drm/amd/display: Revert changes for MPO underflow

[WHY]
The previous changes for fixing MPO underflow with multiple
display connected caused a regression where the machine runs
into a hang when doing multiple driver pnp with multiple displays
connected

[HOW]
Reverted offending change

Reviewed-by: Martin Leung <[e-mail]>
Acked-by: Wayne Lin <[e-mail]>
Signed-off-by: Angus Wang <[e-mail]>
Tested-by: Daniel Wheeler <[e-mail]>
Signed-off-by: Alex Deucher <[e-mail]>

drivers/gpu/drm/amd/display/dc/dcn20/dcn20_resource.c

···
 		.timing_trace = false,
 		.clock_trace = true,
 		.disable_pplib_clock_request = true,
-		.pipe_split_policy = MPC_SPLIT_DYNAMIC,
+		.pipe_split_policy = MPC_SPLIT_AVOID_MULT_DISP,
 		.force_single_disp_pipe_split = false,
 		.disable_dcc = DCC_ENABLE,
 		.vsr_support = true,

drivers/gpu/drm/amd/display/dc/dcn30/dcn30_resource.c

···
 	.timing_trace = false,
 	.clock_trace = true,
 	.disable_pplib_clock_request = true,
-	.pipe_split_policy = MPC_SPLIT_DYNAMIC,
+	.pipe_split_policy = MPC_SPLIT_AVOID_MULT_DISP,
 	.force_single_disp_pipe_split = false,
 	.disable_dcc = DCC_ENABLE,
 	.vsr_support = true,

drivers/gpu/drm/amd/display/dc/dcn302/dcn302_resource.c

···
 		.timing_trace = false,
 		.clock_trace = true,
 		.disable_pplib_clock_request = true,
-		.pipe_split_policy = MPC_SPLIT_DYNAMIC,
+		.pipe_split_policy = MPC_SPLIT_AVOID_MULT_DISP,
 		.force_single_disp_pipe_split = false,
 		.disable_dcc = DCC_ENABLE,
 		.vsr_support = true,

drivers/gpu/drm/amd/display/dc/dcn303/dcn303_resource.c

···
 		.timing_trace = false,
 		.clock_trace = true,
 		.disable_pplib_clock_request = true,
-		.pipe_split_policy = MPC_SPLIT_DYNAMIC,
+		.pipe_split_policy = MPC_SPLIT_AVOID_MULT_DISP,
 		.force_single_disp_pipe_split = false,
 		.disable_dcc = DCC_ENABLE,
 		.vsr_support = true,