[Bug sanitizer/89308] [8 only] The sanitizers do no longer work on GCC 8 with newer kernels

2019-02-20 Thread segher at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=89308

Segher Boessenkool  changed:

   What|Removed |Added

 Status|NEW |SUSPENDED

--- Comment #9 from Segher Boessenkool  ---
We do not currently see problems in practice.  Suspending.

[Bug sanitizer/89308] [8 only] The sanitizers do no longer work on GCC 8 with newer kernels

2019-02-18 Thread rguenth at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=89308

Richard Biener  changed:

   What|Removed |Added

   Priority|P1  |P2

[Bug sanitizer/89308] [8 only] The sanitizers do no longer work on GCC 8 with newer kernels

2019-02-14 Thread seurer at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=89308

--- Comment #8 from seurer at gcc dot gnu.org ---
This is the way it came from upstream (llvm) and the solution for powerpc64 was
copied from what aarch64 did before.

What is really needed is a workable solution from whoever does sanitizer
development that works despite the huge ranges of addresses that ASLR now uses.

[Bug sanitizer/89308] [8 only] The sanitizers do no longer work on GCC 8 with newer kernels

2019-02-14 Thread jakub at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=89308

--- Comment #7 from Jakub Jelinek  ---
Still, a reexec is costly and might break some programs.  If the ASLR makes
problems only sometimes, it might be better to try to map stuff it wants and if
that fails, before reporting failure try this CheckASLR.

[Bug sanitizer/89308] [8 only] The sanitizers do no longer work on GCC 8 with newer kernels

2019-02-14 Thread seurer at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=89308

--- Comment #6 from seurer at gcc dot gnu.org ---
I think it only comes out if you specify the verbose sanitizer option on the
compilation.  If I can remember how to specify that I will try it.

[Bug sanitizer/89308] [8 only] The sanitizers do no longer work on GCC 8 with newer kernels

2019-02-14 Thread jakub at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=89308

--- Comment #5 from Jakub Jelinek  ---
Ugh, does this mean if ASLR is enabled you get
"WARNING: Program is being run with address space layout "
"randomization (ASLR) enabled which prevents the thread and "
"memory sanitizers from working on powerpc64le.\n"
"ASLR will be disabled and the program re-executed.\n"
message from every -fsanitize=address/-fsanitize=thread linked program?
If so, that is extremely nasty.  Couldn't that be done only if you determine
the areas you need to mmap the shadow etc. memory can't be mapped?
Of course, that applies to trunk as well as possible backport.

[Bug sanitizer/89308] [8 only] The sanitizers do no longer work on GCC 8 with newer kernels

2019-02-14 Thread seurer at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=89308

--- Comment #4 from seurer at gcc dot gnu.org ---
The above patch pulls in just enough of the changes from trunk to disable ASLR
for powerpc64 while leaving things alone for everyone else.

[Bug sanitizer/89308] [8 only] The sanitizers do no longer work on GCC 8 with newer kernels

2019-02-14 Thread seurer at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=89308

--- Comment #3 from seurer at gcc dot gnu.org ---
Created attachment 45725
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=45725=edit
Patch to disable ALSR for asan/tsan on powerpc64

[Bug sanitizer/89308] [8 only] The sanitizers do no longer work on GCC 8 with newer kernels

2019-02-12 Thread seurer at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=89308

--- Comment #2 from seurer at gcc dot gnu.org ---
I looked through the various patches for the sanitizer stuff and the important
one that is missing is the one that disables ASLR when the sanitizers are in
use.  This would "solve" the issue where the thread sanitizer tests fail
depending on how ASLR has laid out memory (something I see all the time with
gcc-8 and recent kernels).

However, the thing that prompted this bug report

https://github.com/openbmc/openbmc-build-scripts/commit/282e33011ea62fe80bff2c5ab0e9ef9446d6c551

is dealing with the *address* sanitizer and I haven't seen that sort of
problem.  It looks like it occurs when they are building openbmc which I have
never done.

[Bug sanitizer/89308] [8 only] The sanitizers do no longer work on GCC 8 with newer kernels

2019-02-12 Thread jakub at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=89308

--- Comment #1 from Jakub Jelinek  ---
So, which exact changes (e.g. upstream svn revisions) we are talking about
here?

[Bug sanitizer/89308] [8 only] The sanitizers do no longer work on GCC 8 with newer kernels

2019-02-12 Thread pinskia at gcc dot gnu.org
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=89308

Andrew Pinski  changed:

   What|Removed |Added

 Status|UNCONFIRMED |NEW
   Last reconfirmed||2019-02-12
  Known to work||9.0
   Target Milestone|--- |8.4
Summary|The sanitizers do no longer |[8 only] The sanitizers do
   |work on GCC 8 with newer|no longer work on GCC 8
   |kernels |with newer kernels
 Ever confirmed|0   |1