Opening and closing files is bad for performance, so now that
we accurately track the situation of hitting too many open files
we can keep the band files open, even for the non-readbuf code
path.
This unifies the logic of opening (and caching file descriptors)
between the readbuf and non-readbuf code paths.
The previous logic was flawed anyways, as reading a range could
potentially span multiple bands, but we only checked if we were
just below the max limit.
Using the process group (now that we've got one that's exclusive
to the testrunner) allows us to also clean up children that have
been detached from their original parent (and now have PID 1 as
their parent).
We now reflect the user who mounted the sparsebundle as the owner of
the resulting .dmg file, and try to reflect the state of allow_other
and allow_root through the file permission bits.
These are just informative. The real access control happens in FUSE
based on the allow_other/allow_root options (unless user also passes
the default_permissions option, in which case the permission bits
will also be checked).
Fixes issue #15
The move from off_t to size_t in 90c81b1776 was misguided. Although
it semantically makes sense to use an unsigned datatype for file sizes,
the size_t data type is only 32-bit wide on 32-bit systems, so the move
broke reading of large sparsebundles on 32-bit systems, since we could
no longer represent the file size.
We now use uint64_t explicitly, which is both unsigned, and wide enough
to represent all the files we want.
The uintmax_t datatype has been demoted to be used only for printing,
to make it clearer what width is actually expected and used in the code.
Fixes#26
Tests are run using 'make check', and can be combined with the
existing Makefile targets, e.g. 'make check all' to run checks
on all available platforms.
Pass DEBUG=1 for additional test output.
