On Wed, Jan 23, 2008, Peter Kalbus wrote:
you're right, that make abolutly sense
attached are the corrected patches
one more, since system.h has to be change now, too
Ok, committed. Thanks for your contribution.
Details: http://rpm5.org/cvs/chngview?cn=9423
i've attached two patches, which let's the makedev() compile on qnx
and netbsd
it's build against 5.1.SNAPSHOT-20080120
patch-aa
Description: Binary data
patch-ab
Description: Binary data
i'd like to get a review, if this has a chance to come into the rpm5
source tree for qnx support
On Wed, Jan 23, 2008, Peter Kalbus wrote:
i've attached two patches, which let's the makedev() compile on qnx and
netbsd
it's build against 5.1.SNAPSHOT-20080120
I would move the Makedev() definition to system.h in order to remove
redundancy. Also, similar stuff already is in system.h, so
on qnx makedev() takes three parameters. the third parameter (first in the
list) is needed to address the node in a qnet environment. i got the following
information from the foundry27 community:
Most (all?) 3rd party code isn't QNET aware so to be totally
correct it should
On Tue, Jan 22, 2008, Peter Kalbus wrote:
on qnx makedev() takes three parameters. the third parameter (first in the
list) is needed to address the node in a qnet environment. i got the
following information from the foundry27 community:
Most (all?) 3rd party code isn't QNET aware so
the patches as it works on qnx and passes a netbsd cross check.
-piet
Original-Nachricht
Datum: Tue, 22 Jan 2008 16:41:38 +0100
Von: Ralf S. Engelschall [EMAIL PROTECTED]
An: rpm-devel@rpm5.org
Betreff: Re: makedev() on qnx
On Tue, Jan 22, 2008, Peter Kalbus wrote:
on qnx