Dunno if this is a missmerge but
# fgrep cbe_spu_info arch/powerpc/platforms/cell/ -R | grep -i ini
cell/spufs/sched.c:              mutex_init(&cbe_spu_info[i].list_mutex);
cell/spufs/sched.c:              INIT_LIST_HEAD(&cbe_spu_info[i].spus);
cell/spu_base.c:         mutex_init(&cbe_spu_info[i].list_mutex);
cell/spu_base.c:         INIT_LIST_HEAD(&cbe_spu_info[i].spus);

the result is that no SPU context can be started because no SPE is available
and never will be. I guess this should be only be initialized in spu_base 
since it probes for SPEs and is run before spufs.

Cc: Geoff Levand <geoffrey.levand@am.sony.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Sebastian Siewior <sebastian@breakpoint.cc>

--- a/arch/powerpc/platforms/cell/spufs/sched.c
+++ b/arch/powerpc/platforms/cell/spufs/sched.c
@@ -927,10 +927,6 @@ int __init spu_sched_init(void)
 		INIT_LIST_HEAD(&spu_prio->runq[i]);
 		__clear_bit(i, spu_prio->bitmap);
 	}
-	for (i = 0; i < MAX_NUMNODES; i++) {
-		mutex_init(&cbe_spu_info[i].list_mutex);
-		INIT_LIST_HEAD(&cbe_spu_info[i].spus);
-	}
 	spin_lock_init(&spu_prio->runq_lock);
 
 	setup_timer(&spusched_timer, spusched_wake, 0);

--