DM8168 PG2.1 video decoder problem

Hello,

What is the software release, that you are using?

Are you seeing issue with only  particular stream?

Best Regards,

Margarita

Hello,

 It's nice to hear from you. We use TI 8168 RDK 3.0. The part number of 8168 is x8168x3894CCYG. The video stream is OK. Our code can be run on TI RDK, this problem shows on our HW, the difference is RDK using PG2.0, we use PG2.1. System get h.264 video stream from network camera, decode and display on HDMI output. If we run A8 code, system work well. But system crashes after 20 min. when we run decoder code on HDVPSS. The result is same whatever HDVPSS decode 1CH,2CH,16CH video stream. I want know any way to find the problem

This is unlikely to be due to PG2.1. Are you getting any console logs at point of crash.If so pls share them. There can be many issues related to board design that can cause instability. Pls check the following:

1. DDR byte wise SW leveling is done

2. CHeck for voltage ripple on AVS rail.

3. Check proper heat sink and fan present on board.

4. Check the frequency for various cores like A8/M3/HDVPSS/HDVICP are within spec for your part number.

Hello,

 Thanks for your help.

1. We have done it, result is same. We try to reduce DDR3 speed to check result

2. We use constant voltage

3. Yes, heat sink and fan present on board

4.  We use PG2.1, x8168x3894CCYG, where can I check the frequency for various cores like A8/M3/HDVPSS/HDVICP are within spec ?

816x data sheet has the max frequency for different parts:

http://www.ti.com/litv/pdf/sprs614c (page 320)

Hello,

1.   What is the result with reduced frequency of DDR?
2. Try reducing the load on SOC by reducing the fps.
3. What is the current limit of AVS power supply?
4. What is the constant voltage AVS value?

regards, Abhijit

Hello,

 1.We reduce DDR3 frequency to 533, it is more table, 800 will cause system crash. We wonder why A8 code can run at 800 configuration, but system crashes if run video decoder code

2. We try to reduce FPS and channel number, result is same on 800 configuration

3. The power supply we design is 10A, we don't find voltage dop

4. 1.01V

Hello,

Local display test  with 675MHz DDR clock crashed.

Attached is the message dump when system crashed.

5843092:!!!SLAVE CORE [VIDEO-M3] DOWN!!!
SystemLink_copySlaveCoreExceptionContext:146
mmap of [0xbe9c0000:36864]
mmap virt addresss:0x40072000
munmap of [0x40072000:36864]
SystemLink_copySlaveCoreExceptionContext:153
 [m3video] Unhandled Exception:
 [m3video] Exception occurred in ThreadType_HWI
 [m3video] handle: 0x9dd68bcc.
 [m3video] stack base: 0x9dd622dc.
 [m3video] stack size: 0x4000.
 [m3video] R0 = 0x00008030  R8  = 0x9d26da08
 [m3video] R1 = 0x00008019  R9  = 0x0000000f
 [m3video] R2 = 0xe000e000  R10 = 0x00000000
 [m3video] R3 = 0x9dd67b68  R11 = 0x9d26da0c
 [m3video] R4 = 0x9dd68bcc  R12 = 0x9dd44cf8
 [m3video] R5 = 0x9d26d6c0  SP(R13) = 0x9dd661f8
 [m3video] R6 = 0x00000000  LR(R14) = 0x9d2aa0d5
 [m3video] R7 = 0x9d26da10  PC(R15) = 0x00008018
 [m3video] PSR = 0x6100000f
 [m3video] ICSR = 0x00423003
 [m3video] MMFSR = 0x00
 [m3video] BFSR = 0x01
 [m3video] UFSR = 0x0000
 [m3video] HFSR = 0x40000000
 [m3video] DFSR = 0x00000000
 [m3video] MMAR = 0xe000ed34
 [m3video] BFAR = 0xe000ed38
 [m3video] AFSR = 0x00000000
 [m3video] Terminating Execution...
remove dec ch 0

Close Channel: 0

So far,531Mhz is more stable if we enable video decoder

hello

   I met the same problem. M3 crashed when video decoder(h.264) is runing about 20 min or least. 

   And the log was the same as yours. 

   Have you solved the problem?  

  thank you !!

hi badri,

   I have met once till now and it seems a problem. did you have any idea?