SIGSEGV when moving cursor back to server/pressing any key

[MultisampledNight]

Waynergy crashes with a segmentation fault on moving the cursor back to the server (in this case, barrier), or pressing any key on the server while the cursor is on waynergy's side. I don't know if these are related in any way, but following the procedure in #16 I built in debug-mode and using bt in GDB I produced these two backtraces.

That one is when the cursor moves back to the server:

#0  0x000055ec59956a26 in wlKeyReleaseAll (ctx=0x55ec59978ee0 <wlContext>) at ../src/wl_input.c:134
#1  0x000055ec5995a52f in syn_active_cb (cookie=0x55ec59978fe0 <synNetContext>, active=false) at ../src/main.c:122
#2  0x000055ec5996441b in sProcessMessage (context=0x55ec599689c0 <synContext>, msg=0x7ffdef1644c0) at ../src/uSynergy.c:424
#3  0x000055ec599656de in sUpdateContext (context=0x55ec599689c0 <synContext>) at ../src/uSynergy.c:759
#4  0x000055ec599658d3 in uSynergyUpdate (context=0x55ec599689c0 <synContext>) at ../src/uSynergy.c:823
#5  0x000055ec5995fdb3 in netPoll (snet_ctx=0x55ec59978fe0 <synNetContext>, wl_ctx=0x55ec59978ee0 <wlContext>)
    at ../src/net.c:214
#6  0x000055ec5995ad6e in main (argc=11, argv=0x7ffdef1647c8) at ../src/main.c:315

And that one is when pressing any key server-side:

#0  0x000055db82cc282f in wlKey (ctx=0x55db82ce4ee0 <wlContext>, key=133, state=1) at ../src/wl_input.c:110
#1  0x000055db82cc6269 in syn_key_cb (cookie=0x55db82ce4fe0 <synNetContext>, key=133, mod=0, down=true, repeat=false)
    at ../src/main.c:71
#2  0x000055db82ccf7bf in sSendKeyboardCallback (context=0x55db82cd49c0 <synContext>, key=133, modifiers=0, down=true,
    repeat=false) at ../src/uSynergy.c:214
#3  0x000055db82cd0896 in sProcessMessage (context=0x55db82cd49c0 <synContext>, msg=0x7ffe175ac640) at ../src/uSynergy.c:494
#4  0x000055db82cd16de in sUpdateCon#0  0x000055ec59956a26 in wlKeyReleaseAll (ctx=0x55ec59978ee0 <wlContext>) at ../src/wl_input.c:134
#1  0x000055ec5995a52f in syn_active_cb (cookie=0x55ec59978fe0 <synNetContext>, active=false) at ../src/main.c:122
#2  0x000055ec5996441b in sProcessMessage (context=0x55ec599689c0 <synContext>, msg=0x7ffdef1644c0) at ../src/uSynergy.c:424
#3  0x000055ec599656de in sUpdateContext (context=0x55ec599689c0 <synContext>) at ../src/uSynergy.c:759
#4  0x000055ec599658d3 in uSynergyUpdate (context=0x55ec599689c0 <synContext>) at ../src/uSynergy.c:823
#5  0x000055ec5995fdb3 in netPoll (snet_ctx=0x55ec59978fe0 <synNetContext>, wl_ctx=0x55ec59978ee0 <wlContext>)
    at ../src/net.c:214
#6  0x000055ec5995ad6e in main (argc=11, argv=0x7ffdef1647c8) at ../src/main.c:315text (context=0x55db82cd49c0 <synContext>) at ../src/uSynergy.c:759
#5  0x000055db82cd18d3 in uSynergyUpdate (context=0x55db82cd49c0 <synContext>) at ../src/uSynergy.c:823
#6  0x000055db82ccbdb3 in netPoll (snet_ctx=0x55db82ce4fe0 <synNetContext>, wl_ctx=0x55db82ce4ee0 <wlContext>)
    at ../src/net.c:214
#7  0x000055db82cc6d6e in main (argc=11, argv=0x7ffe175ac948) at ../src/main.c:315

This is an assumption by someone who barely used C in any true fashion, but both seem to involve ctx being NULL, so I figured both backtraces belong in one issue.

EDIT: I'm running sway on Arch Linux, launching waynergy through

waynergy --backend wlr --host spacestation --port 24800 --enable-crypto --enable-tofu --name spacepod

[MultisampledNight]

Just ran waynergy through rr record and I correct myself, no, ctx is not NULL, it's a perfectly fine struct. What's more of a matter is that wlContext->input.key_press_state is indeed NULL:

Breakpoint 1, syn_active_cb (cookie=0x55f04d770fe0 <synNetContext>, active=true) at ../src/main.c:121
121		if (!active) {
(rr) p wlContext
$1 = {registry = 0x55f04ee30c50, display = 0x55f04ee2cab0, seat = 0x55f04ee32b70, input = {state = 0x55f04ee33700,
    key_press_state = 0x0, key_count = 255, xkb_key_offset = 0, xkb_ctx = 0x55f04ee31f60, xkb_map = 0x55f04ee33a20,
    xkb_state = 0x55f04ee46de0, raw_keymap = 0x0, wl_ctx = 0x55f04d770ee0 <wlContext>,
    mouse_rel_motion = 0x55f04d74f49c <mouse_rel_motion>, mouse_motion = 0x55f04d74f528 <mouse_motion>,
    mouse_button = 0x55f04d74f5d2 <mouse_button>, mouse_wheel = 0x55f04d74f6aa <mouse_wheel>, key = 0x55f04d74f3ac <key>,
    key_map = 0x55f04d74f31e <key_map>}, uinput_fd = {-1, 0}, keyboard_manager = 0x55f04ee31920,
  pointer_manager = 0x55f04ee319c0, fake_input = 0x0, output_manager = 0x55f04ee33890, outputs = 0x55f04ee33990,
  idle_manager = 0x55f04ee33930, idle_timeout = 0x55f04ee67680, width = 1128, height = 752, epoch = 0,
  on_output_update = 0x55f04d7523c4 <wl_output_update_cb>}

[r-c-f]

That is now corrected indeed.

[MultisampledNight]

Waynergy works great now, thanks! Only thing that's on my side is that the keymap seems to get lost, but that's not your problem. Thanks again!

EDIT: And I figured out the keymap, too. Now my system successfully gained another screen.