Skip to contentSkip to conversation

PCI fixes for 2.6.25-rc2 git tree

PATCH In review Greg KH 4 participants 7 messages Feb 21, 2008 – Feb 25, 2008 1 message not in the archive

lore
Share & tools
lore

Greg KH

author
#0 lore raw

PATCH 1/1 PCI fixes for 2.6.25-rc2 git tree

Here are some PCI patches against your 2.6.25-rc2 git tree.

They are a collection of PCI quirk additions, build fixes, and some PCI hotplug fixes.

Please pull from:

	master.kernel.org:/pub/scm/linux/kernel/git/gregkh/pci-2.6.git/

The full patches will be sent to the linux-pci mailing list, if anyone wants to see them.

thanks,

greg k-h

-------------

 arch/x86/kernel/acpi/boot.c        |    3 +-
 arch/x86/pci/irq.c                 |    4 ++
 drivers/pci/hotplug/acpiphp_core.c |    6 +++
 drivers/pci/hotplug/acpiphp_ibm.c  |   33 ++++++++-------
 drivers/pci/pci-acpi.c             |    1 +
 drivers/pci/proc.c                 |    2 +-
 drivers/pci/quirks.c               |   79 ++++++++++++++++++++++++++++++++---
 drivers/pci/setup-bus.c            |    6 +--
 drivers/pcmcia/i82092.c            |    7 +++-
 include/linux/pci_ids.h            |    6 +++
 10 files changed, 117 insertions(+), 30 deletions(-)

---------------

Andrew Morton (2):
PCI: drivers/pcmcia/i82092.c: fix up after pci_bus_region changes
PCI: fix up setup-bus.c #ifdef

Crane Cai (1):
PCI: AMD SATA IDE mode quirk

Gary Hade (1):
PCI: hotplug: acpiphp_ibm: Remove get device information

Jason Gaston (2):
PCI: pci_ids: patch for Intel ICH10 DeviceID's
PCI: irq: patch for Intel ICH10 DeviceID's

Kenji Kaneshige (1):
PCI: Fix wrong reference counter check for proc_dir_entry

Peer Chen (1):
PCI: quirks: set 'En' bit of MSI Mapping for devices onHT-based nvidia platform

Randy Dunlap (1):
PCI: kernel-doc: fix pci-acpi warning

Yinghai Lu (1):
PCI: don't load acpi_php when acpi is disabled

A message that is not in the archive (sent off-list or not yet received): [email protected]

#2 lore raw

Re: [PATCH 03/10] PCI: AMD SATA IDE mode quirk

Review

Quoted 11 lines · Greg Kroah-Hartman
(quoted diff) +5 -3
+static void __devinit quirk_amd_ide_mode(struct pci_dev *pdev)
 {
-	/* set sb600 sata to ahci mode */
-	if ((pdev->class >> 8) == PCI_CLASS_STORAGE_IDE) {
-		u8 tmp;
+	/* set sb600/sb700/sb800 sata to ahci mode */
+	u8 tmp;
 
+	pci_read_config_byte(pdev, PCI_CLASS_DEVICE, &tmp);
+	if (tmp == 0x01) {
 		pci_read_config_byte(pdev, 0x40, &tmp);

This seems like a dis-improvement. Why are we reading a config byte for something we already have in the pci_dev? Why are we now checking against 0x01 instead of a symbolic constant? Why are we no longer checking that this is PCI_BASE_CLASS_STORAGE?

Review

Quoted 6 lines
(quoted diff) +4 -2
-DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_ATI, PCI_DEVICE_ID_ATI_IXP600_SATA, quirk_sb600_sata);
-DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_ATI, PCI_DEVICE_ID_ATI_IXP700_SATA, quirk_sb600_sata);
+DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_ATI, PCI_DEVICE_ID_ATI_IXP600_SATA, quirk_amd_ide_mode);
+DECLARE_PCI_FIXUP_RESUME(PCI_VENDOR_ID_ATI, PCI_DEVICE_ID_ATI_IXP600_SATA, quirk_amd_ide_mode);
+DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_ATI, PCI_DEVICE_ID_ATI_IXP700_SATA, quirk_amd_ide_mode);
+DECLARE_PCI_FIXUP_RESUME(PCI_VENDOR_ID_ATI, PCI_DEVICE_ID_ATI_IXP700_SATA, quirk_amd_ide_mode);

Nothing in the changelog entry suggests why we now need FIXUP_RESUME entries when we didn't before.

Signature · 4 lines
Intel are signing my paycheques ... these opinions are still mine
"Bill, look, we understand that you're interested in selling us this
operating system, but compare it to ours.  We can't possibly take such
a retrograde step."
#3 lore raw

RE: [PATCH 03/10] PCI: AMD SATA IDE mode quirk

Quoted 9 lines · Greg Kroah-Hartman
Quoted 8 lines
(quoted diff) +3 -3
+static void __devinit quirk_amd_ide_mode(struct pci_dev *pdev)
 {
-	/* set sb600 sata to ahci mode */
-	if ((pdev->class >> 8) == PCI_CLASS_STORAGE_IDE) {
-		u8 tmp;
+	/* set sb600/sb700/sb800 sata to ahci mode */
+	u8 tmp;
Quoted 9 lines

Review

Quoted 3 lines
(quoted diff) +2 -0
+	pci_read_config_byte(pdev, PCI_CLASS_DEVICE, &tmp);
+	if (tmp == 0x01) {
 		pci_read_config_byte(pdev, 0x40, &tmp);

This seems like a dis-improvement. Why are we reading a config byte for something we already have in the pci_dev? Why are we now checking against 0x01 instead of a symbolic constant? Why are we no longer checking that this is PCI_BASE_CLASS_STORAGE?

It is a quirk. In pci_ids.h did have PCI_CLASS_STORAGE_IDE and PCI_BASE_CLASS_STORAGE, these can not represent the right situation we want to check. 0x01 represents PCI_CLASS_STORAGE_IDE last 2 bit. Also because it
is a quirk, I do not think we need to change pci_ids.h. So 0x01 used.

Quoted 7 lines
Quoted 7 lines

-DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_ATI,
PCI_DEVICE_ID_ATI_IXP600_SATA, quirk_sb600_sata);
-DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_ATI,
PCI_DEVICE_ID_ATI_IXP700_SATA, quirk_sb600_sata);

(quoted diff) +3 -0
+DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_ATI, 
+PCI_DEVICE_ID_ATI_IXP600_SATA, quirk_amd_ide_mode); 
+DECLARE_PCI_FIXUP_RESUME(PCI_VENDOR_ID_ATI, 
Quoted 9 lines

Review

Quoted 5 lines
(quoted diff) +5 -0
+PCI_DEVICE_ID_ATI_IXP600_SATA, quirk_amd_ide_mode); 
+DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_ATI, 
+PCI_DEVICE_ID_ATI_IXP700_SATA, quirk_amd_ide_mode); 
+DECLARE_PCI_FIXUP_RESUME(PCI_VENDOR_ID_ATI, 
+PCI_DEVICE_ID_ATI_IXP700_SATA, quirk_amd_ide_mode);

Nothing in the changelog entry suggests why we now need FIXUP_RESUME entries when we didn't before.

PCI configuration space will be changed by BIOS and then in pci init and restore. So resume also needed.

Quoted 7 lines

--
Intel are signing my paycheques ... these opinions are still mine "Bill, look, we understand that you're interested in selling us this operating system, but compare it to ours. We can't possibly take such a retrograde step."

#4 lore raw

Re: [PATCH 03/10] PCI: AMD SATA IDE mode quirk

Quoted 18 lines · Crane
Quoted 9 lines · Greg Kroah-Hartman
Quoted 8 lines
(quoted diff) +3 -3
+static void __devinit quirk_amd_ide_mode(struct pci_dev *pdev)
 {
-	/* set sb600 sata to ahci mode */
-	if ((pdev->class >> 8) == PCI_CLASS_STORAGE_IDE) {
-		u8 tmp;
+	/* set sb600/sb700/sb800 sata to ahci mode */
+	u8 tmp;
Quoted 9 lines

Review

Quoted 3 lines
(quoted diff) +2 -0
+	pci_read_config_byte(pdev, PCI_CLASS_DEVICE, &tmp);
+	if (tmp == 0x01) {
 		pci_read_config_byte(pdev, 0x40, &tmp);

This seems like a dis-improvement. Why are we reading a config byte for something we already have in the pci_dev? Why are we now checking against 0x01 instead of a symbolic constant? Why are we no longer checking that this is PCI_BASE_CLASS_STORAGE?

Quoted 6 lines

It is a quirk. In pci_ids.h did have PCI_CLASS_STORAGE_IDE and PCI_BASE_CLASS_STORAGE, these can not represent the right situation we want to check. 0x01 represents PCI_CLASS_STORAGE_IDE last 2 bit. Also because it
is a quirk, I do not think we need to change pci_ids.h. So 0x01 used.

You haven't explained what is wrong with the original code:

	if ((pdev->class >> 8) == PCI_CLASS_STORAGE_IDE) {
Quoted 5 lines
Quoted 3 lines

Nothing in the changelog entry suggests why we now need FIXUP_RESUME entries when we didn't before.

PCI configuration space will be changed by BIOS and then in pci init and restore. So resume also needed.

That information needed to be in the changelog.

Signature · 4 lines
Intel are signing my paycheques ... these opinions are still mine
"Bill, look, we understand that you're interested in selling us this
operating system, but compare it to ours.  We can't possibly take such
a retrograde step."
#5 lore raw

RE: [PATCH 03/10] PCI: AMD SATA IDE mode quirk

Quoted 25 lines · Crane
Quoted 1 line
Quoted 1 line

On Thu, Feb 21, 2008 at 03:47:33PM -0800, Greg

Quoted 8 lines · Kroah-Hartman
Quoted 8 lines
Quoted 8 lines
(quoted diff) +3 -3
+static void __devinit quirk_amd_ide_mode(struct pci_dev *pdev)
 {
-	/* set sb600 sata to ahci mode */
-	if ((pdev->class >> 8) == PCI_CLASS_STORAGE_IDE) {
-		u8 tmp;
+	/* set sb600/sb700/sb800 sata to ahci mode */
+	u8 tmp;
Quoted 5 lines
Quoted 5 lines

Review

Quoted 3 lines
(quoted diff) +2 -0
+	pci_read_config_byte(pdev, PCI_CLASS_DEVICE, &tmp);
+	if (tmp == 0x01) {
 		pci_read_config_byte(pdev, 0x40, &tmp);

This seems like a dis-improvement. Why are we reading a

config byte

Quoted 2 lines
Quoted 2 lines

for something we already have in the pci_dev?
Why are we now checking against 0x01 instead of a

symbolic constant?

Quoted 3 lines
Quoted 1 line

Why are we no longer checking that this is PCI_BASE_CLASS_STORAGE?

It is a quirk. In pci_ids.h did have PCI_CLASS_STORAGE_IDE and PCI_BASE_CLASS_STORAGE, these can not represent the right

situation we

Quoted 1 line

want to check. 0x01 represents PCI_CLASS_STORAGE_IDE last 2

Quoted 9 lines

bit. Also

Quoted 1 line

because it is a quirk, I do not think we need to change

pci_ids.h. So

Quoted 1 line

0x01 used.

You haven't explained what is wrong with the original code:

	if ((pdev->class >> 8) == PCI_CLASS_STORAGE_IDE) {

When resume, this pdev->class is quirked, however BIOS has modified pci configuration too. Inconsistance occurs.

Quoted 9 lines
Quoted 1 line
Quoted 1 line

Nothing in the changelog entry suggests why we now need

FIXUP_RESUME

Quoted 3 lines
Quoted 2 lines

entries when we didn't before.

PCI configuration space will be changed by BIOS and then in

pci init

Quoted 1 line

and restore. So resume also needed.

That information needed to be in the changelog.

This info, is a normal info. If maintainer need us to added in source code. I preferred too.

Quoted 8 lines

--
Intel are signing my paycheques ... these opinions are still mine "Bill, look, we understand that you're interested in selling us this operating system, but compare it to ours. We can't possibly take such a retrograde step."

#6 lore raw

Re: [PATCH 03/10] PCI: AMD SATA IDE mode quirk

Quoted 30 lines · Crane
Quoted 30 lines · Crane
Quoted 1 line
Quoted 1 line

On Thu, Feb 21, 2008 at 03:47:33PM -0800, Greg

Quoted 8 lines · Kroah-Hartman
Quoted 8 lines
Quoted 8 lines
(quoted diff) +3 -3
+static void __devinit quirk_amd_ide_mode(struct pci_dev *pdev)
 {
-	/* set sb600 sata to ahci mode */
-	if ((pdev->class >> 8) == PCI_CLASS_STORAGE_IDE) {
-		u8 tmp;
+	/* set sb600/sb700/sb800 sata to ahci mode */
+	u8 tmp;
Quoted 5 lines
Quoted 5 lines

Review

Quoted 3 lines
(quoted diff) +2 -0
+	pci_read_config_byte(pdev, PCI_CLASS_DEVICE, &tmp);
+	if (tmp == 0x01) {
 		pci_read_config_byte(pdev, 0x40, &tmp);

This seems like a dis-improvement. Why are we reading a

config byte

Quoted 2 lines
Quoted 2 lines

for something we already have in the pci_dev?
Why are we now checking against 0x01 instead of a

symbolic constant?

Quoted 3 lines
Quoted 1 line

Why are we no longer checking that this is PCI_BASE_CLASS_STORAGE?

It is a quirk. In pci_ids.h did have PCI_CLASS_STORAGE_IDE and PCI_BASE_CLASS_STORAGE, these can not represent the right

situation we

Quoted 1 line

want to check. 0x01 represents PCI_CLASS_STORAGE_IDE last 2

bit. Also

Quoted 1 line

because it is a quirk, I do not think we need to change

pci_ids.h. So

Quoted 1 line

0x01 used.

Quoted 7 lines
Quoted 4 lines

You haven't explained what is wrong with the original code:

	if ((pdev->class >> 8) == PCI_CLASS_STORAGE_IDE) {

When resume, this pdev->class is quirked, however BIOS has modified pci configuration too. Inconsistance occurs.

Can you update pdev->class from the quirk?
It would be consistent then.
That would leave the code as-is except it's re-reading the field from config space.

hth,
grant

Quoted 19 lines
Quoted 9 lines
Quoted 1 line
Quoted 1 line

Nothing in the changelog entry suggests why we now need

FIXUP_RESUME

Quoted 3 lines
Quoted 2 lines

entries when we didn't before.

PCI configuration space will be changed by BIOS and then in

pci init

Quoted 1 line

and restore. So resume also needed.

That information needed to be in the changelog.

This info, is a normal info. If maintainer need us to added in source code. I preferred too.

Quoted 8 lines

--
Intel are signing my paycheques ... these opinions are still mine "Bill, look, we understand that you're interested in selling us this operating system, but compare it to ours. We can't possibly take such a retrograde step."

#7 lore raw

RE: [PATCH 03/10] PCI: AMD SATA IDE mode quirk

Quoted 35 lines · Crane
Quoted 17 lines
Quoted 17 lines · Crane
Quoted 1 line
Quoted 1 line

On Thu, Feb 21, 2008 at 03:47:33PM -0800, Greg

Quoted 14 lines · Kroah-Hartman
Quoted 14 lines

Review

Quoted 12 lines
(quoted diff) +6 -3
+static void __devinit quirk_amd_ide_mode(struct pci_dev 
+*pdev)
 {
-	/* set sb600 sata to ahci mode */
-	if ((pdev->class >> 8) == PCI_CLASS_STORAGE_IDE) {
-		u8 tmp;
+	/* set sb600/sb700/sb800 sata to ahci mode */
+	u8 tmp;
 
+	pci_read_config_byte(pdev, PCI_CLASS_DEVICE, &tmp);
+	if (tmp == 0x01) {
 		pci_read_config_byte(pdev, 0x40, &tmp);

This seems like a dis-improvement. Why are we reading a

Quoted 5 lines
Quoted 5 lines

config byte

Quoted 2 lines
Quoted 2 lines

for something we already have in the pci_dev?
Why are we now checking against 0x01 instead of a

symbolic constant?

Quoted 1 line
Quoted 1 line

Why are we no longer checking that this is

PCI_BASE_CLASS_STORAGE?

Quoted 11 lines
Quoted 11 lines
Quoted 2 lines

It is a quirk. In pci_ids.h did have PCI_CLASS_STORAGE_IDE and PCI_BASE_CLASS_STORAGE, these can not represent the right

situation we

Quoted 1 line

want to check. 0x01 represents PCI_CLASS_STORAGE_IDE last 2

bit. Also

Quoted 1 line

because it is a quirk, I do not think we need to change

pci_ids.h. So

Quoted 1 line

0x01 used.

You haven't explained what is wrong with the original code:

Quoted 7 lines
Quoted 5 lines
Quoted 2 lines
	if ((pdev->class >> 8) == PCI_CLASS_STORAGE_IDE) {

When resume, this pdev->class is quirked, however BIOS has modified pci configuration too. Inconsistance occurs.

Can you update pdev->class from the quirk?

I think we can, there are many places where pdev->class is changed in quirks.c.

Quoted 31 lines

It would be consistent then.
That would leave the code as-is except it's re-reading the field from config space.

hth,
grant

Quoted 11 lines
Quoted 9 lines
Quoted 1 line
Quoted 1 line

Nothing in the changelog entry suggests why we now need

FIXUP_RESUME

Quoted 3 lines
Quoted 2 lines

entries when we didn't before.

PCI configuration space will be changed by BIOS and then in

pci init

Quoted 1 line

and restore. So resume also needed.

That information needed to be in the changelog.

This info, is a normal info. If maintainer need us to added

in source code. I preferred too.

Quoted 3 lines
Quoted 3 lines

--
Intel are signing my paycheques ... these opinions are still mine "Bill, look, we understand that you're interested in

selling us this

Quoted 5 lines
Quoted 5 lines

operating system, but compare it to ours. We can't possibly take such a retrograde step."