drm/amdkfd: Adding new IOCTL for scratch memory v2

v2:
* Renamed ALLOC_MEMORY_OF_SCRATCH to SET_SCRATCH_BACKING_VA
* Removed size parameter from the ioctl, it was unused
* Removed hole in ioctl number space
* No more call to write_config_static_mem
* Return correct error code from ioctl

Signed-off-by: Moses Reuben <moses.reuben@amd.com>
Signed-off-by: Ben Goz <ben.goz@amd.com>
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>

diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c
index 65b506f19b46275afc4d6b194885b967d32e248d..7436d34b77abd07a0e5fa35eedb5710422309aba 100644 (file)
--- a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c
@@ -848,6 +848,40 @@ static int kfd_ioctl_wait_events(struct file *filp, struct kfd_process *p,
 
        return err;
 }
+static int kfd_ioctl_set_scratch_backing_va(struct file *filep,
+                                       struct kfd_process *p, void *data)
+{
+       struct kfd_ioctl_set_scratch_backing_va_args *args = data;
+       struct kfd_process_device *pdd;
+       struct kfd_dev *dev;
+       long err;
+
+       dev = kfd_device_by_id(args->gpu_id);
+       if (!dev)
+               return -EINVAL;
+
+       mutex_lock(&p->mutex);
+
+       pdd = kfd_bind_process_to_device(dev, p);
+       if (IS_ERR(pdd)) {
+               err = PTR_ERR(pdd);
+               goto bind_process_to_device_fail;
+       }
+
+       pdd->qpd.sh_hidden_private_base = args->va_addr;
+
+       mutex_unlock(&p->mutex);
+
+       if (sched_policy == KFD_SCHED_POLICY_NO_HWS && pdd->qpd.vmid != 0)
+               dev->kfd2kgd->set_scratch_backing_va(
+                       dev->kgd, args->va_addr, pdd->qpd.vmid);
+
+       return 0;
+
+bind_process_to_device_fail:
+       mutex_unlock(&p->mutex);
+       return err;
+}
 
 #define AMDKFD_IOCTL_DEF(ioctl, _func, _flags) \
        [_IOC_NR(ioctl)] = {.cmd = ioctl, .func = _func, .flags = _flags, \
@@ -902,6 +936,9 @@ static const struct amdkfd_ioctl_desc amdkfd_ioctls[] = {
 
        AMDKFD_IOCTL_DEF(AMDKFD_IOC_DBG_WAVE_CONTROL,
                        kfd_ioctl_dbg_wave_control, 0),
+
+       AMDKFD_IOCTL_DEF(AMDKFD_IOC_SET_SCRATCH_BACKING_VA,
+                       kfd_ioctl_set_scratch_backing_va, 0),
 };
 
 #define AMDKFD_CORE_IOCTL_COUNT        ARRAY_SIZE(amdkfd_ioctls)

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 618ac65b61367d7248411b9c1a1e01a30e7147ef..53a66e8216245fb601c7800c8f5d6663228cb767 100644 (file)
--- a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c
@@ -270,6 +270,9 @@ static int create_compute_queue_nocpsch(struct device_queue_manager *dqm,
        pr_debug("Loading mqd to hqd on pipe %d, queue %d\n",
                        q->pipe, q->queue);
 
+       dqm->dev->kfd2kgd->set_scratch_backing_va(
+                       dqm->dev->kgd, qpd->sh_hidden_private_base, qpd->vmid);
+
        retval = mqd->load_mqd(mqd, q->mqd, q->pipe, q->queue, &q->properties,
                               q->process->mm);
        if (retval)

diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager_cik.c b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager_cik.c
index fadc56a8be711140e72387b635bbbd14176bb545..72c3cbabc0a7c6b6b591c3a563a7c66677c28b18 100644 (file)
--- a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager_cik.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager_cik.c
@@ -24,6 +24,7 @@
 #include "kfd_device_queue_manager.h"
 #include "cik_regs.h"
 #include "oss/oss_2_4_sh_mask.h"
+#include "gca/gfx_7_2_sh_mask.h"
 
 static bool set_cache_memory_policy_cik(struct device_queue_manager *dqm,
                                   struct qcm_process_device *qpd,
@@ -123,6 +124,7 @@ static int register_process_cik(struct device_queue_manager *dqm,
        } else {
                temp = get_sh_mem_bases_nybble_64(pdd);
                qpd->sh_mem_bases = compute_sh_mem_bases_64bit(temp);
+               qpd->sh_mem_config |= 1  << SH_MEM_CONFIG__PRIVATE_ATC__SHIFT;
        }
 
        pr_debug("is32bit process: %d sh_mem_bases nybble: 0x%X and register 0x%X\n",

diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager_vi.c b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager_vi.c
index 15e81ae9d2f4f940f0cbb28275e1eafc7eb73099..40e9ddd096cdf7294ad633d8bb021b9a5d0c1a1f 100644 (file)
--- a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager_vi.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager_vi.c
@@ -135,6 +135,8 @@ static int register_process_vi(struct device_queue_manager *dqm,
                qpd->sh_mem_bases = compute_sh_mem_bases_64bit(temp);
                qpd->sh_mem_config |= SH_MEM_ADDRESS_MODE_HSA64 <<
                        SH_MEM_CONFIG__ADDRESS_MODE__SHIFT;
+               qpd->sh_mem_config |= 1  <<
+                       SH_MEM_CONFIG__PRIVATE_ATC__SHIFT;
        }
 
        pr_debug("is32bit process: %d sh_mem_bases nybble: 0x%X and register 0x%X\n",

diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
index 30ce92c6e6a1a77daed7eee9eb7f9c88220b5294..b397ec726400c2a52533356363b36bfd275dfb39 100644 (file)
--- a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
@@ -432,6 +432,7 @@ struct qcm_process_device {
        uint32_t gds_size;
        uint32_t num_gws;
        uint32_t num_oac;
+       uint32_t sh_hidden_private_base;
 };
 
 /* Data that is per-process-per device. */

diff --git a/include/uapi/linux/kfd_ioctl.h b/include/uapi/linux/kfd_ioctl.h
index d6833426fdef15479a0c2f3fe6145a6d5b8b2e15..1b9c5609d5233221adcffe8f67d17fad53f5290b 100644 (file)
--- a/include/uapi/linux/kfd_ioctl.h
+++ b/include/uapi/linux/kfd_ioctl.h
@@ -232,6 +232,12 @@ struct kfd_ioctl_wait_events_args {
        uint32_t wait_result;           /* from KFD */
 };
 
+struct kfd_ioctl_set_scratch_backing_va_args {
+       uint64_t va_addr;       /* to KFD */
+       uint32_t gpu_id;        /* to KFD */
+       uint32_t pad;
+};
+
 #define AMDKFD_IOCTL_BASE 'K'
 #define AMDKFD_IO(nr)                  _IO(AMDKFD_IOCTL_BASE, nr)
 #define AMDKFD_IOR(nr, type)           _IOR(AMDKFD_IOCTL_BASE, nr, type)
@@ -286,7 +292,10 @@ struct kfd_ioctl_wait_events_args {
 #define AMDKFD_IOC_DBG_WAVE_CONTROL            \
                AMDKFD_IOW(0x10, struct kfd_ioctl_dbg_wave_control_args)
 
+#define AMDKFD_IOC_SET_SCRATCH_BACKING_VA      \
+               AMDKFD_IOWR(0x11, struct kfd_ioctl_set_scratch_backing_va_args)
+
 #define AMDKFD_COMMAND_START           0x01
-#define AMDKFD_COMMAND_END             0x11
+#define AMDKFD_COMMAND_END             0x12
 
 #endif