* [Bug bootstrap/102766] [12 Regression][nvptx] internal compiler error: in df_insn_delete, at df-scan.c:949
2021-10-15 8:31 [Bug bootstrap/102766] New: [12 Regression] internal compiler error: in df_insn_delete, at df-scan.c:949 burnus at gcc dot gnu.org
@ 2021-10-15 9:17 ` burnus at gcc dot gnu.org
2021-10-15 9:22 ` marxin at gcc dot gnu.org
` (7 subsequent siblings)
8 siblings, 0 replies; 10+ messages in thread
From: burnus at gcc dot gnu.org @ 2021-10-15 9:17 UTC (permalink / raw)
To: gcc-bugs
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102766
Tobias Burnus <burnus at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
Target Milestone|--- |12.0
CC| |vries at gcc dot gnu.org
Summary|[12 Regression] internal |[12 Regression][nvptx]
|compiler error: in |internal compiler error: in
|df_insn_delete, at |df_insn_delete, at
|df-scan.c:949 |df-scan.c:949
--- Comment #1 from Tobias Burnus <burnus at gcc dot gnu.org> ---
Bisecting shows that the ICE is exposed due to the var-tracking changes:
commit g:4cb52980e5d5fb64a393d385923da1b51ab34606
Author: Martin Liska <mliska@suse.cz>
Date: Tue Oct 12 14:31:50 2021 +0200
Eliminate AUTODETECT_VALUE usage in options.
gcc/ChangeLog:
* common.opt: Stop using AUTODETECT_VALUE
and use EnabledBy where possible.
* opts.c: Enable OPT_fvar_tracking with optimize >= 1.
* toplev.c (AUTODETECT_VALUE): Remove macro.
(process_options): Simplify by using EnabledBy and
OPT_fvar_tracking. Use OPTION_SET_P macro instead of
AUTODETECT_VALUE.
^ permalink raw reply [flat|nested] 10+ messages in thread* [Bug bootstrap/102766] [12 Regression][nvptx] internal compiler error: in df_insn_delete, at df-scan.c:949
2021-10-15 8:31 [Bug bootstrap/102766] New: [12 Regression] internal compiler error: in df_insn_delete, at df-scan.c:949 burnus at gcc dot gnu.org
2021-10-15 9:17 ` [Bug bootstrap/102766] [12 Regression][nvptx] " burnus at gcc dot gnu.org
@ 2021-10-15 9:22 ` marxin at gcc dot gnu.org
2021-10-15 9:39 ` burnus at gcc dot gnu.org
` (6 subsequent siblings)
8 siblings, 0 replies; 10+ messages in thread
From: marxin at gcc dot gnu.org @ 2021-10-15 9:22 UTC (permalink / raw)
To: gcc-bugs
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102766
Martin Liška <marxin at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
Ever confirmed|0 |1
Last reconfirmed| |2021-10-15
Status|UNCONFIRMED |WAITING
Assignee|unassigned at gcc dot gnu.org |marxin at gcc dot gnu.org
--- Comment #2 from Martin Liška <marxin at gcc dot gnu.org> ---
Can you please provide a pre-processed source file?
^ permalink raw reply [flat|nested] 10+ messages in thread* [Bug bootstrap/102766] [12 Regression][nvptx] internal compiler error: in df_insn_delete, at df-scan.c:949
2021-10-15 8:31 [Bug bootstrap/102766] New: [12 Regression] internal compiler error: in df_insn_delete, at df-scan.c:949 burnus at gcc dot gnu.org
2021-10-15 9:17 ` [Bug bootstrap/102766] [12 Regression][nvptx] " burnus at gcc dot gnu.org
2021-10-15 9:22 ` marxin at gcc dot gnu.org
@ 2021-10-15 9:39 ` burnus at gcc dot gnu.org
2021-10-15 9:45 ` marxin at gcc dot gnu.org
` (5 subsequent siblings)
8 siblings, 0 replies; 10+ messages in thread
From: burnus at gcc dot gnu.org @ 2021-10-15 9:39 UTC (permalink / raw)
To: gcc-bugs
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102766
--- Comment #3 from Tobias Burnus <burnus at gcc dot gnu.org> ---
... However, there is no effect when using
-f(no-)var-tracking(-uninit),
it has still an ICE in all cases.
(Well, except that -fvar-tracking-uninit shows the following before the ICE:
cc1: error: variable tracking requested, but not supported by this debug format
[-Werror])
* * *
And I observe that the following option is now set:
1511 if (!OPTION_SET_P (flag_var_tracking_uninit))
1512 flag_var_tracking_uninit = flag_var_tracking;
While the code before is entered and explicitly sets it to 0:
(gdb) p debug_hooks->var_location == do_nothing_debug_hooks.var_location
$5 = true
1481 /* We know which debug output will be used so we can set
flag_var_tracking
1482 and flag_var_tracking_uninit if the user has not specified them.
*/
1483 if (debug_info_level < DINFO_LEVEL_NORMAL
1484 || !dwarf_debuginfo_p ()
1485 || debug_hooks->var_location ==
do_nothing_debug_hooks.var_location)
1486 {
1487 if ((OPTION_SET_P (flag_var_tracking) && flag_var_tracking == 1)
1488 || (OPTION_SET_P (flag_var_tracking_uninit)
1489 && flag_var_tracking_uninit == 1))
1490 {
1491 if (debug_info_level < DINFO_LEVEL_NORMAL)
1492 warning_at (UNKNOWN_LOCATION, 0,
1493 "variable tracking requested, but useless
unless "
1494 "producing debug info");
1495 else
1496 warning_at (UNKNOWN_LOCATION, 0,
1497 "variable tracking requested, but not supported
"
1498 "by this debug format");
1499 }
1500 flag_var_tracking = 0;
1501 flag_var_tracking_uninit = 0;
1502 }
^ permalink raw reply [flat|nested] 10+ messages in thread* [Bug bootstrap/102766] [12 Regression][nvptx] internal compiler error: in df_insn_delete, at df-scan.c:949
2021-10-15 8:31 [Bug bootstrap/102766] New: [12 Regression] internal compiler error: in df_insn_delete, at df-scan.c:949 burnus at gcc dot gnu.org
` (2 preceding siblings ...)
2021-10-15 9:39 ` burnus at gcc dot gnu.org
@ 2021-10-15 9:45 ` marxin at gcc dot gnu.org
2021-10-15 10:32 ` marxin at gcc dot gnu.org
` (4 subsequent siblings)
8 siblings, 0 replies; 10+ messages in thread
From: marxin at gcc dot gnu.org @ 2021-10-15 9:45 UTC (permalink / raw)
To: gcc-bugs
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102766
Martin Liška <marxin at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
See Also| |https://gcc.gnu.org/bugzill
| |a/show_bug.cgi?id=102585
--- Comment #4 from Martin Liška <marxin at gcc dot gnu.org> ---
Ok, so it's something close to PR102585.
^ permalink raw reply [flat|nested] 10+ messages in thread* [Bug bootstrap/102766] [12 Regression][nvptx] internal compiler error: in df_insn_delete, at df-scan.c:949
2021-10-15 8:31 [Bug bootstrap/102766] New: [12 Regression] internal compiler error: in df_insn_delete, at df-scan.c:949 burnus at gcc dot gnu.org
` (3 preceding siblings ...)
2021-10-15 9:45 ` marxin at gcc dot gnu.org
@ 2021-10-15 10:32 ` marxin at gcc dot gnu.org
2021-10-15 10:43 ` burnus at gcc dot gnu.org
` (3 subsequent siblings)
8 siblings, 0 replies; 10+ messages in thread
From: marxin at gcc dot gnu.org @ 2021-10-15 10:32 UTC (permalink / raw)
To: gcc-bugs
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102766
Martin Liška <marxin at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
Status|WAITING |ASSIGNED
--- Comment #5 from Martin Liška <marxin at gcc dot gnu.org> ---
Reduced test-case:
$ cat task.i
void gomp_thread();
void gomp_create_target_task() {
gomp_thread();
__builtin_unreachable();
}
$ /dev/shm/objdir2/gcc/xgcc -B/dev/shm/objdir2/gcc/ task.i -c -O2 -g
during RTL pass: final
task.i: In function ‘gomp_create_target_task’:
task.i:6:1: internal compiler error: in df_insn_delete, at df-scan.c:949
6 | }
| ^
0xcf8940 df_insn_delete(rtx_insn*)
/home/marxin/Programming/gcc/gcc/df-scan.c:949
0xc7fa5d delete_insn(rtx_insn*)
/home/marxin/Programming/gcc/gcc/cfgrtl.c:178
0x181533c reemit_marker_as_note
/home/marxin/Programming/gcc/gcc/var-tracking.c:10024
0x18162a6 delete_vta_debug_insn
/home/marxin/Programming/gcc/gcc/var-tracking.c:10366
0x18165a0 delete_vta_debug_insns(bool)
/home/marxin/Programming/gcc/gcc/var-tracking.c:10411
0xe26655 rest_of_handle_final
/home/marxin/Programming/gcc/gcc/final.c:4279
0xe26ad8 execute
/home/marxin/Programming/gcc/gcc/final.c:4363
Please submit a full bug report,
with preprocessed source if appropriate.
Please include the complete backtrace with any bug report.
See <https://gcc.gnu.org/bugs/> for instructions.
^ permalink raw reply [flat|nested] 10+ messages in thread* [Bug bootstrap/102766] [12 Regression][nvptx] internal compiler error: in df_insn_delete, at df-scan.c:949
2021-10-15 8:31 [Bug bootstrap/102766] New: [12 Regression] internal compiler error: in df_insn_delete, at df-scan.c:949 burnus at gcc dot gnu.org
` (4 preceding siblings ...)
2021-10-15 10:32 ` marxin at gcc dot gnu.org
@ 2021-10-15 10:43 ` burnus at gcc dot gnu.org
2021-10-21 12:31 ` [Bug middle-end/102766] " burnus at gcc dot gnu.org
` (2 subsequent siblings)
8 siblings, 0 replies; 10+ messages in thread
From: burnus at gcc dot gnu.org @ 2021-10-15 10:43 UTC (permalink / raw)
To: gcc-bugs
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102766
--- Comment #6 from Tobias Burnus <burnus at gcc dot gnu.org> ---
[x86_64 build/host with target nvptx]
.../gcc/cc1 -fpreprocessed task.i -quiet -dumpbase task.c -dumpbase-ext .c
-mgomp -g -O2 -Wall -Werror -version -o task.s
* * *
Another variant of reducted testcase:
struct gomp_task { void *fn_data; };
gomp_target_task_fn (void *data) { __builtin_unreachable (); }
gomp_create_target_task ()
{
struct gomp_task *task;
task = gomp_malloc (sizeof (*task));
gomp_target_task_fn (task->fn_data);
}
^ permalink raw reply [flat|nested] 10+ messages in thread* [Bug middle-end/102766] [12 Regression][nvptx] internal compiler error: in df_insn_delete, at df-scan.c:949
2021-10-15 8:31 [Bug bootstrap/102766] New: [12 Regression] internal compiler error: in df_insn_delete, at df-scan.c:949 burnus at gcc dot gnu.org
` (5 preceding siblings ...)
2021-10-15 10:43 ` burnus at gcc dot gnu.org
@ 2021-10-21 12:31 ` burnus at gcc dot gnu.org
2021-10-21 13:20 ` cvs-commit at gcc dot gnu.org
2021-10-21 13:21 ` marxin at gcc dot gnu.org
8 siblings, 0 replies; 10+ messages in thread
From: burnus at gcc dot gnu.org @ 2021-10-21 12:31 UTC (permalink / raw)
To: gcc-bugs
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102766
--- Comment #7 from Tobias Burnus <burnus at gcc dot gnu.org> ---
Cross ref – Martin's patch (current iteration) for this PR and PR102585 is at
https://gcc.gnu.org/pipermail/gcc-patches/2021-October/582103.html
^ permalink raw reply [flat|nested] 10+ messages in thread* [Bug middle-end/102766] [12 Regression][nvptx] internal compiler error: in df_insn_delete, at df-scan.c:949
2021-10-15 8:31 [Bug bootstrap/102766] New: [12 Regression] internal compiler error: in df_insn_delete, at df-scan.c:949 burnus at gcc dot gnu.org
` (6 preceding siblings ...)
2021-10-21 12:31 ` [Bug middle-end/102766] " burnus at gcc dot gnu.org
@ 2021-10-21 13:20 ` cvs-commit at gcc dot gnu.org
2021-10-21 13:21 ` marxin at gcc dot gnu.org
8 siblings, 0 replies; 10+ messages in thread
From: cvs-commit at gcc dot gnu.org @ 2021-10-21 13:20 UTC (permalink / raw)
To: gcc-bugs
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102766
--- Comment #8 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The master branch has been updated by Martin Liska <marxin@gcc.gnu.org>:
https://gcc.gnu.org/g:b4702276615ff8d43bc910e7a54fdd850ad8d461
commit r12-4608-gb4702276615ff8d43bc910e7a54fdd850ad8d461
Author: Martin Liska <mliska@suse.cz>
Date: Thu Oct 14 14:57:18 2021 +0200
options: Fix variable tracking option processing.
PR debug/102585
PR bootstrap/102766
gcc/ChangeLog:
* opts.c (finish_options): Process flag_var_tracking* options
here as they can be adjusted by optimize attribute.
Process also flag_syntax_only and flag_gtoggle.
* toplev.c (process_options): Remove it here.
* common.opt: Make debug_nonbind_markers_p as PerFunction
attribute as it depends on optimization level.
gcc/testsuite/ChangeLog:
* gcc.dg/pr102585.c: New test.
^ permalink raw reply [flat|nested] 10+ messages in thread* [Bug middle-end/102766] [12 Regression][nvptx] internal compiler error: in df_insn_delete, at df-scan.c:949
2021-10-15 8:31 [Bug bootstrap/102766] New: [12 Regression] internal compiler error: in df_insn_delete, at df-scan.c:949 burnus at gcc dot gnu.org
` (7 preceding siblings ...)
2021-10-21 13:20 ` cvs-commit at gcc dot gnu.org
@ 2021-10-21 13:21 ` marxin at gcc dot gnu.org
8 siblings, 0 replies; 10+ messages in thread
From: marxin at gcc dot gnu.org @ 2021-10-21 13:21 UTC (permalink / raw)
To: gcc-bugs
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=102766
Martin Liška <marxin at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
Status|ASSIGNED |RESOLVED
Resolution|--- |FIXED
--- Comment #9 from Martin Liška <marxin at gcc dot gnu.org> ---
Should be fixed now.
^ permalink raw reply [flat|nested] 10+ messages in thread