[dpdk-dev] [PATCH 1/1] ethdev: fix handling of close failure

If a failure happens when closing a port,
it was unnecessarily failing again in the function eth_err(),
because of a check against HW removal cause.
Indeed there is a big chance the port is released at this point.
Given the port is in the middle (or at the end) of a close process,
checking the error cause by accessing the port is a non-sense.
The error check is replaced by a simple return in the close function.

Bugzilla ID: 624
Fixes: 8a5a0aad5d3e ("ethdev: allow close function to return an error")
Cc: stable@dpdk.org

Reported-by: Anatoly Burakov <redacted>
Signed-off-by: Thomas Monjalon <redacted>
---
 lib/librte_ethdev/rte_ethdev.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/lib/librte_ethdev/rte_ethdev.c b/lib/librte_ethdev/rte_ethdev.c
index daf5f24f7e..89455a432e 100644
--- a/lib/librte_ethdev/rte_ethdev.c
+++ b/lib/librte_ethdev/rte_ethdev.c

@@ -1820,7 +1820,7 @@ rte_eth_dev_close(uint16_t port_id)
 	rte_ethdev_trace_close(port_id);
 	*lasterr = rte_eth_dev_release_port(dev);
 
-	return eth_err(port_id, firsterr);
+	return firsterr;
 }
 
 int

-- 
2.30.0

On 1/22/21 8:58 PM, Thomas Monjalon wrote:

If a failure happens when closing a port,
it was unnecessarily failing again in the function eth_err(),
because of a check against HW removal cause.
Indeed there is a big chance the port is released at this point.
Given the port is in the middle (or at the end) of a close process,
checking the error cause by accessing the port is a non-sense.
The error check is replaced by a simple return in the close function.

Bugzilla ID: 624
Fixes: 8a5a0aad5d3e ("ethdev: allow close function to return an error")
Cc: stable@dpdk.org

Reported-by: Anatoly Burakov <redacted>
Signed-off-by: Thomas Monjalon <redacted>

Acked-by: Andrew Rybchenko <redacted>

25/01/2021 10:13, Andrew Rybchenko:

On 1/22/21 8:58 PM, Thomas Monjalon wrote:

quoted

If a failure happens when closing a port,
it was unnecessarily failing again in the function eth_err(),
because of a check against HW removal cause.
Indeed there is a big chance the port is released at this point.
Given the port is in the middle (or at the end) of a close process,
checking the error cause by accessing the port is a non-sense.
The error check is replaced by a simple return in the close function.

Bugzilla ID: 624
Fixes: 8a5a0aad5d3e ("ethdev: allow close function to return an error")
Cc: stable@dpdk.org

Reported-by: Anatoly Burakov <redacted>
Signed-off-by: Thomas Monjalon <redacted>

Acked-by: Andrew Rybchenko <redacted>

Applied

25/01/2021 13:37, Thomas Monjalon:

25/01/2021 10:13, Andrew Rybchenko:

quoted

On 1/22/21 8:58 PM, Thomas Monjalon wrote:

quoted

If a failure happens when closing a port,
it was unnecessarily failing again in the function eth_err(),
because of a check against HW removal cause.
Indeed there is a big chance the port is released at this point.
Given the port is in the middle (or at the end) of a close process,
checking the error cause by accessing the port is a non-sense.
The error check is replaced by a simple return in the close function.

Bugzilla ID: 624
Fixes: 8a5a0aad5d3e ("ethdev: allow close function to return an error")
Cc: stable@dpdk.org

Reported-by: Anatoly Burakov <redacted>
Signed-off-by: Thomas Monjalon <redacted>

Acked-by: Andrew Rybchenko <redacted>

Applied

Sorry please ignore this wrong message, patch not applied.
(will be considered by Ferruh)

On 1/25/2021 9:13 AM, Andrew Rybchenko wrote:

On 1/22/21 8:58 PM, Thomas Monjalon wrote:

quoted

If a failure happens when closing a port,
it was unnecessarily failing again in the function eth_err(),
because of a check against HW removal cause.
Indeed there is a big chance the port is released at this point.
Given the port is in the middle (or at the end) of a close process,
checking the error cause by accessing the port is a non-sense.
The error check is replaced by a simple return in the close function.

Bugzilla ID: 624
Fixes: 8a5a0aad5d3e ("ethdev: allow close function to return an error")
Cc: stable@dpdk.org

Reported-by: Anatoly Burakov <redacted>
Signed-off-by: Thomas Monjalon <redacted>

Acked-by: Andrew Rybchenko <redacted>

Applied to dpdk-next-net/main, thanks.

On 22-Jan-21 5:58 PM, Thomas Monjalon wrote:

If a failure happens when closing a port,
it was unnecessarily failing again in the function eth_err(),
because of a check against HW removal cause.
Indeed there is a big chance the port is released at this point.
Given the port is in the middle (or at the end) of a close process,
checking the error cause by accessing the port is a non-sense.
The error check is replaced by a simple return in the close function.

Bugzilla ID: 624
Fixes: 8a5a0aad5d3e ("ethdev: allow close function to return an error")
Cc: stable@dpdk.org

Reported-by: Anatoly Burakov <redacted>
Signed-off-by: Thomas Monjalon <redacted>
---

Tested-by: Anatoly Burakov <redacted>

-- 
Thanks,
Anatoly