https://gcc.gnu.org/bugzilla/show_bug.cgi?id=88147
--- Comment #25 from Martin Liška ---
Author: marxin
Date: Mon Mar 11 09:37:41 2019
New Revision: 269570
URL: https://gcc.gnu.org/viewcvs?rev=269570=gcc=rev
Log:
Backport r268789
2019-03-11 Martin Liska
Backport from mainline
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=88147
--- Comment #24 from Martin Liška ---
(In reply to David Malcolm from comment #22)
> Should be fixed by r268789.
Nice, thanks for test and cooperation.
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=88147
--- Comment #23 from Martin Liška ---
>
> The purpose of the check is impose a limit on the size of the jumps that
> occur within the location_t representation, so that if there's a big jump in
> line numbers, we start a linemap, where big is
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=88147
David Malcolm changed:
What|Removed |Added
Status|ASSIGNED|RESOLVED
Resolution|---
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=88147
--- Comment #21 from David Malcolm ---
Author: dmalcolm
Date: Tue Feb 12 01:09:31 2019
New Revision: 268789
URL: https://gcc.gnu.org/viewcvs?rev=268789=gcc=rev
Log:
linemap_line_start: protect against location_t overflow (PR lto/88147)
PR
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=88147
--- Comment #20 from David Malcolm ---
(In reply to David Malcolm from comment #19)
> (In reply to David Malcolm from comment #17)
> > Created attachment 45660 [details]
> > Selftest coverage
> >
> > The attached reproduces the problem via a
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=88147
--- Comment #19 from David Malcolm ---
(In reply to David Malcolm from comment #17)
> Created attachment 45660 [details]
> Selftest coverage
>
> The attached reproduces the problem via a minimal selftest, and is also
> fixed by attachment 45653
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=88147
--- Comment #18 from David Malcolm ---
(In reply to Martin Liška from comment #15)
> During the bug investigation I noticed a strange thing in line-map.c:
>
>700if (line_delta < 0
>701|| (line_delta > 10
>702
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=88147
--- Comment #17 from David Malcolm ---
Created attachment 45660
--> https://gcc.gnu.org/bugzilla/attachment.cgi?id=45660=edit
Selftest coverage
The attached reproduces the problem via a minimal selftest, and is also fixed
by attachment 45653.
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=88147
--- Comment #16 from Martin Liška ---
(In reply to Martin Liška from comment #14)
> Created attachment 45653 [details]
> Patch candidate
>
> Patch candidate that survives both reduced and not reduced test-cases.
> David does the patch make
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=88147
--- Comment #15 from Martin Liška ---
During the bug investigation I noticed a strange thing in line-map.c:
700if (line_delta < 0
701|| (line_delta > 10
702&& line_delta * map->m_column_and_range_bits > 1000)
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=88147
--- Comment #14 from Martin Liška ---
Created attachment 45653
--> https://gcc.gnu.org/bugzilla/attachment.cgi?id=45653=edit
Patch candidate
Patch candidate that survives both reduced and not reduced test-cases.
David does the patch make
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=88147
--- Comment #13 from Martin Liška ---
I hope I understand the failure. So what happens:
- first a new map is created with to_line: 2578 (with m_column_and_range_bits
== 13)
- then a new request comes with to_line: 404198,
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=88147
Martin Liška changed:
What|Removed |Added
Status|NEW |ASSIGNED
Assignee|unassigned
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=88147
--- Comment #11 from David Malcolm ---
Created attachment 45610
--> https://gcc.gnu.org/bugzilla/attachment.cgi?id=45610=edit
Partially reduced testcase
This is 975366 bytes (decompressed) and takes 0.3-0.4 seconds to crash
r265875's lto1, as
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=88147
--- Comment #10 from David Malcolm ---
FWIW I'm able to reproduce this with r265875 and am running a reduction script
over this weekend to see if I can isolate what the issue is/was.
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=88147
Jakub Jelinek changed:
What|Removed |Added
CC||jakub at gcc dot gnu.org
--- Comment #9
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=88147
--- Comment #8 from Martin Liška ---
(In reply to David Malcolm from comment #7)
> I've been trying to reproduce this, but failing - I tried with today's
> trunk, and with a build from 2018-11-16.
>
> Do you have a revision that is known to
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=88147
--- Comment #7 from David Malcolm ---
I've been trying to reproduce this, but failing - I tried with today's trunk,
and with a build from 2018-11-16.
Do you have a revision that is known to trigger the ICE?
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=88147
--- Comment #6 from Martin Liška ---
David can you please help us with that? Would it be possible to write a
sanitization patch will catch an overflow?
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=88147
Richard Biener changed:
What|Removed |Added
Priority|P3 |P1
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=88147
--- Comment #5 from Martin Liška ---
Can't reduce that, not a surprise.
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=88147
Jan Hubicka changed:
What|Removed |Added
Status|ASSIGNED|NEW
Assignee|hubicka at gcc dot
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=88147
--- Comment #3 from Jan Hubicka ---
If it is simply location overflow it likely won't reduce into something simple
:(
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=88147
--- Comment #2 from Martin Liška ---
(In reply to Jan Hubicka from comment #1)
> Hmm, this looks like another overfow in line map - my understanding is that
> the assert checks that correct line number is added. I am not quite
> line_map expert
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=88147
--- Comment #1 from Jan Hubicka ---
Hmm, this looks like another overfow in line map - my understanding is that the
assert checks that correct line number is added. I am not quite line_map
expert :)
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=88147
Richard Biener changed:
What|Removed |Added
Target Milestone|--- |9.0
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=88147
Martin Liška changed:
What|Removed |Added
Status|UNCONFIRMED |ASSIGNED
Last reconfirmed|
28 matches
Mail list logo
