stm32l4 usb device debug

Arjun Hary

unread,

Aug 9, 2018, 8:59:39 PM8/9/18

to nu...@googlegroups.com

Hi,

I am working on the nucleo-ll452re board. I am trying to get USB device with mass storage on this board. I did the following.

1) I did a register comparison between the stm32l452 and stm32f303 which are usb device chips. The registers are exactly the same between them with addition of battery charging registers on the stm32l4.

2) I ported the stm_usbdev.c from stm32 folder to the stm32l4 and fixed the interrupts and code and all the compile issues.

3) I compiled the USB msc driver with debug on nuttx and ran msconn and msdis. I am attaching the output from nsh. I don't see any mass storage device appear on the windows PC.

sh> mkrd -m 0 -s 512 64

nsh> mkfatfs /dev/ram0

nsh> msconn

mcsonn_main: Creating block drivers

After boardctl(BOARDIOC_USBDEV_CONTROL):

total used free largest

Before: 137376 45184 92192 69632

After: 137376 45184 92192 69632

mcsonn_main: Configuring with NLUNS=1

mcsonn_main: handle=100027e0

After usbmsc_configure():

total used free largest

Before: 137376 45184 92192 69632

After: 137376 45472 91904 69632

Change: 288 allocated

mcsonn_main: Bind LUN=0 to /dev/ram0

After usbmsc_bindlun():

total used free largest

Before: 137376 45472 91904 69632

After: 137376 46000 91376 69632

Change: 528 allocated

After usbmsc_exportluns():

total used free largest

Before: 137376 46000 91376 69632

After: 137376 50160 87216 69632

Change: 4160 allocated

mcsonn_main: Connected

After MS connection:

total used free largest

Before: 137376 50160 87216 69632

After: 137376 50160 87216 69632

DCD register : Registered 0000

Class bind : Bind class 0000

nsh> msdis

Since MS connection:

total used free largest

Before: 137376 50160 87216 69632

After: 137376 50160 87216 69632

msdis: Disconnected

After usbmsc_disconnect():

total used free largest

Before: 137376 50160 87216 69632

After: 137376 45184 92192 69632

Change: 4976 freed

Final memory usage:

total used free largest

Before: 137376 45184 92192 69632

After: 137376 45184 92192 69632

Class error ae: USBMSC_TRACEERR_IDLERDREQLISTEMPTY 0000

DCD un-register : Un-registered 0000

Class disconnect : Disconnect class 0000

Class unbind : Un-bind class 0000

nsh>

Can you please point me to what is wrong or missing here and how to continue the debug.

--

Arjun Hary

________________________________________

: +1 (425) 381-7722

: zGlue Inc., Mountain View, CA

________________________________________

Alan Carvalho de Assis

unread,

Aug 10, 2018, 5:16:48 AM8/10/18

to nu...@googlegroups.com

Hi Arjun,

Did you see some activity from the host side?

What does dmesg show when you run msconn?

Try to enable USB debug on NuttX.

The Linux kernel also have support for USB monitoring. It could give you hints about what is failing.

BR,

Alan

--
You received this message because you are subscribed to the Google Groups "NuttX" group.
To unsubscribe from this group and stop receiving emails from it, send an email to nuttx+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

David Sidrane

unread,

Aug 10, 2018, 8:52:33 AM8/10/18

to NuttX

Arjun,

Check the polarity of the bit that activates the pull up. ST changed this in later versions of the USB controler.

See the delta in the stm32 USB devices driver for the STM32F446

David

--

You received this message because you are subscribed to the Google Groups "NuttX" group.

To unsubscribe from this group and stop receiving emails from it, send an email to nuttx+un...@googlegroups.com.

patacongo

unread,

Aug 10, 2018, 8:12:23 PM8/10/18

to NuttX

Can you please point me to what is wrong or missing here and how to continue the debug.

As Alan suggests, enabling USB debug might give some insight (but careful, USB debug output also often breaks the USB driver).

In my experience, anything but the most trivial USB bug is impossible to debug without a USB analyzer.

ar...@zglue.com

unread,

Aug 10, 2018, 8:46:42 PM8/10/18

to NuttX

Thanks guys for the inputs. I fixed a few more bugs. Got little further along. I am now seeing this message when i run msconn

EP allocreq : Allocate endpoint request 0000

DCD allocep : Allocate endpoint 0083

DCD allocep : Allocate endpoint 0002

EP allocreq : Allocate endpoint request 0002

EP allocreq : Allocate endpoint request 0003

DCD selfpowered : Self-powered 0001

DCD pullup : Soft connect 0001

I still do not see the device on PC. Next step is a linux kernel monitor or usb analyzer. But does this prints make sense. Sorry for these question but i am new to the nuttx USB driver.

Regards

Arjun

patacongo

unread,

Aug 10, 2018, 8:58:51 PM8/10/18

to NuttX

mcsonn_main: Connected
,,,

This is the device enumeration. The device connection was recognized by the host, the host sent some setup packets, the board provided its device and configuration descriptors. The MSC class has been started and it initializing its endpoints. All good.

DCD register : Registered 0000
Class bind : Bind class 0000
EP allocreq : Allocate endpoint request 0000
DCD allocep : Allocate endpoint 0083
DCD allocep : Allocate endpoint 0002
EP allocreq : Allocate endpoint request 0002
EP allocreq : Allocate endpoint request 0002
EP allocreq : Allocate endpoint request 0002
EP allocreq : Allocate endpoint request 0002
EP allocreq : Allocate endpoint request 0003
EP allocreq : Allocate endpoint request 0003
EP allocreq : Allocate endpoint request 0003
EP allocreq : Allocate endpoint request 0003

DCD selfpowered : Self-powered 0001

I don't understand this (not without looking at the code). The enumeration would not have occurred unless the pullup were in the right state The device should be invisible until it pulls up the data lines. The host side should pull the data lines down (weakly), and the device should pull them up. Then the host should recognize that something is connected and perform the enumeration.

DCD pullup : Soft connect 0001
DCD pullup : Soft connect 0001

I still do not see the device on PC. Next step is a linux kernel monitor or usb analyzer. But does this prints make sense. Sorry for these question but i am new to the nuttx USB driver.

Yes, something seems backward. The enumeration occurred BEFORE the pullup. The communications stopped after the pullup. Kind of seems reversed, right? You should re-read David Sidrane's comment about the polarity of the pull-up control.

David Sidrane

unread,

Aug 11, 2018, 6:56:27 AM8/11/18

to NuttX

Arjun,

If it is that same sort of issue we had on the new USB block (F466,F469, F7 etc) It can be seen in nuttx/arch/arm/src/stm32/stm32_otgfsdev.c

static void stm32_hwinitialize(FAR struct stm32_usbdev_s *priv) look for code wrapped in

#if defined(CONFIG_STM32_STM32F446) || defined(CONFIG_STM32_STM32F469)

Compare and contrast the bits, bit positions and their polarities in STM32_OTGFS_GCCFG and STM32_OTGFS_GOTGCTL to what you see in the manual for the SoC you are using.

David

ar...@zglue.com

unread,

Aug 13, 2018, 6:47:51 PM8/13/18

to NuttX

Turns out stupidity has no bounds. Thanks for the inputs David and Alan

1) I have the nucleo l452re board. I re-read the schematic and turns out the USB connector on that board is only connected to the stm32f013 or the st-link MCU. There is no USB connection to the stm32l452 chip. I ordered a breakout USB to test the USB connection

I will get the USB breakout try it and out and let you know how it goes.

@David: The stm32l452 is a USB device only chip. It has no OTG.

Regards

Arjun

Juha Niskanen (Haltian)

unread,

Mar 29, 2019, 4:37:34 AM3/29/19

to NuttX

Hi Arjun,

Did you get this working? Is your code available somewhere? Could you please send patches to Mr. Greg? I might start with USB FS for STM32L452 soon and it would be waste of effort to do the register definitions and comparisons between STM32 subarchitectures again.

Best Regards,

Juha Niskanen

From: nu...@googlegroups.com <nu...@googlegroups.com> on behalf of David Sidrane <david....@gmail.com>
Sent: Friday, August 10, 2018 3:52 PM
To: NuttX
Subject: Re: [nuttx] stm32l4 usb device debug

Reply all

Reply to author

Forward