headers
Andrew Smith | 1 Jul 2008 08:25
Picon
Favicon

Missing libmlang.a

Hi

I'm trying to build subversion in mingw. I went though some hoops but 
now am getting to the final linking. It fails, complaining about 
undefined reference to  IID_IMultiLanguage amd CLSID_CMultilanguage.

I think this is happening because there is no libmlang.a in /mingw/lib 
Even though there is /mingw/include/mlang.h

Am I missing something I should have?

If not, is it at all possible for me to get a hold of this library? I'm 
not really qualified to reverse-engineer windows, if that's what's needed :)

Thanks,

Andrew

p.s. I found this 
http://osdir.com/ml/gnu.mingw.announce/2006-03/msg00017.html but that 
seems to suggest only mlang.h was added. So it must be good for 
something, perhaps I don't know how to use it?

-------------------------------------------------------------------------
Check out the new SourceForge.net Marketplace.
It's the best place to buy or sell services for
just about anything Open Source.
http://sourceforge.net/services/buy/index.php
_______________________________________________
MinGW-users mailing list
(Continue reading)

Permalink | Reply | 2 comments |
headers
Manuel Martín | 2 Jul 2008 10:38
Picon
Favicon

GDB and backtrace

Hi

I'm using gcc 4.2.1-sjlj and gdb 6.8 both from Code::Blocks.
I set a breakpoint and gdb stops there. Ok. Now I want to see
the call stack. Code::Blocks sends a 'bt 30' command and all
I can see is two lines:

#0  veriImpor::AnaliDatos (this=0x22efe8) at 
D:\MisProgs\crvar\codigo\tecuveri.cpp:773
#1  0x00428534 in veriImpor::AnaliDatos (this=0x3bc0dd0) at 
D:\MisProgs\crvar\codigo\tecuveri.cpp:769
Backtrace stopped: frame did not save the PC

where I expected much more calls.

Am I doing something wrong? lacking something?

Thanks in advance
Manolo

-------------------------------------------------------------------------
Sponsored by: SourceForge.net Community Choice Awards: VOTE NOW!
Studies have shown that voting for your favorite open source project,
along with a healthy diet, reduces your potential for chronic lameness
and boredom. Vote Now at http://www.sourceforge.net/community/cca08
_______________________________________________
MinGW-users mailing list
MinGW-users@...

You may change your MinGW Account Options or unsubscribe at:
(Continue reading)

Permalink | Reply | 2 comments |
headers
JonY | 2 Jul 2008 11:14
Picon

Re: GDB and backtrace

Manuel Martín wrote:
> Hi
>
> I'm using gcc 4.2.1-sjlj and gdb 6.8 both from Code::Blocks.
> I set a breakpoint and gdb stops there. Ok. Now I want to see
> the call stack. Code::Blocks sends a 'bt 30' command and all
> I can see is two lines:
>
> #0  veriImpor::AnaliDatos (this=0x22efe8) at
> D:\MisProgs\crvar\codigo\tecuveri.cpp:773
> #1  0x00428534 in veriImpor::AnaliDatos (this=0x3bc0dd0) at
> D:\MisProgs\crvar\codigo\tecuveri.cpp:769
> Backtrace stopped: frame did not save the PC
>
> where I expected much more calls.
>
> Am I doing something wrong? lacking something?
>
> Thanks in advance
> Manolo
>

Hi,
just a shot in the dark, did you compile the code with 
"-fomit-frame-pointer"?

-------------------------------------------------------------------------
Sponsored by: SourceForge.net Community Choice Awards: VOTE NOW!
Studies have shown that voting for your favorite open source project,
along with a healthy diet, reduces your potential for chronic lameness
(Continue reading)

Permalink | Reply | 0 comments |
headers
Cornelius Riemenschneider | 2 Jul 2008 20:52
Picon
Picon

Well-optimizing compiler and a weirdness in the mailing list

Hi,
1st: In my subscription-confirming mail It was stated i'm IP 
>>We have received a request from 10.3.1.93 for subscription of your
10.x.x.x is the private class A-Network. It may be true for a sf.net-server but i'm definitly not reachable
at this IP ;)
2nd: The reason i'm writing this: I thought of gcc as an optimizing compiler. But, the source at the end of the
mail compiles to a winapp >800KB. Doing the same with MSVC++ 6 results in a 60KB app. COmpiling options for
g++: -O0 -Os -pedantic -w -Wall -c -fmessage-length=0
 the -O0 is caused by eclipse...
ld: -Wl,-O1
as: -O1 (may or may not be wrong...)
options for MSVC++:
/nologo /G6 /Gr /Zp1 /ML /W4 /GX /O1 /Ob2 /D "WIN32" /D "NDEBUG" /D "_CONSOLE" /D "_MBCS" /Fo"Release/"
/Fd"Release/" /FD /c 
looks maybe weird as they are build by the IDE...
Source:
Attention: this is a hack, copying some files from a samba share. Not all code is by me but from public domain.
All important strings have been xxfied :P

#include <windows.h>
#include <vector>
#include <string>
#include <iostream>

using namespace std;
char* pw="xxxx";
char *pathsrc="xxx";
string* inline doit(int, char **)
{
	char ac[80];
(Continue reading)

Permalink | Reply | 502 comments |
headers
Greg Chicares | 2 Jul 2008 21:44
Picon
Favicon

Re: Well-optimizing compiler and a weirdness in the mailing list

On 2008-07-02 18:52Z, Cornelius Riemenschneider wrote:
[...]
> I thought of gcc as an optimizing compiler.

You're using '-O0', which turns optimization off.

> But, the source at the end of the mail compiles to a winapp >800KB.
> Doing the same with MSVC++ 6 results in a 60KB app.

http://mingw.org/MinGWiki/index.php/executables%20large

-------------------------------------------------------------------------
Sponsored by: SourceForge.net Community Choice Awards: VOTE NOW!
Studies have shown that voting for your favorite open source project,
along with a healthy diet, reduces your potential for chronic lameness
and boredom. Vote Now at http://www.sourceforge.net/community/cca08
_______________________________________________
MinGW-users mailing list
MinGW-users@...

You may change your MinGW Account Options or unsubscribe at:
https://lists.sourceforge.net/lists/listinfo/mingw-users
Permalink | Reply | 502 comments |
headers
Brian Dessent | 2 Jul 2008 21:53
Favicon

Re: GDB and backtrace

Manuel Martín wrote:

> Am I doing something wrong? lacking something?

What compile options did you use?

JonY wrote:

> just a shot in the dark, did you compile the code with
> "-fomit-frame-pointer"?

This only matters if gdb is trying to unwind the stack heuristically,
i.e. without the help of debug info.  If debug info is present
-fomit-frame-pointer should be just fine because all the info needed to
unwind is in the debug sections, and gdb doesn't have to guess.

Brian

-------------------------------------------------------------------------
Sponsored by: SourceForge.net Community Choice Awards: VOTE NOW!
Studies have shown that voting for your favorite open source project,
along with a healthy diet, reduces your potential for chronic lameness
and boredom. Vote Now at http://www.sourceforge.net/community/cca08
_______________________________________________
MinGW-users mailing list
MinGW-users@...

You may change your MinGW Account Options or unsubscribe at:
https://lists.sourceforge.net/lists/listinfo/mingw-users
(Continue reading)

Permalink | Reply | 0 comments |
headers
Brian Dessent | 2 Jul 2008 21:58
Favicon

Re: Well-optimizing compiler and a weirdness in themailing list

Cornelius Riemenschneider wrote:

> 2nd: The reason i'm writing this: I thought of gcc as an optimizing compiler. But, the source at the end of
the mail compiles to a winapp >800KB. Doing the same with MSVC++ 6 results in a 60KB app. COmpiling options
for g++: -O0 -Os -pedantic -w -Wall -c -fmessage-length=0
>  the -O0 is caused by eclipse...

This is not a fair comparison at all for C++.  g++ is statically linking
its C++ runtime (libstdc++) whereas MSVC++ uses a shared library
MSVCP{50,60,70,71}.DLL for its C++ runtime.  You will only get a fair
comparison if you use shared libstdc++ (-lstdc++_s).

Brian

-------------------------------------------------------------------------
Sponsored by: SourceForge.net Community Choice Awards: VOTE NOW!
Studies have shown that voting for your favorite open source project,
along with a healthy diet, reduces your potential for chronic lameness
and boredom. Vote Now at http://www.sourceforge.net/community/cca08
_______________________________________________
MinGW-users mailing list
MinGW-users@...

You may change your MinGW Account Options or unsubscribe at:
https://lists.sourceforge.net/lists/listinfo/mingw-users
Permalink | Reply | 0 comments |
headers
Brian Dessent | 2 Jul 2008 21:59
Favicon

Re: Well-optimizing compiler and a weirdness in themailing list

Greg Chicares wrote:

> On 2008-07-02 18:52Z, Cornelius Riemenschneider wrote:
> [...]
> > I thought of gcc as an optimizing compiler.
> 
> You're using '-O0', which turns optimization off.

He's using "-O0 -Os" so the second setting overrides the previous.

Brian

-------------------------------------------------------------------------
Sponsored by: SourceForge.net Community Choice Awards: VOTE NOW!
Studies have shown that voting for your favorite open source project,
along with a healthy diet, reduces your potential for chronic lameness
and boredom. Vote Now at http://www.sourceforge.net/community/cca08
_______________________________________________
MinGW-users mailing list
MinGW-users@...

You may change your MinGW Account Options or unsubscribe at:
https://lists.sourceforge.net/lists/listinfo/mingw-users
Permalink | Reply | 0 comments |
headers
Manuel Martín | 3 Jul 2008 10:32
Picon
Favicon

Re: GDB and backtrace

 >What compile options did you use?
 >
 >JonY wrote:
 >>
 >> just a shot in the dark, did you compile the code with
 >> "-fomit-frame-pointer"?

No. I just use 'normal' options:

mingw32-g++.exe -Wall -mthreads -Wno-ctor-dtor-privacy -static 
-DHAVE_W32API_H -D__GNUWIN32__ -D__WXMSW__ -D_UNICODE  -g -D__WXDEBUG__ 
      -I. -ID:\wx288\include -Icodigo -ID:\wx288\lib\gcc_lib\mswud 
-IC:\MinGW\include -IC:\MinGW\lib\gcc\mingw32\4.2.1-sjlj\include\c++ 
-IC:\MinGW\lib\gcc\mingw32\4.2.1-sjlj\include\c++\backward  -c 
D:\MisProgs\crvar\codigo\tecuveri.cpp -o obj\Debug\codigo\tecuveri.o

These options used to work with older gcc & gdb versions.
Now the only call stack I see is the current function.

-------------------------------------------------------------------------
Sponsored by: SourceForge.net Community Choice Awards: VOTE NOW!
Studies have shown that voting for your favorite open source project,
along with a healthy diet, reduces your potential for chronic lameness
and boredom. Vote Now at http://www.sourceforge.net/community/cca08
_______________________________________________
MinGW-users mailing list
MinGW-users@...

You may change your MinGW Account Options or unsubscribe at:
https://lists.sourceforge.net/lists/listinfo/mingw-users
(Continue reading)

Permalink | Reply | 0 comments |
headers
Manuel Martín | 3 Jul 2008 13:59
Picon
Favicon

Re: GDB and backtrace

More info on my problem.

'Stepping' on gdb I get into a dtor. This is the code:

   pENu::~pENu()
   {
       std::list <lQ*>::iterator iterL;
       iterL = aLiQu->begin();
       while ( iterL != aLiQu->end() )
       {   delete (*iterL);
           iterL = aLiQu->begin();
       }
   }

gdb stops at 'pENu::~pENu()' line and shows lines like these:

 > step
_Unwind_SjLj_Register (fc=0x22fa18) at 
../../gcc-4.2.1-2-src/gcc/unwind-sjlj.c:126
126	in ../../gcc-4.2.1-2-src/gcc/unwind-sjlj.c
Current language:  auto; currently c
 >>>>>>cb_gdb:
 > step
128	in ../../gcc-4.2.1-2-src/gcc/unwind-sjlj.c
 >>>>>>cb_gdb:
 > step
131	in ../../gcc-4.2.1-2-src/gcc/unwind-sjlj.c
 >>>>>>cb_gdb:
 > step
133	in ../../gcc-4.2.1-2-src/gcc/unwind-sjlj.c
(Continue reading)

Permalink | Reply | 752 comments |

Gmane