Move setjmp to lib and add setjmp for sh (imported from newlib) and compile it as separate library.

It's used by both, plugins and codecs, and sh/hwcodec doesn't compile codecs
so it doesn't fit into sources.

git-svn-id: svn://svn.rockbox.org/rockbox/trunk@27869 a1c6a512-1295-4272-9138-f99709370657

2 files changed, 236 insertions, 0 deletions

diff --git a/lib/libsetjmp/sh/asm.h b/lib/libsetjmp/sh/asm.h
new file mode 100644
index 0000000000..db30b149b5
--- /dev/null
+++ b/lib/libsetjmp/sh/asm.h
@@ -0,0 +1,42 @@
+#ifdef __STDC__
+# define _C_LABEL(x)    _ ## x
+#else
+# define _C_LABEL(x)    _/**/x
+#endif
+#define _ASM_LABEL(x)   x
+
+#if __SH5__
+# if __SH5__ == 32 && __SHMEDIA__
+#  define TEXT .section .text..SHmedia32, "ax"
+# else
+#  define TEXT .text
+# endif
+
+# define _ENTRY(name)	\
+	TEXT; .balign 8; .globl name; name:
+#else
+#define _ENTRY(name)	\
+	.text; .align 2; .globl name; name:
+#endif /* __SH5__ */
+
+#define ENTRY(name)	\
+	_ENTRY(_C_LABEL(name))
+
+#if (defined (__sh2__) || defined (__SH2E__) || defined (__sh3__) || defined (__SH3E__) \
+     || defined (__SH4_SINGLE__) || defined (__SH4__)) \
+     || defined (__SH4_SINGLE_ONLY__) || defined (__SH5__) || defined (__SH2A__)
+#define DELAYED_BRANCHES
+#define SL(branch, dest, in_slot, in_slot_arg2) \
+	branch##.s dest; in_slot, in_slot_arg2
+#else
+#define SL(branch, dest, in_slot, in_slot_arg2) \
+	in_slot, in_slot_arg2; branch dest
+#endif
+
+#ifdef __LITTLE_ENDIAN__
+#define SHHI shlld
+#define SHLO shlrd
+#else
+#define SHHI shlrd
+#define SHLO shlld
+#endif
diff --git a/lib/libsetjmp/sh/setjmp.S b/lib/libsetjmp/sh/setjmp.S
new file mode 100644
index 0000000000..7df2054531
--- /dev/null
+++ b/lib/libsetjmp/sh/setjmp.S
@@ -0,0 +1,194 @@
+#include "config.h"
+#ifdef ROCKBOX_LITTLE_ENDIAN
+#define __LITTLE_ENDIAN__
+#endif
+
+/* SH7034, the only supported sh target supported by Rockbox is sh1 */
+#define __SH5__ 0
+#define __SHMEDIA__ 0
+
+/* below is unmodified setjmp.S revision 1.3 from newlib cvs */
+#if __SHMEDIA__
+	.mode	SHmedia
+#endif
+
+#include "asm.h"
+
+ENTRY(setjmp)
+#if __SH5__
+	ptabs	r18, tr0
+	gettr	tr5, r5
+	gettr	tr6, r6
+	gettr	tr7, r7
+	st.q	r2,  0*8, r18
+	st.q	r2,  1*8, r10
+	st.q	r2,  2*8, r11
+	st.q	r2,  3*8, r12
+	st.q	r2,  4*8, r13
+	st.q	r2,  5*8, r14
+	st.q	r2,  6*8, r15
+	st.q	r2,  7*8, r28
+	st.q	r2,  8*8, r29
+	st.q	r2,  9*8, r30
+	st.q	r2, 10*8, r31
+	st.q	r2, 11*8, r32
+	st.q	r2, 12*8, r33
+	st.q	r2, 13*8, r34
+	st.q	r2, 14*8, r35
+	st.q	r2, 15*8, r44
+	st.q	r2, 16*8, r45
+	st.q	r2, 17*8, r46
+	st.q	r2, 18*8, r47
+	st.q	r2, 19*8, r48
+	st.q	r2, 20*8, r49
+	st.q	r2, 21*8, r50
+	st.q	r2, 22*8, r51
+	st.q	r2, 23*8, r52
+	st.q	r2, 24*8, r53
+	st.q	r2, 25*8, r54
+	st.q	r2, 26*8, r55
+	st.q	r2, 27*8, r56
+	st.q	r2, 28*8, r57
+	st.q	r2, 29*8, r58
+	st.q	r2, 30*8, r59
+	st.q	r2, 31*8, r5
+	st.q	r2, 32*8, r6
+	st.q	r2, 33*8, r7
+#if ! __SH4_NOFPU__
+	fst.d	r2, 34*8, dr12
+	fst.d	r2, 35*8, dr14
+	fst.d	r2, 36*8, dr36
+	fst.d	r2, 37*8, dr38
+	fst.d	r2, 38*8, dr40
+	fst.d	r2, 39*8, dr42
+	fst.d	r2, 40*8, dr44
+	fst.d	r2, 41*8, dr46
+	fst.d	r2, 42*8, dr48
+	fst.d	r2, 43*8, dr50
+	fst.d	r2, 44*8, dr52
+	fst.d	r2, 45*8, dr54
+	fst.d	r2, 46*8, dr56
+	fst.d	r2, 47*8, dr58
+	fst.d	r2, 48*8, dr60
+	fst.d	r2, 49*8, dr62
+#endif
+	movi	0, r2
+	blink	tr0, r63
+#else
+#if defined (__SH2E__) || defined (__SH3E__) || defined(__SH4_SINGLE__) || defined(__SH4__) || defined(__SH4_SINGLE_ONLY__)
+	add	#(13*4),r4
+#else
+	add	#(9*4),r4
+#endif
+
+	sts.l	pr,@-r4
+
+#if defined (__SH2E__) || defined (__SH3E__) || defined(__SH4_SINGLE__) || defined(__SH4__) || defined(__SH4_SINGLE_ONLY__)
+	fmov.s	fr15,@-r4	! call saved floating point registers
+	fmov.s	fr14,@-r4
+	fmov.s	fr13,@-r4
+	fmov.s	fr12,@-r4
+#endif
+
+	mov.l	r15,@-r4	! call saved integer registers
+	mov.l	r14,@-r4
+	mov.l	r13,@-r4
+	mov.l	r12,@-r4
+
+	mov.l	r11,@-r4
+	mov.l	r10,@-r4
+	mov.l	r9,@-r4
+	mov.l	r8,@-r4
+
+	rts
+	mov    #0,r0
+#endif /* __SH5__ */
+
+ENTRY(longjmp)
+#if __SH5__
+	ld.q	r2,  0*8, r18
+	ptabs	r18, tr0
+	ld.q	r2,  1*8, r10
+	ld.q	r2,  2*8, r11
+	ld.q	r2,  3*8, r12
+	ld.q	r2,  4*8, r13
+	ld.q	r2,  5*8, r14
+	ld.q	r2,  6*8, r15
+	ld.q	r2,  7*8, r28
+	ld.q	r2,  8*8, r29
+	ld.q	r2,  9*8, r30
+	ld.q	r2, 10*8, r31
+	ld.q	r2, 11*8, r32
+	ld.q	r2, 12*8, r33
+	ld.q	r2, 13*8, r34
+	ld.q	r2, 14*8, r35
+	ld.q	r2, 15*8, r44
+	ld.q	r2, 16*8, r45
+	ld.q	r2, 17*8, r46
+	ld.q	r2, 18*8, r47
+	ld.q	r2, 19*8, r48
+	ld.q	r2, 20*8, r49
+	ld.q	r2, 21*8, r50
+	ld.q	r2, 22*8, r51
+	ld.q	r2, 23*8, r52
+	ld.q	r2, 24*8, r53
+	ld.q	r2, 25*8, r54
+	ld.q	r2, 26*8, r55
+	ld.q	r2, 27*8, r56
+	ld.q	r2, 28*8, r57
+	ld.q	r2, 29*8, r58
+	ld.q	r2, 30*8, r59
+	ld.q	r2, 31*8, r5
+	ld.q	r2, 32*8, r6
+	ld.q	r2, 33*8, r7
+	ptabs	r5, tr5
+	ptabs	r6, tr6
+	ptabs	r7, tr7
+#if ! __SH4_NOFPU__
+	fld.d	r2, 34*8, dr12
+	fld.d	r2, 35*8, dr14
+	fld.d	r2, 36*8, dr36
+	fld.d	r2, 37*8, dr38
+	fld.d	r2, 38*8, dr40
+	fld.d	r2, 39*8, dr42
+	fld.d	r2, 40*8, dr44
+	fld.d	r2, 41*8, dr46
+	fld.d	r2, 42*8, dr48
+	fld.d	r2, 43*8, dr50
+	fld.d	r2, 44*8, dr52
+	fld.d	r2, 45*8, dr54
+	fld.d	r2, 46*8, dr56
+	fld.d	r2, 47*8, dr58
+	fld.d	r2, 48*8, dr60
+	fld.d	r2, 49*8, dr62
+#endif
+	movi	1, r2
+	cmvne	r3, r3, r2
+	blink	tr0, r63
+#else
+	mov.l	@r4+,r8
+	mov.l	@r4+,r9
+	mov.l	@r4+,r10
+	mov.l	@r4+,r11
+
+	mov.l	@r4+,r12
+	mov.l	@r4+,r13
+	mov.l	@r4+,r14
+	mov.l	@r4+,r15
+
+#if defined (__SH2E__) || defined (__SH3E__) || defined(__SH4_SINGLE__) || defined(__SH4__) || defined(__SH4_SINGLE_ONLY__)
+	fmov.s	@r4+,fr12	! call saved floating point registers
+	fmov.s	@r4+,fr13
+	fmov.s	@r4+,fr14
+	fmov.s	@r4+,fr15
+#endif
+
+	lds.l	@r4+,pr
+
+	mov	r5,r0
+	tst	r0,r0
+	bf	retr4
+	movt	r0
+retr4:	rts
+	nop
+#endif /* __SH5__ */