From: Rodrigo Vivi <rodrigo.vivi@intel.com>
To: "Nilawar, Badal" <badal.nilawar@intel.com>
Cc: Jani Nikula <jani.nikula@linux.intel.com>,
<intel-xe@lists.freedesktop.org>, <anshuman.gupta@intel.com>,
<raag.jadav@intel.com>, <riana.tauro@intel.com>,
<mallesh.koujalagi@intel.com>, <aravind.iddamsetty@intel.com>,
<himal.prasad.ghimiray@intel.com>, <michal.wajdeczko@intel.com>
Subject: Re: [PATCH v3] drm/xe/cri: Expose device UID through sysfs
Date: Wed, 30 Sep 2026 09:12:14 -0400 [thread overview]
Message-ID: <ar0Kri7MYtu3oZ_T@intel.com> (raw)
In-Reply-To: <56bce828-8f56-43b2-92a5-1081ce18c595@intel.com>
On Wed, Sep 30, 2026 at 05:23:06PM +0530, Nilawar, Badal wrote:
>
> On 30-09-2026 12:56, Jani Nikula wrote:
> > On Tue, 29 Sep 2026, Badal Nilawar <badal.nilawar@intel.com> wrote:
> > > Expose a read-only sysfs attribute, device_uid, that reports the
> > > GPU SoC's unique identifier.
> > The commit message does not answer the question *why*.
>
> Unfortunately this patch is merged. Use case is to provide alternative to
> PCIe DSN (Device serial number) Capability, which is currently not supported
> by Xe platforms.
>
> In scenarios like hotplug or recovery from wedge, PCIe bdf may change after
> rescan. With device_uid will help userspace app identify recovered device is
> expected device.
Hi Jani, sorry for missing that. I had a biased vision already and my poor brain
assumed universability. Anyway, in large deployments with multiple GPUs in the
system with resets or hotplug cases, BDF is not enough to identify the part.
So, for a proper debugability and server maintenance it is necessary some stable
identification method. UUID would be the ideal, but we don't have this in our parts.
But UID is enough for now...
>
> Thanks,
> Badal
>
> >
> > BR,
> > Jani.
> >
> > > Bspec: 53048, 53049
> > > Assisted-by: Claude:claude-opus-4.8
> > > Signed-off-by: Badal Nilawar <badal.nilawar@intel.com>
> > > ---
> > > v2:
> > > - add has flag instead of platform check to determin uid support (Anshuman)
> > > - Fix the DOC: section (Michal)
> > > - %s/_uuid/uid (Joonas)
> > > v3:
> > > - Restrict device_uid to admin users (Sashiko/Joonas)
> > > - Address review comments (Michal)
> > > ---
> > > .../ABI/testing/sysfs-driver-intel-xe-gpu | 10 ++++++
> > > drivers/gpu/drm/xe/regs/xe_regs.h | 2 ++
> > > drivers/gpu/drm/xe/xe_device.c | 9 +++++
> > > drivers/gpu/drm/xe/xe_device_sysfs.c | 36 +++++++++++++++++++
> > > drivers/gpu/drm/xe/xe_device_types.h | 5 +++
> > > drivers/gpu/drm/xe/xe_pci.c | 2 ++
> > > drivers/gpu/drm/xe/xe_pci_types.h | 1 +
> > > 7 files changed, 65 insertions(+)
> > > create mode 100644 Documentation/ABI/testing/sysfs-driver-intel-xe-gpu
> > >
> > > diff --git a/Documentation/ABI/testing/sysfs-driver-intel-xe-gpu b/Documentation/ABI/testing/sysfs-driver-intel-xe-gpu
> > > new file mode 100644
> > > index 000000000000..4dc4afc3961d
> > > --- /dev/null
> > > +++ b/Documentation/ABI/testing/sysfs-driver-intel-xe-gpu
> > > @@ -0,0 +1,10 @@
> > > +What: /sys/bus/pci/drivers/xe/.../device_uid
> > > +Date: October 2026
> > > +KernelVersion: 7.4
> > > +Contact: intel-xe@lists.freedesktop.org
> > > +Description:
> > > + RO. Unique 64-bit identifier of the GPU SoC device, exposed as
> > > + hexadecimal value.
> > > +
> > > + This sysfs file is present only on supported Intel Xe platforms.
> > > + Accessible only to users with administrative privileges.
> > > diff --git a/drivers/gpu/drm/xe/regs/xe_regs.h b/drivers/gpu/drm/xe/regs/xe_regs.h
> > > index ef4746b7b5d3..437485b5a0af 100644
> > > --- a/drivers/gpu/drm/xe/regs/xe_regs.h
> > > +++ b/drivers/gpu/drm/xe/regs/xe_regs.h
> > > @@ -30,6 +30,8 @@
> > > #define XEHP_MTCFG_ADDR XE_REG(0x101800)
> > > #define TILE_COUNT REG_GENMASK(15, 8)
> > > +#define CRI_DEVICE_UID XE_REG(0x102008)
> > > +
> > > #define GGC XE_REG(0x108040)
> > > #define GMS_MASK REG_GENMASK(15, 8)
> > > #define GGMS_MASK REG_GENMASK(7, 6)
> > > diff --git a/drivers/gpu/drm/xe/xe_device.c b/drivers/gpu/drm/xe/xe_device.c
> > > index 44975e7823be..e1979bbcf7bc 100644
> > > --- a/drivers/gpu/drm/xe/xe_device.c
> > > +++ b/drivers/gpu/drm/xe/xe_device.c
> > > @@ -672,6 +672,7 @@ static void vf_update_device_info(struct xe_device *xe)
> > > xe->info.skip_guc_pc = 1;
> > > xe->info.skip_pcode = 1;
> > > xe->info.has_drm_ras = false;
> > > + xe->info.has_device_uid = false;
> > > }
> > > static int xe_device_vram_alloc(struct xe_device *xe)
> > > @@ -872,6 +873,12 @@ static int xe_debug_page_size_alloc_ctrl_init(struct xe_device *xe)
> > > }
> > > #endif
> > > +static void xe_uid_probe(struct xe_device *xe)
> > > +{
> > > + if (xe->info.has_device_uid)
> > > + xe->device_uid = xe_mmio_read64_2x32(xe_root_tile_mmio(xe), CRI_DEVICE_UID);
> > > +}
> > > +
> > > int xe_device_probe(struct xe_device *xe)
> > > {
> > > struct xe_tile *tile;
> > > @@ -879,6 +886,8 @@ int xe_device_probe(struct xe_device *xe)
> > > int err;
> > > u8 id;
> > > + xe_uid_probe(xe);
> > > +
> > > xe_pat_init_early(xe);
> > > err = xe_sriov_init(xe);
> > > diff --git a/drivers/gpu/drm/xe/xe_device_sysfs.c b/drivers/gpu/drm/xe/xe_device_sysfs.c
> > > index a73e0e957cb0..18783a286a83 100644
> > > --- a/drivers/gpu/drm/xe/xe_device_sysfs.c
> > > +++ b/drivers/gpu/drm/xe/xe_device_sysfs.c
> > > @@ -8,6 +8,7 @@
> > > #include <linux/pci.h>
> > > #include <linux/sysfs.h>
> > > +#include "regs/xe_regs.h"
> > > #include "xe_device.h"
> > > #include "xe_device_sysfs.h"
> > > #include "xe_mmio.h"
> > > @@ -264,6 +265,35 @@ static const struct attribute_group auto_link_downgrade_attr_group = {
> > > .attrs = auto_link_downgrade_attrs,
> > > };
> > > +/**
> > > + * DOC: Device Unique ID
> > > + *
> > > + * On supported platforms, Xe driver exposes a unique 64-bit GPU SOC
> > > + * device identifier through the 'device_uid' sysfs entry.
> > > + *
> > > + * See Documentation/ABI/testing/sysfs-driver-intel-xe-gpu for the ABI
> > > + * specification.
> > > + */
> > > +
> > > +static ssize_t
> > > +device_uid_show(struct device *dev, struct device_attribute *attr, char *buf)
> > > +{
> > > + struct pci_dev *pdev = to_pci_dev(dev);
> > > + struct xe_device *xe = pdev_to_xe_device(pdev);
> > > +
> > > + return sysfs_emit(buf, "0x%016llx\n", xe->device_uid);
> > > +}
> > > +static DEVICE_ATTR_ADMIN_RO(device_uid);
> > > +
> > > +static struct attribute *device_uid_attrs[] = {
> > > + &dev_attr_device_uid.attr,
> > > + NULL
> > > +};
> > > +
> > > +static const struct attribute_group device_uid_attr_group = {
> > > + .attrs = device_uid_attrs,
> > > +};
> > > +
> > > int xe_device_sysfs_init(struct xe_device *xe)
> > > {
> > > struct device *dev = xe->drm.dev;
> > > @@ -285,5 +315,11 @@ int xe_device_sysfs_init(struct xe_device *xe)
> > > return ret;
> > > }
> > > + if (xe->info.has_device_uid) {
> > > + ret = devm_device_add_group(dev, &device_uid_attr_group);
> > > + if (ret)
> > > + return ret;
> > > + }
> > > +
> > > return 0;
> > > }
> > > diff --git a/drivers/gpu/drm/xe/xe_device_types.h b/drivers/gpu/drm/xe/xe_device_types.h
> > > index cd9ed8ff2940..41f703a87916 100644
> > > --- a/drivers/gpu/drm/xe/xe_device_types.h
> > > +++ b/drivers/gpu/drm/xe/xe_device_types.h
> > > @@ -179,6 +179,8 @@ struct xe_device {
> > > u8 has_cached_pt:1;
> > > /** @info.has_device_atomics_on_smem: Supports device atomics on SMEM */
> > > u8 has_device_atomics_on_smem:1;
> > > + /** @info.has_device_uid: Device supports unique 64-bit GPU SOC ID */
> > > + u8 has_device_uid:1;
> > > /** @info.has_drm_ras: Device supports drm_ras (Reliability, Availability, Serviceability) */
> > > u8 has_drm_ras:1;
> > > /** @info.has_fan_control: Device supports fan control */
> > > @@ -265,6 +267,9 @@ struct xe_device {
> > > bool oob_initialized;
> > > } wa_active;
> > > + /** @device_uid: unique 64-bit GPU SOC identifier */
> > > + u64 device_uid;
> > > +
> > > /** @survivability: survivability information for device */
> > > struct xe_survivability survivability;
> > > diff --git a/drivers/gpu/drm/xe/xe_pci.c b/drivers/gpu/drm/xe/xe_pci.c
> > > index 28f37e034e02..032a08fe41e6 100644
> > > --- a/drivers/gpu/drm/xe/xe_pci.c
> > > +++ b/drivers/gpu/drm/xe/xe_pci.c
> > > @@ -472,6 +472,7 @@ static const struct xe_device_desc cri_desc = {
> > > PLATFORM(CRESCENTISLAND),
> > > .dma_mask_size = 52,
> > > .has_display = false,
> > > + .has_device_uid = true,
> > > .has_drm_ras = true,
> > > .has_flat_ccs = false,
> > > .has_gsc_nvm = 1,
> > > @@ -793,6 +794,7 @@ static int xe_info_init_early(struct xe_device *xe,
> > > xe->info.is_dgfx = desc->is_dgfx;
> > > xe->info.has_cached_pt = desc->has_cached_pt;
> > > + xe->info.has_device_uid = desc->has_device_uid;
> > > xe->info.has_drm_ras = desc->has_drm_ras;
> > > xe->info.has_fan_control = desc->has_fan_control;
> > > /* runtime fusing may force flat_ccs to disabled later */
> > > diff --git a/drivers/gpu/drm/xe/xe_pci_types.h b/drivers/gpu/drm/xe/xe_pci_types.h
> > > index 71068cdb3558..448bbe5ba1d7 100644
> > > --- a/drivers/gpu/drm/xe/xe_pci_types.h
> > > +++ b/drivers/gpu/drm/xe/xe_pci_types.h
> > > @@ -40,6 +40,7 @@ struct xe_device_desc {
> > > u8 has_cached_pt:1;
> > > u8 has_display:1;
> > > + u8 has_device_uid:1;
> > > u8 has_drm_ras:1;
> > > u8 has_fan_control:1;
> > > u8 has_flat_ccs:1;
prev parent reply other threads:[~2026-09-30 13:12 UTC|newest]
Thread overview: 9+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-09-29 17:18 [PATCH v3] drm/xe/cri: Expose device UID through sysfs Badal Nilawar
2026-09-29 17:07 ` ✗ CI.checkpatch: warning for " Patchwork
2026-09-29 17:09 ` ✓ CI.KUnit: success " Patchwork
2026-09-29 18:24 ` ✓ Xe.CI.BAT: " Patchwork
2026-09-29 18:37 ` [PATCH v3] " Rodrigo Vivi
2026-09-29 22:36 ` ✗ Xe.CI.FULL: failure for " Patchwork
2026-09-30 7:26 ` [PATCH v3] " Jani Nikula
2026-09-30 11:53 ` Nilawar, Badal
2026-09-30 13:12 ` Rodrigo Vivi [this message]
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=ar0Kri7MYtu3oZ_T@intel.com \
--to=rodrigo.vivi@intel.com \
--cc=anshuman.gupta@intel.com \
--cc=aravind.iddamsetty@intel.com \
--cc=badal.nilawar@intel.com \
--cc=himal.prasad.ghimiray@intel.com \
--cc=intel-xe@lists.freedesktop.org \
--cc=jani.nikula@linux.intel.com \
--cc=mallesh.koujalagi@intel.com \
--cc=michal.wajdeczko@intel.com \
--cc=raag.jadav@intel.com \
--cc=riana.tauro@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