[ANNOUNCEMENT] TEST RELEASE: Cygwin 2.11.0-0.1

[ANNOUNCEMENT] TEST RELEASE: Cygwin 2.11.0-0.1

[Corinna Vinschen]

Hi folks,


I uploaded a new Cygwin test release 2.11.0-0.1

I'm planning for a release end of August.  Please test.

=======================================================================

What's new:
-----------

- New APIs: clearenv, pthread_tryjoin_np, pthread_timedjoin_np,
  sched_getcpu.

- New APIs: aio_cancel, aio_error, aio_fsync, aio_read, aio_return,
  aio_suspend, aio_write, lio_listio.
  New header: <aio.h>.


What changed:
-------------

- SO_RCVTIMEO and SO_SNDTIMEO socket options are now honored.

- /proc/cpuinfo now reports L3 cache size on Intel CPUs.


Bug Fixes
---------

- Fix utils path handling in case cygdrive path is just '/'.
  Addresses: https://cygwin.com/ml/cygwin/2018-02/msg00174.html

- Fix a potential SIGFPE in strtod, if FE_INVALID exceptions are
  enabled.
  Addresses: https://cygwin.com/ml/cygwin/2018-04/msg00055.html

- Fix a CPU affinity problem when creating /proc/cpuinfo output.
  Addresses: https://cygwin.com/ml/cygwin/2018-04/msg00118.html

- Fix a buffer underrun problem in Win32 path normalization.
  Addresses: https://cygwin.com/ml/cygwin/2018-05//msg00017.html

- Fix a stack alignment problem which may lead to spurious crashes after
  fork.
  Addresses: https://cygwin.com/ml/cygwin-patches/2018-q2/msg00016.html

- Fix a g++ compilation problem with -std=c++14 or -std=c++17.
  Addresses: https://cygwin.com/ml/cygwin/2018-05/msg00316.html

- Fix FPE flag handling for division by zero conditions
  Addresses: https://cygwin.com/ml/cygwin/2018-06/msg00281.html

- Fix Unicode table.
  Addresses: https://cygwin.com/ml/cygwin/2018-06/msg00248.html

- Handle a non-standard return value from some tape drives to
  report a "no-media" error.
  Addresses: https://cygwin.com/ml/cygwin/2018-06/msg00245.html

- Fix duration handling in sigtimedwait
  Addresses: https://cygwin.com/ml/cygwin-patches/2018-q3/msg00018.html

- Make FP environment symbols available on x86_64.
  Addresses: https://cygwin.com/ml/cygwin/2018-07/msg00183.html

- Fix fegetenv behaviour.
  Addresses: https://cygwin.com/ml/cygwin/2018-08/msg00000.html

=======================================================================


Have fun,
Corinna

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

--
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 2.11.0-0.1

[Andrey Repin]

Greetings, Corinna Vinschen!

> - Fix utils path handling in case cygdrive path is just '/'.
>   Addresses: https://cygwin.com/ml/cygwin/2018-02/msg00174.html

This one seems fixed all right.


-- 
With best regards,
Andrey Repin
Friday, August 10, 2018 2:49:43

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: [ANNOUNCEMENT] TEST RELEASE: Cygwin 2.11.0-0.1

[Houder]

On Thu, 9 Aug 2018 22:26:04, Corinna Vinschen wrote:
> Hi folks,
> 
> 
> I uploaded a new Cygwin test release 2.11.0-0.1
> 
> I'm planning for a release end of August.  Please test.
[snip]

> - Make FP environment symbols available on x86_64.
>   Addresses: https://cygwin.com/ml/cygwin/2018-07/msg00183.html
> 
> - Fix fegetenv behaviour.
>   Addresses: https://cygwin.com/ml/cygwin/2018-08/msg00000.html

Fixed. (tested both on x86_64 and x86)

Note: the following line must be changed in STC-FENV.c (the STC that was
attached to the bug report -- the last one above).

from:
const int xxx = 0x3d; // FE_ALL_EXCEPT on Linux, i.e. the denormal-operand
                      // exception is excluded on Linux

to:
const int xxx = 0x3f; // Cygwin allows the denormal-operand exception;
                      // Linux (Andreas Jaeger) does not.
                      // This exception is not defined by the IEEE 754
                      // standard (Floating-Point Arithmetic)

(denormal-operand = input of a subnormal value)

Regards,
Henri


--
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 2.11.0-0.1

[Corinna Vinschen]

[-- Attachment #1: Type: text/plain, Size: 1310 bytes --]

On Aug 10 13:05, Houder wrote:
> On Thu, 9 Aug 2018 22:26:04, Corinna Vinschen wrote:
> > Hi folks,
> > 
> > 
> > I uploaded a new Cygwin test release 2.11.0-0.1
> > 
> > I'm planning for a release end of August.  Please test.
> [snip]
> 
> > - Make FP environment symbols available on x86_64.
> >   Addresses: https://cygwin.com/ml/cygwin/2018-07/msg00183.html
> > 
> > - Fix fegetenv behaviour.
> >   Addresses: https://cygwin.com/ml/cygwin/2018-08/msg00000.html
> 
> Fixed. (tested both on x86_64 and x86)
> 
> Note: the following line must be changed in STC-FENV.c (the STC that was
> attached to the bug report -- the last one above).
> 
> from:
> const int xxx = 0x3d; // FE_ALL_EXCEPT on Linux, i.e. the denormal-operand
>                       // exception is excluded on Linux
> 
> to:
> const int xxx = 0x3f; // Cygwin allows the denormal-operand exception;
>                       // Linux (Andreas Jaeger) does not.
>                       // This exception is not defined by the IEEE 754
>                       // standard (Floating-Point Arithmetic)

Shall we change that?  A patch would be nice. ;)


Corinna

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

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

Re: [ANNOUNCEMENT] TEST RELEASE: Cygwin 2.11.0-0.1

[Houder]

On Fri, 10 Aug 2018 13:42:54, Corinna Vinschen wrote:
[snip]

> > Note: the following line must be changed in STC-FENV.c (the STC that was
> > attached to the bug report -- the last one above).
> >
> > from:
> > const int xxx =3D 0x3d; // FE_ALL_EXCEPT on Linux, i.e. the denormal-...
> >                         // exception is excluded on Linux
> >
> > to:
> > const int xxx =3D 0x3f; // Cygwin allows the denormal-operand exception;
> >                         // Linux (Andreas Jaeger) does not.
> >                         // This exception is not defined by the IEEE 754
> >                         // standard (Floating-Point Arithmetic)
> 
> Shall we change that?  A patch would be nice. ;)

That is possible ...

As on Linux, the denormal-operand exception will always be masked after this
change ... (both in the default env. and in the "nomask" env.)

Meaning that, after this change, this exception can * no longer * be enabled
through the API of fenv.h

Objections? Anyone?

The modication would require changing:

winsup/cygwin/fenv.cc (_feinitialise() )
winsup/cygwin/include/fenv.h (FE_ALL_EXCEPT)

/usr/include/fenv.h (which is a copy of the above, I am guessing ?????)

In _feinitialise() this line must be changed:

from:
434	  mxcsr = FE_ALL_EXCEPT << FE_SSE_EXCEPT_MASK_SHIFT;
to:
434	  mxcsr = 0x3f << FE_SSE_EXCEPT_MASK_SHIFT;

Regards,
Henri


--
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 2.11.0-0.1

[Marco Atzeri]

Am 09.08.2018 um 22:26 schrieb Corinna Vinschen:
> Hi folks,
>
>
> I uploaded a new Cygwin test release 2.11.0-0.1
>
> I'm planning for a release end of August.  Please test.
>
> =======================================================================
>
> What's new:
> -----------
>
> - New APIs: clearenv, pthread_tryjoin_np, pthread_timedjoin_np,
>    sched_getcpu.
>
> - New APIs: aio_cancel, aio_error, aio_fsync, aio_read, aio_return,
>    aio_suspend, aio_write, lio_listio.
>    New header: <aio.h>.
>
>
> What changed:
> -------------
>
> - SO_RCVTIMEO and SO_SNDTIMEO socket options are now honored.
>
> - /proc/cpuinfo now reports L3 cache size on Intel CPUs.
>
>
> Bug Fixes
> ---------
>
> - Fix utils path handling in case cygdrive path is just '/'.
>    Addresses: https://cygwin.com/ml/cygwin/2018-02/msg00174.html
>
> - Fix a potential SIGFPE in strtod, if FE_INVALID exceptions are
>    enabled.
>    Addresses: https://cygwin.com/ml/cygwin/2018-04/msg00055.html
>
> - Fix a CPU affinity problem when creating /proc/cpuinfo output.
>    Addresses: https://cygwin.com/ml/cygwin/2018-04/msg00118.html
>
> - Fix a buffer underrun problem in Win32 path normalization.
>    Addresses: https://cygwin.com/ml/cygwin/2018-05//msg00017.html
>
> - Fix a stack alignment problem which may lead to spurious crashes after
>    fork.
>    Addresses: https://cygwin.com/ml/cygwin-patches/2018-q2/msg00016.html
>
> - Fix a g++ compilation problem with -std=c++14 or -std=c++17.
>    Addresses: https://cygwin.com/ml/cygwin/2018-05/msg00316.html
>
> - Fix FPE flag handling for division by zero conditions
>    Addresses: https://cygwin.com/ml/cygwin/2018-06/msg00281.html
>
> - Fix Unicode table.
>    Addresses: https://cygwin.com/ml/cygwin/2018-06/msg00248.html
>
> - Handle a non-standard return value from some tape drives to
>    report a "no-media" error.
>    Addresses: https://cygwin.com/ml/cygwin/2018-06/msg00245.html
>
> - Fix duration handling in sigtimedwait
>    Addresses: https://cygwin.com/ml/cygwin-patches/2018-q3/msg00018.html
>
> - Make FP environment symbols available on x86_64.
>    Addresses: https://cygwin.com/ml/cygwin/2018-07/msg00183.html
>
> - Fix fegetenv behaviour.
>    Addresses: https://cygwin.com/ml/cygwin/2018-08/msg00000.html
>
> =======================================================================
>
>
> Have fun,
> Corinna
>

testing libuv, I see a large number of new failures

not ok 217 - tcp6_ping_pong
# timeout
# Output from process `tcp6_ping_pong`: (no output)
# PING
not ok 218 - tcp6_ping_pong_vec
# timeout
# Output from process `tcp6_ping_pong_vec`: (no output)
# PING
not ok 219 - tcp_alloc_cb_fail
# timeout
# Output from process `tcp_alloc_cb_fail`: (no output)
not ok 234 - tcp_close_accept
# timeout
# Output from process `tcp_close_accept`: (no output)
not ok 235 - tcp_close_while_connecting
# timeout
# Output from process `tcp_close_while_connecting`: (no output)
not ok 238 - tcp_connect_timeout
# timeout
# Output from process `tcp_connect_timeout`: (no out
not ok 245 - tcp_oob
# timeout
# Output from process `tcp_oob`: (no output)
not ok 246 - tcp_open
# timeout
# Output from process `tcp_open`: (no output)
not ok 248 - tcp_open_connected
# timeout
# Output from process `tcp_open_connected`: (no output)
not ok 250 - tcp_ping_pong
# timeout
# Output from process `tcp_ping_pong`: (no output)
# PING
not ok 251 - tcp_ping_pong_vec
# timeout
# Output from process `tcp_ping_pong_vec`: (no output)
# PING
not ok 259 - tcp_try_write
# timeout
# Output from process `tcp_try_write`: (no output)
not ok 260 - tcp_unexpected_read
# timeout
# Output from process `tcp_unexpected_read`: (no output)
not ok 263 - tcp_write_queue_order
# timeout
# Output from process `tcp_write_queue_order`: (no output)
not ok 264 - tcp_write_to_half_open_connection
# timeout
# Output from process `tcp_write_to_half_open_connection`: (no output)
not ok 265 - tcp_writealot
# timeout
# Output from process `tcp_writealot`: (no output)

---
Diese E-Mail wurde von Avast Antivirus-Software auf Viren geprüft.
https://www.avast.com/antivirus


--
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 2.11.0-0.1

[Corinna Vinschen]

[-- Attachment #1: Type: text/plain, Size: 1279 bytes --]

On Aug 10 14:43, Houder wrote:
> On Fri, 10 Aug 2018 13:42:54, Corinna Vinschen wrote:
> [snip]
> 
> > > Note: the following line must be changed in STC-FENV.c (the STC that was
> > > attached to the bug report -- the last one above).
> > >
> > > from:
> > > const int xxx =3D 0x3d; // FE_ALL_EXCEPT on Linux, i.e. the denormal-...
> > >                         // exception is excluded on Linux
> > >
> > > to:
> > > const int xxx =3D 0x3f; // Cygwin allows the denormal-operand exception;
> > >                         // Linux (Andreas Jaeger) does not.
> > >                         // This exception is not defined by the IEEE 754
> > >                         // standard (Floating-Point Arithmetic)
> > 
> > Shall we change that?  A patch would be nice. ;)
> 
> That is possible ...
> 
> As on Linux, the denormal-operand exception will always be masked after this
> change ... (both in the default env. and in the "nomask" env.)
> 
> Meaning that, after this change, this exception can * no longer * be enabled
> through the API of fenv.h
> 
> Objections? Anyone?

Not from me.


Corinna

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

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

Re: [ANNOUNCEMENT] TEST RELEASE: Cygwin 2.11.0-0.1

[Corinna Vinschen]

[-- Attachment #1: Type: text/plain, Size: 411 bytes --]

On Aug 10 18:00, Marco Atzeri wrote:
> Am 09.08.2018 um 22:26 schrieb Corinna Vinschen:
> > Hi folks,
> > 
> > 
> > I uploaded a new Cygwin test release 2.11.0-0.1
> > [...]
> 
> testing libuv, I see a large number of new failures

STC?


Corinna

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

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

Re: [ANNOUNCEMENT] TEST RELEASE: Cygwin 2.11.0-0.1

[Houder]

On 2018-08-10 14:43, Houder wrote:
> On Fri, 10 Aug 2018 13:42:54, Corinna Vinschen wrote:
> [snip]
> 
>> > Note: the following line must be changed in STC-FENV.c (the STC that was
>> > attached to the bug report -- the last one above).
>> >
>> > from:
>> > const int xxx =3D 0x3d; // FE_ALL_EXCEPT on Linux, i.e. the denormal-...
>> >                         // exception is excluded on Linux
>> >
>> > to:
>> > const int xxx =3D 0x3f; // Cygwin allows the denormal-operand exception;
>> >                         // Linux (Andreas Jaeger) does not.
>> >                         // This exception is not defined by the IEEE 754
>> >                         // standard (Floating-Point Arithmetic)
>> 
>> Shall we change that?  A patch would be nice. ;)
> 
> That is possible ...
> 
> As on Linux, the denormal-operand exception will always be masked after 
> this
> change ... (both in the default env. and in the "nomask" env.)
> 
> Meaning that, after this change, this exception can * no longer * be 
> enabled
> through the API of fenv.h
> 
> Objections? Anyone?
> 
> The modication would require changing:
> 
> winsup/cygwin/fenv.cc (_feinitialise() )
> winsup/cygwin/include/fenv.h (FE_ALL_EXCEPT)
> 
> /usr/include/fenv.h (which is a copy of the above, I am guessing ?????)

Stupid question:

Does the modification of winsup/cygwin/include/fenv.h "automatically" 
propogate
to /usr/include/fenv.h

Said differently, does "your" machinery (build) take care of that?

Regards,
Henri

--
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 2.11.0-0.1

[Corinna Vinschen]

[-- Attachment #1: Type: text/plain, Size: 1903 bytes --]

On Aug 13 10:03, Houder wrote:
> On 2018-08-10 14:43, Houder wrote:
> > On Fri, 10 Aug 2018 13:42:54, Corinna Vinschen wrote:
> > [snip]
> > 
> > > > Note: the following line must be changed in STC-FENV.c (the STC that was
> > > > attached to the bug report -- the last one above).
> > > >
> > > > from:
> > > > const int xxx =3D 0x3d; // FE_ALL_EXCEPT on Linux, i.e. the denormal-...
> > > >                         // exception is excluded on Linux
> > > >
> > > > to:
> > > > const int xxx =3D 0x3f; // Cygwin allows the denormal-operand exception;
> > > >                         // Linux (Andreas Jaeger) does not.
> > > >                         // This exception is not defined by the IEEE 754
> > > >                         // standard (Floating-Point Arithmetic)
> > > 
> > > Shall we change that?  A patch would be nice. ;)
> > 
> > That is possible ...
> > 
> > As on Linux, the denormal-operand exception will always be masked after
> > this
> > change ... (both in the default env. and in the "nomask" env.)
> > 
> > Meaning that, after this change, this exception can * no longer * be
> > enabled
> > through the API of fenv.h
> > 
> > Objections? Anyone?
> > 
> > The modication would require changing:
> > 
> > winsup/cygwin/fenv.cc (_feinitialise() )
> > winsup/cygwin/include/fenv.h (FE_ALL_EXCEPT)
> > 
> > /usr/include/fenv.h (which is a copy of the above, I am guessing ?????)
> 
> Stupid question:
> 
> Does the modification of winsup/cygwin/include/fenv.h "automatically"
> propogate
> to /usr/include/fenv.h

Yes, `make install' copies the files where they belong, just usually
not to /usr/include/fenv.h but to ${INSTALL_DIR}/usr/include/fenv.h
for later packaging.


Corinna

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

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

Re: [ANNOUNCEMENT] TEST RELEASE: Cygwin 2.11.0-0.1

[Thomas Wolff]

Am 09.08.2018 um 22:26 schrieb Corinna Vinschen:
> Hi folks,
>
>
> I uploaded a new Cygwin test release 2.11.0-0.1
>
> I'm planning for a release end of August.  Please test.
>
> =======================================================================
>
> What's new:
> -----------
>
> ...
>
> What changed:
> -------------
>
> ...
>
> Bug Fixes
> ---------
>
> ...
>
> - Fix Unicode table.
>    Addresses: https://cygwin.com/ml/cygwin/2018-06/msg00248.html
So basically the update from Unicode 5.0 to Unicode 10.0 will be in the 
release; maybe that's worth an explicit item in the "What changed" section?
Should we make that Unicode 11.0 still on this occasion, or did you 
update already?
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

Re: [ANNOUNCEMENT] TEST RELEASE: Cygwin 2.11.0-0.1

[Corinna Vinschen]

[-- Attachment #1: Type: text/plain, Size: 1113 bytes --]

On Aug 13 10:35, Thomas Wolff wrote:
> Am 09.08.2018 um 22:26 schrieb Corinna Vinschen:
> > Hi folks,
> > 
> > 
> > I uploaded a new Cygwin test release 2.11.0-0.1
> > 
> > I'm planning for a release end of August.  Please test.
> > 
> > =======================================================================
> > 
> > What's new:
> > -----------
> > 
> > ...
> > 
> > What changed:
> > -------------
> > 
> > ...
> > 
> > Bug Fixes
> > ---------
> > 
> > ...
> > 
> > - Fix Unicode table.
> >    Addresses: https://cygwin.com/ml/cygwin/2018-06/msg00248.html
> So basically the update from Unicode 5.0 to Unicode 10.0 will be in the
> release; maybe that's worth an explicit item in the "What changed" section?

Sure, patch welcome.

> Should we make that Unicode 11.0 still on this occasion, or did you update
> already?

No, I didn't.  Let's skip this for the next release.  Patch welcome
afterwards.


Thanks,
Corinna

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

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

Re: [ANNOUNCEMENT] TEST RELEASE: Cygwin 2.11.0-0.1

[Achim Gratz]

Houder writes:
> from:
> const int xxx = 0x3d; // FE_ALL_EXCEPT on Linux, i.e. the denormal-operand
>                       // exception is excluded on Linux
>
> to:
> const int xxx = 0x3f; // Cygwin allows the denormal-operand exception;
>                       // Linux (Andreas Jaeger) does not.
>                       // This exception is not defined by the IEEE 754
>                       // standard (Floating-Point Arithmetic)
>
> (denormal-operand = input of a subnormal value)

Linux supports a range of hardware where denormals are not supported or
a major performance hit.  Windows should be only supported on hardware
where this problem doesn't exist.  That said, numerical software that
relies on correct denormals is really rare and following Linux might be
a good idea.


Regards,
Achim.
-- 
+<[Q+ Matrix-12 WAVE#46+305 Neuron microQkb Andromeda XTk Blofeld]>+

Factory and User Sound Singles for Waldorf rackAttack:
http://Synth.Stromeko.net/Downloads.html#WaldorfSounds

--
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 2.11.0-0.1

[Houder]

On 2018-08-10 14:43, Houder wrote:
> On Fri, 10 Aug 2018 13:42:54, Corinna Vinschen wrote:
> [snip]
> 
>> > Note: the following line must be changed in STC-FENV.c (the STC that was
>> > attached to the bug report -- the last one above).
>> >
>> > from:
>> > const int xxx =3D 0x3d; // FE_ALL_EXCEPT on Linux, i.e. the denormal-...
>> >                         // exception is excluded on Linux
>> >
>> > to:
>> > const int xxx =3D 0x3f; // Cygwin allows the denormal-operand exception;
>> >                         // Linux (Andreas Jaeger) does not.
>> >                         // This exception is not defined by the IEEE 754
>> >                         // standard (Floating-Point Arithmetic)
>> 
>> Shall we change that?  A patch would be nice. ;)
> 
> That is possible ...
> 
> As on Linux, the denormal-operand exception will always be masked after 
> this
> change ... (both in the default env. and in the "nomask" env.)
> 
> Meaning that, after this change, this exception can * no longer * be 
> enabled
> through the API of fenv.h
> 
> Objections? Anyone?
> 
> The modication would require changing:
> 
> winsup/cygwin/fenv.cc (_feinitialise() )
> winsup/cygwin/include/fenv.h (FE_ALL_EXCEPT)

GRRR! The file encoding of fenv.h is "cp1252" because of 2 characters in 
this
line:

      Intel® 64 and IA-32 Architectures Software Developer’s Manuals:

... part of a comment at the beginning of the file.

(the registered trademark sign (u00ae) is encoded as 0xae (cp1252), 
while it
  would be: 0xc2 0xae, in utf-8,
  the right single quotation mark (u2019) is encoded as 0x92 (cp1252), 
but in
  utf-8 it would be: 0xc2 0x80 0x98)

I intend to convert the file encoding of fenv.h to utf-8. Is that a 
"No-No"
or is it allowed? (I assume GIT will notice).

Regards,
Henri

--
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 2.11.0-0.1

[Eric Blake]

On 08/13/2018 04:29 PM, Houder wrote:

>> The modication would require changing:
>>
>> winsup/cygwin/fenv.cc (_feinitialise() )
>> winsup/cygwin/include/fenv.h (FE_ALL_EXCEPT)
> 
> GRRR! The file encoding of fenv.h is "cp1252" because of 2 characters in 
> this
> line:
> 
>       Intel® 64 and IA-32 Architectures Software Developer’s Manuals:
> 
> ... part of a comment at the beginning of the file.
> 
> (the registered trademark sign (u00ae) is encoded as 0xae (cp1252), 
> while it
>   would be: 0xc2 0xae, in utf-8,
>   the right single quotation mark (u2019) is encoded as 0x92 (cp1252), 
> but in
>   utf-8 it would be: 0xc2 0x80 0x98)
> 
> I intend to convert the file encoding of fenv.h to utf-8. Is that a "No-No"
> or is it allowed? (I assume GIT will notice).

In general, git doesn't care if you change a file's encoding - that's 
just another content change.  In practice, you may get weird effects 
when viewing that particular patch (as the patch is not well-formed in 
the new multibyte locale, and looks funky when displayed in the old 
locale), and emailing a patch may require care in telling git which 
encoding to use for the email; but that's cosmetic, and shouldn't matter 
in the long run.  Updating the code base to uniformly use UTF-8 seems 
reasonable to me.

-- 
Eric Blake, Principal Software Engineer
Red Hat, Inc.           +1-919-301-3266
Virtualization:  qemu.org | libvirt.org

--
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 2.11.0-0.1

[Houder]

On 2018-08-13 23:29, Houder wrote:
> On 2018-08-10 14:43, Houder wrote:
>> On Fri, 10 Aug 2018 13:42:54, Corinna Vinschen wrote:
>> [snip]
>> 
>>> > Note: the following line must be changed in STC-FENV.c (the STC that was
>>> > attached to the bug report -- the last one above).
>>> >
>>> > from:
>>> > const int xxx =3D 0x3d; // FE_ALL_EXCEPT on Linux, i.e. the denormal-...
>>> >                         // exception is excluded on Linux
>>> >
>>> > to:
>>> > const int xxx =3D 0x3f; // Cygwin allows the denormal-operand exception;
>>> >                         // Linux (Andreas Jaeger) does not.
>>> >                         // This exception is not defined by the IEEE 754
>>> >                         // standard (Floating-Point Arithmetic)
>>> 
>>> Shall we change that?  A patch would be nice. ;)
>> 
>> That is possible ...
>> 
>> As on Linux, the denormal-operand exception will always be masked 
>> after this
>> change ... (both in the default env. and in the "nomask" env.)
>> 
>> Meaning that, after this change, this exception can * no longer * be 
>> enabled
>> through the API of fenv.h
>> 
>> Objections? Anyone?
>> 
>> The modication would require changing:
>> 
>> winsup/cygwin/fenv.cc (_feinitialise() )
>> winsup/cygwin/include/fenv.h (FE_ALL_EXCEPT)
> 
> GRRR! The file encoding of fenv.h is "cp1252" because of 2 characters 
> in this
> line:
> 
>      Intel® 64 and IA-32 Architectures Software Developer’s Manuals:
> 
> ... part of a comment at the beginning of the file.
> 
> (the registered trademark sign (u00ae) is encoded as 0xae (cp1252), 
> while it
>  would be: 0xc2 0xae, in utf-8,
>  the right single quotation mark (u2019) is encoded as 0x92 (cp1252), 
> but in
>  utf-8 it would be: 0xc2 0x80 0x98)

s/ 0xc2 0x80 0x98 / 0xe2 0x80 0x99 /

Sorry!

--
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 2.11.0-0.1

[Houder]

On 2018-08-14 00:16, Eric Blake wrote:
> On 08/13/2018 04:29 PM, Houder wrote:
> 
>>> The modication would require changing:
>>> 
>>> winsup/cygwin/fenv.cc (_feinitialise() )
>>> winsup/cygwin/include/fenv.h (FE_ALL_EXCEPT)
>> 
>> GRRR! The file encoding of fenv.h is "cp1252" because of 2 characters 
>> in this
>> line:
>> 
>>       Intel® 64 and IA-32 Architectures Software Developer’s Manuals:
>> 
>> ... part of a comment at the beginning of the file.
>> 
>> (the registered trademark sign (u00ae) is encoded as 0xae (cp1252), 
>> while it
>>   would be: 0xc2 0xae, in utf-8,
>>   the right single quotation mark (u2019) is encoded as 0x92 (cp1252), 
>> but in
>>   utf-8 it would be: 0xc2 0x80 0x98)
>> 
>> I intend to convert the file encoding of fenv.h to utf-8. Is that a 
>> "No-No"
>> or is it allowed? (I assume GIT will notice).
> 
> In general, git doesn't care if you change a file's encoding - that's
> just another content change.  In practice, you may get weird effects
> when viewing that particular patch (as the patch is not well-formed in
> the new multibyte locale, and looks funky when displayed in the old
> locale), and emailing a patch may require care in telling git which
> encoding to use for the email; but that's cosmetic, and shouldn't
> matter in the long run.  Updating the code base to uniformly use UTF-8
> seems reasonable to me.

... and emailing a patch may require care in telling git which encoding
to use for the email ... Huh, huh ?????

Last time I used:

  - git format-patch
  - git send-mail

Am I safe here?

Regards,
Henri

--
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 2.11.0-0.1

[john doe]

On 8/14/2018 12:26 AM, Houder wrote:
> On 2018-08-14 00:16, Eric Blake wrote:
>> On 08/13/2018 04:29 PM, Houder wrote:
>>
>>>> The modication would require changing:
>>>>
>>>> winsup/cygwin/fenv.cc (_feinitialise() )
>>>> winsup/cygwin/include/fenv.h (FE_ALL_EXCEPT)
>>>
>>> GRRR! The file encoding of fenv.h is "cp1252" because of 2 characters 
>>> in this
>>> line:
>>>
>>>       Intel® 64 and IA-32 Architectures Software Developer’s Manuals:
>>>
>>> ... part of a comment at the beginning of the file.
>>>
>>> (the registered trademark sign (u00ae) is encoded as 0xae (cp1252), 
>>> while it
>>>   would be: 0xc2 0xae, in utf-8,
>>>   the right single quotation mark (u2019) is encoded as 0x92 
>>> (cp1252), but in
>>>   utf-8 it would be: 0xc2 0x80 0x98)
>>>
>>> I intend to convert the file encoding of fenv.h to utf-8. Is that a 
>>> "No-No"
>>> or is it allowed? (I assume GIT will notice).
>>
>> In general, git doesn't care if you change a file's encoding - that's
>> just another content change.  In practice, you may get weird effects
>> when viewing that particular patch (as the patch is not well-formed in
>> the new multibyte locale, and looks funky when displayed in the old
>> locale), and emailing a patch may require care in telling git which
>> encoding to use for the email; but that's cosmetic, and shouldn't
>> matter in the long run.  Updating the code base to uniformly use UTF-8
>> seems reasonable to me.
> 
> ... and emailing a patch may require care in telling git which encoding
> to use for the email ... Huh, huh ?????
> 
> Last time I used:
> 
>   - git format-patch
>   - git send-mail
> 
> Am I safe here?
> 

To better understand what has happened in the code cosmetic changes 
should be done in separate commits.

- Code changes -- Will only change the code.
- Cosmetic changes -- Will only change the cosmetic aspect of the code 
(encoding, spacing, indentation ...)

In most cases you can simply do 'git send-email' and pass the options 
for 'git format-patch' at the end of the cmd:

$ git send-email master --to=me@example.com --reroll-count 1 --rfc


The options '--reroll-count, --rfc' are format-patch options.

-- 
John Doe

--
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 2.11.0-0.1

[Corinna Vinschen]

[-- Attachment #1: Type: text/plain, Size: 1919 bytes --]

On Aug 13 23:29, Houder wrote:
> On 2018-08-10 14:43, Houder wrote:
> > On Fri, 10 Aug 2018 13:42:54, Corinna Vinschen wrote:
> > [snip]
> > 
> > > > Note: the following line must be changed in STC-FENV.c (the STC that was
> > > > attached to the bug report -- the last one above).
> > > >
> > > > from:
> > > > const int xxx =3D 0x3d; // FE_ALL_EXCEPT on Linux, i.e. the denormal-...
> > > >                         // exception is excluded on Linux
> > > >
> > > > to:
> > > > const int xxx =3D 0x3f; // Cygwin allows the denormal-operand exception;
> > > >                         // Linux (Andreas Jaeger) does not.
> > > >                         // This exception is not defined by the IEEE 754
> > > >                         // standard (Floating-Point Arithmetic)
> > > 
> > > Shall we change that?  A patch would be nice. ;)
> > 
> > That is possible ...
> > 
> > As on Linux, the denormal-operand exception will always be masked after
> > this
> > change ... (both in the default env. and in the "nomask" env.)
> > 
> > Meaning that, after this change, this exception can * no longer * be
> > enabled
> > through the API of fenv.h
> > 
> > Objections? Anyone?
> > 
> > The modication would require changing:
> > 
> > winsup/cygwin/fenv.cc (_feinitialise() )
> > winsup/cygwin/include/fenv.h (FE_ALL_EXCEPT)
> 
> GRRR! The file encoding of fenv.h is "cp1252" because of 2 characters in
> this
> line:
> 
>      Intel® 64 and IA-32 Architectures Software Developer’s Manuals:
> 
> ... part of a comment at the beginning of the file.

Don't worry about it.  I converted the file to ASCII-only and pushed the
change.  ASCII-only is the least common denominator and is preferred for
sources anyway.


Corinna

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

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

Re: [ANNOUNCEMENT] TEST RELEASE: Cygwin 2.11.0-0.1

[Houder]

On 2018-08-14 11:50, Corinna Vinschen wrote:
> On Aug 13 23:29, Houder wrote:
>> On 2018-08-10 14:43, Houder wrote:
[snip]

>> > The modication would require changing:
>> >
>> > winsup/cygwin/fenv.cc (_feinitialise() )
>> > winsup/cygwin/include/fenv.h (FE_ALL_EXCEPT)
>> 
>> GRRR! The file encoding of fenv.h is "cp1252" because of 2 characters 
>> in
>> this
>> line:
>> 
>>      Intel® 64 and IA-32 Architectures Software Developer’s Manuals:
>> 
>> ... part of a comment at the beginning of the file.
> 
> Don't worry about it.  I converted the file to ASCII-only and pushed 
> the
> change.  ASCII-only is the least common denominator and is preferred 
> for
> sources anyway.

... even better! Thank you.

Regards,
Henri

--
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 2.11.0-0.1

[Achim Gratz]

Corinna Vinschen writes:
> I uploaded a new Cygwin test release 2.11.0-0.1
>
> I'm planning for a release end of August.  Please test.

I've tested it for about a week now and have released for my other users
this week.  This release actually fixed two problem I've encountered on
my new Server 2016 machine, both of which looked like races and have
evaded analysis since things worked both when straced and run in the
debugger.  I have no idea which of the changes fixed that, though.


Regards,
Achim.
-- 
+<[Q+ Matrix-12 WAVE#46+305 Neuron microQkb Andromeda XTk Blofeld]>+

SD adaptation for Waldorf Blofeld V1.15B11:
http://Synth.Stromeko.net/Downloads.html#WaldorfSDada

--
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