[Bug c/82050] [8 Regression] ICE on invalid code on x86_64-linux-gnu in column_range, at diagnostic-show-locus.c:1403
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=82050 --- Comment #9 from David Malcolm --- Author: dmalcolm Date: Tue Dec 12 01:31:30 2017 New Revision: 255563 URL: https://gcc.gnu.org/viewcvs?rev=255563=gcc=rev Log: Expensive selftests: torture testing for fix-it boundary conditions (PR c/82050) This patch adds selftest coverage for the fix for PR c/82050. The selftest iterates over various "interesting" column and line-width values to try to shake out bugs in the fix-it printing routines, a kind of "torture" selftest. Unfortunately this selftest is noticably slower than the other selftests; adding it to diagnostic-show-locus.c led to: -fself-test: 40218 pass(es) in 0.172000 seconds slowing down to: -fself-test: 97315 pass(es) in 6.109000 seconds for an unoptimized build (e.g. when hacking with --disable-bootstrap). Given that this affects the compile-edit-test cycle of the "gcc" subdirectory, this felt like an unacceptable amount of overhead to add. I attempted to optimize the test by reducing the amount of coverage, but the test seems useful, and there seems to be a valid role for "torture" selftests. Hence this patch adds a: gcc.dg/plugin/expensive_selftests_plugin.c with the responsibility for running "expensive" selftests, and adds the expensive test there. The patch moves a small amount of code from selftest::run_tests into a helper class so that the plugin can print a useful summary line (to reassure us that the tests are actually being run). With that, the compile-edit-test cycle of the "gcc" subdir is unaffected; the plugin takes: expensive_selftests_plugin: 26641 pass(es) in 3.127000 seconds which seems reasonable within the much longer time taken by "make check" (I optimized some of the overhead away, hence the reduction from 6 seconds above down to 3 seconds). gcc/ChangeLog: PR c/82050 * selftest-run-tests.c (selftest::run_tests): Move start/finish code to... * selftest.c (selftest::test_runner::test_runner): New ctor. (selftest::test_runner::~test_runner): New dtor. * selftest.h (class selftest::test_runner): New class. gcc/testsuite/ChangeLog: PR c/82050 * gcc.dg/plugin/expensive-selftests-1.c: New file. * gcc.dg/plugin/expensive_selftests_plugin.c: New file. * gcc.dg/plugin/plugin.exp (plugin_test_list): Add the above. Added: trunk/gcc/testsuite/gcc.dg/plugin/expensive-selftests-1.c trunk/gcc/testsuite/gcc.dg/plugin/expensive_selftests_plugin.c Modified: trunk/gcc/ChangeLog trunk/gcc/selftest-run-tests.c trunk/gcc/selftest.c trunk/gcc/selftest.h trunk/gcc/testsuite/ChangeLog trunk/gcc/testsuite/gcc.dg/plugin/plugin.exp
[Bug c/82050] [8 Regression] ICE on invalid code on x86_64-linux-gnu in column_range, at diagnostic-show-locus.c:1403
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=82050 --- Comment #8 from David Malcolm --- Author: dmalcolm Date: Tue Nov 28 21:45:56 2017 New Revision: 255219 URL: https://gcc.gnu.org/viewcvs?rev=255219=gcc=rev Log: Handle very long lines when printing fix-it hints When fixing PR c/82050 I noticed a bug in how we print fix-it hints for very long lines: we weren't taking into account the x-offset for the line when printing the fix-it hint. This could lead to output where instead of printing: foo.c:14:3944: error: etc = foo.field ^ replacement where the lines have been offset to start printing at about column 3900, the "replacement" line was erroneously *not* offset, and was thus prefixed by thousands of spaces, leading to large whitespace gaps in the output, and the replacement failing to line up with the source to be replaced. Fixed thusly. gcc/ChangeLog: * diagnostic-show-locus.c (layout::print_trailing_fixits): Handle m_x_offset. (layout::move_to_column): Likewise. gcc/testsuite/ChangeLog: * gcc.dg/plugin/diagnostic-test-show-locus-bw.c (test_very_wide_line): Update expected output to include a fix-it hint. * gcc.dg/plugin/diagnostic-test-show-locus-color.c (test_very_wide_line): Likewise. * gcc.dg/plugin/diagnostic_plugin_test_show_locus.c (test_show_locus): Add a fix-it hint to "test_very_wide_line". Modified: trunk/gcc/ChangeLog trunk/gcc/diagnostic-show-locus.c trunk/gcc/testsuite/ChangeLog trunk/gcc/testsuite/gcc.dg/plugin/diagnostic-test-show-locus-bw.c trunk/gcc/testsuite/gcc.dg/plugin/diagnostic-test-show-locus-color.c trunk/gcc/testsuite/gcc.dg/plugin/diagnostic_plugin_test_show_locus.c
[Bug c/82050] [8 Regression] ICE on invalid code on x86_64-linux-gnu in column_range, at diagnostic-show-locus.c:1403
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=82050 David Malcolm changed: What|Removed |Added Status|ASSIGNED|RESOLVED Resolution|--- |FIXED --- Comment #7 from David Malcolm --- Fixed on trunk for gcc 8 by r255214. Candidate followup patch for test coverage: https://gcc.gnu.org/ml/gcc-patches/2017-11/msg02459.html
[Bug c/82050] [8 Regression] ICE on invalid code on x86_64-linux-gnu in column_range, at diagnostic-show-locus.c:1403
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=82050 --- Comment #6 from David Malcolm --- Author: dmalcolm Date: Tue Nov 28 19:24:35 2017 New Revision: 255214 URL: https://gcc.gnu.org/viewcvs?rev=255214=gcc=rev Log: Reject fix-it hints for various awkward boundary cases (PR c/82050) PR c/82050 reports a failed assertion deep within diagnostic_show_locus's code for printing fix-it hints. The root cause is a fix-it hint suggesting a textual replacement, where the affected column numbers straddle the LINE_MAP_MAX_COLUMN_NUMBER boundary, so that the start of the range has a column number, but the end of the range doesn't. The fix is to verify that the column numbers are sane when adding fix-it hints to a rich_location, rejecting fix-it hints where they are not. libcpp/ChangeLog: PR c/82050 * include/line-map.h (LINE_MAP_MAX_COLUMN_NUMBER): Move here. * line-map.c (LINE_MAP_MAX_COLUMN_NUMBER): ...from here. (rich_location::maybe_add_fixit): Reject fix-it hints in which the start column exceeds the next column. Modified: trunk/libcpp/ChangeLog trunk/libcpp/include/line-map.h trunk/libcpp/line-map.c
[Bug c/82050] [8 Regression] ICE on invalid code on x86_64-linux-gnu in column_range, at diagnostic-show-locus.c:1403
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=82050 David Malcolm changed: What|Removed |Added Status|NEW |ASSIGNED Assignee|unassigned at gcc dot gnu.org |dmalcolm at gcc dot gnu.org --- Comment #5 from David Malcolm --- Thanks!
[Bug c/82050] [8 Regression] ICE on invalid code on x86_64-linux-gnu in column_range, at diagnostic-show-locus.c:1403
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=82050 --- Comment #4 from Jakub Jelinek --- Created attachment 42697 --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=42697=edit pr82050.c Here it is.
[Bug c/82050] [8 Regression] ICE on invalid code on x86_64-linux-gnu in column_range, at diagnostic-show-locus.c:1403
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=82050 --- Comment #3 from David Malcolm --- I attempted to copy from comment #0, but couldn't reproduce the ICE (possibly a spaces vs tabs thing?) Jakub: please can you attach the reproducer file, and I'll take this. Thanks.
[Bug c/82050] [8 Regression] ICE on invalid code on x86_64-linux-gnu in column_range, at diagnostic-show-locus.c:1403
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=82050
Jakub Jelinek changed:
What|Removed |Added
CC||jakub at gcc dot gnu.org
--- Comment #2 from Jakub Jelinek ---
In:
1418static column_range
1419get_affected_columns (const fixit_hint *hint)
1420{
1421 int start_column = LOCATION_COLUMN (hint->get_start_loc ());
1422 int finish_column = LOCATION_COLUMN (hint->get_next_loc ()) - 1;
1423
1424 return column_range (start_column, finish_column);
(gdb) p hint->get_start_loc ()
$3 = 438112
(gdb) p hint->get_next_loc ()
$4 = 307392
(gdb) p expand_location ($3)
$5 = {file = 0x7fffe3d1 "pr82050.c", line = 14, column = 4086, data = 0x0,
sysp = false}
(gdb) p expand_location ($4)
$6 = {file = 0x7fffe3d1 "pr82050.c", line = 14, column = 1, data = 0x0,
sysp = false}
so finish_column is 0 and smaller than start_column, which is what column_range
ICEs on. Is that because we have:
/* Do not track column numbers higher than this one. As a result, the
range of column_bits is [12, 18] (or 0 if column numbers are
disabled). */
const unsigned int LINE_MAP_MAX_COLUMN_NUMBER = (1U << 12);
and thus effectively can only track columns up to 4095 (or 4096 if 1 based)?
If so, something should punt on it rather than ICE it is not possible.
[Bug c/82050] [8 Regression] ICE on invalid code on x86_64-linux-gnu in column_range, at diagnostic-show-locus.c:1403
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=82050 Richard Biener changed: What|Removed |Added Target Milestone|--- |8.0
[Bug c/82050] [8 Regression] ICE on invalid code on x86_64-linux-gnu in column_range, at diagnostic-show-locus.c:1403
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=82050 Martin Liška changed: What|Removed |Added Keywords||ice-on-valid-code Status|UNCONFIRMED |NEW Last reconfirmed||2017-08-31 CC||dmalcolm at gcc dot gnu.org, ||marxin at gcc dot gnu.org Summary|ICE on invalid code on |[8 Regression] ICE on |x86_64-linux-gnu in |invalid code on |column_range, at|x86_64-linux-gnu in |diagnostic-show-locus.c:140 |column_range, at |3 |diagnostic-show-locus.c:140 ||3 Ever confirmed|0 |1 Known to fail||8.0 --- Comment #1 from Martin Liška --- Started with r250187.
