Re: [Bug report] NFS patch breaks TLS device-offloaded TX zerocopy

[Tariq Toukan <ttoukan.linux@gmail.com>]

On 05/08/2024 13:40, Tariq Toukan wrote:
> Hi,
> 
> A recent patch [1] to 'fs' broke the TX TLS device-offloaded flow 
> starting from v6.11-rc1.
> 
> The kernel crashes. Different runs result in different kernel traces.
> See below [2].
> All of them disappear once patch [1] is reverted.
> 
> The issues appears only with "sendfile on and zerocopy on".
> We couldn't repro with "sendfile off", or with "sendfile on and zerocopy 
> off".
> 
> The repro test is as simple as a repeated client/server communication 
> (wrk/nginx), with sendfile on and zc on, and with "tls-hw-tx-offload: on".
> 
> $ for i in `seq 10`; do wrk -b::2:2:2:3 -t10 -c100 -d15 --timeout 5s 
> https://[::2:2:2:2]:20448/16000b.img; done
> 
> We can provide more details if needed, to help with the analysis and debug.
> 
> Regards,
> Tariq
> 
> [1]
> commit 49b29a573da83b65d5f4ecf2db6619bab7aa910c
> Author: Christoph Hellwig <hch@lst.de>
> Date:   Mon May 27 18:36:09 2024 +0200
> 
>      nfs: add support for large folios
> 
>      NFS already is void of folio size assumption, so just pass the 
> chunk size
>      to __filemap_get_folio and set the large folio address_space flag 
> for all
>      regular files.
> 
>      Signed-off-by: Christoph Hellwig <hch@lst.de>
>      Tested-by: Sagi Grimberg <sagi@grimberg.me>
>      Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
> 
>   fs/nfs/file.c  | 4 +++-
>   fs/nfs/inode.c | 1 +
>   2 files changed, 4 insertions(+), 1 deletion(-)
> 
> 

Hi all,

FYI, issue is fixed by:

commit dd6e972cc589 ("net/mlx5e: kTLS, Fix incorrect page refcounting")

https://lore.kernel.org/netdev/20241107183527.676877-5-tariqt@nvidia.com/

Thanks,
Tariq