[RndTbl] Horrific CPU flaws (Meltdown / Spectre)
athompso at athompso.net
Thu Jan 4 18:17:42 CST 2018
I don't know what's being fixed, but there are already microcode updates available for some Haswell Xeon E5 chips, according to OVH's (probably the 4th or 5th largest consumer of them) CEO on Twitter.
In same tweet, said older chips (no idea how old) would get updates within weeks.
On January 4, 2018 6:02:50 PM CST, Trevor Cordes <trevor at tecnopolis.ca> wrote:
>On 2018-01-04 Gilles Detillieux wrote:
>> I just updated our Scientific Linux systems, which means the RHEL
>> updates are out too, and presumably CentOS updates are too or will be
>> shortly. These updates included firmware/microcode packages, which
>Interesting on the fw/microcode as what I read indicated this can't be
>fixed in fw/mc. Though perhaps fw/mc changes could help in some way,
>in tandem with the OS.
>> I'm assuming are loaded on reboot as well. Some of the reports I read
>> suggested that you'd need to reflash your BIOS/UEFI firmware once the
>> PC manufacturers release these updates. Are those reports in error,
>> confusing the two types of firmware, or are we going to have to hunt
>> down PC or mobo-specific firmware updates for this whole debacle too?
>Again, from what I read, this won't be a fw thing, so I don't think
>we'll see the above scenario. If we do -- if fix=fw+os -- then we're
>in for a world of hurt because most mfr's stop updating fw after their
>cycle, usually max of 3 years, some much less. That would leave a
>whole slew of mobos unfixable. Good way for the industry to force a
>rapid upgrade cycle!!! That might really get Intel in trouble,
>though... $$$ lawsuits.
>Of course, it's early days, so I could be way offbase, and even if I'm
>correct now, my info may be wrong in a few days/weeks.
>Roundtable mailing list
>Roundtable at muug.ca
Sent from my Android device with K-9 Mail. Please excuse my brevity.
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Roundtable