[Bug d/103944] New: [12 Regression] Testsuite hang due to libphobos/testsuite/libphobos.gc/forkgc2.d

[Bug d/103944] New: [12 Regression] Testsuite hang due to libphobos/testsuite/libphobos.gc/forkgc2.d

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103944

            Bug ID: 103944
           Summary: [12 Regression] Testsuite hang due to
                    libphobos/testsuite/libphobos.gc/forkgc2.d
           Product: gcc
           Version: 12.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: d
          Assignee: ibuclaw at gdcproject dot org
          Reporter: jakub at gcc dot gnu.org
  Target Milestone: ---

When trying to build gcc 12 snapshot in fedora using rpm, make check hangs, but
what is weird is that the make process has no children and is stuck waiting on
read from a jobserver file descriptor.
Strangely, during my (almost daily) normal bootstraps/regtests I'm not seeing
this.
I think I've narrowed this down to the libphobos.gc/forkgc2.d test.
When running
make -j32 -k check-target-libphobos ALT_CC_UNDER_TEST=gcc
ALT_CXX_UNDER_TEST=g++
RUNTESTFLAGS=--target_board=unix/'{,-fstack-protector-strong}'
I see it hang, again no children, but looking in detail into
libphobos/testsuite/ I see forkgc2.exe in an unfinished log there and
./forkgc2.exe process,
killing it manually make it succeed.

[Bug d/103944] [12 Regression] Testsuite hang due to libphobos/testsuite/libphobos.gc/forkgc2.d

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103944

--- Comment #1 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
Looking at the testcase, in fork child from a multi-threaded process only async
signal safe functions can be called until it execs, are you sure nothing else
is called?

[Bug d/103944] [12 Regression] Testsuite hang due to libphobos/testsuite/libphobos.gc/forkgc2.d

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103944

Andrew Pinski <pinskia at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|---                         |12.0
           Keywords|                            |testsuite-fail

[Bug d/103944] [12 Regression] Testsuite hang due to libphobos/testsuite/libphobos.gc/forkgc2.d

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103944

--- Comment #2 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
Even running it by hand gets stuck:
/builddir/build/BUILD/gcc-12.0.0-20220106/obj-i686-redhat-linux/./gcc/gdc
-B/builddir/build/BUILD/gcc-12.0.0-20220106/obj-i686-redhat-linux/./gcc/
-B/usr/i686-redhat-linux/bin/ -B/usr/i686-redhat-linux/lib/ -isystem
/usr/i686-redhat-linux/include -isystem /usr/i686-redhat-linux/sys-include
-fmessage-length=0 -fno-show-column -Wall -O2 -fexceptions -g
-grecord-gcc-switches -Wall -Wp,-D_GLIBCXX_ASSERTIONS -mtune=generic -msse2
-mfpmath=sse -msse2 -mstackrealign -fasynchronous-unwind-tables
-fstack-clash-protection -fcf-protection -fcf-protection -mshstk -fversion=CET
-fversion=Shared -fpreview=dip1000 -fno-release -funittest -nostdinc
-B/builddir/build/BUILD/gcc-12.0.0-20220106/obj-i686-redhat-linux/i686-redhat-linux/libphobos/src
-I/builddir/build/BUILD/gcc-12.0.0-20220106/obj-i686-redhat-linux/i686-redhat-linux/libphobos/libdruntime
-I/builddir/build/BUILD/gcc-12.0.0-20220106/libphobos/libdruntime
-I/builddir/build/BUILD/gcc-12.0.0-20220106/libphobos/src
/builddir/build/BUILD/gcc-12.0.0-20220106/libphobos/testsuite/libphobos.gc/forkgc2.d
-fstack-protector-strong -fno-diagnostics-show-caret -fdiagnostics-color=never
-B/builddir/build/BUILD/gcc-12.0.0-20220106/obj-i686-redhat-linux/i686-redhat-linux/libphobos/src
-B/builddir/build/BUILD/gcc-12.0.0-20220106/obj-i686-redhat-linux/i686-redhat-linux/libphobos/libdruntime/gcc
-B/builddir/build/BUILD/gcc-12.0.0-20220106/obj-i686-redhat-linux/i686-redhat-linux/libphobos/src/.libs
-L/builddir/build/BUILD/gcc-12.0.0-20220106/obj-i686-redhat-linux/i686-redhat-linux/libphobos/src/.libs
-lm -o ./forkgc2.exe
LD_LIBRARY_PATH=../../src/.libs/ ./forkgc2.exe

[Bug d/103944] [12 Regression] Testsuite hang due to libphobos/testsuite/libphobos.gc/forkgc2.d

[ibuclaw at gdcproject dot org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103944

Iain Buclaw <ibuclaw at gdcproject dot org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Target|                            |i686-redhat-linux

--- Comment #3 from Iain Buclaw <ibuclaw at gdcproject dot org> ---
I'm able to reproduce either, I'll see if there's a 32-bit native machine
around to test.

[Bug d/103944] [12 Regression] Testsuite hang due to libphobos/testsuite/libphobos.gc/forkgc2.d

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103944

--- Comment #4 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
Note, it isn't just i686-linux, I was getting the hangs on x86_64-linux,
s390x-linux or armv7hl-linux-gnueabi too.
Wonder whether it is
-fstack-clash-protection -fcf-protection -fcf-protection
related or perhaps glibc version dependent (in the builds where it reproduces
we are using a very recent glibc snapshot (e.g. one with _dl_find_object in
there in case the test uses unwind info in some way)).

[Bug d/103944] [12 Regression] Testsuite hang due to libphobos/testsuite/libphobos.gc/forkgc2.d

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103944

--- Comment #5 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
The process when it hangs is stuck on pthread_mutex_lock called from
_D4core4sync5mutex5Mutex__T12lock_nothrowTCQBpQBnQBlQBiZQBdMFNbNiNeZv
in libgphobos.so.3.
#0  0xf7ee6579 in __kernel_vsyscall ()
#1  0xf73af173 in __lll_lock_wait () from
/var/lib/mock/fedora-rawhide-i686/root/lib/libc.so.6
#2  0xf73b5f41 in pthread_mutex_lock@@GLIBC_2.0 () from
/var/lib/mock/fedora-rawhide-i686/root/lib/libc.so.6
#3  0xf77b41d0 in ?? ()
#4  0xf7e0cff4 in ?? ()
where 0xf77b41d0 is inside of
_D4core4sync5mutex5Mutex__T12lock_nothrowTCQBpQBnQBlQBiZQBdMFNbNiNeZv

[Bug d/103944] [12 Regression] Testsuite hang due to libphobos/testsuite/libphobos.gc/forkgc2.d

[ibuclaw at gdcproject dot org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103944

--- Comment #6 from Iain Buclaw <ibuclaw at gdcproject dot org> ---
(In reply to Jakub Jelinek from comment #4)
> Note, it isn't just i686-linux, I was getting the hangs on x86_64-linux,
> s390x-linux or armv7hl-linux-gnueabi too.
> Wonder whether it is
> -fstack-clash-protection -fcf-protection -fcf-protection
> related or perhaps glibc version dependent (in the builds where it
> reproduces we are using a very recent glibc snapshot (e.g. one with
> _dl_find_object in there in case the test uses unwind info in some way)).
CET is turned on by default now IIRC for X86/64, so I doubt it'd be
-fcf-protection.

[Bug d/103944] [12 Regression] Testsuite hang due to libphobos/testsuite/libphobos.gc/forkgc2.d

[ibuclaw at gdcproject dot org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103944

--- Comment #7 from Iain Buclaw <ibuclaw at gdcproject dot org> ---
(In reply to Iain Buclaw from comment #6)
> (In reply to Jakub Jelinek from comment #4)
> > Note, it isn't just i686-linux, I was getting the hangs on x86_64-linux,
> > s390x-linux or armv7hl-linux-gnueabi too.
> > Wonder whether it is
> > -fstack-clash-protection -fcf-protection -fcf-protection
> > related or perhaps glibc version dependent (in the builds where it
> > reproduces we are using a very recent glibc snapshot (e.g. one with
> > _dl_find_object in there in case the test uses unwind info in some way)).
> CET is turned on by default now IIRC for X86/64, so I doubt it'd be
> -fcf-protection.
Don't get a hung process on latest rawhide either, which uses this version of
glibc.

https://koji.fedoraproject.org/koji/buildinfo?buildID=1872662

I'll re-run a few times, increasing/decreasing the available memory to see if
anything triggers.

[Bug d/103944] [12/13 Regression] Testsuite hang due to libphobos/testsuite/libphobos.gc/forkgc2.d

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103944

Jakub Jelinek <jakub at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|12.0                        |12.2

--- Comment #8 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
GCC 12.1 is being released, retargeting bugs to GCC 12.2.

[Bug d/103944] [12/13 Regression] Testsuite hang due to libphobos/testsuite/libphobos.gc/forkgc2.d

[doko at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103944

Matthias Klose <doko at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |doko at gcc dot gnu.org

--- Comment #9 from Matthias Klose <doko at gcc dot gnu.org> ---
also seen in both the current Debian and Ubuntu development releases in the
gcc-12 branch and the trunk 20220717.

 - glibc 2.33 (Debian), 2.35 (Ubuntu)
 - binutils 2.39 branch

configured with --enable-default-pie --enable-cet

seen on amd64, aarch64, s390x

[Bug d/103944] [12/13 Regression] Testsuite hang due to libphobos/testsuite/libphobos.gc/forkgc2.d

[iains at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103944

Iain Sandoe <iains at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Last reconfirmed|                            |2023-02-26
     Ever confirmed|0                           |1
             Status|UNCONFIRMED                 |NEW
                 CC|                            |iains at gcc dot gnu.org
             Target|i686-redhat-linux           |i686-redhat-linux,
                   |                            |*-*-darwin*
      Known to fail|                            |12.2.1, 13.0

--- Comment #11 from Iain Sandoe <iains at gcc dot gnu.org> ---
On darwin, I see that the main forkgc2.exe gets killed when it times out but
the child processes are still present - which leads to the testsuite hanging. 
At present, skipping this test seems the best strategy for Darwin, pending a
resolution.

[Bug d/103944] [12/13/14 Regression] Testsuite hang due to libphobos/testsuite/libphobos.gc/forkgc2.d

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103944

Richard Biener <rguenth at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|12.3                        |12.4

--- Comment #12 from Richard Biener <rguenth at gcc dot gnu.org> ---
GCC 12.3 is being released, retargeting bugs to GCC 12.4.

[Bug d/103944] [12/13/14 Regression] Testsuite hang due to libphobos/testsuite/libphobos.gc/forkgc2.d

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103944

--- Comment #13 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The master branch has been updated by Iain D Sandoe <iains@gcc.gnu.org>:

https://gcc.gnu.org/g:fca6d9c12f5bf06469cf9f7db8c42f66ef792fd2

commit r14-2233-gfca6d9c12f5bf06469cf9f7db8c42f66ef792fd2
Author: Iain Sandoe <iainsandoe@Apollo-3-VM-Lion-0-187.local>
Date:   Sun Feb 26 13:53:52 2023 +0000

    libphobos, testsuite: Disable forkgc2 on Darwin [PR103944]

    It hangs the testsuite (requiring manual intervention to kill the
    spawned processes) which breaks CI.  The reason for the hang id not
    clear.  This skips the test for now (xfail does not work).

    Signed-off-by: Iain Sandoe <iain@sandoe.co.uk>

            PR d/103944

    libphobos/ChangeLog:

            * testsuite/libphobos.gc/forkgc2.d: Skip for Darwin.

[Bug d/103944] [12/13/14 Regression] Testsuite hang due to libphobos/testsuite/libphobos.gc/forkgc2.d

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103944

--- Comment #14 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The releases/gcc-13 branch has been updated by Iain D Sandoe
<iains@gcc.gnu.org>:

https://gcc.gnu.org/g:025a3f417899577710eb88527897fc571a0280ff

commit r13-7526-g025a3f417899577710eb88527897fc571a0280ff
Author: Iain Sandoe <iainsandoe@Apollo-3-VM-Lion-0-187.local>
Date:   Sun Feb 26 13:53:52 2023 +0000

    libphobos, testsuite: Disable forkgc2 on Darwin [PR103944]

    It hangs the testsuite (requiring manual intervention to kill the
    spawned processes) which breaks CI.  The reason for the hang id not
    clear.  This skips the test for now (xfail does not work).

    Signed-off-by: Iain Sandoe <iain@sandoe.co.uk>

            PR d/103944

    libphobos/ChangeLog:

            * testsuite/libphobos.gc/forkgc2.d: Skip for Darwin.

    (cherry picked from commit fca6d9c12f5bf06469cf9f7db8c42f66ef792fd2)

[Bug d/103944] [12/13/14 Regression] Testsuite hang due to libphobos/testsuite/libphobos.gc/forkgc2.d

[law at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103944

Jeffrey A. Law <law at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |law at gcc dot gnu.org
           Priority|P3                          |P4

[Bug d/103944] [12/13/14 Regression] Testsuite hang due to libphobos/testsuite/libphobos.gc/forkgc2.d

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103944

--- Comment #15 from GCC Commits <cvs-commit at gcc dot gnu.org> ---
The releases/gcc-12 branch has been updated by Iain D Sandoe
<iains@gcc.gnu.org>:

https://gcc.gnu.org/g:5a72912f9b0a5aa3c5a726ec499137c189921f9b

commit r12-10309-g5a72912f9b0a5aa3c5a726ec499137c189921f9b
Author: Iain Sandoe <iainsandoe@Apollo-3-VM-Lion-0-187.local>
Date:   Sun Feb 26 13:53:52 2023 +0000

    libphobos, testsuite: Disable forkgc2 on Darwin [PR103944]

    It hangs the testsuite (requiring manual intervention to kill the
    spawned processes) which breaks CI.  The reason for the hang id not
    clear.  This skips the test for now (xfail does not work).

    Signed-off-by: Iain Sandoe <iain@sandoe.co.uk>

            PR d/103944

    libphobos/ChangeLog:

            * testsuite/libphobos.gc/forkgc2.d: Skip for Darwin.

    (cherry picked from commit fca6d9c12f5bf06469cf9f7db8c42f66ef792fd2)