On 12 Dec 2014 17:15, "Martin Nowak via Digitalmars-d" < digitalmars-d@puremagic.com> wrote: > > On 12/12/2014 04:47 PM, Joakim wrote: >> >> I asked about this on github but didn't get a good answer, so I'm asking >> here. What's with all the repeated OS blocks in druntime? > > > No, you don't want to accept the answer. That's slightly different than not getting none. > > >> https://github.com/D-Programming-Language/druntime/blob/master/src/core/sys/posix/unistd.d#L945 >> >> https://github.com/D-Programming-Language/druntime/blob/master/src/core/sys/posix/netinet/in_.d#L974 >> >> https://github.com/D-Programming-Language/druntime/blob/master/src/core/sys/posix/arpa/inet.d#L201 >> >> >> It seems like pointless repetition and there are many more examples of >> this, as I keep running into it when adding bionic/Android support. >> Martin suggested that it would be useful to have a default case that >> asserts for an unsupported OS, but many of these blocks don't have that >> either. > > > Because it was written before we spread out to multiple architectures, and learned hard it is to add something. > > >> >> Why not just declare them once for Posix and then specialize for a >> particular OS later on when necessary, as seems to have been done in >> these instances? > > > We've been through this several times, because the poor guy adding support for a new OS or arch has the find all the differences through debugging. > >
One of the worst I've hit was a crash deep in the start-up initialisation of Druntime... caused by a totally not obvious struct size/align mismatch with Cruntime's pthread. Along with not wasting a day switching between druntime and system C headers, having a compile-time error ensures that the porter has vetted the correctness of the ported declarations and structures. It may take a while to port all areas, but at least you can be mostly assured that each step is towards a fully working runtime. Iain.