Re: [Botan-devel] ...and a little help on the mac as well

[Jack L. <ll...@ac...>]

On Thu, 5 Jun 2003, Jeff B wrote:

> If you read my previous message, you'll know that I'm trying to build Botan
> 1.2.4 on Windows using VS .NET 2003.  Now, I'm trying to build Botan 1.2.4
> on the Macintosh (for MacOS X) using Codewarrior 8.  Here's the scoop:
>
> 1. I used perl configure.pl gcc-aix-ppc to set up my configuration.  Does
> this seem right?  I know that the mac & ppc aren't supported but
> http://botan.randombit.net/ states that Botan cam be built for MacOS X
> (albeit using gcc).

The proper string would be <something>-macosx-ppc. There is no explicit
support for using Codewarrior because I don't have access to it. But OS X
is supported, as are PPC machines (OS X, Linux, and *BSD).

> 3. The project compiles and links.  When I run the project, I pass main()
> "512 ThisIsASamplePassphrase", but
>     LibraryInitializer init;
> Fails.

How so? Does an exception get thrown from somewhere?

> when j=0 and we try
>       std::ifstream random_source(sources[j].c_str());
> Sources[0] is "/dev/urandom" and std::ifstream() is failing
>
> Likewise, when j=1
> Sources[0] is "/dev/random" and std::ifstream() is failing

That is bizarre - OS X should have both of those files (at least every
version I have ever seen does).

> Any suggestions on what to do?

First, make sure the files /dev/random and /dev/urandom actually exist.
Open up a Terminal window, and type:

ls -l /dev/*random

Which should produce something along the lines of:

crw-r--r--    1 root     wheel      7,   0 Jun  6 14:35 /dev/random
crw-r--r--    1 root     wheel      7,   1 May  4 12:01 /dev/urandom

In particular, make sure that first bit of "crw-r--r--" is the same. Then,
make sure they work, by typing:

hexdump -n 8 -C /dev/urandom

which should produce something like:

00000000  c1 eb 1f 5d 78 19 c0 85                           |...]x...|
00000008

though the specific byte values will presumably be different.

If the above all checks out, then I'll need to get more information about
exactly what kind of failure occurs.

-Jack