2005-04-16 15:20:36 -07:00
|
|
|
/*
|
|
|
|
* probe.c - PCI detection and setup code
|
|
|
|
*/
|
|
|
|
|
|
|
|
#include <linux/kernel.h>
|
|
|
|
#include <linux/delay.h>
|
|
|
|
#include <linux/init.h>
|
|
|
|
#include <linux/pci.h>
|
|
|
|
#include <linux/slab.h>
|
|
|
|
#include <linux/module.h>
|
|
|
|
#include <linux/cpumask.h>
|
PCI: add PCI Express ASPM support
PCI Express ASPM defines a protocol for PCI Express components in the D0
state to reduce Link power by placing their Links into a low power state
and instructing the other end of the Link to do likewise. This
capability allows hardware-autonomous, dynamic Link power reduction
beyond what is achievable by software-only controlled power management.
However, The device should be configured by software appropriately.
Enabling ASPM will save power, but will introduce device latency.
This patch adds ASPM support in Linux. It introduces a global policy for
ASPM, a sysfs file /sys/module/pcie_aspm/parameters/policy can control
it. The interface can be used as a boot option too. Currently we have
below setting:
-default, BIOS default setting
-powersave, highest power saving mode, enable all available ASPM
state and clock power management
-performance, highest performance, disable ASPM and clock power
management
By default, the 'default' policy is used currently.
In my test, power difference between powersave mode and performance mode
is about 1.3w in a system with 3 PCIE links.
Note: some devices might not work well with aspm, either because chipset
issue or device issue. The patch provide API (pci_disable_link_state),
driver can disable ASPM for specific device.
Signed-off-by: Shaohua Li <shaohua.li@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
2008-02-24 18:46:41 -07:00
|
|
|
#include <linux/pci-aspm.h>
|
2005-04-07 22:53:31 -07:00
|
|
|
#include "pci.h"
|
2005-04-16 15:20:36 -07:00
|
|
|
|
|
|
|
#define CARDBUS_LATENCY_TIMER 176 /* secondary latency timer */
|
|
|
|
#define CARDBUS_RESERVE_BUSNR 3
|
|
|
|
|
|
|
|
/* Ugh. Need to stop exporting this to modules. */
|
|
|
|
LIST_HEAD(pci_root_buses);
|
|
|
|
EXPORT_SYMBOL(pci_root_buses);
|
|
|
|
|
2008-02-13 23:30:39 -07:00
|
|
|
|
|
|
|
static int find_anything(struct device *dev, void *data)
|
|
|
|
{
|
|
|
|
return 1;
|
|
|
|
}
|
2005-04-16 15:20:36 -07:00
|
|
|
|
2007-07-15 23:39:39 -07:00
|
|
|
/*
|
|
|
|
* Some device drivers need know if pci is initiated.
|
|
|
|
* Basically, we think pci is not initiated when there
|
2008-02-13 23:30:39 -07:00
|
|
|
* is no device to be found on the pci_bus_type.
|
2007-07-15 23:39:39 -07:00
|
|
|
*/
|
|
|
|
int no_pci_devices(void)
|
|
|
|
{
|
2008-02-13 23:30:39 -07:00
|
|
|
struct device *dev;
|
|
|
|
int no_devices;
|
2007-07-15 23:39:39 -07:00
|
|
|
|
2008-02-13 23:30:39 -07:00
|
|
|
dev = bus_find_device(&pci_bus_type, NULL, NULL, find_anything);
|
|
|
|
no_devices = (dev == NULL);
|
|
|
|
put_device(dev);
|
|
|
|
return no_devices;
|
|
|
|
}
|
2007-07-15 23:39:39 -07:00
|
|
|
EXPORT_SYMBOL(no_pci_devices);
|
|
|
|
|
2005-04-16 15:20:36 -07:00
|
|
|
/*
|
|
|
|
* PCI Bus Class Devices
|
|
|
|
*/
|
2007-05-22 19:47:54 -07:00
|
|
|
static ssize_t pci_bus_show_cpuaffinity(struct device *dev,
|
2008-04-08 11:43:03 -07:00
|
|
|
int type,
|
2007-05-22 19:47:54 -07:00
|
|
|
struct device_attribute *attr,
|
2005-09-10 00:25:49 -07:00
|
|
|
char *buf)
|
2005-04-16 15:20:36 -07:00
|
|
|
{
|
|
|
|
int ret;
|
2005-09-10 00:25:49 -07:00
|
|
|
cpumask_t cpumask;
|
2005-04-16 15:20:36 -07:00
|
|
|
|
2007-05-22 19:47:54 -07:00
|
|
|
cpumask = pcibus_to_cpumask(to_pci_bus(dev));
|
2008-04-08 11:43:03 -07:00
|
|
|
ret = type?
|
2008-12-13 03:50:25 -07:00
|
|
|
cpulist_scnprintf(buf, PAGE_SIZE-2, &cpumask) :
|
|
|
|
cpumask_scnprintf(buf, PAGE_SIZE-2, &cpumask);
|
2008-04-08 11:43:03 -07:00
|
|
|
buf[ret++] = '\n';
|
|
|
|
buf[ret] = '\0';
|
2005-04-16 15:20:36 -07:00
|
|
|
return ret;
|
|
|
|
}
|
2008-04-08 11:43:03 -07:00
|
|
|
|
|
|
|
static ssize_t inline pci_bus_show_cpumaskaffinity(struct device *dev,
|
|
|
|
struct device_attribute *attr,
|
|
|
|
char *buf)
|
|
|
|
{
|
|
|
|
return pci_bus_show_cpuaffinity(dev, 0, attr, buf);
|
|
|
|
}
|
|
|
|
|
|
|
|
static ssize_t inline pci_bus_show_cpulistaffinity(struct device *dev,
|
|
|
|
struct device_attribute *attr,
|
|
|
|
char *buf)
|
|
|
|
{
|
|
|
|
return pci_bus_show_cpuaffinity(dev, 1, attr, buf);
|
|
|
|
}
|
|
|
|
|
|
|
|
DEVICE_ATTR(cpuaffinity, S_IRUGO, pci_bus_show_cpumaskaffinity, NULL);
|
|
|
|
DEVICE_ATTR(cpulistaffinity, S_IRUGO, pci_bus_show_cpulistaffinity, NULL);
|
2005-04-16 15:20:36 -07:00
|
|
|
|
|
|
|
/*
|
|
|
|
* PCI Bus Class
|
|
|
|
*/
|
2007-05-22 19:47:54 -07:00
|
|
|
static void release_pcibus_dev(struct device *dev)
|
2005-04-16 15:20:36 -07:00
|
|
|
{
|
2007-05-22 19:47:54 -07:00
|
|
|
struct pci_bus *pci_bus = to_pci_bus(dev);
|
2005-04-16 15:20:36 -07:00
|
|
|
|
|
|
|
if (pci_bus->bridge)
|
|
|
|
put_device(pci_bus->bridge);
|
|
|
|
kfree(pci_bus);
|
|
|
|
}
|
|
|
|
|
|
|
|
static struct class pcibus_class = {
|
|
|
|
.name = "pci_bus",
|
2007-05-22 19:47:54 -07:00
|
|
|
.dev_release = &release_pcibus_dev,
|
2005-04-16 15:20:36 -07:00
|
|
|
};
|
|
|
|
|
|
|
|
static int __init pcibus_class_init(void)
|
|
|
|
{
|
|
|
|
return class_register(&pcibus_class);
|
|
|
|
}
|
|
|
|
postcore_initcall(pcibus_class_init);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Translate the low bits of the PCI base
|
|
|
|
* to the resource type
|
|
|
|
*/
|
|
|
|
static inline unsigned int pci_calc_resource_flags(unsigned int flags)
|
|
|
|
{
|
|
|
|
if (flags & PCI_BASE_ADDRESS_SPACE_IO)
|
|
|
|
return IORESOURCE_IO;
|
|
|
|
|
|
|
|
if (flags & PCI_BASE_ADDRESS_MEM_PREFETCH)
|
|
|
|
return IORESOURCE_MEM | IORESOURCE_PREFETCH;
|
|
|
|
|
|
|
|
return IORESOURCE_MEM;
|
|
|
|
}
|
|
|
|
|
2008-07-28 10:38:59 -07:00
|
|
|
static u64 pci_size(u64 base, u64 maxbase, u64 mask)
|
2005-04-16 15:20:36 -07:00
|
|
|
{
|
2008-07-28 10:38:59 -07:00
|
|
|
u64 size = mask & maxbase; /* Find the significant bits */
|
2005-04-16 15:20:36 -07:00
|
|
|
if (!size)
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
/* Get the lowest of them to find the decode size, and
|
|
|
|
from that the extent. */
|
|
|
|
size = (size & ~(size-1)) - 1;
|
|
|
|
|
|
|
|
/* base == maxbase can be valid only if the BAR has
|
|
|
|
already been programmed with all 1s. */
|
|
|
|
if (base == maxbase && ((base | size) & mask) != mask)
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
return size;
|
|
|
|
}
|
|
|
|
|
2008-07-28 10:38:59 -07:00
|
|
|
enum pci_bar_type {
|
|
|
|
pci_bar_unknown, /* Standard PCI BAR probe */
|
|
|
|
pci_bar_io, /* An io port BAR */
|
|
|
|
pci_bar_mem32, /* A 32-bit memory BAR */
|
|
|
|
pci_bar_mem64, /* A 64-bit memory BAR */
|
|
|
|
};
|
2006-11-29 14:53:10 -07:00
|
|
|
|
2008-07-28 10:38:59 -07:00
|
|
|
static inline enum pci_bar_type decode_bar(struct resource *res, u32 bar)
|
|
|
|
{
|
|
|
|
if ((bar & PCI_BASE_ADDRESS_SPACE) == PCI_BASE_ADDRESS_SPACE_IO) {
|
|
|
|
res->flags = bar & ~PCI_BASE_ADDRESS_IO_MASK;
|
|
|
|
return pci_bar_io;
|
|
|
|
}
|
2006-11-29 14:53:10 -07:00
|
|
|
|
2008-07-28 10:38:59 -07:00
|
|
|
res->flags = bar & ~PCI_BASE_ADDRESS_MEM_MASK;
|
2006-11-29 14:53:10 -07:00
|
|
|
|
2008-10-12 17:49:04 -07:00
|
|
|
if (res->flags & PCI_BASE_ADDRESS_MEM_TYPE_64)
|
2008-07-28 10:38:59 -07:00
|
|
|
return pci_bar_mem64;
|
|
|
|
return pci_bar_mem32;
|
2006-11-29 14:53:10 -07:00
|
|
|
}
|
|
|
|
|
2008-07-28 10:38:59 -07:00
|
|
|
/*
|
|
|
|
* If the type is not unknown, we assume that the lowest bit is 'enable'.
|
|
|
|
* Returns 1 if the BAR was 64-bit and 0 if it was 32-bit.
|
|
|
|
*/
|
|
|
|
static int __pci_read_base(struct pci_dev *dev, enum pci_bar_type type,
|
|
|
|
struct resource *res, unsigned int pos)
|
2006-11-29 14:53:10 -07:00
|
|
|
{
|
2008-07-28 10:38:59 -07:00
|
|
|
u32 l, sz, mask;
|
|
|
|
|
|
|
|
mask = type ? ~PCI_ROM_ADDRESS_ENABLE : ~0;
|
|
|
|
|
|
|
|
res->name = pci_name(dev);
|
|
|
|
|
|
|
|
pci_read_config_dword(dev, pos, &l);
|
|
|
|
pci_write_config_dword(dev, pos, mask);
|
|
|
|
pci_read_config_dword(dev, pos, &sz);
|
|
|
|
pci_write_config_dword(dev, pos, l);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* All bits set in sz means the device isn't working properly.
|
|
|
|
* If the BAR isn't implemented, all bits must be 0. If it's a
|
|
|
|
* memory BAR or a ROM, bit 0 must be clear; if it's an io BAR, bit
|
|
|
|
* 1 must be clear.
|
|
|
|
*/
|
|
|
|
if (!sz || sz == 0xffffffff)
|
|
|
|
goto fail;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* I don't know how l can have all bits set. Copied from old code.
|
|
|
|
* Maybe it fixes a bug on some ancient platform.
|
|
|
|
*/
|
|
|
|
if (l == 0xffffffff)
|
|
|
|
l = 0;
|
|
|
|
|
|
|
|
if (type == pci_bar_unknown) {
|
|
|
|
type = decode_bar(res, l);
|
|
|
|
res->flags |= pci_calc_resource_flags(l) | IORESOURCE_SIZEALIGN;
|
|
|
|
if (type == pci_bar_io) {
|
|
|
|
l &= PCI_BASE_ADDRESS_IO_MASK;
|
|
|
|
mask = PCI_BASE_ADDRESS_IO_MASK & 0xffff;
|
|
|
|
} else {
|
|
|
|
l &= PCI_BASE_ADDRESS_MEM_MASK;
|
|
|
|
mask = (u32)PCI_BASE_ADDRESS_MEM_MASK;
|
|
|
|
}
|
|
|
|
} else {
|
|
|
|
res->flags |= (l & IORESOURCE_ROM_ENABLE);
|
|
|
|
l &= PCI_ROM_ADDRESS_MASK;
|
|
|
|
mask = (u32)PCI_ROM_ADDRESS_MASK;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (type == pci_bar_mem64) {
|
|
|
|
u64 l64 = l;
|
|
|
|
u64 sz64 = sz;
|
|
|
|
u64 mask64 = mask | (u64)~0 << 32;
|
|
|
|
|
|
|
|
pci_read_config_dword(dev, pos + 4, &l);
|
|
|
|
pci_write_config_dword(dev, pos + 4, ~0);
|
|
|
|
pci_read_config_dword(dev, pos + 4, &sz);
|
|
|
|
pci_write_config_dword(dev, pos + 4, l);
|
|
|
|
|
|
|
|
l64 |= ((u64)l << 32);
|
|
|
|
sz64 |= ((u64)sz << 32);
|
|
|
|
|
|
|
|
sz64 = pci_size(l64, sz64, mask64);
|
|
|
|
|
|
|
|
if (!sz64)
|
|
|
|
goto fail;
|
|
|
|
|
2008-07-28 10:39:00 -07:00
|
|
|
if ((sizeof(resource_size_t) < 8) && (sz64 > 0x100000000ULL)) {
|
2008-07-28 10:38:59 -07:00
|
|
|
dev_err(&dev->dev, "can't handle 64-bit BAR\n");
|
|
|
|
goto fail;
|
2008-07-28 10:39:00 -07:00
|
|
|
} else if ((sizeof(resource_size_t) < 8) && l) {
|
2008-07-28 10:38:59 -07:00
|
|
|
/* Address above 32-bit boundary; disable the BAR */
|
|
|
|
pci_write_config_dword(dev, pos, 0);
|
|
|
|
pci_write_config_dword(dev, pos + 4, 0);
|
|
|
|
res->start = 0;
|
|
|
|
res->end = sz64;
|
|
|
|
} else {
|
|
|
|
res->start = l64;
|
|
|
|
res->end = l64 + sz64;
|
2008-10-12 03:26:12 -07:00
|
|
|
dev_printk(KERN_DEBUG, &dev->dev,
|
|
|
|
"reg %x 64bit mmio: %pR\n", pos, res);
|
2008-07-28 10:38:59 -07:00
|
|
|
}
|
|
|
|
} else {
|
|
|
|
sz = pci_size(l, sz, mask);
|
|
|
|
|
|
|
|
if (!sz)
|
|
|
|
goto fail;
|
|
|
|
|
|
|
|
res->start = l;
|
|
|
|
res->end = l + sz;
|
2008-10-12 03:26:12 -07:00
|
|
|
|
|
|
|
dev_printk(KERN_DEBUG, &dev->dev, "reg %x %s: %pR\n", pos,
|
|
|
|
(res->flags & IORESOURCE_IO) ? "io port" : "32bit mmio",
|
|
|
|
res);
|
2008-07-28 10:38:59 -07:00
|
|
|
}
|
|
|
|
|
|
|
|
out:
|
|
|
|
return (type == pci_bar_mem64) ? 1 : 0;
|
|
|
|
fail:
|
|
|
|
res->flags = 0;
|
|
|
|
goto out;
|
2006-11-29 14:53:10 -07:00
|
|
|
}
|
|
|
|
|
2005-04-16 15:20:36 -07:00
|
|
|
static void pci_read_bases(struct pci_dev *dev, unsigned int howmany, int rom)
|
|
|
|
{
|
2008-07-28 10:38:59 -07:00
|
|
|
unsigned int pos, reg;
|
2006-11-29 14:53:10 -07:00
|
|
|
|
2008-07-28 10:38:59 -07:00
|
|
|
for (pos = 0; pos < howmany; pos++) {
|
|
|
|
struct resource *res = &dev->resource[pos];
|
2005-04-16 15:20:36 -07:00
|
|
|
reg = PCI_BASE_ADDRESS_0 + (pos << 2);
|
2008-07-28 10:38:59 -07:00
|
|
|
pos += __pci_read_base(dev, pci_bar_unknown, res, reg);
|
2005-04-16 15:20:36 -07:00
|
|
|
}
|
2008-07-28 10:38:59 -07:00
|
|
|
|
2005-04-16 15:20:36 -07:00
|
|
|
if (rom) {
|
2008-07-28 10:38:59 -07:00
|
|
|
struct resource *res = &dev->resource[PCI_ROM_RESOURCE];
|
2005-04-16 15:20:36 -07:00
|
|
|
dev->rom_base_reg = rom;
|
2008-07-28 10:38:59 -07:00
|
|
|
res->flags = IORESOURCE_MEM | IORESOURCE_PREFETCH |
|
|
|
|
IORESOURCE_READONLY | IORESOURCE_CACHEABLE |
|
|
|
|
IORESOURCE_SIZEALIGN;
|
|
|
|
__pci_read_base(dev, pci_bar_mem32, res, rom);
|
2005-04-16 15:20:36 -07:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2008-02-17 02:45:28 -07:00
|
|
|
void __devinit pci_read_bridge_bases(struct pci_bus *child)
|
2005-04-16 15:20:36 -07:00
|
|
|
{
|
|
|
|
struct pci_dev *dev = child->self;
|
|
|
|
u8 io_base_lo, io_limit_lo;
|
|
|
|
u16 mem_base_lo, mem_limit_lo;
|
|
|
|
unsigned long base, limit;
|
|
|
|
struct resource *res;
|
|
|
|
int i;
|
|
|
|
|
|
|
|
if (!dev) /* It's a host bus, nothing to read */
|
|
|
|
return;
|
|
|
|
|
|
|
|
if (dev->transparent) {
|
2008-06-13 09:52:11 -07:00
|
|
|
dev_info(&dev->dev, "transparent bridge\n");
|
2005-06-06 17:07:02 -07:00
|
|
|
for(i = 3; i < PCI_BUS_NUM_RESOURCES; i++)
|
|
|
|
child->resource[i] = child->parent->resource[i - 3];
|
2005-04-16 15:20:36 -07:00
|
|
|
}
|
|
|
|
|
|
|
|
res = child->resource[0];
|
|
|
|
pci_read_config_byte(dev, PCI_IO_BASE, &io_base_lo);
|
|
|
|
pci_read_config_byte(dev, PCI_IO_LIMIT, &io_limit_lo);
|
|
|
|
base = (io_base_lo & PCI_IO_RANGE_MASK) << 8;
|
|
|
|
limit = (io_limit_lo & PCI_IO_RANGE_MASK) << 8;
|
|
|
|
|
|
|
|
if ((io_base_lo & PCI_IO_RANGE_TYPE_MASK) == PCI_IO_RANGE_TYPE_32) {
|
|
|
|
u16 io_base_hi, io_limit_hi;
|
|
|
|
pci_read_config_word(dev, PCI_IO_BASE_UPPER16, &io_base_hi);
|
|
|
|
pci_read_config_word(dev, PCI_IO_LIMIT_UPPER16, &io_limit_hi);
|
|
|
|
base |= (io_base_hi << 16);
|
|
|
|
limit |= (io_limit_hi << 16);
|
|
|
|
}
|
|
|
|
|
|
|
|
if (base <= limit) {
|
|
|
|
res->flags = (io_base_lo & PCI_IO_RANGE_TYPE_MASK) | IORESOURCE_IO;
|
2005-12-05 05:06:43 -07:00
|
|
|
if (!res->start)
|
|
|
|
res->start = base;
|
|
|
|
if (!res->end)
|
|
|
|
res->end = limit + 0xfff;
|
2008-10-12 03:26:12 -07:00
|
|
|
dev_printk(KERN_DEBUG, &dev->dev, "bridge io port: %pR\n", res);
|
2005-04-16 15:20:36 -07:00
|
|
|
}
|
|
|
|
|
|
|
|
res = child->resource[1];
|
|
|
|
pci_read_config_word(dev, PCI_MEMORY_BASE, &mem_base_lo);
|
|
|
|
pci_read_config_word(dev, PCI_MEMORY_LIMIT, &mem_limit_lo);
|
|
|
|
base = (mem_base_lo & PCI_MEMORY_RANGE_MASK) << 16;
|
|
|
|
limit = (mem_limit_lo & PCI_MEMORY_RANGE_MASK) << 16;
|
|
|
|
if (base <= limit) {
|
|
|
|
res->flags = (mem_base_lo & PCI_MEMORY_RANGE_TYPE_MASK) | IORESOURCE_MEM;
|
|
|
|
res->start = base;
|
|
|
|
res->end = limit + 0xfffff;
|
2008-10-12 03:26:12 -07:00
|
|
|
dev_printk(KERN_DEBUG, &dev->dev, "bridge 32bit mmio: %pR\n",
|
|
|
|
res);
|
2005-04-16 15:20:36 -07:00
|
|
|
}
|
|
|
|
|
|
|
|
res = child->resource[2];
|
|
|
|
pci_read_config_word(dev, PCI_PREF_MEMORY_BASE, &mem_base_lo);
|
|
|
|
pci_read_config_word(dev, PCI_PREF_MEMORY_LIMIT, &mem_limit_lo);
|
|
|
|
base = (mem_base_lo & PCI_PREF_RANGE_MASK) << 16;
|
|
|
|
limit = (mem_limit_lo & PCI_PREF_RANGE_MASK) << 16;
|
|
|
|
|
|
|
|
if ((mem_base_lo & PCI_PREF_RANGE_TYPE_MASK) == PCI_PREF_RANGE_TYPE_64) {
|
|
|
|
u32 mem_base_hi, mem_limit_hi;
|
|
|
|
pci_read_config_dword(dev, PCI_PREF_BASE_UPPER32, &mem_base_hi);
|
|
|
|
pci_read_config_dword(dev, PCI_PREF_LIMIT_UPPER32, &mem_limit_hi);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Some bridges set the base > limit by default, and some
|
|
|
|
* (broken) BIOSes do not initialize them. If we find
|
|
|
|
* this, just assume they are not being used.
|
|
|
|
*/
|
|
|
|
if (mem_base_hi <= mem_limit_hi) {
|
|
|
|
#if BITS_PER_LONG == 64
|
|
|
|
base |= ((long) mem_base_hi) << 32;
|
|
|
|
limit |= ((long) mem_limit_hi) << 32;
|
|
|
|
#else
|
|
|
|
if (mem_base_hi || mem_limit_hi) {
|
2008-06-13 09:52:11 -07:00
|
|
|
dev_err(&dev->dev, "can't handle 64-bit "
|
|
|
|
"address space for bridge\n");
|
2005-04-16 15:20:36 -07:00
|
|
|
return;
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
}
|
|
|
|
}
|
|
|
|
if (base <= limit) {
|
|
|
|
res->flags = (mem_base_lo & PCI_MEMORY_RANGE_TYPE_MASK) | IORESOURCE_MEM | IORESOURCE_PREFETCH;
|
|
|
|
res->start = base;
|
|
|
|
res->end = limit + 0xfffff;
|
2008-10-12 03:26:12 -07:00
|
|
|
dev_printk(KERN_DEBUG, &dev->dev, "bridge %sbit mmio pref: %pR\n",
|
|
|
|
(res->flags & PCI_PREF_RANGE_TYPE_64) ? "64" : "32",
|
|
|
|
res);
|
2005-04-16 15:20:36 -07:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2007-03-26 22:53:30 -07:00
|
|
|
static struct pci_bus * pci_alloc_bus(void)
|
2005-04-16 15:20:36 -07:00
|
|
|
{
|
|
|
|
struct pci_bus *b;
|
|
|
|
|
2006-02-28 07:34:49 -07:00
|
|
|
b = kzalloc(sizeof(*b), GFP_KERNEL);
|
2005-04-16 15:20:36 -07:00
|
|
|
if (b) {
|
|
|
|
INIT_LIST_HEAD(&b->node);
|
|
|
|
INIT_LIST_HEAD(&b->children);
|
|
|
|
INIT_LIST_HEAD(&b->devices);
|
2008-06-10 14:28:50 -07:00
|
|
|
INIT_LIST_HEAD(&b->slots);
|
2005-04-16 15:20:36 -07:00
|
|
|
}
|
|
|
|
return b;
|
|
|
|
}
|
|
|
|
|
2008-04-18 13:53:55 -07:00
|
|
|
static struct pci_bus *pci_alloc_child_bus(struct pci_bus *parent,
|
|
|
|
struct pci_dev *bridge, int busnr)
|
2005-04-16 15:20:36 -07:00
|
|
|
{
|
|
|
|
struct pci_bus *child;
|
|
|
|
int i;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Allocate a new bus, and inherit stuff from the parent..
|
|
|
|
*/
|
|
|
|
child = pci_alloc_bus();
|
|
|
|
if (!child)
|
|
|
|
return NULL;
|
|
|
|
|
|
|
|
child->self = bridge;
|
|
|
|
child->parent = parent;
|
|
|
|
child->ops = parent->ops;
|
|
|
|
child->sysdata = parent->sysdata;
|
2006-02-14 09:52:22 -07:00
|
|
|
child->bus_flags = parent->bus_flags;
|
2005-04-16 15:20:36 -07:00
|
|
|
child->bridge = get_device(&bridge->dev);
|
|
|
|
|
2007-05-22 19:47:54 -07:00
|
|
|
/* initialize some portions of the bus device, but don't register it
|
|
|
|
* now as the parent is not properly set up yet. This device will get
|
|
|
|
* registered later in pci_bus_add_devices()
|
|
|
|
*/
|
|
|
|
child->dev.class = &pcibus_class;
|
2008-10-29 18:17:49 -07:00
|
|
|
dev_set_name(&child->dev, "%04x:%02x", pci_domain_nr(child), busnr);
|
2005-04-16 15:20:36 -07:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Set up the primary, secondary and subordinate
|
|
|
|
* bus numbers.
|
|
|
|
*/
|
|
|
|
child->number = child->secondary = busnr;
|
|
|
|
child->primary = parent->secondary;
|
|
|
|
child->subordinate = 0xff;
|
|
|
|
|
|
|
|
/* Set up default resource pointers and names.. */
|
|
|
|
for (i = 0; i < 4; i++) {
|
|
|
|
child->resource[i] = &bridge->resource[PCI_BRIDGE_RESOURCES+i];
|
|
|
|
child->resource[i]->name = child->name;
|
|
|
|
}
|
|
|
|
bridge->subordinate = child;
|
|
|
|
|
|
|
|
return child;
|
|
|
|
}
|
|
|
|
|
2008-02-02 14:33:43 -07:00
|
|
|
struct pci_bus *__ref pci_add_new_bus(struct pci_bus *parent, struct pci_dev *dev, int busnr)
|
2005-04-16 15:20:36 -07:00
|
|
|
{
|
|
|
|
struct pci_bus *child;
|
|
|
|
|
|
|
|
child = pci_alloc_child_bus(parent, dev, busnr);
|
2005-04-28 00:25:48 -07:00
|
|
|
if (child) {
|
2006-06-01 21:35:43 -07:00
|
|
|
down_write(&pci_bus_sem);
|
2005-04-16 15:20:36 -07:00
|
|
|
list_add_tail(&child->node, &parent->children);
|
2006-06-01 21:35:43 -07:00
|
|
|
up_write(&pci_bus_sem);
|
2005-04-28 00:25:48 -07:00
|
|
|
}
|
2005-04-16 15:20:36 -07:00
|
|
|
return child;
|
|
|
|
}
|
|
|
|
|
2007-03-26 22:53:30 -07:00
|
|
|
static void pci_fixup_parent_subordinate_busnr(struct pci_bus *child, int max)
|
2005-06-02 15:41:48 -07:00
|
|
|
{
|
|
|
|
struct pci_bus *parent = child->parent;
|
2005-09-22 21:06:31 -07:00
|
|
|
|
|
|
|
/* Attempts to fix that up are really dangerous unless
|
|
|
|
we're going to re-assign all bus numbers. */
|
|
|
|
if (!pcibios_assign_all_busses())
|
|
|
|
return;
|
|
|
|
|
2005-06-02 15:41:48 -07:00
|
|
|
while (parent->parent && parent->subordinate < max) {
|
|
|
|
parent->subordinate = max;
|
|
|
|
pci_write_config_byte(parent->self, PCI_SUBORDINATE_BUS, max);
|
|
|
|
parent = parent->parent;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2005-04-16 15:20:36 -07:00
|
|
|
/*
|
|
|
|
* If it's a bridge, configure it and scan the bus behind it.
|
|
|
|
* For CardBus bridges, we don't scan behind as the devices will
|
|
|
|
* be handled by the bridge driver itself.
|
|
|
|
*
|
|
|
|
* We need to process bridges in two passes -- first we scan those
|
|
|
|
* already configured by the BIOS and after we are done with all of
|
|
|
|
* them, we proceed to assigning numbers to the remaining buses in
|
|
|
|
* order to avoid overlaps between old and new bus numbers.
|
|
|
|
*/
|
2008-02-17 02:45:28 -07:00
|
|
|
int __devinit pci_scan_bridge(struct pci_bus *bus, struct pci_dev *dev, int max, int pass)
|
2005-04-16 15:20:36 -07:00
|
|
|
{
|
|
|
|
struct pci_bus *child;
|
|
|
|
int is_cardbus = (dev->hdr_type == PCI_HEADER_TYPE_CARDBUS);
|
2005-12-08 08:53:12 -07:00
|
|
|
u32 buses, i, j = 0;
|
2005-04-16 15:20:36 -07:00
|
|
|
u16 bctl;
|
2008-10-20 16:06:29 -07:00
|
|
|
int broken = 0;
|
2005-04-16 15:20:36 -07:00
|
|
|
|
|
|
|
pci_read_config_dword(dev, PCI_PRIMARY_BUS, &buses);
|
|
|
|
|
2008-06-13 09:52:11 -07:00
|
|
|
dev_dbg(&dev->dev, "scanning behind bridge, config %06x, pass %d\n",
|
|
|
|
buses & 0xffffff, pass);
|
2005-04-16 15:20:36 -07:00
|
|
|
|
2008-10-20 16:06:29 -07:00
|
|
|
/* Check if setup is sensible at all */
|
|
|
|
if (!pass &&
|
|
|
|
((buses & 0xff) != bus->number || ((buses >> 8) & 0xff) <= bus->number)) {
|
|
|
|
dev_dbg(&dev->dev, "bus configuration invalid, reconfiguring\n");
|
|
|
|
broken = 1;
|
|
|
|
}
|
|
|
|
|
2005-04-16 15:20:36 -07:00
|
|
|
/* Disable MasterAbortMode during probing to avoid reporting
|
|
|
|
of bus errors (in some architectures) */
|
|
|
|
pci_read_config_word(dev, PCI_BRIDGE_CONTROL, &bctl);
|
|
|
|
pci_write_config_word(dev, PCI_BRIDGE_CONTROL,
|
|
|
|
bctl & ~PCI_BRIDGE_CTL_MASTER_ABORT);
|
|
|
|
|
2008-10-20 16:06:29 -07:00
|
|
|
if ((buses & 0xffff00) && !pcibios_assign_all_busses() && !is_cardbus && !broken) {
|
2005-04-16 15:20:36 -07:00
|
|
|
unsigned int cmax, busnr;
|
|
|
|
/*
|
|
|
|
* Bus already configured by firmware, process it in the first
|
|
|
|
* pass and just note the configuration.
|
|
|
|
*/
|
|
|
|
if (pass)
|
[PATCH] PCI: Avoid leaving MASTER_ABORT disabled permanently when returning from pci_scan_bridge.
> On Mon, Feb 13, 2006 at 05:13:21PM -0800, David S. Miller wrote:
> >
> > In drivers/pci/probe.c:pci_scan_bridge(), if this is not the first
> > pass (pass != 0) we don't restore the PCI_BRIDGE_CONTROL_REGISTER and
> > thus leave PCI_BRIDGE_CTL_MASTER_ABORT off:
> >
> > int __devinit pci_scan_bridge(struct pci_bus *bus, struct pci_dev * dev, int max, int pass)
> > {
> > ...
> > /* Disable MasterAbortMode during probing to avoid reporting
> > of bus errors (in some architectures) */
> > pci_read_config_word(dev, PCI_BRIDGE_CONTROL, &bctl);
> > pci_write_config_word(dev, PCI_BRIDGE_CONTROL,
> > bctl & ~PCI_BRIDGE_CTL_MASTER_ABORT);
> > ...
> > if ((buses & 0xffff00) && !pcibios_assign_all_busses() && !is_cardbus) {
> > unsigned int cmax, busnr;
> > /*
> > * Bus already configured by firmware, process it in the first
> > * pass and just note the configuration.
> > */
> > if (pass)
> > return max;
> > ...
> > }
> >
> > pci_write_config_word(dev, PCI_BRIDGE_CONTROL, bctl);
> > ...
> >
> > This doesn't seem intentional.
Agreed, looks like an accident. The patch [1] originally came from Kip
Walker (Broadcom back then) between 2.6.0-test3 and 2.6.0-test4. As I
recall it was supposed to fix an issue with with PCI aborts being
signalled by the PCI bridge of the Broadcom BCM1250 family of SOCs when
probing behind pci_scan_bridge. It is undeseriable to disable
PCI_BRIDGE_CTL_MASTER_ABORT in pci_{read,write)_config_* and the
behaviour wasn't considered a bug in need of a workaround, so this was
put in probe.c.
I don't have an affected system at hand, so can't really test but I
propose something like the below patch.
[1] http://www.linux-mips.org/git?p=linux.git;a=commit;h=599457e0cb702a31a3247ea6a5d9c6c99c4cf195
[PCI] Avoid leaving MASTER_ABORT disabled permanently when returning from pci_scan_bridge.
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
2006-02-14 09:23:57 -07:00
|
|
|
goto out;
|
2005-04-16 15:20:36 -07:00
|
|
|
busnr = (buses >> 8) & 0xFF;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* If we already got to this bus through a different bridge,
|
|
|
|
* ignore it. This can happen with the i450NX chipset.
|
|
|
|
*/
|
|
|
|
if (pci_find_bus(pci_domain_nr(bus), busnr)) {
|
2008-06-13 09:52:11 -07:00
|
|
|
dev_info(&dev->dev, "bus %04x:%02x already known\n",
|
|
|
|
pci_domain_nr(bus), busnr);
|
[PATCH] PCI: Avoid leaving MASTER_ABORT disabled permanently when returning from pci_scan_bridge.
> On Mon, Feb 13, 2006 at 05:13:21PM -0800, David S. Miller wrote:
> >
> > In drivers/pci/probe.c:pci_scan_bridge(), if this is not the first
> > pass (pass != 0) we don't restore the PCI_BRIDGE_CONTROL_REGISTER and
> > thus leave PCI_BRIDGE_CTL_MASTER_ABORT off:
> >
> > int __devinit pci_scan_bridge(struct pci_bus *bus, struct pci_dev * dev, int max, int pass)
> > {
> > ...
> > /* Disable MasterAbortMode during probing to avoid reporting
> > of bus errors (in some architectures) */
> > pci_read_config_word(dev, PCI_BRIDGE_CONTROL, &bctl);
> > pci_write_config_word(dev, PCI_BRIDGE_CONTROL,
> > bctl & ~PCI_BRIDGE_CTL_MASTER_ABORT);
> > ...
> > if ((buses & 0xffff00) && !pcibios_assign_all_busses() && !is_cardbus) {
> > unsigned int cmax, busnr;
> > /*
> > * Bus already configured by firmware, process it in the first
> > * pass and just note the configuration.
> > */
> > if (pass)
> > return max;
> > ...
> > }
> >
> > pci_write_config_word(dev, PCI_BRIDGE_CONTROL, bctl);
> > ...
> >
> > This doesn't seem intentional.
Agreed, looks like an accident. The patch [1] originally came from Kip
Walker (Broadcom back then) between 2.6.0-test3 and 2.6.0-test4. As I
recall it was supposed to fix an issue with with PCI aborts being
signalled by the PCI bridge of the Broadcom BCM1250 family of SOCs when
probing behind pci_scan_bridge. It is undeseriable to disable
PCI_BRIDGE_CTL_MASTER_ABORT in pci_{read,write)_config_* and the
behaviour wasn't considered a bug in need of a workaround, so this was
put in probe.c.
I don't have an affected system at hand, so can't really test but I
propose something like the below patch.
[1] http://www.linux-mips.org/git?p=linux.git;a=commit;h=599457e0cb702a31a3247ea6a5d9c6c99c4cf195
[PCI] Avoid leaving MASTER_ABORT disabled permanently when returning from pci_scan_bridge.
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
2006-02-14 09:23:57 -07:00
|
|
|
goto out;
|
2005-04-16 15:20:36 -07:00
|
|
|
}
|
|
|
|
|
2005-04-28 00:25:49 -07:00
|
|
|
child = pci_add_new_bus(bus, dev, busnr);
|
2005-04-16 15:20:36 -07:00
|
|
|
if (!child)
|
[PATCH] PCI: Avoid leaving MASTER_ABORT disabled permanently when returning from pci_scan_bridge.
> On Mon, Feb 13, 2006 at 05:13:21PM -0800, David S. Miller wrote:
> >
> > In drivers/pci/probe.c:pci_scan_bridge(), if this is not the first
> > pass (pass != 0) we don't restore the PCI_BRIDGE_CONTROL_REGISTER and
> > thus leave PCI_BRIDGE_CTL_MASTER_ABORT off:
> >
> > int __devinit pci_scan_bridge(struct pci_bus *bus, struct pci_dev * dev, int max, int pass)
> > {
> > ...
> > /* Disable MasterAbortMode during probing to avoid reporting
> > of bus errors (in some architectures) */
> > pci_read_config_word(dev, PCI_BRIDGE_CONTROL, &bctl);
> > pci_write_config_word(dev, PCI_BRIDGE_CONTROL,
> > bctl & ~PCI_BRIDGE_CTL_MASTER_ABORT);
> > ...
> > if ((buses & 0xffff00) && !pcibios_assign_all_busses() && !is_cardbus) {
> > unsigned int cmax, busnr;
> > /*
> > * Bus already configured by firmware, process it in the first
> > * pass and just note the configuration.
> > */
> > if (pass)
> > return max;
> > ...
> > }
> >
> > pci_write_config_word(dev, PCI_BRIDGE_CONTROL, bctl);
> > ...
> >
> > This doesn't seem intentional.
Agreed, looks like an accident. The patch [1] originally came from Kip
Walker (Broadcom back then) between 2.6.0-test3 and 2.6.0-test4. As I
recall it was supposed to fix an issue with with PCI aborts being
signalled by the PCI bridge of the Broadcom BCM1250 family of SOCs when
probing behind pci_scan_bridge. It is undeseriable to disable
PCI_BRIDGE_CTL_MASTER_ABORT in pci_{read,write)_config_* and the
behaviour wasn't considered a bug in need of a workaround, so this was
put in probe.c.
I don't have an affected system at hand, so can't really test but I
propose something like the below patch.
[1] http://www.linux-mips.org/git?p=linux.git;a=commit;h=599457e0cb702a31a3247ea6a5d9c6c99c4cf195
[PCI] Avoid leaving MASTER_ABORT disabled permanently when returning from pci_scan_bridge.
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
2006-02-14 09:23:57 -07:00
|
|
|
goto out;
|
2005-04-16 15:20:36 -07:00
|
|
|
child->primary = buses & 0xFF;
|
|
|
|
child->subordinate = (buses >> 16) & 0xFF;
|
2007-10-08 16:24:16 -07:00
|
|
|
child->bridge_ctl = bctl;
|
2005-04-16 15:20:36 -07:00
|
|
|
|
|
|
|
cmax = pci_scan_child_bus(child);
|
|
|
|
if (cmax > max)
|
|
|
|
max = cmax;
|
|
|
|
if (child->subordinate > max)
|
|
|
|
max = child->subordinate;
|
|
|
|
} else {
|
|
|
|
/*
|
|
|
|
* We need to assign a number to this bus which we always
|
|
|
|
* do in the second pass.
|
|
|
|
*/
|
2005-09-22 21:06:31 -07:00
|
|
|
if (!pass) {
|
2008-10-20 16:06:29 -07:00
|
|
|
if (pcibios_assign_all_busses() || broken)
|
2005-09-22 21:06:31 -07:00
|
|
|
/* Temporarily disable forwarding of the
|
|
|
|
configuration cycles on all bridges in
|
|
|
|
this bus segment to avoid possible
|
|
|
|
conflicts in the second pass between two
|
|
|
|
bridges programmed with overlapping
|
|
|
|
bus ranges. */
|
|
|
|
pci_write_config_dword(dev, PCI_PRIMARY_BUS,
|
|
|
|
buses & ~0xffffff);
|
[PATCH] PCI: Avoid leaving MASTER_ABORT disabled permanently when returning from pci_scan_bridge.
> On Mon, Feb 13, 2006 at 05:13:21PM -0800, David S. Miller wrote:
> >
> > In drivers/pci/probe.c:pci_scan_bridge(), if this is not the first
> > pass (pass != 0) we don't restore the PCI_BRIDGE_CONTROL_REGISTER and
> > thus leave PCI_BRIDGE_CTL_MASTER_ABORT off:
> >
> > int __devinit pci_scan_bridge(struct pci_bus *bus, struct pci_dev * dev, int max, int pass)
> > {
> > ...
> > /* Disable MasterAbortMode during probing to avoid reporting
> > of bus errors (in some architectures) */
> > pci_read_config_word(dev, PCI_BRIDGE_CONTROL, &bctl);
> > pci_write_config_word(dev, PCI_BRIDGE_CONTROL,
> > bctl & ~PCI_BRIDGE_CTL_MASTER_ABORT);
> > ...
> > if ((buses & 0xffff00) && !pcibios_assign_all_busses() && !is_cardbus) {
> > unsigned int cmax, busnr;
> > /*
> > * Bus already configured by firmware, process it in the first
> > * pass and just note the configuration.
> > */
> > if (pass)
> > return max;
> > ...
> > }
> >
> > pci_write_config_word(dev, PCI_BRIDGE_CONTROL, bctl);
> > ...
> >
> > This doesn't seem intentional.
Agreed, looks like an accident. The patch [1] originally came from Kip
Walker (Broadcom back then) between 2.6.0-test3 and 2.6.0-test4. As I
recall it was supposed to fix an issue with with PCI aborts being
signalled by the PCI bridge of the Broadcom BCM1250 family of SOCs when
probing behind pci_scan_bridge. It is undeseriable to disable
PCI_BRIDGE_CTL_MASTER_ABORT in pci_{read,write)_config_* and the
behaviour wasn't considered a bug in need of a workaround, so this was
put in probe.c.
I don't have an affected system at hand, so can't really test but I
propose something like the below patch.
[1] http://www.linux-mips.org/git?p=linux.git;a=commit;h=599457e0cb702a31a3247ea6a5d9c6c99c4cf195
[PCI] Avoid leaving MASTER_ABORT disabled permanently when returning from pci_scan_bridge.
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
2006-02-14 09:23:57 -07:00
|
|
|
goto out;
|
2005-09-22 21:06:31 -07:00
|
|
|
}
|
2005-04-16 15:20:36 -07:00
|
|
|
|
|
|
|
/* Clear errors */
|
|
|
|
pci_write_config_word(dev, PCI_STATUS, 0xffff);
|
|
|
|
|
2005-04-28 00:25:47 -07:00
|
|
|
/* Prevent assigning a bus number that already exists.
|
|
|
|
* This can happen when a bridge is hot-plugged */
|
|
|
|
if (pci_find_bus(pci_domain_nr(bus), max+1))
|
[PATCH] PCI: Avoid leaving MASTER_ABORT disabled permanently when returning from pci_scan_bridge.
> On Mon, Feb 13, 2006 at 05:13:21PM -0800, David S. Miller wrote:
> >
> > In drivers/pci/probe.c:pci_scan_bridge(), if this is not the first
> > pass (pass != 0) we don't restore the PCI_BRIDGE_CONTROL_REGISTER and
> > thus leave PCI_BRIDGE_CTL_MASTER_ABORT off:
> >
> > int __devinit pci_scan_bridge(struct pci_bus *bus, struct pci_dev * dev, int max, int pass)
> > {
> > ...
> > /* Disable MasterAbortMode during probing to avoid reporting
> > of bus errors (in some architectures) */
> > pci_read_config_word(dev, PCI_BRIDGE_CONTROL, &bctl);
> > pci_write_config_word(dev, PCI_BRIDGE_CONTROL,
> > bctl & ~PCI_BRIDGE_CTL_MASTER_ABORT);
> > ...
> > if ((buses & 0xffff00) && !pcibios_assign_all_busses() && !is_cardbus) {
> > unsigned int cmax, busnr;
> > /*
> > * Bus already configured by firmware, process it in the first
> > * pass and just note the configuration.
> > */
> > if (pass)
> > return max;
> > ...
> > }
> >
> > pci_write_config_word(dev, PCI_BRIDGE_CONTROL, bctl);
> > ...
> >
> > This doesn't seem intentional.
Agreed, looks like an accident. The patch [1] originally came from Kip
Walker (Broadcom back then) between 2.6.0-test3 and 2.6.0-test4. As I
recall it was supposed to fix an issue with with PCI aborts being
signalled by the PCI bridge of the Broadcom BCM1250 family of SOCs when
probing behind pci_scan_bridge. It is undeseriable to disable
PCI_BRIDGE_CTL_MASTER_ABORT in pci_{read,write)_config_* and the
behaviour wasn't considered a bug in need of a workaround, so this was
put in probe.c.
I don't have an affected system at hand, so can't really test but I
propose something like the below patch.
[1] http://www.linux-mips.org/git?p=linux.git;a=commit;h=599457e0cb702a31a3247ea6a5d9c6c99c4cf195
[PCI] Avoid leaving MASTER_ABORT disabled permanently when returning from pci_scan_bridge.
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
2006-02-14 09:23:57 -07:00
|
|
|
goto out;
|
2005-04-28 00:25:49 -07:00
|
|
|
child = pci_add_new_bus(bus, dev, ++max);
|
2005-04-16 15:20:36 -07:00
|
|
|
buses = (buses & 0xff000000)
|
|
|
|
| ((unsigned int)(child->primary) << 0)
|
|
|
|
| ((unsigned int)(child->secondary) << 8)
|
|
|
|
| ((unsigned int)(child->subordinate) << 16);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* yenta.c forces a secondary latency timer of 176.
|
|
|
|
* Copy that behaviour here.
|
|
|
|
*/
|
|
|
|
if (is_cardbus) {
|
|
|
|
buses &= ~0xff000000;
|
|
|
|
buses |= CARDBUS_LATENCY_TIMER << 24;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* We need to blast all three values with a single write.
|
|
|
|
*/
|
|
|
|
pci_write_config_dword(dev, PCI_PRIMARY_BUS, buses);
|
|
|
|
|
|
|
|
if (!is_cardbus) {
|
2007-10-08 16:24:16 -07:00
|
|
|
child->bridge_ctl = bctl;
|
2005-06-02 15:41:48 -07:00
|
|
|
/*
|
|
|
|
* Adjust subordinate busnr in parent buses.
|
|
|
|
* We do this before scanning for children because
|
|
|
|
* some devices may not be detected if the bios
|
|
|
|
* was lazy.
|
|
|
|
*/
|
|
|
|
pci_fixup_parent_subordinate_busnr(child, max);
|
2005-04-16 15:20:36 -07:00
|
|
|
/* Now we can scan all subordinate buses... */
|
|
|
|
max = pci_scan_child_bus(child);
|
2006-01-17 17:57:01 -07:00
|
|
|
/*
|
|
|
|
* now fix it up again since we have found
|
|
|
|
* the real value of max.
|
|
|
|
*/
|
|
|
|
pci_fixup_parent_subordinate_busnr(child, max);
|
2005-04-16 15:20:36 -07:00
|
|
|
} else {
|
|
|
|
/*
|
|
|
|
* For CardBus bridges, we leave 4 bus numbers
|
|
|
|
* as cards with a PCI-to-PCI bridge can be
|
|
|
|
* inserted later.
|
|
|
|
*/
|
2005-12-08 08:53:12 -07:00
|
|
|
for (i=0; i<CARDBUS_RESERVE_BUSNR; i++) {
|
|
|
|
struct pci_bus *parent = bus;
|
2005-04-28 00:25:47 -07:00
|
|
|
if (pci_find_bus(pci_domain_nr(bus),
|
|
|
|
max+i+1))
|
|
|
|
break;
|
2005-12-08 08:53:12 -07:00
|
|
|
while (parent->parent) {
|
|
|
|
if ((!pcibios_assign_all_busses()) &&
|
|
|
|
(parent->subordinate > max) &&
|
|
|
|
(parent->subordinate <= max+i)) {
|
|
|
|
j = 1;
|
|
|
|
}
|
|
|
|
parent = parent->parent;
|
|
|
|
}
|
|
|
|
if (j) {
|
|
|
|
/*
|
|
|
|
* Often, there are two cardbus bridges
|
|
|
|
* -- try to leave one valid bus number
|
|
|
|
* for each one.
|
|
|
|
*/
|
|
|
|
i /= 2;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
2005-04-28 00:25:47 -07:00
|
|
|
max += i;
|
2005-06-02 15:41:48 -07:00
|
|
|
pci_fixup_parent_subordinate_busnr(child, max);
|
2005-04-16 15:20:36 -07:00
|
|
|
}
|
|
|
|
/*
|
|
|
|
* Set the subordinate bus number to its real value.
|
|
|
|
*/
|
|
|
|
child->subordinate = max;
|
|
|
|
pci_write_config_byte(dev, PCI_SUBORDINATE_BUS, max);
|
|
|
|
}
|
|
|
|
|
2008-02-08 15:00:52 -07:00
|
|
|
sprintf(child->name,
|
|
|
|
(is_cardbus ? "PCI CardBus %04x:%02x" : "PCI Bus %04x:%02x"),
|
|
|
|
pci_domain_nr(bus), child->number);
|
2005-04-16 15:20:36 -07:00
|
|
|
|
PCI: lets kill the 'PCI hidden behind bridge' message
Adrian Bunk wrote:
> Alois Nešpor wrote
>> PCI: Bus #0b (-#0e) is hidden behind transparent bridge #0a (-#0b) (try 'pci=assign-busses')
>> Please report the result to linux-kernel to fix this permanently"
>>
>> dmesg:
>> "Yenta: Raising subordinate bus# of parent bus (#0a) from #0b to #0e"
>> without pci=assign-busses and nothing with pci=assign-busses.
>
> Bernhard?
Ok, lets kill the message. As Alois Nešpor also saw, that's fixed up by Yenta,
so PCI does not have to warn about it. PCI could still warn about it if
is_cardbus is 0 in that instance of pci_scan_bridge(), but so far I have
not seen a report where this would have been the case so I think we can
spare the kernel of that check (removes ~300 lines of asm) unless debugging
is done.
History: The whole check was added in the days before we had the fixup
for this in Yenta and pci=assign-busses was the only way to get CardBus
cards detected on many (not all) of the machines which give this warning.
In theory, there could be cases when this warning would be triggered and
it's not cardbus, then the warning should still apply, but I think this
should only be the case when working on a completely broken PCI setup,
but one may have already enabled the debug code in drivers/pci and the
patched check would then trigger.
I do not sign this off yet because it's completely untested so far, but
everyone is free to test it (with the #ifdef DEBUG replaced by #if 1 and
pr_debug( changed to printk(.
We may also dump the whole check (remove everything within the #ifdef from
the source) if that's perferred.
On Alois Nešpor's machine this would then (only when debugging) this message:
"PCI: Bus #0b (-#0e) is partially hidden behind transparent bridge #0a (-#0b)"
"partially" should be in the message on his machine because #0b of #0b-#0e
is reachable behind #0a-#0b, but not #0c-#0e.
But that differentiation is now moot anyway because the fixup in Yenta takes
care of it as far as I could see so far, which means that unless somebody
is debugging a totally broken PCI setup, this message is not needed anymore,
not even for debugging PCI.
Ok, here the patch with the following changes:
* Refined to say that the bus is only partially hidden when the parent
bus numbers are not totally way off (outside of) the child bus range
* remove the reference to pci=assign-busses and the plea to report it
We could add a pure source code-only comment to keep a reference to
pci=assign-busses the in case when this is triggered by someone who
is debugging the cause of this message and looking the way to solve it.
From: Bernhard Kaindl <bk@suse.de>
Cc: stable <stable@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
2007-07-30 11:35:13 -07:00
|
|
|
/* Has only triggered on CardBus, fixup is in yenta_socket */
|
2005-12-08 08:53:12 -07:00
|
|
|
while (bus->parent) {
|
|
|
|
if ((child->subordinate > bus->subordinate) ||
|
|
|
|
(child->number > bus->subordinate) ||
|
|
|
|
(child->number < bus->number) ||
|
|
|
|
(child->subordinate < bus->number)) {
|
2007-11-19 18:48:29 -07:00
|
|
|
pr_debug("PCI: Bus #%02x (-#%02x) is %s "
|
PCI: lets kill the 'PCI hidden behind bridge' message
Adrian Bunk wrote:
> Alois Nešpor wrote
>> PCI: Bus #0b (-#0e) is hidden behind transparent bridge #0a (-#0b) (try 'pci=assign-busses')
>> Please report the result to linux-kernel to fix this permanently"
>>
>> dmesg:
>> "Yenta: Raising subordinate bus# of parent bus (#0a) from #0b to #0e"
>> without pci=assign-busses and nothing with pci=assign-busses.
>
> Bernhard?
Ok, lets kill the message. As Alois Nešpor also saw, that's fixed up by Yenta,
so PCI does not have to warn about it. PCI could still warn about it if
is_cardbus is 0 in that instance of pci_scan_bridge(), but so far I have
not seen a report where this would have been the case so I think we can
spare the kernel of that check (removes ~300 lines of asm) unless debugging
is done.
History: The whole check was added in the days before we had the fixup
for this in Yenta and pci=assign-busses was the only way to get CardBus
cards detected on many (not all) of the machines which give this warning.
In theory, there could be cases when this warning would be triggered and
it's not cardbus, then the warning should still apply, but I think this
should only be the case when working on a completely broken PCI setup,
but one may have already enabled the debug code in drivers/pci and the
patched check would then trigger.
I do not sign this off yet because it's completely untested so far, but
everyone is free to test it (with the #ifdef DEBUG replaced by #if 1 and
pr_debug( changed to printk(.
We may also dump the whole check (remove everything within the #ifdef from
the source) if that's perferred.
On Alois Nešpor's machine this would then (only when debugging) this message:
"PCI: Bus #0b (-#0e) is partially hidden behind transparent bridge #0a (-#0b)"
"partially" should be in the message on his machine because #0b of #0b-#0e
is reachable behind #0a-#0b, but not #0c-#0e.
But that differentiation is now moot anyway because the fixup in Yenta takes
care of it as far as I could see so far, which means that unless somebody
is debugging a totally broken PCI setup, this message is not needed anymore,
not even for debugging PCI.
Ok, here the patch with the following changes:
* Refined to say that the bus is only partially hidden when the parent
bus numbers are not totally way off (outside of) the child bus range
* remove the reference to pci=assign-busses and the plea to report it
We could add a pure source code-only comment to keep a reference to
pci=assign-busses the in case when this is triggered by someone who
is debugging the cause of this message and looking the way to solve it.
From: Bernhard Kaindl <bk@suse.de>
Cc: stable <stable@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
2007-07-30 11:35:13 -07:00
|
|
|
"hidden behind%s bridge #%02x (-#%02x)\n",
|
|
|
|
child->number, child->subordinate,
|
|
|
|
(bus->number > child->subordinate &&
|
|
|
|
bus->subordinate < child->number) ?
|
2007-11-19 18:48:29 -07:00
|
|
|
"wholly" : "partially",
|
|
|
|
bus->self->transparent ? " transparent" : "",
|
PCI: lets kill the 'PCI hidden behind bridge' message
Adrian Bunk wrote:
> Alois Nešpor wrote
>> PCI: Bus #0b (-#0e) is hidden behind transparent bridge #0a (-#0b) (try 'pci=assign-busses')
>> Please report the result to linux-kernel to fix this permanently"
>>
>> dmesg:
>> "Yenta: Raising subordinate bus# of parent bus (#0a) from #0b to #0e"
>> without pci=assign-busses and nothing with pci=assign-busses.
>
> Bernhard?
Ok, lets kill the message. As Alois Nešpor also saw, that's fixed up by Yenta,
so PCI does not have to warn about it. PCI could still warn about it if
is_cardbus is 0 in that instance of pci_scan_bridge(), but so far I have
not seen a report where this would have been the case so I think we can
spare the kernel of that check (removes ~300 lines of asm) unless debugging
is done.
History: The whole check was added in the days before we had the fixup
for this in Yenta and pci=assign-busses was the only way to get CardBus
cards detected on many (not all) of the machines which give this warning.
In theory, there could be cases when this warning would be triggered and
it's not cardbus, then the warning should still apply, but I think this
should only be the case when working on a completely broken PCI setup,
but one may have already enabled the debug code in drivers/pci and the
patched check would then trigger.
I do not sign this off yet because it's completely untested so far, but
everyone is free to test it (with the #ifdef DEBUG replaced by #if 1 and
pr_debug( changed to printk(.
We may also dump the whole check (remove everything within the #ifdef from
the source) if that's perferred.
On Alois Nešpor's machine this would then (only when debugging) this message:
"PCI: Bus #0b (-#0e) is partially hidden behind transparent bridge #0a (-#0b)"
"partially" should be in the message on his machine because #0b of #0b-#0e
is reachable behind #0a-#0b, but not #0c-#0e.
But that differentiation is now moot anyway because the fixup in Yenta takes
care of it as far as I could see so far, which means that unless somebody
is debugging a totally broken PCI setup, this message is not needed anymore,
not even for debugging PCI.
Ok, here the patch with the following changes:
* Refined to say that the bus is only partially hidden when the parent
bus numbers are not totally way off (outside of) the child bus range
* remove the reference to pci=assign-busses and the plea to report it
We could add a pure source code-only comment to keep a reference to
pci=assign-busses the in case when this is triggered by someone who
is debugging the cause of this message and looking the way to solve it.
From: Bernhard Kaindl <bk@suse.de>
Cc: stable <stable@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
2007-07-30 11:35:13 -07:00
|
|
|
bus->number, bus->subordinate);
|
2005-12-08 08:53:12 -07:00
|
|
|
}
|
|
|
|
bus = bus->parent;
|
|
|
|
}
|
|
|
|
|
[PATCH] PCI: Avoid leaving MASTER_ABORT disabled permanently when returning from pci_scan_bridge.
> On Mon, Feb 13, 2006 at 05:13:21PM -0800, David S. Miller wrote:
> >
> > In drivers/pci/probe.c:pci_scan_bridge(), if this is not the first
> > pass (pass != 0) we don't restore the PCI_BRIDGE_CONTROL_REGISTER and
> > thus leave PCI_BRIDGE_CTL_MASTER_ABORT off:
> >
> > int __devinit pci_scan_bridge(struct pci_bus *bus, struct pci_dev * dev, int max, int pass)
> > {
> > ...
> > /* Disable MasterAbortMode during probing to avoid reporting
> > of bus errors (in some architectures) */
> > pci_read_config_word(dev, PCI_BRIDGE_CONTROL, &bctl);
> > pci_write_config_word(dev, PCI_BRIDGE_CONTROL,
> > bctl & ~PCI_BRIDGE_CTL_MASTER_ABORT);
> > ...
> > if ((buses & 0xffff00) && !pcibios_assign_all_busses() && !is_cardbus) {
> > unsigned int cmax, busnr;
> > /*
> > * Bus already configured by firmware, process it in the first
> > * pass and just note the configuration.
> > */
> > if (pass)
> > return max;
> > ...
> > }
> >
> > pci_write_config_word(dev, PCI_BRIDGE_CONTROL, bctl);
> > ...
> >
> > This doesn't seem intentional.
Agreed, looks like an accident. The patch [1] originally came from Kip
Walker (Broadcom back then) between 2.6.0-test3 and 2.6.0-test4. As I
recall it was supposed to fix an issue with with PCI aborts being
signalled by the PCI bridge of the Broadcom BCM1250 family of SOCs when
probing behind pci_scan_bridge. It is undeseriable to disable
PCI_BRIDGE_CTL_MASTER_ABORT in pci_{read,write)_config_* and the
behaviour wasn't considered a bug in need of a workaround, so this was
put in probe.c.
I don't have an affected system at hand, so can't really test but I
propose something like the below patch.
[1] http://www.linux-mips.org/git?p=linux.git;a=commit;h=599457e0cb702a31a3247ea6a5d9c6c99c4cf195
[PCI] Avoid leaving MASTER_ABORT disabled permanently when returning from pci_scan_bridge.
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
2006-02-14 09:23:57 -07:00
|
|
|
out:
|
|
|
|
pci_write_config_word(dev, PCI_BRIDGE_CONTROL, bctl);
|
|
|
|
|
2005-04-16 15:20:36 -07:00
|
|
|
return max;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Read interrupt line and base address registers.
|
|
|
|
* The architecture-dependent code can tweak these, of course.
|
|
|
|
*/
|
|
|
|
static void pci_read_irq(struct pci_dev *dev)
|
|
|
|
{
|
|
|
|
unsigned char irq;
|
|
|
|
|
|
|
|
pci_read_config_byte(dev, PCI_INTERRUPT_PIN, &irq);
|
2005-11-02 17:24:32 -07:00
|
|
|
dev->pin = irq;
|
2005-04-16 15:20:36 -07:00
|
|
|
if (irq)
|
|
|
|
pci_read_config_byte(dev, PCI_INTERRUPT_LINE, &irq);
|
|
|
|
dev->irq = irq;
|
|
|
|
}
|
|
|
|
|
2007-04-23 14:19:36 -07:00
|
|
|
#define LEGACY_IO_RESOURCE (IORESOURCE_IO | IORESOURCE_PCI_FIXED)
|
2006-12-29 17:47:29 -07:00
|
|
|
|
2005-04-16 15:20:36 -07:00
|
|
|
/**
|
|
|
|
* pci_setup_device - fill in class and map information of a device
|
|
|
|
* @dev: the device structure to fill
|
|
|
|
*
|
|
|
|
* Initialize the device structure with information about the device's
|
|
|
|
* vendor,class,memory and IO-space addresses,IRQ lines etc.
|
|
|
|
* Called at initialisation of the PCI subsystem and by CardBus services.
|
|
|
|
* Returns 0 on success and -1 if unknown type of device (not normal, bridge
|
|
|
|
* or CardBus).
|
|
|
|
*/
|
|
|
|
static int pci_setup_device(struct pci_dev * dev)
|
|
|
|
{
|
|
|
|
u32 class;
|
|
|
|
|
2008-07-02 13:24:49 -07:00
|
|
|
dev_set_name(&dev->dev, "%04x:%02x:%02x.%d", pci_domain_nr(dev->bus),
|
|
|
|
dev->bus->number, PCI_SLOT(dev->devfn),
|
|
|
|
PCI_FUNC(dev->devfn));
|
2005-04-16 15:20:36 -07:00
|
|
|
|
|
|
|
pci_read_config_dword(dev, PCI_CLASS_REVISION, &class);
|
2007-06-08 15:46:30 -07:00
|
|
|
dev->revision = class & 0xff;
|
2005-04-16 15:20:36 -07:00
|
|
|
class >>= 8; /* upper 3 bytes */
|
|
|
|
dev->class = class;
|
|
|
|
class >>= 8;
|
|
|
|
|
2008-08-25 14:45:20 -07:00
|
|
|
dev_dbg(&dev->dev, "found [%04x:%04x] class %06x header type %02x\n",
|
2005-04-16 15:20:36 -07:00
|
|
|
dev->vendor, dev->device, class, dev->hdr_type);
|
|
|
|
|
|
|
|
/* "Unknown power state" */
|
2005-08-17 15:32:19 -07:00
|
|
|
dev->current_state = PCI_UNKNOWN;
|
2005-04-16 15:20:36 -07:00
|
|
|
|
|
|
|
/* Early fixups, before probing the BARs */
|
|
|
|
pci_fixup_device(pci_fixup_early, dev);
|
|
|
|
class = dev->class >> 8;
|
|
|
|
|
|
|
|
switch (dev->hdr_type) { /* header type */
|
|
|
|
case PCI_HEADER_TYPE_NORMAL: /* standard header */
|
|
|
|
if (class == PCI_CLASS_BRIDGE_PCI)
|
|
|
|
goto bad;
|
|
|
|
pci_read_irq(dev);
|
|
|
|
pci_read_bases(dev, 6, PCI_ROM_ADDRESS);
|
|
|
|
pci_read_config_word(dev, PCI_SUBSYSTEM_VENDOR_ID, &dev->subsystem_vendor);
|
|
|
|
pci_read_config_word(dev, PCI_SUBSYSTEM_ID, &dev->subsystem_device);
|
2006-10-03 16:41:26 -07:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Do the ugly legacy mode stuff here rather than broken chip
|
|
|
|
* quirk code. Legacy mode ATA controllers have fixed
|
|
|
|
* addresses. These are not always echoed in BAR0-3, and
|
|
|
|
* BAR0-3 in a few cases contain junk!
|
|
|
|
*/
|
|
|
|
if (class == PCI_CLASS_STORAGE_IDE) {
|
|
|
|
u8 progif;
|
|
|
|
pci_read_config_byte(dev, PCI_CLASS_PROG, &progif);
|
|
|
|
if ((progif & 1) == 0) {
|
2007-12-10 08:40:54 -07:00
|
|
|
dev->resource[0].start = 0x1F0;
|
|
|
|
dev->resource[0].end = 0x1F7;
|
|
|
|
dev->resource[0].flags = LEGACY_IO_RESOURCE;
|
|
|
|
dev->resource[1].start = 0x3F6;
|
|
|
|
dev->resource[1].end = 0x3F6;
|
|
|
|
dev->resource[1].flags = LEGACY_IO_RESOURCE;
|
2006-10-03 16:41:26 -07:00
|
|
|
}
|
|
|
|
if ((progif & 4) == 0) {
|
2007-12-10 08:40:54 -07:00
|
|
|
dev->resource[2].start = 0x170;
|
|
|
|
dev->resource[2].end = 0x177;
|
|
|
|
dev->resource[2].flags = LEGACY_IO_RESOURCE;
|
|
|
|
dev->resource[3].start = 0x376;
|
|
|
|
dev->resource[3].end = 0x376;
|
|
|
|
dev->resource[3].flags = LEGACY_IO_RESOURCE;
|
2006-10-03 16:41:26 -07:00
|
|
|
}
|
|
|
|
}
|
2005-04-16 15:20:36 -07:00
|
|
|
break;
|
|
|
|
|
|
|
|
case PCI_HEADER_TYPE_BRIDGE: /* bridge header */
|
|
|
|
if (class != PCI_CLASS_BRIDGE_PCI)
|
|
|
|
goto bad;
|
|
|
|
/* The PCI-to-PCI bridge spec requires that subtractive
|
|
|
|
decoding (i.e. transparent) bridge must have programming
|
|
|
|
interface code of 0x01. */
|
2005-11-02 17:55:49 -07:00
|
|
|
pci_read_irq(dev);
|
2005-04-16 15:20:36 -07:00
|
|
|
dev->transparent = ((dev->class & 0xff) == 1);
|
|
|
|
pci_read_bases(dev, 2, PCI_ROM_ADDRESS1);
|
|
|
|
break;
|
|
|
|
|
|
|
|
case PCI_HEADER_TYPE_CARDBUS: /* CardBus bridge header */
|
|
|
|
if (class != PCI_CLASS_BRIDGE_CARDBUS)
|
|
|
|
goto bad;
|
|
|
|
pci_read_irq(dev);
|
|
|
|
pci_read_bases(dev, 1, 0);
|
|
|
|
pci_read_config_word(dev, PCI_CB_SUBSYSTEM_VENDOR_ID, &dev->subsystem_vendor);
|
|
|
|
pci_read_config_word(dev, PCI_CB_SUBSYSTEM_ID, &dev->subsystem_device);
|
|
|
|
break;
|
|
|
|
|
|
|
|
default: /* unknown header */
|
2008-06-13 09:52:11 -07:00
|
|
|
dev_err(&dev->dev, "unknown header type %02x, "
|
|
|
|
"ignoring device\n", dev->hdr_type);
|
2005-04-16 15:20:36 -07:00
|
|
|
return -1;
|
|
|
|
|
|
|
|
bad:
|
2008-06-13 09:52:11 -07:00
|
|
|
dev_err(&dev->dev, "ignoring class %02x (doesn't match header "
|
|
|
|
"type %02x)\n", class, dev->hdr_type);
|
2005-04-16 15:20:36 -07:00
|
|
|
dev->class = PCI_CLASS_NOT_DEFINED;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* We found a fine healthy device, go go go... */
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2008-10-13 04:49:55 -07:00
|
|
|
static void pci_release_capabilities(struct pci_dev *dev)
|
|
|
|
{
|
|
|
|
pci_vpd_release(dev);
|
|
|
|
}
|
|
|
|
|
2005-04-16 15:20:36 -07:00
|
|
|
/**
|
|
|
|
* pci_release_dev - free a pci device structure when all users of it are finished.
|
|
|
|
* @dev: device that's been disconnected
|
|
|
|
*
|
|
|
|
* Will be called only by the device core when all users of this pci device are
|
|
|
|
* done.
|
|
|
|
*/
|
|
|
|
static void pci_release_dev(struct device *dev)
|
|
|
|
{
|
|
|
|
struct pci_dev *pci_dev;
|
|
|
|
|
|
|
|
pci_dev = to_pci_dev(dev);
|
2008-10-13 04:49:55 -07:00
|
|
|
pci_release_capabilities(pci_dev);
|
2005-04-16 15:20:36 -07:00
|
|
|
kfree(pci_dev);
|
|
|
|
}
|
|
|
|
|
2007-10-21 16:41:46 -07:00
|
|
|
static void set_pcie_port_type(struct pci_dev *pdev)
|
|
|
|
{
|
|
|
|
int pos;
|
|
|
|
u16 reg16;
|
|
|
|
|
|
|
|
pos = pci_find_capability(pdev, PCI_CAP_ID_EXP);
|
|
|
|
if (!pos)
|
|
|
|
return;
|
|
|
|
pdev->is_pcie = 1;
|
|
|
|
pci_read_config_word(pdev, pos + PCI_EXP_FLAGS, ®16);
|
|
|
|
pdev->pcie_type = (reg16 & PCI_EXP_FLAGS_TYPE) >> 4;
|
|
|
|
}
|
|
|
|
|
2005-04-16 15:20:36 -07:00
|
|
|
/**
|
|
|
|
* pci_cfg_space_size - get the configuration space size of the PCI device.
|
2005-10-23 11:57:38 -07:00
|
|
|
* @dev: PCI device
|
2005-04-16 15:20:36 -07:00
|
|
|
*
|
|
|
|
* Regular PCI devices have 256 bytes, but PCI-X 2 and PCI Express devices
|
|
|
|
* have 4096 bytes. Even if the device is capable, that doesn't mean we can
|
|
|
|
* access it. Maybe we don't have a way to generate extended config space
|
|
|
|
* accesses, or the device is behind a reverse Express bridge. So we try
|
|
|
|
* reading the dword at 0x100 which must either be 0 or a valid extended
|
|
|
|
* capability header.
|
|
|
|
*/
|
2008-04-28 16:27:23 -07:00
|
|
|
int pci_cfg_space_size_ext(struct pci_dev *dev)
|
2005-04-16 15:20:36 -07:00
|
|
|
{
|
|
|
|
u32 status;
|
2008-10-13 04:18:07 -07:00
|
|
|
int pos = PCI_CFG_SPACE_SIZE;
|
2005-04-16 15:20:36 -07:00
|
|
|
|
2008-10-13 04:18:07 -07:00
|
|
|
if (pci_read_config_dword(dev, pos, &status) != PCIBIOS_SUCCESSFUL)
|
2008-04-28 16:27:23 -07:00
|
|
|
goto fail;
|
|
|
|
if (status == 0xffffffff)
|
|
|
|
goto fail;
|
|
|
|
|
|
|
|
return PCI_CFG_SPACE_EXP_SIZE;
|
|
|
|
|
|
|
|
fail:
|
|
|
|
return PCI_CFG_SPACE_SIZE;
|
|
|
|
}
|
|
|
|
|
|
|
|
int pci_cfg_space_size(struct pci_dev *dev)
|
|
|
|
{
|
|
|
|
int pos;
|
|
|
|
u32 status;
|
2008-02-15 02:32:50 -07:00
|
|
|
|
2005-04-16 15:20:36 -07:00
|
|
|
pos = pci_find_capability(dev, PCI_CAP_ID_EXP);
|
|
|
|
if (!pos) {
|
|
|
|
pos = pci_find_capability(dev, PCI_CAP_ID_PCIX);
|
|
|
|
if (!pos)
|
|
|
|
goto fail;
|
|
|
|
|
|
|
|
pci_read_config_dword(dev, pos + PCI_X_STATUS, &status);
|
|
|
|
if (!(status & (PCI_X_STATUS_266MHZ | PCI_X_STATUS_533MHZ)))
|
|
|
|
goto fail;
|
|
|
|
}
|
|
|
|
|
2008-04-28 16:27:23 -07:00
|
|
|
return pci_cfg_space_size_ext(dev);
|
2005-04-16 15:20:36 -07:00
|
|
|
|
|
|
|
fail:
|
|
|
|
return PCI_CFG_SPACE_SIZE;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void pci_release_bus_bridge_dev(struct device *dev)
|
|
|
|
{
|
|
|
|
kfree(dev);
|
|
|
|
}
|
|
|
|
|
2007-04-05 00:19:08 -07:00
|
|
|
struct pci_dev *alloc_pci_dev(void)
|
|
|
|
{
|
|
|
|
struct pci_dev *dev;
|
|
|
|
|
|
|
|
dev = kzalloc(sizeof(struct pci_dev), GFP_KERNEL);
|
|
|
|
if (!dev)
|
|
|
|
return NULL;
|
|
|
|
|
|
|
|
INIT_LIST_HEAD(&dev->bus_list);
|
|
|
|
|
|
|
|
return dev;
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL(alloc_pci_dev);
|
|
|
|
|
2005-04-16 15:20:36 -07:00
|
|
|
/*
|
|
|
|
* Read the config data for a PCI device, sanity-check it
|
|
|
|
* and fill in the dev structure...
|
|
|
|
*/
|
2008-04-18 13:53:55 -07:00
|
|
|
static struct pci_dev *pci_scan_device(struct pci_bus *bus, int devfn)
|
2005-04-16 15:20:36 -07:00
|
|
|
{
|
|
|
|
struct pci_dev *dev;
|
2008-09-02 08:40:51 -07:00
|
|
|
struct pci_slot *slot;
|
2005-04-16 15:20:36 -07:00
|
|
|
u32 l;
|
|
|
|
u8 hdr_type;
|
|
|
|
int delay = 1;
|
|
|
|
|
|
|
|
if (pci_bus_read_config_dword(bus, devfn, PCI_VENDOR_ID, &l))
|
|
|
|
return NULL;
|
|
|
|
|
|
|
|
/* some broken boards return 0 or ~0 if a slot is empty: */
|
|
|
|
if (l == 0xffffffff || l == 0x00000000 ||
|
|
|
|
l == 0x0000ffff || l == 0xffff0000)
|
|
|
|
return NULL;
|
|
|
|
|
|
|
|
/* Configuration request Retry Status */
|
|
|
|
while (l == 0xffff0001) {
|
|
|
|
msleep(delay);
|
|
|
|
delay *= 2;
|
|
|
|
if (pci_bus_read_config_dword(bus, devfn, PCI_VENDOR_ID, &l))
|
|
|
|
return NULL;
|
|
|
|
/* Card hasn't responded in 60 seconds? Must be stuck. */
|
|
|
|
if (delay > 60 * 1000) {
|
2008-06-13 09:52:11 -07:00
|
|
|
printk(KERN_WARNING "pci %04x:%02x:%02x.%d: not "
|
2005-04-16 15:20:36 -07:00
|
|
|
"responding\n", pci_domain_nr(bus),
|
|
|
|
bus->number, PCI_SLOT(devfn),
|
|
|
|
PCI_FUNC(devfn));
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
if (pci_bus_read_config_byte(bus, devfn, PCI_HEADER_TYPE, &hdr_type))
|
|
|
|
return NULL;
|
|
|
|
|
2007-04-05 00:19:09 -07:00
|
|
|
dev = alloc_pci_dev();
|
2005-04-16 15:20:36 -07:00
|
|
|
if (!dev)
|
|
|
|
return NULL;
|
|
|
|
|
|
|
|
dev->bus = bus;
|
|
|
|
dev->sysdata = bus->sysdata;
|
|
|
|
dev->dev.parent = bus->bridge;
|
|
|
|
dev->dev.bus = &pci_bus_type;
|
|
|
|
dev->devfn = devfn;
|
|
|
|
dev->hdr_type = hdr_type & 0x7f;
|
|
|
|
dev->multifunction = !!(hdr_type & 0x80);
|
|
|
|
dev->vendor = l & 0xffff;
|
|
|
|
dev->device = (l >> 16) & 0xffff;
|
|
|
|
dev->cfg_size = pci_cfg_space_size(dev);
|
2006-07-10 04:44:46 -07:00
|
|
|
dev->error_state = pci_channel_io_normal;
|
2007-10-21 16:41:46 -07:00
|
|
|
set_pcie_port_type(dev);
|
2005-04-16 15:20:36 -07:00
|
|
|
|
2008-09-02 08:40:51 -07:00
|
|
|
list_for_each_entry(slot, &bus->slots, list)
|
|
|
|
if (PCI_SLOT(devfn) == slot->number)
|
|
|
|
dev->slot = slot;
|
|
|
|
|
2005-04-16 15:20:36 -07:00
|
|
|
/* Assume 32-bit PCI; let 64-bit PCI cards (which are far rarer)
|
|
|
|
set this higher, assuming the system even supports it. */
|
|
|
|
dev->dma_mask = 0xffffffff;
|
|
|
|
if (pci_setup_device(dev) < 0) {
|
|
|
|
kfree(dev);
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
return dev;
|
|
|
|
}
|
|
|
|
|
2008-10-13 04:49:55 -07:00
|
|
|
static void pci_init_capabilities(struct pci_dev *dev)
|
|
|
|
{
|
|
|
|
/* MSI/MSI-X list */
|
|
|
|
pci_msi_init_pci_dev(dev);
|
|
|
|
|
|
|
|
/* Power Management */
|
|
|
|
pci_pm_init(dev);
|
|
|
|
|
|
|
|
/* Vital Product Data */
|
|
|
|
pci_vpd_pci22_init(dev);
|
2008-10-13 23:02:53 -07:00
|
|
|
|
|
|
|
/* Alternative Routing-ID Forwarding */
|
|
|
|
pci_enable_ari(dev);
|
2008-10-13 04:49:55 -07:00
|
|
|
}
|
|
|
|
|
2007-03-26 22:53:30 -07:00
|
|
|
void pci_device_add(struct pci_dev *dev, struct pci_bus *bus)
|
2005-04-16 15:20:36 -07:00
|
|
|
{
|
2005-09-05 16:31:03 -07:00
|
|
|
device_initialize(&dev->dev);
|
|
|
|
dev->dev.release = pci_release_dev;
|
|
|
|
pci_dev_get(dev);
|
2005-04-16 15:20:36 -07:00
|
|
|
|
2005-09-05 16:31:03 -07:00
|
|
|
dev->dev.dma_mask = &dev->dma_mask;
|
2008-02-04 23:27:55 -07:00
|
|
|
dev->dev.dma_parms = &dev->dma_parms;
|
2005-09-05 16:31:03 -07:00
|
|
|
dev->dev.coherent_dma_mask = 0xffffffffull;
|
2005-04-16 15:20:36 -07:00
|
|
|
|
2008-02-04 23:27:55 -07:00
|
|
|
pci_set_dma_max_seg_size(dev, 65536);
|
2008-02-04 23:28:14 -07:00
|
|
|
pci_set_dma_seg_boundary(dev, 0xffffffff);
|
2008-02-04 23:27:55 -07:00
|
|
|
|
2005-04-16 15:20:36 -07:00
|
|
|
/* Fix up broken headers */
|
|
|
|
pci_fixup_device(pci_fixup_header, dev);
|
|
|
|
|
2008-10-13 04:49:55 -07:00
|
|
|
/* Initialize various capabilities */
|
|
|
|
pci_init_capabilities(dev);
|
2008-07-06 18:34:48 -07:00
|
|
|
|
2005-04-16 15:20:36 -07:00
|
|
|
/*
|
|
|
|
* Add the device to our list of discovered devices
|
|
|
|
* and the bus list for fixup functions, etc.
|
|
|
|
*/
|
2006-06-01 21:35:43 -07:00
|
|
|
down_write(&pci_bus_sem);
|
2005-04-16 15:20:36 -07:00
|
|
|
list_add_tail(&dev->bus_list, &bus->devices);
|
2006-06-01 21:35:43 -07:00
|
|
|
up_write(&pci_bus_sem);
|
2005-09-05 16:31:03 -07:00
|
|
|
}
|
|
|
|
|
2008-02-02 14:33:43 -07:00
|
|
|
struct pci_dev *__ref pci_scan_single_device(struct pci_bus *bus, int devfn)
|
2005-09-05 16:31:03 -07:00
|
|
|
{
|
|
|
|
struct pci_dev *dev;
|
|
|
|
|
|
|
|
dev = pci_scan_device(bus, devfn);
|
|
|
|
if (!dev)
|
|
|
|
return NULL;
|
|
|
|
|
|
|
|
pci_device_add(dev, bus);
|
2005-04-16 15:20:36 -07:00
|
|
|
|
|
|
|
return dev;
|
|
|
|
}
|
2007-11-21 16:07:11 -07:00
|
|
|
EXPORT_SYMBOL(pci_scan_single_device);
|
2005-04-16 15:20:36 -07:00
|
|
|
|
|
|
|
/**
|
|
|
|
* pci_scan_slot - scan a PCI slot on a bus for devices.
|
|
|
|
* @bus: PCI bus to scan
|
|
|
|
* @devfn: slot number to scan (must have zero function.)
|
|
|
|
*
|
|
|
|
* Scan a PCI slot on the specified PCI bus for devices, adding
|
|
|
|
* discovered devices to the @bus->devices list. New devices
|
2008-02-14 15:56:56 -07:00
|
|
|
* will not have is_added set.
|
2005-04-16 15:20:36 -07:00
|
|
|
*/
|
2007-03-26 22:53:30 -07:00
|
|
|
int pci_scan_slot(struct pci_bus *bus, int devfn)
|
2005-04-16 15:20:36 -07:00
|
|
|
{
|
|
|
|
int func, nr = 0;
|
|
|
|
int scan_all_fns;
|
|
|
|
|
|
|
|
scan_all_fns = pcibios_scan_all_fns(bus, devfn);
|
|
|
|
|
|
|
|
for (func = 0; func < 8; func++, devfn++) {
|
|
|
|
struct pci_dev *dev;
|
|
|
|
|
|
|
|
dev = pci_scan_single_device(bus, devfn);
|
|
|
|
if (dev) {
|
|
|
|
nr++;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* If this is a single function device,
|
|
|
|
* don't scan past the first function.
|
|
|
|
*/
|
|
|
|
if (!dev->multifunction) {
|
|
|
|
if (func > 0) {
|
|
|
|
dev->multifunction = 1;
|
|
|
|
} else {
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
} else {
|
|
|
|
if (func == 0 && !scan_all_fns)
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
PCI: add PCI Express ASPM support
PCI Express ASPM defines a protocol for PCI Express components in the D0
state to reduce Link power by placing their Links into a low power state
and instructing the other end of the Link to do likewise. This
capability allows hardware-autonomous, dynamic Link power reduction
beyond what is achievable by software-only controlled power management.
However, The device should be configured by software appropriately.
Enabling ASPM will save power, but will introduce device latency.
This patch adds ASPM support in Linux. It introduces a global policy for
ASPM, a sysfs file /sys/module/pcie_aspm/parameters/policy can control
it. The interface can be used as a boot option too. Currently we have
below setting:
-default, BIOS default setting
-powersave, highest power saving mode, enable all available ASPM
state and clock power management
-performance, highest performance, disable ASPM and clock power
management
By default, the 'default' policy is used currently.
In my test, power difference between powersave mode and performance mode
is about 1.3w in a system with 3 PCIE links.
Note: some devices might not work well with aspm, either because chipset
issue or device issue. The patch provide API (pci_disable_link_state),
driver can disable ASPM for specific device.
Signed-off-by: Shaohua Li <shaohua.li@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
2008-02-24 18:46:41 -07:00
|
|
|
|
2008-07-22 19:32:31 -07:00
|
|
|
/* only one slot has pcie device */
|
|
|
|
if (bus->self && nr)
|
PCI: add PCI Express ASPM support
PCI Express ASPM defines a protocol for PCI Express components in the D0
state to reduce Link power by placing their Links into a low power state
and instructing the other end of the Link to do likewise. This
capability allows hardware-autonomous, dynamic Link power reduction
beyond what is achievable by software-only controlled power management.
However, The device should be configured by software appropriately.
Enabling ASPM will save power, but will introduce device latency.
This patch adds ASPM support in Linux. It introduces a global policy for
ASPM, a sysfs file /sys/module/pcie_aspm/parameters/policy can control
it. The interface can be used as a boot option too. Currently we have
below setting:
-default, BIOS default setting
-powersave, highest power saving mode, enable all available ASPM
state and clock power management
-performance, highest performance, disable ASPM and clock power
management
By default, the 'default' policy is used currently.
In my test, power difference between powersave mode and performance mode
is about 1.3w in a system with 3 PCIE links.
Note: some devices might not work well with aspm, either because chipset
issue or device issue. The patch provide API (pci_disable_link_state),
driver can disable ASPM for specific device.
Signed-off-by: Shaohua Li <shaohua.li@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
2008-02-24 18:46:41 -07:00
|
|
|
pcie_aspm_init_link_state(bus->self);
|
|
|
|
|
2005-04-16 15:20:36 -07:00
|
|
|
return nr;
|
|
|
|
}
|
|
|
|
|
2008-02-17 02:45:28 -07:00
|
|
|
unsigned int __devinit pci_scan_child_bus(struct pci_bus *bus)
|
2005-04-16 15:20:36 -07:00
|
|
|
{
|
|
|
|
unsigned int devfn, pass, max = bus->secondary;
|
|
|
|
struct pci_dev *dev;
|
|
|
|
|
|
|
|
pr_debug("PCI: Scanning bus %04x:%02x\n", pci_domain_nr(bus), bus->number);
|
|
|
|
|
|
|
|
/* Go find them, Rover! */
|
|
|
|
for (devfn = 0; devfn < 0x100; devfn += 8)
|
|
|
|
pci_scan_slot(bus, devfn);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* After performing arch-dependent fixup of the bus, look behind
|
|
|
|
* all PCI-to-PCI bridges on this bus.
|
|
|
|
*/
|
|
|
|
pr_debug("PCI: Fixups for bus %04x:%02x\n", pci_domain_nr(bus), bus->number);
|
|
|
|
pcibios_fixup_bus(bus);
|
|
|
|
for (pass=0; pass < 2; pass++)
|
|
|
|
list_for_each_entry(dev, &bus->devices, bus_list) {
|
|
|
|
if (dev->hdr_type == PCI_HEADER_TYPE_BRIDGE ||
|
|
|
|
dev->hdr_type == PCI_HEADER_TYPE_CARDBUS)
|
|
|
|
max = pci_scan_bridge(bus, dev, max, pass);
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* We've scanned the bus and so we know all about what's on
|
|
|
|
* the other side of any bridges that may be on this bus plus
|
|
|
|
* any devices.
|
|
|
|
*
|
|
|
|
* Return how far we've got finding sub-buses.
|
|
|
|
*/
|
|
|
|
pr_debug("PCI: Bus scan for %04x:%02x returning with max=%02x\n",
|
|
|
|
pci_domain_nr(bus), bus->number, max);
|
|
|
|
return max;
|
|
|
|
}
|
|
|
|
|
2008-02-19 04:21:20 -07:00
|
|
|
void __attribute__((weak)) set_pci_bus_resources_arch_default(struct pci_bus *b)
|
|
|
|
{
|
|
|
|
}
|
|
|
|
|
2007-03-26 22:53:30 -07:00
|
|
|
struct pci_bus * pci_create_bus(struct device *parent,
|
2005-09-05 16:31:03 -07:00
|
|
|
int bus, struct pci_ops *ops, void *sysdata)
|
2005-04-16 15:20:36 -07:00
|
|
|
{
|
|
|
|
int error;
|
|
|
|
struct pci_bus *b;
|
|
|
|
struct device *dev;
|
|
|
|
|
|
|
|
b = pci_alloc_bus();
|
|
|
|
if (!b)
|
|
|
|
return NULL;
|
|
|
|
|
|
|
|
dev = kmalloc(sizeof(*dev), GFP_KERNEL);
|
|
|
|
if (!dev){
|
|
|
|
kfree(b);
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
b->sysdata = sysdata;
|
|
|
|
b->ops = ops;
|
|
|
|
|
|
|
|
if (pci_find_bus(pci_domain_nr(b), bus)) {
|
|
|
|
/* If we already got to this bus through a different bridge, ignore it */
|
|
|
|
pr_debug("PCI: Bus %04x:%02x already known\n", pci_domain_nr(b), bus);
|
|
|
|
goto err_out;
|
|
|
|
}
|
2006-06-01 21:35:43 -07:00
|
|
|
|
|
|
|
down_write(&pci_bus_sem);
|
2005-04-16 15:20:36 -07:00
|
|
|
list_add_tail(&b->node, &pci_root_buses);
|
2006-06-01 21:35:43 -07:00
|
|
|
up_write(&pci_bus_sem);
|
2005-04-16 15:20:36 -07:00
|
|
|
|
|
|
|
memset(dev, 0, sizeof(*dev));
|
|
|
|
dev->parent = parent;
|
|
|
|
dev->release = pci_release_bus_bridge_dev;
|
2008-10-29 18:17:49 -07:00
|
|
|
dev_set_name(dev, "pci%04x:%02x", pci_domain_nr(b), bus);
|
2005-04-16 15:20:36 -07:00
|
|
|
error = device_register(dev);
|
|
|
|
if (error)
|
|
|
|
goto dev_reg_err;
|
|
|
|
b->bridge = get_device(dev);
|
|
|
|
|
2008-02-19 04:20:41 -07:00
|
|
|
if (!parent)
|
|
|
|
set_dev_node(b->bridge, pcibus_to_node(b));
|
|
|
|
|
2007-05-22 19:47:54 -07:00
|
|
|
b->dev.class = &pcibus_class;
|
|
|
|
b->dev.parent = b->bridge;
|
2008-10-29 18:17:49 -07:00
|
|
|
dev_set_name(&b->dev, "%04x:%02x", pci_domain_nr(b), bus);
|
2007-05-22 19:47:54 -07:00
|
|
|
error = device_register(&b->dev);
|
2005-04-16 15:20:36 -07:00
|
|
|
if (error)
|
|
|
|
goto class_dev_reg_err;
|
2007-05-22 19:47:54 -07:00
|
|
|
error = device_create_file(&b->dev, &dev_attr_cpuaffinity);
|
2005-04-16 15:20:36 -07:00
|
|
|
if (error)
|
2007-05-22 19:47:54 -07:00
|
|
|
goto dev_create_file_err;
|
2005-04-16 15:20:36 -07:00
|
|
|
|
|
|
|
/* Create legacy_io and legacy_mem files for this bus */
|
|
|
|
pci_create_legacy_files(b);
|
|
|
|
|
|
|
|
b->number = b->secondary = bus;
|
|
|
|
b->resource[0] = &ioport_resource;
|
|
|
|
b->resource[1] = &iomem_resource;
|
|
|
|
|
2008-02-19 04:21:20 -07:00
|
|
|
set_pci_bus_resources_arch_default(b);
|
|
|
|
|
2005-04-16 15:20:36 -07:00
|
|
|
return b;
|
|
|
|
|
2007-05-22 19:47:54 -07:00
|
|
|
dev_create_file_err:
|
|
|
|
device_unregister(&b->dev);
|
2005-04-16 15:20:36 -07:00
|
|
|
class_dev_reg_err:
|
|
|
|
device_unregister(dev);
|
|
|
|
dev_reg_err:
|
2006-06-01 21:35:43 -07:00
|
|
|
down_write(&pci_bus_sem);
|
2005-04-16 15:20:36 -07:00
|
|
|
list_del(&b->node);
|
2006-06-01 21:35:43 -07:00
|
|
|
up_write(&pci_bus_sem);
|
2005-04-16 15:20:36 -07:00
|
|
|
err_out:
|
|
|
|
kfree(dev);
|
|
|
|
kfree(b);
|
|
|
|
return NULL;
|
|
|
|
}
|
2005-09-05 16:31:03 -07:00
|
|
|
|
2008-02-17 02:45:28 -07:00
|
|
|
struct pci_bus * __devinit pci_scan_bus_parented(struct device *parent,
|
2005-09-05 16:31:03 -07:00
|
|
|
int bus, struct pci_ops *ops, void *sysdata)
|
|
|
|
{
|
|
|
|
struct pci_bus *b;
|
|
|
|
|
|
|
|
b = pci_create_bus(parent, bus, ops, sysdata);
|
|
|
|
if (b)
|
|
|
|
b->subordinate = pci_scan_child_bus(b);
|
|
|
|
return b;
|
|
|
|
}
|
2005-04-16 15:20:36 -07:00
|
|
|
EXPORT_SYMBOL(pci_scan_bus_parented);
|
|
|
|
|
|
|
|
#ifdef CONFIG_HOTPLUG
|
|
|
|
EXPORT_SYMBOL(pci_add_new_bus);
|
|
|
|
EXPORT_SYMBOL(pci_scan_slot);
|
|
|
|
EXPORT_SYMBOL(pci_scan_bridge);
|
|
|
|
EXPORT_SYMBOL_GPL(pci_scan_child_bus);
|
|
|
|
#endif
|
PCI: optionally sort device lists breadth-first
Problem:
New Dell PowerEdge servers have 2 embedded ethernet ports, which are
labeled NIC1 and NIC2 on the chassis, in the BIOS setup screens, and
in the printed documentation. Assuming no other add-in ethernet ports
in the system, Linux 2.4 kernels name these eth0 and eth1
respectively. Many people have come to expect this naming. Linux 2.6
kernels name these eth1 and eth0 respectively (backwards from
expectations). I also have reports that various Sun and HP servers
have similar behavior.
Root cause:
Linux 2.4 kernels walk the pci_devices list, which happens to be
sorted in breadth-first order (or pcbios_find_device order on i386,
which most often is breadth-first also). 2.6 kernels have both the
pci_devices list and the pci_bus_type.klist_devices list, the latter
is what is walked at driver load time to match the pci_id tables; this
klist happens to be in depth-first order.
On systems where, for physical routing reasons, NIC1 appears on a
lower bus number than NIC2, but NIC2's bridge is discovered first in
the depth-first ordering, NIC2 will be discovered before NIC1. If the
list were sorted breadth-first, NIC1 would be discovered before NIC2.
A PowerEdge 1955 system has the following topology which easily
exhibits the difference between depth-first and breadth-first device
lists.
-[0000:00]-+-00.0 Intel Corporation 5000P Chipset Memory Controller Hub
+-02.0-[0000:03-08]--+-00.0-[0000:04-07]--+-00.0-[0000:05-06]----00.0-[0000:06]----00.0 Broadcom Corporation NetXtreme II BCM5708S Gigabit Ethernet (labeled NIC2, 2.4 kernel name eth1, 2.6 kernel name eth0)
+-1c.0-[0000:01-02]----00.0-[0000:02]----00.0 Broadcom Corporation NetXtreme II BCM5708S Gigabit Ethernet (labeled NIC1, 2.4 kernel name eth0, 2.6 kernel name eth1)
Other factors, such as device driver load order and the presence of
PCI slots at various points in the bus hierarchy further complicate
this problem; I'm not trying to solve those here, just restore the
device order, and thus basic behavior, that 2.4 kernels had.
Solution:
The solution can come in multiple steps.
Suggested fix #1: kernel
Patch below optionally sorts the two device lists into breadth-first
ordering to maintain compatibility with 2.4 kernels. It adds two new
command line options:
pci=bfsort
pci=nobfsort
to force the sort order, or not, as you wish. It also adds DMI checks
for the specific Dell systems which exhibit "backwards" ordering, to
make them "right".
Suggested fix #2: udev rules from userland
Many people also have the expectation that embedded NICs are always
discovered before add-in NICs (which this patch does not try to do).
Using the PCI IRQ Routing Table provided by system BIOS, it's easy to
determine which PCI devices are embedded, or if add-in, which PCI slot
they're in. I'm working on a tool that would allow udev to name
ethernet devices in ascending embedded, slot 1 .. slot N order,
subsort by PCI bus/dev/fn breadth-first. It'll be possible to use it
independent of udev as well for those distributions that don't use
udev in their installers.
Suggested fix #3: system board routing rules
One can constrain the system board layout to put NIC1 ahead of NIC2
regardless of breadth-first or depth-first discovery order. This adds
a significant level of complexity to board routing, and may not be
possible in all instances (witness the above systems from several
major manufacturers). I don't want to encourage this particular train
of thought too far, at the expense of not doing #1 or #2 above.
Feedback appreciated. Patch tested on a Dell PowerEdge 1955 blade
with 2.6.18.
You'll also note I took some liberty and temporarily break the klist
abstraction to simplify and speed up the sort algorithm. I think
that's both safe and appropriate in this instance.
Signed-off-by: Matt Domsch <Matt_Domsch@dell.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
2006-09-29 13:23:23 -07:00
|
|
|
|
2008-08-26 09:00:57 -07:00
|
|
|
static int __init pci_sort_bf_cmp(const struct device *d_a, const struct device *d_b)
|
PCI: optionally sort device lists breadth-first
Problem:
New Dell PowerEdge servers have 2 embedded ethernet ports, which are
labeled NIC1 and NIC2 on the chassis, in the BIOS setup screens, and
in the printed documentation. Assuming no other add-in ethernet ports
in the system, Linux 2.4 kernels name these eth0 and eth1
respectively. Many people have come to expect this naming. Linux 2.6
kernels name these eth1 and eth0 respectively (backwards from
expectations). I also have reports that various Sun and HP servers
have similar behavior.
Root cause:
Linux 2.4 kernels walk the pci_devices list, which happens to be
sorted in breadth-first order (or pcbios_find_device order on i386,
which most often is breadth-first also). 2.6 kernels have both the
pci_devices list and the pci_bus_type.klist_devices list, the latter
is what is walked at driver load time to match the pci_id tables; this
klist happens to be in depth-first order.
On systems where, for physical routing reasons, NIC1 appears on a
lower bus number than NIC2, but NIC2's bridge is discovered first in
the depth-first ordering, NIC2 will be discovered before NIC1. If the
list were sorted breadth-first, NIC1 would be discovered before NIC2.
A PowerEdge 1955 system has the following topology which easily
exhibits the difference between depth-first and breadth-first device
lists.
-[0000:00]-+-00.0 Intel Corporation 5000P Chipset Memory Controller Hub
+-02.0-[0000:03-08]--+-00.0-[0000:04-07]--+-00.0-[0000:05-06]----00.0-[0000:06]----00.0 Broadcom Corporation NetXtreme II BCM5708S Gigabit Ethernet (labeled NIC2, 2.4 kernel name eth1, 2.6 kernel name eth0)
+-1c.0-[0000:01-02]----00.0-[0000:02]----00.0 Broadcom Corporation NetXtreme II BCM5708S Gigabit Ethernet (labeled NIC1, 2.4 kernel name eth0, 2.6 kernel name eth1)
Other factors, such as device driver load order and the presence of
PCI slots at various points in the bus hierarchy further complicate
this problem; I'm not trying to solve those here, just restore the
device order, and thus basic behavior, that 2.4 kernels had.
Solution:
The solution can come in multiple steps.
Suggested fix #1: kernel
Patch below optionally sorts the two device lists into breadth-first
ordering to maintain compatibility with 2.4 kernels. It adds two new
command line options:
pci=bfsort
pci=nobfsort
to force the sort order, or not, as you wish. It also adds DMI checks
for the specific Dell systems which exhibit "backwards" ordering, to
make them "right".
Suggested fix #2: udev rules from userland
Many people also have the expectation that embedded NICs are always
discovered before add-in NICs (which this patch does not try to do).
Using the PCI IRQ Routing Table provided by system BIOS, it's easy to
determine which PCI devices are embedded, or if add-in, which PCI slot
they're in. I'm working on a tool that would allow udev to name
ethernet devices in ascending embedded, slot 1 .. slot N order,
subsort by PCI bus/dev/fn breadth-first. It'll be possible to use it
independent of udev as well for those distributions that don't use
udev in their installers.
Suggested fix #3: system board routing rules
One can constrain the system board layout to put NIC1 ahead of NIC2
regardless of breadth-first or depth-first discovery order. This adds
a significant level of complexity to board routing, and may not be
possible in all instances (witness the above systems from several
major manufacturers). I don't want to encourage this particular train
of thought too far, at the expense of not doing #1 or #2 above.
Feedback appreciated. Patch tested on a Dell PowerEdge 1955 blade
with 2.6.18.
You'll also note I took some liberty and temporarily break the klist
abstraction to simplify and speed up the sort algorithm. I think
that's both safe and appropriate in this instance.
Signed-off-by: Matt Domsch <Matt_Domsch@dell.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
2006-09-29 13:23:23 -07:00
|
|
|
{
|
2008-08-26 09:00:57 -07:00
|
|
|
const struct pci_dev *a = to_pci_dev(d_a);
|
|
|
|
const struct pci_dev *b = to_pci_dev(d_b);
|
|
|
|
|
PCI: optionally sort device lists breadth-first
Problem:
New Dell PowerEdge servers have 2 embedded ethernet ports, which are
labeled NIC1 and NIC2 on the chassis, in the BIOS setup screens, and
in the printed documentation. Assuming no other add-in ethernet ports
in the system, Linux 2.4 kernels name these eth0 and eth1
respectively. Many people have come to expect this naming. Linux 2.6
kernels name these eth1 and eth0 respectively (backwards from
expectations). I also have reports that various Sun and HP servers
have similar behavior.
Root cause:
Linux 2.4 kernels walk the pci_devices list, which happens to be
sorted in breadth-first order (or pcbios_find_device order on i386,
which most often is breadth-first also). 2.6 kernels have both the
pci_devices list and the pci_bus_type.klist_devices list, the latter
is what is walked at driver load time to match the pci_id tables; this
klist happens to be in depth-first order.
On systems where, for physical routing reasons, NIC1 appears on a
lower bus number than NIC2, but NIC2's bridge is discovered first in
the depth-first ordering, NIC2 will be discovered before NIC1. If the
list were sorted breadth-first, NIC1 would be discovered before NIC2.
A PowerEdge 1955 system has the following topology which easily
exhibits the difference between depth-first and breadth-first device
lists.
-[0000:00]-+-00.0 Intel Corporation 5000P Chipset Memory Controller Hub
+-02.0-[0000:03-08]--+-00.0-[0000:04-07]--+-00.0-[0000:05-06]----00.0-[0000:06]----00.0 Broadcom Corporation NetXtreme II BCM5708S Gigabit Ethernet (labeled NIC2, 2.4 kernel name eth1, 2.6 kernel name eth0)
+-1c.0-[0000:01-02]----00.0-[0000:02]----00.0 Broadcom Corporation NetXtreme II BCM5708S Gigabit Ethernet (labeled NIC1, 2.4 kernel name eth0, 2.6 kernel name eth1)
Other factors, such as device driver load order and the presence of
PCI slots at various points in the bus hierarchy further complicate
this problem; I'm not trying to solve those here, just restore the
device order, and thus basic behavior, that 2.4 kernels had.
Solution:
The solution can come in multiple steps.
Suggested fix #1: kernel
Patch below optionally sorts the two device lists into breadth-first
ordering to maintain compatibility with 2.4 kernels. It adds two new
command line options:
pci=bfsort
pci=nobfsort
to force the sort order, or not, as you wish. It also adds DMI checks
for the specific Dell systems which exhibit "backwards" ordering, to
make them "right".
Suggested fix #2: udev rules from userland
Many people also have the expectation that embedded NICs are always
discovered before add-in NICs (which this patch does not try to do).
Using the PCI IRQ Routing Table provided by system BIOS, it's easy to
determine which PCI devices are embedded, or if add-in, which PCI slot
they're in. I'm working on a tool that would allow udev to name
ethernet devices in ascending embedded, slot 1 .. slot N order,
subsort by PCI bus/dev/fn breadth-first. It'll be possible to use it
independent of udev as well for those distributions that don't use
udev in their installers.
Suggested fix #3: system board routing rules
One can constrain the system board layout to put NIC1 ahead of NIC2
regardless of breadth-first or depth-first discovery order. This adds
a significant level of complexity to board routing, and may not be
possible in all instances (witness the above systems from several
major manufacturers). I don't want to encourage this particular train
of thought too far, at the expense of not doing #1 or #2 above.
Feedback appreciated. Patch tested on a Dell PowerEdge 1955 blade
with 2.6.18.
You'll also note I took some liberty and temporarily break the klist
abstraction to simplify and speed up the sort algorithm. I think
that's both safe and appropriate in this instance.
Signed-off-by: Matt Domsch <Matt_Domsch@dell.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
2006-09-29 13:23:23 -07:00
|
|
|
if (pci_domain_nr(a->bus) < pci_domain_nr(b->bus)) return -1;
|
|
|
|
else if (pci_domain_nr(a->bus) > pci_domain_nr(b->bus)) return 1;
|
|
|
|
|
|
|
|
if (a->bus->number < b->bus->number) return -1;
|
|
|
|
else if (a->bus->number > b->bus->number) return 1;
|
|
|
|
|
|
|
|
if (a->devfn < b->devfn) return -1;
|
|
|
|
else if (a->devfn > b->devfn) return 1;
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2008-02-14 15:56:56 -07:00
|
|
|
void __init pci_sort_breadthfirst(void)
|
PCI: optionally sort device lists breadth-first
Problem:
New Dell PowerEdge servers have 2 embedded ethernet ports, which are
labeled NIC1 and NIC2 on the chassis, in the BIOS setup screens, and
in the printed documentation. Assuming no other add-in ethernet ports
in the system, Linux 2.4 kernels name these eth0 and eth1
respectively. Many people have come to expect this naming. Linux 2.6
kernels name these eth1 and eth0 respectively (backwards from
expectations). I also have reports that various Sun and HP servers
have similar behavior.
Root cause:
Linux 2.4 kernels walk the pci_devices list, which happens to be
sorted in breadth-first order (or pcbios_find_device order on i386,
which most often is breadth-first also). 2.6 kernels have both the
pci_devices list and the pci_bus_type.klist_devices list, the latter
is what is walked at driver load time to match the pci_id tables; this
klist happens to be in depth-first order.
On systems where, for physical routing reasons, NIC1 appears on a
lower bus number than NIC2, but NIC2's bridge is discovered first in
the depth-first ordering, NIC2 will be discovered before NIC1. If the
list were sorted breadth-first, NIC1 would be discovered before NIC2.
A PowerEdge 1955 system has the following topology which easily
exhibits the difference between depth-first and breadth-first device
lists.
-[0000:00]-+-00.0 Intel Corporation 5000P Chipset Memory Controller Hub
+-02.0-[0000:03-08]--+-00.0-[0000:04-07]--+-00.0-[0000:05-06]----00.0-[0000:06]----00.0 Broadcom Corporation NetXtreme II BCM5708S Gigabit Ethernet (labeled NIC2, 2.4 kernel name eth1, 2.6 kernel name eth0)
+-1c.0-[0000:01-02]----00.0-[0000:02]----00.0 Broadcom Corporation NetXtreme II BCM5708S Gigabit Ethernet (labeled NIC1, 2.4 kernel name eth0, 2.6 kernel name eth1)
Other factors, such as device driver load order and the presence of
PCI slots at various points in the bus hierarchy further complicate
this problem; I'm not trying to solve those here, just restore the
device order, and thus basic behavior, that 2.4 kernels had.
Solution:
The solution can come in multiple steps.
Suggested fix #1: kernel
Patch below optionally sorts the two device lists into breadth-first
ordering to maintain compatibility with 2.4 kernels. It adds two new
command line options:
pci=bfsort
pci=nobfsort
to force the sort order, or not, as you wish. It also adds DMI checks
for the specific Dell systems which exhibit "backwards" ordering, to
make them "right".
Suggested fix #2: udev rules from userland
Many people also have the expectation that embedded NICs are always
discovered before add-in NICs (which this patch does not try to do).
Using the PCI IRQ Routing Table provided by system BIOS, it's easy to
determine which PCI devices are embedded, or if add-in, which PCI slot
they're in. I'm working on a tool that would allow udev to name
ethernet devices in ascending embedded, slot 1 .. slot N order,
subsort by PCI bus/dev/fn breadth-first. It'll be possible to use it
independent of udev as well for those distributions that don't use
udev in their installers.
Suggested fix #3: system board routing rules
One can constrain the system board layout to put NIC1 ahead of NIC2
regardless of breadth-first or depth-first discovery order. This adds
a significant level of complexity to board routing, and may not be
possible in all instances (witness the above systems from several
major manufacturers). I don't want to encourage this particular train
of thought too far, at the expense of not doing #1 or #2 above.
Feedback appreciated. Patch tested on a Dell PowerEdge 1955 blade
with 2.6.18.
You'll also note I took some liberty and temporarily break the klist
abstraction to simplify and speed up the sort algorithm. I think
that's both safe and appropriate in this instance.
Signed-off-by: Matt Domsch <Matt_Domsch@dell.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
2006-09-29 13:23:23 -07:00
|
|
|
{
|
2008-08-26 09:00:57 -07:00
|
|
|
bus_sort_breadthfirst(&pci_bus_type, &pci_sort_bf_cmp);
|
PCI: optionally sort device lists breadth-first
Problem:
New Dell PowerEdge servers have 2 embedded ethernet ports, which are
labeled NIC1 and NIC2 on the chassis, in the BIOS setup screens, and
in the printed documentation. Assuming no other add-in ethernet ports
in the system, Linux 2.4 kernels name these eth0 and eth1
respectively. Many people have come to expect this naming. Linux 2.6
kernels name these eth1 and eth0 respectively (backwards from
expectations). I also have reports that various Sun and HP servers
have similar behavior.
Root cause:
Linux 2.4 kernels walk the pci_devices list, which happens to be
sorted in breadth-first order (or pcbios_find_device order on i386,
which most often is breadth-first also). 2.6 kernels have both the
pci_devices list and the pci_bus_type.klist_devices list, the latter
is what is walked at driver load time to match the pci_id tables; this
klist happens to be in depth-first order.
On systems where, for physical routing reasons, NIC1 appears on a
lower bus number than NIC2, but NIC2's bridge is discovered first in
the depth-first ordering, NIC2 will be discovered before NIC1. If the
list were sorted breadth-first, NIC1 would be discovered before NIC2.
A PowerEdge 1955 system has the following topology which easily
exhibits the difference between depth-first and breadth-first device
lists.
-[0000:00]-+-00.0 Intel Corporation 5000P Chipset Memory Controller Hub
+-02.0-[0000:03-08]--+-00.0-[0000:04-07]--+-00.0-[0000:05-06]----00.0-[0000:06]----00.0 Broadcom Corporation NetXtreme II BCM5708S Gigabit Ethernet (labeled NIC2, 2.4 kernel name eth1, 2.6 kernel name eth0)
+-1c.0-[0000:01-02]----00.0-[0000:02]----00.0 Broadcom Corporation NetXtreme II BCM5708S Gigabit Ethernet (labeled NIC1, 2.4 kernel name eth0, 2.6 kernel name eth1)
Other factors, such as device driver load order and the presence of
PCI slots at various points in the bus hierarchy further complicate
this problem; I'm not trying to solve those here, just restore the
device order, and thus basic behavior, that 2.4 kernels had.
Solution:
The solution can come in multiple steps.
Suggested fix #1: kernel
Patch below optionally sorts the two device lists into breadth-first
ordering to maintain compatibility with 2.4 kernels. It adds two new
command line options:
pci=bfsort
pci=nobfsort
to force the sort order, or not, as you wish. It also adds DMI checks
for the specific Dell systems which exhibit "backwards" ordering, to
make them "right".
Suggested fix #2: udev rules from userland
Many people also have the expectation that embedded NICs are always
discovered before add-in NICs (which this patch does not try to do).
Using the PCI IRQ Routing Table provided by system BIOS, it's easy to
determine which PCI devices are embedded, or if add-in, which PCI slot
they're in. I'm working on a tool that would allow udev to name
ethernet devices in ascending embedded, slot 1 .. slot N order,
subsort by PCI bus/dev/fn breadth-first. It'll be possible to use it
independent of udev as well for those distributions that don't use
udev in their installers.
Suggested fix #3: system board routing rules
One can constrain the system board layout to put NIC1 ahead of NIC2
regardless of breadth-first or depth-first discovery order. This adds
a significant level of complexity to board routing, and may not be
possible in all instances (witness the above systems from several
major manufacturers). I don't want to encourage this particular train
of thought too far, at the expense of not doing #1 or #2 above.
Feedback appreciated. Patch tested on a Dell PowerEdge 1955 blade
with 2.6.18.
You'll also note I took some liberty and temporarily break the klist
abstraction to simplify and speed up the sort algorithm. I think
that's both safe and appropriate in this instance.
Signed-off-by: Matt Domsch <Matt_Domsch@dell.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
2006-09-29 13:23:23 -07:00
|
|
|
}
|