sfio97 vs sfio2000

I am in the process of making the major overhaul of perlio.h and its
descendants. To check I had not broken sfio I did a build against 
the one I have here - shich happens to be sfio97.

I discovered _my_ changes had not broken it any worse than it was.
The recent change away from sftell() does _NOT_ work right with sfio97.
As it was only to give error rather than something that _might_ be useful
I have reverted that to use sftell().

Also the perlsfio.h that I inherited seems to have turned off 
PerlIO_canset_cnt() - and sfio _can_ set count same way that Linux/glibc
stdio can - set the pointer and count comes too.
It was also broken in that there was a FILE_cnt() macro being used...

With all those tweaked a -Duseperlio -Dusesfio build passes all tests
except for one block of lib/syslfs.t tests with sfio97. 

The block of tests which fail in lib/syslfs.t follow a seek() rather
than a sysseek() - sfio97 is not large file aware so that fails.
It is not clear to me if use of seek() vs sysseek() is a typo 
in the test of if it is trying to prove buffered/unbuffered lfs seeks
interoperate (which seems suspect to me...)

-- 
Nick Ing-Simmons <nik@tiuk.ti.com>
Via, but not speaking for: Texas Instruments Ltd.

• sfio97 vs sfio2000 by Nick Ing-Simmons
• RE: sfio97 vs sfio2000 by Redford, John
• Re: sfio97 vs sfio2000 by Nicholas Clark
• Re: sfio97 vs sfio2000 by Nick Ing-Simmons
• Re: sfio97 vs sfio2000 by Jarkko Hietaniemi