Retry NOP smc call when a guest's VIRQ is pending

In virtualization use case, if a guest's VIRQ is pending at,
the hypervisor, the control returns back control immediately to the guest
without transitioning to TOS so that the guest's IRQ is handled
in time. In such a case, re-try the NOP smc with the same arguments
as the previous call as the SMC wouldn't have reached TOS yet.

JIRA: SEC-366

Change-Id: Iff0f4fd6aab1369848736a1a7f2fa75afd3616b9
Signed-off-by: Amith Nuggehalli <aramachan@nvidia.com>
(cherry picked from commit 585fea50e14410d312c9966f5520af305c56e989)
Reviewed-on: https://git-master.nvidia.com/r/1533177
GVS: Gerrit_Virtual_Submit
Reviewed-by: Varun Wadekar <vwadekar@nvidia.com>

1 files changed, 2 insertions, 0 deletions

diff --git a/include/linux/trusty/sm_err.h b/include/linux/trusty/sm_err.h
index 32ee08e49..fb6f3794d 100644
--- a/include/linux/trusty/sm_err.h
+++ b/include/linux/trusty/sm_err.h
@@ -1,5 +1,6 @@
 /*
  * Copyright (c) 2013 Google Inc. All rights reserved
+ * Copyright (c) 2017 NVIDIA CORPORATION.  All rights reserved.
  *
  * Permission is hereby granted, free of charge, to any person obtaining
  * a copy of this software and associated documentation files
@@ -39,5 +40,6 @@
 #define SM_ERR_CPU_IDLE                 -13     /* SMC call waiting for another CPU */
 #define SM_ERR_NOP_INTERRUPTED          -14     /* Got interrupted. Call back with new SMC_SC_NOP */
 #define SM_ERR_NOP_DONE                 -15     /* Cpu idle after SMC_SC_NOP (not an error) */
+#define SM_ERR_NOP_RETRY                -16     /* Retry the previous NOP smc as the first may not have reached TOS */
 
 #endif