aboutsummaryrefslogtreecommitdiff
path: root/gcc/expmed.c
AgeCommit message (Expand)AuthorFilesLines
1998-01-14alias.c: Change all uses of gen_rtx(FOO...) to gen_rtx_FOO...Richard Henderson1-93/+101
1997-12-15* expmed.c (expand_shift): If SHIFT_COUNT_TRUNCATED, drop a SUBREG.Richard Henderson1-5/+10
1997-12-06Merge from gcc-2.8Jeff Law1-0/+2
1997-11-02Update mainline egcs to gcc2 snapshot 971021.Jeff Law1-3/+9
1997-10-20Patches to fix minor optimizer bugsJim Wilson1-3/+2
1997-09-23expmed.c (expand_divmod): If compute_mode is not the same as mode...David S. Miller1-0/+8
1997-09-07version.c: Bump for snapshot.Jeff Law1-1/+3
1997-05-06reformat a little to match GNU coding standards.Mike Stump1-15/+15
1997-02-15(expand_divmod): prefer divmod in same mode over div in wider mode.Richard Kenner1-5/+12
1997-01-27(emit_store_flag): Properly test for STORE_FLAG_VALUE of just sign bit.Richard Kenner1-2/+2
1997-01-04(store_fixed_bit_field): If not SLOW_UNALIGNED_ACCESS, treatRichard Kenner1-1/+4
1996-12-12* expmed.c (emit_store_flag_force): Fix typos/thinkos.Jeff Law1-2/+2
1996-10-21(emit_store_flag_force): New function.Torbjorn Granlund1-0/+37
1996-09-25(store_bit_field): don't make flag_force_mem disable insvJim Wilson1-6/+6
1996-07-17(expand_mult_highpart): Revert last change.Torbjorn Granlund1-2/+2
1996-07-03formatting tweaksMike Stump1-3/+3
1996-06-28(expand_mult_highpart): Use op1 not wide_op1 in expansion ofRichard Kenner1-2/+2
1996-06-27(extract_bit_field): Check TRULY_NOOP_TRUNCATION beforeJim Wilson1-1/+3
1996-05-17(store_split_bit_field): Don't assume the alignment of VALUE is theRichard Kenner1-6/+21
1996-05-05Fix typo in comment.Richard Kenner1-1/+1
1996-03-30(expand_divmod, case TRUNC_DIV_EXPR): Move some codeTorbjorn Granlund1-70/+74
1996-03-27(negate_rtx): Fix typo in previous change.Richard Kenner1-1/+1
1996-03-24(negate_rtx): Don't try to negate a constant ourself; instead callRichard Kenner1-15/+6
1996-03-18(emit_store_flag): If expanding (GE X 0) will need two insns, don't use subta...Richard Kenner1-2/+4
1996-01-17(extract_bit_field): For multi-word bitfield, clobber target beforeJim Wilson1-0/+3
1996-01-16(extract_bit_field): Don't abort if not MEM_IN_STRUCT_P.Richard Kenner1-4/+1
1995-12-17(expand_mult_highpart): When doing widening multiply,Torbjorn Granlund1-7/+21
1995-12-15(store_bit_field): Don't use insv for BLKmode value.Richard Kenner1-1/+2
1995-12-14(expand_divmod...Torbjorn Granlund1-3/+7
1995-12-13(expand_divmod): Don't use TARGET if it's the wrong mode.Richard Kenner1-3/+9
1995-07-13(SHIFT_COUNT_TRUNCATED): Use #ifdef not #if.Richard Kenner1-1/+1
1995-06-15Update FSF address.Richard Kenner1-1/+2
1995-06-14(emit_store_flag): Always set LAST.Richard Kenner1-3/+2
1995-06-08(store_split_bit_field): When adjust arg in BYTES_BIT_ENDIAN case...Richard Kenner1-6/+14
1995-06-08(expand_divmod): Always check result of emit_store_flag.Richard Kenner1-5/+8
1995-06-01(extract_fixed_bit_field): For REG case, compute total_bits from modeRichard Kenner1-3/+4
1995-05-31(expand_mult_highpart): Use wide_op1 for all multiplies.Richard Kenner1-5/+7
1995-05-16Fix typos in comments.Richard Kenner1-1/+1
1995-05-02If expensive optimizations, use different pseudos when making store flag insns.Michael Meissner1-7/+18
1995-04-03Get correct bit offset for big endian (SUBREG (REG))Michael Meissner1-0/+14
1995-02-13(store_split_bit_field): If value is VOIDmode, pass word_mode to force_reg.Doug Evans1-1/+3
1995-02-11(expand_divmod): Always call expand_twoval_binop with psuedos asRichard Kenner1-6/+14
1995-02-06(mul_cost, div_cost): New variables.Torbjorn Granlund1-50/+113
1995-01-14Add 1995 to copyright.Doug Evans1-1/+1
1995-01-14(extract_bit_field, fields bigger than word case): Count subwords in target.Doug Evans1-3/+29
1994-12-29(expand_mult_highpart): Use convert_modes instead ofJim Wilson1-3/+3
1994-12-12(emit_store_flag): Set mode after canonicalizing op0/op1 so op1 is constant.Doug Evans1-3/+3
1994-11-29(expand_shift): Open-code rotate even if by a variable.Richard Kenner1-21/+20
1994-11-28(store_bit_field, extract_bit_field): Consistently useJim Wilson1-3/+3
1994-11-19(expand_shift): Re-enable truncation of immediate shift count.Jim Wilson1-1/+1