diff options
author | Ralf Baechle <ralf@linux-mips.org> | 2008-01-29 05:14:59 -0500 |
---|---|---|
committer | Ralf Baechle <ralf@linux-mips.org> | 2008-01-29 05:14:59 -0500 |
commit | c5ec1983e45d25446a023e98207e30ab1bf2311a (patch) | |
tree | 68124f5166dc284d7d1c6280f4642a6719a6c554 /arch/mips/lib/csum_partial.S | |
parent | f21d8508080f05a8b05c99fce57cd76ac43666fa (diff) |
[MIPS] Eleminate local symbols from the symbol table.
These symbols appear in oprofile output, stacktraces and similar but only
make the output harder to read. Many identical symbol names such as
"both_aligned" were also being used in multiple source files making it
impossible to see which file actually was meant. So let's get rid of them.
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Diffstat (limited to 'arch/mips/lib/csum_partial.S')
-rw-r--r-- | arch/mips/lib/csum_partial.S | 214 |
1 files changed, 107 insertions, 107 deletions
diff --git a/arch/mips/lib/csum_partial.S b/arch/mips/lib/csum_partial.S index 957a82484e3e..8d7784122c14 100644 --- a/arch/mips/lib/csum_partial.S +++ b/arch/mips/lib/csum_partial.S | |||
@@ -96,13 +96,13 @@ LEAF(csum_partial) | |||
96 | move t7, zero | 96 | move t7, zero |
97 | 97 | ||
98 | sltiu t8, a1, 0x8 | 98 | sltiu t8, a1, 0x8 |
99 | bnez t8, small_csumcpy /* < 8 bytes to copy */ | 99 | bnez t8, .Lsmall_csumcpy /* < 8 bytes to copy */ |
100 | move t2, a1 | 100 | move t2, a1 |
101 | 101 | ||
102 | andi t7, src, 0x1 /* odd buffer? */ | 102 | andi t7, src, 0x1 /* odd buffer? */ |
103 | 103 | ||
104 | hword_align: | 104 | .Lhword_align: |
105 | beqz t7, word_align | 105 | beqz t7, .Lword_align |
106 | andi t8, src, 0x2 | 106 | andi t8, src, 0x2 |
107 | 107 | ||
108 | lbu t0, (src) | 108 | lbu t0, (src) |
@@ -114,8 +114,8 @@ hword_align: | |||
114 | PTR_ADDU src, src, 0x1 | 114 | PTR_ADDU src, src, 0x1 |
115 | andi t8, src, 0x2 | 115 | andi t8, src, 0x2 |
116 | 116 | ||
117 | word_align: | 117 | .Lword_align: |
118 | beqz t8, dword_align | 118 | beqz t8, .Ldword_align |
119 | sltiu t8, a1, 56 | 119 | sltiu t8, a1, 56 |
120 | 120 | ||
121 | lhu t0, (src) | 121 | lhu t0, (src) |
@@ -124,12 +124,12 @@ word_align: | |||
124 | sltiu t8, a1, 56 | 124 | sltiu t8, a1, 56 |
125 | PTR_ADDU src, src, 0x2 | 125 | PTR_ADDU src, src, 0x2 |
126 | 126 | ||
127 | dword_align: | 127 | .Ldword_align: |
128 | bnez t8, do_end_words | 128 | bnez t8, .Ldo_end_words |
129 | move t8, a1 | 129 | move t8, a1 |
130 | 130 | ||
131 | andi t8, src, 0x4 | 131 | andi t8, src, 0x4 |
132 | beqz t8, qword_align | 132 | beqz t8, .Lqword_align |
133 | andi t8, src, 0x8 | 133 | andi t8, src, 0x8 |
134 | 134 | ||
135 | lw t0, 0x00(src) | 135 | lw t0, 0x00(src) |
@@ -138,8 +138,8 @@ dword_align: | |||
138 | PTR_ADDU src, src, 0x4 | 138 | PTR_ADDU src, src, 0x4 |
139 | andi t8, src, 0x8 | 139 | andi t8, src, 0x8 |
140 | 140 | ||
141 | qword_align: | 141 | .Lqword_align: |
142 | beqz t8, oword_align | 142 | beqz t8, .Loword_align |
143 | andi t8, src, 0x10 | 143 | andi t8, src, 0x10 |
144 | 144 | ||
145 | #ifdef USE_DOUBLE | 145 | #ifdef USE_DOUBLE |
@@ -156,8 +156,8 @@ qword_align: | |||
156 | PTR_ADDU src, src, 0x8 | 156 | PTR_ADDU src, src, 0x8 |
157 | andi t8, src, 0x10 | 157 | andi t8, src, 0x10 |
158 | 158 | ||
159 | oword_align: | 159 | .Loword_align: |
160 | beqz t8, begin_movement | 160 | beqz t8, .Lbegin_movement |
161 | LONG_SRL t8, a1, 0x7 | 161 | LONG_SRL t8, a1, 0x7 |
162 | 162 | ||
163 | #ifdef USE_DOUBLE | 163 | #ifdef USE_DOUBLE |
@@ -172,11 +172,11 @@ oword_align: | |||
172 | PTR_ADDU src, src, 0x10 | 172 | PTR_ADDU src, src, 0x10 |
173 | LONG_SRL t8, a1, 0x7 | 173 | LONG_SRL t8, a1, 0x7 |
174 | 174 | ||
175 | begin_movement: | 175 | .Lbegin_movement: |
176 | beqz t8, 1f | 176 | beqz t8, 1f |
177 | andi t2, a1, 0x40 | 177 | andi t2, a1, 0x40 |
178 | 178 | ||
179 | move_128bytes: | 179 | .Lmove_128bytes: |
180 | CSUM_BIGCHUNK(src, 0x00, sum, t0, t1, t3, t4) | 180 | CSUM_BIGCHUNK(src, 0x00, sum, t0, t1, t3, t4) |
181 | CSUM_BIGCHUNK(src, 0x20, sum, t0, t1, t3, t4) | 181 | CSUM_BIGCHUNK(src, 0x20, sum, t0, t1, t3, t4) |
182 | CSUM_BIGCHUNK(src, 0x40, sum, t0, t1, t3, t4) | 182 | CSUM_BIGCHUNK(src, 0x40, sum, t0, t1, t3, t4) |
@@ -184,43 +184,43 @@ move_128bytes: | |||
184 | LONG_SUBU t8, t8, 0x01 | 184 | LONG_SUBU t8, t8, 0x01 |
185 | .set reorder /* DADDI_WAR */ | 185 | .set reorder /* DADDI_WAR */ |
186 | PTR_ADDU src, src, 0x80 | 186 | PTR_ADDU src, src, 0x80 |
187 | bnez t8, move_128bytes | 187 | bnez t8, .Lmove_128bytes |
188 | .set noreorder | 188 | .set noreorder |
189 | 189 | ||
190 | 1: | 190 | 1: |
191 | beqz t2, 1f | 191 | beqz t2, 1f |
192 | andi t2, a1, 0x20 | 192 | andi t2, a1, 0x20 |
193 | 193 | ||
194 | move_64bytes: | 194 | .Lmove_64bytes: |
195 | CSUM_BIGCHUNK(src, 0x00, sum, t0, t1, t3, t4) | 195 | CSUM_BIGCHUNK(src, 0x00, sum, t0, t1, t3, t4) |
196 | CSUM_BIGCHUNK(src, 0x20, sum, t0, t1, t3, t4) | 196 | CSUM_BIGCHUNK(src, 0x20, sum, t0, t1, t3, t4) |
197 | PTR_ADDU src, src, 0x40 | 197 | PTR_ADDU src, src, 0x40 |
198 | 198 | ||
199 | 1: | 199 | 1: |
200 | beqz t2, do_end_words | 200 | beqz t2, .Ldo_end_words |
201 | andi t8, a1, 0x1c | 201 | andi t8, a1, 0x1c |
202 | 202 | ||
203 | move_32bytes: | 203 | .Lmove_32bytes: |
204 | CSUM_BIGCHUNK(src, 0x00, sum, t0, t1, t3, t4) | 204 | CSUM_BIGCHUNK(src, 0x00, sum, t0, t1, t3, t4) |
205 | andi t8, a1, 0x1c | 205 | andi t8, a1, 0x1c |
206 | PTR_ADDU src, src, 0x20 | 206 | PTR_ADDU src, src, 0x20 |
207 | 207 | ||
208 | do_end_words: | 208 | .Ldo_end_words: |
209 | beqz t8, small_csumcpy | 209 | beqz t8, .Lsmall_csumcpy |
210 | andi t2, a1, 0x3 | 210 | andi t2, a1, 0x3 |
211 | LONG_SRL t8, t8, 0x2 | 211 | LONG_SRL t8, t8, 0x2 |
212 | 212 | ||
213 | end_words: | 213 | .Lend_words: |
214 | lw t0, (src) | 214 | lw t0, (src) |
215 | LONG_SUBU t8, t8, 0x1 | 215 | LONG_SUBU t8, t8, 0x1 |
216 | ADDC(sum, t0) | 216 | ADDC(sum, t0) |
217 | .set reorder /* DADDI_WAR */ | 217 | .set reorder /* DADDI_WAR */ |
218 | PTR_ADDU src, src, 0x4 | 218 | PTR_ADDU src, src, 0x4 |
219 | bnez t8, end_words | 219 | bnez t8, .Lend_words |
220 | .set noreorder | 220 | .set noreorder |
221 | 221 | ||
222 | /* unknown src alignment and < 8 bytes to go */ | 222 | /* unknown src alignment and < 8 bytes to go */ |
223 | small_csumcpy: | 223 | .Lsmall_csumcpy: |
224 | move a1, t2 | 224 | move a1, t2 |
225 | 225 | ||
226 | andi t0, a1, 4 | 226 | andi t0, a1, 4 |
@@ -413,48 +413,48 @@ FEXPORT(csum_partial_copy_nocheck) | |||
413 | */ | 413 | */ |
414 | sltu t2, len, NBYTES | 414 | sltu t2, len, NBYTES |
415 | and t1, dst, ADDRMASK | 415 | and t1, dst, ADDRMASK |
416 | bnez t2, copy_bytes_checklen | 416 | bnez t2, .Lcopy_bytes_checklen |
417 | and t0, src, ADDRMASK | 417 | and t0, src, ADDRMASK |
418 | andi odd, dst, 0x1 /* odd buffer? */ | 418 | andi odd, dst, 0x1 /* odd buffer? */ |
419 | bnez t1, dst_unaligned | 419 | bnez t1, .Ldst_unaligned |
420 | nop | 420 | nop |
421 | bnez t0, src_unaligned_dst_aligned | 421 | bnez t0, .Lsrc_unaligned_dst_aligned |
422 | /* | 422 | /* |
423 | * use delay slot for fall-through | 423 | * use delay slot for fall-through |
424 | * src and dst are aligned; need to compute rem | 424 | * src and dst are aligned; need to compute rem |
425 | */ | 425 | */ |
426 | both_aligned: | 426 | .Lboth_aligned: |
427 | SRL t0, len, LOG_NBYTES+3 # +3 for 8 units/iter | 427 | SRL t0, len, LOG_NBYTES+3 # +3 for 8 units/iter |
428 | beqz t0, cleanup_both_aligned # len < 8*NBYTES | 428 | beqz t0, .Lcleanup_both_aligned # len < 8*NBYTES |
429 | nop | 429 | nop |
430 | SUB len, 8*NBYTES # subtract here for bgez loop | 430 | SUB len, 8*NBYTES # subtract here for bgez loop |
431 | .align 4 | 431 | .align 4 |
432 | 1: | 432 | 1: |
433 | EXC( LOAD t0, UNIT(0)(src), l_exc) | 433 | EXC( LOAD t0, UNIT(0)(src), .Ll_exc) |
434 | EXC( LOAD t1, UNIT(1)(src), l_exc_copy) | 434 | EXC( LOAD t1, UNIT(1)(src), .Ll_exc_copy) |
435 | EXC( LOAD t2, UNIT(2)(src), l_exc_copy) | 435 | EXC( LOAD t2, UNIT(2)(src), .Ll_exc_copy) |
436 | EXC( LOAD t3, UNIT(3)(src), l_exc_copy) | 436 | EXC( LOAD t3, UNIT(3)(src), .Ll_exc_copy) |
437 | EXC( LOAD t4, UNIT(4)(src), l_exc_copy) | 437 | EXC( LOAD t4, UNIT(4)(src), .Ll_exc_copy) |
438 | EXC( LOAD t5, UNIT(5)(src), l_exc_copy) | 438 | EXC( LOAD t5, UNIT(5)(src), .Ll_exc_copy) |
439 | EXC( LOAD t6, UNIT(6)(src), l_exc_copy) | 439 | EXC( LOAD t6, UNIT(6)(src), .Ll_exc_copy) |
440 | EXC( LOAD t7, UNIT(7)(src), l_exc_copy) | 440 | EXC( LOAD t7, UNIT(7)(src), .Ll_exc_copy) |
441 | SUB len, len, 8*NBYTES | 441 | SUB len, len, 8*NBYTES |
442 | ADD src, src, 8*NBYTES | 442 | ADD src, src, 8*NBYTES |
443 | EXC( STORE t0, UNIT(0)(dst), s_exc) | 443 | EXC( STORE t0, UNIT(0)(dst), .Ls_exc) |
444 | ADDC(sum, t0) | 444 | ADDC(sum, t0) |
445 | EXC( STORE t1, UNIT(1)(dst), s_exc) | 445 | EXC( STORE t1, UNIT(1)(dst), .Ls_exc) |
446 | ADDC(sum, t1) | 446 | ADDC(sum, t1) |
447 | EXC( STORE t2, UNIT(2)(dst), s_exc) | 447 | EXC( STORE t2, UNIT(2)(dst), .Ls_exc) |
448 | ADDC(sum, t2) | 448 | ADDC(sum, t2) |
449 | EXC( STORE t3, UNIT(3)(dst), s_exc) | 449 | EXC( STORE t3, UNIT(3)(dst), .Ls_exc) |
450 | ADDC(sum, t3) | 450 | ADDC(sum, t3) |
451 | EXC( STORE t4, UNIT(4)(dst), s_exc) | 451 | EXC( STORE t4, UNIT(4)(dst), .Ls_exc) |
452 | ADDC(sum, t4) | 452 | ADDC(sum, t4) |
453 | EXC( STORE t5, UNIT(5)(dst), s_exc) | 453 | EXC( STORE t5, UNIT(5)(dst), .Ls_exc) |
454 | ADDC(sum, t5) | 454 | ADDC(sum, t5) |
455 | EXC( STORE t6, UNIT(6)(dst), s_exc) | 455 | EXC( STORE t6, UNIT(6)(dst), .Ls_exc) |
456 | ADDC(sum, t6) | 456 | ADDC(sum, t6) |
457 | EXC( STORE t7, UNIT(7)(dst), s_exc) | 457 | EXC( STORE t7, UNIT(7)(dst), .Ls_exc) |
458 | ADDC(sum, t7) | 458 | ADDC(sum, t7) |
459 | .set reorder /* DADDI_WAR */ | 459 | .set reorder /* DADDI_WAR */ |
460 | ADD dst, dst, 8*NBYTES | 460 | ADD dst, dst, 8*NBYTES |
@@ -465,44 +465,44 @@ EXC( STORE t7, UNIT(7)(dst), s_exc) | |||
465 | /* | 465 | /* |
466 | * len == the number of bytes left to copy < 8*NBYTES | 466 | * len == the number of bytes left to copy < 8*NBYTES |
467 | */ | 467 | */ |
468 | cleanup_both_aligned: | 468 | .Lcleanup_both_aligned: |
469 | #define rem t7 | 469 | #define rem t7 |
470 | beqz len, done | 470 | beqz len, .Ldone |
471 | sltu t0, len, 4*NBYTES | 471 | sltu t0, len, 4*NBYTES |
472 | bnez t0, less_than_4units | 472 | bnez t0, .Lless_than_4units |
473 | and rem, len, (NBYTES-1) # rem = len % NBYTES | 473 | and rem, len, (NBYTES-1) # rem = len % NBYTES |
474 | /* | 474 | /* |
475 | * len >= 4*NBYTES | 475 | * len >= 4*NBYTES |
476 | */ | 476 | */ |
477 | EXC( LOAD t0, UNIT(0)(src), l_exc) | 477 | EXC( LOAD t0, UNIT(0)(src), .Ll_exc) |
478 | EXC( LOAD t1, UNIT(1)(src), l_exc_copy) | 478 | EXC( LOAD t1, UNIT(1)(src), .Ll_exc_copy) |
479 | EXC( LOAD t2, UNIT(2)(src), l_exc_copy) | 479 | EXC( LOAD t2, UNIT(2)(src), .Ll_exc_copy) |
480 | EXC( LOAD t3, UNIT(3)(src), l_exc_copy) | 480 | EXC( LOAD t3, UNIT(3)(src), .Ll_exc_copy) |
481 | SUB len, len, 4*NBYTES | 481 | SUB len, len, 4*NBYTES |
482 | ADD src, src, 4*NBYTES | 482 | ADD src, src, 4*NBYTES |
483 | EXC( STORE t0, UNIT(0)(dst), s_exc) | 483 | EXC( STORE t0, UNIT(0)(dst), .Ls_exc) |
484 | ADDC(sum, t0) | 484 | ADDC(sum, t0) |
485 | EXC( STORE t1, UNIT(1)(dst), s_exc) | 485 | EXC( STORE t1, UNIT(1)(dst), .Ls_exc) |
486 | ADDC(sum, t1) | 486 | ADDC(sum, t1) |
487 | EXC( STORE t2, UNIT(2)(dst), s_exc) | 487 | EXC( STORE t2, UNIT(2)(dst), .Ls_exc) |
488 | ADDC(sum, t2) | 488 | ADDC(sum, t2) |
489 | EXC( STORE t3, UNIT(3)(dst), s_exc) | 489 | EXC( STORE t3, UNIT(3)(dst), .Ls_exc) |
490 | ADDC(sum, t3) | 490 | ADDC(sum, t3) |
491 | .set reorder /* DADDI_WAR */ | 491 | .set reorder /* DADDI_WAR */ |
492 | ADD dst, dst, 4*NBYTES | 492 | ADD dst, dst, 4*NBYTES |
493 | beqz len, done | 493 | beqz len, .Ldone |
494 | .set noreorder | 494 | .set noreorder |
495 | less_than_4units: | 495 | .Lless_than_4units: |
496 | /* | 496 | /* |
497 | * rem = len % NBYTES | 497 | * rem = len % NBYTES |
498 | */ | 498 | */ |
499 | beq rem, len, copy_bytes | 499 | beq rem, len, .Lcopy_bytes |
500 | nop | 500 | nop |
501 | 1: | 501 | 1: |
502 | EXC( LOAD t0, 0(src), l_exc) | 502 | EXC( LOAD t0, 0(src), .Ll_exc) |
503 | ADD src, src, NBYTES | 503 | ADD src, src, NBYTES |
504 | SUB len, len, NBYTES | 504 | SUB len, len, NBYTES |
505 | EXC( STORE t0, 0(dst), s_exc) | 505 | EXC( STORE t0, 0(dst), .Ls_exc) |
506 | ADDC(sum, t0) | 506 | ADDC(sum, t0) |
507 | .set reorder /* DADDI_WAR */ | 507 | .set reorder /* DADDI_WAR */ |
508 | ADD dst, dst, NBYTES | 508 | ADD dst, dst, NBYTES |
@@ -521,20 +521,20 @@ EXC( STORE t0, 0(dst), s_exc) | |||
521 | * more instruction-level parallelism. | 521 | * more instruction-level parallelism. |
522 | */ | 522 | */ |
523 | #define bits t2 | 523 | #define bits t2 |
524 | beqz len, done | 524 | beqz len, .Ldone |
525 | ADD t1, dst, len # t1 is just past last byte of dst | 525 | ADD t1, dst, len # t1 is just past last byte of dst |
526 | li bits, 8*NBYTES | 526 | li bits, 8*NBYTES |
527 | SLL rem, len, 3 # rem = number of bits to keep | 527 | SLL rem, len, 3 # rem = number of bits to keep |
528 | EXC( LOAD t0, 0(src), l_exc) | 528 | EXC( LOAD t0, 0(src), .Ll_exc) |
529 | SUB bits, bits, rem # bits = number of bits to discard | 529 | SUB bits, bits, rem # bits = number of bits to discard |
530 | SHIFT_DISCARD t0, t0, bits | 530 | SHIFT_DISCARD t0, t0, bits |
531 | EXC( STREST t0, -1(t1), s_exc) | 531 | EXC( STREST t0, -1(t1), .Ls_exc) |
532 | SHIFT_DISCARD_REVERT t0, t0, bits | 532 | SHIFT_DISCARD_REVERT t0, t0, bits |
533 | .set reorder | 533 | .set reorder |
534 | ADDC(sum, t0) | 534 | ADDC(sum, t0) |
535 | b done | 535 | b .Ldone |
536 | .set noreorder | 536 | .set noreorder |
537 | dst_unaligned: | 537 | .Ldst_unaligned: |
538 | /* | 538 | /* |
539 | * dst is unaligned | 539 | * dst is unaligned |
540 | * t0 = src & ADDRMASK | 540 | * t0 = src & ADDRMASK |
@@ -545,25 +545,25 @@ dst_unaligned: | |||
545 | * Set match = (src and dst have same alignment) | 545 | * Set match = (src and dst have same alignment) |
546 | */ | 546 | */ |
547 | #define match rem | 547 | #define match rem |
548 | EXC( LDFIRST t3, FIRST(0)(src), l_exc) | 548 | EXC( LDFIRST t3, FIRST(0)(src), .Ll_exc) |
549 | ADD t2, zero, NBYTES | 549 | ADD t2, zero, NBYTES |
550 | EXC( LDREST t3, REST(0)(src), l_exc_copy) | 550 | EXC( LDREST t3, REST(0)(src), .Ll_exc_copy) |
551 | SUB t2, t2, t1 # t2 = number of bytes copied | 551 | SUB t2, t2, t1 # t2 = number of bytes copied |
552 | xor match, t0, t1 | 552 | xor match, t0, t1 |
553 | EXC( STFIRST t3, FIRST(0)(dst), s_exc) | 553 | EXC( STFIRST t3, FIRST(0)(dst), .Ls_exc) |
554 | SLL t4, t1, 3 # t4 = number of bits to discard | 554 | SLL t4, t1, 3 # t4 = number of bits to discard |
555 | SHIFT_DISCARD t3, t3, t4 | 555 | SHIFT_DISCARD t3, t3, t4 |
556 | /* no SHIFT_DISCARD_REVERT to handle odd buffer properly */ | 556 | /* no SHIFT_DISCARD_REVERT to handle odd buffer properly */ |
557 | ADDC(sum, t3) | 557 | ADDC(sum, t3) |
558 | beq len, t2, done | 558 | beq len, t2, .Ldone |
559 | SUB len, len, t2 | 559 | SUB len, len, t2 |
560 | ADD dst, dst, t2 | 560 | ADD dst, dst, t2 |
561 | beqz match, both_aligned | 561 | beqz match, .Lboth_aligned |
562 | ADD src, src, t2 | 562 | ADD src, src, t2 |
563 | 563 | ||
564 | src_unaligned_dst_aligned: | 564 | .Lsrc_unaligned_dst_aligned: |
565 | SRL t0, len, LOG_NBYTES+2 # +2 for 4 units/iter | 565 | SRL t0, len, LOG_NBYTES+2 # +2 for 4 units/iter |
566 | beqz t0, cleanup_src_unaligned | 566 | beqz t0, .Lcleanup_src_unaligned |
567 | and rem, len, (4*NBYTES-1) # rem = len % 4*NBYTES | 567 | and rem, len, (4*NBYTES-1) # rem = len % 4*NBYTES |
568 | 1: | 568 | 1: |
569 | /* | 569 | /* |
@@ -572,53 +572,53 @@ src_unaligned_dst_aligned: | |||
572 | * It's OK to load FIRST(N+1) before REST(N) because the two addresses | 572 | * It's OK to load FIRST(N+1) before REST(N) because the two addresses |
573 | * are to the same unit (unless src is aligned, but it's not). | 573 | * are to the same unit (unless src is aligned, but it's not). |
574 | */ | 574 | */ |
575 | EXC( LDFIRST t0, FIRST(0)(src), l_exc) | 575 | EXC( LDFIRST t0, FIRST(0)(src), .Ll_exc) |
576 | EXC( LDFIRST t1, FIRST(1)(src), l_exc_copy) | 576 | EXC( LDFIRST t1, FIRST(1)(src), .Ll_exc_copy) |
577 | SUB len, len, 4*NBYTES | 577 | SUB len, len, 4*NBYTES |
578 | EXC( LDREST t0, REST(0)(src), l_exc_copy) | 578 | EXC( LDREST t0, REST(0)(src), .Ll_exc_copy) |
579 | EXC( LDREST t1, REST(1)(src), l_exc_copy) | 579 | EXC( LDREST t1, REST(1)(src), .Ll_exc_copy) |
580 | EXC( LDFIRST t2, FIRST(2)(src), l_exc_copy) | 580 | EXC( LDFIRST t2, FIRST(2)(src), .Ll_exc_copy) |
581 | EXC( LDFIRST t3, FIRST(3)(src), l_exc_copy) | 581 | EXC( LDFIRST t3, FIRST(3)(src), .Ll_exc_copy) |
582 | EXC( LDREST t2, REST(2)(src), l_exc_copy) | 582 | EXC( LDREST t2, REST(2)(src), .Ll_exc_copy) |
583 | EXC( LDREST t3, REST(3)(src), l_exc_copy) | 583 | EXC( LDREST t3, REST(3)(src), .Ll_exc_copy) |
584 | ADD src, src, 4*NBYTES | 584 | ADD src, src, 4*NBYTES |
585 | #ifdef CONFIG_CPU_SB1 | 585 | #ifdef CONFIG_CPU_SB1 |
586 | nop # improves slotting | 586 | nop # improves slotting |
587 | #endif | 587 | #endif |
588 | EXC( STORE t0, UNIT(0)(dst), s_exc) | 588 | EXC( STORE t0, UNIT(0)(dst), .Ls_exc) |
589 | ADDC(sum, t0) | 589 | ADDC(sum, t0) |
590 | EXC( STORE t1, UNIT(1)(dst), s_exc) | 590 | EXC( STORE t1, UNIT(1)(dst), .Ls_exc) |
591 | ADDC(sum, t1) | 591 | ADDC(sum, t1) |
592 | EXC( STORE t2, UNIT(2)(dst), s_exc) | 592 | EXC( STORE t2, UNIT(2)(dst), .Ls_exc) |
593 | ADDC(sum, t2) | 593 | ADDC(sum, t2) |
594 | EXC( STORE t3, UNIT(3)(dst), s_exc) | 594 | EXC( STORE t3, UNIT(3)(dst), .Ls_exc) |
595 | ADDC(sum, t3) | 595 | ADDC(sum, t3) |
596 | .set reorder /* DADDI_WAR */ | 596 | .set reorder /* DADDI_WAR */ |
597 | ADD dst, dst, 4*NBYTES | 597 | ADD dst, dst, 4*NBYTES |
598 | bne len, rem, 1b | 598 | bne len, rem, 1b |
599 | .set noreorder | 599 | .set noreorder |
600 | 600 | ||
601 | cleanup_src_unaligned: | 601 | .Lcleanup_src_unaligned: |
602 | beqz len, done | 602 | beqz len, .Ldone |
603 | and rem, len, NBYTES-1 # rem = len % NBYTES | 603 | and rem, len, NBYTES-1 # rem = len % NBYTES |
604 | beq rem, len, copy_bytes | 604 | beq rem, len, .Lcopy_bytes |
605 | nop | 605 | nop |
606 | 1: | 606 | 1: |
607 | EXC( LDFIRST t0, FIRST(0)(src), l_exc) | 607 | EXC( LDFIRST t0, FIRST(0)(src), .Ll_exc) |
608 | EXC( LDREST t0, REST(0)(src), l_exc_copy) | 608 | EXC( LDREST t0, REST(0)(src), .Ll_exc_copy) |
609 | ADD src, src, NBYTES | 609 | ADD src, src, NBYTES |
610 | SUB len, len, NBYTES | 610 | SUB len, len, NBYTES |
611 | EXC( STORE t0, 0(dst), s_exc) | 611 | EXC( STORE t0, 0(dst), .Ls_exc) |
612 | ADDC(sum, t0) | 612 | ADDC(sum, t0) |
613 | .set reorder /* DADDI_WAR */ | 613 | .set reorder /* DADDI_WAR */ |
614 | ADD dst, dst, NBYTES | 614 | ADD dst, dst, NBYTES |
615 | bne len, rem, 1b | 615 | bne len, rem, 1b |
616 | .set noreorder | 616 | .set noreorder |
617 | 617 | ||
618 | copy_bytes_checklen: | 618 | .Lcopy_bytes_checklen: |
619 | beqz len, done | 619 | beqz len, .Ldone |
620 | nop | 620 | nop |
621 | copy_bytes: | 621 | .Lcopy_bytes: |
622 | /* 0 < len < NBYTES */ | 622 | /* 0 < len < NBYTES */ |
623 | #ifdef CONFIG_CPU_LITTLE_ENDIAN | 623 | #ifdef CONFIG_CPU_LITTLE_ENDIAN |
624 | #define SHIFT_START 0 | 624 | #define SHIFT_START 0 |
@@ -629,14 +629,14 @@ copy_bytes: | |||
629 | #endif | 629 | #endif |
630 | move t2, zero # partial word | 630 | move t2, zero # partial word |
631 | li t3, SHIFT_START # shift | 631 | li t3, SHIFT_START # shift |
632 | /* use l_exc_copy here to return correct sum on fault */ | 632 | /* use .Ll_exc_copy here to return correct sum on fault */ |
633 | #define COPY_BYTE(N) \ | 633 | #define COPY_BYTE(N) \ |
634 | EXC( lbu t0, N(src), l_exc_copy); \ | 634 | EXC( lbu t0, N(src), .Ll_exc_copy); \ |
635 | SUB len, len, 1; \ | 635 | SUB len, len, 1; \ |
636 | EXC( sb t0, N(dst), s_exc); \ | 636 | EXC( sb t0, N(dst), .Ls_exc); \ |
637 | SLLV t0, t0, t3; \ | 637 | SLLV t0, t0, t3; \ |
638 | addu t3, SHIFT_INC; \ | 638 | addu t3, SHIFT_INC; \ |
639 | beqz len, copy_bytes_done; \ | 639 | beqz len, .Lcopy_bytes_done; \ |
640 | or t2, t0 | 640 | or t2, t0 |
641 | 641 | ||
642 | COPY_BYTE(0) | 642 | COPY_BYTE(0) |
@@ -647,14 +647,14 @@ EXC( sb t0, N(dst), s_exc); \ | |||
647 | COPY_BYTE(4) | 647 | COPY_BYTE(4) |
648 | COPY_BYTE(5) | 648 | COPY_BYTE(5) |
649 | #endif | 649 | #endif |
650 | EXC( lbu t0, NBYTES-2(src), l_exc_copy) | 650 | EXC( lbu t0, NBYTES-2(src), .Ll_exc_copy) |
651 | SUB len, len, 1 | 651 | SUB len, len, 1 |
652 | EXC( sb t0, NBYTES-2(dst), s_exc) | 652 | EXC( sb t0, NBYTES-2(dst), .Ls_exc) |
653 | SLLV t0, t0, t3 | 653 | SLLV t0, t0, t3 |
654 | or t2, t0 | 654 | or t2, t0 |
655 | copy_bytes_done: | 655 | .Lcopy_bytes_done: |
656 | ADDC(sum, t2) | 656 | ADDC(sum, t2) |
657 | done: | 657 | .Ldone: |
658 | /* fold checksum */ | 658 | /* fold checksum */ |
659 | .set push | 659 | .set push |
660 | .set noat | 660 | .set noat |
@@ -685,7 +685,7 @@ done: | |||
685 | jr ra | 685 | jr ra |
686 | .set noreorder | 686 | .set noreorder |
687 | 687 | ||
688 | l_exc_copy: | 688 | .Ll_exc_copy: |
689 | /* | 689 | /* |
690 | * Copy bytes from src until faulting load address (or until a | 690 | * Copy bytes from src until faulting load address (or until a |
691 | * lb faults) | 691 | * lb faults) |
@@ -700,7 +700,7 @@ l_exc_copy: | |||
700 | li t2, SHIFT_START | 700 | li t2, SHIFT_START |
701 | LOAD t0, THREAD_BUADDR(t0) | 701 | LOAD t0, THREAD_BUADDR(t0) |
702 | 1: | 702 | 1: |
703 | EXC( lbu t1, 0(src), l_exc) | 703 | EXC( lbu t1, 0(src), .Ll_exc) |
704 | ADD src, src, 1 | 704 | ADD src, src, 1 |
705 | sb t1, 0(dst) # can't fault -- we're copy_from_user | 705 | sb t1, 0(dst) # can't fault -- we're copy_from_user |
706 | SLLV t1, t1, t2 | 706 | SLLV t1, t1, t2 |
@@ -710,7 +710,7 @@ EXC( lbu t1, 0(src), l_exc) | |||
710 | ADD dst, dst, 1 | 710 | ADD dst, dst, 1 |
711 | bne src, t0, 1b | 711 | bne src, t0, 1b |
712 | .set noreorder | 712 | .set noreorder |
713 | l_exc: | 713 | .Ll_exc: |
714 | LOAD t0, TI_TASK($28) | 714 | LOAD t0, TI_TASK($28) |
715 | nop | 715 | nop |
716 | LOAD t0, THREAD_BUADDR(t0) # t0 is just past last good address | 716 | LOAD t0, THREAD_BUADDR(t0) # t0 is just past last good address |
@@ -729,7 +729,7 @@ l_exc: | |||
729 | */ | 729 | */ |
730 | .set reorder /* DADDI_WAR */ | 730 | .set reorder /* DADDI_WAR */ |
731 | SUB src, len, 1 | 731 | SUB src, len, 1 |
732 | beqz len, done | 732 | beqz len, .Ldone |
733 | .set noreorder | 733 | .set noreorder |
734 | 1: sb zero, 0(dst) | 734 | 1: sb zero, 0(dst) |
735 | ADD dst, dst, 1 | 735 | ADD dst, dst, 1 |
@@ -744,10 +744,10 @@ l_exc: | |||
744 | SUB src, src, v1 | 744 | SUB src, src, v1 |
745 | #endif | 745 | #endif |
746 | li v1, -EFAULT | 746 | li v1, -EFAULT |
747 | b done | 747 | b .Ldone |
748 | sw v1, (errptr) | 748 | sw v1, (errptr) |
749 | 749 | ||
750 | s_exc: | 750 | .Ls_exc: |
751 | li v0, -1 /* invalid checksum */ | 751 | li v0, -1 /* invalid checksum */ |
752 | li v1, -EFAULT | 752 | li v1, -EFAULT |
753 | jr ra | 753 | jr ra |