Fix 'tools/statsnoop' from failing to attach kprobes

This fixes 'tools/statsnoop' from failing to attach probes when the expected entry point for a system call cannot be found. This script uses the 'stat', 'statfs' and 'newstat' system calls, all of which must be implemented to be POSIX compliant. However, the names of the actual entry points for their respective implementations in the kernel might vary across architectures. For example, a powerpc64 kernel does not define 'sys_stat' but still provides the 'stat' system call via 'sys_newstat'. This causes the script to fail if it tries to attach a probe at 'sys_stat'. We avoid this by performing some extra checks to see if these entry points exist. Signed-off-by: Sandipan Das <sandipan@linux.vnet.ibm.com>
sjtuhjh · Oct 5, 2017 · 16523a3 · 16523a3
1 parent 782b34f
commit 16523a3
Showing 1 changed file with 16 additions and 6 deletions.
diff --git a/tools/statsnoop.py b/tools/statsnoop.py
@@ -105,12 +105,22 @@
 
 # initialize BPF
 b = BPF(text=bpf_text)
-b.attach_kprobe(event="sys_stat", fn_name="trace_entry")
-b.attach_kprobe(event="sys_statfs", fn_name="trace_entry")
-b.attach_kprobe(event="sys_newstat", fn_name="trace_entry")
-b.attach_kretprobe(event="sys_stat", fn_name="trace_return")
-b.attach_kretprobe(event="sys_statfs", fn_name="trace_return")
-b.attach_kretprobe(event="sys_newstat", fn_name="trace_return")
+
+# for POSIX compliance, all architectures implement these
+# system calls but the name of the actual entry point may
+# be different for which we must check if the entry points
+# actually exist before attaching the probes
+if BPF.ksymname("sys_stat") != -1:
+    b.attach_kprobe(event="sys_stat", fn_name="trace_entry")
+    b.attach_kretprobe(event="sys_stat", fn_name="trace_return")
+
+if BPF.ksymname("sys_statfs") != -1:
+    b.attach_kprobe(event="sys_statfs", fn_name="trace_entry")
+    b.attach_kretprobe(event="sys_statfs", fn_name="trace_return")
+
+if BPF.ksymname("sys_newstat") != -1:
+    b.attach_kprobe(event="sys_newstat", fn_name="trace_entry")
+    b.attach_kretprobe(event="sys_newstat", fn_name="trace_return")
 
 TASK_COMM_LEN = 16    # linux/sched.h
 NAME_MAX = 255        # linux/limits.h