Currently the first thing we do in cxl_probe is to grab a reference
on the pci device. Later on, we call device_register on our adapter,
which also holds the PCI device.
In our remove path, we call device_unregister, but we never call
pci_dev_put. We therefore leak the device every time we do a
reflash.
device_register/unregister is sufficient to hold the reference.
Drop the call to pci_dev_get.
Fixes: f204e0b8cedd ("cxl: Driver code for powernv PCIe based cards for
userspace access")
Cc: sta...@vger.kernel.org
Signed-off-by: Daniel Axtens <d...@axtens.net>
---
This is the cxl bug that caused me to catch this a few weeks back:
e642d11bdbfe ("powerpc/eeh: Probe after unbalanced kref check")
I put an printk in the unbalanced kref path and confirmed that it
was printed with the pci_dev_get in and went away with the
pci_dev_get out.
---
drivers/misc/cxl/pci.c | 2 --
1 file changed, 2 deletions(-)
diff --git a/drivers/misc/cxl/pci.c b/drivers/misc/cxl/pci.c
index 02c85160bfe9..a5e977192b61 100644
--- a/drivers/misc/cxl/pci.c
+++ b/drivers/misc/cxl/pci.c
@@ -1249,8 +1249,6 @@ static int cxl_probe(struct pci_dev *dev, const struct
pci_device_id *id)
int slice;
int rc;
- pci_dev_get(dev);
-
if (cxl_verbose)
dump_cxl_config_space(dev);
--
2.5.0
_______________________________________________
Linuxppc-dev mailing list
Linuxppc-dev@lists.ozlabs.org
https://lists.ozlabs.org/listinfo/linuxppc-dev
