Skip to contentSkip to conversation

spi: spi-sprd-adi: Fix double free in probe error path

PATCH In review Felix Gu 3 participants 5 messages Jan 9 – Jan 13

Subsystems (3)
lore
Share & tools
lore

Felix Gu

author
#0 lore raw

PATCH v2 1/1 spi: spi-sprd-adi: Fix double free in probe error path

The driver currently uses spi_alloc_host() to allocate the controller but registers it using devm_spi_register_controller().

If devm_register_restart_handler() fails, the code jumps to the put_ctlr label and calls spi_controller_put(). However, since the controller was registered via a devm function, the device core will automatically call spi_controller_put() again when the probe fails. This results in a double-free of the spi_controller structure.

Fix this by switching to devm_spi_alloc_host() and removing the manual spi_controller_put() call.

Fixes ac17750 ("spi: sprd: Add the support of restarting the system")

---
Changes in v2:
- Added Fixes: tag.
- Added Baolin's Reviewed-by tag.
- Link to v1: https://lore.kernel.org/lkml/[email protected]/

---
 drivers/spi/spi-sprd-adi.c | 33 ++++++++++-----------------------
 1 file changed, 10 insertions(+), 23 deletions(-)
drivers/spi/spi-sprd-adi.c +10 -23
index 262c11d977ea..f25b34a91756 100644
@@ -528,7 +528,7 @@ static int sprd_adi_probe(struct platform_device *pdev)
 	pdev->id = of_alias_get_id(np, "spi");
 	num_chipselect = of_get_child_count(np);
 
-	ctlr = spi_alloc_host(&pdev->dev, sizeof(struct sprd_adi));
+	ctlr = devm_spi_alloc_host(&pdev->dev, sizeof(struct sprd_adi));
 	if (!ctlr)
 		return -ENOMEM;
 
@@ -536,10 +536,8 @@ static int sprd_adi_probe(struct platform_device *pdev)
 	sadi = spi_controller_get_devdata(ctlr);
 
 	sadi->base = devm_platform_get_and_ioremap_resource(pdev, 0, &res);
-	if (IS_ERR(sadi->base)) {
-		ret = PTR_ERR(sadi->base);
-		goto put_ctlr;
-	}
+	if (IS_ERR(sadi->base))
+		return PTR_ERR(sadi->base);
 
 	sadi->slave_vbase = (unsigned long)sadi->base +
 			    data->slave_offset;
@@ -551,18 +549,15 @@ static int sprd_adi_probe(struct platform_device *pdev)
 	if (ret > 0 || (IS_ENABLED(CONFIG_HWSPINLOCK) && ret == 0)) {
 		sadi->hwlock =
 			devm_hwspin_lock_request_specific(&pdev->dev, ret);
-		if (!sadi->hwlock) {
-			ret = -ENXIO;
-			goto put_ctlr;
-		}
+		if (!sadi->hwlock)
+			return -ENXIO;
 	} else {
 		switch (ret) {
 		case -ENOENT:
 			dev_info(&pdev->dev, "no hardware spinlock supplied\n");
 			break;
 		default:
-			dev_err_probe(&pdev->dev, ret, "failed to find hwlock id\n");
-			goto put_ctlr;
+			return dev_err_probe(&pdev->dev, ret, "failed to find hwlock id\n");
 		}
 	}
 
@@ -579,26 +574,18 @@ static int sprd_adi_probe(struct platform_device *pdev)
 	ctlr->transfer_one = sprd_adi_transfer_one;
 
 	ret = devm_spi_register_controller(&pdev->dev, ctlr);
-	if (ret) {
-		dev_err(&pdev->dev, "failed to register SPI controller\n");
-		goto put_ctlr;
-	}
+	if (ret)
+		return dev_err_probe(&pdev->dev, ret, "failed to register SPI controller\n");
 
 	if (sadi->data->restart) {
 		ret = devm_register_restart_handler(&pdev->dev,
 						    sadi->data->restart,
 						    sadi);
-		if (ret) {
-			dev_err(&pdev->dev, "can not register restart handler\n");
-			goto put_ctlr;
-		}
+		if (ret)
+			return dev_err_probe(&pdev->dev, ret, "can not register restart handler\n");
 	}
 
 	return 0;
-
-put_ctlr:
-	spi_controller_put(ctlr);
-	return ret;
 }
 
 static struct sprd_adi_data sc9860_data = {

---
base-commit: fc4e91c639c0af93d63c3d5bc0ee45515dd7504a
change-id: 20260108-spi-sprd-adi-fix-c071bf124bf6

Best regards,

Signature · 1 line
Felix Gu <[email protected]>
maintainer
#2 lore raw
In reply to Felix Gu
Quoted 10 lines · Felix Gu

The driver currently uses spi_alloc_host() to allocate the controller but registers it using devm_spi_register_controller().

If devm_register_restart_handler() fails, the code jumps to the put_ctlr label and calls spi_controller_put(). However, since the controller was registered via a devm function, the device core will automatically call spi_controller_put() again when the probe fails. This results in a double-free of the spi_controller structure.

[...]

Applied to

https://git.kernel.org/pub/scm/linux/kernel/git/broonie/spi.git for-next

Thanks!

[1/1] spi: spi-sprd-adi: Fix double free in probe error path

      commit: 383d4f5cffcc8df930d95b06518a9d25a6d74aac

All being well this means that it will be integrated into the linux-next tree (usually sometime in the next 24 hours) and sent to Linus during the next merge window (or sooner if it is a bug fix), however if problems are discovered then the patch may be dropped or reverted.

You may get further e-mails resulting from automated or manual testing and review of the tree, please engage with people reporting problems and send followup patches addressing any issues that are reported if needed.

If any updates are required or you are submitting further changes they should be sent as incremental updates against current git, existing patches will not be replaced.

Please add any relevant lists and maintainers to the CCs when replying to this mail.

Thanks,
Mark

#3 lore raw
Quoted 2 lines

[1/1] spi: spi-sprd-adi: Fix double free in probe error path

      commit: 383d4f5cffcc8df930d95b06518a9d25a6d74aac

https://git.kernel.org/pub/scm/linux/kernel/git/broonie/spi.git/commit/?id=383d4f5cffcc8df930d95b06518a9d25a6d74aac

Quoted 4 lines

All being well this means that it will be integrated into the linux-next tree (usually sometime in the next 24 hours) and sent to Linus during the next merge window (or sooner if it is a bug fix), however if problems are discovered then the patch may be dropped or reverted.

See also once more:
Check “BAD_FIXES_TAG”
https://elixir.bootlin.com/linux/v6.19-rc4/source/scripts/checkpatch.pl#L3231-L3269

Regards,
Markus