From: Badal Nilawar <badal.nilawar@intel.com>
To: intel-xe@lists.freedesktop.org, linux-acpi@vger.kernel.org,
linux-pci@vger.kernel.org
Cc: anshuman.gupta@intel.com, rafael@kernel.org, lenb@kernel.org,
bhelgaas@google.com, ilpo.jarvinen@linux.intel.com,
rodrigo.vivi@intel.com, varun.gupta@intel.com,
ville.syrjala@linux.intel.com, uma.shankar@intel.com,
karthik.poosa@intel.com, matthew.auld@intel.com,
sk.anirban@intel.com, raag.jadav@intel.com
Subject: [PATCH v6 01/12] PCI/ACPI: Add D3cold Aux Power Limit_DSM method
Date: Tue, 13 Jan 2026 22:12:02 +0530 [thread overview]
Message-ID: <20260113164200.1151788-15-badal.nilawar@intel.com> (raw)
In-Reply-To: <20260113164200.1151788-14-badal.nilawar@intel.com>
From: Anshuman Gupta <anshuman.gupta@intel.com>
Implement _DSM method 0Ah, as per PCI Firmware r3.3, sec 4.6.10,
to request auxiliary power required by the device when in D3cold state.
Implementation allows only a single device below the Downstream Port to
request for Aux Power Limit under a given Root Port/Downstream Port
because it does not track and aggregate requests from all child devices
below the Downstream Port as required by PCI Firmware r3.3, sec 4.6.10.
Co-developed-by: Badal Nilawar <badal.nilawar@intel.com>
Signed-off-by: Badal Nilawar <badal.nilawar@intel.com>
Signed-off-by: Anshuman Gupta <anshuman.gupta@intel.com>
---
V2(Bjorn/Rafael):
- Call acpi_dsm_check() to find method 0Ah supported
- Return retry interval to caller
V3(Kuppuswamy)
- Add NULL check for retry interval
V4
- Define enums for aux power request status (Rafael)
- Add Co-developed-by and clean up Signed-off-by (Kappuswamy)
(Bjorn)
- Instead of root pci device pass the pci device of driver, traverse
up the tree and discover _DSM
- Allow only function 0 of device to request aux power
- Allow retry_interval to be NULL
- Refine commit message and function description
V5(Rafael)
- Remove function 0 check and allow first caller of the given
downstream port (with _DSM) to requst aux power including
different function
- Squash Patch v5.02 to this patch
- In the logic, to allow single device to req power, use linked list
instead of adding extra variables to acpi device structure
- return positive code for no main power removal to distinguish from
aux power request granted
---
drivers/pci/pci-acpi.c | 136 +++++++++++++++++++++++++++++++++++++++
include/linux/pci-acpi.h | 9 +++
2 files changed, 145 insertions(+)
diff --git a/drivers/pci/pci-acpi.c b/drivers/pci/pci-acpi.c
index 9369377725fa..645d3005ba50 100644
--- a/drivers/pci/pci-acpi.c
+++ b/drivers/pci/pci-acpi.c
@@ -1418,6 +1418,142 @@ static void pci_acpi_optimize_delay(struct pci_dev *pdev,
ACPI_FREE(obj);
}
+static LIST_HEAD(acpi_aux_pwr_list);
+static DEFINE_MUTEX(acpi_aux_pwr_lock);
+
+struct aux_pwr {
+ u32 aux_pwr_limit; /* aux power limit granted by platform firmware */
+ struct device *dev; /* device to which aux power is granted */
+ struct acpi_device *adev; /* root port/downstream port */
+ struct list_head list;
+};
+
+enum aux_pwr_req_status {
+ AUX_PWR_REQ_DENIED = 0x0,
+ AUX_PWR_REQ_GRANTED = 0x1,
+ AUX_PWR_REQ_NO_MAIN_PWR_REMOVAL = 0x2,
+ AUX_PWR_REQ_RETRY_INTERVAL_MIN = 0x11,
+ AUX_PWR_REQ_RETRY_INTERVAL_MAX = 0x1F
+};
+
+/**
+ * pci_acpi_request_d3cold_aux_power - Request aux power while device is in D3cold
+ * @dev: PCI device instance
+ * @requested_mw: Requested auxiliary power in milliwatts
+ * @retry_interval: Retry interval returned by platform to retry auxiliary
+ * power request
+ *
+ * Request auxilary power to platform firmware, via Root Port/Switch Downstream
+ * Port ACPI _DSM Function 0Ah, needed for the PCI device when it is in D3cold.
+ * Evaluate the _DSM and handle the response accordingly.
+ *
+ * For Multi-Function Devices, driver for Function 0 is required to report an
+ * aggregate power requirement covering all functions contained within the
+ * device.
+ *
+ * Return: 0 Aux power request granted
+ * 1 No main power removal
+ * errno on failure.
+ */
+int pci_acpi_request_d3cold_aux_power(struct pci_dev *dev, u32 requested_mw,
+ u32 *retry_interval)
+{
+ union acpi_object in_obj = {
+ .integer.type = ACPI_TYPE_INTEGER,
+ .integer.value = requested_mw,
+ };
+
+ union acpi_object *out_obj;
+ int result;
+ struct pci_dev *bdev;
+ struct acpi_device *adev;
+ acpi_handle handle;
+ struct aux_pwr *apwr, *next;
+
+ if (!dev)
+ return -EINVAL;
+
+ for (bdev = dev; bdev; bdev = pci_upstream_bridge(bdev)) {
+ handle = ACPI_HANDLE(&bdev->dev);
+ if (handle &&
+ acpi_check_dsm(handle, &pci_acpi_dsm_guid, 4,
+ 1 << DSM_PCI_D3COLD_AUX_POWER_LIMIT))
+ break;
+ }
+
+ if (!bdev)
+ return -ENODEV;
+
+ adev = ACPI_COMPANION(&bdev->dev);
+ if (!adev)
+ return -EINVAL;
+
+ guard(mutex)(&acpi_aux_pwr_lock);
+ /* Check if aux power already granted to different device */
+ list_for_each_entry_safe(apwr, next, &acpi_aux_pwr_list, list) {
+ if (apwr->adev == adev && apwr->dev != &dev->dev) {
+ pci_info(to_pci_dev(apwr->dev),
+ "D3cold Aux Power request already granted: %u mW\n",
+ apwr->aux_pwr_limit);
+ return -EALREADY;
+ }
+ if (apwr->adev == adev && apwr->dev == &dev->dev) {
+ list_del(&apwr->list);
+ kfree(apwr);
+ break;
+ }
+ }
+
+ out_obj = acpi_evaluate_dsm_typed(ACPI_HANDLE(&bdev->dev),
+ &pci_acpi_dsm_guid, 4,
+ DSM_PCI_D3COLD_AUX_POWER_LIMIT,
+ &in_obj, ACPI_TYPE_INTEGER);
+ if (!out_obj)
+ return -EINVAL;
+
+ result = out_obj->integer.value;
+ ACPI_FREE(out_obj);
+
+ if (retry_interval)
+ *retry_interval = 0;
+
+ switch (result) {
+ case AUX_PWR_REQ_DENIED:
+ pci_dbg(bdev, "D3cold Aux Power %u mW request denied\n",
+ requested_mw);
+ return -EINVAL;
+ case AUX_PWR_REQ_GRANTED:
+ pci_info(bdev, "D3cold Aux Power request granted: %u mW\n",
+ requested_mw);
+ apwr = kzalloc(sizeof(*apwr), GFP_KERNEL);
+ if (apwr) {
+ apwr->aux_pwr_limit = requested_mw;
+ apwr->dev = &dev->dev;
+ apwr->adev = adev;
+ INIT_LIST_HEAD(&apwr->list);
+ list_add(&acpi_aux_pwr_list,
+ &apwr->list);
+ }
+ return 0;
+ case AUX_PWR_REQ_NO_MAIN_PWR_REMOVAL:
+ pci_info(bdev, "D3cold Aux Power: Main power won't be removed\n");
+ return 2;
+ case AUX_PWR_REQ_RETRY_INTERVAL_MIN ... AUX_PWR_REQ_RETRY_INTERVAL_MAX:
+ pci_info(bdev, "D3cold Aux Power request needs retry, interval: %u seconds\n",
+ result & 0xF);
+ if (retry_interval) {
+ *retry_interval = result & 0xF;
+ return -EAGAIN;
+ }
+ return -EINVAL;
+ default:
+ pci_err(bdev, "D3cold Aux Power: Reserved or unsupported response: 0x%x\n",
+ result);
+ return -EINVAL;
+ }
+}
+EXPORT_SYMBOL_GPL(pci_acpi_request_d3cold_aux_power);
+
static void pci_acpi_set_external_facing(struct pci_dev *dev)
{
u8 val;
diff --git a/include/linux/pci-acpi.h b/include/linux/pci-acpi.h
index 078225b514d4..ba48e7ddb360 100644
--- a/include/linux/pci-acpi.h
+++ b/include/linux/pci-acpi.h
@@ -121,6 +121,7 @@ extern const guid_t pci_acpi_dsm_guid;
#define DSM_PCI_DEVICE_NAME 0x07
#define DSM_PCI_POWER_ON_RESET_DELAY 0x08
#define DSM_PCI_DEVICE_READINESS_DURATIONS 0x09
+#define DSM_PCI_D3COLD_AUX_POWER_LIMIT 0x0A
#ifdef CONFIG_PCIE_EDR
void pci_acpi_add_edr_notifier(struct pci_dev *pdev);
@@ -132,10 +133,18 @@ static inline void pci_acpi_remove_edr_notifier(struct pci_dev *pdev) { }
int pci_acpi_set_companion_lookup_hook(struct acpi_device *(*func)(struct pci_dev *));
void pci_acpi_clear_companion_lookup_hook(void);
+int pci_acpi_request_d3cold_aux_power(struct pci_dev *dev, u32 requested_mw,
+ u32 *retry_interval);
#else /* CONFIG_ACPI */
static inline void acpi_pci_add_bus(struct pci_bus *bus) { }
static inline void acpi_pci_remove_bus(struct pci_bus *bus) { }
+static inline int pci_acpi_request_d3cold_aux_power(struct pci_dev *dev,
+ u32 requested_mw,
+ u32 *retry_interval)
+{
+ return -EOPNOTSUPP;
+}
#endif /* CONFIG_ACPI */
#endif /* _PCI_ACPI_H_ */
--
2.52.0
next prev parent reply other threads:[~2026-01-13 16:32 UTC|newest]
Thread overview: 39+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-01-13 16:42 [PATCH v6 00/12] VRAM Self Refresh Badal Nilawar
2026-01-13 16:40 ` ✓ CI.KUnit: success for VRAM Self Refresh (rev6) Patchwork
2026-01-13 16:42 ` Badal Nilawar [this message]
2026-01-14 20:24 ` [PATCH v6 01/12] PCI/ACPI: Add D3cold Aux Power Limit_DSM method Bjorn Helgaas
2026-01-20 14:03 ` Nilawar, Badal
2026-01-22 20:53 ` Bjorn Helgaas
2026-01-13 16:42 ` [PATCH v6 02/12] PCI/ACPI: Add PERST# Assertion Delay _DSM method Badal Nilawar
2026-01-13 17:04 ` Manivannan Sadhasivam
2026-01-14 13:47 ` Nilawar, Badal
2026-01-14 19:55 ` Bjorn Helgaas
2026-01-14 20:19 ` Bjorn Helgaas
2026-01-20 15:59 ` Nilawar, Badal
2026-01-22 23:27 ` Bjorn Helgaas
2026-01-13 16:42 ` [PATCH v6 03/12] drm/xe/vrsr: Introduce flag has_vrsr Badal Nilawar
2026-01-13 16:42 ` [PATCH v6 04/12] drm/xe/vrsr: Detect VRSR Capability Badal Nilawar
2026-01-13 16:42 ` [PATCH v6 05/12] drm/xe/vrsr: Initialize VRSR feature Badal Nilawar
2026-01-13 16:42 ` [PATCH v6 06/12] drm/xe/vrsr: Enable VRSR on default VGA boot device Badal Nilawar
2026-01-15 14:25 ` Jani Nikula
2026-01-15 15:25 ` Rodrigo Vivi
2026-01-20 13:28 ` Nilawar, Badal
2026-01-20 13:43 ` Jani Nikula
2026-01-20 14:42 ` Shankar, Uma
2026-01-20 15:37 ` Nilawar, Badal
2026-01-20 15:07 ` Vivi, Rodrigo
2026-01-13 16:42 ` [PATCH v6 07/12] drm/xe/vrsr: Refactor d3cold.allowed to a enum Badal Nilawar
2026-01-13 16:42 ` [PATCH v6 08/12] drm/xe/pm: D3cold target state Badal Nilawar
2026-01-13 16:42 ` [PATCH v6 09/12] drm/xe/pm: Refactor PM Sleep Ops Badal Nilawar
2026-01-14 18:00 ` Bjorn Helgaas
2026-01-20 14:05 ` Nilawar, Badal
2026-01-13 16:42 ` [PATCH v6 10/12] drm/xe/vrsr: Enable VRSR Badal Nilawar
2026-01-14 18:02 ` Bjorn Helgaas
2026-01-13 16:42 ` [PATCH v6 11/12] drm/xe/pm/s2idle: Don't evict user BOs D3cold-VRSR state Badal Nilawar
2026-01-13 16:42 ` [PATCH v6 12/12] drm/xe/vrsr: Introduce a debugfs node named vrsr_capable Badal Nilawar
2026-01-13 16:55 ` ✗ CI.checksparse: warning for VRAM Self Refresh (rev6) Patchwork
2026-01-14 2:41 ` ✗ Xe.CI.Full: failure " Patchwork
2026-01-14 5:57 ` ✓ CI.KUnit: success for VRAM Self Refresh (rev7) Patchwork
2026-01-14 6:13 ` ✗ CI.checksparse: warning " Patchwork
2026-01-14 6:31 ` ✓ Xe.CI.BAT: success " Patchwork
2026-01-14 13:55 ` ✗ Xe.CI.Full: failure " Patchwork
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20260113164200.1151788-15-badal.nilawar@intel.com \
--to=badal.nilawar@intel.com \
--cc=anshuman.gupta@intel.com \
--cc=bhelgaas@google.com \
--cc=ilpo.jarvinen@linux.intel.com \
--cc=intel-xe@lists.freedesktop.org \
--cc=karthik.poosa@intel.com \
--cc=lenb@kernel.org \
--cc=linux-acpi@vger.kernel.org \
--cc=linux-pci@vger.kernel.org \
--cc=matthew.auld@intel.com \
--cc=raag.jadav@intel.com \
--cc=rafael@kernel.org \
--cc=rodrigo.vivi@intel.com \
--cc=sk.anirban@intel.com \
--cc=uma.shankar@intel.com \
--cc=varun.gupta@intel.com \
--cc=ville.syrjala@linux.intel.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox