From: Takao Indoh <indou.takao@jp.fujitsu.com>
To: bhelgaas@google.com
Cc: linux-pci@vger.kernel.org, kexec@lists.infradead.org,
iommu@lists.linux-foundation.org, linux-kernel@vger.kernel.org,
alex.williamson@redhat.com, ddutile@redhat.com,
ebiederm@xmission.com, ishii.hironobu@jp.fujitsu.com,
bill.sumner@hp.com
Subject: Re: [PATCH v2] PCI: Reset PCIe devices to stop ongoing DMA
Date: Wed, 22 May 2013 08:46:57 +0900 [thread overview]
Message-ID: <519C0771.2020507@jp.fujitsu.com> (raw)
In-Reply-To: <87d2stw74i.fsf@xmission.com>
Hi Bjorn,
Any comments, ack/nack?
Thanks,
Takao Indoh
(2013/05/15 7:04), Eric W. Biederman wrote:
> Takao Indoh <indou.takao@jp.fujitsu.com> writes:
>
>> This patch resets PCIe devices on boot to stop ongoing DMA. When
>> "pci=pcie_reset_endpoint_devices" is specified, a hot reset is triggered
>> on each PCIe root port and downstream port to reset its downstream
>> endpoint.
>>
>> Problem:
>> This patch solves the problem that kdump can fail when intel_iommu=on is
>> specified. When intel_iommu=on is specified, many dma-remapping errors
>> occur in second kernel and it causes problems like driver error or PCI
>> SERR, at last kdump fails. This problem is caused as follows.
>> 1) Devices are working on first kernel.
>> 2) Switch to second kernel(kdump kernel). The devices are still working
>> and its DMA continues during this switch.
>> 3) iommu is initialized during second kernel boot and ongoing DMA causes
>> dma-remapping errors.
>>
>> Solution:
>> All DMA transactions have to be stopped before iommu is initialized. By
>> this patch devices are reset and in-flight DMA is stopped before
>> pci_iommu_init.
>>
>> To invoke hot reset on an endpoint, its upstream link need to be reset.
>> reset_pcie_endpoints() is called from fs_initcall_sync, and it finds
>> root port/downstream port whose child is PCIe endpoint, and then reset
>> link between them. If the endpoint is VGA device, it is skipped because
>> the monitor blacks out if VGA controller is reset.
>
> At a quick skim this patch looks reasonable.
>
> Acked-by: "Eric W. Biederman" <ebiederm@xmission.com>
>
>> Changelog:
>> v2:
>> - Read pci config before de-assert secondary bus reset to flush previous
>> write
>> - Change function/variable name
>> - Make a list of devices to be reset
>>
>> v1:
>> https://patchwork.kernel.org/patch/2482291/
>>
>> Signed-off-by: Takao Indoh <indou.takao@jp.fujitsu.com>
>> ---
>> Documentation/kernel-parameters.txt | 2 +
>> drivers/pci/pci.c | 113 +++++++++++++++++++++++++++++++++++
>> 2 files changed, 115 insertions(+), 0 deletions(-)
>>
>> diff --git a/Documentation/kernel-parameters.txt b/Documentation/kernel-parameters.txt
>> index c3bfacb..8c9e8e4 100644
>> --- a/Documentation/kernel-parameters.txt
>> +++ b/Documentation/kernel-parameters.txt
>> @@ -2321,6 +2321,8 @@ bytes respectively. Such letter suffixes can also be entirely omitted.
>> any pair of devices, possibly at the cost of
>> reduced performance. This also guarantees
>> that hot-added devices will work.
>> + pcie_reset_endpoint_devices Reset PCIe endpoint on boot by
>> + hot reset
>> cbiosize=nn[KMG] The fixed amount of bus space which is
>> reserved for the CardBus bridge's IO window.
>> The default value is 256 bytes.
>> diff --git a/drivers/pci/pci.c b/drivers/pci/pci.c
>> index a899d8b..70c1205 100644
>> --- a/drivers/pci/pci.c
>> +++ b/drivers/pci/pci.c
>> @@ -3873,6 +3873,116 @@ void __weak pci_fixup_cardbus(struct pci_bus *bus)
>> }
>> EXPORT_SYMBOL(pci_fixup_cardbus);
>>
>> +/*
>> + * Return true if dev is PCIe root port or downstream port whose child is PCIe
>> + * endpoint except VGA device.
>> + */
>> +static int __init need_reset(struct pci_dev *dev)
>> +{
>> + struct pci_bus *subordinate;
>> + struct pci_dev *child;
>> +
>> + if (!pci_is_pcie(dev) || !dev->subordinate ||
>> + list_empty(&dev->subordinate->devices) ||
>> + ((pci_pcie_type(dev) != PCI_EXP_TYPE_ROOT_PORT) &&
>> + (pci_pcie_type(dev) != PCI_EXP_TYPE_DOWNSTREAM)))
>> + return 0;
>> +
>> + subordinate = dev->subordinate;
>> + list_for_each_entry(child, &subordinate->devices, bus_list) {
>> + if ((pci_pcie_type(child) == PCI_EXP_TYPE_UPSTREAM) ||
>> + (pci_pcie_type(child) == PCI_EXP_TYPE_PCI_BRIDGE) ||
>> + ((child->class >> 16) == PCI_BASE_CLASS_DISPLAY))
>> + /* Don't reset switch, bridge, VGA device */
>> + return 0;
>> + }
>> +
>> + return 1;
>> +}
>> +
>> +static void __init save_downstream_configs(struct pci_dev *dev)
>> +{
>> + struct pci_bus *subordinate;
>> + struct pci_dev *child;
>> +
>> + subordinate = dev->subordinate;
>> + list_for_each_entry(child, &subordinate->devices, bus_list) {
>> + dev_info(&child->dev, "save state\n");
>> + pci_save_state(child);
>> + }
>> +}
>> +
>> +static void __init restore_downstream_configs(struct pci_dev *dev)
>> +{
>> + struct pci_bus *subordinate;
>> + struct pci_dev *child;
>> +
>> + subordinate = dev->subordinate;
>> + list_for_each_entry(child, &subordinate->devices, bus_list) {
>> + dev_info(&child->dev, "restore state\n");
>> + pci_restore_state(child);
>> + }
>> +}
>> +
>> +static void __init do_downstream_device_reset(struct pci_dev *dev)
>> +{
>> + u16 ctrl;
>> +
>> + dev_info(&dev->dev, "Reset Secondary bus\n");
>> +
>> + /* Assert Secondary Bus Reset */
>> + pci_read_config_word(dev, PCI_BRIDGE_CONTROL, &ctrl);
>> + ctrl |= PCI_BRIDGE_CTL_BUS_RESET;
>> + pci_write_config_word(dev, PCI_BRIDGE_CONTROL, ctrl);
>> +
>> + /* Read config again to flush previous write */
>> + pci_read_config_word(dev, PCI_BRIDGE_CONTROL, &ctrl);
>> +
>> + msleep(2);
>> +
>> + /* De-assert Secondary Bus Reset */
>> + ctrl &= ~PCI_BRIDGE_CTL_BUS_RESET;
>> + pci_write_config_word(dev, PCI_BRIDGE_CONTROL, ctrl);
>> +}
>> +
>> +struct pci_dev_entry {
>> + struct list_head list;
>> + struct pci_dev *dev;
>> +};
>> +static int __initdata pcie_reset_endpoint_devices;
>> +static int __init reset_pcie_endpoints(void)
>> +{
>> + struct pci_dev *dev = NULL;
>> + struct pci_dev_entry *pdev_entry, *tmp;
>> + LIST_HEAD(pdev_list);
>> +
>> + if (!pcie_reset_endpoint_devices)
>> + return 0;
>> +
>> + for_each_pci_dev(dev)
>> + if (need_reset(dev)) {
>> + pdev_entry = kmalloc(sizeof(*pdev_entry), GFP_KERNEL);
>> + pdev_entry->dev = dev;
>> + list_add(&pdev_entry->list, &pdev_list);
>> + }
>> +
>> + list_for_each_entry(pdev_entry, &pdev_list, list)
>> + save_downstream_configs(pdev_entry->dev);
>> +
>> + list_for_each_entry(pdev_entry, &pdev_list, list)
>> + do_downstream_device_reset(pdev_entry->dev);
>> +
>> + msleep(1000);
>> +
>> + list_for_each_entry_safe(pdev_entry, tmp, &pdev_list, list) {
>> + restore_downstream_configs(pdev_entry->dev);
>> + kfree(pdev_entry);
>> + }
>> +
>> + return 0;
>> +}
>> +fs_initcall_sync(reset_pcie_endpoints);
>> +
>> static int __init pci_setup(char *str)
>> {
>> while (str) {
>> @@ -3915,6 +4025,9 @@ static int __init pci_setup(char *str)
>> pcie_bus_config = PCIE_BUS_PEER2PEER;
>> } else if (!strncmp(str, "pcie_scan_all", 13)) {
>> pci_add_flags(PCI_SCAN_ALL_PCIE_DEVS);
>> + } else if (!strncmp(str, "pcie_reset_endpoint_devices",
>> + 27)) {
>> + pcie_reset_endpoint_devices = 1;
>> } else {
>> printk(KERN_ERR "PCI: Unknown option `%s'\n",
>> str);
>
>
_______________________________________________
kexec mailing list
kexec@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/kexec
next prev parent reply other threads:[~2013-05-21 23:47 UTC|newest]
Thread overview: 34+ messages / expand[flat|nested] mbox.gz Atom feed top
2013-05-14 5:29 [PATCH v2] PCI: Reset PCIe devices to stop ongoing DMA Takao Indoh
2013-05-14 22:04 ` Eric W. Biederman
2013-05-21 23:46 ` Takao Indoh [this message]
2013-06-06 7:25 ` Takao Indoh
2013-06-07 4:14 ` Bjorn Helgaas
2013-06-07 8:46 ` Takao Indoh
2013-06-11 2:20 ` Bjorn Helgaas
2013-06-11 6:08 ` Takao Indoh
2013-06-11 23:19 ` Sumner, William
2013-06-12 0:53 ` Bjorn Helgaas
2013-06-12 13:19 ` Don Dutile
2013-06-13 3:25 ` Takao Indoh
2013-06-12 4:45 ` Bjorn Helgaas
2013-06-13 2:44 ` Takao Indoh
2013-06-13 3:41 ` Bjorn Helgaas
2013-06-14 2:11 ` Takao Indoh
2013-07-24 6:29 ` Takao Indoh
2013-07-25 14:24 ` Vivek Goyal
2013-07-29 0:20 ` Takao Indoh
2013-07-25 17:00 ` Bjorn Helgaas
2013-07-29 0:37 ` Takao Indoh
2013-07-29 14:17 ` Bjorn Helgaas
2013-07-30 6:09 ` Takao Indoh
2013-07-30 15:59 ` Bjorn Helgaas
2013-07-31 0:35 ` Takao Indoh
2013-07-31 3:11 ` Alex Williamson
2013-07-31 5:50 ` Takao Indoh
2013-07-31 21:08 ` Bjorn Helgaas
2013-07-31 21:23 ` Rafael J. Wysocki
2013-08-01 6:34 ` Takao Indoh
2013-08-01 12:42 ` Alex Williamson
2013-08-01 13:20 ` Vivek Goyal
2013-07-31 19:56 ` Vivek Goyal
2013-07-31 16:09 ` Vivek Goyal
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=519C0771.2020507@jp.fujitsu.com \
--to=indou.takao@jp.fujitsu.com \
--cc=alex.williamson@redhat.com \
--cc=bhelgaas@google.com \
--cc=bill.sumner@hp.com \
--cc=ddutile@redhat.com \
--cc=ebiederm@xmission.com \
--cc=iommu@lists.linux-foundation.org \
--cc=ishii.hironobu@jp.fujitsu.com \
--cc=kexec@lists.infradead.org \
--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