2019-11-06 Pat Bernardi <bernardi@adacore.com>
Jerome Lambourg <lambourg@adacore.com>
Olivier Hainque <hainque@adacore.com>
gcc/
* config.gcc: Add comment to introduce the TARGET_VXWORKS
commong macro definitions, conveying VXWORKS7 or 64bit general
variations. Add a block to set gcc_cv_initfini_array
unconditionally to "yes" for VxWorks7.
config/vx-common.h (VXWORKS_CC1_SPEC): New macro, empty string
by default. Update some comments.
config/vxworks.h (VXWORKS_EXTRA_LIBS_RTP): New macro, empty by
default, to be added the end of VXWORKS_LIBS_RTP.
(VXWORKS_LIBS_RTP): Replace hardcoded part by VXWORKS_BASE_LIBS_RTP
and append VXWORKS_EXTRA_LIBS_RTP, both of which specific ports may
redefine.
(VXWORKS_NET_LIBS_RTP): Account for VxWorks7 specificities.
(VXWORKS_CC1_SPEC): Common base definition, with VxWorks7 variation
to account for the now available TLS abilities.
(TARGET_LIBC_HAS_FUNCTION): Account for VxWorks7 abilities.
(VXWORKS_HAVE_TLS): Likewise.
Co-Authored-By: Jerome Lambourg <lambourg@adacore.com>
Co-Authored-By: Olivier Hainque <hainque@adacore.com>
From-SVN: r278247
If the statements in an SLP node aren't similar enough to be vectorised,
or aren't something the vectoriser has code to handle, the BB vectoriser
tries building the vector from scalars instead. This patch does the
same thing if we're able to build a viable-looking tree but fail later
during the analysis phase, e.g. because the target doesn't support a
particular vector operation.
This is needed to avoid regressions with a later patch.
2019-11-14 Richard Sandiford <richard.sandiford@arm.com>
gcc/
* tree-vect-slp.c (vect_contains_pattern_stmt_p): New function.
(vect_slp_convert_to_external): Likewise.
(vect_slp_analyze_node_operations): If analysis fails, try building
the node from scalars instead.
gcc/testsuite/
* gcc.dg/vect/bb-slp-div-2.c: New test.
From-SVN: r278246
This patch adds AArch64 patterns for converting between 64-bit and
128-bit integer vectors, and makes the vectoriser and expand pass
use them.
2019-11-14 Richard Sandiford <richard.sandiford@arm.com>
gcc/
* tree-cfg.c (verify_gimple_assign_unary): Handle conversions
between vector types.
* tree-vect-stmts.c (vectorizable_conversion): Extend the
non-widening and non-narrowing path to handle standard
conversion codes, if the target supports them.
* expr.c (convert_move): Try using the extend and truncate optabs
for vectors.
* optabs-tree.c (supportable_convert_operation): Likewise.
* config/aarch64/iterators.md (Vnarroqw): New iterator.
* config/aarch64/aarch64-simd.md (<optab><Vnarrowq><mode>2)
(trunc<mode><Vnarrowq>2): New patterns.
gcc/testsuite/
* gcc.dg/vect/bb-slp-pr69907.c: Do not expect BB vectorization
to fail for aarch64 targets.
* gcc.dg/vect/no-scevccp-outer-12.c: Expect the test to pass
on aarch64 targets.
* gcc.dg/vect/vect-double-reduc-5.c: Likewise.
* gcc.dg/vect/vect-outer-4e.c: Likewise.
* gcc.target/aarch64/vect_mixed_sizes_5.c: New test.
* gcc.target/aarch64/vect_mixed_sizes_6.c: Likewise.
* gcc.target/aarch64/vect_mixed_sizes_7.c: Likewise.
* gcc.target/aarch64/vect_mixed_sizes_8.c: Likewise.
* gcc.target/aarch64/vect_mixed_sizes_9.c: Likewise.
* gcc.target/aarch64/vect_mixed_sizes_10.c: Likewise.
* gcc.target/aarch64/vect_mixed_sizes_11.c: Likewise.
* gcc.target/aarch64/vect_mixed_sizes_12.c: Likewise.
* gcc.target/aarch64/vect_mixed_sizes_13.c: Likewise.
From-SVN: r278245
Although a previous patch allowed mixed vector sizes within a vector
region, we generally still required equal vector sizes within a vector
stmt. Specifically, vect_get_vector_types_for_stmt computes two vector
types: the vector type corresponding to STMT_VINFO_VECTYPE and the
vector type that determines the minimum vectorisation factor for the
stmt ("nunits_vectype"). It then required these two types to be
the same size.
There doesn't seem to be any need for that restriction though. AFAICT,
all vectorizable_* functions either do their own compatibility checks
or don't need to do them (because gimple guarantees that the scalar
types are compatible).
It should always be the case that nunits_vectype has at least as many
elements as the other vectype, but that's something we can assert for.
I couldn't resist a couple of other tweaks while there:
- there's no need to compute nunits_vectype if its element type is
the same as STMT_VINFO_VECTYPE's.
- it's useful to distinguish the nunits_vectype from the main vectype
in dump messages
- when reusing the existing STMT_VINFO_VECTYPE, it's useful to say so
in the dump, and say what the type is
2019-11-14 Richard Sandiford <richard.sandiford@arm.com>
gcc/
* tree-vect-stmts.c (vect_get_vector_types_for_stmt): Don't
require vectype and nunits_vectype to have the same size;
instead assert that nunits_vectype has at least as many
elements as vectype. Don't compute a separate nunits_vectype
if the scalar type is obviously the same as vectype's.
Tweak dump messages.
From-SVN: r278244
This patch makes the vectoriser try mixtures of 64-bit and 128-bit
vector modes on AArch64. It fixes some existing XFAILs and allows
kernel 24 from the Livermore Loops test to be vectorised (by using
a mixture of V2DF and V2SI).
2019-11-14 Richard Sandiford <richard.sandiford@arm.com>
gcc/
* config/aarch64/aarch64.c (aarch64_vectorize_related_mode): New
function.
(aarch64_autovectorize_vector_modes): Also add V4HImode and V2SImode.
(TARGET_VECTORIZE_RELATED_MODE): Define.
gcc/testsuite/
* gcc.dg/vect/vect-outer-4f.c: Expect the test to pass on aarch64
targets.
* gcc.dg/vect/vect-outer-4g.c: Likewise.
* gcc.dg/vect/vect-outer-4k.c: Likewise.
* gcc.dg/vect/vect-outer-4l.c: Likewise.
* gfortran.dg/vect/vect-8.f90: Expect kernel 24 to be vectorized
for aarch64.
* gcc.target/aarch64/vect_mixed_sizes_1.c: New test.
* gcc.target/aarch64/vect_mixed_sizes_2.c: Likewise.
* gcc.target/aarch64/vect_mixed_sizes_3.c: Likewise.
* gcc.target/aarch64/vect_mixed_sizes_4.c: Likewise.
From-SVN: r278243
A later patch makes the AArch64 port add four entries to
autovectorize_vector_modes. Each entry describes a different
vector mode assignment for vector code that mixes 8-bit, 16-bit,
32-bit and 64-bit elements. But if (as usual) the vector code has
fewer element sizes than that, we could end up trying the same
combination of vector modes multiple times. This patch adds a
check to prevent that.
2019-11-14 Richard Sandiford <richard.sandiford@arm.com>
gcc/
* tree-vectorizer.h (vec_info::mode_set): New typedef.
(vec_info::used_vector_mode): New member variable.
(vect_chooses_same_modes_p): Declare.
* tree-vect-stmts.c (get_vectype_for_scalar_type): Record each
chosen vector mode in vec_info::used_vector_mode.
(vect_chooses_same_modes_p): New function.
* tree-vect-loop.c (vect_analyze_loop): Use it to avoid trying
the same vector statements multiple times.
* tree-vect-slp.c (vect_slp_bb_region): Likewise.
From-SVN: r278242
After previous patches, it's now possible to make the vectoriser
support multiple vector sizes in the same vector region, using
related_vector_mode to pick the right vector mode for a given
element mode. No port yet takes advantage of this, but I have
a follow-on patch for AArch64.
This patch also seemed like a good opportunity to add some more dump
messages: one to make it clear which vector size/mode was being used
when analysis passed or failed, and another to say when we've decided
to skip a redundant vector size/mode.
2019-11-14 Richard Sandiford <richard.sandiford@arm.com>
gcc/
* machmode.h (opt_machine_mode::operator==): New function.
(opt_machine_mode::operator!=): Likewise.
* tree-vectorizer.h (vec_info::vector_mode): Update comment.
(get_related_vectype_for_scalar_type): Delete.
(get_vectype_for_scalar_type_and_size): Declare.
* tree-vect-slp.c (vect_slp_bb_region): Print dump messages to say
whether analysis passed or failed, and with what vector modes.
Use related_vector_mode to check whether trying a particular
vector mode would be redundant with the autodetected mode,
and print a dump message if we decide to skip it.
* tree-vect-loop.c (vect_analyze_loop): Likewise.
(vect_create_epilog_for_reduction): Use
get_related_vectype_for_scalar_type instead of
get_vectype_for_scalar_type_and_size.
* tree-vect-stmts.c (get_vectype_for_scalar_type_and_size): Replace
with...
(get_related_vectype_for_scalar_type): ...this new function.
Take a starting/"prevailing" vector mode rather than a vector size.
Take an optional nunits argument, with the same meaning as for
related_vector_mode. Use related_vector_mode when not
auto-detecting a mode, falling back to mode_for_vector if no
target mode exists.
(get_vectype_for_scalar_type): Update accordingly.
(get_same_sized_vectype): Likewise.
* tree-vectorizer.c (get_vec_alignment_for_array_type): Likewise.
From-SVN: r278240
As explained in the comment, vectorizable_call needs more work to
support mixtures of sizes. This avoids testsuite fallout for
later SVE patches.
2019-11-14 Richard Sandiford <richard.sandiford@arm.com>
gcc/
* tree-vect-stmts.c (vectorizable_call): Require the types
to have the same size.
From-SVN: r278239
Some callers of get_same_sized_vectype were dealing with operands that
are constant or defined externally, and so have no STMT_VINFO_VECTYPE
available. Under the current model, using get_same_sized_vectype for
that case is equivalent to using get_vectype_for_scalar_type, since
get_vectype_for_scalar_type always returns vectors of the same size,
once a size is fixed.
Using get_vectype_for_scalar_type is arguably more obvious though:
if we're using the same scalar type as we would for internal
definitions, we should use the same vector type too. (Constant and
external definitions sometimes let us change the original scalar type
to a "nicer" scalar type, but that isn't what's happening here.)
This is a prerequisite to supporting multiple vector sizes in the same
vec_info.
2019-11-14 Richard Sandiford <richard.sandiford@arm.com>
gcc/
* tree-vect-stmts.c (vectorizable_call): If an operand is
constant or external, use get_vectype_for_scalar_type
rather than get_same_sized_vectype to get its vector type.
(vectorizable_conversion, vectorizable_shift): Likewise.
(vectorizable_operation): Likewise.
From-SVN: r278238
This patch replaces vec_info::vector_size with vec_info::vector_mode,
but for now continues to use it as a way of specifying a single
vector size. This makes it easier for later patches to use
related_vector_mode instead.
2019-11-14 Richard Sandiford <richard.sandiford@arm.com>
gcc/
* tree-vectorizer.h (vec_info::vector_size): Replace with...
(vec_info::vector_mode): ...this new field.
* tree-vect-loop.c (vect_update_vf_for_slp): Update accordingly.
(vect_analyze_loop, vect_transform_loop): Likewise.
* tree-vect-loop-manip.c (vect_do_peeling): Likewise.
* tree-vect-slp.c (can_duplicate_and_interleave_p): Likewise.
(vect_make_slp_decision, vect_slp_bb_region): Likewise.
* tree-vect-stmts.c (get_vectype_for_scalar_type): Likewise.
* tree-vectorizer.c (try_vectorize_loop_1): Likewise.
gcc/testsuite/
* gcc.dg/vect/vect-tail-nomask-1.c: Update expected epilogue
vectorization message.
From-SVN: r278237
This is another patch in the series to remove the assumption that
all modes involved in vectorisation have to be the same size.
Rather than have the target provide a list of vector sizes,
it makes the target provide a list of vector "approaches",
with each approach represented by a mode.
A later patch will pass this mode to targetm.vectorize.related_mode
to get the vector mode for a given element mode. Until then, the modes
simply act as an alternative way of specifying the vector size.
2019-11-14 Richard Sandiford <richard.sandiford@arm.com>
gcc/
* target.h (vector_sizes, auto_vector_sizes): Delete.
(vector_modes, auto_vector_modes): New typedefs.
* target.def (autovectorize_vector_sizes): Replace with...
(autovectorize_vector_modes): ...this new hook.
* doc/tm.texi.in (TARGET_VECTORIZE_AUTOVECTORIZE_VECTOR_SIZES):
Replace with...
(TARGET_VECTORIZE_AUTOVECTORIZE_VECTOR_MODES): ...this new hook.
* doc/tm.texi: Regenerate.
* targhooks.h (default_autovectorize_vector_sizes): Delete.
(default_autovectorize_vector_modes): New function.
* targhooks.c (default_autovectorize_vector_sizes): Delete.
(default_autovectorize_vector_modes): New function.
* omp-general.c (omp_max_vf): Use autovectorize_vector_modes instead
of autovectorize_vector_sizes. Use the number of units in the mode
to calculate the maximum VF.
* omp-low.c (omp_clause_aligned_alignment): Use
autovectorize_vector_modes instead of autovectorize_vector_sizes.
Use a loop based on related_mode to iterate through all supported
vector modes for a given scalar mode.
* optabs-query.c (can_vec_mask_load_store_p): Use
autovectorize_vector_modes instead of autovectorize_vector_sizes.
* tree-vect-loop.c (vect_analyze_loop, vect_transform_loop): Likewise.
* tree-vect-slp.c (vect_slp_bb_region): Likewise.
* config/aarch64/aarch64.c (aarch64_autovectorize_vector_sizes):
Replace with...
(aarch64_autovectorize_vector_modes): ...this new function.
(TARGET_VECTORIZE_AUTOVECTORIZE_VECTOR_SIZES): Delete.
(TARGET_VECTORIZE_AUTOVECTORIZE_VECTOR_MODES): Define.
* config/arc/arc.c (arc_autovectorize_vector_sizes): Replace with...
(arc_autovectorize_vector_modes): ...this new function.
(TARGET_VECTORIZE_AUTOVECTORIZE_VECTOR_SIZES): Delete.
(TARGET_VECTORIZE_AUTOVECTORIZE_VECTOR_MODES): Define.
* config/arm/arm.c (arm_autovectorize_vector_sizes): Replace with...
(arm_autovectorize_vector_modes): ...this new function.
(TARGET_VECTORIZE_AUTOVECTORIZE_VECTOR_SIZES): Delete.
(TARGET_VECTORIZE_AUTOVECTORIZE_VECTOR_MODES): Define.
* config/i386/i386.c (ix86_autovectorize_vector_sizes): Replace with...
(ix86_autovectorize_vector_modes): ...this new function.
(TARGET_VECTORIZE_AUTOVECTORIZE_VECTOR_SIZES): Delete.
(TARGET_VECTORIZE_AUTOVECTORIZE_VECTOR_MODES): Define.
* config/mips/mips.c (mips_autovectorize_vector_sizes): Replace with...
(mips_autovectorize_vector_modes): ...this new function.
(TARGET_VECTORIZE_AUTOVECTORIZE_VECTOR_SIZES): Delete.
(TARGET_VECTORIZE_AUTOVECTORIZE_VECTOR_MODES): Define.
From-SVN: r278236
The validation phase of vectorizable_shift used TYPE_MODE to check
whether the shift amount vector was compatible with the shifted vector:
if ((op1_vectype == NULL_TREE
|| TYPE_MODE (op1_vectype) != TYPE_MODE (vectype))
&& (!slp_node
|| SLP_TREE_DEF_TYPE
(SLP_TREE_CHILDREN (slp_node)[1]) != vect_constant_def))
But the generation phase was stricter and required the element types to
be equivalent:
&& !useless_type_conversion_p (TREE_TYPE (vectype),
TREE_TYPE (op1)))
This difference led to an ICE with a later patch.
The first condition seems a bit too lax given that the function
supports vect_worthwhile_without_simd_p, where two different vector
types could have the same integer mode. But it seems too strict
to reject signed shifts by unsigned amounts or unsigned shifts by
signed amounts; verify_gimple_assign_binary is happy with those.
This patch therefore goes for a middle ground of checking both TYPE_MODE
and TYPE_VECTOR_SUBPARTS, using the same condition in both places.
2019-11-14 Richard Sandiford <richard.sandiford@arm.com>
gcc/
* tree-vect-stmts.c (vectorizable_shift): Check the number
of vector elements as well as the type mode when deciding
whether an op1_vectype is compatible. Reuse the result of
this check when generating vector statements.
From-SVN: r278235
Except for one case, get_vectype_for_scalar_type_and_size calculates
what the vector mode should be and then calls build_vector_type,
which recomputes the mode from scratch. This patch makes it use
build_vector_type_for_mode instead.
The exception mentioned above is when preferred_simd_mode returns
an integer mode, which it does if no appropriate vector mode exists.
The integer mode in question is usually word_mode, although epiphany
can return a doubleword mode in some cases.
There's no guarantee that this integer mode is appropriate, since for
example the scalar type could be a float. The traditional behaviour is
therefore to use the integer mode to determine a size only, and leave
mode_for_vector to pick the TYPE_MODE. (Note that it can actually end
up picking a vector mode if the target defines a disabled vector mode.
We therefore still need to check TYPE_MODE after building the type.)
2019-11-14 Richard Sandiford <richard.sandiford@arm.com>
gcc/
* tree-vect-stmts.c (get_vectype_for_scalar_type_and_size): If
targetm.vectorize.preferred_simd_mode returns an integer mode,
use mode_for_vector to decide what the vector type's mode
should actually be. Use build_vector_type_for_mode instead
of build_vector_type.
From-SVN: r278234
This patch passes the data vector mode to get_mask_mode, rather than its
size and nunits. This is a bit simpler and allows targets to distinguish
between modes that happen to have the same size and number of elements.
2019-11-14 Richard Sandiford <richard.sandiford@arm.com>
gcc/
* target.def (get_mask_mode): Take a vector mode itself as argument,
instead of properties about the vector mode.
* doc/tm.texi: Regenerate.
* targhooks.h (default_get_mask_mode): Update to reflect new
get_mode_mask interface.
* targhooks.c (default_get_mask_mode): Likewise. Use
related_int_vector_mode.
* optabs-query.c (can_vec_mask_load_store_p): Update call
to get_mask_mode.
* tree-vect-stmts.c (check_load_store_masking): Likewise, checking
first that the original mode really is a vector.
* tree.c (build_truth_vector_type_for): Likewise.
* config/aarch64/aarch64.c (aarch64_get_mask_mode): Update for new
get_mode_mask interface.
(aarch64_expand_sve_vcond): Update call accordingly.
* config/gcn/gcn.c (gcn_vectorize_get_mask_mode): Update for new
get_mode_mask interface.
* config/i386/i386.c (ix86_get_mask_mode): Likewise.
From-SVN: r278233
build_same_sized_truth_vector_type was confusingly named, since for
SVE and AVX512 the returned vector isn't the same byte size (although
it does have the same number of elements). What it really returns
is the "truth" vector type for a given data vector type.
The more general truth_type_for provides the same thing when passed
a vector and IMO has a more descriptive name, so this patch replaces
all uses of build_same_sized_truth_vector_type with that. It does
the same for a call to build_truth_vector_type, leaving truth_type_for
itself as the only remaining caller.
It's then more natural to pass build_truth_vector_type the original
vector type rather than its size and nunits, especially since the
given size isn't the size of the returned vector. This in turn allows
a future patch to simplify the interface of get_mask_mode. Doing this
also fixes a bug in which truth_type_for would pass a size of zero for
BLKmode vector types.
2019-11-14 Richard Sandiford <richard.sandiford@arm.com>
gcc/
* tree.h (build_truth_vector_type): Delete.
(build_same_sized_truth_vector_type): Likewise.
* tree.c (build_truth_vector_type): Rename to...
(build_truth_vector_type_for): ...this. Make static and take
a vector type as argument.
(truth_type_for): Update accordingly.
(build_same_sized_truth_vector_type): Delete.
* tree-vect-generic.c (expand_vector_divmod): Use truth_type_for
instead of build_same_sized_truth_vector_type.
* tree-vect-loop.c (vect_create_epilog_for_reduction): Likewise.
(vect_record_loop_mask, vect_get_loop_mask): Likewise.
* tree-vect-patterns.c (build_mask_conversion): Likeise.
* tree-vect-slp.c (vect_get_constant_vectors): Likewise.
* tree-vect-stmts.c (vect_get_vec_def_for_operand): Likewise.
(vect_build_gather_load_calls, vectorizable_call): Likewise.
(scan_store_can_perm_p, vectorizable_scan_store): Likewise.
(vectorizable_store, vectorizable_condition): Likewise.
(get_mask_type_for_scalar_type, get_same_sized_vectype): Likewise.
(vect_get_mask_type_for_stmt): Use truth_type_for instead of
build_truth_vector_type.
* config/aarch64/aarch64-sve-builtins.cc (gimple_folder::convert_pred):
Use truth_type_for instead of build_same_sized_truth_vector_type.
* config/rs6000/rs6000-call.c (fold_build_vec_cmp): Likewise.
gcc/c/
* c-typeck.c (build_conditional_expr): Use truth_type_for instead
of build_same_sized_truth_vector_type.
(build_vec_cmp): Likewise.
gcc/cp/
* call.c (build_conditional_expr_1): Use truth_type_for instead
of build_same_sized_truth_vector_type.
* typeck.c (build_vec_cmp): Likewise.
gcc/d/
* d-codegen.cc (build_boolop): Use truth_type_for instead of
build_same_sized_truth_vector_type.
From-SVN: r278232
Callers of vect_halve_mask_nunits and vect_double_mask_nunits
already know what mode the resulting vector type should have,
so we might as well create the vector type directly with that mode,
just like build_vector_type_for_mode lets us build normal vectors
with a known mode. This avoids the current awkwardness of having
to recompute the mode starting from vec_info::vector_size, which
hard-codes the assumption that all vectors have to be the same size.
A later patch gets rid of build_truth_vector_type and
build_same_sized_truth_vector_type, so the net effect of the
series is to reduce the number of type functions by one.
2019-11-14 Richard Sandiford <richard.sandiford@arm.com>
gcc/
* tree.h (build_truth_vector_type_for_mode): Declare.
* tree.c (build_truth_vector_type_for_mode): New function,
split out from...
(build_truth_vector_type): ...here.
(build_opaque_vector_type): Fix head comment.
* tree-vectorizer.h (supportable_narrowing_operation): Remove
vec_info parameter.
(vect_halve_mask_nunits): Replace vec_info parameter with the
mode of the new vector.
(vect_double_mask_nunits): Likewise.
* tree-vect-loop.c (vect_halve_mask_nunits): Likewise.
(vect_double_mask_nunits): Likewise.
* tree-vect-loop-manip.c: Include insn-config.h, rtl.h and recog.h.
(vect_maybe_permute_loop_masks): Remove vinfo parameter. Update call
to vect_halve_mask_nunits, getting the required mode from the unpack
patterns.
(vect_set_loop_condition_masked): Update call accordingly.
* tree-vect-stmts.c (supportable_narrowing_operation): Remove vec_info
parameter and update call to vect_double_mask_nunits.
(vectorizable_conversion): Update call accordingly.
(simple_integer_narrowing): Likewise. Remove vec_info parameter.
(vectorizable_call): Update call accordingly.
(supportable_widening_operation): Update call to
vect_halve_mask_nunits.
* config/aarch64/aarch64-sve-builtins.cc (register_builtin_types):
Use build_truth_vector_type_mode instead of build_truth_vector_type.
From-SVN: r278231
mode_for_int_vector, like mode_for_vector, can sometimes return
an integer mode or an unsupported vector mode. But no callers
are interested in that case, and only want supported vector modes.
This patch therefore replaces mode_for_int_vector with
related_int_vector_mode, which gives the target a chance to pick
its preferred vector mode for the given element mode and size.
2019-11-14 Richard Sandiford <richard.sandiford@arm.com>
gcc/
* machmode.h (mode_for_int_vector): Delete.
(related_int_vector_mode): Declare.
* stor-layout.c (mode_for_int_vector): Delete.
(related_int_vector_mode): New function.
* optabs.c (expand_vec_perm_1): Use related_int_vector_mode
instead of mode_for_int_vector.
(expand_vec_perm_const): Likewise.
* config/aarch64/aarch64.c (aarch64_emit_approx_sqrt): Likewise.
(aarch64_evpc_sve_tbl): Likewise.
* config/s390/s390.c (s390_expand_vec_compare_cc): Likewise.
(s390_expand_vcond): Likewise.
From-SVN: r278230
This patch is the first of a series that tries to remove two
assumptions:
(1) that all vectors involved in vectorisation must be the same size
(2) that there is only one vector mode for a given element mode and
number of elements
Relaxing (1) helps with targets that support multiple vector sizes or
that require the number of elements to stay the same. E.g. if we're
vectorising code that operates on narrow and wide elements, and the
narrow elements use 64-bit vectors, then on AArch64 it would normally
be better to use 128-bit vectors rather than pairs of 64-bit vectors
for the wide elements.
Relaxing (2) makes it possible for -msve-vector-bits=128 to produce
fixed-length code for SVE. It also allows unpacked/half-size SVE
vectors to work with -msve-vector-bits=256.
The patch adds a new hook that targets can use to control how we
move from one vector mode to another. The hook takes a starting vector
mode, a new element mode, and (optionally) a new number of elements.
The flexibility needed for (1) comes in when the number of elements
isn't specified.
All callers in this patch specify the number of elements, but a later
vectoriser patch doesn't.
2019-11-14 Richard Sandiford <richard.sandiford@arm.com>
gcc/
* target.def (related_mode): New hook.
* doc/tm.texi.in (TARGET_VECTORIZE_RELATED_MODE): New hook.
* doc/tm.texi: Regenerate.
* targhooks.h (default_vectorize_related_mode): Declare.
* targhooks.c (default_vectorize_related_mode): New function.
* machmode.h (related_vector_mode): Declare.
* stor-layout.c (related_vector_mode): New function.
* expmed.c (extract_bit_field_1): Use it instead of mode_for_vector.
* optabs-query.c (qimode_for_vec_perm): Likewise.
* tree-vect-stmts.c (get_group_load_store_type): Likewise.
(vectorizable_store, vectorizable_load): Likewise
From-SVN: r278229
Inspired by the tests in gcc.target/i386. Testing code generation,
diagnostics, and execution.
* gcc.target/aarch64/asm-flag-1.c: New test.
* gcc.target/aarch64/asm-flag-3.c: New test.
* gcc.target/aarch64/asm-flag-5.c: New test.
* gcc.target/aarch64/asm-flag-6.c: New test.
From-SVN: r278228
Inspired by the tests in gcc.target/i386. Testing code generation,
diagnostics, and execution.
* gcc.target/arm/asm-flag-1.c: New test.
* gcc.target/arm/asm-flag-3.c: New test.
* gcc.target/arm/asm-flag-5.c: New test.
* gcc.target/arm/asm-flag-6.c: New test.
From-SVN: r278227
Since all but a couple of lines is shared between the two targets,
enable them both at once.
* config/arm/aarch-common-protos.h (arm_md_asm_adjust): Declare.
* config/arm/aarch-common.c (arm_md_asm_adjust): New.
* config/arm/arm-c.c (arm_cpu_builtins): Define
__GCC_ASM_FLAG_OUTPUTS__.
* config/arm/arm.c (TARGET_MD_ASM_ADJUST): New.
* config/aarch64/aarch64-c.c (aarch64_define_unconditional_macros):
Define __GCC_ASM_FLAG_OUTPUTS__.
* config/aarch64/aarch64.c (TARGET_MD_ASM_ADJUST): New.
* doc/extend.texi (FlagOutputOperands): Add documentation
for ARM and AArch64.
From-SVN: r278226
The existing definition using register class CC_REG does not
work because CC_REGNUM does not support normal modes, and so
fails to match register_operand. Use a non-register constraint
and the cc_register predicate instead.
* config/arm/constraints.md (c): Use cc_register predicate.
From-SVN: r278224
* ipa-cp.c (ipa_vr_operation_and_type_effects): Move up in file.
(ipa_value_range_from_jfunc): New function.
* ipa-fnsummary.c (evaluate_conditions_for_known_args): Add
known_value_ranges parameter; use it to evalulate conditions.
(evaluate_properties_for_edge): Compute known value ranges.
(ipa_fn_summary_t::duplicate): Update use of
evaluate_conditions_for_known_args.
(estimate_ipcp_clone_size_and_time): Likewise.
(ipa_merge_fn_summary_after_inlining): Likewise.
* ipa-prop.h (ipa_value_range_from_jfunc): Declare.
* gcc.dg/ipa/inline-9.c: New testcase.
From-SVN: r278220
2019-11-14 Martin Liska <mliska@suse.cz>
* ipa-inline.c (want_inline_small_function_p): Use
CIF_MAX_INLINE_INSNS_AUTO_LIMIT instead
of CIF_MAX_INLINE_INSNS_SINGLE_O2_LIMIT.
From-SVN: r278219
2019-11-14 Martin Liska <mliska@suse.cz>
* ipa-cp.c (devirtualization_time_bonus): Use opt_for_fn
of a callee to get value of the param.
* ipa-inline.c (inline_insns_auto): Use proper
opt_for_fn.
* opts.c (maybe_default_option): Do not overwrite param
value if optimization level does not match. Note that
params usually have default value set via Init() keyword.
* params.opt: Remove -param=max-inline-insns-auto-O2.
* cif-code.def (MAX_INLINE_INSNS_AUTO_O2_LIMIT): Remove.
* doc/invoke.texi: Remove documentation of
max-inline-insns-auto-O2.
2019-11-14 Martin Liska <mliska@suse.cz>
* c-c++-common/asan/memcmp-1.c: Update expected backtrace.
From-SVN: r278218
2019-11-14 Martin Liska <mliska@suse.cz>
* tree-switch-conversion.c (switch_conversion::switch_conversion):
Do not initialize m_other_count.
(switch_conversion::collect): Do not count m_default_count and
m_other_count as we use frequencies for edges.
* tree-switch-conversion.h: Remove m_default_count and m_other_count.
From-SVN: r278217
2019-11-14 Martin Liska <mliska@suse.cz>
PR target/92389
* config/i386/i386.h: Add PTA_AVX512VPOPCNTDQ to
PTA_ICELAKE_CLIENT which is later interited by
PTA_ICELAKE_SERVER and PTA_TIGERLAKE.
From-SVN: r278210
2019-11-14 Martin Liska <mliska@suse.cz>
* ipa-icf-gimple.c (func_checker::hash_operand): Improve
func_checker::hash_operand by handling of FIELD_DECLs.
From-SVN: r278208
2019-11-14 Martin Liska <mliska@suse.cz>
* ipa-icf-gimple.h (func_checker::func_checker): Add
default constructor.
* ipa-icf.c (sem_function::init): Make operand_equal_p
and hash_operand public.
(sem_item::add_expr): Remove.
(sem_item::add_type): Remove.
(sem_function::hash_stmt): Use m_checker for hashing
of GIMPLE statements.
(sem_function::parse): Init with checker.
(sem_variable::parse): Pass NULL as checker.
(sem_item_optimizer::parse_funcs_and_vars):
Pass checker to ::parse function.
(sem_item_optimizer::parse_nonsingleton_classes): Likewise.
(sem_variable::parse): New function.
(sem_variable::get_hash): Only return computed hash value.
(sem_variable::init): Initialize hash of a variable.
* ipa-icf.h: Remove add_expr, add_type and add func_checker
to couple of functions as a new argument.
From-SVN: r278207
* c-parser.c (c_parser_omp_context_selector): Don't require score
argument to fit into shwi, just to be INTEGER_CST. Diagnose
negative score.
* parser.c (cp_parser_omp_context_selector): Don't require score
argument to fit into shwi, just to be INTEGER_CST. Diagnose
negative score.
* pt.c (tsubst_attribute): Likewise.
* c-c++-common/gomp/declare-variant-2.c: Add test for non-integral
score and for negative score.
* c-c++-common/gomp/declare-variant-3.c: Add test for zero score.
* g++.dg/gomp/declare-variant-8.C: Add test for negative and zero
scores.
From-SVN: r278204
* c-omp.c (c_omp_check_context_selector): Add nvidia to the list of
valid vendors.
* c-c++-common/gomp/declare-variant-3.c: Add testcase for vendor nvidia.
From-SVN: r278203
* omp-general.c (omp_context_name_list_prop): New function.
(omp_context_selector_matches): Use it. Return 0 if it returns
NULL.
(omp_context_selector_props_compare): Allow equivalency of an
identifier and a string literal containing no embedded zeros.
c-family/
* c-omp.c (c_omp_check_context_selector): Handle name lists
containing string literals. Don't diagnose atomic_default_mem_order
with multiple props.
c/
* c-parser.c (c_parser_omp_context_selector): Rename
CTX_PROPERTY_IDLIST to CTX_PROPERTY_NAME_LIST, add CTX_PROPERTY_ID.
Use CTX_PROPERTY_ID for atomic_default_mem_order, only allow a single
identifier in that. For CTX_PROPERTY_NAME_LIST, allow identifiers
and string literals.
cp/
* parser.c (cp_parser_omp_context_selector): Rename
CTX_PROPERTY_IDLIST to CTX_PROPERTY_NAME_LIST, add CTX_PROPERTY_ID.
Use CTX_PROPERTY_ID for atomic_default_mem_order, only allow a single
identifier in that. For CTX_PROPERTY_NAME_LIST, allow identifiers
and string literals.
* pt.c (tsubst_attribute): Fix up STRING_CST handling if allow_string.
testsuite/
* c-c++-common/gomp/declare-variant-2.c: Adjust expected diagnostics,
add a test for atomic_default_mem_order with a string literal.
* c-c++-common/gomp/declare-variant-3.c: Use string literal props
in a few random places, add a few string literal prop related tests.
* c-c++-common/gomp/declare-variant-8.c: Likewise.
* c-c++-common/gomp/declare-variant-9.c: Use string literal props
in a few random places.
* c-c++-common/gomp/declare-variant-10.c: Likewise.
* c-c++-common/gomp/declare-variant-11.c: Likewise.
* c-c++-common/gomp/declare-variant-12.c: Likewise.
* g++.dg/gomp/declare-variant-7.C: Likewise.
From-SVN: r278202
This removes the dependency on m_max_pairs from the selftests, which has
the ultimate effect of allowing us to put the tests in the selftest
namespace as was the original plan.
From-SVN: r278200
The recent vectorization cost adjustment on load leads
the profitable min iteration count to change from 19 to 12.
The case happens to hit the threshold. This patch is to
adjust the loop bound from 16 to 14.
gcc/testsuite/ChangeLog
2019-11-14 Kewen Lin <linkw@gcc.gnu.org>
PR target/92464
* gcc.dg/vect/costmodel/ppc/costmodel-vect-76b.c: Adjust
loop bound due to load cost adjustment.
From-SVN: r278195
This patch adds support for the C2x [[]] attribute syntax to the C
front end. Support is only added for C at this point, not for
Objective-C; I intend to add the unbounded lookahead required to
support it for Objective-C in a followup patch, but maybe not in
development stage 1.
The syntax is supported in all relevant places where the standard says
it is supported, but support is not added for the individual
attributes specified in C2x (all of which are optional to support). I
expect to add support for some of them in followup patches; all except
nodiscard can be mapped directly to the semantics of an existing GNU
attribute (subject to extra checks for invalid usages such as the same
attribute being used more than once inside a single [[]]), and the
fallthrough attribute already works after this patch because of
existing special-case code handling it (but without some of the checks
for invalid usage being present).
Note that the four functions c_token_starts_declspecs,
c_token_starts_declaration, c_parser_next_token_starts_declspecs and
c_parser_next_tokens_start_declaration do *not* accept "[[". This is
analogous with the handling of __extension__: both cases have the
property that they can start either a declaration or some other
statements and so need an unbounded number of tokens to be parsed in
the caller before it can find out what kind of syntactic construct
follows. Note also that, while I updated all places calling those
functions for standard C syntax to handle attributes if applicable, I
did not do anything regarding calls to such functions for OpenMP or
OpenACC constructs. Thus, if there are such constructs using such
functions where "[[" *should* be accepted as a possible start to a
declaration, the code for parsing those constructs should be updated
accordingly.
Although all cases of the syntax are handled, and attributes applied
to the constructs the standard says they should be (with less laxity
than there is for GNU attributes to allow an attribute applied to one
construct to be moved automatically to another one), there is a major
limitation in the existing language-independent code in attribs.c
preventing most cases of type attributes from working. The following
code has been present with minor changes since the first support for
[[]] attributes for C++ was added:
if (TYPE_P (*node)
&& cxx11_attr_p
&& !(flags & ATTR_FLAG_TYPE_IN_PLACE))
{
/* This is a c++11 attribute that appertains to a
type-specifier, outside of the definition of, a class
type. Ignore it. */
auto_diagnostic_group d;
if (warning (OPT_Wattributes, "attribute ignored"))
inform (input_location,
"an attribute that appertains to a type-specifier "
"is ignored");
continue;
}
I see no justification for this in general for either C or C++ and so
propose to remove or restrict it in a followup bug-fix patch. Both C
and C++ are clear about attributes in certain places (at the end of
declaration specifiers, or after function or array declarators)
appertaining to a specific type (and explicitly say, in the case of
attributes at the end of declaration specifiers, that they only apply
for that particular use of that type, not for subsequent uses of the
same type without the attributes). Thus it seems clear to me that,
for example,
int [[gnu::mode(DI)]] x;
ought to be accepted as an analogue in [[]] syntax for
int __attribute__((mode(DI))) x;
(or strictly as an analogue for a version of that with extra
parentheses to make the GNU attribute bind properly to the type rather
than being automatically moved from the declaration to the type).
There are certain cases where an attribute *does* only make sense for
the definition of a type (e.g. "packed" on structure types), but those
should already be handled in the individual attribute handlers (such
as handle_packed_attribute, which already has code to deal with that
issue). So my inclination is that the above-quoted check in attribs.c
should simply be removed, but failing that it should be restricted to
structure and union types (and such a change would be a bug-fix).
That would then allow various cases of [[]] attributes on types to
work properly.
Bootstrapped with no regressions on x86_64-pc-linux-gnu.
gcc/c:
* c-tree.h (enum c_typespec_kind): Add ctsk_tagref_attrs and
ctsk_tagfirstref_attrs.
(struct c_declspecs): Update description of attrs. Add
postfix_attrs and non_std_attrs_seen_p. Increase size of
typespec_kind bit-field.
(c_warn_unused_attributes): New declaration.
(parser_xref_tag): Update prototype.
* c-decl.c (c_warn_unused_attributes): New function.
(shadow_tag_warned): Handle ctsk_tagfirstref_attrs and
ctsk_tagref_attrs. Handle attribute declarations.
(check_compound_literal_type): Handle ctsk_tagfirstref_attrs.
(grokdeclarator): Handle standard attributes.
(parser_xref_tag): Add arguments have_std_attrs and attrs. Apply
attributes to incomplete type reference.
(xref_tag): Update call to parser_xref_tag.
(declspecs_add_addrspace, declspecs_add_type)
(declspecs_add_scspec, declspecs_add_attrs): Set
non_std_attrs_seen_p.
(finish_declspecs): Apply postfix standard attributes to type.
* c-parser.c (c_token_starts_declspecs)
(c_token_starts_declaration, c_parser_next_token_starts_declspecs)
(c_parser_next_tokens_start_declaration): Update comments.
(c_parser_consume_token, c_parser_consume_pragma): Handle moving
parser->tokens[2] to parser->tokens[1].
(c_parser_nth_token_starts_std_attributes)
(c_parser_std_attribute_specifier_sequence): New functions.
(c_parser_declaration_or_fndef): Add arguments have_attrs and
attrs. All callers changed. Handle standard attributes.
(c_parser_parms_declarator, c_parser_parms_list_declarator)
(c_parser_parameter_declaration): Add argument have_gnu_attrs.
All callers changed.
(c_parser_declspecs): Add arguments start_std_attr_ok and
end_std_attr_ok. All callers changed. Handle standard
attributes.
(c_parser_enum_specifier, c_parser_struct_or_union_specifier)
(c_parser_direct_declarator, c_parser_direct_declarator_inner)
(c_parser_compound_statement_nostart, c_parser_all_labels)
(c_parser_label, c_parser_statement, c_parser_for_statement):
Handle standard attributes.
* c-parser.h (c_parser_declspecs): Update prototype.
* gimple-parser.c (c_parser_gimple_declaration): Update call to
c_parser_declspecs.
gcc/testsuite:
* gcc.dg/c2x-attr-fallthrough-1.c, gcc.dg/c2x-attr-syntax-1.c,
gcc.dg/c2x-attr-syntax-2.c, gcc.dg/c2x-attr-syntax-3.c,
gcc.dg/gnu2x-attr-syntax-1.c, gcc.dg/gnu2x-attr-syntax-2.c,
gcc.dg/gnu2x-attrs-1.c: New tests.
From-SVN: r278194
2019-11-14 Feng Xue <fxue@os.amperecomputing.com>
PR ipa/91682
* ipa-prop.h (jump_func_type): New value IPA_JF_LOAD_AGG.
(ipa_load_agg_data, ipa_agg_value, ipa_agg_value_set): New structs.
(ipa_agg_jf_item): Add new field jftype and type, redefine field value.
(ipa_agg_jump_function): Remove member function equal_to.
(ipa_agg_jump_function_p): Remove typedef.
(ipa_copy_agg_values, ipa_release_agg_values): New functions.
* ipa-prop.c (ipa_print_node_jump_functions_for_edge): Dump
information for aggregate jump function.
(get_ssa_def_if_simple_copy): Add new parameter rhs_stmt to
record last definition statement.
(load_from_unmodified_param_or_agg): New function.
(ipa_known_agg_contents_list): Add new field type and value, remove
field constant.
(build_agg_jump_func_from_list): Rename parameter const_count to
value_count, build aggregate jump function from ipa_load_agg_data.
(analyze_agg_content_value): New function.
(extract_mem_content): Analyze memory store assignment to prepare
information for aggregate jump function generation.
(determine_known_aggregate_parts): Add new parameter fbi, remove
parameter aa_walk_budeget_p.
(update_jump_functions_after_inlining): Update aggregate jump function.
(ipa_find_agg_cst_for_param): Change type of parameter agg.
(try_make_edge_direct_simple_call): Add new parameter new_root.
(try_make_edge_direct_virtual_call): Add new parameter new_root and
new_root_info.
(update_indirect_edges_after_inlining): Pass new argument to
try_make_edge_direct_simple_call and try_make_edge_direct_virtual_call.
(ipa_write_jump_function): Write aggregate jump function to file.
(ipa_read_jump_function): Read aggregate jump function from file.
(ipa_agg_value::equal_to): Migrate from ipa_agg_jf_item::equal_to.
* ipa-cp.c (ipa_get_jf_arith_result): New function.
(ipa_agg_value_from_node): Likewise.
(ipa_agg_value_set_from_jfunc): Likewise.
(propagate_vals_across_arith_jfunc): Likewise.
(propagate_aggregate_lattice): Likewise.
(ipa_get_jf_pass_through_result): Call ipa_get_jf_arith_result.
(propagate_vals_across_pass_through): Call
propagate_vals_across_arith_jfunc.
(get_clone_agg_value): Move forward.
(propagate_aggs_across_jump_function): Handle value propagation for
aggregate jump function.
(agg_jmp_p_vec_for_t_vec): Remove.
(context_independent_aggregate_values): Replace vec<ipa_agg_jf_item>
with vec<ipa_agg_value>.
(copy_plats_to_inter, intersect_with_plats): Likewise.
(agg_replacements_to_vector, intersect_with_agg_replacements): Likewise.
(intersect_aggregate_with_edge): Likewise.
(find_aggregate_values_for_callers_subset): Likewise.
(cgraph_edge_brings_all_agg_vals_for_node): Likewise.
(estimate_local_effects): Replace vec<ipa_agg_jump_function> and
vec<ipa_agg_jump_function_p> with vec<ipa_agg_value_set>.
(gather_context_independent_values): Likewise.
(perform_estimation_of_a_value, decide_whether_version_node): Likewise.
* ipa-fnsummary.c (evaluate_conditions_for_known_args): Replace
vec<ipa_agg_jump_function_p> with vec<ipa_agg_value_set>.
(evaluate_properties_for_edge): Likewise.
(estimate_edge_devirt_benefit): Likewise.
(estimate_edge_size_and_time): Likewise.
(estimate_calls_size_and_time): Likewise.
(ipa_call_context::ipa_call_context): Likewise.
(estimate_ipcp_clone_size_and_time): Likewise.
* ipa-fnsummary.h (ipa_call_context): Replace
vec<ipa_agg_jump_function_p> with vec<ipa_agg_value_set>.
* ipa-inline-analysis.c (do_estimate_edge_time): Replace
vec<ipa_agg_jump_function_p> with vec<ipa_agg_value_set>.
(do_estimate_edge_size): Likewise.
(do_estimate_edge_hints): Likewise.
2019-11-14 Feng Xue <fxue@os.amperecomputing.com>
PR ipa/91682
* gcc.dg/ipa/ipcp-agg-10.c: Change dg-scan string.
* gcc.dg/ipa/ipcp-agg-11.c: New test.
From-SVN: r278193
PR ipa/92421
* g++.dg/torture/pr92421.C: Add -Wno-return-type to
dg-additional-options. Avoid -Wwrite-string warnings, most of
-Wreturn-type warnings, define bf ctor. Use struct instead of class
with public: at the start.
From-SVN: r278192