USB Superspeed Peripherals

Dear sir,

Now I am trying the develop a product with CYUSB3014, and I am reading 2 documents to help me understand it.

<<001-76074_EZ-USB_FX3_Technical_Reference_Manual.pdf>> and <<001-92220_AN75779_How_to_Implement_an_Image_Sensor_Interface_with_EZ-USB_FX3_in_a_USB_Video_Class_UVC_Framework.pdf>>

And I find a problem.

In the chapter 7.11 of Techincal reference manual, there is a flag named DMA_RDY_TH0. And DAM_RDY_TH1.

According to my understand, this flag is used to solve the GPIF's problem between data transfer and the DMA buffer switch in different thead. When the thread is swith to used a free DMA buffer, this flag is used to pause the data transfer.

So if the GPIF used 2 thread to receive the data, this flag should be used.

But in the AN75779, I did not find any information about this flag. So does it mean that there is a loss of data during the 2 thread switch in AN75779 example?

Thanks

Chad

你好，我使用一个uvc工具验证发送设置camera 图像对比度，饱和度命令，但是CyU3PUsbRegisterSetupCallback(CyCx3AppUSBSetupCB, CyTrue)注册的回调函数不响应，函数头里加了日志也没有通过串口打印出来。此uvc工具直接测试usb camera摄像头是有效果的。我想咨询一下，fx3需要如何注册回调函数才会响应类似命令？

Hi

I experience an FX3 EP_IN data transfer stop problem. Our design is based on the uvc_cdc example design.

UVC channel is auto_many_to_one DMA, the setting is:

dmaMultiConfig.size = 16384;
dmaMultiConfig.count = 4;
dmaMultiConfig.validSckCount = 2;
dmaMultiConfig.prodSckId [0] = (CyU3PDmaSocketId_t)CY_U3P_PIB_SOCKET_0;
dmaMultiConfig.prodSckId [1] = (CyU3PDmaSocketId_t)CY_U3P_PIB_SOCKET_1;
dmaMultiConfig.dmaMode = CY_U3P_DMA_MODE_BYTE;
dmaMultiConfig.notification = CY_U3P_DMA_CB_PROD_EVENT | CY_U3P_DMA_CB_CONS_EVENT;

dmaMultiConfig.consSckId [0] = (CyU3PDmaSocketId_t)(CY_U3P_UIB_SOCKET_CONS_0 | CY_FX_SOCKET_UVC);
dmaMultiConfig.prodAvailCount = 0;
dmaMultiConfig.prodHeader = 0; /* 0 byte UVC header to be added. */
dmaMultiConfig.prodFooter = 0; /* 0 byte footer to compensate for the 12 byte header. */
dmaMultiConfig.consHeader = 0;
dmaMultiConfig.cb = CyFxUvcDmaCallback; /* use Capture mode's DMA callback */

apiRetStatus = CyU3PDmaMultiChannelCreate (&glChHandleUVCStream,
CY_U3P_DMA_TYPE_AUTO_MANY_TO_ONE,
&dmaMultiConfig);

CDC channel is manual DMA, the setting is:

/* Create a DMA_MANUAL channel between UART producer socket and USB consumer socket
* Use a smaller buffer size (32 bytes) to ensure that packets get filled in a short time.
*/
dmaCfg.size = 32;
dmaCfg.count = 32;
dmaCfg.prodSckId = (CyU3PDmaSocketId_t)(CY_U3P_LPP_SOCKET_UART_PROD);
dmaCfg.consSckId = (CyU3PDmaSocketId_t)(CY_U3P_UIB_SOCKET_CONS_0 | CY_FX_SOCKET_CDC);
dmaCfg.notification = CY_U3P_DMA_CB_PROD_EVENT;
dmaCfg.cb = CyFxUART2USBDmaCallback;

apiRetStatus = CyU3PDmaChannelCreate (&glChHandleUarttoUsb,CY_U3P_DMA_TYPE_MANUAL, &dmaCfg);

Test setting is:

An app from host PC connect with FPGA through CDC channel. This app enable FPGA to generate UVC traffic, after this the app doing nothing, so there is no traffic flow on CDC channel.

Window "Camera" play video FPGA generated (sent to host PC through UVC channel). The rate is about 320M Bytes per second (90% of GPIF bus bandwidth)

The traffic stop is random, some time in a few minute, sometime a few hours, most time never happens. I am able to reproduce this problem by adding 4 hubs between host PC and device. 

Most time when UVC traffic stop, CDC upstream traffic stop as well, CDC downstream traffic is still ok.

If I build FX3 firmware with SDK1.3.3, I can resume CDC upstream traffic by close and re-open host PC app(close and open uart port). UVC traffic can be recovered by changing video channel (in side FX3, uvc related endpiont and DMA channel will be deleted and re-created).

If I build FX3 firmware with SDK1.3.4, this problem will happen much much less often, but once happens, host PC apps freeze, I have to unplug USB cable to close apps, but in one time after about 20 seconds, CDC and UVC channel recovered by itself.

I used a USB analyser to capture traffic to and from FX3, there is no CRC error, the traffic stop is because FX3 send a NRDY transaction packet to host PC (there is no ERDY transaction packet send out by FX3 after NRDY)

Following is a screen shot of some captured data, there are about 30 to 50 unexpected transaction (in class view of usb analyzer) before NRDY, but they look the same as those good one.

Any one has any idea about this please give me a help, Thanks! 

Hi all,

I'm trying to use the AN75779 project as a baseline to set up my MT9P031 as a UVC camera.

Here's what I did:

Modify GPIF state machine: changed to 16 bit wide bus with one clock per pixel, changed LD_ADDR_COUNT and LD_DATA_COUNT limits to 8183 to account for this. Changed the active clock edge to falling.

In order to fake the YUY2 I tied data 15 high and 14:8 low so the upper byte would be 0x80. Then I just attached my most significant 8 bits of my camera bus to data lines 7:0.

Overhaul to the I2C in sensor.c code. The MT9M114 uses two bytes of register addresses in its protocol while the MT9P031 only has one so I had to get rid of the upper address byte in all the functions. I got rid of all of the configuration for MT9M114 and only wrote in commands to set my sensor up to run 720p at whatever rate it runs. I figured initially setting the sensor up to do 720p to match the existing project would be the path of least resistance.

I have a 24MHz clock and the datasheet suggests it would get 60fps with 96MHz so I assume it's actually going to give 15fps instead of 30.

In the UVC Descriptor C file, I've tried just leaving everything alone, as well as increasing the frame time to double or four times the original (so 0xA2C2A or 0x145854) to account for a lower FPS but nothing is helping.

The result is that Windows recognizes the device as a camera and I'm able to select it with the Windows camera app, and I see some response on the debug serial port when I do select or de-select it.

With the additional debug print enabled, I see "UVC: Completed (n) frames and 0 buffers."

Ultimately though I just get a black screen.

Do I need to use the PLL to increase my clock rate to actually achieve 30fps? Or do I need to adjust the minimum bitrate in the descriptor to reflect what the camera actually puts through?

Or am I missing something else?

Thanks in advance!

Hi!

I'm looking for an example that streams parallel data back to a PC using the GPIF on an FX3. Does such an example exist? I'm attempting to build an SRAM sniffer.

Thanks!

Hi,
I used cyfxuartlpdmamode example and modified the channel to UART-> CPU ( MANUAL IN) and another channel CPU ->UART (MANUAL OUT). I attached my new firmware.  I used the UART event(UartIsr). I send data from Teraterm, I got ISR after sending 16-bytes(DMA size). I put a fixed buffer, when I got the ISR, I send that data to Teraterm. Up to here, everything is ok. But I have 2 problems:
1- I don't get another ISR. I mean I sent 16 bytes first time, I got letters A to Q (fixed buffer) in Teraterm, but again I entered another 16 bytes and I expected to get another ISR, but I don't get. I think the previous event hasn't had clear, and I don't know how I should do that.(SendDebugMessage(buf, len))

2- When I try to replace the fixed buffer with the input buffer, I got nothing. (SendDebugMessage(buf_p.buffer, len))

Could you please help me on this matter?

Thanks

Hi,

I try to use JTAG to do on-line update of FX3 firmware (to ram directly).

For this, I have working implementations, both with custom and with OpenOCD.

Both implementations are doing the same thing: stop ARM926, reboot the System (to get to a known state), load binary to relevant places, set firmware entry point in memory at 0x40000000, trigger a reboot to get OTP rom code do platform initialization and jump back to the firmware.

Both implementations run fine, most of the time. But sometimes, it happens that FX3 gets unresponsive, both from USB (not enumerated) and from JTAG (does not react to TLR and chain discovery).

For instance, this happens when running OpenOCD with the following script in a loop:

source [find mem_helper.tcl]
source [find interface/ftdi/digilent_jtag_hs3.cfg]

reset_config trst_only

adapter speed 5000

jtag newtap fx3 cpu -irlen 4 -ircapture 0x1 -irmask 0xf -expected-id 0x07926069

target create fx3.cpu arm926ejs -endian little -chain-position fx3.cpu
fx3.cpu configure -work-area-phys 0x4003f000 -work-area-size 0x1000 -event "reset-assert" { }

init

arm7_9 dcc_downloads enable
arm7_9 fast_memory_access enable

echo "Halting target"
halt
soft_reset_halt

echo "Putting target in SVC mode"
reg cpsr 0xd3

echo "Enabling TCMs"
arm mcr 15 0 9 1 1 0x00000015
arm mcr 15 0 9 1 0 0xf0000019

echo "Setting oscillator parameters"
mww 0xE0052000 0x00080015

echo "Masking IRQs"
mww 0xfffff014 0xffffffff

echo "Loading firmware"
mww 0x40000000 0x00000000 3072
load_image firmware.elf 0 elf

echo "Jumping back to ROM, entry=0x40012b28"
# Firmware entry point, for boot rom
mww 0x40000000 0x40012b28

# Shell code that basically triggers a reset through GCTL_CONTROL
mww 0x40000004 0xe59f0018
mww 0x40000008 0xe5901000
mww 0x4000000c 0xe59f2014
mww 0x40000010 0xe1c11002
mww 0x40000014 0xe59f2010
mww 0x40000018 0xe1811002
mww 0x4000001c 0xe5801000
mww 0x40000020 0xeafffffd
mww 0x40000024 0xe0050000
mww 0x40000028 0x40000007
mww 0x4000002c 0x20000001
reg pc 0x40000004
resume

irscan fx3.cpu 4
runtest 10

shutdown

After some successful iterations giving out:

Open On-Chip Debugger 0.10.0+dev-01411-g051e80812-dirty (2020-09-30-16:01)
Licensed under GNU GPL v2
For bug reports, read
http://openocd.org/doc/doxygen/bugs.html
Info : auto-selecting first available session transport "jtag". To override use 'transport select <transport>'.
Info : clock speed 5000 kHz
Info : JTAG tap: fx3.cpu tap/device found: 0x07926069 (mfg: 0x034 (Cypress), part: 0x7926, ver: 0x0)
Info : Embedded ICE version 6
Info : fx3.cpu: hardware has 2 breakpoint/watchpoint units
Info : starting gdb server for fx3.cpu on 3333
Info : Listening on port 3333 for gdb connections
Halting target
target halted in ARM state due to debug-request, current mode: Supervisor
cpsr: 0x60000013 pc: 0x40012cf0
MMU: enabled, D-Cache: enabled, I-Cache: enabled
requesting target halt and executing a soft reset
target halted in ARM state due to debug-request, current mode: Supervisor
cpsr: 0x600000d3 pc: 0x00000000
MMU: disabled, D-Cache: disabled, I-Cache: disabled
Putting target in SVC mode
Enabling TCMs
Setting oscillator parameters
Masking IRQs
Loading firmware
Jumping back to ROM, entry=0x40012b28
shutdown command invoked

 I get to a point where chain enumeration is broken (basically, FX3 does not drive TDO any more):

Open On-Chip Debugger 0.10.0+dev-01411-g051e80812-dirty (2020-09-30-16:01)
Licensed under GNU GPL v2
For bug reports, read
http://openocd.org/doc/doxygen/bugs.html
Info : auto-selecting first available session transport "jtag". To override use 'transport select <transport>'.
Info : clock speed 5000 kHz
Info : JTAG tap: fx3.cpu tap/device found: 0xffffffff (mfg: 0x7ff (<invalid>), part: 0xffff, ver: 0xf)
Warn : JTAG tap: fx3.cpu UNEXPECTED: 0xffffffff (mfg: 0x7ff (<invalid>), part: 0xffff, ver: 0xf)
Error: JTAG tap: fx3.cpu expected 1 of 1: 0x07926069 (mfg: 0x034 (Cypress), part: 0x7926, ver: 0x0)
Error: Trying to use configured scan chain anyway...
Error: fx3.cpu: IR capture error; saw 0x0f not 0x01
Warn : Bypassing JTAG setup events due to errors
Info : Embedded ICE version 15
Error: unknown EmbeddedICE version (comms ctrl: 0xffffffff)
Info : fx3.cpu: hardware has 2 breakpoint/watchpoint units
Warn : WARNING: unknown debug reason: 0xf
Warn : ThumbEE -- incomplete support
Info : starting gdb server for fx3.cpu on 3333
Info : Listening on port 3333 for gdb connections
Halting target
requesting target halt and executing a soft reset
Error: Jazelle state handling is BROKEN!
target halted in Jazelle state due to debug-request, current mode: Supervisor
cpsr: 0xffffffd3 pc: 0x00000000
MMU: disabled, D-Cache: disabled, I-Cache: disabled
Putting target in SVC mode
Enabling TCMs
Setting oscillator parameters
Masking IRQs
Loading firmware
Warn : WARNING: unknown debug reason: 0xf
Warn : ThumbEE -- incomplete support
Error: timed out while waiting for target debug-running
Warn : WARNING: unknown debug reason: 0xf
Warn : ThumbEE -- incomplete support
Error: timed out while waiting for target debug-running
Warn : WARNING: unknown debug reason: 0xf
Warn : ThumbEE -- incomplete support
Error: timed out while waiting for target debug-running
Warn : WARNING: unknown debug reason: 0xf
Warn : ThumbEE -- incomplete support
Error: timed out while waiting for target debug-running
Jumping back to ROM, entry=0x40012b28
Error: unhandled core state

The only thing that gets it back to a functional state is to cycle power. Sometimes, toggling reset pin also gets it out of limbo, but not always. I cannot rely on reset pin.

Now the questions:

1. Is this a known defect ?
2. Supposedly, JTAG should have precedence on reset pin, not the other way around, is this a known issue ?
3. Is there a known sequence that is known to produce stable results, i.e. something that can be run in a loop without getting to an unrecoverable state at some point ?

I could not find reference of 1. in public documentation or errata, I could not find reference code for 3.

As an illustration of 2., FX3 JTAG TAP does not answer when reset pin (not trst) is asserted, while TAP should still answer in such cases.

Thanks