linux-kernel - Re: [PATCH] PNP: handle sysfs errors

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: Windows password security audit tool. GUI, reports in PDF.

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

Message-Id: <200610112354.06813.dtor@insightbb.com>
Date:	Wed, 11 Oct 2006 23:54:04 -0400
From:	Dmitry Torokhov <dtor@...ightbb.com>
To:	Andrew Morton <akpm@...l.org>
Cc:	Jeff Garzik <jeff@...zik.org>, ambx1@....rr.com,
	LKML <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] PNP: handle sysfs errors

On Wednesday 11 October 2006 18:45, Andrew Morton wrote:
> On Wed, 11 Oct 2006 17:51:09 -0400
> Jeff Garzik <jeff@...zik.org> wrote:
> 
> > +	int rc = device_create_file(&dev->dev,&dev_attr_options);
> > +	if (rc) goto err;
> > +	rc = device_create_file(&dev->dev,&dev_attr_resources);
> > +	if (rc) goto err_opt;
> > +	rc = device_create_file(&dev->dev,&dev_attr_id);
> > +	if (rc) goto err_res;
> > +
> >  	return 0;
> > +
> > +err_res:
> > +	device_remove_file(&dev->dev,&dev_attr_resources);
> > +err_opt:
> > +	device_remove_file(&dev->dev,&dev_attr_options);
> > +err:
> > +	return rc;
> 
> That's a common pattern, isn't it?
> 
> I wonder if we could create some sort of automatic-unwinding engine:
> 
> int pnp_interface_attach_device(struct pnp_dev *dev)
> {
> 	struct unwind_engine *u = NULL;
> 	int err;
> 
> 	u = UNWINDABLE(u, err, device_create_file(&dev->dev,&dev_attr_options),
> 			device_remove_file, &dev->dev, &dev_attr_options);
> 	u = UNWINDABLE(u, err, device_create_file(&dev->dev,&dev_attr_id)
> 			device_remove_file, &dev->dev, &dev_attr_options);
> 	u = UNWINDABLE(u, err, device_create_file(&dev->dev,&dev_attr_id),
> 			device_remove_file, &dev->dev, &dev_attr_id);
> 	err = unwind(err, u);
> 	return err;
> }
> 
> and, umm,
> 
> #define UNWINDABLE(u, err, expr, undo_fn, arg1, arg2)
> 	if (err == 0) {
> 		err = (expr);
> 		if (err == 0)
> 			u = add_unwind(u, undo_fn, arg1, arg2);
> 	}
> 	u;
> 
> int unwind(int err, struct unwind_engine *u)
> {
> 	if (err == 0)
> 		return 0;
> 	for (all entries in u in opposite order) {
> 		u->fn(u->arg0, u>arg1);
> 		kfree(u);
> 	}
> 	return err;
> }
> 
> 
> I dunno - probably too crappy to live, but it'd encourage/help people to
> dtrt.

The best solution is to switch drivers with more than 2 attributes
to use arribute groups which collapses unwinding nicely.

-- 
Dmitry
-
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/