x86, fpu: lazy allocation of FPU area - v5

Only allocate the FPU area when the application actually uses FPU, i.e., in the
first lazy FPU trap. This could save memory for non-fpu using apps.

for example: on my system after boot, there are around 300 processes, with
only 17 using FPU.

Signed-off-by: Suresh Siddha <suresh.b.siddha@intel.com>
Cc: Arjan van de Ven <arjan@linux.intel.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>

1 files changed, 19 insertions, 9 deletions

diff --git a/arch/x86/kernel/process.c b/arch/x86/kernel/process.c
index ead24efbcba0..0e613e7e7b5e 100644
--- a/arch/x86/kernel/process.c
+++ b/arch/x86/kernel/process.c
@@ -5,24 +5,34 @@
 #include <linux/slab.h>
 #include <linux/sched.h>
 
-static struct kmem_cache *task_xstate_cachep;
+struct kmem_cache *task_xstate_cachep;
 
 int arch_dup_task_struct(struct task_struct *dst, struct task_struct *src)
 {
         *dst = *src;
-        dst->thread.xstate = kmem_cache_alloc(task_xstate_cachep, GFP_KERNEL);
-        if (!dst->thread.xstate)
-                return -ENOMEM;
-        WARN_ON((unsigned long)dst->thread.xstate & 15);
-        memcpy(dst->thread.xstate, src->thread.xstate, xstate_size);
+        if (src->thread.xstate) {
+                dst->thread.xstate = kmem_cache_alloc(task_xstate_cachep,
+                                                      GFP_KERNEL);
+                if (!dst->thread.xstate)
+                        return -ENOMEM;
+                WARN_ON((unsigned long)dst->thread.xstate & 15);
+                memcpy(dst->thread.xstate, src->thread.xstate, xstate_size);
+        }
         return 0;
 }
 
-void free_thread_info(struct thread_info *ti)
+void free_thread_xstate(struct task_struct *tsk)
 {
-        kmem_cache_free(task_xstate_cachep, ti->task->thread.xstate);
-        ti->task->thread.xstate = NULL;
+        if (tsk->thread.xstate) {
+                kmem_cache_free(task_xstate_cachep, tsk->thread.xstate);
+                tsk->thread.xstate = NULL;
+        }
+}
+
 
+void free_thread_info(struct thread_info *ti)
+{
+        free_thread_xstate(ti->task);
         free_pages((unsigned long)(ti), get_order(THREAD_SIZE));
 }