[gcc.git] / libffi / src / powerpc / ppc_closure.S

#define LIBFFI_ASM
#include <powerpc/asm.h>

        .file   "ppc_closure.S"

ENTRY(ffi_closure_SYSV)
.LFB1:
	stwu %r1,-144(%r1)
.LCFI0:
	mflr %r0
.LCFI1:
	stw %r0,148(%r1)

# we want to build up an areas for the parameters passed
# in registers (both floating point and integer)
	
	# so first save gpr 3 to gpr 10 (aligned to 4)
	stw   %r3, 16(%r1)
	stw   %r4, 20(%r1)
	stw   %r5, 24(%r1) 
	stw   %r6, 28(%r1)
	stw   %r7, 32(%r1)
	stw   %r8, 36(%r1) 
	stw   %r9, 40(%r1)
	stw   %r10,44(%r1)

	# next save fpr 1 to fpr 8 (aligned to 8)
	stfd  %f1, 48(%r1)
	stfd  %f2, 56(%r1)
	stfd  %f3, 64(%r1)
	stfd  %f4, 72(%r1)
	stfd  %f5, 80(%r1)
	stfd  %f6, 88(%r1)
	stfd  %f7, 96(%r1)
	stfd  %f8, 104(%r1)

	# set up registers for the routine that actually does the work
	# get the context pointer from the trampoline
	mr %r3,%r11
	
        # now load up the pointer to the result storage
	addi %r4,%r1,112
	
	# now load up the pointer to the saved gpr registers
        addi %r5,%r1,16

        # now load up the pointer to the saved fpr registers */
        addi %r6,%r1,48

	# now load up the pointer to the outgoing parameter 
	# stack in the previous frame
	# i.e. the previous frame pointer + 8
	addi %r7,%r1,152
	
        # make the call
	bl JUMPTARGET(ffi_closure_helper_SYSV)

	# now r3 contains the return type
	# so use it to look up in a table
	# so we know how to deal with each type

        # look up the proper starting point in table 
	# by using return type as offset
	addi %r5,%r1,112   # get pointer to results area
	bl .Lget_ret_type0_addr # get pointer to .Lret_type0 into LR
	mflr %r4           # move to r4
	slwi %r3,%r3,4     # now multiply return type by 16
	add %r3,%r3,%r4    # add contents of table to table address
	mtctr %r3
	bctr               # jump to it
.LFE1:

# Each of the ret_typeX code fragments has to be exactly 16 bytes long
# (4 instructions). For cache effectiveness we align to a 16 byte boundary
# first.
	.align 4

	nop
	nop
	nop
.Lget_ret_type0_addr:
	blrl

# case FFI_TYPE_VOID
.Lret_type0:
	b .Lfinish
	nop
	nop
	nop

# case FFI_TYPE_INT
.Lret_type1:
	lwz %r3,0(%r5)
	b .Lfinish
	nop
	nop

# case FFI_TYPE_FLOAT
.Lret_type2:
	lfs %f1,0(%r5)
	b .Lfinish
	nop
	nop

# case FFI_TYPE_DOUBLE
.Lret_type3:
        lfd %f1,0(%r5)
	b .Lfinish
	nop
	nop

# case FFI_TYPE_LONGDOUBLE
.Lret_type4:
        lfd %f1,0(%r5)
	b .Lfinish
	nop
	nop

# case FFI_TYPE_UINT8
.Lret_type5:
        lbz %r3,3(%r5)
	b .Lfinish
	nop
	nop

# case FFI_TYPE_SINT8
.Lret_type6:
	lbz %r3,3(%r5)
	extsb %r3,%r3
	b .Lfinish
	nop

# case FFI_TYPE_UINT16
.Lret_type7:
	lhz %r3,2(%r5)
	b .Lfinish
	nop
	nop

# case FFI_TYPE_SINT16
.Lret_type8:
	lha %r3,2(%r5)
	b .Lfinish
	nop
	nop

# case FFI_TYPE_UINT32
.Lret_type9:
	lwz %r3,0(%r5)
	b .Lfinish
	nop
	nop

# case FFI_TYPE_SINT32
.Lret_type10:
	lwz %r3,0(%r5)
	b .Lfinish
	nop
	nop

# case FFI_TYPE_UINT64
.Lret_type11:
	lwz %r3,0(%r5)
	lwz %r4,4(%r5)
	b .Lfinish
	nop

# case FFI_TYPE_SINT64
.Lret_type12:
	lwz %r3,0(%r5)
	lwz %r4,4(%r5)
	b .Lfinish
	nop

# case FFI_TYPE_STRUCT
.Lret_type13:
	b .Lfinish
	nop
	nop
	nop

# case FFI_TYPE_POINTER
.Lret_type14:
	lwz %r3,0(%r5)
	b .Lfinish
	nop
	nop

# case done	
.Lfinish:
	
	lwz %r0,148(%r1)
	mtlr %r0
	addi %r1,%r1,144
	blr
END(ffi_closure_SYSV)

	.section	".eh_frame","aw"
__FRAME_BEGIN__:
	.4byte	.LECIE1-.LSCIE1	 # Length of Common Information Entry
.LSCIE1:
	.4byte	0x0	 # CIE Identifier Tag
	.byte	0x1	 # CIE Version
	.ascii "\0"	 # CIE Augmentation
	.byte	0x1	 # uleb128 0x1; CIE Code Alignment Factor
	.byte	0x7c	 # sleb128 -4; CIE Data Alignment Factor
	.byte	0x41	 # CIE RA Column
	.byte	0xc	 # DW_CFA_def_cfa
	.byte	0x1	 # uleb128 0x1
	.byte	0x0	 # uleb128 0x0
	.align 2
.LECIE1:
.LSFDE1:
	.4byte	.LEFDE1-.LASFDE1	 # FDE Length
.LASFDE1:
	.4byte	.LASFDE1-__FRAME_BEGIN__	 # FDE CIE offset
	.4byte	.LFB1	 # FDE initial location
	.4byte	.LFE1-.LFB1	 # FDE address range
	.byte	0x4	 # DW_CFA_advance_loc4
	.4byte	.LCFI0-.LFB1
	.byte	0xe	 # DW_CFA_def_cfa_offset
	.byte	144,1	 # uleb128 144
	.byte	0x4	 # DW_CFA_advance_loc4
	.4byte	.LCFI1-.LCFI0
	.byte	0x2f	 # DW_CFA_GNU_negative_offset_extended
	.byte	0x41	 # uleb128 0x41
	.byte	0x1	 # uleb128 0x1
	.align 2
.LEFDE1:
Commit	Line	Data
cc4c8975 KH	1	#define LIBFFI_ASM
	2	#include <powerpc/asm.h>
	3
c33402e2	4	.file "ppc_closure.S"
cc4c8975 KH	5
cc4c8975 KH	6	ENTRY(ffi_closure_SYSV)
552cb45a	7	.LFB1:
cc4c8975	8	stwu %r1,-144(%r1)
552cb45a	9	.LCFI0:
cc4c8975	10	mflr %r0
552cb45a	11	.LCFI1:
cc4c8975 KH	12	stw %r0,148(%r1)
	13
	14	# we want to build up an areas for the parameters passed
	15	# in registers (both floating point and integer)
	16
	17	# so first save gpr 3 to gpr 10 (aligned to 4)
	18	stw %r3, 16(%r1)
	19	stw %r4, 20(%r1)
	20	stw %r5, 24(%r1)
	21	stw %r6, 28(%r1)
	22	stw %r7, 32(%r1)
	23	stw %r8, 36(%r1)
	24	stw %r9, 40(%r1)
	25	stw %r10,44(%r1)
	26
	27	# next save fpr 1 to fpr 8 (aligned to 8)
	28	stfd %f1, 48(%r1)
	29	stfd %f2, 56(%r1)
	30	stfd %f3, 64(%r1)
	31	stfd %f4, 72(%r1)
	32	stfd %f5, 80(%r1)
	33	stfd %f6, 88(%r1)
	34	stfd %f7, 96(%r1)
	35	stfd %f8, 104(%r1)
	36
	37	# set up registers for the routine that actually does the work
	38	# get the context pointer from the trampoline
	39	mr %r3,%r11
	40
	41	# now load up the pointer to the result storage
	42	addi %r4,%r1,112
	43
	44	# now load up the pointer to the saved gpr registers
	45	addi %r5,%r1,16
	46
	47	# now load up the pointer to the saved fpr registers */
	48	addi %r6,%r1,48
	49
	50	# now load up the pointer to the outgoing parameter
	51	# stack in the previous frame
	52	# i.e. the previous frame pointer + 8
	53	addi %r7,%r1,152
	54
	55	# make the call
	56	bl JUMPTARGET(ffi_closure_helper_SYSV)
	57
	58	# now r3 contains the return type
	59	# so use it to look up in a table
	60	# so we know how to deal with each type
	61
	62	# look up the proper starting point in table
	63	# by using return type as offset
	64	addi %r5,%r1,112 # get pointer to results area
c33402e2 FS	65	bl .Lget_ret_type0_addr # get pointer to .Lret_type0 into LR
	66	mflr %r4 # move to r4
	67	slwi %r3,%r3,4 # now multiply return type by 16
	68	add %r3,%r3,%r4 # add contents of table to table address
cc4c8975 KH	69	mtctr %r3
cc4c8975 KH	70	bctr # jump to it
552cb45a	71	.LFE1:
cc4c8975	72
c33402e2 FS	73	# Each of the ret_typeX code fragments has to be exactly 16 bytes long
	74	# (4 instructions). For cache effectiveness we align to a 16 byte boundary
	75	# first.
	76	.align 4
	77
	78	nop
	79	nop
	80	nop
	81	.Lget_ret_type0_addr:
	82	blrl
	83
	84	# case FFI_TYPE_VOID
	85	.Lret_type0:
	86	b .Lfinish
	87	nop
	88	nop
	89	nop
	90
	91	# case FFI_TYPE_INT
	92	.Lret_type1:
	93	lwz %r3,0(%r5)
	94	b .Lfinish
	95	nop
	96	nop
	97
	98	# case FFI_TYPE_FLOAT
	99	.Lret_type2:
cc4c8975	100	lfs %f1,0(%r5)
c33402e2 FS	101	b .Lfinish
	102	nop
	103	nop
	104
	105	# case FFI_TYPE_DOUBLE
	106	.Lret_type3:
	107	lfd %f1,0(%r5)
	108	b .Lfinish
	109	nop
	110	nop
	111
	112	# case FFI_TYPE_LONGDOUBLE
	113	.Lret_type4:
	114	lfd %f1,0(%r5)
	115	b .Lfinish
	116	nop
	117	nop
	118
	119	# case FFI_TYPE_UINT8
	120	.Lret_type5:
	121	lbz %r3,3(%r5)
	122	b .Lfinish
	123	nop
	124	nop
	125
	126	# case FFI_TYPE_SINT8
	127	.Lret_type6:
	128	lbz %r3,3(%r5)
	129	extsb %r3,%r3
	130	b .Lfinish
	131	nop
	132
	133	# case FFI_TYPE_UINT16
	134	.Lret_type7:
	135	lhz %r3,2(%r5)
	136	b .Lfinish
	137	nop
	138	nop
	139
	140	# case FFI_TYPE_SINT16
	141	.Lret_type8:
	142	lha %r3,2(%r5)
	143	b .Lfinish
	144	nop
	145	nop
	146
	147	# case FFI_TYPE_UINT32
	148	.Lret_type9:
	149	lwz %r3,0(%r5)
	150	b .Lfinish
	151	nop
	152	nop
	153
	154	# case FFI_TYPE_SINT32
	155	.Lret_type10:
	156	lwz %r3,0(%r5)
	157	b .Lfinish
	158	nop
	159	nop
	160
	161	# case FFI_TYPE_UINT64
	162	.Lret_type11:
cc4c8975 KH	163	lwz %r3,0(%r5)
cc4c8975 KH	164	lwz %r4,4(%r5)
c33402e2 FS	165	b .Lfinish
	166	nop
	167
	168	# case FFI_TYPE_SINT64
	169	.Lret_type12:
cc4c8975	170	lwz %r3,0(%r5)
c33402e2 FS	171	lwz %r4,4(%r5)
	172	b .Lfinish
	173	nop
	174
	175	# case FFI_TYPE_STRUCT
	176	.Lret_type13:
	177	b .Lfinish
	178	nop
	179	nop
	180	nop
	181
	182	# case FFI_TYPE_POINTER
	183	.Lret_type14:
	184	lwz %r3,0(%r5)
	185	b .Lfinish
	186	nop
	187	nop
	188
	189	# case done
	190	.Lfinish:
cc4c8975	191
c33402e2	192	lwz %r0,148(%r1)
cc4c8975	193	mtlr %r0
c33402e2	194	addi %r1,%r1,144
cc4c8975 KH	195	blr
	196	END(ffi_closure_SYSV)
	197
552cb45a AH	198	.section ".eh_frame","aw"
	199	__FRAME_BEGIN__:
	200	.4byte .LECIE1-.LSCIE1 # Length of Common Information Entry
	201	.LSCIE1:
	202	.4byte 0x0 # CIE Identifier Tag
	203	.byte 0x1 # CIE Version
	204	.ascii "\0" # CIE Augmentation
	205	.byte 0x1 # uleb128 0x1; CIE Code Alignment Factor
	206	.byte 0x7c # sleb128 -4; CIE Data Alignment Factor
	207	.byte 0x41 # CIE RA Column
	208	.byte 0xc # DW_CFA_def_cfa
	209	.byte 0x1 # uleb128 0x1
	210	.byte 0x0 # uleb128 0x0
	211	.align 2
	212	.LECIE1:
	213	.LSFDE1:
	214	.4byte .LEFDE1-.LASFDE1 # FDE Length
	215	.LASFDE1:
	216	.4byte .LASFDE1-__FRAME_BEGIN__ # FDE CIE offset
	217	.4byte .LFB1 # FDE initial location
	218	.4byte .LFE1-.LFB1 # FDE address range
	219	.byte 0x4 # DW_CFA_advance_loc4
	220	.4byte .LCFI0-.LFB1
	221	.byte 0xe # DW_CFA_def_cfa_offset
	222	.byte 144,1 # uleb128 144
	223	.byte 0x4 # DW_CFA_advance_loc4
	224	.4byte .LCFI1-.LCFI0
	225	.byte 0x2f # DW_CFA_GNU_negative_offset_extended
	226	.byte 0x41 # uleb128 0x41
	227	.byte 0x1 # uleb128 0x1
	228	.align 2
	229	.LEFDE1: