feat: add ML-KEM accelerator with AXI slave wrapper

Wrap mlkem_top in an AXI4 burst slave (mlkem_axi_slave) mapped to the
freed FFT slot 6 at 0x1f400000 (virtual 0xbf400000). The core runs on
sys_clk/2 (25 MHz); the wrapper handles the clock-domain crossing:
start toggle->pulse, wait-stated byte-preload handshake, busy/done
sync, and direct readback taps (ek/dk/ct/ss). CPU drives it via a
register map (CTRL/STATUS/CONFIG, d/z/msg, preload + readback windows).

Wire into soc_top on the existing axiOut_6 nets (no crossbar change),
add mlkem_intr on intrpt[3]. Add sdk mlkem_test example running a
KeyGen->Encaps->Decaps self-consistency round trip.

rtl/ip/mlkem/poly_mul/basecase_mul.v

@@ -0,0 +1,45 @@
+// basecase_mul.v - NTT-domain degree-1 polynomial multiplication
+//
+// Computes pointwise product of two degree-1 NTT-domain polynomials:
+//   c0 = (a0*b0 + a1*b1*zeta) mod Q
+//   c1 = (a0*b1 + a1*b0) mod Q
+//
+// Uses three barrett_mul instances for modular multiplications
+// and pure combinational modular addition for the final sums.
+//
+// All inputs/outputs are 12-bit values in [0, Q-1] where Q = 3329.
+
+module basecase_mul (
+    input  [11:0] a0, a1,   // first polynomial: degree-0 and degree-1 coeffs
+    input  [11:0] b0, b1,   // second polynomial: degree-0 and degree-1 coeffs
+    input  [11:0] zeta,     // precomputed twiddle factor (zeta^2 * 17 mod Q)
+    output [11:0] c0, c1    // result polynomial coefficients
+);
+
+    localparam Q = 3329;
+
+    // Barrett modular multiplication outputs
+    wire [11:0] t1;         // a0 * b0 mod Q
+    wire [11:0] t2;         // a1 * b1 mod Q
+    wire [11:0] t3;         // a1 * b0 mod Q
+    wire [11:0] t4;         // a0 * b1 mod Q
+    wire [11:0] t2_zeta;    // (a1 * b1) * zeta mod Q
+
+    // Four Barrett multiplications for the scalar products
+    barrett_mul u_mul1 (.a(a0), .b(b0), .product(t1));
+    barrett_mul u_mul2 (.a(a1), .b(b1), .product(t2));
+    barrett_mul u_mul3 (.a(a1), .b(b0), .product(t3));
+    barrett_mul u_mul4 (.a(a0), .b(b1), .product(t4));
+
+    // Multiply t2 by zeta for the zeta term in c0
+    barrett_mul u_mul_zeta (.a(t2), .b(zeta), .product(t2_zeta));
+
+    // Combinational modular addition: each sum is in [0, 2Q-1)
+    // Subtract Q once if >= Q
+    wire [12:0] sum_c0 = {1'b0, t1} + {1'b0, t2_zeta};
+    wire [12:0] sum_c1 = {1'b0, t3} + {1'b0, t4};
+
+    assign c0 = (sum_c0 >= Q) ? sum_c0[11:0] - Q[11:0] : sum_c0[11:0];
+    assign c1 = (sum_c1 >= Q) ? sum_c1[11:0] - Q[11:0] : sum_c1[11:0];
+
+endmodule