This warning only applies to Linux on Ubuntu on AMD 64-bit chips, as far as I know so far. (minority of a minority of a minority)

Late yesterday installed latest Ubuntu 12.04 kernel update

Linux saguaro 3.13.0-59-generic #98-Ubuntu SMP Fri Jul 24 21:05:26 UTC 2015 x86_64 x86_64 x86_64 GNU/Linux

Installed this Ubuntu kernel update on one Intel and one AMD box. Seems good on Intel, but the AMD box started crashing CPDN every model, see http://climateapps2.oerc.ox.ac.uk/cpdnboinc/results.php?hostid=1310137 with the uninformative "process got signal 11" as all can see.
Wasted a few models, but the only newly downloaded one that didn't start before I rebooted to the old

Linux coffea 3.13.0-58-generic #97-Ubuntu SMP Wed Jul 8 02:56:15 UTC 2015 x86_64 x86_64 x86_64 GNU/Linux

seems to run OK. Will have to wait out the timeout after going crashmaster before get new models on the AMD box, and NO, I'm not going to try this Ubuntu kernel update on my other AMD boxes for at least a few days.

SO, I'm thinking, and I remember some talk on the Linux-kernel list about "fixing" some obscure "x32 on x64 on AMD only" and Linus T saying "we got that wrong"

So, possibly, Ubuntu backported a borken kernel fix to (stable) Trusty.

Anyhow, if anyone watching has UBU (especially) 14.04 watch out for broken kernel update.

PS - if you compile your own kernels (superminority) (and -- I don't anymore) - this should be fixed in in 4.2.rc4 (or earlier) (When the fix of the fix hits the distros and backports - no idea)

---

Well I can confirm that it seems OK on Intel chips.

Ubuntu have published the fix. 3.13.0-61-generic #100-Ubuntu is available. Seems OK on my AMD machines (64-bit)

---

I am running the 4.1.12-1-default kernel on my AMD A10-6700 CPU. The kernel is the default of the Leap 42.1 SuSE release. So far it only runs in a Virtual machine on my Windows 10 PC, and crunches vLHC@home.
Tullio

---