Re: [PATCH v5 1/2] TTY: Add TTY slave enumeration support

[Greg Kroah-Hartman <gregkh@linuxfoundation.org>]

On Thu, Jan 24, 2013 at 06:30:21PM +0800, Lv Zheng wrote:
> In the recent ACPI 5.0 specification updates, firmwares are provided the
> possibilities to enumerate the UART slave devices known to the platform
> vendors.
> There are the needs in Linux to utilize the benefits:
> 1. hotplug uevent
> 2. serial configuration
> Currently, only serial cards on the specific bus (ex. PCMCIA) can be
> enumerated and userspace can obtain the hotplug event of the UART target
> devices. Linux kernel is lack of an overall enumeration mechanism for
> UART slave devices.

Huh?  That's flat out not true, otherwise how would userspace be
creating the proper /dev/tty* nodes for all tty devices?

> In order to send uevent, a device need to be a class device or a bus
> device. This patch introduces a tty_enum bus since the enumerated slave
> devices are expected to be physical devices.

Again, tty devices are already class devices, and they send out uevents.
You can see this today by watching the uevent stream using a tool like
'udevadmin monitor'.

> When the UART slave devices are created, userspace uevent rules can
> pass the creation details to the userspace driver managers
> (ex. hciattach), then the device managers can read hardware IDs and the
> serial configurations from the exported device attributes to match and
> configure a userspace TTY device driver.

What?  We do that today, no kernel changes needed.

> The created slave devices will be automatically unregistered when the
> associated TTY ports are destructed.

Again, this happens today.

Or am I missing something big and major here?  How have people been
seeing and configuring their tty devices for the last 8+ years or so
since the 2.6.0 kernel was released with uevent support for tty devices?

lost,

greg k-h