<div dir="ltr">ping? i saw you added another tar test, but let me know your thoughts on this before i rebase and fix the new test...</div><br><div class="gmail_quote"><div dir="ltr" class="gmail_attr">On Fri, Aug 26, 2022 at 4:29 PM enh <<a href="mailto:enh@google.com">enh@google.com</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><div dir="ltr">macOS doesn't have a 'root' group (it uses 'wheel' instead), which<br>we can work around by using 'sys', since that happens to have the<br>same gid on Linux and BSD. (Android doesn't have a 'sys' group, but<br>I can fix that and give it the same number myself -- we already<br>have similar hacks for Linux Test Project tests -- and this seems<br>like the only way out of our current situation that doesn't require<br>abandoning the "sha" style of test we have here.)<br><br>macOS uses a different device for /dev/null, so work around that too.<br><br>I don't have a good workaround for macOS' weird sparse file behavior<br>(not least because I still haven't understood what it does/doesn't<br>support), so I've just disabled all sparse tests for now.<br><br>I fixed some of the symlink tests (where macOS has group r-x and<br>Linux group rwx) where we're just doing text comparisons, but I've<br>just skipped the "sha" style tests that use symlinks for now.<br><br>With this change, there are no test failures for me on macOS 12.5<br>or on current-ish debian.<br>---<br> tests/tar.test | 101 ++++++++++++++++++++++++++++---------------------<br> 1 file changed, 58 insertions(+), 43 deletions(-)<br><br></div>
</blockquote></div>