LWN: Comments on "Toward a safer sysfs" https://lwn.net/Articles/229774/ This is a special feed containing comments posted to the individual LWN article titled "Toward a safer sysfs". en-us Mon, 29 Sep 2025 23:43:03 +0000 Mon, 29 Sep 2025 23:43:03 +0000 https://www.rssboard.org/rss-specification lwn@lwn.net Toward a safer sysfs https://lwn.net/Articles/231142/ https://lwn.net/Articles/231142/ jd Does that mean a device driver to make weasels erupt from a user's <br> nostrils will be broken? Can we add a special API for such drivers, say <br> next April 1st?<br> <p> Thu, 19 Apr 2007 18:55:05 +0000 Toward a safer sysfs https://lwn.net/Articles/230117/ https://lwn.net/Articles/230117/ corbet Exactly that - it returns ENODEV. Thu, 12 Apr 2007 03:09:15 +0000 Toward a safer sysfs https://lwn.net/Articles/230115/ https://lwn.net/Articles/230115/ flewellyn Okay, so what DOES happen when you try to read from, or write to, a sysfs file for which the <br> device no longer exists?<br> <p> I mean, obviously this new code will prevent the system from causing weasels to erupt from your <br> nostrils or whatever other catastrophic event might occur, but surely there should be some <br> defined behavior here? Something along the lines of "device no longer exists, go away"?<br> Thu, 12 Apr 2007 02:58:22 +0000