[Patch v6 0/4] bnx2x: Fix error recovering in switch configuration

[Patch v6 0/4] bnx2x: Fix error recovering in switch configuration

[Thinh Tran]

While injecting PCIe errors to the upstream PCIe switch of
a BCM57810 NIC, system hangs/crashes were observed.

After several calls to bnx2x_tx_timout() complete,
bnx2x_nic_unload() is called to free up HW resources
and bnx2x_napi_disable() is called to release NAPI objects.
Later, when the EEH driver calls bnx2x_io_slot_reset() to
complete the recovery process, bnx2x attempts to disable
NAPI again by calling bnx2x_napi_disable() and freeing
resources which have already been freed, resulting in a
hang or crash.

This patch set introduces a new flag to track the HW 
resource and NAPI allocation state, refactor duplicated
code into a single function, check page pool allocation
status before freeing, and reduces debug output when
a TX timeout event occurs.


Signed-off-by: Thinh Tran <thinhtr@linux.vnet.ibm.com>
Reviewed-by: Manish Chopra <manishc@marvell.com>
Tested-by: Abdul Haleem <abdhalee@in.ibm.com>
Tested-by: David Christensen <drc@linux.vnet.ibm.com>

Reviewed-by: Simon Horman <simon.horman@corigine.com>
Tested-by: Venkata Sai Duggi <venkata.sai.duggi@ibm.com>

  v6:
   - Clarifying and updating commit messages
  v5:
   - Breaking down into a series of individual patches
  v4:
   - factoring common code into new function bnx2x_stop_nic()
     that disables and releases IRQs and NAPIs
  v3:
   - no changes, just repatched to the latest driver level
   - updated the reviewed-by Manish in October, 2022
  v2:
   - Check the state of the NIC before calling disable nappi
     and freeing the IRQ
   - Prevent recurrence of TX timeout by turning off the carrier,
     calling netif_carrier_off() in bnx2x_tx_timeout()
   - Check and bail out early if fp->page_pool already freed


Thinh Tran (4):
  bnx2x: new the bp->nic_stopped variable for checking NIC status
  bnx2x: factor out common code to bnx2x_stop_nic()
  bnx2x: Prevent access to a freed page in page_pool
  bnx2x: prevent excessive debug information during a TX timeout

 drivers/net/ethernet/broadcom/bnx2x/bnx2x.h   |  2 ++
 .../net/ethernet/broadcom/bnx2x/bnx2x_cmn.c   | 33 ++++++++++++++-----
 .../net/ethernet/broadcom/bnx2x/bnx2x_cmn.h   |  4 +++
 .../net/ethernet/broadcom/bnx2x/bnx2x_main.c  | 26 +++------------
 .../net/ethernet/broadcom/bnx2x/bnx2x_vfpf.c  |  9 ++---
 5 files changed, 37 insertions(+), 37 deletions(-)

-- 
2.27.0

Re: [Patch v6 0/4] bnx2x: Fix error recovering in switch configuration

[Thinh Tran]

Hi,

Could we proceed with advancing these patches? They've been in the 
"Awaiting Upstream" state for a while now. Notably, one of them has 
successfully made it to the mainline kernel:
  [v6,1/4] bnx2x: new flag for tracking HW resource
 
https://github.com/torvalds/linux/commit/bf23ffc8a9a777dfdeb04232e0946b803adbb6a9

As testing the latest kernel, we are still encountering crashes due to 
the absence of one of the patches:
   [v6,3/4] bnx2x: Prevent access to a freed page in page_pool.

Is there anything specific I need to do to help moving these patches 
forward?
We would greatly appreciate if they could be incorporated into the 
mainline kernel.

Thank you,
Thinh Tran

Re: [Patch v6 0/4] bnx2x: Fix error recovering in switch configuration

[Simon Horman]

On Thu, Nov 16, 2023 at 10:08:34AM -0600, Thinh Tran wrote:
> Hi,
> 
> Could we proceed with advancing these patches? They've been in the
> "Awaiting Upstream" state for a while now. Notably, one of them has
> successfully made it to the mainline kernel:
>  [v6,1/4] bnx2x: new flag for tracking HW resource
> 
> https://github.com/torvalds/linux/commit/bf23ffc8a9a777dfdeb04232e0946b803adbb6a9
> 
> As testing the latest kernel, we are still encountering crashes due to
> the absence of one of the patches:
>   [v6,3/4] bnx2x: Prevent access to a freed page in page_pool.
> 
> Is there anything specific I need to do to help moving these patches
> forward?
> We would greatly appreciate if they could be incorporated into the
> mainline kernel.

Hi Thinh Tran,

I'd suggest that the best way to move this forwards would
be to rebase the remaining patches on net-next and posting them as a v7.
It would be useful to include the information above in the cover letter.
And to annotate that they are targeting net-next in the subject of
each patch and the cover letter.

	Subject: [PATCH net-next v8 x/3] ...

Re: [Patch v6 0/4] bnx2x: Fix error recovering in switch configuration

[Thinh Tran]

Hi all,

I hope this message finds you well. I'm reaching out to move forward 
with these patches.  If there are any remaining concerns or if 
additional information is needed from my side, please let me know.
Your guidance on the next steps would be greatly appreciated.

Best regards,
Thinh Tran

On 11/16/2023 10:08 AM, Thinh Tran wrote:
> Hi,
> 
> Could we proceed with advancing these patches? They've been in the 
> "Awaiting Upstream" state for a while now. Notably, one of them has 
> successfully made it to the mainline kernel:
>   [v6,1/4] bnx2x: new flag for tracking HW resource
> 
> https://github.com/torvalds/linux/commit/bf23ffc8a9a777dfdeb04232e0946b803adbb6a9
> 
> As testing the latest kernel, we are still encountering crashes due to 
> the absence of one of the patches:
>    [v6,3/4] bnx2x: Prevent access to a freed page in page_pool.
> 
> Is there anything specific I need to do to help moving these patches 
> forward?
> We would greatly appreciate if they could be incorporated into the 
> mainline kernel.
> 
> Thank you,
> Thinh Tran

Re: [Patch v6 0/4] bnx2x: Fix error recovering in switch configuration

[Jakub Kicinski]

On Wed, 17 Jan 2024 15:56:21 -0600 Thinh Tran wrote:
> I hope this message finds you well. I'm reaching out to move forward 
> with these patches.  If there are any remaining concerns or if 
> additional information is needed from my side, please let me know.
> Your guidance on the next steps would be greatly appreciated.

If there are any patches that got stuck in a limbo for a long time
please repost them in a new thread. If I'm looking this up right in
online archives the thread is 6 months old, I've deleted the old
messages already :(

Re: [Patch v6 0/4] bnx2x: Fix error recovering in switch configuration

[Thinh Tran]

On 1/17/2024 5:55 PM, Jakub Kicinski wrote:
> If there are any patches that got stuck in a limbo for a long time
> please repost them in a new thread. If I'm looking this up right in
> online archives the thread is 6 months old, I've deleted the old
> messages already :(

I will work on re-posting them in a new different thread.
Thank you.
Thinh Tran