spi/spi-bcm2708: respect per-transfer SPI clock speed_hz value

The bcm2708 SPI driver's bcm2708_process_transfer() was ignoring the
per-transfer speed_hz value even when it was provided (it always just
used the spi device's max_speed_hz value).  Now, per-transfer speed_hz
values are respected.

Also added debug print to bcm2708_setup_state() to help keep an eye on
the configured SPI parameters.

Signed-off-by: Kamal Mostafa <kamal@whence.com>

Conflicts:
	drivers/usb/host/dwc_otg/dwc_otg_cil.c
	drivers/usb/host/dwc_otg/dwc_otg_hcd_intr.c

This checkin includes some debug, but its mainly trying to force transactions
to start on frame .7 so that they complete correctly before the end of the
frame.

Update Synopsys USB OTG driver to v2.94a and disable CRYPTOLIB. Third attempt

Revert "Add mechanism to reduce the number of SOF interrupts in dwc_otg USB driver. Enable through /proc/dwc_sof/SOF_reduction"

This reverts commit 85b78218.

The TIMER_PERIOD should be calculated using the timer interrupt frequenc...

Add module parameter for missing_status quirk. sdhci-bcm2708.missing_status=0 may improve interrupt latency

sdhci-bcm2708 enhancements (possible fix for dataloss bug)

There are issues with both single block reads (missed completion)
and writes (data loss in some cases!). Just don't do single block
transfers anymore, and treat them like multiblock transfers. This
adds a quirk for this and uses it.

The additional FIFO might speed up transfers in some cases.

The calculated values somehow don't agree with the sched_clock code and
we end up with warnings like:

sched_clock: wrong multiply/shift: 2097152000>>21 vs calculated 4194304000>>22
sched_clock: fix multiply/shift to avoid scheduler hiccups

So use the constant values much like arch/arm/mach-tegra/timer.c does.

Signed-off-by: Chris Boot <bootc@bootc.net>

Signed-off-by: Chris Boot <bootc@bootc.net>

Conflicts:

	arch/arm/Kconfig.debug
	arch/arm/mach-bcm2708/include/mach/memory.h

Signed-off-by: Chris Boot <bootc@bootc.net>

Not sure what the original code was trying to do as it was completely
wrong on many levels. This patch fixes the macro to return the correct
physical and virtual addresses of the PL011 UART on the RPi.

Note that you need to boot the compressed kernel (zImage) so that the
UART is configured at boot, or your kernel will hang when it tries to
access the UART.

Signed-off-by: Chris Boot <bootc@bootc.net>

Signed-off-by: Chris Boot <bootc@bootc.net>

This was added for the DWC OTG driver in commit d5ef856b

. As we've
removed the dead code from that driver that depends on this, it is no
longer required at all.

Signed-off-by: Chris Boot <bootc@bootc.net>

The dwc_common_port library used by the dwc_otg includes bignumber and
crypto functions which require 64x64 multiplication functions. Remove
this dead code.

Signed-off-by: Chris Boot <bootc@bootc.net>

This reverts commit 7e769589.

Better fix for quitting X hang. Interrupted mailbox reads are not something we can reliably recover from, so down_interruptable is not a safe call.

Change to add SCSI tape support as loadable modules to kernel config after line 264 add these two lines: CONFIG_CHR_DEV_ST=m CONFIG_CHR_DEV_OSST=m

commit 052b1987

 upstream.

When swapon() was not passed the SWAP_FLAG_DISCARD option, sys_swapon()
will still perform a discard operation.  This can cause problems if
discard is slow or buggy.

Reverse the order of the check so that a discard operation is performed
only if the sys_swapon() caller is attempting to enable discard.

Signed-off-by: Shaohua Li <shli@fusionio.com>
Reported-by: Holger Kiehl <Holger.Kiehl@dwd.de>
Tested-by: Holger Kiehl <Holger.Kiehl@dwd.de>
Cc: Hugh Dickins <hughd@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>