From 2e29be2e491595407087ab36a5e5a159be693f7b Mon Sep 17 00:00:00 2001 From: Ira Weiny Date: Tue, 25 May 2021 10:24:26 -0700 Subject: fs/fuse: Remove unneeded kaddr parameter fuse_dax_mem_range_init() does not need the address or the pfn of the memory requested in dax_direct_access(). It is only calling direct access to get the number of pages. Remove the unused variables and stop requesting the kaddr and pfn from dax_direct_access(). Reviewed-by: Dan Williams Signed-off-by: Ira Weiny Reviewed-by: Vivek Goyal Link: https://lore.kernel.org/r/20210525172428.3634316-2-ira.weiny@intel.com Signed-off-by: Dan Williams --- fs/fuse/dax.c | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/fuse/dax.c b/fs/fuse/dax.c index ff99ab2a3c43..34f8a5635c7f 100644 --- a/fs/fuse/dax.c +++ b/fs/fuse/dax.c @@ -1234,8 +1234,6 @@ void fuse_dax_conn_free(struct fuse_conn *fc) static int fuse_dax_mem_range_init(struct fuse_conn_dax *fcd) { long nr_pages, nr_ranges; - void *kaddr; - pfn_t pfn; struct fuse_dax_mapping *range; int ret, id; size_t dax_size = -1; @@ -1247,8 +1245,8 @@ static int fuse_dax_mem_range_init(struct fuse_conn_dax *fcd) INIT_DELAYED_WORK(&fcd->free_work, fuse_dax_free_mem_worker); id = dax_read_lock(); - nr_pages = dax_direct_access(fcd->dev, 0, PHYS_PFN(dax_size), &kaddr, - &pfn); + nr_pages = dax_direct_access(fcd->dev, 0, PHYS_PFN(dax_size), NULL, + NULL); dax_read_unlock(id); if (nr_pages < 0) { pr_debug("dax_direct_access() returned %ld\n", nr_pages); -- cgit v1.3-18-gd494 From 44788591c3cfb81d9315b8ee5c2076e51bfe8a39 Mon Sep 17 00:00:00 2001 From: Ira Weiny Date: Tue, 25 May 2021 10:24:27 -0700 Subject: fs/dax: Clarify nr_pages to dax_direct_access() dax_direct_access() takes a number of pages. PHYS_PFN(PAGE_SIZE) is a very round about way to specify '1'. Change the nr_pages parameter to the explicit value of '1'. Reviewed-by: Dan Williams Signed-off-by: Ira Weiny Link: https://lore.kernel.org/r/20210525172428.3634316-3-ira.weiny@intel.com Signed-off-by: Dan Williams --- fs/dax.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/dax.c b/fs/dax.c index 62352cbcf0f4..8ce3a7e147a8 100644 --- a/fs/dax.c +++ b/fs/dax.c @@ -721,7 +721,7 @@ static int copy_cow_page_dax(struct block_device *bdev, struct dax_device *dax_d return rc; id = dax_read_lock(); - rc = dax_direct_access(dax_dev, pgoff, PHYS_PFN(PAGE_SIZE), &kaddr, NULL); + rc = dax_direct_access(dax_dev, pgoff, 1, &kaddr, NULL); if (rc < 0) { dax_read_unlock(id); return rc; -- cgit v1.3-18-gd494 From bf2ba432213fade50dd39f2e348085b758c0726e Mon Sep 17 00:00:00 2001 From: Luis Henriques Date: Tue, 6 Jul 2021 14:52:41 +0100 Subject: ceph: reduce contention in ceph_check_delayed_caps() Function ceph_check_delayed_caps() is called from the mdsc->delayed_work workqueue and it can be kept looping for quite some time if caps keep being added back to the mdsc->cap_delay_list. This may result in the watchdog tainting the kernel with the softlockup flag. This patch breaks this loop if the caps have been recently (i.e. during the loop execution). Any new caps added to the list will be handled in the next run. Also, allow schedule_delayed() callers to explicitly set the delay value instead of defaulting to 5s, so we can ensure that it runs soon afterward if it looks like there is more work. Cc: stable@vger.kernel.org URL: https://tracker.ceph.com/issues/46284 Signed-off-by: Luis Henriques Reviewed-by: Jeff Layton Signed-off-by: Ilya Dryomov --- fs/ceph/caps.c | 17 ++++++++++++++++- fs/ceph/mds_client.c | 25 ++++++++++++++++--------- fs/ceph/super.h | 2 +- 3 files changed, 33 insertions(+), 11 deletions(-) (limited to 'fs') diff --git a/fs/ceph/caps.c b/fs/ceph/caps.c index 7bdefd0c789a..2a2900903f8c 100644 --- a/fs/ceph/caps.c +++ b/fs/ceph/caps.c @@ -4150,11 +4150,19 @@ bad: /* * Delayed work handler to process end of delayed cap release LRU list. + * + * If new caps are added to the list while processing it, these won't get + * processed in this run. In this case, the ci->i_hold_caps_max will be + * returned so that the work can be scheduled accordingly. */ -void ceph_check_delayed_caps(struct ceph_mds_client *mdsc) +unsigned long ceph_check_delayed_caps(struct ceph_mds_client *mdsc) { struct inode *inode; struct ceph_inode_info *ci; + struct ceph_mount_options *opt = mdsc->fsc->mount_options; + unsigned long delay_max = opt->caps_wanted_delay_max * HZ; + unsigned long loop_start = jiffies; + unsigned long delay = 0; dout("check_delayed_caps\n"); spin_lock(&mdsc->cap_delay_lock); @@ -4162,6 +4170,11 @@ void ceph_check_delayed_caps(struct ceph_mds_client *mdsc) ci = list_first_entry(&mdsc->cap_delay_list, struct ceph_inode_info, i_cap_delay_list); + if (time_before(loop_start, ci->i_hold_caps_max - delay_max)) { + dout("%s caps added recently. Exiting loop", __func__); + delay = ci->i_hold_caps_max; + break; + } if ((ci->i_ceph_flags & CEPH_I_FLUSH) == 0 && time_before(jiffies, ci->i_hold_caps_max)) break; @@ -4177,6 +4190,8 @@ void ceph_check_delayed_caps(struct ceph_mds_client *mdsc) } } spin_unlock(&mdsc->cap_delay_lock); + + return delay; } /* diff --git a/fs/ceph/mds_client.c b/fs/ceph/mds_client.c index 9db1b39df773..afdc20213876 100644 --- a/fs/ceph/mds_client.c +++ b/fs/ceph/mds_client.c @@ -4490,22 +4490,29 @@ void inc_session_sequence(struct ceph_mds_session *s) } /* - * delayed work -- periodically trim expired leases, renew caps with mds + * delayed work -- periodically trim expired leases, renew caps with mds. If + * the @delay parameter is set to 0 or if it's more than 5 secs, the default + * workqueue delay value of 5 secs will be used. */ -static void schedule_delayed(struct ceph_mds_client *mdsc) +static void schedule_delayed(struct ceph_mds_client *mdsc, unsigned long delay) { - int delay = 5; - unsigned hz = round_jiffies_relative(HZ * delay); - schedule_delayed_work(&mdsc->delayed_work, hz); + unsigned long max_delay = HZ * 5; + + /* 5 secs default delay */ + if (!delay || (delay > max_delay)) + delay = max_delay; + schedule_delayed_work(&mdsc->delayed_work, + round_jiffies_relative(delay)); } static void delayed_work(struct work_struct *work) { - int i; struct ceph_mds_client *mdsc = container_of(work, struct ceph_mds_client, delayed_work.work); + unsigned long delay; int renew_interval; int renew_caps; + int i; dout("mdsc delayed_work\n"); @@ -4545,7 +4552,7 @@ static void delayed_work(struct work_struct *work) } mutex_unlock(&mdsc->mutex); - ceph_check_delayed_caps(mdsc); + delay = ceph_check_delayed_caps(mdsc); ceph_queue_cap_reclaim_work(mdsc); @@ -4553,7 +4560,7 @@ static void delayed_work(struct work_struct *work) maybe_recover_session(mdsc); - schedule_delayed(mdsc); + schedule_delayed(mdsc, delay); } int ceph_mdsc_init(struct ceph_fs_client *fsc) @@ -5030,7 +5037,7 @@ void ceph_mdsc_handle_mdsmap(struct ceph_mds_client *mdsc, struct ceph_msg *msg) mdsc->mdsmap->m_epoch); mutex_unlock(&mdsc->mutex); - schedule_delayed(mdsc); + schedule_delayed(mdsc, 0); return; bad_unlock: diff --git a/fs/ceph/super.h b/fs/ceph/super.h index 6b6332a5c113..9215a2f4535c 100644 --- a/fs/ceph/super.h +++ b/fs/ceph/super.h @@ -1167,7 +1167,7 @@ extern void ceph_flush_snaps(struct ceph_inode_info *ci, extern bool __ceph_should_report_size(struct ceph_inode_info *ci); extern void ceph_check_caps(struct ceph_inode_info *ci, int flags, struct ceph_mds_session *session); -extern void ceph_check_delayed_caps(struct ceph_mds_client *mdsc); +extern unsigned long ceph_check_delayed_caps(struct ceph_mds_client *mdsc); extern void ceph_flush_dirty_caps(struct ceph_mds_client *mdsc); extern int ceph_drop_caps_for_unlink(struct inode *inode); extern int ceph_encode_inode_release(void **p, struct inode *inode, -- cgit v1.3-18-gd494 From 8434ffe71c874b9c4e184b88d25de98c2bf5fe3f Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Tue, 3 Aug 2021 12:47:34 -0400 Subject: ceph: take snap_empty_lock atomically with snaprealm refcount change There is a race in ceph_put_snap_realm. The change to the nref and the spinlock acquisition are not done atomically, so you could decrement nref, and before you take the spinlock, the nref is incremented again. At that point, you end up putting it on the empty list when it shouldn't be there. Eventually __cleanup_empty_realms runs and frees it when it's still in-use. Fix this by protecting the 1->0 transition with atomic_dec_and_lock, and just drop the spinlock if we can get the rwsem. Because these objects can also undergo a 0->1 refcount transition, we must protect that change as well with the spinlock. Increment locklessly unless the value is at 0, in which case we take the spinlock, increment and then take it off the empty list if it did the 0->1 transition. With these changes, I'm removing the dout() messages from these functions, as well as in __put_snap_realm. They've always been racy, and it's better to not print values that may be misleading. Cc: stable@vger.kernel.org URL: https://tracker.ceph.com/issues/46419 Reported-by: Mark Nelson Signed-off-by: Jeff Layton Reviewed-by: Luis Henriques Signed-off-by: Ilya Dryomov --- fs/ceph/snap.c | 34 +++++++++++++++++----------------- 1 file changed, 17 insertions(+), 17 deletions(-) (limited to 'fs') diff --git a/fs/ceph/snap.c b/fs/ceph/snap.c index 4ac0606dcbd4..4c6bd1042c94 100644 --- a/fs/ceph/snap.c +++ b/fs/ceph/snap.c @@ -67,19 +67,19 @@ void ceph_get_snap_realm(struct ceph_mds_client *mdsc, { lockdep_assert_held(&mdsc->snap_rwsem); - dout("get_realm %p %d -> %d\n", realm, - atomic_read(&realm->nref), atomic_read(&realm->nref)+1); /* - * since we _only_ increment realm refs or empty the empty - * list with snap_rwsem held, adjusting the empty list here is - * safe. we do need to protect against concurrent empty list - * additions, however. + * The 0->1 and 1->0 transitions must take the snap_empty_lock + * atomically with the refcount change. Go ahead and bump the + * nref here, unless it's 0, in which case we take the spinlock + * and then do the increment and remove it from the list. */ - if (atomic_inc_return(&realm->nref) == 1) { - spin_lock(&mdsc->snap_empty_lock); + if (atomic_inc_not_zero(&realm->nref)) + return; + + spin_lock(&mdsc->snap_empty_lock); + if (atomic_inc_return(&realm->nref) == 1) list_del_init(&realm->empty_item); - spin_unlock(&mdsc->snap_empty_lock); - } + spin_unlock(&mdsc->snap_empty_lock); } static void __insert_snap_realm(struct rb_root *root, @@ -208,28 +208,28 @@ static void __put_snap_realm(struct ceph_mds_client *mdsc, { lockdep_assert_held_write(&mdsc->snap_rwsem); - dout("__put_snap_realm %llx %p %d -> %d\n", realm->ino, realm, - atomic_read(&realm->nref), atomic_read(&realm->nref)-1); + /* + * We do not require the snap_empty_lock here, as any caller that + * increments the value must hold the snap_rwsem. + */ if (atomic_dec_and_test(&realm->nref)) __destroy_snap_realm(mdsc, realm); } /* - * caller needn't hold any locks + * See comments in ceph_get_snap_realm. Caller needn't hold any locks. */ void ceph_put_snap_realm(struct ceph_mds_client *mdsc, struct ceph_snap_realm *realm) { - dout("put_snap_realm %llx %p %d -> %d\n", realm->ino, realm, - atomic_read(&realm->nref), atomic_read(&realm->nref)-1); - if (!atomic_dec_and_test(&realm->nref)) + if (!atomic_dec_and_lock(&realm->nref, &mdsc->snap_empty_lock)) return; if (down_write_trylock(&mdsc->snap_rwsem)) { + spin_unlock(&mdsc->snap_empty_lock); __destroy_snap_realm(mdsc, realm); up_write(&mdsc->snap_rwsem); } else { - spin_lock(&mdsc->snap_empty_lock); list_add(&realm->empty_item, &mdsc->snap_empty); spin_unlock(&mdsc->snap_empty_lock); } -- cgit v1.3-18-gd494 From 7d3fc01796fc895e5fcce45c994c5a8db8120a8d Mon Sep 17 00:00:00 2001 From: Shyam Prasad N Date: Wed, 4 Aug 2021 18:37:22 +0000 Subject: cifs: create sd context must be a multiple of 8 We used to follow the rule earlier that the create SD context always be a multiple of 8. However, with the change: cifs: refactor create_sd_buf() and and avoid corrupting the buffer ...we recompute the length, and we failed that rule. Fixing that with this change. Cc: # v5.10+ Signed-off-by: Shyam Prasad N Signed-off-by: Steve French --- fs/cifs/smb2pdu.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/cifs/smb2pdu.c b/fs/cifs/smb2pdu.c index 781d14e5f2af..b6d2e3591927 100644 --- a/fs/cifs/smb2pdu.c +++ b/fs/cifs/smb2pdu.c @@ -2426,7 +2426,7 @@ create_sd_buf(umode_t mode, bool set_owner, unsigned int *len) memcpy(aclptr, &acl, sizeof(struct cifs_acl)); buf->ccontext.DataLength = cpu_to_le32(ptr - (__u8 *)&buf->sd); - *len = ptr - (__u8 *)buf; + *len = roundup(ptr - (__u8 *)buf, 8); return buf; } -- cgit v1.3-18-gd494 From ef98eb0409c31c39ab55ff46b2721c3b4f84c122 Mon Sep 17 00:00:00 2001 From: Nadav Amit Date: Sat, 7 Aug 2021 17:13:41 -0700 Subject: io_uring: clear TIF_NOTIFY_SIGNAL when running task work When using SQPOLL, the submission queue polling thread calls task_work_run() to run queued work. However, when work is added with TWA_SIGNAL - as done by io_uring itself - the TIF_NOTIFY_SIGNAL remains set afterwards and is never cleared. Consequently, when the submission queue polling thread checks whether signal_pending(), it may always find a pending signal, if task_work_add() was ever called before. The impact of this bug might be different on different kernel versions. It appears that on 5.14 it would only cause unnecessary calculation and prevent the polling thread from sleeping. On 5.13, where the bug was found, it stops the polling thread from finding newly submitted work. Instead of task_work_run(), use tracehook_notify_signal() that clears TIF_NOTIFY_SIGNAL. Test for TIF_NOTIFY_SIGNAL in addition to current->task_works to avoid a race in which task_works is cleared but the TIF_NOTIFY_SIGNAL is set. Fixes: 685fe7feedb96 ("io-wq: eliminate the need for a manager thread") Cc: Jens Axboe Cc: Pavel Begunkov Signed-off-by: Nadav Amit Link: https://lore.kernel.org/r/20210808001342.964634-2-namit@vmware.com Signed-off-by: Jens Axboe --- fs/io_uring.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/io_uring.c b/fs/io_uring.c index bf548af0426c..1093df3977b8 100644 --- a/fs/io_uring.c +++ b/fs/io_uring.c @@ -78,6 +78,7 @@ #include #include #include +#include #define CREATE_TRACE_POINTS #include @@ -2222,9 +2223,9 @@ static inline unsigned int io_put_rw_kbuf(struct io_kiocb *req) static inline bool io_run_task_work(void) { - if (current->task_works) { + if (test_thread_flag(TIF_NOTIFY_SIGNAL) || current->task_works) { __set_current_state(TASK_RUNNING); - task_work_run(); + tracehook_notify_signal(); return true; } -- cgit v1.3-18-gd494 From 20c0b380f971e7d48f5d978bc27d827f7eabb21a Mon Sep 17 00:00:00 2001 From: Nadav Amit Date: Sat, 7 Aug 2021 17:13:42 -0700 Subject: io_uring: Use WRITE_ONCE() when writing to sq_flags The compiler should be forbidden from any strange optimization for async writes to user visible data-structures. Without proper protection, the compiler can cause write-tearing or invent writes that would confuse the userspace. However, there are writes to sq_flags which are not protected by WRITE_ONCE(). Use WRITE_ONCE() for these writes. This is purely a theoretical issue. Presumably, any compiler is very unlikely to do such optimizations. Fixes: 75b28affdd6a ("io_uring: allocate the two rings together") Cc: Jens Axboe Cc: Pavel Begunkov Signed-off-by: Nadav Amit Link: https://lore.kernel.org/r/20210808001342.964634-3-namit@vmware.com Signed-off-by: Jens Axboe --- fs/io_uring.c | 13 +++++++++---- 1 file changed, 9 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/io_uring.c b/fs/io_uring.c index 1093df3977b8..ca064486cb41 100644 --- a/fs/io_uring.c +++ b/fs/io_uring.c @@ -1500,7 +1500,8 @@ static bool __io_cqring_overflow_flush(struct io_ring_ctx *ctx, bool force) all_flushed = list_empty(&ctx->cq_overflow_list); if (all_flushed) { clear_bit(0, &ctx->check_cq_overflow); - ctx->rings->sq_flags &= ~IORING_SQ_CQ_OVERFLOW; + WRITE_ONCE(ctx->rings->sq_flags, + ctx->rings->sq_flags & ~IORING_SQ_CQ_OVERFLOW); } if (posted) @@ -1579,7 +1580,9 @@ static bool io_cqring_event_overflow(struct io_ring_ctx *ctx, u64 user_data, } if (list_empty(&ctx->cq_overflow_list)) { set_bit(0, &ctx->check_cq_overflow); - ctx->rings->sq_flags |= IORING_SQ_CQ_OVERFLOW; + WRITE_ONCE(ctx->rings->sq_flags, + ctx->rings->sq_flags | IORING_SQ_CQ_OVERFLOW); + } ocqe->cqe.user_data = user_data; ocqe->cqe.res = res; @@ -6804,14 +6807,16 @@ static inline void io_ring_set_wakeup_flag(struct io_ring_ctx *ctx) { /* Tell userspace we may need a wakeup call */ spin_lock_irq(&ctx->completion_lock); - ctx->rings->sq_flags |= IORING_SQ_NEED_WAKEUP; + WRITE_ONCE(ctx->rings->sq_flags, + ctx->rings->sq_flags | IORING_SQ_NEED_WAKEUP); spin_unlock_irq(&ctx->completion_lock); } static inline void io_ring_clear_wakeup_flag(struct io_ring_ctx *ctx) { spin_lock_irq(&ctx->completion_lock); - ctx->rings->sq_flags &= ~IORING_SQ_NEED_WAKEUP; + WRITE_ONCE(ctx->rings->sq_flags, + ctx->rings->sq_flags & ~IORING_SQ_NEED_WAKEUP); spin_unlock_irq(&ctx->completion_lock); } -- cgit v1.3-18-gd494 From 769f52676756b8c5feb302d2d95af59577fc69ec Mon Sep 17 00:00:00 2001 From: Bart Van Assche Date: Wed, 4 Aug 2021 21:35:01 -0700 Subject: configfs: restore the kernel v5.13 text attribute write behavior Instead of appending new text attribute data at the offset specified by the write() system call, only pass the newly written data to the .store() callback. Reported-by: Bodo Stroesser Tested-by: Bodo Stroesser Signed-off-by: Bart Van Assche Signed-off-by: Christoph Hellwig --- fs/configfs/file.c | 18 ++++++------------ 1 file changed, 6 insertions(+), 12 deletions(-) (limited to 'fs') diff --git a/fs/configfs/file.c b/fs/configfs/file.c index 5a0be9985bae..0ad32150611e 100644 --- a/fs/configfs/file.c +++ b/fs/configfs/file.c @@ -177,28 +177,22 @@ out: return retval; } -/* Fill [buffer, buffer + pos) with data coming from @from. */ -static int fill_write_buffer(struct configfs_buffer *buffer, loff_t pos, +/* Fill @buffer with data coming from @from. */ +static int fill_write_buffer(struct configfs_buffer *buffer, struct iov_iter *from) { - loff_t to_copy; int copied; - u8 *to; if (!buffer->page) buffer->page = (char *)__get_free_pages(GFP_KERNEL, 0); if (!buffer->page) return -ENOMEM; - to_copy = SIMPLE_ATTR_SIZE - 1 - pos; - if (to_copy <= 0) - return 0; - to = buffer->page + pos; - copied = copy_from_iter(to, to_copy, from); + copied = copy_from_iter(buffer->page, SIMPLE_ATTR_SIZE - 1, from); buffer->needs_read_fill = 1; /* if buf is assumed to contain a string, terminate it by \0, * so e.g. sscanf() can scan the string easily */ - to[copied] = 0; + buffer->page[copied] = 0; return copied ? : -EFAULT; } @@ -227,10 +221,10 @@ static ssize_t configfs_write_iter(struct kiocb *iocb, struct iov_iter *from) { struct file *file = iocb->ki_filp; struct configfs_buffer *buffer = file->private_data; - ssize_t len; + int len; mutex_lock(&buffer->mutex); - len = fill_write_buffer(buffer, iocb->ki_pos, from); + len = fill_write_buffer(buffer, from); if (len > 0) len = flush_write_buffer(file, buffer, len); if (len > 0) -- cgit v1.3-18-gd494 From f153c2246783ba210493054d99c66353f56423c9 Mon Sep 17 00:00:00 2001 From: Sven Schnelle Date: Fri, 30 Jul 2021 08:28:54 +0200 Subject: ucounts: add missing data type changes commit f9c82a4ea89c3 ("Increase size of ucounts to atomic_long_t") changed the data type of ucounts/ucounts_max to long, but missed to adjust a few other places. This is noticeable on big endian platforms from user space because the /proc/sys/user/max_*_names files all contain 0. v4 - Made the min and max constants long so the sysctl values are actually settable on little endian machines. -- EWB Fixes: f9c82a4ea89c ("Increase size of ucounts to atomic_long_t") Signed-off-by: Sven Schnelle Tested-by: Nathan Chancellor Tested-by: Linux Kernel Functional Testing Acked-by: Alexey Gladkov v1: https://lkml.kernel.org/r/20210721115800.910778-1-svens@linux.ibm.com v2: https://lkml.kernel.org/r/20210721125233.1041429-1-svens@linux.ibm.com v3: https://lkml.kernel.org/r/20210730062854.3601635-1-svens@linux.ibm.com Link: https://lkml.kernel.org/r/8735rijqlv.fsf_-_@disp2133 Signed-off-by: Eric W. Biederman --- fs/notify/fanotify/fanotify_user.c | 17 +++++++++++------ fs/notify/inotify/inotify_user.c | 17 +++++++++++------ kernel/ucount.c | 19 +++++++++++-------- 3 files changed, 33 insertions(+), 20 deletions(-) (limited to 'fs') diff --git a/fs/notify/fanotify/fanotify_user.c b/fs/notify/fanotify/fanotify_user.c index 64864fb40b40..28b67cb9458d 100644 --- a/fs/notify/fanotify/fanotify_user.c +++ b/fs/notify/fanotify/fanotify_user.c @@ -54,22 +54,27 @@ static int fanotify_max_queued_events __read_mostly; #include +static long ft_zero = 0; +static long ft_int_max = INT_MAX; + struct ctl_table fanotify_table[] = { { .procname = "max_user_groups", .data = &init_user_ns.ucount_max[UCOUNT_FANOTIFY_GROUPS], - .maxlen = sizeof(int), + .maxlen = sizeof(long), .mode = 0644, - .proc_handler = proc_dointvec_minmax, - .extra1 = SYSCTL_ZERO, + .proc_handler = proc_doulongvec_minmax, + .extra1 = &ft_zero, + .extra2 = &ft_int_max, }, { .procname = "max_user_marks", .data = &init_user_ns.ucount_max[UCOUNT_FANOTIFY_MARKS], - .maxlen = sizeof(int), + .maxlen = sizeof(long), .mode = 0644, - .proc_handler = proc_dointvec_minmax, - .extra1 = SYSCTL_ZERO, + .proc_handler = proc_doulongvec_minmax, + .extra1 = &ft_zero, + .extra2 = &ft_int_max, }, { .procname = "max_queued_events", diff --git a/fs/notify/inotify/inotify_user.c b/fs/notify/inotify/inotify_user.c index 98f61b31745a..62051247f6d2 100644 --- a/fs/notify/inotify/inotify_user.c +++ b/fs/notify/inotify/inotify_user.c @@ -55,22 +55,27 @@ struct kmem_cache *inotify_inode_mark_cachep __read_mostly; #include +static long it_zero = 0; +static long it_int_max = INT_MAX; + struct ctl_table inotify_table[] = { { .procname = "max_user_instances", .data = &init_user_ns.ucount_max[UCOUNT_INOTIFY_INSTANCES], - .maxlen = sizeof(int), + .maxlen = sizeof(long), .mode = 0644, - .proc_handler = proc_dointvec_minmax, - .extra1 = SYSCTL_ZERO, + .proc_handler = proc_doulongvec_minmax, + .extra1 = &it_zero, + .extra2 = &it_int_max, }, { .procname = "max_user_watches", .data = &init_user_ns.ucount_max[UCOUNT_INOTIFY_WATCHES], - .maxlen = sizeof(int), + .maxlen = sizeof(long), .mode = 0644, - .proc_handler = proc_dointvec_minmax, - .extra1 = SYSCTL_ZERO, + .proc_handler = proc_doulongvec_minmax, + .extra1 = &it_zero, + .extra2 = &it_int_max, }, { .procname = "max_queued_events", diff --git a/kernel/ucount.c b/kernel/ucount.c index 77be3bbe3cc4..bb51849e6375 100644 --- a/kernel/ucount.c +++ b/kernel/ucount.c @@ -58,14 +58,17 @@ static struct ctl_table_root set_root = { .permissions = set_permissions, }; -#define UCOUNT_ENTRY(name) \ - { \ - .procname = name, \ - .maxlen = sizeof(int), \ - .mode = 0644, \ - .proc_handler = proc_dointvec_minmax, \ - .extra1 = SYSCTL_ZERO, \ - .extra2 = SYSCTL_INT_MAX, \ +static long ue_zero = 0; +static long ue_int_max = INT_MAX; + +#define UCOUNT_ENTRY(name) \ + { \ + .procname = name, \ + .maxlen = sizeof(long), \ + .mode = 0644, \ + .proc_handler = proc_doulongvec_minmax, \ + .extra1 = &ue_zero, \ + .extra2 = &ue_int_max, \ } static struct ctl_table user_table[] = { UCOUNT_ENTRY("max_user_namespaces"), -- cgit v1.3-18-gd494 From 4956b9eaad456a88b0d56947bef036e086250beb Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Mon, 9 Aug 2021 07:49:41 -0600 Subject: io_uring: rsrc ref lock needs to be IRQ safe Nadav reports running into the below splat on re-enabling softirqs: WARNING: CPU: 2 PID: 1777 at kernel/softirq.c:364 __local_bh_enable_ip+0xaa/0xe0 Modules linked in: CPU: 2 PID: 1777 Comm: umem Not tainted 5.13.1+ #161 Hardware name: VMware, Inc. VMware Virtual Platform/440BX Desktop Reference Platform, BIOS 6.00 07/22/2020 RIP: 0010:__local_bh_enable_ip+0xaa/0xe0 Code: a9 00 ff ff 00 74 38 65 ff 0d a2 21 8c 7a e8 ed 1a 20 00 fb 66 0f 1f 44 00 00 5b 41 5c 5d c3 65 8b 05 e6 2d 8c 7a 85 c0 75 9a <0f> 0b eb 96 e8 2d 1f 20 00 eb a5 4c 89 e7 e8 73 4f 0c 00 eb ae 65 RSP: 0018:ffff88812e58fcc8 EFLAGS: 00010046 RAX: 0000000000000000 RBX: 0000000000000201 RCX: dffffc0000000000 RDX: 0000000000000007 RSI: 0000000000000201 RDI: ffffffff8898c5ac RBP: ffff88812e58fcd8 R08: ffffffff8575dbbf R09: ffffed1028ef14f9 R10: ffff88814778a7c3 R11: ffffed1028ef14f8 R12: ffffffff85c9e9ae R13: ffff88814778a000 R14: ffff88814778a7b0 R15: ffff8881086db890 FS: 00007fbcfee17700(0000) GS:ffff8881e0300000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 000000c0402a5008 CR3: 000000011c1ac003 CR4: 00000000003706e0 Call Trace: _raw_spin_unlock_bh+0x31/0x40 io_rsrc_node_ref_zero+0x13e/0x190 io_dismantle_req+0x215/0x220 io_req_complete_post+0x1b8/0x720 __io_complete_rw.isra.0+0x16b/0x1f0 io_complete_rw+0x10/0x20 where it's clear we end up calling the percpu count release directly from the completion path, as it's in atomic mode and we drop the last ref. For file/block IO, this can be from IRQ context already, and the softirq locking for rsrc isn't enough. Just make the lock fully IRQ safe, and ensure we correctly safe state from the release path as we don't know the full context there. Reported-by: Nadav Amit Tested-by: Nadav Amit Link: https://lore.kernel.org/io-uring/C187C836-E78B-4A31-B24C-D16919ACA093@gmail.com/ Signed-off-by: Jens Axboe --- fs/io_uring.c | 19 +++++-------------- 1 file changed, 5 insertions(+), 14 deletions(-) (limited to 'fs') diff --git a/fs/io_uring.c b/fs/io_uring.c index ca064486cb41..6a8257233061 100644 --- a/fs/io_uring.c +++ b/fs/io_uring.c @@ -7138,16 +7138,6 @@ static void **io_alloc_page_table(size_t size) return table; } -static inline void io_rsrc_ref_lock(struct io_ring_ctx *ctx) -{ - spin_lock_bh(&ctx->rsrc_ref_lock); -} - -static inline void io_rsrc_ref_unlock(struct io_ring_ctx *ctx) -{ - spin_unlock_bh(&ctx->rsrc_ref_lock); -} - static void io_rsrc_node_destroy(struct io_rsrc_node *ref_node) { percpu_ref_exit(&ref_node->refs); @@ -7164,9 +7154,9 @@ static void io_rsrc_node_switch(struct io_ring_ctx *ctx, struct io_rsrc_node *rsrc_node = ctx->rsrc_node; rsrc_node->rsrc_data = data_to_kill; - io_rsrc_ref_lock(ctx); + spin_lock_irq(&ctx->rsrc_ref_lock); list_add_tail(&rsrc_node->node, &ctx->rsrc_ref_list); - io_rsrc_ref_unlock(ctx); + spin_unlock_irq(&ctx->rsrc_ref_lock); atomic_inc(&data_to_kill->refs); percpu_ref_kill(&rsrc_node->refs); @@ -7674,9 +7664,10 @@ static void io_rsrc_node_ref_zero(struct percpu_ref *ref) { struct io_rsrc_node *node = container_of(ref, struct io_rsrc_node, refs); struct io_ring_ctx *ctx = node->rsrc_data->ctx; + unsigned long flags; bool first_add = false; - io_rsrc_ref_lock(ctx); + spin_lock_irqsave(&ctx->rsrc_ref_lock, flags); node->done = true; while (!list_empty(&ctx->rsrc_ref_list)) { @@ -7688,7 +7679,7 @@ static void io_rsrc_node_ref_zero(struct percpu_ref *ref) list_del(&node->node); first_add |= llist_add(&node->llist, &ctx->rsrc_put_llist); } - io_rsrc_ref_unlock(ctx); + spin_unlock_irqrestore(&ctx->rsrc_ref_lock, flags); if (first_add) mod_delayed_work(system_wq, &ctx->rsrc_put_work, HZ); -- cgit v1.3-18-gd494 From 49e7f0c789add1330b111af0b7caeb0e87df063e Mon Sep 17 00:00:00 2001 From: Hao Xu Date: Sun, 8 Aug 2021 21:54:33 +0800 Subject: io-wq: fix bug of creating io-wokers unconditionally The former patch to add check between nr_workers and max_workers has a bug, which will cause unconditionally creating io-workers. That's because the result of the check doesn't affect the call of create_io_worker(), fix it by bringing in a boolean value for it. Fixes: 21698274da5b ("io-wq: fix lack of acct->nr_workers < acct->max_workers judgement") Signed-off-by: Hao Xu Link: https://lore.kernel.org/r/20210808135434.68667-2-haoxu@linux.alibaba.com [axboe: drop hunk that isn't strictly needed] Signed-off-by: Jens Axboe --- fs/io-wq.c | 12 ++++++++++-- 1 file changed, 10 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/io-wq.c b/fs/io-wq.c index 12fc19353bb0..ac8604a35169 100644 --- a/fs/io-wq.c +++ b/fs/io-wq.c @@ -282,16 +282,24 @@ static void create_worker_cb(struct callback_head *cb) struct io_wq *wq; struct io_wqe *wqe; struct io_wqe_acct *acct; + bool do_create = false; cwd = container_of(cb, struct create_worker_data, work); wqe = cwd->wqe; wq = wqe->wq; acct = &wqe->acct[cwd->index]; raw_spin_lock_irq(&wqe->lock); - if (acct->nr_workers < acct->max_workers) + if (acct->nr_workers < acct->max_workers) { acct->nr_workers++; + do_create = true; + } raw_spin_unlock_irq(&wqe->lock); - create_io_worker(wq, cwd->wqe, cwd->index); + if (do_create) { + create_io_worker(wq, cwd->wqe, cwd->index); + } else { + atomic_dec(&acct->nr_running); + io_worker_ref_put(wq); + } kfree(cwd); } -- cgit v1.3-18-gd494 From 47cae0c71f7a126903f930191e6e9f103674aca1 Mon Sep 17 00:00:00 2001 From: Hao Xu Date: Sun, 8 Aug 2021 21:54:34 +0800 Subject: io-wq: fix IO_WORKER_F_FIXED issue in create_io_worker() There may be cases like: A B spin_lock(wqe->lock) nr_workers is 0 nr_workers++ spin_unlock(wqe->lock) spin_lock(wqe->lock) nr_wokers is 1 nr_workers++ spin_unlock(wqe->lock) create_io_worker() acct->worker is 1 create_io_worker() acct->worker is 1 There should be one worker marked IO_WORKER_F_FIXED, but no one is. Fix this by introduce a new agrument for create_io_worker() to indicate if it is the first worker. Fixes: 3d4e4face9c1 ("io-wq: fix no lock protection of acct->nr_worker") Signed-off-by: Hao Xu Link: https://lore.kernel.org/r/20210808135434.68667-3-haoxu@linux.alibaba.com Signed-off-by: Jens Axboe --- fs/io-wq.c | 18 +++++++++++------- 1 file changed, 11 insertions(+), 7 deletions(-) (limited to 'fs') diff --git a/fs/io-wq.c b/fs/io-wq.c index ac8604a35169..7d2ed8c7dd31 100644 --- a/fs/io-wq.c +++ b/fs/io-wq.c @@ -129,7 +129,7 @@ struct io_cb_cancel_data { bool cancel_all; }; -static void create_io_worker(struct io_wq *wq, struct io_wqe *wqe, int index); +static void create_io_worker(struct io_wq *wq, struct io_wqe *wqe, int index, bool first); static void io_wqe_dec_running(struct io_worker *worker); static bool io_worker_get(struct io_worker *worker) @@ -248,18 +248,20 @@ static void io_wqe_wake_worker(struct io_wqe *wqe, struct io_wqe_acct *acct) rcu_read_unlock(); if (!ret) { - bool do_create = false; + bool do_create = false, first = false; raw_spin_lock_irq(&wqe->lock); if (acct->nr_workers < acct->max_workers) { atomic_inc(&acct->nr_running); atomic_inc(&wqe->wq->worker_refs); + if (!acct->nr_workers) + first = true; acct->nr_workers++; do_create = true; } raw_spin_unlock_irq(&wqe->lock); if (do_create) - create_io_worker(wqe->wq, wqe, acct->index); + create_io_worker(wqe->wq, wqe, acct->index, first); } } @@ -282,7 +284,7 @@ static void create_worker_cb(struct callback_head *cb) struct io_wq *wq; struct io_wqe *wqe; struct io_wqe_acct *acct; - bool do_create = false; + bool do_create = false, first = false; cwd = container_of(cb, struct create_worker_data, work); wqe = cwd->wqe; @@ -290,12 +292,14 @@ static void create_worker_cb(struct callback_head *cb) acct = &wqe->acct[cwd->index]; raw_spin_lock_irq(&wqe->lock); if (acct->nr_workers < acct->max_workers) { + if (!acct->nr_workers) + first = true; acct->nr_workers++; do_create = true; } raw_spin_unlock_irq(&wqe->lock); if (do_create) { - create_io_worker(wq, cwd->wqe, cwd->index); + create_io_worker(wq, wqe, cwd->index, first); } else { atomic_dec(&acct->nr_running); io_worker_ref_put(wq); @@ -637,7 +641,7 @@ void io_wq_worker_sleeping(struct task_struct *tsk) raw_spin_unlock_irq(&worker->wqe->lock); } -static void create_io_worker(struct io_wq *wq, struct io_wqe *wqe, int index) +static void create_io_worker(struct io_wq *wq, struct io_wqe *wqe, int index, bool first) { struct io_wqe_acct *acct = &wqe->acct[index]; struct io_worker *worker; @@ -678,7 +682,7 @@ fail: worker->flags |= IO_WORKER_F_FREE; if (index == IO_WQ_ACCT_BOUND) worker->flags |= IO_WORKER_F_BOUND; - if ((acct->nr_workers == 1) && (worker->flags & IO_WORKER_F_BOUND)) + if (first && (worker->flags & IO_WORKER_F_BOUND)) worker->flags |= IO_WORKER_F_FIXED; raw_spin_unlock_irq(&wqe->lock); wake_up_new_task(tsk); -- cgit v1.3-18-gd494 From c018db4a57f3e31a9cb24d528e9f094eda89a499 Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Mon, 9 Aug 2021 08:15:50 -0600 Subject: io_uring: drop ctx->uring_lock before flushing work item Ammar reports that he's seeing a lockdep splat on running test/rsrc_tags from the regression suite: ====================================================== WARNING: possible circular locking dependency detected 5.14.0-rc3-bluetea-test-00249-gc7d102232649 #5 Tainted: G OE ------------------------------------------------------ kworker/2:4/2684 is trying to acquire lock: ffff88814bb1c0a8 (&ctx->uring_lock){+.+.}-{3:3}, at: io_rsrc_put_work+0x13d/0x1a0 but task is already holding lock: ffffc90001c6be70 ((work_completion)(&(&ctx->rsrc_put_work)->work)){+.+.}-{0:0}, at: process_one_work+0x1bc/0x530 which lock already depends on the new lock. the existing dependency chain (in reverse order) is: -> #1 ((work_completion)(&(&ctx->rsrc_put_work)->work)){+.+.}-{0:0}: __flush_work+0x31b/0x490 io_rsrc_ref_quiesce.part.0.constprop.0+0x35/0xb0 __do_sys_io_uring_register+0x45b/0x1060 do_syscall_64+0x35/0xb0 entry_SYSCALL_64_after_hwframe+0x44/0xae -> #0 (&ctx->uring_lock){+.+.}-{3:3}: __lock_acquire+0x119a/0x1e10 lock_acquire+0xc8/0x2f0 __mutex_lock+0x86/0x740 io_rsrc_put_work+0x13d/0x1a0 process_one_work+0x236/0x530 worker_thread+0x52/0x3b0 kthread+0x135/0x160 ret_from_fork+0x1f/0x30 other info that might help us debug this: Possible unsafe locking scenario: CPU0 CPU1 ---- ---- lock((work_completion)(&(&ctx->rsrc_put_work)->work)); lock(&ctx->uring_lock); lock((work_completion)(&(&ctx->rsrc_put_work)->work)); lock(&ctx->uring_lock); *** DEADLOCK *** 2 locks held by kworker/2:4/2684: #0: ffff88810004d938 ((wq_completion)events){+.+.}-{0:0}, at: process_one_work+0x1bc/0x530 #1: ffffc90001c6be70 ((work_completion)(&(&ctx->rsrc_put_work)->work)){+.+.}-{0:0}, at: process_one_work+0x1bc/0x530 stack backtrace: CPU: 2 PID: 2684 Comm: kworker/2:4 Tainted: G OE 5.14.0-rc3-bluetea-test-00249-gc7d102232649 #5 Hardware name: Acer Aspire ES1-421/OLVIA_BE, BIOS V1.05 07/02/2015 Workqueue: events io_rsrc_put_work Call Trace: dump_stack_lvl+0x6a/0x9a check_noncircular+0xfe/0x110 __lock_acquire+0x119a/0x1e10 lock_acquire+0xc8/0x2f0 ? io_rsrc_put_work+0x13d/0x1a0 __mutex_lock+0x86/0x740 ? io_rsrc_put_work+0x13d/0x1a0 ? io_rsrc_put_work+0x13d/0x1a0 ? io_rsrc_put_work+0x13d/0x1a0 ? process_one_work+0x1ce/0x530 io_rsrc_put_work+0x13d/0x1a0 process_one_work+0x236/0x530 worker_thread+0x52/0x3b0 ? process_one_work+0x530/0x530 kthread+0x135/0x160 ? set_kthread_struct+0x40/0x40 ret_from_fork+0x1f/0x30 which is due to holding the ctx->uring_lock when flushing existing pending work, while the pending work flushing may need to grab the uring lock if we're using IOPOLL. Fix this by dropping the uring_lock a bit earlier as part of the flush. Cc: stable@vger.kernel.org Link: https://github.com/axboe/liburing/issues/404 Tested-by: Ammar Faizi Signed-off-by: Jens Axboe --- fs/io_uring.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/io_uring.c b/fs/io_uring.c index 6a8257233061..0c3e90f974e9 100644 --- a/fs/io_uring.c +++ b/fs/io_uring.c @@ -7195,17 +7195,19 @@ static int io_rsrc_ref_quiesce(struct io_rsrc_data *data, struct io_ring_ctx *ct /* kill initial ref, already quiesced if zero */ if (atomic_dec_and_test(&data->refs)) break; + mutex_unlock(&ctx->uring_lock); flush_delayed_work(&ctx->rsrc_put_work); ret = wait_for_completion_interruptible(&data->done); - if (!ret) + if (!ret) { + mutex_lock(&ctx->uring_lock); break; + } atomic_inc(&data->refs); /* wait for all works potentially completing data->done */ flush_delayed_work(&ctx->rsrc_put_work); reinit_completion(&data->done); - mutex_unlock(&ctx->uring_lock); ret = io_run_task_work_sig(); mutex_lock(&ctx->uring_lock); } while (ret >= 0); -- cgit v1.3-18-gd494 From 43597aac1f87230cb565ab354d331682f13d3c7a Mon Sep 17 00:00:00 2001 From: Pavel Begunkov Date: Tue, 10 Aug 2021 02:44:23 +0100 Subject: io_uring: fix ctx-exit io_rsrc_put_work() deadlock __io_rsrc_put_work() might need ->uring_lock, so nobody should wait for rsrc nodes holding the mutex. However, that's exactly what io_ring_ctx_free() does with io_wait_rsrc_data(). Split it into rsrc wait + dealloc, and move the first one out of the lock. Cc: stable@vger.kernel.org Fixes: b60c8dce33895 ("io_uring: preparation for rsrc tagging") Signed-off-by: Pavel Begunkov Link: https://lore.kernel.org/r/0130c5c2693468173ec1afab714e0885d2c9c363.1628559783.git.asml.silence@gmail.com Signed-off-by: Jens Axboe --- fs/io_uring.c | 15 ++++++++------- 1 file changed, 8 insertions(+), 7 deletions(-) (limited to 'fs') diff --git a/fs/io_uring.c b/fs/io_uring.c index 0c3e90f974e9..04c6d059ea94 100644 --- a/fs/io_uring.c +++ b/fs/io_uring.c @@ -8652,13 +8652,10 @@ static void io_req_caches_free(struct io_ring_ctx *ctx) mutex_unlock(&ctx->uring_lock); } -static bool io_wait_rsrc_data(struct io_rsrc_data *data) +static void io_wait_rsrc_data(struct io_rsrc_data *data) { - if (!data) - return false; - if (!atomic_dec_and_test(&data->refs)) + if (data && !atomic_dec_and_test(&data->refs)) wait_for_completion(&data->done); - return true; } static void io_ring_ctx_free(struct io_ring_ctx *ctx) @@ -8670,10 +8667,14 @@ static void io_ring_ctx_free(struct io_ring_ctx *ctx) ctx->mm_account = NULL; } + /* __io_rsrc_put_work() may need uring_lock to progress, wait w/o it */ + io_wait_rsrc_data(ctx->buf_data); + io_wait_rsrc_data(ctx->file_data); + mutex_lock(&ctx->uring_lock); - if (io_wait_rsrc_data(ctx->buf_data)) + if (ctx->buf_data) __io_sqe_buffers_unregister(ctx); - if (io_wait_rsrc_data(ctx->file_data)) + if (ctx->file_data) __io_sqe_files_unregister(ctx); if (ctx->rings) __io_cqring_overflow_flush(ctx, true); -- cgit v1.3-18-gd494 From 9011c2791e63fc05721b545c41ad025d8073566e Mon Sep 17 00:00:00 2001 From: Amir Goldstein Date: Mon, 26 Apr 2021 18:20:21 +0300 Subject: ovl: skip stale entries in merge dir cache iteration On the first getdents call, ovl_iterate() populates the readdir cache with a list of entries, but for upper entries with origin lower inode, p->ino remains zero. Following getdents calls traverse the readdir cache list and call ovl_cache_update_ino() for entries with zero p->ino to lookup the entry in the overlay and return d_ino that is consistent with st_ino. If the upper file was unlinked between the first getdents call and the getdents call that lists the file entry, ovl_cache_update_ino() will not find the entry and fall back to setting d_ino to the upper real st_ino, which is inconsistent with how this object was presented to users. Instead of listing a stale entry with inconsistent d_ino, simply skip the stale entry, which is better for users. xfstest overlay/077 is failing without this patch. Signed-off-by: Amir Goldstein Link: https://lore.kernel.org/fstests/CAOQ4uxgR_cLnC_vdU5=seP3fwqVkuZM_-WfD6maFTMbMYq=a9w@mail.gmail.com/ Signed-off-by: Miklos Szeredi --- fs/overlayfs/readdir.c | 5 +++++ 1 file changed, 5 insertions(+) (limited to 'fs') diff --git a/fs/overlayfs/readdir.c b/fs/overlayfs/readdir.c index e8ad2c2c77dd..150fdf3bc68d 100644 --- a/fs/overlayfs/readdir.c +++ b/fs/overlayfs/readdir.c @@ -481,6 +481,8 @@ static int ovl_cache_update_ino(struct path *path, struct ovl_cache_entry *p) } this = lookup_one_len(p->name, dir, p->len); if (IS_ERR_OR_NULL(this) || !this->d_inode) { + /* Mark a stale entry */ + p->is_whiteout = true; if (IS_ERR(this)) { err = PTR_ERR(this); this = NULL; @@ -776,6 +778,9 @@ static int ovl_iterate(struct file *file, struct dir_context *ctx) if (err) goto out; } + } + /* ovl_cache_update_ino() sets is_whiteout on stale entry */ + if (!p->is_whiteout) { if (!dir_emit(ctx, p->name, p->len, p->ino, p->type)) break; } -- cgit v1.3-18-gd494 From 9b91b6b019fda817eb52f728eb9c79b3579760bc Mon Sep 17 00:00:00 2001 From: Miklos Szeredi Date: Wed, 28 Jul 2021 10:38:43 +0200 Subject: ovl: fix deadlock in splice write There's possibility of an ABBA deadlock in case of a splice write to an overlayfs file and a concurrent splice write to a corresponding real file. The call chain for splice to an overlay file: -> do_splice [takes sb_writers on overlay file] -> do_splice_from -> iter_file_splice_write [takes pipe->mutex] -> vfs_iter_write ... -> ovl_write_iter [takes sb_writers on real file] And the call chain for splice to a real file: -> do_splice [takes sb_writers on real file] -> do_splice_from -> iter_file_splice_write [takes pipe->mutex] Syzbot successfully bisected this to commit 82a763e61e2b ("ovl: simplify file splice"). Fix by reverting the write part of the above commit and by adding missing bits from ovl_write_iter() into ovl_splice_write(). Fixes: 82a763e61e2b ("ovl: simplify file splice") Reported-and-tested-by: syzbot+579885d1a9a833336209@syzkaller.appspotmail.com Signed-off-by: Miklos Szeredi --- fs/overlayfs/file.c | 47 ++++++++++++++++++++++++++++++++++++++++++++++- 1 file changed, 46 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/overlayfs/file.c b/fs/overlayfs/file.c index 4d53d3b7e5fe..d081faa55e83 100644 --- a/fs/overlayfs/file.c +++ b/fs/overlayfs/file.c @@ -392,6 +392,51 @@ out_unlock: return ret; } +/* + * Calling iter_file_splice_write() directly from overlay's f_op may deadlock + * due to lock order inversion between pipe->mutex in iter_file_splice_write() + * and file_start_write(real.file) in ovl_write_iter(). + * + * So do everything ovl_write_iter() does and call iter_file_splice_write() on + * the real file. + */ +static ssize_t ovl_splice_write(struct pipe_inode_info *pipe, struct file *out, + loff_t *ppos, size_t len, unsigned int flags) +{ + struct fd real; + const struct cred *old_cred; + struct inode *inode = file_inode(out); + struct inode *realinode = ovl_inode_real(inode); + ssize_t ret; + + inode_lock(inode); + /* Update mode */ + ovl_copyattr(realinode, inode); + ret = file_remove_privs(out); + if (ret) + goto out_unlock; + + ret = ovl_real_fdget(out, &real); + if (ret) + goto out_unlock; + + old_cred = ovl_override_creds(inode->i_sb); + file_start_write(real.file); + + ret = iter_file_splice_write(pipe, real.file, ppos, len, flags); + + file_end_write(real.file); + /* Update size */ + ovl_copyattr(realinode, inode); + revert_creds(old_cred); + fdput(real); + +out_unlock: + inode_unlock(inode); + + return ret; +} + static int ovl_fsync(struct file *file, loff_t start, loff_t end, int datasync) { struct fd real; @@ -603,7 +648,7 @@ const struct file_operations ovl_file_operations = { .fadvise = ovl_fadvise, .flush = ovl_flush, .splice_read = generic_file_splice_read, - .splice_write = iter_file_splice_write, + .splice_write = ovl_splice_write, .copy_file_range = ovl_copy_file_range, .remap_file_range = ovl_remap_file_range, -- cgit v1.3-18-gd494 From 580c610429b3994e8db24418927747cf28443cde Mon Sep 17 00:00:00 2001 From: Miklos Szeredi Date: Fri, 6 Aug 2021 10:03:12 +0200 Subject: ovl: fix uninitialized pointer read in ovl_lookup_real_one() One error path can result in release_dentry_name_snapshot() being called before "name" was initialized by take_dentry_name_snapshot(). Fix by moving the release_dentry_name_snapshot() to immediately after the only use. Reported-by: Colin Ian King Signed-off-by: Miklos Szeredi --- fs/overlayfs/export.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/overlayfs/export.c b/fs/overlayfs/export.c index 41ebf52f1bbc..ebde05c9cf62 100644 --- a/fs/overlayfs/export.c +++ b/fs/overlayfs/export.c @@ -392,6 +392,7 @@ static struct dentry *ovl_lookup_real_one(struct dentry *connected, */ take_dentry_name_snapshot(&name, real); this = lookup_one_len(name.name.name, connected, name.name.len); + release_dentry_name_snapshot(&name); err = PTR_ERR(this); if (IS_ERR(this)) { goto fail; @@ -406,7 +407,6 @@ static struct dentry *ovl_lookup_real_one(struct dentry *connected, } out: - release_dentry_name_snapshot(&name); dput(parent); inode_unlock(dir); return this; -- cgit v1.3-18-gd494 From 427215d85e8d1476da1a86b8d67aceb485eb3631 Mon Sep 17 00:00:00 2001 From: Miklos Szeredi Date: Mon, 9 Aug 2021 10:19:47 +0200 Subject: ovl: prevent private clone if bind mount is not allowed Add the following checks from __do_loopback() to clone_private_mount() as well: - verify that the mount is in the current namespace - verify that there are no locked children Reported-by: Alois Wohlschlager Fixes: c771d683a62e ("vfs: introduce clone_private_mount()") Cc: # v3.18 Signed-off-by: Miklos Szeredi --- fs/namespace.c | 42 ++++++++++++++++++++++++++++-------------- 1 file changed, 28 insertions(+), 14 deletions(-) (limited to 'fs') diff --git a/fs/namespace.c b/fs/namespace.c index ab4174a3c802..f79d9471cb76 100644 --- a/fs/namespace.c +++ b/fs/namespace.c @@ -1938,6 +1938,20 @@ void drop_collected_mounts(struct vfsmount *mnt) namespace_unlock(); } +static bool has_locked_children(struct mount *mnt, struct dentry *dentry) +{ + struct mount *child; + + list_for_each_entry(child, &mnt->mnt_mounts, mnt_child) { + if (!is_subdir(child->mnt_mountpoint, dentry)) + continue; + + if (child->mnt.mnt_flags & MNT_LOCKED) + return true; + } + return false; +} + /** * clone_private_mount - create a private clone of a path * @path: path to clone @@ -1953,10 +1967,19 @@ struct vfsmount *clone_private_mount(const struct path *path) struct mount *old_mnt = real_mount(path->mnt); struct mount *new_mnt; + down_read(&namespace_sem); if (IS_MNT_UNBINDABLE(old_mnt)) - return ERR_PTR(-EINVAL); + goto invalid; + + if (!check_mnt(old_mnt)) + goto invalid; + + if (has_locked_children(old_mnt, path->dentry)) + goto invalid; new_mnt = clone_mnt(old_mnt, path->dentry, CL_PRIVATE); + up_read(&namespace_sem); + if (IS_ERR(new_mnt)) return ERR_CAST(new_mnt); @@ -1964,6 +1987,10 @@ struct vfsmount *clone_private_mount(const struct path *path) new_mnt->mnt_ns = MNT_NS_INTERNAL; return &new_mnt->mnt; + +invalid: + up_read(&namespace_sem); + return ERR_PTR(-EINVAL); } EXPORT_SYMBOL_GPL(clone_private_mount); @@ -2315,19 +2342,6 @@ static int do_change_type(struct path *path, int ms_flags) return err; } -static bool has_locked_children(struct mount *mnt, struct dentry *dentry) -{ - struct mount *child; - list_for_each_entry(child, &mnt->mnt_mounts, mnt_child) { - if (!is_subdir(child->mnt_mountpoint, dentry)) - continue; - - if (child->mnt.mnt_flags & MNT_LOCKED) - return true; - } - return false; -} - static struct mount *__do_loopback(struct path *old_path, int recurse) { struct mount *mnt = ERR_PTR(-EINVAL), *old = real_mount(old_path->mnt); -- cgit v1.3-18-gd494 From 981567bd965329df7e64b13e92a54da816c1e0a4 Mon Sep 17 00:00:00 2001 From: Ronnie Sahlberg Date: Tue, 10 Aug 2021 16:33:55 +1000 Subject: cifs: use the correct max-length for dentry_path_raw() RHBZ: 1972502 PATH_MAX is 4096 but PAGE_SIZE can be >4096 on some architectures such as ppc and would thus write beyond the end of the actual object. Cc: Reported-by: Xiaoli Feng Suggested-by: Brian foster Reviewed-by: Paulo Alcantara (SUSE) Signed-off-by: Ronnie Sahlberg Signed-off-by: Steve French --- fs/cifs/dir.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/cifs/dir.c b/fs/cifs/dir.c index 79402ca0ddfa..5f8a302ffcb2 100644 --- a/fs/cifs/dir.c +++ b/fs/cifs/dir.c @@ -100,7 +100,7 @@ build_path_from_dentry_optional_prefix(struct dentry *direntry, void *page, if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_USE_PREFIX_PATH) pplen = cifs_sb->prepath ? strlen(cifs_sb->prepath) + 1 : 0; - s = dentry_path_raw(direntry, page, PAGE_SIZE); + s = dentry_path_raw(direntry, page, PATH_MAX); if (IS_ERR(s)) return s; if (!s[1]) // for root we want "", not "/" -- cgit v1.3-18-gd494 From 41535701da3324b80029cabb501e86c4fafe339d Mon Sep 17 00:00:00 2001 From: Rohith Surabattula Date: Thu, 29 Jul 2021 07:45:29 +0000 Subject: cifs: Handle race conditions during rename When rename is executed on directory which has files for which close is deferred, then rename will fail with EACCES. This patch will try to close all deferred files when EACCES is received and retry rename on a directory. Signed-off-by: Rohith Surabattula Cc: stable@vger.kernel.org # 5.13 Reviewed-by: Shyam Prasad N Signed-off-by: Steve French --- fs/cifs/inode.c | 19 +++++++++++++++++-- fs/cifs/misc.c | 16 +++++++++++----- 2 files changed, 28 insertions(+), 7 deletions(-) (limited to 'fs') diff --git a/fs/cifs/inode.c b/fs/cifs/inode.c index b96b253e7635..65f8a70cece3 100644 --- a/fs/cifs/inode.c +++ b/fs/cifs/inode.c @@ -1625,7 +1625,7 @@ int cifs_unlink(struct inode *dir, struct dentry *dentry) goto unlink_out; } - cifs_close_all_deferred_files(tcon); + cifs_close_deferred_file(CIFS_I(inode)); if (cap_unix(tcon->ses) && (CIFS_UNIX_POSIX_PATH_OPS_CAP & le64_to_cpu(tcon->fsUnixInfo.Capability))) { rc = CIFSPOSIXDelFile(xid, tcon, full_path, @@ -2084,6 +2084,7 @@ cifs_rename2(struct user_namespace *mnt_userns, struct inode *source_dir, FILE_UNIX_BASIC_INFO *info_buf_target; unsigned int xid; int rc, tmprc; + int retry_count = 0; if (flags & ~RENAME_NOREPLACE) return -EINVAL; @@ -2113,10 +2114,24 @@ cifs_rename2(struct user_namespace *mnt_userns, struct inode *source_dir, goto cifs_rename_exit; } - cifs_close_all_deferred_files(tcon); + cifs_close_deferred_file(CIFS_I(d_inode(source_dentry))); + if (d_inode(target_dentry) != NULL) + cifs_close_deferred_file(CIFS_I(d_inode(target_dentry))); + rc = cifs_do_rename(xid, source_dentry, from_name, target_dentry, to_name); + if (rc == -EACCES) { + while (retry_count < 3) { + cifs_close_all_deferred_files(tcon); + rc = cifs_do_rename(xid, source_dentry, from_name, target_dentry, + to_name); + if (rc != -EACCES) + break; + retry_count++; + } + } + /* * No-replace is the natural behavior for CIFS, so skip unlink hacks. */ diff --git a/fs/cifs/misc.c b/fs/cifs/misc.c index 844abeb2b48f..cdb1ec1461de 100644 --- a/fs/cifs/misc.c +++ b/fs/cifs/misc.c @@ -723,13 +723,19 @@ void cifs_close_deferred_file(struct cifsInodeInfo *cifs_inode) { struct cifsFileInfo *cfile = NULL; - struct cifs_deferred_close *dclose; + + if (cifs_inode == NULL) + return; list_for_each_entry(cfile, &cifs_inode->openFileList, flist) { - spin_lock(&cifs_inode->deferred_lock); - if (cifs_is_deferred_close(cfile, &dclose)) - mod_delayed_work(deferredclose_wq, &cfile->deferred, 0); - spin_unlock(&cifs_inode->deferred_lock); + if (delayed_work_pending(&cfile->deferred)) { + /* + * If there is no pending work, mod_delayed_work queues new work. + * So, Increase the ref count to avoid use-after-free. + */ + if (!mod_delayed_work(deferredclose_wq, &cfile->deferred, 0)) + cifsFileInfo_get(cfile); + } } } -- cgit v1.3-18-gd494 From 9e992755be8f2d458a0bcbefd19e493483c1dba2 Mon Sep 17 00:00:00 2001 From: Rohith Surabattula Date: Mon, 9 Aug 2021 09:32:46 +0000 Subject: cifs: Call close synchronously during unlink/rename/lease break. During unlink/rename/lease break, deferred work for close is scheduled immediately but in an asynchronous manner which might lead to race with actual(unlink/rename) commands. This change will schedule close synchronously which will avoid the race conditions with other commands. Signed-off-by: Rohith Surabattula Reviewed-by: Shyam Prasad N Cc: stable@vger.kernel.org # 5.13 Signed-off-by: Steve French --- fs/cifs/cifsglob.h | 5 +++++ fs/cifs/file.c | 35 +++++++++++++++++------------------ fs/cifs/misc.c | 46 ++++++++++++++++++++++++++++++++++------------ 3 files changed, 56 insertions(+), 30 deletions(-) (limited to 'fs') diff --git a/fs/cifs/cifsglob.h b/fs/cifs/cifsglob.h index c0bfc2f01030..c6a9542ca281 100644 --- a/fs/cifs/cifsglob.h +++ b/fs/cifs/cifsglob.h @@ -1611,6 +1611,11 @@ struct dfs_info3_param { int ttl; }; +struct file_list { + struct list_head list; + struct cifsFileInfo *cfile; +}; + /* * common struct for holding inode info when searching for or updating an * inode with new info diff --git a/fs/cifs/file.c b/fs/cifs/file.c index 0a72840a88f1..bb98fbdd22a9 100644 --- a/fs/cifs/file.c +++ b/fs/cifs/file.c @@ -4847,17 +4847,6 @@ void cifs_oplock_break(struct work_struct *work) cifs_dbg(VFS, "Push locks rc = %d\n", rc); oplock_break_ack: - /* - * releasing stale oplock after recent reconnect of smb session using - * a now incorrect file handle is not a data integrity issue but do - * not bother sending an oplock release if session to server still is - * disconnected since oplock already released by the server - */ - if (!cfile->oplock_break_cancelled) { - rc = tcon->ses->server->ops->oplock_response(tcon, &cfile->fid, - cinode); - cifs_dbg(FYI, "Oplock release rc = %d\n", rc); - } /* * When oplock break is received and there are no active * file handles but cached, then schedule deferred close immediately. @@ -4865,17 +4854,27 @@ oplock_break_ack: */ spin_lock(&CIFS_I(inode)->deferred_lock); is_deferred = cifs_is_deferred_close(cfile, &dclose); + spin_unlock(&CIFS_I(inode)->deferred_lock); if (is_deferred && cfile->deferred_close_scheduled && delayed_work_pending(&cfile->deferred)) { - /* - * If there is no pending work, mod_delayed_work queues new work. - * So, Increase the ref count to avoid use-after-free. - */ - if (!mod_delayed_work(deferredclose_wq, &cfile->deferred, 0)) - cifsFileInfo_get(cfile); + if (cancel_delayed_work(&cfile->deferred)) { + _cifsFileInfo_put(cfile, false, false); + goto oplock_break_done; + } } - spin_unlock(&CIFS_I(inode)->deferred_lock); + /* + * releasing stale oplock after recent reconnect of smb session using + * a now incorrect file handle is not a data integrity issue but do + * not bother sending an oplock release if session to server still is + * disconnected since oplock already released by the server + */ + if (!cfile->oplock_break_cancelled) { + rc = tcon->ses->server->ops->oplock_response(tcon, &cfile->fid, + cinode); + cifs_dbg(FYI, "Oplock release rc = %d\n", rc); + } +oplock_break_done: _cifsFileInfo_put(cfile, false /* do not wait for ourself */, false); cifs_done_oplock_break(cinode); } diff --git a/fs/cifs/misc.c b/fs/cifs/misc.c index cdb1ec1461de..9469f1cf0b46 100644 --- a/fs/cifs/misc.c +++ b/fs/cifs/misc.c @@ -723,20 +723,32 @@ void cifs_close_deferred_file(struct cifsInodeInfo *cifs_inode) { struct cifsFileInfo *cfile = NULL; + struct file_list *tmp_list, *tmp_next_list; + struct list_head file_head; if (cifs_inode == NULL) return; + INIT_LIST_HEAD(&file_head); + spin_lock(&cifs_inode->open_file_lock); list_for_each_entry(cfile, &cifs_inode->openFileList, flist) { if (delayed_work_pending(&cfile->deferred)) { - /* - * If there is no pending work, mod_delayed_work queues new work. - * So, Increase the ref count to avoid use-after-free. - */ - if (!mod_delayed_work(deferredclose_wq, &cfile->deferred, 0)) - cifsFileInfo_get(cfile); + if (cancel_delayed_work(&cfile->deferred)) { + tmp_list = kmalloc(sizeof(struct file_list), GFP_ATOMIC); + if (tmp_list == NULL) + continue; + tmp_list->cfile = cfile; + list_add_tail(&tmp_list->list, &file_head); + } } } + spin_unlock(&cifs_inode->open_file_lock); + + list_for_each_entry_safe(tmp_list, tmp_next_list, &file_head, list) { + _cifsFileInfo_put(tmp_list->cfile, true, false); + list_del(&tmp_list->list); + kfree(tmp_list); + } } void @@ -744,20 +756,30 @@ cifs_close_all_deferred_files(struct cifs_tcon *tcon) { struct cifsFileInfo *cfile; struct list_head *tmp; + struct file_list *tmp_list, *tmp_next_list; + struct list_head file_head; + INIT_LIST_HEAD(&file_head); spin_lock(&tcon->open_file_lock); list_for_each(tmp, &tcon->openFileList) { cfile = list_entry(tmp, struct cifsFileInfo, tlist); if (delayed_work_pending(&cfile->deferred)) { - /* - * If there is no pending work, mod_delayed_work queues new work. - * So, Increase the ref count to avoid use-after-free. - */ - if (!mod_delayed_work(deferredclose_wq, &cfile->deferred, 0)) - cifsFileInfo_get(cfile); + if (cancel_delayed_work(&cfile->deferred)) { + tmp_list = kmalloc(sizeof(struct file_list), GFP_ATOMIC); + if (tmp_list == NULL) + continue; + tmp_list->cfile = cfile; + list_add_tail(&tmp_list->list, &file_head); + } } } spin_unlock(&tcon->open_file_lock); + + list_for_each_entry_safe(tmp_list, tmp_next_list, &file_head, list) { + _cifsFileInfo_put(tmp_list->cfile, true, false); + list_del(&tmp_list->list); + kfree(tmp_list); + } } /* parses DFS refferal V3 structure -- cgit v1.3-18-gd494 From 21f965221e7c42609521342403e8fb91b8b3e76e Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Sat, 14 Aug 2021 09:04:40 -0600 Subject: io_uring: only assign io_uring_enter() SQPOLL error in actual error case If an SQPOLL based ring is newly created and an application issues an io_uring_enter(2) system call on it, then we can return a spurious -EOWNERDEAD error. This happens because there's nothing to submit, and if the caller doesn't specify any other action, the initial error assignment of -EOWNERDEAD never gets overwritten. This causes us to return it directly, even if it isn't valid. Move the error assignment into the actual failure case instead. Cc: stable@vger.kernel.org Fixes: d9d05217cb69 ("io_uring: stop SQPOLL submit on creator's death") Reported-by: Sherlock Holo sherlockya@gmail.com Link: https://github.com/axboe/liburing/issues/413 Signed-off-by: Jens Axboe --- fs/io_uring.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/io_uring.c b/fs/io_uring.c index 04c6d059ea94..6a092a534d2b 100644 --- a/fs/io_uring.c +++ b/fs/io_uring.c @@ -9370,9 +9370,10 @@ SYSCALL_DEFINE6(io_uring_enter, unsigned int, fd, u32, to_submit, if (ctx->flags & IORING_SETUP_SQPOLL) { io_cqring_overflow_flush(ctx, false); - ret = -EOWNERDEAD; - if (unlikely(ctx->sq_data->thread == NULL)) + if (unlikely(ctx->sq_data->thread == NULL)) { + ret = -EOWNERDEAD; goto out; + } if (flags & IORING_ENTER_SQ_WAKEUP) wake_up(&ctx->sq_data->wait); if (flags & IORING_ENTER_SQ_WAIT) { -- cgit v1.3-18-gd494 From 3f79f6f6247c83f448c8026c3ee16d4636ef8d4f Mon Sep 17 00:00:00 2001 From: NeilBrown Date: Fri, 6 Aug 2021 14:26:24 +1000 Subject: btrfs: prevent rename2 from exchanging a subvol with a directory from different parents Cross-rename lacks a check when that would prevent exchanging a directory and subvolume from different parent subvolume. This causes data inconsistencies and is caught before commit by tree-checker, turning the filesystem to read-only. Calling the renameat2 with RENAME_EXCHANGE flags like renameat2(AT_FDCWD, namesrc, AT_FDCWD, namedest, (1 << 1)) on two paths: namesrc = dir1/subvol1/dir2 namedest = subvol2/subvol3 will cause key order problem with following write time tree-checker report: [1194842.307890] BTRFS critical (device loop1): corrupt leaf: root=5 block=27574272 slot=10 ino=258, invalid previous key objectid, have 257 expect 258 [1194842.322221] BTRFS info (device loop1): leaf 27574272 gen 8 total ptrs 11 free space 15444 owner 5 [1194842.331562] BTRFS info (device loop1): refs 2 lock_owner 0 current 26561 [1194842.338772] item 0 key (256 1 0) itemoff 16123 itemsize 160 [1194842.338793] inode generation 3 size 16 mode 40755 [1194842.338801] item 1 key (256 12 256) itemoff 16111 itemsize 12 [1194842.338809] item 2 key (256 84 2248503653) itemoff 16077 itemsize 34 [1194842.338817] dir oid 258 type 2 [1194842.338823] item 3 key (256 84 2363071922) itemoff 16043 itemsize 34 [1194842.338830] dir oid 257 type 2 [1194842.338836] item 4 key (256 96 2) itemoff 16009 itemsize 34 [1194842.338843] item 5 key (256 96 3) itemoff 15975 itemsize 34 [1194842.338852] item 6 key (257 1 0) itemoff 15815 itemsize 160 [1194842.338863] inode generation 6 size 8 mode 40755 [1194842.338869] item 7 key (257 12 256) itemoff 15801 itemsize 14 [1194842.338876] item 8 key (257 84 2505409169) itemoff 15767 itemsize 34 [1194842.338883] dir oid 256 type 2 [1194842.338888] item 9 key (257 96 2) itemoff 15733 itemsize 34 [1194842.338895] item 10 key (258 12 256) itemoff 15719 itemsize 14 [1194842.339163] BTRFS error (device loop1): block=27574272 write time tree block corruption detected [1194842.339245] ------------[ cut here ]------------ [1194842.443422] WARNING: CPU: 6 PID: 26561 at fs/btrfs/disk-io.c:449 csum_one_extent_buffer+0xed/0x100 [btrfs] [1194842.511863] CPU: 6 PID: 26561 Comm: kworker/u17:2 Not tainted 5.14.0-rc3-git+ #793 [1194842.511870] Hardware name: empty empty/S3993, BIOS PAQEX0-3 02/24/2008 [1194842.511876] Workqueue: btrfs-worker-high btrfs_work_helper [btrfs] [1194842.511976] RIP: 0010:csum_one_extent_buffer+0xed/0x100 [btrfs] [1194842.512068] RSP: 0018:ffffa2c284d77da0 EFLAGS: 00010282 [1194842.512074] RAX: 0000000000000000 RBX: 0000000000001000 RCX: ffff928867bd9978 [1194842.512078] RDX: 0000000000000000 RSI: 0000000000000027 RDI: ffff928867bd9970 [1194842.512081] RBP: ffff92876b958000 R08: 0000000000000001 R09: 00000000000c0003 [1194842.512085] R10: 0000000000000000 R11: 0000000000000001 R12: 0000000000000000 [1194842.512088] R13: ffff92875f989f98 R14: 0000000000000000 R15: 0000000000000000 [1194842.512092] FS: 0000000000000000(0000) GS:ffff928867a00000(0000) knlGS:0000000000000000 [1194842.512095] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [1194842.512099] CR2: 000055f5384da1f0 CR3: 0000000102fe4000 CR4: 00000000000006e0 [1194842.512103] Call Trace: [1194842.512128] ? run_one_async_free+0x10/0x10 [btrfs] [1194842.631729] btree_csum_one_bio+0x1ac/0x1d0 [btrfs] [1194842.631837] run_one_async_start+0x18/0x30 [btrfs] [1194842.631938] btrfs_work_helper+0xd5/0x1d0 [btrfs] [1194842.647482] process_one_work+0x262/0x5e0 [1194842.647520] worker_thread+0x4c/0x320 [1194842.655935] ? process_one_work+0x5e0/0x5e0 [1194842.655946] kthread+0x135/0x160 [1194842.655953] ? set_kthread_struct+0x40/0x40 [1194842.655965] ret_from_fork+0x1f/0x30 [1194842.672465] irq event stamp: 1729 [1194842.672469] hardirqs last enabled at (1735): [] console_trylock_spinning+0x185/0x1a0 [1194842.672477] hardirqs last disabled at (1740): [] console_trylock_spinning+0x15c/0x1a0 [1194842.672482] softirqs last enabled at (1666): [] __do_softirq+0x2e1/0x50a [1194842.672491] softirqs last disabled at (1651): [] __irq_exit_rcu+0xa7/0xd0 The corrupted data will not be written, and filesystem can be unmounted and mounted again (all changes since the last commit will be lost). Add the missing check for new_ino so that all non-subvolumes must reside under the same parent subvolume. There's an exception allowing to exchange two subvolumes from any parents as the directory representing a subvolume is only a logical link and does not have any other structures related to the parent subvolume, unlike files, directories etc, that are always in the inode namespace of the parent subvolume. Fixes: cdd1fedf8261 ("btrfs: add support for RENAME_EXCHANGE and RENAME_WHITEOUT") CC: stable@vger.kernel.org # 4.7+ Reviewed-by: Nikolay Borisov Signed-off-by: NeilBrown Reviewed-by: David Sterba Signed-off-by: David Sterba --- fs/btrfs/inode.c | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 0117d867ecf8..06f9f167222b 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -9226,8 +9226,14 @@ static int btrfs_rename_exchange(struct inode *old_dir, bool dest_log_pinned = false; bool need_abort = false; - /* we only allow rename subvolume link between subvolumes */ - if (old_ino != BTRFS_FIRST_FREE_OBJECTID && root != dest) + /* + * For non-subvolumes allow exchange only within one subvolume, in the + * same inode namespace. Two subvolumes (represented as directory) can + * be exchanged as they're a logical link and have a fixed inode number. + */ + if (root != dest && + (old_ino != BTRFS_FIRST_FREE_OBJECTID || + new_ino != BTRFS_FIRST_FREE_OBJECTID)) return -EXDEV; /* close the race window with snapshot create/destroy ioctl */ -- cgit v1.3-18-gd494 From 9cb0073b302a6b8a8c1015ff31b2b3ab4900f866 Mon Sep 17 00:00:00 2001 From: Pavel Begunkov Date: Tue, 17 Aug 2021 22:36:44 +0100 Subject: io_uring: pin ctx on fallback execution Pin ring in io_fallback_req_func() by briefly elevating ctx->refs in case any task_work handler touches ctx after releasing a request. Fixes: 9011bf9a13e3b ("io_uring: fix stuck fallback reqs") Signed-off-by: Pavel Begunkov Link: https://lore.kernel.org/r/833a494713d235ec144284a9bbfe418df4f6b61c.1629235576.git.asml.silence@gmail.com Signed-off-by: Jens Axboe --- fs/io_uring.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'fs') diff --git a/fs/io_uring.c b/fs/io_uring.c index 6a092a534d2b..979941bcd15a 100644 --- a/fs/io_uring.c +++ b/fs/io_uring.c @@ -2477,8 +2477,10 @@ static void io_fallback_req_func(struct work_struct *work) struct llist_node *node = llist_del_all(&ctx->fallback_llist); struct io_kiocb *req, *tmp; + percpu_ref_get(&ctx->refs); llist_for_each_entry_safe(req, tmp, node, io_task_work.fallback_node) req->io_task_work.func(req); + percpu_ref_put(&ctx->refs); } static void __io_complete_rw(struct io_kiocb *req, long res, long res2, -- cgit v1.3-18-gd494 From 3b844826b6c6affa80755254da322b017358a2f4 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Thu, 5 Aug 2021 10:04:43 -0700 Subject: pipe: avoid unnecessary EPOLLET wakeups under normal loads I had forgotten just how sensitive hackbench is to extra pipe wakeups, and commit 3a34b13a88ca ("pipe: make pipe writes always wake up readers") ended up causing a quite noticeable regression on larger machines. Now, hackbench isn't necessarily a hugely meaningful benchmark, and it's not clear that this matters in real life all that much, but as Mel points out, it's used often enough when comparing kernels and so the performance regression shows up like a sore thumb. It's easy enough to fix at least for the common cases where pipes are used purely for data transfer, and you never have any exciting poll usage at all. So set a special 'poll_usage' flag when there is polling activity, and make the ugly "EPOLLET has crazy legacy expectations" semantics explicit to only that case. I would love to limit it to just the broken EPOLLET case, but the pipe code can't see the difference between epoll and regular select/poll, so any non-read/write waiting will trigger the extra wakeup behavior. That is sufficient for at least the hackbench case. Apart from making the odd extra wakeup cases more explicitly about EPOLLET, this also makes the extra wakeup be at the _end_ of the pipe write, not at the first write chunk. That is actually much saner semantics (as much as you can call any of the legacy edge-triggered expectations for EPOLLET "sane") since it means that you know the wakeup will happen once the write is done, rather than possibly in the middle of one. [ For stable people: I'm putting a "Fixes" tag on this, but I leave it up to you to decide whether you actually want to backport it or not. It likely has no impact outside of synthetic benchmarks - Linus ] Link: https://lore.kernel.org/lkml/20210802024945.GA8372@xsang-OptiPlex-9020/ Fixes: 3a34b13a88ca ("pipe: make pipe writes always wake up readers") Reported-by: kernel test robot Tested-by: Sandeep Patil Tested-by: Mel Gorman Signed-off-by: Linus Torvalds --- fs/pipe.c | 15 +++++++++------ include/linux/pipe_fs_i.h | 2 ++ 2 files changed, 11 insertions(+), 6 deletions(-) (limited to 'fs') diff --git a/fs/pipe.c b/fs/pipe.c index 8e6ef62aeb1c..678dee2a8228 100644 --- a/fs/pipe.c +++ b/fs/pipe.c @@ -444,9 +444,6 @@ pipe_write(struct kiocb *iocb, struct iov_iter *from) #endif /* - * Epoll nonsensically wants a wakeup whether the pipe - * was already empty or not. - * * If it wasn't empty we try to merge new data into * the last buffer. * @@ -455,9 +452,9 @@ pipe_write(struct kiocb *iocb, struct iov_iter *from) * spanning multiple pages. */ head = pipe->head; - was_empty = true; + was_empty = pipe_empty(head, pipe->tail); chars = total_len & (PAGE_SIZE-1); - if (chars && !pipe_empty(head, pipe->tail)) { + if (chars && !was_empty) { unsigned int mask = pipe->ring_size - 1; struct pipe_buffer *buf = &pipe->bufs[(head - 1) & mask]; int offset = buf->offset + buf->len; @@ -590,8 +587,11 @@ out: * This is particularly important for small writes, because of * how (for example) the GNU make jobserver uses small writes to * wake up pending jobs + * + * Epoll nonsensically wants a wakeup whether the pipe + * was already empty or not. */ - if (was_empty) { + if (was_empty || pipe->poll_usage) { wake_up_interruptible_sync_poll(&pipe->rd_wait, EPOLLIN | EPOLLRDNORM); kill_fasync(&pipe->fasync_readers, SIGIO, POLL_IN); } @@ -654,6 +654,9 @@ pipe_poll(struct file *filp, poll_table *wait) struct pipe_inode_info *pipe = filp->private_data; unsigned int head, tail; + /* Epoll has some historical nasty semantics, this enables them */ + pipe->poll_usage = 1; + /* * Reading pipe state only -- no need for acquiring the semaphore. * diff --git a/include/linux/pipe_fs_i.h b/include/linux/pipe_fs_i.h index 5d2705f1d01c..fc5642431b92 100644 --- a/include/linux/pipe_fs_i.h +++ b/include/linux/pipe_fs_i.h @@ -48,6 +48,7 @@ struct pipe_buffer { * @files: number of struct file referring this pipe (protected by ->i_lock) * @r_counter: reader counter * @w_counter: writer counter + * @poll_usage: is this pipe used for epoll, which has crazy wakeups? * @fasync_readers: reader side fasync * @fasync_writers: writer side fasync * @bufs: the circular array of pipe buffers @@ -70,6 +71,7 @@ struct pipe_inode_info { unsigned int files; unsigned int r_counter; unsigned int w_counter; + unsigned int poll_usage; struct page *tmp_page; struct fasync_struct *fasync_readers; struct fasync_struct *fasync_writers; -- cgit v1.3-18-gd494 From a30f895ad3239f45012e860d4f94c1a388b36d14 Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Fri, 20 Aug 2021 14:53:59 -0600 Subject: io_uring: fix xa_alloc_cycle() error return value check We currently check for ret != 0 to indicate error, but '1' is a valid return and just indicates that the allocation succeeded with a wrap. Correct the check to be for < 0, like it was before the xarray conversion. Cc: stable@vger.kernel.org Fixes: 61cf93700fe6 ("io_uring: Convert personality_idr to XArray") Signed-off-by: Jens Axboe --- fs/io_uring.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/io_uring.c b/fs/io_uring.c index 979941bcd15a..a2e20a6fbfed 100644 --- a/fs/io_uring.c +++ b/fs/io_uring.c @@ -9843,10 +9843,11 @@ static int io_register_personality(struct io_ring_ctx *ctx) ret = xa_alloc_cyclic(&ctx->personalities, &id, (void *)creds, XA_LIMIT(0, USHRT_MAX), &ctx->pers_next, GFP_KERNEL); - if (!ret) - return id; - put_cred(creds); - return ret; + if (ret < 0) { + put_cred(creds); + return ret; + } + return id; } static int io_register_restrictions(struct io_ring_ctx *ctx, void __user *arg, -- cgit v1.3-18-gd494 From fdd92b64d15bc4aec973caa25899afd782402e68 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Fri, 20 Aug 2021 09:29:50 -0400 Subject: fs: warn about impending deprecation of mandatory locks We've had CONFIG_MANDATORY_FILE_LOCKING since 2015 and a lot of distros have disabled it. Warn the stragglers that still use "-o mand" that we'll be dropping support for that mount option. Cc: stable@vger.kernel.org Signed-off-by: Jeff Layton --- fs/namespace.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/namespace.c b/fs/namespace.c index ab4174a3c802..2279473d0d6f 100644 --- a/fs/namespace.c +++ b/fs/namespace.c @@ -1716,8 +1716,12 @@ static inline bool may_mount(void) } #ifdef CONFIG_MANDATORY_FILE_LOCKING -static inline bool may_mandlock(void) +static bool may_mandlock(void) { + pr_warn_once("======================================================\n" + "WARNING: the mand mount option is being deprecated and\n" + " will be removed in v5.15!\n" + "======================================================\n"); return capable(CAP_SYS_ADMIN); } #else -- cgit v1.3-18-gd494