strange behaviour of xemacs

[Fabrizio Salvatore]

Hi,

I have a Windows XP laptop where I've installed cygwin. I open an xterm 
using rxvt and log on some linux box:


D:\cygwin\bin\rxvt.exe -ls -si -sk -sb -fg black -bg Wheat -fn 7x14 -g 
120x24 -T "linappserv2.pp.rhul.ac.uk" -e ssh -X -Y -l  
linappserv2.pp.rhul.ac.uk


I now run xemacs on the linux box:

xemacs test.txt

At this point I cannot use any of the xemacs commands; for example if I 
do Ctrl spacebar (that should set the starting point for selecting a 
portion of the file) the xemacs window starts going berserk and I cannot 
even close it using Ctrl x Ctrl c. My only chance to do anything is to 
close the window clicking the X at the upper right corner.
The behaviour is the same whatever linux box I log-in to and also if I 
run xemacs opening an xterm on my laptop. Any idea what it might be ?


Cheers,

Fab

--
#
Dr P-Fabrizio Salvatore| Department of Physics
Phone: +44 (0)1784 44 3479 | Royal Holloway College
Fax: +44 (0)1784 47 2794   | University of London
e-mail: [EMAIL PROTECTED] | Egham, Surrey, TW20 0EX
http://www.pp.rhul.ac.uk/~salvator | United Kingdom (UK)
#

--
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: cygipc-1.13-2.tar.bz2 download

[Owen Rees]

--On Monday, March 20, 2006 12:17:31 -0500 Igor Peshansky wrote:


On Mon, 20 Mar 2006, Rajendra S. Gad wrote:

Please inform me where I will have better instruction for installing and
configuring DSpace on this plateform.


I have no idea what DSpace is, but if it does indeed require Cygwin tools
under Linux, then it's really out of this list's domain.  I'd suggest
asking whoever distributes DSpace for help in configuring it.


DSpace  is a digital repository system. The web 
site has links to code (at SourceForge), installation documentation, FAQ, 
mailing list, Wiki etc. and that would be the best place to look for help. 
There is also a quite long list of installations all around the world.


The SourceForge dspace project mentions Cygwin, but only as one possible 
O/S you could use. (Yes, I know Cygwin is not an O/S in its own right, but 
that is the heading where it appears.)


--
Owen Rees
Hewlett Packard Laboratories, Bristol, UK


--
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/

whats wrong with this compile as it dont work on other pcs

[Usman Khan]

whats wrong with this compile as it dont work on other pcs although it
works where i compiled it. i even included the dlls that the exe was
using
http://utaker.googlepages.com/rhapsody-0.28bwin32-cygwin.rar

i have seen other programs like windrop and others using such methods,
why isnt it working here and how to fix it?how to make better
installers that auto include the needed stuff if any

--
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/

[ANNOUNCEMENT] New Release: octave-forge 2006.03.17

[James R. Phillips]

INFO

The Cygwin octave-forge package has been updated to release 2006.03.17-1.

This is the first Cygwin release to incorporate D. Bateman's fixed-point
package, useful for designing fixed-point signal processing applications.

Changes (from release notes):

2006-03-16 ==
* statistics/regress: new function
2006-01-28 ==
* audio/waveread: big-endian fix; 2-channel file fix
* comm/galois: fix rank calculation for galois array
* general/interp2: cubic interpolation
* image/: return MxNx3 matrices of appropriate numeric class
* io/textread: faster
* io/dlmwrite: control precision and line separator; allow append
* miscellaneous/mfilename: compatibility function
* miscellaneous/listen: subnet patterns; loopback support
* NaN/xcovf: missing data xcorr
* octcdf/: support for netcdf files
* optim/wpolyfit: return row vector; allow centering and scaling
* plot/legend: fix error on first call
* plot/print: add PDF option
* signal/residued,residuez: partial fraction expansion of filters
* signal/qp_kaiser: PPN filter bank coefficients
* statistics/histfit: allow negative data
* statistics/mvnpdf,mvnrnd: multivariate normal distribution
* tsa/: improved speed, style and docs
* tsa/amarma: adaptive mean ARMA modeling
* tsa/y2res: add entropy estimation; add equidistant quantisation
* tsa/mvfreqz: add PDCF
* support for 2.9.4
* support for 64-bit indexing
* support for MinGW build

The octave-forge project contains over 500 contributed functions for GNU Octave
which are not in the main distribution. These functions are grouped according
to the following subdirectories: audio, comm, control, general, geometry,
fixed,
ident, image, io, linear-algebra, miscellaneous, optim, path, plot, set,signal,
sparse, specfun, special-matrix, splines, statistics, strings, struct,
symbolic, time. The Cygwin distribution of octave-forge also incorporates
m-files from the octave-ci collection.

octave-forge home page: http://octave.sourceforge.net

octave-ci collection: ftp://ftp.ci.tuwien.ac.at/pub/octave/octave-ci.tar.gz


UPDATE
==

To update your installation, click on the "Install Cygwin now" link on the
http://cygwin.com/ web page.  This downloads setup.exe to your system.  Save it
and run setup, answer the questions and pick up 'octave-forge' from the 'Math'
category. Then click on the 'skip' field next to 'octave-forge' until
'2006.03.17' is displayed.

DOWNLOAD:
=
Note that downloads from sources.redhat.com (aka cygwin.com) aren't allowed due
to bandwidth limitations.  This means that you will need to find a mirror which
has this update, please choose the one nearest to you: 

http://cygwin.com/mirrors.html


QUESTIONS:
==
If you want to make a point or ask a question the Cygwin mailing list
cygwin@cygwin.com is the appropriate place.


CYGWIN-ANNOUNCE UNSUBSCRIBE INFO:
=
To unsubscribe to the cygwin-announce mailing list, look at the
"List-Unsubscribe: " tag in the email header of this message.  Send email to
the address specified there.  It will be in the format:

[EMAIL PROTECTED]

If you need more information on unsubscribing, start reading here:

http://sources.redhat.com/lists.html#unsubscribe-simple

Please read *all* of the information on unsubscribing that is available
starting at this URL.

MAINTAINER
==
James R. Phillips




--
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: 4355 pipe handlers open at once - is this to be 'expect'ed?

[Dave Korn]

On 21 March 2006 12:21, Dave Korn wrote:

> Subject: 4355 pipe handlers open at once - is this to be 'expect'ed?

  D'oh.  "Handles".  Not "handlers".  D'oh.

>   As you know, most of what expect does involves invoking child processes
> and talking to them via their stdio channels.  It strikes me that something
> may be leaking pipe handles to a popen'd child process 

  Which on five minutes closer inspection would turn out to be a newlib issue,
since cygwin relies on newlib's popen.  Double d'oh, sorry 'bout the noise.


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: cygipc-1.13-2.tar.bz2 download

[Rajendra S. Gad]

Sir,

Thanks for your comment. Infact initially I was installing the DSpace(is a
digital repository system, Linux based application,
http://www.dspace.org/). on the windows Xp hence I was playing with
Cygwin, and hence cygipc.

Presently I am trying on the Linux plateform , since many compatibility
issues came on windows plateform for supporting subtools.
Hence UBUNTU.


Jaguarrr and Jabhua( Its a place in India in M.P. where premitives are
staying).

Buy.
Rajendra




> Ugh, top-posting...  Reformatted.
>
> On Mon, 20 Mar 2006, Rajendra S. Gad wrote:
>
>> > On Fri, 17 Mar 2006, Rajendra S. Gad wrote:
>> >
>> >> Sir,
>> >> I am not in position ot download the above file from the portal
>> >>
>> >> http://www.neuro.gatech.edu/users/cwilson/cygutils/cygipc/cygipc-1.13-2.tar.bz2
>> >>
>> >> Please guide me.
>> >
>> > The question is: why do you need it?  This package has been obsolete
>> for some time, and has been superceded with functionality within
>> Cygwin itself (see /usr/share/doc/Cygwin/cygserver.README).
>> >Igor
>>
>> Sir Igor,
>
> Heh, thank you for inadvertently knighting me... :-)
>
>> I am installing the DSpace application on the UBUNTU version 5.10 .
>
> Umm, UBUNTU is Linux.  The "cyg" in "cygipc" stands for "Cygwin".
> cygipc is a Cygwin tool.  Why would you need to use a *Cygwin* tool
> under *Linux*?
>
>> Please inform me where I will have better instruction for installing
>> and configuring DSpace on this plateform.
>
> I have no idea what DSpace is, but if it does indeed require Cygwin
> tools under Linux, then it's really out of this list's domain.  I'd
> suggest asking whoever distributes DSpace for help in configuring it.
>
>> Also let suggest me other linux plateform which will be user friendly
>> for instyallinf DSpace( Specially which will have all the required
>> supporting tools installed.
>
> You might want to try a Linux forum.  Sorry.
>   Igor
> --
>   http://cs.nyu.edu/~pechtcha/
>   |\  _,,,---,,_  [EMAIL PROTECTED] | [EMAIL PROTECTED]
> ZZZzz /,`.-'`'-.  ;-;;,_  Igor Peshansky, Ph.D. (name changed!)
>  |,4-  ) )-,_. ,\ (  `'-' old name: Igor Pechtchanski
> '---''(_/--'  `-'\_) fL   a.k.a JaguaR-R-R-r-r-r-.-.-.  Meow!
>
> "Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends
> compte." "But no -- you are no fool; you call yourself a fool, there's
> proof enough in that!" -- Rostand, "Cyrano de Bergerac"


-- 
Mr. R. S. Gad
Senior Lecture
Electronics Division
Department of Physics
Goa University
Goa India 403 206



--
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/

4355 pipe handlers open at once - is this to be 'expect'ed?

[Dave Korn]

  :)  Sorry for the terrible pun[*], but I just saw this error message scroll
past in the middle of a gcc (simulator-based) testsuite run:

--
doing compile
pid is 2892 -2892
output is
PASS: gcc.c-torture/execute/921117-1.c compilation,  -O2
Simulator: rc 0, ''
   1896 [main] expect 4872! _pinfo::dup_proc_pipe: DuplicateHandle failed, pid
4872, hProcess 0x46FC, Win32 error 5
   2186 [main] expect 4872! _pinfo::dup_proc_pipe: DuplicateHandle failed, pid
4872, hProcess 0x46FC, Win32 error 5
   2323 [main] expect 4872! pinfo::wait: Couldn't duplicate pipe topid
4872(0x46FC), Win32 error 5
PASS: gcc.c-torture/execute/921117-1.c execution,  -O2
Testing gcc.c-torture/execute/921117-1.c,  -O3 -fomit-frame-pointer
--

  I thought the value for hProcess looked a little suspiciously high, but on
digging through it with Process Explorer (which appears to be playing nicely
with cygwin programs these days) discovered that it was a real process handle
value, because expect.exe had over four thousand anonymous pipes open:

--
Process: expect.exe Pid: 2532

TypeNameHandle  Access  Object Address  Share Flags
File\Device\NamedPipe\Win32Pipes.09e4.0685  0x4 0x00120189
0x888DE5D8  ---
File\Device\NamedPipe\Win32Pipes.09e4.071b  0x8 0x00120189
0x89ADE750  ---
File\Device\NamedPipe\Win32Pipes.09e4.0688  0xC 0x00120189
0x86B40118  ---
File\Device\NamedPipe\Win32Pipes.09e4.06db  0x100x00120196
0x877A30F0  ---
File\Device\NamedPipe\Win32Pipes.09e4.06da  0x140x00120189
0x86736AF0  ---
File\Device\NamedPipe\Win32Pipes.09e4.0709  0x180x00120189
0x880ACEF8  ---
   [ many thousands of similar lines snipped! ]
File\Device\NamedPipe\Win32Pipes.09e4.457d  0x4668  0x00120189
0x8875EB68  ---
File\Device\NamedPipe\Win32Pipes.09e4.457e  0x466C  0x00120196
0x869F64C8  ---
File\Device\NamedPipe\Win32Pipes.09e4.4590  0x467C  0x00120196
0x881899B0  ---
File\Device\NamedPipe\Win32Pipes.09e4.45ab  0x4684  0x00120196
0x88D2A2F0  ---
File\Device\NamedPipe\Win32Pipes.09e4.45a2  0x4688  0x00120196
0x8844EA70  ---
File\Device\NamedPipe\Win32Pipes.09e4.45b4  0x469C  0x00120196
0x87F90F90  ---
--

  As you know, most of what expect does involves invoking child processes and
talking to them via their stdio channels.  It strikes me that something may be
leaking pipe handles to a popen'd child process (or similar), but before I
investigate in depth I'd just like to know if it's perhaps an expected
behaviour that these pipe handles should be kept lying around until the expect
process exist.  The child processes have all definitely exited, but is this
what I'd see if the parent expect wasn't fully draining them in some way?


cheers,
  DaveK

[*] - Oh no I'm not!  Not in the slightest, heh!
-- 
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/

1.5.19-4 g++ error: can't find entry point in cygwin1.dll

[Anton Korneev]

Hello,

  Today I try to use last version of cygwin (1.5.19-4) and g++ for
  cygwin (3.4.4-1), and I take a strange error.

  When I compile any program, appears the window with report on error:

  cc1plus.exe - Entry Point not found
  The procedure entry point _impure_ptr could not be located in the
  dynamic link library cygwin1.dll.

  But, with the old versions (1.5.16) I don't have such problem.

-- 
Best regards,
 Anton  mailto:[EMAIL PROTECTED]


--
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: 1.5.19-4 g++ error: can't find entry point in cygwin1.dll

[Eric Blake]

-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

According to Anton Korneev on 3/20/2006 8:08 PM:
>   cc1plus.exe - Entry Point not found
>   The procedure entry point _impure_ptr could not be located in the
>   dynamic link library cygwin1.dll.
> 
>   But, with the old versions (1.5.16) I don't have such problem.

Without the information requested here:

> Problem reports:   http://cygwin.com/problems.html

(including the output of 'cygcheck -svr' as a text attachment), we can
only guess.  But it would appear that you did not fully install the new
cygwin, or that an older cygwin1.dll is floating around somewhere in your
system; and that during the upgrade you picked up the newer g++ that
depends on the newer cygwin.  Be sure you only have one copy of
cygwin1.dll on your machine, and that all cygwin processes are stopped,
before reusing setup.exe to reinstall the cygwin package for the correct
cygwin1.dll.

- --
Life is short - so eat dessert first!

Eric Blake [EMAIL PROTECTED]
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.2.1 (Cygwin)
Comment: Public key at home.comcast.net/~ericblake/eblake.gpg
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFEIAoZ84KuGfSFAYARAu8pAJ9bEpLpsm/WAE2TiHQIaMcbt/ZROQCgut2B
OUvTqkC2y/E05BQGgsjqd/M=
=82BU
-END PGP SIGNATURE-

--
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: help needed for portable exes

[Eric Blake]

-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

According to Usman Khan on 3/20/2006 12:04 PM:
> suppose i compiled a program and it runs successfully (make install done).
> now that i want that the program runs on any other pc, is there a way
> to find all the dlls that are required, some program, or a program
> that makes an installer that installs all the files with the dlls
> required for that program so i can easily isntall on other systems
> that dont have cygwin? or something?

If your program requires cygwin, then it will only run on systems that
have cygwin installed.  If you want to break the dependence on cygwin,
then consider the mingw project (cygwin is even nice enough to provide the
- -mno-cygwin compiler switch for compiling a mingw app, but using that
means your app is no longer a cygwin app so this mailing list is no longer
the place to discuss it).

Having said all that, the cygcheck utility can show you all static
dependencies of an executable.  For example, 'cygcheck /bin/bash' shows
that bash requires cygwin1.dll, cygintl-3.dll, cygiconv-2.dll,
cygreadline6.dll, and cygncurses-8.dll.  But cygcheck is unable to pick up
dynamic dependencies (for example, if your app uses dlsym).

It may be that the easiest way to get your program installed on any PC is
to get your program accepted as part of the cygwin distribution, then tell
users to use cygwin's setup.exe to install your program.

One other thing - if you are planning on distributing your program that
depends on cygwin, remember to obey the GPL: your program must be open
source, with anyone obtaining your program having the ability to obtain
the source at the same time if they desire.

- --
Life is short - so eat dessert first!

Eric Blake [EMAIL PROTECTED]
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.2.1 (Cygwin)
Comment: Public key at home.comcast.net/~ericblake/eblake.gpg
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFEIAWG84KuGfSFAYARAvUJAJ49oAVPGrm700RxNbGLiDjw+OxoyQCgqHMs
7MFChvWPwBkxBfq+mC6Jv08=
=5E8P
-END PGP SIGNATURE-

--
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: newbie:Missing .bashrc and other startups

[Eric Blake]

-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

According to BeemerBiker on 3/20/2006 8:14 AM:
> I installed cygwin but did not select enough tools.  Took about 4 installs 
> before I thought I got all the stuff I needed, x11, etc, etc.  However, there 
> is no .bashrc nor .bash_profile.  cd /home/xxx does nothing except report 
> home 
> does not exist.  echo $HOME does work.  Is there a fix for this?  What did I 
> fail to setup?   I looked at a friends system and his home directory has all 
> the bash stuff that I am missing.  Shall i uninstall and then select 
> everything 
> and re-install?  I googled but didnt see this problem.

Without the information requested here:
> Problem reports:   http://cygwin.com/problems.html

(including cygcheck -svr output as a text attachment), we can only guess.
 However, it seems like your installation failed to run the base-files
postinstall script properly, otherwise starting a login shell would have
run cygwin's /etc/profile that creates a user environment each time a
previously unknown user starts up.  If this is the case, rerunning
setup.exe and selecting base-files for reinstallation may fix your situation.

- --
Life is short - so eat dessert first!

Eric Blake [EMAIL PROTECTED]
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.2.1 (Cygwin)
Comment: Public key at home.comcast.net/~ericblake/eblake.gpg
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFEIAPB84KuGfSFAYARAq43AKDJdY0oXXw0Jqgk3o0/iGX8gNPSCgCgh4hD
Slir3Nj06KbH/JE0slsm368=
=w285
-END PGP SIGNATURE-

--
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/

1.5.19-4: CVS over SSH - trailing dot mkdir problem

[Rix]

Hi all,

When using cvs over ssh I get the following error:

> cannot mkdir /tmp/cvs-serv3716/.
> No such file or directory

CVSROOT= :ext:[EMAIL PROTECTED]:/a/cvs/root
CVS_RSH= ssh

Cygwin version: 1.5.19-4
ssh: 4.3p2-2
cvs: 1.11.21-1


This problem has already been reported on the mailing
list (see
http://www.cygwin.com/ml/cygwin/2006-01/msg01376.html)
but I can find out if it has been fixed or if there is
a workaround.

Thanks for your help.

Cheers,
Eric Boucher 






___ 
Nouveau : téléphonez moins cher avec Yahoo! Messenger ! Découvez les tarifs 
exceptionnels pour appeler la France et l'international.
Téléchargez sur http://fr.messenger.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: cygipc-1.13-2.tar.bz2 download

[Igor Peshansky]

Ugh, top-posting again...  Reformatted again.

On Tue, 21 Mar 2006, Rajendra S. Gad wrote:

> > Ugh, top-posting...  Reformatted.
> >
> > On Mon, 20 Mar 2006, Rajendra S. Gad wrote:
> >
> >> > On Fri, 17 Mar 2006, Rajendra S. Gad wrote:
> >> >
> >> >> Sir,
> >> >> I am not in position ot download the above file from the portal
> >> >>
> >> >> http://www.neuro.gatech.edu/users/cwilson/cygutils/cygipc/cygipc-1.13-2.tar.bz2
> >> >>
> >> >> Please guide me.
> >> >
> >> > The question is: why do you need it?  This package has been obsolete
> >> > for some time, and has been superceded with functionality within
> >> > Cygwin itself (see /usr/share/doc/Cygwin/cygserver.README).
> >> >  Igor
> >>
> >> Sir Igor,
> >
> > Heh, thank you for inadvertently knighting me... :-)
> >
> >> I am installing the DSpace application on the UBUNTU version 5.10 .
> >
> > Umm, UBUNTU is Linux.  The "cyg" in "cygipc" stands for "Cygwin".
> > cygipc is a Cygwin tool.  Why would you need to use a *Cygwin* tool
> > under *Linux*?
> >
> >> Please inform me where I will have better instruction for installing
> >> and configuring DSpace on this plateform.
> >
> > I have no idea what DSpace is, but if it does indeed require Cygwin
> > tools under Linux, then it's really out of this list's domain.  I'd
> > suggest asking whoever distributes DSpace for help in configuring it.
> >
> >> Also let suggest me other linux plateform which will be user friendly
> >> for instyallinf DSpace( Specially which will have all the required
> >> supporting tools installed.
> >
> > You might want to try a Linux forum.  Sorry.
> > Igor
>
> Sir,
>
> Thanks for your comment. Infact initially I was installing the DSpace(is
> a digital repository system, Linux based application,
> http://www.dspace.org/). on the windows Xp hence I was playing with
> Cygwin, and hence cygipc.

If you need IPC functionality in Cygwin, refer to my earlier answer above.
Cygwin now supports this functionality through cygserver, which makes
cygipc obsolete.  Read the README file.

> Presently I am trying on the Linux plateform , since many compatibility
> issues came on windows plateform for supporting subtools. Hence UBUNTU.

If the code uses standard POSIX APIs, it should build and run on Cygwin.
If it doesn't, it's a Cygwin bug, and should be reported.  The
"compatibility issues" you experienced on the Windows platform (using
Cygwin, I assume) will not be fixed or resolved unless they are reported.

> Jaguarrr and Jabhua( Its a place in India in M.P. where premitives
> are staying).

Huh?
Igor
-- 
http://cs.nyu.edu/~pechtcha/
  |\  _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED]
ZZZzz /,`.-'`'-.  ;-;;,_Igor Peshansky, Ph.D. (name changed!)
 |,4-  ) )-,_. ,\ (  `'-'   old name: Igor Pechtchanski
'---''(_/--'  `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-.  Meow!

"Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte."
"But no -- you are no fool; you call yourself a fool, there's proof enough in
that!" -- Rostand, "Cyrano de Bergerac"

--
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: 1.5.19-4: CVS over SSH - trailing dot mkdir problem

[Igor Peshansky]

On Tue, 21 Mar 2006, Rix wrote:

> Hi all,
>
> When using cvs over ssh I get the following error:
>
> > cannot mkdir /tmp/cvs-serv3716/.
> > No such file or directory
>
> CVSROOT= :ext:[EMAIL PROTECTED]:/a/cvs/root
> CVS_RSH= ssh
>
> Cygwin version: 1.5.19-4
> ssh: 4.3p2-2
> cvs: 1.11.21-1
>
> This problem has already been reported on the mailing list (see
> http://www.cygwin.com/ml/cygwin/2006-01/msg01376.html) but I can find
> out if it has been fixed or if there is a workaround.

IIRC, Cygwin was reporting an incorrect errno in this case, so cvs wasn't
ignoring it properly.  This should be fixed in the snapshots.
Igor
-- 
http://cs.nyu.edu/~pechtcha/
  |\  _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED]
ZZZzz /,`.-'`'-.  ;-;;,_Igor Peshansky, Ph.D. (name changed!)
 |,4-  ) )-,_. ,\ (  `'-'   old name: Igor Pechtchanski
'---''(_/--'  `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-.  Meow!

"Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte."
"But no -- you are no fool; you call yourself a fool, there's proof enough in
that!" -- Rostand, "Cyrano de Bergerac"

--
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: help needed for portable exes

[Wes S]

On 21 Mar 2006 at 6:54, Eric Blake wrote:

[snip]

> then consider the mingw project (cygwin is even nice enough to provide
> the
> - -mno-cygwin compiler switch for compiling a mingw app, but using that
> means your app is no longer a cygwin app so this mailing list is no
> longer
> the place to discuss it).
> 
[snip]

What list is appropiate for discussing mingw? 

Sorry for jumping in,

Wes S


--
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: help needed for portable exes

[Chris Sutcliffe]

Hey,

> What list is appropriate for discussing mingw?

Probably the MinGW-users list.  See the MinGW web site for details,
http://www.mingw.org.

Chris

--
Chris Sutcliffe
http://ir0nh34d.blogspot.com
http://emergedesktop.org

--
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: Redirect stdin & stdout to serial port

[D.Pageau]

Yes it works !

Thanks to both of you for your help.

On 3/20/2006 11:48 PM, Lev Bishop wrote:
> I think the solution might be a combination of both of Igor's suggestions
> cmnd <> /dev/ttyS0 >&0
>
> Ie, open the port read/write on stdin and then dup it to stdout.
>
> Let me know if that works.
>
> Lev


--
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: whats wrong with this compile as it dont work on other pcs

[Eric Blake]

-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

According to Usman Khan on 3/21/2006 3:32 AM:
> whats wrong with this compile as it dont work on other pcs although it
> works where i compiled it. i even included the dlls that the exe was
> using
> http://utaker.googlepages.com/rhapsody-0.28bwin32-cygwin.rar

And you expect us to download random packages to point out your bugs?
Give us some more details (like what is the failure message you are
seeing, what exactly are you copying to the other machine, etc).

> 
> i have seen other programs like windrop and others using such methods,
> why isnt it working here and how to fix it?how to make better
> installers that auto include the needed stuff if any

It sounds like you are trying to set yourself up to be a 3PP
(http://cygwin.com/acronyms/#3PP).  Search this list for that acronym to
find better ways of distributing programs that depend on cygwin
(basically, the best way is to make sure cygwin is installed on the target
machine, using cygwin's setup.exe, and then document that your program
depends on cygwin).

- --
Life is short - so eat dessert first!

Eric Blake [EMAIL PROTECTED]
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.2.1 (Cygwin)
Comment: Public key at home.comcast.net/~ericblake/eblake.gpg
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFEIArT84KuGfSFAYARAlTYAJ40l9m5XxCv/dKeks0b1FMiatQC4wCeNxb4
xuyzndCbewepyXtlTuFoJ/0=
=lkqf
-END PGP SIGNATURE-

--
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: whats wrong with this compile as it dont work on other pcs

[Usman Khan]

Thankyou all. I'll look into the matter more deeply.

On 3/21/06, Eric Blake <[EMAIL PROTECTED]> wrote:
> -BEGIN PGP SIGNED MESSAGE-
> Hash: SHA1
>
> According to Usman Khan on 3/21/2006 3:32 AM:
> > whats wrong with this compile as it dont work on other pcs although it
> > works where i compiled it. i even included the dlls that the exe was
> > using
> > http://utaker.googlepages.com/rhapsody-0.28bwin32-cygwin.rar
>
> And you expect us to download random packages to point out your bugs?
> Give us some more details (like what is the failure message you are
> seeing, what exactly are you copying to the other machine, etc).
>
> >
> > i have seen other programs like windrop and others using such methods,
> > why isnt it working here and how to fix it?how to make better
> > installers that auto include the needed stuff if any
>
> It sounds like you are trying to set yourself up to be a 3PP
> (http://cygwin.com/acronyms/#3PP).  Search this list for that acronym to
> find better ways of distributing programs that depend on cygwin
> (basically, the best way is to make sure cygwin is installed on the target
> machine, using cygwin's setup.exe, and then document that your program
> depends on cygwin).
>
> - --
> Life is short - so eat dessert first!
>
> Eric Blake [EMAIL PROTECTED]
> -BEGIN PGP SIGNATURE-
> Version: GnuPG v1.4.2.1 (Cygwin)
> Comment: Public key at home.comcast.net/~ericblake/eblake.gpg
> Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org
>
> iD8DBQFEIArT84KuGfSFAYARAlTYAJ40l9m5XxCv/dKeks0b1FMiatQC4wCeNxb4
> xuyzndCbewepyXtlTuFoJ/0=
> =lkqf
> -END PGP SIGNATURE-
>
> --
> 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/

pthreads don't scale on windows xp, but does scale on linux, cygwin 1.5.19

[jdeifik]

I have a dual xeon 2.4ghz machine with hypertreading enabled.
This gives me 4 logical processors.
The machine dual boots to windows xp sp2, and linux.
I have a highly parallelizable program I wrote, and I tested it 
running 1 to 8 threads,

running with no source changes on windows and linux.

Here is the performance on linux using gcc-3.4.3
threads
1	1436.41user 0.10system 7:16.37elapsed 100%CPU (0avgtext+0avgdata 
0maxresident)k
2	436.00user 0.02system 3:38.15elapsed 199%CPU (0avgtext+0avgdata 
0maxresident)k
3	369.15user 0.05system 2:03.48elapsed 298%CPU (0avgtext+0avgdata 
0maxresident)k
4	359.77user 0.08system 1:42.95elapsed 349%CPU (0avgtext+0avgdata 
0maxresident)k
6	357.83user 0.09system 1:40.94elapsed 354%CPU (0avgtext+0avgdata 
0maxresident)k
8	358.79user 0.06system 1:41.80elapsed 352%CPU (0avgtext+0avgdata 
0maxresident)k


To compute efficiency, take the single thread elapsed time/(# threads 
* threaded elapsed time)


There is virtually perfect scaling. 4 processors scale with an 
efficiency of about 103%.

For 6 and 8 threads, efficiency goes up a small amount.


Here is the performance on windows xp using cygwin pthreads and gcc-3.4.4
1	434.60user 0.20system 7:16.47elapsed 99%CPU (0avgtext+0avgdata 
509696maxresident)k
2	441.78user 0.24system 3:42.06elapsed 199%CPU (0avgtext+0avgdata 
510208maxresident)k
3	579.68user 0.15system 3:14.50elapsed 298%CPU (0avgtext+0avgdata 
511232maxresident)k
4	675.39user 0.15system 2:51.50elapsed 393%CPU (0avgtext+0avgdata 
512000maxresident)k
6	711.70user 0.18system 3:01.20elapsed 392%CPU (0avgtext+0avgdata 
511488maxresident)k
8	683.35user 0.21system 2:56.05elapsed 388%CPU (0avgtext+0avgdata 
512000maxresident)k


Things are fine for 2 threads, scaling with an efficiency of 96%
For 3 threads, scaling efficiency is 73%
For 4 threads, scaling efficiency is 62%
For 6 threads, scaling efficiency is 39%
For 8 threads, scaling efficiency is 30%

So clearly something is broken with pthreads on windows using cygwin.

I can make the source code and test data available, but it is several 
mbytes and involves a

lot of source files to compile.

Thanks,
Jeff Deifik
Cygwin Configuration Diagnostics
Current System Time: Tue Mar 21 06:44:35 2006

Windows XP Professional Ver 5.1 Build 2600 Service Pack 2

Path:   C:\WINDOWS\system32
C:\WINDOWS
C:\WINDOWS\System32\Wbem
l:\cygwin\bin
l:\cygwin\usr\local\bin
u:\j\bin
C:\Program Files\Executive Software\Diskeeper\
C:\Program Files\Common Files\Sonic Shared
C:\Program Files\QuickTime\QTSystem\

Output from l:\cygwin\bin\id.exe (nontsec)
UID: 1003(jdeifik) GID: 513(None)
513(None)  545(Users)

Output from l:\cygwin\bin\id.exe (ntsec)
UID: 1003(jdeifik) GID: 513(None)
513(None)  545(Users)

SysDir: C:\WINDOWS\system32
WinDir: C:\WINDOWS

HOME = 'u:\j'
Path = 
'C:\WINDOWS\system32;C:\WINDOWS;C:\WINDOWS\System32\Wbem;l:\cygwin\bin;l:\cygwin\usr\local\bin;u:\j\bin;C:\Program
 Files\Executive Software\Diskeeper\;C:\Program Files\Common Files\Sonic 
Shared;C:\Program Files\QuickTime\QTSystem\'

ALLUSERSPROFILE = 'C:\Documents and Settings\All Users'
APPDATA = 'C:\Documents and Settings\jdeifik\Application Data'
CLASSPATH = 'C:\Program Files\Java\jre1.5.0_04\lib\ext\QTJava.zip'
CommonProgramFiles = 'C:\Program Files\Common Files'
COMPUTERNAME = 'DUAL_XEON'
ComSpec = 'C:\WINDOWS\system32\cmd.exe'
CVSROOT = '/cygdrive/u/CVSROOT'
FP_NO_HOST_CHECK = 'NO'
HOMEDRIVE = 'C:'
HOMEPATH = '\Documents and Settings\jdeifik'
LOGONSERVER = '\\DUAL_XEON'
NUMBER_OF_PROCESSORS = '4'
OS = 'Windows_NT'
PATHEXT = '.COM;.EXE;.BAT;.CMD;.VBS;.VBE;.JS;.JSE;.WSF;.WSH'
PROCESSOR_ARCHITECTURE = 'x86'
PROCESSOR_IDENTIFIER = 'x86 Family 15 Model 2 Stepping 7, GenuineIntel'
PROCESSOR_LEVEL = '15'
PROCESSOR_REVISION = '0207'
ProgramFiles = 'C:\Program Files'
PROMPT = '$P$G'
QTJAVA = 'C:\Program Files\Java\jre1.5.0_04\lib\ext\QTJava.zip'
SESSIONNAME = 'Console'
SystemDrive = 'C:'
SystemRoot = 'C:\WINDOWS'
TEMP = 'C:\DOCUME~1\jdeifik\LOCALS~1\Temp'
TERM = 'cygwin'
TMP = 'C:\DOCUME~1\jdeifik\LOCALS~1\Temp'
USERDOMAIN = 'DUAL_XEON'
USERNAME = 'jdeifik'
USERPROFILE = 'C:\Documents and Settings\jdeifik'
windir = 'C:\WINDOWS'
POSIXLY_CORRECT = '1'

HKEY_CURRENT_USER\Software\Cygnus Solutions
HKEY_CURRENT_USER\Software\Cygnus Solutions\Cygwin
HKEY_CURRENT_USER\Software\Cygnus Solutions\Cygwin\mounts v2
HKEY_CURRENT_USER\Software\Cygnus Solutions\Cygwin\Program Options
HKEY_LOCAL_MACHINE\SOFTWARE\Cygnus Solutions
HKEY_LOCAL_MACHINE\SOFTWARE\Cygnus Solutions\Cygwin
HKEY_LOCAL_MACHINE\SOFTWARE\Cygnus Solutions\Cygwin\mounts v2
  (default) = '/cygdrive'
  cygdrive flags = 0x0022
HKEY_LOCAL_MACHINE\SOFTWARE\Cygnus Solutions\Cygwin\mounts v2\/
  (default) = 'l:\cygwin'
  flags = 0x000a
HKEY_LOCAL_MACHINE\SOFTWARE\Cygnus Solutions\Cygwin\mounts v2\/usr/bin
  (default) = 'l:\cygwin/bin'
  flags = 0x000a
HKEY_LOCAL_MACHINE\SOFTWARE\Cygnus Solutions\Cygwin\mounts v2\/usr/lib
  (

Testing snapshots

[Angelo Graziosi]

With recent snapshots I note the following.

If I, as user 'Graziosi' with no admin. permission, create a file
'foo.txt' (for example with emacs) I can see:

$ ls -lrt foo.txt
-rw-r--r--  1 Graziosi Users  728 Mar 21 17:30 foo.txt
^^^

If I, succesively, do a login as Administrator, then I see:

$ ls -lrt
-rw-r--r--  1 Graziosi Administrators 728 Mar 21 17:30 foo.txt  
 ^^^

In Windows (W2K SP4) the user Graziosi has been created as 
 
User Name  Group

 Graziosi   Users 

while Administrator has been created as 

   Administrator   Administrators

(this results from Start | Settings | Users and password)


In Cygwin I have

$ cat /etc/group
root:S-1-5-32-544:0:
SYSTEM:S-1-5-18:18:
Nessuno:S-1-5-21-220523388-842925246-854245398-513:513:
Administrators:S-1-5-32-544:544:
Backup Operators:S-1-5-32-551:551:
Guests:S-1-5-32-546:546:
Power Users:S-1-5-32-547:547:
Replicator:S-1-5-32-552:552:
Users:S-1-5-32-545:545:

i.e. the original group file.


$ cat /etc/passwd
SYSTEM:*:18:544:,S-1-5-18::
Administrators:*:544:544:,S-1-5-32-544::
Administrator:unused_by_nt/2000/xp:500:544:U-GRAZIOSI-IRCI53\Administrator,S-1-5
-21-220523388-842925246-854245398-500:/home/Administrator:/bin/bash
ASPNET:unused_by_nt/2000/xp:1009:545:ASP.NET Machine
Account,U-GRAZIOSI-IRCI53\A
SPNET,S-1-5-21-220523388-842925246-854245398-1009:/home/ASPNET:/bin/bash
Graziosi:unused_by_nt/2000/xp:1000:545:Graziosi,U-GRAZIOSI-IRCI53\Graziosi,S-1-5
-21-220523388-842925246-854245398-1000:/home/Graziosi:/bin/bash
Guest:unused_by_nt/2000/xp:501:546:U-GRAZIOSI-IRCI53\Guest,S-1-5-21-220523388-84
2925246-854245398-501:/home/Guest:/bin/bash
VUSR_GRAZIOSI-IRCI53:unused_by_nt/2000/xp:1001:513:VSA Server
Account,U-GRAZIOSI
-IRCI53\VUSR_GRAZIOSI-IRCI53,S-1-5-21-220523388-842925246-854245398-1001:/home/V
USR_GRAZIOSI-IRCI53:/bin/bash


in which I changed only 

   513 --> 544 for Administrator
   513 --> 545 for Graziosi and ASPNET
   513 --> 546 for Guest



So, foo.txt belongs to 'Users' or 'Administrators' ?



This strange behaviour happens only with file created after the
installation of recent snapshots.

A file created previously as Users belongs to Users, indifferently if one
'see' (ls) it as simple user (of group Users) or Administrator (of group
Administrators).



Cheers,

  Angelo.


--
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[2]: 1.5.19-4 g++ error: can't find entry point in cygwin1.dll

[Anton Korneev]

Hello Eric,

>>   cc1plus.exe - Entry Point not found
>>   The procedure entry point _impure_ptr could not be located in the
>>   dynamic link library cygwin1.dll.
>> 
>>   But, with the old versions (1.5.16) I don't have such problem.

> (including the output of 'cygcheck -svr' as a text attachment), we can
> only guess.  But it would appear that you did not fully install the new
> cygwin, or that an older cygwin1.dll is floating around somewhere in your
> system; and that during the upgrade you picked up the newer g++ that
> depends on the newer cygwin.  Be sure you only have one copy of
> cygwin1.dll on your machine, and that all cygwin processes are stopped,
> before reusing setup.exe to reinstall the cygwin package for the correct
> cygwin1.dll.

Thank you very much for quick answer.
I had a problem which you have described in your answer - several versions
of cygwin1.dll on my computer: one in c:\windows\ and the other in 
c:\cygwin\bin\.
Of course I have read about such problem on www.cygwin.com, but I
thought that my problem is not in this, and I did not check it. I mistakenned 
:).
It appears I installed some program which used the library cygwin1.dll
and it was recorded in directory c:\windows\. Of course it has older
version than 1.5.19-4.


-- 
Best regards,
 Antonmailto:[EMAIL PROTECTED]


--
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/

dlopen() bug

[Bernhard Loos]

Hello

Is there any progress in solving the problem mentioned in this thread:
http://cygwin.com/ml/cygwin/2006-03/msg00274.html ?

Because this is also prevents qt-applications from even starting.


 Bernhard
__
Verschicken Sie romantische, coole und witzige Bilder per SMS!
Jetzt bei WEB.DE FreeMail: http://f.web.de/?mc=021193


--
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: dlopen() bug

[Gary Zablackis]

--- Bernhard Loos <[EMAIL PROTECTED]> wrote:

> Hello
> 
> Is there any progress in solving the problem
> mentioned in this thread:
> http://cygwin.com/ml/cygwin/2006-03/msg00274.html ?
> 
> Because this is also prevents qt-applications from
> even starting.
> 
> 
>  Bernhard
> 
Bernhard,

Not yet. I have built debug versions of the
appropriate dlls (cygwin1) and libraries (libstdc++),
but I have not had time to go in depth into the
problem.

Gary

__
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: 4355 pipe handlers open at once - is this to be 'expect'ed?

[Brian Dessent]

Dave Korn wrote:

>   I thought the value for hProcess looked a little suspiciously high, but on
> digging through it with Process Explorer (which appears to be playing nicely
> with cygwin programs these days) discovered that it was a real process handle
> value, because expect.exe had over four thousand anonymous pipes open:

I noticed this a few months ago.  At the end of running the full gcc
testsuite there were over 60,000 handles in the expect.exe process, the
vast majority of which were these anonymous pipes that you show below.

I tried to reproduce it with a looping dummy "hello world" test harness
but couldn't.  I think it's something particular in the (quite
customized) gcc 'expect' harness.

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/

RE: 4355 pipe handlers open at once - is this to be 'expect'ed?

[Dave Korn]

On 21 March 2006 19:15, Brian Dessent wrote:

> Dave Korn wrote:
> 
>>   I thought the value for hProcess looked a little suspiciously high, but
>> on digging through it with Process Explorer (which appears to be playing
>> nicely with cygwin programs these days) discovered that it was a real
>> process handle value, because expect.exe had over four thousand anonymous
>> pipes open: 
> 
> I noticed this a few months ago.  At the end of running the full gcc
> testsuite there were over 60,000 handles in the expect.exe process, the
> vast majority of which were these anonymous pipes that you show below.
> 
> I tried to reproduce it with a looping dummy "hello world" test harness
> but couldn't.  I think it's something particular in the (quite
> customized) gcc 'expect' harness.
> 
> Brian

  Hmm, I reckon the most likely thing is that expect for some reason isn't
actually fully draining the pipe, or maybe just not closing the fd  I'll
probably have to build it from source and take a look.  Thanks for confirming
it's not just something freaky at my 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: dlopen() bug

[skaller]

On Tue, 2006-03-21 at 19:52 +0100, Bernhard Loos wrote:
> Hello
> 
> Is there any progress in solving the problem mentioned in this thread:
> http://cygwin.com/ml/cygwin/2006-03/msg00274.html ?

The example in that page is bugged and there is no
reason at all to expect it to work.

The mainline ct.c is a C program, it's compiled
as a C program, and so cannot interact with C++
libraries.

The mainline of any body of code using C++ MUST be
compiled as C++. Furthermore you must NOT explicitly
link libstdc++, that is the job of the compiler.

The fact that the example works when libstdc++ is linked
with the mainline simply suggests an implementation artefact
that some startup code is stored in the library and the linker
manages to execute it.

There's no assurance this is actually initialising everything
correctly. Some initialisation may need to be done 'on the
bottom of the stack' and/or before other code, and it is too 
late when a dll is called half way through the program.

-- 
John Skaller 
Felix, successor to C++: http://felix.sf.net


--
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: Where are the home sites of Charles Wilson and Michael A. Chase now?

[Peter A. Castro]

On Tue, 21 Mar 2006, Alexey Lyubimov wrote:


Could anybody tell where are the home sites of
  "cygutils" - Charles Wilson (was 
http://www.neuro.gatech.edu/users/cwilson/cygutils/)


http://cygutils.fruitbat.org

  and "clean_setup.pl" - Michael A. Chase (was 
http://home.ix.netcom.com/~mchase/)

now?
Thank you.

---
Alexey Lyubimov


--
Peter A. Castro <[EMAIL PROTECTED]> or <[EMAIL PROTECTED]>
"Cats are just autistic Dogs" -- Dr. Tony Attwood

--
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/

KillAll SpamAssassin Processes

[Herb Martin]

This should be simple but I haven't found the
magic encantation so far:

I wish to kill SpamAssassin and all child processes
it has started.

[My experience is that trying to signal -s HUP does
not work with SA (on CygWin) but puts SA into some
sort of unresponsive state.]

I have other (than SA) Perl processes running so 
using "killall" with just the process NAME is not
a good choice.

So far no combinations of "killall" switches,
e.g., using Group ID etc, have worked.

My only useful method (other than hitting each process
individually) has been to use -i (interactive) and then
respond to each confirmation.

Thanks for straightening me out -- I am sure this is
very easy but haven't gotten it right yet.

--
Herb Martin


--
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/

cygrunsrv stderr and stdout only to a syslog-ng log?

[Bryan D. Thomas]

I've found in cygrunsrv.README that the -1 and -2
arguments may be used to specify that output of stdout
and stderr go to a log file instead of to the Windows
Event Log.

I'm using cygrunsrv to run two services: sshd and
syslog-ng.  I'm using syslog-ng to collect timestamped
log entries from sshd and do postprocessing with
filters.  What I want to do is stop logging sshd
events also to the Windows Event Log.

Is it possible to specify in the cygrunsrv arguments
to install the sshd service (I suppose by editing the
ssh-host-config script) that the log events will go to
syslog-ng, but will not go to the Windows Event Log?

For example, if I used -1 /var/log/sshd.log and -2
/var/log/sshd.log, would that work nicely with the
syslog-ng which is using those same files as a target,
or would I get contention, or duplicate lines?

Or, could I use -1 /dev/log -2 /dev/log and syslog-ng
would pick up all the output to the unix-dgram but
only enter it once in the log?

Or, could I use -1 /dev/null -2 /dev/null, but
syslog-ng would keep working as it is today, and only
the Windows Event Log would no longer get sshd events?

I also cannot tell from the documentation I've been
able to find whether it is possible to change the -1
and -2 options for an installed service.  Can I simply
stop the service, run cygrunsrv or edit the registry,
then start the service?

Thanks in advance.

__
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/

updatedb broken

[Steve Kelem]

I'm having trouble running updatedb.  I've tried findutils-4.2.27-1 and 
findutils-4.3.0-1.

When I run updatedb manually, I get the message:
updatedb needs to be able to execute /usr/lib/lapack/find, but cannot.
The directory /usr/lib/lapack contains
cygblas.dll   cyglapack.dll

Thanks for your help,
Steve

--
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: KillAll SpamAssassin Processes

[Igor Peshansky]

On Tue, 21 Mar 2006, Herb Martin wrote:

> This should be simple but I haven't found the
> magic encantation so far:
>
> I wish to kill SpamAssassin and all child processes
> it has started.
>
> [My experience is that trying to signal -s HUP does
> not work with SA (on CygWin) but puts SA into some
> sort of unresponsive state.]
>
> I have other (than SA) Perl processes running so
> using "killall" with just the process NAME is not
> a good choice.
>
> So far no combinations of "killall" switches,
> e.g., using Group ID etc, have worked.
>
> My only useful method (other than hitting each process
> individually) has been to use -i (interactive) and then
> respond to each confirmation.
>
> Thanks for straightening me out -- I am sure this is
> very easy but haven't gotten it right yet.

You could put all the processes into their own process group (using
newpgrp() in perl, or bash's job control), and then doing a "kill -SIG
-PGID" to kill all processes in the group PGID.
FWIW, this is not Cygwin-specific.
HTH,
Igor
-- 
http://cs.nyu.edu/~pechtcha/
  |\  _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED]
ZZZzz /,`.-'`'-.  ;-;;,_Igor Peshansky, Ph.D. (name changed!)
 |,4-  ) )-,_. ,\ (  `'-'   old name: Igor Pechtchanski
'---''(_/--'  `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-.  Meow!

"Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte."
"But no -- you are no fool; you call yourself a fool, there's proof enough in
that!" -- Rostand, "Cyrano de Bergerac"

--
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: updatedb broken [attn: lapack maintainer]

[Eric Blake]

Steve Kelem  elementcxi.com> writes:

> 
> I'm having trouble running updatedb.  I've tried findutils-4.2.27-1 and 
> findutils-4.3.0-1.
> When I run updatedb manually, I get the message:
> updatedb needs to be able to execute /usr/lib/lapack/find, but cannot.
> The directory /usr/lib/lapack contains
> cygblas.dll   cyglapack.dll

That's because /etc/profile.d/lapack.sh is polluting the namespace with
BINDIR=/usr/lib/lapack, and then updatedb fails to use its default of
BINDIR=/usr/bin because BINDIR is already set.

I don't know if findutils or lapack is at fault for this conflicting
definition of what $BINDIR represents.  I can probably work around it
in the next findutils release, but wonder if the lapack maintainer should
instead be more careful to not pollute the environment during its profile.d
script.

In the meantime, as a workaround, you can do 'unset $BINDIR' prior
to running updatedb.

-- 
Eric Blake
volunteer cygwin findutils maintainer



--
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: updatedb broken [attn: lapack maintainer]

[Eric Blake]

Eric Blake  byu.net> writes:

> In the meantime, as a workaround, you can do 'unset $BINDIR' prior
> to running updatedb.

Make that 'unset BINDIR'.  My proofreading isn't very good today :)

-- 
Eric Blake





--
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: updatedb broken [attn: lapack maintainer]

[James R. Phillips]

The lapack maintainer would be me.

>That's because /etc/profile.d/lapack.sh is polluting the namespace with
>BINDIR=/usr/lib/lapack

Hm, you are correct.  This is not purposeful, but inadvertent.  I did not note
carefully that the script is being sourced, so that modifications to commonly
used environmental variables like BINDIR are retained when the script
completes.

>I don't know if findutils or lapack is at fault for this conflicting
>definition of what $BINDIR represents.  I can probably work around it
>in the next findutils release, but wonder if the lapack maintainer should
>instead be more careful to not pollute the environment during its profile.d
>script.

lapack.sh is at fault here.  It should use variable names less likely to
produce collisions, perhaps prefixed with "LA_", or some similar device.  I
will fix this in the near future.  Thanks for pointing out the issue.

Jim Phillips




--
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: cygrunsrv stderr and stdout only to a syslog-ng log?

[René Berber]

Bryan D. Thomas wrote:

> I've found in cygrunsrv.README that the -1 and -2
> arguments may be used to specify that output of stdout
> and stderr go to a log file instead of to the Windows
> Event Log.

You are probably misunderstanding, those options change the default which is to
send any cygrunsrv output to /var/log/ which usually is empty
since each service does its own logging.

> I'm using cygrunsrv to run two services: sshd and
> syslog-ng.  I'm using syslog-ng to collect timestamped
> log entries from sshd and do postprocessing with
> filters.  What I want to do is stop logging sshd
> events also to the Windows Event Log.

What you describe is already being done: if sshd, for instance, finds that
syslogd is running when it starts it uses it, if not it uses the Windows event 
log.

So all you need is to start syslogd, it works fine if all services are started
at the same time (automatic in services starup type).

> Is it possible to specify in the cygrunsrv arguments
> to install the sshd service (I suppose by editing the
> ssh-host-config script) that the log events will go to
> syslog-ng, but will not go to the Windows Event Log?

Only one log.  And don't change the script, if you need to change anything, and
you don't, change the sshd configuration.

> For example, if I used -1 /var/log/sshd.log and -2
> /var/log/sshd.log, would that work nicely with the
> syslog-ng which is using those same files as a target,
> or would I get contention, or duplicate lines?

It doesn't work as you describe at all.

> Or, could I use -1 /dev/log -2 /dev/log and syslog-ng
> would pick up all the output to the unix-dgram but
> only enter it once in the log?
> 
> Or, could I use -1 /dev/null -2 /dev/null, but
> syslog-ng would keep working as it is today, and only
> the Windows Event Log would no longer get sshd events?
> 
> I also cannot tell from the documentation I've been
> able to find whether it is possible to change the -1
> and -2 options for an installed service.  Can I simply
> stop the service, run cygrunsrv or edit the registry,
> then start the service?

Don't edit the registry (you are dangerous!), don't run cygrunsrv... just
install the service and start it (which in the case of sshd is done by the
installation script).
-- 
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: Updated: findutils-4.2.27-1

[Steve Kelem]

I'm having trouble running updatedb.  I've tried findutils-4.2.27-1 and 
findutils-4.3.0-1.

When I run updatedb manually, I get the message:
updatedb needs to be able to execute /usr/lib/lapack/find, but cannot.
The directory /usr/lib/lapack contains
cygblas.dll   cyglapack.dll

Thanks for your help,
Steve


--
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: cygrunsrv stderr and stdout only to a syslog-ng log?

[Bryan D. Thomas]

Thank you for your response, René.

"René Berber" <[EMAIL PROTECTED]> wrote:

> if sshd, for instance, finds that syslogd is running when it starts it 
> uses it,
> if not it uses the Windows event log.

> So all you need is to start syslogd, it works fine if all services are 
> started
> at the same time (automatic in services starup type).

What you say is correct. I am getting log lines in my sshd.log with
timestamps, and my filters in syslog-ng.conf are working.  However, I am
also still getting lines from sshd in my Application Event Log.

My goal is to stop the messages going to the Application Event Log.

Perhaps I'm off the track suggesting to change the options sent to 
cygrunsrv.
I only started thinking about that because the ssh-host-config script does 
use cygrunsrv. So, instead of suggesting options, I'll just ask: is there a 
way to stop sshd logging to the Application Event Log? If yes, how?

Thank you.





--
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: cygrunsrv stderr and stdout only to a syslog-ng log?

[René Berber]

Bryan D. Thomas wrote:
[snip]
> I only started thinking about that because the ssh-host-config script does 
> use cygrunsrv. So, instead of suggesting options, I'll just ask: is there a 
> way to stop sshd logging to the Application Event Log? If yes, how?

It shouldn't be doing that.

The problem I guess is that you have more than one sshd process running, are the
syslog messages the same as the ones in the event log?

If you don't have several sshd processes then you did something while installing
the service that caused the problem.  Stop and remove the service (only sshd)
and install it again using the Cygwin provided script.
-- 
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: dlopen() bug

[Bernhard Loos]

> -Ursprüngliche Nachricht-
> Von: skaller  users  sourceforge  net>
> Gesendet: 21.03.06 20:31:08

> On Tue, 2006-03-21 at 19:52 +0100, Bernhard Loos wrote:
> > Hello
> > 
> > Is there any progress in solving the problem mentioned in this thread:
> > http://cygwin.com/ml/cygwin/2006-03/msg00274.html ?
> 
> The example in that page is bugged and there is no
> reason at all to expect it to work.
> 
> The mainline ct.c is a C program, it's compiled
> as a C program, and so cannot interact with C++
> libraries.
 
If the C++library declares its calls as 'extern "C"' (as it is done in the 
example),
AFAIK there shouldn't be any problem.
 
> The mainline of any body of code using C++ MUST be
> compiled as C++. Furthermore you must NOT explicitly
> link libstdc++, that is the job of the compiler.

Yes, but removing the -lstdc++ doesn't make any difference in this example.

> The fact that the example works when libstdc++ is linked
> with the mainline simply suggests an implementation artefact
> that some startup code is stored in the library and the linker
> manages to execute it.
 
It isn't linked against libstdc++. (At least, I don't see, why ist should be.)

> There's no assurance this is actually initialising everything
> correctly. Some initialisation may need to be done 'on the
> bottom of the stack' and/or before other code, and it is too 
> late when a dll is called half way through the program.
> 
> -- 
> John Skaller 
> Felix, successor to C++: http://felix.sf.net


Anyway all this things aren't really related to the crash.

I removed as much things as possible from the testcase and corrected
the things mentioned above. But I wasn't able to find a solution or even an 
explaination.

This is a part of the log from depency walker:
00:00:00.438: LoadLibraryA("H:\cygwin\test\CrashTest\CrashTest.dll") called 
from "CYGWIN1.DLL" at address 0x6100FCB2 by thread 1.
00:00:00.500: Loaded "CRASHTEST.DLL" at address 0x00B4 by thread 1.  
Successfully hooked module.
00:00:00.500: DllMain(0x00B4, DLL_PROCESS_ATTACH, 0x) in 
"CRASHTEST.DLL" called by thread 1.
00:00:00.500: First chance exception 0xC005 (Access Violation) occurred in 
"CYGWIN1.DLL" at address 0x610AE938 by thread 1.
00:00:00.500: Unloaded "CRASHTEST.DLL" at address 0x00B4 by thread 1.
00:00:00.516: LoadLibraryA("H:\cygwin\test\CrashTest\CrashTest.dll") returned 
NULL by thread 1. Error: Unzulssiger Zugriff auf einen Speicherbereich (998).
00:00:00.578: First chance exception 0xC005 (Access Violation) occurred at 
address 0x00B4F37E by thread 1.
00:00:00.578: First chance exception 0xC005 (Access Violation) occurred in 
"CYGWIN1.DLL" at address 0x6108DFFC by thread 1.
00:00:00.578: First chance exception 0xC005 (Access Violation) occurred in 
"CYGWIN1.DLL" at address 0x6108DFFC by thread 1.
00:00:00.578: First chance exception 0xC005 (Access Violation) occurred in 
"CYGWIN1.DLL" at address 0x6108DFFC by thread 1.

The DllMain fails for some reason.
The program crashs with a stack-overflow after a few hundred lines of like the 
last three.

With 1.5.18, everything works as aspected.

file ct.cc:
---
#include 
#include 


extern "C"
void test();

int main(int argc, char** argv)
{
char* pszdll="./CrashTest.dll";

printf("\n\ndlopening %s\n", pszdll);

void* hDLL = dlopen(pszdll, 0);

if(hDLL)
{
printf("OK\n");
}
else
printf("Error dlopening %s\n", pszdll);

return 0;
}

#ifdef HARDLINKTEST
void TestLinked(char* pszdll)
{
test();
}
#endif
--

crashtest.cc:
--
#include 

extern "C" __declspec(dllexport)
void test()
{
return;
}
-

compile with:
g++ -W -Wall -g -DMAKEDLL -DDEBUG CrashTest.cc -shared -o CrashTest.dll

g++ -g -W -Wall -DUSEDLL -DDEBUG ct.cc -o ctCRASH.exe
g++ -W -Wall -DHARDLINKTEST -DUSEDLL -DDEBUG ct.cc -o ctOK.exe -L./  -lCrashTest


  Bernhard Loos
__
XXL-Speicher, PC-Virenschutz, Spartarife & mehr: Nur im WEB.DE Club!
Jetzt gratis testen! http://freemail.web.de/home/landingpad/?mc=021130


--
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: pthreads don't scale on windows xp, but does scale on linux, cygwin 1.5.19

[Tim Prince]

jdeifik wrote:

I have a dual xeon 2.4ghz machine with hypertreading enabled.
This gives me 4 logical processors.
The machine dual boots to windows xp sp2, and linux.
I have a highly parallelizable program I wrote, and I tested it running 
1 to 8 threads,

running with no source changes on windows and linux.

Here is the performance on linux using gcc-3.4.3
threads
11436.41user 0.10system 7:16.37elapsed 100%CPU (0avgtext+0avgdata 
0maxresident)k
2436.00user 0.02system 3:38.15elapsed 199%CPU (0avgtext+0avgdata 
0maxresident)k
3369.15user 0.05system 2:03.48elapsed 298%CPU (0avgtext+0avgdata 
0maxresident)k
4359.77user 0.08system 1:42.95elapsed 349%CPU (0avgtext+0avgdata 
0maxresident)k
6357.83user 0.09system 1:40.94elapsed 354%CPU (0avgtext+0avgdata 
0maxresident)k
8358.79user 0.06system 1:41.80elapsed 352%CPU (0avgtext+0avgdata 
0maxresident)k


To compute efficiency, take the single thread elapsed time/(# threads * 
threaded elapsed time)


There is virtually perfect scaling. 4 processors scale with an 
efficiency of about 103%.

For 6 and 8 threads, efficiency goes up a small amount.


Here is the performance on windows xp using cygwin pthreads and gcc-3.4.4
1434.60user 0.20system 7:16.47elapsed 99%CPU (0avgtext+0avgdata 
509696maxresident)k
2441.78user 0.24system 3:42.06elapsed 199%CPU (0avgtext+0avgdata 
510208maxresident)k
3579.68user 0.15system 3:14.50elapsed 298%CPU (0avgtext+0avgdata 
511232maxresident)k
4675.39user 0.15system 2:51.50elapsed 393%CPU (0avgtext+0avgdata 
512000maxresident)k
6711.70user 0.18system 3:01.20elapsed 392%CPU (0avgtext+0avgdata 
511488maxresident)k
8683.35user 0.21system 2:56.05elapsed 388%CPU (0avgtext+0avgdata 
512000maxresident)k


Things are fine for 2 threads, scaling with an efficiency of 96%
For 3 threads, scaling efficiency is 73%
For 4 threads, scaling efficiency is 62%
For 6 threads, scaling efficiency is 39%
For 8 threads, scaling efficiency is 30%



Windows doesn't have HT aware scheduling, such as recent linux 
schedulers incorporate.  Cygwin doesn't attempt to improve on the 
Windows scheduler.  I won't ask for relevant details about your linux, 
or how you managed to write a program which doesn't deliver close to 
full performance at 2 threads, as that would take this even further Off 
Topic. However, if you are getting good scaling to 2 threads, that 
should enable you to get all the dual processor performance you can 
expect in Windows for practical purposes.  You might try repeating your 
tests with HT disabled in BIOS.


--
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: signal is not delivered to the handler

[D.Pageau]

> If you have a simple test case which illustrates the problem, we'll
> certainly investigate.  Otherwise, I doubt that anyone is going to debug
> lzrz for you.

I was wrong, signal is delivred to handler but read() does not exit with
EINTR on SIGALRM like Linux and FreeBSD does.  Did I miss something ?

#include 
#include 

#define BUFZ5

void alarm_handler(int signo)
{
fprintf(stderr, "SIGNAL: %d\n", signo);
}

int main()
{
charbuff[BUFSIZ];
FILE*logfp = NULL;

siginterrupt(SIGALRM,1);
signal(SIGALRM, alarm_handler);
alarm(5);
fprintf (stderr, "Should timeout by itself in 5 secs\n");
read(0,buff,BUFSIZ);
fprintf (stderr, "Timeout\n");
alarm(0);
signal(SIGALRM,SIG_DFL);
}


--
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: pthreads don't scale on windows xp, but does scale on linux, cygwin 1.5.19

[jdeifik]

At 07:00 PM 3/21/2006, you wrote:


jdeifik wrote:

I have a dual xeon 2.4ghz machine with hypertreading enabled.
This gives me 4 logical processors.
The machine dual boots to windows xp sp2, and linux.
I have a highly parallelizable program I wrote, and I tested it 
running 1 to 8 threads,

running with no source changes on windows and linux.
Here is the performance on linux using gcc-3.4.3
threads
11436.41user 0.10system 7:16.37elapsed 100%CPU 
(0avgtext+0avgdata 0maxresident)k
2436.00user 0.02system 3:38.15elapsed 199%CPU 
(0avgtext+0avgdata 0maxresident)k
3369.15user 0.05system 2:03.48elapsed 298%CPU 
(0avgtext+0avgdata 0maxresident)k
4359.77user 0.08system 1:42.95elapsed 349%CPU 
(0avgtext+0avgdata 0maxresident)k
6357.83user 0.09system 1:40.94elapsed 354%CPU 
(0avgtext+0avgdata 0maxresident)k
8358.79user 0.06system 1:41.80elapsed 352%CPU 
(0avgtext+0avgdata 0maxresident)k
To compute efficiency, take the single thread elapsed time/(# 
threads * threaded elapsed time)
There is virtually perfect scaling. 4 processors scale with an 
efficiency of about 103%.

For 6 and 8 threads, efficiency goes up a small amount.

Here is the performance on windows xp using cygwin pthreads and gcc-3.4.4
1434.60user 0.20system 7:16.47elapsed 99%CPU (0avgtext+0avgdata 
509696maxresident)k
2441.78user 0.24system 3:42.06elapsed 199%CPU 
(0avgtext+0avgdata 510208maxresident)k
3579.68user 0.15system 3:14.50elapsed 298%CPU 
(0avgtext+0avgdata 511232maxresident)k
4675.39user 0.15system 2:51.50elapsed 393%CPU 
(0avgtext+0avgdata 512000maxresident)k
6711.70user 0.18system 3:01.20elapsed 392%CPU 
(0avgtext+0avgdata 511488maxresident)k
8683.35user 0.21system 2:56.05elapsed 388%CPU 
(0avgtext+0avgdata 512000maxresident)k

Things are fine for 2 threads, scaling with an efficiency of 96%
For 3 threads, scaling efficiency is 73%
For 4 threads, scaling efficiency is 62%
For 6 threads, scaling efficiency is 39%
For 8 threads, scaling efficiency is 30%


Windows doesn't have HT aware scheduling, such as recent linux 
schedulers incorporate.  Cygwin doesn't attempt to improve on the 
Windows scheduler.  I won't ask for relevant details about your 
linux, or how you managed to write a program which doesn't deliver 
close to full performance at 2 threads, as that would take this even 
further Off Topic. However, if you are getting good scaling to 2 
threads, that should enable you to get all the dual processor 
performance you can expect in Windows for practical purposes.  You 
might try repeating your tests with HT disabled in BIOS.


My linux is mandrake 10.2, I suspect running kernel 2.6.11-13smp.
My program scales perfectly at 2 threads on linux. It also scales 
perfectly at 4 threads on linux.

The problem isn't with my program.

I am not sure why it is important to have a HT aware scheduler for 
Windows, when there are 4 or more
threads. I can see with 2 threads you would like to have one per 
physical processor.
With 4 or more threads, cygwin phtreads really sucks, 4->62%, 6->39%, 
8->30% efficiency.


It seems to me that more and more apps are turning to threading for 
performance,
and more and more hardware is available with multi-processor, 
multi-core, and multi-threading.


Jeff Deifik 



--
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: pthreads don't scale on windows xp, but does scale on linux, cygwin 1.5.19

[Erich Dollansky]

Hi,

jdeifik wrote:

At 07:00 PM 3/21/2006, you wrote:


jdeifik wrote:

I have a dual xeon 2.4ghz machine with hypertreading enabled.
This gives me 4 logical processors.


I am not sure why it is important to have a HT aware scheduler for 
Windows, when there are 4 or more
threads. I can see with 2 threads you would like to have one per 
physical processor.
With 4 or more threads, cygwin phtreads really sucks, 4->62%, 6->39%, 
8->30% efficiency.


I do not wonder. HT does not give a full CPU to the other thread. If the 
scheduler is not aware of the problem, it actually slows down the 
performance of the real CPU.


It seems to me that more and more apps are turning to threading for 
performance,
and more and more hardware is available with multi-processor, 
multi-core, and multi-threading.


Yes, this is why programs must be written to run with real CPUs and not 
something Intel calls HyperThreading.


Erich

--
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: signal is not delivered to the handler

[Christopher Faylor]

On Tue, Mar 21, 2006 at 10:06:12PM -0500, D.Pageau wrote:
>> If you have a simple test case which illustrates the problem, we'll
>> certainly investigate.  Otherwise, I doubt that anyone is going to debug
>> lzrz for you.
>
>I was wrong, signal is delivred to handler but read() does not exit with
>EINTR on SIGALRM like Linux and FreeBSD does.  Did I miss something ?

Thanks for the test case.

It looks like cygwin's implementation of siginterrupt doesn't match
linux's.  If you switch the order of calls so that siginterrupt comes
second, it will probably work.  You can also use sigaction to set up the
handler, clearing the SA_RESETHAND flag:

  struct sigaction sa;
  sigaction (SIGALRM, NULL, &sa);
  sa.sa_flags &= ~SA_RESTART;
  sa.sa_handler = alarm_handler;
  sigaction (SIGALRM, &sa, NULL);

That also works.

I'll look into fixing this in the next release of cygwin.

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: help needed for portable exes

[Eric Blake]

-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

Please send messages to the list, and not to me - the collective knowledge
of the list is much better (and often faster) than what you will get out
of a single person: http://cygwin.com/acronyms/#PPIOSPE

Please follow list netiquette, and put replies after context, instead of
generating http://cygwin.com/acronyms/#TOFU; reformatted accordingly.

-  Original Message 

>
> On 3/21/06, Eric Blake  wrote:
     ^
Please respect people's right to privacy, by masking raw email addresses:
http://cygwin.com/acronyms/#PCYMTNQREAIYR

> According to Usman Khan on 3/20/2006 12:04 PM:
>> suppose i compiled a program and it runs successfully (make install done).
>> now that i want that the program runs on any other pc, is there a way
>> to find all the dlls that are required, some program, or a program
>> that makes an installer that installs all the files with the dlls
>> required for that program so i can easily isntall on other systems
>> that dont have cygwin? or something?
> 
> If your program requires cygwin, then it will only run on systems that
> have cygwin installed.
> One other thing - if you are planning on distributing your program that
> depends on cygwin, remember to obey the GPL: your program must be open
> source, with anyone obtaining your program having the ability to obtain
> the source at the same time if they desire.

forgot to attach the screenshot, here it is and Thankyou for the
guidance. ---
You are right. I hope I am not bothering you by emailing here to your
email. I compiled a few other proggies and they ran fine with if i
included the dlls with the exes but this one didn't. A screenshot is
included of the same proggie giving an error when ran on another
system if distributed with the dll. If you have any suggestion then
let me know.
Regards

On 3/21/06, Usman Khan  wrote:
> You are right. I hope I am not bothering you by emailing here to your
> email. I compiled a few other proggies and they ran fine with if i
> included the dlls with the exes but this one didn't. A screenshot is
> included of the same proggie giving an error when ran on another
> system if distributed with the dll. If you have any suggestion then
> let me know.
> Regards
> Usman

I suggest that you quit putting copies of cygwin1.dll in non-typical
directories - you are only asking for problems if the version in
c:\rhapsody does not match the (previously or future) installation using
setup.exe in the more typical c:\cygwin.

- --
Life is short - so eat dessert first!

Eric Blake [EMAIL PROTECTED]
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.2.1 (Cygwin)
Comment: Public key at home.comcast.net/~ericblake/eblake.gpg
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFEIMo284KuGfSFAYARApE4AJ9pykpd4laL+XE4zjRMbVSPwnBGvACeKDo0
JEIbz/J2fKtSVNCXK4LxcAc=
=E/C9
-END PGP SIGNATURE-

--
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/

Reading too few bytes

[Paul J. Lucas]

Here's a small test program:

#include 
#include 
#include 

using namespace std;

int main() {
int fd = ::open( "ESLF", O_RDONLY );
if ( fd == -1 ) {
cerr << "could not open ESLF file" << endl;
::exit( 1 );
}
char buf[ 64 ];
ssize_t bytesRead = ::read( fd, buf, sizeof( buf ) );
cout << "read " << bytesRead << " bytes" << endl;
::close( fd );
}

Here's the contents of a small biinary file "ESLF" (printed in hex bytes):

54 39 CA 1A 44

When I compile the program as:

g++ -o test test.cpp

and run it, it prints "read 5 bytes" as one would expect.  When I compile
the program as:

g++ -mno-cygwin -o test test.cpp

and run it, it prints "read 3 bytes".  Why?  How do I teg it to read all
5 bytes?

- Paul


--
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: Creating a custom Cygwin package server

[Joshua Daniel Franklin]

On 3/19/06, Norton Allen wrote:
> The documentation here:
>
> http://sourceware.org/cygwin-apps/package-server.html
>
> suggests it might be possible to set up a custom Cygwin
> package server to install a custom app, but it also says
> the necessary tools are not available.

Just be aware that you are entering unsupported territory,
and get upset from:

http://xarch.tu-graz.ac.at/publ/cygwin/upset

"Unsupported" means don't ask this mailing list if you have
problems. A good alternative might be:

http://sourceforge.net/projects/cygwin-ports/

--
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/

OpenSSH/SSHD Problem - seteuid 1005: Permission denied (WinXP 64-bit)

[Glenn]

I have a problem with "sshd" where it;
  (1) won't allow users to login, users recieve the message "Read from 
socket failed: Connection reset by peer"

  (2) "sshd" terminates prematurely.

The problem I believe (though I'm not really sure) relates to the second 
final line of output from the sshd server (see below)

  "seteuid 1005: Permission denied"
I'm running Windows XP 64-bit Edition -- it's my belief that seteuid cannot 
be called from a 32-bit context?? I could be wrong?? Though 
http://cygwin.com/faq/ seems to suggest that "Cygwin can be expected to run 
on all modern 32 bit versions [including] WOW64 32 bit environment on 
released 64 bit versions of Windows". In my config I've turned off privilege 
seperate ("UsePrivilegeSeparation no") -- however this didn't make any 
difference.


Does anyone know how I can fix this problem??! (see below for output from 
"/usr/sbin/sshd -D -d -e")


Regards,
  Glenn


** Start sshd in debug mode to stdout:

[EMAIL PROTECTED] var [1:221] /usr/sbin/sshd -D -d -e
debug1: sshd version OpenSSH_4.3p2
debug1: private host key: #0 type 0 RSA1
debug1: read PEM private key done: type RSA
debug1: private host key: #1 type 1 RSA
debug1: read PEM private key done: type DSA
debug1: private host key: #2 type 2 DSA
debug1: rexec_argv[0]='/usr/sbin/sshd'
debug1: rexec_argv[1]='-D'
debug1: rexec_argv[2]='-d'
debug1: rexec_argv[3]='-e'
debug1: Bind to port 2 on 0.0.0.0.
Server listening on 0.0.0.0 port 22.
Generating 768 bit RSA key.
RSA key generation complete.

** Execute the command "ssh [EMAIL PROTECTED]" from another machine:

debug1: fd 4 clearing O_NONBLOCK
debug1: Server will not fork when running in debugging mode.
debug1: rexec start in 4 out 4 newsock 4 pipe -1 sock 7
debug1: sshd version OpenSSH_4.3p2
debug1: private host key: #0 type 0 RSA1
debug1: read PEM private key done: type RSA
debug1: private host key: #1 type 1 RSA
debug1: read PEM private key done: type DSA
debug1: private host key: #2 type 2 DSA
debug1: inetd sockets after dupping: 3, 3
Connection from 192.168.0.40 port 1363
debug1: Client protocol version 2.0; client software version OpenSSH_4.3
debug1: match: OpenSSH_4.3 pat OpenSSH*
debug1: Enabling compatibility mode for protocol 2.0
debug1: Local version string SSH-1.99-OpenSSH_4.3
debug1: list_hostkey_types: ssh-rsa,ssh-dss
debug1: SSH2_MSG_KEXINIT sent
debug1: SSH2_MSG_KEXINIT received
debug1: kex: client->server aes128-cbc hmac-md5 none
debug1: kex: server->client aes128-cbc hmac-md5 none
debug1: SSH2_MSG_KEX_DH_GEX_REQUEST received
debug1: SSH2_MSG_KEX_DH_GEX_GROUP sent
debug1: expecting SSH2_MSG_KEX_DH_GEX_INIT
debug1: SSH2_MSG_KEX_DH_GEX_REPLY sent
debug1: SSH2_MSG_NEWKEYS sent
debug1: expecting SSH2_MSG_NEWKEYS
debug1: SSH2_MSG_NEWKEYS received
debug1: KEX done
debug1: userauth-request for user glenn service ssh-connection method none
debug1: attempt 0 failures 0
Failed none for glenn from 192.168.0.40 port 1363 ssh2
debug1: userauth-request for user glenn service ssh-connection method 
publickey

debug1: attempt 1 failures 1
debug1: test whether pkalg/pkblob are acceptable
debug1: temporarily_use_uid: 1005/513 (e=1002/513)
seteuid 1005: Permission denied
debug1: do_cleanup
[EMAIL PROTECTED] var [1:221]

** Sshd terminates prematurely




--
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: Reading too few bytes

[Larry Hall (Cygwin)]

Paul J. Lucas wrote:

Here's a small test program:

#include 
#include 
#include 

using namespace std;

int main() {
int fd = ::open( "ESLF", O_RDONLY );
if ( fd == -1 ) {
cerr << "could not open ESLF file" << endl;
::exit( 1 );
}
char buf[ 64 ];
ssize_t bytesRead = ::read( fd, buf, sizeof( buf ) );
cout << "read " << bytesRead << " bytes" << endl;
::close( fd );
}

Here's the contents of a small biinary file "ESLF" (printed in hex bytes):

54 39 CA 1A 44

When I compile the program as:

g++ -o test test.cpp

and run it, it prints "read 5 bytes" as one would expect.  When I compile
the program as:

g++ -mno-cygwin -o test test.cpp

and run it, it prints "read 3 bytes".  Why?  How do I teg it to read all
5 bytes?


Ask the MinGW llist, since -mno-cygwin simply enables you to run their
compiler "indirectly".  As a result, questions about the MinGW compiler
are off-topic for this list, since it's not Cygwin.
See .

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

--
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: Reading too few bytes

[Eric Blake]

-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

>   int fd = ::open( "ESLF", O_RDONLY );

In the absence of O_TEXT or O_BINARY, open() on cygwin chooses the mode of
the underlying mount point but mingw always uses O_TEXT.

> Here's the contents of a small biinary file "ESLF" (printed in hex bytes):
> 
>   54 39 CA 1A 44

0x1a (ctrl-z) in a text file is Windows EOF.  Read binary data in binary
mode if you want all the bytes.

- --
Life is short - so eat dessert first!

Eric Blake [EMAIL PROTECTED]
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.2.1 (Cygwin)
Comment: Public key at home.comcast.net/~ericblake/eblake.gpg
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFEIN4g84KuGfSFAYARAhr6AJ9m1VrQ8p/0yJEnCNyjEapzDEDr3wCgmeXT
AuP4dUQ+dJyiL+2dxAAUCCU=
=iVtt
-END PGP SIGNATURE-

--
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: updatedb broken

[Eric Blake]

-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

According to Steve Kelem on 3/21/2006 5:29 PM:
> 
> I'm having trouble running updatedb.  I've tried findutils-4.2.27-1 and
> findutils-4.3.0-1.
> When I run updatedb manually, I get the message:
> updatedb needs to be able to execute /usr/lib/lapack/find, but cannot.

Sending the mail multiple times will not speed your answer; it only serves
to annoy people.  Read http://cygwin.com/ml/cygwin/2006-03/msg00594.html

- --
Life is short - so eat dessert first!

Eric Blake [EMAIL PROTECTED]
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.2.1 (Cygwin)
Comment: Public key at home.comcast.net/~ericblake/eblake.gpg
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFEIN8L84KuGfSFAYARApAnAJ48YRp9ruHC3mb3b4qF876UJENXgwCg16t2
bUxnQLnk0Cg8IwvExIPP1dE=
=1/AO
-END PGP SIGNATURE-

--
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: 1.5.19-4 g++ error: can't find entry point in cygwin1.dll

[Eric Blake]

-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

Quit posting raw email addresses: cygwin.com/acronyms/#PCYMTNQREAIYR

> From: "Eric Blake" 
     ^
>
> According to Anton Korneev on 3/20/2006 8:08 PM:
   cc1plus.exe - Entry Point not found
   The procedure entry point _impure_ptr could not be located in the
   dynamic link library cygwin1.dll.

   But, with the old versions (1.5.16) I don't have such problem.
> Without the information requested here:
>
 Problem reports:   http://cygwin.com/problems.html
> (including the output of 'cygcheck -svr' as a text attachment), we can
> only guess.  But it would appear that you did not fully install the new
> cygwin, or that an older cygwin1.dll is floating around somewhere in your
> system; and that during the upgrade you picked up the newer g++ that
> depends on the newer cygwin.  Be sure you only have one copy of
> cygwin1.dll on your machine, and that all cygwin processes are stopped,
> before reusing setup.exe to reinstall the cygwin package for the correct
> cygwin1.dll.


According to Alastair McDonald on 3/21/2006 8:16 AM:
> I suspect that you need to recompile g++ under 1.5.19 so that it finds the
> correct entry points in the new cygwin.dll.

And quit spreading misinformation.  Cygwin strives for backwards
compatibility (older apps linked against newer cygwin1.dll will continue
to work without recompilation of the older app).  The error was from the
reverse direction - the OP had a newer g++ trying to link against an older
cygwin1.dll because he had multiple copies in his filesystem.

- --
Life is short - so eat dessert first!

Eric Blake [EMAIL PROTECTED]
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.2.1 (Cygwin)
Comment: Public key at home.comcast.net/~ericblake/eblake.gpg
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFEIOAs84KuGfSFAYARAoTWAJ903Xvw6Ynk4zaLpYO+1pEjwj2JjwCfcoP8
HOL4gpHirhB6VLskcpneasA=
=raod
-END PGP SIGNATURE-

--
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: dlopen() bug

[skaller]

On Wed, 2006-03-22 at 03:35 +0100, Bernhard Loos wrote:

> If the C++library declares its calls as 'extern "C"' (as it is done in the 
> example),
> AFAIK there shouldn't be any problem.

Doesn't matter what you think you know -- it's not a permitted thing
to do, you never know what implementors are going to change.
 
> Anyway all this things aren't really related to the crash.

> I removed as much things as possible from the testcase and corrected
> the things mentioned above. 

Good. So now you've eliminated some possible causes of the
problem. Now it looks weird. A do nothing program is calling
dlopen() and dying for no reason.

> But I wasn't able to find a solution or even an explaination.
> 
> This is a part of the log from depency walker:
> 00:00:00.438: LoadLibraryA("H:\cygwin\test\CrashTest\CrashTest.dll") called 
> from "CYGWIN1.DLL" at address 0x6100FCB2 by thread 1.
> 00:00:00.500: Loaded "CRASHTEST.DLL" at address 0x00B4 by thread 1.  
> Successfully hooked module.
> 00:00:00.500: DllMain(0x00B4, DLL_PROCESS_ATTACH, 0x) in 
> "CRASHTEST.DLL" called by thread 1.
> 00:00:00.500: First chance exception 0xC005 (Access Violation) occurred 
> in "CYGWIN1.DLL" at address 0x610AE938 by thread 1.
> 00:00:00.500: Unloaded "CRASHTEST.DLL" at address 0x00B4 by thread 1.
> 00:00:00.516: LoadLibraryA("H:\cygwin\test\CrashTest\CrashTest.dll") returned 
> NULL by thread 1. Error: Unzulssiger Zugriff auf einen Speicherbereich (998).
> 00:00:00.578: First chance exception 0xC005 (Access Violation) occurred 
> at address 0x00B4F37E by thread 1.
> 00:00:00.578: First chance exception 0xC005 (Access Violation) occurred 
> in "CYGWIN1.DLL" at address 0x6108DFFC by thread 1.
> 00:00:00.578: First chance exception 0xC005 (Access Violation) occurred 
> in "CYGWIN1.DLL" at address 0x6108DFFC by thread 1.
> 00:00:00.578: First chance exception 0xC005 (Access Violation) occurred 
> in "CYGWIN1.DLL" at address 0x6108DFFC by thread 1.

Ouch. Hmm. But the code isn't doing anything!
This does look weird -- my system is doing what you're doing 
all the time. It was crashing badly for ages until someone pointed
out there was a bug in libstdc++ handling string crossing 
DLL boundaries. Since I fixed that the code works fine.

Perhaps you can try to build it, get Release Candidate 6
from:

http://felix.sf.net/


You'll need to install Ocaml and Python before it will work.
It does 100 or so triangle linkages like

MAIN --loadtime-- > RTL
\   ^
  \/
 dlopenloadtime
\/
  V /
  USER LIB

all C++ calling C++, works on two different Cygwins,
one running on XP32 and the other on XP64.


-- 
John Skaller 
Felix, successor to C++: http://felix.sf.net


--
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: Creating a custom Cygwin package server

[Brian Dessent]

Joshua Daniel Franklin wrote:

> Just be aware that you are entering unsupported territory,
> and get upset from:
> 
> http://xarch.tu-graz.ac.at/publ/cygwin/upset
> 
> "Unsupported" means don't ask this mailing list if you have
> problems. A good alternative might be:

You can also use "genini" which is meant to be used in place of upset
and is still supported, or rather it still exists in sourceware CVS.

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/