[thirdparty/glibc.git] / sysdeps / powerpc / powerpc32 / fpu / setjmp-common.S

/* setjmp for PowerPC.
   Copyright (C) 1995-99, 2000, 2003, 2004 Free Software Foundation, Inc.
   This file is part of the GNU C Library.

   The GNU C Library is free software; you can redistribute it and/or
   modify it under the terms of the GNU Lesser General Public
   License as published by the Free Software Foundation; either
   version 2.1 of the License, or (at your option) any later version.

   The GNU C Library is distributed in the hope that it will be useful,
   but WITHOUT ANY WARRANTY; without even the implied warranty of
   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
   Lesser General Public License for more details.

   You should have received a copy of the GNU Lesser General Public
   License along with the GNU C Library; if not, write to the Free
   Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
   02111-1307 USA.  */

#include <sysdep.h>
#define _ASM
#define _SETJMP_H
#ifdef __NO_VMX__
# include <novmxsetjmp.h>
#else
# include <bits/setjmp.h>
#endif
#include <bp-sym.h>
#include <bp-asm.h>


ENTRY (BP_SYM (__sigsetjmp))
	CHECK_BOUNDS_BOTH_WIDE_LIT (r3, r8, r9, JB_SIZE)

	stw  r1,(JB_GPR1*4)(3)
	mflr r0
	stw  r14,((JB_GPRS+0)*4)(3)
	stfd fp14,((JB_FPRS+0*2)*4)(3)
	stw  r0,(JB_LR*4)(3)
	stw  r15,((JB_GPRS+1)*4)(3)
	stfd fp15,((JB_FPRS+1*2)*4)(3)
	mfcr r0
	stw  r16,((JB_GPRS+2)*4)(3)
	stfd fp16,((JB_FPRS+2*2)*4)(3)
	stw  r0,(JB_CR*4)(3)
	stw  r17,((JB_GPRS+3)*4)(3)
	stfd fp17,((JB_FPRS+3*2)*4)(3)
	stw  r18,((JB_GPRS+4)*4)(3)
	stfd fp18,((JB_FPRS+4*2)*4)(3)
	stw  r19,((JB_GPRS+5)*4)(3)
	stfd fp19,((JB_FPRS+5*2)*4)(3)
	stw  r20,((JB_GPRS+6)*4)(3)
	stfd fp20,((JB_FPRS+6*2)*4)(3)
	stw  r21,((JB_GPRS+7)*4)(3)
	stfd fp21,((JB_FPRS+7*2)*4)(3)
	stw  r22,((JB_GPRS+8)*4)(3)
	stfd fp22,((JB_FPRS+8*2)*4)(3)
	stw  r23,((JB_GPRS+9)*4)(3)
	stfd fp23,((JB_FPRS+9*2)*4)(3)
	stw  r24,((JB_GPRS+10)*4)(3)
	stfd fp24,((JB_FPRS+10*2)*4)(3)
	stw  r25,((JB_GPRS+11)*4)(3)
	stfd fp25,((JB_FPRS+11*2)*4)(3)
	stw  r26,((JB_GPRS+12)*4)(3)
	stfd fp26,((JB_FPRS+12*2)*4)(3)
	stw  r27,((JB_GPRS+13)*4)(3)
	stfd fp27,((JB_FPRS+13*2)*4)(3)
	stw  r28,((JB_GPRS+14)*4)(3)
	stfd fp28,((JB_FPRS+14*2)*4)(3)
	stw  r29,((JB_GPRS+15)*4)(3)
	stfd fp29,((JB_FPRS+15*2)*4)(3)
	stw  r30,((JB_GPRS+16)*4)(3)
	stfd fp30,((JB_FPRS+16*2)*4)(3)
	stw  r31,((JB_GPRS+17)*4)(3)
	stfd fp31,((JB_FPRS+17*2)*4)(3)
#ifndef __NO_VMX__
#ifdef PIC
	mflr    r6
	bl      _GLOBAL_OFFSET_TABLE_@local-4
	mflr    r5
#ifdef SHARED
	lwz     r5,_rtld_global@got(r5)
	mtlr    r6
	lwz     r5,RTLD_GLOBAL_DL_HWCAP_OFFSET(r5)
#else
	lwz     r5,_rtld_global@got(r5)
	mtlr    r6
	lwz     r5,0(r5)
#endif
#else
	lis	r5,_dl_hwcap@ha
	lwz     r5,_dl_hwcap@l(r5)
#endif
	andis.	r5,r5,(PPC_FEATURE_HAS_ALTIVEC >> 16)
	beq	no_vmx
	la	r5,((JB_VRS)*4)(3)
	andi.	r6,r5,0xf
	mfspr	r0,VRSAVE
	stw	r0,((JB_VRSAVE)*4)(3)
	addi	r6,r5,16
	beq+	aligned_save_vmx
	lvsr	v0,0,r5
	vspltisb v1,-1         /* set v1 to all 1's */
	vspltisb v2,0          /* set v2 to all 0's */
	vperm   v3,v2,v1,v0   /* v3 contains shift mask with num all 1 bytes on left = misalignment  */


	/* Special case for v20 we need to preserve what is in save area below v20 before obliterating it */
	lvx     v5,0,r5
	vperm   v20,v20,v20,v0
	vsel    v5,v5,v20,v3
	vsel    v20,v20,v2,v3
	stvx    v5,0,r5

#define save_2vmx_partial(savevr,prev_savevr,hivr,shiftvr,maskvr,savegpr,addgpr) \
	addi    addgpr,addgpr,32; \
	vperm   savevr,savevr,savevr,shiftvr; \
	vsel    hivr,prev_savevr,savevr,maskvr; \
	stvx    hivr,0,savegpr;

	save_2vmx_partial(v21,v20,v5,v0,v3,r6,r5)
	save_2vmx_partial(v22,v21,v5,v0,v3,r5,r6)
	save_2vmx_partial(v23,v22,v5,v0,v3,r6,r5)
	save_2vmx_partial(v24,v23,v5,v0,v3,r5,r6)
	save_2vmx_partial(v25,v24,v5,v0,v3,r6,r5)
	save_2vmx_partial(v26,v25,v5,v0,v3,r5,r6)
	save_2vmx_partial(v27,v26,v5,v0,v3,r6,r5)
	save_2vmx_partial(v28,v27,v5,v0,v3,r5,r6)
	save_2vmx_partial(v29,v28,v5,v0,v3,r6,r5)
	save_2vmx_partial(v30,v29,v5,v0,v3,r5,r6)

	/* Special case for r31 we need to preserve what is in save area above v31 before obliterating it */
	addi    r5,r5,32
	vperm   v31,v31,v31,v0
	lvx     v4,0,r5
	vsel    v5,v30,v31,v3
	stvx    v5,0,r6
	vsel    v4,v31,v4,v3
	stvx    v4,0,r5
	b	no_vmx

aligned_save_vmx:
	stvx	20,0,r5
	addi	r5,r5,32
	stvx	21,0,r6
	addi	r6,r6,32
	stvx	22,0,r5
	addi	r5,r5,32
	stvx	23,0,r6
	addi	r6,r6,32
	stvx	24,0,r5
	addi	r5,r5,32
	stvx	25,0,r6
	addi	r6,r6,32
	stvx	26,0,r5
	addi	r5,r5,32
	stvx	27,0,r6
	addi	r6,r6,32
	stvx	28,0,r5
	addi	r5,r5,32
	stvx	29,0,r6
	addi	r6,r6,32
	stvx	30,0,r5
	stvx	31,0,r6
no_vmx:
#endif
	b JUMPTARGET (BP_SYM (__sigjmp_save))
END (BP_SYM (__sigsetjmp))
Commit	Line	Data
5c76ff27 UD	1	/* setjmp for PowerPC.
	2	Copyright (C) 1995-99, 2000, 2003, 2004 Free Software Foundation, Inc.
	3	This file is part of the GNU C Library.
	4
	5	The GNU C Library is free software; you can redistribute it and/or
	6	modify it under the terms of the GNU Lesser General Public
	7	License as published by the Free Software Foundation; either
	8	version 2.1 of the License, or (at your option) any later version.
	9
	10	The GNU C Library is distributed in the hope that it will be useful,
	11	but WITHOUT ANY WARRANTY; without even the implied warranty of
	12	MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
	13	Lesser General Public License for more details.
	14
	15	You should have received a copy of the GNU Lesser General Public
	16	License along with the GNU C Library; if not, write to the Free
	17	Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
	18	02111-1307 USA. */
	19
	20	#include <sysdep.h>
	21	#define _ASM
	22	#define _SETJMP_H
	23	#ifdef __NO_VMX__
	24	# include <novmxsetjmp.h>
	25	#else
	26	# include <bits/setjmp.h>
	27	#endif
	28	#include <bp-sym.h>
	29	#include <bp-asm.h>
	30
	31
	32	ENTRY (BP_SYM (__sigsetjmp))
	33	CHECK_BOUNDS_BOTH_WIDE_LIT (r3, r8, r9, JB_SIZE)
	34
	35	stw r1,(JB_GPR1*4)(3)
	36	mflr r0
	37	stw r14,((JB_GPRS+0)*4)(3)
	38	stfd fp14,((JB_FPRS+02)4)(3)
	39	stw r0,(JB_LR*4)(3)
	40	stw r15,((JB_GPRS+1)*4)(3)
	41	stfd fp15,((JB_FPRS+12)4)(3)
	42	mfcr r0
	43	stw r16,((JB_GPRS+2)*4)(3)
	44	stfd fp16,((JB_FPRS+22)4)(3)
	45	stw r0,(JB_CR*4)(3)
	46	stw r17,((JB_GPRS+3)*4)(3)
	47	stfd fp17,((JB_FPRS+32)4)(3)
	48	stw r18,((JB_GPRS+4)*4)(3)
	49	stfd fp18,((JB_FPRS+42)4)(3)
	50	stw r19,((JB_GPRS+5)*4)(3)
	51	stfd fp19,((JB_FPRS+52)4)(3)
	52	stw r20,((JB_GPRS+6)*4)(3)
	53	stfd fp20,((JB_FPRS+62)4)(3)
	54	stw r21,((JB_GPRS+7)*4)(3)
	55	stfd fp21,((JB_FPRS+72)4)(3)
	56	stw r22,((JB_GPRS+8)*4)(3)
	57	stfd fp22,((JB_FPRS+82)4)(3)
	58	stw r23,((JB_GPRS+9)*4)(3)
	59	stfd fp23,((JB_FPRS+92)4)(3)
	60	stw r24,((JB_GPRS+10)*4)(3)
	61	stfd fp24,((JB_FPRS+102)4)(3)
	62	stw r25,((JB_GPRS+11)*4)(3)
	63	stfd fp25,((JB_FPRS+112)4)(3)
	64	stw r26,((JB_GPRS+12)*4)(3)
65	stfd fp26,((JB_FPRS+122)4)(3)
66	stw r27,((JB_GPRS+13)*4)(3)
67	stfd fp27,((JB_FPRS+132)4)(3)
68	stw r28,((JB_GPRS+14)*4)(3)
69	stfd fp28,((JB_FPRS+142)4)(3)
70	stw r29,((JB_GPRS+15)*4)(3)
71	stfd fp29,((JB_FPRS+152)4)(3)
72	stw r30,((JB_GPRS+16)*4)(3)
73	stfd fp30,((JB_FPRS+162)4)(3)
74	stw r31,((JB_GPRS+17)*4)(3)
75	stfd fp31,((JB_FPRS+172)4)(3)
76	#ifndef __NO_VMX__
77	#ifdef PIC
78	mflr r6
79	bl _GLOBAL_OFFSET_TABLE_@local-4
80	mflr r5
81	#ifdef SHARED
82	lwz r5,_rtld_global@got(r5)
83	mtlr r6
84	lwz r5,RTLD_GLOBAL_DL_HWCAP_OFFSET(r5)
85	#else
86	lwz r5,_rtld_global@got(r5)
87	mtlr r6
88	lwz r5,0(r5)
89	#endif
90	#else
91	lis r5,_dl_hwcap@ha
92	lwz r5,_dl_hwcap@l(r5)
93	#endif
94	andis. r5,r5,(PPC_FEATURE_HAS_ALTIVEC >> 16)
95	beq no_vmx
96	la r5,((JB_VRS)*4)(3)
97	andi. r6,r5,0xf
98	mfspr r0,VRSAVE
99	stw r0,((JB_VRSAVE)*4)(3)
100	addi r6,r5,16
101	beq+ aligned_save_vmx
102	lvsr v0,0,r5
103	vspltisb v1,-1 /* set v1 to all 1's */
104	vspltisb v2,0 /* set v2 to all 0's */
105	vperm v3,v2,v1,v0 /* v3 contains shift mask with num all 1 bytes on left = misalignment */
106
107
108	/* Special case for v20 we need to preserve what is in save area below v20 before obliterating it */
109	lvx v5,0,r5
110	vperm v20,v20,v20,v0
111	vsel v5,v5,v20,v3
112	vsel v20,v20,v2,v3
113	stvx v5,0,r5
114
115	#define save_2vmx_partial(savevr,prev_savevr,hivr,shiftvr,maskvr,savegpr,addgpr) \
116	addi addgpr,addgpr,32; \
117	vperm savevr,savevr,savevr,shiftvr; \
118	vsel hivr,prev_savevr,savevr,maskvr; \
119	stvx hivr,0,savegpr;
120
121	save_2vmx_partial(v21,v20,v5,v0,v3,r6,r5)
122	save_2vmx_partial(v22,v21,v5,v0,v3,r5,r6)
123	save_2vmx_partial(v23,v22,v5,v0,v3,r6,r5)
124	save_2vmx_partial(v24,v23,v5,v0,v3,r5,r6)
125	save_2vmx_partial(v25,v24,v5,v0,v3,r6,r5)
126	save_2vmx_partial(v26,v25,v5,v0,v3,r5,r6)
127	save_2vmx_partial(v27,v26,v5,v0,v3,r6,r5)
128	save_2vmx_partial(v28,v27,v5,v0,v3,r5,r6)
129	save_2vmx_partial(v29,v28,v5,v0,v3,r6,r5)
130	save_2vmx_partial(v30,v29,v5,v0,v3,r5,r6)
131
132	/* Special case for r31 we need to preserve what is in save area above v31 before obliterating it */
133	addi r5,r5,32
134	vperm v31,v31,v31,v0
135	lvx v4,0,r5
136	vsel v5,v30,v31,v3
137	stvx v5,0,r6
138	vsel v4,v31,v4,v3
139	stvx v4,0,r5
140	b no_vmx
141
142	aligned_save_vmx:
143	stvx 20,0,r5
144	addi r5,r5,32
145	stvx 21,0,r6
146	addi r6,r6,32
147	stvx 22,0,r5
148	addi r5,r5,32
149	stvx 23,0,r6
150	addi r6,r6,32
151	stvx 24,0,r5
152	addi r5,r5,32
153	stvx 25,0,r6
154	addi r6,r6,32
155	stvx 26,0,r5
156	addi r5,r5,32
157	stvx 27,0,r6
158	addi r6,r6,32
159	stvx 28,0,r5
160	addi r5,r5,32
161	stvx 29,0,r6
162	addi r6,r6,32
163	stvx 30,0,r5
164	stvx 31,0,r6
165	no_vmx:
166	#endif
167	b JUMPTARGET (BP_SYM (__sigjmp_save))
168	END (BP_SYM (__sigsetjmp))