rsync server over SSH [includes code patches]

[Martin Pool]

On 28 Nov 2001, Dave Dykstra <dwd at bell-labs.com> wrote:
> Yes, I had seen it but haven't volunteered my build machines because the
> build servers download code to execute over the open internet using only 
> an rsync server, which gives no assurance that it hasn't been spoofed.
> Hopefully after the rsync server over SSH patch is integrated we can use
> that and I won't worry any more.

Yes, that would be an ideal use for it.

You could probably hack it up now using your dwd at samba.org account
by modifying the build scripts if you wanted.

>     1. remove AC_FUNC_MEMCMP from configure.in because it causes Sunos
> 	4.1.4 to die.  memcmp() there fails the "8-bit clean" test, but it
> 	doesn't matter because memcmp() in rsync is only used to test if
> 	something is equal or not equal, not to compare if it is less or
> 	greater.  

What breaks if we leave it in?  Is memcmp a compiler intrinsic or something 
like that?

No objection to taking it out, I'm just curious.

>       AC_FUNC_MEMCMP has always been in configure.in but before
> 	now it has been ignored because @LIBOBJS@ was not used in
> 	Makefile.in now and that was the only thing it affected.

I think it is, isn't it?  LIBOBJS = ... @LIBOBJS@, etc.

>     2. Amdahl's UTS 2.1.2 defines NO_ADDRESS instead of NO_DATA.
> 
>     3. Sunos 4.1.4 and Unixware 1.1.2 don't have stderr defined in unistd.h.

I thought it was meant to be in stdio.h?  OK.

> 	Note: I see that rsync.h has an #ifdef HAVE_UNISTD_H around including
> 	unistd.h, perhaps it would be better to include rsync.h instead of
> 	most of the includes at the beginning of getaddrinfo.c

Right.

>     4. Unixware 1.1.2 and UTS 2.1.2 had missing S_ISLNK and S_ISSOCK
> 	macros which are defined in rsync.h.  Maybe all the lib files
> 	should just include rsync.h.

Yes, I did that in (I think) getnameinfo/getaddrinfo, but we should do it 
in the others as well.

That's great, thankyou.

-- 
Martin