Sort names in build files.

PiperOrigin-RevId: 338575415
diff --git a/BUILD.bazel b/BUILD.bazel
index 6f974d2..cbb3def 100644
--- a/BUILD.bazel
+++ b/BUILD.bazel
@@ -84,8 +84,8 @@
     "src/subgraph/deconvolution-2d.c",
     "src/subgraph/depthwise-convolution-2d.c",
     "src/subgraph/divide.c",
-    "src/subgraph/fully-connected.c",
     "src/subgraph/floor.c",
+    "src/subgraph/fully-connected.c",
     "src/subgraph/global-average-pooling-2d.c",
     "src/subgraph/hardswish.c",
     "src/subgraph/leaky-relu.c",
@@ -97,12 +97,12 @@
     "src/subgraph/prelu.c",
     "src/subgraph/sigmoid.c",
     "src/subgraph/softmax.c",
+    "src/subgraph/square-root.c",
+    "src/subgraph/square.c",
+    "src/subgraph/squared-difference.c",
     "src/subgraph/static-constant-pad.c",
     "src/subgraph/static-reshape.c",
     "src/subgraph/static-resize-bilinear-2d.c",
-    "src/subgraph/square.c",
-    "src/subgraph/square-root.c",
-    "src/subgraph/squared-difference.c",
     "src/subgraph/subtract.c",
     "src/subgraph/unpooling-2d.c",
 ]
@@ -123,79 +123,76 @@
     "src/f32-clamp/gen/scalar-x1.c",
     "src/f32-clamp/gen/scalar-x2.c",
     "src/f32-clamp/gen/scalar-x4.c",
-    "src/f32-conv-hwc/3x3s2p1c3x4-scalar-1x1.c",
     "src/f32-conv-hwc/3x3s2p0p1c3x4-scalar-1x1.c",
+    "src/f32-conv-hwc/3x3s2p1c3x4-scalar-1x1.c",
     "src/f32-conv-hwc2chw/3x3s2p1c3x4-scalar-1x1.c",
     "src/f32-dwconv-chw/3x3p1-scalar-1x1-acc3.c",
     "src/f32-dwconv-chw/3x3s2p1-scalar-1x1-acc3.c",
     "src/f32-dwconv-chw/5x5p2-scalar-1x1-acc5.c",
     "src/f32-dwconv-chw/5x5s2p2-scalar-1x1-acc5.c",
-    "src/f32-dwconv/gen/up1x4-scalar.c",
-    "src/f32-dwconv/gen/up1x4-scalar-acc2.c",
-    "src/f32-dwconv/gen/up1x9-scalar.c",
-    "src/f32-dwconv/gen/up1x9-scalar-acc2.c",
-    "src/f32-dwconv/gen/up1x25-scalar.c",
-    "src/f32-dwconv/gen/up1x25-scalar-acc2.c",
-    "src/f32-dwconv/gen/up2x4-scalar.c",
-    "src/f32-dwconv/gen/up2x4-scalar-acc2.c",
-    "src/f32-dwconv/gen/up2x9-scalar.c",
-    "src/f32-dwconv/gen/up2x9-scalar-acc2.c",
-    "src/f32-dwconv/gen/up2x25-scalar.c",
-    "src/f32-dwconv/gen/up2x25-scalar-acc2.c",
-    "src/f32-dwconv/gen/up1x4-minmax-scalar.c",
     "src/f32-dwconv/gen/up1x4-minmax-scalar-acc2.c",
-    "src/f32-dwconv/gen/up1x9-minmax-scalar.c",
+    "src/f32-dwconv/gen/up1x4-minmax-scalar.c",
+    "src/f32-dwconv/gen/up1x4-scalar-acc2.c",
+    "src/f32-dwconv/gen/up1x4-scalar.c",
     "src/f32-dwconv/gen/up1x9-minmax-scalar-acc2.c",
-    "src/f32-dwconv/gen/up1x25-minmax-scalar.c",
+    "src/f32-dwconv/gen/up1x9-minmax-scalar.c",
+    "src/f32-dwconv/gen/up1x9-scalar-acc2.c",
+    "src/f32-dwconv/gen/up1x9-scalar.c",
     "src/f32-dwconv/gen/up1x25-minmax-scalar-acc2.c",
-    "src/f32-dwconv/gen/up2x4-minmax-scalar.c",
+    "src/f32-dwconv/gen/up1x25-minmax-scalar.c",
+    "src/f32-dwconv/gen/up1x25-scalar-acc2.c",
+    "src/f32-dwconv/gen/up1x25-scalar.c",
     "src/f32-dwconv/gen/up2x4-minmax-scalar-acc2.c",
-    "src/f32-dwconv/gen/up2x9-minmax-scalar.c",
+    "src/f32-dwconv/gen/up2x4-minmax-scalar.c",
+    "src/f32-dwconv/gen/up2x4-scalar-acc2.c",
+    "src/f32-dwconv/gen/up2x4-scalar.c",
     "src/f32-dwconv/gen/up2x9-minmax-scalar-acc2.c",
-    "src/f32-dwconv/gen/up2x25-minmax-scalar.c",
+    "src/f32-dwconv/gen/up2x9-minmax-scalar.c",
+    "src/f32-dwconv/gen/up2x9-scalar-acc2.c",
+    "src/f32-dwconv/gen/up2x9-scalar.c",
     "src/f32-dwconv/gen/up2x25-minmax-scalar-acc2.c",
+    "src/f32-dwconv/gen/up2x25-minmax-scalar.c",
+    "src/f32-dwconv/gen/up2x25-scalar-acc2.c",
+    "src/f32-dwconv/gen/up2x25-scalar.c",
     "src/f32-gavgpool-cw/scalar-x1.c",
     "src/f32-gavgpool/7p7x-minmax-scalar-c1.c",
     "src/f32-gavgpool/7x-minmax-scalar-c1.c",
-    "src/f32-gemm/gen/1x4-scalar.c",
-    "src/f32-gemm/gen/2x4-scalar.c",
-    "src/f32-gemm/gen/4x2-scalar.c",
-    "src/f32-gemm/gen/4x4-scalar.c",
-    "src/f32-gemm/gen/1x4-relu-scalar.c",
-    "src/f32-gemm/gen/2x4-relu-scalar.c",
-    "src/f32-gemm/gen/4x2-relu-scalar.c",
-    "src/f32-gemm/gen/4x4-relu-scalar.c",
-    "src/f32-gemm/gen/1x4-minmax-scalar.c",
-    "src/f32-gemm/gen/2x4-minmax-scalar.c",
-    "src/f32-gemm/gen/4x2-minmax-scalar.c",
-    "src/f32-gemm/gen/4x4-minmax-scalar.c",
     "src/f32-gemm/gen-inc/1x4inc-minmax-scalar.c",
     "src/f32-gemm/gen-inc/2x4inc-minmax-scalar.c",
     "src/f32-gemm/gen-inc/4x4inc-minmax-scalar.c",
+    "src/f32-gemm/gen/1x4-minmax-scalar.c",
+    "src/f32-gemm/gen/1x4-relu-scalar.c",
+    "src/f32-gemm/gen/1x4-scalar.c",
+    "src/f32-gemm/gen/2x4-minmax-scalar.c",
+    "src/f32-gemm/gen/2x4-relu-scalar.c",
+    "src/f32-gemm/gen/2x4-scalar.c",
+    "src/f32-gemm/gen/4x2-minmax-scalar.c",
+    "src/f32-gemm/gen/4x2-relu-scalar.c",
+    "src/f32-gemm/gen/4x2-scalar.c",
+    "src/f32-gemm/gen/4x4-minmax-scalar.c",
+    "src/f32-gemm/gen/4x4-relu-scalar.c",
+    "src/f32-gemm/gen/4x4-scalar.c",
     "src/f32-hswish/gen/hswish-scalar-x1.c",
     "src/f32-hswish/gen/hswish-scalar-x2.c",
     "src/f32-hswish/gen/hswish-scalar-x4.c",
-    "src/f32-ibilinear/gen/scalar-c1.c",
-    "src/f32-ibilinear/gen/scalar-c2.c",
-    "src/f32-ibilinear/gen/scalar-c4.c",
     "src/f32-ibilinear-chw/gen/scalar-p1.c",
     "src/f32-ibilinear-chw/gen/scalar-p2.c",
     "src/f32-ibilinear-chw/gen/scalar-p4.c",
-    "src/f32-igemm/gen/1x4-scalar.c",
-    "src/f32-igemm/gen/2x4-scalar.c",
-    "src/f32-igemm/gen/4x2-scalar.c",
-    "src/f32-igemm/gen/4x4-scalar.c",
-    "src/f32-igemm/gen/1x4-relu-scalar.c",
-    "src/f32-igemm/gen/2x4-relu-scalar.c",
-    "src/f32-igemm/gen/4x2-relu-scalar.c",
-    "src/f32-igemm/gen/4x4-relu-scalar.c",
+    "src/f32-ibilinear/gen/scalar-c1.c",
+    "src/f32-ibilinear/gen/scalar-c2.c",
+    "src/f32-ibilinear/gen/scalar-c4.c",
     "src/f32-igemm/gen/1x4-minmax-scalar.c",
+    "src/f32-igemm/gen/1x4-relu-scalar.c",
+    "src/f32-igemm/gen/1x4-scalar.c",
     "src/f32-igemm/gen/2x4-minmax-scalar.c",
+    "src/f32-igemm/gen/2x4-relu-scalar.c",
+    "src/f32-igemm/gen/2x4-scalar.c",
     "src/f32-igemm/gen/4x2-minmax-scalar.c",
+    "src/f32-igemm/gen/4x2-relu-scalar.c",
+    "src/f32-igemm/gen/4x2-scalar.c",
     "src/f32-igemm/gen/4x4-minmax-scalar.c",
-    "src/f32-vlrelu/gen/vlrelu-scalar-x1.c",
-    "src/f32-vlrelu/gen/vlrelu-scalar-x2.c",
-    "src/f32-vlrelu/gen/vlrelu-scalar-x4.c",
+    "src/f32-igemm/gen/4x4-relu-scalar.c",
+    "src/f32-igemm/gen/4x4-scalar.c",
     "src/f32-maxpool/9p8x-minmax-scalar-c1.c",
     "src/f32-pavgpool/9p8x-minmax-scalar-c1.c",
     "src/f32-pavgpool/9x-minmax-scalar-c1.c",
@@ -206,28 +203,28 @@
     "src/f32-prelu/gen/scalar-2x1.c",
     "src/f32-prelu/gen/scalar-2x4.c",
     "src/f32-raddstoreexpminusmax/gen/scalar-lut64-p2-x1.c",
-    "src/f32-raddstoreexpminusmax/gen/scalar-lut64-p2-x2.c",
     "src/f32-raddstoreexpminusmax/gen/scalar-lut64-p2-x2-acc2.c",
-    "src/f32-raddstoreexpminusmax/gen/scalar-lut64-p2-x4.c",
+    "src/f32-raddstoreexpminusmax/gen/scalar-lut64-p2-x2.c",
     "src/f32-raddstoreexpminusmax/gen/scalar-lut64-p2-x4-acc2.c",
     "src/f32-raddstoreexpminusmax/gen/scalar-lut64-p2-x4-acc4.c",
+    "src/f32-raddstoreexpminusmax/gen/scalar-lut64-p2-x4.c",
     "src/f32-raddstoreexpminusmax/gen/scalar-p5-x1.c",
-    "src/f32-raddstoreexpminusmax/gen/scalar-p5-x2.c",
     "src/f32-raddstoreexpminusmax/gen/scalar-p5-x2-acc2.c",
-    "src/f32-raddstoreexpminusmax/gen/scalar-p5-x4.c",
+    "src/f32-raddstoreexpminusmax/gen/scalar-p5-x2.c",
     "src/f32-raddstoreexpminusmax/gen/scalar-p5-x4-acc2.c",
     "src/f32-raddstoreexpminusmax/gen/scalar-p5-x4-acc4.c",
-    "src/f32-rmax/scalar.c",
+    "src/f32-raddstoreexpminusmax/gen/scalar-p5-x4.c",
     "src/f32-relu/gen/scalar-x1.c",
     "src/f32-relu/gen/scalar-x2.c",
     "src/f32-relu/gen/scalar-x4.c",
     "src/f32-relu/gen/scalar-x8.c",
-    "src/f32-sigmoid/gen/scalar-lut2048-p1-div-x1.c",
-    "src/f32-sigmoid/gen/scalar-lut2048-p1-div-x2.c",
-    "src/f32-sigmoid/gen/scalar-lut2048-p1-div-x4.c",
+    "src/f32-rmax/scalar.c",
     "src/f32-sigmoid/gen/scalar-lut64-p2-div-x1.c",
     "src/f32-sigmoid/gen/scalar-lut64-p2-div-x2.c",
     "src/f32-sigmoid/gen/scalar-lut64-p2-div-x4.c",
+    "src/f32-sigmoid/gen/scalar-lut2048-p1-div-x1.c",
+    "src/f32-sigmoid/gen/scalar-lut2048-p1-div-x2.c",
+    "src/f32-sigmoid/gen/scalar-lut2048-p1-div-x4.c",
     "src/f32-sigmoid/gen/scalar-p5-div-x1.c",
     "src/f32-sigmoid/gen/scalar-p5-div-x2.c",
     "src/f32-sigmoid/gen/scalar-p5-div-x4.c",
@@ -385,6 +382,9 @@
     "src/f32-vbinary/gen/vsubc-scalar-x2.c",
     "src/f32-vbinary/gen/vsubc-scalar-x4.c",
     "src/f32-vbinary/gen/vsubc-scalar-x8.c",
+    "src/f32-vlrelu/gen/vlrelu-scalar-x1.c",
+    "src/f32-vlrelu/gen/vlrelu-scalar-x2.c",
+    "src/f32-vlrelu/gen/vlrelu-scalar-x4.c",
     "src/f32-vmulcaddc/gen/c1-minmax-scalar-2x.c",
     "src/f32-vmulcaddc/gen/c2-minmax-scalar-2x.c",
     "src/f32-vmulcaddc/gen/c4-minmax-scalar-2x.c",
@@ -412,23 +412,23 @@
     "src/f32-vunary/gen/vsqr-scalar-x1.c",
     "src/f32-vunary/gen/vsqr-scalar-x2.c",
     "src/f32-vunary/gen/vsqr-scalar-x4.c",
-    "src/math/expminus-scalar-lut2048-p1.c",
     "src/math/expminus-scalar-lut64-p2.c",
+    "src/math/expminus-scalar-lut2048-p1.c",
     "src/math/expminus-scalar-p5.c",
-    "src/math/roundne-scalar-addsub.c",
-    "src/math/roundne-scalar-nearbyint.c",
-    "src/math/roundne-scalar-rint.c",
     "src/math/roundd-scalar-addsub.c",
     "src/math/roundd-scalar-cvt.c",
     "src/math/roundd-scalar-floor.c",
+    "src/math/roundne-scalar-addsub.c",
+    "src/math/roundne-scalar-nearbyint.c",
+    "src/math/roundne-scalar-rint.c",
     "src/math/roundu-scalar-addsub.c",
     "src/math/roundu-scalar-ceil.c",
     "src/math/roundu-scalar-cvt.c",
     "src/math/roundz-scalar-addsub.c",
     "src/math/roundz-scalar-cvt.c",
     "src/math/roundz-scalar-trunc.c",
-    "src/math/sigmoid-scalar-rr2-lut2048-p1-div.c",
     "src/math/sigmoid-scalar-rr2-lut64-p2-div.c",
+    "src/math/sigmoid-scalar-rr2-lut2048-p1-div.c",
     "src/math/sigmoid-scalar-rr2-p5-div.c",
     "src/qs8-gemm/gen/8x8c4-minmax-scalar.c",
     "src/qs8-gemm/gen/12x4c4-minmax-scalar.c",
@@ -443,9 +443,9 @@
     "src/qu8-dwconv/up1x9-minmax-scalar.c",
     "src/qu8-gavgpool/7p7x-minmax-scalar-c1.c",
     "src/qu8-gavgpool/7x-minmax-scalar-c1.c",
+    "src/qu8-gemm/2x2-minmax-scalar.c",
     "src/qu8-gemm/gen/8x8c4-minmax-scalar.c",
     "src/qu8-gemm/gen/12x4c4-minmax-scalar.c",
-    "src/qu8-gemm/2x2-minmax-scalar.c",
     "src/qu8-igemm/2x2-minmax-scalar.c",
     "src/qu8-requantization/fp32-scalar-lrintf.c",
     "src/qu8-requantization/fp32-scalar-magic.c",
@@ -458,6 +458,11 @@
     "src/u8-lut32norm/scalar.c",
     "src/u8-maxpool/9p8x-minmax-scalar-c1.c",
     "src/u8-rmax/scalar.c",
+    "src/x8-lut/scalar.c",
+    "src/x8-zip/x2-scalar.c",
+    "src/x8-zip/x3-scalar.c",
+    "src/x8-zip/x4-scalar.c",
+    "src/x8-zip/xm-scalar.c",
     "src/x32-fill/scalar-float.c",
     "src/x32-fill/scalar-int.c",
     "src/x32-packx/x2-scalar.c",
@@ -470,11 +475,6 @@
     "src/x32-zip/x3-scalar.c",
     "src/x32-zip/x4-scalar.c",
     "src/x32-zip/xm-scalar.c",
-    "src/x8-lut/scalar.c",
-    "src/x8-zip/x2-scalar.c",
-    "src/x8-zip/x3-scalar.c",
-    "src/x8-zip/x4-scalar.c",
-    "src/x8-zip/xm-scalar.c",
 ]
 
 WASM_UKERNELS = [
@@ -483,62 +483,62 @@
     "src/f32-clamp/gen/wasm-x1.c",
     "src/f32-clamp/gen/wasm-x2.c",
     "src/f32-clamp/gen/wasm-x4.c",
-    "src/f32-dwconv/gen/up1x4-wasm-acc2.c",
-    "src/f32-dwconv/gen/up1x4-wasm.c",
-    "src/f32-dwconv/gen/up1x9-wasm-acc2.c",
-    "src/f32-dwconv/gen/up1x9-wasm.c",
-    "src/f32-dwconv/gen/up1x25-wasm-acc2.c",
-    "src/f32-dwconv/gen/up1x25-wasm.c",
-    "src/f32-dwconv/gen/up2x4-wasm-acc2.c",
-    "src/f32-dwconv/gen/up2x4-wasm.c",
-    "src/f32-dwconv/gen/up2x9-wasm-acc2.c",
-    "src/f32-dwconv/gen/up2x9-wasm.c",
-    "src/f32-dwconv/gen/up2x25-wasm-acc2.c",
-    "src/f32-dwconv/gen/up2x25-wasm.c",
     "src/f32-dwconv/gen/up1x4-minmax-wasm-acc2.c",
     "src/f32-dwconv/gen/up1x4-minmax-wasm.c",
+    "src/f32-dwconv/gen/up1x4-wasm-acc2.c",
+    "src/f32-dwconv/gen/up1x4-wasm.c",
     "src/f32-dwconv/gen/up1x9-minmax-wasm-acc2.c",
     "src/f32-dwconv/gen/up1x9-minmax-wasm.c",
+    "src/f32-dwconv/gen/up1x9-wasm-acc2.c",
+    "src/f32-dwconv/gen/up1x9-wasm.c",
     "src/f32-dwconv/gen/up1x25-minmax-wasm-acc2.c",
     "src/f32-dwconv/gen/up1x25-minmax-wasm.c",
+    "src/f32-dwconv/gen/up1x25-wasm-acc2.c",
+    "src/f32-dwconv/gen/up1x25-wasm.c",
     "src/f32-dwconv/gen/up2x4-minmax-wasm-acc2.c",
     "src/f32-dwconv/gen/up2x4-minmax-wasm.c",
+    "src/f32-dwconv/gen/up2x4-wasm-acc2.c",
+    "src/f32-dwconv/gen/up2x4-wasm.c",
     "src/f32-dwconv/gen/up2x9-minmax-wasm-acc2.c",
     "src/f32-dwconv/gen/up2x9-minmax-wasm.c",
+    "src/f32-dwconv/gen/up2x9-wasm-acc2.c",
+    "src/f32-dwconv/gen/up2x9-wasm.c",
     "src/f32-dwconv/gen/up2x25-minmax-wasm-acc2.c",
     "src/f32-dwconv/gen/up2x25-minmax-wasm.c",
+    "src/f32-dwconv/gen/up2x25-wasm-acc2.c",
+    "src/f32-dwconv/gen/up2x25-wasm.c",
     "src/f32-gavgpool/7p7x-minmax-wasm-c1.c",
     "src/f32-gavgpool/7x-minmax-wasm-c1.c",
     "src/f32-gemm/gen-inc/1x4inc-minmax-wasm.c",
     "src/f32-gemm/gen-inc/2x4inc-minmax-wasm.c",
     "src/f32-gemm/gen-inc/4x4inc-minmax-wasm.c",
-    "src/f32-gemm/gen/1x4-wasm.c",
-    "src/f32-gemm/gen/2x4-wasm.c",
-    "src/f32-gemm/gen/4x2-wasm.c",
-    "src/f32-gemm/gen/4x4-wasm.c",
-    "src/f32-gemm/gen/1x4-relu-wasm.c",
-    "src/f32-gemm/gen/2x4-relu-wasm.c",
-    "src/f32-gemm/gen/4x2-relu-wasm.c",
-    "src/f32-gemm/gen/4x4-relu-wasm.c",
     "src/f32-gemm/gen/1x4-minmax-wasm.c",
+    "src/f32-gemm/gen/1x4-relu-wasm.c",
+    "src/f32-gemm/gen/1x4-wasm.c",
     "src/f32-gemm/gen/2x4-minmax-wasm.c",
+    "src/f32-gemm/gen/2x4-relu-wasm.c",
+    "src/f32-gemm/gen/2x4-wasm.c",
     "src/f32-gemm/gen/4x2-minmax-wasm.c",
+    "src/f32-gemm/gen/4x2-relu-wasm.c",
+    "src/f32-gemm/gen/4x2-wasm.c",
     "src/f32-gemm/gen/4x4-minmax-wasm.c",
+    "src/f32-gemm/gen/4x4-relu-wasm.c",
+    "src/f32-gemm/gen/4x4-wasm.c",
     "src/f32-hswish/gen/hswish-wasm-x1.c",
     "src/f32-hswish/gen/hswish-wasm-x2.c",
     "src/f32-hswish/gen/hswish-wasm-x4.c",
-    "src/f32-igemm/gen/1x4-wasm.c",
-    "src/f32-igemm/gen/2x4-wasm.c",
-    "src/f32-igemm/gen/4x2-wasm.c",
-    "src/f32-igemm/gen/4x4-wasm.c",
-    "src/f32-igemm/gen/1x4-relu-wasm.c",
-    "src/f32-igemm/gen/2x4-relu-wasm.c",
-    "src/f32-igemm/gen/4x2-relu-wasm.c",
-    "src/f32-igemm/gen/4x4-relu-wasm.c",
     "src/f32-igemm/gen/1x4-minmax-wasm.c",
+    "src/f32-igemm/gen/1x4-relu-wasm.c",
+    "src/f32-igemm/gen/1x4-wasm.c",
     "src/f32-igemm/gen/2x4-minmax-wasm.c",
+    "src/f32-igemm/gen/2x4-relu-wasm.c",
+    "src/f32-igemm/gen/2x4-wasm.c",
     "src/f32-igemm/gen/4x2-minmax-wasm.c",
+    "src/f32-igemm/gen/4x2-relu-wasm.c",
+    "src/f32-igemm/gen/4x2-wasm.c",
     "src/f32-igemm/gen/4x4-minmax-wasm.c",
+    "src/f32-igemm/gen/4x4-relu-wasm.c",
+    "src/f32-igemm/gen/4x4-wasm.c",
     "src/f32-maxpool/9p8x-minmax-wasm-c1.c",
     "src/f32-pavgpool/9p8x-minmax-wasm-c1.c",
     "src/f32-pavgpool/9x-minmax-wasm-c1.c",
@@ -552,30 +552,30 @@
     "src/f32-vbinary/gen/vadd-minmax-wasm-x2.c",
     "src/f32-vbinary/gen/vadd-minmax-wasm-x4.c",
     "src/f32-vbinary/gen/vadd-minmax-wasm-x8.c",
-    "src/f32-vbinary/gen/vaddc-minmax-wasm-x1.c",
-    "src/f32-vbinary/gen/vaddc-minmax-wasm-x2.c",
-    "src/f32-vbinary/gen/vaddc-minmax-wasm-x4.c",
-    "src/f32-vbinary/gen/vaddc-minmax-wasm-x8.c",
-    "src/f32-vbinary/gen/vdiv-minmax-wasm-x1.c",
-    "src/f32-vbinary/gen/vdiv-minmax-wasm-x2.c",
-    "src/f32-vbinary/gen/vdiv-minmax-wasm-x4.c",
-    "src/f32-vbinary/gen/vdiv-minmax-wasm-x8.c",
-    "src/f32-vbinary/gen/vdivc-minmax-wasm-x1.c",
-    "src/f32-vbinary/gen/vdivc-minmax-wasm-x2.c",
-    "src/f32-vbinary/gen/vdivc-minmax-wasm-x4.c",
-    "src/f32-vbinary/gen/vdivc-minmax-wasm-x8.c",
     "src/f32-vbinary/gen/vadd-relu-wasm-x1.c",
     "src/f32-vbinary/gen/vadd-relu-wasm-x2.c",
     "src/f32-vbinary/gen/vadd-relu-wasm-x4.c",
     "src/f32-vbinary/gen/vadd-relu-wasm-x8.c",
+    "src/f32-vbinary/gen/vaddc-minmax-wasm-x1.c",
+    "src/f32-vbinary/gen/vaddc-minmax-wasm-x2.c",
+    "src/f32-vbinary/gen/vaddc-minmax-wasm-x4.c",
+    "src/f32-vbinary/gen/vaddc-minmax-wasm-x8.c",
     "src/f32-vbinary/gen/vaddc-relu-wasm-x1.c",
     "src/f32-vbinary/gen/vaddc-relu-wasm-x2.c",
     "src/f32-vbinary/gen/vaddc-relu-wasm-x4.c",
     "src/f32-vbinary/gen/vaddc-relu-wasm-x8.c",
+    "src/f32-vbinary/gen/vdiv-minmax-wasm-x1.c",
+    "src/f32-vbinary/gen/vdiv-minmax-wasm-x2.c",
+    "src/f32-vbinary/gen/vdiv-minmax-wasm-x4.c",
+    "src/f32-vbinary/gen/vdiv-minmax-wasm-x8.c",
     "src/f32-vbinary/gen/vdiv-relu-wasm-x1.c",
     "src/f32-vbinary/gen/vdiv-relu-wasm-x2.c",
     "src/f32-vbinary/gen/vdiv-relu-wasm-x4.c",
     "src/f32-vbinary/gen/vdiv-relu-wasm-x8.c",
+    "src/f32-vbinary/gen/vdivc-minmax-wasm-x1.c",
+    "src/f32-vbinary/gen/vdivc-minmax-wasm-x2.c",
+    "src/f32-vbinary/gen/vdivc-minmax-wasm-x4.c",
+    "src/f32-vbinary/gen/vdivc-minmax-wasm-x8.c",
     "src/f32-vbinary/gen/vdivc-relu-wasm-x1.c",
     "src/f32-vbinary/gen/vdivc-relu-wasm-x2.c",
     "src/f32-vbinary/gen/vdivc-relu-wasm-x4.c",
@@ -600,46 +600,46 @@
     "src/f32-vbinary/gen/vmul-minmax-wasm-x2.c",
     "src/f32-vbinary/gen/vmul-minmax-wasm-x4.c",
     "src/f32-vbinary/gen/vmul-minmax-wasm-x8.c",
-    "src/f32-vbinary/gen/vmulc-minmax-wasm-x1.c",
-    "src/f32-vbinary/gen/vmulc-minmax-wasm-x2.c",
-    "src/f32-vbinary/gen/vmulc-minmax-wasm-x4.c",
-    "src/f32-vbinary/gen/vmulc-minmax-wasm-x8.c",
-    "src/f32-vbinary/gen/vrdivc-minmax-wasm-x1.c",
-    "src/f32-vbinary/gen/vrdivc-minmax-wasm-x2.c",
-    "src/f32-vbinary/gen/vrdivc-minmax-wasm-x4.c",
-    "src/f32-vbinary/gen/vrdivc-minmax-wasm-x8.c",
-    "src/f32-vbinary/gen/vrsubc-minmax-wasm-x1.c",
-    "src/f32-vbinary/gen/vrsubc-minmax-wasm-x2.c",
-    "src/f32-vbinary/gen/vrsubc-minmax-wasm-x4.c",
-    "src/f32-vbinary/gen/vrsubc-minmax-wasm-x8.c",
-    "src/f32-vbinary/gen/vsub-minmax-wasm-x1.c",
-    "src/f32-vbinary/gen/vsub-minmax-wasm-x2.c",
-    "src/f32-vbinary/gen/vsub-minmax-wasm-x4.c",
-    "src/f32-vbinary/gen/vsub-minmax-wasm-x8.c",
-    "src/f32-vbinary/gen/vsubc-minmax-wasm-x1.c",
-    "src/f32-vbinary/gen/vsubc-minmax-wasm-x2.c",
-    "src/f32-vbinary/gen/vsubc-minmax-wasm-x4.c",
-    "src/f32-vbinary/gen/vsubc-minmax-wasm-x8.c",
     "src/f32-vbinary/gen/vmul-relu-wasm-x1.c",
     "src/f32-vbinary/gen/vmul-relu-wasm-x2.c",
     "src/f32-vbinary/gen/vmul-relu-wasm-x4.c",
     "src/f32-vbinary/gen/vmul-relu-wasm-x8.c",
+    "src/f32-vbinary/gen/vmulc-minmax-wasm-x1.c",
+    "src/f32-vbinary/gen/vmulc-minmax-wasm-x2.c",
+    "src/f32-vbinary/gen/vmulc-minmax-wasm-x4.c",
+    "src/f32-vbinary/gen/vmulc-minmax-wasm-x8.c",
     "src/f32-vbinary/gen/vmulc-relu-wasm-x1.c",
     "src/f32-vbinary/gen/vmulc-relu-wasm-x2.c",
     "src/f32-vbinary/gen/vmulc-relu-wasm-x4.c",
     "src/f32-vbinary/gen/vmulc-relu-wasm-x8.c",
+    "src/f32-vbinary/gen/vrdivc-minmax-wasm-x1.c",
+    "src/f32-vbinary/gen/vrdivc-minmax-wasm-x2.c",
+    "src/f32-vbinary/gen/vrdivc-minmax-wasm-x4.c",
+    "src/f32-vbinary/gen/vrdivc-minmax-wasm-x8.c",
     "src/f32-vbinary/gen/vrdivc-relu-wasm-x1.c",
     "src/f32-vbinary/gen/vrdivc-relu-wasm-x2.c",
     "src/f32-vbinary/gen/vrdivc-relu-wasm-x4.c",
     "src/f32-vbinary/gen/vrdivc-relu-wasm-x8.c",
+    "src/f32-vbinary/gen/vrsubc-minmax-wasm-x1.c",
+    "src/f32-vbinary/gen/vrsubc-minmax-wasm-x2.c",
+    "src/f32-vbinary/gen/vrsubc-minmax-wasm-x4.c",
+    "src/f32-vbinary/gen/vrsubc-minmax-wasm-x8.c",
     "src/f32-vbinary/gen/vrsubc-relu-wasm-x1.c",
     "src/f32-vbinary/gen/vrsubc-relu-wasm-x2.c",
     "src/f32-vbinary/gen/vrsubc-relu-wasm-x4.c",
     "src/f32-vbinary/gen/vrsubc-relu-wasm-x8.c",
+    "src/f32-vbinary/gen/vsub-minmax-wasm-x1.c",
+    "src/f32-vbinary/gen/vsub-minmax-wasm-x2.c",
+    "src/f32-vbinary/gen/vsub-minmax-wasm-x4.c",
+    "src/f32-vbinary/gen/vsub-minmax-wasm-x8.c",
     "src/f32-vbinary/gen/vsub-relu-wasm-x1.c",
     "src/f32-vbinary/gen/vsub-relu-wasm-x2.c",
     "src/f32-vbinary/gen/vsub-relu-wasm-x4.c",
     "src/f32-vbinary/gen/vsub-relu-wasm-x8.c",
+    "src/f32-vbinary/gen/vsubc-minmax-wasm-x1.c",
+    "src/f32-vbinary/gen/vsubc-minmax-wasm-x2.c",
+    "src/f32-vbinary/gen/vsubc-minmax-wasm-x4.c",
+    "src/f32-vbinary/gen/vsubc-minmax-wasm-x8.c",
     "src/f32-vbinary/gen/vsubc-relu-wasm-x1.c",
     "src/f32-vbinary/gen/vsubc-relu-wasm-x2.c",
     "src/f32-vbinary/gen/vsubc-relu-wasm-x4.c",
@@ -664,181 +664,163 @@
     "src/f32-clamp/gen/wasmsimd-arm-x8.c",
     "src/f32-clamp/gen/wasmsimd-x86-x4.c",
     "src/f32-clamp/gen/wasmsimd-x86-x8.c",
-    "src/f32-dwconv/gen/up4x25-wasmsimd.c",
-    "src/f32-dwconv/gen/up4x4-wasmsimd.c",
-    "src/f32-dwconv/gen/up4x9-wasmsimd.c",
-    "src/f32-dwconv/gen/up8x25-wasmsimd.c",
-    "src/f32-dwconv/gen/up8x4-wasmsimd.c",
-    "src/f32-dwconv/gen/up8x9-wasmsimd.c",
-    "src/f32-dwconv/gen/up4x25-minmax-wasmsimd-acc2-arm.c",
-    "src/f32-dwconv/gen/up4x25-minmax-wasmsimd-arm.c",
     "src/f32-dwconv/gen/up4x4-minmax-wasmsimd-acc2-arm.c",
-    "src/f32-dwconv/gen/up4x4-minmax-wasmsimd-arm.c",
-    "src/f32-dwconv/gen/up4x9-minmax-wasmsimd-acc2-arm.c",
-    "src/f32-dwconv/gen/up4x9-minmax-wasmsimd-arm.c",
-    "src/f32-dwconv/gen/up8x25-minmax-wasmsimd-acc2-arm.c",
-    "src/f32-dwconv/gen/up8x25-minmax-wasmsimd-arm.c",
-    "src/f32-dwconv/gen/up8x4-minmax-wasmsimd-acc2-arm.c",
-    "src/f32-dwconv/gen/up8x4-minmax-wasmsimd-arm.c",
-    "src/f32-dwconv/gen/up8x9-minmax-wasmsimd-acc2-arm.c",
-    "src/f32-dwconv/gen/up8x9-minmax-wasmsimd-arm.c",
-    "src/f32-dwconv/gen/up4x25-minmax-wasmsimd-acc2-x86.c",
-    "src/f32-dwconv/gen/up4x25-minmax-wasmsimd-x86.c",
     "src/f32-dwconv/gen/up4x4-minmax-wasmsimd-acc2-x86.c",
+    "src/f32-dwconv/gen/up4x4-minmax-wasmsimd-arm.c",
     "src/f32-dwconv/gen/up4x4-minmax-wasmsimd-x86.c",
+    "src/f32-dwconv/gen/up4x4-wasmsimd.c",
+    "src/f32-dwconv/gen/up4x9-minmax-wasmsimd-acc2-arm.c",
     "src/f32-dwconv/gen/up4x9-minmax-wasmsimd-acc2-x86.c",
+    "src/f32-dwconv/gen/up4x9-minmax-wasmsimd-arm.c",
     "src/f32-dwconv/gen/up4x9-minmax-wasmsimd-x86.c",
-    "src/f32-dwconv/gen/up8x25-minmax-wasmsimd-acc2-x86.c",
-    "src/f32-dwconv/gen/up8x25-minmax-wasmsimd-x86.c",
+    "src/f32-dwconv/gen/up4x9-wasmsimd.c",
+    "src/f32-dwconv/gen/up4x25-minmax-wasmsimd-acc2-arm.c",
+    "src/f32-dwconv/gen/up4x25-minmax-wasmsimd-acc2-x86.c",
+    "src/f32-dwconv/gen/up4x25-minmax-wasmsimd-arm.c",
+    "src/f32-dwconv/gen/up4x25-minmax-wasmsimd-x86.c",
+    "src/f32-dwconv/gen/up4x25-wasmsimd.c",
+    "src/f32-dwconv/gen/up8x4-minmax-wasmsimd-acc2-arm.c",
     "src/f32-dwconv/gen/up8x4-minmax-wasmsimd-acc2-x86.c",
+    "src/f32-dwconv/gen/up8x4-minmax-wasmsimd-arm.c",
     "src/f32-dwconv/gen/up8x4-minmax-wasmsimd-x86.c",
+    "src/f32-dwconv/gen/up8x4-wasmsimd.c",
+    "src/f32-dwconv/gen/up8x9-minmax-wasmsimd-acc2-arm.c",
     "src/f32-dwconv/gen/up8x9-minmax-wasmsimd-acc2-x86.c",
+    "src/f32-dwconv/gen/up8x9-minmax-wasmsimd-arm.c",
     "src/f32-dwconv/gen/up8x9-minmax-wasmsimd-x86.c",
-    "src/f32-gemm/gen/1x8-wasmsimd-splat.c",
-    "src/f32-gemm/gen/4x8-wasmsimd-splat.c",
-    "src/f32-gemm/gen/5x8-wasmsimd-splat.c",
-    "src/f32-gemm/gen/4x2c4-wasmsimd.c",
-    "src/f32-gemm/gen/1x8-relu-wasmsimd-splat.c",
-    "src/f32-gemm/gen/4x8-relu-wasmsimd-splat.c",
-    "src/f32-gemm/gen/5x8-relu-wasmsimd-splat.c",
-    "src/f32-gemm/gen/4x2c4-relu-wasmsimd.c",
-    "src/f32-gemm/gen/1x8-minmax-wasmsimd-loadsplat-arm.c",
-    "src/f32-gemm/gen/1x8-minmax-wasmsimd-splat-arm.c",
-    "src/f32-gemm/gen/1x8s4-minmax-wasmsimd-arm.c",
-    "src/f32-gemm/gen/3x8-minmax-wasmsimd-loadsplat-arm.c",
-    "src/f32-gemm/gen/3x8-minmax-wasmsimd-splat-arm.c",
-    "src/f32-gemm/gen/3x8s4-minmax-wasmsimd-arm.c",
-    "src/f32-gemm/gen/4x8-minmax-wasmsimd-loadsplat-arm.c",
-    "src/f32-gemm/gen/4x8-minmax-wasmsimd-splat-arm.c",
-    "src/f32-gemm/gen/4x8s4-minmax-wasmsimd-arm.c",
-    "src/f32-gemm/gen/5x8-minmax-wasmsimd-loadsplat-arm.c",
-    "src/f32-gemm/gen/5x8-minmax-wasmsimd-splat-arm.c",
-    "src/f32-gemm/gen/5x8s4-minmax-wasmsimd-arm.c",
-    "src/f32-gemm/gen/6x8-minmax-wasmsimd-loadsplat-arm.c",
-    "src/f32-gemm/gen/6x8-minmax-wasmsimd-splat-arm.c",
-    "src/f32-gemm/gen/6x8s4-minmax-wasmsimd-arm.c",
-    "src/f32-gemm/gen/1x8-minmax-wasmsimd-loadsplat-x86.c",
-    "src/f32-gemm/gen/1x8-minmax-wasmsimd-splat-x86.c",
-    "src/f32-gemm/gen/1x8s4-minmax-wasmsimd-x86.c",
-    "src/f32-gemm/gen/3x8-minmax-wasmsimd-loadsplat-x86.c",
-    "src/f32-gemm/gen/3x8-minmax-wasmsimd-splat-x86.c",
-    "src/f32-gemm/gen/3x8s4-minmax-wasmsimd-x86.c",
-    "src/f32-gemm/gen/4x8-minmax-wasmsimd-loadsplat-x86.c",
-    "src/f32-gemm/gen/4x8-minmax-wasmsimd-splat-x86.c",
-    "src/f32-gemm/gen/4x8s4-minmax-wasmsimd-x86.c",
-    "src/f32-gemm/gen/5x8-minmax-wasmsimd-loadsplat-x86.c",
-    "src/f32-gemm/gen/5x8-minmax-wasmsimd-splat-x86.c",
-    "src/f32-gemm/gen/5x8s4-minmax-wasmsimd-x86.c",
-    "src/f32-gemm/gen/6x8-minmax-wasmsimd-loadsplat-x86.c",
-    "src/f32-gemm/gen/6x8-minmax-wasmsimd-splat-x86.c",
-    "src/f32-gemm/gen/6x8s4-minmax-wasmsimd-x86.c",
-    "src/f32-gemm/gen/4x2c4-minmax-wasmsimd-arm.c",
-    "src/f32-gemm/gen/4x2c4-minmax-wasmsimd-x86.c",
-    "src/f32-gemm/gen-inc/1x8inc-minmax-wasmsimd-loadsplat-arm.c",
-    "src/f32-gemm/gen-inc/1x8inc-minmax-wasmsimd-splat-arm.c",
-    "src/f32-gemm/gen-inc/1x8s4inc-minmax-wasmsimd-arm.c",
-    "src/f32-gemm/gen-inc/3x8inc-minmax-wasmsimd-loadsplat-arm.c",
-    "src/f32-gemm/gen-inc/3x8inc-minmax-wasmsimd-splat-arm.c",
-    "src/f32-gemm/gen-inc/3x8s4inc-minmax-wasmsimd-arm.c",
-    "src/f32-gemm/gen-inc/4x8inc-minmax-wasmsimd-loadsplat-arm.c",
-    "src/f32-gemm/gen-inc/4x8inc-minmax-wasmsimd-splat-arm.c",
-    "src/f32-gemm/gen-inc/4x8s4inc-minmax-wasmsimd-arm.c",
-    "src/f32-gemm/gen-inc/5x8inc-minmax-wasmsimd-loadsplat-arm.c",
-    "src/f32-gemm/gen-inc/5x8inc-minmax-wasmsimd-splat-arm.c",
-    "src/f32-gemm/gen-inc/5x8s4inc-minmax-wasmsimd-arm.c",
-    "src/f32-gemm/gen-inc/6x8inc-minmax-wasmsimd-loadsplat-arm.c",
-    "src/f32-gemm/gen-inc/6x8inc-minmax-wasmsimd-splat-arm.c",
-    "src/f32-gemm/gen-inc/6x8s4inc-minmax-wasmsimd-arm.c",
-    "src/f32-gemm/gen-inc/1x8inc-minmax-wasmsimd-loadsplat-x86.c",
-    "src/f32-gemm/gen-inc/1x8inc-minmax-wasmsimd-splat-x86.c",
-    "src/f32-gemm/gen-inc/1x8s4inc-minmax-wasmsimd-x86.c",
-    "src/f32-gemm/gen-inc/3x8inc-minmax-wasmsimd-loadsplat-x86.c",
-    "src/f32-gemm/gen-inc/3x8inc-minmax-wasmsimd-splat-x86.c",
-    "src/f32-gemm/gen-inc/3x8s4inc-minmax-wasmsimd-x86.c",
-    "src/f32-gemm/gen-inc/4x8inc-minmax-wasmsimd-loadsplat-x86.c",
-    "src/f32-gemm/gen-inc/4x8inc-minmax-wasmsimd-splat-x86.c",
-    "src/f32-gemm/gen-inc/4x8s4inc-minmax-wasmsimd-x86.c",
-    "src/f32-gemm/gen-inc/5x8inc-minmax-wasmsimd-loadsplat-x86.c",
-    "src/f32-gemm/gen-inc/5x8inc-minmax-wasmsimd-splat-x86.c",
-    "src/f32-gemm/gen-inc/5x8s4inc-minmax-wasmsimd-x86.c",
-    "src/f32-gemm/gen-inc/6x8inc-minmax-wasmsimd-loadsplat-x86.c",
-    "src/f32-gemm/gen-inc/6x8inc-minmax-wasmsimd-splat-x86.c",
-    "src/f32-gemm/gen-inc/6x8s4inc-minmax-wasmsimd-x86.c",
-    "src/f32-ibilinear/gen/wasmsimd-c4.c",
-    "src/f32-ibilinear/gen/wasmsimd-c8.c",
-    "src/f32-igemm/gen/1x8-wasmsimd-splat.c",
-    "src/f32-igemm/gen/4x8-wasmsimd-splat.c",
-    "src/f32-igemm/gen/5x8-wasmsimd-splat.c",
-    "src/f32-igemm/gen/4x2c4-wasmsimd.c",
-    "src/f32-igemm/gen/1x8-relu-wasmsimd-splat.c",
-    "src/f32-igemm/gen/4x8-relu-wasmsimd-splat.c",
-    "src/f32-igemm/gen/5x8-relu-wasmsimd-splat.c",
-    "src/f32-igemm/gen/4x2c4-relu-wasmsimd.c",
-    "src/f32-igemm/gen/1x8-minmax-wasmsimd-loadsplat-arm.c",
-    "src/f32-igemm/gen/1x8-minmax-wasmsimd-splat-arm.c",
-    "src/f32-igemm/gen/1x8s4-minmax-wasmsimd-arm.c",
-    "src/f32-igemm/gen/3x8-minmax-wasmsimd-loadsplat-arm.c",
-    "src/f32-igemm/gen/3x8-minmax-wasmsimd-splat-arm.c",
-    "src/f32-igemm/gen/3x8s4-minmax-wasmsimd-arm.c",
-    "src/f32-igemm/gen/4x8-minmax-wasmsimd-loadsplat-arm.c",
-    "src/f32-igemm/gen/4x8-minmax-wasmsimd-splat-arm.c",
-    "src/f32-igemm/gen/4x8s4-minmax-wasmsimd-arm.c",
-    "src/f32-igemm/gen/5x8-minmax-wasmsimd-loadsplat-arm.c",
-    "src/f32-igemm/gen/5x8-minmax-wasmsimd-splat-arm.c",
-    "src/f32-igemm/gen/5x8s4-minmax-wasmsimd-arm.c",
-    "src/f32-igemm/gen/6x8-minmax-wasmsimd-loadsplat-arm.c",
-    "src/f32-igemm/gen/6x8-minmax-wasmsimd-splat-arm.c",
-    "src/f32-igemm/gen/6x8s4-minmax-wasmsimd-arm.c",
-    "src/f32-igemm/gen/1x8-minmax-wasmsimd-loadsplat-x86.c",
-    "src/f32-igemm/gen/1x8-minmax-wasmsimd-splat-x86.c",
-    "src/f32-igemm/gen/1x8s4-minmax-wasmsimd-x86.c",
-    "src/f32-igemm/gen/3x8-minmax-wasmsimd-loadsplat-x86.c",
-    "src/f32-igemm/gen/3x8-minmax-wasmsimd-splat-x86.c",
-    "src/f32-igemm/gen/3x8s4-minmax-wasmsimd-x86.c",
-    "src/f32-igemm/gen/4x8-minmax-wasmsimd-loadsplat-x86.c",
-    "src/f32-igemm/gen/4x8-minmax-wasmsimd-splat-x86.c",
-    "src/f32-igemm/gen/4x8s4-minmax-wasmsimd-x86.c",
-    "src/f32-igemm/gen/5x8-minmax-wasmsimd-loadsplat-x86.c",
-    "src/f32-igemm/gen/5x8-minmax-wasmsimd-splat-x86.c",
-    "src/f32-igemm/gen/5x8s4-minmax-wasmsimd-x86.c",
-    "src/f32-igemm/gen/6x8-minmax-wasmsimd-loadsplat-x86.c",
-    "src/f32-igemm/gen/6x8-minmax-wasmsimd-splat-x86.c",
-    "src/f32-igemm/gen/6x8s4-minmax-wasmsimd-x86.c",
-    "src/f32-igemm/gen/4x2c4-minmax-wasmsimd-arm.c",
-    "src/f32-igemm/gen/4x2c4-minmax-wasmsimd-x86.c",
+    "src/f32-dwconv/gen/up8x9-wasmsimd.c",
+    "src/f32-dwconv/gen/up8x25-minmax-wasmsimd-acc2-arm.c",
+    "src/f32-dwconv/gen/up8x25-minmax-wasmsimd-acc2-x86.c",
+    "src/f32-dwconv/gen/up8x25-minmax-wasmsimd-arm.c",
+    "src/f32-dwconv/gen/up8x25-minmax-wasmsimd-x86.c",
+    "src/f32-dwconv/gen/up8x25-wasmsimd.c",
+    "src/f32-gavgpool-cw/wasmsimd-arm-x4.c",
+    "src/f32-gavgpool-cw/wasmsimd-x86-x4.c",
     "src/f32-gavgpool/7p7x-minmax-wasmsimd-arm-c4.c",
     "src/f32-gavgpool/7p7x-minmax-wasmsimd-x86-c4.c",
     "src/f32-gavgpool/7x-minmax-wasmsimd-arm-c4.c",
     "src/f32-gavgpool/7x-minmax-wasmsimd-x86-c4.c",
-    "src/f32-gavgpool-cw/wasmsimd-arm-x4.c",
-    "src/f32-gavgpool-cw/wasmsimd-x86-x4.c",
+    "src/f32-gemm/gen-inc/1x8inc-minmax-wasmsimd-loadsplat-arm.c",
+    "src/f32-gemm/gen-inc/1x8inc-minmax-wasmsimd-loadsplat-x86.c",
+    "src/f32-gemm/gen-inc/1x8inc-minmax-wasmsimd-splat-arm.c",
+    "src/f32-gemm/gen-inc/1x8inc-minmax-wasmsimd-splat-x86.c",
+    "src/f32-gemm/gen-inc/1x8s4inc-minmax-wasmsimd-arm.c",
+    "src/f32-gemm/gen-inc/1x8s4inc-minmax-wasmsimd-x86.c",
+    "src/f32-gemm/gen-inc/3x8inc-minmax-wasmsimd-loadsplat-arm.c",
+    "src/f32-gemm/gen-inc/3x8inc-minmax-wasmsimd-loadsplat-x86.c",
+    "src/f32-gemm/gen-inc/3x8inc-minmax-wasmsimd-splat-arm.c",
+    "src/f32-gemm/gen-inc/3x8inc-minmax-wasmsimd-splat-x86.c",
+    "src/f32-gemm/gen-inc/3x8s4inc-minmax-wasmsimd-arm.c",
+    "src/f32-gemm/gen-inc/3x8s4inc-minmax-wasmsimd-x86.c",
+    "src/f32-gemm/gen-inc/4x8inc-minmax-wasmsimd-loadsplat-arm.c",
+    "src/f32-gemm/gen-inc/4x8inc-minmax-wasmsimd-loadsplat-x86.c",
+    "src/f32-gemm/gen-inc/4x8inc-minmax-wasmsimd-splat-arm.c",
+    "src/f32-gemm/gen-inc/4x8inc-minmax-wasmsimd-splat-x86.c",
+    "src/f32-gemm/gen-inc/4x8s4inc-minmax-wasmsimd-arm.c",
+    "src/f32-gemm/gen-inc/4x8s4inc-minmax-wasmsimd-x86.c",
+    "src/f32-gemm/gen-inc/5x8inc-minmax-wasmsimd-loadsplat-arm.c",
+    "src/f32-gemm/gen-inc/5x8inc-minmax-wasmsimd-loadsplat-x86.c",
+    "src/f32-gemm/gen-inc/5x8inc-minmax-wasmsimd-splat-arm.c",
+    "src/f32-gemm/gen-inc/5x8inc-minmax-wasmsimd-splat-x86.c",
+    "src/f32-gemm/gen-inc/5x8s4inc-minmax-wasmsimd-arm.c",
+    "src/f32-gemm/gen-inc/5x8s4inc-minmax-wasmsimd-x86.c",
+    "src/f32-gemm/gen-inc/6x8inc-minmax-wasmsimd-loadsplat-arm.c",
+    "src/f32-gemm/gen-inc/6x8inc-minmax-wasmsimd-loadsplat-x86.c",
+    "src/f32-gemm/gen-inc/6x8inc-minmax-wasmsimd-splat-arm.c",
+    "src/f32-gemm/gen-inc/6x8inc-minmax-wasmsimd-splat-x86.c",
+    "src/f32-gemm/gen-inc/6x8s4inc-minmax-wasmsimd-arm.c",
+    "src/f32-gemm/gen-inc/6x8s4inc-minmax-wasmsimd-x86.c",
+    "src/f32-gemm/gen/1x8-minmax-wasmsimd-loadsplat-arm.c",
+    "src/f32-gemm/gen/1x8-minmax-wasmsimd-loadsplat-x86.c",
+    "src/f32-gemm/gen/1x8-minmax-wasmsimd-splat-arm.c",
+    "src/f32-gemm/gen/1x8-minmax-wasmsimd-splat-x86.c",
+    "src/f32-gemm/gen/1x8-relu-wasmsimd-splat.c",
+    "src/f32-gemm/gen/1x8-wasmsimd-splat.c",
+    "src/f32-gemm/gen/1x8s4-minmax-wasmsimd-arm.c",
+    "src/f32-gemm/gen/1x8s4-minmax-wasmsimd-x86.c",
+    "src/f32-gemm/gen/3x8-minmax-wasmsimd-loadsplat-arm.c",
+    "src/f32-gemm/gen/3x8-minmax-wasmsimd-loadsplat-x86.c",
+    "src/f32-gemm/gen/3x8-minmax-wasmsimd-splat-arm.c",
+    "src/f32-gemm/gen/3x8-minmax-wasmsimd-splat-x86.c",
+    "src/f32-gemm/gen/3x8s4-minmax-wasmsimd-arm.c",
+    "src/f32-gemm/gen/3x8s4-minmax-wasmsimd-x86.c",
+    "src/f32-gemm/gen/4x2c4-minmax-wasmsimd-arm.c",
+    "src/f32-gemm/gen/4x2c4-minmax-wasmsimd-x86.c",
+    "src/f32-gemm/gen/4x2c4-relu-wasmsimd.c",
+    "src/f32-gemm/gen/4x2c4-wasmsimd.c",
+    "src/f32-gemm/gen/4x8-minmax-wasmsimd-loadsplat-arm.c",
+    "src/f32-gemm/gen/4x8-minmax-wasmsimd-loadsplat-x86.c",
+    "src/f32-gemm/gen/4x8-minmax-wasmsimd-splat-arm.c",
+    "src/f32-gemm/gen/4x8-minmax-wasmsimd-splat-x86.c",
+    "src/f32-gemm/gen/4x8-relu-wasmsimd-splat.c",
+    "src/f32-gemm/gen/4x8-wasmsimd-splat.c",
+    "src/f32-gemm/gen/4x8s4-minmax-wasmsimd-arm.c",
+    "src/f32-gemm/gen/4x8s4-minmax-wasmsimd-x86.c",
+    "src/f32-gemm/gen/5x8-minmax-wasmsimd-loadsplat-arm.c",
+    "src/f32-gemm/gen/5x8-minmax-wasmsimd-loadsplat-x86.c",
+    "src/f32-gemm/gen/5x8-minmax-wasmsimd-splat-arm.c",
+    "src/f32-gemm/gen/5x8-minmax-wasmsimd-splat-x86.c",
+    "src/f32-gemm/gen/5x8-relu-wasmsimd-splat.c",
+    "src/f32-gemm/gen/5x8-wasmsimd-splat.c",
+    "src/f32-gemm/gen/5x8s4-minmax-wasmsimd-arm.c",
+    "src/f32-gemm/gen/5x8s4-minmax-wasmsimd-x86.c",
+    "src/f32-gemm/gen/6x8-minmax-wasmsimd-loadsplat-arm.c",
+    "src/f32-gemm/gen/6x8-minmax-wasmsimd-loadsplat-x86.c",
+    "src/f32-gemm/gen/6x8-minmax-wasmsimd-splat-arm.c",
+    "src/f32-gemm/gen/6x8-minmax-wasmsimd-splat-x86.c",
+    "src/f32-gemm/gen/6x8s4-minmax-wasmsimd-arm.c",
+    "src/f32-gemm/gen/6x8s4-minmax-wasmsimd-x86.c",
     "src/f32-hswish/gen/hswish-wasmsimd-x4.c",
     "src/f32-hswish/gen/hswish-wasmsimd-x8.c",
     "src/f32-hswish/gen/hswish-wasmsimd-x16.c",
+    "src/f32-ibilinear/gen/wasmsimd-c4.c",
+    "src/f32-ibilinear/gen/wasmsimd-c8.c",
+    "src/f32-igemm/gen/1x8-minmax-wasmsimd-loadsplat-arm.c",
+    "src/f32-igemm/gen/1x8-minmax-wasmsimd-loadsplat-x86.c",
+    "src/f32-igemm/gen/1x8-minmax-wasmsimd-splat-arm.c",
+    "src/f32-igemm/gen/1x8-minmax-wasmsimd-splat-x86.c",
+    "src/f32-igemm/gen/1x8-relu-wasmsimd-splat.c",
+    "src/f32-igemm/gen/1x8-wasmsimd-splat.c",
+    "src/f32-igemm/gen/1x8s4-minmax-wasmsimd-arm.c",
+    "src/f32-igemm/gen/1x8s4-minmax-wasmsimd-x86.c",
+    "src/f32-igemm/gen/3x8-minmax-wasmsimd-loadsplat-arm.c",
+    "src/f32-igemm/gen/3x8-minmax-wasmsimd-loadsplat-x86.c",
+    "src/f32-igemm/gen/3x8-minmax-wasmsimd-splat-arm.c",
+    "src/f32-igemm/gen/3x8-minmax-wasmsimd-splat-x86.c",
+    "src/f32-igemm/gen/3x8s4-minmax-wasmsimd-arm.c",
+    "src/f32-igemm/gen/3x8s4-minmax-wasmsimd-x86.c",
+    "src/f32-igemm/gen/4x2c4-minmax-wasmsimd-arm.c",
+    "src/f32-igemm/gen/4x2c4-minmax-wasmsimd-x86.c",
+    "src/f32-igemm/gen/4x2c4-relu-wasmsimd.c",
+    "src/f32-igemm/gen/4x2c4-wasmsimd.c",
+    "src/f32-igemm/gen/4x8-minmax-wasmsimd-loadsplat-arm.c",
+    "src/f32-igemm/gen/4x8-minmax-wasmsimd-loadsplat-x86.c",
+    "src/f32-igemm/gen/4x8-minmax-wasmsimd-splat-arm.c",
+    "src/f32-igemm/gen/4x8-minmax-wasmsimd-splat-x86.c",
+    "src/f32-igemm/gen/4x8-relu-wasmsimd-splat.c",
+    "src/f32-igemm/gen/4x8-wasmsimd-splat.c",
+    "src/f32-igemm/gen/4x8s4-minmax-wasmsimd-arm.c",
+    "src/f32-igemm/gen/4x8s4-minmax-wasmsimd-x86.c",
+    "src/f32-igemm/gen/5x8-minmax-wasmsimd-loadsplat-arm.c",
+    "src/f32-igemm/gen/5x8-minmax-wasmsimd-loadsplat-x86.c",
+    "src/f32-igemm/gen/5x8-minmax-wasmsimd-splat-arm.c",
+    "src/f32-igemm/gen/5x8-minmax-wasmsimd-splat-x86.c",
+    "src/f32-igemm/gen/5x8-relu-wasmsimd-splat.c",
+    "src/f32-igemm/gen/5x8-wasmsimd-splat.c",
+    "src/f32-igemm/gen/5x8s4-minmax-wasmsimd-arm.c",
+    "src/f32-igemm/gen/5x8s4-minmax-wasmsimd-x86.c",
+    "src/f32-igemm/gen/6x8-minmax-wasmsimd-loadsplat-arm.c",
+    "src/f32-igemm/gen/6x8-minmax-wasmsimd-loadsplat-x86.c",
+    "src/f32-igemm/gen/6x8-minmax-wasmsimd-splat-arm.c",
+    "src/f32-igemm/gen/6x8-minmax-wasmsimd-splat-x86.c",
+    "src/f32-igemm/gen/6x8s4-minmax-wasmsimd-arm.c",
+    "src/f32-igemm/gen/6x8s4-minmax-wasmsimd-x86.c",
     "src/f32-maxpool/9p8x-minmax-wasmsimd-arm-c4.c",
     "src/f32-maxpool/9p8x-minmax-wasmsimd-x86-c4.c",
     "src/f32-pavgpool/9p8x-minmax-wasmsimd-arm-c4.c",
     "src/f32-pavgpool/9p8x-minmax-wasmsimd-x86-c4.c",
     "src/f32-pavgpool/9x-minmax-wasmsimd-arm-c4.c",
     "src/f32-pavgpool/9x-minmax-wasmsimd-x86-c4.c",
-    "src/f32-spmm/gen/4x1-minmax-wasmsimd-arm.c",
-    "src/f32-spmm/gen/4x1-minmax-wasmsimd-arm-unroll2.c",
-    "src/f32-spmm/gen/4x1-minmax-wasmsimd-arm-unroll4.c",
-    "src/f32-spmm/gen/8x1-minmax-wasmsimd-arm.c",
-    "src/f32-spmm/gen/8x1-minmax-wasmsimd-arm-unroll2.c",
-    "src/f32-spmm/gen/8x1-minmax-wasmsimd-arm-unroll4.c",
-    "src/f32-spmm/gen/16x1-minmax-wasmsimd-arm.c",
-    "src/f32-spmm/gen/16x1-minmax-wasmsimd-arm-unroll2.c",
-    "src/f32-spmm/gen/16x1-minmax-wasmsimd-arm-unroll4.c",
-    "src/f32-spmm/gen/4x1-minmax-wasmsimd-x86.c",
-    "src/f32-spmm/gen/4x1-minmax-wasmsimd-x86-unroll2.c",
-    "src/f32-spmm/gen/4x1-minmax-wasmsimd-x86-unroll4.c",
-    "src/f32-spmm/gen/8x1-minmax-wasmsimd-x86.c",
-    "src/f32-spmm/gen/8x1-minmax-wasmsimd-x86-unroll2.c",
-    "src/f32-spmm/gen/8x1-minmax-wasmsimd-x86-unroll4.c",
-    "src/f32-spmm/gen/16x1-minmax-wasmsimd-x86.c",
-    "src/f32-spmm/gen/16x1-minmax-wasmsimd-x86-unroll2.c",
-    "src/f32-spmm/gen/16x1-minmax-wasmsimd-x86-unroll4.c",
     "src/f32-ppmm/gen/4x8-minmax-wasmsimd-splat-arm.c",
     "src/f32-ppmm/gen/4x8-minmax-wasmsimd-splat-x86.c",
     "src/f32-prelu/gen/wasmsimd-bitselect-1x4.c",
@@ -860,34 +842,52 @@
     "src/f32-prelu/gen/wasmsimd-minmax-4x8.c",
     "src/f32-prelu/gen/wasmsimd-minmax-4x16.c",
     "src/f32-raddstoreexpminusmax/gen/wasmsimd-p5-x4.c",
-    "src/f32-raddstoreexpminusmax/gen/wasmsimd-p5-x8.c",
     "src/f32-raddstoreexpminusmax/gen/wasmsimd-p5-x8-acc2.c",
-    "src/f32-raddstoreexpminusmax/gen/wasmsimd-p5-x12.c",
+    "src/f32-raddstoreexpminusmax/gen/wasmsimd-p5-x8.c",
     "src/f32-raddstoreexpminusmax/gen/wasmsimd-p5-x12-acc2.c",
     "src/f32-raddstoreexpminusmax/gen/wasmsimd-p5-x12-acc3.c",
-    "src/f32-raddstoreexpminusmax/gen/wasmsimd-p5-x16.c",
+    "src/f32-raddstoreexpminusmax/gen/wasmsimd-p5-x12.c",
     "src/f32-raddstoreexpminusmax/gen/wasmsimd-p5-x16-acc2.c",
     "src/f32-raddstoreexpminusmax/gen/wasmsimd-p5-x16-acc4.c",
-    "src/f32-raddstoreexpminusmax/gen/wasmsimd-p5-x20.c",
+    "src/f32-raddstoreexpminusmax/gen/wasmsimd-p5-x16.c",
     "src/f32-raddstoreexpminusmax/gen/wasmsimd-p5-x20-acc2.c",
     "src/f32-raddstoreexpminusmax/gen/wasmsimd-p5-x20-acc5.c",
+    "src/f32-raddstoreexpminusmax/gen/wasmsimd-p5-x20.c",
     "src/f32-relu/gen/wasmsimd-x4.c",
     "src/f32-relu/gen/wasmsimd-x8.c",
     "src/f32-relu/gen/wasmsimd-x16.c",
     "src/f32-rmax/wasmsimd-arm.c",
     "src/f32-rmax/wasmsimd-x86.c",
-    "src/f32-sigmoid/gen/wasmsimd-p5-div-x4.c",
-    "src/f32-sigmoid/gen/wasmsimd-p5-div-x8.c",
-    "src/f32-sigmoid/gen/wasmsimd-p5-div-x12.c",
-    "src/f32-sigmoid/gen/wasmsimd-p5-div-x16.c",
-    "src/f32-sigmoid/gen/wasmsimd-p5-div-x20.c",
-    "src/f32-sigmoid/gen/wasmsimd-p5-div-x24.c",
     "src/f32-sigmoid/gen/wasmsimd-lut64-p2-div-x4.c",
     "src/f32-sigmoid/gen/wasmsimd-lut64-p2-div-x8.c",
     "src/f32-sigmoid/gen/wasmsimd-lut64-p2-div-x12.c",
     "src/f32-sigmoid/gen/wasmsimd-lut64-p2-div-x16.c",
     "src/f32-sigmoid/gen/wasmsimd-lut64-p2-div-x20.c",
     "src/f32-sigmoid/gen/wasmsimd-lut64-p2-div-x24.c",
+    "src/f32-sigmoid/gen/wasmsimd-p5-div-x4.c",
+    "src/f32-sigmoid/gen/wasmsimd-p5-div-x8.c",
+    "src/f32-sigmoid/gen/wasmsimd-p5-div-x12.c",
+    "src/f32-sigmoid/gen/wasmsimd-p5-div-x16.c",
+    "src/f32-sigmoid/gen/wasmsimd-p5-div-x20.c",
+    "src/f32-sigmoid/gen/wasmsimd-p5-div-x24.c",
+    "src/f32-spmm/gen/4x1-minmax-wasmsimd-arm-unroll2.c",
+    "src/f32-spmm/gen/4x1-minmax-wasmsimd-arm-unroll4.c",
+    "src/f32-spmm/gen/4x1-minmax-wasmsimd-arm.c",
+    "src/f32-spmm/gen/4x1-minmax-wasmsimd-x86-unroll2.c",
+    "src/f32-spmm/gen/4x1-minmax-wasmsimd-x86-unroll4.c",
+    "src/f32-spmm/gen/4x1-minmax-wasmsimd-x86.c",
+    "src/f32-spmm/gen/8x1-minmax-wasmsimd-arm-unroll2.c",
+    "src/f32-spmm/gen/8x1-minmax-wasmsimd-arm-unroll4.c",
+    "src/f32-spmm/gen/8x1-minmax-wasmsimd-arm.c",
+    "src/f32-spmm/gen/8x1-minmax-wasmsimd-x86-unroll2.c",
+    "src/f32-spmm/gen/8x1-minmax-wasmsimd-x86-unroll4.c",
+    "src/f32-spmm/gen/8x1-minmax-wasmsimd-x86.c",
+    "src/f32-spmm/gen/16x1-minmax-wasmsimd-arm-unroll2.c",
+    "src/f32-spmm/gen/16x1-minmax-wasmsimd-arm-unroll4.c",
+    "src/f32-spmm/gen/16x1-minmax-wasmsimd-arm.c",
+    "src/f32-spmm/gen/16x1-minmax-wasmsimd-x86-unroll2.c",
+    "src/f32-spmm/gen/16x1-minmax-wasmsimd-x86-unroll4.c",
+    "src/f32-spmm/gen/16x1-minmax-wasmsimd-x86.c",
     "src/f32-vbinary/gen/vadd-minmax-wasmsimd-arm-x4.c",
     "src/f32-vbinary/gen/vadd-minmax-wasmsimd-arm-x8.c",
     "src/f32-vbinary/gen/vadd-minmax-wasmsimd-arm-x16.c",
@@ -1043,23 +1043,23 @@
     "src/f32-vlrelu/gen/vlrelu-wasmsimd-minmax-x4.c",
     "src/f32-vlrelu/gen/vlrelu-wasmsimd-minmax-x8.c",
     "src/f32-vmulcaddc/gen/c4-minmax-wasmsimd-arm-2x.c",
-    "src/f32-vmulcaddc/gen/c8-minmax-wasmsimd-arm-2x.c",
     "src/f32-vmulcaddc/gen/c4-minmax-wasmsimd-x86-2x.c",
+    "src/f32-vmulcaddc/gen/c8-minmax-wasmsimd-arm-2x.c",
     "src/f32-vmulcaddc/gen/c8-minmax-wasmsimd-x86-2x.c",
-    "src/f32-vrnd/gen/vrndne-wasmsimd-addsub-x4.c",
-    "src/f32-vrnd/gen/vrndne-wasmsimd-addsub-x8.c",
-    "src/f32-vrnd/gen/vrndz-wasmsimd-addsub-x4.c",
-    "src/f32-vrnd/gen/vrndz-wasmsimd-addsub-x8.c",
-    "src/f32-vrnd/gen/vrndz-wasmsimd-cvt-x4.c",
-    "src/f32-vrnd/gen/vrndz-wasmsimd-cvt-x8.c",
-    "src/f32-vrnd/gen/vrndu-wasmsimd-addsub-x4.c",
-    "src/f32-vrnd/gen/vrndu-wasmsimd-addsub-x8.c",
-    "src/f32-vrnd/gen/vrndu-wasmsimd-cvt-x4.c",
-    "src/f32-vrnd/gen/vrndu-wasmsimd-cvt-x8.c",
     "src/f32-vrnd/gen/vrndd-wasmsimd-addsub-x4.c",
     "src/f32-vrnd/gen/vrndd-wasmsimd-addsub-x8.c",
     "src/f32-vrnd/gen/vrndd-wasmsimd-cvt-x4.c",
     "src/f32-vrnd/gen/vrndd-wasmsimd-cvt-x8.c",
+    "src/f32-vrnd/gen/vrndne-wasmsimd-addsub-x4.c",
+    "src/f32-vrnd/gen/vrndne-wasmsimd-addsub-x8.c",
+    "src/f32-vrnd/gen/vrndu-wasmsimd-addsub-x4.c",
+    "src/f32-vrnd/gen/vrndu-wasmsimd-addsub-x8.c",
+    "src/f32-vrnd/gen/vrndu-wasmsimd-cvt-x4.c",
+    "src/f32-vrnd/gen/vrndu-wasmsimd-cvt-x8.c",
+    "src/f32-vrnd/gen/vrndz-wasmsimd-addsub-x4.c",
+    "src/f32-vrnd/gen/vrndz-wasmsimd-addsub-x8.c",
+    "src/f32-vrnd/gen/vrndz-wasmsimd-cvt-x4.c",
+    "src/f32-vrnd/gen/vrndz-wasmsimd-cvt-x8.c",
     "src/f32-vsqrt/gen/wasmsimd-sqrt-x4.c",
     "src/f32-vsqrt/gen/wasmsimd-sqrt-x8.c",
     "src/f32-vunary/gen/vabs-wasmsimd-x4.c",
@@ -1068,29 +1068,38 @@
     "src/f32-vunary/gen/vneg-wasmsimd-x8.c",
     "src/f32-vunary/gen/vsqr-wasmsimd-x4.c",
     "src/f32-vunary/gen/vsqr-wasmsimd-x8.c",
+    "src/math/roundd-wasmsimd-addsub.c",
+    "src/math/roundd-wasmsimd-cvt.c",
+    "src/math/roundne-wasmsimd-addsub.c",
+    "src/math/roundu-wasmsimd-addsub.c",
+    "src/math/roundu-wasmsimd-cvt.c",
+    "src/math/roundz-wasmsimd-addsub.c",
+    "src/math/roundz-wasmsimd-cvt.c",
+    "src/math/sigmoid-wasmsimd-rr2-lut64-p2-div.c",
+    "src/math/sigmoid-wasmsimd-rr2-p5-div.c",
     "src/qs8-dwconv/gen/up8x9-minmax-wasmsimd-mul16.c",
     "src/qs8-dwconv/gen/up16x9-minmax-wasmsimd-mul16.c",
     "src/qs8-dwconv/gen/up24x9-minmax-wasmsimd-mul16.c",
-    "src/qs8-gavgpool/gen/7x-minmax-wasmsimd-c8-acc2.c",
-    "src/qs8-gavgpool/gen/7x-minmax-wasmsimd-c16-acc2.c",
-    "src/qs8-gavgpool/gen/7x-minmax-wasmsimd-c24-acc2.c",
     "src/qs8-gavgpool/gen/7p7x-minmax-wasmsimd-c8-acc2.c",
     "src/qs8-gavgpool/gen/7p7x-minmax-wasmsimd-c16-acc2.c",
     "src/qs8-gavgpool/gen/7p7x-minmax-wasmsimd-c24-acc2.c",
+    "src/qs8-gavgpool/gen/7x-minmax-wasmsimd-c8-acc2.c",
+    "src/qs8-gavgpool/gen/7x-minmax-wasmsimd-c16-acc2.c",
+    "src/qs8-gavgpool/gen/7x-minmax-wasmsimd-c24-acc2.c",
     "src/qs8-gemm/gen/1x4c8-minmax-wasmsimd-ld64.c",
-    "src/qs8-gemm/gen/2x4c8-minmax-wasmsimd-ld64.c",
-    "src/qs8-gemm/gen/3x4c8-minmax-wasmsimd-ld64.c",
     "src/qs8-gemm/gen/1x4c8-minmax-wasmsimd-ld128.c",
-    "src/qs8-gemm/gen/2x4c8-minmax-wasmsimd-ld128.c",
-    "src/qs8-gemm/gen/3x4c8-minmax-wasmsimd-ld128.c",
     "src/qs8-gemm/gen/1x4c8-xw-minmax-wasmsimd.c",
+    "src/qs8-gemm/gen/2x4c8-minmax-wasmsimd-ld64.c",
+    "src/qs8-gemm/gen/2x4c8-minmax-wasmsimd-ld128.c",
     "src/qs8-gemm/gen/2x4c8-xw-minmax-wasmsimd.c",
+    "src/qs8-gemm/gen/3x4c8-minmax-wasmsimd-ld64.c",
+    "src/qs8-gemm/gen/3x4c8-minmax-wasmsimd-ld128.c",
     "src/qs8-gemm/gen/3x4c8-xw-minmax-wasmsimd.c",
     "src/qs8-igemm/gen/1x4c8-minmax-wasmsimd-ld64.c",
-    "src/qs8-igemm/gen/2x4c8-minmax-wasmsimd-ld64.c",
-    "src/qs8-igemm/gen/3x4c8-minmax-wasmsimd-ld64.c",
     "src/qs8-igemm/gen/1x4c8-minmax-wasmsimd-ld128.c",
+    "src/qs8-igemm/gen/2x4c8-minmax-wasmsimd-ld64.c",
     "src/qs8-igemm/gen/2x4c8-minmax-wasmsimd-ld128.c",
+    "src/qs8-igemm/gen/3x4c8-minmax-wasmsimd-ld64.c",
     "src/qs8-igemm/gen/3x4c8-minmax-wasmsimd-ld128.c",
     "src/qs8-requantization/fp32-wasmsimd.c",
     "src/qs8-requantization/q31-wasmsimd.c",
@@ -1112,15 +1121,6 @@
     "src/x32-zip/x3-wasmsimd.c",
     "src/x32-zip/x4-wasmsimd.c",
     "src/x32-zip/xm-wasmsimd.c",
-    "src/math/roundne-wasmsimd-addsub.c",
-    "src/math/roundd-wasmsimd-addsub.c",
-    "src/math/roundd-wasmsimd-cvt.c",
-    "src/math/roundu-wasmsimd-addsub.c",
-    "src/math/roundu-wasmsimd-cvt.c",
-    "src/math/roundz-wasmsimd-addsub.c",
-    "src/math/roundz-wasmsimd-cvt.c",
-    "src/math/sigmoid-wasmsimd-rr2-lut64-p2-div.c",
-    "src/math/sigmoid-wasmsimd-rr2-p5-div.c",
 ]
 
 PSIMD_FASTMATH_UKERNELS = [
@@ -1132,10 +1132,10 @@
 ]
 
 PSIMD_ACCMATH_UKERNELS = [
-    "src/qs8-requantization/precise-psimd.c",
     "src/qs8-requantization/fp32-psimd.c",
-    "src/qu8-requantization/precise-psimd.c",
+    "src/qs8-requantization/precise-psimd.c",
     "src/qu8-requantization/fp32-psimd.c",
+    "src/qu8-requantization/precise-psimd.c",
 ]
 
 # ISA-specific micro-kernels
@@ -1148,78 +1148,78 @@
     "src/f32-clamp/gen/neon-x4.c",
     "src/f32-clamp/gen/neon-x8.c",
     "src/f32-conv-hwc/gen/3x3s2p0p1c3x4-neon-2x1.c",
-    "src/f32-conv-hwc/gen/3x3s2p0p1c3x8-neon-2x1.c",
     "src/f32-conv-hwc/gen/3x3s2p0p1c3x4-neon-2x2.c",
+    "src/f32-conv-hwc/gen/3x3s2p0p1c3x8-neon-2x1.c",
     "src/f32-conv-hwc/gen/3x3s2p0p1c3x8-neon-2x2.c",
     "src/f32-conv-hwc/gen/3x3s2p1c3x4-neon-2x1.c",
-    "src/f32-conv-hwc/gen/3x3s2p1c3x8-neon-2x1.c",
     "src/f32-conv-hwc/gen/3x3s2p1c3x4-neon-2x2.c",
+    "src/f32-conv-hwc/gen/3x3s2p1c3x8-neon-2x1.c",
     "src/f32-conv-hwc/gen/3x3s2p1c3x8-neon-2x2.c",
-    "src/f32-dwconv/gen/up4x4-minmax-neon.c",
     "src/f32-dwconv/gen/up4x4-minmax-neon-acc2.c",
-    "src/f32-dwconv/gen/up8x4-minmax-neon.c",
-    "src/f32-dwconv/gen/up8x4-minmax-neon-acc2.c",
-    "src/f32-dwconv/gen/up4x9-minmax-neon.c",
+    "src/f32-dwconv/gen/up4x4-minmax-neon.c",
     "src/f32-dwconv/gen/up4x9-minmax-neon-acc2.c",
-    "src/f32-dwconv/gen/up8x9-minmax-neon.c",
-    "src/f32-dwconv/gen/up8x9-minmax-neon-acc2.c",
-    "src/f32-dwconv/gen/up4x25-minmax-neon.c",
+    "src/f32-dwconv/gen/up4x9-minmax-neon.c",
     "src/f32-dwconv/gen/up4x25-minmax-neon-acc2.c",
-    "src/f32-dwconv/gen/up8x25-minmax-neon.c",
+    "src/f32-dwconv/gen/up4x25-minmax-neon.c",
+    "src/f32-dwconv/gen/up8x4-minmax-neon-acc2.c",
+    "src/f32-dwconv/gen/up8x4-minmax-neon.c",
+    "src/f32-dwconv/gen/up8x9-minmax-neon-acc2.c",
+    "src/f32-dwconv/gen/up8x9-minmax-neon.c",
     "src/f32-dwconv/gen/up8x25-minmax-neon-acc2.c",
+    "src/f32-dwconv/gen/up8x25-minmax-neon.c",
     "src/f32-gavgpool-cw/neon-x4.c",
     "src/f32-gavgpool/7p7x-minmax-neon-c4.c",
     "src/f32-gavgpool/7x-minmax-neon-c4.c",
-    "src/f32-gemm/gen/1x8-minmax-neon-lane-ld64.c",
-    "src/f32-gemm/gen/4x2-minmax-neon-lane-ld64.c",
-    "src/f32-gemm/gen/4x8-minmax-neon-lane-ld128.c",
-    "src/f32-gemm/gen/4x8-minmax-neon-lane-ld64.c",
-    "src/f32-gemm/gen/5x8-minmax-neon-lane-ld64.c",
-    "src/f32-gemm/gen/6x8-minmax-neon-lane-ld64.c",
-    "src/f32-gemm/gen/6x8-minmax-neon-lane-ld128.c",
-    "src/f32-gemm/gen/1x8-minmax-neon-dup-ld64.c",
-    "src/f32-gemm/gen/4x8-minmax-neon-dup-ld128.c",
-    "src/f32-gemm/gen/4x8-minmax-neon-dup-ld64.c",
-    "src/f32-gemm/gen/6x8-minmax-neon-dup-ld64.c",
-    "src/f32-gemm/gen/6x8-minmax-neon-dup-ld128.c",
-    "src/f32-gemm/gen/1x8s4-minmax-neon.c",
-    "src/f32-gemm/gen/4x8s4-minmax-neon.c",
-    "src/f32-gemm/gen/6x8s4-minmax-neon.c",
-    "src/f32-gemm/gen/8x8s4-minmax-neon.c",
-    "src/f32-gemm/gen-inc/1x8inc-minmax-neon-lane-ld64.c",
-    "src/f32-gemm/gen-inc/4x8inc-minmax-neon-lane-ld128.c",
-    "src/f32-gemm/gen-inc/4x8inc-minmax-neon-lane-ld64.c",
-    "src/f32-gemm/gen-inc/5x8inc-minmax-neon-lane-ld64.c",
-    "src/f32-gemm/gen-inc/6x8inc-minmax-neon-lane-ld64.c",
-    "src/f32-gemm/gen-inc/6x8inc-minmax-neon-lane-ld128.c",
     "src/f32-gemm/gen-inc/1x8inc-minmax-neon-dup-ld64.c",
-    "src/f32-gemm/gen-inc/4x8inc-minmax-neon-dup-ld128.c",
+    "src/f32-gemm/gen-inc/1x8inc-minmax-neon-lane-ld64.c",
+    "src/f32-gemm/gen-inc/1x8s4inc-minmax-neon.c",
     "src/f32-gemm/gen-inc/4x8inc-minmax-neon-dup-ld64.c",
+    "src/f32-gemm/gen-inc/4x8inc-minmax-neon-dup-ld128.c",
+    "src/f32-gemm/gen-inc/4x8inc-minmax-neon-lane-ld64.c",
+    "src/f32-gemm/gen-inc/4x8inc-minmax-neon-lane-ld128.c",
+    "src/f32-gemm/gen-inc/4x8s4inc-minmax-neon.c",
+    "src/f32-gemm/gen-inc/5x8inc-minmax-neon-lane-ld64.c",
     "src/f32-gemm/gen-inc/6x8inc-minmax-neon-dup-ld64.c",
     "src/f32-gemm/gen-inc/6x8inc-minmax-neon-dup-ld128.c",
-    "src/f32-gemm/gen-inc/1x8s4inc-minmax-neon.c",
-    "src/f32-gemm/gen-inc/4x8s4inc-minmax-neon.c",
+    "src/f32-gemm/gen-inc/6x8inc-minmax-neon-lane-ld64.c",
+    "src/f32-gemm/gen-inc/6x8inc-minmax-neon-lane-ld128.c",
     "src/f32-gemm/gen-inc/6x8s4inc-minmax-neon.c",
     "src/f32-gemm/gen-inc/8x8s4inc-minmax-neon.c",
+    "src/f32-gemm/gen/1x8-minmax-neon-dup-ld64.c",
+    "src/f32-gemm/gen/1x8-minmax-neon-lane-ld64.c",
+    "src/f32-gemm/gen/1x8s4-minmax-neon.c",
+    "src/f32-gemm/gen/4x2-minmax-neon-lane-ld64.c",
+    "src/f32-gemm/gen/4x8-minmax-neon-dup-ld64.c",
+    "src/f32-gemm/gen/4x8-minmax-neon-dup-ld128.c",
+    "src/f32-gemm/gen/4x8-minmax-neon-lane-ld64.c",
+    "src/f32-gemm/gen/4x8-minmax-neon-lane-ld128.c",
+    "src/f32-gemm/gen/4x8s4-minmax-neon.c",
+    "src/f32-gemm/gen/5x8-minmax-neon-lane-ld64.c",
+    "src/f32-gemm/gen/6x8-minmax-neon-dup-ld64.c",
+    "src/f32-gemm/gen/6x8-minmax-neon-dup-ld128.c",
+    "src/f32-gemm/gen/6x8-minmax-neon-lane-ld64.c",
+    "src/f32-gemm/gen/6x8-minmax-neon-lane-ld128.c",
+    "src/f32-gemm/gen/6x8s4-minmax-neon.c",
+    "src/f32-gemm/gen/8x8s4-minmax-neon.c",
     "src/f32-hswish/gen/hswish-neon-x4.c",
     "src/f32-hswish/gen/hswish-neon-x8.c",
     "src/f32-hswish/gen/hswish-neon-x16.c",
     "src/f32-ibilinear/gen/neon-c4.c",
     "src/f32-ibilinear/gen/neon-c8.c",
+    "src/f32-igemm/gen/1x8-minmax-neon-dup-ld64.c",
     "src/f32-igemm/gen/1x8-minmax-neon-lane-ld64.c",
+    "src/f32-igemm/gen/1x8s4-minmax-neon.c",
     "src/f32-igemm/gen/4x2-minmax-neon-lane-ld64.c",
     "src/f32-igemm/gen/4x4-minmax-neon-lane-ld64.c",
-    "src/f32-igemm/gen/4x8-minmax-neon-lane-ld128.c",
-    "src/f32-igemm/gen/4x8-minmax-neon-lane-ld64.c",
-    "src/f32-igemm/gen/6x8-minmax-neon-lane-ld64.c",
-    "src/f32-igemm/gen/6x8-minmax-neon-lane-ld128.c",
-    "src/f32-igemm/gen/1x8-minmax-neon-dup-ld64.c",
-    "src/f32-igemm/gen/4x8-minmax-neon-dup-ld128.c",
     "src/f32-igemm/gen/4x8-minmax-neon-dup-ld64.c",
+    "src/f32-igemm/gen/4x8-minmax-neon-dup-ld128.c",
+    "src/f32-igemm/gen/4x8-minmax-neon-lane-ld64.c",
+    "src/f32-igemm/gen/4x8-minmax-neon-lane-ld128.c",
+    "src/f32-igemm/gen/4x8s4-minmax-neon.c",
     "src/f32-igemm/gen/6x8-minmax-neon-dup-ld64.c",
     "src/f32-igemm/gen/6x8-minmax-neon-dup-ld128.c",
-    "src/f32-igemm/gen/1x8s4-minmax-neon.c",
-    "src/f32-igemm/gen/4x8s4-minmax-neon.c",
+    "src/f32-igemm/gen/6x8-minmax-neon-lane-ld64.c",
+    "src/f32-igemm/gen/6x8-minmax-neon-lane-ld128.c",
     "src/f32-igemm/gen/6x8s4-minmax-neon.c",
     "src/f32-igemm/gen/8x8s4-minmax-neon.c",
     "src/f32-maxpool/9p8x-minmax-neon-c4.c",
@@ -1236,40 +1236,34 @@
     "src/f32-prelu/gen/neon-4x4.c",
     "src/f32-prelu/gen/neon-4x8.c",
     "src/f32-prelu/gen/neon-4x16.c",
-    "src/f32-raddstoreexpminusmax/gen/neon-p5-x4.c",
-    "src/f32-raddstoreexpminusmax/gen/neon-p5-x8.c",
-    "src/f32-raddstoreexpminusmax/gen/neon-p5-x8-acc2.c",
-    "src/f32-raddstoreexpminusmax/gen/neon-p5-x12.c",
-    "src/f32-raddstoreexpminusmax/gen/neon-p5-x12-acc2.c",
-    "src/f32-raddstoreexpminusmax/gen/neon-p5-x12-acc3.c",
-    "src/f32-raddstoreexpminusmax/gen/neon-p5-x16.c",
-    "src/f32-raddstoreexpminusmax/gen/neon-p5-x16-acc2.c",
-    "src/f32-raddstoreexpminusmax/gen/neon-p5-x16-acc4.c",
-    "src/f32-raddstoreexpminusmax/gen/neon-p5-x20.c",
-    "src/f32-raddstoreexpminusmax/gen/neon-p5-x20-acc2.c",
-    "src/f32-raddstoreexpminusmax/gen/neon-p5-x20-acc5.c",
     "src/f32-raddstoreexpminusmax/gen/neon-lut64-p2-x4.c",
-    "src/f32-raddstoreexpminusmax/gen/neon-lut64-p2-x8.c",
     "src/f32-raddstoreexpminusmax/gen/neon-lut64-p2-x8-acc2.c",
-    "src/f32-raddstoreexpminusmax/gen/neon-lut64-p2-x12.c",
+    "src/f32-raddstoreexpminusmax/gen/neon-lut64-p2-x8.c",
     "src/f32-raddstoreexpminusmax/gen/neon-lut64-p2-x12-acc2.c",
     "src/f32-raddstoreexpminusmax/gen/neon-lut64-p2-x12-acc3.c",
-    "src/f32-raddstoreexpminusmax/gen/neon-lut64-p2-x16.c",
+    "src/f32-raddstoreexpminusmax/gen/neon-lut64-p2-x12.c",
     "src/f32-raddstoreexpminusmax/gen/neon-lut64-p2-x16-acc2.c",
     "src/f32-raddstoreexpminusmax/gen/neon-lut64-p2-x16-acc4.c",
-    "src/f32-raddstoreexpminusmax/gen/neon-lut64-p2-x20.c",
+    "src/f32-raddstoreexpminusmax/gen/neon-lut64-p2-x16.c",
     "src/f32-raddstoreexpminusmax/gen/neon-lut64-p2-x20-acc2.c",
     "src/f32-raddstoreexpminusmax/gen/neon-lut64-p2-x20-acc5.c",
+    "src/f32-raddstoreexpminusmax/gen/neon-lut64-p2-x20.c",
+    "src/f32-raddstoreexpminusmax/gen/neon-p5-x4.c",
+    "src/f32-raddstoreexpminusmax/gen/neon-p5-x8-acc2.c",
+    "src/f32-raddstoreexpminusmax/gen/neon-p5-x8.c",
+    "src/f32-raddstoreexpminusmax/gen/neon-p5-x12-acc2.c",
+    "src/f32-raddstoreexpminusmax/gen/neon-p5-x12-acc3.c",
+    "src/f32-raddstoreexpminusmax/gen/neon-p5-x12.c",
+    "src/f32-raddstoreexpminusmax/gen/neon-p5-x16-acc2.c",
+    "src/f32-raddstoreexpminusmax/gen/neon-p5-x16-acc4.c",
+    "src/f32-raddstoreexpminusmax/gen/neon-p5-x16.c",
+    "src/f32-raddstoreexpminusmax/gen/neon-p5-x20-acc2.c",
+    "src/f32-raddstoreexpminusmax/gen/neon-p5-x20-acc5.c",
+    "src/f32-raddstoreexpminusmax/gen/neon-p5-x20.c",
     "src/f32-relu/gen/neon-x4.c",
     "src/f32-relu/gen/neon-x8.c",
     "src/f32-rmax/neon.c",
     "src/f32-sigmoid/gen/neon-frac-p9-p10-nr1recps-x16.c",
-    "src/f32-sigmoid/gen/neon-rr2-p5-nr2recps-x4.c",
-    "src/f32-sigmoid/gen/neon-rr2-p5-nr2recps-x8.c",
-    "src/f32-sigmoid/gen/neon-rr2-p5-nr2recps-x12.c",
-    "src/f32-sigmoid/gen/neon-rr2-p5-nr2recps-x16.c",
-    "src/f32-sigmoid/gen/neon-rr2-p5-nr2recps-x20.c",
-    "src/f32-sigmoid/gen/neon-rr2-p5-nr2recps-x24.c",
     "src/f32-sigmoid/gen/neon-rr2-lut64-p2-nr2recps-x4.c",
     "src/f32-sigmoid/gen/neon-rr2-lut64-p2-nr2recps-x8.c",
     "src/f32-sigmoid/gen/neon-rr2-lut64-p2-nr2recps-x12.c",
@@ -1282,6 +1276,12 @@
     "src/f32-sigmoid/gen/neon-rr2-lut2048-p1-nr2recps-x16.c",
     "src/f32-sigmoid/gen/neon-rr2-lut2048-p1-nr2recps-x20.c",
     "src/f32-sigmoid/gen/neon-rr2-lut2048-p1-nr2recps-x24.c",
+    "src/f32-sigmoid/gen/neon-rr2-p5-nr2recps-x4.c",
+    "src/f32-sigmoid/gen/neon-rr2-p5-nr2recps-x8.c",
+    "src/f32-sigmoid/gen/neon-rr2-p5-nr2recps-x12.c",
+    "src/f32-sigmoid/gen/neon-rr2-p5-nr2recps-x16.c",
+    "src/f32-sigmoid/gen/neon-rr2-p5-nr2recps-x20.c",
+    "src/f32-sigmoid/gen/neon-rr2-p5-nr2recps-x24.c",
     "src/f32-vbinary/gen/vadd-minmax-neon-x4.c",
     "src/f32-vbinary/gen/vadd-minmax-neon-x8.c",
     "src/f32-vbinary/gen/vaddc-minmax-neon-x4.c",
@@ -1312,42 +1312,59 @@
     "src/f32-vlrelu/gen/vlrelu-neon-x8.c",
     "src/f32-vmulcaddc/gen/c4-minmax-neon-2x.c",
     "src/f32-vmulcaddc/gen/c8-minmax-neon-2x.c",
-    "src/f32-vrnd/gen/vrndne-neon-x4.c",
-    "src/f32-vrnd/gen/vrndne-neon-x8.c",
-    "src/f32-vrnd/gen/vrndz-neon-x4.c",
-    "src/f32-vrnd/gen/vrndz-neon-x8.c",
-    "src/f32-vrnd/gen/vrndu-neon-x4.c",
-    "src/f32-vrnd/gen/vrndu-neon-x8.c",
     "src/f32-vrnd/gen/vrndd-neon-x4.c",
     "src/f32-vrnd/gen/vrndd-neon-x8.c",
+    "src/f32-vrnd/gen/vrndne-neon-x4.c",
+    "src/f32-vrnd/gen/vrndne-neon-x8.c",
+    "src/f32-vrnd/gen/vrndu-neon-x4.c",
+    "src/f32-vrnd/gen/vrndu-neon-x8.c",
+    "src/f32-vrnd/gen/vrndz-neon-x4.c",
+    "src/f32-vrnd/gen/vrndz-neon-x8.c",
     "src/f32-vunary/gen/vabs-neon-x4.c",
     "src/f32-vunary/gen/vabs-neon-x8.c",
     "src/f32-vunary/gen/vneg-neon-x4.c",
     "src/f32-vunary/gen/vneg-neon-x8.c",
     "src/f32-vunary/gen/vsqr-neon-x4.c",
     "src/f32-vunary/gen/vsqr-neon-x8.c",
+    "src/math/roundd-neon-addsub.c",
+    "src/math/roundd-neon-cvt.c",
+    "src/math/roundne-neon-addsub.c",
+    "src/math/roundu-neon-addsub.c",
+    "src/math/roundu-neon-cvt.c",
+    "src/math/roundz-neon-addsub.c",
+    "src/math/roundz-neon-cvt.c",
+    "src/math/sigmoid-neon-frac-p9-p10-nr1recps.c",
+    "src/math/sigmoid-neon-rr1-lut64-p2-nr2recps.c",
+    "src/math/sigmoid-neon-rr1-lut2048-p1-nr2recps.c",
+    "src/math/sigmoid-neon-rr1-p5-nr2recps.c",
+    "src/math/sigmoid-neon-rr2-lut64-p2-nr2recps.c",
+    "src/math/sigmoid-neon-rr2-lut2048-p1-nr2recps.c",
+    "src/math/sigmoid-neon-rr2-p5-nr2recps.c",
+    "src/math/sqrt-neon-nr1rsqrts.c",
+    "src/math/sqrt-neon-nr2rsqrts.c",
+    "src/math/sqrt-neon-nr3rsqrts.c",
     "src/qs8-dwconv/gen/up8x9-minmax-neon-mul16.c",
     "src/qs8-dwconv/gen/up16x9-minmax-neon-mul16.c",
     "src/qs8-dwconv/gen/up24x9-minmax-neon-mul16.c",
     "src/qs8-dwconv/gen/up32x9-minmax-neon-mul16.c",
-    "src/qs8-gavgpool/gen/7x-minmax-neon-c8-acc2.c",
-    "src/qs8-gavgpool/gen/7x-minmax-neon-c16-acc2.c",
-    "src/qs8-gavgpool/gen/7x-minmax-neon-c24-acc2.c",
-    "src/qs8-gavgpool/gen/7x-minmax-neon-c32-acc2.c",
     "src/qs8-gavgpool/gen/7p7x-minmax-neon-c8-acc2.c",
     "src/qs8-gavgpool/gen/7p7x-minmax-neon-c16-acc2.c",
     "src/qs8-gavgpool/gen/7p7x-minmax-neon-c24-acc2.c",
     "src/qs8-gavgpool/gen/7p7x-minmax-neon-c32-acc2.c",
+    "src/qs8-gavgpool/gen/7x-minmax-neon-c8-acc2.c",
+    "src/qs8-gavgpool/gen/7x-minmax-neon-c16-acc2.c",
+    "src/qs8-gavgpool/gen/7x-minmax-neon-c24-acc2.c",
+    "src/qs8-gavgpool/gen/7x-minmax-neon-c32-acc2.c",
     "src/qs8-gemm/gen/1x8-minmax-neon-mlal-lane.c",
-    "src/qs8-gemm/gen/2x8-minmax-neon-mlal-lane.c",
     "src/qs8-gemm/gen/1x16-minmax-neon-mlal-lane.c",
+    "src/qs8-gemm/gen/2x8-minmax-neon-mlal-lane.c",
     "src/qs8-gemm/gen/2x16-minmax-neon-mlal-lane.c",
     "src/qs8-igemm/gen/1x8-minmax-neon-mlal-lane.c",
-    "src/qs8-igemm/gen/2x8-minmax-neon-mlal-lane.c",
     "src/qs8-igemm/gen/1x16-minmax-neon-mlal-lane.c",
+    "src/qs8-igemm/gen/2x8-minmax-neon-mlal-lane.c",
     "src/qs8-igemm/gen/2x16-minmax-neon-mlal-lane.c",
-    "src/qs8-requantization/precise-neon.c",
     "src/qs8-requantization/fp32-neon.c",
+    "src/qs8-requantization/precise-neon.c",
     "src/qs8-requantization/q31-neon.c",
     "src/qs8-vadd/gen/minmax-neon-ld64-x8.c",
     "src/qs8-vadd/gen/minmax-neon-ld64-x16.c",
@@ -1366,13 +1383,17 @@
     "src/qu8-gemm/8x8-minmax-neon.c",
     "src/qu8-igemm/4x8-minmax-neon.c",
     "src/qu8-igemm/8x8-minmax-neon.c",
-    "src/qu8-requantization/precise-neon.c",
     "src/qu8-requantization/fp32-neon.c",
+    "src/qu8-requantization/precise-neon.c",
     "src/qu8-requantization/q31-neon.c",
     "src/qu8-vadd/minmax-neon.c",
     "src/u8-clamp/neon-x64.c",
     "src/u8-maxpool/9p8x-minmax-neon-c16.c",
     "src/u8-rmax/neon.c",
+    "src/x8-zip/x2-neon.c",
+    "src/x8-zip/x3-neon.c",
+    "src/x8-zip/x4-neon.c",
+    "src/x8-zip/xm-neon.c",
     "src/x32-fill/neon.c",
     "src/x32-packx/x4-neon-st4.c",
     "src/x32-pad/neon.c",
@@ -1381,151 +1402,130 @@
     "src/x32-zip/x3-neon.c",
     "src/x32-zip/x4-neon.c",
     "src/x32-zip/xm-neon.c",
-    "src/x8-zip/x2-neon.c",
-    "src/x8-zip/x3-neon.c",
-    "src/x8-zip/x4-neon.c",
-    "src/x8-zip/xm-neon.c",
-    "src/math/roundne-neon-addsub.c",
-    "src/math/roundd-neon-addsub.c",
-    "src/math/roundd-neon-cvt.c",
-    "src/math/roundu-neon-addsub.c",
-    "src/math/roundu-neon-cvt.c",
-    "src/math/roundz-neon-addsub.c",
-    "src/math/roundz-neon-cvt.c",
-    "src/math/sigmoid-neon-frac-p9-p10-nr1recps.c",
-    "src/math/sigmoid-neon-rr1-lut2048-p1-nr2recps.c",
-    "src/math/sigmoid-neon-rr1-lut64-p2-nr2recps.c",
-    "src/math/sigmoid-neon-rr1-p5-nr2recps.c",
-    "src/math/sigmoid-neon-rr2-lut2048-p1-nr2recps.c",
-    "src/math/sigmoid-neon-rr2-lut64-p2-nr2recps.c",
-    "src/math/sigmoid-neon-rr2-p5-nr2recps.c",
-    "src/math/sqrt-neon-nr1rsqrts.c",
-    "src/math/sqrt-neon-nr2rsqrts.c",
-    "src/math/sqrt-neon-nr3rsqrts.c",
 ]
 
 NEONFMA_UKERNELS = [
+    "src/f32-dwconv/gen/up4x4-minmax-neonfma-acc2.c",
+    "src/f32-dwconv/gen/up4x4-minmax-neonfma.c",
+    "src/f32-dwconv/gen/up4x9-minmax-neonfma-acc2.c",
+    "src/f32-dwconv/gen/up4x9-minmax-neonfma.c",
+    "src/f32-dwconv/gen/up4x25-minmax-neonfma-acc2.c",
+    "src/f32-dwconv/gen/up4x25-minmax-neonfma.c",
+    "src/f32-dwconv/gen/up8x4-minmax-neonfma-acc2.c",
+    "src/f32-dwconv/gen/up8x4-minmax-neonfma.c",
+    "src/f32-dwconv/gen/up8x9-minmax-neonfma-acc2.c",
+    "src/f32-dwconv/gen/up8x9-minmax-neonfma.c",
+    "src/f32-dwconv/gen/up8x25-minmax-neonfma-acc2.c",
+    "src/f32-dwconv/gen/up8x25-minmax-neonfma.c",
+    "src/f32-gemm/gen-inc/1x8inc-minmax-neonfma-dup-ld64.c",
+    "src/f32-gemm/gen-inc/1x8s4inc-minmax-neonfma.c",
+    "src/f32-gemm/gen-inc/4x8inc-minmax-neonfma-dup-ld64.c",
+    "src/f32-gemm/gen-inc/4x8inc-minmax-neonfma-dup-ld128.c",
+    "src/f32-gemm/gen-inc/4x8s4inc-minmax-neonfma.c",
+    "src/f32-gemm/gen-inc/6x8inc-minmax-neonfma-dup-ld64.c",
+    "src/f32-gemm/gen-inc/6x8inc-minmax-neonfma-dup-ld128.c",
+    "src/f32-gemm/gen-inc/6x8s4inc-minmax-neonfma.c",
+    "src/f32-gemm/gen-inc/8x8s4inc-minmax-neonfma.c",
+    "src/f32-gemm/gen/1x8-minmax-neonfma-dup-ld64.c",
+    "src/f32-gemm/gen/1x8s4-minmax-neonfma.c",
+    "src/f32-gemm/gen/4x8-minmax-neonfma-dup-ld64.c",
+    "src/f32-gemm/gen/4x8-minmax-neonfma-dup-ld128.c",
+    "src/f32-gemm/gen/4x8s4-minmax-neonfma.c",
+    "src/f32-gemm/gen/6x8-minmax-neonfma-dup-ld64.c",
+    "src/f32-gemm/gen/6x8-minmax-neonfma-dup-ld128.c",
+    "src/f32-gemm/gen/6x8s4-minmax-neonfma.c",
+    "src/f32-gemm/gen/8x8s4-minmax-neonfma.c",
     "src/f32-ibilinear/gen/neonfma-c4.c",
     "src/f32-ibilinear/gen/neonfma-c8.c",
     "src/f32-igemm/gen/1x8-minmax-neonfma-dup-ld64.c",
-    "src/f32-igemm/gen/4x8-minmax-neonfma-dup-ld128.c",
+    "src/f32-igemm/gen/1x8s4-minmax-neonfma.c",
     "src/f32-igemm/gen/4x8-minmax-neonfma-dup-ld64.c",
+    "src/f32-igemm/gen/4x8-minmax-neonfma-dup-ld128.c",
+    "src/f32-igemm/gen/4x8s4-minmax-neonfma.c",
     "src/f32-igemm/gen/6x8-minmax-neonfma-dup-ld64.c",
     "src/f32-igemm/gen/6x8-minmax-neonfma-dup-ld128.c",
-    "src/f32-igemm/gen/1x8s4-minmax-neonfma.c",
-    "src/f32-igemm/gen/4x8s4-minmax-neonfma.c",
     "src/f32-igemm/gen/6x8s4-minmax-neonfma.c",
     "src/f32-igemm/gen/8x8s4-minmax-neonfma.c",
-    "src/f32-dwconv/gen/up4x4-minmax-neonfma.c",
-    "src/f32-dwconv/gen/up4x4-minmax-neonfma-acc2.c",
-    "src/f32-dwconv/gen/up8x4-minmax-neonfma.c",
-    "src/f32-dwconv/gen/up8x4-minmax-neonfma-acc2.c",
-    "src/f32-dwconv/gen/up4x9-minmax-neonfma.c",
-    "src/f32-dwconv/gen/up4x9-minmax-neonfma-acc2.c",
-    "src/f32-dwconv/gen/up8x9-minmax-neonfma.c",
-    "src/f32-dwconv/gen/up8x9-minmax-neonfma-acc2.c",
-    "src/f32-dwconv/gen/up4x25-minmax-neonfma.c",
-    "src/f32-dwconv/gen/up4x25-minmax-neonfma-acc2.c",
-    "src/f32-dwconv/gen/up8x25-minmax-neonfma.c",
-    "src/f32-dwconv/gen/up8x25-minmax-neonfma-acc2.c",
-    "src/f32-gemm/gen/1x8-minmax-neonfma-dup-ld64.c",
-    "src/f32-gemm/gen/4x8-minmax-neonfma-dup-ld128.c",
-    "src/f32-gemm/gen/4x8-minmax-neonfma-dup-ld64.c",
-    "src/f32-gemm/gen/6x8-minmax-neonfma-dup-ld64.c",
-    "src/f32-gemm/gen/6x8-minmax-neonfma-dup-ld128.c",
-    "src/f32-gemm/gen/1x8s4-minmax-neonfma.c",
-    "src/f32-gemm/gen/4x8s4-minmax-neonfma.c",
-    "src/f32-gemm/gen/6x8s4-minmax-neonfma.c",
-    "src/f32-gemm/gen/8x8s4-minmax-neonfma.c",
-    "src/f32-gemm/gen-inc/1x8inc-minmax-neonfma-dup-ld64.c",
-    "src/f32-gemm/gen-inc/4x8inc-minmax-neonfma-dup-ld128.c",
-    "src/f32-gemm/gen-inc/4x8inc-minmax-neonfma-dup-ld64.c",
-    "src/f32-gemm/gen-inc/6x8inc-minmax-neonfma-dup-ld64.c",
-    "src/f32-gemm/gen-inc/6x8inc-minmax-neonfma-dup-ld128.c",
-    "src/f32-gemm/gen-inc/1x8s4inc-minmax-neonfma.c",
-    "src/f32-gemm/gen-inc/4x8s4inc-minmax-neonfma.c",
-    "src/f32-gemm/gen-inc/6x8s4inc-minmax-neonfma.c",
-    "src/f32-gemm/gen-inc/8x8s4inc-minmax-neonfma.c",
     "src/f32-ppmm/gen/4x8-minmax-neonfma.c",
     "src/f32-ppmm/gen/8x8-minmax-neonfma.c",
-    "src/f32-raddstoreexpminusmax/gen/neonfma-p5-x4.c",
-    "src/f32-raddstoreexpminusmax/gen/neonfma-p5-x8.c",
-    "src/f32-raddstoreexpminusmax/gen/neonfma-p5-x8-acc2.c",
-    "src/f32-raddstoreexpminusmax/gen/neonfma-p5-x12.c",
-    "src/f32-raddstoreexpminusmax/gen/neonfma-p5-x12-acc2.c",
-    "src/f32-raddstoreexpminusmax/gen/neonfma-p5-x12-acc3.c",
-    "src/f32-raddstoreexpminusmax/gen/neonfma-p5-x16.c",
-    "src/f32-raddstoreexpminusmax/gen/neonfma-p5-x16-acc2.c",
-    "src/f32-raddstoreexpminusmax/gen/neonfma-p5-x16-acc4.c",
-    "src/f32-raddstoreexpminusmax/gen/neonfma-p5-x20.c",
-    "src/f32-raddstoreexpminusmax/gen/neonfma-p5-x20-acc2.c",
-    "src/f32-raddstoreexpminusmax/gen/neonfma-p5-x20-acc5.c",
     "src/f32-raddstoreexpminusmax/gen/neonfma-lut64-p2-x4.c",
-    "src/f32-raddstoreexpminusmax/gen/neonfma-lut64-p2-x8.c",
     "src/f32-raddstoreexpminusmax/gen/neonfma-lut64-p2-x8-acc2.c",
-    "src/f32-raddstoreexpminusmax/gen/neonfma-lut64-p2-x12.c",
+    "src/f32-raddstoreexpminusmax/gen/neonfma-lut64-p2-x8.c",
     "src/f32-raddstoreexpminusmax/gen/neonfma-lut64-p2-x12-acc2.c",
     "src/f32-raddstoreexpminusmax/gen/neonfma-lut64-p2-x12-acc3.c",
-    "src/f32-raddstoreexpminusmax/gen/neonfma-lut64-p2-x16.c",
+    "src/f32-raddstoreexpminusmax/gen/neonfma-lut64-p2-x12.c",
     "src/f32-raddstoreexpminusmax/gen/neonfma-lut64-p2-x16-acc2.c",
     "src/f32-raddstoreexpminusmax/gen/neonfma-lut64-p2-x16-acc4.c",
-    "src/f32-raddstoreexpminusmax/gen/neonfma-lut64-p2-x20.c",
+    "src/f32-raddstoreexpminusmax/gen/neonfma-lut64-p2-x16.c",
     "src/f32-raddstoreexpminusmax/gen/neonfma-lut64-p2-x20-acc2.c",
     "src/f32-raddstoreexpminusmax/gen/neonfma-lut64-p2-x20-acc5.c",
-    "src/f32-sigmoid/gen/neonfma-rr1-p5-nr2fma-x4.c",
-    "src/f32-sigmoid/gen/neonfma-rr1-p5-nr2fma-x8.c",
-    "src/f32-sigmoid/gen/neonfma-rr1-p5-nr2fma-x12.c",
-    "src/f32-sigmoid/gen/neonfma-rr1-p5-nr2fma-x16.c",
-    "src/f32-sigmoid/gen/neonfma-rr1-p5-nr2fma-x20.c",
-    "src/f32-sigmoid/gen/neonfma-rr1-p5-nr2fma-x24.c",
-    "src/f32-sigmoid/gen/neonfma-rr1-p5-nr1recps1fma-x4.c",
-    "src/f32-sigmoid/gen/neonfma-rr1-p5-nr1recps1fma-x8.c",
-    "src/f32-sigmoid/gen/neonfma-rr1-p5-nr1recps1fma-x12.c",
-    "src/f32-sigmoid/gen/neonfma-rr1-p5-nr1recps1fma-x16.c",
-    "src/f32-sigmoid/gen/neonfma-rr1-p5-nr1recps1fma-x20.c",
-    "src/f32-sigmoid/gen/neonfma-rr1-p5-nr1recps1fma-x24.c",
-    "src/f32-sigmoid/gen/neonfma-rr1-p5-nr2recps-x4.c",
-    "src/f32-sigmoid/gen/neonfma-rr1-p5-nr2recps-x8.c",
-    "src/f32-sigmoid/gen/neonfma-rr1-p5-nr2recps-x12.c",
-    "src/f32-sigmoid/gen/neonfma-rr1-p5-nr2recps-x16.c",
-    "src/f32-sigmoid/gen/neonfma-rr1-p5-nr2recps-x20.c",
-    "src/f32-sigmoid/gen/neonfma-rr1-p5-nr2recps-x24.c",
-    "src/f32-sigmoid/gen/neonfma-rr1-lut64-p2-nr2fma-x4.c",
-    "src/f32-sigmoid/gen/neonfma-rr1-lut64-p2-nr2fma-x8.c",
-    "src/f32-sigmoid/gen/neonfma-rr1-lut64-p2-nr2fma-x12.c",
-    "src/f32-sigmoid/gen/neonfma-rr1-lut64-p2-nr2fma-x16.c",
-    "src/f32-sigmoid/gen/neonfma-rr1-lut64-p2-nr2fma-x20.c",
-    "src/f32-sigmoid/gen/neonfma-rr1-lut64-p2-nr2fma-x24.c",
+    "src/f32-raddstoreexpminusmax/gen/neonfma-lut64-p2-x20.c",
+    "src/f32-raddstoreexpminusmax/gen/neonfma-p5-x4.c",
+    "src/f32-raddstoreexpminusmax/gen/neonfma-p5-x8-acc2.c",
+    "src/f32-raddstoreexpminusmax/gen/neonfma-p5-x8.c",
+    "src/f32-raddstoreexpminusmax/gen/neonfma-p5-x12-acc2.c",
+    "src/f32-raddstoreexpminusmax/gen/neonfma-p5-x12-acc3.c",
+    "src/f32-raddstoreexpminusmax/gen/neonfma-p5-x12.c",
+    "src/f32-raddstoreexpminusmax/gen/neonfma-p5-x16-acc2.c",
+    "src/f32-raddstoreexpminusmax/gen/neonfma-p5-x16-acc4.c",
+    "src/f32-raddstoreexpminusmax/gen/neonfma-p5-x16.c",
+    "src/f32-raddstoreexpminusmax/gen/neonfma-p5-x20-acc2.c",
+    "src/f32-raddstoreexpminusmax/gen/neonfma-p5-x20-acc5.c",
+    "src/f32-raddstoreexpminusmax/gen/neonfma-p5-x20.c",
     "src/f32-sigmoid/gen/neonfma-rr1-lut64-p2-nr1recps1fma-x4.c",
     "src/f32-sigmoid/gen/neonfma-rr1-lut64-p2-nr1recps1fma-x8.c",
     "src/f32-sigmoid/gen/neonfma-rr1-lut64-p2-nr1recps1fma-x12.c",
     "src/f32-sigmoid/gen/neonfma-rr1-lut64-p2-nr1recps1fma-x16.c",
     "src/f32-sigmoid/gen/neonfma-rr1-lut64-p2-nr1recps1fma-x20.c",
     "src/f32-sigmoid/gen/neonfma-rr1-lut64-p2-nr1recps1fma-x24.c",
+    "src/f32-sigmoid/gen/neonfma-rr1-lut64-p2-nr2fma-x4.c",
+    "src/f32-sigmoid/gen/neonfma-rr1-lut64-p2-nr2fma-x8.c",
+    "src/f32-sigmoid/gen/neonfma-rr1-lut64-p2-nr2fma-x12.c",
+    "src/f32-sigmoid/gen/neonfma-rr1-lut64-p2-nr2fma-x16.c",
+    "src/f32-sigmoid/gen/neonfma-rr1-lut64-p2-nr2fma-x20.c",
+    "src/f32-sigmoid/gen/neonfma-rr1-lut64-p2-nr2fma-x24.c",
     "src/f32-sigmoid/gen/neonfma-rr1-lut64-p2-nr2recps-x4.c",
     "src/f32-sigmoid/gen/neonfma-rr1-lut64-p2-nr2recps-x8.c",
     "src/f32-sigmoid/gen/neonfma-rr1-lut64-p2-nr2recps-x12.c",
     "src/f32-sigmoid/gen/neonfma-rr1-lut64-p2-nr2recps-x16.c",
     "src/f32-sigmoid/gen/neonfma-rr1-lut64-p2-nr2recps-x20.c",
     "src/f32-sigmoid/gen/neonfma-rr1-lut64-p2-nr2recps-x24.c",
-    "src/f32-sigmoid/gen/neonfma-rr1-lut2048-p1-nr2fma-x4.c",
-    "src/f32-sigmoid/gen/neonfma-rr1-lut2048-p1-nr2fma-x8.c",
-    "src/f32-sigmoid/gen/neonfma-rr1-lut2048-p1-nr2fma-x12.c",
-    "src/f32-sigmoid/gen/neonfma-rr1-lut2048-p1-nr2fma-x16.c",
-    "src/f32-sigmoid/gen/neonfma-rr1-lut2048-p1-nr2fma-x20.c",
-    "src/f32-sigmoid/gen/neonfma-rr1-lut2048-p1-nr2fma-x24.c",
     "src/f32-sigmoid/gen/neonfma-rr1-lut2048-p1-nr1recps1fma-x4.c",
     "src/f32-sigmoid/gen/neonfma-rr1-lut2048-p1-nr1recps1fma-x8.c",
     "src/f32-sigmoid/gen/neonfma-rr1-lut2048-p1-nr1recps1fma-x12.c",
     "src/f32-sigmoid/gen/neonfma-rr1-lut2048-p1-nr1recps1fma-x16.c",
     "src/f32-sigmoid/gen/neonfma-rr1-lut2048-p1-nr1recps1fma-x20.c",
     "src/f32-sigmoid/gen/neonfma-rr1-lut2048-p1-nr1recps1fma-x24.c",
+    "src/f32-sigmoid/gen/neonfma-rr1-lut2048-p1-nr2fma-x4.c",
+    "src/f32-sigmoid/gen/neonfma-rr1-lut2048-p1-nr2fma-x8.c",
+    "src/f32-sigmoid/gen/neonfma-rr1-lut2048-p1-nr2fma-x12.c",
+    "src/f32-sigmoid/gen/neonfma-rr1-lut2048-p1-nr2fma-x16.c",
+    "src/f32-sigmoid/gen/neonfma-rr1-lut2048-p1-nr2fma-x20.c",
+    "src/f32-sigmoid/gen/neonfma-rr1-lut2048-p1-nr2fma-x24.c",
     "src/f32-sigmoid/gen/neonfma-rr1-lut2048-p1-nr2recps-x4.c",
     "src/f32-sigmoid/gen/neonfma-rr1-lut2048-p1-nr2recps-x8.c",
     "src/f32-sigmoid/gen/neonfma-rr1-lut2048-p1-nr2recps-x12.c",
     "src/f32-sigmoid/gen/neonfma-rr1-lut2048-p1-nr2recps-x16.c",
     "src/f32-sigmoid/gen/neonfma-rr1-lut2048-p1-nr2recps-x20.c",
     "src/f32-sigmoid/gen/neonfma-rr1-lut2048-p1-nr2recps-x24.c",
+    "src/f32-sigmoid/gen/neonfma-rr1-p5-nr1recps1fma-x4.c",
+    "src/f32-sigmoid/gen/neonfma-rr1-p5-nr1recps1fma-x8.c",
+    "src/f32-sigmoid/gen/neonfma-rr1-p5-nr1recps1fma-x12.c",
+    "src/f32-sigmoid/gen/neonfma-rr1-p5-nr1recps1fma-x16.c",
+    "src/f32-sigmoid/gen/neonfma-rr1-p5-nr1recps1fma-x20.c",
+    "src/f32-sigmoid/gen/neonfma-rr1-p5-nr1recps1fma-x24.c",
+    "src/f32-sigmoid/gen/neonfma-rr1-p5-nr2fma-x4.c",
+    "src/f32-sigmoid/gen/neonfma-rr1-p5-nr2fma-x8.c",
+    "src/f32-sigmoid/gen/neonfma-rr1-p5-nr2fma-x12.c",
+    "src/f32-sigmoid/gen/neonfma-rr1-p5-nr2fma-x16.c",
+    "src/f32-sigmoid/gen/neonfma-rr1-p5-nr2fma-x20.c",
+    "src/f32-sigmoid/gen/neonfma-rr1-p5-nr2fma-x24.c",
+    "src/f32-sigmoid/gen/neonfma-rr1-p5-nr2recps-x4.c",
+    "src/f32-sigmoid/gen/neonfma-rr1-p5-nr2recps-x8.c",
+    "src/f32-sigmoid/gen/neonfma-rr1-p5-nr2recps-x12.c",
+    "src/f32-sigmoid/gen/neonfma-rr1-p5-nr2recps-x16.c",
+    "src/f32-sigmoid/gen/neonfma-rr1-p5-nr2recps-x20.c",
+    "src/f32-sigmoid/gen/neonfma-rr1-p5-nr2recps-x24.c",
     "src/f32-vmulcaddc/gen/c4-minmax-neonfma-2x.c",
     "src/f32-vmulcaddc/gen/c8-minmax-neonfma-2x.c",
     "src/f32-vsqrt/gen/neonfma-nr1rsqrts1fma1adj-x4.c",
@@ -1550,80 +1550,68 @@
     "src/f32-vsqrt/gen/neonfma-nr2fma1adj-x40.c",
     "src/math/exp-neonfma-lut64-p2.c",
     "src/math/exp-neonfma-p5.c",
-    "src/math/expminus-neonfma-lut2048-p1.c",
     "src/math/expminus-neonfma-lut64-p2.c",
+    "src/math/expminus-neonfma-lut2048-p1.c",
     "src/math/expminus-neonfma-p5.c",
-    "src/math/sigmoid-neonfma-rr1-lut2048-p1-nr1recps1fma.c",
-    "src/math/sigmoid-neonfma-rr1-lut2048-p1-nr2fma.c",
-    "src/math/sigmoid-neonfma-rr1-lut2048-p1-nr2recps.c",
     "src/math/sigmoid-neonfma-rr1-lut64-p2-nr1recps1fma.c",
     "src/math/sigmoid-neonfma-rr1-lut64-p2-nr2fma.c",
     "src/math/sigmoid-neonfma-rr1-lut64-p2-nr2recps.c",
+    "src/math/sigmoid-neonfma-rr1-lut2048-p1-nr1recps1fma.c",
+    "src/math/sigmoid-neonfma-rr1-lut2048-p1-nr2fma.c",
+    "src/math/sigmoid-neonfma-rr1-lut2048-p1-nr2recps.c",
     "src/math/sigmoid-neonfma-rr1-p5-nr1recps1fma.c",
     "src/math/sigmoid-neonfma-rr1-p5-nr2fma.c",
     "src/math/sigmoid-neonfma-rr1-p5-nr2recps.c",
-    "src/math/sigmoid-neonfma-rr2-lut2048-p1-nr1recps1fma.c",
-    "src/math/sigmoid-neonfma-rr2-lut2048-p1-nr2fma.c",
-    "src/math/sigmoid-neonfma-rr2-lut2048-p1-nr2recps.c",
     "src/math/sigmoid-neonfma-rr2-lut64-p2-nr1recps1fma.c",
     "src/math/sigmoid-neonfma-rr2-lut64-p2-nr2fma.c",
     "src/math/sigmoid-neonfma-rr2-lut64-p2-nr2recps.c",
+    "src/math/sigmoid-neonfma-rr2-lut2048-p1-nr1recps1fma.c",
+    "src/math/sigmoid-neonfma-rr2-lut2048-p1-nr2fma.c",
+    "src/math/sigmoid-neonfma-rr2-lut2048-p1-nr2recps.c",
     "src/math/sigmoid-neonfma-rr2-p5-nr1recps1fma.c",
     "src/math/sigmoid-neonfma-rr2-p5-nr2fma.c",
     "src/math/sigmoid-neonfma-rr2-p5-nr2recps.c",
     "src/math/sqrt-neonfma-nr1fma.c",
-    "src/math/sqrt-neonfma-nr2fma.c",
-    "src/math/sqrt-neonfma-nr3fma.c",
-    "src/math/sqrt-neonfma-nr2fma1adj.c",
     "src/math/sqrt-neonfma-nr1rsqrts1fma1adj.c",
+    "src/math/sqrt-neonfma-nr2fma.c",
+    "src/math/sqrt-neonfma-nr2fma1adj.c",
+    "src/math/sqrt-neonfma-nr3fma.c",
 ]
 
 AARCH64_NEONFMA_UKERNELS = [
-    "src/f32-vbinary/gen/vdiv-minmax-neon-x4.c",
-    "src/f32-vbinary/gen/vdiv-minmax-neon-x8.c",
-    "src/f32-vbinary/gen/vdivc-minmax-neon-x4.c",
-    "src/f32-vbinary/gen/vdivc-minmax-neon-x8.c",
-    "src/f32-vbinary/gen/vrdivc-minmax-neon-x4.c",
-    "src/f32-vbinary/gen/vrdivc-minmax-neon-x8.c",
-    "src/f32-gemm/gen/1x8-minmax-neonfma-lane-ld64.c",
-    "src/f32-gemm/gen/4x2-minmax-neonfma-lane-ld64.c",
-    "src/f32-gemm/gen/4x8-minmax-neonfma-lane-ld128.c",
-    "src/f32-gemm/gen/4x8-minmax-neonfma-lane-ld64.c",
-    "src/f32-gemm/gen/5x8-minmax-neonfma-lane-ld64.c",
-    "src/f32-gemm/gen/6x8-minmax-neonfma-lane-ld64.c",
-    "src/f32-gemm/gen/6x8-minmax-neonfma-lane-ld128.c",
-    "src/f32-gemm/gen-inc/1x8inc-minmax-neonfma-lane-ld64.c",
-    "src/f32-gemm/gen-inc/4x8inc-minmax-neonfma-lane-ld128.c",
-    "src/f32-gemm/gen-inc/4x8inc-minmax-neonfma-lane-ld64.c",
-    "src/f32-gemm/gen-inc/5x8inc-minmax-neonfma-lane-ld64.c",
-    "src/f32-gemm/gen-inc/6x8inc-minmax-neonfma-lane-ld64.c",
-    "src/f32-gemm/gen-inc/6x8inc-minmax-neonfma-lane-ld128.c",
-    "src/f32-igemm/gen/1x8-minmax-neonfma-lane-ld64.c",
-    "src/f32-igemm/gen/4x2-minmax-neonfma-lane-ld64.c",
-    "src/f32-igemm/gen/4x4-minmax-neonfma-lane-ld64.c",
-    "src/f32-igemm/gen/4x8-minmax-neonfma-lane-ld128.c",
-    "src/f32-igemm/gen/4x8-minmax-neonfma-lane-ld64.c",
-    "src/f32-igemm/gen/6x8-minmax-neonfma-lane-ld64.c",
-    "src/f32-igemm/gen/6x8-minmax-neonfma-lane-ld128.c",
     "src/f32-conv-hwc/gen/3x3s2p0p1c3x4-neonfma-2x1.c",
-    "src/f32-conv-hwc/gen/3x3s2p0p1c3x8-neonfma-2x1.c",
     "src/f32-conv-hwc/gen/3x3s2p0p1c3x4-neonfma-2x2.c",
+    "src/f32-conv-hwc/gen/3x3s2p0p1c3x8-neonfma-2x1.c",
     "src/f32-conv-hwc/gen/3x3s2p0p1c3x8-neonfma-2x2.c",
     "src/f32-conv-hwc/gen/3x3s2p1c3x4-neonfma-2x1.c",
-    "src/f32-conv-hwc/gen/3x3s2p1c3x8-neonfma-2x1.c",
     "src/f32-conv-hwc/gen/3x3s2p1c3x4-neonfma-2x2.c",
+    "src/f32-conv-hwc/gen/3x3s2p1c3x8-neonfma-2x1.c",
     "src/f32-conv-hwc/gen/3x3s2p1c3x8-neonfma-2x2.c",
     "src/f32-conv-hwc2chw/3x3s2p1c3x4-neonfma-2x2.c",
     "src/f32-dwconv-chw/3x3p1-neonfma-3x4.c",
     "src/f32-dwconv-chw/3x3s2p1-neonfma-1x4-acc3.c",
     "src/f32-dwconv-chw/5x5p2-neonfma-3x4.c",
     "src/f32-dwconv-chw/5x5s2p2-neonfma-1x4-acc2.c",
-    "src/f32-sigmoid/gen/neonfma-rr1-p5-div-x4.c",
-    "src/f32-sigmoid/gen/neonfma-rr1-p5-div-x8.c",
-    "src/f32-sigmoid/gen/neonfma-rr1-p5-div-x12.c",
-    "src/f32-sigmoid/gen/neonfma-rr1-p5-div-x16.c",
-    "src/f32-sigmoid/gen/neonfma-rr1-p5-div-x20.c",
-    "src/f32-sigmoid/gen/neonfma-rr1-p5-div-x24.c",
+    "src/f32-gemm/gen-inc/1x8inc-minmax-neonfma-lane-ld64.c",
+    "src/f32-gemm/gen-inc/4x8inc-minmax-neonfma-lane-ld64.c",
+    "src/f32-gemm/gen-inc/4x8inc-minmax-neonfma-lane-ld128.c",
+    "src/f32-gemm/gen-inc/5x8inc-minmax-neonfma-lane-ld64.c",
+    "src/f32-gemm/gen-inc/6x8inc-minmax-neonfma-lane-ld64.c",
+    "src/f32-gemm/gen-inc/6x8inc-minmax-neonfma-lane-ld128.c",
+    "src/f32-gemm/gen/1x8-minmax-neonfma-lane-ld64.c",
+    "src/f32-gemm/gen/4x2-minmax-neonfma-lane-ld64.c",
+    "src/f32-gemm/gen/4x8-minmax-neonfma-lane-ld64.c",
+    "src/f32-gemm/gen/4x8-minmax-neonfma-lane-ld128.c",
+    "src/f32-gemm/gen/5x8-minmax-neonfma-lane-ld64.c",
+    "src/f32-gemm/gen/6x8-minmax-neonfma-lane-ld64.c",
+    "src/f32-gemm/gen/6x8-minmax-neonfma-lane-ld128.c",
+    "src/f32-igemm/gen/1x8-minmax-neonfma-lane-ld64.c",
+    "src/f32-igemm/gen/4x2-minmax-neonfma-lane-ld64.c",
+    "src/f32-igemm/gen/4x4-minmax-neonfma-lane-ld64.c",
+    "src/f32-igemm/gen/4x8-minmax-neonfma-lane-ld64.c",
+    "src/f32-igemm/gen/4x8-minmax-neonfma-lane-ld128.c",
+    "src/f32-igemm/gen/6x8-minmax-neonfma-lane-ld64.c",
+    "src/f32-igemm/gen/6x8-minmax-neonfma-lane-ld128.c",
     "src/f32-sigmoid/gen/neonfma-rr1-lut64-p2-div-x4.c",
     "src/f32-sigmoid/gen/neonfma-rr1-lut64-p2-div-x8.c",
     "src/f32-sigmoid/gen/neonfma-rr1-lut64-p2-div-x12.c",
@@ -1636,14 +1624,12 @@
     "src/f32-sigmoid/gen/neonfma-rr1-lut2048-p1-div-x16.c",
     "src/f32-sigmoid/gen/neonfma-rr1-lut2048-p1-div-x20.c",
     "src/f32-sigmoid/gen/neonfma-rr1-lut2048-p1-div-x24.c",
-    "src/f32-spmm/gen/12x1-minmax-neonfma.c",
-    "src/f32-spmm/gen/12x2-minmax-neonfma.c",
-    "src/f32-spmm/gen/12x4-minmax-neonfma.c",
-    "src/f32-spmm/gen/16x1-minmax-neonfma-pipelined.c",
-    "src/f32-spmm/gen/16x1-minmax-neonfma-unroll2.c",
-    "src/f32-spmm/gen/16x1-minmax-neonfma.c",
-    "src/f32-spmm/gen/16x2-minmax-neonfma.c",
-    "src/f32-spmm/gen/16x4-minmax-neonfma.c",
+    "src/f32-sigmoid/gen/neonfma-rr1-p5-div-x4.c",
+    "src/f32-sigmoid/gen/neonfma-rr1-p5-div-x8.c",
+    "src/f32-sigmoid/gen/neonfma-rr1-p5-div-x12.c",
+    "src/f32-sigmoid/gen/neonfma-rr1-p5-div-x16.c",
+    "src/f32-sigmoid/gen/neonfma-rr1-p5-div-x20.c",
+    "src/f32-sigmoid/gen/neonfma-rr1-p5-div-x24.c",
     "src/f32-spmm/gen/4x1-minmax-neonfma-pipelined.c",
     "src/f32-spmm/gen/4x1-minmax-neonfma-unroll2.c",
     "src/f32-spmm/gen/4x1-minmax-neonfma.c",
@@ -1654,137 +1640,151 @@
     "src/f32-spmm/gen/8x1-minmax-neonfma.c",
     "src/f32-spmm/gen/8x2-minmax-neonfma.c",
     "src/f32-spmm/gen/8x4-minmax-neonfma.c",
+    "src/f32-spmm/gen/12x1-minmax-neonfma.c",
+    "src/f32-spmm/gen/12x2-minmax-neonfma.c",
+    "src/f32-spmm/gen/12x4-minmax-neonfma.c",
+    "src/f32-spmm/gen/16x1-minmax-neonfma-pipelined.c",
+    "src/f32-spmm/gen/16x1-minmax-neonfma-unroll2.c",
+    "src/f32-spmm/gen/16x1-minmax-neonfma.c",
+    "src/f32-spmm/gen/16x2-minmax-neonfma.c",
+    "src/f32-spmm/gen/16x4-minmax-neonfma.c",
+    "src/f32-vbinary/gen/vdiv-minmax-neon-x4.c",
+    "src/f32-vbinary/gen/vdiv-minmax-neon-x8.c",
+    "src/f32-vbinary/gen/vdivc-minmax-neon-x4.c",
+    "src/f32-vbinary/gen/vdivc-minmax-neon-x8.c",
+    "src/f32-vbinary/gen/vrdivc-minmax-neon-x4.c",
+    "src/f32-vbinary/gen/vrdivc-minmax-neon-x8.c",
     "src/f32-vsqrt/gen/neon-sqrt-x4.c",
     "src/f32-vsqrt/gen/neon-sqrt-x8.c",
-    "src/math/sigmoid-neonfma-rr1-lut2048-p1-div.c",
     "src/math/sigmoid-neonfma-rr1-lut64-p2-div.c",
+    "src/math/sigmoid-neonfma-rr1-lut2048-p1-div.c",
     "src/math/sigmoid-neonfma-rr1-p5-div.c",
-    "src/math/sigmoid-neonfma-rr2-lut2048-p1-div.c",
     "src/math/sigmoid-neonfma-rr2-lut64-p2-div.c",
+    "src/math/sigmoid-neonfma-rr2-lut2048-p1-div.c",
     "src/math/sigmoid-neonfma-rr2-p5-div.c",
 ]
 
 NEONV8_UKERNELS = [
-    "src/f32-vrnd/gen/vrndne-neonv8-x4.c",
-    "src/f32-vrnd/gen/vrndne-neonv8-x8.c",
-    "src/f32-vrnd/gen/vrndz-neonv8-x4.c",
-    "src/f32-vrnd/gen/vrndz-neonv8-x8.c",
-    "src/f32-vrnd/gen/vrndu-neonv8-x4.c",
-    "src/f32-vrnd/gen/vrndu-neonv8-x8.c",
     "src/f32-vrnd/gen/vrndd-neonv8-x4.c",
     "src/f32-vrnd/gen/vrndd-neonv8-x8.c",
-    "src/math/roundne-neonv8.c",
+    "src/f32-vrnd/gen/vrndne-neonv8-x4.c",
+    "src/f32-vrnd/gen/vrndne-neonv8-x8.c",
+    "src/f32-vrnd/gen/vrndu-neonv8-x4.c",
+    "src/f32-vrnd/gen/vrndu-neonv8-x8.c",
+    "src/f32-vrnd/gen/vrndz-neonv8-x4.c",
+    "src/f32-vrnd/gen/vrndz-neonv8-x8.c",
     "src/math/roundd-neonv8.c",
+    "src/math/roundne-neonv8.c",
     "src/math/roundu-neonv8.c",
     "src/math/roundz-neonv8.c",
 ]
 
 AARCH64_NEONFP16ARITH_UKERNELS = [
-    "src/f16-clamp/gen/neonfp16arith-x16.c",
     "src/f16-clamp/gen/neonfp16arith-x8.c",
-    "src/f16-dwconv/gen/up16x25-minmax-neonfp16arith-acc2.c",
-    "src/f16-dwconv/gen/up16x25-minmax-neonfp16arith.c",
-    "src/f16-dwconv/gen/up16x4-minmax-neonfp16arith-acc2.c",
-    "src/f16-dwconv/gen/up16x4-minmax-neonfp16arith.c",
-    "src/f16-dwconv/gen/up16x9-minmax-neonfp16arith-acc2.c",
-    "src/f16-dwconv/gen/up16x9-minmax-neonfp16arith.c",
-    "src/f16-dwconv/gen/up8x25-minmax-neonfp16arith-acc2.c",
-    "src/f16-dwconv/gen/up8x25-minmax-neonfp16arith.c",
+    "src/f16-clamp/gen/neonfp16arith-x16.c",
     "src/f16-dwconv/gen/up8x4-minmax-neonfp16arith-acc2.c",
     "src/f16-dwconv/gen/up8x4-minmax-neonfp16arith.c",
     "src/f16-dwconv/gen/up8x9-minmax-neonfp16arith-acc2.c",
     "src/f16-dwconv/gen/up8x9-minmax-neonfp16arith.c",
+    "src/f16-dwconv/gen/up8x25-minmax-neonfp16arith-acc2.c",
+    "src/f16-dwconv/gen/up8x25-minmax-neonfp16arith.c",
+    "src/f16-dwconv/gen/up16x4-minmax-neonfp16arith-acc2.c",
+    "src/f16-dwconv/gen/up16x4-minmax-neonfp16arith.c",
+    "src/f16-dwconv/gen/up16x9-minmax-neonfp16arith-acc2.c",
+    "src/f16-dwconv/gen/up16x9-minmax-neonfp16arith.c",
+    "src/f16-dwconv/gen/up16x25-minmax-neonfp16arith-acc2.c",
+    "src/f16-dwconv/gen/up16x25-minmax-neonfp16arith.c",
     "src/f16-gavgpool/7p7x-minmax-neonfp16arith-c8.c",
     "src/f16-gavgpool/7x-minmax-neonfp16arith-c8.c",
-    "src/f16-hswish/gen/hswish-neonfp16arith-x16.c",
+    "src/f16-gemm/gen-inc/1x8inc-minmax-neonfp16arith-ld64.c",
+    "src/f16-gemm/gen-inc/1x16inc-minmax-neonfp16arith-ld64.c",
+    "src/f16-gemm/gen-inc/4x8inc-minmax-neonfp16arith-ld64.c",
+    "src/f16-gemm/gen-inc/4x16inc-minmax-neonfp16arith-ld64.c",
+    "src/f16-gemm/gen-inc/6x8inc-minmax-neonfp16arith-ld64.c",
+    "src/f16-gemm/gen-inc/6x16inc-minmax-neonfp16arith-ld64.c",
+    "src/f16-gemm/gen-inc/8x8inc-minmax-neonfp16arith-ld64.c",
+    "src/f16-gemm/gen-inc/8x16inc-minmax-neonfp16arith-ld64.c",
+    "src/f16-gemm/gen/1x8-minmax-neonfp16arith-ld64.c",
+    "src/f16-gemm/gen/1x16-minmax-neonfp16arith-ld64.c",
+    "src/f16-gemm/gen/4x8-minmax-neonfp16arith-ld64.c",
+    "src/f16-gemm/gen/4x16-minmax-neonfp16arith-ld64.c",
+    "src/f16-gemm/gen/6x8-minmax-neonfp16arith-ld64.c",
+    "src/f16-gemm/gen/6x16-minmax-neonfp16arith-ld64.c",
+    "src/f16-gemm/gen/8x8-minmax-neonfp16arith-ld64.c",
+    "src/f16-gemm/gen/8x16-minmax-neonfp16arith-ld64.c",
     "src/f16-hswish/gen/hswish-neonfp16arith-x8.c",
+    "src/f16-hswish/gen/hswish-neonfp16arith-x16.c",
+    "src/f16-igemm/gen/1x8-minmax-neonfp16arith-ld64.c",
+    "src/f16-igemm/gen/1x16-minmax-neonfp16arith-ld64.c",
+    "src/f16-igemm/gen/4x8-minmax-neonfp16arith-ld64.c",
+    "src/f16-igemm/gen/4x16-minmax-neonfp16arith-ld64.c",
+    "src/f16-igemm/gen/6x8-minmax-neonfp16arith-ld64.c",
+    "src/f16-igemm/gen/6x16-minmax-neonfp16arith-ld64.c",
+    "src/f16-igemm/gen/8x8-minmax-neonfp16arith-ld64.c",
+    "src/f16-igemm/gen/8x16-minmax-neonfp16arith-ld64.c",
     "src/f16-prelu/gen/neonfp16arith-2x8.c",
     "src/f16-prelu/gen/neonfp16arith-2x16.c",
-    "src/f16-relu/gen/neonfp16arith-x16.c",
     "src/f16-relu/gen/neonfp16arith-x8.c",
-    "src/f16-vbinary/gen/vaddc-minmax-neonfp16arith-x16.c",
-    "src/f16-vbinary/gen/vaddc-minmax-neonfp16arith-x8.c",
-    "src/f16-vbinary/gen/vadd-minmax-neonfp16arith-x16.c",
-    "src/f16-vbinary/gen/vadd-minmax-neonfp16arith-x8.c",
-    "src/f16-vbinary/gen/vdivc-minmax-neonfp16arith-x16.c",
-    "src/f16-vbinary/gen/vdivc-minmax-neonfp16arith-x8.c",
-    "src/f16-vbinary/gen/vdiv-minmax-neonfp16arith-x16.c",
-    "src/f16-vbinary/gen/vdiv-minmax-neonfp16arith-x8.c",
-    "src/f16-vbinary/gen/vmaxc-neonfp16arith-x16.c",
-    "src/f16-vbinary/gen/vmaxc-neonfp16arith-x8.c",
-    "src/f16-vbinary/gen/vmax-neonfp16arith-x16.c",
-    "src/f16-vbinary/gen/vmax-neonfp16arith-x8.c",
-    "src/f16-vbinary/gen/vminc-neonfp16arith-x16.c",
-    "src/f16-vbinary/gen/vminc-neonfp16arith-x8.c",
-    "src/f16-vbinary/gen/vmin-neonfp16arith-x16.c",
-    "src/f16-vbinary/gen/vmin-neonfp16arith-x8.c",
-    "src/f16-vbinary/gen/vmulc-minmax-neonfp16arith-x16.c",
-    "src/f16-vbinary/gen/vmulc-minmax-neonfp16arith-x8.c",
-    "src/f16-vbinary/gen/vmul-minmax-neonfp16arith-x16.c",
-    "src/f16-vbinary/gen/vmul-minmax-neonfp16arith-x8.c",
-    "src/f16-vbinary/gen/vrdivc-minmax-neonfp16arith-x16.c",
-    "src/f16-vbinary/gen/vrdivc-minmax-neonfp16arith-x8.c",
-    "src/f16-vbinary/gen/vrsubc-minmax-neonfp16arith-x16.c",
-    "src/f16-vbinary/gen/vrsubc-minmax-neonfp16arith-x8.c",
-    "src/f16-vbinary/gen/vsubc-minmax-neonfp16arith-x16.c",
-    "src/f16-vbinary/gen/vsubc-minmax-neonfp16arith-x8.c",
-    "src/f16-vbinary/gen/vsub-minmax-neonfp16arith-x16.c",
-    "src/f16-vbinary/gen/vsub-minmax-neonfp16arith-x8.c",
-    "src/f16-vmulcaddc/gen/c16-minmax-neonfp16arith-2x.c",
-    "src/f16-vmulcaddc/gen/c8-minmax-neonfp16arith-2x.c",
-    "src/f16-gemm/gen/1x8-minmax-neonfp16arith-ld64.c",
-    "src/f16-gemm/gen/4x8-minmax-neonfp16arith-ld64.c",
-    "src/f16-gemm/gen/6x8-minmax-neonfp16arith-ld64.c",
-    "src/f16-gemm/gen/8x8-minmax-neonfp16arith-ld64.c",
-    "src/f16-gemm/gen-inc/1x8inc-minmax-neonfp16arith-ld64.c",
-    "src/f16-gemm/gen-inc/4x8inc-minmax-neonfp16arith-ld64.c",
-    "src/f16-gemm/gen-inc/6x8inc-minmax-neonfp16arith-ld64.c",
-    "src/f16-gemm/gen-inc/8x8inc-minmax-neonfp16arith-ld64.c",
-    "src/f16-igemm/gen/1x8-minmax-neonfp16arith-ld64.c",
-    "src/f16-igemm/gen/4x8-minmax-neonfp16arith-ld64.c",
-    "src/f16-igemm/gen/6x8-minmax-neonfp16arith-ld64.c",
-    "src/f16-igemm/gen/8x8-minmax-neonfp16arith-ld64.c",
-    "src/f16-gemm/gen/1x16-minmax-neonfp16arith-ld64.c",
-    "src/f16-gemm/gen/4x16-minmax-neonfp16arith-ld64.c",
-    "src/f16-gemm/gen/6x16-minmax-neonfp16arith-ld64.c",
-    "src/f16-gemm/gen/8x16-minmax-neonfp16arith-ld64.c",
-    "src/f16-gemm/gen-inc/1x16inc-minmax-neonfp16arith-ld64.c",
-    "src/f16-gemm/gen-inc/4x16inc-minmax-neonfp16arith-ld64.c",
-    "src/f16-gemm/gen-inc/6x16inc-minmax-neonfp16arith-ld64.c",
-    "src/f16-gemm/gen-inc/8x16inc-minmax-neonfp16arith-ld64.c",
-    "src/f16-igemm/gen/1x16-minmax-neonfp16arith-ld64.c",
-    "src/f16-igemm/gen/4x16-minmax-neonfp16arith-ld64.c",
-    "src/f16-igemm/gen/6x16-minmax-neonfp16arith-ld64.c",
-    "src/f16-igemm/gen/8x16-minmax-neonfp16arith-ld64.c",
-    "src/f16-spmm/gen/8x1-minmax-neonfp16arith.c",
+    "src/f16-relu/gen/neonfp16arith-x16.c",
     "src/f16-spmm/gen/8x1-minmax-neonfp16arith-unroll2.c",
-    "src/f16-spmm/gen/16x1-minmax-neonfp16arith.c",
+    "src/f16-spmm/gen/8x1-minmax-neonfp16arith.c",
     "src/f16-spmm/gen/16x1-minmax-neonfp16arith-unroll2.c",
-    "src/f16-spmm/gen/24x1-minmax-neonfp16arith.c",
+    "src/f16-spmm/gen/16x1-minmax-neonfp16arith.c",
     "src/f16-spmm/gen/24x1-minmax-neonfp16arith-unroll2.c",
-    "src/f16-spmm/gen/32x1-minmax-neonfp16arith.c",
+    "src/f16-spmm/gen/24x1-minmax-neonfp16arith.c",
     "src/f16-spmm/gen/32x1-minmax-neonfp16arith-unroll2.c",
+    "src/f16-spmm/gen/32x1-minmax-neonfp16arith.c",
+    "src/f16-vbinary/gen/vadd-minmax-neonfp16arith-x8.c",
+    "src/f16-vbinary/gen/vadd-minmax-neonfp16arith-x16.c",
+    "src/f16-vbinary/gen/vaddc-minmax-neonfp16arith-x8.c",
+    "src/f16-vbinary/gen/vaddc-minmax-neonfp16arith-x16.c",
+    "src/f16-vbinary/gen/vdiv-minmax-neonfp16arith-x8.c",
+    "src/f16-vbinary/gen/vdiv-minmax-neonfp16arith-x16.c",
+    "src/f16-vbinary/gen/vdivc-minmax-neonfp16arith-x8.c",
+    "src/f16-vbinary/gen/vdivc-minmax-neonfp16arith-x16.c",
+    "src/f16-vbinary/gen/vmax-neonfp16arith-x8.c",
+    "src/f16-vbinary/gen/vmax-neonfp16arith-x16.c",
+    "src/f16-vbinary/gen/vmaxc-neonfp16arith-x8.c",
+    "src/f16-vbinary/gen/vmaxc-neonfp16arith-x16.c",
+    "src/f16-vbinary/gen/vmin-neonfp16arith-x8.c",
+    "src/f16-vbinary/gen/vmin-neonfp16arith-x16.c",
+    "src/f16-vbinary/gen/vminc-neonfp16arith-x8.c",
+    "src/f16-vbinary/gen/vminc-neonfp16arith-x16.c",
+    "src/f16-vbinary/gen/vmul-minmax-neonfp16arith-x8.c",
+    "src/f16-vbinary/gen/vmul-minmax-neonfp16arith-x16.c",
+    "src/f16-vbinary/gen/vmulc-minmax-neonfp16arith-x8.c",
+    "src/f16-vbinary/gen/vmulc-minmax-neonfp16arith-x16.c",
+    "src/f16-vbinary/gen/vrdivc-minmax-neonfp16arith-x8.c",
+    "src/f16-vbinary/gen/vrdivc-minmax-neonfp16arith-x16.c",
+    "src/f16-vbinary/gen/vrsubc-minmax-neonfp16arith-x8.c",
+    "src/f16-vbinary/gen/vrsubc-minmax-neonfp16arith-x16.c",
+    "src/f16-vbinary/gen/vsub-minmax-neonfp16arith-x8.c",
+    "src/f16-vbinary/gen/vsub-minmax-neonfp16arith-x16.c",
+    "src/f16-vbinary/gen/vsubc-minmax-neonfp16arith-x8.c",
+    "src/f16-vbinary/gen/vsubc-minmax-neonfp16arith-x16.c",
+    "src/f16-vmulcaddc/gen/c8-minmax-neonfp16arith-2x.c",
+    "src/f16-vmulcaddc/gen/c16-minmax-neonfp16arith-2x.c",
 ]
 
 NEONDOT_UKERNELS = [
     "src/qs8-gemm/gen/1x8c4-minmax-neondot.c",
-    "src/qs8-gemm/gen/4x8c4-minmax-neondot.c",
-    "src/qs8-gemm/gen/6x8c4-minmax-neondot.c",
-    "src/qs8-gemm/gen/8x8c4-minmax-neondot.c",
-    "src/qs8-gemm/gen/12x8c4-minmax-neondot.c",
     "src/qs8-gemm/gen/1x16c4-minmax-neondot.c",
+    "src/qs8-gemm/gen/4x8c4-minmax-neondot.c",
     "src/qs8-gemm/gen/4x16c4-minmax-neondot.c",
+    "src/qs8-gemm/gen/6x8c4-minmax-neondot.c",
     "src/qs8-gemm/gen/6x16c4-minmax-neondot.c",
+    "src/qs8-gemm/gen/8x8c4-minmax-neondot.c",
     "src/qs8-gemm/gen/8x16c4-minmax-neondot.c",
+    "src/qs8-gemm/gen/12x8c4-minmax-neondot.c",
     "src/qs8-igemm/gen/1x8c4-minmax-neondot.c",
-    "src/qs8-igemm/gen/4x8c4-minmax-neondot.c",
-    "src/qs8-igemm/gen/6x8c4-minmax-neondot.c",
-    "src/qs8-igemm/gen/8x8c4-minmax-neondot.c",
-    "src/qs8-igemm/gen/12x8c4-minmax-neondot.c",
     "src/qs8-igemm/gen/1x16c4-minmax-neondot.c",
+    "src/qs8-igemm/gen/4x8c4-minmax-neondot.c",
     "src/qs8-igemm/gen/4x16c4-minmax-neondot.c",
+    "src/qs8-igemm/gen/6x8c4-minmax-neondot.c",
     "src/qs8-igemm/gen/6x16c4-minmax-neondot.c",
+    "src/qs8-igemm/gen/8x8c4-minmax-neondot.c",
     "src/qs8-igemm/gen/8x16c4-minmax-neondot.c",
+    "src/qs8-igemm/gen/12x8c4-minmax-neondot.c",
 ]
 
 SSE_UKERNELS = [
@@ -1796,21 +1796,27 @@
     "src/f32-conv-hwc2chw/3x3s2p1c3x4-sse-2x2.c",
     "src/f32-dwconv-chw/3x3p1-sse-1x4-acc3.c",
     "src/f32-dwconv-chw/3x3s2p1-sse-1x4-acc3.c",
-    "src/f32-dwconv/gen/up4x25-minmax-sse-acc2.c",
-    "src/f32-dwconv/gen/up4x25-minmax-sse.c",
     "src/f32-dwconv/gen/up4x4-minmax-sse-acc2.c",
     "src/f32-dwconv/gen/up4x4-minmax-sse.c",
     "src/f32-dwconv/gen/up4x9-minmax-sse-acc2.c",
     "src/f32-dwconv/gen/up4x9-minmax-sse.c",
-    "src/f32-dwconv/gen/up8x25-minmax-sse-acc2.c",
-    "src/f32-dwconv/gen/up8x25-minmax-sse.c",
+    "src/f32-dwconv/gen/up4x25-minmax-sse-acc2.c",
+    "src/f32-dwconv/gen/up4x25-minmax-sse.c",
     "src/f32-dwconv/gen/up8x4-minmax-sse-acc2.c",
     "src/f32-dwconv/gen/up8x4-minmax-sse.c",
     "src/f32-dwconv/gen/up8x9-minmax-sse-acc2.c",
     "src/f32-dwconv/gen/up8x9-minmax-sse.c",
+    "src/f32-dwconv/gen/up8x25-minmax-sse-acc2.c",
+    "src/f32-dwconv/gen/up8x25-minmax-sse.c",
     "src/f32-gavgpool-cw/sse-x4.c",
     "src/f32-gavgpool/7p7x-minmax-sse-c4.c",
     "src/f32-gavgpool/7x-minmax-sse-c4.c",
+    "src/f32-gemm/gen-inc/1x8inc-minmax-sse-dup.c",
+    "src/f32-gemm/gen-inc/1x8inc-minmax-sse-load1.c",
+    "src/f32-gemm/gen-inc/1x8s4inc-minmax-sse.c",
+    "src/f32-gemm/gen-inc/4x8inc-minmax-sse-dup.c",
+    "src/f32-gemm/gen-inc/4x8inc-minmax-sse-load1.c",
+    "src/f32-gemm/gen-inc/4x8s4inc-minmax-sse.c",
     "src/f32-gemm/gen/1x8-minmax-sse-dup.c",
     "src/f32-gemm/gen/1x8-minmax-sse-load1.c",
     "src/f32-gemm/gen/1x8s4-minmax-sse.c",
@@ -1818,12 +1824,6 @@
     "src/f32-gemm/gen/4x8-minmax-sse-dup.c",
     "src/f32-gemm/gen/4x8-minmax-sse-load1.c",
     "src/f32-gemm/gen/4x8s4-minmax-sse.c",
-    "src/f32-gemm/gen-inc/1x8inc-minmax-sse-dup.c",
-    "src/f32-gemm/gen-inc/1x8inc-minmax-sse-load1.c",
-    "src/f32-gemm/gen-inc/1x8s4inc-minmax-sse.c",
-    "src/f32-gemm/gen-inc/4x8inc-minmax-sse-dup.c",
-    "src/f32-gemm/gen-inc/4x8inc-minmax-sse-load1.c",
-    "src/f32-gemm/gen-inc/4x8s4inc-minmax-sse.c",
     "src/f32-hswish/gen/hswish-sse-x4.c",
     "src/f32-hswish/gen/hswish-sse-x8.c",
     "src/f32-ibilinear/gen/sse-c4.c",
@@ -1891,36 +1891,36 @@
     "src/f32-vunary/gen/vneg-sse-x8.c",
     "src/f32-vunary/gen/vsqr-sse-x4.c",
     "src/f32-vunary/gen/vsqr-sse-x8.c",
-    "src/x32-fill/sse.c",
-    "src/x32-packx/x4-sse.c",
-    "src/x32-pad/sse.c",
-    "src/math/roundne-sse-addsub.c",
     "src/math/roundd-sse-addsub.c",
+    "src/math/roundne-sse-addsub.c",
     "src/math/roundu-sse-addsub.c",
     "src/math/roundz-sse-addsub.c",
     "src/math/sqrt-sse-hh1mac.c",
     "src/math/sqrt-sse-nr1mac.c",
     "src/math/sqrt-sse-nr2mac.c",
+    "src/x32-fill/sse.c",
+    "src/x32-packx/x4-sse.c",
+    "src/x32-pad/sse.c",
 ]
 
 SSE2_UKERNELS = [
-    "src/f32-argmaxpool/9p8x-sse2-c4.c",
     "src/f32-argmaxpool/4x-sse2-c4.c",
+    "src/f32-argmaxpool/9p8x-sse2-c4.c",
     "src/f32-argmaxpool/9x-sse2-c4.c",
     "src/f32-prelu/gen/sse2-2x4.c",
     "src/f32-prelu/gen/sse2-2x8.c",
     "src/f32-raddstoreexpminusmax/gen/sse2-p5-x4.c",
-    "src/f32-raddstoreexpminusmax/gen/sse2-p5-x8.c",
     "src/f32-raddstoreexpminusmax/gen/sse2-p5-x8-acc2.c",
-    "src/f32-raddstoreexpminusmax/gen/sse2-p5-x12.c",
+    "src/f32-raddstoreexpminusmax/gen/sse2-p5-x8.c",
     "src/f32-raddstoreexpminusmax/gen/sse2-p5-x12-acc2.c",
     "src/f32-raddstoreexpminusmax/gen/sse2-p5-x12-acc3.c",
-    "src/f32-raddstoreexpminusmax/gen/sse2-p5-x16.c",
+    "src/f32-raddstoreexpminusmax/gen/sse2-p5-x12.c",
     "src/f32-raddstoreexpminusmax/gen/sse2-p5-x16-acc2.c",
     "src/f32-raddstoreexpminusmax/gen/sse2-p5-x16-acc4.c",
-    "src/f32-raddstoreexpminusmax/gen/sse2-p5-x20.c",
+    "src/f32-raddstoreexpminusmax/gen/sse2-p5-x16.c",
     "src/f32-raddstoreexpminusmax/gen/sse2-p5-x20-acc2.c",
     "src/f32-raddstoreexpminusmax/gen/sse2-p5-x20-acc5.c",
+    "src/f32-raddstoreexpminusmax/gen/sse2-p5-x20.c",
     "src/f32-sigmoid/gen/sse2-lut64-p2-div-x4.c",
     "src/f32-sigmoid/gen/sse2-lut64-p2-div-x8.c",
     "src/f32-sigmoid/gen/sse2-lut64-p2-div-x12.c",
@@ -1935,50 +1935,63 @@
     "src/f32-sigmoid/gen/sse2-p5-div-x24.c",
     "src/f32-vlrelu/gen/vlrelu-sse2-x4.c",
     "src/f32-vlrelu/gen/vlrelu-sse2-x8.c",
-    "src/f32-vrnd/gen/vrndne-sse2-x4.c",
-    "src/f32-vrnd/gen/vrndne-sse2-x8.c",
-    "src/f32-vrnd/gen/vrndz-sse2-x4.c",
-    "src/f32-vrnd/gen/vrndz-sse2-x8.c",
-    "src/f32-vrnd/gen/vrndu-sse2-x4.c",
-    "src/f32-vrnd/gen/vrndu-sse2-x8.c",
     "src/f32-vrnd/gen/vrndd-sse2-x4.c",
     "src/f32-vrnd/gen/vrndd-sse2-x8.c",
+    "src/f32-vrnd/gen/vrndne-sse2-x4.c",
+    "src/f32-vrnd/gen/vrndne-sse2-x8.c",
+    "src/f32-vrnd/gen/vrndu-sse2-x4.c",
+    "src/f32-vrnd/gen/vrndu-sse2-x8.c",
+    "src/f32-vrnd/gen/vrndz-sse2-x4.c",
+    "src/f32-vrnd/gen/vrndz-sse2-x8.c",
+    "src/math/exp-sse2-lut64-p2.c",
+    "src/math/exp-sse2-p5.c",
+    "src/math/expminus-sse2-p5.c",
+    "src/math/roundd-sse2-cvt.c",
+    "src/math/roundne-sse2-cvt.c",
+    "src/math/roundu-sse2-cvt.c",
+    "src/math/roundz-sse2-cvt.c",
+    "src/math/sigmoid-sse2-rr2-lut64-p2-div.c",
+    "src/math/sigmoid-sse2-rr2-lut64-p2-nr1.c",
+    "src/math/sigmoid-sse2-rr2-lut64-p2-nr2.c",
+    "src/math/sigmoid-sse2-rr2-p5-div.c",
+    "src/math/sigmoid-sse2-rr2-p5-nr1.c",
+    "src/math/sigmoid-sse2-rr2-p5-nr2.c",
     "src/qs8-dwconv/gen/up8x9-minmax-sse2-mul16.c",
     "src/qs8-dwconv/gen/up16x9-minmax-sse2-mul16.c",
     "src/qs8-dwconv/gen/up24x9-minmax-sse2-mul16.c",
-    "src/qs8-gavgpool/gen/7x-minmax-sse2-c8-acc2.c",
-    "src/qs8-gavgpool/gen/7x-minmax-sse2-c16-acc2.c",
-    "src/qs8-gavgpool/gen/7x-minmax-sse2-c24-acc2.c",
     "src/qs8-gavgpool/gen/7p7x-minmax-sse2-c8-acc2.c",
     "src/qs8-gavgpool/gen/7p7x-minmax-sse2-c16-acc2.c",
     "src/qs8-gavgpool/gen/7p7x-minmax-sse2-c24-acc2.c",
+    "src/qs8-gavgpool/gen/7x-minmax-sse2-c8-acc2.c",
+    "src/qs8-gavgpool/gen/7x-minmax-sse2-c16-acc2.c",
+    "src/qs8-gavgpool/gen/7x-minmax-sse2-c24-acc2.c",
     "src/qs8-gemm/gen/1x4c2-minmax-sse2-ld64.c",
-    "src/qs8-gemm/gen/4x4c2-minmax-sse2-ld64.c",
     "src/qs8-gemm/gen/1x4c2-minmax-sse2-ld128.c",
-    "src/qs8-gemm/gen/4x4c2-minmax-sse2-ld128.c",
     "src/qs8-gemm/gen/1x4c2-xw-minmax-sse2.c",
-    "src/qs8-gemm/gen/4x4c2-xw-minmax-sse2.c",
     "src/qs8-gemm/gen/1x4c8-minmax-sse2-ld64.c",
-    "src/qs8-gemm/gen/2x4c8-minmax-sse2-ld64.c",
-    "src/qs8-gemm/gen/3x4c8-minmax-sse2-ld64.c",
     "src/qs8-gemm/gen/1x4c8-minmax-sse2-ld128.c",
-    "src/qs8-gemm/gen/2x4c8-minmax-sse2-ld128.c",
-    "src/qs8-gemm/gen/3x4c8-minmax-sse2-ld128.c",
     "src/qs8-gemm/gen/1x4c8-xw-minmax-sse2.c",
+    "src/qs8-gemm/gen/2x4c8-minmax-sse2-ld64.c",
+    "src/qs8-gemm/gen/2x4c8-minmax-sse2-ld128.c",
     "src/qs8-gemm/gen/2x4c8-xw-minmax-sse2.c",
+    "src/qs8-gemm/gen/3x4c8-minmax-sse2-ld64.c",
+    "src/qs8-gemm/gen/3x4c8-minmax-sse2-ld128.c",
     "src/qs8-gemm/gen/3x4c8-xw-minmax-sse2.c",
+    "src/qs8-gemm/gen/4x4c2-minmax-sse2-ld64.c",
+    "src/qs8-gemm/gen/4x4c2-minmax-sse2-ld128.c",
+    "src/qs8-gemm/gen/4x4c2-xw-minmax-sse2.c",
     "src/qs8-igemm/gen/1x4c2-minmax-sse2-ld64.c",
-    "src/qs8-igemm/gen/4x4c2-minmax-sse2-ld64.c",
     "src/qs8-igemm/gen/1x4c2-minmax-sse2-ld128.c",
-    "src/qs8-igemm/gen/4x4c2-minmax-sse2-ld128.c",
     "src/qs8-igemm/gen/1x4c8-minmax-sse2-ld64.c",
-    "src/qs8-igemm/gen/2x4c8-minmax-sse2-ld64.c",
-    "src/qs8-igemm/gen/3x4c8-minmax-sse2-ld64.c",
     "src/qs8-igemm/gen/1x4c8-minmax-sse2-ld128.c",
+    "src/qs8-igemm/gen/2x4c8-minmax-sse2-ld64.c",
     "src/qs8-igemm/gen/2x4c8-minmax-sse2-ld128.c",
+    "src/qs8-igemm/gen/3x4c8-minmax-sse2-ld64.c",
     "src/qs8-igemm/gen/3x4c8-minmax-sse2-ld128.c",
-    "src/qs8-requantization/precise-sse2.c",
+    "src/qs8-igemm/gen/4x4c2-minmax-sse2-ld64.c",
+    "src/qs8-igemm/gen/4x4c2-minmax-sse2-ld128.c",
     "src/qs8-requantization/fp32-sse2.c",
+    "src/qs8-requantization/precise-sse2.c",
     "src/qs8-requantization/q31-sse2.c",
     "src/qs8-vadd/gen/minmax-sse2-mul16-ld64-x8.c",
     "src/qs8-vadd/gen/minmax-sse2-mul16-ld64-x16.c",
@@ -1990,78 +2003,65 @@
     "src/qs8-vaddc/gen/minmax-sse2-mul16-ld64-x32.c",
     "src/qu8-avgpool/9p8x-minmax-sse2-c8.c",
     "src/qu8-avgpool/9x-minmax-sse2-c8.c",
-    "src/qu8-igemm/4x4c2-minmax-sse2.c",
     "src/qu8-dwconv/up8x9-minmax-sse2.c",
     "src/qu8-gavgpool/7p7x-minmax-sse2-c8.c",
     "src/qu8-gavgpool/7x-minmax-sse2-c8.c",
     "src/qu8-gemm/2x4c8-minmax-sse2.c",
     "src/qu8-gemm/4x4c2-minmax-sse2.c",
-    "src/qu8-requantization/precise-sse2.c",
+    "src/qu8-igemm/4x4c2-minmax-sse2.c",
     "src/qu8-requantization/fp32-sse2.c",
+    "src/qu8-requantization/precise-sse2.c",
     "src/qu8-requantization/q31-sse2.c",
     "src/qu8-vadd/minmax-sse2.c",
     "src/u8-clamp/sse2-x64.c",
     "src/u8-maxpool/9p8x-minmax-sse2-c16.c",
     "src/u8-rmax/sse2.c",
+    "src/x8-zip/x2-sse2.c",
+    "src/x8-zip/x3-sse2.c",
+    "src/x8-zip/x4-sse2.c",
+    "src/x8-zip/xm-sse2.c",
     "src/x32-unpool/sse2.c",
     "src/x32-zip/x2-sse2.c",
     "src/x32-zip/x3-sse2.c",
     "src/x32-zip/x4-sse2.c",
     "src/x32-zip/xm-sse2.c",
-    "src/x8-zip/x2-sse2.c",
-    "src/x8-zip/x3-sse2.c",
-    "src/x8-zip/x4-sse2.c",
-    "src/x8-zip/xm-sse2.c",
-    "src/math/exp-sse2-p5.c",
-    "src/math/exp-sse2-lut64-p2.c",
-    "src/math/expminus-sse2-p5.c",
-    "src/math/roundne-sse2-cvt.c",
-    "src/math/roundd-sse2-cvt.c",
-    "src/math/roundu-sse2-cvt.c",
-    "src/math/roundz-sse2-cvt.c",
-    "src/math/sigmoid-sse2-rr2-p5-div.c",
-    "src/math/sigmoid-sse2-rr2-p5-nr1.c",
-    "src/math/sigmoid-sse2-rr2-p5-nr2.c",
-    "src/math/sigmoid-sse2-rr2-lut64-p2-div.c",
-    "src/math/sigmoid-sse2-rr2-lut64-p2-nr1.c",
-    "src/math/sigmoid-sse2-rr2-lut64-p2-nr2.c",
 ]
 
 SSSE3_UKERNELS = [
     "src/qs8-dwconv/gen/up8x9-minmax-ssse3-mul16.c",
     "src/qs8-dwconv/gen/up16x9-minmax-ssse3-mul16.c",
     "src/qs8-dwconv/gen/up24x9-minmax-ssse3-mul16.c",
-    "src/qs8-gavgpool/gen/7x-minmax-ssse3-c8-acc2.c",
-    "src/qs8-gavgpool/gen/7x-minmax-ssse3-c16-acc2.c",
-    "src/qs8-gavgpool/gen/7x-minmax-ssse3-c24-acc2.c",
     "src/qs8-gavgpool/gen/7p7x-minmax-ssse3-c8-acc2.c",
     "src/qs8-gavgpool/gen/7p7x-minmax-ssse3-c16-acc2.c",
     "src/qs8-gavgpool/gen/7p7x-minmax-ssse3-c24-acc2.c",
+    "src/qs8-gavgpool/gen/7x-minmax-ssse3-c8-acc2.c",
+    "src/qs8-gavgpool/gen/7x-minmax-ssse3-c16-acc2.c",
+    "src/qs8-gavgpool/gen/7x-minmax-ssse3-c24-acc2.c",
     "src/qs8-gemm/gen/1x4c2-minmax-ssse3-ld64.c",
-    "src/qs8-gemm/gen/4x4c2-minmax-ssse3-ld64.c",
     "src/qs8-gemm/gen/1x4c2-minmax-ssse3-ld128.c",
-    "src/qs8-gemm/gen/4x4c2-minmax-ssse3-ld128.c",
     "src/qs8-gemm/gen/1x4c2-xw-minmax-ssse3.c",
-    "src/qs8-gemm/gen/4x4c2-xw-minmax-ssse3.c",
     "src/qs8-gemm/gen/1x4c8-minmax-ssse3-ld64.c",
-    "src/qs8-gemm/gen/2x4c8-minmax-ssse3-ld64.c",
-    "src/qs8-gemm/gen/3x4c8-minmax-ssse3-ld64.c",
     "src/qs8-gemm/gen/1x4c8-minmax-ssse3-ld128.c",
-    "src/qs8-gemm/gen/2x4c8-minmax-ssse3-ld128.c",
-    "src/qs8-gemm/gen/3x4c8-minmax-ssse3-ld128.c",
     "src/qs8-gemm/gen/1x4c8-xw-minmax-ssse3.c",
+    "src/qs8-gemm/gen/2x4c8-minmax-ssse3-ld64.c",
+    "src/qs8-gemm/gen/2x4c8-minmax-ssse3-ld128.c",
     "src/qs8-gemm/gen/2x4c8-xw-minmax-ssse3.c",
+    "src/qs8-gemm/gen/3x4c8-minmax-ssse3-ld64.c",
+    "src/qs8-gemm/gen/3x4c8-minmax-ssse3-ld128.c",
     "src/qs8-gemm/gen/3x4c8-xw-minmax-ssse3.c",
+    "src/qs8-gemm/gen/4x4c2-minmax-ssse3-ld64.c",
+    "src/qs8-gemm/gen/4x4c2-minmax-ssse3-ld128.c",
+    "src/qs8-gemm/gen/4x4c2-xw-minmax-ssse3.c",
     "src/qs8-igemm/gen/1x4c2-minmax-ssse3-ld64.c",
-    "src/qs8-igemm/gen/4x4c2-minmax-ssse3-ld64.c",
     "src/qs8-igemm/gen/1x4c2-minmax-ssse3-ld128.c",
-    "src/qs8-igemm/gen/4x4c2-minmax-ssse3-ld128.c",
     "src/qs8-igemm/gen/1x4c8-minmax-ssse3-ld64.c",
-    "src/qs8-igemm/gen/2x4c8-minmax-ssse3-ld64.c",
-    "src/qs8-igemm/gen/3x4c8-minmax-ssse3-ld64.c",
     "src/qs8-igemm/gen/1x4c8-minmax-ssse3-ld128.c",
+    "src/qs8-igemm/gen/2x4c8-minmax-ssse3-ld64.c",
     "src/qs8-igemm/gen/2x4c8-minmax-ssse3-ld128.c",
+    "src/qs8-igemm/gen/3x4c8-minmax-ssse3-ld64.c",
     "src/qs8-igemm/gen/3x4c8-minmax-ssse3-ld128.c",
+    "src/qs8-igemm/gen/4x4c2-minmax-ssse3-ld64.c",
+    "src/qs8-igemm/gen/4x4c2-minmax-ssse3-ld128.c",
     "src/qs8-requantization/precise-ssse3.c",
     "src/qs8-requantization/q31-ssse3.c",
     "src/qu8-requantization/precise-ssse3.c",
@@ -2085,48 +2085,52 @@
     "src/f32-sigmoid/gen/sse41-p5-div-x24.c",
     "src/f32-vlrelu/gen/vlrelu-sse41-x4.c",
     "src/f32-vlrelu/gen/vlrelu-sse41-x8.c",
-    "src/f32-vrnd/gen/vrndne-sse41-x4.c",
-    "src/f32-vrnd/gen/vrndne-sse41-x8.c",
-    "src/f32-vrnd/gen/vrndz-sse41-x4.c",
-    "src/f32-vrnd/gen/vrndz-sse41-x8.c",
-    "src/f32-vrnd/gen/vrndu-sse41-x4.c",
-    "src/f32-vrnd/gen/vrndu-sse41-x8.c",
     "src/f32-vrnd/gen/vrndd-sse41-x4.c",
     "src/f32-vrnd/gen/vrndd-sse41-x8.c",
+    "src/f32-vrnd/gen/vrndne-sse41-x4.c",
+    "src/f32-vrnd/gen/vrndne-sse41-x8.c",
+    "src/f32-vrnd/gen/vrndu-sse41-x4.c",
+    "src/f32-vrnd/gen/vrndu-sse41-x8.c",
+    "src/f32-vrnd/gen/vrndz-sse41-x4.c",
+    "src/f32-vrnd/gen/vrndz-sse41-x8.c",
+    "src/math/roundd-sse41.c",
+    "src/math/roundne-sse41.c",
+    "src/math/roundu-sse41.c",
+    "src/math/roundz-sse41.c",
     "src/qs8-dwconv/gen/up8x9-minmax-sse41-mul16.c",
     "src/qs8-dwconv/gen/up16x9-minmax-sse41-mul16.c",
     "src/qs8-dwconv/gen/up24x9-minmax-sse41-mul16.c",
-    "src/qs8-gavgpool/gen/7x-minmax-sse41-c8-acc2.c",
-    "src/qs8-gavgpool/gen/7x-minmax-sse41-c16-acc2.c",
-    "src/qs8-gavgpool/gen/7x-minmax-sse41-c24-acc2.c",
     "src/qs8-gavgpool/gen/7p7x-minmax-sse41-c8-acc2.c",
     "src/qs8-gavgpool/gen/7p7x-minmax-sse41-c16-acc2.c",
     "src/qs8-gavgpool/gen/7p7x-minmax-sse41-c24-acc2.c",
+    "src/qs8-gavgpool/gen/7x-minmax-sse41-c8-acc2.c",
+    "src/qs8-gavgpool/gen/7x-minmax-sse41-c16-acc2.c",
+    "src/qs8-gavgpool/gen/7x-minmax-sse41-c24-acc2.c",
     "src/qs8-gemm/gen/1x4c2-minmax-sse41-ld64.c",
-    "src/qs8-gemm/gen/4x4c2-minmax-sse41-ld64.c",
     "src/qs8-gemm/gen/1x4c2-minmax-sse41-ld128.c",
-    "src/qs8-gemm/gen/4x4c2-minmax-sse41-ld128.c",
     "src/qs8-gemm/gen/1x4c2-xw-minmax-sse41.c",
-    "src/qs8-gemm/gen/4x4c2-xw-minmax-sse41.c",
     "src/qs8-gemm/gen/1x4c8-minmax-sse41-ld64.c",
-    "src/qs8-gemm/gen/2x4c8-minmax-sse41-ld64.c",
-    "src/qs8-gemm/gen/3x4c8-minmax-sse41-ld64.c",
     "src/qs8-gemm/gen/1x4c8-minmax-sse41-ld128.c",
-    "src/qs8-gemm/gen/2x4c8-minmax-sse41-ld128.c",
-    "src/qs8-gemm/gen/3x4c8-minmax-sse41-ld128.c",
     "src/qs8-gemm/gen/1x4c8-xw-minmax-sse41.c",
+    "src/qs8-gemm/gen/2x4c8-minmax-sse41-ld64.c",
+    "src/qs8-gemm/gen/2x4c8-minmax-sse41-ld128.c",
     "src/qs8-gemm/gen/2x4c8-xw-minmax-sse41.c",
+    "src/qs8-gemm/gen/3x4c8-minmax-sse41-ld64.c",
+    "src/qs8-gemm/gen/3x4c8-minmax-sse41-ld128.c",
     "src/qs8-gemm/gen/3x4c8-xw-minmax-sse41.c",
+    "src/qs8-gemm/gen/4x4c2-minmax-sse41-ld64.c",
+    "src/qs8-gemm/gen/4x4c2-minmax-sse41-ld128.c",
+    "src/qs8-gemm/gen/4x4c2-xw-minmax-sse41.c",
     "src/qs8-igemm/gen/1x4c2-minmax-sse41-ld64.c",
-    "src/qs8-igemm/gen/4x4c2-minmax-sse41-ld64.c",
     "src/qs8-igemm/gen/1x4c2-minmax-sse41-ld128.c",
-    "src/qs8-igemm/gen/4x4c2-minmax-sse41-ld128.c",
     "src/qs8-igemm/gen/1x4c8-minmax-sse41-ld64.c",
-    "src/qs8-igemm/gen/2x4c8-minmax-sse41-ld64.c",
-    "src/qs8-igemm/gen/3x4c8-minmax-sse41-ld64.c",
     "src/qs8-igemm/gen/1x4c8-minmax-sse41-ld128.c",
+    "src/qs8-igemm/gen/2x4c8-minmax-sse41-ld64.c",
     "src/qs8-igemm/gen/2x4c8-minmax-sse41-ld128.c",
+    "src/qs8-igemm/gen/3x4c8-minmax-sse41-ld64.c",
     "src/qs8-igemm/gen/3x4c8-minmax-sse41-ld128.c",
+    "src/qs8-igemm/gen/4x4c2-minmax-sse41-ld64.c",
+    "src/qs8-igemm/gen/4x4c2-minmax-sse41-ld128.c",
     "src/qs8-requantization/fp32-sse4.c",
     "src/qs8-requantization/precise-sse4.c",
     "src/qs8-requantization/q31-sse4.c",
@@ -2148,60 +2152,56 @@
     "src/qs8-vaddc/gen/minmax-sse41-mul32-ld32-x32.c",
     "src/qu8-requantization/precise-sse4.c",
     "src/qu8-requantization/q31-sse4.c",
-    "src/math/roundne-sse41.c",
-    "src/math/roundd-sse41.c",
-    "src/math/roundu-sse41.c",
-    "src/math/roundz-sse41.c",
 ]
 
 AVX_UKERNELS = [
     "src/f32-clamp/gen/avx-x8.c",
     "src/f32-clamp/gen/avx-x16.c",
-    "src/f32-dwconv/gen/up16x4-minmax-avx-acc2.c",
-    "src/f32-dwconv/gen/up16x4-minmax-avx.c",
     "src/f32-dwconv/gen/up8x4-minmax-avx-acc2.c",
     "src/f32-dwconv/gen/up8x4-minmax-avx.c",
-    "src/f32-dwconv/gen/up16x9-minmax-avx-acc2.c",
-    "src/f32-dwconv/gen/up16x9-minmax-avx.c",
     "src/f32-dwconv/gen/up8x9-minmax-avx-acc2.c",
     "src/f32-dwconv/gen/up8x9-minmax-avx.c",
-    "src/f32-dwconv/gen/up16x25-minmax-avx-acc2.c",
-    "src/f32-dwconv/gen/up16x25-minmax-avx.c",
     "src/f32-dwconv/gen/up8x25-minmax-avx-acc2.c",
     "src/f32-dwconv/gen/up8x25-minmax-avx.c",
-    "src/f32-gemm/gen/1x8-minmax-avx-broadcast.c",
-    "src/f32-gemm/gen/4x8-minmax-avx-broadcast.c",
-    "src/f32-gemm/gen/5x8-minmax-avx-broadcast.c",
-    "src/f32-gemm/gen/6x8-minmax-avx-broadcast.c",
-    "src/f32-gemm/gen/7x8-minmax-avx-broadcast.c",
-    "src/f32-gemm/gen/1x16-minmax-avx-broadcast.c",
-    "src/f32-gemm/gen/3x16-minmax-avx-broadcast.c",
-    "src/f32-gemm/gen/4x16-minmax-avx-broadcast.c",
-    "src/f32-gemm/gen/5x16-minmax-avx-broadcast.c",
+    "src/f32-dwconv/gen/up16x4-minmax-avx-acc2.c",
+    "src/f32-dwconv/gen/up16x4-minmax-avx.c",
+    "src/f32-dwconv/gen/up16x9-minmax-avx-acc2.c",
+    "src/f32-dwconv/gen/up16x9-minmax-avx.c",
+    "src/f32-dwconv/gen/up16x25-minmax-avx-acc2.c",
+    "src/f32-dwconv/gen/up16x25-minmax-avx.c",
     "src/f32-gemm/gen-inc/1x8inc-minmax-avx-broadcast.c",
-    "src/f32-gemm/gen-inc/4x8inc-minmax-avx-broadcast.c",
-    "src/f32-gemm/gen-inc/5x8inc-minmax-avx-broadcast.c",
-    "src/f32-gemm/gen-inc/6x8inc-minmax-avx-broadcast.c",
-    "src/f32-gemm/gen-inc/7x8inc-minmax-avx-broadcast.c",
     "src/f32-gemm/gen-inc/1x16inc-minmax-avx-broadcast.c",
     "src/f32-gemm/gen-inc/3x16inc-minmax-avx-broadcast.c",
+    "src/f32-gemm/gen-inc/4x8inc-minmax-avx-broadcast.c",
     "src/f32-gemm/gen-inc/4x16inc-minmax-avx-broadcast.c",
+    "src/f32-gemm/gen-inc/5x8inc-minmax-avx-broadcast.c",
     "src/f32-gemm/gen-inc/5x16inc-minmax-avx-broadcast.c",
+    "src/f32-gemm/gen-inc/6x8inc-minmax-avx-broadcast.c",
+    "src/f32-gemm/gen-inc/7x8inc-minmax-avx-broadcast.c",
+    "src/f32-gemm/gen/1x8-minmax-avx-broadcast.c",
+    "src/f32-gemm/gen/1x16-minmax-avx-broadcast.c",
+    "src/f32-gemm/gen/3x16-minmax-avx-broadcast.c",
+    "src/f32-gemm/gen/4x8-minmax-avx-broadcast.c",
+    "src/f32-gemm/gen/4x16-minmax-avx-broadcast.c",
+    "src/f32-gemm/gen/5x8-minmax-avx-broadcast.c",
+    "src/f32-gemm/gen/5x16-minmax-avx-broadcast.c",
+    "src/f32-gemm/gen/6x8-minmax-avx-broadcast.c",
+    "src/f32-gemm/gen/7x8-minmax-avx-broadcast.c",
     "src/f32-hswish/gen/hswish-avx-x8.c",
     "src/f32-hswish/gen/hswish-avx-x16.c",
     "src/f32-igemm/gen/1x8-minmax-avx-broadcast.c",
-    "src/f32-igemm/gen/4x8-minmax-avx-broadcast.c",
-    "src/f32-igemm/gen/5x8-minmax-avx-broadcast.c",
-    "src/f32-igemm/gen/6x8-minmax-avx-broadcast.c",
-    "src/f32-igemm/gen/7x8-minmax-avx-broadcast.c",
     "src/f32-igemm/gen/1x16-minmax-avx-broadcast.c",
     "src/f32-igemm/gen/3x16-minmax-avx-broadcast.c",
+    "src/f32-igemm/gen/4x8-minmax-avx-broadcast.c",
     "src/f32-igemm/gen/4x16-minmax-avx-broadcast.c",
+    "src/f32-igemm/gen/5x8-minmax-avx-broadcast.c",
     "src/f32-igemm/gen/5x16-minmax-avx-broadcast.c",
+    "src/f32-igemm/gen/6x8-minmax-avx-broadcast.c",
+    "src/f32-igemm/gen/7x8-minmax-avx-broadcast.c",
     "src/f32-prelu/gen/avx-2x8.c",
     "src/f32-prelu/gen/avx-2x16.c",
-    "src/f32-relu/gen/avx-x16.c",
     "src/f32-relu/gen/avx-x8.c",
+    "src/f32-relu/gen/avx-x16.c",
     "src/f32-rmax/avx.c",
     "src/f32-sigmoid/gen/avx-rr2-p5-div-x8.c",
     "src/f32-sigmoid/gen/avx-rr2-p5-div-x16.c",
@@ -2257,15 +2257,15 @@
     "src/f32-vbinary/gen/vsubc-minmax-avx-x16.c",
     "src/f32-vlrelu/gen/vlrelu-avx-x8.c",
     "src/f32-vlrelu/gen/vlrelu-avx-x16.c",
-    "src/f32-vscale/avx-unroll32.c",
-    "src/f32-vrnd/gen/vrndne-avx-x8.c",
-    "src/f32-vrnd/gen/vrndne-avx-x16.c",
-    "src/f32-vrnd/gen/vrndz-avx-x8.c",
-    "src/f32-vrnd/gen/vrndz-avx-x16.c",
-    "src/f32-vrnd/gen/vrndu-avx-x8.c",
-    "src/f32-vrnd/gen/vrndu-avx-x16.c",
     "src/f32-vrnd/gen/vrndd-avx-x8.c",
     "src/f32-vrnd/gen/vrndd-avx-x16.c",
+    "src/f32-vrnd/gen/vrndne-avx-x8.c",
+    "src/f32-vrnd/gen/vrndne-avx-x16.c",
+    "src/f32-vrnd/gen/vrndu-avx-x8.c",
+    "src/f32-vrnd/gen/vrndu-avx-x16.c",
+    "src/f32-vrnd/gen/vrndz-avx-x8.c",
+    "src/f32-vrnd/gen/vrndz-avx-x16.c",
+    "src/f32-vscale/avx-unroll32.c",
     "src/f32-vsqrt/gen/avx-sqrt-x8.c",
     "src/f32-vsqrt/gen/avx-sqrt-x16.c",
     "src/f32-vunary/gen/vabs-avx-x8.c",
@@ -2275,38 +2275,38 @@
     "src/f32-vunary/gen/vsqr-avx-x8.c",
     "src/f32-vunary/gen/vsqr-avx-x16.c",
     "src/math/exp-avx-p5.c",
+    "src/math/sigmoid-avx-rr2-lut64-p2-div.c",
     "src/math/sigmoid-avx-rr2-p5-div.c",
     "src/math/sigmoid-avx-rr2-p5-nr1.c",
     "src/math/sigmoid-avx-rr2-p5-nr2.c",
-    "src/math/sigmoid-avx-rr2-lut64-p2-div.c",
 ]
 
 XOP_UKERNELS = [
     "src/qs8-gemm/gen/1x4c2-minmax-xop-ld64.c",
-    "src/qs8-gemm/gen/4x4c2-minmax-xop-ld64.c",
     "src/qs8-gemm/gen/1x4c2-minmax-xop-ld128.c",
-    "src/qs8-gemm/gen/4x4c2-minmax-xop-ld128.c",
     "src/qs8-gemm/gen/1x4c2-xw-minmax-xop.c",
-    "src/qs8-gemm/gen/4x4c2-xw-minmax-xop.c",
     "src/qs8-gemm/gen/1x4c8-minmax-xop-ld64.c",
-    "src/qs8-gemm/gen/2x4c8-minmax-xop-ld64.c",
-    "src/qs8-gemm/gen/3x4c8-minmax-xop-ld64.c",
     "src/qs8-gemm/gen/1x4c8-minmax-xop-ld128.c",
-    "src/qs8-gemm/gen/2x4c8-minmax-xop-ld128.c",
-    "src/qs8-gemm/gen/3x4c8-minmax-xop-ld128.c",
     "src/qs8-gemm/gen/1x4c8-xw-minmax-xop.c",
+    "src/qs8-gemm/gen/2x4c8-minmax-xop-ld64.c",
+    "src/qs8-gemm/gen/2x4c8-minmax-xop-ld128.c",
     "src/qs8-gemm/gen/2x4c8-xw-minmax-xop.c",
+    "src/qs8-gemm/gen/3x4c8-minmax-xop-ld64.c",
+    "src/qs8-gemm/gen/3x4c8-minmax-xop-ld128.c",
     "src/qs8-gemm/gen/3x4c8-xw-minmax-xop.c",
+    "src/qs8-gemm/gen/4x4c2-minmax-xop-ld64.c",
+    "src/qs8-gemm/gen/4x4c2-minmax-xop-ld128.c",
+    "src/qs8-gemm/gen/4x4c2-xw-minmax-xop.c",
     "src/qs8-igemm/gen/1x4c2-minmax-xop-ld64.c",
-    "src/qs8-igemm/gen/4x4c2-minmax-xop-ld64.c",
     "src/qs8-igemm/gen/1x4c2-minmax-xop-ld128.c",
-    "src/qs8-igemm/gen/4x4c2-minmax-xop-ld128.c",
     "src/qs8-igemm/gen/1x4c8-minmax-xop-ld64.c",
-    "src/qs8-igemm/gen/2x4c8-minmax-xop-ld64.c",
-    "src/qs8-igemm/gen/3x4c8-minmax-xop-ld64.c",
     "src/qs8-igemm/gen/1x4c8-minmax-xop-ld128.c",
+    "src/qs8-igemm/gen/2x4c8-minmax-xop-ld64.c",
     "src/qs8-igemm/gen/2x4c8-minmax-xop-ld128.c",
+    "src/qs8-igemm/gen/3x4c8-minmax-xop-ld64.c",
     "src/qs8-igemm/gen/3x4c8-minmax-xop-ld128.c",
+    "src/qs8-igemm/gen/4x4c2-minmax-xop-ld64.c",
+    "src/qs8-igemm/gen/4x4c2-minmax-xop-ld128.c",
     "src/qs8-vadd/gen/minmax-xop-mul32-ld32-x8.c",
     "src/qs8-vadd/gen/minmax-xop-mul32-ld32-x16.c",
     "src/qs8-vadd/gen/minmax-xop-mul32-ld32-x24.c",
@@ -2318,62 +2318,62 @@
 ]
 
 FMA3_UKERNELS = [
-    "src/f32-dwconv/gen/up16x4-minmax-fma3-acc2.c",
-    "src/f32-dwconv/gen/up16x4-minmax-fma3.c",
     "src/f32-dwconv/gen/up8x4-minmax-fma3-acc2.c",
     "src/f32-dwconv/gen/up8x4-minmax-fma3.c",
-    "src/f32-dwconv/gen/up16x9-minmax-fma3-acc2.c",
-    "src/f32-dwconv/gen/up16x9-minmax-fma3.c",
     "src/f32-dwconv/gen/up8x9-minmax-fma3-acc2.c",
     "src/f32-dwconv/gen/up8x9-minmax-fma3.c",
-    "src/f32-dwconv/gen/up16x25-minmax-fma3-acc2.c",
-    "src/f32-dwconv/gen/up16x25-minmax-fma3.c",
     "src/f32-dwconv/gen/up8x25-minmax-fma3-acc2.c",
     "src/f32-dwconv/gen/up8x25-minmax-fma3.c",
-    "src/f32-gemm/gen/1x8-minmax-fma3-broadcast.c",
-    "src/f32-gemm/gen/4x8-minmax-fma3-broadcast.c",
-    "src/f32-gemm/gen/5x8-minmax-fma3-broadcast.c",
-    "src/f32-gemm/gen/6x8-minmax-fma3-broadcast.c",
-    "src/f32-gemm/gen/7x8-minmax-fma3-broadcast.c",
-    "src/f32-gemm/gen/8x8-minmax-fma3-broadcast.c",
-    "src/f32-gemm/gen/1x16-minmax-fma3-broadcast.c",
-    "src/f32-gemm/gen/3x16-minmax-fma3-broadcast.c",
-    "src/f32-gemm/gen/4x16-minmax-fma3-broadcast.c",
-    "src/f32-gemm/gen/5x16-minmax-fma3-broadcast.c",
-    "src/f32-gemm/gen/1x16s4-minmax-fma3-broadcast.c",
-    "src/f32-gemm/gen/3x16s4-minmax-fma3-broadcast.c",
-    "src/f32-gemm/gen/4x16s4-minmax-fma3-broadcast.c",
-    "src/f32-gemm/gen/5x16s4-minmax-fma3-broadcast.c",
+    "src/f32-dwconv/gen/up16x4-minmax-fma3-acc2.c",
+    "src/f32-dwconv/gen/up16x4-minmax-fma3.c",
+    "src/f32-dwconv/gen/up16x9-minmax-fma3-acc2.c",
+    "src/f32-dwconv/gen/up16x9-minmax-fma3.c",
+    "src/f32-dwconv/gen/up16x25-minmax-fma3-acc2.c",
+    "src/f32-dwconv/gen/up16x25-minmax-fma3.c",
     "src/f32-gemm/gen-inc/1x8inc-minmax-fma3-broadcast.c",
+    "src/f32-gemm/gen-inc/1x16inc-minmax-fma3-broadcast.c",
+    "src/f32-gemm/gen-inc/1x16s4inc-minmax-fma3-broadcast.c",
+    "src/f32-gemm/gen-inc/3x16inc-minmax-fma3-broadcast.c",
+    "src/f32-gemm/gen-inc/3x16s4inc-minmax-fma3-broadcast.c",
     "src/f32-gemm/gen-inc/4x8inc-minmax-fma3-broadcast.c",
+    "src/f32-gemm/gen-inc/4x16inc-minmax-fma3-broadcast.c",
+    "src/f32-gemm/gen-inc/4x16s4inc-minmax-fma3-broadcast.c",
     "src/f32-gemm/gen-inc/5x8inc-minmax-fma3-broadcast.c",
+    "src/f32-gemm/gen-inc/5x16inc-minmax-fma3-broadcast.c",
+    "src/f32-gemm/gen-inc/5x16s4inc-minmax-fma3-broadcast.c",
     "src/f32-gemm/gen-inc/6x8inc-minmax-fma3-broadcast.c",
     "src/f32-gemm/gen-inc/7x8inc-minmax-fma3-broadcast.c",
     "src/f32-gemm/gen-inc/8x8inc-minmax-fma3-broadcast.c",
-    "src/f32-gemm/gen-inc/1x16inc-minmax-fma3-broadcast.c",
-    "src/f32-gemm/gen-inc/3x16inc-minmax-fma3-broadcast.c",
-    "src/f32-gemm/gen-inc/4x16inc-minmax-fma3-broadcast.c",
-    "src/f32-gemm/gen-inc/5x16inc-minmax-fma3-broadcast.c",
-    "src/f32-gemm/gen-inc/1x16s4inc-minmax-fma3-broadcast.c",
-    "src/f32-gemm/gen-inc/3x16s4inc-minmax-fma3-broadcast.c",
-    "src/f32-gemm/gen-inc/4x16s4inc-minmax-fma3-broadcast.c",
-    "src/f32-gemm/gen-inc/5x16s4inc-minmax-fma3-broadcast.c",
+    "src/f32-gemm/gen/1x8-minmax-fma3-broadcast.c",
+    "src/f32-gemm/gen/1x16-minmax-fma3-broadcast.c",
+    "src/f32-gemm/gen/1x16s4-minmax-fma3-broadcast.c",
+    "src/f32-gemm/gen/3x16-minmax-fma3-broadcast.c",
+    "src/f32-gemm/gen/3x16s4-minmax-fma3-broadcast.c",
+    "src/f32-gemm/gen/4x8-minmax-fma3-broadcast.c",
+    "src/f32-gemm/gen/4x16-minmax-fma3-broadcast.c",
+    "src/f32-gemm/gen/4x16s4-minmax-fma3-broadcast.c",
+    "src/f32-gemm/gen/5x8-minmax-fma3-broadcast.c",
+    "src/f32-gemm/gen/5x16-minmax-fma3-broadcast.c",
+    "src/f32-gemm/gen/5x16s4-minmax-fma3-broadcast.c",
+    "src/f32-gemm/gen/6x8-minmax-fma3-broadcast.c",
+    "src/f32-gemm/gen/7x8-minmax-fma3-broadcast.c",
+    "src/f32-gemm/gen/8x8-minmax-fma3-broadcast.c",
     "src/f32-hswish/gen/hswish-fma3-x8.c",
     "src/f32-hswish/gen/hswish-fma3-x16.c",
     "src/f32-igemm/gen/1x8-minmax-fma3-broadcast.c",
+    "src/f32-igemm/gen/1x16-minmax-fma3-broadcast.c",
+    "src/f32-igemm/gen/1x16s4-minmax-fma3-broadcast.c",
+    "src/f32-igemm/gen/3x16-minmax-fma3-broadcast.c",
+    "src/f32-igemm/gen/3x16s4-minmax-fma3-broadcast.c",
     "src/f32-igemm/gen/4x8-minmax-fma3-broadcast.c",
+    "src/f32-igemm/gen/4x16-minmax-fma3-broadcast.c",
+    "src/f32-igemm/gen/4x16s4-minmax-fma3-broadcast.c",
     "src/f32-igemm/gen/5x8-minmax-fma3-broadcast.c",
+    "src/f32-igemm/gen/5x16-minmax-fma3-broadcast.c",
+    "src/f32-igemm/gen/5x16s4-minmax-fma3-broadcast.c",
     "src/f32-igemm/gen/6x8-minmax-fma3-broadcast.c",
     "src/f32-igemm/gen/7x8-minmax-fma3-broadcast.c",
     "src/f32-igemm/gen/8x8-minmax-fma3-broadcast.c",
-    "src/f32-igemm/gen/1x16-minmax-fma3-broadcast.c",
-    "src/f32-igemm/gen/3x16-minmax-fma3-broadcast.c",
-    "src/f32-igemm/gen/4x16-minmax-fma3-broadcast.c",
-    "src/f32-igemm/gen/5x16-minmax-fma3-broadcast.c",
-    "src/f32-igemm/gen/1x16s4-minmax-fma3-broadcast.c",
-    "src/f32-igemm/gen/3x16s4-minmax-fma3-broadcast.c",
-    "src/f32-igemm/gen/4x16s4-minmax-fma3-broadcast.c",
-    "src/f32-igemm/gen/5x16s4-minmax-fma3-broadcast.c",
     "src/f32-vsqrt/gen/fma3-nr1fma1adj-x8.c",
     "src/f32-vsqrt/gen/fma3-nr1fma1adj-x16.c",
     "src/f32-vsqrt/gen/fma3-nr1fma1adj-x24.c",
@@ -2383,47 +2383,47 @@
     "src/f32-vsqrt/gen/fma3-nr1fma1adj-x56.c",
     "src/f32-vsqrt/gen/fma3-nr1fma1adj-x64.c",
     "src/math/sqrt-fma3-nr1fma.c",
-    "src/math/sqrt-fma3-nr2fma.c",
     "src/math/sqrt-fma3-nr1fma1adj.c",
+    "src/math/sqrt-fma3-nr2fma.c",
 ]
 
 AVX2_UKERNELS = [
-    "src/f32-raddexpminusmax/gen/avx2-p5-x64.c",
     "src/f32-raddexpminusmax/gen/avx2-p5-x64-acc2.c",
     "src/f32-raddexpminusmax/gen/avx2-p5-x64-acc4.c",
-    "src/f32-raddexpminusmax/gen/avx2-p5-x72.c",
+    "src/f32-raddexpminusmax/gen/avx2-p5-x64.c",
     "src/f32-raddexpminusmax/gen/avx2-p5-x72-acc3.c",
-    "src/f32-raddexpminusmax/gen/avx2-p5-x80.c",
+    "src/f32-raddexpminusmax/gen/avx2-p5-x72.c",
     "src/f32-raddexpminusmax/gen/avx2-p5-x80-acc2.c",
     "src/f32-raddexpminusmax/gen/avx2-p5-x80-acc5.c",
-    "src/f32-raddexpminusmax/gen/avx2-p5-x96.c",
+    "src/f32-raddexpminusmax/gen/avx2-p5-x80.c",
     "src/f32-raddexpminusmax/gen/avx2-p5-x96-acc2.c",
     "src/f32-raddexpminusmax/gen/avx2-p5-x96-acc3.c",
     "src/f32-raddexpminusmax/gen/avx2-p5-x96-acc6.c",
-    "src/f32-raddextexp/gen/avx2-p5-x64.c",
+    "src/f32-raddexpminusmax/gen/avx2-p5-x96.c",
     "src/f32-raddextexp/gen/avx2-p5-x64-acc2.c",
     "src/f32-raddextexp/gen/avx2-p5-x64-acc4.c",
-    "src/f32-raddextexp/gen/avx2-p5-x72.c",
+    "src/f32-raddextexp/gen/avx2-p5-x64.c",
     "src/f32-raddextexp/gen/avx2-p5-x72-acc3.c",
-    "src/f32-raddextexp/gen/avx2-p5-x80.c",
+    "src/f32-raddextexp/gen/avx2-p5-x72.c",
     "src/f32-raddextexp/gen/avx2-p5-x80-acc2.c",
     "src/f32-raddextexp/gen/avx2-p5-x80-acc5.c",
-    "src/f32-raddextexp/gen/avx2-p5-x96.c",
+    "src/f32-raddextexp/gen/avx2-p5-x80.c",
     "src/f32-raddextexp/gen/avx2-p5-x96-acc2.c",
     "src/f32-raddextexp/gen/avx2-p5-x96-acc3.c",
     "src/f32-raddextexp/gen/avx2-p5-x96-acc6.c",
-    "src/f32-raddstoreexpminusmax/gen/avx2-p5-x64.c",
+    "src/f32-raddextexp/gen/avx2-p5-x96.c",
     "src/f32-raddstoreexpminusmax/gen/avx2-p5-x64-acc2.c",
     "src/f32-raddstoreexpminusmax/gen/avx2-p5-x64-acc4.c",
-    "src/f32-raddstoreexpminusmax/gen/avx2-p5-x72.c",
+    "src/f32-raddstoreexpminusmax/gen/avx2-p5-x64.c",
     "src/f32-raddstoreexpminusmax/gen/avx2-p5-x72-acc3.c",
-    "src/f32-raddstoreexpminusmax/gen/avx2-p5-x80.c",
+    "src/f32-raddstoreexpminusmax/gen/avx2-p5-x72.c",
     "src/f32-raddstoreexpminusmax/gen/avx2-p5-x80-acc2.c",
     "src/f32-raddstoreexpminusmax/gen/avx2-p5-x80-acc5.c",
-    "src/f32-raddstoreexpminusmax/gen/avx2-p5-x96.c",
+    "src/f32-raddstoreexpminusmax/gen/avx2-p5-x80.c",
     "src/f32-raddstoreexpminusmax/gen/avx2-p5-x96-acc2.c",
     "src/f32-raddstoreexpminusmax/gen/avx2-p5-x96-acc3.c",
     "src/f32-raddstoreexpminusmax/gen/avx2-p5-x96-acc6.c",
+    "src/f32-raddstoreexpminusmax/gen/avx2-p5-x96.c",
     "src/f32-sigmoid/gen/avx2-rr1-p5-div-x8.c",
     "src/f32-sigmoid/gen/avx2-rr1-p5-div-x16.c",
     "src/f32-sigmoid/gen/avx2-rr1-p5-div-x24.c",
@@ -2478,17 +2478,36 @@
     "src/f32-vscaleextexp/gen/avx2-p5-x80.c",
     "src/f32-vscaleextexp/gen/avx2-p5-x88.c",
     "src/f32-vscaleextexp/gen/avx2-p5-x96.c",
-    "src/qs8-dwconv/gen/up16x9-minmax-avx2-mul16.c",
-    "src/qs8-dwconv/gen/up32x9-minmax-avx2-mul16.c",
+    "src/math/exp-avx2-lut8-p3-perm.c",
+    "src/math/exp-avx2-lut8-p4-perm.c",
+    "src/math/exp-avx2-p5.c",
+    "src/math/expminus-avx2-p5.c",
+    "src/math/extexp-avx2-p5.c",
+    "src/math/sigmoid-avx2-rr1-lut64-p2-gather-div.c",
+    "src/math/sigmoid-avx2-rr1-lut64-p2-gather-nr1fma.c",
+    "src/math/sigmoid-avx2-rr1-lut64-p2-gather-nr2fma.c",
+    "src/math/sigmoid-avx2-rr1-lut64-p2-gather-nr2fma1adj.c",
+    "src/math/sigmoid-avx2-rr1-p5-div.c",
+    "src/math/sigmoid-avx2-rr1-p5-nr1fma.c",
+    "src/math/sigmoid-avx2-rr1-p5-nr2fma.c",
+    "src/math/sigmoid-avx2-rr2-lut64-p2-gather-div.c",
+    "src/math/sigmoid-avx2-rr2-lut64-p2-gather-nr1fma.c",
+    "src/math/sigmoid-avx2-rr2-lut64-p2-gather-nr2fma.c",
+    "src/math/sigmoid-avx2-rr2-lut64-p2-gather-nr2fma1adj.c",
+    "src/math/sigmoid-avx2-rr2-p5-div.c",
+    "src/math/sigmoid-avx2-rr2-p5-nr1fma.c",
+    "src/math/sigmoid-avx2-rr2-p5-nr2fma.c",
     "src/qs8-dwconv/gen/up8x9-minmax-avx2-mul32.c",
+    "src/qs8-dwconv/gen/up16x9-minmax-avx2-mul16.c",
     "src/qs8-dwconv/gen/up16x9-minmax-avx2-mul32.c",
     "src/qs8-dwconv/gen/up24x9-minmax-avx2-mul32.c",
+    "src/qs8-dwconv/gen/up32x9-minmax-avx2-mul16.c",
     "src/qs8-dwconv/gen/up32x9-minmax-avx2-mul32.c",
     "src/qs8-gemm/gen/1x8c8-minmax-avx2.c",
-    "src/qs8-gemm/gen/2x8c8-minmax-avx2.c",
-    "src/qs8-gemm/gen/3x8c8-minmax-avx2.c",
     "src/qs8-gemm/gen/1x8c8-xw-minmax-avx2.c",
+    "src/qs8-gemm/gen/2x8c8-minmax-avx2.c",
     "src/qs8-gemm/gen/2x8c8-xw-minmax-avx2.c",
+    "src/qs8-gemm/gen/3x8c8-minmax-avx2.c",
     "src/qs8-gemm/gen/3x8c8-xw-minmax-avx2.c",
     "src/qs8-igemm/gen/1x8c8-minmax-avx2.c",
     "src/qs8-igemm/gen/2x8c8-minmax-avx2.c",
@@ -2501,54 +2520,35 @@
     "src/qs8-vaddc/gen/minmax-avx2-mul32-ld64-x16.c",
     "src/qs8-vaddc/gen/minmax-avx2-mul32-ld64-x24.c",
     "src/qs8-vaddc/gen/minmax-avx2-mul32-ld64-x32.c",
-    "src/math/exp-avx2-p5.c",
-    "src/math/exp-avx2-lut8-p3-perm.c",
-    "src/math/exp-avx2-lut8-p4-perm.c",
-    "src/math/expminus-avx2-p5.c",
-    "src/math/extexp-avx2-p5.c",
-    "src/math/sigmoid-avx2-rr1-p5-div.c",
-    "src/math/sigmoid-avx2-rr2-p5-div.c",
-    "src/math/sigmoid-avx2-rr1-p5-nr2fma.c",
-    "src/math/sigmoid-avx2-rr2-p5-nr2fma.c",
-    "src/math/sigmoid-avx2-rr1-p5-nr1fma.c",
-    "src/math/sigmoid-avx2-rr2-p5-nr1fma.c",
-    "src/math/sigmoid-avx2-rr1-lut64-p2-gather-div.c",
-    "src/math/sigmoid-avx2-rr2-lut64-p2-gather-div.c",
-    "src/math/sigmoid-avx2-rr1-lut64-p2-gather-nr2fma1adj.c",
-    "src/math/sigmoid-avx2-rr2-lut64-p2-gather-nr2fma1adj.c",
-    "src/math/sigmoid-avx2-rr1-lut64-p2-gather-nr2fma.c",
-    "src/math/sigmoid-avx2-rr2-lut64-p2-gather-nr2fma.c",
-    "src/math/sigmoid-avx2-rr1-lut64-p2-gather-nr1fma.c",
-    "src/math/sigmoid-avx2-rr2-lut64-p2-gather-nr1fma.c",
 ]
 
 AVX512F_UKERNELS = [
     "src/f32-clamp/gen/avx512f-x16.c",
     "src/f32-clamp/gen/avx512f-x32.c",
-    "src/f32-dwconv/gen/up32x4-minmax-avx512f-acc2.c",
-    "src/f32-dwconv/gen/up32x4-minmax-avx512f.c",
     "src/f32-dwconv/gen/up16x4-minmax-avx512f-acc2.c",
     "src/f32-dwconv/gen/up16x4-minmax-avx512f.c",
-    "src/f32-dwconv/gen/up32x9-minmax-avx512f-acc2.c",
-    "src/f32-dwconv/gen/up32x9-minmax-avx512f.c",
     "src/f32-dwconv/gen/up16x9-minmax-avx512f-acc2.c",
     "src/f32-dwconv/gen/up16x9-minmax-avx512f.c",
-    "src/f32-dwconv/gen/up32x25-minmax-avx512f-acc2.c",
-    "src/f32-dwconv/gen/up32x25-minmax-avx512f.c",
     "src/f32-dwconv/gen/up16x25-minmax-avx512f-acc2.c",
     "src/f32-dwconv/gen/up16x25-minmax-avx512f.c",
-    "src/f32-gemm/gen/1x16-minmax-avx512f-broadcast.c",
-    "src/f32-gemm/gen/4x16-minmax-avx512f-broadcast.c",
-    "src/f32-gemm/gen/5x16-minmax-avx512f-broadcast.c",
-    "src/f32-gemm/gen/6x16-minmax-avx512f-broadcast.c",
-    "src/f32-gemm/gen/7x16-minmax-avx512f-broadcast.c",
-    "src/f32-gemm/gen/8x16-minmax-avx512f-broadcast.c",
+    "src/f32-dwconv/gen/up32x4-minmax-avx512f-acc2.c",
+    "src/f32-dwconv/gen/up32x4-minmax-avx512f.c",
+    "src/f32-dwconv/gen/up32x9-minmax-avx512f-acc2.c",
+    "src/f32-dwconv/gen/up32x9-minmax-avx512f.c",
+    "src/f32-dwconv/gen/up32x25-minmax-avx512f-acc2.c",
+    "src/f32-dwconv/gen/up32x25-minmax-avx512f.c",
     "src/f32-gemm/gen-inc/1x16inc-minmax-avx512f-broadcast.c",
     "src/f32-gemm/gen-inc/4x16inc-minmax-avx512f-broadcast.c",
     "src/f32-gemm/gen-inc/5x16inc-minmax-avx512f-broadcast.c",
     "src/f32-gemm/gen-inc/6x16inc-minmax-avx512f-broadcast.c",
     "src/f32-gemm/gen-inc/7x16inc-minmax-avx512f-broadcast.c",
     "src/f32-gemm/gen-inc/8x16inc-minmax-avx512f-broadcast.c",
+    "src/f32-gemm/gen/1x16-minmax-avx512f-broadcast.c",
+    "src/f32-gemm/gen/4x16-minmax-avx512f-broadcast.c",
+    "src/f32-gemm/gen/5x16-minmax-avx512f-broadcast.c",
+    "src/f32-gemm/gen/6x16-minmax-avx512f-broadcast.c",
+    "src/f32-gemm/gen/7x16-minmax-avx512f-broadcast.c",
+    "src/f32-gemm/gen/8x16-minmax-avx512f-broadcast.c",
     "src/f32-hswish/gen/hswish-avx512f-x16.c",
     "src/f32-hswish/gen/hswish-avx512f-x32.c",
     "src/f32-igemm/gen/1x16-minmax-avx512f-broadcast.c",
@@ -2559,61 +2559,45 @@
     "src/f32-igemm/gen/8x16-minmax-avx512f-broadcast.c",
     "src/f32-prelu/gen/avx512f-2x16.c",
     "src/f32-prelu/gen/avx512f-2x32.c",
-    "src/f32-raddexpminusmax/gen/avx512f-p5-scalef-x128.c",
     "src/f32-raddexpminusmax/gen/avx512f-p5-scalef-x128-acc2.c",
     "src/f32-raddexpminusmax/gen/avx512f-p5-scalef-x128-acc4.c",
-    "src/f32-raddexpminusmax/gen/avx512f-p5-scalef-x144.c",
+    "src/f32-raddexpminusmax/gen/avx512f-p5-scalef-x128.c",
     "src/f32-raddexpminusmax/gen/avx512f-p5-scalef-x144-acc3.c",
-    "src/f32-raddexpminusmax/gen/avx512f-p5-scalef-x160.c",
+    "src/f32-raddexpminusmax/gen/avx512f-p5-scalef-x144.c",
     "src/f32-raddexpminusmax/gen/avx512f-p5-scalef-x160-acc2.c",
     "src/f32-raddexpminusmax/gen/avx512f-p5-scalef-x160-acc5.c",
-    "src/f32-raddexpminusmax/gen/avx512f-p5-scalef-x192.c",
+    "src/f32-raddexpminusmax/gen/avx512f-p5-scalef-x160.c",
     "src/f32-raddexpminusmax/gen/avx512f-p5-scalef-x192-acc2.c",
     "src/f32-raddexpminusmax/gen/avx512f-p5-scalef-x192-acc3.c",
     "src/f32-raddexpminusmax/gen/avx512f-p5-scalef-x192-acc6.c",
-    "src/f32-raddextexp/gen/avx512f-p5-scalef-x128.c",
+    "src/f32-raddexpminusmax/gen/avx512f-p5-scalef-x192.c",
     "src/f32-raddextexp/gen/avx512f-p5-scalef-x128-acc2.c",
     "src/f32-raddextexp/gen/avx512f-p5-scalef-x128-acc4.c",
-    "src/f32-raddextexp/gen/avx512f-p5-scalef-x144.c",
+    "src/f32-raddextexp/gen/avx512f-p5-scalef-x128.c",
     "src/f32-raddextexp/gen/avx512f-p5-scalef-x144-acc3.c",
-    "src/f32-raddextexp/gen/avx512f-p5-scalef-x160.c",
+    "src/f32-raddextexp/gen/avx512f-p5-scalef-x144.c",
     "src/f32-raddextexp/gen/avx512f-p5-scalef-x160-acc2.c",
     "src/f32-raddextexp/gen/avx512f-p5-scalef-x160-acc5.c",
-    "src/f32-raddextexp/gen/avx512f-p5-scalef-x192.c",
+    "src/f32-raddextexp/gen/avx512f-p5-scalef-x160.c",
     "src/f32-raddextexp/gen/avx512f-p5-scalef-x192-acc2.c",
     "src/f32-raddextexp/gen/avx512f-p5-scalef-x192-acc3.c",
     "src/f32-raddextexp/gen/avx512f-p5-scalef-x192-acc6.c",
-    "src/f32-raddstoreexpminusmax/gen/avx512f-p5-scalef-x128.c",
+    "src/f32-raddextexp/gen/avx512f-p5-scalef-x192.c",
     "src/f32-raddstoreexpminusmax/gen/avx512f-p5-scalef-x128-acc2.c",
     "src/f32-raddstoreexpminusmax/gen/avx512f-p5-scalef-x128-acc4.c",
-    "src/f32-raddstoreexpminusmax/gen/avx512f-p5-scalef-x144.c",
+    "src/f32-raddstoreexpminusmax/gen/avx512f-p5-scalef-x128.c",
     "src/f32-raddstoreexpminusmax/gen/avx512f-p5-scalef-x144-acc3.c",
-    "src/f32-raddstoreexpminusmax/gen/avx512f-p5-scalef-x160.c",
+    "src/f32-raddstoreexpminusmax/gen/avx512f-p5-scalef-x144.c",
     "src/f32-raddstoreexpminusmax/gen/avx512f-p5-scalef-x160-acc2.c",
     "src/f32-raddstoreexpminusmax/gen/avx512f-p5-scalef-x160-acc5.c",
-    "src/f32-raddstoreexpminusmax/gen/avx512f-p5-scalef-x192.c",
+    "src/f32-raddstoreexpminusmax/gen/avx512f-p5-scalef-x160.c",
     "src/f32-raddstoreexpminusmax/gen/avx512f-p5-scalef-x192-acc2.c",
     "src/f32-raddstoreexpminusmax/gen/avx512f-p5-scalef-x192-acc3.c",
     "src/f32-raddstoreexpminusmax/gen/avx512f-p5-scalef-x192-acc6.c",
+    "src/f32-raddstoreexpminusmax/gen/avx512f-p5-scalef-x192.c",
     "src/f32-relu/gen/avx512f-x16.c",
     "src/f32-relu/gen/avx512f-x32.c",
     "src/f32-rmax/avx512f.c",
-    "src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-div-x16.c",
-    "src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-div-x32.c",
-    "src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-div-x48.c",
-    "src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-div-x64.c",
-    "src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-div-x80.c",
-    "src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-div-x96.c",
-    "src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-div-x112.c",
-    "src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-div-x128.c",
-    "src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-nr1fma-x16.c",
-    "src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-nr1fma-x32.c",
-    "src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-nr1fma-x48.c",
-    "src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-nr1fma-x64.c",
-    "src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-nr1fma-x80.c",
-    "src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-nr1fma-x96.c",
-    "src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-nr1fma-x112.c",
-    "src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-nr1fma-x128.c",
     "src/f32-sigmoid/gen/avx512f-rr1-lut16-p3-perm-scalef-div-x16.c",
     "src/f32-sigmoid/gen/avx512f-rr1-lut16-p3-perm-scalef-div-x32.c",
     "src/f32-sigmoid/gen/avx512f-rr1-lut16-p3-perm-scalef-div-x48.c",
@@ -2630,6 +2614,22 @@
     "src/f32-sigmoid/gen/avx512f-rr1-lut16-p3-perm-scalef-nr1fma-x96.c",
     "src/f32-sigmoid/gen/avx512f-rr1-lut16-p3-perm-scalef-nr1fma-x112.c",
     "src/f32-sigmoid/gen/avx512f-rr1-lut16-p3-perm-scalef-nr1fma-x128.c",
+    "src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-div-x16.c",
+    "src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-div-x32.c",
+    "src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-div-x48.c",
+    "src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-div-x64.c",
+    "src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-div-x80.c",
+    "src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-div-x96.c",
+    "src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-div-x112.c",
+    "src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-div-x128.c",
+    "src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-nr1fma-x16.c",
+    "src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-nr1fma-x32.c",
+    "src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-nr1fma-x48.c",
+    "src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-nr1fma-x64.c",
+    "src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-nr1fma-x80.c",
+    "src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-nr1fma-x96.c",
+    "src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-nr1fma-x112.c",
+    "src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-nr1fma-x128.c",
     "src/f32-sigmoid/gen/avx512f-rr2-lut32-p2-perm2-scalef-div-x16.c",
     "src/f32-sigmoid/gen/avx512f-rr2-lut32-p2-perm2-scalef-div-x32.c",
     "src/f32-sigmoid/gen/avx512f-rr2-lut32-p2-perm2-scalef-div-x48.c",
@@ -2680,6 +2680,14 @@
     "src/f32-vbinary/gen/vsubc-minmax-avx512f-x32.c",
     "src/f32-vlrelu/gen/vlrelu-avx512f-x16.c",
     "src/f32-vlrelu/gen/vlrelu-avx512f-x32.c",
+    "src/f32-vrnd/gen/vrndd-avx512f-x16.c",
+    "src/f32-vrnd/gen/vrndd-avx512f-x32.c",
+    "src/f32-vrnd/gen/vrndne-avx512f-x16.c",
+    "src/f32-vrnd/gen/vrndne-avx512f-x32.c",
+    "src/f32-vrnd/gen/vrndu-avx512f-x16.c",
+    "src/f32-vrnd/gen/vrndu-avx512f-x32.c",
+    "src/f32-vrnd/gen/vrndz-avx512f-x16.c",
+    "src/f32-vrnd/gen/vrndz-avx512f-x32.c",
     "src/f32-vscale/avx512f-unroll64.c",
     "src/f32-vscaleexpminusmax/gen/avx512f-p5-scalef-x16.c",
     "src/f32-vscaleexpminusmax/gen/avx512f-p5-scalef-x32.c",
@@ -2705,14 +2713,6 @@
     "src/f32-vscaleextexp/gen/avx512f-p5-scalef-x160.c",
     "src/f32-vscaleextexp/gen/avx512f-p5-scalef-x176.c",
     "src/f32-vscaleextexp/gen/avx512f-p5-scalef-x192.c",
-    "src/f32-vrnd/gen/vrndne-avx512f-x16.c",
-    "src/f32-vrnd/gen/vrndne-avx512f-x32.c",
-    "src/f32-vrnd/gen/vrndz-avx512f-x16.c",
-    "src/f32-vrnd/gen/vrndz-avx512f-x32.c",
-    "src/f32-vrnd/gen/vrndu-avx512f-x16.c",
-    "src/f32-vrnd/gen/vrndu-avx512f-x32.c",
-    "src/f32-vrnd/gen/vrndd-avx512f-x16.c",
-    "src/f32-vrnd/gen/vrndd-avx512f-x32.c",
     "src/f32-vsqrt/gen/avx512f-nr1fma1adj-x16.c",
     "src/f32-vsqrt/gen/avx512f-nr1fma1adj-x32.c",
     "src/f32-vsqrt/gen/avx512f-nr1fma1adj-x48.c",
@@ -2727,40 +2727,40 @@
     "src/f32-vunary/gen/vneg-avx512f-x32.c",
     "src/f32-vunary/gen/vsqr-avx512f-x16.c",
     "src/f32-vunary/gen/vsqr-avx512f-x32.c",
+    "src/math/exp-avx512f-lut16-p3-perm-scalef.c",
+    "src/math/exp-avx512f-lut16-p3-perm.c",
+    "src/math/exp-avx512f-lut32-p2-perm2-scalef.c",
+    "src/math/exp-avx512f-lut32-p2-perm2.c",
     "src/math/exp-avx512f-p5-scalef.c",
     "src/math/exp-avx512f-p5.c",
-    "src/math/exp-avx512f-lut16-p3-perm.c",
-    "src/math/exp-avx512f-lut16-p3-perm-scalef.c",
-    "src/math/exp-avx512f-lut32-p2-perm2.c",
-    "src/math/exp-avx512f-lut32-p2-perm2-scalef.c",
     "src/math/extexp-avx512f-p5.c",
-    "src/math/sigmoid-avx512f-rr1-p5-scalef-div.c",
-    "src/math/sigmoid-avx512f-rr1-p5-scalef-nr1fma1adj.c",
-    "src/math/sigmoid-avx512f-rr1-p5-scalef-nr1fma.c",
-    "src/math/sigmoid-avx512f-rr2-p5-scalef-div.c",
-    "src/math/sigmoid-avx512f-rr2-p5-scalef-nr1fma1adj.c",
-    "src/math/sigmoid-avx512f-rr2-p5-scalef-nr1fma.c",
     "src/math/sigmoid-avx512f-rr1-lut16-p3-perm-scalef-div.c",
-    "src/math/sigmoid-avx512f-rr1-lut16-p3-perm-scalef-nr1fma1adj.c",
     "src/math/sigmoid-avx512f-rr1-lut16-p3-perm-scalef-nr1fma.c",
-    "src/math/sigmoid-avx512f-rr2-lut16-p3-perm-scalef-div.c",
-    "src/math/sigmoid-avx512f-rr2-lut16-p3-perm-scalef-nr1fma1adj.c",
-    "src/math/sigmoid-avx512f-rr2-lut16-p3-perm-scalef-nr1fma.c",
+    "src/math/sigmoid-avx512f-rr1-lut16-p3-perm-scalef-nr1fma1adj.c",
     "src/math/sigmoid-avx512f-rr1-lut32-p2-perm2-scalef-div.c",
-    "src/math/sigmoid-avx512f-rr1-lut32-p2-perm2-scalef-nr1fma1adj.c",
     "src/math/sigmoid-avx512f-rr1-lut32-p2-perm2-scalef-nr1fma.c",
-    "src/math/sigmoid-avx512f-rr2-lut32-p2-perm2-scalef-div.c",
-    "src/math/sigmoid-avx512f-rr2-lut32-p2-perm2-scalef-nr1fma1adj.c",
-    "src/math/sigmoid-avx512f-rr2-lut32-p2-perm2-scalef-nr1fma.c",
+    "src/math/sigmoid-avx512f-rr1-lut32-p2-perm2-scalef-nr1fma1adj.c",
     "src/math/sigmoid-avx512f-rr1-lut64-p2-gather-scalef-div.c",
-    "src/math/sigmoid-avx512f-rr1-lut64-p2-gather-scalef-nr1fma1adj.c",
     "src/math/sigmoid-avx512f-rr1-lut64-p2-gather-scalef-nr1fma.c",
+    "src/math/sigmoid-avx512f-rr1-lut64-p2-gather-scalef-nr1fma1adj.c",
+    "src/math/sigmoid-avx512f-rr1-p5-scalef-div.c",
+    "src/math/sigmoid-avx512f-rr1-p5-scalef-nr1fma.c",
+    "src/math/sigmoid-avx512f-rr1-p5-scalef-nr1fma1adj.c",
+    "src/math/sigmoid-avx512f-rr2-lut16-p3-perm-scalef-div.c",
+    "src/math/sigmoid-avx512f-rr2-lut16-p3-perm-scalef-nr1fma.c",
+    "src/math/sigmoid-avx512f-rr2-lut16-p3-perm-scalef-nr1fma1adj.c",
+    "src/math/sigmoid-avx512f-rr2-lut32-p2-perm2-scalef-div.c",
+    "src/math/sigmoid-avx512f-rr2-lut32-p2-perm2-scalef-nr1fma.c",
+    "src/math/sigmoid-avx512f-rr2-lut32-p2-perm2-scalef-nr1fma1adj.c",
     "src/math/sigmoid-avx512f-rr2-lut64-p2-gather-scalef-div.c",
-    "src/math/sigmoid-avx512f-rr2-lut64-p2-gather-scalef-nr1fma1adj.c",
     "src/math/sigmoid-avx512f-rr2-lut64-p2-gather-scalef-nr1fma.c",
+    "src/math/sigmoid-avx512f-rr2-lut64-p2-gather-scalef-nr1fma1adj.c",
+    "src/math/sigmoid-avx512f-rr2-p5-scalef-div.c",
+    "src/math/sigmoid-avx512f-rr2-p5-scalef-nr1fma.c",
+    "src/math/sigmoid-avx512f-rr2-p5-scalef-nr1fma1adj.c",
     "src/math/sqrt-avx512f-nr1fma.c",
-    "src/math/sqrt-avx512f-nr2fma.c",
     "src/math/sqrt-avx512f-nr1fma1adj.c",
+    "src/math/sqrt-avx512f-nr2fma.c",
 ]
 
 AVX512SKX_UKERNELS = [
@@ -2800,78 +2800,78 @@
 ]
 
 AARCH64_ASM_UKERNELS = [
-    "src/f16-gemm/gen/1x16-minmax-aarch64-neonfp16arith-ld32.S",
-    "src/f16-gemm/gen/4x16-minmax-aarch64-neonfp16arith-ld32.S",
-    "src/f16-gemm/gen/6x16-minmax-aarch64-neonfp16arith-ld32.S",
-    "src/f16-gemm/gen-inc/1x16inc-minmax-aarch64-neonfp16arith-ld32.S",
-    "src/f16-gemm/gen-inc/4x16inc-minmax-aarch64-neonfp16arith-ld32.S",
-    "src/f16-gemm/gen-inc/6x16inc-minmax-aarch64-neonfp16arith-ld32.S",
-    "src/f16-gemm/gen/1x8-minmax-aarch64-neonfp16arith-ld64.S",
-    "src/f16-gemm/gen/4x8-minmax-aarch64-neonfp16arith-ld64.S",
-    "src/f16-gemm/gen/6x8-minmax-aarch64-neonfp16arith-ld64.S",
-    "src/f16-gemm/gen/8x8-minmax-aarch64-neonfp16arith-ld64.S",
     "src/f16-gemm/gen-inc/1x8inc-minmax-aarch64-neonfp16arith-ld64.S",
+    "src/f16-gemm/gen-inc/1x16inc-minmax-aarch64-neonfp16arith-ld32.S",
     "src/f16-gemm/gen-inc/4x8inc-minmax-aarch64-neonfp16arith-ld64.S",
+    "src/f16-gemm/gen-inc/4x16inc-minmax-aarch64-neonfp16arith-ld32.S",
     "src/f16-gemm/gen-inc/6x8inc-minmax-aarch64-neonfp16arith-ld64.S",
+    "src/f16-gemm/gen-inc/6x16inc-minmax-aarch64-neonfp16arith-ld32.S",
     "src/f16-gemm/gen-inc/8x8inc-minmax-aarch64-neonfp16arith-ld64.S",
+    "src/f16-gemm/gen/1x8-minmax-aarch64-neonfp16arith-ld64.S",
+    "src/f16-gemm/gen/1x16-minmax-aarch64-neonfp16arith-ld32.S",
+    "src/f16-gemm/gen/4x8-minmax-aarch64-neonfp16arith-ld64.S",
+    "src/f16-gemm/gen/4x16-minmax-aarch64-neonfp16arith-ld32.S",
+    "src/f16-gemm/gen/6x8-minmax-aarch64-neonfp16arith-ld64.S",
+    "src/f16-gemm/gen/6x16-minmax-aarch64-neonfp16arith-ld32.S",
+    "src/f16-gemm/gen/8x8-minmax-aarch64-neonfp16arith-ld64.S",
     "src/f32-dwconv/up4x9-minmax-aarch64-neonfma-cortex-a55.S",
     "src/f32-dwconv/up4x9-minmax-aarch64-neonfma.S",
-    "src/f32-gemm/gen/1x8-minmax-aarch64-neonfma-ld64.S",
-    "src/f32-gemm/gen/1x12-minmax-aarch64-neonfma-cortex-a53.S",
-    "src/f32-gemm/gen/1x8-minmax-aarch64-neonfma-cortex-a53.S",
-    "src/f32-gemm/gen/1x8-minmax-aarch64-neonfma-cortex-a57.S",
-    "src/f32-gemm/gen/1x8-minmax-aarch64-neonfma-cortex-a75.S",
-    "src/f32-gemm/gen/4x12-minmax-aarch64-neonfma-cortex-a53.S",
-    "src/f32-gemm/gen/4x8-minmax-aarch64-neonfma-cortex-a53.S",
-    "src/f32-gemm/gen/4x8-minmax-aarch64-neonfma-cortex-a55.S",
-    "src/f32-gemm/gen/4x8-minmax-aarch64-neonfma-cortex-a57.S",
-    "src/f32-gemm/gen/4x8-minmax-aarch64-neonfma-cortex-a75.S",
-    "src/f32-gemm/gen/4x8-minmax-aarch64-neonfma-ld128.S",
-    "src/f32-gemm/gen/4x8-minmax-aarch64-neonfma-ld64.S",
-    "src/f32-gemm/gen/5x8-minmax-aarch64-neonfma-cortex-a57.S",
-    "src/f32-gemm/gen/5x8-minmax-aarch64-neonfma-cortex-a75.S",
-    "src/f32-gemm/gen/6x8-minmax-aarch64-neonfma-cortex-a53.S",
-    "src/f32-gemm/gen/6x8-minmax-aarch64-neonfma-cortex-a55.S",
-    "src/f32-gemm/gen/6x8-minmax-aarch64-neonfma-cortex-a73.S",
-    "src/f32-gemm/gen/6x8-minmax-aarch64-neonfma-cortex-a57.S",
-    "src/f32-gemm/gen/6x8-minmax-aarch64-neonfma-cortex-a75.S",
-    "src/f32-gemm/gen/6x8-minmax-aarch64-neonfma-ld128.S",
-    "src/f32-gemm/gen/6x8-minmax-aarch64-neonfma-ld64.S",
-    "src/f32-gemm/gen-inc/1x8inc-minmax-aarch64-neonfma-ld64.S",
-    "src/f32-gemm/gen-inc/1x12inc-minmax-aarch64-neonfma-cortex-a53.S",
     "src/f32-gemm/gen-inc/1x8inc-minmax-aarch64-neonfma-cortex-a53.S",
     "src/f32-gemm/gen-inc/1x8inc-minmax-aarch64-neonfma-cortex-a57.S",
     "src/f32-gemm/gen-inc/1x8inc-minmax-aarch64-neonfma-cortex-a75.S",
-    "src/f32-gemm/gen-inc/4x12inc-minmax-aarch64-neonfma-cortex-a53.S",
+    "src/f32-gemm/gen-inc/1x8inc-minmax-aarch64-neonfma-ld64.S",
+    "src/f32-gemm/gen-inc/1x12inc-minmax-aarch64-neonfma-cortex-a53.S",
     "src/f32-gemm/gen-inc/4x8inc-minmax-aarch64-neonfma-cortex-a53.S",
     "src/f32-gemm/gen-inc/4x8inc-minmax-aarch64-neonfma-cortex-a55.S",
     "src/f32-gemm/gen-inc/4x8inc-minmax-aarch64-neonfma-cortex-a57.S",
     "src/f32-gemm/gen-inc/4x8inc-minmax-aarch64-neonfma-cortex-a75.S",
-    "src/f32-gemm/gen-inc/4x8inc-minmax-aarch64-neonfma-ld128.S",
     "src/f32-gemm/gen-inc/4x8inc-minmax-aarch64-neonfma-ld64.S",
+    "src/f32-gemm/gen-inc/4x8inc-minmax-aarch64-neonfma-ld128.S",
+    "src/f32-gemm/gen-inc/4x12inc-minmax-aarch64-neonfma-cortex-a53.S",
     "src/f32-gemm/gen-inc/5x8inc-minmax-aarch64-neonfma-cortex-a57.S",
     "src/f32-gemm/gen-inc/5x8inc-minmax-aarch64-neonfma-cortex-a75.S",
     "src/f32-gemm/gen-inc/6x8inc-minmax-aarch64-neonfma-cortex-a53.S",
     "src/f32-gemm/gen-inc/6x8inc-minmax-aarch64-neonfma-cortex-a55.S",
-    "src/f32-gemm/gen-inc/6x8inc-minmax-aarch64-neonfma-cortex-a73.S",
     "src/f32-gemm/gen-inc/6x8inc-minmax-aarch64-neonfma-cortex-a57.S",
+    "src/f32-gemm/gen-inc/6x8inc-minmax-aarch64-neonfma-cortex-a73.S",
     "src/f32-gemm/gen-inc/6x8inc-minmax-aarch64-neonfma-cortex-a75.S",
-    "src/f32-gemm/gen-inc/6x8inc-minmax-aarch64-neonfma-ld128.S",
     "src/f32-gemm/gen-inc/6x8inc-minmax-aarch64-neonfma-ld64.S",
-    "src/f32-igemm/1x12-minmax-aarch64-neonfma-cortex-a53.S",
+    "src/f32-gemm/gen-inc/6x8inc-minmax-aarch64-neonfma-ld128.S",
+    "src/f32-gemm/gen/1x8-minmax-aarch64-neonfma-cortex-a53.S",
+    "src/f32-gemm/gen/1x8-minmax-aarch64-neonfma-cortex-a57.S",
+    "src/f32-gemm/gen/1x8-minmax-aarch64-neonfma-cortex-a75.S",
+    "src/f32-gemm/gen/1x8-minmax-aarch64-neonfma-ld64.S",
+    "src/f32-gemm/gen/1x12-minmax-aarch64-neonfma-cortex-a53.S",
+    "src/f32-gemm/gen/4x8-minmax-aarch64-neonfma-cortex-a53.S",
+    "src/f32-gemm/gen/4x8-minmax-aarch64-neonfma-cortex-a55.S",
+    "src/f32-gemm/gen/4x8-minmax-aarch64-neonfma-cortex-a57.S",
+    "src/f32-gemm/gen/4x8-minmax-aarch64-neonfma-cortex-a75.S",
+    "src/f32-gemm/gen/4x8-minmax-aarch64-neonfma-ld64.S",
+    "src/f32-gemm/gen/4x8-minmax-aarch64-neonfma-ld128.S",
+    "src/f32-gemm/gen/4x12-minmax-aarch64-neonfma-cortex-a53.S",
+    "src/f32-gemm/gen/5x8-minmax-aarch64-neonfma-cortex-a57.S",
+    "src/f32-gemm/gen/5x8-minmax-aarch64-neonfma-cortex-a75.S",
+    "src/f32-gemm/gen/6x8-minmax-aarch64-neonfma-cortex-a53.S",
+    "src/f32-gemm/gen/6x8-minmax-aarch64-neonfma-cortex-a55.S",
+    "src/f32-gemm/gen/6x8-minmax-aarch64-neonfma-cortex-a57.S",
+    "src/f32-gemm/gen/6x8-minmax-aarch64-neonfma-cortex-a73.S",
+    "src/f32-gemm/gen/6x8-minmax-aarch64-neonfma-cortex-a75.S",
+    "src/f32-gemm/gen/6x8-minmax-aarch64-neonfma-ld64.S",
+    "src/f32-gemm/gen/6x8-minmax-aarch64-neonfma-ld128.S",
     "src/f32-igemm/1x8-minmax-aarch64-neonfma-cortex-a53.S",
-    "src/f32-igemm/gen/1x8-minmax-aarch64-neonfma-cortex-a57.S",
-    "src/f32-igemm/gen/1x8-minmax-aarch64-neonfma-cortex-a75.S",
-    "src/f32-igemm/4x12-minmax-aarch64-neonfma-cortex-a53.S",
+    "src/f32-igemm/1x12-minmax-aarch64-neonfma-cortex-a53.S",
     "src/f32-igemm/4x8-minmax-aarch64-neonfma-cortex-a53.S",
     "src/f32-igemm/4x8-minmax-aarch64-neonfma-cortex-a55.S",
+    "src/f32-igemm/4x12-minmax-aarch64-neonfma-cortex-a53.S",
+    "src/f32-igemm/6x8-minmax-aarch64-neonfma-cortex-a53.S",
+    "src/f32-igemm/6x8-minmax-aarch64-neonfma-cortex-a55.S",
+    "src/f32-igemm/6x8-minmax-aarch64-neonfma-cortex-a73.S",
+    "src/f32-igemm/gen/1x8-minmax-aarch64-neonfma-cortex-a57.S",
+    "src/f32-igemm/gen/1x8-minmax-aarch64-neonfma-cortex-a75.S",
     "src/f32-igemm/gen/4x8-minmax-aarch64-neonfma-cortex-a57.S",
     "src/f32-igemm/gen/4x8-minmax-aarch64-neonfma-cortex-a75.S",
     "src/f32-igemm/gen/5x8-minmax-aarch64-neonfma-cortex-a57.S",
     "src/f32-igemm/gen/5x8-minmax-aarch64-neonfma-cortex-a75.S",
-    "src/f32-igemm/6x8-minmax-aarch64-neonfma-cortex-a53.S",
-    "src/f32-igemm/6x8-minmax-aarch64-neonfma-cortex-a55.S",
-    "src/f32-igemm/6x8-minmax-aarch64-neonfma-cortex-a73.S",
     "src/f32-igemm/gen/6x8-minmax-aarch64-neonfma-cortex-a57.S",
     "src/f32-igemm/gen/6x8-minmax-aarch64-neonfma-cortex-a75.S",
     "src/qs8-gemm/1x16c4-aarch64-neondot-ld64.S",
@@ -2886,9 +2886,9 @@
     "src/xnnpack/common.h",
     "src/xnnpack/conv.h",
     "src/xnnpack/dwconv.h",
+    "src/xnnpack/fill.h",
     "src/xnnpack/gavgpool.h",
     "src/xnnpack/gemm.h",
-    "src/xnnpack/fill.h",
     "src/xnnpack/hswish.h",
     "src/xnnpack/ibilinear.h",
     "src/xnnpack/igemm.h",
@@ -2953,14 +2953,14 @@
 ]
 
 OPERATOR_TEST_PARAMS_HDRS = [
-    "src/xnnpack/params.h",
     "src/xnnpack/common.h",
+    "src/xnnpack/params.h",
 ]
 
 WEIGHTS_PACK_HDRS = [
-    "src/xnnpack/pack.h",
-    "src/xnnpack/operator.h",
     "src/xnnpack/compute.h",
+    "src/xnnpack/operator.h",
+    "src/xnnpack/pack.h",
 ]
 
 LOGGING_COPTS = select({
@@ -4313,8 +4313,8 @@
     name = "qs8_requantization_bench",
     srcs = [
         "bench/qs8-requantization.cc",
-        "src/xnnpack/requantization-stubs.h",
         "src/xnnpack/AlignedAllocator.h",
+        "src/xnnpack/requantization-stubs.h",
     ] + MICROKERNEL_BENCHMARK_HDRS,
     deps = MICROKERNEL_BENCHMARK_DEPS,
 )
@@ -4334,8 +4334,8 @@
     name = "qu8_requantization_bench",
     srcs = [
         "bench/qu8-requantization.cc",
-        "src/xnnpack/requantization-stubs.h",
         "src/xnnpack/AlignedAllocator.h",
+        "src/xnnpack/requantization-stubs.h",
     ] + MICROKERNEL_BENCHMARK_HDRS,
     deps = MICROKERNEL_BENCHMARK_DEPS,
 )
@@ -4599,8 +4599,8 @@
     name = "rounding_bench",
     srcs = [
         "bench/rounding.cc",
-        "src/xnnpack/math-stubs.h",
         "src/xnnpack/AlignedAllocator.h",
+        "src/xnnpack/math-stubs.h",
     ] + MICROKERNEL_BENCHMARK_HDRS,
     deps = MICROKERNEL_BENCHMARK_DEPS,
 )
diff --git a/CMakeLists.txt b/CMakeLists.txt
index 80a013c..4e3b06d 100755
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@@ -190,8 +190,8 @@
   src/subgraph/deconvolution-2d.c
   src/subgraph/depthwise-convolution-2d.c
   src/subgraph/divide.c
-  src/subgraph/fully-connected.c
   src/subgraph/floor.c
+  src/subgraph/fully-connected.c
   src/subgraph/global-average-pooling-2d.c
   src/subgraph/hardswish.c
   src/subgraph/leaky-relu.c
@@ -203,12 +203,12 @@
   src/subgraph/prelu.c
   src/subgraph/sigmoid.c
   src/subgraph/softmax.c
+  src/subgraph/square-root.c
+  src/subgraph/square.c
+  src/subgraph/squared-difference.c
   src/subgraph/static-constant-pad.c
   src/subgraph/static-reshape.c
   src/subgraph/static-resize-bilinear-2d.c
-  src/subgraph/square.c
-  src/subgraph/square-root.c
-  src/subgraph/squared-difference.c
   src/subgraph/subtract.c
   src/subgraph/unpooling-2d.c)
 
@@ -219,8 +219,8 @@
 SET(XNNPACK_COLD_SRCS ${XNNPACK_OPERATOR_SRCS} ${XNNPACK_SUBGRAPH_SRCS} ${XNNPACK_LOGGING_SRCS})
 LIST(APPEND XNNPACK_COLD_SRCS
   src/init.c
-  src/memory.c
   src/memory-planner.c
+  src/memory.c
   src/operator-delete.c
   src/runtime.c
   src/subgraph.c
@@ -228,8 +228,8 @@
 
 SET(XNNPACK_HOT_SRCS
   src/indirection.c
-  src/packing.c
-  src/operator-run.c)
+  src/operator-run.c
+  src/packing.c)
 
 SET(XNNPACK_TABLE_SRCS
   src/tables/exp2-k-over-64.c
@@ -246,79 +246,76 @@
   src/f32-clamp/gen/scalar-x1.c
   src/f32-clamp/gen/scalar-x2.c
   src/f32-clamp/gen/scalar-x4.c
-  src/f32-conv-hwc/3x3s2p1c3x4-scalar-1x1.c
   src/f32-conv-hwc/3x3s2p0p1c3x4-scalar-1x1.c
+  src/f32-conv-hwc/3x3s2p1c3x4-scalar-1x1.c
   src/f32-conv-hwc2chw/3x3s2p1c3x4-scalar-1x1.c
   src/f32-dwconv-chw/3x3p1-scalar-1x1-acc3.c
   src/f32-dwconv-chw/3x3s2p1-scalar-1x1-acc3.c
   src/f32-dwconv-chw/5x5p2-scalar-1x1-acc5.c
   src/f32-dwconv-chw/5x5s2p2-scalar-1x1-acc5.c
-  src/f32-dwconv/gen/up1x4-scalar.c
-  src/f32-dwconv/gen/up1x4-scalar-acc2.c
-  src/f32-dwconv/gen/up1x9-scalar.c
-  src/f32-dwconv/gen/up1x9-scalar-acc2.c
-  src/f32-dwconv/gen/up1x25-scalar.c
-  src/f32-dwconv/gen/up1x25-scalar-acc2.c
-  src/f32-dwconv/gen/up2x4-scalar.c
-  src/f32-dwconv/gen/up2x4-scalar-acc2.c
-  src/f32-dwconv/gen/up2x9-scalar.c
-  src/f32-dwconv/gen/up2x9-scalar-acc2.c
-  src/f32-dwconv/gen/up2x25-scalar.c
-  src/f32-dwconv/gen/up2x25-scalar-acc2.c
-  src/f32-dwconv/gen/up1x4-minmax-scalar.c
   src/f32-dwconv/gen/up1x4-minmax-scalar-acc2.c
-  src/f32-dwconv/gen/up1x9-minmax-scalar.c
+  src/f32-dwconv/gen/up1x4-minmax-scalar.c
+  src/f32-dwconv/gen/up1x4-scalar-acc2.c
+  src/f32-dwconv/gen/up1x4-scalar.c
   src/f32-dwconv/gen/up1x9-minmax-scalar-acc2.c
-  src/f32-dwconv/gen/up1x25-minmax-scalar.c
+  src/f32-dwconv/gen/up1x9-minmax-scalar.c
+  src/f32-dwconv/gen/up1x9-scalar-acc2.c
+  src/f32-dwconv/gen/up1x9-scalar.c
   src/f32-dwconv/gen/up1x25-minmax-scalar-acc2.c
-  src/f32-dwconv/gen/up2x4-minmax-scalar.c
+  src/f32-dwconv/gen/up1x25-minmax-scalar.c
+  src/f32-dwconv/gen/up1x25-scalar-acc2.c
+  src/f32-dwconv/gen/up1x25-scalar.c
   src/f32-dwconv/gen/up2x4-minmax-scalar-acc2.c
-  src/f32-dwconv/gen/up2x9-minmax-scalar.c
+  src/f32-dwconv/gen/up2x4-minmax-scalar.c
+  src/f32-dwconv/gen/up2x4-scalar-acc2.c
+  src/f32-dwconv/gen/up2x4-scalar.c
   src/f32-dwconv/gen/up2x9-minmax-scalar-acc2.c
-  src/f32-dwconv/gen/up2x25-minmax-scalar.c
+  src/f32-dwconv/gen/up2x9-minmax-scalar.c
+  src/f32-dwconv/gen/up2x9-scalar-acc2.c
+  src/f32-dwconv/gen/up2x9-scalar.c
   src/f32-dwconv/gen/up2x25-minmax-scalar-acc2.c
+  src/f32-dwconv/gen/up2x25-minmax-scalar.c
+  src/f32-dwconv/gen/up2x25-scalar-acc2.c
+  src/f32-dwconv/gen/up2x25-scalar.c
   src/f32-gavgpool-cw/scalar-x1.c
   src/f32-gavgpool/7p7x-minmax-scalar-c1.c
   src/f32-gavgpool/7x-minmax-scalar-c1.c
-  src/f32-gemm/gen/1x4-scalar.c
-  src/f32-gemm/gen/2x4-scalar.c
-  src/f32-gemm/gen/4x2-scalar.c
-  src/f32-gemm/gen/4x4-scalar.c
-  src/f32-gemm/gen/1x4-relu-scalar.c
-  src/f32-gemm/gen/2x4-relu-scalar.c
-  src/f32-gemm/gen/4x2-relu-scalar.c
-  src/f32-gemm/gen/4x4-relu-scalar.c
-  src/f32-gemm/gen/1x4-minmax-scalar.c
-  src/f32-gemm/gen/2x4-minmax-scalar.c
-  src/f32-gemm/gen/4x2-minmax-scalar.c
-  src/f32-gemm/gen/4x4-minmax-scalar.c
   src/f32-gemm/gen-inc/1x4inc-minmax-scalar.c
   src/f32-gemm/gen-inc/2x4inc-minmax-scalar.c
   src/f32-gemm/gen-inc/4x4inc-minmax-scalar.c
+  src/f32-gemm/gen/1x4-minmax-scalar.c
+  src/f32-gemm/gen/1x4-relu-scalar.c
+  src/f32-gemm/gen/1x4-scalar.c
+  src/f32-gemm/gen/2x4-minmax-scalar.c
+  src/f32-gemm/gen/2x4-relu-scalar.c
+  src/f32-gemm/gen/2x4-scalar.c
+  src/f32-gemm/gen/4x2-minmax-scalar.c
+  src/f32-gemm/gen/4x2-relu-scalar.c
+  src/f32-gemm/gen/4x2-scalar.c
+  src/f32-gemm/gen/4x4-minmax-scalar.c
+  src/f32-gemm/gen/4x4-relu-scalar.c
+  src/f32-gemm/gen/4x4-scalar.c
   src/f32-hswish/gen/hswish-scalar-x1.c
   src/f32-hswish/gen/hswish-scalar-x2.c
   src/f32-hswish/gen/hswish-scalar-x4.c
-  src/f32-ibilinear/gen/scalar-c1.c
-  src/f32-ibilinear/gen/scalar-c2.c
-  src/f32-ibilinear/gen/scalar-c4.c
   src/f32-ibilinear-chw/gen/scalar-p1.c
   src/f32-ibilinear-chw/gen/scalar-p2.c
   src/f32-ibilinear-chw/gen/scalar-p4.c
-  src/f32-igemm/gen/1x4-scalar.c
-  src/f32-igemm/gen/2x4-scalar.c
-  src/f32-igemm/gen/4x2-scalar.c
-  src/f32-igemm/gen/4x4-scalar.c
-  src/f32-igemm/gen/1x4-relu-scalar.c
-  src/f32-igemm/gen/2x4-relu-scalar.c
-  src/f32-igemm/gen/4x2-relu-scalar.c
-  src/f32-igemm/gen/4x4-relu-scalar.c
+  src/f32-ibilinear/gen/scalar-c1.c
+  src/f32-ibilinear/gen/scalar-c2.c
+  src/f32-ibilinear/gen/scalar-c4.c
   src/f32-igemm/gen/1x4-minmax-scalar.c
+  src/f32-igemm/gen/1x4-relu-scalar.c
+  src/f32-igemm/gen/1x4-scalar.c
   src/f32-igemm/gen/2x4-minmax-scalar.c
+  src/f32-igemm/gen/2x4-relu-scalar.c
+  src/f32-igemm/gen/2x4-scalar.c
   src/f32-igemm/gen/4x2-minmax-scalar.c
+  src/f32-igemm/gen/4x2-relu-scalar.c
+  src/f32-igemm/gen/4x2-scalar.c
   src/f32-igemm/gen/4x4-minmax-scalar.c
-  src/f32-vlrelu/gen/vlrelu-scalar-x1.c
-  src/f32-vlrelu/gen/vlrelu-scalar-x2.c
-  src/f32-vlrelu/gen/vlrelu-scalar-x4.c
+  src/f32-igemm/gen/4x4-relu-scalar.c
+  src/f32-igemm/gen/4x4-scalar.c
   src/f32-maxpool/9p8x-minmax-scalar-c1.c
   src/f32-pavgpool/9p8x-minmax-scalar-c1.c
   src/f32-pavgpool/9x-minmax-scalar-c1.c
@@ -329,28 +326,28 @@
   src/f32-prelu/gen/scalar-2x1.c
   src/f32-prelu/gen/scalar-2x4.c
   src/f32-raddstoreexpminusmax/gen/scalar-lut64-p2-x1.c
-  src/f32-raddstoreexpminusmax/gen/scalar-lut64-p2-x2.c
   src/f32-raddstoreexpminusmax/gen/scalar-lut64-p2-x2-acc2.c
-  src/f32-raddstoreexpminusmax/gen/scalar-lut64-p2-x4.c
+  src/f32-raddstoreexpminusmax/gen/scalar-lut64-p2-x2.c
   src/f32-raddstoreexpminusmax/gen/scalar-lut64-p2-x4-acc2.c
   src/f32-raddstoreexpminusmax/gen/scalar-lut64-p2-x4-acc4.c
+  src/f32-raddstoreexpminusmax/gen/scalar-lut64-p2-x4.c
   src/f32-raddstoreexpminusmax/gen/scalar-p5-x1.c
-  src/f32-raddstoreexpminusmax/gen/scalar-p5-x2.c
   src/f32-raddstoreexpminusmax/gen/scalar-p5-x2-acc2.c
-  src/f32-raddstoreexpminusmax/gen/scalar-p5-x4.c
+  src/f32-raddstoreexpminusmax/gen/scalar-p5-x2.c
   src/f32-raddstoreexpminusmax/gen/scalar-p5-x4-acc2.c
   src/f32-raddstoreexpminusmax/gen/scalar-p5-x4-acc4.c
-  src/f32-rmax/scalar.c
+  src/f32-raddstoreexpminusmax/gen/scalar-p5-x4.c
   src/f32-relu/gen/scalar-x1.c
   src/f32-relu/gen/scalar-x2.c
   src/f32-relu/gen/scalar-x4.c
   src/f32-relu/gen/scalar-x8.c
-  src/f32-sigmoid/gen/scalar-lut2048-p1-div-x1.c
-  src/f32-sigmoid/gen/scalar-lut2048-p1-div-x2.c
-  src/f32-sigmoid/gen/scalar-lut2048-p1-div-x4.c
+  src/f32-rmax/scalar.c
   src/f32-sigmoid/gen/scalar-lut64-p2-div-x1.c
   src/f32-sigmoid/gen/scalar-lut64-p2-div-x2.c
   src/f32-sigmoid/gen/scalar-lut64-p2-div-x4.c
+  src/f32-sigmoid/gen/scalar-lut2048-p1-div-x1.c
+  src/f32-sigmoid/gen/scalar-lut2048-p1-div-x2.c
+  src/f32-sigmoid/gen/scalar-lut2048-p1-div-x4.c
   src/f32-sigmoid/gen/scalar-p5-div-x1.c
   src/f32-sigmoid/gen/scalar-p5-div-x2.c
   src/f32-sigmoid/gen/scalar-p5-div-x4.c
@@ -508,6 +505,9 @@
   src/f32-vbinary/gen/vsubc-scalar-x2.c
   src/f32-vbinary/gen/vsubc-scalar-x4.c
   src/f32-vbinary/gen/vsubc-scalar-x8.c
+  src/f32-vlrelu/gen/vlrelu-scalar-x1.c
+  src/f32-vlrelu/gen/vlrelu-scalar-x2.c
+  src/f32-vlrelu/gen/vlrelu-scalar-x4.c
   src/f32-vmulcaddc/gen/c1-minmax-scalar-2x.c
   src/f32-vmulcaddc/gen/c2-minmax-scalar-2x.c
   src/f32-vmulcaddc/gen/c4-minmax-scalar-2x.c
@@ -535,23 +535,23 @@
   src/f32-vunary/gen/vsqr-scalar-x1.c
   src/f32-vunary/gen/vsqr-scalar-x2.c
   src/f32-vunary/gen/vsqr-scalar-x4.c
-  src/math/expminus-scalar-lut2048-p1.c
   src/math/expminus-scalar-lut64-p2.c
+  src/math/expminus-scalar-lut2048-p1.c
   src/math/expminus-scalar-p5.c
-  src/math/roundne-scalar-addsub.c
-  src/math/roundne-scalar-nearbyint.c
-  src/math/roundne-scalar-rint.c
   src/math/roundd-scalar-addsub.c
   src/math/roundd-scalar-cvt.c
   src/math/roundd-scalar-floor.c
+  src/math/roundne-scalar-addsub.c
+  src/math/roundne-scalar-nearbyint.c
+  src/math/roundne-scalar-rint.c
   src/math/roundu-scalar-addsub.c
   src/math/roundu-scalar-ceil.c
   src/math/roundu-scalar-cvt.c
   src/math/roundz-scalar-addsub.c
   src/math/roundz-scalar-cvt.c
   src/math/roundz-scalar-trunc.c
-  src/math/sigmoid-scalar-rr2-lut2048-p1-div.c
   src/math/sigmoid-scalar-rr2-lut64-p2-div.c
+  src/math/sigmoid-scalar-rr2-lut2048-p1-div.c
   src/math/sigmoid-scalar-rr2-p5-div.c
   src/qs8-gemm/gen/8x8c4-minmax-scalar.c
   src/qs8-gemm/gen/12x4c4-minmax-scalar.c
@@ -566,9 +566,9 @@
   src/qu8-dwconv/up1x9-minmax-scalar.c
   src/qu8-gavgpool/7p7x-minmax-scalar-c1.c
   src/qu8-gavgpool/7x-minmax-scalar-c1.c
+  src/qu8-gemm/2x2-minmax-scalar.c
   src/qu8-gemm/gen/8x8c4-minmax-scalar.c
   src/qu8-gemm/gen/12x4c4-minmax-scalar.c
-  src/qu8-gemm/2x2-minmax-scalar.c
   src/qu8-igemm/2x2-minmax-scalar.c
   src/qu8-requantization/fp32-scalar-lrintf.c
   src/qu8-requantization/fp32-scalar-magic.c
@@ -581,6 +581,11 @@
   src/u8-lut32norm/scalar.c
   src/u8-maxpool/9p8x-minmax-scalar-c1.c
   src/u8-rmax/scalar.c
+  src/x8-lut/scalar.c
+  src/x8-zip/x2-scalar.c
+  src/x8-zip/x3-scalar.c
+  src/x8-zip/x4-scalar.c
+  src/x8-zip/xm-scalar.c
   src/x32-fill/scalar-float.c
   src/x32-fill/scalar-int.c
   src/x32-packx/x2-scalar.c
@@ -592,12 +597,7 @@
   src/x32-zip/x2-scalar.c
   src/x32-zip/x3-scalar.c
   src/x32-zip/x4-scalar.c
-  src/x32-zip/xm-scalar.c
-  src/x8-lut/scalar.c
-  src/x8-zip/x2-scalar.c
-  src/x8-zip/x3-scalar.c
-  src/x8-zip/x4-scalar.c
-  src/x8-zip/xm-scalar.c)
+  src/x32-zip/xm-scalar.c)
 
 SET(XNNPACK_PSIMD_FASTMATH_MICROKERNEL_SRCS
   src/f32-conv-hwc2chw/3x3s2p1c3x4-psimd-2x2.c
@@ -621,82 +621,80 @@
   src/f32-clamp/gen/neon-x4.c
   src/f32-clamp/gen/neon-x8.c
   src/f32-conv-hwc/gen/3x3s2p0p1c3x4-neon-2x1.c
-  src/f32-conv-hwc/gen/3x3s2p0p1c3x8-neon-2x1.c
   src/f32-conv-hwc/gen/3x3s2p0p1c3x4-neon-2x2.c
+  src/f32-conv-hwc/gen/3x3s2p0p1c3x8-neon-2x1.c
   src/f32-conv-hwc/gen/3x3s2p0p1c3x8-neon-2x2.c
   src/f32-conv-hwc/gen/3x3s2p1c3x4-neon-2x1.c
-  src/f32-conv-hwc/gen/3x3s2p1c3x8-neon-2x1.c
   src/f32-conv-hwc/gen/3x3s2p1c3x4-neon-2x2.c
+  src/f32-conv-hwc/gen/3x3s2p1c3x8-neon-2x1.c
   src/f32-conv-hwc/gen/3x3s2p1c3x8-neon-2x2.c
-  src/f32-dwconv/gen/up4x4-minmax-neon.c
   src/f32-dwconv/gen/up4x4-minmax-neon-acc2.c
-  src/f32-dwconv/gen/up8x4-minmax-neon.c
-  src/f32-dwconv/gen/up8x4-minmax-neon-acc2.c
-  src/f32-dwconv/gen/up4x9-minmax-neon.c
+  src/f32-dwconv/gen/up4x4-minmax-neon.c
   src/f32-dwconv/gen/up4x9-minmax-neon-acc2.c
-  src/f32-dwconv/gen/up8x9-minmax-neon.c
-  src/f32-dwconv/gen/up8x9-minmax-neon-acc2.c
-  src/f32-dwconv/gen/up4x25-minmax-neon.c
+  src/f32-dwconv/gen/up4x9-minmax-neon.c
   src/f32-dwconv/gen/up4x25-minmax-neon-acc2.c
-  src/f32-dwconv/gen/up8x25-minmax-neon.c
+  src/f32-dwconv/gen/up4x25-minmax-neon.c
+  src/f32-dwconv/gen/up8x4-minmax-neon-acc2.c
+  src/f32-dwconv/gen/up8x4-minmax-neon.c
+  src/f32-dwconv/gen/up8x9-minmax-neon-acc2.c
+  src/f32-dwconv/gen/up8x9-minmax-neon.c
   src/f32-dwconv/gen/up8x25-minmax-neon-acc2.c
+  src/f32-dwconv/gen/up8x25-minmax-neon.c
   src/f32-gavgpool-cw/neon-x4.c
   src/f32-gavgpool/7p7x-minmax-neon-c4.c
   src/f32-gavgpool/7x-minmax-neon-c4.c
-  src/f32-gemm/gen/1x8-minmax-neon-lane-ld64.c
-  src/f32-gemm/gen/4x2-minmax-neon-lane-ld64.c
-  src/f32-gemm/gen/4x8-minmax-neon-lane-ld128.c
-  src/f32-gemm/gen/4x8-minmax-neon-lane-ld64.c
-  src/f32-gemm/gen/5x8-minmax-neon-lane-ld64.c
-  src/f32-gemm/gen/6x8-minmax-neon-lane-ld64.c
-  src/f32-gemm/gen/6x8-minmax-neon-lane-ld128.c
-  src/f32-gemm/gen/1x8-minmax-neon-dup-ld64.c
-  src/f32-gemm/gen/4x8-minmax-neon-dup-ld128.c
-  src/f32-gemm/gen/4x8-minmax-neon-dup-ld64.c
-  src/f32-gemm/gen/6x8-minmax-neon-dup-ld64.c
-  src/f32-gemm/gen/6x8-minmax-neon-dup-ld128.c
-  src/f32-gemm/gen/1x8s4-minmax-neon.c
-  src/f32-gemm/gen/4x8s4-minmax-neon.c
-  src/f32-gemm/gen/6x8s4-minmax-neon.c
-  src/f32-gemm/gen/8x8s4-minmax-neon.c
-  src/f32-gemm/gen-inc/1x8inc-minmax-neon-lane-ld64.c
-  src/f32-gemm/gen-inc/4x8inc-minmax-neon-lane-ld128.c
-  src/f32-gemm/gen-inc/4x8inc-minmax-neon-lane-ld64.c
-  src/f32-gemm/gen-inc/5x8inc-minmax-neon-lane-ld64.c
-  src/f32-gemm/gen-inc/6x8inc-minmax-neon-lane-ld64.c
-  src/f32-gemm/gen-inc/6x8inc-minmax-neon-lane-ld128.c
   src/f32-gemm/gen-inc/1x8inc-minmax-neon-dup-ld64.c
-  src/f32-gemm/gen-inc/4x8inc-minmax-neon-dup-ld128.c
+  src/f32-gemm/gen-inc/1x8inc-minmax-neon-lane-ld64.c
+  src/f32-gemm/gen-inc/1x8s4inc-minmax-neon.c
   src/f32-gemm/gen-inc/4x8inc-minmax-neon-dup-ld64.c
+  src/f32-gemm/gen-inc/4x8inc-minmax-neon-dup-ld128.c
+  src/f32-gemm/gen-inc/4x8inc-minmax-neon-lane-ld64.c
+  src/f32-gemm/gen-inc/4x8inc-minmax-neon-lane-ld128.c
+  src/f32-gemm/gen-inc/4x8s4inc-minmax-neon.c
+  src/f32-gemm/gen-inc/5x8inc-minmax-neon-lane-ld64.c
   src/f32-gemm/gen-inc/6x8inc-minmax-neon-dup-ld64.c
   src/f32-gemm/gen-inc/6x8inc-minmax-neon-dup-ld128.c
-  src/f32-gemm/gen-inc/1x8s4inc-minmax-neon.c
-  src/f32-gemm/gen-inc/4x8s4inc-minmax-neon.c
+  src/f32-gemm/gen-inc/6x8inc-minmax-neon-lane-ld64.c
+  src/f32-gemm/gen-inc/6x8inc-minmax-neon-lane-ld128.c
   src/f32-gemm/gen-inc/6x8s4inc-minmax-neon.c
   src/f32-gemm/gen-inc/8x8s4inc-minmax-neon.c
+  src/f32-gemm/gen/1x8-minmax-neon-dup-ld64.c
+  src/f32-gemm/gen/1x8-minmax-neon-lane-ld64.c
+  src/f32-gemm/gen/1x8s4-minmax-neon.c
+  src/f32-gemm/gen/4x2-minmax-neon-lane-ld64.c
+  src/f32-gemm/gen/4x8-minmax-neon-dup-ld64.c
+  src/f32-gemm/gen/4x8-minmax-neon-dup-ld128.c
+  src/f32-gemm/gen/4x8-minmax-neon-lane-ld64.c
+  src/f32-gemm/gen/4x8-minmax-neon-lane-ld128.c
+  src/f32-gemm/gen/4x8s4-minmax-neon.c
+  src/f32-gemm/gen/5x8-minmax-neon-lane-ld64.c
+  src/f32-gemm/gen/6x8-minmax-neon-dup-ld64.c
+  src/f32-gemm/gen/6x8-minmax-neon-dup-ld128.c
+  src/f32-gemm/gen/6x8-minmax-neon-lane-ld64.c
+  src/f32-gemm/gen/6x8-minmax-neon-lane-ld128.c
+  src/f32-gemm/gen/6x8s4-minmax-neon.c
+  src/f32-gemm/gen/8x8s4-minmax-neon.c
   src/f32-hswish/gen/hswish-neon-x4.c
   src/f32-hswish/gen/hswish-neon-x8.c
   src/f32-hswish/gen/hswish-neon-x16.c
   src/f32-ibilinear/gen/neon-c4.c
   src/f32-ibilinear/gen/neon-c8.c
+  src/f32-igemm/gen/1x8-minmax-neon-dup-ld64.c
   src/f32-igemm/gen/1x8-minmax-neon-lane-ld64.c
+  src/f32-igemm/gen/1x8s4-minmax-neon.c
   src/f32-igemm/gen/4x2-minmax-neon-lane-ld64.c
   src/f32-igemm/gen/4x4-minmax-neon-lane-ld64.c
-  src/f32-igemm/gen/4x8-minmax-neon-lane-ld128.c
-  src/f32-igemm/gen/4x8-minmax-neon-lane-ld64.c
-  src/f32-igemm/gen/6x8-minmax-neon-lane-ld64.c
-  src/f32-igemm/gen/6x8-minmax-neon-lane-ld128.c
-  src/f32-igemm/gen/1x8-minmax-neon-dup-ld64.c
-  src/f32-igemm/gen/4x8-minmax-neon-dup-ld128.c
   src/f32-igemm/gen/4x8-minmax-neon-dup-ld64.c
+  src/f32-igemm/gen/4x8-minmax-neon-dup-ld128.c
+  src/f32-igemm/gen/4x8-minmax-neon-lane-ld64.c
+  src/f32-igemm/gen/4x8-minmax-neon-lane-ld128.c
+  src/f32-igemm/gen/4x8s4-minmax-neon.c
   src/f32-igemm/gen/6x8-minmax-neon-dup-ld64.c
   src/f32-igemm/gen/6x8-minmax-neon-dup-ld128.c
-  src/f32-igemm/gen/1x8s4-minmax-neon.c
-  src/f32-igemm/gen/4x8s4-minmax-neon.c
+  src/f32-igemm/gen/6x8-minmax-neon-lane-ld64.c
+  src/f32-igemm/gen/6x8-minmax-neon-lane-ld128.c
   src/f32-igemm/gen/6x8s4-minmax-neon.c
   src/f32-igemm/gen/8x8s4-minmax-neon.c
-  src/f32-vlrelu/gen/vlrelu-neon-x4.c
-  src/f32-vlrelu/gen/vlrelu-neon-x8.c
   src/f32-maxpool/9p8x-minmax-neon-c4.c
   src/f32-pavgpool/9p8x-minmax-neon-c4.c
   src/f32-pavgpool/9x-minmax-neon-c4.c
@@ -711,40 +709,34 @@
   src/f32-prelu/gen/neon-4x4.c
   src/f32-prelu/gen/neon-4x8.c
   src/f32-prelu/gen/neon-4x16.c
-  src/f32-raddstoreexpminusmax/gen/neon-p5-x4.c
-  src/f32-raddstoreexpminusmax/gen/neon-p5-x8.c
-  src/f32-raddstoreexpminusmax/gen/neon-p5-x8-acc2.c
-  src/f32-raddstoreexpminusmax/gen/neon-p5-x12.c
-  src/f32-raddstoreexpminusmax/gen/neon-p5-x12-acc2.c
-  src/f32-raddstoreexpminusmax/gen/neon-p5-x12-acc3.c
-  src/f32-raddstoreexpminusmax/gen/neon-p5-x16.c
-  src/f32-raddstoreexpminusmax/gen/neon-p5-x16-acc2.c
-  src/f32-raddstoreexpminusmax/gen/neon-p5-x16-acc4.c
-  src/f32-raddstoreexpminusmax/gen/neon-p5-x20.c
-  src/f32-raddstoreexpminusmax/gen/neon-p5-x20-acc2.c
-  src/f32-raddstoreexpminusmax/gen/neon-p5-x20-acc5.c
   src/f32-raddstoreexpminusmax/gen/neon-lut64-p2-x4.c
-  src/f32-raddstoreexpminusmax/gen/neon-lut64-p2-x8.c
   src/f32-raddstoreexpminusmax/gen/neon-lut64-p2-x8-acc2.c
-  src/f32-raddstoreexpminusmax/gen/neon-lut64-p2-x12.c
+  src/f32-raddstoreexpminusmax/gen/neon-lut64-p2-x8.c
   src/f32-raddstoreexpminusmax/gen/neon-lut64-p2-x12-acc2.c
   src/f32-raddstoreexpminusmax/gen/neon-lut64-p2-x12-acc3.c
-  src/f32-raddstoreexpminusmax/gen/neon-lut64-p2-x16.c
+  src/f32-raddstoreexpminusmax/gen/neon-lut64-p2-x12.c
   src/f32-raddstoreexpminusmax/gen/neon-lut64-p2-x16-acc2.c
   src/f32-raddstoreexpminusmax/gen/neon-lut64-p2-x16-acc4.c
-  src/f32-raddstoreexpminusmax/gen/neon-lut64-p2-x20.c
+  src/f32-raddstoreexpminusmax/gen/neon-lut64-p2-x16.c
   src/f32-raddstoreexpminusmax/gen/neon-lut64-p2-x20-acc2.c
   src/f32-raddstoreexpminusmax/gen/neon-lut64-p2-x20-acc5.c
-  src/f32-rmax/neon.c
+  src/f32-raddstoreexpminusmax/gen/neon-lut64-p2-x20.c
+  src/f32-raddstoreexpminusmax/gen/neon-p5-x4.c
+  src/f32-raddstoreexpminusmax/gen/neon-p5-x8-acc2.c
+  src/f32-raddstoreexpminusmax/gen/neon-p5-x8.c
+  src/f32-raddstoreexpminusmax/gen/neon-p5-x12-acc2.c
+  src/f32-raddstoreexpminusmax/gen/neon-p5-x12-acc3.c
+  src/f32-raddstoreexpminusmax/gen/neon-p5-x12.c
+  src/f32-raddstoreexpminusmax/gen/neon-p5-x16-acc2.c
+  src/f32-raddstoreexpminusmax/gen/neon-p5-x16-acc4.c
+  src/f32-raddstoreexpminusmax/gen/neon-p5-x16.c
+  src/f32-raddstoreexpminusmax/gen/neon-p5-x20-acc2.c
+  src/f32-raddstoreexpminusmax/gen/neon-p5-x20-acc5.c
+  src/f32-raddstoreexpminusmax/gen/neon-p5-x20.c
   src/f32-relu/gen/neon-x4.c
   src/f32-relu/gen/neon-x8.c
+  src/f32-rmax/neon.c
   src/f32-sigmoid/gen/neon-frac-p9-p10-nr1recps-x16.c
-  src/f32-sigmoid/gen/neon-rr2-p5-nr2recps-x4.c
-  src/f32-sigmoid/gen/neon-rr2-p5-nr2recps-x8.c
-  src/f32-sigmoid/gen/neon-rr2-p5-nr2recps-x12.c
-  src/f32-sigmoid/gen/neon-rr2-p5-nr2recps-x16.c
-  src/f32-sigmoid/gen/neon-rr2-p5-nr2recps-x20.c
-  src/f32-sigmoid/gen/neon-rr2-p5-nr2recps-x24.c
   src/f32-sigmoid/gen/neon-rr2-lut64-p2-nr2recps-x4.c
   src/f32-sigmoid/gen/neon-rr2-lut64-p2-nr2recps-x8.c
   src/f32-sigmoid/gen/neon-rr2-lut64-p2-nr2recps-x12.c
@@ -757,6 +749,12 @@
   src/f32-sigmoid/gen/neon-rr2-lut2048-p1-nr2recps-x16.c
   src/f32-sigmoid/gen/neon-rr2-lut2048-p1-nr2recps-x20.c
   src/f32-sigmoid/gen/neon-rr2-lut2048-p1-nr2recps-x24.c
+  src/f32-sigmoid/gen/neon-rr2-p5-nr2recps-x4.c
+  src/f32-sigmoid/gen/neon-rr2-p5-nr2recps-x8.c
+  src/f32-sigmoid/gen/neon-rr2-p5-nr2recps-x12.c
+  src/f32-sigmoid/gen/neon-rr2-p5-nr2recps-x16.c
+  src/f32-sigmoid/gen/neon-rr2-p5-nr2recps-x20.c
+  src/f32-sigmoid/gen/neon-rr2-p5-nr2recps-x24.c
   src/f32-vbinary/gen/vadd-minmax-neon-x4.c
   src/f32-vbinary/gen/vadd-minmax-neon-x8.c
   src/f32-vbinary/gen/vaddc-minmax-neon-x4.c
@@ -783,44 +781,63 @@
   src/f32-vbinary/gen/vsub-minmax-neon-x8.c
   src/f32-vbinary/gen/vsubc-minmax-neon-x4.c
   src/f32-vbinary/gen/vsubc-minmax-neon-x8.c
+  src/f32-vlrelu/gen/vlrelu-neon-x4.c
+  src/f32-vlrelu/gen/vlrelu-neon-x8.c
   src/f32-vmulcaddc/gen/c4-minmax-neon-2x.c
   src/f32-vmulcaddc/gen/c8-minmax-neon-2x.c
-  src/f32-vrnd/gen/vrndne-neon-x4.c
-  src/f32-vrnd/gen/vrndne-neon-x8.c
-  src/f32-vrnd/gen/vrndz-neon-x4.c
-  src/f32-vrnd/gen/vrndz-neon-x8.c
-  src/f32-vrnd/gen/vrndu-neon-x4.c
-  src/f32-vrnd/gen/vrndu-neon-x8.c
   src/f32-vrnd/gen/vrndd-neon-x4.c
   src/f32-vrnd/gen/vrndd-neon-x8.c
+  src/f32-vrnd/gen/vrndne-neon-x4.c
+  src/f32-vrnd/gen/vrndne-neon-x8.c
+  src/f32-vrnd/gen/vrndu-neon-x4.c
+  src/f32-vrnd/gen/vrndu-neon-x8.c
+  src/f32-vrnd/gen/vrndz-neon-x4.c
+  src/f32-vrnd/gen/vrndz-neon-x8.c
   src/f32-vunary/gen/vabs-neon-x4.c
   src/f32-vunary/gen/vabs-neon-x8.c
   src/f32-vunary/gen/vneg-neon-x4.c
   src/f32-vunary/gen/vneg-neon-x8.c
   src/f32-vunary/gen/vsqr-neon-x4.c
   src/f32-vunary/gen/vsqr-neon-x8.c
-  src/qs8-gavgpool/gen/7x-minmax-neon-c8-acc2.c
-  src/qs8-gavgpool/gen/7x-minmax-neon-c16-acc2.c
-  src/qs8-gavgpool/gen/7x-minmax-neon-c24-acc2.c
-  src/qs8-gavgpool/gen/7x-minmax-neon-c32-acc2.c
-  src/qs8-gavgpool/gen/7p7x-minmax-neon-c8-acc2.c
-  src/qs8-gavgpool/gen/7p7x-minmax-neon-c16-acc2.c
-  src/qs8-gavgpool/gen/7p7x-minmax-neon-c24-acc2.c
-  src/qs8-gavgpool/gen/7p7x-minmax-neon-c32-acc2.c
+  src/math/roundd-neon-addsub.c
+  src/math/roundd-neon-cvt.c
+  src/math/roundne-neon-addsub.c
+  src/math/roundu-neon-addsub.c
+  src/math/roundu-neon-cvt.c
+  src/math/roundz-neon-addsub.c
+  src/math/roundz-neon-cvt.c
+  src/math/sigmoid-neon-frac-p9-p10-nr1recps.c
+  src/math/sigmoid-neon-rr1-lut64-p2-nr2recps.c
+  src/math/sigmoid-neon-rr1-lut2048-p1-nr2recps.c
+  src/math/sigmoid-neon-rr1-p5-nr2recps.c
+  src/math/sigmoid-neon-rr2-lut64-p2-nr2recps.c
+  src/math/sigmoid-neon-rr2-lut2048-p1-nr2recps.c
+  src/math/sigmoid-neon-rr2-p5-nr2recps.c
+  src/math/sqrt-neon-nr1rsqrts.c
+  src/math/sqrt-neon-nr2rsqrts.c
+  src/math/sqrt-neon-nr3rsqrts.c
   src/qs8-dwconv/gen/up8x9-minmax-neon-mul16.c
   src/qs8-dwconv/gen/up16x9-minmax-neon-mul16.c
   src/qs8-dwconv/gen/up24x9-minmax-neon-mul16.c
   src/qs8-dwconv/gen/up32x9-minmax-neon-mul16.c
+  src/qs8-gavgpool/gen/7p7x-minmax-neon-c8-acc2.c
+  src/qs8-gavgpool/gen/7p7x-minmax-neon-c16-acc2.c
+  src/qs8-gavgpool/gen/7p7x-minmax-neon-c24-acc2.c
+  src/qs8-gavgpool/gen/7p7x-minmax-neon-c32-acc2.c
+  src/qs8-gavgpool/gen/7x-minmax-neon-c8-acc2.c
+  src/qs8-gavgpool/gen/7x-minmax-neon-c16-acc2.c
+  src/qs8-gavgpool/gen/7x-minmax-neon-c24-acc2.c
+  src/qs8-gavgpool/gen/7x-minmax-neon-c32-acc2.c
   src/qs8-gemm/gen/1x8-minmax-neon-mlal-lane.c
-  src/qs8-gemm/gen/2x8-minmax-neon-mlal-lane.c
   src/qs8-gemm/gen/1x16-minmax-neon-mlal-lane.c
+  src/qs8-gemm/gen/2x8-minmax-neon-mlal-lane.c
   src/qs8-gemm/gen/2x16-minmax-neon-mlal-lane.c
   src/qs8-igemm/gen/1x8-minmax-neon-mlal-lane.c
-  src/qs8-igemm/gen/2x8-minmax-neon-mlal-lane.c
   src/qs8-igemm/gen/1x16-minmax-neon-mlal-lane.c
+  src/qs8-igemm/gen/2x8-minmax-neon-mlal-lane.c
   src/qs8-igemm/gen/2x16-minmax-neon-mlal-lane.c
-  src/qs8-requantization/precise-neon.c
   src/qs8-requantization/fp32-neon.c
+  src/qs8-requantization/precise-neon.c
   src/qs8-requantization/q31-neon.c
   src/qs8-vadd/gen/minmax-neon-ld64-x8.c
   src/qs8-vadd/gen/minmax-neon-ld64-x16.c
@@ -839,13 +856,17 @@
   src/qu8-gemm/8x8-minmax-neon.c
   src/qu8-igemm/4x8-minmax-neon.c
   src/qu8-igemm/8x8-minmax-neon.c
-  src/qu8-requantization/precise-neon.c
   src/qu8-requantization/fp32-neon.c
+  src/qu8-requantization/precise-neon.c
   src/qu8-requantization/q31-neon.c
   src/qu8-vadd/minmax-neon.c
   src/u8-clamp/neon-x64.c
   src/u8-maxpool/9p8x-minmax-neon-c16.c
   src/u8-rmax/neon.c
+  src/x8-zip/x2-neon.c
+  src/x8-zip/x3-neon.c
+  src/x8-zip/x4-neon.c
+  src/x8-zip/xm-neon.c
   src/x32-fill/neon.c
   src/x32-packx/x4-neon-st4.c
   src/x32-pad/neon.c
@@ -853,151 +874,130 @@
   src/x32-zip/x2-neon.c
   src/x32-zip/x3-neon.c
   src/x32-zip/x4-neon.c
-  src/x32-zip/xm-neon.c
-  src/x8-zip/x2-neon.c
-  src/x8-zip/x3-neon.c
-  src/x8-zip/x4-neon.c
-  src/x8-zip/xm-neon.c
-  src/math/roundne-neon-addsub.c
-  src/math/roundd-neon-addsub.c
-  src/math/roundd-neon-cvt.c
-  src/math/roundu-neon-addsub.c
-  src/math/roundu-neon-cvt.c
-  src/math/roundz-neon-addsub.c
-  src/math/roundz-neon-cvt.c
-  src/math/sigmoid-neon-frac-p9-p10-nr1recps.c
-  src/math/sigmoid-neon-rr1-lut2048-p1-nr2recps.c
-  src/math/sigmoid-neon-rr1-lut64-p2-nr2recps.c
-  src/math/sigmoid-neon-rr1-p5-nr2recps.c
-  src/math/sigmoid-neon-rr2-lut2048-p1-nr2recps.c
-  src/math/sigmoid-neon-rr2-lut64-p2-nr2recps.c
-  src/math/sigmoid-neon-rr2-p5-nr2recps.c
-  src/math/sqrt-neon-nr1rsqrts.c
-  src/math/sqrt-neon-nr2rsqrts.c
-  src/math/sqrt-neon-nr3rsqrts.c)
+  src/x32-zip/xm-neon.c)
 
 SET(XNNPACK_NEONFMA_MICROKERNEL_SRCS
+  src/f32-dwconv/gen/up4x4-minmax-neonfma-acc2.c
+  src/f32-dwconv/gen/up4x4-minmax-neonfma.c
+  src/f32-dwconv/gen/up4x9-minmax-neonfma-acc2.c
+  src/f32-dwconv/gen/up4x9-minmax-neonfma.c
+  src/f32-dwconv/gen/up4x25-minmax-neonfma-acc2.c
+  src/f32-dwconv/gen/up4x25-minmax-neonfma.c
+  src/f32-dwconv/gen/up8x4-minmax-neonfma-acc2.c
+  src/f32-dwconv/gen/up8x4-minmax-neonfma.c
+  src/f32-dwconv/gen/up8x9-minmax-neonfma-acc2.c
+  src/f32-dwconv/gen/up8x9-minmax-neonfma.c
+  src/f32-dwconv/gen/up8x25-minmax-neonfma-acc2.c
+  src/f32-dwconv/gen/up8x25-minmax-neonfma.c
+  src/f32-gemm/gen-inc/1x8inc-minmax-neonfma-dup-ld64.c
+  src/f32-gemm/gen-inc/1x8s4inc-minmax-neonfma.c
+  src/f32-gemm/gen-inc/4x8inc-minmax-neonfma-dup-ld64.c
+  src/f32-gemm/gen-inc/4x8inc-minmax-neonfma-dup-ld128.c
+  src/f32-gemm/gen-inc/4x8s4inc-minmax-neonfma.c
+  src/f32-gemm/gen-inc/6x8inc-minmax-neonfma-dup-ld64.c
+  src/f32-gemm/gen-inc/6x8inc-minmax-neonfma-dup-ld128.c
+  src/f32-gemm/gen-inc/6x8s4inc-minmax-neonfma.c
+  src/f32-gemm/gen-inc/8x8s4inc-minmax-neonfma.c
+  src/f32-gemm/gen/1x8-minmax-neonfma-dup-ld64.c
+  src/f32-gemm/gen/1x8s4-minmax-neonfma.c
+  src/f32-gemm/gen/4x8-minmax-neonfma-dup-ld64.c
+  src/f32-gemm/gen/4x8-minmax-neonfma-dup-ld128.c
+  src/f32-gemm/gen/4x8s4-minmax-neonfma.c
+  src/f32-gemm/gen/6x8-minmax-neonfma-dup-ld64.c
+  src/f32-gemm/gen/6x8-minmax-neonfma-dup-ld128.c
+  src/f32-gemm/gen/6x8s4-minmax-neonfma.c
+  src/f32-gemm/gen/8x8s4-minmax-neonfma.c
   src/f32-ibilinear/gen/neonfma-c4.c
   src/f32-ibilinear/gen/neonfma-c8.c
   src/f32-igemm/gen/1x8-minmax-neonfma-dup-ld64.c
-  src/f32-igemm/gen/4x8-minmax-neonfma-dup-ld128.c
+  src/f32-igemm/gen/1x8s4-minmax-neonfma.c
   src/f32-igemm/gen/4x8-minmax-neonfma-dup-ld64.c
+  src/f32-igemm/gen/4x8-minmax-neonfma-dup-ld128.c
+  src/f32-igemm/gen/4x8s4-minmax-neonfma.c
   src/f32-igemm/gen/6x8-minmax-neonfma-dup-ld64.c
   src/f32-igemm/gen/6x8-minmax-neonfma-dup-ld128.c
-  src/f32-igemm/gen/1x8s4-minmax-neonfma.c
-  src/f32-igemm/gen/4x8s4-minmax-neonfma.c
   src/f32-igemm/gen/6x8s4-minmax-neonfma.c
   src/f32-igemm/gen/8x8s4-minmax-neonfma.c
-  src/f32-dwconv/gen/up4x4-minmax-neonfma.c
-  src/f32-dwconv/gen/up4x4-minmax-neonfma-acc2.c
-  src/f32-dwconv/gen/up8x4-minmax-neonfma.c
-  src/f32-dwconv/gen/up8x4-minmax-neonfma-acc2.c
-  src/f32-dwconv/gen/up4x9-minmax-neonfma.c
-  src/f32-dwconv/gen/up4x9-minmax-neonfma-acc2.c
-  src/f32-dwconv/gen/up8x9-minmax-neonfma.c
-  src/f32-dwconv/gen/up8x9-minmax-neonfma-acc2.c
-  src/f32-dwconv/gen/up4x25-minmax-neonfma.c
-  src/f32-dwconv/gen/up4x25-minmax-neonfma-acc2.c
-  src/f32-dwconv/gen/up8x25-minmax-neonfma.c
-  src/f32-dwconv/gen/up8x25-minmax-neonfma-acc2.c
-  src/f32-gemm/gen/1x8-minmax-neonfma-dup-ld64.c
-  src/f32-gemm/gen/4x8-minmax-neonfma-dup-ld128.c
-  src/f32-gemm/gen/4x8-minmax-neonfma-dup-ld64.c
-  src/f32-gemm/gen/6x8-minmax-neonfma-dup-ld64.c
-  src/f32-gemm/gen/6x8-minmax-neonfma-dup-ld128.c
-  src/f32-gemm/gen/1x8s4-minmax-neonfma.c
-  src/f32-gemm/gen/4x8s4-minmax-neonfma.c
-  src/f32-gemm/gen/6x8s4-minmax-neonfma.c
-  src/f32-gemm/gen/8x8s4-minmax-neonfma.c
-  src/f32-gemm/gen-inc/1x8inc-minmax-neonfma-dup-ld64.c
-  src/f32-gemm/gen-inc/4x8inc-minmax-neonfma-dup-ld128.c
-  src/f32-gemm/gen-inc/4x8inc-minmax-neonfma-dup-ld64.c
-  src/f32-gemm/gen-inc/6x8inc-minmax-neonfma-dup-ld64.c
-  src/f32-gemm/gen-inc/6x8inc-minmax-neonfma-dup-ld128.c
-  src/f32-gemm/gen-inc/1x8s4inc-minmax-neonfma.c
-  src/f32-gemm/gen-inc/4x8s4inc-minmax-neonfma.c
-  src/f32-gemm/gen-inc/6x8s4inc-minmax-neonfma.c
-  src/f32-gemm/gen-inc/8x8s4inc-minmax-neonfma.c
   src/f32-ppmm/gen/4x8-minmax-neonfma.c
   src/f32-ppmm/gen/8x8-minmax-neonfma.c
-  src/f32-raddstoreexpminusmax/gen/neonfma-p5-x4.c
-  src/f32-raddstoreexpminusmax/gen/neonfma-p5-x8.c
-  src/f32-raddstoreexpminusmax/gen/neonfma-p5-x8-acc2.c
-  src/f32-raddstoreexpminusmax/gen/neonfma-p5-x12.c
-  src/f32-raddstoreexpminusmax/gen/neonfma-p5-x12-acc2.c
-  src/f32-raddstoreexpminusmax/gen/neonfma-p5-x12-acc3.c
-  src/f32-raddstoreexpminusmax/gen/neonfma-p5-x16.c
-  src/f32-raddstoreexpminusmax/gen/neonfma-p5-x16-acc2.c
-  src/f32-raddstoreexpminusmax/gen/neonfma-p5-x16-acc4.c
-  src/f32-raddstoreexpminusmax/gen/neonfma-p5-x20.c
-  src/f32-raddstoreexpminusmax/gen/neonfma-p5-x20-acc2.c
-  src/f32-raddstoreexpminusmax/gen/neonfma-p5-x20-acc5.c
   src/f32-raddstoreexpminusmax/gen/neonfma-lut64-p2-x4.c
-  src/f32-raddstoreexpminusmax/gen/neonfma-lut64-p2-x8.c
   src/f32-raddstoreexpminusmax/gen/neonfma-lut64-p2-x8-acc2.c
-  src/f32-raddstoreexpminusmax/gen/neonfma-lut64-p2-x12.c
+  src/f32-raddstoreexpminusmax/gen/neonfma-lut64-p2-x8.c
   src/f32-raddstoreexpminusmax/gen/neonfma-lut64-p2-x12-acc2.c
   src/f32-raddstoreexpminusmax/gen/neonfma-lut64-p2-x12-acc3.c
-  src/f32-raddstoreexpminusmax/gen/neonfma-lut64-p2-x16.c
+  src/f32-raddstoreexpminusmax/gen/neonfma-lut64-p2-x12.c
   src/f32-raddstoreexpminusmax/gen/neonfma-lut64-p2-x16-acc2.c
   src/f32-raddstoreexpminusmax/gen/neonfma-lut64-p2-x16-acc4.c
-  src/f32-raddstoreexpminusmax/gen/neonfma-lut64-p2-x20.c
+  src/f32-raddstoreexpminusmax/gen/neonfma-lut64-p2-x16.c
   src/f32-raddstoreexpminusmax/gen/neonfma-lut64-p2-x20-acc2.c
   src/f32-raddstoreexpminusmax/gen/neonfma-lut64-p2-x20-acc5.c
-  src/f32-sigmoid/gen/neonfma-rr1-p5-nr2fma-x4.c
-  src/f32-sigmoid/gen/neonfma-rr1-p5-nr2fma-x8.c
-  src/f32-sigmoid/gen/neonfma-rr1-p5-nr2fma-x12.c
-  src/f32-sigmoid/gen/neonfma-rr1-p5-nr2fma-x16.c
-  src/f32-sigmoid/gen/neonfma-rr1-p5-nr2fma-x20.c
-  src/f32-sigmoid/gen/neonfma-rr1-p5-nr2fma-x24.c
-  src/f32-sigmoid/gen/neonfma-rr1-p5-nr1recps1fma-x4.c
-  src/f32-sigmoid/gen/neonfma-rr1-p5-nr1recps1fma-x8.c
-  src/f32-sigmoid/gen/neonfma-rr1-p5-nr1recps1fma-x12.c
-  src/f32-sigmoid/gen/neonfma-rr1-p5-nr1recps1fma-x16.c
-  src/f32-sigmoid/gen/neonfma-rr1-p5-nr1recps1fma-x20.c
-  src/f32-sigmoid/gen/neonfma-rr1-p5-nr1recps1fma-x24.c
-  src/f32-sigmoid/gen/neonfma-rr1-p5-nr2recps-x4.c
-  src/f32-sigmoid/gen/neonfma-rr1-p5-nr2recps-x8.c
-  src/f32-sigmoid/gen/neonfma-rr1-p5-nr2recps-x12.c
-  src/f32-sigmoid/gen/neonfma-rr1-p5-nr2recps-x16.c
-  src/f32-sigmoid/gen/neonfma-rr1-p5-nr2recps-x20.c
-  src/f32-sigmoid/gen/neonfma-rr1-p5-nr2recps-x24.c
-  src/f32-sigmoid/gen/neonfma-rr1-lut64-p2-nr2fma-x4.c
-  src/f32-sigmoid/gen/neonfma-rr1-lut64-p2-nr2fma-x8.c
-  src/f32-sigmoid/gen/neonfma-rr1-lut64-p2-nr2fma-x12.c
-  src/f32-sigmoid/gen/neonfma-rr1-lut64-p2-nr2fma-x16.c
-  src/f32-sigmoid/gen/neonfma-rr1-lut64-p2-nr2fma-x20.c
-  src/f32-sigmoid/gen/neonfma-rr1-lut64-p2-nr2fma-x24.c
+  src/f32-raddstoreexpminusmax/gen/neonfma-lut64-p2-x20.c
+  src/f32-raddstoreexpminusmax/gen/neonfma-p5-x4.c
+  src/f32-raddstoreexpminusmax/gen/neonfma-p5-x8-acc2.c
+  src/f32-raddstoreexpminusmax/gen/neonfma-p5-x8.c
+  src/f32-raddstoreexpminusmax/gen/neonfma-p5-x12-acc2.c
+  src/f32-raddstoreexpminusmax/gen/neonfma-p5-x12-acc3.c
+  src/f32-raddstoreexpminusmax/gen/neonfma-p5-x12.c
+  src/f32-raddstoreexpminusmax/gen/neonfma-p5-x16-acc2.c
+  src/f32-raddstoreexpminusmax/gen/neonfma-p5-x16-acc4.c
+  src/f32-raddstoreexpminusmax/gen/neonfma-p5-x16.c
+  src/f32-raddstoreexpminusmax/gen/neonfma-p5-x20-acc2.c
+  src/f32-raddstoreexpminusmax/gen/neonfma-p5-x20-acc5.c
+  src/f32-raddstoreexpminusmax/gen/neonfma-p5-x20.c
   src/f32-sigmoid/gen/neonfma-rr1-lut64-p2-nr1recps1fma-x4.c
   src/f32-sigmoid/gen/neonfma-rr1-lut64-p2-nr1recps1fma-x8.c
   src/f32-sigmoid/gen/neonfma-rr1-lut64-p2-nr1recps1fma-x12.c
   src/f32-sigmoid/gen/neonfma-rr1-lut64-p2-nr1recps1fma-x16.c
   src/f32-sigmoid/gen/neonfma-rr1-lut64-p2-nr1recps1fma-x20.c
   src/f32-sigmoid/gen/neonfma-rr1-lut64-p2-nr1recps1fma-x24.c
+  src/f32-sigmoid/gen/neonfma-rr1-lut64-p2-nr2fma-x4.c
+  src/f32-sigmoid/gen/neonfma-rr1-lut64-p2-nr2fma-x8.c
+  src/f32-sigmoid/gen/neonfma-rr1-lut64-p2-nr2fma-x12.c
+  src/f32-sigmoid/gen/neonfma-rr1-lut64-p2-nr2fma-x16.c
+  src/f32-sigmoid/gen/neonfma-rr1-lut64-p2-nr2fma-x20.c
+  src/f32-sigmoid/gen/neonfma-rr1-lut64-p2-nr2fma-x24.c
   src/f32-sigmoid/gen/neonfma-rr1-lut64-p2-nr2recps-x4.c
   src/f32-sigmoid/gen/neonfma-rr1-lut64-p2-nr2recps-x8.c
   src/f32-sigmoid/gen/neonfma-rr1-lut64-p2-nr2recps-x12.c
   src/f32-sigmoid/gen/neonfma-rr1-lut64-p2-nr2recps-x16.c
   src/f32-sigmoid/gen/neonfma-rr1-lut64-p2-nr2recps-x20.c
   src/f32-sigmoid/gen/neonfma-rr1-lut64-p2-nr2recps-x24.c
-  src/f32-sigmoid/gen/neonfma-rr1-lut2048-p1-nr2fma-x4.c
-  src/f32-sigmoid/gen/neonfma-rr1-lut2048-p1-nr2fma-x8.c
-  src/f32-sigmoid/gen/neonfma-rr1-lut2048-p1-nr2fma-x12.c
-  src/f32-sigmoid/gen/neonfma-rr1-lut2048-p1-nr2fma-x16.c
-  src/f32-sigmoid/gen/neonfma-rr1-lut2048-p1-nr2fma-x20.c
-  src/f32-sigmoid/gen/neonfma-rr1-lut2048-p1-nr2fma-x24.c
   src/f32-sigmoid/gen/neonfma-rr1-lut2048-p1-nr1recps1fma-x4.c
   src/f32-sigmoid/gen/neonfma-rr1-lut2048-p1-nr1recps1fma-x8.c
   src/f32-sigmoid/gen/neonfma-rr1-lut2048-p1-nr1recps1fma-x12.c
   src/f32-sigmoid/gen/neonfma-rr1-lut2048-p1-nr1recps1fma-x16.c
   src/f32-sigmoid/gen/neonfma-rr1-lut2048-p1-nr1recps1fma-x20.c
   src/f32-sigmoid/gen/neonfma-rr1-lut2048-p1-nr1recps1fma-x24.c
+  src/f32-sigmoid/gen/neonfma-rr1-lut2048-p1-nr2fma-x4.c
+  src/f32-sigmoid/gen/neonfma-rr1-lut2048-p1-nr2fma-x8.c
+  src/f32-sigmoid/gen/neonfma-rr1-lut2048-p1-nr2fma-x12.c
+  src/f32-sigmoid/gen/neonfma-rr1-lut2048-p1-nr2fma-x16.c
+  src/f32-sigmoid/gen/neonfma-rr1-lut2048-p1-nr2fma-x20.c
+  src/f32-sigmoid/gen/neonfma-rr1-lut2048-p1-nr2fma-x24.c
   src/f32-sigmoid/gen/neonfma-rr1-lut2048-p1-nr2recps-x4.c
   src/f32-sigmoid/gen/neonfma-rr1-lut2048-p1-nr2recps-x8.c
   src/f32-sigmoid/gen/neonfma-rr1-lut2048-p1-nr2recps-x12.c
   src/f32-sigmoid/gen/neonfma-rr1-lut2048-p1-nr2recps-x16.c
   src/f32-sigmoid/gen/neonfma-rr1-lut2048-p1-nr2recps-x20.c
   src/f32-sigmoid/gen/neonfma-rr1-lut2048-p1-nr2recps-x24.c
+  src/f32-sigmoid/gen/neonfma-rr1-p5-nr1recps1fma-x4.c
+  src/f32-sigmoid/gen/neonfma-rr1-p5-nr1recps1fma-x8.c
+  src/f32-sigmoid/gen/neonfma-rr1-p5-nr1recps1fma-x12.c
+  src/f32-sigmoid/gen/neonfma-rr1-p5-nr1recps1fma-x16.c
+  src/f32-sigmoid/gen/neonfma-rr1-p5-nr1recps1fma-x20.c
+  src/f32-sigmoid/gen/neonfma-rr1-p5-nr1recps1fma-x24.c
+  src/f32-sigmoid/gen/neonfma-rr1-p5-nr2fma-x4.c
+  src/f32-sigmoid/gen/neonfma-rr1-p5-nr2fma-x8.c
+  src/f32-sigmoid/gen/neonfma-rr1-p5-nr2fma-x12.c
+  src/f32-sigmoid/gen/neonfma-rr1-p5-nr2fma-x16.c
+  src/f32-sigmoid/gen/neonfma-rr1-p5-nr2fma-x20.c
+  src/f32-sigmoid/gen/neonfma-rr1-p5-nr2fma-x24.c
+  src/f32-sigmoid/gen/neonfma-rr1-p5-nr2recps-x4.c
+  src/f32-sigmoid/gen/neonfma-rr1-p5-nr2recps-x8.c
+  src/f32-sigmoid/gen/neonfma-rr1-p5-nr2recps-x12.c
+  src/f32-sigmoid/gen/neonfma-rr1-p5-nr2recps-x16.c
+  src/f32-sigmoid/gen/neonfma-rr1-p5-nr2recps-x20.c
+  src/f32-sigmoid/gen/neonfma-rr1-p5-nr2recps-x24.c
   src/f32-vmulcaddc/gen/c4-minmax-neonfma-2x.c
   src/f32-vmulcaddc/gen/c8-minmax-neonfma-2x.c
   src/f32-vsqrt/gen/neonfma-nr1rsqrts1fma1adj-x4.c
@@ -1022,93 +1022,81 @@
   src/f32-vsqrt/gen/neonfma-nr2fma1adj-x40.c
   src/math/exp-neonfma-lut64-p2.c
   src/math/exp-neonfma-p5.c
-  src/math/expminus-neonfma-lut2048-p1.c
   src/math/expminus-neonfma-lut64-p2.c
+  src/math/expminus-neonfma-lut2048-p1.c
   src/math/expminus-neonfma-p5.c
-  src/math/sigmoid-neonfma-rr1-lut2048-p1-nr1recps1fma.c
-  src/math/sigmoid-neonfma-rr1-lut2048-p1-nr2fma.c
-  src/math/sigmoid-neonfma-rr1-lut2048-p1-nr2recps.c
   src/math/sigmoid-neonfma-rr1-lut64-p2-nr1recps1fma.c
   src/math/sigmoid-neonfma-rr1-lut64-p2-nr2fma.c
   src/math/sigmoid-neonfma-rr1-lut64-p2-nr2recps.c
+  src/math/sigmoid-neonfma-rr1-lut2048-p1-nr1recps1fma.c
+  src/math/sigmoid-neonfma-rr1-lut2048-p1-nr2fma.c
+  src/math/sigmoid-neonfma-rr1-lut2048-p1-nr2recps.c
   src/math/sigmoid-neonfma-rr1-p5-nr1recps1fma.c
   src/math/sigmoid-neonfma-rr1-p5-nr2fma.c
   src/math/sigmoid-neonfma-rr1-p5-nr2recps.c
-  src/math/sigmoid-neonfma-rr2-lut2048-p1-nr1recps1fma.c
-  src/math/sigmoid-neonfma-rr2-lut2048-p1-nr2fma.c
-  src/math/sigmoid-neonfma-rr2-lut2048-p1-nr2recps.c
   src/math/sigmoid-neonfma-rr2-lut64-p2-nr1recps1fma.c
   src/math/sigmoid-neonfma-rr2-lut64-p2-nr2fma.c
   src/math/sigmoid-neonfma-rr2-lut64-p2-nr2recps.c
+  src/math/sigmoid-neonfma-rr2-lut2048-p1-nr1recps1fma.c
+  src/math/sigmoid-neonfma-rr2-lut2048-p1-nr2fma.c
+  src/math/sigmoid-neonfma-rr2-lut2048-p1-nr2recps.c
   src/math/sigmoid-neonfma-rr2-p5-nr1recps1fma.c
   src/math/sigmoid-neonfma-rr2-p5-nr2fma.c
   src/math/sigmoid-neonfma-rr2-p5-nr2recps.c
   src/math/sqrt-neonfma-nr1fma.c
+  src/math/sqrt-neonfma-nr1rsqrts1fma1adj.c
   src/math/sqrt-neonfma-nr2fma.c
-  src/math/sqrt-neonfma-nr3fma.c
   src/math/sqrt-neonfma-nr2fma1adj.c
-  src/math/sqrt-neonfma-nr1rsqrts1fma1adj.c)
+  src/math/sqrt-neonfma-nr3fma.c)
 
 SET(XNNPACK_NEONV8_MICROKERNEL_SRCS
-  src/f32-vrnd/gen/vrndne-neonv8-x4.c
-  src/f32-vrnd/gen/vrndne-neonv8-x8.c
-  src/f32-vrnd/gen/vrndz-neonv8-x4.c
-  src/f32-vrnd/gen/vrndz-neonv8-x8.c
-  src/f32-vrnd/gen/vrndu-neonv8-x4.c
-  src/f32-vrnd/gen/vrndu-neonv8-x8.c
   src/f32-vrnd/gen/vrndd-neonv8-x4.c
   src/f32-vrnd/gen/vrndd-neonv8-x8.c
-  src/math/roundne-neonv8.c
+  src/f32-vrnd/gen/vrndne-neonv8-x4.c
+  src/f32-vrnd/gen/vrndne-neonv8-x8.c
+  src/f32-vrnd/gen/vrndu-neonv8-x4.c
+  src/f32-vrnd/gen/vrndu-neonv8-x8.c
+  src/f32-vrnd/gen/vrndz-neonv8-x4.c
+  src/f32-vrnd/gen/vrndz-neonv8-x8.c
   src/math/roundd-neonv8.c
+  src/math/roundne-neonv8.c
   src/math/roundu-neonv8.c
   src/math/roundz-neonv8.c)
 
 SET(XNNPACK_AARCH64_NEONFMA_MICROKERNEL_SRCS
-  src/f32-vbinary/gen/vdiv-minmax-neon-x4.c
-  src/f32-vbinary/gen/vdiv-minmax-neon-x8.c
-  src/f32-vbinary/gen/vdivc-minmax-neon-x4.c
-  src/f32-vbinary/gen/vdivc-minmax-neon-x8.c
-  src/f32-vbinary/gen/vrdivc-minmax-neon-x4.c
-  src/f32-vbinary/gen/vrdivc-minmax-neon-x8.c
-  src/f32-gemm/gen/1x8-minmax-neonfma-lane-ld64.c
-  src/f32-gemm/gen/4x2-minmax-neonfma-lane-ld64.c
-  src/f32-gemm/gen/4x8-minmax-neonfma-lane-ld128.c
-  src/f32-gemm/gen/4x8-minmax-neonfma-lane-ld64.c
-  src/f32-gemm/gen/5x8-minmax-neonfma-lane-ld64.c
-  src/f32-gemm/gen/6x8-minmax-neonfma-lane-ld64.c
-  src/f32-gemm/gen/6x8-minmax-neonfma-lane-ld128.c
-  src/f32-gemm/gen-inc/1x8inc-minmax-neonfma-lane-ld64.c
-  src/f32-gemm/gen-inc/4x8inc-minmax-neonfma-lane-ld128.c
-  src/f32-gemm/gen-inc/4x8inc-minmax-neonfma-lane-ld64.c
-  src/f32-gemm/gen-inc/5x8inc-minmax-neonfma-lane-ld64.c
-  src/f32-gemm/gen-inc/6x8inc-minmax-neonfma-lane-ld64.c
-  src/f32-gemm/gen-inc/6x8inc-minmax-neonfma-lane-ld128.c
-  src/f32-igemm/gen/1x8-minmax-neonfma-lane-ld64.c
-  src/f32-igemm/gen/4x2-minmax-neonfma-lane-ld64.c
-  src/f32-igemm/gen/4x4-minmax-neonfma-lane-ld64.c
-  src/f32-igemm/gen/4x8-minmax-neonfma-lane-ld128.c
-  src/f32-igemm/gen/4x8-minmax-neonfma-lane-ld64.c
-  src/f32-igemm/gen/6x8-minmax-neonfma-lane-ld64.c
-  src/f32-igemm/gen/6x8-minmax-neonfma-lane-ld128.c
   src/f32-conv-hwc/gen/3x3s2p0p1c3x4-neonfma-2x1.c
-  src/f32-conv-hwc/gen/3x3s2p0p1c3x8-neonfma-2x1.c
   src/f32-conv-hwc/gen/3x3s2p0p1c3x4-neonfma-2x2.c
+  src/f32-conv-hwc/gen/3x3s2p0p1c3x8-neonfma-2x1.c
   src/f32-conv-hwc/gen/3x3s2p0p1c3x8-neonfma-2x2.c
   src/f32-conv-hwc/gen/3x3s2p1c3x4-neonfma-2x1.c
-  src/f32-conv-hwc/gen/3x3s2p1c3x8-neonfma-2x1.c
   src/f32-conv-hwc/gen/3x3s2p1c3x4-neonfma-2x2.c
+  src/f32-conv-hwc/gen/3x3s2p1c3x8-neonfma-2x1.c
   src/f32-conv-hwc/gen/3x3s2p1c3x8-neonfma-2x2.c
   src/f32-conv-hwc2chw/3x3s2p1c3x4-neonfma-2x2.c
   src/f32-dwconv-chw/3x3p1-neonfma-3x4.c
   src/f32-dwconv-chw/3x3s2p1-neonfma-1x4-acc3.c
   src/f32-dwconv-chw/5x5p2-neonfma-3x4.c
   src/f32-dwconv-chw/5x5s2p2-neonfma-1x4-acc2.c
-  src/f32-sigmoid/gen/neonfma-rr1-p5-div-x4.c
-  src/f32-sigmoid/gen/neonfma-rr1-p5-div-x8.c
-  src/f32-sigmoid/gen/neonfma-rr1-p5-div-x12.c
-  src/f32-sigmoid/gen/neonfma-rr1-p5-div-x16.c
-  src/f32-sigmoid/gen/neonfma-rr1-p5-div-x20.c
-  src/f32-sigmoid/gen/neonfma-rr1-p5-div-x24.c
+  src/f32-gemm/gen-inc/1x8inc-minmax-neonfma-lane-ld64.c
+  src/f32-gemm/gen-inc/4x8inc-minmax-neonfma-lane-ld64.c
+  src/f32-gemm/gen-inc/4x8inc-minmax-neonfma-lane-ld128.c
+  src/f32-gemm/gen-inc/5x8inc-minmax-neonfma-lane-ld64.c
+  src/f32-gemm/gen-inc/6x8inc-minmax-neonfma-lane-ld64.c
+  src/f32-gemm/gen-inc/6x8inc-minmax-neonfma-lane-ld128.c
+  src/f32-gemm/gen/1x8-minmax-neonfma-lane-ld64.c
+  src/f32-gemm/gen/4x2-minmax-neonfma-lane-ld64.c
+  src/f32-gemm/gen/4x8-minmax-neonfma-lane-ld64.c
+  src/f32-gemm/gen/4x8-minmax-neonfma-lane-ld128.c
+  src/f32-gemm/gen/5x8-minmax-neonfma-lane-ld64.c
+  src/f32-gemm/gen/6x8-minmax-neonfma-lane-ld64.c
+  src/f32-gemm/gen/6x8-minmax-neonfma-lane-ld128.c
+  src/f32-igemm/gen/1x8-minmax-neonfma-lane-ld64.c
+  src/f32-igemm/gen/4x2-minmax-neonfma-lane-ld64.c
+  src/f32-igemm/gen/4x4-minmax-neonfma-lane-ld64.c
+  src/f32-igemm/gen/4x8-minmax-neonfma-lane-ld64.c
+  src/f32-igemm/gen/4x8-minmax-neonfma-lane-ld128.c
+  src/f32-igemm/gen/6x8-minmax-neonfma-lane-ld64.c
+  src/f32-igemm/gen/6x8-minmax-neonfma-lane-ld128.c
   src/f32-sigmoid/gen/neonfma-rr1-lut64-p2-div-x4.c
   src/f32-sigmoid/gen/neonfma-rr1-lut64-p2-div-x8.c
   src/f32-sigmoid/gen/neonfma-rr1-lut64-p2-div-x12.c
@@ -1121,14 +1109,12 @@
   src/f32-sigmoid/gen/neonfma-rr1-lut2048-p1-div-x16.c
   src/f32-sigmoid/gen/neonfma-rr1-lut2048-p1-div-x20.c
   src/f32-sigmoid/gen/neonfma-rr1-lut2048-p1-div-x24.c
-  src/f32-spmm/gen/12x1-minmax-neonfma.c
-  src/f32-spmm/gen/12x2-minmax-neonfma.c
-  src/f32-spmm/gen/12x4-minmax-neonfma.c
-  src/f32-spmm/gen/16x1-minmax-neonfma-pipelined.c
-  src/f32-spmm/gen/16x1-minmax-neonfma-unroll2.c
-  src/f32-spmm/gen/16x1-minmax-neonfma.c
-  src/f32-spmm/gen/16x2-minmax-neonfma.c
-  src/f32-spmm/gen/16x4-minmax-neonfma.c
+  src/f32-sigmoid/gen/neonfma-rr1-p5-div-x4.c
+  src/f32-sigmoid/gen/neonfma-rr1-p5-div-x8.c
+  src/f32-sigmoid/gen/neonfma-rr1-p5-div-x12.c
+  src/f32-sigmoid/gen/neonfma-rr1-p5-div-x16.c
+  src/f32-sigmoid/gen/neonfma-rr1-p5-div-x20.c
+  src/f32-sigmoid/gen/neonfma-rr1-p5-div-x24.c
   src/f32-spmm/gen/4x1-minmax-neonfma-pipelined.c
   src/f32-spmm/gen/4x1-minmax-neonfma-unroll2.c
   src/f32-spmm/gen/4x1-minmax-neonfma.c
@@ -1139,120 +1125,134 @@
   src/f32-spmm/gen/8x1-minmax-neonfma.c
   src/f32-spmm/gen/8x2-minmax-neonfma.c
   src/f32-spmm/gen/8x4-minmax-neonfma.c
+  src/f32-spmm/gen/12x1-minmax-neonfma.c
+  src/f32-spmm/gen/12x2-minmax-neonfma.c
+  src/f32-spmm/gen/12x4-minmax-neonfma.c
+  src/f32-spmm/gen/16x1-minmax-neonfma-pipelined.c
+  src/f32-spmm/gen/16x1-minmax-neonfma-unroll2.c
+  src/f32-spmm/gen/16x1-minmax-neonfma.c
+  src/f32-spmm/gen/16x2-minmax-neonfma.c
+  src/f32-spmm/gen/16x4-minmax-neonfma.c
+  src/f32-vbinary/gen/vdiv-minmax-neon-x4.c
+  src/f32-vbinary/gen/vdiv-minmax-neon-x8.c
+  src/f32-vbinary/gen/vdivc-minmax-neon-x4.c
+  src/f32-vbinary/gen/vdivc-minmax-neon-x8.c
+  src/f32-vbinary/gen/vrdivc-minmax-neon-x4.c
+  src/f32-vbinary/gen/vrdivc-minmax-neon-x8.c
   src/f32-vsqrt/gen/neon-sqrt-x4.c
   src/f32-vsqrt/gen/neon-sqrt-x8.c
-  src/math/sigmoid-neonfma-rr1-lut2048-p1-div.c
   src/math/sigmoid-neonfma-rr1-lut64-p2-div.c
+  src/math/sigmoid-neonfma-rr1-lut2048-p1-div.c
   src/math/sigmoid-neonfma-rr1-p5-div.c
-  src/math/sigmoid-neonfma-rr2-lut2048-p1-div.c
   src/math/sigmoid-neonfma-rr2-lut64-p2-div.c
+  src/math/sigmoid-neonfma-rr2-lut2048-p1-div.c
   src/math/sigmoid-neonfma-rr2-p5-div.c)
 
 SET(XNNPACK_AARCH64_NEONFP16ARITH_MICROKERNEL_SRCS
-  src/f16-clamp/gen/neonfp16arith-x16.c
   src/f16-clamp/gen/neonfp16arith-x8.c
-  src/f16-dwconv/gen/up16x25-minmax-neonfp16arith-acc2.c
-  src/f16-dwconv/gen/up16x25-minmax-neonfp16arith.c
-  src/f16-dwconv/gen/up16x4-minmax-neonfp16arith-acc2.c
-  src/f16-dwconv/gen/up16x4-minmax-neonfp16arith.c
-  src/f16-dwconv/gen/up16x9-minmax-neonfp16arith-acc2.c
-  src/f16-dwconv/gen/up16x9-minmax-neonfp16arith.c
-  src/f16-dwconv/gen/up8x25-minmax-neonfp16arith-acc2.c
-  src/f16-dwconv/gen/up8x25-minmax-neonfp16arith.c
+  src/f16-clamp/gen/neonfp16arith-x16.c
   src/f16-dwconv/gen/up8x4-minmax-neonfp16arith-acc2.c
   src/f16-dwconv/gen/up8x4-minmax-neonfp16arith.c
   src/f16-dwconv/gen/up8x9-minmax-neonfp16arith-acc2.c
   src/f16-dwconv/gen/up8x9-minmax-neonfp16arith.c
+  src/f16-dwconv/gen/up8x25-minmax-neonfp16arith-acc2.c
+  src/f16-dwconv/gen/up8x25-minmax-neonfp16arith.c
+  src/f16-dwconv/gen/up16x4-minmax-neonfp16arith-acc2.c
+  src/f16-dwconv/gen/up16x4-minmax-neonfp16arith.c
+  src/f16-dwconv/gen/up16x9-minmax-neonfp16arith-acc2.c
+  src/f16-dwconv/gen/up16x9-minmax-neonfp16arith.c
+  src/f16-dwconv/gen/up16x25-minmax-neonfp16arith-acc2.c
+  src/f16-dwconv/gen/up16x25-minmax-neonfp16arith.c
   src/f16-gavgpool/7p7x-minmax-neonfp16arith-c8.c
   src/f16-gavgpool/7x-minmax-neonfp16arith-c8.c
-  src/f16-hswish/gen/hswish-neonfp16arith-x16.c
-  src/f16-hswish/gen/hswish-neonfp16arith-x8.c
-  src/f16-prelu/gen/neonfp16arith-2x16.c
-  src/f16-prelu/gen/neonfp16arith-2x8.c
-  src/f16-relu/gen/neonfp16arith-x16.c
-  src/f16-relu/gen/neonfp16arith-x8.c
-  src/f16-vbinary/gen/vaddc-minmax-neonfp16arith-x16.c
-  src/f16-vbinary/gen/vaddc-minmax-neonfp16arith-x8.c
-  src/f16-vbinary/gen/vadd-minmax-neonfp16arith-x16.c
-  src/f16-vbinary/gen/vadd-minmax-neonfp16arith-x8.c
-  src/f16-vbinary/gen/vdivc-minmax-neonfp16arith-x16.c
-  src/f16-vbinary/gen/vdivc-minmax-neonfp16arith-x8.c
-  src/f16-vbinary/gen/vdiv-minmax-neonfp16arith-x16.c
-  src/f16-vbinary/gen/vdiv-minmax-neonfp16arith-x8.c
-  src/f16-vbinary/gen/vmaxc-neonfp16arith-x16.c
-  src/f16-vbinary/gen/vmaxc-neonfp16arith-x8.c
-  src/f16-vbinary/gen/vmax-neonfp16arith-x16.c
-  src/f16-vbinary/gen/vmax-neonfp16arith-x8.c
-  src/f16-vbinary/gen/vminc-neonfp16arith-x16.c
-  src/f16-vbinary/gen/vminc-neonfp16arith-x8.c
-  src/f16-vbinary/gen/vmin-neonfp16arith-x16.c
-  src/f16-vbinary/gen/vmin-neonfp16arith-x8.c
-  src/f16-vbinary/gen/vmulc-minmax-neonfp16arith-x16.c
-  src/f16-vbinary/gen/vmulc-minmax-neonfp16arith-x8.c
-  src/f16-vbinary/gen/vmul-minmax-neonfp16arith-x16.c
-  src/f16-vbinary/gen/vmul-minmax-neonfp16arith-x8.c
-  src/f16-vbinary/gen/vrdivc-minmax-neonfp16arith-x16.c
-  src/f16-vbinary/gen/vrdivc-minmax-neonfp16arith-x8.c
-  src/f16-vbinary/gen/vrsubc-minmax-neonfp16arith-x16.c
-  src/f16-vbinary/gen/vrsubc-minmax-neonfp16arith-x8.c
-  src/f16-vbinary/gen/vsubc-minmax-neonfp16arith-x16.c
-  src/f16-vbinary/gen/vsubc-minmax-neonfp16arith-x8.c
-  src/f16-vbinary/gen/vsub-minmax-neonfp16arith-x16.c
-  src/f16-vbinary/gen/vsub-minmax-neonfp16arith-x8.c
-  src/f16-vmulcaddc/gen/c16-minmax-neonfp16arith-2x.c
-  src/f16-vmulcaddc/gen/c8-minmax-neonfp16arith-2x.c
-  src/f16-gemm/gen/1x8-minmax-neonfp16arith-ld64.c
-  src/f16-gemm/gen/4x8-minmax-neonfp16arith-ld64.c
-  src/f16-gemm/gen/6x8-minmax-neonfp16arith-ld64.c
-  src/f16-gemm/gen/8x8-minmax-neonfp16arith-ld64.c
   src/f16-gemm/gen-inc/1x8inc-minmax-neonfp16arith-ld64.c
-  src/f16-gemm/gen-inc/4x8inc-minmax-neonfp16arith-ld64.c
-  src/f16-gemm/gen-inc/6x8inc-minmax-neonfp16arith-ld64.c
-  src/f16-gemm/gen-inc/8x8inc-minmax-neonfp16arith-ld64.c
-  src/f16-igemm/gen/1x8-minmax-neonfp16arith-ld64.c
-  src/f16-igemm/gen/4x8-minmax-neonfp16arith-ld64.c
-  src/f16-igemm/gen/6x8-minmax-neonfp16arith-ld64.c
-  src/f16-igemm/gen/8x8-minmax-neonfp16arith-ld64.c
-  src/f16-gemm/gen/1x16-minmax-neonfp16arith-ld64.c
-  src/f16-gemm/gen/4x16-minmax-neonfp16arith-ld64.c
-  src/f16-gemm/gen/6x16-minmax-neonfp16arith-ld64.c
-  src/f16-gemm/gen/8x16-minmax-neonfp16arith-ld64.c
   src/f16-gemm/gen-inc/1x16inc-minmax-neonfp16arith-ld64.c
+  src/f16-gemm/gen-inc/4x8inc-minmax-neonfp16arith-ld64.c
   src/f16-gemm/gen-inc/4x16inc-minmax-neonfp16arith-ld64.c
+  src/f16-gemm/gen-inc/6x8inc-minmax-neonfp16arith-ld64.c
   src/f16-gemm/gen-inc/6x16inc-minmax-neonfp16arith-ld64.c
+  src/f16-gemm/gen-inc/8x8inc-minmax-neonfp16arith-ld64.c
   src/f16-gemm/gen-inc/8x16inc-minmax-neonfp16arith-ld64.c
+  src/f16-gemm/gen/1x8-minmax-neonfp16arith-ld64.c
+  src/f16-gemm/gen/1x16-minmax-neonfp16arith-ld64.c
+  src/f16-gemm/gen/4x8-minmax-neonfp16arith-ld64.c
+  src/f16-gemm/gen/4x16-minmax-neonfp16arith-ld64.c
+  src/f16-gemm/gen/6x8-minmax-neonfp16arith-ld64.c
+  src/f16-gemm/gen/6x16-minmax-neonfp16arith-ld64.c
+  src/f16-gemm/gen/8x8-minmax-neonfp16arith-ld64.c
+  src/f16-gemm/gen/8x16-minmax-neonfp16arith-ld64.c
+  src/f16-hswish/gen/hswish-neonfp16arith-x8.c
+  src/f16-hswish/gen/hswish-neonfp16arith-x16.c
+  src/f16-igemm/gen/1x8-minmax-neonfp16arith-ld64.c
   src/f16-igemm/gen/1x16-minmax-neonfp16arith-ld64.c
+  src/f16-igemm/gen/4x8-minmax-neonfp16arith-ld64.c
   src/f16-igemm/gen/4x16-minmax-neonfp16arith-ld64.c
+  src/f16-igemm/gen/6x8-minmax-neonfp16arith-ld64.c
   src/f16-igemm/gen/6x16-minmax-neonfp16arith-ld64.c
+  src/f16-igemm/gen/8x8-minmax-neonfp16arith-ld64.c
   src/f16-igemm/gen/8x16-minmax-neonfp16arith-ld64.c
-  src/f16-spmm/gen/8x1-minmax-neonfp16arith.c
+  src/f16-prelu/gen/neonfp16arith-2x8.c
+  src/f16-prelu/gen/neonfp16arith-2x16.c
+  src/f16-relu/gen/neonfp16arith-x8.c
+  src/f16-relu/gen/neonfp16arith-x16.c
   src/f16-spmm/gen/8x1-minmax-neonfp16arith-unroll2.c
-  src/f16-spmm/gen/16x1-minmax-neonfp16arith.c
+  src/f16-spmm/gen/8x1-minmax-neonfp16arith.c
   src/f16-spmm/gen/16x1-minmax-neonfp16arith-unroll2.c
-  src/f16-spmm/gen/24x1-minmax-neonfp16arith.c
+  src/f16-spmm/gen/16x1-minmax-neonfp16arith.c
   src/f16-spmm/gen/24x1-minmax-neonfp16arith-unroll2.c
+  src/f16-spmm/gen/24x1-minmax-neonfp16arith.c
+  src/f16-spmm/gen/32x1-minmax-neonfp16arith-unroll2.c
   src/f16-spmm/gen/32x1-minmax-neonfp16arith.c
-  src/f16-spmm/gen/32x1-minmax-neonfp16arith-unroll2.c)
+  src/f16-vbinary/gen/vadd-minmax-neonfp16arith-x8.c
+  src/f16-vbinary/gen/vadd-minmax-neonfp16arith-x16.c
+  src/f16-vbinary/gen/vaddc-minmax-neonfp16arith-x8.c
+  src/f16-vbinary/gen/vaddc-minmax-neonfp16arith-x16.c
+  src/f16-vbinary/gen/vdiv-minmax-neonfp16arith-x8.c
+  src/f16-vbinary/gen/vdiv-minmax-neonfp16arith-x16.c
+  src/f16-vbinary/gen/vdivc-minmax-neonfp16arith-x8.c
+  src/f16-vbinary/gen/vdivc-minmax-neonfp16arith-x16.c
+  src/f16-vbinary/gen/vmax-neonfp16arith-x8.c
+  src/f16-vbinary/gen/vmax-neonfp16arith-x16.c
+  src/f16-vbinary/gen/vmaxc-neonfp16arith-x8.c
+  src/f16-vbinary/gen/vmaxc-neonfp16arith-x16.c
+  src/f16-vbinary/gen/vmin-neonfp16arith-x8.c
+  src/f16-vbinary/gen/vmin-neonfp16arith-x16.c
+  src/f16-vbinary/gen/vminc-neonfp16arith-x8.c
+  src/f16-vbinary/gen/vminc-neonfp16arith-x16.c
+  src/f16-vbinary/gen/vmul-minmax-neonfp16arith-x8.c
+  src/f16-vbinary/gen/vmul-minmax-neonfp16arith-x16.c
+  src/f16-vbinary/gen/vmulc-minmax-neonfp16arith-x8.c
+  src/f16-vbinary/gen/vmulc-minmax-neonfp16arith-x16.c
+  src/f16-vbinary/gen/vrdivc-minmax-neonfp16arith-x8.c
+  src/f16-vbinary/gen/vrdivc-minmax-neonfp16arith-x16.c
+  src/f16-vbinary/gen/vrsubc-minmax-neonfp16arith-x8.c
+  src/f16-vbinary/gen/vrsubc-minmax-neonfp16arith-x16.c
+  src/f16-vbinary/gen/vsub-minmax-neonfp16arith-x8.c
+  src/f16-vbinary/gen/vsub-minmax-neonfp16arith-x16.c
+  src/f16-vbinary/gen/vsubc-minmax-neonfp16arith-x8.c
+  src/f16-vbinary/gen/vsubc-minmax-neonfp16arith-x16.c
+  src/f16-vmulcaddc/gen/c8-minmax-neonfp16arith-2x.c
+  src/f16-vmulcaddc/gen/c16-minmax-neonfp16arith-2x.c)
 
 SET(XNNPACK_NEONDOT_MICROKERNEL_SRCS
   src/qs8-gemm/gen/1x8c4-minmax-neondot.c
-  src/qs8-gemm/gen/4x8c4-minmax-neondot.c
-  src/qs8-gemm/gen/6x8c4-minmax-neondot.c
-  src/qs8-gemm/gen/8x8c4-minmax-neondot.c
-  src/qs8-gemm/gen/12x8c4-minmax-neondot.c
   src/qs8-gemm/gen/1x16c4-minmax-neondot.c
+  src/qs8-gemm/gen/4x8c4-minmax-neondot.c
   src/qs8-gemm/gen/4x16c4-minmax-neondot.c
+  src/qs8-gemm/gen/6x8c4-minmax-neondot.c
   src/qs8-gemm/gen/6x16c4-minmax-neondot.c
+  src/qs8-gemm/gen/8x8c4-minmax-neondot.c
   src/qs8-gemm/gen/8x16c4-minmax-neondot.c
+  src/qs8-gemm/gen/12x8c4-minmax-neondot.c
   src/qs8-igemm/gen/1x8c4-minmax-neondot.c
-  src/qs8-igemm/gen/4x8c4-minmax-neondot.c
-  src/qs8-igemm/gen/6x8c4-minmax-neondot.c
-  src/qs8-igemm/gen/8x8c4-minmax-neondot.c
-  src/qs8-igemm/gen/12x8c4-minmax-neondot.c
   src/qs8-igemm/gen/1x16c4-minmax-neondot.c
+  src/qs8-igemm/gen/4x8c4-minmax-neondot.c
   src/qs8-igemm/gen/4x16c4-minmax-neondot.c
+  src/qs8-igemm/gen/6x8c4-minmax-neondot.c
   src/qs8-igemm/gen/6x16c4-minmax-neondot.c
-  src/qs8-igemm/gen/8x16c4-minmax-neondot.c)
+  src/qs8-igemm/gen/8x8c4-minmax-neondot.c
+  src/qs8-igemm/gen/8x16c4-minmax-neondot.c
+  src/qs8-igemm/gen/12x8c4-minmax-neondot.c)
 
 SET(XNNPACK_SSE_MICROKERNEL_SRCS
   src/f32-avgpool/9p8x-minmax-sse-c4.c
@@ -1263,21 +1263,27 @@
   src/f32-conv-hwc2chw/3x3s2p1c3x4-sse-2x2.c
   src/f32-dwconv-chw/3x3p1-sse-1x4-acc3.c
   src/f32-dwconv-chw/3x3s2p1-sse-1x4-acc3.c
-  src/f32-dwconv/gen/up4x25-minmax-sse-acc2.c
-  src/f32-dwconv/gen/up4x25-minmax-sse.c
   src/f32-dwconv/gen/up4x4-minmax-sse-acc2.c
   src/f32-dwconv/gen/up4x4-minmax-sse.c
   src/f32-dwconv/gen/up4x9-minmax-sse-acc2.c
   src/f32-dwconv/gen/up4x9-minmax-sse.c
-  src/f32-dwconv/gen/up8x25-minmax-sse-acc2.c
-  src/f32-dwconv/gen/up8x25-minmax-sse.c
+  src/f32-dwconv/gen/up4x25-minmax-sse-acc2.c
+  src/f32-dwconv/gen/up4x25-minmax-sse.c
   src/f32-dwconv/gen/up8x4-minmax-sse-acc2.c
   src/f32-dwconv/gen/up8x4-minmax-sse.c
   src/f32-dwconv/gen/up8x9-minmax-sse-acc2.c
   src/f32-dwconv/gen/up8x9-minmax-sse.c
+  src/f32-dwconv/gen/up8x25-minmax-sse-acc2.c
+  src/f32-dwconv/gen/up8x25-minmax-sse.c
   src/f32-gavgpool-cw/sse-x4.c
   src/f32-gavgpool/7p7x-minmax-sse-c4.c
   src/f32-gavgpool/7x-minmax-sse-c4.c
+  src/f32-gemm/gen-inc/1x8inc-minmax-sse-dup.c
+  src/f32-gemm/gen-inc/1x8inc-minmax-sse-load1.c
+  src/f32-gemm/gen-inc/1x8s4inc-minmax-sse.c
+  src/f32-gemm/gen-inc/4x8inc-minmax-sse-dup.c
+  src/f32-gemm/gen-inc/4x8inc-minmax-sse-load1.c
+  src/f32-gemm/gen-inc/4x8s4inc-minmax-sse.c
   src/f32-gemm/gen/1x8-minmax-sse-dup.c
   src/f32-gemm/gen/1x8-minmax-sse-load1.c
   src/f32-gemm/gen/1x8s4-minmax-sse.c
@@ -1285,12 +1291,6 @@
   src/f32-gemm/gen/4x8-minmax-sse-dup.c
   src/f32-gemm/gen/4x8-minmax-sse-load1.c
   src/f32-gemm/gen/4x8s4-minmax-sse.c
-  src/f32-gemm/gen-inc/1x8inc-minmax-sse-dup.c
-  src/f32-gemm/gen-inc/1x8inc-minmax-sse-load1.c
-  src/f32-gemm/gen-inc/1x8s4inc-minmax-sse.c
-  src/f32-gemm/gen-inc/4x8inc-minmax-sse-dup.c
-  src/f32-gemm/gen-inc/4x8inc-minmax-sse-load1.c
-  src/f32-gemm/gen-inc/4x8s4inc-minmax-sse.c
   src/f32-hswish/gen/hswish-sse-x4.c
   src/f32-hswish/gen/hswish-sse-x8.c
   src/f32-ibilinear/gen/sse-c4.c
@@ -1302,17 +1302,15 @@
   src/f32-igemm/gen/4x8-minmax-sse-dup.c
   src/f32-igemm/gen/4x8-minmax-sse-load1.c
   src/f32-igemm/gen/4x8s4-minmax-sse.c
-  src/f32-vlrelu/gen/vlrelu-sse-x4.c
-  src/f32-vlrelu/gen/vlrelu-sse-x8.c
   src/f32-maxpool/9p8x-minmax-sse-c4.c
   src/f32-pavgpool/9p8x-minmax-sse-c4.c
   src/f32-pavgpool/9x-minmax-sse-c4.c
   src/f32-ppmm/gen/4x8-minmax-sse.c
   src/f32-prelu/gen/sse-2x4.c
   src/f32-prelu/gen/sse-2x8.c
-  src/f32-rmax/sse.c
   src/f32-relu/gen/sse-x4.c
   src/f32-relu/gen/sse-x8.c
+  src/f32-rmax/sse.c
   src/f32-spmm/gen/4x1-minmax-sse.c
   src/f32-spmm/gen/8x1-minmax-sse.c
   src/f32-spmm/gen/16x1-minmax-sse.c
@@ -1348,6 +1346,8 @@
   src/f32-vbinary/gen/vsub-minmax-sse-x8.c
   src/f32-vbinary/gen/vsubc-minmax-sse-x4.c
   src/f32-vbinary/gen/vsubc-minmax-sse-x8.c
+  src/f32-vlrelu/gen/vlrelu-sse-x4.c
+  src/f32-vlrelu/gen/vlrelu-sse-x8.c
   src/f32-vmulcaddc/gen/c4-minmax-sse-2x.c
   src/f32-vmulcaddc/gen/c8-minmax-sse-2x.c
   src/f32-vsqrt/gen/sse-sqrt-x4.c
@@ -1358,16 +1358,16 @@
   src/f32-vunary/gen/vneg-sse-x8.c
   src/f32-vunary/gen/vsqr-sse-x4.c
   src/f32-vunary/gen/vsqr-sse-x8.c
-  src/x32-fill/sse.c
-  src/x32-packx/x4-sse.c
-  src/x32-pad/sse.c
-  src/math/roundne-sse-addsub.c
   src/math/roundd-sse-addsub.c
+  src/math/roundne-sse-addsub.c
   src/math/roundu-sse-addsub.c
   src/math/roundz-sse-addsub.c
   src/math/sqrt-sse-hh1mac.c
   src/math/sqrt-sse-nr1mac.c
-  src/math/sqrt-sse-nr2mac.c)
+  src/math/sqrt-sse-nr2mac.c
+  src/x32-fill/sse.c
+  src/x32-packx/x4-sse.c
+  src/x32-pad/sse.c)
 
 SET(XNNPACK_SSE2_MICROKERNEL_SRCS
   src/f32-argmaxpool/4x-sse2-c4.c
@@ -1376,17 +1376,17 @@
   src/f32-prelu/gen/sse2-2x4.c
   src/f32-prelu/gen/sse2-2x8.c
   src/f32-raddstoreexpminusmax/gen/sse2-p5-x4.c
-  src/f32-raddstoreexpminusmax/gen/sse2-p5-x8.c
   src/f32-raddstoreexpminusmax/gen/sse2-p5-x8-acc2.c
-  src/f32-raddstoreexpminusmax/gen/sse2-p5-x12.c
+  src/f32-raddstoreexpminusmax/gen/sse2-p5-x8.c
   src/f32-raddstoreexpminusmax/gen/sse2-p5-x12-acc2.c
   src/f32-raddstoreexpminusmax/gen/sse2-p5-x12-acc3.c
-  src/f32-raddstoreexpminusmax/gen/sse2-p5-x16.c
+  src/f32-raddstoreexpminusmax/gen/sse2-p5-x12.c
   src/f32-raddstoreexpminusmax/gen/sse2-p5-x16-acc2.c
   src/f32-raddstoreexpminusmax/gen/sse2-p5-x16-acc4.c
-  src/f32-raddstoreexpminusmax/gen/sse2-p5-x20.c
+  src/f32-raddstoreexpminusmax/gen/sse2-p5-x16.c
   src/f32-raddstoreexpminusmax/gen/sse2-p5-x20-acc2.c
   src/f32-raddstoreexpminusmax/gen/sse2-p5-x20-acc5.c
+  src/f32-raddstoreexpminusmax/gen/sse2-p5-x20.c
   src/f32-sigmoid/gen/sse2-lut64-p2-div-x4.c
   src/f32-sigmoid/gen/sse2-lut64-p2-div-x8.c
   src/f32-sigmoid/gen/sse2-lut64-p2-div-x12.c
@@ -1401,14 +1401,27 @@
   src/f32-sigmoid/gen/sse2-p5-div-x24.c
   src/f32-vlrelu/gen/vlrelu-sse2-x4.c
   src/f32-vlrelu/gen/vlrelu-sse2-x8.c
-  src/f32-vrnd/gen/vrndne-sse2-x4.c
-  src/f32-vrnd/gen/vrndne-sse2-x8.c
-  src/f32-vrnd/gen/vrndz-sse2-x4.c
-  src/f32-vrnd/gen/vrndz-sse2-x8.c
-  src/f32-vrnd/gen/vrndu-sse2-x4.c
-  src/f32-vrnd/gen/vrndu-sse2-x8.c
   src/f32-vrnd/gen/vrndd-sse2-x4.c
   src/f32-vrnd/gen/vrndd-sse2-x8.c
+  src/f32-vrnd/gen/vrndne-sse2-x4.c
+  src/f32-vrnd/gen/vrndne-sse2-x8.c
+  src/f32-vrnd/gen/vrndu-sse2-x4.c
+  src/f32-vrnd/gen/vrndu-sse2-x8.c
+  src/f32-vrnd/gen/vrndz-sse2-x4.c
+  src/f32-vrnd/gen/vrndz-sse2-x8.c
+  src/math/exp-sse2-lut64-p2.c
+  src/math/exp-sse2-p5.c
+  src/math/expminus-sse2-p5.c
+  src/math/roundd-sse2-cvt.c
+  src/math/roundne-sse2-cvt.c
+  src/math/roundu-sse2-cvt.c
+  src/math/roundz-sse2-cvt.c
+  src/math/sigmoid-sse2-rr2-lut64-p2-div.c
+  src/math/sigmoid-sse2-rr2-lut64-p2-nr1.c
+  src/math/sigmoid-sse2-rr2-lut64-p2-nr2.c
+  src/math/sigmoid-sse2-rr2-p5-div.c
+  src/math/sigmoid-sse2-rr2-p5-nr1.c
+  src/math/sigmoid-sse2-rr2-p5-nr2.c
   src/qs8-dwconv/gen/up8x9-minmax-sse2-mul16.c
   src/qs8-dwconv/gen/up16x9-minmax-sse2-mul16.c
   src/qs8-dwconv/gen/up24x9-minmax-sse2-mul16.c
@@ -1419,30 +1432,30 @@
   src/qs8-gavgpool/gen/7x-minmax-sse2-c16-acc2.c
   src/qs8-gavgpool/gen/7x-minmax-sse2-c24-acc2.c
   src/qs8-gemm/gen/1x4c2-minmax-sse2-ld64.c
-  src/qs8-gemm/gen/4x4c2-minmax-sse2-ld64.c
   src/qs8-gemm/gen/1x4c2-minmax-sse2-ld128.c
-  src/qs8-gemm/gen/4x4c2-minmax-sse2-ld128.c
   src/qs8-gemm/gen/1x4c2-xw-minmax-sse2.c
-  src/qs8-gemm/gen/4x4c2-xw-minmax-sse2.c
   src/qs8-gemm/gen/1x4c8-minmax-sse2-ld64.c
-  src/qs8-gemm/gen/2x4c8-minmax-sse2-ld64.c
-  src/qs8-gemm/gen/3x4c8-minmax-sse2-ld64.c
   src/qs8-gemm/gen/1x4c8-minmax-sse2-ld128.c
-  src/qs8-gemm/gen/2x4c8-minmax-sse2-ld128.c
-  src/qs8-gemm/gen/3x4c8-minmax-sse2-ld128.c
   src/qs8-gemm/gen/1x4c8-xw-minmax-sse2.c
+  src/qs8-gemm/gen/2x4c8-minmax-sse2-ld64.c
+  src/qs8-gemm/gen/2x4c8-minmax-sse2-ld128.c
   src/qs8-gemm/gen/2x4c8-xw-minmax-sse2.c
+  src/qs8-gemm/gen/3x4c8-minmax-sse2-ld64.c
+  src/qs8-gemm/gen/3x4c8-minmax-sse2-ld128.c
   src/qs8-gemm/gen/3x4c8-xw-minmax-sse2.c
+  src/qs8-gemm/gen/4x4c2-minmax-sse2-ld64.c
+  src/qs8-gemm/gen/4x4c2-minmax-sse2-ld128.c
+  src/qs8-gemm/gen/4x4c2-xw-minmax-sse2.c
   src/qs8-igemm/gen/1x4c2-minmax-sse2-ld64.c
-  src/qs8-igemm/gen/4x4c2-minmax-sse2-ld64.c
   src/qs8-igemm/gen/1x4c2-minmax-sse2-ld128.c
-  src/qs8-igemm/gen/4x4c2-minmax-sse2-ld128.c
   src/qs8-igemm/gen/1x4c8-minmax-sse2-ld64.c
-  src/qs8-igemm/gen/2x4c8-minmax-sse2-ld64.c
-  src/qs8-igemm/gen/3x4c8-minmax-sse2-ld64.c
   src/qs8-igemm/gen/1x4c8-minmax-sse2-ld128.c
+  src/qs8-igemm/gen/2x4c8-minmax-sse2-ld64.c
   src/qs8-igemm/gen/2x4c8-minmax-sse2-ld128.c
+  src/qs8-igemm/gen/3x4c8-minmax-sse2-ld64.c
   src/qs8-igemm/gen/3x4c8-minmax-sse2-ld128.c
+  src/qs8-igemm/gen/4x4c2-minmax-sse2-ld64.c
+  src/qs8-igemm/gen/4x4c2-minmax-sse2-ld128.c
   src/qs8-requantization/fp32-sse2.c
   src/qs8-requantization/precise-sse2.c
   src/qs8-requantization/q31-sse2.c
@@ -1456,12 +1469,12 @@
   src/qs8-vaddc/gen/minmax-sse2-mul16-ld64-x32.c
   src/qu8-avgpool/9p8x-minmax-sse2-c8.c
   src/qu8-avgpool/9x-minmax-sse2-c8.c
-  src/qu8-igemm/4x4c2-minmax-sse2.c
   src/qu8-dwconv/up8x9-minmax-sse2.c
   src/qu8-gavgpool/7p7x-minmax-sse2-c8.c
   src/qu8-gavgpool/7x-minmax-sse2-c8.c
   src/qu8-gemm/2x4c8-minmax-sse2.c
   src/qu8-gemm/4x4c2-minmax-sse2.c
+  src/qu8-igemm/4x4c2-minmax-sse2.c
   src/qu8-requantization/fp32-sse2.c
   src/qu8-requantization/precise-sse2.c
   src/qu8-requantization/q31-sse2.c
@@ -1469,28 +1482,15 @@
   src/u8-clamp/sse2-x64.c
   src/u8-maxpool/9p8x-minmax-sse2-c16.c
   src/u8-rmax/sse2.c
-  src/x32-unpool/sse2.c
-  src/x32-zip/x2-sse2.c
-  src/x32-zip/x3-sse2.c
-  src/x32-zip/x4-sse2.c
-  src/x32-zip/xm-sse2.c
   src/x8-zip/x2-sse2.c
   src/x8-zip/x3-sse2.c
   src/x8-zip/x4-sse2.c
   src/x8-zip/xm-sse2.c
-  src/math/exp-sse2-p5.c
-  src/math/exp-sse2-lut64-p2.c
-  src/math/expminus-sse2-p5.c
-  src/math/roundne-sse2-cvt.c
-  src/math/roundd-sse2-cvt.c
-  src/math/roundu-sse2-cvt.c
-  src/math/roundz-sse2-cvt.c
-  src/math/sigmoid-sse2-rr2-p5-div.c
-  src/math/sigmoid-sse2-rr2-p5-nr1.c
-  src/math/sigmoid-sse2-rr2-p5-nr2.c
-  src/math/sigmoid-sse2-rr2-lut64-p2-div.c
-  src/math/sigmoid-sse2-rr2-lut64-p2-nr1.c
-  src/math/sigmoid-sse2-rr2-lut64-p2-nr2.c)
+  src/x32-unpool/sse2.c
+  src/x32-zip/x2-sse2.c
+  src/x32-zip/x3-sse2.c
+  src/x32-zip/x4-sse2.c
+  src/x32-zip/xm-sse2.c)
 
 SET(XNNPACK_SSSE3_MICROKERNEL_SRCS
   src/qs8-dwconv/gen/up8x9-minmax-ssse3-mul16.c
@@ -1503,30 +1503,30 @@
   src/qs8-gavgpool/gen/7x-minmax-ssse3-c16-acc2.c
   src/qs8-gavgpool/gen/7x-minmax-ssse3-c24-acc2.c
   src/qs8-gemm/gen/1x4c2-minmax-ssse3-ld64.c
-  src/qs8-gemm/gen/4x4c2-minmax-ssse3-ld64.c
   src/qs8-gemm/gen/1x4c2-minmax-ssse3-ld128.c
-  src/qs8-gemm/gen/4x4c2-minmax-ssse3-ld128.c
   src/qs8-gemm/gen/1x4c2-xw-minmax-ssse3.c
-  src/qs8-gemm/gen/4x4c2-xw-minmax-ssse3.c
   src/qs8-gemm/gen/1x4c8-minmax-ssse3-ld64.c
-  src/qs8-gemm/gen/2x4c8-minmax-ssse3-ld64.c
-  src/qs8-gemm/gen/3x4c8-minmax-ssse3-ld64.c
   src/qs8-gemm/gen/1x4c8-minmax-ssse3-ld128.c
-  src/qs8-gemm/gen/2x4c8-minmax-ssse3-ld128.c
-  src/qs8-gemm/gen/3x4c8-minmax-ssse3-ld128.c
   src/qs8-gemm/gen/1x4c8-xw-minmax-ssse3.c
+  src/qs8-gemm/gen/2x4c8-minmax-ssse3-ld64.c
+  src/qs8-gemm/gen/2x4c8-minmax-ssse3-ld128.c
   src/qs8-gemm/gen/2x4c8-xw-minmax-ssse3.c
+  src/qs8-gemm/gen/3x4c8-minmax-ssse3-ld64.c
+  src/qs8-gemm/gen/3x4c8-minmax-ssse3-ld128.c
   src/qs8-gemm/gen/3x4c8-xw-minmax-ssse3.c
+  src/qs8-gemm/gen/4x4c2-minmax-ssse3-ld64.c
+  src/qs8-gemm/gen/4x4c2-minmax-ssse3-ld128.c
+  src/qs8-gemm/gen/4x4c2-xw-minmax-ssse3.c
   src/qs8-igemm/gen/1x4c2-minmax-ssse3-ld64.c
-  src/qs8-igemm/gen/4x4c2-minmax-ssse3-ld64.c
   src/qs8-igemm/gen/1x4c2-minmax-ssse3-ld128.c
-  src/qs8-igemm/gen/4x4c2-minmax-ssse3-ld128.c
   src/qs8-igemm/gen/1x4c8-minmax-ssse3-ld64.c
-  src/qs8-igemm/gen/2x4c8-minmax-ssse3-ld64.c
-  src/qs8-igemm/gen/3x4c8-minmax-ssse3-ld64.c
   src/qs8-igemm/gen/1x4c8-minmax-ssse3-ld128.c
+  src/qs8-igemm/gen/2x4c8-minmax-ssse3-ld64.c
   src/qs8-igemm/gen/2x4c8-minmax-ssse3-ld128.c
+  src/qs8-igemm/gen/3x4c8-minmax-ssse3-ld64.c
   src/qs8-igemm/gen/3x4c8-minmax-ssse3-ld128.c
+  src/qs8-igemm/gen/4x4c2-minmax-ssse3-ld64.c
+  src/qs8-igemm/gen/4x4c2-minmax-ssse3-ld128.c
   src/qs8-requantization/precise-ssse3.c
   src/qs8-requantization/q31-ssse3.c
   src/qu8-requantization/precise-ssse3.c
@@ -1549,14 +1549,18 @@
   src/f32-sigmoid/gen/sse41-p5-div-x24.c
   src/f32-vlrelu/gen/vlrelu-sse41-x4.c
   src/f32-vlrelu/gen/vlrelu-sse41-x8.c
-  src/f32-vrnd/gen/vrndne-sse41-x4.c
-  src/f32-vrnd/gen/vrndne-sse41-x8.c
-  src/f32-vrnd/gen/vrndz-sse41-x4.c
-  src/f32-vrnd/gen/vrndz-sse41-x8.c
-  src/f32-vrnd/gen/vrndu-sse41-x4.c
-  src/f32-vrnd/gen/vrndu-sse41-x8.c
   src/f32-vrnd/gen/vrndd-sse41-x4.c
   src/f32-vrnd/gen/vrndd-sse41-x8.c
+  src/f32-vrnd/gen/vrndne-sse41-x4.c
+  src/f32-vrnd/gen/vrndne-sse41-x8.c
+  src/f32-vrnd/gen/vrndu-sse41-x4.c
+  src/f32-vrnd/gen/vrndu-sse41-x8.c
+  src/f32-vrnd/gen/vrndz-sse41-x4.c
+  src/f32-vrnd/gen/vrndz-sse41-x8.c
+  src/math/roundd-sse41.c
+  src/math/roundne-sse41.c
+  src/math/roundu-sse41.c
+  src/math/roundz-sse41.c
   src/qs8-dwconv/gen/up8x9-minmax-sse41-mul16.c
   src/qs8-dwconv/gen/up16x9-minmax-sse41-mul16.c
   src/qs8-dwconv/gen/up24x9-minmax-sse41-mul16.c
@@ -1567,30 +1571,30 @@
   src/qs8-gavgpool/gen/7x-minmax-sse41-c16-acc2.c
   src/qs8-gavgpool/gen/7x-minmax-sse41-c24-acc2.c
   src/qs8-gemm/gen/1x4c2-minmax-sse41-ld64.c
-  src/qs8-gemm/gen/4x4c2-minmax-sse41-ld64.c
   src/qs8-gemm/gen/1x4c2-minmax-sse41-ld128.c
-  src/qs8-gemm/gen/4x4c2-minmax-sse41-ld128.c
   src/qs8-gemm/gen/1x4c2-xw-minmax-sse41.c
-  src/qs8-gemm/gen/4x4c2-xw-minmax-sse41.c
   src/qs8-gemm/gen/1x4c8-minmax-sse41-ld64.c
-  src/qs8-gemm/gen/2x4c8-minmax-sse41-ld64.c
-  src/qs8-gemm/gen/3x4c8-minmax-sse41-ld64.c
   src/qs8-gemm/gen/1x4c8-minmax-sse41-ld128.c
-  src/qs8-gemm/gen/2x4c8-minmax-sse41-ld128.c
-  src/qs8-gemm/gen/3x4c8-minmax-sse41-ld128.c
   src/qs8-gemm/gen/1x4c8-xw-minmax-sse41.c
+  src/qs8-gemm/gen/2x4c8-minmax-sse41-ld64.c
+  src/qs8-gemm/gen/2x4c8-minmax-sse41-ld128.c
   src/qs8-gemm/gen/2x4c8-xw-minmax-sse41.c
+  src/qs8-gemm/gen/3x4c8-minmax-sse41-ld64.c
+  src/qs8-gemm/gen/3x4c8-minmax-sse41-ld128.c
   src/qs8-gemm/gen/3x4c8-xw-minmax-sse41.c
+  src/qs8-gemm/gen/4x4c2-minmax-sse41-ld64.c
+  src/qs8-gemm/gen/4x4c2-minmax-sse41-ld128.c
+  src/qs8-gemm/gen/4x4c2-xw-minmax-sse41.c
   src/qs8-igemm/gen/1x4c2-minmax-sse41-ld64.c
-  src/qs8-igemm/gen/4x4c2-minmax-sse41-ld64.c
   src/qs8-igemm/gen/1x4c2-minmax-sse41-ld128.c
-  src/qs8-igemm/gen/4x4c2-minmax-sse41-ld128.c
   src/qs8-igemm/gen/1x4c8-minmax-sse41-ld64.c
-  src/qs8-igemm/gen/2x4c8-minmax-sse41-ld64.c
-  src/qs8-igemm/gen/3x4c8-minmax-sse41-ld64.c
   src/qs8-igemm/gen/1x4c8-minmax-sse41-ld128.c
+  src/qs8-igemm/gen/2x4c8-minmax-sse41-ld64.c
   src/qs8-igemm/gen/2x4c8-minmax-sse41-ld128.c
+  src/qs8-igemm/gen/3x4c8-minmax-sse41-ld64.c
   src/qs8-igemm/gen/3x4c8-minmax-sse41-ld128.c
+  src/qs8-igemm/gen/4x4c2-minmax-sse41-ld64.c
+  src/qs8-igemm/gen/4x4c2-minmax-sse41-ld128.c
   src/qs8-requantization/fp32-sse4.c
   src/qs8-requantization/precise-sse4.c
   src/qs8-requantization/q31-sse4.c
@@ -1611,58 +1615,52 @@
   src/qs8-vaddc/gen/minmax-sse41-mul32-ld32-x24.c
   src/qs8-vaddc/gen/minmax-sse41-mul32-ld32-x32.c
   src/qu8-requantization/precise-sse4.c
-  src/qu8-requantization/q31-sse4.c
-  src/math/roundne-sse41.c
-  src/math/roundd-sse41.c
-  src/math/roundu-sse41.c
-  src/math/roundz-sse41.c)
+  src/qu8-requantization/q31-sse4.c)
 
 SET(XNNPACK_AVX_MICROKERNEL_SRCS
   src/f32-clamp/gen/avx-x8.c
   src/f32-clamp/gen/avx-x16.c
-  src/f32-dwconv/gen/up16x4-minmax-avx-acc2.c
-  src/f32-dwconv/gen/up16x4-minmax-avx.c
   src/f32-dwconv/gen/up8x4-minmax-avx-acc2.c
   src/f32-dwconv/gen/up8x4-minmax-avx.c
-  src/f32-dwconv/gen/up16x9-minmax-avx-acc2.c
-  src/f32-dwconv/gen/up16x9-minmax-avx.c
   src/f32-dwconv/gen/up8x9-minmax-avx-acc2.c
   src/f32-dwconv/gen/up8x9-minmax-avx.c
-  src/f32-dwconv/gen/up16x25-minmax-avx-acc2.c
-  src/f32-dwconv/gen/up16x25-minmax-avx.c
   src/f32-dwconv/gen/up8x25-minmax-avx-acc2.c
   src/f32-dwconv/gen/up8x25-minmax-avx.c
-  src/f32-gemm/gen/1x8-minmax-avx-broadcast.c
-  src/f32-gemm/gen/4x8-minmax-avx-broadcast.c
-  src/f32-gemm/gen/5x8-minmax-avx-broadcast.c
-  src/f32-gemm/gen/6x8-minmax-avx-broadcast.c
-  src/f32-gemm/gen/7x8-minmax-avx-broadcast.c
-  src/f32-gemm/gen/1x16-minmax-avx-broadcast.c
-  src/f32-gemm/gen/3x16-minmax-avx-broadcast.c
-  src/f32-gemm/gen/4x16-minmax-avx-broadcast.c
-  src/f32-gemm/gen/5x16-minmax-avx-broadcast.c
+  src/f32-dwconv/gen/up16x4-minmax-avx-acc2.c
+  src/f32-dwconv/gen/up16x4-minmax-avx.c
+  src/f32-dwconv/gen/up16x9-minmax-avx-acc2.c
+  src/f32-dwconv/gen/up16x9-minmax-avx.c
+  src/f32-dwconv/gen/up16x25-minmax-avx-acc2.c
+  src/f32-dwconv/gen/up16x25-minmax-avx.c
   src/f32-gemm/gen-inc/1x8inc-minmax-avx-broadcast.c
-  src/f32-gemm/gen-inc/4x8inc-minmax-avx-broadcast.c
-  src/f32-gemm/gen-inc/5x8inc-minmax-avx-broadcast.c
-  src/f32-gemm/gen-inc/6x8inc-minmax-avx-broadcast.c
-  src/f32-gemm/gen-inc/7x8inc-minmax-avx-broadcast.c
   src/f32-gemm/gen-inc/1x16inc-minmax-avx-broadcast.c
   src/f32-gemm/gen-inc/3x16inc-minmax-avx-broadcast.c
+  src/f32-gemm/gen-inc/4x8inc-minmax-avx-broadcast.c
   src/f32-gemm/gen-inc/4x16inc-minmax-avx-broadcast.c
+  src/f32-gemm/gen-inc/5x8inc-minmax-avx-broadcast.c
   src/f32-gemm/gen-inc/5x16inc-minmax-avx-broadcast.c
+  src/f32-gemm/gen-inc/6x8inc-minmax-avx-broadcast.c
+  src/f32-gemm/gen-inc/7x8inc-minmax-avx-broadcast.c
+  src/f32-gemm/gen/1x8-minmax-avx-broadcast.c
+  src/f32-gemm/gen/1x16-minmax-avx-broadcast.c
+  src/f32-gemm/gen/3x16-minmax-avx-broadcast.c
+  src/f32-gemm/gen/4x8-minmax-avx-broadcast.c
+  src/f32-gemm/gen/4x16-minmax-avx-broadcast.c
+  src/f32-gemm/gen/5x8-minmax-avx-broadcast.c
+  src/f32-gemm/gen/5x16-minmax-avx-broadcast.c
+  src/f32-gemm/gen/6x8-minmax-avx-broadcast.c
+  src/f32-gemm/gen/7x8-minmax-avx-broadcast.c
   src/f32-hswish/gen/hswish-avx-x8.c
   src/f32-hswish/gen/hswish-avx-x16.c
   src/f32-igemm/gen/1x8-minmax-avx-broadcast.c
-  src/f32-igemm/gen/4x8-minmax-avx-broadcast.c
-  src/f32-igemm/gen/5x8-minmax-avx-broadcast.c
-  src/f32-igemm/gen/6x8-minmax-avx-broadcast.c
-  src/f32-igemm/gen/7x8-minmax-avx-broadcast.c
   src/f32-igemm/gen/1x16-minmax-avx-broadcast.c
   src/f32-igemm/gen/3x16-minmax-avx-broadcast.c
+  src/f32-igemm/gen/4x8-minmax-avx-broadcast.c
   src/f32-igemm/gen/4x16-minmax-avx-broadcast.c
+  src/f32-igemm/gen/5x8-minmax-avx-broadcast.c
   src/f32-igemm/gen/5x16-minmax-avx-broadcast.c
-  src/f32-vlrelu/gen/vlrelu-avx-x8.c
-  src/f32-vlrelu/gen/vlrelu-avx-x16.c
+  src/f32-igemm/gen/6x8-minmax-avx-broadcast.c
+  src/f32-igemm/gen/7x8-minmax-avx-broadcast.c
   src/f32-prelu/gen/avx-2x8.c
   src/f32-prelu/gen/avx-2x16.c
   src/f32-rmax/avx.c
@@ -1718,14 +1716,16 @@
   src/f32-vbinary/gen/vsub-minmax-avx-x16.c
   src/f32-vbinary/gen/vsubc-minmax-avx-x8.c
   src/f32-vbinary/gen/vsubc-minmax-avx-x16.c
-  src/f32-vrnd/gen/vrndne-avx-x8.c
-  src/f32-vrnd/gen/vrndne-avx-x16.c
-  src/f32-vrnd/gen/vrndz-avx-x8.c
-  src/f32-vrnd/gen/vrndz-avx-x16.c
-  src/f32-vrnd/gen/vrndu-avx-x8.c
-  src/f32-vrnd/gen/vrndu-avx-x16.c
+  src/f32-vlrelu/gen/vlrelu-avx-x8.c
+  src/f32-vlrelu/gen/vlrelu-avx-x16.c
   src/f32-vrnd/gen/vrndd-avx-x8.c
   src/f32-vrnd/gen/vrndd-avx-x16.c
+  src/f32-vrnd/gen/vrndne-avx-x8.c
+  src/f32-vrnd/gen/vrndne-avx-x16.c
+  src/f32-vrnd/gen/vrndu-avx-x8.c
+  src/f32-vrnd/gen/vrndu-avx-x16.c
+  src/f32-vrnd/gen/vrndz-avx-x8.c
+  src/f32-vrnd/gen/vrndz-avx-x16.c
   src/f32-vscale/avx-unroll32.c
   src/f32-vsqrt/gen/avx-sqrt-x8.c
   src/f32-vsqrt/gen/avx-sqrt-x16.c
@@ -1736,37 +1736,37 @@
   src/f32-vunary/gen/vsqr-avx-x8.c
   src/f32-vunary/gen/vsqr-avx-x16.c
   src/math/exp-avx-p5.c
+  src/math/sigmoid-avx-rr2-lut64-p2-div.c
   src/math/sigmoid-avx-rr2-p5-div.c
   src/math/sigmoid-avx-rr2-p5-nr1.c
-  src/math/sigmoid-avx-rr2-p5-nr2.c
-  src/math/sigmoid-avx-rr2-lut64-p2-div.c)
+  src/math/sigmoid-avx-rr2-p5-nr2.c)
 
 SET(XNNPACK_XOP_MICROKERNEL_SRCS
   src/qs8-gemm/gen/1x4c2-minmax-xop-ld64.c
-  src/qs8-gemm/gen/4x4c2-minmax-xop-ld64.c
   src/qs8-gemm/gen/1x4c2-minmax-xop-ld128.c
-  src/qs8-gemm/gen/4x4c2-minmax-xop-ld128.c
   src/qs8-gemm/gen/1x4c2-xw-minmax-xop.c
-  src/qs8-gemm/gen/4x4c2-xw-minmax-xop.c
   src/qs8-gemm/gen/1x4c8-minmax-xop-ld64.c
-  src/qs8-gemm/gen/2x4c8-minmax-xop-ld64.c
-  src/qs8-gemm/gen/3x4c8-minmax-xop-ld64.c
   src/qs8-gemm/gen/1x4c8-minmax-xop-ld128.c
-  src/qs8-gemm/gen/2x4c8-minmax-xop-ld128.c
-  src/qs8-gemm/gen/3x4c8-minmax-xop-ld128.c
   src/qs8-gemm/gen/1x4c8-xw-minmax-xop.c
+  src/qs8-gemm/gen/2x4c8-minmax-xop-ld64.c
+  src/qs8-gemm/gen/2x4c8-minmax-xop-ld128.c
   src/qs8-gemm/gen/2x4c8-xw-minmax-xop.c
+  src/qs8-gemm/gen/3x4c8-minmax-xop-ld64.c
+  src/qs8-gemm/gen/3x4c8-minmax-xop-ld128.c
   src/qs8-gemm/gen/3x4c8-xw-minmax-xop.c
+  src/qs8-gemm/gen/4x4c2-minmax-xop-ld64.c
+  src/qs8-gemm/gen/4x4c2-minmax-xop-ld128.c
+  src/qs8-gemm/gen/4x4c2-xw-minmax-xop.c
   src/qs8-igemm/gen/1x4c2-minmax-xop-ld64.c
-  src/qs8-igemm/gen/4x4c2-minmax-xop-ld64.c
   src/qs8-igemm/gen/1x4c2-minmax-xop-ld128.c
-  src/qs8-igemm/gen/4x4c2-minmax-xop-ld128.c
   src/qs8-igemm/gen/1x4c8-minmax-xop-ld64.c
-  src/qs8-igemm/gen/2x4c8-minmax-xop-ld64.c
-  src/qs8-igemm/gen/3x4c8-minmax-xop-ld64.c
   src/qs8-igemm/gen/1x4c8-minmax-xop-ld128.c
+  src/qs8-igemm/gen/2x4c8-minmax-xop-ld64.c
   src/qs8-igemm/gen/2x4c8-minmax-xop-ld128.c
+  src/qs8-igemm/gen/3x4c8-minmax-xop-ld64.c
   src/qs8-igemm/gen/3x4c8-minmax-xop-ld128.c
+  src/qs8-igemm/gen/4x4c2-minmax-xop-ld64.c
+  src/qs8-igemm/gen/4x4c2-minmax-xop-ld128.c
   src/qs8-vadd/gen/minmax-xop-mul32-ld32-x8.c
   src/qs8-vadd/gen/minmax-xop-mul32-ld32-x16.c
   src/qs8-vadd/gen/minmax-xop-mul32-ld32-x24.c
@@ -1777,62 +1777,62 @@
   src/qs8-vaddc/gen/minmax-xop-mul32-ld32-x32.c)
 
 SET(XNNPACK_FMA3_MICROKERNEL_SRCS
-  src/f32-dwconv/gen/up16x4-minmax-fma3-acc2.c
-  src/f32-dwconv/gen/up16x4-minmax-fma3.c
   src/f32-dwconv/gen/up8x4-minmax-fma3-acc2.c
   src/f32-dwconv/gen/up8x4-minmax-fma3.c
-  src/f32-dwconv/gen/up16x9-minmax-fma3-acc2.c
-  src/f32-dwconv/gen/up16x9-minmax-fma3.c
   src/f32-dwconv/gen/up8x9-minmax-fma3-acc2.c
   src/f32-dwconv/gen/up8x9-minmax-fma3.c
-  src/f32-dwconv/gen/up16x25-minmax-fma3-acc2.c
-  src/f32-dwconv/gen/up16x25-minmax-fma3.c
   src/f32-dwconv/gen/up8x25-minmax-fma3-acc2.c
   src/f32-dwconv/gen/up8x25-minmax-fma3.c
-  src/f32-gemm/gen/1x8-minmax-fma3-broadcast.c
-  src/f32-gemm/gen/4x8-minmax-fma3-broadcast.c
-  src/f32-gemm/gen/5x8-minmax-fma3-broadcast.c
-  src/f32-gemm/gen/6x8-minmax-fma3-broadcast.c
-  src/f32-gemm/gen/7x8-minmax-fma3-broadcast.c
-  src/f32-gemm/gen/8x8-minmax-fma3-broadcast.c
-  src/f32-gemm/gen/1x16-minmax-fma3-broadcast.c
-  src/f32-gemm/gen/3x16-minmax-fma3-broadcast.c
-  src/f32-gemm/gen/4x16-minmax-fma3-broadcast.c
-  src/f32-gemm/gen/5x16-minmax-fma3-broadcast.c
-  src/f32-gemm/gen/1x16s4-minmax-fma3-broadcast.c
-  src/f32-gemm/gen/3x16s4-minmax-fma3-broadcast.c
-  src/f32-gemm/gen/4x16s4-minmax-fma3-broadcast.c
-  src/f32-gemm/gen/5x16s4-minmax-fma3-broadcast.c
+  src/f32-dwconv/gen/up16x4-minmax-fma3-acc2.c
+  src/f32-dwconv/gen/up16x4-minmax-fma3.c
+  src/f32-dwconv/gen/up16x9-minmax-fma3-acc2.c
+  src/f32-dwconv/gen/up16x9-minmax-fma3.c
+  src/f32-dwconv/gen/up16x25-minmax-fma3-acc2.c
+  src/f32-dwconv/gen/up16x25-minmax-fma3.c
   src/f32-gemm/gen-inc/1x8inc-minmax-fma3-broadcast.c
+  src/f32-gemm/gen-inc/1x16inc-minmax-fma3-broadcast.c
+  src/f32-gemm/gen-inc/1x16s4inc-minmax-fma3-broadcast.c
+  src/f32-gemm/gen-inc/3x16inc-minmax-fma3-broadcast.c
+  src/f32-gemm/gen-inc/3x16s4inc-minmax-fma3-broadcast.c
   src/f32-gemm/gen-inc/4x8inc-minmax-fma3-broadcast.c
+  src/f32-gemm/gen-inc/4x16inc-minmax-fma3-broadcast.c
+  src/f32-gemm/gen-inc/4x16s4inc-minmax-fma3-broadcast.c
   src/f32-gemm/gen-inc/5x8inc-minmax-fma3-broadcast.c
+  src/f32-gemm/gen-inc/5x16inc-minmax-fma3-broadcast.c
+  src/f32-gemm/gen-inc/5x16s4inc-minmax-fma3-broadcast.c
   src/f32-gemm/gen-inc/6x8inc-minmax-fma3-broadcast.c
   src/f32-gemm/gen-inc/7x8inc-minmax-fma3-broadcast.c
   src/f32-gemm/gen-inc/8x8inc-minmax-fma3-broadcast.c
-  src/f32-gemm/gen-inc/1x16inc-minmax-fma3-broadcast.c
-  src/f32-gemm/gen-inc/3x16inc-minmax-fma3-broadcast.c
-  src/f32-gemm/gen-inc/4x16inc-minmax-fma3-broadcast.c
-  src/f32-gemm/gen-inc/5x16inc-minmax-fma3-broadcast.c
-  src/f32-gemm/gen-inc/1x16s4inc-minmax-fma3-broadcast.c
-  src/f32-gemm/gen-inc/3x16s4inc-minmax-fma3-broadcast.c
-  src/f32-gemm/gen-inc/4x16s4inc-minmax-fma3-broadcast.c
-  src/f32-gemm/gen-inc/5x16s4inc-minmax-fma3-broadcast.c
+  src/f32-gemm/gen/1x8-minmax-fma3-broadcast.c
+  src/f32-gemm/gen/1x16-minmax-fma3-broadcast.c
+  src/f32-gemm/gen/1x16s4-minmax-fma3-broadcast.c
+  src/f32-gemm/gen/3x16-minmax-fma3-broadcast.c
+  src/f32-gemm/gen/3x16s4-minmax-fma3-broadcast.c
+  src/f32-gemm/gen/4x8-minmax-fma3-broadcast.c
+  src/f32-gemm/gen/4x16-minmax-fma3-broadcast.c
+  src/f32-gemm/gen/4x16s4-minmax-fma3-broadcast.c
+  src/f32-gemm/gen/5x8-minmax-fma3-broadcast.c
+  src/f32-gemm/gen/5x16-minmax-fma3-broadcast.c
+  src/f32-gemm/gen/5x16s4-minmax-fma3-broadcast.c
+  src/f32-gemm/gen/6x8-minmax-fma3-broadcast.c
+  src/f32-gemm/gen/7x8-minmax-fma3-broadcast.c
+  src/f32-gemm/gen/8x8-minmax-fma3-broadcast.c
   src/f32-hswish/gen/hswish-fma3-x8.c
   src/f32-hswish/gen/hswish-fma3-x16.c
   src/f32-igemm/gen/1x8-minmax-fma3-broadcast.c
+  src/f32-igemm/gen/1x16-minmax-fma3-broadcast.c
+  src/f32-igemm/gen/1x16s4-minmax-fma3-broadcast.c
+  src/f32-igemm/gen/3x16-minmax-fma3-broadcast.c
+  src/f32-igemm/gen/3x16s4-minmax-fma3-broadcast.c
   src/f32-igemm/gen/4x8-minmax-fma3-broadcast.c
+  src/f32-igemm/gen/4x16-minmax-fma3-broadcast.c
+  src/f32-igemm/gen/4x16s4-minmax-fma3-broadcast.c
   src/f32-igemm/gen/5x8-minmax-fma3-broadcast.c
+  src/f32-igemm/gen/5x16-minmax-fma3-broadcast.c
+  src/f32-igemm/gen/5x16s4-minmax-fma3-broadcast.c
   src/f32-igemm/gen/6x8-minmax-fma3-broadcast.c
   src/f32-igemm/gen/7x8-minmax-fma3-broadcast.c
   src/f32-igemm/gen/8x8-minmax-fma3-broadcast.c
-  src/f32-igemm/gen/1x16-minmax-fma3-broadcast.c
-  src/f32-igemm/gen/3x16-minmax-fma3-broadcast.c
-  src/f32-igemm/gen/4x16-minmax-fma3-broadcast.c
-  src/f32-igemm/gen/5x16-minmax-fma3-broadcast.c
-  src/f32-igemm/gen/1x16s4-minmax-fma3-broadcast.c
-  src/f32-igemm/gen/3x16s4-minmax-fma3-broadcast.c
-  src/f32-igemm/gen/4x16s4-minmax-fma3-broadcast.c
-  src/f32-igemm/gen/5x16s4-minmax-fma3-broadcast.c
   src/f32-vsqrt/gen/fma3-nr1fma1adj-x8.c
   src/f32-vsqrt/gen/fma3-nr1fma1adj-x16.c
   src/f32-vsqrt/gen/fma3-nr1fma1adj-x24.c
@@ -1842,48 +1842,48 @@
   src/f32-vsqrt/gen/fma3-nr1fma1adj-x56.c
   src/f32-vsqrt/gen/fma3-nr1fma1adj-x64.c
   src/math/sqrt-fma3-nr1fma.c
-  src/math/sqrt-fma3-nr2fma.c
-  src/math/sqrt-fma3-nr1fma1adj.c)
+  src/math/sqrt-fma3-nr1fma1adj.c
+  src/math/sqrt-fma3-nr2fma.c)
 
 SET(XNNPACK_AVX2_MICROKERNEL_SRCS
-  src/f32-raddexpminusmax/gen/avx2-p5-x64.c
   src/f32-raddexpminusmax/gen/avx2-p5-x64-acc2.c
   src/f32-raddexpminusmax/gen/avx2-p5-x64-acc4.c
-  src/f32-raddexpminusmax/gen/avx2-p5-x72.c
+  src/f32-raddexpminusmax/gen/avx2-p5-x64.c
   src/f32-raddexpminusmax/gen/avx2-p5-x72-acc3.c
-  src/f32-raddexpminusmax/gen/avx2-p5-x80.c
+  src/f32-raddexpminusmax/gen/avx2-p5-x72.c
   src/f32-raddexpminusmax/gen/avx2-p5-x80-acc2.c
   src/f32-raddexpminusmax/gen/avx2-p5-x80-acc5.c
-  src/f32-raddexpminusmax/gen/avx2-p5-x96.c
+  src/f32-raddexpminusmax/gen/avx2-p5-x80.c
   src/f32-raddexpminusmax/gen/avx2-p5-x96-acc2.c
   src/f32-raddexpminusmax/gen/avx2-p5-x96-acc3.c
   src/f32-raddexpminusmax/gen/avx2-p5-x96-acc6.c
-  src/f32-raddextexp/gen/avx2-p5-x64.c
+  src/f32-raddexpminusmax/gen/avx2-p5-x96.c
   src/f32-raddextexp/gen/avx2-p5-x64-acc2.c
   src/f32-raddextexp/gen/avx2-p5-x64-acc4.c
-  src/f32-raddextexp/gen/avx2-p5-x72.c
+  src/f32-raddextexp/gen/avx2-p5-x64.c
   src/f32-raddextexp/gen/avx2-p5-x72-acc3.c
-  src/f32-raddextexp/gen/avx2-p5-x80.c
+  src/f32-raddextexp/gen/avx2-p5-x72.c
   src/f32-raddextexp/gen/avx2-p5-x80-acc2.c
   src/f32-raddextexp/gen/avx2-p5-x80-acc5.c
-  src/f32-raddextexp/gen/avx2-p5-x96.c
+  src/f32-raddextexp/gen/avx2-p5-x80.c
   src/f32-raddextexp/gen/avx2-p5-x96-acc2.c
   src/f32-raddextexp/gen/avx2-p5-x96-acc3.c
   src/f32-raddextexp/gen/avx2-p5-x96-acc6.c
-  src/f32-raddstoreexpminusmax/gen/avx2-p5-x64.c
+  src/f32-raddextexp/gen/avx2-p5-x96.c
   src/f32-raddstoreexpminusmax/gen/avx2-p5-x64-acc2.c
   src/f32-raddstoreexpminusmax/gen/avx2-p5-x64-acc4.c
-  src/f32-raddstoreexpminusmax/gen/avx2-p5-x72.c
+  src/f32-raddstoreexpminusmax/gen/avx2-p5-x64.c
   src/f32-raddstoreexpminusmax/gen/avx2-p5-x72-acc3.c
-  src/f32-raddstoreexpminusmax/gen/avx2-p5-x80.c
+  src/f32-raddstoreexpminusmax/gen/avx2-p5-x72.c
   src/f32-raddstoreexpminusmax/gen/avx2-p5-x80-acc2.c
   src/f32-raddstoreexpminusmax/gen/avx2-p5-x80-acc5.c
-  src/f32-raddstoreexpminusmax/gen/avx2-p5-x96.c
+  src/f32-raddstoreexpminusmax/gen/avx2-p5-x80.c
   src/f32-raddstoreexpminusmax/gen/avx2-p5-x96-acc2.c
   src/f32-raddstoreexpminusmax/gen/avx2-p5-x96-acc3.c
   src/f32-raddstoreexpminusmax/gen/avx2-p5-x96-acc6.c
-  src/f32-relu/gen/avx-x16.c
+  src/f32-raddstoreexpminusmax/gen/avx2-p5-x96.c
   src/f32-relu/gen/avx-x8.c
+  src/f32-relu/gen/avx-x16.c
   src/f32-sigmoid/gen/avx2-rr1-p5-div-x8.c
   src/f32-sigmoid/gen/avx2-rr1-p5-div-x16.c
   src/f32-sigmoid/gen/avx2-rr1-p5-div-x24.c
@@ -1938,17 +1938,36 @@
   src/f32-vscaleextexp/gen/avx2-p5-x80.c
   src/f32-vscaleextexp/gen/avx2-p5-x88.c
   src/f32-vscaleextexp/gen/avx2-p5-x96.c
-  src/qs8-dwconv/gen/up16x9-minmax-avx2-mul16.c
-  src/qs8-dwconv/gen/up32x9-minmax-avx2-mul16.c
+  src/math/exp-avx2-lut8-p3-perm.c
+  src/math/exp-avx2-lut8-p4-perm.c
+  src/math/exp-avx2-p5.c
+  src/math/expminus-avx2-p5.c
+  src/math/extexp-avx2-p5.c
+  src/math/sigmoid-avx2-rr1-lut64-p2-gather-div.c
+  src/math/sigmoid-avx2-rr1-lut64-p2-gather-nr1fma.c
+  src/math/sigmoid-avx2-rr1-lut64-p2-gather-nr2fma.c
+  src/math/sigmoid-avx2-rr1-lut64-p2-gather-nr2fma1adj.c
+  src/math/sigmoid-avx2-rr1-p5-div.c
+  src/math/sigmoid-avx2-rr1-p5-nr1fma.c
+  src/math/sigmoid-avx2-rr1-p5-nr2fma.c
+  src/math/sigmoid-avx2-rr2-lut64-p2-gather-div.c
+  src/math/sigmoid-avx2-rr2-lut64-p2-gather-nr1fma.c
+  src/math/sigmoid-avx2-rr2-lut64-p2-gather-nr2fma.c
+  src/math/sigmoid-avx2-rr2-lut64-p2-gather-nr2fma1adj.c
+  src/math/sigmoid-avx2-rr2-p5-div.c
+  src/math/sigmoid-avx2-rr2-p5-nr1fma.c
+  src/math/sigmoid-avx2-rr2-p5-nr2fma.c
   src/qs8-dwconv/gen/up8x9-minmax-avx2-mul32.c
+  src/qs8-dwconv/gen/up16x9-minmax-avx2-mul16.c
   src/qs8-dwconv/gen/up16x9-minmax-avx2-mul32.c
   src/qs8-dwconv/gen/up24x9-minmax-avx2-mul32.c
+  src/qs8-dwconv/gen/up32x9-minmax-avx2-mul16.c
   src/qs8-dwconv/gen/up32x9-minmax-avx2-mul32.c
   src/qs8-gemm/gen/1x8c8-minmax-avx2.c
-  src/qs8-gemm/gen/2x8c8-minmax-avx2.c
-  src/qs8-gemm/gen/3x8c8-minmax-avx2.c
   src/qs8-gemm/gen/1x8c8-xw-minmax-avx2.c
+  src/qs8-gemm/gen/2x8c8-minmax-avx2.c
   src/qs8-gemm/gen/2x8c8-xw-minmax-avx2.c
+  src/qs8-gemm/gen/3x8c8-minmax-avx2.c
   src/qs8-gemm/gen/3x8c8-xw-minmax-avx2.c
   src/qs8-igemm/gen/1x8c8-minmax-avx2.c
   src/qs8-igemm/gen/2x8c8-minmax-avx2.c
@@ -1960,54 +1979,35 @@
   src/qs8-vaddc/gen/minmax-avx2-mul32-ld64-x8.c
   src/qs8-vaddc/gen/minmax-avx2-mul32-ld64-x16.c
   src/qs8-vaddc/gen/minmax-avx2-mul32-ld64-x24.c
-  src/qs8-vaddc/gen/minmax-avx2-mul32-ld64-x32.c
-  src/math/exp-avx2-p5.c
-  src/math/exp-avx2-lut8-p3-perm.c
-  src/math/exp-avx2-lut8-p4-perm.c
-  src/math/expminus-avx2-p5.c
-  src/math/extexp-avx2-p5.c
-  src/math/sigmoid-avx2-rr2-p5-div.c
-  src/math/sigmoid-avx2-rr1-p5-div.c
-  src/math/sigmoid-avx2-rr2-p5-nr2fma.c
-  src/math/sigmoid-avx2-rr1-p5-nr2fma.c
-  src/math/sigmoid-avx2-rr2-p5-nr1fma.c
-  src/math/sigmoid-avx2-rr1-p5-nr1fma.c
-  src/math/sigmoid-avx2-rr1-lut64-p2-gather-div.c
-  src/math/sigmoid-avx2-rr2-lut64-p2-gather-div.c
-  src/math/sigmoid-avx2-rr1-lut64-p2-gather-nr2fma1adj.c
-  src/math/sigmoid-avx2-rr2-lut64-p2-gather-nr2fma1adj.c
-  src/math/sigmoid-avx2-rr1-lut64-p2-gather-nr2fma.c
-  src/math/sigmoid-avx2-rr2-lut64-p2-gather-nr2fma.c
-  src/math/sigmoid-avx2-rr1-lut64-p2-gather-nr1fma.c
-  src/math/sigmoid-avx2-rr2-lut64-p2-gather-nr1fma.c)
+  src/qs8-vaddc/gen/minmax-avx2-mul32-ld64-x32.c)
 
 SET(XNNPACK_AVX512F_MICROKERNEL_SRCS
   src/f32-clamp/gen/avx512f-x16.c
   src/f32-clamp/gen/avx512f-x32.c
-  src/f32-dwconv/gen/up32x4-minmax-avx512f-acc2.c
-  src/f32-dwconv/gen/up32x4-minmax-avx512f.c
   src/f32-dwconv/gen/up16x4-minmax-avx512f-acc2.c
   src/f32-dwconv/gen/up16x4-minmax-avx512f.c
-  src/f32-dwconv/gen/up32x9-minmax-avx512f-acc2.c
-  src/f32-dwconv/gen/up32x9-minmax-avx512f.c
   src/f32-dwconv/gen/up16x9-minmax-avx512f-acc2.c
   src/f32-dwconv/gen/up16x9-minmax-avx512f.c
-  src/f32-dwconv/gen/up32x25-minmax-avx512f-acc2.c
-  src/f32-dwconv/gen/up32x25-minmax-avx512f.c
   src/f32-dwconv/gen/up16x25-minmax-avx512f-acc2.c
   src/f32-dwconv/gen/up16x25-minmax-avx512f.c
-  src/f32-gemm/gen/1x16-minmax-avx512f-broadcast.c
-  src/f32-gemm/gen/4x16-minmax-avx512f-broadcast.c
-  src/f32-gemm/gen/5x16-minmax-avx512f-broadcast.c
-  src/f32-gemm/gen/6x16-minmax-avx512f-broadcast.c
-  src/f32-gemm/gen/7x16-minmax-avx512f-broadcast.c
-  src/f32-gemm/gen/8x16-minmax-avx512f-broadcast.c
+  src/f32-dwconv/gen/up32x4-minmax-avx512f-acc2.c
+  src/f32-dwconv/gen/up32x4-minmax-avx512f.c
+  src/f32-dwconv/gen/up32x9-minmax-avx512f-acc2.c
+  src/f32-dwconv/gen/up32x9-minmax-avx512f.c
+  src/f32-dwconv/gen/up32x25-minmax-avx512f-acc2.c
+  src/f32-dwconv/gen/up32x25-minmax-avx512f.c
   src/f32-gemm/gen-inc/1x16inc-minmax-avx512f-broadcast.c
   src/f32-gemm/gen-inc/4x16inc-minmax-avx512f-broadcast.c
   src/f32-gemm/gen-inc/5x16inc-minmax-avx512f-broadcast.c
   src/f32-gemm/gen-inc/6x16inc-minmax-avx512f-broadcast.c
   src/f32-gemm/gen-inc/7x16inc-minmax-avx512f-broadcast.c
   src/f32-gemm/gen-inc/8x16inc-minmax-avx512f-broadcast.c
+  src/f32-gemm/gen/1x16-minmax-avx512f-broadcast.c
+  src/f32-gemm/gen/4x16-minmax-avx512f-broadcast.c
+  src/f32-gemm/gen/5x16-minmax-avx512f-broadcast.c
+  src/f32-gemm/gen/6x16-minmax-avx512f-broadcast.c
+  src/f32-gemm/gen/7x16-minmax-avx512f-broadcast.c
+  src/f32-gemm/gen/8x16-minmax-avx512f-broadcast.c
   src/f32-hswish/gen/hswish-avx512f-x16.c
   src/f32-hswish/gen/hswish-avx512f-x32.c
   src/f32-igemm/gen/1x16-minmax-avx512f-broadcast.c
@@ -2016,65 +2016,47 @@
   src/f32-igemm/gen/6x16-minmax-avx512f-broadcast.c
   src/f32-igemm/gen/7x16-minmax-avx512f-broadcast.c
   src/f32-igemm/gen/8x16-minmax-avx512f-broadcast.c
-  src/f32-vlrelu/gen/vlrelu-avx512f-x16.c
-  src/f32-vlrelu/gen/vlrelu-avx512f-x32.c
   src/f32-prelu/gen/avx512f-2x16.c
   src/f32-prelu/gen/avx512f-2x32.c
-  src/f32-raddexpminusmax/gen/avx512f-p5-scalef-x128.c
   src/f32-raddexpminusmax/gen/avx512f-p5-scalef-x128-acc2.c
   src/f32-raddexpminusmax/gen/avx512f-p5-scalef-x128-acc4.c
-  src/f32-raddexpminusmax/gen/avx512f-p5-scalef-x144.c
+  src/f32-raddexpminusmax/gen/avx512f-p5-scalef-x128.c
   src/f32-raddexpminusmax/gen/avx512f-p5-scalef-x144-acc3.c
-  src/f32-raddexpminusmax/gen/avx512f-p5-scalef-x160.c
+  src/f32-raddexpminusmax/gen/avx512f-p5-scalef-x144.c
   src/f32-raddexpminusmax/gen/avx512f-p5-scalef-x160-acc2.c
   src/f32-raddexpminusmax/gen/avx512f-p5-scalef-x160-acc5.c
-  src/f32-raddexpminusmax/gen/avx512f-p5-scalef-x192.c
+  src/f32-raddexpminusmax/gen/avx512f-p5-scalef-x160.c
   src/f32-raddexpminusmax/gen/avx512f-p5-scalef-x192-acc2.c
   src/f32-raddexpminusmax/gen/avx512f-p5-scalef-x192-acc3.c
   src/f32-raddexpminusmax/gen/avx512f-p5-scalef-x192-acc6.c
-  src/f32-raddextexp/gen/avx512f-p5-scalef-x128.c
+  src/f32-raddexpminusmax/gen/avx512f-p5-scalef-x192.c
   src/f32-raddextexp/gen/avx512f-p5-scalef-x128-acc2.c
   src/f32-raddextexp/gen/avx512f-p5-scalef-x128-acc4.c
-  src/f32-raddextexp/gen/avx512f-p5-scalef-x144.c
+  src/f32-raddextexp/gen/avx512f-p5-scalef-x128.c
   src/f32-raddextexp/gen/avx512f-p5-scalef-x144-acc3.c
-  src/f32-raddextexp/gen/avx512f-p5-scalef-x160.c
+  src/f32-raddextexp/gen/avx512f-p5-scalef-x144.c
   src/f32-raddextexp/gen/avx512f-p5-scalef-x160-acc2.c
   src/f32-raddextexp/gen/avx512f-p5-scalef-x160-acc5.c
-  src/f32-raddextexp/gen/avx512f-p5-scalef-x192.c
+  src/f32-raddextexp/gen/avx512f-p5-scalef-x160.c
   src/f32-raddextexp/gen/avx512f-p5-scalef-x192-acc2.c
   src/f32-raddextexp/gen/avx512f-p5-scalef-x192-acc3.c
   src/f32-raddextexp/gen/avx512f-p5-scalef-x192-acc6.c
-  src/f32-raddstoreexpminusmax/gen/avx512f-p5-scalef-x128.c
+  src/f32-raddextexp/gen/avx512f-p5-scalef-x192.c
   src/f32-raddstoreexpminusmax/gen/avx512f-p5-scalef-x128-acc2.c
   src/f32-raddstoreexpminusmax/gen/avx512f-p5-scalef-x128-acc4.c
-  src/f32-raddstoreexpminusmax/gen/avx512f-p5-scalef-x144.c
+  src/f32-raddstoreexpminusmax/gen/avx512f-p5-scalef-x128.c
   src/f32-raddstoreexpminusmax/gen/avx512f-p5-scalef-x144-acc3.c
-  src/f32-raddstoreexpminusmax/gen/avx512f-p5-scalef-x160.c
+  src/f32-raddstoreexpminusmax/gen/avx512f-p5-scalef-x144.c
   src/f32-raddstoreexpminusmax/gen/avx512f-p5-scalef-x160-acc2.c
   src/f32-raddstoreexpminusmax/gen/avx512f-p5-scalef-x160-acc5.c
-  src/f32-raddstoreexpminusmax/gen/avx512f-p5-scalef-x192.c
+  src/f32-raddstoreexpminusmax/gen/avx512f-p5-scalef-x160.c
   src/f32-raddstoreexpminusmax/gen/avx512f-p5-scalef-x192-acc2.c
   src/f32-raddstoreexpminusmax/gen/avx512f-p5-scalef-x192-acc3.c
   src/f32-raddstoreexpminusmax/gen/avx512f-p5-scalef-x192-acc6.c
-  src/f32-rmax/avx512f.c
+  src/f32-raddstoreexpminusmax/gen/avx512f-p5-scalef-x192.c
   src/f32-relu/gen/avx512f-x16.c
   src/f32-relu/gen/avx512f-x32.c
-  src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-div-x16.c
-  src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-div-x32.c
-  src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-div-x48.c
-  src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-div-x64.c
-  src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-div-x80.c
-  src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-div-x96.c
-  src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-div-x112.c
-  src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-div-x128.c
-  src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-nr1fma-x16.c
-  src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-nr1fma-x32.c
-  src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-nr1fma-x48.c
-  src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-nr1fma-x64.c
-  src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-nr1fma-x80.c
-  src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-nr1fma-x96.c
-  src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-nr1fma-x112.c
-  src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-nr1fma-x128.c
+  src/f32-rmax/avx512f.c
   src/f32-sigmoid/gen/avx512f-rr1-lut16-p3-perm-scalef-div-x16.c
   src/f32-sigmoid/gen/avx512f-rr1-lut16-p3-perm-scalef-div-x32.c
   src/f32-sigmoid/gen/avx512f-rr1-lut16-p3-perm-scalef-div-x48.c
@@ -2091,6 +2073,22 @@
   src/f32-sigmoid/gen/avx512f-rr1-lut16-p3-perm-scalef-nr1fma-x96.c
   src/f32-sigmoid/gen/avx512f-rr1-lut16-p3-perm-scalef-nr1fma-x112.c
   src/f32-sigmoid/gen/avx512f-rr1-lut16-p3-perm-scalef-nr1fma-x128.c
+  src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-div-x16.c
+  src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-div-x32.c
+  src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-div-x48.c
+  src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-div-x64.c
+  src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-div-x80.c
+  src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-div-x96.c
+  src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-div-x112.c
+  src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-div-x128.c
+  src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-nr1fma-x16.c
+  src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-nr1fma-x32.c
+  src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-nr1fma-x48.c
+  src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-nr1fma-x64.c
+  src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-nr1fma-x80.c
+  src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-nr1fma-x96.c
+  src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-nr1fma-x112.c
+  src/f32-sigmoid/gen/avx512f-rr1-p5-scalef-nr1fma-x128.c
   src/f32-sigmoid/gen/avx512f-rr2-lut32-p2-perm2-scalef-div-x16.c
   src/f32-sigmoid/gen/avx512f-rr2-lut32-p2-perm2-scalef-div-x32.c
   src/f32-sigmoid/gen/avx512f-rr2-lut32-p2-perm2-scalef-div-x48.c
@@ -2139,14 +2137,16 @@
   src/f32-vbinary/gen/vsub-minmax-avx512f-x32.c
   src/f32-vbinary/gen/vsubc-minmax-avx512f-x16.c
   src/f32-vbinary/gen/vsubc-minmax-avx512f-x32.c
-  src/f32-vrnd/gen/vrndne-avx512f-x16.c
-  src/f32-vrnd/gen/vrndne-avx512f-x32.c
-  src/f32-vrnd/gen/vrndz-avx512f-x16.c
-  src/f32-vrnd/gen/vrndz-avx512f-x32.c
-  src/f32-vrnd/gen/vrndu-avx512f-x16.c
-  src/f32-vrnd/gen/vrndu-avx512f-x32.c
+  src/f32-vlrelu/gen/vlrelu-avx512f-x16.c
+  src/f32-vlrelu/gen/vlrelu-avx512f-x32.c
   src/f32-vrnd/gen/vrndd-avx512f-x16.c
   src/f32-vrnd/gen/vrndd-avx512f-x32.c
+  src/f32-vrnd/gen/vrndne-avx512f-x16.c
+  src/f32-vrnd/gen/vrndne-avx512f-x32.c
+  src/f32-vrnd/gen/vrndu-avx512f-x16.c
+  src/f32-vrnd/gen/vrndu-avx512f-x32.c
+  src/f32-vrnd/gen/vrndz-avx512f-x16.c
+  src/f32-vrnd/gen/vrndz-avx512f-x32.c
   src/f32-vscale/avx512f-unroll64.c
   src/f32-vscaleexpminusmax/gen/avx512f-p5-scalef-x16.c
   src/f32-vscaleexpminusmax/gen/avx512f-p5-scalef-x32.c
@@ -2186,40 +2186,40 @@
   src/f32-vunary/gen/vneg-avx512f-x32.c
   src/f32-vunary/gen/vsqr-avx512f-x16.c
   src/f32-vunary/gen/vsqr-avx512f-x32.c
-  src/math/exp-avx512f-p5.c
-  src/math/exp-avx512f-p5-scalef.c
-  src/math/exp-avx512f-lut16-p3-perm.c
   src/math/exp-avx512f-lut16-p3-perm-scalef.c
-  src/math/exp-avx512f-lut32-p2-perm2.c
+  src/math/exp-avx512f-lut16-p3-perm.c
   src/math/exp-avx512f-lut32-p2-perm2-scalef.c
+  src/math/exp-avx512f-lut32-p2-perm2.c
+  src/math/exp-avx512f-p5-scalef.c
+  src/math/exp-avx512f-p5.c
   src/math/extexp-avx512f-p5.c
-  src/math/sigmoid-avx512f-rr1-p5-scalef-div.c
-  src/math/sigmoid-avx512f-rr1-p5-scalef-nr1fma1adj.c
-  src/math/sigmoid-avx512f-rr1-p5-scalef-nr1fma.c
-  src/math/sigmoid-avx512f-rr2-p5-scalef-div.c
-  src/math/sigmoid-avx512f-rr2-p5-scalef-nr1fma1adj.c
-  src/math/sigmoid-avx512f-rr2-p5-scalef-nr1fma.c
   src/math/sigmoid-avx512f-rr1-lut16-p3-perm-scalef-div.c
-  src/math/sigmoid-avx512f-rr1-lut16-p3-perm-scalef-nr1fma1adj.c
   src/math/sigmoid-avx512f-rr1-lut16-p3-perm-scalef-nr1fma.c
-  src/math/sigmoid-avx512f-rr2-lut16-p3-perm-scalef-div.c
-  src/math/sigmoid-avx512f-rr2-lut16-p3-perm-scalef-nr1fma1adj.c
-  src/math/sigmoid-avx512f-rr2-lut16-p3-perm-scalef-nr1fma.c
+  src/math/sigmoid-avx512f-rr1-lut16-p3-perm-scalef-nr1fma1adj.c
   src/math/sigmoid-avx512f-rr1-lut32-p2-perm2-scalef-div.c
-  src/math/sigmoid-avx512f-rr1-lut32-p2-perm2-scalef-nr1fma1adj.c
   src/math/sigmoid-avx512f-rr1-lut32-p2-perm2-scalef-nr1fma.c
-  src/math/sigmoid-avx512f-rr2-lut32-p2-perm2-scalef-div.c
-  src/math/sigmoid-avx512f-rr2-lut32-p2-perm2-scalef-nr1fma1adj.c
-  src/math/sigmoid-avx512f-rr2-lut32-p2-perm2-scalef-nr1fma.c
+  src/math/sigmoid-avx512f-rr1-lut32-p2-perm2-scalef-nr1fma1adj.c
   src/math/sigmoid-avx512f-rr1-lut64-p2-gather-scalef-div.c
-  src/math/sigmoid-avx512f-rr1-lut64-p2-gather-scalef-nr1fma1adj.c
   src/math/sigmoid-avx512f-rr1-lut64-p2-gather-scalef-nr1fma.c
+  src/math/sigmoid-avx512f-rr1-lut64-p2-gather-scalef-nr1fma1adj.c
+  src/math/sigmoid-avx512f-rr1-p5-scalef-div.c
+  src/math/sigmoid-avx512f-rr1-p5-scalef-nr1fma.c
+  src/math/sigmoid-avx512f-rr1-p5-scalef-nr1fma1adj.c
+  src/math/sigmoid-avx512f-rr2-lut16-p3-perm-scalef-div.c
+  src/math/sigmoid-avx512f-rr2-lut16-p3-perm-scalef-nr1fma.c
+  src/math/sigmoid-avx512f-rr2-lut16-p3-perm-scalef-nr1fma1adj.c
+  src/math/sigmoid-avx512f-rr2-lut32-p2-perm2-scalef-div.c
+  src/math/sigmoid-avx512f-rr2-lut32-p2-perm2-scalef-nr1fma.c
+  src/math/sigmoid-avx512f-rr2-lut32-p2-perm2-scalef-nr1fma1adj.c
   src/math/sigmoid-avx512f-rr2-lut64-p2-gather-scalef-div.c
-  src/math/sigmoid-avx512f-rr2-lut64-p2-gather-scalef-nr1fma1adj.c
   src/math/sigmoid-avx512f-rr2-lut64-p2-gather-scalef-nr1fma.c
+  src/math/sigmoid-avx512f-rr2-lut64-p2-gather-scalef-nr1fma1adj.c
+  src/math/sigmoid-avx512f-rr2-p5-scalef-div.c
+  src/math/sigmoid-avx512f-rr2-p5-scalef-nr1fma.c
+  src/math/sigmoid-avx512f-rr2-p5-scalef-nr1fma1adj.c
   src/math/sqrt-avx512f-nr1fma.c
-  src/math/sqrt-avx512f-nr2fma.c
-  src/math/sqrt-avx512f-nr1fma1adj.c)
+  src/math/sqrt-avx512f-nr1fma1adj.c
+  src/math/sqrt-avx512f-nr2fma.c)
 
 SET(XNNPACK_AVX512SKX_MICROKERNEL_SRCS
   src/qs8-dwconv/gen/up16x9-minmax-avx512skx-mul32.c
@@ -2250,78 +2250,78 @@
   src/f32-igemm/gen/4x8-minmax-aarch32-neon-pld-cortex-a75.S)
 
 SET(XNNPACK_AARCH64_ASM_MICROKERNEL_SRCS
-  src/f16-gemm/gen/1x16-minmax-aarch64-neonfp16arith-ld32.S
-  src/f16-gemm/gen/4x16-minmax-aarch64-neonfp16arith-ld32.S
-  src/f16-gemm/gen/6x16-minmax-aarch64-neonfp16arith-ld32.S
-  src/f16-gemm/gen-inc/1x16inc-minmax-aarch64-neonfp16arith-ld32.S
-  src/f16-gemm/gen-inc/4x16inc-minmax-aarch64-neonfp16arith-ld32.S
-  src/f16-gemm/gen-inc/6x16inc-minmax-aarch64-neonfp16arith-ld32.S
-  src/f16-gemm/gen/1x8-minmax-aarch64-neonfp16arith-ld64.S
-  src/f16-gemm/gen/4x8-minmax-aarch64-neonfp16arith-ld64.S
-  src/f16-gemm/gen/6x8-minmax-aarch64-neonfp16arith-ld64.S
-  src/f16-gemm/gen/8x8-minmax-aarch64-neonfp16arith-ld64.S
   src/f16-gemm/gen-inc/1x8inc-minmax-aarch64-neonfp16arith-ld64.S
+  src/f16-gemm/gen-inc/1x16inc-minmax-aarch64-neonfp16arith-ld32.S
   src/f16-gemm/gen-inc/4x8inc-minmax-aarch64-neonfp16arith-ld64.S
+  src/f16-gemm/gen-inc/4x16inc-minmax-aarch64-neonfp16arith-ld32.S
   src/f16-gemm/gen-inc/6x8inc-minmax-aarch64-neonfp16arith-ld64.S
+  src/f16-gemm/gen-inc/6x16inc-minmax-aarch64-neonfp16arith-ld32.S
   src/f16-gemm/gen-inc/8x8inc-minmax-aarch64-neonfp16arith-ld64.S
+  src/f16-gemm/gen/1x8-minmax-aarch64-neonfp16arith-ld64.S
+  src/f16-gemm/gen/1x16-minmax-aarch64-neonfp16arith-ld32.S
+  src/f16-gemm/gen/4x8-minmax-aarch64-neonfp16arith-ld64.S
+  src/f16-gemm/gen/4x16-minmax-aarch64-neonfp16arith-ld32.S
+  src/f16-gemm/gen/6x8-minmax-aarch64-neonfp16arith-ld64.S
+  src/f16-gemm/gen/6x16-minmax-aarch64-neonfp16arith-ld32.S
+  src/f16-gemm/gen/8x8-minmax-aarch64-neonfp16arith-ld64.S
   src/f32-dwconv/up4x9-minmax-aarch64-neonfma-cortex-a55.S
   src/f32-dwconv/up4x9-minmax-aarch64-neonfma.S
-  src/f32-gemm/gen/1x8-minmax-aarch64-neonfma-ld64.S
-  src/f32-gemm/gen/1x12-minmax-aarch64-neonfma-cortex-a53.S
-  src/f32-gemm/gen/1x8-minmax-aarch64-neonfma-cortex-a53.S
-  src/f32-gemm/gen/1x8-minmax-aarch64-neonfma-cortex-a57.S
-  src/f32-gemm/gen/1x8-minmax-aarch64-neonfma-cortex-a75.S
-  src/f32-gemm/gen/4x12-minmax-aarch64-neonfma-cortex-a53.S
-  src/f32-gemm/gen/4x8-minmax-aarch64-neonfma-cortex-a53.S
-  src/f32-gemm/gen/4x8-minmax-aarch64-neonfma-cortex-a55.S
-  src/f32-gemm/gen/4x8-minmax-aarch64-neonfma-cortex-a57.S
-  src/f32-gemm/gen/4x8-minmax-aarch64-neonfma-cortex-a75.S
-  src/f32-gemm/gen/4x8-minmax-aarch64-neonfma-ld128.S
-  src/f32-gemm/gen/4x8-minmax-aarch64-neonfma-ld64.S
-  src/f32-gemm/gen/5x8-minmax-aarch64-neonfma-cortex-a57.S
-  src/f32-gemm/gen/5x8-minmax-aarch64-neonfma-cortex-a75.S
-  src/f32-gemm/gen/6x8-minmax-aarch64-neonfma-cortex-a53.S
-  src/f32-gemm/gen/6x8-minmax-aarch64-neonfma-cortex-a55.S
-  src/f32-gemm/gen/6x8-minmax-aarch64-neonfma-cortex-a73.S
-  src/f32-gemm/gen/6x8-minmax-aarch64-neonfma-cortex-a57.S
-  src/f32-gemm/gen/6x8-minmax-aarch64-neonfma-cortex-a75.S
-  src/f32-gemm/gen/6x8-minmax-aarch64-neonfma-ld128.S
-  src/f32-gemm/gen/6x8-minmax-aarch64-neonfma-ld64.S
-  src/f32-gemm/gen-inc/1x8inc-minmax-aarch64-neonfma-ld64.S
-  src/f32-gemm/gen-inc/1x12inc-minmax-aarch64-neonfma-cortex-a53.S
   src/f32-gemm/gen-inc/1x8inc-minmax-aarch64-neonfma-cortex-a53.S
   src/f32-gemm/gen-inc/1x8inc-minmax-aarch64-neonfma-cortex-a57.S
   src/f32-gemm/gen-inc/1x8inc-minmax-aarch64-neonfma-cortex-a75.S
-  src/f32-gemm/gen-inc/4x12inc-minmax-aarch64-neonfma-cortex-a53.S
+  src/f32-gemm/gen-inc/1x8inc-minmax-aarch64-neonfma-ld64.S
+  src/f32-gemm/gen-inc/1x12inc-minmax-aarch64-neonfma-cortex-a53.S
   src/f32-gemm/gen-inc/4x8inc-minmax-aarch64-neonfma-cortex-a53.S
   src/f32-gemm/gen-inc/4x8inc-minmax-aarch64-neonfma-cortex-a55.S
   src/f32-gemm/gen-inc/4x8inc-minmax-aarch64-neonfma-cortex-a57.S
   src/f32-gemm/gen-inc/4x8inc-minmax-aarch64-neonfma-cortex-a75.S
-  src/f32-gemm/gen-inc/4x8inc-minmax-aarch64-neonfma-ld128.S
   src/f32-gemm/gen-inc/4x8inc-minmax-aarch64-neonfma-ld64.S
+  src/f32-gemm/gen-inc/4x8inc-minmax-aarch64-neonfma-ld128.S
+  src/f32-gemm/gen-inc/4x12inc-minmax-aarch64-neonfma-cortex-a53.S
   src/f32-gemm/gen-inc/5x8inc-minmax-aarch64-neonfma-cortex-a57.S
   src/f32-gemm/gen-inc/5x8inc-minmax-aarch64-neonfma-cortex-a75.S
   src/f32-gemm/gen-inc/6x8inc-minmax-aarch64-neonfma-cortex-a53.S
   src/f32-gemm/gen-inc/6x8inc-minmax-aarch64-neonfma-cortex-a55.S
-  src/f32-gemm/gen-inc/6x8inc-minmax-aarch64-neonfma-cortex-a73.S
   src/f32-gemm/gen-inc/6x8inc-minmax-aarch64-neonfma-cortex-a57.S
+  src/f32-gemm/gen-inc/6x8inc-minmax-aarch64-neonfma-cortex-a73.S
   src/f32-gemm/gen-inc/6x8inc-minmax-aarch64-neonfma-cortex-a75.S
-  src/f32-gemm/gen-inc/6x8inc-minmax-aarch64-neonfma-ld128.S
   src/f32-gemm/gen-inc/6x8inc-minmax-aarch64-neonfma-ld64.S
-  src/f32-igemm/1x12-minmax-aarch64-neonfma-cortex-a53.S
+  src/f32-gemm/gen-inc/6x8inc-minmax-aarch64-neonfma-ld128.S
+  src/f32-gemm/gen/1x8-minmax-aarch64-neonfma-cortex-a53.S
+  src/f32-gemm/gen/1x8-minmax-aarch64-neonfma-cortex-a57.S
+  src/f32-gemm/gen/1x8-minmax-aarch64-neonfma-cortex-a75.S
+  src/f32-gemm/gen/1x8-minmax-aarch64-neonfma-ld64.S
+  src/f32-gemm/gen/1x12-minmax-aarch64-neonfma-cortex-a53.S
+  src/f32-gemm/gen/4x8-minmax-aarch64-neonfma-cortex-a53.S
+  src/f32-gemm/gen/4x8-minmax-aarch64-neonfma-cortex-a55.S
+  src/f32-gemm/gen/4x8-minmax-aarch64-neonfma-cortex-a57.S
+  src/f32-gemm/gen/4x8-minmax-aarch64-neonfma-cortex-a75.S
+  src/f32-gemm/gen/4x8-minmax-aarch64-neonfma-ld64.S
+  src/f32-gemm/gen/4x8-minmax-aarch64-neonfma-ld128.S
+  src/f32-gemm/gen/4x12-minmax-aarch64-neonfma-cortex-a53.S
+  src/f32-gemm/gen/5x8-minmax-aarch64-neonfma-cortex-a57.S
+  src/f32-gemm/gen/5x8-minmax-aarch64-neonfma-cortex-a75.S
+  src/f32-gemm/gen/6x8-minmax-aarch64-neonfma-cortex-a53.S
+  src/f32-gemm/gen/6x8-minmax-aarch64-neonfma-cortex-a55.S
+  src/f32-gemm/gen/6x8-minmax-aarch64-neonfma-cortex-a57.S
+  src/f32-gemm/gen/6x8-minmax-aarch64-neonfma-cortex-a73.S
+  src/f32-gemm/gen/6x8-minmax-aarch64-neonfma-cortex-a75.S
+  src/f32-gemm/gen/6x8-minmax-aarch64-neonfma-ld64.S
+  src/f32-gemm/gen/6x8-minmax-aarch64-neonfma-ld128.S
   src/f32-igemm/1x8-minmax-aarch64-neonfma-cortex-a53.S
-  src/f32-igemm/gen/1x8-minmax-aarch64-neonfma-cortex-a57.S
-  src/f32-igemm/gen/1x8-minmax-aarch64-neonfma-cortex-a75.S
+  src/f32-igemm/1x12-minmax-aarch64-neonfma-cortex-a53.S
   src/f32-igemm/4x8-minmax-aarch64-neonfma-cortex-a53.S
   src/f32-igemm/4x8-minmax-aarch64-neonfma-cortex-a55.S
   src/f32-igemm/4x12-minmax-aarch64-neonfma-cortex-a53.S
+  src/f32-igemm/6x8-minmax-aarch64-neonfma-cortex-a53.S
+  src/f32-igemm/6x8-minmax-aarch64-neonfma-cortex-a55.S
+  src/f32-igemm/6x8-minmax-aarch64-neonfma-cortex-a73.S
+  src/f32-igemm/gen/1x8-minmax-aarch64-neonfma-cortex-a57.S
+  src/f32-igemm/gen/1x8-minmax-aarch64-neonfma-cortex-a75.S
   src/f32-igemm/gen/4x8-minmax-aarch64-neonfma-cortex-a57.S
   src/f32-igemm/gen/4x8-minmax-aarch64-neonfma-cortex-a75.S
   src/f32-igemm/gen/5x8-minmax-aarch64-neonfma-cortex-a57.S
   src/f32-igemm/gen/5x8-minmax-aarch64-neonfma-cortex-a75.S
-  src/f32-igemm/6x8-minmax-aarch64-neonfma-cortex-a53.S
-  src/f32-igemm/6x8-minmax-aarch64-neonfma-cortex-a55.S
-  src/f32-igemm/6x8-minmax-aarch64-neonfma-cortex-a73.S
   src/f32-igemm/gen/6x8-minmax-aarch64-neonfma-cortex-a57.S
   src/f32-igemm/gen/6x8-minmax-aarch64-neonfma-cortex-a75.S
   src/qs8-gemm/1x16c4-aarch64-neondot-ld64.S