test size_of_sha1 reports wrong size on i386

[alexanderkjall]

Build failure:

159s 
159s running 1 test
159s test size_of_sha1 ... FAILED
159s 
159s failures:
159s 
159s ---- size_of_sha1 stdout ----
159s thread 'size_of_sha1' panicked at 'assertion failed: `(left == right)`
159s   left: `96`,
159s  right: `104`', tests/hash.rs:12:5
159s stack backtrace:
159s    0: rust_begin_unwind
159s              at /usr/src/rustc-1.70.0/library/std/src/panicking.rs:578:5
159s    1: core::panicking::panic_fmt
159s              at /usr/src/rustc-1.70.0/library/core/src/panicking.rs:67:14
159s    2: core::panicking::assert_failed_inner
159s    3: core::panicking::assert_failed
159s              at /usr/src/rustc-1.70.0/library/core/src/panicking.rs:228:5
159s    4: hash::size_of_sha1
159s              at ./tests/hash.rs:12:5
159s    5: hash::size_of_sha1::{{closure}}
159s              at ./tests/hash.rs:11:19
159s    6: core::ops::function::FnOnce::call_once
159s              at /usr/src/rustc-1.70.0/library/core/src/ops/function.rs:250:5
159s    7: core::ops::function::FnOnce::call_once
159s              at /usr/src/rustc-1.70.0/library/core/src/ops/function.rs:250:5
159s note: Some details are omitted, run with `RUST_BACKTRACE=full` for a verbose backtrace.
159s 
159s 
159s failures:
159s     size_of_sha1

From: https://ci.debian.net/data/autopkgtest/testing/i386/r/rust-gix-features/40167696/log.gz

Can be reproduced with cargo test --all-features

I'm far from certain that this is the correct fix, as this more documents the size differences on the different architectures.

[Byron]

Thanks a lot!

It's strange that you were able to find this failure as I'd expect this to be tested by the 32bit CI job.

Probably I am missing something, somewhere.

[Byron]

Thanks a lot for your help!