Dongsheng "Homer" Xing

SHA3 (KECCAK)

Hi, Can anybody send me the code for SHA 3-512?. The code available here is not generating the correct hash code. I have compared the hash results with matlab code and some online websites but the output is not matching. 
I have tried the two ways provided by others in this Issue List but those suggestions not working.
Kindly someone help to get correct hash code.

Thank you in advance.

SHA 3-512: This verilog code is not providing correct Hash code. Request for Correct Hash Code

This implementation is correctly work Keccak-512 hash function.

If you want to update it to SHA3-512, you need to fix values in padder1.v component as:

case (byte_num)
		0: out =                  64'h0600000000000000;
		1: out = {in[63:56], 56'h06000000000000};
		2: out = {in[63:48], 48'h060000000000};
		3: out = {in[63:40], 40'h0600000000};
		4: out = {in[63:32], 32'h06000000};
		5: out = {in[63:24], 24'h060000};
		6: out = {in[63:16], 16'h0600};
		7: out = {in[63:8],   8'h06};
endcase

WHY?

![](https://cdn.opencores.org/usercontent/132b59891f718e0fbeff3a2fc6cefc23462453c46c4ec0655c6f0dbbad323240.png)

Hash not work correctly as SHA3-512

I noticed that when I put the test strings (The quick brown fox...) in test_keccak.v through a standard SHA3-512, the result is different than the expected value in the test bench for all 3 lengths.
Is there any particular reason for that? Thanks.

Wrong test vectors

Keccak uses 10*1 padding
In "padder.v", the first '1' is padding when is_last is high.
But is is_last comes when buffer_full is high, it sets 'state'=1.
This makes the padding 00*1 instead of 10*1.

The fix is to set 'state' only when is_last & buffer_full is high.

Browse

SHA3 (KECCAK)