pgsql: Use --disable-auto-import linker switch in Mingw builds, too.

On 02/12/2014 12:04 PM, Tom Lane wrote:

Use --disable-auto-import linker switch in Mingw builds, too.

This is evidently the default on buildfarm member narwhal, but that
is a pretty ancient Mingw version, and there is reason to think that
more recent versions of GNU ld have this feature turned on by default.
Since we are trying to achieve consistency of link behavior across
all Windows toolchains, let's just make sure here.

This appears to have broken configure on jacana (probably the most
modern Mingw animal we have).

cheers

andrew

--
Sent via pgsql-committers mailing list (pgsql-committers@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-committers

On 2014-02-14 19:44:11 -0500, Andrew Dunstan wrote:

On 02/12/2014 12:04 PM, Tom Lane wrote:

Use --disable-auto-import linker switch in Mingw builds, too.

This is evidently the default on buildfarm member narwhal, but that
is a pretty ancient Mingw version, and there is reason to think that
more recent versions of GNU ld have this feature turned on by default.
Since we are trying to achieve consistency of link behavior across
all Windows toolchains, let's just make sure here.

This appears to have broken configure on jacana (probably the most modern
Mingw animal we have).

According to a quick search that's caused by a bug in mingw...

http://sourceforge.net/p/mingw/bugs/2020/ and some others.

Any chance you could update to a slighly newer version?

Greetings,

Andres Freund

--
Andres Freund http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services

--
Sent via pgsql-committers mailing list (pgsql-committers@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-committers

On 02/14/2014 08:18 PM, Andres Freund wrote:

On 2014-02-14 19:44:11 -0500, Andrew Dunstan wrote:

On 02/12/2014 12:04 PM, Tom Lane wrote:

Use --disable-auto-import linker switch in Mingw builds, too.

This is evidently the default on buildfarm member narwhal, but that
is a pretty ancient Mingw version, and there is reason to think that
more recent versions of GNU ld have this feature turned on by default.
Since we are trying to achieve consistency of link behavior across
all Windows toolchains, let's just make sure here.

This appears to have broken configure on jacana (probably the most modern
Mingw animal we have).

According to a quick search that's caused by a bug in mingw...

http://sourceforge.net/p/mingw/bugs/2020/ and some others.

Any chance you could update to a slighly newer version?

This is not the mingw compiler, it's the mingw-w64 compiler, a subtle
difference.

I installed the compiler from
<http://sourceforge.net/projects/mingw-w64/files/Toolchains%20targetting%20Win64/Personal%20Builds/rubenvb/gcc-4.8-release&gt;
and it got over this hurdle. Interestingly it says it's gcc 4.8.0, while
the failing version claimed to be 4.8.1, even though this one is dated
quite a lot later.

However, the build now fails like this:

make[1]: Entering directory
`/home/pgrunner/bf/root/HEAD/pgsql.6024/contrib/pg_buffercache'
x86_64-w64-mingw32-gcc -O2 -Wall -Wmissing-prototypes
-Wpointer-arith -Wdeclaration-after-statement -Wendif-labels
-Wmissing-format-attribute -Wformat-security -fno-strict-aliasing
-fwrapv -fexcess-precision=standard -g -I.
-I/home/pgrunner/bf/root/HEAD/pgsql.6024/../pgsql/contrib/pg_buffercache
-I../../src/include
-I/home/pgrunner/bf/root/HEAD/pgsql.6024/../pgsql/src/include
-I../pgsql/src/include/port/win32 -DEXEC_BACKEND
-I/c/prog/3p64/include/libxml2 -I/c/prog/3p64/include
-I/c/prog/3p64/openssl/include
"-I/home/pgrunner/bf/root/HEAD/pgsql.6024/../pgsql/src/include/port/win32"
-c -o pg_buffercache_pages.o
/home/pgrunner/bf/root/HEAD/pgsql.6024/../pgsql/contrib/pg_buffercache/pg_buffercache_pages.c
x86_64-w64-mingw32-gcc -O2 -Wall -Wmissing-prototypes
-Wpointer-arith -Wdeclaration-after-statement -Wendif-labels
-Wmissing-format-attribute -Wformat-security -fno-strict-aliasing
-fwrapv -fexcess-precision=standard -g -shared -o
pg_buffercache.dll pg_buffercache_pages.o -L../../src/port
-L../../src/common -Wl,--allow-multiple-definition
-Wl,--disable-auto-import -L/c/prog/3p64/lib -L/c/prog/3p64/lib
-L/c/prog/3p64/openssl/lib/VC -Wl,--as-needed -L../../src/backend
-lpostgres -lpgcommon -lpgport -lintl -lxslt -lxml2 -lz -lwsock32
-lm -lws2_32 -lshfolder -Wl,--export-all-symbols
-Wl,--out-implib=libpg_buffercache.a
pg_buffercache_pages.o: In function `pg_buffercache_pages':
c:/MinGW/msys/1.0/home/pgrunner/bf/root/HEAD/pgsql.6024/../pgsql/contrib/pg_buffercache/pg_buffercache_pages.c:119:
undefined reference to `MainLWLockArray'
c:/MinGW/msys/1.0/home/pgrunner/bf/root/HEAD/pgsql.6024/../pgsql/contrib/pg_buffercache/pg_buffercache_pages.c:160:
undefined reference to `MainLWLockArray'
collect2.exe: error: ld returned 1 exit status
make[1]: *** [pg_buffercache.dll] Error 1
make[1]: Leaving directory
`/home/pgrunner/bf/root/HEAD/pgsql.6024/contrib/pg_buffercache'

Note that by this time "make check" has already passed.

cheers

andrew

Greetings,

Andres Freund

--
Sent via pgsql-committers mailing list (pgsql-committers@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-committers

On 2014-02-15 09:47:08 -0500, Andrew Dunstan wrote:

However, the build now fails like this:
pg_buffercache_pages.o: In function `pg_buffercache_pages':
c:/MinGW/msys/1.0/home/pgrunner/bf/root/HEAD/pgsql.6024/../pgsql/contrib/pg_buffercache/pg_buffercache_pages.c:119:
undefined reference to `MainLWLockArray'
c:/MinGW/msys/1.0/home/pgrunner/bf/root/HEAD/pgsql.6024/../pgsql/contrib/pg_buffercache/pg_buffercache_pages.c:160:
undefined reference to `MainLWLockArray'
collect2.exe: error: ld returned 1 exit status
make[1]: *** [pg_buffercache.dll] Error 1
make[1]: Leaving directory
`/home/pgrunner/bf/root/HEAD/pgsql.6024/contrib/pg_buffercache'

Note that by this time "make check" has already passed.

That's expected atm, that's breaking all the windows builds right
now. Now that brolga is alive again, it should be able to progress past
this... How often does it build master?

Greetings,

Andres Freund

--
Andres Freund http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services

--
Sent via pgsql-committers mailing list (pgsql-committers@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-committers

On 02/15/2014 09:49 AM, Andres Freund wrote:

On 2014-02-15 09:47:08 -0500, Andrew Dunstan wrote:

However, the build now fails like this:
pg_buffercache_pages.o: In function `pg_buffercache_pages':
c:/MinGW/msys/1.0/home/pgrunner/bf/root/HEAD/pgsql.6024/../pgsql/contrib/pg_buffercache/pg_buffercache_pages.c:119:
undefined reference to `MainLWLockArray'
c:/MinGW/msys/1.0/home/pgrunner/bf/root/HEAD/pgsql.6024/../pgsql/contrib/pg_buffercache/pg_buffercache_pages.c:160:
undefined reference to `MainLWLockArray'
collect2.exe: error: ld returned 1 exit status
make[1]: *** [pg_buffercache.dll] Error 1
make[1]: Leaving directory
`/home/pgrunner/bf/root/HEAD/pgsql.6024/contrib/pg_buffercache'

Note that by this time "make check" has already passed.

That's expected atm, that's breaking all the windows builds right
now. Now that brolga is alive again, it should be able to progress past
this... How often does it build master?

Twice a day as long as currawong and frogmouth aren't running.

But it failed a few hours ago:

gcc-4 -O2 -Wall -Wmissing-prototypes -Wpointer-arith -Wdeclaration-after-statement -Wendif-labels -Wmissing-format-attribute -Wformat-security -fno-strict-aliasing -fwrapv -g zic.o ialloc.o scheck.o localtime.o -L../../src/port -L../../src/common -Wl,--allow-multiple-definition -Wl,--disable-auto-import -L/usr/lib -L/usr/local/lib -Wl,--as-needed -lpgcommon -lpgport -lxslt -lxml2 -lz -lreadline -lcrypt -o zic.exe
zic.o: In function `main':
/cygdrive/d/cygwin-bf/root/HEAD/pgsql.3216/src/timezone/zic.c:511: undefined reference to `_optarg'
/cygdrive/d/cygwin-bf/root/HEAD/pgsql.3216/src/timezone/zic.c:544: undefined reference to `_optarg'
/cygdrive/d/cygwin-bf/root/HEAD/pgsql.3216/src/timezone/zic.c:533: undefined reference to `_optarg'
/cygdrive/d/cygwin-bf/root/HEAD/pgsql.3216/src/timezone/zic.c:522: undefined reference to `_optarg'
/cygdrive/d/cygwin-bf/root/HEAD/pgsql.3216/src/timezone/zic.c:555: undefined reference to `_optarg'
/cygdrive/d/cygwin-bf/root/HEAD/pgsql.3216/src/timezone/zic.c:575: undefined reference to `_optind'
/cygdrive/d/cygwin-bf/root/HEAD/pgsql.3216/src/timezone/zic.c:584: undefined reference to `_optind'
/cygdrive/d/cygwin-bf/root/HEAD/pgsql.3216/src/timezone/zic.c:590: undefined reference to `_optind'
collect2: ld returned 1 exit status

cheers

andrew

cheers

andrew

--
Sent via pgsql-committers mailing list (pgsql-committers@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-committers

On 2014-02-15 10:02:39 -0500, Andrew Dunstan wrote:

On 02/15/2014 09:49 AM, Andres Freund wrote:

On 2014-02-15 09:47:08 -0500, Andrew Dunstan wrote:

However, the build now fails like this:
pg_buffercache_pages.o: In function `pg_buffercache_pages':
c:/MinGW/msys/1.0/home/pgrunner/bf/root/HEAD/pgsql.6024/../pgsql/contrib/pg_buffercache/pg_buffercache_pages.c:119:
undefined reference to `MainLWLockArray'
c:/MinGW/msys/1.0/home/pgrunner/bf/root/HEAD/pgsql.6024/../pgsql/contrib/pg_buffercache/pg_buffercache_pages.c:160:
undefined reference to `MainLWLockArray'
collect2.exe: error: ld returned 1 exit status
make[1]: *** [pg_buffercache.dll] Error 1
make[1]: Leaving directory
`/home/pgrunner/bf/root/HEAD/pgsql.6024/contrib/pg_buffercache'

Note that by this time "make check" has already passed.

That's expected atm, that's breaking all the windows builds right
now. Now that brolga is alive again, it should be able to progress past
this... How often does it build master?

Twice a day as long as currawong and frogmouth aren't running.

But it failed a few hours ago:

/cygdrive/d/cygwin-bf/root/HEAD/pgsql.3216/src/timezone/zic.c:511: undefined reference to `_optarg'

Yes, there's some different fixes required:
http://archives.postgresql.org/message-id/20140215131209.GG20973%40alap3.anarazel.de

Greetings,

Andres Freund

--
Andres Freund http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services

--
Sent via pgsql-committers mailing list (pgsql-committers@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-committers

Andrew Dunstan <andrew@dunslane.net> writes:

However, the build now fails like this:

pg_buffercache_pages.o: In function `pg_buffercache_pages':
c:/MinGW/msys/1.0/home/pgrunner/bf/root/HEAD/pgsql.6024/../pgsql/contrib/pg_buffercache/pg_buffercache_pages.c:119:
undefined reference to `MainLWLockArray'

Yes, this is what we're expecting in HEAD. As soon as all the Windows
critters are doing that, I'll commit PGDLLIMPORT additions to fix it.
We're leaving it deliberately unfixed in the short term until we
confirm that all the toolchains are detecting this type of error now.

regards, tom lane

--
Sent via pgsql-committers mailing list (pgsql-committers@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-committers

On 02/15/2014 10:12 AM, Tom Lane wrote:

Andrew Dunstan <andrew@dunslane.net> writes:

However, the build now fails like this:
pg_buffercache_pages.o: In function `pg_buffercache_pages':
c:/MinGW/msys/1.0/home/pgrunner/bf/root/HEAD/pgsql.6024/../pgsql/contrib/pg_buffercache/pg_buffercache_pages.c:119:
undefined reference to `MainLWLockArray'

Yes, this is what we're expecting in HEAD. As soon as all the Windows
critters are doing that, I'll commit PGDLLIMPORT additions to fix it.
We're leaving it deliberately unfixed in the short term until we
confirm that all the toolchains are detecting this type of error now.

OK. Running jacana now.

cheers

andrew

--
Sent via pgsql-committers mailing list (pgsql-committers@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-committers