From: David Brownell <david-b@pacbell.net>
To: davidm@hpl.hp.com
Cc: Grant Grundler <iod00d@hp.com>, Greg KH <greg@kroah.com>,
vojtech@suse.cz, linux-usb-devel@lists.sourceforge.net,
linux-kernel@vger.kernel.org, linux-ia64@vger.kernel.org,
pochini@shiny.it
Subject: Re: [linux-usb-devel] Re: serious 2.6 bug in USB subsystem?
Date: Tue, 09 Mar 2004 18:53:58 -0800 [thread overview]
Message-ID: <404E8346.4070405@pacbell.net> (raw)
In-Reply-To: <16462.21505.526926.903904@napali.hpl.hp.com>
David Mosberger wrote:
> Anyhow, I have been wondering since about Sunday whether it's really
> safe to write HcControlHeadED (and HcBulkHeadED) with 0. The register
> description itself is ambiguous. Now I'm finding that Figure 6-5
> "List Service Flow" and Section 6.4.2.2 "Locating Endpoint
> Descriptors" are outright contradictory!
OHCI isn't as tightly specified as one might like. There are also
differences in how vendors implemented it (notably initialization);
but at least there don't seem to be major incompatibilities that
crept in between implementations.
> So, if the HC behaves as described in the text, then there is an
> obvious race:
> ....
> So I changed ohci-q.c ... to this:
>
> if (ed->ed_prev == NULL) {
> if (!ed->hwNextED) {
> ohci->hc_control &= ~OHCI_CTRL_CLE;
> writel (ohci->hc_control, &ohci->regs->control);
> } else
(added the ELSE)
> writel (le32_to_cpup (&ed->hwNextED),
> &ohci->regs->ed_controlhead);
> } else ...
>
> and now there are no more crashes when plugging in the BTC keyboard.
Interesting. I had those "else"s in the patch I was testing too,
and they were literally the last "is this really needed?" change I
removed before posting that patch yesterday. I had tested it on
four different implementations of OHCI (two on SMP) and the "else"
didn't make a difference ... as it might not, given the race
you identified. This is a good example of something better
found with a fresh set of eyes looking at spec and code!
Looks like that's really needed then! Whose OHCI silicon are
you testing with, by the way? Good catch, regardless.
- Dave
next prev parent reply other threads:[~2004-03-10 2:57 UTC|newest]
Thread overview: 45+ messages / expand[flat|nested] mbox.gz Atom feed top
2003-10-27 22:35 serious 2.6 bug in USB subsystem? David Mosberger
2003-10-28 1:30 ` Greg KH
2003-10-28 3:00 ` David Mosberger
2003-10-30 15:11 ` [linux-usb-devel] " David Brownell
2003-10-30 20:15 ` David Mosberger
[not found] ` <16289.55171.278494.17172@napali.hpl.hp.com>
2003-10-31 16:23 ` David Brownell
2003-10-31 18:34 ` David Mosberger
2003-10-31 18:50 ` Valdis.Kletnieks
2003-10-31 19:28 ` David Brownell
2003-10-31 19:50 ` David Mosberger
2003-10-31 20:06 ` David S. Miller
2004-03-06 2:08 ` David Mosberger
2004-03-06 2:13 ` David Mosberger
2004-03-06 4:55 ` David Brownell
2004-03-06 5:49 ` David Mosberger
2004-03-06 7:21 ` David Mosberger
2004-03-06 8:39 ` David Mosberger
2004-03-06 16:37 ` David Brownell
2004-03-08 6:18 ` Grant Grundler
2004-03-08 18:58 ` David Mosberger
2004-03-08 21:48 ` David Brownell
2004-03-09 9:15 ` David Mosberger
2004-03-09 17:36 ` David Brownell
2004-03-09 17:58 ` David Mosberger
2004-03-09 20:39 ` David Brownell
2004-03-09 23:32 ` David Mosberger
2004-03-10 2:53 ` David Brownell [this message]
2004-03-10 6:11 ` David Mosberger
2004-03-10 6:59 ` David Mosberger
2004-03-10 7:52 ` Wouter Lueks
2004-03-10 16:49 ` David Mosberger
2004-03-10 19:49 ` Wouter Lueks
2004-03-10 16:22 ` David Brownell
2004-03-10 18:04 ` David Mosberger
2004-03-11 2:43 ` David Brownell
2004-03-11 5:35 ` David Mosberger
2004-03-10 19:29 ` Colin Leroy
2004-03-06 9:17 ` [linux-usb-devel] " David Mosberger
2004-03-06 17:30 ` David Brownell
2004-03-07 13:48 ` Matthias Andree
2004-03-08 18:49 ` David Mosberger
2003-11-03 3:46 ` David Brownell
2003-11-03 21:25 ` David Mosberger
2004-03-03 12:33 ` Wouter Lueks
2004-03-03 15:30 ` Wouter Lueks
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=404E8346.4070405@pacbell.net \
--to=david-b@pacbell.net \
--cc=davidm@hpl.hp.com \
--cc=greg@kroah.com \
--cc=iod00d@hp.com \
--cc=linux-ia64@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-usb-devel@lists.sourceforge.net \
--cc=pochini@shiny.it \
--cc=vojtech@suse.cz \
/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).