From: Oak Zeng <oak.zeng@intel.com>
To: intel-xe@lists.freedesktop.org
Subject: [CI 42/44] drm/xe/svm: Determine a vma is backed by device memory
Date: Fri, 14 Jun 2024 17:58:15 -0400 [thread overview]
Message-ID: <20240614215817.1097633-42-oak.zeng@intel.com> (raw)
In-Reply-To: <20240614215817.1097633-1-oak.zeng@intel.com>
With system allocator, a userptr can now be back by device
memory also. Introduce a helper function xe_vma_is_devmem
to determine whether a range of vma is backed by device memory.
Cc: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Cc: Matthew Brost <matthew.brost@intel.com>
Cc: Brian Welty <brian.welty@intel.com>
Cc: Himal Prasad Ghimiray <himal.prasad.ghimiray@intel.com>
Signed-off-by: Oak Zeng <oak.zeng@intel.com>
---
drivers/gpu/drm/xe/xe_pt.c | 27 +++++++++++++++++++++++++--
1 file changed, 25 insertions(+), 2 deletions(-)
diff --git a/drivers/gpu/drm/xe/xe_pt.c b/drivers/gpu/drm/xe/xe_pt.c
index 3c89a32741f5..a576d5cb7727 100644
--- a/drivers/gpu/drm/xe/xe_pt.c
+++ b/drivers/gpu/drm/xe/xe_pt.c
@@ -3,6 +3,7 @@
* Copyright © 2022 Intel Corporation
*/
+#include <linux/hmm.h>
#include "xe_pt.h"
#include "regs/xe_gtt_defs.h"
@@ -578,6 +579,29 @@ static const struct xe_pt_walk_ops xe_pt_stage_bind_ops = {
.pt_entry = xe_pt_stage_bind_entry,
};
+static bool xe_vma_is_devmem(struct xe_vma *vma, u64 start)
+{
+ struct xe_bo *bo = xe_vma_bo(vma);
+
+ if (xe_vma_is_userptr(vma)) {
+ struct xe_userptr_vma *uvma = to_userptr_vma(vma);
+ struct drm_hmmptr *hmmptr = &uvma->userptr.hmmptr;
+ u64 offset = start - xe_vma_start(vma);
+ u64 page_idx = offset >> PAGE_SHIFT;
+ u64 hmm_pfn = hmmptr->pfn[page_idx];
+ struct page *page = hmm_pfn_to_page(hmm_pfn);
+
+ /**
+ * FIXME: Assume there is no mixture system memory and device
+ * memory placement in the [start, end) range of vma. We might
+ * need to relook at this in the future.
+ */
+ return is_device_private_page(page);
+ }
+
+ return bo && (xe_bo_is_vram(bo) || xe_bo_is_stolen_devmem(bo));
+}
+
/**
* xe_pt_stage_bind() - Build a disconnected page-table tree for a given address
* range.
@@ -604,8 +628,7 @@ xe_pt_stage_bind(struct xe_tile *tile, struct xe_vma *vma, u64 start, u64 end,
{
struct xe_device *xe = tile_to_xe(tile);
struct xe_bo *bo = xe_vma_bo(vma);
- bool is_devmem = !xe_vma_is_userptr(vma) && bo &&
- (xe_bo_is_vram(bo) || xe_bo_is_stolen_devmem(bo));
+ bool is_devmem = xe_vma_is_devmem(vma, start);
struct xe_res_cursor curs;
struct xe_pt_stage_bind_walk xe_walk = {
.base = {
--
2.26.3
next prev parent reply other threads:[~2024-06-14 21:48 UTC|newest]
Thread overview: 60+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-06-14 21:57 [CI 01/44] mm/hmm: let users to tag specific PFNs Oak Zeng
2024-06-14 21:57 ` [CI 02/44] dma-mapping: provide an interface to allocate IOVA Oak Zeng
2024-06-14 21:57 ` [CI 03/44] dma-mapping: provide callbacks to link/unlink pages to specific IOVA Oak Zeng
2024-06-14 21:57 ` [CI 04/44] iommu/dma: Provide an interface to allow preallocate IOVA Oak Zeng
2024-06-14 21:57 ` [CI 05/44] iommu/dma: Prepare map/unmap page functions to receive IOVA Oak Zeng
2024-06-14 21:57 ` [CI 06/44] iommu/dma: Implement link/unlink page callbacks Oak Zeng
2024-06-14 21:57 ` [CI 07/44] mm: Add an empty implementation of zone_device_page_init Oak Zeng
2024-06-14 21:57 ` [CI 08/44] drm: Move GPUVA_START/LAST to drm_gpuvm.h Oak Zeng
2024-06-14 21:57 ` [CI 09/44] drm/svm: Mark drm_gpuvm to participate SVM Oak Zeng
2024-06-14 21:57 ` [CI 10/44] drm/svm: introduce drm_mem_region concept Oak Zeng
2024-06-14 21:57 ` [CI 11/44] drm/svm: introduce hmmptr and helper functions Oak Zeng
2024-06-14 21:57 ` [CI 12/44] drm/svm: Introduce helper to remap drm memory region Oak Zeng
2024-06-14 21:57 ` [CI 13/44] drm/svm: handle CPU page fault Oak Zeng
2024-06-14 21:57 ` [CI 14/44] drm/svm: Migrate a range of hmmptr to vram Oak Zeng
2024-06-14 21:57 ` [CI 15/44] drm/svm: Add DRM SVM documentation Oak Zeng
2024-06-14 21:57 ` [CI 16/44] drm/svm: Introduce DRM_SVM kernel config Oak Zeng
2024-06-14 21:57 ` [CI 17/44] drm/xe: s/xe_tile_migrate_engine/xe_tile_migrate_exec_queue Oak Zeng
2024-06-14 21:57 ` [CI 18/44] drm/xe: Add xe_vm_pgtable_update_op to xe_vma_ops Oak Zeng
2024-06-14 21:57 ` [CI 19/44] drm/xe: Convert multiple bind ops into single job Oak Zeng
2024-06-14 21:57 ` [CI 20/44] drm/xe: Update VM trace events Oak Zeng
2024-06-14 21:57 ` [CI 21/44] drm/xe: Update PT layer with better error handling Oak Zeng
2024-06-14 21:57 ` [CI 22/44] drm/xe: Retry BO allocation Oak Zeng
2024-06-14 21:57 ` [CI 23/44] drm/xe/uapi: Add DRM_XE_VM_BIND_FLAG_SYSTEM_ALLOCATOR flag Oak Zeng
2024-06-14 21:57 ` [CI 24/44] drm/xe: Add a helper to calculate userptr end address Oak Zeng
2024-06-14 21:57 ` [CI 25/44] drm/xe: Add dma_addr res cursor Oak Zeng
2024-06-14 21:57 ` [CI 26/44] drm/xe: Use drm_mem_region for xe Oak Zeng
2024-06-14 21:58 ` [CI 27/44] drm/xe: use drm_hmmptr in xe Oak Zeng
2024-06-14 21:58 ` [CI 28/44] drm/xe: Moving to range based vma invalidation Oak Zeng
2024-06-14 21:58 ` [CI 29/44] drm/xe: Support range based page table update Oak Zeng
2024-06-14 21:58 ` [CI 30/44] drm/xe/uapi: Add DRM_XE_VM_CREATE_FLAG_PARTICIPATE_SVM flag Oak Zeng
2024-06-14 21:58 ` [CI 31/44] drm/xe/svm: Create userptr if page fault occurs on system_allocator VMA Oak Zeng
2024-06-14 21:58 ` [CI 32/44] drm/xe/svm: Add faulted userptr VMA garbage collector Oak Zeng
2024-06-14 21:58 ` [CI 33/44] drm/xe: Introduce helper to get tile from memory region Oak Zeng
2024-06-14 21:58 ` [CI 34/44] drm/xe/svm: implement functions to allocate and free device memory Oak Zeng
2024-06-14 21:58 ` [CI 35/44] drm/xe/svm: Get drm device from drm memory region Oak Zeng
2024-06-14 21:58 ` [CI 36/44] drm/xe/svm: Get page map owner of a " Oak Zeng
2024-06-14 21:58 ` [CI 37/44] drm/xe/svm: Add migrate layer functions for SVM support Oak Zeng
2024-06-14 21:58 ` [CI 38/44] drm/xe/svm: introduce svm migration function Oak Zeng
2024-06-14 21:58 ` [CI 39/44] drm/xe/svm: Register xe memory region to drm layer Oak Zeng
2024-06-14 21:58 ` [CI 40/44] drm/xe/svm: Introduce DRM_XE_SVM kernel config Oak Zeng
2024-06-14 21:58 ` [CI 41/44] drm/xe/svm: Migration from sram to vram for system allocator Oak Zeng
2024-06-14 21:58 ` Oak Zeng [this message]
2024-06-14 21:58 ` [CI 43/44] drm/xe/svm: Introduce hmm_pfn array based resource cursor Oak Zeng
2024-06-14 21:58 ` [CI 44/44] drm/xe: Enable system allocator uAPI Oak Zeng
2024-06-14 22:41 ` ✓ CI.Patch_applied: success for series starting with [CI,01/44] mm/hmm: let users to tag specific PFNs Patchwork
2024-06-14 22:42 ` ✗ CI.checkpatch: warning " Patchwork
2024-06-14 22:43 ` ✓ CI.KUnit: success " Patchwork
2024-06-14 22:55 ` ✓ CI.Build: " Patchwork
2024-06-14 22:57 ` ✗ CI.Hooks: failure " Patchwork
2024-06-14 22:58 ` ✗ CI.checksparse: warning " Patchwork
2024-06-14 23:21 ` ✓ CI.BAT: success " Patchwork
2024-06-15 2:56 ` ✓ CI.Patch_applied: success for series starting with [CI,01/44] mm/hmm: let users to tag specific PFNs (rev2) Patchwork
2024-06-15 2:57 ` ✗ CI.checkpatch: warning " Patchwork
2024-06-15 2:58 ` ✓ CI.KUnit: success " Patchwork
2024-06-15 3:10 ` ✓ CI.Build: " Patchwork
2024-06-15 3:12 ` ✗ CI.Hooks: failure " Patchwork
2024-06-15 3:14 ` ✗ CI.checksparse: warning " Patchwork
2024-06-15 3:36 ` ✓ CI.BAT: success " Patchwork
2024-06-15 13:59 ` ✗ CI.FULL: failure for series starting with [CI,01/44] mm/hmm: let users to tag specific PFNs Patchwork
2024-06-18 8:53 ` ✓ CI.FULL: success for series starting with [CI,01/44] mm/hmm: let users to tag specific PFNs (rev2) Patchwork
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20240614215817.1097633-42-oak.zeng@intel.com \
--to=oak.zeng@intel.com \
--cc=intel-xe@lists.freedesktop.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox