gwyn@smoke.BRL.MIL (Doug Gwyn) (07/07/89)
In article <4896@ficc.uu.net> peter@ficc.uu.net (Peter da Silva) writes: >Does anyone here really know what sort of crud V.4 (not V.3) is going to >have in it? Streams *and* sockets, or so I hear... AT&T presented an overview of SVR4 at a BOF session at the Baltimore USENIX. Therefore lots of people presumably now know what SVR4 will contain and more or less how it's implemented. Sockets are emulated; STREAMS is the basic kernel mechanism.
decot@hpisod2.HP.COM (Dave Decot) (07/19/89)
> >Does anyone here really know what sort of crud V.4 (not V.3) is going to > >have in it? Streams *and* sockets, or so I hear... > > AT&T presented an overview of SVR4 at a BOF session at the Baltimore USENIX. > Therefore lots of people presumably now know what SVR4 will contain and more > or less how it's implemented. I found a reference in a draft of SVID89 to a function called getcontext(BA_OS), but no entry in the book for it. Will such a function exist, and if so, what will it do? Dave Decot
gwyn@smoke.BRL.MIL (Doug Gwyn) (08/15/89)
In article <14020066@hpisod2.HP.COM> decot@hpisod2.HP.COM (Dave Decot) writes: >> AT&T presented an overview of SVR4 at a BOF session at the Baltimore USENIX. >I found a reference in a draft of SVID89 to a function called >getcontext(BA_OS), but no entry in the book for it. >Will such a function exist, and if so, what will it do? My notes from the BOF session say that getcontext() gets current user context (ucontext structure), which is described as being similar to BSD sigcontext: signal mask, machine context, current stack. As near as I can tell from the copy of the VuGraph, this is in addition to the additional siginfo structure that will be fed to three-argument signal handlers when SA_SIGINFO is enabled in the sigaction flag. Apparently you have to make the specific system call getcontext() to retrieve the ucontext information. There is also a setcontext() for the reverse. The notes did not include detailed interface information.