diff --git a/examples/peripherals/spi_slave/receiver/main/app_main.c b/examples/peripherals/spi_slave/receiver/main/app_main.c
index 8032f2b830..5d1da4ea64 100644
--- a/examples/peripherals/spi_slave/receiver/main/app_main.c
+++ b/examples/peripherals/spi_slave/receiver/main/app_main.c
@@ -100,11 +100,10 @@ void app_main(void)
     ret = spi_slave_initialize(RCV_HOST, &buscfg, &slvcfg, SPI_DMA_CH_AUTO);
     assert(ret == ESP_OK);
 
-    WORD_ALIGNED_ATTR char sendbuf[129] = "";
-    WORD_ALIGNED_ATTR char recvbuf[129] = "";
-    memset(recvbuf, 0, 33);
-    spi_slave_transaction_t t;
-    memset(&t, 0, sizeof(t));
+    char *sendbuf = spi_bus_dma_memory_alloc(RCV_HOST, 129, 0);
+    char *recvbuf = spi_bus_dma_memory_alloc(RCV_HOST, 129, 0);
+    assert(sendbuf && recvbuf);
+    spi_slave_transaction_t t = {0};
 
     while (1) {
         //Clear receive buffer, set send buffer to something sane
diff --git a/examples/peripherals/spi_slave_hd/append_mode/master/main/app_main.c b/examples/peripherals/spi_slave_hd/append_mode/master/main/app_main.c
index b16711641a..480921c982 100644
--- a/examples/peripherals/spi_slave_hd/append_mode/master/main/app_main.c
+++ b/examples/peripherals/spi_slave_hd/append_mode/master/main/app_main.c
@@ -24,7 +24,7 @@
 #define GPIO_CS            10
 
 #define HOST_ID            SPI2_HOST
-#define TRANSACTION_LEN    16
+#define TRANSACTION_LEN    64
 //The SPI transaction cycles in this example. You may change the cycle. e.g., use the ``sender`` and change it to a infinite loop
 #define EXAMPLE_CYCLES     10
 
diff --git a/examples/peripherals/spi_slave_hd/append_mode/slave/main/app_main.c b/examples/peripherals/spi_slave_hd/append_mode/slave/main/app_main.c
index 49676693b3..88065a994d 100644
--- a/examples/peripherals/spi_slave_hd/append_mode/slave/main/app_main.c
+++ b/examples/peripherals/spi_slave_hd/append_mode/slave/main/app_main.c
@@ -25,7 +25,7 @@
 
 #define HOST_ID             SPI2_HOST
 #define QUEUE_SIZE          6
-#define TRANSACTION_LEN     16
+#define TRANSACTION_LEN     64
 #define SYNC_REG_FROM_HOST  (14 * 4)
 #define SYNC_REG_TO_HOST    (15 * 4)
 
@@ -72,7 +72,7 @@ static esp_err_t create_transaction_pool(uint8_t **data_buf, trans_link_t *trans
 {
     for (int i = 0; i < times; i++) {
         //malloc data buffers for transaction
-        data_buf[i] = heap_caps_calloc(1, TRANSACTION_LEN, MALLOC_CAP_DMA);
+        data_buf[i] = spi_bus_dma_memory_alloc(HOST_ID, TRANSACTION_LEN, 0);
         if (!data_buf[i]) {
             ESP_LOGI("Create pool:", "No enough memory");
             return ESP_ERR_NO_MEM;
diff --git a/examples/peripherals/spi_slave_hd/segment_mode/seg_slave/main/app_main.c b/examples/peripherals/spi_slave_hd/segment_mode/seg_slave/main/app_main.c
index ad59703d6b..e13c3b1499 100644
--- a/examples/peripherals/spi_slave_hd/segment_mode/seg_slave/main/app_main.c
+++ b/examples/peripherals/spi_slave_hd/segment_mode/seg_slave/main/app_main.c
@@ -198,7 +198,9 @@ void sender(void *arg)
             data_ready = get_tx_data(send_buf[descriptor_id], send_buf_size, &ready_data_size);
             if (data_ready) {
                 slave_trans[descriptor_id].data = send_buf[descriptor_id];
-                slave_trans[descriptor_id].len = ready_data_size;
+                slave_trans[descriptor_id].len = send_buf_size;
+                //To use dma, data buffer address and trans_len should byte align to hardware requirement, or using following flag for auto deal by driver.
+                slave_trans[descriptor_id].flags |= SPI_SLAVE_HD_TRANS_DMA_BUFFER_ALIGN_AUTO;
                 //Due to the `queue_sent_cnt` and `queue_recv_cnt` logic above, we are sure there is space to send data, this will return ESP_OK immediately
                 ESP_ERROR_CHECK(spi_slave_hd_queue_trans(SLAVE_HOST, SPI_SLAVE_CHAN_TX, &slave_trans[descriptor_id], portMAX_DELAY));
                 descriptor_id = (descriptor_id + 1) % QUEUE_SIZE;   //descriptor_id will be: 0, 1, 2, ..., QUEUE_SIZE, 0, 1, ....
@@ -234,7 +236,7 @@ void receiver(void *arg)
     uint8_t *recv_buf[QUEUE_SIZE];
     spi_slave_hd_data_t slave_trans[QUEUE_SIZE];
     for (int i = 0; i < QUEUE_SIZE; i++) {
-        recv_buf[i] = heap_caps_calloc(1, recv_buf_size, MALLOC_CAP_DMA);
+        recv_buf[i] = spi_bus_dma_memory_alloc(SLAVE_HOST, recv_buf_size, MALLOC_CAP_8BIT);
         if (!recv_buf[i]) {
             ESP_LOGE(TAG, "No enough memory!");
             abort();
@@ -249,6 +251,7 @@ void receiver(void *arg)
     for (int i = 0; i < QUEUE_SIZE; i++) {
         slave_trans[descriptor_id].data = recv_buf[descriptor_id];
         slave_trans[descriptor_id].len = recv_buf_size;
+        slave_trans[descriptor_id].flags |= SPI_SLAVE_HD_TRANS_DMA_BUFFER_ALIGN_AUTO;
         ESP_ERROR_CHECK(spi_slave_hd_queue_trans(SLAVE_HOST, SPI_SLAVE_CHAN_RX, &slave_trans[descriptor_id], portMAX_DELAY));
         descriptor_id = (descriptor_id + 1) % QUEUE_SIZE;   //descriptor_id will be: 0, 1, 2, ..., QUEUE_SIZE, 0, 1, ....
     }
@@ -266,7 +269,6 @@ void receiver(void *arg)
          */
         ESP_ERROR_CHECK(spi_slave_hd_get_trans_res(SLAVE_HOST, SPI_SLAVE_CHAN_RX, &ret_trans, portMAX_DELAY));
         //Process the received data in your own code. Here we just print it out.
-        printf("%d bytes are received: \n%s\n", ret_trans->trans_len, ret_trans->data);
         memset(ret_trans->data, 0x0, recv_buf_size);
 
         /**
@@ -288,10 +290,10 @@ void app_main(void)
     uint8_t init_value[SOC_SPI_MAXIMUM_BUFFER_SIZE] = {0x0};
     spi_slave_hd_write_buffer(SLAVE_HOST, 0, init_value, SOC_SPI_MAXIMUM_BUFFER_SIZE);
 
-    static uint32_t send_buf_size = 5000;
+    static uint32_t send_buf_size = 4800;
     spi_slave_hd_write_buffer(SLAVE_HOST, SLAVE_MAX_TX_BUF_LEN_REG, (uint8_t *)&send_buf_size, sizeof(send_buf_size));
 
-    static uint32_t recv_buf_size = 120;
+    static uint32_t recv_buf_size = 128;
     spi_slave_hd_write_buffer(SLAVE_HOST, SLAVE_MAX_RX_BUF_LEN_REG, (uint8_t *)&recv_buf_size, sizeof(recv_buf_size));
 
     uint32_t slave_ready_flag = SLAVE_READY_FLAG;