On 6/17/2024 4:52 PM, Kees Cook wrote:
On Mon, Jun 17, 2024 at 04:41:30PM -0700, Roman Kisel wrote:
Missing, failed, or corrupted core dumps might impede crash
investigations. To improve reliability of that process and consequently
the programs themselves, one needs to trace the path from producing
a core dumpfile to analyzing it. That path starts from the core dump file
written to the disk by the kernel or to the standard input of a user
mode helper program to which the kernel streams the coredump contents.
There are cases where the kernel will interrupt writing the core out or
produce a truncated/not-well-formed core dump.
Hm, I'm all for better diagnostics, but they need to be helpful and not
be a risk to the system. All the added "pr_*()" calls need to use the
_ratelimited variant to avoid a user inducing massive spam to the system
logs. And please standardize the reporting to include information about
the task that is dumping. Otherwise the logging isn't useful for anyone
reading it. Something that includes pid and task->comm at the very
least. :)
Appreciate your suggestions very much! Rate-limiting has definitely
slipped off my mind, my bad. Will also fix the reporting format to make
it useful.
For example, see report_mem_rw_reject() in
https://lore.kernel.org/lkml/20240613133937.2352724-2-adrian.ratiu@xxxxxxxxxxxxx/
Thanks, that's awesome!
-Kees
--
Thank you,
Roman
