Merge branch 'for-3.18' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/libata

Pull libata update from Tejun Heo:
 "AHCI is getting per-port irq handling and locks for better
  scalability.  The gain is not huge but measureable with multiple high
  iops devices connected to the same host; however, the value of
  threaded IRQ handling seems negligible for AHCI and it likely will
  revert to non-threaded handling soon.

  Another noteworthy change is George Spelvin's "libata: Un-break ATA
  blacklist".  During 3.17 devel cycle, the libata blacklist glob
  matching got generalized and rewritten; unfortunately, the patch
  forgot to swap arguments to match the new match function and ended up
  breaking blacklist matching completely.  It got noticed only a couple
  days ago so it couldn't make for-3.17-fixes either.  :(

  Other than the above two, nothing too interesting - the usual cleanup
  churns and device-specific changes"

* 'for-3.18' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/libata: (22 commits)
  pata_serverworks: disable 64-KB DMA transfers on Broadcom OSB4 IDE Controller
  libata: Un-break ATA blacklist
  AHCI: Do not acquire ata_host::lock from single IRQ handler
  AHCI: Optimize single IRQ interrupt processing
  AHCI: Do not read HOST_IRQ_STAT reg in multi-MSI mode
  AHCI: Make few function names more descriptive
  AHCI: Move host activation code into ahci_host_activate()
  AHCI: Move ahci_host_activate() function to libahci.c
  AHCI: Pass SCSI host template as arg to ahci_host_activate()
  ata: pata_imx: Use the SIMPLE_DEV_PM_OPS() macro
  AHCI: Cleanup checking of multiple MSIs/SLM modes
  libata-sff: Fix controllers with no ctl port
  ahci_xgene: Fix the error print invalid resource for APM X-Gene SoC AHCI SATA Host Controller driver.
  libata: change ata_<foo>_printk routines to return void
  ata: qcom: Add device tree bindings information
  ahci-platform: Bump max number of clocks to 5
  ahci: ahci_p5wdh_workaround - constify DMI table
  libahci_platform: Staticize ahci_platform_<en/dis>able_phys()
  pata_platform: Remove useless irq_flags field
  pata_of_platform: Remove "electra-ide" quirk
  ...

1 files changed, 48 insertions, 0 deletions

diff --git a/Documentation/devicetree/bindings/ata/qcom-sata.txt b/Documentation/devicetree/bindings/ata/qcom-sata.txt
new file mode 100644
index 000000000000..094de91cd9fd
--- /dev/null
+++ b/Documentation/devicetree/bindings/ata/qcom-sata.txt
@@ -0,0 +1,48 @@
+* Qualcomm AHCI SATA Controller
+
+SATA nodes are defined to describe on-chip Serial ATA controllers.
+Each SATA controller should have its own node.
+
+Required properties:
+- compatible		: compatible list, must contain "generic-ahci"
+- interrupts		: <interrupt mapping for SATA IRQ>
+- reg			: <registers mapping>
+- phys			: Must contain exactly one entry as specified
+			  in phy-bindings.txt
+- phy-names		: Must be "sata-phy"
+
+Required properties for "qcom,ipq806x-ahci" compatible:
+- clocks		: Must contain an entry for each entry in clock-names.
+- clock-names		: Shall be:
+				"slave_iface" - Fabric port AHB clock for SATA
+				"iface" - AHB clock
+				"core" - core clock
+				"rxoob" - RX out-of-band clock
+				"pmalive" - Power Module Alive clock
+- assigned-clocks	: Shall be:
+				SATA_RXOOB_CLK
+				SATA_PMALIVE_CLK
+- assigned-clock-rates	: Shall be:
+				100Mhz (100000000) for SATA_RXOOB_CLK
+				100Mhz (100000000) for SATA_PMALIVE_CLK
+
+Example:
+	sata@29000000 {
+		compatible = "qcom,ipq806x-ahci", "generic-ahci";
+		reg = <0x29000000 0x180>;
+
+		interrupts = <0 209 0x0>;
+
+		clocks = <&gcc SFAB_SATA_S_H_CLK>,
+			 <&gcc SATA_H_CLK>,
+			 <&gcc SATA_A_CLK>,
+			 <&gcc SATA_RXOOB_CLK>,
+			 <&gcc SATA_PMALIVE_CLK>;
+		clock-names = "slave_iface", "iface", "core",
+				"rxoob", "pmalive";
+		assigned-clocks = <&gcc SATA_RXOOB_CLK>, <&gcc SATA_PMALIVE_CLK>;
+		assigned-clock-rates = <100000000>, <100000000>;
+
+		phys = <&sata_phy>;
+		phy-names = "sata-phy";
+	};