From: Tejun Heo <tj@kernel.org>
To: Bjorn Helgaas <bhelgaas@google.com>
Cc: linux-pci@vger.kernel.org, linux-kernel@vger.kernel.org
Subject: [PATCH] pci: rewrite pci_call_probe() to use workqueue instead of work_on_cpu()
Date: Thu, 23 Aug 2012 14:35:29 -0700 [thread overview]
Message-ID: <20120823213529.GH14962@google.com> (raw)
pci_call_probe() uses work_on_cpu(), which creates and tears down a
full kthread on each call, to invoke ->probe() on node local CPU for
allocation affinity.
The same goal can easily be achieved using a work item. This patch
rewrites pci_call_probe() so that it uses a work item instead of
work_on_cpu().
Note that the function is restructured for simplicity. This adds
get/put_online_cpus() pair for devices without node but the overhead
of doing so isn't anything material at this level.
Signed-off-by: Tejun Heo <tj@kernel.org>
---
I tested both paths but had to simulate device on foreign node.
Thanks.
drivers/pci/pci-driver.c | 48 +++++++++++++++++++++++++++--------------------
1 file changed, 28 insertions(+), 20 deletions(-)
diff --git a/drivers/pci/pci-driver.c b/drivers/pci/pci-driver.c
index 5270f1a..554c15a 100644
--- a/drivers/pci/pci-driver.c
+++ b/drivers/pci/pci-driver.c
@@ -274,11 +274,14 @@ struct drv_dev_and_id {
struct pci_driver *drv;
struct pci_dev *dev;
const struct pci_device_id *id;
+ struct work_struct work;
+ int error;
};
-static long local_pci_probe(void *_ddi)
+static void pci_probe_fn(struct work_struct *work)
{
- struct drv_dev_and_id *ddi = _ddi;
+ struct drv_dev_and_id *ddi = container_of(work, struct drv_dev_and_id,
+ work);
struct device *dev = &ddi->dev->dev;
int rc;
@@ -298,33 +301,38 @@ static long local_pci_probe(void *_ddi)
pm_runtime_set_suspended(dev);
pm_runtime_put_noidle(dev);
}
- return rc;
+
+ ddi->error = rc;
}
static int pci_call_probe(struct pci_driver *drv, struct pci_dev *dev,
const struct pci_device_id *id)
{
- int error, node;
+ int node;
+ int cpu = nr_cpu_ids;
struct drv_dev_and_id ddi = { drv, dev, id };
- /* Execute driver initialization on node where the device's
- bus is attached to. This way the driver likely allocates
- its local memory on the right node without any need to
- change it. */
- node = dev_to_node(&dev->dev);
- if (node >= 0) {
- int cpu;
+ /*
+ * Execute driver initialization on node where the device's bus is
+ * attached to. This way the driver likely allocates its local
+ * memory on the right node without any need to change it.
+ */
+ get_online_cpus();
- get_online_cpus();
+ node = dev_to_node(&dev->dev);
+ if (node >= 0)
cpu = cpumask_any_and(cpumask_of_node(node), cpu_online_mask);
- if (cpu < nr_cpu_ids)
- error = work_on_cpu(cpu, local_pci_probe, &ddi);
- else
- error = local_pci_probe(&ddi);
- put_online_cpus();
- } else
- error = local_pci_probe(&ddi);
- return error;
+
+ if (cpu < nr_cpu_ids) {
+ INIT_WORK_ONSTACK(&ddi.work, pci_probe_fn);
+ schedule_work_on(cpu, &ddi.work);
+ flush_work(&ddi.work);
+ } else {
+ pci_probe_fn(&ddi.work);
+ }
+
+ put_online_cpus();
+ return ddi.error;
}
/**
next reply other threads:[~2012-08-23 21:35 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
2012-08-23 21:35 Tejun Heo [this message]
2012-09-14 18:13 ` [PATCH] pci: rewrite pci_call_probe() to use workqueue instead of work_on_cpu() Bjorn Helgaas
2012-09-17 21:41 ` Tejun Heo
2012-09-17 21:45 ` Bjorn Helgaas
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=20120823213529.GH14962@google.com \
--to=tj@kernel.org \
--cc=bhelgaas@google.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-pci@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).