Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

exoticSpice

Suspended
Jan 9, 2022
1,242
1,952
Any attack that gains kernel access is potentially bad.

What's popular with real security professionals is chaining multiple attacks. So, combining zero-day remote exploit + PACMAN kernel attack. You can get a glimpse of chaining attacks at Pwn2Own. Time will tell if it's as bad as Pegasus.

https://www.zerodayinitiative.com/blog/2022/5/18/pwn2own-vancouver-2022-the-results

https://www.zerodayinitiative.com/blog/2021/4/2/pwn2own-2021-schedule-and-live-results
Yep it could also effect other ARM chips. Hopefully they test other ARM chips too. But I think it would take 2 years for those chips to come ro market
 
Last edited:

mi7chy

macrumors G4
Oct 24, 2014
10,622
11,294
This is a hardware hacking attack requires physical access.

https://pacmanattack.com/

Does this attack require physical access?​

Nope! We actually did all our experiments over the network on a machine in another room. PACMAN works just fine remotely if you have unprivileged code execution.

Yep it also effects other ARM chips. Hopefully it gets patched from all ARM chips on a hardware soon. But I think it would take 2 years for those chips to come ro market

That's just an assumption. Where does it say that on web site or white paper that they performed successful proof-of-concept against other SoCs than Apple M1? All it says is pointer authentication was introduced with ARM v8.3 but doesn't say if it's proven on other ARM implementations from Qualcomm, Samsung, Mediatek, etc. or if it affects newer ARM extensions up to v8.8 and v9.3.
 
  • Wow
Reactions: addamas

exoticSpice

Suspended
Jan 9, 2022
1,242
1,952

Should I be worried?​

As long as you keep your software up to date, no. PACMAN is an exploitation technique- on its own it cannot compromise your system. While the hardware mechanisms used by PACMAN cannot be patched with software features, memory corruption bugs can be.

https://pacmanattack.com/ - Source

There we go keep your software updated.
 

exoticSpice

Suspended
Jan 9, 2022
1,242
1,952
That's just an assumption. Where does it say that on web site or white paper that they performed successful proof-of-concept against other SoCs than Apple M1? All it says is pointer authentication was introduced with ARM v8.3 but doesn't say if it's proven on other ARM implementations from Qualcomm, Samsung, Mediatek, etc. or if it affects newer ARM extensions up to v8.8 and v9.3.

"and the MIT team believes the vulnerability could impact future Arm mobile devices, and likely even future Arm desktop PCs, if it isn't mitigated in future architectures. "Any chip that uses speculative execution to evaluate and operate on pointer authentication signed pointers (and handles nested mispredicts eagerly) could potentially be vulnerable to PACMAN," said Joseph Ravichandran, a researcher with the MIT team. That means this could possibly impact chips from other Arm vendors that support pointer authentication, such as Qualcomm and Samsung, but those chips haven't been tested yet."

Those other chips have not been tested it but they could be effected if they have support for pointer authentication.
 

throAU

macrumors G3
Feb 13, 2012
9,199
7,354
Perth, Western Australia
The article has been amended as follows:
"
The real-world risk is low because PACMAN requires physical access to a Mac; the attack cannot be carried out remotely.
Macworld stated that “Because PACMAN requires a hardware device, a hacker has to have physical access to a Mac, which limits how a PACMAN can be executed,” but the research team advises me that this is incorrect. No physical access is needed.
"

Also, it states that all ARM chips are affected (not just the M1), and if this is true it means that potentially all Apple devices could be compromised (as well as those of many other vendors).
Cheers for the update.

Still,

  1. this is a security mechanism that x86 machines don't even have
  2. it requires compromise of the machine via other method first. i.e., your machine has to be compromised via another exploit to even get at this protection mechanism.
As always, stay up to date.

No, the sky is not falling.
 

throAU

macrumors G3
Feb 13, 2012
9,199
7,354
Perth, Western Australia
You obviously have not seen Mission Impossible and unaware how hackers can physically access your machine without you knowing

Obviously that is possible.

What i'm saying is that if they have the ability to do that, this exploit existing makes no difference. Like i said, beating the password/touchid/key out of you with a $5 wrench is far easier than this, most likely (assuming that physical access is required). i.e., you wouldn't bother wasting your time with this exploit, if you're the type of organization likely in a position to make use of it.
 

Unregistered 4U

macrumors G4
Jul 22, 2002
10,610
8,628
It's a vulnerability, sure - but not one that has any practical application that couldn't just be achieved by more conventional means. I'm sure apple will fix it in the M3 if it hasn't been fixed in M2, but in the mean-time there's little to worry about.
I actually liked how a couple stories I linked to indicated up front that this requires some “greater than normal” effort to set up.
 

Unregistered 4U

macrumors G4
Jul 22, 2002
10,610
8,628

Does this attack require physical access?​

Nope! We actually did all our experiments over the network on a machine in another room. PACMAN works just fine remotely if you have unprivileged code execution.
That’s a BIT of a stretch. It was THEIR machine that was in the other room, not a stranger’s machine where they didn’t know where it was, what it’s IP address was, what it’s security posture was, etc. So, while the attack might not require physical access, SETTING up the target to be pwned certainly would.

There are a lot of exploits that don’t require physical access in the same way. I can just have access to a monkey wrench and to someone that the owner of the system is fond of and I’d likely be able to obtain critical information from that computer in the other room WITHOUT PHYSICAL ACCESS. :)
 

Freeangel1

Suspended
Jan 13, 2020
1,191
1,755
always keep an Intel chip Mac machine around if you're worried about security on ARM.

it will probably be in M2 chips too. Maybe might be fixed by 3nm chips next year.
 
  • Haha
Reactions: jdb8167

h0ndaf4n

Suspended
Jun 1, 2022
14
8
Here's what i found with 2 seconds of google:



And here's the summary:


And here's my comments as a network security guy:

If someone has unfettered physical access to your machine, you're boned.

This is a hardware hacking attack requires physical access.

The short version is this: nothing to worry about unless you leave your machine unattended where bad guys can get access to it - and you're interesting enough for someone to go to the trouble to do this.

And even then, there's plenty of other things they can do with physical access. The likelyhood of this being used is probably limited to nation-state sponsored attacks (NSA/FSB and the like), where the attacker could just as easily beat your password/touchId/etc. out of you with a pipe. Or drop a camera in your home to record you typing your password, etc.

It's a vulnerability, sure - but not one that has any practical application that couldn't just be achieved by more conventional means. I'm sure apple will fix it in the M3 if it hasn't been fixed in M2, but in the mean-time there's little to worry about.
Macworld stated that “Because PACMAN requires a hardware device, a hacker has to have physical access to a Mac, which limits how a PACMAN can be executed,” but the research team advises me that this is incorrect. No physical access is needed.
 

jav6454

macrumors Core
Nov 14, 2007
22,303
6,264
1 Geostationary Tower Plaza
Macworld stated that “Because PACMAN requires a hardware device, a hacker has to have physical access to a Mac, which limits how a PACMAN can be executed,” but the research team advises me that this is incorrect. No physical access is needed.
They also advise that a certain bug must be present in memory for it to work. Lots of "Ifs" in this flaw.
 

eltoslightfoot

macrumors 68030
Feb 25, 2011
2,547
3,099
The intel architecture has multiple hardware flaws, so it isn’t like we can just avoid by going intel…
 

Cognizant.

Suspended
May 15, 2022
427
723
The intel architecture has multiple hardware flaws, so it isn’t like we can just avoid by going intel…
Intel has had massive security holes that are constantly being patched ... and unfortunately the patches degrade performance. I'm not sure if these apply to the brand new Intel chips, but both AMD and Intel have been plagued with this nonsense lately.
 
  • Like
Reactions: eltoslightfoot
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.