[KJ] another cleanup TO DO: is_power_of_2() check

[KJ] another cleanup TO DO: is_power_of_2() check

[Robert P. J. Day]

  there are *numerous* places in the code that check if some value is
a power of 2 using the basic test "((x & (x - 1)) = 0)":

  $ grep -Er "(.*) ?\& ?\(\1 ?- ?1\)" .

as you can see, that test is all over the place.

  occasionally, someone decides to do the right thing and define a
macro:

  $ grep -r "#define is_power_of_2" .

i'm thinking it might be cleaner to define that macro in kernel.h and
clean up the source to use that macro, rather than constantly
rewriting that expression.

  thoughts?

rday
_______________________________________________
Kernel-janitors mailing list
Kernel-janitors@lists.osdl.org
https://lists.osdl.org/mailman/listinfo/kernel-janitors

Re: [KJ] another cleanup TO DO: is_power_of_2() check

[Frederik Deweerdt]

On Mon, Jan 08, 2007 at 05:37:16PM -0500, Robert P. J. Day wrote:
> 
>   there are *numerous* places in the code that check if some value is
> a power of 2 using the basic test "((x & (x - 1)) = 0)":
> 
>   $ grep -Er "(.*) ?\& ?\(\1 ?- ?1\)" .
> 
> as you can see, that test is all over the place.
> 
>   occasionally, someone decides to do the right thing and define a
> macro:
> 
>   $ grep -r "#define is_power_of_2" .
Some of which really are is_power_of_2_but_not_0 :). That corner case
set aside, but it is handled differently throughout the source, it seems
fine to me.

Regards,
Frederik
> 
> i'm thinking it might be cleaner to define that macro in kernel.h and
> clean up the source to use that macro, rather than constantly
> rewriting that expression.
> 
>   thoughts?
> 
> rday
> _______________________________________________
> Kernel-janitors mailing list
> Kernel-janitors@lists.osdl.org
> https://lists.osdl.org/mailman/listinfo/kernel-janitors
> 
_______________________________________________
Kernel-janitors mailing list
Kernel-janitors@lists.osdl.org
https://lists.osdl.org/mailman/listinfo/kernel-janitors

Re: [KJ] another cleanup TO DO: is_power_of_2() check

[Robert P. J. Day]

On Mon, 8 Jan 2007, Frederik Deweerdt wrote:

> On Mon, Jan 08, 2007 at 05:37:16PM -0500, Robert P. J. Day wrote:
> >
> >   there are *numerous* places in the code that check if some value is
> > a power of 2 using the basic test "((x & (x - 1)) = 0)":
> >
> >   $ grep -Er "(.*) ?\& ?\(\1 ?- ?1\)" .
> >
> > as you can see, that test is all over the place.
> >
> >   occasionally, someone decides to do the right thing and define a
> > macro:
> >
> >   $ grep -r "#define is_power_of_2" .

> Some of which really are is_power_of_2_but_not_0 :).

yes, i had noticed that.

> That corner case set aside, but it is handled differently throughout
> the source, it seems fine to me.

*what* seems fine?  making the change, or leaving it as is?

rday
_______________________________________________
Kernel-janitors mailing list
Kernel-janitors@lists.osdl.org
https://lists.osdl.org/mailman/listinfo/kernel-janitors

Re: [KJ] another cleanup TO DO: is_power_of_2() check

[Frederik Deweerdt]

> *what* seems fine?  making the change, or leaving it as is?
> 
Making the change.

Regards,
Frederik
_______________________________________________
Kernel-janitors mailing list
Kernel-janitors@lists.osdl.org
https://lists.osdl.org/mailman/listinfo/kernel-janitors