* Re: [PATCH 1/4] Add ps3_storage module alias
From: David Woodhouse @ 2010-02-23 12:47 UTC (permalink / raw)
To: Geert Uytterhoeven; +Cc: linuxppc-dev
In-Reply-To: <alpine.LRH.2.00.1002051446280.5981@mink.sonytel.be>
On Fri, 2010-02-05 at 14:49 +0100, Geert Uytterhoeven wrote:
> On Fri, 5 Feb 2010, Josh Boyer wrote:
> > This allows the upgrade path from older kernel releases to work
> > with the Fedora mkinitrd tool.
> >
> > ---
> >
> > --- linux-2.6.22.ppc64/drivers/block/ps3disk.c~ 2007-07-25 16:06:16.000000000 +0100
> > +++ linux-2.6.22.ppc64/drivers/block/ps3disk.c 2007-07-26 08:49:44.000000000 +0100
> > @@ -628,3 +628,4 @@ MODULE_LICENSE("GPL");
> > MODULE_DESCRIPTION("PS3 Disk Storage Driver");
> > MODULE_AUTHOR("Sony Corporation");
> > MODULE_ALIAS(PS3_MODULE_ALIAS_STOR_DISK);
> > +MODULE_ALIAS("ps3_storage");
>
> NAKed-by: Geert Uytterhoeven <Geert.Uytterhoeven@sonycom.com>
>
> ps3_storage was replaced by 3 drivers: ps3disk, ps3rom, and ps3flash. These are
> automatically loaded by udev.
>
> How does mkinitrd determine the driver for the root file system on other
> systems? Does it really contain a big list of e.g. all SCSI drivers?
It pokes about in sysfs and works out the driver which is used for the
current root devices.
This hack was only needed for the upgrade -- when you were running a
kernel with the ps3_storage kernel and you first installed a kernel with
the new ps3disk driver, you needed it to find the right module.
I think we can drop it now.
--
David Woodhouse Open Source Technology Centre
David.Woodhouse@intel.com Intel Corporation
^ permalink raw reply
* Re: [PATCH 2/4] Fix G5 thermal shutdown
From: David Woodhouse @ 2010-02-23 12:48 UTC (permalink / raw)
To: Benjamin Herrenschmidt; +Cc: linuxppc-dev
In-Reply-To: <1265676905.8287.254.camel@pasglop>
On Tue, 2010-02-09 at 11:55 +1100, Benjamin Herrenschmidt wrote:
> On Fri, 2010-02-05 at 08:52 -0500, Josh Boyer wrote:
> > This changes the thresholds for the liquid cooled G5 thermal
> > shutdown mechanism to prevent an errant shutdown.
> >
> > This has been carried since about Fedora Core 5. I have no idea
> > if it's really needed or not.
It's kind of a dirty hack, but it was sufficient to stop my machine
constantly powering itself off.
> Can we have a SoB ?
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
--
David Woodhouse Open Source Technology Centre
David.Woodhouse@intel.com Intel Corporation
^ permalink raw reply
* Re: [PATCH 3/4] Provide VIO modalias
From: David Woodhouse @ 2010-02-23 12:50 UTC (permalink / raw)
To: Josh Boyer; +Cc: linuxppc-dev
In-Reply-To: <20100205135354.GI12001@hansolo.jdub.homelinux.org>
On Fri, 2010-02-05 at 08:53 -0500, Josh Boyer wrote:
> Provide a modalias entry for VIO devices in sysfs. I believe
> this was another initrd generation bugfix for anaconda.
>
> ---
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
Not entirely sure why this didn't get upstream when I first did it;
maybe I just posted it once and then forgot about it?
> diff --git a/arch/powerpc/kernel/vio.c b/arch/powerpc/kernel/vio.c
> index f988672..12a0851 100644
> --- a/arch/powerpc/kernel/vio.c
> +++ b/arch/powerpc/kernel/vio.c
> @@ -294,9 +294,27 @@ static ssize_t devspec_show(struct device *dev,
> return sprintf(buf, "%s\n", of_node ? of_node->full_name : "none");
> }
>
> +static ssize_t modalias_show(struct device *dev, struct device_attribute *attr,
> + char *buf)
> +{
> + const struct vio_dev *vio_dev = to_vio_dev(dev);
> + struct device_node *dn;
> + const char *cp;
> +
> + dn = dev->archdata.of_node;
> + if (!dn)
> + return -ENODEV;
> + cp = of_get_property(dn, "compatible", NULL);
> + if (!cp)
> + return -ENODEV;
> +
> + return sprintf(buf, "vio:T%sS%s\n", vio_dev->type, cp);
> +}
> +
> static struct device_attribute vio_dev_attrs[] = {
> __ATTR_RO(name),
> __ATTR_RO(devspec),
> + __ATTR_RO(modalias),
> __ATTR_NULL
> };
>
> _______________________________________________
> Linuxppc-dev mailing list
> Linuxppc-dev@lists.ozlabs.org
> https://lists.ozlabs.org/listinfo/linuxppc-dev
--
David Woodhouse Open Source Technology Centre
David.Woodhouse@intel.com Intel Corporation
^ permalink raw reply
* Re: [PATCH 4/4] Fix iMac iSight PCI bridge setup
From: David Woodhouse @ 2010-02-23 12:50 UTC (permalink / raw)
To: Josh Boyer; +Cc: linuxppc-dev
In-Reply-To: <20100206010314.GP12001@hansolo.jdub.homelinux.org>
On Fri, 2010-02-05 at 20:03 -0500, Josh Boyer wrote:
> On Sat, Feb 06, 2010 at 07:57:46AM +1100, Benjamin Herrenschmidt wrote:
> >On Fri, 2010-02-05 at 08:55 -0500, Josh Boyer wrote:
> >> This works around the PCIe bridge setup on the iMac iSight boxen.
> >
> >Is this still needed ? I though I had fixed the root cause (incorrect
> >setup of the bridge) a while back...
>
> I honestly have no clue, and I have no means of actually testing this. I have
> an iMac running Fedora, but not the iSight version. If you think it's fixed
> then I can just drop the patch from the Fedora kernel and we can wait for
> either silence or bug reports :)
I'll dig out the machine and test. Some time next month would be the
current estimate.
--
David Woodhouse Open Source Technology Centre
David.Woodhouse@intel.com Intel Corporation
^ permalink raw reply
* Re: [PATCH v4 04/11] mtd: Add MPC5121 NAND Flash Controller driver
From: Kári Davíðsson @ 2010-02-23 15:54 UTC (permalink / raw)
To: dedekind1@gmail.com
Cc: Piotr Ziecik, Detlev Zundel, linuxppc-dev@ozlabs.org,
linux-mtd@lists.infradead.org, Anatolij Gustschin,
David Woodhouse
In-Reply-To: <1266307868.11659.177.camel@localhost>
Any special reason this in MPC5121 Revision 2 only?
I did not see anything that contradicted the rev 3 manual in the code.
But there is one harmless bug (since the NFC_CMD_FAIL symbol is never used), i.e.
git diff
diff --git a/drivers/mtd/nand/mpc5121_nfc.c b/drivers/mtd/nand/mpc5121_nfc.c
index 191bf99..38a99ef 100644
--- a/drivers/mtd/nand/mpc5121_nfc.c
+++ b/drivers/mtd/nand/mpc5121_nfc.c
@@ -100,7 +100,7 @@
#define NFC_OUTPUT (1 << 3)
#define NFC_ID (1 << 4)
#define NFC_STATUS (1 << 5)
-#define NFC_CMD_FAIL (1 << 15)
+#define NFC_CMD_FAIL (1 << 14)
#define NFC_INT (1 << 15)
/* Bit Definitions: NFC_WRPROT */
The NAND chip is is detected (given that I disable the Revison check),
but the /dev7mtd0 is not accessable as block device. Is that normal?
The UBIFS also gives up I am assuming because the /dev/mtd0 is not block device.
rg
kd
On 02/16/2010 08:11 AM, Artem Bityutskiy wrote:
> On Mon, 2010-02-15 at 18:35 +0100, Anatolij Gustschin wrote:
>> Adds NAND Flash Controller driver for MPC5121 Revision 2.
>> All device features, except hardware ECC and power management,
>> are supported.
>>
>> Signed-off-by: Piotr Ziecik<kosmo@semihalf.com>
>> Signed-off-by: Wolfgang Denk<wd@denx.de>
>> Signed-off-by: Anatolij Gustschin<agust@denx.de>
>> Acked-by: Grant Likely<grant.likely@secretlab.ca>
>> Cc:<linux-mtd@lists.infradead.org>
>> Cc: Grant Likely<grant.likely@secretlab.ca>
>> Cc: John Rigby<jcrigby@gmail.com>
>> ---
>> Changes since v3:
>> - include 'asm/mpc5121.h' header instead of 'asm/mpc5xxx.h'.
>> This change is needed because arch patch adding reset
>> module definition was reworked and doesn't add mpc5121
>> specific definitions to common header for mpc52xx/mpc5121.
>>
>> Changes since v2:
>> - move the arch bits into separate patch
>> (it is the next patch in this series now)
>> - use __devinit/__devexit/__devexit_p and __devinitdata
>>
>> Changes since v1:
>> - add logfile with changes since previous version
>>
>> Changes since the patch version submitted in May 2009:
>>
>> - move mpc5121_nfc.h to the driver .c as there is only one user
>> - remove DRV_VERSION macro
>> - replace printk() by dev_*()
>> - drop unnecessary .suspend and .resume initializations
>> - remove duplicate .name/.owner settings
>> - fix mpc5121_nfc_init() to "return of_register_platform_driver(&mpc5121_nfc_driver);"
>> - move module_init() to just below the init function
>> - remove MODULE_VERSION
>> - use "mtd: Add MPC5121 NAND Flash Controller driver" as the subject,
>> previously it was "mpc5121: Added NAND Flash Controller driver.
>>
>> drivers/mtd/nand/Kconfig | 7 +
>> drivers/mtd/nand/Makefile | 1 +
>> drivers/mtd/nand/mpc5121_nfc.c | 916 ++++++++++++++++++++++++++++++++++++++++
>> 3 files changed, 924 insertions(+), 0 deletions(-)
>> create mode 100644 drivers/mtd/nand/mpc5121_nfc.c
>
> Pushed to my l2-mtd-2.6.git / dunno.
>
^ permalink raw reply related
* Re: [PATCH v4 4/5] PCI: augment bus resource table with a list
From: Bjorn Helgaas @ 2010-02-23 16:08 UTC (permalink / raw)
To: Jesse Barnes
Cc: Matthew Garrett, Tony Luck, linuxppc-dev, linux-pci, Peter Haight,
Gary Hade, linux-kernel, linux-acpi, linux-am33-list,
Linus Torvalds, Ingo Molnar, Yinghai Lu, Larry Finger
In-Reply-To: <20100217202008.4013.56938.stgit@bob.kio>
Here's the update (only added EXPORT_SYMBOL_GPL for pci_resource_n).
I'll repost the whole series if that's easier for you.
Bjorn
commit 72b4d74167af9cec8e68cd6d54eda5d3cb894c57
Author: Bjorn Helgaas <bjorn.helgaas@hp.com>
Date: Tue Feb 9 15:33:27 2010 -0700
PCI: augment bus resource table with a list
Previously we used a table of size PCI_BUS_NUM_RESOURCES (16) for resources
forwarded to a bus by its upstream bridge. We've increased this size
several times when the table overflowed.
But there's no good limit on the number of resources because host bridges
and subtractive decode bridges can forward any number of ranges to their
secondary buses.
This patch reduces the table to only PCI_BRIDGE_RESOURCE_NUM (4) entries,
which corresponds to the number of windows a PCI-to-PCI (3) or CardBus (4)
bridge can positively decode. Any additional resources, e.g., PCI host
bridge windows or subtractively-decoded regions, are kept in a list.
I'd prefer a single list rather than this split table/list approach, but
that requires simultaneous changes to every architecture. This approach
only requires immediate changes where we set up (a) host bridges with more
than four windows and (b) subtractive-decode P2P bridges, and we can
incrementally change other architectures to use the list.
Signed-off-by: Bjorn Helgaas <bjorn.helgaas@hp.com>
diff --git a/arch/ia64/pci/pci.c b/arch/ia64/pci/pci.c
index 89f957c..64aff52 100644
--- a/arch/ia64/pci/pci.c
+++ b/arch/ia64/pci/pci.c
@@ -320,9 +320,9 @@ static __devinit acpi_status add_window(struct acpi_resource *res, void *data)
static void __devinit
pcibios_setup_root_windows(struct pci_bus *bus, struct pci_controller *ctrl)
{
- int i, j;
+ int i;
- j = 0;
+ pci_bus_remove_resources(bus);
for (i = 0; i < ctrl->windows; i++) {
struct resource *res = &ctrl->window[i].resource;
/* HP's firmware has a hack to work around a Windows bug.
@@ -330,13 +330,7 @@ pcibios_setup_root_windows(struct pci_bus *bus, struct pci_controller *ctrl)
if ((res->flags & IORESOURCE_MEM) &&
(res->end - res->start < 16))
continue;
- if (j >= PCI_BUS_NUM_RESOURCES) {
- dev_warn(&bus->dev,
- "ignoring host bridge window %pR (no space)\n",
- res);
- continue;
- }
- bus->resource[j++] = res;
+ pci_bus_add_resource(bus, res, 0);
}
}
diff --git a/arch/x86/pci/acpi.c b/arch/x86/pci/acpi.c
index 959e548..a2f8cdb 100644
--- a/arch/x86/pci/acpi.c
+++ b/arch/x86/pci/acpi.c
@@ -45,20 +45,6 @@ count_resource(struct acpi_resource *acpi_res, void *data)
return AE_OK;
}
-static int
-bus_has_transparent_bridge(struct pci_bus *bus)
-{
- struct pci_dev *dev;
-
- list_for_each_entry(dev, &bus->devices, bus_list) {
- u16 class = dev->class >> 8;
-
- if (class == PCI_CLASS_BRIDGE_PCI && dev->transparent)
- return true;
- }
- return false;
-}
-
static void
align_resource(struct acpi_device *bridge, struct resource *res)
{
@@ -92,12 +78,8 @@ setup_resource(struct acpi_resource *acpi_res, void *data)
acpi_status status;
unsigned long flags;
struct resource *root;
- int max_root_bus_resources = PCI_BUS_NUM_RESOURCES;
u64 start, end;
- if (bus_has_transparent_bridge(info->bus))
- max_root_bus_resources -= 3;
-
status = resource_to_addr(acpi_res, &addr);
if (!ACPI_SUCCESS(status))
return AE_OK;
@@ -115,15 +97,6 @@ setup_resource(struct acpi_resource *acpi_res, void *data)
start = addr.minimum + addr.translation_offset;
end = start + addr.address_length - 1;
- if (info->res_num >= max_root_bus_resources) {
- if (pci_probe & PCI_USE__CRS)
- printk(KERN_WARNING "PCI: Failed to allocate "
- "0x%lx-0x%lx from %s for %s due to _CRS "
- "returning more than %d resource descriptors\n",
- (unsigned long) start, (unsigned long) end,
- root->name, info->name, max_root_bus_resources);
- return AE_OK;
- }
res = &info->res[info->res_num];
res->name = info->name;
@@ -143,7 +116,7 @@ setup_resource(struct acpi_resource *acpi_res, void *data)
dev_err(&info->bridge->dev,
"can't allocate host bridge window %pR\n", res);
} else {
- info->bus->resource[info->res_num] = res;
+ pci_bus_add_resource(info->bus, res, 0);
info->res_num++;
if (addr.translation_offset)
dev_info(&info->bridge->dev, "host bridge window %pR "
@@ -164,7 +137,9 @@ get_current_resources(struct acpi_device *device, int busnum,
struct pci_root_info info;
size_t size;
- if (!(pci_probe & PCI_USE__CRS))
+ if (pci_probe & PCI_USE__CRS)
+ pci_bus_remove_resources(bus);
+ else
dev_info(&device->dev,
"ignoring host bridge windows from ACPI; "
"boot with \"pci=use_crs\" to use them\n");
diff --git a/arch/x86/pci/bus_numa.c b/arch/x86/pci/bus_numa.c
index f939d60..12d54ff 100644
--- a/arch/x86/pci/bus_numa.c
+++ b/arch/x86/pci/bus_numa.c
@@ -36,13 +36,14 @@ void x86_pci_root_bus_res_quirks(struct pci_bus *b)
printk(KERN_DEBUG "PCI: peer root bus %02x res updated from pci conf\n",
b->number);
+ pci_bus_remove_resources(b);
info = &pci_root_info[i];
for (j = 0; j < info->res_num; j++) {
struct resource *res;
struct resource *root;
res = &info->res[j];
- b->resource[j] = res;
+ pci_bus_add_resource(b, res, 0);
if (res->flags & IORESOURCE_IO)
root = &ioport_resource;
else
diff --git a/arch/x86/pci/bus_numa.h b/arch/x86/pci/bus_numa.h
index adbc23f..731b64e 100644
--- a/arch/x86/pci/bus_numa.h
+++ b/arch/x86/pci/bus_numa.h
@@ -2,8 +2,7 @@
/*
* sub bus (transparent) will use entres from 3 to store extra from
- * root, so need to make sure we have enough slot there, Should we
- * increase PCI_BUS_NUM_RESOURCES?
+ * root, so need to make sure we have enough slot there.
*/
#define RES_NUM 16
struct pci_root_info {
diff --git a/drivers/pci/bus.c b/drivers/pci/bus.c
index e75d219..712250f 100644
--- a/drivers/pci/bus.c
+++ b/drivers/pci/bus.c
@@ -17,6 +17,52 @@
#include "pci.h"
+void pci_bus_add_resource(struct pci_bus *bus, struct resource *res,
+ unsigned int flags)
+{
+ struct pci_bus_resource *bus_res;
+
+ bus_res = kzalloc(sizeof(struct pci_bus_resource), GFP_KERNEL);
+ if (!bus_res) {
+ dev_err(&bus->dev, "can't add %pR resource\n", res);
+ return;
+ }
+
+ bus_res->res = res;
+ bus_res->flags = flags;
+ list_add_tail(&bus_res->list, &bus->resources);
+}
+
+struct resource *pci_bus_resource_n(const struct pci_bus *bus, int n)
+{
+ struct pci_bus_resource *bus_res;
+
+ if (n < PCI_BRIDGE_RESOURCE_NUM)
+ return bus->resource[n];
+
+ n -= PCI_BRIDGE_RESOURCE_NUM;
+ list_for_each_entry(bus_res, &bus->resources, list) {
+ if (n-- == 0)
+ return bus_res->res;
+ }
+ return NULL;
+}
+EXPORT_SYMBOL_GPL(pci_bus_resource_n);
+
+void pci_bus_remove_resources(struct pci_bus *bus)
+{
+ struct pci_bus_resource *bus_res, *tmp;
+ int i;
+
+ for (i = 0; i < PCI_BRIDGE_RESOURCE_NUM; i++)
+ bus->resource[i] = 0;
+
+ list_for_each_entry_safe(bus_res, tmp, &bus->resources, list) {
+ list_del(&bus_res->list);
+ kfree(bus_res);
+ }
+}
+
/**
* pci_bus_alloc_resource - allocate a resource from a parent bus
* @bus: PCI bus
diff --git a/drivers/pci/probe.c b/drivers/pci/probe.c
index 70c4ed2..270d069 100644
--- a/drivers/pci/probe.c
+++ b/drivers/pci/probe.c
@@ -89,6 +89,7 @@ static void release_pcibus_dev(struct device *dev)
if (pci_bus->bridge)
put_device(pci_bus->bridge);
+ pci_bus_remove_resources(pci_bus);
kfree(pci_bus);
}
@@ -394,6 +395,7 @@ static void __devinit pci_read_bridge_mmio_pref(struct pci_bus *child)
void __devinit pci_read_bridge_bases(struct pci_bus *child)
{
struct pci_dev *dev = child->self;
+ struct resource *res;
int i;
if (pci_is_root_bus(child)) /* It's a host bus, nothing to read */
@@ -403,17 +405,23 @@ void __devinit pci_read_bridge_bases(struct pci_bus *child)
child->secondary, child->subordinate,
dev->transparent ? " (subtractive decode)" : "");
+ pci_bus_remove_resources(child);
+ for (i = 0; i < PCI_BRIDGE_RESOURCE_NUM; i++)
+ child->resource[i] = &dev->resource[PCI_BRIDGE_RESOURCES+i];
+
pci_read_bridge_io(child);
pci_read_bridge_mmio(child);
pci_read_bridge_mmio_pref(child);
if (dev->transparent) {
- for (i = 3; i < PCI_BUS_NUM_RESOURCES; i++) {
- child->resource[i] = child->parent->resource[i - 3];
- if (child->resource[i])
+ pci_bus_for_each_resource(child->parent, res, i) {
+ if (res) {
+ pci_bus_add_resource(child, res,
+ PCI_SUBTRACTIVE_DECODE);
dev_printk(KERN_DEBUG, &dev->dev,
" bridge window %pR (subtractive decode)\n",
- child->resource[i]);
+ res);
+ }
}
}
}
@@ -428,6 +436,7 @@ static struct pci_bus * pci_alloc_bus(void)
INIT_LIST_HEAD(&b->children);
INIT_LIST_HEAD(&b->devices);
INIT_LIST_HEAD(&b->slots);
+ INIT_LIST_HEAD(&b->resources);
b->max_bus_speed = PCI_SPEED_UNKNOWN;
b->cur_bus_speed = PCI_SPEED_UNKNOWN;
}
diff --git a/include/linux/pci.h b/include/linux/pci.h
index 8891d3b..2af5eab 100644
--- a/include/linux/pci.h
+++ b/include/linux/pci.h
@@ -363,9 +363,26 @@ static inline void pci_add_saved_cap(struct pci_dev *pci_dev,
hlist_add_head(&new_cap->next, &pci_dev->saved_cap_space);
}
-#ifndef PCI_BUS_NUM_RESOURCES
-#define PCI_BUS_NUM_RESOURCES 16
-#endif
+/*
+ * The first PCI_BRIDGE_RESOURCE_NUM PCI bus resources (those that correspond
+ * to P2P or CardBus bridge windows) go in a table. Additional ones (for
+ * buses below host bridges or subtractive decode bridges) go in the list.
+ * Use pci_bus_for_each_resource() to iterate through all the resources.
+ */
+
+/*
+ * PCI_SUBTRACTIVE_DECODE means the bridge forwards the window implicitly
+ * and there's no way to program the bridge with the details of the window.
+ * This does not apply to ACPI _CRS windows, even with the _DEC subtractive-
+ * decode bit set, because they are explicit and can be programmed with _SRS.
+ */
+#define PCI_SUBTRACTIVE_DECODE 0x1
+
+struct pci_bus_resource {
+ struct list_head list;
+ struct resource *res;
+ unsigned int flags;
+};
#define PCI_REGION_FLAG_MASK 0x0fU /* These bits of resource flags tell us the PCI region flags */
@@ -376,8 +393,8 @@ struct pci_bus {
struct list_head devices; /* list of devices on this bus */
struct pci_dev *self; /* bridge device as seen by parent */
struct list_head slots; /* list of slots on this bus */
- struct resource *resource[PCI_BUS_NUM_RESOURCES];
- /* address space routed to this bus */
+ struct resource *resource[PCI_BRIDGE_RESOURCE_NUM];
+ struct list_head resources; /* address space routed to this bus */
struct pci_ops *ops; /* configuration access functions */
void *sysdata; /* hook for sys-specific extension */
@@ -828,8 +845,14 @@ int pci_request_selected_regions_exclusive(struct pci_dev *, int, const char *);
void pci_release_selected_regions(struct pci_dev *, int);
/* drivers/pci/bus.c */
+void pci_bus_add_resource(struct pci_bus *bus, struct resource *res, unsigned int flags);
+struct resource *pci_bus_resource_n(const struct pci_bus *bus, int n);
+void pci_bus_remove_resources(struct pci_bus *bus);
+
#define pci_bus_for_each_resource(bus, res, i) \
- for (i = 0; res = bus->resource[i], i < PCI_BUS_NUM_RESOURCES; i++)
+ for (i = 0; \
+ (res = pci_bus_resource_n(bus, i)) || i < PCI_BRIDGE_RESOURCE_NUM; \
+ i++)
int __must_check pci_bus_alloc_resource(struct pci_bus *bus,
struct resource *res, resource_size_t size,
^ permalink raw reply related
* Re: [PATCH] powerpc: Set a smaller value for RECLAIM_DISTANCE to enable zone reclaim
From: Mel Gorman @ 2010-02-23 16:23 UTC (permalink / raw)
To: Anton Blanchard; +Cc: cl, linuxppc-dev
In-Reply-To: <20100223015551.GG31681@kryten>
On Tue, Feb 23, 2010 at 12:55:51PM +1100, Anton Blanchard wrote:
>
> Hi Mel,
>
I'm afraid I'm on vacation at the moment. This mail is costing me shots with
penaltys every minute it's open. It'll be early next week before I can look
at this closely.
Sorry.
> > You're pretty much on the button here. Only one thread at a time enters
> > zone_reclaim. The others back off and try the next zone in the zonelist
> > instead. I'm not sure what the original intention was but most likely it
> > was to prevent too many parallel reclaimers in the same zone potentially
> > dumping out way more data than necessary.
> >
> > > I'm not sure if there is an easy way to fix this without penalising other
> > > workloads though.
> > >
> >
> > You could experiment with waiting on the bit if the GFP flags allowi it? The
> > expectation would be that the reclaim operation does not take long. Wait
> > on the bit, if you are making the forward progress, recheck the
> > watermarks before continueing.
>
> Thanks to you and Christoph for some suggestions to try. Attached is a
> chart showing the results of the following tests:
>
>
> baseline.txt
> The current ppc64 default of zone_reclaim_mode = 0. As expected we see
> no change in remote node memory usage even after 10 iterations.
>
> zone_reclaim_mode.txt
> Now we set zone_reclaim_mode = 1. On each iteration we continue to improve,
> but even after 10 runs of stream we have > 10% remote node memory usage.
>
> reclaim_4096_pages.txt
> Instead of reclaiming 32 pages at a time, we try for a much larger batch
> of 4096. The slope is much steeper but it still takes around 6 iterations
> to get almost all local node memory.
>
> wait_on_busy_flag.txt
> Here we busy wait if the ZONE_RECLAIM_LOCKED flag is set. As you suggest
> we would need to check the GFP flags etc, but so far it looks the most
> promising. We only get a few percent of remote node memory on the first
> iteration and get all local node by the second.
>
>
> Perhaps a combination of larger batch size and waiting on the busy
> flag is the way to go?
>
> Anton
> --- mm/vmscan.c~ 2010-02-21 23:47:14.000000000 -0600
> +++ mm/vmscan.c 2010-02-22 03:22:01.000000000 -0600
> @@ -2534,7 +2534,7 @@
> .may_unmap = !!(zone_reclaim_mode & RECLAIM_SWAP),
> .may_swap = 1,
> .nr_to_reclaim = max_t(unsigned long, nr_pages,
> - SWAP_CLUSTER_MAX),
> + 4096),
> .gfp_mask = gfp_mask,
> .swappiness = vm_swappiness,
> .order = order,
> --- mm/vmscan.c~ 2010-02-21 23:47:14.000000000 -0600
> +++ mm/vmscan.c 2010-02-21 23:47:31.000000000 -0600
> @@ -2634,8 +2634,8 @@
> if (node_state(node_id, N_CPU) && node_id != numa_node_id())
> return ZONE_RECLAIM_NOSCAN;
>
> - if (zone_test_and_set_flag(zone, ZONE_RECLAIM_LOCKED))
> - return ZONE_RECLAIM_NOSCAN;
> + while (zone_test_and_set_flag(zone, ZONE_RECLAIM_LOCKED))
> + cpu_relax();
>
> ret = __zone_reclaim(zone, gfp_mask, order);
> zone_clear_flag(zone, ZONE_RECLAIM_LOCKED);
--
Mel Gorman
Part-time Phd Student Linux Technology Center
University of Limerick IBM Dublin Software Lab
^ permalink raw reply
* Re: [PATCHv4 2/2] powerpc: implement arch_scale_smt_power for Power7
From: Peter Zijlstra @ 2010-02-23 16:24 UTC (permalink / raw)
To: Michael Neuling; +Cc: Ingo Molnar, linuxppc-dev, linux-kernel, ego
In-Reply-To: <23662.1266905307@neuling.org>
On Tue, 2010-02-23 at 17:08 +1100, Michael Neuling wrote:
> I have some comments on the code inline but...
>
> So when I run this, I don't get processes pulled down to the lower
> threads. A simple test case of running 1 CPU intensive process at
> SCHED_OTHER on a machine with 2 way SMT system (a POWER6 but enabling
> SD_ASYM_PACKING). The single processes doesn't move to lower threads as
> I'd hope.
>
> Also, are you sure you want to put this in generic code? It seem to be
> quite POWER7 specific functionality, so would be logically better in
> arch/powerpc. I guess some other arch *might* need it, but seems
> unlikely.
Well, there are no arch hooks in the load-balancing (aside from the
recent cpu_power stuff, and that really is the wrong thing to poke at
for this), and I did hear some other people express interest in such a
constraint.
Also, load-balancing is complex enough as it is, so I prefer to keep
everything in the generic code where possible, clearly things like
sched_domain creation need arch topology bits, and the arch_scale*
things require other arch information like cpu frequency.
> > @@ -2493,6 +2494,28 @@ static inline void update_sg_lb_stats(st
> > DIV_ROUND_CLOSEST(group->cpu_power, SCHED_LOAD_SCALE);
> > }
> >
> > +static int update_sd_pick_busiest(struct sched_domain *sd,
> > + struct sd_lb_stats *sds,
> > + struct sched_group *sg,
> > + struct sg_lb_stats *sgs)
> > +{
> > + if (sgs->sum_nr_running > sgs->group_capacity)
> > + return 1;
> > +
> > + if (sgs->group_imb)
> > + return 1;
> > +
> > + if ((sd->flags & SD_ASYM_PACKING) && sgs->sum_nr_running) {
>
> If we are asymetric packing...
>
>
> > + if (!sds->busiest)
> > + return 1;
>
> This just seems to be a null pointer check.
>
> From the tracing I've done, this is always true (always NULL) at this
> point so we return here.
Right, so we need to have a busiest group to take a task from, if there
is no busiest yet, take this group.
And in your scenario, with there being only a single task, we'd only hit
this once at most, so yes it makes sense this is always NULL.
> > +
> > + if (group_first_cpu(sds->busiest) < group_first_cpu(sg))
> > + return 1;
>
> I'm a bit lost as to what this is for. Any clues you could provide
> would be appreciated. :-)
>
> Is the first cpu in this domain's busiest group before the first cpu in
> this group. If, so pick this as the busiest?
>
> Should this be the other way around if we want to pack the busiest to
> the first cpu? Mark it as the busiest if it's after (not before).
>
> Is group_first_cpu guaranteed to give us the first physical cpu (ie.
> thread 0 in our case) or are these virtualised at this point?
>
> I'm not seeing this hit anyway due to the null pointer check above.
So this says, if all things being equal, and we already have a busiest,
but this candidate (sg) is higher than the current (busiest) take this
one.
The idea is to move the highest SMT task down.
> > @@ -2562,6 +2585,38 @@ static inline void update_sd_lb_stats(st
> > } while (group != sd->groups);
> > }
> >
> > +int __weak sd_asym_packing_arch(void)
> > +{
> > + return 0;
> > +}
arch_sd_asym_packing() is what you used in topology.h
> > +static int check_asym_packing(struct sched_domain *sd,
> > + struct sd_lb_stats *sds,
> > + unsigned long *imbalance)
> > +{
> > + int i, cpu, busiest_cpu;
> > +
> > + if (!(sd->flags & SD_ASYM_PACKING))
> > + return 0;
> > +
> > + if (!sds->busiest)
> > + return 0;
> > +
> > + i = 0;
> > + busiest_cpu = group_first_cpu(sds->busiest);
> > + for_each_cpu(cpu, sched_domain_span(sd)) {
> > + i++;
> > + if (cpu == busiest_cpu)
> > + break;
> > + }
> > +
> > + if (sds->total_nr_running > i)
> > + return 0;
>
> This seems to be the core of the packing logic.
>
> We make sure the busiest_cpu is not past total_nr_running. If it is we
> mark as imbalanced. Correct?
>
> It seems if a non zero thread/group had a pile of processes running on
> it and a lower thread had much less, this wouldn't fire, but I'm
> guessing normal load balancing would kick in that case to fix the
> imbalance.
>
> Any corrections to my ramblings appreciated :-)
Right, so we're concerned the scenario where there's less tasks than SMT
siblings, if there's more they should all be running and the regular
load-balancer will deal with it.
If there's less the group will normally be balanced and we fall out and
end up in check_asym_packing().
So what I tried doing with that loop is detect if there's a hole in the
packing before busiest. Now that I think about it, what we need to check
is if this_cpu (the removed cpu argument) is idle and less than busiest.
So something like:
static int check_asym_pacing(struct sched_domain *sd,
struct sd_lb_stats *sds,
int this_cpu, unsigned long *imbalance)
{
int busiest_cpu;
if (!(sd->flags & SD_ASYM_PACKING))
return 0;
if (!sds->busiest)
return 0;
busiest_cpu = group_first_cpu(sds->busiest);
if (cpu_rq(this_cpu)->nr_running || this_cpu > busiest_cpu)
return 0;
*imbalance = (sds->max_load * sds->busiest->cpu_power) /
SCHED_LOAD_SCALE;
return 1;
}
Does that make sense?
I still see two problems with this though,.. regular load-balancing only
balances on the first cpu of a domain (see the *balance = 0, condition
in update_sg_lb_stats()), this means that if SMT[12] are idle we'll not
pull properly. Also, nohz balancing might mess this up further.
We could maybe play some games with the balance decision in
update_sg_lb_stats() for SD_ASYM_PACKING domains and idle == CPU_IDLE,
no ideas yet on nohz though.
^ permalink raw reply
* Re: [PATCHv4 2/2] powerpc: implement arch_scale_smt_power for Power7
From: Peter Zijlstra @ 2010-02-23 16:30 UTC (permalink / raw)
To: Michael Neuling; +Cc: Ingo Molnar, linuxppc-dev, linux-kernel, ego
In-Reply-To: <1266942281.11845.521.camel@laptop>
On Tue, 2010-02-23 at 17:24 +0100, Peter Zijlstra wrote:
>
> busiest_cpu = group_first_cpu(sds->busiest);
> if (cpu_rq(this_cpu)->nr_running || this_cpu > busiest_cpu)
> return 0;
Hmm, we could change the bit in find_busiest_group() to:
if (idle == CPU_IDLE && check_asym_packing())
and skip the nr_running test..
^ permalink raw reply
* [PATCH v5 0/5] PCI: try "pci=use_crs" again
From: Bjorn Helgaas @ 2010-02-23 17:24 UTC (permalink / raw)
To: Jesse Barnes
Cc: Matthew Garrett, Tony Luck, linuxppc-dev, linux-pci, Peter Haight,
Gary Hade, linux-kernel, linux-acpi, linux-am33-list,
Linus Torvalds, Ingo Molnar, Yinghai Lu, Larry Finger
Historically, Linux has assumed a single PCI host bridge, with that bridge
claiming all the address space left after RAM and legacy devices are taken out.
If the system contains multiple host bridges, we can no longer operate under
that assumption. We have to know what parts of the address space are claimed
by each bridge so that when we assign resources to a PCI device, we take them
from a range claimed by the upstream host bridge.
On x86 and ia64, we use ACPI to enumerate all the PCI host bridges in the
system, and part of the host bridge description is the "_CRS" (current resource
settings" property, which lists the address space used by the bridge. On x86,
we currently ignore most of the _CRS information. This patch series changes
this, so we will use _CRS to learn about the host bridge windows.
Since most x86 machines with multiple host bridges are relatively new, this
series only turns this on for machines with BIOS dates of 2008 or newer.
Changes from v4 to v5:
- Add EXPORT_SYMBOL_GPL(pci_bus_resource_n) for module builds.
Changes from v3 to v4:
- Keep the bus->resource[] table to avoid having to change all arches
at once, but reduce it to size 4 (minimum required for P2P & CardBus
bridges).
- Store additional host bridge or subtractive decode windows in a list.
- Add pci_bus_for_each_resource() iterator to hide PCI_BUS_NUM_RESOURCES
and the table/list split. This reduces the size of the main patch.
- Remove tidy-up patches completely; they feel like distractions to this
series.
Changes from v2 to v3:
- Fix alpha, powerpc, and mn10300 references to pci_bus resource table
- Turn on "pci=use_crs" for 2008 and newer, not 2010 and newer
- Remove Toshiba A355 quirk (BIOS date is 2009, so automatically included)
- Remove IBM x3850 and x3950 quirks (BIOS dates in 2008, so automatically
included)
- Leave IBM x3800 quirk (pre-2008 BIOS)
- Use "bool" for pci_use_crs (new to me, but I see akpm suggesting it)
- Reorder so the important patches are first
Changes from v1 to v2:
- Rebase to be6e9f7853e
- Add patch to clean up "disabled window" printk
- Add bugzilla reference comment in use_crs DMI quirk
---
Bjorn Helgaas (5):
PCI: split up pci_read_bridge_bases()
PCI: read bridge windows before filling in subtractive decode resources
PCI: add pci_bus_for_each_resource(), remove direct bus->resource[] refs
PCI: augment bus resource table with a list
x86/PCI: use host bridge _CRS info by default on 2008 and newer machines
Documentation/kernel-parameters.txt | 8 ++-
arch/ia64/include/asm/acpi.h | 1
arch/ia64/pci/pci.c | 17 ++-----
arch/mn10300/unit-asb2305/pci.c | 6 +-
arch/powerpc/kernel/pci-common.c | 11 ++---
arch/powerpc/platforms/fsl_uli1575.c | 12 ++---
arch/x86/include/asm/pci_x86.h | 1
arch/x86/pci/acpi.c | 82 ++++++++++++++++++++--------------
arch/x86/pci/bus_numa.c | 3 +
arch/x86/pci/bus_numa.h | 3 -
arch/x86/pci/common.c | 3 +
drivers/acpi/pci_root.c | 1
drivers/pci/bus.c | 49 +++++++++++++++++++-
drivers/pci/hotplug/shpchp_sysfs.c | 9 +---
drivers/pci/pci.c | 5 +-
drivers/pci/probe.c | 68 ++++++++++++++++++++++------
drivers/pci/setup-bus.c | 10 ++--
drivers/pcmcia/rsrc_nonstatic.c | 3 -
drivers/pcmcia/yenta_socket.c | 5 +-
include/acpi/acpi_drivers.h | 1
include/linux/pci.h | 36 +++++++++++++--
21 files changed, 226 insertions(+), 108 deletions(-)
^ permalink raw reply
* [PATCH v5 1/5] PCI: split up pci_read_bridge_bases()
From: Bjorn Helgaas @ 2010-02-23 17:24 UTC (permalink / raw)
To: Jesse Barnes
Cc: Matthew Garrett, Tony Luck, linuxppc-dev, linux-pci, Peter Haight,
Gary Hade, linux-kernel, linux-acpi, linux-am33-list,
Linus Torvalds, Ingo Molnar, Yinghai Lu, Larry Finger
In-Reply-To: <20100223172355.17242.55927.stgit@bob.kio>
No functional change; this breaks up pci_read_bridge_bases() into separate
pieces for the I/O, memory, and prefetchable memory windows, similar to how
Yinghai recently split up pci_setup_bridge() in 68e84ff3bdc.
Signed-off-by: Bjorn Helgaas <bjorn.helgaas@hp.com>
---
drivers/pci/probe.c | 54 +++++++++++++++++++++++++++++++++++++--------------
1 files changed, 39 insertions(+), 15 deletions(-)
diff --git a/drivers/pci/probe.c b/drivers/pci/probe.c
index d300943..4b47b4b 100644
--- a/drivers/pci/probe.c
+++ b/drivers/pci/probe.c
@@ -281,26 +281,12 @@ static void pci_read_bases(struct pci_dev *dev, unsigned int howmany, int rom)
}
}
-void __devinit pci_read_bridge_bases(struct pci_bus *child)
+static void __devinit pci_read_bridge_io(struct pci_bus *child)
{
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 (pci_is_root_bus(child)) /* It's a host bus, nothing to read */
- return;
-
- dev_info(&dev->dev, "PCI bridge to [bus %02x-%02x]%s\n",
- child->secondary, child->subordinate,
- dev->transparent ? " (subtractive decode)": "");
-
- if (dev->transparent) {
- for(i = 3; i < PCI_BUS_NUM_RESOURCES; i++)
- child->resource[i] = child->parent->resource[i - 3];
- }
res = child->resource[0];
pci_read_config_byte(dev, PCI_IO_BASE, &io_base_lo);
@@ -328,6 +314,14 @@ void __devinit pci_read_bridge_bases(struct pci_bus *child)
" bridge window [io %04lx - %04lx] reg reading\n",
base, limit);
}
+}
+
+static void __devinit pci_read_bridge_mmio(struct pci_bus *child)
+{
+ struct pci_dev *dev = child->self;
+ u16 mem_base_lo, mem_limit_lo;
+ unsigned long base, limit;
+ struct resource *res;
res = child->resource[1];
pci_read_config_word(dev, PCI_MEMORY_BASE, &mem_base_lo);
@@ -344,6 +338,14 @@ void __devinit pci_read_bridge_bases(struct pci_bus *child)
" bridge window [mem 0x%08lx - 0x%08lx] reg reading\n",
base, limit + 0xfffff);
}
+}
+
+static void __devinit pci_read_bridge_mmio_pref(struct pci_bus *child)
+{
+ struct pci_dev *dev = child->self;
+ u16 mem_base_lo, mem_limit_lo;
+ unsigned long base, limit;
+ struct resource *res;
res = child->resource[2];
pci_read_config_word(dev, PCI_PREF_MEMORY_BASE, &mem_base_lo);
@@ -389,6 +391,28 @@ void __devinit pci_read_bridge_bases(struct pci_bus *child)
}
}
+void __devinit pci_read_bridge_bases(struct pci_bus *child)
+{
+ struct pci_dev *dev = child->self;
+ int i;
+
+ if (pci_is_root_bus(child)) /* It's a host bus, nothing to read */
+ return;
+
+ dev_info(&dev->dev, "PCI bridge to [bus %02x-%02x]%s\n",
+ child->secondary, child->subordinate,
+ dev->transparent ? " (subtractive decode)" : "");
+
+ if (dev->transparent) {
+ for (i = 3; i < PCI_BUS_NUM_RESOURCES; i++)
+ child->resource[i] = child->parent->resource[i - 3];
+ }
+
+ pci_read_bridge_io(child);
+ pci_read_bridge_mmio(child);
+ pci_read_bridge_mmio_pref(child);
+}
+
static struct pci_bus * pci_alloc_bus(void)
{
struct pci_bus *b;
^ permalink raw reply related
* [PATCH v5 2/5] PCI: read bridge windows before filling in subtractive decode resources
From: Bjorn Helgaas @ 2010-02-23 17:24 UTC (permalink / raw)
To: Jesse Barnes
Cc: Matthew Garrett, Tony Luck, linuxppc-dev, linux-pci, Peter Haight,
Gary Hade, linux-kernel, linux-acpi, linux-am33-list,
Linus Torvalds, Ingo Molnar, Yinghai Lu, Larry Finger
In-Reply-To: <20100223172355.17242.55927.stgit@bob.kio>
No functional change; this fills in the bus subtractive decode resources
after reading the bridge window information rather than before. Also,
print out the subtractive decode resources as we already do for the
positive decode windows.
Signed-off-by: Bjorn Helgaas <bjorn.helgaas@hp.com>
---
drivers/pci/probe.c | 15 ++++++++++-----
1 files changed, 10 insertions(+), 5 deletions(-)
diff --git a/drivers/pci/probe.c b/drivers/pci/probe.c
index 4b47b4b..70c4ed2 100644
--- a/drivers/pci/probe.c
+++ b/drivers/pci/probe.c
@@ -403,14 +403,19 @@ void __devinit pci_read_bridge_bases(struct pci_bus *child)
child->secondary, child->subordinate,
dev->transparent ? " (subtractive decode)" : "");
- if (dev->transparent) {
- for (i = 3; i < PCI_BUS_NUM_RESOURCES; i++)
- child->resource[i] = child->parent->resource[i - 3];
- }
-
pci_read_bridge_io(child);
pci_read_bridge_mmio(child);
pci_read_bridge_mmio_pref(child);
+
+ if (dev->transparent) {
+ for (i = 3; i < PCI_BUS_NUM_RESOURCES; i++) {
+ child->resource[i] = child->parent->resource[i - 3];
+ if (child->resource[i])
+ dev_printk(KERN_DEBUG, &dev->dev,
+ " bridge window %pR (subtractive decode)\n",
+ child->resource[i]);
+ }
+ }
}
static struct pci_bus * pci_alloc_bus(void)
^ permalink raw reply related
* [PATCH v5 3/5] PCI: add pci_bus_for_each_resource(), remove direct bus->resource[] refs
From: Bjorn Helgaas @ 2010-02-23 17:24 UTC (permalink / raw)
To: Jesse Barnes
Cc: Matthew Garrett, Tony Luck, linuxppc-dev, linux-pci, Peter Haight,
Gary Hade, linux-kernel, linux-acpi, linux-am33-list,
Linus Torvalds, Ingo Molnar, Yinghai Lu, Larry Finger
In-Reply-To: <20100223172355.17242.55927.stgit@bob.kio>
No functional change; this converts loops that iterate from 0 to
PCI_BUS_NUM_RESOURCES through pci_bus resource[] table to use the
pci_bus_for_each_resource() iterator instead.
This doesn't change the way resources are stored; it merely removes
dependencies on the fact that they're in a table.
Signed-off-by: Bjorn Helgaas <bjorn.helgaas@hp.com>
---
arch/ia64/pci/pci.c | 5 ++---
arch/mn10300/unit-asb2305/pci.c | 6 ++----
arch/powerpc/kernel/pci-common.c | 11 ++++-------
arch/powerpc/platforms/fsl_uli1575.c | 12 ++++++------
drivers/pci/bus.c | 4 ++--
drivers/pci/hotplug/shpchp_sysfs.c | 9 +++------
drivers/pci/pci.c | 5 ++---
drivers/pci/setup-bus.c | 10 ++++------
drivers/pcmcia/rsrc_nonstatic.c | 3 +--
drivers/pcmcia/yenta_socket.c | 5 +++--
include/linux/pci.h | 3 +++
11 files changed, 32 insertions(+), 41 deletions(-)
diff --git a/arch/ia64/pci/pci.c b/arch/ia64/pci/pci.c
index 783c83b..89f957c 100644
--- a/arch/ia64/pci/pci.c
+++ b/arch/ia64/pci/pci.c
@@ -452,13 +452,12 @@ EXPORT_SYMBOL(pcibios_bus_to_resource);
static int __devinit is_valid_resource(struct pci_dev *dev, int idx)
{
unsigned int i, type_mask = IORESOURCE_IO | IORESOURCE_MEM;
- struct resource *devr = &dev->resource[idx];
+ struct resource *devr = &dev->resource[idx], *busr;
if (!dev->bus)
return 0;
- for (i=0; i<PCI_BUS_NUM_RESOURCES; i++) {
- struct resource *busr = dev->bus->resource[i];
+ pci_bus_for_each_resource(dev->bus, busr, i) {
if (!busr || ((busr->flags ^ devr->flags) & type_mask))
continue;
if ((devr->start) && (devr->start >= busr->start) &&
diff --git a/arch/mn10300/unit-asb2305/pci.c b/arch/mn10300/unit-asb2305/pci.c
index 2cb7e75..6d8720a 100644
--- a/arch/mn10300/unit-asb2305/pci.c
+++ b/arch/mn10300/unit-asb2305/pci.c
@@ -331,12 +331,10 @@ static int __init pci_check_direct(void)
static int __devinit is_valid_resource(struct pci_dev *dev, int idx)
{
unsigned int i, type_mask = IORESOURCE_IO | IORESOURCE_MEM;
- struct resource *devr = &dev->resource[idx];
+ struct resource *devr = &dev->resource[idx], *busr;
if (dev->bus) {
- for (i = 0; i < PCI_BUS_NUM_RESOURCES; i++) {
- struct resource *busr = dev->bus->resource[i];
-
+ pci_bus_for_each_resource(dev->bus, busr, i) {
if (!busr || (busr->flags ^ devr->flags) & type_mask)
continue;
diff --git a/arch/powerpc/kernel/pci-common.c b/arch/powerpc/kernel/pci-common.c
index e640810..2597f95 100644
--- a/arch/powerpc/kernel/pci-common.c
+++ b/arch/powerpc/kernel/pci-common.c
@@ -1047,10 +1047,8 @@ static void __devinit pcibios_fixup_bridge(struct pci_bus *bus)
struct pci_dev *dev = bus->self;
- for (i = 0; i < PCI_BUS_NUM_RESOURCES; ++i) {
- if ((res = bus->resource[i]) == NULL)
- continue;
- if (!res->flags)
+ pci_bus_for_each_resource(bus, res, i) {
+ if (!res || !res->flags)
continue;
if (i >= 3 && bus->self->transparent)
continue;
@@ -1277,9 +1275,8 @@ void pcibios_allocate_bus_resources(struct pci_bus *bus)
pr_debug("PCI: Allocating bus resources for %04x:%02x...\n",
pci_domain_nr(bus), bus->number);
- for (i = 0; i < PCI_BUS_NUM_RESOURCES; ++i) {
- if ((res = bus->resource[i]) == NULL || !res->flags
- || res->start > res->end || res->parent)
+ pci_bus_for_each_resource(bus, res, i) {
+ if (!res || !res->flags || res->start > res->end || res->parent)
continue;
if (bus->parent == NULL)
pr = (res->flags & IORESOURCE_IO) ?
diff --git a/arch/powerpc/platforms/fsl_uli1575.c b/arch/powerpc/platforms/fsl_uli1575.c
index fd23a1d..8b0c208 100644
--- a/arch/powerpc/platforms/fsl_uli1575.c
+++ b/arch/powerpc/platforms/fsl_uli1575.c
@@ -222,6 +222,7 @@ static void __devinit quirk_final_uli5249(struct pci_dev *dev)
int i;
u8 *dummy;
struct pci_bus *bus = dev->bus;
+ struct resource *res;
resource_size_t end = 0;
for (i = PCI_BRIDGE_RESOURCES; i < PCI_BRIDGE_RESOURCES+3; i++) {
@@ -230,13 +231,12 @@ static void __devinit quirk_final_uli5249(struct pci_dev *dev)
end = pci_resource_end(dev, i);
}
- for (i = 0; i < PCI_BUS_NUM_RESOURCES; i++) {
- if ((bus->resource[i]) &&
- (bus->resource[i]->flags & IORESOURCE_MEM)) {
- if (bus->resource[i]->end == end)
- dummy = ioremap(bus->resource[i]->start, 0x4);
+ pci_bus_for_each_resource(bus, res, i) {
+ if (res && res->flags & IORESOURCE_MEM) {
+ if (res->end == end)
+ dummy = ioremap(res->start, 0x4);
else
- dummy = ioremap(bus->resource[i]->end - 3, 0x4);
+ dummy = ioremap(res->end - 3, 0x4);
if (dummy) {
in_8(dummy);
iounmap(dummy);
diff --git a/drivers/pci/bus.c b/drivers/pci/bus.c
index a26135b..e75d219 100644
--- a/drivers/pci/bus.c
+++ b/drivers/pci/bus.c
@@ -43,6 +43,7 @@ pci_bus_alloc_resource(struct pci_bus *bus, struct resource *res,
void *alignf_data)
{
int i, ret = -ENOMEM;
+ struct resource *r;
resource_size_t max = -1;
type_mask |= IORESOURCE_IO | IORESOURCE_MEM;
@@ -51,8 +52,7 @@ pci_bus_alloc_resource(struct pci_bus *bus, struct resource *res,
if (!(res->flags & IORESOURCE_MEM_64))
max = PCIBIOS_MAX_MEM_32;
- for (i = 0; i < PCI_BUS_NUM_RESOURCES; i++) {
- struct resource *r = bus->resource[i];
+ pci_bus_for_each_resource(bus, r, i) {
if (!r)
continue;
diff --git a/drivers/pci/hotplug/shpchp_sysfs.c b/drivers/pci/hotplug/shpchp_sysfs.c
index 29fa9d2..071b7dc 100644
--- a/drivers/pci/hotplug/shpchp_sysfs.c
+++ b/drivers/pci/hotplug/shpchp_sysfs.c
@@ -47,8 +47,7 @@ static ssize_t show_ctrl (struct device *dev, struct device_attribute *attr, cha
bus = pdev->subordinate;
out += sprintf(buf, "Free resources: memory\n");
- for (index = 0; index < PCI_BUS_NUM_RESOURCES; index++) {
- res = bus->resource[index];
+ pci_bus_for_each_resource(bus, res, index) {
if (res && (res->flags & IORESOURCE_MEM) &&
!(res->flags & IORESOURCE_PREFETCH)) {
out += sprintf(out, "start = %8.8llx, "
@@ -58,8 +57,7 @@ static ssize_t show_ctrl (struct device *dev, struct device_attribute *attr, cha
}
}
out += sprintf(out, "Free resources: prefetchable memory\n");
- for (index = 0; index < PCI_BUS_NUM_RESOURCES; index++) {
- res = bus->resource[index];
+ pci_bus_for_each_resource(bus, res, index) {
if (res && (res->flags & IORESOURCE_MEM) &&
(res->flags & IORESOURCE_PREFETCH)) {
out += sprintf(out, "start = %8.8llx, "
@@ -69,8 +67,7 @@ static ssize_t show_ctrl (struct device *dev, struct device_attribute *attr, cha
}
}
out += sprintf(out, "Free resources: IO\n");
- for (index = 0; index < PCI_BUS_NUM_RESOURCES; index++) {
- res = bus->resource[index];
+ pci_bus_for_each_resource(bus, res, index) {
if (res && (res->flags & IORESOURCE_IO)) {
out += sprintf(out, "start = %8.8llx, "
"length = %8.8llx\n",
diff --git a/drivers/pci/pci.c b/drivers/pci/pci.c
index d62a5de..f4a2738 100644
--- a/drivers/pci/pci.c
+++ b/drivers/pci/pci.c
@@ -386,10 +386,9 @@ pci_find_parent_resource(const struct pci_dev *dev, struct resource *res)
{
const struct pci_bus *bus = dev->bus;
int i;
- struct resource *best = NULL;
+ struct resource *best = NULL, *r;
- for(i = 0; i < PCI_BUS_NUM_RESOURCES; i++) {
- struct resource *r = bus->resource[i];
+ pci_bus_for_each_resource(bus, r, i) {
if (!r)
continue;
if (res->start && !(res->start >= r->start && res->end <= r->end))
diff --git a/drivers/pci/setup-bus.c b/drivers/pci/setup-bus.c
index 743ed8c..bf32f07 100644
--- a/drivers/pci/setup-bus.c
+++ b/drivers/pci/setup-bus.c
@@ -387,8 +387,7 @@ static struct resource *find_free_bus_resource(struct pci_bus *bus, unsigned lon
unsigned long type_mask = IORESOURCE_IO | IORESOURCE_MEM |
IORESOURCE_PREFETCH;
- for (i = 0; i < PCI_BUS_NUM_RESOURCES; i++) {
- r = bus->resource[i];
+ pci_bus_for_each_resource(bus, r, i) {
if (r == &ioport_resource || r == &iomem_resource)
continue;
if (r && (r->flags & type_mask) == type && !r->parent)
@@ -803,11 +802,10 @@ static void __ref pci_bus_release_bridge_resources(struct pci_bus *bus,
static void pci_bus_dump_res(struct pci_bus *bus)
{
- int i;
-
- for (i = 0; i < PCI_BUS_NUM_RESOURCES; i++) {
- struct resource *res = bus->resource[i];
+ struct resource *res;
+ int i;
+ pci_bus_for_each_resource(bus, res, i) {
if (!res || !res->end || !res->flags)
continue;
diff --git a/drivers/pcmcia/rsrc_nonstatic.c b/drivers/pcmcia/rsrc_nonstatic.c
index 45d75dc..c67638f 100644
--- a/drivers/pcmcia/rsrc_nonstatic.c
+++ b/drivers/pcmcia/rsrc_nonstatic.c
@@ -803,8 +803,7 @@ static int nonstatic_autoadd_resources(struct pcmcia_socket *s)
return -EINVAL;
#endif
- for (i = 0; i < PCI_BUS_NUM_RESOURCES; i++) {
- res = s->cb_dev->bus->resource[i];
+ pci_bus_for_each_resource(s->cb_dev->bus, res, i) {
if (!res)
continue;
diff --git a/drivers/pcmcia/yenta_socket.c b/drivers/pcmcia/yenta_socket.c
index e4d12ac..1f2039d 100644
--- a/drivers/pcmcia/yenta_socket.c
+++ b/drivers/pcmcia/yenta_socket.c
@@ -649,9 +649,10 @@ static int yenta_search_one_res(struct resource *root, struct resource *res,
static int yenta_search_res(struct yenta_socket *socket, struct resource *res,
u32 min)
{
+ struct resource *root;
int i;
- for (i = 0; i < PCI_BUS_NUM_RESOURCES; i++) {
- struct resource *root = socket->dev->bus->resource[i];
+
+ pci_bus_for_each_resource(socket->dev->bus, root, i) {
if (!root)
continue;
diff --git a/include/linux/pci.h b/include/linux/pci.h
index df2a12f..8891d3b 100644
--- a/include/linux/pci.h
+++ b/include/linux/pci.h
@@ -828,6 +828,9 @@ int pci_request_selected_regions_exclusive(struct pci_dev *, int, const char *);
void pci_release_selected_regions(struct pci_dev *, int);
/* drivers/pci/bus.c */
+#define pci_bus_for_each_resource(bus, res, i) \
+ for (i = 0; res = bus->resource[i], i < PCI_BUS_NUM_RESOURCES; i++)
+
int __must_check pci_bus_alloc_resource(struct pci_bus *bus,
struct resource *res, resource_size_t size,
resource_size_t align, resource_size_t min,
^ permalink raw reply related
* [PATCH v5 4/5] PCI: augment bus resource table with a list
From: Bjorn Helgaas @ 2010-02-23 17:24 UTC (permalink / raw)
To: Jesse Barnes
Cc: Matthew Garrett, Tony Luck, linuxppc-dev, linux-pci, Peter Haight,
Gary Hade, linux-kernel, linux-acpi, linux-am33-list,
Linus Torvalds, Ingo Molnar, Yinghai Lu, Larry Finger
In-Reply-To: <20100223172355.17242.55927.stgit@bob.kio>
Previously we used a table of size PCI_BUS_NUM_RESOURCES (16) for resources
forwarded to a bus by its upstream bridge. We've increased this size
several times when the table overflowed.
But there's no good limit on the number of resources because host bridges
and subtractive decode bridges can forward any number of ranges to their
secondary buses.
This patch reduces the table to only PCI_BRIDGE_RESOURCE_NUM (4) entries,
which corresponds to the number of windows a PCI-to-PCI (3) or CardBus (4)
bridge can positively decode. Any additional resources, e.g., PCI host
bridge windows or subtractively-decoded regions, are kept in a list.
I'd prefer a single list rather than this split table/list approach, but
that requires simultaneous changes to every architecture. This approach
only requires immediate changes where we set up (a) host bridges with more
than four windows and (b) subtractive-decode P2P bridges, and we can
incrementally change other architectures to use the list.
Signed-off-by: Bjorn Helgaas <bjorn.helgaas@hp.com>
---
arch/ia64/pci/pci.c | 12 +++---------
arch/x86/pci/acpi.c | 33 ++++-----------------------------
arch/x86/pci/bus_numa.c | 3 ++-
arch/x86/pci/bus_numa.h | 3 +--
drivers/pci/bus.c | 46 ++++++++++++++++++++++++++++++++++++++++++++++
drivers/pci/probe.c | 17 +++++++++++++----
include/linux/pci.h | 35 +++++++++++++++++++++++++++++------
7 files changed, 98 insertions(+), 51 deletions(-)
diff --git a/arch/ia64/pci/pci.c b/arch/ia64/pci/pci.c
index 89f957c..64aff52 100644
--- a/arch/ia64/pci/pci.c
+++ b/arch/ia64/pci/pci.c
@@ -320,9 +320,9 @@ static __devinit acpi_status add_window(struct acpi_resource *res, void *data)
static void __devinit
pcibios_setup_root_windows(struct pci_bus *bus, struct pci_controller *ctrl)
{
- int i, j;
+ int i;
- j = 0;
+ pci_bus_remove_resources(bus);
for (i = 0; i < ctrl->windows; i++) {
struct resource *res = &ctrl->window[i].resource;
/* HP's firmware has a hack to work around a Windows bug.
@@ -330,13 +330,7 @@ pcibios_setup_root_windows(struct pci_bus *bus, struct pci_controller *ctrl)
if ((res->flags & IORESOURCE_MEM) &&
(res->end - res->start < 16))
continue;
- if (j >= PCI_BUS_NUM_RESOURCES) {
- dev_warn(&bus->dev,
- "ignoring host bridge window %pR (no space)\n",
- res);
- continue;
- }
- bus->resource[j++] = res;
+ pci_bus_add_resource(bus, res, 0);
}
}
diff --git a/arch/x86/pci/acpi.c b/arch/x86/pci/acpi.c
index 959e548..a2f8cdb 100644
--- a/arch/x86/pci/acpi.c
+++ b/arch/x86/pci/acpi.c
@@ -45,20 +45,6 @@ count_resource(struct acpi_resource *acpi_res, void *data)
return AE_OK;
}
-static int
-bus_has_transparent_bridge(struct pci_bus *bus)
-{
- struct pci_dev *dev;
-
- list_for_each_entry(dev, &bus->devices, bus_list) {
- u16 class = dev->class >> 8;
-
- if (class == PCI_CLASS_BRIDGE_PCI && dev->transparent)
- return true;
- }
- return false;
-}
-
static void
align_resource(struct acpi_device *bridge, struct resource *res)
{
@@ -92,12 +78,8 @@ setup_resource(struct acpi_resource *acpi_res, void *data)
acpi_status status;
unsigned long flags;
struct resource *root;
- int max_root_bus_resources = PCI_BUS_NUM_RESOURCES;
u64 start, end;
- if (bus_has_transparent_bridge(info->bus))
- max_root_bus_resources -= 3;
-
status = resource_to_addr(acpi_res, &addr);
if (!ACPI_SUCCESS(status))
return AE_OK;
@@ -115,15 +97,6 @@ setup_resource(struct acpi_resource *acpi_res, void *data)
start = addr.minimum + addr.translation_offset;
end = start + addr.address_length - 1;
- if (info->res_num >= max_root_bus_resources) {
- if (pci_probe & PCI_USE__CRS)
- printk(KERN_WARNING "PCI: Failed to allocate "
- "0x%lx-0x%lx from %s for %s due to _CRS "
- "returning more than %d resource descriptors\n",
- (unsigned long) start, (unsigned long) end,
- root->name, info->name, max_root_bus_resources);
- return AE_OK;
- }
res = &info->res[info->res_num];
res->name = info->name;
@@ -143,7 +116,7 @@ setup_resource(struct acpi_resource *acpi_res, void *data)
dev_err(&info->bridge->dev,
"can't allocate host bridge window %pR\n", res);
} else {
- info->bus->resource[info->res_num] = res;
+ pci_bus_add_resource(info->bus, res, 0);
info->res_num++;
if (addr.translation_offset)
dev_info(&info->bridge->dev, "host bridge window %pR "
@@ -164,7 +137,9 @@ get_current_resources(struct acpi_device *device, int busnum,
struct pci_root_info info;
size_t size;
- if (!(pci_probe & PCI_USE__CRS))
+ if (pci_probe & PCI_USE__CRS)
+ pci_bus_remove_resources(bus);
+ else
dev_info(&device->dev,
"ignoring host bridge windows from ACPI; "
"boot with \"pci=use_crs\" to use them\n");
diff --git a/arch/x86/pci/bus_numa.c b/arch/x86/pci/bus_numa.c
index f939d60..12d54ff 100644
--- a/arch/x86/pci/bus_numa.c
+++ b/arch/x86/pci/bus_numa.c
@@ -36,13 +36,14 @@ void x86_pci_root_bus_res_quirks(struct pci_bus *b)
printk(KERN_DEBUG "PCI: peer root bus %02x res updated from pci conf\n",
b->number);
+ pci_bus_remove_resources(b);
info = &pci_root_info[i];
for (j = 0; j < info->res_num; j++) {
struct resource *res;
struct resource *root;
res = &info->res[j];
- b->resource[j] = res;
+ pci_bus_add_resource(b, res, 0);
if (res->flags & IORESOURCE_IO)
root = &ioport_resource;
else
diff --git a/arch/x86/pci/bus_numa.h b/arch/x86/pci/bus_numa.h
index adbc23f..731b64e 100644
--- a/arch/x86/pci/bus_numa.h
+++ b/arch/x86/pci/bus_numa.h
@@ -2,8 +2,7 @@
/*
* sub bus (transparent) will use entres from 3 to store extra from
- * root, so need to make sure we have enough slot there, Should we
- * increase PCI_BUS_NUM_RESOURCES?
+ * root, so need to make sure we have enough slot there.
*/
#define RES_NUM 16
struct pci_root_info {
diff --git a/drivers/pci/bus.c b/drivers/pci/bus.c
index e75d219..712250f 100644
--- a/drivers/pci/bus.c
+++ b/drivers/pci/bus.c
@@ -17,6 +17,52 @@
#include "pci.h"
+void pci_bus_add_resource(struct pci_bus *bus, struct resource *res,
+ unsigned int flags)
+{
+ struct pci_bus_resource *bus_res;
+
+ bus_res = kzalloc(sizeof(struct pci_bus_resource), GFP_KERNEL);
+ if (!bus_res) {
+ dev_err(&bus->dev, "can't add %pR resource\n", res);
+ return;
+ }
+
+ bus_res->res = res;
+ bus_res->flags = flags;
+ list_add_tail(&bus_res->list, &bus->resources);
+}
+
+struct resource *pci_bus_resource_n(const struct pci_bus *bus, int n)
+{
+ struct pci_bus_resource *bus_res;
+
+ if (n < PCI_BRIDGE_RESOURCE_NUM)
+ return bus->resource[n];
+
+ n -= PCI_BRIDGE_RESOURCE_NUM;
+ list_for_each_entry(bus_res, &bus->resources, list) {
+ if (n-- == 0)
+ return bus_res->res;
+ }
+ return NULL;
+}
+EXPORT_SYMBOL_GPL(pci_bus_resource_n);
+
+void pci_bus_remove_resources(struct pci_bus *bus)
+{
+ struct pci_bus_resource *bus_res, *tmp;
+ int i;
+
+ for (i = 0; i < PCI_BRIDGE_RESOURCE_NUM; i++)
+ bus->resource[i] = 0;
+
+ list_for_each_entry_safe(bus_res, tmp, &bus->resources, list) {
+ list_del(&bus_res->list);
+ kfree(bus_res);
+ }
+}
+
/**
* pci_bus_alloc_resource - allocate a resource from a parent bus
* @bus: PCI bus
diff --git a/drivers/pci/probe.c b/drivers/pci/probe.c
index 70c4ed2..270d069 100644
--- a/drivers/pci/probe.c
+++ b/drivers/pci/probe.c
@@ -89,6 +89,7 @@ static void release_pcibus_dev(struct device *dev)
if (pci_bus->bridge)
put_device(pci_bus->bridge);
+ pci_bus_remove_resources(pci_bus);
kfree(pci_bus);
}
@@ -394,6 +395,7 @@ static void __devinit pci_read_bridge_mmio_pref(struct pci_bus *child)
void __devinit pci_read_bridge_bases(struct pci_bus *child)
{
struct pci_dev *dev = child->self;
+ struct resource *res;
int i;
if (pci_is_root_bus(child)) /* It's a host bus, nothing to read */
@@ -403,17 +405,23 @@ void __devinit pci_read_bridge_bases(struct pci_bus *child)
child->secondary, child->subordinate,
dev->transparent ? " (subtractive decode)" : "");
+ pci_bus_remove_resources(child);
+ for (i = 0; i < PCI_BRIDGE_RESOURCE_NUM; i++)
+ child->resource[i] = &dev->resource[PCI_BRIDGE_RESOURCES+i];
+
pci_read_bridge_io(child);
pci_read_bridge_mmio(child);
pci_read_bridge_mmio_pref(child);
if (dev->transparent) {
- for (i = 3; i < PCI_BUS_NUM_RESOURCES; i++) {
- child->resource[i] = child->parent->resource[i - 3];
- if (child->resource[i])
+ pci_bus_for_each_resource(child->parent, res, i) {
+ if (res) {
+ pci_bus_add_resource(child, res,
+ PCI_SUBTRACTIVE_DECODE);
dev_printk(KERN_DEBUG, &dev->dev,
" bridge window %pR (subtractive decode)\n",
- child->resource[i]);
+ res);
+ }
}
}
}
@@ -428,6 +436,7 @@ static struct pci_bus * pci_alloc_bus(void)
INIT_LIST_HEAD(&b->children);
INIT_LIST_HEAD(&b->devices);
INIT_LIST_HEAD(&b->slots);
+ INIT_LIST_HEAD(&b->resources);
b->max_bus_speed = PCI_SPEED_UNKNOWN;
b->cur_bus_speed = PCI_SPEED_UNKNOWN;
}
diff --git a/include/linux/pci.h b/include/linux/pci.h
index 8891d3b..2af5eab 100644
--- a/include/linux/pci.h
+++ b/include/linux/pci.h
@@ -363,9 +363,26 @@ static inline void pci_add_saved_cap(struct pci_dev *pci_dev,
hlist_add_head(&new_cap->next, &pci_dev->saved_cap_space);
}
-#ifndef PCI_BUS_NUM_RESOURCES
-#define PCI_BUS_NUM_RESOURCES 16
-#endif
+/*
+ * The first PCI_BRIDGE_RESOURCE_NUM PCI bus resources (those that correspond
+ * to P2P or CardBus bridge windows) go in a table. Additional ones (for
+ * buses below host bridges or subtractive decode bridges) go in the list.
+ * Use pci_bus_for_each_resource() to iterate through all the resources.
+ */
+
+/*
+ * PCI_SUBTRACTIVE_DECODE means the bridge forwards the window implicitly
+ * and there's no way to program the bridge with the details of the window.
+ * This does not apply to ACPI _CRS windows, even with the _DEC subtractive-
+ * decode bit set, because they are explicit and can be programmed with _SRS.
+ */
+#define PCI_SUBTRACTIVE_DECODE 0x1
+
+struct pci_bus_resource {
+ struct list_head list;
+ struct resource *res;
+ unsigned int flags;
+};
#define PCI_REGION_FLAG_MASK 0x0fU /* These bits of resource flags tell us the PCI region flags */
@@ -376,8 +393,8 @@ struct pci_bus {
struct list_head devices; /* list of devices on this bus */
struct pci_dev *self; /* bridge device as seen by parent */
struct list_head slots; /* list of slots on this bus */
- struct resource *resource[PCI_BUS_NUM_RESOURCES];
- /* address space routed to this bus */
+ struct resource *resource[PCI_BRIDGE_RESOURCE_NUM];
+ struct list_head resources; /* address space routed to this bus */
struct pci_ops *ops; /* configuration access functions */
void *sysdata; /* hook for sys-specific extension */
@@ -828,8 +845,14 @@ int pci_request_selected_regions_exclusive(struct pci_dev *, int, const char *);
void pci_release_selected_regions(struct pci_dev *, int);
/* drivers/pci/bus.c */
+void pci_bus_add_resource(struct pci_bus *bus, struct resource *res, unsigned int flags);
+struct resource *pci_bus_resource_n(const struct pci_bus *bus, int n);
+void pci_bus_remove_resources(struct pci_bus *bus);
+
#define pci_bus_for_each_resource(bus, res, i) \
- for (i = 0; res = bus->resource[i], i < PCI_BUS_NUM_RESOURCES; i++)
+ for (i = 0; \
+ (res = pci_bus_resource_n(bus, i)) || i < PCI_BRIDGE_RESOURCE_NUM; \
+ i++)
int __must_check pci_bus_alloc_resource(struct pci_bus *bus,
struct resource *res, resource_size_t size,
^ permalink raw reply related
* [PATCH v5 5/5] x86/PCI: use host bridge _CRS info by default on 2008 and newer machines
From: Bjorn Helgaas @ 2010-02-23 17:24 UTC (permalink / raw)
To: Jesse Barnes
Cc: Matthew Garrett, Tony Luck, linuxppc-dev, linux-pci, Peter Haight,
Gary Hade, linux-kernel, linux-acpi, linux-am33-list,
Linus Torvalds, Ingo Molnar, Yinghai Lu, Larry Finger
In-Reply-To: <20100223172355.17242.55927.stgit@bob.kio>
The main benefit of using ACPI host bridge window information is that
we can do better resource allocation in systems with multiple host bridges,
e.g., http://bugzilla.kernel.org/show_bug.cgi?id=14183
Sometimes we need _CRS information even if we only have one host bridge,
e.g., https://bugs.launchpad.net/ubuntu/+source/linux/+bug/341681
Most of these systems are relatively new, so this patch turns on
"pci=use_crs" only on machines with a BIOS date of 2008 or newer.
Signed-off-by: Bjorn Helgaas <bjorn.helgaas@hp.com>
---
Documentation/kernel-parameters.txt | 8 ++++-
arch/ia64/include/asm/acpi.h | 1 +
arch/x86/include/asm/pci_x86.h | 1 +
arch/x86/pci/acpi.c | 53 +++++++++++++++++++++++++++++++----
arch/x86/pci/common.c | 3 ++
drivers/acpi/pci_root.c | 1 +
include/acpi/acpi_drivers.h | 1 +
7 files changed, 60 insertions(+), 8 deletions(-)
diff --git a/Documentation/kernel-parameters.txt b/Documentation/kernel-parameters.txt
index 01e2a98..6913e9a 100644
--- a/Documentation/kernel-parameters.txt
+++ b/Documentation/kernel-parameters.txt
@@ -1939,8 +1939,12 @@ and is between 256 and 4096 characters. It is defined in the file
IRQ routing is enabled.
noacpi [X86] Do not use ACPI for IRQ routing
or for PCI scanning.
- use_crs [X86] Use _CRS for PCI resource
- allocation.
+ use_crs [X86] Use PCI host bridge window information
+ from ACPI. On BIOSes from 2008 or later, this
+ is enabled by default. If you need to use this,
+ please report a bug.
+ nocrs [X86] Ignore PCI host bridge windows from ACPI.
+ If you need to use this, please report a bug.
routeirq Do IRQ routing for all PCI devices.
This is normally done in pci_enable_device(),
so this option is a temporary workaround
diff --git a/arch/ia64/include/asm/acpi.h b/arch/ia64/include/asm/acpi.h
index 7ae5889..7f2d7f2 100644
--- a/arch/ia64/include/asm/acpi.h
+++ b/arch/ia64/include/asm/acpi.h
@@ -97,6 +97,7 @@ ia64_acpi_release_global_lock (unsigned int *lock)
#endif
#define acpi_processor_cstate_check(x) (x) /* no idle limits on IA64 :) */
static inline void disable_acpi(void) { }
+static inline void pci_acpi_crs_quirks(void) { }
const char *acpi_get_sysname (void);
int acpi_request_vector (u32 int_type);
diff --git a/arch/x86/include/asm/pci_x86.h b/arch/x86/include/asm/pci_x86.h
index b4bf9a9..05b58cc 100644
--- a/arch/x86/include/asm/pci_x86.h
+++ b/arch/x86/include/asm/pci_x86.h
@@ -29,6 +29,7 @@
#define PCI_CHECK_ENABLE_AMD_MMCONF 0x20000
#define PCI_HAS_IO_ECS 0x40000
#define PCI_NOASSIGN_ROMS 0x80000
+#define PCI_ROOT_NO_CRS 0x100000
extern unsigned int pci_probe;
extern unsigned long pirq_table_addr;
diff --git a/arch/x86/pci/acpi.c b/arch/x86/pci/acpi.c
index a2f8cdb..5f11ff6 100644
--- a/arch/x86/pci/acpi.c
+++ b/arch/x86/pci/acpi.c
@@ -15,6 +15,51 @@ struct pci_root_info {
int busnum;
};
+static bool pci_use_crs = true;
+
+static int __init set_use_crs(const struct dmi_system_id *id)
+{
+ pci_use_crs = true;
+ return 0;
+}
+
+static const struct dmi_system_id pci_use_crs_table[] __initconst = {
+ /* http://bugzilla.kernel.org/show_bug.cgi?id=14183 */
+ {
+ .callback = set_use_crs,
+ .ident = "IBM System x3800",
+ .matches = {
+ DMI_MATCH(DMI_SYS_VENDOR, "IBM"),
+ DMI_MATCH(DMI_PRODUCT_NAME, "x3800"),
+ },
+ },
+ {}
+};
+
+void __init pci_acpi_crs_quirks(void)
+{
+ int year;
+
+ if (dmi_get_date(DMI_BIOS_DATE, &year, NULL, NULL) && year < 2008)
+ pci_use_crs = false;
+
+ dmi_check_system(pci_use_crs_table);
+
+ /*
+ * If the user specifies "pci=use_crs" or "pci=nocrs" explicitly, that
+ * takes precedence over anything we figured out above.
+ */
+ if (pci_probe & PCI_ROOT_NO_CRS)
+ pci_use_crs = false;
+ else if (pci_probe & PCI_USE__CRS)
+ pci_use_crs = true;
+
+ printk(KERN_INFO "PCI: %s host bridge windows from ACPI; "
+ "if necessary, use \"pci=%s\" and report a bug\n",
+ pci_use_crs ? "Using" : "Ignoring",
+ pci_use_crs ? "nocrs" : "use_crs");
+}
+
static acpi_status
resource_to_addr(struct acpi_resource *resource,
struct acpi_resource_address64 *addr)
@@ -106,7 +151,7 @@ setup_resource(struct acpi_resource *acpi_res, void *data)
res->child = NULL;
align_resource(info->bridge, res);
- if (!(pci_probe & PCI_USE__CRS)) {
+ if (!pci_use_crs) {
dev_printk(KERN_DEBUG, &info->bridge->dev,
"host bridge window %pR (ignored)\n", res);
return AE_OK;
@@ -137,12 +182,8 @@ get_current_resources(struct acpi_device *device, int busnum,
struct pci_root_info info;
size_t size;
- if (pci_probe & PCI_USE__CRS)
+ if (pci_use_crs)
pci_bus_remove_resources(bus);
- else
- dev_info(&device->dev,
- "ignoring host bridge windows from ACPI; "
- "boot with \"pci=use_crs\" to use them\n");
info.bridge = device;
info.bus = bus;
diff --git a/arch/x86/pci/common.c b/arch/x86/pci/common.c
index d2552c6..3736176 100644
--- a/arch/x86/pci/common.c
+++ b/arch/x86/pci/common.c
@@ -520,6 +520,9 @@ char * __devinit pcibios_setup(char *str)
} else if (!strcmp(str, "use_crs")) {
pci_probe |= PCI_USE__CRS;
return NULL;
+ } else if (!strcmp(str, "nocrs")) {
+ pci_probe |= PCI_ROOT_NO_CRS;
+ return NULL;
} else if (!strcmp(str, "earlydump")) {
pci_early_dump_regs = 1;
return NULL;
diff --git a/drivers/acpi/pci_root.c b/drivers/acpi/pci_root.c
index 64ae2f2..4a46051 100644
--- a/drivers/acpi/pci_root.c
+++ b/drivers/acpi/pci_root.c
@@ -566,6 +566,7 @@ static int __init acpi_pci_root_init(void)
if (acpi_pci_disabled)
return 0;
+ pci_acpi_crs_quirks();
if (acpi_bus_register_driver(&acpi_pci_root_driver) < 0)
return -ENODEV;
diff --git a/include/acpi/acpi_drivers.h b/include/acpi/acpi_drivers.h
index f4906f6..3a4767c 100644
--- a/include/acpi/acpi_drivers.h
+++ b/include/acpi/acpi_drivers.h
@@ -104,6 +104,7 @@ int acpi_pci_bind_root(struct acpi_device *device);
struct pci_bus *pci_acpi_scan_root(struct acpi_device *device, int domain,
int bus);
+void pci_acpi_crs_quirks(void);
/* --------------------------------------------------------------------------
Processor
^ permalink raw reply related
* Re: [PATCH v5 0/5] PCI: try "pci=use_crs" again
From: Jesse Barnes @ 2010-02-23 18:06 UTC (permalink / raw)
To: Bjorn Helgaas
Cc: Matthew Garrett, Tony Luck, linuxppc-dev, linux-pci, Peter Haight,
Gary Hade, linux-kernel, linux-acpi, linux-am33-list,
Linus Torvalds, Ingo Molnar, Yinghai Lu, Larry Finger
In-Reply-To: <20100223172355.17242.55927.stgit@bob.kio>
On Tue, 23 Feb 2010 10:24:15 -0700
Bjorn Helgaas <bjorn.helgaas@hp.com> wrote:
> Historically, Linux has assumed a single PCI host bridge, with that
> bridge claiming all the address space left after RAM and legacy
> devices are taken out.
>
> If the system contains multiple host bridges, we can no longer
> operate under that assumption. We have to know what parts of the
> address space are claimed by each bridge so that when we assign
> resources to a PCI device, we take them from a range claimed by the
> upstream host bridge.
>
> On x86 and ia64, we use ACPI to enumerate all the PCI host bridges in
> the system, and part of the host bridge description is the
> "_CRS" (current resource settings" property, which lists the address
> space used by the bridge. On x86, we currently ignore most of the
> _CRS information. This patch series changes this, so we will use
> _CRS to learn about the host bridge windows.
>
> Since most x86 machines with multiple host bridges are relatively
> new, this series only turns this on for machines with BIOS dates of
> 2008 or newer.
>
> Changes from v4 to v5:
> - Add EXPORT_SYMBOL_GPL(pci_bus_resource_n) for module builds.
Thanks, it built this time. Pushed to my linux-next branch.
--
Jesse Barnes, Intel Open Source Technology Center
^ permalink raw reply
* [PATCH] powerpc/booke: Fix a couple typos in the advanced ptrace code
From: Dave Kleikamp @ 2010-02-23 19:43 UTC (permalink / raw)
To: linuxppc-dev list
In-Reply-To: <20100208215049.17930.8586.sendpatchset@norville.austin.ibm.com>
powerpc/booke: Fix a couple typos in the advanced ptrace code
Found and fixed a couple typos in the advanced ptrace patches.
(These patches are currently in benh's next tree.)
Signed-off-by: Dave Kleikamp <shaggy@linux.vnet.ibm.com>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: linuxppc-dev list <Linuxppc-dev@ozlabs.org>
---
diff --git a/arch/powerpc/include/asm/reg_booke.h
b/arch/powerpc/include/asm/reg_booke.h
index 126cb05..96c2ce8 100644
--- a/arch/powerpc/include/asm/reg_booke.h
+++ b/arch/powerpc/include/asm/reg_booke.h
@@ -421,8 +421,8 @@
/* Bit definitions related to the DBCR2. */
#define DBCR2_DAC1US 0xC0000000 /* Data Addr Cmp 1 Sup/User */
#define DBCR2_DAC1ER 0x30000000 /* Data Addr Cmp 1 Eff/Real */
-#define DBCR2_DAC2US 0x00000000 /* Data Addr Cmp 2 Sup/User */
-#define DBCR2_DAC2ER 0x00000000 /* Data Addr Cmp 2 Eff/Real */
+#define DBCR2_DAC2US 0x0C000000 /* Data Addr Cmp 2 Sup/User */
+#define DBCR2_DAC2ER 0x03000000 /* Data Addr Cmp 2 Eff/Real */
#define DBCR2_DAC12M 0x00800000 /* DAC 1-2 range enable */
#define DBCR2_DAC12MM 0x00400000 /* DAC 1-2 Mask mode*/
#define DBCR2_DAC12MX 0x00C00000 /* DAC 1-2 range eXclusive */
diff --git a/arch/powerpc/kernel/ptrace.c b/arch/powerpc/kernel/ptrace.c
index d9b0586..0efa2e3 100644
--- a/arch/powerpc/kernel/ptrace.c
+++ b/arch/powerpc/kernel/ptrace.c
@@ -1070,7 +1070,7 @@ static int del_dac(struct task_struct *child, int
slot)
child->thread.dvc1 = 0;
#endif
} else if (slot == 2) {
- if (child->thread.dac1 == 0)
+ if (child->thread.dac2 == 0)
return -ENOENT;
#ifdef CONFIG_PPC_ADV_DEBUG_DAC_RANGE
^ permalink raw reply related
* [PATCH 1/2] powerpc: Partition hibernation support
From: Brian King @ 2010-02-24 0:02 UTC (permalink / raw)
To: benh; +Cc: brking, linuxppc-dev
Enables support for HMC initiated partition hibernation. This is
a firmware assisted hibernation, since the firmware handles writing
the memory out to disk, along with other partition information,
so we just mimic suspend to ram.
Signed-off-by: Brian King <brking@linux.vnet.ibm.com>
---
arch/powerpc/Kconfig | 2
arch/powerpc/include/asm/hvcall.h | 1
arch/powerpc/include/asm/machdep.h | 1
arch/powerpc/include/asm/rtas.h | 10 +
arch/powerpc/kernel/rtas.c | 118 ++++++++++-----
arch/powerpc/platforms/pseries/Makefile | 1
arch/powerpc/platforms/pseries/hotplug-cpu.c | 6
arch/powerpc/platforms/pseries/suspend.c | 209 +++++++++++++++++++++++++++
8 files changed, 312 insertions(+), 36 deletions(-)
diff -puN /dev/null arch/powerpc/platforms/pseries/suspend.c
--- /dev/null 2009-12-15 17:58:07.000000000 -0600
+++ linux-2.6-bjking1/arch/powerpc/platforms/pseries/suspend.c 2010-02-23 16:29:25.000000000 -0600
@@ -0,0 +1,209 @@
+/*
+ * Copyright (C) 2010 Brian King IBM Corporation
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ */
+
+#include <linux/suspend.h>
+#include <asm/hvcall.h>
+#include <asm/machdep.h>
+#include <asm/mmu.h>
+#include <asm/rtas.h>
+
+static u64 stream_id;
+static struct sys_device suspend_sysdev;
+static DECLARE_COMPLETION(suspend_work);
+static struct rtas_suspend_me_data suspend_data;
+static atomic_t suspending;
+
+/**
+ * pseries_suspend_begin - First phase of hibernation
+ *
+ * Check to ensure we are in a valid state to hibernate
+ *
+ * Return value:
+ * 0 on success / other on failure
+ **/
+static int pseries_suspend_begin(suspend_state_t state)
+{
+ long vs, rc;
+ unsigned long retbuf[PLPAR_HCALL_BUFSIZE];
+
+ if (!rtas_service_present("ibm,suspend-me"))
+ return -ENOSYS;
+
+ /* Make sure the state is valid */
+ rc = plpar_hcall(H_VASI_STATE, retbuf, stream_id);
+
+ vs = retbuf[0];
+
+ if (rc) {
+ pr_err("pseries_suspend_begin: vasi_state returned %ld\n",rc);
+ return rc;
+ } else if (vs == H_VASI_ENABLED) {
+ return RTAS_NOT_SUSPENDABLE;
+ } else if (vs != H_VASI_SUSPENDING) {
+ pr_err("pseries_suspend_begin: vasi_state returned state %ld\n", vs);
+ return -EIO;
+ }
+
+ return 0;
+}
+
+/**
+ * pseries_suspend_cpu - Suspend a single CPU
+ *
+ * Makes the H_JOIN call to suspend the CPU
+ *
+ **/
+static int pseries_suspend_cpu(void)
+{
+ if (atomic_read(&suspending))
+ return rtas_suspend_cpu(&suspend_data);
+ return 0;
+}
+
+/**
+ * pseries_suspend_enter - Final phase of hibernation
+ *
+ * Return value:
+ * 0 on success / other on failure
+ **/
+static int pseries_suspend_enter(suspend_state_t state)
+{
+ int rc = rtas_suspend_last_cpu(&suspend_data);
+
+ atomic_set(&suspending, 0);
+ atomic_set(&suspend_data.done, 1);
+ return rc;
+}
+
+/**
+ * pseries_prepare_late - Suspend all other CPUs
+ *
+ * Call H_JOIN for all CPUs except for this CPU
+ *
+ * Return value:
+ * 0 on success / other on failure
+ **/
+static int pseries_prepare_late(void)
+{
+ atomic_set(&suspending, 1);
+ atomic_set(&suspend_data.working, 0);
+ atomic_set(&suspend_data.done, 0);
+ atomic_set(&suspend_data.error, 0);
+ suspend_data.token = rtas_token("ibm,suspend-me");
+ suspend_data.complete = &suspend_work;
+ INIT_COMPLETION(suspend_work);
+ return 0;
+}
+
+/**
+ * store_hibernate - Initiate partition hibernation
+ * @dev: device struct
+ * @attr: device attribute struct
+ * @buf: buffer
+ * @count: buffer size
+ *
+ * Write the stream ID received from the HMC to this file
+ * to trigger hibernating the partition
+ *
+ * Return value:
+ * number of bytes printed to buffer / other on failure
+ **/
+static ssize_t store_hibernate(struct sys_device *dev,
+ struct sysdev_attribute *attr,
+ const char *buf, size_t count)
+{
+ unsigned long long val = simple_strtoull (buf, NULL, 16);
+ int rc;
+
+ if (!capable(CAP_SYS_ADMIN))
+ return -EPERM;
+
+ stream_id = val;
+ rc = pseries_suspend_begin(PM_SUSPEND_MEM);
+ if (!rc)
+ rc = pm_suspend(PM_SUSPEND_MEM);
+
+ stream_id = 0;
+
+ if (!rc)
+ rc = count;
+ return rc;
+}
+
+static SYSDEV_ATTR(hibernate, S_IWUSR, NULL, store_hibernate);
+
+static struct sysdev_class suspend_sysdev_class = {
+ .name = "power",
+};
+
+static struct platform_suspend_ops pseries_suspend_ops = {
+ .valid = suspend_valid_only_mem,
+ .begin = pseries_suspend_begin,
+ .prepare_late = pseries_prepare_late,
+ .enter = pseries_suspend_enter,
+};
+
+/**
+ * pseries_suspend_sysfs_register - Register with sysfs
+ *
+ * Return value:
+ * 0 on success / other on failure
+ **/
+static int pseries_suspend_sysfs_register(struct sys_device *sysdev)
+{
+ int rc;
+
+ if ((rc = sysdev_class_register(&suspend_sysdev_class)))
+ return rc;
+
+ sysdev->id = 0;
+ sysdev->cls = &suspend_sysdev_class;
+
+ if ((rc = sysdev_register(sysdev)))
+ goto class_unregister;
+ if ((rc = sysdev_create_file(sysdev, &attr_hibernate)))
+ goto fail;
+
+ return 0;
+
+fail:
+ sysdev_unregister(sysdev);
+class_unregister:
+ sysdev_class_unregister(&suspend_sysdev_class);
+ return rc;
+}
+
+/**
+ * pseries_suspend_init - initcall for pSeries suspend
+ *
+ * Return value:
+ * 0 on success / other on failure
+ **/
+static int __init pseries_suspend_init(void)
+{
+ int rc;
+
+ if ((rc = pseries_suspend_sysfs_register(&suspend_sysdev)))
+ return rc;
+
+ ppc_md.suspend_disable_cpu = pseries_suspend_cpu;
+ suspend_set_ops(&pseries_suspend_ops);
+ return 0;
+}
+
+__initcall(pseries_suspend_init);
diff -puN arch/powerpc/kernel/rtas.c~powerpc_allarch_pseries_hibernation arch/powerpc/kernel/rtas.c
--- linux-2.6/arch/powerpc/kernel/rtas.c~powerpc_allarch_pseries_hibernation 2010-02-23 15:33:06.000000000 -0600
+++ linux-2.6-bjking1/arch/powerpc/kernel/rtas.c 2010-02-23 15:33:06.000000000 -0600
@@ -46,14 +46,6 @@ struct rtas_t rtas = {
};
EXPORT_SYMBOL(rtas);
-struct rtas_suspend_me_data {
- atomic_t working; /* number of cpus accessing this struct */
- atomic_t done;
- int token; /* ibm,suspend-me */
- int error;
- struct completion *complete; /* wait on this until working == 0 */
-};
-
DEFINE_SPINLOCK(rtas_data_buf_lock);
EXPORT_SYMBOL(rtas_data_buf_lock);
@@ -710,14 +702,54 @@ void rtas_os_term(char *str)
static int ibm_suspend_me_token = RTAS_UNKNOWN_SERVICE;
#ifdef CONFIG_PPC_PSERIES
-static void rtas_percpu_suspend_me(void *info)
+static int __rtas_suspend_last_cpu(struct rtas_suspend_me_data *data, int wake_when_done)
+{
+ u16 slb_size = mmu_slb_size;
+ int rc = H_MULTI_THREADS_ACTIVE;
+ int cpu;
+
+ atomic_inc(&data->working);
+
+ slb_set_size(SLB_MIN_SIZE);
+ printk(KERN_DEBUG "calling ibm,suspend-me on cpu %i\n", smp_processor_id());
+
+ while (rc == H_MULTI_THREADS_ACTIVE && !atomic_read(&data->done) &&
+ !atomic_read(&data->error))
+ rc = rtas_call(data->token, 0, 1, NULL);
+
+ if (rc || atomic_read(&data->error)) {
+ printk(KERN_DEBUG "ibm,suspend-me returned %d\n", rc);
+ slb_set_size(slb_size);
+ }
+
+ if (atomic_read(&data->error))
+ rc = atomic_read(&data->error);
+
+ atomic_set(&data->error, rc);
+
+ if (wake_when_done) {
+ atomic_set(&data->done, 1);
+
+ for_each_online_cpu(cpu)
+ plpar_hcall_norets(H_PROD, get_hard_smp_processor_id(cpu));
+ }
+
+ if (atomic_dec_return(&data->working) == 0)
+ complete(data->complete);
+
+ return rc;
+}
+
+int rtas_suspend_last_cpu(struct rtas_suspend_me_data *data)
+{
+ return __rtas_suspend_last_cpu(data, 0);
+}
+
+static int __rtas_suspend_cpu(struct rtas_suspend_me_data *data, int wake_when_done)
{
long rc = H_SUCCESS;
unsigned long msr_save;
- u16 slb_size = mmu_slb_size;
int cpu;
- struct rtas_suspend_me_data *data =
- (struct rtas_suspend_me_data *)info;
atomic_inc(&data->working);
@@ -725,7 +757,7 @@ static void rtas_percpu_suspend_me(void
msr_save = mfmsr();
mtmsr(msr_save & ~(MSR_EE));
- while (rc == H_SUCCESS && !atomic_read(&data->done))
+ while (rc == H_SUCCESS && !atomic_read(&data->done) && !atomic_read(&data->error))
rc = plpar_hcall_norets(H_JOIN);
mtmsr(msr_save);
@@ -737,33 +769,37 @@ static void rtas_percpu_suspend_me(void
/* All other cpus are in H_JOIN, this cpu does
* the suspend.
*/
- slb_set_size(SLB_MIN_SIZE);
- printk(KERN_DEBUG "calling ibm,suspend-me on cpu %i\n",
- smp_processor_id());
- data->error = rtas_call(data->token, 0, 1, NULL);
-
- if (data->error) {
- printk(KERN_DEBUG "ibm,suspend-me returned %d\n",
- data->error);
- slb_set_size(slb_size);
- }
+ return __rtas_suspend_last_cpu(data, wake_when_done);
} else {
printk(KERN_ERR "H_JOIN on cpu %i failed with rc = %ld\n",
smp_processor_id(), rc);
- data->error = rc;
+ atomic_set(&data->error, rc);
}
- atomic_set(&data->done, 1);
+ if (wake_when_done) {
+ atomic_set(&data->done, 1);
- /* This cpu did the suspend or got an error; in either case,
- * we need to prod all other other cpus out of join state.
- * Extra prods are harmless.
- */
- for_each_online_cpu(cpu)
- plpar_hcall_norets(H_PROD, get_hard_smp_processor_id(cpu));
+ /* This cpu did the suspend or got an error; in either case,
+ * we need to prod all other other cpus out of join state.
+ * Extra prods are harmless.
+ */
+ for_each_online_cpu(cpu)
+ plpar_hcall_norets(H_PROD, get_hard_smp_processor_id(cpu));
+ }
out:
if (atomic_dec_return(&data->working) == 0)
complete(data->complete);
+ return rc;
+}
+
+int rtas_suspend_cpu(struct rtas_suspend_me_data *data)
+{
+ return __rtas_suspend_cpu(data, 0);
+}
+
+static void rtas_percpu_suspend_me(void *info)
+{
+ __rtas_suspend_cpu((struct rtas_suspend_me_data *)info, 1);
}
static int rtas_ibm_suspend_me(struct rtas_args *args)
@@ -798,29 +834,41 @@ static int rtas_ibm_suspend_me(struct rt
atomic_set(&data.working, 0);
atomic_set(&data.done, 0);
+ atomic_set(&data.error, 0);
data.token = rtas_token("ibm,suspend-me");
- data.error = 0;
data.complete = &done;
/* Call function on all CPUs. One of us will make the
* rtas call
*/
if (on_each_cpu(rtas_percpu_suspend_me, &data, 0))
- data.error = -EINVAL;
+ atomic_set(&data.error, -EINVAL);
wait_for_completion(&done);
- if (data.error != 0)
+ if (atomic_read(&data.error) != 0)
printk(KERN_ERR "Error doing global join\n");
- return data.error;
+ return atomic_read(&data.error);
}
#else /* CONFIG_PPC_PSERIES */
static int rtas_ibm_suspend_me(struct rtas_args *args)
{
return -ENOSYS;
}
+
+int rtas_suspend_cpu(struct rtas_suspend_me_data *data)
+{
+ return -ENOSYS;
+}
+
+int rtas_suspend_last_cpu(struct rtas_suspend_me_data *data)
+{
+ return -ENOSYS;
+}
#endif
+EXPORT_SYMBOL_GPL(rtas_suspend_cpu);
+EXPORT_SYMBOL_GPL(rtas_suspend_last_cpu);
asmlinkage int ppc_rtas(struct rtas_args __user *uargs)
{
diff -puN arch/powerpc/include/asm/rtas.h~powerpc_allarch_pseries_hibernation arch/powerpc/include/asm/rtas.h
--- linux-2.6/arch/powerpc/include/asm/rtas.h~powerpc_allarch_pseries_hibernation 2010-02-23 15:33:06.000000000 -0600
+++ linux-2.6-bjking1/arch/powerpc/include/asm/rtas.h 2010-02-23 15:33:06.000000000 -0600
@@ -63,6 +63,14 @@ struct rtas_t {
struct device_node *dev; /* virtual address pointer */
};
+struct rtas_suspend_me_data {
+ atomic_t working; /* number of cpus accessing this struct */
+ atomic_t done;
+ int token; /* ibm,suspend-me */
+ atomic_t error;
+ struct completion *complete; /* wait on this until working == 0 */
+};
+
/* RTAS event classes */
#define RTAS_INTERNAL_ERROR 0x80000000 /* set bit 0 */
#define RTAS_EPOW_WARNING 0x40000000 /* set bit 1 */
@@ -174,6 +182,8 @@ extern int rtas_set_indicator(int indica
extern int rtas_set_indicator_fast(int indicator, int index, int new_value);
extern void rtas_progress(char *s, unsigned short hex);
extern void rtas_initialize(void);
+extern int rtas_suspend_cpu(struct rtas_suspend_me_data *data);
+extern int rtas_suspend_last_cpu(struct rtas_suspend_me_data *data);
struct rtc_time;
extern unsigned long rtas_get_boot_time(void);
diff -puN arch/powerpc/Kconfig~powerpc_allarch_pseries_hibernation arch/powerpc/Kconfig
--- linux-2.6/arch/powerpc/Kconfig~powerpc_allarch_pseries_hibernation 2010-02-23 15:33:06.000000000 -0600
+++ linux-2.6-bjking1/arch/powerpc/Kconfig 2010-02-23 15:33:06.000000000 -0600
@@ -216,7 +216,7 @@ config ARCH_HIBERNATION_POSSIBLE
config ARCH_SUSPEND_POSSIBLE
def_bool y
depends on ADB_PMU || PPC_EFIKA || PPC_LITE5200 || PPC_83xx || \
- PPC_85xx || PPC_86xx
+ PPC_85xx || PPC_86xx || PPC_PSERIES
config PPC_DCR_NATIVE
bool
diff -puN arch/powerpc/platforms/pseries/Makefile~powerpc_allarch_pseries_hibernation arch/powerpc/platforms/pseries/Makefile
--- linux-2.6/arch/powerpc/platforms/pseries/Makefile~powerpc_allarch_pseries_hibernation 2010-02-23 15:33:06.000000000 -0600
+++ linux-2.6-bjking1/arch/powerpc/platforms/pseries/Makefile 2010-02-23 15:33:06.000000000 -0600
@@ -26,3 +26,4 @@ obj-$(CONFIG_HCALL_STATS) += hvCall_inst
obj-$(CONFIG_PHYP_DUMP) += phyp_dump.o
obj-$(CONFIG_CMM) += cmm.o
obj-$(CONFIG_DTL) += dtl.o
+obj-$(CONFIG_SUSPEND) += suspend.o
diff -puN arch/powerpc/include/asm/hvcall.h~powerpc_allarch_pseries_hibernation arch/powerpc/include/asm/hvcall.h
--- linux-2.6/arch/powerpc/include/asm/hvcall.h~powerpc_allarch_pseries_hibernation 2010-02-23 15:33:06.000000000 -0600
+++ linux-2.6-bjking1/arch/powerpc/include/asm/hvcall.h 2010-02-23 15:33:06.000000000 -0600
@@ -74,6 +74,7 @@
#define H_NOT_ENOUGH_RESOURCES -44
#define H_R_STATE -45
#define H_RESCINDEND -46
+#define H_MULTI_THREADS_ACTIVE -9005
/* Long Busy is a condition that can be returned by the firmware
diff -puN arch/powerpc/include/asm/machdep.h~powerpc_allarch_pseries_hibernation arch/powerpc/include/asm/machdep.h
--- linux-2.6/arch/powerpc/include/asm/machdep.h~powerpc_allarch_pseries_hibernation 2010-02-23 15:33:06.000000000 -0600
+++ linux-2.6-bjking1/arch/powerpc/include/asm/machdep.h 2010-02-23 15:33:06.000000000 -0600
@@ -265,6 +265,7 @@ struct machdep_calls {
*/
void (*suspend_disable_irqs)(void);
void (*suspend_enable_irqs)(void);
+ int (*suspend_disable_cpu)(void);
#endif
#ifdef CONFIG_ARCH_CPU_PROBE_RELEASE
diff -puN arch/powerpc/platforms/pseries/hotplug-cpu.c~powerpc_allarch_pseries_hibernation arch/powerpc/platforms/pseries/hotplug-cpu.c
--- linux-2.6/arch/powerpc/platforms/pseries/hotplug-cpu.c~powerpc_allarch_pseries_hibernation 2010-02-23 15:33:06.000000000 -0600
+++ linux-2.6-bjking1/arch/powerpc/platforms/pseries/hotplug-cpu.c 2010-02-23 15:33:06.000000000 -0600
@@ -116,6 +116,12 @@ static void pseries_mach_cpu_die(void)
if (get_preferred_offline_state(cpu) == CPU_STATE_INACTIVE) {
set_cpu_current_state(cpu, CPU_STATE_INACTIVE);
+ if (ppc_md.suspend_disable_cpu)
+ ppc_md.suspend_disable_cpu();
+ }
+
+ if (get_preferred_offline_state(cpu) == CPU_STATE_INACTIVE) {
+ set_cpu_current_state(cpu, CPU_STATE_INACTIVE);
cede_latency_hint = 2;
get_lppaca()->idle = 1;
_
^ permalink raw reply
* [PATCH 2/2] powerpc: Add power management support to VIO bus
From: Brian King @ 2010-02-24 0:02 UTC (permalink / raw)
To: benh; +Cc: brking, linuxppc-dev
Adds support for suspend/resume for VIO devices. This is needed for
support for HMC initiated hibernation.
Signed-off-by: Brian King <brking@linux.vnet.ibm.com>
---
arch/powerpc/kernel/vio.c | 24 ++++++++++++++++++++++++
1 file changed, 24 insertions(+)
diff -puN arch/powerpc/kernel/vio.c~powerpc_vio_bus_pm arch/powerpc/kernel/vio.c
--- linux-2.6/arch/powerpc/kernel/vio.c~powerpc_vio_bus_pm 2010-02-21 09:56:18.000000000 -0600
+++ linux-2.6-bjking1/arch/powerpc/kernel/vio.c 2010-02-21 09:57:53.000000000 -0600
@@ -1357,6 +1357,29 @@ static int vio_hotplug(struct device *de
return 0;
}
+static int vio_pm_suspend(struct device *dev)
+{
+ const struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL;
+
+ if (pm && pm->suspend)
+ return pm->suspend(dev);
+ return 0;
+}
+
+static int vio_pm_resume(struct device *dev)
+{
+ const struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL;
+
+ if (pm && pm->resume)
+ return pm->resume(dev);
+ return 0;
+}
+
+const struct dev_pm_ops vio_dev_pm_ops = {
+ .suspend = vio_pm_suspend,
+ .resume = vio_pm_resume,
+};
+
static struct bus_type vio_bus_type = {
.name = "vio",
.dev_attrs = vio_dev_attrs,
@@ -1364,6 +1387,7 @@ static struct bus_type vio_bus_type = {
.match = vio_bus_match,
.probe = vio_bus_probe,
.remove = vio_bus_remove,
+ .pm = &vio_dev_pm_ops,
};
/**
_
^ permalink raw reply
* Re: [PATCH 3/4] Provide VIO modalias
From: Brian King @ 2010-02-24 0:05 UTC (permalink / raw)
To: David Woodhouse; +Cc: linuxppc-dev
In-Reply-To: <1266929400.30247.2585.camel@macbook.infradead.org>
On 02/23/2010 06:50 AM, David Woodhouse wrote:
>> +static ssize_t modalias_show(struct device *dev, struct device_attribute *attr,
>> + char *buf)
>> +{
>> + const struct vio_dev *vio_dev = to_vio_dev(dev);
>> + struct device_node *dn;
>> + const char *cp;
>> +
>> + dn = dev->archdata.of_node;
>> + if (!dn)
>> + return -ENODEV;
>> + cp = of_get_property(dn, "compatible", NULL);
>> + if (!cp)
>> + return -ENODEV;
>> +
>> + return sprintf(buf, "vio:T%sS%s\n", vio_dev->type, cp);
>> +}
>> +
>> static struct device_attribute vio_dev_attrs[] = {
>> __ATTR_RO(name),
>> __ATTR_RO(devspec),
>> + __ATTR_RO(modalias),
>> __ATTR_NULL
>> };
You'll want to add this to vio_cmo_dev_attrs as well, as that is the
device attribute array that gets used when active memory sharing is enabled.
Thanks,
Brian
--
Brian King
Linux on Power Virtualization
IBM Linux Technology Center
^ permalink raw reply
* Re: simpleImage.XX and large kernels
From: David Gibson @ 2010-02-24 0:27 UTC (permalink / raw)
To: John Williams; +Cc: linuxppc-dev
In-Reply-To: <1d3f23371002080247j17138famcf4e4aebc1ae04f5@mail.gmail.com>
On Mon, Feb 08, 2010 at 08:47:44PM +1000, John Williams wrote:
> Hi,
>
> I'm looking at the simpleImage.XXX make target (PPC 405/440), and it
> seems that by default the arch/powerpc/boot/wrapper script places the
> bootwrapper at 0x400000, effectively setting a limit on the maximum
> bootable kernel size.
>
> For various reasons we'd like to be able to put a fairly complete
> rootfs as an initramfs, which obviously blows past a 4Mbyte limit very
> quickly. Short of adding a new 'platform' option and associated
> hackery through the powerpc/boot Makefiles and wrapper, is there a
> quick and clean way I can tell the boot wrapper to link at a higher
> address?
Not really, although altering the wrapper script to change the link
address is really not difficult.
Having the wrapper automatically adjust the link address based on the
size of the decompressed kernel has been on my todo list for ages, but
it's unlikely I'll get around to implementing it any time soon.
--
David Gibson | I'll have my music baroque, and my code
david AT gibson.dropbear.id.au | minimalist, thank you. NOT _the_ _other_
| _way_ _around_!
http://www.ozlabs.org/~dgibson
^ permalink raw reply
* [Fwd: [PATCH] windfarm: init sysfs attributes]
From: Benjamin Herrenschmidt @ 2010-02-24 2:49 UTC (permalink / raw)
To: linuxppc-dev
[-- Attachment #1: Type: text/plain, Size: 1 bytes --]
[-- Attachment #2: Forwarded message - [PATCH] windfarm: init sysfs attributes --]
[-- Type: message/rfc822, Size: 2892 bytes --]
From: Johannes Berg <johannes@sipsolutions.net>
To: "Américo Wang" <xiyou.wangcong@gmail.com>
Cc: Kernel development list <linux-kernel@vger.kernel.org>, "Eric W. Biederman" <ebiederm@xmission.com>, Greg KH <greg@kroah.com>, Benjamin Herrenschmidt <benh@kernel.crashing.org>
Subject: [PATCH] windfarm: init sysfs attributes
Date: Sat, 20 Feb 2010 16:43:02 +0100
Message-ID: <1266680582.17363.7.camel@jlt3.sipsolutions.net>
This is required for lockdep.
Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
---
drivers/macintosh/windfarm_core.c | 1 +
drivers/macintosh/windfarm_smu_controls.c | 1 +
--- wireless-testing.orig/drivers/macintosh/windfarm_smu_controls.c 2010-02-20 16:05:55.738721565 +0100
+++ wireless-testing/drivers/macintosh/windfarm_smu_controls.c 2010-02-20 16:07:51.996717974 +0100
@@ -173,6 +173,7 @@ static struct smu_fan_control *smu_fan_c
fct->fan_type = pwm_fan;
fct->ctrl.type = pwm_fan ? WF_CONTROL_PWM_FAN : WF_CONTROL_RPM_FAN;
+ sysfs_attr_init(&fct->ctrl.attr.attr);
/* We use the name & location here the same way we do for SMU sensors,
* see the comment in windfarm_smu_sensors.c. The locations are a bit
--- wireless-testing.orig/drivers/macintosh/windfarm_core.c 2010-02-20 16:15:14.701580166 +0100
+++ wireless-testing/drivers/macintosh/windfarm_core.c 2010-02-20 16:15:38.416369684 +0100
@@ -321,6 +321,7 @@ int wf_register_sensor(struct wf_sensor
kref_init(&new_sr->ref);
list_add(&new_sr->link, &wf_sensors);
+ sysfs_attr_init(&new_sr->attr.attr);
new_sr->attr.attr.name = new_sr->name;
new_sr->attr.attr.mode = 0444;
new_sr->attr.show = wf_show_sensor;
^ permalink raw reply
* Re: [PATCH 2/4] Fix G5 thermal shutdown
From: Benjamin Herrenschmidt @ 2010-02-24 2:54 UTC (permalink / raw)
To: David Woodhouse; +Cc: linuxppc-dev
In-Reply-To: <1266929284.30247.2581.camel@macbook.infradead.org>
On Tue, 2010-02-23 at 12:48 +0000, David Woodhouse wrote:
> On Tue, 2010-02-09 at 11:55 +1100, Benjamin Herrenschmidt wrote:
> > On Fri, 2010-02-05 at 08:52 -0500, Josh Boyer wrote:
> > > This changes the thresholds for the liquid cooled G5 thermal
> > > shutdown mechanism to prevent an errant shutdown.
> > >
> > > This has been carried since about Fedora Core 5. I have no idea
> > > if it's really needed or not.
>
> It's kind of a dirty hack, but it was sufficient to stop my machine
> constantly powering itself off.
>
> > Can we have a SoB ?
>
> Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
Well, I think you've been testing it long enough now :-)
Cheers,
Ben.
^ permalink raw reply
* Re: [PATCHv4 2/2] powerpc: implement arch_scale_smt_power for Power7
From: Michael Neuling @ 2010-02-24 6:07 UTC (permalink / raw)
To: Peter Zijlstra; +Cc: Ingo Molnar, linuxppc-dev, linux-kernel, ego
In-Reply-To: <1266942281.11845.521.camel@laptop>
In message <1266942281.11845.521.camel@laptop> you wrote:
> On Tue, 2010-02-23 at 17:08 +1100, Michael Neuling wrote:
>
> > I have some comments on the code inline but...
> >
> > So when I run this, I don't get processes pulled down to the lower
> > threads. A simple test case of running 1 CPU intensive process at
> > SCHED_OTHER on a machine with 2 way SMT system (a POWER6 but enabling
> > SD_ASYM_PACKING). The single processes doesn't move to lower threads as
> > I'd hope.
> >
> > Also, are you sure you want to put this in generic code? It seem to be
> > quite POWER7 specific functionality, so would be logically better in
> > arch/powerpc. I guess some other arch *might* need it, but seems
> > unlikely.
>
> Well, there are no arch hooks in the load-balancing (aside from the
> recent cpu_power stuff, and that really is the wrong thing to poke at
> for this), and I did hear some other people express interest in such a
> constraint.
Interesting. Can I ask, were people interesting in this at the SMT
level or higher in the hierarchy?
> Also, load-balancing is complex enough as it is, so I prefer to keep
> everything in the generic code where possible, clearly things like
> sched_domain creation need arch topology bits, and the arch_scale*
> things require other arch information like cpu frequency.
OK
> > > @@ -2493,6 +2494,28 @@ static inline void update_sg_lb_stats(st
> > > DIV_ROUND_CLOSEST(group->cpu_power, SCHED_LOAD_SCALE);
> > > }
> > >
> > > +static int update_sd_pick_busiest(struct sched_domain *sd,
> > > + struct sd_lb_stats *sds,
> > > + struct sched_group *sg,
> > > + struct sg_lb_stats *sgs)
> > > +{
> > > + if (sgs->sum_nr_running > sgs->group_capacity)
> > > + return 1;
> > > +
> > > + if (sgs->group_imb)
> > > + return 1;
> > > +
> > > + if ((sd->flags & SD_ASYM_PACKING) && sgs->sum_nr_running) {
> >
> > If we are asymetric packing...
> >
> >
> > > + if (!sds->busiest)
> > > + return 1;
> >
> > This just seems to be a null pointer check.
> >
> > From the tracing I've done, this is always true (always NULL) at this
> > point so we return here.
>
> Right, so we need to have a busiest group to take a task from, if there
> is no busiest yet, take this group.
>
> And in your scenario, with there being only a single task, we'd only hit
> this once at most, so yes it makes sense this is always NULL.
OK
>
> > > +
> > > + if (group_first_cpu(sds->busiest) < group_first_cpu(sg))
> > > + return 1;
> >
> > I'm a bit lost as to what this is for. Any clues you could provide
> > would be appreciated. :-)
> >
> > Is the first cpu in this domain's busiest group before the first cpu in
> > this group. If, so pick this as the busiest?
> >
> > Should this be the other way around if we want to pack the busiest to
> > the first cpu? Mark it as the busiest if it's after (not before).
> >
> > Is group_first_cpu guaranteed to give us the first physical cpu (ie.
> > thread 0 in our case) or are these virtualised at this point?
> >
> > I'm not seeing this hit anyway due to the null pointer check above.
>
> So this says, if all things being equal, and we already have a busiest,
> but this candidate (sg) is higher than the current (busiest) take this
> one.
>
> The idea is to move the highest SMT task down.
So in the p7 case, I don't think this is required, as the threads are
all of the same performance when in a given SMT mode. So we don't need
to change the order if the lower groups are busy anyway. It's only once
they became idle that we'd need to rebalanced.
That being said, it probably doesn't hurt?
>
> > > @@ -2562,6 +2585,38 @@ static inline void update_sd_lb_stats(st
> > > } while (group != sd->groups);
> > > }
> > >
> > > +int __weak sd_asym_packing_arch(void)
> > > +{
> > > + return 0;
> > > +}
>
> arch_sd_asym_packing() is what you used in topology.h
Oops, thanks. That would make the function even weaker than I'd
intended :-)
> > > +static int check_asym_packing(struct sched_domain *sd,
> > > + struct sd_lb_stats *sds,
> > > + unsigned long *imbalance)
> > > +{
> > > + int i, cpu, busiest_cpu;
> > > +
> > > + if (!(sd->flags & SD_ASYM_PACKING))
> > > + return 0;
> > > +
> > > + if (!sds->busiest)
> > > + return 0;
> > > +
> > > + i = 0;
> > > + busiest_cpu = group_first_cpu(sds->busiest);
> > > + for_each_cpu(cpu, sched_domain_span(sd)) {
> > > + i++;
> > > + if (cpu == busiest_cpu)
> > > + break;
> > > + }
> > > +
> > > + if (sds->total_nr_running > i)
> > > + return 0;
> >
> > This seems to be the core of the packing logic.
> >
> > We make sure the busiest_cpu is not past total_nr_running. If it is we
> > mark as imbalanced. Correct?
> >
> > It seems if a non zero thread/group had a pile of processes running on
> > it and a lower thread had much less, this wouldn't fire, but I'm
> > guessing normal load balancing would kick in that case to fix the
> > imbalance.
> >
> > Any corrections to my ramblings appreciated :-)
>
> Right, so we're concerned the scenario where there's less tasks than SMT
> siblings, if there's more they should all be running and the regular
> load-balancer will deal with it.
Yes
> If there's less the group will normally be balanced and we fall out and
> end up in check_asym_packing().
>
> So what I tried doing with that loop is detect if there's a hole in the
> packing before busiest. Now that I think about it, what we need to check
> is if this_cpu (the removed cpu argument) is idle and less than busiest.
>
> So something like:
>
> static int check_asym_pacing(struct sched_domain *sd,
> struct sd_lb_stats *sds,
> int this_cpu, unsigned long *imbalance)
> {
> int busiest_cpu;
>
> if (!(sd->flags & SD_ASYM_PACKING))
> return 0;
>
> if (!sds->busiest)
> return 0;
>
> busiest_cpu = group_first_cpu(sds->busiest);
> if (cpu_rq(this_cpu)->nr_running || this_cpu > busiest_cpu)
> return 0;
>
> *imbalance = (sds->max_load * sds->busiest->cpu_power) /
> SCHED_LOAD_SCALE;
> return 1;
> }
>
> Does that make sense?
I think so.
I'm seeing check_asym_packing do the right thing with the simple SMT2
with 1 process case. It marks cpu0 as imbalanced when cpu0 is idle and
cpu1 is busy.
Unfortunately the process doesn't seem to be get migrated down though.
Do we need to give *imbalance a higher value?
FYI this version doesn't use sgs->total_nr_running anymore so I've
removed it.
>
> I still see two problems with this though,.. regular load-balancing only
> balances on the first cpu of a domain (see the *balance = 0, condition
> in update_sg_lb_stats()), this means that if SMT[12] are idle we'll not
> pull properly. Also, nohz balancing might mess this up further.
I do have CONFIG_NO_HZ set but turning it off doesn't help with the
above issue.
> We could maybe play some games with the balance decision in
> update_sg_lb_stats() for SD_ASYM_PACKING domains and idle == CPU_IDLE,
> no ideas yet on nohz though.
OK
<from followup email>
> Hmm, we could change the bit in find_busiest_group() to:
>
> if (idle == CPU_IDLE && check_asym_packing())
>
> and skip the nr_running test..
ok, changed.
This is the patch so far minus the trivial PPC bits.
Thanks!
Mikey
---
include/linux/sched.h | 4 ++
include/linux/topology.h | 1
kernel/sched_fair.c | 65 ++++++++++++++++++++++++++++++++++++++++++++---
3 files changed, 66 insertions(+), 4 deletions(-)
Index: linux-next/include/linux/sched.h
===================================================================
--- linux-next.orig/include/linux/sched.h
+++ linux-next/include/linux/sched.h
@@ -849,7 +849,7 @@ enum cpu_idle_type {
#define SD_POWERSAVINGS_BALANCE 0x0100 /* Balance for power savings */
#define SD_SHARE_PKG_RESOURCES 0x0200 /* Domain members share cpu pkg resources */
#define SD_SERIALIZE 0x0400 /* Only a single load balancing instance */
-
+#define SD_ASYM_PACKING 0x0800 /* Asymmetric SMT packing */
#define SD_PREFER_SIBLING 0x1000 /* Prefer to place tasks in a sibling domain */
enum powersavings_balance_level {
@@ -881,6 +881,8 @@ static inline int sd_balance_for_package
return SD_PREFER_SIBLING;
}
+extern int arch_sd_asym_packing(void);
+
/*
* Optimise SD flags for power savings:
* SD_BALANCE_NEWIDLE helps agressive task consolidation and power savings.
Index: linux-next/include/linux/topology.h
===================================================================
--- linux-next.orig/include/linux/topology.h
+++ linux-next/include/linux/topology.h
@@ -102,6 +102,7 @@ int arch_update_cpu_topology(void);
| 1*SD_SHARE_PKG_RESOURCES \
| 0*SD_SERIALIZE \
| 0*SD_PREFER_SIBLING \
+ | arch_sd_asym_packing() \
, \
.last_balance = jiffies, \
.balance_interval = 1, \
Index: linux-next/kernel/sched_fair.c
===================================================================
--- linux-next.orig/kernel/sched_fair.c
+++ linux-next/kernel/sched_fair.c
@@ -2494,6 +2494,32 @@ static inline void update_sg_lb_stats(st
}
/**
+ * update_sd_pick_busiest - return 1 on busiest group
+ */
+static int update_sd_pick_busiest(struct sched_domain *sd,
+ struct sd_lb_stats *sds,
+ struct sched_group *sg,
+ struct sg_lb_stats *sgs)
+{
+ if (sgs->sum_nr_running > sgs->group_capacity)
+ return 1;
+
+ if (sgs->group_imb)
+ return 1;
+
+ /* Check packing mode for this domain */
+ if ((sd->flags & SD_ASYM_PACKING) && sgs->sum_nr_running) {
+ if (!sds->busiest)
+ return 1;
+
+ if (group_first_cpu(sds->busiest) > group_first_cpu(sg))
+ return 1;
+ }
+
+ return 0;
+}
+
+/**
* update_sd_lb_stats - Update sched_group's statistics for load balancing.
* @sd: sched_domain whose statistics are to be updated.
* @this_cpu: Cpu for which load balance is currently performed.
@@ -2547,9 +2573,8 @@ static inline void update_sd_lb_stats(st
sds->this = group;
sds->this_nr_running = sgs.sum_nr_running;
sds->this_load_per_task = sgs.sum_weighted_load;
- } else if (sgs.avg_load > sds->max_load &&
- (sgs.sum_nr_running > sgs.group_capacity ||
- sgs.group_imb)) {
+ } else if (sgs.avg_load >= sds->max_load &&
+ update_sd_pick_busiest(sd, sds, group, &sgs)) {
sds->max_load = sgs.avg_load;
sds->busiest = group;
sds->busiest_nr_running = sgs.sum_nr_running;
@@ -2562,6 +2587,36 @@ static inline void update_sd_lb_stats(st
} while (group != sd->groups);
}
+int __weak arch_sd_asym_packing(void)
+{
+ return 0*SD_ASYM_PACKING;
+}
+
+/**
+ * check_asym_packing - Check to see if we the group is packed into
+ * the sched doman
+ */
+static int check_asym_packing(struct sched_domain *sd,
+ struct sd_lb_stats *sds,
+ int this_cpu, unsigned long *imbalance)
+{
+ int busiest_cpu;
+
+ if (!(sd->flags & SD_ASYM_PACKING))
+ return 0;
+
+ if (!sds->busiest)
+ return 0;
+
+ busiest_cpu = group_first_cpu(sds->busiest);
+ if (this_cpu > busiest_cpu)
+ return 0;
+
+ *imbalance = (sds->max_load * sds->busiest->cpu_power) /
+ SCHED_LOAD_SCALE;
+ return 1;
+}
+
/**
* fix_small_imbalance - Calculate the minor imbalance that exists
* amongst the groups of a sched_domain, during
@@ -2761,6 +2816,10 @@ find_busiest_group(struct sched_domain *
return sds.busiest;
out_balanced:
+ if (idle == CPU_IDLE &&
+ check_asym_packing(sd, &sds, this_cpu, imbalance))
+ return sds.busiest;
+
/*
* There is no obvious imbalance. But check if we can do some balancing
* to save power.
^ permalink raw reply
* [PATCH] Fixed statement without effect by using a correct empty instruction
From: Marcus Meissner @ 2010-02-24 8:51 UTC (permalink / raw)
To: linuxppc-dev, albrecht.dress, grant.likely
This fixes this build failure:
cc1: warnings being treated as errors
/usr/src/packages/BUILD/kernel-default-2.6.33/linux-2.6.32/arch/powerpc/platforms/52xx/mpc52xx_gpt.c: In function 'mpc52xx_gpt_probe':
/usr/src/packages/BUILD/kernel-default-2.6.33/linux-2.6.32/arch/powerpc/platforms/52xx/mpc52xx_gpt.c:761:3: error: statement with no effect
Signed-off-by: Marcus Meissner <meissner@suse.de>
---
arch/powerpc/platforms/52xx/mpc52xx_gpt.c | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/arch/powerpc/platforms/52xx/mpc52xx_gpt.c b/arch/powerpc/platforms/52xx/mpc52xx_gpt.c
index 6f8ebe1..88231d8 100644
--- a/arch/powerpc/platforms/52xx/mpc52xx_gpt.c
+++ b/arch/powerpc/platforms/52xx/mpc52xx_gpt.c
@@ -711,7 +711,7 @@ static int __devinit mpc52xx_gpt_wdt_init(void)
return 0;
}
-#define mpc52xx_gpt_wdt_setup(x, y) (0)
+#define mpc52xx_gpt_wdt_setup(x, y) do { } while (0)
#endif /* CONFIG_MPC5200_WDT */
--
1.6.6.1
^ permalink raw reply related
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox