[PATCH 4.9 095/119] powerpc/kernel: Fix FP and vector register restoration

[Greg Kroah-Hartman]

4.9-stable review patch.  If anyone has any objections, please let me know.

------------------

From: Breno Leitao <leitao@xxxxxxxxxx>

commit 1195892c091a15cc862f4e202482a36adc924e12 upstream.

Currently tsk->thread->load_vec and load_fp are not initialized during
task creation, which can lead to garbage values in these variables (non-zero
values).

These variables will be checked later in restore_math() to validate if the
FP and vector registers are being utilized. Since these values might be
non-zero, the restore_math() will continue to save the FP and vectors even if
they were never utilized by the userspace application. load_fp and load_vec
counters will then overflow (they wrap at 255) and the FP and Altivec will be
finally disabled, but before that condition is reached (counter overflow)
several context switches will have restored FP and vector registers without
need, causing a performance degradation.

Fixes: 70fe3d980f5f ("powerpc: Restore FPU/VEC/VSX if previously used")
Signed-off-by: Breno Leitao <leitao@xxxxxxxxxx>
Signed-off-by: Gustavo Romero <gusbromero@xxxxxxxxx>
Acked-by: Anton Blanchard <anton@xxxxxxxxx>
Signed-off-by: Michael Ellerman <mpe@xxxxxxxxxxxxxx>
Signed-off-by: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx>

---
 arch/powerpc/kernel/process.c |    2 ++
 1 file changed, 2 insertions(+)

--- a/arch/powerpc/kernel/process.c
+++ b/arch/powerpc/kernel/process.c
@@ -1659,6 +1659,7 @@ void start_thread(struct pt_regs *regs,
 #ifdef CONFIG_VSX
 	current->thread.used_vsr = 0;
 #endif
+	current->thread.load_fp = 0;
 	memset(&current->thread.fp_state, 0, sizeof(current->thread.fp_state));
 	current->thread.fp_save_area = NULL;
 #ifdef CONFIG_ALTIVEC
@@ -1667,6 +1668,7 @@ void start_thread(struct pt_regs *regs,
 	current->thread.vr_save_area = NULL;
 	current->thread.vrsave = 0;
 	current->thread.used_vr = 0;
+	current->thread.load_vec = 0;
 #endif /* CONFIG_ALTIVEC */
 #ifdef CONFIG_SPE
 	memset(current->thread.evr, 0, sizeof(current->thread.evr));