So if it helps to shed any light, right after I performed the OS upgrade and before reinstalling all my ports I got back into the system and started up X and ran a few things to see if everything was good to proceed I did try wsjtx as well and to my surprise it worked fine and didn't crash between switching contacts, so I went ahead with a pkg-static upgrade -f to bring everything level with the 13 version and then started having this issue again. Maybe some conflict with another port? Just my guess.
Thanks Jason On Mon, Apr 11, 2022, 10:52 PM Alex Lelievre <a...@foinc.com> wrote: > Sorry let me be a bit more specific: > > The message field size that the runtime is complaining about has a length > of 22. > > The message passed in has a size of -4294967274. > > -4294967274 is 0xFFFFFFFF00000016 in hexadecimal. > > And if you ignore the top part of that 64 bit value, you get 0x16 which in > decimal is 22. So there’s some sanity. > > Unfortunately there could be any number of issues that cause the value to > become negative. > > alex > K6LOT > > > > On Apr 11, 2022, at 7:14 PM, Alex Lelievre <a...@foinc.com> wrote: > > > > Yes that is helpful- I ran into this same problem but I don’t remember > exactly what I did to solve it and my source code is not showing a diff in > that file. > > > > I believe what I had to do was to make sure that the caller to that > routine was passing the right length string. > > > > However the “-4294967274” length you have makes me believe that the > length has underflowed or is improperly encoded because in hex that value > is 0xFFFFFFFF00000016 and typically lengths of things aren’t negative! > > > > Sorry I don’t have a better answer- I plan to build the debug version > of the source code soon for Apple M1 to look into a Hamlib crash that was > reported to me (only happens on M1) and I have a funny feeling I’m going to > be hitting this problem as well. > > > > alex > > K6LOT > > > > > >> On Apr 11, 2022, at 6:48 PM, Jason Selwitz <jselw...@gmail.com> wrote: > >> > >> is this what you were refering to? > >> > >> At line 1 of file > /usr/ports/comms/wsjtx/work/.build/wsjtx-prefix/src/wsjtx/lib/gen65.f90 > >> Fortran runtime error: Actual string length is shorter than the > declared one for dummy argument 'msgsent' (-4294967274/22) > >> > >> Thanks again > >> > >> On 4/11/22 17:28, Jason Selwitz wrote: > >>> Hi Alex, Yes this is the exact same issue nothing has changed.. > >>> > >>> And I would be happy to provide the backtrace, is there a particular > way you need it generated, sorry I don't do it much. > >>> > >>> Thanks again for the assistance. > >>> > >>> Jason > >>> > >>> On 4/11/22 17:09, Alex Lelievre wrote: > >>>> Hi Jason, > >>>> > >>>> Can you provide a backtrace from the crash you are seeing? > >>>> Is this the same problem you reported back in January with a Fortran > runtime error? > >>>> > >>>> alex > >>>> K6LOT > >>>> > >>>>> On Apr 11, 2022, at 1:41 PM, Jason Selwitz via wsjt-devel < > wsjt-devel@lists.sourceforge.net> wrote: > >>>>> > >>>>> Just following up again on this issue, I went ahead and finally > upgraded to 13.0 and reinstalled all my ports yet the crash still persists > with the same error, I have tried recompiling with both GCC 10 and 8 to see > if it would make a difference and no appreciable improvements have resulted > .. I also tried the suggestion to adjust the lib/grid2deg.f90 file to read > as follows > >>>>> > >>>>> Remove the line: > >>>>> character*6 grid0,grid > >>>>> > >>>>> And replace with: > >>>>> character(len=*) :: grid0 > >>>>> character*6 grid > >>>>> > >>>>> This also has not resulted in any improvement, any further > suggestions would be very much appreciated. > >>>>> > >>>>> If there is any additional information I can provide or > troubleshooting steps to try I would be happy to do so. > >>>>> > >>>>> Thanks again everyone.. > >>>>> > >>>>> Jason > >>>>> > >>>>> > >>>>> On 1/7/22 15:21, Jeff Stillinger via wsjt-devel wrote: > >>>>>> One of the beautiful things about FreeBSD is building out of the > ports collection. Everything is built from source, so it should not matter > if you are on 11, 12 or 13. I would encourage an update to 13 if you have > a non-production machine, simply because that is where people seem to be > doing all the software building these days. Otherwise, leave a production > machine alone. > >>>>>> > >>>>>> You will see warnings as WSJT-X compiles with several declared but > unused variables. This is true on all platforms. Joe assures us that many > of these warnings will be cleaned up soon. The warnings do not effect the > operation of WSJT-X or the build. > >>>>>> > >>>>>> If you make the jump to 13. Install: GCC-10_4, QT5, Ruby and Gems > prior to starting your build. This will save time, dependencies will be > built from source if not installed. For example, if gcc is not installed, > it will be built from source. Which takes most of the day. Yes, I found > out from the school of hard knocks... :-) The other minor dependencies > like pulseaudio build quickly enough to not worry about. > >>>>>> > >>>>>> If your build on 12 falls over, please document it. I can forward > it upstream to the FreeBSD team to fix a broken port. > >>>>>> > >>>>>> > >>>>>> > >>>>>> On 1/7/22 09:13, Jason Selwitz via wsjt-devel wrote: > >>>>>>> Thanks Jeff. > >>>>>>> > >>>>>>> I'm running 12.X currently but 2.5.4 seems to build fine for me at > least I could try to re-run the build and watch for any warnings or > errors.. is it a matter of needing to jump to the 13.X version train? or > should it still work under 12.X? > >>>>>>> > >>>>>>> > >>>>>>> Thanks again > >>>>>>> > >>>>>>> Jason > >>>>>>> > >>>>>>> > >>>>>>> On 1/7/22 09:50, Jeff Stillinger via wsjt-devel wrote: > >>>>>>>> That has been fixed and WSJT-X 2.5.4 builds out of the ports > collection on FreeBSD 13. > >>>>>>>> > >>>>>>>> portsnap fetch update > >>>>>>>> > >>>>>>>> make install clean > >>>>>>>> > >>>>>>>> WSJT-X 2.5.4 builds and installs without error. > >>>>>>>> > >>>>>>>> > >>>>>>>> Oh, and thanks for the reminder to send Diane the request to add > the binary to the package collection. > >>>>>>>> > >>>>>>>> > >>>>>>>> > >>>>>>>> On 1/6/22 16:22, Jason Selwitz via wsjt-devel wrote: > >>>>>>>>> Hi there, > >>>>>>>>> > >>>>>>>>> I was wondering if anyone had seen the following error it > usually happens when I try to double click on another callsign to start > attempting to make a contact Ive tried removing and rebuilding the > configuration and tried changing a few settings in the software to try to > find a work around but so far no luck. if anyone has any tip, suggestions, > etc it would be much appreciated, if this is not the right place I can > always try to follow up with the FreeBSD porters team. > >>>>>>>>> > >>>>>>>>> Fortran runtime error: Actual string length is shorter than the > declared one for dummy argument 'msgsent' (-4294967274/22) > >>>>>>>>> > >>>>>>>>> Thanks, again! > >>>>>>>>> > >>>>>>>>> Jason > >>>>>>>>> > >>>>>>>>> > >>>>>>>>> > >>>>>>>>> _______________________________________________ > >>>>>>>>> wsjt-devel mailing list > >>>>>>>>> wsjt-devel@lists.sourceforge.net > >>>>>>>>> https://lists.sourceforge.net/lists/listinfo/wsjt-devel > >>>>>>> > >>>>>>> _______________________________________________ > >>>>>>> wsjt-devel mailing list > >>>>>>> wsjt-devel@lists.sourceforge.net > >>>>>>> https://lists.sourceforge.net/lists/listinfo/wsjt-devel > >>>>> > >>>>> _______________________________________________ > >>>>> wsjt-devel mailing list > >>>>> wsjt-devel@lists.sourceforge.net > >>>>> https://lists.sourceforge.net/lists/listinfo/wsjt-devel > > > >
_______________________________________________ wsjt-devel mailing list wsjt-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/wsjt-devel
