[PATCH] xtensa: Architecture support for Tensilica Xtensa Part 4
The attached patches provides part 4 of an architecture implementation for the
Tensilica Xtensa CPU series.
Signed-off-by: Chris Zankel <chris@zankel.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
diff --git a/arch/xtensa/lib/usercopy.S b/arch/xtensa/lib/usercopy.S
new file mode 100644
index 0000000..265db26
--- /dev/null
+++ b/arch/xtensa/lib/usercopy.S
@@ -0,0 +1,321 @@
+/*
+ * arch/xtensa/lib/usercopy.S
+ *
+ * Copy to/from user space (derived from arch/xtensa/lib/hal/memcopy.S)
+ *
+ * DO NOT COMBINE this function with <arch/xtensa/lib/hal/memcopy.S>.
+ * It needs to remain separate and distinct. The hal files are part
+ * of the the Xtensa link-time HAL, and those files may differ per
+ * processor configuration. Patching the kernel for another
+ * processor configuration includes replacing the hal files, and we
+ * could loose the special functionality for accessing user-space
+ * memory during such a patch. We sacrifice a little code space here
+ * in favor to simplify code maintenance.
+ *
+ * This file is subject to the terms and conditions of the GNU General
+ * Public License. See the file "COPYING" in the main directory of
+ * this archive for more details.
+ *
+ * Copyright (C) 2002 Tensilica Inc.
+ */
+
+
+/*
+ * size_t __xtensa_copy_user (void *dst, const void *src, size_t len);
+ *
+ * The returned value is the number of bytes not copied. Implies zero
+ * is success.
+ *
+ * The general case algorithm is as follows:
+ * If the destination and source are both aligned,
+ * do 16B chunks with a loop, and then finish up with
+ * 8B, 4B, 2B, and 1B copies conditional on the length.
+ * If destination is aligned and source unaligned,
+ * do the same, but use SRC to align the source data.
+ * If destination is unaligned, align it by conditionally
+ * copying 1B and 2B and then retest.
+ * This code tries to use fall-through braches for the common
+ * case of aligned destinations (except for the branches to
+ * the alignment label).
+ *
+ * Register use:
+ * a0/ return address
+ * a1/ stack pointer
+ * a2/ return value
+ * a3/ src
+ * a4/ length
+ * a5/ dst
+ * a6/ tmp
+ * a7/ tmp
+ * a8/ tmp
+ * a9/ tmp
+ * a10/ tmp
+ * a11/ original length
+ */
+
+#include <xtensa/coreasm.h>
+
+#ifdef __XTENSA_EB__
+#define ALIGN(R, W0, W1) src R, W0, W1
+#define SSA8(R) ssa8b R
+#else
+#define ALIGN(R, W0, W1) src R, W1, W0
+#define SSA8(R) ssa8l R
+#endif
+
+/* Load or store instructions that may cause exceptions use the EX macro. */
+
+#define EX(insn,reg1,reg2,offset,handler) \
+9: insn reg1, reg2, offset; \
+ .section __ex_table, "a"; \
+ .word 9b, handler; \
+ .previous
+
+
+ .text
+ .align 4
+ .global __xtensa_copy_user
+ .type __xtensa_copy_user,@function
+__xtensa_copy_user:
+ entry sp, 16 # minimal stack frame
+ # a2/ dst, a3/ src, a4/ len
+ mov a5, a2 # copy dst so that a2 is return value
+ mov a11, a4 # preserve original len for error case
+.Lcommon:
+ bbsi.l a2, 0, .Ldst1mod2 # if dst is 1 mod 2
+ bbsi.l a2, 1, .Ldst2mod4 # if dst is 2 mod 4
+.Ldstaligned: # return here from .Ldstunaligned when dst is aligned
+ srli a7, a4, 4 # number of loop iterations with 16B
+ # per iteration
+ movi a8, 3 # if source is also aligned,
+ bnone a3, a8, .Laligned # then use word copy
+ SSA8( a3) # set shift amount from byte offset
+ bnez a4, .Lsrcunaligned
+ movi a2, 0 # return success for len==0
+ retw
+
+/*
+ * Destination is unaligned
+ */
+
+.Ldst1mod2: # dst is only byte aligned
+ bltui a4, 7, .Lbytecopy # do short copies byte by byte
+
+ # copy 1 byte
+ EX(l8ui, a6, a3, 0, l_fixup)
+ addi a3, a3, 1
+ EX(s8i, a6, a5, 0, s_fixup)
+ addi a5, a5, 1
+ addi a4, a4, -1
+ bbci.l a5, 1, .Ldstaligned # if dst is now aligned, then
+ # return to main algorithm
+.Ldst2mod4: # dst 16-bit aligned
+ # copy 2 bytes
+ bltui a4, 6, .Lbytecopy # do short copies byte by byte
+ EX(l8ui, a6, a3, 0, l_fixup)
+ EX(l8ui, a7, a3, 1, l_fixup)
+ addi a3, a3, 2
+ EX(s8i, a6, a5, 0, s_fixup)
+ EX(s8i, a7, a5, 1, s_fixup)
+ addi a5, a5, 2
+ addi a4, a4, -2
+ j .Ldstaligned # dst is now aligned, return to main algorithm
+
+/*
+ * Byte by byte copy
+ */
+ .align 4
+ .byte 0 # 1 mod 4 alignment for LOOPNEZ
+ # (0 mod 4 alignment for LBEG)
+.Lbytecopy:
+#if XCHAL_HAVE_LOOPS
+ loopnez a4, .Lbytecopydone
+#else /* !XCHAL_HAVE_LOOPS */
+ beqz a4, .Lbytecopydone
+ add a7, a3, a4 # a7 = end address for source
+#endif /* !XCHAL_HAVE_LOOPS */
+.Lnextbyte:
+ EX(l8ui, a6, a3, 0, l_fixup)
+ addi a3, a3, 1
+ EX(s8i, a6, a5, 0, s_fixup)
+ addi a5, a5, 1
+#if !XCHAL_HAVE_LOOPS
+ blt a3, a7, .Lnextbyte
+#endif /* !XCHAL_HAVE_LOOPS */
+.Lbytecopydone:
+ movi a2, 0 # return success for len bytes copied
+ retw
+
+/*
+ * Destination and source are word-aligned.
+ */
+ # copy 16 bytes per iteration for word-aligned dst and word-aligned src
+ .align 4 # 1 mod 4 alignment for LOOPNEZ
+ .byte 0 # (0 mod 4 alignment for LBEG)
+.Laligned:
+#if XCHAL_HAVE_LOOPS
+ loopnez a7, .Loop1done
+#else /* !XCHAL_HAVE_LOOPS */
+ beqz a7, .Loop1done
+ slli a8, a7, 4
+ add a8, a8, a3 # a8 = end of last 16B source chunk
+#endif /* !XCHAL_HAVE_LOOPS */
+.Loop1:
+ EX(l32i, a6, a3, 0, l_fixup)
+ EX(l32i, a7, a3, 4, l_fixup)
+ EX(s32i, a6, a5, 0, s_fixup)
+ EX(l32i, a6, a3, 8, l_fixup)
+ EX(s32i, a7, a5, 4, s_fixup)
+ EX(l32i, a7, a3, 12, l_fixup)
+ EX(s32i, a6, a5, 8, s_fixup)
+ addi a3, a3, 16
+ EX(s32i, a7, a5, 12, s_fixup)
+ addi a5, a5, 16
+#if !XCHAL_HAVE_LOOPS
+ blt a3, a8, .Loop1
+#endif /* !XCHAL_HAVE_LOOPS */
+.Loop1done:
+ bbci.l a4, 3, .L2
+ # copy 8 bytes
+ EX(l32i, a6, a3, 0, l_fixup)
+ EX(l32i, a7, a3, 4, l_fixup)
+ addi a3, a3, 8
+ EX(s32i, a6, a5, 0, s_fixup)
+ EX(s32i, a7, a5, 4, s_fixup)
+ addi a5, a5, 8
+.L2:
+ bbci.l a4, 2, .L3
+ # copy 4 bytes
+ EX(l32i, a6, a3, 0, l_fixup)
+ addi a3, a3, 4
+ EX(s32i, a6, a5, 0, s_fixup)
+ addi a5, a5, 4
+.L3:
+ bbci.l a4, 1, .L4
+ # copy 2 bytes
+ EX(l16ui, a6, a3, 0, l_fixup)
+ addi a3, a3, 2
+ EX(s16i, a6, a5, 0, s_fixup)
+ addi a5, a5, 2
+.L4:
+ bbci.l a4, 0, .L5
+ # copy 1 byte
+ EX(l8ui, a6, a3, 0, l_fixup)
+ EX(s8i, a6, a5, 0, s_fixup)
+.L5:
+ movi a2, 0 # return success for len bytes copied
+ retw
+
+/*
+ * Destination is aligned, Source is unaligned
+ */
+
+ .align 4
+ .byte 0 # 1 mod 4 alignement for LOOPNEZ
+ # (0 mod 4 alignment for LBEG)
+.Lsrcunaligned:
+ # copy 16 bytes per iteration for word-aligned dst and unaligned src
+ and a10, a3, a8 # save unalignment offset for below
+ sub a3, a3, a10 # align a3 (to avoid sim warnings only; not needed for hardware)
+ EX(l32i, a6, a3, 0, l_fixup) # load first word
+#if XCHAL_HAVE_LOOPS
+ loopnez a7, .Loop2done
+#else /* !XCHAL_HAVE_LOOPS */
+ beqz a7, .Loop2done
+ slli a10, a7, 4
+ add a10, a10, a3 # a10 = end of last 16B source chunk
+#endif /* !XCHAL_HAVE_LOOPS */
+.Loop2:
+ EX(l32i, a7, a3, 4, l_fixup)
+ EX(l32i, a8, a3, 8, l_fixup)
+ ALIGN( a6, a6, a7)
+ EX(s32i, a6, a5, 0, s_fixup)
+ EX(l32i, a9, a3, 12, l_fixup)
+ ALIGN( a7, a7, a8)
+ EX(s32i, a7, a5, 4, s_fixup)
+ EX(l32i, a6, a3, 16, l_fixup)
+ ALIGN( a8, a8, a9)
+ EX(s32i, a8, a5, 8, s_fixup)
+ addi a3, a3, 16
+ ALIGN( a9, a9, a6)
+ EX(s32i, a9, a5, 12, s_fixup)
+ addi a5, a5, 16
+#if !XCHAL_HAVE_LOOPS
+ blt a3, a10, .Loop2
+#endif /* !XCHAL_HAVE_LOOPS */
+.Loop2done:
+ bbci.l a4, 3, .L12
+ # copy 8 bytes
+ EX(l32i, a7, a3, 4, l_fixup)
+ EX(l32i, a8, a3, 8, l_fixup)
+ ALIGN( a6, a6, a7)
+ EX(s32i, a6, a5, 0, s_fixup)
+ addi a3, a3, 8
+ ALIGN( a7, a7, a8)
+ EX(s32i, a7, a5, 4, s_fixup)
+ addi a5, a5, 8
+ mov a6, a8
+.L12:
+ bbci.l a4, 2, .L13
+ # copy 4 bytes
+ EX(l32i, a7, a3, 4, l_fixup)
+ addi a3, a3, 4
+ ALIGN( a6, a6, a7)
+ EX(s32i, a6, a5, 0, s_fixup)
+ addi a5, a5, 4
+ mov a6, a7
+.L13:
+ add a3, a3, a10 # readjust a3 with correct misalignment
+ bbci.l a4, 1, .L14
+ # copy 2 bytes
+ EX(l8ui, a6, a3, 0, l_fixup)
+ EX(l8ui, a7, a3, 1, l_fixup)
+ addi a3, a3, 2
+ EX(s8i, a6, a5, 0, s_fixup)
+ EX(s8i, a7, a5, 1, s_fixup)
+ addi a5, a5, 2
+.L14:
+ bbci.l a4, 0, .L15
+ # copy 1 byte
+ EX(l8ui, a6, a3, 0, l_fixup)
+ EX(s8i, a6, a5, 0, s_fixup)
+.L15:
+ movi a2, 0 # return success for len bytes copied
+ retw
+
+
+ .section .fixup, "ax"
+ .align 4
+
+/* a2 = original dst; a5 = current dst; a11= original len
+ * bytes_copied = a5 - a2
+ * retval = bytes_not_copied = original len - bytes_copied
+ * retval = a11 - (a5 - a2)
+ *
+ * Clearing the remaining pieces of kernel memory plugs security
+ * holes. This functionality is the equivalent of the *_zeroing
+ * functions that some architectures provide.
+ */
+
+.Lmemset:
+ .word memset
+
+s_fixup:
+ sub a2, a5, a2 /* a2 <-- bytes copied */
+ sub a2, a11, a2 /* a2 <-- bytes not copied */
+ retw
+
+l_fixup:
+ sub a2, a5, a2 /* a2 <-- bytes copied */
+ sub a2, a11, a2 /* a2 <-- bytes not copied == return value */
+
+ /* void *memset(void *s, int c, size_t n); */
+ mov a6, a5 /* s */
+ movi a7, 0 /* c */
+ mov a8, a2 /* n */
+ l32r a4, .Lmemset
+ callx4 a4
+ /* Ignore memset return value in a6. */
+ /* a2 still contains bytes not copied. */
+ retw
+