From: Alexander Gordeev <agordeev@redhat.com>
To: Bjorn Helgaas <bhelgaas@google.com>
Cc: linux-kernel@vger.kernel.org,
Michael Ellerman <michael@ellerman.id.au>,
Benjamin Herrenschmidt <benh@kernel.crashing.org>,
Tejun Heo <tj@kernel.org>,
Ben Hutchings <bhutchings@solarflare.com>,
David Laight <David.Laight@ACULAB.COM>,
Mark Lord <kernel@start.ca>, "H. Peter Anvin" <hpa@zytor.com>,
linux-pci@vger.kernel.org
Subject: Re: [PATCH v3 04/11] PCI/MSI/pSeries: Make quota traversing and requesting race-safe
Date: Fri, 13 Dec 2013 11:29:25 +0100 [thread overview]
Message-ID: <20131213102924.GA14728@dhcp-26-207.brq.redhat.com> (raw)
In-Reply-To: <20131210223020.GF4699@google.com>
On Tue, Dec 10, 2013 at 03:30:20PM -0700, Bjorn Helgaas wrote:
> Can you outline the race and the scenario that leads to incorrect results
> or a crash? I looked through rtas_setup_msi_irqs() (briefly) and I didn't
> see the way that concurrent calls for different devices could interfere
> with each other.
>
> I was looking for some place that modifies state, where concurrent calls
> might trample on each other, but it looks like msi_quota_for_device() is
> pretty safe: it traverses a tree, but everything it computes is on the
> stack and it doesn't seem to save results anywhere. Maybe I'm barking up
> the wrong tree?
Hmm. I've assumed the number of MSIs for a device is cached, and therefore
concurrent calls to msi_quota_for_device() and rtas_change_msi() could race.
But it seems msi_quota_for_device() indeed computes a quota while reading
only device's properties and gains constant result (well, assuming the device
tree is not updated, but this is a different story). Which makes me confused
about this note from a earlier thread:
[quote]
On Sat, 2013-10-05 at 16:20 +0200, Alexander Gordeev wrote:
> So my point is - drivers should first obtain a number of MSIs they *can*
> get, then *derive* a number of MSIs the device is fine with and only then
> request that number. Not terribly different from memory or any other type
> of resource allocation ;)
What if the limit is for a group of devices ? Your interface is racy in
that case, another driver could have eaten into the limit in between the
calls.
Ben.
[/quote]
Some comment from Ben would be nice, but I think the patch could be dropped
for now.
Thanks, Bjorn!
> Bjorn
--
Regards,
Alexander Gordeev
agordeev@redhat.com
next prev parent reply other threads:[~2013-12-13 11:33 UTC|newest]
Thread overview: 24+ messages / expand[flat|nested] mbox.gz Atom feed top
2013-11-26 9:09 [PATCH v3 00/11] Introduce pcim_enable_msi*() family helpers Alexander Gordeev
2013-11-26 9:09 ` [PATCH v3 01/11] PCI/MSI/s390: Fix single MSI only check Alexander Gordeev
2013-11-26 9:09 ` [PATCH v3 02/11] PCI/MSI/s390: Remove superfluous check of MSI type Alexander Gordeev
2013-11-26 9:09 ` [PATCH v3 03/11] PCI/MSI/pSeries: Fix wrong error code reporting Alexander Gordeev
2013-11-26 9:09 ` [PATCH v3 04/11] PCI/MSI/pSeries: Make quota traversing and requesting race-safe Alexander Gordeev
2013-12-10 22:30 ` Bjorn Helgaas
2013-12-13 10:29 ` Alexander Gordeev [this message]
2013-11-26 9:09 ` [PATCH v3 05/11] PCI/MSI: Fix return value when populate_msi_sysfs() failed Alexander Gordeev
2013-11-26 9:09 ` [PATCH v3 06/11] PCI/MSI: Return -ENOSYS for unimplemented interfaces, not -1 Alexander Gordeev
2013-11-26 9:09 ` [PATCH v3 07/11] PCI/MSI: Make pci_enable_msi/msix() 'nvec' argument type as int Alexander Gordeev
2013-11-26 20:58 ` Tejun Heo
2013-11-26 9:09 ` [PATCH v3 08/11] PCI/MSI: Factor out pci_get_msi_cap() interface Alexander Gordeev
2013-11-26 9:09 ` [PATCH v3 09/11] PCI/MSI: Get rid of pci_enable_msi_block_auto() interface Alexander Gordeev
2013-11-26 9:09 ` [PATCH v3 10/11] PCI/MSI: Convert pci_msix_table_size() to a public interface Alexander Gordeev
2013-12-10 23:08 ` Bjorn Helgaas
2013-12-12 16:06 ` Alexander Gordeev
2013-12-12 21:16 ` Bjorn Helgaas
2013-11-26 9:10 ` [PATCH v3 11/11] PCI/MSI: Introduce pcim_enable_msi*() family helpers Alexander Gordeev
2013-12-10 23:16 ` Bjorn Helgaas
2013-12-12 16:11 ` Alexander Gordeev
2013-12-12 16:11 ` Tejun Heo
2013-12-12 21:16 ` Bjorn Helgaas
2013-11-26 21:00 ` [PATCH v3 00/11] " Tejun Heo
2013-12-06 8:39 ` Alexander Gordeev
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=20131213102924.GA14728@dhcp-26-207.brq.redhat.com \
--to=agordeev@redhat.com \
--cc=David.Laight@ACULAB.COM \
--cc=benh@kernel.crashing.org \
--cc=bhelgaas@google.com \
--cc=bhutchings@solarflare.com \
--cc=hpa@zytor.com \
--cc=kernel@start.ca \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-pci@vger.kernel.org \
--cc=michael@ellerman.id.au \
--cc=tj@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).