John E Hein
jhein at timing.com
Wed Apr 9 17:59:39 UTC 2008
M. Warner Losh wrote at 10:18 -0600 on Apr 9, 2008:
> Yes. I almost just quietly committed it to FreeBSD at the time, but I
> got busy on another project and never got back to it. Maybe I should
> have just done it and saw if phk noticed :-)
Never try to accomplish through planning what can be done more simply
> : Well, I could go either way on this issue - 'specious' might be a bit
> : too strong. I could see issues with pass-through device-specific
> : ioctls on tty devs - especially due to the fact that it's a tty device
> : is somewhat obscured in the case of ucom children.
> I'm not sure I follow what you are saying here...
I'm not sure which part your asking about, so I'll just try to be a
bit more verbose.
- re: issues with pass-thru ioctls... I'm just commenting that I
understand phk's concerns about potentially ill-advised ioctls on
the tty dev.
- re: ucom-based driver ioctls... those drivers just set the
ucomioctl method and so don't know (without digging) that it hooks
up to the tty ioctl method... hence the 'obscured' comment and
since the relationship to the tty device is somewhat obscured
through an indirect relationship, the ucom-based driver writer
may not be aware of the potential problems with exposing ioctls
on the tty dev.
More information about the freebsd-arch