TEEC_InvokeCommand(forward) failed 0xffff3024 origin 0x3

[HaoXuan-a]

Hi,mofan!
When i execute darknetp classifier train -pp_start 4 -pp_end 10 cfg/mnist.dataset cfg/mnist_lenet.cfg to train a model from scratch, I got the TEEC_InvokeCommand(forward) failed 0xffff3024 origin 0x3 error.

Here's the log:

4524, 150.800: 0.000000, 0.000000 avg, 0.000899 rate, 4.800005 seconds, 452400 images
user CPU start: 15782.053947; end: 15785.557877
kernel CPU start: 9762.548833; end: 9764.112507
Max: 16960 kilobytes
vmsize:2085752; vmrss:16516; vmdata:56060; vmstk:132; vmexe:428; vmlib:2280
Loaded: 0.000248 seconds
D/TC:2 0 abort_handler:524 [abort] abort in User mode (TA will panic)
E/TC:? 0
E/TC:? 0 User TA data-abort at address 0x40029fc0 (write permission fault)
E/TC:? 0 esr 0x9200004f ttbr0 0x200001019d040 ttbr1 0x00000000 cidr 0x0
E/TC:? 0 cpu #2 cpsr 0x30000100
E/TC:? 0 x0 0000000000000001 x1 0000000040028000
E/TC:? 0 x2 0000000043b64ef0 x3 0000000043b67f10
E/TC:? 0 x4 0000000043b69ed0 x5 000000000000000a
E/TC:? 0 x6 0000000000000001 x7 0000000043b68ef0
E/TC:? 0 x8 0000000000000001 x9 0000000043b6e9b0
E/TC:? 0 x10 00000000000003e8 x11 000000000000000a
E/TC:? 0 x12 0000000043b6e9b0 x13 00000000000003e8
E/TC:? 0 x14 0000000000000064 x15 0000000043bab880
E/TC:? 0 x16 000000000000000a x17 0000000043b6bae0
E/TC:? 0 x18 0000000043bf1218 x19 0000000000000001
E/TC:? 0 x20 00000000000003e8 x21 0000000043b68ef0
E/TC:? 0 x22 0000000043b64ef0 x23 0000000043b67f10
E/TC:? 0 x24 0000000043b69ed0 x25 0000000043d2db80
E/TC:? 0 x26 0000000043b69ed0 x27 0000000043b64ef0
E/TC:? 0 x28 0000000043b6da10 x29 000000004002a000
E/TC:? 0 x30 0000000040014280 elr 00000000400141a0
E/TC:? 0 sp_el0 000000004002a000
E/LD: Status of TA 7fc5c039-0542-4ee1-80af-b4eab2f1998d
E/LD: arch: aarch64
E/LD: region 0: va 0x40004000 pa 0x10300000 size 0x002000 flags rw-s (ldelf)
E/LD: region 1: va 0x40006000 pa 0x10302000 size 0x007000 flags r-xs (ldelf)
E/LD: region 2: va 0x4000d000 pa 0x10309000 size 0x001000 flags rw-s (ldelf)
E/LD: region 3: va 0x4000e000 pa 0x1030a000 size 0x003000 flags rw-s (ldelf)
E/LD: region 4: va 0x40011000 pa 0x1030d000 size 0x001000 flags r--s
E/LD: region 5: va 0x40012000 pa 0x00001000 size 0x018000 flags r-xs [0]
E/LD: region 6: va 0x4002a000 pa 0x00019000 size 0x3c04000 flags rw-s [0]
E/LD: region 7: va 0x43c2e000 pa 0x13f2a000 size 0x101000 flags rw-s (stack)
E/LD: region 8: va 0x43d30000 pa 0x08001000 size 0x01d000 flags rw-- (param)
E/LD: [0] 7fc5c039-0542-4ee1-80af-b4eab2f1998d @ 0x40012000
E/LD: Call stack:
E/LD: 0x00000000400141a0
D/TC:? 0 user_ta_enter:164 tee_user_ta_enter: TA panicked with code 0xdeadbeef
D/TC:? 0 destroy_ta_ctx_from_session:313 Remove references to context (0x10170b00)
D/TC:? 0 destroy_context:298 Destroy TA ctx (0x10170b00)
darknetp: TEEC_InvokeCommand(forward) failed 0xffff3024 origin 0x3
D/TC:? 0 tee_ta_close_session:499 csess 0x10170b60 id 1
D/TC:? 0 tee_ta_close_session:518 Destroy session

I looked up the files /optee/optee_os/lib/libutee/include/tee_api_defines.h, found that 0xFFFF3024 represents TEE_ERROR_TARGET_DEAD, but i dont't know the cause of the problem.
Looking forward to your reply, thank you!

[mofanv]

Hi, have you increased the TA_STACK_SIZE and TA_DATA_SIZE ? This error may be caused by insufficient secure memory. Also please check #3 #14

[HaoXuan-a]

yes, I have set the size of TA_STACK_SIZE and TA_DATA_SIZE.
#define TA_STACK_SIZE (1 * 1024 * 1024)
#define TA_DATA_SIZE (60 * 1024 * 1024)

[yhl8611]

Hi, how did you finally solve it?