* [PATCH v2] PCI/sysfs: Ensure devices are powered for config reads
@ 2025-09-24 16:57 Brian Norris
2025-09-24 19:05 ` Bjorn Helgaas
0 siblings, 1 reply; 2+ messages in thread
From: Brian Norris @ 2025-09-24 16:57 UTC (permalink / raw)
To: Bjorn Helgaas
Cc: linux-pci, Andrey Ryabinin, Ethan Zhao, linux-kernel,
Brian Norris, stable, Brian Norris
From: Brian Norris <briannorris@google.com>
max_link_width, current_link_speed, current_link_width,
secondary_bus_number, and subordinate_bus_number all access config
registers, but they don't check the runtime PM state. If the device is
in D3cold or a parent bridge is suspended, we may see -EINVAL, bogus
values, or worse, depending on implementation details.
Wrap these access in pci_config_pm_runtime_{get,put}() like most of the
rest of the similar sysfs attributes.
Notably, max_link_speed does not access config registers; it returns a
cached value [1]. So it needs no changes.
[1] Caching was added to pcie_get_speed_cap() in v6.13 via commit
d2bd39c0456b ("PCI: Store all PCIe Supported Link Speeds").
Fixes: 56c1af4606f0 ("PCI: Add sysfs max_link_speed/width, current_link_speed/width, etc")
Cc: stable@vger.kernel.org
Signed-off-by: Brian Norris <briannorris@google.com>
Signed-off-by: Brian Norris <briannorris@chromium.org>
---
Changes in v2:
* Don't touch max_link_speed; it's cached, so we don't actually touch
the hardware
* Improve commit message
drivers/pci/pci-sysfs.c | 20 +++++++++++++++++++-
1 file changed, 19 insertions(+), 1 deletion(-)
diff --git a/drivers/pci/pci-sysfs.c b/drivers/pci/pci-sysfs.c
index f28fdf6dfa02..af74cf02bb90 100644
--- a/drivers/pci/pci-sysfs.c
+++ b/drivers/pci/pci-sysfs.c
@@ -209,8 +209,14 @@ static ssize_t max_link_width_show(struct device *dev,
struct device_attribute *attr, char *buf)
{
struct pci_dev *pdev = to_pci_dev(dev);
+ ssize_t ret;
- return sysfs_emit(buf, "%u\n", pcie_get_width_cap(pdev));
+ /* We read PCI_EXP_LNKCAP, so we need the device to be accessible. */
+ pci_config_pm_runtime_get(pdev);
+ ret = sysfs_emit(buf, "%u\n", pcie_get_width_cap(pdev));
+ pci_config_pm_runtime_put(pdev);
+
+ return ret;
}
static DEVICE_ATTR_RO(max_link_width);
@@ -222,7 +228,10 @@ static ssize_t current_link_speed_show(struct device *dev,
int err;
enum pci_bus_speed speed;
+ pci_config_pm_runtime_get(pci_dev);
err = pcie_capability_read_word(pci_dev, PCI_EXP_LNKSTA, &linkstat);
+ pci_config_pm_runtime_put(pci_dev);
+
if (err)
return -EINVAL;
@@ -239,7 +248,10 @@ static ssize_t current_link_width_show(struct device *dev,
u16 linkstat;
int err;
+ pci_config_pm_runtime_get(pci_dev);
err = pcie_capability_read_word(pci_dev, PCI_EXP_LNKSTA, &linkstat);
+ pci_config_pm_runtime_put(pci_dev);
+
if (err)
return -EINVAL;
@@ -255,7 +267,10 @@ static ssize_t secondary_bus_number_show(struct device *dev,
u8 sec_bus;
int err;
+ pci_config_pm_runtime_get(pci_dev);
err = pci_read_config_byte(pci_dev, PCI_SECONDARY_BUS, &sec_bus);
+ pci_config_pm_runtime_put(pci_dev);
+
if (err)
return -EINVAL;
@@ -271,7 +286,10 @@ static ssize_t subordinate_bus_number_show(struct device *dev,
u8 sub_bus;
int err;
+ pci_config_pm_runtime_get(pci_dev);
err = pci_read_config_byte(pci_dev, PCI_SUBORDINATE_BUS, &sub_bus);
+ pci_config_pm_runtime_put(pci_dev);
+
if (err)
return -EINVAL;
--
2.51.0.536.g15c5d4f767-goog
^ permalink raw reply related [flat|nested] 2+ messages in thread
* Re: [PATCH v2] PCI/sysfs: Ensure devices are powered for config reads
2025-09-24 16:57 [PATCH v2] PCI/sysfs: Ensure devices are powered for config reads Brian Norris
@ 2025-09-24 19:05 ` Bjorn Helgaas
0 siblings, 0 replies; 2+ messages in thread
From: Bjorn Helgaas @ 2025-09-24 19:05 UTC (permalink / raw)
To: Brian Norris
Cc: Bjorn Helgaas, linux-pci, Andrey Ryabinin, Ethan Zhao,
linux-kernel, Brian Norris, stable
On Wed, Sep 24, 2025 at 09:57:11AM -0700, Brian Norris wrote:
> From: Brian Norris <briannorris@google.com>
>
> max_link_width, current_link_speed, current_link_width,
> secondary_bus_number, and subordinate_bus_number all access config
> registers, but they don't check the runtime PM state. If the device is
> in D3cold or a parent bridge is suspended, we may see -EINVAL, bogus
> values, or worse, depending on implementation details.
>
> Wrap these access in pci_config_pm_runtime_{get,put}() like most of the
> rest of the similar sysfs attributes.
>
> Notably, max_link_speed does not access config registers; it returns a
> cached value [1]. So it needs no changes.
>
> [1] Caching was added to pcie_get_speed_cap() in v6.13 via commit
> d2bd39c0456b ("PCI: Store all PCIe Supported Link Speeds").
>
> Fixes: 56c1af4606f0 ("PCI: Add sysfs max_link_speed/width, current_link_speed/width, etc")
> Cc: stable@vger.kernel.org
> Signed-off-by: Brian Norris <briannorris@google.com>
> Signed-off-by: Brian Norris <briannorris@chromium.org>
Applied to pci/pm for v 6.18, thanks!
> ---
>
> Changes in v2:
> * Don't touch max_link_speed; it's cached, so we don't actually touch
> the hardware
> * Improve commit message
>
> drivers/pci/pci-sysfs.c | 20 +++++++++++++++++++-
> 1 file changed, 19 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/pci/pci-sysfs.c b/drivers/pci/pci-sysfs.c
> index f28fdf6dfa02..af74cf02bb90 100644
> --- a/drivers/pci/pci-sysfs.c
> +++ b/drivers/pci/pci-sysfs.c
> @@ -209,8 +209,14 @@ static ssize_t max_link_width_show(struct device *dev,
> struct device_attribute *attr, char *buf)
> {
> struct pci_dev *pdev = to_pci_dev(dev);
> + ssize_t ret;
>
> - return sysfs_emit(buf, "%u\n", pcie_get_width_cap(pdev));
> + /* We read PCI_EXP_LNKCAP, so we need the device to be accessible. */
> + pci_config_pm_runtime_get(pdev);
> + ret = sysfs_emit(buf, "%u\n", pcie_get_width_cap(pdev));
> + pci_config_pm_runtime_put(pdev);
> +
> + return ret;
> }
> static DEVICE_ATTR_RO(max_link_width);
>
> @@ -222,7 +228,10 @@ static ssize_t current_link_speed_show(struct device *dev,
> int err;
> enum pci_bus_speed speed;
>
> + pci_config_pm_runtime_get(pci_dev);
> err = pcie_capability_read_word(pci_dev, PCI_EXP_LNKSTA, &linkstat);
> + pci_config_pm_runtime_put(pci_dev);
> +
> if (err)
> return -EINVAL;
>
> @@ -239,7 +248,10 @@ static ssize_t current_link_width_show(struct device *dev,
> u16 linkstat;
> int err;
>
> + pci_config_pm_runtime_get(pci_dev);
> err = pcie_capability_read_word(pci_dev, PCI_EXP_LNKSTA, &linkstat);
> + pci_config_pm_runtime_put(pci_dev);
> +
> if (err)
> return -EINVAL;
>
> @@ -255,7 +267,10 @@ static ssize_t secondary_bus_number_show(struct device *dev,
> u8 sec_bus;
> int err;
>
> + pci_config_pm_runtime_get(pci_dev);
> err = pci_read_config_byte(pci_dev, PCI_SECONDARY_BUS, &sec_bus);
> + pci_config_pm_runtime_put(pci_dev);
> +
> if (err)
> return -EINVAL;
>
> @@ -271,7 +286,10 @@ static ssize_t subordinate_bus_number_show(struct device *dev,
> u8 sub_bus;
> int err;
>
> + pci_config_pm_runtime_get(pci_dev);
> err = pci_read_config_byte(pci_dev, PCI_SUBORDINATE_BUS, &sub_bus);
> + pci_config_pm_runtime_put(pci_dev);
> +
> if (err)
> return -EINVAL;
>
> --
> 2.51.0.536.g15c5d4f767-goog
>
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2025-09-24 19:05 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2025-09-24 16:57 [PATCH v2] PCI/sysfs: Ensure devices are powered for config reads Brian Norris
2025-09-24 19:05 ` Bjorn Helgaas
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox