Reattach/redetect allways connected umass device - is it possible ?

[Vladimir Grebenschikov]

В пн, 28/03/2005 в 13:46 +0200, Bernd Walter пишет:
> On Mon, Mar 28, 2005 at 03:35:10PM +0400, Vladimir Grebenschikov wrote:
> > ? ??, 28/03/2005 ? 13:11 +0200, Bernd Walter ?????:
> > > > 
> > > > Unfortunately, it seen as not-ejectable media.
> > > > 
> > > > # camcontrol inquiry da0
> > > > pass0: <Sony MSC-U04 3.00> Removable Direct Access SCSI-0 device 
> > > > pass0: Serial Number 3
> > > > pass0: 40.000MB/s transfers 
> > > > #
> > > 
> > > It's _removeable_ direct access.
> > 
> > oops, sorry. yes it is mentioned as removable.
> > 
> > > Since this is a memory stick - I don't know how one could remove the
> > > flash media from the USB drive.
> > 
> > Just mechanically.
> > 
> > > But this device bug won't hurt as long as you can remove the drive.
> > 
> > I can remove even mounted device. So it is not problem.
> 
> So in fact it is a flash card reader?
> In that case removeable is correct.
> You shouldn't eject mounted media - FreeBSD locks mounted media, but
> since your device can't physically lock it...
> 
> > Problem is that I should do something after replace card by another card
> > of different size to be able mount new card.
> 
> Yes - as already mentioned on the list by diffrent people:
> cp /dev/null > /dev/da0

Yes, this works also, I've already relayed about it. 

> > camcontrol detach da0; camcontrol rescan all 
> > helps, but, it should be much better if it will be issued automatically.
> 
> Yes - GEOM seems to ignore media change signals from drives.
> I've added PHK to the recipient list - maybe he has an idea about this
> problem.

I am even not sure Is there such signal.

> However - lots of cheap flash card readers are to stupid for even that.

This is on-board in my notebook. But m$win detects disconnection.

> You just can't rely on an automatic reread.

I already do mount/umount and also should do redetect ...

> > > > # usbdevs -v
> > > > ...
> > > > Controller /dev/usb3:
> > > > addr 1: high speed, self powered, config 1, EHCI root hub(0x0000),
> > > > Intel(0x0000), rev 1.00
> > > >  port 1 powered
> > > >  port 2 powered
> > > >  port 3 powered
> > > >  port 4 powered
> > > >  port 5 addr 2: high speed, self powered, config 1, USB Memory Stick
> > > > Slot(0x014d), Sony(0x054c), rev 1.10
> > > >  port 6 powered
> > > > #
> > > > 
> > > > I guess usb device should be removed on skick removal and back, but it
> > > > does not works :(.
> > > 
> > > So you say that the device still exists after unplugging the USB stick?
> > 
> > Yes.
> 
> That shouldn't happen, unless you still have references to that device.
> Are you running with latest code?

Latest does not build due to error in ndis. 
Now I have code of 24 March 05.

> There were lots of ehci fixes lately.

I've seen commits, But not sure that they are related - USB device stays
connected even when stick itself removed.

> > > This is not the intended behavour. Is it still mounted when you unplug?
> > 
> > No. (If be more exact it does not depends on mount state).
> > USB device never disappear. Probably it reflect real state, I guess,
> > card-reader itself hardwired to USB hub inside notebook and never
> > disconnected.
> > 
> > Probably there is kind of disconnect notification on CAM level ? 
> 
> The first disconnect notification is from your root hub.
> The uhub driver then destroys the umass, which first destroys the
> scbus and so on.
> If one instance on the chain blocks the whole thing has to wait.
> 
-- 
Vladimir B. Grebenchikov
vova at fbsd.ru