Merge main (CRC32 fix)

sdk/software/examples/asm/user-sample.c

@@ -146,8 +146,6 @@ int main(void)
     /* ── Start marker (must be first output) ──────────────────── */
     printf("MATMUL_START\n");
 
-    unsigned int tick_start = get_ns();
-
     // 清理寄存器状态
     REG_WR(MATMUL_STATUS, STATUS_DONE | STATUS_ERROR); 
 
@@ -187,16 +185,6 @@ int main(void)
 
         // 6. 立即启动当前组在 IP 中的计算
         REG_WR(MATMUL_CTRL, ctrl_start[cur_bank]);
-        
-        // 调试输出逻辑 (通过无缓存指针直接从内存读取刚刚 DMA 写回的数据)
-        // 注意：因为 Channel 1 刚启动，需要等它搬运完才能打印 C[0..2]
-        if (g - 1 == 1) { // 也就是当刚刚计算完 Group 1 并发起 DMA 后
-            dma_wait_polling(1); // 强制等一下，供打印使用
-            printf("DBG C[0..2]=%08lx %08lx %08lx\n",
-                   (unsigned long)extram[DST_BASE_WORDS + 1 * 48 + 0],
-                   (unsigned long)extram[DST_BASE_WORDS + 1 * 48 + 1],
-                   (unsigned long)extram[DST_BASE_WORDS + 1 * 48 + 2]);
-        }
     }
 
     /* ── Epilogue: 收尾最后一组 ───────────────────────────── */
@@ -207,17 +195,6 @@ int main(void)
     dma_start_transfer(1, phys_bank_C[(MATMUL_GROUP_NUM - 1) % 2], EXTRAM_PHYS_C(MATMUL_GROUP_NUM - 1), 192, 20);
     dma_wait_polling(1); // 必须等待最后一组写回完成
 
-    unsigned int tick_end = get_ns();
-
-    // 检查测试输出要求
-    printf("DBG wrote C[0]=%08lx @ ExtRAM[%d]\n",
-            (unsigned long)extram[DST_BASE_WORDS], DST_BASE_WORDS);
-
-    printf("\n--- Performance Data ---\n");
-    unsigned int total_time = tick_end - tick_start;
-    printf("Matmul %d groups total time: %u ns (%.3f ms)\n", 
-           MATMUL_GROUP_NUM, total_time, (float)total_time / 1000000.0);
-
     /* ── CRC32 over result region ─────────────────────────────── */
     crc = compute_crc32(extram);