Hi Bernhard, Looks fine to me, although I'd prefer to continue to print the "-1" in xen_hyper_dump_xen_hyper_offset_table(), the same way that "help -o" does for all the Linux offsets. CC'ing the xencrash maintainer Itsuro Oda to get his ACK. Thanks, Dave ----- "Bernhard Walle" <bwalle(a)suse.de&gt; wrote: > According to that change set in the Xen unstable branch > > changeset: 18428:ae9b223a675d > user: Keir Fraser <keir.fraser(a)citrix.com&gt; > date: Thu Sep 04 14:38:26 2008 +0100 > files: xen/common/domain.c > xen/common/event_channel.c > xen/common/schedule.c > xen/include/xen/sched.h > description: > More efficient implementation of SCHEDOP_poll when polling a > single port. > Signed-off-by: Keir Fraser <keir.fraser(a)citrix.com&gt; > > the member "is_polling" of struct domain will be removed in Xen 4.4. > This patch fixes the problem that a dump cannot be opened in > Hypervisor mode in > crash when produced by such a Xen version. That's the error message of > Crash: > > crash: invalid structure member offset: domain_is_polling > FILE: xen_hyper.c LINE: 1182 FUNCTION: > xen_hyper_store_domain_context() > > [./crash] error trace: 52bfc2 => 52e401 => 52ee07 => 508062 > > 508062: OFFSET_verify+168 > 52ee07: xen_hyper_store_domain_context+742 > 52e401: xen_hyper_refresh_domain_context_space+196 > 52bfc2: xen_hyper_domain_init+959 > > > Signed-off-by: Bernhard Walle <bwalle(a)suse.de&gt; > > --- > xen_hyper.c | 9 ++++++++- > xen_hyper_dump_tables.c | 6 ++++-- > 2 files changed, 12 insertions(+), 3 deletions(-) > > --- a/xen_hyper.c > +++ b/xen_hyper.c > @@ -188,7 +188,13 @@ xen_hyper_domain_init(void) > XEN_HYPER_MEMBER_OFFSET_INIT(domain_is_hvm, "domain", "is_hvm"); > XEN_HYPER_MEMBER_OFFSET_INIT(domain_is_privileged, "domain", > "is_privileged"); > XEN_HYPER_MEMBER_OFFSET_INIT(domain_debugger_attached, "domain", > "debugger_attached"); > + > + /* > + * Will be removed in Xen 4.4 (hg ae9b223a675d), > + * need to check that with XEN_HYPER_VALID_MEMBER() before using > + */ > XEN_HYPER_MEMBER_OFFSET_INIT(domain_is_polling, "domain", > "is_polling"); > + > XEN_HYPER_MEMBER_OFFSET_INIT(domain_is_dying, "domain", > "is_dying"); > XEN_HYPER_MEMBER_OFFSET_INIT(domain_is_paused_by_controller, > "domain", "is_paused_by_controller"); > XEN_HYPER_MEMBER_OFFSET_INIT(domain_is_shutting_down, "domain", > "is_shutting_down"); > @@ -1179,7 +1185,8 @@ xen_hyper_store_domain_context(struct xe > dc->domain_flags |= XEN_HYPER_DOMS_privileged; > } else if (*(dp + XEN_HYPER_OFFSET(domain_debugger_attached))) { > dc->domain_flags |= XEN_HYPER_DOMS_debugging; > - } else if (*(dp + XEN_HYPER_OFFSET(domain_is_polling))) { > + } else if (XEN_HYPER_VALID_MEMBER(domain_is_polling) && > + *(dp + XEN_HYPER_OFFSET(domain_is_polling))) { > dc->domain_flags |= XEN_HYPER_DOMS_polling; > } else if (*(dp + > XEN_HYPER_OFFSET(domain_is_paused_by_controller))) { > dc->domain_flags |= XEN_HYPER_DOMS_ctrl_pause; > --- a/xen_hyper_dump_tables.c > +++ b/xen_hyper_dump_tables.c > @@ -775,8 +775,10 @@ xen_hyper_dump_xen_hyper_offset_table(ch > (buf, "%ld\n", xen_hyper_offset_table.domain_is_privileged)); > XEN_HYPER_PRI(fp, len, "domain_debugger_attached: ", buf, flag, > (buf, "%ld\n", xen_hyper_offset_table.domain_debugger_attached)); > - XEN_HYPER_PRI(fp, len, "domain_is_polling: ", buf, flag, > - (buf, "%ld\n", xen_hyper_offset_table.domain_is_polling)); > + if (XEN_HYPER_VALID_MEMBER(domain_is_polling)) { > + XEN_HYPER_PRI(fp, len, "domain_is_polling: ", buf, flag, > + (buf, "%ld\n", xen_hyper_offset_table.domain_is_polling)); > + } > XEN_HYPER_PRI(fp, len, "domain_is_dying: ", buf, flag, > (buf, "%ld\n", xen_hyper_offset_table.domain_is_dying)); > XEN_HYPER_PRI(fp, len, "domain_is_paused_by_controller: ", buf, > flag, > > -- > Crash-utility mailing list > Crash-utility(a)redhat.com > https://www.redhat.com/mailman/listinfo/crash-utility