[PATCH] PA-RISC update for 2.6.0-test5

PA-RISC updates for 2.6.0-test5.

Contributions from Paul Bame, James Bottomley, Randolph Chung, Helge
Deller, Grant Grundler, LaMont Jones, Matthew Wilcox

10 files changed, 177 insertions, 102 deletions

diff --git a/include/asm-parisc/atomic.h b/include/asm-parisc/atomic.h
index d50c506dd8a4..808373512d6c 100644
--- a/include/asm-parisc/atomic.h
+++ b/include/asm-parisc/atomic.h
@@ -49,27 +49,14 @@ extern spinlock_t __atomic_hash[ATOMIC_HASH_SIZE];
  * Cache-line alignment would conflict with, for example, linux/module.h
  */
 
-typedef struct {
-	volatile int counter;
-} atomic_t;
+typedef struct { volatile long counter; } atomic_t;
 
 
-/*
-** xchg/cmpxchg moved from asm/system.h - ggg
-*/
-
-#if 1
 /* This should get optimized out since it's never called.
 ** Or get a link error if xchg is used "wrong".
 */
 extern void __xchg_called_with_bad_pointer(void);
-#else
-static inline void __xchg_called_with_bad_pointer(void)
-{
-	extern void panic(const char * fmt, ...);
-	panic("xchg called with bad pointer");
-}
-#endif
+
 
 /* __xchg32/64 defined in arch/parisc/lib/bitops.c */
 extern unsigned long __xchg8(char, char *);
diff --git a/include/asm-parisc/bitops.h b/include/asm-parisc/bitops.h
index 711e7d154afa..51f4871ac6ce 100644
--- a/include/asm-parisc/bitops.h
+++ b/include/asm-parisc/bitops.h
@@ -203,55 +203,102 @@ static __inline__ int test_bit(int nr, const void *address)
 	return !!(*addr & mask);
 }
 
-extern __inline__ unsigned long ffz(unsigned long word)
-{
-	unsigned long result;
-
-	result = 0;
-	while (word & 1) {
-		result++;
-		word >>= 1;
-	}
-
-	return result;
-}
-
 #ifdef __KERNEL__
 
 /**
- * __ffs - find first bit in word.
+ * __ffs - find first bit in word. returns 0 to "BITS_PER_LONG-1".
  * @word: The word to search
  *
- * Undefined if no bit exists, so code should check against 0 first.
+ * __ffs() return is undefined if no bit is set.
+ *
+ * 32-bit fast __ffs by LaMont Jones "lamont At hp com".
+ * 64-bit enhancement by Grant Grundler "grundler At parisc-linux org".
+ * (with help from willy/jejb to get the semantics right)
+ *
+ * This algorithm avoids branches by making use of nullification.
+ * One side effect of "extr" instructions is it sets PSW[N] bit.
+ * How PSW[N] (nullify next insn) gets set is determined by the 
+ * "condition" field (eg "<>" or "TR" below) in the extr* insn.
+ * Only the 1st and one of either the 2cd or 3rd insn will get executed.
+ * Each set of 3 insn will get executed in 2 cycles on PA8x00 vs 16 or so
+ * cycles for each mispredicted branch.
  */
-static __inline__ unsigned long __ffs(unsigned long word)
-{
-	unsigned long result = 0;
 
-	while (!(word & 1UL)) {
-		result++;
-		word >>= 1;
-	}
-	return result;
+static __inline__ unsigned long __ffs(unsigned long x)
+{
+	unsigned long ret;
+
+	__asm__(
+#if BITS_PER_LONG > 32
+		" ldi       63,%1\n"
+		" extrd,u,*<>  %0,63,32,%%r0\n"
+		" extrd,u,*TR  %0,31,32,%0\n"
+		" addi    -32,%1,%1\n"
+#else
+		" ldi       31,%1\n"
+#endif
+		" extru,<>  %0,31,16,%%r0\n"
+		" extru,TR  %0,15,16,%0\n"
+		" addi    -16,%1,%1\n"
+		" extru,<>  %0,31,8,%%r0\n"
+		" extru,TR  %0,23,8,%0\n"
+		" addi    -8,%1,%1\n"
+		" extru,<>  %0,31,4,%%r0\n"
+		" extru,TR  %0,27,4,%0\n"
+		" addi    -4,%1,%1\n"
+		" extru,<>  %0,31,2,%%r0\n"
+		" extru,TR  %0,29,2,%0\n"
+		" addi    -2,%1,%1\n"
+		" extru,=  %0,31,1,%%r0\n"
+		" addi    -1,%1,%1\n"
+			: "+r" (x), "=r" (ret) );
+	return ret;
 }
 
+/* Undefined if no bit is zero. */
+#define ffz(x)	__ffs(~x)
+
 /*
- * ffs: find first bit set. This is defined the same way as
- * the libc and compiler builtin ffs routines, therefore
- * differs in spirit from the above ffz (man ffs).
+ * ffs: find first bit set. returns 1 to BITS_PER_LONG or 0 (if none set)
+ * This is defined the same way as the libc and compiler builtin
+ * ffs routines, therefore differs in spirit from the above ffz (man ffs).
  */
 static __inline__ int ffs(int x)
 {
-	if (!x)
-		return 0;
-	return __ffs((unsigned long)x);
+	return x ? (__ffs((unsigned long)x) + 1) : 0;
 }
 
 /*
- * fls: find last bit set.
+ * fls: find last (most significant) bit set.
+ * fls(0) = 0, fls(1) = 1, fls(0x80000000) = 32.
  */
 
-#define fls(x) generic_fls(x)
+static __inline__ int fls(int x)
+{
+	int ret;
+	if (!x)
+		return 0;
+
+	__asm__(
+	"	ldi		1,%1\n"
+	"	extru,<>	%0,15,16,%%r0\n"
+	"	zdep,TR		%0,15,16,%0\n"		/* xxxx0000 */
+	"	addi		16,%1,%1\n"
+	"	extru,<>	%0,7,8,%%r0\n"
+	"	zdep,TR		%0,23,24,%0\n"		/* xx000000 */
+	"	addi		8,%1,%1\n"
+	"	extru,<>	%0,3,4,%%r0\n"
+	"	zdep,TR		%0,27,28,%0\n"		/* x0000000 */
+	"	addi		4,%1,%1\n"
+	"	extru,<>	%0,1,2,%%r0\n"
+	"	zdep,TR		%0,29,30,%0\n"		/* y0000000 (y&3 = 0 */
+	"	addi		2,%1,%1\n"
+	"	extru,=		%0,0,1,%%r0\n"
+	"	addi		1,%1,%1\n"		/* if y & 8, add 1 */
+		: "+r" (x), "=r" (ret) );
+
+	return ret;
+}
 
 /*
  * hweightN: returns the hamming weight (i.e. the number
diff --git a/include/asm-parisc/byteorder.h b/include/asm-parisc/byteorder.h
index 9ddeab1777a5..352b3146f7b0 100644
--- a/include/asm-parisc/byteorder.h
+++ b/include/asm-parisc/byteorder.h
@@ -5,6 +5,25 @@
 
 #ifdef __GNUC__
 
+static __inline__ __const__ __u16 ___arch__swab16(__u16 x)
+{
+	__asm__("dep %0, 15, 8, %0\n\t"		/* deposit 00ab -> 0bab */
+		"shd %%r0, %0, 8, %0"		/* shift 000000ab -> 00ba */
+		: "=r" (x)
+		: "0" (x));
+	return x;
+}
+
+static __inline__ __const__ __u32 ___arch__swab24(__u32 x)
+{
+	__asm__("shd %0, %0, 8, %0\n\t"		/* shift xabcxabc -> cxab */
+		"dep %0, 15, 8, %0\n\t"		/* deposit cxab -> cbab */
+		"shd %%r0, %0, 8, %0"		/* shift 0000cbab -> 0cba */
+		: "=r" (x)
+		: "0" (x));
+	return x;
+}
+
 static __inline__ __const__ __u32 ___arch__swab32(__u32 x)
 {
 	unsigned int temp;
@@ -39,34 +58,21 @@ static __inline__ __const__ __u64 ___arch__swab64(__u64 x) {
 	return x;
 }
 #define __arch__swab64(x) ___arch__swab64(x)
-#else
+#define __BYTEORDER_HAS_U64__
+#elif !defined(__STRICT_ANSI__)
 static __inline__ __const__ __u64 ___arch__swab64(__u64 x)
 {
-	__u32 t1 = (__u32) x;
-	__u32 t2 = (__u32) ((x) >> 32);
-	___arch__swab32(t1);
-	___arch__swab32(t2);
-	return (((__u64) t1 << 32) + ((__u64) t2));
+	__u32 t1 = ___arch__swab32((__u32) x);
+	__u32 t2 = ___arch__swab32((__u32) (x >> 32));
+	return (((__u64) t1 << 32) | t2);
 }
+#define __arch__swab64(x) ___arch__swab64(x)
+#define __BYTEORDER_HAS_U64__
 #endif
 
-
-static __inline__ __const__ __u16 ___arch__swab16(__u16 x)
-{
-	__asm__("dep %0, 15, 8, %0\n\t"		/* deposit 00ab -> 0bab */
-		"shd %r0, %0, 8, %0"		/* shift 000000ab -> 00ba */
-		: "=r" (x)
-		: "0" (x));
-	return x;
-}
-
-#define __arch__swab32(x) ___arch__swab32(x)
 #define __arch__swab16(x) ___arch__swab16(x)
-
-#if !defined(__STRICT_ANSI__) || defined(__KERNEL__)
-#  define __BYTEORDER_HAS_U64__
-#  define __SWAB_64_THRU_32__
-#endif
+#define __arch__swab24(x) ___arch__swab24(x)
+#define __arch__swab32(x) ___arch__swab32(x)
 
 #endif /* __GNUC__ */
 
diff --git a/include/asm-parisc/cacheflush.h b/include/asm-parisc/cacheflush.h
index 5392339e549b..b88200337cca 100644
--- a/include/asm-parisc/cacheflush.h
+++ b/include/asm-parisc/cacheflush.h
@@ -78,8 +78,9 @@ static inline void flush_dcache_page(struct page *page)
 
 #define flush_icache_range(s,e)		do { flush_kernel_dcache_range_asm(s,e); flush_kernel_icache_range_asm(s,e); } while (0)
 
-#define flush_icache_user_range(vma, page, addr, len) \
-	flush_icache_page((vma), (page))
+#define flush_icache_user_range(vma, page, addr, len) do { \
+        flush_user_dcache_range(addr, addr + len); \
+	flush_user_icache_range(addr, addr + len); } while (0)
 
 static inline void flush_cache_range(struct vm_area_struct *vma,
 		unsigned long start, unsigned long end)
diff --git a/include/asm-parisc/elf.h b/include/asm-parisc/elf.h
index b9f6090b78f2..a01e946dfc57 100644
--- a/include/asm-parisc/elf.h
+++ b/include/asm-parisc/elf.h
@@ -283,6 +283,7 @@ struct pt_regs;	/* forward declaration... */
  */
 #define ELF_DATA	ELFDATA2MSB
 #define ELF_ARCH	EM_PARISC
+#define ELF_OSABI 	ELFOSABI_LINUX
 
 /* %r23 is set by ld.so to a pointer to a function which might be 
    registered using atexit.  This provides a mean for the dynamic
diff --git a/include/asm-parisc/hil.h b/include/asm-parisc/hil.h
deleted file mode 100644
index 97cd5553f22f..000000000000
--- a/include/asm-parisc/hil.h
+++ /dev/null
@@ -1,26 +0,0 @@
-#ifndef _ASM_HIL_H
-#define _ASM_HIL_H
-
-/*
- *	linux/asm-parisc/hil.h
- *
- *	(c) 1999 Matthew Wilcox
- */
-
-extern unsigned long hil_base;	/* declared in drivers/parisc/hil.c */
-extern unsigned int hil_irq;
-
-#define HILBASE			hil_base /* 0xf0821000 (old) or 0xf0201000 (new) */
-#define HIL_DATA		0x800
-#define HIL_CMD			0x801
-
-#define HIL_IRQ			hil_irq
-
-#define hil_busy()		(gsc_readb(HILBASE + HIL_CMD) & HIL_BUSY)
-#define hil_data_available()	(gsc_readb(HILBASE + HIL_CMD) & HIL_DATA_RDY)
-#define hil_status()		(gsc_readb(HILBASE + HIL_CMD))
-#define hil_command(x)		do { gsc_writeb((x), HILBASE + HIL_CMD); } while (0)
-#define hil_read_data()		(gsc_readb(HILBASE + HIL_DATA))
-#define hil_write_data(x)	do { gsc_writeb((x), HILBASE + HIL_DATA); } while (0)
-
-#endif
diff --git a/include/asm-parisc/io.h b/include/asm-parisc/io.h
index 69933ab23159..0350f0198995 100644
--- a/include/asm-parisc/io.h
+++ b/include/asm-parisc/io.h
@@ -291,4 +291,11 @@ extern void outsl (unsigned long port, const void *src, unsigned long count);
 #define dma_cache_wback(_start,_size)		do { flush_kernel_dcache_range(_start,_size); } while (0)
 #define dma_cache_wback_inv(_start,_size)	do { flush_kernel_dcache_range(_start,_size); } while (0)
 
+/* PA machines have an MM I/O space from 0xf0000000-0xffffffff in 32
+ * bit mode and from 0xfffffffff0000000-0xfffffffffffffff in 64 bit
+ * mode (essentially just sign extending.  This macro takes in a 32
+ * bit I/O address (still with the leading f) and outputs the correct
+ * value for either 32 or 64 bit mode */
+#define F_EXTEND(x) ((unsigned long)((x) | (0xffffffff00000000ULL)))
+
 #endif
diff --git a/include/asm-parisc/local.h b/include/asm-parisc/local.h
new file mode 100644
index 000000000000..892b3b2c4962
--- /dev/null
+++ b/include/asm-parisc/local.h
@@ -0,0 +1,40 @@
+#ifndef _ARCH_PARISC_LOCAL_H
+#define _ARCH_PARISC_LOCAL_H
+
+#include <linux/percpu.h>
+#include <asm/atomic.h>
+
+typedef atomic_t local_t;
+
+#define LOCAL_INIT(i)	ATOMIC_INIT(i)
+#define local_read(v)	atomic_read(v)
+#define local_set(v,i)	atomic_set(v,i)
+
+#define local_inc(v)	atomic_inc(v)
+#define local_dec(v)	atomic_dec(v)
+#define local_add(i, v)	atomic_add(i, v)
+#define local_sub(i, v)	atomic_sub(i, v)
+
+#define __local_inc(v)		((v)->counter++)
+#define __local_dec(v)		((v)->counter--)
+#define __local_add(i,v)	((v)->counter+=(i))
+#define __local_sub(i,v)	((v)->counter-=(i))
+
+/* Use these for per-cpu local_t variables: on some archs they are
+ * much more efficient than these naive implementations.  Note they take
+ * a variable, not an address.
+ */
+#define cpu_local_read(v)	local_read(&__get_cpu_var(v))
+#define cpu_local_set(v, i)	local_set(&__get_cpu_var(v), (i))
+
+#define cpu_local_inc(v)	local_inc(&__get_cpu_var(v))
+#define cpu_local_dec(v)	local_dec(&__get_cpu_var(v))
+#define cpu_local_add(i, v)	local_add((i), &__get_cpu_var(v))
+#define cpu_local_sub(i, v)	local_sub((i), &__get_cpu_var(v))
+
+#define __cpu_local_inc(v)	__local_inc(&__get_cpu_var(v))
+#define __cpu_local_dec(v)	__local_dec(&__get_cpu_var(v))
+#define __cpu_local_add(i, v)	__local_add((i), &__get_cpu_var(v))
+#define __cpu_local_sub(i, v)	__local_sub((i), &__get_cpu_var(v))
+
+#endif /* _ARCH_PARISC_LOCAL_H */
diff --git a/include/asm-parisc/sections.h b/include/asm-parisc/sections.h
new file mode 100644
index 000000000000..fdd43ec42ec5
--- /dev/null
+++ b/include/asm-parisc/sections.h
@@ -0,0 +1,7 @@
+#ifndef _PARISC_SECTIONS_H
+#define _PARISC_SECTIONS_H
+
+/* nothing to see, move along */
+#include <asm-generic/sections.h>
+
+#endif
diff --git a/include/asm-parisc/uaccess.h b/include/asm-parisc/uaccess.h
index 8bd1067c4a50..cc9e311cabd1 100644
--- a/include/asm-parisc/uaccess.h
+++ b/include/asm-parisc/uaccess.h
@@ -28,6 +28,11 @@
  * that put_user is the same as __put_user, etc.
  */
 
+extern int __get_kernel_bad(void);
+extern int __get_user_bad(void);
+extern int __put_kernel_bad(void);
+extern int __put_user_bad(void);
+
 #define access_ok(type,addr,size)   (1)
 #define verify_area(type,addr,size) (0)
 
@@ -35,8 +40,8 @@
 #define get_user __get_user
 
 #if BITS_PER_LONG == 32
-#define LDD_KERNEL(ptr)		BUG()
-#define LDD_USER(ptr)		BUG()
+#define LDD_KERNEL(ptr) __get_kernel_bad();
+#define LDD_USER(ptr) __get_user_bad();
 #define STD_KERNEL(x, ptr) __put_kernel_asm64((u32)x,ptr)
 #define STD_USER(x, ptr) __put_user_asm64((u32)x,ptr)
 #else
@@ -72,7 +77,7 @@ struct exception_table_entry {
 	    case 2: __get_kernel_asm("ldh",ptr); break; \
 	    case 4: __get_kernel_asm("ldw",ptr); break; \
 	    case 8: LDD_KERNEL(ptr); break;		\
-	    default: BUG(); break;                      \
+	    default: __get_kernel_bad(); break;         \
 	    }                                           \
 	}                                               \
 	else {                                          \
@@ -81,7 +86,7 @@ struct exception_table_entry {
 	    case 2: __get_user_asm("ldh",ptr); break;   \
 	    case 4: __get_user_asm("ldw",ptr); break;   \
 	    case 8: LDD_USER(ptr);  break;		\
-	    default: BUG(); break;                      \
+	    default: __get_user_bad(); break;           \
 	    }                                           \
 	}                                               \
 							\
@@ -141,7 +146,7 @@ struct exception_table_entry {
 	    case 2: __put_kernel_asm("sth",x,ptr); break;       \
 	    case 4: __put_kernel_asm("stw",x,ptr); break;       \
 	    case 8: STD_KERNEL(x,ptr); break;			\
-	    default: BUG(); break;                              \
+	    default: __put_kernel_bad(); break;			\
 	    }                                                   \
 	}                                                       \
 	else {                                                  \
@@ -150,7 +155,7 @@ struct exception_table_entry {
 	    case 2: __put_user_asm("sth",x,ptr); break;         \
 	    case 4: __put_user_asm("stw",x,ptr); break;         \
 	    case 8: STD_USER(x,ptr); break;			\
-	    default: BUG(); break;                              \
+	    default: __put_user_bad(); break;			\
 	    }                                                   \
 	}                                                       \
 								\