Hi Roger,
thanks for the prompt response. I can confirm the patched file fixed
the problem in Fedora. There are still some warnings regarding the
fortran code, but as the gfortran build flags aren't hardened in
Fedora it doesn't block the build. Adding the fortran warnings for the
reference:
Warning: ‘snr1’ may be used uninitialized [-Wmaybe-uninitialized]
/builddir/build/BUILD/wsjtx-3.0.0_rc1-build/wsjtx-3.0.0-rc1/wsjtx/lib/sh65.f90:85:49:
85 | call sh65snr(ss(ia2,nbest),ib2-ia2+1,snr1)
--
Warning: ‘snr1’ may be used uninitialized [-Wmaybe-uninitialized]
/builddir/build/BUILD/wsjtx-3.0.0_rc1-build/wsjtx-3.0.0-rc1/wsjtx/lib/sh65.f90:85:49:
85 | call sh65snr(ss(ia2,nbest),ib2-ia2+1,snr1)
--
Warning: ‘REALPART_EXPR <z4>’ may be used uninitialized [-Wmaybe-uninitialized]
/builddir/build/BUILD/wsjtx-3.0.0_rc1-build/wsjtx-3.0.0-rc1/wsjtx/lib/ft8var/sync8dvar.f90:7:47:
7 | complex
csync1(0:18,32),csync2(32),z1,z2,z3,z4,zt1(0:6),zt2(0:6),zt3(0:6),zt4(0:7),zt5(0:18),
&
--
Warning: ‘IMAGPART_EXPR <z4>’ may be used uninitialized [-Wmaybe-uninitialized]
/builddir/build/BUILD/wsjtx-3.0.0_rc1-build/wsjtx-3.0.0-rc1/wsjtx/lib/ft8var/sync8dvar.f90:7:47:
7 | complex
csync1(0:18,32),csync2(32),z1,z2,z3,z4,zt1(0:6),zt2(0:6),zt3(0:6),zt4(0:7),zt5(0:18),
&
--
Warning: ‘npfx’ may be used uninitialized [-Wmaybe-uninitialized]
/builddir/build/BUILD/wsjtx-3.0.0_rc1-build/wsjtx-3.0.0-rc1/wsjtx/lib/ft8var/packjt77var.f90:1212:9:
1212 | npfx=index(a2,w(1)(1:1))-1
--
Warning: ‘REALPART_EXPR <z4>’ may be used uninitialized [-Wmaybe-uninitialized]
/builddir/build/BUILD/wsjtx-3.0.0_rc1-build/wsjtx-3.0.0-rc1/wsjtx/lib/ft8var/sync8dvar.f90:7:47:
7 | complex
csync1(0:18,32),csync2(32),z1,z2,z3,z4,zt1(0:6),zt2(0:6),zt3(0:6),zt4(0:7),zt5(0:18),
&
--
Warning: ‘IMAGPART_EXPR <z4>’ may be used uninitialized [-Wmaybe-uninitialized]
/builddir/build/BUILD/wsjtx-3.0.0_rc1-build/wsjtx-3.0.0-rc1/wsjtx/lib/ft8var/sync8dvar.f90:7:47:
7 | complex
csync1(0:18,32),csync2(32),z1,z2,z3,z4,zt1(0:6),zt2(0:6),zt3(0:6),zt4(0:7),zt5(0:18),
&
--
Warning: ‘npfx’ may be used uninitialized [-Wmaybe-uninitialized]
/builddir/build/BUILD/wsjtx-3.0.0_rc1-build/wsjtx-3.0.0-rc1/wsjtx/lib/ft8var/packjt77var.f90:1212:9:
1212 | npfx=index(a2,w(1)(1:1))-1
--
Warning: ‘npfx’ may be used uninitialized [-Wmaybe-uninitialized]
/builddir/build/BUILD/wsjtx-3.0.0_rc1-build/wsjtx-3.0.0-rc1/wsjtx/lib/77bit/packjt77.f90:1088:9:
1088 | npfx=index(a2,w(1)(1:1))-1
--
Warning: ‘npfx’ may be used uninitialized [-Wmaybe-uninitialized]
/builddir/build/BUILD/wsjtx-3.0.0_rc1-build/wsjtx-3.0.0-rc1/wsjtx/lib/77bit/packjt77.f90:1088:9:
1088 | npfx=index(a2,w(1)(1:1))-1
thanks & regards
Jaroslav
On Sun, Jan 11, 2026 at 5:13 AM Roger Rehr via wsjt-devel
<[email protected]> wrote:
>
> HI Jaroslav,
>
> I just got a chance to look at this.
>
> Please see if dropping in the attached HRDTransceiver.cpp into your
> source and compiling fixes the issue, and let me know.
>
> Here, with Qt 6.10.0 and gcc 13.1.0 I of course don't see the initial
> error so I can't test the potential fix.
>
> Thanks and 73,
>
> Roger
> W3SZ
>
> On 1/8/2026 10:27 AM, Jaroslav Škarvada via wsjt-devel wrote:
> > Fedora rawhide, gcc-16, it emits some maybe uninitialized warnings,
> > which with hardened build flags (several distros use) may turn into
> > errors:
> > /builddir/build/BUILD/wsjtx-3.0.0_rc1-build/wsjtx-3.0.0-rc1/wsjtx/Transceiver/HRDTransceiver.cpp:1067:86:
> > error: ‘*<unknown>.HRDMessage::size_’ may be used uninitialized
> > [-Werror=maybe-uninitialized]
> > 1067 | if (!write_to_port (reinterpret_cast<char const *>
> > (message.data ()), message->size_))
> > |
> > ~~~~~~~~~^~~~~
> > /builddir/build/BUILD/wsjtx-3.0.0_rc1-build/wsjtx-3.0.0-rc1/wsjtx/Transceiver/HRDTransceiver.cpp:1084:43:
> > error: ‘*(const HRDMessage*)<unknown>.HRDMessage::magic_1_’ may be
> > used uninitialized [-Werror=maybe-uninitialized]
> > 1084 | if (reply->magic_1_value_ != reply->magic_1_ &&
> > reply->magic_2_value_ != reply->magic_2_)
> > | ~~~~~~~^~~~~~~~
> > /builddir/build/BUILD/wsjtx-3.0.0_rc1-build/wsjtx-3.0.0-rc1/wsjtx/Transceiver/HRDTransceiver.cpp:1084:87:
> > error: ‘*(const HRDMessage*)<unknown>.HRDMessage::magic_2_’ may be
> > used uninitialized [-Werror=maybe-uninitialized]
> > 1084 | if (reply->magic_1_value_ != reply->magic_1_ &&
> > reply->magic_2_value_ != reply->magic_2_)
> > |
> > ~~~~~~~^~~~~~~~
> > /builddir/build/BUILD/wsjtx-3.0.0_rc1-build/wsjtx-3.0.0-rc1/wsjtx/Transceiver/HRDTransceiver.cpp:1094:69:
> > error: ‘*(const HRDMessage*)<unknown>.HRDMessage::size_’ may be used
> > uninitialized [-Werror=maybe-uninitialized]
> > 1094 | while (buffer.size () - offsetof (HRDMessage, size_) <
> > reply->size_)
> > |
> > ~~~~~~~^~~~~
> > /builddir/build/BUILD/wsjtx-3.0.0_rc1-build/wsjtx-3.0.0-rc1/wsjtx/Transceiver/HRDTransceiver.cpp:1094:69:
> > error: ‘*(const HRDMessage*)<unknown>.HRDMessage::size_’ may be used
> > uninitialized [-Werror=maybe-uninitialized]
> >
> > I didn't check the code whether it's a real problem, but reporting
> >
> > thanks & regards
> >
> > Jaroslav
> >
> >
> >
> > _______________________________________________
> > wsjt-devel mailing list
> > [email protected]
> > https://lists.sourceforge.net/lists/listinfo/wsjt-devel
> _______________________________________________
> wsjt-devel mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/wsjt-devel
_______________________________________________
wsjt-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/wsjt-devel
