Merge from trunk revision 277462.

From-SVN: r277464

5 files changed, 137 insertions, 87 deletions

diff --git a/libgcc/config/msp430/lib2hw_mul.S b/libgcc/config/msp430/lib2hw_mul.S
index 1a0e6e8..894c551 100644
--- a/libgcc/config/msp430/lib2hw_mul.S
+++ b/libgcc/config/msp430/lib2hw_mul.S
@@ -81,9 +81,9 @@
 	.type \gcc_name , @function
 \gcc_name:
 #ifdef __MSP430X_LARGE__
-	BRA	\eabi_soft_name
+	BRA	#\eabi_soft_name
 #else
-	BR	\eabi_soft_name
+	BR	#\eabi_soft_name
 #endif
 	.size \gcc_name , . - \gcc_name
 	.popsection
@@ -109,7 +109,7 @@
 	MOV.W	&\RESULT, r12		; Move result into return register
 .endm
 
-.macro mult1632 OP1, OP2, RESULT_LO, RESULT_HI
+.macro mult1632 OP1, OP2, RESLO, RESHI
 ;* * 16-bit hardware multiply with a 32-bit result:
 ;*	int32 = int16 * int16
 ;* 	uint32 = uint16 * uint16
@@ -127,11 +127,11 @@
 	
 	MOV.W	r12, &\OP1		; Load operand 1 into multiplier
 	MOV.W	r13, &\OP2		; Load operand 2 which triggers MPY
-	MOV.W	&\RESULT_LO, r12	; Move low result into return register
-	MOV.W	&\RESULT_HI, r13	; Move high result into return register
+	MOV.W	&\RESLO, r12		; Move low result into return register
+	MOV.W	&\RESHI, r13		; Move high result into return register
 .endm
 
-.macro mult32 OP1, OP2, MAC_OP1, MAC_OP2, RESULT_LO, RESULT_HI
+.macro mult32 OP1, OP2, MAC_OP1, MAC_OP2, RESLO, RESHI
 ;* * 32-bit hardware multiply with a 32-bit result using 16 multiply and accumulate:
 ;*	int32 = int32 * int32
 ;*  
@@ -149,16 +149,16 @@
 	MOV.W	r12, &\OP1		; Load operand 1 Low into multiplier
 	MOV.W	r14, &\OP2		; Load operand 2 Low which triggers MPY
 	MOV.W	r12, &\MAC_OP1		; Load operand 1 Low into mac
-	MOV.W   &\RESULT_LO, r12	; Low 16-bits of result ready for return
-	MOV.W   &\RESULT_HI, &\RESULT_LO; MOV intermediate mpy high into low
+	MOV.W   &\RESLO, r12		; Low 16-bits of result ready for return
+	MOV.W   &\RESHI, &\RESLO	; MOV intermediate mpy high into low
 	MOV.W	r15, &\MAC_OP2		; Load operand 2 High, trigger MAC
 	MOV.W	r13, &\MAC_OP1		; Load operand 1 High
 	MOV.W	r14, &\MAC_OP2		; Load operand 2 Lo, trigger MAC
-	MOV.W	&\RESULT_LO, r13        ; Upper 16-bits result ready for return
+	MOV.W	&\RESLO, r13		; Upper 16-bits result ready for return
 .endm
 
 
-.macro mult32_hw  OP1_LO  OP1_HI  OP2_LO  OP2_HI  RESULT_LO  RESULT_HI
+.macro mult32_hw  OP1_LO  OP1_HI  OP2_LO  OP2_HI  RESLO  RESHI
 ;* * 32-bit hardware multiply with a 32-bit result
 ;*	int32 = int32 * int32
 ;*  
@@ -177,8 +177,8 @@
 	MOV.W	r13, &\OP1_HI		; Load operand 1 High into multiplier
 	MOV.W	r14, &\OP2_LO		; Load operand 2 Low into multiplier
 	MOV.W	r15, &\OP2_HI		; Load operand 2 High, trigger MPY
-	MOV.W	&\RESULT_LO, r12	; Ready low 16-bits for return
-	MOV.W   &\RESULT_HI, r13	; Ready high 16-bits for return
+	MOV.W	&\RESLO, r12		; Ready low 16-bits for return
+	MOV.W   &\RESHI, r13		; Ready high 16-bits for return
 .endm
 
 .macro mult3264_hw  OP1_LO  OP1_HI  OP2_LO  OP2_HI  RES0 RES1 RES2 RES3
@@ -264,105 +264,141 @@
 ;; 	      Multiply unsigned long by unsigned long; result is unsigned long long. Uses hardware MPY16
 ;; uint64 __mspabi_mpyull_hw32(uint32 x, uint32 y)
 ;; 	      Multiply unsigned long by unsigned long; result is unsigned long long. Uses hardware MPY32 (F4xx devices).
-;; uint64 _ _mspabi_mpyull_f5hw(uint32 x, uint32 y)
+;; uint64 __mspabi_mpyull_f5hw(uint32 x, uint32 y)
 ;;            Multiply unsigned long by unsigned long; result is unsigned long long. Uses hardware MPY32 (F5xx devices and up)
 
-
-
-.set MPY_OP1,   0x0130
-.set MPY_OP1_S, 0x0132
-.set MAC_OP1, 	0x0134
-.set MPY_OP2, 	0x0138
-.set MAC_OP2, 	0x0138
-.set RESULT_LO, 0x013A
-.set RESULT_HI, 0x013C
+;;;; The register names below are the standardised versions used across TI
+;;;; literature.
+
+;; Hardware multiply register addresses for devices with 16-bit hardware
+;; multiply.
+.set MPY,	0x0130
+.set MPYS,	0x0132
+.set MAC, 	0x0134
+.set OP2, 	0x0138
+.set RESLO,	0x013A
+.set RESHI,	0x013C
+;; Hardware multiply register addresses for devices with 32-bit (non-f5)
+;; hardware multiply.
+.set MPY32L,	0x0140
+.set MPY32H,	0x0142
+.set MPYS32L,	0x0144
+.set MPYS32H,	0x0146
+.set OP2L,	0x0150
+.set OP2H,	0x0152
+.set RES0,	0x0154
+.set RES1,	0x0156
+.set RES2,	0x0158
+.set RES3,	0x015A
+;; Hardware multiply register addresses for devices with f5series hardware
+;; multiply.
+;; The F5xxx series of MCUs support the same 16-bit and 32-bit multiply
+;; as the second generation hardware, but they are accessed from different
+;; memory registers.
+;; These names AREN'T standard.  We've appended _F5 to the standard names.
+.set MPY_F5,		0x04C0
+.set MPYS_F5,		0x04C2
+.set MAC_F5,		0x04C4
+.set OP2_F5,		0x04C8
+.set RESLO_F5,		0x04CA
+.set RESHI_F5,		0x04CC
+.set MPY32L_F5,		0x04D0
+.set MPY32H_F5,		0x04D2
+.set MPYS32L_F5,	0x04D4
+.set MPYS32H_F5,	0x04D6
+.set OP2L_F5,		0x04E0
+.set OP2H_F5,		0x04E2
+.set RES0_F5,		0x04E4
+.set RES1_F5,		0x04E6
+.set RES2_F5,		0x04E8
+.set RES3_F5,		0x04EA
 
 #if defined MUL_16
 ;;  First generation MSP430 hardware multiplies ...
 
 	start_func __mulhi2 __mspabi_mpyi  __mspabi_mpyi_hw
-	mult16 MPY_OP1, MPY_OP2, RESULT_LO
+	mult16 MPY, OP2, RESLO
 	end_func   __mulhi2
 
-	start_func __mulsihi2  __mspabi_mpysl  __mspabi_mpysl_hw
-	mult1632 MPY_OP1_S, MPY_OP2, RESULT_LO, RESULT_HI
-	end_func   __mulsihi2
+	start_func __mulhisi2  __mspabi_mpysl  __mspabi_mpysl_hw
+	mult1632 MPYS, OP2, RESLO, RESHI
+	end_func   __mulhisi2
 
-	start_func __umulsihi2  __mspabi_mpyul  _mspabi_mpyul_hw
-	mult1632 MPY_OP1, MPY_OP2, RESULT_LO, RESULT_HI
-	end_func   __umulsihi2
+	start_func __umulhisi2  __mspabi_mpyul  __mspabi_mpyul_hw
+	mult1632 MPY, OP2, RESLO, RESHI
+	end_func   __umulhisi2
 
 	start_func __mulsi2  __mspabi_mpyl  __mspabi_mpyl_hw
-	mult32 MPY_OP1, MPY_OP2, MAC_OP1, MAC_OP2, RESULT_LO, RESULT_HI
+	mult32 MPY, OP2, MAC, OP2, RESLO, RESHI
 	end_func   __mulsi2
 
 	;; FIXME: We do not have hardware implementations of these
 	;; routines, so just jump to the software versions instead.
-	fake_func __muldisi2   __mspabi_mpysll  __mspabi_mpysll_hw   
-	fake_func __umuldisi2  __mspabi_mpyull  __mspabi_mpyull_hw  
-	fake_func __muldi3     __mspabi_mpyll   __mspabi_mpyll_hw	
+	fake_func __mulsidi2   __mspabi_mpysll  __mspabi_mpysll_hw
+	fake_func __umulsidi2  __mspabi_mpyull  __mspabi_mpyull_hw
+	fake_func __muldi3     __mspabi_mpyll   __mspabi_mpyll_hw
 
 #elif defined MUL_32
 ;;  Second generation MSP430 hardware multiplies ...
 
 	start_func __mulhi2  __mspabi_mpyi  __mspabi_mpyi_hw
-	mult16 MPY_OP1, MPY_OP2, RESULT_LO
+	mult16 MPY, OP2, RESLO
 	end_func   __mulhi2
 
-	start_func __mulsihi2  __mspabi_mpysl  __mspabi_mpysl_hw
-	mult1632 MPY_OP1_S, MPY_OP2, RESULT_LO, RESULT_HI
-	end_func   __mulsihi2
+	start_func __mulhisi2  __mspabi_mpysl  __mspabi_mpysl_hw
+	mult1632 MPYS, OP2, RESLO, RESHI
+	end_func   __mulhisi2
 
-	start_func __umulsihi2  __mspabi_mpyul  _mspabi_mpyul_hw
-	mult1632 MPY_OP1, MPY_OP2, RESULT_LO, RESULT_HI
-	end_func   __umulsihi2
+	start_func __umulhisi2  __mspabi_mpyul  __mspabi_mpyul_hw
+	mult1632 MPY, OP2, RESLO, RESHI
+	end_func   __umulhisi2
 
 	start_func __mulsi2_hw32  __mspabi_mpyl  __mspabi_mpyl_hw32
-	mult32_hw 0x0140, 0x0142, 0x0150, 0x0152, 0x0154, 0x0156
+	mult32_hw MPY32L, MPY32H, OP2L, OP2H, RES0, RES1
 	end_func   __mulsi2_hw32
 
-	start_func __muldisi2  __mspabi_mpysll  __mspabi_mpysll_hw32
-	mult3264_hw 0x0144, 0x146, 0x0150, 0x0152, 0x0154, 0x0156, 0x0158, 0x015A
-	end_func   __muldisi2
+	start_func __mulsidi2  __mspabi_mpysll  __mspabi_mpysll_hw32
+	mult3264_hw MPYS32L, MPYS32H, OP2L, OP2H, RES0, RES1, RES2, RES3
+	end_func   __mulsidi2
 
-	start_func __umuldisi2 __mspabi_mpyull  __mspabi_mpyull_hw32
-	mult3264_hw 0x0140, 0x142, 0x0150, 0x0152, 0x0154, 0x0156, 0x0158, 0x015A
-	end_func   __umuldisi2
+	start_func __umulsidi2 __mspabi_mpyull  __mspabi_mpyull_hw32
+	mult3264_hw MPY32L, MPY32H, OP2L, OP2H, RES0, RES1, RES2, RES3
+	end_func   __umulsidi2
 
 	;; FIXME: Add a hardware version of this function.
-	fake_func __muldi3    __mspabi_mpyll  __mspabi_mpyll_hw32 
-	
+	fake_func __muldi3    __mspabi_mpyll  __mspabi_mpyll_hw32
+
 #elif defined MUL_F5
 /* The F5xxx series of MCUs support the same 16-bit and 32-bit multiply
    as the second generation hardware, but they are accessed from different
    memory registers.  */
 
 	start_func __mulhi2_f5 __mspabi_mpyi  __mspabi_mpyi_f5hw
-	mult16 0x04C0, 0x04C8, 0x04CA
+	mult16 MPY_F5, OP2_F5, RESLO_F5
 	end_func   __mulhi2_f5
 
-	start_func __mulsihi2  __mspabi_mpysl  __mspabi_mpysl_f5hw
-	mult1632 0x04C2, 0x04C8, 0x04CA, 0x04CC
-	end_func   __mulsihi2
-	
-	start_func __umulsihi2  __mspabi_mpyul  _mspabi_mpyul_f5hw
-	mult1632 0x04C0, 0x04C8, 0x04CA, 0x04CC
-	end_func   __umulsihi2
+	start_func __mulhisi2  __mspabi_mpysl  __mspabi_mpysl_f5hw
+	mult1632 MPYS_F5, OP2_F5, RESLO_F5, RESHI_F5
+	end_func   __mulhisi2
+
+	start_func __umulhisi2  __mspabi_mpyul  __mspabi_mpyul_f5hw
+	mult1632 MPY_F5, OP2_F5, RESLO_F5, RESHI_F5
+	end_func   __umulhisi2
 
 	start_func __mulsi2_f5  __mspabi_mpyl  __mspabi_mpyl_f5hw
-	mult32_hw 0x04D0, 0x04D2, 0x04E0, 0x04E2, 0x04E4, 0x04E6
+	mult32_hw MPY32L_F5, MPY32H_F5, OP2L_F5, OP2H_F5, RES0_F5, RES1_F5
 	end_func   __mulsi2_f5
-	
-	start_func __muldisi2  __mspabi_mpysll  __mspabi_mpysll_f5hw
-	mult3264_hw 0x04D4, 0x04D6, 0x04E0, 0x04E2, 0x04E4, 0x04E6, 0x04E8, 0x04EA
-	end_func   __muldisi2
-	
-	start_func __umuldisi2  __mspabi_mpyull  __mspabi_mpyull_f5hw
-	mult3264_hw 0x04D0, 0x04D2, 0x04E0, 0x04E2, 0x04E4, 0x04E6, 0x04E8, 0x04EA
-	end_func   __umuldisi2
+
+	start_func __mulsidi2  __mspabi_mpysll  __mspabi_mpysll_f5hw
+	mult3264_hw MPYS32L_F5, MPYS32H_F5, OP2L_F5, OP2H_F5, RES0_F5, RES1_F5, RES2_F5, RES3_F5
+	end_func   __mulsidi2
+
+	start_func __umulsidi2  __mspabi_mpyull  __mspabi_mpyull_f5hw
+	mult3264_hw MPY32L_F5, MPY32H_F5, OP2L_F5, OP2H_F5, RES0_F5, RES1_F5, RES2_F5, RES3_F5
+	end_func   __umulsidi2
 
 	;; FIXME: Add a hardware version of this function.
-	fake_func __muldi3   __mspabi_mpyll __mspabi_mpyll_f5hw 
+	fake_func __muldi3   __mspabi_mpyll __mspabi_mpyll_f5hw
 
 #else
 #error MUL type not defined
diff --git a/libgcc/config/msp430/lib2mul.c b/libgcc/config/msp430/lib2mul.c
index 4ceb147..95a2f1a 100644
--- a/libgcc/config/msp430/lib2mul.c
+++ b/libgcc/config/msp430/lib2mul.c
@@ -46,6 +46,9 @@ typedef unsigned int  uint08_type   __attribute__ ((mode (QI)));
 
 #elif defined MUL_16
 
+/* The 16-bit multiply library needs a software version of SI->DI widening
+   multiplication.  */
+
 signed long long
 __mspabi_mpysll (signed long a, signed long b)
 {
diff --git a/libgcc/config/pa/fptr.c b/libgcc/config/pa/fptr.c
index 6cca747..9d2bb73 100644
--- a/libgcc/config/pa/fptr.c
+++ b/libgcc/config/pa/fptr.c
@@ -53,7 +53,7 @@ typedef int (*fixup_t) (struct link_map *, unsigned int);
 extern unsigned int _GLOBAL_OFFSET_TABLE_;
 
 static inline int
-_dl_read_access_allowed (unsigned int *addr)
+_dl_read_access_allowed (unsigned int addr)
 {
   int result;
 
@@ -79,7 +79,8 @@ __canonicalize_funcptr_for_compare (fptr_t fptr)
 {
   static unsigned int fixup_plabel[2] __attribute__((used));
   fixup_t fixup;
-  unsigned int *got, *iptr, *plabel;
+  volatile unsigned int *plabel;
+  unsigned int *got, *iptr, reloc_offset;
   int i;
 
   /* -1 and page 0 are special.  -1 is used in crtend to mark the end of
@@ -94,17 +95,20 @@ __canonicalize_funcptr_for_compare (fptr_t fptr)
      to the entry of the PLT stub just before the global offset table.
      The second word in the plabel contains the relocation offset for the
      function.  */
-  plabel = (unsigned int *) ((unsigned int) fptr & ~3);
-  if (!_dl_read_access_allowed (plabel))
+  plabel = (volatile unsigned int *) ((unsigned int) fptr & ~3);
+  if (!_dl_read_access_allowed ((unsigned int)plabel))
     return (unsigned int) fptr;
 
   /* Load first word of candidate descriptor.  It should be a pointer
      with word alignment and point to memory that can be read.  */
   got = (unsigned int *) plabel[0];
   if (((unsigned int) got & 3) != 0
-      || !_dl_read_access_allowed (got))
+      || !_dl_read_access_allowed ((unsigned int)got))
     return (unsigned int) fptr;
 
+  /* We need to load the relocation offset before the function address.  */
+  reloc_offset = plabel[1];
+  __sync_synchronize();
   got = (unsigned int *) (plabel[0] + GOT_FROM_PLT_STUB);
 
   /* Return the address of the function if the plabel has been resolved.  */
@@ -140,7 +144,7 @@ __canonicalize_funcptr_for_compare (fptr_t fptr)
 
   /* Call fixup to resolve the function address.  got[1] contains the
      link_map pointer and plabel[1] the relocation offset.  */
-  fixup ((struct link_map *) got[1], plabel[1]);
+  fixup ((struct link_map *) got[1], reloc_offset);
 
   return plabel[0];
 }
diff --git a/libgcc/config/pa/lib2funcs.S b/libgcc/config/pa/lib2funcs.S
index b401b61..a2db5b3 100644
--- a/libgcc/config/pa/lib2funcs.S
+++ b/libgcc/config/pa/lib2funcs.S
@@ -55,13 +55,13 @@ __gcc_plt_call
 	; An inline version of dyncall so we don't have to worry
 	; about long calls to millicode, PIC and other complexities.
 	bb,>=,n %r22,30,L$foo
-        depi 0,31,2,%r22
-        ldw 4(%r22),%r19
-        ldw 0(%r22),%r22
+	depi 0,31,2,%r22
+	ldw 0(%r22),%r21
+	ldw 4(%r22),%r19
 L$foo
-        ldsid (%r22),%r1
-        mtsp %r1,%sr0
-        ble 0(%sr0,%r22)
+	ldsid (%r21),%r1
+	mtsp %r1,%sr0
+	ble 0(%sr0,%r21)
 	copy %r31,%r2
 	ldw -8(%r30),%r2
 
diff --git a/libgcc/config/pa/milli64.S b/libgcc/config/pa/milli64.S
index 1e46f50..36040e9 100644
--- a/libgcc/config/pa/milli64.S
+++ b/libgcc/config/pa/milli64.S
@@ -222,19 +222,26 @@ GSYM($$dyncall)
 	.proc
 	.callinfo	millicode
 	.entry
-	bb,>=,n %r22,30,LREF(1)		; branch if not plabel address
-	depi	0,31,2,%r22		; clear the two least significant bits
-	ldw	4(%r22),%r19		; load new LTP value
-	ldw	0(%r22),%r22		; load address of target
-LSYM(1)
 #ifdef LINUX
-	bv	%r0(%r22)		; branch to the real target
+	extru,<>	%r22,30,1,%r0	; nullify if plabel bit set
+	bv,n	%r0(%r22)		; branch to target
+	ldw	-2(%r22),%r21		; load address of target
+	bv	%r0(%r21)		; branch to the real target
+	ldw	2(%r22),%r19		; load new LTP value
 #else
+	bb,>=,n %r22,30,LREF(1)		; branch if not plabel address
+	ldw	-2(%r22),%r21		; load address of target to r21
+	ldsid	(%sr0,%r21),%r1		; get the "space ident" selected by r21
+	ldw	2(%r22),%r19		; load new LTP value
+	mtsp	%r1,%sr0		; move that space identifier into sr0
+	be	0(%sr0,%r21)		; branch to the real target
+	stw	%r2,-24(%r30)		; save return address into frame marker
+LSYM(1)
 	ldsid	(%sr0,%r22),%r1		; get the "space ident" selected by r22
 	mtsp	%r1,%sr0		; move that space identifier into sr0
-	be	0(%sr0,%r22)		; branch to the real target
-#endif
+	be	0(%sr0,%r22)		; branch to the target
 	stw	%r2,-24(%r30)		; save return address into frame marker
+#endif
 	.exit
 	.procend
 #endif