[SyncEvolution] Syncing to Memotoo w SyncEvolution post-Fedora 37 upgrade (fwd)

[Milan Crha]

On Wed, 2023-02-15 at 17:05 -0500, Max Pyziur wrote:
> Core was generated by `/usr/libexec/syncevo-dbus-helper --dbus-
> verbosity 
> 3'.
> Program terminated with signal SIGSEGV, Segmentation fault.
> #0  0x00007f28c1252d70 in ?? () from /lib64/libcurl.so.4
> [Current thread is 1 (Thread 0x7f28c082c1c0 (LWP 65192))]
> (gdb)

	Hi,
I saw in my tests that it also prints the backtrace, but it did not
happen for you - or at least it did not print a complete backtrace.
Maybe I'm wrong (incorrect observation on my side). Anyway, when you
re-run that:

   $ coredumpctl debug 65192

ignore everything down to the place quoted above, then execute in the
gdb prompt:

   (gdb) bt
   (gdb) t a a bt

to print the current backtrace and then backtraces for each thread.

It's shown there the crash happens in the /lib64/libcurl.so.4 . I do
not understand why the debuginfod could not find the debuginfo package
for it. Maybe you updated the libcurl meanwhile? Even then the
debuginfo should be available on the servers.

You can try one thing: if there's still running syncevo-dbus-helper in
the bakcground, then kill it. After that run it under gdb:

   $ gdb /usr/libexec/syncevo-dbus-helper --ex "r --dbus-verbosity 3"

and then wait for the debuginfo download and then reproduce the crash.
The gdb prompt will stop at that place (similar to the quoted part
above) and you can run:

   (gdb) bt
   (gdb) t a a bt

to get the baktraces from the current session, with the current
packages and debug infos.

It can be SyncEvolution passes some odd pointer to the libcurl, or
there is a bug in the libcurl. It's hard to tell right now.

By the way, did you fill an ABRT bug already? If so, do you have a link
to it, please?

	Bye,
	Milan