https://bugs.kde.org/show_bug.cgi?id=427433
--- Comment #11 from Paul Floyd <pa...@free.fr> --- Before my relating my debug efforts, I did notice that there was one change in git - 22aa8640e6c44c78c228ffa726cfacf918455343 - that might affect this problem, adding PIDFDs, but that was back in Feb 2020, so it should be in 3.16.1. So, debugging the problem. I just tried with Valgrind built from git HEAD and there was no problem. paulf@localhost:~/Downloads> ../valgrind/vg-in-place -q ./test Qt Version: 5.15.2 ==16771== Syscall param waitid(infop) points to unaddressable byte(s) ==16771== at 0x5DE334D: syscall (in /lib64/libc-2.32.so) ==16771== by 0x53B1996: sys_waitid (forkfd_linux.c:65) ==16771== by 0x53B1996: detect_clone_pidfd_support (forkfd_linux.c:125) ==16771== by 0x53B1996: system_forkfd (forkfd_linux.c:141) ==16771== by 0x53B1996: forkfd (forkfd.c:651) ==16771== by 0x53954EC: QProcessPrivate::startProcess() (qprocess_unix.cpp:465) ==16771== by 0x538FABD: QProcessPrivate::start(QFlags<QIODevice::OpenModeFlag>) (qprocess.cpp:2247) ==16771== by 0x538FCF2: QProcess::start(QString const&, QStringList const&, QFlags<QIODevice::OpenModeFlag>) (qprocess.cpp:2095) ==16771== by 0x4012A4: main (in /home/paulf/Downloads/test) ==16771== Address 0x0 is not stack'd, malloc'd or (recently) free'd ==16771== "Linux localhost.localdomain 5.9.8-2-default #1 SMP Thu Nov 12 07:43:32 UTC 2020 (ea93937) x86_64 x86_64 x86_64 GNU/Linux\n" Then I checked out the 3.16.1 code, same thing. Next I tried debugging the package installed Valgrind. No joy, it's been stripped and has no debug info. paulf@localhost:~/Downloads> ll /usr/lib64/valgrind/memcheck-amd64-linux ../valgrind/memcheck/memcheck-amd64-linux -rwxr-xr-x 1 root root 2743648 Sep 30 20:10 /usr/lib64/valgrind/memcheck-amd64-linux -rwxr-xr-x 1 paulf users 16555896 Nov 22 16:42 ../valgrind/memcheck/memcheck-amd64-linux paulf@localhost:~/Downloads> file /usr/lib64/valgrind/memcheck-amd64-linux ../valgrind/memcheck/memcheck-amd64-linux /usr/lib64/valgrind/memcheck-amd64-linux: ELF 64-bit LSB executable, x86-64, version 1 (SYSV), statically linked, BuildID[sha1]=fb2fe27acb715941050c3806c632d51fe2c03455, stripped ../valgrind/memcheck/memcheck-amd64-linux: ELF 64-bit LSB executable, x86-64, version 1 (SYSV), statically linked, BuildID[sha1]=8192d34835865874742478127c0276f5fc75b277, with debug_info, not stripped I had a look here https://build.opensuse.org/package/view_file/devel:tools/valgrind/valgrind.spec?expand=1 I don't know what their optflags are. I tried adding the LTO patch and configuring with --enable-lto=yes, and still could not reproduce the issue. That's about all that I can try for the moment. -- You are receiving this mail because: You are watching all bug changes.
