Re: [PATCH 00/10] KVM PCIe/MSI passthrough on ARM/ARM64
On Wed, Feb 03, 2016 at 01:10:58PM +, Will Deacon wrote: > On Wed, Feb 03, 2016 at 01:50:47PM +0100, Christoffer Dall wrote: > > On Mon, Feb 01, 2016 at 02:03:51PM +, Will Deacon wrote: > > > On Fri, Jan 29, 2016 at 10:25:52PM +0100, Eric Auger wrote: > > > > On 01/29/2016 08:33 PM, Alex Williamson wrote: > > > > >>> We know that x86 handles MSI vectors specially, so there is some > > > > >>> hardware that helps the situation. It's not just that x86 has a > > > > >>> fixed > > > > >>> range for MSI, it's how it manages that range when interrupt > > > > >>> remapping > > > > >>> hardware is enabled. A device table indexed by source-ID > > > > >>> references a > > > > >>> per device table indexed by data from the MSI write itself. So we > > > > >>> get > > > > >>> much, much finer granularity, > > > > >> About the granularity, I think ARM GICv3 now provides a similar > > > > >> capability with GICv3 ITS (interrupt translation service). Along with > > > > >> the MSI MSG write transaction, the device outputs a DeviceID > > > > >> conveyed on > > > > >> the bus. This DeviceID (~ your source-ID) enables to index a device > > > > >> table. The entry in the device table points to a DeviceId interrupt > > > > >> translation table indexed by the EventID found in the msi msg. So the > > > > >> entry in the interrupt translation table eventually gives you the > > > > >> eventual interrupt ID targeted by the MSI MSG. > > > > >> This translation capability if not available in GICv2M though, ie. > > > > >> the > > > > >> one I am currently using. > > > > >> > > > > >> Those tables currently are built by the ITS irqchip > > > > >> (irq-gic-v3-its.c) > > > > > > That's right. GICv3/ITS disambiguates the interrupt source using the > > > DeviceID, which for PCI is derived from the Requester ID of the endpoint. > > > GICv2m is less flexible and requires a separate physical frame per guest > > > to achieve isolation. > > > > > We should still support MSI passthrough with a single MSI frame host > > system though, right? > > I think we should treat the frame as an exclusive resource and assign it > to a single VM. so on a single frame GICv2m system, either your host or a single VM gets to do MSIs... > > > (Users should just be aware that guests are not fully protected against > > misbehaving hardware in that case). > > Is it confined to misbehaving hardware? What if a malicious/buggy guest > configures its device to DMA all over the doorbell? > I guess not, I suppose we can't trap any configuration access and mediate that for any device. Bummer. -Christoffer
Re: [PATCH 00/10] KVM PCIe/MSI passthrough on ARM/ARM64
On Wed, Feb 03, 2016 at 01:50:47PM +0100, Christoffer Dall wrote: > On Mon, Feb 01, 2016 at 02:03:51PM +, Will Deacon wrote: > > On Fri, Jan 29, 2016 at 10:25:52PM +0100, Eric Auger wrote: > > > On 01/29/2016 08:33 PM, Alex Williamson wrote: > > > >>> We know that x86 handles MSI vectors specially, so there is some > > > >>> hardware that helps the situation. It's not just that x86 has a fixed > > > >>> range for MSI, it's how it manages that range when interrupt remapping > > > >>> hardware is enabled. A device table indexed by source-ID references a > > > >>> per device table indexed by data from the MSI write itself. So we get > > > >>> much, much finer granularity, > > > >> About the granularity, I think ARM GICv3 now provides a similar > > > >> capability with GICv3 ITS (interrupt translation service). Along with > > > >> the MSI MSG write transaction, the device outputs a DeviceID conveyed > > > >> on > > > >> the bus. This DeviceID (~ your source-ID) enables to index a device > > > >> table. The entry in the device table points to a DeviceId interrupt > > > >> translation table indexed by the EventID found in the msi msg. So the > > > >> entry in the interrupt translation table eventually gives you the > > > >> eventual interrupt ID targeted by the MSI MSG. > > > >> This translation capability if not available in GICv2M though, ie. the > > > >> one I am currently using. > > > >> > > > >> Those tables currently are built by the ITS irqchip (irq-gic-v3-its.c) > > > > That's right. GICv3/ITS disambiguates the interrupt source using the > > DeviceID, which for PCI is derived from the Requester ID of the endpoint. > > GICv2m is less flexible and requires a separate physical frame per guest > > to achieve isolation. > > > We should still support MSI passthrough with a single MSI frame host > system though, right? I think we should treat the frame as an exclusive resource and assign it to a single VM. > (Users should just be aware that guests are not fully protected against > misbehaving hardware in that case). Is it confined to misbehaving hardware? What if a malicious/buggy guest configures its device to DMA all over the doorbell? Will
Re: [PATCH 00/10] KVM PCIe/MSI passthrough on ARM/ARM64
On Mon, Feb 01, 2016 at 02:03:51PM +, Will Deacon wrote: > On Fri, Jan 29, 2016 at 10:25:52PM +0100, Eric Auger wrote: > > On 01/29/2016 08:33 PM, Alex Williamson wrote: > > >>> We know that x86 handles MSI vectors specially, so there is some > > >>> hardware that helps the situation. It's not just that x86 has a fixed > > >>> range for MSI, it's how it manages that range when interrupt remapping > > >>> hardware is enabled. A device table indexed by source-ID references a > > >>> per device table indexed by data from the MSI write itself. So we get > > >>> much, much finer granularity, > > >> About the granularity, I think ARM GICv3 now provides a similar > > >> capability with GICv3 ITS (interrupt translation service). Along with > > >> the MSI MSG write transaction, the device outputs a DeviceID conveyed on > > >> the bus. This DeviceID (~ your source-ID) enables to index a device > > >> table. The entry in the device table points to a DeviceId interrupt > > >> translation table indexed by the EventID found in the msi msg. So the > > >> entry in the interrupt translation table eventually gives you the > > >> eventual interrupt ID targeted by the MSI MSG. > > >> This translation capability if not available in GICv2M though, ie. the > > >> one I am currently using. > > >> > > >> Those tables currently are built by the ITS irqchip (irq-gic-v3-its.c) > > That's right. GICv3/ITS disambiguates the interrupt source using the > DeviceID, which for PCI is derived from the Requester ID of the endpoint. > GICv2m is less flexible and requires a separate physical frame per guest > to achieve isolation. > We should still support MSI passthrough with a single MSI frame host system though, right? (Users should just be aware that guests are not fully protected against misbehaving hardware in that case). -Christoffer
Re: [PATCH 00/10] KVM PCIe/MSI passthrough on ARM/ARM64
On Wed, Feb 03, 2016 at 01:50:47PM +0100, Christoffer Dall wrote: > On Mon, Feb 01, 2016 at 02:03:51PM +, Will Deacon wrote: > > On Fri, Jan 29, 2016 at 10:25:52PM +0100, Eric Auger wrote: > > > On 01/29/2016 08:33 PM, Alex Williamson wrote: > > > >>> We know that x86 handles MSI vectors specially, so there is some > > > >>> hardware that helps the situation. It's not just that x86 has a fixed > > > >>> range for MSI, it's how it manages that range when interrupt remapping > > > >>> hardware is enabled. A device table indexed by source-ID references a > > > >>> per device table indexed by data from the MSI write itself. So we get > > > >>> much, much finer granularity, > > > >> About the granularity, I think ARM GICv3 now provides a similar > > > >> capability with GICv3 ITS (interrupt translation service). Along with > > > >> the MSI MSG write transaction, the device outputs a DeviceID conveyed > > > >> on > > > >> the bus. This DeviceID (~ your source-ID) enables to index a device > > > >> table. The entry in the device table points to a DeviceId interrupt > > > >> translation table indexed by the EventID found in the msi msg. So the > > > >> entry in the interrupt translation table eventually gives you the > > > >> eventual interrupt ID targeted by the MSI MSG. > > > >> This translation capability if not available in GICv2M though, ie. the > > > >> one I am currently using. > > > >> > > > >> Those tables currently are built by the ITS irqchip (irq-gic-v3-its.c) > > > > That's right. GICv3/ITS disambiguates the interrupt source using the > > DeviceID, which for PCI is derived from the Requester ID of the endpoint. > > GICv2m is less flexible and requires a separate physical frame per guest > > to achieve isolation. > > > We should still support MSI passthrough with a single MSI frame host > system though, right? I think we should treat the frame as an exclusive resource and assign it to a single VM. > (Users should just be aware that guests are not fully protected against > misbehaving hardware in that case). Is it confined to misbehaving hardware? What if a malicious/buggy guest configures its device to DMA all over the doorbell? Will
Re: [PATCH 00/10] KVM PCIe/MSI passthrough on ARM/ARM64
On Mon, Feb 01, 2016 at 02:03:51PM +, Will Deacon wrote: > On Fri, Jan 29, 2016 at 10:25:52PM +0100, Eric Auger wrote: > > On 01/29/2016 08:33 PM, Alex Williamson wrote: > > >>> We know that x86 handles MSI vectors specially, so there is some > > >>> hardware that helps the situation. It's not just that x86 has a fixed > > >>> range for MSI, it's how it manages that range when interrupt remapping > > >>> hardware is enabled. A device table indexed by source-ID references a > > >>> per device table indexed by data from the MSI write itself. So we get > > >>> much, much finer granularity, > > >> About the granularity, I think ARM GICv3 now provides a similar > > >> capability with GICv3 ITS (interrupt translation service). Along with > > >> the MSI MSG write transaction, the device outputs a DeviceID conveyed on > > >> the bus. This DeviceID (~ your source-ID) enables to index a device > > >> table. The entry in the device table points to a DeviceId interrupt > > >> translation table indexed by the EventID found in the msi msg. So the > > >> entry in the interrupt translation table eventually gives you the > > >> eventual interrupt ID targeted by the MSI MSG. > > >> This translation capability if not available in GICv2M though, ie. the > > >> one I am currently using. > > >> > > >> Those tables currently are built by the ITS irqchip (irq-gic-v3-its.c) > > That's right. GICv3/ITS disambiguates the interrupt source using the > DeviceID, which for PCI is derived from the Requester ID of the endpoint. > GICv2m is less flexible and requires a separate physical frame per guest > to achieve isolation. > We should still support MSI passthrough with a single MSI frame host system though, right? (Users should just be aware that guests are not fully protected against misbehaving hardware in that case). -Christoffer
Re: [PATCH 00/10] KVM PCIe/MSI passthrough on ARM/ARM64
On Wed, Feb 03, 2016 at 01:10:58PM +, Will Deacon wrote: > On Wed, Feb 03, 2016 at 01:50:47PM +0100, Christoffer Dall wrote: > > On Mon, Feb 01, 2016 at 02:03:51PM +, Will Deacon wrote: > > > On Fri, Jan 29, 2016 at 10:25:52PM +0100, Eric Auger wrote: > > > > On 01/29/2016 08:33 PM, Alex Williamson wrote: > > > > >>> We know that x86 handles MSI vectors specially, so there is some > > > > >>> hardware that helps the situation. It's not just that x86 has a > > > > >>> fixed > > > > >>> range for MSI, it's how it manages that range when interrupt > > > > >>> remapping > > > > >>> hardware is enabled. A device table indexed by source-ID > > > > >>> references a > > > > >>> per device table indexed by data from the MSI write itself. So we > > > > >>> get > > > > >>> much, much finer granularity, > > > > >> About the granularity, I think ARM GICv3 now provides a similar > > > > >> capability with GICv3 ITS (interrupt translation service). Along with > > > > >> the MSI MSG write transaction, the device outputs a DeviceID > > > > >> conveyed on > > > > >> the bus. This DeviceID (~ your source-ID) enables to index a device > > > > >> table. The entry in the device table points to a DeviceId interrupt > > > > >> translation table indexed by the EventID found in the msi msg. So the > > > > >> entry in the interrupt translation table eventually gives you the > > > > >> eventual interrupt ID targeted by the MSI MSG. > > > > >> This translation capability if not available in GICv2M though, ie. > > > > >> the > > > > >> one I am currently using. > > > > >> > > > > >> Those tables currently are built by the ITS irqchip > > > > >> (irq-gic-v3-its.c) > > > > > > That's right. GICv3/ITS disambiguates the interrupt source using the > > > DeviceID, which for PCI is derived from the Requester ID of the endpoint. > > > GICv2m is less flexible and requires a separate physical frame per guest > > > to achieve isolation. > > > > > We should still support MSI passthrough with a single MSI frame host > > system though, right? > > I think we should treat the frame as an exclusive resource and assign it > to a single VM. so on a single frame GICv2m system, either your host or a single VM gets to do MSIs... > > > (Users should just be aware that guests are not fully protected against > > misbehaving hardware in that case). > > Is it confined to misbehaving hardware? What if a malicious/buggy guest > configures its device to DMA all over the doorbell? > I guess not, I suppose we can't trap any configuration access and mediate that for any device. Bummer. -Christoffer
Re: [PATCH 00/10] KVM PCIe/MSI passthrough on ARM/ARM64
On Fri, Jan 29, 2016 at 10:25:52PM +0100, Eric Auger wrote: > On 01/29/2016 08:33 PM, Alex Williamson wrote: > >>> We know that x86 handles MSI vectors specially, so there is some > >>> hardware that helps the situation. It's not just that x86 has a fixed > >>> range for MSI, it's how it manages that range when interrupt remapping > >>> hardware is enabled. A device table indexed by source-ID references a > >>> per device table indexed by data from the MSI write itself. So we get > >>> much, much finer granularity, > >> About the granularity, I think ARM GICv3 now provides a similar > >> capability with GICv3 ITS (interrupt translation service). Along with > >> the MSI MSG write transaction, the device outputs a DeviceID conveyed on > >> the bus. This DeviceID (~ your source-ID) enables to index a device > >> table. The entry in the device table points to a DeviceId interrupt > >> translation table indexed by the EventID found in the msi msg. So the > >> entry in the interrupt translation table eventually gives you the > >> eventual interrupt ID targeted by the MSI MSG. > >> This translation capability if not available in GICv2M though, ie. the > >> one I am currently using. > >> > >> Those tables currently are built by the ITS irqchip (irq-gic-v3-its.c) That's right. GICv3/ITS disambiguates the interrupt source using the DeviceID, which for PCI is derived from the Requester ID of the endpoint. GICv2m is less flexible and requires a separate physical frame per guest to achieve isolation. Will
Re: [PATCH 00/10] KVM PCIe/MSI passthrough on ARM/ARM64
On Fri, Jan 29, 2016 at 10:25:52PM +0100, Eric Auger wrote: > On 01/29/2016 08:33 PM, Alex Williamson wrote: > >>> We know that x86 handles MSI vectors specially, so there is some > >>> hardware that helps the situation. It's not just that x86 has a fixed > >>> range for MSI, it's how it manages that range when interrupt remapping > >>> hardware is enabled. A device table indexed by source-ID references a > >>> per device table indexed by data from the MSI write itself. So we get > >>> much, much finer granularity, > >> About the granularity, I think ARM GICv3 now provides a similar > >> capability with GICv3 ITS (interrupt translation service). Along with > >> the MSI MSG write transaction, the device outputs a DeviceID conveyed on > >> the bus. This DeviceID (~ your source-ID) enables to index a device > >> table. The entry in the device table points to a DeviceId interrupt > >> translation table indexed by the EventID found in the msi msg. So the > >> entry in the interrupt translation table eventually gives you the > >> eventual interrupt ID targeted by the MSI MSG. > >> This translation capability if not available in GICv2M though, ie. the > >> one I am currently using. > >> > >> Those tables currently are built by the ITS irqchip (irq-gic-v3-its.c) That's right. GICv3/ITS disambiguates the interrupt source using the DeviceID, which for PCI is derived from the Requester ID of the endpoint. GICv2m is less flexible and requires a separate physical frame per guest to achieve isolation. Will
Re: [PATCH 00/10] KVM PCIe/MSI passthrough on ARM/ARM64
Hi Alex, On 01/29/2016 08:33 PM, Alex Williamson wrote: > On Fri, 2016-01-29 at 15:35 +0100, Eric Auger wrote: >> Hi Alex, >> On 01/28/2016 10:51 PM, Alex Williamson wrote: >>> On Tue, 2016-01-26 at 13:12 +, Eric Auger wrote: This series addresses KVM PCIe passthrough with MSI enabled on ARM/ARM64. It pursues the efforts done on [1], [2], [3]. It also aims at covering the same need on some PowerPC platforms. On x86 all accesses to the 1MB PA region [FEE0_h - FEF0_000h] are directed as interrupt messages: accesses to this special PA window directly target the APIC configuration space and not DRAM, meaning the downstream IOMMU is bypassed. This is not the case on above mentionned platforms where MSI messages emitted by devices are conveyed through the IOMMU. This means an IOVA/host PA mapping must exist for the MSI to reach the MSI controller. Normal way to create IOVA bindings consists in using VFIO DMA MAP API. However in this case the MSI IOVA is not mapped onto guest RAM but on host physical page (the MSI controller frame). Following first comments, the spirit of [2] is kept: the guest registers an IOVA range reserved for MSI mapping. When the VFIO-PCIe driver allocates its MSI vectors, it overwrites the MSI controller physical address with an IOVA, allocated within the window provided by the userspace. This IOVA is mapped onto the MSI controller frame physical page. The series does not address yet the problematic of telling the userspace how much IOVA he should provision. >>> >>> I'm sort of on a think-different approach today, so bear with me; how is >>> it that x86 can make interrupt remapping so transparent to drivers like >>> vfio-pci while for ARM and ppc we seem to be stuck with doing these >>> fixups of the physical vector ourselves, implying ugly (no offense) >>> paths bouncing through vfio to connect the driver and iommu backends? >>> >>> We know that x86 handles MSI vectors specially, so there is some >>> hardware that helps the situation. It's not just that x86 has a fixed >>> range for MSI, it's how it manages that range when interrupt remapping >>> hardware is enabled. A device table indexed by source-ID references a >>> per device table indexed by data from the MSI write itself. So we get >>> much, much finer granularity, >> About the granularity, I think ARM GICv3 now provides a similar >> capability with GICv3 ITS (interrupt translation service). Along with >> the MSI MSG write transaction, the device outputs a DeviceID conveyed on >> the bus. This DeviceID (~ your source-ID) enables to index a device >> table. The entry in the device table points to a DeviceId interrupt >> translation table indexed by the EventID found in the msi msg. So the >> entry in the interrupt translation table eventually gives you the >> eventual interrupt ID targeted by the MSI MSG. >> This translation capability if not available in GICv2M though, ie. the >> one I am currently using. >> >> Those tables currently are built by the ITS irqchip (irq-gic-v3-its.c) > > So it sounds like the interrupt remapping plumbing needs to be > implemented for those chips. How does ITS identify an MSI versus any > other DMA write? Does it need to be within a preconfigured address > space like on x86 or does it know this implicitly by the transaction > (which doesn't seem possible on PCIe)? It seems there is a kind of misunderstanding here. Assuming a "simple" system with a single ITS, all devices likely to produce MSI must write those messages in a single register, located in the ITS MSI 64kB frame (this register is called GITS_TRANSLATER). Then the ITS discriminates between senders using the DeviceID conveyed out-of-band on the bus (or by other implementation defined means). For those DeviceId, a deviceId interrupt translation table is supposed to exist, else it is going to fault. If any "undeclared" device is writing into that register, its deviceid will be unknown. It looks like on Intel the interrupt remapping HW rather is abstracted on the IOMMU side; I did not take time yet to carefully read the VT-d spec but maybe the Intel interrupt remapping HW rather acts as an IOMMU that takes an input MSI address within the famous window and apply a translation scheme based on the MSI address & data? On ARM the input MSI address always is the GITS_TRANSLATER and then the translation scheme is based on out-of-band info (deviceid) + data content(eventid). I Hope this clarifies. > > Along with this discussion, we should probably be revisiting whether > existing ARM SMMUs should be exposing the IOMMU_CAP_INTR_REMAP > capability. so according to the above explanation not sure it is relevant. Will/Marc might correct me if I told some wrong things. This capability is meant to indicate interrupt isolation, > but if an entire page of IOVA space is
Re: [PATCH 00/10] KVM PCIe/MSI passthrough on ARM/ARM64
On Fri, 2016-01-29 at 15:35 +0100, Eric Auger wrote: > Hi Alex, > On 01/28/2016 10:51 PM, Alex Williamson wrote: > > On Tue, 2016-01-26 at 13:12 +, Eric Auger wrote: > > > This series addresses KVM PCIe passthrough with MSI enabled on ARM/ARM64. > > > It pursues the efforts done on [1], [2], [3]. It also aims at covering the > > > same need on some PowerPC platforms. > > > > > > On x86 all accesses to the 1MB PA region [FEE0_h - FEF0_000h] are > > > directed > > > as interrupt messages: accesses to this special PA window directly target > > > the > > > APIC configuration space and not DRAM, meaning the downstream IOMMU is > > > bypassed. > > > > > > This is not the case on above mentionned platforms where MSI messages > > > emitted > > > by devices are conveyed through the IOMMU. This means an IOVA/host PA > > > mapping > > > must exist for the MSI to reach the MSI controller. Normal way to create > > > IOVA bindings consists in using VFIO DMA MAP API. However in this case > > > the MSI IOVA is not mapped onto guest RAM but on host physical page (the > > > MSI > > > controller frame). > > > > > > Following first comments, the spirit of [2] is kept: the guest registers > > > an IOVA range reserved for MSI mapping. When the VFIO-PCIe driver > > > allocates > > > its MSI vectors, it overwrites the MSI controller physical address with > > > an IOVA, > > > allocated within the window provided by the userspace. This IOVA is mapped > > > onto the MSI controller frame physical page. > > > > > > The series does not address yet the problematic of telling the userspace > > > how > > > much IOVA he should provision. > > > > I'm sort of on a think-different approach today, so bear with me; how is > > it that x86 can make interrupt remapping so transparent to drivers like > > vfio-pci while for ARM and ppc we seem to be stuck with doing these > > fixups of the physical vector ourselves, implying ugly (no offense) > > paths bouncing through vfio to connect the driver and iommu backends? > > > > We know that x86 handles MSI vectors specially, so there is some > > hardware that helps the situation. It's not just that x86 has a fixed > > range for MSI, it's how it manages that range when interrupt remapping > > hardware is enabled. A device table indexed by source-ID references a > > per device table indexed by data from the MSI write itself. So we get > > much, much finer granularity, > About the granularity, I think ARM GICv3 now provides a similar > capability with GICv3 ITS (interrupt translation service). Along with > the MSI MSG write transaction, the device outputs a DeviceID conveyed on > the bus. This DeviceID (~ your source-ID) enables to index a device > table. The entry in the device table points to a DeviceId interrupt > translation table indexed by the EventID found in the msi msg. So the > entry in the interrupt translation table eventually gives you the > eventual interrupt ID targeted by the MSI MSG. > This translation capability if not available in GICv2M though, ie. the > one I am currently using. > > Those tables currently are built by the ITS irqchip (irq-gic-v3-its.c) So it sounds like the interrupt remapping plumbing needs to be implemented for those chips. How does ITS identify an MSI versus any other DMA write? Does it need to be within a preconfigured address space like on x86 or does it know this implicitly by the transaction (which doesn't seem possible on PCIe)? Along with this discussion, we should probably be revisiting whether existing ARM SMMUs should be exposing the IOMMU_CAP_INTR_REMAP capability. This capability is meant to indicate interrupt isolation, but if an entire page of IOVA space is mapped through the IOMMU to a range of interrupts and some of those interrupts are shared with host devices or other VMs, then we really don't have that isolation and the system is susceptible to one VM interfering with another or with the host. If that's the case, the SMMU should not be claiming IOMMU_CAP_INTR_REMAP. > but there's still effectively an interrupt > > domain per device that's being transparently managed under the covers > > whenever we request an MSI vector for a device. > > > > So why can't we do something more like that here? There's no predefined > > MSI vector range, so defining an interface for the user to specify that > > is unavoidable. > Do you confirm that VFIO user API still still is the good choice to > provide that IOVA range? I don't see that we have an option there unless ARM wants to retroactively reserve a range of IOVA space in the spec, which is certainly not going to happen. The only other thing that comes to mind would be if there was an existing address space which could never be backed by RAM or other DMA capable targets. But that seems far fetched as well. > But why shouldn't everything else be transparent? We > > could add an interface to the IOMMU API that allows us to register that > > reserved range for the
Re: [PATCH 00/10] KVM PCIe/MSI passthrough on ARM/ARM64
Hi Alex, On 01/28/2016 10:51 PM, Alex Williamson wrote: > On Tue, 2016-01-26 at 13:12 +, Eric Auger wrote: >> This series addresses KVM PCIe passthrough with MSI enabled on ARM/ARM64. >> It pursues the efforts done on [1], [2], [3]. It also aims at covering the >> same need on some PowerPC platforms. >> >> On x86 all accesses to the 1MB PA region [FEE0_h - FEF0_000h] are >> directed >> as interrupt messages: accesses to this special PA window directly target the >> APIC configuration space and not DRAM, meaning the downstream IOMMU is >> bypassed. >> >> This is not the case on above mentionned platforms where MSI messages emitted >> by devices are conveyed through the IOMMU. This means an IOVA/host PA mapping >> must exist for the MSI to reach the MSI controller. Normal way to create >> IOVA bindings consists in using VFIO DMA MAP API. However in this case >> the MSI IOVA is not mapped onto guest RAM but on host physical page (the MSI >> controller frame). >> >> Following first comments, the spirit of [2] is kept: the guest registers >> an IOVA range reserved for MSI mapping. When the VFIO-PCIe driver allocates >> its MSI vectors, it overwrites the MSI controller physical address with an >> IOVA, >> allocated within the window provided by the userspace. This IOVA is mapped >> onto the MSI controller frame physical page. >> >> The series does not address yet the problematic of telling the userspace how >> much IOVA he should provision. > > I'm sort of on a think-different approach today, so bear with me; how is > it that x86 can make interrupt remapping so transparent to drivers like > vfio-pci while for ARM and ppc we seem to be stuck with doing these > fixups of the physical vector ourselves, implying ugly (no offense) > paths bouncing through vfio to connect the driver and iommu backends? > > We know that x86 handles MSI vectors specially, so there is some > hardware that helps the situation. It's not just that x86 has a fixed > range for MSI, it's how it manages that range when interrupt remapping > hardware is enabled. A device table indexed by source-ID references a > per device table indexed by data from the MSI write itself. So we get > much, much finer granularity, About the granularity, I think ARM GICv3 now provides a similar capability with GICv3 ITS (interrupt translation service). Along with the MSI MSG write transaction, the device outputs a DeviceID conveyed on the bus. This DeviceID (~ your source-ID) enables to index a device table. The entry in the device table points to a DeviceId interrupt translation table indexed by the EventID found in the msi msg. So the entry in the interrupt translation table eventually gives you the eventual interrupt ID targeted by the MSI MSG. This translation capability if not available in GICv2M though, ie. the one I am currently using. Those tables currently are built by the ITS irqchip (irq-gic-v3-its.c) but there's still effectively an interrupt > domain per device that's being transparently managed under the covers > whenever we request an MSI vector for a device. > > So why can't we do something more like that here? There's no predefined > MSI vector range, so defining an interface for the user to specify that > is unavoidable. Do you confirm that VFIO user API still still is the good choice to provide that IOVA range? But why shouldn't everything else be transparent? We > could add an interface to the IOMMU API that allows us to register that > reserved range for the IOMMU domain. IOMMU-core (or maybe interrupt > remapping) code might allocate an IOVA domain for this just as you've > done in the type1 code here. I have no objection to move that iova allocation scheme somewhere else. I just need to figure out how to deal with the fact iova.c is not compiled everywhere as I noticed too late ;-) But rather than having any interaction > with vfio-pci, why not do this at lower levels such that the platform > interrupt vector allocation code automatically uses one of those IOVA > ranges and returns the IOVA rather than the physical address for the PCI > code to program into the device? I think we know what needs to be done, > but we're taking the approach of managing the space ourselves and doing > a fixup of the device after the core code has done its job when we > really ought to be letting the core code manage a space that we define > and programming the device so that it doesn't need a fixup in the > vfio-pci code. Wouldn't it be nicer if pci_enable_msix_range() returned > with the device properly programmed or generate an error if there's not > enough reserved mapping space in IOMMU domain? Can it be done? I agree with you on the fact it would be cleaner to manage that natively at MSI controller level instead of patching the address value in vfio_pci_intrs.c. I will investigate in that direction but I need some more time to understand the links between the MSI controller, the PCI device and the IOMMU. Best Regards
Re: [PATCH 00/10] KVM PCIe/MSI passthrough on ARM/ARM64
Hi Alex, On 01/28/2016 10:51 PM, Alex Williamson wrote: > On Tue, 2016-01-26 at 13:12 +, Eric Auger wrote: >> This series addresses KVM PCIe passthrough with MSI enabled on ARM/ARM64. >> It pursues the efforts done on [1], [2], [3]. It also aims at covering the >> same need on some PowerPC platforms. >> >> On x86 all accesses to the 1MB PA region [FEE0_h - FEF0_000h] are >> directed >> as interrupt messages: accesses to this special PA window directly target the >> APIC configuration space and not DRAM, meaning the downstream IOMMU is >> bypassed. >> >> This is not the case on above mentionned platforms where MSI messages emitted >> by devices are conveyed through the IOMMU. This means an IOVA/host PA mapping >> must exist for the MSI to reach the MSI controller. Normal way to create >> IOVA bindings consists in using VFIO DMA MAP API. However in this case >> the MSI IOVA is not mapped onto guest RAM but on host physical page (the MSI >> controller frame). >> >> Following first comments, the spirit of [2] is kept: the guest registers >> an IOVA range reserved for MSI mapping. When the VFIO-PCIe driver allocates >> its MSI vectors, it overwrites the MSI controller physical address with an >> IOVA, >> allocated within the window provided by the userspace. This IOVA is mapped >> onto the MSI controller frame physical page. >> >> The series does not address yet the problematic of telling the userspace how >> much IOVA he should provision. > > I'm sort of on a think-different approach today, so bear with me; how is > it that x86 can make interrupt remapping so transparent to drivers like > vfio-pci while for ARM and ppc we seem to be stuck with doing these > fixups of the physical vector ourselves, implying ugly (no offense) > paths bouncing through vfio to connect the driver and iommu backends? > > We know that x86 handles MSI vectors specially, so there is some > hardware that helps the situation. It's not just that x86 has a fixed > range for MSI, it's how it manages that range when interrupt remapping > hardware is enabled. A device table indexed by source-ID references a > per device table indexed by data from the MSI write itself. So we get > much, much finer granularity, About the granularity, I think ARM GICv3 now provides a similar capability with GICv3 ITS (interrupt translation service). Along with the MSI MSG write transaction, the device outputs a DeviceID conveyed on the bus. This DeviceID (~ your source-ID) enables to index a device table. The entry in the device table points to a DeviceId interrupt translation table indexed by the EventID found in the msi msg. So the entry in the interrupt translation table eventually gives you the eventual interrupt ID targeted by the MSI MSG. This translation capability if not available in GICv2M though, ie. the one I am currently using. Those tables currently are built by the ITS irqchip (irq-gic-v3-its.c) but there's still effectively an interrupt > domain per device that's being transparently managed under the covers > whenever we request an MSI vector for a device. > > So why can't we do something more like that here? There's no predefined > MSI vector range, so defining an interface for the user to specify that > is unavoidable. Do you confirm that VFIO user API still still is the good choice to provide that IOVA range? But why shouldn't everything else be transparent? We > could add an interface to the IOMMU API that allows us to register that > reserved range for the IOMMU domain. IOMMU-core (or maybe interrupt > remapping) code might allocate an IOVA domain for this just as you've > done in the type1 code here. I have no objection to move that iova allocation scheme somewhere else. I just need to figure out how to deal with the fact iova.c is not compiled everywhere as I noticed too late ;-) But rather than having any interaction > with vfio-pci, why not do this at lower levels such that the platform > interrupt vector allocation code automatically uses one of those IOVA > ranges and returns the IOVA rather than the physical address for the PCI > code to program into the device? I think we know what needs to be done, > but we're taking the approach of managing the space ourselves and doing > a fixup of the device after the core code has done its job when we > really ought to be letting the core code manage a space that we define > and programming the device so that it doesn't need a fixup in the > vfio-pci code. Wouldn't it be nicer if pci_enable_msix_range() returned > with the device properly programmed or generate an error if there's not > enough reserved mapping space in IOMMU domain? Can it be done? I agree with you on the fact it would be cleaner to manage that natively at MSI controller level instead of patching the address value in vfio_pci_intrs.c. I will investigate in that direction but I need some more time to understand the links between the MSI controller, the PCI device and the IOMMU. Best Regards
Re: [PATCH 00/10] KVM PCIe/MSI passthrough on ARM/ARM64
On Fri, 2016-01-29 at 15:35 +0100, Eric Auger wrote: > Hi Alex, > On 01/28/2016 10:51 PM, Alex Williamson wrote: > > On Tue, 2016-01-26 at 13:12 +, Eric Auger wrote: > > > This series addresses KVM PCIe passthrough with MSI enabled on ARM/ARM64. > > > It pursues the efforts done on [1], [2], [3]. It also aims at covering the > > > same need on some PowerPC platforms. > > > > > > On x86 all accesses to the 1MB PA region [FEE0_h - FEF0_000h] are > > > directed > > > as interrupt messages: accesses to this special PA window directly target > > > the > > > APIC configuration space and not DRAM, meaning the downstream IOMMU is > > > bypassed. > > > > > > This is not the case on above mentionned platforms where MSI messages > > > emitted > > > by devices are conveyed through the IOMMU. This means an IOVA/host PA > > > mapping > > > must exist for the MSI to reach the MSI controller. Normal way to create > > > IOVA bindings consists in using VFIO DMA MAP API. However in this case > > > the MSI IOVA is not mapped onto guest RAM but on host physical page (the > > > MSI > > > controller frame). > > > > > > Following first comments, the spirit of [2] is kept: the guest registers > > > an IOVA range reserved for MSI mapping. When the VFIO-PCIe driver > > > allocates > > > its MSI vectors, it overwrites the MSI controller physical address with > > > an IOVA, > > > allocated within the window provided by the userspace. This IOVA is mapped > > > onto the MSI controller frame physical page. > > > > > > The series does not address yet the problematic of telling the userspace > > > how > > > much IOVA he should provision. > > > > I'm sort of on a think-different approach today, so bear with me; how is > > it that x86 can make interrupt remapping so transparent to drivers like > > vfio-pci while for ARM and ppc we seem to be stuck with doing these > > fixups of the physical vector ourselves, implying ugly (no offense) > > paths bouncing through vfio to connect the driver and iommu backends? > > > > We know that x86 handles MSI vectors specially, so there is some > > hardware that helps the situation. It's not just that x86 has a fixed > > range for MSI, it's how it manages that range when interrupt remapping > > hardware is enabled. A device table indexed by source-ID references a > > per device table indexed by data from the MSI write itself. So we get > > much, much finer granularity, > About the granularity, I think ARM GICv3 now provides a similar > capability with GICv3 ITS (interrupt translation service). Along with > the MSI MSG write transaction, the device outputs a DeviceID conveyed on > the bus. This DeviceID (~ your source-ID) enables to index a device > table. The entry in the device table points to a DeviceId interrupt > translation table indexed by the EventID found in the msi msg. So the > entry in the interrupt translation table eventually gives you the > eventual interrupt ID targeted by the MSI MSG. > This translation capability if not available in GICv2M though, ie. the > one I am currently using. > > Those tables currently are built by the ITS irqchip (irq-gic-v3-its.c) So it sounds like the interrupt remapping plumbing needs to be implemented for those chips. How does ITS identify an MSI versus any other DMA write? Does it need to be within a preconfigured address space like on x86 or does it know this implicitly by the transaction (which doesn't seem possible on PCIe)? Along with this discussion, we should probably be revisiting whether existing ARM SMMUs should be exposing the IOMMU_CAP_INTR_REMAP capability. This capability is meant to indicate interrupt isolation, but if an entire page of IOVA space is mapped through the IOMMU to a range of interrupts and some of those interrupts are shared with host devices or other VMs, then we really don't have that isolation and the system is susceptible to one VM interfering with another or with the host. If that's the case, the SMMU should not be claiming IOMMU_CAP_INTR_REMAP. > but there's still effectively an interrupt > > domain per device that's being transparently managed under the covers > > whenever we request an MSI vector for a device. > > > > So why can't we do something more like that here? There's no predefined > > MSI vector range, so defining an interface for the user to specify that > > is unavoidable. > Do you confirm that VFIO user API still still is the good choice to > provide that IOVA range? I don't see that we have an option there unless ARM wants to retroactively reserve a range of IOVA space in the spec, which is certainly not going to happen. The only other thing that comes to mind would be if there was an existing address space which could never be backed by RAM or other DMA capable targets. But that seems far fetched as well. > But why shouldn't everything else be transparent? We > > could add an interface to the IOMMU API that allows us to register that > > reserved range for the
Re: [PATCH 00/10] KVM PCIe/MSI passthrough on ARM/ARM64
Hi Alex, On 01/29/2016 08:33 PM, Alex Williamson wrote: > On Fri, 2016-01-29 at 15:35 +0100, Eric Auger wrote: >> Hi Alex, >> On 01/28/2016 10:51 PM, Alex Williamson wrote: >>> On Tue, 2016-01-26 at 13:12 +, Eric Auger wrote: This series addresses KVM PCIe passthrough with MSI enabled on ARM/ARM64. It pursues the efforts done on [1], [2], [3]. It also aims at covering the same need on some PowerPC platforms. On x86 all accesses to the 1MB PA region [FEE0_h - FEF0_000h] are directed as interrupt messages: accesses to this special PA window directly target the APIC configuration space and not DRAM, meaning the downstream IOMMU is bypassed. This is not the case on above mentionned platforms where MSI messages emitted by devices are conveyed through the IOMMU. This means an IOVA/host PA mapping must exist for the MSI to reach the MSI controller. Normal way to create IOVA bindings consists in using VFIO DMA MAP API. However in this case the MSI IOVA is not mapped onto guest RAM but on host physical page (the MSI controller frame). Following first comments, the spirit of [2] is kept: the guest registers an IOVA range reserved for MSI mapping. When the VFIO-PCIe driver allocates its MSI vectors, it overwrites the MSI controller physical address with an IOVA, allocated within the window provided by the userspace. This IOVA is mapped onto the MSI controller frame physical page. The series does not address yet the problematic of telling the userspace how much IOVA he should provision. >>> >>> I'm sort of on a think-different approach today, so bear with me; how is >>> it that x86 can make interrupt remapping so transparent to drivers like >>> vfio-pci while for ARM and ppc we seem to be stuck with doing these >>> fixups of the physical vector ourselves, implying ugly (no offense) >>> paths bouncing through vfio to connect the driver and iommu backends? >>> >>> We know that x86 handles MSI vectors specially, so there is some >>> hardware that helps the situation. It's not just that x86 has a fixed >>> range for MSI, it's how it manages that range when interrupt remapping >>> hardware is enabled. A device table indexed by source-ID references a >>> per device table indexed by data from the MSI write itself. So we get >>> much, much finer granularity, >> About the granularity, I think ARM GICv3 now provides a similar >> capability with GICv3 ITS (interrupt translation service). Along with >> the MSI MSG write transaction, the device outputs a DeviceID conveyed on >> the bus. This DeviceID (~ your source-ID) enables to index a device >> table. The entry in the device table points to a DeviceId interrupt >> translation table indexed by the EventID found in the msi msg. So the >> entry in the interrupt translation table eventually gives you the >> eventual interrupt ID targeted by the MSI MSG. >> This translation capability if not available in GICv2M though, ie. the >> one I am currently using. >> >> Those tables currently are built by the ITS irqchip (irq-gic-v3-its.c) > > So it sounds like the interrupt remapping plumbing needs to be > implemented for those chips. How does ITS identify an MSI versus any > other DMA write? Does it need to be within a preconfigured address > space like on x86 or does it know this implicitly by the transaction > (which doesn't seem possible on PCIe)? It seems there is a kind of misunderstanding here. Assuming a "simple" system with a single ITS, all devices likely to produce MSI must write those messages in a single register, located in the ITS MSI 64kB frame (this register is called GITS_TRANSLATER). Then the ITS discriminates between senders using the DeviceID conveyed out-of-band on the bus (or by other implementation defined means). For those DeviceId, a deviceId interrupt translation table is supposed to exist, else it is going to fault. If any "undeclared" device is writing into that register, its deviceid will be unknown. It looks like on Intel the interrupt remapping HW rather is abstracted on the IOMMU side; I did not take time yet to carefully read the VT-d spec but maybe the Intel interrupt remapping HW rather acts as an IOMMU that takes an input MSI address within the famous window and apply a translation scheme based on the MSI address & data? On ARM the input MSI address always is the GITS_TRANSLATER and then the translation scheme is based on out-of-band info (deviceid) + data content(eventid). I Hope this clarifies. > > Along with this discussion, we should probably be revisiting whether > existing ARM SMMUs should be exposing the IOMMU_CAP_INTR_REMAP > capability. so according to the above explanation not sure it is relevant. Will/Marc might correct me if I told some wrong things. This capability is meant to indicate interrupt isolation, > but if an entire page of IOVA space is
Re: [PATCH 00/10] KVM PCIe/MSI passthrough on ARM/ARM64
On Tue, 2016-01-26 at 13:12 +, Eric Auger wrote: > This series addresses KVM PCIe passthrough with MSI enabled on ARM/ARM64. > It pursues the efforts done on [1], [2], [3]. It also aims at covering the > same need on some PowerPC platforms. > > On x86 all accesses to the 1MB PA region [FEE0_h - FEF0_000h] are directed > as interrupt messages: accesses to this special PA window directly target the > APIC configuration space and not DRAM, meaning the downstream IOMMU is > bypassed. > > This is not the case on above mentionned platforms where MSI messages emitted > by devices are conveyed through the IOMMU. This means an IOVA/host PA mapping > must exist for the MSI to reach the MSI controller. Normal way to create > IOVA bindings consists in using VFIO DMA MAP API. However in this case > the MSI IOVA is not mapped onto guest RAM but on host physical page (the MSI > controller frame). > > Following first comments, the spirit of [2] is kept: the guest registers > an IOVA range reserved for MSI mapping. When the VFIO-PCIe driver allocates > its MSI vectors, it overwrites the MSI controller physical address with an > IOVA, > allocated within the window provided by the userspace. This IOVA is mapped > onto the MSI controller frame physical page. > > The series does not address yet the problematic of telling the userspace how > much IOVA he should provision. I'm sort of on a think-different approach today, so bear with me; how is it that x86 can make interrupt remapping so transparent to drivers like vfio-pci while for ARM and ppc we seem to be stuck with doing these fixups of the physical vector ourselves, implying ugly (no offense) paths bouncing through vfio to connect the driver and iommu backends? We know that x86 handles MSI vectors specially, so there is some hardware that helps the situation. It's not just that x86 has a fixed range for MSI, it's how it manages that range when interrupt remapping hardware is enabled. A device table indexed by source-ID references a per device table indexed by data from the MSI write itself. So we get much, much finer granularity, but there's still effectively an interrupt domain per device that's being transparently managed under the covers whenever we request an MSI vector for a device. So why can't we do something more like that here? There's no predefined MSI vector range, so defining an interface for the user to specify that is unavoidable. But why shouldn't everything else be transparent? We could add an interface to the IOMMU API that allows us to register that reserved range for the IOMMU domain. IOMMU-core (or maybe interrupt remapping) code might allocate an IOVA domain for this just as you've done in the type1 code here. But rather than having any interaction with vfio-pci, why not do this at lower levels such that the platform interrupt vector allocation code automatically uses one of those IOVA ranges and returns the IOVA rather than the physical address for the PCI code to program into the device? I think we know what needs to be done, but we're taking the approach of managing the space ourselves and doing a fixup of the device after the core code has done its job when we really ought to be letting the core code manage a space that we define and programming the device so that it doesn't need a fixup in the vfio-pci code. Wouldn't it be nicer if pci_enable_msix_range() returned with the device properly programmed or generate an error if there's not enough reserved mapping space in IOMMU domain? Can it be done? Thanks, Alex
Re: [PATCH 00/10] KVM PCIe/MSI passthrough on ARM/ARM64
Hi Pavel, On 01/28/2016 08:13 AM, Pavel Fedin wrote: > Hello! > >> x86 isn't problem-free in this space. An x86 VM is going to know that >> the 0xfee0 address range is special, it won't be backed by RAM and >> won't be a DMA target, thus we'll never attempt to map it for an iova >> address. However, if we run a non-x86 VM or a userspace driver, it >> doesn't necessarily know that there's anything special about that range >> of iovas. I intend to resolve this with an extension to the iommu info >> ioctl that describes the available iova space for the iommu. The >> interrupt region would simply be excluded. > > I see now, but i still don't understand how it would work. How can we tell > the guest OS that we cannot do DMA to this particular > area? Just exclude it from RAM at all? But this means we would have to modify > machine's model... > I know that this is a bit different story from what we are implementing now. > Just curious. Well in QEMU mach-virt we have a static guest PA memory map. Maybe in some other virt machines this is different and it is possible to take into account the fact an IOVA range cannot be used? Regards Eric > > Kind regards, > Pavel Fedin > Senior Engineer > Samsung Electronics Research center Russia > >
Re: [PATCH 00/10] KVM PCIe/MSI passthrough on ARM/ARM64
On Tue, 2016-01-26 at 13:12 +, Eric Auger wrote: > This series addresses KVM PCIe passthrough with MSI enabled on ARM/ARM64. > It pursues the efforts done on [1], [2], [3]. It also aims at covering the > same need on some PowerPC platforms. > > On x86 all accesses to the 1MB PA region [FEE0_h - FEF0_000h] are directed > as interrupt messages: accesses to this special PA window directly target the > APIC configuration space and not DRAM, meaning the downstream IOMMU is > bypassed. > > This is not the case on above mentionned platforms where MSI messages emitted > by devices are conveyed through the IOMMU. This means an IOVA/host PA mapping > must exist for the MSI to reach the MSI controller. Normal way to create > IOVA bindings consists in using VFIO DMA MAP API. However in this case > the MSI IOVA is not mapped onto guest RAM but on host physical page (the MSI > controller frame). > > Following first comments, the spirit of [2] is kept: the guest registers > an IOVA range reserved for MSI mapping. When the VFIO-PCIe driver allocates > its MSI vectors, it overwrites the MSI controller physical address with an > IOVA, > allocated within the window provided by the userspace. This IOVA is mapped > onto the MSI controller frame physical page. > > The series does not address yet the problematic of telling the userspace how > much IOVA he should provision. I'm sort of on a think-different approach today, so bear with me; how is it that x86 can make interrupt remapping so transparent to drivers like vfio-pci while for ARM and ppc we seem to be stuck with doing these fixups of the physical vector ourselves, implying ugly (no offense) paths bouncing through vfio to connect the driver and iommu backends? We know that x86 handles MSI vectors specially, so there is some hardware that helps the situation. It's not just that x86 has a fixed range for MSI, it's how it manages that range when interrupt remapping hardware is enabled. A device table indexed by source-ID references a per device table indexed by data from the MSI write itself. So we get much, much finer granularity, but there's still effectively an interrupt domain per device that's being transparently managed under the covers whenever we request an MSI vector for a device. So why can't we do something more like that here? There's no predefined MSI vector range, so defining an interface for the user to specify that is unavoidable. But why shouldn't everything else be transparent? We could add an interface to the IOMMU API that allows us to register that reserved range for the IOMMU domain. IOMMU-core (or maybe interrupt remapping) code might allocate an IOVA domain for this just as you've done in the type1 code here. But rather than having any interaction with vfio-pci, why not do this at lower levels such that the platform interrupt vector allocation code automatically uses one of those IOVA ranges and returns the IOVA rather than the physical address for the PCI code to program into the device? I think we know what needs to be done, but we're taking the approach of managing the space ourselves and doing a fixup of the device after the core code has done its job when we really ought to be letting the core code manage a space that we define and programming the device so that it doesn't need a fixup in the vfio-pci code. Wouldn't it be nicer if pci_enable_msix_range() returned with the device properly programmed or generate an error if there's not enough reserved mapping space in IOMMU domain? Can it be done? Thanks, Alex
Re: [PATCH 00/10] KVM PCIe/MSI passthrough on ARM/ARM64
Hi Pavel, On 01/28/2016 08:13 AM, Pavel Fedin wrote: > Hello! > >> x86 isn't problem-free in this space. An x86 VM is going to know that >> the 0xfee0 address range is special, it won't be backed by RAM and >> won't be a DMA target, thus we'll never attempt to map it for an iova >> address. However, if we run a non-x86 VM or a userspace driver, it >> doesn't necessarily know that there's anything special about that range >> of iovas. I intend to resolve this with an extension to the iommu info >> ioctl that describes the available iova space for the iommu. The >> interrupt region would simply be excluded. > > I see now, but i still don't understand how it would work. How can we tell > the guest OS that we cannot do DMA to this particular > area? Just exclude it from RAM at all? But this means we would have to modify > machine's model... > I know that this is a bit different story from what we are implementing now. > Just curious. Well in QEMU mach-virt we have a static guest PA memory map. Maybe in some other virt machines this is different and it is possible to take into account the fact an IOVA range cannot be used? Regards Eric > > Kind regards, > Pavel Fedin > Senior Engineer > Samsung Electronics Research center Russia > >
RE: [PATCH 00/10] KVM PCIe/MSI passthrough on ARM/ARM64
Hello! > x86 isn't problem-free in this space. An x86 VM is going to know that > the 0xfee0 address range is special, it won't be backed by RAM and > won't be a DMA target, thus we'll never attempt to map it for an iova > address. However, if we run a non-x86 VM or a userspace driver, it > doesn't necessarily know that there's anything special about that range > of iovas. I intend to resolve this with an extension to the iommu info > ioctl that describes the available iova space for the iommu. The > interrupt region would simply be excluded. I see now, but i still don't understand how it would work. How can we tell the guest OS that we cannot do DMA to this particular area? Just exclude it from RAM at all? But this means we would have to modify machine's model... I know that this is a bit different story from what we are implementing now. Just curious. Kind regards, Pavel Fedin Senior Engineer Samsung Electronics Research center Russia
Re: [PATCH 00/10] KVM PCIe/MSI passthrough on ARM/ARM64
Hi Pavel, On 01/26/2016 06:25 PM, Pavel Fedin wrote: > Hello! > I'd like just to clarify some things for myself and better wrap my head > around it... > >> On x86 all accesses to the 1MB PA region [FEE0_h - FEF0_000h] are >> directed >> as interrupt messages: accesses to this special PA window directly target the >> APIC configuration space and not DRAM, meaning the downstream IOMMU is >> bypassed. > > So, this is effectively the same as always having hardwired 1:1 mappings on > all IOMMUs, isn't it ? > If so, then we can't we just do the same, just by forcing similar 1:1 > mapping? This is what i tried to do in my patchset. All of > you are talking about a situation which arises when we are emulating > different machine with different physical addresses layout. And > e. g. if our host has MSI at 0xABADCAFE, our target could have valid RAM at > the same location, and we need to handle it somehow, > therefore we have to move our MSI window out of target's RAM. But how does > this work on a PC then? What if our host is PC, and we > want to emulate some ARM board, which has RAM at FE00 ? Or does it mean > that PC architecture is flawed and can reliably handle > PCI passthrough only for itself ? Alex answered to this I think: " x86 isn't problem-free in this space. An x86 VM is going to know that the 0xfee0 address range is special, it won't be backed by RAM and won't be a DMA target, thus we'll never attempt to map it for an iova address. However, if we run a non-x86 VM or a userspace driver, it doesn't necessarily know that there's anything special about that range of iovas. I intend to resolve this with an extension to the iommu info ioctl that describes the available iova space for the iommu. The interrupt region would simply be excluded. " I am not sure I've addressed this requirement yet but it seems more future proof to have an IOMMU mapping for those addresses. For the ARM use case I think Marc gave guidance: " We want userspace to be in control of the memory map, and it is the kernel's job to tell us whether or not this matches the HW capabilities or not. A fixed mapping may completely clash with the memory map I want (think emulating HW x on platform y), and there is no reason why we should have the restrictions x86 has. " That's the rationale behind respining that way. Waiting for other comments & discussions, I am going to address the iova and dma_addr_t kbuilt reported compilation issues. Please apologize for those. Best Regards Eric > > Kind regards, > Pavel Fedin > Senior Engineer > Samsung Electronics Research center Russia > >
RE: [PATCH 00/10] KVM PCIe/MSI passthrough on ARM/ARM64
Hello! > x86 isn't problem-free in this space. An x86 VM is going to know that > the 0xfee0 address range is special, it won't be backed by RAM and > won't be a DMA target, thus we'll never attempt to map it for an iova > address. However, if we run a non-x86 VM or a userspace driver, it > doesn't necessarily know that there's anything special about that range > of iovas. I intend to resolve this with an extension to the iommu info > ioctl that describes the available iova space for the iommu. The > interrupt region would simply be excluded. I see now, but i still don't understand how it would work. How can we tell the guest OS that we cannot do DMA to this particular area? Just exclude it from RAM at all? But this means we would have to modify machine's model... I know that this is a bit different story from what we are implementing now. Just curious. Kind regards, Pavel Fedin Senior Engineer Samsung Electronics Research center Russia
Re: [PATCH 00/10] KVM PCIe/MSI passthrough on ARM/ARM64
Hi Pavel, On 01/26/2016 06:25 PM, Pavel Fedin wrote: > Hello! > I'd like just to clarify some things for myself and better wrap my head > around it... > >> On x86 all accesses to the 1MB PA region [FEE0_h - FEF0_000h] are >> directed >> as interrupt messages: accesses to this special PA window directly target the >> APIC configuration space and not DRAM, meaning the downstream IOMMU is >> bypassed. > > So, this is effectively the same as always having hardwired 1:1 mappings on > all IOMMUs, isn't it ? > If so, then we can't we just do the same, just by forcing similar 1:1 > mapping? This is what i tried to do in my patchset. All of > you are talking about a situation which arises when we are emulating > different machine with different physical addresses layout. And > e. g. if our host has MSI at 0xABADCAFE, our target could have valid RAM at > the same location, and we need to handle it somehow, > therefore we have to move our MSI window out of target's RAM. But how does > this work on a PC then? What if our host is PC, and we > want to emulate some ARM board, which has RAM at FE00 ? Or does it mean > that PC architecture is flawed and can reliably handle > PCI passthrough only for itself ? Alex answered to this I think: " x86 isn't problem-free in this space. An x86 VM is going to know that the 0xfee0 address range is special, it won't be backed by RAM and won't be a DMA target, thus we'll never attempt to map it for an iova address. However, if we run a non-x86 VM or a userspace driver, it doesn't necessarily know that there's anything special about that range of iovas. I intend to resolve this with an extension to the iommu info ioctl that describes the available iova space for the iommu. The interrupt region would simply be excluded. " I am not sure I've addressed this requirement yet but it seems more future proof to have an IOMMU mapping for those addresses. For the ARM use case I think Marc gave guidance: " We want userspace to be in control of the memory map, and it is the kernel's job to tell us whether or not this matches the HW capabilities or not. A fixed mapping may completely clash with the memory map I want (think emulating HW x on platform y), and there is no reason why we should have the restrictions x86 has. " That's the rationale behind respining that way. Waiting for other comments & discussions, I am going to address the iova and dma_addr_t kbuilt reported compilation issues. Please apologize for those. Best Regards Eric > > Kind regards, > Pavel Fedin > Senior Engineer > Samsung Electronics Research center Russia > >
RE: [PATCH 00/10] KVM PCIe/MSI passthrough on ARM/ARM64
Hello! I'd like just to clarify some things for myself and better wrap my head around it... > On x86 all accesses to the 1MB PA region [FEE0_h - FEF0_000h] are directed > as interrupt messages: accesses to this special PA window directly target the > APIC configuration space and not DRAM, meaning the downstream IOMMU is > bypassed. So, this is effectively the same as always having hardwired 1:1 mappings on all IOMMUs, isn't it ? If so, then we can't we just do the same, just by forcing similar 1:1 mapping? This is what i tried to do in my patchset. All of you are talking about a situation which arises when we are emulating different machine with different physical addresses layout. And e. g. if our host has MSI at 0xABADCAFE, our target could have valid RAM at the same location, and we need to handle it somehow, therefore we have to move our MSI window out of target's RAM. But how does this work on a PC then? What if our host is PC, and we want to emulate some ARM board, which has RAM at FE00 ? Or does it mean that PC architecture is flawed and can reliably handle PCI passthrough only for itself ? Kind regards, Pavel Fedin Senior Engineer Samsung Electronics Research center Russia
RE: [PATCH 00/10] KVM PCIe/MSI passthrough on ARM/ARM64
Hello! I'd like just to clarify some things for myself and better wrap my head around it... > On x86 all accesses to the 1MB PA region [FEE0_h - FEF0_000h] are directed > as interrupt messages: accesses to this special PA window directly target the > APIC configuration space and not DRAM, meaning the downstream IOMMU is > bypassed. So, this is effectively the same as always having hardwired 1:1 mappings on all IOMMUs, isn't it ? If so, then we can't we just do the same, just by forcing similar 1:1 mapping? This is what i tried to do in my patchset. All of you are talking about a situation which arises when we are emulating different machine with different physical addresses layout. And e. g. if our host has MSI at 0xABADCAFE, our target could have valid RAM at the same location, and we need to handle it somehow, therefore we have to move our MSI window out of target's RAM. But how does this work on a PC then? What if our host is PC, and we want to emulate some ARM board, which has RAM at FE00 ? Or does it mean that PC architecture is flawed and can reliably handle PCI passthrough only for itself ? Kind regards, Pavel Fedin Senior Engineer Samsung Electronics Research center Russia