Re: kern/55777: more rump fixes

[Christos Zoulas <christos%zoulas.com@localhost>]

The following reply was made to PR kern/55777; it has been noted by GNATS.

From: Christos Zoulas <christos%zoulas.com@localhost>
To: Ruslan Nikolaev <nruslan_devel%yahoo.com@localhost>
Cc: gnats-bugs%netbsd.org@localhost,
 kern-bug-people%netbsd.org@localhost,
 gnats-admin%netbsd.org@localhost,
 netbsd-bugs%netbsd.org@localhost
Subject: Re: kern/55777: more rump fixes
Date: Tue, 3 Nov 2020 08:42:21 -0500

 --Apple-Mail=_742F619D-B2DB-450C-927E-29588387AD95
 Content-Transfer-Encoding: quoted-printable
 Content-Type: text/plain;
 	charset=us-ascii
 
 
 
 > On Nov 2, 2020, at 11:48 PM, Ruslan Nikolaev <nruslan_devel%yahoo.com@localhost> =
 wrote:
 >=20
 > I see, let me try that. But what is the rationale to have it defined =
 in *both*
 >=20
 > librumpnet_local (through uipc_usrreq.c) and librumpnet (through =
 net_stub.c)?
 
 rump tries to split the kernel into components so that programs that =
 just need a
 particular subsystem from the kernel don't need to link against the =
 whole kernel.
 Unfortunately the kernel was not designed with such a split in mind, so =
 that variables
 are shared between components. The hacky way to make this work is to =
 define them
 in both components. A better way is to put them in a third component and =
 have the other
 two require it, but that introduces overhead.
 
 christos
 
 --Apple-Mail=_742F619D-B2DB-450C-927E-29588387AD95
 Content-Transfer-Encoding: 7bit
 Content-Disposition: attachment;
 	filename=signature.asc
 Content-Type: application/pgp-signature;
 	name=signature.asc
 Content-Description: Message signed with OpenPGP
 
 -----BEGIN PGP SIGNATURE-----
 Comment: GPGTools - http://gpgtools.org
 
 iF0EARECAB0WIQS+BJlbqPkO0MDBdsRxESqxbLM7OgUCX6FePQAKCRBxESqxbLM7
 OmuXAKDo4W1J2H0ELVTXWVqd3eZuUpHVpQCguj9XlqXzItMWiQLLi+lbKSsuzZQ=
 =u1xK
 -----END PGP SIGNATURE-----
 
 --Apple-Mail=_742F619D-B2DB-450C-927E-29588387AD95--