If the changes truely are easy to review and fairly contained I have no objections in fixing this in 7.4. Care to make a patch tar?
BTW, it would be nice to have some... uhm... preprocessor constant (or anything else) to determine native int size (and perhaps float size). I couldn't find how to do this (except of hack with sprintf("%u", ~0)
- but this is ugly, IMHO).
I totally agree. You should be able to fix it by adding a new preprocessor symbol in cpp.c.
/ Martin Nilsson (saturator)
Previous text:
2003-09-20 01:42: Subject: Re: bug: casts to string for long double and long long int are incorrect (7.4.28 rel)
On Sat, Sep 20, 2003 at 01:05:02AM +0200, Martin Nilsson (saturator) @ Pike (-) developers forum wrote:
feasible to backport all that to make longer ints and doubles work flawlessly in 7.4". Are you of another opinion, and if so, why?
Well... Currently, so far all the problems with long longs are related to testsuite, which is, in turn, dependent on casting to string.
64 bits int are useful in many apps which otherwise would require Gmp - it includes byte counters, etc. May be it is a bit specific to my projects (log processing, traffic accounting in particular), but I see no real problems to fix all this stuff in 7.4 too, and I don't think that this feature might be as unstable as something more sophisticated (on some architectures native int size is 64 bits anyway, or this arch support isn't planned/available in 7.4?).
BTW, it would be nice to have some... uhm... preprocessor constant (or anything else) to determine native int size (and perhaps float size). I couldn't find how to do this (except of hack with sprintf("%u", ~0) - but this is ugly, IMHO).
Regards, /Al
/ Brevbäraren