Hi!
I've committed following backports to 4.8 branch, after
bootstrapping/regtesting on x86_64-linux and i686-linux.
Jakub
2014-11-28 Jakub Jelinek <ja...@redhat.com>
Backported from mainline
2014-10-03 Jakub Jelinek <ja...@redhat.com>
PR libgomp/61200
* omp-low.c (taskreg_contexts): New variable.
(scan_omp_parallel): Push newly created context into taskreg_contexts
vector and move record layout code to finish_taskreg_scan.
(scan_omp_task): Likewise.
(finish_taskreg_scan): New function.
(execute_lower_omp): Call finish_taskreg_scan on all taskreg_contexts
vector elements and release it.
* c-c++-common/gomp/pr61200.c: New test.
* testsuite/libgomp.c/pr61200.c: New test.
--- gcc/omp-low.c (revision 215834)
+++ gcc/omp-low.c (revision 215835)
@@ -128,6 +128,7 @@ static splay_tree all_contexts;
static int taskreg_nesting_level;
struct omp_region *root_omp_region;
static bitmap task_shared_vars;
+static vec<omp_context *> taskreg_contexts;
static void scan_omp (gimple_seq *, omp_context *);
static tree scan_omp_1_op (tree *, int *, void *);
@@ -1655,6 +1656,7 @@ scan_omp_parallel (gimple_stmt_iterator
}
ctx = new_omp_context (stmt, outer_ctx);
+ taskreg_contexts.safe_push (ctx);
if (taskreg_nesting_level > 1)
ctx->is_nested = true;
ctx->field_map = splay_tree_new (splay_tree_compare_pointers, 0, 0);
@@ -1674,11 +1676,6 @@ scan_omp_parallel (gimple_stmt_iterator
if (TYPE_FIELDS (ctx->record_type) == NULL)
ctx->record_type = ctx->receiver_decl = NULL;
- else
- {
- layout_type (ctx->record_type);
- fixup_child_record_type (ctx);
- }
}
/* Scan an OpenMP task directive. */
@@ -1689,7 +1686,6 @@ scan_omp_task (gimple_stmt_iterator *gsi
omp_context *ctx;
tree name, t;
gimple stmt = gsi_stmt (*gsi);
- location_t loc = gimple_location (stmt);
/* Ignore task directives with empty bodies. */
if (optimize > 0
@@ -1700,6 +1696,7 @@ scan_omp_task (gimple_stmt_iterator *gsi
}
ctx = new_omp_context (stmt, outer_ctx);
+ taskreg_contexts.safe_push (ctx);
if (taskreg_nesting_level > 1)
ctx->is_nested = true;
ctx->field_map = splay_tree_new (splay_tree_compare_pointers, 0, 0);
@@ -1737,8 +1734,71 @@ scan_omp_task (gimple_stmt_iterator *gsi
t = build_int_cst (long_integer_type_node, 1);
gimple_omp_task_set_arg_align (stmt, t);
}
+}
+
+
+/* If any decls have been made addressable during scan_omp,
+ adjust their fields if needed, and layout record types
+ of parallel/task constructs. */
+
+static void
+finish_taskreg_scan (omp_context *ctx)
+{
+ if (ctx->record_type == NULL_TREE)
+ return;
+
+ /* If any task_shared_vars were needed, verify all
+ OMP_CLAUSE_SHARED clauses on GIMPLE_OMP_{PARALLEL,TASK}
+ statements if use_pointer_for_field hasn't changed
+ because of that. If it did, update field types now. */
+ if (task_shared_vars)
+ {
+ tree c;
+
+ for (c = gimple_omp_taskreg_clauses (ctx->stmt);
+ c; c = OMP_CLAUSE_CHAIN (c))
+ if (OMP_CLAUSE_CODE (c) == OMP_CLAUSE_SHARED)
+ {
+ tree decl = OMP_CLAUSE_DECL (c);
+
+ /* Global variables don't need to be copied,
+ the receiver side will use them directly. */
+ if (is_global_var (maybe_lookup_decl_in_outer_ctx (decl, ctx)))
+ continue;
+ if (!bitmap_bit_p (task_shared_vars, DECL_UID (decl))
+ || !use_pointer_for_field (decl, ctx))
+ continue;
+ tree field = lookup_field (decl, ctx);
+ if (TREE_CODE (TREE_TYPE (field)) == POINTER_TYPE
+ && TREE_TYPE (TREE_TYPE (field)) == TREE_TYPE (decl))
+ continue;
+ TREE_TYPE (field) = build_pointer_type (TREE_TYPE (decl));
+ TREE_THIS_VOLATILE (field) = 0;
+ DECL_USER_ALIGN (field) = 0;
+ DECL_ALIGN (field) = TYPE_ALIGN (TREE_TYPE (field));
+ if (TYPE_ALIGN (ctx->record_type) < DECL_ALIGN (field))
+ TYPE_ALIGN (ctx->record_type) = DECL_ALIGN (field);
+ if (ctx->srecord_type)
+ {
+ tree sfield = lookup_sfield (decl, ctx);
+ TREE_TYPE (sfield) = TREE_TYPE (field);
+ TREE_THIS_VOLATILE (sfield) = 0;
+ DECL_USER_ALIGN (sfield) = 0;
+ DECL_ALIGN (sfield) = DECL_ALIGN (field);
+ if (TYPE_ALIGN (ctx->srecord_type) < DECL_ALIGN (sfield))
+ TYPE_ALIGN (ctx->srecord_type) = DECL_ALIGN (sfield);
+ }
+ }
+ }
+
+ if (gimple_code (ctx->stmt) == GIMPLE_OMP_PARALLEL)
+ {
+ layout_type (ctx->record_type);
+ fixup_child_record_type (ctx);
+ }
else
{
+ location_t loc = gimple_location (ctx->stmt);
tree *p, vla_fields = NULL_TREE, *q = &vla_fields;
/* Move VLA fields to the end. */
p = &TYPE_FIELDS (ctx->record_type);
@@ -1758,12 +1818,12 @@ scan_omp_task (gimple_stmt_iterator *gsi
fixup_child_record_type (ctx);
if (ctx->srecord_type)
layout_type (ctx->srecord_type);
- t = fold_convert_loc (loc, long_integer_type_node,
- TYPE_SIZE_UNIT (ctx->record_type));
- gimple_omp_task_set_arg_size (stmt, t);
+ tree t = fold_convert_loc (loc, long_integer_type_node,
+ TYPE_SIZE_UNIT (ctx->record_type));
+ gimple_omp_task_set_arg_size (ctx->stmt, t);
t = build_int_cst (long_integer_type_node,
TYPE_ALIGN_UNIT (ctx->record_type));
- gimple_omp_task_set_arg_align (stmt, t);
+ gimple_omp_task_set_arg_align (ctx->stmt, t);
}
}
@@ -7112,6 +7172,8 @@ static unsigned int
execute_lower_omp (void)
{
gimple_seq body;
+ int i;
+ omp_context *ctx;
/* This pass always runs, to provide PROP_gimple_lomp.
But there is nothing to do unless -fopenmp is given. */
@@ -7124,6 +7186,9 @@ execute_lower_omp (void)
body = gimple_body (current_function_decl);
scan_omp (&body, NULL);
gcc_assert (taskreg_nesting_level == 0);
+ FOR_EACH_VEC_ELT (taskreg_contexts, i, ctx)
+ finish_taskreg_scan (ctx);
+ taskreg_contexts.release ();
if (all_contexts->root)
{
--- gcc/testsuite/c-c++-common/gomp/pr61200.c (revision 0)
+++ gcc/testsuite/c-c++-common/gomp/pr61200.c (revision 215835)
@@ -0,0 +1,13 @@
+/* PR libgomp/61200 */
+
+int
+main ()
+{
+ int var = 1;
+ #pragma omp parallel
+ if (var != 1)
+ __builtin_abort ();
+ #pragma omp task shared(var)
+ var = 2;
+ return 0;
+}
--- libgomp/testsuite/libgomp.c/pr61200.c (revision 0)
+++ libgomp/testsuite/libgomp.c/pr61200.c (revision 215835)
@@ -0,0 +1,87 @@
+/* PR libgomp/61200 */
+/* { dg-do run } */
+
+#include <omp.h>
+#include <stdlib.h>
+#include <unistd.h>
+
+volatile int x;
+
+void
+foo ()
+{
+ int var = 1;
+ int i;
+
+ for (i = 0; i < 2; i++)
+ {
+ if (i == 1)
+ {
+ #pragma omp parallel num_threads(2)
+ if (x)
+ var++;
+ else
+ {
+ #pragma omp single
+ sleep (2);
+ }
+ }
+ else
+ {
+ #pragma omp task shared(var)
+ {
+ sleep (1);
+ var = 2;
+ }
+ }
+ }
+ #pragma omp taskwait
+ if (var != 2)
+ abort ();
+}
+
+void
+bar ()
+{
+ int var = 1;
+ int i;
+
+ for (i = 0; i < 2; i++)
+ {
+ if (i == 0)
+ {
+ #pragma omp task shared(var)
+ {
+ sleep (1);
+ var = 2;
+ }
+ }
+ else
+ {
+ #pragma omp parallel num_threads(2)
+ if (x)
+ var++;
+ else
+ {
+ #pragma omp single
+ sleep (2);
+ }
+ }
+ }
+ #pragma omp taskwait
+ if (var != 2)
+ abort ();
+}
+
+int
+main ()
+{
+ omp_set_nested (1);
+ #pragma omp parallel num_threads(2)
+ #pragma omp single
+ foo ();
+ #pragma omp parallel num_threads(2)
+ #pragma omp single
+ bar ();
+ return 0;
+}
2014-11-28 Jakub Jelinek <ja...@redhat.com>
Backported from mainline
2014-10-31 Jakub Jelinek <ja...@redhat.com>
PR rtl-optimization/63659
* ree.c (update_reg_equal_equiv_notes): New function.
(combine_set_extension, transform_ifelse): Use it.
* gcc.c-torture/execute/pr63659.c: New test.
--- gcc/ree.c (revision 216984)
+++ gcc/ree.c (revision 216985)
@@ -261,6 +261,50 @@ typedef struct ext_cand
static int max_insn_uid;
+/* Update or remove REG_EQUAL or REG_EQUIV notes for INSN. */
+
+static bool
+update_reg_equal_equiv_notes (rtx insn, enum machine_mode new_mode,
+ enum machine_mode old_mode, enum rtx_code code)
+{
+ rtx *loc = ®_NOTES (insn);
+ while (*loc)
+ {
+ enum reg_note kind = REG_NOTE_KIND (*loc);
+ if (kind == REG_EQUAL || kind == REG_EQUIV)
+ {
+ rtx orig_src = XEXP (*loc, 0);
+ /* Update equivalency constants. Recall that RTL constants are
+ sign-extended. */
+ if (GET_CODE (orig_src) == CONST_INT
+ && HOST_BITS_PER_WIDE_INT >= GET_MODE_BITSIZE (new_mode))
+ {
+ if (INTVAL (orig_src) >= 0 || code == SIGN_EXTEND)
+ /* Nothing needed. */;
+ else
+ {
+ /* Zero-extend the negative constant by masking out the
+ bits outside the source mode. */
+ rtx new_const_int
+ = gen_int_mode (INTVAL (orig_src)
+ & GET_MODE_MASK (old_mode),
+ new_mode);
+ if (!validate_change (insn, &XEXP (*loc, 0),
+ new_const_int, true))
+ return false;
+ }
+ loc = &XEXP (*loc, 1);
+ }
+ /* Drop all other notes, they assume a wrong mode. */
+ else if (!validate_change (insn, loc, XEXP (*loc, 1), true))
+ return false;
+ }
+ else
+ loc = &XEXP (*loc, 1);
+ }
+ return true;
+}
+
/* Given a insn (CURR_INSN), an extension candidate for removal (CAND)
and a pointer to the SET rtx (ORIG_SET) that needs to be modified,
this code modifies the SET rtx to a new SET rtx that extends the
@@ -282,6 +326,7 @@ static bool
combine_set_extension (ext_cand *cand, rtx curr_insn, rtx *orig_set)
{
rtx orig_src = SET_SRC (*orig_set);
+ enum machine_mode orig_mode = GET_MODE (SET_DEST (*orig_set));
rtx new_reg = gen_rtx_REG (cand->mode, REGNO (SET_DEST (*orig_set)));
rtx new_set;
@@ -296,9 +341,8 @@ combine_set_extension (ext_cand *cand, r
{
/* Zero-extend the negative constant by masking out the bits outside
the source mode. */
- enum machine_mode src_mode = GET_MODE (SET_DEST (*orig_set));
rtx new_const_int
- = GEN_INT (INTVAL (orig_src) & GET_MODE_MASK (src_mode));
+ = GEN_INT (INTVAL (orig_src) & GET_MODE_MASK (orig_mode));
new_set = gen_rtx_SET (VOIDmode, new_reg, new_const_int);
}
}
@@ -336,7 +380,9 @@ combine_set_extension (ext_cand *cand, r
/* This change is a part of a group of changes. Hence,
validate_change will not try to commit the change. */
- if (validate_change (curr_insn, orig_set, new_set, true))
+ if (validate_change (curr_insn, orig_set, new_set, true)
+ && update_reg_equal_equiv_notes (curr_insn, cand->mode, orig_mode,
+ cand->code))
{
if (dump_file)
{
@@ -385,7 +431,9 @@ transform_ifelse (ext_cand *cand, rtx de
ifexpr = gen_rtx_IF_THEN_ELSE (cand->mode, cond, map_srcreg, map_srcreg2);
new_set = gen_rtx_SET (VOIDmode, map_dstreg, ifexpr);
- if (validate_change (def_insn, &PATTERN (def_insn), new_set, true))
+ if (validate_change (def_insn, &PATTERN (def_insn), new_set, true)
+ && update_reg_equal_equiv_notes (def_insn, cand->mode, GET_MODE (dstreg),
+ cand->code))
{
if (dump_file)
{
--- gcc/testsuite/gcc.c-torture/execute/pr63659.c (revision 0)
+++ gcc/testsuite/gcc.c-torture/execute/pr63659.c (revision 216985)
@@ -0,0 +1,29 @@
+/* PR rtl-optimization/63659 */
+
+int a, b, c, *d = &b, g, h, i;
+unsigned char e;
+char f;
+
+int
+main ()
+{
+ while (a)
+ {
+ for (a = 0; a; a++)
+ for (; c; c++)
+ ;
+ if (i)
+ break;
+ }
+
+ char j = c, k = -1, l;
+ l = g = j >> h;
+ f = l == 0 ? k : k % l;
+ e = 0 ? 0 : f;
+ *d = e;
+
+ if (b != 255)
+ __builtin_abort ();
+
+ return 0;
+}
2014-11-28 Jakub Jelinek <ja...@redhat.com>
Backported from mainline
2014-11-24 Jakub Jelinek <ja...@redhat.com>
PR fortran/63938
* trans-openmp.c (gfc_trans_omp_atomic): Make sure lhsaddr is
simple enough for goa_lhs_expr_p.
* libgomp.fortran/pr63938-1.f90: New test.
* libgomp.fortran/pr63938-2.f90: New test.
--- gcc/fortran/trans-openmp.c (revision 218030)
+++ gcc/fortran/trans-openmp.c (revision 218031)
@@ -1227,6 +1227,18 @@ gfc_trans_omp_atomic (gfc_code *code)
}
lhsaddr = save_expr (lhsaddr);
+ if (TREE_CODE (lhsaddr) != SAVE_EXPR
+ && (TREE_CODE (lhsaddr) != ADDR_EXPR
+ || TREE_CODE (TREE_OPERAND (lhsaddr, 0)) != VAR_DECL))
+ {
+ /* Make sure LHS is simple enough so that goa_lhs_expr_p can recognize
+ it even after unsharing function body. */
+ tree var = create_tmp_var_raw (TREE_TYPE (lhsaddr), NULL);
+ DECL_CONTEXT (var) = current_function_decl;
+ lhsaddr = build4 (TARGET_EXPR, TREE_TYPE (lhsaddr), var, lhsaddr,
+ NULL_TREE, NULL_TREE);
+ }
+
rhs = gfc_evaluate_now (rse.expr, &block);
if (atomic_code->ext.omp_atomic == GFC_OMP_ATOMIC_WRITE)
--- libgomp/testsuite/libgomp.fortran/pr63938-1.f90 (revision 0)
+++ libgomp/testsuite/libgomp.fortran/pr63938-1.f90 (revision 218031)
@@ -0,0 +1,14 @@
+! PR fortran/63938
+! { dg-do run }
+
+program pr63938_1
+ integer :: i, x(1)
+ x(1) = 0
+!$omp parallel do
+ do i = 1, 1000
+ !$omp atomic
+ x(1) = x(1) + 1
+ end do
+!$omp end parallel do
+ if (x(1) .ne. 1000) call abort
+end program pr63938_1
--- libgomp/testsuite/libgomp.fortran/pr63938-2.f90 (revision 0)
+++ libgomp/testsuite/libgomp.fortran/pr63938-2.f90 (revision 218031)
@@ -0,0 +1,18 @@
+! PR fortran/63938
+! { dg-do run }
+
+program pr63938_2
+ type t
+ integer :: x
+ end type
+ integer :: i
+ type(t) :: x
+ x%x = 0
+!$omp parallel do
+ do i = 1, 1000
+ !$omp atomic
+ x%x = x%x + 1
+ end do
+!$omp end parallel do
+ if (x%x .ne. 1000) call abort
+end program pr63938_2
2014-11-28 Jakub Jelinek <ja...@redhat.com>
Backported from mainline
2014-11-25 Jakub Jelinek <ja...@redhat.com>
PR preprocessor/60436
* line-map.c (linemap_line_start): If highest is above 0x60000000
and we are still tracking columns or highest is above 0x70000000,
force add_map.
--- libcpp/line-map.c (revision 218041)
+++ libcpp/line-map.c (revision 218042)
@@ -529,10 +529,10 @@ linemap_line_start (struct line_maps *se
&& line_delta * ORDINARY_MAP_NUMBER_OF_COLUMN_BITS (map) > 1000)
|| (max_column_hint >= (1U << ORDINARY_MAP_NUMBER_OF_COLUMN_BITS (map)))
|| (max_column_hint <= 80
- && ORDINARY_MAP_NUMBER_OF_COLUMN_BITS (map) >= 10))
- {
- add_map = true;
- }
+ && ORDINARY_MAP_NUMBER_OF_COLUMN_BITS (map) >= 10)
+ || (highest > 0x60000000
+ && (set->max_column_hint || highest > 0x70000000)))
+ add_map = true;
else
max_column_hint = set->max_column_hint;
if (add_map)
@@ -543,7 +543,7 @@ linemap_line_start (struct line_maps *se
/* If the column number is ridiculous or we've allocated a huge
number of source_locations, give up on column numbers. */
max_column_hint = 0;
- if (highest >0x70000000)
+ if (highest > 0x70000000)
return 0;
column_bits = 0;
}
2014-11-28 Jakub Jelinek <ja...@redhat.com>
Backported from mainline
2014-11-27 Jakub Jelinek <ja...@redhat.com>
PR middle-end/64067
* expr.c (expand_expr_addr_expr_1) <case COMPOUND_LITERAL_EXPR>:
Handle it by returning address of COMPOUND_LITERAL_EXPR_DECL
not only if modifier is EXPAND_INITIALIZER, but whenever
COMPOUND_LITERAL_EXPR_DECL is non-NULL and TREE_STATIC.
* gcc.c-torture/compile/pr64067.c: New test.
--- gcc/expr.c (revision 218120)
+++ gcc/expr.c (revision 218121)
@@ -7677,11 +7677,13 @@ expand_expr_addr_expr_1 (tree exp, rtx t
break;
case COMPOUND_LITERAL_EXPR:
- /* Allow COMPOUND_LITERAL_EXPR in initializers, if e.g.
- rtl_for_decl_init is called on DECL_INITIAL with
- COMPOUNT_LITERAL_EXPRs in it, they aren't gimplified. */
- if (modifier == EXPAND_INITIALIZER
- && COMPOUND_LITERAL_EXPR_DECL (exp))
+ /* Allow COMPOUND_LITERAL_EXPR in initializers or coming from
+ initializers, if e.g. rtl_for_decl_init is called on DECL_INITIAL
+ with COMPOUND_LITERAL_EXPRs in it, or ARRAY_REF on a const static
+ array with address of COMPOUND_LITERAL_EXPR in DECL_INITIAL;
+ the initializers aren't gimplified. */
+ if (COMPOUND_LITERAL_EXPR_DECL (exp)
+ && TREE_STATIC (COMPOUND_LITERAL_EXPR_DECL (exp)))
return expand_expr_addr_expr_1 (COMPOUND_LITERAL_EXPR_DECL (exp),
target, tmode, modifier, as);
/* FALLTHRU */
--- gcc/testsuite/gcc.c-torture/compile/pr64067.c (revision 0)
+++ gcc/testsuite/gcc.c-torture/compile/pr64067.c (revision 218121)
@@ -0,0 +1,10 @@
+/* PR middle-end/64067 */
+
+struct S { int s; };
+int *const v[1] = { &((struct S) { .s = 42 }).s };
+
+int *
+foo (void)
+{
+ return v[0];
+}
