[Bug runtime/30534] program received "Illegal instruction" when using systemtap set a probe and hit the probe

wcohen at redhat dot com sourceware-bugzilla@sourceware.org
Thu Oct 12 20:43:57 GMT 2023


https://sourceware.org/bugzilla/show_bug.cgi?id=30534

William Cohen <wcohen at redhat dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         Resolution|---                         |WORKSFORME
                 CC|                            |wcohen at redhat dot com
             Status|UNCONFIRMED                 |RESOLVED

--- Comment #1 from William Cohen <wcohen at redhat dot com> ---
As the upstream systemtap doesn't currently build on RHEL7 because of missing
debuginfod support in elfutils resorted to trying things with the existing
systemtap available in RHEL7.9.  Things appeared to work correctly:

[root@hpe-dl160gen8-01 ~]# rpm -q kernel glibc systemtap
kernel-3.10.0-1160.el7.x86_64
glibc-2.17-317.el7.x86_64
glibc-2.17-317.el7.i686
systemtap-4.0-13.el7.x86_64
[root@hpe-dl160gen8-01 ~]# stap dbg-core.stp -c /usr/bin/ls
anaconda-ks.cfg  NETBOOT_METHOD.TXT  rpmbuild
dbg-core.stp     original-ks.cfg     systemtap-5.0-1.el8.src.rpm
dbg-core.stp~    RECIPE.TXT          systemtap_write
call _IO_new_fclose()call _IO_new_fclose()[root@hpe-dl160gen8-01 ~]# cat
dbg-core.stp
probe process("/lib64/libc.so.6").function("_IO_new_fclose").return
{
  printf ("call _IO_new_fclose()")
}


The only difference I can see is that this is the kernel is a bit newer. 
Doesn't look like we are able to replicate this.

-- 
You are receiving this mail because:
You are the assignee for the bug.


More information about the Systemtap mailing list