AMD General
kfd_mqd_manager_v12.c (GFX 12.0) and kfd_mqd_manager_v12_1.c (GFX 12.1)
initialize their MQD managers without assigning restore_mqd or checkpoint_mqd
callbacks, leaving both function pointers NULL.
Both create_queue_cpsch() and create_queue_nocpsch() in
kfd_device_queue_manager.c unconditionally call mqd_mgr->restore_mqd() when
a CRIU restore is in progress (qd != NULL), with no NULL guard. On any system
with a GFX 12.x GPU, a user holding CAP_CHECKPOINT_RESTORE can trigger a
kernel NULL pointer dereference and panic the machine by issuing
KFD_IOC_CRIU_OP_RESTORE with a crafted queue restore object.
Fix this by:
1. Implementing restore_mqd and checkpoint_mqd callbacks for both compute
and SDMA queues in kfd_mqd_manager_v12.c and kfd_mqd_manager_v12_1.c,
modeled after the GFX 11 implementation.
2. Adding a NULL guard at both call sites in kfd_device_queue_manager.c
as a defensive measure against future regressions.
Fixes: 48f0bdf4e38e ("drm/amdkfd: Added MQD manager files for GFX12.")
Fixes: 01bbc4a4b947 ("drm/amdkfd: Add MQD manager for GFX 12.1.0")
Signed-off-by: Vladimir Marioukhine <[email protected]>
---
drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c | 10 ++++--
drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v12.c | 69
+++++++++++++++++++++++++++++++++++++++
drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v12_1.c | 69
++++++++++++++++++++++++++++++++++++++
3 files changed, 146 insertions(+), 2 deletions(-)
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c
b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c
index 51ee9c39104b..31bf479212e5 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c
@@ -794,11 +794,16 @@ static int create_queue_nocpsch(struct
device_queue_manager *dqm,
goto out_deallocate_doorbell;
}
- if (qd)
+ if (qd) {
+ if (!mqd_mgr->restore_mqd) {
+ pr_err("amdkfd: restore_mqd not implemented for this
GPU\n");
+ retval = -EINVAL;
+ goto out_deallocate_doorbell;
+ }
mqd_mgr->restore_mqd(mqd_mgr, &q->mqd, q->mqd_mem_obj,
&q->gart_mqd_addr,
&q->properties, restore_mqd,
restore_ctl_stack,
qd->ctl_stack_size);
- else
+ } else
mqd_mgr->init_mqd(mqd_mgr, &q->mqd, q->mqd_mem_obj,
&q->gart_mqd_addr, &q->properties);
@@ -2153,11 +2158,16 @@ static int create_queue_cpsch(struct
device_queue_manager *dqm, struct queue *q,
q->properties.is_dbg_wa = qpd->pqm->process->debug_trap_enabled &&
kfd_dbg_has_cwsr_workaround(q->device);
- if (qd)
+ if (qd) {
+ if (!mqd_mgr->restore_mqd) {
+ pr_err("amdkfd: restore_mqd not implemented for this
GPU\n");
+ retval = -EINVAL;
+ goto out;
+ }
mqd_mgr->restore_mqd(mqd_mgr, &q->mqd, q->mqd_mem_obj,
&q->gart_mqd_addr,
&q->properties, restore_mqd,
restore_ctl_stack,
qd->ctl_stack_size);
- else
+ } else
mqd_mgr->init_mqd(mqd_mgr, &q->mqd, q->mqd_mem_obj,
&q->gart_mqd_addr, &q->properties);
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v12.c
b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v12.c
index 7c387fa90076..c6a8e99ad069 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v12.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v12.c
@@ -380,6 +380,75 @@ static int debugfs_show_mqd_sdma(struct seq_file *m, void
*data)
#endif
+static void checkpoint_mqd(struct mqd_manager *mm, void *mqd,
+ void *mqd_dst, void *ctl_stack_dst)
+{
+ struct v12_compute_mqd *m;
+
+ m = get_mqd(mqd);
+
+ memcpy(mqd_dst, m, sizeof(struct v12_compute_mqd));
+}
+
+static void restore_mqd(struct mqd_manager *mm, void **mqd,
+ struct kfd_mem_obj *mqd_mem_obj, uint64_t *gart_addr,
+ struct queue_properties *qp, const void *mqd_src,
+ const void *ctl_stack_src, const u32 ctl_stack_size)
+{
+ uint64_t addr;
+ struct v12_compute_mqd *m;
+
+ m = (struct v12_compute_mqd *) mqd_mem_obj->cpu_ptr;
+ addr = mqd_mem_obj->gpu_addr;
+
+ memcpy(m, mqd_src, sizeof(*m));
+
+ *mqd = m;
+ if (gart_addr)
+ *gart_addr = addr;
+
+ m->cp_hqd_pq_doorbell_control =
+ qp->doorbell_off <<
CP_HQD_PQ_DOORBELL_CONTROL__DOORBELL_OFFSET__SHIFT;
+ pr_debug("cp_hqd_pq_doorbell_control 0x%x\n",
m->cp_hqd_pq_doorbell_control);
+
+ qp->is_active = 0;
+}
+
+static void checkpoint_mqd_sdma(struct mqd_manager *mm, void *mqd,
+ void *mqd_dst, void *ctl_stack_dst)
+{
+ struct v12_sdma_mqd *m;
+
+ m = get_sdma_mqd(mqd);
+
+ memcpy(mqd_dst, m, sizeof(struct v12_sdma_mqd));
+}
+
+static void restore_mqd_sdma(struct mqd_manager *mm, void **mqd,
+ struct kfd_mem_obj *mqd_mem_obj, uint64_t
*gart_addr,
+ struct queue_properties *qp,
+ const void *mqd_src,
+ const void *ctl_stack_src,
+ const u32 ctl_stack_size)
+{
+ uint64_t addr;
+ struct v12_sdma_mqd *m;
+
+ m = (struct v12_sdma_mqd *) mqd_mem_obj->cpu_ptr;
+ addr = mqd_mem_obj->gpu_addr;
+
+ memcpy(m, mqd_src, sizeof(*m));
+
+ m->sdmax_rlcx_doorbell_offset =
+ qp->doorbell_off << SDMA0_QUEUE0_DOORBELL_OFFSET__OFFSET__SHIFT;
+
+ *mqd = m;
+ if (gart_addr)
+ *gart_addr = addr;
+
+ qp->is_active = 0;
+}
+
struct mqd_manager *mqd_manager_init_v12(enum KFD_MQD_TYPE type,
struct kfd_node *dev)
{
@@ -407,6 +476,8 @@ struct mqd_manager *mqd_manager_init_v12(enum KFD_MQD_TYPE
type,
mqd->mqd_size = sizeof(struct v12_compute_mqd);
mqd->get_wave_state = get_wave_state;
mqd->mqd_stride = kfd_mqd_stride;
+ mqd->checkpoint_mqd = checkpoint_mqd;
+ mqd->restore_mqd = restore_mqd;
#if defined(CONFIG_DEBUG_FS)
mqd->debugfs_show_mqd = debugfs_show_mqd;
#endif
@@ -453,6 +524,8 @@ struct mqd_manager *mqd_manager_init_v12(enum KFD_MQD_TYPE
type,
mqd->is_occupied = kfd_is_occupied_sdma;
mqd->mqd_size = sizeof(struct v12_sdma_mqd);
mqd->mqd_stride = kfd_mqd_stride;
+ mqd->checkpoint_mqd = checkpoint_mqd_sdma;
+ mqd->restore_mqd = restore_mqd_sdma;
#if defined(CONFIG_DEBUG_FS)
mqd->debugfs_show_mqd = debugfs_show_mqd_sdma;
#endif
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v12_1.c
b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v12_1.c
index 431a940f91f3..7ee98ef084ea 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v12_1.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v12_1.c
@@ -641,6 +641,75 @@ static int debugfs_show_mqd_sdma(struct seq_file *m, void
*data)
#endif
+static void checkpoint_mqd_v12_1(struct mqd_manager *mm, void *mqd,
+ void *mqd_dst, void *ctl_stack_dst)
+{
+ struct v12_1_compute_mqd *m;
+
+ m = get_mqd(mqd);
+
+ memcpy(mqd_dst, m, sizeof(struct v12_1_compute_mqd));
+}
+
+static void restore_mqd_v12_1(struct mqd_manager *mm, void **mqd,
+ struct kfd_mem_obj *mqd_mem_obj, uint64_t
*gart_addr,
+ struct queue_properties *qp, const void *mqd_src,
+ const void *ctl_stack_src, const u32
ctl_stack_size)
+{
+ uint64_t addr;
+ struct v12_1_compute_mqd *m;
+
+ m = (struct v12_1_compute_mqd *) mqd_mem_obj->cpu_ptr;
+ addr = mqd_mem_obj->gpu_addr;
+
+ memcpy(m, mqd_src, sizeof(*m));
+
+ *mqd = m;
+ if (gart_addr)
+ *gart_addr = addr;
+
+ m->cp_hqd_pq_doorbell_control =
+ qp->doorbell_off <<
CP_HQD_PQ_DOORBELL_CONTROL__DOORBELL_OFFSET__SHIFT;
+ pr_debug("cp_hqd_pq_doorbell_control 0x%x\n",
m->cp_hqd_pq_doorbell_control);
+
+ qp->is_active = 0;
+}
+
+static void checkpoint_mqd_sdma_v12_1(struct mqd_manager *mm, void *mqd,
+ void *mqd_dst, void *ctl_stack_dst)
+{
+ struct v12_sdma_mqd *m;
+
+ m = get_sdma_mqd(mqd);
+
+ memcpy(mqd_dst, m, sizeof(struct v12_sdma_mqd));
+}
+
+static void restore_mqd_sdma_v12_1(struct mqd_manager *mm, void **mqd,
+ struct kfd_mem_obj *mqd_mem_obj, uint64_t
*gart_addr,
+ struct queue_properties *qp,
+ const void *mqd_src,
+ const void *ctl_stack_src,
+ const u32 ctl_stack_size)
+{
+ uint64_t addr;
+ struct v12_sdma_mqd *m;
+
+ m = (struct v12_sdma_mqd *) mqd_mem_obj->cpu_ptr;
+ addr = mqd_mem_obj->gpu_addr;
+
+ memcpy(m, mqd_src, sizeof(*m));
+
+ m->sdmax_rlcx_doorbell_offset =
+ qp->doorbell_off << SDMA0_QUEUE0_DOORBELL_OFFSET__OFFSET__SHIFT;
+
+ *mqd = m;
+ if (gart_addr)
+ *gart_addr = addr;
+
+ qp->is_active = 0;
+}
+
struct mqd_manager *mqd_manager_init_v12_1(enum KFD_MQD_TYPE type,
struct kfd_node *dev)
{
@@ -668,6 +737,8 @@ struct mqd_manager *mqd_manager_init_v12_1(enum
KFD_MQD_TYPE type,
mqd->mqd_size = sizeof(struct v12_1_compute_mqd);
mqd->get_wave_state = get_wave_state_v12_1;
mqd->mqd_stride = kfd_mqd_stride;
+ mqd->checkpoint_mqd = checkpoint_mqd_v12_1;
+ mqd->restore_mqd = restore_mqd_v12_1;
#if defined(CONFIG_DEBUG_FS)
mqd->debugfs_show_mqd = debugfs_show_mqd;
#endif
@@ -714,6 +785,8 @@ struct mqd_manager *mqd_manager_init_v12_1(enum
KFD_MQD_TYPE type,
mqd->is_occupied = kfd_is_occupied_sdma;
mqd->mqd_size = sizeof(struct v12_sdma_mqd);
mqd->mqd_stride = kfd_mqd_stride;
+ mqd->checkpoint_mqd = checkpoint_mqd_sdma_v12_1;
+ mqd->restore_mqd = restore_mqd_sdma_v12_1;
#if defined(CONFIG_DEBUG_FS)
mqd->debugfs_show_mqd = debugfs_show_mqd_sdma;
#endif