[PATCH v3] drm/amdgpu: Add debugfs entry for printing VM info
Patel, Mihir
Mihir.Patel at amd.com
Mon Oct 12 09:37:35 UTC 2020
[AMD Official Use Only - Internal Distribution Only]
-----Original Message-----
From: Koenig, Christian <Christian.Koenig at amd.com>
Sent: Monday, October 12, 2020 2:41 PM
To: Patel, Mihir <Mihir.Patel at amd.com>; amd-gfx at lists.freedesktop.org
Cc: Chauhan, Madhav <Madhav.Chauhan at amd.com>; Deucher, Alexander <Alexander.Deucher at amd.com>; Surampalli, Kishore <Kishore.Surampalli at amd.com>; Kamliya, Prakash <Prakash.Kamliya at amd.com>
Subject: Re: [PATCH v3] drm/amdgpu: Add debugfs entry for printing VM info
Am 12.10.20 um 11:01 schrieb Mihir Patel:
> From: Mihir Bhogilal Patel <Mihir.Patel at amd.com>
>
> Create new debugfs entry to print memory info using VM buffer objects.
>
> Pending:
> - Consolidated memory utilization info like total, swap etc.
>
> V2: Added Common function for printing BO info.
> Dump more VM lists for evicted, moved, relocated, invalidated.
> Removed dumping VM mapped BOs.
> V3: Fixed coding style comments, renamed print API and variables.
>
> Signed-off-by: Mihir Bhogilal Patel <Mihir.Patel at amd.com>
> ---
> drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c | 30 +++++++
> drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c | 69 ++--------------
> drivers/gpu/drm/amd/amdgpu/amdgpu_object.c | 74 +++++++++++++++++
> drivers/gpu/drm/amd/amdgpu/amdgpu_object.h | 1 +
> drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 88 +++++++++++++++++++++
> drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h | 4 +
> 6 files changed, 204 insertions(+), 62 deletions(-)
>
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c
> b/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c
> index 2d125b8b15ee..0b41b8b72ba3 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c
> @@ -1335,11 +1335,41 @@ static int amdgpu_debugfs_evict_gtt(struct seq_file *m, void *data)
> return 0;
> }
>
> +static int amdgpu_debugfs_vm_info(struct seq_file *m, void *data) {
> + struct drm_info_node *node = (struct drm_info_node *)m->private;
> + struct drm_device *dev = node->minor->dev;
> + struct drm_file *file;
> + int r;
> +
> + r = mutex_lock_interruptible(&dev->filelist_mutex);
> + if (r)
> + return r;
> +
> + list_for_each_entry(file, &dev->filelist, lhead) {
> + struct amdgpu_fpriv *fpriv = file->driver_priv;
> + struct amdgpu_vm *vm = &fpriv->vm;
> +
> + seq_printf(m, "pid:%d\tProcess:%s ----------\n",
> + vm->task_info.pid, vm->task_info.process_name);
> + r = amdgpu_bo_reserve(vm->root.base.bo, true);
> + if (r)
> + continue;
I would rather say break here and return r from the function as well.
When reading this debugfs file is aborted using ctrl+c we would otherwise just continue with the loop.
> + amdgpu_debugfs_vm_bo_info(vm, m);
> + amdgpu_bo_unreserve(vm->root.base.bo);
> + }
> +
> + mutex_unlock(&dev->filelist_mutex);
> +
> + return 0;
> +}
> +
> static const struct drm_info_list amdgpu_debugfs_list[] = {
> {"amdgpu_vbios", amdgpu_debugfs_get_vbios_dump},
> {"amdgpu_test_ib", &amdgpu_debugfs_test_ib},
> {"amdgpu_evict_vram", &amdgpu_debugfs_evict_vram},
> {"amdgpu_evict_gtt", &amdgpu_debugfs_evict_gtt},
> + {"amdgpu_vm_info", &amdgpu_debugfs_vm_info},
> };
>
> static void amdgpu_ib_preempt_fences_swap(struct amdgpu_ring *ring,
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
> b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
> index f4c2e2e75b8f..6197c5ce744c 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
> @@ -826,67 +826,6 @@ int amdgpu_mode_dumb_create(struct drm_file *file_priv,
> }
>
> #if defined(CONFIG_DEBUG_FS)
> -
> -#define amdgpu_debugfs_gem_bo_print_flag(m, bo, flag) \
> - if (bo->flags & (AMDGPU_GEM_CREATE_ ## flag)) { \
> - seq_printf((m), " " #flag); \
> - }
> -
> -static int amdgpu_debugfs_gem_bo_info(int id, void *ptr, void *data)
> -{
> - struct drm_gem_object *gobj = ptr;
> - struct amdgpu_bo *bo = gem_to_amdgpu_bo(gobj);
> - struct seq_file *m = data;
> -
> - struct dma_buf_attachment *attachment;
> - struct dma_buf *dma_buf;
> - unsigned domain;
> - const char *placement;
> - unsigned pin_count;
> -
> - domain = amdgpu_mem_type_to_domain(bo->tbo.mem.mem_type);
> - switch (domain) {
> - case AMDGPU_GEM_DOMAIN_VRAM:
> - placement = "VRAM";
> - break;
> - case AMDGPU_GEM_DOMAIN_GTT:
> - placement = " GTT";
> - break;
> - case AMDGPU_GEM_DOMAIN_CPU:
> - default:
> - placement = " CPU";
> - break;
> - }
> - seq_printf(m, "\t0x%08x: %12ld byte %s",
> - id, amdgpu_bo_size(bo), placement);
> -
> - pin_count = READ_ONCE(bo->pin_count);
> - if (pin_count)
> - seq_printf(m, " pin count %d", pin_count);
> -
> - dma_buf = READ_ONCE(bo->tbo.base.dma_buf);
> - attachment = READ_ONCE(bo->tbo.base.import_attach);
> -
> - if (attachment)
> - seq_printf(m, " imported from %p%s", dma_buf,
> - attachment->peer2peer ? " P2P" : "");
> - else if (dma_buf)
> - seq_printf(m, " exported as %p", dma_buf);
> -
> - amdgpu_debugfs_gem_bo_print_flag(m, bo, CPU_ACCESS_REQUIRED);
> - amdgpu_debugfs_gem_bo_print_flag(m, bo, NO_CPU_ACCESS);
> - amdgpu_debugfs_gem_bo_print_flag(m, bo, CPU_GTT_USWC);
> - amdgpu_debugfs_gem_bo_print_flag(m, bo, VRAM_CLEARED);
> - amdgpu_debugfs_gem_bo_print_flag(m, bo, SHADOW);
> - amdgpu_debugfs_gem_bo_print_flag(m, bo, VRAM_CONTIGUOUS);
> - amdgpu_debugfs_gem_bo_print_flag(m, bo, VM_ALWAYS_VALID);
> - amdgpu_debugfs_gem_bo_print_flag(m, bo, EXPLICIT_SYNC);
> -
> - seq_printf(m, "\n");
> -
> - return 0;
> -}
> -
> static int amdgpu_debugfs_gem_info(struct seq_file *m, void *data)
> {
> struct drm_info_node *node = (struct drm_info_node *)m->private; @@
> -900,6 +839,8 @@ static int amdgpu_debugfs_gem_info(struct seq_file
> *m, void *data)
>
> list_for_each_entry(file, &dev->filelist, lhead) {
> struct task_struct *task;
> + struct drm_gem_object *gobj;
> + int id;
>
> /*
> * Although we have a valid reference on file->pid, that does @@
> -914,7 +855,11 @@ static int amdgpu_debugfs_gem_info(struct seq_file *m, void *data)
> rcu_read_unlock();
>
> spin_lock(&file->table_lock);
> - idr_for_each(&file->object_idr, amdgpu_debugfs_gem_bo_info, m);
> + idr_for_each_entry(&file->object_idr, gobj, id) {
> + struct amdgpu_bo *bo = gem_to_amdgpu_bo(gobj);
> +
> + amdgpu_bo_print_info(id, bo, m);
> + }
> spin_unlock(&file->table_lock);
> }
>
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
> b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
> index 2ce79bccfc30..f4142b7a3c59 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
> @@ -1528,3 +1528,77 @@ uint32_t amdgpu_bo_get_preferred_pin_domain(struct amdgpu_device *adev,
> }
> return domain;
> }
> +
> +#if defined(CONFIG_DEBUG_FS)
> +#define amdgpu_bo_print_flag(m, bo, flag) \
> + do { \
> + if (bo->flags & (AMDGPU_GEM_CREATE_ ## flag)) { \
> + seq_printf((m), " " #flag); \
> + } \
> + } while (0)
> +
> +/**
> + * amdgpu_debugfs_print_bo_info - print BO info in debugfs file
> + *
> + * @id: Index or Id of the BO
> + * @bo: Requested BO for printing info
> + * @data: debugfs file
> + *
> + * Print BO information in debugfs file
> + *
> + * Returns:
> + * Size of the BO in bytes.
> + */
> +unsigned long amdgpu_bo_print_info(int id, struct amdgpu_bo *bo,
> +struct seq_file *m)
Please use uint64_t or u64 for the return type.
> +{
> + struct dma_buf_attachment *attachment;
> + struct dma_buf *dma_buf;
> + unsigned int domain;
> + const char *placement;
> + unsigned int pin_count;
> + unsigned long size;
Same here of course.
> +
> + domain = amdgpu_mem_type_to_domain(bo->tbo.mem.mem_type);
> + switch (domain) {
> + case AMDGPU_GEM_DOMAIN_VRAM:
> + placement = "VRAM";
> + break;
> + case AMDGPU_GEM_DOMAIN_GTT:
> + placement = " GTT";
> + break;
> + case AMDGPU_GEM_DOMAIN_CPU:
> + default:
> + placement = " CPU";
> + break;
> + }
> +
> + size = amdgpu_bo_size(bo);
> + seq_printf(m, "\t\t0x%08x: %12ld byte %s",
> + id++, size, placement);
Incrementing id here is pointless.
Maybe move this outside the function as well, but this is not a hard requirement.
> +
> + pin_count = READ_ONCE(bo->pin_count);
> + if (pin_count)
> + seq_printf(m, " pin count %d", pin_count);
> +
> + dma_buf = READ_ONCE(bo->tbo.base.dma_buf);
> + attachment = READ_ONCE(bo->tbo.base.import_attach);
> +
> + if (attachment)
> + seq_printf(m, " imported from %p", dma_buf);
> + else if (dma_buf)
> + seq_printf(m, " exported as %p", dma_buf);
> +
> + amdgpu_bo_print_flag(m, bo, CPU_ACCESS_REQUIRED);
> + amdgpu_bo_print_flag(m, bo, NO_CPU_ACCESS);
> + amdgpu_bo_print_flag(m, bo, CPU_GTT_USWC);
> + amdgpu_bo_print_flag(m, bo, VRAM_CLEARED);
> + amdgpu_bo_print_flag(m, bo, SHADOW);
> + amdgpu_bo_print_flag(m, bo, VRAM_CONTIGUOUS);
> + amdgpu_bo_print_flag(m, bo, VM_ALWAYS_VALID);
> + amdgpu_bo_print_flag(m, bo, EXPLICIT_SYNC);
> +
> + seq_puts(m, "\n");
> +
> + return size;
> +}
> +#endif
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h
> b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h
> index afa5189dba7d..06fbff49958d 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h
> @@ -330,6 +330,7 @@ void amdgpu_sa_bo_free(struct amdgpu_device *adev,
> #if defined(CONFIG_DEBUG_FS)
> void amdgpu_sa_bo_dump_debug_info(struct amdgpu_sa_manager *sa_manager,
> struct seq_file *m);
> +unsigned long amdgpu_bo_print_info(int id, struct amdgpu_bo *bo,
> +struct seq_file *m);
> #endif
> int amdgpu_debugfs_sa_init(struct amdgpu_device *adev);
>
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
> b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
> index 3cd949aad500..0e1cb399e508 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
> @@ -3392,3 +3392,91 @@ bool amdgpu_vm_handle_fault(struct
> amdgpu_device *adev, unsigned int pasid,
>
> return false;
> }
> +
> +#if defined(CONFIG_DEBUG_FS)
> +/**
> + * amdgpu_debugfs_vm_bo_info - print BO info for the VM
> + *
> + * @vm: Requested VM for printing BO info
> + * @data: debugfs file
> + *
> + * Print BO information in debugfs file for the VM */ void
> +amdgpu_debugfs_vm_bo_info(struct amdgpu_vm *vm, struct seq_file *m) {
> + struct amdgpu_bo_va *bo_va, *tmp;
> + u64 total_idle = 0;
> + u64 total_evicted = 0;
> + u64 total_relocated = 0;
> + u64 total_moved = 0;
> + u64 total_invalidated = 0;
> + unsigned int total_idle_objs = 0;
> + unsigned int total_evicted_objs = 0;
> + unsigned int total_relocated_objs = 0;
> + unsigned int total_moved_objs = 0;
> + unsigned int total_invalidated_objs = 0;
> + unsigned int id = 0;
> +
> + /* Print info for Idle BOs */
I would drop those comments, they just duplicate what the printed text says anyway.
> + seq_puts(m, "\tIdle BOs:\n");
> + list_for_each_entry_safe(bo_va, tmp, &vm->idle, base.vm_status) {
> + if (!bo_va->base.bo)
> + continue;
> + total_idle += amdgpu_bo_print_info(id++, bo_va->base.bo, m);
> + }
> + total_idle_objs = id;
> + id = 0;
> +
> + /* Print info for Evicted BOs */
> + seq_puts(m, "\tEvicted BOs:\n");
> + list_for_each_entry_safe(bo_va, tmp, &vm->evicted, base.vm_status) {
> + if (!bo_va->base.bo)
> + continue;
> + total_evicted += amdgpu_bo_print_info(id++, bo_va->base.bo, m);
> + }
> + total_evicted_objs = id;
> + id = 0;
> +
> + /* Print info for Relocated BOs */
> + seq_puts(m, "\tRelocated BOs:\n");
> + list_for_each_entry_safe(bo_va, tmp, &vm->relocated, base.vm_status) {
> + if (!bo_va->base.bo)
> + continue;
> + total_relocated += amdgpu_bo_print_info(id++, bo_va->base.bo, m);
> + }
> + total_relocated_objs = id;
> + id = 0;
> +
> + /* Print info for Moved BOs */
> + seq_puts(m, "\tMoved BOs:\n");
> + list_for_each_entry_safe(bo_va, tmp, &vm->moved, base.vm_status) {
> + if (!bo_va->base.bo)
> + continue;
> + total_moved += amdgpu_bo_print_info(id++, bo_va->base.bo, m);
> + }
> + total_moved_objs = id;
> + id = 0;
> +
> + /* Print info for Invalidated BOs */
> + seq_puts(m, "\tInvalidated BOs:\n");
> + spin_lock(&vm->invalidated_lock);
> + list_for_each_entry_safe(bo_va, tmp, &vm->invalidated, base.vm_status) {
> + if (!bo_va->base.bo)
> + continue;
> + total_invalidated += amdgpu_bo_print_info(id++, bo_va->base.bo, m);
> + }
> + spin_unlock(&vm->invalidated_lock);
> + total_invalidated_objs = id;
> +
> + seq_printf(m, "\tTotal idle size: %12lld\tobjs:\t%d\n", total_idle,
> + total_idle_objs);
The coding style here is still odd, the second line should start after the ( of the first line.
Apart from that looks good to me,
Christian.
Hi Christian,
One question regarding getting total allocations by app and also swapped size for the app.
Shouldn’t we print mapped entries as well to get total allocation by the process ?
Also which list would represent swapped memory?
Thanks,
Mihir
> + seq_printf(m, "\tTotal evicted size: %12lld\tobjs:\t%d\n", total_evicted,
> + total_evicted_objs);
> + seq_printf(m, "\tTotal relocated size: %12lld\tobjs:\t%d\n", total_relocated,
> + total_relocated_objs);
> + seq_printf(m, "\tTotal moved size: %12lld\tobjs:\t%d\n", total_moved,
> + total_moved_objs);
> + seq_printf(m, "\tTotal invalidated size: %12lld\tobjs:\t%d\n", total_invalidated,
> + total_invalidated_objs);
> +}
> +#endif
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h
> b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h
> index 7c46937c1c0e..74cc14179c41 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h
> @@ -441,4 +441,8 @@ void amdgpu_vm_move_to_lru_tail(struct amdgpu_device *adev,
> struct amdgpu_vm *vm);
> void amdgpu_vm_del_from_lru_notify(struct ttm_buffer_object *bo);
>
> +#if defined(CONFIG_DEBUG_FS)
> +void amdgpu_debugfs_vm_bo_info(struct amdgpu_vm *vm, struct seq_file
> +*m); #endif
> +
> #endif
More information about the amd-gfx
mailing list