rustytorch

Files

T

Omar SobhandClaude Opus 5.5 0150f79a4a embedded3 R6-2: the moving mask's classification on the device (RTX_E3_MASK_DEVICE=1, needs RTX_E3_GEOM_DEVICE=1)

- e3_mask.cu: the changed set (touched by either build, dilated), the faces of
  the changed cells, per changed cell the fluid flag / space-time activity /
  GCL entry and the merging master, per face the kind, the step aperture and
  open flag and the open-part centroid shift (compute_face_shifts' cv
  geometry), the GCL table's wall cells / areas / table with the correction,
  the merged cells' CSR (slave ranks from the master's distinct face
  neighbours), the imposition band's faces; block scans for ascending stream
  compaction and the CSR's exclusive scan. fp64, FMA contraction off.
- step/device/mask.rs DeviceMask: snapshots the previous apertures / volumes /
  evaluated cells before the R6-1 geometry kernels, runs the classification
  into DeviceCut's persistent tables (a, open, open_pred, active,
  active_pred, owner, shift, wall_flux, fold_ptr/fold_idx) and returns the
  compact values (MaskUpdate); the GCL sums in ascending order on the host.
  DeviceCut::update_after_device_mask: only the band's surface velocities
  remain (band list compacted on the device).
- maskupdate.rs: Mask::from_update — the host mirror from the old mask's step
  arrays and merging map (moved) and the mask retired a step earlier
  (instantaneous arrays, pooled), the compact values scattered; incremental
  fluid count / anchor / side check; the fresh-cell refill over the changed set.
- RTX_E3_BAND_CHECK=1 with the knob: the host rebuild on copies (from_cut,
  face shifts, step apertures, merging, GCL, refill) against the mirror, bit
  for bit, and every device table against a full build.
- Knob off: the host path unchanged (Mask::from_parts, Solver::configure_mask
  and retire_mask are the same statements); profile laps under
  RTX_E3_MOVING_PROFILE only.

Co-Authored-By: Claude Opus 5.5 (1M context) <[email protected]>

2026-09-24 21:00:25 -05:00

poisson

embedded3 PERF-3 P1-5 (c) part 1: scatter kernels (e3_scat_{f64,f32,u32,i32}) in the CG module; the refresh's seven f64 coefficient arrays and z links, and the V-cycle fine level's f32 arrays and maps, scattered on the changed rows into persistent buffers (DeviceVcycle::refresh_fine_rows) — slab CSV byte-identical, band check passed, device moving/cg green; ny 124 rebuild block 1,895 → 1,748 ms per step. The first form redirected the V-cycle's own e3_cg_scatter_f64 launch (a replace-all on the launch name) and failed every gate — the symbol table is the first thing to read when adding to a module

2026-09-20 18:09:02 -05:00

step

embedded3 R6-2: the moving mask's classification on the device (RTX_E3_MASK_DEVICE=1, needs RTX_E3_GEOM_DEVICE=1)

2026-09-24 21:00:25 -05:00

body.rs

embedded3 R6-1: the body's φ and the cut geometry on the device (RTX_E3_GEOM_DEVICE=1)

2026-09-24 08:15:34 -05:00

closure.rs

embedded3 PERF-3 P1-3 (f): the mask build's whole-grid scans as parallel per-entry maps (cell classification with the anchor as the smallest fluid index, the three face-kind tables, the merging scan's small flags, the centroid-shift table per face) — slab flag ny 62 CSV byte-identical, device moving/cg green; ny 124 rebuild block 2,680 → 2,569 ms per step (build_mask 409 → 309)