[PATCH net-next 0/3] ptp: ocp: minor updates

Fix errors spotted by automated tools.

Add myself to the MAINTAINERS for the ptp_ocp driver.

Jonathan Lemon (3):
  ptp: ocp: Fix uninitialized variable warning spotted by clang.
  ptp: ocp: Fix error path for pci_ocp_device_init()
  MAINTAINERS: Update for ptp_ocp driver.

 MAINTAINERS           |  6 ++++++
 drivers/ptp/ptp_ocp.c | 11 +++++------
 2 files changed, 11 insertions(+), 6 deletions(-)

-- 
2.31.1

If attempting to flash the firmware with a blob of size 0,
the entire write loop is skipped and the uninitialized err
is returned.  Fix by setting to 0 first.

Also remove a now-unused error handling statement.

Signed-off-by: Jonathan Lemon <redacted>
---
 drivers/ptp/ptp_ocp.c | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)

diff --git a/drivers/ptp/ptp_ocp.c b/drivers/ptp/ptp_ocp.c
index 92edf772feed..9b2ba06ebf97 100644
--- a/drivers/ptp/ptp_ocp.c
+++ b/drivers/ptp/ptp_ocp.c

@@ -763,7 +763,7 @@ ptp_ocp_devlink_flash(struct devlink *devlink, struct device *dev,
 	size_t off, len, resid, wrote;
 	struct erase_info erase;
 	size_t base, blksz;
-	int err;
+	int err = 0;
 
 	off = 0;
 	base = bp->flash_start;

@@ -847,8 +847,6 @@ ptp_ocp_devlink_info_get(struct devlink *devlink, struct devlink_info_req *req,
 							       "loader",
 							       buf);
 		}
-		if (err)
-			return err;
 	}
 
 	if (!bp->has_serial)

-- 
2.31.1

If ptp_ocp_device_init() fails, pci_disable_device() is skipped.
Fix the error handling so this case is covered.  Update ptp_ocp_remove()
so the normal exit path is identical.

Reported-by: Hulk Robot <redacted>
Signed-off-by: Jonathan Lemon <redacted>
---
 drivers/ptp/ptp_ocp.c | 7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)

diff --git a/drivers/ptp/ptp_ocp.c b/drivers/ptp/ptp_ocp.c
index 9b2ba06ebf97..a8d390314897 100644
--- a/drivers/ptp/ptp_ocp.c
+++ b/drivers/ptp/ptp_ocp.c

@@ -1436,7 +1436,7 @@ ptp_ocp_probe(struct pci_dev *pdev, const struct pci_device_id *id)
 	bp = devlink_priv(devlink);
 	err = ptp_ocp_device_init(bp, pdev);
 	if (err)
-		goto out_unregister;
+		goto out_disable;
 
 	/* compat mode.
 	 * Older FPGA firmware only returns 2 irq's.

@@ -1474,8 +1474,9 @@ ptp_ocp_probe(struct pci_dev *pdev, const struct pci_device_id *id)
 
 out:
 	ptp_ocp_detach(bp);
-	pci_disable_device(pdev);
 	pci_set_drvdata(pdev, NULL);
+out_disable:
+	pci_disable_device(pdev);
 out_unregister:
 	devlink_unregister(devlink);
 out_free:

@@ -1491,8 +1492,8 @@ ptp_ocp_remove(struct pci_dev *pdev)
 	struct devlink *devlink = priv_to_devlink(bp);
 
 	ptp_ocp_detach(bp);
-	pci_disable_device(pdev);
 	pci_set_drvdata(pdev, NULL);
+	pci_disable_device(pdev);
 
 	devlink_unregister(devlink);
 	devlink_free(devlink);

-- 
2.31.1

Add maintainer info for the OpenCompute PTP driver.

Signed-off-by: Jonathan Lemon <redacted>
---
 MAINTAINERS | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/MAINTAINERS b/MAINTAINERS
index 41fcfdb24a81..01632d3ca1f5 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS

@@ -13890,6 +13890,12 @@ F:	Documentation/devicetree/
 F:	arch/*/boot/dts/
 F:	include/dt-bindings/
 
+OPENCOMPUTE PTP CLOCK DRIVER
+M:	Jonathan Lemon <jonathan.lemon@gmail.com>
+L:	netdev@vger.kernel.org
+S:	Maintained
+F:	drivers/ptp/ptp_ocp.c
+
 OPENCORES I2C BUS DRIVER
 M:	Peter Korsgaard <peter@korsgaard.com>
 M:	Andrew Lunn <andrew@lunn.ch>

-- 
2.31.1

On Wed, 11 Aug 2021 11:31:31 -0700 Jonathan Lemon wrote:

Looks like an accidental change, but it's mentioned in the commit log?

 	}
 
 	if (!bp->has_serial)

On Wed, 11 Aug 2021 11:31:32 -0700 Jonathan Lemon wrote:

If ptp_ocp_device_init() fails, pci_disable_device() is skipped.
Fix the error handling so this case is covered.  Update ptp_ocp_remove()
so the normal exit path is identical.

Reported-by: Hulk Robot <redacted>
Signed-off-by: Jonathan Lemon <redacted>

Fixes tag would be useful on this and previous patch to make it clear
the problems are only present in net-next.

On Aug 12, 2021, at 2:58 PM, Jakub Kicinski [off-list ref] wrote:

On Wed, 11 Aug 2021 11:31:31 -0700 Jonathan Lemon wrote:

quoted

If attempting to flash the firmware with a blob of size 0,
the entire write loop is skipped and the uninitialized err
is returned.  Fix by setting to 0 first.

Also remove a now-unused error handling statement.

Signed-off-by: Jonathan Lemon <redacted>
---
drivers/ptp/ptp_ocp.c | 4 +---
1 file changed, 1 insertion(+), 3 deletions(-)

diff --git a/drivers/ptp/ptp_ocp.c b/drivers/ptp/ptp_ocp.c
index 92edf772feed..9b2ba06ebf97 100644
--- a/drivers/ptp/ptp_ocp.c
+++ b/drivers/ptp/ptp_ocp.c

@@ -763,7 +763,7 @@ ptp_ocp_devlink_flash(struct devlink *devlink, struct device *dev,
   size_t off, len, resid, wrote;
   struct erase_info erase;
   size_t base, blksz;
-    int err;
+    int err = 0;

   off = 0;
   base = bp->flash_start;

@@ -847,8 +847,6 @@ ptp_ocp_devlink_info_get(struct devlink *devlink, struct devlink_info_req *req,
                                  "loader",
                                  buf);
       }
-        if (err)
-            return err;

Looks like an accidental change, but it's mentioned in the commit log?

Yes, intentional, I spotted this NOP when fixing the compiler warning. 
— 
Jonathan