Hello All,
A colleague runs the included c++ program on his ARM system and gets times around 3.6 seconds. On my logicpd development board I'm getting about 16s. I know there are lots of variables here but I'm hoping to get an idea from the group as to whether we should expect more from the coldfire or not.
Thanks

The arm system:
# uname -a
Linux (none) 2.6.17 #7 Fri Jun 23 15:54:11 UTC 2006 armv4tl unknown
# cat /proc/cpuinfo
Processor : ARM920Tid(wb) rev 0 (v4l)
BogoMIPS : 88.47
Features : swp half thumb
CPU implementer : 0x41
CPU architecture: 4T
CPU variant : 0x1
CPU part : 0x920
CPU revision : 0
Cache type : write-back
Cache clean : cp15 c7 ops
Cache lockdown : format A
Cache format : Harvard
I size : 16384
I assoc : 64
I line length : 32
I sets : 8
D size : 16384
D assoc : 64
D line length : 32
D sets : 8


The logicpd coldfire system:
[root@localhost perftest]# uname -a
Linux localhost 2.4.26 #3 Tue May 2 11:58:17 EDT 2006 m68k unknown
[root@localhost perftest]# cat /proc/cpuinfo
CPU: MCF5484 Rev. 1
MMU: ColdFire V4e
FPU: ColdFire V4e
Clocking: 99.9MHz
BogoMips: 199.88
Calibration: 999424 loops

The benchmark:
#include <cstdio>
#include <cstdlib>
#include <string>
#include <list>

#include <sys/time.h>


static void fill_list(std::list<std::string> &slist)
{
char tmpbuf[256];

// Create string list with dupes
for (int i=0; i<2000; i++) {
sprintf(tmpbuf, "String number %d\n", rand()%500);
slist.push_back(tmpbuf);
}

// Remove duplicate entries
slist.unique();

// Sort the list
slist.sort();
}


int main()
{
std::list<std::string> slist;
struct timeval tv1, tv2;
float time;

srand(0);

gettimeofday(&tv1, NULL);

// Fill out the list then eliminate dupes 50 times
for (int i=0; i<50; i++)
fill_list(slist);

gettimeofday(&tv2, NULL);

time=(tv2.tv_sec-tv1.tv_sec)*1000000;
time+=(tv2.tv_usec-tv1.tv_usec);
time/=1000000;
printf("Total time: %.4f seconds\n", time);

return 0;
}