docs: add internal render scale FPS findings

Co-authored-by: codegirl007 <s.raide@gmail.com>

.github/issues/add-configurable-internal-render-scale.md

@@ -0,0 +1,208 @@
+# Add configurable internal render scale to improve fill-bound FPS
+
+## Summary
+
+Complete-frame benchmarks show that the current sprite renderer is primarily
+pixel/fill bound on Vulkan/Lavapipe, not CPU draw-preparation bound. Rendering
+the world into a smaller offscreen color target and nearest-blitting it to the
+native swapchain produced the largest measured FPS improvement.
+
+This should be an opt-in render-quality setting with a native-resolution
+fallback. For pixel art, a 50% scale is especially useful because it maps to an
+exact 2× nearest-neighbor upscale.
+
+## Evidence
+
+Environment:
+
+- Commit `dc66203259e92ce39291556f7b6c26c8dd999b84`
+- Odin `dev-2026-05-nightly:ea5175d`
+- `-debug -o:speed`
+- SDL 3.4.12, Vulkan/Lavapipe, immediate present
+- 800×600 swapchain
+- Five 150-frame trials per invocation after 50 warm-up frames
+- Order-balanced baseline/candidate invocations
+- GPU idle wait included before stopping each trial timer
+
+### Fill and overdraw scaling
+
+The committed `perf-frame` harness produced:
+
+| Workload | Median frame time | Median FPS |
+| --- | ---: | ---: |
+| 1 visible sprite | 0.592 ms | 1,690 |
+| 24 visible sprites | 2.249 ms | 445 |
+| 64 visible sprites | 4.017 ms | 249 |
+| 128 visible sprites | 6.255 ms | 160 |
+| 128 stacked sprites | 7.692 ms | 130 |
+
+Stacking the same 128 sprites increased frame time by approximately 23%,
+confirming that overdraw matters.
+
+With one centered sprite, render-target scaling produced:
+
+| Window size | Pixels | Median frame time |
+| --- | ---: | ---: |
+| 400×300 | 120,000 | 0.268 ms |
+| 800×600 | 480,000 | 0.545 ms |
+| 1600×1200 | 1,920,000 | 2.075 ms |
+
+The near-linear increase at larger sizes is further evidence of a pixel-bound
+workload.
+
+### Internal render-scale prototype
+
+The temporary candidate kept the physical swapchain at 800×600, rendered
+sprites into a smaller `COLOR_TARGET | SAMPLER` texture, and used
+`BlitGPUTexture` with nearest filtering to upscale into the swapchain.
+
+Aggregate medians across ten trials per mode:
+
+| Workload | Native | 75% scale | 50% scale |
+| --- | ---: | ---: | ---: |
+| 128 spread sprites | 6.082 ms / 164 FPS | 4.082 ms / 245 FPS | 3.082 ms / 325 FPS |
+| 128 stacked sprites | 7.912 ms / 126 FPS | 5.116 ms / 195 FPS | 3.034 ms / 330 FPS |
+
+Compared with native resolution:
+
+- 75% reduced frame time by 33–35% and increased FPS by 49–55%.
+- 50% reduced frame time by 49–62% and increased FPS by 97–161%.
+
+A visual smoke test confirmed that the 50% path rendered the complete scene at
+the correct orientation and 800×600 output size. It was visibly coarser, as
+expected. A 75% scale at 800×600 does not produce an integer upscale and can
+create uneven pixel sizing with nearest filtering.
+
+## Reproduction harness
+
+Native spread and stacked workloads:
+
+```bash
+make perf-frame \
+	PERF_FRAME_SCENARIO=0 \
+	PERF_FRAME_SPRITES=128 \
+	PERF_FRAME_WIDTH=800 \
+	PERF_FRAME_HEIGHT=600
+
+make perf-frame \
+	PERF_FRAME_SCENARIO=3 \
+	PERF_FRAME_SPRITES=128 \
+	PERF_FRAME_WIDTH=800 \
+	PERF_FRAME_HEIGHT=600
+```
+
+Run the same commands on the candidate branch with internal render scale set to
+75% and 50%. Keep every `PERF_FRAME_*` value unchanged between comparisons.
+
+## Suggested fix
+
+Add explicit logical/output and internal-render dimensions to `App`:
+
+```odin
+Render_Scale :: enum {
+	Native,
+	Three_Quarter,
+	Half,
+}
+
+App :: struct {
+	// Existing swapchain fields remain the logical/output dimensions.
+	swapchain_texture: ^sdl.GPUTexture,
+	swapchain_w:       u32,
+	swapchain_h:       u32,
+
+	render_scale:   Render_Scale,
+	scene_texture:  ^sdl.GPUTexture,
+	scene_w:        u32,
+	scene_h:        u32,
+}
+```
+
+Create the offscreen target with the swapchain format and both usages required
+by SDL's blit path:
+
+```odin
+app.scene_texture = sdl.CreateGPUTexture(
+	app.device,
+	{
+		type = .D2,
+		format = sdl.GetGPUSwapchainTextureFormat(app.device, app.window),
+		usage = {.COLOR_TARGET, .SAMPLER},
+		width = app.scene_w,
+		height = app.scene_h,
+		layer_count_or_depth = 1,
+		num_levels = 1,
+		sample_count = ._1,
+	},
+)
+```
+
+Render the sprite pass into `scene_texture`. Continue using the logical
+swapchain dimensions for camera and clip-space calculations so world layout
+does not change with render scale:
+
+```odin
+color_info := sdl.GPUColorTargetInfo {
+	texture = app.scene_texture,
+	clear_color = app.clear_color,
+	load_op = .CLEAR,
+	store_op = .STORE,
+	cycle = true,
+}
+```
+
+After ending the render pass, upscale into the full swapchain:
+
+```odin
+blit := sdl.GPUBlitInfo {
+	source = {
+		texture = app.scene_texture,
+		w = app.scene_w,
+		h = app.scene_h,
+	},
+	destination = {
+		texture = app.swapchain_texture,
+		w = app.swapchain_w,
+		h = app.swapchain_h,
+	},
+	load_op = .DONT_CARE,
+	filter = .NEAREST,
+}
+sdl.BlitGPUTexture(app.cmd, blit)
+```
+
+Use the existing direct-to-swapchain path at native scale to avoid an
+unnecessary blit. Recreate the offscreen texture whenever the swapchain size,
+format, or render-scale setting changes. If native-resolution UI or text is
+added later, render it after the world blit in a separate swapchain pass.
+
+## Rejected experiment: fragment alpha discard
+
+Approximately 49.7% of pixels inside the baked frame rectangles have exact
+alpha zero. A temporary fragment shader discarded those texels:
+
+```glsl
+vec4 texel = texture(u_tex, v_uv);
+if (texel.a == 0.0) {
+    discard;
+}
+out_color = texel;
+```
+
+Despite the high transparent coverage, this regressed frame time by roughly
+5–6% in both spread and stacked workloads. Do not add alpha discard without
+contradictory hardware-GPU evidence.
+
+## Acceptance criteria
+
+- Native, 75%, and 50% internal render-scale settings are available.
+- Native scale retains the current direct-to-swapchain path.
+- The scene target is recreated safely on resize, format change, or scale
+  change and released during shutdown.
+- Camera/world coordinates remain stable when render scale changes.
+- Nearest filtering is used for pixel-art output.
+- Add screenshot-based checks for output orientation, viewport coverage, and
+  stable sprite placement at every scale.
+- Benchmark spread and stacked scenarios on at least one hardware GPU.
+- Document the quality tradeoff and recommend integer upscale ratios for pixel
+  art.

.github/issues/add-deterministic-sprite-performance-benchmark.md

@@ -0,0 +1,218 @@
+# Add deterministic CPU and frame benchmarks for sprite rendering
+
+## Summary
+
+The current `make flame` workflow is valuable for finding call stacks, but it
+does not provide a reproducible performance metric:
+
+- `examples/crowd` draws only 24 sprites.
+- `MAX_SPRITES` limits the renderer to 128 sprites.
+- Recording starts an interactive application and asks the user to play for
+  10–20 seconds before quitting.
+- The two checked-in reports contain only 499 and 469 samples.
+- CPU queue construction and GPU submission are combined in one profile.
+- The profiler currently builds unoptimized code.
+
+These limitations make it difficult to tell whether a change made
+`draw_sprite` faster, changed driver behavior, or merely changed sampling noise.
+
+## Evidence
+
+A temporary deterministic benchmark exposed two very different results:
+
+1. CPU-only `draw_sprite`, two million calls and seven trials:
+   - `-debug`: median 192.560 ns/draw
+   - `-debug -o:speed`: median 19.154 ns/draw
+2. Full 128-sprite frames through SDL GPU on Lavapipe, 1,000 measured frames
+   and five trials:
+   - `-debug`: median 2.486 ms/frame
+   - `-debug -o:speed`: median 2.447 ms/frame
+
+At the current cap, optimized CPU queue construction is approximately 3.2
+microseconds for 128 sprites. The full software-rendered frame is around 2.45
+milliseconds, so optimizing `draw_sprite` cannot materially improve that
+specific end-to-end workload. A hardware GPU or a larger future sprite limit
+may have a different balance.
+
+This split also explains why percentages from the current unoptimized
+flamegraphs overstate small helper functions.
+
+### Follow-up enhancement benchmarks
+
+Six proposed renderer changes were implemented temporarily and measured before
+being discarded. Full-frame tests used:
+
+- Odin `dev-2026-05-nightly:ea5175d`
+- `-debug -o:speed`
+- SDL 3.4.12 with Vulkan/Lavapipe
+- 128 animated sprites using real baked toad metadata and textures
+- Warm-up before measurement
+- Paired baseline/change samples on the same device with alternating order
+- Ten 400-frame samples per mode, except culling, which used seven 750-frame
+  samples per mode
+
+Each row is a separate paired run, so absolute frame times should only be
+compared within that row.
+
+| Enhancement | Baseline median | Changed median | Result |
+| --- | ---: | ---: | ---: |
+| Viewport culling, all visible | 4.845 ms | 4.822 ms | 0.5% faster |
+| Viewport culling, 50% offscreen | 2.856 ms | 2.878 ms | 0.8% slower |
+| SDL transfer and vertex buffer cycling | 4.869 ms | 4.834 ms | 0.7% faster |
+| Contiguous vertex queue and one upload-side copy | 4.793 ms | 4.809 ms | 0.3% slower |
+| Four-vertex indexed quads | 4.745 ms | 4.830 ms | 1.8% slower |
+| GPU instancing with 32-byte instance records | 5.527 ms | 5.467 ms | 1.1% faster |
+| Texture sorting, including sort cost, 128 runs to 2 | 5.765 ms | 5.654 ms | 1.9% faster |
+
+Interpretation:
+
+- Viewport culling is neutral at the current cap. The GPU already clips
+  offscreen triangles, and the sprites remain in one batched draw.
+- SDL buffer cycling is a small performance improvement and is also the
+  documented way to avoid overwriting resources still bound by prior frames.
+- Repacking the CPU queue does not help at 128 sprites; extra dynamic-array
+  work offsets the saved small-copy loop.
+- Indexed quads regress performance despite reducing dynamic vertex data.
+- Instancing reduces per-sprite upload data from 96 to 32 bytes, but the 1.1%
+  gain does not justify a second pipeline and shader path at the current cap.
+- Texture sorting has the largest full-frame gain, but unrestricted sorting can
+  change alpha compositing. It is only safe within compatible layer/order
+  groups.
+
+The recommended order is:
+
+1. Profile optimized builds and establish the deterministic benchmark.
+2. Apply the clip-space math simplification documented in the related issue.
+3. Enable SDL buffer cycling for correct cross-frame resource reuse.
+4. Consider layer-aware texture grouping if a 1.9% workload-specific gain is
+   worth the ordering complexity.
+5. Defer culling, queue repacking, indexed quads, and instancing until the
+   sprite limit or measured workload grows substantially.
+
+These results are from a software Vulkan backend. Hardware drivers may have a
+different balance, which is another reason to keep the benchmark reproducible
+and report backend details.
+
+## Suggested fix
+
+Add a non-interactive benchmark target with two explicitly separate workloads.
+
+The committed harnesses are:
+
+```make
+PERF_ODIN_FLAGS ?= -debug -o:speed
+
+perf-draw:
+	odin run benchmarks/draw_sprite \
+		-collection:pkg=. \
+		$(PERF_ODIN_FLAGS) \
+		-define:PERF_ITERATIONS=$(PERF_DRAW_ITERATIONS)
+
+perf-frame:
+	odin run benchmarks/sprite_frame \
+		-collection:pkg=. \
+		$(PERF_ODIN_FLAGS) \
+		-define:PERF_FRAMES=$(PERF_FRAME_FRAMES) \
+		-define:PERF_SCENARIO=$(PERF_FRAME_SCENARIO)
+```
+
+Run the standard workloads with:
+
+```bash
+# CPU-only draw preparation.
+make perf-draw
+
+# Complete frame: 128 visible sprites sharing one texture.
+make perf-frame PERF_FRAME_SCENARIO=0
+
+# Complete frame: every second sprite is fully offscreen.
+make perf-frame PERF_FRAME_SCENARIO=1
+
+# Complete frame: 128 sprites alternate between two texture objects.
+make perf-frame PERF_FRAME_SCENARIO=2
+```
+
+Every invocation prints the Git commit, Odin version, compiler flags, workload
+configuration, every trial, and the median. `perf-frame` waits for GPU idle
+after warm-up and after each measured frame batch so outstanding work is
+included.
+
+### CPU queue benchmark
+
+- Construct `App`, `Character_Data`, and `Sprite` with real baked metadata.
+- Use safe fake non-null GPU handles; `draw_sprite` only checks/stores these.
+- Preallocate the draw list.
+- Clear the queue whenever it reaches `MAX_SPRITES`.
+- Vary sprite position between calls so the compiler cannot hoist the work.
+- Warm up before timing.
+- Run at least one million calls and report nanoseconds per draw.
+- Build with `-o:speed` by default.
+
+The measured loop should clear the queue at its cap, vary input to prevent
+compiler hoisting, and report time per draw:
+
+```odin
+PERF_ITERATIONS :: #config(PERF_ITERATIONS, 2_000_000)
+
+start := sdl.GetTicksNS()
+for i in 0 ..< PERF_ITERATIONS {
+	if len(app.draw_list) == eng.MAX_SPRITES {
+		clear(&app.draw_list)
+	}
+	sprite.position.x = f32(i & 1023)
+	eng.draw_sprite(&app, &sprite)
+}
+elapsed := sdl.GetTicksNS() - start
+
+fmt.printfln(
+	"%.3f ns/draw",
+	f64(elapsed) / f64(PERF_ITERATIONS),
+)
+```
+
+### Full-frame benchmark
+
+- Use a real SDL GPU device and baked texture.
+- Warm up before timing.
+- Run a fixed number of frames without interactive input.
+- Report milliseconds per frame and sprites per second.
+- Record GPU backend, present mode, compiler version, compiler flags, and sprite
+  count.
+
+Use a fixed frame count rather than an interactive quit time:
+
+```odin
+PERF_FRAMES :: #config(PERF_FRAMES, 1_000)
+
+for _ in 0 ..< 100 {
+	draw_benchmark_frame(&app, sprites[:]) // warm-up
+}
+
+start := sdl.GetTicksNS()
+for _ in 0 ..< PERF_FRAMES {
+	draw_benchmark_frame(&app, sprites[:])
+}
+elapsed := sdl.GetTicksNS() - start
+
+fmt.printfln(
+	"%.3f ms/frame",
+	f64(elapsed) / f64(PERF_FRAMES) / 1_000_000.0,
+)
+```
+
+The CPU benchmark should be available without a display or GPU. The full-frame
+benchmark may remain opt-in where a suitable GPU backend is unavailable.
+
+Do not add a strict CI regression threshold initially; hosted runner variance
+will make a single threshold flaky. CI can still compile the benchmark and
+verify that it completes.
+
+## Acceptance criteria
+
+- A Makefile target runs the optimized CPU benchmark non-interactively.
+- Results include compiler flags, iteration count, median, and per-trial values.
+- CPU queue time is reported separately from complete frame time.
+- Sprite positions or frames vary during the measured loop.
+- The draw list never silently exceeds `MAX_SPRITES`.
+- The benchmark has documented commands for repeatable local comparison.
+- `make check` and `make test` continue to pass.

.github/issues/add-order-safe-texture-batching.md

@@ -0,0 +1,116 @@
+# Add opt-in order-safe texture batching
+
+## Summary
+
+`end_frame` batches only consecutive sprites that use the same texture.
+Alternating two textures therefore produces one sampler bind and draw call per
+sprite even when some sprites could safely be regrouped.
+
+Globally sorting transparent sprites by texture is not correct: overlapping
+sprites may blend differently when submission order changes. Batching should
+therefore be opt-in within explicit groups whose members are safe to reorder.
+
+## Evidence
+
+A temporary paired benchmark used:
+
+- Odin `dev-2026-05-nightly:ea5175d` with `-debug -o:speed`
+- SDL 3.4.12 and Vulkan/Lavapipe
+- 128 animated sprites alternating between two equivalent textures
+- Sorting cost included in the measured frame
+- Ten order-alternated samples of 400 frames per mode
+
+| Mode | Texture runs | Median frame time |
+| --- | ---: | ---: |
+| Submission order | 128 | 5.765 ms |
+| Texture grouped | 2 | 5.654 ms |
+
+Sorting and grouping improved median frame time by approximately **1.9%**.
+Hardware drivers with higher draw-call overhead may show a different result.
+
+## Reproduction harness
+
+Run the committed alternating-texture workload on the baseline and candidate
+commits:
+
+```bash
+make perf-frame \
+	PERF_FRAME_SCENARIO=2 \
+	PERF_ODIN_FLAGS="-debug -o:speed"
+```
+
+Keep all `PERF_FRAME_*` values unchanged. The baseline should produce one
+texture run per sprite; the candidate should reduce runs only inside explicit
+reorder-safe groups. Compare `median_ms_per_frame` and verify rendered output.
+
+## Suggested fix
+
+Add an explicit batch group to queued sprites. Group `0` keeps strict submission
+order; nonzero groups may be reordered only when the caller guarantees that
+their members are order-independent.
+
+```odin
+Queued_Sprite :: struct {
+	texture:     ^sdl.GPUTexture,
+	verts:       [SPRITE_VERT_COUNT]Vertex,
+	batch_group: u32, // 0 = strict order; nonzero = caller permits regrouping
+}
+```
+
+Sort each contiguous, nonzero group by texture immediately before upload:
+
+```odin
+group_texture_runs :: proc(list: []Queued_Sprite) {
+	start := 0
+	for start < len(list) {
+		group := list[start].batch_group
+		if group == 0 {
+			start += 1
+			continue
+		}
+
+		end := start + 1
+		for end < len(list) && list[end].batch_group == group {
+			end += 1
+		}
+
+		// Stable insertion sort is sufficient while MAX_SPRITES is 128.
+		for i in start + 1 ..< end {
+			item := list[i]
+			j := i
+			for j > start {
+				if uintptr(list[j - 1].texture) <= uintptr(item.texture) {
+					break
+				}
+				list[j] = list[j - 1]
+				j -= 1
+			}
+			list[j] = item
+		}
+		start = end
+	}
+}
+```
+
+Call it after all sprites are queued and before the transfer-buffer copy:
+
+```odin
+group_texture_runs(app.draw_list[:])
+```
+
+Expose batching through a separate API or explicit parameter so existing
+`draw_sprite` calls remain strict-order by default:
+
+```odin
+draw_sprite_batched(&app, &sprite, batch_group = 1)
+```
+
+## Acceptance criteria
+
+- Existing `draw_sprite` behavior preserves exact submission order.
+- Reordering requires an explicit nonzero batch group.
+- Sorting never moves a sprite across a strict-order entry or group boundary.
+- Add tests for strict order, group boundaries, stable same-texture ordering,
+  and reduced texture-run count.
+- Add a visual overlap test confirming default alpha compositing is unchanged.
+- Benchmark sorting cost and draw-call reduction on a hardware GPU.

.github/issues/enable-sdl-gpu-buffer-cycling.md

@@ -0,0 +1,83 @@
+# Enable SDL GPU buffer cycling for per-frame sprite uploads
+
+## Summary
+
+`end_frame` overwrites the same transfer buffer and vertex buffer every frame,
+but both SDL calls currently pass `cycle = false`:
+
+```odin
+sdl.MapGPUTransferBuffer(app.device, app.transfer_buffer, false)
+sdl.UploadToGPUBuffer(copy_pass, src, dst, false)
+```
+
+SDL documents cycling as the mechanism that rotates to an unbound internal
+resource when the previous frame still references the current one. Enabling it
+avoids an unnecessary resource dependency and makes the overwrite pattern
+explicitly safe.
+
+## Evidence
+
+A temporary paired benchmark used:
+
+- Odin `dev-2026-05-nightly:ea5175d` with `-debug -o:speed`
+- SDL 3.4.12 and Vulkan/Lavapipe
+- 128 animated sprites
+- Ten order-alternated samples of 400 frames per mode
+
+| Mode | Median frame time |
+| --- | ---: |
+| Cycling disabled | 4.869 ms |
+| Cycling enabled | 4.834 ms |
+
+Cycling improved median frame time by approximately **0.7%**. This is a small
+performance change, but it also follows SDL's documented resource-reuse model.
+
+## Reproduction harness
+
+Run the committed full-frame harness on the baseline commit and candidate
+commit:
+
+```bash
+make perf-frame \
+	PERF_FRAME_SCENARIO=0 \
+	PERF_ODIN_FLAGS="-debug -o:speed"
+```
+
+Keep all `PERF_FRAME_*` values unchanged. Compare `median_ms_per_frame`; the
+harness waits for GPU idle before stopping each trial timer.
+
+## Suggested fix
+
+Cycle both resources that are fully overwritten each frame:
+
+```odin
+map_ptr := sdl.MapGPUTransferBuffer(
+	app.device,
+	app.transfer_buffer,
+	true, // rotate if the previous frame still binds this transfer buffer
+)
+
+// Write the complete [0, n * SPRITE_VERTS_SIZE) range, then unmap.
+sdl.UnmapGPUTransferBuffer(app.device, app.transfer_buffer)
+
+copy_pass := sdl.BeginGPUCopyPass(cmd)
+sdl.UploadToGPUBuffer(
+	copy_pass,
+	src,
+	dst,
+	true, // rotate the destination vertex buffer if it is still bound
+)
+sdl.EndGPUCopyPass(copy_pass)
+```
+
+Cycling makes previous contents undefined, so this remains correct only because
+the renderer writes the complete vertex range used by the frame before drawing.
+Do not enable cycling for partial updates that depend on untouched data.
+
+## Acceptance criteria
+
+- Both `MapGPUTransferBuffer` and `UploadToGPUBuffer` use `cycle = true`.
+- The complete submitted vertex range is rewritten every frame.
+- Existing engine tests and examples continue to pass.
+- A full-frame benchmark confirms no regression on a hardware GPU backend.
+- Add a comment explaining why cycling is safe for this full-overwrite path.

.github/issues/profile-optimized-sprite-builds.md

@@ -0,0 +1,98 @@
+# Profile optimized sprite builds instead of unoptimized debug code
+
+## Summary
+
+`make flame-build` currently compiles the selected example with `-debug` but
+without an optimization mode:
+
+```make
+odin build examples/$(FLAME_EXAMPLE) -collection:pkg=. -out:$(FLAME_BIN) -debug
+```
+
+This makes the flamegraph useful for debugging but misleading for performance
+decisions. The current profiles largely describe code that will disappear or
+be inlined in an optimized build.
+
+## Evidence
+
+The two checked-in `crowd` profiles were captured from this unoptimized binary.
+They report:
+
+- `engine::draw_sprite`: 9.56% and 12.22% self time
+- `engine::to_clip`: 5.33% and 5.51% self time
+- `engine::sprite_feet_quad`: 3.98% and 3.30% self time
+- Additional time in string hashing/map lookup, bounds checks, and dynamic
+  array append helpers
+
+A CPU-only benchmark using the real `draw_sprite`, real baked toad metadata,
+preallocated draw list, changing sprite positions, and two million draws per
+trial produced:
+
+| Build | Median time per draw | Trials |
+| --- | ---: | ---: |
+| `-debug` (current Makefile behavior) | 192.560 ns | 7 |
+| `-debug -o:speed` | 19.154 ns | 7 |
+
+The optimized build is about **10.1x faster** without an engine code change.
+
+An end-to-end 128-sprite benchmark on SDL 3.4.12 with Lavapipe showed only a
+small full-frame difference (median 2.486 ms debug versus 2.447 ms optimized)
+because software GPU/driver work dominated. This does not invalidate the CPU
+result; it shows why CPU queue time and full-frame time must be reported
+separately.
+
+Environment:
+
+- Odin `dev-2026-05-nightly:ea5175d` (the version pinned by CI)
+- SDL 3.4.12
+- Linux x86-64
+
+## Suggested fix
+
+Compile profiling binaries with optimization while retaining symbols:
+
+```make
+FLAME_ODIN_FLAGS ?= -debug -o:speed
+
+flame-build:
+	odin build examples/$(FLAME_EXAMPLE) \
+		-collection:pkg=. \
+		-out:$(FLAME_BIN) \
+		$(FLAME_ODIN_FLAGS)
+```
+
+Keeping the flags configurable allows an explicitly unoptimized diagnostic run
+without making it the performance default.
+
+Example usage:
+
+```bash
+# Representative performance profile: optimized code with debug symbols.
+make flame FLAME_EXAMPLE=crowd
+
+# Explicitly profile unoptimized code when investigating debug-only behavior.
+make flame FLAME_EXAMPLE=crowd FLAME_ODIN_FLAGS="-debug -o:none"
+```
+
+Reproduce the isolated build-mode comparison with the committed CPU harness:
+
+```bash
+make perf-draw PERF_ODIN_FLAGS="-debug -o:none"
+make perf-draw PERF_ODIN_FLAGS="-debug -o:speed"
+```
+
+Both runs print the Git commit, Odin version, compiler flags, every trial, and
+the median nanoseconds per draw.
+
+Consider applying an explicit optimization mode to performance-oriented example
+runs as well. Plain `odin run` currently uses Odin's unoptimized default.
+
+## Acceptance criteria
+
+- `make flame-build` produces an optimized binary with debug symbols.
+- `FLAME_ODIN_FLAGS` can override the default for diagnostic builds.
+- `make check` and `make test` continue to pass.
+- A new `crowd` profile records the exact compiler flags in its report or
+  accompanying documentation.
+- Performance conclusions distinguish CPU `draw_sprite` cost from complete
+  frame/GPU submission cost.

.github/issues/prototype-instanced-sprite-rendering.md

@@ -0,0 +1,139 @@
+# Prototype instanced sprite rendering for larger batches
+
+## Summary
+
+The renderer currently generates and uploads six complete vertices per sprite:
+
+```text
+6 vertices × (position float2 + UV float2) = 96 bytes/sprite/frame
+```
+
+Instancing can keep one immutable six-corner unit quad on the GPU and upload one
+rectangle/UV record per sprite:
+
+```text
+clip rectangle float4 + UV rectangle float4 = 32 bytes/sprite/frame
+```
+
+This reduces dynamic upload volume by two thirds, but requires another pipeline
+and backend-specific vertex shader. At the current 128-sprite cap, the measured
+gain is too small to justify enabling it unconditionally.
+
+## Evidence
+
+A complete temporary instanced path was implemented with:
+
+- A static six-corner vertex buffer
+- A 32-byte per-instance buffer
+- Vertex-rate and instance-rate pipeline inputs
+- One instanced draw per texture run
+- Validated SPIR-V and SDL's debug GPU device
+
+The paired benchmark used ten order-alternated samples of 400 frames:
+
+| Mode | Median frame time |
+| --- | ---: |
+| Six dynamic vertices per sprite | 5.527 ms |
+| 32-byte instance per sprite | 5.467 ms |
+
+Instancing improved median frame time by approximately **1.1%** on
+Vulkan/Lavapipe with 128 sprites.
+
+## Reproduction harness
+
+Run the committed visible full-frame workload on baseline and candidate commits:
+
+```bash
+make perf-frame \
+	PERF_FRAME_SCENARIO=0 \
+	PERF_FRAME_SPRITES=128 \
+	PERF_ODIN_FLAGS="-debug -o:speed"
+```
+
+Keep all other `PERF_FRAME_*` values unchanged and compare
+`median_ms_per_frame`. To test 512 or more sprites, first raise the engine's
+`MAX_SPRITES` and associated buffer capacities on the candidate branch, then
+set `PERF_FRAME_SPRITES` to the same value.
+
+## Suggested fix
+
+Treat this as a prototype gated by larger sprite counts or a demonstrated
+hardware bottleneck, not as an immediate replacement.
+
+Define the compact instance payload:
+
+```odin
+Sprite_Instance :: struct {
+	clip_rect: [4]f32, // left, top, right, bottom
+	uv_rect:   [4]f32, // u0, v0, u1, v1
+}
+```
+
+Queue one record after the existing quad and UV calculations:
+
+```odin
+instance := Sprite_Instance {
+	clip_rect = {p0.x, p0.y, p2.x, p2.y},
+	uv_rect   = {u0, v0, u1, v1},
+}
+append(&app.instance_list, instance)
+```
+
+Use a static unit quad:
+
+```odin
+UNIT_QUAD := [6]Vec2 {
+	{0, 0}, {1, 0}, {1, 1},
+	{0, 0}, {1, 1}, {0, 1},
+}
+```
+
+The instanced vertex shader reconstructs position and UV:
+
+```glsl
+#version 450
+
+layout(location = 0) in vec2 in_corner;
+layout(location = 1) in vec4 in_clip_rect;
+layout(location = 2) in vec4 in_uv_rect;
+
+layout(location = 0) out vec2 v_uv;
+
+void main() {
+    vec2 position = mix(in_clip_rect.xy, in_clip_rect.zw, in_corner);
+    v_uv = mix(in_uv_rect.xy, in_uv_rect.zw, in_corner);
+    gl_Position = vec4(position, 0.0, 1.0);
+}
+```
+
+Configure slot 0 as vertex-rate and slot 1 as instance-rate, then draw each
+texture run with six vertices and `run` instances:
+
+```odin
+vb_descs := [2]sdl.GPUVertexBufferDescription {
+	{slot = 0, pitch = u32(size_of(Vec2)), input_rate = .VERTEX},
+	{slot = 1, pitch = u32(size_of(Sprite_Instance)), input_rate = .INSTANCE},
+}
+
+sdl.DrawGPUPrimitives(
+	app.render_pass,
+	6,        // unit-quad vertices
+	u32(run), // sprite instances in this texture run
+	0,
+	0,
+)
+```
+
+Keep the current path as a fallback until the instanced implementation exists
+for Vulkan, D3D12, and Metal and demonstrates a meaningful hardware win.
+
+## Acceptance criteria
+
+- Benchmark at 128, 512, 2,048, and 10,000 sprites or the highest supported
+  counts.
+- Report CPU queue time, bytes uploaded, and complete frame time separately.
+- Require a meaningful hardware improvement before changing the default path.
+- Supply equivalent Vulkan, D3D12, and Metal shaders.
+- Preserve texture-run batching and sprite flip/trim behavior.
+- Add visual equivalence tests for position, UVs, animation frames, and flip.
+- Retain the existing six-vertex path as a fallback during evaluation.

.github/issues/simplify-draw-sprite-clip-math.md

@@ -0,0 +1,130 @@
+# Reduce repeated clip-space work in `draw_sprite`
+
+## Summary
+
+`draw_sprite` calls `to_clip` four times for an axis-aligned quad:
+
+```odin
+p0 := to_clip(x0_px, y0_px, sw, sh)
+p1 := to_clip(x1_px, y0_px, sw, sh)
+p2 := to_clip(x1_px, y1_px, sw, sh)
+p3 := to_clip(x0_px, y1_px, sw, sh)
+```
+
+This repeats the same divisions and converts duplicate x/y coordinates. An
+axis-aligned sprite has only two unique x values and two unique y values.
+
+This is a measurable optimization, but it is low priority at the current
+128-sprite limit because the absolute saving is small.
+
+## Evidence
+
+The checked-in unoptimized profiles report `engine::to_clip` at 5.33% and 5.51%
+self time. Those percentages are inflated by the unoptimized profiling build,
+so the change was also measured with `-debug -o:speed`.
+
+A temporary benchmark used the real baked toad metadata, changed sprite
+position on every iteration, preallocated the queue, and performed two million
+draws per mode over seven trials:
+
+| Mode | Median time per draw |
+| --- | ---: |
+| Current `draw_sprite` | 25.128 ns |
+| Precomputed clip scale and reused coordinates | 21.531 ns |
+| Same math plus cached `Frame_Def` | 21.778 ns |
+
+Simplifying the math improved isolated draw time by approximately **14.3%**.
+Caching the resolved frame did not provide an additional benefit and should not
+be included without new evidence.
+
+At `MAX_SPRITES == 128`, the measured math saving is only about 0.46
+microseconds per completely full frame. GPU/driver work dominated the
+end-to-end benchmark, so this should follow the profiling and deterministic
+benchmark improvements.
+
+Follow-up paired full-frame experiments at the 128-sprite cap found no larger
+renderer-architecture win: buffer cycling improved median frame time by 0.7%,
+GPU instancing by 1.1%, and texture sorting by 1.9%, while culling, queue
+repacking, and indexed quads were neutral or slower. The clip-space change
+therefore remains the strongest measured optimization specifically inside
+`draw_sprite`, although its absolute frame impact is still small.
+
+Environment:
+
+- Odin `dev-2026-05-nightly:ea5175d`
+- Optimized with `-debug -o:speed`
+- Linux x86-64
+
+## Reproduction harness
+
+Run the committed CPU harness on the baseline commit and again after applying
+the suggested fix:
+
+```bash
+make perf-draw PERF_ODIN_FLAGS="-debug -o:speed"
+```
+
+Keep `PERF_DRAW_ITERATIONS`, `PERF_DRAW_WARMUP`, and `PERF_DRAW_TRIALS`
+unchanged between commits. Compare `median_ns_per_draw`.
+
+## Suggested fix
+
+Compute clip scaling once and construct corners from the unique coordinates:
+
+```odin
+sprite_quad_to_clip :: proc(x0, y0, x1, y1, sw, sh: f32) -> [4]Vec2 {
+	sx := 2.0 / sw
+	sy := 2.0 / sh
+
+	left   := x0 * sx - 1
+	right  := x1 * sx - 1
+	top    := 1 - y0 * sy
+	bottom := 1 - y1 * sy
+
+	return {
+		{left, top},
+		{right, top},
+		{right, bottom},
+		{left, bottom},
+	}
+}
+
+points := sprite_quad_to_clip(x0_px, y0_px, x1_px, y1_px, sw, sh)
+p0, p1, p2, p3 := points[0], points[1], points[2], points[3]
+```
+
+Keep `to_clip` for general callers and its existing tests; this change only
+specializes quad construction inside `draw_sprite`.
+
+Add an equivalence test before replacing the current calls:
+
+```odin
+@(test)
+sprite_quad_clip_math_matches_to_clip :: proc(t: ^testing.T) {
+	x0, y0 := f32(125), f32(80)
+	x1, y1 := f32(325), f32(280)
+	sw, sh := f32(800), f32(600)
+
+	expected := [4]Vec2 {
+		to_clip(x0, y0, sw, sh),
+		to_clip(x1, y0, sw, sh),
+		to_clip(x1, y1, sw, sh),
+		to_clip(x0, y1, sw, sh),
+	}
+	actual := sprite_quad_to_clip(x0, y0, x1, y1, sw, sh)
+
+	for i in 0 ..< 4 {
+		testing.expect_value(t, actual[i], expected[i])
+	}
+}
+```
+
+## Acceptance criteria
+
+- Existing sprite geometry, camera, UV, and batching tests pass.
+- Add or extend a test that compares all four generated corners against
+  `to_clip` for representative viewport and sprite coordinates.
+- Flipped and unflipped sprites produce identical vertices to the current code.
+- An optimized deterministic benchmark shows at least a 10% improvement in
+  isolated `draw_sprite` time under comparable conditions.
+- Do not add a per-sprite frame cache as part of this issue.

.github/workflows/ci.yml

@@ -48,6 +48,10 @@ jobs:
       - name: Odin version
         run: odin version
 
+      # assetbake tests need vendor:stb native libs
+      - name: Build Odin STB libraries
+        run: make -C "$(odin root)/vendor/stb/src"
+
       - name: Unit tests
         run: make test
 

.gitignore

@@ -1,2 +1,14 @@
 *.bin
 tutorials/
+
+# Flamegraph / perf profiling artifacts
+tools/FlameGraph/
+perf.data
+perf.data.old
+flame-*.svg
+flame-*.jpg
+flame-*-report.txt
+flame.svg
+flame.jpg
+flame-report.txt
+*_perf

Makefile

@@ -1,15 +1,52 @@
-.PHONY: shaders-vulkan shaders-d3d12 shaders-metal shaders-all bake toad check test help
+.PHONY: shaders-vulkan shaders-d3d12 shaders-metal shaders-all bake toad hello_sprite crowd camera_sandbox clips check test help
+.PHONY: flame flame-build flame-record flame-svg flame-report flame-tools
+.PHONY: perf-draw perf-frame
+
+# Flamegraph profiling (needs: pacman -S perf). Example: make flame  or  make flame FLAME_EXAMPLE=toad
+FLAME_EXAMPLE ?= crowd
+FLAME_BIN := $(FLAME_EXAMPLE)_perf
+FLAMEGRAPH_DIR ?= tools/FlameGraph
+FLAME_OUT_DIR ?= flame
+# One stamp per `make` invocation so svg/jpg/report share a name.
+ifndef FLAME_STAMP
+FLAME_STAMP := $(shell date +%Y%m%d-%H%M%S)
+endif
+FLAME_PREFIX := $(FLAME_OUT_DIR)/$(FLAME_EXAMPLE)-$(FLAME_STAMP)
+
+PERF_DRAW_ITERATIONS ?= 2000000
+PERF_DRAW_WARMUP ?= 10000
+PERF_DRAW_TRIALS ?= 7
+PERF_ODIN_FLAGS ?= -debug -o:speed
+PERF_FRAME_SPRITES ?= 128
+PERF_FRAME_FRAMES ?= 400
+PERF_FRAME_WARMUP ?= 100
+PERF_FRAME_TRIALS ?= 10
+# 0=visible, 1=half offscreen, 2=alternating textures, 3=stacked
+PERF_FRAME_SCENARIO ?= 0
+PERF_FRAME_WIDTH ?= 800
+PERF_FRAME_HEIGHT ?= 600
 
 help:
 	@echo "Targets:"
-	@echo "  bake            Bake all characters under content/characters/"
-	@echo "  shaders-vulkan  Compile SPIR-V into shaders/vulkan/"
-	@echo "  shaders-d3d12   Compile DXIL into shaders/d3d12/ (needs shadercross)"
-	@echo "  shaders-metal   Compile MSL into shaders/metal/ (needs shadercross)"
-	@echo "  shaders-all     Build all shader backends"
-	@echo "  test            Run engine unit tests"
-	@echo "  check           Typecheck examples/toad"
-	@echo "  toad            Run the toad example"
+	@echo "  bake             Bake all characters under content/characters/"
+	@echo "  shaders-vulkan   Compile SPIR-V into shaders/vulkan/"
+	@echo "  shaders-d3d12    Compile DXIL into shaders/d3d12/ (needs shadercross)"
+	@echo "  shaders-metal    Compile MSL into shaders/metal/ (needs shadercross)"
+	@echo "  shaders-all      Build all shader backends"
+	@echo "  test             Run engine unit tests"
+	@echo "  check            Typecheck all examples"
+	@echo "  toad             Run the toad example (full demo)"
+	@echo "  hello_sprite     Minimal load + draw"
+	@echo "  crowd            Many sprites, one Character_Data"
+	@echo "  camera_sandbox   Pan camera / Space toggles follow"
+	@echo "  clips            Keys 1/2 switch idle/walk"
+	@echo "  flame            Build+record+SVG+JPG+text report (FLAME_EXAMPLE=$(FLAME_EXAMPLE))"
+	@echo "  flame-build      Debug binary only ($(FLAME_BIN))"
+	@echo "  flame-record     perf record (play, then quit)"
+	@echo "  flame-svg        Convert perf.data -> $(FLAME_PREFIX).svg/.jpg"
+	@echo "  flame-report     Convert perf.data -> $(FLAME_PREFIX)-report.txt"
+	@echo "  perf-draw        Deterministic CPU draw benchmark"
+	@echo "  perf-frame       Deterministic SDL GPU frame benchmark (scenario 0/1/2/3)"
 
 bake:
 	./scripts/bake_all.sh
@@ -27,9 +64,78 @@ shaders-all: shaders-vulkan shaders-d3d12 shaders-metal
 
 test:
 	odin test engine
+	odin test assetbake
 
 check:
 	odin check examples/toad -collection:pkg=.
+	odin check examples/hello_sprite -collection:pkg=.
+	odin check examples/crowd -collection:pkg=.
+	odin check examples/camera_sandbox -collection:pkg=.
+	odin check examples/clips -collection:pkg=.
+
+perf-draw:
+	@echo "git_commit=$$(git rev-parse HEAD)"
+	@echo "odin_version=$$(odin version)"
+	@echo "odin_flags=$(PERF_ODIN_FLAGS)"
+	odin run benchmarks/draw_sprite -collection:pkg=. $(PERF_ODIN_FLAGS) \
+		-define:PERF_ITERATIONS=$(PERF_DRAW_ITERATIONS) \
+		-define:PERF_WARMUP=$(PERF_DRAW_WARMUP) \
+		-define:PERF_TRIALS=$(PERF_DRAW_TRIALS)
+
+perf-frame:
+	@echo "git_commit=$$(git rev-parse HEAD)"
+	@echo "odin_version=$$(odin version)"
+	@echo "odin_flags=$(PERF_ODIN_FLAGS)"
+	odin run benchmarks/sprite_frame -collection:pkg=. $(PERF_ODIN_FLAGS) \
+		-define:PERF_SPRITES=$(PERF_FRAME_SPRITES) \
+		-define:PERF_FRAMES=$(PERF_FRAME_FRAMES) \
+		-define:PERF_WARMUP_FRAMES=$(PERF_FRAME_WARMUP) \
+		-define:PERF_TRIALS=$(PERF_FRAME_TRIALS) \
+		-define:PERF_SCENARIO=$(PERF_FRAME_SCENARIO) \
+		-define:PERF_WIDTH=$(PERF_FRAME_WIDTH) \
+		-define:PERF_HEIGHT=$(PERF_FRAME_HEIGHT)
 
 toad:
 	odin run examples/toad -collection:pkg=.
+
+hello_sprite:
+	odin run examples/hello_sprite -collection:pkg=.
+
+crowd:
+	odin run examples/crowd -collection:pkg=.
+
+camera_sandbox:
+	odin run examples/camera_sandbox -collection:pkg=.
+
+clips:
+	odin run examples/clips -collection:pkg=.
+
+flame-tools:
+	@if [ ! -x "$(FLAMEGRAPH_DIR)/stackcollapse-perf.pl" ] || [ ! -x "$(FLAMEGRAPH_DIR)/flamegraph.pl" ]; then \
+		echo "Cloning FlameGraph scripts into $(FLAMEGRAPH_DIR)..."; \
+		git clone --depth 1 https://github.com/brendangregg/FlameGraph.git "$(FLAMEGRAPH_DIR)"; \
+	fi
+
+flame-build:
+	odin build examples/$(FLAME_EXAMPLE) -collection:pkg=. -out:$(FLAME_BIN) -debug
+
+flame-record: flame-build
+	@echo ">>> Profiling ./$(FLAME_BIN) — play for ~10–20s under load, then quit the window."
+	@echo ">>> If perf fails with permissions: sudo sysctl kernel.perf_event_paranoid=1"
+	perf record -F 99 -g --call-graph dwarf -- ./$(FLAME_BIN)
+
+flame-svg: flame-tools
+	@test -f perf.data || { echo "No perf.data — run: make flame-record"; exit 1; }
+	@mkdir -p "$(FLAME_OUT_DIR)"
+	perf script | "$(FLAMEGRAPH_DIR)/stackcollapse-perf.pl" | "$(FLAMEGRAPH_DIR)/flamegraph.pl" > "$(FLAME_PREFIX).svg"
+	magick "$(FLAME_PREFIX).svg" "$(FLAME_PREFIX).jpg"
+	@echo "Wrote $(FLAME_PREFIX).svg and $(FLAME_PREFIX).jpg"
+
+flame-report:
+	@test -f perf.data || { echo "No perf.data — run: make flame-record"; exit 1; }
+	@mkdir -p "$(FLAME_OUT_DIR)"
+	perf report --stdio --no-children > "$(FLAME_PREFIX)-report.txt"
+	@echo "Wrote $(FLAME_PREFIX)-report.txt"
+
+flame: flame-record flame-svg flame-report
+	@echo "Artifacts: $(FLAME_PREFIX).{svg,jpg} $(FLAME_PREFIX)-report.txt"

assetbake/main.odin

@@ -56,10 +56,12 @@ Char_Def :: struct {
 
 // Packed atlas pixels + per-frame rects (indexed like the combined frames list)
 Atlas :: struct {
-	pixels: []u8,
-	width:  int,
-	height: int,
-	rects:  [][4]int, // [x, y, w, h] per frame; sizes may differ across clips
+	pixels:        []u8,
+	width:         int,
+	height:        int,
+	rects:         [][4]int,
+	source_sizes:  [][2]int,
+	trim_offsets:  [][2]int,
 }
 
 // Where one clip's frames sit inside the packed atlas frame list
@@ -95,6 +97,8 @@ bake_character :: proc(in_dir, out_dir: string) {
 	atlas := pack_atlas(frames)
 	defer delete(atlas.pixels)
 	defer delete(atlas.rects)
+	defer delete(atlas.source_sizes)
+	defer delete(atlas.trim_offsets)
 
 	ensure_dir(out_dir)
 
@@ -283,22 +287,60 @@ destroy_frames :: proc(frames: []^image.Image) {
 	delete(frames)
 }
 
-// pack_atlas packs variable-size frames with stb_rect_pack and blits them into one sheet.
-// Caller owns atlas.pixels and atlas.rects.
+opaque_bounds_rgba :: proc(pixels: []u8, iw, ih: int, alpha_min: u8 = 1) -> (x, y, w, h: int) {
+	min_x, min_y := iw, ih
+	max_x, max_y := -1, -1
+
+	for py in 0 ..< ih {
+		for px in 0 ..< iw {
+			i := (py * iw + px) * 4
+			if i + 3 >= len(pixels) do continue
+			a := pixels[i + 3]
+			if a >= alpha_min {
+				if px < min_x do min_x = px
+				if py < min_y do min_y = py
+				if px > max_x do max_x = px
+				if py > max_y do max_y = py
+			}
+		}
+	}
+
+	if max_x < min_x {
+		return 0, 0, 1, 1
+	}
+	return min_x, min_y, max_x - min_x + 1, max_y - min_y + 1
+}
+
+opaque_bounds :: proc(img: ^image.Image, alpha_min: u8 = 1) -> (x, y, w, h: int) {
+	iw, ih := int(img.width), int(img.height)
+	pixels := bytes.buffer_to_bytes(&img.pixels)
+	return opaque_bounds_rgba(pixels, iw, ih, alpha_min)
+}
+
 pack_atlas :: proc(frames: []^image.Image) -> Atlas {
-	PADDING :: 1 // 1px gap to reduce bleeding if filtered later
+	PADDING :: 1
 
 	pack_rects := make([]stbrp.Rect, len(frames))
 	defer delete(pack_rects)
 
+	trims := make([][4]int, len(frames))
+	defer delete(trims)
+	source_sizes := make([][2]int, len(frames))
+	trim_offsets := make([][2]int, len(frames))
+
 	total_area := 0
 	max_w := 0
 	max_h := 0
 	for frame, i in frames {
-		w := frame.width + PADDING
-		h := frame.height + PADDING
+		tx, ty, tw, th := opaque_bounds(frame)
+		trims[i] = {tx, ty, tw, th}
+		source_sizes[i] = {int(frame.width), int(frame.height)}
+		trim_offsets[i] = {tx, ty}
+
+		w := tw + PADDING
+		h := th + PADDING
 		pack_rects[i] = stbrp.Rect {
-			id = c.int(i), // pack_rects may reorder; id maps back to frame index
+			id = c.int(i),
 			w  = stbrp.Coord(w),
 			h  = stbrp.Coord(h),
 		}
@@ -307,7 +349,6 @@ pack_atlas :: proc(frames: []^image.Image) -> Atlas {
 		max_h = max(max_h, h)
 	}
 
-	// Grow a square-ish atlas until everything fits
 	side := max(max_w, max_h, int(math.ceil(math.sqrt(f64(total_area)))))
 	atlas_w, atlas_h := 0, 0
 	packed := false
@@ -326,10 +367,9 @@ pack_atlas :: proc(frames: []^image.Image) -> Atlas {
 		os.exit(1)
 	}
 
-	pixels := make([]u8, atlas_w * atlas_h * 4) // transparent
+	pixels := make([]u8, atlas_w * atlas_h * 4)
 	rects := make([][4]int, len(frames))
 
-	// Blit using each rect's id (array order may have changed during packing)
 	for pr in pack_rects {
 		if !pr.was_packed {
 			fmt.eprintfln("frame %d was not packed", pr.id)
@@ -337,27 +377,30 @@ pack_atlas :: proc(frames: []^image.Image) -> Atlas {
 		}
 		idx := int(pr.id)
 		frame := frames[idx]
-		fw := frame.width
-		fh := frame.height
+		trim := trims[idx]
+		tx, ty, tw, th := trim[0], trim[1], trim[2], trim[3]
 		dst_x := int(pr.x)
 		dst_y := int(pr.y)
 
 		src := bytes.buffer_to_bytes(&frame.pixels)
-		for y in 0 ..< fh {
-			src_row := src[y * fw * 4:(y + 1) * fw * 4]
+		src_w := int(frame.width)
+		for y in 0 ..< th {
+			src_row := src[((ty + y) * src_w + tx) * 4:][:tw * 4]
 			dst_i := ((dst_y + y) * atlas_w + dst_x) * 4
 			copy(pixels[dst_i:], src_row)
 		}
 
-		rects[idx] = {dst_x, dst_y, fw, fh}
+		rects[idx] = {dst_x, dst_y, tw, th}
 	}
 
 	fmt.printfln("packed atlas %dx%d", atlas_w, atlas_h)
 	return Atlas {
-		pixels = pixels,
-		width  = atlas_w,
-		height = atlas_h,
-		rects  = rects,
+		pixels       = pixels,
+		width        = atlas_w,
+		height       = atlas_h,
+		rects        = rects,
+		source_sizes = source_sizes,
+		trim_offsets = trim_offsets,
 	}
 }
 
@@ -422,8 +465,8 @@ build_char_def :: proc(
 			r := atlas.rects[slot]
 			frame_defs[i] = Frame_Def {
 				rect        = r,
-				source_size = {r[2], r[3]},
-				trim_offset = {0, 0}, // no trimming in v1
+				source_size = atlas.source_sizes[slot],
+				trim_offset = atlas.trim_offsets[slot],
 			}
 		}
 

assetbake/opaque_bounds_test.odin

@@ -0,0 +1,46 @@
+package main
+
+import "core:testing"
+
+@(test)
+opaque_bounds_inset :: proc(t: ^testing.T) {
+	// 4x4, opaque 2x2 at (1,1)
+	pixels := make([]u8, 4 * 4 * 4)
+	defer delete(pixels)
+	for y in 1 ..= 2 {
+		for x in 1 ..= 2 {
+			i := (y * 4 + x) * 4
+			pixels[i + 3] = 255
+		}
+	}
+	x, y, w, h := opaque_bounds_rgba(pixels, 4, 4)
+	testing.expect_value(t, x, 1)
+	testing.expect_value(t, y, 1)
+	testing.expect_value(t, w, 2)
+	testing.expect_value(t, h, 2)
+}
+
+@(test)
+opaque_bounds_full :: proc(t: ^testing.T) {
+	pixels := make([]u8, 2 * 2 * 4)
+	defer delete(pixels)
+	for i := 3; i < len(pixels); i += 4 {
+		pixels[i] = 255
+	}
+	x, y, w, h := opaque_bounds_rgba(pixels, 2, 2)
+	testing.expect_value(t, x, 0)
+	testing.expect_value(t, y, 0)
+	testing.expect_value(t, w, 2)
+	testing.expect_value(t, h, 2)
+}
+
+@(test)
+opaque_bounds_empty :: proc(t: ^testing.T) {
+	pixels := make([]u8, 3 * 3 * 4)
+	defer delete(pixels)
+	x, y, w, h := opaque_bounds_rgba(pixels, 3, 3)
+	testing.expect_value(t, x, 0)
+	testing.expect_value(t, y, 0)
+	testing.expect_value(t, w, 1)
+	testing.expect_value(t, h, 1)
+}

assets_baked/characters/toad/toad.char.json

@@ -14,274 +14,274 @@
             "frames": [
                 {
                     "rect": [
-                        0,
-                        0,
-                        268,
-                        326
+                        1253,
+                        630,
+                        204,
+                        311
                     ],
                     "source_size": [
                         268,
                         326
                     ],
                     "trim_offset": [
-                        0,
-                        0
+                        9,
+                        14
                     ]
                 },
                 {
                     "rect": [
-                        269,
+                        1256,
                         0,
-                        268,
-                        326
+                        218,
+                        314
                     ],
                     "source_size": [
                         268,
                         326
                     ],
                     "trim_offset": [
-                        0,
-                        0
+                        9,
+                        10
                     ]
                 },
                 {
                     "rect": [
-                        538,
-                        0,
-                        268,
-                        326
+                        1458,
+                        630,
+                        231,
+                        310
                     ],
                     "source_size": [
                         268,
                         326
                     ],
                     "trim_offset": [
-                        0,
-                        0
+                        9,
+                        8
                     ]
                 },
                 {
                     "rect": [
-                        807,
-                        0,
-                        268,
-                        326
+                        643,
+                        943,
+                        243,
+                        295
                     ],
                     "source_size": [
                         268,
                         326
                     ],
                     "trim_offset": [
-                        0,
-                        0
+                        10,
+                        7
                     ]
                 },
                 {
                     "rect": [
-                        1076,
-                        0,
-                        268,
-                        326
+                        881,
+                        1245,
+                        254,
+                        278
                     ],
                     "source_size": [
                         268,
                         326
                     ],
                     "trim_offset": [
-                        0,
-                        0
-                    ]
-                },
-                {
-                    "rect": [
-                        1345,
-                        0,
-                        268,
-                        326
-                    ],
-                    "source_size": [
-                        268,
-                        326
-                    ],
-                    "trim_offset": [
-                        0,
-                        0
-                    ]
-                },
-                {
-                    "rect": [
-                        1614,
-                        0,
-                        268,
-                        326
-                    ],
-                    "source_size": [
-                        268,
-                        326
-                    ],
-                    "trim_offset": [
-                        0,
-                        0
+                        11,
+                        8
                     ]
                 },
                 {
                     "rect": [
                         0,
-                        327,
-                        268,
-                        326
+                        944,
+                        247,
+                        288
                     ],
                     "source_size": [
                         268,
                         326
                     ],
                     "trim_offset": [
-                        0,
-                        0
+                        10,
+                        7
                     ]
                 },
                 {
                     "rect": [
-                        269,
-                        327,
-                        268,
-                        326
+                        1669,
+                        941,
+                        236,
+                        305
                     ],
                     "source_size": [
                         268,
                         326
                     ],
                     "trim_offset": [
-                        0,
-                        0
-                    ]
-                },
-                {
-                    "rect": [
-                        538,
-                        327,
-                        268,
-                        326
-                    ],
-                    "source_size": [
-                        268,
-                        326
-                    ],
-                    "trim_offset": [
-                        0,
-                        0
-                    ]
-                },
-                {
-                    "rect": [
-                        807,
-                        327,
-                        268,
-                        326
-                    ],
-                    "source_size": [
-                        268,
-                        326
-                    ],
-                    "trim_offset": [
-                        0,
-                        0
-                    ]
-                },
-                {
-                    "rect": [
-                        1076,
-                        327,
-                        268,
-                        326
-                    ],
-                    "source_size": [
-                        268,
-                        326
-                    ],
-                    "trim_offset": [
-                        0,
-                        0
-                    ]
-                },
-                {
-                    "rect": [
-                        1345,
-                        327,
-                        268,
-                        326
-                    ],
-                    "source_size": [
-                        268,
-                        326
-                    ],
-                    "trim_offset": [
-                        0,
-                        0
-                    ]
-                },
-                {
-                    "rect": [
-                        1614,
-                        327,
-                        268,
-                        326
-                    ],
-                    "source_size": [
-                        268,
-                        326
-                    ],
-                    "trim_offset": [
-                        0,
-                        0
+                        9,
+                        8
                     ]
                 },
                 {
                     "rect": [
                         0,
-                        654,
-                        268,
-                        326
+                        316,
+                        223,
+                        313
                     ],
                     "source_size": [
                         268,
                         326
                     ],
                     "trim_offset": [
-                        0,
-                        0
+                        9,
+                        9
                     ]
                 },
                 {
                     "rect": [
-                        269,
-                        654,
-                        268,
-                        326
+                        1458,
+                        941,
+                        210,
+                        310
                     ],
                     "source_size": [
                         268,
                         326
                     ],
                     "trim_offset": [
-                        0,
-                        0
+                        9,
+                        13
                     ]
                 },
                 {
                     "rect": [
-                        538,
-                        654,
-                        268,
-                        326
+                        224,
+                        316,
+                        211,
+                        313
                     ],
                     "source_size": [
                         268,
                         326
                     ],
                     "trim_offset": [
+                        11,
+                        11
+                    ]
+                },
+                {
+                    "rect": [
+                        1032,
+                        630,
+                        220,
+                        311
+                    ],
+                    "source_size": [
+                        268,
+                        326
+                    ],
+                    "trim_offset": [
+                        13,
+                        7
+                    ]
+                },
+                {
+                    "rect": [
+                        1032,
+                        942,
+                        228,
+                        302
+                    ],
+                    "source_size": [
+                        268,
+                        326
+                    ],
+                    "trim_offset": [
+                        16,
+                        4
+                    ]
+                },
+                {
+                    "rect": [
                         0,
-                        0
+                        1233,
+                        237,
+                        286
+                    ],
+                    "source_size": [
+                        268,
+                        326
+                    ],
+                    "trim_offset": [
+                        18,
+                        2
+                    ]
+                },
+                {
+                    "rect": [
+                        643,
+                        1239,
+                        237,
+                        286
+                    ],
+                    "source_size": [
+                        268,
+                        326
+                    ],
+                    "trim_offset": [
+                        18,
+                        2
+                    ]
+                },
+                {
+                    "rect": [
+                        414,
+                        943,
+                        228,
+                        302
+                    ],
+                    "source_size": [
+                        268,
+                        326
+                    ],
+                    "trim_offset": [
+                        16,
+                        4
+                    ]
+                },
+                {
+                    "rect": [
+                        1690,
+                        630,
+                        220,
+                        310
+                    ],
+                    "source_size": [
+                        268,
+                        326
+                    ],
+                    "trim_offset": [
+                        13,
+                        7
+                    ]
+                },
+                {
+                    "rect": [
+                        436,
+                        316,
+                        210,
+                        313
+                    ],
+                    "source_size": [
+                        268,
+                        326
+                    ],
+                    "trim_offset": [
+                        11,
+                        11
                     ]
                 }
             ]
@@ -292,290 +292,290 @@
             "frames": [
                 {
                     "rect": [
-                        807,
-                        654,
-                        213,
-                        319
+                        414,
+                        630,
+                        205,
+                        312
                     ],
                     "source_size": [
                         213,
                         319
                     ],
                     "trim_offset": [
-                        0,
-                        0
+                        6,
+                        7
                     ]
                 },
                 {
                     "rect": [
-                        1021,
-                        654,
-                        213,
-                        319
+                        647,
+                        316,
+                        206,
+                        313
                     ],
                     "source_size": [
                         213,
                         319
                     ],
                     "trim_offset": [
-                        0,
-                        0
+                        5,
+                        6
                     ]
                 },
                 {
                     "rect": [
-                        1235,
-                        654,
-                        213,
-                        319
+                        854,
+                        316,
+                        206,
+                        313
                     ],
                     "source_size": [
                         213,
                         319
                     ],
                     "trim_offset": [
-                        0,
-                        0
+                        5,
+                        6
                     ]
                 },
                 {
                     "rect": [
-                        1449,
-                        654,
-                        213,
-                        319
+                        1684,
+                        0,
+                        207,
+                        314
                     ],
                     "source_size": [
                         213,
                         319
                     ],
                     "trim_offset": [
-                        0,
-                        0
+                        4,
+                        5
                     ]
                 },
                 {
                     "rect": [
-                        1663,
-                        654,
-                        213,
-                        319
+                        1256,
+                        315,
+                        207,
+                        314
                     ],
                     "source_size": [
                         213,
                         319
                     ],
                     "trim_offset": [
-                        0,
-                        0
+                        4,
+                        5
                     ]
                 },
                 {
                     "rect": [
-                        807,
-                        974,
-                        213,
-                        319
+                        420,
+                        0,
+                        208,
+                        315
                     ],
                     "source_size": [
                         213,
                         319
                     ],
                     "trim_offset": [
-                        0,
-                        0
+                        3,
+                        4
                     ]
                 },
                 {
                     "rect": [
-                        1021,
-                        974,
-                        213,
-                        319
+                        1475,
+                        0,
+                        208,
+                        314
                     ],
                     "source_size": [
                         213,
                         319
                     ],
                     "trim_offset": [
-                        0,
-                        0
+                        3,
+                        4
                     ]
                 },
                 {
                     "rect": [
-                        1235,
-                        974,
-                        213,
-                        319
+                        629,
+                        0,
+                        208,
+                        315
                     ],
                     "source_size": [
                         213,
                         319
                     ],
                     "trim_offset": [
-                        0,
-                        0
-                    ]
-                },
-                {
-                    "rect": [
-                        1449,
-                        974,
-                        213,
-                        319
-                    ],
-                    "source_size": [
-                        213,
-                        319
-                    ],
-                    "trim_offset": [
-                        0,
-                        0
-                    ]
-                },
-                {
-                    "rect": [
-                        1663,
-                        974,
-                        213,
-                        319
-                    ],
-                    "source_size": [
-                        213,
-                        319
-                    ],
-                    "trim_offset": [
-                        0,
-                        0
+                        3,
+                        3
                     ]
                 },
                 {
                     "rect": [
                         0,
-                        981,
-                        213,
-                        319
+                        0,
+                        209,
+                        315
                     ],
                     "source_size": [
                         213,
                         319
                     ],
                     "trim_offset": [
-                        0,
-                        0
+                        2,
+                        3
                     ]
                 },
                 {
                     "rect": [
-                        214,
-                        981,
-                        213,
-                        319
+                        210,
+                        0,
+                        209,
+                        315
                     ],
                     "source_size": [
                         213,
                         319
                     ],
                     "trim_offset": [
-                        0,
-                        0
+                        2,
+                        3
                     ]
                 },
                 {
                     "rect": [
-                        428,
-                        981,
-                        213,
-                        319
+                        838,
+                        0,
+                        208,
+                        315
                     ],
                     "source_size": [
                         213,
                         319
                     ],
                     "trim_offset": [
-                        0,
-                        0
+                        3,
+                        3
                     ]
                 },
                 {
                     "rect": [
-                        642,
-                        1294,
-                        213,
-                        319
+                        1047,
+                        0,
+                        208,
+                        315
                     ],
                     "source_size": [
                         213,
                         319
                     ],
                     "trim_offset": [
-                        0,
-                        0
+                        3,
+                        4
                     ]
                 },
                 {
                     "rect": [
-                        856,
-                        1294,
-                        213,
-                        319
+                        1464,
+                        315,
+                        207,
+                        314
                     ],
                     "source_size": [
                         213,
                         319
                     ],
                     "trim_offset": [
-                        0,
-                        0
+                        4,
+                        5
                     ]
                 },
                 {
                     "rect": [
-                        1070,
-                        1294,
-                        213,
-                        319
+                        1672,
+                        315,
+                        207,
+                        314
                     ],
                     "source_size": [
                         213,
                         319
                     ],
                     "trim_offset": [
-                        0,
-                        0
+                        4,
+                        5
                     ]
                 },
                 {
                     "rect": [
-                        1284,
-                        1294,
-                        213,
-                        319
+                        0,
+                        630,
+                        206,
+                        313
                     ],
                     "source_size": [
                         213,
                         319
                     ],
                     "trim_offset": [
-                        0,
-                        0
+                        5,
+                        6
                     ]
                 },
                 {
                     "rect": [
-                        1498,
-                        1294,
-                        213,
-                        319
+                        207,
+                        630,
+                        206,
+                        313
                     ],
                     "source_size": [
                         213,
                         319
                     ],
                     "trim_offset": [
-                        0,
-                        0
+                        5,
+                        6
+                    ]
+                },
+                {
+                    "rect": [
+                        620,
+                        630,
+                        205,
+                        312
+                    ],
+                    "source_size": [
+                        213,
+                        319
+                    ],
+                    "trim_offset": [
+                        6,
+                        7
+                    ]
+                },
+                {
+                    "rect": [
+                        826,
+                        630,
+                        205,
+                        312
+                    ],
+                    "source_size": [
+                        213,
+                        319
+                    ],
+                    "trim_offset": [
+                        6,
+                        7
                     ]
                 }
             ]

benchmarks/draw_sprite/main.odin

@@ -0,0 +1,102 @@
+package main
+
+import "core:fmt"
+import "core:os"
+import eng "pkg:engine"
+import sdl "vendor:sdl3"
+
+PERF_ITERATIONS :: #config(PERF_ITERATIONS, 2_000_000)
+PERF_WARMUP     :: #config(PERF_WARMUP, 10_000)
+PERF_TRIALS     :: #config(PERF_TRIALS, 7)
+
+@(private)
+median :: proc(values: []f64) -> f64 {
+	for i in 1 ..< len(values) {
+		value := values[i]
+		j := i
+		for j > 0 {
+			if values[j - 1] <= value do break
+			values[j] = values[j - 1]
+			j -= 1
+		}
+		values[j] = value
+	}
+
+	middle := len(values) / 2
+	if len(values) & 1 == 1 do return values[middle]
+	return (values[middle - 1] + values[middle]) / 2
+}
+
+@(private)
+run_draws :: proc(app: ^eng.App, sprite: ^eng.Sprite, iterations: int) -> u64 {
+	clear(&app.draw_list)
+	start := sdl.GetTicksNS()
+	for i in 0 ..< iterations {
+		if len(app.draw_list) >= eng.MAX_SPRITES {
+			clear(&app.draw_list)
+		}
+
+		// Vary an input so optimized builds cannot hoist the draw calculations.
+		sprite.position.x = f32(i & 1023)
+		eng.draw_sprite(app, sprite)
+	}
+	return sdl.GetTicksNS() - start
+}
+
+main :: proc() {
+	#assert(PERF_ITERATIONS > 0)
+	#assert(PERF_WARMUP > 0)
+	#assert(PERF_TRIALS > 0)
+
+	file_data, err := os.read_entire_file(
+		"assets_baked/characters/toad/toad.char.json",
+		context.allocator,
+	)
+	if err != nil {
+		fmt.eprintfln("benchmark asset read failed: %v", err)
+		return
+	}
+	defer delete(file_data)
+
+	def, ok := eng.parse_char_def(file_data)
+	if !ok do return
+
+	data := eng.Character_Data {
+		def     = def,
+		texture = cast(^sdl.GPUTexture)uintptr(1),
+		width   = 1911,
+		height  = 1526,
+	}
+	app := eng.App {
+		cmd               = cast(^sdl.GPUCommandBuffer)uintptr(1),
+		swapchain_texture = cast(^sdl.GPUTexture)uintptr(2),
+		swapchain_w       = 800,
+		swapchain_h       = 600,
+		camera             = eng.camera_default(),
+		draw_list          = make(
+			[dynamic]eng.Queued_Sprite,
+			0,
+			eng.MAX_SPRITES,
+		),
+	}
+	defer delete(app.draw_list)
+
+	sprite := eng.spawn_sprite(&data, {400, 400}, "walk", 4)
+	_ = run_draws(&app, &sprite, PERF_WARMUP)
+
+	samples: [PERF_TRIALS]f64
+	fmt.printfln(
+		"benchmark=draw_sprite iterations=%d warmup=%d trials=%d sprites_per_queue=%d",
+		PERF_ITERATIONS,
+		PERF_WARMUP,
+		PERF_TRIALS,
+		eng.MAX_SPRITES,
+	)
+	for trial in 0 ..< PERF_TRIALS {
+		elapsed := run_draws(&app, &sprite, PERF_ITERATIONS)
+		samples[trial] = f64(elapsed) / f64(PERF_ITERATIONS)
+		fmt.printfln("trial=%d ns_per_draw=%.3f", trial + 1, samples[trial])
+	}
+
+	fmt.printfln("median_ns_per_draw=%.3f", median(samples[:]))
+}

benchmarks/sprite_frame/main.odin

@@ -0,0 +1,174 @@
+package main
+
+import "core:fmt"
+import eng "pkg:engine"
+import sdl "vendor:sdl3"
+
+PERF_SPRITES      :: #config(PERF_SPRITES, 128)
+PERF_FRAMES       :: #config(PERF_FRAMES, 400)
+PERF_WARMUP_FRAMES :: #config(PERF_WARMUP_FRAMES, 100)
+PERF_TRIALS       :: #config(PERF_TRIALS, 10)
+PERF_SCENARIO     :: #config(PERF_SCENARIO, 0)
+PERF_WIDTH        :: #config(PERF_WIDTH, 800)
+PERF_HEIGHT       :: #config(PERF_HEIGHT, 600)
+
+Scenario :: enum {
+	Visible,
+	Half_Offscreen,
+	Alternating_Textures,
+	Stacked,
+}
+
+@(private)
+median :: proc(values: []f64) -> f64 {
+	for i in 1 ..< len(values) {
+		value := values[i]
+		j := i
+		for j > 0 {
+			if values[j - 1] <= value do break
+			values[j] = values[j - 1]
+			j -= 1
+		}
+		values[j] = value
+	}
+
+	middle := len(values) / 2
+	if len(values) & 1 == 1 do return values[middle]
+	return (values[middle - 1] + values[middle]) / 2
+}
+
+@(private)
+scenario_name :: proc(scenario: Scenario) -> string {
+	switch scenario {
+	case .Visible:
+		return "visible"
+	case .Half_Offscreen:
+		return "half_offscreen"
+	case .Alternating_Textures:
+		return "alternating_textures"
+	case .Stacked:
+		return "stacked"
+	}
+	return "unknown"
+}
+
+@(private)
+present_mode_name :: proc(app: ^eng.App) -> string {
+	if sdl.WindowSupportsGPUPresentMode(app.device, app.window, .IMMEDIATE) {
+		return "immediate"
+	}
+	if sdl.WindowSupportsGPUPresentMode(app.device, app.window, .MAILBOX) {
+		return "mailbox"
+	}
+	return "vsync"
+}
+
+@(private)
+draw_frame :: proc(app: ^eng.App, sprites: []eng.Sprite) {
+	for &sprite in sprites {
+		eng.update_sprite(&sprite, 1.0 / 60.0)
+	}
+
+	eng.begin_frame(app)
+	for &sprite in sprites {
+		eng.draw_sprite(app, &sprite)
+	}
+	eng.end_frame(app)
+}
+
+main :: proc() {
+	#assert(PERF_SPRITES > 0)
+	#assert(PERF_SPRITES <= eng.MAX_SPRITES)
+	#assert(PERF_FRAMES > 0)
+	#assert(PERF_WARMUP_FRAMES > 0)
+	#assert(PERF_TRIALS > 0)
+	#assert(PERF_SCENARIO >= 0 && PERF_SCENARIO <= 3)
+	#assert(PERF_WIDTH > 0)
+	#assert(PERF_HEIGHT > 0)
+
+	scenario := Scenario(PERF_SCENARIO)
+
+	app: eng.App
+	if !eng.init(&app, "sprite frame benchmark", PERF_WIDTH, PERF_HEIGHT) {
+		return
+	}
+	defer eng.shutdown(&app)
+
+	path := "assets_baked/characters/toad/toad.char.json"
+	data_a, ok := eng.load_character_data(&app, path)
+	if !ok do return
+	defer eng.destroy_character_data(&app, &data_a)
+
+	data_b: eng.Character_Data
+	data_b, ok = eng.load_character_data(&app, path)
+	if !ok do return
+	defer eng.destroy_character_data(&app, &data_b)
+
+	sprites: [PERF_SPRITES]eng.Sprite
+	for i in 0 ..< PERF_SPRITES {
+		data := &data_a
+		if scenario == .Alternating_Textures && (i & 1) == 1 {
+			data = &data_b
+		}
+
+		position := eng.Vec2 {
+			f32(40 + (i % 16) * 48),
+			f32(120 + (i / 16) * 60),
+		}
+		if scenario == .Half_Offscreen && (i & 1) == 1 {
+			position = {-10_000, -10_000}
+		} else if scenario == .Stacked {
+			position = {f32(PERF_WIDTH / 2), f32(PERF_HEIGHT / 2)}
+		}
+		sprites[i] = eng.spawn_sprite(data, position, "walk", i % 17)
+	}
+
+	for _ in 0 ..< PERF_WARMUP_FRAMES {
+		draw_frame(&app, sprites[:])
+	}
+	if !sdl.WaitForGPUIdle(app.device) {
+		fmt.eprintfln("GPU wait failed after warm-up: %s", sdl.GetError())
+		return
+	}
+
+	fmt.printfln(
+		"benchmark=sprite_frame scenario=%s sprites=%d resolution=%dx%d frames=%d warmup=%d trials=%d backend=%v driver=%s present=%s",
+		scenario_name(scenario),
+		PERF_SPRITES,
+		PERF_WIDTH,
+		PERF_HEIGHT,
+		PERF_FRAMES,
+		PERF_WARMUP_FRAMES,
+		PERF_TRIALS,
+		app.shader.backend,
+		sdl.GetGPUDeviceDriver(app.device),
+		present_mode_name(&app),
+	)
+
+	samples: [PERF_TRIALS]f64
+	for trial in 0 ..< PERF_TRIALS {
+		start := sdl.GetTicksNS()
+		for _ in 0 ..< PERF_FRAMES {
+			draw_frame(&app, sprites[:])
+		}
+		if !sdl.WaitForGPUIdle(app.device) {
+			fmt.eprintfln("GPU wait failed after trial %d: %s", trial + 1, sdl.GetError())
+			return
+		}
+		elapsed := sdl.GetTicksNS() - start
+		samples[trial] = f64(elapsed) / f64(PERF_FRAMES) / 1_000_000.0
+		fmt.printfln(
+			"trial=%d ms_per_frame=%.3f fps=%.1f",
+			trial + 1,
+			samples[trial],
+			1_000.0 / samples[trial],
+		)
+	}
+
+	median_ms := median(samples[:])
+	fmt.printfln(
+		"median_ms_per_frame=%.3f median_fps=%.1f",
+		median_ms,
+		1_000.0 / median_ms,
+	)
+}

engine/animation_test.odin

@@ -74,13 +74,13 @@ character_clip_found_and_missing :: proc(t: ^testing.T) {
 	defer destroy_test_character(&data)
 
 	clip, ok := character_clip(&data, "idle")
-	testing.expect(t, ok)
+	testing.expect(t, ok, "character_clip should find idle")
 	testing.expect_value(t, len(clip.frames), 3)
-	testing.expect(t, clip.loop)
+	testing.expect(t, clip.loop, "idle clip should loop")
 	testing.expect_value(t, clip.fps, f32(10))
 
 	_, ok = character_clip(&data, "missing")
-	testing.expect(t, !ok)
+	testing.expect(t, !ok, "character_clip should miss unknown name")
 
 	data.def.clips["empty"] = Clip_Def {
 		loop   = true,
@@ -88,7 +88,7 @@ character_clip_found_and_missing :: proc(t: ^testing.T) {
 		frames = nil,
 	}
 	_, ok = character_clip(&data, "empty")
-	testing.expect(t, !ok)
+	testing.expect(t, !ok, "character_clip should reject empty frames")
 }
 
 @(test)
@@ -161,7 +161,7 @@ update_sprite_advances_multiple_frames :: proc(t: ^testing.T) {
 	s := spawn_sprite(&data, {}, "idle", 0)
 	update_sprite(&s, 0.25)
 	testing.expect_value(t, s.frame, 2)
-	testing.expect(t, s.time > 0.049 && s.time < 0.051)
+	testing.expect(t, s.time > 0.049 && s.time < 0.051, "leftover time after multi-frame advance should be ~0.05")
 }
 
 @(test)

engine/app.odin

@@ -1,6 +1,7 @@
 package engine
 
 import "core:fmt"
+import "core:mem"
 import "core:path/filepath"
 import sdl "vendor:sdl3"
 
@@ -10,7 +11,14 @@ Vertex :: struct {
 }
 
 SPRITE_VERT_COUNT :: 6
-VERTEX_BUFFER_SIZE :: SPRITE_VERT_COUNT * size_of(Vertex)
+MAX_SPRITES :: 128
+SPRITE_VERTS_SIZE :: SPRITE_VERT_COUNT * size_of(Vertex)
+VERTEX_BUFFER_SIZE :: MAX_SPRITES * SPRITE_VERTS_SIZE
+
+Queued_Sprite :: struct {
+	texture: ^sdl.GPUTexture,
+	verts:   [SPRITE_VERT_COUNT]Vertex,
+}
 
 App :: struct {
 	window:            ^sdl.Window,
@@ -25,6 +33,9 @@ App :: struct {
 	swapchain_h:       u32,
 	vertex_buffer:     ^sdl.GPUBuffer,
 	transfer_buffer:   ^sdl.GPUTransferBuffer,
+	draw_list:         [dynamic]Queued_Sprite,
+	clear_color:       sdl.FColor,
+	camera:            Camera,
 }
 
 Shader_Backend :: enum {
@@ -65,6 +76,19 @@ init :: proc(app: ^App, title: cstring, width, height: i32) -> bool {
 		return false
 	}
 
+	// Prefer uncapped present for profiling; fall back if unsupported.
+	present := sdl.GPUPresentMode.VSYNC
+	if sdl.WindowSupportsGPUPresentMode(app.device, app.window, .IMMEDIATE) {
+		present = .IMMEDIATE
+	} else if sdl.WindowSupportsGPUPresentMode(app.device, app.window, .MAILBOX) {
+		present = .MAILBOX
+	}
+	if present != .VSYNC {
+		if !sdl.SetGPUSwapchainParameters(app.device, app.window, .SDR, present) {
+			fmt.eprintfln("SetGPUSwapchainParameters failed: %s", sdl.GetError())
+		}
+	}
+
 	ok: bool
 	app.shader, ok = choose_shader_runtime(app.device)
 	if !ok do return false
@@ -112,10 +136,16 @@ init :: proc(app: ^App, title: cstring, width, height: i32) -> bool {
 		return false
 	}
 
+	app.draw_list = make([dynamic]Queued_Sprite)
+
+	app.camera = camera_default()
+
 	return true
 }
 
 shutdown :: proc(app: ^App) {
+	delete(app.draw_list)
+
 	if app.device != nil {
 		ok := sdl.WaitForGPUIdle(app.device)
 		if !ok {
@@ -164,9 +194,13 @@ events :: proc() -> bool {
 	return true
 }
 
-begin_frame :: proc(app: ^App, clear: sdl.FColor = {0.12, 0.12, 0.16, 1}) {
-	app.cmd = sdl.AcquireGPUCommandBuffer(app.device)
+begin_frame :: proc(app: ^App, clear_color: sdl.FColor = {0.12, 0.12, 0.16, 1}) {
+	clear(&app.draw_list)
+	app.clear_color = clear_color
+	app.render_pass = nil
+	app.swapchain_texture = nil
 
+	app.cmd = sdl.AcquireGPUCommandBuffer(app.device)
 	if app.cmd == nil {
 		fmt.eprintfln("AcquireGPUCommandBuffer failed: %s", sdl.GetError())
 		return
@@ -181,36 +215,104 @@ begin_frame :: proc(app: ^App, clear: sdl.FColor = {0.12, 0.12, 0.16, 1}) {
 	)
 
 	if !ok || app.swapchain_texture == nil {
+		app.swapchain_texture = nil
 		return
 	}
+}
+
+end_frame :: proc(app: ^App) {
+	if app.cmd == nil {
+		clear(&app.draw_list)
+		return
+	}
+
+	cmd := app.cmd
+	defer {
+		app.render_pass = nil
+		app.swapchain_texture = nil
+		app.cmd = nil
+		clear(&app.draw_list)
+	}
+
+	if app.swapchain_texture == nil {
+		if !sdl.SubmitGPUCommandBuffer(cmd) {
+			fmt.eprintfln("SubmitGPUCommandBuffer failed")
+		}
+		return
+	}
+
+	n := len(app.draw_list)
+	if n > 0 {
+		map_ptr := sdl.MapGPUTransferBuffer(app.device, app.transfer_buffer, false)
+		if map_ptr == nil {
+			fmt.eprintfln("MapGPUTransferBuffer failed: %s", sdl.GetError())
+			if !sdl.SubmitGPUCommandBuffer(cmd) {
+				fmt.eprintfln("SubmitGPUCommandBuffer failed")
+			}
+			return
+		}
+
+		for i in 0 ..< n {
+			q := &app.draw_list[i]
+			offset := i * SPRITE_VERTS_SIZE
+			mem.copy(
+				rawptr(uintptr(map_ptr) + uintptr(offset)),
+				raw_data(q.verts[:]),
+				SPRITE_VERTS_SIZE,
+			)
+		}
+		sdl.UnmapGPUTransferBuffer(app.device, app.transfer_buffer)
+
+		copy_pass := sdl.BeginGPUCopyPass(cmd)
+		src := sdl.GPUTransferBufferLocation {
+			transfer_buffer = app.transfer_buffer,
+			offset          = 0,
+		}
+		dst := sdl.GPUBufferRegion {
+			buffer = app.vertex_buffer,
+			offset = 0,
+			size   = u32(n * SPRITE_VERTS_SIZE),
+		}
+		sdl.UploadToGPUBuffer(copy_pass, src, dst, false)
+		sdl.EndGPUCopyPass(copy_pass)
+	}
 
 	color_info := sdl.GPUColorTargetInfo {
 		texture     = app.swapchain_texture,
-		clear_color = clear,
+		clear_color = app.clear_color,
 		load_op     = .CLEAR,
 		store_op    = .STORE,
 	}
-
-	app.render_pass = sdl.BeginGPURenderPass(app.cmd, &color_info, 1, nil)
-
+	app.render_pass = sdl.BeginGPURenderPass(cmd, &color_info, 1, nil)
 	sdl.BindGPUGraphicsPipeline(app.render_pass, app.pipeline)
-}
 
-end_frame :: proc(app: ^App) {
-	if app.render_pass != nil {
-		sdl.EndGPURenderPass(app.render_pass)
-		app.render_pass = nil
-	}
+	i := 0
+	for i < n {
+		run := texture_run_len(app.draw_list[:], i)
+		q0 := app.draw_list[i]
 
-	if app.cmd != nil {
-		ok := sdl.SubmitGPUCommandBuffer(app.cmd)
-		if !ok {
-			fmt.eprintfln("SubmitGPUCommandBuffer failed")
+		sampler_binding := sdl.GPUTextureSamplerBinding {
+			texture = q0.texture,
+			sampler = app.sampler,
 		}
-		app.cmd = nil
-	}
+		sdl.BindGPUFragmentSamplers(app.render_pass, 0, &sampler_binding, 1)
 
-	app.swapchain_texture = nil
+		vb_binding := sdl.GPUBufferBinding {
+			buffer = app.vertex_buffer,
+			offset = u32(i * SPRITE_VERTS_SIZE),
+		}
+		sdl.BindGPUVertexBuffers(app.render_pass, 0, &vb_binding, 1)
+
+		sdl.DrawGPUPrimitives(app.render_pass, u32(run * SPRITE_VERT_COUNT), 1, 0, 0)
+
+		i += run
+	}
+	sdl.EndGPURenderPass(app.render_pass)
+	app.render_pass = nil
+
+	if !sdl.SubmitGPUCommandBuffer(cmd) {
+		fmt.eprintfln("SubmitGPUCommandBuffer failed")
+	}
 }
 
 load_gpu_shader :: proc(
@@ -338,3 +440,17 @@ create_sprite_pipeline :: proc(app: ^App) -> ^sdl.GPUGraphicsPipeline {
 
 	return pipeline
 }
+
+texture_run_len :: proc(list: []Queued_Sprite, start: int) -> int {
+	if start < 0 || start >= len(list) do return 0
+
+	tex := list[start].texture
+	n := 1
+
+	for i in start + 1 ..< len(list) {
+		if list[i].texture != tex do break
+		n += 1
+	}
+
+	return n
+}

engine/batching_test.odin

@@ -0,0 +1,64 @@
+package engine
+
+import "core:testing"
+import sdl "vendor:sdl3"
+
+fake_tex :: proc(id: uintptr) -> ^sdl.GPUTexture {
+	return cast(^sdl.GPUTexture)id
+}
+
+@(test)
+texture_run_len_empty_or_oob :: proc(t: ^testing.T) {
+	testing.expect_value(t, texture_run_len(nil, 0), 0)
+	list := []Queued_Sprite{}
+	testing.expect_value(t, texture_run_len(list, 0), 0)
+	list = make([]Queued_Sprite, 1)
+	defer delete(list)
+	list[0] = {texture = fake_tex(1)}
+	testing.expect_value(t, texture_run_len(list, -1), 0)
+	testing.expect_value(t, texture_run_len(list, 1), 0)
+}
+
+@(test)
+texture_run_len_single :: proc(t: ^testing.T) {
+	list := make([]Queued_Sprite, 1)
+	defer delete(list)
+	list[0] = {texture = fake_tex(1)}
+	testing.expect_value(t, texture_run_len(list, 0), 1)
+}
+
+@(test)
+texture_run_len_same_texture :: proc(t: ^testing.T) {
+	tex := fake_tex(1)
+	list := make([]Queued_Sprite, 3)
+	defer delete(list)
+	list[0] = {texture = tex}
+	list[1] = {texture = tex}
+	list[2] = {texture = tex}
+	testing.expect_value(t, texture_run_len(list, 0), 3)
+}
+
+@(test)
+texture_run_len_breaks_on_change :: proc(t: ^testing.T) {
+	a := fake_tex(1)
+	b := fake_tex(2)
+	list := make([]Queued_Sprite, 3)
+	defer delete(list)
+	list[0] = {texture = a}
+	list[1] = {texture = a}
+	list[2] = {texture = b}
+	testing.expect_value(t, texture_run_len(list, 0), 2)
+	testing.expect_value(t, texture_run_len(list, 2), 1)
+}
+
+@(test)
+texture_run_len_all_different :: proc(t: ^testing.T) {
+	list := make([]Queued_Sprite, 3)
+	defer delete(list)
+	list[0] = {texture = fake_tex(1)}
+	list[1] = {texture = fake_tex(2)}
+	list[2] = {texture = fake_tex(3)}
+	testing.expect_value(t, texture_run_len(list, 0), 1)
+	testing.expect_value(t, texture_run_len(list, 1), 1)
+	testing.expect_value(t, texture_run_len(list, 2), 1)
+}

engine/camera.odin

@@ -0,0 +1,17 @@
+package engine
+
+Camera :: struct {
+	position: Vec2,
+	anchor:   Vec2,
+}
+
+camera_default :: proc() -> Camera {
+	return Camera{position = {0, 0}, anchor = {0.5, 0.5}}
+}
+
+world_to_screen :: proc(cam: Camera, world: Vec2, viewport: Vec2) -> Vec2 {
+	return {
+		world.x - cam.position.x + viewport.x * cam.anchor[0],
+		world.y - cam.position.y + viewport.y * cam.anchor[1],
+	}
+}

engine/camera_test.odin

@@ -0,0 +1,44 @@
+package engine
+
+import "core:testing"
+
+@(test)
+camera_default_values :: proc(t: ^testing.T) {
+	cam := camera_default()
+	testing.expect_value(t, cam.position, Vec2{0, 0})
+	testing.expect_value(t, cam.anchor, Vec2{0.5, 0.5})
+}
+
+@(test)
+world_to_screen_centered_identity :: proc(t: ^testing.T) {
+	cam := Camera {
+		position = {400, 500},
+		anchor   = {0.5, 0.5},
+	}
+	viewport := Vec2{800, 600}
+	screen := world_to_screen(cam, cam.position, viewport)
+	testing.expect_value(t, screen.x, f32(400))
+	testing.expect_value(t, screen.y, f32(300))
+}
+
+@(test)
+world_to_screen_origin_cam :: proc(t: ^testing.T) {
+	cam := Camera {
+		position = {0, 0},
+		anchor   = {0.5, 0.5},
+	}
+	screen := world_to_screen(cam, {10, 20}, {200, 100})
+	testing.expect_value(t, screen.x, f32(110))
+	testing.expect_value(t, screen.y, f32(70))
+}
+
+@(test)
+world_to_screen_top_left_anchor :: proc(t: ^testing.T) {
+	cam := Camera {
+		position = {5, 7},
+		anchor   = {0, 0},
+	}
+	screen := world_to_screen(cam, {15, 27}, {800, 600})
+	testing.expect_value(t, screen.x, f32(10))
+	testing.expect_value(t, screen.y, f32(20))
+}

engine/character_data.odin

@@ -210,6 +210,22 @@ character_clip :: proc(data: ^Character_Data, clip_name: string) -> (clip: Clip_
 	return character, true
 }
 
+character_frame :: proc(
+	data: ^Character_Data,
+	clip_name: string,
+	frame_index: int,
+) -> (
+	frame: Frame_Def,
+	ok: bool,
+) {
+	if data == nil do return {}, false
+	clip, found := data.def.clips[clip_name]
+	if !found || frame_index < 0 || frame_index >= len(clip.frames) {
+		return {}, false
+	}
+	return clip.frames[frame_index], true
+}
+
 character_frame_rect :: proc(
 	data: ^Character_Data,
 	clip_name: string,
@@ -218,10 +234,7 @@ character_frame_rect :: proc(
 	rect: [4]int,
 	ok: bool,
 ) {
-	clip, found := data.def.clips[clip_name]
-	if !found || frame_index < 0 || frame_index >= len(clip.frames) {
-		return {}, false
-	}
-
-	return clip.frames[frame_index].rect, true
+	frame, found := character_frame(data, clip_name, frame_index)
+	if !found do return {}, false
+	return frame.rect, true
 }

engine/character_data_test.odin

@@ -11,7 +11,7 @@ parse_char_def_happy_path :: proc(t: ^testing.T) {
 	def, ok := parse_char_def(src)
 	defer destroy_char_def(&def)
 
-	testing.expect(t, ok)
+	testing.expect(t, ok, "parse_char_def should succeed on valid JSON")
 	testing.expect_value(t, def.id, "toad")
 	testing.expect_value(t, def.atlas, "toad.atlas.png")
 	testing.expect_value(t, def.version, 1)
@@ -20,8 +20,8 @@ parse_char_def_happy_path :: proc(t: ^testing.T) {
 	testing.expect_value(t, def.pivot[1], f32(1.0))
 
 	idle, found := def.clips["idle"]
-	testing.expect(t, found)
-	testing.expect(t, idle.loop)
+	testing.expect(t, found, "expected idle clip in parsed def")
+	testing.expect(t, idle.loop, "idle clip should loop")
 	testing.expect_value(t, idle.fps, f32(8.0))
 	testing.expect_value(t, len(idle.frames), 1)
 	testing.expect_value(t, idle.frames[0].rect, [4]int{1, 2, 3, 4})
@@ -30,7 +30,7 @@ parse_char_def_happy_path :: proc(t: ^testing.T) {
 @(test)
 parse_char_def_bad_json :: proc(t: ^testing.T) {
 	_, ok := parse_char_def(transmute([]u8)string("{ not json"))
-	testing.expect(t, !ok)
+	testing.expect(t, !ok, "parse_char_def should fail on invalid JSON")
 }
 
 @(test)
@@ -51,17 +51,66 @@ character_frame_rect_hit_and_miss :: proc(t: ^testing.T) {
 	}
 
 	rect, ok := character_frame_rect(&data, "idle", 1)
-	testing.expect(t, ok)
+	testing.expect(t, ok, "character_frame_rect should find idle frame 1")
 	testing.expect_value(t, rect, [4]int{50, 60, 70, 80})
 
 	_, ok = character_frame_rect(&data, "missing", 0)
-	testing.expect(t, !ok)
+	testing.expect(t, !ok, "character_frame_rect should miss unknown clip")
 
 	_, ok = character_frame_rect(&data, "idle", -1)
-	testing.expect(t, !ok)
+	testing.expect(t, !ok, "character_frame_rect should reject negative index")
 
 	_, ok = character_frame_rect(&data, "idle", 2)
-	testing.expect(t, !ok)
+	testing.expect(t, !ok, "character_frame_rect should reject out-of-range index")
+}
+
+@(test)
+character_frame_returns_full_def :: proc(t: ^testing.T) {
+	data: Character_Data
+	data.def.clips = make(map[string]Clip_Def)
+	defer delete(data.def.clips)
+
+	frames := make([]Frame_Def, 1)
+	frames[0] = Frame_Def {
+		rect        = {10, 20, 30, 40},
+		source_size = {100, 200},
+		trim_offset = {5, 7},
+	}
+	defer delete(frames)
+
+	data.def.clips["idle"] = Clip_Def {
+		loop   = true,
+		fps    = 10,
+		frames = frames,
+	}
+
+	frame, ok := character_frame(&data, "idle", 0)
+	testing.expect(t, ok, "character_frame should find idle frame 0")
+	testing.expect_value(t, frame.rect, [4]int{10, 20, 30, 40})
+	testing.expect_value(t, frame.source_size, [2]int{100, 200})
+	testing.expect_value(t, frame.trim_offset, [2]int{5, 7})
+}
+
+@(test)
+character_frame_nil_and_bad_index :: proc(t: ^testing.T) {
+	_, ok := character_frame(nil, "idle", 0)
+	testing.expect(t, !ok, "character_frame should fail on nil data")
+
+	data: Character_Data
+	data.def.clips = make(map[string]Clip_Def)
+	defer delete(data.def.clips)
+
+	frames := make([]Frame_Def, 1)
+	frames[0] = {rect = {1, 2, 3, 4}}
+	defer delete(frames)
+	data.def.clips["idle"] = Clip_Def{frames = frames}
+
+	_, ok = character_frame(&data, "idle", -1)
+	testing.expect(t, !ok, "character_frame should reject negative index")
+	_, ok = character_frame(&data, "idle", 1)
+	testing.expect(t, !ok, "character_frame should reject out-of-range index")
+	_, ok = character_frame(&data, "missing", 0)
+	testing.expect(t, !ok, "character_frame should miss unknown clip")
 }
 
 destroy_char_def :: proc(def: ^Char_Def) {

engine/input.odin

@@ -0,0 +1,43 @@
+package engine
+
+import sdl "vendor:sdl3"
+
+Key :: enum {
+	A,
+	D,
+	Left,
+	Right,
+	Up,
+	Down,
+	Space,
+	N1,
+	N2,
+}
+
+key_down :: proc(key: Key) -> bool {
+	keys := sdl.GetKeyboardState(nil)
+	if keys == nil do return false
+
+	scan_code: sdl.Scancode
+	switch key {
+	case .A:
+		scan_code = .A
+	case .D:
+		scan_code = .D
+	case .Left:
+		scan_code = .LEFT
+	case .Right:
+		scan_code = .RIGHT
+	case .Up:
+		scan_code = .UP
+	case .Down:
+		scan_code = .DOWN
+	case .Space:
+		scan_code = .SPACE
+	case .N1:
+		scan_code = ._1
+	case .N2:
+		scan_code = ._2
+	}
+	return keys[scan_code]
+}

engine/shader_backend_test.odin

@@ -21,7 +21,7 @@ shader_filenames_per_backend :: proc(t: ^testing.T) {
 @(test)
 choose_shader_runtime_prefers_msl :: proc(t: ^testing.T) {
 	rt, ok := choose_shader_runtime_from_formats({.MSL, .SPIRV, .DXIL})
-	testing.expect(t, ok)
+	testing.expect(t, ok, "should pick a runtime when MSL is available")
 	testing.expect_value(t, rt.backend, Shader_Backend.Metal_MSL)
 	testing.expect_value(t, rt.shader_dir, "shaders/metal")
 	testing.expect_value(t, rt.format, sdl.GPUShaderFormat{.MSL})
@@ -30,7 +30,7 @@ choose_shader_runtime_prefers_msl :: proc(t: ^testing.T) {
 @(test)
 choose_shader_runtime_spirv_only :: proc(t: ^testing.T) {
 	rt, ok := choose_shader_runtime_from_formats({.SPIRV})
-	testing.expect(t, ok)
+	testing.expect(t, ok, "should pick SPIR-V when it is the only format")
 	testing.expect_value(t, rt.backend, Shader_Backend.Vulkan_SPIRV)
 	testing.expect_value(t, rt.shader_dir, "shaders/vulkan")
 	testing.expect_value(t, rt.format, sdl.GPUShaderFormat{.SPIRV})
@@ -39,5 +39,5 @@ choose_shader_runtime_spirv_only :: proc(t: ^testing.T) {
 @(test)
 choose_shader_runtime_empty_fails :: proc(t: ^testing.T) {
 	_, ok := choose_shader_runtime_from_formats({})
-	testing.expect(t, !ok)
+	testing.expect(t, !ok, "empty format set should fail")
 }

engine/sprite.odin

@@ -1,6 +1,5 @@
 package engine
 
-import "core:mem"
 import sdl "vendor:sdl3"
 
 Vec2 :: [2]f32
@@ -11,6 +10,7 @@ Sprite :: struct {
 	clip:     string,
 	frame:    int,
 	time:     f32,
+	flip_x:   bool,
 }
 
 spawn_sprite :: proc(
@@ -39,7 +39,6 @@ spawn_sprite :: proc(
 	return sprite
 }
 
-// stubbed for later
 update_sprite :: proc(sprite: ^Sprite, dt: f32) {
 	if sprite == nil || sprite.data == nil do return
 	if dt <= 0 do return
@@ -84,43 +83,57 @@ to_clip :: proc(px, py, sw, sh: f32) -> [2]f32 {
 }
 
 draw_sprite :: proc(app: ^App, sprite: ^Sprite) {
-	if app.render_pass == nil || app.cmd == nil || app.swapchain_texture == nil {
-		return // begin_frame may have skipped (minimized window, etc.)
+	if app.cmd == nil || app.swapchain_texture == nil {
+		return
 	}
 	if sprite == nil || sprite.data == nil || sprite.data.texture == nil {
 		return
 	}
+	if len(app.draw_list) >= MAX_SPRITES {
+		return
+	}
 
-	rect, ok := character_frame_rect(sprite.data, sprite.clip, sprite.frame)
+	frame, ok := character_frame(sprite.data, sprite.clip, sprite.frame)
 	if !ok {
 		return
 	}
 
-	fw := f32(rect[2])
-	fh := f32(rect[3])
+	src_w := f32(frame.source_size[0])
+	src_h := f32(frame.source_size[1])
+	if src_w <= 0 do src_w = f32(frame.rect[2])
+	if src_h <= 0 do src_h = f32(frame.rect[3])
 
-	// Feet-centered placement, same as renderer version
-	x0_px := sprite.position.x - fw * 0.5
-	y0_px := sprite.position.y - fh
-	x1_px := x0_px + fw
-	y1_px := y0_px + fh
+	fw := f32(frame.rect[2])
+	fh := f32(frame.rect[3])
+	trim_x := f32(frame.trim_offset[0])
+	trim_y := f32(frame.trim_offset[1])
+	pivot := sprite.data.def.pivot
+
+	viewport := Vec2{f32(app.swapchain_w), f32(app.swapchain_h)}
+	feet := world_to_screen(app.camera, sprite.position, viewport)
+
+	x0_px, y0_px, x1_px, y1_px := sprite_feet_quad(
+		feet,
+		src_w,
+		src_h,
+		{trim_x, trim_y},
+		{fw, fh},
+		pivot,
+		sprite.flip_x,
+	)
 
 	sw := f32(app.swapchain_w)
 	sh := f32(app.swapchain_h)
-	p0 := to_clip(x0_px, y0_px, sw, sh) // top-left
-	p1 := to_clip(x1_px, y0_px, sw, sh) // top-right
-	p2 := to_clip(x1_px, y1_px, sw, sh) // bottom-right
-	p3 := to_clip(x0_px, y1_px, sw, sh) // bottom-left
+	p0 := to_clip(x0_px, y0_px, sw, sh)
+	p1 := to_clip(x1_px, y0_px, sw, sh)
+	p2 := to_clip(x1_px, y1_px, sw, sh)
+	p3 := to_clip(x0_px, y1_px, sw, sh)
 
 	tex_w := f32(sprite.data.width)
 	tex_h := f32(sprite.data.height)
-	u0 := f32(rect[0]) / tex_w
-	v0 := f32(rect[1]) / tex_h
-	u1 := f32(rect[0] + rect[2]) / tex_w
-	v1 := f32(rect[1] + rect[3]) / tex_h
+	u0, v0, u1, v1 := frame_uvs(frame.rect, tex_w, tex_h, sprite.flip_x)
 
-	// Two triangles: (0,1,2) and (0,2,3)
-	verts := [6]Vertex {
+	verts := [SPRITE_VERT_COUNT]Vertex {
 		{pos = p0, uv = {u0, v0}},
 		{pos = p1, uv = {u1, v0}},
 		{pos = p2, uv = {u1, v1}},
@@ -129,54 +142,7 @@ draw_sprite :: proc(app: ^App, sprite: ^Sprite) {
 		{pos = p3, uv = {u0, v1}},
 	}
 
-	map_ptr := sdl.MapGPUTransferBuffer(app.device, app.transfer_buffer, false)
-	if map_ptr == nil {
-		return
-	}
-	mem.copy(map_ptr, raw_data(verts[:]), size_of(verts))
-	sdl.UnmapGPUTransferBuffer(app.device, app.transfer_buffer)
-
-	// IMPORTANT: SDL does not allow beginning a copy pass while a render pass
-	// is active on the same command buffer. Upload on a separate command buffer.
-	copy_cmd := sdl.AcquireGPUCommandBuffer(app.device)
-	if copy_cmd == nil {
-		return
-	}
-	copy_pass := sdl.BeginGPUCopyPass(copy_cmd)
-
-	src := sdl.GPUTransferBufferLocation {
-		transfer_buffer = app.transfer_buffer,
-		offset          = 0,
-	}
-	dst := sdl.GPUBufferRegion {
-		buffer = app.vertex_buffer,
-		offset = 0,
-		size   = u32(size_of(verts)),
-	}
-	sdl.UploadToGPUBuffer(copy_pass, src, dst, false)
-	sdl.EndGPUCopyPass(copy_pass)
-	fence := sdl.SubmitGPUCommandBufferAndAcquireFence(copy_cmd)
-	if fence == nil {
-		return
-	}
-	defer sdl.ReleaseGPUFence(app.device, fence)
-	if !sdl.WaitForGPUFences(app.device, true, &fence, 1) {
-		return
-	}
-
-	sampler_binding := sdl.GPUTextureSamplerBinding {
-		texture = sprite.data.texture,
-		sampler = app.sampler,
-	}
-	sdl.BindGPUFragmentSamplers(app.render_pass, 0, &sampler_binding, 1)
-
-	vb_binding := sdl.GPUBufferBinding {
-		buffer = app.vertex_buffer,
-		offset = 0,
-	}
-	sdl.BindGPUVertexBuffers(app.render_pass, 0, &vb_binding, 1)
-
-	sdl.DrawGPUPrimitives(app.render_pass, 6, 1, 0, 0)
+	append(&app.draw_list, Queued_Sprite{texture = sprite.data.texture, verts = verts})
 }
 
 set_sprite_clip :: proc(sprite: ^Sprite, clip: string) {
@@ -191,3 +157,46 @@ set_sprite_clip :: proc(sprite: ^Sprite, clip: string) {
 	sprite.frame = 0
 	sprite.time = 0
 }
+
+sprite_quad_origin :: proc(position: Vec2, size: Vec2, pivot: [2]f32) -> Vec2 {
+	return {position.x - size.x * pivot[0], position.y - size.y * pivot[1]}
+}
+
+frame_uvs :: proc(rect: [4]int, tex_w, tex_h: f32, flip_x: bool) -> (u0, v0, u1, v1: f32) {
+	u0 = f32(rect[0]) / tex_w
+	v0 = f32(rect[1]) / tex_h
+	u1 = f32(rect[0] + rect[2]) / tex_w
+	v1 = f32(rect[1] + rect[3]) / tex_h
+	if flip_x do u0, u1 = u1, u0
+	return
+}
+
+sprite_feet_quad :: proc(
+	feet: Vec2,
+	src_w, src_h: f32,
+	trim: Vec2,
+	size: Vec2,
+	pivot: [2]f32,
+	flip_x: bool,
+) -> (
+	x0, y0, x1, y1: f32,
+) {
+	canvas_top := feet.y - src_h * pivot[1]
+	if flip_x {
+		canvas_left := feet.x - (1.0 - pivot[0]) * src_w
+		x0 = canvas_left + (src_w - trim.x - size.x)
+		y0 = canvas_top + trim.y
+	} else {
+		canvas_left := feet.x - src_w * pivot[0]
+		x0 = canvas_left + trim.x
+		y0 = canvas_top + trim.y
+	}
+	x1 = x0 + size.x
+	y1 = y0 + size.y
+	return
+}
+
+set_sprite_flip_x :: proc(sprite: ^Sprite, flip: bool) {
+	if sprite == nil do return
+	sprite.flip_x = flip
+}

engine/sprite_test.odin

@@ -19,3 +19,73 @@ to_clip_corners :: proc(t: ^testing.T) {
 	testing.expect_value(t, p[0], f32(0))
 	testing.expect_value(t, p[1], f32(0))
 }
+
+@(test)
+sprite_quad_origin_feet :: proc(t: ^testing.T) {
+	o := sprite_quad_origin({400, 500}, {100, 200}, {0.5, 1.0})
+	testing.expect_value(t, o.x, f32(350))
+	testing.expect_value(t, o.y, f32(300))
+}
+
+@(test)
+sprite_quad_origin_center :: proc(t: ^testing.T) {
+	o := sprite_quad_origin({400, 500}, {100, 200}, {0.5, 0.5})
+	testing.expect_value(t, o.x, f32(350))
+	testing.expect_value(t, o.y, f32(400))
+}
+
+@(test)
+frame_uvs_unflipped :: proc(t: ^testing.T) {
+	u0, v0, u1, v1 := frame_uvs({10, 20, 30, 40}, 100, 200, false)
+	testing.expect_value(t, u0, f32(0.1))
+	testing.expect_value(t, v0, f32(0.1))
+	testing.expect_value(t, u1, f32(0.4))
+	testing.expect_value(t, v1, f32(0.3))
+	testing.expect(t, u0 < u1, "unflipped UVs should have u0 < u1")
+}
+
+@(test)
+frame_uvs_flipped :: proc(t: ^testing.T) {
+	a0, _, a1, _ := frame_uvs({10, 20, 30, 40}, 100, 200, false)
+	b0, v0, b1, v1 := frame_uvs({10, 20, 30, 40}, 100, 200, true)
+	testing.expect_value(t, b0, a1)
+	testing.expect_value(t, b1, a0)
+	testing.expect_value(t, v0, f32(0.1))
+	testing.expect_value(t, v1, f32(0.3))
+}
+
+@(test)
+sprite_feet_quad_no_flip :: proc(t: ^testing.T) {
+	// feet (400,500), source 100x200, pivot feet, trim (10,20), packed 80x160
+	x0, y0, x1, y1 := sprite_feet_quad(
+		{400, 500},
+		100,
+		200,
+		{10, 20},
+		{80, 160},
+		{0.5, 1.0},
+		false,
+	)
+	testing.expect_value(t, x0, f32(360)) // 400 - 50 + 10
+	testing.expect_value(t, y0, f32(320)) // 500 - 200 + 20
+	testing.expect_value(t, x1, f32(440))
+	testing.expect_value(t, y1, f32(480))
+}
+
+@(test)
+sprite_feet_quad_flip_x :: proc(t: ^testing.T) {
+	x0, y0, x1, y1 := sprite_feet_quad(
+		{400, 500},
+		100,
+		200,
+		{5, 20},
+		{80, 160},
+		{0.5, 1.0},
+		true,
+	)
+	// canvas_left = 350; trim_x_draw = 100 - 5 - 80 = 15 → x0 = 365
+	testing.expect_value(t, x0, f32(365))
+	testing.expect_value(t, y0, f32(320))
+	testing.expect_value(t, x1, f32(445))
+	testing.expect_value(t, y1, f32(480))
+}

examples/camera_sandbox/main.odin

@@ -0,0 +1,58 @@
+package main
+
+import eng "pkg:engine"
+
+// Camera sandbox: toad stays put; arrows pan the camera. Space toggles follow mode.
+main :: proc() {
+	app: eng.App
+	if !eng.init(&app, "camera sandbox", 800, 600) do return
+	defer eng.shutdown(&app)
+
+	data, ok := eng.load_character_data(&app, "assets_baked/characters/toad/toad.char.json")
+	if !ok do return
+	defer eng.destroy_character_data(&app, &data)
+
+	toad := eng.spawn_sprite(&data, {0, 0}, "idle", 0)
+	app.camera.position = {0, -100}
+
+	follow := false
+	space_was_down := false
+	CAM_SPEED :: f32(300)
+
+	last := eng.now_seconds()
+
+	for eng.events() {
+		now := eng.now_seconds()
+		dt := f32(now - last)
+		last = now
+
+		space := eng.key_down(.Space)
+		if space && !space_was_down {
+			follow = !follow
+		}
+		space_was_down = space
+
+		if follow {
+			app.camera.position = toad.position - {0, 100}
+		} else {
+			if eng.key_down(.Left) || eng.key_down(.A) {
+				app.camera.position.x -= CAM_SPEED * dt
+			}
+			if eng.key_down(.Right) || eng.key_down(.D) {
+				app.camera.position.x += CAM_SPEED * dt
+			}
+			if eng.key_down(.Up) {
+				app.camera.position.y -= CAM_SPEED * dt
+			}
+			if eng.key_down(.Down) {
+				app.camera.position.y += CAM_SPEED * dt
+			}
+		}
+
+		eng.update_sprite(&toad, dt)
+
+		eng.begin_frame(&app)
+		eng.draw_sprite(&app, &toad)
+		eng.end_frame(&app)
+	}
+}

examples/clips/main.odin

@@ -0,0 +1,38 @@
+package main
+
+import eng "pkg:engine"
+
+// Clip switcher: 1 = idle, 2 = walk. No movement — animation policy only.
+main :: proc() {
+	app: eng.App
+	if !eng.init(&app, "clips", 800, 600) do return
+	defer eng.shutdown(&app)
+
+	data, ok := eng.load_character_data(&app, "assets_baked/characters/toad/toad.char.json")
+	if !ok do return
+	defer eng.destroy_character_data(&app, &data)
+
+	toad := eng.spawn_sprite(&data, {400, 500}, "idle", 0)
+	app.camera.position = toad.position - {0, 100}
+
+	last := eng.now_seconds()
+
+	for eng.events() {
+		now := eng.now_seconds()
+		dt := f32(now - last)
+		last = now
+
+		if eng.key_down(.N1) {
+			eng.set_sprite_clip(&toad, "idle")
+		}
+		if eng.key_down(.N2) {
+			eng.set_sprite_clip(&toad, "walk")
+		}
+
+		eng.update_sprite(&toad, dt)
+
+		eng.begin_frame(&app)
+		eng.draw_sprite(&app, &toad)
+		eng.end_frame(&app)
+	}
+}

examples/crowd/main.odin

@@ -0,0 +1,49 @@
+package main
+
+import eng "pkg:engine"
+
+// Many sprites sharing one Character_Data (Flyweight) — good batching demo.
+COUNT :: 24
+
+main :: proc() {
+	app: eng.App
+	if !eng.init(&app, "crowd", 800, 600) do return
+	defer eng.shutdown(&app)
+
+	data, ok := eng.load_character_data(&app, "assets_baked/characters/toad/toad.char.json")
+	if !ok do return
+	defer eng.destroy_character_data(&app, &data)
+
+	sprites: [COUNT]eng.Sprite
+	for i in 0 ..< COUNT {
+		col := i % 8
+		row := i / 8
+		pos := eng.Vec2 {
+			f32(120 + col * 80),
+			f32(280 + row * 120),
+		}
+		clip := "idle" if (i % 2) == 0 else "walk"
+		sprites[i] = eng.spawn_sprite(&data, pos, clip, i % 5)
+	}
+
+	// Look at the middle of the grid
+	app.camera.position = {400, 400}
+
+	last := eng.now_seconds()
+
+	for eng.events() {
+		now := eng.now_seconds()
+		dt := f32(now - last)
+		last = now
+
+		for &s in sprites {
+			eng.update_sprite(&s, dt)
+		}
+
+		eng.begin_frame(&app)
+		for &s in sprites {
+			eng.draw_sprite(&app, &s)
+		}
+		eng.end_frame(&app)
+	}
+}

examples/hello_sprite/main.odin

@@ -0,0 +1,34 @@
+package main
+
+import eng "pkg:engine"
+
+// Minimal: load one baked character and draw idle. No input, no camera follow.
+main :: proc() {
+	app: eng.App
+	if !eng.init(&app, "hello sprite", 800, 600) do return
+	defer eng.shutdown(&app)
+
+	data, ok := eng.load_character_data(&app, "assets_baked/characters/toad/toad.char.json")
+	if !ok do return
+	defer eng.destroy_character_data(&app, &data)
+
+	sprite := eng.spawn_sprite(&data, {400, 500}, "idle", 0)
+
+	// Screen-centered framing without gameplay camera follow:
+	// treat spawn position as world; look slightly above feet.
+	app.camera.position = sprite.position - {0, 100}
+
+	last := eng.now_seconds()
+
+	for eng.events() {
+		now := eng.now_seconds()
+		dt := f32(now - last)
+		last = now
+
+		eng.update_sprite(&sprite, dt)
+
+		eng.begin_frame(&app)
+		eng.draw_sprite(&app, &sprite)
+		eng.end_frame(&app)
+	}
+}

examples/toad/main.odin

@@ -15,12 +15,32 @@ main :: proc() {
 	toad := eng.spawn_sprite(&data, {400, 500}, "idle", 0)
 
 	last := eng.now_seconds()
+	SPEED :: f32(200)
 
 	for eng.events() {
 		now := eng.now_seconds()
 		dt := f32(now - last)
 		last = now
 
+		left := eng.key_down(.A) || eng.key_down(.Left)
+		right := eng.key_down(.D) || eng.key_down(.Right)
+
+		if left || right {
+			eng.set_sprite_clip(&toad, "walk")
+			if left {
+				toad.position.x -= SPEED * dt
+				toad.flip_x = true
+			}
+
+			if right {
+				toad.position.x += SPEED * dt
+				toad.flip_x = false
+			}
+		} else {
+			eng.set_sprite_clip(&toad, "idle")
+		}
+
+		app.camera.position = toad.position - {0, 100}
 		eng.update_sprite(&toad, dt)
 
 		eng.begin_frame(&app)

flame/crowd-20260803-003404-report.txt

@@ -0,0 +1,645 @@
+# To display the perf.data header info, please use --header/--header-only options.
+#
+#
+# Total Lost Samples: 0
+#
+# Samples: 499  of event 'cpu/cycles/Pu'
+# Event count (approx.): 2684910393
+#
+# Overhead  Command          Shared Object                  Symbol                                     
+# ........  ...............  .............................  ...........................................
+#
+     9.56%  crowd_perf       crowd_perf                     [.] engine::draw_sprite
+            |
+            ---engine::draw_sprite
+               main::main
+               main
+               0x7fe3b6c27740
+               __libc_start_main
+               _start
+
+     5.33%  crowd_perf       crowd_perf                     [.] engine::to_clip
+            |
+            ---engine::to_clip
+               engine::draw_sprite
+               main::main
+               main
+               0x7fe3b6c27740
+               __libc_start_main
+               _start
+
+     3.98%  crowd_perf       crowd_perf                     [.] engine::sprite_feet_quad
+            |
+            ---engine::sprite_feet_quad
+               engine::draw_sprite
+               main::main
+               main
+               0x7fe3b6c27740
+               __libc_start_main
+               _start
+
+     3.01%  crowd_perf       crowd_perf                     [.] engine::update_sprite
+            |
+            ---engine::update_sprite
+               main::main
+               main
+               0x7fe3b6c27740
+               __libc_start_main
+               _start
+
+     2.56%  crowd_perf       crowd_perf                     [.] runtime::default_hasher
+            |
+            ---runtime::default_hasher
+               runtime::default_hasher_string
+               __$hasher$$string
+               |          
+               |--1.36%--engine::character_frame
+               |          engine::draw_sprite
+               |          main::main
+               |          main
+               |          0x7fe3b6c27740
+               |          __libc_start_main
+               |          _start
+               |          
+                --1.20%--engine::character_clip
+                          engine::update_sprite
+                          main::main
+                          main
+                          0x7fe3b6c27740
+                          __libc_start_main
+                          _start
+
+     2.49%  crowd_perf       libc.so.6                      [.] 0x000000000018a4c4
+            |
+            ---0x7fe3b6d8a4c4
+               engine::draw_sprite
+               main::main
+               main
+               0x7fe3b6c27740
+               __libc_start_main
+               _start
+
+     2.47%  crowd_perf       crowd_perf                     [.] runtime::_append_elem
+            |
+            ---runtime::_append_elem
+               runtime::append_elem:proc(array:^[dynamic]engine::Queued_Sprite,arg:engine::Queued_Sprite,loc:runtime::Source_Code_Location)->(n:int,err:runtime::Allocator_Error)
+               engine::draw_sprite
+               main::main
+               main
+               0x7fe3b6c27740
+               __libc_start_main
+               _start
+
+     1.97%  crowd_perf       crowd_perf                     [.] runtime::bounds_check_error
+            |
+            ---runtime::bounds_check_error
+               engine::character_frame
+               engine::draw_sprite
+               main::main
+               main
+               0x7fe3b6c27740
+               __libc_start_main
+               _start
+
+     1.70%  crowd_perf       libwayland-client.so.0.25.0    [.] wl_proxy_marshal_array_flags
+            |
+            ---wl_proxy_marshal_array_flags
+               wl_proxy_marshal_flags
+               |          
+                --0.87%--0x7fe396d97539
+                          0x7fe396d8a4f8
+                          0x7fe3b724ffc9
+                          engine::end_frame
+                          main::main
+                          main
+                          0x7fe3b6c27740
+                          __libc_start_main
+                          _start
+
+     1.61%  crowd_perf       crowd_perf                     [.] __$map_get$$map[string]engine::Clip_Def
+            |
+            ---__$map_get$$map[string]engine::Clip_Def
+               |          
+               |--1.00%--engine::character_frame
+               |          engine::draw_sprite
+               |          main::main
+               |          main
+               |          0x7fe3b6c27740
+               |          __libc_start_main
+               |          _start
+               |          
+                --0.61%--engine::character_clip
+                          engine::update_sprite
+                          main::main
+                          main
+                          0x7fe3b6c27740
+                          __libc_start_main
+                          _start
+
+     1.57%  wl_cursor_surfa  libc.so.6                      [.] 0x000000000009ca5c
+     1.50%  crowd_perf       crowd_perf                     [.] engine::world_to_screen
+            |
+            ---engine::world_to_screen
+               engine::draw_sprite
+               main::main
+               main
+               0x7fe3b6c27740
+               __libc_start_main
+               _start
+
+     1.35%  crowd_perf       crowd_perf                     [.] runtime::memory_equal
+            |
+            ---runtime::memory_equal
+               runtime::string_eq
+               __$map_get$$map[string]engine::Clip_Def
+               |          
+                --1.06%--engine::character_clip
+                          engine::update_sprite
+                          main::main
+                          main
+                          0x7fe3b6c27740
+                          __libc_start_main
+                          _start
+
+     1.33%  crowd_perf       libc.so.6                      [.] pthread_mutex_lock
+            |
+            ---pthread_mutex_lock
+
+     1.31%  crowd_perf       crowd_perf                     [.] engine::character_frame
+            |
+            ---engine::character_frame
+               engine::draw_sprite
+               main::main
+               main
+               0x7fe3b6c27740
+               __libc_start_main
+               _start
+
+     1.19%  crowd_perf       crowd_perf                     [.] runtime::string_eq
+            |
+            ---runtime::string_eq
+               __$map_get$$map[string]engine::Clip_Def
+               |          
+               |--0.62%--engine::character_frame
+               |          engine::draw_sprite
+               |          main::main
+               |          main
+               |          0x7fe3b6c27740
+               |          __libc_start_main
+               |          _start
+               |          
+                --0.57%--engine::character_clip
+                          engine::update_sprite
+                          main::main
+                          main
+                          0x7fe3b6c27740
+                          __libc_start_main
+                          _start
+
+     1.18%  crowd_perf       [unknown]                      [k] 0xffffffffab355a81
+            |
+            ---0xffffffffab355a81
+               0xffffffffa9e0012f
+               ioctl
+               drmIoctl
+               |          
+                --0.59%--drmSyncobjCreate
+                          0x7fe396d8c878
+                          0x7fe396d893c8
+                          0x7fe396d89241
+                          0x7fe3b72544bd
+                          0x7fe3b7069860
+                          engine::begin_frame
+                          main::main
+                          main
+                          0x7fe3b6c27740
+                          __libc_start_main
+                          _start
+
+     1.11%  crowd_perf       libc.so.6                      [.] clock_gettime
+            |
+            ---clock_gettime
+               |          
+                --0.68%--0x7fe396f7a09a
+
+     1.08%  crowd_perf       libc.so.6                      [.] 0x0000000000189c40
+            |
+            ---0x7fe3b6d89c40
+               0x7fe3b6e6ee65
+               wl_proxy_marshal_array_flags
+               wl_proxy_marshal_flags
+               0x7fe396d97614
+               0x7fe396d8a4f8
+               0x7fe3b724ffc9
+               engine::end_frame
+               main::main
+               main
+               0x7fe3b6c27740
+               __libc_start_main
+               _start
+
+     1.03%  crowd_perf       crowd_perf                     [.] main::main
+            |
+            ---main::main
+               main
+               0x7fe3b6c27740
+               __libc_start_main
+               _start
+
+     0.94%  crowd_perf       libc.so.6                      [.] ioctl
+            |
+            ---ioctl
+               |          
+                --0.94%--drmIoctl
+
+     0.91%  crowd_perf       libSDL3.so.0.4.12              [.] 0x00000000001428a1
+            |
+            ---0x7fe3b71428a1
+               0x7fe3b714928f
+               0x7fe3b7141b9a
+               0x7fe3b71426da
+
+     0.89%  crowd_perf       ld-linux-x86-64.so.2           [.] 0x000000000000ac3a
+            |
+            ---0x7fe3b739bc3a
+               0x7fe3b739c855
+               0x7fe3b73a0f57
+               0x7fe3b73a2e00
+               0x7fe3b739eeda
+               _dl_catch_exception
+               0x7fe3b739e242
+               _dl_catch_exception
+               0x7fe3b739e7a9
+               0x7fe3b6c93bb3
+               _dl_catch_exception
+               0x7fe3b73935c8
+               0x7fe3b6c936a2
+               dlopen
+               0x7fe3b61b2c10
+               0x7fe3b61b7b4b
+
+     0.85%  crowd_perf       libc.so.6                      [.] 0x000000000018a4c0
+            |
+            ---0x7fe3b6d8a4c0
+               engine::draw_sprite
+               main::main
+               main
+               0x7fe3b6c27740
+               __libc_start_main
+               _start
+
+     0.84%  crowd_perf       libSDL3.so.0.4.12              [.] 0x00000000000691d6
+            |
+            ---0x7fe3b70691d6
+               engine::end_frame
+               main::main
+               main
+               0x7fe3b6c27740
+               __libc_start_main
+               _start
+
+     0.76%  crowd_perf       libc.so.6                      [.] pthread_rwlock_wrlock
+            |
+            ---pthread_rwlock_wrlock
+
+     0.74%  crowd_perf       libSDL3.so.0.4.12              [.] 0x0000000000142863
+            |
+            ---0x7fe3b7142863
+               0x7fe3b71429b4
+               0x7fe3b714921a
+               0x7fe3b7141b9a
+               0x7fe3b71426da
+
+     0.74%  crowd_perf       libSDL3.so.0.4.12              [.] 0x0000000000149304
+            |
+            ---0x7fe3b7149304
+               0x7fe3b7141b9a
+               0x7fe3b71426da
+
+     0.74%  crowd_perf       libc.so.6                      [.] 0x000000000018a0e8
+            |
+            ---0x7fe3b6d8a0e8
+               mem::copy
+               engine::load_character_data
+               main::main
+               main
+               0x7fe3b6c27740
+               __libc_start_main
+               _start
+
+     0.74%  crowd_perf       libc.so.6                      [.] 0x000000000009ca5c
+            |
+            ---0x7fe3b6c9ca5c
+
+     0.73%  crowd_perf       libSDL3.so.0.4.12              [.] 0x0000000000143ffe
+            |
+            ---0x7fe3b7143ffe
+               0x7fe3b714a028
+               0x7fe3b7141b9a
+
+     0.72%  crowd_perf       libSDL3.so.0.4.12              [.] 0x0000000000143fe9
+            |
+            ---0x7fe3b7143fe9
+               0x7fe3b714a028
+               0x7fe3b7141b9a
+
+     0.71%  crowd_perf       libwayland-client.so.0.25.0    [.] 0x0000000000003263
+            |
+            ---0x7fe3b6e6d263
+               wl_proxy_marshal_flags
+               0x7fe396d97937
+               0x7fe396d8a4f8
+               0x7fe3b724ffc9
+               engine::end_frame
+               main::main
+               main
+               0x7fe3b6c27740
+               __libc_start_main
+               _start
+
+     0.70%  crowd_perf       crowd_perf                     [.] runtime::default_hasher_string
+            |
+            ---runtime::default_hasher_string
+               __$hasher$$string
+
+     0.69%  crowd_perf       libdrm.so.2.134.0              [.] drmSyncobjTimelineWait
+            |
+            ---drmSyncobjTimelineWait
+
+     0.68%  crowd_perf       libc.so.6                      [.] 0x00000000000a590c
+            |
+            ---0x7fe3b6ca590c
+               0x7fe396d827d6
+               0x7fe396d3cff4
+               0x7fe396ddcaf6
+               0x7fe396ddccd4
+               0x7fe396de6f2c
+               0x7fe3b724fec1
+               engine::end_frame
+               main::main
+               main
+               0x7fe3b6c27740
+               __libc_start_main
+               _start
+
+     0.66%  crowd_perf       libSDL3.so.0.4.12              [.] 0x000000000024d030
+            |
+            ---0x7fe3b724d030
+               0x7fe3b724f5c0
+               0x7fe3b72545c5
+               0x7fe3b7069860
+               engine::begin_frame
+               main::main
+               main
+               0x7fe3b6c27740
+               __libc_start_main
+               _start
+
+     0.66%  crowd_perf       libc.so.6                      [.] 0x000000000018a547
+            |
+            ---0x7fe3b6d8a547
+
+     0.65%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000000d27b4
+            |
+            ---0x7fe396cd27b4
+               0x7fe396de6325
+               0x7fe3b72546f1
+               0x7fe3b7069860
+               engine::begin_frame
+               main::main
+               main
+               0x7fe3b6c27740
+               __libc_start_main
+               _start
+
+     0.64%  crowd_perf       libvulkan_radeon.so            [.] 0x0000000000022d88
+            |
+            ---0x7fe396c22d88
+               0x7fe396cd2f6f
+               0x7fe396de6325
+
+     0.61%  crowd_perf       crowd_perf                     [.] engine::texture_run_len
+            |
+            ---engine::texture_run_len
+               engine::end_frame
+               main::main
+               main
+               0x7fe3b6c27740
+               __libc_start_main
+               _start
+
+     0.55%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000000bbf60
+            |
+            ---0x7fe396cbbf60
+               0x7fe396cc6c98
+               0x7fe396cc985a
+               0x7fe396cd7932
+               engine::end_frame
+               main::main
+               main
+               0x7fe3b6c27740
+               __libc_start_main
+               _start
+
+     0.54%  crowd_perf       libvulkan_radeon.so            [.] 0x000000000022daad
+            |
+            ---0x7fe396e2daad
+               0x7fe396d15bcb
+               0x7fe396d162ff
+
+     0.53%  crowd_perf       libSDL3.so.0.4.12              [.] 0x000000000023af6a
+            |
+            ---0x7fe3b723af6a
+               0x7fe3b7249f38
+               engine::end_frame
+               main::main
+               main
+               0x7fe3b6c27740
+               __libc_start_main
+               _start
+
+     0.48%  crowd_perf       libwayland-client.so.0.25.0    [.] wl_display_flush
+     0.48%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000000cf827
+     0.47%  crowd_perf       libSDL3.so.0.4.12              [.] 0x00000000000166ae
+     0.44%  crowd_perf       libvulkan_radeon.so            [.] 0x0000000000022daa
+     0.43%  crowd_perf       libc.so.6                      [.] cfree
+     0.43%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000000d2c23
+     0.41%  crowd_perf       libc.so.6                      [.] 0x00000000000a58f9
+     0.41%  crowd_perf       libvulkan_radeon.so            [.] 0x000000000018d44f
+     0.40%  crowd_perf       libvulkan_radeon.so            [.] 0x000000000002b9b1
+     0.40%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000001e4771
+     0.40%  crowd_perf       libwayland-client.so.0.25.0    [.] 0x00000000000055ca
+     0.39%  crowd_perf       libvulkan_radeon.so            [.] 0x0000000000116639
+     0.39%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000001d1196
+     0.39%  crowd_perf       libwayland-client.so.0.25.0    [.] 0x000000000000315e
+     0.39%  crowd_perf       libvulkan_radeon.so            [.] 0x000000000002d957
+     0.39%  crowd_perf       libvulkan_radeon.so            [.] 0x000000000013d072
+     0.39%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000000cb17b
+     0.38%  crowd_perf       libSDL3.so.0.4.12              [.] 0x000000000005485a
+     0.38%  crowd_perf       libc.so.6                      [.] malloc
+     0.38%  crowd_perf       libc.so.6                      [.] 0x0000000000180625
+     0.38%  crowd_perf       libvulkan_radeon.so            [.] 0x000000000033b85f
+     0.38%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000000ccd20
+     0.38%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000000c9899
+     0.38%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000001deb23
+     0.38%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000000284e5
+     0.38%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000001f41b7
+     0.37%  crowd_perf       libvulkan_radeon.so            [.] 0x0000000000027cca
+     0.37%  crowd_perf       libSDL3.so.0.4.12              [.] 0x00000000002476b9
+     0.37%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000001f41d6
+     0.37%  crowd_perf       libc.so.6                      [.] 0x000000000018a52e
+     0.37%  crowd_perf       crowd_perf                     [.] runtime::map_seed_from_map_data
+     0.36%  crowd_perf       libdbus-1.so.3.38.3            [.] _dbus_rmutex_lock
+     0.36%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000001e2e46
+     0.36%  crowd_perf       libvulkan_radeon.so            [.] 0x000000000022e017
+     0.36%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000000b7b5f
+     0.35%  crowd_perf       libdbus-1.so.3.38.3            [.] _dbus_message_loader_queue_messages
+     0.35%  crowd_perf       libc.so.6                      [.] pthread_rwlock_rdlock
+     0.35%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000000be015
+     0.35%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000000c49da
+     0.34%  crowd_perf       libc.so.6                      [.] 0x000000000018a507
+     0.34%  crowd_perf       libwayland-client.so.0.25.0    [.] wl_display_dispatch_queue_pending
+     0.34%  crowd_perf       libvulkan_radeon.so            [.] 0x0000000000197523
+     0.34%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000000ec905
+     0.34%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000000c4a56
+     0.34%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000000cb864
+     0.34%  crowd_perf       libdrm.so.2.134.0              [.] drmIoctl
+     0.34%  crowd_perf       crowd_perf                     [.] engine::character_clip
+     0.34%  crowd_perf       libvulkan_radeon.so            [.] 0x000000000002d8f7
+     0.33%  crowd_perf       libvulkan_radeon.so            [.] 0x000000000002f104
+     0.33%  crowd_perf       libc.so.6                      [.] pthread_mutex_unlock
+     0.33%  crowd_perf       libSDL3.so.0.4.12              [.] 0x00000000000f7a2c
+     0.33%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000000b92b8
+     0.33%  crowd_perf       crowd_perf                     [.] engine::begin_frame
+     0.33%  crowd_perf       libSDL3.so.0.4.12              [.] 0x000000000023b8f4
+     0.33%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000000c2207
+     0.33%  crowd_perf       libvulkan_radeon.so            [.] 0x000000000018c7d5
+     0.32%  wl_cursor_surfa  [unknown]                      [k] 0xffffffffa9e000ad
+     0.32%  crowd_perf       libSDL3.so.0.4.12              [.] 0x000000000024ab50
+     0.32%  crowd_perf       libvulkan_radeon.so            [.] 0x0000000000022a03
+     0.32%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000000c9cd7
+     0.32%  crowd_perf       libvulkan_radeon.so            [.] 0x0000000000116770
+     0.32%  crowd_perf       libvulkan_radeon.so            [.] 0x0000000000182332
+     0.32%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000000ebd64
+     0.32%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000001f3c19
+     0.31%  crowd_perf       libc.so.6                      [.] __errno_location
+     0.31%  crowd_perf       libc.so.6                      [.] 0x00000000000a57e4
+     0.31%  crowd_perf       libc.so.6                      [.] 0x000000000018a4b0
+     0.31%  wl_cursor_surfa  libwayland-client.so.0.25.0    [.] wl_display_dispatch_queue_pending
+     0.30%  wl_cursor_surfa  libc.so.6                      [.] 0x00000000000a7491
+     0.30%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000001801f0
+     0.30%  crowd_perf       libwayland-client.so.0.25.0    [.] 0x00000000000039f1
+     0.30%  crowd_perf       libSDL3.so.0.4.12              [.] 0x000000000023c110
+     0.30%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000001dd301
+     0.30%  crowd_perf       libSDL3.so.0.4.12              [.] 0x00000000000176c9
+     0.30%  crowd_perf       libc.so.6                      [.] 0x0000000000189c53
+     0.30%  crowd_perf       libc.so.6                      [.] 0x0000000000180600
+     0.30%  crowd_perf       crowd_perf                     [.] __$hasher$$string
+     0.30%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000000c8b16
+     0.30%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000000cd419
+     0.30%  crowd_perf       libvulkan_radeon.so            [.] 0x000000000021b321
+     0.30%  crowd_perf       libSDL3.so.0.4.12              [.] 0x000000000023b39a
+     0.30%  crowd_perf       libc.so.6                      [.] 0x000000000018a4ea
+     0.30%  crowd_perf       libSDL3.so.0.4.12              [.] 0x000000000023be20
+     0.29%  crowd_perf       libvulkan_radeon.so            [.] 0x0000000000027f50
+     0.29%  crowd_perf       libvulkan_radeon.so            [.] 0x0000000000022e83
+     0.28%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000000d178c
+     0.28%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000000d6734
+     0.28%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000000bb103
+     0.28%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000000b70b2
+     0.28%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000000cc0c4
+     0.26%  crowd_perf       libSDL3.so.0.4.12              [.] 0x0000000000017694
+     0.26%  crowd_perf       libvulkan_radeon.so            [.] 0x000000000017e9d4
+     0.26%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000000d64a8
+     0.26%  crowd_perf       libvulkan_radeon.so            [.] 0x0000000000335bd4
+     0.26%  crowd_perf       libSDL3.so.0.4.12              [.] 0x0000000000243e20
+     0.26%  wl_cursor_surfa  libwayland-client.so.0.25.0    [.] 0x0000000000004951
+     0.25%  crowd_perf       libvulkan_radeon.so            [.] 0x0000000000189397
+     0.25%  crowd_perf       libwayland-client.so.0.25.0    [.] 0x0000000000003147
+     0.24%  crowd_perf       libvulkan_radeon.so            [.] 0x000000000068d01a
+     0.24%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000001e379c
+     0.24%  crowd_perf       libdrm.so.2.134.0              [.] drmSyncobjWait
+     0.23%  crowd_perf       [unknown]                      [k] 0xffffffffa9e000ca
+     0.23%  crowd_perf       libvulkan_radeon.so            [.] 0x000000000028eca3
+     0.20%  crowd_perf       libvulkan_radeon.so            [.] 0x000000000002d851
+     0.20%  wl_cursor_surfa  libc.so.6                      [.] 0x00000000000a73ff
+     0.18%  crowd_perf       libxkbcommon.so.0.13.2         [.] xkb_keymap_key_get_syms_by_level
+     0.18%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000001dce94
+     0.10%  crowd_perf       libdbus-1.so.3.38.3            [.] _dbus_connection_unlock
+     0.08%  wl_cursor_surfa  libc.so.6                      [.] 0x00000000000a6761
+     0.01%  crowd_perf       libc.so.6                      [.] 0x0000000000189d47
+     0.00%  wl_cursor_surfa  libc.so.6                      [.] 0x0000000000094084
+     0.00%  wl_cursor_surfa  libc.so.6                      [.] 0x000000000018a507
+     0.00%  wl_cursor_surfa  libc.so.6                      [.] pthread_mutex_lock
+     0.00%  crowd_perf       libnvidia-glcore.so.610.43.03  [.] 0x0000000000e922c9
+     0.00%  crowd_perf       [unknown]                      [k] 0xffffffffab35a5c1
+     0.00%  crowd_perf       libnvidia-glcore.so.610.43.03  [.] 0x0000000000fcd461
+     0.00%  wl_cursor_surfa  libSDL3.so.0.4.12              [.] 0x00000000001e3f74
+     0.00%  wl_cursor_surfa  libSDL3.so.0.4.12              [.] 0x00000000001cd7b4
+     0.00%  crowd_perf       libvulkan_radeon.so            [.] 0x000000000017ff33
+     0.00%  crowd_perf       ld-linux-x86-64.so.2           [.] 0x000000000000af32
+     0.00%  wl_cursor_surfa  libwayland-client.so.0.25.0    [.] wl_display_read_events
+     0.00%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000000d65a4
+     0.00%  crowd_perf       libc.so.6                      [.] vsnprintf
+     0.00%  crowd_perf       libc.so.6                      [.] 0x0000000000184dd2
+     0.00%  crowd_perf       libvulkan_radeon.so            [.] 0x0000000000193ec9
+     0.00%  crowd_perf       ld-linux-x86-64.so.2           [.] 0x00000000000147f8
+     0.00%  crowd_perf       libvulkan_radeon.so            [.] 0x0000000000345adb
+     0.00%  crowd_perf       libc.so.6                      [.] __ctype_init
+     0.00%  crowd_perf       libc.so.6                      [.] 0x0000000000064834
+     0.00%  crowd_perf       libvulkan_radeon.so            [.] 0x0000000000345ad6
+     0.00%  crowd_perf       libvulkan_radeon.so            [.] 0x0000000000344ddb
+     0.00%  crowd_perf       ld-linux-x86-64.so.2           [.] 0x0000000000005d0b
+     0.00%  crowd_perf       libvulkan_radeon.so            [.] 0x0000000000345b18
+     0.00%  crowd_perf       libvulkan_radeon.so            [.] 0x0000000000345bb0
+     0.00%  wl_cursor_surfa  libSDL3.so.0.4.12              [.] 0x00000000001e3f76
+     0.00%  crowd_perf       libvulkan_radeon.so            [.] 0x0000000000345ce7
+     0.00%  crowd_perf       ld-linux-x86-64.so.2           [.] 0x000000000000616e
+     0.00%  crowd_perf       libc.so.6                      [.] 0x0000000000094040
+     0.00%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000001f3362
+     0.00%  crowd_perf       libc.so.6                      [.] realloc
+     0.00%  crowd_perf       libvulkan_radeon.so            [.] 0x0000000000345bb3
+     0.00%  crowd_perf       ld-linux-x86-64.so.2           [.] 0x000000000000ac94
+     0.00%  crowd_perf       libc.so.6                      [.] pthread_setaffinity_np
+     0.00%  crowd_perf       libc.so.6                      [.] 0x0000000000097417
+     0.00%  crowd_perf       libc.so.6                      [.] 0x00000000000973bb
+     0.00%  crowd_perf       libdrm_amdgpu.so.1.134.0       [.] 0x00000000000078b0
+     0.00%  crowd_perf       [unknown]                      [k] 0xffffffffa9e01284
+     0.00%  crowd_perf       libc.so.6                      [.] 0x0000000000103a40
+     0.00%  crowd_perf       ld-linux-x86-64.so.2           [.] 0x000000000001fcc4
+     0.00%  crowd_perf       libvulkan_radeon.so            [.] 0x0000000000344ccb
+     0.00%  crowd_perf       libvulkan_radeon.so            [.] 0x0000000000344cd4
+     0.00%  wl_cursor_surfa  libc.so.6                      [.] ppoll
+     0.00%  wl_cursor_surfa  libc.so.6                      [.] 0x0000000000094040
+     0.00%  crowd_perf       libc.so.6                      [.] 0x000000000011bed0
+     0.00%  crowd_perf       [unknown]                      [k] 0xffffffffa9e01280
+     0.00%  wl_cursor_surfa  libc.so.6                      [.] 0x0000000000094047
+     0.00%  wl_cursor_surfa  libc.so.6                      [.] 0x0000000000094085
+     0.00%  crowd_perf       libc.so.6                      [.] 0x00000000000973a0
+     0.00%  crowd_perf       libc.so.6                      [.] 0x000000000011bed7
+     0.00%  crowd_perf       ld-linux-x86-64.so.2           [.] 0x0000000000002ede
+     0.00%  crowd_perf       ld-linux-x86-64.so.2           [.] 0x0000000000006167
+     0.00%  crowd_perf       ld-linux-x86-64.so.2           [.] 0x000000000000ac75
+     0.00%  crowd_perf       ld-linux-x86-64.so.2           [.] 0x000000000000b7f3
+     0.00%  crowd_perf       ld-linux-x86-64.so.2           [.] 0x000000000000b84a
+     0.00%  crowd_perf       ld-linux-x86-64.so.2           [.] 0x0000000000019a7d
+     0.00%  crowd_perf       ld-linux-x86-64.so.2           [.] 0x0000000000019a84
+     0.00%  crowd_perf       ld-linux-x86-64.so.2           [.] 0x000000000001f103
+     0.00%  crowd_perf       ld-linux-x86-64.so.2           [.] 0x000000000001fcc0
+     0.00%  crowd_perf       libc.so.6                      [.] 0x000000000011bed2
+     0.00%  crowd_perf       libc.so.6                      [.] 0x000000000011bed5
+     0.00%  crowd_perf       libdrm.so.2.134.0              [.] drmSyncobjImportSyncFile
+     0.00%  crowd_perf       libdrm_amdgpu.so.1.134.0       [.] amdgpu_device_initialize
+     0.00%  crowd_perf       libnvidia-glcore.so.610.43.03  [.] 0x0000000000fcd464
+     0.00%  crowd_perf       libnvidia-glcore.so.610.43.03  [.] 0x0000000000fcd46a
+     0.00%  crowd_perf       libvulkan_radeon.so            [.] 0x0000000000344ce2
+     0.00%  crowd_perf       libvulkan_radeon.so            [.] 0x0000000000344ddd
+     0.00%  crowd_perf       libz.so.1.3.2                  [.] 0x0000000000003004
+     0.00%  crowd_perf       libz.so.1.3.2                  [.] 0x000000000000300f
+     0.00%  crowd_perf       [unknown]                      [k] 0xffffffffa9e01670
+     0.00%  wl_cursor_surfa  libc.so.6                      [.] recvmsg
+
+
+#
+# (Tip: To change sampling frequency to 100 Hz: perf record -F 100)
+#

flame/crowd-20260803-134209-report.txt

@@ -0,0 +1,700 @@
+# To display the perf.data header info, please use --header/--header-only options.
+#
+#
+# Total Lost Samples: 0
+#
+# Samples: 469  of event 'cpu/cycles/Pu'
+# Event count (approx.): 2304609671
+#
+# Overhead  Command          Shared Object                  Symbol                                                                                                                                                                
+# ........  ...............  .............................  ......................................................................................................................................................................
+#
+    12.22%  crowd_perf       crowd_perf                     [.] engine::draw_sprite
+            |
+            ---engine::draw_sprite
+               main::main
+               main
+               0x7fe9b3c27740
+               __libc_start_main
+               _start
+
+     5.51%  crowd_perf       crowd_perf                     [.] engine::to_clip
+            |
+            ---engine::to_clip
+               engine::draw_sprite
+               main::main
+               main
+               0x7fe9b3c27740
+               __libc_start_main
+               _start
+
+     4.04%  crowd_perf       libc.so.6                      [.] pthread_mutex_lock
+            |
+            ---pthread_mutex_lock
+               |          
+               |--0.81%--0x7fe9b424f559
+               |          0x7fe9b42545c5
+               |          0x7fe9b4069860
+               |          engine::begin_frame
+               |          main::main
+               |          main
+               |          0x7fe9b3c27740
+               |          __libc_start_main
+               |          _start
+               |          
+               |--0.79%--wl_display_prepare_read_queue
+               |          wl_display_dispatch_queue_timeout
+               |          0x7fe9b228fead
+               |          0x7fe9b1d936f3
+               |          0x7fe9b1d97300
+               |          0x7fe9b1d8a4f8
+               |          0x7fe9b424ffc9
+               |          engine::end_frame
+               |          main::main
+               |          main
+               |          0x7fe9b3c27740
+               |          __libc_start_main
+               |          _start
+               |          
+               |--0.76%--0x7fe9b3e94a51
+               |          0x7fe9b4054fca
+               |          0x7fe9b4055bbd
+               |          engine::events
+               |          main::main
+               |          main
+               |          0x7fe9b3c27740
+               |          __libc_start_main
+               |          _start
+               |          
+               |--0.76%--wl_display_cancel_read
+               |          
+                --0.52%--0x7fe9b3eab81f
+                          0x7fe9b3e905f6
+                          dbus_connection_dispatch
+                          0x7fe9b4054fe2
+                          0x7fe9b4055bbd
+                          engine::events
+                          main::main
+                          main
+                          0x7fe9b3c27740
+                          __libc_start_main
+                          _start
+
+     3.30%  crowd_perf       crowd_perf                     [.] engine::sprite_feet_quad
+            |
+            ---engine::sprite_feet_quad
+               engine::draw_sprite
+               main::main
+               main
+               0x7fe9b3c27740
+               __libc_start_main
+               _start
+
+     3.12%  crowd_perf       crowd_perf                     [.] main::main
+            |
+            ---main::main
+               main
+               0x7fe9b3c27740
+               __libc_start_main
+               _start
+
+     3.01%  crowd_perf       crowd_perf                     [.] engine::update_sprite
+            |
+            ---engine::update_sprite
+               main::main
+               main
+               0x7fe9b3c27740
+               __libc_start_main
+               _start
+
+     2.34%  crowd_perf       crowd_perf                     [.] engine::world_to_screen
+            |
+            ---engine::world_to_screen
+               engine::draw_sprite
+               main::main
+               main
+               0x7fe9b3c27740
+               __libc_start_main
+               _start
+
+     1.95%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000001f41d6
+            |
+            ---0x7fe9b1df41d6
+               |          
+                --1.18%--0x7fe9b1df44e5
+                          0x7fe9b1c27d3c
+                          0x7fe9b1c28070
+                          0x7fe9b1c2d63c
+                          |          
+                           --0.87%--0x7fe9b1cd16b8
+                                     0x7fe9b1cd2520
+                                     0x7fe9b1cd2d0a
+                                     0x7fe9b1de6325
+                                     0x7fe9b42546f1
+                                     0x7fe9b4069860
+                                     engine::begin_frame
+                                     main::main
+                                     main
+                                     0x7fe9b3c27740
+                                     __libc_start_main
+                                     _start
+
+     1.95%  crowd_perf       crowd_perf                     [.] runtime::default_hasher
+            |
+            ---runtime::default_hasher
+               runtime::default_hasher_string
+               __$hasher$$string
+               |          
+               |--1.28%--engine::character_clip
+               |          engine::update_sprite
+               |          main::main
+               |          main
+               |          0x7fe9b3c27740
+               |          __libc_start_main
+               |          _start
+               |          
+                --0.66%--engine::character_frame
+                          engine::draw_sprite
+                          main::main
+                          main
+                          0x7fe9b3c27740
+                          __libc_start_main
+                          _start
+
+     1.60%  crowd_perf       libc.so.6                      [.] clock_gettime
+            |
+            ---clock_gettime
+               |          
+                --0.83%--wl_display_dispatch_queue_timeout
+                          0x7fe9b228fead
+                          0x7fe9b1d936f3
+                          0x7fe9b1d97300
+                          0x7fe9b1d8a4f8
+                          0x7fe9b424ffc9
+                          engine::end_frame
+                          main::main
+                          main
+                          0x7fe9b3c27740
+                          __libc_start_main
+                          _start
+
+     1.50%  crowd_perf       crowd_perf                     [.] __$map_get$$map[string]engine::Clip_Def
+            |
+            ---__$map_get$$map[string]engine::Clip_Def
+               |          
+               |--0.79%--engine::character_frame
+               |          engine::draw_sprite
+               |          main::main
+               |          main
+               |          0x7fe9b3c27740
+               |          __libc_start_main
+               |          _start
+               |          
+                --0.71%--engine::character_clip
+                          engine::update_sprite
+                          main::main
+                          main
+                          0x7fe9b3c27740
+                          __libc_start_main
+                          _start
+
+     1.49%  wl_cursor_surfa  [unknown]                      [k] 0xffffffffab041cef
+            |
+            ---0xffffffffab041cef
+               0xffffffffab355aba
+               0xffffffffa9e0012f
+
+     1.41%  crowd_perf       crowd_perf                     [.] engine::character_frame
+            |
+            ---engine::character_frame
+               engine::draw_sprite
+               main::main
+               main
+               0x7fe9b3c27740
+               __libc_start_main
+               _start
+
+     1.28%  crowd_perf       [unknown]                      [k] 0xffffffffab355a81
+            |
+            ---0xffffffffab355a81
+               0xffffffffa9e0012f
+               |          
+               |--0.77%--ioctl
+               |          drmIoctl
+               |          
+                --0.51%--0x7fe9b3c9fff2
+                          0x7fe9b3c9403b
+                          0x7fe9b3c94083
+                          __close
+                          0x7fe9b1d8cacd
+                          0x7fe9b1d893c8
+                          0x7fe9b1d89241
+                          0x7fe9b42544bd
+                          0x7fe9b4069860
+                          engine::begin_frame
+                          main::main
+                          main
+                          0x7fe9b3c27740
+                          __libc_start_main
+                          _start
+
+     1.26%  wl_cursor_surfa  libwayland-client.so.0.25.0    [.] wl_display_read_events
+     1.25%  crowd_perf       libwayland-client.so.0.25.0    [.] wl_proxy_marshal_array_flags
+            |
+            ---wl_proxy_marshal_array_flags
+               wl_proxy_marshal_flags
+               |          
+                --0.83%--0x7fe9b1d97614
+                          0x7fe9b1d8a4f8
+                          0x7fe9b424ffc9
+                          engine::end_frame
+                          main::main
+                          main
+                          0x7fe9b3c27740
+                          __libc_start_main
+                          _start
+
+     1.22%  crowd_perf       crowd_perf                     [.] runtime::string_eq
+            |
+            ---runtime::string_eq
+               __$map_get$$map[string]engine::Clip_Def
+               |          
+               |--0.72%--engine::character_clip
+               |          engine::update_sprite
+               |          main::main
+               |          main
+               |          0x7fe9b3c27740
+               |          __libc_start_main
+               |          _start
+               |          
+                --0.50%--engine::character_frame
+                          engine::draw_sprite
+                          main::main
+                          main
+                          0x7fe9b3c27740
+                          __libc_start_main
+                          _start
+
+     1.22%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000000ce007
+            |
+            ---0x7fe9b1cce007
+               0x7fe9b1dcae1a
+               0x7fe9b424437c
+               0x7fe9b406920d
+               engine::end_frame
+               main::main
+
+     1.17%  crowd_perf       crowd_perf                     [.] runtime::memory_equal
+            |
+            ---runtime::memory_equal
+               runtime::string_eq
+               __$map_get$$map[string]engine::Clip_Def
+               |          
+                --0.72%--engine::character_frame
+                          engine::draw_sprite
+                          main::main
+                          main
+                          0x7fe9b3c27740
+                          __libc_start_main
+                          _start
+
+     1.05%  crowd_perf       crowd_perf                     [.] engine::character_clip
+            |
+            ---engine::character_clip
+               engine::update_sprite
+               main::main
+               main
+               0x7fe9b3c27740
+               __libc_start_main
+               _start
+
+     0.99%  crowd_perf       crowd_perf                     [.] runtime::bounds_check_error
+            |
+            ---runtime::bounds_check_error
+               |          
+                --0.54%--engine::texture_run_len
+                          engine::end_frame
+                          main::main
+                          main
+                          0x7fe9b3c27740
+                          __libc_start_main
+                          _start
+
+     0.95%  crowd_perf       libc.so.6                      [.] 0x000000000018a4c4
+            |
+            ---0x7fe9b3d8a4c4
+               |          
+                --0.52%--engine::draw_sprite
+                          main::main
+                          main
+                          0x7fe9b3c27740
+                          __libc_start_main
+                          _start
+
+     0.94%  crowd_perf       crowd_perf                     [.] engine::texture_run_len
+            |
+            ---engine::texture_run_len
+               engine::end_frame
+               main::main
+               main
+               0x7fe9b3c27740
+               __libc_start_main
+               _start
+
+     0.93%  crowd_perf       libSDL3.so.0.4.12              [.] 0x00000000001428e7
+            |
+            ---0x7fe9b41428e7
+
+     0.85%  crowd_perf       libc.so.6                      [.] 0x00000000000a44b7
+            |
+            ---0x7fe9b3ca44b7
+               0x7fe9b3ca6ed2
+
+     0.85%  crowd_perf       libc.so.6                      [.] 0x000000000009ca5c
+            |
+            ---0x7fe9b3c9ca5c
+
+     0.84%  crowd_perf       libzstd.so.1.5.7               [.] 0x0000000000092a16
+            |
+            ---0x7fe9b0567a16
+               ZSTD_decompressSequences (inlined)
+               ZSTD_decompressBlock_internal
+               ZSTD_decompressBlock_internal
+               0x7fe9b0562e85
+               ZSTD_decompress_usingDDict
+               ZSTD_decompress
+               0x7fe9b1f5f8cf
+               0x7fe9b1f380f5
+               0x7fe9b1f3978d
+               0x7fe9b1f37a86
+               0x7fe9b1e05d1b
+               0x7fe9b1d28dcd
+               0x7fe9b1e0583a
+               0x7fe9b1e05d49
+               0x7fe9b1d28abb
+               0x7fe9b1d292b4
+               0x7fe9b1d2ab09
+
+     0.83%  crowd_perf       libSDL3.so.0.4.12              [.] 0x0000000000143ffe
+            |
+            ---0x7fe9b4143ffe
+               0x7fe9b414a028
+               0x7fe9b4141b9a
+
+     0.83%  crowd_perf       libwayland-client.so.0.25.0    [.] 0x0000000000003263
+            |
+            ---0x7fe9b433a263
+               wl_proxy_marshal_flags
+               0x7fe9b1d97937
+               0x7fe9b1d8a4f8
+               0x7fe9b424ffc9
+               engine::end_frame
+               main::main
+               main
+               0x7fe9b3c27740
+               __libc_start_main
+               _start
+
+     0.83%  crowd_perf       libvulkan_radeon.so            [.] 0x000000000033b851
+            |
+            ---0x7fe9b1f3b851
+               0x7fe9b1df41ce
+               0x7fe9b1df44e5
+               0x7fe9b1c27d3c
+               0x7fe9b1c28070
+               0x7fe9b1c2d63c
+
+     0.83%  crowd_perf       libSDL3.so.0.4.12              [.] 0x000000000014400b
+            |
+            ---0x7fe9b414400b
+               0x7fe9b414a028
+               0x7fe9b4141b9a
+
+     0.82%  crowd_perf       crowd_perf                     [.] engine::frame_uvs
+            |
+            ---engine::frame_uvs
+               engine::draw_sprite
+               main::main
+               main
+               0x7fe9b3c27740
+               __libc_start_main
+               _start
+
+     0.82%  crowd_perf       libSDL3.so.0.4.12              [.] 0x0000000000144008
+            |
+            ---0x7fe9b4144008
+               0x7fe9b414a028
+               0x7fe9b4141b9a
+
+     0.81%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000000d2c2b
+            |
+            ---0x7fe9b1cd2c2b
+               0x7fe9b1de6325
+               0x7fe9b42546f1
+               0x7fe9b4069860
+               engine::begin_frame
+               main::main
+               main
+               0x7fe9b3c27740
+               __libc_start_main
+               _start
+
+     0.79%  crowd_perf       libc.so.6                      [.] pthread_rwlock_rdlock
+            |
+            ---pthread_rwlock_rdlock
+
+     0.78%  crowd_perf       libSDL3.so.0.4.12              [.] 0x00000000001428a1
+            |
+            ---0x7fe9b41428a1
+               0x7fe9b414921a
+               0x7fe9b4141b9a
+               0x7fe9b41426da
+
+     0.76%  wl_cursor_surfa  libc.so.6                      [.] 0x00000000000a66e2
+     0.69%  crowd_perf       libc.so.6                      [.] 0x000000000018a547
+            |
+            ---0x7fe9b3d8a547
+               0x7fe9b1cd1554
+               0x7fe9b1de165e
+               0x7fe9b1ddee47
+               0x7fe9b1de35f1
+               0x7fe9b1de157e
+               0x7fe9b423be46
+               0x7fe9b4068a74
+               engine::end_frame
+               main::main
+               main
+               0x7fe9b3c27740
+               __libc_start_main
+               _start
+
+     0.67%  crowd_perf       libSDL3.so.0.4.12              [.] 0x000000000014288a
+            |
+            ---0x7fe9b414288a
+               0x7fe9b41429b4
+               0x7fe9b414921a
+               0x7fe9b4141b9a
+               0x7fe9b41426da
+
+     0.61%  crowd_perf       libvulkan_radeon.so            [.] 0x0000000000180ab2
+            |
+            ---0x7fe9b1d80ab2
+
+     0.58%  crowd_perf       libvulkan_radeon.so            [.] 0x0000000000114c26
+            |
+            ---0x7fe9b1d14c26
+               0x7fe9b1d15976
+               0x7fe9b1d162ff
+               0x7fe9b1d1692c
+               0x7fe9b1c2bd93
+               0x7fe9b1c2da03
+               0x7fe9b1c2e4ca
+               0x7fe9b1c2f277
+               0x7fe9b1de052c
+               0x7fe9b1de1521
+               0x7fe9b4247581
+               0x7fe9b4064bab
+               engine::end_frame
+               main::main
+
+     0.52%  crowd_perf       libvulkan_radeon.so            [.] 0x0000000000022daa
+            |
+            ---0x7fe9b1c22daa
+               0x7fe9b1cd2f6f
+               0x7fe9b1de6325
+               0x7fe9b423bc48
+               0x7fe9b423beb3
+               0x7fe9b4068a74
+               engine::end_frame
+               main::main
+               main
+               0x7fe9b3c27740
+               __libc_start_main
+               _start
+
+     0.51%  wl_cursor_surfa  libc.so.6                      [.] 0x00000000000a6747
+     0.49%  crowd_perf       libc.so.6                      [.] 0x00000000001904ba
+     0.48%  crowd_perf       libc.so.6                      [.] 0x000000000018a52e
+     0.48%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000000d17cc
+     0.47%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000001f3c6b
+     0.47%  crowd_perf       [vdso]                         [.] __vdso_clock_gettime
+     0.47%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000001d6365
+     0.46%  crowd_perf       libc.so.6                      [.] 0x000000000018a486
+     0.46%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000001e4c30
+     0.46%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000001f3ff4
+     0.45%  crowd_perf       libdbus-1.so.3.38.3            [.] dbus_connection_get_dispatch_status
+     0.45%  crowd_perf       libc.so.6                      [.] 0x0000000000093fe4
+     0.44%  crowd_perf       libvulkan_radeon.so            [.] 0x000000000017e9d4
+     0.44%  crowd_perf       libvulkan_radeon.so            [.] 0x0000000000114197
+     0.43%  crowd_perf       libnvidia-glcore.so.610.43.03  [.] 0x0000000000fcd46c
+     0.43%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000000ce927
+     0.43%  crowd_perf       crowd_perf                     [.] runtime::map_seed_from_map_data
+     0.43%  crowd_perf       libvulkan_radeon.so            [.] 0x000000000033cb4a
+     0.42%  crowd_perf       libc.so.6                      [.] 0x0000000000189c4b
+     0.42%  crowd_perf       libSDL3.so.0.4.12              [.] 0x0000000000018567
+     0.42%  crowd_perf       libc.so.6                      [.] 0x0000000000189cca
+     0.41%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000001de7c1
+     0.41%  crowd_perf       libc.so.6                      [.] 0x000000000009ca21
+     0.41%  crowd_perf       libSDL3.so.0.4.12              [.] 0x00000000000176c6
+     0.41%  crowd_perf       crowd_perf                     [.] __$hasher$$string
+     0.41%  crowd_perf       libxkbcommon.so.0.13.2         [.] 0x0000000000005fe7
+     0.40%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000000d2c23
+     0.40%  crowd_perf       libSDL3.so.0.4.12              [.] 0x00000000000185a0
+     0.40%  crowd_perf       libSDL3.so.0.4.12              [.] 0x000000000024f5ea
+     0.40%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000000e662b
+     0.40%  crowd_perf       libwayland-client.so.0.25.0    [.] 0x0000000000005555
+     0.40%  crowd_perf       libSDL3.so.0.4.12              [.] 0x00000000000541ae
+     0.40%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000000e6624
+     0.40%  crowd_perf       libvulkan_radeon.so            [.] 0x000000000033e954
+     0.40%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000001ded02
+     0.40%  crowd_perf       [unknown]                      [k] 0xffffffffa9e000ab
+     0.39%  crowd_perf       libSDL3.so.0.4.12              [.] 0x0000000000054314
+     0.39%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000001f34c7
+     0.39%  crowd_perf       libSDL3.so.0.4.12              [.] 0x0000000000064bba
+     0.39%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000000ebaa4
+     0.39%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000000b8cb1
+     0.38%  crowd_perf       libc.so.6                      [.] pthread_rwlock_wrlock
+     0.38%  wl_cursor_surfa  libSDL3.so.0.4.12              [.] 0x00000000001e3f76
+     0.38%  crowd_perf       libwayland-client.so.0.25.0    [.] 0x000000000000326a
+     0.38%  crowd_perf       libSDL3.so.0.4.12              [.] 0x0000000000064a44
+     0.37%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000000cf7ea
+     0.37%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000000c4a65
+     0.37%  crowd_perf       libSDL3.so.0.4.12              [.] 0x00000000002501f2
+     0.37%  crowd_perf       libSDL3.so.0.4.12              [.] 0x000000000024e78a
+     0.37%  crowd_perf       libvulkan_radeon.so            [.] 0x000000000035f600
+     0.37%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000001f41b7
+     0.37%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000000bdd41
+     0.37%  crowd_perf       libvulkan_radeon.so            [.] 0x000000000013c80b
+     0.37%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000001e61b8
+     0.37%  crowd_perf       libSDL3.so.0.4.12              [.] 0x000000000023bc49
+     0.37%  crowd_perf       libc.so.6                      [.] 0x0000000000189cc4
+     0.36%  crowd_perf       libSDL3.so.0.4.12              [.] 0x0000000000247677
+     0.36%  wl_cursor_surfa  libc.so.6                      [.] pthread_mutex_lock
+     0.36%  crowd_perf       libdbus-1.so.3.38.3            [.] 0x000000000003186f
+     0.36%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000000d19ec
+     0.36%  crowd_perf       libvulkan_radeon.so            [.] 0x000000000017ff7a
+     0.36%  crowd_perf       libSDL3.so.0.4.12              [.] 0x0000000000053d7e
+     0.36%  crowd_perf       libc.so.6                      [.] 0x000000000018a4ca
+     0.36%  crowd_perf       libSDL3.so.0.4.12              [.] 0x0000000000246f63
+     0.36%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000001e6473
+     0.35%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000000d6725
+     0.34%  crowd_perf       libc.so.6                      [.] pthread_mutex_unlock
+     0.32%  crowd_perf       libSDL3.so.0.4.12              [.] 0x000000000024fc98
+     0.32%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000000cf078
+     0.32%  crowd_perf       libwayland-client.so.0.25.0    [.] 0x0000000000004dc1
+     0.31%  wl_cursor_surfa  libc.so.6                      [.] 0x0000000000094047
+     0.31%  crowd_perf       crowd_perf                     [.] runtime::_append_elem
+     0.29%  crowd_perf       libvulkan_radeon.so            [.] 0x0000000000022d80
+     0.28%  crowd_perf       libSDL3.so.0.4.12              [.] 0x000000000023b5cb
+     0.24%  crowd_perf       crowd_perf                     [.] runtime::append_elem:proc(array:^[dynamic]engine::Queued_Sprite,arg:engine::Queued_Sprite,loc:runtime::Source_Code_Location)->(n:int,err:runtime::Allocator_Error)
+     0.23%  wl_cursor_surfa  libwayland-client.so.0.25.0    [.] wl_list_insert
+     0.23%  crowd_perf       libc.so.6                      [.] 0x00000000000a718d
+     0.19%  crowd_perf       libvulkan_radeon.so            [.] 0x000000000033b80b
+     0.17%  crowd_perf       ld-linux-x86-64.so.2           [.] 0x000000000000ffcb
+     0.16%  crowd_perf       libc.so.6                      [.] 0x0000000000189c40
+     0.09%  crowd_perf       libvulkan_radeon.so            [.] 0x0000000000335bdd
+     0.02%  crowd_perf       libc.so.6                      [.] 0x00000000000a4451
+     0.02%  wl_cursor_surfa  [unknown]                      [k] 0xffffffffa9e000ad
+     0.02%  wl_cursor_surfa  libc.so.6                      [.] 0x00000000000a6ea4
+     0.00%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000000e5f87
+     0.00%  crowd_perf       libGLX_nvidia.so.610.43.03     [.] 0x00000000000ab456
+     0.00%  crowd_perf       libc.so.6                      [.] 0x00000000000a6968
+     0.00%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000001e4d25
+     0.00%  crowd_perf       libc.so.6                      [.] 0x000000000018325e
+     0.00%  crowd_perf       ld-linux-x86-64.so.2           [.] 0x00000000000144d1
+     0.00%  crowd_perf       libvulkan_radeon.so            [.] 0x000000000018d5e9
+     0.00%  wl_cursor_surfa  libwayland-client.so.0.25.0    [.] wl_display_dispatch_queue_pending
+     0.00%  wl_cursor_surfa  libc.so.6                      [.] 0x00000000000a44b7
+     0.00%  crowd_perf       libvulkan_radeon.so            [.] 0x0000000000344db9
+     0.00%  wl_cursor_surfa  libc.so.6                      [.] 0x00000000000a74c7
+     0.00%  crowd_perf       libc.so.6                      [.] 0x0000000000185195
+     0.00%  wl_cursor_surfa  [unknown]                      [k] 0xffffffffab041d04
+     0.00%  wl_cursor_surfa  libffi.so.8.4.1                [.] 0x0000000000002807
+     0.00%  crowd_perf       libdrm.so.2.134.0              [.] drmSyncobjTimelineWait
+     0.00%  crowd_perf       libexpat.so.1.12.2             [.] 0x0000000000002ea9
+     0.00%  crowd_perf       libvulkan_radeon.so            [.] 0x0000000000344c64
+     0.00%  crowd_perf       libc.so.6                      [.] 0x00000000000a5735
+     0.00%  crowd_perf       libdbus-1.so.3.38.3            [.] _dbus_string_equal_c_str
+     0.00%  wl_cursor_surfa  libwayland-client.so.0.25.0    [.] wl_display_get_fd
+     0.00%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000001f2fa5
+     0.00%  crowd_p:disk$0   libc.so.6                      [.] 0x00000000000a590f
+     0.00%  wl_cursor_surfa  libc.so.6                      [.] 0x00000000000a6732
+     0.00%  crowd_perf       ld-linux-x86-64.so.2           [.] 0x000000000002647b
+     0.00%  crowd_perf       libc.so.6                      [.] realloc
+     0.00%  crowd_perf       libc.so.6                      [.] 0x00000000000534a3
+     0.00%  crowd_perf       libc.so.6                      [.] 0x00000000000649f8
+     0.00%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000001e4d24
+     0.00%  crowd_perf       libvulkan_radeon.so            [.] 0x0000000000344ccb
+     0.00%  wl_cursor_surfa  libc.so.6                      [.] 0x00000000000a66c4
+     0.00%  wl_cursor_surfa  libc.so.6                      [.] 0x0000000000189cf7
+     0.00%  crowd_perf       libvulkan_radeon.so            [.] 0x0000000000345b18
+     0.00%  crowd_perf       libc.so.6                      [.] strchrnul@plt
+     0.00%  crowd_perf       libc.so.6                      [.] 0x000000000008b802
+     0.00%  crowd_perf       [unknown]                      [k] 0xffffffffab355a39
+     0.00%  crowd_perf       [unknown]                      [k] 0xffffffffab355a10
+     0.00%  wl_cursor_surfa  libc.so.6                      [.] ppoll
+     0.00%  crowd_perf       libvulkan_radeon.so            [.] 0x0000000000345ad6
+     0.00%  crowd_perf       libvulkan_radeon.so            [.] 0x0000000000345adb
+     0.00%  crowd_perf       libvulkan_radeon.so            [.] 0x0000000000345ce7
+     0.00%  crowd_perf       ld-linux-x86-64.so.2           [.] 0x0000000000014464
+     0.00%  wl_cursor_surfa  libc.so.6                      [.] recvmsg
+     0.00%  crowd_perf       libxcb.so.1.1.0                [.] 0x000000000000caad
+     0.00%  crowd_perf       libc.so.6                      [.] __ctype_init
+     0.00%  crowd_perf       libc.so.6                      [.] 0x00000000000a713d
+     0.00%  crowd_perf       [unknown]                      [k] 0xffffffffa9e01690
+     0.00%  wl_cursor_surfa  libSDL3.so.0.4.12              [.] 0x00000000001cd7b1
+     0.00%  crowd_perf       libc.so.6                      [.] 0x0000000000097417
+     0.00%  crowd_perf       libc.so.6                      [.] 0x0000000000103a63
+     0.00%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000001dd5e5
+     0.00%  crowd_perf       libdrm.so.2.134.0              [.] drmSyncobjWait
+     0.00%  crowd_perf       libexpat.so.1.12.2             [.] 0x000000000000b5db
+     0.00%  crowd_perf       libc.so.6                      [.] __errno_location
+     0.00%  crowd_p:disk$0   libc.so.6                      [.] 0x000000000009404d
+     0.00%  crowd_perf       ld-linux-x86-64.so.2           [.] 0x000000000001fce9
+     0.00%  crowd_perf       libvulkan_radeon.so            [.] 0x0000000000344ddd
+     0.00%  crowd_perf       libvulkan_radeon.so            [.] 0x0000000000345bb3
+     0.00%  crowd_perf       [unknown]                      [k] 0xffffffffa9e01280
+     0.00%  crowd_perf       libvulkan_radeon.so            [.] 0x0000000000344cd4
+     0.00%  wl_cursor_surfa  libc.so.6                      [.] 0x0000000000094085
+     0.00%  crowd_perf       libc.so.6                      [.] 0x000000000011bed0
+     0.00%  crowd_perf       libvulkan_radeon.so            [.] 0x0000000000344ddb
+     0.00%  wl_cursor_surfa  libc.so.6                      [.] 0x0000000000094040
+     0.00%  crowd_perf       libdrm.so.2.134.0              [.] drmIoctl
+     0.00%  crowd_perf       crowd_perf                     [.] runtime::multi_pointer_slice_expr_error
+     0.00%  crowd_perf       libc.so.6                      [.] ioctl
+     0.00%  crowd_perf       libc.so.6                      [.] 0x00000000000973a0
+     0.00%  crowd_perf       libc.so.6                      [.] 0x0000000000103a40
+     0.00%  crowd_perf       [unknown]                      [k] 0xffffffffa9e01670
+     0.00%  wl_cursor_surfa  libc.so.6                      [.] 0x0000000000094050
+     0.00%  wl_cursor_surfa  libc.so.6                      [.] 0x0000000000094084
+     0.00%  crowd_p:disk$0   libc.so.6                      [.] 0x0000000000094047
+     0.00%  crowd_p:disk$0   libc.so.6                      [.] 0x0000000000094048
+     0.00%  crowd_p:disk$0   libc.so.6                      [.] 0x0000000000094049
+     0.00%  crowd_perf       crowd_perf                     [.] main
+     0.00%  crowd_perf       crowd_perf                     [.] memset@plt
+     0.00%  crowd_perf       ld-linux-x86-64.so.2           [.] 0x000000000001f103
+     0.00%  crowd_perf       libc.so.6                      [.] __poll
+     0.00%  crowd_perf       libc.so.6                      [.] malloc
+     0.00%  crowd_perf       libc.so.6                      [.] 0x0000000000094047
+     0.00%  crowd_perf       libc.so.6                      [.] 0x000000000009741b
+     0.00%  crowd_perf       libc.so.6                      [.] 0x00000000000a710f
+     0.00%  crowd_perf       libdbus-1.so.3.38.3            [.] 0x0000000000032197
+     0.00%  crowd_perf       libexpat.so.1.12.2             [.] 0x0000000000021792
+     0.00%  crowd_perf       libexpat.so.1.12.2             [.] 0x000000000002179b
+     0.00%  crowd_perf       libexpat.so.1.12.2             [.] 0x00000000000217aa
+     0.00%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000001dd564
+     0.00%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000001dd592
+     0.00%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000001dd61c
+     0.00%  crowd_perf       libvulkan_radeon.so            [.] 0x00000000003458a3
+     0.00%  crowd_perf       libxcb.so.1.1.0                [.] 0x000000000000c8db
+     0.00%  crowd_perf       libxcb.so.1.1.0                [.] 0x000000000000ce6f
+     0.00%  crowd_perf       [unknown]                      [k] 0xffffffffa9e015f0
+     0.00%  wl_cursor_surfa  libSDL3.so.0.4.12              [.] 0x00000000001e3f74
+     0.00%  wl_cursor_surfa  libc.so.6                      [.] 0x0000000000094048
+     0.00%  wl_cursor_surfa  libc.so.6                      [.] 0x0000000000094049
+     0.00%  wl_cursor_surfa  libc.so.6                      [.] 0x0000000000094097
+
+
+#
+# (Tip: To show IPC for sampling periods use perf record -e '{cycles,instructions}:S' and then browse context)
+#