cachegrind output help

[John]

Can someone help me decipher what the output of cachegrind
means. Is this "c" code good or bad. What can I infer from
this output? The source code is too long to post here.

Does the l1 and L2i miss rate say that my code is very cache
friendly?

Thanks,
--j

==3978==
==3978== I refs: 193,569,732,574
==3978== I1 misses: 1,460,592,844
==3978== L2i misses: 48,987
==3978== I1 miss rate: 0.75%
==3978== L2i miss rate: 0.00%
==3978==
==3978== D refs: 128,486,842,763 (69,275,893,182 rd +
59,210,949,581 wr)

==3978== D1 misses: 260,193,784 ( 194,255,891 rd +
65,937,893 wr)

==3978== L2d misses: 2,891,578 ( 2,512,228 rd +
379,350 wr)

==3978== D1 miss rate: 0.2% ( 0.2% +
0.1% )

==3978== L2d miss rate: 0.0% ( 0.0% +
0.0% )

==3978==
==3978== L2 refs: 1,720,786,628 ( 1,654,848,735 rd +
65,937,893 wr)

==3978== L2 misses: 2,940,565 ( 2,561,215 rd +
379,350 wr)

==3978== L2 miss rate: 0.0% ( 0.0% +
0.0% )

 

[Flash Gordon]

Can someone help me decipher what the output of cachegrind
means. Is this "c" code good or bad. What can I infer from
this output? The source code is too long to post here.

<snip>

I'm sure someone could, but not on comp.lang.c (I don't know about
gnu.gcc.help). Here on comp.lang.c we only deal with the C program
language, not the may tools and processors it is used on. For instance,
I have done a lot of C programming for processors that do not have any
cache. I suggest you ask on a Linux group since cachgrind is a Linux tool.