+Luwei, who developed PT for KVM and is the best one who can help
review VMX changes from Intel side. Please include him in future
post or discussion.

> -----Original Message-----
> From: Michał Leszczyński <michal.leszczynski@cert.pl>
> Sent: Wednesday, June 17, 2020 2:48 AM
> To: Andrew Cooper <andrew.cooper3@citrix.com>
> Cc: Xen-devel <xen-devel@lists.xenproject.org>; Jan Beulich
> <jbeulich@suse.com>; Wei Liu <wl@xen.org>; Roger Pau Monné
> <roger.pau@citrix.com>; Nakajima, Jun <jun.nakajima@intel.com>; Tian,
> Kevin <kevin.tian@intel.com>; George Dunlap <george.dunlap@citrix.com>;
> Ian Jackson <ian.jackson@eu.citrix.com>; Julien Grall <julien@xen.org>;
> Stefano Stabellini <sstabellini@kernel.org>
> Subject: Re: [PATCH v1 0/7] Implement support for external IPT monitoring
> 
> ----- 16 cze 2020 o 20:17, Andrew Cooper andrew.cooper3@citrix.com
> napisał(a):
> 
> > On 16/06/2020 16:16, Michał Leszczyński wrote:
> >> Intel Processor Trace is an architectural extension available in modern
> Intel
> >> family CPUs. It allows recording the detailed trace of activity while the
> >> processor executes the code. One might use the recorded trace to
> reconstruct
> >> the code flow. It means, to find out the executed code paths, determine
> >> branches taken, and so forth.
> >>
> >> The abovementioned feature is described in Intel(R) 64 and IA-32
> Architectures
> >> Software Developer's Manual Volume 3C: System Programming Guide,
> Part 3,
> >> Chapter 36: "Intel Processor Trace."
> >>
> >> This patch series implements an interface that Dom0 could use in order to
> enable
> >> IPT for particular vCPUs in DomU, allowing for external monitoring. Such a
> >> feature has numerous applications like malware monitoring, fuzzing, or
> >> performance testing.
> >
> > Hello,
> >
> > I'm very excited to see support like this appearing.  However, be aware
> > that we're currently in code freeze for the 4.14 release, so in-depth
> > reviews will probably be delayed somewhat due to our bug queue and
> > release activities.
> 
> Sure, take your time :)
> 
> 
> >
> > That said, I've had a very quick look through the series, and have a few
> > general questions first.
> >
> > AFAICT, this is strictly for external monitoring of the VM, not for the
> > VM to use itself?  If so, it shouldn't have the H tag here:
> >
> > XEN_CPUFEATURE(IPT,           5*32+25) /*H  Intel Processor Trace */
> >
> > because that exposes the feature to the guest, with the implication that
> > all other parts of the feature work as advertised.
> 
> Ok, I will remove the H tag.
> 
> 
> >
> >
> > Are there any restrictions on EPT being enabled in the first place?  I'm
> > not aware of any, and in principle we could use this functionality for
> > PV guests as well (using the CPL filter).  Therefore, I think it would
> > be helpful to not tie the functionality to HVM guests, even if that is
> > the only option enabled to start with.
> 
> I think at the moment it's not required to have EPT. This patch series doesn't
> use any translation feature flags, so the output address is always a machine
> physical address, regardless of context. I will check if it could be easily used
> with PV.
> 
> 
> >
> > The buffer mapping and creation logic is fairly problematic.  Instead of
> > fighting with another opencoded example, take a look at the IOREQ
> > server's use of "acquire resource" which is a mapping interface which
> > supports allocating memory on behalf of the guest, outside of the guest
> > memory, for use by control tools.
> >
> > I think what this wants is a bit somewhere in domain_create to indicate
> > that external tracing is used for this domain (and allocate whatever
> > structures/buffers are necessary), acquire resource to map the buffers
> > themselves, and a domctl for any necessary runtime controls.
> >
> 
> I will check this out, this sounds like a good option as it would remove lots of
> complexity from the existing ipt_enable domctl.
> 
> >
> > What semantics do you want for the buffer becoming full?  Given that
> > debugging/tracing is the goal, I presume "pause vcpu on full" is the
> > preferred behaviour, rather than drop packets on full?
> >
> 
> Right now this is a ring-style buffer and when it would become full it would
> simply wrap and override the old data.
> 
> >
> > When this subject was broached on xen-devel before, one issue was the
> > fact that all actions which are intercepted don't end up writing any
> > appropriate packets.  This is perhaps less of an issue for this example,
> > where the external agent can see VMExits in the trace, but it still
> > results in missing information.  (It is a major problem for PT within
> > the guest, and needs Xen's intercept/emulation framework being updated
> > to be PT-aware so it can fill in the same packets which hardware would
> > have done for equivalent actions.)
> 
> Ok, this sounds like a hard issue. Could you point out what could be the
> particular problematic cases? For instance, if something would alter EIP/RIP
> or CR3 then I belive it would still be recorded in PT trace (i.e. these values will
> be logged on VM entry).
> 
> >
> >
> > Thanks,
> >
> > ~Andrew
> 
> 
> Best regards,
> Michał Leszczyński
> CERT Polska

> -----Original Message-----
> From: Tian, Kevin <kevin.tian@intel.com>
> Sent: Wednesday, June 17, 2020 9:35 AM
> To: Michał Leszczyński <michal.leszczynski@cert.pl>; Andrew Cooper
> <andrew.cooper3@citrix.com>
> Cc: Xen-devel <xen-devel@lists.xenproject.org>; Jan Beulich
> <jbeulich@suse.com>; Wei Liu <wl@xen.org>; Roger Pau Monné
> <roger.pau@citrix.com>; Nakajima, Jun <jun.nakajima@intel.com>; George
> Dunlap <george.dunlap@citrix.com>; Ian Jackson <ian.jackson@eu.citrix.com>;
> Julien Grall <julien@xen.org>; Stefano Stabellini <sstabellini@kernel.org>;
> Kang, Luwei <luwei.kang@intel.com>
> Subject: RE: [PATCH v1 0/7] Implement support for external IPT monitoring
> 
> +Luwei, who developed PT for KVM and is the best one who can help
> review VMX changes from Intel side. Please include him in future post or
> discussion.
> 
> > -----Original Message-----
> > From: Michał Leszczyński <michal.leszczynski@cert.pl>
> > Sent: Wednesday, June 17, 2020 2:48 AM
> > To: Andrew Cooper <andrew.cooper3@citrix.com>
> > Cc: Xen-devel <xen-devel@lists.xenproject.org>; Jan Beulich
> > <jbeulich@suse.com>; Wei Liu <wl@xen.org>; Roger Pau Monné
> > <roger.pau@citrix.com>; Nakajima, Jun <jun.nakajima@intel.com>; Tian,
> > Kevin <kevin.tian@intel.com>; George Dunlap
> > <george.dunlap@citrix.com>; Ian Jackson <ian.jackson@eu.citrix.com>;
> > Julien Grall <julien@xen.org>; Stefano Stabellini
> > <sstabellini@kernel.org>
> > Subject: Re: [PATCH v1 0/7] Implement support for external IPT
> > monitoring
> >
> > ----- 16 cze 2020 o 20:17, Andrew Cooper andrew.cooper3@citrix.com
> > napisał(a):
> >
> > > On 16/06/2020 16:16, Michał Leszczyński wrote:
> > >> Intel Processor Trace is an architectural extension available in
> > >> modern
> > Intel
> > >> family CPUs. It allows recording the detailed trace of activity
> > >> while the processor executes the code. One might use the recorded
> > >> trace to
> > reconstruct
> > >> the code flow. It means, to find out the executed code paths,
> > >> determine branches taken, and so forth.
> > >>
> > >> The abovementioned feature is described in Intel(R) 64 and IA-32
> > Architectures
> > >> Software Developer's Manual Volume 3C: System Programming Guide,
> > Part 3,
> > >> Chapter 36: "Intel Processor Trace."
> > >>
> > >> This patch series implements an interface that Dom0 could use in
> > >> order to
> > enable
> > >> IPT for particular vCPUs in DomU, allowing for external monitoring.
> > >> Such a feature has numerous applications like malware monitoring,
> > >> fuzzing, or performance testing.
> > >
> > > Hello,
> > >
> > > I'm very excited to see support like this appearing.  However, be
> > > aware that we're currently in code freeze for the 4.14 release, so
> > > in-depth reviews will probably be delayed somewhat due to our bug
> > > queue and release activities.
> >
> > Sure, take your time :)
> >
> >
> > >
> > > That said, I've had a very quick look through the series, and have a
> > > few general questions first.
> > >
> > > AFAICT, this is strictly for external monitoring of the VM, not for
> > > the VM to use itself?  If so, it shouldn't have the H tag here:
> > >
> > > XEN_CPUFEATURE(IPT,           5*32+25) /*H  Intel Processor Trace */
> > >
> > > because that exposes the feature to the guest, with the implication
> > > that all other parts of the feature work as advertised.
> >
> > Ok, I will remove the H tag.
> >
> >
> > >
> > >
> > > Are there any restrictions on EPT being enabled in the first place?
> > > I'm not aware of any, and in principle we could use this
> > > functionality for PV guests as well (using the CPL filter).
> > > Therefore, I think it would be helpful to not tie the functionality
> > > to HVM guests, even if that is the only option enabled to start with.
> >
> > I think at the moment it's not required to have EPT. This patch series
> > doesn't use any translation feature flags, so the output address is
> > always a machine physical address, regardless of context. I will check
> > if it could be easily used with PV.
> >
> >
> > >
> > > The buffer mapping and creation logic is fairly problematic.
> > > Instead of fighting with another opencoded example, take a look at
> > > the IOREQ server's use of "acquire resource" which is a mapping
> > > interface which supports allocating memory on behalf of the guest,
> > > outside of the guest memory, for use by control tools.
> > >
> > > I think what this wants is a bit somewhere in domain_create to
> > > indicate that external tracing is used for this domain (and allocate
> > > whatever structures/buffers are necessary), acquire resource to map
> > > the buffers themselves, and a domctl for any necessary runtime controls.
> > >
> >
> > I will check this out, this sounds like a good option as it would
> > remove lots of complexity from the existing ipt_enable domctl.
> >
> > >
> > > What semantics do you want for the buffer becoming full?  Given that
> > > debugging/tracing is the goal, I presume "pause vcpu on full" is the
> > > preferred behaviour, rather than drop packets on full?
> > >
> >
> > Right now this is a ring-style buffer and when it would become full it
> > would simply wrap and override the old data.
> >
> > >
> > > When this subject was broached on xen-devel before, one issue was
> > > the fact that all actions which are intercepted don't end up writing
> > > any appropriate packets.  This is perhaps less of an issue for this
> > > example, where the external agent can see VMExits in the trace, but
> > > it still results in missing information.  (It is a major problem for
> > > PT within the guest, and needs Xen's intercept/emulation framework
> > > being updated to be PT-aware so it can fill in the same packets
> > > which hardware would have done for equivalent actions.)
> >
> > Ok, this sounds like a hard issue. Could you point out what could be
> > the particular problematic cases? For instance, if something would
> > alter EIP/RIP or CR3 then I belive it would still be recorded in PT
> > trace (i.e. these values will be logged on VM entry).

e.g. If a VM exit is taken on a guest write to CR3 (including “MOV CR3” as well as task switches), the PIP packet
normally generated on the CR3 write will be missing. The PIP packet needs to be written to the PT buffer by software. Another example is VM-exit taken on RDTSC. 

For VM introspection, all the Intel PT packets may need to emulated by software. Some description in SDM as below:
If a VMM emulates an element of processor state by taking a VM exit on reads and/or writes to that piece of state, and the state element impacts Intel PT packet generation or values, it may be incumbent upon the VMM to insert or modify the output trace data.

Thanks,
Luwei Kang

> >
> > >
> > >
> > > Thanks,
> > >
> > > ~Andrew
> >
> >
> > Best regards,
> > Michał Leszczyński
> > CERT Polska

On Wed, Jun 17, 2020 at 06:45:22AM +0000, Kang, Luwei wrote:
> > -----Original Message-----
> > From: Tian, Kevin <kevin.tian@intel.com>
> > Sent: Wednesday, June 17, 2020 9:35 AM
> > To: Michał Leszczyński <michal.leszczynski@cert.pl>; Andrew Cooper
> > <andrew.cooper3@citrix.com>
> > Cc: Xen-devel <xen-devel@lists.xenproject.org>; Jan Beulich
> > <jbeulich@suse.com>; Wei Liu <wl@xen.org>; Roger Pau Monné
> > <roger.pau@citrix.com>; Nakajima, Jun <jun.nakajima@intel.com>; George
> > Dunlap <george.dunlap@citrix.com>; Ian Jackson <ian.jackson@eu.citrix.com>;
> > Julien Grall <julien@xen.org>; Stefano Stabellini <sstabellini@kernel.org>;
> > Kang, Luwei <luwei.kang@intel.com>
> > Subject: RE: [PATCH v1 0/7] Implement support for external IPT monitoring
> > 
> > +Luwei, who developed PT for KVM and is the best one who can help
> > review VMX changes from Intel side. Please include him in future post or
> > discussion.
> > 
> > > -----Original Message-----
> > > From: Michał Leszczyński <michal.leszczynski@cert.pl>
> > > Sent: Wednesday, June 17, 2020 2:48 AM
> > > To: Andrew Cooper <andrew.cooper3@citrix.com>
> > > Cc: Xen-devel <xen-devel@lists.xenproject.org>; Jan Beulich
> > > <jbeulich@suse.com>; Wei Liu <wl@xen.org>; Roger Pau Monné
> > > <roger.pau@citrix.com>; Nakajima, Jun <jun.nakajima@intel.com>; Tian,
> > > Kevin <kevin.tian@intel.com>; George Dunlap
> > > <george.dunlap@citrix.com>; Ian Jackson <ian.jackson@eu.citrix.com>;
> > > Julien Grall <julien@xen.org>; Stefano Stabellini
> > > <sstabellini@kernel.org>
> > > Subject: Re: [PATCH v1 0/7] Implement support for external IPT
> > > monitoring
> > >
> > > ----- 16 cze 2020 o 20:17, Andrew Cooper andrew.cooper3@citrix.com
> > > napisał(a):
> > >
> > > > On 16/06/2020 16:16, Michał Leszczyński wrote:
> > > > When this subject was broached on xen-devel before, one issue was
> > > > the fact that all actions which are intercepted don't end up writing
> > > > any appropriate packets.  This is perhaps less of an issue for this
> > > > example, where the external agent can see VMExits in the trace, but
> > > > it still results in missing information.  (It is a major problem for
> > > > PT within the guest, and needs Xen's intercept/emulation framework
> > > > being updated to be PT-aware so it can fill in the same packets
> > > > which hardware would have done for equivalent actions.)
> > >
> > > Ok, this sounds like a hard issue. Could you point out what could be
> > > the particular problematic cases? For instance, if something would
> > > alter EIP/RIP or CR3 then I belive it would still be recorded in PT
> > > trace (i.e. these values will be logged on VM entry).
> 
> e.g. If a VM exit is taken on a guest write to CR3 (including “MOV CR3” as well as task switches), the PIP packet
> normally generated on the CR3 write will be missing. The PIP packet needs to be written to the PT buffer by software. Another example is VM-exit taken on RDTSC. 
> 
> For VM introspection, all the Intel PT packets may need to emulated by software. Some description in SDM as below:
> If a VMM emulates an element of processor state by taking a VM exit on reads and/or writes to that piece of state, and the state element impacts Intel PT packet generation or values, it may be incumbent upon the VMM to insert or modify the output trace data.

I got the impression that IPT was mostly useful together with
introspection, as you can then get events from trapped instructions
(and likely emulated) from the introspection interface, while being
able to get the processor trace for non-trapped events.

I'm not sure whether there would be corner cases with trapped
instructions not being handled by the introspection framework.

How does KVM deal with this, do they insert/modify trace packets on
trapped and emulated instructions by the VMM?

Roger.

> > > -----Original Message-----
> > > From: Tian, Kevin <kevin.tian@intel.com>
> > > Sent: Wednesday, June 17, 2020 9:35 AM
> > > To: Michał Leszczyński <michal.leszczynski@cert.pl>; Andrew Cooper
> > > <andrew.cooper3@citrix.com>
> > > Cc: Xen-devel <xen-devel@lists.xenproject.org>; Jan Beulich
> > > <jbeulich@suse.com>; Wei Liu <wl@xen.org>; Roger Pau Monné
> > > <roger.pau@citrix.com>; Nakajima, Jun <jun.nakajima@intel.com>;
> > > George Dunlap <george.dunlap@citrix.com>; Ian Jackson
> > > <ian.jackson@eu.citrix.com>; Julien Grall <julien@xen.org>; Stefano
> > > Stabellini <sstabellini@kernel.org>; Kang, Luwei
> > > <luwei.kang@intel.com>
> > > Subject: RE: [PATCH v1 0/7] Implement support for external IPT
> > > monitoring
> > >
> > > +Luwei, who developed PT for KVM and is the best one who can help
> > > review VMX changes from Intel side. Please include him in future
> > > post or discussion.
> > >
> > > > -----Original Message-----
> > > > From: Michał Leszczyński <michal.leszczynski@cert.pl>
> > > > Sent: Wednesday, June 17, 2020 2:48 AM
> > > > To: Andrew Cooper <andrew.cooper3@citrix.com>
> > > > Cc: Xen-devel <xen-devel@lists.xenproject.org>; Jan Beulich
> > > > <jbeulich@suse.com>; Wei Liu <wl@xen.org>; Roger Pau Monné
> > > > <roger.pau@citrix.com>; Nakajima, Jun <jun.nakajima@intel.com>;
> > > > Tian, Kevin <kevin.tian@intel.com>; George Dunlap
> > > > <george.dunlap@citrix.com>; Ian Jackson
> > > > <ian.jackson@eu.citrix.com>; Julien Grall <julien@xen.org>;
> > > > Stefano Stabellini <sstabellini@kernel.org>
> > > > Subject: Re: [PATCH v1 0/7] Implement support for external IPT
> > > > monitoring
> > > >
> > > > ----- 16 cze 2020 o 20:17, Andrew Cooper andrew.cooper3@citrix.com
> > > > napisał(a):
> > > >
> > > > > On 16/06/2020 16:16, Michał Leszczyński wrote:
> > > > > When this subject was broached on xen-devel before, one issue
> > > > > was the fact that all actions which are intercepted don't end up
> > > > > writing any appropriate packets.  This is perhaps less of an
> > > > > issue for this example, where the external agent can see VMExits
> > > > > in the trace, but it still results in missing information.  (It
> > > > > is a major problem for PT within the guest, and needs Xen's
> > > > > intercept/emulation framework being updated to be PT-aware so it
> > > > > can fill in the same packets which hardware would have done for
> > > > > equivalent actions.)
> > > >
> > > > Ok, this sounds like a hard issue. Could you point out what could
> > > > be the particular problematic cases? For instance, if something
> > > > would alter EIP/RIP or CR3 then I belive it would still be
> > > > recorded in PT trace (i.e. these values will be logged on VM entry).
> >
> > e.g. If a VM exit is taken on a guest write to CR3 (including “MOV
> > CR3” as well as task switches), the PIP packet normally generated on the CR3
> write will be missing. The PIP packet needs to be written to the PT buffer by
> software. Another example is VM-exit taken on RDTSC.
> >
> > For VM introspection, all the Intel PT packets may need to emulated by
> software. Some description in SDM as below:
> > If a VMM emulates an element of processor state by taking a VM exit on
> reads and/or writes to that piece of state, and the state element impacts Intel
> PT packet generation or values, it may be incumbent upon the VMM to insert
> or modify the output trace data.
> 
> I got the impression that IPT was mostly useful together with introspection, as
> you can then get events from trapped instructions (and likely emulated) from
> the introspection interface, while being able to get the processor trace for non-
> trapped events.
> 
> I'm not sure whether there would be corner cases with trapped instructions
> not being handled by the introspection framework.
> 
> How does KVM deal with this, do they insert/modify trace packets on trapped
> and emulated instructions by the VMM?

The KVM includes instruction decoder and emulator(arch/x86/kvm/emulate.c), and the guest's memory can be set to write-protect as well. But it doesn't support Intel PT packets software emulator. For KVM, the Intel PT feature will be exposed to KVM guest and KVM guest can use Intel PT feature like native.

Thanks,
Luwei Kang

On Wed, Jun 17, 2020 at 12:37:13PM +0000, Kang, Luwei wrote:
> > How does KVM deal with this, do they insert/modify trace packets on trapped
> > and emulated instructions by the VMM?
> 
> The KVM includes instruction decoder and emulator(arch/x86/kvm/emulate.c), and the guest's memory can be set to write-protect as well. But it doesn't support Intel PT packets software emulator. For KVM, the Intel PT feature will be exposed to KVM guest and KVM guest can use Intel PT feature like native.

But if such feature is exposed to the guest for it's own usage, won't
it be missing packets for instructions emulated by the VMM?

Thanks, Roger.

> > > How does KVM deal with this, do they insert/modify trace packets on
> > > trapped and emulated instructions by the VMM?
> >
> > The KVM includes instruction decoder and
> emulator(arch/x86/kvm/emulate.c), and the guest's memory can be set to
> write-protect as well. But it doesn't support Intel PT packets software emulator.
> For KVM, the Intel PT feature will be exposed to KVM guest and KVM guest can
> use Intel PT feature like native.
> 
> But if such feature is exposed to the guest for it's own usage, won't it be
> missing packets for instructions emulated by the VMM?

If setting the guest's memory write-protect, I think yes. 

Thanks,
Luwei Kang

> 
> Thanks, Roger.

----- 18 cze 2020 o 1:29, Kang, Luwei luwei.kang@intel.com napisał(a):

>> > > How does KVM deal with this, do they insert/modify trace packets on
>> > > trapped and emulated instructions by the VMM?
>> >
>> > The KVM includes instruction decoder and
>> emulator(arch/x86/kvm/emulate.c), and the guest's memory can be set to
>> write-protect as well. But it doesn't support Intel PT packets software
>> emulator.
>> For KVM, the Intel PT feature will be exposed to KVM guest and KVM guest can
>> use Intel PT feature like native.
>> 
>> But if such feature is exposed to the guest for it's own usage, won't it be
>> missing packets for instructions emulated by the VMM?
> 
> If setting the guest's memory write-protect, I think yes.


Thus, I propose to leave it as it is right now. If somebody is purposely altering the VM state then he/she should consult not only the IPT but also understand what was done "in the meantime" by additional features, e.g. when something was altered by vm_event callback. As Tamas said previously, we usually just want to see certain path leading to vmexit.

Please also note that there is a PTWRITE instruction that could be used in the future in order to add custom payloads/hints to the PT trace, when needed.


> 
> Thanks,
> Luwei Kang
> 
>> 
> > Thanks, Roger.

On Thu, Jun 18, 2020 at 02:56:17AM +0200, Michał Leszczyński wrote:
> ----- 18 cze 2020 o 1:29, Kang, Luwei luwei.kang@intel.com napisał(a):
> 
> >> > > How does KVM deal with this, do they insert/modify trace packets on
> >> > > trapped and emulated instructions by the VMM?
> >> >
> >> > The KVM includes instruction decoder and
> >> emulator(arch/x86/kvm/emulate.c), and the guest's memory can be set to
> >> write-protect as well. But it doesn't support Intel PT packets software
> >> emulator.
> >> For KVM, the Intel PT feature will be exposed to KVM guest and KVM guest can
> >> use Intel PT feature like native.
> >> 
> >> But if such feature is exposed to the guest for it's own usage, won't it be
> >> missing packets for instructions emulated by the VMM?
> > 
> > If setting the guest's memory write-protect, I think yes.
> 
> 
> Thus, I propose to leave it as it is right now. If somebody is purposely altering the VM state then he/she should consult not only the IPT but also understand what was done "in the meantime" by additional features, e.g. when something was altered by vm_event callback. As Tamas said previously, we usually just want to see certain path leading to vmexit.
> 
> Please also note that there is a PTWRITE instruction that could be used in the future in order to add custom payloads/hints to the PT trace, when needed.

Yes, I think the usage of IPT by a third party against a guest is
fine, as such third party can also use introspection and get the
information about the emulated instructions.

OTOH exposing the feature to the guest itself for it's own usage seems
wrong without adding the packets related to the instructions emulated.

I understand the current series only cares about the first option, so
that's perfectly fine.

Roger.