[MIPS] R4000/R4400 daddiu erratum workaround

 This complements the generic R4000/R4400 errata workaround code and adds 
bits for the daddiu problem.  In most places it just modifies handwritten 
assembly code so that the assembler is allowed to use a temporary register 
as daddiu may now be treated as a macro that expands to a sequence of li 
and daddu.  It is the AT register or, where AT is unavailable or used 
explicitly for another purpose, an explicitly-named register is selected, 
using the .set at=<reg> feature added recently to gas.  This feature is 
only used if CONFIG_CPU_DADDI_WORKAROUNDS has been set, so if the 
workaround remains disabled, the required version of binutils stays 
unchanged.

 Similarly, daddiu instructions put in branch delay slots in noreorder 
fragments are now taken out of them and the assembler is allowed to 
reorder them itself as possible (which it does making the whole idea of 
scheduling them into delay slots manually questionable).

 Also in the very few places where such a simple conversion was not 
possible, a handcoded longer sequence is implemented.

 Other than that there are changes to code responsible for building the 
TLB fault and page clear/copy handlers to avoid daddiu as appropriate.  
These are only effective if the erratum is verified to be present at the 
run time.

 Finally there is a trivial update to __delay(), because it uses daddiu in 
a branch delay slot.

Signed-off-by: Maciej W. Rozycki <macro@linux-mips.org>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>

3 files changed, 29 insertions, 4 deletions

diff --git a/include/asm-mips/delay.h b/include/asm-mips/delay.h
index de5105d05f1e..b0bccd2c4ed5 100644
--- a/include/asm-mips/delay.h
+++ b/include/asm-mips/delay.h
@@ -28,7 +28,7 @@ static inline void __delay(unsigned long loops)
 		"	.set	reorder					\n"
 		: "=r" (loops)
 		: "0" (loops));
-	else if (sizeof(long) == 8)
+	else if (sizeof(long) == 8 && !DADDI_WAR)
 		__asm__ __volatile__ (
 		"	.set	noreorder				\n"
 		"	.align	3					\n"
@@ -37,6 +37,15 @@ static inline void __delay(unsigned long loops)
 		"	.set	reorder					\n"
 		: "=r" (loops)
 		: "0" (loops));
+	else if (sizeof(long) == 8 && DADDI_WAR)
+		__asm__ __volatile__ (
+		"	.set	noreorder				\n"
+		"	.align	3					\n"
+		"1:	bnez	%0, 1b					\n"
+		"	dsubu	%0, %2					\n"
+		"	.set	reorder					\n"
+		: "=r" (loops)
+		: "0" (loops), "r" (1));
 }
 
 
diff --git a/include/asm-mips/stackframe.h b/include/asm-mips/stackframe.h
index fb41a8d76392..051e1af0bb95 100644
--- a/include/asm-mips/stackframe.h
+++ b/include/asm-mips/stackframe.h
@@ -6,6 +6,7 @@
  * Copyright (C) 1994, 95, 96, 99, 2001 Ralf Baechle
  * Copyright (C) 1994, 1995, 1996 Paul M. Antoine.
  * Copyright (C) 1999 Silicon Graphics, Inc.
+ * Copyright (C) 2007  Maciej W. Rozycki
  */
 #ifndef _ASM_STACKFRAME_H
 #define _ASM_STACKFRAME_H
@@ -145,8 +146,16 @@
 		.set	reorder
 		/* Called from user mode, new stack. */
 		get_saved_sp
+#ifndef CONFIG_CPU_DADDI_WORKAROUNDS
 8:		move	k0, sp
 		PTR_SUBU sp, k1, PT_SIZE
+#else
+		.set	at=k0
+8:		PTR_SUBU k1, PT_SIZE
+		.set	noat
+		move	k0, sp
+		move	sp, k1
+#endif
 		LONG_S	k0, PT_R29(sp)
 		LONG_S	$3, PT_R3(sp)
 		/*
diff --git a/include/asm-mips/uaccess.h b/include/asm-mips/uaccess.h
index c30c718994c9..66523d610950 100644
--- a/include/asm-mips/uaccess.h
+++ b/include/asm-mips/uaccess.h
@@ -5,6 +5,7 @@
  *
  * Copyright (C) 1996, 1997, 1998, 1999, 2000, 03, 04 by Ralf Baechle
  * Copyright (C) 1999, 2000 Silicon Graphics, Inc.
+ * Copyright (C) 2007  Maciej W. Rozycki
  */
 #ifndef _ASM_UACCESS_H
 #define _ASM_UACCESS_H
@@ -387,6 +388,12 @@ extern void __put_user_unknown(void);
 	"jal\t" #destination "\n\t"
 #endif
 
+#ifndef CONFIG_CPU_DADDI_WORKAROUNDS
+#define DADDI_SCRATCH "$0"
+#else
+#define DADDI_SCRATCH "$3"
+#endif
+
 extern size_t __copy_user(void *__to, const void *__from, size_t __n);
 
 #define __invoke_copy_to_user(to, from, n)				\
@@ -403,7 +410,7 @@ extern size_t __copy_user(void *__to, const void *__from, size_t __n);
 	: "+r" (__cu_to_r), "+r" (__cu_from_r), "+r" (__cu_len_r)	\
 	:								\
 	: "$8", "$9", "$10", "$11", "$12", "$15", "$24", "$31",		\
-	  "memory");							\
+	  DADDI_SCRATCH, "memory");					\
 	__cu_len_r;							\
 })
 
@@ -512,7 +519,7 @@ extern size_t __copy_user_inatomic(void *__to, const void *__from, size_t __n);
 	: "+r" (__cu_to_r), "+r" (__cu_from_r), "+r" (__cu_len_r)	\
 	:								\
 	: "$8", "$9", "$10", "$11", "$12", "$15", "$24", "$31",		\
-	  "memory");							\
+	  DADDI_SCRATCH, "memory");					\
 	__cu_len_r;							\
 })
 
@@ -535,7 +542,7 @@ extern size_t __copy_user_inatomic(void *__to, const void *__from, size_t __n);
 	: "+r" (__cu_to_r), "+r" (__cu_from_r), "+r" (__cu_len_r)	\
 	:								\
 	: "$8", "$9", "$10", "$11", "$12", "$15", "$24", "$31",		\
-	  "memory");							\
+	  DADDI_SCRATCH, "memory");					\
 	__cu_len_r;							\
 })