On 19.03.22 16:54, Jeremy Harris via Exim-users wrote: > To work towards a fix will however need better info. A first cut
> would be just a "-ggdb -O0" build, and inspect the stacktrace for
> obvious null-pointers in function args.
We created a build with the mentioned flags. But with this version, no
segfaults are directly reported and no core file created.
Instead the following is shown:
LOG: MAIN PANIC DIE
bad memory allocation requested (-83643 bytes) at string_localpart_utf8_to_alabel 146
LOG: MAIN PANIC
Delivery status for ������@somehost.net: got 0 of 7 bytes (pipeheader) from transport process 108 for transport smtp
Compared to:
LOG: MAIN PANIC
SIGSEGV (maybe attempt to write to immutable memory)
LOG: MAIN PANIC
Delivery status for ������@somehost.net: got 0 of 7 bytes (pipeheader) from transport process 92288 for transport smtp
I guess that's no big help.
If desired, I can provide the spool files causing the segfaults.
