nds32: add support for new architecture

Add support for Andes Technology NDS32 architecture.
See here http://www.andestech.com/en/index/index.htm for more
informaton. Verification of the port from an older uClibc
port was done on a sponsored AG101p board.

The testsuite only has 5 errors, three are related to
an existing bug in dlclose() with LT.old, also happening
on cris32 and m68k.

Failures to fallocate/posix_fallocate are unresolved.

Thanks to Andes Technology sponsoring the hardware and
being very helpful while doing the uClibc-ng porting.

Signed-off-by: Waldemar Brodkorb <wbx@uclibc-ng.org>

1 files changed, 88 insertions, 0 deletions

diff --git a/libc/string/nds32/memcpy.S b/libc/string/nds32/memcpy.S
new file mode 100644
index 000000000..4f285b5ee
--- /dev/null
+++ b/libc/string/nds32/memcpy.S
@@ -0,0 +1,88 @@
+/*
+ * Copyright (C) 2016 Andes Technology, Inc.
+ * Licensed under the LGPL v2.1, see the file COPYING.LIB in this tarball.
+ */
+
+#include <sysdep.h>
+!==========================================================
+!  void *memcpy(void *dst, const void *src, int n);
+!
+!        dst: $r0
+!        src: $r1
+!        n  : $r2
+!        ret: $r0 - pointer to the memory area dst.
+!==========================================================
+.weak memcpy
+ENTRY(memcpy)
+	move    $r5, $r0
+	beq     $r0, $r1, .Lquit_memcpy
+	beqz    $r2, .Lquit_memcpy
+	srli	$r3, $r2, #5			! check if len < cache-line size 32
+	beqz	$r3, .Lword_copy_entry
+	andi	$r4, $r0, #0x3			! check byte-align
+	beqz	$r4, .Lunalign_word_copy_entry
+
+	addi    $r4, $r4, #-4
+	abs		$r4, $r4				! check how many un-align byte to copy
+	sub		$r2, $r2, $r4			! update $R2
+
+.Lunalign_byte_copy:
+	lbi.bi	$r3, [$r1], #1
+	addi	$r4, $r4, #-1
+	sbi.bi	$r3, [$r0], #1
+	bnez	$r4, .Lunalign_byte_copy
+	beqz	$r2, .Lquit_memcpy
+
+.Lunalign_word_copy_entry:
+	andi	$r3, $r0, 0x1f			! check cache-line unaligncount
+	beqz	$r3, .Lcache_copy
+
+	addi	$r3, $r3, #-32
+	abs		$r3, $r3
+	sub		$r2, $r2, $r3			! update $R2
+
+.Lunalign_word_copy:
+	lmw.bim	$r4, [$r1], $r4
+	addi	$r3, $r3, #-4
+	smw.bim	$r4, [$r0], $r4
+	bnez	$r3, .Lunalign_word_copy
+	beqz	$r2, .Lquit_memcpy
+
+	addi	$r3, $r2, #-32			! to check $r2 < cache_line, than go to .Lword_copy
+	bltz	$r3, .Lword_copy_entry
+.Lcache_copy:
+	srli	$r3, $r2, #5
+	beqz	$r3, .Lword_copy_entry
+	pushm	$r6, $r13
+.L3:
+	lmw.bim	$r6, [$r1], $r13
+	addi	$r3, $r3, #-1
+	smw.bim	$r6, [$r0], $r13
+	bnez	$r3, .L3
+	popm	$r6, $r13
+
+.Lword_copy_entry:
+	andi	$r2, $r2, #31
+	beqz	$r2, .Lquit_memcpy
+	srli	$r3, $r2, #2
+	beqz	$r3, .Lbyte_copy
+.Lword_copy:
+	lmw.bim	$r4, [$r1], $r4
+	addi	$r3, $r3, #-1
+	smw.bim	$r4, [$r0], $r4
+	bnez	$r3, .Lword_copy
+	andi	$r2, $r2, #3
+	beqz	$r2, .Lquit_memcpy
+
+.Lbyte_copy:
+	lbi.bi	$r3, [$r1], #1
+	addi	$r2, $r2, #-1
+	sbi.bi	$r3, [$r0], #1
+	bnez	$r2, .Lbyte_copy
+
+.Lquit_memcpy:
+	move	$r0, $r5
+	ret
+
+END(memcpy)
+libc_hidden_def(memcpy)