linux-kernel - Re: [PATCH 2/6] block: push down BKL into .open and .release

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-Id: <201007042309.09727.arnd@arndb.de>
Date:	Sun, 4 Jul 2010 23:09:09 +0200
From:	Arnd Bergmann <arnd@...db.de>
To:	Sam Ravnborg <sam@...nborg.org>
Cc:	Jens Axboe <axboe@...nel.dk>,
	Christoph Hellwig <hch@...radead.org>,
	linux-kernel@...r.kernel.org, linux-fsdevel@...r.kernel.org,
	John Kacur <jkacur@...hat.com>,
	Frederic Weisbecker <fweisbec@...il.com>,
	linux-scsi@...r.kernel.org
Subject: Re: [PATCH 2/6] block: push down BKL into .open and .release

On Sunday 04 July 2010 10:01:46 Sam Ravnborg wrote:
> > --- a/drivers/block/amiflop.c
> > +++ b/drivers/block/amiflop.c
> > @@ -1555,10 +1555,13 @@ static int floppy_open(struct block_device *bdev, fmode_t mode)
> >  	int old_dev;
> >  	unsigned long flags;
> >  
> > +	lock_kernel();
> >  	old_dev = fd_device[drive];
> >  
> > -	if (fd_ref[drive] && old_dev != system)
> > +	if (fd_ref[drive] && old_dev != system) {
> > +		unlock_kernel();
> >  		return -EBUSY;
> > +	}
> >  
> >  	if (mode & (FMODE_READ|FMODE_WRITE)) {
> >  		check_disk_change(bdev);
> > @@ -1571,8 +1574,10 @@ static int floppy_open(struct block_device *bdev, fmode_t mode)
> >  			fd_deselect (drive);
> >  			rel_fdc();
> >  
> > -			if (wrprot)
> > +			if (wrprot) {
> > +				unlock_kernel();
> >  				return -EROFS;
> > +			}
> >  		}
> >  	}
> >  
> > @@ -1589,6 +1594,7 @@ static int floppy_open(struct block_device *bdev, fmode_t mode)
> >  	printk(KERN_INFO "fd%d: accessing %s-disk with %s-layout\n",drive,
> >  	       unit[drive].type->name, data_types[system].name);
> >  
> > +	unlock_kernel();
> >  	return 0;
> >  }
> 
> Using goto for errorhandling here would have been nicer.

I tried to minimize the chance for breaking stuff in code I cannot easily
test build. As shown by the bug you found in my pktcdvd patch, changing the
control flow of a function has a higher potential of introducing bugs,
so I didn't do it for drivers that people don't care much about any more.

> Also did you forget to include smp_locks.h?

No, that was already there from the first patch.

> Lot's of other places could benefit from improved goto error handling.
> The driver maintainers should do this (if there is a maintainer).

If that makes Jens accept my patches, I'd gladly change them this way.
I agree that it's cleaner and I always write my own code like that,
but when modifying (mostly) unmaintained code, my preference is on
trying to ensure the patch is correct.


> I did not find anything else going through this patch.

Ok, thanks a lot for looking through this!

	Arnd
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/