Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 09 Nov 2005 08:01:33 +0100
From:      "Poul-Henning Kamp" <phk@phk.freebsd.dk>
To:        Nate Lawson <nate@root.org>
Cc:        Craig Rodrigues <rodrigc@FreeBSD.org>, cvs-src@FreeBSD.org, src-committers@FreeBSD.org, cvs-all@FreeBSD.org
Subject:   Re: cvs commit: src/sys/kern vfs_mount.c 
Message-ID:  <26185.1131519693@critter.freebsd.dk>
In-Reply-To: Your message of "Tue, 08 Nov 2005 19:24:06 PST." <43716BD6.5050604@root.org> 

next in thread | previous in thread | raw e-mail | index | archive | help
In message <43716BD6.5050604@root.org>, Nate Lawson writes:

>While I don't have ideas for a better general mechanism for this, I 
>think it sets a bad precedent.  We can't have every complex syscall 
>transporting its own error message strings back to the user program. 
>And we can't expand errno to be the union of every single API-specific 
>error either.

That was one the main points of my ioctl talk at BSDcan.

One way to solve it:

Hang a sbuf pointer in each thread and APIs cound stuff their error
message there.

Add a new syscall to return the string to userland.

Modify perror(3), err(3) and similar to pull out the "extended"
error, if there is one.

-- 
Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
phk@FreeBSD.ORG         | TCP/IP since RFC 956
FreeBSD committer       | BSD since 4.3-tahoe    
Never attribute to malice what can adequately be explained by incompetence.



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?26185.1131519693>