[parted-devel] [Parted-edge] [RFC][PATCH] Remove alloca from libparted/arch/linux.c

[Jim Meyering]

"Frodo Baggins" <frodo.drogo at gmail.com> wrote:

> On 9/6/07, Jim Meyering <jim at meyering.net> wrote:
>> Otavio Salvador <otavio at debian.org> wrote:
>> > Jim Meyering <jim at meyering.net> writes:
>> >
>> >> I think it would be worthwhile (long term) to convert the other way,
>> >> e.g., ped_free -> free, ped_malloc -> malloc, etc.
>> >
>> > But wouldn't the exception handling be a good thing in this case?
>>
>> ?!?  This sort of "exception handling" is never a good thing.
>> Its use makes for ugly, unmaintainable code.
>>
>> One problem is that what libparted calls "exception handling"
>> is an ugly kludge that can make the library code produce output
>> and/or try to read an interactive reply from the user
>> (things libraries must not do, in general).  In addition,
>> how the handler behaves depends on static state,
>> which is another big no-no for a library.
>
> Do we have a state machine which could help decide what to do when an
> "exception" occurs?

I'm not sure what you mean.
The problem isn't deciding what to do, but that there
is even an opportunity to decide.  Adding state and acting
on it would just make things worse.  Libraries are supposed
to be stateless.

IMHO, when something exceptional happens in a library, the
affected function should simply return with an indication of
what went wrong.  Unfortunately, making this particular
library clean would be a big and risky task.