FAILED: patch "[PATCH] spi: ti-qspi: fix controller deregistration" failed to apply to 5.15-stable tree

The patch below does not apply to the 5.15-stable tree.
If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <[email protected]>.

To reproduce the conflict and resubmit, you may use the following commands:

git fetch https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/ linux-5.15.y
git checkout FETCH_HEAD
git cherry-pick -x 0c18a1bacbb1d8b8aa34d3d004a2cb8226c8b1ea
# <resolve conflicts, build, test, etc.>
git commit -s
git send-email --to '<[email protected]>' --in-reply-to '2026051215-rebound-collie-a3aa@gregkh' --subject-prefix 'PATCH 5.15.y' HEAD^..

Possible dependencies:

thanks,

greg k-h

------------------ original commit in Linus's tree ------------------

From 0c18a1bacbb1d8b8aa34d3d004a2cb8226c8b1ea Mon Sep 17 00:00:00 2001
From: Johan Hovold <[email protected]>
Date: Fri, 10 Apr 2026 10:17:53 +0200
Subject: [PATCH] spi: ti-qspi: fix controller deregistration

Make sure to deregister the controller before disabling underlying resources like clocks during driver unbind.

Note that the controller is suspended before disabling and releasing resources since commit 3ac066e2227c ("spi: spi-ti-qspi: Suspend the queue before removing the device") which avoids issues like unclocked accesses but prevents SPI device drivers from doing I/O during deregistration.

index d1d880a8ed7d..1fbd710d616f 100644
@@ -888,7 +888,7 @@ static int ti_qspi_probe(struct platform_device *pdev)
 	qspi->mmap_enabled = false;
 	qspi->current_cs = -1;
 
-	ret = devm_spi_register_controller(&pdev->dev, host);
+	ret = spi_register_controller(host);
 	if (!ret)
 		return 0;
 

@@ -903,19 +903,17 @@ static int ti_qspi_probe(struct platform_device *pdev)
 static void ti_qspi_remove(struct platform_device *pdev)
 {
 	struct ti_qspi *qspi = platform_get_drvdata(pdev);
-	int rc;
 
-	rc = spi_controller_suspend(qspi->host);
-	if (rc) {
-		dev_alert(&pdev->dev, "spi_controller_suspend() failed (%pe)\n",
-			  ERR_PTR(rc));
-		return;
-	}
+	spi_controller_get(qspi->host);
+
+	spi_unregister_controller(qspi->host);
 
 	pm_runtime_put_sync(&pdev->dev);
 	pm_runtime_disable(&pdev->dev);
 
 	ti_qspi_dma_cleanup(qspi);
+
+	spi_controller_put(qspi->host);
 }
 
 static const struct dev_pm_ops ti_qspi_pm_ops = {

From: Uwe Kleine-König <[email protected]>

[ Upstream commit 2f2802d1a59d79a3d00cb429841db502c2bbc3df ]

The .remove() callback for a platform driver returns an int which makes many driver authors wrongly assume it's possible to do error handling by returning an error code. However the value returned is ignored (apart from emitting a warning) and this typically results in resource leaks.

To improve here there is a quest to make the remove callback return void. In the first step of this quest all drivers are converted to .remove_new(), which already returns void. Eventually after all drivers are converted, .remove_new() will be renamed to .remove().

Add an error message to the error path that returned an error before to replace the core's error message with more information. Apart from the different wording of the error message, this patch doesn't introduce a semantic difference.

Stable-dep-of: 0c18a1bacbb1 ("spi: ti-qspi: fix controller deregistration")

---
 drivers/spi/spi-ti-qspi.c | 13 +++++++------
 1 file changed, 7 insertions(+), 6 deletions(-)

index 081da1fd3fd7e..f2ccd264230f7 100644
@@ -906,21 +906,22 @@ static int ti_qspi_probe(struct platform_device *pdev)
 	return ret;
 }
 
-static int ti_qspi_remove(struct platform_device *pdev)
+static void ti_qspi_remove(struct platform_device *pdev)
 {
 	struct ti_qspi *qspi = platform_get_drvdata(pdev);
 	int rc;
 
 	rc = spi_master_suspend(qspi->master);
-	if (rc)
-		return rc;
+	if (rc) {
+		dev_alert(&pdev->dev, "spi_master_suspend() failed (%pe)\n",
+			  ERR_PTR(rc));
+		return;
+	}
 
 	pm_runtime_put_sync(&pdev->dev);
 	pm_runtime_disable(&pdev->dev);
 
 	ti_qspi_dma_cleanup(qspi);
-
-	return 0;
 }
 
 static const struct dev_pm_ops ti_qspi_pm_ops = {

@@ -929,7 +930,7 @@ static const struct dev_pm_ops ti_qspi_pm_ops = {
 
 static struct platform_driver ti_qspi_driver = {
 	.probe	= ti_qspi_probe,
-	.remove = ti_qspi_remove,
+	.remove_new = ti_qspi_remove,
 	.driver = {
 		.name	= "ti-qspi",
 		.pm =   &ti_qspi_pm_ops,

2.53.0

From: Johan Hovold <[email protected]>

[ Upstream commit 0c18a1bacbb1d8b8aa34d3d004a2cb8226c8b1ea ]

Make sure to deregister the controller before disabling underlying resources like clocks during driver unbind.

Note that the controller is suspended before disabling and releasing resources since commit 3ac066e2227c ("spi: spi-ti-qspi: Suspend the queue before removing the device") which avoids issues like unclocked accesses but prevents SPI device drivers from doing I/O during deregistration.

---
 drivers/spi/spi-ti-qspi.c | 14 ++++++--------
 1 file changed, 6 insertions(+), 8 deletions(-)

index f2ccd264230f7..d5a4fc70c45f8 100644
@@ -894,7 +894,7 @@ static int ti_qspi_probe(struct platform_device *pdev)
 	qspi->mmap_enabled = false;
 	qspi->current_cs = -1;
 
-	ret = devm_spi_register_master(&pdev->dev, master);
+	ret = spi_register_master(master);
 	if (!ret)
 		return 0;
 

@@ -909,19 +909,17 @@ static int ti_qspi_probe(struct platform_device *pdev)
 static void ti_qspi_remove(struct platform_device *pdev)
 {
 	struct ti_qspi *qspi = platform_get_drvdata(pdev);
-	int rc;
 
-	rc = spi_master_suspend(qspi->master);
-	if (rc) {
-		dev_alert(&pdev->dev, "spi_master_suspend() failed (%pe)\n",
-			  ERR_PTR(rc));
-		return;
-	}
+	spi_master_get(qspi->master);
+
+	spi_unregister_master(qspi->master);
 
 	pm_runtime_put_sync(&pdev->dev);
 	pm_runtime_disable(&pdev->dev);
 
 	ti_qspi_dma_cleanup(qspi);
+
+	spi_master_put(qspi->master);
 }
 
 static const struct dev_pm_ops ti_qspi_pm_ops = {

2.53.0