| Message ID | CAAgBjMk43ewF+CjrCKnEr60WuvhvAWppsRpfJS4kMg74BEr1wQ@mail.gmail.com |
|---|---|
| State | New |
| Headers | show |
On Sat, 4 Feb 2017, Prathamesh Kulkarni wrote: > Hi, > The following test-case ICE's with -fgimple: > > int __GIMPLE foo(int a) > { > int t1; > t1_1 = __builtin_abs (a); > return t1_1; > } > > gimplefe-2.c:4:3: internal compiler error: in get_callee_fndecl, at tree.c:9500 > t1_1 = __builtin_abs (a); > ^~~~ > 0xe96e8d get_callee_fndecl(tree_node const*) > ../../gcc/gcc/tree.c:9500 > 0x924d75 gimple_build_call_from_tree(tree_node*) > ../../gcc/gcc/gimple.c:351 > 0x6c86b3 c_parser_gimple_statement > ../../gcc/gcc/c/gimple-parser.c:393 > 0x6c86b3 c_parser_gimple_compound_statement > ../../gcc/gcc/c/gimple-parser.c:216 > 0x6c86b3 c_parser_parse_gimple_body(c_parser*) > ../../gcc/gcc/c/gimple-parser.c:93 > 0x6b04f1 c_parser_declaration_or_fndef > ../../gcc/gcc/c/c-parser.c:2081 > 0x6b883b c_parser_external_declaration > ../../gcc/gcc/c/c-parser.c:1464 > 0x6b92a1 c_parser_translation_unit > ../../gcc/gcc/c/c-parser.c:1344 > 0x6b92a1 c_parse_file() > ../../gcc/gcc/c/c-parser.c:18141 > 0x717832 c_common_parse_file() > ../../gcc/gcc/c-family/c-opts.c:1102 > > This happens because __builtin_abs(a) gets folded to <nop_expr<abs_expr<a>> > and get_callee_fndecl expects CALL_EXPR. > > The attached patch tries to fix the issue by building gimple_assign > with appropriate subcode > for functions that get folded to expression instead of trying to build > it as a function-call. > Is it OK to commit after bootstrap+test ? No. The proper fix is to not use the C frontend call-expr parsing and building -- it does have many more issues I think. Richard. > Thanks, > Prathamesh > -- Richard Biener <rguenther@suse.de> SUSE LINUX GmbH, GF: Felix Imendoerffer, Jane Smithard, Graham Norton, HRB 21284 (AG Nuernberg)
On Mon, 6 Feb 2017, Richard Biener wrote: > On Sat, 4 Feb 2017, Prathamesh Kulkarni wrote: > > > Hi, > > The following test-case ICE's with -fgimple: > > > > int __GIMPLE foo(int a) > > { > > int t1; > > t1_1 = __builtin_abs (a); > > return t1_1; > > } > > > > gimplefe-2.c:4:3: internal compiler error: in get_callee_fndecl, at tree.c:9500 > > t1_1 = __builtin_abs (a); > > ^~~~ > > 0xe96e8d get_callee_fndecl(tree_node const*) > > ../../gcc/gcc/tree.c:9500 > > 0x924d75 gimple_build_call_from_tree(tree_node*) > > ../../gcc/gcc/gimple.c:351 > > 0x6c86b3 c_parser_gimple_statement > > ../../gcc/gcc/c/gimple-parser.c:393 > > 0x6c86b3 c_parser_gimple_compound_statement > > ../../gcc/gcc/c/gimple-parser.c:216 > > 0x6c86b3 c_parser_parse_gimple_body(c_parser*) > > ../../gcc/gcc/c/gimple-parser.c:93 > > 0x6b04f1 c_parser_declaration_or_fndef > > ../../gcc/gcc/c/c-parser.c:2081 > > 0x6b883b c_parser_external_declaration > > ../../gcc/gcc/c/c-parser.c:1464 > > 0x6b92a1 c_parser_translation_unit > > ../../gcc/gcc/c/c-parser.c:1344 > > 0x6b92a1 c_parse_file() > > ../../gcc/gcc/c/c-parser.c:18141 > > 0x717832 c_common_parse_file() > > ../../gcc/gcc/c-family/c-opts.c:1102 > > > > This happens because __builtin_abs(a) gets folded to <nop_expr<abs_expr<a>> > > and get_callee_fndecl expects CALL_EXPR. > > > > The attached patch tries to fix the issue by building gimple_assign > > with appropriate subcode > > for functions that get folded to expression instead of trying to build > > it as a function-call. > > Is it OK to commit after bootstrap+test ? > > No. The proper fix is to not use the C frontend call-expr parsing > and building -- it does have many more issues I think. Sth as simple as for example fixes the bogus promotions inserted for short int __GIMPLE () foo (short int s) { short int D_1803; bb_2: D_1803 = s; L0: return D_1803; } int __GIMPLE () main (int argc, char * * argv) { short int s; int D_1805; int _1; short _2; bb_2: s = (short int) argc; _1 = (int) s; _2 = foo (_1); D_1805 = (int) _2; L0: return D_1805; } it should also fix the folding you see. Otherwise untested, of course (and we shouldn't build a CALL_EXPR but instead refactor this so we can build a GIMPLE_CALL directly) Richard.Index: gcc/c/gimple-parser.c =================================================================== --- gcc/c/gimple-parser.c (revision 245203) +++ gcc/c/gimple-parser.c (working copy) @@ -946,17 +946,11 @@ orig_expr = expr; start = expr.get_start (); finish = c_parser_tokens_buf (parser, 0)->get_finish (); - expr.value = c_build_function_call_vec (expr_loc, arg_loc, - expr.value, - exprlist, origtypes); + expr.value = build_call_vec (TREE_TYPE (TREE_TYPE (expr.value)), + expr.value, exprlist); + SET_EXPR_LOCATION (expr.value, expr_loc); set_c_expr_source_range (&expr, start, finish); - expr.original_code = ERROR_MARK; - if (TREE_CODE (expr.value) == INTEGER_CST - && TREE_CODE (orig_expr.value) == FUNCTION_DECL - && DECL_BUILT_IN_CLASS (orig_expr.value) == BUILT_IN_NORMAL - && DECL_FUNCTION_CODE (orig_expr.value) == BUILT_IN_CONSTANT_P) - expr.original_code = C_MAYBE_CONST_EXPR; expr.original_type = NULL; if (exprlist) {
diff --git a/gcc/c/gimple-parser.c b/gcc/c/gimple-parser.c index 7feb6d0..4e2ae37 100644 --- a/gcc/c/gimple-parser.c +++ b/gcc/c/gimple-parser.c @@ -401,10 +401,22 @@ c_parser_gimple_statement (c_parser *parser, gimple_seq *seq) rhs = c_parser_gimple_unary_expression (parser); if (rhs.value != error_mark_node) { - gimple *call = gimple_build_call_from_tree (rhs.value); - gimple_call_set_lhs (call, lhs.value); - gimple_seq_add_stmt (seq, call); - gimple_set_location (call, loc); + if (TREE_CODE (rhs.value) == CALL_EXPR) + { + gimple *call = gimple_build_call_from_tree (rhs.value); + gimple_call_set_lhs (call, lhs.value); + gimple_seq_add_stmt (seq, call); + gimple_set_location (call, loc); + } + else + { + tree value = tree_strip_nop_conversions (rhs.value); + gcc_assert (get_gimple_rhs_class (TREE_CODE (value)) + != GIMPLE_INVALID_RHS); + gassign *assign = gimple_build_assign (lhs.value, value); + gimple_seq_add_stmt (seq, assign); + gimple_set_location (assign, loc); + } } return; } diff --git a/gcc/gimple-pretty-print.c b/gcc/gimple-pretty-print.c index 91c839d..f773d85 100644 --- a/gcc/gimple-pretty-print.c +++ b/gcc/gimple-pretty-print.c @@ -323,9 +323,18 @@ dump_unary_rhs (pretty_printer *buffer, gassign *gs, int spc, int flags) break; case ABS_EXPR: - pp_string (buffer, "ABS_EXPR <"); - dump_generic_node (buffer, rhs, spc, flags, false); - pp_greater (buffer); + if (flags & TDF_GIMPLE) + { + pp_string (buffer, "__builtin_abs ("); + dump_generic_node (buffer, rhs, spc, flags, false); + pp_right_paren (buffer); + } + else + { + pp_string (buffer, "ABS_EXPR <"); + dump_generic_node (buffer, rhs, spc, flags, false); + pp_greater (buffer); + } break; default: diff --git a/gcc/testsuite/gcc.dg/gimplefe-23.c b/gcc/testsuite/gcc.dg/gimplefe-23.c new file mode 100644 index 0000000..1448030 --- /dev/null +++ b/gcc/testsuite/gcc.dg/gimplefe-23.c @@ -0,0 +1,11 @@ +/* { dg-do compile } */ +/* { dg-options "-O -fgimple -fdump-tree-ssa-gimple" } */ + +int __GIMPLE foo(int a) +{ + int t1; + t1_1 = __builtin_abs (a); + return t1_1; +} + +/* { dg-final { scan-tree-dump "__builtin_abs" "ssa" } } */ diff --git a/gcc/tree-pretty-print.c b/gcc/tree-pretty-print.c index d823c2e..f1a5bdd 100644 --- a/gcc/tree-pretty-print.c +++ b/gcc/tree-pretty-print.c @@ -2441,9 +2441,18 @@ dump_generic_node (pretty_printer *pp, tree node, int spc, int flags, break; case ABS_EXPR: - pp_string (pp, "ABS_EXPR <"); - dump_generic_node (pp, TREE_OPERAND (node, 0), spc, flags, false); - pp_greater (pp); + if (flags & TDF_GIMPLE) + { + pp_string (pp, "__builtin_abs ("); + dump_generic_node (pp, TREE_OPERAND (node, 0), spc, flags, false); + pp_right_paren (pp); + } + else + { + pp_string (pp, "ABS_EXPR <"); + dump_generic_node (pp, TREE_OPERAND (node, 0), spc, flags, false); + pp_greater (pp); + } break; case RANGE_EXPR:
Hi, The following test-case ICE's with -fgimple: int __GIMPLE foo(int a) { int t1; t1_1 = __builtin_abs (a); return t1_1; } gimplefe-2.c:4:3: internal compiler error: in get_callee_fndecl, at tree.c:9500 t1_1 = __builtin_abs (a); ^~~~ 0xe96e8d get_callee_fndecl(tree_node const*) ../../gcc/gcc/tree.c:9500 0x924d75 gimple_build_call_from_tree(tree_node*) ../../gcc/gcc/gimple.c:351 0x6c86b3 c_parser_gimple_statement ../../gcc/gcc/c/gimple-parser.c:393 0x6c86b3 c_parser_gimple_compound_statement ../../gcc/gcc/c/gimple-parser.c:216 0x6c86b3 c_parser_parse_gimple_body(c_parser*) ../../gcc/gcc/c/gimple-parser.c:93 0x6b04f1 c_parser_declaration_or_fndef ../../gcc/gcc/c/c-parser.c:2081 0x6b883b c_parser_external_declaration ../../gcc/gcc/c/c-parser.c:1464 0x6b92a1 c_parser_translation_unit ../../gcc/gcc/c/c-parser.c:1344 0x6b92a1 c_parse_file() ../../gcc/gcc/c/c-parser.c:18141 0x717832 c_common_parse_file() ../../gcc/gcc/c-family/c-opts.c:1102 This happens because __builtin_abs(a) gets folded to <nop_expr<abs_expr<a>> and get_callee_fndecl expects CALL_EXPR. The attached patch tries to fix the issue by building gimple_assign with appropriate subcode for functions that get folded to expression instead of trying to build it as a function-call. Is it OK to commit after bootstrap+test ? Thanks, Prathamesh 2017-02-04 Prathamesh Kulkarni <prathamesh.kulkarni@linaro.org> * gimple-pretty-print.c (dump_unary_rhs): Change dump format for ABS_EXPR for gimple dump. * tree-pretty-print.c (dump_generic_node): Likewise. c/ * gimple-parser.c (c_parser_gimple_statement): Check if rhs.value is CALL_EXPR and build gimple_assign if it isn't. testsuite/ * gcc.dg/gimplefe-23.c: New test-case.