heimdal headers ordering (gssapi/gssapi.h, crypto-headers.h)

[Andrew Bartlett]

On Mon, 2022-03-28 at 12:53 +0300, Michael Tokarev via samba-technical
wrote:
> 28.03.2022 11:21, Andrew Bartlett wrote:
> > On Mon, 2022-03-28 at 11:13 +0300, Michael Tokarev wrote:
> > > It'd be very nice to have this fixed inside waf too :)
> 
> If not fixing it in waf, but at least waf can check if
> PYTHONHASHSEED is set and refuse to run if it is not.
> This will save a lot of grief like in that bugreport,
> like I encountered too, and like another bugreport
> against upstream heimdal with proposals to fix their
> headers.. :)

https://gitlab.com/samba-team/samba/-/merge_requests/2463

> > Just so you know how that works, like the other 'third_party'
> > components, to do that we would need to work it with upstream, then
> > import a new version.
> 
> Is waf also an external tool?  Okay... :)  Anyway, it can
> be patched in samba locally just to prevent this from
> happening. :)  Hopefully there will be less issues now
> once we release 4.16 in debian (and I hope to address
> this very issue in the next debian stable upload too).

So, waf is a tool to build build systems.  The core of waf is a
third_party library that we consume, and then we provide scripts that
run as part of it (the wscript files and scripts under wafsamba).

So some things we fix in our use of waf, others like how dependencies
are ordered etc we need to fix upstream first, lest we end up
maintaining our own version of waf forever.

Andrew,

-- 
Andrew Bartlett (he/him)       https://samba.org/~abartlet/
Samba Team Member (since 2001) https://samba.org
Samba Team Lead, Catalyst IT   https://catalyst.net.nz/services/samba

Samba Development and Support, Catalyst IT - Expert Open Source
Solutions