Re: Squidfs code

http://www.bofh.net.au/~darius/squidfs/sfs-0.1.tgz contains the below fixes,
plus fixes for the mentioned problems - it no longer cores (or shouldn't),
-Wall produces no warnings, and several memory leaks have been shut off.
Next step is to go over the various interface routines, make sure they're
sane - now that my other major in-house project has settled down (and I've
found a house to live in, yay ;), I can sink my teeth into this properly
again.

More to come.

KevinL

>>> oskar@linux.org.za wrote
>
> --zhXaljGHf11kAtnf
> Content-Type: text/plain; charset=us-ascii
>
> Hi
>
> Attached is a bunch of cosmetic changes to the sfs code. (I know, I know,
> it's not really useful, but I am tired... getting rid of warnings etc is
> the most I can do right now ;)
>
> Includes:
>
> o A real makefile
> o Makefile includes linux and solaris sections. Defines _REENTRANT
> for Linux. You have to manually specify this.
> Currently linux just core-dumps with me immediately. I will try
> and track this down.
> o Compiles almost without warnings with 'gcc -Wall'. There are a
> couple of things that I was not up to fixing immediately: noted them
> with 'XXX'. I removed some unused variables (and commented out
> others): if they were 'for future use', sorry.
> o Now have $Id$ entries in every file, so that changes can be
> tracked.
> o Fixed recursive includes of header files
> o sfs_seek code doesn't match documentation. Kludged it's variable
> list in the meantime
> o replaced broken squid_curtime definition. was the same as saying
> 'exit + 1;'
> o Included headers to get rid of silly warnings
>
> Hopefully this makes the code more 'coder friendly'... I will actually
> work through the code and documentation sometime....
>
> Oskar
Received on Tue Jul 29 2003 - 13:15:56 MDT