Re: fix for 37878 and 37550

[christos%astron.com@localhost (Christos Zoulas)]

In article <grpj80$6go$1%serpens.de@localhost>,
Michael van Elst <mlelstv%serpens.de@localhost> wrote:
>christos%zoulas.com@localhost (Christos Zoulas) writes:
>
>>Hello,
>
>>I've implemented stat functions for all the devices that called fbadop_stat
>>and tried to fill the fields as best as I could. I am concerned about the
>>cost of getnanotime() and the extra fields in the softc struct. Is it worth
>>it? We could also simply fix this by implementing fnullop_stat, which is
>>effectively what soo_stat() ends up being for non-unix sockets.
>
>>Comments?
>
>POSIX says we need the timestamps, but then it seems to only talk about
>maintaining it for read() and write() but not for send()/recv(), so
>that's not a very strong reason.

I am not doing this to comply with POSIX. I just want to give more information
to userland.

>Pipes use getmicrotime() instead of getnanotime(). Both functions
>are very fast. Cost shouldn't be an issue here.

I will fix that.

>The PRs are about fstat(1) not printing descriptors which was already
>fixed by implementing output routines for the missing descriptor types
>(e.g. DTYPE_MISC). The fix hasn't been pulled up.

I wanted to fix fstat(2) not fstat(1) so that it never returns an error
for active file descriptor (unless there is a system error EIO or programming
error EFAULT).

>N.B. fstat shows the missing descriptors only when you use -v
>because the error message about an unsupported descriptor type
>is printed as a debug message. I think that should be a regular
>error message.

That will not be needed anymore.

christos