-Takashi Iwai and sound :
Still a slightly high amount of changes than wished, but they are all
good regression and/or device-specific fixes. Majority of commits are
for HD-audio, an HDMI ctl index fix that hits old graphics boards,
regression fixes for AD codecs and a few quirks. Other than that, two
major fixes are included: a 64bit ABI fix for compress offload, and
64bit dma_addr_t truncation fix, which had hit on PAE kernels.
-Vinod Koul has slave-dma fixes :
Here is the common fixes PULL for dmaengine. Dan has been working on fixing the
build issues in bunch of drivers. Here we have one fixing s3c24xx-dma, along
with fix from Russell on pl08x. Also we have Kuninori rcar dma fixes. The
s3c24xx-dma which was added in last merge window missed updates to usage of
DMA_COMPLETE so converting the last driver
-MTD fixes from Brian Norris :
Two MTD fixes, for the pxa3xx-nand driver:
– This driver was not ready to fully support Armada 370 NAND, with
particularly notable problems seen on flash with 2KB page sizes. This
“compatible” entry really should have been held back until 3.14 or
later.– Fix a bug seen in rare cases on the error path of a failed probe attempt,
where we free unallocated DMA resources
-Nicolas Ferre has additions to at91 :
This is a “drivers” series that adds Device Tree support to
Atmel crypto drivers (AES/[T]DES/SHA). As the DT part of this
addition is in at91-3.14-dt I thought it would be simpler to take
this series through arm-soc. I asked and got Herbert’s blessing.The Device Tree entries documented in these patches are very simple
and only declare the reg/irq values and the link to DMA.
-Mark Brown has some regulator fixes :
A small set of driver fixes plus one larger core change which changes
the way we check to see if we’re using DT so that there aren’t any races
between deciding we’re using DT and the regulator subsystem noticing.
This makes the new support for substituting a dummy regulator and
optional regulators work a lot better on DT systems since it ensures
that we don’t trigger probe deferral when we shouldn’t which was causing
bugs in clients.
