* [PATCH 0/2] Make device links between KFD and GPU device @ 2025-12-07 14:04 Mario Limonciello (AMD) 2025-12-07 14:04 ` [PATCH 1/2] amdkfd: Only ignore -ENOENT for KFD init failuires Mario Limonciello (AMD) 2025-12-07 14:04 ` [PATCH 2/2] amdkfd: Add device links between kfd device and amdgpu device Mario Limonciello (AMD) 0 siblings, 2 replies; 4+ messages in thread From: Mario Limonciello (AMD) @ 2025-12-07 14:04 UTC (permalink / raw) To: amd-gfx; +Cc: Mario Limonciello (AMD) Discovering which KFD device is associated with a GPU is relatively awkward right now in userspace. This series creates sysfs links between the devices to simplify it for userspace. Mario Limonciello (AMD) (2): amdkfd: Only ignore -ENOENT for KFD init failuires amdkfd: Add device links between kfd device and amdgpu device drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c | 8 +++++ drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h | 1 + drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 4 +++ drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c | 6 ++-- drivers/gpu/drm/amd/amdkfd/kfd_topology.c | 36 +++++++++++++++++++ .../gpu/drm/amd/include/kgd_kfd_interface.h | 2 ++ 6 files changed, 55 insertions(+), 2 deletions(-) -- 2.43.0 ^ permalink raw reply [flat|nested] 4+ messages in thread
* [PATCH 1/2] amdkfd: Only ignore -ENOENT for KFD init failuires 2025-12-07 14:04 [PATCH 0/2] Make device links between KFD and GPU device Mario Limonciello (AMD) @ 2025-12-07 14:04 ` Mario Limonciello (AMD) 2025-12-10 17:06 ` Russell, Kent 2025-12-07 14:04 ` [PATCH 2/2] amdkfd: Add device links between kfd device and amdgpu device Mario Limonciello (AMD) 1 sibling, 1 reply; 4+ messages in thread From: Mario Limonciello (AMD) @ 2025-12-07 14:04 UTC (permalink / raw) To: amd-gfx; +Cc: Mario Limonciello (AMD) When compiled without CONFIG_HSA_AMD KFD will return -ENOENT. As other errors will cause KFD functionality issues this is the only error code that should be ignored at init. Signed-off-by: Mario Limonciello (AMD) <superm1@kernel.org> --- drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c index 16adeba4d7e68..e804461e5f272 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c @@ -3169,8 +3169,10 @@ static int __init amdgpu_init(void) amdgpu_register_atpx_handler(); amdgpu_acpi_detect(); - /* Ignore KFD init failures. Normal when CONFIG_HSA_AMD is not set. */ - amdgpu_amdkfd_init(); + /* Ignore KFD init failures when CONFIG_HSA_AMD is not set. */ + r = amdgpu_amdkfd_init(); + if (r && r != -ENOENT) + goto error_fence; if (amdgpu_pp_feature_mask & PP_OVERDRIVE_MASK) { add_taint(TAINT_CPU_OUT_OF_SPEC, LOCKDEP_STILL_OK); -- 2.43.0 ^ permalink raw reply related [flat|nested] 4+ messages in thread
* RE: [PATCH 1/2] amdkfd: Only ignore -ENOENT for KFD init failuires 2025-12-07 14:04 ` [PATCH 1/2] amdkfd: Only ignore -ENOENT for KFD init failuires Mario Limonciello (AMD) @ 2025-12-10 17:06 ` Russell, Kent 0 siblings, 0 replies; 4+ messages in thread From: Russell, Kent @ 2025-12-10 17:06 UTC (permalink / raw) To: Mario Limonciello (AMD), amd-gfx@lists.freedesktop.org [AMD Official Use Only - AMD Internal Distribution Only] Reviewed-by: Kent Russell <kent.russell@amd.com> > -----Original Message----- > From: amd-gfx <amd-gfx-bounces@lists.freedesktop.org> On Behalf Of Mario > Limonciello (AMD) > Sent: Sunday, December 7, 2025 9:04 AM > To: amd-gfx@lists.freedesktop.org > Cc: Mario Limonciello (AMD) <superm1@kernel.org> > Subject: [PATCH 1/2] amdkfd: Only ignore -ENOENT for KFD init failuires > > When compiled without CONFIG_HSA_AMD KFD will return -ENOENT. > As other errors will cause KFD functionality issues this is the > only error code that should be ignored at init. > > Signed-off-by: Mario Limonciello (AMD) <superm1@kernel.org> > --- > drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c | 6 ++++-- > 1 file changed, 4 insertions(+), 2 deletions(-) > > diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c > b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c > index 16adeba4d7e68..e804461e5f272 100644 > --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c > +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c > @@ -3169,8 +3169,10 @@ static int __init amdgpu_init(void) > amdgpu_register_atpx_handler(); > amdgpu_acpi_detect(); > > - /* Ignore KFD init failures. Normal when CONFIG_HSA_AMD is not set. */ > - amdgpu_amdkfd_init(); > + /* Ignore KFD init failures when CONFIG_HSA_AMD is not set. */ > + r = amdgpu_amdkfd_init(); > + if (r && r != -ENOENT) > + goto error_fence; > > if (amdgpu_pp_feature_mask & PP_OVERDRIVE_MASK) { > add_taint(TAINT_CPU_OUT_OF_SPEC, LOCKDEP_STILL_OK); > -- > 2.43.0 ^ permalink raw reply [flat|nested] 4+ messages in thread
* [PATCH 2/2] amdkfd: Add device links between kfd device and amdgpu device 2025-12-07 14:04 [PATCH 0/2] Make device links between KFD and GPU device Mario Limonciello (AMD) 2025-12-07 14:04 ` [PATCH 1/2] amdkfd: Only ignore -ENOENT for KFD init failuires Mario Limonciello (AMD) @ 2025-12-07 14:04 ` Mario Limonciello (AMD) 1 sibling, 0 replies; 4+ messages in thread From: Mario Limonciello (AMD) @ 2025-12-07 14:04 UTC (permalink / raw) To: amd-gfx; +Cc: Mario Limonciello (AMD) Mapping out a KFD device to a GPU can be done manually by looking at the domain and location properties. To make it easier to discover which KFD device goes with what GPU add bidirectional links. Signed-off-by: Mario Limonciello (AMD) <superm1@kernel.org> --- drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c | 8 +++++ drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h | 1 + drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 4 +++ drivers/gpu/drm/amd/amdkfd/kfd_topology.c | 36 +++++++++++++++++++ .../gpu/drm/amd/include/kgd_kfd_interface.h | 2 ++ 5 files changed, 51 insertions(+) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c index a2879d2b7c8ec..5d6cf3adfa7b8 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c @@ -910,3 +910,11 @@ int amdgpu_amdkfd_config_sq_perfmon(struct amdgpu_device *adev, uint32_t xcp_id, return r; } + +int amdgpu_amdkfd_create_sysfs_links(struct amdgpu_device *adev) +{ + if (!adev->kfd.init_complete || !adev->kfd.dev) + return 0; + + return kgd2kfd_create_sysfs_links(adev->kfd.dev); +} diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h index 2fa5f1925f5a3..542f5bc2dd189 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h @@ -270,6 +270,7 @@ int amdgpu_amdkfd_stop_sched(struct amdgpu_device *adev, uint32_t node_id); int amdgpu_amdkfd_config_sq_perfmon(struct amdgpu_device *adev, uint32_t xcp_id, bool core_override_enable, bool reg_override_enable, bool perfmon_override_enable); bool amdgpu_amdkfd_compute_active(struct amdgpu_device *adev, uint32_t node_id); +int amdgpu_amdkfd_create_sysfs_links(struct amdgpu_device *adev); /* Read user wptr from a specified user address space with page fault diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c index 7a0213a07023d..44c9320d72a56 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c @@ -4947,6 +4947,10 @@ int amdgpu_device_init(struct amdgpu_device *adev, */ r = amdgpu_device_sys_interface_init(adev); + r = amdgpu_amdkfd_create_sysfs_links(adev); + if (r) + dev_err(adev->dev, "Failed to create KFD sysfs link: %d\n", r); + if (IS_ENABLED(CONFIG_PERF_EVENTS)) r = amdgpu_pmu_init(adev); if (r) diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_topology.c b/drivers/gpu/drm/amd/amdkfd/kfd_topology.c index 9c3e8f946a3d5..be673e35978eb 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_topology.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_topology.c @@ -79,6 +79,37 @@ struct kfd_topology_device *kfd_topology_device_by_proximity_domain( return device; } +int kgd2kfd_create_sysfs_links(struct kfd_dev *kfd) +{ + struct kfd_topology_device *top_dev; + int ret = -ENODEV; + + if (!kfd) + return -EINVAL; + + down_read(&topology_lock); + + list_for_each_entry(top_dev, &topology_device_list, list) { + struct kobject *amdgpu_kobj; + + if (!top_dev->gpu || top_dev->gpu->kfd != kfd || !top_dev->kobj_node) + continue; + + amdgpu_kobj = &top_dev->gpu->adev->dev->kobj; + ret = sysfs_create_link(top_dev->kobj_node, amdgpu_kobj, "device"); + if (ret) + break; + + ret = sysfs_create_link(amdgpu_kobj, top_dev->kobj_node, "kfd"); + if (ret) + sysfs_remove_link(top_dev->kobj_node, "device"); + break; + } + + up_read(&topology_lock); + return ret; +} + struct kfd_topology_device *kfd_topology_device_by_id(uint32_t gpu_id) { struct kfd_topology_device *top_dev = NULL; @@ -567,6 +598,11 @@ static void kfd_remove_sysfs_node_entry(struct kfd_topology_device *dev) struct kfd_mem_properties *mem; struct kfd_perf_properties *perf; + if (dev->gpu) { + sysfs_remove_link(dev->kobj_node, "device"); + sysfs_remove_link(&dev->gpu->adev->dev->kobj, "kfd"); + } + if (dev->kobj_iolink) { list_for_each_entry(iolink, &dev->io_link_props, list) if (iolink->kobj) { diff --git a/drivers/gpu/drm/amd/include/kgd_kfd_interface.h b/drivers/gpu/drm/amd/include/kgd_kfd_interface.h index 9aba8596faa7e..f6db1dc634399 100644 --- a/drivers/gpu/drm/amd/include/kgd_kfd_interface.h +++ b/drivers/gpu/drm/amd/include/kgd_kfd_interface.h @@ -335,4 +335,6 @@ struct kfd2kgd_calls { int engine, int queue); }; +int kgd2kfd_create_sysfs_links(struct kfd_dev *kfd); + #endif /* KGD_KFD_INTERFACE_H_INCLUDED */ -- 2.43.0 ^ permalink raw reply related [flat|nested] 4+ messages in thread
end of thread, other threads:[~2025-12-10 17:06 UTC | newest] Thread overview: 4+ messages (download: mbox.gz follow: Atom feed -- links below jump to the message on this page -- 2025-12-07 14:04 [PATCH 0/2] Make device links between KFD and GPU device Mario Limonciello (AMD) 2025-12-07 14:04 ` [PATCH 1/2] amdkfd: Only ignore -ENOENT for KFD init failuires Mario Limonciello (AMD) 2025-12-10 17:06 ` Russell, Kent 2025-12-07 14:04 ` [PATCH 2/2] amdkfd: Add device links between kfd device and amdgpu device Mario Limonciello (AMD)
This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox