Re: [wsjt-devel] OpenBSD build problem

2019-11-28 Thread Larry Gadallah
On Sun, 10 Nov 2019 at 04:14, Bill Somerville  wrote:

> Message: 3
> Date: Sun, 10 Nov 2019 00:22:42 +
> From: Bill Somerville 
> To: wsjt-devel@lists.sourceforge.net
> Subject: Re: [wsjt-devel] OpenBSD build problem
> Message-ID: <32c76b4f-02d7-8f43-ab42-8b51e6048...@classdesign.com>
> Content-Type: text/plain; charset=utf-8; format=flowed
>
> On 10/11/2019 00:02, Larry Gadallah wrote:
> > Modifying CMakeCache.txt to point at gcc 8.3.0 and egfortan over in
> > /usr/local/bin, and adding the global option -fopenmp to the compilers
>
> Hi Larry,
>
> don't do this, CMake will figure out how to build WSJT-X itself, just
> set the compiler(s) by defining CMAKE__COMPILER to the
> compiler driver path where  are the compiler names (C, CXX,
> Fortran), or simply put the compiler on the path.
>
> Having said that the compiler driver you are using (egcc) is configured
> without the OpenMP suppprt (--disable-libgomp). Are you sure that is a
> general desktop system compiler, egcc sounds like some embedded compiler
> designed to make minimal size executables for embedded systems.
>
> 73
> Bill
> G4WJS.
>

Bill:

As near as I can tell, the OpenBSD people have become so militantly
anti-Gnu, that the egcc port is a version of GCC et al from 2007, before
the GPL v3 license was applied. Unfortunately this might mean that features
from newer versions of GCC (like OpenMP) may not be present. I'll follow
your advice regarding CMAKE_*_COMPILER and see how much further I can get.

Thanks,



-- 
Larry Gadallah, lgadallah AT gmail DOT com
PGP Sig: AE93 1785 6874 7111 48AD  63A6 2136 3651 981C F87B
___
wsjt-devel mailing list
wsjt-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/wsjt-devel


Re: [wsjt-devel] OpenBSD build problem

2019-11-09 Thread Larry Gadallah
On 18/01/2019 01:12. Bill Somerville wrote:

> On 17/01/2019 22:27, Jon Tabor wrote:
> > Target: x86_64-unknown-openbsd6.4
> > Configured with: /usr/obj/ports/gcc-4.9.4/gcc-4.9.4/configure
> > --enable-libgcj --without-jar --verbose
> > --program-transform-name='s,^,e,' --disable-nls --with-system-zlib
> > --disable-libmudflap *--disable-libgomp* --disable-tls --with-gnu-ld
> > --with-gnu-as --enable-threads=posix --enable-wchar_t
> > --with-gmp=/usr/local --enable-languages=c,c++,fortran,objc,java,ada
> > --disable-libstdcxx-pch --enable-cpp --prefix=/usr/local
> > --sysconfdir=/etc --mandir=/usr/local/man --infodir=/usr/local/info
> > --localstatedir=/var --disable-silent-rules --disable-gtk-doc
> > Thread model: posix
> > gcc version 4.9.4 (GCC)
>
>
> Hi Jon,
>
> and there is the issue, for some reason the author of the OpenBSD recipe
> to build gcc/g++/gfortran has elected to disable OpenMP, or at lest the
> support library for it.
>
> 73
> Bill
> G4WJS.

I've picked this up again with OpenBSD 6.6 amd64, and after installing Gnu
CC 8.3.0 et al:

openbsd66$ egcc -v
Using built-in specs.
COLLECT_GCC=egcc
COLLECT_LTO_WRAPPER=/usr/local/libexec/gcc/x86_64-unknown-openbsd6.6/8.3.0/lto-wrapper
Target: x86_64-unknown-openbsd6.6
Configured with: /usr/obj/ports/gcc-8.3.0/gcc-8.3.0/configure
--with-stage1-ldflags=-L/usr/obj/ports/gcc-8.3.0/bootstrap/lib --verbose
--program-transform-name='s,^,e,' --disable-nls --with-system-zlib
--disable-libmudflap --disable-libgomp --disable-libssp --disable-tls
--with-gnu-ld --with-gnu-as --enable-threads=posix --enable-wchar_t
--with-gmp=/usr/local --enable-languages=c,c++,fortran,objc,ada
--disable-libstdcxx-pch --enable-default-ssp --enable-default-pie
--without-isl --enable-cpp --prefix=/usr/local --sysconfdir=/etc
--mandir=/usr/local/man --infodir=/usr/local/info --localstatedir=/var
--disable-silent-rules --disable-gtk-doc
Thread model: posix
gcc version 8.3.0 (GCC)

Modifying CMakeCache.txt to point at gcc 8.3.0 and egfortan over in
/usr/local/bin, and adding the global option -fopenmp to the compilers, we
get this result:

openbsd66$ cmake ../src
-- The C compiler identification is GNU 8.3.0
-- The CXX compiler identification is Clang 8.0.1
-- The Fortran compiler identification is GNU 8.3.0
-- Check for working C compiler: /usr/local/bin/egcc
-- Check for working C compiler: /usr/local/bin/egcc -- broken
CMake Error at /usr/local/share/cmake/Modules/CMakeTestCCompiler.cmake:60
(message):
  The C compiler

"/usr/local/bin/egcc"

  is not able to compile a simple test program.

  It fails with the following output:

Change Dir: /home/larry/projects/wsjt-wsjtx/build/CMakeFiles/CMakeTmp

Run Build Command(s):/usr/bin/make cmTC_9fdaf/fast && /usr/bin/make -f
CMakeFiles/cmTC_9fdaf.dir/build.make CMakeFiles/cmTC_9fdaf.dir/build
Building C object CMakeFiles/cmTC_9fdaf.dir/testCCompiler.c.o
/usr/local/bin/egcc   -fopenmp-o
CMakeFiles/cmTC_9fdaf.dir/testCCompiler.c.o   -c
/home/larry/projects/wsjt-wsjtx/build/CMakeFiles/CMakeTmp/testCCompiler.c
Linking C executable cmTC_9fdaf
/usr/local/bin/cmake -E cmake_link_script
CMakeFiles/cmTC_9fdaf.dir/link.txt --verbose=1
/usr/local/bin/egcc -fopenmp
CMakeFiles/cmTC_9fdaf.dir/testCCompiler.c.o  -o cmTC_9fdaf
-Wl,-rpath-link,/usr/X11R6/lib:/usr/local/lib
egcc: error: libgomp.spec: No such file or directory
*** Error 1 in . (CMakeFiles/cmTC_9fdaf.dir/build.make:87 'cmTC_9fdaf')
*** Error 1 in
/home/larry/projects/wsjt-wsjtx/build/CMakeFiles/CMakeTmp (Makefile:121
'cmTC_9fdaf/fast')





  CMake will not be able to correctly generate this project.
Call Stack (most recent call first):
  CMakeLists.txt:25 (project)


-- Configuring incomplete, errors occurred!
See also "/home/larry/projects/wsjt-wsjtx/build/CMakeFiles/CMakeOutput.log".
See also "/home/larry/projects/wsjt-wsjtx/build/CMakeFiles/CMakeError.log".

I grepped about for a bit for information about where the missing
libgomp.spec file should live, and I couldn't find anything. Searching the
system for .spec files yields the following:

openbsd66# find / -depth -name "*.spec"
/home/larry/projects/wsjt-wsjtx/src/boost/tools/build/src/engine/boost-jam.spec
/usr/local/lib/libitm.spec
/usr/local/lib/libgfortran.spec
/usr/local/share/libxml2/docbook.spec
/usr/local/share/libxml2/docbook-xsl.spec

Makes it look like maybe OpenMP support is only partially present?

--
Larry Gadallah, lgadallah AT gmail DOT com
PGP Sig: AE93 1785 6874 7111 48AD  63A6 2136 3651 981C F87B
___
wsjt-devel mailing list
wsjt-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/wsjt-devel


Re: [wsjt-devel] OS X Audio Device Issues

2017-09-24 Thread Larry Gadallah
I can now report that on my older MacBookPro 8,2 (early 2011) laptop
running OS X Mavericks 10.9.5, this problem no longer occurs.

However, I noticed in some earlier postings on the WSJT list, there
was some speculation that this might be caused by some interaction
between the SiLabs CP210x USB serial port driver and the USB codec
driver. I have seen messages in the log like this that indicate the
possible presence of this problem:

9/24/17 11:22:49.000 kernel[0]: WARNING: AppleUSBAudio has detected
that a connected USB audio device is sending too much audio data.
9/24/17 11:22:49.000 kernel[0]: WARNING: This USB audio device may not
function properly. Please notify the device manufacturer.

The newest versions of that driver (5.0 and later) were updated for OS
X 10.11 and later. Anyone trying this trick on OS X 10.10 and earlier
would have to use the earlier 4.x versions of the SiLabs driver, so it
may have more to do with the SiLabs driver than the OS X version?

I'm contemplating bumping up to Yosemite (OS X 10.10) and seeing what happens.

73,

On 22 September 2017 at 19:56,
<wsjt-devel-requ...@lists.sourceforge.net> wrote:
> Message: 3
> Date: Fri, 22 Sep 2017 22:55:39 +0100
> From: Bill Somerville <g4...@classdesign.com>
> To: wsjt-devel@lists.sourceforge.net
> Subject: Re: [wsjt-devel] OS X Audio Device Issues
> Message-ID: <81a71c94-8873-27f1-d050-ecff71006...@classdesign.com>
> Content-Type: text/plain; charset=utf-8; format=flowed
>
> On 22/09/2017 22:49, Larry Gadallah wrote:
>> Does anyone have any references to any Apple bug reports/known issues
>> that document the issue of USB audio devices becoming unreliable,
>> and/or which versions of OS X where this problem emerged?
>
> Hi Larry,
>
> I don't have any Apple ticket numbers but the issues started in Yosemite
> and only seem to effect certain motherboards. Several users have
> reported that not using the USB ports by using a Thunderbolt to USB
> adapter or Thunderbolt sound card avoids the problems, if your Mac
> suffers them and has a Thunderbolt port.
>
> 73
> Bill
> G4WJS.

-- 
Larry Gadallah, VE6VQ/W7  lgadallah AT gmail DOT com
PGP Sig: B5F9 C4A8 8517 82AC 16B6  02B6 0645 69F0 1F29 A512

--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
wsjt-devel mailing list
wsjt-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/wsjt-devel


Re: [wsjt-devel] OS X Audio Device Issues

2017-09-24 Thread Larry Gadallah
Hi all:

I can now report that, thanks to having access to another Apple
laptop, that the combination of WSJT-X v1.8.0-rc2 r8069 and:

- OS X El Capitan 10.11.6
- MacBookPro11,4 (Mid 2015)

works fine with the USB Codec in the IC-7300 and the SiLab CP210x USB
serial drivers.

I have reverted my older MacBook Pro to OS X Mavericks 10.9.5, and I'm
going to try it next and see if this is a software or a hardware
issue. I did see many reports of people using all sorts of USB audio
devices having increasing problems with the Yosemite -> El Capitan ->
Sierra progression, particularly with the older hardware.

73,

On 22 September 2017 at 19:56,
<wsjt-devel-requ...@lists.sourceforge.net> wrote:
> Message: 3
> Date: Fri, 22 Sep 2017 22:55:39 +0100
> From: Bill Somerville <g4...@classdesign.com>
> To: wsjt-devel@lists.sourceforge.net
> Subject: Re: [wsjt-devel] OS X Audio Device Issues
> Message-ID: <81a71c94-8873-27f1-d050-ecff71006...@classdesign.com>
> Content-Type: text/plain; charset=utf-8; format=flowed
>
> On 22/09/2017 22:49, Larry Gadallah wrote:
>> Does anyone have any references to any Apple bug reports/known issues
>> that document the issue of USB audio devices becoming unreliable,
>> and/or which versions of OS X where this problem emerged?
>
> Hi Larry,
>
> I don't have any Apple ticket numbers but the issues started in Yosemite
> and only seem to effect certain motherboards. Several users have
> reported that not using the USB ports by using a Thunderbolt to USB
> adapter or Thunderbolt sound card avoids the problems, if your Mac
> suffers them and has a Thunderbolt port.
>
> 73
> Bill
> G4WJS.

-- 
Larry Gadallah, VE6VQ/W7  lgadallah AT gmail DOT com
PGP Sig: B5F9 C4A8 8517 82AC 16B6  02B6 0645 69F0 1F29 A512

--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
wsjt-devel mailing list
wsjt-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/wsjt-devel


Re: [wsjt-devel] OS X Audio Device Issues

2017-09-22 Thread Larry Gadallah
Does anyone have any references to any Apple bug reports/known issues
that document the issue of USB audio devices becoming unreliable,
and/or which versions of OS X where this problem emerged?

Thanks,

On 22 September 2017 at 06:21,
<wsjt-devel-requ...@lists.sourceforge.net> wrote:
> Message: 3
> Date: Fri, 22 Sep 2017 06:07:20 -0700
> From: Robert Nobis <n7...@nobis.net>
> To: astro.biolog...@virgin.net
> Cc: "Larry Gadallah lgadal...@gmail.com [wsjtgroup]"
> <lgadal...@gmail.com>, Bill Turner via wsjt-devel
> <wsjt-devel@lists.sourceforge.net>, wsjtgr...@yahoogroups.com
> Subject: Re: [wsjt-devel] [wsjtgroup] OS X Audio Device Issues
> Message-ID: <ab61d7b3-d602-448c-b38c-d251b98a3...@nobis.net>
> Content-Type: text/plain; charset="utf-8"
>
> Use two different Macs with WSJT-X. A MacBook Air the I operate portable, and 
> a MacMini in my shack.  I also previously used an iMac.
>
> Interestingly, I have never had any of these USB issues when using the USB 
> ports on the Macs.
>
> My rigs are Elecraft K3 and KX3.
>
>
> Bob Nobis
> n7...@nobis.net
>
>
>> On Sep 22, 2017, at 01:02, astro.biolog...@virgin.net [wsjtgroup] 
>> <wsjtgroup-nore...@yahoogroups.com> wrote:
>>
>>
>> The problem is with USB Audio devices and MacOS, not WSJTX.
>> Basically recent versions of MacOS do not reliably support USB Audio devices 
>> any more, this has been noticed elsewhere.
>>
>> You may have more luck if you buy a Thunderbolt to USB adapter, instead of 
>> plugging your Icom straight into the Mac's USB port (again, this has been 
>> reported elsewhere).
>>
>> In general, I don't think any external USB soundcards work reliably with 
>> Macs any more, you would be better off with an external Thunderbolt 
>> soundcard.
>>
>> Oliver
>> M0LVR

-- 
Larry Gadallah, VE6VQ/W7  lgadallah AT gmail DOT com
PGP Sig: B5F9 C4A8 8517 82AC 16B6  02B6 0645 69F0 1F29 A512

--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
wsjt-devel mailing list
wsjt-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/wsjt-devel


[wsjt-devel] OS X Audio Device Issues

2017-09-21 Thread Larry Gadallah
Hi all:

I have found a few more interesting problems with the Mac version of
WSJT-X 1.80-rc2 r8069:

I am running a Mac laptop with OS X 10.12.6, and I installed the
SiLabs USB serial port drivers. I am using an Icom IC-7300 via the USB
cable, and after considerable tinkering, I think I have it to the
point where the various levels are close to correct.

Now as long as the computer and the screen don't go to sleep, WSPR
decoding works fine. However, when I go to setup transmitting, things
get odd: First of all, if I hit the Tune button, the transmitter will
key, but there won't be any audio.

If I go into Preferences -> Audio and set the Output Device to
Built-in Output and then set it back to USB Audio Codec, then the
transmitter will key and there will be audio at the radio's input.
Now, once that is done, if I click on the Enable Tx button and then
the Tx Next button, I will get a frame sent out during the next
period.

If I wait for another period or two, and press the Tx Next button, the
transmitter will be keyed, but there won't be any audio again. I have
to go back to preferences and "flip" the output device back to USB
Audio Codec to get it to provide transmit audio again.

Everything seems very flaky on the transmit audio side. Has anyone
been able to get this kind of setup running reliably?

A friend of mine, also using MacOS Sierra Version 10.12.6 and WSJT-X
v1.8.0-rc2 r8069 but with an Icom IC-7410 reports the following:

"I too experience the same problems with WSJT-X. I am hooking my Mac
to my Icom 7410 with just a USB cable. The transmit audio is good for
about 3 key-ups then it just goes to the computer speaker and NOT
through the USB cable to the radio. I have no problem with either
receive or transmit when running fldigi."

"I’ve done some further troubleshooting and here’s what I have found.
The problem does not appear to be related to Icom radios. It appears
to be related to USB sound cards and the Mac version WSJT-X. To prove
this I got out my SignaLink USB and connected it to the accessory jack
on the back of my IC 7410. In WSJT-X configuration I selected NONE for
radio and I selected VOX for PTT. I manually set the radio to the 20
meter frequency for FT8. I set the 7410 up so that data audio went in
and out of the accessory jack for data. FT8 signals were received fine
and an audio signal from the Mac/WSJT-X program was sent back through
the USB cable to the radio and would key the transmitter with VOX for
3 or 4 times and then revert to the problem sending outbound
(transmitter bound) audio to the iMac’s speaker (built-in output) when
it was supposed to have sent it via USB to the SignaLink USB which
would key the transmitter. I conclude the problem has to be in the
software and not  related to Icom or any other radios.  The thing that
I have not yet tried but I don’t see why it would work any better is
to connect the radio or SignaLink USB to the Mac via a powered USB
hub."

"I happened to have a powered USB  hub from another project so I tried
it. The results are the same. The audio changes path after keying the
transmitter 3 or 4 times. It’s what I figured would be the result of
the experiment but I figured that I had to just give it a try anyway."

Are there any references about how to build a "debug version" of
WSJT-X to help figure out what's going on in cases like this?

Thanks es 73,
-- 
Larry Gadallah, VE6VQ/W7  lgadallah AT gmail DOT com
PGP Sig: B5F9 C4A8 8517 82AC 16B6  02B6 0645 69F0 1F29 A512

--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
wsjt-devel mailing list
wsjt-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/wsjt-devel