ptrace testsuite

Please post to utrace-devel@redhat.com with any questions, suggestions, or contributions to this test suite. (Or edit this page!)

Most of the tests target ptrace(2) implementation. Most of the ptrace(2) tests have ever failed only on the ptrace-by-utrace implementation.

Get the sources by:

cvs -d :pserver:anoncvs:anoncvs@sourceware.org:/cvs/systemtap co ptrace-tests

Browse CVS sources

Simple safe run:

Runs only the tests which crashed older kernels (crasher flag below):

Runs all the tests:

Environment variable TESTTIME=<seconds> sets the testing time for the unreliable tests.

Testcases

Information on individual testcases and their status can be found in STATUS.tests file.

The table below was used for this purpose, but maintaining it outside of CVS which contains tests themselves proved to be difficult. This data is now outdated:

testcase

flags

Fedora 8 (2.6.26.6-49.fc8)

Fedora 9 (2.6.26.6-79.fc9)

Fedora 10

others

attach-into-signal

?

/!\ rh470249

attach-sigcont-wait

unreliable

{OK} rh248532

{OK}

attach-wait-on-stopped

{OK}

{OK}

{OK} rh232837 {OK} frysk4217

block-step

{OK}

{OK}

/!\ rh437028 KVM host

clone-get-signal

crasher, unreliable

{OK} rh312951

{OK}

fixed upstream (1 2 3)

clone-multi-ptrace

unreliable

{OK}

{OK}

RHEL 5.2 ia64 2.6.18-118.el5

clone-ptrace

{OK}

{OK} rh461552

detach-can-signal

{OK}

{OK}

{OK} frysk3381 frysk3595

detach-stopped

/!\ rh456335

/!\ rh454404

broken upstream in v2.6.26-5982-g0988c37 (and presumably earlier)

erestart-debugger
erestart-debuggee

{OK}

{OK}

{OK} rh436183 Rawhide

erestartsys

{OK} rh434995

{OK}

fixed upstream

event-exit-proc-environ
event-exit-proc-maps

{OK} rh285761

{OK}

fixed upstream

late-ptrace-may-attach-check

crasher, unreliable

{OK}

{OK} rh449294

o_tracevfork

{OK} rh464520

{OK}

o_tracevforkdone

/!\ rh468395

/!\ rh468089

/!\

x86_64 specific

peekpokeusr

{OK}

{OK}

powerpc-altivec

{OK} rh387821

?

ppc specific

ppc-dabr-race

/!\ rh439619

?

ppc specific fixed upstream

ppc-ptrace-exec-full-regs

crasher

{OK} rh301791

?

ppc specific

ptrace-cont-sigstop-detach

{OK}

{OK}

{OK} rh244162

ptrace_event_clone

?

/!\ rh469693

ptrace-on-job-control-stopped

{OK} rh235521

{OK}

{OK} frysk4330

reparent-zombie

{OK}

{OK}

fixed upstream after 2.6.26

reparent-zombie-clone

crasher

{OK} rh450454

{OK} rh443552

sa-resethand-on-cont-signal

{OK}

{OK}

{OK} rh284311

signal-loss

?

/!\ rh469684

step-into-handler

{OK}

{OK}

step-jump-cont

{OK}

{OK}

step-jump-cont-strict

{OK} rh450439

{OK} rh450379

pending fix

step-through-sigret

{OK} rh441406

{OK} rh450378

/!\ frysk6044

step-to-breakpoint

{OK}

{OK}

/!\ rh439668

stop-attach-then-wait

{OK} rh454157

{OK} rh454156

syscall-reset

{OK}

{OK}

{OK} rh438614 Rawhide, fixed upstream

tif-syscall-trace-after-detach

unreliable

{OK}

{OK}

{OK} RHEL-4.6 (rh276861 rh276091)

tracer-lockup-on-sighandler-kill

crasher, unreliable

{OK}

{OK}

{OK} rh267161

user-area-access

{OK}

?

s390 specific

user-area-padding

crasher

{OK}

?

s390 specific

user-regs-peekpoke

{OK}

{OK}

watchpoint

{OK} rh387741

{OK}

x86_64-cs

crasher

{OK} rh431314

{OK}

fixed upstream in 2.6.25-rc1

x86_64-gsbase

{OK}

{OK}

fixed upstream in 2.6.25-rc1

x86_64-ia32-gs

{OK}

{OK}

{OK} rh247561

{OK} The bug was never present in this distribution or it is already fixed in the latest distribution release.

/!\ The bug still needs to be fixed for the latest distribution release.

crasher = There exist (possibly older) kernel releases which could crash on this test.

unreliable = Race conditions may need a long time testing runs for a reasonable reproducibility. Please set TESTTIME to values like 3600 (seconds), still in some cases the problem reproducibility may depend also on the kernel arch and other kernel configuration options.