From: linas@austin.ibm.com (Linas Vepstas)
To: linux-hotplug@vger.kernel.org
Subject: Re: Hardware error reporting [was Re: PCI Error reporting]
Date: Tue, 03 Oct 2006 16:01:47 +0000 [thread overview]
Message-ID: <20061003160147.GC4381@austin.ibm.com> (raw)
In-Reply-To: <20061003152636.GA4381@austin.ibm.com>
On Tue, Oct 03, 2006 at 04:32:29PM +0100, johnflux@gmail.com wrote:
> Hi,
> Say you added a card to your system, but didn't quite push it in right.
> Or say there's a dirty fingerprint on the pci connecters. The PCI bus
> detects that there is an error here, and reports this.
Only if you have an IBM PowerPC-based pSeries server (or
the pSeries desk-side machines) which have the custom
pci host bridges in them that can detect and report such errors.
There aren't any Intel/PC desktop machines that have this
capability, although I understand that some of the
next-generation PCI express busses will have this feature,
presumably in the server-class machines first. I haven't
seen, touched or smelled PCI express hardware yet.
> This is the sort of
> thing I want to report to the user.
If a card is not seated correctly, or the connector is dirty,
(or the voltage is low, etc.) there will be PCI bus parity
errors. If the problem is severe, then the PCI subsystem
will not be able to detect what kind of card it is, and so
no device driver will be selected. I'm not sure what happens
in this case. I think what happens is that the kernel assumes
that the slot is empty, as it has no particular way of
distinguishing empty slots from slots with hard errors.
I'm not sure what would happen. One could plug in a known-dead card,
and see what happens, or take a cheap card, ad razor-blade off
one of the signal pins. For PC-class hardware, I don't think
there's any generic way of finding out if something is wrong,
although maybe one could make "smart guesses". I'll try some
experiments.
--linas
-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys -- and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CIDÞVDEV
_______________________________________________
Linux-hotplug-devel mailing list http://linux-hotplug.sourceforge.net
Linux-hotplug-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linux-hotplug-devel
next prev parent reply other threads:[~2006-10-03 16:01 UTC|newest]
Thread overview: 6+ messages / expand[flat|nested] mbox.gz Atom feed top
2006-10-03 15:26 Hardware error reporting [was Re: PCI Error reporting] Linas Vepstas
2006-10-03 15:57 ` Kay Sievers
2006-10-03 16:01 ` Linas Vepstas [this message]
2006-10-03 16:26 ` Linas Vepstas
2006-10-03 21:52 ` Kay Sievers
2006-10-03 23:00 ` Greg KH
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20061003160147.GC4381@austin.ibm.com \
--to=linas@austin.ibm.com \
--cc=linux-hotplug@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).