drm/amdgpu: fix number of fence calculations

Since adding gang submit we need to take the gang size into account
while reserving fences.

Signed-off-by: Christian König <christian.koenig@amd.com>
Fixes: 4624459c84d7 ("drm/amdgpu: add gang submit frontend v6")
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
index d950388..b34f9f8 100644 (file)
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
@@ -136,9 +136,6 @@ static int amdgpu_cs_p1_user_fence(struct amdgpu_cs_parser *p,
        bo = amdgpu_bo_ref(gem_to_amdgpu_bo(gobj));
        p->uf_entry.priority = 0;
        p->uf_entry.tv.bo = &bo->tbo;
-       /* One for TTM and two for the CS job */
-       p->uf_entry.tv.num_shared = 3;
-
        drm_gem_object_put(gobj);
 
        size = amdgpu_bo_size(bo);
@@ -912,15 +909,19 @@ static int amdgpu_cs_parser_bos(struct amdgpu_cs_parser *p,
 
        mutex_lock(&p->bo_list->bo_list_mutex);
 
-       /* One for TTM and one for the CS job */
+       /* One for TTM and one for each CS job */
        amdgpu_bo_list_for_each_entry(e, p->bo_list)
-               e->tv.num_shared = 2;
+               e->tv.num_shared = 1 + p->gang_size;
+       p->uf_entry.tv.num_shared = 1 + p->gang_size;
 
        amdgpu_bo_list_get_list(p->bo_list, &p->validated);
 
        INIT_LIST_HEAD(&duplicates);
        amdgpu_vm_get_pd_bo(&fpriv->vm, &p->validated, &p->vm_pd);
 
+       /* Two for VM updates, one for TTM and one for each CS job */
+       p->vm_pd.tv.num_shared = 3 + p->gang_size;
+
        if (p->uf_entry.tv.bo && !ttm_to_amdgpu_bo(p->uf_entry.tv.bo)->parent)
                list_add(&p->uf_entry.tv.head, &p->validated);