Re: POSIX::sigprocmask implemented incorrectly

Rafael Garcia-Suarez wrote: I've a patch - see attached. I've also replaced all the occurences of sv_isa with sv_derived_from so that inheritance will work properly for POSIX::SigAction and POSIX::SigSet, and I've changed the behaviour of sigprocmask so that if the third argument (oldsigset) is

Re: Patch acceptance/rejection (Win32)

H.Merijn Brand wrote: >On Tue 27 Jul 2004 17:26, Artiom <[EMAIL PROTECTED]> wrote: > > >>Hello All, >> >> Some time ago i've posted a problem report and a patch for it. The >>problem is not unsual (network operations failures when LSP (lost of >>antiviruses, firewalls etc) installed). How

Re: is there an API to check whether it's safe to snatch SV away?

Nick Ing-Simmons <[EMAIL PROTECTED]> wrote: :Stas Bekman <[EMAIL PROTECTED]> writes: :> :>OK, so are there any other cases where I have to do the copy? : :Yes, ALL other cases, bar none. As soon as your XS returns, or :calls back into perl (or calls PerlIO or ...) the value can be toasted. : :Arg