From: Sasha Levin <sashal@kernel.org>
To: linux-kernel@vger.kernel.org, stable@vger.kernel.org
Cc: Xiubo Li <xiubli@redhat.com>, Jeff Layton <jlayton@kernel.org>,
Ilya Dryomov <idryomov@gmail.com>,
Sasha Levin <sashal@kernel.org>,
ceph-devel@vger.kernel.org
Subject: [PATCH AUTOSEL 5.18 43/68] ceph: fix possible deadlock when holding Fwb to get inline_data
Date: Tue, 7 Jun 2022 13:48:09 -0400 [thread overview]
Message-ID: <20220607174846.477972-43-sashal@kernel.org> (raw)
In-Reply-To: <20220607174846.477972-1-sashal@kernel.org>
From: Xiubo Li <xiubli@redhat.com>
[ Upstream commit 825978fd6a0defc3c29d8a38b6cea76a0938d21e ]
1, mount with wsync.
2, create a file with O_RDWR, and the request was sent to mds.0:
ceph_atomic_open()-->
ceph_mdsc_do_request(openc)
finish_open(file, dentry, ceph_open)-->
ceph_open()-->
ceph_init_file()-->
ceph_init_file_info()-->
ceph_uninline_data()-->
{
...
if (inline_version == 1 || /* initial version, no data */
inline_version == CEPH_INLINE_NONE)
goto out_unlock;
...
}
The inline_version will be 1, which is the initial version for the
new create file. And here the ci->i_inline_version will keep with 1,
it's buggy.
3, buffer write to the file immediately:
ceph_write_iter()-->
ceph_get_caps(file, need=Fw, want=Fb, ...);
generic_perform_write()-->
a_ops->write_begin()-->
ceph_write_begin()-->
netfs_write_begin()-->
netfs_begin_read()-->
netfs_rreq_submit_slice()-->
netfs_read_from_server()-->
rreq->netfs_ops->issue_read()-->
ceph_netfs_issue_read()-->
{
...
if (ci->i_inline_version != CEPH_INLINE_NONE &&
ceph_netfs_issue_op_inline(subreq))
return;
...
}
ceph_put_cap_refs(ci, Fwb);
The ceph_netfs_issue_op_inline() will send a getattr(Fsr) request to
mds.1.
4, then the mds.1 will request the rd lock for CInode::filelock from
the auth mds.0, the mds.0 will do the CInode::filelock state transation
from excl --> sync, but it need to revoke the Fxwb caps back from the
clients.
While the kernel client has aleady held the Fwb caps and waiting for
the getattr(Fsr).
It's deadlock!
URL: https://tracker.ceph.com/issues/55377
Signed-off-by: Xiubo Li <xiubli@redhat.com>
Reviewed-by: Jeff Layton <jlayton@kernel.org>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
fs/ceph/addr.c | 33 +++++++++++++++++++--------------
1 file changed, 19 insertions(+), 14 deletions(-)
diff --git a/fs/ceph/addr.c b/fs/ceph/addr.c
index b6edcf89a429..adef10a6e5c7 100644
--- a/fs/ceph/addr.c
+++ b/fs/ceph/addr.c
@@ -1644,7 +1644,7 @@ int ceph_uninline_data(struct file *file)
struct inode *inode = file_inode(file);
struct ceph_inode_info *ci = ceph_inode(inode);
struct ceph_fs_client *fsc = ceph_inode_to_client(inode);
- struct ceph_osd_request *req;
+ struct ceph_osd_request *req = NULL;
struct ceph_cap_flush *prealloc_cf;
struct folio *folio = NULL;
u64 inline_version = CEPH_INLINE_NONE;
@@ -1652,10 +1652,23 @@ int ceph_uninline_data(struct file *file)
int err = 0;
u64 len;
+ spin_lock(&ci->i_ceph_lock);
+ inline_version = ci->i_inline_version;
+ spin_unlock(&ci->i_ceph_lock);
+
+ dout("uninline_data %p %llx.%llx inline_version %llu\n",
+ inode, ceph_vinop(inode), inline_version);
+
+ if (inline_version == CEPH_INLINE_NONE)
+ return 0;
+
prealloc_cf = ceph_alloc_cap_flush();
if (!prealloc_cf)
return -ENOMEM;
+ if (inline_version == 1) /* initial version, no data */
+ goto out_uninline;
+
folio = read_mapping_folio(inode->i_mapping, 0, file);
if (IS_ERR(folio)) {
err = PTR_ERR(folio);
@@ -1664,17 +1677,6 @@ int ceph_uninline_data(struct file *file)
folio_lock(folio);
- spin_lock(&ci->i_ceph_lock);
- inline_version = ci->i_inline_version;
- spin_unlock(&ci->i_ceph_lock);
-
- dout("uninline_data %p %llx.%llx inline_version %llu\n",
- inode, ceph_vinop(inode), inline_version);
-
- if (inline_version == 1 || /* initial version, no data */
- inline_version == CEPH_INLINE_NONE)
- goto out_unlock;
-
len = i_size_read(inode);
if (len > folio_size(folio))
len = folio_size(folio);
@@ -1739,6 +1741,7 @@ int ceph_uninline_data(struct file *file)
ceph_update_write_metrics(&fsc->mdsc->metric, req->r_start_latency,
req->r_end_latency, len, err);
+out_uninline:
if (!err) {
int dirty;
@@ -1757,8 +1760,10 @@ int ceph_uninline_data(struct file *file)
if (err == -ECANCELED)
err = 0;
out_unlock:
- folio_unlock(folio);
- folio_put(folio);
+ if (folio) {
+ folio_unlock(folio);
+ folio_put(folio);
+ }
out:
ceph_free_cap_flush(prealloc_cf);
dout("uninline_data %p %llx.%llx inline_version %llu = %d\n",
--
2.35.1
next prev parent reply other threads:[~2022-06-07 18:51 UTC|newest]
Thread overview: 73+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-06-07 17:47 [PATCH AUTOSEL 5.18 01/68] iio: dummy: iio_simple_dummy: check the return value of kstrdup() Sasha Levin
2022-06-07 17:47 ` [PATCH AUTOSEL 5.18 02/68] staging: rtl8712: fix a potential memory leak in r871xu_drv_init() Sasha Levin
2022-06-07 17:47 ` [PATCH AUTOSEL 5.18 03/68] iio: st_sensors: Add a local lock for protecting odr Sasha Levin
2022-06-08 9:26 ` Jonathan Cameron
2022-06-09 13:56 ` Sasha Levin
2022-06-07 17:47 ` [PATCH AUTOSEL 5.18 04/68] lkdtm/usercopy: Expand size of "out of frame" object Sasha Levin
2022-06-07 17:47 ` [PATCH AUTOSEL 5.18 05/68] drivers: staging: rtl8723bs: Fix deadlock in rtw_surveydone_event_callback() Sasha Levin
2022-06-07 17:47 ` [PATCH AUTOSEL 5.18 06/68] drivers: staging: rtl8192bs: Fix deadlock in rtw_joinbss_event_prehandle() Sasha Levin
2022-06-07 17:47 ` [PATCH AUTOSEL 5.18 07/68] drivers: staging: rtl8192eu: Fix deadlock in rtw_joinbss_event_prehandle Sasha Levin
2022-06-07 17:47 ` [PATCH AUTOSEL 5.18 08/68] tty: synclink_gt: Fix null-pointer-dereference in slgt_clean() Sasha Levin
2022-06-07 17:47 ` [PATCH AUTOSEL 5.18 09/68] tty: Fix a possible resource leak in icom_probe Sasha Levin
2022-06-07 17:47 ` [PATCH AUTOSEL 5.18 10/68] thunderbolt: Use different lane for second DisplayPort tunnel Sasha Levin
2022-06-07 17:47 ` [PATCH AUTOSEL 5.18 11/68] drivers: staging: rtl8192u: Fix deadlock in ieee80211_beacons_stop() Sasha Levin
2022-06-07 17:47 ` [PATCH AUTOSEL 5.18 12/68] drivers: staging: rtl8192e: Fix deadlock in rtllib_beacons_stop() Sasha Levin
2022-06-07 17:47 ` [PATCH AUTOSEL 5.18 13/68] USB: host: isp116x: check return value after calling platform_get_resource() Sasha Levin
2022-06-07 17:47 ` [PATCH AUTOSEL 5.18 14/68] drivers: tty: serial: Fix deadlock in sa1100_set_termios() Sasha Levin
2022-06-07 17:47 ` [PATCH AUTOSEL 5.18 15/68] drivers: usb: host: Fix deadlock in oxu_bus_suspend() Sasha Levin
2022-06-07 17:47 ` [PATCH AUTOSEL 5.18 16/68] USB: hcd-pci: Fully suspend across freeze/thaw cycle Sasha Levin
2022-06-07 17:47 ` [PATCH AUTOSEL 5.18 17/68] char: xillybus: fix a refcount leak in cleanup_dev() Sasha Levin
2022-06-07 17:47 ` [PATCH AUTOSEL 5.18 18/68] sysrq: do not omit current cpu when showing backtrace of all active CPUs Sasha Levin
2022-06-07 17:47 ` [PATCH AUTOSEL 5.18 19/68] usb: dwc2: gadget: don't reset gadget's driver->bus Sasha Levin
2022-06-07 17:47 ` [PATCH AUTOSEL 5.18 20/68] usb: dwc3: host: Stop setting the ACPI companion Sasha Levin
2022-06-07 17:47 ` [PATCH AUTOSEL 5.18 21/68] usb: dwc3: gadget: Only End Transfer for ep0 data phase Sasha Levin
2022-06-07 17:47 ` [PATCH AUTOSEL 5.18 22/68] soundwire: qcom: adjust autoenumeration timeout Sasha Levin
2022-06-07 17:47 ` [PATCH AUTOSEL 5.18 23/68] misc: rtsx: set NULL intfdata when probe fails Sasha Levin
2022-06-07 17:47 ` [PATCH AUTOSEL 5.18 24/68] extcon: Fix extcon_get_extcon_dev() error handling Sasha Levin
2022-06-07 17:47 ` [PATCH AUTOSEL 5.18 25/68] extcon: Modify extcon device to be created after driver data is set Sasha Levin
2022-06-07 17:47 ` [PATCH AUTOSEL 5.18 26/68] clocksource/drivers/sp804: Avoid error on multiple instances Sasha Levin
2022-06-07 17:47 ` [PATCH AUTOSEL 5.18 27/68] staging: rtl8723bs: Fix alignment to match open parenthesis Sasha Levin
2022-06-07 17:47 ` [PATCH AUTOSEL 5.18 28/68] staging: rtl8712: fix uninit-value in usb_read8() and friends Sasha Levin
2022-06-07 17:47 ` [PATCH AUTOSEL 5.18 29/68] staging: rtl8712: fix uninit-value in r871xu_drv_init() Sasha Levin
2022-06-07 17:47 ` [PATCH AUTOSEL 5.18 30/68] serial: msm_serial: disable interrupts in __msm_console_write() Sasha Levin
2022-06-07 17:47 ` [PATCH AUTOSEL 5.18 31/68] accessiblity: speakup: Add missing misc_deregister in softsynth_probe Sasha Levin
2022-06-09 12:28 ` Samuel Thibault
2022-06-07 17:47 ` [PATCH AUTOSEL 5.18 32/68] kernfs: Separate kernfs_pr_cont_buf and rename_lock Sasha Levin
2022-06-07 17:47 ` [PATCH AUTOSEL 5.18 33/68] watchdog: wdat_wdt: Stop watchdog when rebooting the system Sasha Levin
2022-06-07 17:48 ` [PATCH AUTOSEL 5.18 34/68] ksmbd: smbd: fix connection dropped issue Sasha Levin
2022-06-07 17:48 ` [PATCH AUTOSEL 5.18 35/68] md: don't unregister sync_thread with reconfig_mutex held Sasha Levin
2022-06-08 8:43 ` [dm-devel] " Guoqing Jiang
2022-06-12 17:53 ` Sasha Levin
2022-06-07 17:48 ` [PATCH AUTOSEL 5.18 36/68] md: protect md_unregister_thread from reentrancy Sasha Levin
2022-06-07 17:48 ` [PATCH AUTOSEL 5.18 37/68] ASoC: SOF: amd: Fixed Build error Sasha Levin
2022-06-07 17:48 ` [PATCH AUTOSEL 5.18 38/68] scsi: myrb: Fix up null pointer access on myrb_cleanup() Sasha Levin
2022-06-07 17:48 ` [PATCH AUTOSEL 5.18 39/68] ASoC: rt5640: Do not manipulate pin "Platform Clock" if the "Platform Clock" is not in the DAPM Sasha Levin
2022-06-07 17:48 ` [PATCH AUTOSEL 5.18 40/68] Revert "net: af_key: add check for pfkey_broadcast in function pfkey_process" Sasha Levin
2022-06-07 17:48 ` [PATCH AUTOSEL 5.18 41/68] ceph: allow ceph.dir.rctime xattr to be updatable Sasha Levin
2022-06-07 17:48 ` [PATCH AUTOSEL 5.18 42/68] ceph: flush the mdlog for filesystem sync Sasha Levin
2022-06-07 17:48 ` Sasha Levin [this message]
2022-06-07 17:48 ` [PATCH AUTOSEL 5.18 44/68] net, neigh: Set lower cap for neigh_managed_work rearming Sasha Levin
2022-06-07 17:48 ` [PATCH AUTOSEL 5.18 45/68] drm/amd/display: Check if modulo is 0 before dividing Sasha Levin
2022-06-07 17:48 ` [PATCH AUTOSEL 5.18 46/68] drm/amd/display: Check zero planes for OTG disable W/A on clock change Sasha Levin
2022-06-07 17:48 ` [PATCH AUTOSEL 5.18 47/68] drm/radeon: fix a possible null pointer dereference Sasha Levin
2022-06-07 17:48 ` [PATCH AUTOSEL 5.18 48/68] drm/amd/pm: fix a potential gpu_metrics_table memory leak Sasha Levin
2022-06-07 17:48 ` [PATCH AUTOSEL 5.18 49/68] drm/amd/pm: Fix missing thermal throttler status Sasha Levin
2022-06-07 17:48 ` [PATCH AUTOSEL 5.18 50/68] drm/amd/pm: correct the metrics version for SMU 11.0.11/12/13 Sasha Levin
2022-06-07 17:48 ` [PATCH AUTOSEL 5.18 51/68] um: line: Use separate IRQs per line Sasha Levin
2022-06-07 17:48 ` [PATCH AUTOSEL 5.18 52/68] modpost: fix undefined behavior of is_arm_mapping_symbol() Sasha Levin
2022-06-07 17:48 ` [PATCH AUTOSEL 5.18 53/68] objtool: Mark __ubsan_handle_builtin_unreachable() as noreturn Sasha Levin
2022-06-07 17:48 ` [PATCH AUTOSEL 5.18 54/68] x86/cpu: Elide KCSAN for cpu_has() and friends Sasha Levin
2022-06-07 17:48 ` [PATCH AUTOSEL 5.18 55/68] jump_label,noinstr: Avoid instrumentation for JUMP_LABEL=n builds Sasha Levin
2022-06-07 17:48 ` [PATCH AUTOSEL 5.18 56/68] nbd: call genl_unregister_family() first in nbd_cleanup() Sasha Levin
2022-06-07 17:48 ` [PATCH AUTOSEL 5.18 57/68] nbd: fix race between nbd_alloc_config() and module removal Sasha Levin
2022-06-07 17:48 ` [PATCH AUTOSEL 5.18 58/68] nbd: fix io hung while disconnecting device Sasha Levin
2022-06-07 17:48 ` [PATCH AUTOSEL 5.18 59/68] fs/ntfs3: Fix invalid free in log_replay Sasha Levin
2022-06-07 17:48 ` [PATCH AUTOSEL 5.18 60/68] Revert "PCI: brcmstb: Do not turn off WOL regulators on suspend" Sasha Levin
2022-06-07 17:48 ` [PATCH AUTOSEL 5.18 61/68] Revert "PCI: brcmstb: Add control of subdevice voltage regulators" Sasha Levin
2022-06-07 17:48 ` [PATCH AUTOSEL 5.18 62/68] Revert "PCI: brcmstb: Add mechanism to turn on subdev regulators" Sasha Levin
2022-06-07 17:48 ` [PATCH AUTOSEL 5.18 63/68] Revert "PCI: brcmstb: Split brcm_pcie_setup() into two funcs" Sasha Levin
2022-06-07 17:48 ` [PATCH AUTOSEL 5.18 64/68] cifs: fix potential deadlock in direct reclaim Sasha Levin
2022-06-07 17:48 ` [PATCH AUTOSEL 5.18 65/68] s390/gmap: voluntarily schedule during key setting Sasha Levin
2022-06-07 17:48 ` [PATCH AUTOSEL 5.18 66/68] cifs: version operations for smb20 unneeded when legacy support disabled Sasha Levin
2022-06-07 17:48 ` [PATCH AUTOSEL 5.18 67/68] drm/amd/pm: use bitmap_{from,to}_arr32 where appropriate Sasha Levin
2022-06-07 17:48 ` [PATCH AUTOSEL 5.18 68/68] nodemask: Fix return values to be unsigned Sasha Levin
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20220607174846.477972-43-sashal@kernel.org \
--to=sashal@kernel.org \
--cc=ceph-devel@vger.kernel.org \
--cc=idryomov@gmail.com \
--cc=jlayton@kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=stable@vger.kernel.org \
--cc=xiubli@redhat.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox