[PARISC] Update spinlocks from parisc tree

Neaten up the CONFIG_PA20 ifdefs

More merge fixes, this time for SMP

Signed-off-by: Matthew Wilcox <willy@parisc-linux.org>

Prettify the CONFIG_DEBUG_SPINLOCK __SPIN_LOCK_UNLOCKED initializers.

Clean up some warnings with CONFIG_DEBUG_SPINLOCK enabled.

Fix build with spinlock debugging turned on. Patch is cleaner like this,
too.

Remove mandatory 16-byte alignment requirement on PA2.0 processors by
using the ldcw,CO completer. Provides a nice insn savings.

Signed-off-by: Kyle McMartin <kyle@parisc-linux.org>

3 files changed, 30 insertions, 14 deletions

diff --git a/include/asm-parisc/spinlock.h b/include/asm-parisc/spinlock.h
index 43eaa6e742e0..7c3f406a746a 100644
--- a/include/asm-parisc/spinlock.h
+++ b/include/asm-parisc/spinlock.h
@@ -5,11 +5,6 @@
 #include <asm/processor.h>
 #include <asm/spinlock_types.h>
 
-/* Note that PA-RISC has to use `1' to mean unlocked and `0' to mean locked
- * since it only has load-and-zero. Moreover, at least on some PA processors,
- * the semaphore address has to be 16-byte aligned.
- */
-
 static inline int __raw_spin_is_locked(raw_spinlock_t *x)
 {
         volatile unsigned int *a = __ldcw_align(x);
diff --git a/include/asm-parisc/spinlock_types.h b/include/asm-parisc/spinlock_types.h
index 785bba822fbf..d6b479bdb886 100644
--- a/include/asm-parisc/spinlock_types.h
+++ b/include/asm-parisc/spinlock_types.h
@@ -6,11 +6,15 @@
 #endif
 
 typedef struct {
+#ifdef CONFIG_PA20
+        volatile unsigned int slock;
+# define __RAW_SPIN_LOCK_UNLOCKED { 1 }
+#else
         volatile unsigned int lock[4];
+# define __RAW_SPIN_LOCK_UNLOCKED       { { 1, 1, 1, 1 } }
+#endif
 } raw_spinlock_t;
 
-#define __RAW_SPIN_LOCK_UNLOCKED        { { 1, 1, 1, 1 } }
-
 typedef struct {
         raw_spinlock_t lock;
         volatile int counter;
diff --git a/include/asm-parisc/system.h b/include/asm-parisc/system.h
index a25e9dcd2e79..f3928d3a80cb 100644
--- a/include/asm-parisc/system.h
+++ b/include/asm-parisc/system.h
@@ -138,13 +138,7 @@ static inline void set_eiem(unsigned long val)
 #define set_wmb(var, value)             do { var = value; wmb(); } while (0)
 
 
-/* LDCW, the only atomic read-write operation PA-RISC has. *sigh*.  */
-#define __ldcw(a) ({ \
-        unsigned __ret; \
-        __asm__ __volatile__("ldcw 0(%1),%0" : "=r" (__ret) : "r" (a)); \
-        __ret; \
-})
-
+#ifndef CONFIG_PA20
 /* Because kmalloc only guarantees 8-byte alignment for kmalloc'd data,
    and GCC only guarantees 8-byte alignment for stack locals, we can't
    be assured of 16-byte alignment for atomic lock data even if we
@@ -152,12 +146,35 @@ static inline void set_eiem(unsigned long val)
    we use a struct containing an array of four ints for the atomic lock
    type and dynamically select the 16-byte aligned int from the array
    for the semaphore.  */
+
 #define __PA_LDCW_ALIGNMENT 16
 #define __ldcw_align(a) ({ \
   unsigned long __ret = (unsigned long) &(a)->lock[0];                  \
   __ret = (__ret + __PA_LDCW_ALIGNMENT - 1) & ~(__PA_LDCW_ALIGNMENT - 1); \
   (volatile unsigned int *) __ret;                                      \
 })
+#define LDCW    "ldcw"
+
+#else /*CONFIG_PA20*/
+/* From: "Jim Hull" <jim.hull of hp.com>
+   I've attached a summary of the change, but basically, for PA 2.0, as
+   long as the ",CO" (coherent operation) completer is specified, then the
+   16-byte alignment requirement for ldcw and ldcd is relaxed, and instead
+   they only require "natural" alignment (4-byte for ldcw, 8-byte for
+   ldcd). */
+
+#define __PA_LDCW_ALIGNMENT 4
+#define __ldcw_align(a) ((volatile unsigned int *)a)
+#define LDCW    "ldcw,co"
+
+#endif /*!CONFIG_PA20*/
+
+/* LDCW, the only atomic read-write operation PA-RISC has. *sigh*.  */
+#define __ldcw(a) ({ \
+        unsigned __ret; \
+        __asm__ __volatile__(LDCW " 0(%1),%0" : "=r" (__ret) : "r" (a)); \
+        __ret; \
+})
 
 #ifdef CONFIG_SMP
 # define __lock_aligned __attribute__((__section__(".data.lock_aligned")))