Add 32x1 32x2 32x4 SPMM microkernels and remove 4x1 4x2 4x4 for WASMSIMD, Neon and SSE

PiperOrigin-RevId: 339125492
diff --git a/BUILD.bazel b/BUILD.bazel
index 760dd4f..e2b7202 100644
--- a/BUILD.bazel
+++ b/BUILD.bazel
@@ -915,6 +915,12 @@
     "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-spmm/gen/32x1-minmax-wasmsimd-arm-unroll2.c",
+    "src/f32-spmm/gen/32x1-minmax-wasmsimd-arm-unroll4.c",
+    "src/f32-spmm/gen/32x1-minmax-wasmsimd-arm.c",
+    "src/f32-spmm/gen/32x1-minmax-wasmsimd-x86-unroll2.c",
+    "src/f32-spmm/gen/32x1-minmax-wasmsimd-x86-unroll4.c",
+    "src/f32-spmm/gen/32x1-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",
@@ -1734,6 +1740,11 @@
     "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-spmm/gen/32x1-minmax-neonfma-pipelined.c",
+    "src/f32-spmm/gen/32x1-minmax-neonfma-unroll2.c",
+    "src/f32-spmm/gen/32x1-minmax-neonfma.c",
+    "src/f32-spmm/gen/32x2-minmax-neonfma.c",
+    "src/f32-spmm/gen/32x4-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",
@@ -1949,6 +1960,7 @@
     "src/f32-spmm/gen/4x1-minmax-sse.c",
     "src/f32-spmm/gen/8x1-minmax-sse.c",
     "src/f32-spmm/gen/16x1-minmax-sse.c",
+    "src/f32-spmm/gen/32x1-minmax-sse.c",
     "src/f32-vbinary/gen/vadd-minmax-sse-x4.c",
     "src/f32-vbinary/gen/vadd-minmax-sse-x8.c",
     "src/f32-vbinary/gen/vaddc-minmax-sse-x4.c",