Ed Schonberg [Fri, 25 Jun 2021 13:54:17 +0000 (06:54 -0700)]
[Ada] Spurious link error with child unit and different Assertion modes.
gcc/ada/
* exp_util.ads (Force_Evaluation): Add formal parameter
Discr_Number, to indicate discriminant expression for which an
external name must be created.
(Remove_Side_Effects): Ditto.
* exp_util.adb (Force_Evaluation): Call Remove_Side_Effects with
added parameter.
(Remove_Side_Effects, Build_Temporary): If Discr_Number is
positive, create an external name with suffix DISCR and the
given discriminant number, analogous to what is done for
temporaries for array type bounds.
* sem_ch3.adb (Process_Discriminant_Expressions): If the
constraint is for an object or component declaration and the
corresponding entity may be visible in another unit, invoke
Force_Evaluation with the new parameter.
Gary Dismukes [Tue, 22 Jun 2021 04:47:00 +0000 (00:47 -0400)]
[Ada] Implementation of Preelaborable_Initialization attribute for AI12-0409
gcc/ada/
* exp_attr.adb (Expand_N_Attribute_Reference): Fold
Preelaborable_Initialization attribute in cases where it hasn't
been folded by the analyzer.
* exp_disp.adb (Original_View_In_Visible_Part): This function is
removed and moved to sem_util.adb.
* sem_attr.adb (Attribute_22): Add
Attribute_Preelaborable_Initialization as an Ada 2022 attribute.
(Analyze_Attribute, Attribute_Preelaborable_Initialization):
Check that the prefix of the attribute is either a formal
private or derived type, or a composite type declared within the
visible part of a package or generic package.
(Eval_Attribute): Perform folding of
Preelaborable_Initialization attribute based on
Has_Preelaborable_Initialization applied to the prefix type.
* sem_ch3.adb (Resolve_Aspects): Add specialized code for
Preelaborable_Initialization used at the end of a package
visible part for setting Known_To_Have_Preelab_Init on types
that are specified with True or that have a conjunction of one
or more P_I attributes applied to formal types.
* sem_ch7.adb (Analyze_Package_Specification): On call to
Has_Preelaborable_Initialization, pass True for new formal
Formal_Types_Have_Preelab_Init, so that error checking treats
subcomponents that are declared within types in generics as
having preelaborable initialization when the subcomponents are
of formal types.
* sem_ch13.adb (Analyze_Aspects_At_Freeze_Point): Add test for
P_I to prevent calling Make_Pragma_From_Boolean_Aspect, since
this aspect is handled specially and the
Known_To_Have_Preelab_Init flag will get set on types that have
the aspect by other means.
(Analyze_Aspect_Specifications.Analyze_One_Aspect): Add test for
Aspect_Preelaborable_Initialization for allowing the aspect to
be specified on formal type declarations.
(Is_Operational_Item): Treat Attribute_Put_Image as an
operational attribute. The need for this was encountered while
working on these changes.
* sem_util.ads (Has_Preelaborable_Initialization): Add
Formal_Types_Have_Preelab_Init as a new formal parameter that
defaults to False.
(Is_Conjunction_Of_Formal_Preelab_Init_Attributes): New
function.
(Original_View_In_Visible_Part): Moved here from exp_disp.adb,
so it can be called by Analyze_Attribute.
* sem_util.adb (Has_Preelaborable_Initialization): Return True
for formal private and derived types when new formal
Formal_Types_Have_Preelab_Init is True, and pass along the
Formal_Types_Have_Preelab_Init flag in the array component case.
(Check_Components): Pass along Formal_Types_Have_Preelab_Init
flag on call to Has_Preelaborable_Initialization.
(Is_Conjunction_Of_Formal_Preelab_Init_Attributes): New function
that returns True when passed an expression that includes one or
more attributes for Preelaborable_Initialization applied to
prefixes that denote formal types.
(Is_Formal_Preelab_Init_Attribute): New utility function nested
within Is_Conjunction_Of_Formal_Preelab_Init_Attributes that
determines whether a node is a P_I attribute applied to a
generic formal type.
(Original_View_In_Visible_Part): Moved here from exp_util.adb,
so it can be called by Analyze_Attribute.
* snames.ads-tmpl: Add note near the start of spec giving
details about what needs to be done when adding a name that
corresponds to both an attribute and a pragma. Delete existing
occurrence of Name_Preelaborable_Initialization, and add a note
comment in the list of Name_* constants at that place,
indicating that it's included in type Pragma_Id, etc., echoing
other such comments for names that are both an attribute and a
pragma. Insert Name_Preelaborable_Initialization in the
alphabetized set of Name_* constants corresponding to
attributes (between First_Attribute_Name and
Last_Attribute_Name).
(type Attribute_Id): Add new literal
Attribute_Preelaborable_Initialization.
(type Pragma_Id): Move Pragma_Preelaborable_Initialization from
its current position to the end of the type, in the special set
of pragma literals that have corresponding atttributes. Add to
accompanying comment, indicating that functions Get_Pragma_Id
and Is_Pragma_Name need to be updated when adding a pragma
literal to the special set.
* snames.adb-tmpl (Get_Pragma_Id): Add case alternative for
Pragma_Preelaborable_Initialization.
(Is_Pragma_Name): Add test for
Name_Preelaborable_Initialization.
Eric Botcazou [Tue, 22 Jun 2021 22:21:00 +0000 (00:21 +0200)]
[Ada] Small cleanup in System.Dwarf_Line
gcc/ada/
* libgnat/s-dwalin.ads: Remove clause for Ada.Exceptions.Traceback,
add clause for System.Traceback_Entries and alphabetize.
(AET): Delete.
(STE): New package renaming.
(Symbolic_Traceback): Adjust.
* libgnat/s-dwalin.adb: Remove clauses for Ada.Exceptions.Traceback
and System.Traceback_Entries.
(Symbolic_Traceback): Adjust.
Ghjuvan Lacambre [Wed, 16 Jun 2021 13:17:13 +0000 (15:17 +0200)]
[Ada] Move Build_And_Insert_Cuda_Initialization to Expand_CUDA_Package
gcc/ada/
* exp_ch7.adb (Expand_N_Package_Body): Replace
Build_And_Insert_Cuda_Initialization with Expand_CUDA_Package.
* gnat_cuda.adb (Expand_CUDA_Package): New procedure.
(Build_And_Insert_Cuda_Initialization): Make internal.
* gnat_cuda.ads (Expand_CUDA_Package): New procedure.
(Build_And_Insert_Cuda_Initialization): Remove from spec.
* gfortran.h (gfc_omp_clauses): Add order_unconstrained.
* dump-parse-tree.c (show_omp_clauses): Dump it.
* openmp.c (gfc_match_omp_clauses): Match unconstrained/reproducible
modifiers to ordered(concurrent).
(OMP_DISTRIBUTE_CLAUSES): Accept ordered clause.
(resolve_omp_clauses): Reject ordered + order on same directive.
* trans-openmp.c (gfc_trans_omp_clauses, gfc_split_omp_clauses): Pass
on unconstrained modifier of ordered(concurrent).
gcc/testsuite/ChangeLog:
* gfortran.dg/gomp/order-5.f90: New test.
* gfortran.dg/gomp/order-6.f90: New test.
* gfortran.dg/gomp/order-7.f90: New test.
* gfortran.dg/gomp/order-8.f90: New test.
* gfortran.dg/gomp/order-9.f90: New test.
Richard Biener [Mon, 20 Sep 2021 09:20:19 +0000 (11:20 +0200)]
Avoid premature alignment setting in vect_duplicate_ssa_name_ptr_info
This removes adjusting alignment based on the vectorized accesses
and instead keeps what was set on the original access. The
code generating the actual accesses make sure to properly align
the vectorized accesses based on the generated pointer already
and the vectorizers alignment is always based of the desired
alignment of a vector type and thus will reset alignment to
unknown this way for example when doing strided accesses.
2021-09-20 Richard Biener <rguenther@suse.de>
* tree-vect-data-refs.c (vect_duplicate_ssa_name_ptr_info):
Do not compute alignment of the vectorized access here.
When running the testsuite under Windows, we noticed failures in
testcase which attempt to match compiler error messages containing the
name of the executable.
For instance, gcc.dg/analyzer/signal-4a.c tries to match 'cc1:' which
obviously fails when the executable is called cc1.exe.
This patch removes the .exe suffix from various toolchain executables
to avoid this problem.
Thomas Schwinge [Fri, 3 Sep 2021 16:25:10 +0000 (18:25 +0200)]
Don't record string concatenation data for 'RESERVED_LOCATION_P'
'RESERVED_LOCATION_P' means 'UNKNOWN_LOCATION' or 'BUILTINS_LOCATION'.
We're using 'UNKNOWN_LOCATION' as a spare value for 'Empty', so should
ascertain that we don't use it as a key additionally. Similarly for
'BUILTINS_LOCATION' that we'd later like to use as a spare value for
'Deleted'.
As discussed in the source code comment added, for these we didn't have
stable behavior anyway.
Richard Biener [Wed, 8 Sep 2021 12:42:31 +0000 (14:42 +0200)]
tree-optimization/65206 - dependence analysis on mixed pointer/array
This adds the capability to analyze the dependence of mixed
pointer/array accesses. The example is from where using a masked
load/store creates the pointer-based access when an otherwise
unconditional access is array based. Other examples would include
accesses to an array mixed with accesses from inlined helpers
that work on pointers.
The idea is quite simple and old - analyze the data-ref indices
as if the reference was pointer-based. The following change does
this by changing dr_analyze_indices to work on the indices
sub-structure and storing an alternate indices substructure in
each data reference. That alternate set of indices is analyzed
lazily by initialize_data_dependence_relation when it fails to
match-up the main set of indices of two data references.
initialize_data_dependence_relation is refactored into a head
and a tail worker and changed to work on one of the indices
structures and thus away from using DR_* access macros which
continue to reference the main indices substructure.
There are quite some vectorization and loop distribution opportunities
unleashed in SPEC CPU 2017, notably 520.omnetpp_r, 548.exchange2_r,
510.parest_r, 511.povray_r, 521.wrf_r, 526.blender_r, 527.cam4_r and
544.nab_r see amendments in what they report with -fopt-info-loop while
the rest of the specrate set sees no changes there. Measuring runtime
for the set where changes were reported reveals nothing off-noise
besides 511.povray_r which seems to regress slightly for me
(on a Zen2 machine with -Ofast -march=native).
2021-09-08 Richard Biener <rguenther@suse.de>
PR tree-optimization/65206
* tree-data-ref.h (struct data_reference): Add alt_indices,
order it last.
* tree-data-ref.c (free_data_ref): Release alt_indices.
(dr_analyze_indices): Work on struct indices and get DR_REF as tree.
(create_data_ref): Adjust.
(initialize_data_dependence_relation): Split into head
and tail. When the base objects fail to match up try
again with pointer-based analysis of indices.
* tree-vectorizer.c (vec_info_shared::check_datarefs): Do
not compare the lazily computed alternate set of indices.
* gcc.dg/torture/20210916.c: New testcase.
* gcc.dg/vect/pr65206.c: Likewise.
Driver: Fix bootstrap with DEFAULT_{ASSEMBLER,LINKER,DSYMUTIL}.
The patch at r12-3662-g5fee8a0a9223d factored the code for
printing the names of programes into a separate function.
However the moved editions that print out the names of the
assembler, linker (and dsymutil on Darwin) when those are
specified at configure-time were not adjusted accordingly,
leading to a bootstrap fail.
Fixed by testing specifically for execute OK, since we know
these are programs.
These were used on older systems to equate the FAT libgcc_s
library to single-slice equivalents. Unused for any current
system and never emitted by GCC.
This reorganises the Darwin symbol vers files to include
the generic ones at the top level; allowing for arch ports
to override (via either exclusion or inclusion as needed).
We add an X86-specific vers file containing the new HF
symbols. Note that although Darwin does not use ELF-style
symbol versioning - the parser that produces the map can
consume it. Using the ELF-style description will help us
know at which rev the symbols were introduced.
* config/i386/t-darwin: Add in a vers file for X86-specific
symbols.
* config/t-darwin: Add the generic symbol maps here...
* config/t-slibgcc-darwin: ... removing from here.
* config/i386/libgcc-darwin.ver: New file.
Darwin, crts: Build Darwin10 unwinder shim as a library.
We have a small unwinder shim that is only used for Darwin10
(and only then in quite specific cases). To avoid linking
this code for every executable or DSO, we can present the crt
as a convenience library (rather than a .o file).
Jeff Law [Sun, 19 Sep 2021 17:31:32 +0000 (13:31 -0400)]
[committed] Make test names unique for a couple of goacc tests
gcc/testsuite
* gfortran.dg/goacc/privatization-1-compute.f90: Make test names
unique.
* gfortran.dg/goacc/routine-external-level-of-parallelism-2.f:
Likewise.
Andrew Pinski [Sat, 18 Sep 2021 01:08:17 +0000 (01:08 +0000)]
Fix PR bootstrap/102389: --with-build-config=bootstrap-lto is broken
So the problem here is that now the lto-plugin requires NM that works
with LTO to work so we need to pass down NM just like we do for ranlib
and ar.
OK? Bootstrapped and tested with --with-build-config=bootstrap-lto on aarch64-linux-gnu.
Note you need to use binutils 2.35 or later too due to ttps://sourceware.org/PR25355
(I will submit another patch to improve the installation instructions too).
config/ChangeLog:
PR bootstrap/102389
* bootstrap-lto-lean.mk: Handle NM like RANLIB AND AR.
* bootstrap-lto.mk: Likewise.
Every time we allocate a threading edge we push it onto the path in a
distinct step. There's no need to do this in two steps, and avoiding
this, keeps us from exposing the internals of the registry.
I've also did some tiny cleanups in thread_across_edge, most importantly
removing the bitmap in favor of an auto_bitmap.
There are no functional changes.
gcc/ChangeLog:
* tree-ssa-threadbackward.c
(back_threader_registry::register_path): Use push_edge.
* tree-ssa-threadedge.c
(jump_threader::thread_around_empty_blocks): Same.
(jump_threader::thread_through_normal_block): Same.
(jump_threader::thread_across_edge): Same. Also, use auto_bitmap.
Tidy up code.
* tree-ssa-threadupdate.c
(jt_path_registry::allocate_thread_edge): Remove.
(jt_path_registry::push_edge): New.
(dump_jump_thread_path): Make static.
* tree-ssa-threadupdate.h (allocate_thread_edge): Remove.
(push_edge): New.
Iain Sandoe [Thu, 19 Aug 2021 12:01:14 +0000 (13:01 +0100)]
Jit, testsuite: Amend expect processing to tolerate more platforms.
The current 'fixed_host_execute' implementation fails on Darwin
platforms for a number of reasons:
1/ If the sub-process spawn fails (e.g. because of missing or mal-
formed params); rather than reporting the fail output into the
match stream, as indicated by the expect manual, it terminates
the script.
- We fix this by (a) checking that the executable is valid as well
as existing (b) we put the spawn into a catch block and report
a failure.
2/ There is no recovery path at all for a buffer-full case (and we
do see buffer-full events with the default sizes).
- Added by the patch here, however it is not as sophisticated as
the methods used by dejagnu internally. Here we set the process
to be "nowait" and then close the connection - with the intent
that this will terminate the spawned process.
3/ The expect logic assumes that 'Totals:' is a valid indicator
for the end of the spawned process output. This is not true
even for the default dejagnu header (there are a number of
additional reporting lines after). In addition to this, there
are some tests that intentionally produce more output after
the totals report (and there are tests that do not use that
mechanism at all).
The effect is the we might arrive at the "wait" for the spawned
process to finish - but that process might not have completed
all its output. For Darwin, at least that causes a deadlock
between expect and the spawnee - the latter is doing a non-
cancellable write and the former is waiting for the latter to
terminate. For some reason this does not seem to affect Linux
perhaps the pty implementation allows the write(s) are able to
proceed even though there is no reader.
- This is fixed by modifying the loop termination condition to be
either EOF (which will be the 'correct' condition) or a timeout
which would represent an error either in the runtime or in the
parsing of the output. As added precautions, we only try to
wait if there is a correcly-spawned process, and we are also
specific about which process we are waiting for.
4/ Darwin appears to have a bug in either the tcl or termios
'cooking' code that ocassionally inserts an additional CR char
into the stream - thus '\n' => '\r\r\n' instead of '\r\n'. The
original program output is correct (it only contains a single
\n) - the additional character is being inserted somewhere in
the translations applied before the output reaches expect.
The logic of this expect implementation does not tolerate single
\r or \n characters (it will fail with a timeout or buffer-full
if that occurs).
- This is fixed by having a line-end match that is adjusted for
Darwin.
5/ The default buffer size does seem to be too small in some cases
noting that GCC uses 10000 as the match buffer size and the
default is 2000.
- Fixed by increasing the size to 8192.
6/ There is a somewhat arbitrary dumping of output where we match
^$prefix\tSOMETHING... and then process the something. This
essentially allows the match to start at any place in the buffer
following any collection of non-line-end chars.
- Fixed by amending the match for 'general' lines to accommodate
these cases, and reporting such lines to the log. At least this
should allow debugging of any cases where output that should be
recognized is being dropped.
There was an inline extern declaration for dump_ranger that was a bit of
a hack. I've removed it in favor of an actual prototype. There are
also some trivial changes to the dumping code in the path solver.
John Ericson [Sun, 19 Sep 2021 15:08:32 +0000 (11:08 -0400)]
[PATCH] Factor out `find_a_program` helper around `find_a_file`
gcc/
* gcc.c (find_a_program): New function, factored out of...
(find_a_file): Here.
(execute): Use find_a_program when looking for programs rather
than find_a_file.
Roger Sayle [Sun, 19 Sep 2021 08:07:01 +0000 (09:07 +0100)]
PR middle-end/88173: More constant folding of NaN comparisons.
This patch tackles PR middle-end/88173 where the order of operands in
a comparison affects constant folding. As diagnosed by Jason Merrill,
"match.pd handles these comparisons very differently". The history is
that the middle end, typically canonicalizes comparisons to place
constants on the right, but when a comparison contains two constants
we need to check/transform both constants, i.e. on both the left and the
right. Hence the added lines below duplicate for @0 the same transform
applied a few lines above for @1.
Whilst preparing the testcase, I noticed that this transformation is
incorrectly disabled with -fsignaling-nans even when both operands are
known not be be signaling NaNs, so I've corrected that and added a
second test case. Unfortunately, c-c++-common/pr57371-4.c then starts
failing, as it doesn't distinguish QNaNs (which are quiet) from SNaNs
(which signal), so this patch includes a minor tweak to the expected
behaviour for QNaNs in that existing test.
2021-09-19 Roger Sayle <roger@nextmovesoftware.com>
gcc/ChangeLog
PR middle-end/88173
* match.pd (cmp @0 REAL_CST@1): When @0 is also REAL_CST, apply
the same transformations as to @1. For comparisons against NaN,
don't check HONOR_SNANS but confirm that neither operand is a
signaling NaN.
gcc/testsuite/ChangeLog
PR middle-end/88173
* c-c++-common/pr57371-4.c: Tweak/correct test case for QNaNs.
* g++.dg/pr88173-1.C: New test case.
* g++.dg/pr88173-2.C: New test case.
Andrew Pinski [Sat, 18 Sep 2021 08:20:16 +0000 (08:20 +0000)]
Fix middle-end/102395: reg_class having only NO_REGS and ALL_REGS.
So this is a simple fix is to just add to the assert that
sclass and dclass are both greater than or equal to NO_REGS.
NO_REGS is documented as the first register class so it should
have the value of 0.
gcc/ChangeLog:
* lra-constraints.c (check_and_process_move): Assert
that dclass and sclass are greater than or equal to NO_REGS.
Jakub Jelinek [Sat, 18 Sep 2021 07:50:14 +0000 (09:50 +0200)]
openmp: Handle unconstrained and reproducible modifiers on order(concurrent)
This patch adds handling for unconstrained and reproducible modifiers on
order(concurrent) clause. For all static schedules (including auto and
no schedule or dist_schedule clauses) I believe what we implement is
reproducible, so the patch doesn't do much beyond recognizing those.
Note, there is an OpenMP/spec issue that needs resolution on what
should happen with the dynamic schedules (whether it should be an error
to mix such clauses, or silently make it non-reproducible, and in which
exact cases), so it might need some follow-up.
Besides that, this patch allows order(concurrent) clause on the distribute
construct which is something also added in OpenMP 5.1, and finally
check the newly added restriction that at most one order clause
can appear on a construct.
The allowing of order clause on distribute has a side-effect that
order(concurrent) copyin(thrpriv) is no longer allowed on combined/composite
constructs with distribute parallel for{, simd} in it, previously the
order applied only to for/simd and so a threadprivate var could be seen
in the construct, but now it also applies to distribute and so on the parallel
we shouldn't refer to a threadprivate var.
2021-09-18 Jakub Jelinek <jakub@redhat.com>
gcc/
* tree.h (OMP_CLAUSE_ORDER_UNCONSTRAINED): Define.
* tree-pretty-print.c (dump_omp_clause): Print unconstrained:
for OMP_CLAUSE_ORDER_UNCONSTRAINED.
gcc/c-family/
* c-omp.c (c_omp_split_clauses): Split order clause also to
distribute construct. Copy over OMP_CLAUSE_ORDER_UNCONSTRAINED.
gcc/c/
* c-parser.c (c_parser_omp_clause_order): Parse unconstrained
and reproducible modifiers.
(OMP_DISTRIBUTE_CLAUSE_MASK): Add order clause.
gcc/cp/
* parser.c (cp_parser_omp_clause_order): Parse unconstrained
and reproducible modifiers.
(OMP_DISTRIBUTE_CLAUSE_MASK): Add order clause.
gcc/testsuite/
* c-c++-common/gomp/order-1.c (f2): Add tests for distribute
with order clause.
(f3): Remove.
* c-c++-common/gomp/order-2.c: Don't expect error for distribute
with order clause.
* c-c++-common/gomp/order-5.c: New test.
* c-c++-common/gomp/order-6.c: New test.
* c-c++-common/gomp/clause-dups-1.c (f1): Add tests for
duplicated order clause.
(f9): New function.
* c-c++-common/gomp/clauses-1.c (baz, bar): Don't mix copyin and
order(concurrent) clauses on the same composite construct combined
with distribute, instead split it into two tests, one without
copyin and one without order(concurrent). Add order(concurrent)
clauses to {,{,target} teams} distribute.
* g++.dg/gomp/attrs-1.C (baz, bar): Likewise.
* g++.dg/gomp/attrs-2.C (baz, bar): Likewise.
Besides conversion instructions, pass_rpad also handles scalar
sqrt/rsqrt/rcp/round instructions, while r12-3614 should only want to
handle conversion instructions, so fix it.
gcc/ChangeLog:
* config/i386/i386-features.c (remove_partial_avx_dependency):
Restrict TARGET_USE_VECTOR_FP_CONVERTS and
TARGET_USE_VECTOR_CONVERTS to conversion instructions only.
Jakub Jelinek [Sat, 18 Sep 2021 07:47:25 +0000 (09:47 +0200)]
openmp: Allow private or firstprivate arguments to default clause even for C/C++
OpenMP 5.1 allows default(private) or default(firstprivate) even in C/C++,
but it behaves the same way as in Fortran only for variables not declared at
namespace or file scope. For the namespace/file scope variables it instead
behaves as default(none).
2021-09-18 Jakub Jelinek <jakub@redhat.com>
gcc/
* gimplify.c (omp_default_clause): For C/C++ default({,first}private),
if file/namespace scope variable doesn't have predetermined sharing,
treat it as if there was default(none).
gcc/c/
* c-parser.c (c_parser_omp_clause_default): Handle private and
firstprivate arguments, adjust diagnostics on unknown argument.
gcc/cp/
* parser.c (cp_parser_omp_clause_default): Handle private and
firstprivate arguments, adjust diagnostics on unknown argument.
* cp-gimplify.c (cxx_omp_finish_clause): Handle OMP_CLAUSE_PRIVATE.
gcc/testsuite/
* c-c++-common/gomp/default-2.c: New test.
* c-c++-common/gomp/default-3.c: New test.
* g++.dg/gomp/default-1.C: New test.
libgomp/
* testsuite/libgomp.c++/default-1.C: New test.
* testsuite/libgomp.c-c++-common/default-1.c: New test.
* libgomp.texi (OpenMP 5.1): Mark "private and firstprivate argument
to default clause in C and C++" as implemented.
* gcc.target/i386/avx-1.c: Add test for new builtins.
* gcc.target/i386/sse-13.c: Ditto.
* gcc.target/i386/sse-23.c: Ditto.
* gcc.target/i386/sse-14.c: Add test for new intrinsics.
* gcc.target/i386/sse-22.c: Ditto.
* gcc.target/i386/avx-1.c: Add test for new builtins.
* gcc.target/i386/sse-13.c: Ditto.
* gcc.target/i386/sse-23.c: Ditto.
* gcc.target/i386/sse-14.c: Add test fot new intrinsics.
* gcc.target/i386/sse-22.c: Ditto.
* gcc.target/i386/avx-1.c: Add test for new builtins.
* gcc.target/i386/sse-13.c: Ditto.
* gcc.target/i386/sse-23.c: Ditto.
* gcc.target/i386/sse-14.c: Add test for new intrinsics.
* gcc.target/i386/sse-22.c: Ditto.
Jason Merrill [Fri, 17 Sep 2021 18:18:55 +0000 (14:18 -0400)]
c++: improve lookup of member-qualified names
I've been working on the resolution of CWG1835 by P1787, which among many
other things clarified that a name after -> or . is looked up first in the
class of the object expression even if it's dependent. This patch does not
make that change; this is a smaller change extracted from that work in
progress to make the lookup in the object type work better in cases where
unqualified lookup doesn't find anything.
Basically, if we see "t.foo::" we know that looking up foo in t needs to
find a type, so we build an implicit TYPENAME_TYPE for it.
This also implements the change from P1787 to assume that a name followed by
< in a type-only context names a template, since the less-than operator
can't appear in a type context. This makes some of the lines in dtor11.C
work.
I introduce the predicate 'dependentish_scope_p' for the case where the
current instantiation has dependent bases, so even though we can perform
name lookup, we can't conclude that a lookup failure is conclusive.
gcc/cp/ChangeLog:
* cp-tree.h (dependentish_scope_p): Declare.
* pt.c (dependentish_scope_p): New.
* parser.c (cp_parser_lookup_name): Return a TYPENAME_TYPE
for lookup of a type in a dependent object.
(cp_parser_template_id): Handle TYPENAME_TYPE.
(cp_parser_template_name): If we're looking for a type,
a name followed by < names a template.
gcc/testsuite/ChangeLog:
* g++.dg/template/dtor5.C: Adjust expected error.
* g++.dg/cpp23/lookup2.C: New test.
* g++.dg/template/dtor11.C: New test.
Harald Anlauf [Fri, 17 Sep 2021 19:45:33 +0000 (21:45 +0200)]
Fortran - (large) arrays in the main shall be static
gcc/fortran/ChangeLog:
PR fortran/102366
* trans-decl.c (gfc_finish_var_decl): Disable the warning message
for variables moved from stack to static storange if they are
declared in the main, but allow the move to happen.
gcc/testsuite/ChangeLog:
PR fortran/102366
* gfortran.dg/pr102366.f90: New test.
Jonathan Wakely [Fri, 17 Sep 2021 11:28:35 +0000 (12:28 +0100)]
libstdc++: Add 'noexcept' to path::iterator members
All path::iterator operations are non-throwing.
Signed-off-by: Jonathan Wakely <jwakely@redhat.com>
libstdc++-v3/ChangeLog:
* include/bits/fs_path.h (path::iterator): Add noexcept to all
member functions and friend functions.
(distance): Add noexcept.
(advance): Add noexcept and inline.
* include/experimental/bits/fs_path.h (path::iterator):
Add noexcept to all member functions.
Jonathan Wakely [Fri, 17 Sep 2021 11:27:02 +0000 (12:27 +0100)]
libstdc++: Fix last std::tuple constructor missing 'constexpr' [PR102270]
Also rename the test so it actually runs.
Signed-off-by: Jonathan Wakely <jwakely@redhat.com>
libstdc++-v3/ChangeLog:
PR libstdc++/102270
* include/std/tuple (_Tuple_impl): Add constexpr to constructor
missed in previous patch.
* testsuite/20_util/tuple/cons/102270.C: Moved to...
* testsuite/20_util/tuple/cons/102270.cc: ...here.
* testsuite/util/testsuite_allocator.h (SimpleAllocator): Add
constexpr to constructor so it can be used for C++20 tests.
This is an optimisation for middle-end worker-partitioning support (used
to support multiple workers on AMD GCN). At present, barriers may be
emitted in cases where they aren't needed and cannot be optimised away.
This patch stops the extraneous barriers from being emitted in the
first place.
One exception to the above (where the barrier is still needed) is for
predicated blocks of code that perform a write to gang-private shared
memory from one worker. We must execute a barrier before other workers
read that shared memory location.
gcc/
* config/gcn/gcn.c (gimple.h): Include.
(gcn_fork_join): Emit barrier for worker-level joins.
* omp-oacc-neuter-broadcast.cc (find_local_vars_to_propagate): Add
writes_gang_private bitmap parameter. Set bit for blocks
containing gang-private variable writes.
(worker_single_simple): Don't emit barrier after predicated block.
(worker_single_copy): Don't emit barrier if we're not broadcasting
anything and the block contains no gang-private writes.
(neuter_worker_single): Don't predicate blocks that only contain
NOPs or internal marker functions. Pass has_gang_private_write
argument to worker_single_copy.
(oacc_do_neutering): Add writes_gang_private bitmap handling.
Julian Brown [Mon, 29 Jun 2020 20:16:52 +0000 (13:16 -0700)]
openacc: Shared memory layout optimisation
This patch implements an algorithm to lay out local data-share (LDS)
space. It currently works for AMD GCN. At the moment, LDS is used for
three things:
After the patch is applied, (2) and (3) are placed at preallocated
locations in LDS, and (1) continues to be handled by the backend (as it
is at present prior to this patch being applied). LDS now looks like this:
So, gang-private space is fixed at a constant amount at compile time
(which can be increased with a command-line switch if necessary
for some given code). The layout algorithm takes out a slice of the
remainder of usable space for reduction vars, and uses the rest for
worker partitioning.
The partitioning algorithm works as follows.
1. An "adjacency" set is built up for each basic block that might
do a broadcast. This is calculated by starting at each such block,
and doing a recursive DFS walk over successors to find the next
block (or blocks) that *also* does a broadcast
(dfs_broadcast_reachable_1).
2. The adjacency set is inverted to get adjacent predecessor blocks also.
3. Blocks that will perform a broadcast are sorted by size of that
broadcast: the biggest blocks are handled first.
4. A splay tree structure is used to calculate the spans of LDS memory
that are already allocated by the blocks adjacent to this one
(merge_ranges{,_1}.
5. The current block's broadcast space is allocated from the first free
span not allocated in the splay tree structure calculated above
(first_fit_range). This seems to work quite nicely and efficiently
with the splay tree structure.
6. Continue with the next-biggest broadcast block until we're done.
In this way, "adjacent" broadcasts will not use the same piece of
LDS memory.
PR96334 "openacc: Unshare reduction temporaries for GCN" got merged in:
The GCN backend uses tree nodes like MEM((__lds TYPE *) <constant>)
for reduction temporaries. Unlike e.g. var decls and SSA names, these
nodes cannot be shared during gimplification, but are so in some
circumstances. This is detected when appropriate --enable-checking
options are used. This patch unshares such nodes when they are reused
more than once.
gcc/
* config/gcn/gcn-protos.h
(gcn_goacc_create_worker_broadcast_record): Update prototype.
* config/gcn/gcn-tree.c (gcn_goacc_get_worker_red_decl): Use
preallocated block of LDS memory. Do not cache/share decls for
reduction temporaries between invocations.
(gcn_goacc_reduction_teardown): Unshare VAR on second use.
(gcn_goacc_create_worker_broadcast_record): Add OFFSET parameter
and return temporary LDS space at that offset. Return pointer in
"sender" case.
* config/gcn/gcn.c (acc_lds_size, gang_private_hwm, lds_allocs):
New global vars.
(ACC_LDS_SIZE): Define as acc_lds_size.
(gcn_init_machine_status): Don't initialise lds_allocated,
lds_allocs, reduc_decls fields of machine function struct.
(gcn_option_override): Handle default size for gang-private
variables and -mgang-private-size option.
(gcn_expand_prologue): Use LDS_SIZE instead of LDS_SIZE-1 when
initialising M0_REG.
(gcn_shared_mem_layout): New function.
(gcn_print_lds_decl): Update comment. Use global lds_allocs map and
gang_private_hwm variable.
(TARGET_GOACC_SHARED_MEM_LAYOUT): Define target hook.
* config/gcn/gcn.h (machine_function): Remove lds_allocated,
lds_allocs, reduc_decls. Add reduction_base, reduction_limit.
* config/gcn/gcn.opt (gang_private_size_opt): New global.
(mgang-private-size=): New option.
* doc/tm.texi.in (TARGET_GOACC_SHARED_MEM_LAYOUT): Place
documentation hook.
* doc/tm.texi: Regenerate.
* omp-oacc-neuter-broadcast.cc (targhooks.h, diagnostic-core.h):
Add includes.
(build_sender_ref): Handle sender_decl being pointer.
(worker_single_copy): Add PLACEMENT and ISOLATE_BROADCASTS
parameters. Pass placement argument to
create_worker_broadcast_record hook invocations. Handle
sender_decl being pointer and isolate_broadcasts inserting extra
barriers.
(blk_offset_map_t): Add typedef.
(neuter_worker_single): Add BLK_OFFSET_MAP parameter. Pass
preallocated range to worker_single_copy call.
(dfs_broadcast_reachable_1): New function.
(idx_decl_pair_t, used_range_vec_t): New typedefs.
(sort_size_descending): New function.
(addr_range): New class.
(splay_tree_compare_addr_range, splay_tree_free_key)
(first_fit_range, merge_ranges_1, merge_ranges): New functions.
(execute_omp_oacc_neuter_broadcast): Rename to...
(oacc_do_neutering): ... this. Add BOUNDS_LO, BOUNDS_HI
parameters. Arrange layout of shared memory for broadcast
operations.
(execute_omp_oacc_neuter_broadcast): New function.
(pass_omp_oacc_neuter_broadcast::gate): Remove num_workers==1
handling from here. Enable pass for all OpenACC routines in order
to call shared memory-layout hook.
* target.def (create_worker_broadcast_record): Add OFFSET
parameter.
(shared_mem_layout): New hook.
libgomp/
* testsuite/libgomp.oacc-c-c++-common/broadcast-many.c: Update.
Julian Brown [Mon, 29 Jun 2020 20:16:51 +0000 (13:16 -0700)]
openacc: Turn off worker partitioning if num_workers==1
This patch turns off the middle-end worker-partitioning support if the
number of workers for an outlined offload function is one. In that case,
we do not need to perform the broadcasting/neutering code transformation.
gcc/
* omp-oacc-neuter-broadcast.cc
(pass_omp_oacc_neuter_broadcast::gate): Disable if num_workers is
1.
(execute_omp_oacc_neuter_broadcast): Adjust.
Co-Authored-By: Thomas Schwinge <thomas@codesourcery.com>
Andrew MacLeod [Wed, 15 Sep 2021 18:43:51 +0000 (14:43 -0400)]
Virtualize relation oracle and various cleanups.
Standardize equiv_oracle API onto the new relation_oracle virtual base, and
then have dom_oracle inherit from that.
equiv_set always returns an equivalency set now, never NULL.
EQ_EXPR requires symmetry now. Each SSA name must be in the other equiv set.
Shuffle some routines around, simplify.
* gimple-range-cache.cc (ranger_cache::ranger_cache): Create a DOM
based oracle.
* gimple-range-fold.cc (fur_depend::register_relation): Use
register_stmt/edge routines.
* value-relation.cc (equiv_chain::find): Relocate from equiv_oracle.
(equiv_oracle::equiv_oracle): Create self equivalence cache.
(equiv_oracle::~equiv_oracle): Release same.
(equiv_oracle::equiv_set): Return entry from self equiv cache if there
are no equivalences.
(equiv_oracle::find_equiv_block): Move list find to equiv_chain.
(equiv_oracle::register_relation): Rename from register_equiv.
(relation_chain_head::find_relation): Relocate from dom_oracle.
(relation_oracle::register_stmt): New.
(relation_oracle::register_edge): New.
(dom_oracle::*): Rename from relation_oracle.
(dom_oracle::register_relation): Adjust to call equiv_oracle.
(dom_oracle::set_one_relation): Split from register_relation.
(dom_oracle::register_transitives): Consolidate 2 methods.
(dom_oracle::find_relation_block): Move core to relation_chain.
(dom_oracle::query_relation): Rename from find_relation_dom and adjust.
* value-relation.h (class relation_oracle): New pure virtual base.
(class equiv_oracle): Inherit from relation_oracle and adjust.
(class dom_oracle): Rename from old relation_oracle and adjust.
This set of tests failed on many different combination of -march, -mtune.
some of them failed with -fstack-protestor-all, or -mno-sse. And the
pattern matches are also different on lp64 or ia32.
The reason for these failures is that the RTL or assembly level patten
matches are only valid for -march=x86-64 -mtune=generic.
We restrict the testing only for -march=x86-64 and -mtune=generic. Also
add -fno-stack-protector or -msse for some of the testing cases.
gcc/testsuite/ChangeLog:
2021-09-17 qing zhao <qing.zhao@oracle.com>
* gcc.target/i386/auto-init-1.c: Restrict the testing only for
-march=x86-64 and -mtune=generic. Add -fno-stack-protector.
* gcc.target/i386/auto-init-2.c: Restrict the testing only for
-march=x86-64 and -mtune=generic -msse.
* gcc.target/i386/auto-init-3.c: Likewise.
* gcc.target/i386/auto-init-4.c: Likewise.
* gcc.target/i386/auto-init-5.c: Different pattern match for lp64 and
ia32.
* gcc.target/i386/auto-init-6.c: Restrict the testing only for
-march=x86-64 and -mtune-generic -msse. Add -fno-stack-protector.
* gcc.target/i386/auto-init-7.c: Likewise.
* gcc.target/i386/auto-init-8.c: Restrict the testing only for
-march=x86-64 and -mtune=generic -msse..
* gcc.target/i386/auto-init-padding-1.c: Likewise.
* gcc.target/i386/auto-init-padding-10.c: Likewise.
* gcc.target/i386/auto-init-padding-11.c: Likewise.
* gcc.target/i386/auto-init-padding-12.c: Likewise.
* gcc.target/i386/auto-init-padding-2.c: Likewise.
* gcc.target/i386/auto-init-padding-3.c: Restrict the testing only for
-march=x86-64. Different pattern match for lp64 and ia32.
* gcc.target/i386/auto-init-padding-4.c: Restrict the testing only for
-march=x86-64 and -mtune-generic -msse.
* gcc.target/i386/auto-init-padding-5.c: Likewise.
* gcc.target/i386/auto-init-padding-6.c: Likewise.
* gcc.target/i386/auto-init-padding-7.c: Restrict the testing only for
-march=x86-64 and -mtune-generic -msse. Add -fno-stack-protector.
* gcc.target/i386/auto-init-padding-8.c: Likewise.
* gcc.target/i386/auto-init-padding-9.c: Restrict the testing only for
-march=x86-64. Different pattern match for lp64 and ia32.
Martin Sebor [Fri, 17 Sep 2021 16:36:54 +0000 (10:36 -0600)]
Better handle MIN/MAX_EXPR of unrelated objects [PR102200].
Resolves:
PR middle-end/102200 - ICE on a min of a decl and pointer in a loop
gcc/ChangeLog:
PR middle-end/102200
* pointer-query.cc (access_ref::inform_access): Handle MIN/MAX_EXPR.
(handle_min_max_size): Change argument. Store original SSA_NAME for
operands to potentially distinct (sub)objects.
(compute_objsize_r): Adjust call to the above.
gcc/testsuite/ChangeLog:
PR middle-end/102200
* gcc.dg/Wstringop-overflow-62.c: Adjust text of an expected note.
* gcc.dg/Warray-bounds-89.c: New test.
* gcc.dg/Wstringop-overflow-74.c: New test.
* gcc.dg/Wstringop-overflow-75.c: New test.
* gcc.dg/Wstringop-overflow-76.c: New test.
Bill Schmidt [Fri, 17 Sep 2021 15:32:59 +0000 (10:32 -0500)]
rs6000: Handle some recent MMA builtin changes
Peter Bergner recently added two new builtins __builtin_vsx_lxvp and
__builtin_vsx_stxvp. These happened to break a pattern in MMA builtins that
I had been using to automate gimple folding of MMA builtins. Previously,
every MMA function that could be folded had an associated internal function
that it was folded into. The LXVP/STXVP builtins are just folded directly
into memory operations.
Instead of relying on this pattern, this patch adds a new attribute to
builtins called "mmaint," which is set for all MMA builtins that have an
associated internal builtin. The naming convention that adds _INTERNAL to
the builtin index name remains.
The rest of the patch is just duplicating Peter's patch, using the new
builtin infrastructure.
Bill Schmidt [Fri, 17 Sep 2021 15:06:22 +0000 (10:06 -0500)]
rs6000: Handle gimple folding of target built-ins
This is another patch that looks bigger than it really is. Because we
have a new namespace for the builtins, allowing us to have both the old
and new builtin infrastructure supported at once, we need versions of
these functions that use the new builtin namespace. Otherwise the code is
unchanged.
2021-09-17 Bill Schmidt <wschmidt@linux.ibm.com>
gcc/
* config/rs6000/rs6000-call.c (rs6000_gimple_fold_new_builtin): New
forward decl.
(rs6000_gimple_fold_builtin): Call rs6000_gimple_fold_new_builtin.
(rs6000_new_builtin_valid_without_lhs): New function.
(rs6000_gimple_fold_new_mma_builtin): Likewise.
(rs6000_gimple_fold_new_builtin): Likewise.
Fortran: Use _Float128 rather than __float128 for c_float128 kind.
The GNU Fortran manual documents that the c_float128 kind corresponds
to __float128, but in fact the implementation uses float128_type_node,
which is _Float128. Both refer to the 128-bit IEEE/ISO encoding, but
some targets including aarch64 only define _Float128 and not __float128,
and do not provide quadmath.h. This caused errors in some test cases
referring to __float128.
This patch changes the documentation (including code comments) and
test cases to use _Float128 to match the implementation.
* intrinsic.texi (ISO_C_BINDING): Change C_FLOAT128 to correspond
to _Float128 rather than __float128.
* iso-c-binding.def (c_float128): Update comments.
* trans-intrinsic.c (gfc_builtin_decl_for_float_kind): Likewise.
(build_round_expr): Likewise.
(gfc_build_intrinsic_lib_fndcecls): Likewise.
* trans-types.h (gfc_real16_is_float128): Likewise.
gcc/testsuite/
* gfortran.dg/PR100914.c: Do not include quadmath.h. Use
_Float128 _Complex instead of __complex128.
* gfortran.dg/PR100914.f90: Add -Wno-pedantic to suppress error
about use of _Float128.
* gfortran.dg/c-interop/typecodes-array-float128-c.c: Use
_Float128 instead of __float128.
* gfortran.dg/c-interop/typecodes-sanity-c.c: Likewise.
* gfortran.dg/c-interop/typecodes-scalar-float128-c.c: Likewise.
* lib/target-supports.exp
(check_effective_target_fortran_real_c_float128): Update comments.
Roger Sayle [Fri, 17 Sep 2021 14:57:34 +0000 (15:57 +0100)]
PR c/102245: Disable sign-changing optimization for shifts by zero.
Respecting Jakub's suggestion that it may be better to warn-on-valid for
"if (x << 0)" as the author might have intended "if (x < 0)" [which will
also warn when x is _Bool], the simplest way to resolve this regression
is to disable the recently added fold transformation for shifts by zero;
these will be optimized later (elsewhere). Guarding against integer_zerop
is the simplest of three alternatives; the second being to only apply
this transformation to GIMPLE and not GENERIC, and the third (potentially)
being to explicitly handle shifts by zero here, with an (if cond then else),
optimizing the expression to a convert, but awkwardly duplicating a
more general transformation earlier in match.pd's shift simplifications.
2021-09-17 Roger Sayle <roger@nextmovesoftware.com>
gcc/ChangeLog
PR c/102245
* match.pd (shift optimizations): Disable recent sign-changing
optimization for shifts by zero, these will be folded later.
gcc/testsuite/ChangeLog
PR c/102245
* gcc.dg/Wint-in-bool-context-4.c: New test case.
Bill Schmidt [Tue, 31 Aug 2021 20:13:01 +0000 (15:13 -0500)]
rs6000: Move __builtin_mffsl to the [always] stanza
I over-restricted use of __builtin_mffsl, since I was unaware that it
automatically uses mffs when mffsl is not available. Paul Clarke pointed
this out in discussion of his SSE 4.1 compatibility patches.
2021-08-31 Bill Schmidt <wschmidt@linux.ibm.com>
gcc/
* config/rs6000/rs6000-builtin-new.def (__builtin_mffsl): Move from
[power9] to [always].
libgfortran/
* ISO_Fortran_binding.h: Only include float.h if the C compiler
doesn't have predefined __LDBL_* and __DBL_* macros. Handle
LDBL_MANT_DIG == 53 for FreeBSD.
Iain Sandoe [Sat, 7 Aug 2021 13:45:56 +0000 (14:45 +0100)]
configure, jit: Allow for 'make check-gcc-jit'.
This is a convenience feature that allows the user to
do "make check-gcc-jit" at the top level of the build
to check that facility in isolation from others.
Jakub Jelinek [Fri, 17 Sep 2021 09:28:31 +0000 (11:28 +0200)]
openmp: Add support for OpenMP 5.1 atomics for C++
Besides the C++ FE changes, I've noticed that the C FE didn't reject
#pragma omp atomic capture compare
{ v = x; x = y; }
and other forms of atomic swap, this patch fixes that too. And the
c-family/ routine needed quite a few changes so that the new code
in it works fine with both FEs.
2021-09-17 Jakub Jelinek <jakub@redhat.com>
gcc/c-family/
* c-omp.c (c_finish_omp_atomic): Avoid creating
TARGET_EXPR if test is true, use create_tmp_var_raw instead of
create_tmp_var and add a zero initializer to TARGET_EXPRs that
had NULL initializer. When omitting operands after v = x,
use type of v rather than type of x. Fix type of vtmp
TARGET_EXPR.
gcc/c/
* c-parser.c (c_parser_omp_atomic): Reject atomic swap if capture
is true.
gcc/cp/
* cp-tree.h (finish_omp_atomic): Add r and weak arguments.
* parser.c (cp_parser_omp_atomic): Update function comment for
OpenMP 5.1 atomics, parse OpenMP 5.1 atomics and fail, compare and
weak clauses.
* semantics.c (finish_omp_atomic): Add r and weak arguments, handle
them, handle COND_EXPRs.
* pt.c (tsubst_expr): Adjust for COND_EXPR forms that
finish_omp_atomic can now produce.
gcc/testsuite/
* c-c++-common/gomp/atomic-18.c: Expect same diagnostics in C++ as in
C.
* c-c++-common/gomp/atomic-25.c: Drop c effective target.
* c-c++-common/gomp/atomic-26.c: Likewise.
* c-c++-common/gomp/atomic-27.c: Likewise.
* c-c++-common/gomp/atomic-28.c: Likewise.
* c-c++-common/gomp/atomic-29.c: Likewise.
* c-c++-common/gomp/atomic-30.c: Likewise. Adjust expected diagnostics
for C++ when it differs from C.
(foo): Change return type from double to void.
* g++.dg/gomp/atomic-5.C: Adjust expected diagnostics wording.
* g++.dg/gomp/atomic-20.C: New test.
libgomp/
* testsuite/libgomp.c-c++-common/atomic-19.c: Drop c effective target.
Use /* */ comments instead of //.
* testsuite/libgomp.c-c++-common/atomic-20.c: Likewise.
* testsuite/libgomp.c-c++-common/atomic-21.c: Likewise.
* testsuite/libgomp.c++/atomic-16.C: New test.
* testsuite/libgomp.c++/atomic-17.C: New test.
1. Replace TARGET_SSE_PARTIAL_REG_DEPENDENCY with
TARGET_SSE_PARTIAL_REG_FP_CONVERTS_DEPENDENCY in SSE FP to FP splitters.
2. Replace TARGET_SSE_PARTIAL_REG_DEPENDENCY with
TARGET_SSE_PARTIAL_REG_CONVERTS_DEPENDENCY in SSE INT to FP splitters.
3. Also check TARGET_SSE_PARTIAL_REG_FP_CONVERTS_DEPENDENCY and
TARGET_SSE_PARTIAL_REG_DEPENDENCY when handling avx_partial_xmm_update
attribute. Don't convert AVX partial XMM register update if there is no
partial SSE register dependency for SSE conversion.
gcc/
* config/i386/i386-features.c (remove_partial_avx_dependency):
Also check TARGET_SSE_PARTIAL_REG_FP_CONVERTS_DEPENDENCY and
and TARGET_SSE_PARTIAL_REG_CONVERTS_DEPENDENCY before generating
vxorps.
* config/i386/i386.h (TARGET_SSE_PARTIAL_REG_FP_CONVERTS_DEPENDENCY):
New.
(TARGET_SSE_PARTIAL_REG_CONVERTS_DEPENDENCY): Likewise.
* config/i386/i386.md (SSE FP to FP splitters): Replace
TARGET_SSE_PARTIAL_REG_DEPENDENCY with
TARGET_SSE_PARTIAL_REG_FP_CONVERTS_DEPENDENCY.
(SSE INT to FP splitter): Replace TARGET_SSE_PARTIAL_REG_DEPENDENCY
with TARGET_SSE_PARTIAL_REG_CONVERTS_DEPENDENCY.
* config/i386/x86-tune.def
(X86_TUNE_SSE_PARTIAL_REG_FP_CONVERTS_DEPENDENCY): New.
(X86_TUNE_SSE_PARTIAL_REG_CONVERTS_DEPENDENCY): Likewise.
Check TARGET_USE_VECTOR_FP_CONVERTS or TARGET_USE_VECTOR_CONVERTS when
handling avx_partial_xmm_update attribute. Don't convert AVX partial
XMM register update if vector packed SSE conversion should be used.
gcc/
PR target/101900
* config/i386/i386-features.c (remove_partial_avx_dependency):
Check TARGET_USE_VECTOR_FP_CONVERTS and TARGET_USE_VECTOR_CONVERTS
before generating vxorps.
H.J. Lu [Wed, 15 Sep 2021 06:17:08 +0000 (14:17 +0800)]
x86: Update memcpy/memset inline strategies for -mtune=tremont
Simply memcpy and memset inline strategies to avoid branches for
-mtune=tremont:
1. Create Tremont cost model from generic cost model.
2. With MOVE_RATIO and CLEAR_RATIO == 17, GCC will use integer/vector
load and store for up to 16 * 16 (256) bytes when the data size is
fixed and known.
3. Inline only if data size is known to be <= 256.
a. Use "rep movsb/stosb" with simple code sequence if the data size
is a constant.
b. Use loop if data size is not a constant.
4. Use memcpy/memset libray function if data size is unknown or > 256.
* config/i386/i386-options.c (processor_cost_table): Use
tremont_cost for Tremont.
* config/i386/x86-tune-costs.h (tremont_memcpy): New.
(tremont_memset): Likewise.
(tremont_cost): Likewise.
* config/i386/x86-tune.def (X86_TUNE_PREFER_KNOWN_REP_MOVSB_STOSB):
Enable for Tremont.
Eric Botcazou [Fri, 17 Sep 2021 08:12:12 +0000 (10:12 +0200)]
Fix PR rtl-optimization/102306
This is a duplication of volatile loads introduced during GCC 9 development
by the 2->2 mechanism of the RTL combiner. There is already a substantial
checking for volatile references in can_combine_p but it implicitly assumes
that the combination reduces the number of instructions, which is of course
not the case here. So the fix teaches try_combine to abort the combination
when it is about to make a copy of volatile references to preserve them.
gcc/
PR rtl-optimization/102306
* combine.c (try_combine): Abort the combination if we are about to
duplicate volatile references.
gcc/testsuite/
* gcc.target/sparc/20210917-1.c: New test.
* gcc.target/i386/avx-1.c: Add test for new builtins.
* gcc.target/i386/sse-13.c: Ditto.
* gcc.target/i386/sse-23.c: Ditto.
* gcc.target/i386/sse-14.c: Add test for new intrinsics.
* gcc.target/i386/sse-22.c: Ditto.
* gcc.target/i386/avx-1.c: Add test for new builtins.
* gcc.target/i386/sse-13.c: Ditto.
* gcc.target/i386/sse-23.c: Ditto.
* gcc.target/i386/sse-14.c: Add test for new intrinsics.
* gcc.target/i386/sse-22.c: Ditto.
* gcc.target/i386/avx-1.c: Add test for new builtins.
* gcc.target/i386/sse-13.c: Ditto.
* gcc.target/i386/sse-23.c: Ditto.
* gcc.target/i386/sse-14.c: Add test for new intrinsics.
* gcc.target/i386/sse-22.c: Ditto.
* gcc.target/i386/avx-1.c: Add test for new builtins.
* gcc.target/i386/sse-13.c: Ditto.
* gcc.target/i386/sse-23.c: Ditto.
* gcc.target/i386/sse-14.c: Add test for new intrinsics.
* gcc.target/i386/sse-22.c: Ditto.
projects / thirdparty / gcc.git / log
