RNG Job Descriptor to copy random number to system RAM

[fmajeric]

Hello,

I read a resolved discussion about RNG Job Descriptor:

P4080 SEC4.0 Job Descriptor to copy random numbers to system RAM​

I've the same need : generate a simple random number and store it at  the address 0x10000170 in RAM.

So I use the descriptor provided in the answer of the discussion:

### rnggen_1

[00] B0800006    jobhdr: stidx=0 len=6

[01] 12820004    ld: ccb1-datasz len=4 offs=0 imm

[02] 00000020    data:0x00000020

[03] 82500000    operation: cls1-op rng (SH0) generate random

[04] 60300020    fifostr: msg len=32

[05] 10000170    @ptr->@0x10000170

But when I load the Job Descriptor Base Address into the Input Job Ring Register,

I have no Random number at the address  @0x10000170...

... what is the problem with this descriptor ?

What I don't do well ?

Please, could you help me ?

Fabien.

[bpe]

What is the Job termination status returned for this JD?

Have a great day,
Platon

-----------------------------------------------------------------------------------------------------------------------
Note: If this post answers your question, please click the Correct Answer button. Thank you!
-----------------------------------------------------------------------------------------------------------------------

[fmajeric]

Hi Platon,

I will try to explain what happened:

1- First, I load the descriptor in RAM memory. I initiate all the Job Ring registers in CAAM (size, addresses, ...)

Addresses in RAM	Comments	Values
0x10000000	JR0_input_Base_Addr	0x10000100
...	... ...	...
0x10000040	JR0_Output_Base_Addr	0x00000000
...	... ...	...
0x10000100	jobhdr: stidx=0 len=6	0xB0800006
0x10000104	ld: ccb1-datasz len=4 offs=0 imm	0x12820004
0x10000108	data:0x00000020	0x00000020
0x1000010C	operation: cls1-op rng (SH0) generate random	0x82500000
0x10000110	fifostr: msg len=32	0x60300020
0x10000114	@ptr->@0x10000170	0x10000170
...	... ...	...
0x10000170	where the descriptor tell the CAAM to store the result.	0x00000000
...	... ...	...

--> At this moment, the status register are:

Addresses in RAM	Comments	Values
0x02100DF4	JRJDS — Job Ring Job-Done Status Register	0x00000000
...	... ...	...
0x02100FD4	CSTA — CAAM Status Register	0x00000006
...	... ...	...
0x02101044	JROSR0 — Job Ring Output Status Register for Job Ring 0	0x00000000
...	... ...	...
0x02101FD4	CSTA — CAAM Status Register	0x00000006

2- Then I write into the IRJAR0 of the CAAM the value 0x1 in order it begin its work:

0x0210101C

IRJAR0 — Input Ring Jobs Added Register for Job Ring 0

0x00000001

3- Last it returns this:

Addresses in RAM	Comments	Values
0x10000000	JR0_input_Base_Addr	0x10000100
...	... ...	...
0x10000040	JR0_Output_Base_Addr	0x10000100
0x10000044		0x20000454
...	... ...	...
0x10000100	jobhdr: stidx=0 len=6	0xB0800006
0x10000104	ld: ccb1-datasz len=4 offs=0 imm	0x12820004
0x10000108	data:0x00000020	0x00000020
0x1000010C	operation: cls1-op rng (SH0) generate random	0x82500000
0x10000110	fifostr: msg len=32	0x60300020
0x10000114	@ptr->@0x10000170	0x10000170
...	... ...	...
0x10000170	where the descriptor tell the CAAM to store the result.	0x00000000
...	... ...	...

--> then the status registers are:

Addresses in RAM	Comments	Values
0x02100DF4	JRJDS — Job Ring Job-Done Status Register	0x00000000
...	... ...	...
0x02100DF4	CSTA — CAAM Status Register	0x00000005
...	... ...	...
0x02100DF4	JROSR0 — Job Ring Output Status Register for Job Ring 0	0x20000454
...	... ...	...
0x02101FD4	CSTA — CAAM Status Register	0x00000005

I hope it explains quite well the situation. If not don't hesitate to ask me some specifications.

Thank you for your help,

Fabien.

[bpe · Accepted solution]

The job status tells you that  RNG has not been instantiated. Refer to Security RM, Sections 5.7.3.2.1, 5.7.3.2.2,

Table 5-68.

Have a great day,
Platon

-----------------------------------------------------------------------------------------------------------------------
Note: If this post answers your question, please click the Correct Answer button. Thank you!
-----------------------------------------------------------------------------------------------------------------------

[fmajeric]

Ok Platon,

Thank you very much. It sould be the reason for this dysfunction :smileyhappy: ...

Best Regards,

Fabien.