017 11:25:45 -0700 X-Spam-Status: No, score=-1.4 required=5.0 tests=BAYES_00,MISSING_MID, UNPARSEABLE_RELAY autolearn=no autolearn_force=no version=3.4.1 X-Spam-Checker-Version: SpamAssassin 3.4.1 (2015-04-28) on chez.mckusick.com X-BeenThere: freebsd-fs@freebsd.org X-Mailman-Version: 2.1.23 Precedence: list List-Id: Filesystems List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sun, 29 Oct 2017 18:28:25 -0000 Your patch will require some additional work now that we have journalled soft updates. The issue is that when journalled soft updates are in use, the FS_DOSOFTDEP flag will be set for the filesystem as it is a prerequisite for journalling. If you turn off the FS_DOSOFTDEP flag while journalling is enabled very bad things will happen (hopefully just a kernel panic and not irrecoverable filesystem state). So, additional changes are needed to check for the FS_SUJ flag being set, and if it is, to disallow the disabling of the FS_DOSOFTDEP flag by the mount of the filesystem. Kirk McKusick From owner-freebsd-fs@freebsd.org Tue Oct 31 12:08:07 2017 Return-Path: Delivered-To: freebsd-fs@mailman.ysv.freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:1900:2254:206a::19:1]) by mailman.ysv.freebsd.org (Postfix) with ESMTP id 05873E57EDF for ; Tue, 31 Oct 2017 12:08:07 +0000 (UTC) (envelope-from bugzilla-noreply@freebsd.org) Received: from kenobi.freebsd.org (kenobi.freebsd.org [IPv6:2001:1900:2254:206a::16:76]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (Client did not present a certificate) by mx1.freebsd.org (Postfix) with ESMTPS id E77547268F for ; Tue, 31 Oct 2017 12:08:06 +0000 (UTC) (envelope-from bugzilla-noreply@freebsd.org) Received: from bugs.freebsd.org ([127.0.1.118]) by kenobi.freebsd.org (8.15.2/8.15.2) with ESMTP id v9VC86BX073787 for ; Tue, 31 Oct 2017 12:08:06 GMT (envelope-from bugzilla-noreply@freebsd.org) From: bugzilla-noreply@freebsd.org To: freebsd-fs@FreeBSD.org Subject: [Bug 223287] posix_fallocate(2) attempts to allocate impossibly large files and cannot be killed on ZFS Date: Tue, 31 Oct 2017 12:08:07 +0000 X-Bugzilla-Reason: AssignedTo X-Bugzilla-Type: changed X-Bugzilla-Watch-Reason: None X-Bugzilla-Product: Base System X-Bugzilla-Component: kern X-Bugzilla-Version: 11.0-RELEASE X-Bugzilla-Keywords: X-Bugzilla-Severity: Affects Some People X-Bugzilla-Who: avg@FreeBSD.org X-Bugzilla-Status: New X-Bugzilla-Resolution: X-Bugzilla-Priority: --- X-Bugzilla-Assigned-To: freebsd-fs@FreeBSD.org X-Bugzilla-Flags: X-Bugzilla-Changed-Fields: Message-ID: In-Reply-To: References: Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable X-Bugzilla-URL: https://bugs.freebsd.org/bugzilla/ Auto-Submitted: auto-generated MIME-Version: 1.0 X-BeenThere: freebsd-fs@freebsd.org X-Mailman-Version: 2.1.23 Precedence: list List-Id: Filesystems List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 31 Oct 2017 12:08:07 -0000 https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=3D223287 --- Comment #1 from Andriy Gapon --- I plan on disabling posix_fallocate for ZFS altogether, same as on illumos. --=20 You are receiving this mail because: You are the assignee for the bug.= From owner-freebsd-fs@freebsd.org Tue Oct 31 16:42:40 2017 Return-Path: Delivered-To: freebsd-fs@mailman.ysv.freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:1900:2254:206a::19:1]) by mailman.ysv.freebsd.org (Postfix) with ESMTP id 5AF7CE5E7C0 for ; Tue, 31 Oct 2017 16:42:40 +0000 (UTC) (envelope-from bugzilla-noreply@freebsd.org) Received: from kenobi.freebsd.org (kenobi.freebsd.org [IPv6:2001:1900:2254:206a::16:76]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (Client did not present a certificate) by mx1.freebsd.org (Postfix) with ESMTPS id 4984980C5C for ; Tue, 31 Oct 2017 16:42:40 +0000 (UTC) (envelope-from bugzilla-noreply@freebsd.org) Received: from bugs.freebsd.org ([127.0.1.118]) by kenobi.freebsd.org (8.15.2/8.15.2) with ESMTP id v9VGgebq091819 for ; Tue, 31 Oct 2017 16:42:40 GMT (envelope-from bugzilla-noreply@freebsd.org) From: bugzilla-noreply@freebsd.org To: freebsd-fs@FreeBSD.org Subject: [Bug 221077] Boot from ZFS fails following freebsd-update Date: Tue, 31 Oct 2017 16:42:40 +0000 X-Bugzilla-Reason: AssignedTo X-Bugzilla-Type: changed X-Bugzilla-Watch-Reason: None X-Bugzilla-Product: Base System X-Bugzilla-Component: kern X-Bugzilla-Version: 10.3-RELEASE X-Bugzilla-Keywords: regression X-Bugzilla-Severity: Affects Only Me X-Bugzilla-Who: bacon4000@gmail.com X-Bugzilla-Status: New X-Bugzilla-Resolution: X-Bugzilla-Priority: --- X-Bugzilla-Assigned-To: freebsd-fs@FreeBSD.org X-Bugzilla-Flags: X-Bugzilla-Changed-Fields: Message-ID: In-Reply-To: References: Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable X-Bugzilla-URL: https://bugs.freebsd.org/bugzilla/ Auto-Submitted: auto-generated MIME-Version: 1.0 X-BeenThere: freebsd-fs@freebsd.org X-Mailman-Version: 2.1.23 Precedence: list List-Id: Filesystems List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 31 Oct 2017 16:42:40 -0000 https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=3D221077 --- Comment #9 from Jason Bacon --- Add one more data point: I experienced the issue on a Proliant server with hardware RAID 1 immediately after installation. Switching to UFS2 resolved= the issue. I have still not seen a problem where ZFS is doing the volume management, so this is looking more like an issue of ZFS on top of hardware RAIDs. --=20 You are receiving this mail because: You are the assignee for the bug.= From owner-freebsd-fs@freebsd.org Tue Oct 31 18:59:20 2017 Return-Path: Delivered-To: freebsd-fs@mailman.ysv.freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:1900:2254:206a::19:1]) by mailman.ysv.freebsd.org (Postfix) with ESMTP id AC2E7E60F88 for ; Tue, 31 Oct 2017 18:59:20 +0000 (UTC) (envelope-from bugzilla-noreply@freebsd.org) Received: from kenobi.freebsd.org (kenobi.freebsd.org [IPv6:2001:1900:2254:206a::16:76]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (Client did not present a certificate) by mx1.freebsd.org (Postfix) with ESMTPS id 998436AA for ; Tue, 31 Oct 2017 18:59:20 +0000 (UTC) (envelope-from bugzilla-noreply@freebsd.org) Received: from bugs.freebsd.org ([127.0.1.118]) by kenobi.freebsd.org (8.15.2/8.15.2) with ESMTP id v9VIxJES043242 for ; Tue, 31 Oct 2017 18:59:20 GMT (envelope-from bugzilla-noreply@freebsd.org) From: bugzilla-noreply@freebsd.org To: freebsd-fs@FreeBSD.org Subject: [Bug 223348] Inode number for directories change to inode number of directory two levels up causing hierarchy loop Date: Tue, 31 Oct 2017 18:59:20 +0000 X-Bugzilla-Reason: AssignedTo X-Bugzilla-Type: changed X-Bugzilla-Watch-Reason: None X-Bugzilla-Product: Base System X-Bugzilla-Component: gnu X-Bugzilla-Version: CURRENT X-Bugzilla-Keywords: X-Bugzilla-Severity: Affects Only Me X-Bugzilla-Who: linimon@FreeBSD.org X-Bugzilla-Status: New X-Bugzilla-Resolution: X-Bugzilla-Priority: --- X-Bugzilla-Assigned-To: freebsd-fs@FreeBSD.org X-Bugzilla-Flags: X-Bugzilla-Changed-Fields: assigned_to Message-ID: In-Reply-To: References: Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable X-Bugzilla-URL: https://bugs.freebsd.org/bugzilla/ Auto-Submitted: auto-generated MIME-Version: 1.0 X-BeenThere: freebsd-fs@freebsd.org X-Mailman-Version: 2.1.23 Precedence: list List-Id: Filesystems List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 31 Oct 2017 18:59:20 -0000 https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=3D223348 Mark Linimon changed: What |Removed |Added ---------------------------------------------------------------------------- Assignee|freebsd-bugs@FreeBSD.org |freebsd-fs@FreeBSD.org --=20 You are receiving this mail because: You are the assignee for the bug.= From owner-freebsd-fs@freebsd.org Tue Oct 31 19:41:04 2017 Return-Path: Delivered-To: freebsd-fs@mailman.ysv.freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:1900:2254:206a::19:1]) by mailman.ysv.freebsd.org (Postfix) with ESMTP id 86C26E61E3E for ; Tue, 31 Oct 2017 19:41:04 +0000 (UTC) (envelope-from bugzilla-noreply@freebsd.org) Received: from kenobi.freebsd.org (kenobi.freebsd.org [IPv6:2001:1900:2254:206a::16:76]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (Client did not present a certificate) by mx1.freebsd.org (Postfix) with ESMTPS id 74C75244D for ; Tue, 31 Oct 2017 19:41:04 +0000 (UTC) (envelope-from bugzilla-noreply@freebsd.org) Received: from bugs.freebsd.org ([127.0.1.118]) by kenobi.freebsd.org (8.15.2/8.15.2) with ESMTP id v9VJf4os046197 for ; Tue, 31 Oct 2017 19:41:04 GMT (envelope-from bugzilla-noreply@freebsd.org) From: bugzilla-noreply@freebsd.org To: freebsd-fs@FreeBSD.org Subject: [Bug 223348] Inode number for directories change to inode number of directory two levels up causing hierarchy loop Date: Tue, 31 Oct 2017 19:41:04 +0000 X-Bugzilla-Reason: AssignedTo X-Bugzilla-Type: changed X-Bugzilla-Watch-Reason: None X-Bugzilla-Product: Base System X-Bugzilla-Component: gnu X-Bugzilla-Version: CURRENT X-Bugzilla-Keywords: X-Bugzilla-Severity: Affects Only Me X-Bugzilla-Who: avg@FreeBSD.org X-Bugzilla-Status: New X-Bugzilla-Resolution: X-Bugzilla-Priority: --- X-Bugzilla-Assigned-To: freebsd-fs@FreeBSD.org X-Bugzilla-Flags: X-Bugzilla-Changed-Fields: Message-ID: In-Reply-To: References: Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable X-Bugzilla-URL: https://bugs.freebsd.org/bugzilla/ Auto-Submitted: auto-generated MIME-Version: 1.0 X-BeenThere: freebsd-fs@freebsd.org X-Mailman-Version: 2.1.23 Precedence: list List-Id: Filesystems List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 31 Oct 2017 19:41:04 -0000 https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=3D223348 --- Comment #1 from Andriy Gapon --- Another hypothesis is that the name cache may have something to do with it. --=20 You are receiving this mail because: You are the assignee for the bug.= From owner-freebsd-fs@freebsd.org Tue Oct 31 21:45:04 2017 Return-Path: Delivered-To: freebsd-fs@mailman.ysv.freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:1900:2254:206a::19:1]) by mailman.ysv.freebsd.org (Postfix) with ESMTP id 7C88BE63D0C for ; Tue, 31 Oct 2017 21:45:04 +0000 (UTC) (envelope-from matthew.ahrens@delphix.com) Received: from mail-lf0-x236.google.com (mail-lf0-x236.google.com [IPv6:2a00:1450:4010:c07::236]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (Client CN "smtp.gmail.com", Issuer "Google Internet Authority G2" (verified OK)) by mx1.freebsd.org (Postfix) with ESMTPS id E725365317 for ; Tue, 31 Oct 2017 21:45:03 +0000 (UTC) (envelope-from matthew.ahrens@delphix.com) Received: by mail-lf0-x236.google.com with SMTP id e143so433431lfg.12 for ; Tue, 31 Oct 2017 14:45:03 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=delphix.com; s=google; h=mime-version:from:date:message-id:subject:to; bh=B/JTm4Y9BinFJQxU8o8/5EcLhyMy7esixFHtz/9LGts=; b=UQsuB/5X80hsxoVtB5sGBHjxS9svrUAKCyPQJFX1exhCwhHY0sI3U114D2B7SGZ0Wu eUSVUoDJua9xKdl94B5JW82Ys5qor/6eto0q1PRbtU2IdXP/2sbP9baOAX0Hwgj0VuYk 4aF27UJ6YCHNpq1BHcDUVawDwQUnOAUJQwx3E= X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:mime-version:from:date:message-id:subject:to; bh=B/JTm4Y9BinFJQxU8o8/5EcLhyMy7esixFHtz/9LGts=; b=IwNBY0fUMJ1pi6g9Hv1nYCnWiuFNL/ncOINWRb6sHy0BBDLANGfXrxmxSzXxhyf+W2 NyV1glshWlQ/ZkQqmwoVl/tsKYQJt5NNqvD34CKplEYzQZtTYA5lOnxDqP2+zna+vKCE Ng8nXQJEoVbGGMaKWf6PBV0ZoGBh2lBw43sjlXbU5fHgx7R1G+X3TAaKtjL3Dpy9lTZc of173OdIZCdvx3hYNv7Z6lWs7dnMHChneOIylanUUmX2kWO+E0b9Qu6pZ5Ego2oRsWco js1qvm5KJh1Cu23OWfwFJDFyMD8H2d4XBwvvO3B37h1UOlTpNflMvYwctzRlexf2PwTv a8MA== X-Gm-Message-State: AMCzsaVUB3Bw5eH78cRMFt0cwztpP5JkNmsHeSAiKvpw2UtJyqDmNYWJ +1AuoIMdQLRLEpHgA078W1vho4nuC/DQAamQ1ZDOHg== X-Google-Smtp-Source: ABhQp+R6klxTJTe0oj/Voa53nXeZEb0J8g9ZkVzv3Gl38RT7vDRo0hNZC0osDn9/6/tSwsIxB2/G4QB3nXDsngU3Y1U= X-Received: by 10.25.115.14 with SMTP id o14mr1114668lfc.79.1509486300730; Tue, 31 Oct 2017 14:45:00 -0700 (PDT) MIME-Version: 1.0 Received: by 10.25.198.71 with HTTP; Tue, 31 Oct 2017 14:44:59 -0700 (PDT) From: Matthew Ahrens Date: Tue, 31 Oct 2017 14:44:59 -0700 Message-ID: Subject: OpenZFS Developer Summit 2017 - slides & video posted To: "admin@open-zfs.org" , developer , illumos-zfs , zfs-devel@list.zfsonlinux.org, zfs-devel@freebsd.org, freebsd-fs , zfs-discuss Content-Type: text/plain; charset="UTF-8" X-Content-Filtered-By: Mailman/MimeDel 2.1.23 X-BeenThere: freebsd-fs@freebsd.org X-Mailman-Version: 2.1.23 Precedence: list List-Id: Filesystems List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 31 Oct 2017 21:45:04 -0000 Thank you to our speakers and participants for making this year's OpenZFS Developer Summit truly excellent! The slides and videos are now posted on our website, and reproduced below: http://www.open-zfs.org/wiki/OpenZFS_Developer_Summit_2017 TitleSpeakerCompanyVideoSlides State of the Union Matt Ahrens Delphix Video Slides Keynote: ZFS Past & Future Mark Maybee Oracle Not Available N/A ZSTD Compression Allan Jude ScaleEngine Video Slides Fast Clone Deletion Sara Hartse Delphix Video Slides MMP: Safe "zpool import" for Clusters Olaf Faaland LLNL Video Slides Porting With OSX Jorgen Lundman GMO Video Slides Faster Allocation with the Log Spacemap Serapheim Dimitropoulos Delphix Video Slides iFlash: Dynamic Adaptive L2ARC Caching Shailendra Tripathi Tegile Video Slides DRAID Isaac Huang Intel Video Slides ZIL Performance: How I Doubled Sync Write Speed Prakash Surya Delphix Video Slides TitleSpeakerCompanyVideoSlides "Oh Shift!" changing the allocation size George Wilson Delphix Video Slides A proposal for 1,000x better dedup performance Matt Ahrens Delphix Video Slides New prefetcher for sequential scrub Tom Caputi Datto Video Slides Storage Pool Checkpoint Serapheim Dimitropoulos Delphix Video Slides Improving resilver: results & operational impacts Saso Kiselkov Nexenta Video Slides RAID-Z Expansion Matt Ahrens Delphix + FreeBSD Foundation Video Slides From owner-freebsd-fs@freebsd.org Wed Nov 1 07:46:17 2017 Return-Path: Delivered-To: freebsd-fs@mailman.ysv.freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:1900:2254:206a::19:1]) by mailman.ysv.freebsd.org (Postfix) with ESMTP id 0A674E66ACD for ; Wed, 1 Nov 2017 07:46:17 +0000 (UTC) (envelope-from bugzilla-noreply@freebsd.org) Received: from kenobi.freebsd.org (kenobi.freebsd.org [IPv6:2001:1900:2254:206a::16:76]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (Client did not present a certificate) by mx1.freebsd.org (Postfix) with ESMTPS id EC6A3228F for ; Wed, 1 Nov 2017 07:46:16 +0000 (UTC) (envelope-from bugzilla-noreply@freebsd.org) Received: from bugs.freebsd.org ([127.0.1.118]) by kenobi.freebsd.org (8.15.2/8.15.2) with ESMTP id vA17kGGV067112 for ; Wed, 1 Nov 2017 07:46:16 GMT (envelope-from bugzilla-noreply@freebsd.org) From: bugzilla-noreply@freebsd.org To: freebsd-fs@FreeBSD.org Subject: [Bug 223348] Inode number for directories change to inode number of directory two levels up causing hierarchy loop Date: Wed, 01 Nov 2017 07:46:17 +0000 X-Bugzilla-Reason: AssignedTo X-Bugzilla-Type: changed X-Bugzilla-Watch-Reason: None X-Bugzilla-Product: Base System X-Bugzilla-Component: gnu X-Bugzilla-Version: CURRENT X-Bugzilla-Keywords: X-Bugzilla-Severity: Affects Only Me X-Bugzilla-Who: avg@FreeBSD.org X-Bugzilla-Status: New X-Bugzilla-Resolution: X-Bugzilla-Priority: --- X-Bugzilla-Assigned-To: freebsd-fs@FreeBSD.org X-Bugzilla-Flags: X-Bugzilla-Changed-Fields: Message-ID: In-Reply-To: References: Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable X-Bugzilla-URL: https://bugs.freebsd.org/bugzilla/ Auto-Submitted: auto-generated MIME-Version: 1.0 X-BeenThere: freebsd-fs@freebsd.org X-Mailman-Version: 2.1.23 Precedence: list List-Id: Filesystems List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 01 Nov 2017 07:46:17 -0000 https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=3D223348 --- Comment #2 from Andriy Gapon --- (In reply to Andriy Gapon from comment #1) To expand on my previous comment, I mean that your report is the first of i= ts kind and the affected directories have rather peculiar names of form "?.". = So, my suspicion is that some generic lookup or cache code may have a flaw of misinterpreting those names as "..". --=20 You are receiving this mail because: You are the assignee for the bug.= From owner-freebsd-fs@freebsd.org Wed Nov 1 08:41:16 2017 Return-Path: Delivered-To: freebsd-fs@mailman.ysv.freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:1900:2254:206a::19:1]) by mailman.ysv.freebsd.org (Postfix) with ESMTP id BD10EE4D447 for ; Wed, 1 Nov 2017 08:41:16 +0000 (UTC) (envelope-from bugzilla-noreply@freebsd.org) Received: from kenobi.freebsd.org (kenobi.freebsd.org [IPv6:2001:1900:2254:206a::16:76]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (Client did not present a certificate) by mx1.freebsd.org (Postfix) with ESMTPS id AAE1F3978 for ; Wed, 1 Nov 2017 08:41:16 +0000 (UTC) (envelope-from bugzilla-noreply@freebsd.org) Received: from bugs.freebsd.org ([127.0.1.118]) by kenobi.freebsd.org (8.15.2/8.15.2) with ESMTP id vA18fGxP097889 for ; Wed, 1 Nov 2017 08:41:16 GMT (envelope-from bugzilla-noreply@freebsd.org) From: bugzilla-noreply@freebsd.org To: freebsd-fs@FreeBSD.org Subject: [Bug 223348] Inode number for directories change to inode number of directory two levels up causing hierarchy loop Date: Wed, 01 Nov 2017 08:41:16 +0000 X-Bugzilla-Reason: AssignedTo X-Bugzilla-Type: changed X-Bugzilla-Watch-Reason: None X-Bugzilla-Product: Base System X-Bugzilla-Component: gnu X-Bugzilla-Version: CURRENT X-Bugzilla-Keywords: X-Bugzilla-Severity: Affects Only Me X-Bugzilla-Who: mjg@FreeBSD.org X-Bugzilla-Status: New X-Bugzilla-Resolution: X-Bugzilla-Priority: --- X-Bugzilla-Assigned-To: freebsd-fs@FreeBSD.org X-Bugzilla-Flags: X-Bugzilla-Changed-Fields: Message-ID: In-Reply-To: References: Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable X-Bugzilla-URL: https://bugs.freebsd.org/bugzilla/ Auto-Submitted: auto-generated MIME-Version: 1.0 X-BeenThere: freebsd-fs@freebsd.org X-Mailman-Version: 2.1.23 Precedence: list List-Id: Filesystems List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 01 Nov 2017 08:41:16 -0000 https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=3D223348 --- Comment #5 from Mateusz Guzik --- The obvious bug is fixed in https://svnweb.freebsd.org/changeset/base/325268 --=20 You are receiving this mail because: You are the assignee for the bug.= From owner-freebsd-fs@freebsd.org Wed Nov 1 07:53:50 2017 Return-Path: Delivered-To: freebsd-fs@mailman.ysv.freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:1900:2254:206a::19:1]) by mailman.ysv.freebsd.org (Postfix) with ESMTP id E618DE67505 for ; Wed, 1 Nov 2017 07:53:50 +0000 (UTC) (envelope-from bugzilla-noreply@freebsd.org) Received: from kenobi.freebsd.org (kenobi.freebsd.org [IPv6:2001:1900:2254:206a::16:76]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (Client did not present a certificate) by mx1.freebsd.org (Postfix) with ESMTPS id D471026F0 for ; Wed, 1 Nov 2017 07:53:50 +0000 (UTC) (envelope-from bugzilla-noreply@freebsd.org) Received: from bugs.freebsd.org ([127.0.1.118]) by kenobi.freebsd.org (8.15.2/8.15.2) with ESMTP id vA17roHi087274 for ; Wed, 1 Nov 2017 07:53:50 GMT (envelope-from bugzilla-noreply@freebsd.org) From: bugzilla-noreply@freebsd.org To: freebsd-fs@FreeBSD.org Subject: [Bug 223348] Inode number for directories change to inode number of directory two levels up causing hierarchy loop Date: Wed, 01 Nov 2017 07:53:50 +0000 X-Bugzilla-Reason: AssignedTo X-Bugzilla-Type: changed X-Bugzilla-Watch-Reason: None X-Bugzilla-Product: Base System X-Bugzilla-Component: gnu X-Bugzilla-Version: CURRENT X-Bugzilla-Keywords: X-Bugzilla-Severity: Affects Only Me X-Bugzilla-Who: avg@FreeBSD.org X-Bugzilla-Status: New X-Bugzilla-Resolution: X-Bugzilla-Priority: --- X-Bugzilla-Assigned-To: freebsd-fs@FreeBSD.org X-Bugzilla-Flags: X-Bugzilla-Changed-Fields: cc Message-ID: In-Reply-To: References: Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable X-Bugzilla-URL: https://bugs.freebsd.org/bugzilla/ Auto-Submitted: auto-generated MIME-Version: 1.0 X-BeenThere: freebsd-fs@freebsd.org X-Mailman-Version: 2.1.23 Precedence: list List-Id: Filesystems List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 01 Nov 2017 07:53:51 -0000 https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=3D223348 Andriy Gapon changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |mjg@FreeBSD.org --- Comment #3 from Andriy Gapon --- And when you know what you are looking for: static __noinline int cache_lookup_nomakeentry(struct vnode *dvp, struct vnode **vpp, struct componentname *cnp, struct timespec *tsp, int *ticksp) { struct namecache *ncp; struct rwlock *blp; struct mtx *dvlp, *dvlp2; uint32_t hash; int error; if (cnp->cn_namelen =3D=3D 2 && cnp->cn_nameptr[1] =3D=3D '.') { There is a similar condition in cache_lookup() as well. I do not see any code to pre-check that cnp->cn_nameptr[0] =3D=3D '.' Seems like bad refactoring in r324378. --=20 You are receiving this mail because: You are the assignee for the bug.= From owner-freebsd-fs@freebsd.org Wed Nov 1 07:59:15 2017 Return-Path: Delivered-To: freebsd-fs@mailman.ysv.freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:1900:2254:206a::19:1]) by mailman.ysv.freebsd.org (Postfix) with ESMTP id ED701E67C59 for ; Wed, 1 Nov 2017 07:59:15 +0000 (UTC) (envelope-from bugzilla-noreply@freebsd.org) Received: from kenobi.freebsd.org (kenobi.freebsd.org [IPv6:2001:1900:2254:206a::16:76]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (Client did not present a certificate) by mx1.freebsd.org (Postfix) with ESMTPS id DBAE028D2 for ; Wed, 1 Nov 2017 07:59:15 +0000 (UTC) (envelope-from bugzilla-noreply@freebsd.org) Received: from bugs.freebsd.org ([127.0.1.118]) by kenobi.freebsd.org (8.15.2/8.15.2) with ESMTP id vA17xFYs095093 for ; Wed, 1 Nov 2017 07:59:15 GMT (envelope-from bugzilla-noreply@freebsd.org) From: bugzilla-noreply@freebsd.org To: freebsd-fs@FreeBSD.org Subject: [Bug 223348] Inode number for directories change to inode number of directory two levels up causing hierarchy loop Date: Wed, 01 Nov 2017 07:59:16 +0000 X-Bugzilla-Reason: AssignedTo X-Bugzilla-Type: changed X-Bugzilla-Watch-Reason: None X-Bugzilla-Product: Base System X-Bugzilla-Component: gnu X-Bugzilla-Version: CURRENT X-Bugzilla-Keywords: X-Bugzilla-Severity: Affects Only Me X-Bugzilla-Who: mjg@FreeBSD.org X-Bugzilla-Status: New X-Bugzilla-Resolution: X-Bugzilla-Priority: --- X-Bugzilla-Assigned-To: freebsd-fs@FreeBSD.org X-Bugzilla-Flags: X-Bugzilla-Changed-Fields: Message-ID: In-Reply-To: References: Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable X-Bugzilla-URL: https://bugs.freebsd.org/bugzilla/ Auto-Submitted: auto-generated MIME-Version: 1.0 X-BeenThere: freebsd-fs@freebsd.org X-Mailman-Version: 2.1.23 Precedence: list List-Id: Filesystems List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 01 Nov 2017 07:59:16 -0000 https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=3D223348 --- Comment #4 from Mateusz Guzik --- wow, that's a fucking retarded bug if i ever seen one. sorry for the inconvenience. So to be clear, does this solve the problem for you? diff --git a/sys/kern/vfs_cache.c b/sys/kern/vfs_cache.c index 3da1c97..23c80a9 100644 --- a/sys/kern/vfs_cache.c +++ b/sys/kern/vfs_cache.c @@ -1126,7 +1126,8 @@ cache_lookup_nomakeentry(struct vnode *dvp, struct vn= ode **vpp, uint32_t hash; int error; - if (cnp->cn_namelen =3D=3D 2 && cnp->cn_nameptr[1] =3D=3D '.') { + if (cnp->cn_namelen =3D=3D 2 && + cnp->cn_nameptr[0] =3D=3D '.' && cnp->cn_nameptr[1] =3D=3D '.')= { counter_u64_add(dotdothits, 1); dvlp =3D VP2VNODELOCK(dvp); dvlp2 =3D NULL; --=20 You are receiving this mail because: You are the assignee for the bug.= From owner-freebsd-fs@freebsd.org Wed Nov 1 12:34:27 2017 Return-Path: Delivered-To: freebsd-fs@mailman.ysv.freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:1900:2254:206a::19:1]) by mailman.ysv.freebsd.org (Postfix) with ESMTP id 72B75E5A983 for ; Wed, 1 Nov 2017 12:34:27 +0000 (UTC) (envelope-from bugzilla-noreply@freebsd.org) Received: from kenobi.freebsd.org (kenobi.freebsd.org [IPv6:2001:1900:2254:206a::16:76]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (Client did not present a certificate) by mx1.freebsd.org (Postfix) with ESMTPS id 5B1926CC20 for ; Wed, 1 Nov 2017 12:34:27 +0000 (UTC) (envelope-from bugzilla-noreply@freebsd.org) Received: from bugs.freebsd.org ([127.0.1.118]) by kenobi.freebsd.org (8.15.2/8.15.2) with ESMTP id vA1CYR2T029560 for ; Wed, 1 Nov 2017 12:34:27 GMT (envelope-from bugzilla-noreply@freebsd.org) From: bugzilla-noreply@freebsd.org To: freebsd-fs@FreeBSD.org Subject: [Bug 223348] Inode number for directories change to inode number of directory two levels up causing hierarchy loop Date: Wed, 01 Nov 2017 12:34:27 +0000 X-Bugzilla-Reason: AssignedTo X-Bugzilla-Type: changed X-Bugzilla-Watch-Reason: None X-Bugzilla-Product: Base System X-Bugzilla-Component: gnu X-Bugzilla-Version: CURRENT X-Bugzilla-Keywords: X-Bugzilla-Severity: Affects Only Me X-Bugzilla-Who: iz-rpi03@hs-karlsruhe.de X-Bugzilla-Status: Closed X-Bugzilla-Resolution: FIXED X-Bugzilla-Priority: --- X-Bugzilla-Assigned-To: freebsd-fs@FreeBSD.org X-Bugzilla-Flags: X-Bugzilla-Changed-Fields: bug_status resolution Message-ID: In-Reply-To: References: Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable X-Bugzilla-URL: https://bugs.freebsd.org/bugzilla/ Auto-Submitted: auto-generated MIME-Version: 1.0 X-BeenThere: freebsd-fs@freebsd.org X-Mailman-Version: 2.1.23 Precedence: list List-Id: Filesystems List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 01 Nov 2017 12:34:27 -0000 https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=3D223348 iz-rpi03@hs-karlsruhe.de changed: What |Removed |Added ---------------------------------------------------------------------------- Status|New |Closed Resolution|--- |FIXED --- Comment #6 from iz-rpi03@hs-karlsruhe.de --- Sorry for the wait. I can confirm, that base r325268 fixes the bug. Tested with the faster restore-umount-mount case. Compiling the arm will take some more hours, but I do not expect a surprise there. So I like to close the bug as fixed.=20 Thank you all. Ralf --=20 You are receiving this mail because: You are the assignee for the bug.= From owner-freebsd-fs@freebsd.org Wed Nov 1 15:46:41 2017 Return-Path: Delivered-To: freebsd-fs@mailman.ysv.freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:1900:2254:206a::19:1]) by mailman.ysv.freebsd.org (Postfix) with ESMTP id B96E5E5EC00 for ; Wed, 1 Nov 2017 15:46:41 +0000 (UTC) (envelope-from bugzilla-noreply@freebsd.org) Received: from kenobi.freebsd.org (kenobi.freebsd.org [IPv6:2001:1900:2254:206a::16:76]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (Client did not present a certificate) by mx1.freebsd.org (Postfix) with ESMTPS id A809573492 for ; Wed, 1 Nov 2017 15:46:41 +0000 (UTC) (envelope-from bugzilla-noreply@freebsd.org) Received: from bugs.freebsd.org ([127.0.1.118]) by kenobi.freebsd.org (8.15.2/8.15.2) with ESMTP id vA1Fkfp2053888 for ; Wed, 1 Nov 2017 15:46:41 GMT (envelope-from bugzilla-noreply@freebsd.org) From: bugzilla-noreply@freebsd.org To: freebsd-fs@FreeBSD.org Subject: [Bug 219972] Unable to zpool export following some zfs recv Date: Wed, 01 Nov 2017 15:46:41 +0000 X-Bugzilla-Reason: AssignedTo X-Bugzilla-Type: changed X-Bugzilla-Watch-Reason: None X-Bugzilla-Product: Base System X-Bugzilla-Component: kern X-Bugzilla-Version: 11.0-RELEASE X-Bugzilla-Keywords: X-Bugzilla-Severity: Affects Some People X-Bugzilla-Who: pfribeiro@gmail.com X-Bugzilla-Status: New X-Bugzilla-Resolution: X-Bugzilla-Priority: --- X-Bugzilla-Assigned-To: freebsd-fs@FreeBSD.org X-Bugzilla-Flags: X-Bugzilla-Changed-Fields: Message-ID: In-Reply-To: References: Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable X-Bugzilla-URL: https://bugs.freebsd.org/bugzilla/ Auto-Submitted: auto-generated MIME-Version: 1.0 X-BeenThere: freebsd-fs@freebsd.org X-Mailman-Version: 2.1.23 Precedence: list List-Id: Filesystems List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 01 Nov 2017 15:46:41 -0000 https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=3D219972 --- Comment #8 from pfribeiro@gmail.com --- (In reply to Matthias Pfaller from comment #7) Hi Matthias, Sorry for having abandoned this thread for a while.. I haven't had much tim= e to come back to examining this problem in more detail. I did not have any zvol= s on the disk, just a regular dataset. Apologies if I'm mistaken with the terminology of ZFS. I also haven't had time to retest on recent releases but I suppose the behaviour is still the same. Unless someone else is willing to dig into thi= s, I don't think I will be able to offer much help, even after examining the sou= rce code and the call trace at the time. Concurrency issues are tricky. --=20 You are receiving this mail because: You are the assignee for the bug.=