Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 25 Mar 2013 22:58:47 +0200
From:      Kimmo Paasiala <kpaasial@gmail.com>
To:        Andriy Gapon <avg@freebsd.org>
Cc:        freebsd-hackers@freebsd.org, FreeBSD Current <freebsd-current@freebsd.org>
Subject:   Re: kern/122838: [devfs] devfs doesn't handle complex paths (like zvol/pool/vms) good
Message-ID:  <CA%2B7WWScTkf=sFA73vb_=QEObRbDhg9DkG9rw3-xb7OF19KJmDw@mail.gmail.com>
In-Reply-To: <5150B71B.6060106@FreeBSD.org>
References:  <5150B598.7050700@FreeBSD.org> <5150B71B.6060106@FreeBSD.org>

next in thread | previous in thread | raw e-mail | index | archive | help
On Mon, Mar 25, 2013 at 10:44 PM, Andriy Gapon <avg@freebsd.org> wrote:
>
> Would like to ask for opinions on this topic...
> Please read this PR for context:
> http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/122838
> Especially Jaakko's insightful description of the problem.
>
>
> -------- Original Message --------
> Message-ID: <5150B598.7050700@FreeBSD.org>
> Date: Mon, 25 Mar 2013 22:37:44 +0200
> From: Andriy Gapon <avg@FreeBSD.org>
> Subject: Re: kern/122838: [devfs] devfs doesn&#39;t handle complex paths (like
> zvol/pool/vms) good
>
>
> Can't believe that we are still where we were more than two years ago...
>
> I think that we have to make this change even if it _might_ break some existing
> rulesets.
>
> Rationale:
> - current behavior is contrary to any documentation
> - current behavior is contrary to common sense
> - current behavior is very hard to describe and account for
> - I presume that very few people actually fully understand the current behavior
> - I presume that even fewer people made a conscious choice to depend or make use
> of its non-trivial features of the current behavior
>
> So, we should make the behavior of devfs pattern consistent with the
> documentation and the common sense.
>
> In addition to Jaakko's patch I propose that we pass FNM_PATHNAME to fnmatch(9),
> so that the matching is indeed consistent with glob(3) / shell glob-ing rules
> for filesystem paths.
>
> --
> Andriy Gapon
>
>
>
Absolutely yes. Any kind of rule based matching should default to
matching full strings. The rule syntax should then offer options to
narrow down the matching to a specific part(s) of the matched strings.

-Kimmo



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CA%2B7WWScTkf=sFA73vb_=QEObRbDhg9DkG9rw3-xb7OF19KJmDw>