[SRU][Wily][Xenial][PATCH 2/2] UBUNTU: SAUCE: (no-up) scsi: storvsc: Filter out storvsc messages CD-ROM medium not present

Andy Whitcroft apw at canonical.com
Thu Jul 7 11:46:34 UTC 2016


On Wed, Jul 06, 2016 at 04:03:49PM -0400, Joseph Salisbury wrote:
> On 07/04/2016 12:32 PM, Andy Whitcroft wrote:
> > On Fri, Jul 01, 2016 at 09:44:14PM -0400, Joseph Salisbury wrote:
> >> From: Cathy Avery <cavery at redhat.com>
> >>
> >> BugLink: http://bugs.launchpad.net/bugs/1590655
> >>
> >> Signed-off-by: Joseph Salisbury <joseph.salisbury at canonical.com>
> >> ---
> >>  drivers/scsi/storvsc_drv.c | 4 +++-
> >>  1 file changed, 3 insertions(+), 1 deletion(-)
> >>
> >> diff --git a/drivers/scsi/storvsc_drv.c b/drivers/scsi/storvsc_drv.c
> >> index 95462c5..9948745 100644
> >> --- a/drivers/scsi/storvsc_drv.c
> >> +++ b/drivers/scsi/storvsc_drv.c
> >> @@ -963,7 +963,9 @@ static void storvsc_command_completion(struct storvsc_cmd_request *cmd_request)
> >>  	if (scmnd->result) {
> >>  		if (scsi_normalize_sense(scmnd->sense_buffer,
> >>  				SCSI_SENSE_BUFFERSIZE, &sense_hdr) &&
> >> -		    do_logging(STORVSC_LOGGING_ERROR))
> >> +		    !(sense_hdr.sense_key == NOT_READY &&
> >> +				 sense_hdr.asc == 0x03A) &&
> >> +		    do_logging(STORVSC_LOGGING_ERROR))
> >>  			scsi_print_sense_hdr(scmnd->device, "storvsc",
> >>  					     &sense_hdr);
> > Do we really need that foundation patch.  This appears to just add a
> > couple of additional clauses to this if which would apply with wiggling
> > pretty directly to the original code?
> >
> > -apw
> I took another look at this patch set.  It looks like we do actually
> need some of the first prereq patch.  It defines some constants and the
> do_logging function:
> 
> 
> +#define STORVSC_LOGGING_ERROR  1
>  
> AND
> 
> +static inline bool do_logging(int level)
> +{
> +       return logging_level >= level;
> +}
> 
> 
> Without these definitions, there will be build failures:
> 
> 
> /home/jsalisbury/bugs/lp1590655/ubuntu-xenial/drivers/scsi/storvsc_drv.c: In
> function 'storvsc_command_completion':
> /home/jsalisbury/bugs/lp1590655/ubuntu-xenial/drivers/scsi/storvsc_drv.c:948:7:
> error: implicit declaration of function 'do_logging'
> [-Werror=implicit-function-declaration]
>        do_logging(STORVSC_LOGGING_ERROR))
>        ^
> /home/jsalisbury/bugs/lp1590655/ubuntu-xenial/drivers/scsi/storvsc_drv.c:948:18:
> error: 'STORVSC_LOGGING_ERROR' undeclared (first use in this function)
>        do_logging(STORVSC_LOGGING_ERROR))
>                   ^
> /home/jsalisbury/bugs/lp1590655/ubuntu-xenial/drivers/scsi/storvsc_drv.c:948:18:
> note: each undeclared identifier is reported only once for each function
> it appears in
> /home/jsalisbury/bugs/lp1590655/ubuntu-xenial/drivers/scsi/storvsc_drv.c:948:40:
> error: expected ';' before ')' token
>        do_logging(STORVSC_LOGGING_ERROR))
> 
> 
> 
> 
> I can move these declarations into a single backport of the new SAUCE
> patch, or we can keep them separate by applying mainline commit
> f8aea701b as a prereq.  Which way would you prefer?

The patch you are trying apply here actually doesn't have do_loggin() in
it, it looks like it does because for unknown reasons the version you
submitted removes and readds the same line.  Applying that patch here it
actually is just:

+                   !(sense_hdr.sense_key == NOT_READY &&
+                                sense_hdr.asc == 0x03A) &&

Separatly it looks like tim has applied this foundation patch as part of
a larger batch on Xenial at least, so this discussion may well be moot.

-apw




More information about the kernel-team mailing list