Yet another Intel CPU security vulnerability!

Rroff · 20 Sep 2019 at 15:41

zx128k said:
Dude you have no argument. My source is the white paper and also most of the tech articles on the internet. I have no burden of proof to meet. So why would I help you to attack me? I've given reasonable proof.

And there we have it. If you were so confident in your understanding of the white paper you'd have no problem in answering my question there which is a pivotal point in whether my post #261 is correct or not. If you aren't prepared to answer it then you should retract any contention to my post #261.

EDIT: PS next time just buy the 9900K or be happy with your 3800X (which isn't a bad CPU at all aside from the platform needs some maturity) and save yourself some pointless bickering on the internet

zx128k · 20 Sep 2019 at 15:49

Rroff said:
And there we have it. If you were so confident in your understanding of the white paper you'd have no problem in answering my question there which is a pivotal point in whether my post #261 is correct or not. If you aren't prepared to answer it then you should retract any contention to my post #261.

I don't have to prove to anyone what I understand or answer constant straw man arguments. If you can't disprove my source via pier review you have got nothing. This is what a debate really is, so long as the white paper is logically sound then I an okay. If their finding state that it wise to disable SMT, then that is reasonable. It's up to you to present a full coherent argument. When you can then there would be a debate. Burden of proof is on you, make the grade or get lost.

Rroff · 20 Sep 2019 at 15:51

zx128k said:
I don't have to prove to anyone what I understand or answer constant straw man arguments. If you can't disprove my source via pier review you have got nothing. This is what a debate really is, so long as the white paper is logically sound then I an okay. If their finding state that it wise to disable SMT, then that is reasonable. It's up to you to present a full coherent argument. When you can then there would be a debate. Burden of proof is on you, make the grade or get lost.

As I said nothing in my posts disproves what is in the white paper - and that fairly simple question would prove it one way or another - yet you refuse to answer it despite disagreeing with what I'm saying. That doesn't leave your argument in a very good position.

Post #261 presents a coherent argument as to why you need to understand what is in the white paper to understand the application of their advice.

I'll repeat it again as it very simply proves or disproves what I've been posting:

Rroff said:
Lets make it simple - do they or do they not need the victim data to be repeatedly hitting the relevant buffers (for instance line fill buffer) at an unrealistically high tempo so that their exploit can retrieve it?

All the information needed to answer that is in the white paper.

zx128k · 20 Sep 2019 at 15:55

Rroff said:
As I said nothing in my posts disproves what is in the white paper - and that fairly simple question would prove it one way or another - yet you refuse to answer it despite disagreeing with what I'm saying. That doesn't leave your argument in a very good position.

Post #261 presents a coherent argument as to why you need to understand what is in the white paper to understand the application of their advice.

My argument is sound, if you agree with the white paper then you agree with me because my argument is sourced there. All arguments stand on their own merits, they don't require your opinion on the matter.

IT Troll · 20 Sep 2019 at 15:56

zx128k said:
This argument was answered above. I don't have to cover all points. In a debate I just have to show one of your points are wrong. I did that. So there is no need for me to revisit a failed argument.

You claim that the JavaScript example can be served direct into a browser whilst a home user is simply browsing the web. There is no evidence of this. The proof of concept clearly shows the researcher executing the attack locally, in ideal conditions, having used local system tools and knowledge to prepare the victim. But sure, just be selective on the points you want to address.

zx128k · 20 Sep 2019 at 15:58

Rroff said:
As I said nothing in my posts disproves what is in the white paper - and that fairly simple question would prove it one way or another - yet you refuse to answer it despite disagreeing with what I'm saying. That doesn't leave your argument in a very good position.

Post #261 presents a coherent argument as to why you need to understand what is in the white paper to understand the application of their advice.

I'll repeat it again as it very simply proves or disproves what I've been posting:

All the information needed to answer that is in the white paper.

What is this the 64 million dollar question, your the host and I am the contestant. Grow up.

zx128k · 20 Sep 2019 at 15:58

IT Troll said:
You claim that the JavaScript example can be served direct into a browser whilst a home user is simply browsing the web. There is no evidence of this. The proof of concept clearly shows the researcher executing the attack in ideal conditions having used local system tools to prepare the victim. But sure, just be selective on the points you want to address.

Everything I state is sourced and referenced. This was answered above.

Rroff · 20 Sep 2019 at 16:00

zx128k said:
My argument is sound, if you agree with the white paper then you agree with me because my argument is sourced there. All arguments stand on their own merits, they don't require your opinion on the matter.

So you will have no problem saying I'm wrong for claiming that they need the victim data to be repeatedly hitting the relevant buffers (for instance line fill buffer) at an unrealistically high tempo so that their exploit can retrieve it?

Or are you going to yet again refer me to the white paper in an unspecified manner because you know I'm right?

zx128k · 20 Sep 2019 at 16:00

Rroff said:
So you will have no problem saying I'm wrong for claiming that they need the victim data to be repeatedly hitting the relevant buffers (for instance line fill buffer) at an unrealistically high tempo so that their exploit can retrieve it?

I don't have to do anything.

Rroff · 20 Sep 2019 at 16:01

zx128k said:
I don't have to do anything.

Well you kind of do because if you won't answer that one it is pretty bad form for the rest of your disagreements with my posts to stand.

I've given you the chance to conclusively prove one way or another whether what I'm saying is correct yet you are backpedaling hard - it doesn't look good on you.

zx128k · 20 Sep 2019 at 16:02

Rroff said:
Well you kind of do because if you won't answer that one it is pretty bad form for the rest of your disagreements with my posts to stand.

I really don't have to do anything. This is how debates work. You need to provide some evidence my source is wrong.

Vince · 20 Sep 2019 at 16:03

zx128k said:
I don't have to prove to anyone what I understand or answer constant straw man arguments. If you can't disprove my source via pier review you have got nothing. This is what a debate really is, so long as the white paper is logically sound then I an okay. If their finding state that it wise to disable SMT, then that is reasonable. It's up to you to present a full coherent argument. When you can then there would be a debate. Burden of proof is on you, make the grade or get lost.

Problem is you don't understand the paper you are quoting. 15 (not man size pages) attest to this. Lads please give up this is painful.

zx128k · 20 Sep 2019 at 16:04

Vince said:
Problem is you don't understand the paper you are quoting. 15 (not man size pages) attest to this. Lads please give up this is painful.

I don't have too, I can be right in saying the white paper advises disabling HT and not know or understand anything. If you can't prove the white paper is wrong then it's you who will look like a fool.

Rroff · 20 Sep 2019 at 16:04

zx128k said:
I really don't have to do anything. This is how debates work. You need to provide some evidence my source is wrong.

I am not in any shape or form claiming your source is wrong - but your interpretation and understanding of it. And given you a chance to prove it yet you are backpedalling like crazy now and just pointing at the white paper in an unspecific manner.

zx128k said:
and not know or understand anything

Yet to disagree with the points I've made you do need to know and understand the subject - so are you going to retract your disagreement with my posts? or answer the simple question I put above.

zx128k · 20 Sep 2019 at 16:05

Rroff said:
I am not in any shape or form claiming your source is wrong - but your interpretation and understanding of it. And given you a chance to prove it yet you are backpedalling like crazy now and just pointing at the white paper in an unspecific manner.

Then get lost. That your argument completely dead.

Vince · 20 Sep 2019 at 16:05

zx128k said:
I don't have too, I can be right in saying the white paper advises disabling HT and not know or understand anything. If you can't prove the white paper is wrong then it's you who will look like a fool.

Ignorance is bliss right?

Rroff · 20 Sep 2019 at 16:07

zx128k said:
Then get lost. That your argument completely dead.

Weren't you the one talking about debate?

I take it at this point you aren't standing behind your earlier posts counter to what I've said as you don't (by your own admission) understand the subject at the level required to engage with them.

zx128k · 20 Sep 2019 at 16:07

Vince said:
Ignorance is bliss right?

Only for you. Like I stated if the white paper is right, then citing it as a source would be correct. There is no other criteria needed.

zx128k · 20 Sep 2019 at 16:07

Rroff said:
Weren't you the one talking about debate?

You still here? There's the door.

Rroff · 20 Sep 2019 at 16:08

zx128k said:
Only for you. Like I stated if the white paper is right, then citing it as a source would be correct. There is no other criteria needed.

You don't get how white papers or citing works.