Hi Ed & Nils,
Everything looks reasonable to me.  Here's my few sentence summary of the
code I reviewed.  Please let me know if I got anything wrong.

Problem: Solaris lockd & statd need to run in the Linux zone for various
architectural reasons (covered in the design doc), but they also need to
perform some Solaris getservby* and getnameby* (and syslog) APIs using
Linux's name service switch/lookup code.

The solution for the getservby* and getnameby* are to intercept the
requests via a Solaris netconfig module that sits in the Linux zone,
packs up, and sends the get calls to the lx_thunker via a door so it can
perform the native Linux lookup and return the results, which are then
mapped back into Solaris structures and returned to lockd/statd.

I did not do a thorough code review, but everything looked straight
forward, and I don't anticipate any issues with the work we are doing
in sparks at this time.

Doug.


Edward Pilatowicz wrote:
> [ note that i'm not subscribed to this alias so please be sure to
>   include me directly on any replies. ]
> 
> hey all,
> 
> so when brandz recently went to psarc they noticed that we are delivering
> a new libnsl nametoaddr plugin library.  they asked us to contact you (the
> sparks project) to make sure that you are aware of what we're doing.
> that's this email.
> 
> back in feb (before the sparks project was public) i talked to Doug
> Leavitt and Michen Chang about what i was planning on doing and you
> guys seemed ok it.  since what i implemented has not changed from what
> we discussed back then i'm assuming you'll be ok with the changes.
> but just to make sure there are no surprises i'm sending this email
> as a reminder that brandz is planning on introducing a new libnsl
> nametoaddr plugin.  ;)
> 
> if you'd like to take another look at what we've done you can check out
> the nfs section of our design doc on opensolaris.org:
>       http://www.opensolaris.org/os/community/brandz/design/
> 
> specifically, the last paragraph in the following section:
>       3.8.1.4 Allowing lockd and statd to communicate with Linux ...
> 
> if you'd like to see our code you can check out a webrev at (note that this
> is an internal url):
>       http://jurassic.eng/net/antilock.sfbay/export/nils/codereview.1/webrev
> 
> if you do look at this webrev here are the netconfig(4) plugin library
> related files we're creating.
> 
>         usr/src/lib/brand/lx/netfiles/Makefile
>         usr/src/lib/brand/lx/netfiles/netconfig
>         usr/src/lib/brand/lx/lx_nametoaddr/Makefile
>         usr/src/lib/brand/lx/lx_nametoaddr/Makefile.com
>         usr/src/lib/brand/lx/lx_nametoaddr/amd64/Makefile
>         usr/src/lib/brand/lx/lx_nametoaddr/common/lx_nametoaddr.c
>         usr/src/lib/brand/lx/lx_nametoaddr/common/mapfile-vers
>         usr/src/lib/brand/lx/lx_nametoaddr/i386/Makefile
> 
> thanks
> ed
> _______________________________________________
> sparks-discuss mailing list
> sparks-discuss at opensolaris.org
> http://opensolaris.org/mailman/listinfo/sparks-discuss

Reply via email to