[Drbd-dev] [PATCH] drbd: Fix kernel_sendmsg() usage

Christoph Lechleitner christoph.lechleitner at iteg.at
Tue Nov 8 15:03:08 CET 2016


Am 2016-11-08 um 14:43 schrieb Lars Ellenberg:
> From 3a5859e696178e31a25e65de58c461046fc52beb Mon Sep 17 00:00:00 2001
> From: Richard Weinberger <richard at nod.at>
> Date: Tue, 8 Nov 2016 11:43:09 +0100
> Subject: [PATCH] drbd: Fix kernel_sendmsg() usage - potential NULL deref
> drbd: Fix kernel_sendmsg() usage - potential NULL deref

> Even with this error exposed, to trigger the bug, we would need to have
> an environment (config or otherwise) causing us to not use sendpage()
> for larger transfers, a flaky connection, and have it fail "just at the
> right time".  Apparently that was unlikely enough for most, so this went
> unnoticed for years.

Our drbd configuration was created some 8 years ago.
Maybe I should have read more migration tips when upgrading again and
again, sorry ;-)

But a 30cm Cat6 cable directly connecting 2 dedicated ethernet ports
should not match the term "flaky connection".

FYI:
I co-own the company that hired Richard to track down this bug, that
repeatedly (~15 times) forced us to hard-reset servers hosting tens of
virtual root servers for our customers.

Regards, Christoph Lechleitner



More information about the drbd-dev mailing list