Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Message-ID: <2d3b0e8f422b7ff08a5c4ce804a1884eaf9b5d60.camel@puri.sm>
Subject: Re: [PATCH] scsi: sd: print write through due to no caching mode
 page as warning
From:   Martin Kepplinger <martin.kepplinger@puri.sm>
To:     jejb@linux.ibm.com, martin.petersen@oracle.com
Cc:     dgilbert@interlog.com, bvanassche@acm.org,
        linux-scsi@vger.kernel.org, linux-kernel@vger.kernel.org
Date:   Tue, 12 Oct 2021 17:18:26 +0200
In-Reply-To: <20210122083000.32598-1-martin.kepplinger@puri.sm>
References: <20210122083000.32598-1-martin.kepplinger@puri.sm>
Content-Type: text/plain; charset="UTF-8"
User-Agent: Evolution 3.38.3-1 
MIME-Version: 1.0
Content-Transfer-Encoding: 8bit
Precedence: bulk

Am Freitag, dem 22.01.2021 um 09:30 +0100 schrieb Martin Kepplinger:
> For SD cardreaders it's extremely common not to find cache on disk.
> The following error messages are thus very common and don't point
> to a real error one could try to fix but rather describe how the disk
> works:
> 
> sd 0:0:0:0: [sda] No Caching mode page found
> sd 0:0:0:0: [sda] Assuming drive cache: write through
> 
> Print these messages as warnings instead of errors.
> 
> Signed-off-by: Martin Kepplinger <martin.kepplinger@puri.sm>
> ---
>  drivers/scsi/sd.c | 5 +++--
>  1 file changed, 3 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/scsi/sd.c b/drivers/scsi/sd.c
> index e7c52d6df4dc..db0171c81c5b 100644
> --- a/drivers/scsi/sd.c
> +++ b/drivers/scsi/sd.c
> @@ -2808,7 +2808,8 @@ sd_read_cache_type(struct scsi_disk *sdkp,
> unsigned char *buffer)
>                         }
>                 }
>  
> -               sd_first_printk(KERN_ERR, sdkp, "No Caching mode page
> found\n");
> +               sd_first_printk(KERN_WARNING, sdkp,
> +                               "No Caching mode page found\n");
>                 goto defaults;
>  
>         Page_found:
> @@ -2863,7 +2864,7 @@ sd_read_cache_type(struct scsi_disk *sdkp,
> unsigned char *buffer)
>                                 "Assuming drive cache: write
> back\n");
>                 sdkp->WCE = 1;
>         } else {
> -               sd_first_printk(KERN_ERR, sdkp,
> +               sd_first_printk(KERN_WARNING, sdkp,
>                                 "Assuming drive cache: write
> through\n");
>                 sdkp->WCE = 0;
>         }


hi Bart and all who it may concern,

does this "consmetic" change have any chance of being acceptible? At
least it'd be nice if messages sent as error are real errors that needs
fixing.

the patch still applies.

thank you,

                               martin