Re: Using XLogFileNameP in critical section - Mailing list pgsql-hackers

From Michael Paquier
Subject Re: Using XLogFileNameP in critical section
Date
Msg-id 20191204045523.GE2381@paquier.xyz
Whole thread Raw
In response to Re: Using XLogFileNameP in critical section  (Tom Lane <tgl@sss.pgh.pa.us>)
List pgsql-hackers
On Tue, Dec 03, 2019 at 11:24:57AM -0500, Tom Lane wrote:
> So it seems that the problem might really be a faulty rule in our
> MSVC build script about when postgres.def needs to be regenerated?
> Or else it's some weird caching problem on drongo --- the lack of
> complaints from other Windows critters might point the finger
> that way.

Yes, I saw the failure from the buildfarm logs, but I got to the
conclusion that the animal just got crazy with a portion of its
caching because there are no more references to routine removed.  So I
did not bother much.

FWIW, I have seen sometimes similar warnings regarding conflicting
projects when doing a compilation on MSVC, applying a patch and then
recompiling (because that's just faster than recompiling the whole).
Perhaps we could do better here.  I am not completely sure what
though, it's not a problem I have spent much brain-time on.
--
Michael

Attachment

pgsql-hackers by date:

Previous
From: Kyotaro Horiguchi
Date:
Subject: Re: [HACKERS] WAL logging problem in 9.4.3?
Next
From: Amit Kapila
Date:
Subject: Re: Windows buildfarm members vs. new async-notify isolation test