I'm looking for reviewers to review my changes for:
PSARC 2007/296
<http://www.opensolaris.org/os/community/arc/caselog/2007/296> nemo
ethernet stats update, part 2
PSARC 2007/298
<http://www.opensolaris.org/os/community/arc/caselog/2007/298> GLDv3
link status logging
6561815
<http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6561815>
nemo needs a few more kstats
6561809
<http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6561809>
nemo should perform link status logging on behalf of drivers
6561860
<http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6561860> rge
kstats duplication
There are two reviews, one open and one closed. (The closed one has
everything that is in the open one, plus the required changes to the
closed ixgb and nge drivers.)
The reviews are located at:
http://cr.grommit.com/~gdamore/nemo-stats-log
and
http://jurassic.eng/~gd78059/webrev/nemo-stats-log
Hopefully its clear that the first review is the open one, and that you
have to be on SWAN to look at the second (closed) review.
I'm testing the mac layer changes, dmfe, bge, and rge right now. I'll
test e1000g, nge, nxge later. I cannot test easily ixgb or xge as I
don't have the requisite hardware. But the drivers compile cleanly,
lint clean, and the changes are obvious enough that I'm hoping I can
punt on actually testing the changes. (I think the worst case that is
likely to happen is that I may have missed a log entry somewhere, so one
of the untested drivers may still have extra log contents. I don't
_think_ even this is likely. I was pretty careful about checking.)
Note that a cursory analysis of the wifi drivers seems to indicate that
they don't log link status changes anyway, so they need no changes/no
testing. One upshot of this is that they will get link notifications in
the log where before they had none. But only if they report those link
status changes to the upper layers. (In other words, if NWAM/IPMP will
notice, then it will get logged.)
-- Garrett
PS: My apologies to those of you I BCC'd if you get 2 copies of this
message. I want the widest review possible, and I wasn't sure who reads
networking-discuss@ and who does not.
_______________________________________________
networking-discuss mailing list
[email protected]
