Merge branch 'master' into renesas_ra_hs_rebased

41 files changed, 613 insertions, 585 deletions

diff --git a/src/CMakeLists.txt b/src/CMakeLists.txt
index 8f7f38589..076e1e1eb 100644
--- a/src/CMakeLists.txt
+++ b/src/CMakeLists.txt
@@ -64,6 +64,7 @@ function(add_tinyusb TARGET)
       -Wnull-dereference
       -Wuninitialized
       -Wunused
+      -Wunused-function
       -Wreturn-type
       -Wredundant-decls
       )
diff --git a/src/class/audio/audio_device.c b/src/class/audio/audio_device.c
index e5dbe988a..1a7ce7870 100644
--- a/src/class/audio/audio_device.c
+++ b/src/class/audio/audio_device.c
@@ -113,21 +113,21 @@
 // EP IN software buffers and mutexes
 #if CFG_TUD_AUDIO_ENABLE_EP_IN && !CFG_TUD_AUDIO_ENABLE_ENCODING
   #if CFG_TUD_AUDIO_FUNC_1_EP_IN_SW_BUF_SZ > 0
-    CFG_TUSB_MEM_SECTION CFG_TUSB_MEM_ALIGN uint8_t audio_ep_in_sw_buf_1[CFG_TUD_AUDIO_FUNC_1_EP_IN_SW_BUF_SZ];
+    CFG_TUD_MEM_SECTION CFG_TUSB_MEM_ALIGN uint8_t audio_ep_in_sw_buf_1[CFG_TUD_AUDIO_FUNC_1_EP_IN_SW_BUF_SZ];
     #if CFG_FIFO_MUTEX
     osal_mutex_def_t ep_in_ff_mutex_wr_1; // No need for read mutex as only USB driver reads from FIFO
     #endif
   #endif // CFG_TUD_AUDIO_FUNC_1_EP_IN_SW_BUF_SZ > 0
 
   #if CFG_TUD_AUDIO > 1 && CFG_TUD_AUDIO_FUNC_2_EP_IN_SW_BUF_SZ > 0
-    CFG_TUSB_MEM_SECTION CFG_TUSB_MEM_ALIGN uint8_t audio_ep_in_sw_buf_2[CFG_TUD_AUDIO_FUNC_2_EP_IN_SW_BUF_SZ];
+    CFG_TUD_MEM_SECTION CFG_TUSB_MEM_ALIGN uint8_t audio_ep_in_sw_buf_2[CFG_TUD_AUDIO_FUNC_2_EP_IN_SW_BUF_SZ];
     #if CFG_FIFO_MUTEX
     osal_mutex_def_t ep_in_ff_mutex_wr_2; // No need for read mutex as only USB driver reads from FIFO
     #endif
   #endif // CFG_TUD_AUDIO > 1 && CFG_TUD_AUDIO_FUNC_2_EP_IN_SW_BUF_SZ > 0
 
   #if CFG_TUD_AUDIO > 2 && CFG_TUD_AUDIO_FUNC_3_EP_IN_SW_BUF_SZ > 0
-    CFG_TUSB_MEM_SECTION CFG_TUSB_MEM_ALIGN uint8_t audio_ep_in_sw_buf_3[CFG_TUD_AUDIO_FUNC_3_EP_IN_SW_BUF_SZ];
+    CFG_TUD_MEM_SECTION CFG_TUSB_MEM_ALIGN uint8_t audio_ep_in_sw_buf_3[CFG_TUD_AUDIO_FUNC_3_EP_IN_SW_BUF_SZ];
     #if CFG_FIFO_MUTEX
     osal_mutex_def_t ep_in_ff_mutex_wr_3; // No need for read mutex as only USB driver reads from FIFO
     #endif
@@ -139,36 +139,36 @@
 // - the software encoding is used - in this case the linear buffers serve as a target memory where logical channels are encoded into
 #if CFG_TUD_AUDIO_ENABLE_EP_IN && (USE_LINEAR_BUFFER || CFG_TUD_AUDIO_ENABLE_ENCODING)
   #if CFG_TUD_AUDIO_FUNC_1_EP_IN_SZ_MAX > 0
-    CFG_TUSB_MEM_SECTION CFG_TUSB_MEM_ALIGN uint8_t lin_buf_in_1[CFG_TUD_AUDIO_FUNC_1_EP_IN_SZ_MAX];
+    CFG_TUD_MEM_SECTION CFG_TUSB_MEM_ALIGN uint8_t lin_buf_in_1[CFG_TUD_AUDIO_FUNC_1_EP_IN_SZ_MAX];
   #endif
 
   #if CFG_TUD_AUDIO > 1 && CFG_TUD_AUDIO_FUNC_2_EP_IN_SZ_MAX > 0
-    CFG_TUSB_MEM_SECTION CFG_TUSB_MEM_ALIGN uint8_t lin_buf_in_2[CFG_TUD_AUDIO_FUNC_2_EP_IN_SZ_MAX];
+    CFG_TUD_MEM_SECTION CFG_TUSB_MEM_ALIGN uint8_t lin_buf_in_2[CFG_TUD_AUDIO_FUNC_2_EP_IN_SZ_MAX];
   #endif
 
   #if CFG_TUD_AUDIO > 2 && CFG_TUD_AUDIO_FUNC_3_EP_IN_SZ_MAX > 0
-    CFG_TUSB_MEM_SECTION CFG_TUSB_MEM_ALIGN uint8_t lin_buf_in_3[CFG_TUD_AUDIO_FUNC_3_EP_IN_SZ_MAX];
+    CFG_TUD_MEM_SECTION CFG_TUSB_MEM_ALIGN uint8_t lin_buf_in_3[CFG_TUD_AUDIO_FUNC_3_EP_IN_SZ_MAX];
   #endif
 #endif // CFG_TUD_AUDIO_ENABLE_EP_IN && (USE_LINEAR_BUFFER || CFG_TUD_AUDIO_ENABLE_DECODING)
 
 // EP OUT software buffers and mutexes
 #if CFG_TUD_AUDIO_ENABLE_EP_OUT && !CFG_TUD_AUDIO_ENABLE_DECODING
   #if CFG_TUD_AUDIO_FUNC_1_EP_OUT_SW_BUF_SZ > 0
-    CFG_TUSB_MEM_SECTION CFG_TUSB_MEM_ALIGN uint8_t audio_ep_out_sw_buf_1[CFG_TUD_AUDIO_FUNC_1_EP_OUT_SW_BUF_SZ];
+    CFG_TUD_MEM_SECTION CFG_TUSB_MEM_ALIGN uint8_t audio_ep_out_sw_buf_1[CFG_TUD_AUDIO_FUNC_1_EP_OUT_SW_BUF_SZ];
     #if CFG_FIFO_MUTEX
     osal_mutex_def_t ep_out_ff_mutex_rd_1; // No need for write mutex as only USB driver writes into FIFO
     #endif
   #endif // CFG_TUD_AUDIO_FUNC_1_EP_OUT_SW_BUF_SZ > 0
 
   #if CFG_TUD_AUDIO > 1 && CFG_TUD_AUDIO_FUNC_2_EP_OUT_SW_BUF_SZ > 0
-    CFG_TUSB_MEM_SECTION CFG_TUSB_MEM_ALIGN uint8_t audio_ep_out_sw_buf_2[CFG_TUD_AUDIO_FUNC_2_EP_OUT_SW_BUF_SZ];
+    CFG_TUD_MEM_SECTION CFG_TUSB_MEM_ALIGN uint8_t audio_ep_out_sw_buf_2[CFG_TUD_AUDIO_FUNC_2_EP_OUT_SW_BUF_SZ];
     #if CFG_FIFO_MUTEX
     osal_mutex_def_t ep_out_ff_mutex_rd_2; // No need for write mutex as only USB driver writes into FIFO
     #endif
   #endif // CFG_TUD_AUDIO > 1 && CFG_TUD_AUDIO_FUNC_2_EP_OUT_SW_BUF_SZ > 0
 
   #if CFG_TUD_AUDIO > 2 && CFG_TUD_AUDIO_FUNC_3_EP_OUT_SW_BUF_SZ > 0
-    CFG_TUSB_MEM_SECTION CFG_TUSB_MEM_ALIGN uint8_t audio_ep_out_sw_buf_3[CFG_TUD_AUDIO_FUNC_3_EP_OUT_SW_BUF_SZ];
+    CFG_TUD_MEM_SECTION CFG_TUSB_MEM_ALIGN uint8_t audio_ep_out_sw_buf_3[CFG_TUD_AUDIO_FUNC_3_EP_OUT_SW_BUF_SZ];
     #if CFG_FIFO_MUTEX
     osal_mutex_def_t ep_out_ff_mutex_rd_3; // No need for write mutex as only USB driver writes into FIFO
     #endif
@@ -180,27 +180,27 @@
 // - the software encoding is used - in this case the linear buffers serve as a target memory where logical channels are encoded into
 #if CFG_TUD_AUDIO_ENABLE_EP_OUT && (USE_LINEAR_BUFFER || CFG_TUD_AUDIO_ENABLE_DECODING)
   #if CFG_TUD_AUDIO_FUNC_1_EP_OUT_SZ_MAX > 0
-    CFG_TUSB_MEM_SECTION CFG_TUSB_MEM_ALIGN uint8_t lin_buf_out_1[CFG_TUD_AUDIO_FUNC_1_EP_OUT_SZ_MAX];
+    CFG_TUD_MEM_SECTION CFG_TUSB_MEM_ALIGN uint8_t lin_buf_out_1[CFG_TUD_AUDIO_FUNC_1_EP_OUT_SZ_MAX];
   #endif
 
   #if CFG_TUD_AUDIO > 1 && CFG_TUD_AUDIO_FUNC_2_EP_OUT_SZ_MAX > 0
-    CFG_TUSB_MEM_SECTION CFG_TUSB_MEM_ALIGN uint8_t lin_buf_out_2[CFG_TUD_AUDIO_FUNC_2_EP_OUT_SZ_MAX];
+    CFG_TUD_MEM_SECTION CFG_TUSB_MEM_ALIGN uint8_t lin_buf_out_2[CFG_TUD_AUDIO_FUNC_2_EP_OUT_SZ_MAX];
   #endif
 
   #if CFG_TUD_AUDIO > 2 && CFG_TUD_AUDIO_FUNC_3_EP_OUT_SZ_MAX > 0
-    CFG_TUSB_MEM_SECTION CFG_TUSB_MEM_ALIGN uint8_t lin_buf_out_3[CFG_TUD_AUDIO_FUNC_3_EP_OUT_SZ_MAX];
+    CFG_TUD_MEM_SECTION CFG_TUSB_MEM_ALIGN uint8_t lin_buf_out_3[CFG_TUD_AUDIO_FUNC_3_EP_OUT_SZ_MAX];
   #endif
 #endif // CFG_TUD_AUDIO_ENABLE_EP_OUT && (USE_LINEAR_BUFFER || CFG_TUD_AUDIO_ENABLE_DECODING)
 
 // Control buffers
-CFG_TUSB_MEM_SECTION CFG_TUSB_MEM_ALIGN uint8_t ctrl_buf_1[CFG_TUD_AUDIO_FUNC_1_CTRL_BUF_SZ];
+CFG_TUD_MEM_SECTION CFG_TUSB_MEM_ALIGN uint8_t ctrl_buf_1[CFG_TUD_AUDIO_FUNC_1_CTRL_BUF_SZ];
 
 #if CFG_TUD_AUDIO > 1
-CFG_TUSB_MEM_SECTION CFG_TUSB_MEM_ALIGN uint8_t ctrl_buf_2[CFG_TUD_AUDIO_FUNC_2_CTRL_BUF_SZ];
+CFG_TUD_MEM_SECTION CFG_TUSB_MEM_ALIGN uint8_t ctrl_buf_2[CFG_TUD_AUDIO_FUNC_2_CTRL_BUF_SZ];
 #endif
 
 #if CFG_TUD_AUDIO > 2
-CFG_TUSB_MEM_SECTION CFG_TUSB_MEM_ALIGN uint8_t ctrl_buf_3[CFG_TUD_AUDIO_FUNC_3_CTRL_BUF_SZ];
+CFG_TUD_MEM_SECTION CFG_TUSB_MEM_ALIGN uint8_t ctrl_buf_3[CFG_TUD_AUDIO_FUNC_3_CTRL_BUF_SZ];
 #endif
 
 // Active alternate setting of interfaces
@@ -217,7 +217,7 @@ uint8_t alt_setting_3[CFG_TUD_AUDIO_FUNC_3_N_AS_INT];
 // Software encoding/decoding support FIFOs
 #if CFG_TUD_AUDIO_ENABLE_EP_IN && CFG_TUD_AUDIO_ENABLE_ENCODING
   #if CFG_TUD_AUDIO_FUNC_1_TX_SUPP_SW_FIFO_SZ > 0
-    CFG_TUSB_MEM_SECTION CFG_TUSB_MEM_ALIGN uint8_t tx_supp_ff_buf_1[CFG_TUD_AUDIO_FUNC_1_N_TX_SUPP_SW_FIFO][CFG_TUD_AUDIO_FUNC_1_TX_SUPP_SW_FIFO_SZ];
+    CFG_TUD_MEM_SECTION CFG_TUSB_MEM_ALIGN uint8_t tx_supp_ff_buf_1[CFG_TUD_AUDIO_FUNC_1_N_TX_SUPP_SW_FIFO][CFG_TUD_AUDIO_FUNC_1_TX_SUPP_SW_FIFO_SZ];
     tu_fifo_t tx_supp_ff_1[CFG_TUD_AUDIO_FUNC_1_N_TX_SUPP_SW_FIFO];
     #if CFG_FIFO_MUTEX
     osal_mutex_def_t tx_supp_ff_mutex_wr_1[CFG_TUD_AUDIO_FUNC_1_N_TX_SUPP_SW_FIFO]; // No need for read mutex as only USB driver reads from FIFO
@@ -225,7 +225,7 @@ uint8_t alt_setting_3[CFG_TUD_AUDIO_FUNC_3_N_AS_INT];
   #endif
 
   #if CFG_TUD_AUDIO > 1 && CFG_TUD_AUDIO_FUNC_2_TX_SUPP_SW_FIFO_SZ > 0
-    CFG_TUSB_MEM_SECTION CFG_TUSB_MEM_ALIGN uint8_t tx_supp_ff_buf_2[CFG_TUD_AUDIO_FUNC_2_N_TX_SUPP_SW_FIFO][CFG_TUD_AUDIO_FUNC_2_TX_SUPP_SW_FIFO_SZ];
+    CFG_TUD_MEM_SECTION CFG_TUSB_MEM_ALIGN uint8_t tx_supp_ff_buf_2[CFG_TUD_AUDIO_FUNC_2_N_TX_SUPP_SW_FIFO][CFG_TUD_AUDIO_FUNC_2_TX_SUPP_SW_FIFO_SZ];
     tu_fifo_t tx_supp_ff_2[CFG_TUD_AUDIO_FUNC_2_N_TX_SUPP_SW_FIFO];
     #if CFG_FIFO_MUTEX
     osal_mutex_def_t tx_supp_ff_mutex_wr_2[CFG_TUD_AUDIO_FUNC_2_N_TX_SUPP_SW_FIFO]; // No need for read mutex as only USB driver reads from FIFO
@@ -233,7 +233,7 @@ uint8_t alt_setting_3[CFG_TUD_AUDIO_FUNC_3_N_AS_INT];
   #endif
 
   #if CFG_TUD_AUDIO > 2 && CFG_TUD_AUDIO_FUNC_3_TX_SUPP_SW_FIFO_SZ > 0
-    CFG_TUSB_MEM_SECTION CFG_TUSB_MEM_ALIGN uint8_t tx_supp_ff_buf_3[CFG_TUD_AUDIO_FUNC_3_N_TX_SUPP_SW_FIFO][CFG_TUD_AUDIO_FUNC_3_TX_SUPP_SW_FIFO_SZ];
+    CFG_TUD_MEM_SECTION CFG_TUSB_MEM_ALIGN uint8_t tx_supp_ff_buf_3[CFG_TUD_AUDIO_FUNC_3_N_TX_SUPP_SW_FIFO][CFG_TUD_AUDIO_FUNC_3_TX_SUPP_SW_FIFO_SZ];
     tu_fifo_t tx_supp_ff_3[CFG_TUD_AUDIO_FUNC_3_N_TX_SUPP_SW_FIFO];
     #if CFG_FIFO_MUTEX
     osal_mutex_def_t tx_supp_ff_mutex_wr_3[CFG_TUD_AUDIO_FUNC_3_N_TX_SUPP_SW_FIFO]; // No need for read mutex as only USB driver reads from FIFO
@@ -243,7 +243,7 @@ uint8_t alt_setting_3[CFG_TUD_AUDIO_FUNC_3_N_AS_INT];
 
 #if CFG_TUD_AUDIO_ENABLE_EP_OUT && CFG_TUD_AUDIO_ENABLE_DECODING
   #if CFG_TUD_AUDIO_FUNC_1_RX_SUPP_SW_FIFO_SZ > 0
-    CFG_TUSB_MEM_SECTION CFG_TUSB_MEM_ALIGN uint8_t rx_supp_ff_buf_1[CFG_TUD_AUDIO_FUNC_1_N_RX_SUPP_SW_FIFO][CFG_TUD_AUDIO_FUNC_1_RX_SUPP_SW_FIFO_SZ];
+    CFG_TUD_MEM_SECTION CFG_TUSB_MEM_ALIGN uint8_t rx_supp_ff_buf_1[CFG_TUD_AUDIO_FUNC_1_N_RX_SUPP_SW_FIFO][CFG_TUD_AUDIO_FUNC_1_RX_SUPP_SW_FIFO_SZ];
     tu_fifo_t rx_supp_ff_1[CFG_TUD_AUDIO_FUNC_1_N_RX_SUPP_SW_FIFO];
     #if CFG_FIFO_MUTEX
     osal_mutex_def_t rx_supp_ff_mutex_rd_1[CFG_TUD_AUDIO_FUNC_1_N_RX_SUPP_SW_FIFO]; // No need for write mutex as only USB driver writes into FIFO
@@ -251,7 +251,7 @@ uint8_t alt_setting_3[CFG_TUD_AUDIO_FUNC_3_N_AS_INT];
   #endif
 
   #if CFG_TUD_AUDIO > 1 && CFG_TUD_AUDIO_FUNC_2_RX_SUPP_SW_FIFO_SZ > 0
-    CFG_TUSB_MEM_SECTION CFG_TUSB_MEM_ALIGN uint8_t rx_supp_ff_buf_2[CFG_TUD_AUDIO_FUNC_2_N_RX_SUPP_SW_FIFO][CFG_TUD_AUDIO_FUNC_2_RX_SUPP_SW_FIFO_SZ];
+    CFG_TUD_MEM_SECTION CFG_TUSB_MEM_ALIGN uint8_t rx_supp_ff_buf_2[CFG_TUD_AUDIO_FUNC_2_N_RX_SUPP_SW_FIFO][CFG_TUD_AUDIO_FUNC_2_RX_SUPP_SW_FIFO_SZ];
     tu_fifo_t rx_supp_ff_2[CFG_TUD_AUDIO_FUNC_2_N_RX_SUPP_SW_FIFO];
     #if CFG_FIFO_MUTEX
     osal_mutex_def_t rx_supp_ff_mutex_rd_2[CFG_TUD_AUDIO_FUNC_2_N_RX_SUPP_SW_FIFO]; // No need for write mutex as only USB driver writes into FIFO
@@ -259,7 +259,7 @@ uint8_t alt_setting_3[CFG_TUD_AUDIO_FUNC_3_N_AS_INT];
   #endif
 
   #if CFG_TUD_AUDIO > 2 && CFG_TUD_AUDIO_FUNC_3_RX_SUPP_SW_FIFO_SZ > 0
-    CFG_TUSB_MEM_SECTION CFG_TUSB_MEM_ALIGN uint8_t rx_supp_ff_buf_3[CFG_TUD_AUDIO_FUNC_3_N_RX_SUPP_SW_FIFO][CFG_TUD_AUDIO_FUNC_3_RX_SUPP_SW_FIFO_SZ];
+    CFG_TUD_MEM_SECTION CFG_TUSB_MEM_ALIGN uint8_t rx_supp_ff_buf_3[CFG_TUD_AUDIO_FUNC_3_N_RX_SUPP_SW_FIFO][CFG_TUD_AUDIO_FUNC_3_RX_SUPP_SW_FIFO_SZ];
     tu_fifo_t rx_supp_ff_3[CFG_TUD_AUDIO_FUNC_3_N_RX_SUPP_SW_FIFO];
     #if CFG_FIFO_MUTEX
     osal_mutex_def_t rx_supp_ff_mutex_rd_3[CFG_TUD_AUDIO_FUNC_3_N_RX_SUPP_SW_FIFO]; // No need for write mutex as only USB driver writes into FIFO
@@ -416,7 +416,7 @@ typedef struct
 //--------------------------------------------------------------------+
 // INTERNAL OBJECT & FUNCTION DECLARATION
 //--------------------------------------------------------------------+
-CFG_TUSB_MEM_SECTION audiod_function_t _audiod_fct[CFG_TUD_AUDIO];
+CFG_TUD_MEM_SECTION audiod_function_t _audiod_fct[CFG_TUD_AUDIO];
 
 #if CFG_TUD_AUDIO_ENABLE_EP_OUT
 static bool audiod_rx_done_cb(uint8_t rhport, audiod_function_t* audio, uint16_t n_bytes_received);
diff --git a/src/class/bth/bth_device.c b/src/class/bth/bth_device.c
index f96bb3552..c5c74d26a 100755
--- a/src/class/bth/bth_device.c
+++ b/src/class/bth/bth_device.c
@@ -55,7 +55,7 @@ typedef struct
 //--------------------------------------------------------------------+
 // INTERNAL OBJECT & FUNCTION DECLARATION
 //--------------------------------------------------------------------+
-CFG_TUSB_MEM_SECTION btd_interface_t _btd_itf;
+CFG_TUD_MEM_SECTION btd_interface_t _btd_itf;
 
 static bool bt_tx_data(uint8_t ep, void *data, uint16_t len)
 {
diff --git a/src/class/cdc/cdc_device.c b/src/class/cdc/cdc_device.c
index 5adce521d..f658df4d0 100644
--- a/src/class/cdc/cdc_device.c
+++ b/src/class/cdc/cdc_device.c
@@ -76,7 +76,7 @@ typedef struct
 //--------------------------------------------------------------------+
 // INTERNAL OBJECT & FUNCTION DECLARATION
 //--------------------------------------------------------------------+
-CFG_TUSB_MEM_SECTION tu_static cdcd_interface_t _cdcd_itf[CFG_TUD_CDC];
+CFG_TUD_MEM_SECTION tu_static cdcd_interface_t _cdcd_itf[CFG_TUD_CDC];
 
 static bool _prep_out_transaction (cdcd_interface_t* p_cdc)
 {
diff --git a/src/class/dfu/dfu_device.c b/src/class/dfu/dfu_device.c
index 464c4bd6b..bffdb8376 100644
--- a/src/class/dfu/dfu_device.c
+++ b/src/class/dfu/dfu_device.c
@@ -56,7 +56,7 @@ typedef struct
 } dfu_state_ctx_t;
 
 // Only a single dfu state is allowed
-CFG_TUSB_MEM_SECTION tu_static dfu_state_ctx_t _dfu_ctx;
+CFG_TUD_MEM_SECTION tu_static dfu_state_ctx_t _dfu_ctx;
 
 static void reset_state(void)
 {
diff --git a/src/class/hid/hid_device.c b/src/class/hid/hid_device.c
index 9240fe2ca..5637ea6b4 100644
--- a/src/class/hid/hid_device.c
+++ b/src/class/hid/hid_device.c
@@ -58,7 +58,7 @@ typedef struct
   tusb_hid_descriptor_hid_t const * hid_descriptor;
 } hidd_interface_t;
 
-CFG_TUSB_MEM_SECTION tu_static hidd_interface_t _hidd_itf[CFG_TUD_HID];
+CFG_TUD_MEM_SECTION tu_static hidd_interface_t _hidd_itf[CFG_TUD_HID];
 
 /*------------- Helpers -------------*/
 static inline uint8_t get_index_by_itfnum(uint8_t itf_num)
diff --git a/src/class/midi/midi_device.c b/src/class/midi/midi_device.c
index e3e7826da..052372a93 100644
--- a/src/class/midi/midi_device.c
+++ b/src/class/midi/midi_device.c
@@ -82,7 +82,7 @@ typedef struct
 //--------------------------------------------------------------------+
 // INTERNAL OBJECT & FUNCTION DECLARATION
 //--------------------------------------------------------------------+
-CFG_TUSB_MEM_SECTION midid_interface_t _midid_itf[CFG_TUD_MIDI];
+CFG_TUD_MEM_SECTION midid_interface_t _midid_itf[CFG_TUD_MIDI];
 
 bool tud_midi_n_mounted (uint8_t itf)
 {
diff --git a/src/class/msc/msc_device.c b/src/class/msc/msc_device.c
index 159a11259..a602e9672 100644
--- a/src/class/msc/msc_device.c
+++ b/src/class/msc/msc_device.c
@@ -34,13 +34,16 @@
 
 #include "msc_device.h"
 
+// Level where CFG_TUSB_DEBUG must be at least for this driver is logged
+#ifndef CFG_TUD_MSC_LOG_LEVEL
+  #define CFG_TUD_MSC_LOG_LEVEL   2
+#endif
+
+#define TU_LOG_DRV(...)   TU_LOG(CFG_TUD_MSC_LOG_LEVEL, __VA_ARGS__)
+
 //--------------------------------------------------------------------+
 // MACRO CONSTANT TYPEDEF
 //--------------------------------------------------------------------+
-
-// Can be selectively disabled to reduce logging when troubleshooting other driver
-#define MSC_DEBUG   2
-
 enum
 {
   MSC_STAGE_CMD  = 0,
@@ -71,8 +74,8 @@ typedef struct
   uint8_t add_sense_qualifier;
 }mscd_interface_t;
 
-CFG_TUSB_MEM_SECTION CFG_TUSB_MEM_ALIGN tu_static mscd_interface_t _mscd_itf;
-CFG_TUSB_MEM_SECTION CFG_TUSB_MEM_ALIGN tu_static uint8_t _mscd_buf[CFG_TUD_MSC_EP_BUFSIZE];
+CFG_TUD_MEM_SECTION CFG_TUSB_MEM_ALIGN tu_static mscd_interface_t _mscd_itf;
+CFG_TUD_MEM_SECTION CFG_TUSB_MEM_ALIGN tu_static uint8_t _mscd_buf[CFG_TUD_MSC_EP_BUFSIZE];
 
 //--------------------------------------------------------------------+
 // INTERNAL OBJECT & FUNCTION DECLARATION
@@ -164,7 +167,7 @@ uint8_t rdwr10_validate_cmd(msc_cbw_t const* cbw)
   {
     if ( block_count )
     {
-      TU_LOG(MSC_DEBUG, "  SCSI case 2 (Hn < Di) or case 3 (Hn < Do) \r\n");
+      TU_LOG_DRV("  SCSI case 2 (Hn < Di) or case 3 (Hn < Do) \r\n");
       status = MSC_CSW_STATUS_PHASE_ERROR;
     }else
     {
@@ -174,22 +177,22 @@ uint8_t rdwr10_validate_cmd(msc_cbw_t const* cbw)
   {
     if ( SCSI_CMD_READ_10 == cbw->command[0] && !is_data_in(cbw->dir) )
     {
-      TU_LOG(MSC_DEBUG, "  SCSI case 10 (Ho <> Di)\r\n");
+      TU_LOG_DRV("  SCSI case 10 (Ho <> Di)\r\n");
       status = MSC_CSW_STATUS_PHASE_ERROR;
     }
     else if ( SCSI_CMD_WRITE_10 == cbw->command[0] && is_data_in(cbw->dir) )
     {
-      TU_LOG(MSC_DEBUG, "  SCSI case 8 (Hi <> Do)\r\n");
+      TU_LOG_DRV("  SCSI case 8 (Hi <> Do)\r\n");
       status = MSC_CSW_STATUS_PHASE_ERROR;
     }
     else if ( 0 == block_count )
     {
-      TU_LOG(MSC_DEBUG, "  SCSI case 4 Hi > Dn (READ10) or case 9 Ho > Dn (WRITE10) \r\n");
+      TU_LOG_DRV("  SCSI case 4 Hi > Dn (READ10) or case 9 Ho > Dn (WRITE10) \r\n");
       status =  MSC_CSW_STATUS_FAILED;
     }
     else if ( cbw->total_bytes / block_count == 0 )
     {
-      TU_LOG(MSC_DEBUG, " Computed block size = 0. SCSI case 7 Hi < Di (READ10) or case 13 Ho < Do (WRIT10)\r\n");
+      TU_LOG_DRV(" Computed block size = 0. SCSI case 7 Hi < Di (READ10) or case 13 Ho < Do (WRIT10)\r\n");
       status = MSC_CSW_STATUS_PHASE_ERROR;
     }
   }
@@ -352,7 +355,7 @@ bool mscd_control_xfer_cb(uint8_t rhport, uint8_t stage, tusb_control_request_t
   switch ( request->bRequest )
   {
     case MSC_REQ_RESET:
-      TU_LOG(MSC_DEBUG, "  MSC BOT Reset\r\n");
+      TU_LOG_DRV("  MSC BOT Reset\r\n");
       TU_VERIFY(request->wValue == 0 && request->wLength == 0);
 
       // driver state reset
@@ -363,7 +366,7 @@ bool mscd_control_xfer_cb(uint8_t rhport, uint8_t stage, tusb_control_request_t
 
     case MSC_REQ_GET_MAX_LUN:
     {
-      TU_LOG(MSC_DEBUG, "  MSC Get Max Lun\r\n");
+      TU_LOG_DRV("  MSC Get Max Lun\r\n");
       TU_VERIFY(request->wValue == 0 && request->wLength == 1);
 
       uint8_t maxlun = 1;
@@ -400,7 +403,7 @@ bool mscd_xfer_cb(uint8_t rhport, uint8_t ep_addr, xfer_result_t event, uint32_t
 
       if ( !(xferred_bytes == sizeof(msc_cbw_t) && p_cbw->signature == MSC_CBW_SIGNATURE) )
       {
-        TU_LOG(MSC_DEBUG, "  SCSI CBW is not valid\r\n");
+        TU_LOG_DRV("  SCSI CBW is not valid\r\n");
 
         // BOT 6.6.1 If CBW is not valid stall both endpoints until reset recovery
         p_msc->stage = MSC_STAGE_NEED_RESET;
@@ -412,7 +415,7 @@ bool mscd_xfer_cb(uint8_t rhport, uint8_t ep_addr, xfer_result_t event, uint32_t
         return false;
       }
 
-      TU_LOG(MSC_DEBUG, "  SCSI Command [Lun%u]: %s\r\n", p_cbw->lun, tu_lookup_find(&_msc_scsi_cmd_table, p_cbw->command[0]));
+      TU_LOG_DRV("  SCSI Command [Lun%u]: %s\r\n", p_cbw->lun, tu_lookup_find(&_msc_scsi_cmd_table, p_cbw->command[0]));
       //TU_LOG_MEM(MSC_DEBUG, p_cbw, xferred_bytes, 2);
 
       p_csw->signature    = MSC_CSW_SIGNATURE;
@@ -457,7 +460,7 @@ bool mscd_xfer_cb(uint8_t rhport, uint8_t ep_addr, xfer_result_t event, uint32_t
         {
           if (p_cbw->total_bytes > sizeof(_mscd_buf))
           {
-            TU_LOG(MSC_DEBUG, "  SCSI reject non READ10/WRITE10 with large data\r\n");
+            TU_LOG_DRV("  SCSI reject non READ10/WRITE10 with large data\r\n");
             fail_scsi_op(rhport, p_msc, MSC_CSW_STATUS_FAILED);
           }else
           {
@@ -479,7 +482,7 @@ bool mscd_xfer_cb(uint8_t rhport, uint8_t ep_addr, xfer_result_t event, uint32_t
           if ( resplen < 0 )
           {
             // unsupported command
-            TU_LOG(MSC_DEBUG, "  SCSI unsupported or failed command\r\n");
+            TU_LOG_DRV("  SCSI unsupported or failed command\r\n");
             fail_scsi_op(rhport, p_msc, MSC_CSW_STATUS_FAILED);
           }
           else if (resplen == 0)
@@ -514,7 +517,7 @@ bool mscd_xfer_cb(uint8_t rhport, uint8_t ep_addr, xfer_result_t event, uint32_t
     break;
 
     case MSC_STAGE_DATA:
-      TU_LOG(MSC_DEBUG, "  SCSI Data [Lun%u]\r\n", p_cbw->lun);
+      TU_LOG_DRV("  SCSI Data [Lun%u]\r\n", p_cbw->lun);
       //TU_LOG_MEM(MSC_DEBUG, _mscd_buf, xferred_bytes, 2);
 
       if (SCSI_CMD_READ_10 == p_cbw->command[0])
@@ -546,7 +549,7 @@ bool mscd_xfer_cb(uint8_t rhport, uint8_t ep_addr, xfer_result_t event, uint32_t
           if ( cb_result < 0 )
           {
             // unsupported command
-            TU_LOG(MSC_DEBUG, "  SCSI unsupported command\r\n");
+            TU_LOG_DRV("  SCSI unsupported command\r\n");
             fail_scsi_op(rhport, p_msc, MSC_CSW_STATUS_FAILED);
           }else
           {
@@ -575,7 +578,7 @@ bool mscd_xfer_cb(uint8_t rhport, uint8_t ep_addr, xfer_result_t event, uint32_t
       // Wait for the Status phase to complete
       if( (ep_addr == p_msc->ep_in) && (xferred_bytes == sizeof(msc_csw_t)) )
       {
-        TU_LOG(MSC_DEBUG, "  SCSI Status [Lun%u] = %u\r\n", p_cbw->lun, p_csw->status);
+        TU_LOG_DRV("  SCSI Status [Lun%u] = %u\r\n", p_cbw->lun, p_csw->status);
         // TU_LOG_MEM(MSC_DEBUG, p_csw, xferred_bytes, 2);
 
         // Invoke complete callback if defined
@@ -845,7 +848,7 @@ static void proc_read10_cmd(uint8_t rhport, mscd_interface_t* p_msc)
   if ( nbytes < 0 )
   {
     // negative means error -> endpoint is stalled & status in CSW set to failed
-    TU_LOG(MSC_DEBUG, "  tud_msc_read10_cb() return -1\r\n");
+    TU_LOG_DRV("  tud_msc_read10_cb() return -1\r\n");
 
     // set sense
     set_sense_medium_not_present(p_cbw->lun);
@@ -907,7 +910,7 @@ static void proc_write10_new_data(uint8_t rhport, mscd_interface_t* p_msc, uint3
   if ( nbytes < 0 )
   {
     // negative means error -> failed this scsi op
-    TU_LOG(MSC_DEBUG, "  tud_msc_write10_cb() return -1\r\n");
+    TU_LOG_DRV("  tud_msc_write10_cb() return -1\r\n");
 
     // update actual byte before failed
     p_msc->xferred_len += xferred_bytes;
diff --git a/src/class/net/ecm_rndis_device.c b/src/class/net/ecm_rndis_device.c
index 8ac7cbd01..762425732 100644
--- a/src/class/net/ecm_rndis_device.c
+++ b/src/class/net/ecm_rndis_device.c
@@ -61,10 +61,10 @@ typedef struct
 #define CFG_TUD_NET_PACKET_PREFIX_LEN sizeof(rndis_data_packet_t)
 #define CFG_TUD_NET_PACKET_SUFFIX_LEN 0
 
-CFG_TUSB_MEM_SECTION CFG_TUSB_MEM_ALIGN tu_static
+CFG_TUD_MEM_SECTION CFG_TUSB_MEM_ALIGN tu_static
 uint8_t received[CFG_TUD_NET_PACKET_PREFIX_LEN + CFG_TUD_NET_MTU + CFG_TUD_NET_PACKET_PREFIX_LEN];
 
-CFG_TUSB_MEM_SECTION CFG_TUSB_MEM_ALIGN tu_static
+CFG_TUD_MEM_SECTION CFG_TUSB_MEM_ALIGN tu_static
 uint8_t transmitted[CFG_TUD_NET_PACKET_PREFIX_LEN + CFG_TUD_NET_MTU + CFG_TUD_NET_PACKET_PREFIX_LEN];
 
 struct ecm_notify_struct
@@ -94,8 +94,8 @@ tu_static const struct ecm_notify_struct ecm_notify_csc =
   .uplink = 9728000,
 };
 
-// TODO remove CFG_TUSB_MEM_SECTION, control internal buffer is already in this special section
-CFG_TUSB_MEM_SECTION CFG_TUSB_MEM_ALIGN tu_static union
+// TODO remove CFG_TUD_MEM_SECTION, control internal buffer is already in this special section
+CFG_TUD_MEM_SECTION CFG_TUSB_MEM_ALIGN tu_static union
 {
   uint8_t rndis_buf[120];
   struct ecm_notify_struct ecm_buf;
@@ -104,8 +104,8 @@ CFG_TUSB_MEM_SECTION CFG_TUSB_MEM_ALIGN tu_static union
 //--------------------------------------------------------------------+
 // INTERNAL OBJECT & FUNCTION DECLARATION
 //--------------------------------------------------------------------+
-// TODO remove CFG_TUSB_MEM_SECTION
-CFG_TUSB_MEM_SECTION tu_static netd_interface_t _netd_itf;
+// TODO remove CFG_TUD_MEM_SECTION
+CFG_TUD_MEM_SECTION tu_static netd_interface_t _netd_itf;
 
 tu_static bool can_xmit;
 
diff --git a/src/class/net/ncm_device.c b/src/class/net/ncm_device.c
index 9e9580249..611dc3348 100644
--- a/src/class/net/ncm_device.c
+++ b/src/class/net/ncm_device.c
@@ -130,7 +130,7 @@ typedef struct
 // INTERNAL OBJECT & FUNCTION DECLARATION
 //--------------------------------------------------------------------+
 
-CFG_TUSB_MEM_SECTION CFG_TUSB_MEM_ALIGN tu_static const ntb_parameters_t ntb_parameters = {
+CFG_TUD_MEM_SECTION CFG_TUSB_MEM_ALIGN tu_static const ntb_parameters_t ntb_parameters = {
     .wLength                 = sizeof(ntb_parameters_t),
     .bmNtbFormatsSupported   = 0x01,
     .dwNtbInMaxSize          = CFG_TUD_NCM_IN_NTB_MAX_SIZE,
@@ -145,9 +145,9 @@ CFG_TUSB_MEM_SECTION CFG_TUSB_MEM_ALIGN tu_static const ntb_parameters_t ntb_par
     .wNtbOutMaxDatagrams     = 0
 };
 
-CFG_TUSB_MEM_SECTION CFG_TUSB_MEM_ALIGN tu_static transmit_ntb_t transmit_ntb[2];
+CFG_TUD_MEM_SECTION CFG_TUSB_MEM_ALIGN tu_static transmit_ntb_t transmit_ntb[2];
 
-CFG_TUSB_MEM_SECTION CFG_TUSB_MEM_ALIGN tu_static uint8_t receive_ntb[CFG_TUD_NCM_OUT_NTB_MAX_SIZE];
+CFG_TUD_MEM_SECTION CFG_TUSB_MEM_ALIGN tu_static uint8_t receive_ntb[CFG_TUD_NCM_OUT_NTB_MAX_SIZE];
 
 tu_static ncm_interface_t ncm_interface;
 
diff --git a/src/class/usbtmc/usbtmc_device.c b/src/class/usbtmc/usbtmc_device.c
index 4e320a778..573654d58 100644
--- a/src/class/usbtmc/usbtmc_device.c
+++ b/src/class/usbtmc/usbtmc_device.c
@@ -143,7 +143,7 @@ typedef struct
   usbtmc_capabilities_specific_t const * capabilities;
 } usbtmc_interface_state_t;
 
-CFG_TUSB_MEM_SECTION tu_static usbtmc_interface_state_t usbtmc_state =
+CFG_TUD_MEM_SECTION tu_static usbtmc_interface_state_t usbtmc_state =
 {
     .itf_id = 0xFF,
 };
diff --git a/src/class/vendor/vendor_device.c b/src/class/vendor/vendor_device.c
index 93596ee33..389a29696 100644
--- a/src/class/vendor/vendor_device.c
+++ b/src/class/vendor/vendor_device.c
@@ -59,7 +59,7 @@ typedef struct
   CFG_TUSB_MEM_ALIGN uint8_t epin_buf[CFG_TUD_VENDOR_EPSIZE];
 } vendord_interface_t;
 
-CFG_TUSB_MEM_SECTION tu_static vendord_interface_t _vendord_itf[CFG_TUD_VENDOR];
+CFG_TUD_MEM_SECTION tu_static vendord_interface_t _vendord_itf[CFG_TUD_VENDOR];
 
 #define ITF_MEM_RESET_SIZE   offsetof(vendord_interface_t, rx_ff)
 
diff --git a/src/class/video/video_device.c b/src/class/video/video_device.c
index 91f452afc..e6fadf41b 100644
--- a/src/class/video/video_device.c
+++ b/src/class/video/video_device.c
@@ -130,8 +130,8 @@ typedef struct TU_ATTR_PACKED {
 //--------------------------------------------------------------------+
 // INTERNAL OBJECT & FUNCTION DECLARATION
 //--------------------------------------------------------------------+
-CFG_TUSB_MEM_SECTION tu_static videod_interface_t _videod_itf[CFG_TUD_VIDEO];
-CFG_TUSB_MEM_SECTION tu_static videod_streaming_interface_t _videod_streaming_itf[CFG_TUD_VIDEO_STREAMING];
+CFG_TUD_MEM_SECTION tu_static videod_interface_t _videod_itf[CFG_TUD_VIDEO];
+CFG_TUD_MEM_SECTION tu_static videod_streaming_interface_t _videod_streaming_itf[CFG_TUD_VIDEO_STREAMING];
 
 tu_static uint8_t const _cap_get     = 0x1u; /* support for GET */
 tu_static uint8_t const _cap_get_set = 0x3u; /* support for GET and SET */
diff --git a/src/common/tusb_common.h b/src/common/tusb_common.h
index 576cb123f..25f85e501 100644
--- a/src/common/tusb_common.h
+++ b/src/common/tusb_common.h
@@ -53,6 +53,8 @@
 #define U32_TO_U8S_LE(_u32)   TU_U32_BYTE0(_u32), TU_U32_BYTE1(_u32), TU_U32_BYTE2(_u32), TU_U32_BYTE3(_u32)
 
 #define TU_BIT(n)             (1UL << (n))
+
+// Generate a mask with bit from high (31) to low (0) set, e.g TU_GENMASK(3, 0) = 0b1111
 #define TU_GENMASK(h, l)      ( (UINT32_MAX << (l)) & (UINT32_MAX >> (31 - (h))) )
 
 //--------------------------------------------------------------------+
@@ -99,10 +101,9 @@ TU_ATTR_WEAK extern void* tusb_app_phys_to_virt(void *phys_addr);
 #define tu_varclr(_var)          tu_memclr(_var, sizeof(*(_var)))
 
 // This is a backport of memset_s from c11
-TU_ATTR_ALWAYS_INLINE static inline int tu_memset_s(void *dest, size_t destsz, int ch, size_t count)
-{
+TU_ATTR_ALWAYS_INLINE static inline int tu_memset_s(void *dest, size_t destsz, int ch, size_t count) {
   // TODO may check if desst and src is not NULL
-  if (count > destsz) {
+  if ( count > destsz ) {
     return -1;
   }
   memset(dest, ch, count);
@@ -110,10 +111,9 @@ TU_ATTR_ALWAYS_INLINE static inline int tu_memset_s(void *dest, size_t destsz, i
 }
 
 // This is a backport of memcpy_s from c11
-TU_ATTR_ALWAYS_INLINE static inline int tu_memcpy_s(void *dest, size_t destsz, const void * src, size_t count )
-{
+TU_ATTR_ALWAYS_INLINE static inline int tu_memcpy_s(void *dest, size_t destsz, const void *src, size_t count) {
   // TODO may check if desst and src is not NULL
-  if (count > destsz) {
+  if ( count > destsz ) {
     return -1;
   }
   memcpy(dest, src, count);
@@ -169,6 +169,9 @@ TU_ATTR_ALWAYS_INLINE static inline uint32_t tu_align32 (uint32_t value) { retur
 TU_ATTR_ALWAYS_INLINE static inline uint32_t tu_align4k (uint32_t value) { return (value & 0xFFFFF000UL); }
 TU_ATTR_ALWAYS_INLINE static inline uint32_t tu_offset4k(uint32_t value) { return (value & 0xFFFUL); }
 
+TU_ATTR_ALWAYS_INLINE static inline bool tu_is_aligned32(uint32_t value) { return (value & 0x1FUL) == 0; }
+TU_ATTR_ALWAYS_INLINE static inline bool tu_is_aligned64(uint64_t value) { return (value & 0x3FUL) == 0; }
+
 //------------- Mathematics -------------//
 TU_ATTR_ALWAYS_INLINE static inline uint32_t tu_div_ceil(uint32_t v, uint32_t d) { return (v + d -1)/d; }
 
diff --git a/src/device/usbd.c b/src/device/usbd.c
index 44c2530ce..9429cf664 100644
--- a/src/device/usbd.c
+++ b/src/device/usbd.c
@@ -43,9 +43,6 @@
   #define CFG_TUD_TASK_QUEUE_SZ   16
 #endif
 
-// Debug level of USBD
-#define USBD_DBG   2
-
 //--------------------------------------------------------------------+
 // Device Data
 //--------------------------------------------------------------------+
@@ -81,7 +78,7 @@ tu_static usbd_device_t _usbd_dev;
 //--------------------------------------------------------------------+
 // Class Driver
 //--------------------------------------------------------------------+
-#if CFG_TUSB_DEBUG >= 2
+#if CFG_TUSB_DEBUG >= CFG_TUD_LOG_LEVEL
   #define DRIVER_NAME(_name)    .name = _name,
 #else
   #define DRIVER_NAME(_name)
@@ -308,7 +305,7 @@ bool usbd_control_xfer_cb (uint8_t rhport, uint8_t ep_addr, xfer_result_t event,
 //--------------------------------------------------------------------+
 // Debug
 //--------------------------------------------------------------------+
-#if CFG_TUSB_DEBUG >= 2
+#if CFG_TUSB_DEBUG >= CFG_TUD_LOG_LEVEL
 tu_static char const* const _usbd_event_str[DCD_EVENT_COUNT] =
 {
   "Invalid"        ,
@@ -330,7 +327,7 @@ void usbd_driver_print_control_complete_name(usbd_control_xfer_cb_t callback)
     usbd_class_driver_t const * driver = get_driver(i);
     if ( driver && driver->control_xfer_cb == callback )
     {
-      TU_LOG(USBD_DBG, "  %s control complete\r\n", driver->name);
+      TU_LOG_USBD("  %s control complete\r\n", driver->name);
       return;
     }
   }
@@ -396,10 +393,10 @@ bool tud_init (uint8_t rhport)
   // skip if already initialized
   if ( tud_inited() ) return true;
 
-  TU_LOG(USBD_DBG, "USBD init on controller %u\r\n", rhport);
-  TU_LOG_INT(USBD_DBG, sizeof(usbd_device_t));
-  TU_LOG_INT(USBD_DBG, sizeof(tu_fifo_t));
-  TU_LOG_INT(USBD_DBG, sizeof(tu_edpt_stream_t));
+  TU_LOG_USBD("USBD init on controller %u\r\n", rhport);
+  TU_LOG_INT(CFG_TUD_LOG_LEVEL, sizeof(usbd_device_t));
+  TU_LOG_INT(CFG_TUD_LOG_LEVEL, sizeof(tu_fifo_t));
+  TU_LOG_INT(CFG_TUD_LOG_LEVEL, sizeof(tu_edpt_stream_t));
 
   tu_varclr(&_usbd_dev);
 
@@ -424,7 +421,7 @@ bool tud_init (uint8_t rhport)
   {
     usbd_class_driver_t const * driver = get_driver(i);
     TU_ASSERT(driver);
-    TU_LOG(USBD_DBG, "%s init\r\n", driver->name);
+    TU_LOG_USBD("%s init\r\n", driver->name);
     driver->init();
   }
 
@@ -496,21 +493,21 @@ void tud_task_ext(uint32_t timeout_ms, bool in_isr)
     dcd_event_t event;
     if ( !osal_queue_receive(_usbd_q, &event, timeout_ms) ) return;
 
-#if CFG_TUSB_DEBUG >= 2
-    if (event.event_id == DCD_EVENT_SETUP_RECEIVED) TU_LOG(USBD_DBG, "\r\n"); // extra line for setup
-    TU_LOG(USBD_DBG, "USBD %s ", event.event_id < DCD_EVENT_COUNT ? _usbd_event_str[event.event_id] : "CORRUPTED");
+#if CFG_TUSB_DEBUG >= CFG_TUD_LOG_LEVEL
+    if (event.event_id == DCD_EVENT_SETUP_RECEIVED) TU_LOG_USBD("\r\n"); // extra line for setup
+    TU_LOG_USBD("USBD %s ", event.event_id < DCD_EVENT_COUNT ? _usbd_event_str[event.event_id] : "CORRUPTED");
 #endif
 
     switch ( event.event_id )
     {
       case DCD_EVENT_BUS_RESET:
-        TU_LOG(USBD_DBG, ": %s Speed\r\n", tu_str_speed[event.bus_reset.speed]);
+        TU_LOG_USBD(": %s Speed\r\n", tu_str_speed[event.bus_reset.speed]);
         usbd_reset(event.rhport);
         _usbd_dev.speed = event.bus_reset.speed;
       break;
 
       case DCD_EVENT_UNPLUGGED:
-        TU_LOG(USBD_DBG, "\r\n");
+        TU_LOG_USBD("\r\n");
         usbd_reset(event.rhport);
 
         // invoke callback
@@ -518,8 +515,8 @@ void tud_task_ext(uint32_t timeout_ms, bool in_isr)
       break;
 
       case DCD_EVENT_SETUP_RECEIVED:
-        TU_LOG_PTR(USBD_DBG, &event.setup_received);
-        TU_LOG(USBD_DBG, "\r\n");
+        TU_LOG_PTR(CFG_TUD_LOG_LEVEL, &event.setup_received);
+        TU_LOG_USBD("\r\n");
 
         // Mark as connected after receiving 1st setup packet.
         // But it is easier to set it every time instead of wasting time to check then set
@@ -534,7 +531,7 @@ void tud_task_ext(uint32_t timeout_ms, bool in_isr)
         // Process control request
         if ( !process_control_request(event.rhport, &event.setup_received) )
         {
-          TU_LOG(USBD_DBG, "  Stall EP0\r\n");
+          TU_LOG_USBD("  Stall EP0\r\n");
           // Failed -> stall both control endpoint IN and OUT
           dcd_edpt_stall(event.rhport, 0);
           dcd_edpt_stall(event.rhport, 0 | TUSB_DIR_IN_MASK);
@@ -548,7 +545,7 @@ void tud_task_ext(uint32_t timeout_ms, bool in_isr)
         uint8_t const epnum   = tu_edpt_number(ep_addr);
         uint8_t const ep_dir  = tu_edpt_dir(ep_addr);
 
-        TU_LOG(USBD_DBG, "on EP %02X with %u bytes\r\n", ep_addr, (unsigned int) event.xfer_complete.len);
+        TU_LOG_USBD("on EP %02X with %u bytes\r\n", ep_addr, (unsigned int) event.xfer_complete.len);
 
         _usbd_dev.ep_status[epnum][ep_dir].busy = 0;
         _usbd_dev.ep_status[epnum][ep_dir].claimed = 0;
@@ -563,7 +560,7 @@ void tud_task_ext(uint32_t timeout_ms, bool in_isr)
           usbd_class_driver_t const * driver = get_driver( _usbd_dev.ep2drv[epnum][ep_dir] );
           TU_ASSERT(driver, );
 
-          TU_LOG(USBD_DBG, "  %s xfer callback\r\n", driver->name);
+          TU_LOG_USBD("  %s xfer callback\r\n", driver->name);
           driver->xfer_cb(event.rhport, ep_addr, (xfer_result_t) event.xfer_complete.result, event.xfer_complete.len);
         }
       }
@@ -575,27 +572,27 @@ void tud_task_ext(uint32_t timeout_ms, bool in_isr)
         // e.g suspend -> resume -> unplug/plug. Skip suspend/resume if not connected
         if ( _usbd_dev.connected )
         {
-          TU_LOG(USBD_DBG, ": Remote Wakeup = %u\r\n", _usbd_dev.remote_wakeup_en);
+          TU_LOG_USBD(": Remote Wakeup = %u\r\n", _usbd_dev.remote_wakeup_en);
           if (tud_suspend_cb) tud_suspend_cb(_usbd_dev.remote_wakeup_en);
         }else
         {
-          TU_LOG(USBD_DBG, " Skipped\r\n");
+          TU_LOG_USBD(" Skipped\r\n");
         }
       break;
 
       case DCD_EVENT_RESUME:
         if ( _usbd_dev.connected )
         {
-          TU_LOG(USBD_DBG, "\r\n");
+          TU_LOG_USBD("\r\n");
           if (tud_resume_cb) tud_resume_cb();
         }else
         {
-          TU_LOG(USBD_DBG, " Skipped\r\n");
+          TU_LOG_USBD(" Skipped\r\n");
         }
       break;
 
       case USBD_EVENT_FUNC_CALL:
-        TU_LOG(USBD_DBG, "\r\n");
+        TU_LOG_USBD("\r\n");
         if ( event.func_call.func ) event.func_call.func(event.func_call.param);
       break;
 
@@ -620,7 +617,7 @@ void tud_task_ext(uint32_t timeout_ms, bool in_isr)
 static bool invoke_class_control(uint8_t rhport, usbd_class_driver_t const * driver, tusb_control_request_t const * request)
 {
   usbd_control_set_complete_callback(driver->control_xfer_cb);
-  TU_LOG(USBD_DBG, "  %s control request\r\n", driver->name);
+  TU_LOG_USBD("  %s control request\r\n", driver->name);
   return driver->control_xfer_cb(rhport, CONTROL_STAGE_SETUP, request);
 }
 
@@ -641,11 +638,11 @@ static bool process_control_request(uint8_t rhport, tusb_control_request_t const
     return tud_vendor_control_xfer_cb(rhport, CONTROL_STAGE_SETUP, p_request);
   }
 
-#if CFG_TUSB_DEBUG >= 2
+#if CFG_TUSB_DEBUG >= CFG_TUD_LOG_LEVEL
   if (TUSB_REQ_TYPE_STANDARD == p_request->bmRequestType_bit.type && p_request->bRequest <= TUSB_REQ_SYNCH_FRAME)
   {
-    TU_LOG(USBD_DBG, "  %s", tu_str_std_request[p_request->bRequest]);
-    if (TUSB_REQ_GET_DESCRIPTOR != p_request->bRequest) TU_LOG(USBD_DBG, "\r\n");
+    TU_LOG_USBD("  %s", tu_str_std_request[p_request->bRequest]);
+    if (TUSB_REQ_GET_DESCRIPTOR != p_request->bRequest) TU_LOG_USBD("\r\n");
   }
 #endif
 
@@ -701,7 +698,7 @@ static bool process_control_request(uint8_t rhport, tusb_control_request_t const
             if ( _usbd_dev.cfg_num )
             {
               // already configured: need to clear all endpoints and driver first
-              TU_LOG(USBD_DBG, "  Clear current Configuration (%u) before switching\r\n", _usbd_dev.cfg_num);
+              TU_LOG_USBD("  Clear current Configuration (%u) before switching\r\n", _usbd_dev.cfg_num);
 
               // close all non-control endpoints, cancel all pending transfers if any
               dcd_edpt_close_all(rhport);
@@ -730,7 +727,7 @@ static bool process_control_request(uint8_t rhport, tusb_control_request_t const
           // Only support remote wakeup for device feature
           TU_VERIFY(TUSB_REQ_FEATURE_REMOTE_WAKEUP == p_request->wValue);
 
-          TU_LOG(USBD_DBG, "    Enable Remote Wakeup\r\n");
+          TU_LOG_USBD("    Enable Remote Wakeup\r\n");
 
           // Host may enable remote wake up before suspending especially HID device
           _usbd_dev.remote_wakeup_en = true;
@@ -741,7 +738,7 @@ static bool process_control_request(uint8_t rhport, tusb_control_request_t const
           // Only support remote wakeup for device feature
           TU_VERIFY(TUSB_REQ_FEATURE_REMOTE_WAKEUP == p_request->wValue);
 
-          TU_LOG(USBD_DBG, "    Disable Remote Wakeup\r\n");
+          TU_LOG_USBD("    Disable Remote Wakeup\r\n");
 
           // Host may disable remote wake up after resuming
           _usbd_dev.remote_wakeup_en = false;
@@ -924,7 +921,7 @@ static bool process_set_config(uint8_t rhport, uint8_t cfg_num)
       if ( (sizeof(tusb_desc_interface_t) <= drv_len)  && (drv_len <= remaining_len) )
       {
         // Open successfully
-        TU_LOG(USBD_DBG, "  %s opened\r\n", driver->name);
+        TU_LOG_USBD("  %s opened\r\n", driver->name);
 
         // Some drivers use 2 or more interfaces but may not have IAD e.g MIDI (always) or
         // BTH (even CDC) with class in device descriptor (single interface)
@@ -983,7 +980,7 @@ static bool process_get_descriptor(uint8_t rhport, tusb_control_request_t const
   {
     case TUSB_DESC_DEVICE:
     {
-      TU_LOG(USBD_DBG, " Device\r\n");
+      TU_LOG_USBD(" Device\r\n");
 
       void* desc_device = (void*) (uintptr_t) tud_descriptor_device_cb();
 
@@ -1007,7 +1004,7 @@ static bool process_get_descriptor(uint8_t rhport, tusb_control_request_t const
 
     case TUSB_DESC_BOS:
     {
-      TU_LOG(USBD_DBG, " BOS\r\n");
+      TU_LOG_USBD(" BOS\r\n");
 
       // requested by host if USB > 2.0 ( i.e 2.1 or 3.x )
       if (!tud_descriptor_bos_cb) return false;
@@ -1029,12 +1026,12 @@ static bool process_get_descriptor(uint8_t rhport, tusb_control_request_t const
 
       if ( desc_type == TUSB_DESC_CONFIGURATION )
       {
-        TU_LOG(USBD_DBG, " Configuration[%u]\r\n", desc_index);
+        TU_LOG_USBD(" Configuration[%u]\r\n", desc_index);
         desc_config = (uintptr_t) tud_descriptor_configuration_cb(desc_index);
       }else
       {
         // Host only request this after getting Device Qualifier descriptor
-        TU_LOG(USBD_DBG, " Other Speed Configuration\r\n");
+        TU_LOG_USBD(" Other Speed Configuration\r\n");
         TU_VERIFY( tud_descriptor_other_speed_configuration_cb );
         desc_config = (uintptr_t) tud_descriptor_other_speed_configuration_cb(desc_index);
       }
@@ -1050,7 +1047,7 @@ static bool process_get_descriptor(uint8_t rhport, tusb_control_request_t const
 
     case TUSB_DESC_STRING:
     {
-      TU_LOG(USBD_DBG, " String[%u]\r\n", desc_index);
+      TU_LOG_USBD(" String[%u]\r\n", desc_index);
 
       // String Descriptor always uses the desc set from user
       uint8_t const* desc_str = (uint8_t const*) tud_descriptor_string_cb(desc_index, tu_le16toh(p_request->wIndex));
@@ -1063,7 +1060,7 @@ static bool process_get_descriptor(uint8_t rhport, tusb_control_request_t const
 
     case TUSB_DESC_DEVICE_QUALIFIER:
     {
-      TU_LOG(USBD_DBG, " Device Qualifier\r\n");
+      TU_LOG_USBD(" Device Qualifier\r\n");
 
       TU_VERIFY( tud_descriptor_device_qualifier_cb );
 
@@ -1248,7 +1245,7 @@ bool usbd_edpt_xfer(uint8_t rhport, uint8_t ep_addr, uint8_t * buffer, uint16_t
   // TODO skip ready() check for now since enumeration also use this API
   // TU_VERIFY(tud_ready());
 
-  TU_LOG(USBD_DBG, "  Queue EP %02X with %u bytes ...\r\n", ep_addr, total_bytes);
+  TU_LOG_USBD("  Queue EP %02X with %u bytes ...\r\n", ep_addr, total_bytes);
 
   // Attempt to transfer on a busy endpoint, sound like an race condition !
   TU_ASSERT(_usbd_dev.ep_status[epnum][dir].busy == 0);
@@ -1265,7 +1262,7 @@ bool usbd_edpt_xfer(uint8_t rhport, uint8_t ep_addr, uint8_t * buffer, uint16_t
     // DCD error, mark endpoint as ready to allow next transfer
     _usbd_dev.ep_status[epnum][dir].busy = 0;
     _usbd_dev.ep_status[epnum][dir].claimed = 0;
-    TU_LOG(USBD_DBG, "FAILED\r\n");
+    TU_LOG_USBD("FAILED\r\n");
     TU_BREAKPOINT();
     return false;
   }
@@ -1282,7 +1279,7 @@ bool usbd_edpt_xfer_fifo(uint8_t rhport, uint8_t ep_addr, tu_fifo_t * ff, uint16
   uint8_t const epnum = tu_edpt_number(ep_addr);
   uint8_t const dir   = tu_edpt_dir(ep_addr);
 
-  TU_LOG(USBD_DBG, "  Queue ISO EP %02X with %u bytes ... ", ep_addr, total_bytes);
+  TU_LOG_USBD("  Queue ISO EP %02X with %u bytes ... ", ep_addr, total_bytes);
 
   // Attempt to transfer on a busy endpoint, sound like an race condition !
   TU_ASSERT(_usbd_dev.ep_status[epnum][dir].busy == 0);
@@ -1293,14 +1290,14 @@ bool usbd_edpt_xfer_fifo(uint8_t rhport, uint8_t ep_addr, tu_fifo_t * ff, uint16
 
   if (dcd_edpt_xfer_fifo(rhport, ep_addr, ff, total_bytes))
   {
-    TU_LOG(USBD_DBG, "OK\r\n");
+    TU_LOG_USBD("OK\r\n");
     return true;
   }else
   {
     // DCD error, mark endpoint as ready to allow next transfer
     _usbd_dev.ep_status[epnum][dir].busy = 0;
     _usbd_dev.ep_status[epnum][dir].claimed = 0;
-    TU_LOG(USBD_DBG, "failed\r\n");
+    TU_LOG_USBD("failed\r\n");
     TU_BREAKPOINT();
     return false;
   }
@@ -1326,7 +1323,7 @@ void usbd_edpt_stall(uint8_t rhport, uint8_t ep_addr)
   // only stalled if currently cleared
   if ( !_usbd_dev.ep_status[epnum][dir].stalled )
   {
-    TU_LOG(USBD_DBG, "    Stall EP %02X\r\n", ep_addr);
+    TU_LOG_USBD("    Stall EP %02X\r\n", ep_addr);
     dcd_edpt_stall(rhport, ep_addr);
     _usbd_dev.ep_status[epnum][dir].stalled = 1;
     _usbd_dev.ep_status[epnum][dir].busy = 1;
@@ -1343,7 +1340,7 @@ void usbd_edpt_clear_stall(uint8_t rhport, uint8_t ep_addr)
   // only clear if currently stalled
   if ( _usbd_dev.ep_status[epnum][dir].stalled )
   {
-    TU_LOG(USBD_DBG, "    Clear Stall EP %02X\r\n", ep_addr);
+    TU_LOG_USBD("    Clear Stall EP %02X\r\n", ep_addr);
     dcd_edpt_clear_stall(rhport, ep_addr);
     _usbd_dev.ep_status[epnum][dir].stalled = 0;
     _usbd_dev.ep_status[epnum][dir].busy = 0;
@@ -1371,7 +1368,7 @@ void usbd_edpt_close(uint8_t rhport, uint8_t ep_addr)
   rhport = _usbd_rhport;
 
   TU_ASSERT(dcd_edpt_close, /**/);
-  TU_LOG(USBD_DBG, "  CLOSING Endpoint: 0x%02X\r\n", ep_addr);
+  TU_LOG_USBD("  CLOSING Endpoint: 0x%02X\r\n", ep_addr);
 
   uint8_t const epnum = tu_edpt_number(ep_addr);
   uint8_t const dir   = tu_edpt_dir(ep_addr);
diff --git a/src/device/usbd_control.c b/src/device/usbd_control.c
index 2afe967b5..76d062e40 100644
--- a/src/device/usbd_control.c
+++ b/src/device/usbd_control.c
@@ -32,10 +32,7 @@
 #include "tusb.h"
 #include "device/usbd_pvt.h"
 
-// Debug level of USBD Control
-#define USBD_CONTROL_DEBUG   2
-
-#if CFG_TUSB_DEBUG >= USBD_CONTROL_DEBUG
+#if CFG_TUSB_DEBUG >= CFG_TUD_LOG_LEVEL
 extern void usbd_driver_print_control_complete_name(usbd_control_xfer_cb_t callback);
 #endif
 
@@ -58,7 +55,7 @@ typedef struct
 
 tu_static usbd_control_xfer_t _ctrl_xfer;
 
-CFG_TUSB_MEM_SECTION CFG_TUSB_MEM_ALIGN
+CFG_TUD_MEM_SECTION CFG_TUSB_MEM_ALIGN
 tu_static uint8_t _usbd_ctrl_buf[CFG_TUD_ENDPOINT0_SIZE];
 
 //--------------------------------------------------------------------+
@@ -191,7 +188,7 @@ bool usbd_control_xfer_cb (uint8_t rhport, uint8_t ep_addr, xfer_result_t result
   {
     TU_VERIFY(_ctrl_xfer.buffer);
     memcpy(_ctrl_xfer.buffer, _usbd_ctrl_buf, xferred_bytes);
-    TU_LOG_MEM(USBD_CONTROL_DEBUG, _usbd_ctrl_buf, xferred_bytes, 2);
+    TU_LOG_MEM(CFG_TUD_LOG_LEVEL, _usbd_ctrl_buf, xferred_bytes, 2);
   }
 
   _ctrl_xfer.total_xferred += (uint16_t) xferred_bytes;
@@ -208,7 +205,7 @@ bool usbd_control_xfer_cb (uint8_t rhport, uint8_t ep_addr, xfer_result_t result
     // callback can still stall control in status phase e.g out data does not make sense
     if ( _ctrl_xfer.complete_cb )
     {
-      #if CFG_TUSB_DEBUG >= USBD_CONTROL_DEBUG
+      #if CFG_TUSB_DEBUG >= CFG_TUD_LOG_LEVEL
       usbd_driver_print_control_complete_name(_ctrl_xfer.complete_cb);
       #endif
 
diff --git a/src/device/usbd_pvt.h b/src/device/usbd_pvt.h
index 8393d3469..940b2858b 100644
--- a/src/device/usbd_pvt.h
+++ b/src/device/usbd_pvt.h
@@ -33,13 +33,20 @@
  extern "C" {
 #endif
 
+// Level where CFG_TUSB_DEBUG must be at least for USBD is logged
+#ifndef CFG_TUD_LOG_LEVEL
+#define CFG_TUD_LOG_LEVEL   2
+#endif
+
+#define TU_LOG_USBD(...)   TU_LOG(CFG_TUD_LOG_LEVEL, __VA_ARGS__)
+
 //--------------------------------------------------------------------+
 // Class Driver API
 //--------------------------------------------------------------------+
 
 typedef struct
 {
-  #if CFG_TUSB_DEBUG >= 2
+  #if CFG_TUSB_DEBUG >= CFG_TUD_LOG_LEVEL
   char const* name;
   #endif
 
diff --git a/src/host/hcd.h b/src/host/hcd.h
index 3355c18b2..d77dd6e2f 100644
--- a/src/host/hcd.h
+++ b/src/host/hcd.h
@@ -110,15 +110,15 @@ typedef struct
 
 // clean/flush data cache: write cache -> memory.
 // Required before an DMA TX transfer to make sure data is in memory
-void hcd_dcache_clean(void const* addr, uint32_t data_size) TU_ATTR_WEAK;
+bool hcd_dcache_clean(void const* addr, uint32_t data_size) TU_ATTR_WEAK;
 
 // invalidate data cache: mark cache as invalid, next read will read from memory
 // Required BOTH before and after an DMA RX transfer
-void hcd_dcache_invalidate(void const* addr, uint32_t data_size) TU_ATTR_WEAK;
+bool hcd_dcache_invalidate(void const* addr, uint32_t data_size) TU_ATTR_WEAK;
 
 // clean and invalidate data cache
 // Required before an DMA transfer where memory is both read/write by DMA
-void hcd_dcache_clean_invalidate(void const* addr, uint32_t data_size) TU_ATTR_WEAK;
+bool hcd_dcache_clean_invalidate(void const* addr, uint32_t data_size) TU_ATTR_WEAK;
 
 //--------------------------------------------------------------------+
 // Controller API
@@ -171,11 +171,15 @@ bool hcd_edpt_open(uint8_t rhport, uint8_t dev_addr, tusb_desc_endpoint_t const
 // Submit a transfer, when complete hcd_event_xfer_complete() must be invoked
 bool hcd_edpt_xfer(uint8_t rhport, uint8_t dev_addr, uint8_t ep_addr, uint8_t * buffer, uint16_t buflen);
 
+// Abort a queued transfer. Note: it can only abort transfer that has not been started
+// Return true if a queued transfer is aborted, false if there is no transfer to abort
+bool hcd_edpt_abort_xfer(uint8_t rhport, uint8_t dev_addr, uint8_t ep_addr);
+
 // Submit a special transfer to send 8-byte Setup Packet, when complete hcd_event_xfer_complete() must be invoked
 bool hcd_setup_send(uint8_t rhport, uint8_t dev_addr, uint8_t const setup_packet[8]);
 
 // clear stall, data toggle is also reset to DATA0
-bool hcd_edpt_clear_stall(uint8_t daddr, uint8_t ep_addr);
+bool hcd_edpt_clear_stall(uint8_t rhport, uint8_t dev_addr, uint8_t ep_addr);
 
 //--------------------------------------------------------------------+
 // USBH implemented API
diff --git a/src/host/hub.c b/src/host/hub.c
index 182bd6ce8..16dd534d2 100644
--- a/src/host/hub.c
+++ b/src/host/hub.c
@@ -45,8 +45,8 @@ typedef struct
   uint8_t itf_num;
   uint8_t ep_in;
   uint8_t port_count;
-  uint8_t status_change; // data from status change interrupt endpoint
 
+  CFG_TUH_MEM_ALIGN uint8_t status_change;
   CFG_TUH_MEM_ALIGN hub_port_status_response_t port_status;
   CFG_TUH_MEM_ALIGN hub_status_response_t hub_status;
 } hub_interface_t;
diff --git a/src/host/usbh.c b/src/host/usbh.c
index 1418cfb79..c8eb91be3 100644
--- a/src/host/usbh.c
+++ b/src/host/usbh.c
@@ -38,17 +38,19 @@
 //--------------------------------------------------------------------+
 
 #ifndef CFG_TUH_TASK_QUEUE_SZ
-#define CFG_TUH_TASK_QUEUE_SZ   16
+  #define CFG_TUH_TASK_QUEUE_SZ   16
 #endif
 
 #ifndef CFG_TUH_INTERFACE_MAX
-#define CFG_TUH_INTERFACE_MAX   8
+  #define CFG_TUH_INTERFACE_MAX   8
 #endif
 
-// Debug level, TUSB_CFG_DEBUG must be at least this level for debug message
-#define USBH_DEBUG   2
+// Level where CFG_TUSB_DEBUG must be at least for USBH is logged
+#ifndef CFG_TUH_LOG_LEVEL
+  #define CFG_TUH_LOG_LEVEL   2
+#endif
 
-#define TU_LOG_USBH(...)   TU_LOG(USBH_DEBUG, __VA_ARGS__)
+#define TU_LOG_USBH(...)   TU_LOG(CFG_TUH_LOG_LEVEL, __VA_ARGS__)
 
 //--------------------------------------------------------------------+
 // USBH-HCD common data structure
@@ -322,12 +324,12 @@ bool tuh_init(uint8_t controller_id)
   if ( tuh_inited() ) return true;
 
   TU_LOG_USBH("USBH init on controller %u\r\n", controller_id);
-  TU_LOG_INT(USBH_DEBUG, sizeof(usbh_device_t));
-  TU_LOG_INT(USBH_DEBUG, sizeof(hcd_event_t));
-  TU_LOG_INT(USBH_DEBUG, sizeof(_ctrl_xfer));
-  TU_LOG_INT(USBH_DEBUG, sizeof(tuh_xfer_t));
-  TU_LOG_INT(USBH_DEBUG, sizeof(tu_fifo_t));
-  TU_LOG_INT(USBH_DEBUG, sizeof(tu_edpt_stream_t));
+  TU_LOG_INT(CFG_TUH_LOG_LEVEL, sizeof(usbh_device_t));
+  TU_LOG_INT(CFG_TUH_LOG_LEVEL, sizeof(hcd_event_t));
+  TU_LOG_INT(CFG_TUH_LOG_LEVEL, sizeof(_ctrl_xfer));
+  TU_LOG_INT(CFG_TUH_LOG_LEVEL, sizeof(tuh_xfer_t));
+  TU_LOG_INT(CFG_TUH_LOG_LEVEL, sizeof(tu_fifo_t));
+  TU_LOG_INT(CFG_TUH_LOG_LEVEL, sizeof(tu_edpt_stream_t));
 
   // Event queue
   _usbh_q = osal_queue_create( &_usbh_qdef );
@@ -450,39 +452,28 @@ void tuh_task_ext(uint32_t timeout_ms, bool in_isr)
         TU_LOG_USBH("on EP %02X with %u bytes: %s\r\n", ep_addr, (unsigned int) event.xfer_complete.len,
                     tu_str_xfer_result[event.xfer_complete.result]);
 
-        if (event.dev_addr == 0)
-        {
+        if (event.dev_addr == 0) {
           // device 0 only has control endpoint
           TU_ASSERT(epnum == 0, );
           usbh_control_xfer_cb(event.dev_addr, ep_addr, (xfer_result_t) event.xfer_complete.result, event.xfer_complete.len);
-        }
-        else
-        {
+        } else {
           usbh_device_t* dev = get_device(event.dev_addr);
           TU_VERIFY(dev && dev->connected, );
 
           dev->ep_status[epnum][ep_dir].busy    = 0;
           dev->ep_status[epnum][ep_dir].claimed = 0;
 
-          if ( 0 == epnum )
-          {
-            usbh_control_xfer_cb(event.dev_addr, ep_addr, (xfer_result_t) event.xfer_complete.result, event.xfer_complete.len);
-          }else
-          {
-            uint8_t drv_id = dev->ep2drv[epnum][ep_dir];
-            if(drv_id < USBH_CLASS_DRIVER_COUNT)
-            {
-              TU_LOG_USBH("%s xfer callback\r\n", usbh_class_drivers[drv_id].name);
-              usbh_class_drivers[drv_id].xfer_cb(event.dev_addr, ep_addr, (xfer_result_t) event.xfer_complete.result, event.xfer_complete.len);
-            }
-            else
-            {
-#if CFG_TUH_API_EDPT_XFER
-              tuh_xfer_cb_t complete_cb = dev->ep_callback[epnum][ep_dir].complete_cb;
-              if ( complete_cb )
-              {
-                tuh_xfer_t xfer =
-                {
+          if ( 0 == epnum ) {
+            usbh_control_xfer_cb(event.dev_addr, ep_addr, (xfer_result_t) event.xfer_complete.result,
+                                 event.xfer_complete.len);
+          }else {
+            // Prefer application callback over built-in one if available. This occurs when tuh_edpt_xfer() is used
+            // with enabled driver e.g HID endpoint
+            #if CFG_TUH_API_EDPT_XFER
+            tuh_xfer_cb_t const complete_cb = dev->ep_callback[epnum][ep_dir].complete_cb;
+            if ( complete_cb ) {
+              // re-construct xfer info
+              tuh_xfer_t xfer = {
                   .daddr       = event.dev_addr,
                   .ep_addr     = ep_addr,
                   .result      = event.xfer_complete.result,
@@ -491,16 +482,21 @@ void tuh_task_ext(uint32_t timeout_ms, bool in_isr)
                   .buffer      = NULL, // not available
                   .complete_cb = complete_cb,
                   .user_data   = dev->ep_callback[epnum][ep_dir].user_data
-                };
+              };
 
-                complete_cb(&xfer);
-              }else
-#endif
-              {
+              complete_cb(&xfer);
+            }else
+            #endif
+            {
+              uint8_t const drv_id = dev->ep2drv[epnum][ep_dir];
+              if ( drv_id < USBH_CLASS_DRIVER_COUNT ) {
+                TU_LOG_USBH("%s xfer callback\r\n", usbh_class_drivers[drv_id].name);
+                usbh_class_drivers[drv_id].xfer_cb(event.dev_addr, ep_addr, (xfer_result_t) event.xfer_complete.result,
+                                                   event.xfer_complete.len);
+              } else {
                 // no driver/callback responsible for this transfer
-                TU_ASSERT(false, );
+                TU_ASSERT(false,);
               }
-
             }
           }
         }
@@ -565,7 +561,7 @@ bool tuh_control_xfer (tuh_xfer_t* xfer)
   TU_LOG_USBH("[%u:%u] %s: ", rhport, daddr,
               (xfer->setup->bmRequestType_bit.type == TUSB_REQ_TYPE_STANDARD && xfer->setup->bRequest <= TUSB_REQ_SYNCH_FRAME) ?
                   tu_str_std_request[xfer->setup->bRequest] : "Class Request");
-  TU_LOG_PTR(USBH_DEBUG, xfer->setup);
+  TU_LOG_PTR(CFG_TUH_LOG_LEVEL, xfer->setup);
   TU_LOG_USBH("\r\n");
 
   if (xfer->complete_cb)
@@ -671,7 +667,7 @@ static bool usbh_control_xfer_cb (uint8_t dev_addr, uint8_t ep_addr, xfer_result
         if (request->wLength)
         {
           TU_LOG_USBH("[%u:%u] Control data:\r\n", rhport, dev_addr);
-          TU_LOG_MEM(USBH_DEBUG, _ctrl_xfer.buffer, xferred_bytes, 2);
+          TU_LOG_MEM(CFG_TUH_LOG_LEVEL, _ctrl_xfer.buffer, xferred_bytes, 2);
         }
 
         _ctrl_xfer.actual_len = (uint16_t) xferred_bytes;
@@ -714,6 +710,25 @@ bool tuh_edpt_xfer(tuh_xfer_t* xfer)
   return true;
 }
 
+bool tuh_edpt_abort_xfer(uint8_t daddr, uint8_t ep_addr) {
+  usbh_device_t* dev = get_device(daddr);
+  TU_VERIFY(dev);
+
+  uint8_t const epnum = tu_edpt_number(ep_addr);
+  uint8_t const dir   = tu_edpt_dir(ep_addr);
+
+  // skip if not busy
+  TU_VERIFY(dev->ep_status[epnum][dir].busy);
+
+  bool const ret = hcd_edpt_abort_xfer(dev->rhport, daddr, ep_addr);
+  if (ret) {
+    // mark as ready if transfer is aborted
+    dev->ep_status[epnum][dir].busy = false;
+  }
+
+  return ret;
+}
+
 //--------------------------------------------------------------------+
 // USBH API For Class Driver
 //--------------------------------------------------------------------+
@@ -745,7 +760,7 @@ void usbh_int_set(bool enabled)
 // Endpoint API
 //--------------------------------------------------------------------+
 
-// TODO has some duplication code with device, refactor later
+// Claim an endpoint for transfer
 bool usbh_edpt_claim(uint8_t dev_addr, uint8_t ep_addr)
 {
   // Note: addr0 only use tuh_control_xfer
@@ -761,7 +776,7 @@ bool usbh_edpt_claim(uint8_t dev_addr, uint8_t ep_addr)
   return true;
 }
 
-// TODO has some duplication code with device, refactor later
+// Release an claimed endpoint due to failed transfer attempt
 bool usbh_edpt_release(uint8_t dev_addr, uint8_t ep_addr)
 {
   // Note: addr0 only use tuh_control_xfer
@@ -777,7 +792,7 @@ bool usbh_edpt_release(uint8_t dev_addr, uint8_t ep_addr)
   return true;
 }
 
-// TODO has some duplication code with device, refactor later
+// Submit an transfer
 bool usbh_edpt_xfer_with_callback(uint8_t dev_addr, uint8_t ep_addr, uint8_t * buffer, uint16_t total_bytes,
                                   tuh_xfer_cb_t complete_cb, uintptr_t user_data)
 {
@@ -844,14 +859,13 @@ bool tuh_edpt_open(uint8_t dev_addr, tusb_desc_endpoint_t const * desc_ep)
   return hcd_edpt_open(usbh_get_rhport(dev_addr), dev_addr, desc_ep);
 }
 
-bool usbh_edpt_busy(uint8_t dev_addr, uint8_t ep_addr)
-{
-  uint8_t const epnum = tu_edpt_number(ep_addr);
-  uint8_t const dir   = tu_edpt_dir(ep_addr);
-
+bool usbh_edpt_busy(uint8_t dev_addr, uint8_t ep_addr) {
   usbh_device_t* dev = get_device(dev_addr);
   TU_VERIFY(dev);
 
+  uint8_t const epnum = tu_edpt_number(ep_addr);
+  uint8_t const dir   = tu_edpt_dir(ep_addr);
+
   return dev->ep_status[epnum][dir].busy;
 }
 
@@ -1186,7 +1200,7 @@ static void process_removing_device(uint8_t rhport, uint8_t hub_addr, uint8_t hu
       TU_LOG_USBH("Device unplugged address = %u\r\n", daddr);
 
       if (is_hub_addr(daddr)) {
-        TU_LOG(USBH_DEBUG, "  is a HUB device %u\r\n", daddr);
+        TU_LOG(CFG_TUH_LOG_LEVEL, "  is a HUB device %u\r\n", daddr);
 
         // Submit removed event If the device itself is a hub (un-rolled recursive)
         // TODO a better to unroll recursrive is using array of removing_hubs and mark it here
@@ -1657,7 +1671,7 @@ static bool _parse_configuration_descriptor(uint8_t dev_addr, tusb_desc_configur
 
       if( drv_id >= USBH_CLASS_DRIVER_COUNT )
       {
-        TU_LOG(USBH_DEBUG, "Interface %u: class = %u subclass = %u protocol = %u is not supported\r\n",
+        TU_LOG(CFG_TUH_LOG_LEVEL, "Interface %u: class = %u subclass = %u protocol = %u is not supported\r\n",
                desc_itf->bInterfaceNumber, desc_itf->bInterfaceClass, desc_itf->bInterfaceSubClass, desc_itf->bInterfaceProtocol);
       }
     }
@@ -1695,7 +1709,7 @@ void usbh_driver_set_config_complete(uint8_t dev_addr, uint8_t itf_num)
 
     if (is_hub_addr(dev_addr))
     {
-      TU_LOG(USBH_DEBUG, "HUB address = %u is mounted\r\n", dev_addr);
+      TU_LOG(CFG_TUH_LOG_LEVEL, "HUB address = %u is mounted\r\n", dev_addr);
     }else
     {
       // Invoke callback if available
diff --git a/src/host/usbh.h b/src/host/usbh.h
index 0f969a46a..0d56f40c9 100644
--- a/src/host/usbh.h
+++ b/src/host/usbh.h
@@ -172,8 +172,12 @@ bool tuh_control_xfer(tuh_xfer_t* xfer);
 //  - sync : blocking if complete callback is NULL.
 bool tuh_edpt_xfer(tuh_xfer_t* xfer);
 
-// Open an non-control endpoint
-bool tuh_edpt_open(uint8_t dev_addr, tusb_desc_endpoint_t const * desc_ep);
+// Open a non-control endpoint
+bool tuh_edpt_open(uint8_t daddr, tusb_desc_endpoint_t const * desc_ep);
+
+// Abort a queued transfer. Note: it can only abort transfer that has not been started
+// Return true if a queued transfer is aborted, false if there is no transfer to abort
+bool tuh_edpt_abort_xfer(uint8_t daddr, uint8_t ep_addr);
 
 // Set Configuration (control transfer)
 // config_num = 0 will un-configure device. Note: config_num = config_descriptor_index + 1
diff --git a/src/portable/bridgetek/ft9xx/dcd_ft9xx.c b/src/portable/bridgetek/ft9xx/dcd_ft9xx.c
index 95dfda49c..f02415904 100644
--- a/src/portable/bridgetek/ft9xx/dcd_ft9xx.c
+++ b/src/portable/bridgetek/ft9xx/dcd_ft9xx.c
@@ -51,7 +51,7 @@ extern int8_t board_ft9xx_vbus(void);
 extern int board_uart_write(void const *buf, int len);
 
 // Static array to store an incoming SETUP request for processing by tinyusb.
-CFG_TUSB_MEM_SECTION CFG_TUSB_MEM_ALIGN
+CFG_TUD_MEM_SECTION CFG_TUSB_MEM_ALIGN
 static uint8_t _ft9xx_setup_packet[8];
 
 struct ft9xx_xfer_state
diff --git a/src/portable/chipidea/ci_fs/dcd_ci_fs.c b/src/portable/chipidea/ci_fs/dcd_ci_fs.c
index 37265df8b..9327e09d8 100644
--- a/src/portable/chipidea/ci_fs/dcd_ci_fs.c
+++ b/src/portable/chipidea/ci_fs/dcd_ci_fs.c
@@ -116,7 +116,7 @@ typedef struct
 // INTERNAL OBJECT & FUNCTION DECLARATION
 //--------------------------------------------------------------------+
 // BDT(Buffer Descriptor Table) must be 256-byte aligned
-CFG_TUSB_MEM_SECTION TU_ATTR_ALIGNED(512) static dcd_data_t _dcd;
+CFG_TUD_MEM_SECTION TU_ATTR_ALIGNED(512) static dcd_data_t _dcd;
 
 TU_VERIFY_STATIC( sizeof(_dcd.bdt) == 512, "size is not correct" );
 
diff --git a/src/portable/chipidea/ci_hs/ci_hs_imxrt.h b/src/portable/chipidea/ci_hs/ci_hs_imxrt.h
index c14f00431..c59c107ff 100644
--- a/src/portable/chipidea/ci_hs/ci_hs_imxrt.h
+++ b/src/portable/chipidea/ci_hs/ci_hs_imxrt.h
@@ -68,25 +68,34 @@ TU_ATTR_ALWAYS_INLINE static inline bool imxrt_is_cache_mem(uintptr_t addr) {
   return !(0x20000000 <= addr && addr < 0x20100000);
 }
 
-TU_ATTR_ALWAYS_INLINE static inline void imxrt_dcache_clean(void const* addr, uint32_t data_size) {
+TU_ATTR_ALWAYS_INLINE static inline bool imxrt_dcache_clean(void const* addr, uint32_t data_size) {
   const uintptr_t addr32 = (uintptr_t) addr;
   if (imxrt_is_cache_mem(addr32)) {
+    TU_ASSERT(tu_is_aligned32(addr32));
     SCB_CleanDCache_by_Addr((uint32_t *) addr32, (int32_t) data_size);
   }
+  return true;
 }
 
-TU_ATTR_ALWAYS_INLINE static inline void imxrt_dcache_invalidate(void const* addr, uint32_t data_size) {
+TU_ATTR_ALWAYS_INLINE static inline bool imxrt_dcache_invalidate(void const* addr, uint32_t data_size) {
   const uintptr_t addr32 = (uintptr_t) addr;
   if (imxrt_is_cache_mem(addr32)) {
+    // Invalidating does not push cached changes back to RAM so we need to be
+    // *very* careful when we do it. If we're not aligned, then we risk resetting
+    // values back to their RAM state.
+    TU_ASSERT(tu_is_aligned32(addr32));
     SCB_InvalidateDCache_by_Addr((void*) addr32, (int32_t) data_size);
   }
+  return true;
 }
 
-TU_ATTR_ALWAYS_INLINE static inline void imxrt_dcache_clean_invalidate(void const* addr, uint32_t data_size) {
+TU_ATTR_ALWAYS_INLINE static inline bool imxrt_dcache_clean_invalidate(void const* addr, uint32_t data_size) {
   const uintptr_t addr32 = (uintptr_t) addr;
   if (imxrt_is_cache_mem(addr32)) {
+    TU_ASSERT(tu_is_aligned32(addr32));
     SCB_CleanInvalidateDCache_by_Addr((uint32_t *) addr32, (int32_t) data_size);
   }
+  return true;
 }
 
 #endif
diff --git a/src/portable/chipidea/ci_hs/dcd_ci_hs.c b/src/portable/chipidea/ci_hs/dcd_ci_hs.c
index f50550d33..f9ec666e5 100644
--- a/src/portable/chipidea/ci_hs/dcd_ci_hs.c
+++ b/src/portable/chipidea/ci_hs/dcd_ci_hs.c
@@ -175,7 +175,7 @@ typedef struct {
   dcd_qtd_t qtd[TUP_DCD_ENDPOINT_MAX][2] TU_ATTR_ALIGNED(32);
 }dcd_data_t;
 
-CFG_TUSB_MEM_SECTION TU_ATTR_ALIGNED(2048)
+CFG_TUD_MEM_SECTION TU_ATTR_ALIGNED(2048)
 static dcd_data_t _dcd_data;
 
 //--------------------------------------------------------------------+
diff --git a/src/portable/chipidea/ci_hs/hcd_ci_hs.c b/src/portable/chipidea/ci_hs/hcd_ci_hs.c
index 56167b8f6..6ee91e18e 100644
--- a/src/portable/chipidea/ci_hs/hcd_ci_hs.c
+++ b/src/portable/chipidea/ci_hs/hcd_ci_hs.c
@@ -41,16 +41,16 @@
 #if CFG_TUSB_MCU == OPT_MCU_MIMXRT1XXX
   #include "ci_hs_imxrt.h"
 
-  void hcd_dcache_clean(void const* addr, uint32_t data_size) {
-    imxrt_dcache_clean(addr, data_size);
+  bool hcd_dcache_clean(void const* addr, uint32_t data_size) {
+    return imxrt_dcache_clean(addr, data_size);
   }
 
-  void hcd_dcache_invalidate(void const* addr, uint32_t data_size) {
-    imxrt_dcache_invalidate(addr, data_size);
+  bool hcd_dcache_invalidate(void const* addr, uint32_t data_size) {
+    return imxrt_dcache_invalidate(addr, data_size);
   }
 
-  void hcd_dcache_clean_invalidate(void const* addr, uint32_t data_size) {
-    imxrt_dcache_clean_invalidate(addr, data_size);
+  bool hcd_dcache_clean_invalidate(void const* addr, uint32_t data_size) {
+    return imxrt_dcache_clean_invalidate(addr, data_size);
   }
 
 #elif TU_CHECK_MCU(OPT_MCU_LPC18XX, OPT_MCU_LPC43XX)
diff --git a/src/portable/ehci/ehci.c b/src/portable/ehci/ehci.c
index 38711e382..9be24c888 100644
--- a/src/portable/ehci/ehci.c
+++ b/src/portable/ehci/ehci.c
@@ -48,8 +48,8 @@
 #ifdef TUP_USBIP_CHIPIDEA_HS
   // NXP Transdimension: 8 elements
   #define FRAMELIST_SIZE_BIT_VALUE      7u
-  #define FRAMELIST_SIZE_USBCMD_VALUE   (((FRAMELIST_SIZE_BIT_VALUE &  3) << EHCI_USBCMD_POS_FRAMELIST_SIZE) | \
-                                         ((FRAMELIST_SIZE_BIT_VALUE >> 2) << EHCI_USBCMD_POS_NXP_FRAMELIST_SIZE_MSB))
+  #define FRAMELIST_SIZE_USBCMD_VALUE   (((FRAMELIST_SIZE_BIT_VALUE &  3) << EHCI_USBCMD_FRAMELIST_SIZE_SHIFT) | \
+                                         ((FRAMELIST_SIZE_BIT_VALUE >> 2) << EHCI_USBCMD_CHIPIDEA_FRAMELIST_SIZE_MSB_SHIFT))
 #else
   // STD EHCI: 256 elements
   #define FRAMELIST_SIZE_BIT_VALUE      2u
@@ -126,52 +126,50 @@ static inline void print_intr(uint32_t intr) {
 //--------------------------------------------------------------------+
 // PROTOTYPE
 //--------------------------------------------------------------------+
-static inline ehci_link_t* get_period_head(uint8_t rhport, uint32_t interval_ms)
-{
-  (void) rhport;
-  return (ehci_link_t*) &ehci_data.period_head_arr[ tu_log2( tu_min32(FRAMELIST_SIZE, interval_ms) ) ];
-}
 
-static inline ehci_qhd_t* qhd_control(uint8_t dev_addr)
-{
-  return &ehci_data.control[dev_addr].qhd;
-}
+// weak dcache for non-cacheable MCU
+TU_ATTR_WEAK bool hcd_dcache_clean(void const* addr, uint32_t data_size) { (void) addr; (void) data_size; return true; }
+TU_ATTR_WEAK bool hcd_dcache_invalidate(void const* addr, uint32_t data_size) { (void) addr; (void) data_size; return true; }
+TU_ATTR_WEAK bool hcd_dcache_clean_invalidate(void const* addr, uint32_t data_size) { (void) addr; (void) data_size; return true; }
 
-static inline ehci_qhd_t* qhd_async_head(uint8_t rhport)
-{
-  (void) rhport;
-  // control qhd of dev0 is used as async head
-  return qhd_control(0);
-}
-
-static inline ehci_qtd_t* qtd_control(uint8_t dev_addr)
-{
-  return &ehci_data.control[dev_addr].qtd;
-}
-
-
-static inline ehci_qhd_t* qhd_next (ehci_qhd_t const * p_qhd);
-static inline ehci_qhd_t* qhd_find_free (void);
-static inline ehci_qhd_t* qhd_get_from_addr (uint8_t dev_addr, uint8_t ep_addr);
+TU_ATTR_ALWAYS_INLINE static inline ehci_qhd_t* qhd_control(uint8_t dev_addr);
+TU_ATTR_ALWAYS_INLINE static inline ehci_qhd_t* qhd_next (ehci_qhd_t const * p_qhd);
+TU_ATTR_ALWAYS_INLINE static inline ehci_qhd_t* qhd_find_free (void);
+static ehci_qhd_t* qhd_get_from_addr (uint8_t dev_addr, uint8_t ep_addr);
 static void qhd_init(ehci_qhd_t *p_qhd, uint8_t dev_addr, tusb_desc_endpoint_t const * ep_desc);
 static void qhd_attach_qtd(ehci_qhd_t *qhd, ehci_qtd_t *qtd);
+static void qhd_remove_qtd(ehci_qhd_t *qhd);
 
-static inline ehci_qtd_t* qtd_find_free (void);
+TU_ATTR_ALWAYS_INLINE static inline ehci_qtd_t* qtd_control(uint8_t dev_addr);
+TU_ATTR_ALWAYS_INLINE static inline ehci_qtd_t* qtd_find_free (void);
 static void qtd_init (ehci_qtd_t* qtd, void const* buffer, uint16_t total_bytes);
 
-static inline void list_insert (ehci_link_t *current, ehci_link_t *new, uint8_t new_type);
-static inline ehci_link_t* list_next (ehci_link_t const *p_link);
+TU_ATTR_ALWAYS_INLINE static inline ehci_link_t* list_get_period_head(uint8_t rhport, uint32_t interval_ms);
+TU_ATTR_ALWAYS_INLINE static inline ehci_qhd_t* list_get_async_head(uint8_t rhport);
+TU_ATTR_ALWAYS_INLINE static inline void list_insert (ehci_link_t *current, ehci_link_t *new, uint8_t new_type);
+TU_ATTR_ALWAYS_INLINE static inline ehci_link_t* list_next (ehci_link_t const *p_link);
+static void list_remove_qhd_by_daddr(ehci_link_t* list_head, uint8_t dev_addr);
 
-TU_ATTR_WEAK void hcd_dcache_clean(void const* addr, uint32_t data_size) {
-  (void) addr; (void) data_size;
-}
-
-TU_ATTR_WEAK void hcd_dcache_invalidate(void const* addr, uint32_t data_size) {
-  (void) addr; (void) data_size;
+static void ehci_disable_schedule(ehci_registers_t* regs, bool is_period) {
+  // maybe have a timeout for status
+  if (is_period) {
+    regs->command_bm.periodic_enable = 0;
+    while(regs->status_bm.periodic_status) {}
+  } else {
+    regs->command_bm.async_enable = 0;
+    while(regs->status_bm.async_status) {} // should have a timeout
+  }
 }
 
-TU_ATTR_WEAK void hcd_dcache_clean_invalidate(void const* addr, uint32_t data_size) {
-  (void) addr; (void) data_size;
+static void ehci_enable_schedule(ehci_registers_t* regs, bool is_period) {
+  // maybe have a timeout for status
+  if (is_period) {
+    regs->command_bm.periodic_enable = 1;
+    while ( 0 == regs->status_bm.periodic_status ) {}
+  } else {
+    regs->command_bm.async_enable = 1;
+    while( 0 == regs->status_bm.async_status ) {}
+  }
 }
 
 //--------------------------------------------------------------------+
@@ -228,43 +226,6 @@ tusb_speed_t hcd_port_speed_get(uint8_t rhport)
   return (tusb_speed_t) ehci_data.regs->portsc_bm.nxp_port_speed; // NXP specific port speed
 }
 
-static void list_remove_qhd_by_daddr(ehci_link_t* list_head, uint8_t dev_addr) {
-  ehci_link_t* prev = list_head;
-
-  while (prev && !prev->terminate) {
-    ehci_qhd_t* qhd = (ehci_qhd_t*) (uintptr_t) list_next(prev);
-
-    // done if loop back to head
-    if ( (uintptr_t) qhd == (uintptr_t) list_head) {
-      break;
-    }
-
-    if ( qhd->dev_addr == dev_addr ) {
-      // TODO deactivate all TD, wait for QHD to inactive before removal
-      prev->address = qhd->next.address;
-
-      // EHCI 4.8.2 link the removed qhd's next to async head (which always reachable by Host Controller)
-      qhd->next.address = ((uint32_t) list_head) | (EHCI_QTYPE_QHD << 1);
-
-      if ( qhd->int_smask )
-      {
-        // period list queue element is guarantee to be free in the next frame (1 ms)
-        qhd->used = 0;
-      }else
-      {
-        // async list use async advance handshake
-        // mark as removing, will completely re-usable when async advance isr occurs
-        qhd->removing = 1;
-      }
-
-      hcd_dcache_clean(qhd, sizeof(ehci_qhd_t));
-      hcd_dcache_clean(prev, sizeof(ehci_qhd_t));
-    }else {
-      prev = list_next(prev);
-    }
-  }
-}
-
 // Close all opened endpoint belong to this device
 void hcd_device_close(uint8_t rhport, uint8_t daddr)
 {
@@ -274,7 +235,7 @@ void hcd_device_close(uint8_t rhport, uint8_t daddr)
   }
 
   // Remove from async list
-  list_remove_qhd_by_daddr((ehci_link_t *) qhd_async_head(rhport), daddr);
+  list_remove_qhd_by_daddr((ehci_link_t *) list_get_async_head(rhport), daddr);
 
   // Remove from all interval period list
   for(uint8_t i = 0; i < TU_ARRAY_SIZE(ehci_data.period_head_arr); i++) {
@@ -286,37 +247,42 @@ void hcd_device_close(uint8_t rhport, uint8_t daddr)
 }
 
 static void init_periodic_list(uint8_t rhport) {
+  (void) rhport;
+
   // Build the polling interval tree with 1 ms, 2 ms, 4 ms and 8 ms (framesize) only
   for ( uint32_t i = 0; i < TU_ARRAY_SIZE(ehci_data.period_head_arr); i++ ) {
     ehci_data.period_head_arr[i].int_smask          = 1; // queue head in period list must have smask non-zero
     ehci_data.period_head_arr[i].qtd_overlay.halted = 1; // dummy node, always inactive
   }
 
-  ehci_link_t * const framelist  = ehci_data.period_framelist;
-  ehci_link_t * const period_1ms = get_period_head(rhport, 1u);
-
+  // TODO EHCI_FRAMELIST_SIZE with other size than 8
   // all links --> period_head_arr[0] (1ms)
   // 0, 2, 4, 6 etc --> period_head_arr[1] (2ms)
   // 1, 5 --> period_head_arr[2] (4ms)
   // 3 --> period_head_arr[3] (8ms)
 
-  // TODO EHCI_FRAMELIST_SIZE with other size than 8
+  ehci_link_t * const framelist  = ehci_data.period_framelist;
+  ehci_link_t * const head_1ms = (ehci_link_t *) &ehci_data.period_head_arr[0];
+  ehci_link_t * const head_2ms = (ehci_link_t *) &ehci_data.period_head_arr[1];
+  ehci_link_t * const head_4ms = (ehci_link_t *) &ehci_data.period_head_arr[2];
+  ehci_link_t * const head_8ms = (ehci_link_t *) &ehci_data.period_head_arr[3];
+
   for (uint32_t i = 0; i < FRAMELIST_SIZE; i++) {
-    framelist[i].address = (uint32_t) period_1ms;
+    framelist[i].address = (uint32_t) head_1ms;
     framelist[i].type = EHCI_QTYPE_QHD;
   }
 
   for (uint32_t i = 0; i < FRAMELIST_SIZE; i += 2) {
-    list_insert(framelist + i, get_period_head(rhport, 2u), EHCI_QTYPE_QHD);
+    list_insert(framelist + i, head_2ms, EHCI_QTYPE_QHD);
   }
 
   for (uint32_t i = 1; i < FRAMELIST_SIZE; i += 4) {
-    list_insert(framelist + i, get_period_head(rhport, 4u), EHCI_QTYPE_QHD);
+    list_insert(framelist + i, head_4ms, EHCI_QTYPE_QHD);
   }
 
-  list_insert(framelist + 3, get_period_head(rhport, 8u), EHCI_QTYPE_QHD);
+  list_insert(framelist + 3, head_8ms, EHCI_QTYPE_QHD);
 
-  period_1ms->terminate    = 1;
+  head_1ms->terminate = 1;
 }
 
 bool ehci_init(uint8_t rhport, uint32_t capability_reg, uint32_t operatial_reg)
@@ -341,18 +307,18 @@ bool ehci_init(uint8_t rhport, uint32_t capability_reg, uint32_t operatial_reg)
   regs->status = (EHCI_INT_MASK_ALL & ~EHCI_INT_MASK_PORT_CHANGE);
 
   // Enable interrupts
-  regs->inten  = EHCI_INT_MASK_ERROR | EHCI_INT_MASK_PORT_CHANGE | EHCI_INT_MASK_ASYNC_ADVANCE |
-                 EHCI_INT_MASK_NXP_PERIODIC | EHCI_INT_MASK_NXP_ASYNC | EHCI_INT_MASK_FRAMELIST_ROLLOVER;
+  regs->inten  = EHCI_INT_MASK_USB | EHCI_INT_MASK_ERROR | EHCI_INT_MASK_PORT_CHANGE |
+                 EHCI_INT_MASK_ASYNC_ADVANCE | EHCI_INT_MASK_FRAMELIST_ROLLOVER;
 
   //------------- Asynchronous List -------------//
-  ehci_qhd_t * const async_head = qhd_async_head(rhport);
+  ehci_qhd_t * const async_head = list_get_async_head(rhport);
   tu_memclr(async_head, sizeof(ehci_qhd_t));
 
-  async_head->next.address                    = (uint32_t) async_head; // circular list, next is itself
-  async_head->next.type                       = EHCI_QTYPE_QHD;
-  async_head->head_list_flag                  = 1;
-  async_head->qtd_overlay.halted              = 1; // inactive most of time
-  async_head->qtd_overlay.next.terminate      = 1; // TODO removed if verified
+  async_head->next.address               = (uint32_t) async_head; // circular list, next is itself
+  async_head->next.type                  = EHCI_QTYPE_QHD;
+  async_head->head_list_flag             = 1;
+  async_head->qtd_overlay.halted         = 1; // inactive most of time
+  async_head->qtd_overlay.next.terminate = 1; // TODO removed if verified
 
   regs->async_list_addr = (uint32_t) async_head;
 
@@ -366,8 +332,7 @@ bool ehci_init(uint8_t rhport, uint32_t capability_reg, uint32_t operatial_reg)
   regs->nxp_tt_control = 0;
 
   //------------- USB CMD Register -------------//
-  regs->command |= TU_BIT(EHCI_USBCMD_POS_RUN_STOP) | TU_BIT(EHCI_USBCMD_POS_ASYNC_ENABLE) |
-                   TU_BIT(EHCI_USBCMD_POS_PERIOD_ENABLE) |  // TODO enable period list only there is int/iso endpoint
+  regs->command |= EHCI_USBCMD_RUN_STOP | EHCI_USBCMD_PERIOD_SCHEDULE_ENABLE | EHCI_USBCMD_ASYNC_SCHEDULE_ENABLE |
                    FRAMELIST_SIZE_USBCMD_VALUE;
 
   //------------- ConfigFlag Register (skip) -------------//
@@ -377,7 +342,7 @@ bool ehci_init(uint8_t rhport, uint32_t capability_reg, uint32_t operatial_reg)
   if (ehci_data.cap_regs->hcsparams_bm.port_power_control) {
     // mask out all change bits since they are Write 1 to clear
     uint32_t portsc = (regs->portsc & ~EHCI_PORTSC_MASK_W1C);
-    portsc |= ECHI_PORTSC_MASK_PORT_POWER;
+    portsc |= EHCI_PORTSC_MASK_PORT_POWER;
 
     regs->portsc = portsc;
   }
@@ -426,11 +391,11 @@ bool hcd_edpt_open(uint8_t rhport, uint8_t dev_addr, tusb_desc_endpoint_t const
   {
     case TUSB_XFER_CONTROL:
     case TUSB_XFER_BULK:
-      list_head = (ehci_link_t*) qhd_async_head(rhport);
+      list_head = (ehci_link_t*) list_get_async_head(rhport);
     break;
 
     case TUSB_XFER_INTERRUPT:
-      list_head = get_period_head(rhport, p_qhd->interval_ms);
+      list_head = list_get_period_head(rhport, p_qhd->interval_ms);
     break;
 
     case TUSB_XFER_ISOCHRONOUS:
@@ -439,10 +404,8 @@ bool hcd_edpt_open(uint8_t rhport, uint8_t dev_addr, tusb_desc_endpoint_t const
 
     default: break;
   }
-
   TU_ASSERT(list_head);
 
-  // TODO might need to disable async/period list
   list_insert(list_head, (ehci_link_t*) p_qhd, EHCI_QTYPE_QHD);
 
   hcd_dcache_clean(p_qhd, sizeof(ehci_qhd_t));
@@ -476,20 +439,25 @@ bool hcd_edpt_xfer(uint8_t rhport, uint8_t dev_addr, uint8_t ep_addr, uint8_t *
   uint8_t const epnum = tu_edpt_number(ep_addr);
   uint8_t const dir   = tu_edpt_dir(ep_addr);
 
-  ehci_qhd_t* qhd;
+  ehci_qhd_t* qhd = qhd_get_from_addr(dev_addr, ep_addr);
   ehci_qtd_t* qtd;
 
   if (epnum == 0) {
-    qhd = qhd_control(dev_addr);
-    qtd = qtd_control(dev_addr);
+    // Control endpoint never be stalled. Skip reset Data Toggle since it is fixed per stage
+    if (qhd->qtd_overlay.halted) {
+      qhd->qtd_overlay.halted = false;
+    }
 
+    qtd = qtd_control(dev_addr);
     qtd_init(qtd, buffer, buflen);
 
     // first data toggle is always 1 (data & setup stage)
     qtd->data_toggle = 1;
     qtd->pid = dir ? EHCI_PID_IN : EHCI_PID_OUT;
   } else {
-    qhd = qhd_get_from_addr(dev_addr, ep_addr);
+    // skip if endpoint is halted
+    TU_VERIFY(!qhd->qtd_overlay.halted);
+
     qtd = qtd_find_free();
     TU_ASSERT(qtd);
 
@@ -510,12 +478,45 @@ bool hcd_edpt_xfer(uint8_t rhport, uint8_t dev_addr, uint8_t ep_addr, uint8_t *
   return true;
 }
 
-bool hcd_edpt_clear_stall(uint8_t daddr, uint8_t ep_addr)
-{
+bool hcd_edpt_abort_xfer(uint8_t rhport, uint8_t dev_addr, uint8_t ep_addr) {
+  (void) rhport;
+
+  // TODO ISO not supported yet
+  ehci_qhd_t* qhd = qhd_get_from_addr(dev_addr, ep_addr);
+  ehci_qtd_t * volatile qtd = qhd->attached_qtd;
+  TU_VERIFY(qtd != NULL); // no queued transfer
+
+  hcd_dcache_invalidate(qtd, sizeof(ehci_qtd_t));
+  TU_VERIFY(qtd->active); // transfer is already complete
+
+  // HC is still processing, disable HC list schedule before making changes
+  bool const is_period = (qhd->interval_ms > 0);
+
+  ehci_disable_schedule(ehci_data.regs, is_period);
+
+  // check active bit again just in case HC has just processed the TD
+  bool const still_active = qtd->active;
+  if (still_active) {
+    // remove TD from QH overlay
+    qhd->qtd_overlay.next.terminate = 1;
+    hcd_dcache_clean(qhd, sizeof(ehci_qhd_t));
+
+    // remove TD from QH software list
+    qhd_remove_qtd(qhd);
+  }
+
+  ehci_enable_schedule(ehci_data.regs, is_period);
+
+  return still_active; // true if removed an active transfer
+}
+
+bool hcd_edpt_clear_stall(uint8_t rhport, uint8_t daddr, uint8_t ep_addr) {
+  (void) rhport;
   ehci_qhd_t *qhd = qhd_get_from_addr(daddr, ep_addr);
   qhd->qtd_overlay.halted = 0;
+  qhd->qtd_overlay.data_toggle = 0;
   hcd_dcache_clean_invalidate(qhd, sizeof(ehci_qhd_t));
-  // TODO reset data toggle ?
+
   return true;
 }
 
@@ -541,71 +542,77 @@ void async_advance_isr(uint8_t rhport)
 }
 
 TU_ATTR_ALWAYS_INLINE static inline
-void port_connect_status_change_isr(uint8_t rhport)
-{
+void port_connect_status_change_isr(uint8_t rhport) {
   // NOTE There is an sequence plug->unplug->…..-> plug if device is powering with pre-plugged device
-  if (ehci_data.regs->portsc_bm.current_connect_status)
-  {
+  if ( ehci_data.regs->portsc_bm.current_connect_status ) {
     hcd_port_reset(rhport);
     hcd_event_device_attach(rhport, true);
-  }else // device unplugged
+  } else // device unplugged
   {
     hcd_event_device_remove(rhport, true);
   }
 }
 
+// Check queue head for potential transfer complete (successful or error)
 TU_ATTR_ALWAYS_INLINE static inline
 void qhd_xfer_complete_isr(ehci_qhd_t * qhd) {
-  // examine TD attached to queue head
-  ehci_qtd_t * volatile qtd = (ehci_qtd_t * volatile) qhd->attached_qtd;
-  if (qtd == NULL) return; // no TD attached
-  hcd_dcache_invalidate(qtd, sizeof(ehci_qtd_t));
+  hcd_dcache_invalidate(qhd, sizeof(ehci_qhd_t)); // HC may have updated the overlay
+  volatile ehci_qtd_t *qtd_overlay = &qhd->qtd_overlay;
 
-  // TD is still active, no need to process
-  if (qtd->active) {
-    return;
-  }
+  // process non-active (completed) QHD with attached (scheduled) TD
+  if ( !qtd_overlay->active && qhd->attached_qtd != NULL ) {
+    xfer_result_t xfer_result;
 
-  uint8_t dir = (qtd->pid == EHCI_PID_IN) ? 1 : 0;
-  uint32_t const xferred_bytes = qtd->expected_bytes - qtd->total_bytes;
+    if ( qtd_overlay->halted ) {
+      if (qtd_overlay->xact_err || qtd_overlay->err_count == 0 || qtd_overlay->buffer_err || qtd_overlay->babble_err) {
+        // Error count = 0 often occurs when device disconnected, or other bus-related error
+        xfer_result = XFER_RESULT_FAILED;
+        TU_LOG3("  QHD xfer err count: %d\n", qtd_overlay->err_count);
+        // TU_BREAKPOINT(); // TODO skip unplugged device
+      }else {
+        // no error bits are set, endpoint is halted due to STALL
+        xfer_result = XFER_RESULT_STALLED;
+      }
+    } else {
+      xfer_result = XFER_RESULT_SUCCESS;
+    }
 
-  // invalidate dcache if IN transfer
-  if (dir == 1 && qhd->attached_buffer != 0 && xferred_bytes > 0) {
-    hcd_dcache_invalidate((void*) qhd->attached_buffer, xferred_bytes);
-  }
+    ehci_qtd_t * volatile qtd = qhd->attached_qtd;
+    hcd_dcache_invalidate(qtd, sizeof(ehci_qtd_t)); // HC may have written back TD
 
-  // remove and free TD before invoking callback
-  qhd->attached_qtd = NULL;
-  qhd->attached_buffer = 0;
-  qtd->used = 0; // free QTD
+    uint8_t const dir = (qtd->pid == EHCI_PID_IN) ? 1 : 0;
+    uint32_t const xferred_bytes = qtd->expected_bytes - qtd->total_bytes;
+
+    // invalidate dcache if IN transfer with data
+    if (dir == 1 && qhd->attached_buffer != 0 && xferred_bytes > 0) {
+      hcd_dcache_invalidate((void*) qhd->attached_buffer, xferred_bytes);
+    }
+
+    // remove and free TD before invoking callback
+    qhd_remove_qtd(qhd);
 
-  // notify usbh
-  uint8_t const ep_addr = tu_edpt_addr(qhd->ep_number, dir);
-  hcd_event_xfer_complete(qhd->dev_addr, ep_addr, xferred_bytes, XFER_RESULT_SUCCESS, true);
+    // notify usbh
+    uint8_t const ep_addr = tu_edpt_addr(qhd->ep_number, dir);
+    hcd_event_xfer_complete(qhd->dev_addr, ep_addr, xferred_bytes, xfer_result, true);
+  }
 }
 
 TU_ATTR_ALWAYS_INLINE static inline
-void async_list_xfer_complete_isr(ehci_qhd_t * const async_head)
+void proccess_async_xfer_isr(ehci_qhd_t * const list_head)
 {
-  ehci_qhd_t *p_qhd = async_head;
-  do
-  {
-    hcd_dcache_invalidate(p_qhd, sizeof(ehci_qhd_t));
-
-    // halted or error is processed in error isr
-    if ( !p_qhd->qtd_overlay.halted ) {
-      qhd_xfer_complete_isr(p_qhd);
-    }
+  ehci_qhd_t *qhd = list_head;
 
-    p_qhd = qhd_next(p_qhd);
-  }while(p_qhd != async_head); // async list traversal, stop if loop around
+  do {
+    qhd_xfer_complete_isr(qhd);
+    qhd = qhd_next(qhd);
+  } while ( qhd != list_head ); // async list traversal, stop if loop around
 }
 
 TU_ATTR_ALWAYS_INLINE static inline
-void period_list_xfer_complete_isr(uint8_t rhport, uint32_t interval_ms)
+void process_period_xfer_isr(uint8_t rhport, uint32_t interval_ms)
 {
-  uint32_t const period_1ms_addr = (uint32_t) get_period_head(rhport, 1u);
-  ehci_link_t next_link = * get_period_head(rhport, interval_ms);
+  uint32_t const period_1ms_addr = (uint32_t) list_get_period_head(rhport, 1u);
+  ehci_link_t next_link = *list_get_period_head(rhport, interval_ms);
 
   while (!next_link.terminate) {
     if (interval_ms > 1 && period_1ms_addr == tu_align32(next_link.address)) {
@@ -618,22 +625,13 @@ void period_list_xfer_complete_isr(uint8_t rhport, uint32_t interval_ms)
     switch (next_link.type) {
       case EHCI_QTYPE_QHD: {
         ehci_qhd_t *qhd = (ehci_qhd_t *) entry_addr;
-        hcd_dcache_invalidate(qhd, sizeof(ehci_qhd_t));
-
-        if (!qhd->qtd_overlay.halted) {
-          qhd_xfer_complete_isr(qhd);
-        }
+        qhd_xfer_complete_isr(qhd);
       }
         break;
 
+      // TODO support hs/fs ISO
       case EHCI_QTYPE_ITD:
-        // TODO support hs ISO
-        break;
-
       case EHCI_QTYPE_SITD:
-        // TODO support split ISO
-        break;
-
       case EHCI_QTYPE_FSTN:
       default:
         break;
@@ -643,108 +641,6 @@ void period_list_xfer_complete_isr(uint8_t rhport, uint32_t interval_ms)
   }
 }
 
-// TODO merge with qhd_xfer_complete_isr()
-TU_ATTR_ALWAYS_INLINE static inline
-void qhd_xfer_error_isr(ehci_qhd_t * qhd)
-{
-  volatile ehci_qtd_t *qtd_overlay = &qhd->qtd_overlay;
-
-  // TD has error
-  if (qtd_overlay->halted) {
-    xfer_result_t xfer_result;
-
-    if (qtd_overlay->xact_err || qtd_overlay->err_count == 0 || qtd_overlay->buffer_err || qtd_overlay->babble_err) {
-      // Error count = 0 often occurs when device disconnected, or other bus-related error
-      xfer_result = XFER_RESULT_FAILED;
-    }else {
-      // no error bits are set, endpoint is halted due to STALL
-      xfer_result = XFER_RESULT_STALLED;
-    }
-
-//    if (XFER_RESULT_FAILED == xfer_result ) {
-//      TU_LOG1("  QHD xfer err count: %d\n", qtd_overlay->err_count);
-//      TU_BREAKPOINT(); // TODO skip unplugged device
-//    }
-
-    ehci_qtd_t * volatile qtd = (ehci_qtd_t * volatile) qhd->attached_qtd;
-    TU_ASSERT(qtd, ); // No TD yet, probably a race condition or cache issue !?
-
-    hcd_dcache_invalidate(qtd, sizeof(ehci_qtd_t));
-
-    uint8_t dir = (qtd->pid == EHCI_PID_IN) ? 1 : 0;
-    uint32_t const xferred_bytes = qtd->expected_bytes - qtd->total_bytes;
-
-    // invalidate dcache if IN transfer
-    if (dir == 1 && qhd->attached_buffer != 0 && xferred_bytes > 0) {
-      hcd_dcache_invalidate((void*) qhd->attached_buffer, xferred_bytes);
-    }
-
-    // remove and free TD before invoking callback
-    qhd->attached_qtd = NULL;
-    qhd->attached_buffer = 0;
-    qtd->used = 0; // free QTD
-
-    if (0 == qhd->ep_number ) {
-      // control cannot be halted
-      qhd->qtd_overlay.next.terminate      = 1;
-      qhd->qtd_overlay.alternate.terminate = 1;
-      qhd->qtd_overlay.halted              = 0;
-
-      hcd_dcache_clean(qhd, sizeof(ehci_qhd_t));
-    }
-
-    // notify usbh
-    uint8_t const ep_addr = tu_edpt_addr(qhd->ep_number, dir);
-    hcd_event_xfer_complete(qhd->dev_addr, ep_addr, xferred_bytes, xfer_result, true);
-  }
-}
-
-TU_ATTR_ALWAYS_INLINE static inline
-void xfer_error_isr(uint8_t rhport)
-{
-  //------------- async list -------------//
-  ehci_qhd_t * const async_head = qhd_async_head(rhport);
-  ehci_qhd_t *p_qhd = async_head;
-  do
-  {
-    hcd_dcache_invalidate(p_qhd, sizeof(ehci_qhd_t));
-    qhd_xfer_error_isr( p_qhd );
-    p_qhd = qhd_next(p_qhd);
-  }while(p_qhd != async_head); // async list traversal, stop if loop around
-
-  //------------- TODO refractor period list -------------//
-  uint32_t const period_1ms_addr = (uint32_t) get_period_head(rhport, 1u);
-  for (uint32_t interval_ms=1; interval_ms <= FRAMELIST_SIZE; interval_ms *= 2)
-  {
-    ehci_link_t next_item = * get_period_head(rhport, interval_ms);
-
-    // TODO abstract max loop guard for period
-    while( !next_item.terminate &&
-           !(interval_ms > 1 && period_1ms_addr == tu_align32(next_item.address)) )
-    {
-      switch ( next_item.type )
-      {
-        case EHCI_QTYPE_QHD:
-        {
-          ehci_qhd_t *p_qhd_int = (ehci_qhd_t *) tu_align32(next_item.address);
-          hcd_dcache_invalidate(p_qhd_int, sizeof(ehci_qhd_t));
-
-          qhd_xfer_error_isr(p_qhd_int);
-        }
-        break;
-
-				// TODO support hs/fs ISO
-        case EHCI_QTYPE_ITD:
-        case EHCI_QTYPE_SITD:
-        case EHCI_QTYPE_FSTN:
-        default: break;
-      }
-
-      next_item = *list_next(&next_item);
-    }
-  }
-}
-
 //------------- Host Controller Driver's Interrupt Handler -------------//
 void hcd_int_handler(uint8_t rhport)
 {
@@ -776,29 +672,16 @@ void hcd_int_handler(uint8_t rhport)
     regs->status = EHCI_INT_MASK_PORT_CHANGE; // Acknowledge
   }
 
-  if (int_status & EHCI_INT_MASK_ERROR) {
-    xfer_error_isr(rhport);
-    regs->status = EHCI_INT_MASK_ERROR; // Acknowledge
-  }
-
-  //------------- some QTD/SITD/ITD with IOC set is completed -------------//
-  if (int_status & EHCI_INT_MASK_NXP_ASYNC) {
-    async_list_xfer_complete_isr(qhd_async_head(rhport));
-    regs->status = EHCI_INT_MASK_NXP_ASYNC; // Acknowledge
-  }
+  // A USB transfer is completed (OK or error)
+  uint32_t const usb_int = int_status & (EHCI_INT_MASK_USB | EHCI_INT_MASK_ERROR);
+  if (usb_int) {
+    proccess_async_xfer_isr(list_get_async_head(rhport));
 
-  if (int_status & EHCI_INT_MASK_NXP_PERIODIC)
-  {
-    for (uint32_t i=1; i <= FRAMELIST_SIZE; i *= 2)
-    {
-      period_list_xfer_complete_isr(rhport, i);
+    for ( uint32_t i = 1; i <= FRAMELIST_SIZE; i *= 2 ) {
+      process_period_xfer_isr(rhport, i);
     }
-    regs->status = EHCI_INT_MASK_NXP_PERIODIC; // Acknowledge
-  }
 
-  if (int_status & EHCI_INT_MASK_USB) {
-    // TODO standard EHCI xfer complete
-    regs->status = EHCI_INT_MASK_USB; // Acknowledge
+    regs->status = usb_int; // Acknowledge
   }
 
   //------------- There is some removed async previously -------------//
@@ -810,35 +693,103 @@ void hcd_int_handler(uint8_t rhport)
 }
 
 //--------------------------------------------------------------------+
-// HELPER
+// List Managing Helper
 //--------------------------------------------------------------------+
 
+// Get head of periodic list
+TU_ATTR_ALWAYS_INLINE static inline ehci_link_t* list_get_period_head(uint8_t rhport, uint32_t interval_ms) {
+  (void) rhport;
+  return (ehci_link_t*) &ehci_data.period_head_arr[ tu_log2( tu_min32(FRAMELIST_SIZE, interval_ms) ) ];
+}
+
+// Get head of async list
+TU_ATTR_ALWAYS_INLINE static inline ehci_qhd_t* list_get_async_head(uint8_t rhport) {
+  (void) rhport;
+  return qhd_control(0); // control qhd of dev0 is used as async head
+}
+
+TU_ATTR_ALWAYS_INLINE static inline ehci_link_t* list_next(ehci_link_t const *p_link) {
+  return (ehci_link_t*) tu_align32(p_link->address);
+}
 
-//------------- queue head helper -------------//
-static inline ehci_qhd_t* qhd_find_free (void)
+TU_ATTR_ALWAYS_INLINE static inline void list_insert(ehci_link_t *current, ehci_link_t *new, uint8_t new_type)
 {
-  for (uint32_t i=0; i<QHD_MAX; i++)
-  {
-    if ( !ehci_data.qhd_pool[i].used ) return &ehci_data.qhd_pool[i];
+  new->address = current->address;
+  current->address = ((uint32_t) new) | (new_type << 1);
+}
+
+// Remove all queue head belong to this device address
+static void list_remove_qhd_by_daddr(ehci_link_t* list_head, uint8_t dev_addr) {
+  ehci_link_t* prev = list_head;
+
+  while (prev && !prev->terminate) {
+    ehci_qhd_t* qhd = (ehci_qhd_t*) (uintptr_t) list_next(prev);
+
+    // done if loop back to head
+    if ( (uintptr_t) qhd == (uintptr_t) list_head) {
+      break;
+    }
+
+    if ( qhd->dev_addr == dev_addr ) {
+      // TODO deactivate all TD, wait for QHD to inactive before removal
+      prev->address = qhd->next.address;
+
+      // EHCI 4.8.2 link the removed qhd's next to async head (which always reachable by Host Controller)
+      qhd->next.address = ((uint32_t) list_head) | (EHCI_QTYPE_QHD << 1);
+
+      if ( qhd->int_smask )
+      {
+        // period list queue element is guarantee to be free in the next frame (1 ms)
+        qhd->used = 0;
+      }else
+      {
+        // async list use async advance handshake
+        // mark as removing, will completely re-usable when async advance isr occurs
+        qhd->removing = 1;
+      }
+
+      hcd_dcache_clean(qhd, sizeof(ehci_qhd_t));
+      hcd_dcache_clean(prev, sizeof(ehci_qhd_t));
+    }else {
+      prev = list_next(prev);
+    }
   }
+}
 
+
+//--------------------------------------------------------------------+
+// Queue Header helper
+//--------------------------------------------------------------------+
+
+// Get queue head for control transfer (always available)
+TU_ATTR_ALWAYS_INLINE static inline ehci_qhd_t* qhd_control(uint8_t dev_addr) {
+  return &ehci_data.control[dev_addr].qhd;
+}
+
+// Find a free queue head
+TU_ATTR_ALWAYS_INLINE static inline ehci_qhd_t *qhd_find_free(void) {
+  for ( uint32_t i = 0; i < QHD_MAX; i++ ) {
+    if ( !ehci_data.qhd_pool[i].used ) return &ehci_data.qhd_pool[i];
+  }
   return NULL;
 }
 
-static inline ehci_qhd_t* qhd_next(ehci_qhd_t const * p_qhd)
-{
-  return (ehci_qhd_t*) tu_align32(p_qhd->next.address);
+// Next queue head link
+TU_ATTR_ALWAYS_INLINE static inline ehci_qhd_t *qhd_next(ehci_qhd_t const *p_qhd) {
+  return (ehci_qhd_t *) tu_align32(p_qhd->next.address);
 }
 
-static inline ehci_qhd_t* qhd_get_from_addr(uint8_t dev_addr, uint8_t ep_addr)
-{
-  ehci_qhd_t* qhd_pool = ehci_data.qhd_pool;
+// Get queue head from device + endpoint address
+static ehci_qhd_t *qhd_get_from_addr(uint8_t dev_addr, uint8_t ep_addr) {
+  if ( 0 == tu_edpt_number(ep_addr) ) {
+    return qhd_control(dev_addr);
+  }
 
-  for(uint32_t i=0; i<QHD_MAX; i++)
-  {
+  ehci_qhd_t *qhd_pool = ehci_data.qhd_pool;
+
+  for ( uint32_t i = 0; i < QHD_MAX; i++ ) {
     if ( (qhd_pool[i].dev_addr == dev_addr) &&
-          ep_addr == tu_edpt_addr(qhd_pool[i].ep_number, qhd_pool[i].pid) )
-    {
+         ep_addr == tu_edpt_addr(qhd_pool[i].ep_number, qhd_pool[i].pid) ) {
       return &qhd_pool[i];
     }
   }
@@ -846,6 +797,7 @@ static inline ehci_qhd_t* qhd_get_from_addr(uint8_t dev_addr, uint8_t ep_addr)
   return NULL;
 }
 
+// Init queue head with endpoint descriptor
 static void qhd_init(ehci_qhd_t *p_qhd, uint8_t dev_addr, tusb_desc_endpoint_t const * ep_desc)
 {
   // address 0 is used as async head, which always on the list --> cannot be cleared (ehci halted otherwise)
@@ -920,6 +872,7 @@ static void qhd_init(ehci_qhd_t *p_qhd, uint8_t dev_addr, tusb_desc_endpoint_t c
   }
 }
 
+// Attach a TD to queue head
 static void qhd_attach_qtd(ehci_qhd_t *qhd, ehci_qtd_t *qtd) {
   qhd->attached_qtd = qtd;
   qhd->attached_buffer = qtd->buffer[0];
@@ -931,17 +884,35 @@ static void qhd_attach_qtd(ehci_qhd_t *qhd, ehci_qtd_t *qtd) {
   hcd_dcache_clean_invalidate(qhd, sizeof(ehci_qhd_t));
 }
 
+// Remove an attached TD from queue head
+static void qhd_remove_qtd(ehci_qhd_t *qhd) {
+  ehci_qtd_t * volatile qtd = qhd->attached_qtd;
+
+  qhd->attached_qtd = NULL;
+  qhd->attached_buffer = 0;
+  hcd_dcache_clean(qhd, sizeof(ehci_qhd_t));
+
+  qtd->used = 0; // free QTD
+  hcd_dcache_clean(qtd, sizeof(ehci_qtd_t));
+}
+
+//--------------------------------------------------------------------+
+// Queue TD helper
+//--------------------------------------------------------------------+
+
+// Get TD for control transfer (always available)
+TU_ATTR_ALWAYS_INLINE static inline ehci_qtd_t* qtd_control(uint8_t dev_addr) {
+  return &ehci_data.control[dev_addr].qtd;
+}
 
-//------------- TD helper -------------//
-static inline ehci_qtd_t *qtd_find_free(void) {
+TU_ATTR_ALWAYS_INLINE static inline ehci_qtd_t *qtd_find_free(void) {
   for (uint32_t i = 0; i < QTD_MAX; i++) {
     if (!ehci_data.qtd_pool[i].used) return &ehci_data.qtd_pool[i];
   }
   return NULL;
 }
 
-static void qtd_init(ehci_qtd_t* qtd, void const* buffer, uint16_t total_bytes)
-{
+static void qtd_init(ehci_qtd_t* qtd, void const* buffer, uint16_t total_bytes) {
   tu_memclr(qtd, sizeof(ehci_qtd_t));
   qtd->used                = 1;
 
@@ -955,24 +926,9 @@ static void qtd_init(ehci_qtd_t* qtd, void const* buffer, uint16_t total_bytes)
   qtd->expected_bytes      = total_bytes;
 
   qtd->buffer[0] = (uint32_t) buffer;
-  for(uint8_t i=1; i<5; i++)
-  {
+  for(uint8_t i=1; i<5; i++) {
     qtd->buffer[i] |= tu_align4k(qtd->buffer[i - 1] ) + 4096;
   }
 }
 
-//------------- List Managing Helper -------------//
-
-// insert at head
-static inline void list_insert(ehci_link_t *current, ehci_link_t *new, uint8_t new_type)
-{
-  new->address = current->address;
-  current->address = ((uint32_t) new) | (new_type << 1);
-}
-
-static inline ehci_link_t* list_next(ehci_link_t const *p_link)
-{
-  return (ehci_link_t*) tu_align32(p_link->address);
-}
-
 #endif
diff --git a/src/portable/ehci/ehci.h b/src/portable/ehci/ehci.h
index 8338fb419..457adc1d3 100644
--- a/src/portable/ehci/ehci.h
+++ b/src/portable/ehci/ehci.h
@@ -278,23 +278,24 @@ enum {
   EHCI_INT_MASK_PERIODIC_SCHED_STATUS = TU_BIT(14),
   EHCI_INT_MASK_ASYNC_SCHED_STATUS    = TU_BIT(15),
 
-  EHCI_INT_MASK_NXP_ASYNC             = TU_BIT(18),
-  EHCI_INT_MASK_NXP_PERIODIC          = TU_BIT(19),
-
   EHCI_INT_MASK_ALL                   =
       EHCI_INT_MASK_USB | EHCI_INT_MASK_ERROR | EHCI_INT_MASK_PORT_CHANGE |
       EHCI_INT_MASK_FRAMELIST_ROLLOVER | EHCI_INT_MASK_PCI_HOST_SYSTEM_ERROR |
-      EHCI_INT_MASK_ASYNC_ADVANCE | EHCI_INT_MASK_NXP_SOF |
-      EHCI_INT_MASK_NXP_ASYNC | EHCI_INT_MASK_NXP_PERIODIC
+      EHCI_INT_MASK_ASYNC_ADVANCE | EHCI_INT_MASK_NXP_SOF
+};
+
+enum {
+  EHCI_USBCMD_FRAMELIST_SIZE_SHIFT              = 2, // [2..3]
+  EHCI_USBCMD_CHIPIDEA_FRAMELIST_SIZE_MSB_SHIFT = 15,
+  EHCI_USBCMD_INTERRUPT_THRESHOLD_SHIFT         = 16
 };
 
 enum {
-  EHCI_USBCMD_POS_RUN_STOP               = 0,
-  EHCI_USBCMD_POS_FRAMELIST_SIZE         = 2,
-  EHCI_USBCMD_POS_PERIOD_ENABLE          = 4,
-  EHCI_USBCMD_POS_ASYNC_ENABLE           = 5,
-  EHCI_USBCMD_POS_NXP_FRAMELIST_SIZE_MSB = 15,
-  EHCI_USBCMD_POS_INTERRUPT_THRESHOLD    = 16
+  EHCI_USBCMD_RUN_STOP                       = TU_BIT(0), // [0..0] 1 = Run, 0 = Stop
+  EHCI_USBCMD_HCRESET                        = TU_BIT(1), // [1..1] SW write 1 to reset HC, clear by HC when complete
+  EHCI_USBCMD_PERIOD_SCHEDULE_ENABLE         = TU_BIT(4), // [4..4] Enable periodic schedule
+  EHCI_USBCMD_ASYNC_SCHEDULE_ENABLE          = TU_BIT(5), // [5..5] Enable async schedule
+  EHCI_USBCMD_INTR_ON_ASYNC_ADVANCE_DOORBELL = TU_BIT(6), // [6..6] Tell HC to interrupt next time it advances async list. Clear by HC
 };
 
 enum {
@@ -306,7 +307,7 @@ enum {
   EHCI_PORTSC_MASK_FORCE_RESUME           = TU_BIT(6),
   EHCI_PORTSC_MASK_PORT_SUSPEND           = TU_BIT(7),
   EHCI_PORTSC_MASK_PORT_RESET             = TU_BIT(8),
-  ECHI_PORTSC_MASK_PORT_POWER             = TU_BIT(12),
+  EHCI_PORTSC_MASK_PORT_POWER             = TU_BIT(12),
 
   EHCI_PORTSC_MASK_W1C =
     EHCI_PORTSC_MASK_CONNECT_STATUS_CHANGE |
diff --git a/src/portable/mentor/musb/hcd_musb.c b/src/portable/mentor/musb/hcd_musb.c
index 9eb2e005e..5d1fba6d9 100644
--- a/src/portable/mentor/musb/hcd_musb.c
+++ b/src/portable/mentor/musb/hcd_musb.c
@@ -822,9 +822,17 @@ bool hcd_edpt_xfer(uint8_t rhport, uint8_t dev_addr, uint8_t ep_addr, uint8_t *b
   return ret;
 }
 
+bool hcd_edpt_abort_xfer(uint8_t rhport, uint8_t dev_addr, uint8_t ep_addr) {
+  (void) rhport;
+  (void) dev_addr;
+  (void) ep_addr;
+  // TODO not implemented yet
+  return false;
+}
+
 // clear stall, data toggle is also reset to DATA0
-bool hcd_edpt_clear_stall(uint8_t dev_addr, uint8_t ep_addr)
-{
+bool hcd_edpt_clear_stall(uint8_t rhport, uint8_t dev_addr, uint8_t ep_addr) {
+  (void) rhport;
   unsigned const pipenum = find_pipe(dev_addr, ep_addr);
   if (!pipenum) return false;
   hw_endpoint_t volatile *regs = edpt_regs(pipenum - 1);
diff --git a/src/portable/microchip/pic/dcd_pic.c b/src/portable/microchip/pic/dcd_pic.c
index 6986c8317..ccc27c3c9 100644
--- a/src/portable/microchip/pic/dcd_pic.c
+++ b/src/portable/microchip/pic/dcd_pic.c
@@ -189,7 +189,7 @@ typedef struct
 // INTERNAL OBJECT & FUNCTION DECLARATION
 //--------------------------------------------------------------------+
 // BDT(Buffer Descriptor Table) must be 256-byte aligned
-CFG_TUSB_MEM_SECTION TU_ATTR_ALIGNED(512) volatile static dcd_data_t _dcd;
+CFG_TUD_MEM_SECTION TU_ATTR_ALIGNED(512) volatile static dcd_data_t _dcd;
 
 #if TU_PIC_INT_SIZE == 4
 TU_VERIFY_STATIC( sizeof(_dcd.bdt) == 512, "size is not correct" );
diff --git a/src/portable/microchip/samx7x/dcd_samx7x.c b/src/portable/microchip/samx7x/dcd_samx7x.c
index 2bbb345d6..9586df84d 100644
--- a/src/portable/microchip/samx7x/dcd_samx7x.c
+++ b/src/portable/microchip/samx7x/dcd_samx7x.c
@@ -77,7 +77,7 @@ static tusb_speed_t get_speed(void);
 static void dcd_transmit_packet(xfer_ctl_t * xfer, uint8_t ep_ix);
 
 // DMA descriptors shouldn't be placed in ITCM !
-CFG_TUSB_MEM_SECTION static dma_desc_t dma_desc[6];
+CFG_TUD_MEM_SECTION static dma_desc_t dma_desc[6];
 
 static xfer_ctl_t xfer_status[EP_MAX];
 
diff --git a/src/portable/mindmotion/mm32/dcd_mm32f327x_otg.c b/src/portable/mindmotion/mm32/dcd_mm32f327x_otg.c
index 39b09db68..c3d0c7297 100644
--- a/src/portable/mindmotion/mm32/dcd_mm32f327x_otg.c
+++ b/src/portable/mindmotion/mm32/dcd_mm32f327x_otg.c
@@ -110,7 +110,7 @@ typedef struct
 // INTERNAL OBJECT & FUNCTION DECLARATION
 //--------------------------------------------------------------------+
 // BDT(Buffer Descriptor Table) must be 256-byte aligned
-CFG_TUSB_MEM_SECTION TU_ATTR_ALIGNED(512) static dcd_data_t _dcd;
+CFG_TUD_MEM_SECTION TU_ATTR_ALIGNED(512) static dcd_data_t _dcd;
 
 TU_VERIFY_STATIC( sizeof(_dcd.bdt) == 512, "size is not correct" );
 
diff --git a/src/portable/nxp/khci/dcd_khci.c b/src/portable/nxp/khci/dcd_khci.c
index 52f4145f2..5c65ea33d 100644
--- a/src/portable/nxp/khci/dcd_khci.c
+++ b/src/portable/nxp/khci/dcd_khci.c
@@ -114,7 +114,7 @@ typedef struct
 // INTERNAL OBJECT & FUNCTION DECLARATION
 //--------------------------------------------------------------------+
 // BDT(Buffer Descriptor Table) must be 256-byte aligned
-CFG_TUSB_MEM_SECTION TU_ATTR_ALIGNED(512) static dcd_data_t _dcd;
+CFG_TUD_MEM_SECTION TU_ATTR_ALIGNED(512) static dcd_data_t _dcd;
 
 TU_VERIFY_STATIC( sizeof(_dcd.bdt) == 512, "size is not correct" );
 
diff --git a/src/portable/nxp/khci/hcd_khci.c b/src/portable/nxp/khci/hcd_khci.c
index f6029ee0d..928564e26 100644
--- a/src/portable/nxp/khci/hcd_khci.c
+++ b/src/portable/nxp/khci/hcd_khci.c
@@ -562,8 +562,16 @@ bool hcd_edpt_xfer(uint8_t rhport, uint8_t dev_addr, uint8_t ep_addr, uint8_t *
   return true;
 }
 
-bool hcd_edpt_clear_stall(uint8_t dev_addr, uint8_t ep_addr)
-{
+bool hcd_edpt_abort_xfer(uint8_t rhport, uint8_t dev_addr, uint8_t ep_addr) {
+  (void) rhport;
+  (void) dev_addr;
+  (void) ep_addr;
+  // TODO not implemented yet
+  return false;
+}
+
+bool hcd_edpt_clear_stall(uint8_t rhport, uint8_t dev_addr, uint8_t ep_addr) {
+  (void) rhport;
   if (!tu_edpt_number(ep_addr)) return true;
   int num = find_pipe(dev_addr, ep_addr);
   if (num < 0) return false;
diff --git a/src/portable/nxp/lpc17_40/dcd_lpc17_40.c b/src/portable/nxp/lpc17_40/dcd_lpc17_40.c
index 86149afd8..b880c2870 100644
--- a/src/portable/nxp/lpc17_40/dcd_lpc17_40.c
+++ b/src/portable/nxp/lpc17_40/dcd_lpc17_40.c
@@ -92,7 +92,7 @@ typedef struct
 
 } dcd_data_t;
 
-CFG_TUSB_MEM_SECTION TU_ATTR_ALIGNED(128) static dcd_data_t _dcd;
+CFG_TUD_MEM_SECTION TU_ATTR_ALIGNED(128) static dcd_data_t _dcd;
 
 
 //--------------------------------------------------------------------+
diff --git a/src/portable/nxp/lpc_ip3511/dcd_lpc_ip3511.c b/src/portable/nxp/lpc_ip3511/dcd_lpc_ip3511.c
index 5368ef868..7ca698a93 100644
--- a/src/portable/nxp/lpc_ip3511/dcd_lpc_ip3511.c
+++ b/src/portable/nxp/lpc_ip3511/dcd_lpc_ip3511.c
@@ -176,11 +176,11 @@ typedef struct
 // EP list must be 256-byte aligned
 //    Some MCU controller may require this variable to be placed in specific SRAM region.
 //    For example: LPC55s69 port1 Highspeed must be USB_RAM (0x40100000)
-//    Use CFG_TUSB_MEM_SECTION to place it accordingly.
-CFG_TUSB_MEM_SECTION TU_ATTR_ALIGNED(256) static dcd_data_t _dcd;
+//    Use CFG_TUD_MEM_SECTION to place it accordingly.
+CFG_TUD_MEM_SECTION TU_ATTR_ALIGNED(256) static dcd_data_t _dcd;
 
 // Dummy buffer to fix ZLPs overwriting the buffer (probably an USB/DMA controller bug)
-CFG_TUSB_MEM_SECTION TU_ATTR_ALIGNED(64) static uint8_t dummy[8];
+CFG_TUD_MEM_SECTION TU_ATTR_ALIGNED(64) static uint8_t dummy[8];
 
 //--------------------------------------------------------------------+
 // Multiple Controllers
diff --git a/src/portable/ohci/ohci.c b/src/portable/ohci/ohci.c
index 3f702d22c..413e72037 100644
--- a/src/portable/ohci/ohci.c
+++ b/src/portable/ohci/ohci.c
@@ -341,19 +341,24 @@ static void ed_init(ohci_ed_t *p_ed, uint8_t dev_addr, uint16_t ep_size, uint8_t
   p_ed->is_interrupt_xfer = (xfer_type == TUSB_XFER_INTERRUPT ? 1 : 0);
 }
 
-static void gtd_init(ohci_gtd_t* p_td, uint8_t* data_ptr, uint16_t total_bytes)
-{
+static void gtd_init(ohci_gtd_t *p_td, uint8_t *data_ptr, uint16_t total_bytes) {
   tu_memclr(p_td, sizeof(ohci_gtd_t));
 
-  p_td->used                   = 1;
-  p_td->expected_bytes         = total_bytes;
+  p_td->used = 1;
+  p_td->expected_bytes = total_bytes;
+
+  p_td->buffer_rounding = 1; // less than queued length is not a error
+  p_td->delay_interrupt = OHCI_INT_ON_COMPLETE_NO;
+  p_td->condition_code = OHCI_CCODE_NOT_ACCESSED;
 
-  p_td->buffer_rounding        = 1; // less than queued length is not a error
-  p_td->delay_interrupt        = OHCI_INT_ON_COMPLETE_NO;
-  p_td->condition_code         = OHCI_CCODE_NOT_ACCESSED;
+  uint8_t *cbp = (uint8_t *) _phys_addr(data_ptr);
 
-  p_td->current_buffer_pointer = _phys_addr(data_ptr);
-  p_td->buffer_end             = total_bytes ? (_phys_addr(data_ptr + total_bytes - 1)) : (uint8_t *)p_td->current_buffer_pointer;
+  p_td->current_buffer_pointer = cbp;
+  if ( total_bytes ) {
+    p_td->buffer_end = _phys_addr(data_ptr + total_bytes - 1);
+  } else {
+    p_td->buffer_end = cbp;
+  }
 }
 
 static ohci_ed_t * ed_from_addr(uint8_t dev_addr, uint8_t ep_addr)
@@ -487,7 +492,7 @@ bool hcd_setup_send(uint8_t rhport, uint8_t dev_addr, uint8_t const setup_packet
   ohci_ed_t* ed   = &ohci_data.control[dev_addr].ed;
   ohci_gtd_t *qtd = &ohci_data.control[dev_addr].gtd;
 
-  gtd_init(qtd, (uint8_t*) setup_packet, 8);
+  gtd_init(qtd, (uint8_t*)(uintptr_t) setup_packet, 8);
   qtd->index           = dev_addr;
   qtd->pid             = PID_SETUP;
   qtd->data_toggle     = GTD_DT_DATA0;
@@ -543,8 +548,16 @@ bool hcd_edpt_xfer(uint8_t rhport, uint8_t dev_addr, uint8_t ep_addr, uint8_t *
   return true;
 }
 
-bool hcd_edpt_clear_stall(uint8_t dev_addr, uint8_t ep_addr)
-{
+bool hcd_edpt_abort_xfer(uint8_t rhport, uint8_t dev_addr, uint8_t ep_addr) {
+  (void) rhport;
+  (void) dev_addr;
+  (void) ep_addr;
+  // TODO not implemented yet
+  return false;
+}
+
+bool hcd_edpt_clear_stall(uint8_t rhport, uint8_t dev_addr, uint8_t ep_addr) {
+  (void) rhport;
   ohci_ed_t * const p_ed = ed_from_addr(dev_addr, ep_addr);
 
   p_ed->is_stalled = 0;
diff --git a/src/portable/raspberrypi/pio_usb/hcd_pio_usb.c b/src/portable/raspberrypi/pio_usb/hcd_pio_usb.c
index 264af2e7a..45a3e80ef 100644
--- a/src/portable/raspberrypi/pio_usb/hcd_pio_usb.c
+++ b/src/portable/raspberrypi/pio_usb/hcd_pio_usb.c
@@ -138,6 +138,11 @@ bool hcd_edpt_xfer(uint8_t rhport, uint8_t dev_addr, uint8_t ep_addr, uint8_t *
   return pio_usb_host_endpoint_transfer(pio_rhport, dev_addr, ep_addr, buffer, buflen);
 }
 
+bool hcd_edpt_abort_xfer(uint8_t rhport, uint8_t dev_addr, uint8_t ep_addr) {
+  uint8_t const pio_rhport = RHPORT_PIO(rhport);
+  return pio_usb_host_endpoint_abort_transfer(pio_rhport, dev_addr, ep_addr);
+}
+
 bool hcd_setup_send(uint8_t rhport, uint8_t dev_addr, uint8_t const setup_packet[8])
 {
   uint8_t const pio_rhport = RHPORT_PIO(rhport);
@@ -158,8 +163,8 @@ bool hcd_setup_send(uint8_t rhport, uint8_t dev_addr, uint8_t const setup_packet
 //    return busy;
 //}
 
-bool hcd_edpt_clear_stall(uint8_t dev_addr, uint8_t ep_addr)
-{
+bool hcd_edpt_clear_stall(uint8_t rhport, uint8_t dev_addr, uint8_t ep_addr) {
+  (void) rhport;
   (void) dev_addr;
   (void) ep_addr;
 
diff --git a/src/portable/raspberrypi/rp2040/hcd_rp2040.c b/src/portable/raspberrypi/rp2040/hcd_rp2040.c
index 02f9968a7..6db8e2a0a 100644
--- a/src/portable/raspberrypi/rp2040/hcd_rp2040.c
+++ b/src/portable/raspberrypi/rp2040/hcd_rp2040.c
@@ -576,6 +576,14 @@ bool hcd_edpt_xfer(uint8_t rhport, uint8_t dev_addr, uint8_t ep_addr, uint8_t *
   return true;
 }
 
+bool hcd_edpt_abort_xfer(uint8_t rhport, uint8_t dev_addr, uint8_t ep_addr) {
+  (void) rhport;
+  (void) dev_addr;
+  (void) ep_addr;
+  // TODO not implemented yet
+  return false;
+}
+
 bool hcd_setup_send(uint8_t rhport, uint8_t dev_addr, uint8_t const setup_packet[8])
 {
   (void) rhport;
@@ -617,8 +625,8 @@ bool hcd_setup_send(uint8_t rhport, uint8_t dev_addr, uint8_t const setup_packet
   return true;
 }
 
-bool hcd_edpt_clear_stall(uint8_t dev_addr, uint8_t ep_addr)
-{
+bool hcd_edpt_clear_stall(uint8_t rhport, uint8_t dev_addr, uint8_t ep_addr) {
+  (void) rhport;
   (void) dev_addr;
   (void) ep_addr;
 
diff --git a/src/portable/renesas/rusb2/hcd_rusb2.c b/src/portable/renesas/rusb2/hcd_rusb2.c
index a4b534a05..fef8bdf19 100644
--- a/src/portable/renesas/rusb2/hcd_rusb2.c
+++ b/src/portable/renesas/rusb2/hcd_rusb2.c
@@ -752,9 +752,15 @@ bool hcd_edpt_xfer(uint8_t rhport, uint8_t dev_addr, uint8_t ep_addr, uint8_t *b
   return r;
 }
 
-bool hcd_edpt_clear_stall(uint8_t dev_addr, uint8_t ep_addr)
-{
-  uint8_t rhport = 0; // FIXME change API
+bool hcd_edpt_abort_xfer(uint8_t rhport, uint8_t dev_addr, uint8_t ep_addr) {
+  (void) rhport;
+  (void) dev_addr;
+  (void) ep_addr;
+  // TODO not implemented yet
+  return false;
+}
+
+bool hcd_edpt_clear_stall(uint8_t rhport, uint8_t dev_addr, uint8_t ep_addr) {
   uint16_t volatile *ctr = addr_to_pipectr(rhport, dev_addr, ep_addr);
   TU_ASSERT(ctr);
 
diff --git a/src/tusb_option.h b/src/tusb_option.h
index f16d8a5d0..7a4cfe51c 100644
--- a/src/tusb_option.h
+++ b/src/tusb_option.h
@@ -295,15 +295,14 @@
   #define CFG_TUSB_DEBUG 0
 #endif
 
-// TODO MEM_SECTION can be different for host and device controller
-// should use CFG_TUD_MEM_SECTION, CFG_TUH_MEM_SECTION
+// Memory section for placing buffer used for usb transferring. If MEM_SECTION is different for
+// host and device use: CFG_TUD_MEM_SECTION, CFG_TUH_MEM_SECTION instead
 #ifndef CFG_TUSB_MEM_SECTION
   #define CFG_TUSB_MEM_SECTION
 #endif
 
-// alignment requirement of buffer used for endpoint transferring
-// TODO MEM_ALIGN can be different for host and device controller
-// should use CFG_TUD_MEM_ALIGN, CFG_TUH_MEM_ALIGN
+// Alignment requirement of buffer used for usb transferring. if MEM_ALIGN is different for
+// host and device controller use: CFG_TUD_MEM_ALIGN, CFG_TUH_MEM_ALIGN instead
 #ifndef CFG_TUSB_MEM_ALIGN
   #define CFG_TUSB_MEM_ALIGN      TU_ATTR_ALIGNED(4)
 #endif
@@ -321,24 +320,14 @@
 // Device Options (Default)
 //--------------------------------------------------------------------
 
-// Attribute to place data in accessible RAM for device controller
-// default to CFG_TUSB_MEM_SECTION for backward-compatible
+// Attribute to place data in accessible RAM for device controller (default: CFG_TUSB_MEM_SECTION)
 #ifndef CFG_TUD_MEM_SECTION
-  #ifdef CFG_TUSB_MEM_SECTION
-    #define CFG_TUD_MEM_SECTION   CFG_TUSB_MEM_SECTION
-  #else
-    #define CFG_TUD_MEM_SECTION
-  #endif
+  #define CFG_TUD_MEM_SECTION     CFG_TUSB_MEM_SECTION
 #endif
 
-// Attribute to align memory for device controller
-// default to CFG_TUSB_MEM_ALIGN for backward-compatible
+// Attribute to align memory for device controller (default: CFG_TUSB_MEM_ALIGN)
 #ifndef CFG_TUD_MEM_ALIGN
-  #ifdef CFG_TUSB_MEM_ALIGN
-    #define CFG_TUD_MEM_ALIGN   CFG_TUSB_MEM_ALIGN
-  #else
-    #define CFG_TUD_MEM_ALIGN   TU_ATTR_ALIGNED(4)
-  #endif
+  #define CFG_TUD_MEM_ALIGN       CFG_TUSB_MEM_ALIGN
 #endif
 
 #ifndef CFG_TUD_ENDPOINT0_SIZE
@@ -419,19 +408,14 @@
   #endif
 #endif // CFG_TUH_ENABLED
 
-// Attribute to place data in accessible RAM for host controller
-// default to CFG_TUSB_MEM_SECTION for backward-compatible
+// Attribute to place data in accessible RAM for host controller (default: CFG_TUSB_MEM_SECTION)
 #ifndef CFG_TUH_MEM_SECTION
-  #ifdef CFG_TUSB_MEM_SECTION
-    #define CFG_TUH_MEM_SECTION   CFG_TUSB_MEM_SECTION
-  #else
-    #define CFG_TUH_MEM_SECTION
-  #endif
+  #define CFG_TUH_MEM_SECTION   CFG_TUSB_MEM_SECTION
 #endif
 
 // Attribute to align memory for host controller
 #ifndef CFG_TUH_MEM_ALIGN
-  #define CFG_TUH_MEM_ALIGN   TU_ATTR_ALIGNED(4)
+  #define CFG_TUH_MEM_ALIGN     CFG_TUSB_MEM_ALIGN
 #endif
 
 //------------- CLASS -------------//