| View previous topic :: View next topic |
| Author |
Message |
gwr
Apprentice
Joined: 19 Nov 2014
Posts: 194
|
|
| Back to top |
|
 |
Anon-E-moose
Watchman
Joined: 23 May 2008
Posts: 6098
Location: Dallas area
|
 Posted: Tue Apr 28, 2015 12:28 pm Post subject: |
 |
|
It's funny to watch the different linux threads on the news reporting sites.
It seems the kdbus and dbus devs are commenting heavily that "kdbus is good to go".
Seems they've given up trying to get it past the kernel devs and all their commentary
and is trying to get the court of public opinion "to put pressure on the kernel devs" to get it in.
Of course if it should get in and cause the havoc with the kernel that it seems it might,
then the kernel devs would be the ones getting booed not the kdbus/dbus devs nor their mindless cheering section. 
_________________
PRIME x570-pro, 3700x, 6.1 zen kernel
gcc 13, profile 17.0 (custom bare multilib), openrc, wayland |
|
| Back to top |
|
|
Naib
Watchman
Joined: 21 May 2004
Posts: 6051
Location: Removed by Neddy
|
| Posted: Tue Apr 28, 2015 12:37 pm Post subject: |
|
|
appeal to the masses unfortunately works... especially if the PR team is better.
_________________
| Quote: | | Removed by Chiitoo |
|
|
| Back to top |
|
|
asturm
Developer
Joined: 05 Apr 2007
Posts: 8936
|
| Posted: Tue Apr 28, 2015 1:26 pm Post subject: |
|
|
| Anon-E-moose wrote: | | It seems the kdbus and dbus devs are commenting heavily that "kdbus is good to go". |
Can you provide links? |
|
| Back to top |
|
|
Anon-E-moose
Watchman
Joined: 23 May 2008
Posts: 6098
Location: Dallas area
|
| Posted: Tue Apr 28, 2015 1:48 pm Post subject: |
|
|
| genstorm wrote: | | Anon-E-moose wrote: | | It seems the kdbus and dbus devs are commenting heavily that "kdbus is good to go". |
Can you provide links? |
I don't have the links at hand, I was reading from lwn, linuxtoday, slashdot and a few others, so I don't remember where I saw the comments.
I wouldn't even have known that any of them were kdbus/dbus devs until one of them
answered to defend them (devs), by way of saying "no we don't" in response to criticism.
Which I took to mean they were in the (k)dbus dev circle.
But the number of (k)dbus devs is far outnumbered by the amount of people (who knows who they are)
commenting about it should be in the kernel now, and they pooh pooh the concerns of the kernel devs.
Using the same responses and tactics (by (k)dbus devs) that we see in the lkml discussions.
Edit to add: the whole "it's good to go" seems to be the party line of those (k)dbus devs on the lkml discussions.
It's either "that's a minor problem" or "just put it in and we'll work on those problems later"
And the last week Havoc Pennington has joined the fray (on the side of (k)dbus seemingly)
even though he only worked on the original dbus code and admits that he knows nothing
of the current implementation.
What I don't hear from any user space devs is that perhaps they should relook at the design and see if it can be improved.
And that's scary.
_________________
PRIME x570-pro, 3700x, 6.1 zen kernel
gcc 13, profile 17.0 (custom bare multilib), openrc, wayland |
|
| Back to top |
|
|
Anon-E-moose
Watchman
Joined: 23 May 2008
Posts: 6098
Location: Dallas area
|
| Posted: Tue Apr 28, 2015 6:08 pm Post subject: |
|
|
Ted Tso has this to say on the lkml
| Quote: | So the question is if one of the justifications for moving the daemon
into kernel space is that it's performance is crap, then I think it is
useful to determine whether a fully optimized userspace daemon would
be good enough.
After all, we can go down the Novell Netware path and push arbitrary
web servers, ldap servers, etc. all into the kernel on the excuse of
"the performance would be faster". But that begs the question of how
much performance improvements can be made purely in userspace, and
ignores all of the security and stability costs of moving more and
more code into the kernel.
So the question I have is why in the world do we want to be able to
support 1.5x raw sockets for a bus speed? What's the use case where
that kind of performance is required for a bus based system, and is
that a world we really want to live in? I find dbus to be extremely
hard to debug when my desktop starts doing things I don't want it to
do. The fact that it might be flinging around hundreds of thousands
of messages, and that this is something we want to encourage, doesn't
make me feel any more kindly inclined towards dbus or kdbus.... |
Note: I have a non-dbus system and it works perfectly well.
_________________
PRIME x570-pro, 3700x, 6.1 zen kernel
gcc 13, profile 17.0 (custom bare multilib), openrc, wayland |
|
| Back to top |
|
|
toralf
Developer
Joined: 01 Feb 2004
Posts: 3922
Location: Hamburg
|
| Posted: Tue Apr 28, 2015 6:15 pm Post subject: |
|
|
| Linus used to use less words to describe the current implementation of kdbus : "full of shit" |
|
| Back to top |
|
|
arnvidr
l33t
Joined: 19 Aug 2004
Posts: 629
Location: Oslo, Norway
|
| Posted: Wed Apr 29, 2015 6:17 am Post subject: |
|
|
Last reply from Linus seems to lay the discussion dead | Quote: | Now, there may be *other* reasons why kdbus is a good idea. But quite
frankly, every time somebody asks "why", performance seems to be one
of the main answers.
And quite frankly, that *stinks*.
Do proper optimizations of the actual real costs before starting to
work on kernel stuff. It's *stupid* to add a kernel driver to get 2x
improvement, when there's a 10x bloat in user space.
Is that really so hard to see? I don't think it is at *all*
appropriate to say "we're a f*cking bloated pig, but we're too lazy to
fix the bloat and the primary performance problems, so we'll add a
kernel interface to partially hide the issue".
That is particularly true because if you fix the user-level
performance problems, you may notice that there was something stupid
in the interfaces, and some of the kernel interface design was wrong.
|
Haven't seen many responses from anyone but Havoc lately, and he rightly defers any questions not about the current dbus to the kdbus developers.
Maybe they'll be forced to look at the dbus mess now.
_________________
|
|
| Back to top |
|
|
Anon-E-moose
Watchman
Joined: 23 May 2008
Posts: 6098
Location: Dallas area
|
| Posted: Wed Apr 29, 2015 8:59 am Post subject: |
|
|
| arnvidr wrote: | | Haven't seen many responses from anyone but Havoc lately, and he rightly defers any questions not about the current dbus to the kdbus developers. |
On the one hand Havoc says he doesn't know much about the current implementation of dbus,
but he still continues on trying to be an apologist for it. He's just a tool, IMO.
| Quote: | | Maybe they'll be forced to look at the dbus mess now. |
That would be nice but I wouldn't hold my breath waiting for it.
Since this last push to get kdbus into the kernel, it's been mentioned, several times, by kernel devs to relook at their dbus code.
They just keep on trying to get kdbus into the kernel (real soon) and then they'll work on the problems of it.
_________________
PRIME x570-pro, 3700x, 6.1 zen kernel
gcc 13, profile 17.0 (custom bare multilib), openrc, wayland |
|
| Back to top |
|
|
Naib
Watchman
Joined: 21 May 2004
Posts: 6051
Location: Removed by Neddy
|
| Posted: Wed Apr 29, 2015 9:06 am Post subject: |
|
|
so the question is ... why does it need to be in the kernel. if it is just "dbus in the kernel" then it would work in userland.
If it is for speed, "dbus in userland" still works
But speed has been shown to be a possible red herring. If it was about needing dbus and speed then there would be a drive for userland optimisation, which isn't occurring... why does it need to be in the kernel?
_________________
| Quote: | | Removed by Chiitoo |
|
|
| Back to top |
|
|
EmaRsk
Apprentice
Joined: 07 Sep 2004
Posts: 158
Location: Italy
|
| Posted: Wed Apr 29, 2015 1:07 pm Post subject: |
|
|
There something I don't understand. (Well, there A LOT I don't understand, but if Linus says kdbus is shit, I believe him, and I dislike dbus - and systemd - enough already).
If they want kdbus in the kernel, why don't they just maintain a patchset?
I mean, the Con Kolivas patches aren't in the official kernel, too, and still they're available for everyone.
Debian maintains its own kernel patchset, Gentoo does the same, as almost every other distro, I suppose.
Couldn't they just maintain a kdbus patchset?
It seems to me that the only reasons to rush this into the official kernel tree, must be political reasons.
Am I missing something? Could there be a real technical reason? |
|
| Back to top |
|
|
steveL
Watchman
Joined: 13 Sep 2006
Posts: 5153
Location: The Peanut Gallery
|
| Posted: Wed Apr 29, 2015 1:26 pm Post subject: |
|
|
| Why does it need to be in the kernel? Because then a nice GPL-evasion mechanism is baked-in, forever, and no-one is going to dream of arguing that that's what it's main purpose is, because Linux. |
|
| Back to top |
|
|
gwr
Apprentice
Joined: 19 Nov 2014
Posts: 194
|
| Posted: Wed Apr 29, 2015 1:37 pm Post subject: |
|
|
| steveL wrote: | | Why does it need to be in the kernel? Because then a nice GPL-evasion mechanism is baked-in, forever, and no-one is going to dream of arguing that that's what it's main purpose is, because Linux. |
Isn't dbus already a gpl-evasion mechanism? Why concern themselves of baking it right into the kernel if systemd is already forcing most major players to use it? |
|
| Back to top |
|
|
steveL
Watchman
Joined: 13 Sep 2006
Posts: 5153
Location: The Peanut Gallery
|
| Posted: Thu Apr 30, 2015 3:19 am Post subject: |
|
|
| gwr wrote: | | Isn't dbus already a gpl-evasion mechanism? Why concern themselves of baking it right into the kernel if systemd is already forcing most major players to use it? |
So no one will "dream of arguing that that is what it's main purpose is, because Linux."
Once it's in the kernel, it's in forever effectively, or at least the medium-term (10-15 years minimum.)
Everyone else will be responsible for it, so it won't be associated with RedHat, Sievers and Poeterring any more; it'll just be part of the accepted background of Linux itself, rather than a woefully ill-designed userland project.
IMO it would in fact be a stain on the reputation of Linux, one which Windoze newbs will find familiar, so in user-acceptability terms it won't be such a big deal for RedHat and other corporations seeking to "monetise" the userbase into an "income-stream."
They're after people on the level of a mobile-phone or television user, not any of us.
ATM, systemd can be avoided because it's userland, as so many of us already do, and so do other distros.
Linux is still about choice, as it has been from the beginning.
Cheers, Mr. Torvalds; I owe you a beer. :-) |
|
| Back to top |
|
|
Anon-E-moose
Watchman
Joined: 23 May 2008
Posts: 6098
Location: Dallas area
|
| Posted: Thu Apr 30, 2015 9:39 am Post subject: |
|
|
Looking at the traffic (lkml) over the last few days and even last week, it seems that there are
a great number of people who doubt the wisdom of putting kbdus into the kernel,
at least for the moment and for the current design of it. So there's hope on that front.
I think what's going to happen, is that the ones that want (k)dbus in the kernel
will start screeds disguised as a blog about how the kernel devs are keeping
the kernel in the dark ages, because they're all luddites who are stuck in the past.
The CoC might even be invoked if they can figure out how it would apply.
There have been a few kernel devs who even have the temerity to say that
there only needs to be a minimum in the kernel, ie the bus transport, not all of kdbus.
We'll see.
_________________
PRIME x570-pro, 3700x, 6.1 zen kernel
gcc 13, profile 17.0 (custom bare multilib), openrc, wayland |
|
| Back to top |
|
|
depontius
Advocate
Joined: 05 May 2004
Posts: 3509
|
| Posted: Thu Apr 30, 2015 3:40 pm Post subject: |
|
|
| toralf wrote: | | Linus used to use less words to describe the current implementation of kdbus : "full of shit" |
That guy? Whoever the heck he is, he's just a hater!
I wonder if this whole kernel push was a miscalculation by the systemd people, or a deliberate part of a plan. Hanlon's Razor would say the former.
I suspect that Linus and several others are going to start to dig in their heels shortly. If the systemd folks keep up with their normal "public opinion" approach, I see bad things. What would happen to Linux if Linus walked away saying, "This has quit being fun, goodbye!"
_________________
.sigs waste space and bandwidth |
|
| Back to top |
|
|
saellaven
l33t
Joined: 23 Jul 2006
Posts: 646
|
| Posted: Thu Apr 30, 2015 6:46 pm Post subject: |
|
|
| depontius wrote: | | toralf wrote: | | Linus used to use less words to describe the current implementation of kdbus : "full of shit" |
That guy? Whoever the heck he is, he's just a hater!
I wonder if this whole kernel push was a miscalculation by the systemd people, or a deliberate part of a plan. Hanlon's Razor would say the former.
I suspect that Linus and several others are going to start to dig in their heels shortly. If the systemd folks keep up with their normal "public opinion" approach, I see bad things. What would happen to Linux if Linus walked away saying, "This has quit being fun, goodbye!" |
systemd rolled just about everyone else in the linux world... I'm sure they figured they had the momentum and a big enough key player given GKH's kernel role (not to mention how many other kernel devs are working at redhatd) to push the kernel devs into adopting it regardless of the need or quality.
I have to admit that, for as much of a technological pragmatist as Linus is, I thought that there was a good chance he might roll over myself... I can still envision a scenario where GKH and friends harass him with the CoC policy to the point where he just says "f it, I've made enough money off linux to support my family for the rest of my life, I'm walking away... you want it so bad, you're in charge Greg." |
|
| Back to top |
|
|
gwr
Apprentice
Joined: 19 Nov 2014
Posts: 194
|
| Posted: Thu Apr 30, 2015 8:19 pm Post subject: |
|
|
| Quote: | | "f it, I've made enough money off linux to support my family for the rest of my life, I'm walking away... you want it so bad, you're in charge Greg." |
Every project needs people who have authority to say no. Once you lose that person, then all is lost and it becomes a free-for-all. If they don't like Linus' management, then they can fork the kernel. That is the entire point of open source software. |
|
| Back to top |
|
|
Anon-E-moose
Watchman
Joined: 23 May 2008
Posts: 6098
Location: Dallas area
|
| Posted: Thu Apr 30, 2015 8:34 pm Post subject: |
|
|
Have to chuckle at this. http://lkml.iu.edu/hypermail/linux/kernel/1504.3/04680.html
| Quote: | From: Eric W. Biederman
Date: Thu Apr 30 2015 - 16:14:25 EST
On April 29, 2015 7:47:53 AM CDT, Harald Hoyer <harald@xxxxxxxxxx> wrote:
...
kdbus seems to be the NIH "we know better better" approach. Many of it's design decisions we have chosen to differently elsewhere in the kernel because the have caused problems. When these issues have been pointed out in review people have blown off leading to the current mess.
Furthermore I don't know that I have seen people arguing for transporting something other than dbus messages but rather I have seen people pointing out there have been many excellent IPC mechanisms that are simpler and faster for the same kind of task and suggesting mapping dbus to better kernel primitives might be productive.
But seriously if you want to have one IPC mechanism to rule them all you won't succeed in convincing everyone with the currently sloppily designed kdbus code.
Performance matters, simplicity matters, being able to explain design decisions matter.
Eric |
Once again a kernel dev explains what's needed, sadly I don't think it'll sink into the "kdbus needs to be the way it is" crowd.
_________________
PRIME x570-pro, 3700x, 6.1 zen kernel
gcc 13, profile 17.0 (custom bare multilib), openrc, wayland |
|
| Back to top |
|
|
asturm
Developer
Joined: 05 Apr 2007
Posts: 8936
|
| Posted: Thu Apr 30, 2015 10:39 pm Post subject: |
|
|
Yeah, the list was quite entertaining those last few days.  |
|
| Back to top |
|
|
krinn
Watchman
Joined: 02 May 2003
Posts: 7470
|
| Posted: Fri May 01, 2015 9:35 am Post subject: |
|
|
am i the one thinking how kernel devs speak about kdbus show the same conceptual spirit systemd was built?
And that this pic would explain everyone really shortly the problem by just changing systemd with kdbus word in it :
http://www.microlinux.fr/download/systemd.jpg |
|
| Back to top |
|
|
ulenrich
Veteran
Joined: 10 Oct 2010
Posts: 1480
|
| Posted: Fri May 01, 2015 10:32 am Post subject: |
|
|
just found Linus say in the year 2022 | Quote: | The apparent inability (and perhaps more importantly - total unwilling[n]ess) from the kdbus team to be able to see what makes sense in a long-term general kernel and what does not, and split things up and try to push the sensible things up (and know which things are too ugly or too specialized to make sense), caused many kdbus features to never be merged.
Much of it did get merged over the years (mostly because some people spent the time to separate things out), but no, we're not going to suddenly start merging code like that just because the project is in trouble. None of the basic issues have been solved. |
[edit] faked kdbus for a name in two places.
This little piece makes me wonder GKH does not know nothing about Linux history |
|
| Back to top |
|
|
steveL
Watchman
Joined: 13 Sep 2006
Posts: 5153
Location: The Peanut Gallery
|
| Posted: Fri May 01, 2015 11:27 am Post subject: |
|
|
Heh, 2022.
This was an interesting tidbit from grsecurity:
| PAX wrote: | | the original intention of those backing LSM was to allow for the existence of proprietary security modules that used the kernel's GPL'd interfaces |
especially combined with this:
| RSBAC wrote: | | It seems that LSM is about to be removed from the official kernel and SELinux allowed to have their individual hooks. Reason: All other LSM users live outside the official tree. ..wonder why |
Granted that's from April 2006, and in 2009 the former page mentions:
| PAX wrote: | | New security modules are finally being allowed mainline, so hopefully the push from certain individuals to make SELinux the 'one true security model' for Linux is over. |
However that does not change the intent, nor the efforts taken to pursue that intent.
It's interesting because the same putsch to combine proprietary code with GPL, is now underway with the kdbust-GPL-evasion-or-nothing campaign.
And OFC, we all know who the main backer of SELinux (which seems far too complex to make a decent security infra, imo) is: RedHat.
It's at this point, we really should decide to treat RedHat as hostile, both to Gentoo and the wider Free Linux infrastructure, ie our community.
If it's not in tune with this I don't want it on machines I am liable for. |
|
| Back to top |
|
|
FreeAsInFreedom
n00b
Joined: 04 May 2015
Posts: 2
|
| Posted: Mon May 04, 2015 7:09 am Post subject: |
|
|
This is my first post here on Gentoo forums.
About 10 to 15 years ago I was a big Linux enthusiast. I've been using it work and at home since then.
Gradually, over the years that tapered until the last couple of years when I had no idea what's going on.
On a whim a couple of weeks ago I felt like getting back up to date, and then I stumbled upon the whole systemd fiasco.
Just wanted to give that intro and say I'm glad that I eventually found this forum for people who actually get it. |
|
| Back to top |
|
|
FreeAsInFreedom
n00b
Joined: 04 May 2015
Posts: 2
|
| Posted: Mon May 04, 2015 8:00 am Post subject: |
|
|
So, I've done a load of research and looking around to figure out what's going on with systemd/kdbus.
Here are some observations:
Linus is sponsored by the Linux Foundation, and is listed as a "Fellow".
Right beneath him as another "Fellow". It's that familiar name I've seen LKML threads, Greg Kroah-Hartman.
That annoying guy trying to stick his filthy kdbus in the kernel like a used car salesman. "Trust me. It's all great stuff we really need."
Here's a simplified explanation for what's going on.
You have a group of Red Hat employees - Pottering et al pushing systemd in all the corporate distros.
And then right in the kernel itself you have Greg Kroah-Hartman pushing kdbus.
Why? Because they're getting paid to. It's as simple as that.
Well actually in Pottering's case it seems more like an ego trip being enabled by a Red Hat position and paycheck.
If you look at GKH's blog you can see all the buddy-buddy references he makes to the systemd developers.
Having the systemd/kdbus monstrosity in Linux serves Red Hat's interest primarily.
I noticed that Jim Zemlin (Linus' "boss") has not voiced any opinion at all on the systemd.
Linus himself seems muzzled.
The angle that's being taken is to just shove systemd/kdbus on the community like it's inevitable.
The way these tech sites are pushing it.
Looks like it's been a sockpuppet social media blitz looks like another part of it I bet. I've dealt with this in completely different areas like GMOs, fluoridation, etc. It appears similar.
Anyway, just a fewf observations.
What else do people make of the specific political/money factors?
To me it looks like it's definitely a factor. Sort of the elephant in the LKML mailing list is not addressing how corporate money is behind it all.
It's like an unspoken rule that it must only be revolve around technical issues.
That's good to a point, but eventually I think you need to realize these people have bad character and bad intentions.
I would love for Linux to give a unequivocable and permanent "NO" on this kdbus.
Unfortunately Linus himself gets paid from the some source of funding of systemd/kdbus invasion.
A final note: maybe it seems kind of extreme but I think the concerns about NSA wanting a backdoor in Linux/RedHat machines are a legitimate concern.
This blog gets into that
https://igurublog.wordpress.com/2014/04/03/tso-and-linus-and-the-impotent-rage-against-systemd/ |
|
| Back to top |
|
|
|
|
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
|
|
Kernel & Hardware
