Hello Marcel,

On Wed, 30 Oct 2002, Marcel Pol wrote:

> Hello,
> I started using codafs very recently, and I noticed it only builds with
> gcc-2.9x.

I cannot give you any guarantee of course,
but I am running recent Coda (cvs, later than 5.3.19) and using gcc 3.2.
So far I have no good reason to believe gcc-3.2 breaks the code, as when I
wanted sometimes to exclude it as a reason for errors, I recompiled with
2.95.4 and got the same behavior pattern.

It does not mean the compile is bug-free, but the possible bugs are well
hidden :)

I have seen similar - while still different - compile problems once and
had to make small changes to the code, but my recent compile tree does not
contain such fixes, that means they have been fixed in the cvs repository.

I think most of gcc-3.2-related issues have been sorted out,
at least in the development branch, or I am extremely lucky :-)

> gcc-3.2-2mdk
> gcc-c++-3.2-2mdk
> glibc-2.2.5-16mdk
> kernel-headers-2.4.18-41mdk

My setup is (all compiled from source) :
Linux 2.4.19
gcc-3.2
compile time glibc 2.2.4 or 2.2.5 (unsure)
coda runtime glibc 2.2.5

In other words, for your question about future Coda versions, it should be
ok to compile with 3.2.
Any incompatibility with 3.2 if not a gcc bug (that means - when Coda
really violates C specification) is essentially a Coda bug and as such
will be fixed surely.

> The message I got was that even if the build errors are fixed, there
> might be other problems which could cause data corruption.

Sure, there is no safety in this world :-/ but 3.2 looks much better that
3.1, and I think Jan has fixed a lot in Coda since that.

Best regards,
--
Ivan