svn commit: r191816 - in head/sys: conf dev/cxgb/ulp/tom kern net net80211 netgraph netinet netinet6 netipsec nfsclient sys

Marko Zec zec at freebsd.org
Wed May 6 09:31:57 UTC 2009


On Tuesday 05 May 2009 10:33:52 Alexander Leidinger wrote:
> Quoting Marko Zec <zec at FreeBSD.org> (from Tue, 5 May 2009 10:56:12
>
> +0000 (UTC)):
> >   The exact placement of the CURVNET_SET() / CURVNET_RESTORE() macros
> >   was a result of an empirical iterative process,
>
> Was the linuxulator taken into account in this process?

Hmm I have no simple answer to that and similar questions, so:

- this change has no effect on nooptions VIMAGE builds whatsoever, including 
linuxulator;

- with options VIMAGE enabled, linuxulator builds fine and kldloads as well;

- on few occasions in 2008. I ran linux binaries (firefox I think, in hope to 
get flash working) on VIMAGE kernels from p4/vimage branch, and they had no 
trouble communicating over network sockets;

- I don't run linux binaries these days on -CURRENT, hence do not know whether 
any issues would be observable with linuxulator and options VIMAGE kernels;

- there are _many_ kernel subsystems that yet have to be made compatible with 
options VIMAGE, so I'd join Julian in his call to encourage people to start 
playing with VIMAGE kernels and report back any issues or suspicious 
behavior.

Cheers,

Marko


More information about the svn-src-all mailing list