NVIDIA/open-gpu-kernel-modules
1
0
Fork 0
mirror of https://github.com/NVIDIA/open-gpu-kernel-modules.git synced 2026-02-28 19:03:58 +00:00
Code Issues Packages Projects Releases Wiki Activity

570.86.15

Browse Source
This commit is contained in:
Bernhard Stoeckner
2025-01-27 19:36:56 +01:00
parent 9d0b0414a5
commit 54d69484da
1166 changed files with 318863 additions and 182687 deletions
Download Patch File Download Diff File Expand all files Collapse all files

13
kernel-open/nvidia-uvm/uvm_ats_faults.h
View File

@@ -29,12 +29,13 @@
// Service ATS faults in the range (base, base + UVM_VA_BLOCK_SIZE) with service
// type for individual pages in the range requested by page masks set in
// ats_context->fault.read_fault_mask/write_fault_mask. base must be aligned to
// UVM_VA_BLOCK_SIZE. The caller is responsible for ensuring that faulting
// addresses fall completely within the VMA. The caller is also responsible for
// ensuring that the faulting addresses don't overlap a GMMU region. (See
// uvm_ats_check_in_gmmu_region). The caller is also responsible for handling
// any errors returned by this function (fault cancellations etc.).
// ats_context->fault.read_fault_mask/write_fault_mask/prefetch_only_mask.
// base must be aligned to UVM_VA_BLOCK_SIZE. The caller is responsible for
// ensuring that faulting addresses fall completely within the VMA. The caller
// is also responsible for ensuring that the faulting addresses don't overlap
// a GMMU region. (See uvm_ats_check_in_gmmu_region). The caller is also
// responsible for handling any errors returned by this function (fault
// cancellations etc.).
//
// Returns the fault service status in ats_context->fault.faults_serviced_mask.
// In addition, ats_context->fault.reads_serviced_mask returns whether read