Merge tag 'xtensa-20121218' of git://github.com/czankel/xtensa-linux

Pull Xtensa patchset from Chris Zankel:
 "This contains support of device trees, many fixes, and code clean-ups"

* tag 'xtensa-20121218' of git://github.com/czankel/xtensa-linux: (33 commits)
  xtensa: don't try to build DTB when OF is disabled
  xtensa: set the correct ethernet address for xtfpga
  xtensa: clean up files to make them code-style compliant
  xtensa: provide endianness macro for sparse
  xtensa: fix RASID SR initialization
  xtensa: initialize CPENABLE SR when core has one
  xtensa: reset all timers on initialization
  Use for_each_compatible_node() macro.
  xtensa: add XTFPGA DTS
  xtensa: add support for the XTFPGA boards
  xtensa: add device trees support
  xtensa: add IRQ domains support
  xtensa: add U-Boot image support (uImage).
  xtensa: clean up boot make rules
  xtensa: fix mb and wmb definitions
  xtensa: add s32c1i-based spinlock implementations
  xtensa: add s32c1i-based bitops implementations
  xtensa: add s32c1i-based atomic ops implementations
  xtensa: add s32c1i sanity check
  xtensa: add trap_set_handler function
  ...

Documentation/xtensa/atomctl.txt

+We Have Atomic Operation Control (ATOMCTL) Register.
+This register determines the effect of using a S32C1I instruction
+with various combinations of:
+
+     1. With and without an Coherent Cache Controller which
+        can do Atomic Transactions to the memory internally.
+
+     2. With and without An Intelligent Memory Controller which
+        can do Atomic Transactions itself.
+
+The Core comes up with a default value of for the three types of cache ops:
+
+      0x28: (WB: Internal, WT: Internal, BY:Exception)
+
+On the FPGA Cards we typically simulate an Intelligent Memory controller
+which can implement  RCW transactions. For FPGA cards with an External
+Memory controller we let it to the atomic operations internally while
+doing a Cached (WB) transaction and use the Memory RCW for un-cached
+operations.
+
+For systems without an coherent cache controller, non-MX, we always
+use the memory controllers RCW, thought non-MX controlers likely
+support the Internal Operation.
+
+CUSTOMER-WARNING:
+   Virtually all customers buy their memory controllers from vendors that
+   don't support atomic RCW memory transactions and will likely want to
+   configure this register to not use RCW.
+
+Developers might find using RCW in Bypass mode convenient when testing
+with the cache being bypassed; for example studying cache alias problems.
+
+See Section 4.3.12.4 of ISA; Bits:
+
+                             WB     WT      BY
+                           5   4 | 3   2 | 1   0
+  2 Bit
+  Field
+  Values     WB - Write Back         WT - Write Thru         BY - Bypass
+---------    ---------------         -----------------     ----------------
+    0        Exception               Exception               Exception
+    1        RCW Transaction         RCW Transaction         RCW Transaction
+    2        Internal Operation      Exception               Reserved
+    3        Reserved                Reserved                Reserved

arch/xtensa/Kconfig

@@ -17,6 +17,7 @@ config XTENSA
 	select GENERIC_KERNEL_EXECVE
 	select ARCH_WANT_OPTIONAL_GPIOLIB
 	select CLONE_BACKWARDS
+	select IRQ_DOMAIN
 	help
 	  Xtensa processors are 32-bit RISC machines designed by Tensilica
 	  primarily for embedded systems.  These processors are both
@@ -150,6 +151,15 @@ config XTENSA_PLATFORM_S6105
 	select SERIAL_CONSOLE
 	select NO_IOPORT
 
+config XTENSA_PLATFORM_XTFPGA
+	bool "XTFPGA"
+	select SERIAL_CONSOLE
+	select ETHOC
+	select XTENSA_CALIBRATE_CCOUNT
+	help
+	  XTFPGA is the name of Tensilica board family (LX60, LX110, LX200, ML605).
+	  This hardware is capable of running a full Linux distribution.
+
 endchoice
 
 
@@ -177,6 +187,17 @@ config CMDLINE
 	  time by entering them here. As a minimum, you should specify the
 	  memory size and the root device (e.g., mem=64M root=/dev/nfs).
 
+config USE_OF
+	bool "Flattened Device Tree support"
+	select OF
+	select OF_EARLY_FLATTREE
+	help
+	  Include support for flattened device tree machine descriptions.
+
+config BUILTIN_DTB
+	string "DTB to build into the kernel image"
+	depends on OF
+
 source "mm/Kconfig"
 
 source "drivers/pcmcia/Kconfig"

arch/xtensa/Kconfig.debug

@@ -2,6 +2,26 @@ menu "Kernel hacking"
 
 source "lib/Kconfig.debug"
 
-endmenu
+config LD_NO_RELAX
+	bool "Disable linker relaxation"
+	default n
+	help
+	  Enable this function to disable link-time optimizations.
+	  The default linker behavior is to combine identical literal
+	  values to reduce code size and remove unnecessary overhead from
+	  assembler-generated 'longcall' sequences.
+	  Enabling this option improves the link time but increases the
+	  code size, and possibly execution time.
+
+config S32C1I_SELFTEST
+	bool "Perform S32C1I instruction self-test at boot"
+	default y
+	help
+	  Enable this option to test S32C1I instruction behavior at boot.
+	  Correct operation of this instruction requires some cooperation from hardware
+	  external to the processor (such as bus bridge, bus fabric, or memory controller).
+	  It is easy to make wrong hardware configuration, this test should catch it early.
 
+	  Say 'N' on stable hardware.
 
+endmenu

arch/xtensa/Makefile

@@ -38,6 +38,7 @@ endif
 platform-$(CONFIG_XTENSA_PLATFORM_XT2000)	:= xt2000
 platform-$(CONFIG_XTENSA_PLATFORM_ISS)		:= iss
 platform-$(CONFIG_XTENSA_PLATFORM_S6105)	:= s6105
+platform-$(CONFIG_XTENSA_PLATFORM_XTFPGA)	:= xtfpga
 
 PLATFORM = $(platform-y)
 export PLATFORM
@@ -49,6 +50,17 @@ KBUILD_CFLAGS += -pipe -mlongcalls
 
 KBUILD_CFLAGS += $(call cc-option,-mforce-no-pic,)
 
+ifneq ($(CONFIG_LD_NO_RELAX),)
+LDFLAGS := --no-relax
+endif
+
+ifeq ($(shell echo -e __XTENSA_EB__ | $(CC) -E - | grep -v "\#"),1)
+CHECKFLAGS += -D__XTENSA_EB__
+endif
+ifeq ($(shell echo -e __XTENSA_EL__ | $(CC) -E - | grep -v "\#"),1)
+CHECKFLAGS += -D__XTENSA_EL__
+endif
+
 vardirs := $(patsubst %,arch/xtensa/variants/%/,$(variant-y))
 plfdirs := $(patsubst %,arch/xtensa/platforms/%/,$(platform-y))
 
@@ -75,6 +87,10 @@ core-y		+= $(buildvar) $(buildplf)
 
 libs-y		+= arch/xtensa/lib/ $(LIBGCC)
 
+ifneq ($(CONFIG_BUILTIN_DTB),"")
+core-$(CONFIG_OF) += arch/xtensa/boot/
+endif
+
 boot		:= arch/xtensa/boot
 
 all: zImage
@@ -84,7 +100,9 @@ bzImage : zImage
 zImage: vmlinux
 	$(Q)$(MAKE) $(build)=$(boot) $@
 
+%.dtb:
+	$(Q)$(MAKE) $(build)=$(boot) $(boot)/$@
+
 define archhelp
   @echo '* zImage      - Compressed kernel image (arch/xtensa/boot/images/zImage.*)'
 endef
-

arch/xtensa/boot/Makefile

@@ -22,12 +22,35 @@ subdir-y	:= lib
 # Subdirs for the boot loader(s)
 
 bootdir-$(CONFIG_XTENSA_PLATFORM_ISS)	 += boot-elf
-bootdir-$(CONFIG_XTENSA_PLATFORM_XT2000) += boot-redboot boot-elf
+bootdir-$(CONFIG_XTENSA_PLATFORM_XT2000) += boot-redboot boot-elf boot-uboot
+bootdir-$(CONFIG_XTENSA_PLATFORM_XTFPGA) += boot-redboot boot-elf boot-uboot
 
 
+BUILTIN_DTB := $(patsubst "%",%,$(CONFIG_BUILTIN_DTB)).dtb.o
+ifneq ($(CONFIG_BUILTIN_DTB),"")
+obj-$(CONFIG_OF) += $(BUILTIN_DTB)
+endif
+
+# Rule to build device tree blobs
+$(obj)/%.dtb: $(src)/dts/%.dts FORCE
+	$(call if_changed_dep,dtc)
+
+clean-files := *.dtb.S
+
 zImage Image: $(bootdir-y)
 
 $(bootdir-y): $(addprefix $(obj)/,$(subdir-y)) \
 	      $(addprefix $(obj)/,$(host-progs))
 	$(Q)$(MAKE) $(build)=$(obj)/$@ $(MAKECMDGOALS)
 
+OBJCOPYFLAGS = --strip-all -R .comment -R .note.gnu.build-id -O binary
+
+vmlinux.bin: vmlinux FORCE
+	$(call if_changed,objcopy)
+
+vmlinux.bin.gz: vmlinux.bin FORCE
+	$(call if_changed,gzip)
+
+boot-elf: vmlinux.bin
+boot-redboot: vmlinux.bin.gz
+boot-uboot: vmlinux.bin.gz

arch/xtensa/boot/boot-elf/Makefile

@@ -4,9 +4,6 @@
 # for more details.
 #
 
-GZIP = gzip
-GZIP_FLAGS = -v9fc
-
 ifeq ($(BIG_ENDIAN),1)
 OBJCOPY_ARGS    := -O elf32-xtensa-be
 else
@@ -20,18 +17,17 @@ boot-y		:= bootstrap.o
 
 OBJS		:= $(addprefix $(obj)/,$(boot-y))
 
-vmlinux.tmp: vmlinux
-	$(OBJCOPY) --strip-all -R .comment -R .note.gnu.build-id -O binary \
-		$^ $@
-
-Image:	vmlinux.tmp $(OBJS) arch/$(ARCH)/boot/boot-elf/boot.lds
-	$(OBJCOPY) $(OBJCOPY_ARGS) -R .comment \
-		--add-section image=vmlinux.tmp \
+$(obj)/Image.o: vmlinux.bin $(OBJS)
+	$(Q)$(OBJCOPY) $(OBJCOPY_ARGS) -R .comment \
+		--add-section image=vmlinux.bin \
 		--set-section-flags image=contents,alloc,load,load,data \
-		$(OBJS) $@.tmp
-	$(LD) $(LDFLAGS) $(LDFLAGS_vmlinux) \
-		-T arch/$(ARCH)/boot/boot-elf/boot.lds \
-		-o arch/$(ARCH)/boot/$@.elf $@.tmp
+		$(OBJS) $@
 
-zImage:	Image
+$(obj)/../Image.elf: $(obj)/Image.o $(obj)/boot.lds
+	$(Q)$(LD) $(LDFLAGS) $(LDFLAGS_vmlinux) \
+		-T $(obj)/boot.lds \
+		--build-id=none \
+		-o $@ $(obj)/Image.o
+	$(Q)$(kecho) '  Kernel: $@ is ready'
 
+zImage:	$(obj)/../Image.elf

arch/xtensa/boot/boot-redboot/Makefile

@@ -4,8 +4,6 @@
 # for more details.
 #
 
-GZIP = gzip
-GZIP_FLAGS = -v9fc
 ifeq ($(BIG_ENDIAN),1)
 OBJCOPY_ARGS 	:= -O elf32-xtensa-be
 else
@@ -21,17 +19,17 @@ LIBS	:= arch/xtensa/boot/lib/lib.a arch/xtensa/lib/lib.a
 
 LIBGCC := $(shell $(CC) $(KBUILD_CFLAGS) -print-libgcc-file-name)
 
-vmlinux.tmp: vmlinux
-	$(OBJCOPY) --strip-all -R .comment -R .note.gnu.build-id -O binary \
-		$^ $@
+$(obj)/zImage.o: vmlinux.bin.gz $(OBJS)
+	$(Q)$(OBJCOPY) $(OBJCOPY_ARGS) -R .comment \
+		--add-section image=vmlinux.bin.gz \
+		--set-section-flags image=contents,alloc,load,load,data \
+		$(OBJS) $@
 
-vmlinux.tmp.gz: vmlinux.tmp
-	$(GZIP) $(GZIP_FLAGS) $^ > $@
+$(obj)/zImage.elf: $(obj)/zImage.o $(LIBS)
+	$(Q)$(LD) $(LD_ARGS) -o $@ $^ -L/xtensa-elf/lib $(LIBGCC)
 
-zImage: vmlinux.tmp.gz $(OBJS) $(LIBS)
-	$(OBJCOPY) $(OBJCOPY_ARGS) -R .comment \
-		--add-section image=vmlinux.tmp.gz \
-		--set-section-flags image=contents,alloc,load,load,data \
-		$(OBJS) $@.tmp
-	$(LD) $(LD_ARGS) -o $@.elf $@.tmp $(LIBS) -L/xtensa-elf/lib $(LIBGCC)
-	$(OBJCOPY) -S -O binary $@.elf arch/$(ARCH)/boot/$@.redboot
+$(obj)/../zImage.redboot: $(obj)/zImage.elf
+	$(Q)$(OBJCOPY) -S -O binary $< $@
+	$(Q)$(kecho) '  Kernel: $@ is ready'
+
+zImage: $(obj)/../zImage.redboot

arch/xtensa/boot/boot-uboot/Makefile

+#
+# This file is subject to the terms and conditions of the GNU General Public
+# License.  See the file "COPYING" in the main directory of this archive
+# for more details.
+#
+
+UIMAGE_LOADADDR = 0xd0001000
+UIMAGE_COMPRESSION = gzip
+
+$(obj)/../uImage: vmlinux.bin.gz FORCE
+	$(call if_changed,uimage)
+	$(Q)$(kecho) '  Kernel: $@ is ready'
+
+zImage: $(obj)/../uImage

arch/xtensa/boot/dts/lx60.dts

+/dts-v1/;
+/include/ "xtfpga.dtsi"
+/include/ "xtfpga-flash-4m.dtsi"
+
+/ {
+	compatible = "xtensa,lx60";
+	memory@0 {
+		device_type = "memory";
+		reg = <0x00000000 0x04000000>;
+	};
+};

arch/xtensa/boot/dts/ml605.dts

+/dts-v1/;
+/include/ "xtfpga.dtsi"
+/include/ "xtfpga-flash-16m.dtsi"
+
+/ {
+	compatible = "xtensa,ml605";
+	memory@0 {
+		device_type = "memory";
+		reg = <0x00000000 0x08000000>;
+	};
+};

arch/xtensa/boot/dts/xtfpga-flash-16m.dtsi

+/ {
+	flash: flash@f8000000 {
+		#address-cells = <1>;
+		#size-cells = <1>;
+		compatible = "cfi-flash";
+		reg = <0xf8000000 0x01000000>;
+		bank-width = <2>;
+		device-width = <2>;
+		partition@0x0 {
+			label = "boot loader area";
+			reg = <0x00000000 0x00400000>;
+		};
+		partition@0x400000 {
+			label = "kernel image";
+			reg = <0x00400000 0x00600000>;
+		};
+		partition@0xa00000 {
+			label = "data";
+			reg = <0x00a00000 0x005e0000>;
+		};
+		partition@0xfe0000 {
+			label = "boot environment";
+			reg = <0x00fe0000 0x00020000>;
+		};
+        };
+};

arch/xtensa/boot/dts/xtfpga-flash-4m.dtsi

+/ {
+	flash: flash@f8000000 {
+		#address-cells = <1>;
+		#size-cells = <1>;
+		compatible = "cfi-flash";
+		reg = <0xf8000000 0x00400000>;
+		bank-width = <2>;
+		device-width = <2>;
+		partition@0x0 {
+			label = "boot loader area";
+			reg = <0x00000000 0x003f0000>;
+		};
+		partition@0x3f0000 {
+			label = "boot environment";
+			reg = <0x003f0000 0x00010000>;
+		};
+        };
+};

arch/xtensa/boot/dts/xtfpga.dtsi

+/ {
+	compatible = "xtensa,xtfpga";
+	#address-cells = <1>;
+	#size-cells = <1>;
+	interrupt-parent = <&pic>;
+
+	chosen {
+		bootargs = "earlycon=uart8250,mmio32,0xfd050020,115200n8 console=ttyS0,115200n8 ip=dhcp root=/dev/nfs rw debug";
+	};
+
+	memory@0 {
+		device_type = "memory";
+		reg = <0x00000000 0x06000000>;
+	};
+
+	cpus {
+		#address-cells = <1>;
+		#size-cells = <0>;
+		cpu@0 {
+			compatible = "xtensa,cpu";
+			reg = <0>;
+			/* Filled in by platform_setup from FPGA register
+			 * clock-frequency = <100000000>;
+			 */
+		};
+	};
+
+	pic: pic {
+		compatible = "xtensa,pic";
+		/* one cell: internal irq number,
+		 * two cells: second cell == 0: internal irq number
+		 *            second cell == 1: external irq number
+		 */
+		#interrupt-cells = <2>;
+		interrupt-controller;
+	};
+
+	serial0: serial@fd050020 {
+		device_type = "serial";
+		compatible = "ns16550a";
+		no-loopback-test;
+		reg = <0xfd050020 0x20>;
+		reg-shift = <2>;
+		interrupts = <0 1>; /* external irq 0 */
+		/* Filled in by platform_setup from FPGA register
+		 * clock-frequency = <100000000>;
+		 */
+	};
+
+	enet0: ethoc@fd030000 {
+		compatible = "opencores,ethoc";
+		reg = <0xfd030000 0x4000 0xfd800000 0x4000>;
+		interrupts = <1 1>; /* external irq 1 */
+		local-mac-address = [00 50 c2 13 6f 00];
+	};
+};

arch/xtensa/include/asm/atomic.h

@@ -66,19 +66,35 @@
  */
 static inline void atomic_add(int i, atomic_t * v)
 {
+#if XCHAL_HAVE_S32C1I
+	unsigned long tmp;
+	int result;
+
+	__asm__ __volatile__(
+			"1:     l32i    %1, %3, 0\n"
+			"       wsr     %1, scompare1\n"
+			"       add     %0, %1, %2\n"
+			"       s32c1i  %0, %3, 0\n"
+			"       bne     %0, %1, 1b\n"
+			: "=&a" (result), "=&a" (tmp)
+			: "a" (i), "a" (v)
+			: "memory"
+			);
+#else
 	unsigned int vval;
 
 	__asm__ __volatile__(
-	"rsil    a15, "__stringify(LOCKLEVEL)"\n\t"
-	"l32i    %0, %2, 0              \n\t"
-	"add     %0, %0, %1             \n\t"
-	"s32i    %0, %2, 0              \n\t"
-	"wsr     a15, ps                \n\t"
-	"rsync                          \n"
+			"       rsil    a15, "__stringify(LOCKLEVEL)"\n"
+			"       l32i    %0, %2, 0\n"
+			"       add     %0, %0, %1\n"
+			"       s32i    %0, %2, 0\n"
+			"       wsr     a15, ps\n"
+			"       rsync\n"
 			: "=&a" (vval)
 			: "a" (i), "a" (v)
 			: "a15", "memory"
 			);
+#endif
 }
 
 /**
@@ -90,19 +106,35 @@ static inline void atomic_add(int i, atomic_t * v)
  */
 static inline void atomic_sub(int i, atomic_t *v)
 {
+#if XCHAL_HAVE_S32C1I
+	unsigned long tmp;
+	int result;
+
+	__asm__ __volatile__(
+			"1:     l32i    %1, %3, 0\n"
+			"       wsr     %1, scompare1\n"
+			"       sub     %0, %1, %2\n"
+			"       s32c1i  %0, %3, 0\n"
+			"       bne     %0, %1, 1b\n"
+			: "=&a" (result), "=&a" (tmp)
+			: "a" (i), "a" (v)
+			: "memory"
+			);
+#else
 	unsigned int vval;
 
 	__asm__ __volatile__(
-	"rsil    a15, "__stringify(LOCKLEVEL)"\n\t"
-	"l32i    %0, %2, 0              \n\t"
-	"sub     %0, %0, %1             \n\t"
-	"s32i    %0, %2, 0              \n\t"
-	"wsr     a15, ps                \n\t"
-	"rsync                          \n"
+			"       rsil    a15, "__stringify(LOCKLEVEL)"\n"
+			"       l32i    %0, %2, 0\n"
+			"       sub     %0, %0, %1\n"
+			"       s32i    %0, %2, 0\n"
+			"       wsr     a15, ps\n"
+			"       rsync\n"
 			: "=&a" (vval)
 			: "a" (i), "a" (v)
 			: "a15", "memory"
 			);
+#endif
 }
 
 /*
@@ -111,40 +143,78 @@ static inline void atomic_sub(int i, atomic_t *v)
 
 static inline int atomic_add_return(int i, atomic_t * v)
 {
+#if XCHAL_HAVE_S32C1I
+	unsigned long tmp;
+	int result;
+
+	__asm__ __volatile__(
+			"1:     l32i    %1, %3, 0\n"
+			"       wsr     %1, scompare1\n"
+			"       add     %0, %1, %2\n"
+			"       s32c1i  %0, %3, 0\n"
+			"       bne     %0, %1, 1b\n"
+			"       add     %0, %0, %2\n"
+			: "=&a" (result), "=&a" (tmp)
+			: "a" (i), "a" (v)
+			: "memory"
+			);
+
+	return result;
+#else
 	unsigned int vval;
 
 	__asm__ __volatile__(
-	"rsil    a15,"__stringify(LOCKLEVEL)"\n\t"
-	"l32i    %0, %2, 0             \n\t"
-	"add     %0, %0, %1            \n\t"
-	"s32i    %0, %2, 0             \n\t"
-	"wsr     a15, ps               \n\t"
-	"rsync                         \n"
+			"       rsil    a15,"__stringify(LOCKLEVEL)"\n"
+			"       l32i    %0, %2, 0\n"
+			"       add     %0, %0, %1\n"
+			"       s32i    %0, %2, 0\n"
+			"       wsr     a15, ps\n"
+			"       rsync\n"
 			: "=&a" (vval)
 			: "a" (i), "a" (v)
 			: "a15", "memory"
 			);
 
 	return vval;
+#endif
 }
 
 static inline int atomic_sub_return(int i, atomic_t * v)
 {
+#if XCHAL_HAVE_S32C1I
+	unsigned long tmp;
+	int result;
+
+	__asm__ __volatile__(
+			"1:     l32i    %1, %3, 0\n"
+			"       wsr     %1, scompare1\n"
+			"       sub     %0, %1, %2\n"
+			"       s32c1i  %0, %3, 0\n"
+			"       bne     %0, %1, 1b\n"
+			"       sub     %0, %0, %2\n"
+			: "=&a" (result), "=&a" (tmp)
+			: "a" (i), "a" (v)
+			: "memory"
+			);
+
+	return result;
+#else
 	unsigned int vval;
 
 	__asm__ __volatile__(
-	"rsil    a15,"__stringify(LOCKLEVEL)"\n\t"
-	"l32i    %0, %2, 0             \n\t"
-	"sub     %0, %0, %1            \n\t"
-	"s32i    %0, %2, 0             \n\t"
-	"wsr     a15, ps               \n\t"
-	"rsync                         \n"
+			"       rsil    a15,"__stringify(LOCKLEVEL)"\n"
+			"       l32i    %0, %2, 0\n"
+			"       sub     %0, %0, %1\n"
+			"       s32i    %0, %2, 0\n"
+			"       wsr     a15, ps\n"
+			"       rsync\n"
 			: "=&a" (vval)
 			: "a" (i), "a" (v)
 			: "a15", "memory"
 			);
 
 	return vval;
+#endif
 }
 
 /**
@@ -251,38 +321,70 @@ static __inline__ int __atomic_add_unless(atomic_t *v, int a, int u)
 
 static inline void atomic_clear_mask(unsigned int mask, atomic_t *v)
 {
+#if XCHAL_HAVE_S32C1I
+	unsigned long tmp;
+	int result;
+
+	__asm__ __volatile__(
+			"1:     l32i    %1, %3, 0\n"
+			"       wsr     %1, scompare1\n"
+			"       and     %0, %1, %2\n"
+			"       s32c1i  %0, %3, 0\n"
+			"       bne     %0, %1, 1b\n"
+			: "=&a" (result), "=&a" (tmp)
+			: "a" (~mask), "a" (v)
+			: "memory"
+			);
+#else
 	unsigned int all_f = -1;
 	unsigned int vval;
 
 	__asm__ __volatile__(
-	"rsil    a15,"__stringify(LOCKLEVEL)"\n\t"
-	"l32i    %0, %2, 0             \n\t"
-	"xor     %1, %4, %3            \n\t"
-	"and     %0, %0, %4            \n\t"
-	"s32i    %0, %2, 0             \n\t"
-	"wsr     a15, ps               \n\t"
-	"rsync                         \n"
+			"       rsil    a15,"__stringify(LOCKLEVEL)"\n"
+			"       l32i    %0, %2, 0\n"
+			"       xor     %1, %4, %3\n"
+			"       and     %0, %0, %4\n"
+			"       s32i    %0, %2, 0\n"
+			"       wsr     a15, ps\n"
+			"       rsync\n"
 			: "=&a" (vval), "=a" (mask)
 			: "a" (v), "a" (all_f), "1" (mask)
 			: "a15", "memory"
 			);
+#endif
 }
 
 static inline void atomic_set_mask(unsigned int mask, atomic_t *v)
 {
+#if XCHAL_HAVE_S32C1I
+	unsigned long tmp;
+	int result;
+
+	__asm__ __volatile__(
+			"1:     l32i    %1, %3, 0\n"
+			"       wsr     %1, scompare1\n"
+			"       or      %0, %1, %2\n"
+			"       s32c1i  %0, %3, 0\n"
+			"       bne     %0, %1, 1b\n"
+			: "=&a" (result), "=&a" (tmp)
+			: "a" (mask), "a" (v)
+			: "memory"
+			);
+#else
 	unsigned int vval;
 
 	__asm__ __volatile__(
-	"rsil    a15,"__stringify(LOCKLEVEL)"\n\t"
-	"l32i    %0, %2, 0             \n\t"
-	"or      %0, %0, %1            \n\t"
-	"s32i    %0, %2, 0             \n\t"
-	"wsr     a15, ps               \n\t"
-	"rsync                         \n"
+			"       rsil    a15,"__stringify(LOCKLEVEL)"\n"
+			"       l32i    %0, %2, 0\n"
+			"       or      %0, %0, %1\n"
+			"       s32i    %0, %2, 0\n"
+			"       wsr     a15, ps\n"
+			"       rsync\n"
 			: "=&a" (vval)
 			: "a" (mask), "a" (v)
 			: "a15", "memory"
 			);
+#endif
 }
 
 /* Atomic operations are already serializing */
@@ -294,4 +396,3 @@ static inline void atomic_set_mask(unsigned int mask, atomic_t *v)
 #endif /* __KERNEL__ */
 
 #endif /* _XTENSA_ATOMIC_H */
-

arch/xtensa/include/asm/barrier.h

@@ -3,7 +3,7 @@
  * License.  See the file "COPYING" in the main directory of this archive
  * for more details.
  *
- * Copyright (C) 2001 - 2005 Tensilica Inc.
+ * Copyright (C) 2001 - 2012 Tensilica Inc.
  */
 
 #ifndef _XTENSA_SYSTEM_H
@@ -12,8 +12,8 @@
 #define smp_read_barrier_depends() do { } while(0)
 #define read_barrier_depends() do { } while(0)
 
-#define mb()  barrier()
-#define rmb() mb()
+#define mb()  ({ __asm__ __volatile__("memw" : : : "memory"); })
+#define rmb() barrier()
 #define wmb() mb()
 
 #ifdef CONFIG_SMP

arch/xtensa/include/asm/bitops.h

@@ -29,7 +29,6 @@
 #define smp_mb__before_clear_bit()	barrier()
 #define smp_mb__after_clear_bit()	barrier()
 
-#include <asm-generic/bitops/atomic.h>
 #include <asm-generic/bitops/non-atomic.h>
 
 #if XCHAL_HAVE_NSA
@@ -104,6 +103,132 @@ static inline unsigned long __fls(unsigned long word)
 #endif
 
 #include <asm-generic/bitops/fls64.h>
+
+#if XCHAL_HAVE_S32C1I
+
+static inline void set_bit(unsigned int bit, volatile unsigned long *p)
+{
+	unsigned long tmp, value;
+	unsigned long mask = 1UL << (bit & 31);
+
+	p += bit >> 5;
+
+	__asm__ __volatile__(
+			"1:     l32i    %1, %3, 0\n"
+			"       wsr     %1, scompare1\n"
+			"       or      %0, %1, %2\n"
+			"       s32c1i  %0, %3, 0\n"
+			"       bne     %0, %1, 1b\n"
+			: "=&a" (tmp), "=&a" (value)
+			: "a" (mask), "a" (p)
+			: "memory");
+}
+
+static inline void clear_bit(unsigned int bit, volatile unsigned long *p)
+{
+	unsigned long tmp, value;
+	unsigned long mask = 1UL << (bit & 31);
+
+	p += bit >> 5;
+
+	__asm__ __volatile__(
+			"1:     l32i    %1, %3, 0\n"
+			"       wsr     %1, scompare1\n"
+			"       and     %0, %1, %2\n"
+			"       s32c1i  %0, %3, 0\n"
+			"       bne     %0, %1, 1b\n"
+			: "=&a" (tmp), "=&a" (value)
+			: "a" (~mask), "a" (p)
+			: "memory");
+}
+
+static inline void change_bit(unsigned int bit, volatile unsigned long *p)
+{
+	unsigned long tmp, value;
+	unsigned long mask = 1UL << (bit & 31);
+
+	p += bit >> 5;
+
+	__asm__ __volatile__(
+			"1:     l32i    %1, %3, 0\n"
+			"       wsr     %1, scompare1\n"
+			"       xor     %0, %1, %2\n"
+			"       s32c1i  %0, %3, 0\n"
+			"       bne     %0, %1, 1b\n"
+			: "=&a" (tmp), "=&a" (value)
+			: "a" (mask), "a" (p)
+			: "memory");
+}
+
+static inline int
+test_and_set_bit(unsigned int bit, volatile unsigned long *p)
+{
+	unsigned long tmp, value;
+	unsigned long mask = 1UL << (bit & 31);
+
+	p += bit >> 5;
+
+	__asm__ __volatile__(
+			"1:     l32i    %1, %3, 0\n"
+			"       wsr     %1, scompare1\n"
+			"       or      %0, %1, %2\n"
+			"       s32c1i  %0, %3, 0\n"
+			"       bne     %0, %1, 1b\n"
+			: "=&a" (tmp), "=&a" (value)
+			: "a" (mask), "a" (p)
+			: "memory");
+
+	return tmp & mask;
+}
+
+static inline int
+test_and_clear_bit(unsigned int bit, volatile unsigned long *p)
+{
+	unsigned long tmp, value;
+	unsigned long mask = 1UL << (bit & 31);
+
+	p += bit >> 5;
+
+	__asm__ __volatile__(
+			"1:     l32i    %1, %3, 0\n"
+			"       wsr     %1, scompare1\n"
+			"       and     %0, %1, %2\n"
+			"       s32c1i  %0, %3, 0\n"
+			"       bne     %0, %1, 1b\n"
+			: "=&a" (tmp), "=&a" (value)
+			: "a" (~mask), "a" (p)
+			: "memory");
+
+	return tmp & mask;
+}
+
+static inline int
+test_and_change_bit(unsigned int bit, volatile unsigned long *p)
+{
+	unsigned long tmp, value;
+	unsigned long mask = 1UL << (bit & 31);
+
+	p += bit >> 5;
+
+	__asm__ __volatile__(
+			"1:     l32i    %1, %3, 0\n"
+			"       wsr     %1, scompare1\n"
+			"       xor     %0, %1, %2\n"
+			"       s32c1i  %0, %3, 0\n"
+			"       bne     %0, %1, 1b\n"
+			: "=&a" (tmp), "=&a" (value)
+			: "a" (mask), "a" (p)
+			: "memory");
+
+	return tmp & mask;
+}
+
+#else
+
+#include <asm-generic/bitops/atomic.h>
+
+#endif /* XCHAL_HAVE_S32C1I */
+
 #include <asm-generic/bitops/find.h>
 #include <asm-generic/bitops/le.h>
 

arch/xtensa/include/asm/bootparam.h

@@ -22,6 +22,7 @@
 #define BP_TAG_MEMORY		0x1003	/* memory addr and size (bp_meminfo) */
 #define BP_TAG_SERIAL_BAUSRATE	0x1004	/* baud rate of current console. */
 #define BP_TAG_SERIAL_PORT	0x1005	/* serial device of current console */
+#define BP_TAG_FDT		0x1006	/* flat device tree addr */
 
 #define BP_TAG_FIRST		0x7B0B  /* first tag with a version number */
 #define BP_TAG_LAST 		0x7E0B	/* last tag */
@@ -56,6 +57,3 @@ extern sysmem_info_t sysmem;
 
 #endif
 #endif
-
-
-

arch/xtensa/include/asm/cacheasm.h

@@ -174,4 +174,3 @@
 	__loop_cache_page \ar \as ihi XCHAL_ICACHE_LINEWIDTH
 
 	.endm
-

arch/xtensa/include/asm/cacheflush.h

@@ -104,7 +104,8 @@ static inline void __invalidate_icache_page_alias(unsigned long virt,
 #define ARCH_IMPLEMENTS_FLUSH_DCACHE_PAGE 1
 extern void flush_dcache_page(struct page*);
 extern void flush_cache_range(struct vm_area_struct*, ulong, ulong);
-extern void flush_cache_page(struct vm_area_struct*, unsigned long, unsigned long);
+extern void flush_cache_page(struct vm_area_struct*,
+			     unsigned long, unsigned long);
 
 #else
 

arch/xtensa/include/asm/checksum.h

@@ -36,7 +36,8 @@ asmlinkage __wsum csum_partial(const void *buff, int len, __wsum sum);
  * better 64-bit) boundary
  */
 
-asmlinkage __wsum csum_partial_copy_generic(const void *src, void *dst, int len, __wsum sum,
+asmlinkage __wsum csum_partial_copy_generic(const void *src, void *dst,
+					    int len, __wsum sum,
 					    int *src_err_ptr, int *dst_err_ptr);
 
 /*
@@ -112,7 +113,8 @@ static __inline__ __sum16 ip_fast_csum(const void *iph, unsigned int ihl)
 	/* Since the input registers which are loaded with iph and ihl
 	   are modified, we must also specify them as outputs, or gcc
 	   will assume they contain their original values. */
-		: "=r" (sum), "=r" (iph), "=r" (ihl), "=&r" (tmp), "=&r" (endaddr)
+		: "=r" (sum), "=r" (iph), "=r" (ihl), "=&r" (tmp),
+		  "=&r" (endaddr)
 		: "1" (iph), "2" (ihl)
 		: "memory");
 
@@ -238,11 +240,12 @@ static __inline__ __sum16 csum_ipv6_magic(const struct in6_addr *saddr,
  *	Copy and checksum to user
  */
 #define HAVE_CSUM_COPY_USER
-static __inline__ __wsum csum_and_copy_to_user(const void *src, void __user *dst,
-				    int len, __wsum sum, int *err_ptr)
+static __inline__ __wsum csum_and_copy_to_user(const void *src,
+					       void __user *dst, int len,
+					       __wsum sum, int *err_ptr)
 {
 	if (access_ok(VERIFY_WRITE, dst, len))
-		return csum_partial_copy_generic(src, dst, len, sum, NULL, err_ptr);
+		return csum_partial_copy_generic(src,dst,len,sum,NULL,err_ptr);
 
 	if (len)
 		*err_ptr = -EFAULT;

arch/xtensa/include/asm/cmpxchg.h

@@ -22,17 +22,30 @@
 static inline unsigned long
 __cmpxchg_u32(volatile int *p, int old, int new)
 {
-  __asm__ __volatile__("rsil    a15, "__stringify(LOCKLEVEL)"\n\t"
-		       "l32i    %0, %1, 0              \n\t"
-		       "bne	%0, %2, 1f             \n\t"
-		       "s32i    %3, %1, 0              \n\t"
-		       "1:                             \n\t"
-		       "wsr     a15, ps                \n\t"
-		       "rsync                          \n\t"
+#if XCHAL_HAVE_S32C1I
+	__asm__ __volatile__(
+			"       wsr     %2, scompare1\n"
+			"       s32c1i  %0, %1, 0\n"
+			: "+a" (new)
+			: "a" (p), "a" (old)
+			: "memory"
+			);
+
+	return new;
+#else
+	__asm__ __volatile__(
+			"       rsil    a15, "__stringify(LOCKLEVEL)"\n"
+			"       l32i    %0, %1, 0\n"
+			"       bne     %0, %2, 1f\n"
+			"       s32i    %3, %1, 0\n"
+			"1:\n"
+			"       wsr     a15, ps\n"
+			"       rsync\n"
 			: "=&a" (old)
 			: "a" (p), "a" (old), "r" (new)
 			: "a15", "memory");
 	return old;
+#endif
 }
 /* This function doesn't exist, so you'll get a linker error
  * if something tries to do an invalid cmpxchg(). */
@@ -93,19 +106,36 @@ static inline unsigned long __cmpxchg_local(volatile void *ptr,
 
 static inline unsigned long xchg_u32(volatile int * m, unsigned long val)
 {
+#if XCHAL_HAVE_S32C1I
+	unsigned long tmp, result;
+	__asm__ __volatile__(
+			"1:     l32i    %1, %2, 0\n"
+			"       mov     %0, %3\n"
+			"       wsr     %1, scompare1\n"
+			"       s32c1i  %0, %2, 0\n"
+			"       bne     %0, %1, 1b\n"
+			: "=&a" (result), "=&a" (tmp)
+			: "a" (m), "a" (val)
+			: "memory"
+			);
+	return result;
+#else
 	unsigned long tmp;
-  __asm__ __volatile__("rsil    a15, "__stringify(LOCKLEVEL)"\n\t"
-		       "l32i    %0, %1, 0              \n\t"
-		       "s32i    %2, %1, 0              \n\t"
-		       "wsr     a15, ps                \n\t"
-		       "rsync                          \n\t"
+	__asm__ __volatile__(
+			"       rsil    a15, "__stringify(LOCKLEVEL)"\n"
+			"       l32i    %0, %1, 0\n"
+			"       s32i    %2, %1, 0\n"
+			"       wsr     a15, ps\n"
+			"       rsync\n"
 			: "=&a" (tmp)
 			: "a" (m), "a" (val)
 			: "a15", "memory");
 	return tmp;
+#endif
 }
 
-#define xchg(ptr,x) ((__typeof__(*(ptr)))__xchg((unsigned long)(x),(ptr),sizeof(*(ptr))))
+#define xchg(ptr,x) \
+	((__typeof__(*(ptr)))__xchg((unsigned long)(x),(ptr),sizeof(*(ptr))))
 
 /*
  * This only works if the compiler isn't horribly bad at optimizing.

arch/xtensa/include/asm/delay.h

@@ -46,4 +46,3 @@ static __inline__ void udelay (unsigned long usecs)
 }
 
 #endif
-

arch/xtensa/include/asm/dma-mapping.h

@@ -16,6 +16,8 @@
 #include <linux/mm.h>
 #include <linux/scatterlist.h>
 
+#define DMA_ERROR_CODE		(~(dma_addr_t)0x0)
+
 /*
  * DMA-consistent mapping functions.
  */
@@ -98,8 +100,8 @@ dma_sync_single_for_cpu(struct device *dev, dma_addr_t dma_handle, size_t size,
 }
 
 static inline void
-dma_sync_single_for_device(struct device *dev, dma_addr_t dma_handle, size_t size,
-		enum dma_data_direction direction)
+dma_sync_single_for_device(struct device *dev, dma_addr_t dma_handle,
+		           size_t size, enum dma_data_direction direction)
 {
 	consistent_sync((void *)bus_to_virt(dma_handle), size, direction);
 }

arch/xtensa/include/asm/highmem.h

@@ -14,4 +14,3 @@
 extern void flush_cache_kmaps(void);
 
 #endif
-

arch/xtensa/include/asm/initialize_mmu.h

+/*
+ * arch/xtensa/include/asm/initialize_mmu.h
+ *
+ * Initializes MMU:
+ *
+ *      For the new V3 MMU we remap the TLB from virtual == physical
+ *      to the standard Linux mapping used in earlier MMU's.
+ *
+ *      The the MMU we also support a new configuration register that
+ *      specifies how the S32C1I instruction operates with the cache
+ *      controller.
+ *
+ * This file is subject to the terms and conditions of the GNU General
+ * Public License.  See the file "COPYING" in the main directory of
+ * this archive for more details.
+ *
+ * Copyright (C) 2008 - 2012 Tensilica, Inc.
+ *
+ *   Marc Gauthier <marc@tensilica.com>
+ *   Pete Delaney <piet@tensilica.com>
+ */
+
+#ifndef _XTENSA_INITIALIZE_MMU_H
+#define _XTENSA_INITIALIZE_MMU_H
+
+#ifdef __ASSEMBLY__
+
+#define XTENSA_HWVERSION_RC_2009_0 230000
+
+	.macro	initialize_mmu
+
+#if XCHAL_HAVE_S32C1I && (XCHAL_HW_MIN_VERSION >= XTENSA_HWVERSION_RC_2009_0)
+/*
+ * We Have Atomic Operation Control (ATOMCTL) Register; Initialize it.
+ * For details see Documentation/xtensa/atomctl.txt
+ */
+#if XCHAL_DCACHE_IS_COHERENT
+	movi	a3, 0x25	/* For SMP/MX -- internal for writeback,
+				 * RCW otherwise
+				 */
+#else
+	movi	a3, 0x29	/* non-MX -- Most cores use Std Memory
+				 * Controlers which usually can't use RCW
+				 */
+#endif
+	wsr	a3, atomctl
+#endif  /* XCHAL_HAVE_S32C1I &&
+	 * (XCHAL_HW_MIN_VERSION >= XTENSA_HWVERSION_RC_2009_0)
+	 */
+
+	.endm
+
+#endif /*__ASSEMBLY__*/
+
+#endif /* _XTENSA_INITIALIZE_MMU_H */

arch/xtensa/include/asm/nommu_context.h

@@ -2,7 +2,7 @@ static inline void enter_lazy_tlb(struct mm_struct *mm, struct task_struct *tsk)
 {
 }
 
-static inline int init_new_context(struct task_struct *tsk, struct mm_struct *mm)
+static inline int init_new_context(struct task_struct *tsk,struct mm_struct *mm)
 {
 	return 0;
 }

arch/xtensa/include/asm/page.h

@@ -161,7 +161,9 @@ extern void copy_user_page(void*, void*, unsigned long, struct page*);
 
 #define __pa(x)			((unsigned long) (x) - PAGE_OFFSET)
 #define __va(x)			((void *)((unsigned long) (x) + PAGE_OFFSET))
-#define pfn_valid(pfn)		((pfn) >= ARCH_PFN_OFFSET && ((pfn) - ARCH_PFN_OFFSET) < max_mapnr)
+#define pfn_valid(pfn) \
+	((pfn) >= ARCH_PFN_OFFSET && ((pfn) - ARCH_PFN_OFFSET) < max_mapnr)
+
 #ifdef CONFIG_DISCONTIGMEM
 # error CONFIG_DISCONTIGMEM not supported
 #endif

arch/xtensa/include/asm/platform.h

@@ -75,4 +75,3 @@ extern int platform_pcibios_fixup (void);
 extern void platform_calibrate_ccount (void);
 
 #endif	/* _XTENSA_PLATFORM_H */
-

arch/xtensa/include/asm/prom.h

+#ifndef _XTENSA_ASM_PROM_H
+#define _XTENSA_ASM_PROM_H
+
+#define HAVE_ARCH_DEVTREE_FIXUPS
+
+#endif /* _XTENSA_ASM_PROM_H */

arch/xtensa/include/asm/ptrace.h

@@ -37,7 +37,7 @@ struct pt_regs {
 	unsigned long windowstart;	/*  52 */
 	unsigned long syscall;		/*  56 */
 	unsigned long icountlevel;	/*  60 */
-	int reserved[1];		/*  64 */
+	unsigned long scompare1;	/*  64 */
 
 	/* Additional configurable registers that are used by the compiler. */
 	xtregs_opt_t xtregs_opt;

arch/xtensa/include/asm/regs.h

@@ -52,6 +52,10 @@
 #define EXCCAUSE_SPECULATION			7
 #define EXCCAUSE_PRIVILEGED			8
 #define EXCCAUSE_UNALIGNED			9
+#define EXCCAUSE_INSTR_DATA_ERROR		12
+#define EXCCAUSE_LOAD_STORE_DATA_ERROR		13
+#define EXCCAUSE_INSTR_ADDR_ERROR		14
+#define EXCCAUSE_LOAD_STORE_ADDR_ERROR		15
 #define EXCCAUSE_ITLB_MISS			16
 #define EXCCAUSE_ITLB_MULTIHIT			17
 #define EXCCAUSE_ITLB_PRIVILEGE			18
@@ -105,4 +109,3 @@
 #define DEBUGCAUSE_ICOUNT_BIT		0	/* ICOUNT would incr. to zero */
 
 #endif /* _XTENSA_SPECREG_H */
-

arch/xtensa/include/asm/spinlock.h

@@ -11,6 +11,192 @@
 #ifndef _XTENSA_SPINLOCK_H
 #define _XTENSA_SPINLOCK_H
 
-#include <linux/spinlock.h>
+/*
+ * spinlock
+ *
+ * There is at most one owner of a spinlock.  There are not different
+ * types of spinlock owners like there are for rwlocks (see below).
+ *
+ * When trying to obtain a spinlock, the function "spins" forever, or busy-
+ * waits, until the lock is obtained.  When spinning, presumably some other
+ * owner will soon give up the spinlock making it available to others.  Use
+ * the trylock functions to avoid spinning forever.
+ *
+ * possible values:
+ *
+ *    0         nobody owns the spinlock
+ *    1         somebody owns the spinlock
+ */
+
+#define __raw_spin_is_locked(x) ((x)->slock != 0)
+#define __raw_spin_unlock_wait(lock) \
+	do { while (__raw_spin_is_locked(lock)) cpu_relax(); } while (0)
+
+#define __raw_spin_lock_flags(lock, flags) __raw_spin_lock(lock)
+
+static inline void __raw_spin_lock(raw_spinlock_t *lock)
+{
+	unsigned long tmp;
+
+	__asm__ __volatile__(
+			"       movi    %0, 0\n"
+			"       wsr     %0, scompare1\n"
+			"1:     movi    %0, 1\n"
+			"       s32c1i  %0, %1, 0\n"
+			"       bnez    %0, 1b\n"
+			: "=&a" (tmp)
+			: "a" (&lock->slock)
+			: "memory");
+}
+
+/* Returns 1 if the lock is obtained, 0 otherwise. */
+
+static inline int __raw_spin_trylock(raw_spinlock_t *lock)
+{
+	unsigned long tmp;
+
+	__asm__ __volatile__(
+			"       movi    %0, 0\n"
+			"       wsr     %0, scompare1\n"
+			"       movi    %0, 1\n"
+			"       s32c1i  %0, %1, 0\n"
+			: "=&a" (tmp)
+			: "a" (&lock->slock)
+			: "memory");
+
+	return tmp == 0 ? 1 : 0;
+}
+
+static inline void __raw_spin_unlock(raw_spinlock_t *lock)
+{
+	unsigned long tmp;
+
+	__asm__ __volatile__(
+			"       movi    %0, 0\n"
+			"       s32ri   %0, %1, 0\n"
+			: "=&a" (tmp)
+			: "a" (&lock->slock)
+			: "memory");
+}
+
+/*
+ * rwlock
+ *
+ * Read-write locks are really a more flexible spinlock.  They allow
+ * multiple readers but only one writer.  Write ownership is exclusive
+ * (i.e., all other readers and writers are blocked from ownership while
+ * there is a write owner).  These rwlocks are unfair to writers.  Writers
+ * can be starved for an indefinite time by readers.
+ *
+ * possible values:
+ *
+ *   0          nobody owns the rwlock
+ *  >0          one or more readers own the rwlock
+ *                (the positive value is the actual number of readers)
+ *  0x80000000  one writer owns the rwlock, no other writers, no readers
+ */
+
+#define __raw_write_can_lock(x)  ((x)->lock == 0)
+
+static inline void __raw_write_lock(raw_rwlock_t *rw)
+{
+	unsigned long tmp;
+
+	__asm__ __volatile__(
+			"       movi    %0, 0\n"
+			"       wsr     %0, scompare1\n"
+			"1:     movi    %0, 1\n"
+			"       slli    %0, %0, 31\n"
+			"       s32c1i  %0, %1, 0\n"
+			"       bnez    %0, 1b\n"
+			: "=&a" (tmp)
+			: "a" (&rw->lock)
+			: "memory");
+}
+
+/* Returns 1 if the lock is obtained, 0 otherwise. */
+
+static inline int __raw_write_trylock(raw_rwlock_t *rw)
+{
+	unsigned long tmp;
+
+	__asm__ __volatile__(
+			"       movi    %0, 0\n"
+			"       wsr     %0, scompare1\n"
+			"       movi    %0, 1\n"
+			"       slli    %0, %0, 31\n"
+			"       s32c1i  %0, %1, 0\n"
+			: "=&a" (tmp)
+			: "a" (&rw->lock)
+			: "memory");
+
+	return tmp == 0 ? 1 : 0;
+}
+
+static inline void __raw_write_unlock(raw_rwlock_t *rw)
+{
+	unsigned long tmp;
+
+	__asm__ __volatile__(
+			"       movi    %0, 0\n"
+			"       s32ri   %0, %1, 0\n"
+			: "=&a" (tmp)
+			: "a" (&rw->lock)
+			: "memory");
+}
+
+static inline void __raw_read_lock(raw_rwlock_t *rw)
+{
+	unsigned long tmp;
+	unsigned long result;
+
+	__asm__ __volatile__(
+			"1:     l32i    %1, %2, 0\n"
+			"       bltz    %1, 1b\n"
+			"       wsr     %1, scompare1\n"
+			"       addi    %0, %1, 1\n"
+			"       s32c1i  %0, %2, 0\n"
+			"       bne     %0, %1, 1b\n"
+			: "=&a" (result), "=&a" (tmp)
+			: "a" (&rw->lock)
+			: "memory");
+}
+
+/* Returns 1 if the lock is obtained, 0 otherwise. */
+
+static inline int __raw_read_trylock(raw_rwlock_t *rw)
+{
+	unsigned long result;
+	unsigned long tmp;
+
+	__asm__ __volatile__(
+			"       l32i    %1, %2, 0\n"
+			"       addi    %0, %1, 1\n"
+			"       bltz    %0, 1f\n"
+			"       wsr     %1, scompare1\n"
+			"       s32c1i  %0, %2, 0\n"
+			"       sub     %0, %0, %1\n"
+			"1:\n"
+			: "=&a" (result), "=&a" (tmp)
+			: "a" (&rw->lock)
+			: "memory");
+
+	return result == 0;
+}
+
+static inline void __raw_read_unlock(raw_rwlock_t *rw)
+{
+	unsigned long tmp1, tmp2;
+
+	__asm__ __volatile__(
+			"1:     l32i    %1, %2, 0\n"
+			"       addi    %0, %1, -1\n"
+			"       wsr     %1, scompare1\n"
+			"       s32c1i  %0, %2, 0\n"
+			"       bne     %0, %1, 1b\n"
+			: "=&a" (tmp1), "=&a" (tmp2)
+			: "a" (&rw->lock)
+			: "memory");
+}
 
 #endif	/* _XTENSA_SPINLOCK_H */

arch/xtensa/include/asm/syscall.h

@@ -25,9 +25,10 @@ asmlinkage long xtensa_fadvise64_64(int, int,
 /* Should probably move to linux/syscalls.h */
 struct pollfd;
 asmlinkage long sys_pselect6(int n, fd_set __user *inp, fd_set __user *outp,
-	fd_set __user *exp, struct timespec __user *tsp, void __user *sig);
+			     fd_set __user *exp, struct timespec __user *tsp,
+			     void __user *sig);
 asmlinkage long sys_ppoll(struct pollfd __user *ufds, unsigned int nfds,
-	struct timespec __user *tsp, const sigset_t __user *sigmask,
-	size_t sigsetsize);
-asmlinkage long sys_rt_sigsuspend(sigset_t __user *unewset,
+			  struct timespec __user *tsp,
+			  const sigset_t __user *sigmask,
 			  size_t sigsetsize);
+asmlinkage long sys_rt_sigsuspend(sigset_t __user *unewset, size_t sigsetsize);

arch/xtensa/include/asm/traps.h

+/*
+ * arch/xtensa/include/asm/traps.h
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2012 Tensilica Inc.
+ */
+#ifndef _XTENSA_TRAPS_H
+#define _XTENSA_TRAPS_H
+
+#include <asm/ptrace.h>
+
+/*
+ * handler must be either of the following:
+ *  void (*)(struct pt_regs *regs);
+ *  void (*)(struct pt_regs *regs, unsigned long exccause);
+ */
+extern void * __init trap_set_handler(int cause, void *handler);
+extern void do_unhandled(struct pt_regs *regs, unsigned long exccause);
+
+#endif /* _XTENSA_TRAPS_H */

arch/xtensa/include/asm/uaccess.h

@@ -180,7 +180,8 @@
 #define segment_eq(a,b)	((a).seg == (b).seg)
 
 #define __kernel_ok (segment_eq(get_fs(), KERNEL_DS))
-#define __user_ok(addr,size) (((size) <= TASK_SIZE)&&((addr) <= TASK_SIZE-(size)))
+#define __user_ok(addr,size) \
+		(((size) <= TASK_SIZE)&&((addr) <= TASK_SIZE-(size)))
 #define __access_ok(addr,size) (__kernel_ok || __user_ok((addr),(size)))
 #define access_ok(type,addr,size) __access_ok((unsigned long)(addr),(size))
 
@@ -291,7 +292,7 @@ do {									\
  * __check_align_* macros still work.
  */
 #define __put_user_asm(x, addr, err, align, insn, cb)	\
-   __asm__ __volatile__(				\
+__asm__ __volatile__(					\
 	__check_align_##align				\
 	"1: "insn"  %2, %3, 0		\n"		\
 	"2:				\n"		\
@@ -349,7 +350,7 @@ do {									\
  * __check_align_* macros still work.
  */
 #define __get_user_asm(x, addr, err, align, insn, cb) \
-   __asm__ __volatile__(			\
+__asm__ __volatile__(			\
 	__check_align_##align			\
 	"1: "insn"  %2, %3, 0		\n"	\
 	"2:				\n"	\
@@ -421,8 +422,10 @@ __generic_copy_from_user(void *to, const void *from, unsigned long n)
 
 #define copy_to_user(to,from,n) __generic_copy_to_user((to),(from),(n))
 #define copy_from_user(to,from,n) __generic_copy_from_user((to),(from),(n))
-#define __copy_to_user(to,from,n) __generic_copy_to_user_nocheck((to),(from),(n))
-#define __copy_from_user(to,from,n) __generic_copy_from_user_nocheck((to),(from),(n))
+#define __copy_to_user(to,from,n) \
+	__generic_copy_to_user_nocheck((to),(from),(n))
+#define __copy_from_user(to,from,n) \
+	__generic_copy_from_user_nocheck((to),(from),(n))
 #define __copy_to_user_inatomic __copy_to_user
 #define __copy_from_user_inatomic __copy_from_user
 

arch/xtensa/kernel/Makefile

@@ -28,7 +28,7 @@ sed-y = -e 's/\*(\(\.[a-z]*it\|\.ref\|\)\.text)/*(\1.literal \1.text)/g'    \
 	-e 's/\*(\(\.text\.[a-z]*\))/*(\1.literal \1)/g'
 
 quiet_cmd__cpp_lds_S = LDS     $@
-      cmd__cpp_lds_S = $(CPP) $(cpp_flags) -P -C -Uxtensa -D__ASSEMBLY__ $< \
+cmd__cpp_lds_S = $(CPP) $(cpp_flags) -P -C -Uxtensa -D__ASSEMBLY__ $<    \
                  | sed $(sed-y) >$@
 
 $(obj)/vmlinux.lds: $(src)/vmlinux.lds.S FORCE

arch/xtensa/kernel/align.S

@@ -450,6 +450,6 @@ ENTRY(fast_unaligned)
 1:	movi	a0, _user_exception
 	jx	a0
 
+ENDPROC(fast_unaligned)
 
 #endif /* XCHAL_UNALIGNED_LOAD_EXCEPTION || XCHAL_UNALIGNED_STORE_EXCEPTION */
-

arch/xtensa/kernel/asm-offsets.c

@@ -41,6 +41,7 @@ int main(void)
 	DEFINE(PT_SAR, offsetof (struct pt_regs, sar));
 	DEFINE(PT_ICOUNTLEVEL, offsetof (struct pt_regs, icountlevel));
 	DEFINE(PT_SYSCALL, offsetof (struct pt_regs, syscall));
+	DEFINE(PT_SCOMPARE1, offsetof(struct pt_regs, scompare1));
 	DEFINE(PT_AREG, offsetof (struct pt_regs, areg[0]));
 	DEFINE(PT_AREG0, offsetof (struct pt_regs, areg[0]));
 	DEFINE(PT_AREG1, offsetof (struct pt_regs, areg[1]));
@@ -91,7 +92,8 @@ int main(void)
 #endif
 	DEFINE(THREAD_XTREGS_USER, offsetof (struct thread_info, xtregs_user));
 	DEFINE(XTREGS_USER_SIZE, sizeof(xtregs_user_t));
-	DEFINE(THREAD_CURRENT_DS, offsetof (struct task_struct, thread.current_ds));
+	DEFINE(THREAD_CURRENT_DS, offsetof (struct task_struct, \
+	       thread.current_ds));
 
 	/* struct mm_struct */
 	DEFINE(MM_USERS, offsetof(struct mm_struct, mm_users));
@@ -108,4 +110,3 @@ int main(void)
 
 	return 0;
 }
-

arch/xtensa/kernel/coprocessor.S

@@ -43,10 +43,13 @@
 /* IO protection is currently unsupported. */
 
 ENTRY(fast_io_protect)
+
 	wsr	a0, excsave1
 	movi	a0, unrecoverable_exception
 	callx0	a0
 
+ENDPROC(fast_io_protect)
+
 #if XTENSA_HAVE_COPROCESSORS
 
 /*
@@ -139,6 +142,7 @@ ENTRY(fast_io_protect)
  */
 
 ENTRY(coprocessor_save)
+
 	entry	a1, 32
 	s32i	a0, a1, 0
 	movi	a0, .Lsave_cp_regs_jump_table
@@ -150,7 +154,10 @@ ENTRY(coprocessor_save)
 1:	l32i	a0, a1, 0
 	retw
 
+ENDPROC(coprocessor_save)
+
 ENTRY(coprocessor_load)
+
 	entry	a1, 32
 	s32i	a0, a1, 0
 	movi	a0, .Lload_cp_regs_jump_table
@@ -162,6 +169,8 @@ ENTRY(coprocessor_load)
 1:	l32i	a0, a1, 0
 	retw
 
+ENDPROC(coprocessor_load)
+
 /*
  * coprocessor_flush(struct task_info*, index)
  *                             a2        a3
@@ -178,6 +187,7 @@ ENTRY(coprocessor_load)
 
 
 ENTRY(coprocessor_flush)
+
 	entry	a1, 32
 	s32i	a0, a1, 0
 	movi	a0, .Lsave_cp_regs_jump_table
@@ -191,6 +201,8 @@ ENTRY(coprocessor_flush)
 1:	l32i	a0, a1, 0
 	retw
 
+ENDPROC(coprocessor_flush)
+
 ENTRY(coprocessor_restore)
 	entry	a1, 32
 	s32i	a0, a1, 0
@@ -205,6 +217,8 @@ ENTRY(coprocessor_restore)
 1:	l32i	a0, a1, 0
 	retw
 
+ENDPROC(coprocessor_restore)
+
 /*
  * Entry condition:
  *
@@ -220,10 +234,12 @@ ENTRY(coprocessor_restore)
  */
 
 ENTRY(fast_coprocessor_double)
+
 	wsr	a0, excsave1
 	movi	a0, unrecoverable_exception
 	callx0	a0
 
+ENDPROC(fast_coprocessor_double)
 
 ENTRY(fast_coprocessor)
 
@@ -327,9 +343,14 @@ ENTRY(fast_coprocessor)
 
 	rfe
 
+ENDPROC(fast_coprocessor)
+
 	.data
+
 ENTRY(coprocessor_owner)
+
 	.fill XCHAL_CP_MAX, 4, 0
 
-#endif /* XTENSA_HAVE_COPROCESSORS */
+END(coprocessor_owner)
 
+#endif /* XTENSA_HAVE_COPROCESSORS */

arch/xtensa/kernel/entry.S

@@ -219,6 +219,7 @@ _user_exception:
 
 	j	common_exception
 
+ENDPROC(user_exception)
 
 /*
  * First-level exit handler for kernel exceptions
@@ -371,6 +372,13 @@ common_exception:
 	s32i	a2, a1, PT_LBEG
 	s32i	a3, a1, PT_LEND
 
+	/* Save SCOMPARE1 */
+
+#if XCHAL_HAVE_S32C1I
+	rsr     a2, scompare1
+	s32i    a2, a1, PT_SCOMPARE1
+#endif
+
 	/* Save optional registers. */
 
 	save_xtregs_opt a1 a2 a4 a5 a6 a7 PT_XTREGS_OPT
@@ -432,6 +440,12 @@ common_exception_return:
 
 	load_xtregs_opt a1 a2 a4 a5 a6 a7 PT_XTREGS_OPT
 
+	/* Restore SCOMPARE1 */
+
+#if XCHAL_HAVE_S32C1I
+	l32i    a2, a1, PT_SCOMPARE1
+	wsr     a2, scompare1
+#endif
 	wsr	a3, ps		/* disable interrupts */
 
 	_bbci.l	a3, PS_UM_BIT, kernel_exception_exit
@@ -641,6 +655,8 @@ common_exception_exit:
 	l32i	a1, a1, PT_AREG1
 	rfde
 
+ENDPROC(kernel_exception)
+
 /*
  * Debug exception handler.
  *
@@ -701,6 +717,7 @@ ENTRY(debug_exception)
 	/* Debug exception while in exception mode. */
 1:	j	1b	// FIXME!!
 
+ENDPROC(debug_exception)
 
 /*
  * We get here in case of an unrecoverable exception.
@@ -751,6 +768,7 @@ ENTRY(unrecoverable_exception)
 
 1:	j	1b
 
+ENDPROC(unrecoverable_exception)
 
 /* -------------------------- FAST EXCEPTION HANDLERS ----------------------- */
 
@@ -929,6 +947,7 @@ ENTRY(fast_alloca)
 	l32i	a2, a2, PT_AREG2
 	rfe
 
+ENDPROC(fast_alloca)
 
 /*
  * fast system calls.
@@ -966,6 +985,8 @@ ENTRY(fast_syscall_kernel)
 
 	j	kernel_exception
 
+ENDPROC(fast_syscall_kernel)
+
 ENTRY(fast_syscall_user)
 
 	/* Skip syscall. */
@@ -983,6 +1004,8 @@ ENTRY(fast_syscall_user)
 
 	j	user_exception
 
+ENDPROC(fast_syscall_user)
+
 ENTRY(fast_syscall_unrecoverable)
 
 	/* Restore all states. */
@@ -995,7 +1018,7 @@ ENTRY(fast_syscall_unrecoverable)
 	movi    a0, unrecoverable_exception
 	callx0  a0
 
-
+ENDPROC(fast_syscall_unrecoverable)
 
 /*
  * sysxtensa syscall handler
@@ -1101,7 +1124,7 @@ CATCH
 	movi	a2, -EINVAL
 	rfe
 
-
+ENDPROC(fast_syscall_xtensa)
 
 
 /* fast_syscall_spill_registers.
@@ -1160,6 +1183,8 @@ ENTRY(fast_syscall_spill_registers)
 	movi	a2, 0
 	rfe
 
+ENDPROC(fast_syscall_spill_registers)
+
 /* Fixup handler.
  *
  * We get here if the spill routine causes an exception, e.g. tlb miss.
@@ -1464,6 +1489,8 @@ ENTRY(_spill_registers)
 	callx0	a0		# should not return
 1:	j	1b
 
+ENDPROC(_spill_registers)
+
 #ifdef CONFIG_MMU
 /*
  * We should never get here. Bail out!
@@ -1475,6 +1502,8 @@ ENTRY(fast_second_level_miss_double_kernel)
 	callx0	a0		# should not return
 1:	j	1b
 
+ENDPROC(fast_second_level_miss_double_kernel)
+
 /* First-level entry handler for user, kernel, and double 2nd-level
  * TLB miss exceptions.  Note that for now, user and kernel miss
  * exceptions share the same entry point and are handled identically.
@@ -1682,6 +1711,7 @@ ENTRY(fast_second_level_miss)
 	j	_kernel_exception
 1:	j	_user_exception
 
+ENDPROC(fast_second_level_miss)
 
 /*
  * StoreProhibitedException
@@ -1777,6 +1807,9 @@ ENTRY(fast_store_prohibited)
 	bbsi.l	a2, PS_UM_BIT, 1f
 	j	_kernel_exception
 1:	j	_user_exception
+
+ENDPROC(fast_store_prohibited)
+
 #endif /* CONFIG_MMU */
 
 /*
@@ -1787,6 +1820,7 @@ ENTRY(fast_store_prohibited)
  */
 
 ENTRY(system_call)
+
 	entry	a1, 32
 
 	/* regs->syscall = regs->areg[2] */
@@ -1831,6 +1865,8 @@ ENTRY(system_call)
 	callx4	a4
 	retw
 
+ENDPROC(system_call)
+
 
 /*
  * Task switch.
@@ -1899,6 +1935,7 @@ ENTRY(_switch_to)
 
 	retw
 
+ENDPROC(_switch_to)
 
 ENTRY(ret_from_fork)
 
@@ -1914,6 +1951,8 @@ ENTRY(ret_from_fork)
 
 	j	common_exception_return
 
+ENDPROC(ret_from_fork)
+
 /*
  * Kernel thread creation helper
  * On entry, set up by copy_thread: a2 = thread_fn, a3 = thread_fn arg

arch/xtensa/kernel/head.S

@@ -18,6 +18,7 @@
 #include <asm/processor.h>
 #include <asm/page.h>
 #include <asm/cacheasm.h>
+#include <asm/initialize_mmu.h>
 
 #include <linux/init.h>
 #include <linux/linkage.h>
@@ -47,16 +48,19 @@
 	 */
 
 	__HEAD
-	.globl _start
-_start:	_j	2f
+ENTRY(_start)
+
+	_j	2f
 	.align	4
 1:	.word	_startup
 2:	l32r	a0, 1b
 	jx	a0
 
+ENDPROC(_start)
+
 	.section .init.text, "ax"
-	.align 4
-_startup:
+
+ENTRY(_startup)
 
 	/* Disable interrupts and exceptions. */
 
@@ -107,7 +111,7 @@ _startup:
 	/* Disable all timers. */
 
 	.set	_index, 0
-	.rept	XCHAL_NUM_TIMERS - 1
+	.rept	XCHAL_NUM_TIMERS
 	wsr	a0, SREG_CCOMPARE + _index
 	.set	_index, _index + 1
 	.endr
@@ -120,7 +124,7 @@ _startup:
 
 	/* Disable coprocessors. */
 
-#if XCHAL_CP_NUM > 0
+#if XCHAL_HAVE_CP
 	wsr	a0, cpenable
 #endif
 
@@ -152,6 +156,8 @@ _startup:
 
 	isync
 
+	initialize_mmu
+
 	/* Unpack data sections
 	 *
 	 * The linker script used to build the Linux kernel image
@@ -230,6 +236,7 @@ _startup:
 should_never_return:
 	j	should_never_return
 
+ENDPROC(_startup)
 
 /*
  * BSS section
@@ -239,6 +246,8 @@ __PAGE_ALIGNED_BSS
 #ifdef CONFIG_MMU
 ENTRY(swapper_pg_dir)
 	.fill	PAGE_SIZE, 1, 0
+END(swapper_pg_dir)
 #endif
 ENTRY(empty_zero_page)
 	.fill	PAGE_SIZE, 1, 0
+END(empty_zero_page)

arch/xtensa/kernel/irq.c

@@ -18,6 +18,8 @@
 #include <linux/interrupt.h>
 #include <linux/irq.h>
 #include <linux/kernel_stat.h>
+#include <linux/irqdomain.h>
+#include <linux/of.h>
 
 #include <asm/uaccess.h>
 #include <asm/platform.h>
@@ -26,19 +28,22 @@ static unsigned int cached_irq_mask;
 
 atomic_t irq_err_count;
 
+static struct irq_domain *root_domain;
+
 /*
  * do_IRQ handles all normal device IRQ's (the special
  * SMP cross-CPU interrupts have their own specific
  * handlers).
  */
 
-asmlinkage void do_IRQ(int irq, struct pt_regs *regs)
+asmlinkage void do_IRQ(int hwirq, struct pt_regs *regs)
 {
 	struct pt_regs *old_regs = set_irq_regs(regs);
+	int irq = irq_find_mapping(root_domain, hwirq);
 
-	if (irq >= NR_IRQS) {
+	if (hwirq >= NR_IRQS) {
 		printk(KERN_EMERG "%s: cannot handle IRQ %d\n",
-				__func__, irq);
+				__func__, hwirq);
 	}
 
 	irq_enter();
@@ -71,40 +76,39 @@ int arch_show_interrupts(struct seq_file *p, int prec)
 
 static void xtensa_irq_mask(struct irq_data *d)
 {
-	cached_irq_mask &= ~(1 << d->irq);
+	cached_irq_mask &= ~(1 << d->hwirq);
 	set_sr (cached_irq_mask, intenable);
 }
 
 static void xtensa_irq_unmask(struct irq_data *d)
 {
-	cached_irq_mask |= 1 << d->irq;
+	cached_irq_mask |= 1 << d->hwirq;
 	set_sr (cached_irq_mask, intenable);
 }
 
 static void xtensa_irq_enable(struct irq_data *d)
 {
-	variant_irq_enable(d->irq);
+	variant_irq_enable(d->hwirq);
 	xtensa_irq_unmask(d);
 }
 
 static void xtensa_irq_disable(struct irq_data *d)
 {
 	xtensa_irq_mask(d);
-	variant_irq_disable(d->irq);
+	variant_irq_disable(d->hwirq);
 }
 
 static void xtensa_irq_ack(struct irq_data *d)
 {
-	set_sr(1 << d->irq, intclear);
+	set_sr(1 << d->hwirq, intclear);
 }
 
 static int xtensa_irq_retrigger(struct irq_data *d)
 {
-	set_sr (1 << d->irq, INTSET);
+	set_sr(1 << d->hwirq, intset);
 	return 1;
 }
 
-
 static struct irq_chip xtensa_irq_chip = {
 	.name		= "xtensa",
 	.irq_enable	= xtensa_irq_enable,
@@ -115,37 +119,99 @@ static struct irq_chip xtensa_irq_chip = {
 	.irq_retrigger	= xtensa_irq_retrigger,
 };
 
-void __init init_IRQ(void)
+static int xtensa_irq_map(struct irq_domain *d, unsigned int irq,
+		irq_hw_number_t hw)
 {
-	int index;
-
-	for (index = 0; index < XTENSA_NR_IRQS; index++) {
-		int mask = 1 << index;
+	u32 mask = 1 << hw;
+
+	if (mask & XCHAL_INTTYPE_MASK_SOFTWARE) {
+		irq_set_chip_and_handler_name(irq, &xtensa_irq_chip,
+				handle_simple_irq, "level");
+		irq_set_status_flags(irq, IRQ_LEVEL);
+	} else if (mask & XCHAL_INTTYPE_MASK_EXTERN_EDGE) {
+		irq_set_chip_and_handler_name(irq, &xtensa_irq_chip,
+				handle_edge_irq, "edge");
+		irq_clear_status_flags(irq, IRQ_LEVEL);
+	} else if (mask & XCHAL_INTTYPE_MASK_EXTERN_LEVEL) {
+		irq_set_chip_and_handler_name(irq, &xtensa_irq_chip,
+				handle_level_irq, "level");
+		irq_set_status_flags(irq, IRQ_LEVEL);
+	} else if (mask & XCHAL_INTTYPE_MASK_TIMER) {
+		irq_set_chip_and_handler_name(irq, &xtensa_irq_chip,
+				handle_edge_irq, "edge");
+		irq_clear_status_flags(irq, IRQ_LEVEL);
+	} else {/* XCHAL_INTTYPE_MASK_WRITE_ERROR */
+		/* XCHAL_INTTYPE_MASK_NMI */
 
-		if (mask & XCHAL_INTTYPE_MASK_SOFTWARE)
-			irq_set_chip_and_handler(index, &xtensa_irq_chip,
-						 handle_simple_irq);
+		irq_set_chip_and_handler_name(irq, &xtensa_irq_chip,
+				handle_level_irq, "level");
+		irq_set_status_flags(irq, IRQ_LEVEL);
+	}
+	return 0;
+}
 
-		else if (mask & XCHAL_INTTYPE_MASK_EXTERN_EDGE)
-			irq_set_chip_and_handler(index, &xtensa_irq_chip,
-						 handle_edge_irq);
+static unsigned map_ext_irq(unsigned ext_irq)
+{
+	unsigned mask = XCHAL_INTTYPE_MASK_EXTERN_EDGE |
+		XCHAL_INTTYPE_MASK_EXTERN_LEVEL;
+	unsigned i;
 
-		else if (mask & XCHAL_INTTYPE_MASK_EXTERN_LEVEL)
-			irq_set_chip_and_handler(index, &xtensa_irq_chip,
-						 handle_level_irq);
+	for (i = 0; mask; ++i, mask >>= 1) {
+		if ((mask & 1) && ext_irq-- == 0)
+			return i;
+	}
+	return XCHAL_NUM_INTERRUPTS;
+}
 
-		else if (mask & XCHAL_INTTYPE_MASK_TIMER)
-			irq_set_chip_and_handler(index, &xtensa_irq_chip,
-						 handle_edge_irq);
+/*
+ * Device Tree IRQ specifier translation function which works with one or
+ * two cell bindings. First cell value maps directly to the hwirq number.
+ * Second cell if present specifies whether hwirq number is external (1) or
+ * internal (0).
+ */
+int xtensa_irq_domain_xlate(struct irq_domain *d, struct device_node *ctrlr,
+		const u32 *intspec, unsigned int intsize,
+		unsigned long *out_hwirq, unsigned int *out_type)
+{
+	if (WARN_ON(intsize < 1 || intsize > 2))
+		return -EINVAL;
+	if (intsize == 2 && intspec[1] == 1) {
+		unsigned int_irq = map_ext_irq(intspec[0]);
+		if (int_irq < XCHAL_NUM_INTERRUPTS)
+			*out_hwirq = int_irq;
+		else
+			return -EINVAL;
+	} else {
+		*out_hwirq = intspec[0];
+	}
+	*out_type = IRQ_TYPE_NONE;
+	return 0;
+}
 
-		else	/* XCHAL_INTTYPE_MASK_WRITE_ERROR */
-			/* XCHAL_INTTYPE_MASK_NMI */
+static const struct irq_domain_ops xtensa_irq_domain_ops = {
+	.xlate = xtensa_irq_domain_xlate,
+	.map = xtensa_irq_map,
+};
 
-			irq_set_chip_and_handler(index, &xtensa_irq_chip,
-						 handle_level_irq);
-	}
+void __init init_IRQ(void)
+{
+	struct device_node *intc = NULL;
 
 	cached_irq_mask = 0;
+	set_sr(~0, intclear);
+
+#ifdef CONFIG_OF
+	/* The interrupt controller device node is mandatory */
+	intc = of_find_compatible_node(NULL, NULL, "xtensa,pic");
+	BUG_ON(!intc);
+
+	root_domain = irq_domain_add_linear(intc, NR_IRQS,
+			&xtensa_irq_domain_ops, NULL);
+#else
+	root_domain = irq_domain_add_legacy(intc, NR_IRQS, 0, 0,
+			&xtensa_irq_domain_ops, NULL);
+#endif
+	irq_set_default_host(root_domain);
 
 	variant_init_irq();
 }

arch/xtensa/kernel/platform.c

@@ -44,4 +44,3 @@ _F(void, calibrate_ccount, (void),
 	ccount_per_jiffy = 10 * (1000000UL/HZ);
 });
 #endif
-

arch/xtensa/kernel/ptrace.c

@@ -343,4 +343,3 @@ void do_syscall_trace_leave(struct pt_regs *regs)
 			&& (current->ptrace & PT_PTRACED))
 		do_syscall_trace();
 }
-

arch/xtensa/kernel/setup.c

@@ -22,6 +22,11 @@
 #include <linux/bootmem.h>
 #include <linux/kernel.h>
 
+#ifdef CONFIG_OF
+#include <linux/of_fdt.h>
+#include <linux/of_platform.h>
+#endif
+
 #if defined(CONFIG_VGA_CONSOLE) || defined(CONFIG_DUMMY_CONSOLE)
 # include <linux/console.h>
 #endif
@@ -42,6 +47,7 @@
 #include <asm/page.h>
 #include <asm/setup.h>
 #include <asm/param.h>
+#include <asm/traps.h>
 
 #include <platform/hardware.h>
 
@@ -64,6 +70,11 @@ int initrd_is_mapped = 0;
 extern int initrd_below_start_ok;
 #endif
 
+#ifdef CONFIG_OF
+extern u32 __dtb_start[];
+void *dtb_start = __dtb_start;
+#endif
+
 unsigned char aux_device_present;
 extern unsigned long loops_per_jiffy;
 
@@ -83,6 +94,8 @@ extern void init_mmu(void);
 static inline void init_mmu(void) { }
 #endif
 
+extern int mem_reserve(unsigned long, unsigned long, int);
+extern void bootmem_init(void);
 extern void zones_init(void);
 
 /*
@@ -104,28 +117,33 @@ typedef struct tagtable {
 
 /* parse current tag */
 
-static int __init parse_tag_mem(const bp_tag_t *tag)
+static int __init add_sysmem_bank(unsigned long type, unsigned long start,
+		unsigned long end)
 {
-	meminfo_t *mi = (meminfo_t*)(tag->data);
-
-	if (mi->type != MEMORY_TYPE_CONVENTIONAL)
-		return -1;
-
 	if (sysmem.nr_banks >= SYSMEM_BANKS_MAX) {
 		printk(KERN_WARNING
 				"Ignoring memory bank 0x%08lx size %ldKB\n",
-		       (unsigned long)mi->start,
-		       (unsigned long)mi->end - (unsigned long)mi->start);
+				start, end - start);
 		return -EINVAL;
 	}
-	sysmem.bank[sysmem.nr_banks].type  = mi->type;
-	sysmem.bank[sysmem.nr_banks].start = PAGE_ALIGN(mi->start);
-	sysmem.bank[sysmem.nr_banks].end   = mi->end & PAGE_MASK;
+	sysmem.bank[sysmem.nr_banks].type  = type;
+	sysmem.bank[sysmem.nr_banks].start = PAGE_ALIGN(start);
+	sysmem.bank[sysmem.nr_banks].end   = end & PAGE_MASK;
 	sysmem.nr_banks++;
 
 	return 0;
 }
 
+static int __init parse_tag_mem(const bp_tag_t *tag)
+{
+	meminfo_t *mi = (meminfo_t *)(tag->data);
+
+	if (mi->type != MEMORY_TYPE_CONVENTIONAL)
+		return -1;
+
+	return add_sysmem_bank(mi->type, mi->start, mi->end);
+}
+
 __tagtable(BP_TAG_MEMORY, parse_tag_mem);
 
 #ifdef CONFIG_BLK_DEV_INITRD
@@ -142,12 +160,31 @@ static int __init parse_tag_initrd(const bp_tag_t* tag)
 
 __tagtable(BP_TAG_INITRD, parse_tag_initrd);
 
+#ifdef CONFIG_OF
+
+static int __init parse_tag_fdt(const bp_tag_t *tag)
+{
+	dtb_start = (void *)(tag->data[0]);
+	return 0;
+}
+
+__tagtable(BP_TAG_FDT, parse_tag_fdt);
+
+void __init early_init_dt_setup_initrd_arch(unsigned long start,
+		unsigned long end)
+{
+	initrd_start = (void *)__va(start);
+	initrd_end = (void *)__va(end);
+	initrd_below_start_ok = 1;
+}
+
+#endif /* CONFIG_OF */
+
 #endif /* CONFIG_BLK_DEV_INITRD */
 
 static int __init parse_tag_cmdline(const bp_tag_t* tag)
 {
-	strncpy(command_line, (char*)(tag->data), COMMAND_LINE_SIZE);
-	command_line[COMMAND_LINE_SIZE - 1] = '\0';
+	strlcpy(command_line, (char *)(tag->data), COMMAND_LINE_SIZE);
 	return 0;
 }
 
@@ -185,6 +222,58 @@ static int __init parse_bootparam(const bp_tag_t* tag)
 	return 0;
 }
 
+#ifdef CONFIG_OF
+
+void __init early_init_dt_add_memory_arch(u64 base, u64 size)
+{
+	size &= PAGE_MASK;
+	add_sysmem_bank(MEMORY_TYPE_CONVENTIONAL, base, base + size);
+}
+
+void * __init early_init_dt_alloc_memory_arch(u64 size, u64 align)
+{
+	return __alloc_bootmem(size, align, 0);
+}
+
+void __init early_init_devtree(void *params)
+{
+	/* Setup flat device-tree pointer */
+	initial_boot_params = params;
+
+	/* Retrieve various informations from the /chosen node of the
+	 * device-tree, including the platform type, initrd location and
+	 * size, TCE reserve, and more ...
+	 */
+	if (!command_line[0])
+		of_scan_flat_dt(early_init_dt_scan_chosen, command_line);
+
+	/* Scan memory nodes and rebuild MEMBLOCKs */
+	of_scan_flat_dt(early_init_dt_scan_root, NULL);
+	if (sysmem.nr_banks == 0)
+		of_scan_flat_dt(early_init_dt_scan_memory, NULL);
+}
+
+static void __init copy_devtree(void)
+{
+	void *alloc = early_init_dt_alloc_memory_arch(
+			be32_to_cpu(initial_boot_params->totalsize), 0);
+	if (alloc) {
+		memcpy(alloc, initial_boot_params,
+				be32_to_cpu(initial_boot_params->totalsize));
+		initial_boot_params = alloc;
+	}
+}
+
+static int __init xtensa_device_probe(void)
+{
+	of_platform_populate(NULL, NULL, NULL, NULL);
+	return 0;
+}
+
+device_initcall(xtensa_device_probe);
+
+#endif /* CONFIG_OF */
+
 /*
  * Initialize architecture. (Early stage)
  */
@@ -193,15 +282,15 @@ void __init init_arch(bp_tag_t *bp_start)
 {
 	sysmem.nr_banks = 0;
 
-#ifdef CONFIG_CMDLINE_BOOL
-	strcpy(command_line, default_command_line);
-#endif
-
 	/* Parse boot parameters */
 
 	if (bp_start)
 		parse_bootparam(bp_start);
 
+#ifdef CONFIG_OF
+	early_init_devtree(dtb_start);
+#endif
+
 	if (sysmem.nr_banks == 0) {
 		sysmem.nr_banks = 1;
 		sysmem.bank[0].start = PLATFORM_DEFAULT_MEM_START;
@@ -209,6 +298,11 @@ void __init init_arch(bp_tag_t *bp_start)
 				     + PLATFORM_DEFAULT_MEM_SIZE;
 	}
 
+#ifdef CONFIG_CMDLINE_BOOL
+	if (!command_line[0])
+		strlcpy(command_line, default_command_line, COMMAND_LINE_SIZE);
+#endif
+
 	/* Early hook for platforms */
 
 	platform_init(bp_start);
@@ -235,15 +329,130 @@ extern char _UserExceptionVector_text_end;
 extern char _DoubleExceptionVector_literal_start;
 extern char _DoubleExceptionVector_text_end;
 
-void __init setup_arch(char **cmdline_p)
+
+#ifdef CONFIG_S32C1I_SELFTEST
+#if XCHAL_HAVE_S32C1I
+
+static int __initdata rcw_word, rcw_probe_pc, rcw_exc;
+
+/*
+ * Basic atomic compare-and-swap, that records PC of S32C1I for probing.
+ *
+ * If *v == cmp, set *v = set.  Return previous *v.
+ */
+static inline int probed_compare_swap(int *v, int cmp, int set)
+{
+	int tmp;
+
+	__asm__ __volatile__(
+			"	movi	%1, 1f\n"
+			"	s32i	%1, %4, 0\n"
+			"	wsr	%2, scompare1\n"
+			"1:	s32c1i	%0, %3, 0\n"
+			: "=a" (set), "=&a" (tmp)
+			: "a" (cmp), "a" (v), "a" (&rcw_probe_pc), "0" (set)
+			: "memory"
+			);
+	return set;
+}
+
+/* Handle probed exception */
+
+void __init do_probed_exception(struct pt_regs *regs, unsigned long exccause)
+{
+	if (regs->pc == rcw_probe_pc) {	/* exception on s32c1i ? */
+		regs->pc += 3;		/* skip the s32c1i instruction */
+		rcw_exc = exccause;
+	} else {
+		do_unhandled(regs, exccause);
+	}
+}
+
+/* Simple test of S32C1I (soc bringup assist) */
+
+void __init check_s32c1i(void)
 {
-	extern int mem_reserve(unsigned long, unsigned long, int);
-	extern void bootmem_init(void);
+	int n, cause1, cause2;
+	void *handbus, *handdata, *handaddr; /* temporarily saved handlers */
+
+	rcw_probe_pc = 0;
+	handbus  = trap_set_handler(EXCCAUSE_LOAD_STORE_ERROR,
+			do_probed_exception);
+	handdata = trap_set_handler(EXCCAUSE_LOAD_STORE_DATA_ERROR,
+			do_probed_exception);
+	handaddr = trap_set_handler(EXCCAUSE_LOAD_STORE_ADDR_ERROR,
+			do_probed_exception);
+
+	/* First try an S32C1I that does not store: */
+	rcw_exc = 0;
+	rcw_word = 1;
+	n = probed_compare_swap(&rcw_word, 0, 2);
+	cause1 = rcw_exc;
+
+	/* took exception? */
+	if (cause1 != 0) {
+		/* unclean exception? */
+		if (n != 2 || rcw_word != 1)
+			panic("S32C1I exception error");
+	} else if (rcw_word != 1 || n != 1) {
+		panic("S32C1I compare error");
+	}
+
+	/* Then an S32C1I that stores: */
+	rcw_exc = 0;
+	rcw_word = 0x1234567;
+	n = probed_compare_swap(&rcw_word, 0x1234567, 0xabcde);
+	cause2 = rcw_exc;
+
+	if (cause2 != 0) {
+		/* unclean exception? */
+		if (n != 0xabcde || rcw_word != 0x1234567)
+			panic("S32C1I exception error (b)");
+	} else if (rcw_word != 0xabcde || n != 0x1234567) {
+		panic("S32C1I store error");
+	}
+
+	/* Verify consistency of exceptions: */
+	if (cause1 || cause2) {
+		pr_warn("S32C1I took exception %d, %d\n", cause1, cause2);
+		/* If emulation of S32C1I upon bus error gets implemented,
+		   we can get rid of this panic for single core (not SMP) */
+		panic("S32C1I exceptions not currently supported");
+	}
+	if (cause1 != cause2)
+		panic("inconsistent S32C1I exceptions");
+
+	trap_set_handler(EXCCAUSE_LOAD_STORE_ERROR, handbus);
+	trap_set_handler(EXCCAUSE_LOAD_STORE_DATA_ERROR, handdata);
+	trap_set_handler(EXCCAUSE_LOAD_STORE_ADDR_ERROR, handaddr);
+}
 
-	memcpy(boot_command_line, command_line, COMMAND_LINE_SIZE);
-	boot_command_line[COMMAND_LINE_SIZE-1] = '\0';
+#else /* XCHAL_HAVE_S32C1I */
+
+/* This condition should not occur with a commercially deployed processor.
+   Display reminder for early engr test or demo chips / FPGA bitstreams */
+void __init check_s32c1i(void)
+{
+	pr_warn("Processor configuration lacks atomic compare-and-swap support!\n");
+}
+
+#endif /* XCHAL_HAVE_S32C1I */
+#else /* CONFIG_S32C1I_SELFTEST */
+
+void __init check_s32c1i(void)
+{
+}
+
+#endif /* CONFIG_S32C1I_SELFTEST */
+
+
+void __init setup_arch(char **cmdline_p)
+{
+	strlcpy(boot_command_line, command_line, COMMAND_LINE_SIZE);
 	*cmdline_p = command_line;
 
+	check_s32c1i();
+
 	/* Reserve some memory regions */
 
 #ifdef CONFIG_BLK_DEV_INITRD
@@ -275,8 +484,12 @@ void __init setup_arch(char **cmdline_p)
 
 	bootmem_init();
 
-	platform_setup(cmdline_p);
+#ifdef CONFIG_OF
+	copy_devtree();
+	unflatten_device_tree();
+#endif
 
+	platform_setup(cmdline_p);
 
 	paging_init();
 	zones_init();
@@ -380,6 +593,9 @@ c_show(struct seq_file *f, void *slot)
 #endif
 #if XCHAL_HAVE_FP
 		     "fpu "
+#endif
+#if XCHAL_HAVE_S32C1I
+		     "s32c1i "
 #endif
 		     "\n");
 
@@ -412,7 +628,7 @@ c_show(struct seq_file *f, void *slot)
 		     "icache size\t: %d\n"
 		     "icache flags\t: "
 #if XCHAL_ICACHE_LINE_LOCKABLE
-		     "lock"
+		     "lock "
 #endif
 		     "\n"
 		     "dcache line size: %d\n"
@@ -420,10 +636,10 @@ c_show(struct seq_file *f, void *slot)
 		     "dcache size\t: %d\n"
 		     "dcache flags\t: "
 #if XCHAL_DCACHE_IS_WRITEBACK
-		     "writeback"
+		     "writeback "
 #endif
 #if XCHAL_DCACHE_LINE_LOCKABLE
-		     "lock"
+		     "lock "
 #endif
 		     "\n",
 		     XCHAL_ICACHE_LINESIZE,
@@ -465,4 +681,3 @@ const struct seq_operations cpuinfo_op =
 };
 
 #endif /* CONFIG_PROC_FS */
-

arch/xtensa/kernel/syscall.c

@@ -52,4 +52,3 @@ asmlinkage long xtensa_fadvise64_64(int fd, int advice,
 {
 	return sys_fadvise64_64(fd, offset, len, advice);
 }
-

arch/xtensa/kernel/time.c

@@ -22,6 +22,7 @@
 #include <linux/irq.h>
 #include <linux/profile.h>
 #include <linux/delay.h>
+#include <linux/irqdomain.h>
 
 #include <asm/timex.h>
 #include <asm/platform.h>
@@ -31,7 +32,7 @@ unsigned long ccount_per_jiffy;		/* per 1/HZ */
 unsigned long nsec_per_ccount;		/* nsec per ccount increment */
 #endif
 
-static cycle_t ccount_read(void)
+static cycle_t ccount_read(struct clocksource *cs)
 {
 	return (cycle_t)get_ccount();
 }
@@ -52,6 +53,7 @@ static struct irqaction timer_irqaction = {
 
 void __init time_init(void)
 {
+	unsigned int irq;
 #ifdef CONFIG_XTENSA_CALIBRATE_CCOUNT
 	printk("Calibrating CPU frequency ");
 	platform_calibrate_ccount();
@@ -62,7 +64,8 @@ void __init time_init(void)
 
 	/* Initialize the linux timer interrupt. */
 
-	setup_irq(LINUX_TIMER_INT, &timer_irqaction);
+	irq = irq_create_mapping(NULL, LINUX_TIMER_INT);
+	setup_irq(irq, &timer_irqaction);
 	set_linux_timer(get_ccount() + CCOUNT_PER_JIFFY);
 }
 

arch/xtensa/kernel/traps.c

@@ -293,6 +293,17 @@ do_debug(struct pt_regs *regs)
 }
 
 
+/* Set exception C handler - for temporary use when probing exceptions */
+
+void * __init trap_set_handler(int cause, void *handler)
+{
+	unsigned long *entry = &exc_table[EXC_TABLE_DEFAULT / 4 + cause];
+	void *previous = (void *)*entry;
+	*entry = (unsigned long)handler;
+	return previous;
+}
+
+
 /*
  * Initialize dispatch tables.
  *
@@ -397,7 +408,8 @@ static inline void spill_registers(void)
 		"wsr	a13, sar\n\t"
 		"wsr	a14, ps\n\t"
 		:: "a" (&a0), "a" (&ps)
-		: "a2", "a3", "a4", "a7", "a11", "a12", "a13", "a14", "a15", "memory");
+		: "a2", "a3", "a4", "a7", "a11", "a12", "a13", "a14", "a15",
+		  "memory");
 }
 
 void show_trace(struct task_struct *task, unsigned long *sp)
@@ -523,5 +535,3 @@ void die(const char * str, struct pt_regs * regs, long err)
 
 	do_exit(err);
 }
-
-

arch/xtensa/kernel/vectors.S

@@ -79,6 +79,8 @@ ENTRY(_UserExceptionVector)
 	l32i	a0, a0, EXC_TABLE_FAST_USER	# load handler
 	jx	a0
 
+ENDPROC(_UserExceptionVector)
+
 /*
  * Kernel exception vector. (Exceptions with PS.UM == 0, PS.EXCM == 0)
  *
@@ -103,6 +105,7 @@ ENTRY(_KernelExceptionVector)
 	l32i	a0, a0, EXC_TABLE_FAST_KERNEL	# load handler address
 	jx	a0
 
+ENDPROC(_KernelExceptionVector)
 
 /*
  * Double exception vector (Exceptions with PS.EXCM == 1)
@@ -225,7 +228,13 @@ ENTRY(_DoubleExceptionVector)
 	/* Window overflow/underflow exception. Get stack pointer. */
 
 	mov	a3, a2
-	movi	a2, exc_table
+	/* This explicit literal and the following references to it are made
+	 * in order to fit DoubleExceptionVector.literals into the available
+	 * 16-byte gap before DoubleExceptionVector.text in the absence of
+	 * link time relaxation. See kernel/vmlinux.lds.S
+	 */
+	.literal .Lexc_table, exc_table
+	l32r	a2, .Lexc_table
 	l32i	a2, a2, EXC_TABLE_KSTK
 
 	/* Check for overflow/underflow exception, jump if overflow. */
@@ -255,7 +264,7 @@ ENTRY(_DoubleExceptionVector)
 	s32i	a0, a2, PT_AREG0
 
 	wsr	a3, excsave1		# save a3
-	movi	a3, exc_table
+	l32r	a3, .Lexc_table
 
 	rsr	a0, exccause
 	s32i	a0, a2, PT_DEPC		# mark it as a regular exception
@@ -267,7 +276,7 @@ ENTRY(_DoubleExceptionVector)
 
 	/* a0: depc, a1: a1, a2: a2, a3: trashed, depc: a0, excsave1: a3 */
 
-	movi	a3, exc_table
+	l32r	a3, .Lexc_table
 	s32i	a2, a3, EXC_TABLE_DOUBLE_SAVE	# temporary variable
 
 	/* Enter critical section. */
@@ -296,7 +305,7 @@ ENTRY(_DoubleExceptionVector)
 
 	/* a0: avail, a1: a1, a2: kstk, a3: avail, depc: a2, excsave: a3 */
 
-	movi	a3, exc_table
+	l32r	a3, .Lexc_table
 	rsr	a0, exccause
 	addx4	a0, a0, a3
 	l32i	a0, a0, EXC_TABLE_FAST_USER
@@ -338,6 +347,7 @@ ENTRY(_DoubleExceptionVector)
 
 	.end literal_prefix
 
+ENDPROC(_DoubleExceptionVector)
 
 /*
  * Debug interrupt vector
@@ -349,9 +359,11 @@ ENTRY(_DoubleExceptionVector)
 	.section .DebugInterruptVector.text, "ax"
 
 ENTRY(_DebugInterruptVector)
+
 	xsr	a0, SREG_EXCSAVE + XCHAL_DEBUGLEVEL
 	jx	a0
 
+ENDPROC(_DebugInterruptVector)
 
 
 /* Window overflow and underflow handlers.
@@ -363,38 +375,43 @@ ENTRY(_DebugInterruptVector)
  *	 we try to access any page that would cause a page fault early.
  */
 
+#define ENTRY_ALIGN64(name)	\
+	.globl name;		\
+	.align 64;		\
+	name:
+
 	.section		.WindowVectors.text, "ax"
 
 
 /* 4-Register Window Overflow Vector (Handler) */
 
-	.align 64
-.global _WindowOverflow4
-_WindowOverflow4:
+ENTRY_ALIGN64(_WindowOverflow4)
+
 	s32e	a0, a5, -16
 	s32e	a1, a5, -12
 	s32e	a2, a5,  -8
 	s32e	a3, a5,  -4
 	rfwo
 
+ENDPROC(_WindowOverflow4)
+
 
 /* 4-Register Window Underflow Vector (Handler) */
 
-	.align 64
-.global _WindowUnderflow4
-_WindowUnderflow4:
+ENTRY_ALIGN64(_WindowUnderflow4)
+
 	l32e	a0, a5, -16
 	l32e	a1, a5, -12
 	l32e	a2, a5,  -8
 	l32e	a3, a5,  -4
 	rfwu
 
+ENDPROC(_WindowUnderflow4)
 
 /* 8-Register Window Overflow Vector (Handler) */
 
-	.align 64
-.global _WindowOverflow8
-_WindowOverflow8:
+ENTRY_ALIGN64(_WindowOverflow8)
+
 	s32e	a0, a9, -16
 	l32e	a0, a1, -12
 	s32e	a2, a9,  -8
@@ -406,11 +423,12 @@ _WindowOverflow8:
 	s32e	a7, a0, -20
 	rfwo
 
+ENDPROC(_WindowOverflow8)
+
 /* 8-Register Window Underflow Vector (Handler) */
 
-	.align 64
-.global _WindowUnderflow8
-_WindowUnderflow8:
+ENTRY_ALIGN64(_WindowUnderflow8)
+
 	l32e	a1, a9, -12
 	l32e	a0, a9, -16
 	l32e	a7, a1, -12
@@ -422,12 +440,12 @@ _WindowUnderflow8:
 	l32e	a7, a7, -20
 	rfwu
 
+ENDPROC(_WindowUnderflow8)
 
 /* 12-Register Window Overflow Vector (Handler) */
 
-	.align 64
-.global _WindowOverflow12
-_WindowOverflow12:
+ENTRY_ALIGN64(_WindowOverflow12)
+
 	s32e	a0,  a13, -16
 	l32e	a0,  a1,  -12
 	s32e	a1,  a13, -12
@@ -443,11 +461,12 @@ _WindowOverflow12:
 	s32e	a11, a0,  -20
 	rfwo
 
+ENDPROC(_WindowOverflow12)
+
 /* 12-Register Window Underflow Vector (Handler) */
 
-	.align 64
-.global _WindowUnderflow12
-_WindowUnderflow12:
+ENTRY_ALIGN64(_WindowUnderflow12)
+
 	l32e	a1,  a13, -12
 	l32e	a0,  a13, -16
 	l32e	a11, a1,  -12
@@ -463,6 +482,6 @@ _WindowUnderflow12:
 	l32e	a11, a11, -20
 	rfwu
 
-	.text
-
+ENDPROC(_WindowUnderflow12)
 
+	.text

arch/xtensa/lib/checksum.S

@@ -41,6 +41,7 @@
 
 .text
 ENTRY(csum_partial)
+
 	/*
 	 * Experiments with Ethernet and SLIP connections show that buf
 	 * is aligned on either a 2-byte or 4-byte boundary.
@@ -170,7 +171,7 @@ ENTRY(csum_partial)
 3:
 	j	5b		/* branch to handle the remaining byte */
 
-
+ENDPROC(csum_partial)
 
 /*
  * Copy from ds while checksumming, otherwise like csum_partial
@@ -211,6 +212,7 @@ unsigned int csum_partial_copy_generic (const char *src, char *dst, int len,
  */
 
 ENTRY(csum_partial_copy_generic)
+
 	entry	sp, 32
 	mov	a12, a3
 	mov	a11, a4
@@ -367,6 +369,8 @@ DST(	s8i	a8, a3, 1	)
 6:
 	j	4b		/* process the possible trailing odd byte */
 
+ENDPROC(csum_partial_copy_generic)
+
 
 # Exception handler:
 .section .fixup, "ax"
@@ -406,4 +410,3 @@ DST(	s8i	a8, a3, 1	)
 	retw
 
 .previous
-

arch/xtensa/lib/memcopy.S

@@ -210,8 +210,10 @@ memcpy:
 	_beqz	a4, .Ldone	# avoid loading anything for zero-length copies
 	# copy 16 bytes per iteration for word-aligned dst and unaligned src
 	ssa8	a3		# set shift amount from byte offset
-#define SIM_CHECKS_ALIGNMENT	1	/* set to 1 when running on ISS (simulator) with the
+
+/* set to 1 when running on ISS (simulator) with the
    lint or ferret client, or 0 to save a few cycles */
+#define SIM_CHECKS_ALIGNMENT	1
 #if XCHAL_UNALIGNED_LOAD_EXCEPTION || SIM_CHECKS_ALIGNMENT
 	and	a11, a3, a8	# save unalignment offset for below
 	sub	a3, a3, a11	# align a3

arch/xtensa/lib/pci-auto.c

@@ -345,8 +345,3 @@ int __init pciauto_bus_scan(struct pci_controller *pci_ctrl, int current_bus)
 	}
 	return sub_bus;
 }
-
-
-
-
-

arch/xtensa/lib/strnlen_user.S

@@ -145,4 +145,3 @@ __strnlen_user:
 lenfixup:
 	movi	a2, 0
 	retw
-

arch/xtensa/lib/usercopy.S

@@ -318,4 +318,3 @@ l_fixup:
 	/* Ignore memset return value in a6. */
 	/* a2 still contains bytes not copied. */
 	retw
-

arch/xtensa/mm/cache.c

@@ -166,14 +166,14 @@ update_mmu_cache(struct vm_area_struct * vma, unsigned long addr, pte_t *ptep)
 
 	if (!PageReserved(page) && test_bit(PG_arch_1, &page->flags)) {
 
-		unsigned long vaddr = TLBTEMP_BASE_1 + (addr & DCACHE_ALIAS_MASK);
 		unsigned long paddr = (unsigned long) page_address(page);
 		unsigned long phys = page_to_phys(page);
+		unsigned long tmp = TLBTEMP_BASE_1 + (addr & DCACHE_ALIAS_MASK);
 
 		__flush_invalidate_dcache_page(paddr);
 
-		__flush_invalidate_dcache_page_alias(vaddr, phys);
-		__invalidate_icache_page_alias(vaddr, phys);
+		__flush_invalidate_dcache_page_alias(tmp, phys);
+		__invalidate_icache_page_alias(tmp, phys);
 
 		clear_bit(PG_arch_1, &page->flags);
 	}
@@ -205,8 +205,8 @@ void copy_to_user_page(struct vm_area_struct *vma, struct page *page,
 	/* Flush and invalidate user page if aliased. */
 
 	if (alias) {
-		unsigned long temp = TLBTEMP_BASE_1 + (vaddr & DCACHE_ALIAS_MASK);
-		__flush_invalidate_dcache_page_alias(temp, phys);
+		unsigned long t = TLBTEMP_BASE_1 + (vaddr & DCACHE_ALIAS_MASK);
+		__flush_invalidate_dcache_page_alias(t, phys);
 	}
 
 	/* Copy data */
@@ -219,12 +219,11 @@ void copy_to_user_page(struct vm_area_struct *vma, struct page *page,
 	 */
 
 	if (alias) {
-		unsigned long temp = TLBTEMP_BASE_1 + (vaddr & DCACHE_ALIAS_MASK);
+		unsigned long t = TLBTEMP_BASE_1 + (vaddr & DCACHE_ALIAS_MASK);
 
 		__flush_invalidate_dcache_range((unsigned long) dst, len);
-		if ((vma->vm_flags & VM_EXEC) != 0) {
-			__invalidate_icache_page_alias(temp, phys);
-		}
+		if ((vma->vm_flags & VM_EXEC) != 0)
+			__invalidate_icache_page_alias(t, phys);
 
 	} else if ((vma->vm_flags & VM_EXEC) != 0) {
 		__flush_dcache_range((unsigned long)dst,len);
@@ -245,8 +244,8 @@ extern void copy_from_user_page(struct vm_area_struct *vma, struct page *page,
 	 */
 
 	if (alias) {
-		unsigned long temp = TLBTEMP_BASE_1 + (vaddr & DCACHE_ALIAS_MASK);
-		__flush_invalidate_dcache_page_alias(temp, phys);
+		unsigned long t = TLBTEMP_BASE_1 + (vaddr & DCACHE_ALIAS_MASK);
+		__flush_invalidate_dcache_page_alias(t, phys);
 	}
 
 	memcpy(dst, src, len);

arch/xtensa/mm/fault.c

@@ -254,4 +254,3 @@ bad_page_fault(struct pt_regs *regs, unsigned long address, int sig)
 	die("Oops", regs, sig);
 	do_exit(sig);
 }
-

arch/xtensa/mm/init.c

@@ -75,16 +75,16 @@ int __init mem_reserve(unsigned long start, unsigned long end, int must_exist)
 			sysmem.nr_banks++;
 		}
 		sysmem.bank[i].end = start;
-	} else {
-		if (end < sysmem.bank[i].end)
+
+	} else if (end < sysmem.bank[i].end) {
 		sysmem.bank[i].start = end;
-		else {
+
+	} else {
 		/* remove entry */
 		sysmem.nr_banks--;
 		sysmem.bank[i].start = sysmem.bank[sysmem.nr_banks].start;
 		sysmem.bank[i].end   = sysmem.bank[sysmem.nr_banks].end;
 	}
-	}
 	return -1;
 }
 

arch/xtensa/mm/misc.S

@@ -29,6 +29,7 @@
  */
 
 ENTRY(clear_page)
+
 	entry	a1, 16
 
 	movi	a3, 0
@@ -45,6 +46,8 @@ ENTRY(clear_page)
 
 	retw
 
+ENDPROC(clear_page)
+
 /*
  * copy_page and copy_user_page are the same for non-cache-aliased configs.
  *
@@ -53,6 +56,7 @@ ENTRY(clear_page)
  */
 
 ENTRY(copy_page)
+
 	entry	a1, 16
 
 	__loopi a2, a4, PAGE_SIZE, 32
@@ -84,6 +88,8 @@ ENTRY(copy_page)
 
 	retw
 
+ENDPROC(copy_page)
+
 #ifdef CONFIG_MMU
 /*
  * If we have to deal with cache aliasing, we use temporary memory mappings
@@ -109,6 +115,7 @@ ENTRY(__tlbtemp_mapping_start)
  */
 
 ENTRY(clear_user_page)
+
 	entry	a1, 32
 
 	/* Mark page dirty and determine alias. */
@@ -164,6 +171,8 @@ ENTRY(clear_user_page)
 
 	retw
 
+ENDPROC(clear_user_page)
+
 /*
  * copy_page_user (void *to, void *from, unsigned long vaddr, struct page *page)
  *                    a2          a3	        a4		    a5
@@ -262,6 +271,8 @@ ENTRY(copy_user_page)
 
 	retw
 
+ENDPROC(copy_user_page)
+
 #endif
 
 #if (DCACHE_WAY_SIZE > PAGE_SIZE)
@@ -272,6 +283,7 @@ ENTRY(copy_user_page)
  */
 
 ENTRY(__flush_invalidate_dcache_page_alias)
+
 	entry	sp, 16
 
 	movi	a7, 0			# required for exception handler
@@ -287,6 +299,7 @@ ENTRY(__flush_invalidate_dcache_page_alias)
 
 	retw
 
+ENDPROC(__flush_invalidate_dcache_page_alias)
 #endif
 
 ENTRY(__tlbtemp_mapping_itlb)
@@ -294,6 +307,7 @@ ENTRY(__tlbtemp_mapping_itlb)
 #if (ICACHE_WAY_SIZE > PAGE_SIZE)
 	
 ENTRY(__invalidate_icache_page_alias)
+
 	entry	sp, 16
 
 	addi	a6, a3, (PAGE_KERNEL_EXEC | _PAGE_HW_WRITE)
@@ -307,11 +321,14 @@ ENTRY(__invalidate_icache_page_alias)
 	isync
 	retw
 
+ENDPROC(__invalidate_icache_page_alias)
+
 #endif
 
 /* End of special treatment in tlb miss exception */
 
 ENTRY(__tlbtemp_mapping_end)
+
 #endif /* CONFIG_MMU
 
 /*
@@ -319,6 +336,7 @@ ENTRY(__tlbtemp_mapping_end)
  */
 
 ENTRY(__invalidate_icache_page)
+
 	entry	sp, 16
 
 	___invalidate_icache_page a2 a3
@@ -326,11 +344,14 @@ ENTRY(__invalidate_icache_page)
 
 	retw
 
+ENDPROC(__invalidate_icache_page)
+
 /*
  * void __invalidate_dcache_page(ulong start)
  */
 
 ENTRY(__invalidate_dcache_page)
+
 	entry	sp, 16
 
 	___invalidate_dcache_page a2 a3
@@ -338,11 +359,14 @@ ENTRY(__invalidate_dcache_page)
 
 	retw
 
+ENDPROC(__invalidate_dcache_page)
+
 /*
  * void __flush_invalidate_dcache_page(ulong start)
  */
 
 ENTRY(__flush_invalidate_dcache_page)
+
 	entry	sp, 16
 
 	___flush_invalidate_dcache_page a2 a3
@@ -350,11 +374,14 @@ ENTRY(__flush_invalidate_dcache_page)
 	dsync
 	retw
 
+ENDPROC(__flush_invalidate_dcache_page)
+
 /*
  * void __flush_dcache_page(ulong start)
  */
 
 ENTRY(__flush_dcache_page)
+
 	entry	sp, 16
 
 	___flush_dcache_page a2 a3
@@ -362,11 +389,14 @@ ENTRY(__flush_dcache_page)
 	dsync
 	retw
 
+ENDPROC(__flush_dcache_page)
+
 /*
  * void __invalidate_icache_range(ulong start, ulong size)
  */
 
 ENTRY(__invalidate_icache_range)
+
 	entry	sp, 16
 
 	___invalidate_icache_range a2 a3 a4
@@ -374,11 +404,14 @@ ENTRY(__invalidate_icache_range)
 
 	retw
 
+ENDPROC(__invalidate_icache_range)
+
 /*
  * void __flush_invalidate_dcache_range(ulong start, ulong size)
  */
 
 ENTRY(__flush_invalidate_dcache_range)
+
 	entry	sp, 16
 
 	___flush_invalidate_dcache_range a2 a3 a4
@@ -386,11 +419,14 @@ ENTRY(__flush_invalidate_dcache_range)
 
 	retw
 
+ENDPROC(__flush_invalidate_dcache_range)
+
 /*
  * void _flush_dcache_range(ulong start, ulong size)
  */
 
 ENTRY(__flush_dcache_range)
+
 	entry	sp, 16
 
 	___flush_dcache_range a2 a3 a4
@@ -398,22 +434,28 @@ ENTRY(__flush_dcache_range)
 
 	retw
 
+ENDPROC(__flush_dcache_range)
+
 /*
  * void _invalidate_dcache_range(ulong start, ulong size)
  */
 
 ENTRY(__invalidate_dcache_range)
+
 	entry	sp, 16
 
 	___invalidate_dcache_range a2 a3 a4
 
 	retw
 
+ENDPROC(__invalidate_dcache_range)
+
 /*
  * void _invalidate_icache_all(void)
  */
 
 ENTRY(__invalidate_icache_all)
+
 	entry	sp, 16
 
 	___invalidate_icache_all a2 a3
@@ -421,11 +463,14 @@ ENTRY(__invalidate_icache_all)
 
 	retw
 
+ENDPROC(__invalidate_icache_all)
+
 /*
  * void _flush_invalidate_dcache_all(void)
  */
 
 ENTRY(__flush_invalidate_dcache_all)
+
 	entry	sp, 16
 
 	___flush_invalidate_dcache_all a2 a3
@@ -433,11 +478,14 @@ ENTRY(__flush_invalidate_dcache_all)
 
 	retw
 
+ENDPROC(__flush_invalidate_dcache_all)
+
 /*
  * void _invalidate_dcache_all(void)
  */
 
 ENTRY(__invalidate_dcache_all)
+
 	entry	sp, 16
 
 	___invalidate_dcache_all a2 a3
@@ -445,3 +493,4 @@ ENTRY(__invalidate_dcache_all)
 
 	retw
 
+ENDPROC(__invalidate_dcache_all)

arch/xtensa/mm/mmu.c

@@ -37,7 +37,7 @@ void __init init_mmu(void)
 
 	/* Set rasid register to a known value. */
 
-	set_rasid_register(ASID_USER_FIRST);
+	set_rasid_register(ASID_INSERT(ASID_USER_FIRST));
 
 	/* Set PTEVADDR special register to the start of the page
 	 * table, which is in kernel mappable space (ie. not

arch/xtensa/mm/tlb.c

@@ -63,7 +63,7 @@ void flush_tlb_all (void)
 void flush_tlb_mm(struct mm_struct *mm)
 {
 	if (mm == current->active_mm) {
-		int flags;
+		unsigned long flags;
 		local_save_flags(flags);
 		__get_new_mmu_context(mm);
 		__load_mmu_context(mm);
@@ -140,4 +140,3 @@ void flush_tlb_page (struct vm_area_struct *vma, unsigned long page)
 
 	local_irq_restore(flags);
 }
-

arch/xtensa/platforms/iss/include/platform/serial.h

+/*
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2012 Tensilica Inc.
+ */
+
+#ifndef __ASM_XTENSA_ISS_SERIAL_H
+#define __ASM_XTENSA_ISS_SERIAL_H
+
+/* Have no meaning on ISS, but needed for 8250_early.c */
+#define BASE_BAUD 0
+
+#endif /* __ASM_XTENSA_ISS_SERIAL_H */

arch/xtensa/platforms/iss/include/platform/simcall.h

@@ -74,13 +74,12 @@ static inline int __simc(int a, int b, int c, int d, int e, int f)
 			"mov %1, a3\n"
 			: "=a" (ret), "=a" (errno), "+r"(a1), "+r"(b1)
 			: "r"(c1), "r"(d1), "r"(e1), "r"(f1)
-			: );
+			: "memory");
 	return ret;
 }
 
 static inline int simc_open(const char *file, int flags, int mode)
 {
-	wmb();
 	return __simc(SYS_open, (int) file, flags, mode, 0, 0);
 }
 
@@ -91,19 +90,16 @@ static inline int simc_close(int fd)
 
 static inline int simc_ioctl(int fd, int request, void *arg)
 {
-	wmb();
 	return __simc(SYS_ioctl, fd, request, (int) arg, 0, 0);
 }
 
 static inline int simc_read(int fd, void *buf, size_t count)
 {
-	rmb();
 	return __simc(SYS_read, fd, (int) buf, count, 0, 0);
 }
 
 static inline int simc_write(int fd, const void *buf, size_t count)
 {
-	wmb();
 	return __simc(SYS_write, fd, (int) buf, count, 0, 0);
 }
 
@@ -111,7 +107,6 @@ static inline int simc_poll(int fd)
 {
 	struct timeval tv = { .tv_sec = 0, .tv_usec = 0 };
 
-	wmb();
 	return __simc(SYS_select_one, fd, XTISS_SELECT_ONE_READ, (int)&tv,
 			0, 0);
 }

arch/xtensa/platforms/xtfpga/Makefile

+# Makefile for the Tensilica xtavnet Emulation Board
+#
+# Note! Dependencies are done automagically by 'make dep', which also
+# removes any old dependencies. DON'T put your own dependencies here
+# unless it's something special (ie not a .c file).
+#
+# Note 2! The CFLAGS definitions are in the main makefile...
+
+obj-y			= setup.o lcd.o

arch/xtensa/platforms/xtfpga/include/platform/hardware.h

+/*
+ * arch/xtensa/platform/xtavnet/include/platform/hardware.h
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2006 Tensilica Inc.
+ */
+
+/*
+ * This file contains the hardware configuration of the XTAVNET boards.
+ */
+
+#ifndef __XTENSA_XTAVNET_HARDWARE_H
+#define __XTENSA_XTAVNET_HARDWARE_H
+
+/* By default NO_IRQ is defined to 0 in Linux, but we use the
+   interrupt 0 for UART... */
+#define NO_IRQ                 -1
+
+/* Memory configuration. */
+
+#define PLATFORM_DEFAULT_MEM_START 0x00000000
+#define PLATFORM_DEFAULT_MEM_SIZE  0x04000000
+
+/* Interrupt configuration. */
+
+#define PLATFORM_NR_IRQS	10
+
+/* Default assignment of LX60 devices to external interrupts. */
+
+#ifdef CONFIG_ARCH_HAS_SMP
+#define DUART16552_INTNUM	XCHAL_EXTINT3_NUM
+#define OETH_IRQ		XCHAL_EXTINT4_NUM
+#else
+#define DUART16552_INTNUM	XCHAL_EXTINT0_NUM
+#define OETH_IRQ		XCHAL_EXTINT1_NUM
+#endif
+
+/*
+ *  Device addresses and parameters.
+ */
+
+/* UART */
+#define DUART16552_PADDR	(XCHAL_KIO_PADDR + 0x0D050020)
+/* LCD instruction and data addresses. */
+#define LCD_INSTR_ADDR		((char *)IOADDR(0x0D040000))
+#define LCD_DATA_ADDR		((char *)IOADDR(0x0D040004))
+
+/* Misc. */
+#define XTFPGA_FPGAREGS_VADDR	IOADDR(0x0D020000)
+/* Clock frequency in Hz (read-only):  */
+#define XTFPGA_CLKFRQ_VADDR	(XTFPGA_FPGAREGS_VADDR + 0x04)
+/* Setting of 8 DIP switches:  */
+#define DIP_SWITCHES_VADDR	(XTFPGA_FPGAREGS_VADDR + 0x0C)
+/* Software reset (write 0xdead):  */
+#define XTFPGA_SWRST_VADDR	(XTFPGA_FPGAREGS_VADDR + 0x10)
+
+/*  OpenCores Ethernet controller:  */
+				/* regs + RX/TX descriptors */
+#define OETH_REGS_PADDR		(XCHAL_KIO_PADDR + 0x0D030000)
+#define OETH_REGS_SIZE		0x1000
+#define OETH_SRAMBUFF_PADDR	(XCHAL_KIO_PADDR + 0x0D800000)
+
+				/* 5*rx buffs + 5*tx buffs */
+#define OETH_SRAMBUFF_SIZE	(5 * 0x600 + 5 * 0x600)
+
+#endif /* __XTENSA_XTAVNET_HARDWARE_H */

arch/xtensa/platforms/xtfpga/include/platform/lcd.h

+/*
+ * arch/xtensa/platform/xtavnet/include/platform/lcd.h
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001, 2006 Tensilica Inc.
+ */
+
+#ifndef __XTENSA_XTAVNET_LCD_H
+#define __XTENSA_XTAVNET_LCD_H
+
+/* Display string STR at position POS on the LCD. */
+void lcd_disp_at_pos(char *str, unsigned char pos);
+
+/* Shift the contents of the LCD display left or right. */
+void lcd_shiftleft(void);
+void lcd_shiftright(void);
+#endif

arch/xtensa/platforms/xtfpga/include/platform/serial.h

+/*
+ * arch/xtensa/platform/xtavnet/include/platform/serial.h
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001, 2006 Tensilica Inc.
+ */
+
+#ifndef __ASM_XTENSA_XTAVNET_SERIAL_H
+#define __ASM_XTENSA_XTAVNET_SERIAL_H
+
+#include <platform/hardware.h>
+
+#define BASE_BAUD (*(long *)XTFPGA_CLKFRQ_VADDR / 16)
+
+#endif /* __ASM_XTENSA_XTAVNET_SERIAL_H */

arch/xtensa/platforms/xtfpga/lcd.c

+/*
+ * Driver for the LCD display on the Tensilica LX60 Board.
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001, 2006 Tensilica Inc.
+ */
+
+/*
+ *
+ * FIXME: this code is from the examples from the LX60 user guide.
+ *
+ * The lcd_pause function does busy waiting, which is probably not
+ * great. Maybe the code could be changed to use kernel timers, or
+ * change the hardware to not need to wait.
+ */
+
+#include <linux/init.h>
+#include <linux/io.h>
+
+#include <platform/hardware.h>
+#include <platform/lcd.h>
+#include <linux/delay.h>
+
+#define LCD_PAUSE_ITERATIONS	4000
+#define LCD_CLEAR		0x1
+#define LCD_DISPLAY_ON		0xc
+
+/* 8bit and 2 lines display */
+#define LCD_DISPLAY_MODE8BIT	0x38
+#define LCD_DISPLAY_POS		0x80
+#define LCD_SHIFT_LEFT		0x18
+#define LCD_SHIFT_RIGHT		0x1c
+
+static int __init lcd_init(void)
+{
+	*LCD_INSTR_ADDR = LCD_DISPLAY_MODE8BIT;
+	mdelay(5);
+	*LCD_INSTR_ADDR = LCD_DISPLAY_MODE8BIT;
+	udelay(200);
+	*LCD_INSTR_ADDR = LCD_DISPLAY_MODE8BIT;
+	udelay(50);
+	*LCD_INSTR_ADDR = LCD_DISPLAY_ON;
+	udelay(50);
+	*LCD_INSTR_ADDR = LCD_CLEAR;
+	mdelay(10);
+	lcd_disp_at_pos("XTENSA LINUX", 0);
+	return 0;
+}
+
+void lcd_disp_at_pos(char *str, unsigned char pos)
+{
+	*LCD_INSTR_ADDR = LCD_DISPLAY_POS | pos;
+	udelay(100);
+	while (*str != 0) {
+		*LCD_DATA_ADDR = *str;
+		udelay(200);
+		str++;
+	}
+}
+
+void lcd_shiftleft(void)
+{
+	*LCD_INSTR_ADDR = LCD_SHIFT_LEFT;
+	udelay(50);
+}
+
+void lcd_shiftright(void)
+{
+	*LCD_INSTR_ADDR = LCD_SHIFT_RIGHT;
+	udelay(50);
+}
+
+arch_initcall(lcd_init);

arch/xtensa/platforms/xtfpga/setup.c

+/*
+ *
+ * arch/xtensa/platform/xtavnet/setup.c
+ *
+ * ...
+ *
+ * Authors:	Chris Zankel <chris@zankel.net>
+ *		Joe Taylor <joe@tensilica.com>
+ *
+ * Copyright 2001 - 2006 Tensilica Inc.
+ *
+ * This program is free software; you can redistribute  it and/or modify it
+ * under  the terms of  the GNU General  Public License as published by the
+ * Free Software Foundation;  either version 2 of the  License, or (at your
+ * option) any later version.
+ *
+ */
+#include <linux/stddef.h>
+#include <linux/kernel.h>
+#include <linux/init.h>
+#include <linux/errno.h>
+#include <linux/reboot.h>
+#include <linux/kdev_t.h>
+#include <linux/types.h>
+#include <linux/major.h>
+#include <linux/console.h>
+#include <linux/delay.h>
+#include <linux/of.h>
+
+#include <asm/timex.h>
+#include <asm/processor.h>
+#include <asm/platform.h>
+#include <asm/bootparam.h>
+#include <platform/lcd.h>
+#include <platform/hardware.h>
+
+void platform_halt(void)
+{
+	lcd_disp_at_pos(" HALT ", 0);
+	local_irq_disable();
+	while (1)
+		cpu_relax();
+}
+
+void platform_power_off(void)
+{
+	lcd_disp_at_pos("POWEROFF", 0);
+	local_irq_disable();
+	while (1)
+		cpu_relax();
+}
+
+void platform_restart(void)
+{
+	/* Flush and reset the mmu, simulate a processor reset, and
+	 * jump to the reset vector. */
+
+
+	__asm__ __volatile__ ("movi	a2, 15\n\t"
+			      "wsr	a2, icountlevel\n\t"
+			      "movi	a2, 0\n\t"
+			      "wsr	a2, icount\n\t"
+			      "wsr	a2, ibreakenable\n\t"
+			      "wsr	a2, lcount\n\t"
+			      "movi	a2, 0x1f\n\t"
+			      "wsr	a2, ps\n\t"
+			      "isync\n\t"
+			      "jx	%0\n\t"
+			      :
+			      : "a" (XCHAL_RESET_VECTOR_VADDR)
+			      : "a2"
+			      );
+
+	/* control never gets here */
+}
+
+void __init platform_setup(char **cmdline)
+{
+}
+
+#ifdef CONFIG_OF
+
+static void __init update_clock_frequency(struct device_node *node)
+{
+	struct property *newfreq;
+	u32 freq;
+
+	if (!of_property_read_u32(node, "clock-frequency", &freq) && freq != 0)
+		return;
+
+	newfreq = kzalloc(sizeof(*newfreq) + sizeof(u32), GFP_KERNEL);
+	if (!newfreq)
+		return;
+	newfreq->value = newfreq + 1;
+	newfreq->length = sizeof(freq);
+	newfreq->name = kstrdup("clock-frequency", GFP_KERNEL);
+	if (!newfreq->name) {
+		kfree(newfreq);
+		return;
+	}
+
+	*(u32 *)newfreq->value = cpu_to_be32(*(u32 *)XTFPGA_CLKFRQ_VADDR);
+	prom_update_property(node, newfreq);
+}
+
+#define MAC_LEN 6
+static void __init update_local_mac(struct device_node *node)
+{
+	struct property *newmac;
+	const u8* macaddr;
+	int prop_len;
+
+	macaddr = of_get_property(node, "local-mac-address", &prop_len);
+	if (macaddr == NULL || prop_len != MAC_LEN)
+		return;
+
+	newmac = kzalloc(sizeof(*newmac) + MAC_LEN, GFP_KERNEL);
+	if (newmac == NULL)
+		return;
+
+	newmac->value = newmac + 1;
+	newmac->length = MAC_LEN;
+	newmac->name = kstrdup("local-mac-address", GFP_KERNEL);
+	if (newmac->name == NULL) {
+		kfree(newmac);
+		return;
+	}
+
+	memcpy(newmac->value, macaddr, MAC_LEN);
+	((u8*)newmac->value)[5] = (*(u32*)DIP_SWITCHES_VADDR) & 0x3f;
+	prom_update_property(node, newmac);
+}
+
+static int __init machine_setup(void)
+{
+	struct device_node *serial;
+	struct device_node *eth = NULL;
+
+	for_each_compatible_node(serial, NULL, "ns16550a")
+		update_clock_frequency(serial);
+
+	if ((eth = of_find_compatible_node(eth, NULL, "opencores,ethoc")))
+		update_local_mac(eth);
+	return 0;
+}
+arch_initcall(machine_setup);
+
+#endif
+
+/* early initialization */
+
+void __init platform_init(bp_tag_t *first)
+{
+}
+
+/* Heartbeat. */
+
+void platform_heartbeat(void)
+{
+}
+
+#ifdef CONFIG_XTENSA_CALIBRATE_CCOUNT
+
+void platform_calibrate_ccount(void)
+{
+	long clk_freq = 0;
+#ifdef CONFIG_OF
+	struct device_node *cpu =
+		of_find_compatible_node(NULL, NULL, "xtensa,cpu");
+	if (cpu) {
+		u32 freq;
+		update_clock_frequency(cpu);
+		if (!of_property_read_u32(cpu, "clock-frequency", &freq))
+			clk_freq = freq;
+	}
+#endif
+	if (!clk_freq)
+		clk_freq = *(long *)XTFPGA_CLKFRQ_VADDR;
+
+	ccount_per_jiffy = clk_freq / HZ;
+	nsec_per_ccount = 1000000000UL / clk_freq;
+}
+
+#endif
+
+#ifndef CONFIG_OF
+
+#include <linux/serial_8250.h>
+#include <linux/if.h>
+#include <net/ethoc.h>
+
+/*----------------------------------------------------------------------------
+ *  Ethernet -- OpenCores Ethernet MAC (ethoc driver)
+ */
+
+static struct resource ethoc_res[] __initdata = {
+	[0] = { /* register space */
+		.start = OETH_REGS_PADDR,
+		.end   = OETH_REGS_PADDR + OETH_REGS_SIZE - 1,
+		.flags = IORESOURCE_MEM,
+	},
+	[1] = { /* buffer space */
+		.start = OETH_SRAMBUFF_PADDR,
+		.end   = OETH_SRAMBUFF_PADDR + OETH_SRAMBUFF_SIZE - 1,
+		.flags = IORESOURCE_MEM,
+	},
+	[2] = { /* IRQ number */
+		.start = OETH_IRQ,
+		.end   = OETH_IRQ,
+		.flags = IORESOURCE_IRQ,
+	},
+};
+
+static struct ethoc_platform_data ethoc_pdata __initdata = {
+	/*
+	 * The MAC address for these boards is 00:50:c2:13:6f:xx.
+	 * The last byte (here as zero) is read from the DIP switches on the
+	 * board.
+	 */
+	.hwaddr = { 0x00, 0x50, 0xc2, 0x13, 0x6f, 0 },
+	.phy_id = -1,
+};
+
+static struct platform_device ethoc_device __initdata = {
+	.name = "ethoc",
+	.id = -1,
+	.num_resources = ARRAY_SIZE(ethoc_res),
+	.resource = ethoc_res,
+	.dev = {
+		.platform_data = &ethoc_pdata,
+	},
+};
+
+/*----------------------------------------------------------------------------
+ *  UART
+ */
+
+static struct resource serial_resource __initdata = {
+	.start	= DUART16552_PADDR,
+	.end	= DUART16552_PADDR + 0x1f,
+	.flags	= IORESOURCE_MEM,
+};
+
+static struct plat_serial8250_port serial_platform_data[] __initdata = {
+	[0] = {
+		.mapbase	= DUART16552_PADDR,
+		.irq		= DUART16552_INTNUM,
+		.flags		= UPF_BOOT_AUTOCONF | UPF_SKIP_TEST |
+				  UPF_IOREMAP,
+		.iotype		= UPIO_MEM32,
+		.regshift	= 2,
+		.uartclk	= 0,    /* set in xtavnet_init() */
+	},
+	{ },
+};
+
+static struct platform_device xtavnet_uart __initdata = {
+	.name		= "serial8250",
+	.id		= PLAT8250_DEV_PLATFORM,
+	.dev		= {
+		.platform_data	= serial_platform_data,
+	},
+	.num_resources	= 1,
+	.resource	= &serial_resource,
+};
+
+/* platform devices */
+static struct platform_device *platform_devices[] __initdata = {
+	&ethoc_device,
+	&xtavnet_uart,
+};
+
+
+static int __init xtavnet_init(void)
+{
+	/* Ethernet MAC address.  */
+	ethoc_pdata.hwaddr[5] = *(u32 *)DIP_SWITCHES_VADDR;
+
+	/* Clock rate varies among FPGA bitstreams; board specific FPGA register
+	 * reports the actual clock rate.
+	 */
+	serial_platform_data[0].uartclk = *(long *)XTFPGA_CLKFRQ_VADDR;
+
+
+	/* register platform devices */
+	platform_add_devices(platform_devices, ARRAY_SIZE(platform_devices));
+
+	/* ETHOC driver is a bit quiet; at least display Ethernet MAC, so user
+	 * knows whether they set it correctly on the DIP switches.
+	 */
+	pr_info("XTFPGA: Ethernet MAC %pM\n", ethoc_pdata.hwaddr);
+
+	return 0;
+}
+
+/*
+ * Register to be done during do_initcalls().
+ */
+arch_initcall(xtavnet_init);
+
+#endif /* CONFIG_OF */

arch/xtensa/variants/s6000/gpio.c

@@ -164,7 +164,7 @@ static void demux_irqs(unsigned int irq, struct irq_desc *desc)
 	int cirq;
 
 	chip->irq_mask(&desc->irq_data);
-	chip->irq_ack(&desc->irq_data));
+	chip->irq_ack(&desc->irq_data);
 	pending = readb(S6_REG_GPIO + S6_GPIO_BANK(0) + S6_GPIO_MIS) & *mask;
 	cirq = IRQ_BASE - 1;
 	while (pending) {
@@ -173,7 +173,7 @@ static void demux_irqs(unsigned int irq, struct irq_desc *desc)
 		pending >>= n;
 		generic_handle_irq(cirq);
 	}
-	chip->irq_unmask(&desc->irq_data));
+	chip->irq_unmask(&desc->irq_data);
 }
 
 extern const signed char *platform_irq_mappings[XTENSA_NR_IRQS];