libusb-devel Mailing List for libusb

Have you tried running your code on a different platform?  I have issues
with my product on RPi4, but it uses Isochronous.

Have you tried analyzing the USB packets with Wireshark?  I know on both
Windows and macOS this can be done - I'd be surprised if Linux couldn't do
it too.

At the very least, those two things with give you some sanity.  :)

-Chris

On Wed., 29 Apr. 2020, 1:14 am Thomas Demeillers, <ata...@li...> wrote:

> Hello everyone,
>
> I am new to USB and I am working on a project in which I need to receive
> data from a PIC micro-controller to a Raspberry Pi 4 with USB. The code on
> the micro-controller is already done and cannot be modified. I can’t access
> the source, but I know the basics. The device provided me with two
> endpoints.
>
> I am supposed to receive USB binary frames of 512bit / 64bytes every 10th
> or 100th of a second.
>
> The goal is first to receive one frame and then to try with a higher
> number of frames.
>
> Everything goes well until the bulk transfer which returns no error. The
> received size is matching the predicted one, but the data buffer remains
> empty.
>
> I have been trying to bypass this bug for the past week.
>
> I have tried every modification I can think of, can somebody help me find
> mistakes in my code?
>
> Thank you in advance !
>
>
> Thomas
>
>
>
>
>
>
>  Here is the pastebin of my code in case file attachment doesn't work
> https://pastebin.com/xQdX1Xjj
>
> Configuration:
>
> Hardware: Raspberry Pi 4
>
> OS: Raspbian Lite kernel 4.19
>
> Libusb-1.0.23
>
>
>
>
> The output running the program is the following:
>
>
>
> The module has been detected! (id = 19780 = 4D44)
>
> Dev (bus 1, device 8): 4747 - 4D44
>
> Reset successful!
>
> Kernel driver already disconnected                        OK
>
> Success
>
> Data received:
>
> Lengh of data received: 64
>
> Value of the firsts characters received:  , , ,
>
>
>
>
>
> Here is the output from sudo lsusb -v -d 4747:4D44 where 4747:4D44 are the
> manufacturer and product id :
>
>
> Bus 001 Device 007: ID 4747:4d44
>
> Device Descriptor:
>
>   bLength                18
>
>   bDescriptorType         1
>
>   bcdUSB               2.00
>
>   bDeviceClass            0
>
>   bDeviceSubClass         0
>
>   bDeviceProtocol         0
>
>   bMaxPacketSize0         8
>
>   idVendor           0x4747
>
>   idProduct          0x4d44
>
>   bcdDevice            0.01
>
>   iManufacturer           1 TopSecretManufacturer
>
>   iProduct                2 TopSecretProduct
>
>   iSerial                 0
>
>   bNumConfigurations      1
>
>   Configuration Descriptor:
>
>     bLength                 9
>
>     bDescriptorType         2
>
>     wTotalLength       0x0020
>
>     bNumInterfaces          1
>
>     bConfigurationValue     1
>
>     iConfiguration          0
>
>     bmAttributes         0xc0
>
>       Self Powered
>
>     MaxPower              100mA
>
>     Interface Descriptor:
>
>       bLength                 9
>
>       bDescriptorType         4
>
>       bInterfaceNumber        0
>
>       bAlternateSetting       0
>
>       bNumEndpoints           2
>
>       bInterfaceClass       255 Vendor Specific Class
>
>       bInterfaceSubClass    255 Vendor Specific Subclass
>
>       bInterfaceProtocol    255 Vendor Specific Protocol
>
>       iInterface              0
>
>       Endpoint Descriptor:
>
>         bLength                 7
>
>         bDescriptorType         5
>
>         bEndpointAddress     0x01  EP 1 OUT
>
>         bmAttributes            2
>
>           Transfer Type            Bulk
>
>           Synch Type               None
>
>           Usage Type               Data
>
>         wMaxPacketSize     0x0040  1x 64 bytes
>
>         bInterval               1
>
>       Endpoint Descriptor:
>
>         bLength                 7
>
>         bDescriptorType         5
>
>         bEndpointAddress     0x81  EP 1 IN
>
>         bmAttributes            2
>
>           Transfer Type            Bulk
>
>           Synch Type               None
>
>           Usage Type               Data
>
>         wMaxPacketSize     0x0040  1x 64 bytes
>
>         bInterval               1
>
> Device Status:     0x0001
>
>   Self Powered
> _______________________________________________
> libusb-devel mailing list
> lib...@li...
> https://lists.sourceforge.net/lists/listinfo/libusb-devel
>

Hello everyone,

I am new to USB and I am working on a project in which I need to receive data from a PIC micro-controller to a Raspberry Pi 4 with USB. The code on the micro-controller is already done and cannot be modified. I can’t access the source, but I know the basics. The device provided me with two endpoints.

I am supposed to receive USB binary frames of 512bit / 64bytes every 10th or 100th of a second.

The goal is first to receive one frame and then to try with a higher number of frames.

Everything goes well until the bulk transfer which returns no error. The received size is matching the predicted one, but the data buffer remains empty.

I have been trying to bypass this bug for the past week.

I have tried every modification I can think of, can somebody help me find mistakes in my code?

Thank you in advance !


Thomas






 Here is the pastebin of my code in case file attachment doesn't work https://pastebin.com/xQdX1Xjj


Configuration:

Hardware: Raspberry Pi 4

OS: Raspbian Lite kernel 4.19

Libusb-1.0.23




The output running the program is the following:



The module has been detected! (id = 19780 = 4D44)

Dev (bus 1, device 8): 4747 - 4D44

Reset successful!

Kernel driver already disconnected                        OK

Success

Data received:

Lengh of data received: 64

Value of the firsts characters received:  , , ,





Here is the output from sudo lsusb -v -d 4747:4D44 where 4747:4D44 are the manufacturer and product id :


Bus 001 Device 007: ID 4747:4d44

Device Descriptor:

  bLength                18

  bDescriptorType         1

  bcdUSB               2.00

  bDeviceClass            0

  bDeviceSubClass         0

  bDeviceProtocol         0

  bMaxPacketSize0         8

  idVendor           0x4747

  idProduct          0x4d44

  bcdDevice            0.01

  iManufacturer           1 TopSecretManufacturer

  iProduct                2 TopSecretProduct

  iSerial                 0

  bNumConfigurations      1

  Configuration Descriptor:

    bLength                 9

    bDescriptorType         2

    wTotalLength       0x0020

    bNumInterfaces          1

    bConfigurationValue     1

    iConfiguration          0

    bmAttributes         0xc0

      Self Powered

    MaxPower              100mA

    Interface Descriptor:

      bLength                 9

      bDescriptorType         4

      bInterfaceNumber        0

      bAlternateSetting       0

      bNumEndpoints           2

      bInterfaceClass       255 Vendor Specific Class

      bInterfaceSubClass    255 Vendor Specific Subclass

      bInterfaceProtocol    255 Vendor Specific Protocol

      iInterface              0

      Endpoint Descriptor:

        bLength                 7

        bDescriptorType         5

        bEndpointAddress     0x01  EP 1 OUT

        bmAttributes            2

          Transfer Type            Bulk

          Synch Type               None

          Usage Type               Data

        wMaxPacketSize     0x0040  1x 64 bytes

        bInterval               1

      Endpoint Descriptor:

        bLength                 7

        bDescriptorType         5

        bEndpointAddress     0x81  EP 1 IN

        bmAttributes            2

          Transfer Type            Bulk

          Synch Type               None

          Usage Type               Data

        wMaxPacketSize     0x0040  1x 64 bytes

        bInterval               1

Device Status:     0x0001

  Self Powered

Praveen R wrote:
>
> Could you please let me know if suerspeed steams are supported on 
> windows and if its not Is it possible to add this support and is there 
> any plan to support it in the near future.


This doesn't help answer your question, but I'm curious to know why you 
would want this.  What's the use case?  SuperSpeed streams are a hack 
that was added for the USB Mass Storage folks, and it's really hard to 
think of any other good uses for it that can't be done a lot easier in 
another way.

-- 
Tim Roberts, ti...@pr...
Providenza & Boekelheide, Inc.

Hi Team,

Than you for the continuous support of Libusb library.

Could you please let me know if suerspeed steams are supported on windows and if its not Is it possible to add this support and is there any plan to support it in the near future.


Best Regards
Praveen

I'm struggling with libusb_get_device_list(NULL, &devs)which crashes mine
application.

So it's my question, is it possible to run this library on Android 9?

Please let me know.

Turned out to be all my fault.  Adding the serial number caused it to 
try and totally re-enumerate.  Without the serial number it was using 
an old, successful enumeration.  Fixed the extended descriptors stuff 
and it worked.
Thanks for the hints.

>Date: Fri, 17 Apr 2020 22:10:09 -0700
>From: Tim Roberts <ti...@pr...>
>To: Libusb Mailing List <lib...@li...>
>Subject: Re: [libusb] Serial Number Messes Up Enumeration
>Message-ID: <36D...@pr...>
>Content-Type: text/plain;       charset=utf-8
>
>On Apr 17, 2020, at 1:26 PM, Robert Ritchey 
><R.R...@Qu...> wrote:
> >
> > Hi, this is a bit off topic for this list but I was hoping 
> someone might have some idea about my problem.
> > I am using libusb with my device and its been working great but I 
> decided to add a serial number to my board now that I am getting 
> closer to production.  I am using the Microsoft compatible ID 
> descriptors to use WinUSB.
> > This has been working fine but when I added the serial number 
> string Windows goes looking for a driver and fails.  When I check 
> the registry it seems that Windows is no longer trying to match me to WinUSB.
> > I don't know why just adding the serial number string messes 
> things up.  Any thoughts?
>
>As I?m sure you know, that shouldn?t make any difference.  Have you 
>double-checked that you didn?t mess something up when you added the 
>string descriptor?  You?re sure it?s still responding to the 
>Microsoft extended descriptors?  Those are also string descriptors, 
>and could have interfered.
>?
>Tim Roberts, ti...@pr...
>Providenza & Boekelheide, Inc.

-Bob
--------------------------------------------------------------------
Robert Ritchey
Quest Engineering & Development, Inc.
Suite B-8
6125 South Ash Avenue
Tempe, AZ 85283
Tel: (480) 460-2652
E-Mail: R.R...@Sm...
E-Mail: R.R...@Qu...
E-Mail: R.R...@AC...
WWW: www.QuestEngDev.com/
WWW: www.Smart-Fly.com/
-------------------------------------------------------------------- 

On Apr 17, 2020, at 1:26 PM, Robert Ritchey <R.R...@Qu...> wrote:
> 
> Hi, this is a bit off topic for this list but I was hoping someone might have some idea about my problem.
> I am using libusb with my device and its been working great but I decided to add a serial number to my board now that I am getting closer to production.  I am using the Microsoft compatible ID descriptors to use WinUSB.
> This has been working fine but when I added the serial number string Windows goes looking for a driver and fails.  When I check the registry it seems that Windows is no longer trying to match me to WinUSB.
> I don't know why just adding the serial number string messes things up.  Any thoughts?

As I’m sure you know, that shouldn’t make any difference.  Have you double-checked that you didn’t mess something up when you added the string descriptor?  You’re sure it’s still responding to the Microsoft extended descriptors?  Those are also string descriptors, and could have interfered.
— 
Tim Roberts, ti...@pr...
Providenza & Boekelheide, Inc.

Hi, this is a bit off topic for this list but I was hoping someone 
might have some idea about my problem.
I am using libusb with my device and its been working great but I 
decided to add a serial number to my board now that I am getting 
closer to production.  I am using the Microsoft compatible ID 
descriptors to use WinUSB.
This has been working fine but when I added the serial number string 
Windows goes looking for a driver and fails.  When I check the 
registry it seems that Windows is no longer trying to match me to WinUSB.
I don't know why just adding the serial number string messes things 
up.  Any thoughts?
Thank you,

-Bob
--------------------------------------------------------------------
Robert Ritchey
Quest Engineering & Development, Inc.
Suite B-8
6125 South Ash Avenue
Tempe, AZ 85283
Tel: (480) 460-2652
E-Mail: R.R...@Sm...
E-Mail: R.R...@Qu...
E-Mail: R.R...@AC...
WWW: www.QuestEngDev.com/
WWW: www.Smart-Fly.com/
-------------------------------------------------------------------- 

On Apr 5, 2020, at 1:38 AM, li...@su... wrote:
> 
> Now I can get my touch screen report descriptor, but it doesn't work with usb mouse or usb keyboard when I  use "libusb_get_descriptor(handle,LIBUSB_DT_REPORT,0,buff,sizeof(buff))" with -9 (LIBUSB_ERROR_PIPE) returned (also, I don't know how to get the desc_index so I always set it 0). 
> 
> But I can get all the HID devices' HID descriptor in 'configuration descriptor -> interface configurator-> extra' (e.g. 09 21 1101 00 01 22 3b00 ,it should have  003b -> 59 byte of report descriptor )
> 
> What's the problem with them.Should I re-configure them so that the default pipe can be used?

As it turns out, you can’t use libusb_get_descriptor to fetch a report descriptor.  Report descriptors have to be fetched using  a GET_DESCRIPTOR request aimed at the interface, and libusb always sends descriptor requests aimed at the device.  This kind of thing should work:

libusb_control_transfer (handle,
                          LIBUSB_ENDPOINT_IN |
                            LIBUSB_REQUEST_TYPE_STANDARD |
                            LIBUSB_RECIPIENT_INTERFACE,
                          LIBUSB_REQUEST_GET_DESCRIPTOR,
                          (LIBUSB_DT_REPORT << 8),
                          0,
                          buf, 2014,
                          5000);

— 
Tim Roberts, ti...@pr...
Providenza & Boekelheide, Inc.

On Apr 5, 2020, at 1:38 AM, li...@su... wrote:
> 
> Now I can get my touch screen report descriptor, but it doesn't work with usb mouse or usb keyboard when I  use "libusb_get_descriptor(handle,LIBUSB_DT_REPORT,0,buff,sizeof(buff))" with -9 (LIBUSB_ERROR_PIPE) returned (also, I don't know how to get the desc_index so I always set it 0). 
> 
> But I can get all the HID devices' HID descriptor in 'configuration descriptor -> interface configurator-> extra' (e.g. 09 21 1101 00 01 22 3b00 ,it should have  003b -> 59 byte of report descriptor )
> 
> What's the problem with them.Should I re-configure them so that the default pipe can be used?

As it turns out, you can’t use libusb_get_descriptor to fetch a report descriptor.  Report descriptors have to be fetched using  a GET_DESCRIPTOR request aimed at the interface, and libusb always sends descriptor requests aimed at the device.  This kind of thing should work:

libusb_control_transfer (handle,
                          LIBUSB_ENDPOINT_IN |
                            LIBUSB_REQUEST_TYPE_STANDARD |
                            LIBUSB_RECIPIENT_INTERFACE,
                          LIBUSB_REQUEST_GET_DESCRIPTOR,
                          (LIBUSB_DT_REPORT << 8),
                          0,
                          buf, 2014,
                          5000);

— 
Tim Roberts, ti...@pr...
Providenza & Boekelheide, Inc.

On Apr 5, 2020, at 1:38 AM, li...@su... wrote:
> 
> Now I can get my touch screen report descriptor, but it doesn't work with usb mouse or usb keyboard when I  use "libusb_get_descriptor(handle,LIBUSB_DT_REPORT,0,buff,sizeof(buff))" with -9 (LIBUSB_ERROR_PIPE) returned (also, I don't know how to get the desc_index so I always set it 0). 
> 
> But I can get all the HID devices' HID descriptor in 'configuration descriptor -> interface configurator-> extra' (e.g. 09 21 1101 00 01 22 3b00 ,it should have  003b -> 59 byte of report descriptor )
> 
> What's the problem with them.Should I re-configure them so that the default pipe can be used?

As it turns out, you can’t use libusb_get_descriptor to fetch a report descriptor.  Report descriptors have to be fetched using  a GET_DESCRIPTOR request aimed at the interface, and libusb always sends descriptor requests aimed at the device.  This kind of thing should work:

libusb_control_transfer (handle,
                          LIBUSB_ENDPOINT_IN |
                            LIBUSB_REQUEST_TYPE_STANDARD |
                            LIBUSB_RECIPIENT_INTERFACE,
                          LIBUSB_REQUEST_GET_DESCRIPTOR,
                          (LIBUSB_DT_REPORT << 8),
                          0,
                          buf, 2014,
                          5000);

— 
Tim Roberts, ti...@pr...
Providenza & Boekelheide, Inc.

On Apr 5, 2020, at 2:00 AM, li...@su... wrote:
> 
> Must I use claim an interfece?

Descriptor access all happens over the control pipe, so you shouldn’t need to claim an interface.  Windows’ rules are slightly different, but I think you’re running on on Linux.
— 
Tim Roberts, ti...@pr...
Providenza & Boekelheide, Inc.

On Apr 5, 2020, at 2:00 AM, li...@su... wrote:
> 
> Must I use claim an interfece?

Descriptor access all happens over the control pipe, so you shouldn’t need to claim an interface.  Windows’ rules are slightly different, but I think you’re running on on Linux.
— 
Tim Roberts, ti...@pr...
Providenza & Boekelheide, Inc.

li...@su... wrote:
> Sorry I'm late.
> (I study it when I'm free)
> I just want to analyze usb devices(including HID device and so on) on 
> linux.
> e.g.
>     1) a mouse plugged in.
>     2) use libusb check out this plug-in event
>     3) analyze it by libusb -> it's HID
>     4) analyze it by hidapi -> it's a mouse
>
> If it doesn't get usage & usage page on linux,how can I 
> get hidraw_report_descriptor to analyze devices.
> 1. devices will not add the device file in /dev/, rigth? So I can't 
> use 'open()'.

Well, you can open the device in libusb, and once you have done that you 
can fetch the report descriptors yourself.  The HID descriptor (part of 
the configuration descriptor) tells you how many report descriptors 
there are, and libusb_get_descriptor can fetch the report descriptors.


> 2. I have read hidapi 
> on https://www.kernel.org/doc/Documentation/hid/hidraw.txt, it 
> says Hidraw supports the following ioctls. but I don't know what's the 
> target(  ioctl( ?, HIDIOCGREPORT, hidraw_report_descriptor&) ).

The HIDAPI library talks to the hidraw driver.  The driver handle is 
internal to the library.  A suitably motivated person could figure out 
pretty easily how to get the driver handle from the hid_device structure.

-- 
Tim Roberts, ti...@pr...
Providenza & Boekelheide, Inc.

li...@su... wrote:
> Sorry I'm late.
> (I study it when I'm free)
> I just want to analyze usb devices(including HID device and so on) on 
> linux.
> e.g.
>     1) a mouse plugged in.
>     2) use libusb check out this plug-in event
>     3) analyze it by libusb -> it's HID
>     4) analyze it by hidapi -> it's a mouse
>
> If it doesn't get usage & usage page on linux,how can I 
> get hidraw_report_descriptor to analyze devices.
> 1. devices will not add the device file in /dev/, rigth? So I can't 
> use 'open()'.

Well, you can open the device in libusb, and once you have done that you 
can fetch the report descriptors yourself.  The HID descriptor (part of 
the configuration descriptor) tells you how many report descriptors 
there are, and libusb_get_descriptor can fetch the report descriptors.


> 2. I have read hidapi 
> on https://www.kernel.org/doc/Documentation/hid/hidraw.txt, it 
> says Hidraw supports the following ioctls. but I don't know what's the 
> target(  ioctl( ?, HIDIOCGREPORT, hidraw_report_descriptor&) ).

The HIDAPI library talks to the hidraw driver.  The driver handle is 
internal to the library.  A suitably motivated person could figure out 
pretty easily how to get the driver handle from the hid_device structure.

-- 
Tim Roberts, ti...@pr...
Providenza & Boekelheide, Inc.

OT as this is not libusb but if you allow me that would be great:
 
My problem is that USB on PIC18F45K50 stops working on some chips after some time. The USB firmware is something I botched together closer to 10 years ago from various sources. The code does not look pretty at all but it has worked for me all that time so I have not bothered to do anything with having been otherwise engaged as they say.
 
But now I have all of a suddenly on some PIC chips the USB stops working in anything from minutes to tens of minutes.
 
With some chips the identical code runs as long as I have the patience,  I just completed a 7 day and night run without problems.

All the working chips have manufacturing code 1321VYE and all the failing chips have manufacturing code 1718D51.

So sounds like a hardware problem right. 

OTH I see no errata or reporting on problems that correspond to my problems so could I be the only people with bad chips?

Also the bootloader code (which uses USB HID and was not written by me) seems to work no problem as long as I test it.

So this suggest to me that the problem is with the man on the mirror and his code.

I have access to Beagle 480 analyzer and I can easily capture the situation when it fails.

Here is what the beagle shows when things are running fine:
# Level	Sp	Index	m:s.ms.us	Dur	Len	Err	Dev	Ep	Record	Summary
0	FS	536639	0:58.793.206	2.183 us					[1 SOF]	[Frame: 1980] 
0	FS	536640	0:58.793.221	50.166 us	64 B		6	2	IN txn	00 00 00 00 80 00 00 00 00 00 00 00 80 00 00 00 00 00 00 00 80 00 00 00…
0	FS	536644	0:58.794.206	2.083 us					[1 SOF]	[Frame: 1981] 
0	FS	536645	0:58.794.212	50.000 us	64 B		6	2	OUT txn	00 01 00 00 00 00 00 00 00 02 00 00 00 00 00 00 00 13 01 00 20 00 00 00…
0	FS	536649	0:58.794.263	50.083 us	64 B		6	2	IN txn	00 00 00 00 80 00 00 00 00 00 00 00 80 00 00 00 00 00 00 00 80 00 00 00…
0	FS	536653	0:58.795.206	2.083 us					[1 SOF]	[Frame: 1982] 
0	FS	536654	0:58.795.212	50.166 us	64 B		6	2	IN txn	00 00 00 00 80 00 00 00 00 00 00 00 80 00 00 00 00 00 00 00 80 00 00 00…
0	FS	536658	0:58.796.206	2.250 us					[1 SOF]	[Frame: 1983] 
0	FS	536659	0:58.796.211	50.166 us	64 B		6	2	IN txn	00 00 00 00 80 00 00 00 00 00 00 00 80 00 00 00 00 00 00 00 80 00 00 00…
0	FS	536663	0:58.797.206	2.083 us					[1 SOF]	[Frame: 1984] 
0	FS	536664	0:58.797.211	50.166 us	64 B		6	2	IN txn	00 00 00 00 80 00 00 00 00 00 00 00 80 00 00 00 00 00 00 00 80 00 00 00…
0	FS	536668	0:58.798.206	2.183 us					[1 SOF]	[Frame: 1985] 
0	FS	536669	0:58.798.211	50.083 us	64 B		6	2	IN txn	00 00 00 00 80 00 00 00 00 00 00 00 80 00 00 00 00 00 00 00 80 00 00 00…
0	FS	536673	0:58.799.206	2.183 us					[1 SOF]	[Frame: 1986] 
0	FS	536674	0:58.799.211	50.166 us	64 B		6	2	IN txn	00 00 00 00 80 00 00 00 00 00 00 00 80 00 00 00 00 00 00 00 80 00 00 00…
0	FS	536678	0:58.800.206	2.083 us					[1 SOF]	[Frame: 1987] 
0	FS	536679	0:58.800.211	50.166 us	64 B		6	2	IN txn	00 00 00 00 80 00 00 00 00 00 00 00 80 00 00 00 00 00 00 00 80 00 00 00…
0	FS	536683	0:58.801.207	2.183 us					[1 SOF]	[Frame: 1988] 

repeats unti it fails.

This is what it looks like when the communication stops fail:
# Level	Sp	Index	m:s.ms.us	Dur	Len	Err	Dev	Ep	Record	Summary
0	FS	592699	1:10.520.254	1.999.172.416 s			6	2	[2000 IN-NAK]	[Periodic Timeout]
0	FS	592700	1:10.521.214	1.999.206.916 s			6	2	[2000 OUT-DATA-NAK]	[Periodic Timeout]
0	FS	600701	1:12.520.366	1.999.171.083 s					[2000 SOF]	[Frames: 1370 - 1321] [Periodic Timeout]
0	FS	600702	1:12.520.422	1.999.183.750 s			6	2	[2000 IN-NAK]	[Periodic Timeout]
0	FS	600703	1:12.521.371	1.999.219.750 s			6	2	[2000 OUT-DATA-NAK]	[Periodic Timeout]
0	FS	608704	1:14.520.535	1.999.171.083 s					[2000 SOF]	[Frames: 1322 - 1273] [Periodic Timeout]
0	FS	608705	1:14.520.592	1.999.172.250 s			6	2	[2000 IN-NAK]	[Periodic Timeout]
0	FS	608706	1:14.521.552	1.999.206.916 s			6	2	[2000 OUT-DATA-NAK]	[Periodic Timeout]

Now my question:

AFAIU what happens is that the host tries to do alternatively an IN and OUT transaction which are both NAKed by the SIE in the PIC.

Is that correct interpretation?

What is the periodic timeout?

Ok, so assuming my interpretation is right, there is very little software interaction involved here on the device side:
the SIE simply NAKs if the firmware has not processed  the data received (OUT dir) or prepared more data to send (IN dir).

The way this happens in PIC is that the firmware does what it does with USB memory and signals the SIE that the
the data in the buffer has been processed or prepared by setting the UOWN bit in the STAT register of the EP in question.

So basically above Beagle log tells me that the firmware has not set UOWN for neither in or out EP, right?

My main loop (the data processing happens at main program, not in USB interrupt) looks like this:

	while (1) {
		if (!(ep2_o.STAT & UOWN)) {  // new data from host, so process it
			<snip>
			ep2_o.CNT = 64;
			if (ep2_o.STAT & DTS) // toggle DATA0/DATA1
				ep2_o.STAT = DTSEN;
			else
				ep2_o.STAT = DTS | DTSEN;
			ep2_o.STAT |= UOWN;
			}
		if (!(ep2_i.STAT & UOWN)) {	// we own the USB buffer, so update data going to the host
			<snip>
			// turn the buffer over to the SIE so the host will pick it up
			ep2_i.CNT = 64;
			if (ep2_i.STAT & DTS) // toggle DATA0/DATA1 
				ep2_i.STAT = DTSEN;
			else
				ep2_i.STAT = DTS | DTSEN;
			ep2_i.STAT |= UOWN;
			}
		blink_led();
		}

From the LED blinking I know this loops gets executed so it looks like the SIE never clears the UOWN bit
because if it would then my firmware would set it in short order.

Does that look like a sensical description of the situation?

Any idea how to debug this further?

I've shown the main loop, the USB interrupt of course has a lot of code but AFAIU none of
that should get involved or do nothing unless there is some setup packet on the USB bus
and I see no evidence of that on Beagle log.

So the actual code involved is just what I've show above, right?

For completeness: this is a HID device with two 64 byte endpoints and the higher level protocol
is simply asking the host USB stack to write and read 64 bytes at a time.

wbr Kusti

li...@su... wrote:
> Sorry I'm late.
> (I study it when I'm free)
> I just want to analyze usb devices(including HID device and so on) on 
> linux.
> e.g.
>     1) a mouse plugged in.
>     2) use libusb check out this plug-in event
>     3) analyze it by libusb -> it's HID
>     4) analyze it by hidapi -> it's a mouse
>
> If it doesn't get usage & usage page on linux,how can I 
> get hidraw_report_descriptor to analyze devices.
> 1. devices will not add the device file in /dev/, rigth? So I can't 
> use 'open()'.

Well, you can open the device in libusb, and once you have done that you 
can fetch the report descriptors yourself.  The HID descriptor (part of 
the configuration descriptor) tells you how many report descriptors 
there are, and libusb_get_descriptor can fetch the report descriptors.


> 2. I have read hidapi 
> on https://www.kernel.org/doc/Documentation/hid/hidraw.txt, it 
> says Hidraw supports the following ioctls. but I don't know what's the 
> target(  ioctl( ?, HIDIOCGREPORT, hidraw_report_descriptor&) ).

The HIDAPI library talks to the hidraw driver.  The driver handle is 
internal to the library.  A suitably motivated person could figure out 
pretty easily how to get the driver handle from the hid_device structure.

-- 
Tim Roberts, ti...@pr...
Providenza & Boekelheide, Inc.

On Mar 31, 2020, at 11:07 PM, li...@su... wrote:
> 
> Now, I make a simple test to use it, but I always get 0 when I plug in a mouse or keyboard and get their 'usage' nad 'usage page'
> I'm using ubuntu 14.04 (linux) and imx6 device(4.1.15).
> ...
> Here is output of a mouse device on linux:
> It's a hid device

Well, sort of.  It’s fake device created by VMWare.


> what's the problem? 

Did you read the documentation, where it says that “usage_page” and  “usage” are available on Windows and Mac only?


> OTOH, I can't get list by hid_enumerate() on ARM but it return success when hid_init(). Also I found the program linking to 'libhidapi-hidraw.so' not 'libhidapi-libusb.so' on ARM, why and what's the 'libhidapi-hidraw.so’?

Google. is much faster than this mailing list.  “libhidapi-libusb” accesses the device using the libusb driver.  “libhidapi-hidraw” accesses the device using the Linux raw HID device.


> BTW, could somebody give me Definition file of Miscellaneous, because it doesn't in libusb but I also have a wifi & BT module. The class is 0xEF (USB-IF : https://www.usb.org/defined-class-codes#anchor_BaseClassEFh). <https://www.usb.org/defined-class-codes#anchor_BaseClassEFh).> 
> I hope you also can give me a table as it says "Wire Adapter Multifunction Peripheral programming interface. Definition can be found in the Wireless USB Specification in Chapter 8." I can't find this file.

I’m really wondering how hard you looked.  I Googled for “wireless USB specification”, and it was the second hit.  That’s a device for crossing between wired USB and wireless USB.

What on earth are you trying to do?  Because what you have looks like a very large waste of time.
— 
Tim Roberts, ti...@pr...
Providenza & Boekelheide, Inc.

Hi everyone

A few days ago, I initiated a discussion about how to use hidapi.

Now, I make a simple test to use it, but I always get 0 when I plug in a mouse or keyboard and get their 'usage' nad 'usage page'
I'm using ubuntu 14.04 (linux) and imx6 device(4.1.15).

Here is output of a mouse device on linux:
It's a hid device
libusb: VID: 3599 PID: 3
device: PID:3 VID:3599 usage:0 usage page:0
release: 258 interface: 0 path: 0002:0002:00
device: PID:3 VID:3599 usage:0 usage page:0
release: 258 interface: 1 path: 0002:0002:01
Manufacturer: VMware

Product: VMware Virtual USB Mouse

what's the problem? 

OTOH, I can't get list by hid_enumerate() on ARM but it return success when hid_init(). Also I found the program linking to 'libhidapi-hidraw.so' not 'libhidapi-libusb.so' on ARM, why and what's the 'libhidapi-hidraw.so'?

BTW, could somebody give me Definition file of Miscellaneous, because it doesn't in libusb but I also have a wifi & BT module. The class is 0xEF (USB-IF : https://www.usb.org/defined-class-codes#anchor_BaseClassEFh). 
I hope you also can give me a table as it says "Wire Adapter Multifunction Peripheral programming interface. Definition can be found in the Wireless USB Specification in Chapter 8." I can't find this file.

PS:Here is my test program of libusb and hidapi
Regards
Mihan

On Mar 31, 2020, at 6:13 AM, Jim Flanagan <jrf...@fr...> wrote:
> 
> So, either we set up a Win 7 system for calibrating, etc. or I thought I would write my own calibration routine for Win 10.  In order to initiate a communication link to the device, I was going to use a filter-driver.  At the moment I am using libusb-win32 under Win 10 setup as a filter and it works just fine.  Since libusb-win32 is a bit old, I thought it might be better to use Libusb-1.0 but I need the filtering capability.  Thus my questioning.  If you have any suggestions, I would appreciate any advice.

It’s hard to argue with success.  If libusb_win32 works, why would you switch away from it?
— 
Tim Roberts, ti...@pr...
Providenza & Boekelheide, Inc.