Currently we allocate all 3 levels of radix3 page tables using
nvkm_gsp_mem_ctor(), which uses dma_alloc_coherent() for allocating all of
the relevant memory. This can end up failing in scenarios where the system
has very high memory fragmentation, and we can't find enough contiguous
memory to allocate level 2 of the page table.

Currently, this can result in runtime PM issues on systems where memory
fragmentation is high - as we'll fail to allocate the page table for our
suspend/resume buffer:

  kworker/10:2: page allocation failure: order:7, mode:0xcc0(GFP_KERNEL),
  nodemask=(null),cpuset=/,mems_allowed=0
  CPU: 10 PID: 479809 Comm: kworker/10:2 Not tainted
  6.8.6-201.ChopperV6.fc39.x86_64 #1
  Hardware name: SLIMBOOK Executive/Executive, BIOS N.1.10GRU06 02/02/2024
  Workqueue: pm pm_runtime_work
  Call Trace:
   <TASK>
   dump_stack_lvl+0x64/0x80
   warn_alloc+0x165/0x1e0
   ? __alloc_pages_direct_compact+0xb3/0x2b0
   __alloc_pages_slowpath.constprop.0+0xd7d/0xde0
   __alloc_pages+0x32d/0x350
   __dma_direct_alloc_pages.isra.0+0x16a/0x2b0
   dma_direct_alloc+0x70/0x270
   nvkm_gsp_radix3_sg+0x5e/0x130 [nouveau]
   r535_gsp_fini+0x1d4/0x350 [nouveau]
   nvkm_subdev_fini+0x67/0x150 [nouveau]
   nvkm_device_fini+0x95/0x1e0 [nouveau]
   nvkm_udevice_fini+0x53/0x70 [nouveau]
   nvkm_object_fini+0xb9/0x240 [nouveau]
   nvkm_object_fini+0x75/0x240 [nouveau]
   nouveau_do_suspend+0xf5/0x280 [nouveau]
   nouveau_pmops_runtime_suspend+0x3e/0xb0 [nouveau]
   pci_pm_runtime_suspend+0x67/0x1e0
   ? __pfx_pci_pm_runtime_suspend+0x10/0x10
   __rpm_callback+0x41/0x170
   ? __pfx_pci_pm_runtime_suspend+0x10/0x10
   rpm_callback+0x5d/0x70
   ? __pfx_pci_pm_runtime_suspend+0x10/0x10
   rpm_suspend+0x120/0x6a0
   pm_runtime_work+0x98/0xb0
   process_one_work+0x171/0x340
   worker_thread+0x27b/0x3a0
   ? __pfx_worker_thread+0x10/0x10
   kthread+0xe5/0x120
   ? __pfx_kthread+0x10/0x10
   ret_from_fork+0x31/0x50
   ? __pfx_kthread+0x10/0x10
   ret_from_fork_asm+0x1b/0x30

Luckily, we don't actually need to allocate coherent memory for the page
table thanks to being able to pass the GPU a radix3 page table for
suspend/resume data. So, let's rewrite nvkm_gsp_radix3_sg() to use the sg
allocator for level 2. We continue using coherent allocations for lvl0 and
1, since they only take a single page.

V2:
* Don't forget to actually jump to the next scatterlist when we reach the
  end of the scatterlist we're currently on when writing out the page table
  for level 2
Signed-off-by: Lyude Paul <lyude@redhat.com>
Cc: stable@vger.kernel.org
Reviewed-by: Ben Skeggs <bskeggs@nvidia.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240429182318.189668-2-lyude@redhat.com