>> [ proposed fix ]
>> As you can see, this is quite a bit more complicated than the simple
>> CreateProcess() call we have now.
>> ...
>> If this seems like a reasonable approach, I can see if I can get
>> something together. But it's a fairly large change..
>
>It sounds reasonable to me, in the sense that it is a localized change,
>even if rather messy. (Perhaps this chunk of code should be
>pushed into
>src/port someplace, instead of being dropped into postmaster.c?)
That's an idea. However, since this has to be tied in with the creationi
of the backend-parameter-file, it's going to have to hit
write_backend_variables() in some way, and have access to postmaster
variables. I think it might be even more messy to push it out.
Basically, I think internal_forkexec() needs to be split up into two -
one win32 and one other. For win32 version, it needs to CreateProcess()
*before* it does write_backend_variables(), and then pass the process id
as a parameter to write_backend_vars().
I think if it should be broken out, we need to break the whole package
of internal_forkexec, read/write_backend_vars, and probably
SubPostmasterMain() out somewhere. Not sure it's worth it.
//Magnus
