Fix function call to rec_backtrace_ctx when using USE_SYSTEM_LIBUNWIND=1

[rbvermaa]

When building julia-1.5.0-rc1 from source, we ran into the following error. It looks like one call to rec_backtrace_ctx was not changed when an extra arg was added.

Not 100% this patch is correct (whether to pass 1 for the lockless argument), but it does fix the compilation error, and is similar to the other branch of the ifndef.

In file included from /private/tmp/nix-build-julia-1.5.0-rc1-patched.drv-0/source/src/signal-handling.c:117:
In file included from ./signals-unix.c:218:
./signals-mach.c:462:135: error: too few arguments to function call, expected 5, have 4
                bt_size_cur += rec_backtrace_ctx((jl_bt_element_t*)bt_data_prof + bt_size_cur, bt_size_max - bt_size_cur - 1, uc, NULL);
                               ~~~~~~~~~~~~~~~~~                                                                                      ^
./julia_internal.h:797:1: note: 'rec_backtrace_ctx' declared here
size_t rec_backtrace_ctx(jl_bt_element_t *bt_data, size_t maxsize, bt_context_t *ctx,
^
1 error generated.
make[1]: *** [Makefile:161: signal-handling.o] Error 1

[NHDaly]

Note: this also needs 1.5 backport.

[Sacha0]

It looks like this patch is no longer correct after #37002. Whether the issue persists I do not know, but will try to find out :). Edit: Taking another look at the issue description, this issue of course no longer exists as the relevant argument is gone again 😄.

[Sacha0]

This issue has been obviated on master. Perhaps close? :)

[NHDaly]

@Sacha0 would it still make sense to backport this to 1.5 though?

[Sacha0]

I'm operating under the assumption that with 1.6 inbound, there will be no more 1.5 releases. But if there are additional 1.5 releases, agreed! :)