[mirror_edk2.git] / MdePkg / Library / BaseLib / Arm / Math64.S

#------------------------------------------------------------------------------ \r
#\r
# Replacement for Math64.c that is coded to use older GCC intrinsics. \r
# Doing this reduces the number of intrinsics that are required when\r
# you port to a new version of gcc. \r
#\r
# Need to split this into multple files to size optimize the image.\r
#\r
# Copyright (c) 2009 - 2010, Apple Inc. All rights reserved.<BR>\r
# This program and the accompanying materials\r
# are licensed and made available under the terms and conditions of the BSD License\r
# which accompanies this distribution.  The full text of the license may be found at\r
# http://opensource.org/licenses/bsd-license.php.\r
#\r
# THE PROGRAM IS DISTRIBUTED UNDER THE BSD LICENSE ON AN "AS IS" BASIS,\r
# WITHOUT WARRANTIES OR REPRESENTATIONS OF ANY KIND, EITHER EXPRESS OR IMPLIED.\r
#\r
#------------------------------------------------------------------------------\r
\r
	.text\r
	.align 2\r
	GCC_ASM_EXPORT(InternalMathLShiftU64)\r
\r
ASM_PFX(InternalMathLShiftU64):\r
	stmfd	sp!, {r4, r5, r6}\r
	mov	r6, r1\r
	rsb	ip, r2, #32\r
	mov	r4, r6, asl r2\r
	subs	r1, r2, #32\r
	orr	r4, r4, r0, lsr ip\r
	mov	r3, r0, asl r2\r
	movpl	r4, r0, asl r1\r
	mov	r5, r0\r
	mov	r0, r3\r
	mov	r1, r4\r
	ldmfd	sp!, {r4, r5, r6}\r
	bx	lr\r
\r
	.align 2\r
	GCC_ASM_EXPORT(InternalMathRShiftU64)\r
\r
ASM_PFX(InternalMathRShiftU64):\r
	stmfd	sp!, {r4, r5, r6}\r
	mov	r5, r0\r
	rsb	ip, r2, #32\r
	mov	r3, r5, lsr r2\r
	subs	r0, r2, #32\r
	orr	r3, r3, r1, asl ip\r
	mov	r4, r1, lsr r2\r
	movpl	r3, r1, lsr r0\r
	mov	r6, r1\r
	mov	r0, r3\r
	mov	r1, r4\r
	ldmfd	sp!, {r4, r5, r6}\r
	bx	lr\r
\r
	.align 2\r
	GCC_ASM_EXPORT(InternalMathARShiftU64)\r
\r
ASM_PFX(InternalMathARShiftU64):\r
	stmfd	sp!, {r4, r5, r6}\r
	mov	r5, r0\r
	rsb	ip, r2, #32\r
	mov	r3, r5, lsr r2\r
	subs	r0, r2, #32\r
	orr	r3, r3, r1, asl ip\r
	mov	r4, r1, asr r2\r
	movpl	r3, r1, asr r0\r
	mov	r6, r1\r
	mov	r0, r3\r
	mov	r1, r4\r
	ldmfd	sp!, {r4, r5, r6}\r
	bx	lr\r
\r
	.align 2\r
	GCC_ASM_EXPORT(InternalMathLRotU64)\r
\r
ASM_PFX(InternalMathLRotU64):\r
	stmfd	sp!, {r4, r5, r6, r7, lr}\r
	add	r7, sp, #12\r
	mov	r6, r1\r
	rsb	ip, r2, #32\r
	mov	r4, r6, asl r2\r
	rsb	lr, r2, #64\r
	subs	r1, r2, #32\r
	orr	r4, r4, r0, lsr ip\r
	mov	r3, r0, asl r2\r
	movpl	r4, r0, asl r1\r
	sub	ip, r2, #32\r
	mov	r5, r0\r
	mov	r0, r0, lsr lr\r
	rsbs	r2, r2, #32\r
	orr	r0, r0, r6, asl ip\r
	mov	r1, r6, lsr lr\r
	movpl	r0, r6, lsr r2\r
	orr	r1, r1, r4\r
	orr	r0, r0, r3\r
	ldmfd	sp!, {r4, r5, r6, r7, pc}\r
\r
\r
	.align 2\r
	GCC_ASM_EXPORT(InternalMathRRotU64)\r
\r
ASM_PFX(InternalMathRRotU64):\r
	stmfd	sp!, {r4, r5, r6, r7, lr}\r
	add	r7, sp, #12\r
	mov	r5, r0\r
	rsb	ip, r2, #32\r
	mov	r3, r5, lsr r2\r
	rsb	lr, r2, #64\r
	subs	r0, r2, #32\r
	orr	r3, r3, r1, asl ip\r
	mov	r4, r1, lsr r2\r
	movpl	r3, r1, lsr r0\r
	sub	ip, r2, #32\r
	mov	r6, r1\r
	mov	r1, r1, asl lr\r
	rsbs	r2, r2, #32\r
	orr	r1, r1, r5, lsr ip\r
	mov	r0, r5, asl lr\r
	movpl	r1, r5, asl r2\r
	orr	r0, r0, r3\r
	orr	r1, r1, r4\r
	ldmfd	sp!, {r4, r5, r6, r7, pc}\r
\r
	.align 2\r
	GCC_ASM_EXPORT(InternalMathMultU64x32)\r
\r
ASM_PFX(InternalMathMultU64x32):\r
	stmfd	sp!, {r7, lr}\r
	add	r7, sp, #0\r
	mov	r3, #0\r
	mov	ip, r0\r
	mov	lr, r1\r
	umull	r0, r1, ip, r2\r
	mla	r1, lr, r2, r1\r
	mla	r1, ip, r3, r1\r
	ldmfd	sp!, {r7, pc}\r
\r
	.align 2\r
	GCC_ASM_EXPORT(InternalMathMultU64x64)\r
\r
ASM_PFX(InternalMathMultU64x64):\r
	stmfd	sp!, {r7, lr}\r
	add	r7, sp, #0\r
	mov	ip, r0\r
	mov	lr, r1\r
	umull	r0, r1, ip, r2\r
	mla	r1, lr, r2, r1\r
	mla	r1, ip, r3, r1\r
	ldmfd	sp!, {r7, pc}\r
\r
	.align 2\r
	GCC_ASM_EXPORT(InternalMathDivU64x32)\r
\r
ASM_PFX(InternalMathDivU64x32):\r
	stmfd	sp!, {r7, lr}\r
	add	r7, sp, #0\r
	mov	r3, #0\r
	bl 	ASM_PFX(__udivdi3)\r
	ldmfd	sp!, {r7, pc}\r
	\r
	\r
	.align 2\r
	GCC_ASM_EXPORT(InternalMathModU64x32)\r
\r
ASM_PFX(InternalMathModU64x32):\r
	stmfd	sp!, {r7, lr}\r
	add	r7, sp, #0\r
	mov	r3, #0\r
	bl 	ASM_PFX(__umoddi3)\r
	ldmfd	sp!, {r7, pc}\r
	\r
	\r
	.align 2\r
	GCC_ASM_EXPORT(InternalMathDivRemU64x32)\r
\r
ASM_PFX(InternalMathDivRemU64x32):\r
	stmfd	sp!, {r4, r5, r6, r7, lr}\r
	add	r7, sp, #12\r
	stmfd	sp!, {r10, r11}\r
	subs	r6, r3, #0\r
	mov	r10, r0\r
	mov	r11, r1\r
	moveq	r4, r2\r
	moveq	r5, #0\r
	beq	L22\r
	mov	r4, r2\r
	mov	r5, #0\r
	mov	r3, #0\r
	bl 	ASM_PFX(__umoddi3)\r
	str	r0, [r6, #0]\r
L22:\r
	mov	r0, r10\r
	mov	r1, r11\r
	mov	r2, r4\r
	mov	r3, r5\r
	bl 	ASM_PFX(__udivdi3)\r
	ldmfd	sp!, {r10, r11}\r
	ldmfd	sp!, {r4, r5, r6, r7, pc}\r
	\r
	\r
	.align 2\r
	GCC_ASM_EXPORT(InternalMathDivRemU64x64)\r
\r
ASM_PFX(InternalMathDivRemU64x64):\r
	stmfd	sp!, {r4, r5, r6, r7, lr}\r
	add	r7, sp, #12\r
	stmfd	sp!, {r10, r11}\r
	ldr	r6, [sp, #28]\r
	mov	r4, r0\r
	cmp	r6, #0\r
	mov	r5, r1\r
	mov	r10, r2\r
	mov	r11, r3\r
	beq	L26\r
	bl 	ASM_PFX(__umoddi3)\r
	stmia	r6, {r0-r1}\r
L26:\r
	mov	r0, r4\r
	mov	r1, r5\r
	mov	r2, r10\r
	mov	r3, r11\r
	bl 	ASM_PFX(__udivdi3)\r
	ldmfd	sp!, {r10, r11}\r
	ldmfd	sp!, {r4, r5, r6, r7, pc}\r
	\r
	\r
	.align 2\r
	GCC_ASM_EXPORT(InternalMathDivRemS64x64)\r
\r
ASM_PFX(InternalMathDivRemS64x64):\r
	stmfd	sp!, {r4, r5, r6, r7, lr}\r
	add	r7, sp, #12\r
	stmfd	sp!, {r10, r11}\r
	ldr	r6, [sp, #28]\r
	mov	r4, r0\r
	cmp	r6, #0\r
	mov	r5, r1\r
	mov	r10, r2\r
	mov	r11, r3\r
	beq	L30\r
	bl 	ASM_PFX(__moddi3)\r
	stmia	r6, {r0-r1}\r
L30:\r
	mov	r0, r4\r
	mov	r1, r5\r
	mov	r2, r10\r
	mov	r3, r11\r
	bl 	ASM_PFX(__divdi3)\r
	ldmfd	sp!, {r10, r11}\r
	ldmfd	sp!, {r4, r5, r6, r7, pc}\r
	\r
	\r
	.align 2\r
	GCC_ASM_EXPORT(InternalMathSwapBytes64)\r
\r
ASM_PFX(InternalMathSwapBytes64):\r
	stmfd	sp!, {r4, r5, r7, lr}\r
	mov	r5, r1\r
	bl	ASM_PFX(SwapBytes32)\r
	mov	r4, r0\r
	mov	r0, r5\r
	bl	ASM_PFX(SwapBytes32)\r
	mov	r1, r4\r
	ldmfd	sp!, {r4, r5, r7, pc}\r
\r
\r
ASM_FUNCTION_REMOVE_IF_UNREFERENCED
Commit	Line	Data
8db92ab5 HT	1	#------------------------------------------------------------------------------ \r
	2	#\r
	3	# Replacement for Math64.c that is coded to use older GCC intrinsics. \r
	4	# Doing this reduces the number of intrinsics that are required when\r
	5	# you port to a new version of gcc. \r
	6	#\r
	7	# Need to split this into multple files to size optimize the image.\r
	8	#\r
	9	# Copyright (c) 2009 - 2010, Apple Inc. All rights reserved.<BR>\r
	10	# This program and the accompanying materials\r
	11	# are licensed and made available under the terms and conditions of the BSD License\r
	12	# which accompanies this distribution. The full text of the license may be found at\r
	13	# http://opensource.org/licenses/bsd-license.php.\r
	14	#\r
	15	# THE PROGRAM IS DISTRIBUTED UNDER THE BSD LICENSE ON AN "AS IS" BASIS,\r
	16	# WITHOUT WARRANTIES OR REPRESENTATIONS OF ANY KIND, EITHER EXPRESS OR IMPLIED.\r
	17	#\r
	18	#------------------------------------------------------------------------------\r
	19	\r
	20	.text\r
	21	.align 2\r
	22	GCC_ASM_EXPORT(InternalMathLShiftU64)\r
	23	\r
	24	ASM_PFX(InternalMathLShiftU64):\r
	25	stmfd sp!, {r4, r5, r6}\r
	26	mov r6, r1\r
	27	rsb ip, r2, #32\r
	28	mov r4, r6, asl r2\r
	29	subs r1, r2, #32\r
	30	orr r4, r4, r0, lsr ip\r
	31	mov r3, r0, asl r2\r
	32	movpl r4, r0, asl r1\r
	33	mov r5, r0\r
	34	mov r0, r3\r
	35	mov r1, r4\r
	36	ldmfd sp!, {r4, r5, r6}\r
	37	bx lr\r
	38	\r
	39	.align 2\r
	40	GCC_ASM_EXPORT(InternalMathRShiftU64)\r
	41	\r
	42	ASM_PFX(InternalMathRShiftU64):\r
	43	stmfd sp!, {r4, r5, r6}\r
	44	mov r5, r0\r
	45	rsb ip, r2, #32\r
	46	mov r3, r5, lsr r2\r
	47	subs r0, r2, #32\r
	48	orr r3, r3, r1, asl ip\r
	49	mov r4, r1, lsr r2\r
	50	movpl r3, r1, lsr r0\r
	51	mov r6, r1\r
	52	mov r0, r3\r
	53	mov r1, r4\r
	54	ldmfd sp!, {r4, r5, r6}\r
	55	bx lr\r
	56	\r
	57	.align 2\r
	58	GCC_ASM_EXPORT(InternalMathARShiftU64)\r
	59	\r
	60	ASM_PFX(InternalMathARShiftU64):\r
	61	stmfd sp!, {r4, r5, r6}\r
	62	mov r5, r0\r
	63	rsb ip, r2, #32\r
	64	mov r3, r5, lsr r2\r
65	subs r0, r2, #32\r
66	orr r3, r3, r1, asl ip\r
67	mov r4, r1, asr r2\r
68	movpl r3, r1, asr r0\r
69	mov r6, r1\r
70	mov r0, r3\r
71	mov r1, r4\r
72	ldmfd sp!, {r4, r5, r6}\r
73	bx lr\r
74	\r
75	.align 2\r
76	GCC_ASM_EXPORT(InternalMathLRotU64)\r
77	\r
78	ASM_PFX(InternalMathLRotU64):\r
79	stmfd sp!, {r4, r5, r6, r7, lr}\r
80	add r7, sp, #12\r
81	mov r6, r1\r
82	rsb ip, r2, #32\r
83	mov r4, r6, asl r2\r
84	rsb lr, r2, #64\r
85	subs r1, r2, #32\r
86	orr r4, r4, r0, lsr ip\r
87	mov r3, r0, asl r2\r
88	movpl r4, r0, asl r1\r
89	sub ip, r2, #32\r
90	mov r5, r0\r
91	mov r0, r0, lsr lr\r
92	rsbs r2, r2, #32\r
93	orr r0, r0, r6, asl ip\r
94	mov r1, r6, lsr lr\r
95	movpl r0, r6, lsr r2\r
96	orr r1, r1, r4\r
97	orr r0, r0, r3\r
98	ldmfd sp!, {r4, r5, r6, r7, pc}\r
99	\r
100	\r
101	.align 2\r
102	GCC_ASM_EXPORT(InternalMathRRotU64)\r
103	\r
104	ASM_PFX(InternalMathRRotU64):\r
105	stmfd sp!, {r4, r5, r6, r7, lr}\r
106	add r7, sp, #12\r
107	mov r5, r0\r
108	rsb ip, r2, #32\r
109	mov r3, r5, lsr r2\r
110	rsb lr, r2, #64\r
111	subs r0, r2, #32\r
112	orr r3, r3, r1, asl ip\r
113	mov r4, r1, lsr r2\r
114	movpl r3, r1, lsr r0\r
115	sub ip, r2, #32\r
116	mov r6, r1\r
117	mov r1, r1, asl lr\r
118	rsbs r2, r2, #32\r
119	orr r1, r1, r5, lsr ip\r
120	mov r0, r5, asl lr\r
121	movpl r1, r5, asl r2\r
122	orr r0, r0, r3\r
123	orr r1, r1, r4\r
124	ldmfd sp!, {r4, r5, r6, r7, pc}\r
125	\r
126	.align 2\r
127	GCC_ASM_EXPORT(InternalMathMultU64x32)\r
128	\r
129	ASM_PFX(InternalMathMultU64x32):\r
130	stmfd sp!, {r7, lr}\r
131	add r7, sp, #0\r
132	mov r3, #0\r
133	mov ip, r0\r
134	mov lr, r1\r
135	umull r0, r1, ip, r2\r
136	mla r1, lr, r2, r1\r
137	mla r1, ip, r3, r1\r
138	ldmfd sp!, {r7, pc}\r
139	\r
140	.align 2\r
141	GCC_ASM_EXPORT(InternalMathMultU64x64)\r
142	\r
143	ASM_PFX(InternalMathMultU64x64):\r
144	stmfd sp!, {r7, lr}\r
145	add r7, sp, #0\r
146	mov ip, r0\r
147	mov lr, r1\r
148	umull r0, r1, ip, r2\r
149	mla r1, lr, r2, r1\r
150	mla r1, ip, r3, r1\r
151	ldmfd sp!, {r7, pc}\r
152	\r
153	.align 2\r
154	GCC_ASM_EXPORT(InternalMathDivU64x32)\r
155	\r
156	ASM_PFX(InternalMathDivU64x32):\r
157	stmfd sp!, {r7, lr}\r
158	add r7, sp, #0\r
159	mov r3, #0\r
160	bl ASM_PFX(__udivdi3)\r
161	ldmfd sp!, {r7, pc}\r
162	\r
163	\r
164	.align 2\r
165	GCC_ASM_EXPORT(InternalMathModU64x32)\r
166	\r
167	ASM_PFX(InternalMathModU64x32):\r
168	stmfd sp!, {r7, lr}\r
169	add r7, sp, #0\r
170	mov r3, #0\r
171	bl ASM_PFX(__umoddi3)\r
172	ldmfd sp!, {r7, pc}\r
173	\r
174	\r
175	.align 2\r
176	GCC_ASM_EXPORT(InternalMathDivRemU64x32)\r
177	\r
178	ASM_PFX(InternalMathDivRemU64x32):\r
179	stmfd sp!, {r4, r5, r6, r7, lr}\r
180	add r7, sp, #12\r
181	stmfd sp!, {r10, r11}\r
182	subs r6, r3, #0\r
183	mov r10, r0\r
184	mov r11, r1\r
185	moveq r4, r2\r
186	moveq r5, #0\r
187	beq L22\r
188	mov r4, r2\r
189	mov r5, #0\r
190	mov r3, #0\r
191	bl ASM_PFX(__umoddi3)\r
192	str r0, [r6, #0]\r
193	L22:\r
194	mov r0, r10\r
195	mov r1, r11\r
196	mov r2, r4\r
197	mov r3, r5\r
198	bl ASM_PFX(__udivdi3)\r
199	ldmfd sp!, {r10, r11}\r
200	ldmfd sp!, {r4, r5, r6, r7, pc}\r
201	\r
202	\r
203	.align 2\r
204	GCC_ASM_EXPORT(InternalMathDivRemU64x64)\r
205	\r
206	ASM_PFX(InternalMathDivRemU64x64):\r
207	stmfd sp!, {r4, r5, r6, r7, lr}\r
208	add r7, sp, #12\r
209	stmfd sp!, {r10, r11}\r
210	ldr r6, [sp, #28]\r
211	mov r4, r0\r
212	cmp r6, #0\r
213	mov r5, r1\r
214	mov r10, r2\r
215	mov r11, r3\r
216	beq L26\r
217	bl ASM_PFX(__umoddi3)\r
218	stmia r6, {r0-r1}\r
219	L26:\r
220	mov r0, r4\r
221	mov r1, r5\r
222	mov r2, r10\r
223	mov r3, r11\r
224	bl ASM_PFX(__udivdi3)\r
225	ldmfd sp!, {r10, r11}\r
226	ldmfd sp!, {r4, r5, r6, r7, pc}\r
227	\r
228	\r
229	.align 2\r
230	GCC_ASM_EXPORT(InternalMathDivRemS64x64)\r
231	\r
232	ASM_PFX(InternalMathDivRemS64x64):\r
233	stmfd sp!, {r4, r5, r6, r7, lr}\r
234	add r7, sp, #12\r
235	stmfd sp!, {r10, r11}\r
236	ldr r6, [sp, #28]\r
237	mov r4, r0\r
238	cmp r6, #0\r
239	mov r5, r1\r
240	mov r10, r2\r
241	mov r11, r3\r
242	beq L30\r
243	bl ASM_PFX(__moddi3)\r
244	stmia r6, {r0-r1}\r
245	L30:\r
246	mov r0, r4\r
247	mov r1, r5\r
248	mov r2, r10\r
249	mov r3, r11\r
250	bl ASM_PFX(__divdi3)\r
251	ldmfd sp!, {r10, r11}\r
252	ldmfd sp!, {r4, r5, r6, r7, pc}\r
253	\r
254	\r
255	.align 2\r
256	GCC_ASM_EXPORT(InternalMathSwapBytes64)\r
257	\r
258	ASM_PFX(InternalMathSwapBytes64):\r
55f5af18	259	stmfd sp!, {r4, r5, r7, lr}\r
8db92ab5 HT	260	mov r5, r1\r
8db92ab5 HT	261	bl ASM_PFX(SwapBytes32)\r
55f5af18	262	mov r4, r0\r
8db92ab5 HT	263	mov r0, r5\r
8db92ab5 HT	264	bl ASM_PFX(SwapBytes32)\r
55f5af18 AB	265	mov r1, r4\r
55f5af18 AB	266	ldmfd sp!, {r4, r5, r7, pc}\r
8db92ab5 HT	267	\r
8db92ab5 HT	268	\r
703f1d09	269	ASM_FUNCTION_REMOVE_IF_UNREFERENCED