* optabs.c (expand_vec_perm): Use the correct mode for scaling the
selector. Save the qimode constant selector for later use by the
qimode vec_perm pattern.
From-SVN: r180567
The can_vec_perm_p interface changed to use a C integer array. This
allows easy re-use from the rtl level and the gimple level within
the vectorizer. It allows both to determine if a given permutation
is (un-)supported without having to create tree/rtl garbage.
The expand_vec_perm interface changed to use rtl. This allows easy
re-use from the rtl level, so that expand_vec_perm can be used in the
fallback implementation of other optabs.
* target.def (vec_perm_const_ok): Change parameters to mode and
array of indicies.
* doc/tm.texi: Rebuild.
* config/i386/i386.c (ix86_vectorize_vec_perm_const_ok): Change
parameters to mode and array of indicies.
* expr.c (expand_expr_real_2) [VEC_PERM_EXPR]: Expand operands here.
* optabs.c (can_vec_perm_p): Rename from can_vec_perm_expr_p.
Change parameters to mode and array of indicies.
(expand_vec_perm_1): Rename from expand_vec_perm_expr_1.
(expand_vec_perm): Rename from expand_vec_perm_expr. Change
parameters to mode and rtx inputs. Try lowering to QImode
vec_perm_const before trying fully variable permutation.
* optabs.h: Update decls.
* tree-vect-generic.c (lower_vec_perm): Extract array of indices from
VECTOR_CST to pass to can_vec_perm_p.
* tree-vect-slp.c (vect_get_mask_element): Change mask parameter type
from int pointer to unsigned char pointer.
(vect_transform_slp_perm_load): Update for change to can_vec_perm_p.
* tree-vect-stmts.c (perm_mask_for_reverse): Likewise.
From-SVN: r180449
* tree-vect-slp.c: Include langhooks.h.
(vect_create_mask_and_perm): Emit VEC_PERM_EXPR, not a builtin.
(vect_transform_slp_perm_load): Use can_vec_perm_expr_p. Simplify
mask creation for VEC_PERM_EXPR.
* tree-vect-stmts.c (perm_mask_for_reverse): Return the mask,
not the builtin.
(reverse_vec_elements): Emit VEC_PERM_EXPR not a builtin.
* Makefile.in (tree-vect-slp.o): Update dependency.
* optabs.c (can_vec_perm_expr_p): Allow NULL as unknown constant.
From-SVN: r180047
* doc/extend.texi (__builtin_shuffle): Improve the description to
include the modulus of the selector. Mention OpenCL.
* doc/md.texi (vec_perm, vec_perm_const): Document named patterns.
* tree.def (VEC_PERM_EXPR): Rename from VEC_SHUFFLE_EXPR.
* genopinit.c (optabs): Rename vshuffle to vec_perm.
* c-typeck.c (c_build_vec_perm_expr): Rename from
c_build_vec_shuffle_expr. Update for name changes.
* optabs.c (expand_vec_perm_expr_p): Rename from
expand_vec_shuffle_expr_p.
(expand_vec_perm_expr): Rename from expand_vec_shuffle_expr.
* optabs.h (OTI_vec_perm): Rename from DOI_vshuffle.
(vec_perm_optab): Rename from vshuffle_optab.
* expr.c, gimple-pretty-print.c, gimple.c, gimplify.c,
c-tree.h, c-parser.c, tree-cfg.c, tree-inline.c, tree-pretty-print.c,
tree-ssa-operands.c, tree-vect-generic.c: Update for name changes.
* config/i386/i386.c (ix86_expand_vec_perm): Rename from
ix86_expand_vshuffle.
* config/i386/i386-protos.h: Update.
* config/i386/sse.md (VEC_PERM_AVX2): Rename from VSHUFFLE_AVX2.
(vec_perm<VEC_PERM_AVX2>): Rename from vshuffle<VSHUFFLE_AVX2>.
From-SVN: r179701
PR middle-end/50161
* simplify-rtx.c (simplify_const_unary_operation): If
op is CONST_INT, don't look at op_mode, but use instead
mode.
* optabs.c (add_equal_note): For FFS, CLZ, CTZ,
CLRSB, POPCOUNT, PARITY and BSWAP use operand mode for
operation and TRUNCATE/ZERO_EXTEND if needed.
* doc/rtl.texi (ffs, clrsb, clz, ctz, popcount, parity, bswap):
Document that operand mode must be same as operation mode,
or VOIDmode.
* config/avr/avr.md (paritysi2, *parityqihi2.libgcc,
*paritysihi2.libgcc, popcountsi2, *popcountsi2.libgcc,
*popcountqihi2.libgcc, clzsi2, *clzsihi2.libgcc, ctzsi2,
*ctzsihi2.libgcc, ffssi2, *ffssihi2.libgcc): For unary ops
use the mode of operand for the operation and add truncate
or zero_extend around if needed.
* config/c6x/c6x.md (ctzdi2): Likewise.
* config/bfin/bfin.md (clrsbsi2, signbitssi2): Likewise.
* gcc.dg/pr50161.c: New test.
From-SVN: r177991
2011-08-19 Andrew Stubbs <ams@codesourcery.com>
gcc/
* config/arm/arm.md (maddhidi4): Remove '*' from name.
* expr.c (expand_expr_real_2): Use find_widening_optab_handler.
* optabs.c (find_widening_optab_handler_and_mode): New function.
(expand_widen_pattern_expr): Use find_widening_optab_handler.
(expand_binop_directly): Likewise.
(expand_binop): Likewise.
* optabs.h (find_widening_optab_handler): New macro define.
(find_widening_optab_handler_and_mode): New prototype.
* tree-cfg.c (verify_gimple_assign_binary): Adjust WIDEN_MULT_EXPR
type precision rules.
(verify_gimple_assign_ternary): Likewise for WIDEN_MULT_PLUS_EXPR.
* tree-ssa-math-opts.c (build_and_insert_cast): New function.
(is_widening_mult_rhs_p): Allow widening by more than one mode.
Explicitly disallow mis-matched input types.
(convert_mult_to_widen): Use find_widening_optab_handler, and cast
input types to fit the new handler.
(convert_plusminus_to_widen): Likewise.
gcc/testsuite/
* gcc.target/arm/wmul-bitfield-1.c: New file.
From-SVN: r177902
2011-08-19 Andrew Stubbs <ams@codesourcery.com>
gcc/
* expr.c (expand_expr_real_2): Use widening_optab_handler.
* genopinit.c (optabs): Use set_widening_optab_handler for $N.
(gen_insn): $N now means $a must be wider than $b, not consecutive.
* optabs.c (widened_mode): New function.
(expand_widen_pattern_expr): Use widening_optab_handler.
(expand_binop_directly): Likewise.
(expand_binop): Likewise.
* optabs.h (widening_optab_handlers): New struct.
(optab_d): New member, 'widening'.
(widening_optab_handler): New function.
(set_widening_optab_handler): New function.
* tree-ssa-math-opts.c (convert_mult_to_widen): Use
widening_optab_handler.
(convert_plusminus_to_widen): Likewise.
From-SVN: r177901
* optabs.c (expand_binop): Tighten conditions for doubleword
expansions.
(widen_bswap): Assert that mode bitsize and precision are the
same.
* stor-layout.c (get_best_mode): Skip modes that have lower
precision than bitsize.
* recog.c (simplify_while_replacing): Assert that bitsize and
precision are the same.
From-SVN: r176040
PR middle-end/49489
* builtins.c (expand_builtin_unop): Call expand_unop with 0 as
unsignedp argument instead of 1 for clrsb_optab.
(fold_builtin_bitop): Fix masking for width > HOST_BITS_PER_WIDE_INT
and < 2 * HOST_BITS_PER_WIDE_INT. Optimize BUILT_IN_CLRSB*.
(fold_builtin_1): Call fold_builtin_binop for BUILT_IN_CLRSB*.
* optabs.c (widen_leading): Call widen_operand and expand_unop
with 0 as unsignedp argument instead of 1 for clrsb_optab.
(expand_unop): Subtract difference of mode sizes also for
clrsb_optab.
From-SVN: r175265
gcc/
* recog.h (insn_operand_data): Add an "allows_mem" field.
* genoutput.c (output_operand_data): Initialize it.
* optabs.c (maybe_legitimize_operand_same_code): New function.
(maybe_legitimize_operand): Use it when matching the original
op->value.
From-SVN: r172478
gcc/
* recog.h (insn_operand_data): Add an "allows_mem" field.
* genoutput.c (output_operand_data): Initialize it.
* optabs.c (maybe_legitimize_operand_same_code): New function.
(maybe_legitimize_operand): Use it when matching the original
op->value.
From-SVN: r172316
gcc/
PR rtl-optimization/48332
* optabs.c (expand_binop_directly): Set xmodeN to the target-mandated
mode of input operand N and modeN to its actual mode.
From-SVN: r171733
gcc/
PR rtl-optimization/48263
* optabs.c (expand_binop_directly): Reinstate convert_modes code
and original commutative_p handling. Use maybe_gen_insn.
From-SVN: r171418
2011-03-04 Richard Guenther <rguenther@suse.de>
PR middle-end/47975
* optabs.c (optab_for_tree_code): Do not use VECTOR_MODE_P.
* gcc.dg/torture/pr47975.c: New testcase.
From-SVN: r170672
gcc/
* Makefile.in (LIBFUNCS_H): Add $(HASHTAB_H).
(target-globals.o): Depend on $(LIBFUNCS_H).
* libfuncs.h: Include hashtab.h.
(libfunc_entry): Moved from optabs.c.
(target_libfuncs): New structure.
(default_target_libfuncs): Declare.
(this_target_libfuncs): Declare as a variable or define as a macro.
(libfunc_table): Redefine as a macro.
* optabs.c (default_target_libfuncs): New variable.
(this_target_libfuncs): New conditional variable.
(libfunc_table): Delete.
(libfunc_entry): Moved to optabs.h.
(libfunc_hash): Redefine as a macro.
(hash_libfunc, eq_libfunc): Fix comments.
(init_optabs): Use libfunc_hash to detect cases where the function
has already been called. Clear the hash table instead of
recreating it.
* target-globals.h (this_target_libfuncs): Declare.
(target_globals): Add a libfuncs field.
(restore_target_globals): Copy the libfuncs field to
this_target_libfuncs.
* target-globals.c: Include libfuncs.h.
(default_target_globals): Initialize the libfuncs field.
(save_target_globals): Likewise.
From-SVN: r162099
gcc/
* Makefile.in (target-globals.o): Depend on $(EXPR_H) and $(OPTABS_H).
* optabs.h (target_optabs): New structure.
(default_target_optabs): Declare.
(this_target_optabs): Declare as a variable or define as a macro.
(optab_table, convert_optab_table, direct_optab_table): Redefine
as macros.
* optabs.c (default_target_optabs): New variable.
(this_target_optabs): New conditional variable.
(optab_table, convert_optab_table, direct_optab_table): Delete.
* target-globals.h (this_target_optabs): Declare.
(target_globals): Add a optabs field.
(restore_target_globals): Copy the optabs field to
this_target_optabs.
* target-globals.c: Include expr.h and optabs.h.
(default_target_globals): Initialize the optabs field.
(save_target_globals): Likewise.
From-SVN: r162097
* optabs.h (reload_in_optab, reload_out_optab, code_to_optab)
(vcond_gen_code, vcondu_gen_code, movmem_optab, setmem_optab)
(cmpstr_optab, cmpstrn_optab, cmpmem_optab, sync_add_optab)
(sync_sub_optab, sync_ior_optab, sync_and_optab, sync_xor_optab)
(sync_nand_optab, sync_old_add_optab, sync_old_sub_optab)
(sync_old_ior_optab, sync_old_and_optab, sync_old_xor_optab)
(sync_old_nand_optab, sync_new_add_optab, sync_new_sub_optab)
(sync_new_ior_optab, sync_new_and_optab, sync_new_xor_optab)
(sync_new_nand_optab): Redefine as macros.
(sync_compare_and_swap, sync_lock_test_and_set, sync_lock_release):
Delete.
(direct_optab_index): New enum.
(direct_optab_d): New structure.
(direct_optab): New typedef.
(direct_optab_table): Declare.
(direct_optab_handler, set_direct_optab_handler): New functions.
(sync_compare_and_swap_optab, sync_lock_test_and_set_optab)
(sync_lock_release_optab): New macros.
* optabs.c (direct_optab_table): New variable.
(movcc_gen_code, vcond_gen_code, vcondu_gen_code): Delete.
(prepare_cmp_insn): Use direct_optab_handler for cmpmem_optab,
cmpstr_optab and cmpstrn_optab.
(emit_conditional_move): Likewise for movcc_optab.
(can_conditionally_move_p): Likewise for movcc_gen_code.
(init_insn_codes): Clear direct_optab_table.
(init_optabs): Don't initialize the new "direct optabs" here.
(get_vcond_icode): Use direct_optab_handler for vcondu_gen_code and
vcond_gen_code.
(expand_val_compare_and_swap): Likewise sync_compare_and_swap_optab.
(expand_bool_compare_and_swap): Likewise sync_compare_and_swap_optab.
(expand_compare_and_swap_loop): Likewise sync_compare_and_swap_optab.
(expand_sync_operation): Likewise other sync_*_optabs.
(expand_sync_fetch_operation): Likewise. Rename sync_compare_and_swap
to sync_compare_and_swap_optab.
(expand_sync_lock_test_and_set): Use direct_optab_handler for
sync_lock_test_and_set and sync_compare_and_swap, adding "_optab"
to the names of both.
* builtins.c (expand_builtin_strcmp): Use direct_optab_handler for
cmpstr_optab and cmpstrn_optab.
(expand_builtin_lock_release): Likewise sync_lock_release.
* expr.c (movmem_optab, setmem_optab, cmpstr_optab, cmpstrn_optab)
(cmpmem_optab, sync_add_optab, sync_sub_optab, sync_ior_optab)
(sync_and_optab, sync_xor_optab, sync_nand_optab, sync_old_add_optab)
(sync_old_sub_optab, sync_old_ior_optab, sync_old_and_optab)
(sync_old_xor_optab, sync_old_nand_optab, sync_new_add_optab)
(sync_new_sub_optab, sync_new_ior_optab, sync_new_and_optab)
(sync_new_xor_optab, sync_new_nand_optab, sync_compare_and_swap)
(sync_lock_test_and_set, sync_lock_release): Delete.
(emit_block_move_via_movmem): Use direct_optab_handler for movmem_optab.
(emit_block_move_via_setmem): Use direct_optab_handler for setmem_optab.
* genopinit.c (optabs): Use set_direct_optab_handler for the new
macro optabs.
* omp-low.c (expand_omp_atomic_fetch_op): Update the type of
the "optab" local variable. Use direct_optab_handler for optab and
sync_compare_and_swap_optab.
* reload1.c (reload_in_optab, reload_out_optab): Delete.
* targhooks.c (default_secondary_reload): Use direct_optab_handler for
reload_in_optab and reload_out_optab.
* config/alpha/alpha.c (alpha_secondary_reload): Likewise.
* config/frv/frv.c (frv_alloc_temp_reg): Likewise.
* config/pa/pa.c (pa_secondary_reload): Likewise.
* java/builtins.c (compareAndSwapInt_builtin): Use direct_optab_handler
for sync_compare_and_swap, renaming it to sync_compare_and_swap_optab.
(compareAndSwapLong_builtin, compareAndSwapObject_builtin): Likewise.
(VMSupportsCS8_builtin): Likewise.
From-SVN: r161810
gcc/
* optabs.h (optab_handlers): Change type of insn_code to int.
(optab_handler, set_optab_handler, convert_optab_handler)
(set_convert_optab_handler): Treat the insn_code field as "insn_code -
CODE_FOR_nothing".
* optabs.c (optab_table, convert_optab_table): Always zero-initialize.
(init_insn_codes): Zero both the above arrays.
(init_optabs): Never call init_insn_codes first time around.
From-SVN: r161809