logging: fix mtrace backend activation on IPC log enable

This patch addresses an issue where logs were not being collected after
an IPC message when the mtrace autostart was disabled in regular builds.

The problem was resolved by replacing `log_backend_activate` and
`log_backend_deactivate` with `log_backend_enable` and
`log_backend_disable`, respectively. These functions ensure that the
backend is properly enabled with the initial maximum filtering level and
disabled when necessary. This change ensures that logging behavior
aligns with the intended configuration, allowing logs to be collected as
expected after an IPC message.

Signed-off-by: Tomasz Leman <tomasz.m.leman@intel.com>

Author: tmleman

src/ipc/ipc4/logging.c

@@ -15,8 +15,8 @@
 #include <ipc4/error_status.h>
 #include <ipc4/logging.h>
 #if !CONFIG_LIBRARY
-#include <zephyr/logging/log_backend.h>
 #include <zephyr/logging/log.h>
+#include <zephyr/logging/log_ctrl.h>
 #endif
 #if CONFIG_LOG_BACKEND_SOF_PROBE
 #include <sof/probe/probe.h>
@@ -135,8 +135,7 @@ int ipc4_logging_enable_logs(bool first_block,
 
 	if (log_state->enable) {
 		adsp_mtrace_log_init(mtrace_log_hook);
-
-		log_backend_activate(log_backend, mtrace_log_hook);
+		log_backend_enable(log_backend, mtrace_log_hook, CONFIG_SOF_LOG_LEVEL);
 
 		mtrace_aging_timer = log_state->aging_timer_period;
 		if (mtrace_aging_timer < IPC4_MTRACE_AGING_TIMER_MIN_MS) {
@@ -150,7 +149,7 @@ int ipc4_logging_enable_logs(bool first_block,
 		schedule_task(&mtrace_task, mtrace_aging_timer * 1000, 0);
 	} else  {
 		adsp_mtrace_log_init(NULL);
-		log_backend_deactivate(log_backend);
+		log_backend_disable(log_backend);
 		schedule_task_cancel(&mtrace_task);
 	}