NUMA systems with ACPI normally describe the physical topology via _PXM
methods. But many BIOSes don't implement _PXM, which leaves the kernel
with no way to discover the device topology, which reduces performance
because we can't put memory and processes close to the device.
The NUMA node of a PCI device is already exported in the sysfs "numa_node"
file. Make that file writable so users can workaround the lack of _PXM
methods in the BIOS. For example:
echo 3 > /sys/devices/pci0000:ff/0000:03:1f.3/numa_node
sets the node for PCI device 0000:03:1f.3.
Writing the file emits a FW_BUG warning to encourage users to request
firmware updates. It also taints the kernel with TAINT_FIRMWARE_WORKAROUND
because overriding the node incorrectly can cause performance issues.
[bhelgaas: changelog, documentation text]
Signed-off-by: Prarit Bhargava <prarit@redhat.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
CC: Myron Stowe <mstowe@redhat.com>
CC: Alexander Ducyk <alexander.h.duyck@redhat.com>
CC: Jiang Liu <jiang.liu@linux.intel.com>
opt-out of driver binding using a driver_override name such as
"none". Only a single driver may be specified in the override,
there is no support for parsing delimiters.
opt-out of driver binding using a driver_override name such as
"none". Only a single driver may be specified in the override,
there is no support for parsing delimiters.
+
+What: /sys/bus/pci/devices/.../numa_node
+Date: Oct 2014
+Contact: Prarit Bhargava <prarit@redhat.com>
+Description:
+ This file contains the NUMA node to which the PCI device is
+ attached, or -1 if the node is unknown. The initial value
+ comes from an ACPI _PXM method or a similar firmware
+ source. If that is missing or incorrect, this file can be
+ written to override the node. In that case, please report
+ a firmware bug to the system vendor. Writing to this file
+ taints the kernel with TAINT_FIRMWARE_WORKAROUND, which
+ reduces the supportability of your system.
static DEVICE_ATTR_RW(enabled);
#ifdef CONFIG_NUMA
static DEVICE_ATTR_RW(enabled);
#ifdef CONFIG_NUMA
+static ssize_t numa_node_store(struct device *dev,
+ struct device_attribute *attr, const char *buf,
+ size_t count)
+{
+ struct pci_dev *pdev = to_pci_dev(dev);
+ int node, ret;
+
+ if (!capable(CAP_SYS_ADMIN))
+ return -EPERM;
+
+ ret = kstrtoint(buf, 0, &node);
+ if (ret)
+ return ret;
+
+ if (!node_online(node))
+ return -EINVAL;
+
+ add_taint(TAINT_FIRMWARE_WORKAROUND, LOCKDEP_STILL_OK);
+ dev_alert(&pdev->dev, FW_BUG "Overriding NUMA node to %d. Contact your vendor for updates.",
+ node);
+
+ dev->numa_node = node;
+ return count;
+}
+
static ssize_t numa_node_show(struct device *dev, struct device_attribute *attr,
char *buf)
{
return sprintf(buf, "%d\n", dev->numa_node);
}
static ssize_t numa_node_show(struct device *dev, struct device_attribute *attr,
char *buf)
{
return sprintf(buf, "%d\n", dev->numa_node);
}
-static DEVICE_ATTR_RO(numa_node);
+static DEVICE_ATTR_RW(numa_node);
#endif
static ssize_t dma_mask_bits_show(struct device *dev,
#endif
static ssize_t dma_mask_bits_show(struct device *dev,