On Tue, Mar 29, 2016 at 11:20 PM, Tom Lane <tgl@sss.pgh.pa.us> wrote:
> Christian Ullrich <chris@chrullrich.net> writes:
>> * Tom Lane wrote:
>>> But then, should not this code make sure that errno *always* gets set?
>
>> A library function that does not fail does not touch errno.
>
> Right, I meant "always in the error path".
>
>>> I'd be inclined to think we should use _dosmaperr(), too, rather than
>>> hand-coding it.
>
>> Yes, of course. If only I had known about it ...
>> New patch attached.
>
> This looks good, will push shortly.
Thanks for digging more than I did regarding this issue, things are
fixed on my side, and the buildfarm looks happy. Seeing EEXIST being
returned for a non-existing entry when calling link() really bugged
me, and the errno my patch looked at was set from another function
call and not link() itself...
--
Michael