[Not Solved] The “Meltdown” bug is hitting Intel

[Pat D]

How serious is this story?

https://qz.com/1171391/the-intel-int...md-is-soaring/

Looking to replace Intel MB, but I'm now concerned that I should maybe switch to a new MB and CPU from AMD?

Any thoughts?

[Pjotr]

Considerably cheaper to wait for the redeeming update.

[Pat D]

This article about

"the redeemimg update" https://lwn.net/Articles/738975/

seems to suggest up to a 30% performance hit is possible - this is what's worrying me about it.

[Pjotr]

Well, possibly even that update can be improved to eliminate that performance hit.... It's not as if there's a lack of pressure on Intel right now.

[ClixTrix]

Any thoughts?

My sig should explain where my thoughts operate.

[ibbill]

Google researchers said one of the "serious security flaws", dubbed "Spectre", was found in chips made by Intel, AMD and ARM.

The other, known as "Meltdown" affects Intel-made chips.

http://www.bbc.com/news/technology-42561169

[Pjotr]

The rescue operation is ongoing:
viewtopic.php?f=58&t=260764&start=40#p1409187

[karlchen]

<moderator on>
Changed the [solved] in the thread title to [Not solved]. The Intel / AMD / ARM hardware flaws have not been solved.
The software workarounds are in the development stage, no more.
The Spectre and Meltdown vulnerabilities are far far away from being solved.
Therefore marking this thread as [solved] is inadequate.
</moderator off>

[Pat D]

FWIW, I marked it as solving MY problem, with the referral to the proper thread (which may not have been done well.)

Mini-rant:
I like this forum. You're so much better than what I call "Orange Ubuntu" , which has only had it's "crimes" ameliorated by the newer and hugely more friendly "Green Ubuntu" for Ubuntu Mate.

I'm quite sure that critical and extreme nit-picking attitude "Orange Ubuntu " forums chased away 2 to 3 % of the PC market from Linux. If not more.
Seriously. Those people are such (can't say this on a family show)-holes. I only came back because I'm a stubborn one of those (can't say this on a family show)-holes.

Cheers.

[jimallyn]

It's not as if there's a lack of pressure on Intel right now.

Are there people at Intel's door with torches and pitchforks? There probably should be.

[brianshel]

Quite prepared to recognise that I may be naïve, but why does this matter provided that you have a firewall active on your computer? I have my ufw firewall set up so that only specified machines on my local network can gain access. Probably I'm missing something, so perhaps someone could enlighten me - not too complicated please!

[Pjotr]

Firefox 57.0.4 contains a (partial) fix for this:
https://www.mozilla.org/en-US/firefox/5 ... ox-browser

[harryw]

Quite prepared to recognise that I may be naïve, but why does this matter provided that you have a firewall active on your computer? I have my ufw firewall set up so that only specified machines on my local network can gain access. Probably I'm missing something, so perhaps someone could enlighten me - not too complicated please!

Mozilla has shown a proof of concept that this can be exploited via JavaScript. So if you venture to the wrong webpage without a script blocker - poof!

https://www.bleepingcomputer.com/news/s ... e-attacks/

Therefore, keep a lookout for browser updates (see above post).

[ClixTrix]

Kernel 4.14.12 was released at kernel.org with the following update for those with AMD processors, i.e. that aren't subject to Meltdown. Note, the Meltdown fix was added at kernel.org to 4.14.11, so this follow-up is definitely going to help the AMD users avoid the slowdown caused by the Intel KPTI fix.

Code: Select all

commit 151d7039757b71ebd9d170af0944562f51149372
Author: Tom Lendacky <thomas.lendacky@amd.com>
Date:   Tue Dec 26 23:43:54 2017 -0600

    x86/cpu, x86/pti: Do not enable PTI on AMD processors
    
    commit 694d99d40972f12e59a3696effee8a376b79d7c8 upstream.
    
    AMD processors are not subject to the types of attacks that the kernel
    page table isolation feature protects against.  The AMD microarchitecture
    does not allow memory references, including speculative references, that
    access higher privileged data when running in a lesser privileged mode
    when that access would result in a page fault.
    
    Disable page table isolation by default on AMD processors by not setting
    the X86_BUG_CPU_INSECURE feature, which controls whether X86_FEATURE_PTI
    is set.
    
    Signed-off-by: Tom Lendacky <thomas.lendacky@amd.com>
    Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
    Reviewed-by: Borislav Petkov <bp@suse.de>
    Cc: Dave Hansen <dave.hansen@linux.intel.com>
    Cc: Andy Lutomirski <luto@kernel.org>
    Link: https://lkml.kernel.org/r/20171227054354.20369.94587.stgit@tlendack-t1.amdoffice.net
    Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

[WiltshireJon]

It's going to be interesting to see what this mess will do to the retail computer market for a while.
I can't see anyone who has any knowledge of it buying new in the near future unless to replace something that dies. I've been thinking about buying a replacement for an aged home server, but I think I'll patch and wait until new processors come out, hoping the old one holds together. Which could be years. There could be a whole lot of very unhappy manufacturers and retailers (and maybe some cheap deals if you think the patches will work as advertised).

[Mattyboy]

It's going to be interesting to see what this mess will do to the retail computer market for a while.

I've been keeping an eye on processor prices recently simply because I was 'this close' to ordering parts for a new rig.. looks like I got lucky and will hold off for next gen... intel chips have dropped here in the UK, depending on version, anything from 15% to 30% on coffee lake. They're practically giving away older models. AMD have also dropped slightly but this probably is due to the market.

[Madmogone]

Well bugger me, (SORRY FOR THE FRENCH) as they say here in blighty. About 10 months ago I built a firewall/router computer using an INTEL cpu (J1900C,2.0GHz,quad core) running Smoothwall Express.The first time I have ever used an INTEL cpu and they screwed me like a lot of other people out there. I normally only use AMD cpu,s.There,s a lot of angry people out there and I am one of them.Will NEVER use them again.It makes you wonder whether INTEL instead of working to MOORES LAW, was instead working to MURPHYS LAW .Any how thats my rant over with for today, there may be many more in the coming weeks and months.So bye for now and all stay safe out there.

[j_silv]

The 30% performance drop probably wont be noticeable on home users but for server users it will be very noticeable

[AndyMH]

Latest from the which website (UK consumer group):

Microsoft has released security patches for almost every version of Windows it currently supports. However, the extra Intel microcode fixes (see above) also need to be installed, and it’s these – in combination with Windows updates – that are likely to lead to slowdown on devices.

Microsoft has revealed that performance for millions of computer users will be noticeably worse following the release of security patches to protect PCs against an Intel processor security flaw. In a blog post on Microsoft’s website, Windows and Devices vice president Terry Myerson revealed that different combinations of hardware and software will be affected by the security patches in different ways.

Windows 10 devices released in 2016 or later: Microsoft says users will experience ‘Single-digit [percentage] slowdowns’ that most won’t notice.

Windows 10 devices released in 2015 or earlier: If you updated an older PC to Windows 10, Microsoft says some people will notice ‘more significant slowdowns’.

Windows 7 and 8 devices released in 2015 or earlier: If you didn’t upgrade to Windows 10, you’ll experience a much more significant performance drop on these machines. Microsoft says ‘most users’ will feel the difference here.

Note: All the above is based on a combination of Windows and Intel ‘microcode’ updates. Since the microcode update hasn’t yet been issued to Intel chips, anybody who’s downloaded the fix via Windows Update will not yet notice a significant performance drop

Now I updgraded my kernel on Friday and as a first time experience, it wasn't completely painless. I'm not looking forward to updating the microcode and, as a linux user, will I have to?

[Pjotr]

I'm not looking forward to updating the microcode and, as a linux user, will I have to?

Yes.