Skip to contentSkip to conversation

PCI IRQ resource deallocation support [2/3]

PATCH In review Kenji Kaneshige 5 participants 7 messages Sep 21, 2004 – Sep 27, 2004

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 PCI IRQ resource deallocation support [2/3]

This patch is ACPI portion of IRQ deallocation. This patch defines the following new interface. The implementation of this interface depends on each platform.

o void acpi_unregister_gsi(int irq)

This is a opposite portion of acpi_register_gsi(). This has a responsibility for deallocating IRQ resources associated with the specified linux IRQ number.

We need to consider the case of shared interrupt. In the case of shared interrupt, acpi_register_gsi() is called multiple times for one gsi. That is, registrations and unregistrations can be nested.

This function undoes the effect of one call to acpi_register_gsi(). If this matches the last registration, IRQ resources associated with the specified linux IRQ number are freed.

This patch also adds the following new function.

o void acpi_pci_irq_disable (struct pci_dev *dev)

This function is a opposite portion of acpi_pci_enable_irq(). It clears the device's linux IRQ number and calls acpi_unregister_gsi() to deallocate IRQ resources.

---

 linux-2.6.9-rc2-mm1-kanesige/arch/i386/kernel/acpi/boot.c |   11 +++++
 linux-2.6.9-rc2-mm1-kanesige/arch/ia64/kernel/acpi.c      |   11 +++++
 linux-2.6.9-rc2-mm1-kanesige/drivers/acpi/pci_irq.c       |   26 ++++++++++++++
 linux-2.6.9-rc2-mm1-kanesige/include/linux/acpi.h         |    2 +
 4 files changed, 50 insertions(+)
arch/i386/kernel/acpi/boot.c renamed from arch/i386/kernel/acpi/boot.c~IRQ_deallocation_acpi +11 -0
diff -puN arch/i386/kernel/acpi/boot.c~IRQ_deallocation_acpi arch/i386/kernel/acpi/boot.c
@@ -480,6 +480,17 @@ unsigned int acpi_register_gsi(u32 gsi, 
 }
 EXPORT_SYMBOL(acpi_register_gsi);
 
+/*
+ * This function undoes the effect of one call to acpi_register_gsi().
+ * If this matches the last regstration, any IRQ resources for gsi
+ * associated with the irq are freed.
+ */
+void
+acpi_unregister_gsi (unsigned int irq)
+{
+}
+EXPORT_SYMBOL(acpi_unregister_gsi);
+
 static unsigned long __init
 acpi_scan_rsdp (
 	unsigned long		start,
arch/ia64/kernel/acpi.c renamed from arch/ia64/kernel/acpi.c~IRQ_deallocation_acpi +11 -0
diff -puN arch/ia64/kernel/acpi.c~IRQ_deallocation_acpi arch/ia64/kernel/acpi.c
@@ -516,6 +516,17 @@ acpi_register_gsi (u32 gsi, int edge_lev
 }
 EXPORT_SYMBOL(acpi_register_gsi);
 
+/*
+ * This function undoes the effect of one call to acpi_register_gsi().
+ * If this matches the last regstration, any IRQ resources for gsi
+ * associated with the irq are freed.
+ */
+void
+acpi_unregister_gsi (unsigned int irq)
+{
+}
+EXPORT_SYMBOL(acpi_unregister_gsi);
+
 static int __init
 acpi_parse_fadt (unsigned long phys_addr, unsigned long size)
 {
drivers/acpi/pci_irq.c renamed from drivers/acpi/pci_irq.c~IRQ_deallocation_acpi +26 -0
diff -puN drivers/acpi/pci_irq.c~IRQ_deallocation_acpi drivers/acpi/pci_irq.c
@@ -390,3 +390,29 @@ acpi_pci_irq_enable (
 
 	return_VALUE(dev->irq);
 }
+
+void
+acpi_pci_irq_disable (
+	struct pci_dev		*dev)
+{
+	unsigned char irq_disabled, irq;
+
+	ACPI_FUNCTION_TRACE("acpi_pci_irq_disable");
+
+	irq_disabled = dev->irq;
+
+	/*
+	 * dev->irq is cleared by BIOS-assigned IRQ set during boot.
+	 */
+	pci_read_config_byte(dev, PCI_INTERRUPT_PIN, &irq);
+	if (irq)
+		pci_read_config_byte(dev, PCI_INTERRUPT_LINE, &irq);
+	dev->irq = irq;
+
+	printk(KERN_INFO PREFIX "PCI interrupt for device %s disabled\n",
+	       pci_name(dev));
+
+	acpi_unregister_gsi(irq_disabled);
+
+	return_VOID;
+}
include/linux/acpi.h renamed from include/linux/acpi.h~IRQ_deallocation_acpi +2 -0
diff -puN include/linux/acpi.h~IRQ_deallocation_acpi include/linux/acpi.h
@@ -414,6 +414,7 @@ static inline int acpi_boot_init(void)
 #endif 	/*!CONFIG_ACPI_BOOT*/
 
 unsigned int acpi_register_gsi (u32 gsi, int edge_level, int active_high_low);
+void acpi_unregister_gsi (unsigned int);
 int acpi_gsi_to_irq (u32 gsi, unsigned int *irq);
 
 #ifdef CONFIG_ACPI_PCI
@@ -439,6 +440,7 @@ extern struct acpi_prt_list	acpi_prt;
 struct pci_dev;
 
 int acpi_pci_irq_enable (struct pci_dev *dev);
+void acpi_pci_irq_disable (struct pci_dev *dev);
 
 struct acpi_pci_driver {
 	struct acpi_pci_driver *next;

_

-------------------------------------------------------
This SF.Net email is sponsored by: YOU BE THE JUDGE. Be one of 170 Project Admins to receive an Apple iPod Mini FREE for your judgement on who ports your project to Linux PPC the best. Sponsored by IBM.
Deadline: Sept. 24. Go here: http://sf.net/ppc_contest.php

#1 lore raw
Quoted 2 lines · Kenji Kaneshige

+ * This function undoes the effect of one call to acpi_register_gsi().
+ * If this matches the last regstration, any IRQ resources for gsi

s/regstration/registration/ (also other occurrences below).

drivers/acpi/pci_irq.c acpi_pci_irq_enable ( · line 398

Quoted 5 lines
drivers/acpi/pci_irq.c renamed from drivers/acpi/pci_irq.c~IRQ_deallocation_acpi +5 -0
+void
+acpi_pci_irq_disable (
+ struct pci_dev  *dev)
+{
+ unsigned char irq_disabled, irq;

pci_dev.irq is unsigned int, not unsigned char, so irq_disabled should be unsigned int as well.

drivers/acpi/pci_irq.c acpi_pci_irq_enable ( · line 410

Quoted 6 lines
drivers/acpi/pci_irq.c renamed from drivers/acpi/pci_irq.c~IRQ_deallocation_acpi +6 -0
+  * dev->irq is cleared by BIOS-assigned IRQ set during boot.
+  */
+ pci_read_config_byte(dev, PCI_INTERRUPT_PIN, &irq);
+ if (irq)
+  pci_read_config_byte(dev, PCI_INTERRUPT_LINE, &irq);
+ dev->irq = irq;

Why do we need to fiddle with dev->irq? I think it should just be undefined after acpi_pci_irq_disable().

#2 lore raw

Hi Bjorn,

Thank you for your feedbacks.

Quoted 5 lines · Bjorn Helgaas
Quoted 2 lines · Kenji Kaneshige

+ * This function undoes the effect of one call to acpi_register_gsi().
+ * If this matches the last regstration, any IRQ resources for gsi

s/regstration/registration/ (also other occurrences below).

Oops..
I'll fix these.

Quoted 10 lines

drivers/acpi/pci_irq.c acpi_pci_irq_enable ( · line 398

Quoted 6 lines
drivers/acpi/pci_irq.c renamed from drivers/acpi/pci_irq.c~IRQ_deallocation_acpi +5 -0
+void
+acpi_pci_irq_disable (
+ struct pci_dev  *dev)
+{
+ unsigned char irq_disabled, irq;

pci_dev.irq is unsigned int, not unsigned char, so irq_disabled should be unsigned int as well.

I'll fix this, thanks.

Quoted 9 lines

drivers/acpi/pci_irq.c acpi_pci_irq_enable ( · line 410

Quoted 6 lines
drivers/acpi/pci_irq.c renamed from drivers/acpi/pci_irq.c~IRQ_deallocation_acpi +6 -0
+  * dev->irq is cleared by BIOS-assigned IRQ set during boot.
+  */
+ pci_read_config_byte(dev, PCI_INTERRUPT_PIN, &irq);
+ if (irq)
+  pci_read_config_byte(dev, PCI_INTERRUPT_LINE, &irq);
+ dev->irq = irq;

Why do we need to fiddle with dev->irq? I think it should just be undefined after acpi_pci_irq_disable().

I had been considering what the "undefined dev->irq" was. In fact, I had other ideas that was clearing it by zero or -1 (0xffffffff). But I didn't know if we can use these values as a undefined IRQ number. So I'm clearing it by the value which was assigned by PCI core code (pci_read_irq()) before acpi_pci_irq_enable() was called.

How do you think?

Thanks,
Kenji Kaneshige

#3 lore raw

From: Kenji Kaneshige <[email protected]>
Subject: Re: [ACPI] [PATCH] PCI IRQ resource deallocation support [2/3]
Date: Wed, 22 Sep 2004 10:24:40 +0900

Quoted 10 lines
Quoted 9 lines

drivers/acpi/pci_irq.c acpi_pci_irq_enable ( · line 410

Quoted 6 lines
drivers/acpi/pci_irq.c renamed from drivers/acpi/pci_irq.c~IRQ_deallocation_acpi +6 -0
+  * dev->irq is cleared by BIOS-assigned IRQ set during boot.
+  */
+ pci_read_config_byte(dev, PCI_INTERRUPT_PIN, &irq);
+ if (irq)
+  pci_read_config_byte(dev, PCI_INTERRUPT_LINE, &irq);
+ dev->irq = irq;

Why do we need to fiddle with dev->irq? I think it should just be undefined after acpi_pci_irq_disable().

Quoted 6 lines

I had been considering what the "undefined dev->irq" was. In fact, I had other ideas that was clearing it by zero or -1 (0xffffffff). But I didn't know if we can use these values as a undefined IRQ number. So I'm clearing it by the value which was assigned by PCI core code (pci_read_irq()) before acpi_pci_irq_enable() was called.

I think it has little sense in restoring value from the configuration space to dev->irq or clearing it.

If we do preventive programming, it may be worth
trying to define some magic constant (e.g. PCI_UNDEFINED_IRQ) and panic/BUG when the irq is being enabled.
Otherwise, leaving dev->irq as it is would be ok.

---
Takayoshi Kochi <t-kochi-UDFczIW9X1d8UrSeD/[email protected]>

-------------------------------------------------------
This SF.Net email is sponsored by: YOU BE THE JUDGE. Be one of 170 Project Admins to receive an Apple iPod Mini FREE for your judgement on who ports your project to Linux PPC the best. Sponsored by IBM.
Deadline: Sept. 24. Go here: http://sf.net/ppc_contest.php

#4 lore raw

Hi Takayoshi,

Thank you for feedback.

Quoted 10 lines · Takayoshi Kochi
Quoted 9 lines
Quoted 2 lines

Why do we need to fiddle with dev->irq? I think it should just be undefined after acpi_pci_irq_disable().

I had been considering what the "undefined dev->irq" was. In fact, I had other ideas that was clearing it by zero or -1 (0xffffffff). But I didn't know if we can use these values as a undefined IRQ number. So I'm clearing it by the value which was assigned by PCI core code (pci_read_irq()) before acpi_pci_irq_enable() was called.

Quoted 7 lines

I think it has little sense in restoring value from the configuration space to dev->irq or clearing it.

If we do preventive programming, it may be worth
trying to define some magic constant (e.g. PCI_UNDEFINED_IRQ) and panic/BUG when the irq is being enabled.
Otherwise, leaving dev->irq as it is would be ok.

Hmm, I think you are right.
I'll change my patch to leave dev->irq as it is. And then I'll investigate about defining PCI_UNDEFINED_IRQ.

In fact, I posted updated patches a little before. I'll re-post it with your feedback :-)

Thanks,
Kenji Kaneshige

-------------------------------------------------------
This SF.Net email is sponsored by: YOU BE THE JUDGE. Be one of 170 Project Admins to receive an Apple iPod Mini FREE for your judgement on who ports your project to Linux PPC the best. Sponsored by IBM.
Deadline: Sept. 24. Go here: http://sf.net/ppc_contest.php

#5 lore raw
Quoted 3 lines · Kenji Kaneshige

Takayoshi Kochi wrote:
I'll change my patch to leave dev->irq as it is. And then I'll investigate about defining PCI_UNDEFINED_IRQ.

Some platforms (arm, arm26, ppc64) define a macro NO_IRQ:

include/asm-arm/irq.h:#define NO_IRQ        ((unsigned int)(-1))
include/asm-arm26/irq.h:#define NO_IRQ      ((unsigned int)(-1))
include/asm-ppc64/irq.h:#define NO_IRQ      (-1)

Thanks
Guennadi
---
Guennadi Liakhovetski

#6 lore raw

Hi Guennadi Liakhovetski,

Thank you for the information.
I refer to arm, arm26 and ppc64 codes.

Thanks,
Kenji Kaneshige

Quoted 30 lines · Guennadi Liakhovetski
Quoted 3 lines · Kenji Kaneshige

Takayoshi Kochi wrote:
I'll change my patch to leave dev->irq as it is. And then I'll investigate about defining PCI_UNDEFINED_IRQ.

Some platforms (arm, arm26, ppc64) define a macro NO_IRQ:

include/asm-arm/irq.h:#define NO_IRQ        ((unsigned int)(-1))
include/asm-arm26/irq.h:#define NO_IRQ      ((unsigned int)(-1))
include/asm-ppc64/irq.h:#define NO_IRQ      (-1)

Thanks
Guennadi
---
Guennadi Liakhovetski

-------------------------------------------------------
This SF.Net email is sponsored by: YOU BE THE JUDGE. Be one of 170 Project Admins to receive an Apple iPod Mini FREE for your judgement on who ports your project to Linux PPC the best. Sponsored by IBM.
Deadline: Sept. 24. Go here: http://sf.net/ppc_contest.php
_______________________________________________
Acpi-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/acpi-devel