gfapi: backport handles to 3.4 for nfs-ganeshav3.4.2qa2

nfs-ganesha-2.0 will ship shortly. It depends on the gfapi handle support,
which won't otherwise be released until 3.5 sometime in 2014.

Change-Id: I104d6fb275bb2c710790340fdc7d998446403026
Signed-off-by: Kaleb S. KEITHLEY <kkeithle@redhat.com>
Reviewed-on: http://review.gluster.org/6394
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Shyamsundar Ranganathan <srangana@redhat.com>
Reviewed-by: Vijay Bellur <vbellur@redhat.com>

1 files changed, 26 insertions, 0 deletions

diff --git a/api/src/glfs.h b/api/src/glfs.h
index 460edf1d0..bf34ec6bf 100644
--- a/api/src/glfs.h
+++ b/api/src/glfs.h
@@ -269,6 +269,32 @@ int glfs_fini (glfs_t *fs);
 struct glfs_fd;
 typedef struct glfs_fd glfs_fd_t;
 
+/*
+ * PER THREAD IDENTITY MODIFIERS
+ *
+ * The following operations enable to set a per thread identity context
+ * for the glfs APIs to perform operations as. The calls here are kept as close
+ * to POSIX equivalents as possible.
+ *
+ * NOTES:
+ *
+ *  - setgroups is a per thread setting, hence this is named as fsgroups to be
+ *    close in naming to the fs(u/g)id APIs
+ *  - Typical mode of operation is to set the IDs as required, with the
+ *    supplementary groups being optionally set, make the glfs call and post the
+ *    glfs operation set them back to eu/gid or uid/gid as appropriate to the
+ *    caller
+ *  - The groups once set, need to be unset by setting the size to 0 (in which
+ *    case the list argument is a do not care)
+ *  - Once a process for a thread of operation choses to set the IDs, all glfs
+ *    calls made from that thread would default to the IDs set for the thread.
+ *    As a result use these APIs with care and ensure that the set IDs are
+ *    reverted to global process defaults as required.
+ *
+ */
+int glfs_setfsuid (uid_t fsuid);
+int glfs_setfsgid (gid_t fsgid);
+int glfs_setfsgroups (size_t size, const gid_t *list);
 
 /*
   SYNOPSIS