I don''t think this is expected, is it? I can no longer move a subvolume into another subvolume. I can move a subvolume into a directory. This happens with 3.11.3, and 3.12 rc4. I''m not sure yet when the regression first appeared. Example move subvol to subvol, where home is a subvolume that contains files, and f20inprogress is a newly created (empty) subvolume. [root@f20s btrfs]# mv home f20inprogress mv: cannot move ‘home’ to ‘f20inprogress/home’: File exists [ 1505.577231] ------------[ cut here ]------------ [ 1505.577270] WARNING: CPU: 1 PID: 2958 at fs/btrfs/inode.c:8049 btrfs_rename+0x157/0x740 [btrfs]() [ 1505.577272] Modules linked in: nls_utf8 isofs fuse ebtable_nat nf_conntrack_netbios_ns nf_conntrack_broadcast ipt_MASQUERADE ip6table_nat nf_nat_ipv6 ip6table_mangle ip6table_security ip6table_raw ip6t_REJECT nf_conntrack_ipv6 nf_defrag_ipv6 iptable_nat nf_nat_ipv4 nf_nat iptable_mangle iptable_security iptable_raw nf_conntrack_ipv4 nf_defrag_ipv4 xt_conntrack nf_conntrack ebtable_filter ebtables ip6table_filter ip6_tables bnep b43 snd_hda_codec_realtek snd_hda_intel coretemp snd_hda_codec kvm_intel bcma snd_hwdep mac80211 kvm uvcvideo videobuf2_vmalloc videobuf2_memops videobuf2_core cfg80211 videodev iTCO_wdt joydev applesmc iTCO_vendor_support i2c_i801 media input_polldev microcode hid_appleir btusb bcm5974 bluetooth snd_seq rfkill snd_seq_device lpc_ich snd_pcm mfd_core snd_page_alloc snd_timer [ 1505.577307] snd apple_bl shpchp soundcore acpi_cpufreq mperf nfsd auth_rpcgss nfs_acl lockd vfat fat squashfs btrfs libcrc32c xor zlib_deflate raid6_pq nouveau firewire_ohci mxm_wmi firewire_core ssb wmi i2c_algo_bit ata_generic pata_acpi drm_kms_helper usb_storage crc_itu_t sky2 ttm mmc_core drm i2c_core video sunrpc loop [ 1505.577333] CPU: 1 PID: 2958 Comm: mv Tainted: G W 3.11.3-301.fc20.x86_64 #1 [ 1505.577334] Hardware name: Apple Inc. MacBookPro4,1/Mac-F42C89C8, BIOS MBP41.88Z.00C1.B03.0802271651 02/27/08 [ 1505.577336] 0000000000000009 ffff8800ba46fc78 ffffffff8164807b 0000000000000000 [ 1505.577339] ffff8800ba46fcb0 ffffffff8106715d ffff8801009b15f8 0000000000000100 [ 1505.577342] 0000000000000000 ffff8801009b09e0 ffff8800a8586a01 ffff8800ba46fcc0 [ 1505.577345] Call Trace: [ 1505.577351] [<ffffffff8164807b>] dump_stack+0x45/0x56 [ 1505.577354] [<ffffffff8106715d>] warn_slowpath_common+0x7d/0xa0 [ 1505.577357] [<ffffffff8106723a>] warn_slowpath_null+0x1a/0x20 [ 1505.577370] [<ffffffffa02d16e7>] btrfs_rename+0x157/0x740 [btrfs] [ 1505.577374] [<ffffffff811b5b4f>] vfs_rename+0x16f/0x480 [ 1505.577377] [<ffffffff811b61e1>] SYSC_renameat+0x381/0x420 [ 1505.577380] [<ffffffff811b7f81>] ? user_path_at+0x11/0x20 [ 1505.577383] [<ffffffff811ad092>] ? SYSC_newlstat+0x22/0x40 [ 1505.577386] [<ffffffff811b87de>] SyS_renameat+0xe/0x10 [ 1505.577388] [<ffffffff811b87fb>] SyS_rename+0x1b/0x20 [ 1505.577392] [<ffffffff81657259>] system_call_fastpath+0x16/0x1b [ 1505.577394] ---[ end trace 84e316366ba447b6 ]--- Chris Murphy-- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Oct 8, 2013, at 3:58 PM, Chris Murphy <lists@colorremedies.com> wrote:> I don''t think this is expected, is it? I can no longer move a subvolume into another subvolume. I can move a subvolume into a directory.So far this isn''t working with 3.10.14 or 3.9.11. The file system and subvolumes were created while btrfs-progs-0.20.rc1.20130917git194aa4a-1.fc20.x86_64 was installed. So I think I need to go back to an older btrfs-progs. Or I need a new brain. Chris Murphy-- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Chris Murphy
2013-Oct-08 23:35 UTC
Re: regression, can''t move subvols into subvols, fs/btrfs/inode.c:8049 btrfs_rename
On Oct 8, 2013, at 4:29 PM, Chris Murphy <lists@colorremedies.com> wrote:> So I think I need to go back to an older btrfs-progs. Or I need a new brain.OK I''m losing it. New file system created with: btrfs-progs-0.20.rc1.20130501git7854c8b-4.fc20.x86_64 3.11.3-301.fc20.x86_64 I still can''t move a subvolume into a subvolume. Using identical syntax with a directory as target works OK. [root@f20s mnt]# mv subvol1 subvol2 mv: cannot move ‘subvol1’ to ‘subvol2/subvol1’: File exists Same kernel message: [ 2108.277456] ------------[ cut here ]------------ [ 2108.277532] WARNING: CPU: 0 PID: 2127 at fs/btrfs/inode.c:8049 btrfs_rename+0x157/0x740 [btrfs]() [ 2108.277537] Modules linked in: nls_utf8 hfsplus fuse ebtable_nat nf_conntrack_netbios_ns nf_conntrack_broadcast ipt_MASQUERADE ip6table_nat nf_nat_ipv6 ip6table_mangle ip6table_security ip6table_raw ip6t_REJECT nf_conntrack_ipv6 nf_defrag_ipv6 iptable_nat nf_nat_ipv4 nf_nat iptable_mangle iptable_security iptable_raw nf_conntrack_ipv4 nf_defrag_ipv4 xt_conntrack nf_conntrack ebtable_filter ebtables ip6table_filter ip6_tables bnep coretemp kvm_intel kvm iTCO_wdt iTCO_vendor_support joydev applesmc input_polldev uvcvideo b43 videobuf2_vmalloc bcma mac80211 videobuf2_memops videobuf2_core microcode videodev snd_hda_codec_realtek media snd_hda_intel snd_hda_codec i2c_i801 snd_hwdep lpc_ich mfd_core snd_seq cfg80211 snd_seq_device bcm5974 snd_pcm btusb hid_appleir snd_page_alloc snd_timer snd soundcore [ 2108.277641] bluetooth rfkill shpchp apple_bl acpi_cpufreq mperf nfsd auth_rpcgss nfs_acl lockd vfat fat squashfs btrfs libcrc32c xor zlib_deflate raid6_pq nouveau mxm_wmi wmi i2c_algo_bit firewire_ohci ata_generic pata_acpi firewire_core ssb crc_itu_t sky2 usb_storage drm_kms_helper mmc_core ttm drm i2c_core video sunrpc loop [ 2108.277712] CPU: 0 PID: 2127 Comm: mv Tainted: G W 3.11.3-301.fc20.x86_64 #1 [ 2108.277717] Hardware name: Apple Inc. MacBookPro4,1/Mac-F42C89C8, BIOS MBP41.88Z.00C1.B03.0802271651 02/27/08 [ 2108.277721] 0000000000000009 ffff88013463bc78 ffffffff8164807b 0000000000000000 [ 2108.277731] ffff88013463bcb0 ffffffff8106715d ffff88004e0c05d8 0000000000000100 [ 2108.277739] 0000000000000000 ffff88004e0c15f8 ffff8800b4ce9001 ffff88013463bcc0 [ 2108.277748] Call Trace: [ 2108.277761] [<ffffffff8164807b>] dump_stack+0x45/0x56 [ 2108.277771] [<ffffffff8106715d>] warn_slowpath_common+0x7d/0xa0 [ 2108.277779] [<ffffffff8106723a>] warn_slowpath_null+0x1a/0x20 [ 2108.277821] [<ffffffffa02c36e7>] btrfs_rename+0x157/0x740 [btrfs] [ 2108.277832] [<ffffffff811b5b4f>] vfs_rename+0x16f/0x480 [ 2108.277840] [<ffffffff811b61e1>] SYSC_renameat+0x381/0x420 [ 2108.277850] [<ffffffff811b7f81>] ? user_path_at+0x11/0x20 [ 2108.277858] [<ffffffff811ad092>] ? SYSC_newlstat+0x22/0x40 [ 2108.277867] [<ffffffff811b87de>] SyS_renameat+0xe/0x10 [ 2108.277874] [<ffffffff811b87fb>] SyS_rename+0x1b/0x20 [ 2108.277884] [<ffffffff81657259>] system_call_fastpath+0x16/0x1b [ 2108.277889] ---[ end trace c377166ca4c37375 ]----- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Tue, Oct 08, 2013 at 03:58:23PM -0600, Chris Murphy wrote:> I don''t think this is expected, is it? I can no longer move a > subvolume into another subvolume. I can move a subvolume into a > directory. This happens with 3.11.3, and 3.12 rc4. I''m not sure yet > when the regression first appeared.If you revert commit 9c52057c698fb96f8f07e7a4bcf4801a092bda89 Author: Chris Mason <chris.mason@fusionio.com> Date: Mon Dec 17 14:26:57 2012 -0500 Btrfs: fix hash overflow handling Does it still fail? I have some guesses, but I really don''t know this code, and I''m needed upstairs to feed a toddler at this *precise* moment :). - z -- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Tue, Oct 08, 2013 at 03:58:23PM -0600, Chris Murphy wrote:> I don''t think this is expected, is it? I can no longer move a subvolume into another subvolume. I can move a subvolume into a directory. This happens with 3.11.3, and 3.12 rc4. I''m not sure yet when the regression first appeared. > > Example move subvol to subvol, where home is a subvolume that contains files, and f20inprogress is a newly created (empty) subvolume. > > [root@f20s btrfs]# mv home f20inprogress > mv: cannot move ‘home’ to ‘f20inprogress/home’: File exists >Fixed it with [PATCH] Btrfs: use right root when checking for hash collision Sorry about that. Josef -- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Oct 9, 2013, at 10:27 AM, Josef Bacik <jbacik@fusionio.com> wrote:> On Tue, Oct 08, 2013 at 03:58:23PM -0600, Chris Murphy wrote: >> I don''t think this is expected, is it? I can no longer move a subvolume into another subvolume. I can move a subvolume into a directory. This happens with 3.11.3, and 3.12 rc4. I''m not sure yet when the regression first appeared. >> >> Example move subvol to subvol, where home is a subvolume that contains files, and f20inprogress is a newly created (empty) subvolume. >> >> [root@f20s btrfs]# mv home f20inprogress >> mv: cannot move ‘home’ to ‘f20inprogress/home’: File exists >> > > Fixed it with > > [PATCH] Btrfs: use right root when checking for hash collision > > Sorry about that.This behavior seems to go back quite far. At least 3.9.5 is affected; 3.9 is EOL but does it make sense to fix this for 3.10 and 3.11? Chris Murphy-- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Wed, Oct 09, 2013 at 11:17:43AM -0600, Chris Murphy wrote:> > On Oct 9, 2013, at 10:27 AM, Josef Bacik <jbacik@fusionio.com> wrote: > > > On Tue, Oct 08, 2013 at 03:58:23PM -0600, Chris Murphy wrote: > >> I don''t think this is expected, is it? I can no longer move a subvolume into another subvolume. I can move a subvolume into a directory. This happens with 3.11.3, and 3.12 rc4. I''m not sure yet when the regression first appeared. > >> > >> Example move subvol to subvol, where home is a subvolume that contains files, and f20inprogress is a newly created (empty) subvolume. > >> > >> [root@f20s btrfs]# mv home f20inprogress > >> mv: cannot move ‘home’ to ‘f20inprogress/home’: File exists > >> > > > > Fixed it with > > > > [PATCH] Btrfs: use right root when checking for hash collision > > > > Sorry about that. > > This behavior seems to go back quite far. At least 3.9.5 is affected; 3.9 is EOL but does it make sense to fix this for 3.10 and 3.11? >Yeah I''ll mark it for stable. Thanks, Josef -- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html