My actual use case (on a different OS) was for a proxy.

If you call getsockopt with SO_ERROR, it clears out so_error, so the ability to asynchronously clear the error already exists, and hasn't broken things.

Any comments or reviews? 😄

I was paranoid about the lock, as is only correct.

I forgot to edit the man page.

Changes based on Allan's comments:

• Moved a comment around
• Changed the name of a variable to have its purpose clearer
• Updated a comment to reflect the current prototype

Thanks

The code looks fine (small change, after all), other than my request for comments. :)

lgtm.

I still use an 80 character limit, and an 80 character-wide Terminal window.

I think the best way forward is to start with -@ first and leave -l output as it is, at first.

Ping? Anyone?

Long long long overdue.

Ping?

Oh, an example output:

I have not forgotten about this; I'm working on another aspect of this (dataset properties), and am holding off on committing this a bit.

Ping anyone?

New version coming now. Thanks!

My only concern involves the /etc symlink possibly pointing at a non-existent file. This is a small concern not worth preventing this.

Great, thanks! :) I'll go over it in more detail over the next few days.

A quick skim and the main question I had is: no acl check?

In D21646#475415, @mjg wrote:

The new op is *not* called by anything but the new handler which has to be explicitly set by filesystems which want the feature, i.e. the change is a no-op for anyone who does not opt in.

There is no way to get a "default" implementation for this.

Perhaps I'm confused, but since you've added a new VFS operation, shouldn't there be a default implementation? You mention doing this for ZFS, but it's only there for tmpfs and UFS; does this need to be a VFS function, or should it be done as part of the mount structure and maintenance functions?

Remove thread argument to vfs_emptydir. Also remove a blank line. Per kib.

The td arg is still there.

Any further comments on this? I'd gotten an OOB message from Josh about this. If not, I'll make sure it's still functional and plan on committing this weekend or Monday.

I forgot to add the assertion kib suggested. (NB, still doesn't have the td changes we discussed.)

More discussion about it than it's worth. Ah well.

In D21660#472249, @avg wrote:

Or add a sleep instead of additional sync-s. And how do you know that 3 will always be enough?

BTW, you marked it as needing revision, but didn't say what revision.

Could you please tell more about this?

If you need more than one sync, then you've broken posix semantics.

If you think that it's useful and sync 3 would be an improvement, then could you please explain why?

Small change from kib feedback.

More changes due to feedback from kib. Thanks!

New update coming in.

Feedback from kib.

Updated diff coming in a few seconds.

Ping, anyone?

Feedback from Alan.

Ping?

Add support for having "cover" and "noemptydir" to the options parsing.

Update the man page dates per reminder.

Still no examples, but linking to the consumer helps.

As for rounding - the bugs section, which mentioned them, is gone. The code is just a fixed-point library, so the rounding is what one would expect, I suppose: stuff gets truncated according to the radix point. Should I document it somehow?

This much work in the kernel for what should be user-space makes me twitchy... but there is lots of precedent for it, so I obviously have to get with the times.

I'm too old for the world.

The spilt-up man pages are much better. The examples definitely help a lot. But now I'm wondering about things like rounding -- the obvious usage that occurred to me here was, ahem, ZZZZZZ9.99, to borrow from a long-ago past :).

This mostly seems ok -- it's pretty much nothing but a man page and header file, after all -- but my biggest complaint is that the man page is exceedingly complicated, and does not provide any real examples. The examples would also be a good way to show the justification for the code.

I'd like to see some example usages.

A bit late, but the updated diff promised. Thanks cem & jhb!

Doing testing with the changes, then I'll upload the diffs.

I'll get to those tomorrow, thanks! (Various stuff going on today has thrown my schedule off.)

The promised updated diffs.

I'm updating and build testing before uploading new diffs.

Poke?

Something made me think that it was not acceptable, but now I can't remember what it was, so it's clearly wrong.