Re: [ANNOUNCEMENT] TEST RELEASE: Cygwin 2.6.0-0.11

[Brian Inglis]

On 2016-08-23 22:15, Brian Inglis wrote:

On 2016-08-23 10:11, Corinna Vinschen wrote:

On Aug 23 07:27, Brian Inglis wrote:

Compared lists of locale_t headers and functions for POSIX, Cygwin,
and glibc, attached below for comparison, and found:

* missed string.h(strerror_l) on my first check;
   not sure if you can implement that easily on Windows?

* GNU also supports wchar.h(wcsftime_l) and time.h(strptime_l);

* GNU also defines string.h(str[n]casecmp_l) functions as an extension,
   as well as in POSIX specified strings.h.


I just applied a couple of patches to add the missing strerror_l,
strptime_l and wcsftime_l.  I also added the missing str[n]casecmp_l
prototypes to strings.h.  I'll create a new test release in a bit.


GNU duplicates the POSIX strings.h(str[n]casecmp_l) in string.h also.


i.e. str[n]casecmp_l should be defined under #if __POSIX_VISIBLE >= 200809
but not defined under #ifdef __GNU_VISIBLE in string*s*.h,
and defined under #ifdef __GNU_VISIBLE but not defined under
#if __POSIX_VISIBLE >= 200809 in *string*.h;
strerror_l should be under #if __POSIX_VISIBLE  >= 200809 in *string*.h,
or its #includes.

--
Take care. Thanks, Brian Inglis, Calgary, Alberta, Canada

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

Re: [ANNOUNCEMENT] TEST RELEASE: Cygwin 2.6.0-0.11

[Brian Inglis]

On 2016-08-23 10:11, Corinna Vinschen wrote:

On Aug 23 07:27, Brian Inglis wrote:

Compared lists of locale_t headers and functions for POSIX, Cygwin,
and glibc, attached below for comparison, and found:

* missed string.h(strerror_l) on my first check;
   not sure if you can implement that easily on Windows?

* GNU also supports wchar.h(wcsftime_l) and time.h(strptime_l);

* GNU also defines string.h(str[n]casecmp_l) functions as an extension,
   as well as in POSIX specified strings.h.


I just applied a couple of patches to add the missing strerror_l,
strptime_l and wcsftime_l.  I also added the missing str[n]casecmp_l
prototypes to strings.h.  I'll create a new test release in a bit.


GNU duplicates the POSIX strings.h(str[n]casecmp_l) in string.h also.

--
Take care. Thanks, Brian Inglis, Calgary, Alberta, Canada

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

Cygwin 2.6.0-0.12 (?)

[Jim Reisert AD1C]

I just updated Cygwin-64 and got 2.6.0-0.12:

CYGWIN_NT-10.0 JJR 2.6.0(0.304/5/3) 2016-08-23 18:30 x86_64 Cygwin

I can not find any corresponding announcement E-mail.

-- 
Jim Reisert AD1C, , http://www.ad1c.us

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

Won't let me load the new software

[Walter Rutledge]

[main] Intellycheckupdates 5948 find_fast_cwd: Warning: coulnd't compute
FAST_CWD pointer.  

 

 

Please let me know what I need to do to get the program loaded and working
again please

 

 

Thanks

Walter Rutledge



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

Re: ImageMagick -6 not wotrking in WXP

[Ken Brown]

On 8/23/2016 7:28 PM, Rodrigo Medina wrote:

 Hi,

 From an xterm the command
 display picture.jpg
 exits without doing anything.


This is a known issue (on 32-bit Cygwin only), triggered by the recent 
update of libgd3:


  https://www.cygwin.com/ml/cygwin/2016-08/msg00339.html

The workaround is to revert to the previous version (2.1.1-2) of libgd3.

Ken


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

Native build fails in winsup/cygwin/libc/strptime.cc because of strtou?l_l()

[Hans-Bernhard Bröker]

Hello everyone,

building the git version of cygwin natively on Cygwin-2.5.3, I'm running 
into a problem with strtoul_l / strtol_l. 
winsup/cygwin/libc/strptime.cc fails to build:


hbbro@NB4 ~/prg/cygwin/bld/x86_64-unknown-cygwin/winsup/cygwin
$ make
c++wrap -O2 -g -fno-rtti -fno-exceptions -fno-use-cxa-atexit -Wall 
-Wstrict-aliasing -Wwrite-strings -fno-common -pipe -fbuiltin 
-fmessage-length=0 -MMD -Werror -fmerge-constants -ftracer 
-mcmodel=small -c -o strptime.o 
../../../../newlib-cygwin/winsup/cygwin/libc/strptime.cc
../../../../newlib-cygwin/winsup/cygwin/libc/strptime.cc: In function 
'era_info_t* get_era_info(const char*, locale_t)':
../../../../newlib-cygwin/winsup/cygwin/libc/strptime.cc:125:53: error: 
'strtol_l' was not declared in this scope

   ei[cur].offset = strtol_l (era, , 10, locale);
 ^
../../../../newlib-cygwin/winsup/cygwin/libc/strptime.cc: In function 
'char* __strptime(const char*, const char*, tm*, era_info_t**, 
alt_digits_t**, locale_t)':
../../../../newlib-cygwin/winsup/cygwin/libc/strptime.cc:390:50: error: 
'strtoul_l' was not declared in this scope

  width = strtoul_l (fmt - 1, , 10, locale);
  ^
make: *** 
[../../../../newlib-cygwin/winsup/cygwin/../Makefile.common:41: 
strptime.o] Fehler 1



Looking at the .ii file, it appears __GNU_VISIBLE is not active for this 
C++ source file?


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

Re: [ANNOUNCEMENT] Updated: mintty 2.4.3 and test release mintty 2.5.0

[Thomas Wolff]

Am 23.08.2016 um 11:30 schrieb Frank Fesevur:

2016-08-23 1:29 GMT+02:00 Thomas Wolff:

I have also uploaded mintty 2.5.0 as a test release with the following
change:

   * Revise DPI handling (#470; #492, #487); always consider individual
monitor DPI.

Note that this release introduces a slightly incompatible change in mintty.
The issue is that the interpretation of font size used to be suitable for
traditional resolution monitors and did not consider higher-DPI monitors or
the Windows monitor "zooming" feature which adjusts a virtual DPI.
As a result, the handling of changed DPI (when moving the window to another
monitor) introduced in 2.2.1 interfered with other aspects, the initial DPI
was not considered, and a number of unpleasant side effects were
occasionally observed.

I intend to change DPI handling to be consistent with the respective monitor
DPI (as optionally configured by "zooming factor"), and to comply with font
size interpreation of other applications, e.g. notepad. In consequence the
initial font size may be smaller (or larger) than before, depending on the
actual monitor geometry and configuration. Nothing would change on a
"standard" monitor configuration. To compensate, some people may have to
change their font size configuration.

I was about to write a bug report that I encountered with a maximized
mintty on a 125% secondary screen. When I minimized it to the taskbar
it would restore to my primary screen. But that problem is fixed now,
so for me 2.5.0 solves that problem even before reporting it. The
draft mail is thrown away ;-)

Thanks for the fix!
Thanks for the report anyway; I wasn't aware of that problem and I have 
no idea how it was magically resolved...

Thomas

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

[newlib-cygwin] Clarify Cygwin's support for Win32 paths

[Corinna Vinschen]

https://sourceware.org/git/gitweb.cgi?p=newlib-cygwin.git;h=8ad852ac1209ed34ea19126097e63b1e7a4e4d9d

commit 8ad852ac1209ed34ea19126097e63b1e7a4e4d9d
Author: Ken Brown 
Date:   Tue Aug 23 13:24:49 2016 -0400

Clarify Cygwin's support for Win32 paths

See https://www.cygwin.com/ml/cygwin/2016-08/msg00431.html.

Diff:
---
 winsup/doc/pathnames.xml | 14 +-
 1 file changed, 9 insertions(+), 5 deletions(-)

diff --git a/winsup/doc/pathnames.xml b/winsup/doc/pathnames.xml
index 3c0bdc1..6f9fefa 100644
--- a/winsup/doc/pathnames.xml
+++ b/winsup/doc/pathnames.xml
@@ -6,15 +6,18 @@
 
 Introduction
 
-Cygwin supports both POSIX- and Win32-style paths.  Directory
+The Cygwin DLL supports both POSIX- and Win32-style paths.  Directory
 delimiters may be either forward slashes or backslashes.  Paths using
 backslashes or starting with a drive letter are always handled as
 Win32 paths.  POSIX paths must only use forward slashes as delimiter,
 otherwise they are treated as Win32 paths and file access might fail
 in surprising ways.
 
-The usage of Win32 paths, though possible, is deprecated,
-since it circumvents important internal path handling mechanisms. 
+Although the Cygwin DLL supports Win32 paths, not all
+Cygwin applications support them.  Moreover, the usage of Win32 paths
+circumvents important internal path handling mechanisms.  This usage
+is therefore strongly deprecated and may be removed in a future
+release of Cygwin.
 See  and
  for more information.
 
@@ -451,11 +454,12 @@ filesystem borders by commands like find 
-xdev.
 
 Using native Win32 paths
 
-Using native Win32 paths in Cygwin, while possible, is generally
+Using native Win32 paths in Cygwin, while often possible, is generally
 inadvisable.  Those paths circumvent all internal integrity checking and
 bypass the information given in the Cygwin mount table.
 
-The following paths are treated as native Win32 paths in Cygwin:
+The following paths are treated as native Win32 paths by the
+Cygwin DLL (but not necessarily by Cygwin applications):
 
 
   

Re: #!/bin/sh and #!/bin/bash is not the same

[cyg Simple]

On 8/22/2016 3:31 PM, Morten Kjærulff wrote:
> 
> This:
> 
> date
> echo a | while read ; do
>   sleep 3 &
> done
> wait
> date
> 
> will run in 0 seconds, with BOTH /bin/sh and /bin/bash
> 
> This:
> 
> date
> while read ; do
>   sleep 3 &
> done < <(echo a)
> wait
> date
> 
> will run in 3 seconds with /bin/bash and get syntax errors with /bin/sh:
> syntax error near unexpected token `<'
> `done < <(echo a)'
> 

See
http://stackoverflow.com/questions/12120598/syntax-error-in-shell-script-with-process-substitution
for the answer to your quandary.

-- 
cyg Simple

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

Documentation patch: Clarify Cygwin's support for Win32 paths

[Ken Brown]

The attached patch is the one promised in

  https://www.cygwin.com/ml/cygwin/2016-08/msg00431.html

Ken
From e46d88d4141737f7f75ec1fb60b6e404f598a965 Mon Sep 17 00:00:00 2001
From: Ken Brown 
Date: Tue, 23 Aug 2016 13:24:49 -0400
Subject: [PATCH] Clarify Cygwin's support for Win32 paths

See https://www.cygwin.com/ml/cygwin/2016-08/msg00431.html.
---
 winsup/doc/pathnames.xml | 14 +-
 1 file changed, 9 insertions(+), 5 deletions(-)

diff --git a/winsup/doc/pathnames.xml b/winsup/doc/pathnames.xml
index 3c0bdc1..6f9fefa 100644
--- a/winsup/doc/pathnames.xml
+++ b/winsup/doc/pathnames.xml
@@ -6,15 +6,18 @@
 
 Introduction
 
-Cygwin supports both POSIX- and Win32-style paths.  Directory
+The Cygwin DLL supports both POSIX- and Win32-style paths.  Directory
 delimiters may be either forward slashes or backslashes.  Paths using
 backslashes or starting with a drive letter are always handled as
 Win32 paths.  POSIX paths must only use forward slashes as delimiter,
 otherwise they are treated as Win32 paths and file access might fail
 in surprising ways.
 
-The usage of Win32 paths, though possible, is deprecated,
-since it circumvents important internal path handling mechanisms. 
+Although the Cygwin DLL supports Win32 paths, not all
+Cygwin applications support them.  Moreover, the usage of Win32 paths
+circumvents important internal path handling mechanisms.  This usage
+is therefore strongly deprecated and may be removed in a future
+release of Cygwin.
 See  and
  for more information.
 
@@ -451,11 +454,12 @@ filesystem borders by commands like find 
-xdev.
 
 Using native Win32 paths
 
-Using native Win32 paths in Cygwin, while possible, is generally
+Using native Win32 paths in Cygwin, while often possible, is generally
 inadvisable.  Those paths circumvent all internal integrity checking and
 bypass the information given in the Cygwin mount table.
 
-The following paths are treated as native Win32 paths in Cygwin:
+The following paths are treated as native Win32 paths by the
+Cygwin DLL (but not necessarily by Cygwin applications):
 
 
   
-- 
2.8.3

[newlib-cygwin] Add missing LF in release text

[Corinna Vinschen]

https://sourceware.org/git/gitweb.cgi?p=newlib-cygwin.git;h=801fffcb7edda8f0e20d4bc02342c1f0df6aef27

commit 801fffcb7edda8f0e20d4bc02342c1f0df6aef27
Author: Corinna Vinschen 
Date:   Tue Aug 23 18:57:16 2016 +0200

Add missing LF in release text

Signed-off-by: Corinna Vinschen 

Diff:
---
 winsup/cygwin/release/2.6.0 | 1 +
 1 file changed, 1 insertion(+)

diff --git a/winsup/cygwin/release/2.6.0 b/winsup/cygwin/release/2.6.0
index aa25796..e13a31e 100644
--- a/winsup/cygwin/release/2.6.0
+++ b/winsup/cygwin/release/2.6.0
@@ -22,6 +22,7 @@ What's new:
 
 - New API: pthread_getname_np, pthread_setname_np.
 
+
 What changed:
 -

Re: Cygwin SSHD 32bit

[Erik Soderquist]

On Tue, Aug 23, 2016 at 12:43 PM, Marcel Grandemange wrote:
> I have entirely disabled the firewall at one point and this didn't help.
> Also keep in mind I did install bitvise ssh server to do a comparison and
> this worked.

Then I don't know, though I will be watching this thread for a
possible solution to my issue as well.

Also, please do not top post

-- Erik

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

RE: Cygwin SSHD 32bit

[Marcel Grandemange]

I have entirely disabled the firewall at one point and this didn't help. 
Also keep in mind I did install bitvise ssh server to do a comparison and 
this worked.

I would also like to add to this I also went through the effort of removing 
the SSH packages from Cygwin and built from source OpenSSH as well at one 
point which also I might add , had ther exact same issues.


-Original Message-
From: cygwin-ow...@cygwin.com [mailto:cygwin-ow...@cygwin.com] On Behalf Of 
Erik Soderquist
Sent: Tuesday, August 23, 2016 6:35 PM
To: cygwin@cygwin.com
Subject: Re: Cygwin SSHD 32bit

On Tue, Aug 23, 2016 at 12:24 PM, Marcel Grandemange  wrote:
> Good Day
>
> I would really appreciate it if someone can assist me.
>
> I have a Windows server 2003 server with the latest Cygwin installed.
>
> CYGWIN_NT-5.2  2.5.2(0.297/5/3) 2016-06-23 14:27 i686 Cygwin
>
> I have setup the SSH service as I have many a time before on other
> servers. The service successfully starts however I cannot SSH into the
> machine and am presented with the following error:
>
> ssh_exchange_identification: read: Connection reset by peer
>
> Note that I have allowed it through the firewall and this error is
> even present on the machine itself

I had this error on a Windows Home Server v1 (server 2003 SB base) and the 
only way I was able to resolve it was to disable the Microsoft firewall on 
the host and use an external firewall.  It seems to be a problem with the 
Microsoft firewall in my experience.  I had tried both opening tcp/22 and 
the executable (either/or as well as together) to no avail.

-- Erik

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


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

Re: Cygwin SSHD 32bit

[Erik Soderquist]

On Tue, Aug 23, 2016 at 12:24 PM, Marcel Grandemange  wrote:
> Good Day
>
> I would really appreciate it if someone can assist me.
>
> I have a Windows server 2003 server with the latest Cygwin installed.
>
> CYGWIN_NT-5.2  2.5.2(0.297/5/3) 2016-06-23 14:27 i686 Cygwin
>
> I have setup the SSH service as I have many a time before on other
> servers. The service successfully starts however I cannot SSH into the
> machine and am presented with the following error:
>
> ssh_exchange_identification: read: Connection reset by peer
>
> Note that I have allowed it through the firewall and this error is even
> present on the machine itself

I had this error on a Windows Home Server v1 (server 2003 SB base) and
the only way I was able to resolve it was to disable the Microsoft
firewall on the host and use an external firewall.  It seems to be a
problem with the Microsoft firewall in my experience.  I had tried
both opening tcp/22 and the executable (either/or as well as together)
to no avail.

-- Erik

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

Cygwin SSHD 32bit

[Marcel Grandemange]

Good Day

I would really appreciate it if someone can assist me.

I have a Windows server 2003 server with the latest Cygwin installed.

CYGWIN_NT-5.2  2.5.2(0.297/5/3) 2016-06-23 14:27 i686 Cygwin

I have setup the SSH service as I have many a time before on other
servers. The service successfully starts however I cannot SSH into the
machine and am presented with the following error:

ssh_exchange_identification: read: Connection reset by peer

Note that I have allowed it through the firewall and this error is even
present on the machine itself

$ ssh - localhost
OpenSSH_7.3p1, OpenSSL 1.0.2h  3 May 2016
debug1: Reading configuration data /etc/ssh_config
debug2: resolving "localhost" port 22
debug2: ssh_connect_direct: needpriv 0
debug1: Connecting to localhost [127.0.0.1] port 22.
debug1: Connection established.
debug1: identity file /home/Marcel/.ssh/id_rsa type 1
debug1: key_load_public: No such file or directory
debug1: identity file /home/Marcel/.ssh/id_rsa-cert type -1
debug1: key_load_public: No such file or directory
debug1: identity file /home/Marcel/.ssh/id_dsa type -1
debug1: key_load_public: No such file or directory
debug1: identity file /home/Marcel/.ssh/id_dsa-cert type -1
debug1: key_load_public: No such file or directory
debug1: identity file /home/Marcel/.ssh/id_ecdsa type -1
debug1: key_load_public: No such file or directory
debug1: identity file /home/Marcel/.ssh/id_ecdsa-cert type -1
debug1: key_load_public: No such file or directory
debug1: identity file /home/Marcel/.ssh/id_ed25519 type -1
debug1: key_load_public: No such file or directory
debug1: identity file /home/Marcel/.ssh/id_ed25519-cert type -1
debug1: Enabling compatibility mode for protocol 2.0
debug1: Local version string SSH-2.0-OpenSSH_7.3
ssh_exchange_identification: read: Connection reset by peer


There are no hosts.allow or hosts.deny files present on this machine and
Cygwin has been reinstalled several times under different users to test
permission theories I had.

When I change the permission of the /var/empty folder to my current user
to allow me to run the sshd Daemon as my current user I see the following:

$ /usr/sbin/sshd -d
debug2: load_server_config: filename /etc/sshd_config
debug2: load_server_config: done config len = 288
debug2: parse_server_config: config /etc/sshd_config len 288
debug3: /etc/sshd_config:19 setting Protocol 2
debug3: /etc/sshd_config:43 setting PermitRootLogin yes
debug3: /etc/sshd_config:53 setting AuthorizedKeysFile
.ssh/authorized_keys
debug3: /etc/sshd_config:79 setting KerberosOrLocalPasswd yes
debug3: /etc/sshd_config:109 setting UsePrivilegeSeparation yes
debug3: /etc/sshd_config:125 setting Subsystem sftp
/usr/sbin/sftp-server
debug1: sshd version OpenSSH_7.3, OpenSSL 1.0.2h  3 May 2016
debug1: private host key #0: ssh-rsa
SHA256:eH6f4o7GOW6xajSHXQywMtsWmocl4PzgNy6o1Dmbxoc
debug1: private host key #1: ssh-dss
SHA256:pZxX5Ddzn6mR5sOlRPjNV3mtXgJTSpZjkSgpySz0wsI
debug1: private host key #2: ecdsa-sha2-nistp256
SHA256:dyTs9OY0gaXfEv09jx3DZYlgEUjBQKKGkQJ/ioTG2qQ
debug1: private host key #3: ssh-ed25519
SHA256:KeIpfsnyBDf2V7OH9BFjx8xmhh/HwqJ2QUOyUPy6hsM
debug1: rexec_argv[0]='/usr/sbin/sshd'
debug1: rexec_argv[1]='-d'
debug2: fd 3 setting O_NONBLOCK
debug1: Bind to port 22 on 0.0.0.0.
Server listening on 0.0.0.0 port 22.
debug1: fd 4 clearing O_NONBLOCK
debug1: Server will not fork when running in debugging mode.
debug3: send_rexec_state: entering fd = 7 config len 288
debug3: ssh_msg_send: type 0
debug3: send_rexec_state: done
debug1: rexec start in 4 out 4 newsock 4 pipe -1 sock 7
debug1: inetd sockets after dupping: 3, 3

It seems that the SSH daemon never gets to the “SSH2_MSG_KEXINIT sent”
stage.

I have attempted numerous modifications of the setup and daemon config,
eventually just to confirm all networking aspects are correct I installed
bitvise ssh server and this worked, however I really need to get Cygwin
working correctly. I have also tried changing the default port to 2200 but
all yield same result.

Any Ideas?

Kind Regards
Marcel Grandemange
OPS Manager
Antfarm (PTY) Ltd
Tel: 012 346 7511
Direct : 0100072640

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

Re: The Cygwin User Guide on path names

[Corinna Vinschen]

On Aug 23 12:16, Ken Brown wrote:
> The section on path names in the user guide
> (https://cygwin.com/cygwin-ug-net/using.html#using-pathnames) strikes me as
> misleading when it says in the first sentence, "Cygwin supports both POSIX-
> and Win32-style paths."  I think it would be better to say "The Cygwin DLL"
> and to emphasize that Cygwin applications do *not* necessarily support Win32
> paths.  See
> 
>   https://www.cygwin.com/ml/cygwin/2016-08/msg00409.html
> 
> for a recent example where this has come up.  It's also come up in
> connection with git and emacs and probably many other applications.
> 
> The documentation also says, "The usage of Win32 paths, though possible, is
> deprecated"  I wonder if this should be strengthened to say something
> like, "The usage of Win32 paths, though possible, is strongly deprecated and
> may be removed in a future release of Cygwin."
> 
> I'll be glad to prepare a documentation patch.

Yes, please.  Thanks for the offer,
Corinna

-- 
Corinna Vinschen  Please, send mails regarding Cygwin to
Cygwin Maintainer cygwin AT cygwin DOT com
Red Hat


signature.asc
Description: PGP signature

The Cygwin User Guide on path names

[Ken Brown]

The section on path names in the user guide 
(https://cygwin.com/cygwin-ug-net/using.html#using-pathnames) strikes me 
as misleading when it says in the first sentence, "Cygwin supports both 
POSIX- and Win32-style paths."  I think it would be better to say "The 
Cygwin DLL" and to emphasize that Cygwin applications do *not* 
necessarily support Win32 paths.  See


  https://www.cygwin.com/ml/cygwin/2016-08/msg00409.html

for a recent example where this has come up.  It's also come up in 
connection with git and emacs and probably many other applications.


The documentation also says, "The usage of Win32 paths, though possible, 
is deprecated"  I wonder if this should be strengthened to say 
something like, "The usage of Win32 paths, though possible, is strongly 
deprecated and may be removed in a future release of Cygwin."


I'll be glad to prepare a documentation patch.

Ken

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

Re: [ANNOUNCEMENT] TEST RELEASE: Cygwin 2.6.0-0.11

[Corinna Vinschen]

On Aug 23 07:27, Brian Inglis wrote:
> Compared lists of locale_t headers and functions for POSIX, Cygwin,
> and glibc, attached below for comparison, and found:
> 
> * missed string.h(strerror_l) on my first check;
>not sure if you can implement that easily on Windows?
> 
> * GNU also supports wchar.h(wcsftime_l) and time.h(strptime_l);
> 
> * GNU also defines string.h(str[n]casecmp_l) functions as an extension,
>as well as in POSIX specified strings.h.

I just applied a couple of patches to add the missing strerror_l,
strptime_l and wcsftime_l.  I also added the missing str[n]casecmp_l
prototypes to strings.h.  I'll create a new test release in a bit.


Thanks a lot,
Corinna

-- 
Corinna Vinschen  Please, send mails regarding Cygwin to
Cygwin Maintainer cygwin AT cygwin DOT com
Red Hat


signature.asc
Description: PGP signature

[newlib-cygwin] Export and document strerror_l, strptime_l, wcsftime_l from Cygwin

[Corinna Vinschen]

https://sourceware.org/git/gitweb.cgi?p=newlib-cygwin.git;h=2ea3993619206a8f87593d35f2e823223793be0d

commit 2ea3993619206a8f87593d35f2e823223793be0d
Author: Corinna Vinschen 
Date:   Tue Aug 23 17:50:21 2016 +0200

Export and document strerror_l, strptime_l, wcsftime_l from Cygwin

Bump Cygwin API minor number.

Signed-off-by: Corinna Vinschen 

Diff:
---
 winsup/cygwin/include/cygwin/version.h |  3 ++-
 winsup/cygwin/release/2.6.0| 12 ++--
 winsup/doc/new-features.xml| 10 +-
 winsup/doc/posix.xml   |  3 +++
 4 files changed, 16 insertions(+), 12 deletions(-)

diff --git a/winsup/cygwin/include/cygwin/version.h 
b/winsup/cygwin/include/cygwin/version.h
index a1b8a62..820d28b 100644
--- a/winsup/cygwin/include/cygwin/version.h
+++ b/winsup/cygwin/include/cygwin/version.h
@@ -468,12 +468,13 @@ details. */
wcstoul_l, wcstoull_l.
   302: Export nl_langinfo_l.
   303: Export pthread_getname_np, pthread_setname_np.
+  304: Export strerror_l, strptime_l, wcsftime_l.
 
   Note that we forgot to bump the api for ualarm, strtoll, strtoull,
   sigaltstack, sethostname. */
 
 #define CYGWIN_VERSION_API_MAJOR 0
-#define CYGWIN_VERSION_API_MINOR 303
+#define CYGWIN_VERSION_API_MINOR 304
 
 /* There is also a compatibity version number associated with the shared memory
regions.  It is incremented when incompatible changes are made to the shared
diff --git a/winsup/cygwin/release/2.6.0 b/winsup/cygwin/release/2.6.0
index 8255450..aa25796 100644
--- a/winsup/cygwin/release/2.6.0
+++ b/winsup/cygwin/release/2.6.0
@@ -9,13 +9,13 @@ What's new:
   iswalpha_l, iswblank_l, iswcntrl_l, iswctype_l, iswdigit_l, iswgraph_l,
   iswlower_l, iswprint_l, iswpunct_l, iswspace_l, iswupper_l, iswxdigit_l,
   isxdigit_l, tolower_l, toupper_l, towctrans_l, towlower_l, towupper_l,
-  wctrans_l, wctype_l, strcasecmp_l, strcoll_l, strfmon_l, strftime_l,
-  strncasecmp_l, strxfrm_l, wcscasecmp_l, wcscoll_l, wcstrncasecmp_l,
-  wcstrxfrm_l.
+  wctrans_l, wctype_l, strcasecmp_l, strcoll_l, strerror_l, strfmon_l,
+  strftime_l, strncasecmp_l, strxfrm_l, wcscasecmp_l, wcscoll_l,
+  wcstrncasecmp_l, wcstrxfrm_l.
 
-  New API, GNU extensions: isascii_l, toascii_l, strtod_l, strtof_l, strtol_l,
-  strtold_l, strtoll_l, strtoul_l, strtoull_l, wcstod_l, wcstof_l, wcstol_l,
-  wcstold_l, wcstoll_l, wcstoul_l, wcstoull_l.
+  New API, GNU extensions: isascii_l, toascii_l, strptime_l, strtod_l,
+  strtof_l, strtol_l, strtold_l, strtoll_l, strtoul_l, strtoull_l, wcsftime_l,
+  wcstod_l, wcstof_l, wcstol_l, wcstold_l, wcstoll_l, wcstoul_l, wcstoull_l.
 
 - locale(1) now supports a -i/--input option to fetch the current input
   locale (this is basically equivalent to the current keyboard layout setting).
diff --git a/winsup/doc/new-features.xml b/winsup/doc/new-features.xml
index 7e39316..edd1466 100644
--- a/winsup/doc/new-features.xml
+++ b/winsup/doc/new-features.xml
@@ -19,15 +19,15 @@ isgraph_l, islower_l, isprint_l, ispunct_l, isspace_l, 
isupper_l, iswalnum_l,
 iswalpha_l, iswblank_l, iswcntrl_l, iswctype_l, iswdigit_l, iswgraph_l,
 iswlower_l, iswprint_l, iswpunct_l, iswspace_l, iswupper_l, iswxdigit_l,
 isxdigit_l, tolower_l, toupper_l, towctrans_l, towlower_l, towupper_l,
-wctrans_l, wctype_l, strcasecmp_l, strcoll_l, strfmon_l, strftime_l,
-strncasecmp_l, strxfrm_l, wcscasecmp_l, wcscoll_l, wcstrncasecmp_l,
+wctrans_l, wctype_l, strcasecmp_l, strcoll_l, strerror_l, strfmon_l,
+strftime_l, strncasecmp_l, strxfrm_l, wcscasecmp_l, wcscoll_l, wcstrncasecmp_l,
 wcstrxfrm_l.
 
 
 
-New API, GNU extensions: isascii_l, toascii_l, strtod_l, strtof_l, strtol_l,
-strtold_l, strtoll_l, strtoul_l, strtoull_l, wcstod_l, wcstof_l, wcstol_l,
-wcstold_l, wcstoll_l, wcstoul_l, wcstoull_l.
+New API, GNU extensions: isascii_l, toascii_l, strptime_l, strtod_l, strtof_l,
+strtol_l, strtold_l, strtoll_l, strtoul_l, strtoull_l, wcsftime_l, wcstod_l,
+wcstof_l, wcstol_l, wcstold_l, wcstoll_l, wcstoul_l, wcstoull_l.
 
 
 
diff --git a/winsup/doc/posix.xml b/winsup/doc/posix.xml
index babf115..e80da4d 100644
--- a/winsup/doc/posix.xml
+++ b/winsup/doc/posix.xml
@@ -924,6 +924,7 @@ also IEEE Std 1003.1-2008 (POSIX.1-2008).
 strcspn
 strdup
 strerror
+strerror_l
 strerror_r
 strfmon
 strfmon_l
@@ -1353,6 +1354,7 @@ also IEEE Std 1003.1-2008 (POSIX.1-2008).
 sincosf
 sincosl
 strchrnul
+strptime_l
 strtod_l
 strtof_l
 strtol_l
@@ -1370,6 +1372,7 @@ also IEEE Std 1003.1-2008 (POSIX.1-2008).
 vasnprintf
 vasprintf
 vasprintf_r
+wcsftime_l
 wcstod_l
 wcstof_l
 wcstol_l

[newlib-cygwin] Implement GNU extension wcsftime_l

[Corinna Vinschen]

https://sourceware.org/git/gitweb.cgi?p=newlib-cygwin.git;h=e636fe3d489ca60d171ae1fcf5fabecdd1665af7

commit e636fe3d489ca60d171ae1fcf5fabecdd1665af7
Author: Corinna Vinschen 
Date:   Tue Aug 23 17:41:50 2016 +0200

Implement GNU extension wcsftime_l

Signed-off-by: Corinna Vinschen 

Diff:
---
 newlib/libc/include/wchar.h | 4 
 newlib/libc/time/strftime.c | 5 ++---
 winsup/cygwin/common.din| 1 +
 3 files changed, 7 insertions(+), 3 deletions(-)

diff --git a/newlib/libc/include/wchar.h b/newlib/libc/include/wchar.h
index b7f8374..4766ca9 100644
--- a/newlib/libc/include/wchar.h
+++ b/newlib/libc/include/wchar.h
@@ -132,6 +132,10 @@ wchar_t*_EXFUN(_wcsdup_r, (struct _reent *, const 
wchar_t * ));
 size_t _EXFUN(wcscspn, (const wchar_t *, const wchar_t *));
 size_t  _EXFUN(wcsftime, (wchar_t *__restrict, size_t,
const wchar_t *__restrict, const struct tm 
*__restrict));
+#if __GNU_VISIBLE
+size_t  wcsftime_l (wchar_t *__restrict, size_t, const wchar_t *__restrict,
+   const struct tm *__restrict, locale_t);
+#endif
 size_t _EXFUN(wcslcat, (wchar_t *, const wchar_t *, size_t));
 size_t _EXFUN(wcslcpy, (wchar_t *, const wchar_t *, size_t));
 size_t _EXFUN(wcslen, (const wchar_t *));
diff --git a/newlib/libc/time/strftime.c b/newlib/libc/time/strftime.c
index 46c2d95..f11a002 100644
--- a/newlib/libc/time/strftime.c
+++ b/newlib/libc/time/strftime.c
@@ -311,6 +311,7 @@ locale, hard-coding the "C" locale settings.
 #  define STRLEN(a)strlen(a)
 # else
 #  define strftime wcsftime/* Alternate function name */
+#  define strftime_l   wcsftime_l  /* Alternate function name */
 #  define CHAR wchar_t /* string type basis */
 #  define CQ(a)L##a/* character constant qualifier 
*/
 #  define snprintf swprintf/* wide-char equivalent function name */
@@ -1460,9 +1461,8 @@ _DEFUN (strftime, (s, maxsize, format, tim_p),
 #endif /* !_WANT_C99_TIME_FORMATS */
 }
 
-#if !defined(MAKE_WCSFTIME)
 size_t
-strftime_l (char *__restrict s, size_t maxsize, const char *__restrict format,
+strftime_l (CHAR *__restrict s, size_t maxsize, const CHAR *__restrict format,
const struct tm *__restrict tim_p, struct __locale_t *locale)
 {
 #ifdef _WANT_C99_TIME_FORMATS
@@ -1479,7 +1479,6 @@ strftime_l (char *__restrict s, size_t maxsize, const 
char *__restrict format,
   return __strftime (s, maxsize, format, tim_p, locale, NULL, NULL);
 #endif /* !_WANT_C99_TIME_FORMATS */
 }
-#endif
 
 /* The remainder of this file can serve as a regression test.  Compile
  *  with -D_REGRESSION_TEST.  */
diff --git a/winsup/cygwin/common.din b/winsup/cygwin/common.din
index 7fce587..12705fe 100644
--- a/winsup/cygwin/common.din
+++ b/winsup/cygwin/common.din
@@ -1522,6 +1522,7 @@ wcscpy NOSIGFE
 wcscspn NOSIGFE
 wcsdup NOSIGFE
 wcsftime SIGFE
+wcsftime_l SIGFE
 wcslcat NOSIGFE
 wcslcpy NOSIGFE
 wcslen NOSIGFE

[newlib-cygwin] Implement GNU extension strptime_l

[Corinna Vinschen]

https://sourceware.org/git/gitweb.cgi?p=newlib-cygwin.git;h=0ecb846d2ba2525ec1ec7842a8291da2a2f8f823

commit 0ecb846d2ba2525ec1ec7842a8291da2a2f8f823
Author: Corinna Vinschen 
Date:   Tue Aug 23 17:42:47 2016 +0200

Implement GNU extension strptime_l

Signed-off-by: Corinna Vinschen 

Diff:
---
 newlib/libc/include/time.h |  5 +++
 newlib/libc/time/strptime.c| 80 ++
 winsup/cygwin/common.din   |  1 +
 winsup/cygwin/libc/strptime.cc | 74 --
 4 files changed, 94 insertions(+), 66 deletions(-)

diff --git a/newlib/libc/include/time.h b/newlib/libc/include/time.h
index 47ab578..ec67cb2 100644
--- a/newlib/libc/include/time.h
+++ b/newlib/libc/include/time.h
@@ -92,6 +92,11 @@ char  *_EXFUN(strptime, (const char *__restrict,
 const char *__restrict,
 struct tm *__restrict));
 #endif
+#if __GNU_VISIBLE
+char *strptime_l (const char *__restrict, const char *__restrict,
+ struct tm *__restrict, locale_t);
+#endif
+
 #if __POSIX_VISIBLE
 _VOID  _EXFUN(tzset,   (_VOID));
 #endif
diff --git a/newlib/libc/time/strptime.c b/newlib/libc/time/strptime.c
index a057450..c0861eb 100644
--- a/newlib/libc/time/strptime.c
+++ b/newlib/libc/time/strptime.c
@@ -30,6 +30,7 @@
  * OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF
  * ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. */
 
+#define _GNU_SOURCE
 #include 
 #include 
 #include 
@@ -68,14 +69,15 @@ is_leap_year (int year)
 
 /* Needed for strptime. */
 static int
-match_string (const char *__restrict *buf, const char * const*strs)
+match_string (const char *__restrict *buf, const char * const*strs,
+ locale_t locale)
 {
 int i = 0;
 
 for (i = 0; strs[i] != NULL; ++i) {
int len = strlen (strs[i]);
 
-   if (strncasecmp (*buf, strs[i], len) == 0) {
+   if (strncasecmp_l (*buf, strs[i], len, locale) == 0) {
*buf += len;
return i;
}
@@ -148,25 +150,20 @@ set_week_number_mon4 (struct tm *timeptr, int wnum)
 }
 }
 
-/* strptime: roken */
-//extern "C"
 char *
-//strptime (const char *buf, const char *format, struct tm *timeptr)
-_DEFUN (strptime, (buf, format, timeptr),
-   _CONST char *__restrict buf _AND
-   _CONST char *__restrict format _AND
-   struct tm *__restrict timeptr)
+strptime_l (const char *buf, const char *format, struct tm *timeptr,
+   locale_t locale)
 {
 char c;
 int ymd = 0;
 
-const struct lc_time_T *_CurrentTimeLocale = __get_current_time_locale ();
+const struct lc_time_T *_CurrentTimeLocale = __get_time_locale (locale);
 for (; (c = *format) != '\0'; ++format) {
char *s;
int ret;
 
-   if (isspace ((unsigned char) c)) {
-   while (isspace ((unsigned char) *buf))
+   if (isspace_l ((unsigned char) c, locale)) {
+   while (isspace_l ((unsigned char) *buf, locale))
++buf;
} else if (c == '%' && format[1] != '\0') {
c = *++format;
@@ -174,21 +171,21 @@ _DEFUN (strptime, (buf, format, timeptr),
c = *++format;
switch (c) {
case 'A' :
-   ret = match_string (, _ctloc (weekday));
+   ret = match_string (, _ctloc (weekday), locale);
if (ret < 0)
return NULL;
timeptr->tm_wday = ret;
ymd |= SET_WDAY;
break;
case 'a' :
-   ret = match_string (, _ctloc (wday));
+   ret = match_string (, _ctloc (wday), locale);
if (ret < 0)
return NULL;
timeptr->tm_wday = ret;
ymd |= SET_WDAY;
break;
case 'B' :
-   ret = match_string (, _ctloc (month));
+   ret = match_string (, _ctloc (month), locale);
if (ret < 0)
return NULL;
timeptr->tm_mon = ret;
@@ -196,14 +193,14 @@ _DEFUN (strptime, (buf, format, timeptr),
break;
case 'b' :
case 'h' :
-   ret = match_string (, _ctloc (mon));
+   ret = match_string (, _ctloc (mon), locale);
if (ret < 0)
return NULL;
timeptr->tm_mon = ret;
ymd |= SET_MON;
break;
case 'C' :
-   ret = strtol (buf, , 10);
+   ret = strtol_l (buf, , 10, locale);
if (s == buf)
return NULL;
timeptr->tm_year = (ret * 100) - tm_year_base;
@@ -211,14 +208,14 @@ _DEFUN (strptime, (buf, format, timeptr),
ymd |= SET_YEAR;
break;
case 'c' :  /* %a %b %e %H:%M:%S %Y */
-   s = strptime (buf, 

[newlib-cygwin] Implement missing POSIX-1.2008 function strerror_l

[Corinna Vinschen]

https://sourceware.org/git/gitweb.cgi?p=newlib-cygwin.git;h=463a8afaa58b80926b0236e14a6171f45c1b784f

commit 463a8afaa58b80926b0236e14a6171f45c1b784f
Author: Corinna Vinschen 
Date:   Tue Aug 23 17:49:24 2016 +0200

Implement missing POSIX-1.2008 function strerror_l

Signed-off-by: Corinna Vinschen 

Diff:
---
 newlib/libc/string/strerror.c | 19 ++-
 winsup/cygwin/common.din  |  1 +
 winsup/cygwin/errno.cc|  7 +++
 3 files changed, 26 insertions(+), 1 deletion(-)

diff --git a/newlib/libc/string/strerror.c b/newlib/libc/string/strerror.c
index fcef33e..49e5f7e 100644
--- a/newlib/libc/string/strerror.c
+++ b/newlib/libc/string/strerror.c
@@ -7,14 +7,18 @@
 
 /*
 FUNCTION
-   <>---convert error number to string
+   <>, <>---convert error number to string
 
 INDEX
strerror
 
+INDEX
+   strerror_l
+
 ANSI_SYNOPSIS
#include 
char *strerror(int <[errnum]>);
+   char *strerror_l(int <[errnum]>, locale_t locale);
char *_strerror_r(struct _reent <[ptr]>, int <[errnum]>,
  int <[internal]>, int *<[error]>);
 
@@ -29,6 +33,10 @@ string.  The value of <[errnum]> is usually a copy of 
<>.
 If <> is not a known error number, the result points to an
 empty string.
 
+<> is like <> but creates a string in a format
+as expected in locale <[locale]>.  If <[locale]> is LC_GLOBAL_LOCALE or
+not a valid locale object, the behaviour is undefined.
+
 This implementation of <> prints out the following strings
 for each of the values defined in `<>':
 
@@ -330,6 +338,8 @@ PORTABILITY
 ANSI C requires <>, but does not specify the strings used
 for each error number.
 
+<> is POSIX-1.2008.
+
 Although this implementation of <> is reentrant (depending
 on <<_user_strerror>>), ANSI C declares that subsequent calls to
 <> may overwrite the result string; therefore portable
@@ -893,3 +903,10 @@ _DEFUN(strerror, (int),
 {
   return _strerror_r (_REENT, errnum, 0, NULL);
 }
+
+char *
+strerror_l (int errnum, locale_t locale)
+{
+  /* We don't support per-locale error messages. */
+  return _strerror_r (_REENT, errnum, 0, NULL);
+}
diff --git a/winsup/cygwin/common.din b/winsup/cygwin/common.din
index 8f7a282..7fce587 100644
--- a/winsup/cygwin/common.din
+++ b/winsup/cygwin/common.din
@@ -1345,6 +1345,7 @@ strcpy NOSIGFE
 strcspn NOSIGFE
 strdup SIGFE
 strerror SIGFE
+strerror_l SIGFE
 strerror_r SIGFE
 strfmon SIGFE
 strfmon_l SIGFE
diff --git a/winsup/cygwin/errno.cc b/winsup/cygwin/errno.cc
index 7e07a61..9168e9b 100644
--- a/winsup/cygwin/errno.cc
+++ b/winsup/cygwin/errno.cc
@@ -403,6 +403,13 @@ strerror (int errnum)
   return result;
 }
 
+extern "C" char *
+strerror_l (int errnum, locale_t locale)
+{
+  /* We don't provide localized system error messages (yet?). */
+  return strerror (errnum);
+}
+
 /* Newlib's  provides declarations for two strerror_r
variants, according to preprocessor feature macros.  However, it
returns "" instead of "Unknown error ...", so we override both

[ANNOUNCEMENT] Updated: {libEMF/libEMF1/libEMF-devel}-1.0.8-1: A library for generating Enhanced Metafiles

[Dr . Volker Zell]

Hi

New versions of 'libEMF/libEMF1/libEMF-devel' have been uploaded to a server 
near you.

 o Updated to latest upstream release
 o Build for cygwin 2.5.2 with gcc-5.4.0
 o [x86] Removed .la file from /usr/lib

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

Updated: {libEMF/libEMF1/libEMF-devel}-1.0.8-1: A library for generating Enhanced Metafiles

[Dr . Volker Zell]

Hi

New versions of 'libEMF/libEMF1/libEMF-devel' have been uploaded to a server 
near you.

 o Updated to latest upstream release
 o Build for cygwin 2.5.2 with gcc-5.4.0
 o [x86] Removed .la file from /usr/lib

Re: zsh bug: Incorrect (misleading) error message in zsh when using rm and drive letter is specified

[Ken Brown]

On 8/23/2016 3:57 AM, Ronald Fischer wrote:

I observed this:

-0-1- ~/exp  > rm e:/media/*
zsh: sure you want to delete all the files in /home/FISRONA/exp/e:/media
[yn]? n

We see that zsh is catching the case that I want to rm all files in the
specified directory, and warns me about it (note that it is a zsh error
message, not a error message from rm), which is fine and expected
behaviour (which can be controlled by the zsh option RM_STAR_SILENT).
However, the path printed by zsh is wrong: It interprets e:/... as
relative path, while it is an absolute path. It seems that zsh is not
translating windows pathes using a drive letter to /cygdrive/... pathes
in this case.


Some Cygwin applications do this translation, but not all.  You're 
better off just using Posix paths.


Ken


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

[newlib-cygwin] Send thread names to debugger

[Jon TURNEY]

https://sourceware.org/git/gitweb.cgi?p=newlib-cygwin.git;h=9e0f9ec7aedb59cacaf5831a27b925d791090113

commit 9e0f9ec7aedb59cacaf5831a27b925d791090113
Author: Jon Turney 
Date:   Thu Jul 28 00:40:23 2016 +0100

Send thread names to debugger

GDB since commit 24cdb46e [1] can report and use these names.

Add utility function SetThreadName(), which sends a thread name to the
debugger.

Use that:
- to set the default thread name for main thread and newly created pthreads.
- in pthread_setname_np() for user thread names.
- for helper thread names in cygthread::create()
- for helper threads which are created directly with CreateThread.

Note that there can still be anonymous threads, created by system or
injected DLLs.

[1] 
https://sourceware.org/git/gitweb.cgi?p=binutils-gdb.git;h=24cdb46e9f0a694b4fbc11085e094857f08c0419

Diff:
---
 winsup/cygwin/cygthread.cc  |  2 ++
 winsup/cygwin/dcrt0.cc  |  1 +
 winsup/cygwin/exceptions.cc |  2 +-
 winsup/cygwin/miscfuncs.cc  | 26 ++
 winsup/cygwin/miscfuncs.h   |  2 ++
 winsup/cygwin/net.cc|  1 +
 winsup/cygwin/thread.cc |  5 +
 7 files changed, 38 insertions(+), 1 deletion(-)

diff --git a/winsup/cygwin/cygthread.cc b/winsup/cygwin/cygthread.cc
index b9d706b..4404e4a 100644
--- a/winsup/cygwin/cygthread.cc
+++ b/winsup/cygwin/cygthread.cc
@@ -213,6 +213,8 @@ cygthread::create ()
this, 0, );
   if (!htobe)
api_fatal ("CreateThread failed for %s - %p<%y>, %E", __name, h, id);
+  else
+   SetThreadName (GetThreadId (htobe), __name);
   thread_printf ("created name '%s', thread %p, id %y", __name, h, id);
 #ifdef DEBUGGING
   terminated = false;
diff --git a/winsup/cygwin/dcrt0.cc b/winsup/cygwin/dcrt0.cc
index 2328411..8ddee0c 100644
--- a/winsup/cygwin/dcrt0.cc
+++ b/winsup/cygwin/dcrt0.cc
@@ -964,6 +964,7 @@ dll_crt0_1 (void *)
   if (cp > __progname && ascii_strcasematch (cp, ".exe"))
*cp = '\0';
 }
+  SetThreadName (GetCurrentThreadId (), program_invocation_short_name);
 
   (void) xdr_set_vprintf (_vwarnx);
   cygwin_finished_initializing = true;
diff --git a/winsup/cygwin/exceptions.cc b/winsup/cygwin/exceptions.cc
index d65f56e..0f5a890 100644
--- a/winsup/cygwin/exceptions.cc
+++ b/winsup/cygwin/exceptions.cc
@@ -1288,7 +1288,7 @@ DWORD WINAPI
 dumpstack_overflow_wrapper (PVOID arg)
 {
   cygwin_exception *exc = (cygwin_exception *) arg;
-
+  SetThreadName (GetCurrentThreadId (), "__dumpstack_overflow");
   exc->dumpstack ();
   return 0;
 }
diff --git a/winsup/cygwin/miscfuncs.cc b/winsup/cygwin/miscfuncs.cc
index d0e4bf7..5a63b26 100644
--- a/winsup/cygwin/miscfuncs.cc
+++ b/winsup/cygwin/miscfuncs.cc
@@ -1110,3 +1110,29 @@ wmemcpy: 
\n\
.seh_endproc\n\
 ");
 #endif
+
+/* Signal the thread name to any attached debugger
+
+   (See "How to: Set a Thread Name in Native Code"
+   https://msdn.microsoft.com/en-us/library/xcb2z8hs.aspx) */
+
+#define MS_VC_EXCEPTION 0x406D1388
+
+void
+SetThreadName(DWORD dwThreadID, const char* threadName)
+{
+  if (!IsDebuggerPresent ())
+return;
+
+  ULONG_PTR info[] =
+{
+  0x1000, /* type, must be 0x1000 */
+  (ULONG_PTR) threadName, /* pointer to threadname */
+  dwThreadID, /* thread ID (+ flags on x86_64) */
+#ifdef __X86__
+  0,  /* flags, must be zero */
+#endif
+};
+
+  RaiseException (MS_VC_EXCEPTION, 0, sizeof (info)/sizeof (ULONG_PTR), 
(ULONG_PTR *) );
+}
diff --git a/winsup/cygwin/miscfuncs.h b/winsup/cygwin/miscfuncs.h
index a885dcf..5087299 100644
--- a/winsup/cygwin/miscfuncs.h
+++ b/winsup/cygwin/miscfuncs.h
@@ -85,4 +85,6 @@ extern "C" HANDLE WINAPI CygwinCreateThread 
(LPTHREAD_START_ROUTINE thread_func,
 DWORD creation_flags,
 LPDWORD thread_id);
 
+void SetThreadName (DWORD dwThreadID, const char* threadName);
+
 #endif /*_MISCFUNCS_H*/
diff --git a/winsup/cygwin/net.cc b/winsup/cygwin/net.cc
index 52b3d98..e4805d3 100644
--- a/winsup/cygwin/net.cc
+++ b/winsup/cygwin/net.cc
@@ -1819,6 +1819,7 @@ get_adapters_addresses (PIP_ADAPTER_ADDRESSES *pa_ret, 
ULONG family)
 The OS allocates stacks bottom up, so chances are good that the new
 stack will be located in the lower address area. */
   HANDLE thr = CreateThread (NULL, 0, call_gaa, , 0, NULL);
+  SetThreadName (GetThreadId (thr), "__call_gaa");
   if (!thr)
{
  debug_printf ("CreateThread: %E");
diff --git a/winsup/cygwin/thread.cc b/winsup/cygwin/thread.cc
index e8622f9..7084657 100644
--- a/winsup/cygwin/thread.cc
+++ b/winsup/cygwin/thread.cc
@@ -1992,6 +1992,9 @@ pthread::thread_init_wrapper (void *arg)
   _my_tls.sigmask = 

RE: Trouble installing to UNC path and/or mapped network drive

[Andreas Wettergren]

> -Original Message-
> From: Andrey Repin [mailto:anrdae...@yandex.ru]
> Sent: Tuesday, August 23, 2016 12:01 PM
> To: Andreas Wettergren ;
> cygwin@cygwin.com
> Subject: Re: Trouble installing to UNC path and/or mapped network drive
> 
> Greetings, Andreas Wettergren!
> 
> > Hi,
> > We have a shared Cygwin environment installed to a Windows network
> > share at my workplace, and I have been assigned the task to update it
> > (along with installing a Perl program I wrote + dependencies).
> 
> I'll just throw in a note that the way Cygwin operates discourages any
> "shared" setup scenario.
> If you want a stable set of packages, your only real solution is a local 
> package
> mirror.
> For more information, look up the "rebase" in Cygwin documentation.
> 
> 
> --
> With best regards,
> Andrey Repin
> Tuesday, August 23, 2016 12:58:57
> 
> Sorry for my terrible english...

Thanks for your reply.

Noted, but maybe we are talking about different types of sharing?
In our scenario we are actually using the Cygwin installation as a shared 
multi-user system,
we are not sharing the just the installation packages for multiple 
installations like a mirror.

It's a bit wild if you ask me :) 
If I didn't use it every day I wouldn't think that sort of setup would work in 
practice.
There are some valid reasons it has been set up this way, but it does come with 
some special headaches as well...

I'll look into rebase anyway, thanks for the tip.

(BTW, my original issue has been resolved, for details see my reply to Erik)

Best regards

Andreas Wettergren
Translation Support
Office: +46 31-360 98 30   

Re: Wrong file permissions

[Erik Soderquist]

On Tue, Aug 23, 2016 at 10:23 AM, Eliot Moss  wrote:
> Thank you, Erik -- you answered better than I could have!  I don't
> use a domain login, so the fact that my local SID is different has
> been very plain to me!

Multiple years as a network admin managing Windows/Linux/UNIX hybrid
environments both with and without Windows AD involved has lead me to
many many interesting twists in how the SIDs are handled... I
particularly ran into the machine specific SIDs when ghosting client
machines for faster reload and deployment... You can imagine the fun I
had sorting out why adding a freshly ghosted machine to the domain
caused things to break on another machine...

-- Erik

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

RE: Trouble installing to UNC path and/or mapped network drive

[Andreas Wettergren]

Thanks for your reply, the problem has been resolved, details below and in my 
previous reply to Erik.

> -Original Message-
> From: cygwin-ow...@cygwin.com [mailto:cygwin-ow...@cygwin.com] On
> Behalf Of Achim Gratz
> Sent: Monday, August 22, 2016 7:37 PM
> To: cygwin@cygwin.com
> Subject: Re: Trouble installing to UNC path and/or mapped network drive
> 
> Andreas Wettergren writes:
> > I’m running this as a network user that has full read & write access
> > to the target directory and subdirectories.  This user is also a
> > member of the local administrator group.
> […]
> > I’ve tried to run the setup to an empty network path (no prior cygwin
> > files, just an empty directory) as well, with identical results.
> > Nothing gets written to this directory.
> 
> I think the problem is either the local administrator group (sometimes local
> administrators are expressedly forbidden to do anything on the
> network) or your network user cannot change the DACL on the network
> share (another common setup in large organizations).
> 

Yes, it was something like this (I haven't pinpointed the exact cause yet),
running the setup as the network administrator seem to have done the trick.

> > I’ve also tried to run the install to a local drive on my computer.
> > This works as expected, no problems.
> 
> You could try to move / copy that installation to the network share, then.
> 

I had this in mind as something to try, but I wasn't sure if anything in the 
installation might get incorrect paths or similar.

> > Finally, this sort of install routine has apparently worked in the
> > past, my colleague tells me this is the way he has updated our Cygwin
> > environment before.  It’s possible he did something slightly
> > differently to what I have done this time though.
> 
> Or meanwhile some new security policy got installed that throws a spanner in
> the works.  If your colleague is still able to update Cygwin, you'll know it's
> some sort of access right you're missing and can work it out from there.
> 

Correct, it was some sort of permissions issue, possibly related to underlying 
Samba problems (see my reply to Erik for details).
I think this Cygwin account I was using was created for a different purpose
(updating a specific in-house program inside Cygwin), and the full install 
scenario had not been tested before.

Thanks for your suggestions, much appreciated!


Andreas Wettergren
Translation Support
Amesto Translations




> 
> Regards,
> Achim.
> --
> +<[Q+ Matrix-12 WAVE#46+305 Neuron microQkb Andromeda XTk
> Blofeld]>+
> 
> Samples for the Waldorf Blofeld:
> http://Synth.Stromeko.net/Downloads.html#BlofeldSamplesExtra
> 
> --
> Problem reports:   http://cygwin.com/problems.html
> FAQ:   http://cygwin.com/faq/
> Documentation: http://cygwin.com/docs.html
> Unsubscribe info:  http://cygwin.com/ml/#unsubscribe-simple

Re: Wrong file permissions

[Eliot Moss]

On 8/23/2016 9:41 AM, Erik Soderquist wrote:

On Mon, Aug 22, 2016 at 3:59 AM, Björn Kautler  wrote:



Your domain user SID will remain the same, however, all of the local
machine's domain SIDs will be different as it is a different host on
the domain, so I would expect permissions problems for any/all machine
specific installations when doing a raw perms copy.




I would imagine all of the *files* are present in the cygwin root, but
I've never checked, and a simple file copy, even if the adjustments
are made to result in the correct permissions, will not get things
like registry entries.  (I don't know how many there are, but do know
I've seen some).


Thank you, Erik -- you answered better than I could have!  I don't
use a domain login, so the fact that my local SID is different has
been very plain to me!

Regards -- EM

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

Re: Trouble installing to UNC path and/or mapped network drive

[Erik Soderquist]

On Tue, Aug 23, 2016 at 10:06 AM, Andreas Wettergren  wrote:
>> Why 32 bit rather than 64 bit in a 64 bit environment?  (more curious than
>> anything else, I don't think this affects the outcome).
>>
>
> This is partially for historical reasons, and partially because I'm 
> personally not
> 100% sure every client that uses the environment is 64-bit yet.

Valid reasons  :)

>> > To work around this, I mapped a network “drive” in Windows Explorer.
>> > The drive is mapped so that my target directory is Z:\Cygwin.
>>
>> Did you set the drive mapping to be remembered on future logins?  If not, a
>> new instance after privilege elevation may also not have the drive mapping.
>> (Still might not, I've had UAC/network drive mapping problems with other
>> programs as well).  Is disabling UAC and running entirely as local admin to 
>> test
>> an option?  For that matter, is installinng on the host the installation is 
>> local to
>> an option?
>
> Good points.
> I got some help from my colleague in the end, and it turns out he ran the 
> install as network administrator after all.
> In hindsight I should have tried this, but his instructions specifically 
> mentioned this "Cygwin" user, so I assumed they were accurate.

Assumptions can be dangerous  ;)

> A detail I forgot to mention is that the network share is not really a 
> Windows share, it's actually on a Samba server.
> (Sorry, I should have mentioned this, but I forgot).

That adds additional twists... I have seen Samba configured (both
accidentally and intentionally) to report "full control" when queried
for what permissions a user has and still have everything locked as
read only for most users.  Windows can now do this from AD Group
Policies as well, but is rarely configured to do so.

> Although at first glance it seems the Cygwin user I was using looks like it 
> has the correct permissions,
> something was obviously missing or incorrectly mapped, running as the network 
> admin worked fine.
> I suspect there might be some underlying issue on the Linux file system level 
> or it may be a Samba permissions mapping issue
> for the Cygwin user.
>
> UAC should not have been an issue, the account was in the local admins group,
> but having the mapping be remembered would make sense anyhow, I agree.
> (And running as the actual local administrator account may perhaps make some 
> minute difference that the group membership
> doesn't bring.)

Actually, unless the cygwin user was a domain account rather than a
local machine account, I would be surprised if it was not receiving
the ghost "full control" I mentioned above.

>> > The symptoms begin with an error saying something like “Unable to extract
>> /etc/ -- the file is in use.” plus a suspicion about a corrupt package.
>> > Skipping past this error gives an almost identical error but about
>> > “/etc/postinstall”, and skipping that gives another about
>> “/etc/postinstall/000-cygwin-post-install.sh”.
>>
>> When was the most recent filesystems check on the host with the actual
>> files?  A corrupt directory tree can cause similar issues.
>>
>
> Will do. Even if my immediate problem is solved, I think you may be right,
> there seems to be some lingering permissions issue that caused my confusion.
>
> Thanks again, even if my problem turned out to really be 
> account/permissions-related,
> your answers might give some useful hints for others with similar problems.



You are very welcome  :)

-- Erik

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

[newlib-cygwin] Add pthread_getname_np and pthread_setname_np

[Jon TURNEY]

https://sourceware.org/git/gitweb.cgi?p=newlib-cygwin.git;h=fdb7df230ded019be0472565d79de14fe677dea7

commit fdb7df230ded019be0472565d79de14fe677dea7
Author: Jon Turney 
Date:   Mon Aug 22 18:43:05 2016 +0100

Add pthread_getname_np and pthread_setname_np

This patch adds pthread_getname_np and pthread_setname_np.

These were added to glibc in 2.12[1] and are also present in some form on
NetBSD and several UNIXes.

The code is based on NetBSD's implementation with changes to better match
Linux behaviour.

Implementation quirks:

* pthread_setname_np with a NULL pointer segfaults (as linux)

* pthread_setname_np returns ERANGE for names longer than 16 characters (as
linux)

* pthread_getname_np with a NULL pointer returns EFAULT (as linux)

* pthread_getname_np with a buffer length of less than 16 returns ERANGE (as
linux)

* pthread_getname_np truncates the thread name to fit the buffer length.
This guarantees success even when the default thread name is longer than 16
characters, but means there is no way to discover the actual length of the
thread name. (Linux always truncates the thread name to 16 characters)

* Changing program_invocation_short_name changes the default thread name (on
linux, it has no effect on the default thread name)

I'll leave it up to you to decide if any of these matter.

This is implemented via class pthread_attr to make it easier to add
pthread_attr_[gs]etname_np (present in NetBSD and some UNIXes) should it
ever be added to Linux (or we decide we want it anyway).

[1] https://sourceware.org/git/?p=glibc.git;a=blob;f=NEWS

Diff:
---
 winsup/cygwin/common.din   |  2 ++
 winsup/cygwin/include/cygwin/version.h |  3 +-
 winsup/cygwin/include/pthread.h|  2 ++
 winsup/cygwin/release/2.6.0|  1 +
 winsup/cygwin/thread.cc| 65 +-
 winsup/cygwin/thread.h |  1 +
 winsup/doc/new-features.xml|  4 +++
 winsup/doc/posix.xml   |  2 ++
 8 files changed, 78 insertions(+), 2 deletions(-)

diff --git a/winsup/cygwin/common.din b/winsup/cygwin/common.din
index d54b70a..8f7a282 100644
--- a/winsup/cygwin/common.din
+++ b/winsup/cygwin/common.din
@@ -1046,6 +1046,7 @@ pthread_exit SIGFE
 pthread_getattr_np SIGFE
 pthread_getconcurrency SIGFE
 pthread_getcpuclockid SIGFE
+pthread_getname_np SIGFE
 pthread_getschedparam SIGFE
 pthread_getsequence_np SIGFE
 pthread_getspecific SIGFE
@@ -1086,6 +1087,7 @@ pthread_self SIGFE
 pthread_setcancelstate SIGFE
 pthread_setcanceltype SIGFE
 pthread_setconcurrency SIGFE
+pthread_setname_np SIGFE
 pthread_setschedparam SIGFE
 pthread_setschedprio SIGFE
 pthread_setspecific SIGFE
diff --git a/winsup/cygwin/include/cygwin/version.h 
b/winsup/cygwin/include/cygwin/version.h
index 2782c32..a1b8a62 100644
--- a/winsup/cygwin/include/cygwin/version.h
+++ b/winsup/cygwin/include/cygwin/version.h
@@ -467,12 +467,13 @@ details. */
strtoull_l, wcstod_l, wcstof_l, wcstol_l, wcstold_l, wcstoll_l,
wcstoul_l, wcstoull_l.
   302: Export nl_langinfo_l.
+  303: Export pthread_getname_np, pthread_setname_np.
 
   Note that we forgot to bump the api for ualarm, strtoll, strtoull,
   sigaltstack, sethostname. */
 
 #define CYGWIN_VERSION_API_MAJOR 0
-#define CYGWIN_VERSION_API_MINOR 302
+#define CYGWIN_VERSION_API_MINOR 303
 
 /* There is also a compatibity version number associated with the shared memory
regions.  It is incremented when incompatible changes are made to the shared
diff --git a/winsup/cygwin/include/pthread.h b/winsup/cygwin/include/pthread.h
index 609eac2..47ee6bd 100644
--- a/winsup/cygwin/include/pthread.h
+++ b/winsup/cygwin/include/pthread.h
@@ -222,6 +222,8 @@ void pthread_testcancel (void);
 
 #if __GNU_VISIBLE
 int pthread_getattr_np (pthread_t, pthread_attr_t *);
+int pthread_getname_np (pthread_t, char *, size_t) __attribute__((nonnull(2)));
+int pthread_setname_np (pthread_t, const char *) __attribute__((nonnull(2)));
 int pthread_sigqueue (pthread_t *, int, const union sigval);
 int pthread_yield (void);
 #endif
diff --git a/winsup/cygwin/release/2.6.0 b/winsup/cygwin/release/2.6.0
index 3f64577..8255450 100644
--- a/winsup/cygwin/release/2.6.0
+++ b/winsup/cygwin/release/2.6.0
@@ -20,6 +20,7 @@ What's new:
 - locale(1) now supports a -i/--input option to fetch the current input
   locale (this is basically equivalent to the current keyboard layout setting).
 
+- New API: pthread_getname_np, pthread_setname_np.
 
 What changed:
 -
diff --git a/winsup/cygwin/thread.cc b/winsup/cygwin/thread.cc
index 4414785..e8622f9 100644
--- a/winsup/cygwin/thread.cc
+++ b/winsup/cygwin/thread.cc
@@ -1106,7 +1106,7 @@ pthread::resume ()
 pthread_attr::pthread_attr ():verifyable_object (PTHREAD_ATTR_MAGIC),
 joinable (PTHREAD_CREATE_JOINABLE), 

RE: Trouble installing to UNC path and/or mapped network drive

[Andreas Wettergren]

Thank you very much for the reply, my issue has been resolved, details in my 
answers below.

> -Original Message-
> From: cygwin-ow...@cygwin.com [mailto:cygwin-ow...@cygwin.com] On
> Behalf Of Erik Soderquist
> Sent: Monday, August 22, 2016 6:04 PM
> To: cygwin@cygwin.com
> Subject: Re: Trouble installing to UNC path and/or mapped network drive
> 
> On Mon, Aug 22, 2016 at 5:57 AM, Andreas Wettergren wrote:
> > The computer I’m using to launch the setup from is running Windows 7
> Professsional (64 - bit).
> > I’m using setup-x86.exe version 2.874.
> 
> Why 32 bit rather than 64 bit in a 64 bit environment?  (more curious than
> anything else, I don't think this affects the outcome).
> 

This is partially for historical reasons, and partially because I'm personally 
not 
100% sure every client that uses the environment is 64-bit yet.

> > Antivirus (Webroot SecureAnywhere) is disabled, but not completely
> unloaded, during setup.
> > (I have permissions to disable Webroot via a control applet, but I
> > don’t think I’m allowed to unload it completely)
> 
> Do you have the option of creating a couple virtual machines to test with
> virgin Windows installs?  That could relatively quickly determine if it is
> something in the exiting environment...
> 
> > First I tried using an UNC path, but this does not work at all, the
> > setup program complains that the UNC path is not an absolute path.
> > (Should this be considered a bug?  If so I will rerun the setup, get
> > the exact error message and report it separately.)
> 
> I believe this is expected, but will leave the final determination to others 
> with
> greater knowledge than I.
> 

Yep, I suspect network installs aren't really a supported scenario anyway, so 
filing a bug might not be useful.

> > To work around this, I mapped a network “drive” in Windows Explorer.
> > The drive is mapped so that my target directory is Z:\Cygwin.
> 
> Did you set the drive mapping to be remembered on future logins?  If not, a
> new instance after privilege elevation may also not have the drive mapping.
> (Still might not, I've had UAC/network drive mapping problems with other
> programs as well).  Is disabling UAC and running entirely as local admin to 
> test
> an option?  For that matter, is installinng on the host the installation is 
> local to
> an option?
> 

Good points. 
I got some help from my colleague in the end, and it turns out he ran the 
install as network administrator after all.
In hindsight I should have tried this, but his instructions specifically 
mentioned this "Cygwin" user, so I assumed they were accurate.

A detail I forgot to mention is that the network share is not really a Windows 
share, it's actually on a Samba server.
(Sorry, I should have mentioned this, but I forgot).
Although at first glance it seems the Cygwin user I was using looks like it has 
the correct permissions,
something was obviously missing or incorrectly mapped, running as the network 
admin worked fine.
I suspect there might be some underlying issue on the Linux file system level 
or it may be a Samba permissions mapping issue
for the Cygwin user.

UAC should not have been an issue, the account was in the local admins group, 
but having the mapping be remembered would make sense anyhow, I agree.
(And running as the actual local administrator account may perhaps make some 
minute difference that the group membership 
doesn't bring.)

> > The symptoms begin with an error saying something like “Unable to extract
> /etc/ -- the file is in use.” plus a suspicion about a corrupt package.
> > Skipping past this error gives an almost identical error but about
> > “/etc/postinstall”, and skipping that gives another about
> “/etc/postinstall/000-cygwin-post-install.sh”.
> 
> When was the most recent filesystems check on the host with the actual
> files?  A corrupt directory tree can cause similar issues.
> 

Will do. Even if my immediate problem is solved, I think you may be right,
there seems to be some lingering permissions issue that caused my confusion.

Thanks again, even if my problem turned out to really be 
account/permissions-related,
your answers might give some useful hints for others with similar problems.

Best regards

Andreas Wettergren
Translation Support
Amesto Translations



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

Re: Wrong file permissions

[Erik Soderquist]

On Mon, Aug 22, 2016 at 3:59 AM, Björn Kautler  wrote:
> Hi Eliot,
>
> thanks for your answer.
>
> It seems this was an issue with the NTFS permissions.
> I also was not able to access the folders via Windows Explorer.
> After also fixing the Windows permissions it works now as expected so far.
>
> Do you really think I got a new SID on the new box when logging in
> with the same user of the same domain?
> Can I check this somehow?

Your domain user SID will remain the same, however, all of the local
machine's domain SIDs will be different as it is a different host on
the domain, so I would expect permissions problems for any/all machine
specific installations when doing a raw perms copy.

> Why should the setup be redone on a new box?
> I'd need to do redo all installations, configuration, and so on.

I usually do this via script.  Besides, after copying, unless you
first do a full update on your source every time you'd still need to
update to current for the latest fixes.

> Besides ths permsissions issue I had, why shouldn't I just copy over
> the whole cygwin root directory?
> Everything cygwin related should be in there, isn't it?

I would imagine all of the *files* are present in the cygwin root, but
I've never checked, and a simple file copy, even if the adjustments
are made to result in the correct permissions, will not get things
like registry entries.  (I don't know how many there are, but do know
I've seen some).

-- Erik


ps:  this is a "no top posting" email list; please put your comments
and responses inline with what your are responding to so you do not
disrupt the normal flow of reading.

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

Non-Privileged Execution of cygrunsrv/cygserver

[Brian Inglis]

Hi folks,

Does anyone know how to start cygrunsrv/cygserver from a non-elevated shell?
Starting them in an elevated system startup scheduled task works, as does
shutting them down from a non-elevated shell before running setup.

The problem is that neither the scheduled task nor service startup can be
run from a non-elevated shell.
This used to be doable in a profile script on Windows 7 with earlier Cygwin,
but no longer works after Win 10 and/or security and/or Cygwin updates.

Alternatively (main use case) could setup-x86{,-64} be changed to shut down
cygrunsrv/cygserver if they are running when it starts, then restart them
if setup completes successfully, and they were running when setup started?

--
Take care. Thanks, Brian Inglis, Calgary, Alberta, Canada

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

Re: [ANNOUNCEMENT] TEST RELEASE: Cygwin 2.6.0-0.11

[Brian Inglis]

On 2016-08-20 15:37, Corinna Vinschen wrote:

I uploaded a new Cygwin test release 2.6.0-0.11.
(Still) two interesting changes in this test release:
- Fix the bug reported in
  https://cygwin.com/ml/cygwin/2016-08/msg00357.html
- Add missing nl_langinfo_l functions as reported in
  https://cygwin.com/ml/cygwin/2016-08/msg00349.html
The 2.6.0 release is going to introducing the locale_t datatype, as well
as all functions related to locale_t locales and per-thread locales per
POSIX-1.2008.
So, rather than just providing a single, per-process locale, you can now
create new locales ("newlocale") and set it as locale for the current
thread ("uselocale") or use it directly with one of the new functions
taking a locale_t as parameter (i.e. isalpha_l).
The full list of new interfaces is:
- POSIX:
newlocale, freelocale, duplocale, uselocale, nl_langinfo_l
isalnum_l, isalpha_l, isblank_l, iscntrl_l, isdigit_l, isgraph_l,
islower_l, isprint_l, ispunct_l, isspace_l, isupper_l,
iswalnum_l, iswalpha_l, iswblank_l, iswcntrl_l, iswctype_l, iswdigit_l,
iswgraph_l, iswlower_l, iswprint_l, iswpunct_l, iswspace_l, iswupper_l,
iswxdigit_l, isxdigit_l
tolower_l, toupper_l, towctrans_l, towlower_l, towupper_l,
wctrans_l, wctype_l
strcasecmp_l, strcoll_l, strncasecmp_l, strxfrm_l
wcscasecmp_l, wcscoll_l, wcstrncasecmp_l, wcstrxfrm_l
strfmon_l, strftime_l
- GNU extensions:
isascii_l, toascii_l
strtol_l, stroll_l, strtoul_l, strtoull_l
strtod_l, strtofl, strtold_l
wcstol_l, wcsoll_l, wcstoul_l, wcstoull_l
wcstod_l, wcstofl, wcstold_l


Hi folks,

Compared lists of locale_t headers and functions for POSIX, Cygwin,
and glibc, attached below for comparison, and found:

* missed string.h(strerror_l) on my first check;
   not sure if you can implement that easily on Windows?

* GNU also supports wchar.h(wcsftime_l) and time.h(strptime_l);

* GNU also defines string.h(str[n]casecmp_l) functions as an extension,
   as well as in POSIX specified strings.h.

--
Take care. Thanks, Brian Inglis, Calgary, Alberta, Canada

POSIX   CYGWIN  GLIBC
headers ctype.h
functions   isalnum_l   isalnum_l   isalnum_l
functions   isalpha_l   isalpha_l   isalpha_l
functions   isblank_l   isblank_l   isblank_l
functions   iscntrl_l   iscntrl_l   iscntrl_l
functions   isdigit_l   isdigit_l   isdigit_l
functions   isgraph_l   isgraph_l   isgraph_l
functions   islower_l   islower_l   islower_l
functions   isprint_l   isprint_l   isprint_l
functions   ispunct_l   ispunct_l   ispunct_l
functions   isspace_l   isspace_l   isspace_l
functions   isupper_l   isupper_l   isupper_l
functions   isxdigit_l  isxdigit_l  isxdigit_l
functions   tolower_l   tolower_l   tolower_l
functions   toupper_l   toupper_l   toupper_l
functions   toascii_l   toascii_l
functions   isascii_l   isascii_l
headers wctype.h
functions   iswalnum_l  iswalnum_l  iswalnum_l
functions   iswalpha_l  iswalpha_l  iswalpha_l
functions   iswblank_l  iswblank_l  iswblank_l
functions   iswcntrl_l  iswcntrl_l  iswcntrl_l
functions   iswctype_l  iswctype_l  iswctype_l
functions   iswdigit_l  iswdigit_l  iswdigit_l
functions   iswgraph_l  iswgraph_l  iswgraph_l
functions   iswlower_l  iswlower_l  iswlower_l
functions   iswprint_l  iswprint_l  iswprint_l
functions   iswpunct_l  iswpunct_l  iswpunct_l
functions   iswspace_l  iswspace_l  iswspace_l
functions   iswupper_l  iswupper_l  iswupper_l
functions   iswxdigit_l iswxdigit_l iswxdigit_l
functions   towctrans_l towctrans_l towctrans_l
functions   towlower_l  towlower_l  towlower_l
functions   towupper_l  towupper_l  towupper_l
functions   wctrans_l   wctrans_l   wctrans_l
functions   wctype_lwctype_lwctype_l
headers locale.h
functions   duplocale   duplocale   duplocale
functions   freelocale  freelocale  freelocale
functions   newlocale   newlocale   newlocale
functions   uselocale   uselocale   uselocale
headers wchar.h
functions   wcscasecmp_lwcscasecmp_lwcscasecmp_l
functions   wcscoll_l   wcscoll_l   wcscoll_l
functions   wcsncasecmp_l   wcsncasecmp_l   wcsncasecmp_l
functions 

Re: building libgeos [solved]

[Felipe Vieira]

Hey guys,

after a couple more trials was able to solve the issue. I am not sure
exactly what was the problem but I'll be reporting the full
workaround.

Neither I am certain that the libgeos shipped with cygwin has an
issue; it looks like it was a matplotlib basemap issue instead.

Anyways here is the full process to get matplotlib basemap on cygwin:

1) <> git clone https://github.com/matplotlib/basemap .
Instead use their releases page:
https://github.com/matplotlib/basemap/releases . Follow the procedure
indicated in their github home page.

2) After the "> ./configure --prefix=$GEOS_DIR" but before the " >
make; make install" edit the file include/geos/platform.h on line 87
from:

'''
#if defined(HAVE_ISNAN)
# define ISNAN(x) (isnan(x))
#else
# if defined(_MSC_VER)
#  define ISNAN(x) _isnan(x)
# elif defined(__MINGW32__)
// sandro furieri: sanitizing MinGW32
#  define ISNAN(x) (std::isnan(x))
# elif defined(__OSX__) || defined(__APPLE__)
   // Hack for OS/X  incorrectly re-defining isnan() into oblivion.
   // It does leave a version in std.
#  define ISNAN(x) (std::isnan(x))
# elif defined(__sun) || defined(__sun__)
#  include 
#  define ISNAN(x) (::isnan(x))
# endif
#endif
'''
to
'''
# define ISNAN(x) (std::isnan(x))
# include 
# include 
'''
now proceed to the " > make; make install" and continue the procedure
as informed.

Thanks for all the help!

Best regards,

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

Re: [ANNOUNCEMENT] Updated: sqlite3-3.14.1-1 for Cygwin/Cygwin64

[Keith Christian]

Thank you for the update, Jan!

On Tue, Aug 23, 2016 at 3:31 AM, Jan Nijtmans  wrote:
> SQLite is a software library that implements a self-contained,
> serverless, zero-configuration, transactional SQL database engine
>
> Changes since 3.13.0-1
> =
> * Update to upstream SQLite 3.14.1, also known as the π release.
> 
> * Maximum default page length enhanced from 8192 to 32768
> * Maximum Number Of Attached Databases enhanced from 10 to 62
> 
> Additional SQLite bug-fixes, rejected (apparently) upstream
> but important (and simple) enough for Cygwin:
>   * VFS filename truncation issues
> 
>   * Wrong filename handling in sqlite3_load_extension() for Cygwin
> 
>   * pragma database_list returns win32 paths on Cygwin
> 
>   * ISO time leap second
> 
>   * update to Unicode 9.0 for FTS3/4/5 tokenizer
>
> --
> Jan Nijtmans
>
> --
> Problem reports:   http://cygwin.com/problems.html
> FAQ:   http://cygwin.com/faq/
> Documentation: http://cygwin.com/docs.html
> Unsubscribe info:  http://cygwin.com/ml/#unsubscribe-simple
>
> --
> Problem reports:   http://cygwin.com/problems.html
> FAQ:   http://cygwin.com/faq/
> Documentation: http://cygwin.com/docs.html
> Unsubscribe info:  http://cygwin.com/ml/#unsubscribe-simple
>

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

[newlib-cygwin] Reference __global_locale only via __get_global_locale.

[Corinna Vinschen]

https://sourceware.org/git/gitweb.cgi?p=newlib-cygwin.git;h=2fb5e3dfb2603988640c87d69596f470900e1ff5

commit 2fb5e3dfb2603988640c87d69596f470900e1ff5
Author: Corinna Vinschen 
Date:   Tue Aug 23 12:38:28 2016 +0200

Reference __global_locale only via __get_global_locale.

Signed-off-by: Corinna Vinschen 

Diff:
---
 newlib/libc/locale/setlocale.h | 6 +++---
 winsup/cygwin/nlsfuncs.cc  | 6 +++---
 2 files changed, 6 insertions(+), 6 deletions(-)

diff --git a/newlib/libc/locale/setlocale.h b/newlib/libc/locale/setlocale.h
index 99bd466..5bf9edc 100644
--- a/newlib/libc/locale/setlocale.h
+++ b/newlib/libc/locale/setlocale.h
@@ -195,8 +195,6 @@ struct __locale_t
 };
 
 extern const struct __locale_t __C_locale;
-extern struct __locale_t __global_locale;
-
 extern char *__loadlocale (struct __locale_t *, int, const char *);
 extern const char *__get_locale_env(struct _reent *, int);
 
@@ -210,6 +208,7 @@ extern size_t _wcsnrtombs_l (struct _reent *, char *, const 
wchar_t **,
 _ELIDABLE_INLINE struct __locale_t *
 __get_global_locale ()
 {
+  extern struct __locale_t __global_locale;
   return &__global_locale;
 }
 
@@ -227,7 +226,8 @@ __get_locale_r (struct _reent *r)
 _ELIDABLE_INLINE struct __locale_t *
 __get_current_locale ()
 {
-  return _REENT->_locale ?: &__global_locale;
+  return _REENT->_locale ?: __get_global_locale ();
+}
 }
 
 #ifdef __CYGWIN__
diff --git a/winsup/cygwin/nlsfuncs.cc b/winsup/cygwin/nlsfuncs.cc
index 3e3b5f4..60ad2d3 100644
--- a/winsup/cygwin/nlsfuncs.cc
+++ b/winsup/cygwin/nlsfuncs.cc
@@ -1524,11 +1524,11 @@ internal_setlocale ()
   wchar_t *w_path = NULL, *w_cwd;
 
   /* Don't do anything if the charset hasn't actually changed. */
-  if (cygheap->locale.mbtowc == __global_locale.mbtowc)
+  if (cygheap->locale.mbtowc == __get_global_locale ()->mbtowc)
 return;
 
   debug_printf ("Global charset set to %s",
-   __locale_charset (&__global_locale));
+   __locale_charset (__get_global_locale ()));
   /* Fetch PATH and CWD and convert to wchar_t in previous charset. */
   path = getenv ("PATH");
   if (path && *path)   /* $PATH can be potentially unset. */
@@ -1541,7 +1541,7 @@ internal_setlocale ()
   sys_cp_mbstowcs (cygheap->locale.mbtowc, w_cwd, 32768,
   cygheap->cwd.get_posix ());
   /* Set charset for internal conversion functions. */
-  cygheap->locale.mbtowc = __global_locale.mbtowc;
+  cygheap->locale.mbtowc = __get_global_locale ()->mbtowc;
   if (cygheap->locale.mbtowc == __ascii_mbtowc)
 cygheap->locale.mbtowc = __utf8_mbtowc;
   /* Restore CWD and PATH in new charset. */

Re: CALM: error while merging uploaded packages for db (obsoleting db4.8)

[Jon Turney]

On 23/08/2016 07:57, Dr. Volker Zell wrote:

Achim Gratz writes:


> Dr. Volker Zell writes:
>> Could somebody please check out what's wrong with my latest "db" upload
>> for x86_64 and x86 and the "db4.8" upload for x86 (should be obsoleted
>> by "db"). It somehow seems to be stuck in my staging area without
>> getting moved to the release area.

> Did you place a !ready file?

I thought it's not needed when doing a cygport upload. At least that's
the case with my other uploads so far.


cygport upload creates the !ready file for you.

I'm afraid calm isn't very intelligent and removes all the !ready files 
after an error has occurred, so you need to replace it to make it try to 
process the uploads again.


Anyhow, on to today's error:


ERROR: package 'db4.8' version '4.8.30-2' source has no non-empty install 
tarfiles
ERROR: error while validating merged x86 packages for Dr. Volker Zell


So, this error is caused by the fact that db4.8-4.8.30-2 source pacakge 
exists, and is non-empty, but the db4.8, db4.8-doc and libdb4.8-devel 
packages it creates are all empty obsoletion packages.


This is unusual, as in all existing cases, the superseding package (in 
this case 'db' is replacing 'db4.8') source would create the obsoletion 
packages.


I've moved db4.8-4.8.30-2-src.tar.xz aside and reset !ready to allow the 
upload to proceed.


(This is disallowed because we had a few cases where source for removed 
packages was being retained in the distro only because it was referenced 
by an obsoletion package.  This seems to be the case here with the 
db-4.8.30 source.)


I'll see if I can find a better way to handle these cases in the future.

Sorry for the inconvenience.

Re: issue report

[Andrey Repin]

Greetings, Stanislav Bivolarski!

> C:\wamp\www\Leanplum-Marketing>grunt install
> (node) sys is deprecated. Use util instead.
> Running "exec:seedDB" (exec) task
>   0 [main] sh 728 find_fast_cwd: WARNING: Couldn't compute
> FAST_CWD pointer.  Please report this problem to
> the public mailing list cygwin@cygwin.com

Please do not report issues in versions of Cygwin years, if not decades, old.
Only the current version is supported.

>>> 0 [main] sh 728 find_fast_cwd: WARNING: Couldn't compute FAST_CWD pointer.  
>>> Please report this problem to
>>> the public mailing list cygwin@cygwin.com
>>> tools/db_create.sh: line 3: $'\r': command not found
> : No such file or directory 4: tools/config.sh
>>> Exited with code: 1.
> Warning: Task "exec:seedDB" failed. Use --force to continue.

> Aborted due to warnings.

> --
> Problem reports:   http://cygwin.com/problems.html
> FAQ:   http://cygwin.com/faq/
> Documentation: http://cygwin.com/docs.html

Also this ^

P.S.
And as a general rule, "paste error text in google first".


-- 
With best regards,
Andrey Repin
Tuesday, August 23, 2016 13:01:51

Sorry for my terrible english...


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

Re: Trouble installing to UNC path and/or mapped network drive

[Andrey Repin]

Greetings, Andreas Wettergren!

> Hi,
> We have a shared Cygwin environment installed to a Windows network share at
> my workplace, and I have been assigned the task to update it (along with
> installing a Perl program I wrote + dependencies).

I'll just throw in a note that the way Cygwin operates discourages any
"shared" setup scenario.
If you want a stable set of packages, your only real solution is a local
package mirror.
For more information, look up the "rebase" in Cygwin documentation.


-- 
With best regards,
Andrey Repin
Tuesday, August 23, 2016 12:58:57

Sorry for my terrible english...


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

[ANNOUNCEMENT] Updated: sqlite3-3.14.1-1 for Cygwin/Cygwin64

[Jan Nijtmans]

SQLite is a software library that implements a self-contained,
serverless, zero-configuration, transactional SQL database engine

Changes since 3.13.0-1
=
* Update to upstream SQLite 3.14.1, also known as the π release.

* Maximum default page length enhanced from 8192 to 32768
* Maximum Number Of Attached Databases enhanced from 10 to 62

Additional SQLite bug-fixes, rejected (apparently) upstream
but important (and simple) enough for Cygwin:
  * VFS filename truncation issues

  * Wrong filename handling in sqlite3_load_extension() for Cygwin

  * pragma database_list returns win32 paths on Cygwin

  * ISO time leap second

  * update to Unicode 9.0 for FTS3/4/5 tokenizer

--
Jan Nijtmans

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

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

Re: [ANNOUNCEMENT] Updated: mintty 2.4.3 and test release mintty 2.5.0

[Frank Fesevur]

2016-08-23 1:29 GMT+02:00 Thomas Wolff:
> I have also uploaded mintty 2.5.0 as a test release with the following
> change:
>
>   * Revise DPI handling (#470; #492, #487); always consider individual
> monitor DPI.
>
> Note that this release introduces a slightly incompatible change in mintty.
> The issue is that the interpretation of font size used to be suitable for
> traditional resolution monitors and did not consider higher-DPI monitors or
> the Windows monitor "zooming" feature which adjusts a virtual DPI.
> As a result, the handling of changed DPI (when moving the window to another
> monitor) introduced in 2.2.1 interfered with other aspects, the initial DPI
> was not considered, and a number of unpleasant side effects were
> occasionally observed.
>
> I intend to change DPI handling to be consistent with the respective monitor
> DPI (as optionally configured by "zooming factor"), and to comply with font
> size interpreation of other applications, e.g. notepad. In consequence the
> initial font size may be smaller (or larger) than before, depending on the
> actual monitor geometry and configuration. Nothing would change on a
> "standard" monitor configuration. To compensate, some people may have to
> change their font size configuration.

I was about to write a bug report that I encountered with a maximized
mintty on a 125% secondary screen. When I minimized it to the taskbar
it would restore to my primary screen. But that problem is fixed now,
so for me 2.5.0 solves that problem even before reporting it. The
draft mail is thrown away ;-)

Thanks for the fix!

Frank

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

[ANNOUNCEMENT] Updated: {t1lib/t1lib-x11/t1lib5/t1libx5/t1lib-devel/t1libx-devel}-5.1.2-13: Rasterizer library for Adobe Type 1 Fonts

[Dr . Volker Zell]

Hi

New versions of 't1lib/t1lib-x11/t1lib5/t1libx5/t1lib-devel/t1libx-devel' have 
been uploaded to a server near you.

 o Build for cygwin 2.5.2 with gcc-5.4.0
 o [x86] removed .la files from /usr/lib

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

Updated: {t1lib/t1lib-x11/t1lib5/t1libx5/t1lib-devel/t1libx-devel}-5.1.2-13: Rasterizer library for Adobe Type 1 Fonts

[Dr . Volker Zell]

Hi

New versions of 't1lib/t1lib-x11/t1lib5/t1libx5/t1lib-devel/t1libx-devel' have 
been uploaded to a server near you.

 o Build for cygwin 2.5.2 with gcc-5.4.0
 o [x86] removed .la files from /usr/lib

[ANNOUNCEMENT] Updated: mintty 2.4.3 and test release mintty 2.5.0

[Thomas Wolff]

I have uploaded mintty 2.4.3 with the following changes:

  * Fix trails when moving other window over mintty under certain 
conditions (#576).

  * Fix format of double child creation error message.

I have also uploaded mintty 2.5.0 as a test release with the following 
change:


  * Revise DPI handling (#470; #492, #487); always consider individual 
monitor DPI.


Note that this release introduces a slightly incompatible change in mintty.
The issue is that the interpretation of font size used to be suitable 
for traditional resolution monitors and did not consider higher-DPI 
monitors or the Windows monitor "zooming" feature which adjusts a 
virtual DPI.
As a result, the handling of changed DPI (when moving the window to 
another monitor) introduced in 2.2.1 interfered with other aspects, the 
initial DPI was not considered, and a number of unpleasant side effects 
were occasionally observed.


I intend to change DPI handling to be consistent with the respective 
monitor DPI (as optionally configured by "zooming factor"), and to 
comply with font size interpreation of other applications, e.g. notepad. 
In consequence the initial font size may be smaller (or larger) than 
before, depending on the actual monitor geometry and configuration. 
Nothing would change on a "standard" monitor configuration. To 
compensate, some people may have to change their font size configuration.


The homepage is at http://mintty.github.io/
It also links to the issue tracker.

--
Thomas

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

[ANNOUNCEMENT] Updated: {libtasn1_6/libtasn1-devel/libtasn1-doc}-4.9-1: ASN.1 library

[Dr . Volker Zell]

Hi

New versions of 'libtasn1_6/libtasn1-devel/libtasn1-doc' have been uploaded to 
a server near you.

 o Update to latest upstream release
 o Build for cygwin 2.5.2 with gcc-5.4.0
 o [x86] Removed the /usr/lib/libtasn1.la file


libtasn1 NEWS:
==
  
 o Fixes to OID encoding of OIDs which have elements which exceed 2^32

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

Updated: mintty 2.4.3 and test release mintty 2.5.0

[Thomas Wolff]

I have uploaded mintty 2.4.3 with the following changes:

  * Fix trails when moving other window over mintty under certain 
conditions (#576).

  * Fix format of double child creation error message.

I have also uploaded mintty 2.5.0 as a test release with the following 
change:


  * Revise DPI handling (#470; #492, #487); always consider individual 
monitor DPI.


Note that this release introduces a slightly incompatible change in mintty.
The issue is that the interpretation of font size used to be suitable 
for traditional resolution monitors and did not consider higher-DPI 
monitors or the Windows monitor "zooming" feature which adjusts a 
virtual DPI.
As a result, the handling of changed DPI (when moving the window to 
another monitor) introduced in 2.2.1 interfered with other aspects, the 
initial DPI was not considered, and a number of unpleasant side effects 
were occasionally observed.


I intend to change DPI handling to be consistent with the respective 
monitor DPI (as optionally configured by "zooming factor"), and to 
comply with font size interpreation of other applications, e.g. notepad. 
In consequence the initial font size may be smaller (or larger) than 
before, depending on the actual monitor geometry and configuration. 
Nothing would change on a "standard" monitor configuration. To 
compensate, some people may have to change their font size configuration.


The homepage is at http://mintty.github.io/
It also links to the issue tracker.

--
Thomas

Updated: {libtasn1_6/libtasn1-devel/libtasn1-doc}-4.9-1: ASN.1 library

[Dr . Volker Zell]

Hi

New versions of 'libtasn1_6/libtasn1-devel/libtasn1-doc' have been uploaded to 
a server near you.

 o Update to latest upstream release
 o Build for cygwin 2.5.2 with gcc-5.4.0
 o [x86] Removed the /usr/lib/libtasn1.la file


libtasn1 NEWS:
==
  
 o Fixes to OID encoding of OIDs which have elements which exceed 2^32

Re: [PATCH 1/2] Add pthread_getname_np and pthread_setname_np

[Corinna Vinschen]

Hi Jon,

On Aug 22 19:08, Jon Turney wrote:
> +#define NAMELEN 16

Please use a more descriptive name here, say  THR_NAMEMAX or so, and add
a "for Linux compat" comment.  With this change the patch is ok to apply.


Thanks,
Corinna

-- 
Corinna Vinschen  Please, send mails regarding Cygwin to
Cygwin Maintainer cygwin AT cygwin DOT com
Red Hat


signature.asc
Description: PGP signature

zsh bug: Incorrect (misleading) error message in zsh when using rm and drive letter is specified

[Ronald Fischer]

I observed this:

-0-1- ~/exp  > rm e:/media/*
zsh: sure you want to delete all the files in /home/FISRONA/exp/e:/media
[yn]? n

We see that zsh is catching the case that I want to rm all files in the
specified directory, and warns me about it (note that it is a zsh error
message, not a error message from rm), which is fine and expected
behaviour (which can be controlled by the zsh option RM_STAR_SILENT).
However, the path printed by zsh is wrong: It interprets e:/... as
relative path, while it is an absolute path. It seems that zsh is not
translating windows pathes using a drive letter to /cygdrive/... pathes
in this case.

Version information:
zsh 5.1.1 (x86_64-unknown-cygwin)
 
Ronald

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

Re: CALM: error while merging uploaded packages for db (obsoleting db4.8)

[Dr. Volker Zell]

> Achim Gratz writes:

> Dr. Volker Zell writes:
>> Could somebody please check out what's wrong with my latest "db" upload
>> for x86_64 and x86 and the "db4.8" upload for x86 (should be obsoleted
>> by "db"). It somehow seems to be stuck in my staging area without
>> getting moved to the release area.

> Did you place a !ready file?

I thought it's not needed when doing a cygport upload. At least that's
the case with my other uploads so far.

Ciao
  Volker