Re: [PATCH] Fix Trivial Warnining in drivers/ata/libata-scsi.c

[Subrata Modak <subrata@linux.vnet.ibm.com>]

Hi Jeff,

On Mon, 2009-04-27 at 08:03 -0400, Jeff Garzik wrote:
> Subrata Modak wrote:
> > Hi Jeff,
> > 
> > I observed the following warning for drivers/ata/libata-scsi.c:
> > 
> > drivers/ata/libata-scsi.c: In function ‘ata_scsi_scan_host’:
> > drivers/ata/libata-scsi.c:3272: warning: ‘dev’ may be used uninitialized in this function
> > ---
> > 
> > In most instances=>
> > 	struct ata_device *dev
> > has been initialized except the above instance. My trivial patch below fixes this.
> > Please include it if you like it.
> > 
> > To: Jeff Garzik <jgarzik@pobox.com>
> > Cc: linux-ide@vger.kernel.org
> > Cc: Balbir Singh <balbir@linux.vnet.ibm.com>
> > Cc: Sachin P Sant <sachinp@linux.vnet.ibm.com>
> > Signed-Off-By: Subrata Modak <subrata@linux.vnet.ibm.com>
> > ---
> > 
> > --- a/drivers/ata/libata-scsi.c	2009-04-27 11:39:27.000000000 +0530
> > +++ b/drivers/ata/libata-scsi.c	2009-04-27 15:52:17.000000000 +0530
> > @@ -3269,7 +3269,7 @@ void ata_scsi_scan_host(struct ata_port 
> >  	int tries = 5;
> >  	struct ata_device *last_failed_dev = NULL;
> >  	struct ata_link *link;
> > -	struct ata_device *dev;
> > +	struct ata_device *dev = NULL;
> >  
> >  	if (ap->flags & ATA_FLAG_DISABLED)
> 
> What was the conclusion of the analysis?  :)
> 
> Is this a kernel bug, or gcc bug?

Oops. I just wrote the patch to disable the build warning. As, i found
that in other instances of use of the same declaration, it has been
initialized, and the compiler was complaining for this instance alone.
Hence, thought of addressing through a small fix.

Regards--
Subrata

> 
> 	Jeff
> 
> 
> 
>