[Bug middle-end/80423] [7 Regression] GC related -fcompare-debug failure
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=80423 Jakub Jelinek changed: What|Removed |Added Status|NEW |RESOLVED Resolution|--- |FIXED --- Comment #14 from Jakub Jelinek --- Fixed.
[Bug middle-end/80423] [7 Regression] GC related -fcompare-debug failure
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=80423 --- Comment #13 from Jakub Jelinek --- Author: jakub Date: Thu Apr 20 09:18:02 2017 New Revision: 247014 URL: https://gcc.gnu.org/viewcvs?rev=247014&root=gcc&view=rev Log: PR middle-end/80423 * tree.h (build_array_type): Add typeless_storage default argument. * tree.c (type_cache_hasher::equal): Also compare TYPE_TYPELESS_STORAGE flag for ARRAY_TYPEs. (build_array_type): Add typeless_storage argument, set TYPE_TYPELESS_STORAGE to it, if shared also hash it, and pass to recursive call. (build_nonshared_array_type): Adjust build_array_type_1 caller. (build_array_type): Likewise. Add typeless_storage argument. c-family/ * c-common.c (complete_array_type): Preserve TYPE_TYPELESS_STORAGE. cp/ * tree.c (build_cplus_array_type): Call build_array_type with the intended TYPE_TYPELESS_STORAGE flag value, instead of calling build_array_type and modifying later TYPE_TYPELESS_STORAGE on the shared type. testsuite/ * g++.dg/other/pr80423.C: New test. Added: trunk/gcc/testsuite/g++.dg/other/pr80423.C Modified: trunk/gcc/ChangeLog trunk/gcc/c-family/ChangeLog trunk/gcc/c-family/c-common.c trunk/gcc/cp/ChangeLog trunk/gcc/cp/tree.c trunk/gcc/testsuite/ChangeLog trunk/gcc/tree.c trunk/gcc/tree.h
[Bug middle-end/80423] [7 Regression] GC related -fcompare-debug failure
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=80423 Jakub Jelinek changed: What|Removed |Added Priority|P3 |P1
[Bug middle-end/80423] [7 Regression] GC related -fcompare-debug failure
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=80423 Jakub Jelinek changed: What|Removed |Added Attachment #41215|0 |1 is obsolete|| --- Comment #12 from Jakub Jelinek --- Created attachment 41218 --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=41218&action=edit gcc7-pr80423.patch Updated patch (still untested).
[Bug middle-end/80423] [7 Regression] GC related -fcompare-debug failure
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=80423
--- Comment #11 from Markus Trippelsdorf ---
With the proposed patch:
trippels@gcc2-power8 llvm_build % cat ClangOffloadBundler.ii
typedef unsigned char uint8_t;
struct A {
template A(unsigned char (&)[N]);
};
void fn1(A) {
uint8_t a[]{0};
fn1(a);
}
trippels@gcc2-power8 llvm_build % /home/trippels/gcc_test/usr/local/bin/g++ -c
ClangOffloadBundler.ii
ClangOffloadBundler.ii: In function ‘void fn1(A)’:
ClangOffloadBundler.ii:7:8: internal compiler error: canonical types differ for
identical types unsigned char [1] and uint8_t [1] {aka unsigned char [1]}
fn1(a);
^
0x1038d76b comptypes(tree_node*, tree_node*, int)
../../gcc/gcc/cp/typeck.c:1433
0x10191a53 reference_related_p(tree_node*, tree_node*)
../../gcc/gcc/cp/call.c:1442
0x101a68b3 reference_binding
../../gcc/gcc/cp/call.c:1595
0x101a793b implicit_conversion
../../gcc/gcc/cp/call.c:1837
0x101a9c5b add_function_candidate
../../gcc/gcc/cp/call.c:2196
0x101ab0c7 add_template_candidate_real
../../gcc/gcc/cp/call.c:3208
0x101abc57 add_template_candidate
../../gcc/gcc/cp/call.c:3252
0x101abc57 add_candidates
../../gcc/gcc/cp/call.c:5483
0x101a5933 build_user_type_conversion_1
../../gcc/gcc/cp/call.c:3799
0x101a7853 implicit_conversion
../../gcc/gcc/cp/call.c:1898
0x101a9c5b add_function_candidate
../../gcc/gcc/cp/call.c:2196
0x101aba8f add_candidates
../../gcc/gcc/cp/call.c:5496
0x101ac593 perform_overload_resolution
../../gcc/gcc/cp/call.c:4143
0x101afef3 build_new_function_call(tree_node*, vec**, bool, int)
../../gcc/gcc/cp/call.c:4232
0x103fd9f3 finish_call_expr(tree_node*, vec**,
bool, bool, int)
../../gcc/gcc/cp/semantics.c:2441
0x10348197 cp_parser_postfix_expression
../../gcc/gcc/cp/parser.c:6992
0x10348f9f cp_parser_unary_expression
../../gcc/gcc/cp/parser.c:8103
0x1034a613 cp_parser_cast_expression
../../gcc/gcc/cp/parser.c:8781
0x1034af87 cp_parser_binary_expression
../../gcc/gcc/cp/parser.c:8882
0x1034bab7 cp_parser_assignment_expression
../../gcc/gcc/cp/parser.c:9169
[Bug middle-end/80423] [7 Regression] GC related -fcompare-debug failure
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=80423 --- Comment #10 from Jakub Jelinek --- Created attachment 41215 --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=41215&action=edit gcc7-pr80423.patch Untested fix.
[Bug middle-end/80423] [7 Regression] GC related -fcompare-debug failure
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=80423
--- Comment #9 from Jakub Jelinek ---
I think the bug is in:
t = build_array_type (elt_type, index_type);
if (elt_type == unsigned_char_type_node
|| elt_type == signed_char_type_node
|| elt_type == char_type_node
|| (TREE_CODE (elt_type) == ENUMERAL_TYPE
&& TYPE_CONTEXT (elt_type) == std_node
&& !strcmp ("byte", TYPE_NAME_STRING (elt_type
TYPE_TYPELESS_STORAGE (t) = 1;
We modify a flag on a cached shared type. Now, as the flag is controlled
solely by the element type, if all the build_array_type{,_nelts} calls were
done only through this routine, it would just be set everywhere and not a
problem.
But that is not the case, the middle-end etc. also build array types and don't
set the typeless storage flag on it anywhere.
So, either the above needs to be a langhook and we'd need to force that flag on
every array building (but then we need to handle it somehow with LTO too),
or we must not share array types with/without the TYPE_TYPELESS_STORAGE flag.
I'll try to write an untested patch for the latter.
[Bug middle-end/80423] [7 Regression] GC related -fcompare-debug failure
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=80423 --- Comment #8 from Markus Trippelsdorf --- Created attachment 41212 --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=41212&action=edit somewhat reduced testcase
[Bug middle-end/80423] [7 Regression] GC related -fcompare-debug failure
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=80423 --- Comment #7 from Markus Trippelsdorf --- (In reply to Jakub Jelinek from comment #6) > Unfortunately can't reproduce, neither with a cross, nor on > powerpc64le-linux configured with ../configure > --enable-languages=c,c++,fortran . > Can this be reproduced on some GCCFarm box somewhere? Yes, on gcc112: trippels@gcc2-power8 ~ % /home/trippels/gcc_test/usr/local/bin/g++ -fcompare-debug --param ggc-min-expand=20 --param ggc-min-heapsize=0 -c -O3 SampleProfReader.ii -w
[Bug middle-end/80423] [7 Regression] GC related -fcompare-debug failure
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=80423 Jakub Jelinek changed: What|Removed |Added CC||jakub at gcc dot gnu.org --- Comment #6 from Jakub Jelinek --- Unfortunately can't reproduce, neither with a cross, nor on powerpc64le-linux configured with ../configure --enable-languages=c,c++,fortran . Can this be reproduced on some GCCFarm box somewhere?
[Bug middle-end/80423] [7 Regression] GC related -fcompare-debug failure
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=80423 Jeffrey A. Law changed: What|Removed |Added CC||law at redhat dot com Target Milestone|--- |7.0
[Bug middle-end/80423] [7 Regression] GC related -fcompare-debug failure
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=80423 Markus Trippelsdorf changed: What|Removed |Added Status|UNCONFIRMED |NEW Last reconfirmed||2017-04-14 Component|debug |middle-end Ever confirmed|0 |1 --- Comment #5 from Markus Trippelsdorf --- Started with r246866, aka PR79671 fix.
