linux-kernel - Re: [PATCH v1 1/1] x86/PCI: Introduce pcibios_is_irq

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20210802231643.GA1474171@bjorn-Precision-5520>
Date:   Mon, 2 Aug 2021 18:16:43 -0500
From:   Bjorn Helgaas <helgaas@...nel.org>
To:     Andy Shevchenko <andriy.shevchenko@...ux.intel.com>
Cc:     Thomas Gleixner <tglx@...utronix.de>, linux-kernel@...r.kernel.org,
        linux-pci@...r.kernel.org, linux-acpi@...r.kernel.org,
        Ingo Molnar <mingo@...hat.com>, Borislav Petkov <bp@...en8.de>,
        x86@...nel.org, "H. Peter Anvin" <hpa@...or.com>,
        Bjorn Helgaas <bhelgaas@...gle.com>,
        "Rafael J. Wysocki" <rjw@...ysocki.net>,
        Len Brown <lenb@...nel.org>
Subject: Re: [PATCH v1 1/1] x86/PCI: Introduce pcibios_is_irq_managed() helper

On Fri, Jul 30, 2021 at 11:53:55PM +0300, Andy Shevchenko wrote:
> The check for irq_managed flag along with non-zero irq is an idiom
> for x86 PCI implementation. Introduce helper and switch users over
> using it.
> 
> Signed-off-by: Andy Shevchenko <andriy.shevchenko@...ux.intel.com>
> ---
>  arch/x86/include/asm/pci.h   | 4 ++++
>  arch/x86/pci/intel_mid_pci.c | 5 ++---
>  arch/x86/pci/irq.c           | 4 ++--
>  drivers/acpi/pci_irq.c       | 4 ++--
>  4 files changed, 10 insertions(+), 7 deletions(-)
> 
> diff --git a/arch/x86/include/asm/pci.h b/arch/x86/include/asm/pci.h
> index d2c76c8d8cfd..ac25470c9558 100644
> --- a/arch/x86/include/asm/pci.h
> +++ b/arch/x86/include/asm/pci.h
> @@ -92,6 +92,10 @@ void pcibios_scan_root(int bus);
>  struct irq_routing_table *pcibios_get_irq_routing_table(void);
>  int pcibios_set_irq_routing(struct pci_dev *dev, int pin, int irq);
>  
> +static inline bool pcibios_irq_is_managed(struct pci_dev *dev)
> +{
> +	return dev->irq_managed && dev->irq > 0;
> +}
>  
>  #define HAVE_PCI_MMAP
>  #define arch_can_pci_mmap_wc()	pat_enabled()
> diff --git a/arch/x86/pci/intel_mid_pci.c b/arch/x86/pci/intel_mid_pci.c
> index f04742caf62b..0da287bcabf5 100644
> --- a/arch/x86/pci/intel_mid_pci.c
> +++ b/arch/x86/pci/intel_mid_pci.c
> @@ -230,7 +230,7 @@ static int intel_mid_pci_irq_enable(struct pci_dev *dev)
>  	int ret;
>  	u8 gsi;
>  
> -	if (dev->irq_managed && dev->irq > 0)
> +	if (pcibios_irq_is_managed(dev))
>  		return 0;
>  
>  	ret = pci_read_config_byte(dev, PCI_INTERRUPT_LINE, &gsi);
> @@ -290,8 +290,7 @@ static int intel_mid_pci_irq_enable(struct pci_dev *dev)
>  
>  static void intel_mid_pci_irq_disable(struct pci_dev *dev)
>  {
> -	if (!mp_should_keep_irq(&dev->dev) && dev->irq_managed &&
> -	    dev->irq > 0) {
> +	if (pcibios_irq_is_managed(dev) && !mp_should_keep_irq(&dev->dev)) {
>  		mp_unmap_irq(dev->irq);
>  		dev->irq_managed = 0;
>  	}
> diff --git a/arch/x86/pci/irq.c b/arch/x86/pci/irq.c
> index d3a73f9335e1..ce3927b68f9e 100644
> --- a/arch/x86/pci/irq.c
> +++ b/arch/x86/pci/irq.c
> @@ -1210,7 +1210,7 @@ static int pirq_enable_irq(struct pci_dev *dev)
>  			struct pci_dev *temp_dev;
>  			int irq;
>  
> -			if (dev->irq_managed && dev->irq > 0)
> +			if (pcibios_irq_is_managed(dev))
>  				return 0;
>  
>  			irq = IO_APIC_get_PCI_irq_vector(dev->bus->number,
> @@ -1280,7 +1280,7 @@ bool mp_should_keep_irq(struct device *dev)
>  static void pirq_disable_irq(struct pci_dev *dev)
>  {
>  	if (io_apic_assign_pci_irqs && !mp_should_keep_irq(&dev->dev) &&
> -	    dev->irq_managed && dev->irq) {
> +	    pcibios_irq_is_managed(dev)) {
>  		mp_unmap_irq(dev->irq);
>  		dev->irq = 0;
>  		dev->irq_managed = 0;
> diff --git a/drivers/acpi/pci_irq.c b/drivers/acpi/pci_irq.c
> index b63954c36e86..b463bdd2dbb5 100644
> --- a/drivers/acpi/pci_irq.c
> +++ b/drivers/acpi/pci_irq.c
> @@ -397,7 +397,7 @@ int __acpi_pci_irq_enable(struct pci_dev *dev, int polarity)
>  		return 0;
>  	}
>  
> -	if (dev->irq_managed && dev->irq > 0)
> +	if (pcibios_irq_is_managed(dev))
>  		return 0;
>  
>  	entry = acpi_pci_irq_lookup(dev, pin);
> @@ -486,7 +486,7 @@ void acpi_pci_irq_disable(struct pci_dev *dev)
>  	u8 pin;
>  
>  	pin = dev->pin;
> -	if (!pin || !dev->irq_managed || dev->irq <= 0)
> +	if (!pin || !pcibios_irq_is_managed(dev))

This file may be compiled for x86, arm64, and ia64, but it looks like
you only defined pcibios_irq_is_managed() for x86.

We used to have a generic pci_has_managed_irq() that was the same as
your pcibios_irq_is_managed(), but it was removed by 67b4eab91caf
("Revert "PCI: Add helpers to manage pci_dev->irq and
pci_dev->irq_managed"").

I'm sorry that pdev->irq_managed got added without a comment by
cffe0a2b5a34 ("x86, irq: Keep balance of IOAPIC pin reference count").
Even reading that commit log, I'm not sure exactly what it means.  The
log says:

  So flag irq_managed is introduced to track whether IRQ number is
  assigned by OS and to protect pirq_enable_irq(), acpi_pci_irq_enable()
  and intel_mid_pci_irq_enable() from reentrance.

>  		return;
>  
>  	/* Keep IOAPIC pin configuration when suspending */
> -- 
> 2.30.2
>