diff options
author | Max Filippov <jcmvbkbc@gmail.com> | 2019-05-12 20:28:25 -0700 |
---|---|---|
committer | Max Filippov <jcmvbkbc@gmail.com> | 2019-07-08 10:04:48 -0700 |
commit | d6d5f19e21d98c0607ff029e4e2e508d4cdd1d5a (patch) | |
tree | 1780d36ac99c16f8ed506948a020017e84a89a9f /arch/xtensa | |
parent | 831c4f3da83e260df943dfb982d77cef5cba2c49 (diff) |
xtensa: abstract 'entry' and 'retw' in assembly code
Provide abi_entry, abi_entry_default, abi_ret and abi_ret_default macros
that allocate aligned stack frame in windowed and call0 ABIs.
Provide XTENSA_SPILL_STACK_RESERVE macro that specifies required stack
frame size when register spilling is involved.
Replace all uses of 'entry' and 'retw' with the above macros.
This makes most of the xtensa assembly code ready for XEA3 and call0 ABI.
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
Diffstat (limited to 'arch/xtensa')
-rw-r--r-- | arch/xtensa/include/asm/asmmacro.h | 46 | ||||
-rw-r--r-- | arch/xtensa/kernel/coprocessor.S | 7 | ||||
-rw-r--r-- | arch/xtensa/kernel/entry.S | 11 | ||||
-rw-r--r-- | arch/xtensa/kernel/mcount.S | 11 | ||||
-rw-r--r-- | arch/xtensa/lib/checksum.S | 12 | ||||
-rw-r--r-- | arch/xtensa/lib/memcopy.S | 38 | ||||
-rw-r--r-- | arch/xtensa/lib/memset.S | 10 | ||||
-rw-r--r-- | arch/xtensa/lib/strncpy_user.S | 16 | ||||
-rw-r--r-- | arch/xtensa/lib/strnlen_user.S | 14 | ||||
-rw-r--r-- | arch/xtensa/lib/usercopy.S | 12 | ||||
-rw-r--r-- | arch/xtensa/mm/misc.S | 78 |
11 files changed, 153 insertions, 102 deletions
diff --git a/arch/xtensa/include/asm/asmmacro.h b/arch/xtensa/include/asm/asmmacro.h index 8308a9c3abb2..71a7e846bc1f 100644 --- a/arch/xtensa/include/asm/asmmacro.h +++ b/arch/xtensa/include/asm/asmmacro.h @@ -191,4 +191,50 @@ #endif .endm +#define XTENSA_STACK_ALIGNMENT 16 + +#if defined(__XTENSA_WINDOWED_ABI__) +#define XTENSA_FRAME_SIZE_RESERVE 16 +#define XTENSA_SPILL_STACK_RESERVE 32 + +#define abi_entry(frame_size) \ + entry sp, (XTENSA_FRAME_SIZE_RESERVE + \ + (((frame_size) + XTENSA_STACK_ALIGNMENT - 1) & \ + -XTENSA_STACK_ALIGNMENT)) +#define abi_entry_default abi_entry(0) + +#define abi_ret(frame_size) retw +#define abi_ret_default retw + +#elif defined(__XTENSA_CALL0_ABI__) + +#define XTENSA_SPILL_STACK_RESERVE 0 + +#define abi_entry(frame_size) __abi_entry (frame_size) + + .macro __abi_entry frame_size + .ifgt \frame_size + addi sp, sp, -(((\frame_size) + XTENSA_STACK_ALIGNMENT - 1) & \ + -XTENSA_STACK_ALIGNMENT) + .endif + .endm + +#define abi_entry_default + +#define abi_ret(frame_size) __abi_ret (frame_size) + + .macro __abi_ret frame_size + .ifgt \frame_size + addi sp, sp, (((\frame_size) + XTENSA_STACK_ALIGNMENT - 1) & \ + -XTENSA_STACK_ALIGNMENT) + .endif + ret + .endm + +#define abi_ret_default ret + +#else +#error Unsupported Xtensa ABI +#endif + #endif /* _XTENSA_ASMMACRO_H */ diff --git a/arch/xtensa/kernel/coprocessor.S b/arch/xtensa/kernel/coprocessor.S index 92bf24a9da92..60c220020054 100644 --- a/arch/xtensa/kernel/coprocessor.S +++ b/arch/xtensa/kernel/coprocessor.S @@ -121,7 +121,9 @@ ENTRY(coprocessor_flush) - entry a1, 32 + /* reserve 4 bytes on stack to save a0 */ + abi_entry(4) + s32i a0, a1, 0 movi a0, .Lsave_cp_regs_jump_table addx8 a3, a3, a0 @@ -131,7 +133,8 @@ ENTRY(coprocessor_flush) beqz a3, 1f callx0 a3 1: l32i a0, a1, 0 - retw + + abi_ret(4) ENDPROC(coprocessor_flush) diff --git a/arch/xtensa/kernel/entry.S b/arch/xtensa/kernel/entry.S index e54af8b7e0f8..183fa8e0bb5b 100644 --- a/arch/xtensa/kernel/entry.S +++ b/arch/xtensa/kernel/entry.S @@ -1842,7 +1842,8 @@ ENDPROC(fast_store_prohibited) ENTRY(system_call) - entry a1, 32 + /* reserve 4 bytes on stack for function parameter */ + abi_entry(4) /* regs->syscall = regs->areg[2] */ @@ -1892,7 +1893,7 @@ ENTRY(system_call) s32i a6, a2, PT_AREG2 bnez a3, 1f - retw + abi_ret(4) 1: l32i a4, a1, 4 @@ -1901,7 +1902,7 @@ ENTRY(system_call) mov a6, a2 call4 do_syscall_trace_leave s32i a3, a2, PT_SYSCALL - retw + abi_ret(4) ENDPROC(system_call) @@ -1952,7 +1953,7 @@ ENDPROC(system_call) ENTRY(_switch_to) - entry a1, 48 + abi_entry(XTENSA_SPILL_STACK_RESERVE) mov a11, a3 # and 'next' (a3) @@ -2013,7 +2014,7 @@ ENTRY(_switch_to) wsr a14, ps rsync - retw + abi_ret(XTENSA_SPILL_STACK_RESERVE) ENDPROC(_switch_to) diff --git a/arch/xtensa/kernel/mcount.S b/arch/xtensa/kernel/mcount.S index 0eeda2e4a25e..5e4619f52858 100644 --- a/arch/xtensa/kernel/mcount.S +++ b/arch/xtensa/kernel/mcount.S @@ -11,6 +11,7 @@ */ #include <linux/linkage.h> +#include <asm/asmmacro.h> #include <asm/ftrace.h> /* @@ -21,13 +22,13 @@ ENTRY(_mcount) - entry a1, 16 + abi_entry_default movi a4, ftrace_trace_function l32i a4, a4, 0 movi a3, ftrace_stub bne a3, a4, 1f - retw + abi_ret_default 1: xor a7, a2, a1 movi a3, 0x3fffffff @@ -40,11 +41,11 @@ ENTRY(_mcount) addi a6, a6, -MCOUNT_INSN_SIZE callx4 a4 - retw + abi_ret_default ENDPROC(_mcount) ENTRY(ftrace_stub) - entry a1, 16 - retw + abi_entry_default + abi_ret_default ENDPROC(ftrace_stub) diff --git a/arch/xtensa/lib/checksum.S b/arch/xtensa/lib/checksum.S index c6e73b12e519..4cb9ca58d9ad 100644 --- a/arch/xtensa/lib/checksum.S +++ b/arch/xtensa/lib/checksum.S @@ -43,7 +43,7 @@ ENTRY(csum_partial) * Experiments with Ethernet and SLIP connections show that buf * is aligned on either a 2-byte or 4-byte boundary. */ - entry sp, 32 + abi_entry_default extui a5, a2, 0, 2 bnez a5, 8f /* branch if 2-byte aligned */ /* Fall-through on common case, 4-byte alignment */ @@ -107,7 +107,7 @@ ENTRY(csum_partial) ONES_ADD(a4, a6) 7: mov a2, a4 - retw + abi_ret_default /* uncommon case, buf is 2-byte aligned */ 8: @@ -195,7 +195,7 @@ unsigned int csum_partial_copy_generic (const char *src, char *dst, int len, ENTRY(csum_partial_copy_generic) - entry sp, 32 + abi_entry_default mov a12, a3 mov a11, a4 or a10, a2, a3 @@ -316,7 +316,7 @@ EX(11f) s8i a9, a3, 0 ONES_ADD(a5, a9) 8: mov a2, a5 - retw + abi_ret_default 5: /* Control branch to here when either src or dst is odd. We @@ -383,12 +383,12 @@ ENDPROC(csum_partial_copy_generic) blt a12, a11, .Leloop #endif 2: - retw + abi_ret_default 11: movi a2, -EFAULT s32i a2, a7, 0 /* dst_err_ptr */ movi a2, 0 - retw + abi_ret_default .previous diff --git a/arch/xtensa/lib/memcopy.S b/arch/xtensa/lib/memcopy.S index efecfd7ed8cc..582d817979ed 100644 --- a/arch/xtensa/lib/memcopy.S +++ b/arch/xtensa/lib/memcopy.S @@ -79,7 +79,7 @@ bne a3, a7, .Lnextbyte # continue loop if $a3:src != $a7:src_end #endif /* !XCHAL_HAVE_LOOPS */ .Lbytecopydone: - retw + abi_ret_default /* * Destination is unaligned @@ -112,7 +112,7 @@ ENTRY(__memcpy) WEAK(memcpy) - entry sp, 16 # minimal stack frame + abi_entry_default # a2/ dst, a3/ src, a4/ len mov a5, a2 # copy dst so that a2 is return value .Lcommon: @@ -161,7 +161,7 @@ WEAK(memcpy) bbsi.l a4, 2, .L3 bbsi.l a4, 1, .L4 bbsi.l a4, 0, .L5 - retw + abi_ret_default .L3: # copy 4 bytes l32i a6, a3, 0 @@ -170,7 +170,7 @@ WEAK(memcpy) addi a5, a5, 4 bbsi.l a4, 1, .L4 bbsi.l a4, 0, .L5 - retw + abi_ret_default .L4: # copy 2 bytes l16ui a6, a3, 0 @@ -178,12 +178,12 @@ WEAK(memcpy) s16i a6, a5, 0 addi a5, a5, 2 bbsi.l a4, 0, .L5 - retw + abi_ret_default .L5: # copy 1 byte l8ui a6, a3, 0 s8i a6, a5, 0 - retw + abi_ret_default /* * Destination is aligned, Source is unaligned @@ -255,7 +255,7 @@ WEAK(memcpy) #endif bbsi.l a4, 1, .L14 bbsi.l a4, 0, .L15 -.Ldone: retw +.Ldone: abi_ret_default .L14: # copy 2 bytes l8ui a6, a3, 0 @@ -265,12 +265,12 @@ WEAK(memcpy) s8i a7, a5, 1 addi a5, a5, 2 bbsi.l a4, 0, .L15 - retw + abi_ret_default .L15: # copy 1 byte l8ui a6, a3, 0 s8i a6, a5, 0 - retw + abi_ret_default ENDPROC(__memcpy) @@ -280,7 +280,7 @@ ENDPROC(__memcpy) ENTRY(bcopy) - entry sp, 16 # minimal stack frame + abi_entry_default # a2=src, a3=dst, a4=len mov a5, a3 mov a3, a2 @@ -346,7 +346,7 @@ ENDPROC(bcopy) # $a3:src != $a7:src_start #endif /* !XCHAL_HAVE_LOOPS */ .Lbackbytecopydone: - retw + abi_ret_default /* * Destination is unaligned @@ -380,7 +380,7 @@ ENDPROC(bcopy) ENTRY(__memmove) WEAK(memmove) - entry sp, 16 # minimal stack frame + abi_entry_default # a2/ dst, a3/ src, a4/ len mov a5, a2 # copy dst so that a2 is return value .Lmovecommon: @@ -435,7 +435,7 @@ WEAK(memmove) bbsi.l a4, 2, .Lback3 bbsi.l a4, 1, .Lback4 bbsi.l a4, 0, .Lback5 - retw + abi_ret_default .Lback3: # copy 4 bytes addi a3, a3, -4 @@ -444,7 +444,7 @@ WEAK(memmove) s32i a6, a5, 0 bbsi.l a4, 1, .Lback4 bbsi.l a4, 0, .Lback5 - retw + abi_ret_default .Lback4: # copy 2 bytes addi a3, a3, -2 @@ -452,14 +452,14 @@ WEAK(memmove) addi a5, a5, -2 s16i a6, a5, 0 bbsi.l a4, 0, .Lback5 - retw + abi_ret_default .Lback5: # copy 1 byte addi a3, a3, -1 l8ui a6, a3, 0 addi a5, a5, -1 s8i a6, a5, 0 - retw + abi_ret_default /* * Destination is aligned, Source is unaligned @@ -531,7 +531,7 @@ WEAK(memmove) bbsi.l a4, 1, .Lback14 bbsi.l a4, 0, .Lback15 .Lbackdone: - retw + abi_ret_default .Lback14: # copy 2 bytes addi a3, a3, -2 @@ -541,13 +541,13 @@ WEAK(memmove) s8i a6, a5, 0 s8i a7, a5, 1 bbsi.l a4, 0, .Lback15 - retw + abi_ret_default .Lback15: # copy 1 byte addi a3, a3, -1 addi a5, a5, -1 l8ui a6, a3, 0 s8i a6, a5, 0 - retw + abi_ret_default ENDPROC(__memmove) diff --git a/arch/xtensa/lib/memset.S b/arch/xtensa/lib/memset.S index 8632eacbdc80..59b1524fd601 100644 --- a/arch/xtensa/lib/memset.S +++ b/arch/xtensa/lib/memset.S @@ -34,7 +34,7 @@ ENTRY(__memset) WEAK(memset) - entry sp, 16 # minimal stack frame + abi_entry_default # a2/ dst, a3/ c, a4/ length extui a3, a3, 0, 8 # mask to just 8 bits slli a7, a3, 8 # duplicate character in all bytes of word @@ -48,7 +48,7 @@ WEAK(memset) srli a7, a4, 4 # number of loop iterations with 16B # per iteration bnez a4, .Laligned - retw + abi_ret_default /* * Destination is word-aligned. @@ -95,7 +95,7 @@ EX(10f) s16i a3, a5, 0 EX(10f) s8i a3, a5, 0 .L5: .Lret1: - retw + abi_ret_default /* * Destination is unaligned @@ -139,7 +139,7 @@ EX(10f) s8i a3, a5, 0 blt a5, a6, .Lbyteloop #endif /* !XCHAL_HAVE_LOOPS */ .Lbytesetdone: - retw + abi_ret_default ENDPROC(__memset) @@ -150,4 +150,4 @@ ENDPROC(__memset) 10: movi a2, 0 - retw + abi_ret_default diff --git a/arch/xtensa/lib/strncpy_user.S b/arch/xtensa/lib/strncpy_user.S index c4c6c8578d59..4faf46fe3f38 100644 --- a/arch/xtensa/lib/strncpy_user.S +++ b/arch/xtensa/lib/strncpy_user.S @@ -50,7 +50,7 @@ .text ENTRY(__strncpy_user) - entry sp, 16 # minimal stack frame + abi_entry_default # a2/ dst, a3/ src, a4/ len mov a11, a2 # leave dst in return value register beqz a4, .Lret # if len is zero @@ -93,7 +93,7 @@ EX(10f) s8i a9, a11, 0 # store byte 0 bnez a4, .Lsrcaligned # if len is nonzero .Lret: sub a2, a11, a2 # compute strlen - retw + abi_ret_default /* * dst is word-aligned, src is word-aligned @@ -148,14 +148,14 @@ EX(10f) s8i a9, a11, 0 .Lz3: # byte 3 is zero addi a11, a11, 3 # advance dst pointer sub a2, a11, a2 # compute strlen - retw + abi_ret_default .Lz0: # byte 0 is zero #ifdef __XTENSA_EB__ movi a9, 0 #endif /* __XTENSA_EB__ */ EX(10f) s8i a9, a11, 0 sub a2, a11, a2 # compute strlen - retw + abi_ret_default .Lz1: # byte 1 is zero #ifdef __XTENSA_EB__ extui a9, a9, 16, 16 @@ -163,7 +163,7 @@ EX(10f) s8i a9, a11, 0 EX(10f) s16i a9, a11, 0 addi a11, a11, 1 # advance dst pointer sub a2, a11, a2 # compute strlen - retw + abi_ret_default .Lz2: # byte 2 is zero #ifdef __XTENSA_EB__ extui a9, a9, 16, 16 @@ -173,7 +173,7 @@ EX(10f) s16i a9, a11, 0 EX(10f) s8i a9, a11, 2 addi a11, a11, 2 # advance dst pointer sub a2, a11, a2 # compute strlen - retw + abi_ret_default .align 4 # 1 mod 4 alignment for LOOPNEZ .byte 0 # (0 mod 4 alignment for LBEG) @@ -199,7 +199,7 @@ EX(10f) s8i a9, a11, 0 .Lunalignedend: sub a2, a11, a2 # compute strlen - retw + abi_ret_default ENDPROC(__strncpy_user) @@ -214,4 +214,4 @@ ENDPROC(__strncpy_user) 10: 11: movi a2, -EFAULT - retw + abi_ret_default diff --git a/arch/xtensa/lib/strnlen_user.S b/arch/xtensa/lib/strnlen_user.S index 1f2ca2bb2ab3..3d391dca3efb 100644 --- a/arch/xtensa/lib/strnlen_user.S +++ b/arch/xtensa/lib/strnlen_user.S @@ -45,7 +45,7 @@ .text ENTRY(__strnlen_user) - entry sp, 16 # minimal stack frame + abi_entry_default # a2/ s, a3/ len addi a4, a2, -4 # because we overincrement at the end; # we compensate with load offsets of 4 @@ -96,7 +96,7 @@ EX(10f) l32i a9, a4, 4 # load 4 bytes for remaining checks addi a4, a4, 1 # advance string pointer .L101: sub a2, a4, a2 # compute length - retw + abi_ret_default # NOTE that in several places below, we point to the byte just after # the zero byte in order to include the NULL terminator in the count. @@ -106,15 +106,15 @@ EX(10f) l32i a9, a4, 4 # load 4 bytes for remaining checks .Lz0: # byte 0 is zero addi a4, a4, 1 # point just beyond zero byte sub a2, a4, a2 # subtract to get length - retw + abi_ret_default .Lz1: # byte 1 is zero addi a4, a4, 1+1 # point just beyond zero byte sub a2, a4, a2 # subtract to get length - retw + abi_ret_default .Lz2: # byte 2 is zero addi a4, a4, 2+1 # point just beyond zero byte sub a2, a4, a2 # subtract to get length - retw + abi_ret_default .L1mod2: # address is odd EX(10f) l8ui a9, a4, 4 # get byte 0 @@ -130,7 +130,7 @@ EX(10f) l32i a9, a4, 0 # get word with first two bytes of string # byte 3 is zero addi a4, a4, 3+1 # point just beyond zero byte sub a2, a4, a2 # subtract to get length - retw + abi_ret_default ENDPROC(__strnlen_user) @@ -138,4 +138,4 @@ ENDPROC(__strnlen_user) .align 4 10: movi a2, 0 - retw + abi_ret_default diff --git a/arch/xtensa/lib/usercopy.S b/arch/xtensa/lib/usercopy.S index 228607e30bc2..a0aa4047f94a 100644 --- a/arch/xtensa/lib/usercopy.S +++ b/arch/xtensa/lib/usercopy.S @@ -60,7 +60,7 @@ .text ENTRY(__xtensa_copy_user) - entry sp, 16 # minimal stack frame + abi_entry_default # 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 @@ -75,7 +75,7 @@ ENTRY(__xtensa_copy_user) __ssa8 a3 # set shift amount from byte offset bnez a4, .Lsrcunaligned movi a2, 0 # return success for len==0 - retw + abi_ret_default /* * Destination is unaligned @@ -127,7 +127,7 @@ EX(10f) s8i a6, a5, 0 #endif /* !XCHAL_HAVE_LOOPS */ .Lbytecopydone: movi a2, 0 # return success for len bytes copied - retw + abi_ret_default /* * Destination and source are word-aligned. @@ -187,7 +187,7 @@ EX(10f) l8ui a6, a3, 0 EX(10f) s8i a6, a5, 0 .L5: movi a2, 0 # return success for len bytes copied - retw + abi_ret_default /* * Destination is aligned, Source is unaligned @@ -264,7 +264,7 @@ EX(10f) l8ui a6, a3, 0 EX(10f) s8i a6, a5, 0 .L15: movi a2, 0 # return success for len bytes copied - retw + abi_ret_default ENDPROC(__xtensa_copy_user) @@ -281,4 +281,4 @@ ENDPROC(__xtensa_copy_user) 10: sub a2, a5, a2 /* a2 <-- bytes copied */ sub a2, a11, a2 /* a2 <-- bytes not copied */ - retw + abi_ret_default diff --git a/arch/xtensa/mm/misc.S b/arch/xtensa/mm/misc.S index 11a01c3e9cea..6aa036c427c3 100644 --- a/arch/xtensa/mm/misc.S +++ b/arch/xtensa/mm/misc.S @@ -30,7 +30,7 @@ ENTRY(clear_page) - entry a1, 16 + abi_entry_default movi a3, 0 __loopi a2, a7, PAGE_SIZE, 32 @@ -44,7 +44,7 @@ ENTRY(clear_page) s32i a3, a2, 28 __endla a2, a7, 32 - retw + abi_ret_default ENDPROC(clear_page) @@ -57,7 +57,7 @@ ENDPROC(clear_page) ENTRY(copy_page) - entry a1, 16 + abi_entry_default __loopi a2, a4, PAGE_SIZE, 32 @@ -86,7 +86,7 @@ ENTRY(copy_page) __endl a2, a4 - retw + abi_ret_default ENDPROC(copy_page) @@ -116,7 +116,7 @@ ENTRY(__tlbtemp_mapping_start) ENTRY(clear_page_alias) - entry a1, 32 + abi_entry_default /* Skip setting up a temporary DTLB if not aliased low page. */ @@ -144,14 +144,14 @@ ENTRY(clear_page_alias) __endla a2, a7, 32 bnez a6, 1f - retw + abi_ret_default /* We need to invalidate the temporary idtlb entry, if any. */ 1: idtlb a4 dsync - retw + abi_ret_default ENDPROC(clear_page_alias) @@ -164,7 +164,7 @@ ENDPROC(clear_page_alias) ENTRY(copy_page_alias) - entry a1, 32 + abi_entry_default /* Skip setting up a temporary DTLB for destination if not aliased. */ @@ -221,19 +221,19 @@ ENTRY(copy_page_alias) bnez a6, 1f bnez a7, 2f - retw + abi_ret_default 1: addi a2, a2, -PAGE_SIZE idtlb a2 dsync bnez a7, 2f - retw + abi_ret_default 2: addi a3, a3, -PAGE_SIZE+1 idtlb a3 dsync - retw + abi_ret_default ENDPROC(copy_page_alias) @@ -248,7 +248,7 @@ ENDPROC(copy_page_alias) ENTRY(__flush_invalidate_dcache_page_alias) - entry sp, 16 + abi_entry_default movi a7, 0 # required for exception handler addi a6, a3, (PAGE_KERNEL | _PAGE_HW_WRITE) @@ -261,7 +261,7 @@ ENTRY(__flush_invalidate_dcache_page_alias) idtlb a4 dsync - retw + abi_ret_default ENDPROC(__flush_invalidate_dcache_page_alias) @@ -272,7 +272,7 @@ ENDPROC(__flush_invalidate_dcache_page_alias) ENTRY(__invalidate_dcache_page_alias) - entry sp, 16 + abi_entry_default movi a7, 0 # required for exception handler addi a6, a3, (PAGE_KERNEL | _PAGE_HW_WRITE) @@ -285,7 +285,7 @@ ENTRY(__invalidate_dcache_page_alias) idtlb a4 dsync - retw + abi_ret_default ENDPROC(__invalidate_dcache_page_alias) #endif @@ -296,7 +296,7 @@ ENTRY(__tlbtemp_mapping_itlb) ENTRY(__invalidate_icache_page_alias) - entry sp, 16 + abi_entry_default addi a6, a3, (PAGE_KERNEL_EXEC | _PAGE_HW_WRITE) mov a4, a2 @@ -307,7 +307,7 @@ ENTRY(__invalidate_icache_page_alias) iitlb a4 isync - retw + abi_ret_default ENDPROC(__invalidate_icache_page_alias) @@ -325,12 +325,12 @@ ENTRY(__tlbtemp_mapping_end) ENTRY(__invalidate_icache_page) - entry sp, 16 + abi_entry_default ___invalidate_icache_page a2 a3 isync - retw + abi_ret_default ENDPROC(__invalidate_icache_page) @@ -340,12 +340,12 @@ ENDPROC(__invalidate_icache_page) ENTRY(__invalidate_dcache_page) - entry sp, 16 + abi_entry_default ___invalidate_dcache_page a2 a3 dsync - retw + abi_ret_default ENDPROC(__invalidate_dcache_page) @@ -355,12 +355,12 @@ ENDPROC(__invalidate_dcache_page) ENTRY(__flush_invalidate_dcache_page) - entry sp, 16 + abi_entry_default ___flush_invalidate_dcache_page a2 a3 dsync - retw + abi_ret_default ENDPROC(__flush_invalidate_dcache_page) @@ -370,12 +370,12 @@ ENDPROC(__flush_invalidate_dcache_page) ENTRY(__flush_dcache_page) - entry sp, 16 + abi_entry_default ___flush_dcache_page a2 a3 dsync - retw + abi_ret_default ENDPROC(__flush_dcache_page) @@ -385,12 +385,12 @@ ENDPROC(__flush_dcache_page) ENTRY(__invalidate_icache_range) - entry sp, 16 + abi_entry_default ___invalidate_icache_range a2 a3 a4 isync - retw + abi_ret_default ENDPROC(__invalidate_icache_range) @@ -400,12 +400,12 @@ ENDPROC(__invalidate_icache_range) ENTRY(__flush_invalidate_dcache_range) - entry sp, 16 + abi_entry_default ___flush_invalidate_dcache_range a2 a3 a4 dsync - retw + abi_ret_default ENDPROC(__flush_invalidate_dcache_range) @@ -415,12 +415,12 @@ ENDPROC(__flush_invalidate_dcache_range) ENTRY(__flush_dcache_range) - entry sp, 16 + abi_entry_default ___flush_dcache_range a2 a3 a4 dsync - retw + abi_ret_default ENDPROC(__flush_dcache_range) @@ -430,11 +430,11 @@ ENDPROC(__flush_dcache_range) ENTRY(__invalidate_dcache_range) - entry sp, 16 + abi_entry_default ___invalidate_dcache_range a2 a3 a4 - retw + abi_ret_default ENDPROC(__invalidate_dcache_range) @@ -444,12 +444,12 @@ ENDPROC(__invalidate_dcache_range) ENTRY(__invalidate_icache_all) - entry sp, 16 + abi_entry_default ___invalidate_icache_all a2 a3 isync - retw + abi_ret_default ENDPROC(__invalidate_icache_all) @@ -459,12 +459,12 @@ ENDPROC(__invalidate_icache_all) ENTRY(__flush_invalidate_dcache_all) - entry sp, 16 + abi_entry_default ___flush_invalidate_dcache_all a2 a3 dsync - retw + abi_ret_default ENDPROC(__flush_invalidate_dcache_all) @@ -474,11 +474,11 @@ ENDPROC(__flush_invalidate_dcache_all) ENTRY(__invalidate_dcache_all) - entry sp, 16 + abi_entry_default ___invalidate_dcache_all a2 a3 dsync - retw + abi_ret_default ENDPROC(__invalidate_dcache_all) |