Linux 4.2: total_link_count is no longer accessible

The value is now stored in the nameidata structure which
is private to fs/namei.c, so we can't modify it here.

The effect is that using a path that contains 40+ directories
may fail with ELOOP, depending on which directories in the
path were previously used.  After a directory is accessed once
its D_AUTOMOUNT flag is reset and it will no longer count
against the symlink limit in later path lookups.

Reviewed-on: http://gerrit.openafs.org/11926
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Jeffrey Altman <jaltman@your-file-system.com>
(cherry picked from commit 89aeb71a3e23c944f58cfa9572e9eae4d2130d37)

Change-Id: Ib6282f3029c4112c510217eacf270a56d679fccd
Reviewed-on: http://gerrit.openafs.org/11949
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Benjamin Kaduk <kaduk@mit.edu>
Reviewed-by: Stephan Wiesand <stephan.wiesand@desy.de>

diff --git a/acinclude.m4 b/acinclude.m4
index 0d1bb613fd111e44c17e2a41202e3c8a91a81d24..935fd00086f35ebac39891f1b3e13b5b1498914e 100644 (file)
--- a/acinclude.m4
+++ b/acinclude.m4
@@ -894,6 +894,7 @@ case $AFS_SYSNAME in *_linux* | *_umlinux*)
                 AC_CHECK_LINUX_STRUCT([task_struct], [sigmask_lock], [sched.h])
                 AC_CHECK_LINUX_STRUCT([task_struct], [tgid], [sched.h])
                 AC_CHECK_LINUX_STRUCT([task_struct], [thread_info], [sched.h])
+                AC_CHECK_LINUX_STRUCT([task_struct], [total_link_count], [sched.h])
                 LINUX_SCHED_STRUCT_TASK_STRUCT_HAS_SIGNAL_RLIM
 
                 dnl Check for typed structure elements

diff --git a/src/afs/LINUX/osi_vnodeops.c b/src/afs/LINUX/osi_vnodeops.c
index a8f00746a73c3cc5f0a310f09484ac67c3aa5ee4..a878eb42154370a87b4976ea5fa7b91f511491c3 100644 (file)
--- a/src/afs/LINUX/osi_vnodeops.c
+++ b/src/afs/LINUX/osi_vnodeops.c
@@ -1385,9 +1385,17 @@ afs_dentry_automount(afs_linux_path_t *path)
 {
     struct dentry *target;
 
-    /* avoid symlink resolution limits when resolving; we cannot contribute to
-     * an infinite symlink loop */
+    /* 
+     * Avoid symlink resolution limits when resolving; we cannot contribute to
+     * an infinite symlink loop.
+     *
+     * On newer kernels the field has moved to the private nameidata structure
+     * so we can't adjust it here.  This may cause ELOOP when using a path with
+     * 40 or more directories that are not already in the dentry cache.
+     */
+#if defined(STRUCT_TASK_STRUCT_HAS_TOTAL_LINK_COUNT)
     current->total_link_count--;
+#endif
 
     target = canonical_dentry(path->dentry->d_inode);