IRQFD support with GICv3 ITS (WAS: RE: [PATCH 00/13] arm64: KVM: GICv3 ITS emulation)
Hello guys! Currently on ARM, irqfd supports routing an host eventfd towards a virtual SPI: eventfd - vSPI = gsi+32 parameters of irqfd are the eventfd and the gsi. Yes, but this works only with GICv2m, because it actually turns MSI data into SPI number. ITS works in a completely different way. 2) now we have virtual msi injection, we could use msi routing to inject virtual LPI's. But is it what you need for your qemu integration? Actually this is what i wanted to discuss here... I have studied a little bit IRQ routing mechanism... And it comes to a question what is 'GSI'. As far as i could understand, on x86 GSI is a completely virtual entity, which can be tied either to irqchip's pin (physical IRQ) or MSI event. There is totally no correspondence between GSI numbers and guest IRQ numbers. GSIs are just allocated by the userspace starting from 0 and on. Is my understanding correct? On ARM, i see, completely different approach is used. For KVM_IRQ_LINE ioctl GSI is actually a raw GIC IRQ number plus some extra bits for target and type. For KVM_IRQFD with GICv2m GSI is actually SPI number (starting from zero, so that IRQ = GSI + 32). First of all, i would say that we already have an inconsistence in ARM API. The same thing called GSI has two different meanings for different functions. I think it would be a bad idea to introduce a third, separate meaning for MSIs. However, this is what we could do: Approach 1: GICv2m way. We could add one more IOCTL which would decode MSI data into IRQ (in our case it's LPI) number. What it would return is LPI - 32, to keep in line with existing convention. Pros: does not bring any more inconsistence into KVM API. Cons: requires adding one more IOCTL and one more MSI handling mechanism. Isn't there too many of them already? Approach 2: IRQ routing. We could implement MSI routing using virtual GSI numbers. In order to stay compatible with what we have, we could say that GSI numbers below 8192 are SPI GSIs, and everything starting from 8192 is MSI. Then we could use KVM_SET_GSI_ROUTING ioctl to assign these GSIs to actual MSIs which then will go full-cycle through ITS. Pros: Does not introduce any new APIs. Cons: - Introduces third meaning for GSI on ARM. - Slower than approach 1 because in that case every interrupt is pre-translated while here we engage ITS every time. Personally i have already tried approach 1 and i can say that it works. There is no problem with target specification because current ITS code stores everything in a single bunch so that i anyway have to locate a particular ITTE corresponding to an LPI and get collection ID from there. However, yes, i agree, this approach has the same performance drawback as my suggested approach 2. Any thoughts / ideas ? Kind regards, Pavel Fedin Expert Engineer Samsung Electronics Research center Russia ___ kvmarm mailing list kvmarm@lists.cs.columbia.edu https://lists.cs.columbia.edu/mailman/listinfo/kvmarm
Re: IRQFD support with GICv3 ITS (WAS: RE: [PATCH 00/13] arm64: KVM: GICv3 ITS emulation)
Hi, On 06/10/2015 10:31 AM, Pavel Fedin wrote: Hello guys! Currently on ARM, irqfd supports routing an host eventfd towards a virtual SPI: eventfd - vSPI = gsi+32 parameters of irqfd are the eventfd and the gsi. Yes, but this works only with GICv2m, because it actually turns MSI data into SPI number. ITS works in a completely different way. 2) now we have virtual msi injection, we could use msi routing to inject virtual LPI's. But is it what you need for your qemu integration? Actually this is what i wanted to discuss here... I have studied a little bit IRQ routing mechanism... And it comes to a question what is 'GSI'. As far as i could understand, on x86 GSI is a completely virtual entity, which can be tied either to irqchip's pin (physical IRQ) or MSI event. There is totally no correspondence between GSI numbers and guest IRQ numbers. GSIs are just allocated by the userspace starting from 0 and on. Is my understanding correct? Well I think as long as you use irqchip routing, gsi is not random. When looking at arch/x86/kvm/irq_comm.c and in kvm_set_routing_entry you can see there is an offset applied on the gsi and irqchip/pin depending on type of irqchip (pic_master, pic_slave, ioapic). originally done by BIOS? This is the default routing. Now in qemu the irqchip routing entries are built by kvm_irqchip_add_irq_route in hw/intc/openpic_kvm.c, hw/i386/kvm/ioapic.c, ... and then committed. When looking at MSI routing and qemu integration, in case kvm_gsi_direct_mapping is NOT used, kvm_irqchp_get_virq indeed finds out a gsi belonging to the gsi range and not mapped with irqchip entries. if kvm_gsi_direct_mapping is used, an irqchip mapped gsi is used instead. At least this is my understanding On ARM, i see, completely different approach is used. For KVM_IRQ_LINE ioctl GSI is actually a raw GIC IRQ number plus some extra bits for target and type. For KVM_IRQFD with GICv2m GSI is actually SPI number (starting from zero, so that IRQ = GSI + 32). First of all, i would say that we already have an inconsistence in ARM API. The same thing called GSI has two different meanings for different functions. well that's true. This avoided to have ARM archi specific adaptations for VFIO and MMIO VHOST-NET proto at that time. Now with the MSI advent, those adaptations becomes needed anyway. Also we concluded it was not meaningful to inject PPIs. gsi irqfd argument directly matches the IRQ number found in guest device tree. I think it would be a bad idea to introduce a third, separate meaning for MSIs. However, this is what we could do: Approach 1: GICv2m way. We could add one more IOCTL which would decode MSI data into IRQ (in our case it's LPI) number. What it would return is LPI - 32, to keep in line with existing convention. Pros: does not bring any more inconsistence into KVM API. Cons: requires adding one more IOCTL and one more MSI handling mechanism. Isn't there too many of them already? indeed in newly added qemu kvm-all.c kvm_arch_msi_data_to_gsi we could call a new ioctl that translates the data + deviceid? into an LPI and program irqfd with that LPI. This is done once when setting irqfd up. This also means extending irqfd support to lpi injection, gsi being the LPI index if gsi = 8192. in that case we continue using kvm_gsi_direct_mapping and gsi still is an IRQ index. Approach 2: IRQ routing. We could implement MSI routing using virtual GSI numbers. In order to stay compatible with what we have, we could say that GSI numbers below 8192 are SPI GSIs, and everything starting from 8192 is MSI. I think the gsi can be considered as an index: 0 - 1020: SPI index = 8192: LPI index Then we could use KVM_SET_GSI_ROUTING ioctl to assign these GSIs to actual MSIs which then will go full-cycle through ITS. Pros: Does not introduce any new APIs. Cons: - Introduces third meaning for GSI on ARM. - Slower than approach 1 because in that case every interrupt is pre-translated while here we engage ITS every time. KVM GSI routing, even if only used for MSI routing then mandates to build entries for non MSI IRQs, using irqchip routing entries. Then you draw the irqchip.c kvm_irq_routing_table chip[KVM_NR_IRQCHIPS][KVM_IRQCHIP_NUM_PINS] static allocation issue. I guess this code would need to be revisited to accomodate large space and variable pin number of GIC. Hope it helps Best Regards Eric Personally i have already tried approach 1 and i can say that it works. There is no problem with target specification because current ITS code stores everything in a single bunch so that i anyway have to locate a particular ITTE corresponding to an LPI and get collection ID from there. However, yes, i agree, this approach has the same performance drawback as my suggested approach 2. Any thoughts / ideas ? Kind regards, Pavel Fedin Expert Engineer Samsung Electronics Research center Russia
RE: IRQFD support with GICv3 ITS (WAS: RE: [PATCH 00/13] arm64: KVM: GICv3 ITS emulation)
Hi! indeed in newly added qemu kvm-all.c kvm_arch_msi_data_to_gsi we could call a new ioctl that translates the data + deviceid? into an LPI and program irqfd with that LPI. This is done once when setting irqfd up. This also means extending irqfd support to lpi injection, gsi being the LPI index if gsi = 8192. in that case we continue using kvm_gsi_direct_mapping and gsi still is an IRQ index. This is exactly what i have done in my kernel + qemu. I have added a new KVM capability and then in qemu i do this: --- cut --- if (kvm_gsi_kernel_mapping()) { struct kvm_msi msi; msi.address_lo = (uint32_t)msg.address; msi.address_hi = msg.address 32; msi.data = le32_to_cpu(msg.data); memset(msi.pad, 0, sizeof(msi.pad)); if (dev) { msi.devid = (pci_bus_num(dev-bus) 8) | dev-devfn; msi.flags = KVM_MSI_VALID_DEVID; } else { msi.devid = 0; msi.flags = 0; } return kvm_vm_ioctl(s, KVM_TRANSLATE_MSI, msi); } --- cut --- KVM_TRANSLATE_MSI returns an LPI number. This seemed to be the simplest and fastest thing to do. If someone is interested, i could prepare an RFC patch series for this, which would apply on top of Andre's ITS implementation. Kind regards, Pavel Fedin Expert Engineer Samsung Electronics Research center Russia ___ kvmarm mailing list kvmarm@lists.cs.columbia.edu https://lists.cs.columbia.edu/mailman/listinfo/kvmarm
RE: IRQFD support with GICv3 ITS (WAS: RE: [PATCH 00/13] arm64: KVM: GICv3 ITS emulation)
Hello! KVM GSI routing, even if only used for MSI routing then mandates to build entries for non MSI IRQs, using irqchip routing entries. Then you draw the irqchip.c kvm_irq_routing_table chip[KVM_NR_IRQCHIPS][KVM_IRQCHIP_NUM_PINS] static allocation issue. Sorry for this add-on, needed time to look at the code. Actually, if we don't use this code at all, and implement our own kvm_set_irq_routing() and kvm_free_irq_routing(), we don't have to bother about all these limitations. The simplest thing to do there would be to store GSI number in struct its_itte. In this case raising an MSI by GSI would not differ from what i currently do. Kind regards, Pavel Fedin Expert Engineer Samsung Electronics Research center Russia ___ kvmarm mailing list kvmarm@lists.cs.columbia.edu https://lists.cs.columbia.edu/mailman/listinfo/kvmarm
Re: [PATCH 00/13] arm64: KVM: GICv3 ITS emulation
On 08/06/15 11:54, Pavel Fedin wrote: Hi! I'm afraid this is not enough. A write to GICR_TRANSLATER (DID+EID) results in a (LPI,CPU) pair. Can you easily express the CPU part in irqfd (this is a genuine question, I'm not familiar enough with that part of the core)? But... As far as i could understand, LPI is added to a collection as a part of setup. And collection actually represents a destination CPU, doesn't it? And we can't have multiple LPIs sharing the same number and going to different CPUs. Or am i wrong? Unfortunately i don't have GICv3 arch reference manual. This is true to some extent. But the point is that the result of the translation is both an LPI and a CPU. My question was how you would indicate convey the notion of a target vcpu when using irqfd. As far as I know this doesn't really fit, unless we start introducing the dreaded GSI routing... Do we really want to go down that road? Another concern would be the support of GICv4, which relies on the command queue handling to be handled in the kernel Wow, i didn't know about GICv4. I wish I didn't know about it. M. -- Jazz is not dead. It just smells funny... ___ kvmarm mailing list kvmarm@lists.cs.columbia.edu https://lists.cs.columbia.edu/mailman/listinfo/kvmarm
RE: [PATCH 00/13] arm64: KVM: GICv3 ITS emulation
Hello everybody! The GICv3 ITS (Interrupt Translation Service) is a part of the ARM GICv3 interrupt controller used for implementing MSIs. It specifies a new kind of interrupts (LPIs), which are mapped to establish a connection between a device, its MSI payload value and the target processor the IRQ is eventually delivered to. In order to allow using MSIs in an ARM64 KVM guest, we emulate this ITS widget in the kernel. I have tested the patch and got some more ideas for future extension... First of all, it would be nice to have a possibility to directly inject LPIs by number. This will be useful for irqfd support in qemu. Next, irqfd support currently poses a problem. We need to somehow know IRQ number from MSI-X data (device ID plus event ID). ITS has all this information, so it would be nice to be able to query for the translation from within userspace. The question is - how to do it? Should we add some ioctl for this purpose? Currently i am experimenting with extra KVM_TRANSLATE_MSI ioctl which, given MSI data, would return LPI number. Actually before your patch came out i have almost done the same thing. But instead i decided to implement ITS in qemu while leaving LPI handling to kernel. In this case my qemu would have everything needed. By the way, why did you decide to put everything into kernel? Yes, in-kernel emulation is faster, but ITS is not accessed frequently. Kind regards, Pavel Fedin Expert Engineer Samsung Electronics Research center Russia ___ kvmarm mailing list kvmarm@lists.cs.columbia.edu https://lists.cs.columbia.edu/mailman/listinfo/kvmarm
RE: [PATCH 00/13] arm64: KVM: GICv3 ITS emulation
Hi! I'm afraid this is not enough. A write to GICR_TRANSLATER (DID+EID) results in a (LPI,CPU) pair. Can you easily express the CPU part in irqfd (this is a genuine question, I'm not familiar enough with that part of the core)? But... As far as i could understand, LPI is added to a collection as a part of setup. And collection actually represents a destination CPU, doesn't it? And we can't have multiple LPIs sharing the same number and going to different CPUs. Or am i wrong? Unfortunately i don't have GICv3 arch reference manual. Another concern would be the support of GICv4, which relies on the command queue handling to be handled in the kernel Wow, i didn't know about GICv4. Kind regards, Pavel Fedin Expert Engineer Samsung Electronics Research center Russia ___ kvmarm mailing list kvmarm@lists.cs.columbia.edu https://lists.cs.columbia.edu/mailman/listinfo/kvmarm