Re: port conflicts when running tests concurrently on windows.

On 13.12.21 20:33, Andres Freund wrote:
>> pg_regress.c at remove_temp() is still there. These things should probably
>> be addressed before we can consider making this the default.
>
> Hm, not immediately obvious what to do about this. Do you know if windows has
> restrictions around the length of unix domain sockets? If not, I wonder if it
> could be worth using the data directory as the socket path on windows instead
> of the separate temp directory?

According to src/include/port/win32.h, it's the same 108 or so bytes
that everyone else uses. That might not be the kernel limit, however,
just the way the external API is defined.

After the reading the material again, I think that comment might be
overly cautious. The list of things not to do in a signal handler on
Windows is not significantly different than those for say Linux, yet we
do a lot of them anyway. I'm tempted to just ignore the advice and do
it anyway, while ignoring errors, which is what it already does.