search for: nvkm_vmm_ptes_get

...u] [ 8.209278] nvkm_mmu_ptc_get.part.4+0x1ae/0x230 [nouveau] [ 8.209311] nvkm_vmm_iter.constprop.13+0x458/0x790 [nouveau] [ 8.209344] ? nvkm_vmm_map_choose+0xb0/0xb0 [nouveau] [ 8.209348] ? kvmalloc_node+0x3e/0x70 [ 8.209381] ? nvkm_vmm_ctor+0x1ce/0x250 [nouveau] [ 8.209413] nvkm_vmm_ptes_get+0x2e/0x80 [nouveau] [ 8.209446] nvkm_vmm_boot+0x3e/0x80 [nouveau] [ 8.209475] nv50_bar_oneinit+0xe8/0x310 [nouveau] [ 8.209504] nvkm_subdev_init+0x4b/0x1f0 [nouveau] [ 8.209538] nvkm_device_init+0x12c/0x260 [nouveau] [ 8.209573] nvkm_udevice_init+0x41/0x60 [nouveau] [ 8.20960...

...e allocation happens through nouveau_uvmm_vmm_get() which either allocates the corresponding page tables or increases the reference count, in case they already exist, accordingly. The call goes all the way through nvif into the nvkm layer (not the easiest to follow the call chain) and ends up in nvkm_vmm_ptes_get(). There are multiple reasons for updating the VA space at submit time in Nouveau. 1) Subsequent EXEC ioctl() calls would need to wait for the bind jobs they depend on within the ioctl() rather than in the scheduler queue, because at the point of time where the ioctl() happens the VA space wo...

...eau_uvmm_vmm_get() which >> either allocates the corresponding page tables or increases the >> reference count, in case they already exist, accordingly. >> The call goes all the way through nvif into the nvkm layer (not the >> easiest to follow the call chain) and ends up in nvkm_vmm_ptes_get(). >> >> There are multiple reasons for updating the VA space at submit time in >> Nouveau. >> >> 1) Subsequent EXEC ioctl() calls would need to wait for the bind jobs >> they depend on within the ioctl() rather than in the scheduler queue, >> because at th...