;; Machine description for AArch64 SVE. ;; Copyright (C) 2009-2021 Free Software Foundation, Inc. ;; Contributed by ARM Ltd. ;; ;; This file is part of GCC. ;; ;; GCC is free software; you can redistribute it and/or modify it ;; under the terms of the GNU General Public License as published by ;; the Free Software Foundation; either version 3, or (at your option) ;; any later version. ;; ;; GCC is distributed in the hope that it will be useful, but ;; WITHOUT ANY WARRANTY; without even the implied warranty of ;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU ;; General Public License for more details. ;; ;; You should have received a copy of the GNU General Public License ;; along with GCC; see the file COPYING3. If not see ;; http://www.gnu.org/licenses/.

;; The file is organised into the following sections (search for the full ;; line): ;; ;; == General notes ;; ---- Note on the handling of big-endian SVE ;; ---- Description of UNSPEC_PTEST ;; ---- Description of UNSPEC_PRED_Z ;; ---- Note on predicated integer arithemtic and UNSPEC_PRED_X ;; ---- Note on predicated FP arithmetic patterns and GP “strictness” ;; ---- Note on FFR handling ;; ;; == Moves ;; ---- Moves of single vectors ;; ---- Moves of multiple vectors ;; ---- Moves of predicates ;; ---- Moves relating to the FFR ;; ;; == Loads ;; ---- Normal contiguous loads ;; ---- Extending contiguous loads ;; ---- First-faulting contiguous loads ;; ---- First-faulting extending contiguous loads ;; ---- Non-temporal contiguous loads ;; ---- Normal gather loads ;; ---- Extending gather loads ;; ---- First-faulting gather loads ;; ---- First-faulting extending gather loads ;; ;; == Prefetches ;; ---- Contiguous prefetches ;; ---- Gather prefetches ;; ;; == Stores ;; ---- Normal contiguous stores ;; ---- Truncating contiguous stores ;; ---- Non-temporal contiguous stores ;; ---- Normal scatter stores ;; ---- Truncating scatter stores ;; ;; == Vector creation ;; ---- [INT,FP] Duplicate element ;; ---- [INT,FP] Initialize from individual elements ;; ---- [INT] Linear series ;; ---- [PRED] Duplicate element ;; ;; == Vector decomposition ;; ---- [INT,FP] Extract index ;; ---- [INT,FP] Extract active element ;; ---- [PRED] Extract index ;; ;; == Unary arithmetic ;; ---- [INT] General unary arithmetic corresponding to rtx codes ;; ---- [INT] General unary arithmetic corresponding to unspecs ;; ---- [INT] Sign and zero extension ;; ---- [INT] Truncation ;; ---- [INT] Logical inverse ;; ---- [FP<-INT] General unary arithmetic that maps to unspecs ;; ---- [FP] General unary arithmetic corresponding to unspecs ;; ---- [FP] Square root ;; ---- [FP] Reciprocal square root ;; ---- [PRED] Inverse

;; == Binary arithmetic ;; ---- [INT] General binary arithmetic corresponding to rtx codes ;; ---- [INT] Addition ;; ---- [INT] Subtraction ;; ---- [INT] Take address ;; ---- [INT] Absolute difference ;; ---- [INT] Saturating addition and subtraction ;; ---- [INT] Highpart multiplication ;; ---- [INT] Division ;; ---- [INT] Binary logical operations ;; ---- [INT] Binary logical operations (inverted second input) ;; ---- [INT] Shifts (rounding towards -Inf) ;; ---- [INT] Shifts (rounding towards 0) ;; ---- [FP<-INT] General binary arithmetic corresponding to unspecs ;; ---- [FP] General binary arithmetic corresponding to rtx codes ;; ---- [FP] General binary arithmetic corresponding to unspecs ;; ---- [FP] Addition ;; ---- [FP] Complex addition ;; ---- [FP] Subtraction ;; ---- [FP] Absolute difference ;; ---- [FP] Multiplication ;; ---- [FP] Division ;; ---- [FP] Binary logical operations ;; ---- [FP] Sign copying ;; ---- [FP] Maximum and minimum ;; ---- [PRED] Binary logical operations ;; ---- [PRED] Binary logical operations (inverted second input) ;; ---- [PRED] Binary logical operations (inverted result) ;; ;; == Ternary arithmetic ;; ---- [INT] MLA and MAD ;; ---- [INT] MLS and MSB ;; ---- [INT] Dot product ;; ---- [INT] Sum of absolute differences ;; ---- [INT] Matrix multiply-accumulate ;; ---- [FP] General ternary arithmetic corresponding to unspecs ;; ---- [FP] Complex multiply-add ;; ---- [FP] Trigonometric multiply-add ;; ---- [FP] Bfloat16 long ternary arithmetic (SF,BF,BF) ;; ---- [FP] Matrix multiply-accumulate ;; ;; == Comparisons and selects ;; ---- [INT,FP] Select based on predicates ;; ---- [INT,FP] Compare and select ;; ---- [INT] Comparisons ;; ---- [INT] While tests ;; ---- [FP] Direct comparisons ;; ---- [FP] Absolute comparisons ;; ---- [PRED] Select ;; ---- [PRED] Test bits ;; ;; == Reductions ;; ---- [INT,FP] Conditional reductions ;; ---- [INT] Tree reductions ;; ---- [FP] Tree reductions ;; ---- [FP] Left-to-right reductions ;; ;; == Permutes ;; ---- [INT,FP] General permutes ;; ---- [INT,FP] Special-purpose unary permutes ;; ---- [INT,FP] Special-purpose binary permutes ;; ---- [PRED] Special-purpose unary permutes ;; ---- [PRED] Special-purpose binary permutes ;; ;; == Conversions ;; ---- [INT<-INT] Packs ;; ---- [INT<-INT] Unpacks ;; ---- [INT<-FP] Conversions ;; ---- [INT<-FP] Packs ;; ---- [INT<-FP] Unpacks ;; ---- [FP<-INT] Conversions ;; ---- [FP<-INT] Packs ;; ---- [FP<-INT] Unpacks ;; ---- [FP<-FP] Packs ;; ---- [FP<-FP] Packs (bfloat16) ;; ---- [FP<-FP] Unpacks ;; ---- [PRED<-PRED] Packs ;; ---- [PRED<-PRED] Unpacks ;; ;; == Vector partitioning ;; ---- [PRED] Unary partitioning ;; ---- [PRED] Binary partitioning ;; ---- [PRED] Scalarization ;; ;; == Counting elements ;; ---- [INT] Count elements in a pattern (scalar) ;; ---- [INT] Increment by the number of elements in a pattern (scalar) ;; ---- [INT] Increment by the number of elements in a pattern (vector) ;; ---- [INT] Decrement by the number of elements in a pattern (scalar) ;; ---- [INT] Decrement by the number of elements in a pattern (vector) ;; ---- [INT] Count elements in a predicate (scalar) ;; ---- [INT] Increment by the number of elements in a predicate (scalar) ;; ---- [INT] Increment by the number of elements in a predicate (vector) ;; ---- [INT] Decrement by the number of elements in a predicate (scalar) ;; ---- [INT] Decrement by the number of elements in a predicate (vector)

;; ========================================================================= ;; == General notes ;; ========================================================================= ;; ;; ------------------------------------------------------------------------- ;; ---- Note on the handling of big-endian SVE ;; ------------------------------------------------------------------------- ;; ;; On big-endian systems, Advanced SIMD mov patterns act in the ;; same way as movdi or movti would: the first byte of memory goes ;; into the most significant byte of the register and the last byte ;; of memory goes into the least significant byte of the register. ;; This is the most natural ordering for Advanced SIMD and matches ;; the ABI layout for 64-bit and 128-bit vector types. ;; ;; As a result, the order of bytes within the register is what GCC ;; expects for a big-endian target, and subreg offsets therefore work ;; as expected, with the first element in memory having subreg offset 0 ;; and the last element in memory having the subreg offset associated ;; with a big-endian lowpart. However, this ordering also means that ;; GCC‘s lane numbering does not match the architecture’s numbering: ;; GCC always treats the element at the lowest address in memory ;; (subreg offset 0) as element 0, while the architecture treats ;; the least significant end of the register as element 0. ;; ;; The situation for SVE is different. We want the layout of the ;; SVE register to be same for mov as it is for maskload: ;; logically, a mov load must be indistinguishable from a ;; maskload whose mask is all true. We therefore need the ;; register layout to match LD1 rather than LDR. The ABI layout of ;; SVE types also matches LD1 byte ordering rather than LDR byte ordering. ;; ;; As a result, the architecture lane numbering matches GCC‘s lane ;; numbering, with element 0 always being the first in memory. ;; However: ;; ;; - Applying a subreg offset to a register does not give the element ;; that GCC expects: the first element in memory has the subreg offset ;; associated with a big-endian lowpart while the last element in memory ;; has subreg offset 0. We handle this via TARGET_CAN_CHANGE_MODE_CLASS. ;; ;; - We cannot use LDR and STR for spill slots that might be accessed ;; via subregs, since although the elements have the order GCC expects, ;; the order of the bytes within the elements is different. We instead ;; access spill slots via LD1 and ST1, using secondary reloads to ;; reserve a predicate register. ;; ;; ------------------------------------------------------------------------- ;; ---- Description of UNSPEC_PTEST ;; ------------------------------------------------------------------------- ;; ;; SVE provides a PTEST instruction for testing the active lanes of a ;; predicate and setting the flags based on the result. The associated ;; condition code tests are: ;; ;; - any (= ne): at least one active bit is set ;; - none (= eq): all active bits are clear () ;; - first (= mi): the first active bit is set ;; - nfrst (= pl): the first active bit is clear () ;; - last (= cc): the last active bit is set ;; - nlast (= cs): the last active bit is clear () ;; ;; where the conditions marked () are also true when there are no active ;; lanes (i.e. when the governing predicate is a PFALSE). The flags results ;; of a PTEST use the condition code mode CC_NZC. ;; ;; PTEST is always a .B operation (i.e. it always operates on VNx16BI). ;; This means that for other predicate modes, we need a governing predicate ;; in which all bits are defined. ;; ;; For example, most predicated .H operations ignore the odd bits of the ;; governing predicate, so that an active lane is represented by the ;; bits “1x” and an inactive lane by the bits “0x”, where “x” can be ;; any value. To test a .H predicate, we instead need “10” and “00” ;; respectively, so that the condition only tests the even bits of the ;; predicate. ;; ;; Several instructions set the flags as a side-effect, in the same way ;; that a separate PTEST would. It’s important for code quality that we ;; use these flags results as often as possible, particularly in the case ;; of WHILE* and RDFFR. ;; ;; Also, some of the instructions that set the flags are unpredicated ;; and instead implicitly test all .B, .H, .S or .D elements, as though ;; they were predicated on a PTRUE of that size. For example, a .S ;; WHILELO sets the flags in the same way as a PTEST with a .S PTRUE ;; would. ;; ;; We therefore need to represent PTEST operations in a way that ;; makes it easy to combine them with both predicated and unpredicated ;; operations, while using a VNx16BI governing predicate for all ;; predicate modes. We do this using: ;; ;; (unspec:CC_NZC [gp cast_gp ptrue_flag op] UNSPEC_PTEST) ;; ;; where: ;; ;; - GP is the real VNx16BI governing predicate ;; ;; - CAST_GP is GP cast to the mode of OP. All bits dropped by casting ;; GP to CAST_GP are guaranteed to be clear in GP. ;; ;; - PTRUE_FLAG is a CONST_INT (conceptually of mode SI) that has the value ;; SVE_KNOWN_PTRUE if we know that CAST_GP (rather than GP) is all-true and ;; SVE_MAYBE_NOT_PTRUE otherwise. ;; ;; - OP is the predicate we want to test, of the same mode as CAST_GP. ;; ;; ------------------------------------------------------------------------- ;; ---- Description of UNSPEC_PRED_Z ;; ------------------------------------------------------------------------- ;; ;; SVE integer comparisons are predicated and return zero for inactive ;; lanes. Sometimes we use them with predicates that are all-true and ;; sometimes we use them with general predicates. ;; ;; The integer comparisons also set the flags and so build-in the effect ;; of a PTEST. We therefore want to be able to combine integer comparison ;; patterns with PTESTs of the result. One difficulty with doing this is ;; that (as noted above) the PTEST is always a .B operation and so can place ;; stronger requirements on the governing predicate than the comparison does. ;; ;; For example, when applying a separate PTEST to the result of a full-vector ;; .H comparison, the PTEST must be predicated on a .H PTRUE instead of a ;; .B PTRUE. In constrast, the comparison might be predicated on either ;; a .H PTRUE or a .B PTRUE, since the values of odd-indexed predicate ;; bits don‘t matter for .H operations. ;; ;; We therefore can’t rely on a full-vector comparison using the same ;; predicate register as a following PTEST. We instead need to remember ;; whether a comparison is known to be a full-vector comparison and use ;; this information in addition to a check for equal predicate registers. ;; At the same time, it‘s useful to have a common representation for all ;; integer comparisons, so that they can be handled by a single set of ;; patterns. ;; ;; We therefore take a similar approach to UNSPEC_PTEST above and use: ;; ;; (unspec:<M:VPRED> [gp ptrue_flag (code:M op0 op1)] UNSPEC_PRED_Z) ;; ;; where: ;; ;; - GP is the governing predicate, of mode <M:VPRED> ;; ;; - PTRUE_FLAG is a CONST_INT (conceptually of mode SI) that has the value ;; SVE_KNOWN_PTRUE if we know that GP is all-true and SVE_MAYBE_NOT_PTRUE ;; otherwise ;; ;; - CODE is the comparison code ;; ;; - OP0 and OP1 are the values being compared, of mode M ;; ;; The “Z” in UNSPEC_PRED_Z indicates that inactive lanes are zero. ;; ;; ------------------------------------------------------------------------- ;; ---- Note on predicated integer arithemtic and UNSPEC_PRED_X ;; ------------------------------------------------------------------------- ;; ;; Many SVE integer operations are predicated. We can generate them ;; from four sources: ;; ;; (1) Using normal unpredicated optabs. In this case we need to create ;; an all-true predicate register to act as the governing predicate ;; for the SVE instruction. There are no inactive lanes, and thus ;; the values of inactive lanes don’t matter. ;; ;; (2) Using x ACLE functions. In this case the function provides a ;; specific predicate and some lanes might be inactive. However, ;; as for (1), the values of the inactive lanes don't matter. ;; We can make extra lanes active without changing the behavior ;; (although for code-quality reasons we should avoid doing so ;; needlessly). ;; ;; (3) Using cond* optabs that correspond to IFN_COND_* internal functions. ;; These optabs have a predicate operand that specifies which lanes are ;; active and another operand that provides the values of inactive lanes. ;; ;; (4) Using _m and _z ACLE functions. These functions map to the same ;; patterns as (3), with the z functions setting inactive lanes to zero ;; and the m functions setting the inactive lanes to one of the function ;; arguments. ;; ;; For (1) and (2) we need a way of attaching the predicate to a normal ;; unpredicated integer operation. We do this using: ;; ;; (unspec:M [pred (code:M (op0 op1 ...))] UNSPEC_PRED_X) ;; ;; where (code:M (op0 op1 ...)) is the normal integer operation and PRED ;; is a predicate of mode <M:VPRED>. PRED might or might not be a PTRUE; ;; it always is for (1), but might not be for (2). ;; ;; The unspec as a whole has the same value as (code:M ...) when PRED is ;; all-true. It is always semantically valid to replace PRED with a PTRUE, ;; but as noted above, we should only do so if there‘s a specific benefit. ;; ;; (The “_X” in the unspec is named after the ACLE functions in (2).) ;; ;; For (3) and (4) we can simply use the SVE port’s normal representation ;; of a predicate-based select: ;; ;; (unspec:M [pred (code:M (op0 op1 ...)) inactive] UNSPEC_SEL) ;; ;; where INACTIVE specifies the values of inactive lanes. ;; ;; We can also use the UNSPEC_PRED_X wrapper in the UNSPEC_SEL rather ;; than inserting the integer operation directly. This is mostly useful ;; if we want the combine pass to merge an integer operation with an explicit ;; vcond_mask (in other words, with a following SEL instruction). However, ;; it‘s generally better to merge such operations at the gimple level ;; using (3). ;; ;; ------------------------------------------------------------------------- ;; ---- Note on predicated FP arithmetic patterns and GP “strictness” ;; ------------------------------------------------------------------------- ;; ;; Most SVE floating-point operations are predicated. We can generate ;; them from four sources: ;; ;; (1) Using normal unpredicated optabs. In this case we need to create ;; an all-true predicate register to act as the governing predicate ;; for the SVE instruction. There are no inactive lanes, and thus ;; the values of inactive lanes don’t matter. ;; ;; (2) Using x ACLE functions. In this case the function provides a ;; specific predicate and some lanes might be inactive. However, ;; as for (1), the values of the inactive lanes don‘t matter. ;; ;; The instruction must have the same exception behavior as the ;; function call unless things like command-line flags specifically ;; allow otherwise. For example, with -ffast-math, it is OK to ;; raise exceptions for inactive lanes, but normally it isn’t. ;; ;; (3) Using cond* optabs that correspond to IFN_COND* internal functions. ;; These optabs have a predicate operand that specifies which lanes are ;; active and another operand that provides the values of inactive lanes. ;; ;; (4) Using m and z ACLE functions. These functions map to the same ;; patterns as (3), with the z functions setting inactive lanes to zero ;; and the m functions setting the inactive lanes to one of the function ;; arguments. ;; ;; So: ;; ;; - In (1), the predicate is known to be all true and the pattern can use ;; unpredicated operations where available. ;; ;; - In (2), the predicate might or might not be all true. The pattern can ;; use unpredicated instructions if the predicate is all-true or if things ;; like command-line flags allow exceptions for inactive lanes. ;; ;; - (3) and (4) represent a native SVE predicated operation. Some lanes ;; might be inactive and inactive lanes of the result must have specific ;; values. There is no scope for using unpredicated instructions (and no ;; reason to want to), so the question about command-line flags doesn't ;; arise. ;; ;; It would be inaccurate to model (2) as an rtx code like (sqrt ...) ;; in combination with a separate predicate operand, e.g. ;; ;; (unspec [(match_operand: 1 “register_operand” “Upl”) ;; (sqrt:SVE_FULL_F 2 “register_operand” “w”)] ;; ....) ;; ;; because (sqrt ...) can raise an exception for any lane, including ;; inactive ones. We therefore need to use an unspec instead. ;; ;; Also, (2) requires some way of distinguishing the case in which the ;; predicate might have inactive lanes and cannot be changed from the ;; case in which the predicate has no inactive lanes or can be changed. ;; This information is also useful when matching combined FP patterns ;; in which the predicates might not be equal. ;; ;; We therefore model FP operations as an unspec of the form: ;; ;; (unspec [pred strictness op0 op1 ...] UNSPEC_COND) ;; ;; where: ;; ;; - PRED is the governing predicate. ;; ;; - STRICTNESS is a CONST_INT that conceptually has mode SI. It has the ;; value SVE_STRICT_GP if PRED might have inactive lanes and if those ;; lanes must remain inactive. It has the value SVE_RELAXED_GP otherwise. ;; ;; - OP0 OP1 ... are the normal input operands to the operation. ;; ;; - MNEMONIC is the mnemonic of the associated SVE instruction. ;; ;; For (3) and (4), we combine these operations with an UNSPEC_SEL ;; that selects between the result of the FP operation and the “else” ;; value. (This else value is a merge input for m ACLE functions ;; and zero for z ACLE functions.) The outer pattern then has the form: ;; ;; (unspec [pred fp_operation else_value] UNSPEC_SEL) ;; ;; This means that the patterns for (3) and (4) have two predicates: ;; one for the FP operation itself and one for the UNSPEC_SEL. ;; This pattern is equivalent to the result of combining an instance ;; of (1) or (2) with a separate vcond instruction, so these patterns ;; are useful as combine targets too. ;; ;; However, in the combine case, the instructions that we want to ;; combine might use different predicates. Then: ;; ;; - Some of the active lanes of the FP operation might be discarded ;; by the UNSPEC_SEL. It's OK to drop the FP operation on those lanes, ;; even for SVE_STRICT_GP, since the operations on those lanes are ;; effectively dead code. ;; ;; - Some of the inactive lanes of the FP operation might be selected ;; by the UNSPEC_SEL, giving unspecified values for those lanes. ;; SVE_RELAXED_GP lets us extend the FP operation to cover these ;; extra lanes, but SVE_STRICT_GP does not. ;; ;; Thus SVE_RELAXED_GP allows us to ignore the predicate on the FP operation ;; and operate on exactly the lanes selected by the UNSPEC_SEL predicate. ;; This typically leads to patterns like: ;; ;; (unspec [(match_operand 1 “register_operand” “Upl”) ;; (unspec [(match_operand N) ;; (const_int SVE_RELAXED_GP) ;; ...] ;; UNSPEC_COND) ;; ...]) ;; ;; where operand N is allowed to be anything. These instructions then ;; have rewrite rules to replace operand N with operand 1, which gives the ;; instructions a canonical form and means that the original operand N is ;; not kept live unnecessarily. ;; ;; In contrast, SVE_STRICT_GP only allows the UNSPEC_SEL predicate to be ;; a subset of the FP operation predicate. This case isn't interesting ;; for FP operations that have an all-true predicate, since such operations ;; use SVE_RELAXED_GP instead. And it is not possible for instruction ;; conditions to track the subset relationship for arbitrary registers. ;; So in practice, the only useful case for SVE_STRICT_GP is the one ;; in which the predicates match: ;; ;; (unspec [(match_operand 1 “register_operand” “Upl”) ;; (unspec [(match_dup 1) ;; (const_int SVE_STRICT_GP) ;; ...] ;; UNSPEC_COND) ;; ...]) ;; ;; This pattern would also be correct for SVE_RELAXED_GP, but it would ;; be redundant with the one above. However, if the combine pattern ;; has multiple FP operations, using a match_operand allows combinations ;; of SVE_STRICT_GP and SVE_RELAXED_GP in the same operation, provided ;; that the predicates are the same: ;; ;; (unspec [(match_operand 1 “register_operand” “Upl”) ;; (... ;; (unspec [(match_dup 1) ;; (match_operand:SI N “aarch64_sve_gp_strictness”) ;; ...] ;; UNSPEC_COND) ;; (unspec [(match_dup 1) ;; (match_operand:SI M “aarch64_sve_gp_strictness”) ;; ...] ;; UNSPEC_COND) ...) ;; ...]) ;; ;; The fully-relaxed version of this pattern is: ;; ;; (unspec [(match_operand 1 “register_operand” “Upl”) ;; (... ;; (unspec [(match_operand:SI N) ;; (const_int SVE_RELAXED_GP) ;; ...] ;; UNSPEC_COND) ;; (unspec [(match_operand:SI M) ;; (const_int SVE_RELAXED_GP) ;; ...] ;; UNSPEC_COND) ...) ;; ...]) ;; ;; ------------------------------------------------------------------------- ;; ---- Note on FFR handling ;; ------------------------------------------------------------------------- ;; ;; Logically we want to divide FFR-related instructions into regions ;; that contain exactly one of: ;; ;; - a single write to the FFR ;; - any number of reads from the FFR (but only one read is likely) ;; - any number of LDFF1 and LDNF1 instructions ;; ;; However, LDFF1 and LDNF1 instructions should otherwise behave like ;; normal loads as far as possible. This means that they should be ;; schedulable within a region in the same way that LD1 would be, ;; and they should be deleted as dead if the result is unused. The loads ;; should therefore not write to the FFR, since that would both serialize ;; the loads with respect to each other and keep the loads live for any ;; later RDFFR. ;; ;; We get around this by using a fake “FFR token” (FFRT) to help describe ;; the dependencies. Writing to the FFRT starts a new “FFRT region”, ;; while using the FFRT keeps the instruction within its region. ;; Specifically: ;; ;; - Writes start a new FFRT region as well as setting the FFR: ;; ;; W1: parallel (FFRT = , FFR = ) ;; ;; - Loads use an LD1-like instruction that also uses the FFRT, so that the ;; loads stay within the same FFRT region: ;; ;; L1: load data while using the FFRT ;; ;; In addition, any FFRT region that includes a load also has at least one ;; instance of: ;; ;; L2: FFR = update(FFR, FFRT) [type == no_insn] ;; ;; to make it clear that the region both reads from and writes to the FFR. ;; ;; - Reads do the following: ;; ;; R1: FFRT = FFR [type == no_insn] ;; R2: read from the FFRT ;; R3: FFRT = update(FFRT) [type == no_insn] ;; ;; R1 and R3 both create new FFRT regions, so that previous LDFF1s and ;; LDNF1s cannot move forwards across R1 and later LDFF1s and LDNF1s ;; cannot move backwards across R3. ;; ;; This way, writes are only kept alive by later loads or reads, ;; and write/read pairs fold normally. For two consecutive reads, ;; the first R3 is made dead by the second R1, which in turn becomes ;; redundant with the first R1. We then have: ;; ;; first R1: FFRT = FFR ;; first read from the FFRT ;; second read from the FFRT ;; second R3: FFRT = update(FFRT) ;; ;; i.e. the two FFRT regions collapse into a single one with two ;; independent reads. ;; ;; The model still prevents some valid optimizations though. For example, ;; if all loads in an FFRT region are deleted as dead, nothing would remove ;; the L2 instructions.

;; ========================================================================= ;; == Moves ;; =========================================================================

;; ------------------------------------------------------------------------- ;; ---- Moves of single vectors ;; ------------------------------------------------------------------------- ;; Includes: ;; - MOV (including aliases) ;; - LD1B (contiguous form) ;; - LD1D ( " " ) ;; - LD1H ( " " ) ;; - LD1W ( " " ) ;; - LDR ;; - ST1B (contiguous form) ;; - ST1D ( " " ) ;; - ST1H ( " " ) ;; - ST1W ( " " ) ;; - STR ;; -------------------------------------------------------------------------

(define_expand “mov” [(set (match_operand:SVE_ALL 0 “nonimmediate_operand”) (match_operand:SVE_ALL 1 “general_operand”))] “TARGET_SVE” { /* Use the predicated load and store patterns where possible. This is required for big-endian targets (see the comment at the head of the file) and increases the addressing choices for little-endian. */ if ((MEM_P (operands[0]) || MEM_P (operands[1])) && can_create_pseudo_p ()) { aarch64_expand_sve_mem_move (operands[0], operands[1], mode); DONE; }

if (CONSTANT_P (operands[1]))
  {
aarch64_expand_mov_immediate (operands[0], operands[1]);
DONE;
  }

/* Optimize subregs on big-endian targets: we can use REV[BHW]
   instead of going through memory.  */
if (BYTES_BIG_ENDIAN
&& aarch64_maybe_expand_sve_subreg_move (operands[0], operands[1]))
  DONE;

} )

(define_expand “movmisalign” [(set (match_operand:SVE_ALL 0 “nonimmediate_operand”) (match_operand:SVE_ALL 1 “general_operand”))] “TARGET_SVE” { /* Equivalent to a normal move for our purpooses. */ emit_move_insn (operands[0], operands[1]); DONE; } )

;; Unpredicated moves that can use LDR and STR, i.e. full vectors for which ;; little-endian ordering is acceptable. Only allow memory operations during ;; and after RA; before RA we want the predicated load and store patterns to ;; be used instead. (define_insn “*aarch64_sve_mov_ldr_str” [(set (match_operand:SVE_FULL 0 “aarch64_sve_nonimmediate_operand” “=w, Utr, w, w”) (match_operand:SVE_FULL 1 “aarch64_sve_general_operand” “Utr, w, w, Dn”))] “TARGET_SVE && (mode == VNx16QImode || !BYTES_BIG_ENDIAN) && ((lra_in_progress || reload_completed) || (register_operand (operands[0], mode) && nonmemory_operand (operands[1], mode)))” "@ ldr\t%0, %1 str\t%1, %0 mov\t%0.d, %1.d

  • return aarch64_output_sve_mov_immediate (operands[1]);" )

;; Unpredicated moves that cannot use LDR and STR, i.e. partial vectors ;; or vectors for which little-endian ordering isn't acceptable. Memory ;; accesses require secondary reloads. (define_insn “*aarch64_sve_mov_no_ldr_str” [(set (match_operand:SVE_ALL 0 “register_operand” “=w, w”) (match_operand:SVE_ALL 1 “aarch64_nonmemory_operand” “w, Dn”))] “TARGET_SVE && mode != VNx16QImode && (BYTES_BIG_ENDIAN || maybe_ne (BYTES_PER_SVE_VECTOR, GET_MODE_SIZE (mode)))” "@ mov\t%0.d, %1.d

  • return aarch64_output_sve_mov_immediate (operands[1]);" )

;; Handle memory reloads for modes that can't use LDR and STR. We use ;; byte PTRUE for all modes to try to encourage reuse. This pattern ;; needs constraints because it is returned by TARGET_SECONDARY_RELOAD. (define_expand “aarch64_sve_reload_mem” [(parallel [(set (match_operand 0) (match_operand 1)) (clobber (match_operand:VNx16BI 2 “register_operand” “=Upl”))])] “TARGET_SVE” { /* Create a PTRUE. */ emit_move_insn (operands[2], CONSTM1_RTX (VNx16BImode));

/* Refer to the PTRUE in the appropriate mode for this move.  */
machine_mode mode = GET_MODE (operands[0]);
rtx pred = gen_lowpart (aarch64_sve_pred_mode (mode), operands[2]);

/* Emit a predicated load or store.  */
aarch64_emit_sve_pred_move (operands[0], pred, operands[1]);
DONE;

} )

;; A predicated move in which the predicate is known to be all-true. ;; Note that this pattern is generated directly by aarch64_emit_sve_pred_move, ;; so changes to this pattern will need changes there as well. (define_insn_and_split “@aarch64_pred_mov” [(set (match_operand:SVE_ALL 0 “nonimmediate_operand” “=w, w, m”) (unspec:SVE_ALL [(match_operand: 1 “register_operand” “Upl, Upl, Upl”) (match_operand:SVE_ALL 2 “nonimmediate_operand” “w, m, w”)] UNSPEC_PRED_X))] “TARGET_SVE && (register_operand (operands[0], mode) || register_operand (operands[2], mode))” "@

ld1\t%0., %1/z, %2 st1\t%2., %1, %0" “&& register_operand (operands[0], mode) && register_operand (operands[2], mode)” [(set (match_dup 0) (match_dup 2))] )

;; A pattern for optimizing SUBREGs that have a reinterpreting effect ;; on big-endian targets; see aarch64_maybe_expand_sve_subreg_move ;; for details. We use a special predicate for operand 2 to reduce ;; the number of patterns. (define_insn_and_split “*aarch64_sve_mov_subreg_be” [(set (match_operand:SVE_ALL 0 “aarch64_sve_nonimmediate_operand” “=w”) (unspec:SVE_ALL [(match_operand:VNx16BI 1 “register_operand” “Upl”) (match_operand 2 “aarch64_any_register_operand” “w”)] UNSPEC_REV_SUBREG))] “TARGET_SVE && BYTES_BIG_ENDIAN” “#” “&& reload_completed” [(const_int 0)] { aarch64_split_sve_subreg_move (operands[0], operands[1], operands[2]); DONE; } )

;; Reinterpret operand 1 in operand 0's mode, without changing its contents. ;; This is equivalent to a subreg on little-endian targets but not for ;; big-endian; see the comment at the head of the file for details. (define_expand “@aarch64_sve_reinterpret” [(set (match_operand:SVE_ALL 0 “register_operand”) (unspec:SVE_ALL [(match_operand 1 “aarch64_any_register_operand”)] UNSPEC_REINTERPRET))] “TARGET_SVE” { machine_mode src_mode = GET_MODE (operands[1]); if (targetm.can_change_mode_class (mode, src_mode, FP_REGS)) { emit_move_insn (operands[0], gen_lowpart (mode, operands[1])); DONE; } } )

;; A pattern for handling type punning on big-endian targets. We use a ;; special predicate for operand 1 to reduce the number of patterns. (define_insn_and_split “*aarch64_sve_reinterpret” [(set (match_operand:SVE_ALL 0 “register_operand” “=w”) (unspec:SVE_ALL [(match_operand 1 “aarch64_any_register_operand” “w”)] UNSPEC_REINTERPRET))] “TARGET_SVE” “#” “&& reload_completed” [(set (match_dup 0) (match_dup 1))] { operands[1] = aarch64_replace_reg_mode (operands[1], mode); } )

;; ------------------------------------------------------------------------- ;; ---- Moves of multiple vectors ;; ------------------------------------------------------------------------- ;; All patterns in this section are synthetic and split to real ;; instructions after reload. ;; -------------------------------------------------------------------------

(define_expand “mov” [(set (match_operand:SVE_STRUCT 0 “nonimmediate_operand”) (match_operand:SVE_STRUCT 1 “general_operand”))] “TARGET_SVE” { /* Big-endian loads and stores need to be done via LD1 and ST1; see the comment at the head of the file for details. */ if ((MEM_P (operands[0]) || MEM_P (operands[1])) && BYTES_BIG_ENDIAN) { gcc_assert (can_create_pseudo_p ()); aarch64_expand_sve_mem_move (operands[0], operands[1], mode); DONE; }

if (CONSTANT_P (operands[1]))
  {
aarch64_expand_mov_immediate (operands[0], operands[1]);
DONE;
  }

} )

;; Unpredicated structure moves (little-endian). (define_insn “*aarch64_sve_mov_le” [(set (match_operand:SVE_STRUCT 0 “aarch64_sve_nonimmediate_operand” “=w, Utr, w, w”) (match_operand:SVE_STRUCT 1 “aarch64_sve_general_operand” “Utr, w, w, Dn”))] “TARGET_SVE && !BYTES_BIG_ENDIAN” “#” [(set_attr “length” “<insn_length>”)] )

;; Unpredicated structure moves (big-endian). Memory accesses require ;; secondary reloads. (define_insn “*aarch64_sve_mov_be” [(set (match_operand:SVE_STRUCT 0 “register_operand” “=w, w”) (match_operand:SVE_STRUCT 1 “aarch64_nonmemory_operand” “w, Dn”))] “TARGET_SVE && BYTES_BIG_ENDIAN” “#” [(set_attr “length” “<insn_length>”)] )

;; Split unpredicated structure moves into pieces. This is the same ;; for both big-endian and little-endian code, although it only needs ;; to handle memory operands for little-endian code. (define_split [(set (match_operand:SVE_STRUCT 0 “aarch64_sve_nonimmediate_operand”) (match_operand:SVE_STRUCT 1 “aarch64_sve_general_operand”))] “TARGET_SVE && reload_completed” [(const_int 0)] { rtx dest = operands[0]; rtx src = operands[1]; if (REG_P (dest) && REG_P (src)) aarch64_simd_emit_reg_reg_move (operands, mode, <vector_count>); else for (unsigned int i = 0; i < <vector_count>; ++i) { rtx subdest = simplify_gen_subreg (mode, dest, mode, i * BYTES_PER_SVE_VECTOR); rtx subsrc = simplify_gen_subreg (mode, src, mode, i * BYTES_PER_SVE_VECTOR); emit_insn (gen_rtx_SET (subdest, subsrc)); } DONE; } )

;; Predicated structure moves. This works for both endiannesses but in ;; practice is only useful for big-endian. (define_insn_and_split “@aarch64_pred_mov” [(set (match_operand:SVE_STRUCT 0 “aarch64_sve_struct_nonimmediate_operand” “=w, w, Utx”) (unspec:SVE_STRUCT [(match_operand: 1 “register_operand” “Upl, Upl, Upl”) (match_operand:SVE_STRUCT 2 “aarch64_sve_struct_nonimmediate_operand” “w, Utx, w”)] UNSPEC_PRED_X))] “TARGET_SVE && (register_operand (operands[0], mode) || register_operand (operands[2], mode))” “#” “&& reload_completed” [(const_int 0)] { for (unsigned int i = 0; i < <vector_count>; ++i) { rtx subdest = simplify_gen_subreg (mode, operands[0], mode, i * BYTES_PER_SVE_VECTOR); rtx subsrc = simplify_gen_subreg (mode, operands[2], mode, i * BYTES_PER_SVE_VECTOR); aarch64_emit_sve_pred_move (subdest, operands[1], subsrc); } DONE; } [(set_attr “length” “<insn_length>”)] )

;; ------------------------------------------------------------------------- ;; ---- Moves of predicates ;; ------------------------------------------------------------------------- ;; Includes: ;; - MOV ;; - LDR ;; - PFALSE ;; - PTRUE ;; - PTRUES ;; - STR ;; -------------------------------------------------------------------------

(define_expand “mov” [(set (match_operand:PRED_ALL 0 “nonimmediate_operand”) (match_operand:PRED_ALL 1 “general_operand”))] “TARGET_SVE” { if (GET_CODE (operands[0]) == MEM) operands[1] = force_reg (mode, operands[1]);

if (CONSTANT_P (operands[1]))
  {
aarch64_expand_mov_immediate (operands[0], operands[1]);
DONE;
  }

} )

(define_insn “*aarch64_sve_mov” [(set (match_operand:PRED_ALL 0 “nonimmediate_operand” “=Upa, m, Upa, Upa”) (match_operand:PRED_ALL 1 “aarch64_mov_operand” “Upa, Upa, m, Dn”))] “TARGET_SVE && (register_operand (operands[0], mode) || register_operand (operands[1], mode))” "@ mov\t%0.b, %1.b str\t%1, %0 ldr\t%0, %1

  • return aarch64_output_sve_mov_immediate (operands[1]);" )

;; Match PTRUES Pn.B when both the predicate and flags are useful. (define_insn_and_rewrite “*aarch64_sve_ptruevnx16bi_cc” [(set (reg:CC_NZC CC_REGNUM) (unspec:CC_NZC [(match_operand 2) (match_operand 3) (const_int SVE_KNOWN_PTRUE) (match_operator:VNx16BI 1 “aarch64_sve_ptrue_svpattern_immediate” [(unspec:VNx16BI [(match_operand:SI 4 “const_int_operand”) (match_operand:VNx16BI 5 “aarch64_simd_imm_zero”)] UNSPEC_PTRUE)])] UNSPEC_PTEST)) (set (match_operand:VNx16BI 0 “register_operand” “=Upa”) (match_dup 1))] “TARGET_SVE” { return aarch64_output_sve_ptrues (operands[1]); } “&& (!CONSTANT_P (operands[2]) || !CONSTANT_P (operands[3]))” { operands[2] = operands[3] = CONSTM1_RTX (VNx16BImode); } )

;; Match PTRUES Pn.[HSD] when both the predicate and flags are useful. (define_insn_and_rewrite “*aarch64_sve_ptrue_cc” [(set (reg:CC_NZC CC_REGNUM) (unspec:CC_NZC [(match_operand 2) (match_operand 3) (const_int SVE_KNOWN_PTRUE) (subreg:PRED_HSD (match_operator:VNx16BI 1 “aarch64_sve_ptrue_svpattern_immediate” [(unspec:VNx16BI [(match_operand:SI 4 “const_int_operand”) (match_operand:PRED_HSD 5 “aarch64_simd_imm_zero”)] UNSPEC_PTRUE)]) 0)] UNSPEC_PTEST)) (set (match_operand:VNx16BI 0 “register_operand” “=Upa”) (match_dup 1))] “TARGET_SVE” { return aarch64_output_sve_ptrues (operands[1]); } “&& (!CONSTANT_P (operands[2]) || !CONSTANT_P (operands[3]))” { operands[2] = CONSTM1_RTX (VNx16BImode); operands[3] = CONSTM1_RTX (mode); } )

;; Match PTRUES Pn.B when only the flags result is useful (which is ;; a way of testing VL). (define_insn_and_rewrite “*aarch64_sve_ptruevnx16bi_ptest” [(set (reg:CC_NZC CC_REGNUM) (unspec:CC_NZC [(match_operand 2) (match_operand 3) (const_int SVE_KNOWN_PTRUE) (match_operator:VNx16BI 1 “aarch64_sve_ptrue_svpattern_immediate” [(unspec:VNx16BI [(match_operand:SI 4 “const_int_operand”) (match_operand:VNx16BI 5 “aarch64_simd_imm_zero”)] UNSPEC_PTRUE)])] UNSPEC_PTEST)) (clobber (match_scratch:VNx16BI 0 “=Upa”))] “TARGET_SVE” { return aarch64_output_sve_ptrues (operands[1]); } “&& (!CONSTANT_P (operands[2]) || !CONSTANT_P (operands[3]))” { operands[2] = operands[3] = CONSTM1_RTX (VNx16BImode); } )

;; Match PTRUES Pn.[HWD] when only the flags result is useful (which is ;; a way of testing VL). (define_insn_and_rewrite “*aarch64_sve_ptrue_ptest” [(set (reg:CC_NZC CC_REGNUM) (unspec:CC_NZC [(match_operand 2) (match_operand 3) (const_int SVE_KNOWN_PTRUE) (subreg:PRED_HSD (match_operator:VNx16BI 1 “aarch64_sve_ptrue_svpattern_immediate” [(unspec:VNx16BI [(match_operand:SI 4 “const_int_operand”) (match_operand:PRED_HSD 5 “aarch64_simd_imm_zero”)] UNSPEC_PTRUE)]) 0)] UNSPEC_PTEST)) (clobber (match_scratch:VNx16BI 0 “=Upa”))] “TARGET_SVE” { return aarch64_output_sve_ptrues (operands[1]); } “&& (!CONSTANT_P (operands[2]) || !CONSTANT_P (operands[3]))” { operands[2] = CONSTM1_RTX (VNx16BImode); operands[3] = CONSTM1_RTX (mode); } )

;; ------------------------------------------------------------------------- ;; ---- Moves relating to the FFR ;; ------------------------------------------------------------------------- ;; RDFFR ;; RDFFRS ;; SETFFR ;; WRFFR ;; -------------------------------------------------------------------------

;; [W1 in the block comment above about FFR handling] ;; ;; Write to the FFR and start a new FFRT scheduling region. (define_insn “aarch64_wrffr” [(set (reg:VNx16BI FFR_REGNUM) (match_operand:VNx16BI 0 “aarch64_simd_reg_or_minus_one” “Dm, Upa”)) (set (reg:VNx16BI FFRT_REGNUM) (unspec:VNx16BI [(match_dup 0)] UNSPEC_WRFFR))] “TARGET_SVE” “@ setffr wrffr\t%0.b” )

;; [L2 in the block comment above about FFR handling] ;; ;; Introduce a read from and write to the FFR in the current FFRT region, ;; so that the FFR value is live on entry to the region and so that the FFR ;; value visibly changes within the region. This is used (possibly multiple ;; times) in an FFRT region that includes LDFF1 or LDNF1 instructions. (define_insn “aarch64_update_ffr_for_load” [(set (reg:VNx16BI FFR_REGNUM) (unspec:VNx16BI [(reg:VNx16BI FFRT_REGNUM) (reg:VNx16BI FFR_REGNUM)] UNSPEC_UPDATE_FFR))] “TARGET_SVE” "" [(set_attr “type” “no_insn”)] )

;; [R1 in the block comment above about FFR handling] ;; ;; Notionally copy the FFR to the FFRT, so that the current FFR value ;; can be read from there by the RDFFR instructions below. This acts ;; as a scheduling barrier for earlier LDFF1 and LDNF1 instructions and ;; creates a natural dependency with earlier writes. (define_insn “aarch64_copy_ffr_to_ffrt” [(set (reg:VNx16BI FFRT_REGNUM) (reg:VNx16BI FFR_REGNUM))] “TARGET_SVE” "" [(set_attr “type” “no_insn”)] )

;; [R2 in the block comment above about FFR handling] ;; ;; Read the FFR via the FFRT. (define_insn “aarch64_rdffr” [(set (match_operand:VNx16BI 0 “register_operand” “=Upa”) (reg:VNx16BI FFRT_REGNUM))] “TARGET_SVE” “rdffr\t%0.b” )

;; Likewise with zero predication. (define_insn “aarch64_rdffr_z” [(set (match_operand:VNx16BI 0 “register_operand” “=Upa”) (and:VNx16BI (reg:VNx16BI FFRT_REGNUM) (match_operand:VNx16BI 1 “register_operand” “Upa”)))] “TARGET_SVE” “rdffr\t%0.b, %1/z” )

;; Read the FFR to test for a fault, without using the predicate result. (define_insn “*aarch64_rdffr_z_ptest” [(set (reg:CC_NZC CC_REGNUM) (unspec:CC_NZC [(match_operand:VNx16BI 1 “register_operand” “Upa”) (match_dup 1) (match_operand:SI 2 “aarch64_sve_ptrue_flag”) (and:VNx16BI (reg:VNx16BI FFRT_REGNUM) (match_dup 1))] UNSPEC_PTEST)) (clobber (match_scratch:VNx16BI 0 “=Upa”))] “TARGET_SVE” “rdffrs\t%0.b, %1/z” )

;; Same for unpredicated RDFFR when tested with a known PTRUE. (define_insn “*aarch64_rdffr_ptest” [(set (reg:CC_NZC CC_REGNUM) (unspec:CC_NZC [(match_operand:VNx16BI 1 “register_operand” “Upa”) (match_dup 1) (const_int SVE_KNOWN_PTRUE) (reg:VNx16BI FFRT_REGNUM)] UNSPEC_PTEST)) (clobber (match_scratch:VNx16BI 0 “=Upa”))] “TARGET_SVE” “rdffrs\t%0.b, %1/z” )

;; Read the FFR with zero predication and test the result. (define_insn “*aarch64_rdffr_z_cc” [(set (reg:CC_NZC CC_REGNUM) (unspec:CC_NZC [(match_operand:VNx16BI 1 “register_operand” “Upa”) (match_dup 1) (match_operand:SI 2 “aarch64_sve_ptrue_flag”) (and:VNx16BI (reg:VNx16BI FFRT_REGNUM) (match_dup 1))] UNSPEC_PTEST)) (set (match_operand:VNx16BI 0 “register_operand” “=Upa”) (and:VNx16BI (reg:VNx16BI FFRT_REGNUM) (match_dup 1)))] “TARGET_SVE” “rdffrs\t%0.b, %1/z” )

;; Same for unpredicated RDFFR when tested with a known PTRUE. (define_insn “*aarch64_rdffr_cc” [(set (reg:CC_NZC CC_REGNUM) (unspec:CC_NZC [(match_operand:VNx16BI 1 “register_operand” “Upa”) (match_dup 1) (const_int SVE_KNOWN_PTRUE) (reg:VNx16BI FFRT_REGNUM)] UNSPEC_PTEST)) (set (match_operand:VNx16BI 0 “register_operand” “=Upa”) (reg:VNx16BI FFRT_REGNUM))] “TARGET_SVE” “rdffrs\t%0.b, %1/z” )

;; [R3 in the block comment above about FFR handling] ;; ;; Arbitrarily update the FFRT after a read from the FFR. This acts as ;; a scheduling barrier for later LDFF1 and LDNF1 instructions. (define_insn “aarch64_update_ffrt” [(set (reg:VNx16BI FFRT_REGNUM) (unspec:VNx16BI [(reg:VNx16BI FFRT_REGNUM)] UNSPEC_UPDATE_FFRT))] “TARGET_SVE” "" [(set_attr “type” “no_insn”)] )

;; ========================================================================= ;; == Loads ;; =========================================================================

;; ------------------------------------------------------------------------- ;; ---- Normal contiguous loads ;; ------------------------------------------------------------------------- ;; Includes contiguous forms of: ;; - LD1B ;; - LD1D ;; - LD1H ;; - LD1W ;; - LD2B ;; - LD2D ;; - LD2H ;; - LD2W ;; - LD3B ;; - LD3D ;; - LD3H ;; - LD3W ;; - LD4B ;; - LD4D ;; - LD4H ;; - LD4W ;; -------------------------------------------------------------------------

;; Predicated LD1. (define_insn “maskload” [(set (match_operand:SVE_ALL 0 “register_operand” “=w”) (unspec:SVE_ALL [(match_operand: 2 “register_operand” “Upl”) (match_operand:SVE_ALL 1 “memory_operand” “m”)] UNSPEC_LD1_SVE))] “TARGET_SVE” “ld1\t%0., %2/z, %1” )

;; Unpredicated LD[234]. (define_expand “vec_load_lanes” [(set (match_operand:SVE_STRUCT 0 “register_operand”) (unspec:SVE_STRUCT [(match_dup 2) (match_operand:SVE_STRUCT 1 “memory_operand”)] UNSPEC_LDN))] “TARGET_SVE” { operands[2] = aarch64_ptrue_reg (mode); } )

;; Predicated LD[234]. (define_insn “vec_mask_load_lanes” [(set (match_operand:SVE_STRUCT 0 “register_operand” “=w”) (unspec:SVE_STRUCT [(match_operand: 2 “register_operand” “Upl”) (match_operand:SVE_STRUCT 1 “memory_operand” “m”)] UNSPEC_LDN))] “TARGET_SVE” “ld<vector_count>\t%0, %2/z, %1” )

;; ------------------------------------------------------------------------- ;; ---- Extending contiguous loads ;; ------------------------------------------------------------------------- ;; Includes contiguous forms of: ;; LD1B ;; LD1H ;; LD1SB ;; LD1SH ;; LD1SW ;; LD1W ;; -------------------------------------------------------------------------

;; Predicated load and extend, with 8 elements per 128-bit block. (define_insn_and_rewrite “@aarch64_load<SVE_PRED_LOAD:pred_load>_<ANY_EXTEND:optab><SVE_HSDI:mode><SVE_PARTIAL_I:mode>” [(set (match_operand:SVE_HSDI 0 “register_operand” “=w”) (unspec:SVE_HSDI [(match_operand:<SVE_HSDI:VPRED> 3 “general_operand” “UplDnm”) (ANY_EXTEND:SVE_HSDI (unspec:SVE_PARTIAL_I [(match_operand:<SVE_PARTIAL_I:VPRED> 2 “register_operand” “Upl”) (match_operand:SVE_PARTIAL_I 1 “memory_operand” “m”)] SVE_PRED_LOAD))] UNSPEC_PRED_X))] “TARGET_SVE && (~<SVE_HSDI:narrower_mask> & <SVE_PARTIAL_I:self_mask>) == 0” “ld1<ANY_EXTEND:s><SVE_PARTIAL_I:Vesize>\t%0.<SVE_HSDI:Vctype>, %2/z, %1” “&& !CONSTANT_P (operands[3])” { operands[3] = CONSTM1_RTX (<SVE_HSDI:VPRED>mode); } )

;; ------------------------------------------------------------------------- ;; ---- First-faulting contiguous loads ;; ------------------------------------------------------------------------- ;; Includes contiguous forms of: ;; - LDFF1B ;; - LDFF1D ;; - LDFF1H ;; - LDFF1W ;; - LDNF1B ;; - LDNF1D ;; - LDNF1H ;; - LDNF1W ;; -------------------------------------------------------------------------

;; Contiguous non-extending first-faulting or non-faulting loads. (define_insn “@aarch64_ldf1” [(set (match_operand:SVE_FULL 0 “register_operand” “=w”) (unspec:SVE_FULL [(match_operand: 2 “register_operand” “Upl”) (match_operand:SVE_FULL 1 “aarch64_sve_ldf1_operand” “Ut”) (reg:VNx16BI FFRT_REGNUM)] SVE_LDFF1_LDNF1))] “TARGET_SVE” “ldf1\t%0., %2/z, %1” )

;; ------------------------------------------------------------------------- ;; ---- First-faulting extending contiguous loads ;; ------------------------------------------------------------------------- ;; Includes contiguous forms of: ;; - LDFF1B ;; - LDFF1H ;; - LDFF1SB ;; - LDFF1SH ;; - LDFF1SW ;; - LDFF1W ;; - LDNF1B ;; - LDNF1H ;; - LDNF1SB ;; - LDNF1SH ;; - LDNF1SW ;; - LDNF1W ;; -------------------------------------------------------------------------

;; Predicated first-faulting or non-faulting load and extend. (define_insn_and_rewrite “@aarch64_ldf1_<ANY_EXTEND:optab><SVE_HSDI:mode><SVE_PARTIAL_I:mode>” [(set (match_operand:SVE_HSDI 0 “register_operand” “=w”) (unspec:SVE_HSDI [(match_operand:<SVE_HSDI:VPRED> 3 “general_operand” “UplDnm”) (ANY_EXTEND:SVE_HSDI (unspec:SVE_PARTIAL_I [(match_operand:<SVE_PARTIAL_I:VPRED> 2 “register_operand” “Upl”) (match_operand:SVE_PARTIAL_I 1 “aarch64_sve_ldf1_operand” “Ut”) (reg:VNx16BI FFRT_REGNUM)] SVE_LDFF1_LDNF1))] UNSPEC_PRED_X))] “TARGET_SVE && (~<SVE_HSDI:narrower_mask> & <SVE_PARTIAL_I:self_mask>) == 0” “ldf1<ANY_EXTEND:s><SVE_PARTIAL_I:Vesize>\t%0.<SVE_HSDI:Vctype>, %2/z, %1” “&& !CONSTANT_P (operands[3])” { operands[3] = CONSTM1_RTX (<SVE_HSDI:VPRED>mode); } )

;; ------------------------------------------------------------------------- ;; ---- Non-temporal contiguous loads ;; ------------------------------------------------------------------------- ;; Includes: ;; - LDNT1B ;; - LDNT1D ;; - LDNT1H ;; - LDNT1W ;; -------------------------------------------------------------------------

;; Predicated contiguous non-temporal load. (define_insn “@aarch64_ldnt1” [(set (match_operand:SVE_FULL 0 “register_operand” “=w”) (unspec:SVE_FULL [(match_operand: 2 “register_operand” “Upl”) (match_operand:SVE_FULL 1 “memory_operand” “m”)] UNSPEC_LDNT1_SVE))] “TARGET_SVE” “ldnt1\t%0., %2/z, %1” )

;; ------------------------------------------------------------------------- ;; ---- Normal gather loads ;; ------------------------------------------------------------------------- ;; Includes gather forms of: ;; - LD1D ;; - LD1W ;; -------------------------------------------------------------------------

;; Unpredicated gather loads. (define_expand “gather_load<v_int_container>” [(set (match_operand:SVE_24 0 “register_operand”) (unspec:SVE_24 [(match_dup 5) (match_operand:DI 1 “aarch64_sve_gather_offset_”) (match_operand:<V_INT_CONTAINER> 2 “register_operand”) (match_operand:DI 3 “const_int_operand”) (match_operand:DI 4 “aarch64_gather_scale_operand_”) (mem:BLK (scratch))] UNSPEC_LD1_GATHER))] “TARGET_SVE” { operands[5] = aarch64_ptrue_reg (mode); } )

;; Predicated gather loads for 32-bit elements. Operand 3 is true for ;; unsigned extension and false for signed extension. (define_insn “mask_gather_load<v_int_container>” [(set (match_operand:SVE_4 0 “register_operand” “=w, w, w, w, w, w”) (unspec:SVE_4 [(match_operand:VNx4BI 5 “register_operand” “Upl, Upl, Upl, Upl, Upl, Upl”) (match_operand:DI 1 “aarch64_sve_gather_offset_” “Z, vgw, rk, rk, rk, rk”) (match_operand:VNx4SI 2 “register_operand” “w, w, w, w, w, w”) (match_operand:DI 3 “const_int_operand” “Ui1, Ui1, Z, Ui1, Z, Ui1”) (match_operand:DI 4 “aarch64_gather_scale_operand_” “Ui1, Ui1, Ui1, Ui1, i, i”) (mem:BLK (scratch))] UNSPEC_LD1_GATHER))] “TARGET_SVE” “@ ld1\t%0.s, %5/z, [%2.s] ld1\t%0.s, %5/z, [%2.s, #%1] ld1\t%0.s, %5/z, [%1, %2.s, sxtw] ld1\t%0.s, %5/z, [%1, %2.s, uxtw] ld1\t%0.s, %5/z, [%1, %2.s, sxtw %p4] ld1\t%0.s, %5/z, [%1, %2.s, uxtw %p4]” )

;; Predicated gather loads for 64-bit elements. The value of operand 3 ;; doesn't matter in this case. (define_insn “mask_gather_load<v_int_container>” [(set (match_operand:SVE_2 0 “register_operand” “=w, w, w, w”) (unspec:SVE_2 [(match_operand:VNx2BI 5 “register_operand” “Upl, Upl, Upl, Upl”) (match_operand:DI 1 “aarch64_sve_gather_offset_” “Z, vgd, rk, rk”) (match_operand:VNx2DI 2 “register_operand” “w, w, w, w”) (match_operand:DI 3 “const_int_operand”) (match_operand:DI 4 “aarch64_gather_scale_operand_” “Ui1, Ui1, Ui1, i”) (mem:BLK (scratch))] UNSPEC_LD1_GATHER))] “TARGET_SVE” “@ ld1\t%0.d, %5/z, [%2.d] ld1\t%0.d, %5/z, [%2.d, #%1] ld1\t%0.d, %5/z, [%1, %2.d] ld1\t%0.d, %5/z, [%1, %2.d, lsl %p4]” )

;; Likewise, but with the offset being extended from 32 bits. (define_insn_and_rewrite “*mask_gather_load<v_int_container>_xtw_unpacked” [(set (match_operand:SVE_2 0 “register_operand” “=w, w”) (unspec:SVE_2 [(match_operand:VNx2BI 5 “register_operand” “Upl, Upl”) (match_operand:DI 1 “register_operand” “rk, rk”) (unspec:VNx2DI [(match_operand 6) (ANY_EXTEND:VNx2DI (match_operand:VNx2SI 2 “register_operand” “w, w”))] UNSPEC_PRED_X) (match_operand:DI 3 “const_int_operand”) (match_operand:DI 4 “aarch64_gather_scale_operand_” “Ui1, i”) (mem:BLK (scratch))] UNSPEC_LD1_GATHER))] “TARGET_SVE” “@ ld1\t%0.d, %5/z, [%1, %2.d, xtw] ld1\t%0.d, %5/z, [%1, %2.d, xtw %p4]” “&& !CONSTANT_P (operands[6])” { operands[6] = CONSTM1_RTX (VNx2BImode); } )

;; Likewise, but with the offset being truncated to 32 bits and then ;; sign-extended. (define_insn_and_rewrite “*mask_gather_load<v_int_container>_sxtw” [(set (match_operand:SVE_2 0 “register_operand” “=w, w”) (unspec:SVE_2 [(match_operand:VNx2BI 5 “register_operand” “Upl, Upl”) (match_operand:DI 1 “register_operand” “rk, rk”) (unspec:VNx2DI [(match_operand 6) (sign_extend:VNx2DI (truncate:VNx2SI (match_operand:VNx2DI 2 “register_operand” “w, w”)))] UNSPEC_PRED_X) (match_operand:DI 3 “const_int_operand”) (match_operand:DI 4 “aarch64_gather_scale_operand_” “Ui1, i”) (mem:BLK (scratch))] UNSPEC_LD1_GATHER))] “TARGET_SVE” “@ ld1\t%0.d, %5/z, [%1, %2.d, sxtw] ld1\t%0.d, %5/z, [%1, %2.d, sxtw %p4]” “&& !CONSTANT_P (operands[6])” { operands[6] = CONSTM1_RTX (VNx2BImode); } )

;; Likewise, but with the offset being truncated to 32 bits and then ;; zero-extended. (define_insn “*mask_gather_load<v_int_container>_uxtw” [(set (match_operand:SVE_2 0 “register_operand” “=w, w”) (unspec:SVE_2 [(match_operand:VNx2BI 5 “register_operand” “Upl, Upl”) (match_operand:DI 1 “register_operand” “rk, rk”) (and:VNx2DI (match_operand:VNx2DI 2 “register_operand” “w, w”) (match_operand:VNx2DI 6 “aarch64_sve_uxtw_immediate”)) (match_operand:DI 3 “const_int_operand”) (match_operand:DI 4 “aarch64_gather_scale_operand_” “Ui1, i”) (mem:BLK (scratch))] UNSPEC_LD1_GATHER))] “TARGET_SVE” “@ ld1\t%0.d, %5/z, [%1, %2.d, uxtw] ld1\t%0.d, %5/z, [%1, %2.d, uxtw %p4]” )

;; ------------------------------------------------------------------------- ;; ---- Extending gather loads ;; ------------------------------------------------------------------------- ;; Includes gather forms of: ;; - LD1B ;; - LD1H ;; - LD1SB ;; - LD1SH ;; - LD1SW ;; - LD1W ;; -------------------------------------------------------------------------

;; Predicated extending gather loads for 32-bit elements. Operand 3 is ;; true for unsigned extension and false for signed extension. (define_insn_and_rewrite “@aarch64_gather_load_<ANY_EXTEND:optab><SVE_4HSI:mode><SVE_4BHI:mode>” [(set (match_operand:SVE_4HSI 0 “register_operand” “=w, w, w, w, w, w”) (unspec:SVE_4HSI [(match_operand:VNx4BI 6 “general_operand” “UplDnm, UplDnm, UplDnm, UplDnm, UplDnm, UplDnm”) (ANY_EXTEND:SVE_4HSI (unspec:SVE_4BHI [(match_operand:VNx4BI 5 “register_operand” “Upl, Upl, Upl, Upl, Upl, Upl”) (match_operand:DI 1 “aarch64_sve_gather_offset_<SVE_4BHI:Vesize>” “Z, vg<SVE_4BHI:Vesize>, rk, rk, rk, rk”) (match_operand:VNx4SI 2 “register_operand” “w, w, w, w, w, w”) (match_operand:DI 3 “const_int_operand” “Ui1, Ui1, Z, Ui1, Z, Ui1”) (match_operand:DI 4 “aarch64_gather_scale_operand_<SVE_4BHI:Vesize>” “Ui1, Ui1, Ui1, Ui1, i, i”) (mem:BLK (scratch))] UNSPEC_LD1_GATHER))] UNSPEC_PRED_X))] “TARGET_SVE && (~<SVE_4HSI:narrower_mask> & <SVE_4BHI:self_mask>) == 0” “@ ld1<ANY_EXTEND:s><SVE_4BHI:Vesize>\t%0.s, %5/z, [%2.s] ld1<ANY_EXTEND:s><SVE_4BHI:Vesize>\t%0.s, %5/z, [%2.s, #%1] ld1<ANY_EXTEND:s><SVE_4BHI:Vesize>\t%0.s, %5/z, [%1, %2.s, sxtw] ld1<ANY_EXTEND:s><SVE_4BHI:Vesize>\t%0.s, %5/z, [%1, %2.s, uxtw] ld1<ANY_EXTEND:s><SVE_4BHI:Vesize>\t%0.s, %5/z, [%1, %2.s, sxtw %p4] ld1<ANY_EXTEND:s><SVE_4BHI:Vesize>\t%0.s, %5/z, [%1, %2.s, uxtw %p4]” “&& !CONSTANT_P (operands[6])” { operands[6] = CONSTM1_RTX (VNx4BImode); } )

;; Predicated extending gather loads for 64-bit elements. The value of ;; operand 3 doesn't matter in this case. (define_insn_and_rewrite “@aarch64_gather_load_<ANY_EXTEND:optab><SVE_2HSDI:mode><SVE_2BHSI:mode>” [(set (match_operand:SVE_2HSDI 0 “register_operand” “=w, w, w, w”) (unspec:SVE_2HSDI [(match_operand:VNx2BI 6 “general_operand” “UplDnm, UplDnm, UplDnm, UplDnm”) (ANY_EXTEND:SVE_2HSDI (unspec:SVE_2BHSI [(match_operand:VNx2BI 5 “register_operand” “Upl, Upl, Upl, Upl”) (match_operand:DI 1 “aarch64_sve_gather_offset_<SVE_2BHSI:Vesize>” “Z, vg<SVE_2BHSI:Vesize>, rk, rk”) (match_operand:VNx2DI 2 “register_operand” “w, w, w, w”) (match_operand:DI 3 “const_int_operand”) (match_operand:DI 4 “aarch64_gather_scale_operand_<SVE_2BHSI:Vesize>” “Ui1, Ui1, Ui1, i”) (mem:BLK (scratch))] UNSPEC_LD1_GATHER))] UNSPEC_PRED_X))] “TARGET_SVE && (~<SVE_2HSDI:narrower_mask> & <SVE_2BHSI:self_mask>) == 0” “@ ld1<ANY_EXTEND:s><SVE_2BHSI:Vesize>\t%0.d, %5/z, [%2.d] ld1<ANY_EXTEND:s><SVE_2BHSI:Vesize>\t%0.d, %5/z, [%2.d, #%1] ld1<ANY_EXTEND:s><SVE_2BHSI:Vesize>\t%0.d, %5/z, [%1, %2.d] ld1<ANY_EXTEND:s><SVE_2BHSI:Vesize>\t%0.d, %5/z, [%1, %2.d, lsl %p4]” “&& !CONSTANT_P (operands[6])” { operands[6] = CONSTM1_RTX (VNx2BImode); } )

;; Likewise, but with the offset being extended from 32 bits. (define_insn_and_rewrite “*aarch64_gather_load_<ANY_EXTEND:optab><SVE_2HSDI:mode><SVE_2BHSI:mode>_<ANY_EXTEND2:su>xtw_unpacked” [(set (match_operand:SVE_2HSDI 0 “register_operand” “=w, w”) (unspec:SVE_2HSDI [(match_operand 6) (ANY_EXTEND:SVE_2HSDI (unspec:SVE_2BHSI [(match_operand:VNx2BI 5 “register_operand” “Upl, Upl”) (match_operand:DI 1 “aarch64_reg_or_zero” “rk, rk”) (unspec:VNx2DI [(match_operand 7) (ANY_EXTEND2:VNx2DI (match_operand:VNx2SI 2 “register_operand” “w, w”))] UNSPEC_PRED_X) (match_operand:DI 3 “const_int_operand”) (match_operand:DI 4 “aarch64_gather_scale_operand_<SVE_2BHSI:Vesize>” “Ui1, i”) (mem:BLK (scratch))] UNSPEC_LD1_GATHER))] UNSPEC_PRED_X))] “TARGET_SVE && (~<SVE_2HSDI:narrower_mask> & <SVE_2BHSI:self_mask>) == 0” “@ ld1<ANY_EXTEND:s><SVE_2BHSI:Vesize>\t%0.d, %5/z, [%1, %2.d, <ANY_EXTEND2:su>xtw] ld1<ANY_EXTEND:s><SVE_2BHSI:Vesize>\t%0.d, %5/z, [%1, %2.d, <ANY_EXTEND2:su>xtw %p4]” “&& (!CONSTANT_P (operands[6]) || !CONSTANT_P (operands[7]))” { operands[6] = CONSTM1_RTX (VNx2BImode); operands[7] = CONSTM1_RTX (VNx2BImode); } )

;; Likewise, but with the offset being truncated to 32 bits and then ;; sign-extended. (define_insn_and_rewrite “*aarch64_gather_load_<ANY_EXTEND:optab><SVE_2HSDI:mode><SVE_2BHSI:mode>_sxtw” [(set (match_operand:SVE_2HSDI 0 “register_operand” “=w, w”) (unspec:SVE_2HSDI [(match_operand 6) (ANY_EXTEND:SVE_2HSDI (unspec:SVE_2BHSI [(match_operand:VNx2BI 5 “register_operand” “Upl, Upl”) (match_operand:DI 1 “aarch64_reg_or_zero” “rk, rk”) (unspec:VNx2DI [(match_operand 7) (sign_extend:VNx2DI (truncate:VNx2SI (match_operand:VNx2DI 2 “register_operand” “w, w”)))] UNSPEC_PRED_X) (match_operand:DI 3 “const_int_operand”) (match_operand:DI 4 “aarch64_gather_scale_operand_<SVE_2BHSI:Vesize>” “Ui1, i”) (mem:BLK (scratch))] UNSPEC_LD1_GATHER))] UNSPEC_PRED_X))] “TARGET_SVE && (~<SVE_2HSDI:narrower_mask> & <SVE_2BHSI:self_mask>) == 0” “@ ld1<ANY_EXTEND:s><SVE_2BHSI:Vesize>\t%0.d, %5/z, [%1, %2.d, sxtw] ld1<ANY_EXTEND:s><SVE_2BHSI:Vesize>\t%0.d, %5/z, [%1, %2.d, sxtw %p4]” “&& (!CONSTANT_P (operands[6]) || !CONSTANT_P (operands[7]))” { operands[6] = CONSTM1_RTX (VNx2BImode); operands[7] = CONSTM1_RTX (VNx2BImode); } )

;; Likewise, but with the offset being truncated to 32 bits and then ;; zero-extended. (define_insn_and_rewrite “*aarch64_gather_load_<ANY_EXTEND:optab><SVE_2HSDI:mode><SVE_2BHSI:mode>_uxtw” [(set (match_operand:SVE_2HSDI 0 “register_operand” “=w, w”) (unspec:SVE_2HSDI [(match_operand 7) (ANY_EXTEND:SVE_2HSDI (unspec:SVE_2BHSI [(match_operand:VNx2BI 5 “register_operand” “Upl, Upl”) (match_operand:DI 1 “aarch64_reg_or_zero” “rk, rk”) (and:VNx2DI (match_operand:VNx2DI 2 “register_operand” “w, w”) (match_operand:VNx2DI 6 “aarch64_sve_uxtw_immediate”)) (match_operand:DI 3 “const_int_operand”) (match_operand:DI 4 “aarch64_gather_scale_operand_<SVE_2BHSI:Vesize>” “Ui1, i”) (mem:BLK (scratch))] UNSPEC_LD1_GATHER))] UNSPEC_PRED_X))] “TARGET_SVE && (~<SVE_2HSDI:narrower_mask> & <SVE_2BHSI:self_mask>) == 0” “@ ld1<ANY_EXTEND:s><SVE_2BHSI:Vesize>\t%0.d, %5/z, [%1, %2.d, uxtw] ld1<ANY_EXTEND:s><SVE_2BHSI:Vesize>\t%0.d, %5/z, [%1, %2.d, uxtw %p4]” “&& !CONSTANT_P (operands[7])” { operands[7] = CONSTM1_RTX (VNx2BImode); } )

;; ------------------------------------------------------------------------- ;; ---- First-faulting gather loads ;; ------------------------------------------------------------------------- ;; Includes gather forms of: ;; - LDFF1D ;; - LDFF1W ;; -------------------------------------------------------------------------

;; Predicated first-faulting gather loads for 32-bit elements. Operand ;; 3 is true for unsigned extension and false for signed extension. (define_insn “@aarch64_ldff1_gather” [(set (match_operand:SVE_FULL_S 0 “register_operand” “=w, w, w, w, w, w”) (unspec:SVE_FULL_S [(match_operand:VNx4BI 5 “register_operand” “Upl, Upl, Upl, Upl, Upl, Upl”) (match_operand:DI 1 “aarch64_sve_gather_offset_w” “Z, vgw, rk, rk, rk, rk”) (match_operand:VNx4SI 2 “register_operand” “w, w, w, w, w, w”) (match_operand:DI 3 “const_int_operand” “i, i, Z, Ui1, Z, Ui1”) (match_operand:DI 4 “aarch64_gather_scale_operand_w” “Ui1, Ui1, Ui1, Ui1, i, i”) (mem:BLK (scratch)) (reg:VNx16BI FFRT_REGNUM)] UNSPEC_LDFF1_GATHER))] “TARGET_SVE” “@ ldff1w\t%0.s, %5/z, [%2.s] ldff1w\t%0.s, %5/z, [%2.s, #%1] ldff1w\t%0.s, %5/z, [%1, %2.s, sxtw] ldff1w\t%0.s, %5/z, [%1, %2.s, uxtw] ldff1w\t%0.s, %5/z, [%1, %2.s, sxtw %p4] ldff1w\t%0.s, %5/z, [%1, %2.s, uxtw %p4]” )

;; Predicated first-faulting gather loads for 64-bit elements. The value ;; of operand 3 doesn't matter in this case. (define_insn “@aarch64_ldff1_gather” [(set (match_operand:SVE_FULL_D 0 “register_operand” “=w, w, w, w”) (unspec:SVE_FULL_D [(match_operand:VNx2BI 5 “register_operand” “Upl, Upl, Upl, Upl”) (match_operand:DI 1 “aarch64_sve_gather_offset_d” “Z, vgd, rk, rk”) (match_operand:VNx2DI 2 “register_operand” “w, w, w, w”) (match_operand:DI 3 “const_int_operand”) (match_operand:DI 4 “aarch64_gather_scale_operand_d” “Ui1, Ui1, Ui1, i”) (mem:BLK (scratch)) (reg:VNx16BI FFRT_REGNUM)] UNSPEC_LDFF1_GATHER))] “TARGET_SVE” “@ ldff1d\t%0.d, %5/z, [%2.d] ldff1d\t%0.d, %5/z, [%2.d, #%1] ldff1d\t%0.d, %5/z, [%1, %2.d] ldff1d\t%0.d, %5/z, [%1, %2.d, lsl %p4]” )

;; Likewise, but with the offset being sign-extended from 32 bits. (define_insn_and_rewrite “*aarch64_ldff1_gather_sxtw” [(set (match_operand:SVE_FULL_D 0 “register_operand” “=w, w”) (unspec:SVE_FULL_D [(match_operand:VNx2BI 5 “register_operand” “Upl, Upl”) (match_operand:DI 1 “register_operand” “rk, rk”) (unspec:VNx2DI [(match_operand 6) (sign_extend:VNx2DI (truncate:VNx2SI (match_operand:VNx2DI 2 “register_operand” “w, w”)))] UNSPEC_PRED_X) (match_operand:DI 3 “const_int_operand”) (match_operand:DI 4 “aarch64_gather_scale_operand_d” “Ui1, i”) (mem:BLK (scratch)) (reg:VNx16BI FFRT_REGNUM)] UNSPEC_LDFF1_GATHER))] “TARGET_SVE” “@ ldff1d\t%0.d, %5/z, [%1, %2.d, sxtw] ldff1d\t%0.d, %5/z, [%1, %2.d, sxtw %p4]” “&& !CONSTANT_P (operands[6])” { operands[6] = CONSTM1_RTX (VNx2BImode); } )

;; Likewise, but with the offset being zero-extended from 32 bits. (define_insn “*aarch64_ldff1_gather_uxtw” [(set (match_operand:SVE_FULL_D 0 “register_operand” “=w, w”) (unspec:SVE_FULL_D [(match_operand:VNx2BI 5 “register_operand” “Upl, Upl”) (match_operand:DI 1 “register_operand” “rk, rk”) (and:VNx2DI (match_operand:VNx2DI 2 “register_operand” “w, w”) (match_operand:VNx2DI 6 “aarch64_sve_uxtw_immediate”)) (match_operand:DI 3 “const_int_operand”) (match_operand:DI 4 “aarch64_gather_scale_operand_d” “Ui1, i”) (mem:BLK (scratch)) (reg:VNx16BI FFRT_REGNUM)] UNSPEC_LDFF1_GATHER))] “TARGET_SVE” “@ ldff1d\t%0.d, %5/z, [%1, %2.d, uxtw] ldff1d\t%0.d, %5/z, [%1, %2.d, uxtw %p4]” )

;; ------------------------------------------------------------------------- ;; ---- First-faulting extending gather loads ;; ------------------------------------------------------------------------- ;; Includes gather forms of: ;; - LDFF1B ;; - LDFF1H ;; - LDFF1SB ;; - LDFF1SH ;; - LDFF1SW ;; - LDFF1W ;; -------------------------------------------------------------------------

;; Predicated extending first-faulting gather loads for 32-bit elements. ;; Operand 3 is true for unsigned extension and false for signed extension. (define_insn_and_rewrite “@aarch64_ldff1_gather_<ANY_EXTEND:optab><VNx4_WIDE:mode><VNx4_NARROW:mode>” [(set (match_operand:VNx4_WIDE 0 “register_operand” “=w, w, w, w, w, w”) (unspec:VNx4_WIDE [(match_operand:VNx4BI 6 “general_operand” “UplDnm, UplDnm, UplDnm, UplDnm, UplDnm, UplDnm”) (ANY_EXTEND:VNx4_WIDE (unspec:VNx4_NARROW [(match_operand:VNx4BI 5 “register_operand” “Upl, Upl, Upl, Upl, Upl, Upl”) (match_operand:DI 1 “aarch64_sve_gather_offset_<VNx4_NARROW:Vesize>” “Z, vg<VNx4_NARROW:Vesize>, rk, rk, rk, rk”) (match_operand:VNx4_WIDE 2 “register_operand” “w, w, w, w, w, w”) (match_operand:DI 3 “const_int_operand” “i, i, Z, Ui1, Z, Ui1”) (match_operand:DI 4 “aarch64_gather_scale_operand_<VNx4_NARROW:Vesize>” “Ui1, Ui1, Ui1, Ui1, i, i”) (mem:BLK (scratch)) (reg:VNx16BI FFRT_REGNUM)] UNSPEC_LDFF1_GATHER))] UNSPEC_PRED_X))] “TARGET_SVE” “@ ldff1<ANY_EXTEND:s><VNx4_NARROW:Vesize>\t%0.s, %5/z, [%2.s] ldff1<ANY_EXTEND:s><VNx4_NARROW:Vesize>\t%0.s, %5/z, [%2.s, #%1] ldff1<ANY_EXTEND:s><VNx4_NARROW:Vesize>\t%0.s, %5/z, [%1, %2.s, sxtw] ldff1<ANY_EXTEND:s><VNx4_NARROW:Vesize>\t%0.s, %5/z, [%1, %2.s, uxtw] ldff1<ANY_EXTEND:s><VNx4_NARROW:Vesize>\t%0.s, %5/z, [%1, %2.s, sxtw %p4] ldff1<ANY_EXTEND:s><VNx4_NARROW:Vesize>\t%0.s, %5/z, [%1, %2.s, uxtw %p4]” “&& !CONSTANT_P (operands[6])” { operands[6] = CONSTM1_RTX (VNx4BImode); } )

;; Predicated extending first-faulting gather loads for 64-bit elements. ;; The value of operand 3 doesn't matter in this case. (define_insn_and_rewrite “@aarch64_ldff1_gather_<ANY_EXTEND:optab><VNx2_WIDE:mode><VNx2_NARROW:mode>” [(set (match_operand:VNx2_WIDE 0 “register_operand” “=w, w, w, w”) (unspec:VNx2_WIDE [(match_operand:VNx2BI 6 “general_operand” “UplDnm, UplDnm, UplDnm, UplDnm”) (ANY_EXTEND:VNx2_WIDE (unspec:VNx2_NARROW [(match_operand:VNx2BI 5 “register_operand” “Upl, Upl, Upl, Upl”) (match_operand:DI 1 “aarch64_sve_gather_offset_<VNx2_NARROW:Vesize>” “Z, vg<VNx2_NARROW:Vesize>, rk, rk”) (match_operand:VNx2_WIDE 2 “register_operand” “w, w, w, w”) (match_operand:DI 3 “const_int_operand”) (match_operand:DI 4 “aarch64_gather_scale_operand_<VNx2_NARROW:Vesize>” “Ui1, Ui1, Ui1, i”) (mem:BLK (scratch)) (reg:VNx16BI FFRT_REGNUM)] UNSPEC_LDFF1_GATHER))] UNSPEC_PRED_X))] “TARGET_SVE” “@ ldff1<ANY_EXTEND:s><VNx2_NARROW:Vesize>\t%0.d, %5/z, [%2.d] ldff1<ANY_EXTEND:s><VNx2_NARROW:Vesize>\t%0.d, %5/z, [%2.d, #%1] ldff1<ANY_EXTEND:s><VNx2_NARROW:Vesize>\t%0.d, %5/z, [%1, %2.d] ldff1<ANY_EXTEND:s><VNx2_NARROW:Vesize>\t%0.d, %5/z, [%1, %2.d, lsl %p4]” “&& !CONSTANT_P (operands[6])” { operands[6] = CONSTM1_RTX (VNx2BImode); } )

;; Likewise, but with the offset being sign-extended from 32 bits. (define_insn_and_rewrite “*aarch64_ldff1_gather_<ANY_EXTEND:optab><VNx2_WIDE:mode><VNx2_NARROW:mode>_sxtw” [(set (match_operand:VNx2_WIDE 0 “register_operand” “=w, w”) (unspec:VNx2_WIDE [(match_operand 6) (ANY_EXTEND:VNx2_WIDE (unspec:VNx2_NARROW [(match_operand:VNx2BI 5 “register_operand” “Upl, Upl”) (match_operand:DI 1 “aarch64_reg_or_zero” “rk, rk”) (unspec:VNx2DI [(match_operand 7) (sign_extend:VNx2DI (truncate:VNx2SI (match_operand:VNx2DI 2 “register_operand” “w, w”)))] UNSPEC_PRED_X) (match_operand:DI 3 “const_int_operand”) (match_operand:DI 4 “aarch64_gather_scale_operand_<VNx2_NARROW:Vesize>” “Ui1, i”) (mem:BLK (scratch)) (reg:VNx16BI FFRT_REGNUM)] UNSPEC_LDFF1_GATHER))] UNSPEC_PRED_X))] “TARGET_SVE” “@ ldff1<ANY_EXTEND:s><VNx2_NARROW:Vesize>\t%0.d, %5/z, [%1, %2.d, sxtw] ldff1<ANY_EXTEND:s><VNx2_NARROW:Vesize>\t%0.d, %5/z, [%1, %2.d, sxtw %p4]” “&& (!CONSTANT_P (operands[6]) || !CONSTANT_P (operands[7]))” { operands[6] = CONSTM1_RTX (VNx2BImode); operands[7] = CONSTM1_RTX (VNx2BImode); } )

;; Likewise, but with the offset being zero-extended from 32 bits. (define_insn_and_rewrite “*aarch64_ldff1_gather_<ANY_EXTEND:optab><VNx2_WIDE:mode><VNx2_NARROW:mode>_uxtw” [(set (match_operand:VNx2_WIDE 0 “register_operand” “=w, w”) (unspec:VNx2_WIDE [(match_operand 7) (ANY_EXTEND:VNx2_WIDE (unspec:VNx2_NARROW [(match_operand:VNx2BI 5 “register_operand” “Upl, Upl”) (match_operand:DI 1 “aarch64_reg_or_zero” “rk, rk”) (and:VNx2DI (match_operand:VNx2DI 2 “register_operand” “w, w”) (match_operand:VNx2DI 6 “aarch64_sve_uxtw_immediate”)) (match_operand:DI 3 “const_int_operand”) (match_operand:DI 4 “aarch64_gather_scale_operand_<VNx2_NARROW:Vesize>” “Ui1, i”) (mem:BLK (scratch)) (reg:VNx16BI FFRT_REGNUM)] UNSPEC_LDFF1_GATHER))] UNSPEC_PRED_X))] “TARGET_SVE” “@ ldff1<ANY_EXTEND:s><VNx2_NARROW:Vesize>\t%0.d, %5/z, [%1, %2.d, uxtw] ldff1<ANY_EXTEND:s><VNx2_NARROW:Vesize>\t%0.d, %5/z, [%1, %2.d, uxtw %p4]” “&& !CONSTANT_P (operands[7])” { operands[7] = CONSTM1_RTX (VNx2BImode); } )

;; ========================================================================= ;; == Prefetches ;; =========================================================================

;; ------------------------------------------------------------------------- ;; ---- Contiguous prefetches ;; ------------------------------------------------------------------------- ;; Includes contiguous forms of: ;; - PRFB ;; - PRFD ;; - PRFH ;; - PRFW ;; -------------------------------------------------------------------------

;; Contiguous predicated prefetches. Operand 2 gives the real prefetch ;; operation (as an svprfop), with operands 3 and 4 providing distilled ;; information. (define_insn “@aarch64_sve_prefetch” [(prefetch (unspec:DI [(match_operand: 0 “register_operand” “Upl”) (match_operand:SVE_FULL_I 1 “aarch64_sve_prefetch_operand” “UP”) (match_operand:DI 2 “const_int_operand”)] UNSPEC_SVE_PREFETCH) (match_operand:DI 3 “const_int_operand”) (match_operand:DI 4 “const_int_operand”))] “TARGET_SVE” { operands[1] = gen_rtx_MEM (mode, operands[1]); return aarch64_output_sve_prefetch (“prf”, operands[2], “%0, %1”); } )

;; ------------------------------------------------------------------------- ;; ---- Gather prefetches ;; ------------------------------------------------------------------------- ;; Includes gather forms of: ;; - PRFB ;; - PRFD ;; - PRFH ;; - PRFW ;; -------------------------------------------------------------------------

;; Predicated gather prefetches for 32-bit bases and offsets. The operands ;; are: ;; 0: the governing predicate ;; 1: the scalar component of the address ;; 2: the vector component of the address ;; 3: 1 for zero extension, 0 for sign extension ;; 4: the scale multiplier ;; 5: a vector zero that identifies the mode of data being accessed ;; 6: the prefetch operator (an svprfop) ;; 7: the normal RTL prefetch rw flag ;; 8: the normal RTL prefetch locality value (define_insn “@aarch64_sve_gather_prefetch<SVE_FULL_I:mode><VNx4SI_ONLY:mode>” [(prefetch (unspec:DI [(match_operand:VNx4BI 0 “register_operand” “Upl, Upl, Upl, Upl, Upl, Upl”) (match_operand:DI 1 “aarch64_sve_gather_offset_<SVE_FULL_I:Vesize>” “Z, vg<SVE_FULL_I:Vesize>, rk, rk, rk, rk”) (match_operand:VNx4SI_ONLY 2 “register_operand” “w, w, w, w, w, w”) (match_operand:DI 3 “const_int_operand” “i, i, Z, Ui1, Z, Ui1”) (match_operand:DI 4 “aarch64_gather_scale_operand_<SVE_FULL_I:Vesize>” “Ui1, Ui1, Ui1, Ui1, i, i”) (match_operand:SVE_FULL_I 5 “aarch64_simd_imm_zero”) (match_operand:DI 6 “const_int_operand”)] UNSPEC_SVE_PREFETCH_GATHER) (match_operand:DI 7 “const_int_operand”) (match_operand:DI 8 “const_int_operand”))] “TARGET_SVE” { static const char *const insns[][2] = { “prf<SVE_FULL_I:Vesize>”, “%0, [%2.s]”, “prf<SVE_FULL_I:Vesize>”, “%0, [%2.s, #%1]”, “prfb”, “%0, [%1, %2.s, sxtw]”, “prfb”, “%0, [%1, %2.s, uxtw]”, “prf<SVE_FULL_I:Vesize>”, “%0, [%1, %2.s, sxtw %p4]”, “prf<SVE_FULL_I:Vesize>”, “%0, [%1, %2.s, uxtw %p4]” }; const char *const *parts = insns[which_alternative]; return aarch64_output_sve_prefetch (parts[0], operands[6], parts[1]); } )

;; Predicated gather prefetches for 64-bit elements. The value of operand 3 ;; doesn't matter in this case. (define_insn “@aarch64_sve_gather_prefetch<SVE_FULL_I:mode><VNx2DI_ONLY:mode>” [(prefetch (unspec:DI [(match_operand:VNx2BI 0 “register_operand” “Upl, Upl, Upl, Upl”) (match_operand:DI 1 “aarch64_sve_gather_offset_<SVE_FULL_I:Vesize>” “Z, vg<SVE_FULL_I:Vesize>, rk, rk”) (match_operand:VNx2DI_ONLY 2 “register_operand” “w, w, w, w”) (match_operand:DI 3 “const_int_operand”) (match_operand:DI 4 “aarch64_gather_scale_operand_<SVE_FULL_I:Vesize>” “Ui1, Ui1, Ui1, i”) (match_operand:SVE_FULL_I 5 “aarch64_simd_imm_zero”) (match_operand:DI 6 “const_int_operand”)] UNSPEC_SVE_PREFETCH_GATHER) (match_operand:DI 7 “const_int_operand”) (match_operand:DI 8 “const_int_operand”))] “TARGET_SVE” { static const char *const insns[][2] = { “prf<SVE_FULL_I:Vesize>”, “%0, [%2.d]”, “prf<SVE_FULL_I:Vesize>”, “%0, [%2.d, #%1]”, “prfb”, “%0, [%1, %2.d]”, “prf<SVE_FULL_I:Vesize>”, “%0, [%1, %2.d, lsl %p4]” }; const char *const *parts = insns[which_alternative]; return aarch64_output_sve_prefetch (parts[0], operands[6], parts[1]); } )

;; Likewise, but with the offset being sign-extended from 32 bits. (define_insn_and_rewrite “*aarch64_sve_gather_prefetch<SVE_FULL_I:mode><VNx2DI_ONLY:mode>_sxtw” [(prefetch (unspec:DI [(match_operand:VNx2BI 0 “register_operand” “Upl, Upl”) (match_operand:DI 1 “register_operand” “rk, rk”) (unspec:VNx2DI_ONLY [(match_operand 9) (sign_extend:VNx2DI (truncate:VNx2SI (match_operand:VNx2DI 2 “register_operand” “w, w”)))] UNSPEC_PRED_X) (match_operand:DI 3 “const_int_operand”) (match_operand:DI 4 “aarch64_gather_scale_operand_<SVE_FULL_I:Vesize>” “Ui1, i”) (match_operand:SVE_FULL_I 5 “aarch64_simd_imm_zero”) (match_operand:DI 6 “const_int_operand”)] UNSPEC_SVE_PREFETCH_GATHER) (match_operand:DI 7 “const_int_operand”) (match_operand:DI 8 “const_int_operand”))] “TARGET_SVE” { static const char *const insns[][2] = { “prfb”, “%0, [%1, %2.d, sxtw]”, “prf<SVE_FULL_I:Vesize>”, “%0, [%1, %2.d, sxtw %p4]” }; const char *const *parts = insns[which_alternative]; return aarch64_output_sve_prefetch (parts[0], operands[6], parts[1]); } “&& !rtx_equal_p (operands[0], operands[9])” { operands[9] = copy_rtx (operands[0]); } )

;; Likewise, but with the offset being zero-extended from 32 bits. (define_insn “*aarch64_sve_gather_prefetch<SVE_FULL_I:mode><VNx2DI_ONLY:mode>_uxtw” [(prefetch (unspec:DI [(match_operand:VNx2BI 0 “register_operand” “Upl, Upl”) (match_operand:DI 1 “register_operand” “rk, rk”) (and:VNx2DI_ONLY (match_operand:VNx2DI 2 “register_operand” “w, w”) (match_operand:VNx2DI 9 “aarch64_sve_uxtw_immediate”)) (match_operand:DI 3 “const_int_operand”) (match_operand:DI 4 “aarch64_gather_scale_operand_<SVE_FULL_I:Vesize>” “Ui1, i”) (match_operand:SVE_FULL_I 5 “aarch64_simd_imm_zero”) (match_operand:DI 6 “const_int_operand”)] UNSPEC_SVE_PREFETCH_GATHER) (match_operand:DI 7 “const_int_operand”) (match_operand:DI 8 “const_int_operand”))] “TARGET_SVE” { static const char *const insns[][2] = { “prfb”, “%0, [%1, %2.d, uxtw]”, “prf<SVE_FULL_I:Vesize>”, “%0, [%1, %2.d, uxtw %p4]” }; const char *const *parts = insns[which_alternative]; return aarch64_output_sve_prefetch (parts[0], operands[6], parts[1]); } )

;; ========================================================================= ;; == Stores ;; =========================================================================

;; ------------------------------------------------------------------------- ;; ---- Normal contiguous stores ;; ------------------------------------------------------------------------- ;; Includes contiguous forms of: ;; - ST1B ;; - ST1D ;; - ST1H ;; - ST1W ;; - ST2B ;; - ST2D ;; - ST2H ;; - ST2W ;; - ST3B ;; - ST3D ;; - ST3H ;; - ST3W ;; - ST4B ;; - ST4D ;; - ST4H ;; - ST4W ;; -------------------------------------------------------------------------

;; Predicated ST1. (define_insn “maskstore” [(set (match_operand:SVE_ALL 0 “memory_operand” “+m”) (unspec:SVE_ALL [(match_operand: 2 “register_operand” “Upl”) (match_operand:SVE_ALL 1 “register_operand” “w”) (match_dup 0)] UNSPEC_ST1_SVE))] “TARGET_SVE” “st1\t%1., %2, %0” )

;; Unpredicated ST[234]. This is always a full update, so the dependence ;; on the old value of the memory location (via (match_dup 0)) is redundant. ;; There doesn‘t seem to be any obvious benefit to treating the all-true ;; case differently though. In particular, it’s very unlikely that we'll ;; only find out during RTL that a store_lanes is dead. (define_expand “vec_store_lanes” [(set (match_operand:SVE_STRUCT 0 “memory_operand”) (unspec:SVE_STRUCT [(match_dup 2) (match_operand:SVE_STRUCT 1 “register_operand”) (match_dup 0)] UNSPEC_STN))] “TARGET_SVE” { operands[2] = aarch64_ptrue_reg (mode); } )

;; Predicated ST[234]. (define_insn “vec_mask_store_lanes” [(set (match_operand:SVE_STRUCT 0 “memory_operand” “+m”) (unspec:SVE_STRUCT [(match_operand: 2 “register_operand” “Upl”) (match_operand:SVE_STRUCT 1 “register_operand” “w”) (match_dup 0)] UNSPEC_STN))] “TARGET_SVE” “st<vector_count>\t%1, %2, %0” )

;; ------------------------------------------------------------------------- ;; ---- Truncating contiguous stores ;; ------------------------------------------------------------------------- ;; Includes: ;; - ST1B ;; - ST1H ;; - ST1W ;; -------------------------------------------------------------------------

;; Predicated truncate and store, with 8 elements per 128-bit block. (define_insn “@aarch64_store_trunc<VNx8_NARROW:mode><VNx8_WIDE:mode>” [(set (match_operand:VNx8_NARROW 0 “memory_operand” “+m”) (unspec:VNx8_NARROW [(match_operand:VNx8BI 2 “register_operand” “Upl”) (truncate:VNx8_NARROW (match_operand:VNx8_WIDE 1 “register_operand” “w”)) (match_dup 0)] UNSPEC_ST1_SVE))] “TARGET_SVE” “st1<VNx8_NARROW:Vesize>\t%1.<VNx8_WIDE:Vetype>, %2, %0” )

;; Predicated truncate and store, with 4 elements per 128-bit block. (define_insn “@aarch64_store_trunc<VNx4_NARROW:mode><VNx4_WIDE:mode>” [(set (match_operand:VNx4_NARROW 0 “memory_operand” “+m”) (unspec:VNx4_NARROW [(match_operand:VNx4BI 2 “register_operand” “Upl”) (truncate:VNx4_NARROW (match_operand:VNx4_WIDE 1 “register_operand” “w”)) (match_dup 0)] UNSPEC_ST1_SVE))] “TARGET_SVE” “st1<VNx4_NARROW:Vesize>\t%1.<VNx4_WIDE:Vetype>, %2, %0” )

;; Predicated truncate and store, with 2 elements per 128-bit block. (define_insn “@aarch64_store_trunc<VNx2_NARROW:mode><VNx2_WIDE:mode>” [(set (match_operand:VNx2_NARROW 0 “memory_operand” “+m”) (unspec:VNx2_NARROW [(match_operand:VNx2BI 2 “register_operand” “Upl”) (truncate:VNx2_NARROW (match_operand:VNx2_WIDE 1 “register_operand” “w”)) (match_dup 0)] UNSPEC_ST1_SVE))] “TARGET_SVE” “st1<VNx2_NARROW:Vesize>\t%1.<VNx2_WIDE:Vetype>, %2, %0” )

;; ------------------------------------------------------------------------- ;; ---- Non-temporal contiguous stores ;; ------------------------------------------------------------------------- ;; Includes: ;; - STNT1B ;; - STNT1D ;; - STNT1H ;; - STNT1W ;; -------------------------------------------------------------------------

(define_insn “@aarch64_stnt1” [(set (match_operand:SVE_FULL 0 “memory_operand” “+m”) (unspec:SVE_FULL [(match_operand: 2 “register_operand” “Upl”) (match_operand:SVE_FULL 1 “register_operand” “w”) (match_dup 0)] UNSPEC_STNT1_SVE))] “TARGET_SVE” “stnt1\t%1., %2, %0” )

;; ------------------------------------------------------------------------- ;; ---- Normal scatter stores ;; ------------------------------------------------------------------------- ;; Includes scatter forms of: ;; - ST1D ;; - ST1W ;; -------------------------------------------------------------------------

;; Unpredicated scatter stores. (define_expand “scatter_store<v_int_container>” [(set (mem:BLK (scratch)) (unspec:BLK [(match_dup 5) (match_operand:DI 0 “aarch64_sve_gather_offset_”) (match_operand:<V_INT_CONTAINER> 1 “register_operand”) (match_operand:DI 2 “const_int_operand”) (match_operand:DI 3 “aarch64_gather_scale_operand_”) (match_operand:SVE_24 4 “register_operand”)] UNSPEC_ST1_SCATTER))] “TARGET_SVE” { operands[5] = aarch64_ptrue_reg (mode); } )

;; Predicated scatter stores for 32-bit elements. Operand 2 is true for ;; unsigned extension and false for signed extension. (define_insn “mask_scatter_store<v_int_container>” [(set (mem:BLK (scratch)) (unspec:BLK [(match_operand:VNx4BI 5 “register_operand” “Upl, Upl, Upl, Upl, Upl, Upl”) (match_operand:DI 0 “aarch64_sve_gather_offset_” “Z, vgw, rk, rk, rk, rk”) (match_operand:VNx4SI 1 “register_operand” “w, w, w, w, w, w”) (match_operand:DI 2 “const_int_operand” “Ui1, Ui1, Z, Ui1, Z, Ui1”) (match_operand:DI 3 “aarch64_gather_scale_operand_” “Ui1, Ui1, Ui1, Ui1, i, i”) (match_operand:SVE_4 4 “register_operand” “w, w, w, w, w, w”)] UNSPEC_ST1_SCATTER))] “TARGET_SVE” “@ st1\t%4.s, %5, [%1.s] st1\t%4.s, %5, [%1.s, #%0] st1\t%4.s, %5, [%0, %1.s, sxtw] st1\t%4.s, %5, [%0, %1.s, uxtw] st1\t%4.s, %5, [%0, %1.s, sxtw %p3] st1\t%4.s, %5, [%0, %1.s, uxtw %p3]” )

;; Predicated scatter stores for 64-bit elements. The value of operand 2 ;; doesn't matter in this case. (define_insn “mask_scatter_store<v_int_container>” [(set (mem:BLK (scratch)) (unspec:BLK [(match_operand:VNx2BI 5 “register_operand” “Upl, Upl, Upl, Upl”) (match_operand:DI 0 “aarch64_sve_gather_offset_” “Z, vgd, rk, rk”) (match_operand:VNx2DI 1 “register_operand” “w, w, w, w”) (match_operand:DI 2 “const_int_operand”) (match_operand:DI 3 “aarch64_gather_scale_operand_” “Ui1, Ui1, Ui1, i”) (match_operand:SVE_2 4 “register_operand” “w, w, w, w”)] UNSPEC_ST1_SCATTER))] “TARGET_SVE” “@ st1\t%4.d, %5, [%1.d] st1\t%4.d, %5, [%1.d, #%0] st1\t%4.d, %5, [%0, %1.d] st1\t%4.d, %5, [%0, %1.d, lsl %p3]” )

;; Likewise, but with the offset being extended from 32 bits. (define_insn_and_rewrite “*mask_scatter_store<v_int_container>_xtw_unpacked” [(set (mem:BLK (scratch)) (unspec:BLK [(match_operand:VNx2BI 5 “register_operand” “Upl, Upl”) (match_operand:DI 0 “register_operand” “rk, rk”) (unspec:VNx2DI [(match_operand 6) (ANY_EXTEND:VNx2DI (match_operand:VNx2SI 1 “register_operand” “w, w”))] UNSPEC_PRED_X) (match_operand:DI 2 “const_int_operand”) (match_operand:DI 3 “aarch64_gather_scale_operand_” “Ui1, i”) (match_operand:SVE_2 4 “register_operand” “w, w”)] UNSPEC_ST1_SCATTER))] “TARGET_SVE” “@ st1\t%4.d, %5, [%0, %1.d, xtw] st1\t%4.d, %5, [%0, %1.d, xtw %p3]” “&& !CONSTANT_P (operands[6])” { operands[6] = CONSTM1_RTX (mode); } )

;; Likewise, but with the offset being truncated to 32 bits and then ;; sign-extended. (define_insn_and_rewrite “*mask_scatter_store<v_int_container>_sxtw” [(set (mem:BLK (scratch)) (unspec:BLK [(match_operand:VNx2BI 5 “register_operand” “Upl, Upl”) (match_operand:DI 0 “register_operand” “rk, rk”) (unspec:VNx2DI [(match_operand 6) (sign_extend:VNx2DI (truncate:VNx2SI (match_operand:VNx2DI 1 “register_operand” “w, w”)))] UNSPEC_PRED_X) (match_operand:DI 2 “const_int_operand”) (match_operand:DI 3 “aarch64_gather_scale_operand_” “Ui1, i”) (match_operand:SVE_2 4 “register_operand” “w, w”)] UNSPEC_ST1_SCATTER))] “TARGET_SVE” “@ st1\t%4.d, %5, [%0, %1.d, sxtw] st1\t%4.d, %5, [%0, %1.d, sxtw %p3]” “&& !CONSTANT_P (operands[6])” { operands[6] = CONSTM1_RTX (mode); } )

;; Likewise, but with the offset being truncated to 32 bits and then ;; zero-extended. (define_insn “*mask_scatter_store<v_int_container>_uxtw” [(set (mem:BLK (scratch)) (unspec:BLK [(match_operand:VNx2BI 5 “register_operand” “Upl, Upl”) (match_operand:DI 0 “aarch64_reg_or_zero” “rk, rk”) (and:VNx2DI (match_operand:VNx2DI 1 “register_operand” “w, w”) (match_operand:VNx2DI 6 “aarch64_sve_uxtw_immediate”)) (match_operand:DI 2 “const_int_operand”) (match_operand:DI 3 “aarch64_gather_scale_operand_” “Ui1, i”) (match_operand:SVE_2 4 “register_operand” “w, w”)] UNSPEC_ST1_SCATTER))] “TARGET_SVE” “@ st1\t%4.d, %5, [%0, %1.d, uxtw] st1\t%4.d, %5, [%0, %1.d, uxtw %p3]” )

;; ------------------------------------------------------------------------- ;; ---- Truncating scatter stores ;; ------------------------------------------------------------------------- ;; Includes scatter forms of: ;; - ST1B ;; - ST1H ;; - ST1W ;; -------------------------------------------------------------------------

;; Predicated truncating scatter stores for 32-bit elements. Operand 2 is ;; true for unsigned extension and false for signed extension. (define_insn “@aarch64_scatter_store_trunc<VNx4_NARROW:mode><VNx4_WIDE:mode>” [(set (mem:BLK (scratch)) (unspec:BLK [(match_operand:VNx4BI 5 “register_operand” “Upl, Upl, Upl, Upl, Upl, Upl”) (match_operand:DI 0 “aarch64_sve_gather_offset_<VNx4_NARROW:Vesize>” “Z, vg<VNx4_NARROW:Vesize>, rk, rk, rk, rk”) (match_operand:VNx4SI 1 “register_operand” “w, w, w, w, w, w”) (match_operand:DI 2 “const_int_operand” “Ui1, Ui1, Z, Ui1, Z, Ui1”) (match_operand:DI 3 “aarch64_gather_scale_operand_<VNx4_NARROW:Vesize>” “Ui1, Ui1, Ui1, Ui1, i, i”) (truncate:VNx4_NARROW (match_operand:VNx4_WIDE 4 “register_operand” “w, w, w, w, w, w”))] UNSPEC_ST1_SCATTER))] “TARGET_SVE” “@ st1<VNx4_NARROW:Vesize>\t%4.s, %5, [%1.s] st1<VNx4_NARROW:Vesize>\t%4.s, %5, [%1.s, #%0] st1<VNx4_NARROW:Vesize>\t%4.s, %5, [%0, %1.s, sxtw] st1<VNx4_NARROW:Vesize>\t%4.s, %5, [%0, %1.s, uxtw] st1<VNx4_NARROW:Vesize>\t%4.s, %5, [%0, %1.s, sxtw %p3] st1<VNx4_NARROW:Vesize>\t%4.s, %5, [%0, %1.s, uxtw %p3]” )

;; Predicated truncating scatter stores for 64-bit elements. The value of ;; operand 2 doesn't matter in this case. (define_insn “@aarch64_scatter_store_trunc<VNx2_NARROW:mode><VNx2_WIDE:mode>” [(set (mem:BLK (scratch)) (unspec:BLK [(match_operand:VNx2BI 5 “register_operand” “Upl, Upl, Upl, Upl”) (match_operand:DI 0 “aarch64_sve_gather_offset_<VNx2_NARROW:Vesize>” “Z, vg<VNx2_NARROW:Vesize>, rk, rk”) (match_operand:VNx2DI 1 “register_operand” “w, w, w, w”) (match_operand:DI 2 “const_int_operand”) (match_operand:DI 3 “aarch64_gather_scale_operand_<VNx2_NARROW:Vesize>” “Ui1, Ui1, Ui1, i”) (truncate:VNx2_NARROW (match_operand:VNx2_WIDE 4 “register_operand” “w, w, w, w”))] UNSPEC_ST1_SCATTER))] “TARGET_SVE” “@ st1<VNx2_NARROW:Vesize>\t%4.d, %5, [%1.d] st1<VNx2_NARROW:Vesize>\t%4.d, %5, [%1.d, #%0] st1<VNx2_NARROW:Vesize>\t%4.d, %5, [%0, %1.d] st1<VNx2_NARROW:Vesize>\t%4.d, %5, [%0, %1.d, lsl %p3]” )

;; Likewise, but with the offset being sign-extended from 32 bits. (define_insn_and_rewrite “*aarch64_scatter_store_trunc<VNx2_NARROW:mode><VNx2_WIDE:mode>_sxtw” [(set (mem:BLK (scratch)) (unspec:BLK [(match_operand:VNx2BI 5 “register_operand” “Upl, Upl”) (match_operand:DI 0 “register_operand” “rk, rk”) (unspec:VNx2DI [(match_operand 6) (sign_extend:VNx2DI (truncate:VNx2SI (match_operand:VNx2DI 1 “register_operand” “w, w”)))] UNSPEC_PRED_X) (match_operand:DI 2 “const_int_operand”) (match_operand:DI 3 “aarch64_gather_scale_operand_<VNx2_NARROW:Vesize>” “Ui1, i”) (truncate:VNx2_NARROW (match_operand:VNx2_WIDE 4 “register_operand” “w, w”))] UNSPEC_ST1_SCATTER))] “TARGET_SVE” “@ st1<VNx2_NARROW:Vesize>\t%4.d, %5, [%0, %1.d, sxtw] st1<VNx2_NARROW:Vesize>\t%4.d, %5, [%0, %1.d, sxtw %p3]” “&& !rtx_equal_p (operands[5], operands[6])” { operands[6] = copy_rtx (operands[5]); } )

;; Likewise, but with the offset being zero-extended from 32 bits. (define_insn “*aarch64_scatter_store_trunc<VNx2_NARROW:mode><VNx2_WIDE:mode>_uxtw” [(set (mem:BLK (scratch)) (unspec:BLK [(match_operand:VNx2BI 5 “register_operand” “Upl, Upl”) (match_operand:DI 0 “aarch64_reg_or_zero” “rk, rk”) (and:VNx2DI (match_operand:VNx2DI 1 “register_operand” “w, w”) (match_operand:VNx2DI 6 “aarch64_sve_uxtw_immediate”)) (match_operand:DI 2 “const_int_operand”) (match_operand:DI 3 “aarch64_gather_scale_operand_<VNx2_NARROW:Vesize>” “Ui1, i”) (truncate:VNx2_NARROW (match_operand:VNx2_WIDE 4 “register_operand” “w, w”))] UNSPEC_ST1_SCATTER))] “TARGET_SVE” “@ st1<VNx2_NARROW:Vesize>\t%4.d, %5, [%0, %1.d, uxtw] st1<VNx2_NARROW:Vesize>\t%4.d, %5, [%0, %1.d, uxtw %p3]” )

;; ========================================================================= ;; == Vector creation ;; =========================================================================

;; ------------------------------------------------------------------------- ;; ---- [INT,FP] Duplicate element ;; ------------------------------------------------------------------------- ;; Includes: ;; - DUP ;; - MOV ;; - LD1RB ;; - LD1RD ;; - LD1RH ;; - LD1RW ;; - LD1ROB (F64MM) ;; - LD1ROD (F64MM) ;; - LD1ROH (F64MM) ;; - LD1ROW (F64MM) ;; - LD1RQB ;; - LD1RQD ;; - LD1RQH ;; - LD1RQW ;; -------------------------------------------------------------------------

(define_expand “vec_duplicate” [(parallel [(set (match_operand:SVE_ALL 0 “register_operand”) (vec_duplicate:SVE_ALL (match_operand: 1 “aarch64_sve_dup_operand”))) (clobber (scratch:VNx16BI))])] “TARGET_SVE” { if (MEM_P (operands[1])) { rtx ptrue = aarch64_ptrue_reg (mode); emit_insn (gen_sve_ld1r (operands[0], ptrue, operands[1], CONST0_RTX (mode))); DONE; } } )

;; Accept memory operands for the benefit of combine, and also in case ;; the scalar input gets spilled to memory during RA. We want to split ;; the load at the first opportunity in order to allow the PTRUE to be ;; optimized with surrounding code. (define_insn_and_split “*vec_duplicate_reg” [(set (match_operand:SVE_ALL 0 “register_operand” “=w, w, w”) (vec_duplicate:SVE_ALL (match_operand: 1 “aarch64_sve_dup_operand” “r, w, Uty”))) (clobber (match_scratch:VNx16BI 2 “=X, X, Upl”))] “TARGET_SVE” “@ mov\t%0., %1 mov\t%0., %1 #” “&& MEM_P (operands[1])” [(const_int 0)] { if (GET_CODE (operands[2]) == SCRATCH) operands[2] = gen_reg_rtx (VNx16BImode); emit_move_insn (operands[2], CONSTM1_RTX (VNx16BImode)); rtx gp = gen_lowpart (mode, operands[2]); emit_insn (gen_sve_ld1r (operands[0], gp, operands[1], CONST0_RTX (mode))); DONE; } [(set_attr “length” “4,4,8”)] )

;; Duplicate an Advanced SIMD vector to fill an SVE vector (LE version). (define_insn “@aarch64_vec_duplicate_vq_le” [(set (match_operand:SVE_FULL 0 “register_operand” “=w”) (vec_duplicate:SVE_FULL (match_operand: 1 “register_operand” “w”)))] “TARGET_SVE && !BYTES_BIG_ENDIAN” { operands[1] = gen_rtx_REG (mode, REGNO (operands[1])); return “dup\t%0.q, %1.q[0]”; } )

;; Duplicate an Advanced SIMD vector to fill an SVE vector (BE version). ;; The SVE register layout puts memory lane N into (architectural) ;; register lane N, whereas the Advanced SIMD layout puts the memory ;; lsb into the register lsb. We therefore have to describe this in rtl ;; terms as a reverse of the V128 vector followed by a duplicate. (define_insn “@aarch64_vec_duplicate_vq_be” [(set (match_operand:SVE_FULL 0 “register_operand” “=w”) (vec_duplicate:SVE_FULL (vec_select: (match_operand: 1 “register_operand” “w”) (match_operand 2 “descending_int_parallel”))))] “TARGET_SVE && BYTES_BIG_ENDIAN && known_eq (INTVAL (XVECEXP (operands[2], 0, 0)), GET_MODE_NUNITS (mode) - 1)” { operands[1] = gen_rtx_REG (mode, REGNO (operands[1])); return “dup\t%0.q, %1.q[0]”; } )

;; This is used for vec_duplicates from memory, but can also ;; be used by combine to optimize selects of a vec_duplicate ;; with zero. (define_insn “sve_ld1r” [(set (match_operand:SVE_ALL 0 “register_operand” “=w”) (unspec:SVE_ALL [(match_operand: 1 “register_operand” “Upl”) (vec_duplicate:SVE_ALL (match_operand: 2 “aarch64_sve_ld1r_operand” “Uty”)) (match_operand:SVE_ALL 3 “aarch64_simd_imm_zero”)] UNSPEC_SEL))] “TARGET_SVE” “ld1r\t%0., %1/z, %2” )

;; Load 128 bits from memory under predicate control and duplicate to ;; fill a vector. (define_insn “@aarch64_sve_ld1rq” [(set (match_operand:SVE_FULL 0 “register_operand” “=w”) (unspec:SVE_FULL [(match_operand: 2 “register_operand” “Upl”) (match_operand: 1 “aarch64_sve_ld1rq_operand” “UtQ”)] UNSPEC_LD1RQ))] “TARGET_SVE” { operands[1] = gen_rtx_MEM (mode, XEXP (operands[1], 0)); return “ld1rq\t%0., %2/z, %1”; } )

(define_insn “@aarch64_sve_ld1ro” [(set (match_operand:SVE_FULL 0 “register_operand” “=w”) (unspec:SVE_FULL [(match_operand: 2 “register_operand” “Upl”) (match_operand:OI 1 “aarch64_sve_ld1ro_operand_” “UO”)] UNSPEC_LD1RO))] “TARGET_SVE_F64MM” { operands[1] = gen_rtx_MEM (mode, XEXP (operands[1], 0)); return “ld1ro\t%0., %2/z, %1”; } )

;; ------------------------------------------------------------------------- ;; ---- [INT,FP] Initialize from individual elements ;; ------------------------------------------------------------------------- ;; Includes: ;; - INSR ;; -------------------------------------------------------------------------

(define_expand “vec_init” [(match_operand:SVE_FULL 0 “register_operand”) (match_operand 1 "")] “TARGET_SVE” { aarch64_sve_expand_vector_init (operands[0], operands[1]); DONE; } )

;; Shift an SVE vector left and insert a scalar into element 0. (define_insn “vec_shl_insert_” [(set (match_operand:SVE_FULL 0 “register_operand” “=?w, w, ??&w, ?&w”) (unspec:SVE_FULL [(match_operand:SVE_FULL 1 “register_operand” “0, 0, w, w”) (match_operand: 2 “aarch64_reg_or_zero” “rZ, w, rZ, w”)] UNSPEC_INSR))] “TARGET_SVE” “@ insr\t%0., %2 insr\t%0., %2 movprfx\t%0, %1;insr\t%0., %2 movprfx\t%0, %1;insr\t%0., %2” [(set_attr “movprfx” “,,yes,yes”)] )

;; ------------------------------------------------------------------------- ;; ---- [INT] Linear series ;; ------------------------------------------------------------------------- ;; Includes: ;; - INDEX ;; -------------------------------------------------------------------------

(define_insn “vec_series” [(set (match_operand:SVE_I 0 “register_operand” “=w, w, w”) (vec_series:SVE_I (match_operand: 1 “aarch64_sve_index_operand” “Usi, r, r”) (match_operand: 2 “aarch64_sve_index_operand” “r, Usi, r”)))] “TARGET_SVE” “@ index\t%0., #%1, %2 index\t%0., %1, #%2 index\t%0., %1, %2” )

;; Optimize {x, x, x, x, ...} + {0, n, 2n, 3n, ...} if n is in range ;; of an INDEX instruction. (define_insn “*vec_series_plus” [(set (match_operand:SVE_I 0 “register_operand” “=w”) (plus:SVE_I (vec_duplicate:SVE_I (match_operand: 1 “register_operand” “r”)) (match_operand:SVE_I 2 “immediate_operand”)))] “TARGET_SVE && aarch64_check_zero_based_sve_index_immediate (operands[2])” { operands[2] = aarch64_check_zero_based_sve_index_immediate (operands[2]); return “index\t%0., %1, #%2”; } )

;; ------------------------------------------------------------------------- ;; ---- [PRED] Duplicate element ;; ------------------------------------------------------------------------- ;; The patterns in this section are synthetic. ;; -------------------------------------------------------------------------

;; Implement a predicate broadcast by shifting the low bit of the scalar ;; input into the top bit and using a WHILELO. An alternative would be to ;; duplicate the input and do a compare with zero. (define_expand “vec_duplicate” [(set (match_operand:PRED_ALL 0 “register_operand”) (vec_duplicate:PRED_ALL (match_operand:QI 1 “register_operand”)))] “TARGET_SVE” { rtx tmp = gen_reg_rtx (DImode); rtx op1 = gen_lowpart (DImode, operands[1]); emit_insn (gen_ashldi3 (tmp, op1, gen_int_mode (63, DImode))); emit_insn (gen_while_ultdi (operands[0], const0_rtx, tmp)); DONE; } )

;; ========================================================================= ;; == Vector decomposition ;; =========================================================================

;; ------------------------------------------------------------------------- ;; ---- [INT,FP] Extract index ;; ------------------------------------------------------------------------- ;; Includes: ;; - DUP (Advanced SIMD) ;; - DUP (SVE) ;; - EXT (SVE) ;; - ST1 (Advanced SIMD) ;; - UMOV (Advanced SIMD) ;; -------------------------------------------------------------------------

(define_expand “vec_extract” [(set (match_operand: 0 “register_operand”) (vec_select: (match_operand:SVE_FULL 1 “register_operand”) (parallel [(match_operand:SI 2 “nonmemory_operand”)])))] “TARGET_SVE” { poly_int64 val; if (poly_int_rtx_p (operands[2], &val) && known_eq (val, GET_MODE_NUNITS (mode) - 1)) { /* The last element can be extracted with a LASTB and a false predicate. / rtx sel = aarch64_pfalse_reg (mode); emit_insn (gen_extract_last_ (operands[0], sel, operands[1])); DONE; } if (!CONST_INT_P (operands[2])) { / Create an index with operand[2] as the base and -1 as the step. It will then be zero for the element we care about. */ rtx index = gen_lowpart (<VEL_INT>mode, operands[2]); index = force_reg (<VEL_INT>mode, index); rtx series = gen_reg_rtx (<V_INT_EQUIV>mode); emit_insn (gen_vec_series<v_int_equiv> (series, index, constm1_rtx));

/* Get a predicate that is true for only that element.  */
rtx zero = CONST0_RTX (<V_INT_EQUIV>mode);
rtx cmp = gen_rtx_EQ (<V_INT_EQUIV>mode, series, zero);
rtx sel = gen_reg_rtx (<VPRED>mode);
emit_insn (gen_vec_cmp<v_int_equiv><vpred> (sel, cmp, series, zero));

/* Select the element using LASTB.  */
emit_insn (gen_extract_last_<mode> (operands[0], sel, operands[1]));
DONE;
  }

} )

;; Extract element zero. This is a special case because we want to force ;; the registers to be the same for the second alternative, and then ;; split the instruction into nothing after RA. (define_insn_and_split “*vec_extract_0” [(set (match_operand: 0 “aarch64_simd_nonimmediate_operand” “=r, w, Utv”) (vec_select: (match_operand:SVE_FULL 1 “register_operand” “w, 0, w”) (parallel [(const_int 0)])))] “TARGET_SVE” { operands[1] = gen_rtx_REG (mode, REGNO (operands[1])); switch (which_alternative) { case 0: return “umov\t%0, %1.[0]”; case 1: return “#”; case 2: return “st1\t{%1.}[0], %0”; default: gcc_unreachable (); } } “&& reload_completed && REG_P (operands[0]) && REGNO (operands[0]) == REGNO (operands[1])” [(const_int 0)] { emit_note (NOTE_INSN_DELETED); DONE; } [(set_attr “type” “neon_to_gp_q, untyped, neon_store1_one_lane_q”)] )

;; Extract an element from the Advanced SIMD portion of the register. ;; We don‘t just reuse the aarch64-simd.md pattern because we don’t ;; want any change in lane number on big-endian targets. (define_insn “*vec_extract_v128” [(set (match_operand: 0 “aarch64_simd_nonimmediate_operand” “=r, w, Utv”) (vec_select: (match_operand:SVE_FULL 1 “register_operand” “w, w, w”) (parallel [(match_operand:SI 2 “const_int_operand”)])))] “TARGET_SVE && IN_RANGE (INTVAL (operands[2]) * GET_MODE_SIZE (mode), 1, 15)” { operands[1] = gen_rtx_REG (mode, REGNO (operands[1])); switch (which_alternative) { case 0: return “umov\t%0, %1.[%2]”; case 1: return “dup\t%0, %1.[%2]”; case 2: return “st1\t{%1.}[%2], %0”; default: gcc_unreachable (); } } [(set_attr “type” “neon_to_gp_q, neon_dup_q, neon_store1_one_lane_q”)] )

;; Extract an element in the range of DUP. This pattern allows the ;; source and destination to be different. (define_insn “*vec_extract_dup” [(set (match_operand: 0 “register_operand” “=w”) (vec_select: (match_operand:SVE_FULL 1 “register_operand” “w”) (parallel [(match_operand:SI 2 “const_int_operand”)])))] “TARGET_SVE && IN_RANGE (INTVAL (operands[2]) * GET_MODE_SIZE (mode), 16, 63)” { operands[0] = gen_rtx_REG (mode, REGNO (operands[0])); return “dup\t%0., %1.[%2]”; } )

;; Extract an element outside the range of DUP. This pattern requires the ;; source and destination to be the same. (define_insn “*vec_extract_ext” [(set (match_operand: 0 “register_operand” “=w, ?&w”) (vec_select: (match_operand:SVE_FULL 1 “register_operand” “0, w”) (parallel [(match_operand:SI 2 “const_int_operand”)])))] “TARGET_SVE && INTVAL (operands[2]) * GET_MODE_SIZE (mode) >= 64” { operands[0] = gen_rtx_REG (mode, REGNO (operands[0])); operands[2] = GEN_INT (INTVAL (operands[2]) * GET_MODE_SIZE (mode)); return (which_alternative == 0 ? “ext\t%0.b, %0.b, %0.b, #%2” : “movprfx\t%0, %1;ext\t%0.b, %0.b, %1.b, #%2”); } [(set_attr “movprfx” “*,yes”)] )

;; ------------------------------------------------------------------------- ;; ---- [INT,FP] Extract active element ;; ------------------------------------------------------------------------- ;; Includes: ;; - LASTA ;; - LASTB ;; -------------------------------------------------------------------------

;; Extract the last active element of operand 1 into operand 0. ;; If no elements are active, extract the last inactive element instead. (define_insn “@extract_<last_op>_” [(set (match_operand: 0 “register_operand” “=?r, w”) (unspec: [(match_operand: 1 “register_operand” “Upl, Upl”) (match_operand:SVE_FULL 2 “register_operand” “w, w”)] LAST))] “TARGET_SVE” “@ last\t%0, %1, %2. last\t%0, %1, %2.” )

;; ------------------------------------------------------------------------- ;; ---- [PRED] Extract index ;; ------------------------------------------------------------------------- ;; The patterns in this section are synthetic. ;; -------------------------------------------------------------------------

;; Handle extractions from a predicate by converting to an integer vector ;; and extracting from there. (define_expand “vec_extract” [(match_operand: 0 “register_operand”) (match_operand: 1 “register_operand”) (match_operand:SI 2 “nonmemory_operand”) ;; Dummy operand to which we can attach the iterator. (reg:SVE_FULL_I V0_REGNUM)] “TARGET_SVE” { rtx tmp = gen_reg_rtx (mode); emit_insn (gen_vcond_mask_ (tmp, operands[1], CONST1_RTX (mode), CONST0_RTX (mode))); emit_insn (gen_vec_extract (operands[0], tmp, operands[2])); DONE; } )

;; ========================================================================= ;; == Unary arithmetic ;; =========================================================================

;; ------------------------------------------------------------------------- ;; ---- [INT] General unary arithmetic corresponding to rtx codes ;; ------------------------------------------------------------------------- ;; Includes: ;; - ABS ;; - CLS (= clrsb) ;; - CLZ ;; - CNT (= popcount) ;; - NEG ;; - NOT ;; -------------------------------------------------------------------------

;; Unpredicated integer unary arithmetic. (define_expand “2” [(set (match_operand:SVE_I 0 “register_operand”) (unspec:SVE_I [(match_dup 2) (SVE_INT_UNARY:SVE_I (match_operand:SVE_I 1 “register_operand”))] UNSPEC_PRED_X))] “TARGET_SVE” { operands[2] = aarch64_ptrue_reg (mode); } )

;; Integer unary arithmetic predicated with a PTRUE. (define_insn “@aarch64_pred_” [(set (match_operand:SVE_I 0 “register_operand” “=w, ?&w”) (unspec:SVE_I [(match_operand: 1 “register_operand” “Upl, Upl”) (SVE_INT_UNARY:SVE_I (match_operand:SVE_I 2 “register_operand” “0, w”))] UNSPEC_PRED_X))] “TARGET_SVE” “@ <sve_int_op>\t%0., %1/m, %2. movprfx\t%0, %2;<sve_int_op>\t%0., %1/m, %2.” [(set_attr “movprfx” “*,yes”)] )

;; Predicated integer unary arithmetic with merging. (define_expand “@cond_” [(set (match_operand:SVE_I 0 “register_operand”) (unspec:SVE_I [(match_operand: 1 “register_operand”) (SVE_INT_UNARY:SVE_I (match_operand:SVE_I 2 “register_operand”)) (match_operand:SVE_I 3 “aarch64_simd_reg_or_zero”)] UNSPEC_SEL))] “TARGET_SVE” )

;; Predicated integer unary arithmetic, merging with the first input. (define_insn “*cond__2” [(set (match_operand:SVE_I 0 “register_operand” “=w, ?&w”) (unspec:SVE_I [(match_operand: 1 “register_operand” “Upl, Upl”) (SVE_INT_UNARY:SVE_I (match_operand:SVE_I 2 “register_operand” “0, w”)) (match_dup 2)] UNSPEC_SEL))] “TARGET_SVE” “@ <sve_int_op>\t%0., %1/m, %0. movprfx\t%0, %2;<sve_int_op>\t%0., %1/m, %2.” [(set_attr “movprfx” “*,yes”)] )

;; Predicated integer unary arithmetic, merging with an independent value. ;; ;; The earlyclobber isn't needed for the first alternative, but omitting ;; it would only help the case in which operands 2 and 3 are the same, ;; which is handled above rather than here. Marking all the alternatives ;; as earlyclobber helps to make the instruction more regular to the ;; register allocator. (define_insn “*cond__any” [(set (match_operand:SVE_I 0 “register_operand” “=&w, ?&w, ?&w”) (unspec:SVE_I [(match_operand: 1 “register_operand” “Upl, Upl, Upl”) (SVE_INT_UNARY:SVE_I (match_operand:SVE_I 2 “register_operand” “w, w, w”)) (match_operand:SVE_I 3 “aarch64_simd_reg_or_zero” “0, Dz, w”)] UNSPEC_SEL))] “TARGET_SVE && !rtx_equal_p (operands[2], operands[3])” “@ <sve_int_op>\t%0., %1/m, %2. movprfx\t%0., %1/z, %2.;<sve_int_op>\t%0., %1/m, %2. movprfx\t%0, %3;<sve_int_op>\t%0., %1/m, %2.” [(set_attr “movprfx” “*,yes,yes”)] )

;; ------------------------------------------------------------------------- ;; ---- [INT] General unary arithmetic corresponding to unspecs ;; ------------------------------------------------------------------------- ;; Includes ;; - RBIT ;; - REVB ;; - REVH ;; - REVW ;; -------------------------------------------------------------------------

;; Predicated integer unary operations. (define_insn “@aarch64_pred_” [(set (match_operand:SVE_FULL_I 0 “register_operand” “=w, ?&w”) (unspec:SVE_FULL_I [(match_operand: 1 “register_operand” “Upl, Upl”) (unspec:SVE_FULL_I [(match_operand:SVE_FULL_I 2 “register_operand” “0, w”)] SVE_INT_UNARY)] UNSPEC_PRED_X))] “TARGET_SVE && <elem_bits> >= <min_elem_bits>” “@ <sve_int_op>\t%0., %1/m, %2. movprfx\t%0, %2;<sve_int_op>\t%0., %1/m, %2.” [(set_attr “movprfx” “*,yes”)] )

;; Another way of expressing the REVB, REVH and REVW patterns, with this ;; form being easier for permutes. The predicate mode determines the number ;; of lanes and the data mode decides the granularity of the reversal within ;; each lane. (define_insn “@aarch64_sve_revbhw_<SVE_ALL:mode><PRED_HSD:mode>” [(set (match_operand:SVE_ALL 0 “register_operand” “=w, ?&w”) (unspec:SVE_ALL [(match_operand:PRED_HSD 1 “register_operand” “Upl, Upl”) (unspec:SVE_ALL [(match_operand:SVE_ALL 2 “register_operand” “0, w”)] UNSPEC_REVBHW)] UNSPEC_PRED_X))] “TARGET_SVE && <PRED_HSD:elem_bits> > <SVE_ALL:container_bits>” “@ rev<SVE_ALL:Vcwtype>\t%0.<PRED_HSD:Vetype>, %1/m, %2.<PRED_HSD:Vetype> movprfx\t%0, %2;rev<SVE_ALL:Vcwtype>\t%0.<PRED_HSD:Vetype>, %1/m, %2.<PRED_HSD:Vetype>” [(set_attr “movprfx” “*,yes”)] )

;; Predicated integer unary operations with merging. (define_insn “@cond_” [(set (match_operand:SVE_FULL_I 0 “register_operand” “=w, ?&w, ?&w”) (unspec:SVE_FULL_I [(match_operand: 1 “register_operand” “Upl, Upl, Upl”) (unspec:SVE_FULL_I [(match_operand:SVE_FULL_I 2 “register_operand” “w, w, w”)] SVE_INT_UNARY) (match_operand:SVE_FULL_I 3 “aarch64_simd_reg_or_zero” “0, Dz, w”)] UNSPEC_SEL))] “TARGET_SVE && <elem_bits> >= <min_elem_bits>” “@ <sve_int_op>\t%0., %1/m, %2. movprfx\t%0., %1/z, %2.;<sve_int_op>\t%0., %1/m, %2. movprfx\t%0, %3;<sve_int_op>\t%0., %1/m, %2.” [(set_attr “movprfx” “*,yes,yes”)] )

;; ------------------------------------------------------------------------- ;; ---- [INT] Sign and zero extension ;; ------------------------------------------------------------------------- ;; Includes: ;; - SXTB ;; - SXTH ;; - SXTW ;; - UXTB ;; - UXTH ;; - UXTW ;; -------------------------------------------------------------------------

;; Unpredicated sign and zero extension from a narrower mode. (define_expand “<SVE_PARTIAL_I:mode><SVE_HSDI:mode>2” [(set (match_operand:SVE_HSDI 0 “register_operand”) (unspec:SVE_HSDI [(match_dup 2) (ANY_EXTEND:SVE_HSDI (match_operand:SVE_PARTIAL_I 1 “register_operand”))] UNSPEC_PRED_X))] “TARGET_SVE && (~<SVE_HSDI:narrower_mask> & <SVE_PARTIAL_I:self_mask>) == 0” { operands[2] = aarch64_ptrue_reg (<SVE_HSDI:VPRED>mode); } )

;; Predicated sign and zero extension from a narrower mode. (define_insn “*<SVE_PARTIAL_I:mode><SVE_HSDI:mode>2” [(set (match_operand:SVE_HSDI 0 “register_operand” “=w, ?&w”) (unspec:SVE_HSDI [(match_operand:<SVE_HSDI:VPRED> 1 “register_operand” “Upl, Upl”) (ANY_EXTEND:SVE_HSDI (match_operand:SVE_PARTIAL_I 2 “register_operand” “0, w”))] UNSPEC_PRED_X))] “TARGET_SVE && (~<SVE_HSDI:narrower_mask> & <SVE_PARTIAL_I:self_mask>) == 0” “@ xt<SVE_PARTIAL_I:Vesize>\t%0.<SVE_HSDI:Vetype>, %1/m, %2.<SVE_HSDI:Vetype> movprfx\t%0, %2;xt<SVE_PARTIAL_I:Vesize>\t%0.<SVE_HSDI:Vetype>, %1/m, %2.<SVE_HSDI:Vetype>” [(set_attr “movprfx” “*,yes”)] )

;; Predicated truncate-and-sign-extend operations. (define_insn “@aarch64_pred_sxt<SVE_FULL_HSDI:mode><SVE_PARTIAL_I:mode>” [(set (match_operand:SVE_FULL_HSDI 0 “register_operand” “=w, ?&w”) (unspec:SVE_FULL_HSDI [(match_operand:<SVE_FULL_HSDI:VPRED> 1 “register_operand” “Upl, Upl”) (sign_extend:SVE_FULL_HSDI (truncate:SVE_PARTIAL_I (match_operand:SVE_FULL_HSDI 2 “register_operand” “0, w”)))] UNSPEC_PRED_X))] “TARGET_SVE && (~<SVE_FULL_HSDI:narrower_mask> & <SVE_PARTIAL_I:self_mask>) == 0” “@ sxt<SVE_PARTIAL_I:Vesize>\t%0.<SVE_FULL_HSDI:Vetype>, %1/m, %2.<SVE_FULL_HSDI:Vetype> movprfx\t%0, %2;sxt<SVE_PARTIAL_I:Vesize>\t%0.<SVE_FULL_HSDI:Vetype>, %1/m, %2.<SVE_FULL_HSDI:Vetype>” [(set_attr “movprfx” “*,yes”)] )

;; Predicated truncate-and-sign-extend operations with merging. (define_insn “@aarch64_cond_sxt<SVE_FULL_HSDI:mode><SVE_PARTIAL_I:mode>” [(set (match_operand:SVE_FULL_HSDI 0 “register_operand” “=w, ?&w, ?&w”) (unspec:SVE_FULL_HSDI [(match_operand:<SVE_FULL_HSDI:VPRED> 1 “register_operand” “Upl, Upl, Upl”) (sign_extend:SVE_FULL_HSDI (truncate:SVE_PARTIAL_I (match_operand:SVE_FULL_HSDI 2 “register_operand” “w, w, w”))) (match_operand:SVE_FULL_HSDI 3 “aarch64_simd_reg_or_zero” “0, Dz, w”)] UNSPEC_SEL))] “TARGET_SVE && (~<SVE_FULL_HSDI:narrower_mask> & <SVE_PARTIAL_I:self_mask>) == 0” “@ sxt<SVE_PARTIAL_I:Vesize>\t%0.<SVE_FULL_HSDI:Vetype>, %1/m, %2.<SVE_FULL_HSDI:Vetype> movprfx\t%0.<SVE_FULL_HSDI:Vetype>, %1/z, %2.<SVE_FULL_HSDI:Vetype>;sxt<SVE_PARTIAL_I:Vesize>\t%0.<SVE_FULL_HSDI:Vetype>, %1/m, %2.<SVE_FULL_HSDI:Vetype> movprfx\t%0, %3;sxt<SVE_PARTIAL_I:Vesize>\t%0.<SVE_FULL_HSDI:Vetype>, %1/m, %2.<SVE_FULL_HSDI:Vetype>” [(set_attr “movprfx” “*,yes,yes”)] )

;; Predicated truncate-and-zero-extend operations, merging with the ;; first input. ;; ;; The canonical form of this operation is an AND of a constant rather ;; than (zero_extend (truncate ...)). (define_insn “*cond_uxt_2” [(set (match_operand:SVE_I 0 “register_operand” “=w, ?&w”) (unspec:SVE_I [(match_operand: 1 “register_operand” “Upl, Upl”) (and:SVE_I (match_operand:SVE_I 2 “register_operand” “0, w”) (match_operand:SVE_I 3 “aarch64_sve_uxt_immediate”)) (match_dup 2)] UNSPEC_SEL))] “TARGET_SVE” “@ uxt%e3\t%0., %1/m, %0. movprfx\t%0, %2;uxt%e3\t%0., %1/m, %2.” [(set_attr “movprfx” “*,yes”)] )

;; Predicated truncate-and-zero-extend operations, merging with an ;; independent value. ;; ;; The earlyclobber isn't needed for the first alternative, but omitting ;; it would only help the case in which operands 2 and 4 are the same, ;; which is handled above rather than here. Marking all the alternatives ;; as early-clobber helps to make the instruction more regular to the ;; register allocator. (define_insn “*cond_uxt_any” [(set (match_operand:SVE_I 0 “register_operand” “=&w, ?&w, ?&w”) (unspec:SVE_I [(match_operand: 1 “register_operand” “Upl, Upl, Upl”) (and:SVE_I (match_operand:SVE_I 2 “register_operand” “w, w, w”) (match_operand:SVE_I 3 “aarch64_sve_uxt_immediate”)) (match_operand:SVE_I 4 “aarch64_simd_reg_or_zero” “0, Dz, w”)] UNSPEC_SEL))] “TARGET_SVE && !rtx_equal_p (operands[2], operands[4])” “@ uxt%e3\t%0., %1/m, %2. movprfx\t%0., %1/z, %2.;uxt%e3\t%0., %1/m, %2. movprfx\t%0, %4;uxt%e3\t%0., %1/m, %2.” [(set_attr “movprfx” “*,yes,yes”)] )

;; ------------------------------------------------------------------------- ;; ---- [INT] Truncation ;; ------------------------------------------------------------------------- ;; The patterns in this section are synthetic. ;; -------------------------------------------------------------------------

;; Truncate to a partial SVE vector from either a full vector or a ;; wider partial vector. This is a no-op, because we can just ignore ;; the unused upper bits of the source. (define_insn_and_split “trunc<SVE_HSDI:mode><SVE_PARTIAL_I:mode>2” [(set (match_operand:SVE_PARTIAL_I 0 “register_operand” “=w”) (truncate:SVE_PARTIAL_I (match_operand:SVE_HSDI 1 “register_operand” “w”)))] “TARGET_SVE && (~<SVE_HSDI:narrower_mask> & <SVE_PARTIAL_I:self_mask>) == 0” “#” “&& reload_completed” [(set (match_dup 0) (match_dup 1))] { operands[1] = aarch64_replace_reg_mode (operands[1], <SVE_PARTIAL_I:MODE>mode); } )

;; ------------------------------------------------------------------------- ;; ---- [INT] Logical inverse ;; ------------------------------------------------------------------------- ;; Includes: ;; - CNOT ;; -------------------------------------------------------------------------

;; Predicated logical inverse. (define_expand “@aarch64_pred_cnot” [(set (match_operand:SVE_FULL_I 0 “register_operand”) (unspec:SVE_FULL_I [(unspec: [(match_operand: 1 “register_operand”) (match_operand:SI 2 “aarch64_sve_ptrue_flag”) (eq: (match_operand:SVE_FULL_I 3 “register_operand”) (match_dup 4))] UNSPEC_PRED_Z) (match_dup 5) (match_dup 4)] UNSPEC_SEL))] “TARGET_SVE” { operands[4] = CONST0_RTX (mode); operands[5] = CONST1_RTX (mode); } )

(define_insn “*cnot” [(set (match_operand:SVE_I 0 “register_operand” “=w, ?&w”) (unspec:SVE_I [(unspec: [(match_operand: 1 “register_operand” “Upl, Upl”) (match_operand:SI 5 “aarch64_sve_ptrue_flag”) (eq: (match_operand:SVE_I 2 “register_operand” “0, w”) (match_operand:SVE_I 3 “aarch64_simd_imm_zero”))] UNSPEC_PRED_Z) (match_operand:SVE_I 4 “aarch64_simd_imm_one”) (match_dup 3)] UNSPEC_SEL))] “TARGET_SVE” “@ cnot\t%0., %1/m, %2. movprfx\t%0, %2;cnot\t%0., %1/m, %2.” [(set_attr “movprfx” “*,yes”)] )

;; Predicated logical inverse with merging. (define_expand “@cond_cnot” [(set (match_operand:SVE_FULL_I 0 “register_operand”) (unspec:SVE_FULL_I [(match_operand: 1 “register_operand”) (unspec:SVE_FULL_I [(unspec: [(match_dup 4) (const_int SVE_KNOWN_PTRUE) (eq: (match_operand:SVE_FULL_I 2 “register_operand”) (match_dup 5))] UNSPEC_PRED_Z) (match_dup 6) (match_dup 5)] UNSPEC_SEL) (match_operand:SVE_FULL_I 3 “aarch64_simd_reg_or_zero”)] UNSPEC_SEL))] “TARGET_SVE” { operands[4] = CONSTM1_RTX (mode); operands[5] = CONST0_RTX (mode); operands[6] = CONST1_RTX (mode); } )

;; Predicated logical inverse, merging with the first input. (define_insn_and_rewrite “*cond_cnot_2” [(set (match_operand:SVE_I 0 “register_operand” “=w, ?&w”) (unspec:SVE_I [(match_operand: 1 “register_operand” “Upl, Upl”) ;; Logical inverse of operand 2 (as above). (unspec:SVE_I [(unspec: [(match_operand 5) (const_int SVE_KNOWN_PTRUE) (eq: (match_operand:SVE_I 2 “register_operand” “0, w”) (match_operand:SVE_I 3 “aarch64_simd_imm_zero”))] UNSPEC_PRED_Z) (match_operand:SVE_I 4 “aarch64_simd_imm_one”) (match_dup 3)] UNSPEC_SEL) (match_dup 2)] UNSPEC_SEL))] “TARGET_SVE” “@ cnot\t%0., %1/m, %0. movprfx\t%0, %2;cnot\t%0., %1/m, %2.” “&& !CONSTANT_P (operands[5])” { operands[5] = CONSTM1_RTX (mode); } [(set_attr “movprfx” “*,yes”)] )

;; Predicated logical inverse, merging with an independent value. ;; ;; The earlyclobber isn't needed for the first alternative, but omitting ;; it would only help the case in which operands 2 and 6 are the same, ;; which is handled above rather than here. Marking all the alternatives ;; as earlyclobber helps to make the instruction more regular to the ;; register allocator. (define_insn_and_rewrite “*cond_cnot_any” [(set (match_operand:SVE_I 0 “register_operand” “=&w, ?&w, ?&w”) (unspec:SVE_I [(match_operand: 1 “register_operand” “Upl, Upl, Upl”) ;; Logical inverse of operand 2 (as above). (unspec:SVE_I [(unspec: [(match_operand 5) (const_int SVE_KNOWN_PTRUE) (eq: (match_operand:SVE_I 2 “register_operand” “w, w, w”) (match_operand:SVE_I 3 “aarch64_simd_imm_zero”))] UNSPEC_PRED_Z) (match_operand:SVE_I 4 “aarch64_simd_imm_one”) (match_dup 3)] UNSPEC_SEL) (match_operand:SVE_I 6 “aarch64_simd_reg_or_zero” “0, Dz, w”)] UNSPEC_SEL))] “TARGET_SVE && !rtx_equal_p (operands[2], operands[6])” “@ cnot\t%0., %1/m, %2. movprfx\t%0., %1/z, %2.;cnot\t%0., %1/m, %2. movprfx\t%0, %6;cnot\t%0., %1/m, %2.” “&& !CONSTANT_P (operands[5])” { operands[5] = CONSTM1_RTX (mode); } [(set_attr “movprfx” “*,yes,yes”)] )

;; ------------------------------------------------------------------------- ;; ---- [FP<-INT] General unary arithmetic that maps to unspecs ;; ------------------------------------------------------------------------- ;; Includes: ;; - FEXPA ;; -------------------------------------------------------------------------

;; Unpredicated unary operations that take an integer and return a float. (define_insn “@aarch64_sve_” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=w”) (unspec:SVE_FULL_F [(match_operand:<V_INT_EQUIV> 1 “register_operand” “w”)] SVE_FP_UNARY_INT))] “TARGET_SVE” “<sve_fp_op>\t%0., %1.” )

;; ------------------------------------------------------------------------- ;; ---- [FP] General unary arithmetic corresponding to unspecs ;; ------------------------------------------------------------------------- ;; Includes: ;; - FABS ;; - FNEG ;; - FRECPE ;; - FRECPX ;; - FRINTA ;; - FRINTI ;; - FRINTM ;; - FRINTN ;; - FRINTP ;; - FRINTX ;; - FRINTZ ;; - FRSQRTE ;; - FSQRT ;; -------------------------------------------------------------------------

;; Unpredicated floating-point unary operations. (define_insn “@aarch64_sve_” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=w”) (unspec:SVE_FULL_F [(match_operand:SVE_FULL_F 1 “register_operand” “w”)] SVE_FP_UNARY))] “TARGET_SVE” “<sve_fp_op>\t%0., %1.” )

;; Unpredicated floating-point unary operations. (define_expand “2” [(set (match_operand:SVE_FULL_F 0 “register_operand”) (unspec:SVE_FULL_F [(match_dup 2) (const_int SVE_RELAXED_GP) (match_operand:SVE_FULL_F 1 “register_operand”)] SVE_COND_FP_UNARY_OPTAB))] “TARGET_SVE” { operands[2] = aarch64_ptrue_reg (mode); } )

;; Predicated floating-point unary operations. (define_insn “@aarch64_pred_” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=w, ?&w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl”) (match_operand:SI 3 “aarch64_sve_gp_strictness”) (match_operand:SVE_FULL_F 2 “register_operand” “0, w”)] SVE_COND_FP_UNARY))] “TARGET_SVE” “@ <sve_fp_op>\t%0., %1/m, %2. movprfx\t%0, %2;<sve_fp_op>\t%0., %1/m, %2.” [(set_attr “movprfx” “*,yes”)] )

;; Predicated floating-point unary arithmetic with merging. (define_expand “@cond_” [(set (match_operand:SVE_FULL_F 0 “register_operand”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand”) (unspec:SVE_FULL_F [(match_dup 1) (const_int SVE_STRICT_GP) (match_operand:SVE_FULL_F 2 “register_operand”)] SVE_COND_FP_UNARY) (match_operand:SVE_FULL_F 3 “aarch64_simd_reg_or_zero”)] UNSPEC_SEL))] “TARGET_SVE” )

;; Predicated floating-point unary arithmetic, merging with the first input. (define_insn_and_rewrite “*cond__2_relaxed” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=w, ?&w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl”) (unspec:SVE_FULL_F [(match_operand 3) (const_int SVE_RELAXED_GP) (match_operand:SVE_FULL_F 2 “register_operand” “0, w”)] SVE_COND_FP_UNARY) (match_dup 2)] UNSPEC_SEL))] “TARGET_SVE” “@ <sve_fp_op>\t%0., %1/m, %0. movprfx\t%0, %2;<sve_fp_op>\t%0., %1/m, %2.” “&& !rtx_equal_p (operands[1], operands[3])” { operands[3] = copy_rtx (operands[1]); } [(set_attr “movprfx” “*,yes”)] )

(define_insn “*cond__2_strict” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=w, ?&w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl”) (unspec:SVE_FULL_F [(match_dup 1) (const_int SVE_STRICT_GP) (match_operand:SVE_FULL_F 2 “register_operand” “0, w”)] SVE_COND_FP_UNARY) (match_dup 2)] UNSPEC_SEL))] “TARGET_SVE” “@ <sve_fp_op>\t%0., %1/m, %0. movprfx\t%0, %2;<sve_fp_op>\t%0., %1/m, %2.” [(set_attr “movprfx” “*,yes”)] )

;; Predicated floating-point unary arithmetic, merging with an independent ;; value. ;; ;; The earlyclobber isn't needed for the first alternative, but omitting ;; it would only help the case in which operands 2 and 3 are the same, ;; which is handled above rather than here. Marking all the alternatives ;; as earlyclobber helps to make the instruction more regular to the ;; register allocator. (define_insn_and_rewrite “*cond__any_relaxed” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=&w, ?&w, ?&w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl, Upl”) (unspec:SVE_FULL_F [(match_operand 4) (const_int SVE_RELAXED_GP) (match_operand:SVE_FULL_F 2 “register_operand” “w, w, w”)] SVE_COND_FP_UNARY) (match_operand:SVE_FULL_F 3 “aarch64_simd_reg_or_zero” “0, Dz, w”)] UNSPEC_SEL))] “TARGET_SVE && !rtx_equal_p (operands[2], operands[3])” “@ <sve_fp_op>\t%0., %1/m, %2. movprfx\t%0., %1/z, %2.;<sve_fp_op>\t%0., %1/m, %2. movprfx\t%0, %3;<sve_fp_op>\t%0., %1/m, %2.” “&& !rtx_equal_p (operands[1], operands[4])” { operands[4] = copy_rtx (operands[1]); } [(set_attr “movprfx” “*,yes,yes”)] )

(define_insn “*cond__any_strict” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=&w, ?&w, ?&w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl, Upl”) (unspec:SVE_FULL_F [(match_dup 1) (const_int SVE_STRICT_GP) (match_operand:SVE_FULL_F 2 “register_operand” “w, w, w”)] SVE_COND_FP_UNARY) (match_operand:SVE_FULL_F 3 “aarch64_simd_reg_or_zero” “0, Dz, w”)] UNSPEC_SEL))] “TARGET_SVE && !rtx_equal_p (operands[2], operands[3])” “@ <sve_fp_op>\t%0., %1/m, %2. movprfx\t%0., %1/z, %2.;<sve_fp_op>\t%0., %1/m, %2. movprfx\t%0, %3;<sve_fp_op>\t%0., %1/m, %2.” [(set_attr “movprfx” “*,yes,yes”)] )

;; ------------------------------------------------------------------------- ;; ---- [FP] Square root ;; -------------------------------------------------------------------------

(define_expand “sqrt2” [(set (match_operand:SVE_FULL_F 0 “register_operand”) (unspec:SVE_FULL_F [(match_dup 2) (const_int SVE_RELAXED_GP) (match_operand:SVE_FULL_F 1 “register_operand”)] UNSPEC_COND_FSQRT))] “TARGET_SVE” { if (aarch64_emit_approx_sqrt (operands[0], operands[1], false)) DONE; operands[2] = aarch64_ptrue_reg (mode); })

;; ------------------------------------------------------------------------- ;; ---- [FP] Reciprocal square root ;; -------------------------------------------------------------------------

(define_expand “rsqrt2” [(set (match_operand:SVE_FULL_SDF 0 “register_operand”) (unspec:SVE_FULL_SDF [(match_operand:SVE_FULL_SDF 1 “register_operand”)] UNSPEC_RSQRT))] “TARGET_SVE” { aarch64_emit_approx_sqrt (operands[0], operands[1], true); DONE; })

(define_expand “@aarch64_rsqrte” [(set (match_operand:SVE_FULL_SDF 0 “register_operand”) (unspec:SVE_FULL_SDF [(match_operand:SVE_FULL_SDF 1 “register_operand”)] UNSPEC_RSQRTE))] “TARGET_SVE” )

(define_expand “@aarch64_rsqrts” [(set (match_operand:SVE_FULL_SDF 0 “register_operand”) (unspec:SVE_FULL_SDF [(match_operand:SVE_FULL_SDF 1 “register_operand”) (match_operand:SVE_FULL_SDF 2 “register_operand”)] UNSPEC_RSQRTS))] “TARGET_SVE” )

;; ------------------------------------------------------------------------- ;; ---- [PRED] Inverse ;; ------------------------------------------------------------------------- ;; Includes: ;; - NOT ;; -------------------------------------------------------------------------

;; Unpredicated predicate inverse. (define_expand “one_cmpl2” [(set (match_operand:PRED_ALL 0 “register_operand”) (and:PRED_ALL (not:PRED_ALL (match_operand:PRED_ALL 1 “register_operand”)) (match_dup 2)))] “TARGET_SVE” { operands[2] = aarch64_ptrue_reg (mode); } )

;; Predicated predicate inverse. (define_insn “*one_cmpl3” [(set (match_operand:PRED_ALL 0 “register_operand” “=Upa”) (and:PRED_ALL (not:PRED_ALL (match_operand:PRED_ALL 2 “register_operand” “Upa”)) (match_operand:PRED_ALL 1 “register_operand” “Upa”)))] “TARGET_SVE” “not\t%0.b, %1/z, %2.b” )

;; ========================================================================= ;; == Binary arithmetic ;; =========================================================================

;; ------------------------------------------------------------------------- ;; ---- [INT] General binary arithmetic corresponding to rtx codes ;; ------------------------------------------------------------------------- ;; Includes: ;; - ADD (merging form only) ;; - AND (merging form only) ;; - ASR (merging form only) ;; - EOR (merging form only) ;; - LSL (merging form only) ;; - LSR (merging form only) ;; - MUL ;; - ORR (merging form only) ;; - SMAX ;; - SMIN ;; - SQADD (SVE2 merging form only) ;; - SQSUB (SVE2 merging form only) ;; - SUB (merging form only) ;; - UMAX ;; - UMIN ;; - UQADD (SVE2 merging form only) ;; - UQSUB (SVE2 merging form only) ;; -------------------------------------------------------------------------

;; Unpredicated integer binary operations that have an immediate form. (define_expand “3” [(set (match_operand:SVE_I 0 “register_operand”) (unspec:SVE_I [(match_dup 3) (SVE_INT_BINARY_IMM:SVE_I (match_operand:SVE_I 1 “register_operand”) (match_operand:SVE_I 2 “aarch64_sve_<sve_imm_con>_operand”))] UNSPEC_PRED_X))] “TARGET_SVE” { operands[3] = aarch64_ptrue_reg (mode); } )

;; Integer binary operations that have an immediate form, predicated ;; with a PTRUE. We don‘t actually need the predicate for the first ;; and third alternatives, but using Upa or X isn’t likely to gain much ;; and would make the instruction seem less uniform to the register ;; allocator. (define_insn_and_split “@aarch64_pred_” [(set (match_operand:SVE_I 0 “register_operand” “=w, w, ?&w, ?&w”) (unspec:SVE_I [(match_operand: 1 “register_operand” “Upl, Upl, Upl, Upl”) (SVE_INT_BINARY_IMM:SVE_I (match_operand:SVE_I 2 “register_operand” “%0, 0, w, w”) (match_operand:SVE_I 3 “aarch64_sve_<sve_imm_con>_operand” “<sve_imm_con>, w, <sve_imm_con>, w”))] UNSPEC_PRED_X))] “TARGET_SVE” "@

<sve_int_op>\t%0., %1/m, %0., %3.

movprfx\t%0, %2;<sve_int_op>\t%0., %1/m, %0., %3." ; Split the unpredicated form after reload, so that we don't have ; the unnecessary PTRUE. “&& reload_completed && !register_operand (operands[3], mode)” [(set (match_dup 0) (SVE_INT_BINARY_IMM:SVE_I (match_dup 2) (match_dup 3)))] "" [(set_attr “movprfx” “,,yes,yes”)] )

;; Unpredicated binary operations with a constant (post-RA only). ;; These are generated by splitting a predicated instruction whose ;; predicate is unused. (define_insn “*post_ra_3” [(set (match_operand:SVE_I 0 “register_operand” “=w, ?&w”) (SVE_INT_BINARY_IMM:SVE_I (match_operand:SVE_I 1 “register_operand” “0, w”) (match_operand:SVE_I 2 “aarch64_sve_<sve_imm_con>_immediate”)))] “TARGET_SVE && reload_completed” “@ <sve_int_op>\t%0., %0., #%<sve_imm_prefix>2 movprfx\t%0, %1;<sve_int_op>\t%0., %0., #%<sve_imm_prefix>2” [(set_attr “movprfx” “*,yes”)] )

;; Predicated integer operations with merging. (define_expand “@cond_” [(set (match_operand:SVE_I 0 “register_operand”) (unspec:SVE_I [(match_operand: 1 “register_operand”) (SVE_INT_BINARY:SVE_I (match_operand:SVE_I 2 “register_operand”) (match_operand:SVE_I 3 “<sve_pred_int_rhs2_operand>”)) (match_operand:SVE_I 4 “aarch64_simd_reg_or_zero”)] UNSPEC_SEL))] “TARGET_SVE” )

;; Predicated integer operations, merging with the first input. (define_insn “*cond__2” [(set (match_operand:SVE_I 0 “register_operand” “=w, ?&w”) (unspec:SVE_I [(match_operand: 1 “register_operand” “Upl, Upl”) (SVE_INT_BINARY:SVE_I (match_operand:SVE_I 2 “register_operand” “0, w”) (match_operand:SVE_I 3 “register_operand” “w, w”)) (match_dup 2)] UNSPEC_SEL))] “TARGET_SVE” “@ <sve_int_op>\t%0., %1/m, %0., %3. movprfx\t%0, %2;<sve_int_op>\t%0., %1/m, %0., %3.” [(set_attr “movprfx” “*,yes”)] )

;; Predicated integer operations, merging with the second input. (define_insn “*cond__3” [(set (match_operand:SVE_I 0 “register_operand” “=w, ?&w”) (unspec:SVE_I [(match_operand: 1 “register_operand” “Upl, Upl”) (SVE_INT_BINARY:SVE_I (match_operand:SVE_I 2 “register_operand” “w, w”) (match_operand:SVE_I 3 “register_operand” “0, w”)) (match_dup 3)] UNSPEC_SEL))] “TARGET_SVE” “@ <sve_int_op_rev>\t%0., %1/m, %0., %2. movprfx\t%0, %3;<sve_int_op_rev>\t%0., %1/m, %0., %2.” [(set_attr “movprfx” “*,yes”)] )

;; Predicated integer operations, merging with an independent value. (define_insn_and_rewrite “*cond__any” [(set (match_operand:SVE_I 0 “register_operand” “=&w, &w, &w, &w, ?&w”) (unspec:SVE_I [(match_operand: 1 “register_operand” “Upl, Upl, Upl, Upl, Upl”) (SVE_INT_BINARY:SVE_I (match_operand:SVE_I 2 “register_operand” “0, w, w, w, w”) (match_operand:SVE_I 3 “register_operand” “w, 0, w, w, w”)) (match_operand:SVE_I 4 “aarch64_simd_reg_or_zero” “Dz, Dz, Dz, 0, w”)] UNSPEC_SEL))] “TARGET_SVE && !rtx_equal_p (operands[2], operands[4]) && !rtx_equal_p (operands[3], operands[4])” “@ movprfx\t%0., %1/z, %0.;<sve_int_op>\t%0., %1/m, %0., %3. movprfx\t%0., %1/z, %0.;<sve_int_op_rev>\t%0., %1/m, %0., %2. movprfx\t%0., %1/z, %2.;<sve_int_op>\t%0., %1/m, %0., %3. movprfx\t%0., %1/m, %2.;<sve_int_op>\t%0., %1/m, %0., %3. #” “&& reload_completed && register_operand (operands[4], mode) && !rtx_equal_p (operands[0], operands[4])” { emit_insn (gen_vcond_mask_ (operands[0], operands[2], operands[4], operands[1])); operands[4] = operands[2] = operands[0]; } [(set_attr “movprfx” “yes”)] )

;; ------------------------------------------------------------------------- ;; ---- [INT] Addition ;; ------------------------------------------------------------------------- ;; Includes: ;; - ADD ;; - DECB ;; - DECD ;; - DECH ;; - DECW ;; - INCB ;; - INCD ;; - INCH ;; - INCW ;; - SUB ;; -------------------------------------------------------------------------

(define_insn “add3” [(set (match_operand:SVE_I 0 “register_operand” “=w, w, w, ?w, ?w, w”) (plus:SVE_I (match_operand:SVE_I 1 “register_operand” “%0, 0, 0, w, w, w”) (match_operand:SVE_I 2 “aarch64_sve_add_operand” “vsa, vsn, vsi, vsa, vsn, w”)))] “TARGET_SVE” "@ add\t%0., %0., #%D2 sub\t%0., %0., #%N2

  • return aarch64_output_sve_vector_inc_dec ("%0.", operands[2]); movprfx\t%0, %1;add\t%0., %0., #%D2 movprfx\t%0, %1;sub\t%0., %0., #%N2 add\t%0., %1., %2." [(set_attr “movprfx” “,,,yes,yes,”)] )

;; Merging forms are handled through SVE_INT_BINARY.

;; ------------------------------------------------------------------------- ;; ---- [INT] Subtraction ;; ------------------------------------------------------------------------- ;; Includes: ;; - SUB ;; - SUBR ;; -------------------------------------------------------------------------

(define_insn “sub3” [(set (match_operand:SVE_I 0 “register_operand” “=w, w, ?&w”) (minus:SVE_I (match_operand:SVE_I 1 “aarch64_sve_arith_operand” “w, vsa, vsa”) (match_operand:SVE_I 2 “register_operand” “w, 0, w”)))] “TARGET_SVE” “@ sub\t%0., %1., %2. subr\t%0., %0., #%D1 movprfx\t%0, %2;subr\t%0., %0., #%D1” [(set_attr “movprfx” “,,yes”)] )

;; Merging forms are handled through SVE_INT_BINARY.

;; ------------------------------------------------------------------------- ;; ---- [INT] Take address ;; ------------------------------------------------------------------------- ;; Includes: ;; - ADR ;; -------------------------------------------------------------------------

;; An unshifted and unscaled ADR. This is functionally equivalent to an ADD, ;; but the svadrb intrinsics should preserve the user's choice. (define_insn “@aarch64_adr” [(set (match_operand:SVE_FULL_SDI 0 “register_operand” “=w”) (unspec:SVE_FULL_SDI [(match_operand:SVE_FULL_SDI 1 “register_operand” “w”) (match_operand:SVE_FULL_SDI 2 “register_operand” “w”)] UNSPEC_ADR))] “TARGET_SVE” “adr\t%0., [%1., %2.]” )

;; Same, but with the offset being sign-extended from the low 32 bits. (define_insn_and_rewrite “*aarch64_adr_sxtw” [(set (match_operand:VNx2DI 0 “register_operand” “=w”) (unspec:VNx2DI [(match_operand:VNx2DI 1 “register_operand” “w”) (unspec:VNx2DI [(match_operand 3) (sign_extend:VNx2DI (truncate:VNx2SI (match_operand:VNx2DI 2 “register_operand” “w”)))] UNSPEC_PRED_X)] UNSPEC_ADR))] “TARGET_SVE” “adr\t%0.d, [%1.d, %2.d, sxtw]” “&& !CONSTANT_P (operands[3])” { operands[3] = CONSTM1_RTX (VNx2BImode); } )

;; Same, but with the offset being zero-extended from the low 32 bits. (define_insn “*aarch64_adr_uxtw_unspec” [(set (match_operand:VNx2DI 0 “register_operand” “=w”) (unspec:VNx2DI [(match_operand:VNx2DI 1 “register_operand” “w”) (and:VNx2DI (match_operand:VNx2DI 2 “register_operand” “w”) (match_operand:VNx2DI 3 “aarch64_sve_uxtw_immediate”))] UNSPEC_ADR))] “TARGET_SVE” “adr\t%0.d, [%1.d, %2.d, uxtw]” )

;; Same, matching as a PLUS rather than unspec. (define_insn “*aarch64_adr_uxtw_and” [(set (match_operand:VNx2DI 0 “register_operand” “=w”) (plus:VNx2DI (and:VNx2DI (match_operand:VNx2DI 2 “register_operand” “w”) (match_operand:VNx2DI 3 “aarch64_sve_uxtw_immediate”)) (match_operand:VNx2DI 1 “register_operand” “w”)))] “TARGET_SVE” “adr\t%0.d, [%1.d, %2.d, uxtw]” )

;; ADR with a nonzero shift. (define_expand “@aarch64_adr_shift” [(set (match_operand:SVE_FULL_SDI 0 “register_operand”) (plus:SVE_FULL_SDI (unspec:SVE_FULL_SDI [(match_dup 4) (ashift:SVE_FULL_SDI (match_operand:SVE_FULL_SDI 2 “register_operand”) (match_operand:SVE_FULL_SDI 3 “const_1_to_3_operand”))] UNSPEC_PRED_X) (match_operand:SVE_FULL_SDI 1 “register_operand”)))] “TARGET_SVE” { operands[4] = CONSTM1_RTX (mode); } )

(define_insn_and_rewrite “*aarch64_adr_shift” [(set (match_operand:SVE_24I 0 “register_operand” “=w”) (plus:SVE_24I (unspec:SVE_24I [(match_operand 4) (ashift:SVE_24I (match_operand:SVE_24I 2 “register_operand” “w”) (match_operand:SVE_24I 3 “const_1_to_3_operand”))] UNSPEC_PRED_X) (match_operand:SVE_24I 1 “register_operand” “w”)))] “TARGET_SVE” “adr\t%0., [%1., %2., lsl %3]” “&& !CONSTANT_P (operands[4])” { operands[4] = CONSTM1_RTX (mode); } )

;; Same, but with the index being sign-extended from the low 32 bits. (define_insn_and_rewrite “*aarch64_adr_shift_sxtw” [(set (match_operand:VNx2DI 0 “register_operand” “=w”) (plus:VNx2DI (unspec:VNx2DI [(match_operand 4) (ashift:VNx2DI (unspec:VNx2DI [(match_operand 5) (sign_extend:VNx2DI (truncate:VNx2SI (match_operand:VNx2DI 2 “register_operand” “w”)))] UNSPEC_PRED_X) (match_operand:VNx2DI 3 “const_1_to_3_operand”))] UNSPEC_PRED_X) (match_operand:VNx2DI 1 “register_operand” “w”)))] “TARGET_SVE” “adr\t%0.d, [%1.d, %2.d, sxtw %3]” “&& (!CONSTANT_P (operands[4]) || !CONSTANT_P (operands[5]))” { operands[5] = operands[4] = CONSTM1_RTX (VNx2BImode); } )

;; Same, but with the index being zero-extended from the low 32 bits. (define_insn_and_rewrite “*aarch64_adr_shift_uxtw” [(set (match_operand:VNx2DI 0 “register_operand” “=w”) (plus:VNx2DI (unspec:VNx2DI [(match_operand 5) (ashift:VNx2DI (and:VNx2DI (match_operand:VNx2DI 2 “register_operand” “w”) (match_operand:VNx2DI 4 “aarch64_sve_uxtw_immediate”)) (match_operand:VNx2DI 3 “const_1_to_3_operand”))] UNSPEC_PRED_X) (match_operand:VNx2DI 1 “register_operand” “w”)))] “TARGET_SVE” “adr\t%0.d, [%1.d, %2.d, uxtw %3]” “&& !CONSTANT_P (operands[5])” { operands[5] = CONSTM1_RTX (VNx2BImode); } )

;; ------------------------------------------------------------------------- ;; ---- [INT] Absolute difference ;; ------------------------------------------------------------------------- ;; Includes: ;; - SABD ;; - UABD ;; -------------------------------------------------------------------------

;; Unpredicated integer absolute difference. (define_expand “abd_3” [(use (match_operand:SVE_I 0 “register_operand”)) (USMAX:SVE_I (match_operand:SVE_I 1 “register_operand”) (match_operand:SVE_I 2 “register_operand”))] “TARGET_SVE” { rtx pred = aarch64_ptrue_reg (mode); emit_insn (gen_aarch64_pred_abd (operands[0], pred, operands[1], operands[2])); DONE; } )

;; Predicated integer absolute difference. (define_insn “@aarch64_pred_abd” [(set (match_operand:SVE_I 0 “register_operand” “=w, ?&w”) (minus:SVE_I (unspec:SVE_I [(match_operand: 1 “register_operand” “Upl, Upl”) (USMAX:SVE_I (match_operand:SVE_I 2 “register_operand” “%0, w”) (match_operand:SVE_I 3 “register_operand” “w, w”))] UNSPEC_PRED_X) (unspec:SVE_I [(match_dup 1) (<max_opp>:SVE_I (match_dup 2) (match_dup 3))] UNSPEC_PRED_X)))] “TARGET_SVE” “@ abd\t%0., %1/m, %0., %3. movprfx\t%0, %2;abd\t%0., %1/m, %0., %3.” [(set_attr “movprfx” “*,yes”)] )

(define_expand “@aarch64_cond_abd” [(set (match_operand:SVE_FULL_I 0 “register_operand”) (unspec:SVE_FULL_I [(match_operand: 1 “register_operand”) (minus:SVE_FULL_I (unspec:SVE_FULL_I [(match_dup 1) (USMAX:SVE_FULL_I (match_operand:SVE_FULL_I 2 “register_operand”) (match_operand:SVE_FULL_I 3 “register_operand”))] UNSPEC_PRED_X) (unspec:SVE_FULL_I [(match_dup 1) (<max_opp>:SVE_FULL_I (match_dup 2) (match_dup 3))] UNSPEC_PRED_X)) (match_operand:SVE_FULL_I 4 “aarch64_simd_reg_or_zero”)] UNSPEC_SEL))] “TARGET_SVE” { if (rtx_equal_p (operands[3], operands[4])) std::swap (operands[2], operands[3]); })

;; Predicated integer absolute difference, merging with the first input. (define_insn_and_rewrite “*aarch64_cond_abd_2” [(set (match_operand:SVE_I 0 “register_operand” “=w, ?&w”) (unspec:SVE_I [(match_operand: 1 “register_operand” “Upl, Upl”) (minus:SVE_I (unspec:SVE_I [(match_operand 4) (USMAX:SVE_I (match_operand:SVE_I 2 “register_operand” “0, w”) (match_operand:SVE_I 3 “register_operand” “w, w”))] UNSPEC_PRED_X) (unspec:SVE_I [(match_operand 5) (<max_opp>:SVE_I (match_dup 2) (match_dup 3))] UNSPEC_PRED_X)) (match_dup 2)] UNSPEC_SEL))] “TARGET_SVE” “@ abd\t%0., %1/m, %0., %3. movprfx\t%0, %2;abd\t%0., %1/m, %0., %3.” “&& (!CONSTANT_P (operands[4]) || !CONSTANT_P (operands[5]))” { operands[4] = operands[5] = CONSTM1_RTX (mode); } [(set_attr “movprfx” “*,yes”)] )

;; Predicated integer absolute difference, merging with the second input. (define_insn_and_rewrite “*aarch64_cond_abd_3” [(set (match_operand:SVE_I 0 “register_operand” “=w, ?&w”) (unspec:SVE_I [(match_operand: 1 “register_operand” “Upl, Upl”) (minus:SVE_I (unspec:SVE_I [(match_operand 4) (USMAX:SVE_I (match_operand:SVE_I 2 “register_operand” “w, w”) (match_operand:SVE_I 3 “register_operand” “0, w”))] UNSPEC_PRED_X) (unspec:SVE_I [(match_operand 5) (<max_opp>:SVE_I (match_dup 2) (match_dup 3))] UNSPEC_PRED_X)) (match_dup 3)] UNSPEC_SEL))] “TARGET_SVE” “@ abd\t%0., %1/m, %0., %2. movprfx\t%0, %3;abd\t%0., %1/m, %0., %2.” “&& (!CONSTANT_P (operands[4]) || !CONSTANT_P (operands[5]))” { operands[4] = operands[5] = CONSTM1_RTX (mode); } [(set_attr “movprfx” “*,yes”)] )

;; Predicated integer absolute difference, merging with an independent value. (define_insn_and_rewrite “*aarch64_cond_abd_any” [(set (match_operand:SVE_I 0 “register_operand” “=&w, &w, &w, &w, ?&w”) (unspec:SVE_I [(match_operand: 1 “register_operand” “Upl, Upl, Upl, Upl, Upl”) (minus:SVE_I (unspec:SVE_I [(match_operand 5) (USMAX:SVE_I (match_operand:SVE_I 2 “register_operand” “0, w, w, w, w”) (match_operand:SVE_I 3 “register_operand” “w, 0, w, w, w”))] UNSPEC_PRED_X) (unspec:SVE_I [(match_operand 6) (<max_opp>:SVE_I (match_dup 2) (match_dup 3))] UNSPEC_PRED_X)) (match_operand:SVE_I 4 “aarch64_simd_reg_or_zero” “Dz, Dz, Dz, 0, w”)] UNSPEC_SEL))] “TARGET_SVE && !rtx_equal_p (operands[2], operands[4]) && !rtx_equal_p (operands[3], operands[4])” “@ movprfx\t%0., %1/z, %0.;abd\t%0., %1/m, %0., %3. movprfx\t%0., %1/z, %0.;abd\t%0., %1/m, %0., %2. movprfx\t%0., %1/z, %2.;abd\t%0., %1/m, %0., %3. movprfx\t%0., %1/m, %2.;abd\t%0., %1/m, %0., %3. #” “&& 1” { if (!CONSTANT_P (operands[5]) || !CONSTANT_P (operands[6])) operands[5] = operands[6] = CONSTM1_RTX (mode); else if (reload_completed && register_operand (operands[4], mode) && !rtx_equal_p (operands[0], operands[4])) { emit_insn (gen_vcond_mask_ (operands[0], operands[2], operands[4], operands[1])); operands[4] = operands[2] = operands[0]; } else FAIL; } [(set_attr “movprfx” “yes”)] )

;; ------------------------------------------------------------------------- ;; ---- [INT] Saturating addition and subtraction ;; ------------------------------------------------------------------------- ;; - SQADD ;; - SQSUB ;; - UQADD ;; - UQSUB ;; -------------------------------------------------------------------------

;; Unpredicated saturating signed addition and subtraction. (define_insn “@aarch64_sve_” [(set (match_operand:SVE_FULL_I 0 “register_operand” “=w, w, ?&w, ?&w, w”) (SBINQOPS:SVE_FULL_I (match_operand:SVE_FULL_I 1 “register_operand” “0, 0, w, w, w”) (match_operand:SVE_FULL_I 2 “aarch64_sve_sqadd_operand” “vsQ, vsS, vsQ, vsS, w”)))] “TARGET_SVE” “@ <binqops_op>\t%0., %0., #%D2 <binqops_op_rev>\t%0., %0., #%N2 movprfx\t%0, %1;<binqops_op>\t%0., %0., #%D2 movprfx\t%0, %1;<binqops_op_rev>\t%0., %0., #%N2 <binqops_op>\t%0., %1., %2.” [(set_attr “movprfx” “,,yes,yes,*”)] )

;; Unpredicated saturating unsigned addition and subtraction. (define_insn “@aarch64_sve_” [(set (match_operand:SVE_FULL_I 0 “register_operand” “=w, ?&w, w”) (UBINQOPS:SVE_FULL_I (match_operand:SVE_FULL_I 1 “register_operand” “0, w, w”) (match_operand:SVE_FULL_I 2 “aarch64_sve_arith_operand” “vsa, vsa, w”)))] “TARGET_SVE” “@ <binqops_op>\t%0., %0., #%D2 movprfx\t%0, %1;<binqops_op>\t%0., %0., #%D2 <binqops_op>\t%0., %1., %2.” [(set_attr “movprfx” “,yes,”)] )

;; ------------------------------------------------------------------------- ;; ---- [INT] Highpart multiplication ;; ------------------------------------------------------------------------- ;; Includes: ;; - SMULH ;; - UMULH ;; -------------------------------------------------------------------------

;; Unpredicated highpart multiplication. (define_expand “mul3_highpart” [(set (match_operand:SVE_I 0 “register_operand”) (unspec:SVE_I [(match_dup 3) (unspec:SVE_I [(match_operand:SVE_I 1 “register_operand”) (match_operand:SVE_I 2 “register_operand”)] MUL_HIGHPART)] UNSPEC_PRED_X))] “TARGET_SVE” { operands[3] = aarch64_ptrue_reg (mode); } )

;; Predicated highpart multiplication. (define_insn “@aarch64_pred_” [(set (match_operand:SVE_I 0 “register_operand” “=w, ?&w”) (unspec:SVE_I [(match_operand: 1 “register_operand” “Upl, Upl”) (unspec:SVE_I [(match_operand:SVE_I 2 “register_operand” “%0, w”) (match_operand:SVE_I 3 “register_operand” “w, w”)] MUL_HIGHPART)] UNSPEC_PRED_X))] “TARGET_SVE” “@ mulh\t%0., %1/m, %0., %3. movprfx\t%0, %2;mulh\t%0., %1/m, %0., %3.” [(set_attr “movprfx” “*,yes”)] )

;; Predicated highpart multiplications with merging. (define_expand “@cond_” [(set (match_operand:SVE_FULL_I 0 “register_operand”) (unspec:SVE_FULL_I [(match_operand: 1 “register_operand”) (unspec:SVE_FULL_I [(match_operand:SVE_FULL_I 2 “register_operand”) (match_operand:SVE_FULL_I 3 “register_operand”)] MUL_HIGHPART) (match_operand:SVE_FULL_I 4 “aarch64_simd_reg_or_zero”)] UNSPEC_SEL))] “TARGET_SVE” { /* Only target code is aware of these operations, so we don't need to handle the fully-general case. */ gcc_assert (rtx_equal_p (operands[2], operands[4]) || CONSTANT_P (operands[4])); })

;; Predicated highpart multiplications, merging with the first input. (define_insn “*cond__2” [(set (match_operand:SVE_FULL_I 0 “register_operand” “=w, ?&w”) (unspec:SVE_FULL_I [(match_operand: 1 “register_operand” “Upl, Upl”) (unspec:SVE_FULL_I [(match_operand:SVE_FULL_I 2 “register_operand” “0, w”) (match_operand:SVE_FULL_I 3 “register_operand” “w, w”)] MUL_HIGHPART) (match_dup 2)] UNSPEC_SEL))] “TARGET_SVE” “@ <sve_int_op>\t%0., %1/m, %0., %3. movprfx\t%0, %2;<sve_int_op>\t%0., %1/m, %0., %3.” [(set_attr “movprfx” “*,yes”)])

;; Predicated highpart multiplications, merging with zero. (define_insn “*cond__z” [(set (match_operand:SVE_FULL_I 0 “register_operand” “=&w, &w”) (unspec:SVE_FULL_I [(match_operand: 1 “register_operand” “Upl, Upl”) (unspec:SVE_FULL_I [(match_operand:SVE_FULL_I 2 “register_operand” “%0, w”) (match_operand:SVE_FULL_I 3 “register_operand” “w, w”)] MUL_HIGHPART) (match_operand:SVE_FULL_I 4 “aarch64_simd_imm_zero”)] UNSPEC_SEL))] “TARGET_SVE” “@ movprfx\t%0., %1/z, %0.;<sve_int_op>\t%0., %1/m, %0., %3. movprfx\t%0., %1/z, %2.;<sve_int_op>\t%0., %1/m, %0., %3.” [(set_attr “movprfx” “yes”)])

;; ------------------------------------------------------------------------- ;; ---- [INT] Division ;; ------------------------------------------------------------------------- ;; Includes: ;; - SDIV ;; - SDIVR ;; - UDIV ;; - UDIVR ;; -------------------------------------------------------------------------

;; Unpredicated integer division. (define_expand “3” [(set (match_operand:SVE_FULL_SDI 0 “register_operand”) (unspec:SVE_FULL_SDI [(match_dup 3) (SVE_INT_BINARY_SD:SVE_FULL_SDI (match_operand:SVE_FULL_SDI 1 “register_operand”) (match_operand:SVE_FULL_SDI 2 “register_operand”))] UNSPEC_PRED_X))] “TARGET_SVE” { operands[3] = aarch64_ptrue_reg (mode); } )

;; Integer division predicated with a PTRUE. (define_insn “@aarch64_pred_” [(set (match_operand:SVE_FULL_SDI 0 “register_operand” “=w, w, ?&w”) (unspec:SVE_FULL_SDI [(match_operand: 1 “register_operand” “Upl, Upl, Upl”) (SVE_INT_BINARY_SD:SVE_FULL_SDI (match_operand:SVE_FULL_SDI 2 “register_operand” “0, w, w”) (match_operand:SVE_FULL_SDI 3 “register_operand” “w, 0, w”))] UNSPEC_PRED_X))] “TARGET_SVE” “@ <sve_int_op>\t%0., %1/m, %0., %3. <sve_int_op>r\t%0., %1/m, %0., %2. movprfx\t%0, %2;<sve_int_op>\t%0., %1/m, %0., %3.” [(set_attr “movprfx” “,,yes”)] )

;; Predicated integer division with merging. (define_expand “@cond_” [(set (match_operand:SVE_FULL_SDI 0 “register_operand”) (unspec:SVE_FULL_SDI [(match_operand: 1 “register_operand”) (SVE_INT_BINARY_SD:SVE_FULL_SDI (match_operand:SVE_FULL_SDI 2 “register_operand”) (match_operand:SVE_FULL_SDI 3 “register_operand”)) (match_operand:SVE_FULL_SDI 4 “aarch64_simd_reg_or_zero”)] UNSPEC_SEL))] “TARGET_SVE” )

;; Predicated integer division, merging with the first input. (define_insn “*cond__2” [(set (match_operand:SVE_FULL_SDI 0 “register_operand” “=w, ?&w”) (unspec:SVE_FULL_SDI [(match_operand: 1 “register_operand” “Upl, Upl”) (SVE_INT_BINARY_SD:SVE_FULL_SDI (match_operand:SVE_FULL_SDI 2 “register_operand” “0, w”) (match_operand:SVE_FULL_SDI 3 “register_operand” “w, w”)) (match_dup 2)] UNSPEC_SEL))] “TARGET_SVE” “@ <sve_int_op>\t%0., %1/m, %0., %3. movprfx\t%0, %2;<sve_int_op>\t%0., %1/m, %0., %3.” [(set_attr “movprfx” “*,yes”)] )

;; Predicated integer division, merging with the second input. (define_insn “*cond__3” [(set (match_operand:SVE_FULL_SDI 0 “register_operand” “=w, ?&w”) (unspec:SVE_FULL_SDI [(match_operand: 1 “register_operand” “Upl, Upl”) (SVE_INT_BINARY_SD:SVE_FULL_SDI (match_operand:SVE_FULL_SDI 2 “register_operand” “w, w”) (match_operand:SVE_FULL_SDI 3 “register_operand” “0, w”)) (match_dup 3)] UNSPEC_SEL))] “TARGET_SVE” “@ <sve_int_op_rev>\t%0., %1/m, %0., %2. movprfx\t%0, %3;<sve_int_op_rev>\t%0., %1/m, %0., %2.” [(set_attr “movprfx” “*,yes”)] )

;; Predicated integer division, merging with an independent value. (define_insn_and_rewrite “*cond__any” [(set (match_operand:SVE_FULL_SDI 0 “register_operand” “=&w, &w, &w, &w, ?&w”) (unspec:SVE_FULL_SDI [(match_operand: 1 “register_operand” “Upl, Upl, Upl, Upl, Upl”) (SVE_INT_BINARY_SD:SVE_FULL_SDI (match_operand:SVE_FULL_SDI 2 “register_operand” “0, w, w, w, w”) (match_operand:SVE_FULL_SDI 3 “register_operand” “w, 0, w, w, w”)) (match_operand:SVE_FULL_SDI 4 “aarch64_simd_reg_or_zero” “Dz, Dz, Dz, 0, w”)] UNSPEC_SEL))] “TARGET_SVE && !rtx_equal_p (operands[2], operands[4]) && !rtx_equal_p (operands[3], operands[4])” “@ movprfx\t%0., %1/z, %0.;<sve_int_op>\t%0., %1/m, %0., %3. movprfx\t%0., %1/z, %0.;<sve_int_op_rev>\t%0., %1/m, %0., %2. movprfx\t%0., %1/z, %2.;<sve_int_op>\t%0., %1/m, %0., %3. movprfx\t%0., %1/m, %2.;<sve_int_op>\t%0., %1/m, %0., %3. #” “&& reload_completed && register_operand (operands[4], mode) && !rtx_equal_p (operands[0], operands[4])” { emit_insn (gen_vcond_mask_ (operands[0], operands[2], operands[4], operands[1])); operands[4] = operands[2] = operands[0]; } [(set_attr “movprfx” “yes”)] )

;; ------------------------------------------------------------------------- ;; ---- [INT] Binary logical operations ;; ------------------------------------------------------------------------- ;; Includes: ;; - AND ;; - EOR ;; - ORR ;; -------------------------------------------------------------------------

;; Unpredicated integer binary logical operations. (define_insn “3” [(set (match_operand:SVE_I 0 “register_operand” “=w, ?w, w”) (LOGICAL:SVE_I (match_operand:SVE_I 1 “register_operand” “%0, w, w”) (match_operand:SVE_I 2 “aarch64_sve_logical_operand” “vsl, vsl, w”)))] “TARGET_SVE” “@ \t%0., %0., #%C2 movprfx\t%0, %1;\t%0., %0., #%C2 \t%0.d, %1.d, %2.d” [(set_attr “movprfx” “,yes,”)] )

;; Merging forms are handled through SVE_INT_BINARY.

;; ------------------------------------------------------------------------- ;; ---- [INT] Binary logical operations (inverted second input) ;; ------------------------------------------------------------------------- ;; Includes: ;; - BIC ;; -------------------------------------------------------------------------

;; Unpredicated BIC. (define_expand “@aarch64_bic” [(set (match_operand:SVE_I 0 “register_operand”) (and:SVE_I (unspec:SVE_I [(match_dup 3) (not:SVE_I (match_operand:SVE_I 2 “register_operand”))] UNSPEC_PRED_X) (match_operand:SVE_I 1 “register_operand”)))] “TARGET_SVE” { operands[3] = CONSTM1_RTX (mode); } )

;; Predicated BIC. (define_insn_and_rewrite “*bic3” [(set (match_operand:SVE_I 0 “register_operand” “=w”) (and:SVE_I (unspec:SVE_I [(match_operand 3) (not:SVE_I (match_operand:SVE_I 2 “register_operand” “w”))] UNSPEC_PRED_X) (match_operand:SVE_I 1 “register_operand” “w”)))] “TARGET_SVE” “bic\t%0.d, %1.d, %2.d” “&& !CONSTANT_P (operands[3])” { operands[3] = CONSTM1_RTX (mode); } )

;; Predicated BIC with merging. (define_expand “@cond_bic” [(set (match_operand:SVE_FULL_I 0 “register_operand”) (unspec:SVE_FULL_I [(match_operand: 1 “register_operand”) (and:SVE_FULL_I (not:SVE_FULL_I (match_operand:SVE_FULL_I 3 “register_operand”)) (match_operand:SVE_FULL_I 2 “register_operand”)) (match_operand:SVE_FULL_I 4 “aarch64_simd_reg_or_zero”)] UNSPEC_SEL))] “TARGET_SVE” )

;; Predicated integer BIC, merging with the first input. (define_insn “*cond_bic_2” [(set (match_operand:SVE_I 0 “register_operand” “=w, ?&w”) (unspec:SVE_I [(match_operand: 1 “register_operand” “Upl, Upl”) (and:SVE_I (not:SVE_I (match_operand:SVE_I 3 “register_operand” “w, w”)) (match_operand:SVE_I 2 “register_operand” “0, w”)) (match_dup 2)] UNSPEC_SEL))] “TARGET_SVE” “@ bic\t%0., %1/m, %0., %3. movprfx\t%0, %2;bic\t%0., %1/m, %0., %3.” [(set_attr “movprfx” “*,yes”)] )

;; Predicated integer BIC, merging with an independent value. (define_insn_and_rewrite “*cond_bic_any” [(set (match_operand:SVE_I 0 “register_operand” “=&w, &w, &w, ?&w”) (unspec:SVE_I [(match_operand: 1 “register_operand” “Upl, Upl, Upl, Upl”) (and:SVE_I (not:SVE_I (match_operand:SVE_I 3 “register_operand” “w, w, w, w”)) (match_operand:SVE_I 2 “register_operand” “0, w, w, w”)) (match_operand:SVE_I 4 “aarch64_simd_reg_or_zero” “Dz, Dz, 0, w”)] UNSPEC_SEL))] “TARGET_SVE && !rtx_equal_p (operands[2], operands[4])” “@ movprfx\t%0., %1/z, %0.;bic\t%0., %1/m, %0., %3. movprfx\t%0., %1/z, %2.;bic\t%0., %1/m, %0., %3. movprfx\t%0., %1/m, %2.;bic\t%0., %1/m, %0., %3. #” “&& reload_completed && register_operand (operands[4], mode) && !rtx_equal_p (operands[0], operands[4])” { emit_insn (gen_vcond_mask_ (operands[0], operands[2], operands[4], operands[1])); operands[4] = operands[2] = operands[0]; } [(set_attr “movprfx” “yes”)] )

;; ------------------------------------------------------------------------- ;; ---- [INT] Shifts (rounding towards -Inf) ;; ------------------------------------------------------------------------- ;; Includes: ;; - ASR ;; - ASRR ;; - LSL ;; - LSLR ;; - LSR ;; - LSRR ;; -------------------------------------------------------------------------

;; Unpredicated shift by a scalar, which expands into one of the vector ;; shifts below. (define_expand “ASHIFT:optab3” [(set (match_operand:SVE_I 0 “register_operand”) (ASHIFT:SVE_I (match_operand:SVE_I 1 “register_operand”) (match_operand: 2 “general_operand”)))] “TARGET_SVE” { rtx amount; if (CONST_INT_P (operands[2])) { amount = gen_const_vec_duplicate (mode, operands[2]); if (!aarch64_sve_shift_operand (operands[2], mode)) amount = force_reg (mode, amount); } else { amount = convert_to_mode (mode, operands[2], 0); amount = expand_vector_broadcast (mode, amount); } emit_insn (gen_v3 (operands[0], operands[1], amount)); DONE; } )

;; Unpredicated shift by a vector. (define_expand “v3” [(set (match_operand:SVE_I 0 “register_operand”) (unspec:SVE_I [(match_dup 3) (ASHIFT:SVE_I (match_operand:SVE_I 1 “register_operand”) (match_operand:SVE_I 2 “aarch64_sve_shift_operand”))] UNSPEC_PRED_X))] “TARGET_SVE” { operands[3] = aarch64_ptrue_reg (mode); } )

;; Shift by a vector, predicated with a PTRUE. We don‘t actually need ;; the predicate for the first alternative, but using Upa or X isn’t ;; likely to gain much and would make the instruction seem less uniform ;; to the register allocator. (define_insn_and_split “@aarch64_pred_” [(set (match_operand:SVE_I 0 “register_operand” “=w, w, w, ?&w”) (unspec:SVE_I [(match_operand: 1 “register_operand” “Upl, Upl, Upl, Upl”) (ASHIFT:SVE_I (match_operand:SVE_I 2 “register_operand” “w, 0, w, w”) (match_operand:SVE_I 3 “aarch64_sve_shift_operand” “D, w, 0, w”))] UNSPEC_PRED_X))] “TARGET_SVE” "@

\t%0., %1/m, %0., %3. r\t%0., %1/m, %3., %2. movprfx\t%0, %2;\t%0., %1/m, %0., %3." “&& reload_completed && !register_operand (operands[3], mode)” [(set (match_dup 0) (ASHIFT:SVE_I (match_dup 2) (match_dup 3)))] "" [(set_attr “movprfx” “,,*,yes”)] )

;; Unpredicated shift operations by a constant (post-RA only). ;; These are generated by splitting a predicated instruction whose ;; predicate is unused. (define_insn “*post_ra_v3” [(set (match_operand:SVE_I 0 “register_operand” “=w”) (ASHIFT:SVE_I (match_operand:SVE_I 1 “register_operand” “w”) (match_operand:SVE_I 2 “aarch64_simd_shift_imm”)))] “TARGET_SVE && reload_completed” “\t%0., %1., #%2” )

;; Predicated integer shift, merging with the first input. (define_insn “*cond__2_const” [(set (match_operand:SVE_I 0 “register_operand” “=w, ?&w”) (unspec:SVE_I [(match_operand: 1 “register_operand” “Upl, Upl”) (ASHIFT:SVE_I (match_operand:SVE_I 2 “register_operand” “0, w”) (match_operand:SVE_I 3 “aarch64_simd_shift_imm”)) (match_dup 2)] UNSPEC_SEL))] “TARGET_SVE” “@ \t%0., %1/m, %0., #%3 movprfx\t%0, %2;\t%0., %1/m, %0., #%3” [(set_attr “movprfx” “*,yes”)] )

;; Predicated integer shift, merging with an independent value. (define_insn_and_rewrite “*cond__any_const” [(set (match_operand:SVE_I 0 “register_operand” “=w, &w, ?&w”) (unspec:SVE_I [(match_operand: 1 “register_operand” “Upl, Upl, Upl”) (ASHIFT:SVE_I (match_operand:SVE_I 2 “register_operand” “w, w, w”) (match_operand:SVE_I 3 “aarch64_simd_shift_imm”)) (match_operand:SVE_I 4 “aarch64_simd_reg_or_zero” “Dz, 0, w”)] UNSPEC_SEL))] “TARGET_SVE && !rtx_equal_p (operands[2], operands[4])” “@ movprfx\t%0., %1/z, %2.;\t%0., %1/m, %0., #%3 movprfx\t%0., %1/m, %2.;\t%0., %1/m, %0., #%3 #” “&& reload_completed && register_operand (operands[4], mode) && !rtx_equal_p (operands[0], operands[4])” { emit_insn (gen_vcond_mask_ (operands[0], operands[2], operands[4], operands[1])); operands[4] = operands[2] = operands[0]; } [(set_attr “movprfx” “yes”)] )

;; Unpredicated shifts of narrow elements by 64-bit amounts. (define_insn “@aarch64_sve_<sve_int_op>” [(set (match_operand:SVE_FULL_BHSI 0 “register_operand” “=w”) (unspec:SVE_FULL_BHSI [(match_operand:SVE_FULL_BHSI 1 “register_operand” “w”) (match_operand:VNx2DI 2 “register_operand” “w”)] SVE_SHIFT_WIDE))] “TARGET_SVE” “<sve_int_op>\t%0., %1., %2.d” )

;; Merging predicated shifts of narrow elements by 64-bit amounts. (define_expand “@cond_<sve_int_op>” [(set (match_operand:SVE_FULL_BHSI 0 “register_operand”) (unspec:SVE_FULL_BHSI [(match_operand: 1 “register_operand”) (unspec:SVE_FULL_BHSI [(match_operand:SVE_FULL_BHSI 2 “register_operand”) (match_operand:VNx2DI 3 “register_operand”)] SVE_SHIFT_WIDE) (match_operand:SVE_FULL_BHSI 4 “aarch64_simd_reg_or_zero”)] UNSPEC_SEL))] “TARGET_SVE” )

;; Predicated shifts of narrow elements by 64-bit amounts, merging with ;; the first input. (define_insn “*cond_<sve_int_op>_m” [(set (match_operand:SVE_FULL_BHSI 0 “register_operand” “=w, ?&w”) (unspec:SVE_FULL_BHSI [(match_operand: 1 “register_operand” “Upl, Upl”) (unspec:SVE_FULL_BHSI [(match_operand:SVE_FULL_BHSI 2 “register_operand” “0, w”) (match_operand:VNx2DI 3 “register_operand” “w, w”)] SVE_SHIFT_WIDE) (match_dup 2)] UNSPEC_SEL))] “TARGET_SVE” “@ <sve_int_op>\t%0., %1/m, %0., %3.d movprfx\t%0, %2;<sve_int_op>\t%0., %1/m, %0., %3.d” [(set_attr “movprfx” “*, yes”)])

;; Predicated shifts of narrow elements by 64-bit amounts, merging with zero. (define_insn “*cond_<sve_int_op>_z” [(set (match_operand:SVE_FULL_BHSI 0 “register_operand” “=&w, &w”) (unspec:SVE_FULL_BHSI [(match_operand: 1 “register_operand” “Upl, Upl”) (unspec:SVE_FULL_BHSI [(match_operand:SVE_FULL_BHSI 2 “register_operand” “0, w”) (match_operand:VNx2DI 3 “register_operand” “w, w”)] SVE_SHIFT_WIDE) (match_operand:SVE_FULL_BHSI 4 “aarch64_simd_imm_zero”)] UNSPEC_SEL))] “TARGET_SVE” “@ movprfx\t%0., %1/z, %0.;<sve_int_op>\t%0., %1/m, %0., %3.d movprfx\t%0., %1/z, %2.;<sve_int_op>\t%0., %1/m, %0., %3.d” [(set_attr “movprfx” “yes”)])

;; ------------------------------------------------------------------------- ;; ---- [INT] Shifts (rounding towards 0) ;; ------------------------------------------------------------------------- ;; Includes: ;; - ASRD ;; - SQSHLU (SVE2) ;; - SRSHR (SVE2) ;; - URSHR (SVE2) ;; -------------------------------------------------------------------------

;; Unpredicated ASRD. (define_expand “sdiv_pow23” [(set (match_operand:SVE_I 0 “register_operand”) (unspec:SVE_I [(match_dup 3) (unspec:SVE_I [(match_operand:SVE_I 1 “register_operand”) (match_operand 2 “aarch64_simd_rshift_imm”)] UNSPEC_ASRD)] UNSPEC_PRED_X))] “TARGET_SVE” { operands[3] = aarch64_ptrue_reg (mode); } )

;; Predicated ASRD. (define_insn “*sdiv_pow23” [(set (match_operand:SVE_I 0 “register_operand” “=w, ?&w”) (unspec:SVE_I [(match_operand: 1 “register_operand” “Upl, Upl”) (unspec:SVE_I [(match_operand:SVE_I 2 “register_operand” “0, w”) (match_operand:SVE_I 3 “aarch64_simd_rshift_imm”)] UNSPEC_ASRD)] UNSPEC_PRED_X))] “TARGET_SVE” “@ asrd\t%0., %1/m, %0., #%3 movprfx\t%0, %2;asrd\t%0., %1/m, %0., #%3” [(set_attr “movprfx” “*,yes”)])

;; Predicated shift with merging. (define_expand “@cond_<sve_int_op>” [(set (match_operand:SVE_I 0 “register_operand”) (unspec:SVE_I [(match_operand: 1 “register_operand”) (unspec:SVE_I [(match_dup 5) (unspec:SVE_I [(match_operand:SVE_I 2 “register_operand”) (match_operand:SVE_I 3 “aarch64_simd_shift_imm”)] SVE_INT_SHIFT_IMM)] UNSPEC_PRED_X) (match_operand:SVE_I 4 “aarch64_simd_reg_or_zero”)] UNSPEC_SEL))] “TARGET_SVE” { operands[5] = aarch64_ptrue_reg (mode); } )

;; Predicated shift, merging with the first input. (define_insn_and_rewrite “*cond_<sve_int_op>_2” [(set (match_operand:SVE_I 0 “register_operand” “=w, ?&w”) (unspec:SVE_I [(match_operand: 1 “register_operand” “Upl, Upl”) (unspec:SVE_I [(match_operand 4) (unspec:SVE_I [(match_operand:SVE_I 2 “register_operand” “0, w”) (match_operand:SVE_I 3 “aarch64_simd_shift_imm”)] SVE_INT_SHIFT_IMM)] UNSPEC_PRED_X) (match_dup 2)] UNSPEC_SEL))] “TARGET_SVE” “@ <sve_int_op>\t%0., %1/m, %0., #%3 movprfx\t%0, %2;<sve_int_op>\t%0., %1/m, %0., #%3” “&& !CONSTANT_P (operands[4])” { operands[4] = CONSTM1_RTX (mode); } [(set_attr “movprfx” “*,yes”)])

;; Predicated shift, merging with an independent value. (define_insn_and_rewrite “*cond_<sve_int_op>_any” [(set (match_operand:SVE_I 0 “register_operand” “=w, &w, ?&w”) (unspec:SVE_I [(match_operand: 1 “register_operand” “Upl, Upl, Upl”) (unspec:SVE_I [(match_operand 5) (unspec:SVE_I [(match_operand:SVE_I 2 “register_operand” “w, w, w”) (match_operand:SVE_I 3 “aarch64_simd_shift_imm”)] SVE_INT_SHIFT_IMM)] UNSPEC_PRED_X) (match_operand:SVE_I 4 “aarch64_simd_reg_or_zero” “Dz, 0, w”)] UNSPEC_SEL))] “TARGET_SVE && !rtx_equal_p (operands[2], operands[4])” “@ movprfx\t%0., %1/z, %2.;<sve_int_op>\t%0., %1/m, %0., #%3 movprfx\t%0., %1/m, %2.;<sve_int_op>\t%0., %1/m, %0., #%3 #” “&& reload_completed && register_operand (operands[4], mode) && !rtx_equal_p (operands[0], operands[4])” { emit_insn (gen_vcond_mask_ (operands[0], operands[2], operands[4], operands[1])); operands[4] = operands[2] = operands[0]; } [(set_attr “movprfx” “yes”)] )

;; ------------------------------------------------------------------------- ;; ---- [FP<-INT] General binary arithmetic corresponding to unspecs ;; ------------------------------------------------------------------------- ;; Includes: ;; - FSCALE ;; - FTSMUL ;; - FTSSEL ;; -------------------------------------------------------------------------

;; Unpredicated floating-point binary operations that take an integer as ;; their second operand. (define_insn “@aarch64_sve_” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=w”) (unspec:SVE_FULL_F [(match_operand:SVE_FULL_F 1 “register_operand” “w”) (match_operand:<V_INT_EQUIV> 2 “register_operand” “w”)] SVE_FP_BINARY_INT))] “TARGET_SVE” “<sve_fp_op>\t%0., %1., %2.” )

;; Predicated floating-point binary operations that take an integer ;; as their second operand. (define_insn “@aarch64_pred_” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=w, ?&w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl”) (match_operand:SI 4 “aarch64_sve_gp_strictness”) (match_operand:SVE_FULL_F 2 “register_operand” “0, w”) (match_operand:<V_INT_EQUIV> 3 “register_operand” “w, w”)] SVE_COND_FP_BINARY_INT))] “TARGET_SVE” “@ <sve_fp_op>\t%0., %1/m, %0., %3. movprfx\t%0, %2;<sve_fp_op>\t%0., %1/m, %0., %3.” [(set_attr “movprfx” “*,yes”)] )

;; Predicated floating-point binary operations with merging, taking an ;; integer as their second operand. (define_expand “@cond_” [(set (match_operand:SVE_FULL_F 0 “register_operand”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand”) (unspec:SVE_FULL_F [(match_dup 1) (const_int SVE_STRICT_GP) (match_operand:SVE_FULL_F 2 “register_operand”) (match_operand:<V_INT_EQUIV> 3 “register_operand”)] SVE_COND_FP_BINARY_INT) (match_operand:SVE_FULL_F 4 “aarch64_simd_reg_or_zero”)] UNSPEC_SEL))] “TARGET_SVE” )

;; Predicated floating-point binary operations that take an integer as their ;; second operand, with inactive lanes coming from the first operand. (define_insn_and_rewrite “*cond__2_relaxed” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=w, ?&w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl”) (unspec:SVE_FULL_F [(match_operand 4) (const_int SVE_RELAXED_GP) (match_operand:SVE_FULL_F 2 “register_operand” “0, w”) (match_operand:<V_INT_EQUIV> 3 “register_operand” “w, w”)] SVE_COND_FP_BINARY_INT) (match_dup 2)] UNSPEC_SEL))] “TARGET_SVE” “@ <sve_fp_op>\t%0., %1/m, %0., %3. movprfx\t%0, %2;<sve_fp_op>\t%0., %1/m, %0., %3.” “&& !rtx_equal_p (operands[1], operands[4])” { operands[4] = copy_rtx (operands[1]); } [(set_attr “movprfx” “*,yes”)] )

(define_insn “*cond__2_strict” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=w, ?&w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl”) (unspec:SVE_FULL_F [(match_dup 1) (const_int SVE_STRICT_GP) (match_operand:SVE_FULL_F 2 “register_operand” “0, w”) (match_operand:<V_INT_EQUIV> 3 “register_operand” “w, w”)] SVE_COND_FP_BINARY_INT) (match_dup 2)] UNSPEC_SEL))] “TARGET_SVE” “@ <sve_fp_op>\t%0., %1/m, %0., %3. movprfx\t%0, %2;<sve_fp_op>\t%0., %1/m, %0., %3.” [(set_attr “movprfx” “*,yes”)] )

;; Predicated floating-point binary operations that take an integer as ;; their second operand, with the values of inactive lanes being distinct ;; from the other inputs. (define_insn_and_rewrite “*cond__any_relaxed” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=&w, &w, &w, ?&w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl, Upl, Upl”) (unspec:SVE_FULL_F [(match_operand 5) (const_int SVE_RELAXED_GP) (match_operand:SVE_FULL_F 2 “register_operand” “0, w, w, w”) (match_operand:<V_INT_EQUIV> 3 “register_operand” “w, w, w, w”)] SVE_COND_FP_BINARY_INT) (match_operand:SVE_FULL_F 4 “aarch64_simd_reg_or_zero” “Dz, Dz, 0, w”)] UNSPEC_SEL))] “TARGET_SVE && !rtx_equal_p (operands[2], operands[4])” “@ movprfx\t%0., %1/z, %2.;<sve_fp_op>\t%0., %1/m, %0., %3. movprfx\t%0., %1/z, %2.;<sve_fp_op>\t%0., %1/m, %0., %3. movprfx\t%0., %1/m, %2.;<sve_fp_op>\t%0., %1/m, %0., %3. #” “&& 1” { if (reload_completed && register_operand (operands[4], mode) && !rtx_equal_p (operands[0], operands[4])) { emit_insn (gen_vcond_mask_ (operands[0], operands[2], operands[4], operands[1])); operands[4] = operands[2] = operands[0]; } else if (!rtx_equal_p (operands[1], operands[5])) operands[5] = copy_rtx (operands[1]); else FAIL; } [(set_attr “movprfx” “yes”)] )

(define_insn_and_rewrite “*cond__any_strict” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=&w, &w, &w, ?&w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl, Upl, Upl”) (unspec:SVE_FULL_F [(match_dup 1) (const_int SVE_STRICT_GP) (match_operand:SVE_FULL_F 2 “register_operand” “0, w, w, w”) (match_operand:<V_INT_EQUIV> 3 “register_operand” “w, w, w, w”)] SVE_COND_FP_BINARY_INT) (match_operand:SVE_FULL_F 4 “aarch64_simd_reg_or_zero” “Dz, Dz, 0, w”)] UNSPEC_SEL))] “TARGET_SVE && !rtx_equal_p (operands[2], operands[4])” “@ movprfx\t%0., %1/z, %2.;<sve_fp_op>\t%0., %1/m, %0., %3. movprfx\t%0., %1/z, %2.;<sve_fp_op>\t%0., %1/m, %0., %3. movprfx\t%0., %1/m, %2.;<sve_fp_op>\t%0., %1/m, %0., %3. #” “&& reload_completed && register_operand (operands[4], mode) && !rtx_equal_p (operands[0], operands[4])” { emit_insn (gen_vcond_mask_ (operands[0], operands[2], operands[4], operands[1])); operands[4] = operands[2] = operands[0]; } [(set_attr “movprfx” “yes”)] )

;; ------------------------------------------------------------------------- ;; ---- [FP] General binary arithmetic corresponding to rtx codes ;; ------------------------------------------------------------------------- ;; Includes post-RA forms of: ;; - FADD ;; - FMUL ;; - FSUB ;; -------------------------------------------------------------------------

;; Unpredicated floating-point binary operations (post-RA only). ;; These are generated by splitting a predicated instruction whose ;; predicate is unused. (define_insn “*post_ra_<sve_fp_op>3” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=w”) (SVE_UNPRED_FP_BINARY:SVE_FULL_F (match_operand:SVE_FULL_F 1 “register_operand” “w”) (match_operand:SVE_FULL_F 2 “register_operand” “w”)))] “TARGET_SVE && reload_completed” “<sve_fp_op>\t%0., %1., %2.”)

;; ------------------------------------------------------------------------- ;; ---- [FP] General binary arithmetic corresponding to unspecs ;; ------------------------------------------------------------------------- ;; Includes merging forms of: ;; - FADD (constant forms handled in the “Addition” section) ;; - FDIV ;; - FDIVR ;; - FMAX ;; - FMAXNM (including #0.0 and #1.0) ;; - FMIN ;; - FMINNM (including #0.0 and #1.0) ;; - FMUL (including #0.5 and #2.0) ;; - FMULX ;; - FRECPS ;; - FRSQRTS ;; - FSUB (constant forms handled in the “Addition” section) ;; - FSUBR (constant forms handled in the “Subtraction” section) ;; -------------------------------------------------------------------------

;; Unpredicated floating-point binary operations. (define_insn “@aarch64_sve_” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=w”) (unspec:SVE_FULL_F [(match_operand:SVE_FULL_F 1 “register_operand” “w”) (match_operand:SVE_FULL_F 2 “register_operand” “w”)] SVE_FP_BINARY))] “TARGET_SVE” “<sve_fp_op>\t%0., %1., %2.” )

;; Unpredicated floating-point binary operations that need to be predicated ;; for SVE. (define_expand “3” [(set (match_operand:SVE_FULL_F 0 “register_operand”) (unspec:SVE_FULL_F [(match_dup 3) (const_int SVE_RELAXED_GP) (match_operand:SVE_FULL_F 1 “<sve_pred_fp_rhs1_operand>”) (match_operand:SVE_FULL_F 2 “<sve_pred_fp_rhs2_operand>”)] SVE_COND_FP_BINARY_OPTAB))] “TARGET_SVE” { operands[3] = aarch64_ptrue_reg (mode); } )

;; Predicated floating-point binary operations that have no immediate forms. (define_insn “@aarch64_pred_” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=w, w, ?&w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl, Upl”) (match_operand:SI 4 “aarch64_sve_gp_strictness”) (match_operand:SVE_FULL_F 2 “register_operand” “0, w, w”) (match_operand:SVE_FULL_F 3 “register_operand” “w, 0, w”)] SVE_COND_FP_BINARY_REG))] “TARGET_SVE” “@ <sve_fp_op>\t%0., %1/m, %0., %3. <sve_fp_op_rev>\t%0., %1/m, %0., %2. movprfx\t%0, %2;<sve_fp_op>\t%0., %1/m, %0., %3.” [(set_attr “movprfx” “,,yes”)] )

;; Predicated floating-point operations with merging. (define_expand “@cond_” [(set (match_operand:SVE_FULL_F 0 “register_operand”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand”) (unspec:SVE_FULL_F [(match_dup 1) (const_int SVE_STRICT_GP) (match_operand:SVE_FULL_F 2 “<sve_pred_fp_rhs1_operand>”) (match_operand:SVE_FULL_F 3 “<sve_pred_fp_rhs2_operand>”)] SVE_COND_FP_BINARY) (match_operand:SVE_FULL_F 4 “aarch64_simd_reg_or_zero”)] UNSPEC_SEL))] “TARGET_SVE” )

;; Predicated floating-point operations, merging with the first input. (define_insn_and_rewrite “*cond__2_relaxed” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=w, ?&w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl”) (unspec:SVE_FULL_F [(match_operand 4) (const_int SVE_RELAXED_GP) (match_operand:SVE_FULL_F 2 “register_operand” “0, w”) (match_operand:SVE_FULL_F 3 “register_operand” “w, w”)] SVE_COND_FP_BINARY) (match_dup 2)] UNSPEC_SEL))] “TARGET_SVE” “@ <sve_fp_op>\t%0., %1/m, %0., %3. movprfx\t%0, %2;<sve_fp_op>\t%0., %1/m, %0., %3.” “&& !rtx_equal_p (operands[1], operands[4])” { operands[4] = copy_rtx (operands[1]); } [(set_attr “movprfx” “*,yes”)] )

(define_insn “*cond__2_strict” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=w, ?&w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl”) (unspec:SVE_FULL_F [(match_dup 1) (const_int SVE_STRICT_GP) (match_operand:SVE_FULL_F 2 “register_operand” “0, w”) (match_operand:SVE_FULL_F 3 “register_operand” “w, w”)] SVE_COND_FP_BINARY) (match_dup 2)] UNSPEC_SEL))] “TARGET_SVE” “@ <sve_fp_op>\t%0., %1/m, %0., %3. movprfx\t%0, %2;<sve_fp_op>\t%0., %1/m, %0., %3.” [(set_attr “movprfx” “*,yes”)] )

;; Same for operations that take a 1-bit constant. (define_insn_and_rewrite “*cond__2_const_relaxed” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=w, ?w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl”) (unspec:SVE_FULL_F [(match_operand 4) (const_int SVE_RELAXED_GP) (match_operand:SVE_FULL_F 2 “register_operand” “0, w”) (match_operand:SVE_FULL_F 3 “<sve_pred_fp_rhs2_immediate>”)] SVE_COND_FP_BINARY_I1) (match_dup 2)] UNSPEC_SEL))] “TARGET_SVE” “@ <sve_fp_op>\t%0., %1/m, %0., #%3 movprfx\t%0, %2;<sve_fp_op>\t%0., %1/m, %0., #%3” “&& !rtx_equal_p (operands[1], operands[4])” { operands[4] = copy_rtx (operands[1]); } [(set_attr “movprfx” “*,yes”)] )

(define_insn “*cond__2_const_strict” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=w, ?w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl”) (unspec:SVE_FULL_F [(match_dup 1) (const_int SVE_STRICT_GP) (match_operand:SVE_FULL_F 2 “register_operand” “0, w”) (match_operand:SVE_FULL_F 3 “<sve_pred_fp_rhs2_immediate>”)] SVE_COND_FP_BINARY_I1) (match_dup 2)] UNSPEC_SEL))] “TARGET_SVE” “@ <sve_fp_op>\t%0., %1/m, %0., #%3 movprfx\t%0, %2;<sve_fp_op>\t%0., %1/m, %0., #%3” [(set_attr “movprfx” “*,yes”)] )

;; Predicated floating-point operations, merging with the second input. (define_insn_and_rewrite “*cond__3_relaxed” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=w, ?&w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl”) (unspec:SVE_FULL_F [(match_operand 4) (const_int SVE_RELAXED_GP) (match_operand:SVE_FULL_F 2 “register_operand” “w, w”) (match_operand:SVE_FULL_F 3 “register_operand” “0, w”)] SVE_COND_FP_BINARY) (match_dup 3)] UNSPEC_SEL))] “TARGET_SVE” “@ <sve_fp_op_rev>\t%0., %1/m, %0., %2. movprfx\t%0, %3;<sve_fp_op_rev>\t%0., %1/m, %0., %2.” “&& !rtx_equal_p (operands[1], operands[4])” { operands[4] = copy_rtx (operands[1]); } [(set_attr “movprfx” “*,yes”)] )

(define_insn “*cond__3_strict” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=w, ?&w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl”) (unspec:SVE_FULL_F [(match_dup 1) (const_int SVE_STRICT_GP) (match_operand:SVE_FULL_F 2 “register_operand” “w, w”) (match_operand:SVE_FULL_F 3 “register_operand” “0, w”)] SVE_COND_FP_BINARY) (match_dup 3)] UNSPEC_SEL))] “TARGET_SVE” “@ <sve_fp_op_rev>\t%0., %1/m, %0., %2. movprfx\t%0, %3;<sve_fp_op_rev>\t%0., %1/m, %0., %2.” [(set_attr “movprfx” “*,yes”)] )

;; Predicated floating-point operations, merging with an independent value. (define_insn_and_rewrite “*cond__any_relaxed” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=&w, &w, &w, &w, ?&w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl, Upl, Upl, Upl”) (unspec:SVE_FULL_F [(match_operand 5) (const_int SVE_RELAXED_GP) (match_operand:SVE_FULL_F 2 “register_operand” “0, w, w, w, w”) (match_operand:SVE_FULL_F 3 “register_operand” “w, 0, w, w, w”)] SVE_COND_FP_BINARY) (match_operand:SVE_FULL_F 4 “aarch64_simd_reg_or_zero” “Dz, Dz, Dz, 0, w”)] UNSPEC_SEL))] “TARGET_SVE && !rtx_equal_p (operands[2], operands[4]) && !rtx_equal_p (operands[3], operands[4])” “@ movprfx\t%0., %1/z, %0.;<sve_fp_op>\t%0., %1/m, %0., %3. movprfx\t%0., %1/z, %0.;<sve_fp_op_rev>\t%0., %1/m, %0., %2. movprfx\t%0., %1/z, %2.;<sve_fp_op>\t%0., %1/m, %0., %3. movprfx\t%0., %1/m, %2.;<sve_fp_op>\t%0., %1/m, %0., %3. #” “&& 1” { if (reload_completed && register_operand (operands[4], mode) && !rtx_equal_p (operands[0], operands[4])) { emit_insn (gen_vcond_mask_ (operands[0], operands[2], operands[4], operands[1])); operands[4] = operands[2] = operands[0]; } else if (!rtx_equal_p (operands[1], operands[5])) operands[5] = copy_rtx (operands[1]); else FAIL; } [(set_attr “movprfx” “yes”)] )

(define_insn_and_rewrite “*cond__any_strict” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=&w, &w, &w, &w, ?&w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl, Upl, Upl, Upl”) (unspec:SVE_FULL_F [(match_dup 1) (const_int SVE_STRICT_GP) (match_operand:SVE_FULL_F 2 “register_operand” “0, w, w, w, w”) (match_operand:SVE_FULL_F 3 “register_operand” “w, 0, w, w, w”)] SVE_COND_FP_BINARY) (match_operand:SVE_FULL_F 4 “aarch64_simd_reg_or_zero” “Dz, Dz, Dz, 0, w”)] UNSPEC_SEL))] “TARGET_SVE && !rtx_equal_p (operands[2], operands[4]) && !rtx_equal_p (operands[3], operands[4])” “@ movprfx\t%0., %1/z, %0.;<sve_fp_op>\t%0., %1/m, %0., %3. movprfx\t%0., %1/z, %0.;<sve_fp_op_rev>\t%0., %1/m, %0., %2. movprfx\t%0., %1/z, %2.;<sve_fp_op>\t%0., %1/m, %0., %3. movprfx\t%0., %1/m, %2.;<sve_fp_op>\t%0., %1/m, %0., %3. #” “&& reload_completed && register_operand (operands[4], mode) && !rtx_equal_p (operands[0], operands[4])” { emit_insn (gen_vcond_mask_ (operands[0], operands[2], operands[4], operands[1])); operands[4] = operands[2] = operands[0]; } [(set_attr “movprfx” “yes”)] )

;; Same for operations that take a 1-bit constant. (define_insn_and_rewrite “*cond__any_const_relaxed” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=w, w, ?w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl, Upl”) (unspec:SVE_FULL_F [(match_operand 5) (const_int SVE_RELAXED_GP) (match_operand:SVE_FULL_F 2 “register_operand” “w, w, w”) (match_operand:SVE_FULL_F 3 “<sve_pred_fp_rhs2_immediate>”)] SVE_COND_FP_BINARY_I1) (match_operand:SVE_FULL_F 4 “aarch64_simd_reg_or_zero” “Dz, 0, w”)] UNSPEC_SEL))] “TARGET_SVE && !rtx_equal_p (operands[2], operands[4])” “@ movprfx\t%0., %1/z, %2.;<sve_fp_op>\t%0., %1/m, %0., #%3 movprfx\t%0., %1/m, %2.;<sve_fp_op>\t%0., %1/m, %0., #%3 #” “&& 1” { if (reload_completed && register_operand (operands[4], mode) && !rtx_equal_p (operands[0], operands[4])) { emit_insn (gen_vcond_mask_ (operands[0], operands[2], operands[4], operands[1])); operands[4] = operands[2] = operands[0]; } else if (!rtx_equal_p (operands[1], operands[5])) operands[5] = copy_rtx (operands[1]); else FAIL; } [(set_attr “movprfx” “yes”)] )

(define_insn_and_rewrite “*cond__any_const_strict” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=w, w, ?w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl, Upl”) (unspec:SVE_FULL_F [(match_dup 1) (const_int SVE_STRICT_GP) (match_operand:SVE_FULL_F 2 “register_operand” “w, w, w”) (match_operand:SVE_FULL_F 3 “<sve_pred_fp_rhs2_immediate>”)] SVE_COND_FP_BINARY_I1) (match_operand:SVE_FULL_F 4 “aarch64_simd_reg_or_zero” “Dz, 0, w”)] UNSPEC_SEL))] “TARGET_SVE && !rtx_equal_p (operands[2], operands[4])” “@ movprfx\t%0., %1/z, %2.;<sve_fp_op>\t%0., %1/m, %0., #%3 movprfx\t%0., %1/m, %2.;<sve_fp_op>\t%0., %1/m, %0., #%3 #” “&& reload_completed && register_operand (operands[4], mode) && !rtx_equal_p (operands[0], operands[4])” { emit_insn (gen_vcond_mask_ (operands[0], operands[2], operands[4], operands[1])); operands[4] = operands[2] = operands[0]; } [(set_attr “movprfx” “yes”)] )

;; ------------------------------------------------------------------------- ;; ---- [FP] Addition ;; ------------------------------------------------------------------------- ;; Includes: ;; - FADD ;; - FSUB ;; -------------------------------------------------------------------------

;; Predicated floating-point addition. (define_insn_and_split “@aarch64_pred_” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=w, w, w, w, ?&w, ?&w, ?&w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl, Upl, Upl, Upl, Upl, Upl”) (match_operand:SI 4 “aarch64_sve_gp_strictness” “i, i, Z, Ui1, i, i, Ui1”) (match_operand:SVE_FULL_F 2 “register_operand” “%0, 0, w, 0, w, w, w”) (match_operand:SVE_FULL_F 3 “aarch64_sve_float_arith_with_sub_operand” “vsA, vsN, w, w, vsA, vsN, w”)] SVE_COND_FP_ADD))] “TARGET_SVE” "@ fadd\t%0., %1/m, %0., #%3 fsub\t%0., %1/m, %0., #%N3

fadd\t%0., %1/m, %0., %3. movprfx\t%0, %2;fadd\t%0., %1/m, %0., #%3 movprfx\t%0, %2;fsub\t%0., %1/m, %0., #%N3 movprfx\t%0, %2;fadd\t%0., %1/m, %0., %3." ; Split the unpredicated form after reload, so that we don't have ; the unnecessary PTRUE. “&& reload_completed && register_operand (operands[3], mode) && INTVAL (operands[4]) == SVE_RELAXED_GP” [(set (match_dup 0) (plus:SVE_FULL_F (match_dup 2) (match_dup 3)))] "" [(set_attr “movprfx” “,,,,yes,yes,yes”)] )

;; Predicated floating-point addition of a constant, merging with the ;; first input. (define_insn_and_rewrite “*cond_add_2_const_relaxed” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=w, w, ?w, ?w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl, Upl, Upl”) (unspec:SVE_FULL_F [(match_operand 4) (const_int SVE_RELAXED_GP) (match_operand:SVE_FULL_F 2 “register_operand” “0, 0, w, w”) (match_operand:SVE_FULL_F 3 “aarch64_sve_float_arith_with_sub_immediate” “vsA, vsN, vsA, vsN”)] UNSPEC_COND_FADD) (match_dup 2)] UNSPEC_SEL))] “TARGET_SVE” “@ fadd\t%0., %1/m, %0., #%3 fsub\t%0., %1/m, %0., #%N3 movprfx\t%0, %2;fadd\t%0., %1/m, %0., #%3 movprfx\t%0, %2;fsub\t%0., %1/m, %0., #%N3” “&& !rtx_equal_p (operands[1], operands[4])” { operands[4] = copy_rtx (operands[1]); } [(set_attr “movprfx” “,,yes,yes”)] )

(define_insn “*cond_add_2_const_strict” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=w, w, ?w, ?w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl, Upl, Upl”) (unspec:SVE_FULL_F [(match_dup 1) (const_int SVE_STRICT_GP) (match_operand:SVE_FULL_F 2 “register_operand” “0, 0, w, w”) (match_operand:SVE_FULL_F 3 “aarch64_sve_float_arith_with_sub_immediate” “vsA, vsN, vsA, vsN”)] UNSPEC_COND_FADD) (match_dup 2)] UNSPEC_SEL))] “TARGET_SVE” “@ fadd\t%0., %1/m, %0., #%3 fsub\t%0., %1/m, %0., #%N3 movprfx\t%0, %2;fadd\t%0., %1/m, %0., #%3 movprfx\t%0, %2;fsub\t%0., %1/m, %0., #%N3” [(set_attr “movprfx” “,,yes,yes”)] )

;; Predicated floating-point addition of a constant, merging with an ;; independent value. (define_insn_and_rewrite “*cond_add_any_const_relaxed” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=w, w, w, w, ?w, ?w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl, Upl, Upl, Upl, Upl”) (unspec:SVE_FULL_F [(match_operand 5) (const_int SVE_RELAXED_GP) (match_operand:SVE_FULL_F 2 “register_operand” “w, w, w, w, w, w”) (match_operand:SVE_FULL_F 3 “aarch64_sve_float_arith_with_sub_immediate” “vsA, vsN, vsA, vsN, vsA, vsN”)] UNSPEC_COND_FADD) (match_operand:SVE_FULL_F 4 “aarch64_simd_reg_or_zero” “Dz, Dz, 0, 0, w, w”)] UNSPEC_SEL))] “TARGET_SVE && !rtx_equal_p (operands[2], operands[4])” "@ movprfx\t%0., %1/z, %2.;fadd\t%0., %1/m, %0., #%3 movprfx\t%0., %1/z, %2.;fsub\t%0., %1/m, %0., #%N3 movprfx\t%0., %1/m, %2.;fadd\t%0., %1/m, %0., #%3 movprfx\t%0., %1/m, %2.;fsub\t%0., %1/m, %0., #%N3

#" “&& 1” { if (reload_completed && register_operand (operands[4], mode) && !rtx_equal_p (operands[0], operands[4])) { emit_insn (gen_vcond_mask_ (operands[0], operands[2], operands[4], operands[1])); operands[4] = operands[2] = operands[0]; } else if (!rtx_equal_p (operands[1], operands[5])) operands[5] = copy_rtx (operands[1]); else FAIL; } [(set_attr “movprfx” “yes”)] )

(define_insn_and_rewrite “*cond_add_any_const_strict” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=w, w, w, w, ?w, ?w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl, Upl, Upl, Upl, Upl”) (unspec:SVE_FULL_F [(match_dup 1) (const_int SVE_STRICT_GP) (match_operand:SVE_FULL_F 2 “register_operand” “w, w, w, w, w, w”) (match_operand:SVE_FULL_F 3 “aarch64_sve_float_arith_with_sub_immediate” “vsA, vsN, vsA, vsN, vsA, vsN”)] UNSPEC_COND_FADD) (match_operand:SVE_FULL_F 4 “aarch64_simd_reg_or_zero” “Dz, Dz, 0, 0, w, w”)] UNSPEC_SEL))] “TARGET_SVE && !rtx_equal_p (operands[2], operands[4])” "@ movprfx\t%0., %1/z, %2.;fadd\t%0., %1/m, %0., #%3 movprfx\t%0., %1/z, %2.;fsub\t%0., %1/m, %0., #%N3 movprfx\t%0., %1/m, %2.;fadd\t%0., %1/m, %0., #%3 movprfx\t%0., %1/m, %2.;fsub\t%0., %1/m, %0., #%N3

#" “&& reload_completed && register_operand (operands[4], mode) && !rtx_equal_p (operands[0], operands[4])” { emit_insn (gen_vcond_mask_ (operands[0], operands[2], operands[4], operands[1])); operands[4] = operands[2] = operands[0]; } [(set_attr “movprfx” “yes”)] )

;; Register merging forms are handled through SVE_COND_FP_BINARY.

;; ------------------------------------------------------------------------- ;; ---- [FP] Complex addition ;; ------------------------------------------------------------------------- ;; Includes: ;; - FCADD ;; -------------------------------------------------------------------------

;; Predicated FCADD. (define_insn “@aarch64_pred_” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=w, ?&w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl”) (match_operand:SI 4 “aarch64_sve_gp_strictness”) (match_operand:SVE_FULL_F 2 “register_operand” “0, w”) (match_operand:SVE_FULL_F 3 “register_operand” “w, w”)] SVE_COND_FCADD))] “TARGET_SVE” “@ fcadd\t%0., %1/m, %0., %3., # movprfx\t%0, %2;fcadd\t%0., %1/m, %0., %3., #” [(set_attr “movprfx” “*,yes”)] )

;; Predicated FCADD with merging. (define_expand “@cond_” [(set (match_operand:SVE_FULL_F 0 “register_operand”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand”) (unspec:SVE_FULL_F [(match_dup 1) (const_int SVE_STRICT_GP) (match_operand:SVE_FULL_F 2 “register_operand”) (match_operand:SVE_FULL_F 3 “register_operand”)] SVE_COND_FCADD) (match_operand:SVE_FULL_F 4 “aarch64_simd_reg_or_zero”)] UNSPEC_SEL))] “TARGET_SVE” )

;; Predicated FCADD using ptrue for unpredicated optab for auto-vectorizer (define_expand “@cadd3” [(set (match_operand:SVE_FULL_F 0 “register_operand”) (unspec:SVE_FULL_F [(match_dup 3) (const_int SVE_RELAXED_GP) (match_operand:SVE_FULL_F 1 “register_operand”) (match_operand:SVE_FULL_F 2 “register_operand”)] SVE_COND_FCADD))] “TARGET_SVE” { operands[3] = aarch64_ptrue_reg (mode); })

;; Predicated FCADD, merging with the first input. (define_insn_and_rewrite “*cond__2_relaxed” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=w, ?&w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl”) (unspec:SVE_FULL_F [(match_operand 4) (const_int SVE_RELAXED_GP) (match_operand:SVE_FULL_F 2 “register_operand” “0, w”) (match_operand:SVE_FULL_F 3 “register_operand” “w, w”)] SVE_COND_FCADD) (match_dup 2)] UNSPEC_SEL))] “TARGET_SVE” “@ fcadd\t%0., %1/m, %0., %3., # movprfx\t%0, %2;fcadd\t%0., %1/m, %0., %3., #” “&& !rtx_equal_p (operands[1], operands[4])” { operands[4] = copy_rtx (operands[1]); } [(set_attr “movprfx” “*,yes”)] )

(define_insn “*cond__2_strict” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=w, ?&w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl”) (unspec:SVE_FULL_F [(match_dup 1) (const_int SVE_STRICT_GP) (match_operand:SVE_FULL_F 2 “register_operand” “0, w”) (match_operand:SVE_FULL_F 3 “register_operand” “w, w”)] SVE_COND_FCADD) (match_dup 2)] UNSPEC_SEL))] “TARGET_SVE” “@ fcadd\t%0., %1/m, %0., %3., # movprfx\t%0, %2;fcadd\t%0., %1/m, %0., %3., #” [(set_attr “movprfx” “*,yes”)] )

;; Predicated FCADD, merging with an independent value. (define_insn_and_rewrite “*cond__any_relaxed” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=&w, &w, &w, ?&w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl, Upl, Upl”) (unspec:SVE_FULL_F [(match_operand 5) (const_int SVE_RELAXED_GP) (match_operand:SVE_FULL_F 2 “register_operand” “w, 0, w, w”) (match_operand:SVE_FULL_F 3 “register_operand” “w, w, w, w”)] SVE_COND_FCADD) (match_operand:SVE_FULL_F 4 “aarch64_simd_reg_or_zero” “Dz, Dz, 0, w”)] UNSPEC_SEL))] “TARGET_SVE && !rtx_equal_p (operands[2], operands[4])” “@ movprfx\t%0., %1/z, %2.;fcadd\t%0., %1/m, %0., %3., # movprfx\t%0., %1/z, %0.;fcadd\t%0., %1/m, %0., %3., # movprfx\t%0., %1/m, %2.;fcadd\t%0., %1/m, %0., %3., # #” “&& 1” { if (reload_completed && register_operand (operands[4], mode) && !rtx_equal_p (operands[0], operands[4])) { emit_insn (gen_vcond_mask_ (operands[0], operands[2], operands[4], operands[1])); operands[4] = operands[2] = operands[0]; } else if (!rtx_equal_p (operands[1], operands[5])) operands[5] = copy_rtx (operands[1]); else FAIL; } [(set_attr “movprfx” “yes”)] )

(define_insn_and_rewrite “*cond__any_strict” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=&w, &w, &w, ?&w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl, Upl, Upl”) (unspec:SVE_FULL_F [(match_dup 1) (const_int SVE_STRICT_GP) (match_operand:SVE_FULL_F 2 “register_operand” “w, 0, w, w”) (match_operand:SVE_FULL_F 3 “register_operand” “w, w, w, w”)] SVE_COND_FCADD) (match_operand:SVE_FULL_F 4 “aarch64_simd_reg_or_zero” “Dz, Dz, 0, w”)] UNSPEC_SEL))] “TARGET_SVE && !rtx_equal_p (operands[2], operands[4])” “@ movprfx\t%0., %1/z, %2.;fcadd\t%0., %1/m, %0., %3., # movprfx\t%0., %1/z, %0.;fcadd\t%0., %1/m, %0., %3., # movprfx\t%0., %1/m, %2.;fcadd\t%0., %1/m, %0., %3., # #” “&& reload_completed && register_operand (operands[4], mode) && !rtx_equal_p (operands[0], operands[4])” { emit_insn (gen_vcond_mask_ (operands[0], operands[2], operands[4], operands[1])); operands[4] = operands[2] = operands[0]; } [(set_attr “movprfx” “yes”)] )

;; ------------------------------------------------------------------------- ;; ---- [FP] Subtraction ;; ------------------------------------------------------------------------- ;; Includes: ;; - FSUB ;; - FSUBR ;; -------------------------------------------------------------------------

;; Predicated floating-point subtraction. (define_insn_and_split “@aarch64_pred_” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=w, w, w, w, ?&w, ?&w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl, Upl, Upl, Upl, Upl”) (match_operand:SI 4 “aarch64_sve_gp_strictness” “i, Z, Ui1, Ui1, i, Ui1”) (match_operand:SVE_FULL_F 2 “aarch64_sve_float_arith_operand” “vsA, w, 0, w, vsA, w”) (match_operand:SVE_FULL_F 3 “register_operand” “0, w, w, 0, w, w”)] SVE_COND_FP_SUB))] “TARGET_SVE” "@ fsubr\t%0., %1/m, %0., #%2

fsub\t%0., %1/m, %0., %3. fsubr\t%0., %1/m, %0., %2. movprfx\t%0, %3;fsubr\t%0., %1/m, %0., #%2 movprfx\t%0, %2;fsub\t%0., %1/m, %0., %3." ; Split the unpredicated form after reload, so that we don't have ; the unnecessary PTRUE. “&& reload_completed && register_operand (operands[2], mode) && INTVAL (operands[4]) == SVE_RELAXED_GP” [(set (match_dup 0) (minus:SVE_FULL_F (match_dup 2) (match_dup 3)))] "" [(set_attr “movprfx” “,,,,yes,yes”)] )

;; Predicated floating-point subtraction from a constant, merging with the ;; second input. (define_insn_and_rewrite “*cond_sub_3_const_relaxed” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=w, ?w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl”) (unspec:SVE_FULL_F [(match_operand 4) (const_int SVE_RELAXED_GP) (match_operand:SVE_FULL_F 2 “aarch64_sve_float_arith_immediate”) (match_operand:SVE_FULL_F 3 “register_operand” “0, w”)] UNSPEC_COND_FSUB) (match_dup 3)] UNSPEC_SEL))] “TARGET_SVE” “@ fsubr\t%0., %1/m, %0., #%2 movprfx\t%0, %3;fsubr\t%0., %1/m, %0., #%2” “&& !rtx_equal_p (operands[1], operands[4])” { operands[4] = copy_rtx (operands[1]); } [(set_attr “movprfx” “*,yes”)] )

(define_insn “*cond_sub_3_const_strict” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=w, ?w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl”) (unspec:SVE_FULL_F [(match_dup 1) (const_int SVE_STRICT_GP) (match_operand:SVE_FULL_F 2 “aarch64_sve_float_arith_immediate”) (match_operand:SVE_FULL_F 3 “register_operand” “0, w”)] UNSPEC_COND_FSUB) (match_dup 3)] UNSPEC_SEL))] “TARGET_SVE” “@ fsubr\t%0., %1/m, %0., #%2 movprfx\t%0, %3;fsubr\t%0., %1/m, %0., #%2” [(set_attr “movprfx” “*,yes”)] )

;; Predicated floating-point subtraction from a constant, merging with an ;; independent value. (define_insn_and_rewrite “*cond_sub_const_relaxed” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=w, w, ?w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl, Upl”) (unspec:SVE_FULL_F [(match_operand 5) (const_int SVE_RELAXED_GP) (match_operand:SVE_FULL_F 2 “aarch64_sve_float_arith_immediate”) (match_operand:SVE_FULL_F 3 “register_operand” “w, w, w”)] UNSPEC_COND_FSUB) (match_operand:SVE_FULL_F 4 “aarch64_simd_reg_or_zero” “Dz, 0, w”)] UNSPEC_SEL))] “TARGET_SVE && !rtx_equal_p (operands[3], operands[4])” “@ movprfx\t%0., %1/z, %3.;fsubr\t%0., %1/m, %0., #%2 movprfx\t%0., %1/m, %3.;fsubr\t%0., %1/m, %0., #%2 #” “&& 1” { if (reload_completed && register_operand (operands[4], mode) && !rtx_equal_p (operands[0], operands[4])) { emit_insn (gen_vcond_mask_ (operands[0], operands[3], operands[4], operands[1])); operands[4] = operands[3] = operands[0]; } else if (!rtx_equal_p (operands[1], operands[5])) operands[5] = copy_rtx (operands[1]); else FAIL; } [(set_attr “movprfx” “yes”)] )

(define_insn_and_rewrite “*cond_sub_const_strict” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=w, w, ?w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl, Upl”) (unspec:SVE_FULL_F [(match_dup 1) (const_int SVE_STRICT_GP) (match_operand:SVE_FULL_F 2 “aarch64_sve_float_arith_immediate”) (match_operand:SVE_FULL_F 3 “register_operand” “w, w, w”)] UNSPEC_COND_FSUB) (match_operand:SVE_FULL_F 4 “aarch64_simd_reg_or_zero” “Dz, 0, w”)] UNSPEC_SEL))] “TARGET_SVE && !rtx_equal_p (operands[3], operands[4])” “@ movprfx\t%0., %1/z, %3.;fsubr\t%0., %1/m, %0., #%2 movprfx\t%0., %1/m, %3.;fsubr\t%0., %1/m, %0., #%2 #” “&& reload_completed && register_operand (operands[4], mode) && !rtx_equal_p (operands[0], operands[4])” { emit_insn (gen_vcond_mask_ (operands[0], operands[3], operands[4], operands[1])); operands[4] = operands[3] = operands[0]; } [(set_attr “movprfx” “yes”)] ) ;; Register merging forms are handled through SVE_COND_FP_BINARY.

;; ------------------------------------------------------------------------- ;; ---- [FP] Absolute difference ;; ------------------------------------------------------------------------- ;; Includes: ;; - FABD ;; -------------------------------------------------------------------------

;; Predicated floating-point absolute difference. (define_expand “@aarch64_pred_abd” [(set (match_operand:SVE_FULL_F 0 “register_operand”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand”) (match_operand:SI 4 “aarch64_sve_gp_strictness”) (unspec:SVE_FULL_F [(match_dup 1) (match_dup 4) (match_operand:SVE_FULL_F 2 “register_operand”) (match_operand:SVE_FULL_F 3 “register_operand”)] UNSPEC_COND_FSUB)] UNSPEC_COND_FABS))] “TARGET_SVE” )

;; Predicated floating-point absolute difference. (define_insn_and_rewrite “*aarch64_pred_abd_relaxed” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=w, ?&w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl”) (match_operand:SI 4 “aarch64_sve_gp_strictness”) (unspec:SVE_FULL_F [(match_operand 5) (const_int SVE_RELAXED_GP) (match_operand:SVE_FULL_F 2 “register_operand” “%0, w”) (match_operand:SVE_FULL_F 3 “register_operand” “w, w”)] UNSPEC_COND_FSUB)] UNSPEC_COND_FABS))] “TARGET_SVE” “@ fabd\t%0., %1/m, %0., %3. movprfx\t%0, %2;fabd\t%0., %1/m, %0., %3.” “&& !rtx_equal_p (operands[1], operands[5])” { operands[5] = copy_rtx (operands[1]); } [(set_attr “movprfx” “*,yes”)] )

(define_insn “*aarch64_pred_abd_strict” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=w, ?&w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl”) (match_operand:SI 4 “aarch64_sve_gp_strictness”) (unspec:SVE_FULL_F [(match_dup 1) (const_int SVE_STRICT_GP) (match_operand:SVE_FULL_F 2 “register_operand” “%0, w”) (match_operand:SVE_FULL_F 3 “register_operand” “w, w”)] UNSPEC_COND_FSUB)] UNSPEC_COND_FABS))] “TARGET_SVE” “@ fabd\t%0., %1/m, %0., %3. movprfx\t%0, %2;fabd\t%0., %1/m, %0., %3.” [(set_attr “movprfx” “*,yes”)] )

(define_expand “@aarch64_cond_abd” [(set (match_operand:SVE_FULL_F 0 “register_operand”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand”) (unspec:SVE_FULL_F [(match_dup 1) (const_int SVE_STRICT_GP) (unspec:SVE_FULL_F [(match_dup 1) (const_int SVE_STRICT_GP) (match_operand:SVE_FULL_F 2 “register_operand”) (match_operand:SVE_FULL_F 3 “register_operand”)] UNSPEC_COND_FSUB)] UNSPEC_COND_FABS) (match_operand:SVE_FULL_F 4 “aarch64_simd_reg_or_zero”)] UNSPEC_SEL))] “TARGET_SVE” { if (rtx_equal_p (operands[3], operands[4])) std::swap (operands[2], operands[3]); })

;; Predicated floating-point absolute difference, merging with the first ;; input. (define_insn_and_rewrite “*aarch64_cond_abd_2_relaxed” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=w, ?&w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl”) (unspec:SVE_FULL_F [(match_operand 4) (const_int SVE_RELAXED_GP) (unspec:SVE_FULL_F [(match_operand 5) (const_int SVE_RELAXED_GP) (match_operand:SVE_FULL_F 2 “register_operand” “0, w”) (match_operand:SVE_FULL_F 3 “register_operand” “w, w”)] UNSPEC_COND_FSUB)] UNSPEC_COND_FABS) (match_dup 2)] UNSPEC_SEL))] “TARGET_SVE” “@ fabd\t%0., %1/m, %0., %3. movprfx\t%0, %2;fabd\t%0., %1/m, %0., %3.” “&& (!rtx_equal_p (operands[1], operands[4]) || !rtx_equal_p (operands[1], operands[5]))” { operands[4] = copy_rtx (operands[1]); operands[5] = copy_rtx (operands[1]); } [(set_attr “movprfx” “*,yes”)] )

(define_insn “*aarch64_cond_abd_2_strict” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=w, ?&w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl”) (unspec:SVE_FULL_F [(match_dup 1) (match_operand:SI 4 “aarch64_sve_gp_strictness”) (unspec:SVE_FULL_F [(match_dup 1) (match_operand:SI 5 “aarch64_sve_gp_strictness”) (match_operand:SVE_FULL_F 2 “register_operand” “0, w”) (match_operand:SVE_FULL_F 3 “register_operand” “w, w”)] UNSPEC_COND_FSUB)] UNSPEC_COND_FABS) (match_dup 2)] UNSPEC_SEL))] “TARGET_SVE” “@ fabd\t%0., %1/m, %0., %3. movprfx\t%0, %2;fabd\t%0., %1/m, %0., %3.” [(set_attr “movprfx” “*,yes”)] )

;; Predicated floating-point absolute difference, merging with the second ;; input. (define_insn_and_rewrite “*aarch64_cond_abd_3_relaxed” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=w, ?&w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl”) (unspec:SVE_FULL_F [(match_operand 4) (const_int SVE_RELAXED_GP) (unspec:SVE_FULL_F [(match_operand 5) (const_int SVE_RELAXED_GP) (match_operand:SVE_FULL_F 2 “register_operand” “w, w”) (match_operand:SVE_FULL_F 3 “register_operand” “0, w”)] UNSPEC_COND_FSUB)] UNSPEC_COND_FABS) (match_dup 3)] UNSPEC_SEL))] “TARGET_SVE” “@ fabd\t%0., %1/m, %0., %2. movprfx\t%0, %3;fabd\t%0., %1/m, %0., %2.” “&& (!rtx_equal_p (operands[1], operands[4]) || !rtx_equal_p (operands[1], operands[5]))” { operands[4] = copy_rtx (operands[1]); operands[5] = copy_rtx (operands[1]); } [(set_attr “movprfx” “*,yes”)] )

(define_insn “*aarch64_cond_abd_3_strict” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=w, ?&w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl”) (unspec:SVE_FULL_F [(match_dup 1) (match_operand:SI 4 “aarch64_sve_gp_strictness”) (unspec:SVE_FULL_F [(match_dup 1) (match_operand:SI 5 “aarch64_sve_gp_strictness”) (match_operand:SVE_FULL_F 2 “register_operand” “w, w”) (match_operand:SVE_FULL_F 3 “register_operand” “0, w”)] UNSPEC_COND_FSUB)] UNSPEC_COND_FABS) (match_dup 3)] UNSPEC_SEL))] “TARGET_SVE” “@ fabd\t%0., %1/m, %0., %2. movprfx\t%0, %3;fabd\t%0., %1/m, %0., %2.” [(set_attr “movprfx” “*,yes”)] )

;; Predicated floating-point absolute difference, merging with an ;; independent value. (define_insn_and_rewrite “*aarch64_cond_abd_any_relaxed” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=&w, &w, &w, &w, ?&w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl, Upl, Upl, Upl”) (unspec:SVE_FULL_F [(match_operand 5) (const_int SVE_RELAXED_GP) (unspec:SVE_FULL_F [(match_operand 6) (const_int SVE_RELAXED_GP) (match_operand:SVE_FULL_F 2 “register_operand” “0, w, w, w, w”) (match_operand:SVE_FULL_F 3 “register_operand” “w, 0, w, w, w”)] UNSPEC_COND_FSUB)] UNSPEC_COND_FABS) (match_operand:SVE_FULL_F 4 “aarch64_simd_reg_or_zero” “Dz, Dz, Dz, 0, w”)] UNSPEC_SEL))] “TARGET_SVE && !rtx_equal_p (operands[2], operands[4]) && !rtx_equal_p (operands[3], operands[4])” “@ movprfx\t%0., %1/z, %0.;fabd\t%0., %1/m, %0., %3. movprfx\t%0., %1/z, %0.;fabd\t%0., %1/m, %0., %2. movprfx\t%0., %1/z, %2.;fabd\t%0., %1/m, %0., %3. movprfx\t%0., %1/m, %2.;fabd\t%0., %1/m, %0., %3. #” “&& 1” { if (reload_completed && register_operand (operands[4], mode) && !rtx_equal_p (operands[0], operands[4])) { emit_insn (gen_vcond_mask_ (operands[0], operands[3], operands[4], operands[1])); operands[4] = operands[3] = operands[0]; } else if (!rtx_equal_p (operands[1], operands[5]) || !rtx_equal_p (operands[1], operands[6])) { operands[5] = copy_rtx (operands[1]); operands[6] = copy_rtx (operands[1]); } else FAIL; } [(set_attr “movprfx” “yes”)] )

(define_insn_and_rewrite “*aarch64_cond_abd_any_strict” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=&w, &w, &w, &w, ?&w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl, Upl, Upl, Upl”) (unspec:SVE_FULL_F [(match_dup 1) (match_operand:SI 5 “aarch64_sve_gp_strictness”) (unspec:SVE_FULL_F [(match_dup 1) (match_operand:SI 6 “aarch64_sve_gp_strictness”) (match_operand:SVE_FULL_F 2 “register_operand” “0, w, w, w, w”) (match_operand:SVE_FULL_F 3 “register_operand” “w, 0, w, w, w”)] UNSPEC_COND_FSUB)] UNSPEC_COND_FABS) (match_operand:SVE_FULL_F 4 “aarch64_simd_reg_or_zero” “Dz, Dz, Dz, 0, w”)] UNSPEC_SEL))] “TARGET_SVE && !rtx_equal_p (operands[2], operands[4]) && !rtx_equal_p (operands[3], operands[4])” “@ movprfx\t%0., %1/z, %0.;fabd\t%0., %1/m, %0., %3. movprfx\t%0., %1/z, %0.;fabd\t%0., %1/m, %0., %2. movprfx\t%0., %1/z, %2.;fabd\t%0., %1/m, %0., %3. movprfx\t%0., %1/m, %2.;fabd\t%0., %1/m, %0., %3. #” “&& reload_completed && register_operand (operands[4], mode) && !rtx_equal_p (operands[0], operands[4])” { emit_insn (gen_vcond_mask_ (operands[0], operands[3], operands[4], operands[1])); operands[4] = operands[3] = operands[0]; } [(set_attr “movprfx” “yes”)] )

;; ------------------------------------------------------------------------- ;; ---- [FP] Multiplication ;; ------------------------------------------------------------------------- ;; Includes: ;; - FMUL ;; -------------------------------------------------------------------------

;; Predicated floating-point multiplication. (define_insn_and_split “@aarch64_pred_” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=w, w, w, ?&w, ?&w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl, Upl, Upl, Upl”) (match_operand:SI 4 “aarch64_sve_gp_strictness” “i, Z, Ui1, i, Ui1”) (match_operand:SVE_FULL_F 2 “register_operand” “%0, w, 0, w, w”) (match_operand:SVE_FULL_F 3 “aarch64_sve_float_mul_operand” “vsM, w, w, vsM, w”)] SVE_COND_FP_MUL))] “TARGET_SVE” "@ fmul\t%0., %1/m, %0., #%3

fmul\t%0., %1/m, %0., %3. movprfx\t%0, %2;fmul\t%0., %1/m, %0., #%3 movprfx\t%0, %2;fmul\t%0., %1/m, %0., %3." ; Split the unpredicated form after reload, so that we don't have ; the unnecessary PTRUE. “&& reload_completed && register_operand (operands[3], mode) && INTVAL (operands[4]) == SVE_RELAXED_GP” [(set (match_dup 0) (mult:SVE_FULL_F (match_dup 2) (match_dup 3)))] "" [(set_attr “movprfx” “,,*,yes,yes”)] )

;; Merging forms are handled through SVE_COND_FP_BINARY and ;; SVE_COND_FP_BINARY_I1.

;; Unpredicated multiplication by selected lanes. (define_insn “@aarch64_mul_lane_” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=w”) (mult:SVE_FULL_F (unspec:SVE_FULL_F [(match_operand:SVE_FULL_F 2 “register_operand” “<sve_lane_con>”) (match_operand:SI 3 “const_int_operand”)] UNSPEC_SVE_LANE_SELECT) (match_operand:SVE_FULL_F 1 “register_operand” “w”)))] “TARGET_SVE” “fmul\t%0., %1., %2.[%3]” )

;; ------------------------------------------------------------------------- ;; ---- [FP] Division ;; ------------------------------------------------------------------------- ;; The patterns in this section are synthetic. ;; -------------------------------------------------------------------------

(define_expand “div3” [(set (match_operand:SVE_FULL_F 0 “register_operand”) (unspec:SVE_FULL_F [(match_dup 3) (const_int SVE_RELAXED_GP) (match_operand:SVE_FULL_F 1 “nonmemory_operand”) (match_operand:SVE_FULL_F 2 “register_operand”)] UNSPEC_COND_FDIV))] “TARGET_SVE” { if (aarch64_emit_approx_div (operands[0], operands[1], operands[2])) DONE;

operands[1] = force_reg (<MODE>mode, operands[1]);
operands[3] = aarch64_ptrue_reg (<VPRED>mode);

} )

(define_expand “@aarch64_frecpe” [(set (match_operand:SVE_FULL_F 0 “register_operand”) (unspec:SVE_FULL_F [(match_operand:SVE_FULL_F 1 “register_operand”)] UNSPEC_FRECPE))] “TARGET_SVE” )

(define_expand “@aarch64_frecps” [(set (match_operand:SVE_FULL_F 0 “register_operand”) (unspec:SVE_FULL_F [(match_operand:SVE_FULL_F 1 “register_operand”) (match_operand:SVE_FULL_F 2 “register_operand”)] UNSPEC_FRECPS))] “TARGET_SVE” )

;; ------------------------------------------------------------------------- ;; ---- [FP] Binary logical operations ;; ------------------------------------------------------------------------- ;; Includes ;; - AND ;; - EOR ;; - ORR ;; -------------------------------------------------------------------------

;; Binary logical operations on floating-point modes. We avoid subregs ;; by providing this, but we need to use UNSPECs since rtx logical ops ;; aren't defined for floating-point modes. (define_insn “*3” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=w”) (unspec:SVE_FULL_F [(match_operand:SVE_FULL_F 1 “register_operand” “w”) (match_operand:SVE_FULL_F 2 “register_operand” “w”)] LOGICALF))] “TARGET_SVE” “<logicalf_op>\t%0.d, %1.d, %2.d” )

;; ------------------------------------------------------------------------- ;; ---- [FP] Sign copying ;; ------------------------------------------------------------------------- ;; The patterns in this section are synthetic. ;; -------------------------------------------------------------------------

(define_expand “copysign3” [(match_operand:SVE_FULL_F 0 “register_operand”) (match_operand:SVE_FULL_F 1 “register_operand”) (match_operand:SVE_FULL_F 2 “register_operand”)] “TARGET_SVE” { rtx sign = gen_reg_rtx (<V_INT_EQUIV>mode); rtx mant = gen_reg_rtx (<V_INT_EQUIV>mode); rtx int_res = gen_reg_rtx (<V_INT_EQUIV>mode); int bits = GET_MODE_UNIT_BITSIZE (mode) - 1;

rtx arg1 = lowpart_subreg (<V_INT_EQUIV>mode, operands[1], <MODE>mode);
rtx arg2 = lowpart_subreg (<V_INT_EQUIV>mode, operands[2], <MODE>mode);

emit_insn (gen_and<v_int_equiv>3
       (sign, arg2,
	aarch64_simd_gen_const_vector_dup (<V_INT_EQUIV>mode,
					   HOST_WIDE_INT_M1U
					   << bits)));
emit_insn (gen_and<v_int_equiv>3
       (mant, arg1,
	aarch64_simd_gen_const_vector_dup (<V_INT_EQUIV>mode,
					   ~(HOST_WIDE_INT_M1U
					     << bits))));
emit_insn (gen_ior<v_int_equiv>3 (int_res, sign, mant));
emit_move_insn (operands[0], gen_lowpart (<MODE>mode, int_res));
DONE;

} )

(define_expand “xorsign3” [(match_operand:SVE_FULL_F 0 “register_operand”) (match_operand:SVE_FULL_F 1 “register_operand”) (match_operand:SVE_FULL_F 2 “register_operand”)] “TARGET_SVE” { rtx sign = gen_reg_rtx (<V_INT_EQUIV>mode); rtx int_res = gen_reg_rtx (<V_INT_EQUIV>mode); int bits = GET_MODE_UNIT_BITSIZE (mode) - 1;

rtx arg1 = lowpart_subreg (<V_INT_EQUIV>mode, operands[1], <MODE>mode);
rtx arg2 = lowpart_subreg (<V_INT_EQUIV>mode, operands[2], <MODE>mode);

emit_insn (gen_and<v_int_equiv>3
       (sign, arg2,
	aarch64_simd_gen_const_vector_dup (<V_INT_EQUIV>mode,
					   HOST_WIDE_INT_M1U
					   << bits)));
emit_insn (gen_xor<v_int_equiv>3 (int_res, arg1, sign));
emit_move_insn (operands[0], gen_lowpart (<MODE>mode, int_res));
DONE;

} )

;; ------------------------------------------------------------------------- ;; ---- [FP] Maximum and minimum ;; ------------------------------------------------------------------------- ;; Includes: ;; - FMAX ;; - FMAXNM ;; - FMIN ;; - FMINNM ;; -------------------------------------------------------------------------

;; Unpredicated fmax/fmin (the libm functions). The optabs for the ;; smin/smax rtx codes are handled in the generic section above. (define_expand “<maxmin_uns>3” [(set (match_operand:SVE_FULL_F 0 “register_operand”) (unspec:SVE_FULL_F [(match_dup 3) (const_int SVE_RELAXED_GP) (match_operand:SVE_FULL_F 1 “register_operand”) (match_operand:SVE_FULL_F 2 “aarch64_sve_float_maxmin_operand”)] SVE_COND_FP_MAXMIN_PUBLIC))] “TARGET_SVE” { operands[3] = aarch64_ptrue_reg (mode); } )

;; Predicated floating-point maximum/minimum. (define_insn “@aarch64_pred_” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=w, w, ?&w, ?&w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl, Upl, Upl”) (match_operand:SI 4 “aarch64_sve_gp_strictness”) (match_operand:SVE_FULL_F 2 “register_operand” “%0, 0, w, w”) (match_operand:SVE_FULL_F 3 “aarch64_sve_float_maxmin_operand” “vsB, w, vsB, w”)] SVE_COND_FP_MAXMIN))] “TARGET_SVE” “@ <sve_fp_op>\t%0., %1/m, %0., #%3 <sve_fp_op>\t%0., %1/m, %0., %3. movprfx\t%0, %2;<sve_fp_op>\t%0., %1/m, %0., #%3 movprfx\t%0, %2;<sve_fp_op>\t%0., %1/m, %0., %3.” [(set_attr “movprfx” “,,yes,yes”)] )

;; Merging forms are handled through SVE_COND_FP_BINARY and ;; SVE_COND_FP_BINARY_I1.

;; ------------------------------------------------------------------------- ;; ---- [PRED] Binary logical operations ;; ------------------------------------------------------------------------- ;; Includes: ;; - AND ;; - ANDS ;; - EOR ;; - EORS ;; - ORR ;; - ORRS ;; -------------------------------------------------------------------------

;; Predicate AND. We can reuse one of the inputs as the GP. ;; Doubling the second operand is the preferred implementation ;; of the MOV alias, so we use that instead of %1/z, %1, %2. (define_insn “and3” [(set (match_operand:PRED_ALL 0 “register_operand” “=Upa”) (and:PRED_ALL (match_operand:PRED_ALL 1 “register_operand” “Upa”) (match_operand:PRED_ALL 2 “register_operand” “Upa”)))] “TARGET_SVE” “and\t%0.b, %1/z, %2.b, %2.b” )

;; Unpredicated predicate EOR and ORR. (define_expand “3” [(set (match_operand:PRED_ALL 0 “register_operand”) (and:PRED_ALL (LOGICAL_OR:PRED_ALL (match_operand:PRED_ALL 1 “register_operand”) (match_operand:PRED_ALL 2 “register_operand”)) (match_dup 3)))] “TARGET_SVE” { operands[3] = aarch64_ptrue_reg (mode); } )

;; Predicated predicate AND, EOR and ORR. (define_insn “@aarch64_pred__z” [(set (match_operand:PRED_ALL 0 “register_operand” “=Upa”) (and:PRED_ALL (LOGICAL:PRED_ALL (match_operand:PRED_ALL 2 “register_operand” “Upa”) (match_operand:PRED_ALL 3 “register_operand” “Upa”)) (match_operand:PRED_ALL 1 “register_operand” “Upa”)))] “TARGET_SVE” “\t%0.b, %1/z, %2.b, %3.b” )

;; Perform a logical operation on operands 2 and 3, using operand 1 as ;; the GP. Store the result in operand 0 and set the flags in the same ;; way as for PTEST. (define_insn “*3_cc” [(set (reg:CC_NZC CC_REGNUM) (unspec:CC_NZC [(match_operand:VNx16BI 1 “register_operand” “Upa”) (match_operand 4) (match_operand:SI 5 “aarch64_sve_ptrue_flag”) (and:PRED_ALL (LOGICAL:PRED_ALL (match_operand:PRED_ALL 2 “register_operand” “Upa”) (match_operand:PRED_ALL 3 “register_operand” “Upa”)) (match_dup 4))] UNSPEC_PTEST)) (set (match_operand:PRED_ALL 0 “register_operand” “=Upa”) (and:PRED_ALL (LOGICAL:PRED_ALL (match_dup 2) (match_dup 3)) (match_dup 4)))] “TARGET_SVE” “s\t%0.b, %1/z, %2.b, %3.b” )

;; Same with just the flags result. (define_insn “*3_ptest” [(set (reg:CC_NZC CC_REGNUM) (unspec:CC_NZC [(match_operand:VNx16BI 1 “register_operand” “Upa”) (match_operand 4) (match_operand:SI 5 “aarch64_sve_ptrue_flag”) (and:PRED_ALL (LOGICAL:PRED_ALL (match_operand:PRED_ALL 2 “register_operand” “Upa”) (match_operand:PRED_ALL 3 “register_operand” “Upa”)) (match_dup 4))] UNSPEC_PTEST)) (clobber (match_scratch:VNx16BI 0 “=Upa”))] “TARGET_SVE” “s\t%0.b, %1/z, %2.b, %3.b” )

;; ------------------------------------------------------------------------- ;; ---- [PRED] Binary logical operations (inverted second input) ;; ------------------------------------------------------------------------- ;; Includes: ;; - BIC ;; - ORN ;; -------------------------------------------------------------------------

;; Predicated predicate BIC and ORN. (define_insn “aarch64_pred__z” [(set (match_operand:PRED_ALL 0 “register_operand” “=Upa”) (and:PRED_ALL (NLOGICAL:PRED_ALL (not:PRED_ALL (match_operand:PRED_ALL 3 “register_operand” “Upa”)) (match_operand:PRED_ALL 2 “register_operand” “Upa”)) (match_operand:PRED_ALL 1 “register_operand” “Upa”)))] “TARGET_SVE” “\t%0.b, %1/z, %2.b, %3.b” )

;; Same, but set the flags as a side-effect. (define_insn “*3_cc” [(set (reg:CC_NZC CC_REGNUM) (unspec:CC_NZC [(match_operand:VNx16BI 1 “register_operand” “Upa”) (match_operand 4) (match_operand:SI 5 “aarch64_sve_ptrue_flag”) (and:PRED_ALL (NLOGICAL:PRED_ALL (not:PRED_ALL (match_operand:PRED_ALL 3 “register_operand” “Upa”)) (match_operand:PRED_ALL 2 “register_operand” “Upa”)) (match_dup 4))] UNSPEC_PTEST)) (set (match_operand:PRED_ALL 0 “register_operand” “=Upa”) (and:PRED_ALL (NLOGICAL:PRED_ALL (not:PRED_ALL (match_dup 3)) (match_dup 2)) (match_dup 4)))] “TARGET_SVE” “s\t%0.b, %1/z, %2.b, %3.b” )

;; Same with just the flags result. (define_insn “*3_ptest” [(set (reg:CC_NZC CC_REGNUM) (unspec:CC_NZC [(match_operand:VNx16BI 1 “register_operand” “Upa”) (match_operand 4) (match_operand:SI 5 “aarch64_sve_ptrue_flag”) (and:PRED_ALL (NLOGICAL:PRED_ALL (not:PRED_ALL (match_operand:PRED_ALL 3 “register_operand” “Upa”)) (match_operand:PRED_ALL 2 “register_operand” “Upa”)) (match_dup 4))] UNSPEC_PTEST)) (clobber (match_scratch:VNx16BI 0 “=Upa”))] “TARGET_SVE” “s\t%0.b, %1/z, %2.b, %3.b” )

;; ------------------------------------------------------------------------- ;; ---- [PRED] Binary logical operations (inverted result) ;; ------------------------------------------------------------------------- ;; Includes: ;; - NAND ;; - NOR ;; -------------------------------------------------------------------------

;; Predicated predicate NAND and NOR. (define_insn “aarch64_pred_<logical_nn>_z” [(set (match_operand:PRED_ALL 0 “register_operand” “=Upa”) (and:PRED_ALL (NLOGICAL:PRED_ALL (not:PRED_ALL (match_operand:PRED_ALL 2 “register_operand” “Upa”)) (not:PRED_ALL (match_operand:PRED_ALL 3 “register_operand” “Upa”))) (match_operand:PRED_ALL 1 “register_operand” “Upa”)))] “TARGET_SVE” “<logical_nn>\t%0.b, %1/z, %2.b, %3.b” )

;; Same, but set the flags as a side-effect. (define_insn “*<logical_nn>3_cc” [(set (reg:CC_NZC CC_REGNUM) (unspec:CC_NZC [(match_operand:VNx16BI 1 “register_operand” “Upa”) (match_operand 4) (match_operand:SI 5 “aarch64_sve_ptrue_flag”) (and:PRED_ALL (NLOGICAL:PRED_ALL (not:PRED_ALL (match_operand:PRED_ALL 2 “register_operand” “Upa”)) (not:PRED_ALL (match_operand:PRED_ALL 3 “register_operand” “Upa”))) (match_dup 4))] UNSPEC_PTEST)) (set (match_operand:PRED_ALL 0 “register_operand” “=Upa”) (and:PRED_ALL (NLOGICAL:PRED_ALL (not:PRED_ALL (match_dup 2)) (not:PRED_ALL (match_dup 3))) (match_dup 4)))] “TARGET_SVE” “<logical_nn>s\t%0.b, %1/z, %2.b, %3.b” )

;; Same with just the flags result. (define_insn “*<logical_nn>3_ptest” [(set (reg:CC_NZC CC_REGNUM) (unspec:CC_NZC [(match_operand:VNx16BI 1 “register_operand” “Upa”) (match_operand 4) (match_operand:SI 5 “aarch64_sve_ptrue_flag”) (and:PRED_ALL (NLOGICAL:PRED_ALL (not:PRED_ALL (match_operand:PRED_ALL 2 “register_operand” “Upa”)) (not:PRED_ALL (match_operand:PRED_ALL 3 “register_operand” “Upa”))) (match_dup 4))] UNSPEC_PTEST)) (clobber (match_scratch:VNx16BI 0 “=Upa”))] “TARGET_SVE” “<logical_nn>s\t%0.b, %1/z, %2.b, %3.b” )

;; ========================================================================= ;; == Ternary arithmetic ;; =========================================================================

;; ------------------------------------------------------------------------- ;; ---- [INT] MLA and MAD ;; ------------------------------------------------------------------------- ;; Includes: ;; - MAD ;; - MLA ;; -------------------------------------------------------------------------

;; Unpredicated integer addition of product. (define_expand “fma4” [(set (match_operand:SVE_I 0 “register_operand”) (plus:SVE_I (unspec:SVE_I [(match_dup 4) (mult:SVE_I (match_operand:SVE_I 1 “register_operand”) (match_operand:SVE_I 2 “nonmemory_operand”))] UNSPEC_PRED_X) (match_operand:SVE_I 3 “register_operand”)))] “TARGET_SVE” { if (aarch64_prepare_sve_int_fma (operands, PLUS)) DONE; operands[4] = aarch64_ptrue_reg (mode); } )

;; Predicated integer addition of product. (define_insn “@aarch64_pred_fma” [(set (match_operand:SVE_I 0 “register_operand” “=w, w, ?&w”) (plus:SVE_I (unspec:SVE_I [(match_operand: 1 “register_operand” “Upl, Upl, Upl”) (mult:SVE_I (match_operand:SVE_I 2 “register_operand” “%0, w, w”) (match_operand:SVE_I 3 “register_operand” “w, w, w”))] UNSPEC_PRED_X) (match_operand:SVE_I 4 “register_operand” “w, 0, w”)))] “TARGET_SVE” “@ mad\t%0., %1/m, %3., %4. mla\t%0., %1/m, %2., %3. movprfx\t%0, %4;mla\t%0., %1/m, %2., %3.” [(set_attr “movprfx” “,,yes”)] )

;; Predicated integer addition of product with merging. (define_expand “cond_fma” [(set (match_operand:SVE_I 0 “register_operand”) (unspec:SVE_I [(match_operand: 1 “register_operand”) (plus:SVE_I (mult:SVE_I (match_operand:SVE_I 2 “register_operand”) (match_operand:SVE_I 3 “general_operand”)) (match_operand:SVE_I 4 “register_operand”)) (match_operand:SVE_I 5 “aarch64_simd_reg_or_zero”)] UNSPEC_SEL))] “TARGET_SVE” { if (aarch64_prepare_sve_cond_int_fma (operands, PLUS)) DONE; /* Swap the multiplication operands if the fallback value is the second of the two. */ if (rtx_equal_p (operands[3], operands[5])) std::swap (operands[2], operands[3]); } )

;; Predicated integer addition of product, merging with the first input. (define_insn “*cond_fma_2” [(set (match_operand:SVE_I 0 “register_operand” “=w, ?&w”) (unspec:SVE_I [(match_operand: 1 “register_operand” “Upl, Upl”) (plus:SVE_I (mult:SVE_I (match_operand:SVE_I 2 “register_operand” “0, w”) (match_operand:SVE_I 3 “register_operand” “w, w”)) (match_operand:SVE_I 4 “register_operand” “w, w”)) (match_dup 2)] UNSPEC_SEL))] “TARGET_SVE” “@ mad\t%0., %1/m, %3., %4. movprfx\t%0, %2;mad\t%0., %1/m, %3., %4.” [(set_attr “movprfx” “*,yes”)] )

;; Predicated integer addition of product, merging with the third input. (define_insn “*cond_fma_4” [(set (match_operand:SVE_I 0 “register_operand” “=w, ?&w”) (unspec:SVE_I [(match_operand: 1 “register_operand” “Upl, Upl”) (plus:SVE_I (mult:SVE_I (match_operand:SVE_I 2 “register_operand” “w, w”) (match_operand:SVE_I 3 “register_operand” “w, w”)) (match_operand:SVE_I 4 “register_operand” “0, w”)) (match_dup 4)] UNSPEC_SEL))] “TARGET_SVE” “@ mla\t%0., %1/m, %2., %3. movprfx\t%0, %4;mla\t%0., %1/m, %2., %3.” [(set_attr “movprfx” “*,yes”)] )

;; Predicated integer addition of product, merging with an independent value. (define_insn_and_rewrite “*cond_fma_any” [(set (match_operand:SVE_I 0 “register_operand” “=&w, &w, &w, &w, &w, ?&w”) (unspec:SVE_I [(match_operand: 1 “register_operand” “Upl, Upl, Upl, Upl, Upl, Upl”) (plus:SVE_I (mult:SVE_I (match_operand:SVE_I 2 “register_operand” “w, w, 0, w, w, w”) (match_operand:SVE_I 3 “register_operand” “w, w, w, 0, w, w”)) (match_operand:SVE_I 4 “register_operand” “w, 0, w, w, w, w”)) (match_operand:SVE_I 5 “aarch64_simd_reg_or_zero” “Dz, Dz, Dz, Dz, 0, w”)] UNSPEC_SEL))] “TARGET_SVE && !rtx_equal_p (operands[2], operands[5]) && !rtx_equal_p (operands[3], operands[5]) && !rtx_equal_p (operands[4], operands[5])” “@ movprfx\t%0., %1/z, %4.;mla\t%0., %1/m, %2., %3. movprfx\t%0., %1/z, %0.;mla\t%0., %1/m, %2., %3. movprfx\t%0., %1/z, %0.;mad\t%0., %1/m, %3., %4. movprfx\t%0., %1/z, %0.;mad\t%0., %1/m, %2., %4. movprfx\t%0., %1/m, %4.;mla\t%0., %1/m, %2., %3. #” “&& reload_completed && register_operand (operands[5], mode) && !rtx_equal_p (operands[0], operands[5])” { emit_insn (gen_vcond_mask_ (operands[0], operands[4], operands[5], operands[1])); operands[5] = operands[4] = operands[0]; } [(set_attr “movprfx” “yes”)] )

;; ------------------------------------------------------------------------- ;; ---- [INT] MLS and MSB ;; ------------------------------------------------------------------------- ;; Includes: ;; - MLS ;; - MSB ;; -------------------------------------------------------------------------

;; Unpredicated integer subtraction of product. (define_expand “fnma4” [(set (match_operand:SVE_I 0 “register_operand”) (minus:SVE_I (match_operand:SVE_I 3 “register_operand”) (unspec:SVE_I [(match_dup 4) (mult:SVE_I (match_operand:SVE_I 1 “register_operand”) (match_operand:SVE_I 2 “general_operand”))] UNSPEC_PRED_X)))] “TARGET_SVE” { if (aarch64_prepare_sve_int_fma (operands, MINUS)) DONE; operands[4] = aarch64_ptrue_reg (mode); } )

;; Predicated integer subtraction of product. (define_insn “@aarch64_pred_fnma” [(set (match_operand:SVE_I 0 “register_operand” “=w, w, ?&w”) (minus:SVE_I (match_operand:SVE_I 4 “register_operand” “w, 0, w”) (unspec:SVE_I [(match_operand: 1 “register_operand” “Upl, Upl, Upl”) (mult:SVE_I (match_operand:SVE_I 2 “register_operand” “%0, w, w”) (match_operand:SVE_I 3 “register_operand” “w, w, w”))] UNSPEC_PRED_X)))] “TARGET_SVE” “@ msb\t%0., %1/m, %3., %4. mls\t%0., %1/m, %2., %3. movprfx\t%0, %4;mls\t%0., %1/m, %2., %3.” [(set_attr “movprfx” “,,yes”)] )

;; Predicated integer subtraction of product with merging. (define_expand “cond_fnma” [(set (match_operand:SVE_I 0 “register_operand”) (unspec:SVE_I [(match_operand: 1 “register_operand”) (minus:SVE_I (match_operand:SVE_I 4 “register_operand”) (mult:SVE_I (match_operand:SVE_I 2 “register_operand”) (match_operand:SVE_I 3 “general_operand”))) (match_operand:SVE_I 5 “aarch64_simd_reg_or_zero”)] UNSPEC_SEL))] “TARGET_SVE” { if (aarch64_prepare_sve_cond_int_fma (operands, MINUS)) DONE; /* Swap the multiplication operands if the fallback value is the second of the two. */ if (rtx_equal_p (operands[3], operands[5])) std::swap (operands[2], operands[3]); } )

;; Predicated integer subtraction of product, merging with the first input. (define_insn “*cond_fnma_2” [(set (match_operand:SVE_I 0 “register_operand” “=w, ?&w”) (unspec:SVE_I [(match_operand: 1 “register_operand” “Upl, Upl”) (minus:SVE_I (match_operand:SVE_I 4 “register_operand” “w, w”) (mult:SVE_I (match_operand:SVE_I 2 “register_operand” “0, w”) (match_operand:SVE_I 3 “register_operand” “w, w”))) (match_dup 2)] UNSPEC_SEL))] “TARGET_SVE” “@ msb\t%0., %1/m, %3., %4. movprfx\t%0, %2;msb\t%0., %1/m, %3., %4.” [(set_attr “movprfx” “*,yes”)] )

;; Predicated integer subtraction of product, merging with the third input. (define_insn “*cond_fnma_4” [(set (match_operand:SVE_I 0 “register_operand” “=w, ?&w”) (unspec:SVE_I [(match_operand: 1 “register_operand” “Upl, Upl”) (minus:SVE_I (match_operand:SVE_I 4 “register_operand” “0, w”) (mult:SVE_I (match_operand:SVE_I 2 “register_operand” “w, w”) (match_operand:SVE_I 3 “register_operand” “w, w”))) (match_dup 4)] UNSPEC_SEL))] “TARGET_SVE” “@ mls\t%0., %1/m, %2., %3. movprfx\t%0, %4;mls\t%0., %1/m, %2., %3.” [(set_attr “movprfx” “*,yes”)] )

;; Predicated integer subtraction of product, merging with an ;; independent value. (define_insn_and_rewrite “*cond_fnma_any” [(set (match_operand:SVE_I 0 “register_operand” “=&w, &w, &w, &w, &w, ?&w”) (unspec:SVE_I [(match_operand: 1 “register_operand” “Upl, Upl, Upl, Upl, Upl, Upl”) (minus:SVE_I (match_operand:SVE_I 4 “register_operand” “w, 0, w, w, w, w”) (mult:SVE_I (match_operand:SVE_I 2 “register_operand” “w, w, 0, w, w, w”) (match_operand:SVE_I 3 “register_operand” “w, w, w, 0, w, w”))) (match_operand:SVE_I 5 “aarch64_simd_reg_or_zero” “Dz, Dz, Dz, Dz, 0, w”)] UNSPEC_SEL))] “TARGET_SVE && !rtx_equal_p (operands[2], operands[5]) && !rtx_equal_p (operands[3], operands[5]) && !rtx_equal_p (operands[4], operands[5])” “@ movprfx\t%0., %1/z, %4.;mls\t%0., %1/m, %2., %3. movprfx\t%0., %1/z, %0.;mls\t%0., %1/m, %2., %3. movprfx\t%0., %1/z, %0.;msb\t%0., %1/m, %3., %4. movprfx\t%0., %1/z, %0.;msb\t%0., %1/m, %2., %4. movprfx\t%0., %1/m, %4.;mls\t%0., %1/m, %2., %3. #” “&& reload_completed && register_operand (operands[5], mode) && !rtx_equal_p (operands[0], operands[5])” { emit_insn (gen_vcond_mask_ (operands[0], operands[4], operands[5], operands[1])); operands[5] = operands[4] = operands[0]; } [(set_attr “movprfx” “yes”)] )

;; ------------------------------------------------------------------------- ;; ---- [INT] Dot product ;; ------------------------------------------------------------------------- ;; Includes: ;; - SDOT ;; - SUDOT (I8MM) ;; - UDOT ;; - USDOT (I8MM) ;; -------------------------------------------------------------------------

;; Four-element integer dot-product with accumulation. (define_insn “dot_prod” [(set (match_operand:SVE_FULL_SDI 0 “register_operand” “=w, ?&w”) (plus:SVE_FULL_SDI (unspec:SVE_FULL_SDI [(match_operand: 1 “register_operand” “w, w”) (match_operand: 2 “register_operand” “w, w”)] DOTPROD) (match_operand:SVE_FULL_SDI 3 “register_operand” “0, w”)))] “TARGET_SVE” “@ dot\t%0., %1.<Vetype_fourth>, %2.<Vetype_fourth> movprfx\t%0, %3;dot\t%0., %1.<Vetype_fourth>, %2.<Vetype_fourth>” [(set_attr “movprfx” “*,yes”)] )

;; Four-element integer dot-product by selected lanes with accumulation. (define_insn “@aarch64_dot_prod_lane” [(set (match_operand:SVE_FULL_SDI 0 “register_operand” “=w, ?&w”) (plus:SVE_FULL_SDI (unspec:SVE_FULL_SDI [(match_operand: 1 “register_operand” “w, w”) (unspec: [(match_operand: 2 “register_operand” “<sve_lane_con>, <sve_lane_con>”) (match_operand:SI 3 “const_int_operand”)] UNSPEC_SVE_LANE_SELECT)] DOTPROD) (match_operand:SVE_FULL_SDI 4 “register_operand” “0, w”)))] “TARGET_SVE” “@ dot\t%0., %1.<Vetype_fourth>, %2.<Vetype_fourth>[%3] movprfx\t%0, %4;dot\t%0., %1.<Vetype_fourth>, %2.<Vetype_fourth>[%3]” [(set_attr “movprfx” “*,yes”)] )

(define_insn “@dot_prod” [(set (match_operand:VNx4SI_ONLY 0 “register_operand” “=w, ?&w”) (plus:VNx4SI_ONLY (unspec:VNx4SI_ONLY [(match_operand: 1 “register_operand” “w, w”) (match_operand: 2 “register_operand” “w, w”)] DOTPROD_US_ONLY) (match_operand:VNx4SI_ONLY 3 “register_operand” “0, w”)))] “TARGET_SVE_I8MM” “@ dot\t%0.s, %1.b, %2.b movprfx\t%0, %3;dot\t%0.s, %1.b, %2.b” [(set_attr “movprfx” “*,yes”)] )

(define_insn “@aarch64_dot_prod_lane” [(set (match_operand:VNx4SI_ONLY 0 “register_operand” “=w, ?&w”) (plus:VNx4SI_ONLY (unspec:VNx4SI_ONLY [(match_operand: 1 “register_operand” “w, w”) (unspec: [(match_operand: 2 “register_operand” “y, y”) (match_operand:SI 3 “const_int_operand”)] UNSPEC_SVE_LANE_SELECT)] DOTPROD_I8MM) (match_operand:VNx4SI_ONLY 4 “register_operand” “0, w”)))] “TARGET_SVE_I8MM” “@ dot\t%0.s, %1.b, %2.b[%3] movprfx\t%0, %4;dot\t%0.s, %1.b, %2.b[%3]” [(set_attr “movprfx” “*,yes”)] )

;; ------------------------------------------------------------------------- ;; ---- [INT] Sum of absolute differences ;; ------------------------------------------------------------------------- ;; The patterns in this section are synthetic. ;; -------------------------------------------------------------------------

;; Emit a sequence to produce a sum-of-absolute-differences of the inputs in ;; operands 1 and 2. The sequence also has to perform a widening reduction of ;; the difference into a vector and accumulate that into operand 3 before ;; copying that into the result operand 0. ;; Perform that with a sequence of: ;; MOV ones.b, #1 ;; [SU]ABD diff.b, p0/m, op1.b, op2.b ;; MOVPRFX op0, op3 // If necessary ;; UDOT op0.s, diff.b, ones.b (define_expand “sad” [(use (match_operand:SVE_FULL_SDI 0 “register_operand”)) (unspec: [(use (match_operand: 1 “register_operand”)) (use (match_operand: 2 “register_operand”))] ABAL) (use (match_operand:SVE_FULL_SDI 3 “register_operand”))] “TARGET_SVE” { rtx ones = force_reg (mode, CONST1_RTX (mode)); rtx diff = gen_reg_rtx (mode); emit_insn (gen_abd_3 (diff, operands[1], operands[2])); emit_insn (gen_udot_prod (operands[0], diff, ones, operands[3])); DONE; } )

;; ------------------------------------------------------------------------- ;; ---- [INT] Matrix multiply-accumulate ;; ------------------------------------------------------------------------- ;; Includes: ;; - SMMLA (I8MM) ;; - UMMLA (I8MM) ;; - USMMLA (I8MM) ;; -------------------------------------------------------------------------

(define_insn “@aarch64_sve_add_” [(set (match_operand:VNx4SI_ONLY 0 “register_operand” “=w, ?&w”) (plus:VNx4SI_ONLY (unspec:VNx4SI_ONLY [(match_operand: 2 “register_operand” “w, w”) (match_operand: 3 “register_operand” “w, w”)] MATMUL) (match_operand:VNx4SI_ONLY 1 “register_operand” “0, w”)))] “TARGET_SVE_I8MM” “@ mmla\t%0.s, %2.b, %3.b movprfx\t%0, %1;mmla\t%0.s, %2.b, %3.b” [(set_attr “movprfx” “*,yes”)] )

;; ------------------------------------------------------------------------- ;; ---- [FP] General ternary arithmetic corresponding to unspecs ;; ------------------------------------------------------------------------- ;; Includes merging patterns for: ;; - FMAD ;; - FMLA ;; - FMLS ;; - FMSB ;; - FNMAD ;; - FNMLA ;; - FNMLS ;; - FNMSB ;; -------------------------------------------------------------------------

;; Unpredicated floating-point ternary operations. (define_expand “4” [(set (match_operand:SVE_FULL_F 0 “register_operand”) (unspec:SVE_FULL_F [(match_dup 4) (const_int SVE_RELAXED_GP) (match_operand:SVE_FULL_F 1 “register_operand”) (match_operand:SVE_FULL_F 2 “register_operand”) (match_operand:SVE_FULL_F 3 “register_operand”)] SVE_COND_FP_TERNARY))] “TARGET_SVE” { operands[4] = aarch64_ptrue_reg (mode); } )

;; Predicated floating-point ternary operations. (define_insn “@aarch64_pred_” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=w, w, ?&w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl, Upl”) (match_operand:SI 5 “aarch64_sve_gp_strictness”) (match_operand:SVE_FULL_F 2 “register_operand” “%w, 0, w”) (match_operand:SVE_FULL_F 3 “register_operand” “w, w, w”) (match_operand:SVE_FULL_F 4 “register_operand” “0, w, w”)] SVE_COND_FP_TERNARY))] “TARGET_SVE” “@ <sve_fmla_op>\t%0., %1/m, %2., %3. <sve_fmad_op>\t%0., %1/m, %3., %4. movprfx\t%0, %4;<sve_fmla_op>\t%0., %1/m, %2., %3.” [(set_attr “movprfx” “,,yes”)] )

;; Predicated floating-point ternary operations with merging. (define_expand “@cond_” [(set (match_operand:SVE_FULL_F 0 “register_operand”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand”) (unspec:SVE_FULL_F [(match_dup 1) (const_int SVE_STRICT_GP) (match_operand:SVE_FULL_F 2 “register_operand”) (match_operand:SVE_FULL_F 3 “register_operand”) (match_operand:SVE_FULL_F 4 “register_operand”)] SVE_COND_FP_TERNARY) (match_operand:SVE_FULL_F 5 “aarch64_simd_reg_or_zero”)] UNSPEC_SEL))] “TARGET_SVE” { /* Swap the multiplication operands if the fallback value is the second of the two. */ if (rtx_equal_p (operands[3], operands[5])) std::swap (operands[2], operands[3]); })

;; Predicated floating-point ternary operations, merging with the ;; first input. (define_insn_and_rewrite “*cond__2_relaxed” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=w, ?&w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl”) (unspec:SVE_FULL_F [(match_operand 5) (const_int SVE_RELAXED_GP) (match_operand:SVE_FULL_F 2 “register_operand” “0, w”) (match_operand:SVE_FULL_F 3 “register_operand” “w, w”) (match_operand:SVE_FULL_F 4 “register_operand” “w, w”)] SVE_COND_FP_TERNARY) (match_dup 2)] UNSPEC_SEL))] “TARGET_SVE” “@ <sve_fmad_op>\t%0., %1/m, %3., %4. movprfx\t%0, %2;<sve_fmad_op>\t%0., %1/m, %3., %4.” “&& !rtx_equal_p (operands[1], operands[5])” { operands[5] = copy_rtx (operands[1]); } [(set_attr “movprfx” “*,yes”)] )

(define_insn “*cond__2_strict” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=w, ?&w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl”) (unspec:SVE_FULL_F [(match_dup 1) (const_int SVE_STRICT_GP) (match_operand:SVE_FULL_F 2 “register_operand” “0, w”) (match_operand:SVE_FULL_F 3 “register_operand” “w, w”) (match_operand:SVE_FULL_F 4 “register_operand” “w, w”)] SVE_COND_FP_TERNARY) (match_dup 2)] UNSPEC_SEL))] “TARGET_SVE” “@ <sve_fmad_op>\t%0., %1/m, %3., %4. movprfx\t%0, %2;<sve_fmad_op>\t%0., %1/m, %3., %4.” [(set_attr “movprfx” “*,yes”)] )

;; Predicated floating-point ternary operations, merging with the ;; third input. (define_insn_and_rewrite “*cond__4_relaxed” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=w, ?&w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl”) (unspec:SVE_FULL_F [(match_operand 5) (const_int SVE_RELAXED_GP) (match_operand:SVE_FULL_F 2 “register_operand” “w, w”) (match_operand:SVE_FULL_F 3 “register_operand” “w, w”) (match_operand:SVE_FULL_F 4 “register_operand” “0, w”)] SVE_COND_FP_TERNARY) (match_dup 4)] UNSPEC_SEL))] “TARGET_SVE” “@ <sve_fmla_op>\t%0., %1/m, %2., %3. movprfx\t%0, %4;<sve_fmla_op>\t%0., %1/m, %2., %3.” “&& !rtx_equal_p (operands[1], operands[5])” { operands[5] = copy_rtx (operands[1]); } [(set_attr “movprfx” “*,yes”)] )

(define_insn “*cond__4_strict” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=w, ?&w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl”) (unspec:SVE_FULL_F [(match_dup 1) (const_int SVE_STRICT_GP) (match_operand:SVE_FULL_F 2 “register_operand” “w, w”) (match_operand:SVE_FULL_F 3 “register_operand” “w, w”) (match_operand:SVE_FULL_F 4 “register_operand” “0, w”)] SVE_COND_FP_TERNARY) (match_dup 4)] UNSPEC_SEL))] “TARGET_SVE” “@ <sve_fmla_op>\t%0., %1/m, %2., %3. movprfx\t%0, %4;<sve_fmla_op>\t%0., %1/m, %2., %3.” [(set_attr “movprfx” “*,yes”)] )

;; Predicated floating-point ternary operations, merging with an ;; independent value. (define_insn_and_rewrite “*cond__any_relaxed” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=&w, &w, &w, &w, &w, ?&w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl, Upl, Upl, Upl, Upl”) (unspec:SVE_FULL_F [(match_operand 6) (const_int SVE_RELAXED_GP) (match_operand:SVE_FULL_F 2 “register_operand” “w, w, 0, w, w, w”) (match_operand:SVE_FULL_F 3 “register_operand” “w, w, w, 0, w, w”) (match_operand:SVE_FULL_F 4 “register_operand” “w, 0, w, w, w, w”)] SVE_COND_FP_TERNARY) (match_operand:SVE_FULL_F 5 “aarch64_simd_reg_or_zero” “Dz, Dz, Dz, Dz, 0, w”)] UNSPEC_SEL))] “TARGET_SVE && !rtx_equal_p (operands[2], operands[5]) && !rtx_equal_p (operands[3], operands[5]) && !rtx_equal_p (operands[4], operands[5])” “@ movprfx\t%0., %1/z, %4.;<sve_fmla_op>\t%0., %1/m, %2., %3. movprfx\t%0., %1/z, %0.;<sve_fmla_op>\t%0., %1/m, %2., %3. movprfx\t%0., %1/z, %0.;<sve_fmad_op>\t%0., %1/m, %3., %4. movprfx\t%0., %1/z, %0.;<sve_fmad_op>\t%0., %1/m, %2., %4. movprfx\t%0., %1/m, %4.;<sve_fmla_op>\t%0., %1/m, %2., %3. #” “&& 1” { if (reload_completed && register_operand (operands[5], mode) && !rtx_equal_p (operands[0], operands[5])) { emit_insn (gen_vcond_mask_ (operands[0], operands[4], operands[5], operands[1])); operands[5] = operands[4] = operands[0]; } else if (!rtx_equal_p (operands[1], operands[6])) operands[6] = copy_rtx (operands[1]); else FAIL; } [(set_attr “movprfx” “yes”)] )

(define_insn_and_rewrite “*cond__any_strict” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=&w, &w, &w, &w, &w, ?&w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl, Upl, Upl, Upl, Upl”) (unspec:SVE_FULL_F [(match_dup 1) (const_int SVE_STRICT_GP) (match_operand:SVE_FULL_F 2 “register_operand” “w, w, 0, w, w, w”) (match_operand:SVE_FULL_F 3 “register_operand” “w, w, w, 0, w, w”) (match_operand:SVE_FULL_F 4 “register_operand” “w, 0, w, w, w, w”)] SVE_COND_FP_TERNARY) (match_operand:SVE_FULL_F 5 “aarch64_simd_reg_or_zero” “Dz, Dz, Dz, Dz, 0, w”)] UNSPEC_SEL))] “TARGET_SVE && !rtx_equal_p (operands[2], operands[5]) && !rtx_equal_p (operands[3], operands[5]) && !rtx_equal_p (operands[4], operands[5])” “@ movprfx\t%0., %1/z, %4.;<sve_fmla_op>\t%0., %1/m, %2., %3. movprfx\t%0., %1/z, %0.;<sve_fmla_op>\t%0., %1/m, %2., %3. movprfx\t%0., %1/z, %0.;<sve_fmad_op>\t%0., %1/m, %3., %4. movprfx\t%0., %1/z, %0.;<sve_fmad_op>\t%0., %1/m, %2., %4. movprfx\t%0., %1/m, %4.;<sve_fmla_op>\t%0., %1/m, %2., %3. #” “&& reload_completed && register_operand (operands[5], mode) && !rtx_equal_p (operands[0], operands[5])” { emit_insn (gen_vcond_mask_ (operands[0], operands[4], operands[5], operands[1])); operands[5] = operands[4] = operands[0]; } [(set_attr “movprfx” “yes”)] )

;; Unpredicated FMLA and FMLS by selected lanes. It doesn‘t seem worth using ;; (fma ...) since target-independent code won’t understand the indexing. (define_insn “@aarch64_lane” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=w, ?&w”) (unspec:SVE_FULL_F [(match_operand:SVE_FULL_F 1 “register_operand” “w, w”) (unspec:SVE_FULL_F [(match_operand:SVE_FULL_F 2 “register_operand” “<sve_lane_con>, <sve_lane_con>”) (match_operand:SI 3 “const_int_operand”)] UNSPEC_SVE_LANE_SELECT) (match_operand:SVE_FULL_F 4 “register_operand” “0, w”)] SVE_FP_TERNARY_LANE))] “TARGET_SVE” “@ <sve_fp_op>\t%0., %1., %2.[%3] movprfx\t%0, %4;<sve_fp_op>\t%0., %1., %2.[%3]” [(set_attr “movprfx” “*,yes”)] )

;; ------------------------------------------------------------------------- ;; ---- [FP] Complex multiply-add ;; ------------------------------------------------------------------------- ;; Includes merging patterns for: ;; - FCMLA ;; -------------------------------------------------------------------------

;; Predicated FCMLA. (define_insn “@aarch64_pred_” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=w, ?&w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl”) (match_operand:SI 5 “aarch64_sve_gp_strictness”) (match_operand:SVE_FULL_F 2 “register_operand” “w, w”) (match_operand:SVE_FULL_F 3 “register_operand” “w, w”) (match_operand:SVE_FULL_F 4 “register_operand” “0, w”)] SVE_COND_FCMLA))] “TARGET_SVE” “@ fcmla\t%0., %1/m, %2., %3., # movprfx\t%0, %4;fcmla\t%0., %1/m, %2., %3., #” [(set_attr “movprfx” “*,yes”)] )

;; unpredicated optab pattern for auto-vectorizer ;; The complex mla/mls operations always need to expand to two instructions. ;; The first operation does half the computation and the second does the ;; remainder. Because of this, expand early. (define_expand “cml<conj_op>4” [(set (match_operand:SVE_FULL_F 0 “register_operand”) (unspec:SVE_FULL_F [(match_dup 4) (match_dup 5) (match_operand:SVE_FULL_F 1 “register_operand”) (match_operand:SVE_FULL_F 2 “register_operand”) (match_operand:SVE_FULL_F 3 “register_operand”)] FCMLA_OP))] “TARGET_SVE” { operands[4] = aarch64_ptrue_reg (mode); operands[5] = gen_int_mode (SVE_RELAXED_GP, SImode); rtx tmp = gen_reg_rtx (mode); emit_insn (gen_aarch64_pred_fcmla<sve_rot1> (tmp, operands[4], operands[3], operands[2], operands[1], operands[5])); emit_insn (gen_aarch64_pred_fcmla<sve_rot2> (operands[0], operands[4], operands[3], operands[2], tmp, operands[5])); DONE; })

;; unpredicated optab pattern for auto-vectorizer ;; The complex mul operations always need to expand to two instructions. ;; The first operation does half the computation and the second does the ;; remainder. Because of this, expand early. (define_expand “cmul<conj_op>3” [(set (match_operand:SVE_FULL_F 0 “register_operand”) (unspec:SVE_FULL_F [(match_operand:SVE_FULL_F 1 “register_operand”) (match_operand:SVE_FULL_F 2 “register_operand”)] FCMUL_OP))] “TARGET_SVE” { rtx pred_reg = aarch64_ptrue_reg (mode); rtx gp_mode = gen_int_mode (SVE_RELAXED_GP, SImode); rtx accum = force_reg (mode, CONST0_RTX (mode)); rtx tmp = gen_reg_rtx (mode); emit_insn (gen_aarch64_pred_fcmla<sve_rot1> (tmp, pred_reg, operands[2], operands[1], accum, gp_mode)); emit_insn (gen_aarch64_pred_fcmla<sve_rot2> (operands[0], pred_reg, operands[2], operands[1], tmp, gp_mode)); DONE; })

;; Predicated FCMLA with merging. (define_expand “@cond_” [(set (match_operand:SVE_FULL_F 0 “register_operand”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand”) (unspec:SVE_FULL_F [(match_dup 1) (const_int SVE_STRICT_GP) (match_operand:SVE_FULL_F 2 “register_operand”) (match_operand:SVE_FULL_F 3 “register_operand”) (match_operand:SVE_FULL_F 4 “register_operand”)] SVE_COND_FCMLA) (match_operand:SVE_FULL_F 5 “aarch64_simd_reg_or_zero”)] UNSPEC_SEL))] “TARGET_SVE” )

;; Predicated FCMLA, merging with the third input. (define_insn_and_rewrite “*cond__4_relaxed” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=w, ?&w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl”) (unspec:SVE_FULL_F [(match_operand 5) (const_int SVE_RELAXED_GP) (match_operand:SVE_FULL_F 2 “register_operand” “w, w”) (match_operand:SVE_FULL_F 3 “register_operand” “w, w”) (match_operand:SVE_FULL_F 4 “register_operand” “0, w”)] SVE_COND_FCMLA) (match_dup 4)] UNSPEC_SEL))] “TARGET_SVE” “@ fcmla\t%0., %1/m, %2., %3., # movprfx\t%0, %4;fcmla\t%0., %1/m, %2., %3., #” “&& !rtx_equal_p (operands[1], operands[5])” { operands[5] = copy_rtx (operands[1]); } [(set_attr “movprfx” “*,yes”)] )

(define_insn “*cond__4_strict” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=w, ?&w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl”) (unspec:SVE_FULL_F [(match_dup 1) (const_int SVE_STRICT_GP) (match_operand:SVE_FULL_F 2 “register_operand” “w, w”) (match_operand:SVE_FULL_F 3 “register_operand” “w, w”) (match_operand:SVE_FULL_F 4 “register_operand” “0, w”)] SVE_COND_FCMLA) (match_dup 4)] UNSPEC_SEL))] “TARGET_SVE” “@ fcmla\t%0., %1/m, %2., %3., # movprfx\t%0, %4;fcmla\t%0., %1/m, %2., %3., #” [(set_attr “movprfx” “*,yes”)] )

;; Predicated FCMLA, merging with an independent value. (define_insn_and_rewrite “*cond__any_relaxed” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=&w, &w, &w, ?&w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl, Upl, Upl”) (unspec:SVE_FULL_F [(match_operand 6) (const_int SVE_RELAXED_GP) (match_operand:SVE_FULL_F 2 “register_operand” “w, w, w, w”) (match_operand:SVE_FULL_F 3 “register_operand” “w, w, w, w”) (match_operand:SVE_FULL_F 4 “register_operand” “w, 0, w, w”)] SVE_COND_FCMLA) (match_operand:SVE_FULL_F 5 “aarch64_simd_reg_or_zero” “Dz, Dz, 0, w”)] UNSPEC_SEL))] “TARGET_SVE && !rtx_equal_p (operands[4], operands[5])” “@ movprfx\t%0., %1/z, %4.;fcmla\t%0., %1/m, %2., %3., # movprfx\t%0., %1/z, %0.;fcmla\t%0., %1/m, %2., %3., # movprfx\t%0., %1/m, %4.;fcmla\t%0., %1/m, %2., %3., # #” “&& 1” { if (reload_completed && register_operand (operands[5], mode) && !rtx_equal_p (operands[0], operands[5])) { emit_insn (gen_vcond_mask_ (operands[0], operands[4], operands[5], operands[1])); operands[5] = operands[4] = operands[0]; } else if (!rtx_equal_p (operands[1], operands[6])) operands[6] = copy_rtx (operands[1]); else FAIL; } [(set_attr “movprfx” “yes”)] )

(define_insn_and_rewrite “*cond__any_strict” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=&w, &w, &w, ?&w”) (unspec:SVE_FULL_F [(match_operand: 1 “register_operand” “Upl, Upl, Upl, Upl”) (unspec:SVE_FULL_F [(match_dup 1) (const_int SVE_STRICT_GP) (match_operand:SVE_FULL_F 2 “register_operand” “w, w, w, w”) (match_operand:SVE_FULL_F 3 “register_operand” “w, w, w, w”) (match_operand:SVE_FULL_F 4 “register_operand” “w, 0, w, w”)] SVE_COND_FCMLA) (match_operand:SVE_FULL_F 5 “aarch64_simd_reg_or_zero” “Dz, Dz, 0, w”)] UNSPEC_SEL))] “TARGET_SVE && !rtx_equal_p (operands[4], operands[5])” “@ movprfx\t%0., %1/z, %4.;fcmla\t%0., %1/m, %2., %3., # movprfx\t%0., %1/z, %0.;fcmla\t%0., %1/m, %2., %3., # movprfx\t%0., %1/m, %4.;fcmla\t%0., %1/m, %2., %3., # #” “&& reload_completed && register_operand (operands[5], mode) && !rtx_equal_p (operands[0], operands[5])” { emit_insn (gen_vcond_mask_ (operands[0], operands[4], operands[5], operands[1])); operands[5] = operands[4] = operands[0]; } [(set_attr “movprfx” “yes”)] )

;; Unpredicated FCMLA with indexing. (define_insn “@aarch64_lane” [(set (match_operand:SVE_FULL_HSF 0 “register_operand” “=w, ?&w”) (unspec:SVE_FULL_HSF [(match_operand:SVE_FULL_HSF 1 “register_operand” “w, w”) (unspec:SVE_FULL_HSF [(match_operand:SVE_FULL_HSF 2 “register_operand” “<sve_lane_pair_con>, <sve_lane_pair_con>”) (match_operand:SI 3 “const_int_operand”)] UNSPEC_SVE_LANE_SELECT) (match_operand:SVE_FULL_HSF 4 “register_operand” “0, w”)] FCMLA))] “TARGET_SVE” “@ fcmla\t%0., %1., %2.[%3], # movprfx\t%0, %4;fcmla\t%0., %1., %2.[%3], #” [(set_attr “movprfx” “*,yes”)] )

;; ------------------------------------------------------------------------- ;; ---- [FP] Trigonometric multiply-add ;; ------------------------------------------------------------------------- ;; Includes: ;; - FTMAD ;; -------------------------------------------------------------------------

(define_insn “@aarch64_sve_tmad” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=w, ?&w”) (unspec:SVE_FULL_F [(match_operand:SVE_FULL_F 1 “register_operand” “0, w”) (match_operand:SVE_FULL_F 2 “register_operand” “w, w”) (match_operand:DI 3 “const_int_operand”)] UNSPEC_FTMAD))] “TARGET_SVE” “@ ftmad\t%0., %0., %2., #%3 movprfx\t%0, %1;ftmad\t%0., %0., %2., #%3” [(set_attr “movprfx” “*,yes”)] )

;; ------------------------------------------------------------------------- ;; ---- [FP] Bfloat16 long ternary arithmetic (SF,BF,BF) ;; ------------------------------------------------------------------------- ;; Includes: ;; - BFDOT (BF16) ;; - BFMLALB (BF16) ;; - BFMLALT (BF16) ;; - BFMMLA (BF16) ;; -------------------------------------------------------------------------

(define_insn “@aarch64_sve_<sve_fp_op>vnx4sf” [(set (match_operand:VNx4SF 0 “register_operand” “=w, ?&w”) (unspec:VNx4SF [(match_operand:VNx4SF 1 “register_operand” “0, w”) (match_operand:VNx8BF 2 “register_operand” “w, w”) (match_operand:VNx8BF 3 “register_operand” “w, w”)] SVE_BFLOAT_TERNARY_LONG))] “TARGET_SVE_BF16” “@ <sve_fp_op>\t%0.s, %2.h, %3.h movprfx\t%0, %1;<sve_fp_op>\t%0.s, %2.h, %3.h” [(set_attr “movprfx” “*,yes”)] )

;; The immediate range is enforced before generating the instruction. (define_insn “@aarch64_sve_<sve_fp_op>_lanevnx4sf” [(set (match_operand:VNx4SF 0 “register_operand” “=w, ?&w”) (unspec:VNx4SF [(match_operand:VNx4SF 1 “register_operand” “0, w”) (match_operand:VNx8BF 2 “register_operand” “w, w”) (match_operand:VNx8BF 3 “register_operand” “y, y”) (match_operand:SI 4 “const_int_operand”)] SVE_BFLOAT_TERNARY_LONG_LANE))] “TARGET_SVE_BF16” “@ <sve_fp_op>\t%0.s, %2.h, %3.h[%4] movprfx\t%0, %1;<sve_fp_op>\t%0.s, %2.h, %3.h[%4]” [(set_attr “movprfx” “*,yes”)] )

;; ------------------------------------------------------------------------- ;; ---- [FP] Matrix multiply-accumulate ;; ------------------------------------------------------------------------- ;; Includes: ;; - FMMLA (F32MM,F64MM) ;; -------------------------------------------------------------------------

;; The mode iterator enforces the target requirements. (define_insn “@aarch64_sve_<sve_fp_op>” [(set (match_operand:SVE_MATMULF 0 “register_operand” “=w, ?&w”) (unspec:SVE_MATMULF [(match_operand:SVE_MATMULF 2 “register_operand” “w, w”) (match_operand:SVE_MATMULF 3 “register_operand” “w, w”) (match_operand:SVE_MATMULF 1 “register_operand” “0, w”)] FMMLA))] “TARGET_SVE” “@ <sve_fp_op>\t%0., %2., %3. movprfx\t%0, %1;<sve_fp_op>\t%0., %2., %3.” [(set_attr “movprfx” “*,yes”)] )

;; ========================================================================= ;; == Comparisons and selects ;; =========================================================================

;; ------------------------------------------------------------------------- ;; ---- [INT,FP] Select based on predicates ;; ------------------------------------------------------------------------- ;; Includes merging patterns for: ;; - FMOV ;; - MOV ;; - SEL ;; -------------------------------------------------------------------------

;; vcond_mask operand order: true, false, mask ;; UNSPEC_SEL operand order: mask, true, false (as for VEC_COND_EXPR) ;; SEL operand order: mask, true, false (define_expand “@vcond_mask_” [(set (match_operand:SVE_ALL 0 “register_operand”) (unspec:SVE_ALL [(match_operand: 3 “register_operand”) (match_operand:SVE_ALL 1 “aarch64_sve_reg_or_dup_imm”) (match_operand:SVE_ALL 2 “aarch64_simd_reg_or_zero”)] UNSPEC_SEL))] “TARGET_SVE” { if (register_operand (operands[1], mode)) operands[2] = force_reg (mode, operands[2]); } )

;; Selects between: ;; - two registers ;; - a duplicated immediate and a register ;; - a duplicated immediate and zero ;; ;; For unpacked vectors, it doesn't really matter whether SEL uses the ;; the container size or the element size. If SEL used the container size, ;; it would ignore undefined bits of the predicate but would copy the ;; upper (undefined) bits of each container along with the defined bits. ;; If SEL used the element size, it would use undefined bits of the predicate ;; to select between undefined elements in each input vector. Thus the only ;; difference is whether the undefined bits in a container always come from ;; the same input as the defined bits, or whether the choice can vary ;; independently of the defined bits. ;; ;; For the other instructions, using the element size is more natural, ;; so we do that for SEL as well. (define_insn “*vcond_mask_” [(set (match_operand:SVE_ALL 0 “register_operand” “=w, w, w, w, ?w, ?&w, ?&w”) (unspec:SVE_ALL [(match_operand: 3 “register_operand” “Upa, Upa, Upa, Upa, Upl, Upl, Upl”) (match_operand:SVE_ALL 1 “aarch64_sve_reg_or_dup_imm” “w, vss, vss, Ufc, Ufc, vss, Ufc”) (match_operand:SVE_ALL 2 “aarch64_simd_reg_or_zero” “w, 0, Dz, 0, Dz, w, w”)] UNSPEC_SEL))] “TARGET_SVE && (!register_operand (operands[1], mode) || register_operand (operands[2], mode))” “@ sel\t%0., %3, %1., %2. mov\t%0., %3/m, #%I1 mov\t%0., %3/z, #%I1 fmov\t%0., %3/m, #%1 movprfx\t%0., %3/z, %0.;fmov\t%0., %3/m, #%1 movprfx\t%0, %2;mov\t%0., %3/m, #%I1 movprfx\t%0, %2;fmov\t%0., %3/m, #%1” [(set_attr “movprfx” “,,,,yes,yes,yes”)] )

;; Optimize selects between a duplicated scalar variable and another vector, ;; the latter of which can be a zero constant or a variable. Treat duplicates ;; of GPRs as being more expensive than duplicates of FPRs, since they ;; involve a cross-file move. (define_insn “@aarch64_sel_dup” [(set (match_operand:SVE_ALL 0 “register_operand” “=?w, w, ??w, ?&w, ??&w, ?&w”) (unspec:SVE_ALL [(match_operand: 3 “register_operand” “Upl, Upl, Upl, Upl, Upl, Upl”) (vec_duplicate:SVE_ALL (match_operand: 1 “register_operand” “r, w, r, w, r, w”)) (match_operand:SVE_ALL 2 “aarch64_simd_reg_or_zero” “0, 0, Dz, Dz, w, w”)] UNSPEC_SEL))] “TARGET_SVE” “@ mov\t%0., %3/m, %1 mov\t%0., %3/m, %1 movprfx\t%0., %3/z, %0.;mov\t%0., %3/m, %1 movprfx\t%0., %3/z, %0.;mov\t%0., %3/m, %1 movprfx\t%0, %2;mov\t%0., %3/m, %1 movprfx\t%0, %2;mov\t%0., %3/m, %1” [(set_attr “movprfx” “,,yes,yes,yes,yes”)] )

;; ------------------------------------------------------------------------- ;; ---- [INT,FP] Compare and select ;; ------------------------------------------------------------------------- ;; The patterns in this section are synthetic. ;; -------------------------------------------------------------------------

;; Integer (signed) vcond. Don't enforce an immediate range here, since it ;; depends on the comparison; leave it to aarch64_expand_sve_vcond instead. (define_expand “vcond<SVE_ALL:mode><SVE_I:mode>” [(set (match_operand:SVE_ALL 0 “register_operand”) (if_then_else:SVE_ALL (match_operator 3 “comparison_operator” [(match_operand:SVE_I 4 “register_operand”) (match_operand:SVE_I 5 “nonmemory_operand”)]) (match_operand:SVE_ALL 1 “nonmemory_operand”) (match_operand:SVE_ALL 2 “nonmemory_operand”)))] “TARGET_SVE && <SVE_ALL:container_bits> == <SVE_I:container_bits>” { aarch64_expand_sve_vcond (<SVE_ALL:MODE>mode, <SVE_I:MODE>mode, operands); DONE; } )

;; Integer vcondu. Don't enforce an immediate range here, since it ;; depends on the comparison; leave it to aarch64_expand_sve_vcond instead. (define_expand “vcondu<SVE_ALL:mode><SVE_I:mode>” [(set (match_operand:SVE_ALL 0 “register_operand”) (if_then_else:SVE_ALL (match_operator 3 “comparison_operator” [(match_operand:SVE_I 4 “register_operand”) (match_operand:SVE_I 5 “nonmemory_operand”)]) (match_operand:SVE_ALL 1 “nonmemory_operand”) (match_operand:SVE_ALL 2 “nonmemory_operand”)))] “TARGET_SVE && <SVE_ALL:container_bits> == <SVE_I:container_bits>” { aarch64_expand_sve_vcond (<SVE_ALL:MODE>mode, <SVE_I:MODE>mode, operands); DONE; } )

;; Floating-point vcond. All comparisons except FCMUO allow a zero operand; ;; aarch64_expand_sve_vcond handles the case of an FCMUO with zero. (define_expand “vcond<v_fp_equiv>” [(set (match_operand:SVE_FULL_HSD 0 “register_operand”) (if_then_else:SVE_FULL_HSD (match_operator 3 “comparison_operator” [(match_operand:<V_FP_EQUIV> 4 “register_operand”) (match_operand:<V_FP_EQUIV> 5 “aarch64_simd_reg_or_zero”)]) (match_operand:SVE_FULL_HSD 1 “nonmemory_operand”) (match_operand:SVE_FULL_HSD 2 “nonmemory_operand”)))] “TARGET_SVE” { aarch64_expand_sve_vcond (mode, <V_FP_EQUIV>mode, operands); DONE; } )

;; ------------------------------------------------------------------------- ;; ---- [INT] Comparisons ;; ------------------------------------------------------------------------- ;; Includes: ;; - CMPEQ ;; - CMPGE ;; - CMPGT ;; - CMPHI ;; - CMPHS ;; - CMPLE ;; - CMPLO ;; - CMPLS ;; - CMPLT ;; - CMPNE ;; -------------------------------------------------------------------------

;; Signed integer comparisons. Don't enforce an immediate range here, since ;; it depends on the comparison; leave it to aarch64_expand_sve_vec_cmp_int ;; instead. (define_expand “vec_cmp” [(parallel [(set (match_operand: 0 “register_operand”) (match_operator: 1 “comparison_operator” [(match_operand:SVE_I 2 “register_operand”) (match_operand:SVE_I 3 “nonmemory_operand”)])) (clobber (reg:CC_NZC CC_REGNUM))])] “TARGET_SVE” { aarch64_expand_sve_vec_cmp_int (operands[0], GET_CODE (operands[1]), operands[2], operands[3]); DONE; } )

;; Unsigned integer comparisons. Don't enforce an immediate range here, since ;; it depends on the comparison; leave it to aarch64_expand_sve_vec_cmp_int ;; instead. (define_expand “vec_cmpu” [(parallel [(set (match_operand: 0 “register_operand”) (match_operator: 1 “comparison_operator” [(match_operand:SVE_I 2 “register_operand”) (match_operand:SVE_I 3 “nonmemory_operand”)])) (clobber (reg:CC_NZC CC_REGNUM))])] “TARGET_SVE” { aarch64_expand_sve_vec_cmp_int (operands[0], GET_CODE (operands[1]), operands[2], operands[3]); DONE; } )

;; Predicated integer comparisons. ;; ;; For unpacked vectors, only the lowpart element in each input container ;; has a defined value, and only the predicate bits associated with ;; those elements are defined. For example, when comparing two VNx2SIs: ;; ;; - The VNx2SIs can be seem as VNx2DIs in which the low halves of each ;; DI container store an SI element. The upper bits of each DI container ;; are undefined. ;; ;; - Alternatively, the VNx2SIs can be seen as VNx4SIs in which the ;; even elements are defined and the odd elements are undefined. ;; ;; - The associated predicate mode is VNx2BI. This means that only the ;; low bit in each predicate byte is defined (on input and on output). ;; ;; - We use a .s comparison to compare VNx2SIs, under the control of a ;; VNx2BI governing predicate, to produce a VNx2BI result. If we view ;; the .s operation as operating on VNx4SIs then for odd lanes: ;; ;; - the input governing predicate bit is undefined ;; - the SI elements being compared are undefined ;; - the predicate result bit is therefore undefined, but ;; - the predicate result bit is in the undefined part of a VNx2BI, ;; so its value doesn't matter anyway. (define_insn “@aarch64_pred_cmp<cmp_op>” [(set (match_operand: 0 “register_operand” “=Upa, Upa”) (unspec: [(match_operand: 1 “register_operand” “Upl, Upl”) (match_operand:SI 2 “aarch64_sve_ptrue_flag”) (SVE_INT_CMP: (match_operand:SVE_I 3 “register_operand” “w, w”) (match_operand:SVE_I 4 “aarch64_sve_cmp_<sve_imm_con>_operand” “<sve_imm_con>, w”))] UNSPEC_PRED_Z)) (clobber (reg:CC_NZC CC_REGNUM))] “TARGET_SVE” “@ cmp<cmp_op>\t%0., %1/z, %3., #%4 cmp<cmp_op>\t%0., %1/z, %3., %4.” )

;; Predicated integer comparisons in which both the flag and predicate ;; results are interesting. (define_insn_and_rewrite “*cmp<cmp_op>_cc” [(set (reg:CC_NZC CC_REGNUM) (unspec:CC_NZC [(match_operand:VNx16BI 1 “register_operand” “Upl, Upl”) (match_operand 4) (match_operand:SI 5 “aarch64_sve_ptrue_flag”) (unspec: [(match_operand 6) (match_operand:SI 7 “aarch64_sve_ptrue_flag”) (SVE_INT_CMP: (match_operand:SVE_I 2 “register_operand” “w, w”) (match_operand:SVE_I 3 “aarch64_sve_cmp_<sve_imm_con>_operand” “<sve_imm_con>, w”))] UNSPEC_PRED_Z)] UNSPEC_PTEST)) (set (match_operand: 0 “register_operand” “=Upa, Upa”) (unspec: [(match_dup 6) (match_dup 7) (SVE_INT_CMP: (match_dup 2) (match_dup 3))] UNSPEC_PRED_Z))] “TARGET_SVE && aarch64_sve_same_pred_for_ptest_p (&operands[4], &operands[6])” “@ cmp<cmp_op>\t%0., %1/z, %2., #%3 cmp<cmp_op>\t%0., %1/z, %2., %3.” “&& !rtx_equal_p (operands[4], operands[6])” { operands[6] = copy_rtx (operands[4]); operands[7] = operands[5]; } )

;; Predicated integer comparisons in which only the flags result is ;; interesting. (define_insn_and_rewrite “*cmp<cmp_op>_ptest” [(set (reg:CC_NZC CC_REGNUM) (unspec:CC_NZC [(match_operand:VNx16BI 1 “register_operand” “Upl, Upl”) (match_operand 4) (match_operand:SI 5 “aarch64_sve_ptrue_flag”) (unspec: [(match_operand 6) (match_operand:SI 7 “aarch64_sve_ptrue_flag”) (SVE_INT_CMP: (match_operand:SVE_I 2 “register_operand” “w, w”) (match_operand:SVE_I 3 “aarch64_sve_cmp_<sve_imm_con>_operand” “<sve_imm_con>, w”))] UNSPEC_PRED_Z)] UNSPEC_PTEST)) (clobber (match_scratch: 0 “=Upa, Upa”))] “TARGET_SVE && aarch64_sve_same_pred_for_ptest_p (&operands[4], &operands[6])” “@ cmp<cmp_op>\t%0., %1/z, %2., #%3 cmp<cmp_op>\t%0., %1/z, %2., %3.” “&& !rtx_equal_p (operands[4], operands[6])” { operands[6] = copy_rtx (operands[4]); operands[7] = operands[5]; } )

;; Predicated integer comparisons, formed by combining a PTRUE-predicated ;; comparison with an AND. Split the instruction into its preferred form ;; at the earliest opportunity, in order to get rid of the redundant ;; operand 4. (define_insn_and_split “*cmp<cmp_op>_and” [(set (match_operand: 0 “register_operand” “=Upa, Upa”) (and: (unspec: [(match_operand 4) (const_int SVE_KNOWN_PTRUE) (SVE_INT_CMP: (match_operand:SVE_I 2 “register_operand” “w, w”) (match_operand:SVE_I 3 “aarch64_sve_cmp_<sve_imm_con>_operand” “<sve_imm_con>, w”))] UNSPEC_PRED_Z) (match_operand: 1 “register_operand” “Upl, Upl”))) (clobber (reg:CC_NZC CC_REGNUM))] “TARGET_SVE” “#” “&& 1” [(parallel [(set (match_dup 0) (unspec: [(match_dup 1) (const_int SVE_MAYBE_NOT_PTRUE) (SVE_INT_CMP: (match_dup 2) (match_dup 3))] UNSPEC_PRED_Z)) (clobber (reg:CC_NZC CC_REGNUM))])] )

;; Predicated integer wide comparisons. (define_insn “@aarch64_pred_cmp<cmp_op>_wide” [(set (match_operand: 0 “register_operand” “=Upa”) (unspec: [(match_operand:VNx16BI 1 “register_operand” “Upl”) (match_operand:SI 2 “aarch64_sve_ptrue_flag”) (unspec: [(match_operand:SVE_FULL_BHSI 3 “register_operand” “w”) (match_operand:VNx2DI 4 “register_operand” “w”)] SVE_COND_INT_CMP_WIDE)] UNSPEC_PRED_Z)) (clobber (reg:CC_NZC CC_REGNUM))] “TARGET_SVE” “cmp<cmp_op>\t%0., %1/z, %3., %4.d” )

;; Predicated integer wide comparisons in which both the flag and ;; predicate results are interesting. (define_insn “*aarch64_pred_cmp<cmp_op>_wide_cc” [(set (reg:CC_NZC CC_REGNUM) (unspec:CC_NZC [(match_operand:VNx16BI 1 “register_operand” “Upl”) (match_operand 4) (match_operand:SI 5 “aarch64_sve_ptrue_flag”) (unspec: [(match_operand:VNx16BI 6 “register_operand” “Upl”) (match_operand:SI 7 “aarch64_sve_ptrue_flag”) (unspec: [(match_operand:SVE_FULL_BHSI 2 “register_operand” “w”) (match_operand:VNx2DI 3 “register_operand” “w”)] SVE_COND_INT_CMP_WIDE)] UNSPEC_PRED_Z)] UNSPEC_PTEST)) (set (match_operand: 0 “register_operand” “=Upa”) (unspec: [(match_dup 6) (match_dup 7) (unspec: [(match_dup 2) (match_dup 3)] SVE_COND_INT_CMP_WIDE)] UNSPEC_PRED_Z))] “TARGET_SVE && aarch64_sve_same_pred_for_ptest_p (&operands[4], &operands[6])” “cmp<cmp_op>\t%0., %1/z, %2., %3.d” )

;; Predicated integer wide comparisons in which only the flags result ;; is interesting. (define_insn “*aarch64_pred_cmp<cmp_op>_wide_ptest” [(set (reg:CC_NZC CC_REGNUM) (unspec:CC_NZC [(match_operand:VNx16BI 1 “register_operand” “Upl”) (match_operand 4) (match_operand:SI 5 “aarch64_sve_ptrue_flag”) (unspec: [(match_operand:VNx16BI 6 “register_operand” “Upl”) (match_operand:SI 7 “aarch64_sve_ptrue_flag”) (unspec: [(match_operand:SVE_FULL_BHSI 2 “register_operand” “w”) (match_operand:VNx2DI 3 “register_operand” “w”)] SVE_COND_INT_CMP_WIDE)] UNSPEC_PRED_Z)] UNSPEC_PTEST)) (clobber (match_scratch: 0 “=Upa”))] “TARGET_SVE && aarch64_sve_same_pred_for_ptest_p (&operands[4], &operands[6])” “cmp<cmp_op>\t%0., %1/z, %2., %3.d” )

;; ------------------------------------------------------------------------- ;; ---- [INT] While tests ;; ------------------------------------------------------------------------- ;; Includes: ;; - WHILEGE (SVE2) ;; - WHILEGT (SVE2) ;; - WHILEHI (SVE2) ;; - WHILEHS (SVE2) ;; - WHILELE ;; - WHILELO ;; - WHILELS ;; - WHILELT ;; - WHILERW (SVE2) ;; - WHILEWR (SVE2) ;; -------------------------------------------------------------------------

;; Set element I of the result if (cmp (plus operand1 J) operand2) is ;; true for all J in [0, I]. (define_insn “@while_<while_optab_cmp>GPI:mode<PRED_ALL:mode>” [(set (match_operand:PRED_ALL 0 “register_operand” “=Upa”) (unspec:PRED_ALL [(match_operand:GPI 1 “aarch64_reg_or_zero” “rZ”) (match_operand:GPI 2 “aarch64_reg_or_zero” “rZ”)] SVE_WHILE)) (clobber (reg:CC_NZC CC_REGNUM))] “TARGET_SVE” “while<cmp_op>\t%0.<PRED_ALL:Vetype>, %1, %2” )

;; The WHILE instructions set the flags in the same way as a PTEST with ;; a PTRUE GP. Handle the case in which both results are useful. The GP ;; operands to the PTEST aren‘t needed, so we allow them to be anything. (define_insn_and_rewrite “*while_<while_optab_cmp>GPI:mode<PRED_ALL:mode>_cc” [(set (reg:CC_NZC CC_REGNUM) (unspec:CC_NZC [(match_operand 3) (match_operand 4) (const_int SVE_KNOWN_PTRUE) (unspec:PRED_ALL [(match_operand:GPI 1 “aarch64_reg_or_zero” “rZ”) (match_operand:GPI 2 “aarch64_reg_or_zero” “rZ”)] SVE_WHILE)] UNSPEC_PTEST)) (set (match_operand:PRED_ALL 0 “register_operand” “=Upa”) (unspec:PRED_ALL [(match_dup 1) (match_dup 2)] SVE_WHILE))] “TARGET_SVE” “while<cmp_op>\t%0.<PRED_ALL:Vetype>, %1, %2” ;; Force the compiler to drop the unused predicate operand, so that we ;; don’t have an unnecessary PTRUE. “&& (!CONSTANT_P (operands[3]) || !CONSTANT_P (operands[4]))” { operands[3] = CONSTM1_RTX (VNx16BImode); operands[4] = CONSTM1_RTX (<PRED_ALL:MODE>mode); } )

;; Same, but handle the case in which only the flags result is useful. (define_insn_and_rewrite “@while_<while_optab_cmp>GPI:mode<PRED_ALL:mode>_ptest” [(set (reg:CC_NZC CC_REGNUM) (unspec:CC_NZC [(match_operand 3) (match_operand 4) (const_int SVE_KNOWN_PTRUE) (unspec:PRED_ALL [(match_operand:GPI 1 “aarch64_reg_or_zero” “rZ”) (match_operand:GPI 2 “aarch64_reg_or_zero” “rZ”)] SVE_WHILE)] UNSPEC_PTEST)) (clobber (match_scratch:PRED_ALL 0 “=Upa”))] “TARGET_SVE” “while<cmp_op>\t%0.<PRED_ALL:Vetype>, %1, %2” ;; Force the compiler to drop the unused predicate operand, so that we ;; don't have an unnecessary PTRUE. “&& (!CONSTANT_P (operands[3]) || !CONSTANT_P (operands[4]))” { operands[3] = CONSTM1_RTX (VNx16BImode); operands[4] = CONSTM1_RTX (<PRED_ALL:MODE>mode); } )

;; ------------------------------------------------------------------------- ;; ---- [FP] Direct comparisons ;; ------------------------------------------------------------------------- ;; Includes: ;; - FCMEQ ;; - FCMGE ;; - FCMGT ;; - FCMLE ;; - FCMLT ;; - FCMNE ;; - FCMUO ;; -------------------------------------------------------------------------

;; Floating-point comparisons. All comparisons except FCMUO allow a zero ;; operand; aarch64_expand_sve_vec_cmp_float handles the case of an FCMUO ;; with zero. (define_expand “vec_cmp” [(set (match_operand: 0 “register_operand”) (match_operator: 1 “comparison_operator” [(match_operand:SVE_FULL_F 2 “register_operand”) (match_operand:SVE_FULL_F 3 “aarch64_simd_reg_or_zero”)]))] “TARGET_SVE” { aarch64_expand_sve_vec_cmp_float (operands[0], GET_CODE (operands[1]), operands[2], operands[3], false); DONE; } )

;; Predicated floating-point comparisons. (define_insn “@aarch64_pred_fcm<cmp_op>” [(set (match_operand: 0 “register_operand” “=Upa, Upa”) (unspec: [(match_operand: 1 “register_operand” “Upl, Upl”) (match_operand:SI 2 “aarch64_sve_ptrue_flag”) (match_operand:SVE_FULL_F 3 “register_operand” “w, w”) (match_operand:SVE_FULL_F 4 “aarch64_simd_reg_or_zero” “Dz, w”)] SVE_COND_FP_CMP_I0))] “TARGET_SVE” “@ fcm<cmp_op>\t%0., %1/z, %3., #0.0 fcm<cmp_op>\t%0., %1/z, %3., %4.” )

;; Same for unordered comparisons. (define_insn “@aarch64_pred_fcmuo” [(set (match_operand: 0 “register_operand” “=Upa”) (unspec: [(match_operand: 1 “register_operand” “Upl”) (match_operand:SI 2 “aarch64_sve_ptrue_flag”) (match_operand:SVE_FULL_F 3 “register_operand” “w”) (match_operand:SVE_FULL_F 4 “register_operand” “w”)] UNSPEC_COND_FCMUO))] “TARGET_SVE” “fcmuo\t%0., %1/z, %3., %4.” )

;; Floating-point comparisons predicated on a PTRUE, with the results ANDed ;; with another predicate P. This does not have the same trapping behavior ;; as predicating the comparison itself on P, but it's a legitimate fold, ;; since we can drop any potentially-trapping operations whose results ;; are not needed. ;; ;; Split the instruction into its preferred form (below) at the earliest ;; opportunity, in order to get rid of the redundant operand 1. (define_insn_and_split “*fcm<cmp_op>_and_combine” [(set (match_operand: 0 “register_operand” “=Upa, Upa”) (and: (unspec: [(match_operand: 1) (const_int SVE_KNOWN_PTRUE) (match_operand:SVE_FULL_F 2 “register_operand” “w, w”) (match_operand:SVE_FULL_F 3 “aarch64_simd_reg_or_zero” “Dz, w”)] SVE_COND_FP_CMP_I0) (match_operand: 4 “register_operand” “Upl, Upl”)))] “TARGET_SVE” “#” “&& 1” [(set (match_dup 0) (unspec: [(match_dup 4) (const_int SVE_MAYBE_NOT_PTRUE) (match_dup 2) (match_dup 3)] SVE_COND_FP_CMP_I0))] )

;; Same for unordered comparisons. (define_insn_and_split “*fcmuo_and_combine” [(set (match_operand: 0 “register_operand” “=Upa”) (and: (unspec: [(match_operand: 1) (const_int SVE_KNOWN_PTRUE) (match_operand:SVE_FULL_F 2 “register_operand” “w”) (match_operand:SVE_FULL_F 3 “register_operand” “w”)] UNSPEC_COND_FCMUO) (match_operand: 4 “register_operand” “Upl”)))] “TARGET_SVE” “#” “&& 1” [(set (match_dup 0) (unspec: [(match_dup 4) (const_int SVE_MAYBE_NOT_PTRUE) (match_dup 2) (match_dup 3)] UNSPEC_COND_FCMUO))] )

;; ------------------------------------------------------------------------- ;; ---- [FP] Absolute comparisons ;; ------------------------------------------------------------------------- ;; Includes: ;; - FACGE ;; - FACGT ;; - FACLE ;; - FACLT ;; -------------------------------------------------------------------------

;; Predicated floating-point absolute comparisons. (define_expand “@aarch64_pred_fac<cmp_op>” [(set (match_operand: 0 “register_operand”) (unspec: [(match_operand: 1 “register_operand”) (match_operand:SI 2 “aarch64_sve_ptrue_flag”) (unspec:SVE_FULL_F [(match_dup 1) (match_dup 2) (match_operand:SVE_FULL_F 3 “register_operand”)] UNSPEC_COND_FABS) (unspec:SVE_FULL_F [(match_dup 1) (match_dup 2) (match_operand:SVE_FULL_F 4 “register_operand”)] UNSPEC_COND_FABS)] SVE_COND_FP_ABS_CMP))] “TARGET_SVE” )

(define_insn_and_rewrite “*aarch64_pred_fac<cmp_op>_relaxed” [(set (match_operand: 0 “register_operand” “=Upa”) (unspec: [(match_operand: 1 “register_operand” “Upl”) (match_operand:SI 4 “aarch64_sve_ptrue_flag”) (unspec:SVE_FULL_F [(match_operand 5) (const_int SVE_RELAXED_GP) (match_operand:SVE_FULL_F 2 “register_operand” “w”)] UNSPEC_COND_FABS) (unspec:SVE_FULL_F [(match_operand 6) (const_int SVE_RELAXED_GP) (match_operand:SVE_FULL_F 3 “register_operand” “w”)] UNSPEC_COND_FABS)] SVE_COND_FP_ABS_CMP))] “TARGET_SVE” “fac<cmp_op>\t%0., %1/z, %2., %3.” “&& (!rtx_equal_p (operands[1], operands[5]) || !rtx_equal_p (operands[1], operands[6]))” { operands[5] = copy_rtx (operands[1]); operands[6] = copy_rtx (operands[1]); } )

(define_insn “*aarch64_pred_fac<cmp_op>_strict” [(set (match_operand: 0 “register_operand” “=Upa”) (unspec: [(match_operand: 1 “register_operand” “Upl”) (match_operand:SI 4 “aarch64_sve_ptrue_flag”) (unspec:SVE_FULL_F [(match_dup 1) (match_operand:SI 5 “aarch64_sve_gp_strictness”) (match_operand:SVE_FULL_F 2 “register_operand” “w”)] UNSPEC_COND_FABS) (unspec:SVE_FULL_F [(match_dup 1) (match_operand:SI 6 “aarch64_sve_gp_strictness”) (match_operand:SVE_FULL_F 3 “register_operand” “w”)] UNSPEC_COND_FABS)] SVE_COND_FP_ABS_CMP))] “TARGET_SVE” “fac<cmp_op>\t%0., %1/z, %2., %3.” )

;; ------------------------------------------------------------------------- ;; ---- [PRED] Select ;; ------------------------------------------------------------------------- ;; Includes: ;; - SEL ;; -------------------------------------------------------------------------

(define_insn “@vcond_mask_” [(set (match_operand:PRED_ALL 0 “register_operand” “=Upa”) (ior:PRED_ALL (and:PRED_ALL (match_operand:PRED_ALL 3 “register_operand” “Upa”) (match_operand:PRED_ALL 1 “register_operand” “Upa”)) (and:PRED_ALL (not (match_dup 3)) (match_operand:PRED_ALL 2 “register_operand” “Upa”))))] “TARGET_SVE” “sel\t%0.b, %3, %1.b, %2.b” )

;; ------------------------------------------------------------------------- ;; ---- [PRED] Test bits ;; ------------------------------------------------------------------------- ;; Includes: ;; - PTEST ;; -------------------------------------------------------------------------

;; Branch based on predicate equality or inequality. (define_expand “cbranch4” [(set (pc) (if_then_else (match_operator 0 “aarch64_equality_operator” [(match_operand:PRED_ALL 1 “register_operand”) (match_operand:PRED_ALL 2 “aarch64_simd_reg_or_zero”)]) (label_ref (match_operand 3 "")) (pc)))] "" { rtx ptrue = force_reg (VNx16BImode, aarch64_ptrue_all (<data_bytes>)); rtx cast_ptrue = gen_lowpart (mode, ptrue); rtx ptrue_flag = gen_int_mode (SVE_KNOWN_PTRUE, SImode); rtx pred; if (operands[2] == CONST0_RTX (mode)) pred = operands[1]; else { pred = gen_reg_rtx (mode); emit_insn (gen_aarch64_pred_xor_z (pred, cast_ptrue, operands[1], operands[2])); } emit_insn (gen_aarch64_ptest (ptrue, cast_ptrue, ptrue_flag, pred)); operands[1] = gen_rtx_REG (CC_NZCmode, CC_REGNUM); operands[2] = const0_rtx; } )

;; See “Description of UNSPEC_PTEST” above for details. (define_insn “aarch64_ptest” [(set (reg:CC_NZC CC_REGNUM) (unspec:CC_NZC [(match_operand:VNx16BI 0 “register_operand” “Upa”) (match_operand 1) (match_operand:SI 2 “aarch64_sve_ptrue_flag”) (match_operand:PRED_ALL 3 “register_operand” “Upa”)] UNSPEC_PTEST))] “TARGET_SVE” “ptest\t%0, %3.b” )

;; ========================================================================= ;; == Reductions ;; =========================================================================

;; ------------------------------------------------------------------------- ;; ---- [INT,FP] Conditional reductions ;; ------------------------------------------------------------------------- ;; Includes: ;; - CLASTA ;; - CLASTB ;; -------------------------------------------------------------------------

;; Set operand 0 to the last active element in operand 3, or to tied ;; operand 1 if no elements are active. (define_insn “@fold_extract_<last_op>_” [(set (match_operand: 0 “register_operand” “=?r, w”) (unspec: [(match_operand: 1 “register_operand” “0, 0”) (match_operand: 2 “register_operand” “Upl, Upl”) (match_operand:SVE_FULL 3 “register_operand” “w, w”)] CLAST))] “TARGET_SVE” “@ clast\t%0, %2, %0, %3. clast\t%0, %2, %0, %3.” )

(define_insn “@aarch64_fold_extract_vector_<last_op>_” [(set (match_operand:SVE_FULL 0 “register_operand” “=w, ?&w”) (unspec:SVE_FULL [(match_operand:SVE_FULL 1 “register_operand” “0, w”) (match_operand: 2 “register_operand” “Upl, Upl”) (match_operand:SVE_FULL 3 “register_operand” “w, w”)] CLAST))] “TARGET_SVE” “@ clast\t%0., %2, %0., %3. movprfx\t%0, %1;clast\t%0., %2, %0., %3.” )

;; ------------------------------------------------------------------------- ;; ---- [INT] Tree reductions ;; ------------------------------------------------------------------------- ;; Includes: ;; - ANDV ;; - EORV ;; - ORV ;; - SADDV ;; - SMAXV ;; - SMINV ;; - UADDV ;; - UMAXV ;; - UMINV ;; -------------------------------------------------------------------------

;; Unpredicated integer add reduction. (define_expand “reduc_plus_scal_” [(match_operand: 0 “register_operand”) (match_operand:SVE_FULL_I 1 “register_operand”)] “TARGET_SVE” { rtx pred = aarch64_ptrue_reg (mode); rtx tmp = mode == DImode ? operands[0] : gen_reg_rtx (DImode); emit_insn (gen_aarch64_pred_reduc_uadd_ (tmp, pred, operands[1])); if (tmp != operands[0]) emit_move_insn (operands[0], gen_lowpart (mode, tmp)); DONE; } )

;; Predicated integer add reduction. The result is always 64-bits. (define_insn “@aarch64_pred_reduc__” [(set (match_operand:DI 0 “register_operand” “=w”) (unspec:DI [(match_operand: 1 “register_operand” “Upl”) (match_operand:SVE_FULL_I 2 “register_operand” “w”)] SVE_INT_ADDV))] “TARGET_SVE && <max_elem_bits> >= <elem_bits>” “addv\t%d0, %1, %2.” )

;; Unpredicated integer reductions. (define_expand “reduc_scal” [(set (match_operand: 0 “register_operand”) (unspec: [(match_dup 2) (match_operand:SVE_FULL_I 1 “register_operand”)] SVE_INT_REDUCTION))] “TARGET_SVE” { operands[2] = aarch64_ptrue_reg (mode); } )

;; Predicated integer reductions. (define_insn “@aarch64_pred_reduc__” [(set (match_operand: 0 “register_operand” “=w”) (unspec: [(match_operand: 1 “register_operand” “Upl”) (match_operand:SVE_FULL_I 2 “register_operand” “w”)] SVE_INT_REDUCTION))] “TARGET_SVE” “<sve_int_op>\t%0, %1, %2.” )

;; ------------------------------------------------------------------------- ;; ---- [FP] Tree reductions ;; ------------------------------------------------------------------------- ;; Includes: ;; - FADDV ;; - FMAXNMV ;; - FMAXV ;; - FMINNMV ;; - FMINV ;; -------------------------------------------------------------------------

;; Unpredicated floating-point tree reductions. (define_expand “reduc_scal” [(set (match_operand: 0 “register_operand”) (unspec: [(match_dup 2) (match_operand:SVE_FULL_F 1 “register_operand”)] SVE_FP_REDUCTION))] “TARGET_SVE” { operands[2] = aarch64_ptrue_reg (mode); } )

;; Predicated floating-point tree reductions. (define_insn “@aarch64_pred_reduc__” [(set (match_operand: 0 “register_operand” “=w”) (unspec: [(match_operand: 1 “register_operand” “Upl”) (match_operand:SVE_FULL_F 2 “register_operand” “w”)] SVE_FP_REDUCTION))] “TARGET_SVE” “<sve_fp_op>\t%0, %1, %2.” )

;; ------------------------------------------------------------------------- ;; ---- [FP] Left-to-right reductions ;; ------------------------------------------------------------------------- ;; Includes: ;; - FADDA ;; -------------------------------------------------------------------------

;; Unpredicated in-order FP reductions. (define_expand “fold_left_plus_” [(set (match_operand: 0 “register_operand”) (unspec: [(match_dup 3) (match_operand: 1 “register_operand”) (match_operand:SVE_FULL_F 2 “register_operand”)] UNSPEC_FADDA))] “TARGET_SVE” { operands[3] = aarch64_ptrue_reg (mode); } )

;; Predicated in-order FP reductions. (define_insn “mask_fold_left_plus_” [(set (match_operand: 0 “register_operand” “=w”) (unspec: [(match_operand: 3 “register_operand” “Upl”) (match_operand: 1 “register_operand” “0”) (match_operand:SVE_FULL_F 2 “register_operand” “w”)] UNSPEC_FADDA))] “TARGET_SVE” “fadda\t%0, %3, %0, %2.” )

;; ========================================================================= ;; == Permutes ;; =========================================================================

;; ------------------------------------------------------------------------- ;; ---- [INT,FP] General permutes ;; ------------------------------------------------------------------------- ;; Includes: ;; - TBL ;; -------------------------------------------------------------------------

(define_expand “vec_perm” [(match_operand:SVE_FULL 0 “register_operand”) (match_operand:SVE_FULL 1 “register_operand”) (match_operand:SVE_FULL 2 “register_operand”) (match_operand:<V_INT_EQUIV> 3 “aarch64_sve_vec_perm_operand”)] “TARGET_SVE && GET_MODE_NUNITS (mode).is_constant ()” { aarch64_expand_sve_vec_perm (operands[0], operands[1], operands[2], operands[3]); DONE; } )

(define_insn “@aarch64_sve_tbl” [(set (match_operand:SVE_FULL 0 “register_operand” “=w”) (unspec:SVE_FULL [(match_operand:SVE_FULL 1 “register_operand” “w”) (match_operand:<V_INT_EQUIV> 2 “register_operand” “w”)] UNSPEC_TBL))] “TARGET_SVE” “tbl\t%0., %1., %2.” )

;; ------------------------------------------------------------------------- ;; ---- [INT,FP] Special-purpose unary permutes ;; ------------------------------------------------------------------------- ;; Includes: ;; - COMPACT ;; - DUP ;; - REV ;; -------------------------------------------------------------------------

;; Compact active elements and pad with zeros. (define_insn “@aarch64_sve_compact” [(set (match_operand:SVE_FULL_SD 0 “register_operand” “=w”) (unspec:SVE_FULL_SD [(match_operand: 1 “register_operand” “Upl”) (match_operand:SVE_FULL_SD 2 “register_operand” “w”)] UNSPEC_SVE_COMPACT))] “TARGET_SVE” “compact\t%0., %1, %2.” )

;; Duplicate one element of a vector. (define_insn “@aarch64_sve_dup_lane” [(set (match_operand:SVE_ALL 0 “register_operand” “=w”) (vec_duplicate:SVE_ALL (vec_select: (match_operand:SVE_ALL 1 “register_operand” “w”) (parallel [(match_operand:SI 2 “const_int_operand”)]))))] “TARGET_SVE && IN_RANGE (INTVAL (operands[2]) * <container_bits> / 8, 0, 63)” “dup\t%0., %1.[%2]” )

;; Use DUP.Q to duplicate a 128-bit segment of a register. ;; ;; The vec_select: sets memory lane number N of the V128 to lane ;; number op2 + N of op1. (We don't need to distinguish between memory ;; and architectural register lane numbering for op1 or op0, since the ;; two numbering schemes are the same for SVE.) ;; ;; The vec_duplicate:SVE_FULL then copies memory lane number N of the ;; V128 (and thus lane number op2 + N of op1) to lane numbers N + I * STEP ;; of op0. We therefore get the correct result for both endiannesses. ;; ;; The wrinkle is that for big-endian V128 registers, memory lane numbering ;; is in the opposite order to architectural register lane numbering. ;; Thus if we were to do this operation via a V128 temporary register, ;; the vec_select and vec_duplicate would both involve a reverse operation ;; for big-endian targets. In this fused pattern the two reverses cancel ;; each other out. (define_insn “@aarch64_sve_dupq_lane” [(set (match_operand:SVE_FULL 0 “register_operand” “=w”) (vec_duplicate:SVE_FULL (vec_select: (match_operand:SVE_FULL 1 “register_operand” “w”) (match_operand 2 “ascending_int_parallel”))))] “TARGET_SVE && (INTVAL (XVECEXP (operands[2], 0, 0)) * GET_MODE_SIZE (mode)) % 16 == 0 && IN_RANGE (INTVAL (XVECEXP (operands[2], 0, 0)) * GET_MODE_SIZE (mode), 0, 63)” { unsigned int byte = (INTVAL (XVECEXP (operands[2], 0, 0)) * GET_MODE_SIZE (mode)); operands[2] = gen_int_mode (byte / 16, DImode); return “dup\t%0.q, %1.q[%2]”; } )

;; Reverse the order of elements within a full vector. (define_insn “@aarch64_sve_rev” [(set (match_operand:SVE_ALL 0 “register_operand” “=w”) (unspec:SVE_ALL [(match_operand:SVE_ALL 1 “register_operand” “w”)] UNSPEC_REV))] “TARGET_SVE” “rev\t%0., %1.”)

;; ------------------------------------------------------------------------- ;; ---- [INT,FP] Special-purpose binary permutes ;; ------------------------------------------------------------------------- ;; Includes: ;; - EXT ;; - SPLICE ;; - TRN1 ;; - TRN2 ;; - UZP1 ;; - UZP2 ;; - ZIP1 ;; - ZIP2 ;; -------------------------------------------------------------------------

;; Like EXT, but start at the first active element. (define_insn “@aarch64_sve_splice” [(set (match_operand:SVE_FULL 0 “register_operand” “=w, ?&w”) (unspec:SVE_FULL [(match_operand: 1 “register_operand” “Upl, Upl”) (match_operand:SVE_FULL 2 “register_operand” “0, w”) (match_operand:SVE_FULL 3 “register_operand” “w, w”)] UNSPEC_SVE_SPLICE))] “TARGET_SVE” “@ splice\t%0., %1, %0., %3. movprfx\t%0, %2;splice\t%0., %1, %0., %3.” [(set_attr “movprfx” “*, yes”)] )

;; Permutes that take half the elements from one vector and half the ;; elements from the other. (define_insn “@aarch64_sve_<perm_insn>” [(set (match_operand:SVE_ALL 0 “register_operand” “=w”) (unspec:SVE_ALL [(match_operand:SVE_ALL 1 “register_operand” “w”) (match_operand:SVE_ALL 2 “register_operand” “w”)] PERMUTE))] “TARGET_SVE” “<perm_insn>\t%0., %1., %2.” )

;; Apply PERMUTE to 128-bit sequences. The behavior of these patterns ;; doesn't depend on the mode. (define_insn “@aarch64_sve_” [(set (match_operand:SVE_FULL 0 “register_operand” “=w”) (unspec:SVE_FULL [(match_operand:SVE_FULL 1 “register_operand” “w”) (match_operand:SVE_FULL 2 “register_operand” “w”)] PERMUTEQ))] “TARGET_SVE_F64MM” “<perm_insn>\t%0.q, %1.q, %2.q” )

;; Concatenate two vectors and extract a subvector. Note that the ;; immediate (third) operand is the lane index not the byte index. (define_insn “@aarch64_sve_ext” [(set (match_operand:SVE_ALL 0 “register_operand” “=w, ?&w”) (unspec:SVE_ALL [(match_operand:SVE_ALL 1 “register_operand” “0, w”) (match_operand:SVE_ALL 2 “register_operand” “w, w”) (match_operand:SI 3 “const_int_operand”)] UNSPEC_EXT))] “TARGET_SVE && IN_RANGE (INTVAL (operands[3]) * <container_bits> / 8, 0, 255)” { operands[3] = GEN_INT (INTVAL (operands[3]) * <container_bits> / 8); return (which_alternative == 0 ? “ext\t%0.b, %0.b, %2.b, #%3” : “movprfx\t%0, %1;ext\t%0.b, %0.b, %2.b, #%3”); } [(set_attr “movprfx” “*,yes”)] )

;; ------------------------------------------------------------------------- ;; ---- [PRED] Special-purpose unary permutes ;; ------------------------------------------------------------------------- ;; Includes: ;; - REV ;; -------------------------------------------------------------------------

(define_insn “@aarch64_sve_rev” [(set (match_operand:PRED_ALL 0 “register_operand” “=Upa”) (unspec:PRED_ALL [(match_operand:PRED_ALL 1 “register_operand” “Upa”)] UNSPEC_REV))] “TARGET_SVE” “rev\t%0., %1.”)

;; ------------------------------------------------------------------------- ;; ---- [PRED] Special-purpose binary permutes ;; ------------------------------------------------------------------------- ;; Includes: ;; - TRN1 ;; - TRN2 ;; - UZP1 ;; - UZP2 ;; - ZIP1 ;; - ZIP2 ;; -------------------------------------------------------------------------

;; Permutes that take half the elements from one vector and half the ;; elements from the other. (define_insn “@aarch64_sve_<perm_insn>” [(set (match_operand:PRED_ALL 0 “register_operand” “=Upa”) (unspec:PRED_ALL [(match_operand:PRED_ALL 1 “register_operand” “Upa”) (match_operand:PRED_ALL 2 “register_operand” “Upa”)] PERMUTE))] “TARGET_SVE” “<perm_insn>\t%0., %1., %2.” )

;; Special purpose permute used by the predicate generation instructions. ;; Unlike the normal permute patterns, these instructions operate on VNx16BI ;; regardless of the element size, so that all input and output bits are ;; well-defined. Operand 3 then indicates the size of the permute. (define_insn “@aarch64_sve_trn1_conv” [(set (match_operand:VNx16BI 0 “register_operand” “=Upa”) (unspec:VNx16BI [(match_operand:VNx16BI 1 “register_operand” “Upa”) (match_operand:VNx16BI 2 “register_operand” “Upa”) (match_operand:PRED_ALL 3 “aarch64_simd_imm_zero”)] UNSPEC_TRN1_CONV))] “TARGET_SVE” “trn1\t%0.<PRED_ALL:Vetype>, %1.<PRED_ALL:Vetype>, %2.<PRED_ALL:Vetype>” )

;; ========================================================================= ;; == Conversions ;; =========================================================================

;; ------------------------------------------------------------------------- ;; ---- [INT<-INT] Packs ;; ------------------------------------------------------------------------- ;; Includes: ;; - UZP1 ;; -------------------------------------------------------------------------

;; Integer pack. Use UZP1 on the narrower type, which discards ;; the high part of each wide element. (define_insn “vec_pack_trunc_” [(set (match_operand:SVE_FULL_BHSI 0 “register_operand” “=w”) (unspec:SVE_FULL_BHSI [(match_operand: 1 “register_operand” “w”) (match_operand: 2 “register_operand” “w”)] UNSPEC_PACK))] “TARGET_SVE” “uzp1\t%0., %1., %2.” )

;; ------------------------------------------------------------------------- ;; ---- [INT<-INT] Unpacks ;; ------------------------------------------------------------------------- ;; Includes: ;; - SUNPKHI ;; - SUNPKLO ;; - UUNPKHI ;; - UUNPKLO ;; -------------------------------------------------------------------------

;; Unpack the low or high half of a vector, where “high” refers to ;; the low-numbered lanes for big-endian and the high-numbered lanes ;; for little-endian. (define_expand “vec_unpack<perm_hilo><SVE_FULL_BHSI:mode>” [(match_operand: 0 “register_operand”) (unspec: [(match_operand:SVE_FULL_BHSI 1 “register_operand”)] UNPACK)] “TARGET_SVE” { emit_insn ((<hi_lanes_optab> ? gen_aarch64_sve_unpkhi_<SVE_FULL_BHSI:mode> : gen_aarch64_sve_unpklo_<SVE_FULL_BHSI:mode>) (operands[0], operands[1])); DONE; } )

(define_insn “@aarch64_sve_unpk<perm_hilo>_<SVE_FULL_BHSI:mode>” [(set (match_operand: 0 “register_operand” “=w”) (unspec: [(match_operand:SVE_FULL_BHSI 1 “register_operand” “w”)] UNPACK))] “TARGET_SVE” “unpk<perm_hilo>\t%0., %1.” )

;; ------------------------------------------------------------------------- ;; ---- [INT<-FP] Conversions ;; ------------------------------------------------------------------------- ;; Includes: ;; - FCVTZS ;; - FCVTZU ;; -------------------------------------------------------------------------

;; Unpredicated conversion of floats to integers of the same size (HF to HI, ;; SF to SI or DF to DI). (define_expand “<v_int_equiv>2” [(set (match_operand:<V_INT_EQUIV> 0 “register_operand”) (unspec:<V_INT_EQUIV> [(match_dup 2) (const_int SVE_RELAXED_GP) (match_operand:SVE_FULL_F 1 “register_operand”)] SVE_COND_FCVTI))] “TARGET_SVE” { operands[2] = aarch64_ptrue_reg (mode); } )

;; Predicated float-to-integer conversion, either to the same width or wider. (define_insn “@aarch64_sve__nontrunc<SVE_FULL_F:mode><SVE_FULL_HSDI:mode>” [(set (match_operand:SVE_FULL_HSDI 0 “register_operand” “=w, ?&w”) (unspec:SVE_FULL_HSDI [(match_operand:<SVE_FULL_HSDI:VPRED> 1 “register_operand” “Upl, Upl”) (match_operand:SI 3 “aarch64_sve_gp_strictness”) (match_operand:SVE_FULL_F 2 “register_operand” “0, w”)] SVE_COND_FCVTI))] “TARGET_SVE && <SVE_FULL_HSDI:elem_bits> >= <SVE_FULL_F:elem_bits>” “@ fcvtz\t%0.<SVE_FULL_HSDI:Vetype>, %1/m, %2.<SVE_FULL_F:Vetype> movprfx\t%0, %2;fcvtz\t%0.<SVE_FULL_HSDI:Vetype>, %1/m, %2.<SVE_FULL_F:Vetype>” [(set_attr “movprfx” “*,yes”)] )

;; Predicated narrowing float-to-integer conversion. (define_insn “@aarch64_sve__trunc<VNx2DF_ONLY:mode><VNx4SI_ONLY:mode>” [(set (match_operand:VNx4SI_ONLY 0 “register_operand” “=w, ?&w”) (unspec:VNx4SI_ONLY [(match_operand:VNx2BI 1 “register_operand” “Upl, Upl”) (match_operand:SI 3 “aarch64_sve_gp_strictness”) (match_operand:VNx2DF_ONLY 2 “register_operand” “0, w”)] SVE_COND_FCVTI))] “TARGET_SVE” “@ fcvtz\t%0.<VNx4SI_ONLY:Vetype>, %1/m, %2.<VNx2DF_ONLY:Vetype> movprfx\t%0, %2;fcvtz\t%0.<VNx4SI_ONLY:Vetype>, %1/m, %2.<VNx2DF_ONLY:Vetype>” [(set_attr “movprfx” “*,yes”)] )

;; Predicated float-to-integer conversion with merging, either to the same ;; width or wider. (define_expand “@cond__nontrunc<SVE_FULL_F:mode><SVE_FULL_HSDI:mode>” [(set (match_operand:SVE_FULL_HSDI 0 “register_operand”) (unspec:SVE_FULL_HSDI [(match_operand:<SVE_FULL_HSDI:VPRED> 1 “register_operand”) (unspec:SVE_FULL_HSDI [(match_dup 1) (const_int SVE_STRICT_GP) (match_operand:SVE_FULL_F 2 “register_operand”)] SVE_COND_FCVTI) (match_operand:SVE_FULL_HSDI 3 “aarch64_simd_reg_or_zero”)] UNSPEC_SEL))] “TARGET_SVE && <SVE_FULL_HSDI:elem_bits> >= <SVE_FULL_F:elem_bits>” )

;; The first alternative doesn't need the earlyclobber, but the only case ;; it would help is the uninteresting one in which operands 2 and 3 are ;; the same register (despite having different modes). Making all the ;; alternatives earlyclobber makes things more consistent for the ;; register allocator. (define_insn_and_rewrite “*cond__nontrunc<SVE_FULL_F:mode><SVE_FULL_HSDI:mode>_relaxed” [(set (match_operand:SVE_FULL_HSDI 0 “register_operand” “=&w, &w, ?&w”) (unspec:SVE_FULL_HSDI [(match_operand:<SVE_FULL_HSDI:VPRED> 1 “register_operand” “Upl, Upl, Upl”) (unspec:SVE_FULL_HSDI [(match_operand 4) (const_int SVE_RELAXED_GP) (match_operand:SVE_FULL_F 2 “register_operand” “w, w, w”)] SVE_COND_FCVTI) (match_operand:SVE_FULL_HSDI 3 “aarch64_simd_reg_or_zero” “0, Dz, w”)] UNSPEC_SEL))] “TARGET_SVE && <SVE_FULL_HSDI:elem_bits> >= <SVE_FULL_F:elem_bits>” “@ fcvtz\t%0.<SVE_FULL_HSDI:Vetype>, %1/m, %2.<SVE_FULL_F:Vetype> movprfx\t%0.<SVE_FULL_HSDI:Vetype>, %1/z, %2.<SVE_FULL_HSDI:Vetype>;fcvtz\t%0.<SVE_FULL_HSDI:Vetype>, %1/m, %2.<SVE_FULL_F:Vetype> movprfx\t%0, %3;fcvtz\t%0.<SVE_FULL_HSDI:Vetype>, %1/m, %2.<SVE_FULL_F:Vetype>” “&& !rtx_equal_p (operands[1], operands[4])” { operands[4] = copy_rtx (operands[1]); } [(set_attr “movprfx” “*,yes,yes”)] )

(define_insn “*cond__nontrunc<SVE_FULL_F:mode><SVE_FULL_HSDI:mode>_strict” [(set (match_operand:SVE_FULL_HSDI 0 “register_operand” “=&w, &w, ?&w”) (unspec:SVE_FULL_HSDI [(match_operand:<SVE_FULL_HSDI:VPRED> 1 “register_operand” “Upl, Upl, Upl”) (unspec:SVE_FULL_HSDI [(match_dup 1) (const_int SVE_STRICT_GP) (match_operand:SVE_FULL_F 2 “register_operand” “w, w, w”)] SVE_COND_FCVTI) (match_operand:SVE_FULL_HSDI 3 “aarch64_simd_reg_or_zero” “0, Dz, w”)] UNSPEC_SEL))] “TARGET_SVE && <SVE_FULL_HSDI:elem_bits> >= <SVE_FULL_F:elem_bits>” “@ fcvtz\t%0.<SVE_FULL_HSDI:Vetype>, %1/m, %2.<SVE_FULL_F:Vetype> movprfx\t%0.<SVE_FULL_HSDI:Vetype>, %1/z, %2.<SVE_FULL_HSDI:Vetype>;fcvtz\t%0.<SVE_FULL_HSDI:Vetype>, %1/m, %2.<SVE_FULL_F:Vetype> movprfx\t%0, %3;fcvtz\t%0.<SVE_FULL_HSDI:Vetype>, %1/m, %2.<SVE_FULL_F:Vetype>” [(set_attr “movprfx” “*,yes,yes”)] )

;; Predicated narrowing float-to-integer conversion with merging. (define_expand “@cond__trunc<VNx2DF_ONLY:mode><VNx4SI_ONLY:mode>” [(set (match_operand:VNx4SI_ONLY 0 “register_operand”) (unspec:VNx4SI_ONLY [(match_operand:VNx2BI 1 “register_operand”) (unspec:VNx4SI_ONLY [(match_dup 1) (const_int SVE_STRICT_GP) (match_operand:VNx2DF_ONLY 2 “register_operand”)] SVE_COND_FCVTI) (match_operand:VNx4SI_ONLY 3 “aarch64_simd_reg_or_zero”)] UNSPEC_SEL))] “TARGET_SVE” )

(define_insn “*cond__trunc<VNx2DF_ONLY:mode><VNx4SI_ONLY:mode>” [(set (match_operand:VNx4SI_ONLY 0 “register_operand” “=&w, &w, ?&w”) (unspec:VNx4SI_ONLY [(match_operand:VNx2BI 1 “register_operand” “Upl, Upl, Upl”) (unspec:VNx4SI_ONLY [(match_dup 1) (match_operand:SI 4 “aarch64_sve_gp_strictness”) (match_operand:VNx2DF_ONLY 2 “register_operand” “w, w, w”)] SVE_COND_FCVTI) (match_operand:VNx4SI_ONLY 3 “aarch64_simd_reg_or_zero” “0, Dz, w”)] UNSPEC_SEL))] “TARGET_SVE” “@ fcvtz\t%0.<VNx4SI_ONLY:Vetype>, %1/m, %2.<VNx2DF_ONLY:Vetype> movprfx\t%0.<VNx2DF_ONLY:Vetype>, %1/z, %2.<VNx2DF_ONLY:Vetype>;fcvtz\t%0.<VNx4SI_ONLY:Vetype>, %1/m, %2.<VNx2DF_ONLY:Vetype> movprfx\t%0, %3;fcvtz\t%0.<VNx4SI_ONLY:Vetype>, %1/m, %2.<VNx2DF_ONLY:Vetype>” [(set_attr “movprfx” “*,yes,yes”)] )

;; ------------------------------------------------------------------------- ;; ---- [INT<-FP] Packs ;; ------------------------------------------------------------------------- ;; The patterns in this section are synthetic. ;; -------------------------------------------------------------------------

;; Convert two vectors of DF to SI and pack the results into a single vector. (define_expand “vec_pack_fix_trunc_vnx2df” [(set (match_dup 4) (unspec:VNx4SI [(match_dup 3) (const_int SVE_RELAXED_GP) (match_operand:VNx2DF 1 “register_operand”)] SVE_COND_FCVTI)) (set (match_dup 5) (unspec:VNx4SI [(match_dup 3) (const_int SVE_RELAXED_GP) (match_operand:VNx2DF 2 “register_operand”)] SVE_COND_FCVTI)) (set (match_operand:VNx4SI 0 “register_operand”) (unspec:VNx4SI [(match_dup 4) (match_dup 5)] UNSPEC_UZP1))] “TARGET_SVE” { operands[3] = aarch64_ptrue_reg (VNx2BImode); operands[4] = gen_reg_rtx (VNx4SImode); operands[5] = gen_reg_rtx (VNx4SImode); } )

;; ------------------------------------------------------------------------- ;; ---- [INT<-FP] Unpacks ;; ------------------------------------------------------------------------- ;; No patterns here yet! ;; -------------------------------------------------------------------------

;; ------------------------------------------------------------------------- ;; ---- [FP<-INT] Conversions ;; ------------------------------------------------------------------------- ;; Includes: ;; - SCVTF ;; - UCVTF ;; -------------------------------------------------------------------------

;; Unpredicated conversion of integers to floats of the same size ;; (HI to HF, SI to SF or DI to DF). (define_expand “<v_int_equiv>2” [(set (match_operand:SVE_FULL_F 0 “register_operand”) (unspec:SVE_FULL_F [(match_dup 2) (const_int SVE_RELAXED_GP) (match_operand:<V_INT_EQUIV> 1 “register_operand”)] SVE_COND_ICVTF))] “TARGET_SVE” { operands[2] = aarch64_ptrue_reg (mode); } )

;; Predicated integer-to-float conversion, either to the same width or ;; narrower. (define_insn “@aarch64_sve__nonextend<SVE_FULL_HSDI:mode><SVE_FULL_F:mode>” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=w, ?&w”) (unspec:SVE_FULL_F [(match_operand:<SVE_FULL_HSDI:VPRED> 1 “register_operand” “Upl, Upl”) (match_operand:SI 3 “aarch64_sve_gp_strictness”) (match_operand:SVE_FULL_HSDI 2 “register_operand” “0, w”)] SVE_COND_ICVTF))] “TARGET_SVE && <SVE_FULL_HSDI:elem_bits> >= <SVE_FULL_F:elem_bits>” “@ cvtf\t%0.<SVE_FULL_F:Vetype>, %1/m, %2.<SVE_FULL_HSDI:Vetype> movprfx\t%0, %2;cvtf\t%0.<SVE_FULL_F:Vetype>, %1/m, %2.<SVE_FULL_HSDI:Vetype>” [(set_attr “movprfx” “*,yes”)] )

;; Predicated widening integer-to-float conversion. (define_insn “@aarch64_sve__extend<VNx4SI_ONLY:mode><VNx2DF_ONLY:mode>” [(set (match_operand:VNx2DF_ONLY 0 “register_operand” “=w, ?&w”) (unspec:VNx2DF_ONLY [(match_operand:VNx2BI 1 “register_operand” “Upl, Upl”) (match_operand:SI 3 “aarch64_sve_gp_strictness”) (match_operand:VNx4SI_ONLY 2 “register_operand” “0, w”)] SVE_COND_ICVTF))] “TARGET_SVE” “@ cvtf\t%0.<VNx2DF_ONLY:Vetype>, %1/m, %2.<VNx4SI_ONLY:Vetype> movprfx\t%0, %2;cvtf\t%0.<VNx2DF_ONLY:Vetype>, %1/m, %2.<VNx4SI_ONLY:Vetype>” [(set_attr “movprfx” “*,yes”)] )

;; Predicated integer-to-float conversion with merging, either to the same ;; width or narrower. (define_expand “@cond__nonextend<SVE_FULL_HSDI:mode><SVE_FULL_F:mode>” [(set (match_operand:SVE_FULL_F 0 “register_operand”) (unspec:SVE_FULL_F [(match_operand:<SVE_FULL_HSDI:VPRED> 1 “register_operand”) (unspec:SVE_FULL_F [(match_dup 1) (const_int SVE_STRICT_GP) (match_operand:SVE_FULL_HSDI 2 “register_operand”)] SVE_COND_ICVTF) (match_operand:SVE_FULL_F 3 “aarch64_simd_reg_or_zero”)] UNSPEC_SEL))] “TARGET_SVE && <SVE_FULL_HSDI:elem_bits> >= <SVE_FULL_F:elem_bits>” )

;; The first alternative doesn't need the earlyclobber, but the only case ;; it would help is the uninteresting one in which operands 2 and 3 are ;; the same register (despite having different modes). Making all the ;; alternatives earlyclobber makes things more consistent for the ;; register allocator. (define_insn_and_rewrite “*cond__nonextend<SVE_FULL_HSDI:mode><SVE_FULL_F:mode>_relaxed” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=&w, &w, ?&w”) (unspec:SVE_FULL_F [(match_operand:<SVE_FULL_HSDI:VPRED> 1 “register_operand” “Upl, Upl, Upl”) (unspec:SVE_FULL_F [(match_operand 4) (const_int SVE_RELAXED_GP) (match_operand:SVE_FULL_HSDI 2 “register_operand” “w, w, w”)] SVE_COND_ICVTF) (match_operand:SVE_FULL_F 3 “aarch64_simd_reg_or_zero” “0, Dz, w”)] UNSPEC_SEL))] “TARGET_SVE && <SVE_FULL_HSDI:elem_bits> >= <SVE_FULL_F:elem_bits>” “@ cvtf\t%0.<SVE_FULL_F:Vetype>, %1/m, %2.<SVE_FULL_HSDI:Vetype> movprfx\t%0.<SVE_FULL_HSDI:Vetype>, %1/z, %2.<SVE_FULL_HSDI:Vetype>;cvtf\t%0.<SVE_FULL_F:Vetype>, %1/m, %2.<SVE_FULL_HSDI:Vetype> movprfx\t%0, %3;cvtf\t%0.<SVE_FULL_F:Vetype>, %1/m, %2.<SVE_FULL_HSDI:Vetype>” “&& !rtx_equal_p (operands[1], operands[4])” { operands[4] = copy_rtx (operands[1]); } [(set_attr “movprfx” “*,yes,yes”)] )

(define_insn “*cond__nonextend<SVE_FULL_HSDI:mode><SVE_FULL_F:mode>_strict” [(set (match_operand:SVE_FULL_F 0 “register_operand” “=&w, &w, ?&w”) (unspec:SVE_FULL_F [(match_operand:<SVE_FULL_HSDI:VPRED> 1 “register_operand” “Upl, Upl, Upl”) (unspec:SVE_FULL_F [(match_dup 1) (const_int SVE_STRICT_GP) (match_operand:SVE_FULL_HSDI 2 “register_operand” “w, w, w”)] SVE_COND_ICVTF) (match_operand:SVE_FULL_F 3 “aarch64_simd_reg_or_zero” “0, Dz, w”)] UNSPEC_SEL))] “TARGET_SVE && <SVE_FULL_HSDI:elem_bits> >= <SVE_FULL_F:elem_bits>” “@ cvtf\t%0.<SVE_FULL_F:Vetype>, %1/m, %2.<SVE_FULL_HSDI:Vetype> movprfx\t%0.<SVE_FULL_HSDI:Vetype>, %1/z, %2.<SVE_FULL_HSDI:Vetype>;cvtf\t%0.<SVE_FULL_F:Vetype>, %1/m, %2.<SVE_FULL_HSDI:Vetype> movprfx\t%0, %3;cvtf\t%0.<SVE_FULL_F:Vetype>, %1/m, %2.<SVE_FULL_HSDI:Vetype>” [(set_attr “movprfx” “*,yes,yes”)] )

;; Predicated widening integer-to-float conversion with merging. (define_expand “@cond__extend<VNx4SI_ONLY:mode><VNx2DF_ONLY:mode>” [(set (match_operand:VNx2DF_ONLY 0 “register_operand”) (unspec:VNx2DF_ONLY [(match_operand:VNx2BI 1 “register_operand”) (unspec:VNx2DF_ONLY [(match_dup 1) (const_int SVE_STRICT_GP) (match_operand:VNx4SI_ONLY 2 “register_operand”)] SVE_COND_ICVTF) (match_operand:VNx2DF_ONLY 3 “aarch64_simd_reg_or_zero”)] UNSPEC_SEL))] “TARGET_SVE” )

(define_insn “*cond__extend<VNx4SI_ONLY:mode><VNx2DF_ONLY:mode>” [(set (match_operand:VNx2DF_ONLY 0 “register_operand” “=w, ?&w, ?&w”) (unspec:VNx2DF_ONLY [(match_operand:VNx2BI 1 “register_operand” “Upl, Upl, Upl”) (unspec:VNx2DF_ONLY [(match_dup 1) (match_operand:SI 4 “aarch64_sve_gp_strictness”) (match_operand:VNx4SI_ONLY 2 “register_operand” “w, w, w”)] SVE_COND_ICVTF) (match_operand:VNx2DF_ONLY 3 “aarch64_simd_reg_or_zero” “0, Dz, w”)] UNSPEC_SEL))] “TARGET_SVE” “@ cvtf\t%0.<VNx2DF_ONLY:Vetype>, %1/m, %2.<VNx4SI_ONLY:Vetype> movprfx\t%0.<VNx2DF_ONLY:Vetype>, %1/z, %2.<VNx2DF_ONLY:Vetype>;cvtf\t%0.<VNx2DF_ONLY:Vetype>, %1/m, %2.<VNx4SI_ONLY:Vetype> movprfx\t%0, %3;cvtf\t%0.<VNx2DF_ONLY:Vetype>, %1/m, %2.<VNx4SI_ONLY:Vetype>” [(set_attr “movprfx” “*,yes,yes”)] )

;; ------------------------------------------------------------------------- ;; ---- [FP<-INT] Packs ;; ------------------------------------------------------------------------- ;; No patterns here yet! ;; -------------------------------------------------------------------------

;; ------------------------------------------------------------------------- ;; ---- [FP<-INT] Unpacks ;; ------------------------------------------------------------------------- ;; The patterns in this section are synthetic. ;; -------------------------------------------------------------------------

;; Unpack one half of a VNx4SI to VNx2DF. First unpack from VNx4SI ;; to VNx2DI, reinterpret the VNx2DI as a VNx4SI, then convert the ;; unpacked VNx4SI to VNx2DF. (define_expand “vec_unpack<su_optab>float<perm_hilo>_vnx4si” [(match_operand:VNx2DF 0 “register_operand”) (FLOATUORS:VNx2DF (unspec:VNx2DI [(match_operand:VNx4SI 1 “register_operand”)] UNPACK_UNSIGNED))] “TARGET_SVE” { /* Use ZIP to do the unpack, since we don't care about the upper halves and since it has the nice property of not needing any subregs. If using UUNPK* turns out to be preferable, we could model it as a ZIP whose first operand is zero. */ rtx temp = gen_reg_rtx (VNx4SImode); emit_insn ((<hi_lanes_optab> ? gen_aarch64_sve_zip2vnx4si : gen_aarch64_sve_zip1vnx4si) (temp, operands[1], operands[1])); rtx ptrue = aarch64_ptrue_reg (VNx2BImode); rtx strictness = gen_int_mode (SVE_RELAXED_GP, SImode); emit_insn (gen_aarch64_sve_FLOATUORS:optab_extendvnx4sivnx2df (operands[0], ptrue, temp, strictness)); DONE; } )

;; ------------------------------------------------------------------------- ;; ---- [FP<-FP] Packs ;; ------------------------------------------------------------------------- ;; Includes: ;; - FCVT ;; -------------------------------------------------------------------------

;; Convert two vectors of DF to SF, or two vectors of SF to HF, and pack ;; the results into a single vector. (define_expand “vec_pack_trunc_” [(set (match_dup 4) (unspec:SVE_FULL_HSF [(match_dup 3) (const_int SVE_RELAXED_GP) (match_operand: 1 “register_operand”)] UNSPEC_COND_FCVT)) (set (match_dup 5) (unspec:SVE_FULL_HSF [(match_dup 3) (const_int SVE_RELAXED_GP) (match_operand: 2 “register_operand”)] UNSPEC_COND_FCVT)) (set (match_operand:SVE_FULL_HSF 0 “register_operand”) (unspec:SVE_FULL_HSF [(match_dup 4) (match_dup 5)] UNSPEC_UZP1))] “TARGET_SVE” { operands[3] = aarch64_ptrue_reg (<VWIDE_PRED>mode); operands[4] = gen_reg_rtx (mode); operands[5] = gen_reg_rtx (mode); } )

;; Predicated float-to-float truncation. (define_insn “@aarch64_sve__trunc<SVE_FULL_SDF:mode><SVE_FULL_HSF:mode>” [(set (match_operand:SVE_FULL_HSF 0 “register_operand” “=w, ?&w”) (unspec:SVE_FULL_HSF [(match_operand:<SVE_FULL_SDF:VPRED> 1 “register_operand” “Upl, Upl”) (match_operand:SI 3 “aarch64_sve_gp_strictness”) (match_operand:SVE_FULL_SDF 2 “register_operand” “0, w”)] SVE_COND_FCVT))] “TARGET_SVE && <SVE_FULL_SDF:elem_bits> > <SVE_FULL_HSF:elem_bits>” “@ fcvt\t%0.<SVE_FULL_HSF:Vetype>, %1/m, %2.<SVE_FULL_SDF:Vetype> movprfx\t%0, %2;fcvt\t%0.<SVE_FULL_HSF:Vetype>, %1/m, %2.<SVE_FULL_SDF:Vetype>” [(set_attr “movprfx” “*,yes”)] )

;; Predicated float-to-float truncation with merging. (define_expand “@cond__trunc<SVE_FULL_SDF:mode><SVE_FULL_HSF:mode>” [(set (match_operand:SVE_FULL_HSF 0 “register_operand”) (unspec:SVE_FULL_HSF [(match_operand:<SVE_FULL_SDF:VPRED> 1 “register_operand”) (unspec:SVE_FULL_HSF [(match_dup 1) (const_int SVE_STRICT_GP) (match_operand:SVE_FULL_SDF 2 “register_operand”)] SVE_COND_FCVT) (match_operand:SVE_FULL_HSF 3 “aarch64_simd_reg_or_zero”)] UNSPEC_SEL))] “TARGET_SVE && <SVE_FULL_SDF:elem_bits> > <SVE_FULL_HSF:elem_bits>” )

(define_insn “*cond__trunc<SVE_FULL_SDF:mode><SVE_FULL_HSF:mode>” [(set (match_operand:SVE_FULL_HSF 0 “register_operand” “=w, ?&w, ?&w”) (unspec:SVE_FULL_HSF [(match_operand:<SVE_FULL_SDF:VPRED> 1 “register_operand” “Upl, Upl, Upl”) (unspec:SVE_FULL_HSF [(match_dup 1) (match_operand:SI 4 “aarch64_sve_gp_strictness”) (match_operand:SVE_FULL_SDF 2 “register_operand” “w, w, w”)] SVE_COND_FCVT) (match_operand:SVE_FULL_HSF 3 “aarch64_simd_reg_or_zero” “0, Dz, w”)] UNSPEC_SEL))] “TARGET_SVE && <SVE_FULL_SDF:elem_bits> > <SVE_FULL_HSF:elem_bits>” “@ fcvt\t%0.<SVE_FULL_HSF:Vetype>, %1/m, %2.<SVE_FULL_SDF:Vetype> movprfx\t%0.<SVE_FULL_SDF:Vetype>, %1/z, %2.<SVE_FULL_SDF:Vetype>;fcvt\t%0.<SVE_FULL_HSF:Vetype>, %1/m, %2.<SVE_FULL_SDF:Vetype> movprfx\t%0, %3;fcvt\t%0.<SVE_FULL_HSF:Vetype>, %1/m, %2.<SVE_FULL_SDF:Vetype>” [(set_attr “movprfx” “*,yes,yes”)] )

;; ------------------------------------------------------------------------- ;; ---- [FP<-FP] Packs (bfloat16) ;; ------------------------------------------------------------------------- ;; Includes: ;; - BFCVT (BF16) ;; - BFCVTNT (BF16) ;; -------------------------------------------------------------------------

;; Predicated BFCVT. (define_insn “@aarch64_sve__trunc<VNx4SF_ONLY:mode><VNx8BF_ONLY:mode>” [(set (match_operand:VNx8BF_ONLY 0 “register_operand” “=w, ?&w”) (unspec:VNx8BF_ONLY [(match_operand:VNx4BI 1 “register_operand” “Upl, Upl”) (match_operand:SI 3 “aarch64_sve_gp_strictness”) (match_operand:VNx4SF_ONLY 2 “register_operand” “0, w”)] SVE_COND_FCVT))] “TARGET_SVE_BF16” “@ bfcvt\t%0.h, %1/m, %2.s movprfx\t%0, %2;bfcvt\t%0.h, %1/m, %2.s” [(set_attr “movprfx” “*,yes”)] )

;; Predicated BFCVT with merging. (define_expand “@cond__trunc<VNx4SF_ONLY:mode><VNx8BF_ONLY:mode>” [(set (match_operand:VNx8BF_ONLY 0 “register_operand”) (unspec:VNx8BF_ONLY [(match_operand:VNx4BI 1 “register_operand”) (unspec:VNx8BF_ONLY [(match_dup 1) (const_int SVE_STRICT_GP) (match_operand:VNx4SF_ONLY 2 “register_operand”)] SVE_COND_FCVT) (match_operand:VNx8BF_ONLY 3 “aarch64_simd_reg_or_zero”)] UNSPEC_SEL))] “TARGET_SVE_BF16” )

(define_insn “*cond__trunc<VNx4SF_ONLY:mode><VNx8BF_ONLY:mode>” [(set (match_operand:VNx8BF_ONLY 0 “register_operand” “=w, ?&w, ?&w”) (unspec:VNx8BF_ONLY [(match_operand:VNx4BI 1 “register_operand” “Upl, Upl, Upl”) (unspec:VNx8BF_ONLY [(match_dup 1) (match_operand:SI 4 “aarch64_sve_gp_strictness”) (match_operand:VNx4SF_ONLY 2 “register_operand” “w, w, w”)] SVE_COND_FCVT) (match_operand:VNx8BF_ONLY 3 “aarch64_simd_reg_or_zero” “0, Dz, w”)] UNSPEC_SEL))] “TARGET_SVE_BF16” “@ bfcvt\t%0.h, %1/m, %2.s movprfx\t%0.s, %1/z, %2.s;bfcvt\t%0.h, %1/m, %2.s movprfx\t%0, %3;bfcvt\t%0.h, %1/m, %2.s” [(set_attr “movprfx” “*,yes,yes”)] )

;; Predicated BFCVTNT. This doesn‘t give a natural aarch64_pred_/cond_ ;; pair because the even elements always have to be supplied for active ;; elements, even if the inactive elements don’t matter. ;; ;; This instructions does not take MOVPRFX. (define_insn “@aarch64_sve_cvtnt” [(set (match_operand:VNx8BF_ONLY 0 “register_operand” “=w”) (unspec:VNx8BF_ONLY [(match_operand:VNx4BI 2 “register_operand” “Upl”) (const_int SVE_STRICT_GP) (match_operand:VNx8BF_ONLY 1 “register_operand” “0”) (match_operand:VNx4SF 3 “register_operand” “w”)] UNSPEC_COND_FCVTNT))] “TARGET_SVE_BF16” “bfcvtnt\t%0.h, %2/m, %3.s” )

;; ------------------------------------------------------------------------- ;; ---- [FP<-FP] Unpacks ;; ------------------------------------------------------------------------- ;; Includes: ;; - FCVT ;; -------------------------------------------------------------------------

;; Unpack one half of a VNx4SF to VNx2DF, or one half of a VNx8HF to VNx4SF. ;; First unpack the source without conversion, then float-convert the ;; unpacked source. (define_expand “vec_unpacks_<perm_hilo>_” [(match_operand: 0 “register_operand”) (unspec:SVE_FULL_HSF [(match_operand:SVE_FULL_HSF 1 “register_operand”)] UNPACK_UNSIGNED)] “TARGET_SVE” { /* Use ZIP to do the unpack, since we don't care about the upper halves and since it has the nice property of not needing any subregs. If using UUNPK* turns out to be preferable, we could model it as a ZIP whose first operand is zero. */ rtx temp = gen_reg_rtx (mode); emit_insn ((<hi_lanes_optab> ? gen_aarch64_sve_zip2 : gen_aarch64_sve_zip1) (temp, operands[1], operands[1])); rtx ptrue = aarch64_ptrue_reg (<VWIDE_PRED>mode); rtx strictness = gen_int_mode (SVE_RELAXED_GP, SImode); emit_insn (gen_aarch64_sve_fcvt_nontrunc (operands[0], ptrue, temp, strictness)); DONE; } )

;; Predicated float-to-float extension. (define_insn “@aarch64_sve__nontrunc<SVE_FULL_HSF:mode><SVE_FULL_SDF:mode>” [(set (match_operand:SVE_FULL_SDF 0 “register_operand” “=w, ?&w”) (unspec:SVE_FULL_SDF [(match_operand:<SVE_FULL_SDF:VPRED> 1 “register_operand” “Upl, Upl”) (match_operand:SI 3 “aarch64_sve_gp_strictness”) (match_operand:SVE_FULL_HSF 2 “register_operand” “0, w”)] SVE_COND_FCVT))] “TARGET_SVE && <SVE_FULL_SDF:elem_bits> > <SVE_FULL_HSF:elem_bits>” “@ fcvt\t%0.<SVE_FULL_SDF:Vetype>, %1/m, %2.<SVE_FULL_HSF:Vetype> movprfx\t%0, %2;fcvt\t%0.<SVE_FULL_SDF:Vetype>, %1/m, %2.<SVE_FULL_HSF:Vetype>” [(set_attr “movprfx” “*,yes”)] )

;; Predicated float-to-float extension with merging. (define_expand “@cond__nontrunc<SVE_FULL_HSF:mode><SVE_FULL_SDF:mode>” [(set (match_operand:SVE_FULL_SDF 0 “register_operand”) (unspec:SVE_FULL_SDF [(match_operand:<SVE_FULL_SDF:VPRED> 1 “register_operand”) (unspec:SVE_FULL_SDF [(match_dup 1) (const_int SVE_STRICT_GP) (match_operand:SVE_FULL_HSF 2 “register_operand”)] SVE_COND_FCVT) (match_operand:SVE_FULL_SDF 3 “aarch64_simd_reg_or_zero”)] UNSPEC_SEL))] “TARGET_SVE && <SVE_FULL_SDF:elem_bits> > <SVE_FULL_HSF:elem_bits>” )

(define_insn “*cond__nontrunc<SVE_FULL_HSF:mode><SVE_FULL_SDF:mode>” [(set (match_operand:SVE_FULL_SDF 0 “register_operand” “=w, ?&w, ?&w”) (unspec:SVE_FULL_SDF [(match_operand:<SVE_FULL_SDF:VPRED> 1 “register_operand” “Upl, Upl, Upl”) (unspec:SVE_FULL_SDF [(match_dup 1) (match_operand:SI 4 “aarch64_sve_gp_strictness”) (match_operand:SVE_FULL_HSF 2 “register_operand” “w, w, w”)] SVE_COND_FCVT) (match_operand:SVE_FULL_SDF 3 “aarch64_simd_reg_or_zero” “0, Dz, w”)] UNSPEC_SEL))] “TARGET_SVE && <SVE_FULL_SDF:elem_bits> > <SVE_FULL_HSF:elem_bits>” “@ fcvt\t%0.<SVE_FULL_SDF:Vetype>, %1/m, %2.<SVE_FULL_HSF:Vetype> movprfx\t%0.<SVE_FULL_SDF:Vetype>, %1/z, %2.<SVE_FULL_SDF:Vetype>;fcvt\t%0.<SVE_FULL_SDF:Vetype>, %1/m, %2.<SVE_FULL_HSF:Vetype> movprfx\t%0, %3;fcvt\t%0.<SVE_FULL_SDF:Vetype>, %1/m, %2.<SVE_FULL_HSF:Vetype>” [(set_attr “movprfx” “*,yes,yes”)] )

;; ------------------------------------------------------------------------- ;; ---- [PRED<-PRED] Packs ;; ------------------------------------------------------------------------- ;; Includes: ;; - UZP1 ;; -------------------------------------------------------------------------

;; Predicate pack. Use UZP1 on the narrower type, which discards ;; the high part of each wide element. (define_insn “vec_pack_trunc_” [(set (match_operand:PRED_BHS 0 “register_operand” “=Upa”) (unspec:PRED_BHS [(match_operand: 1 “register_operand” “Upa”) (match_operand: 2 “register_operand” “Upa”)] UNSPEC_PACK))] “TARGET_SVE” “uzp1\t%0., %1., %2.” )

;; ------------------------------------------------------------------------- ;; ---- [PRED<-PRED] Unpacks ;; ------------------------------------------------------------------------- ;; Includes: ;; - PUNPKHI ;; - PUNPKLO ;; -------------------------------------------------------------------------

;; Unpack the low or high half of a predicate, where “high” refers to ;; the low-numbered lanes for big-endian and the high-numbered lanes ;; for little-endian. (define_expand “vec_unpack<perm_hilo>” [(match_operand: 0 “register_operand”) (unspec: [(match_operand:PRED_BHS 1 “register_operand”)] UNPACK)] “TARGET_SVE” { emit_insn ((<hi_lanes_optab> ? gen_aarch64_sve_punpkhi_<PRED_BHS:mode> : gen_aarch64_sve_punpklo_<PRED_BHS:mode>) (operands[0], operands[1])); DONE; } )

(define_insn “@aarch64_sve_punpk<perm_hilo>_” [(set (match_operand: 0 “register_operand” “=Upa”) (unspec: [(match_operand:PRED_BHS 1 “register_operand” “Upa”)] UNPACK_UNSIGNED))] “TARGET_SVE” “punpk<perm_hilo>\t%0.h, %1.b” )

;; ========================================================================= ;; == Vector partitioning ;; =========================================================================

;; ------------------------------------------------------------------------- ;; ---- [PRED] Unary partitioning ;; ------------------------------------------------------------------------- ;; Includes: ;; - BRKA ;; - BRKAS ;; - BRKB ;; - BRKBS ;; -------------------------------------------------------------------------

;; Note that unlike most other instructions that have both merging and ;; zeroing forms, these instructions don‘t operate elementwise and so ;; don’t fit the IFN_COND model. (define_insn “@aarch64_brk<brk_op>” [(set (match_operand:VNx16BI 0 “register_operand” “=Upa, Upa”) (unspec:VNx16BI [(match_operand:VNx16BI 1 “register_operand” “Upa, Upa”) (match_operand:VNx16BI 2 “register_operand” “Upa, Upa”) (match_operand:VNx16BI 3 “aarch64_simd_reg_or_zero” “Dz, 0”)] SVE_BRK_UNARY))] “TARGET_SVE” “@ brk<brk_op>\t%0.b, %1/z, %2.b brk<brk_op>\t%0.b, %1/m, %2.b” )

;; Same, but also producing a flags result. (define_insn “*aarch64_brk<brk_op>_cc” [(set (reg:CC_NZC CC_REGNUM) (unspec:CC_NZC [(match_operand:VNx16BI 1 “register_operand” “Upa, Upa”) (match_dup 1) (match_operand:SI 4 “aarch64_sve_ptrue_flag”) (unspec:VNx16BI [(match_dup 1) (match_operand:VNx16BI 2 “register_operand” “Upa, Upa”) (match_operand:VNx16BI 3 “aarch64_simd_reg_or_zero” “Dz, 0”)] SVE_BRK_UNARY)] UNSPEC_PTEST)) (set (match_operand:VNx16BI 0 “register_operand” “=Upa, Upa”) (unspec:VNx16BI [(match_dup 1) (match_dup 2) (match_dup 3)] SVE_BRK_UNARY))] “TARGET_SVE” “@ brk<brk_op>s\t%0.b, %1/z, %2.b brk<brk_op>s\t%0.b, %1/m, %2.b” )

;; Same, but with only the flags result being interesting. (define_insn “*aarch64_brk<brk_op>_ptest” [(set (reg:CC_NZC CC_REGNUM) (unspec:CC_NZC [(match_operand:VNx16BI 1 “register_operand” “Upa, Upa”) (match_dup 1) (match_operand:SI 4 “aarch64_sve_ptrue_flag”) (unspec:VNx16BI [(match_dup 1) (match_operand:VNx16BI 2 “register_operand” “Upa, Upa”) (match_operand:VNx16BI 3 “aarch64_simd_reg_or_zero” “Dz, 0”)] SVE_BRK_UNARY)] UNSPEC_PTEST)) (clobber (match_scratch:VNx16BI 0 “=Upa, Upa”))] “TARGET_SVE” “@ brk<brk_op>s\t%0.b, %1/z, %2.b brk<brk_op>s\t%0.b, %1/m, %2.b” )

;; ------------------------------------------------------------------------- ;; ---- [PRED] Binary partitioning ;; ------------------------------------------------------------------------- ;; Includes: ;; - BRKN ;; - BRKNS ;; - BRKPA ;; - BRKPAS ;; - BRKPB ;; - BRKPBS ;; -------------------------------------------------------------------------

;; Binary BRKs (BRKN, BRKPA, BRKPB). (define_insn “@aarch64_brk<brk_op>” [(set (match_operand:VNx16BI 0 “register_operand” “=Upa”) (unspec:VNx16BI [(match_operand:VNx16BI 1 “register_operand” “Upa”) (match_operand:VNx16BI 2 “register_operand” “Upa”) (match_operand:VNx16BI 3 “register_operand” “<brk_reg_con>”)] SVE_BRK_BINARY))] “TARGET_SVE” “brk<brk_op>\t%0.b, %1/z, %2.b, %<brk_reg_opno>.b” )

;; Same, but also producing a flags result. (define_insn “*aarch64_brk<brk_op>_cc” [(set (reg:CC_NZC CC_REGNUM) (unspec:CC_NZC [(match_operand:VNx16BI 1 “register_operand” “Upa”) (match_dup 1) (match_operand:SI 4 “aarch64_sve_ptrue_flag”) (unspec:VNx16BI [(match_dup 1) (match_operand:VNx16BI 2 “register_operand” “Upa”) (match_operand:VNx16BI 3 “register_operand” “<brk_reg_con>”)] SVE_BRK_BINARY)] UNSPEC_PTEST)) (set (match_operand:VNx16BI 0 “register_operand” “=Upa”) (unspec:VNx16BI [(match_dup 1) (match_dup 2) (match_dup 3)] SVE_BRK_BINARY))] “TARGET_SVE” “brk<brk_op>s\t%0.b, %1/z, %2.b, %<brk_reg_opno>.b” )

;; Same, but with only the flags result being interesting. (define_insn “*aarch64_brk<brk_op>_ptest” [(set (reg:CC_NZC CC_REGNUM) (unspec:CC_NZC [(match_operand:VNx16BI 1 “register_operand” “Upa”) (match_dup 1) (match_operand:SI 4 “aarch64_sve_ptrue_flag”) (unspec:VNx16BI [(match_dup 1) (match_operand:VNx16BI 2 “register_operand” “Upa”) (match_operand:VNx16BI 3 “register_operand” “<brk_reg_con>”)] SVE_BRK_BINARY)] UNSPEC_PTEST)) (clobber (match_scratch:VNx16BI 0 “=Upa”))] “TARGET_SVE” “brk<brk_op>s\t%0.b, %1/z, %2.b, %<brk_reg_opno>.b” )

;; ------------------------------------------------------------------------- ;; ---- [PRED] Scalarization ;; ------------------------------------------------------------------------- ;; Includes: ;; - PFIRST ;; - PNEXT ;; -------------------------------------------------------------------------

(define_insn “@aarch64_sve_<sve_pred_op>” [(set (match_operand:PRED_ALL 0 “register_operand” “=Upa”) (unspec:PRED_ALL [(match_operand:PRED_ALL 1 “register_operand” “Upa”) (match_operand:SI 2 “aarch64_sve_ptrue_flag”) (match_operand:PRED_ALL 3 “register_operand” “0”)] SVE_PITER)) (clobber (reg:CC_NZC CC_REGNUM))] “TARGET_SVE && <max_elem_bits> >= <elem_bits>” “<sve_pred_op>\t%0., %1, %0.” )

;; Same, but also producing a flags result. (define_insn_and_rewrite “*aarch64_sve_<sve_pred_op>_cc” [(set (reg:CC_NZC CC_REGNUM) (unspec:CC_NZC [(match_operand:VNx16BI 1 “register_operand” “Upa”) (match_operand 2) (match_operand:SI 3 “aarch64_sve_ptrue_flag”) (unspec:PRED_ALL [(match_operand 4) (match_operand:SI 5 “aarch64_sve_ptrue_flag”) (match_operand:PRED_ALL 6 “register_operand” “0”)] SVE_PITER)] UNSPEC_PTEST)) (set (match_operand:PRED_ALL 0 “register_operand” “=Upa”) (unspec:PRED_ALL [(match_dup 4) (match_dup 5) (match_dup 6)] SVE_PITER))] “TARGET_SVE && <max_elem_bits> >= <elem_bits> && aarch64_sve_same_pred_for_ptest_p (&operands[2], &operands[4])” “<sve_pred_op>\t%0., %1, %0.” “&& !rtx_equal_p (operands[2], operands[4])” { operands[4] = operands[2]; operands[5] = operands[3]; } )

;; Same, but with only the flags result being interesting. (define_insn_and_rewrite “*aarch64_sve_<sve_pred_op>_ptest” [(set (reg:CC_NZC CC_REGNUM) (unspec:CC_NZC [(match_operand:VNx16BI 1 “register_operand” “Upa”) (match_operand 2) (match_operand:SI 3 “aarch64_sve_ptrue_flag”) (unspec:PRED_ALL [(match_operand 4) (match_operand:SI 5 “aarch64_sve_ptrue_flag”) (match_operand:PRED_ALL 6 “register_operand” “0”)] SVE_PITER)] UNSPEC_PTEST)) (clobber (match_scratch:PRED_ALL 0 “=Upa”))] “TARGET_SVE && <max_elem_bits> >= <elem_bits> && aarch64_sve_same_pred_for_ptest_p (&operands[2], &operands[4])” “<sve_pred_op>\t%0., %1, %0.” “&& !rtx_equal_p (operands[2], operands[4])” { operands[4] = operands[2]; operands[5] = operands[3]; } )

;; ========================================================================= ;; == Counting elements ;; =========================================================================

;; ------------------------------------------------------------------------- ;; ---- [INT] Count elements in a pattern (scalar) ;; ------------------------------------------------------------------------- ;; Includes: ;; - CNTB ;; - CNTD ;; - CNTH ;; - CNTW ;; -------------------------------------------------------------------------

;; Count the number of elements in an svpattern. Operand 1 is the pattern, ;; operand 2 is the number of elements that fit in a 128-bit block, and ;; operand 3 is a multiplier in the range [1, 16]. ;; ;; Note that this pattern isn't used for SV_ALL (but would work for that too). (define_insn “aarch64_sve_cnt_pat” [(set (match_operand:DI 0 “register_operand” “=r”) (zero_extend:DI (unspec:SI [(match_operand:DI 1 “const_int_operand”) (match_operand:DI 2 “const_int_operand”) (match_operand:DI 3 “const_int_operand”)] UNSPEC_SVE_CNT_PAT)))] “TARGET_SVE” { return aarch64_output_sve_cnt_pat_immediate (“cnt”, “%x0”, operands + 1); } )

;; ------------------------------------------------------------------------- ;; ---- [INT] Increment by the number of elements in a pattern (scalar) ;; ------------------------------------------------------------------------- ;; Includes: ;; - INC ;; - SQINC ;; - UQINC ;; -------------------------------------------------------------------------

;; Increment a DImode register by the number of elements in an svpattern. ;; See aarch64_sve_cnt_pat for the counting behavior. (define_insn “@aarch64_sve_<inc_dec>_pat” [(set (match_operand:DI 0 “register_operand” “=r”) (ANY_PLUS:DI (zero_extend:DI (unspec:SI [(match_operand:DI 2 “const_int_operand”) (match_operand:DI 3 “const_int_operand”) (match_operand:DI 4 “const_int_operand”)] UNSPEC_SVE_CNT_PAT)) (match_operand:DI_ONLY 1 “register_operand” “0”)))] “TARGET_SVE” { return aarch64_output_sve_cnt_pat_immediate (“<inc_dec>”, “%x0”, operands + 2); } )

;; Increment an SImode register by the number of elements in an svpattern ;; using modular arithmetic. See aarch64_sve_cnt_pat for the counting ;; behavior. (define_insn “*aarch64_sve_incsi_pat” [(set (match_operand:SI 0 “register_operand” “=r”) (plus:SI (unspec:SI [(match_operand:DI 2 “const_int_operand”) (match_operand:DI 3 “const_int_operand”) (match_operand:DI 4 “const_int_operand”)] UNSPEC_SVE_CNT_PAT) (match_operand:SI 1 “register_operand” “0”)))] “TARGET_SVE” { return aarch64_output_sve_cnt_pat_immediate (“inc”, “%x0”, operands + 2); } )

;; Increment an SImode register by the number of elements in an svpattern ;; using saturating arithmetic, extending the result to 64 bits. ;; ;; See aarch64_sve_cnt_pat for the counting behavior. (define_insn “@aarch64_sve_<inc_dec>_pat” [(set (match_operand:DI 0 “register_operand” “=r”) (<paired_extend>:DI (SAT_PLUS:SI (unspec:SI [(match_operand:DI 2 “const_int_operand”) (match_operand:DI 3 “const_int_operand”) (match_operand:DI 4 “const_int_operand”)] UNSPEC_SVE_CNT_PAT) (match_operand:SI_ONLY 1 “register_operand” “0”))))] “TARGET_SVE” { const char *registers = ( == SS_PLUS ? “%x0, %w0” : “%w0”); return aarch64_output_sve_cnt_pat_immediate (“<inc_dec>”, registers, operands + 2); } )

;; ------------------------------------------------------------------------- ;; ---- [INT] Increment by the number of elements in a pattern (vector) ;; ------------------------------------------------------------------------- ;; Includes: ;; - INC ;; - SQINC ;; - UQINC ;; -------------------------------------------------------------------------

;; Increment a vector of DIs by the number of elements in an svpattern. ;; See aarch64_sve_cnt_pat for the counting behavior. (define_insn “@aarch64_sve_<inc_dec>_pat” [(set (match_operand:VNx2DI 0 “register_operand” “=w, ?&w”) (ANY_PLUS:VNx2DI (vec_duplicate:VNx2DI (zero_extend:DI (unspec:SI [(match_operand:DI 2 “const_int_operand”) (match_operand:DI 3 “const_int_operand”) (match_operand:DI 4 “const_int_operand”)] UNSPEC_SVE_CNT_PAT))) (match_operand:VNx2DI_ONLY 1 “register_operand” “0, w”)))] “TARGET_SVE” { if (which_alternative == 1) output_asm_insn (“movprfx\t%0, %1”, operands); return aarch64_output_sve_cnt_pat_immediate (“<inc_dec>”, “%0.”, operands + 2); } [(set_attr “movprfx” “*,yes”)] )

;; Increment a vector of SIs by the number of elements in an svpattern. ;; See aarch64_sve_cnt_pat for the counting behavior. (define_insn “@aarch64_sve_<inc_dec>_pat” [(set (match_operand:VNx4SI 0 “register_operand” “=w, ?&w”) (ANY_PLUS:VNx4SI (vec_duplicate:VNx4SI (unspec:SI [(match_operand:DI 2 “const_int_operand”) (match_operand:DI 3 “const_int_operand”) (match_operand:DI 4 “const_int_operand”)] UNSPEC_SVE_CNT_PAT)) (match_operand:VNx4SI_ONLY 1 “register_operand” “0, w”)))] “TARGET_SVE” { if (which_alternative == 1) output_asm_insn (“movprfx\t%0, %1”, operands); return aarch64_output_sve_cnt_pat_immediate (“<inc_dec>”, “%0.”, operands + 2); } [(set_attr “movprfx” “*,yes”)] )

;; Increment a vector of HIs by the number of elements in an svpattern. ;; See aarch64_sve_cnt_pat for the counting behavior. (define_expand “@aarch64_sve_<inc_dec>_pat” [(set (match_operand:VNx8HI 0 “register_operand”) (ANY_PLUS:VNx8HI (vec_duplicate:VNx8HI (truncate:HI (unspec:SI [(match_operand:DI 2 “const_int_operand”) (match_operand:DI 3 “const_int_operand”) (match_operand:DI 4 “const_int_operand”)] UNSPEC_SVE_CNT_PAT))) (match_operand:VNx8HI_ONLY 1 “register_operand”)))] “TARGET_SVE” )

(define_insn “*aarch64_sve_<inc_dec>_pat” [(set (match_operand:VNx8HI 0 “register_operand” “=w, ?&w”) (ANY_PLUS:VNx8HI (vec_duplicate:VNx8HI (match_operator:HI 5 “subreg_lowpart_operator” [(unspec:SI [(match_operand:DI 2 “const_int_operand”) (match_operand:DI 3 “const_int_operand”) (match_operand:DI 4 “const_int_operand”)] UNSPEC_SVE_CNT_PAT)])) (match_operand:VNx8HI_ONLY 1 “register_operand” “0, w”)))] “TARGET_SVE” { if (which_alternative == 1) output_asm_insn (“movprfx\t%0, %1”, operands); return aarch64_output_sve_cnt_pat_immediate (“<inc_dec>”, “%0.”, operands + 2); } [(set_attr “movprfx” “*,yes”)] )

;; ------------------------------------------------------------------------- ;; ---- [INT] Decrement by the number of elements in a pattern (scalar) ;; ------------------------------------------------------------------------- ;; Includes: ;; - DEC ;; - SQDEC ;; - UQDEC ;; -------------------------------------------------------------------------

;; Decrement a DImode register by the number of elements in an svpattern. ;; See aarch64_sve_cnt_pat for the counting behavior. (define_insn “@aarch64_sve_<inc_dec>_pat” [(set (match_operand:DI 0 “register_operand” “=r”) (ANY_MINUS:DI (match_operand:DI_ONLY 1 “register_operand” “0”) (zero_extend:DI (unspec:SI [(match_operand:DI 2 “const_int_operand”) (match_operand:DI 3 “const_int_operand”) (match_operand:DI 4 “const_int_operand”)] UNSPEC_SVE_CNT_PAT))))] “TARGET_SVE” { return aarch64_output_sve_cnt_pat_immediate (“<inc_dec>”, “%x0”, operands + 2); } )

;; Decrement an SImode register by the number of elements in an svpattern ;; using modular arithmetic. See aarch64_sve_cnt_pat for the counting ;; behavior. (define_insn “*aarch64_sve_decsi_pat” [(set (match_operand:SI 0 “register_operand” “=r”) (minus:SI (match_operand:SI 1 “register_operand” “0”) (unspec:SI [(match_operand:DI 2 “const_int_operand”) (match_operand:DI 3 “const_int_operand”) (match_operand:DI 4 “const_int_operand”)] UNSPEC_SVE_CNT_PAT)))] “TARGET_SVE” { return aarch64_output_sve_cnt_pat_immediate (“dec”, “%x0”, operands + 2); } )

;; Decrement an SImode register by the number of elements in an svpattern ;; using saturating arithmetic, extending the result to 64 bits. ;; ;; See aarch64_sve_cnt_pat for the counting behavior. (define_insn “@aarch64_sve_<inc_dec>_pat” [(set (match_operand:DI 0 “register_operand” “=r”) (<paired_extend>:DI (SAT_MINUS:SI (match_operand:SI_ONLY 1 “register_operand” “0”) (unspec:SI [(match_operand:DI 2 “const_int_operand”) (match_operand:DI 3 “const_int_operand”) (match_operand:DI 4 “const_int_operand”)] UNSPEC_SVE_CNT_PAT))))] “TARGET_SVE” { const char *registers = ( == SS_MINUS ? “%x0, %w0” : “%w0”); return aarch64_output_sve_cnt_pat_immediate (“<inc_dec>”, registers, operands + 2); } )

;; ------------------------------------------------------------------------- ;; ---- [INT] Decrement by the number of elements in a pattern (vector) ;; ------------------------------------------------------------------------- ;; Includes: ;; - DEC ;; - SQDEC ;; - UQDEC ;; -------------------------------------------------------------------------

;; Decrement a vector of DIs by the number of elements in an svpattern. ;; See aarch64_sve_cnt_pat for the counting behavior. (define_insn “@aarch64_sve_<inc_dec>_pat” [(set (match_operand:VNx2DI 0 “register_operand” “=w, ?&w”) (ANY_MINUS:VNx2DI (match_operand:VNx2DI_ONLY 1 “register_operand” “0, w”) (vec_duplicate:VNx2DI (zero_extend:DI (unspec:SI [(match_operand:DI 2 “const_int_operand”) (match_operand:DI 3 “const_int_operand”) (match_operand:DI 4 “const_int_operand”)] UNSPEC_SVE_CNT_PAT)))))] “TARGET_SVE” { if (which_alternative == 1) output_asm_insn (“movprfx\t%0, %1”, operands); return aarch64_output_sve_cnt_pat_immediate (“<inc_dec>”, “%0.”, operands + 2); } [(set_attr “movprfx” “*,yes”)] )

;; Decrement a vector of SIs by the number of elements in an svpattern. ;; See aarch64_sve_cnt_pat for the counting behavior. (define_insn “@aarch64_sve_<inc_dec>_pat” [(set (match_operand:VNx4SI 0 “register_operand” “=w, ?&w”) (ANY_MINUS:VNx4SI (match_operand:VNx4SI_ONLY 1 “register_operand” “0, w”) (vec_duplicate:VNx4SI (unspec:SI [(match_operand:DI 2 “const_int_operand”) (match_operand:DI 3 “const_int_operand”) (match_operand:DI 4 “const_int_operand”)] UNSPEC_SVE_CNT_PAT))))] “TARGET_SVE” { if (which_alternative == 1) output_asm_insn (“movprfx\t%0, %1”, operands); return aarch64_output_sve_cnt_pat_immediate (“<inc_dec>”, “%0.”, operands + 2); } [(set_attr “movprfx” “*,yes”)] )

;; Decrement a vector of HIs by the number of elements in an svpattern. ;; See aarch64_sve_cnt_pat for the counting behavior. (define_expand “@aarch64_sve_<inc_dec>_pat” [(set (match_operand:VNx8HI 0 “register_operand”) (ANY_MINUS:VNx8HI (match_operand:VNx8HI_ONLY 1 “register_operand”) (vec_duplicate:VNx8HI (truncate:HI (unspec:SI [(match_operand:DI 2 “const_int_operand”) (match_operand:DI 3 “const_int_operand”) (match_operand:DI 4 “const_int_operand”)] UNSPEC_SVE_CNT_PAT)))))] “TARGET_SVE” )

(define_insn “*aarch64_sve_<inc_dec>_pat” [(set (match_operand:VNx8HI 0 “register_operand” “=w, ?&w”) (ANY_MINUS:VNx8HI (match_operand:VNx8HI_ONLY 1 “register_operand” “0, w”) (vec_duplicate:VNx8HI (match_operator:HI 5 “subreg_lowpart_operator” [(unspec:SI [(match_operand:DI 2 “const_int_operand”) (match_operand:DI 3 “const_int_operand”) (match_operand:DI 4 “const_int_operand”)] UNSPEC_SVE_CNT_PAT)]))))] “TARGET_SVE” { if (which_alternative == 1) output_asm_insn (“movprfx\t%0, %1”, operands); return aarch64_output_sve_cnt_pat_immediate (“<inc_dec>”, “%0.”, operands + 2); } [(set_attr “movprfx” “*,yes”)] )

;; ------------------------------------------------------------------------- ;; ---- [INT] Count elements in a predicate (scalar) ;; ------------------------------------------------------------------------- ;; Includes: ;; - CNTP ;; -------------------------------------------------------------------------

;; Count the number of set bits in a predicate. Operand 3 is true if ;; operand 1 is known to be all-true. (define_insn “@aarch64_pred_cntp” [(set (match_operand:DI 0 “register_operand” “=r”) (zero_extend:DI (unspec:SI [(match_operand:PRED_ALL 1 “register_operand” “Upl”) (match_operand:SI 2 “aarch64_sve_ptrue_flag”) (match_operand:PRED_ALL 3 “register_operand” “Upa”)] UNSPEC_CNTP)))] “TARGET_SVE” “cntp\t%x0, %1, %3.”)

;; ------------------------------------------------------------------------- ;; ---- [INT] Increment by the number of elements in a predicate (scalar) ;; ------------------------------------------------------------------------- ;; Includes: ;; - INCP ;; - SQINCP ;; - UQINCP ;; -------------------------------------------------------------------------

;; Increment a DImode register by the number of set bits in a predicate. ;; See aarch64_sve_cntp for a description of the operands. (define_expand “@aarch64_sve_<inc_dec><DI_ONLY:mode><PRED_ALL:mode>_cntp” [(set (match_operand:DI 0 “register_operand”) (ANY_PLUS:DI (zero_extend:DI (unspec:SI [(match_dup 3) (const_int SVE_KNOWN_PTRUE) (match_operand:PRED_ALL 2 “register_operand”)] UNSPEC_CNTP)) (match_operand:DI_ONLY 1 “register_operand”)))] “TARGET_SVE” { operands[3] = CONSTM1_RTX (<PRED_ALL:MODE>mode); } )

(define_insn_and_rewrite “*aarch64_sve_<inc_dec><DI_ONLY:mode><PRED_ALL:mode>_cntp” [(set (match_operand:DI 0 “register_operand” “=r”) (ANY_PLUS:DI (zero_extend:DI (unspec:SI [(match_operand 3) (const_int SVE_KNOWN_PTRUE) (match_operand:PRED_ALL 2 “register_operand” “Upa”)] UNSPEC_CNTP)) (match_operand:DI_ONLY 1 “register_operand” “0”)))] “TARGET_SVE” “<inc_dec>p\t%x0, %2.<PRED_ALL:Vetype>” “&& !CONSTANT_P (operands[3])” { operands[3] = CONSTM1_RTX (<PRED_ALL:MODE>mode); } )

;; Increment an SImode register by the number of set bits in a predicate ;; using modular arithmetic. See aarch64_sve_cntp for a description of ;; the operands. (define_insn_and_rewrite “*aarch64_incsi_cntp” [(set (match_operand:SI 0 “register_operand” “=r”) (plus:SI (unspec:SI [(match_operand 3) (const_int SVE_KNOWN_PTRUE) (match_operand:PRED_ALL 2 “register_operand” “Upa”)] UNSPEC_CNTP) (match_operand:SI 1 “register_operand” “0”)))] “TARGET_SVE” “incp\t%x0, %2.” “&& !CONSTANT_P (operands[3])” { operands[3] = CONSTM1_RTX (mode); } )

;; Increment an SImode register by the number of set bits in a predicate ;; using saturating arithmetic, extending the result to 64 bits. ;; ;; See aarch64_sve_cntp for a description of the operands. (define_expand “@aarch64_sve_<inc_dec><SI_ONLY:mode><PRED_ALL:mode>_cntp” [(set (match_operand:DI 0 “register_operand”) (<paired_extend>:DI (SAT_PLUS:SI (unspec:SI [(match_dup 3) (const_int SVE_KNOWN_PTRUE) (match_operand:PRED_ALL 2 “register_operand”)] UNSPEC_CNTP) (match_operand:SI_ONLY 1 “register_operand”))))] “TARGET_SVE” { operands[3] = CONSTM1_RTX (<PRED_ALL:MODE>mode); } )

(define_insn_and_rewrite “*aarch64_sve_<inc_dec><SI_ONLY:mode><PRED_ALL:mode>_cntp” [(set (match_operand:DI 0 “register_operand” “=r”) (<paired_extend>:DI (SAT_PLUS:SI (unspec:SI [(match_operand 3) (const_int SVE_KNOWN_PTRUE) (match_operand:PRED_ALL 2 “register_operand” “Upa”)] UNSPEC_CNTP) (match_operand:SI_ONLY 1 “register_operand” “0”))))] “TARGET_SVE” { if ( == SS_PLUS) return “<inc_dec>p\t%x0, %2.<PRED_ALL:Vetype>, %w0”; else return “<inc_dec>p\t%w0, %2.<PRED_ALL:Vetype>”; } “&& !CONSTANT_P (operands[3])” { operands[3] = CONSTM1_RTX (<PRED_ALL:MODE>mode); } )

;; ------------------------------------------------------------------------- ;; ---- [INT] Increment by the number of elements in a predicate (vector) ;; ------------------------------------------------------------------------- ;; Includes: ;; - INCP ;; - SQINCP ;; - UQINCP ;; -------------------------------------------------------------------------

;; Increment a vector of DIs by the number of set bits in a predicate. ;; See aarch64_sve_cntp for a description of the operands. (define_expand “@aarch64_sve_<inc_dec>_cntp” [(set (match_operand:VNx2DI 0 “register_operand”) (ANY_PLUS:VNx2DI (vec_duplicate:VNx2DI (zero_extend:DI (unspec:SI [(match_dup 3) (const_int SVE_KNOWN_PTRUE) (match_operand: 2 “register_operand”)] UNSPEC_CNTP))) (match_operand:VNx2DI_ONLY 1 “register_operand”)))] “TARGET_SVE” { operands[3] = CONSTM1_RTX (mode); } )

(define_insn_and_rewrite “*aarch64_sve_<inc_dec>_cntp” [(set (match_operand:VNx2DI 0 “register_operand” “=w, ?&w”) (ANY_PLUS:VNx2DI (vec_duplicate:VNx2DI (zero_extend:DI (unspec:SI [(match_operand 3) (const_int SVE_KNOWN_PTRUE) (match_operand: 2 “register_operand” “Upa, Upa”)] UNSPEC_CNTP))) (match_operand:VNx2DI_ONLY 1 “register_operand” “0, w”)))] “TARGET_SVE” “@ <inc_dec>p\t%0.d, %2 movprfx\t%0, %1;<inc_dec>p\t%0.d, %2” “&& !CONSTANT_P (operands[3])” { operands[3] = CONSTM1_RTX (mode); } [(set_attr “movprfx” “*,yes”)] )

;; Increment a vector of SIs by the number of set bits in a predicate. ;; See aarch64_sve_cntp for a description of the operands. (define_expand “@aarch64_sve_<inc_dec>_cntp” [(set (match_operand:VNx4SI 0 “register_operand”) (ANY_PLUS:VNx4SI (vec_duplicate:VNx4SI (unspec:SI [(match_dup 3) (const_int SVE_KNOWN_PTRUE) (match_operand: 2 “register_operand”)] UNSPEC_CNTP)) (match_operand:VNx4SI_ONLY 1 “register_operand”)))] “TARGET_SVE” { operands[3] = CONSTM1_RTX (mode); } )

(define_insn_and_rewrite “*aarch64_sve_<inc_dec>_cntp” [(set (match_operand:VNx4SI 0 “register_operand” “=w, ?&w”) (ANY_PLUS:VNx4SI (vec_duplicate:VNx4SI (unspec:SI [(match_operand 3) (const_int SVE_KNOWN_PTRUE) (match_operand: 2 “register_operand” “Upa, Upa”)] UNSPEC_CNTP)) (match_operand:VNx4SI_ONLY 1 “register_operand” “0, w”)))] “TARGET_SVE” “@ <inc_dec>p\t%0.s, %2 movprfx\t%0, %1;<inc_dec>p\t%0.s, %2” “&& !CONSTANT_P (operands[3])” { operands[3] = CONSTM1_RTX (mode); } [(set_attr “movprfx” “*,yes”)] )

;; Increment a vector of HIs by the number of set bits in a predicate. ;; See aarch64_sve_cntp for a description of the operands. (define_expand “@aarch64_sve_<inc_dec>_cntp” [(set (match_operand:VNx8HI 0 “register_operand”) (ANY_PLUS:VNx8HI (vec_duplicate:VNx8HI (truncate:HI (unspec:SI [(match_dup 3) (const_int SVE_KNOWN_PTRUE) (match_operand: 2 “register_operand”)] UNSPEC_CNTP))) (match_operand:VNx8HI_ONLY 1 “register_operand”)))] “TARGET_SVE” { operands[3] = CONSTM1_RTX (mode); } )

(define_insn_and_rewrite “*aarch64_sve_<inc_dec>_cntp” [(set (match_operand:VNx8HI 0 “register_operand” “=w, ?&w”) (ANY_PLUS:VNx8HI (vec_duplicate:VNx8HI (match_operator:HI 3 “subreg_lowpart_operator” [(unspec:SI [(match_operand 4) (const_int SVE_KNOWN_PTRUE) (match_operand: 2 “register_operand” “Upa, Upa”)] UNSPEC_CNTP)])) (match_operand:VNx8HI_ONLY 1 “register_operand” “0, w”)))] “TARGET_SVE” “@ <inc_dec>p\t%0.h, %2 movprfx\t%0, %1;<inc_dec>p\t%0.h, %2” “&& !CONSTANT_P (operands[4])” { operands[4] = CONSTM1_RTX (mode); } [(set_attr “movprfx” “*,yes”)] )

;; ------------------------------------------------------------------------- ;; ---- [INT] Decrement by the number of elements in a predicate (scalar) ;; ------------------------------------------------------------------------- ;; Includes: ;; - DECP ;; - SQDECP ;; - UQDECP ;; -------------------------------------------------------------------------

;; Decrement a DImode register by the number of set bits in a predicate. ;; See aarch64_sve_cntp for a description of the operands. (define_expand “@aarch64_sve_<inc_dec><DI_ONLY:mode><PRED_ALL:mode>_cntp” [(set (match_operand:DI 0 “register_operand”) (ANY_MINUS:DI (match_operand:DI_ONLY 1 “register_operand”) (zero_extend:DI (unspec:SI [(match_dup 3) (const_int SVE_KNOWN_PTRUE) (match_operand:PRED_ALL 2 “register_operand”)] UNSPEC_CNTP))))] “TARGET_SVE” { operands[3] = CONSTM1_RTX (<PRED_ALL:MODE>mode); } )

(define_insn_and_rewrite “*aarch64_sve_<inc_dec><DI_ONLY:mode><PRED_ALL:mode>_cntp” [(set (match_operand:DI 0 “register_operand” “=r”) (ANY_MINUS:DI (match_operand:DI_ONLY 1 “register_operand” “0”) (zero_extend:DI (unspec:SI [(match_operand 3) (const_int SVE_KNOWN_PTRUE) (match_operand:PRED_ALL 2 “register_operand” “Upa”)] UNSPEC_CNTP))))] “TARGET_SVE” “<inc_dec>p\t%x0, %2.<PRED_ALL:Vetype>” “&& !CONSTANT_P (operands[3])” { operands[3] = CONSTM1_RTX (<PRED_ALL:MODE>mode); } )

;; Decrement an SImode register by the number of set bits in a predicate ;; using modular arithmetic. See aarch64_sve_cntp for a description of the ;; operands. (define_insn_and_rewrite “*aarch64_decsi_cntp” [(set (match_operand:SI 0 “register_operand” “=r”) (minus:SI (match_operand:SI 1 “register_operand” “0”) (unspec:SI [(match_operand 3) (const_int SVE_KNOWN_PTRUE) (match_operand:PRED_ALL 2 “register_operand” “Upa”)] UNSPEC_CNTP)))] “TARGET_SVE” “decp\t%x0, %2.” “&& !CONSTANT_P (operands[3])” { operands[3] = CONSTM1_RTX (mode); } )

;; Decrement an SImode register by the number of set bits in a predicate ;; using saturating arithmetic, extending the result to 64 bits. ;; ;; See aarch64_sve_cntp for a description of the operands. (define_expand “@aarch64_sve_<inc_dec><SI_ONLY:mode><PRED_ALL:mode>_cntp” [(set (match_operand:DI 0 “register_operand”) (<paired_extend>:DI (SAT_MINUS:SI (match_operand:SI_ONLY 1 “register_operand”) (unspec:SI [(match_dup 3) (const_int SVE_KNOWN_PTRUE) (match_operand:PRED_ALL 2 “register_operand”)] UNSPEC_CNTP))))] “TARGET_SVE” { operands[3] = CONSTM1_RTX (<PRED_ALL:MODE>mode); } )

(define_insn_and_rewrite “*aarch64_sve_<inc_dec><SI_ONLY:mode><PRED_ALL:mode>_cntp” [(set (match_operand:DI 0 “register_operand” “=r”) (<paired_extend>:DI (SAT_MINUS:SI (match_operand:SI_ONLY 1 “register_operand” “0”) (unspec:SI [(match_operand 3) (const_int SVE_KNOWN_PTRUE) (match_operand:PRED_ALL 2 “register_operand” “Upa”)] UNSPEC_CNTP))))] “TARGET_SVE” { if ( == SS_MINUS) return “<inc_dec>p\t%x0, %2.<PRED_ALL:Vetype>, %w0”; else return “<inc_dec>p\t%w0, %2.<PRED_ALL:Vetype>”; } “&& !CONSTANT_P (operands[3])” { operands[3] = CONSTM1_RTX (<PRED_ALL:MODE>mode); } )

;; ------------------------------------------------------------------------- ;; ---- [INT] Decrement by the number of elements in a predicate (vector) ;; ------------------------------------------------------------------------- ;; Includes: ;; - DECP ;; - SQDECP ;; - UQDECP ;; -------------------------------------------------------------------------

;; Decrement a vector of DIs by the number of set bits in a predicate. ;; See aarch64_sve_cntp for a description of the operands. (define_expand “@aarch64_sve_<inc_dec>_cntp” [(set (match_operand:VNx2DI 0 “register_operand”) (ANY_MINUS:VNx2DI (match_operand:VNx2DI_ONLY 1 “register_operand”) (vec_duplicate:VNx2DI (zero_extend:DI (unspec:SI [(match_dup 3) (const_int SVE_KNOWN_PTRUE) (match_operand: 2 “register_operand”)] UNSPEC_CNTP)))))] “TARGET_SVE” { operands[3] = CONSTM1_RTX (mode); } )

(define_insn_and_rewrite “*aarch64_sve_<inc_dec>_cntp” [(set (match_operand:VNx2DI 0 “register_operand” “=w, ?&w”) (ANY_MINUS:VNx2DI (match_operand:VNx2DI_ONLY 1 “register_operand” “0, w”) (vec_duplicate:VNx2DI (zero_extend:DI (unspec:SI [(match_operand 3) (const_int SVE_KNOWN_PTRUE) (match_operand: 2 “register_operand” “Upa, Upa”)] UNSPEC_CNTP)))))] “TARGET_SVE” “@ <inc_dec>p\t%0.d, %2 movprfx\t%0, %1;<inc_dec>p\t%0.d, %2” “&& !CONSTANT_P (operands[3])” { operands[3] = CONSTM1_RTX (mode); } [(set_attr “movprfx” “*,yes”)] )

;; Decrement a vector of SIs by the number of set bits in a predicate. ;; See aarch64_sve_cntp for a description of the operands. (define_expand “@aarch64_sve_<inc_dec>_cntp” [(set (match_operand:VNx4SI 0 “register_operand”) (ANY_MINUS:VNx4SI (match_operand:VNx4SI_ONLY 1 “register_operand”) (vec_duplicate:VNx4SI (unspec:SI [(match_dup 3) (const_int SVE_KNOWN_PTRUE) (match_operand: 2 “register_operand”)] UNSPEC_CNTP))))] “TARGET_SVE” { operands[3] = CONSTM1_RTX (mode); } )

(define_insn_and_rewrite “*aarch64_sve_<inc_dec>_cntp” [(set (match_operand:VNx4SI 0 “register_operand” “=w, ?&w”) (ANY_MINUS:VNx4SI (match_operand:VNx4SI_ONLY 1 “register_operand” “0, w”) (vec_duplicate:VNx4SI (unspec:SI [(match_operand 3) (const_int SVE_KNOWN_PTRUE) (match_operand: 2 “register_operand” “Upa, Upa”)] UNSPEC_CNTP))))] “TARGET_SVE” “@ <inc_dec>p\t%0.s, %2 movprfx\t%0, %1;<inc_dec>p\t%0.s, %2” “&& !CONSTANT_P (operands[3])” { operands[3] = CONSTM1_RTX (mode); } [(set_attr “movprfx” “*,yes”)] )

;; Decrement a vector of HIs by the number of set bits in a predicate. ;; See aarch64_sve_cntp for a description of the operands. (define_expand “@aarch64_sve_<inc_dec>_cntp” [(set (match_operand:VNx8HI 0 “register_operand”) (ANY_MINUS:VNx8HI (match_operand:VNx8HI_ONLY 1 “register_operand”) (vec_duplicate:VNx8HI (truncate:HI (unspec:SI [(match_dup 3) (const_int SVE_KNOWN_PTRUE) (match_operand: 2 “register_operand”)] UNSPEC_CNTP)))))] “TARGET_SVE” { operands[3] = CONSTM1_RTX (mode); } )

(define_insn_and_rewrite “*aarch64_sve_<inc_dec>_cntp” [(set (match_operand:VNx8HI 0 “register_operand” “=w, ?&w”) (ANY_MINUS:VNx8HI (match_operand:VNx8HI_ONLY 1 “register_operand” “0, w”) (vec_duplicate:VNx8HI (match_operator:HI 3 “subreg_lowpart_operator” [(unspec:SI [(match_operand 4) (const_int SVE_KNOWN_PTRUE) (match_operand: 2 “register_operand” “Upa, Upa”)] UNSPEC_CNTP)]))))] “TARGET_SVE” “@ <inc_dec>p\t%0.h, %2 movprfx\t%0, %1;<inc_dec>p\t%0.h, %2” “&& !CONSTANT_P (operands[4])” { operands[4] = CONSTM1_RTX (mode); } [(set_attr “movprfx” “*,yes”)] )