Re: [PATCH 1/1] net/mlx5: update debug log level for remote access error syndromes

[Aru]

Hi Leon,

On 10/25/22 10:48 PM, Leon Romanovsky wrote:
> On Tue, Oct 25, 2022 at 02:22:01AM -0700, Arumugam Kolappan wrote:
> > The mlx5 driver dumps the entire CQE buffer by default for few syndromes.
> > Some syndromes are expected due to the application behavior [ex:
> > MLX5_CQE_SYNDROME_REMOTE_ACCESS_ERR, MLX5_CQE_SYNDROME_REMOTE_OP_ERR and
> > MLX5_CQE_SYNDROME_LOCAL_PROT_ERR]. Hence, for these syndromes, the patch
> > converts the log level from KERN_WARNING to KERN_DEBUG. This enables the
> > application to get the CQE buffer dump by changing to KERN_DEBUG level
> > as and when needed.
> >
> > Suggested-by: Leon Romanovsky <leon@xxxxxxxxxx>
> > Signed-off-by: Arumugam Kolappan <aru.kolappan@xxxxxxxxxx>
> > ---
> >   drivers/infiniband/hw/mlx5/cq.c | 30 ++++++++++++++++++++++--------
> >   1 file changed, 22 insertions(+), 8 deletions(-)
> >
> > diff --git a/drivers/infiniband/hw/mlx5/cq.c b/drivers/infiniband/hw/mlx5/cq.c
> > index be189e0..d665129 100644
> > --- a/drivers/infiniband/hw/mlx5/cq.c
> > +++ b/drivers/infiniband/hw/mlx5/cq.c
> > @@ -267,10 +267,25 @@ static void handle_responder(struct ib_wc *wc, struct mlx5_cqe64 *cqe,
> >   	wc->wc_flags |= IB_WC_WITH_NETWORK_HDR_TYPE;
> >   }
> >   
> > -static void dump_cqe(struct mlx5_ib_dev *dev, struct mlx5_err_cqe *cqe)
> > +static void dump_cqe(struct mlx5_ib_dev *dev, struct mlx5_err_cqe *cqe,
> > +		     struct ib_wc *wc, int dump)
> >   {
> > -	mlx5_ib_warn(dev, "dump error cqe\n");
> > -	mlx5_dump_err_cqe(dev->mdev, cqe);
> > +	const char *level;
> > +
> > +	if (!dump)
> > +		return;
> > +
> > +	mlx5_ib_warn(dev, "WC error: %d, Message: %s\n", wc->status,
> > +		     ib_wc_status_msg(wc->status));
> Aren't you interested "to hide" this print too? Right now, it will
> be printed without relation to your "dump" variable value.

Thanks for pointing out this. Yes. This line also needs to be covered by 
debug log level.

Current existing functions ("mlx5_ib_warn(), mlx5_ib_err() ...) do not 
accept log-level as argument.

So I've added a new fn: mlx5_ib_log(..) which takes log-level as the 
first argument and print it accordingly.


The updated patch will be posted in the next email for your review.

> > +
> > +	if (dump == 1)
> > +		level = KERN_WARNING;
> > +
> > +	if (dump == 2)
> > +		level = KERN_DEBUG;
> Please change dump_cqe() arguments to receive level directly, so you
> will set "dump = KERN_DEBUG" and not not "dump = 2" in
> mlx5_handle_error_cqe().

Yes. This is also taken care in the updated patch.
Also I've updated the subject line as you mentioned in the other email.


Thanks

Aru

> Thanks