You are not logged in.

#1 2017-10-02 14:31:29

jwhendy
Member
Registered: 2010-04-01
Posts: 621

Request/suggestion: add note about intel_iommu with 4.13 to Arch news

If you search the forums, there are a ton of threads recently about issues with the 4.13 upgrades. I think Arch is generally excellent with it's handling of breakage-level issues, and archlinux.org is my goto when I experience them. This issue is not currently mentioned there, and it would seem more efficient to make users aware of a kernel change like this via Arch news instead of forum posts one by one.

I tried to suggest the same via this bug report but it was closed (no a bug). It's not clear how one puts forth suggestions for the arch news section... I just wanted the idea to reach those who could implement it if they agreed.

@V1del is chiming in on nearly every issue, directing people to try intel_iommu=off, or pointing them to this this thread. There would seem to be a better way. Reusing my thoughts from the bug report, I think handling this from a visible, official place like archlinux.org would be great for these reasons:

1) this seems widespread enough to make the arch news page
2) putting something in a highly visible place will hopefully save work in the forums
3) whatever the definitive fix/workaround should be is better than having users try lesser/inferior solutions

Offline

#2 2017-10-02 15:09:43

jasonwryan
Anarchist
From: .nz
Registered: 2009-05-09
Posts: 30,426
Website

Re: Request/suggestion: add note about intel_iommu with 4.13 to Arch news

The wave is mostly over now anyway. All we have left to look forward to is the people that don't read the News, search the boards, etc...


Arch + dwm   •   Mercurial repos  •   Surfraw

Registered Linux User #482438

Offline

#3 2017-10-02 15:45:40

jwhendy
Member
Registered: 2010-04-01
Posts: 621

Re: Request/suggestion: add note about intel_iommu with 4.13 to Arch news

I wondered about that. If there's a sense for how long the spillover is between event X and people finding out about X, and that time has passed... maybe we're good.

I still think this was the type of event that would make a prime candidate for arch news, though! Kind of surprised it wasn't, as I definitely got a scary boot screen. Also stinks that it prevented iwlwifi from loading, which meant I was searching on my phone for what in the world was going on!

Offline

#4 2017-10-02 17:11:20

jasonwryan
Anarchist
From: .nz
Registered: 2009-05-09
Posts: 30,426
Website

Re: Request/suggestion: add note about intel_iommu with 4.13 to Arch news

Well, for many of us who use [testing] it wasn't an issue so it would not have been possible to flag it as newsworthy. If you think about the number of installs there are, the actual percentage of people experiencing issues is pretty small.


Arch + dwm   •   Mercurial repos  •   Surfraw

Registered Linux User #482438

Offline

#5 2017-10-02 20:57:01

schnedan
Member
Registered: 2012-11-02
Posts: 19

Re: Request/suggestion: add note about intel_iommu with 4.13 to Arch news

As I am suffering from this issue, a List of known incompatible hardware would be helpfull. I have at least one more computer waiting to go to 4.13. I don't know if this box needs the iommu hack or not and not to be forced to suffer some crashes to find out would be nice.

PS: is there a (sane and uncomplicated) method to test a new kernel in a safe, read only manner, to prevent trashed file systems... installing and setting up systems is not a hobby of mine

Last edited by schnedan (2017-10-02 20:59:47)

Offline

#6 2017-10-02 21:39:32

loqs
Member
Registered: 2014-03-06
Posts: 18,961

Re: Request/suggestion: add note about intel_iommu with 4.13 to Arch news

schnedan wrote:

As I am suffering from this issue, a List of known incompatible hardware would be helpfull.

Where would such a list come from.  For example which components on your system that you upgraded are the cause of the issue?

schnedan wrote:

PS: is there a (sane and uncomplicated) method to test a new kernel in a safe, read only manner, to prevent trashed file systems... installing and setting up systems is not a hobby of mine

Not really visualizing the system affects its interactions with hardware alternatively you would need an isolated system from the one you want to protect or accept the prospect of having to fix a system that will not boot.
Would be nice if the related issues such as FS#55629 were forwarded to upstream so they could be fixed but if those affected choose not to do so,  so be it.

Offline

#7 2017-10-02 21:51:19

Trilby
Inspector Parrot
Registered: 2011-11-29
Posts: 30,471
Website

Re: Request/suggestion: add note about intel_iommu with 4.13 to Arch news

The news is for announcing upcoming issues that are known about ahead of time.  By the time this issue was known about, it had been covered extensively on the forums, so why post it in the "news"?  Perhaps we need another rss feed, for "olds".


"UNIX is simple and coherent" - Dennis Ritchie; "GNU's Not Unix" - Richard Stallman

Offline

#8 2017-10-02 22:03:29

bulletmark
Member
From: Brisbane, Australia
Registered: 2013-10-22
Posts: 713

Re: Request/suggestion: add note about intel_iommu with 4.13 to Arch news

jasonwryan wrote:

If you think about the number of installs there are, the actual percentage of people experiencing issues is pretty small.

I don't think you can assume that. I had problems on both my Arch machines after the kernel update but soon came to the forums and found the issues with the common fix. I didn't post anywhere about it and I suspect a silent majority of other Arch users experiencing problems did the same thing. There are probably others out there still scratching their head why their machine hangs up on resume from suspend and/or odd UI hangups so I agree with the OP that a news post is warranted for this serious bug.

Offline

#9 2017-10-02 22:37:11

jasonwryan
Anarchist
From: .nz
Registered: 2009-05-09
Posts: 30,426
Website

Re: Request/suggestion: add note about intel_iommu with 4.13 to Arch news

bulletmark wrote:
jasonwryan wrote:

If you think about the number of installs there are, the actual percentage of people experiencing issues is pretty small.

I don't think you can assume that. I had problems on both my Arch machines after the kernel update but soon came to the forums and found the issues with the common fix. I didn't post anywhere about it and I suspect a silent majority of other Arch users experiencing problems did the same thing. There are probably others out there still scratching their head why their machine hangs up on resume from suspend and/or odd UI hangups so I agree with the OP that a news post is warranted for this serious bug.

The same silent majority that don't enable [testing] and help to identify these issue before they are pushed to the standard repos?

Also, it isn't an Arch bug, but an upstream one. This is one of the benefits of a rolling release.


Arch + dwm   •   Mercurial repos  •   Surfraw

Registered Linux User #482438

Offline

Board footer

Powered by FluxBB