Re: [ADT] bug report?

View: New views
7 Messages — Rating Filter:   Alert me  

Parent Message unknown Re: [ADT] bug report?

by Stefan Manegold :: Rate this Message:

Reply to Author | View Threaded | Show Only this Message

Dear Hendrike,

thank you very much for your detailed report.

I suppose you sucessfully compiled the latest released "Stable" "Aug2008"
version of the Super tarball, right?

Then, indeed the compilation problems seem to be triggered by the (our) code
in the latest "cutting-edge" "Current" development version from the CVS
heads.

Although we do quite some testing (cf.
http://monetdb.cwi.nl/Development/TestWeb/), we do not (yet?) have a
"Leopard" machine in our testing pool (we actually also lost access to the
"Tiger" machine we used to have...).

We will look into the problem ASAP and try to fix it.
Could you please also file your problem as bug report in the MonetDB bug
tracker at SF:
http://sourceforge.net/tracker/?atid=482468&group_id=56967&func=browse
?

For the time being you might want to try the latest "Stable" version from
the CVS branches as listed at
http://monetdb.cwi.nl/projects/monetdb/Development/Releases/Version4.24/index.html
.

What puzzles me, though, is your initial point that no tests were included
in the tests were not included in the Super tarball. They should actaully be
in there. I'll check again ...

Stefan


On Mon, Sep 01, 2008 at 09:27:37PM +0200, Hendrike Peetz wrote:

> Dear people in charge,
>
> I have been spending a lot of time today trying to install Monet on  
> Leopard 10.5.4, 64 bit, core 2 duo.
>
> Before I did anything with the cvs, I tried to work with the Super  
> tarball version. Compiling was not a problem but the tests were not  
> installed and I was not too sure wether everything is working. I thus  
> wanted to install it properly:
>
> I checked out the latest version (friday I assume) from the cvs.
> bootstrap and configure worked fine, but I had problems with the  
> compiling.
>
> I tried gcc4.1 (apple) and gcc4.2 (apple) (Yes, I recompiled my c-
> compiler!). Both of them returned the attached error.
> I am not sure wether this is my mistake or some weird code, but as I was
> able to compile the tarball version I assume its something in the code.
>
> Best regards,
>
> Hendrike
>
>

> $ make
> make  all-recursive
> Making all in src
> make  all-recursive
> Making all in common
> make  all-am
> make[5]: Nothing to be done for `all-am'.
> Making all in gdk
> make  all-am
> /bin/sh ../../libtool --tag=CC --mode=compile gcc -DHAVE_CONFIG_H -I. -I../.. -I../../../MonetDB/src/gdk -I../../../MonetDB/src/gdk -I../common -I../../../MonetDB/src/gdk/../common   -DLIBGDK  -Wall -Wextra -std=c99 -g -Werror-implicit-function-declaration -Werror -Wpointer-arith -Wdeclaration-after-statement -Wundef -Wp,-D_FORTIFY_SOURCE=2 -Wno-uninitialized  -D_REENTRANT  -c -o libbat_la-gdk_system.lo `test -f 'gdk_system.c' || echo '../../../MonetDB/src/gdk/'`gdk_system.c
>  gcc -DHAVE_CONFIG_H -I. -I../.. -I../../../MonetDB/src/gdk -I../../../MonetDB/src/gdk -I../common -I../../../MonetDB/src/gdk/../common -DLIBGDK -Wall -Wextra -std=c99 -g -Werror-implicit-function-declaration -Werror -Wpointer-arith -Wdeclaration-after-statement -Wundef -Wp,-D_FORTIFY_SOURCE=2 -Wno-uninitialized -D_REENTRANT -c gdk_system.c  -fno-common -DPIC -o .libs/libbat_la-gdk_system.o
> cc1: warnings being treated as errors
> ../../../MonetDB/src/gdk/gdk_system.mx: In function 'MT_check_nr_cores':
> ../../../MonetDB/src/gdk/gdk_system.mx:575: warning: ISO C90 forbids mixed declarations and code
> ../../../MonetDB/src/gdk/gdk_system.mx:576: error: implicit declaration of function 'sysctlbyname'
> make[5]: *** [libbat_la-gdk_system.lo] Error 1
> make[4]: *** [all] Error 2
> make[3]: *** [all-recursive] Error 1
> make[2]: *** [all] Error 2
> make[1]: *** [all-recursive] Error 1
> make: *** [all] Error 2

>
>
>
>


--
| Dr. Stefan Manegold | mailto:Stefan.Manegold@... |
| CWI,  P.O.Box 94079 | http://www.cwi.nl/~manegold/  |
| 1090 GB Amsterdam   | Tel.: +31 (20) 592-4212       |
| The Netherlands     | Fax : +31 (20) 592-4312       |

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
MonetDB-users mailing list
MonetDB-users@...
https://lists.sourceforge.net/lists/listinfo/monetdb-users

Re: [ADT] bug report?

by Fabian Groffen :: Rate this Message:

Reply to Author | View Threaded | Show Only this Message

Just from the side...

> On Mon, Sep 01, 2008 at 09:27:37PM +0200, Hendrike Peetz wrote:
> > Dear people in charge,
> >
> > I have been spending a lot of time today trying to install Monet on  
> > Leopard 10.5.4, 64 bit, core 2 duo.
> >
> > Before I did anything with the cvs, I tried to work with the Super  
> > tarball version. Compiling was not a problem but the tests were not  
> > installed and I was not too sure wether everything is working. I thus  
> > wanted to install it properly:

How did you try to "test"?  What was the error you got which made you
conclude that the tests weren't included?

> > I checked out the latest version (friday I assume) from the cvs.
> > bootstrap and configure worked fine, but I had problems with the  
> > compiling.
> >
> > I tried gcc4.1 (apple) and gcc4.2 (apple) (Yes, I recompiled my c-
> > compiler!). Both of them returned the attached error.

Apple never released any 4.1 compiler (only 4.0.1 and very recently
4.2.1 from Xcode Tools 3.1 -- a few revisions later than the beta
versions)

> > I am not sure wether this is my mistake or some weird code, but as I was
> > able to compile the tarball version I assume its something in the code.

> > cc1: warnings being treated as errors
> > ../../../MonetDB/src/gdk/gdk_system.mx: In function 'MT_check_nr_cores':
> > ../../../MonetDB/src/gdk/gdk_system.mx:575: warning: ISO C90 forbids mixed declarations and code
> > ../../../MonetDB/src/gdk/gdk_system.mx:576: error: implicit declaration of function 'sysctlbyname'

No.  The MonetDB sources aren't tested on Leopard on a daily basis, and
this change/code in particular isn't at all.  I tried it on Tiger, but
didn't care to check on Leopard.

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
MonetDB-users mailing list
MonetDB-users@...
https://lists.sourceforge.net/lists/listinfo/monetdb-users

Re: [ADT] bug report?

by Fabian Groffen :: Rate this Message:

Reply to Author | View Threaded | Show Only this Message

On 01-09-2008 21:57:13 +0200, Fabian Groffen wrote:
> > > cc1: warnings being treated as errors
> > > ../../../MonetDB/src/gdk/gdk_system.mx: In function 'MT_check_nr_cores':
> > > ../../../MonetDB/src/gdk/gdk_system.mx:575: warning: ISO C90 forbids mixed declarations and code
> > > ../../../MonetDB/src/gdk/gdk_system.mx:576: error: implicit declaration of function 'sysctlbyname'
>
> No.  The MonetDB sources aren't tested on Leopard on a daily basis, and
> this change/code in particular isn't at all.  I tried it on Tiger, but
> didn't care to check on Leopard.

Actually, the actual code was never tested, as it's just missing an
include ;)

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
MonetDB-users mailing list
MonetDB-users@...
https://lists.sourceforge.net/lists/listinfo/monetdb-users

Re: [ADT] bug report?

by Fabian Groffen :: Rate this Message:

Reply to Author | View Threaded | Show Only this Message

I tried to fix the issue.  Please report if it works.

On 01-09-2008 22:01:01 +0200, Fabian Groffen (Project Account) wrote:

> On 01-09-2008 21:57:13 +0200, Fabian Groffen wrote:
> > > > cc1: warnings being treated as errors
> > > > ../../../MonetDB/src/gdk/gdk_system.mx: In function 'MT_check_nr_cores':
> > > > ../../../MonetDB/src/gdk/gdk_system.mx:575: warning: ISO C90 forbids mixed declarations and code
> > > > ../../../MonetDB/src/gdk/gdk_system.mx:576: error: implicit declaration of function 'sysctlbyname'
> >
> > No.  The MonetDB sources aren't tested on Leopard on a daily basis, and
> > this change/code in particular isn't at all.  I tried it on Tiger, but
> > didn't care to check on Leopard.
>
> Actually, the actual code was never tested, as it's just missing an
> include ;)

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
MonetDB-users mailing list
MonetDB-users@...
https://lists.sourceforge.net/lists/listinfo/monetdb-users

Re: [ADT] bug report?

by Hendrike Peetz :: Rate this Message:

Reply to Author | View Threaded | Show Only this Message

Dear Stefan,

If it is not tested on Leopard, I have to admit that the impression of  
being a "lab rat" does not really comfort me. I have a faible for  
finding weird bugs that might cost me days because I am never assuming  
anyone else to have done something wrong. I will continue to try to  
install it though.

Can I do the work via ssh and if so, which computer may I use?


Hendrike

On Sep 1, 2008, at 9:50 PM, Stefan Manegold wrote:

> Dear Hendrike,
>
> thank you very much for your detailed report.
>
> I suppose you sucessfully compiled the latest released "Stable"  
> "Aug2008"
> version of the Super tarball, right?
>
> Then, indeed the compilation problems seem to be triggered by the  
> (our) code
> in the latest "cutting-edge" "Current" development version from the  
> CVS
> heads.
>
> Although we do quite some testing (cf.
> http://monetdb.cwi.nl/Development/TestWeb/), we do not (yet?) have a
> "Leopard" machine in our testing pool (we actually also lost access  
> to the
> "Tiger" machine we used to have...).
>
> We will look into the problem ASAP and try to fix it.
> Could you please also file your problem as bug report in the MonetDB  
> bug
> tracker at SF:
> http://sourceforge.net/tracker/?atid=482468&group_id=56967&func=browse
> ?
>
> For the time being you might want to try the latest "Stable" version  
> from
> the CVS branches as listed at
> http://monetdb.cwi.nl/projects/monetdb/Development/Releases/Version4.24/index.html
> .
>
> What puzzles me, though, is your initial point that no tests were  
> included
> in the tests were not included in the Super tarball. They should  
> actaully be
> in there. I'll check again ...
>
> Stefan
>
>
> On Mon, Sep 01, 2008 at 09:27:37PM +0200, Hendrike Peetz wrote:
>> Dear people in charge,
>>
>> I have been spending a lot of time today trying to install Monet on
>> Leopard 10.5.4, 64 bit, core 2 duo.
>>
>> Before I did anything with the cvs, I tried to work with the Super
>> tarball version. Compiling was not a problem but the tests were not
>> installed and I was not too sure wether everything is working. I thus
>> wanted to install it properly:
>>
>> I checked out the latest version (friday I assume) from the cvs.
>> bootstrap and configure worked fine, but I had problems with the
>> compiling.
>>
>> I tried gcc4.1 (apple) and gcc4.2 (apple) (Yes, I recompiled my c-
>> compiler!). Both of them returned the attached error.
>> I am not sure wether this is my mistake or some weird code, but as  
>> I was
>> able to compile the tarball version I assume its something in the  
>> code.
>>
>> Best regards,
>>
>> Hendrike
>>
>>
>
>> $ make
>> make  all-recursive
>> Making all in src
>> make  all-recursive
>> Making all in common
>> make  all-am
>> make[5]: Nothing to be done for `all-am'.
>> Making all in gdk
>> make  all-am
>> /bin/sh ../../libtool --tag=CC --mode=compile gcc -DHAVE_CONFIG_H -
>> I. -I../.. -I../../../MonetDB/src/gdk -I../../../MonetDB/src/gdk -
>> I../common -I../../../MonetDB/src/gdk/../common   -DLIBGDK  -Wall -
>> Wextra -std=c99 -g -Werror-implicit-function-declaration -Werror -
>> Wpointer-arith -Wdeclaration-after-statement -Wundef -Wp,-
>> D_FORTIFY_SOURCE=2 -Wno-uninitialized  -D_REENTRANT  -c -o  
>> libbat_la-gdk_system.lo `test -f 'gdk_system.c' || echo '../../../
>> MonetDB/src/gdk/'`gdk_system.c
>> gcc -DHAVE_CONFIG_H -I. -I../.. -I../../../MonetDB/src/gdk -
>> I../../../MonetDB/src/gdk -I../common -I../../../MonetDB/src/gdk/../
>> common -DLIBGDK -Wall -Wextra -std=c99 -g -Werror-implicit-function-
>> declaration -Werror -Wpointer-arith -Wdeclaration-after-statement -
>> Wundef -Wp,-D_FORTIFY_SOURCE=2 -Wno-uninitialized -D_REENTRANT -c  
>> gdk_system.c  -fno-common -DPIC -o .libs/libbat_la-gdk_system.o
>> cc1: warnings being treated as errors
>> ../../../MonetDB/src/gdk/gdk_system.mx: In function  
>> 'MT_check_nr_cores':
>> ../../../MonetDB/src/gdk/gdk_system.mx:575: warning: ISO C90  
>> forbids mixed declarations and code
>> ../../../MonetDB/src/gdk/gdk_system.mx:576: error: implicit  
>> declaration of function 'sysctlbyname'
>> make[5]: *** [libbat_la-gdk_system.lo] Error 1
>> make[4]: *** [all] Error 2
>> make[3]: *** [all-recursive] Error 1
>> make[2]: *** [all] Error 2
>> make[1]: *** [all-recursive] Error 1
>> make: *** [all] Error 2
>
>>
>>
>>
>>
>
>
> --
> | Dr. Stefan Manegold | mailto:Stefan.Manegold@... |
> | CWI,  P.O.Box 94079 | http://www.cwi.nl/~manegold/  |
> | 1090 GB Amsterdam   | Tel.: +31 (20) 592-4212       |
> | The Netherlands     | Fax : +31 (20) 592-4312       |


-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
MonetDB-users mailing list
MonetDB-users@...
https://lists.sourceforge.net/lists/listinfo/monetdb-users

Parent Message unknown Re: [ADT] bug report?

by Stefan Manegold :: Rate this Message:

Reply to Author | View Threaded | Show Only this Message

On Mon, Sep 01, 2008 at 10:08:09PM +0200, Hendrike Peetz wrote:
> Ok, I assume I should have set compiler options to --disable-strict.

well, for us, it should also work with --enable-strict (default in CVS
version as that is meant for developers --- and as a reminder to us to
produce clean code ... ;-)) --- but for "users", --disable-strict (default
in tarballs) is indeed just fine.

> But now I can still not check, because it says:
>
> $ make check
> make  check-recursive
> Making check in src
> make  check-recursive
> Making check in common
> make  check-am
> make[5]: Nothing to be done for `check-am'.
> Making check in gdk
> make  check-am
> make[5]: Nothing to be done for `check-am'.
> Making check in testing
> make  check-am
> rm -f Mapprove.py
> ln -s Mtest.py Mapprove.py
> make[4]: Nothing to be done for `check-am'.
> Making check in conf
> make  check-am
> make[3]: Nothing to be done for `check-am'.

for which part did you call `make check`?
for "MonetDB" (Common) there are indeed not tests; neither for
(MonetDB/)"java"; only for clients, MonetDB4, MonetDB5, sql, geom,
pathfinder, template there are tests and `make check` should work --- be
aware though, that some tests are known to fail on some platforms ...

> Sorry for bothering you, but we were two people (who do know the basics
> of working with unix) trying to get this running today and couldn't make
> it work.

No problem --- in fact, `make check` is not "mendatory" for getting it to
work --- just `make && make install` (or use monetdb-install.sh) and then
try to run `Mserver` (MonetDB4) or `mserver5` (MonetDB5) as a very basic
"sanitycheck".

Stefan

> Hendrike
>
>
> Begin forwarded message:
>
>> From: Hendrike Peetz <hendrike@...>
>> Date: September 1, 2008 9:27:37 PM GMT+02:00
>> To: Monet@...
>> Subject: [ADT] bug report?
>>
>> Dear people in charge,
>>
>> I have been spending a lot of time today trying to install Monet on  
>> Leopard 10.5.4, 64 bit, core 2 duo.
>>
>> Before I did anything with the cvs, I tried to work with the Super  
>> tarball version. Compiling was not a problem but the tests were not  
>> installed and I was not too sure wether everything is working. I thus
>> wanted to install it properly:
>>
>> I checked out the latest version (friday I assume) from the cvs.
>> bootstrap and configure worked fine, but I had problems with the  
>> compiling.
>>
>> I tried gcc4.1 (apple) and gcc4.2 (apple) (Yes, I recompiled my c-
>> compiler!). Both of them returned the attached error.
>> I am not sure wether this is my mistake or some weird code, but as I  
>> was able to compile the tarball version I assume its something in the
>> code.
>>
>> Best regards,
>>
>> Hendrike
>>
>>
>>
>>
>>
>>
>

--
| Dr. Stefan Manegold | mailto:Stefan.Manegold@... |
| CWI,  P.O.Box 94079 | http://www.cwi.nl/~manegold/  |
| 1090 GB Amsterdam   | Tel.: +31 (20) 592-4212       |
| The Netherlands     | Fax : +31 (20) 592-4312       |

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
MonetDB-users mailing list
MonetDB-users@...
https://lists.sourceforge.net/lists/listinfo/monetdb-users

Parent Message unknown Re: [ADT] bug report?

by Stefan Manegold :: Rate this Message:

Reply to Author | View Threaded | Show Only this Message

On Mon, Sep 01, 2008 at 11:22:10PM +0200, Ying Zhang wrote:

> Hi,
>
> I have just updated my monet source from souceforge and did a from  
> scratch compilation on my leopard.
>
> Fabian's fix gets rid of the error "implicit declaration of..." (thanks!)
> However, the other error "ISO C90 forbids.." is still there.  With the  
> following patch, I was able to compile the module MonetDB (a complete  
> rebuild of monet modules is still ongoing).
>
> I simply moved the definition of len to the begin of the function:
>
> $ cvs diff src/gdk/gdk_system.mx
> Index: src/gdk/gdk_system.mx
> ===================================================================
> RCS file: /cvsroot/monetdb/MonetDB/src/gdk/gdk_system.mx,v
> retrieving revision 1.115
> diff -r1.115 gdk_system.mx
> 570a571,574
> > #if defined(__FreeBSD__) || defined(__NetBSD__) ||  
> defined(__OpenBSD__) || defined(__MACH__)
> > size_t len = sizeof(int);
> > #endif
> >
> 578d581
> < size_t len = sizeof(int);

Thanks.
I just checked in (my version of ;-)) your patch.

Stefan

> Jennie
>
>
> On Sep 01, 2008, at 22:35 , Hendrike Peetz wrote:
>
>> nope. still the same problem. I checked out from the svn right after I
>> got your mail.
>>
>> On Sep 1, 2008, at 10:13 PM, Fabian Groffen (Project Account) wrote:
>>
>>> I tried to fix the issue.  Please report if it works.
>>>
>>> On 01-09-2008 22:01:01 +0200, Fabian Groffen (Project Account) wrote:
>>>> On 01-09-2008 21:57:13 +0200, Fabian Groffen wrote:
>>>>>>> cc1: warnings being treated as errors
>>>>>>> ../../../MonetDB/src/gdk/gdk_system.mx: In function  
>>>>>>> 'MT_check_nr_cores':
>>>>>>> ../../../MonetDB/src/gdk/gdk_system.mx:575: warning: ISO C90  
>>>>>>> forbids mixed declarations and code
>>>>>>> ../../../MonetDB/src/gdk/gdk_system.mx:576: error: implicit  
>>>>>>> declaration of function 'sysctlbyname'
>>>>>
>>>>> No.  The MonetDB sources aren't tested on Leopard on a daily  
>>>>> basis, and
>>>>> this change/code in particular isn't at all.  I tried it on  
>>>>> Tiger, but
>>>>> didn't care to check on Leopard.
>>>>
>>>> Actually, the actual code was never tested, as it's just missing an
>>>> include ;)
>>
>
>

--
| Dr. Stefan Manegold | mailto:Stefan.Manegold@... |
| CWI,  P.O.Box 94079 | http://www.cwi.nl/~manegold/  |
| 1090 GB Amsterdam   | Tel.: +31 (20) 592-4212       |
| The Netherlands     | Fax : +31 (20) 592-4312       |

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
MonetDB-users mailing list
MonetDB-users@...
https://lists.sourceforge.net/lists/listinfo/monetdb-users
LightInTheBox - Buy quality products at wholesale price!