From: Lizhi Hou <lizhi.hou@amd.com>
To: "Ilpo Järvinen" <ilpo.jarvinen@linux.intel.com>
Cc: Hans de Goede <hansg@kernel.org>, <ogabbay@kernel.org>,
<quic_jhugo@quicinc.com>, <maciej.falkowski@linux.intel.com>,
LKML <linux-kernel@vger.kernel.org>, <max.zhen@amd.com>,
<sonal.santan@amd.com>, <mario.limonciello@amd.com>,
<platform-driver-x86@vger.kernel.org>,
<dri-devel@lists.freedesktop.org>, <Shyam-sundar.S-k@amd.com>,
<VinitKumar.Shukla@amd.com>
Subject: Re: [PATCH V2 2/2] accel/amdxdna: Add IOCTL to retrieve realtime NPU power estimate
Date: Thu, 11 Dec 2025 11:06:58 -0800 [thread overview]
Message-ID: <88a18f23-04d9-db8c-9c2c-8bf36afa8d63@amd.com> (raw)
In-Reply-To: <9dd39a22-608b-12a8-ffa1-3937be62f47d@linux.intel.com>
On 12/11/25 10:24, Ilpo Järvinen wrote:
> On Thu, 11 Dec 2025, Lizhi Hou wrote:
>
>> The AMD PMF driver provides an interface to obtain realtime power
>> estimates for the NPU. Expose this information to userspace through a
>> new DRM_IOCTL_AMDXDNA_GET_INFO parameter, allowing applications to query
>> the current NPU power level.
>>
>> Signed-off-by: Lizhi Hou <lizhi.hou@amd.com>
>> ---
>> drivers/accel/amdxdna/aie2_pci.c | 29 +++++++++++++++++++++++++
>> drivers/accel/amdxdna/aie2_pci.h | 18 +++++++++++++++
>> drivers/accel/amdxdna/amdxdna_pci_drv.c | 3 ++-
>> 3 files changed, 49 insertions(+), 1 deletion(-)
>>
>> diff --git a/drivers/accel/amdxdna/aie2_pci.c b/drivers/accel/amdxdna/aie2_pci.c
>> index 81a8e4137bfd..4a2c7addcd79 100644
>> --- a/drivers/accel/amdxdna/aie2_pci.c
>> +++ b/drivers/accel/amdxdna/aie2_pci.c
>> @@ -10,6 +10,7 @@
>> #include <drm/drm_managed.h>
>> #include <drm/drm_print.h>
>> #include <drm/gpu_scheduler.h>
>> +#include <linux/amd-pmf-io.h>
>> #include <linux/cleanup.h>
>> #include <linux/errno.h>
>> #include <linux/firmware.h>
>> @@ -777,6 +778,31 @@ static int aie2_get_clock_metadata(struct amdxdna_client *client,
>> return ret;
>> }
>>
>> +static int aie2_get_sensors(struct amdxdna_client *client,
>> + struct amdxdna_drm_get_info *args)
>> +{
>> + struct amdxdna_drm_query_sensor sensor = { 0 };
> = {} is enough to initialize to defaults.
Will fix.
>
>> + int ret;
>> +
>> + if (args->buffer_size < sizeof(sensor))
>> + return -EINVAL;
>> +
>> + ret = AIE2_GET_PMF_NPU_DATA(npu_power, sensor.input);
>> + if (ret)
>> + return ret;
>> + sensor.type = AMDXDNA_SENSOR_TYPE_POWER;
>> + sensor.unitm = -3;
>> + snprintf(sensor.label, sizeof(sensor.label), "Total Power");
>> + snprintf(sensor.units, sizeof(sensor.units), "mW");
> scnprintf() x2
Will fix.
>> +
>> + if (copy_to_user(u64_to_user_ptr(args->buffer), &sensor, sizeof(sensor)))
>> + return -EFAULT;
>> +
>> + args->buffer_size = sizeof(sensor);
>> +
>> + return 0;
>> +}
>> +
>> static int aie2_hwctx_status_cb(struct amdxdna_hwctx *hwctx, void *arg)
>> {
>> struct amdxdna_drm_hwctx_entry *tmp __free(kfree) = NULL;
>> @@ -980,6 +1006,9 @@ static int aie2_get_info(struct amdxdna_client *client, struct amdxdna_drm_get_i
>> case DRM_AMDXDNA_QUERY_CLOCK_METADATA:
>> ret = aie2_get_clock_metadata(client, args);
>> break;
>> + case DRM_AMDXDNA_QUERY_SENSORS:
>> + ret = aie2_get_sensors(client, args);
>> + break;
>> case DRM_AMDXDNA_QUERY_HW_CONTEXTS:
>> ret = aie2_get_hwctx_status(client, args);
>> break;
>> diff --git a/drivers/accel/amdxdna/aie2_pci.h b/drivers/accel/amdxdna/aie2_pci.h
>> index c6b5cf4ae5c4..edf6f2e00dea 100644
>> --- a/drivers/accel/amdxdna/aie2_pci.h
>> +++ b/drivers/accel/amdxdna/aie2_pci.h
>> @@ -46,6 +46,24 @@
>> pci_resource_len(NDEV2PDEV(_ndev), (_ndev)->xdna->dev_info->mbox_bar); \
>> })
>>
>> +#if IS_ENABLED(CONFIG_AMD_PMF)
>> +#define AIE2_GET_PMF_NPU_DATA(field, val) \
>> +({ \
>> + struct amd_pmf_npu_metrics _npu_metrics; \
>> + int _ret; \
>> + _ret = amd_pmf_get_npu_data(&_npu_metrics); \
>> + val = _ret ? U32_MAX : _npu_metrics.field; \
> #include needed for U32_MAX.
Will fix.
>
> Unrelated to this patch, this files is also missing include for at least
> SZ_* (so likely works by chance) and types.h.
Thanks for pointing this out. I will fix it in future patch.
>
>> + (_ret); \
>> +})
> Please try to align the backslashed to right so that they don't mix up
> with the code.
>
> Add an empty lines after variables declaration (with the backslash in
> the end obviously) so these almost looks like "normal" coding style.
Will fix.
Thanks,
Lizhi
>
>> +#else
>> +#define SENSOR_DEFAULT_npu_power U32_MAX
>> +#define AIE2_GET_PMF_NPU_DATA(field, val) \
>> +({ \
>> + val = SENSOR_DEFAULT_##field; \
>> + (-EOPNOTSUPP); \
>> +})
>> +#endif
>> +
>> enum aie2_smu_reg_idx {
>> SMU_CMD_REG = 0,
>> SMU_ARG_REG,
>> diff --git a/drivers/accel/amdxdna/amdxdna_pci_drv.c b/drivers/accel/amdxdna/amdxdna_pci_drv.c
>> index 1973ab67721b..643ebd387074 100644
>> --- a/drivers/accel/amdxdna/amdxdna_pci_drv.c
>> +++ b/drivers/accel/amdxdna/amdxdna_pci_drv.c
>> @@ -32,9 +32,10 @@ MODULE_FIRMWARE("amdnpu/17f0_20/npu.sbin");
>> * 0.4: Support getting resource information
>> * 0.5: Support getting telemetry data
>> * 0.6: Support preemption
>> + * 0.7: Support getting power data
>> */
>> #define AMDXDNA_DRIVER_MAJOR 0
>> -#define AMDXDNA_DRIVER_MINOR 6
>> +#define AMDXDNA_DRIVER_MINOR 7
>>
>> /*
>> * Bind the driver base on (vendor_id, device_id) pair and later use the
>>
next prev parent reply other threads:[~2025-12-11 19:07 UTC|newest]
Thread overview: 13+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-12-11 17:58 [PATCH V1 0/2] Get real time power input via AMD PMF Lizhi Hou
2025-12-11 17:58 ` [PATCH V1 1/2] platform/x86/amd/pmf: Introduce new interface to export NPU metrics Lizhi Hou
2025-12-11 18:28 ` Ilpo Järvinen
2025-12-11 19:07 ` Lizhi Hou
2025-12-12 6:00 ` Shyam Sundar S K
2025-12-11 17:58 ` [PATCH V2 2/2] accel/amdxdna: Add IOCTL to retrieve realtime NPU power estimate Lizhi Hou
2025-12-11 18:20 ` Mario Limonciello
2025-12-11 18:24 ` Ilpo Järvinen
2025-12-11 19:06 ` Lizhi Hou [this message]
2025-12-12 6:01 ` Shyam Sundar S K
2025-12-11 18:22 ` [PATCH V1 0/2] Get real time power input via AMD PMF Mario Limonciello
2025-12-11 19:16 ` Lizhi Hou
-- strict thread matches above, loose matches on Subject: below --
2025-12-12 18:18 [PATCH V2 " Lizhi Hou
2025-12-12 18:18 ` [PATCH V2 2/2] accel/amdxdna: Add IOCTL to retrieve realtime NPU power estimate Lizhi Hou
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=88a18f23-04d9-db8c-9c2c-8bf36afa8d63@amd.com \
--to=lizhi.hou@amd.com \
--cc=Shyam-sundar.S-k@amd.com \
--cc=VinitKumar.Shukla@amd.com \
--cc=dri-devel@lists.freedesktop.org \
--cc=hansg@kernel.org \
--cc=ilpo.jarvinen@linux.intel.com \
--cc=linux-kernel@vger.kernel.org \
--cc=maciej.falkowski@linux.intel.com \
--cc=mario.limonciello@amd.com \
--cc=max.zhen@amd.com \
--cc=ogabbay@kernel.org \
--cc=platform-driver-x86@vger.kernel.org \
--cc=quic_jhugo@quicinc.com \
--cc=sonal.santan@amd.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