Thanks for reporting this bug. The same bug exists in the driver for OF
(xillybus_of.c).
As for the choice of -EIO, I've surveyed several drivers calling
of_iomap() and pci_iomap(), and it turns out that some fail on -ENOMEM,
some on -EIO and some on -EBUSY.
Personally, I'm inclined towards -ENOMEM, since the problem is a memory
mapping failure.
I'll submit a patch fixing both files soon.
Thanks again,
Eli
On 16/12/13 12:06, Jingoo Han wrote:
On Monday, December 16, 2013 2:51 PM, Wei Yongjun wrote:
From: Wei Yongjun<yongjun_wei@xxxxxxxxxxxxxxxxx>
Fix to return negative error code -EIO from the error handling
case instead of 0.
Signed-off-by: Wei Yongjun<yongjun_wei@xxxxxxxxxxxxxxxxx>
---
drivers/staging/xillybus/xillybus_pcie.c | 2 +-
It looks good.
Reviewed-by: Jingoo Han<jg1.han@xxxxxxxxxxx>
Best regards,
Jingoo Han
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/staging/xillybus/xillybus_pcie.c b/drivers/staging/xillybus/xillybus_pcie.c
index 0564f97..51426d8 100644
--- a/drivers/staging/xillybus/xillybus_pcie.c
+++ b/drivers/staging/xillybus/xillybus_pcie.c
@@ -168,9 +168,9 @@ static int xilly_probe(struct pci_dev *pdev,
}
endpoint->registers = pci_iomap(pdev, 0, 128);
-
if (!endpoint->registers) {
dev_err(endpoint->dev, "Failed to map BAR 0. Aborting.\n");
+ rc = -EIO;
goto failed_iomap0;
}
_______________________________________________
devel mailing list
devel@xxxxxxxxxxxxxxxxxxxxxx
http://driverdev.linuxdriverproject.org/mailman/listinfo/driverdev-devel
