dustman31 / sha-3 Goto Github PK

Its complete junk.
1)state->length isnt a uint64_t count.
2)A is a 25 uint64
3)'for' really terrible

'Keccak.exe' (Win32): Unloaded 'C:\Windows\SysWOW64\kernel32.dll'
The thread 0x1184 has exited with code 0 (0x0).
The program '[8172] Keccak.exe' has exited with code 0 (0x0).

What version you use VS ?
Or how I can solve this problem ?

Do you have a favourite unit testing framework? I can hopefully get the time to write a couple of simple tests. I note you are developing in visual studio- therefore maybe one of https://docs.microsoft.com/en-us/visualstudio/test/writing-unit-tests-for-c-cpp?view=vs-2019

The code looks like it computes the digest properly, but sha3Digest truncates the output to 24 bytes (192 bits), because math is 64-bit. The last 4 bytes are missing from the result.

fairly self explanatory. I have used the boost/beast implementation, which has a very open (but bespoke) licence here

Bjorn Nilsson pointed out the following:

In the doFile function, the keccakUpdate call looks different for the three hash types:

For hashType 0: keccakUpdate((uint8_t*)buf, 0, bytesRead, st);
For hashType 1: keccakUpdate((uint8_t*)buf, 0, bufferSize, st);
For hashType 2: keccakUpdate((uint8_t*)buf, 0, bufferSize, st);

Is the difference bytesRead vs bufferSize intentional? It seems to me the code might feed uninitialized buffer for hashType 1 and 2 if bytesRead<bufferSize.

This is correct - it should have been bytesRead. This was likely missed when I converted the project from reading the whole file at once. I have pushed the fix, and encourage users to update to this version.

My Code：

Input ”A“

Debug（Correct）：
421a35a60054e5f383b6137e43d44e998f496748cc77258240ccfaa8730b51f40cf47c1bc09c728a8cd4f096731298d51463f15af89543fed478053346260c38

Release：
fff33bdfffbfdbffb5712b45afaa596feaf59e440bf66b0f18948e4cc9ebc144e7c9f0a0aa3c8fc7375e2a4fc8add620f28210b2cb87f1d040448b62fbe7392a