Front page | perl.perl5.porters |
Postings from January 2001
On Fri, Jan 26, 2001 at 02:40:05PM -0600, Jarkko Hietaniemi wrote:
> > It depends on your platform--some have threadsafe gethost* calls and don't
> > need the _r version, and some don't. (And some aren't threadsafe but don't
> > have _r versions...)
>
> And don't foolishly think that there is only one set of _r variants
> out there. There are several different signatures for getpwnam_r,
> for example.
>
> > It's pretty much a crapshoot, unfortunately.
>
> Alan would drop -shoot :-)
Alan has expressed his feelings on C++, Perl's "API" and the implementation
of threads most everywhere. Dan doesn't like C, and I'm starting to agree.
Jarkko and I both despise the standards committee for long long in C99
(but I think I start to understand how a standards committee works, and hence
can see why it happens)
Is there any language we actually all like? :-)
> >From Todo:
>
> the _r-problem: for all the {set,get,end}*() system database
> calls (and a couple more: readdir, *rand*, crypt, *time,
> tmpnam) there are in many systems the _r versions
> to be used in re-entrant (=multithreaded) code
> Icky things: the _r API is not standardized and
> the _r-forms require per-thread data to store their state
So we've traded the insecurity of csh globbing for tread non-safety
of readdir calls in File::Glob?
[I don't have a problem with this - I think it's the right choice. But I
like to know what assumptions are underlying things]
Is it worth adding a question to Configure, default to [n], to ask if
the gethostent() is thread safe?
And if not, conditionally protect all the pp ops with a mutex if threading
is compiled in?
Or are there more important things to put effort into?
[like doing the whole thing properly for perl6]
Nicholas Clark
Thread Previous
|
Thread Next