You are not logged in.

#1 2017-02-19 12:45:30

nannerpussy
Member
Registered: 2017-02-15
Posts: 96

Affected by very old, fairly common bug - report or not?

The issue is a priority 3 error on boot: "sp5100_tco: I/O address 0x0cd6 already in use" which is logged, system then boots fine. A quick search yields results from Fedora and RH as well as Arch going back to kernel 3.0 or earlier. I followed the advice from an Arch thread for it and blacklisted it in /etc/modprobe.d/, but a dirty systemctl output kind of drives me nuts. I have not seen this error on any other distro, but Arch and the kernel are obviously not something I could try anywhere else, so I can't say for certain if it's a universal problem, but from what I'm reading it's tied to my CPU which is an AMD FX-6100 or my Radeon HD7750 GPU, or some combination of the two.

I read the How-to Report A Bug page, but not sure if I should waste anyone's time with this. I cannot find another report in the Bug listings, but like I said, it's an old bug and old to Arch. Thanks.

Offline

#2 2017-02-19 13:56:27

Trilby
Inspector Parrot
Registered: 2011-11-29
Posts: 29,444
Website

Re: Affected by very old, fairly common bug - report or not?

I really don't have an answer, but I'm sure anyone who might be able to answer would appreciate if you'd share links to all these references you refer to so they could actually know what you are talking about.


"UNIX is simple and coherent..." - Dennis Ritchie, "GNU's Not UNIX" -  Richard Stallman

Offline

#3 2017-02-19 21:07:09

seth
Member
Registered: 2012-09-03
Posts: 49,981

Re: Affected by very old, fairly common bug - report or not?

Offline

#4 2017-02-20 03:29:20

nannerpussy
Member
Registered: 2017-02-15
Posts: 96

Re: Affected by very old, fairly common bug - report or not?

Trilby wrote:

I really don't have an answer, but I'm sure anyone who might be able to answer would appreciate if you'd share links to all these references you refer to so they could actually know what you are talking about.

Yeah, sorry about that. I was aiming for the shortest, most generalized question so it didn't come off as my asking for technical support. Just wanted an answer in general I guess, when it concerns well-known older bugs.

seth wrote:

<lsmod cmd and link>

Following up on Trilby's request here too, but you've got the most likely suspect. It turns out AMD and the ATI GPU I use(d) have had several major issues crop up around the early KMS sequence in booting over quite a few kernels. Without tracking those forums back down right now, here are other results from the "i2x-piix4" query that also feature AMD hardware identical or close to what I've used at various points that I can remember having errors or worse with:

https://bugzilla.kernel.org/show_bug.cgi?id=105311
https://bugzilla.kernel.org/show_bug.cgi?id=70711
https://bugzilla.kernel.org/show_bug.cgi?id=58621

But as I said, your link is the most likely culprit. Maybe if I get back into give-a-shit-and-persevere mode I will re-investigate and scour Google for precise replications of the error and kernel version. For anyone who may see this post and is having the issue, begin at this Bugzilla post and decide if the minor annoyance is worth the investment of time to fix it. A cursory Google search with your error output (if it's anything like the errors featured here) will land you on a whole bunch of Fedora forums, which is where I decided that in my case and best judgement, this is an annoyance rather than an issue:

https://bugzilla.kernel.org/show_bug.cgi?id=170741

TL;DR
Based on what it is and where it comes from, combined with AMD's history of hardware that likes to get upset by the kernel, I don't forsee this being something I can personally fix, and I'm fine with a lone error in my systemctrl log when it's harmless and probably won't lead to anything. Just an annoyance and thought a bug report might help someone. Not to mention my very novice level of skill and total ineptitude at code dissection or heavy debugging of things like possible kernel bugs.

If it's OK to post that report I'll comb my history and find those Fedora forum posts that had several identical situations in them.

Last edited by nannerpussy (2017-02-20 03:38:27)

Offline

#5 2017-02-20 05:54:52

Pyntux
Member
From: Serbia
Registered: 2008-12-21
Posts: 391

Re: Affected by very old, fairly common bug - report or not?

I have same output when system is booting, but I forgot about that because I set up "silent boot" (look at wiki), but that does not change anything beside I do not see that error.

Last edited by Pyntux (2017-02-20 05:55:10)


I do not speak English, but I understand...

Offline

#6 2017-02-20 07:23:07

seth
Member
Registered: 2012-09-03
Posts: 49,981

Re: Affected by very old, fairly common bug - report or not?

The consequence is, that you've no hardware watchdog around.
There's a software watchdog and the watchdog is ideally never required ;-)

Offline

#7 2017-02-20 11:39:51

Trilby
Inspector Parrot
Registered: 2011-11-29
Posts: 29,444
Website

Re: Affected by very old, fairly common bug - report or not?

For the general answer then, age is irrelevant: if it 1) really is a bug, 2) affects current versions, 3) there are no currently open bug reports for it, and 4) you want to see it fixed, then report it.

But do be careful to check on points 1-3 well.  This is where the references help.


"UNIX is simple and coherent..." - Dennis Ritchie, "GNU's Not UNIX" -  Richard Stallman

Offline

#8 2017-02-20 13:54:32

Scimmia
Fellow
Registered: 2012-09-01
Posts: 11,463

Re: Affected by very old, fairly common bug - report or not?

Specifically, report it upstream. Reporting a kernel bug to Arch won't do anything.

Offline

Board footer

Powered by FluxBB