On 5/1/20 11:20 AM, Jason Gunthorpe wrote:
From: Jason Gunthorpe <jgg@xxxxxxxxxxxx>
Presumably the intent here was that hmm_range_fault() could put the data
into some HW specific format and thus avoid some work. However, nothing
actually does that, and it isn't clear how anything actually could do that
as hmm_range_fault() provides CPU addresses which must be DMA mapped.
Perhaps there is some special HW that does not need DMA mapping, but we
don't have any examples of this, and the theoretical performance win of
avoiding an extra scan over the pfns array doesn't seem worth the
complexity. Plus pfns needs to be scanned anyhow to sort out any
DEVICE_PRIVATE pages.
This version replaces the uint64_t with an usigned long containing a pfn
and fixed flags. On input flags is filled with the HMM_PFN_REQ_* values,
on successful output it is filled with HMM_PFN_* values, describing the
state of the pages.
amdgpu is simple to convert, it doesn't use snapshot and doesn't use
per-page flags.
nouveau uses only 16 hmm_pte entries at most (ie fits in a few cache
lines), and it sweeps over its pfns array a couple of times anyhow. It
also has a nasty call chain before it reaches the dma map and hardware
suggesting performance isn't important:
nouveau_svm_fault():
args.i.m.method = NVIF_VMM_V0_PFNMAP
nouveau_range_fault()
nvif_object_ioctl()
client->driver->ioctl()
struct nvif_driver nvif_driver_nvkm:
.ioctl = nvkm_client_ioctl
nvkm_ioctl()
nvkm_ioctl_path()
nvkm_ioctl_v0[type].func(..)
nvkm_ioctl_mthd()
nvkm_object_mthd()
struct nvkm_object_func nvkm_uvmm:
.mthd = nvkm_uvmm_mthd
nvkm_uvmm_mthd()
nvkm_uvmm_mthd_pfnmap()
nvkm_vmm_pfn_map()
nvkm_vmm_ptes_get_map()
func == gp100_vmm_pgt_pfn
struct nvkm_vmm_desc_func gp100_vmm_desc_spt:
.pfn = gp100_vmm_pgt_pfn
nvkm_vmm_iter()
REF_PTES == func == gp100_vmm_pgt_pfn()
dma_map_page()
Acked-by: Felix Kuehling <Felix.Kuehling@xxxxxxx>
Tested-by: Ralph Campbell <rcampbell@xxxxxxxxxx>
Signed-off-by: Jason Gunthorpe <jgg@xxxxxxxxxxxx>
Signed-off-by: Christoph Hellwig <hch@xxxxxx>
---
Documentation/vm/hmm.rst | 26 ++--
drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c | 35 ++----
drivers/gpu/drm/nouveau/nouveau_dmem.c | 27 +---
drivers/gpu/drm/nouveau/nouveau_dmem.h | 3 +-
drivers/gpu/drm/nouveau/nouveau_svm.c | 87 ++++++++-----
include/linux/hmm.h | 99 ++++++---------
mm/hmm.c | 160 +++++++++++-------------
7 files changed, 192 insertions(+), 245 deletions(-)
...snip...
+static void nouveau_hmm_convert_pfn(struct nouveau_drm *drm,
+ struct hmm_range *range, u64 *ioctl_addr)
+{
+ unsigned long i, npages;
+
+ /*
+ * The ioctl_addr prepared here is passed through nvif_object_ioctl()
+ * to an eventual DMA map in something like gp100_vmm_pgt_pfn()
+ *
+ * This is all just encoding the internal hmm reprensetation into a
s/reprensetation/representation/
Looks good and still tests OK with nouveau.
