Re: BUG #15804: Assertion failure when using logging_collector with EXEC_BACKEND

Andres Freund <andres@anarazel.de>

From: Andres Freund <andres@anarazel.de>
To: Michael Paquier <michael@paquier.xyz>, Peter Eisentraut <peter.eisentraut@2ndquadrant.com>
Cc: Tom Lane <tgl@sss.pgh.pa.us>, Yuli Khodorkovskiy <yuli.khodorkovskiy@crunchydata.com>, pgsql-bugs@lists.postgresql.org
Date: 2019-05-17T20:24:43Z
Lists: pgsql-bugs
Hi,

Peter, it'd be really good if you could chime in here PDQ.


On 2019-05-17 13:25:36 +0900, Michael Paquier wrote:
> This may sound rough, but based on the timing and the time I can spend
> on that, I really think that we should revert 57431a9, as this has not
> been completely thought through for EXEC_BACKEND, and the patch I am
> attaching is a set of rough counter-measures which keep piling up,
> giving a result I am not comfortable with at all.  I am not sure that
> I could actually go to the bottom of it without reworking the way we
> attach, detach and reattach to shared memory a process, particularly
> on Windows where we don't actually need to allocate a region for a
> process we know will never use shared memory (right?).  And this is no
> time to do so for v12.

I'm not sure I see this as sufficient reason to revert.


> Last year I did the same kind of mistake by
> underestimating the Windows-related stuff for v11 just before a
> release which has resulted in an urgent revert as of df8b5f3.  I don't
> want to do the same mistake again.

I don't think a windows-only patch is the same thing as a feature, one
that makes debugging easier to boot, that's applicable to all platforms.



> diff --git a/src/tools/msvc/config_default.pl b/src/tools/msvc/config_default.pl
> index d7a9fc5039..de070e26e5 100644
> --- a/src/tools/msvc/config_default.pl
> +++ b/src/tools/msvc/config_default.pl
> @@ -3,7 +3,7 @@ use strict;
>  use warnings;
>
>  our $config = {
> -	asserts => 0,    # --enable-cassert
> +	asserts => 1,    # --enable-cassert
>  	     # float4byval=>1,         # --disable-float4-byval, on by default
>
>  	# float8byval=> $platformbits == 64, # --disable-float8-byval,
> @@ -17,7 +17,7 @@ our $config = {
>  	gss       => undef,    # --with-gssapi=<path>
>  	icu       => undef,    # --with-icu=<path>
>  	nls       => undef,    # --enable-nls=<path>
> -	tap_tests => undef,    # --enable-tap-tests
> +	tap_tests => 1,    # --enable-tap-tests
>  	tcl       => undef,    # --with-tcl=<path>
>  	perl      => undef,    # --with-perl
>  	python    => undef,    # --with-python=<path>

Hm? Is this to coerce cfbot?

Greetings,

Andres Freund



Commits

  1. Rearrange postmaster's startup sequence for better syslogger results.

  2. Improve logrotate test so that it meaningfully exercises syslogger.

  3. Revert "postmaster: Start syslogger earlier".

  4. postmaster: Start syslogger earlier