Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.debian.user > #190656 > unrolled thread

“Meltdown” and “Spectre”: Every modern processor has unfixable security flaws

Back to article view | Back to linux.debian.user

Contents

  “Meltdown” and  “Spectre”: Every modern processor has  unfixable security flaws Curt <curty@free.fr> - 2018-01-04 18:30 +0100
    Re: “Meltdown” and “Spectre”: Every modern processor has unfixable security flaws The Wanderer <wanderer@fastmail.fm> - 2018-01-04 18:40 +0100
    Re: “Meltdown” and “Spectre”: Every modern processor has unfixable security flaws Michael Fothergill <michael.fothergill@gmail.com> - 2018-01-04 18:50 +0100
      Re: “Meltdown” and “Spectre”: Every modern processor has unfixable security flaws The Wanderer <wanderer@fastmail.fm> - 2018-01-04 19:00 +0100
        Re: “Meltdown” and  “Spectre”: Every modern processor has  unfixable security flaws Tixy <tixy@yxit.co.uk> - 2018-01-04 19:20 +0100
          Re: “Meltdown” and “Spectre”: Every modern processor has unfixable security flaws The Wanderer <wanderer@fastmail.fm> - 2018-01-04 19:30 +0100
        Re: “Meltdown” and “Spectre”: Every modern processor has unfixable security flaws Michael Fothergill <michael.fothergill@gmail.com> - 2018-01-04 19:40 +0100
          Re: “Meltdown” and “Spectre”: Every modern processor has unfixable security flaws Nicholas Geovanis <nickgeovanis@gmail.com> - 2018-01-04 20:00 +0100
        Re: “Meltdown” and “Spectre”: Every modern processor has unfixable security flaws Jack Dangler <tdldev@gmail.com> - 2018-01-05 13:50 +0100
          Re: “Meltdown” and “Spectre”: Every modern processor has  unfixable security flaws <tomas@tuxteam.de> - 2018-01-05 14:00 +0100
            Re: "Meltdown" and "Spectre": Every modern processor has unfixable security flaws "Thomas Schmitt" <scdbackup@gmx.net> - 2018-01-05 14:50 +0100
              Re: "Meltdown" and "Spectre": Every modern processor has unfixable  security flaws <tomas@tuxteam.de> - 2018-01-05 15:40 +0100
                Re: "Meltdown" and "Spectre": Every modern processor has unfixable security flaws "Thomas Schmitt" <scdbackup@gmx.net> - 2018-01-05 16:50 +0100
                  Re: "Meltdown" and "Spectre": Every modern processor has unfixable  security flaws <tomas@tuxteam.de> - 2018-01-05 17:10 +0100
                  [OFFTOPIC] Re: "Meltdown" and "Spectre": Every modern processor has unfixable security flaws Stefan Monnier <monnier@iro.umontreal.ca> - 2018-01-05 19:10 +0100
                    Re: [OFFTOPIC] Re: "Meltdown" and "Spectre": Every modern processor  has unfixable security flaws SDA <marathon.durandal@gmail.com> - 2018-01-08 02:40 +0100
                      Re: [OFFTOPIC] Re: "Meltdown" and "Spectre": Every modern processor  has unfixable security flaws Curt <curty@free.fr> - 2018-01-08 11:30 +0100
                      Re: [OFFTOPIC] Re: "Meltdown" and "Spectre": Every modern processor  has unfixable security flaws Tom Furie <tom@furie.org.uk> - 2018-01-08 13:10 +0100
            Re: “Meltdown” and “Spectre”: Every modern processor has unfixable security flaws Nicholas Geovanis <nickgeovanis@gmail.com> - 2018-01-05 17:40 +0100
              Re: “Meltdown” and “Spectre”: Every modern processor has  unfixable security flaws <tomas@tuxteam.de> - 2018-01-05 17:50 +0100
              [OFFTOPIC] Re: “Meltdown” and  “Spectre”: Every modern processor has unfixable  security flaws Stefan Monnier <monnier@iro.umontreal.ca> - 2018-01-05 19:10 +0100
    Re: “Meltdown” and “Spectre”: Every modern processor has unfixable security flaws francis picabia <fpicabia@gmail.com> - 2018-01-04 19:10 +0100
      Re: “Meltdown” and “Spectre”: Every modern processor has unfixable security flaws bw <bwtnguy@yahoo.com> - 2018-01-04 19:20 +0100
      Re: “Meltdown” and “Spectre”: Every modern processor has unfixable security flaws The Wanderer <wanderer@fastmail.fm> - 2018-01-04 19:20 +0100

Page 1 of 2  [1] 2  Next page →

#190656 — “Meltdown” and “Spectre”: Every modern processor has unfixable security flaws

https://arstechnica.com/gadgets/2018/01/meltdown-and-spectre-every-modern-processor-has-unfixable-security-fladdws/U


TL;DR

 Windows, Linux, and macOS have all received security patches that
 significantly alter how the operating systems handle virtual memory in
 order to protect against a hitherto undisclosed flaw.
...
 In the immediate term, it looks like most systems will shortly have
 patches for Meltdown. At least for Linux and Windows, these patches
 allow end-users to opt out if they would prefer. The most vulnerable
 users are probably cloud service providers; Meltdown and Spectre can
 both in principle be used to further attacks against hypervisors,
 making it easier for malicious users to break out of their virtual
 machines.
...
 For typical desktop users, the risk is arguably less significant. While
 both Meltdown and Spectre can have value in expanding the scope of an
 existing flaw, neither one is sufficient on its own to, for example,
 break out of a Web browser.

Apparent moral of story for CPU: don't speculate (but it's significantly
*slower*).

-- 
"An autobiography is only to be trusted when it reveals something disgraceful.
A man who gives a good account of himself is probably lying, since any life
when viewed from the inside is simply a series of defeats."
— George Orwell

[toc] | [next] | [standalone]

#190657 — Re: “Meltdown” and “Spectre”: Every modern processor has unfixable security flaws

[Multipart message — attachments visible in raw view] — view raw

On 2018-01-04 at 12:22, Curt wrote:

> https://arstechnica.com/gadgets/2018/01/meltdown-and-spectre-every-modern-processor-has-unfixable-security-fladdws/U
> 
> 
> TL;DR
> 
> Windows, Linux, and macOS have all received security patches that 
> significantly alter how the operating systems handle virtual memory
> in order to protect against a hitherto undisclosed flaw.
> ...
> In the immediate term, it looks like most systems will shortly have 
> patches for Meltdown. At least for Linux and Windows, these patches 
> allow end-users to opt out if they would prefer. The most vulnerable 
> users are probably cloud service providers; Meltdown and Spectre can 
> both in principle be used to further attacks against hypervisors, 
> making it easier for malicious users to break out of their virtual 
> machines.
> ...
> For typical desktop users, the risk is arguably less significant.
> While both Meltdown and Spectre can have value in expanding the scope
> of an existing flaw, neither one is sufficient on its own to, for
> example, break out of a Web browser.
> 
> Apparent moral of story for CPU: don't speculate (but it's
> significantly *slower*).

https://spectreattack.com/ has the best and most concise summary I've
seen yet, as well as links to various other places for in-depth details
(whitepapers and blog posts).


The situation as I understand it is basically that:

* Meltdown affects all Intel CPUs released since about 1995, except for
Itanium models and pre-2013 Atom models.

* Spectre appears to affect Intel, AMD, and ARM chips alike, going back
who-knows-how-far.

* Meltdown can be mitigated by security patches which people are rushing
out, but doing so has a performance cost, anywhere from 5% to 30%
depending on various factors.

* Spectre can only be fixed (or even meaningfully mitigated)
per-program, and doing so isn't necessarily trivial.

* Meltdown is relatively easy and straightforward to trigger, once you
know how.

* Spectre is harder to trigger and/or exploit, but still possible.

* Nobody knows whether either of these is being exploited in the wild.


Also,
https://techreport.com/news/33026/researchers-reveal-meltdown-and-spectre-cpu-exploits?post=1064251
is what looks to me like an "as near to layman's terms as we're likely
to get" explanation of how the Spectre exploit actually works.

-- 
   The Wanderer

The reasonable man adapts himself to the world; the unreasonable one
persists in trying to adapt the world to himself. Therefore all
progress depends on the unreasonable man.         -- George Bernard Shaw

[toc] | [prev] | [next] | [standalone]

#190658 — Re: “Meltdown” and “Spectre”: Every modern processor has unfixable security flaws

[Multipart message — attachments visible in raw view] — view raw

On 4 January 2018 at 17:22, Curt <curty@free.fr> wrote:

> https://arstechnica.com/gadgets/2018/01/meltdown-and-spectre-every-modern-
> processor-has-unfixable-security-fladdws/U
>
>
> TL;DR
>
>  Windows, Linux, and macOS have all received security patches that
>  significantly alter how the operating systems handle virtual memory in
>  order to protect against a hitherto undisclosed flaw.
> ...
>  In the immediate term, it looks like most systems will shortly have
>  patches for Meltdown. At least for Linux and Windows, these patches
>  allow end-users to opt out if they would prefer. The most vulnerable
>  users are probably cloud service providers; Meltdown and Spectre can
>  both in principle be used to further attacks against hypervisors,
>  making it easier for malicious users to break out of their virtual
>  machines.
> ...
>  For typical desktop users, the risk is arguably less significant. While
>  both Meltdown and Spectre can have value in expanding the scope of an
>  existing flaw, neither one is sufficient on its own to, for example,
>  break out of a Web browser.
>
> Apparent moral of story for CPU: don't speculate (but it's significantly
> *slower*).
>

​Isn't this mainly an Intel problem?  I use AMD chipsets.  I would go for
Ryzen nowadays anyway.

Regards

Michael Fothergill​



>
> --
> "An autobiography is only to be trusted when it reveals something
> disgraceful.
> A man who gives a good account of himself is probably lying, since any life
> when viewed from the inside is simply a series of defeats."
> — George Orwell
>
>

[toc] | [prev] | [next] | [standalone]

#190659 — Re: “Meltdown” and “Spectre”: Every modern processor has unfixable security flaws

[Multipart message — attachments visible in raw view] — view raw

On 2018-01-04 at 12:30, Michael Fothergill wrote:

> On 4 January 2018 at 17:22, Curt <curty@free.fr> wrote:
> 
>> https://arstechnica.com/gadgets/2018/01/meltdown-and-spectre-every-modern-
>> processor-has-unfixable-security-fladdws/U
>>
>>
>> TL;DR
>>
>>  Windows, Linux, and macOS have all received security patches that
>>  significantly alter how the operating systems handle virtual memory in
>>  order to protect against a hitherto undisclosed flaw.
>> ...
>>  In the immediate term, it looks like most systems will shortly have
>>  patches for Meltdown. At least for Linux and Windows, these patches
>>  allow end-users to opt out if they would prefer. The most vulnerable
>>  users are probably cloud service providers; Meltdown and Spectre can
>>  both in principle be used to further attacks against hypervisors,
>>  making it easier for malicious users to break out of their virtual
>>  machines.
>> ...
>>  For typical desktop users, the risk is arguably less significant. While
>>  both Meltdown and Spectre can have value in expanding the scope of an
>>  existing flaw, neither one is sufficient on its own to, for example,
>>  break out of a Web browser.
>>
>> Apparent moral of story for CPU: don't speculate (but it's significantly
>> *slower*).
> 
> ​Isn't this mainly an Intel problem?  I use AMD chipsets.  I would go for
> Ryzen nowadays anyway.

Meltdown so far is not known to affect anything other than Intel.

Spectre, however, is confirmed to affect AMD CPUs - and Ryzen CPUs are
specifically stated to be affected.

-- 
   The Wanderer

The reasonable man adapts himself to the world; the unreasonable one
persists in trying to adapt the world to himself. Therefore all
progress depends on the unreasonable man.         -- George Bernard Shaw

[toc] | [prev] | [next] | [standalone]

#190662

On Thu, 2018-01-04 at 12:55 -0500, The Wanderer wrote:
> Meltdown so far is not known to affect anything other than Intel.

And ARM's Cortex-A75 [1] which according to The Register [2] "Qualcomm's
upcoming Snapdragon 845 is an example part that uses the A75"

[1] https://developer.arm.com/support/security-update
[2] https://www.theregister.co.uk/2018/01/04/intel_amd_arm_cpu_vulnerability/

-- 
Tixy

[toc] | [prev] | [next] | [standalone]

#190664 — Re: “Meltdown” and “Spectre”: Every modern processor has unfixable security flaws

[Multipart message — attachments visible in raw view] — view raw

On 2018-01-04 at 13:17, Tixy wrote:

> On Thu, 2018-01-04 at 12:55 -0500, The Wanderer wrote:
> 
>> Meltdown so far is not known to affect anything other than Intel.
> 
> And ARM's Cortex-A75 [1] which according to The Register [2]
> "Qualcomm's upcoming Snapdragon 845 is an example part that uses the
> A75"

Thanks for pointing that out; I'd read the The Reg article, but missed
retaining that part, since they didn't list it all in one place.

Looks like the spectreattack.com overview is inaccurate or out-of-date
in that detail, then; at the moment, it still states that "At the
moment, it is unclear whether ARM and AMD processors are also affected
by Meltdown.".

At least the A75 isn't out yet (according to those same articles), so
it's not something end users currently need to be concerned about.

-- 
   The Wanderer

The reasonable man adapts himself to the world; the unreasonable one
persists in trying to adapt the world to himself. Therefore all
progress depends on the unreasonable man.         -- George Bernard Shaw

[toc] | [prev] | [next] | [standalone]

#190665 — Re: “Meltdown” and “Spectre”: Every modern processor has unfixable security flaws

[Multipart message — attachments visible in raw view] — view raw

On 4 January 2018 at 17:55, The Wanderer <wanderer@fastmail.fm> wrote:

> On 2018-01-04 at 12:30, Michael Fothergill wrote:
>
> > On 4 January 2018 at 17:22, Curt <curty@free.fr> wrote:
> >
> >> https://arstechnica.com/gadgets/2018/01/meltdown-and-
> spectre-every-modern-
> >> processor-has-unfixable-security-fladdws/U
> >>
> >>
> >> TL;DR
> >>
> >>  Windows, Linux, and macOS have all received security patches that
> >>  significantly alter how the operating systems handle virtual memory in
> >>  order to protect against a hitherto undisclosed flaw.
> >> ...
> >>  In the immediate term, it looks like most systems will shortly have
> >>  patches for Meltdown. At least for Linux and Windows, these patches
> >>  allow end-users to opt out if they would prefer. The most vulnerable
> >>  users are probably cloud service providers; Meltdown and Spectre can
> >>  both in principle be used to further attacks against hypervisors,
> >>  making it easier for malicious users to break out of their virtual
> >>  machines.
> >> ...
> >>  For typical desktop users, the risk is arguably less significant. While
> >>  both Meltdown and Spectre can have value in expanding the scope of an
> >>  existing flaw, neither one is sufficient on its own to, for example,
> >>  break out of a Web browser.
> >>
> >> Apparent moral of story for CPU: don't speculate (but it's significantly
> >> *slower*).
> >
> > ​Isn't this mainly an Intel problem?  I use AMD chipsets.  I would go for
> > Ryzen nowadays anyway.
>
> Meltdown so far is not known to affect anything other than Intel.
>
> Spectre, however, is confirmed to affect AMD CPUs - and Ryzen CPUs are
> specifically stated to be affected.
>

​But if the spectre vulnerability is hard exploit in practice and  even in
artificial test situations devised e.g. in the world of Linus Torvalds then
​AMD could turn out to be relatively cyber-kosher in the end.

MF


>
> --
>    The Wanderer
>
> The reasonable man adapts himself to the world; the unreasonable one
> persists in trying to adapt the world to himself. Therefore all
> progress depends on the unreasonable man.         -- George Bernard Shaw
>
>

[toc] | [prev] | [next] | [standalone]

#190666 — Re: “Meltdown” and “Spectre”: Every modern processor has unfixable security flaws

I was hoping to be retired before this happened......
All of AWS EC2 is rebooting today by 4pm UTC
AppArmor everywhere: Can't trust the hardware to do it right. Clowns! Buffo!

On Thu, Jan 4, 2018 at 12:19 PM, Michael Fothergill
<michael.fothergill@gmail.com> wrote:
>
>
> On 4 January 2018 at 17:55, The Wanderer <wanderer@fastmail.fm> wrote:
>>
>> On 2018-01-04 at 12:30, Michael Fothergill wrote:
>>
>> > On 4 January 2018 at 17:22, Curt <curty@free.fr> wrote:
>> >
>> >>
>> >> https://arstechnica.com/gadgets/2018/01/meltdown-and-spectre-every-modern-
>> >> processor-has-unfixable-security-fladdws/U
>> >>
>> >>
>> >> TL;DR
>> >>
>> >>  Windows, Linux, and macOS have all received security patches that
>> >>  significantly alter how the operating systems handle virtual memory in
>> >>  order to protect against a hitherto undisclosed flaw.
>> >> ...
>> >>  In the immediate term, it looks like most systems will shortly have
>> >>  patches for Meltdown. At least for Linux and Windows, these patches
>> >>  allow end-users to opt out if they would prefer. The most vulnerable
>> >>  users are probably cloud service providers; Meltdown and Spectre can
>> >>  both in principle be used to further attacks against hypervisors,
>> >>  making it easier for malicious users to break out of their virtual
>> >>  machines.
>> >> ...
>> >>  For typical desktop users, the risk is arguably less significant.
>> >> While
>> >>  both Meltdown and Spectre can have value in expanding the scope of an
>> >>  existing flaw, neither one is sufficient on its own to, for example,
>> >>  break out of a Web browser.
>> >>
>> >> Apparent moral of story for CPU: don't speculate (but it's
>> >> significantly
>> >> *slower*).
>> >
>> > Isn't this mainly an Intel problem?  I use AMD chipsets.  I would go for
>> > Ryzen nowadays anyway.
>>
>> Meltdown so far is not known to affect anything other than Intel.
>>
>> Spectre, however, is confirmed to affect AMD CPUs - and Ryzen CPUs are
>> specifically stated to be affected.
>
>
> But if the spectre vulnerability is hard exploit in practice and  even in
> artificial test situations devised e.g. in the world of Linus Torvalds then
> AMD could turn out to be relatively cyber-kosher in the end.
>
> MF
>
>>
>>
>> --
>>    The Wanderer
>>
>> The reasonable man adapts himself to the world; the unreasonable one
>> persists in trying to adapt the world to himself. Therefore all
>> progress depends on the unreasonable man.         -- George Bernard Shaw
>>
>

[toc] | [prev] | [next] | [standalone]

#190672 — Re: “Meltdown” and “Spectre”: Every modern processor has unfixable security flaws

On 01/04/2018 12:55 PM, The Wanderer wrote:
> On 2018-01-04 at 12:30, Michael Fothergill wrote:
>
>> On 4 January 2018 at 17:22, Curt <curty@free.fr> wrote:
>>
>>> https://arstechnica.com/gadgets/2018/01/meltdown-and-spectre-every-modern-
>>> processor-has-unfixable-security-fladdws/U
>>>
>>>
>>> TL;DR
>>>
>>>   Windows, Linux, and macOS have all received security patches that
>>>   significantly alter how the operating systems handle virtual memory in
>>>   order to protect against a hitherto undisclosed flaw.
>>> ...
>>>   In the immediate term, it looks like most systems will shortly have
>>>   patches for Meltdown. At least for Linux and Windows, these patches
>>>   allow end-users to opt out if they would prefer. The most vulnerable
>>>   users are probably cloud service providers; Meltdown and Spectre can
>>>   both in principle be used to further attacks against hypervisors,
>>>   making it easier for malicious users to break out of their virtual
>>>   machines.
>>> ...
>>>   For typical desktop users, the risk is arguably less significant. While
>>>   both Meltdown and Spectre can have value in expanding the scope of an
>>>   existing flaw, neither one is sufficient on its own to, for example,
>>>   break out of a Web browser.
>>>
>>> Apparent moral of story for CPU: don't speculate (but it's significantly
>>> *slower*).
>> ​Isn't this mainly an Intel problem?  I use AMD chipsets.  I would go for
>> Ryzen nowadays anyway.
> Meltdown so far is not known to affect anything other than Intel.
>
> Spectre, however, is confirmed to affect AMD CPUs - and Ryzen CPUs are
> specifically stated to be affected.
>
Did this also affect Motorola chipsets? I know they haven't been popular 
in a while, but I believe they are still in use (i.e. 68000)

[toc] | [prev] | [next] | [standalone]

#190673 — Re: “Meltdown” and “Spectre”: Every modern processor has unfixable security flaws

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Fri, Jan 05, 2018 at 07:39:23AM -0500, Jack Dangler wrote:

[...]

> Did this also affect Motorola chipsets? I know they haven't been
> popular in a while, but I believe they are still in use (i.e. 68000)

You can answer this question yourself:

 - Melt is an attack on virtual memory infrastructure. Does
   any of the processors in the M68K family support VM?

 - Spectre is an attack on speculative prefetching. Does any
   of the M68K family members prefetch speculatively?

The excellent Wikipedia article on 68K [1] might be a good starting
point on that.

My bet would be on "no to both", since M68K is from that happy,
good ol' time where (micro) CPUs were so excruciatingly slow
that there was little pressure to implement those monstrosities
which basically hide the facts of too little/too slow physical
memory (mainframes of that time had at least VM, possibly
speculative prefetch).

Cheers

[1] https://en.wikipedia.org/wiki/Motorola_68000

- -- tomás
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.12 (GNU/Linux)

iEYEARECAAYFAlpPdckACgkQBcgs9XrR2kae7wCfVk+Nn+5+Eib/Oiy3RD8+/sJ0
uRcAmwWVgJ9Hke8AwClhpCZNpqdqzxtp
=sDdP
-----END PGP SIGNATURE-----

[toc] | [prev] | [next] | [standalone]

#190674 — Re: "Meltdown" and "Spectre": Every modern processor has unfixable security flaws

Hi,

tomas@tuxteam.de wrote:
> Does any of the processors in the M68K family support VM?

They did since the early 1980s when i wondered what the advantage of an
68010 would be over an 68000 (with HP BASIC: none).
  http://gunkies.org/wiki/MC68010

After all, early Sun, HP and Apollo Unix workstations were based on 680X0.


> Does any of the M68K family members prefetch speculatively?

The last one (68060) had Branch Prediction, which implies some kind of
speculative execution and data prefetching.
  https://en.wikipedia.org/wiki/Motorola_68000_series


Good news: Google claims to have a software remedy against Spectre,
called the "Retpoline". It keeps the CPU speculator busy with harmless
things while vulnerable commands are executed. I guess we will see it soon
as compiler option.
  https://support.google.com/faqs/answer/7625886
(I love the image of a hyperactive 7-year old child wrapped in a warehouse
 of trampolines.)

Man against hardware. Who will finally win ?


Have a nice day :)

Thomas

[toc] | [prev] | [next] | [standalone]

#190675 — Re: "Meltdown" and "Spectre": Every modern processor has unfixable security flaws

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Fri, Jan 05, 2018 at 02:41:57PM +0100, Thomas Schmitt wrote:
> Hi,
> 
> tomas@tuxteam.de wrote:
> > Does any of the processors in the M68K family support VM?
> 
> They did since the early 1980s when i wondered what the advantage of an
> 68010 would be over an 68000 (with HP BASIC: none).
>   http://gunkies.org/wiki/MC68010
> 
> After all, early Sun, HP and Apollo Unix workstations were based on 680X0.

With TLB cache and all that? Pretty impressive :)

> > Does any of the M68K family members prefetch speculatively?
> 
> The last one (68060) had Branch Prediction, which implies some kind of
> speculative execution and data prefetching.
>   https://en.wikipedia.org/wiki/Motorola_68000_series

I remember at the time to be pretty discouraging to see 8086 and ilk
eating all the market.

> Man against hardware. Who will finally win ?

Hardware.

Cheers
- -- t
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.12 (GNU/Linux)

iEYEARECAAYFAlpPjgsACgkQBcgs9XrR2kat8wCfTIjdez6O0MMe+5dCvc1rntUL
bskAniAbwBnxHumIdRpojWpSp9o0iSaJ
=mZjm
-----END PGP SIGNATURE-----

[toc] | [prev] | [next] | [standalone]

#190681 — Re: "Meltdown" and "Spectre": Every modern processor has unfixable security flaws

Hi,

tomas@tuxteam.de wrote:
> > > Does any of the processors in the M68K family support VM?

I wrote:
> >   http://gunkies.org/wiki/MC68010

> With TLB cache and all that? Pretty impressive :)

I am not sure about the 68010 and its separate MMU. But beginning with 68020
there surely was memory space separation per process and cache memory in the
CPU. Looking at
  https://en.wikipedia.org/wiki/Motorola_68000_series#Improvement_history
i'd say that the family participated in the development of those tricks
for microprocessors. (Mainframe is said to have invented many of the modern
stuff but failed to make it affordable.)


> I remember at the time to be pretty discouraging to see 8086 and ilk
> eating all the market.

Well, MC680X0 first fell victim to RISC CPUs like DEC Alpha, MIPS or SPARC.
Then x386 made possible ultra-low-cost Unix machines (not more than 5000 USD),
x486 drilled holes into the SPARCstation market, Pentium began to compete
with the faster RISC architectures, and Pentium II decided the race.
PowerPC and MIPS still offered some resistence, but the system prices of
fat x86 machines with GNU/Linux were at most half those of the cheapest
RISC based Unix workstations.


> > Man against hardware. Who will finally win ?

> Hardware.

The more we should enjoy our intermediate victories.


Have a nice day :)

Thomas

[toc] | [prev] | [next] | [standalone]

#190683 — Re: "Meltdown" and "Spectre": Every modern processor has unfixable security flaws

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Fri, Jan 05, 2018 at 04:39:41PM +0100, Thomas Schmitt wrote:
> Hi,
> 
> tomas@tuxteam.de wrote:

(thanks for this walk down the memory (pun? me?) lane.

[...]

> > > Man against hardware. Who will finally win ?
> 
> > Hardware.
> 
> The more we should enjoy our intermediate victories.

Optimist as always. Thomas, you're my hero :)

Cheers
- -- t
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.12 (GNU/Linux)

iEYEARECAAYFAlpPotwACgkQBcgs9XrR2kYLDACdHaLSYNUXwy8T159mBnjbsdR5
Yn0An01el2YYjhpfoyhVNAJ1c2sVzHhq
=wbrL
-----END PGP SIGNATURE-----

[toc] | [prev] | [next] | [standalone]

#190692 — [OFFTOPIC] Re: "Meltdown" and "Spectre": Every modern processor has unfixable security flaws

>> With TLB cache and all that? Pretty impressive :)
> I am not sure about the 68010 and its separate MMU. But beginning with 68020
> there surely was memory space separation per process and cache memory in the
> CPU.

The 68020 didn't have an MMU on chip (it required a separate chip
(MC68851) if you wanted the virtual memory features of an MMU).  The MMU
came on-board only in the 68030 (and then the FPU came on board for the
68040).


        Stefan

[toc] | [prev] | [next] | [standalone]

#190779 — Re: [OFFTOPIC] Re: "Meltdown" and "Spectre": Every modern processor has unfixable security flaws

Show who you're quoting with an attribution line, please!

[toc] | [prev] | [next] | [standalone]

#190789 — Re: [OFFTOPIC] Re: "Meltdown" and "Spectre": Every modern processor has unfixable security flaws

On 2018-01-08, SDA <marathon.durandal@gmail.com> wrote:

<?>

> Show who you're quoting with an attribution line, please!
>

Tit for tat, unintended irony, blatant hypocrisy, or something else (I'm
leaning toward the foremost, but you never know)?

Apropos, as revealed in another thread, I'm dying to learn why SM never
includes an attributive prefacing line (while *signing* his own posts at
the bottom so we're absolutely clear about who *he* is), but you've
eliminated that interrogative step and gone directly to demanding
corrective measures.

-- 
"An autobiography is only to be trusted when it reveals something disgraceful.
A man who gives a good account of himself is probably lying, since any life
when viewed from the inside is simply a series of defeats."
— George Orwell

[toc] | [prev] | [next] | [standalone]

#190796 — Re: [OFFTOPIC] Re: "Meltdown" and "Spectre": Every modern processor has unfixable security flaws

[Multipart message — attachments visible in raw view] — view raw

On Sun, Jan 07, 2018 at 08:32:17PM -0500, SDA wrote:

> Show who you're quoting with an attribution line, please!

With proper attribution, we might know who you are addressing with this
statement...

Cheers,
Tom

-- 
What's the matter with the world?  Why, there ain't but one thing wrong
with every one of us -- and that's "selfishness."
		-- The Best of Will Rogers

[toc] | [prev] | [next] | [standalone]

#190687 — Re: “Meltdown” and “Spectre”: Every modern processor has unfixable security flaws

On Fri, Jan 5, 2018 at 6:55 AM,  <tomas@tuxteam.de> wrote:
> (mainframes of that time had at least VM, possibly
> speculative prefetch).

Is it correct to call branch prediction the same as speculative execution?
If so, then "yes" they had it, but I don't honestly know if that's correct.
Pipeline rewinding was necessary on prediction failure, etc, similarly.

> Cheers
>
> [1] https://en.wikipedia.org/wiki/Motorola_68000
>
> - -- tomás

[toc] | [prev] | [next] | [standalone]

#190688 — Re: “Meltdown” and “Spectre”: Every modern processor has unfixable security flaws

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Fri, Jan 05, 2018 at 10:33:45AM -0600, Nicholas Geovanis wrote:
> On Fri, Jan 5, 2018 at 6:55 AM,  <tomas@tuxteam.de> wrote:
> > (mainframes of that time had at least VM, possibly
> > speculative prefetch).
> 
> Is it correct to call branch prediction the same as speculative execution?

I'm far from a processor expert myself but yes, that sounds plausible: the
memory unit is prefeching stuff on the speculation that the branch will
go one way, to keep the pipeline full. When things go the other way...

> If so, then "yes" they had it, but I don't honestly know if that's correct.
> Pipeline rewinding was necessary on prediction failure, etc, similarly.

... this happens.

I guess that this, in combination with a cache, could be enough for the
kind of timing attacks demonstrated by Spectre. But I'm speculating myself
here :)

Cheers
- -- tomás
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.12 (GNU/Linux)

iEYEARECAAYFAlpPq9QACgkQBcgs9XrR2kZPhQCeNpZYsjInyrAh5HNPqTkJ6w33
s9gAn1th2QyrDsjqu/dOlZcOjRFTC3Mo
=dV+r
-----END PGP SIGNATURE-----

[toc] | [prev] | [next] | [standalone]

Page 1 of 2  [1] 2  Next page →

Back to top | Article view | linux.debian.user

csiph-web