> 在 2018年2月9日，下午10:28，Yan, Zheng <zyan@xxxxxxxxxx> 写道：
>
> I’m curious how did you make ceph_fs_debugfs_init fail
Hi Yan,
Actually I got some error messages during my hacking work
in mount process, I did some diagnostics for the reason and
found this issue. Let ceph_fs_debugfs_init fail in a natural
way may be difficult, so I explicitly let it fail in the code
to see what will happen in this context.
>
> Yan, Zheng
>
>> On 9 Feb 2018, at 20:40, Chengguang Xu <cgxu519@xxxxxxxxxx> wrote:
>>
>> When failing from ceph_fs_debugfs_init() in ceph_real_mount(),
>> there is lack of dput of root_dentry and it causes slab errors,
>> so change the calling order of ceph_fs_debugfs_init() and
>> open_root_dentry() and do some cleanups to avoid this issue.
>>
>> Detail error messages:
>> =============================================================================
>> BUG ceph_inode_info (Not tainted): Objects remaining in ceph_inode_info on __kmem_cache_shutdown()
>> -----------------------------------------------------------------------------
>> Disabling lock debugging due to kernel taint
>> INFO: Slab 0x000000001037cfd2 objects=23 used=1 fp=0x000000006c0dc265 flags=0x2fffff80008100
>> CPU: 0 PID: 27086 Comm: rmmod Tainted: G B 4.15.0-rc8+ #7
>> Hardware name: Fedora Project OpenStack Nova, BIOS seabios-1.7.5-11.el7 04/01/2014
>> Call Trace:
>> dump_stack+0x5c/0x7c
>> slab_err+0xb4/0xe0
>> ? on_each_cpu_mask+0x24/0x60
>> ? on_each_cpu_cond+0xad/0x140
>> ? __kmalloc+0x1be/0x1f0
>> ? __kmem_cache_shutdown+0x175/0x3c0
>> __kmem_cache_shutdown+0x196/0x3c0
>> shutdown_cache+0x13/0x1a0
>> kmem_cache_destroy+0x1e4/0x220
>> destroy_caches+0x16/0x46 [ceph]
>> SyS_delete_module+0x168/0x240
>> ? exit_to_usermode_loop+0x6e/0x93
>> entry_SYSCALL_64_fastpath+0x20/0x83
>> RIP: 0033:0x7f8650f87c27
>> RSP: 002b:00007ffda546ff88 EFLAGS: 00000206
>> INFO: Object 0x00000000568b3053 @offset=0
>> kmem_cache_destroy ceph_inode_info: Slab cache still has objects
>> CPU: 0 PID: 27086 Comm: rmmod Tainted: G B 4.15.0-rc8+ #7
>> Hardware name: Fedora Project OpenStack Nova, BIOS seabios-1.7.5-11.el7 04/01/2014
>> Call Trace:
>> dump_stack+0x5c/0x7c
>> kmem_cache_destroy+0x203/0x220
>> destroy_caches+0x16/0x46 [ceph]
>> SyS_delete_module+0x168/0x240
>> ? exit_to_usermode_loop+0x6e/0x93
>> entry_SYSCALL_64_fastpath+0x20/0x83
>> RIP: 0033:0x7f8650f87c27
>> RSP: 002b:00007ffda546ff88 EFLAGS: 00000206
>> =============================================================================
>> BUG ceph_dentry_info (Tainted: G B ): Objects remaining in ceph_dentry_info on __kmem_cache_shutdown()
>> -----------------------------------------------------------------------------
>> INFO: Slab 0x00000000aabb3f88 objects=51 used=1 fp=0x000000002436421a flags=0x2fffff80000100
>> CPU: 0 PID: 27086 Comm: rmmod Tainted: G B 4.15.0-rc8+ #7
>> Hardware name: Fedora Project OpenStack Nova, BIOS seabios-1.7.5-11.el7 04/01/2014
>> Call Trace:
>> dump_stack+0x5c/0x7c
>> slab_err+0xb4/0xe0
>> ? on_each_cpu_cond+0xad/0x140
>> ? __kmalloc+0x1be/0x1f0
>> ? __kmem_cache_shutdown+0x175/0x3c0
>> __kmem_cache_shutdown+0x196/0x3c0
>> shutdown_cache+0x13/0x1a0
>> kmem_cache_destroy+0x1e4/0x220
>> destroy_caches+0x3a/0x46 [ceph]
>> SyS_delete_module+0x168/0x240
>> ? exit_to_usermode_loop+0x6e/0x93
>> entry_SYSCALL_64_fastpath+0x20/0x83
>> RIP: 0033:0x7f8650f87c27
>> RSP: 002b:00007ffda546ff88 EFLAGS: 00000206
>> INFO: Object 0x0000000034be48a6 @offset=0
>> kmem_cache_destroy ceph_dentry_info: Slab cache still has objects
>> CPU: 0 PID: 27086 Comm: rmmod Tainted: G B 4.15.0-rc8+ #7
>> Hardware name: Fedora Project OpenStack Nova, BIOS seabios-1.7.5-11.el7 04/01/2014
>> Call Trace:
>> dump_stack+0x5c/0x7c
>> kmem_cache_destroy+0x203/0x220
>> destroy_caches+0x3a/0x46 [ceph]
>> SyS_delete_module+0x168/0x240
>> ? exit_to_usermode_loop+0x6e/0x93
>> entry_SYSCALL_64_fastpath+0x20/0x83
>> RIP: 0033:0x7f8650f87c27
>>
>> Signed-off-by: Chengguang Xu <cgxu519@xxxxxxxxxx>
>> ---
>> fs/ceph/super.c | 16 +++++-----------
>> 1 file changed, 5 insertions(+), 11 deletions(-)
>>
>> diff --git a/fs/ceph/super.c b/fs/ceph/super.c
>> index a62d2a9..26526601 100644
>> --- a/fs/ceph/super.c
>> +++ b/fs/ceph/super.c
>> @@ -836,7 +836,6 @@ static struct dentry *ceph_real_mount(struct ceph_fs_client *fsc)
>> int err;
>> unsigned long started = jiffies; /* note the start time */
>> struct dentry *root;
>> - int first = 0; /* first vfsmount for this super_block */
>>
>> dout("mount start %p\n", fsc);
>> mutex_lock(&fsc->client->mount_mutex);
>> @@ -861,17 +860,17 @@ static struct dentry *ceph_real_mount(struct ceph_fs_client *fsc)
>> path = fsc->mount_options->server_path + 1;
>> dout("mount opening path %s\n", path);
>> }
>> +
>> + err = ceph_fs_debugfs_init(fsc);
>> + if (err < 0)
>> + goto out;
>> +
>> root = open_root_dentry(fsc, path, started);
>> if (IS_ERR(root)) {
>> err = PTR_ERR(root);
>> goto out;
>> }
>> fsc->sb->s_root = dget(root);
>> - first = 1;
>> -
>> - err = ceph_fs_debugfs_init(fsc);
>> - if (err < 0)
>> - goto fail;
>> } else {
>> root = dget(fsc->sb->s_root);
>> }
>> @@ -881,11 +880,6 @@ static struct dentry *ceph_real_mount(struct ceph_fs_client *fsc)
>> mutex_unlock(&fsc->client->mount_mutex);
>> return root;
>>
>> -fail:
>> - if (first) {
>> - dput(fsc->sb->s_root);
>> - fsc->sb->s_root = NULL;
>> - }
>> out:
>> mutex_unlock(&fsc->client->mount_mutex);
>> return ERR_PTR(err);
>> --
>> 1.8.3.1
>>
>
> --
> To unsubscribe from this list: send the line "unsubscribe ceph-devel" in
> the body of a message to majordomo@xxxxxxxxxxxxxxx
> More majordomo info at http://vger.kernel.org/majordomo-info.html
--
To unsubscribe from this list: send the line "unsubscribe ceph-devel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
