[MinGW-notify] [ mingw-Bugs-3156957 ] install -s strips export tables from dll's

[SF/projects/mingw notification list <mingw-notify@li...>]

Bugs item #3156957, was opened at 2011-01-13 18:41
Message generated for change (Tracker Item Submitted) made by donaldd8
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=102435&aid=3156957&group_id=2435

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: None
Group: None
Status: Open
Resolution: None
Priority: 5
Private: No
Submitted By: Donald Douwsma (donaldd8)
Assigned to: Nobody/Anonymous (nobody)
Summary: install -s strips export tables from dll's

Initial Comment:
I've been working on a project that uses the install command when
creating its installation package. The program builds both a Linux 
and a Windows version from shared soruce. Both are native builds.

Unfortunately I've found that 'install -s' causes Windows dll's to have their
export tables removed, rendering them useless. Currently I'm working 
around this with a platform specific override that drops the -s for Windows
builds.

Steps to reproduce.
Open a dll with depends  http://www.dependencywalker.com/
and observe that there is are exports. Run install -s to copy the
dll somewhere. Open the copied dll with depends and observe
that it has trashed the export symbols. 

depends /bin/iconv.dll
install -s /bin/iconv.dll ~/
depends ~/iconv.dll

I believe this is a regression as the project I'm working on previously
used install -s without these problems. 

OS is Windows XP Professional
uname -a
MINGW32_NT-5.1 DD 1.0.16(0.48/3/2) 2010-09-29 00:07 i686 Msys

install --version
install (GNU coreutils) 5.97
Copyright (C) 2006 Free Software Foundation, Inc.
This is free software.  You may redistribute copies of it under the terms of
the GNU General Public License <http://www.gnu.org/licenses/gpl.html>;.
There is NO WARRANTY, to the extent permitted by law.



----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=102435&aid=3156957&group_id=2435

[MinGW-notify] [ mingw-Bugs-3147803 ] No error message on failed download

[SF/projects/mingw notification list <mingw-notify@li...>]

Bugs item #3147803, was opened at 2010-12-29 16:17
Message generated for change (Comment added) made by belgarionthekin
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=102435&aid=3147803&group_id=2435

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: MinGW Installer
Group: Known Feature
Status: Open
Resolution: None
Priority: 3
Private: No
Submitted By: Mike Hudson (belgarionthekin)
Assigned to: Keith Marshall (keithmarshall)
Summary: No error message on failed download

Initial Comment:
I apologize if this is a duplicate bug.

I spent over 20 minutes wondering why I couldn't use MinGW.  Finally it dawned on me.

The download of the files with mingw-get failed because I am behind a corporate firewall/proxy that requires authentication.  EVERYTHING that wants to connect to the internet needs to authenticate to the proxy.  All the browsers pop up an authentication window so I can put in my creds.

The problem is that the mingw-get window's messages flew by so fast that I couldn't see what was said.  Then, the shell window just went away and I couldn't scroll up to see the errors.

I was able to connect to a different network to download the proper file, but it would have been nice to know that I needed to 20 minutes ago.

----------------------------------------------------------------------

>Comment By: Mike Hudson (belgarionthekin)
Date: 2011-01-11 10:25

Message:
Thanks for your response Keith.

A little background on my proxy.  The method used in the installer might
have worked with the old way the proxy works, but not the new way.

Old way: The domain and realm were always the same.  One login would
authenticate to everything.

New way: The domain isn't always, the same, but the realm has some
uniformity.  I keep having to authenticate on pages that I've already
received, like when flash ads or flash vids come up and request content
from other sites.  I'm not 100% sure why this is, but it's pretty
annoying.

I don't normally use IE but I have saved my proxy creds so I don't have to
fill them in every time in IE.  

I'd be willing to test things and give you log files if/when you get
around to this.  

----------------------------------------------------------------------

Comment By: Keith Marshall (keithmarshall)
Date: 2011-01-08 18:30

Message:
Thank you for the report; I am aware of some download issues, but a formal
bug report does no harm.

You actually raise three separate, (although somewhat related issues):--

1) If downloads are blocked by a firewall, then presumably that's because
the firewall has been configured to block the host domain; it's doing its
intended job, and we can do nothing to work around it.  (I don't think this
is the problem, in your case).

2) Currently, mingw-get supports only one mechanism for authenticating to
a proxy -- the Windows standard method where the credentials for each user
are preconfigured, (e.g. by Internet Explorer setup), and recorded in the
registry.  That method works absolutely fine for me, behind the
authenticating Squid proxy in my place of work; apparently it doesn't work
for you, so you must be faced by a more draconian proxy configuration than
I am.  Ultimately, I would like to be able to support users with more
difficult proxy configurations, but since I cannot reproduce your problem
myself, I may need some assistance with specifying, coding and testing the
alternatives.  In the present alpha phase of development, this isn't a high
priority for me, but I will willingly consider patches.

3) mingw-get's diagnostics, particularly those related to download issues,
could be improved; in some respects, there is too much informational noise,
while some more serious issues may not be adequately reported.  This is on
my to-do list, but again, it isn't my highest priority at present.

----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=102435&aid=3147803&group_id=2435

[MinGW-notify] [ mingw-Patches-3154551 ] Define PAGE_WRITECOMBINE, THREAD_MODE_BACKGROUND_*

[SF/projects/mingw notification list <mingw-notify@li...>]

Patches item #3154551, was opened at 2011-01-10 14:57
Message generated for change (Tracker Item Submitted) made by sethj
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=302435&aid=3154551&group_id=2435

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: w32api
Group: None
Status: Open
Resolution: None
Priority: 5
Private: No
Submitted By: Seth Jackson (sethj)
Assigned to: Nobody/Anonymous (nobody)
Summary: Define PAGE_WRITECOMBINE, THREAD_MODE_BACKGROUND_*

Initial Comment:
This patch adds the PAGE_WRITECOMBINE, THREAD_MODE_BACKGROUND_BEGIN, and THREAD_MODE_BACKGROUND_END flags.

PAGE_WRITECOMBINE source: http://msdn.microsoft.com/en-us/library/aa366786%28v=vs.85%29.aspx
[Fixes bug: 3152142]
  => https://sourceforge.net/tracker/?func=detail&aid=3152142&group_id=2435&atid=102435

THREAD_MODE_BACKGROUND_* source: http://msdn.microsoft.com/en-us/library/ms686277(v=vs.85).aspx
[Source patch for patch: 3148129]
  => https://sourceforge.net/tracker/?func=detail&aid=3148129&group_id=2435&atid=302435

----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=302435&aid=3154551&group_id=2435

[MinGW-notify] [ mingw-Bugs-3067290 ] There are no #defines for reason of ExitWindowsEx

[SF/projects/mingw notification list <mingw-notify@li...>]

Bugs item #3067290, was opened at 2010-09-16 00:03
Message generated for change (Comment added) made by sethj
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=102435&aid=3067290&group_id=2435

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: w32api
Group: None
Status: Open
Resolution: None
Priority: 5
Private: No
Submitted By: Коренберг Марк (mmarkk)
Assigned to: Nobody/Anonymous (nobody)
Summary: There are no #defines for reason of ExitWindowsEx

Initial Comment:
http://msdn.microsoft.com/en-us/library/aa376885.aspx

----------------------------------------------------------------------

Comment By: Seth Jackson (sethj)
Date: 2011-01-10 14:40

Message:
They are there in "reason.h". Added in 2005...

http://cygwin.com/cgi-bin/cvsweb.cgi/src/winsup/w32api/include/reason.h?cvsroot=src

----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=102435&aid=3067290&group_id=2435

[MinGW-notify] [ mingw-Bugs-3154437 ] _chkstk Windows compatibility

[SF/projects/mingw notification list <mingw-notify@li...>]

Bugs item #3154437, was opened at 2011-01-10 13:01
Message generated for change (Tracker Item Submitted) made by selinger
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=102435&aid=3154437&group_id=2435

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: gcc
Group: None
Status: Open
Resolution: None
Priority: 5
Private: No
Submitted By: Peter Selinger (selinger)
Assigned to: Nobody/Anonymous (nobody)
Summary: _chkstk Windows compatibility

Initial Comment:
Background: Windows defines a function called _chkstk, calls to which are automatically inserted by the compiler at the beginning of functions that use more than a certain amount of stack space. GCC also defines such a function; it was formerly called _alloca and is now called __chkstk, presumably for better Windows compatibility.

Problem: When creating an object file with GCC under MSYS/MinGW, it may include undefined calls to __chkstk. However, linking this using another compiler (such as MS Visual Studio) is not possible, because in Windows, the function is called _chkstk (note: only one underscore, not two). 

So what was meant to be compatible, is not really.

To be clear: in GCC, the function is called __chkstk (with two underscores), but the symbol in the object code is called ___chkstk (with 3 underscores), because of the extra "_" added to every symbol.

In Windows, the function is called _chkstk (with one underscore), but the symbol in the object code is called __chkstk (with 2 underscores), because of the extra "_" added to all symbols.

Presumably this lead to the confusion in the first place.

Suggested fix: the source file in question is gcc-3.4.4/gcc/config/i386/msys.asm.
Replace "___chkstk" (3 underscores) by "__chkstk" (2 underscores). Note that this has 
already been done in another source file uwin.asm.

Presumably also GCC itself would need to be updated to insert calls to _chkstk and not __chkstk. 

If backward compatibility is desired, ___chkstk can be retained as an alias for __chkstk



----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=102435&aid=3154437&group_id=2435

[MinGW-notify] [ mingw-Bugs-3153453 ] Segmentation fault building Qt Creator 2.0.1

[SF/projects/mingw notification list <mingw-notify@li...>]

Bugs item #3153453, was opened at 2011-01-08 19:20
Message generated for change (Comment added) made by cstrauss
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=102435&aid=3153453&group_id=2435

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: gcc
Group: None
>Status: Pending
Resolution: None
Priority: 5
Private: No
Submitted By: Bob Scott (bscott45)
>Assigned to: Cesar Strauss (cstrauss)
Summary: Segmentation fault building Qt Creator 2.0.1

Initial Comment:
Using Qt Creator 2.0.1 source

>g++ -v
Using built-in specs.
COLLECT_GCC=g++
COLLECT_LTO_WRAPPER=c:/mingw/bin/../libexec/gcc/mingw32/4.5.0/lto-wrapper.exe
Target: mingw32
Configured with: ../gcc-4.5.0/configure --enable-languages=c,c++,ada,fortran,obj
c,obj-c++ --disable-sjlj-exceptions --with-dwarf2 --enable-shared --enable-libgo
mp --disable-win32-registry --enable-libstdcxx-debug --enable-version-specific-r
untime-libs --disable-werror --build=mingw32 --prefix=/mingw
Thread model: win32
gcc version 4.5.0 (GCC)

>ld -v
GNU ld (GNU Binutils) 2.21

gives error:-

mingw32-make[6]: Entering directory `E:/work/qt-creator-build/src/libs/3rdparty/
botan/src'
g++ -c -fpermissive -finline-functions -Wno-long-long -O2 -frtti -fexceptions -m
threads -Wall -DUNICODE -DQT_LARGEFILE_SUPPORT -DIDE_LIBRARY_BASENAME=\"lib\" -D
QT_NO_CAST_TO_ASCII -DQT_USE_FAST_OPERATOR_PLUS -DQT_USE_FAST_CONCATENATION -DQT
_DLL -DQT_NO_DEBUG -DQT_GUI_LIB -DQT_CORE_LIB -DQT_HAVE_MMX -DQT_HAVE_3DNOW -DQT
_HAVE_SSE -DQT_HAVE_MMXEXT -DQT_HAVE_SSE2 -DQT_THREAD_SUPPORT -I"c:\Qt\4.7.1\inc
lude\QtCore" -I"c:\Qt\4.7.1\include\QtGui" -I"c:\Qt\4.7.1\include" -I"e:\work\qt
-creator-2.0.1-src\src\libs" -I"e:\work\qt-creator-2.0.1-src\tools" -I"e:\work\q
t-creator-2.0.1-src\src\libs\3rdparty\botan\src" -I"e:\work\qt-creator-2.0.1-src
\src\libs\3rdparty\botan\build" -I"e:\work\qt-creator-2.0.1-src\src\libs\3rdpart
y\botan\build\botan" -I"c:\Qt\4.7.1\include\ActiveQt" -I"release" -I"e:\work\qt-
creator-2.0.1-src\src\libs\3rdparty\botan\src" -I"." -I"c:\qt\4.7.1\mkspecs\win3
2-g++" -o release\pk_algs.o e:\work\qt-creator-2.0.1-src\src\libs\3rdparty\botan
\src\pubkey\pk_algs.cpp
e:\work\qt-creator-2.0.1-src\src\libs\3rdparty\botan\src\pubkey\pk_algs.cpp: In
function 'Botan::Private_Key* Botan::get_private_key(const std::string&)':
e:\work\qt-creator-2.0.1-src\src\libs\3rdparty\botan\src\pubkey\pk_algs.cpp:79:1
4: internal compiler error: Segmentation fault
Please submit a full bug report,
with preprocessed source if appropriate.




----------------------------------------------------------------------

>Comment By: Cesar Strauss (cstrauss)
Date: 2011-01-10 14:41

Message:
Could you send the preprocessed source, please? In the g++ command line
above, just replace -c by -E, and replace the output file option by -o
bug3153453.i

Thanks,
Cesar


----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=102435&aid=3153453&group_id=2435

[MinGW-notify] [ mingw-Support Requests-3153615 ] CVS not user friendly

[SF/projects/mingw notification list <mingw-notify@li...>]

Support Requests item #3153615, was opened at 2011-01-09 07:30
Message generated for change (Comment added) made by earnie
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=202435&aid=3153615&group_id=2435

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
>Category: Other
Group: None
>Status: Closed
Priority: 5
Private: No
Submitted By: Jonas (jonasno)
Assigned to: Nobody/Anonymous (nobody)
Summary: CVS not user friendly

Initial Comment:
Have there been any talk about converting to SVN or GIT ?
CVS is old and have begun to have age issues (on the program support side) (TortoiseCVS)

So would there be a possibility for you too convert to SVN or GIT ?
or perhaps setup a mirror of the CVS that is in SVN or GIT ?

----------------------------------------------------------------------

>Comment By: Earnie Boyd (earnie)
Date: 2011-01-10 07:39

Message:
CVS is just as user friendly as SVN or GIT.  I've used all three and know
the differences are not all that different from an end-user point of view. 
The bike shed colour will remain CVS until there is sufficient reason to
cause the developers to spend time repainting the bike shed.

----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=202435&aid=3153615&group_id=2435

[MinGW-notify] [ mingw-Support Requests-3153615 ] CVS not user friendly

[SF/projects/mingw notification list <mingw-notify@li...>]

Support Requests item #3153615, was opened at 2011-01-09 13:30
Message generated for change (Tracker Item Submitted) made by jonasno
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=202435&aid=3153615&group_id=2435

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: None
Group: None
Status: Open
Priority: 5
Private: No
Submitted By: Jonas (jonasno)
Assigned to: Nobody/Anonymous (nobody)
Summary: CVS not user friendly

Initial Comment:
Have there been any talk about converting to SVN or GIT ?
CVS is old and have begun to have age issues (on the program support side) (TortoiseCVS)

So would there be a possibility for you too convert to SVN or GIT ?
or perhaps setup a mirror of the CVS that is in SVN or GIT ?

----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=202435&aid=3153615&group_id=2435

[MinGW-notify] [ mingw-Bugs-3153599 ] g++ doesn't use default constructor properly

[SF/projects/mingw notification list <mingw-notify@li...>]

Bugs item #3153599, was opened at 2011-01-09 14:05
Message generated for change (Tracker Item Submitted) made by mixpro
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=102435&aid=3153599&group_id=2435

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: gcc
Group: None
Status: Open
Resolution: None
Priority: 5
Private: No
Submitted By: mIXpRo mixpro (mixpro)
Assigned to: Nobody/Anonymous (nobody)
Summary: g++ doesn't use default constructor properly

Initial Comment:
hii , 

#include<iostream>
using namespace std;
class rational{
        int a;
public:
        rational(int a=1):a(a){}
        ~rational(){}
        friend bool operator ==(const rational&,const rational&);
};
 
bool operator==(const rational& a,const rational& b){
        return  (a.a==b.a) ;
}
int main(){

        rational c,d;
        bool x = c==d;
        bool y = 1==c;
        cout << x << y <<endl;
        return 0;
}

i tried to compile this code on g++ without any other flag , but gave me :
couldn't find operator== for (1==c) or something like that . 
but this code should work properly , i have a default constructor .
so when callinng (1==c) 1 should be casted as a rational and compared .
i tried this in linux terminal and it worked fine .

----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=102435&aid=3153599&group_id=2435

[MinGW-notify] [ mingw-Bugs-3147803 ] No error message on failed download

[SF/projects/mingw notification list <mingw-notify@li...>]

Bugs item #3147803, was opened at 2010-12-29 21:17
Message generated for change (Comment added) made by keithmarshall
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=102435&aid=3147803&group_id=2435

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: MinGW Installer
>Group: Known Feature
Status: Open
Resolution: None
>Priority: 3
Private: No
Submitted By: Mike Hudson (belgarionthekin)
>Assigned to: Keith Marshall (keithmarshall)
Summary: No error message on failed download

Initial Comment:
I apologize if this is a duplicate bug.

I spent over 20 minutes wondering why I couldn't use MinGW.  Finally it dawned on me.

The download of the files with mingw-get failed because I am behind a corporate firewall/proxy that requires authentication.  EVERYTHING that wants to connect to the internet needs to authenticate to the proxy.  All the browsers pop up an authentication window so I can put in my creds.

The problem is that the mingw-get window's messages flew by so fast that I couldn't see what was said.  Then, the shell window just went away and I couldn't scroll up to see the errors.

I was able to connect to a different network to download the proper file, but it would have been nice to know that I needed to 20 minutes ago.

----------------------------------------------------------------------

>Comment By: Keith Marshall (keithmarshall)
Date: 2011-01-08 23:30

Message:
Thank you for the report; I am aware of some download issues, but a formal
bug report does no harm.

You actually raise three separate, (although somewhat related issues):--

1) If downloads are blocked by a firewall, then presumably that's because
the firewall has been configured to block the host domain; it's doing its
intended job, and we can do nothing to work around it.  (I don't think this
is the problem, in your case).

2) Currently, mingw-get supports only one mechanism for authenticating to
a proxy -- the Windows standard method where the credentials for each user
are preconfigured, (e.g. by Internet Explorer setup), and recorded in the
registry.  That method works absolutely fine for me, behind the
authenticating Squid proxy in my place of work; apparently it doesn't work
for you, so you must be faced by a more draconian proxy configuration than
I am.  Ultimately, I would like to be able to support users with more
difficult proxy configurations, but since I cannot reproduce your problem
myself, I may need some assistance with specifying, coding and testing the
alternatives.  In the present alpha phase of development, this isn't a high
priority for me, but I will willingly consider patches.

3) mingw-get's diagnostics, particularly those related to download issues,
could be improved; in some respects, there is too much informational noise,
while some more serious issues may not be adequately reported.  This is on
my to-do list, but again, it isn't my highest priority at present.

----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=102435&aid=3147803&group_id=2435

[MinGW-notify] [ mingw-Bugs-3153453 ] Segmentation fault building Qt Creator 2.0.1

[SF/projects/mingw notification list <mingw-notify@li...>]

Bugs item #3153453, was opened at 2011-01-08 21:20
Message generated for change (Tracker Item Submitted) made by bscott45
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=102435&aid=3153453&group_id=2435

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: gcc
Group: None
Status: Open
Resolution: None
Priority: 5
Private: No
Submitted By: Bob Scott (bscott45)
Assigned to: Nobody/Anonymous (nobody)
Summary: Segmentation fault building Qt Creator 2.0.1

Initial Comment:
Using Qt Creator 2.0.1 source

>g++ -v
Using built-in specs.
COLLECT_GCC=g++
COLLECT_LTO_WRAPPER=c:/mingw/bin/../libexec/gcc/mingw32/4.5.0/lto-wrapper.exe
Target: mingw32
Configured with: ../gcc-4.5.0/configure --enable-languages=c,c++,ada,fortran,obj
c,obj-c++ --disable-sjlj-exceptions --with-dwarf2 --enable-shared --enable-libgo
mp --disable-win32-registry --enable-libstdcxx-debug --enable-version-specific-r
untime-libs --disable-werror --build=mingw32 --prefix=/mingw
Thread model: win32
gcc version 4.5.0 (GCC)

>ld -v
GNU ld (GNU Binutils) 2.21

gives error:-

mingw32-make[6]: Entering directory `E:/work/qt-creator-build/src/libs/3rdparty/
botan/src'
g++ -c -fpermissive -finline-functions -Wno-long-long -O2 -frtti -fexceptions -m
threads -Wall -DUNICODE -DQT_LARGEFILE_SUPPORT -DIDE_LIBRARY_BASENAME=\"lib\" -D
QT_NO_CAST_TO_ASCII -DQT_USE_FAST_OPERATOR_PLUS -DQT_USE_FAST_CONCATENATION -DQT
_DLL -DQT_NO_DEBUG -DQT_GUI_LIB -DQT_CORE_LIB -DQT_HAVE_MMX -DQT_HAVE_3DNOW -DQT
_HAVE_SSE -DQT_HAVE_MMXEXT -DQT_HAVE_SSE2 -DQT_THREAD_SUPPORT -I"c:\Qt\4.7.1\inc
lude\QtCore" -I"c:\Qt\4.7.1\include\QtGui" -I"c:\Qt\4.7.1\include" -I"e:\work\qt
-creator-2.0.1-src\src\libs" -I"e:\work\qt-creator-2.0.1-src\tools" -I"e:\work\q
t-creator-2.0.1-src\src\libs\3rdparty\botan\src" -I"e:\work\qt-creator-2.0.1-src
\src\libs\3rdparty\botan\build" -I"e:\work\qt-creator-2.0.1-src\src\libs\3rdpart
y\botan\build\botan" -I"c:\Qt\4.7.1\include\ActiveQt" -I"release" -I"e:\work\qt-
creator-2.0.1-src\src\libs\3rdparty\botan\src" -I"." -I"c:\qt\4.7.1\mkspecs\win3
2-g++" -o release\pk_algs.o e:\work\qt-creator-2.0.1-src\src\libs\3rdparty\botan
\src\pubkey\pk_algs.cpp
e:\work\qt-creator-2.0.1-src\src\libs\3rdparty\botan\src\pubkey\pk_algs.cpp: In
function 'Botan::Private_Key* Botan::get_private_key(const std::string&)':
e:\work\qt-creator-2.0.1-src\src\libs\3rdparty\botan\src\pubkey\pk_algs.cpp:79:1
4: internal compiler error: Segmentation fault
Please submit a full bug report,
with preprocessed source if appropriate.




----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=102435&aid=3153453&group_id=2435

[MinGW-notify] [ mingw-Bugs-3153300 ] gettextize refers to missing files in

[SF/projects/mingw notification list <mingw-notify@li...>]

Bugs item #3153300, was opened at 2011-01-08 22:31
Message generated for change (Comment added) made by vultaire
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=102435&aid=3153300&group_id=2435

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: None
Group: None
Status: Open
Resolution: None
Priority: 5
Private: No
Submitted By: Paul Goins (vultaire)
Assigned to: Nobody/Anonymous (nobody)
Summary: gettextize refers to missing files in 

Initial Comment:
I'm having problems running gettextize on a project.  

I've installed MinGW and Msys via the mingw-get-inst-20101030.exe package.  I selected the option to use more recent packages than those in the installer's packaged list.

----------------------------------------------------------------------

>Comment By: Paul Goins (vultaire)
Date: 2011-01-08 23:14

Message:
I've redone everything w/ a fresh install and after babying some
environment variables.  Although I have a problem which is kind of related,
it doesn't seem to be from gettextize.

That being said, at least let me document something strange which did
occur.

Detailed output is in the attachment.  It seems that gettextize looks for
some m4 macros, and is unable to find them.  This doesn't seem to break
things, but it at least makes me a bit concerned.

Specifically, gettextize is trying to copy files from
C:/msys/1.0/var/tmp/gettext-reloc-yA4108/mingw/share/aclocal/.  I checked
the gettextize source, and the path is specified directly in the vars at
the top of the script.  (msys, by the way, is at c:/mingw/msys on my
system, so it's a total miss...)

As I say, this seems to be a "works for me" issue even though the m4
macros are not found.  But in case investigation is desired, here's more
details:

-----

gettextize modification date: 2009-JUL-26, 9:20 AM

Gettext version info:

  /mingw/bin/gettextize (GNU gettext-tools) 0.17
  Copyright (C) 1995-1998, 2000-2007 Free Software Foundation, Inc.
  License GPLv3+: GNU GPL version 3 or later
<http://gnu.org/licenses/gpl.html>;
  This is free software: you are free to change and redistribute it.
  There is NO WARRANTY, to the extent permitted by law.
  Written by Ulrich Drepper

MinGW shell "uname -a", in case it's relevant:

  MINGW32_NT-6.1 CORE 1.0.16(0.48/3/2) 2010-09-29 00:07 i686 Msysw


----------------------------------------------------------------------

Comment By: Paul Goins (vultaire)
Date: 2011-01-08 22:32

Message:
Sorry, not used to the tracker here.  I seem to have committed the ticket
before I was finished documenting it.  Please wait...

----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=102435&aid=3153300&group_id=2435

[MinGW-notify] [ mingw-Bugs-3153300 ] gettextize refers to missing files in

[SF/projects/mingw notification list <mingw-notify@li...>]

Bugs item #3153300, was opened at 2011-01-08 22:31
Message generated for change (Comment added) made by vultaire
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=102435&aid=3153300&group_id=2435

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: None
Group: None
Status: Open
Resolution: None
Priority: 5
Private: No
Submitted By: Paul Goins (vultaire)
Assigned to: Nobody/Anonymous (nobody)
Summary: gettextize refers to missing files in 

Initial Comment:
I'm having problems running gettextize on a project.  

I've installed MinGW and Msys via the mingw-get-inst-20101030.exe package.  I selected the option to use more recent packages than those in the installer's packaged list.

----------------------------------------------------------------------

>Comment By: Paul Goins (vultaire)
Date: 2011-01-08 22:32

Message:
Sorry, not used to the tracker here.  I seem to have committed the ticket
before I was finished documenting it.  Please wait...

----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=102435&aid=3153300&group_id=2435

[MinGW-notify] [ mingw-Bugs-3153300 ] gettextize refers to missing files in

[SF/projects/mingw notification list <mingw-notify@li...>]

Bugs item #3153300, was opened at 2011-01-08 22:31
Message generated for change (Tracker Item Submitted) made by vultaire
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=102435&aid=3153300&group_id=2435

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: None
Group: None
Status: Open
Resolution: None
Priority: 5
Private: No
Submitted By: Paul Goins (vultaire)
Assigned to: Nobody/Anonymous (nobody)
Summary: gettextize refers to missing files in 

Initial Comment:
I'm having problems running gettextize on a project.  

I've installed MinGW and Msys via the mingw-get-inst-20101030.exe package.  I selected the option to use more recent packages than those in the installer's packaged list.

----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=102435&aid=3153300&group_id=2435

[MinGW-notify] [ mingw-Patches-3015199 ] pformat: fix %a for small double (bug 2997007)

[SF/projects/mingw notification list <mingw-notify@li...>]

Patches item #3015199, was opened at 2010-06-12 13:36
Message generated for change (Comment added) made by keithmarshall
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=302435&aid=3015199&group_id=2435

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: runtime
>Group: Patch committed
>Status: Closed
>Resolution: Accepted
Priority: 5
Private: No
Submitted By: Jacky Lai (crazyjacky)
Assigned to: Keith Marshall (keithmarshall)
Summary: pformat: fix %a for small double (bug 2997007)

Initial Comment:
Refer to bug 2997007: https://sourceforge.net/tracker/?func=detail&aid=2997007&group_id=2435&atid=102435
The problem is in __pformat_xdouble(), which checks denormalized number incorrectly .
But perhaps the bug is best tackled by directly using __pformat_xldouble() to handle all floating point types when %a specifier is detected
Attached patch do this.

----------------------------------------------------------------------

>Comment By: Keith Marshall (keithmarshall)
Date: 2011-01-07 23:04

Message:
Thanks for the reminder.

I hadn't forgotten it; just hadn't found time to get around to it.  It's
committed now.

----------------------------------------------------------------------

Comment By: Keith Marshall (keithmarshall)
Date: 2011-01-07 23:04

Message:
Your patch has been accepted and is now commited to CVS.  You should expect
to see it in the next release.

----------------------------------------------------------------------

Comment By: Jacky Lai (crazyjacky)
Date: 2011-01-06 14:08

Message:
Ping, before this patch is forgotten

----------------------------------------------------------------------

Comment By: Jacky Lai (crazyjacky)
Date: 2010-06-14 15:32

Message:
> > The two patches are independent.
> 
> Thanks. I can see that. What I really wanted was confirmation that the
> second patch will not actually fix the problem it addresses, if I don't
> also apply the first?

The first patch only fixes the case of "%La" with denormal long double.
And the reason is that denormal long doubles is not handled correctly in
__pformat_xldouble().
This second patch aims at fixing the case of "%a" with normal/denormal
double. These doubles are normal numbers in long double, which is correctly
handled in __pformat_xldouble(), even without the first patch.
So without the first one applied, this patch can still do the job.

> > And both patch files are prepared against
> > the current svn revision.
> 
> Really? We have an SVN repository? It's the first I've heard of it; our
> master code is in a CVS repository, on sourceware.org

Yes you are right. I must be typing too much "svn" in these few days.


----------------------------------------------------------------------

Comment By: Keith Marshall (keithmarshall)
Date: 2010-06-13 20:29

Message:
> The two patches are independent.

Thanks.  I can see that.  What I really wanted was confirmation that the
second patch will not actually fix the problem it addresses, if I don't
also apply the first?

> And both patch files are prepared against
> the current svn revision.

Really?  We have an SVN repository?  It's the first I've heard of it; our
master code is in a CVS repository, on sourceware.org

----------------------------------------------------------------------

Comment By: Jacky Lai (crazyjacky)
Date: 2010-06-13 13:03

Message:
The two patches are independent. And both patch files are prepared against
the current svn revision.

----------------------------------------------------------------------

Comment By: Keith Marshall (keithmarshall)
Date: 2010-06-12 22:32

Message:
Thanks again, Jacky.

I assume this also requires me to apply your earlier patch?
http://sourceforge.net/tracker/?func=detail&aid=3000040&group_id=2435&atid=302435

----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=302435&aid=3015199&group_id=2435

[MinGW-notify] [ mingw-Patches-3000040 ] pformat: fix long double subnormal handling

[SF/projects/mingw notification list <mingw-notify@li...>]

Patches item #3000040, was opened at 2010-05-11 15:52
Message generated for change (Comment added) made by keithmarshall
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=302435&aid=3000040&group_id=2435

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: runtime
>Group: Patch committed
>Status: Closed
>Resolution: Accepted
Priority: 5
Private: No
Submitted By: Jacky Lai (crazyjacky)
Assigned to: Keith Marshall (keithmarshall)
Summary: pformat: fix long double subnormal handling

Initial Comment:
The attached test case, when compiled with:
 gcc -D__USE_MINGW_ANSI_STDIO -o test testcase.c
gives incorrect output:
0x8p-16385
0x2p-16384
when the correct output should be something equivalent to:
0x8p-16385
0x8p-16387

The subnormal number is handled incorrectly by treating it as the normal ones, in __pformat_xldouble(). The attached patch fixes this.

----------------------------------------------------------------------

Comment By: Keith Marshall (keithmarshall)
Date: 2011-01-07 22:59

Message:
Your patch has been accepted and is now commited to CVS.  You should expect
to see it in the next release.

----------------------------------------------------------------------

Comment By: Keith Marshall (keithmarshall)
Date: 2010-06-12 22:25

Message:
Don't know why the previous version didn't work for me; the test case
downloaded okay, at the time.  Anyway, I can read it now.  Thanks.

----------------------------------------------------------------------

Comment By: Jacky Lai (crazyjacky)
Date: 2010-06-11 15:58

Message:
I can download and read the original patch file. Maybe intermittent network
problem?
Anyway I uploaded a regenerated patch.

----------------------------------------------------------------------

Comment By: Keith Marshall (keithmarshall)
Date: 2010-06-11 08:54

Message:
Thanks for this, Jacky; unfortunately, the patch file seems to be
inaccessible.

----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=302435&aid=3000040&group_id=2435

[MinGW-notify] [ mingw-Patches-3015199 ] pformat: fix %a for small double (bug 2997007)

[SF/projects/mingw notification list <mingw-notify@li...>]

Patches item #3015199, was opened at 2010-06-12 21:36
Message generated for change (Comment added) made by crazyjacky
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=302435&aid=3015199&group_id=2435

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: runtime
Group: Patch under review
Status: Open
Resolution: None
Priority: 5
Private: No
Submitted By: Jacky Lai (crazyjacky)
Assigned to: Keith Marshall (keithmarshall)
Summary: pformat: fix %a for small double (bug 2997007)

Initial Comment:
Refer to bug 2997007: https://sourceforge.net/tracker/?func=detail&aid=2997007&group_id=2435&atid=102435
The problem is in __pformat_xdouble(), which checks denormalized number incorrectly .
But perhaps the bug is best tackled by directly using __pformat_xldouble() to handle all floating point types when %a specifier is detected
Attached patch do this.

----------------------------------------------------------------------

>Comment By: Jacky Lai (crazyjacky)
Date: 2011-01-06 22:08

Message:
Ping, before this patch is forgotten

----------------------------------------------------------------------

Comment By: Jacky Lai (crazyjacky)
Date: 2010-06-14 23:32

Message:
> > The two patches are independent.
> 
> Thanks. I can see that. What I really wanted was confirmation that the
> second patch will not actually fix the problem it addresses, if I don't
> also apply the first?

The first patch only fixes the case of "%La" with denormal long double.
And the reason is that denormal long doubles is not handled correctly in
__pformat_xldouble().
This second patch aims at fixing the case of "%a" with normal/denormal
double. These doubles are normal numbers in long double, which is correctly
handled in __pformat_xldouble(), even without the first patch.
So without the first one applied, this patch can still do the job.

> > And both patch files are prepared against
> > the current svn revision.
> 
> Really? We have an SVN repository? It's the first I've heard of it; our
> master code is in a CVS repository, on sourceware.org

Yes you are right. I must be typing too much "svn" in these few days.


----------------------------------------------------------------------

Comment By: Keith Marshall (keithmarshall)
Date: 2010-06-14 04:29

Message:
> The two patches are independent.

Thanks.  I can see that.  What I really wanted was confirmation that the
second patch will not actually fix the problem it addresses, if I don't
also apply the first?

> And both patch files are prepared against
> the current svn revision.

Really?  We have an SVN repository?  It's the first I've heard of it; our
master code is in a CVS repository, on sourceware.org

----------------------------------------------------------------------

Comment By: Jacky Lai (crazyjacky)
Date: 2010-06-13 21:03

Message:
The two patches are independent. And both patch files are prepared against
the current svn revision.

----------------------------------------------------------------------

Comment By: Keith Marshall (keithmarshall)
Date: 2010-06-13 06:32

Message:
Thanks again, Jacky.

I assume this also requires me to apply your earlier patch?
http://sourceforge.net/tracker/?func=detail&aid=3000040&group_id=2435&atid=302435

----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=302435&aid=3015199&group_id=2435

[MinGW-notify] [ mingw-Bugs-3152142 ] PAGE_WRITECOMBINE missing in winnt.h

[SF/projects/mingw notification list <mingw-notify@li...>]

Bugs item #3152142, was opened at 2011-01-05 23:44
Message generated for change (Tracker Item Submitted) made by jonasno
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=102435&aid=3152142&group_id=2435

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: w32api
Group: None
Status: Open
Resolution: None
Priority: 5
Private: No
Submitted By: Jonas (jonasno)
Assigned to: Nobody/Anonymous (nobody)
Summary: PAGE_WRITECOMBINE missing in winnt.h

Initial Comment:

PAGE_WRITECOMBINE is missing in the 'WinNT.h' file.

"#define PAGE_WRITECOMBINE 0x400"

It should be under the line "#define PAGE_NOCACHE		0x0200"

More information see link below.
Link: http://msdn.microsoft.com/en-us/library/aa366786%28v=vs.85%29.aspx


----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=102435&aid=3152142&group_id=2435

[MinGW-notify] [ mingw-Bugs-3092806 ] gcc: CreateProcess: No such file or directory

[SF/projects/mingw notification list <mingw-notify@li...>]

Bugs item #3092806, was opened at 2010-10-22 10:25
Message generated for change (Comment added) made by jonasno
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=102435&aid=3092806&group_id=2435

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: gcc
Group: None
Status: Open
Resolution: None
Priority: 5
Private: No
Submitted By: grischka (grischka)
Assigned to: Nobody/Anonymous (nobody)
Summary: gcc: CreateProcess: No such file or directory

Initial Comment:
Is it maybe possible that instead of some low-level win32 API function name at least the real file name is included in this error message, that is WHAT(TF) "file or directory" gcc actually cannot find?  After all to start with gcc on windows you're not required to know CreateProcess.


----------------------------------------------------------------------

Comment By: Jonas (jonasno)
Date: 2011-01-05 23:43

Message:
I'm guessing you want GCC / g++ to report what file that is missing ?

Then i agree, it should report the file name.
Ex,
"gcc: CreateProcess( <file name here> ): No such file or directory "
Ex 2
"gcc: CreateProcess: <file name here>, no such file or directory "

----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=102435&aid=3092806&group_id=2435

[MinGW-notify] [ mingw-Bugs-3149468 ] Incorrect name of FreeLibraryAndExitThread in libkernel32.a

[SF/projects/mingw notification list <mingw-notify@li...>]

Bugs item #3149468, was opened at 2011-01-02 10:08
Message generated for change (Tracker Item Submitted) made by yak1ex
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=102435&aid=3149468&group_id=2435

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: w32api
Group: None
Status: Open
Resolution: None
Priority: 5
Private: No
Submitted By: Yak! (yak1ex)
Assigned to: Nobody/Anonymous (nobody)
Summary: Incorrect name of FreeLibraryAndExitThread in libkernel32.a

Initial Comment:
FreeLibraryAndExitThread is exported as FreeLibraryAndExitThread@... in libkernel32.a
It should be FreeLibraryAndExitThread@... because its function signature is "VOID WINAPI FreeLibraryAndExitThread(HMODULE, DWORD)".

Reference:
http://msdn.microsoft.com/en-us/library/ms683153(VS.85).aspx
http://cygwin.com/cgi-bin/cvsweb.cgi/src/winsup/w32api/lib/kernel32.def?rev=1.32&content-type=text/x-cvsweb-markup&cvsroot=src

----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=102435&aid=3149468&group_id=2435