Spectre and Meltdown

IBM workstations/servers/mainframes and operating systems (AIX, OS/2, etc.)
Forum rules
Any posts concerning pirated software or offering to buy/sell/trade commercial software are subject to removal.
astouffer
Posts: 207
Joined: Sat Jan 29, 2011 5:52 pm

Spectre and Meltdown

Unread postby astouffer » Thu Jan 11, 2018 11:52 am

For curiosity's sake I just ran the Spectre and Meltdown test script on my 8203-E4A box and here are the results:
Attachments
spectre.jpg

robespierre
Posts: 1667
Joined: Mon Sep 12, 2011 2:28 pm
Location: Boston

Re: Spectre and Meltdown

Unread postby robespierre » Thu Jan 11, 2018 12:05 pm

I'm not intimately familiar with that tool, but from the messages it appears to be simply scanning for known mitigations and not an actual Proof of Concept attack.
:PI: :O2: :Indigo2IMP: :Indigo2IMP:

User avatar
vishnu
Donor
Donor
Posts: 3361
Joined: Sun Mar 18, 2007 3:25 pm
Location: Minneapolis, Minnesota USA

Re: Spectre and Meltdown

Unread postby vishnu » Thu Jan 11, 2018 10:17 pm

From what I understand (admittedly not much), to use spectre you've got to have local access, and to use meltdown, you can do it over a network if you can figure out a way to crack into a box to install the malware. As of two days ago google says they haven't seen anyone attacking anywhere using either. So for the moment I'm not going to worry about it. I am thinking seriously about changing my home LAN's firewall from Intel/Linux to Sparc/Solaris tho... :shock:
Project:
Temporarily lost at sea...
Plan:
World domination! Or something...

:Tezro: :Octane2:

Shiunbird
Donor
Donor
Posts: 488
Joined: Fri May 06, 2016 1:43 pm
Location: Czech Republic

Re: Spectre and Meltdown

Unread postby Shiunbird » Fri Jan 12, 2018 3:42 am

I'll try on my 9111-285 once I'm back to Europe.

@vishnu: My firewall currently is Dodoid's box. =p I'm not sure whether the Odroid box is affected though. The problem with non-x86 things is the super high energy consumption. Hardware I have laying around, but my power bill is already too high as it is.

My whole setup draws 135W (2 x86 systems, 8 hard drives, a switch, a modem, 2 ARM boxes). If I would go non-x86 (use my POWER5 as a server, for example), I'd go to 500W+. The POWER5 eats 35W on standby!

If you are aware of low power, non-x86, relatively modern stuff, I'd love to hear about.
ImageImage

User avatar
johnnym
Donor
Donor
Posts: 219
Joined: Sun Sep 04, 2016 9:53 pm

Re: Spectre and Meltdown

Unread postby johnnym » Fri Jan 12, 2018 8:53 am

Shiunbird wrote:My whole setup draws 135W (2 x86 systems, 8 hard drives, a switch, a modem, 2 ARM boxes). If I would go non-x86 (use my POWER5 as a server, for example), I'd go to 500W+. The POWER5 eats 35W on standby!

If you are aware of low power, non-x86, relatively modern stuff, I'd love to hear about.


Maybe a little off-topic, so I answered in a new thread...
:Indigo: :Indy: :O2: :Octane: :Octane2: :O200: = :O200: - :O200: = :O200: (O200 cluster w/2 GIGAchannel cabinets)
[ ( hp ) ] 712/80 c3000 (dead) :hpserv: (J5600) c3700 c3750 c8000 rp2470 rp3440 :rx2600: (rx2620) rx2660 rx4640
| d | i | g | i | t | a | l | AXPpci33 AlphaStation 200 AlphaStation 255 PWS 500au AlphaServer DS20E AlphaServer DS25
C O B A L T Qube 2 Qube 3 RaQ RaQ 2 RaQ 4r RaQ XTR

User avatar
ClassicHasClass
Donor
Donor
Posts: 2189
Joined: Wed Jul 25, 2012 7:12 pm
Location: Sunny So Cal
Contact:

Re: Spectre and Meltdown

Unread postby ClassicHasClass » Fri Jan 12, 2018 9:08 pm

The POWER6 is not vulnerable to Meltdown.

It _is_ vulnerable to Spectre. It's not clear whether IBM will offer firmware patches for it yet.

source: tested my own system with the PoC
smit happens.

:Fuel: bigred, 900MHz R16K, 4GB RAM, V12 DCD, 6.5.30
:Indy: indy, 150MHz R4400SC, 256MB RAM, XL24, 6.5.10
:Indigo2IMP: purplehaze, 175MHz R10000, Solid IMPACT
probably posted from Image bruce, Quad 2.5GHz PowerPC 970MP, 16GB RAM, Mac OS X 10.4.11
plus IBM POWER6 p520 * Apple Network Server 500 * RDI PrecisionBook * BeBox * Solbourne S3000 * Commodore 128 * many more...

User avatar
uunix
Donor
Donor
Posts: 1989
Joined: Sun Mar 27, 2011 12:48 pm
Location: Stourbridge / England / UK

Re: Spectre and Meltdown

Unread postby uunix » Sat Jan 13, 2018 1:08 am

ClassicHasClass wrote:The POWER6 is not vulnerable to Meltdown.

It _is_ vulnerable to Spectre. It's not clear whether IBM will offer firmware patches for it yet.

source: tested my own system with the PoC

Hurrah for my POWER5 then.
-----------------------------------------------------------------------
Hey Ho! Pip & Dandy!
:Tezro: :Octane2: :O2: :Indigo: :Indy:
-----------------------------------------------------------------------

User avatar
johnnym
Donor
Donor
Posts: 219
Joined: Sun Sep 04, 2016 9:53 pm

Re: Spectre and Meltdown

Unread postby johnnym » Sat Jan 13, 2018 3:38 am

uunix wrote:
ClassicHasClass wrote:The POWER6 is not vulnerable to Meltdown.

It _is_ vulnerable to Spectre. It's not clear whether IBM will offer firmware patches for it yet.

source: tested my own system with the PoC

Hurrah for my POWER5 then.

Not necessarily: POWER4 (and then most likely also PowerPC 970 and derivatives) and POWER5 make use of out-of-order and speculative execution, whereas POWER6 uses in-order execution.
:Indigo: :Indy: :O2: :Octane: :Octane2: :O200: = :O200: - :O200: = :O200: (O200 cluster w/2 GIGAchannel cabinets)
[ ( hp ) ] 712/80 c3000 (dead) :hpserv: (J5600) c3700 c3750 c8000 rp2470 rp3440 :rx2600: (rx2620) rx2660 rx4640
| d | i | g | i | t | a | l | AXPpci33 AlphaStation 200 AlphaStation 255 PWS 500au AlphaServer DS20E AlphaServer DS25
C O B A L T Qube 2 Qube 3 RaQ RaQ 2 RaQ 4r RaQ XTR

User avatar
uunix
Donor
Donor
Posts: 1989
Joined: Sun Mar 27, 2011 12:48 pm
Location: Stourbridge / England / UK

Re: Spectre and Meltdown

Unread postby uunix » Sat Jan 13, 2018 5:36 am

johnnym wrote:
uunix wrote:
ClassicHasClass wrote:The POWER6 is not vulnerable to Meltdown.

It _is_ vulnerable to Spectre. It's not clear whether IBM will offer firmware patches for it yet.

source: tested my own system with the PoC

Hurrah for my POWER5 then.

Not necessarily: POWER4 (and then most likely also PowerPC 970 and derivatives) and POWER5 make use of out-of-order and speculative execution, whereas POWER6 uses in-order execution.

Ahh maaaaaan! :x
-----------------------------------------------------------------------
Hey Ho! Pip & Dandy!
:Tezro: :Octane2: :O2: :Indigo: :Indy:
-----------------------------------------------------------------------

astouffer
Posts: 207
Joined: Sat Jan 29, 2011 5:52 pm

Re: Spectre and Meltdown

Unread postby astouffer » Sat Jan 13, 2018 9:57 am

Someone was telling me even the later SPARC cpus and even some Alphas did speculative execution. Itanium might be safe :lol:

User avatar
ClassicHasClass
Donor
Donor
Posts: 2189
Joined: Wed Jul 25, 2012 7:12 pm
Location: Sunny So Cal
Contact:

Re: Spectre and Meltdown

Unread postby ClassicHasClass » Sat Jan 13, 2018 9:07 pm

johnnym wrote:
uunix wrote:
ClassicHasClass wrote:The POWER6 is not vulnerable to Meltdown.

It _is_ vulnerable to Spectre. It's not clear whether IBM will offer firmware patches for it yet.

source: tested my own system with the PoC

Hurrah for my POWER5 then.

Not necessarily: POWER4 (and then most likely also PowerPC 970 and derivatives) and POWER5 make use of out-of-order and speculative execution, whereas POWER6 uses in-order execution.


Yes. The 970 is definitely vulnerable, so the POWER4 is also almost certainly vulnerable, and if the POWER6 is, then I would be surprised if the POWER5 weren't. Even though the POWER6 is in-order, it still speculates; out-of-order execution only has to do with how instructions are issued.
smit happens.

:Fuel: bigred, 900MHz R16K, 4GB RAM, V12 DCD, 6.5.30
:Indy: indy, 150MHz R4400SC, 256MB RAM, XL24, 6.5.10
:Indigo2IMP: purplehaze, 175MHz R10000, Solid IMPACT
probably posted from Image bruce, Quad 2.5GHz PowerPC 970MP, 16GB RAM, Mac OS X 10.4.11
plus IBM POWER6 p520 * Apple Network Server 500 * RDI PrecisionBook * BeBox * Solbourne S3000 * Commodore 128 * many more...

User avatar
skywriter
Posts: 3303
Joined: Fri Mar 14, 2003 5:22 am
Location: Trump-proof bunker
Contact:

Re: Spectre and Meltdown

Unread postby skywriter » Wed Jan 31, 2018 6:22 pm

Any performance processor used branch prediction, speculative execution, and out of order execution. Before attacks were based on them they were the best tools to reduce memory latency - the major performance inhibiter in any high performance system.

Assuming google could even detect this, they’re never going to admit it. A vendors first response is always “none of our other customers have this problem”. It’s like “I never had any contact with Russia”, you know how that turned out - EVERYONE had contact with Russia. Lies make the world go round.
:Skywriter:

DECUS Member 368596


Return to “IBM”

Who is online

Users browsing this forum: No registered users and 3 guests