Re: cleaner/better zlib sources?

From: Josef Weidendorfer <josef.weidendorfer@gmx.de>
Date: Mar 16, 2007, 15:51 UTC
Message-ID: <200703161651.17478.Josef.Weidendorfer@gmx.de>
In-Reply-To: <20070316060841.19946.qmail@science.horizon.com>

On Friday 16 March 2007, linux@horizon.com wrote:

> Er, it's a little hard to see, but zlib spends the bulk of its time
> in inflate_fast().  This is the fast path, invoked as long as there's
> at least 6 bytes of input and 258 bytes of output space available.

Hmm...

When running "git log >/dev/null" a few times on my Pentium-M here, I get the following from oprofile:

CPU: PIII, speed 1600 MHz (estimated)
Counted CPU_CLK_UNHALTED events (clocks processor is not halted) with a unit mask of 0x00 (No unit mask) count 100000
Counted INST_RETIRED events (number of instructions retired) with a unit mask of 0x00 (No unit mask) count 400000
samples  %        samples  %        image name               app name                 symbol name
12816    19.6492  3221     19.0772  libz.so.1.2.3            git-log                  inflate_table
12186    18.6833  3640     21.5589  libz.so.1.2.3            git-log                  inflate
5228      8.0155  1395      8.2623  libz.so.1.2.3            git-log                  inflate_fast
3855      5.9104  738       4.3710  libc-2.5.so              git-log                  strlen
3782      5.7985  1361      8.0609  git-log                  git-log                  pretty_print_commit
2022      3.1001  600       3.5537  libc-2.5.so              git-log                  vfprintf

inflate_fast() is definitly below inflate().

Josef