[BusyBox] The pidof patch from the bug thing.

Bernhard Fischer rep.nop at aon.at
Sun Jun 5 10:48:58 UTC 2005


On Sat, Jun 04, 2005 at 11:43:03PM -0400, Rob Landley wrote:
>On Saturday 04 June 2005 11:00 pm, Rob Landley wrote:
>
>> +#ifdef CONFIG_FEATURE_PIDOF_SINGLE
>> +               if (!(opt & SINGLE))
>> +#endif
>> +#ifdef CONFIG_FEATURE_PIDOF_OMIT
>> +               if (!(opt & OMIT))
>> +#endif
>> +               bb_show_usage();
>>
>> And if they're both enabled...?
>
>Nevermind, that one makes a strange sort of sense now.

ok.

Wrt SINGLE ifdef: I thought to let the user decide whether -s was wanted
or not (same with -o) in order not to add functionality one might not
want to have compiled in unconditionally.
I was not aware that it is generally not desired to make code conditional
if it only adds an unspecified amount of bytes to busybox. If so, please
let me know what percentage / number of bytes are the margin,
approximately.

Wrt _..COMPL: They are prepended with an underscore to express that
they are only private helpers. Regarding the concatenation: yes, i'm
aware of that. Purpose is to build up a single string depending on which
features are enabled.

Where should bb_err_msg_and_die() be used? If you are referring to the
"illegal omit pid value" then i thought that warning about this should
not be necessary but would only make the object file bigger for no real
gain.
>
>I'll look at this patch again later.  Possibly clean it up myself...
TIA.
>
>Rob



More information about the busybox mailing list