[PATCH V2] VECT: Apply COND_LEN_* into vectorizable_operation
Richard Sandiford
richard.sandiford@arm.com
Wed Jul 12 10:53:23 GMT 2023
juzhe.zhong@rivai.ai writes:
> From: Ju-Zhe Zhong <juzhe.zhong@rivai.ai>
>
> Hi, Richard and Richi.
> As we disscussed before, COND_LEN_* patterns were added for multiple situations.
> This patch apply CON_LEN_* for the following situation:
>
> Support for the situation that in "vectorizable_operation":
> /* If operating on inactive elements could generate spurious traps,
> we need to restrict the operation to active lanes. Note that this
> specifically doesn't apply to unhoisted invariants, since they
> operate on the same value for every lane.
>
> Similarly, if this operation is part of a reduction, a fully-masked
> loop should only change the active lanes of the reduction chain,
> keeping the inactive lanes as-is. */
> bool mask_out_inactive = ((!is_invariant && gimple_could_trap_p (stmt))
> || reduc_idx >= 0);
>
> For mask_out_inactive is true with length loop control.
>
> So, we can these 2 following cases:
>
> 1. Integer division:
>
> #define TEST_TYPE(TYPE) \
> __attribute__((noipa)) \
> void vrem_##TYPE (TYPE *dst, TYPE *a, TYPE *b, int n) \
> { \
> for (int i = 0; i < n; i++) \
> dst[i] = a[i] % b[i]; \
> }
> #define TEST_ALL() \
> TEST_TYPE(int8_t) \
> TEST_ALL()
>
> With this patch:
>
> _61 = .SELECT_VL (ivtmp_59, POLY_INT_CST [4, 4]);
> ivtmp_45 = _61 * 4;
> vect__4.8_48 = .LEN_MASK_LOAD (vectp_a.6_46, 32B, _61, 0, { -1, ... });
> vect__6.11_52 = .LEN_MASK_LOAD (vectp_b.9_50, 32B, _61, 0, { -1, ... });
> vect__8.12_53 = .COND_LEN_ADD ({ -1, ... }, vect__4.8_48, vect__6.11_52, vect__4.8_48, _61, 0);
> .LEN_MASK_STORE (vectp_dst.13_55, 32B, _61, 0, { -1, ... }, vect__8.12_53);
>
> 2. Floating-point arithmetic **WITHOUT** -ffast-math
>
> #define TEST_TYPE(TYPE) \
> __attribute__((noipa)) \
> void vadd_##TYPE (TYPE *dst, TYPE *a, TYPE *b, int n) \
> { \
> for (int i = 0; i < n; i++) \
> dst[i] = a[i] + b[i]; \
> }
> #define TEST_ALL() \
> TEST_TYPE(float) \
> TEST_ALL()
>
> With this patch:
>
> _61 = .SELECT_VL (ivtmp_59, POLY_INT_CST [4, 4]);
> ivtmp_45 = _61 * 4;
> vect__4.8_48 = .LEN_MASK_LOAD (vectp_a.6_46, 32B, _61, 0, { -1, ... });
> vect__6.11_52 = .LEN_MASK_LOAD (vectp_b.9_50, 32B, _61, 0, { -1, ... });
> vect__8.12_53 = .COND_LEN_ADD ({ -1, ... }, vect__4.8_48, vect__6.11_52, vect__4.8_48, _61, 0);
> .LEN_MASK_STORE (vectp_dst.13_55, 32B, _61, 0, { -1, ... }, vect__8.12_53);
>
> With this patch, we can make sure operations won't trap for elements that "mask_out_inactive".
>
> gcc/ChangeLog:
>
> * internal-fn.cc (FOR_EACH_CODE_MAPPING): Adapt for COND_LEN_* support.
> (CASE): Ditto.
> (get_conditional_len_internal_fn): New function.
> * internal-fn.h (get_conditional_len_internal_fn): Ditto.
> * tree-vect-stmts.cc (vectorizable_operation): Adapt for COND_LEN_* support.
>
> ---
> gcc/internal-fn.cc | 65 ++++++++++++++++++++++++++++++------------
> gcc/internal-fn.h | 1 +
> gcc/tree-vect-stmts.cc | 48 ++++++++++++++++++++++++-------
> 3 files changed, 85 insertions(+), 29 deletions(-)
>
> diff --git a/gcc/internal-fn.cc b/gcc/internal-fn.cc
> index f9aaf66cf2a..7e3a8cc8412 100644
> --- a/gcc/internal-fn.cc
> +++ b/gcc/internal-fn.cc
> @@ -4276,23 +4276,24 @@ static void (*const internal_fn_expanders[]) (internal_fn, gcall *) = {
> 0
> };
>
> -/* Invoke T(CODE, IFN) for each conditional function IFN that maps to a
> - tree code CODE. */
> +/* Invoke T(CODE, SUFFIX) for each conditional function IFN_COND_##SUFFIX
> + that maps to a tree code CODE. There is also an IFN_COND_LEN_##SUFFIX
> + for each such IFN_COND_##SUFFIX. */
> #define FOR_EACH_CODE_MAPPING(T) \
> - T (PLUS_EXPR, IFN_COND_ADD) \
> - T (MINUS_EXPR, IFN_COND_SUB) \
> - T (MULT_EXPR, IFN_COND_MUL) \
> - T (TRUNC_DIV_EXPR, IFN_COND_DIV) \
> - T (TRUNC_MOD_EXPR, IFN_COND_MOD) \
> - T (RDIV_EXPR, IFN_COND_RDIV) \
> - T (MIN_EXPR, IFN_COND_MIN) \
> - T (MAX_EXPR, IFN_COND_MAX) \
> - T (BIT_AND_EXPR, IFN_COND_AND) \
> - T (BIT_IOR_EXPR, IFN_COND_IOR) \
> - T (BIT_XOR_EXPR, IFN_COND_XOR) \
> - T (LSHIFT_EXPR, IFN_COND_SHL) \
> - T (RSHIFT_EXPR, IFN_COND_SHR) \
> - T (NEGATE_EXPR, IFN_COND_NEG)
> + T (PLUS_EXPR, ADD) \
> + T (MINUS_EXPR, SUB) \
> + T (MULT_EXPR, MUL) \
> + T (TRUNC_DIV_EXPR, DIV) \
> + T (TRUNC_MOD_EXPR, MOD) \
> + T (RDIV_EXPR, RDIV) \
> + T (MIN_EXPR, MIN) \
> + T (MAX_EXPR, MAX) \
> + T (BIT_AND_EXPR, AND) \
> + T (BIT_IOR_EXPR, IOR) \
> + T (BIT_XOR_EXPR, XOR) \
> + T (LSHIFT_EXPR, SHL) \
> + T (RSHIFT_EXPR, SHR) \
> + T (NEGATE_EXPR, NEG)
>
> /* Return a function that only performs CODE when a certain condition is met
> and that uses a given fallback value otherwise. For example, if CODE is
> @@ -4313,7 +4314,7 @@ get_conditional_internal_fn (tree_code code)
> {
> switch (code)
> {
> -#define CASE(CODE, IFN) case CODE: return IFN;
> +#define CASE(CODE, IFN) case CODE: return IFN_COND_##IFN;
> FOR_EACH_CODE_MAPPING(CASE)
> #undef CASE
> default:
> @@ -4329,7 +4330,7 @@ conditional_internal_fn_code (internal_fn ifn)
> {
> switch (ifn)
> {
> -#define CASE(CODE, IFN) case IFN: return CODE;
> +#define CASE(CODE, IFN) case IFN_COND_##IFN: return CODE;
> FOR_EACH_CODE_MAPPING(CASE)
> #undef CASE
> default:
> @@ -4337,6 +4338,34 @@ conditional_internal_fn_code (internal_fn ifn)
> }
> }
>
> +/* Return a function that only performs CODE when a certain condition is met
> + and that uses a given fallback value otherwise. For example, if CODE is
This describes COND_* rather than COND_LEN_*. How about:
/* Like get_conditional_internal_fn, but return a function that
additionally restricts the operation to the leading elements
of a vector. The number of elements to process is given by
a length and bias pair, as for IFN_LOAD_LEN. The values of
the remaining elements are undefined.
For example, if CODE is [...]
(Please confirm that the last sentence of the first paragraph is true.)
> + a binary operation associated with conditional function FN:
> +
> + LHS = FN (COND, A, B, ELSE, LEN, BIAS)
> +
> + is equivalent to the C expression:
> +
> + for (int i = 0; i < LEN + BIAS; i++)
> + LHS[i] = COND[i] ? A[i] CODE B[i] : ELSE[i];
> +
> + operating elementwise if the operands are vectors.
> +
The "operating elementwise" doesn't make sense in this context,
since the example is a loop that always operates elementwise.
Seems better to drop that line.
OK with those changes, thanks.
Richard
> + Return IFN_LAST if no such function exists. */
> +
> +internal_fn
> +get_conditional_len_internal_fn (tree_code code)
> +{
> + switch (code)
> + {
> +#define CASE(CODE, IFN) case CODE: return IFN_COND_LEN_##IFN;
> + FOR_EACH_CODE_MAPPING(CASE)
> +#undef CASE
> + default:
> + return IFN_LAST;
> + }
> +}
> +
> /* Invoke T(IFN) for each internal function IFN that also has an
> IFN_COND_* form. */
> #define FOR_EACH_COND_FN_PAIR(T) \
> diff --git a/gcc/internal-fn.h b/gcc/internal-fn.h
> index 4234bbfed87..dd1bab0bddf 100644
> --- a/gcc/internal-fn.h
> +++ b/gcc/internal-fn.h
> @@ -224,6 +224,7 @@ extern bool set_edom_supported_p (void);
>
> extern internal_fn get_conditional_internal_fn (tree_code);
> extern internal_fn get_conditional_internal_fn (internal_fn);
> +extern internal_fn get_conditional_len_internal_fn (tree_code);
> extern tree_code conditional_internal_fn_code (internal_fn);
> extern internal_fn get_unconditional_internal_fn (internal_fn);
> extern bool can_interpret_as_conditional_op_p (gimple *, tree *,
> diff --git a/gcc/tree-vect-stmts.cc b/gcc/tree-vect-stmts.cc
> index 10e71178ce7..dd24f017235 100644
> --- a/gcc/tree-vect-stmts.cc
> +++ b/gcc/tree-vect-stmts.cc
> @@ -6711,7 +6711,9 @@ vectorizable_operation (vec_info *vinfo,
>
> int reduc_idx = STMT_VINFO_REDUC_IDX (stmt_info);
> vec_loop_masks *masks = (loop_vinfo ? &LOOP_VINFO_MASKS (loop_vinfo) : NULL);
> + vec_loop_lens *lens = (loop_vinfo ? &LOOP_VINFO_LENS (loop_vinfo) : NULL);
> internal_fn cond_fn = get_conditional_internal_fn (code);
> + internal_fn cond_len_fn = get_conditional_len_internal_fn (code);
>
> /* If operating on inactive elements could generate spurious traps,
> we need to restrict the operation to active lanes. Note that this
> @@ -6730,9 +6732,17 @@ vectorizable_operation (vec_info *vinfo,
> && LOOP_VINFO_CAN_USE_PARTIAL_VECTORS_P (loop_vinfo)
> && mask_out_inactive)
> {
> - if (cond_fn == IFN_LAST
> - || !direct_internal_fn_supported_p (cond_fn, vectype,
> - OPTIMIZE_FOR_SPEED))
> + if (cond_fn != IFN_LAST
> + && direct_internal_fn_supported_p (cond_fn, vectype,
> + OPTIMIZE_FOR_SPEED))
> + vect_record_loop_mask (loop_vinfo, masks, ncopies * vec_num,
> + vectype, NULL);
> + else if (cond_len_fn != IFN_LAST
> + && direct_internal_fn_supported_p (cond_len_fn, vectype,
> + OPTIMIZE_FOR_SPEED))
> + vect_record_loop_len (loop_vinfo, lens, ncopies * vec_num, vectype,
> + 1);
> + else
> {
> if (dump_enabled_p ())
> dump_printf_loc (MSG_MISSED_OPTIMIZATION, vect_location,
> @@ -6740,9 +6750,6 @@ vectorizable_operation (vec_info *vinfo,
> " conditional operation is available.\n");
> LOOP_VINFO_CAN_USE_PARTIAL_VECTORS_P (loop_vinfo) = false;
> }
> - else
> - vect_record_loop_mask (loop_vinfo, masks, ncopies * vec_num,
> - vectype, NULL);
> }
>
> /* Put types on constant and invariant SLP children. */
> @@ -6805,6 +6812,7 @@ vectorizable_operation (vec_info *vinfo,
> "transform binary/unary operation.\n");
>
> bool masked_loop_p = loop_vinfo && LOOP_VINFO_FULLY_MASKED_P (loop_vinfo);
> + bool len_loop_p = loop_vinfo && LOOP_VINFO_FULLY_WITH_LENGTH_P (loop_vinfo);
>
> /* POINTER_DIFF_EXPR has pointer arguments which are vectorized as
> vectors with unsigned elements, but the result is signed. So, we
> @@ -6971,11 +6979,16 @@ vectorizable_operation (vec_info *vinfo,
> gimple_assign_set_lhs (new_stmt, new_temp);
> vect_finish_stmt_generation (vinfo, stmt_info, new_stmt, gsi);
> }
> - else if (masked_loop_p && mask_out_inactive)
> + else if ((masked_loop_p || len_loop_p) && mask_out_inactive)
> {
> - tree mask = vect_get_loop_mask (loop_vinfo, gsi, masks,
> - vec_num * ncopies, vectype, i);
> - auto_vec<tree> vops (5);
> + tree mask;
> + if (masked_loop_p)
> + mask = vect_get_loop_mask (loop_vinfo, gsi, masks,
> + vec_num * ncopies, vectype, i);
> + else
> + /* Dummy mask. */
> + mask = build_minus_one_cst (truth_type_for (vectype));
> + auto_vec<tree> vops (6);
> vops.quick_push (mask);
> vops.quick_push (vop0);
> if (vop1)
> @@ -6995,7 +7008,20 @@ vectorizable_operation (vec_info *vinfo,
> (cond_fn, vectype, vops.length () - 1, &vops[1]);
> vops.quick_push (else_value);
> }
> - gcall *call = gimple_build_call_internal_vec (cond_fn, vops);
> + if (len_loop_p)
> + {
> + tree len = vect_get_loop_len (loop_vinfo, gsi, lens,
> + vec_num * ncopies, vectype, i, 1);
> + signed char biasval
> + = LOOP_VINFO_PARTIAL_LOAD_STORE_BIAS (loop_vinfo);
> + tree bias = build_int_cst (intQI_type_node, biasval);
> + vops.quick_push (len);
> + vops.quick_push (bias);
> + }
> + gcall *call
> + = gimple_build_call_internal_vec (masked_loop_p ? cond_fn
> + : cond_len_fn,
> + vops);
> new_temp = make_ssa_name (vec_dest, call);
> gimple_call_set_lhs (call, new_temp);
> gimple_call_set_nothrow (call, true);
More information about the Gcc-patches
mailing list