linux-toradex.git/drivers, branch v2.6.28.9

drm/i915: Don't allow objects to get bound while VT switched.

2009-03-23T21:55:32+00:00

commit 9bb2d6f94aeb9a185d69aedbd19421b6da4e3309 upstream.

This avoids a BUG_ON in the enter_vt path due to objects being in the GTT
when we shouldn't have ever let them be (as we're not supposed to touch the
device during that time).

This was triggered by a change in the 2D driver to use the GTT mapping of
objects after pinning them to improve software fallback performance.

Signed-off-by: Eric Anholt 
Signed-off-by: Dave Airlie 
Signed-off-by: Greg Kroah-Hartman

drm/i915: Don't print to dmesg when taking signal during object_pin.

2009-03-23T21:55:30+00:00

commit f1acec933848219c402c165686677b1c307407f8 upstream.

This showed up in logs where people had a hung chip, so pinning was blocked
on the chip unpinning other buffers, and the X Server took its scheduler
signal during that time.

Signed-off-by: Eric Anholt 
Signed-off-by: Dave Airlie 
Signed-off-by: Greg Kroah-Hartman

drm/i915: Don't double-unpin buffers if we take a signal in evict_everything().

2009-03-23T21:55:30+00:00

commit b117763627ef4d24086801dd5f74c9eb2f487790 upstream.

We haven't seen this in practice, but it was visible when looking at a bug
report from when i915_gem_evict_everything() was broken and would always
return error.

Signed-off-by: Eric Anholt 
Signed-off-by: Dave Airlie 
Signed-off-by: Greg Kroah-Hartman

drm/i915: don't enable vblanks on disabled pipes

2009-03-23T21:55:30+00:00

commit 71e0ffa599f54058d9b8724b4b14d0486751681d upstream.

In some cases userland may be confused and try to wait on vblank events from
pipes that aren't actually enabled.  We shouldn't allow this, so return
-EINVAL if the pipe isn't on.

Signed-off-by: Jesse Barnes 
Signed-off-by: Eric Anholt 
Signed-off-by: Dave Airlie 
Signed-off-by: Greg Kroah-Hartman

drm/i915: set vblank enabled flag correctly across IRQ install/uninstall

2009-03-23T21:55:30+00:00

commit dc1336ff4fe08ae7cfe8301bfd7f0b2cfd31d20a upstream.

In the absence of kernel mode setting, many drivers disable IRQs across VT
switch.  The core DRM vblank code is missing a check for this case however;
even after IRQ disable, the vblank code will still have the vblank_enabled
flag set, so unless we track the fact that they're disabled at IRQ uninstall
time, when we VT switch back in we won't actually re-enable them, which means
any apps waiting on vblank before the switch will hang.

This patch does that and also adds a sanity check to the wait condition to
look for the irq_enabled flag in general, as well as adding a wakeup to the
IRQ uninstall path.

Fixes fdo bug #18879 with compiz hangs at VT switch.

Signed-off-by: Jesse Barnes 
Signed-off-by: Eric Anholt 
Signed-off-by: Dave Airlie 
Signed-off-by: Greg Kroah-Hartman

USB: EHCI: Fix isochronous URB leak

2009-03-23T21:55:29+00:00

commit 508db8c954d55ed30f870d2c24d741ba6269d13c upstream.

ehci-hcd uses usb_get_urb() and usb_put_urb() in an unbalanced way causing
isochronous URB's kref.counts incrementing once per usb_submit_urb() call.
The culprit is *usb being set to NULL when usb_put_urb() is called after URB
is given back.
Due to other fixes there is no need for ehci-hcd to deal with usb_get_urb()
nor usb_put_urb() anymore, so patch removes their usages in ehci-hcd.
Patch also makes ehci_to_hcd(ehci)->self.bandwidth_allocated adjust, if a
stream finishes.

Signed-off-by: Karsten Wiese 
Cc: David Brownell 
Signed-off-by: Greg Kroah-Hartman

USB: EHCI: expedite unlinks when the root hub is suspended

2009-03-23T21:55:29+00:00

commit 391016f6e2fe3b9979b4c6880a76e5e434d6947c upstream.

This patch (as1225) fixes a bug in ehci-hcd.  The condition for
whether unlinked QHs can become IDLE should not be that the controller
is halted, but rather that the controller isn't running.  In other
words when the root hub is suspended, the hardware doesn't own any
QHs.

This fixes a problem that can show up during hibernation: If a QH is
only partially unlinked when the root hub is frozen, then when the
root hub is thawed the QH won't be in the IDLE state.  As a result it
can't be used properly for new URB submissions.

Signed-off-by: Alan Stern 
Reported-by: Brandon Philips 
Tested-by: Brandon Philips 
Acked-by: David Brownell 
Signed-off-by: Greg Kroah-Hartman

USB: usbfs: keep async URBs until the device file is closed

2009-03-23T21:55:29+00:00

commit 6ff10464096540e14d7575a72c50d0316d003714 upstream.

The usbfs driver manages a list of completed asynchronous URBs.  But
it is too eager to free the entries on this list: destroy_async() gets
called whenever an interface is unbound or a device is removed, and it
deallocates the outstanding struct async entries for all URBs on that
interface or device.  This is wrong; the user program should be able
to reap an URB any time after it has completed, regardless of whether
or not the interface is still bound or the device is still present.

This patch (as1222) moves the code for deallocating the completed list
entries from destroy_async() to usbdev_release().  The outstanding
entries won't be freed until the user program has closed the device
file, thereby eliminating any possibility that the remaining URBs
might still be reaped.

This fixes a bug in which a program can hang in the USBDEVFS_REAPURB
ioctl when the device is unplugged.

Reported-and-tested-by: Martin Poupe 
Signed-off-by: Alan Stern 
Signed-off-by: Greg Kroah-Hartman

USB: usbtmc: add protocol 1 support

2009-03-23T21:55:29+00:00

commit 228dd05dbfdd0fced8ab1a28ed73b500ba6bb0a6 upstream.

The driver already supports the 1 protocol support, so just add it to
the MODULE_DEVICE_TABLE entry so it properly picks up these devices.

Thanks to Jouni Rynö for pointing this out.

Reported-by: Jouni Ryno 
Signed-off-by: Greg Kroah-Hartman

USB: usbtmc: fix stupid bug in open()

2009-03-23T21:55:29+00:00

commit 5b10916ea0a62920204517e1c4ce14560b4f96ab upstream.

open() will never succeed, as we always return -ENODEV.  Fix this
obvious bug.

Thanks to Jouni Ryno for reporting it.

Reported-by: Jouni Ryno 
Signed-off-by: Greg Kroah-Hartman