[PATCH v4 21/30] context_tracking: Explicitely use CT_STATE_KERNEL where it is missing

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



CT_STATE_KERNEL being zero means it can be (and is) omitted in a handful of
places. A later patch will change CT_STATE_KERNEL into a non-zero value,
prepare that by using it where it should be:

o In the initial CT state
o At kernel entry / exit

No change in functionality intended.

Signed-off-by: Valentin Schneider <vschneid@xxxxxxxxxx>
---
 kernel/context_tracking.c | 10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)

diff --git a/kernel/context_tracking.c b/kernel/context_tracking.c
index 938c48952d265..a61498a8425e2 100644
--- a/kernel/context_tracking.c
+++ b/kernel/context_tracking.c
@@ -31,7 +31,7 @@ DEFINE_PER_CPU(struct context_tracking, context_tracking) = {
 	.nesting = 1,
 	.nmi_nesting = CT_NESTING_IRQ_NONIDLE,
 #endif
-	.state = ATOMIC_INIT(CT_RCU_WATCHING),
+	.state = ATOMIC_INIT(CT_RCU_WATCHING | CT_STATE_KERNEL),
 };
 EXPORT_SYMBOL_GPL(context_tracking);
 
@@ -147,7 +147,7 @@ static void noinstr ct_kernel_exit(bool user, int offset)
 	instrumentation_end();
 	WRITE_ONCE(ct->nesting, 0); /* Avoid irq-access tearing. */
 	// RCU is watching here ...
-	ct_kernel_exit_state(offset);
+	ct_kernel_exit_state(offset - CT_STATE_KERNEL);
 	// ... but is no longer watching here.
 	rcu_task_exit();
 }
@@ -175,7 +175,7 @@ static void noinstr ct_kernel_enter(bool user, int offset)
 	}
 	rcu_task_enter();
 	// RCU is not watching here ...
-	ct_kernel_enter_state(offset);
+	ct_kernel_enter_state(offset + CT_STATE_KERNEL);
 	// ... but is watching here.
 	instrumentation_begin();
 
@@ -537,7 +537,7 @@ void noinstr __ct_user_enter(enum ctx_state state)
 				 * RCU only requires CT_RCU_WATCHING increments to be fully
 				 * ordered.
 				 */
-				raw_atomic_add(state, &ct->state);
+				raw_atomic_add(state - CT_STATE_KERNEL, &ct->state);
 			}
 		}
 	}
@@ -647,7 +647,7 @@ void noinstr __ct_user_exit(enum ctx_state state)
 				 * RCU only requires CT_RCU_WATCHING increments to be fully
 				 * ordered.
 				 */
-				raw_atomic_sub(state, &ct->state);
+				raw_atomic_sub(state - CT_STATE_KERNEL, &ct->state);
 			}
 		}
 	}
-- 
2.43.0





[Index of Archives]     [Linux Samsung SoC]     [Linux Rockchip SoC]     [Linux Actions SoC]     [Linux for Synopsys ARC Processors]     [Linux NFS]     [Linux NILFS]     [Linux USB Devel]     [Video for Linux]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]


  Powered by Linux