PCI: allow assignment of memory resources with a specified alignment

This patch allows memory resources to be assigned with a specified
alignment at boot-time or run-time. The patch is useful when we use PCI
pass-through, because page-aligned memory resources are required to
securely share PCI resources with guest drivers.

If you want to assign the resource at boot time, please set
"pci=resource_alignment=" boot parameter.

This is format of "pci=resource_alignment=" boot parameter:

        [<order of align>@][<domain>:]<bus>:<slot>.<func>[; ...]
                Specifies alignment and device to reassign
                aligned memory resources.
                If <order of align> is not specified, PAGE_SIZE is
                used as alignment.
                PCI-PCI bridge can be specified, if resource
                windows need to be expanded.

This is example:

        pci=resource_alignment=20@07:00.0;18@0f:00.0;00:1d.7

If you want to assign the resource at run-time, please set
"/sys/bus/pci/resource_alignment" file, and hot-remove the device and
hot-add the device.  For this purpose, fakephp or PCI hotplug interfaces
can be used.

The format of "/sys/bus/pci/resource_alignment" file is the same with
boot parameter. You can use "," instead of ";".

For example:

        # cd /sys/bus/pci
        # echo -n 20@12:00.0 > resource_alignment
        # echo 1 > devices/0000:12:00.0/remove
        # echo 1 > rescan
Reviewed-by: NAlex Chiang <achiang@hp.com>
Reviewed-by: NYu Zhao <yu.zhao@intel.com>
Signed-off-by: NYuji Shimada <shimada-yxb@necst.nec.co.jp>
Signed-off-by: NJesse Barnes <jbarnes@virtuousgeek.org>

Documentation/kernel-parameters.txt

@@ -1760,6 +1760,15 @@ and is between 256 and 4096 characters. It is defined in the file
 		cbmemsize=nn[KMG]	The fixed amount of bus space which is
 				reserved for the CardBus bridge's memory
 				window. The default value is 64 megabytes.
+		resource_alignment=
+				Format:
+				[<order of align>@][<domain>:]<bus>:<slot>.<func>[; ...]
+				Specifies alignment and device to reassign
+				aligned memory resources.
+				If <order of align> is not specified,
+				PAGE_SIZE is used as alignment.
+				PCI-PCI bridge can be specified, if resource
+				windows need to be expanded.
 
 	pcie_aspm=	[PCIE] Forcibly enable or disable PCIe Active State Power
 			Management.

drivers/pci/pci.c

@@ -20,6 +20,8 @@
 #include <linux/pm_wakeup.h>
 #include <linux/interrupt.h>
 #include <asm/dma.h>	/* isa_dma_bridge_buggy */
+#include <linux/device.h>
+#include <asm/setup.h>
 #include "pci.h"
 
 unsigned int pci_pm_d3_delay = PCI_PM_D3_WAIT;
@@ -2370,6 +2372,121 @@ int pci_resource_bar(struct pci_dev *dev, int resno, enum pci_bar_type *type)
 	return 0;
 }
 
+#define RESOURCE_ALIGNMENT_PARAM_SIZE COMMAND_LINE_SIZE
+static char resource_alignment_param[RESOURCE_ALIGNMENT_PARAM_SIZE] = {0};
+spinlock_t resource_alignment_lock = SPIN_LOCK_UNLOCKED;
+
+/**
+ * pci_specified_resource_alignment - get resource alignment specified by user.
+ * @dev: the PCI device to get
+ *
+ * RETURNS: Resource alignment if it is specified.
+ *          Zero if it is not specified.
+ */
+resource_size_t pci_specified_resource_alignment(struct pci_dev *dev)
+{
+	int seg, bus, slot, func, align_order, count;
+	resource_size_t align = 0;
+	char *p;
+
+	spin_lock(&resource_alignment_lock);
+	p = resource_alignment_param;
+	while (*p) {
+		count = 0;
+		if (sscanf(p, "%d%n", &align_order, &count) == 1 &&
+							p[count] == '@') {
+			p += count + 1;
+		} else {
+			align_order = -1;
+		}
+		if (sscanf(p, "%x:%x:%x.%x%n",
+			&seg, &bus, &slot, &func, &count) != 4) {
+			seg = 0;
+			if (sscanf(p, "%x:%x.%x%n",
+					&bus, &slot, &func, &count) != 3) {
+				/* Invalid format */
+				printk(KERN_ERR "PCI: Can't parse resource_alignment parameter: %s\n",
+					p);
+				break;
+			}
+		}
+		p += count;
+		if (seg == pci_domain_nr(dev->bus) &&
+			bus == dev->bus->number &&
+			slot == PCI_SLOT(dev->devfn) &&
+			func == PCI_FUNC(dev->devfn)) {
+			if (align_order == -1) {
+				align = PAGE_SIZE;
+			} else {
+				align = 1 << align_order;
+			}
+			/* Found */
+			break;
+		}
+		if (*p != ';' && *p != ',') {
+			/* End of param or invalid format */
+			break;
+		}
+		p++;
+	}
+	spin_unlock(&resource_alignment_lock);
+	return align;
+}
+
+/**
+ * pci_is_reassigndev - check if specified PCI is target device to reassign
+ * @dev: the PCI device to check
+ *
+ * RETURNS: non-zero for PCI device is a target device to reassign,
+ *          or zero is not.
+ */
+int pci_is_reassigndev(struct pci_dev *dev)
+{
+	return (pci_specified_resource_alignment(dev) != 0);
+}
+
+ssize_t pci_set_resource_alignment_param(const char *buf, size_t count)
+{
+	if (count > RESOURCE_ALIGNMENT_PARAM_SIZE - 1)
+		count = RESOURCE_ALIGNMENT_PARAM_SIZE - 1;
+	spin_lock(&resource_alignment_lock);
+	strncpy(resource_alignment_param, buf, count);
+	resource_alignment_param[count] = '\0';
+	spin_unlock(&resource_alignment_lock);
+	return count;
+}
+
+ssize_t pci_get_resource_alignment_param(char *buf, size_t size)
+{
+	size_t count;
+	spin_lock(&resource_alignment_lock);
+	count = snprintf(buf, size, "%s", resource_alignment_param);
+	spin_unlock(&resource_alignment_lock);
+	return count;
+}
+
+static ssize_t pci_resource_alignment_show(struct bus_type *bus, char *buf)
+{
+	return pci_get_resource_alignment_param(buf, PAGE_SIZE);
+}
+
+static ssize_t pci_resource_alignment_store(struct bus_type *bus,
+					const char *buf, size_t count)
+{
+	return pci_set_resource_alignment_param(buf, count);
+}
+
+BUS_ATTR(resource_alignment, 0644, pci_resource_alignment_show,
+					pci_resource_alignment_store);
+
+static int __init pci_resource_alignment_sysfs_init(void)
+{
+	return bus_create_file(&pci_bus_type,
+					&bus_attr_resource_alignment);
+}
+
+late_initcall(pci_resource_alignment_sysfs_init);
+
 static void __devinit pci_no_domains(void)
 {
 #ifdef CONFIG_PCI_DOMAINS
@@ -2418,6 +2535,9 @@ static int __init pci_setup(char *str)
 				pci_cardbus_io_size = memparse(str + 9, &str);
 			} else if (!strncmp(str, "cbmemsize=", 10)) {
 				pci_cardbus_mem_size = memparse(str + 10, &str);
+			} else if (!strncmp(str, "resource_alignment=", 19)) {
+				pci_set_resource_alignment_param(str + 19,
+							strlen(str + 19));
 			} else {
 				printk(KERN_ERR "PCI: Unknown option `%s'\n",
 						str);

drivers/pci/pci.h

@@ -195,4 +195,10 @@ static inline int pci_ari_enabled(struct pci_bus *bus)
 	return bus->self && bus->self->ari_enabled;
 }
 
+#ifdef CONFIG_PCI_QUIRKS
+extern int pci_is_reassigndev(struct pci_dev *dev);
+resource_size_t pci_specified_resource_alignment(struct pci_dev *dev);
+extern void pci_disable_bridge_window(struct pci_dev *dev);
+#endif
+
 #endif /* DRIVERS_PCI_H */

drivers/pci/quirks.c

@@ -24,6 +24,7 @@
 #include <linux/kallsyms.h>
 #include <linux/dmi.h>
 #include <linux/pci-aspm.h>
+#include <linux/ioport.h>
 #include "pci.h"
 
 int isa_dma_bridge_buggy;
@@ -34,6 +35,65 @@ int pcie_mch_quirk;
 EXPORT_SYMBOL(pcie_mch_quirk);
 
 #ifdef CONFIG_PCI_QUIRKS
+/*
+ * This quirk function disables the device and releases resources
+ * which is specified by kernel's boot parameter 'pci=resource_alignment='.
+ * It also rounds up size to specified alignment.
+ * Later on, the kernel will assign page-aligned memory resource back
+ * to that device.
+ */
+static void __devinit quirk_resource_alignment(struct pci_dev *dev)
+{
+	int i;
+	struct resource *r;
+	resource_size_t align, size;
+
+	if (!pci_is_reassigndev(dev))
+		return;
+
+	if (dev->hdr_type == PCI_HEADER_TYPE_NORMAL &&
+	    (dev->class >> 8) == PCI_CLASS_BRIDGE_HOST) {
+		dev_warn(&dev->dev,
+			"Can't reassign resources to host bridge.\n");
+		return;
+	}
+
+	dev_info(&dev->dev, "Disabling device and release resources.\n");
+	pci_disable_device(dev);
+
+	align = pci_specified_resource_alignment(dev);
+	for (i=0; i < PCI_BRIDGE_RESOURCES; i++) {
+		r = &dev->resource[i];
+		if (!(r->flags & IORESOURCE_MEM))
+			continue;
+		size = resource_size(r);
+		if (size < align) {
+			size = align;
+			dev_info(&dev->dev,
+				"Rounding up size of resource #%d to %#llx.\n",
+				i, (unsigned long long)size);
+		}
+		r->end = size - 1;
+		r->start = 0;
+	}
+	/* Need to disable bridge's resource window,
+	 * to enable the kernel to reassign new resource
+	 * window later on.
+	 */
+	if (dev->hdr_type == PCI_HEADER_TYPE_BRIDGE &&
+	    (dev->class >> 8) == PCI_CLASS_BRIDGE_PCI) {
+		for (i = PCI_BRIDGE_RESOURCES; i < PCI_NUM_RESOURCES; i++) {
+			r = &dev->resource[i];
+			if (!(r->flags & IORESOURCE_MEM))
+				continue;
+			r->end = resource_size(r) - 1;
+			r->start = 0;
+		}
+		pci_disable_bridge_window(dev);
+	}
+}
+DECLARE_PCI_FIXUP_HEADER(PCI_ANY_ID, PCI_ANY_ID, quirk_resource_alignment);
+
 /* The Mellanox Tavor device gives false positive parity errors
  * Mark this device with a broken_parity_status, to allow
  * PCI scanning code to "skip" this now blacklisted device.

drivers/pci/setup-res.c

@@ -120,6 +120,21 @@ int pci_claim_resource(struct pci_dev *dev, int resource)
 	return err;
 }
 
+#ifdef CONFIG_PCI_QUIRKS
+void pci_disable_bridge_window(struct pci_dev *dev)
+{
+	dev_dbg(&dev->dev, "Disabling bridge window.\n");
+
+	/* MMIO Base/Limit */
+	pci_write_config_dword(dev, PCI_MEMORY_BASE, 0x0000fff0);
+
+	/* Prefetchable MMIO Base/Limit */
+	pci_write_config_dword(dev, PCI_PREF_LIMIT_UPPER32, 0);
+	pci_write_config_dword(dev, PCI_PREF_MEMORY_BASE, 0x0000fff0);
+	pci_write_config_dword(dev, PCI_PREF_BASE_UPPER32, 0xffffffff);
+}
+#endif	/* CONFIG_PCI_QUIRKS */
+
 int pci_assign_resource(struct pci_dev *dev, int resno)
 {
 	struct pci_bus *bus = dev->bus;