On 24-Jan-26 2:21 AM, Alex Deucher wrote:
On Thu, Jan 22, 2026 at 5:52 AM Lijo Lazar <[email protected]> wrote:
Add cwsr parameters to userqueue ioctl. User should pass the GPU virtual
address for save/restore buffer, and size allocated. They are supported
only for user compute queues.
Signed-off-by: Lijo Lazar <[email protected]>
---
drivers/gpu/drm/amd/amdgpu/mes_userqueue.c | 13 +++++++++----
include/uapi/drm/amdgpu_drm.h | 16 ++++++++++++++++
2 files changed, 25 insertions(+), 4 deletions(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/mes_userqueue.c
b/drivers/gpu/drm/amd/amdgpu/mes_userqueue.c
index 7ad8297eb0d8..2765317f04df 100644
--- a/drivers/gpu/drm/amd/amdgpu/mes_userqueue.c
+++ b/drivers/gpu/drm/amd/amdgpu/mes_userqueue.c
@@ -343,16 +343,21 @@ static int mes_userq_mqd_create(struct
amdgpu_usermode_queue *queue,
if (amdgpu_cwsr_is_enabled(adev)) {
cwsr_params.ctx_save_area_address =
- userq_props->ctx_save_area_addr;
- cwsr_params.cwsr_sz = userq_props->ctx_save_area_size;
- cwsr_params.ctl_stack_sz = userq_props->ctl_stack_size;
-
+ compute_mqd->ctx_save_area_va;
+ cwsr_params.cwsr_sz = compute_mqd->ctx_save_area_size;
+ cwsr_params.ctl_stack_sz = compute_mqd->ctl_stack_size;
r = amdgpu_userq_input_cwsr_params_validate(
queue, &cwsr_params);
if (r) {
kfree(compute_mqd);
goto free_mqd;
}
+ userq_props->ctx_save_area_addr =
+ compute_mqd->ctx_save_area_va;
+ userq_props->ctx_save_area_size =
+ compute_mqd->ctx_save_area_size;
+ userq_props->ctl_stack_size =
+ compute_mqd->ctl_stack_size;
}
kfree(compute_mqd);
diff --git a/include/uapi/drm/amdgpu_drm.h b/include/uapi/drm/amdgpu_drm.h
index c178b8e0bd3f..b7a858365174 100644
--- a/include/uapi/drm/amdgpu_drm.h
+++ b/include/uapi/drm/amdgpu_drm.h
@@ -460,6 +460,22 @@ struct drm_amdgpu_userq_mqd_compute_gfx11 {
* to get the size.
*/
__u64 eop_va;
+ /**
+ * @ctx_save_area_va: Virtual address of the GPU memory for
save/restore buffer.
+ * This must be from a separate GPU object, and use AMDGPU_INFO IOCTL
+ * to get the size. This includes control stack, wave context and
debugger memory.
+ */
+ __u64 ctx_save_area_va;
+ /**
+ * @ctx_save_area_size: Total size (in bytes) allocated for
save/restore buffer.
+ * Use AMDGPU_INFO IOCTL to get the size.
+ */
+ __u32 ctx_save_area_size;
+ /**
+ * @ctl_stack_size: Size (in bytes) of control stack region in the
save/restore buffer.
+ * Use AMDGPU_INFO IOCTL to get the size.
+ */
+ __u32 ctl_stack_size;
Does it matter where the ctl_stack is within the save area?
This is the legacy way. Probably, this can be avoided. Adding David and
Lancelot.
Hi David/Lancelot,
Do you have the background of userspace passing back control stack size?
https://github.com/torvalds/linux/blob/master/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c#L260
Can driver assume that context save area takes care of everything and
assume that user allotted as per the right control stack size?
Thanks,
Lijo
Alex
};
/* userq signal/wait ioctl */
--
2.49.0
