Open LewisR opened 2 years ago
BTW, these reports all ended up in the root of C: (boot volume) and not ticked away where expected. Not sure what's up with that.
Judging from the report, it looks like DCCLIENT frees some memory block twice. The reason you could not see it before is because the .TRP creation in LIBC is very new. 0.1.7 and earlier didn't have it. W/o that you'd only have a console output from LIBC with panic and this you will not see for background or daemon processes.
Re reports ending up in the root of C: The boot drive is chosen as a target dir for them when:
%LOGINFO%/app
dir does not exist and can't be created, or%UNIXROOT%/var/log/app
dir does not exist and can't be created, orSo I guess your system triggers one of these conditions.
Regarding reports ending up in the root of the boot drive. It could happen even on a properly configured system if the child process is started with the environment cleared up (so that LOGFILES and UNIXROOT are removed too). I bet it's the case with DHCLIENT). I've fixed it because it's annoying indeed. A commit will follow.
Since upgrading to libc-0.1.9-1, I've noticed several exceptq reports while running dhclient-3.1-3. The first was a panic, then later, different libc failures. I've attached three exceptq reports (I have more, but perhaps they are not necessary). 613aae3e-008e_01-DHCLIENT-exceptq.txt 613d13ab-0086_01-DHCLIENT-exceptq.txt 61300c7a-0088_01-DHCLIENT-exceptq.txt