Cygwin update hanging on mingw-runtime

[Michael Ash]

We are attempting to use setup.exe to update our previously working
Cygwin installation.  The download completes but the installation
hangs at 25% on the following:

Installing
mingw-runtime-3.12-4
/usr/include/mingw/assert.h

After half an hour hanging at this point, the following appear in a Windows box:

"Dr. Watson Postmortem Debugger has encountered a problem and needs to
close.  We are sorry for the inconvenience.  If you were in the middle
of something, the information you were working on might be lost."

The error report gave the following information, in case it's of any use to you:

EventType: BEX
P1: drwtsn32.exe
P2: 5.1.2600.0
P3: 3b7d84a2
P4: dbghelp.dll
P5: 5.1.2600.2180
P6: 4110969a
P7: 0001295d
P8: c409
P9: 

At this point, the screen is frozen, and the Task Manager indicates
that the Cygwin installation is not responding.


I have a couple of suspicions:
(1) I also installed MINGW directly on the computer.  Maybe the
Cygwin/MINGW and the main MINGW are interacting badly.
(2) Maybe there Cygwin services running that are interfering with the
update.  We used cygrunsrv to stop sshd which was the only service
that we thought we were running.
(3) We have occasionally used the command-line program cyg-apt to
update and install.  Maybe this has fouled the database.


Our failed update seems to have messed up other parts of cygwin.  We
can no longer start sshd with 'cygrunsrv --start sshd'  and we can't
seem to locate ssh-host-config

We have attempted to select "skip" for mingw in setup.exe, but this
doesn't affect the hang described above.

Advice welcome.

Best,
Michael

--
Michael Ash, Associate Professor
 of Economics and Public Policy
Department of Economics and CPPA
University of Massachusetts

--
Unsubscribe info:  http://cygwin.com/ml/#unsubscribe-simple
Problem reports:   http://cygwin.com/problems.html
Documentation: http://cygwin.com/docs.html
FAQ:   http://cygwin.com/faq/

Re: Cygwin update hanging on mingw-runtime

[Brian Dessent]

Michael Ash wrote:

> We are attempting to use setup.exe to update our previously working
> Cygwin installation.  The download completes but the installation
> hangs at 25% on the following:
> 
> Installing
> mingw-runtime-3.12-4
> /usr/include/mingw/assert.h
> 
> After half an hour hanging at this point, the following appear in a Windows 
> box:
> 
> "Dr. Watson Postmortem Debugger has encountered a problem and needs to
> close.  We are sorry for the inconvenience.  If you were in the middle
> of something, the information you were working on might be lost."
> 
> The error report gave the following information, in case it's of any use to 
> you:
> 
> EventType: BEX
> P1: drwtsn32.exe
> P2: 5.1.2600.0
> P3: 3b7d84a2
> P4: dbghelp.dll
> P5: 5.1.2600.2180
> P6: 4110969a
> P7: 0001295d
> P8: c409
> P9: 

Sorry, but the above is meaningless.  You'll need to debug the actual
problem a little farther.  You can also try a setup.exe snapshot to see
if it works any better.  Alternatively, you can easily install this
package by hand:

tar jxvf path/to/mingw-runtime-3.12-4.tar.bz2 -C /
sed -ie 's/\(mingw-runtime\)-.*\(\.tar\.bz2\)/\1-3.12-4\2/' \
/etc/setup/installed.db

> At this point, the screen is frozen, and the Task Manager indicates
> that the Cygwin installation is not responding.

My guess is that this is triggering some kind of virus scanner's false
positive.

> I have a couple of suspicions:
> (1) I also installed MINGW directly on the computer.  Maybe the
> Cygwin/MINGW and the main MINGW are interacting badly.

This seems extremely unlikely because this package contains only headers
and static import libraries, both of which are only used when compiling
and linking MinGW objects, not at runtime.  (Okay, it also contains the
stub DLL mingwm10.dll which is the TLS-destructor stub helper, but if
this were in use at the time setup ran, it would not be able to touch
it, only schedule it for replacement at reboot.  If you're using the
latest version of setup you'll get a dialog in this situation.  Not to
mention that any separate installation of MinGW would have its own copy
of mingwm10.dll.)

> (2) Maybe there Cygwin services running that are interfering with the
> update.  We used cygrunsrv to stop sshd which was the only service
> that we thought we were running.

I can't see how this would happen.  Nor do I see how sshd could possibly
be relevant at all since it's not even a MinGW app.

> (3) We have occasionally used the command-line program cyg-apt to
> update and install.  Maybe this has fouled the database.

That is possible but I don't see why it would cause setup to foul up at
this point in the process.

> Our failed update seems to have messed up other parts of cygwin.  We
> can no longer start sshd with 'cygrunsrv --start sshd'  and we can't
> seem to locate ssh-host-config

Uh, suddenly you're describing a much different problem than just being
able to unpack the the mingw-runtime package.  It sounds like you have
bigger issues that need solving.

> We have attempted to select "skip" for mingw in setup.exe, but this
> doesn't affect the hang described above.

If you selected 'skip' then setup should not be unpacking anything from
the package, so I don't see how it could fail trying to unpack from the
package, unless you're saying it fails trying to unpack some other
package.  If that's the case then the fact that it happens to foul up
unpacking the mingw-runtime package is a complete red herring, and you
have some kind of bigger problem that's interfering with setup that
you'll need to diagnose.

Brian

--
Unsubscribe info:  http://cygwin.com/ml/#unsubscribe-simple
Problem reports:   http://cygwin.com/problems.html
Documentation: http://cygwin.com/docs.html
FAQ:   http://cygwin.com/faq/

using regtool to modify remote registries?

[Richard Foulk]

Can someone point me to documentation or examples of using regtool to
modify remote user registries?

The man pages and other documentation don't quite explain things.  The
get and set commands seem likely, but don't work by themselves.  The
load, unload and save commands seem likely, but the documentation on
them is cryptic and terse.  Plenty of searching with google didn't help
much either ...


Thanks,

Richard


--
Unsubscribe info:  http://cygwin.com/ml/#unsubscribe-simple
Problem reports:   http://cygwin.com/problems.html
Documentation: http://cygwin.com/docs.html
FAQ:   http://cygwin.com/faq/

SIGINT not delivered on Ctrl-C

[Chuck Taylor]

Does anyone know why a SIGINT signal would fail to be
delivered to a Cygwin-linked process (via Ctrl-C) in
the scenario where all of the following conditions
hold?

1) Process in question is launched from a
non-Cygwin-linked program (such as native Windows
command prompt), with all of stdin, stdout, and stderr
redirected.

2) Process registers a SIGINT handler via Cygwin's
implementation of signal() and waits for signal to be
delivered.

Attaching a WinDBG to the process in question shows
that the process is indeed receiving the Ctrl-C
exception from Windows.  However, program behavior
proves that the SIGINT handler is not being called. 
If any of the stdio channels are left unredirected, or
if process is launched from a Cygwin-linked program
(e.g. from bash instead of non-Cygwin-linked program),
then SIGINT is delivered successfully.


__
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 

--
Unsubscribe info:  http://cygwin.com/ml/#unsubscribe-simple
Problem reports:   http://cygwin.com/problems.html
Documentation: http://cygwin.com/docs.html
FAQ:   http://cygwin.com/faq/

Re: DualCores and Current Cygwin problems

[William Sutton]

I'm running Cygwin, Windows XP professiona 2002 SP2, and Norton AntiVirus 
(with the firewall turned off) without problems on an Intel Core2 Duo 2.13 
GHz system.

-- 
William Sutton


On Thu, 10 May 2007, Brian Salter-Duke wrote:

> On Wed, May 09, 2007 at 10:07:20AM -0400, Christopher Faylor wrote:
> > On Wed, May 09, 2007 at 09:54:36PM +0800, Chee Kiang Goh wrote:
> > >Looking forward nevertheless to a better co-existence solution with
> > >WindowXP/DualCoreCPU.  This feature had been the primary reason why I
> > >stick to cygwin over the years :>
> > 
> > FYI, there is no better coexistence solution contemplated.
> > 
> > PTC, although, you have to wonder why we'd have to accommodate virus
> > checkers or firewalls, which are supposed to be unobtrusive.
> 
> It is indeed a bit of a worry. I have just installed Cygwin and Norton
> securities on a new dual-core laptop. I have seen no problems so far. Is
> there anything in particular I should look for?
> 
> Brian.
>  
> > cgf
> > 
> > --
> > Unsubscribe info:  http://cygwin.com/ml/#unsubscribe-simple
> > Problem reports:   http://cygwin.com/problems.html
> > Documentation: http://cygwin.com/docs.html
> > FAQ:   http://cygwin.com/faq/
> > 
> 
> 

--
Unsubscribe info:  http://cygwin.com/ml/#unsubscribe-simple
Problem reports:   http://cygwin.com/problems.html
Documentation: http://cygwin.com/docs.html
FAQ:   http://cygwin.com/faq/

Re: can not exec cc1.exe

[Jim Marshall]

Brian Dessent wrote:

Jim Marshall wrote:


Here is the cygcheck -svr output.


Hmm, you're using a local NTFS drive and $CYGWIN is not set to anything
strage.  So no more clues there.

What are the attributes of the .lnk files?  What happens if you manually
delete them and recreate them as working symlinks by hand?

Brian


Brian,
 Again many thanks for the help. I manually created the links and it 
resulted in the same behavior.I did some more looking around and it 
appeared that the /usr/lib/gcc/3.4.4/i686-pc-mingw directory suffered 
the same problem (it had .lnk files). So I went into the cygwin 
setup.exe and uninstalled all of the mingw stuff.  Setup complained that 
parts of the main gcc stuff required mingw tools, I unchecked the box so 
that setup would remove the mingw stuff anyway. When I looked on the HD 
all of the mingw directories were still there. I manually deleted all 
the files  in these directories (/usr/i686-pc-mingw and 
/usr/lib/gcc/i686-pc-mingw32).  I then reran the cygwin setup and had it 
install the mingw tools. This fixed the problem. I am presuming that the 
setup was not actually erasing the mingw files from the HD (as evidenced 
by what I saw) so that was causing the problem in that when setup was 
not properly over-writing the existing files.


Again thanks.
-Jim



--
Unsubscribe info:  http://cygwin.com/ml/#unsubscribe-simple
Problem reports:   http://cygwin.com/problems.html
Documentation: http://cygwin.com/docs.html
FAQ:   http://cygwin.com/faq/

Change SFTP UMask

[John J. Culkin]

Hello

Does anyone know of an easy way to change the UMask of SFTP on CYGWIN? 
For one user and/or globally?


For example when I create a via via SFTP, I want its permissions to be 
774 instead 664 (the default)


Did anyone try anything like this? http://sftplogging.sourceforge.net/

Any information would be very helpful.

Thanks,

-- John C.

--
John J. Culkin  Systems Administrator
[EMAIL PROTECTED]   The University of Scranton
Phone: (570) 941-7665


--
Unsubscribe info:  http://cygwin.com/ml/#unsubscribe-simple
Problem reports:   http://cygwin.com/problems.html
Documentation: http://cygwin.com/docs.html
FAQ:   http://cygwin.com/faq/

RE: Change SFTP UMask

[Dave Korn]

On 10 May 2007 15:58, John J. Culkin wrote:

> Hello
> 
> Does anyone know of an easy way to change the UMask of SFTP on CYGWIN?
> For one user and/or globally?
> 
> For example when I create a via via SFTP, I want its permissions to be
> 774 instead 664 (the default)

  Set it in the .bashrc of the user you're logging in as at the destination
end.


cheers,
  DaveK
-- 
Can't think of a witty .sigline today


--
Unsubscribe info:  http://cygwin.com/ml/#unsubscribe-simple
Problem reports:   http://cygwin.com/problems.html
Documentation: http://cygwin.com/docs.html
FAQ:   http://cygwin.com/faq/

RE: Change SFTP UMask

[Dave Korn]

On 10 May 2007 16:04, Dave Korn wrote:

> On 10 May 2007 15:58, John J. Culkin wrote:
> 
>> Hello
>> 
>> Does anyone know of an easy way to change the UMask of SFTP on CYGWIN? For
>> one user and/or globally? 
>> 
>> For example when I create a via via SFTP, I want its permissions to be
>> 774 instead 664 (the default)
> 
>   Set it in the .bashrc of the user you're logging in as at the destination
> end. 

  Erk.  Or possibly the .bash_profile; I'm not sure whether sftp counts as an
interactive login and therefore precisely which startup scripts it would run.


cheers,
  DaveK
-- 
Can't think of a witty .sigline today


--
Unsubscribe info:  http://cygwin.com/ml/#unsubscribe-simple
Problem reports:   http://cygwin.com/problems.html
Documentation: http://cygwin.com/docs.html
FAQ:   http://cygwin.com/faq/

Re: Change SFTP UMask

[John J. Culkin]

I tried putting UMask commands in both places and neither seem to have 
any effect on files created by sftp


-- John C.

--
John J. Culkin  Systems Administrator
[EMAIL PROTECTED]   The University of Scranton
Phone: (570) 941-7665


--
Unsubscribe info:  http://cygwin.com/ml/#unsubscribe-simple
Problem reports:   http://cygwin.com/problems.html
Documentation: http://cygwin.com/docs.html
FAQ:   http://cygwin.com/faq/

gcc packaging bug [was: Re: can not exec cc1.exe]

[Brian Dessent]

Jim Marshall wrote:

> all of the mingw directories were still there. I manually deleted all
> the files  in these directories (/usr/i686-pc-mingw and
> /usr/lib/gcc/i686-pc-mingw32).  I then reran the cygwin setup and had it
> install the mingw tools. This fixed the problem. I am presuming that the
> setup was not actually erasing the mingw files from the HD (as evidenced
> by what I saw) so that was causing the problem in that when setup was
> not properly over-writing the existing files.

I'm glad it's now fixed.  But for the mingw-gcc-* packages, setup does
not actually create or delete any of those links, it is all done by
postinstall/preremove files.  If you take a look (see
/etc/{preremove,postinstall}/gcc-mingw*) you will see that the preremove
script only deletes the files contained in the manifest, and the
symlinks are not included there; and the postinstall only tries to
create the symlinks if they don't exist.  This could be considered a
slight packaging bug I suppose, since once created, those symlinks will
never be removed/recreated on reinstalls.

Brian

--
Unsubscribe info:  http://cygwin.com/ml/#unsubscribe-simple
Problem reports:   http://cygwin.com/problems.html
Documentation: http://cygwin.com/docs.html
FAQ:   http://cygwin.com/faq/

su: cannot set groups: Invalid argument

[Ben Atkin]

I've been using the standard cygwin distribution for setting up
ssh/sftp on Windows Server 2003. It's been working fine for the
Administrator account, but now I'd like to set up another account and
use that, just in case people see ssh running on a windows computer
and try running a password guesser on the Administrator account.

There is a domain account that I'd like to use. I can get to it if I
ssh into the domain controller, but it won't work for other computers
on the domain. I searched for a solution, and I found mkpasswd. I ran
"mkpasswd -d (domain) -u (user)" and it worked without error and so I
appended the entry to /etc/passwd.

To test it, I tried doing "su (username)". It prompts me for the
password, and after I enter the password, it says "su: cannot set
groups: Invalid argument". Does anyone have any idea why this error
might occur? I searched for my error message to no avail.

Thanks,

Ben Atkin

--
Unsubscribe info:  http://cygwin.com/ml/#unsubscribe-simple
Problem reports:   http://cygwin.com/problems.html
Documentation: http://cygwin.com/docs.html
FAQ:   http://cygwin.com/faq/

Re: Please Help!! Calling Socket function in a dll file (that is created using cygwin library), by a Microsoft Visual C++ program resulting in infinite loop

[kalasad mailu]

This question is also regarding my efforts to create a dll file using
g++ in cygwin and use it in VC++.

The below program works fine and the output is also shown below. But
IF I UNCOMMENT THE TWO LINES in my file dtest.cpp i.e.

//#include
//std::cout<<"hello naumskara\n";

and do the same steps. Dll is created successfully, and the MSVC++
compiles fine. But when I try to run it, I get the following msg when
it tries to execute the init(); (to initialize the cygwin environment)
function(MSVC++ file):

First-chance exception at 0x610b48b6 in new3.exe: 0xC005: Access
violation reading location 0x0004.
-

I am running the MSVC++ program from the IDE.

Could you please suggest me how to make the program(MSVC++) work after
un-commenting the line.

Thanks.


--
dtest.cpp FILE THAT I AM USING TO CREATE THE DLL FILE
--
#include
//#include
#define SOCKETTEST_BUILD_DLL
#include "dtest.h"
int called()
{
printf("Ctor 400 called\n");
//std::cout<<"hello naumskara\n";
return 400;
}

int c = called();

DLL_IMPORT_EXPORT int test_dll()
{
printf("%d printed from dll",c);
return c;
}

--
dtest.h FILE THAT I AM USING TO CREATE THE DLL FILE
--
#ifdef SOCKETTEST_BUILD_DLL
#define DLL_IMPORT_EXPORT __declspec(dllexport)
#else
#define DLL_IMPORT_EXPORT __declspec(dllimport)
#endif

extern "C" {
DLL_IMPORT_EXPORT int test_dll();
}

---
COMMANDS THAT I USED TO CREATE THE DLL AND LIB FILES
---

$ g++ -shared -o dtestdll.dll dtest.cpp
-Wl,--output-def,dtestdll.def,--out-implib,libdtestdll.a

C:\Program Files\Microsoft Visual Studio .NET 2003\Vc7\bin\lib"
/machine:i386 /DEF:dtestdll.def


--
MSVC++ PROGRAM USING THE DLL FILE, THAT IS CREATED IN CYGWIN USING g++
--

#include
#include "dtest.h"
#include

void main(){
HMODULE h = LoadLibrary("cygwin1.dll");
void (*init)() = (void(__cdecl *)(void))GetProcAddress(h, 
"cygwin_dll_init");
init();
   int k= test_dll();
printf("\nthe OSA value returned is %d\n", k);
Sleep(5000);

}

-
OUTPUT FROM MY MSVC++ PROGRAM:

Ctor 400 called
400 printed from dll
the OSA value returned is 400



On 5/9/07, Brian Dessent <[EMAIL PROTECTED]> wrote:

kalasad mailu wrote:

> I used the socket.h file provided by cygwin(/usr/include/cygwin/.) and
> wrote a program to create socket, made this a dll file and used this
> dll in VC++.
>
> I assume using of the socket.h( form cygwin) did all the conversion
> form the linux system calls to the windows system calls and made my
> socket program work. Please correct me if my understanding is wrong.

A header file does not implement anything.  It contains no code at all
(except in the case of C++ or inlined functions.)  It only describes an
interface that is actually implemented in a library.  In the case of
Cygwin, the standard C library is implemented by cygwin1.dll, along with
many POSIX functions.

> Now I want to try the same process for simple c++ functions like
> "cout". I don't find any cygwin header files like iostream.h or
> stream.h in the cygwin directory.

cout is part of the C++ standard library (STL) and is implemented by
libstdc++ which is part of gcc.

> When I created this stand alone program.
>
> #include 
>
> int main () {
> std::cout << "Hello World\n";
> return 0;
> }
>
> I guess this picked the header file from
> "cygwin\lib\gcc\i686-pc-cygwin\3.4.4\include\c++\" (a gcc header file)
>
> Is there no iostream header file by cygwin (/usr/include/cygwin)?

First of all, stop looking in /usr/include/cygwin for things, and stop
worrying about which directory header files are in.  That has no bearing
on what library implements a specific function, as all headers for all
installed libraries are in /usr/include.  Some of these headers are
provided by Cygwin.  The ones in the "cygwin" subdirectory are only for
Cygwin-specific things, but this is only a small fraction of the
functionality provided by Cygwin.  For example, all of the stadard C
library functions (such as stdio.h, stdlib.h, io.h, stdint.h, and on and
on) are all implemented by Cygwin, i.e. cygwin1.dll, and are in
/usr/include.  And many other libraries as well, such as zlib, gettext,
libintl, and so on.  So, stop confusing a header file with the library
that implements the code and r

Re: su: cannot set groups: Invalid argument

[Christopher Faylor]

On Thu, May 10, 2007 at 10:48:55AM -0700, Ben Atkin wrote:
>I've been using the standard cygwin distribution for setting up
>ssh/sftp on Windows Server 2003. It's been working fine for the
>Administrator account, but now I'd like to set up another account and
>use that, just in case people see ssh running on a windows computer
>and try running a password guesser on the Administrator account.
>
>There is a domain account that I'd like to use. I can get to it if I
>ssh into the domain controller, but it won't work for other computers
>on the domain. I searched for a solution, and I found mkpasswd. I ran
>"mkpasswd -d (domain) -u (user)" and it worked without error and so I
>appended the entry to /etc/passwd.
>
>To test it, I tried doing "su (username)". It prompts me for the
>password, and after I enter the password, it says "su: cannot set
>groups: Invalid argument". Does anyone have any idea why this error
>might occur? I searched for my error message to no avail.

http://cygwin.com/faq/faq-nochunks.html#faq.using.su

cgf

--
Unsubscribe info:  http://cygwin.com/ml/#unsubscribe-simple
Problem reports:   http://cygwin.com/problems.html
Documentation: http://cygwin.com/docs.html
FAQ:   http://cygwin.com/faq/

Re: DualCores and Current Cygwin problems

[Joseph Kowalski]

I think the following sad story should shed a little light on this issue.

I'm attempting to use cygwin to provide the UNIX-like build environment 
to build Java SE on a DualCore system.


Relevant configuration details:
   AMD X2 5200+
   Asus M2NPV-VM
   2 Gb ECC memory (std. clocks)

I'm working to improve the documentation for the Java SE build, so I 
know *exactly* what's on the system.  The

installations are:

   1) Windows XP, fully updated
   2) Visual Studio .NET Professional (2003)
   3) Microsoft Platform SDK (2004 - *not* R2)
   4) Microsoft DirectX SDK (Summer 2004)
   5) Sun Java 6 SDK (1.6.0_01)
   6) Cygwin (current)

That's it.  No additional software components.  None.

With this configuration, I get random "can not fork: Resource 
temporarily unavailable" errors when trying to perform the fairly large 
and complex product build.  I also get "dup_proc_pipe" failures, which 
are fairly random, but tend to be understandably associated with long 
pipes in the build process.


If I add /ONECPU to boot.ini, neatly turning my DualCore system into a 
single core system, the failures all magically disappear.


This would tend to indicate that there is a multi-threading issue either 
in cygwin or in the underlying Windows XP operating environment.


It hopefully also provides a work-around (abet a somewhat painful one).  
It's probably worth trying by all who have been reporting this failure.  
If it doesn't eliminate the failure for you, its an indication that 
there may be multiple failures involved.


- cheers and good luck (these are hard problems to track down),

- Joseph Kowalski


--
Unsubscribe info:  http://cygwin.com/ml/#unsubscribe-simple
Problem reports:   http://cygwin.com/problems.html
Documentation: http://cygwin.com/docs.html
FAQ:   http://cygwin.com/faq/

Re: DualCores and Current Cygwin problems

[René Berber]

Joseph Kowalski wrote:

I'm attempting to use cygwin to provide the UNIX-like build environment 
to build Java SE on a DualCore system.


Relevant configuration details:
   AMD X2 5200+

[snip]
With this configuration, I get random "can not fork: Resource 
temporarily unavailable" errors when trying to perform the fairly large 
and complex product build.  [snip]


The problem does not only happen with double cores, lately I'm seeing this (P4 
single core, no simulated multi-processor either):


2007-04-29 15:02:36 daemon: accept process fork failed: Resource temporarily 
unavailable
2007-04-29 15:03:36 daemon: accept process fork failed: Resource temporarily 
unavailable
2007-04-29 15:04:37 daemon: accept process fork failed: Resource temporarily 
unavailable


which is Exim complaining after it started.

My problem, I think, is a bad network driver... or at least it takes too long 
for it to "start up".  The network card is surely completely different from what 
you use (an Intel PRO/1000 MT with driver from Intel).

--
René Berber


--
Unsubscribe info:  http://cygwin.com/ml/#unsubscribe-simple
Problem reports:   http://cygwin.com/problems.html
Documentation: http://cygwin.com/docs.html
FAQ:   http://cygwin.com/faq/

Re: DualCores and Current Cygwin problems

[Chee Kiang Goh]

Good day, all,

Still no luck trying out Larry's Suggestion... my
system is a bit edgy and unstable at the moment... is
gonna be a long re-installation ... :< fortunately I
am trying this @ home and only have to live with my
wife's complaints. 

After reading more of the current and past postings, I
gather that there are lots of reports with similar
symptoms... some regarding the vista installation,
some like mine and Joseph, on the dual cores,  and
lots of others with build-failures. The interesting
part is that the underlying system/OS are typically
quite varied. 

I am not sure if these are somehow connected.
Appreciate if you can share your opinions.  

Will update as soon as my PC bounced back to normal to
continue my trials...:P

Thanks again all round!

Best Regards
Chee Kiang
>From my wife's notebook...



--- Joseph Kowalski <[EMAIL PROTECTED]> wrote:

> 
> I think the following sad story should shed a little
> light on this issue.
> 
> I'm attempting to use cygwin to provide the
> UNIX-like build environment 
> to build Java SE on a DualCore system.
> 
> Relevant configuration details:
> AMD X2 5200+
> Asus M2NPV-VM
> 2 Gb ECC memory (std. clocks)
> 
> I'm working to improve the documentation for the
> Java SE build, so I 
> know *exactly* what's on the system.  The
> installations are:
> 
> 1) Windows XP, fully updated
> 2) Visual Studio .NET Professional (2003)
> 3) Microsoft Platform SDK (2004 - *not* R2)
> 4) Microsoft DirectX SDK (Summer 2004)
> 5) Sun Java 6 SDK (1.6.0_01)
> 6) Cygwin (current)
> 
> That's it.  No additional software components. 
> None.
> 
> With this configuration, I get random "can not fork:
> Resource 
> temporarily unavailable" errors when trying to
> perform the fairly large 
> and complex product build.  I also get
> "dup_proc_pipe" failures, which 
> are fairly random, but tend to be understandably
> associated with long 
> pipes in the build process.
> 
> If I add /ONECPU to boot.ini, neatly turning my
> DualCore system into a 
> single core system, the failures all magically
> disappear.
> 
> This would tend to indicate that there is a
> multi-threading issue either 
> in cygwin or in the underlying Windows XP operating
> environment.
> 
> It hopefully also provides a work-around (abet a
> somewhat painful one).  
> It's probably worth trying by all who have been
> reporting this failure.  
> If it doesn't eliminate the failure for you, its an
> indication that 
> there may be multiple failures involved.
> 
> - cheers and good luck (these are hard problems to
> track down),
> 
> - Joseph Kowalski
> 
> 
> --
> Unsubscribe info: 
> http://cygwin.com/ml/#unsubscribe-simple
> Problem reports:  
> http://cygwin.com/problems.html
> Documentation: http://cygwin.com/docs.html
> FAQ:   http://cygwin.com/faq/
> 
> 




__ 
Yahoo! Movies - Search movie info and celeb profiles and photos. 
http://sg.movies.yahoo.com/

--
Unsubscribe info:  http://cygwin.com/ml/#unsubscribe-simple
Problem reports:   http://cygwin.com/problems.html
Documentation: http://cygwin.com/docs.html
FAQ:   http://cygwin.com/faq/

Trademark rights and copyright for "Cygwin" and logo.

[ls-cygwin-2006]

Hi All,

If I understand it right, Cygwin is trademark of redhat and the Cygwin
logo is copyright protected.

Some people I know (including me) like to have their software on CD
for archiving (to get repeatable results or be able to understand
errors in former setups), so I've written a small tool that creates a
(minimum) 3 CD-Set (1 binary, 2 source) of Cygwin. Since I already
have the tool, I'd like to distribute the CD set at installation
parties and trade fairs. Two questions arise naturally:

 1. May I use the word / product name on the cover as description /
label of the content (probably not)?

 2. May I use the logo (probably not, too) on the cover?

I've found sentences to the effect that logo and the term "Cygwin" are
copyrighted respectively trademarked, but I haven't seen a license
that would allow distributors to use those. I might have missed
something, so

 - If somebody knows about a text I've missed, would you please point
   me to it?

 - Else whom would I write for more information?

BTW, if I can't do (1) that would be especially sad: As I understand
it I can distribute binaries and source according to the respective
licenses. But I'd also like to give proper credit to the Cygwin
project, the more, the better. At the moment I don't want even to
pretend that what we distribute / give away is a product of ours: Ours
is only the service to download it and fit it on a CD (as I said:
Esp. useful for archiving or to install baselines).

( And yes I understand that Cygwin is also a commercial product --
  http://www.redhat.com/software/cygwin/. Redhat is welcome to their
  right and welcome to protecting them if/when they think the
  necessity arises. That is nothing I want to draw in doubt, just to
  forestall a discussion in that direction for the moment. I just
  wouldn't want to label the CD "Footools for Windows" or something to
  that effect and thus _not_ attribute to the project if the necessity
  doesn't arise. And it would still be strange if setup came up with
  the cygwin banner :-).

Regards -- Markus


--
Unsubscribe info:  http://cygwin.com/ml/#unsubscribe-simple
Problem reports:   http://cygwin.com/problems.html
Documentation: http://cygwin.com/docs.html
FAQ:   http://cygwin.com/faq/

Re: DualCores and Current Cygwin problems

[Joseph Kowalski]

Rene Berber wrote:

> My problem, I think, is a bad network driver... or at least it takes 
too long for it to "start up". The network card is surely completely 
different from what you use (an Intel PRO/1000 MT with driver from Intel.


Don't be too sure.  The configuration I described also has an Intel 
PRO/1000 MT with driver from Intel (via Microsoft).


Another experiment I didn't mention is I loaded the same software stack 
on to another machine with a single core processor (Shuttle FX41).  It 
also has an Intel PRO/1000 MT.  It showed no problems.


However, I suspect this is just coincidence.  My build is "network 
silent".  The network adapter shouldn't be involved.


- Joseph Kowalski

--
Unsubscribe info:  http://cygwin.com/ml/#unsubscribe-simple
Problem reports:   http://cygwin.com/problems.html
Documentation: http://cygwin.com/docs.html
FAQ:   http://cygwin.com/faq/

Re: Please Help!! Calling Socket function in a dll file (that is created using cygwin library), by a Microsoft Visual C++ program resulting in infinite loop

[Larry Hall (Cygwin)]

kalasad mailu wrote:
> This question is also regarding my efforts to create a dll file using
> g++ in cygwin and use it in VC++.
> 
> The below program works fine and the output is also shown below. But
> IF I UNCOMMENT THE TWO LINES in my file dtest.cpp i.e.
> 
> //#include
> //std::cout<<"hello naumskara\n";
> 
> and do the same steps. Dll is created successfully, and the MSVC++
> compiles fine. But when I try to run it, I get the following msg when
> it tries to execute the init(); (to initialize the cygwin environment)
> function(MSVC++ file):
> 
> 
> First-chance exception at 0x610b48b6 in new3.exe: 0xC005: Access
> violation reading location 0x0004.
> -
> 
> 
> I am running the MSVC++ program from the IDE.
> 
> Could you please suggest me how to make the program(MSVC++) work after
> un-commenting the line.

See


You're mixing and matching I/O from two different CRTs.  You can't do that.


-- 
Larry Hall  http://www.rfk.com
RFK Partners, Inc.  (508) 893-9779 - RFK Office
216 Dalton Rd.  (508) 893-9889 - FAX
Holliston, MA 01746

_

A: Yes.
> Q: Are you sure?
>> A: Because it reverses the logical flow of conversation.
>>> Q: Why is top posting annoying in email?

--
Unsubscribe info:  http://cygwin.com/ml/#unsubscribe-simple
Problem reports:   http://cygwin.com/problems.html
Documentation: http://cygwin.com/docs.html
FAQ:   http://cygwin.com/faq/

Re: Trademark rights and copyright for "Cygwin" and logo.

[Larry Hall (Cygwin)]

[EMAIL PROTECTED] wrote:
> Hi All,
> 
> If I understand it right, Cygwin is trademark of redhat and the Cygwin
> logo is copyright protected.





You should probably take this discussion to cygwin-licensing list.


-- 
Larry Hall  http://www.rfk.com
RFK Partners, Inc.  (508) 893-9779 - RFK Office
216 Dalton Rd.  (508) 893-9889 - FAX
Holliston, MA 01746

_

A: Yes.
> Q: Are you sure?
>> A: Because it reverses the logical flow of conversation.
>>> Q: Why is top posting annoying in email?

--
Unsubscribe info:  http://cygwin.com/ml/#unsubscribe-simple
Problem reports:   http://cygwin.com/problems.html
Documentation: http://cygwin.com/docs.html
FAQ:   http://cygwin.com/faq/

Re: Trademark rights and copyright for "Cygwin" and logo.

[ls-cygwin-2006]

"Larry Hall (Cygwin)" <[EMAIL PROTECTED]> writes:

> [EMAIL PROTECTED] wrote:
>> Hi All,
>> 
>> If I understand it right, Cygwin is trademark of redhat and the Cygwin
>> logo is copyright protected.
>
>
> 
>
>
> You should probably take this discussion to cygwin-licensing list.


Ooops, yes thanks. I didn't realize that there is such a list -- and
if I had, I probably wouldn't have been sure, wether it is the right
list (after all I don't want to license cygwin as you can usually
license it for closed source distribution). But yes, I'll do that 1st
thing tomorrow. 

Nonetheless I continue to be thankful for anybody's input (esp. if I
just missed a plainly visible and well known document on that topic
somewhere).

Regards -- Markus







--
Unsubscribe info:  http://cygwin.com/ml/#unsubscribe-simple
Problem reports:   http://cygwin.com/problems.html
Documentation: http://cygwin.com/docs.html
FAQ:   http://cygwin.com/faq/

Re: Please Help!! Calling Socket function in a dll file (that is created using cygwin library), by a Microsoft Visual C++ program resulting in infinite loop

[kalasad mailu]

On 5/10/07, Larry Hall (Cygwin) <[EMAIL PROTECTED]> wrote:

kalasad mailu wrote:
> This question is also regarding my efforts to create a dll file using
> g++ in cygwin and use it in VC++.
>
> The below program works fine and the output is also shown below. But
> IF I UNCOMMENT THE TWO LINES in my file dtest.cpp i.e.
>
> //#include
> //std::cout<<"hello naumskara\n";
>
> and do the same steps. Dll is created successfully, and the MSVC++
> compiles fine. But when I try to run it, I get the following msg when
> it tries to execute the init(); (to initialize the cygwin environment)
> function(MSVC++ file):
> 

>
> First-chance exception at 0x610b48b6 in new3.exe: 0xC005: Access
> violation reading location 0x0004.
> 
-
>
>
> I am running the MSVC++ program from the IDE.
>
> Could you please suggest me how to make the program(MSVC++) work after
> un-commenting the line.

See


You're mixing and matching I/O from two different CRTs.  You can't do that.


Why will it work if I comment these line. I am still mixing the
cygwin1 and msvcrt.dll?




--
Larry Hall  http://www.rfk.com
RFK Partners, Inc.  (508) 893-9779 - RFK Office
216 Dalton Rd.  (508) 893-9889 - FAX
Holliston, MA 01746

_

A: Yes.
> Q: Are you sure?
>> A: Because it reverses the logical flow of conversation.
>>> Q: Why is top posting annoying in email?

--
Unsubscribe info:  http://cygwin.com/ml/#unsubscribe-simple
Problem reports:   http://cygwin.com/problems.html
Documentation: http://cygwin.com/docs.html
FAQ:   http://cygwin.com/faq/




--
Unsubscribe info:  http://cygwin.com/ml/#unsubscribe-simple
Problem reports:   http://cygwin.com/problems.html
Documentation: http://cygwin.com/docs.html
FAQ:   http://cygwin.com/faq/

Re: DualCores and Current Cygwin problems

[René Berber]

Joseph Kowalski wrote:

[snip]
Don't be too sure.  The configuration I described also has an Intel 
PRO/1000 MT with driver from Intel (via Microsoft).


Another experiment I didn't mention is I loaded the same software stack 
on to another machine with a single core processor (Shuttle FX41).  It 
also has an Intel PRO/1000 MT.  It showed no problems.


However, I suspect this is just coincidence.  My build is "network 
silent".  The network adapter shouldn't be involved.


Interesting.  The problems I have seen only happen after booting up, so yes it 
looks like different problems.


On the other hand every thread in Cygwin has an associated UDP connection (the 
implementation uses it), so "network silent" is not quite true... but I don't 
know how fork is implemented and if it uses UDP connections also.  You could 
monitor with TCPView if any such UDP ports are being open.


The interesting part would be to know what resource is the message "Resource 
temporarily unavailable" talking about.  It could be a Windows XP-SP2 
limitation: can't open more than 10 connections in a given time span (the 
default after SP2 set in TCPIP.SYS), it was different before SP2 but that patch 
is very old and I've only seen these messages recently.


For me, it started when I updated Exim to 4.66, about a month or two ago, 
yesterday I installed version 4.67 so the problem may change again.

--
René Berber


--
Unsubscribe info:  http://cygwin.com/ml/#unsubscribe-simple
Problem reports:   http://cygwin.com/problems.html
Documentation: http://cygwin.com/docs.html
FAQ:   http://cygwin.com/faq/

Re: Please Help!! Calling Socket function in a dll file (that is created using cygwin library), by a Microsoft Visual C++ program resulting in infinite loop

[Larry Hall (Cygwin)]

kalasad mailu wrote:
> On 5/10/07, Larry Hall (Cygwin)  cygwin  com> 
> wrote:
   
 
>> kalasad mailu wrote:
>> > This question is also regarding my efforts to create a dll file using
>> > g++ in cygwin and use it in VC++.
>> >
>> > The below program works fine and the output is also shown below. But
>> > IF I UNCOMMENT THE TWO LINES in my file dtest.cpp i.e.
>> >
>> > //#include
>> > //std::cout<<"hello naumskara\n";
>> >
>> > and do the same steps. Dll is created successfully, and the MSVC++
>> > compiles fine. But when I try to run it, I get the following msg when
>> > it tries to execute the init(); (to initialize the cygwin environment)
>> > function(MSVC++ file):
>> >
>> 
>>
>> >
>> > First-chance exception at 0x610b48b6 in new3.exe: 0xC005: Access
>> > violation reading location 0x0004.
>> >
>> -
>>
>> >
>> >
>> > I am running the MSVC++ program from the IDE.
>> >
>> > Could you please suggest me how to make the program(MSVC++) work after
>> > un-commenting the line.
>>
>> See
>> 
>>
>>
>> You're mixing and matching I/O from two different CRTs.  You can't do
>> that.
> 
> Why will it work if I comment these line. I am still mixing the
> cygwin1 and msvcrt.dll?


You're only playing with fire then. ;-)  Seriously, if you know what you're
doing, you can skirt the danger of all kinds of things... if you know what
you're doing.


-- 
Larry Hall  http://www.rfk.com
RFK Partners, Inc.  (508) 893-9779 - RFK Office
216 Dalton Rd.  (508) 893-9889 - FAX
Holliston, MA 01746

_

A: Yes.
> Q: Are you sure?
>> A: Because it reverses the logical flow of conversation.
>>> Q: Why is top posting annoying in email?

--
Unsubscribe info:  http://cygwin.com/ml/#unsubscribe-simple
Problem reports:   http://cygwin.com/problems.html
Documentation: http://cygwin.com/docs.html
FAQ:   http://cygwin.com/faq/

Re: Trademark rights and copyright for "Cygwin" and logo.

[Christopher Faylor]

On Fri, May 11, 2007 at 04:13:12AM +0200, [EMAIL PROTECTED] wrote:
>Nonetheless I continue to be thankful for anybody's input (esp.  if I
>just missed a plainly visible and well known document on that topic
>somewhere).

It is unclear to me what kind of insight you expect to get about matters
like these which would be definitive enough to actually allow you to
distribute a product.

If the 75% of the mailing list thinks you're ok do you think that Red Hat
will just take that as gospel?

If you want to get definitive answers about this, then you should be asking
Red Hat, not this mailing list.

--
Unsubscribe info:  http://cygwin.com/ml/#unsubscribe-simple
Problem reports:   http://cygwin.com/problems.html
Documentation: http://cygwin.com/docs.html
FAQ:   http://cygwin.com/faq/