Supporting new Windows UDP offload features

2022-05-02 Thread Rony Ross
Hello,

Starting from Windows 10 version 2004, Windows supports UDP
segmentation offload and UDP receive coalescing as described in
https://docs.microsoft.com/en-us/windows-hardware/drivers/network/udp-segmentation-offload-uso-#ndis-interface-changes
and in 
https://docs.microsoft.com/en-us/windows/win32/winsock/ipproto-udp-socket-options
It seems that in order to support UDP segmentaion offload and UDP
receive coalescing, cygwin needs to add support for UDP_SEND_MSG_SIZE
and UDP_RECV_MAX_COALESCED_SIZE socket options for IPPROTO_UDP

Could you please add it? If so, could you please estimate when it will
be available?

Thanks you,
Rony

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


Re: [Attn: mercurial maintainer] hg fails with python3.9

2022-05-02 Thread Jari Aalto
On 2022-04-30 15:16, Jon Turney wrote:
> > My mistake. I had the development environment in wrong server, so only 5.7
> > was in there and I forgot to check the latest in Cygwin.
> > 
> > 6.1.1-1 uploaded
> 
> No problem.  I'll see if I can improve that error message sometime.
> 
> I notice that mercurial appears to be pure python, so maybe we can upload it
> as noarch in future?

Sure. I'll adjust it for the next upload.

Jari

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


Re: Supporting new Windows UDP offload features

2022-05-02 Thread Brian Inglis

On 2022-05-02 08:11, Rony Ross wrote:

Starting from Windows 10 version 2004, Windows supports UDP
segmentation offload and UDP receive coalescing as described in
https://docs.microsoft.com/en-us/windows-hardware/drivers/network/udp-segmentation-offload-uso-#ndis-interface-changes
and in
https://docs.microsoft.com/en-us/windows/win32/winsock/ipproto-udp-socket-options

It seems that in order to support UDP segmentaion offload and UDP
receive coalescing, cygwin needs to add support for UDP_SEND_MSG_SIZE
and UDP_RECV_MAX_COALESCED_SIZE socket options for IPPROTO_UDP

Could you please add it? If so, could you please estimate when it will
be available?


Cygwin is an all-volunteer non-commercial open source POSIX/Linux 
emulation project, typically driven by developers' needs and interests, 
sometimes offering patchsets to add or enable features previously 
available in BSD flavours.


If you are not a developer capable of doing so, you will have to wait 
and see if any developer is inspired by your post, and has the free time 
to work on it.


You might want to help this along by expanding on some description of 
use cases, impacts, advantages, benefits, downsides, Unix parameters and 
implementations to be supported for emulation, Unix and Windows 
configuration of features, and availability on NICs, of those network 
features.


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

This email may be disturbing to some readers as it contains
too much technical detail. Reader discretion is advised.
[Data in binary units and prefixes, physical quantities in SI.]

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


Re: Supporting new Windows UDP offload features

2022-05-02 Thread Corinna Vinschen
On May  2 17:11, Rony Ross wrote:
> Hello,
> 
> Starting from Windows 10 version 2004, Windows supports UDP
> segmentation offload and UDP receive coalescing as described in
> https://docs.microsoft.com/en-us/windows-hardware/drivers/network/udp-segmentation-offload-uso-#ndis-interface-changes
> and in 
> https://docs.microsoft.com/en-us/windows/win32/winsock/ipproto-udp-socket-options
> It seems that in order to support UDP segmentaion offload and UDP
> receive coalescing, cygwin needs to add support for UDP_SEND_MSG_SIZE
> and UDP_RECV_MAX_COALESCED_SIZE socket options for IPPROTO_UDP
> 
> Could you please add it?

Reading about those, it looks like they might be drop-in solutions
to implement the Linux UDP_SEGMENT and UDP_GRO socket options.
Knowing Windows however, I'm sure there's a twist somewhere...

> If so, could you please estimate when it will be available?

In the hopes they work the same way, I can make the UDP_SEGMENT and
UDP_GRO socket options available in the next major Cygwin version.

Availability, probably at some point late this year. Or later.
We're not at a point where this is safe to say.

However, test versions can be made available rather soon, I think.


Corinna

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


[ANNOUNCEMENT] Updated: {,mingw64-{i686,x86_64}-}binutils-2.38-1

2022-05-02 Thread Achim Gratz


The binutils packages for Cygwin and MingW64 cross-compilation
toolchains have been updated to the upstream feature release 2.38, with
additional patches applied to remove upstream changes that are
incompatible with Cygwin.

 binutils-2.38-1
 mingw64-i686-binutils-2.38-1
 mingw64-x86_64-binutils-2.38-1

The GNU Binutils are a collection of binary tools.  Most of these
programs use BFD, the Binary File Descriptor library, to do low-level
manipulation. Many of them also use the opcodes library to assemble and
disassemble machine instructions.

Note


This release still allows one to apparently use weak symbols (depending
on exactly how you test for them), which however don't actually work on
Windows nor Cygwin as they are unsupported by the loader.  These symbols
should always resolve as NULL, but since release 2.37 they sometimes
don't.

https://sourceware.org/bugzilla/show_bug.cgi?id=28719

After further changes and bugfixes in this area a workaround that was
identified earlier no longer has the desired effect and investigation
into the real root cause is still ongoing.

This change in behaviour triggered a latent bug in gnulib's detection of
weak symbols for pthreads that leads to non-debuggable crashes of
applications erroneously built with weak symbols. Either update gnulib
(or at least threadlib.m4) or add

export gl_cv_have_weak=no

to the cygport (or set it as an environment variable during configure).


-- 
  *** CYGWIN-ANNOUNCE UNSUBSCRIBE INFO ***

If you want to unsubscribe from the cygwin-announce mailing list, look
at the "List-Unsubscribe: " tag in the email header of this message.
Send email to the address specified there. It will be in the format:

cygwin-announce-unsubscribe-you=yourdomain@cygwin.com

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

http://sourceware.org/lists.html#unsubscribe-simple

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

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


[ANNOUNCEMENT] Updated: {,mingw64-{i686,x86_64}-}gcc-11.3.0-1

2022-05-02 Thread Achim Gratz
 
The native and mingw-w64 cross compilers have been updated for both
architectures to the latest upstream release version of the gcc-11
branch:

 gcc-11.3.0-1
 mingw64-i686-gcc-11.3.0-1
 mingw64-x86_64-gcc-11.3.0-1

This release includes libgccjit as a separate package for the native
toolchain on both architectures.  Since Cygwin can't use ASLR, any
nontrivial dynamic objects that get created in this way will likely need
to get rebased before they can be used (especially on 32bit).  It is
unlikely that build systems recognize the need for doing that at the
moment.

The support libraries and binutils have been rebuilt with the new
compiler.


-- 
  *** CYGWIN-ANNOUNCE UNSUBSCRIBE INFO ***

If you want to unsubscribe from the cygwin-announce mailing list, look
at the "List-Unsubscribe: " tag in the email header of this message.
Send email to the address specified there. It will be in the format:

cygwin-announce-unsubscribe-you=yourdomain@cygwin.com

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

http://sourceware.org/lists.html#unsubscribe-simple

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

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


[ANNOUNCEMENT] cygport 0.35.0-1

2022-05-02 Thread Jon Turney

The following packages have been uploaded to the Cygwin distribution:

* cygport-0.35.0-1

cygport is the standard method for building and maintaining
packages for the Cygwin distribution.

Highlights of this release:

* New subcommands 'vars' and 'srcpackage'
* cygtest() no longer ignores errors from 'make check' (making the 
default src_test function consistent with more specialized versions)
* Automatic dependency detection now generates a dependency on the 
perl5_0xy provide for packages which install files into a versioned 
vendor_perl directory or link with cygperl5_xy.dll.


Shortlog:

Achim Gratz (2):
  cygclass/perl.cygclass: do not clobber HOMEPAGE and provide a 
correct default
  lib/pkg_info.cygport: implement automatic determination of the 
appropriate perl5_0xy requirement


Adam Dinwoodie (1):
  autotools.cygclass: correctly detect Autoconf 2.70+

Jon Turney (11):
  Updates to tests for further rusting of python2
  CI: Don't use a relative path with hashfiles()
  testsuite: Add more expected hints
  Don't ignore errors from make in cygtest()
  Add 'vars' command to output arbitrary .cygport variables
  Add srcpackage command
  Add git-archive-all to build-depends
  Updates to tests for package updates
  xorg.cygclass: Allow configuration of default SRC_URI compression
  autotools.cygclass: Report an error if ./configure version could 
not be detected

  Bump version to 0.35.0

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


mintty phantom key presses

2022-05-02 Thread Eric Adams
Hi,

I had previously reported this issue as "Possible phantom control-key state..."

I observe that moving between cygwin mintty windows and Windows
windows causes unexpected behavior in the cygwin world.

I took a new approach, using vi :) . In cygwin, I open a new file,
enter insert mode, hit Ctrl-V, and mouse out of the window. When I
mouse back into the cygwin edit window, my screen contains the display
string "^[[O" (note that the "^[" is vi-speak for "esc"). Examining
the resulting file with od shows:
> od -ah fdsa
000 esc   [   O  nl
   5b1b0a4f

Here, the nl character is inserted by vi.

This smells like an incomplete escape sequence. If it's left at the
command line, just waiting for the user to type something, there might
be trouble.

Am I completely off?
Thanks,
Eric Adams.


fdsa
Description: Binary data


cygcheck.out
Description: Binary data

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


Offer to collaborate with cygwin com

2022-05-02 Thread william . wood

Hello!

My name is William, I am a marketing specialist of Softorino.com

I liked the content of your site and would like to offer you cooperation.

We have a product that solves the problem of users who need to organize their 
workspace.


We could give you our lifetime Universal License 
https://softorino.com/universal-license/ (which already includes 16 premium 
apps) for FREE to organize your workspace and make your workflow pleasant.


All we ask in return, is that you write a review-article about one of our 
apps that the Universal License includes, and include a link to product in 
the article.


It would be great if you could write a review within 1 working week of 
receiving your license.


Write to me if you are interested in my proposal.

Best,
William

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


Re: mintty phantom key presses

2022-05-02 Thread Thomas Wolff



Am 02.05.2022 um 23:54 schrieb Eric Adams:

Hi,

I had previously reported this issue as "Possible phantom control-key state..."

I observe that moving between cygwin mintty windows and Windows
windows causes unexpected behavior in the cygwin world.

I took a new approach, using vi :) . In cygwin, I open a new file,
enter insert mode, hit Ctrl-V, and mouse out of the window. When I
mouse back into the cygwin edit window, my screen contains the display
string "^[[O" (note that the "^[" is vi-speak for "esc"). Examining
the resulting file with od shows:

od -ah fdsa

000 esc   [   O  nl
5b1b0a4f

Here, the nl character is inserted by vi.

This smells like an incomplete escape sequence. If it's left at the
command line, just waiting for the user to type something, there might
be trouble.

Am I completely off?
Thanks,
Eric Adams.

CSI O is the focus off notification (CSI I is the focus in notification).
Someone has switched on focus reporting mode (CSI ? 1004 h) in your 
session (and isn't catching the notifications).

Run your application in a fresh mintty, with no other software, to test.
Thomas

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