[Patch] 2.5.70-bk11 zlib reduce deflate workspace by 128k

[Patch] 2.5.70-bk11 zlib reduce deflate workspace by 128k

[Jörn Engel]

Hi Linus!

Free lunch time.  There are currently no users in the kernel that use
a memlevel >8, but we reserve space for 9.  The patch saves 128k at no
cost.

Jörn

-- 
The cost of changing business rules is much more expensive for software
than for a secretaty.
-- unknown

--- linux-2.5.70-bk11/include/linux/zconf.h~zlib_memlevel	2003-06-06 15:56:15.000000000 +0200
+++ linux-2.5.70-bk11/include/linux/zconf.h	2003-06-06 15:57:28.000000000 +0200
@@ -23,7 +23,7 @@
 
 /* Maximum value for memLevel in deflateInit2 */
 #ifndef MAX_MEM_LEVEL
-#  define MAX_MEM_LEVEL 9
+#  define MAX_MEM_LEVEL 8
 #endif
 
 /* Maximum value for windowBits in deflateInit2 and inflateInit2.

Re: [Patch] 2.5.70-bk11 zlib reduce deflate workspace by 128k

[Jörn Engel]

On Fri, 6 June 2003 17:00:43 +0200, art_k@o2.pl wrote:
> Jörn Engel <joern@wohnheim.fh-wedel.de> (030606.160149,<20030606140149.GA20168@wohnheim.fh-wedel.de>):
> > Free lunch time.  There are currently no users in the kernel that use
> > a memlevel >8, but we reserve space for 9.  The patch saves 128k at no
> > cost.
> 
> >  /* Maximum value for memLevel in deflateInit2 */
> >  #ifndef MAX_MEM_LEVEL
> > -#  define MAX_MEM_LEVEL 9
> > +#  define MAX_MEM_LEVEL 8
> >  #endif
> 
> i like the direction this is going, but when making these kind of changes
> it would be even better if the reasoning would be noted in a short comment
> (eg. "/* no users of memlevel>8 in kernel; saves 128k RAM */").
> That way the next person looking at that code would immediately know why
> that number was chosen.

Usually, you would be right.  But in this case, the existing comments
in zconf.h match the situation *after* my patch.  Looks good to me.

Jörn

-- 
But this is not to say that the main benefit of Linux and other GPL
software is lower-cost. Control is the main benefit--cost is secondary.
-- Bruce Perens