Re: [PATCH v3 3/3] remoteproc: xilinx: add mailbox channels for rpmsg

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

 




On 2/22/23 11:06 AM, Mathieu Poirier wrote:
On Mon, Feb 13, 2023 at 01:18:26PM -0800, Tanmay Shah wrote:
[ ... ]
+
  /*
   * zynqmp_r5_set_mode()
   *
@@ -617,7 +819,7 @@ static int zynqmp_r5_rproc_unprepare(struct rproc *rproc)
  	return 0;
  }
-static const struct rproc_ops zynqmp_r5_rproc_ops = {
+static struct rproc_ops zynqmp_r5_rproc_ops = {
  	.prepare	= zynqmp_r5_rproc_prepare,
  	.unprepare	= zynqmp_r5_rproc_unprepare,
  	.start		= zynqmp_r5_rproc_start,
@@ -642,6 +844,7 @@ static struct zynqmp_r5_core *zynqmp_r5_add_rproc_core(struct device *cdev)
  {
  	struct zynqmp_r5_core *r5_core;
  	struct rproc *r5_rproc;
+	struct mbox_info *ipi;
  	int ret;
/* Set up DMA mask */
@@ -649,12 +852,23 @@ static struct zynqmp_r5_core *zynqmp_r5_add_rproc_core(struct device *cdev)
  	if (ret)
  		return ERR_PTR(ret);
+ /*
+	 * If mailbox nodes are disabled using "status" property then setting up
+	 * mailbox channels will be failed. In that case we don't really need
+	 * kick() operation. Include .kick() only if mbox channels are acquired
+	 * successfully.
+	 */
+	ipi = zynqmp_r5_setup_mbox(cdev);
+	if (ipi)
+		zynqmp_r5_rproc_ops.kick = zynqmp_r5_rproc_kick;
+
  	/* Allocate remoteproc instance */
  	r5_rproc = rproc_alloc(cdev, dev_name(cdev),
  			       &zynqmp_r5_rproc_ops,
  			       NULL, sizeof(struct zynqmp_r5_core));
  	if (!r5_rproc) {
  		dev_err(cdev, "failed to allocate memory for rproc instance\n");
+		zynqmp_r5_free_mbox(ipi);
  		return ERR_PTR(-ENOMEM);
  	}
@@ -665,6 +879,7 @@ static struct zynqmp_r5_core *zynqmp_r5_add_rproc_core(struct device *cdev)
  	if (!r5_core->np) {
  		dev_err(cdev, "can't get device node for r5 core\n");
  		ret = -EINVAL;
+		zynqmp_r5_free_mbox(ipi);
  		goto free_rproc;
  	}
@@ -672,10 +887,17 @@ static struct zynqmp_r5_core *zynqmp_r5_add_rproc_core(struct device *cdev)
  	ret = rproc_add(r5_rproc);
  	if (ret) {
  		dev_err(cdev, "failed to add r5 remoteproc\n");
+		zynqmp_r5_free_mbox(ipi);
  		goto free_rproc;
  	}
+ if (ipi) {
+		r5_core->ipi = ipi;
+		ipi->r5_core = r5_core;
+	}
+
  	r5_core->rproc = r5_rproc;
+
  	return r5_core;
free_rproc:
@@ -918,6 +1140,7 @@ static int zynqmp_r5_cluster_init(struct zynqmp_r5_cluster *cluster)
  	while (i >= 0) {
  		put_device(child_devs[i]);
  		if (r5_cores[i]) {
+			zynqmp_r5_free_mbox(r5_cores[i]->ipi);
The mailboxes are initialized in zynqmp_r5_add_rproc_core() but free'd here in
case of trouble, which introduces coupling between the two functions.  I suggest
moving zynqmp_r5_setup_mbox() in zynqmp_r5_cluster_init() and initialize both
mailboxes in it.

I am done reviewing this set.


Ack. Yes that makes sense.

Thanks,

Tanmay



Thanks,
Mathieu

Thanks,
Mathieu

  			of_reserved_mem_device_release(r5_cores[i]->dev);
  			rproc_del(r5_cores[i]->rproc);
  			rproc_free(r5_cores[i]->rproc);
@@ -942,6 +1165,7 @@ static void zynqmp_r5_cluster_exit(void *data)
for (i = 0; i < cluster->core_count; i++) {
  		r5_core = cluster->r5_cores[i];
+		zynqmp_r5_free_mbox(r5_core->ipi);
  		of_reserved_mem_device_release(r5_core->dev);
  		put_device(r5_core->dev);
  		rproc_del(r5_core->rproc);
--
2.25.1




[Index of Archives]     [Linux Sound]     [ALSA Users]     [ALSA Devel]     [Linux Audio Users]     [Linux Media]     [Kernel]     [Photo Sharing]     [Gimp]     [Yosemite News]     [Linux Media]

  Powered by Linux