procfs: fdinfo -- Extend information about epoll target files

Since it is possbile to have same number in tfd field (say file added, closed, then nother file dup'ed to same number and added back) it is imposible to distinguish such target files solely by their numbers.

Strictly speaking regular applications don't need to recognize these targets at all but for checkpoint/restore sake we need to collect targets to be able to push them back on restore stage in a proper order.

Thus lets add file position, inode and device number where this target lays. This three fields can be used as a primary key for sorting, and together with kcmp help CRIU can find out an exact file target (from the whole set of processes being checkpointed).

Signed-off-by Cyrill Gorcunov <[email protected]>

CC Al Viro <viro-RmSDqhL/[email protected]>

CC Andrew Morton <[email protected]>

CC Andrey Vagin <[email protected]>

CC Pavel Emelyanov <[email protected]>

CC Michael Kerrisk <[email protected]>

CC Kir Kolyshkin <[email protected]>

CC Jason Baron <[email protected]>

CC Andy Lutomirski <[email protected]>

---
 Documentation/filesystems/proc.txt |    6 +++++-
 fs/eventpoll.c                     |    8 ++++++--
 2 files changed, 11 insertions(+), 3 deletions(-)

Index: linux-ml.git/Documentation/filesystems/proc.txt
===================================================================

Documentation/filesystems/proc.txt +4 -0

 	where 'tfd' is a target file descriptor number in decimal form,
 	'events' is events mask being watched and the 'data' is data
 	associated with a target [see epoll(7) for more details].
 
+	The 'pos' is current offset of the target file in decimal form
+	[see lseek(2)], 'ino' and 'sdev' are inode and device numbers
+	where target file resides, all in hex format.
+
 	Fsnotify files
 	~~~~~~~~~~~~~~
 	For inotify files the format is the following

Index: linux-ml.git/fs/eventpoll.c
===================================================================

fs/eventpoll.c +6 -2

@@ -883,10 +883,14 @@ static void ep_show_fdinfo(struct seq_fi
 	mutex_lock(&ep->mtx);
 	for (rbp = rb_first(&ep->rbr); rbp; rbp = rb_next(rbp)) {
 		struct epitem *epi = rb_entry(rbp, struct epitem, rbn);
+		struct inode *inode = file_inode(epi->ffd.file);
 
-		seq_printf(m, "tfd: %8d events: %8x data: %16llx\n",
+		seq_printf(m, "tfd: %8d events: %8x data: %16llx "
+			   " pos:%lli ino:%lx sdev:%x\n",
 			   epi->ffd.fd, epi->event.events,
-			   (long long)epi->event.data);
+			   (long long)epi->event.data,
+			   (long long)epi->ffd.file->f_pos,
+			   inode->i_ino, inode->i_sb->s_dev);
 		if (seq_has_overflowed(m))
 			break;
 	}

Quoted 30 lines · Cyrill Gorcunov

Since it is possbile to have same number in tfd field (say file added, closed, then nother file dup'ed to same number and added back) it is imposible to distinguish such target files solely by their numbers.

Strictly speaking regular applications don't need to recognize these targets at all but for checkpoint/restore sake we need to collect targets to be able to push them back on restore stage in a proper order.

Thus lets add file position, inode and device number where this target lays. This three fields can be used as a primary key for sorting, and together with kcmp help CRIU can find out an exact file target (from the whole set of processes being checkpointed).

Signed-off-by Cyrill Gorcunov <[email protected]>

CC Al Viro <[email protected]>

CC Andrew Morton <[email protected]>

CC Andrey Vagin <[email protected]>

CC Pavel Emelyanov <[email protected]>

CC Michael Kerrisk <[email protected]>

CC Kir Kolyshkin <[email protected]>

CC Jason Baron <[email protected]>

CC Andy Lutomirski <[email protected]>

---
 Documentation/filesystems/proc.txt |    6 +++++-
 fs/eventpoll.c                     |    8 ++++++--
 2 files changed, 11 insertions(+), 3 deletions(-)

Documentation/filesystems/proc.txt pair provide additional information part · line 1782

Quoted 10 lines

Index: linux-ml.git/Documentation/filesystems/proc.txt
===================================================================

Documentation/filesystems/proc.txt +1 -1

@@ -1779,12 +1779,16 @@ pair provide additional information part
 	pos:	0
 	flags:	02
 	mnt_id:	9
-	tfd:        5 events:       1d data: ffffffffffffffff
+	tfd:        5 events:       1d data: ffffffffffffffff pos:0 ino:61af sdev:7

I think it may be better to print mnt_id instead of sdev, because there may be two file descriptors opened from different bind mounts.