FAQs
cancel
Turn on suggestions
Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.
Showing results for 
Search instead for 
Did you mean: 

Known bugs in the OTG / USB library v2.1.0

ivous2
Associate

‎2014-01-21 05:04 AM

Posted on January 21, 2014 at 14:04

Here is the list of known bugs in the USB / OTG library v2.1.0:

files - usbd_cdc_core.c / usbd_conf.h

ep descriptors should not use 

#ifdef USE_USB_OTG_HS

but use instead

#ifndef USE_EMBEDDED_PHY 

file - usb_bsp.c

function USB_OTG_BSP_Init 

For OTH-HS using Internal Phy

RCC_AHB1PeriphClockLPModeCmd( RCC_AHB1Periph_OTG_HS_ULPI, DISABLE);  

This problem was discovered some time ago, when OTG-HS is in use and device is put in sleep mode, the ULPI clocks must be switched off, otherwise the OTG functionality is broken.

Whenever GPIO_InitStructure is used, it must be properly initialized first (otherwise it is causing strange side effects):

GPIO_StructInit(&GPIO_InitStructure);

file - usb_dcd_int.c

Search for ''emptyintr'' occurence, where the source code is trying to clear this flag. There are two occurrences:

CLEAR_IN_EP_INTR(1, emptyintr);

CLEAR_IN_EP_INTR( epnum , emptyintr);

It is read only flag, which cannot be cleared, so this interrupt must be masked instead:

fifoemptymsk = 0x1 << epnum;

USB_OTG_MODIFY_REG32(&pdev->regs.DREGS->DIEPEMPMSK, fifoemptymsk, 0)

This problem is causing high CPU load, but the USB is working.

file - usbd_core.c

USBD_Init function

correct order:  

/* Force Device Mode*/

USB_OTG_SetCurrentMode(pdev, DEVICE_MODE);

/* Enable Interrupts */

USB_OTG_BSP_EnableInterrupt(pdev);

On some hosts this is blocking proper enumeration of our device.

file usbh_core.c

missing proper error handling 

USBH_HandleControl function

case CTRL_STATUS_IN_WAIT:

   

    URB_Status = HCD_GetURB_State(pdev , phost->Control.hc_num_in);

    

    if  ( URB_Status == URB_DONE)

    { /* Control transfers completed, Exit the State Machine */

      phost->gState =   phost->gStateBkp;

      phost->Control.state = CTRL_COMPLETE;

    }

   

    else if (URB_Status == URB_ERROR)

    {

      phost->Control.state = CTRL_ERROR; 

    }

    

    else if((HCD_GetCurrentFrame(pdev)\

      - phost->Control.timer) > timeout)

    {

      phost->Control.state = CTRL_ERROR;

    }

     else if(URB_Status == URB_STALL)

    {

      /* Control transfers completed, Exit the State Machine */

      phost->gState =   phost->gStateBkp;

      phost->Control.state = CTRL_STALL; //added

 

      status = USBH_NOT_SUPPORTED; //added

    }

    break;

0 Kudos
13 REPLIES 13
waclawek.jan
Guru

‎2014-01-21 11:59 PM

Posted on January 22, 2014 at 08:59

Thanks.

Couldn't you get this post get sticky (to appear always at the top of the list)?

JW

0 Kudos
garberman
Associate

‎2014-04-02 03:56 AM

Posted on April 02, 2014 at 12:56

Hi,

Can I receive USB OTG FS WakeUp interrupt under Standby Mode for the processor STM32F437?

According to the datasheet the answer is ''Yes'' but I could not practical!

Can you help me by practical advice?

0 Kudos
tsuneo
Senior

‎2014-04-02 10:25 AM

Posted on April 02, 2014 at 19:25

> garberman.sergey:

> Can I receive USB OTG FS WakeUp interrupt under Standby Mode for the processor STM32F437? USB Wakeup event is passed to NVIC over EXTI 2.5 External interrupt/event line mapping (p380, DM00031pdf) EXTI line 18 is connected to the USB OTG FS Wakeup event - NVIC position:42 OTG_FS_WKUP EXTI line 20 is connected to the USB OTG HS (configured in FS) Wakeup event - NVIC position:76 OTG_HS_WKUP STM32_USB-Host-Device_Lib_V2.1.0 has already implemented these events. a) initialization 

\STM32_USB-Host-Device_Lib_V2.1.0\Project\USB_Device_Examples\HID\inc\usb_conf.h
/****************** USB OTG FS CONFIGURATION **********************************/
#ifdef USB_OTG_FS_CORE
...
#define USB_OTG_FS_LOW_PWR_MGMT_SUPPORT
/****************** USB OTG HS CONFIGURATION **********************************/
#ifdef USB_OTG_HS_CORE
...
#define USB_OTG_HS_LOW_PWR_MGMT_SUPPORT
#endif

\STM32_USB-Host-Device_Lib_V2.1.0\Project\USB_Device_Examples\HID\src\usb_bsp.c
void USB_OTG_BSP_Init(USB_OTG_CORE_HANDLE *pdev)
{
...
#ifdef USB_OTG_FS_LOW_PWR_MGMT_SUPPORT
EXTI_ClearITPendingBit(EXTI_Line18);
EXTI_InitStructure.EXTI_Line = EXTI_Line18; 
EXTI_InitStructure.EXTI_Mode = EXTI_Mode_Interrupt;
EXTI_InitStructure.EXTI_Trigger = EXTI_Trigger_Rising;
EXTI_InitStructure.EXTI_LineCmd = ENABLE;
EXTI_Init(&EXTI_InitStructure);
EXTI_ClearITPendingBit(EXTI_Line18); 
NVIC_InitStructure.NVIC_IRQChannel = OTG_FS_WKUP_IRQn;
NVIC_InitStructure.NVIC_IRQChannelPreemptionPriority = 1;
NVIC_InitStructure.NVIC_IRQChannelSubPriority = 5;
NVIC_InitStructure.NVIC_IRQChannelCmd = ENABLE;
NVIC_Init(&NVIC_InitStructure);
EXTI_ClearITPendingBit(EXTI_Line18); 
#endif 
#ifdef USB_OTG_HS_LOW_PWR_MGMT_SUPPORT
EXTI_ClearITPendingBit(EXTI_Line20);
EXTI_InitStructure.EXTI_Line = EXTI_Line20; 
EXTI_InitStructure.EXTI_Mode = EXTI_Mode_Interrupt;
EXTI_InitStructure.EXTI_Trigger = EXTI_Trigger_Rising;
EXTI_InitStructure.EXTI_LineCmd = ENABLE;
EXTI_Init(&EXTI_InitStructure);
EXTI_ClearITPendingBit(EXTI_Line20); 
NVIC_InitStructure.NVIC_IRQChannel = OTG_HS_WKUP_IRQn;
NVIC_InitStructure.NVIC_IRQChannelPreemptionPriority = 1;
NVIC_InitStructure.NVIC_IRQChannelSubPriority = 5;
NVIC_InitStructure.NVIC_IRQChannelCmd = ENABLE;
NVIC_Init(&NVIC_InitStructure);
EXTI_ClearITPendingBit(EXTI_Line20); 
#endif

b) ISRs 

\STM32_USB-Host-Device_Lib_V2.1.0\Project\USB_Device_Examples\HID\src\stm32fxxx_it.c
#ifdef USE_USB_OTG_FS 
void OTG_FS_WKUP_IRQHandler(void)
{
...
}
#ifdef USE_USB_OTG_HS 
void OTG_HS_WKUP_IRQHandler(void)
{
...
]

Tsuneo
0 Kudos
garberman
Associate

‎2014-04-03 12:53 AM

Posted on April 03, 2014 at 09:53

Many thanks, Tsuneo!

But I had been tried this sample and didn't get results!

Of course, I implemented it into my project... which not trivial.

May be, you know (link) a simple project for example?

0 Kudos
chmorgan
Associate II

‎2014-04-30 03:45 AM

Posted on April 30, 2014 at 12:45 

--- a/Libraries/STM32_USB_Device_Library/Class/cdc/src/usbd_cdc_core.c
+++ b/Libraries/STM32_USB_Device_Library/Class/cdc/src/usbd_cdc_core.c
@@ -625,7 +625,7 @@ static uint8_t usbd_cdc_Setup (void *pdev,
else /* No Data request */
{
/* Transfer the command to the interface layer */
- APP_FOPS.pIf_Ctrl(req->bRequest, NULL, 0);
+ APP_FOPS.pIf_Ctrl(req->bRequest, (uint8_t*)req, 0);
}

Here is another library fix I picked up from the forums, '' Modify usbd_cdc_Setup() to pass the request buffer into the APP_FOPS.pIf_Ctrl() function pointer callback. Change is per instructions by chinzei.tsuneo on the st forums to enable passing the ''No data request'' setup packets through to the callback function for processing. ''
0 Kudos
chmorgan
Associate II

‎2014-04-30 03:55 AM

Posted on April 30, 2014 at 12:55

Is it within the license for the library to be able to push it to some place like github? It seems a waste for everyone to be applying the same set of fixes to the library over and over.

0 Kudos
chmorgan
Associate II

‎2014-04-30 05:45 AM

Posted on April 30, 2014 at 14:45

file - usb_dcd_int.c

Search for ''emptyintr'' occurence, where the source code is trying to clear this flag. There are two occurrences:

CLEAR_IN_EP_INTR(1, emptyintr);

CLEAR_IN_EP_INTR( epnum , emptyintr);

It is read only flag, which cannot be cleared, so this interrupt must be masked instead:

fifoemptymsk = 0x1 << epnum;

USB_OTG_MODIFY_REG32(&pdev->regs.DREGS->DIEPEMPMSK, fifoemptymsk, 0)

This problem is causing high CPU load, but the USB is working.

Re: this fix, I've applied this change to DCD_WriteEmptyTxFifo() that I found from another forum discussion. This function appears to be called at all locations where the CLEAR_*** calls are made so it looked like a better fix: 

/**

* @brief DCD_WriteEmptyTxFifo

* check FIFO for the next packet to be loaded

* @param pdev: device instance

* @retval status

*/

static uint32_t DCD_WriteEmptyTxFifo(USB_OTG_CORE_HANDLE *pdev, uint32_t epnum)

{

USB_OTG_DTXFSTSn_TypeDef txstatus;

USB_OTG_EP *ep;

uint32_t len = 0;

uint32_t len32b;

txstatus.d32 = 0;

ep = &pdev->dev.in_ep[epnum]; 

len = ep->xfer_len - ep->xfer_count;

if (len > ep->maxpacket)

{

len = ep->maxpacket;

}

len32b = (len + 3) / 4;

txstatus.d32 = USB_OTG_READ_REG32( &pdev->regs.INEP_REGS[epnum]->DTXFSTS);

while (txstatus.b.txfspcavail > len32b &&

ep->xfer_count < 
ep-
>xfer_len &&

ep->xfer_len != 0)

{

/* Write the FIFO */

len = ep->xfer_len - ep->xfer_count;

if (len > ep->maxpacket)

{

len = ep->maxpacket;

}

len32b = (len + 3) / 4;

USB_OTG_WritePacket (pdev , ep->xfer_buff, epnum, len);

ep->xfer_buff += len;

ep->xfer_count += len;

#if 1

// NOTE: Fix from [DEAD LINK /public/STe2ecommunities/mcu/Lists/cortex_mx_stm32/Flat.aspx?RootFolder=/public/STe2ecommunities/mcu/Lists/cortex_mx_stm32/Flooded%20empty%20FIFO%20interrupts%20still%20make%20USB%20devices%20crawl&currentviews=413]https://my.st.com/public/STe2ecommunities/mcu/Lists/cortex_mx_stm32/Flat.aspx?RootFolder=/public/STe2ecommunities/mcu/Lists/cortex_mx_stm32/Flooded%20empty%20FIFO%20interrupts%20still%20make%20USB%20devices%20crawl¤tviews=413

if( ep->xfer_count >= ep->xfer_len){

uint32_t fifoemptymsk = 1 << 
epnum
;

USB_OTG_MODIFY_REG32(&pdev->regs.DREGS->DIEPEMPMSK, fifoemptymsk, 0);

break;

}

#endif

txstatus.d32 = USB_OTG_READ_REG32(&pdev->regs.INEP_REGS[epnum]->DTXFSTS);

}

return 1;

}

0 Kudos
chmorgan
Associate II

‎2014-04-30 06:02 AM

Posted on April 30, 2014 at 15:02

file - usbd_core.c

USBD_Init function

correct order:

/* Force Device Mode*/

USB_OTG_SetCurrentMode(pdev, DEVICE_MODE); /* Enable Interrupts */ USB_OTG_BSP_EnableInterrupt(pdev);

On some hosts this is blocking proper enumeration of our device.

I'm not seeing the call to USB_OTG_SetCurrentMode() in my version of USBD_Init(): 

/**

* @brief USBD_Init

* Initailizes the device stack and load the class driver

* @param pdev: device instance

* @param core_address: USB OTG core ID

* @param class_cb: Class callback structure address

* @param usr_cb: User callback structure address

* @retval None

*/

void USBD_Init(USB_OTG_CORE_HANDLE *pdev,

USB_OTG_CORE_ID_TypeDef coreID,

USBD_DEVICE *pDevice, 

USBD_Class_cb_TypeDef *class_cb, 

USBD_Usr_cb_TypeDef *usr_cb)

{

/* Hardware Init */

USB_OTG_BSP_Init(pdev); 

USBD_DeInit(pdev);

/*Register class and user callbacks */

pdev->dev.class_cb = class_cb;

pdev->dev.usr_cb = usr_cb; 

pdev->dev.usr_device = pDevice; 

/* set USB OTG core params */

DCD_Init(pdev , coreID);

/* Upon Init call usr callback */

pdev->dev.usr_cb->Init();

/* Enable Interrupts */

USB_OTG_BSP_EnableInterrupt(pdev);

}

Is this fix suggesting that a call to USB_OTG_SetCurrentMode() be added prior to enabling interrupts?
0 Kudos
Top