This is the mail archive of the systemtap@sourceware.org mailing list for the systemtap project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

[Bug tapsets/21353] New: syscall nesting in fcntl


https://sourceware.org/bugzilla/show_bug.cgi?id=21353

            Bug ID: 21353
           Summary: syscall nesting in fcntl
           Product: systemtap
           Version: unspecified
            Status: NEW
          Severity: normal
          Priority: P2
         Component: tapsets
          Assignee: systemtap at sourceware dot org
          Reporter: dsmith at redhat dot com
  Target Milestone: ---

In [nd_]syscall.fcntl, there is some syscall nesting going on. The example
program from the syscall testsuite demonstrates the problem:

====
# gcc -m32 -o net1 net1.c
stap -e 'probe syscall.fcntl { printf("%s\n", ppfunc()) } probe
syscall.fcntl.return { printf("return\n") }' -c ./net1
SyS_fcntl
return
SyS_fcntl
return
SyS_fcntl
return
SyS_fcntl
return
compat_SyS_fcntl64
SyS_fcntl
return
return
compat_SyS_fcntl64
SyS_fcntl
return
return
compat_SyS_fcntl64
SyS_fcntl
return
return
compat_SyS_fcntl64
SyS_fcntl
return
return
compat_SyS_fcntl64
SyS_fcntl
return
return
====

So, both the compat call and the "real" call are getting hits. So, we're
getting 2 sets of probe hits for 1 fcntl() call.

I saw this with kernel-4.11.0-0.rc4.git2.1.fc27.x86_64 and HEAD systemtap.

This was originally reported by 'georgj' on #systemtap

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

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]