[PATCH 68/81] drm/amd/display: Preserve refcount for S3 case.

sunpeng.li at amd.com sunpeng.li at amd.com
Tue Jul 25 13:54:28 UTC 2017


From: Andrey Grodzovsky <Andrey.Grodzovsky at amd.com>

Curent_context is zerroed out for suspend, keep the refcount.
Minor code move in dc_commit_context_no_check

Change-Id: I45b7fa4b0a7df54f1621f6306ef9e31c9ebe25eb
Signed-off-by: Andrey Grodzovsky <Andrey.Grodzovsky at amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng at amd.com>
Acked-by: Harry Wentland <Harry.Wentland at amd.com>
---
 drivers/gpu/drm/amd/display/dc/core/dc.c | 12 ++++++++----
 1 file changed, 8 insertions(+), 4 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/core/dc.c b/drivers/gpu/drm/amd/display/dc/core/dc.c
index 8e580ac..20f4199 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc.c
@@ -977,10 +977,10 @@ static bool dc_commit_context_no_check(struct dc *dc, struct validate_context *c
 
 	dc_release_validate_context(core_dc->current_context);
 
-	dc_retain_validate_context(context);
-
 	core_dc->current_context = context;
 
+	dc_retain_validate_context(core_dc->current_context);
+
 	return (result == DC_OK);
 }
 
@@ -1065,8 +1065,6 @@ bool dc_commit_streams(
 
 	result = dc_commit_context_no_check(dc, context);
 
-	return (result == DC_OK);
-
 fail:
 	dc_release_validate_context(context);
 
@@ -1736,6 +1734,7 @@ void dc_set_power_state(
 	enum dc_acpi_cm_power_state power_state)
 {
 	struct core_dc *core_dc = DC_TO_CORE(dc);
+	int ref_count;
 
 	switch (power_state) {
 	case DC_ACPI_CM_POWER_STATE_D0:
@@ -1749,8 +1748,13 @@ void dc_set_power_state(
 		 * clean state, and dc hw programming optimizations will not
 		 * cause any trouble.
 		 */
+
+		/* Preserve refcount */
+		ref_count = core_dc->current_context->ref_count;
+		dc_resource_validate_ctx_destruct(core_dc->current_context);
 		memset(core_dc->current_context, 0,
 				sizeof(*core_dc->current_context));
+		core_dc->current_context->ref_count = ref_count;
 
 		break;
 	}
-- 
2.7.4



More information about the amd-gfx mailing list