Build fails on big-endian

[n4ss]

Hey!

We're currently running into the following (as expected) on big-endian machines:

   |
10 | / compile_error!(
11 | |     r#"
12 | | This crate doesn't support big-endian targets, since I didn't
13 | | have one to test correctness on.  If you're seeing this message,
14 | | please file an issue!
15 | | "#
16 | | );
   | |__^

error: aborting due to previous error

error: could not compile `merlin`.

[n4ss]

FYI, there is a pretty complete tutorial on how to test endian aware code through Qemu https://create.stephan-brumme.com/big-endian/

[hdevalence]

@n4ss Cool, if you'd like to write a PR that adds endianness testing to CI (could be Github Actions instead of Travis, whichever is easier), I'd be happy to merge it!

[vishwabmc]

Hello,

We are building https://github.com/paritytech/polkadot on s390x platform, which is a Big Endian. Polkadot is using merlin and it is failing with the same error that has been reported in this.

To get going, I have commented that block and got the polkadot compiled. I have not tested that yet, but because I commented this, I am foreseeing some issues to come while running.

Wonder if you there is any plan on supporting Big Endian platforms ?

Running unittests (target/debug/deps/merlin-244c4f32d2c2636d)

running 4 tests
test transcript::tests::equivalence_simple ... FAILED
test transcript::tests::transcript_rng_is_bound_to_transcript_and_witnesses ... ok
test strobe::tests::test_conformance ... FAILED
test transcript::tests::equivalence_complex ... FAILED

failures:

---- transcript::tests::equivalence_simple stdout ----
thread 'transcript::tests::equivalence_simple' panicked at 'assertion failed: `(left == right)`
  left: `[113, 14, 135, 252, 119, 93, 73, 162, 144, 233, 98, 44, 115, 204, 183, 193, 14, 24, 106, 72, 213, 247, 144, 93, 189, 93, 204, 25, 37, 102, 237, 70]`,
 right: `[213, 162, 25, 114, 208, 213, 254, 50, 12, 13, 38, 63, 172, 127, 255, 184, 20, 90, 166, 64, 175, 110, 155, 202, 23, 124, 3, 199, 239, 207, 6, 21]`', src/transcript.rs:444:9
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace

---- strobe::tests::test_conformance stdout ----
thread 'strobe::tests::test_conformance' panicked at 'assertion failed: `(left == right)`
  left: `[58, 171, 197, 41, 152, 177, 135, 146, 172, 35, 191, 55, 143, 111, 197, 172, 220, 196, 217, 218, 56, 23, 228, 19, 48, 158, 103, 83, 95, 40, 106, 184]`,
 right: `[180, 142, 100, 92, 161, 124, 102, 127, 213, 32, 107, 165, 122, 106, 34, 141, 114, 216, 225, 144, 56, 20, 211, 241, 127, 98, 41, 150, 215, 207, 239, 176]`', src/strobe.rs:211:9

---- transcript::tests::equivalence_complex stdout ----
thread 'transcript::tests::equivalence_complex' panicked at 'assertion failed: `(left == right)`
  left: `[12, 106, 162, 156, 230, 4, 178, 155, 223, 154, 239, 86, 20, 223, 244, 166, 163, 112, 78, 178, 163, 220, 5, 53, 46, 222, 11, 142, 179, 14, 201, 54]`,
 right: `[182, 26, 85, 64, 195, 220, 22, 229, 215, 70, 194, 46, 11, 253, 53, 9, 98, 30, 55, 224, 136, 161, 68, 85, 129, 129, 58, 116, 110, 156, 18, 234]`', src/transcript.rs:467:13


failures:
    strobe::tests::test_conformance
    transcript::tests::equivalence_complex
    transcript::tests::equivalence_simple

test result: FAILED. 1 passed; 3 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.02s

Thanks,
!! Vishwa !!

[vishwabmc]

Hello, is there any feedback on this ?