Build is broken for target aarch64-linux-android

[flxo]

The bionic interface slightly differs from glibc and others. The build is broken due to some type conflict:

error[E0605]: non-primitive cast: `&[u8; 129]` as `&[i8]`
  --> src/core/deviceinfo.rs:44:35
   |
44 |         let uuid = str_from_c_str(&ioctl.uuid as &[i8]).ok_or_else(|| {
   |                                   ^^^^^^^^^^^^^^^^^^^^ an `as` expression can only be used to convert between primitive types or to coerce to a specific trait object

error[E0605]: non-primitive cast: `&[u8; 128]` as `&[i8]`
  --> src/core/deviceinfo.rs:71:32
   |
71 |                 str_from_c_str(&ioctl.name as &[i8])
   |                                ^^^^^^^^^^^^^^^^^^^^ an `as` expression can only be used to convert between primitive types or to coerce to a specific trait object

error[E0308]: mismatched types
  --> src/core/dm.rs:98:59
   |
98 |         let _ = name.as_bytes().read(mut_slice_from_c_str(&mut hdr.name))?;
   |                                                           ^^^^^^^^^^^^^ expected slice `[i8]`, found array `[u8; 128]`
   |
   = note: expected mutable reference `&mut [i8]`
              found mutable reference `&mut [u8; 128]`

error[E0308]: mismatched types
   --> src/core/dm.rs:103:59
    |
103 |         let _ = uuid.as_bytes().read(mut_slice_from_c_str(&mut hdr.uuid))?;
    |                                                           ^^^^^^^^^^^^^ expected slice `[i8]`, found array `[u8; 129]`
    |
    = note: expected mutable reference `&mut [i8]`
               found mutable reference `&mut [u8; 129]`

error[E0308]: mismatched types
   --> src/core/dm.rs:481:44
    |
481 |             let dst = mut_slice_from_c_str(&mut targ.target_type);
    |                                            ^^^^^^^^^^^^^^^^^^^^^ expected slice `[i8]`, found array `[u8; 16]`
    |
    = note: expected mutable reference `&mut [i8]`
               found mutable reference `&mut [u8; 16]`

error[E0308]: mismatched types
   --> src/core/dm.rs:586:50
    |
586 |                 let target_type = str_from_c_str(&targ.target_type)
    |                                                  ^^^^^^^^^^^^^^^^^ expected slice `[i8]`, found array `[u8; 16]`
    |
    = note: expected reference `&[i8]`
               found reference `&[u8; 16]`

error[E0308]: mismatched types
  --> src/core/dm_ioctl.rs:41:33
   |
41 |                 &str_from_c_str(&self.name as &[libc::c_char]).unwrap_or("Could not parse string"),
   |                                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ expected `i8`, found `u8`
   |
   = note: expected reference `&[i8]`
              found reference `&[u8]`

error[E0308]: mismatched types
  --> src/core/dm_ioctl.rs:45:33
   |
45 |                 &str_from_c_str(&self.uuid as &[libc::c_char]).unwrap_or("Could not parse string"),
   |                                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ expected `i8`, found `u8`
   |
   = note: expected reference `&[i8]`
              found reference `&[u8]`

error[E0308]: mismatched types
  --> src/core/dm_ioctl.rs:49:33
   |
49 |                 &str_from_c_str(&self.data as &[libc::c_char]).unwrap_or("Could not parse string"),
   |                                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ expected `i8`, found `u8`
   |
   = note: expected reference `&[i8]`
              found reference `&[u8]`

error: aborting due to 9 previous errors

Some errors have detailed explanations: E0308, E0605.
For more information about an error, try `rustc --explain E0308`.
error: could not compile `devicemapper`

To learn more, run the command again with --verbose.

To reproduce run a cargo check --target aarch64-linux-android I tested with NDK r19c.

[mulkieran]

Note: Perhaps libc::char would have been better as i8, which would work with at least two targets.

[jbaublitz]

@mulkieran If you want me to take a look at this, let me know! I may even be able to set up CI to test for musl and android (compile only) if that would be desirable.

[mulkieran]

@jbaublitz Please go ahead and attempt to set up CI for other architectures. Perhaps you can learn something from the record of failures so far in #641.