* [PATCH net] i40e: fix endless loop under rtnl
@ 2021-08-30 10:34 Jiri Benc
2021-09-14 23:37 ` [Intel-wired-lan] " Jesse Brandeburg
0 siblings, 1 reply; 5+ messages in thread
From: Jiri Benc @ 2021-08-30 10:34 UTC (permalink / raw)
To: netdev; +Cc: Jesse Brandeburg, Tony Nguyen
The loop in i40e_get_capabilities can never end. The problem is that
although i40e_aq_discover_capabilities returns with an error if there's
a firmware problem, the returned error is not checked. There is a check for
pf->hw.aq.asq_last_status but that value is set to I40E_AQ_RC_OK on most
firmware problems.
When i40e_aq_discover_capabilities encounters a firmware problem, it will
enocunter the same problem on its next invocation. As the result, the loop
becomes endless. We hit this with I40E_ERR_ADMIN_QUEUE_TIMEOUT but looking
at the code, it can happen with a range of other firmware errors.
I don't know what the correct behavior should be: whether the firmware
should be retried a few times, or whether pf->hw.aq.asq_last_status should
be always set to the encountered firmware error (but then it would be
pointless and can be just replaced by the i40e_aq_discover_capabilities
return value). However, the current behavior with an endless loop under the
rtnl mutex(!) is unacceptable and Intel has not submitted a fix, although we
explained the bug to them 7 months ago.
This may not be the best possible fix but it's better than hanging the whole
system on a firmware bug.
Tested-by: Stefan Assmann <sassmann@redhat.com>
Signed-off-by: Jiri Benc <jbenc@redhat.com>
---
drivers/net/ethernet/intel/i40e/i40e_main.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/net/ethernet/intel/i40e/i40e_main.c b/drivers/net/ethernet/intel/i40e/i40e_main.c
index 1d1f52756a93..772dd05a0ae8 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_main.c
+++ b/drivers/net/ethernet/intel/i40e/i40e_main.c
@@ -10110,7 +10110,7 @@ static int i40e_get_capabilities(struct i40e_pf *pf,
if (pf->hw.aq.asq_last_status == I40E_AQ_RC_ENOMEM) {
/* retry with a larger buffer */
buf_len = data_size;
- } else if (pf->hw.aq.asq_last_status != I40E_AQ_RC_OK) {
+ } else if (pf->hw.aq.asq_last_status != I40E_AQ_RC_OK || err) {
dev_info(&pf->pdev->dev,
"capability discovery failed, err %s aq_err %s\n",
i40e_stat_str(&pf->hw, err),
--
2.18.1
^ permalink raw reply related [flat|nested] 5+ messages in thread
* Re: [PATCH net] i40e: fix endless loop under rtnl
2021-08-30 10:34 [PATCH net] i40e: fix endless loop under rtnl Jiri Benc
@ 2021-09-14 23:37 ` Jesse Brandeburg
0 siblings, 0 replies; 5+ messages in thread
From: Jesse Brandeburg @ 2021-09-14 23:37 UTC (permalink / raw)
To: Jiri Benc, netdev; +Cc: Tony Nguyen, intel-wired-lan
+intel-wired-lan list
On 8/30/2021 3:34 AM, Jiri Benc wrote:
> The loop in i40e_get_capabilities can never end. The problem is that
> although i40e_aq_discover_capabilities returns with an error if there's
> a firmware problem, the returned error is not checked. There is a check for
> pf->hw.aq.asq_last_status but that value is set to I40E_AQ_RC_OK on most
> firmware problems.
>
> When i40e_aq_discover_capabilities encounters a firmware problem, it will
> enocunter the same problem on its next invocation. As the result, the loop
encounter
> becomes endless. We hit this with I40E_ERR_ADMIN_QUEUE_TIMEOUT but looking
> at the code, it can happen with a range of other firmware errors.
>
> I don't know what the correct behavior should be: whether the firmware
> should be retried a few times, or whether pf->hw.aq.asq_last_status should
> be always set to the encountered firmware error (but then it would be
> pointless and can be just replaced by the i40e_aq_discover_capabilities
> return value).
If the firmware doesn't reply we have a lot of problems.
> However, the current behavior with an endless loop under the
> rtnl mutex(!) is unacceptable and Intel has not submitted a fix, although we
> explained the bug to them 7 months ago.
That's really frustrating, sorry! We are actively working to do better
by working with you more closely. This one must have slipped through the
cracks.
> This may not be the best possible fix but it's better than hanging the whole
> system on a firmware bug.
>
The fix seems fine to me on initial review.
you can add my
Reviewed-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
but...
Missing Fixes: tag?
> Tested-by: Stefan Assmann <sassmann@redhat.com>
> Signed-off-by: Jiri Benc <jbenc@redhat.com>
> ---
> drivers/net/ethernet/intel/i40e/i40e_main.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/net/ethernet/intel/i40e/i40e_main.c b/drivers/net/ethernet/intel/i40e/i40e_main.c
> index 1d1f52756a93..772dd05a0ae8 100644
> --- a/drivers/net/ethernet/intel/i40e/i40e_main.c
> +++ b/drivers/net/ethernet/intel/i40e/i40e_main.c
> @@ -10110,7 +10110,7 @@ static int i40e_get_capabilities(struct i40e_pf *pf,
> if (pf->hw.aq.asq_last_status == I40E_AQ_RC_ENOMEM) {
> /* retry with a larger buffer */
> buf_len = data_size;
> - } else if (pf->hw.aq.asq_last_status != I40E_AQ_RC_OK) {
> + } else if (pf->hw.aq.asq_last_status != I40E_AQ_RC_OK || err) {
> dev_info(&pf->pdev->dev,
> "capability discovery failed, err %s aq_err %s\n",
> i40e_stat_str(&pf->hw, err),
>
^ permalink raw reply [flat|nested] 5+ messages in thread
* [Intel-wired-lan] [PATCH net] i40e: fix endless loop under rtnl
@ 2021-09-14 23:37 ` Jesse Brandeburg
0 siblings, 0 replies; 5+ messages in thread
From: Jesse Brandeburg @ 2021-09-14 23:37 UTC (permalink / raw)
To: intel-wired-lan
+intel-wired-lan list
On 8/30/2021 3:34 AM, Jiri Benc wrote:
> The loop in i40e_get_capabilities can never end. The problem is that
> although i40e_aq_discover_capabilities returns with an error if there's
> a firmware problem, the returned error is not checked. There is a check for
> pf->hw.aq.asq_last_status but that value is set to I40E_AQ_RC_OK on most
> firmware problems.
>
> When i40e_aq_discover_capabilities encounters a firmware problem, it will
> enocunter the same problem on its next invocation. As the result, the loop
encounter
> becomes endless. We hit this with I40E_ERR_ADMIN_QUEUE_TIMEOUT but looking
> at the code, it can happen with a range of other firmware errors.
>
> I don't know what the correct behavior should be: whether the firmware
> should be retried a few times, or whether pf->hw.aq.asq_last_status should
> be always set to the encountered firmware error (but then it would be
> pointless and can be just replaced by the i40e_aq_discover_capabilities
> return value).
If the firmware doesn't reply we have a lot of problems.
> However, the current behavior with an endless loop under the
> rtnl mutex(!) is unacceptable and Intel has not submitted a fix, although we
> explained the bug to them 7 months ago.
That's really frustrating, sorry! We are actively working to do better
by working with you more closely. This one must have slipped through the
cracks.
> This may not be the best possible fix but it's better than hanging the whole
> system on a firmware bug.
>
The fix seems fine to me on initial review.
you can add my
Reviewed-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
but...
Missing Fixes: tag?
> Tested-by: Stefan Assmann <sassmann@redhat.com>
> Signed-off-by: Jiri Benc <jbenc@redhat.com>
> ---
> drivers/net/ethernet/intel/i40e/i40e_main.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/net/ethernet/intel/i40e/i40e_main.c b/drivers/net/ethernet/intel/i40e/i40e_main.c
> index 1d1f52756a93..772dd05a0ae8 100644
> --- a/drivers/net/ethernet/intel/i40e/i40e_main.c
> +++ b/drivers/net/ethernet/intel/i40e/i40e_main.c
> @@ -10110,7 +10110,7 @@ static int i40e_get_capabilities(struct i40e_pf *pf,
> if (pf->hw.aq.asq_last_status == I40E_AQ_RC_ENOMEM) {
> /* retry with a larger buffer */
> buf_len = data_size;
> - } else if (pf->hw.aq.asq_last_status != I40E_AQ_RC_OK) {
> + } else if (pf->hw.aq.asq_last_status != I40E_AQ_RC_OK || err) {
> dev_info(&pf->pdev->dev,
> "capability discovery failed, err %s aq_err %s\n",
> i40e_stat_str(&pf->hw, err),
>
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [PATCH net] i40e: fix endless loop under rtnl
2021-09-14 23:37 ` [Intel-wired-lan] " Jesse Brandeburg
@ 2021-09-15 6:40 ` Jiri Benc
-1 siblings, 0 replies; 5+ messages in thread
From: Jiri Benc @ 2021-09-15 6:40 UTC (permalink / raw)
To: Jesse Brandeburg; +Cc: netdev, Tony Nguyen, intel-wired-lan
On Tue, 14 Sep 2021 16:37:55 -0700, Jesse Brandeburg wrote:
> The fix seems fine to me on initial review.
> you can add my
> Reviewed-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
Thanks!
> but...
> Missing Fixes: tag?
I should have added that, yes. I sent a v2 yesterday with the tag
included,
https://patchwork.kernel.org/project/netdevbpf/patch/452ff4ddfef7fc8f558a8c8eb7a8050688760e11.1631609537.git.jbenc@redhat.com/
Could you reply with your Reviewed-by to that one? Alternatively, I can
send a v3.
Thanks,
Jiri
^ permalink raw reply [flat|nested] 5+ messages in thread
* [Intel-wired-lan] [PATCH net] i40e: fix endless loop under rtnl
@ 2021-09-15 6:40 ` Jiri Benc
0 siblings, 0 replies; 5+ messages in thread
From: Jiri Benc @ 2021-09-15 6:40 UTC (permalink / raw)
To: intel-wired-lan
On Tue, 14 Sep 2021 16:37:55 -0700, Jesse Brandeburg wrote:
> The fix seems fine to me on initial review.
> you can add my
> Reviewed-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
Thanks!
> but...
> Missing Fixes: tag?
I should have added that, yes. I sent a v2 yesterday with the tag
included,
https://patchwork.kernel.org/project/netdevbpf/patch/452ff4ddfef7fc8f558a8c8eb7a8050688760e11.1631609537.git.jbenc at redhat.com/
Could you reply with your Reviewed-by to that one? Alternatively, I can
send a v3.
Thanks,
Jiri
^ permalink raw reply [flat|nested] 5+ messages in thread
end of thread, other threads:[~2021-09-15 6:41 UTC | newest]
Thread overview: 5+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2021-08-30 10:34 [PATCH net] i40e: fix endless loop under rtnl Jiri Benc
2021-09-14 23:37 ` Jesse Brandeburg
2021-09-14 23:37 ` [Intel-wired-lan] " Jesse Brandeburg
2021-09-15 6:40 ` Jiri Benc
2021-09-15 6:40 ` [Intel-wired-lan] " Jiri Benc
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.