Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Date:   Tue, 12 Apr 2022 15:43:47 -0700
In-Reply-To: <20220412224348.1038613-2-tansuresh@google.com>
Message-Id: <20220412224348.1038613-3-tansuresh@google.com>
Mime-Version: 1.0
References: <20220412224348.1038613-1-tansuresh@google.com> <20220412224348.1038613-2-tansuresh@google.com>
Subject: [PATCH v2 2/3] PCI: Support asynchronous shutdown
From:   Tanjore Suresh <tansuresh@google.com>
To:     Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
        "Rafael J . Wysocki" <rafael@kernel.org>,
        Christoph Hellwig <hch@lst.de>,
        Sagi Grimberg <sagi@grimberg.me>,
        Bjorn Helgaas <bhelgaas@google.com>
Cc:     linux-kernel@vger.kernel.org, linux-nvme@lists.infradead.org,
        linux-pci@vger.kernel.org, Tanjore Suresh <tansuresh@google.com>
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

Enhances the base PCI driver to add support for asynchronous
shutdown.

Assume a device takes n secs to shutdown. If a machine has been
populated with M such devices, the total time spent in shutting down
all the devices will be M * n secs, if the shutdown is done
synchronously. For example, if NVMe PCI Controllers take 5 secs
to shutdown and if there are 16 such NVMe controllers in a system,
system will spend a total of 80 secs to shutdown all
NVMe devices in that system.

In order to speed up the shutdown time, asynchronous interface to
shutdown has been implemented. This will significantly reduce
the machine reboot time.

Signed-off-by: Tanjore Suresh <tansuresh@google.com>
---
 drivers/pci/pci-driver.c | 20 ++++++++++++++++----
 include/linux/pci.h      |  4 ++++
 2 files changed, 20 insertions(+), 4 deletions(-)

diff --git a/drivers/pci/pci-driver.c b/drivers/pci/pci-driver.c
index 4ceeb75fc899..63f49a8dff8e 100644
--- a/drivers/pci/pci-driver.c
+++ b/drivers/pci/pci-driver.c
@@ -501,16 +501,28 @@ static void pci_device_remove(struct device *dev)
 	pci_dev_put(pci_dev);
 }
 
-static void pci_device_shutdown(struct device *dev)
+static void pci_device_async_shutdown_start(struct device *dev)
 {
 	struct pci_dev *pci_dev = to_pci_dev(dev);
 	struct pci_driver *drv = pci_dev->driver;
 
 	pm_runtime_resume(dev);
 
-	if (drv && drv->shutdown)
+	if (drv && drv->async_shutdown_start)
+		drv->async_shutdown_start(pci_dev);
+	else if (drv && drv->shutdown)
 		drv->shutdown(pci_dev);
 
+}
+
+static void pci_device_async_shutdown_end(struct device *dev)
+{
+	struct pci_dev *pci_dev = to_pci_dev(dev);
+	struct pci_driver *drv = pci_dev->driver;
+
+	if (drv && drv->async_shutdown_end)
+		drv->async_shutdown_end(pci_dev);
+
 	/*
 	 * If this is a kexec reboot, turn off Bus Master bit on the
 	 * device to tell it to not continue to do DMA. Don't touch
@@ -521,7 +533,6 @@ static void pci_device_shutdown(struct device *dev)
 	if (kexec_in_progress && (pci_dev->current_state <= PCI_D3hot))
 		pci_clear_master(pci_dev);
 }
-
 #ifdef CONFIG_PM
 
 /* Auxiliary functions used for system resume and run-time resume. */
@@ -1625,7 +1636,8 @@ struct bus_type pci_bus_type = {
 	.uevent		= pci_uevent,
 	.probe		= pci_device_probe,
 	.remove		= pci_device_remove,
-	.shutdown	= pci_device_shutdown,
+	.async_shutdown_start	= pci_device_async_shutdown_start,
+	.async_shutdown_end	= pci_device_async_shutdown_end,
 	.dev_groups	= pci_dev_groups,
 	.bus_groups	= pci_bus_groups,
 	.drv_groups	= pci_drv_groups,
diff --git a/include/linux/pci.h b/include/linux/pci.h
index b957eeb89c7a..bbdf7d52e87b 100644
--- a/include/linux/pci.h
+++ b/include/linux/pci.h
@@ -881,6 +881,8 @@ struct module;
  *		Useful for enabling wake-on-lan (NIC) or changing
  *		the power state of a device before reboot.
  *		e.g. drivers/net/e100.c.
+ * @async_shutdown_start:	This starts the asynchronous shutdown
+ * @async_shutdown_end:	This completes the started asynchronous shutdown
  * @sriov_configure: Optional driver callback to allow configuration of
  *		number of VFs to enable via sysfs "sriov_numvfs" file.
  * @sriov_set_msix_vec_count: PF Driver callback to change number of MSI-X
@@ -905,6 +907,8 @@ struct pci_driver {
 	int  (*suspend)(struct pci_dev *dev, pm_message_t state);	/* Device suspended */
 	int  (*resume)(struct pci_dev *dev);	/* Device woken up */
 	void (*shutdown)(struct pci_dev *dev);
+	void (*async_shutdown_start)(struct pci_dev *dev);
+	void (*async_shutdown_end)(struct pci_dev *dev);
 	int  (*sriov_configure)(struct pci_dev *dev, int num_vfs); /* On PF */
 	int  (*sriov_set_msix_vec_count)(struct pci_dev *vf, int msix_vec_count); /* On PF */
 	u32  (*sriov_get_vf_total_msix)(struct pci_dev *pf);
-- 
2.36.0.rc0.470.gd361397f0d-goog