Al Viro adbecb128c [PATCH] dup_fd() part 4 - race fix
Parent _can_ be a clone task, contrary to the comment.  Moreover,
more files could be opened while we allocate a copy, in which case
we end up copying only part into new descriptor table.  Since what
we get _is_ affected by all changes in the old range, we can get
rather weird effects - e.g.
	dup2(0, 1024); close(0);
in parallel with fork() resulting in child that sees the effect of
close(), but not that of dup2() done just before that close().

What we need is to recalculate the open_count after having reacquired
->file_lock and if external fdtable we'd just allocated is too small for
it, free the sucker and redo allocation.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2008-05-16 17:22:46 -04:00
..
2008-05-14 19:23:27 -05:00
2008-02-08 09:22:40 -08:00
2008-05-01 08:04:01 -07:00
2008-04-29 08:06:04 -07:00
2008-04-29 08:06:00 -07:00
2008-05-13 08:02:26 -07:00
2008-04-30 08:29:52 -07:00
2008-02-08 09:22:40 -08:00
2008-05-13 08:02:21 -07:00
2008-04-29 08:06:18 -07:00
2007-10-18 14:37:31 -07:00
2008-04-30 08:29:51 -07:00
2008-04-30 08:29:52 -07:00
2008-04-29 08:06:17 -07:00
2008-05-07 18:35:03 +02:00
2008-01-30 13:31:46 +01:00
2008-04-30 08:29:47 -07:00
2008-05-01 13:08:16 -04:00
2008-05-01 13:08:16 -04:00
2008-05-01 13:08:50 -04:00
2008-05-01 13:08:16 -04:00
2008-05-01 13:08:16 -04:00
2008-05-16 17:22:46 -04:00
2008-05-06 13:13:37 -07:00
2008-02-06 10:41:07 -08:00
2008-04-29 08:06:00 -07:00
2008-04-30 13:38:47 +02:00
2008-04-29 08:06:01 -07:00
2008-01-30 13:31:46 +01:00
2008-04-29 08:06:09 -07:00
2008-02-14 21:13:33 -08:00
2008-05-01 13:08:16 -04:00
2008-05-08 10:46:56 -07:00
2008-04-30 08:29:51 -07:00
2008-05-01 13:08:16 -04:00
2008-05-01 13:08:50 -04:00
2008-05-08 14:06:19 +02:00
2008-02-14 21:13:33 -08:00
2008-04-29 08:06:00 -07:00
2008-05-01 13:08:50 -04:00