Re: PCI Express support for 2.4 kernel

[Kevin P. Fleming]

Vladimir Kondratiev wrote:

> To illustrate zero cost, I did the following test:
> [tmp]$ cat t.c; gcc -S t.c; cat t.s
> static int a1=0;
> static int a2;
> /* EOF */
>
>    .file    "t.c"
>    .local    a1
>    .comm    a1,4,4
>    .local    a2
>    .comm    a2,4,4
>    .section    .note.GNU-stack,"",@progbits
>    .ident    "GCC: (GNU) 3.3.1 20030811 (Red Hat Linux 3.3.1-1)"
>
> As you can see, assembly code is identical, compiler did this trivial 
> optimization for me.

You've missed the point, though. Initializing a static variable to zero 
causes space to be consumed in the resulting object file (not 
instruction code to be generated). This is wasted space, because if you 
don't initialize to zero the variable will be allocated out of space 
that is _automatically_ zeroed for you. This reduces the size of the 
kernel image by not filling it with unnecessary zeroes.

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/