Comments on: FG-fed Linux RNG http://bvt-trace.net/2019/10/fg-fed-linux-rng/ Fri, 21 Aug 2020 17:09:17 +0000 http://polimedia.us hourly 1 By: Implementing TMSR OS « Dorion Mode http://bvt-trace.net/2019/10/fg-fed-linux-rng/#comment-97 Implementing TMSR OS « Dorion Mode Fri, 29 Nov 2019 17:49:55 +0000 http://bvt-trace.net/?p=51#comment-97 [...] Linux 4.9.95 was genesis'd and feeding the RNG with FG is [...] [...] Linux 4.9.95 was genesis'd and feeding the RNG with FG is [...]

]]> By: bvt http://bvt-trace.net/2019/10/fg-fed-linux-rng/#comment-83 bvt Tue, 05 Nov 2019 18:10:32 +0000 http://bvt-trace.net/?p=51#comment-83 Yeah, I see: can't stop operation because of that. Thinking about the markup filtering problem a bit more, a guess some sort of preview functionality for comments would help? Yeah, I see: can't stop operation because of that.

Thinking about the markup filtering problem a bit more, a guess some sort of preview functionality for comments would help?

]]> By: Mircea Popescu http://bvt-trace.net/2019/10/fg-fed-linux-rng/#comment-82 Mircea Popescu Mon, 04 Nov 2019 19:34:54 +0000 http://bvt-trace.net/?p=51#comment-82 You realise I could turn comment #5 right back on #4, point out how teh kernel is certainly going to be a neverending source of issues, and inquire the same thing. What can we do, practically, turn off the machine ? You realise I could turn comment #5 right back on #4, point out how teh kernel is certainly going to be a neverending source of issues, and inquire the same thing. What can we do, practically, turn off the machine ?

]]> By: bvt http://bvt-trace.net/2019/10/fg-fed-linux-rng/#comment-81 bvt Wed, 23 Oct 2019 15:43:20 +0000 http://bvt-trace.net/?p=51#comment-81 And the first spotted error: unprivileged users can write to inner ring through /dev/urandom just fine. And the first spotted error: unprivileged users can write to inner ring through /dev/urandom just fine.

]]> By: bvt http://bvt-trace.net/2019/10/fg-fed-linux-rng/#comment-80 bvt Tue, 22 Oct 2019 18:22:59 +0000 http://bvt-trace.net/?p=51#comment-80 @Mircea Popescu: (1) Good that I got it right. (2) Re blog eating parts of comments, it seems to be a never ending issue. Is it really worth having this functionality? @Stanislav: I am aware of <a href="http://www.nihamkin.com/tag/kernel.html" rel="nofollow">this</a> but never tried in practice. This runtime thing seems to be designed specifically for writing modules, and it is not clear how it will work for non-module code (code is this post is not a module, for example). Something would have to call adainit(), at the very least. I can give it a shot, I guess. @Mircea Popescu: (1) Good that I got it right.
(2) Re blog eating parts of comments, it seems to be a never ending issue. Is it really worth having this functionality?

@Stanislav: I am aware of this but never tried in practice. This runtime thing seems to be designed specifically for writing modules, and it is not clear how it will work for non-module code (code is this post is not a module, for example). Something would have to call adainit(), at the very least.

I can give it a shot, I guess.

]]> By: Stanislav Datskovskiy http://bvt-trace.net/2019/10/fg-fed-linux-rng/#comment-79 Stanislav Datskovskiy Tue, 22 Oct 2019 17:18:22 +0000 http://bvt-trace.net/?p=51#comment-79 Possibly orthogonal to the immediate subj : what's the obstacle to writing kernel module in Ada? Has anyone tried yet? Possibly orthogonal to the immediate subj : what's the obstacle to writing kernel module in Ada? Has anyone tried yet?

]]> By: Mircea Popescu http://bvt-trace.net/2019/10/fg-fed-linux-rng/#comment-78 Mircea Popescu Tue, 22 Oct 2019 09:32:39 +0000 http://bvt-trace.net/?p=51#comment-78 God damn the blog ate my thing. -the criteria is properly written as (O.W - O.R) +the criteria is properly written as (O.W - O.R) < O / 2. God damn the blog ate my thing.

-the criteria is properly written as (O.W - O.R)
+the criteria is properly written as (O.W - O.R) < O / 2.

]]> By: Mircea Popescu http://bvt-trace.net/2019/10/fg-fed-linux-rng/#comment-77 Mircea Popescu Tue, 22 Oct 2019 05:54:28 +0000 http://bvt-trace.net/?p=51#comment-77 > If I'm in the error here, I will fix this ASAP. I believe you aren't, but that my spec was hastily overshort. Suppose instead of the incorrectly fixedpoint O.R >= O.W/2 the criteria is properly written as (O.W - O.R) To implement something for this I do not believe the kernel should watch the FG. In closing -- you did an impressive amt of work here ; it will take multiple digestion passes on my part. > If I'm in the error here, I will fix this ASAP.

I believe you aren't, but that my spec was hastily overshort. Suppose instead of the incorrectly fixedpoint O.R >= O.W/2 the criteria is properly written as (O.W - O.R) To implement something for this

I do not believe the kernel should watch the FG.

In closing -- you did an impressive amt of work here ; it will take multiple digestion passes on my part.

]]>