This is the mail archive of the
systemtap@sourceware.org
mailing list for the systemtap project.
[Bug bpf/24812] New: stapbpf: support order-parametrized begin/end probes
- From: "me at serhei dot io" <sourceware-bugzilla at sourceware dot org>
- To: systemtap at sourceware dot org
- Date: Mon, 15 Jul 2019 18:44:03 +0000
- Subject: [Bug bpf/24812] New: stapbpf: support order-parametrized begin/end probes
- Auto-submitted: auto-generated
https://sourceware.org/bugzilla/show_bug.cgi?id=24812
Bug ID: 24812
Summary: stapbpf: support order-parametrized begin/end probes
Product: systemtap
Version: unspecified
Status: NEW
Severity: normal
Priority: P2
Component: bpf
Assignee: systemtap at sourceware dot org
Reporter: me at serhei dot io
Target Milestone: ---
found this while trying to work around probe size requirement
-- so this will be important for complex bpf programs until/unless we bump the
userspace interpreter's stack size
I want two end probes to run in a specific order. Normally begin/end probe
order is unspecified but in fact you can constrain it with sequence numbers.
global flag1 = 0
global flag2 = 0
probe begin(1) {
printf("BEGIN\n")
flag1 = 2
}
probe begin(2) {
flag1 = 1
exit()
}
probe end(1) {
flag2 = 2
}
probe end(2) {
flag2 = 1
}
probe end(3) {
printf("got flag1==%d flag2==%d\n", flag1, flag2)
if (flag1 == 1 && flag2 == 1)
printf("END PASS\n")
else
printf("END FAIL\n")
}
but it doesn't work on bpf.
$ sudo ~/stap-install/bin/stap probe_order.stp
BEGIN
got flag1==1 flag2==1
END PASS
$ sudo ~/stap-install/bin/stap --bpf probe_order.stp
BEGIN
got flag1==1 flag2==0
END FAIL
Mirrors pr3624 for the kernel backend.
--
You are receiving this mail because:
You are the assignee for the bug.