KVM platform device passthrough
v16 cover KVM platform device passthrough
Now that the platform-bus is instantiated in virt, the ARM sysbus-fdt infrastructure, the VFIO platform device and its first Calxeda xgmac specialization also are on the master, let's complete the work by
- enabling the instantiation of the VFIO Calxeda xgmac device
- enabling irqfd usage, which is vital for performance.
So the resend only relates to patches 4-9 of the v16 series.
http://git.linaro.org/people/eric.auger/qemu.git
(branch vfio_integ_v16_partial_resend)
Tested by Vikram Sethi on a SATA device featuring 2 MMIO regions. Code also exercised successfully by VOSYS for AMBA derivative series.
Thanks
Eric
v15 -> v16:
- no code change
- add Vikram's T-b
- rebase on 0915aed
- remove linux header update patch file
v14 -> v15:
- add Peter R-b on sysbus: add irq_routing_notifier
- correct g_malloc0_n usage in skeleton
- correct return values of reset related functions
- include Cornelia's patch for header update
v13 -> v14:
- remove v13 9, 10, 11 patch files and replace them by a single patch file "sysbus: add irq_routing_notifier".
- in skeleton, fix ENAMETOOLONG sign
- remove VFIOINTp virtualID in "add irq assignment" patch file
- removed trace_vfio_platform_start_eventfd
v12 -> v13:
- header update but same update was already sent by Cornelia
- Rework VFIO signaling & irqfd setup: restored 2-step setup featuring eventfd setup on realize and then irqfd setup on irq binding.
- irqfd setup now uses kvm_irqchip_add_irqfd_notifier and sysbus irq_set_hook override. This leads to the introduction of 6 patch files enabling those 2 features. Paolo advised to introduce kvm_irqchip_add_irqfd_notifier series in the VFIO one. I did the same for irq_set_hook series but if it is better I can submit it aside.
- above changes made possible to remove
x hw/vfio/platform: add capability to start IRQ propagation" x hw/arm/virt: start VFIO IRQ propagation
- in sysbus-fdt.c, use platform_bus_get_mmio_addr instead of deprecated mmio[0] property. Thanks to Bharat who pointed this issue out. also cpu_to_be32 was used for size and base (Vikram input) .
- in skeleton misc corrections following Alex
v11->v12:
- add x-mmap property definition, without which the default value of vbasedev.allow_mmap is false, hence preventing the reg space from being mmapped.
v10->v11:
- rebase onto v2.3.0-rc0 (mainly related to PCIe support in virt)
- add dma-coherent property for calxeda midway (fix revealed by removal of kernel-side "vfio: type1: support for ARM SMMUS with VFIO_IOMMU_TYPE1")
- virt modifications to start VFIO IRQ forwarding are now in a separate patch
- rearrange linux header exports (those are still partial exports waiting for definitive 4.1-rc0)
- take into account Alex Bennécomments:
- use g_malloc0_n instead of g_malloc0
- use block declarations when possible
- rework readlink returned value treatment
- use g_strlcat in place strncat
- re-arrange mutex locking for multiple IRQ support (user-side handled eventfds)
- use g_snprintf instead of snprintf
- change the order of functions to avoid pre-declaration in platform.c
- add flags in VFIOINTp struct to detect whether the IRQ is automasked
- some comment rewriting
v9->v10:
- rebase on "vfio: cleanup vfio_get_device error path, remove vfio_populate_device": vfio_populate_device no more called in vfio_get_device but in vfio_base_device_init
- update VFIO header according to vfio platform driver v13 (no AMBA)
v8->v9:
- rebase on 2.2.0 and machvirt dynamic sysbus instantiation v10
- v8 1-11 were pulled
- patch files related to forwarding are moved in a seperate series since it depends on kernel series still in RFC.
- introduction of basic VFIO platform device split into 3 patch files to ease the review (hope it will help).
- add an author in platform.c
- add deallocation in vfio_populate_device error case
- add patch file doing the VFIO header sync
- use VFIO_DEVICE_FLAGS_PLATFORM in vfio_populate_device
- rename calxeda_xgmac.c into calxeda-xgmac.c
- sysbus-fdt: add_calxeda_midway_xgmac_fdt_node g_free in case of errors
- reword of linux-headers patch files
v7->v8:
- rebase on v2.2.0-rc3 and integrate
"Add skip_dump flag to ignore memory region during dump"
- KVM header evolution with subindex addition in kvm_arch_forwarded_irq
- split [PATCH v7 03/16] hw/vfio/pci: introduce VFIODevice into 4 patches
- vfio_compute_needs_reset does not return bool anymore
- add some comments about exposed MMIO region and IRQ in calxeda xgmac device
- vfio_[un]mask_irqindex renamed into vfio_[un]mask_single_irqindex
- rework IRQ startup: former machine init done notifier is replaced by a reset notifier. machine file passes the interrupt controller DeviceState handle (not the platform bus first irq parameter).
- sysbus-fdt:
- move the add_fdt_node_functions array declaration between the device specific code and the generic code to avoid forward declarations of decice specific functions
- rename add_basic_vfio_fdt_node into add_calxeda_midway_xgmac_fdt_node emphasizing the fact it is xgmac specific
v6->v7:
- fake injection test modality removed
- VFIO_DEVICE_TYPE_PLATFORM only introduced with VFIO platform
- new helper functions to start VFIO IRQ on machine init done notifier (introduced in hw/vfio/platform: add vfio-platform support and notifier registration invoked in hw/arm/virt: add support for VFIO devices). vfio_start_irq_injection is replaced by vfio_register_irq_starter.
v5->v6:
- rebase on 2.1rc5 PCI code
- forwarded IRQ first integraton
- vfio_device property renamed into host property
- split IRQ setup in different functions that match the 3 supported injection techniques (user handled eventfd, irqfd, forwarded IRQ):
removes dynamic switch between injection methods
- introduce fake interrupts as a test modality:
x makes possible to test multiple IRQ user-side handling. x this is a test feature only: enable to trigger a fd as if the
real physical IRQ hit. No virtual IRQ is injected into the guest but handling is simulated so that the state machine can be tested
- user handled eventfd:
x add mutex to protect IRQ state & list manipulation,
x correct misleading comment in vfio_intp_interrupt.
x Fix bugs using fake interrupt modality
- irqfd no more advertised in this patchset (handled in [3])
- VFIOPlatformDeviceClass becomes abstract and Calxeda xgmac device and class is re-introduced (as per v4)
- all DPRINTF removed in platform and replaced by trace-points
- corrects compilation with configure --disable-kvm
- simplifies the split for vfio_get_device and introduce a unique specialized function named vfio_populate_device
- group_list renamed into vfio_group_list
- hw/arm/dyn_sysbus_devtree.c currently only support vfio-calxeda-xgmac instantiation. Needs to be specialized for other VFIO devices
- fix 2 bugs in dyn_sysbus_devtree(reg_attr index and compat)
v4->v5:
- rebase on v2.1.0 PCI code
- take into account Alex Williamson comments on PCI code rework
- trace updates in vfio_region_write/read
- remove fd from VFIORegion
- get/put ckeanup
- bug fix: bar region's vbasedev field duly initialization
- misc cleanups in platform device
- device tree node generation removed from device and handled in hw/arm/dyn_sysbus_devtree.c
- remove "hw/vfio: add an example calxeda_xgmac": with removal of device tree node generation we do not have so many things to implement in that derived device yet. May be re-introduced later on if needed typically for reset/migration.
- no GSI routing table anymore
v3->v4 changes (Eric Auger, Alvise Rigo)
- rebase on last VFIO PCI code (v2.1.0-rc0)
- full git history rework to ease PCI code change review
- mv include files in hw/vfio
- DPRINTF reformatting temporarily moved out
- support of VFIO virq (removal of resamplefd handler on user-side)
- integration with sysbus dynamic instantiation framwork
- removal of unrealize and cleanup routines until it is better understood what is really needed
- Support of VFIO for Amba devices should be handled in an inherited device to specialize the device tree generation (clock handle currently missing in framework however)
- "Always use eventfd as notifying mechanism" temporarily moved out
- static instantiation is not mainstream (although it remains possible) note if static instantiation is used, irqfd must be setup in machine file when virtual IRQ is known
- create the GSI routing table on qemu side
v2->v3 changes (Alvise Rigo, Eric Auger):
- Following Alex W recommandations, further efforts to factorize the code between PCI:introduction of VFIODevice and VFIORegion as base classes
- unique reset handler for platform and PCI
- cleanup following Kim's comments
- multiple IRQ support mechanics should be in place although not tested
- Better handling of MMIO multiple regions
- New features and fixes by Alvise (multiple compat string, exec flag, force eventfd usage, amba device tree support)
- irqfd support
v1->v2 changes (Kim Phillips, Eric Auger):
- IRQ initial support (legacy mode where eventfds are handled on user side)
- hacked dynamic instantiation
v1 (Kim Phillips):
- initial split between PCI and platform
- MMIO support only
- static instantiation
Best Regards
Eric
Eric Auger (6):
hw/arm/sysbus-fdt: enable vfio-calxeda-xgmac dynamic instantiation
kvm: rename kvm_irqchip_[add,remove]_irqfd_notifier with gsi suffix kvm-all.c: add qemu_irq/gsi hash table and utility routines
intc: arm_gic_kvm: set the qemu_irq/gsi mapping
sysbus: add irq_routing_notifier
hw/vfio/platform: add irqfd support
hw/arm/sysbus-fdt.c | 72 +++++++++++++++++++++++++++ hw/core/sysbus.c | 6 +++ hw/intc/arm_gic_kvm.c | 6 +++ hw/s390x/virtio-ccw.c | 8 +-- hw/vfio/pci.c | 6 +-- hw/vfio/platform.c | 107 ++++++++++++++++++++++++++++++++++++++++ hw/virtio/virtio-pci.c | 4 +- include/hw/sysbus.h | 1 + include/hw/vfio/vfio-platform.h | 2 + include/sysemu/kvm.h | 11 ++++- kvm-all.c | 44 +++++++++++++++-- kvm-stub.c | 7 +-- trace-events | 1 + 13 files changed, 258 insertions(+), 17 deletions(-)
Signature · 1 line
1.8.3.2
PATCH RESEND v16 1/6 hw/arm/sysbus-fdt: enable vfio-calxeda-xgmac dynamic instantiation
This patch allows the instantiation of the vfio-calxeda-xgmac device from the QEMU command line (-device vfio-calxeda-xgmac,host="<device>").
A specialized device tree node is created for the guest, containing compat, dma-coherent, reg and interrupts properties.
---
v12 -> v13:
- use platform_bus_get_mmio_addr instead of deprecated mmio[0] property. Thanks to Bharat who pointed this issue out.
- use cpu_to_be32 to mmio_base & size (Vikram report)
v10 -> v11:
- add dma-coherent property to calxeda midway xgmac node (fix)
- use qemu_fdt_setprop to add reg property instead of qemu_fdt_setprop_sized_cells_from_array
- commit message rewording
v8 -> v9:
- properly free resources in case of errors in
add_calxeda_midway_xgmac_fdt_node
v7 -> v8:
- move the add_fdt_node_functions array declaration between the device specific code and the generic code to avoid forward declarations of decice specific functions
- rename add_basic_vfio_fdt_node into add_calxeda_midway_xgmac_fdt_node
v6 -> v7:
- compat string re-formatting removed since compat string is not exposed anymore as a user option
- VFIO IRQ kick-off removed from sysbus-fdt and moved to VFIO platform device
--- hw/arm/sysbus-fdt.c | 72 +++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 72 insertions(+)
hw/arm/sysbus-fdt.c +72 -0
@@ -26,6 +26,8 @@#include "sysemu/device_tree.h" #include "hw/platform-bus.h" #include "sysemu/sysemu.h" +#include "hw/vfio/vfio-platform.h" +#include "hw/vfio/vfio-calxeda-xgmac.h" /* * internal struct that contains the information to create dynamic
@@ -53,11 +55,81 @@ typedef struct NodeCreationPair {int (*add_fdt_node_fn)(SysBusDevice *sbdev,void *opaque); } NodeCreationPair; +/* Device Specific Code */ + +/** +* add_calxeda_midway_xgmac_fdt_node +* +* Generates a simple node with following properties: +* compatible string, regs, interrupts, dma-coherent +*/ +static int add_calxeda_midway_xgmac_fdt_node (SysBusDevice *sbdev,void *opaque) +{ +PlatformBusFDTData *data = opaque; +PlatformBusDevice *pbus = data->pbus; +void *fdt = data->fdt; +const char *parent_node = data->pbus_node_name; +int compat_str_len, i, ret =-1 ; +char *nodename; +uint32_t *irq_attr, *reg_attr; +uint64_t mmio_base, irq_number; +VFIOPlatformDevice *vdev =VFIO_PLATFORM_DEVICE (sbdev); +VFIODevice *vbasedev = &vdev->vbasedev; + + mmio_base =platform_bus_get_mmio_addr (pbus, sbdev,0 ); + nodename =g_strdup_printf ("%s/%s@%" PRIx64, parent_node, + vbasedev->name, mmio_base); +qemu_fdt_add_subnode (fdt, nodename); + + compat_str_len =strlen (vdev->compat) +1 ; +qemu_fdt_setprop (fdt, nodename,"compatible" , + vdev->compat, compat_str_len); + +qemu_fdt_setprop (fdt, nodename,"dma-coherent" ,"" ,0 ); + + reg_attr =g_new (uint32_t, vbasedev->num_regions*2 ); +for (i =0 ; i < vbasedev->num_regions; i++) { + mmio_base =platform_bus_get_mmio_addr (pbus, sbdev, i); + reg_attr[2 *i] =cpu_to_be32 (mmio_base); + reg_attr[2 *i+1 ] =cpu_to_be32 ( +memory_region_size (&vdev->regions[i]->mem)); + } + ret =qemu_fdt_setprop (fdt, nodename,"reg" , reg_attr, + vbasedev->num_regions*2 *sizeof (uint32_t )); +if (ret) { +error_report ("could not set reg property of node %s" , nodename); + goto fail_reg; + } + + irq_attr =g_new (uint32_t, vbasedev->num_irqs*3 ); +for (i =0 ; i < vbasedev->num_irqs; i++) { + irq_number =platform_bus_get_irqn (pbus, sbdev , i) + + data->irq_start; + irq_attr[3 *i] =cpu_to_be32 (0 ); + irq_attr[3 *i+1 ] =cpu_to_be32 (irq_number); + irq_attr[3 *i+2 ] =cpu_to_be32 (0x4 ); + } + ret =qemu_fdt_setprop (fdt, nodename,"interrupts" , + irq_attr, vbasedev->num_irqs*3 *sizeof (uint32_t )); +if (ret) { +error_report ("could not set interrupts property of node %s" , + nodename); + } +g_free (irq_attr); +fail_reg: +g_free (reg_attr); +g_free (nodename); +return ret; +} +/* list of supported dynamic sysbus devices */ static const NodeCreationPair add_fdt_node_functions[] = { + {TYPE_VFIO_CALXEDA_XGMAC , add_calxeda_midway_xgmac_fdt_node}, {"" ,NULL },/* last element */ }; +/* Generic Code */ +/** * add_fdt_node - add the device tree node of a dynamic sysbus device *
Signature · 1 line
1.8.3.2
Re: [Qemu-devel] [RESEND PATCH v16 1/6] hw/arm/sysbus-fdt: enable vfio-calxeda-xgmac dynamic instantiation
Quoted 8 lines · Eric Auger
This patch allows the instantiation of the vfio-calxeda-xgmac device from the QEMU command line (-device vfio-calxeda-xgmac,host="<device>").
A specialized device tree node is created for the guest, containing compat, dma-coherent, reg and interrupts properties.
Signed-off-by Eric Auger <[email protected]>
Minor nits:
Quoted 23 lines
hw/arm/sysbus-fdt.c +23
-0+/* Device Specific Code */ + +/** +* add_calxeda_midway_xgmac_fdt_node +* +* Generates a simple node with following properties: +* compatible string, regs, interrupts, dma-coherent +*/ +static int add_calxeda_midway_xgmac_fdt_node (SysBusDevice *sbdev,void *opaque) +{ +PlatformBusFDTData *data = opaque; +PlatformBusDevice *pbus = data->pbus; +void *fdt = data->fdt; +const char *parent_node = data->pbus_node_name; +int compat_str_len, i, ret =-1 ; +char *nodename; +uint32_t *irq_attr, *reg_attr; +uint64_t mmio_base, irq_number; +VFIOPlatformDevice *vdev =VFIO_PLATFORM_DEVICE (sbdev); +VFIODevice *vbasedev = &vdev->vbasedev; + + mmio_base =platform_bus_get_mmio_addr (pbus, sbdev,0 ); + nodename = g_strdup_printf("%s/%s@%" PRIx64, parent_node,
hw/arm/sysbus-fdt.c typedef struct NodeCreationPair { · line 90
Quoted 10 lines
hw/arm/sysbus-fdt.c +10
-0+ vbasedev->name, mmio_base); +qemu_fdt_add_subnode (fdt, nodename); + + compat_str_len =strlen (vdev->compat) +1 ; +qemu_fdt_setprop (fdt, nodename,"compatible" , + vdev->compat, compat_str_len); + +qemu_fdt_setprop (fdt, nodename,"dma-coherent" ,"" ,0 ); + + reg_attr =g_new (uint32_t, vbasedev->num_regions*2 );
You should have spaces around the '*' operator here (and in a couple of other places below).
Quoted 10 lines
hw/arm/sysbus-fdt.c +10
-0+for (i =0 ; i < vbasedev->num_regions; i++) { + mmio_base =platform_bus_get_mmio_addr (pbus, sbdev, i); + reg_attr[2 *i] =cpu_to_be32 (mmio_base); + reg_attr[2 *i+1 ] =cpu_to_be32 ( +memory_region_size (&vdev->regions[i]->mem)); + } + ret =qemu_fdt_setprop (fdt, nodename,"reg" , reg_attr, + vbasedev->num_regions*2 *sizeof (uint32_t )); +if (ret) { +error_report ("could not set reg property of node %s" , nodename);
hw/arm/sysbus-fdt.c typedef struct NodeCreationPair { · line 110
Quoted 10 lines
hw/arm/sysbus-fdt.c +10
-0+ goto fail_reg; + } + + irq_attr =g_new (uint32_t, vbasedev->num_irqs*3 ); +for (i =0 ; i < vbasedev->num_irqs; i++) { + irq_number =platform_bus_get_irqn (pbus, sbdev , i) + + data->irq_start; + irq_attr[3 *i] =cpu_to_be32 (0 ); + irq_attr[3 *i+1 ] =cpu_to_be32 (irq_number); + irq_attr[3 *i+2 ] =cpu_to_be32 (0x4 );
In virt.c the magic numbers in cells 0 and 2 have #defines (GIC_FDT_IRQ_TYPE_SPI and GIC_FDT_IRQ_FLAGS_LEVEL_HI).
(1) maybe we should pull those out into a header so you can use them here
(2) is this device really using a level-triggered interrupt?
Quoted 9 lines
hw/arm/sysbus-fdt.c +9
-0+ } + ret =qemu_fdt_setprop (fdt, nodename,"interrupts" , + irq_attr, vbasedev->num_irqs*3 *sizeof (uint32_t )); +if (ret) { +error_report ("could not set interrupts property of node %s" , + nodename); + } +g_free (irq_attr); +fail_reg:
hw/arm/sysbus-fdt.c typedef struct NodeCreationPair { · line 129
Quoted 10 lines
hw/arm/sysbus-fdt.c +6
-0+g_free (reg_attr); +g_free (nodename); +return ret; +} +/* list of supported dynamic sysbus devices */ static const NodeCreationPair add_fdt_node_functions[] = { + {TYPE_VFIO_CALXEDA_XGMAC , add_calxeda_midway_xgmac_fdt_node}, {"" ,NULL },/* last element */ };
This file could get big if we have a lot of platform passthrough devices, but let's hope we don't...
Otherwise:
-- PMM
Re: [Qemu-devel] [RESEND PATCH v16 1/6] hw/arm/sysbus-fdt: enable vfio-calxeda-xgmac dynamic instantiation
Quoted 40 lines · Peter Maydell
Quoted 8 lines · Eric Auger
This patch allows the instantiation of the vfio-calxeda-xgmac device from the QEMU command line (-device vfio-calxeda-xgmac,host="<device>").
A specialized device tree node is created for the guest, containing compat, dma-coherent, reg and interrupts properties.
Signed-off-by Eric Auger <[email protected]>Minor nits:
Quoted 28 lines
hw/arm/sysbus-fdt.c +28
-0+/* Device Specific Code */ + +/** +* add_calxeda_midway_xgmac_fdt_node +* +* Generates a simple node with following properties: +* compatible string, regs, interrupts, dma-coherent +*/ +static int add_calxeda_midway_xgmac_fdt_node (SysBusDevice *sbdev,void *opaque) +{ +PlatformBusFDTData *data = opaque; +PlatformBusDevice *pbus = data->pbus; +void *fdt = data->fdt; +const char *parent_node = data->pbus_node_name; +int compat_str_len, i, ret =-1 ; +char *nodename; +uint32_t *irq_attr, *reg_attr; +uint64_t mmio_base, irq_number; +VFIOPlatformDevice *vdev =VFIO_PLATFORM_DEVICE (sbdev); +VFIODevice *vbasedev = &vdev->vbasedev; + + mmio_base =platform_bus_get_mmio_addr (pbus, sbdev,0 ); + nodename =g_strdup_printf ("%s/%s@%" PRIx64, parent_node, + vbasedev->name, mmio_base); +qemu_fdt_add_subnode (fdt, nodename); + + compat_str_len =strlen (vdev->compat) +1 ; + qemu_fdt_setprop(fdt, nodename, "compatible",
Quoted 8 lines
hw/arm/sysbus-fdt.c typedef struct NodeCreationPair { · line 90
Quoted 5 lines
hw/arm/sysbus-fdt.c +5
-0+ vdev->compat, compat_str_len); + +qemu_fdt_setprop (fdt, nodename,"dma-coherent" ,"" ,0 ); + + reg_attr =g_new (uint32_t, vbasedev->num_regions*2 );You should have spaces around the '*' operator here (and in a couple of other places below).
OK
Quoted 22 lines
Quoted 11 lines
hw/arm/sysbus-fdt.c +10
-0+for (i =0 ; i < vbasedev->num_regions; i++) { + mmio_base =platform_bus_get_mmio_addr (pbus, sbdev, i); + reg_attr[2 *i] =cpu_to_be32 (mmio_base); + reg_attr[2 *i+1 ] =cpu_to_be32 ( +memory_region_size (&vdev->regions[i]->mem)); + } + ret =qemu_fdt_setprop (fdt, nodename,"reg" , reg_attr, + vbasedev->num_regions*2 *sizeof (uint32_t )); +if (ret) { +error_report ("could not set reg property of node %s" , nodename);hw/arm/sysbus-fdt.c typedef struct NodeCreationPair { · line 110
Quoted 10 lines
hw/arm/sysbus-fdt.c +10
-0+ goto fail_reg; + } + + irq_attr =g_new (uint32_t, vbasedev->num_irqs*3 ); +for (i =0 ; i < vbasedev->num_irqs; i++) { + irq_number =platform_bus_get_irqn (pbus, sbdev , i) + + data->irq_start; + irq_attr[3 *i] =cpu_to_be32 (0 ); + irq_attr[3 *i+1 ] =cpu_to_be32 (irq_number); + irq_attr[3 *i+2 ] =cpu_to_be32 (0x4 );
Quoted 4 lines
In
virt.cthe magic numbers in cells 0 and 2 have #defines (GIC_FDT_IRQ_TYPE_SPIandGIC_FDT_IRQ_FLAGS_LEVEL_HI).(1) maybe we should pull those out into a header so you can use them here
OK maybe I can put them in newly created include/hw/arm/virt.h
Quoted 1 line
(2) is this device really using a level-triggered interrupt?
yes it does
Quoted 15 lines
Quoted 15 lines
hw/arm/sysbus-fdt.c +14
-0+ } + ret =qemu_fdt_setprop (fdt, nodename,"interrupts" , + irq_attr, vbasedev->num_irqs*3 *sizeof (uint32_t )); +if (ret) { +error_report ("could not set interrupts property of node %s" , + nodename); + } +g_free (irq_attr); +fail_reg: +g_free (reg_attr); +g_free (nodename); +return ret; +} +
Quoted 8 lines
Quoted 5 lines
/* list of supported dynamic sysbus devices */ static const NodeCreationPair add_fdt_node_functions[] = {
+ {TYPE_VFIO_CALXEDA_XGMAC,add_calxeda_midway_xgmac_fdt_node},{"", NULL}, /* last element */};
This file could get big if we have a lot of platform passthrough devices, but let's hope we don't...
Yes let's see how this feature gets used ;-)
Quoted 3 lines
Otherwise:
Acked-by Peter Maydell <[email protected]>
If you allow me I will resend the patch file separately with the above modifications.
Thanks!
Best Regards
Eric
Quoted 3 lines
-- PMM
Re: [Qemu-devel] [RESEND PATCH v16 1/6] hw/arm/sysbus-fdt: enable vfio-calxeda-xgmac dynamic instantiation
Quoted 6 lines · Eric Auger
Quoted 4 lines · Peter Maydell
In
virt.cthe magic numbers in cells 0 and 2 have #defines (GIC_FDT_IRQ_TYPE_SPIandGIC_FDT_IRQ_FLAGS_LEVEL_HI).(1) maybe we should pull those out into a header so you can use them here
OK maybe I can put them in newly created
include/hw/arm/virt.h
They're not really virt-board specific though, a new header would probably be better; maybe include/hw/arm/fdt.h ?
-- PMM
PATCH RESEND v16 2/6 kvm: rename kvm_irqchip_[add, remove]_irqfd_notifier with gsi suffix
In reply to Eric AugerAnticipating for the introduction of new add/remove functions taking a qemu_irq parameter, let's rename existing ones with a gsi suffix.
---
v15 -> v16:
- added Vikram's T-b
- resolve rebase conflict in kvm.h
--- hw/s390x/virtio-ccw.c | 8 ++++---- hw/vfio/pci.c | 6 +++--- hw/virtio/virtio-pci.c | 4 ++-- include/sysemu/kvm.h | 7 ++++--- kvm-all.c | 7 ++++--- kvm-stub.c | 7 ++++--- 6 files changed, 21 insertions(+), 18 deletions(-)
hw/s390x/virtio-ccw.c +4 -4
@@ -1186,8 +1186,8 @@ static int virtio_ccw_add_irqfd(VirtioCcwDevice *dev, int n)VirtQueue *vq =virtio_get_queue (vdev, n);EventNotifier *notifier =virtio_queue_get_guest_notifier (vq); -return kvm_irqchip_add_irqfd_notifier (kvm_state, notifier,NULL , - dev->routes.gsi[n]); +return kvm_irqchip_add_irqfd_notifier_gsi (kvm_state, notifier,NULL , + dev->routes.gsi[n]); }static void virtio_ccw_remove_irqfd (VirtioCcwDevice *dev,int n)
@@ -1197,8 +1197,8 @@ static void virtio_ccw_remove_irqfd(VirtioCcwDevice *dev, int n)EventNotifier *notifier =virtio_queue_get_guest_notifier (vq);int ret; - ret =kvm_irqchip_remove_irqfd_notifier (kvm_state, notifier, - dev->routes.gsi[n]); + ret =kvm_irqchip_remove_irqfd_notifier_gsi (kvm_state, notifier, + dev->routes.gsi[n]);assert (ret ==0 ); }
hw/vfio/pci.c +3 -3
@@ -597,7 +597,7 @@ static void vfio_add_kvm_msi_virq(VFIOMSIVector *vector, MSIMessage *msg,return ; } -if (kvm_irqchip_add_irqfd_notifier (kvm_state, &vector->kvm_interrupt, +if (kvm_irqchip_add_irqfd_notifier_gsi (kvm_state, &vector->kvm_interrupt,NULL , virq) <0 ) {kvm_irqchip_release_virq (kvm_state, virq);event_notifier_cleanup (&vector->kvm_interrupt);
@@ -609,8 +609,8 @@ static void vfio_add_kvm_msi_virq(VFIOMSIVector *vector, MSIMessage *msg,static void vfio_remove_kvm_msi_virq (VFIOMSIVector *vector) { -kvm_irqchip_remove_irqfd_notifier (kvm_state, &vector->kvm_interrupt, - vector->virq); +kvm_irqchip_remove_irqfd_notifier_gsi (kvm_state, &vector->kvm_interrupt, + vector->virq);kvm_irqchip_release_virq (kvm_state, vector->virq); vector->virq =-1 ;event_notifier_cleanup (&vector->kvm_interrupt);
hw/virtio/virtio-pci.c +2 -2
@@ -495,7 +495,7 @@ static int kvm_virtio_pci_irqfd_use(VirtIOPCIProxy *proxy,VirtQueue *vq =virtio_get_queue (vdev, queue_no);EventNotifier *n =virtio_queue_get_guest_notifier (vq);int ret; - ret =kvm_irqchip_add_irqfd_notifier (kvm_state, n,NULL , irqfd->virq); + ret =kvm_irqchip_add_irqfd_notifier_gsi (kvm_state, n,NULL , irqfd->virq);return ret; }
@@ -509,7 +509,7 @@ static void kvm_virtio_pci_irqfd_release(VirtIOPCIProxy *proxy,VirtIOIRQFD *irqfd = &proxy->vector_irqfd[vector];int ret; - ret =kvm_irqchip_remove_irqfd_notifier (kvm_state, n, irqfd->virq); + ret =kvm_irqchip_remove_irqfd_notifier_gsi (kvm_state, n, irqfd->virq);assert (ret ==0 ); }
include/sysemu/kvm.h +4 -3
@@ -416,9 +416,10 @@ void kvm_irqchip_release_virq(KVMState *s, int virq);int kvm_irqchip_add_adapter_route (KVMState *s,AdapterInfo *adapter); -int kvm_irqchip_add_irqfd_notifier (KVMState *s,EventNotifier *n, -EventNotifier *rn,int virq); -int kvm_irqchip_remove_irqfd_notifier (KVMState *s,EventNotifier *n,int virq); +int kvm_irqchip_add_irqfd_notifier_gsi (KVMState *s,EventNotifier *n, +EventNotifier *rn,int virq); +int kvm_irqchip_remove_irqfd_notifier_gsi (KVMState *s,EventNotifier *n, +int virq);void kvm_pc_gsi_handler (void *opaque,int n,int level);void kvm_pc_setup_irq_routing (bool pci_enabled);void kvm_init_irq_routing (KVMState *s);
kvm-all.c +4 -3
@@ -1325,14 +1325,15 @@ int kvm_irqchip_update_msi_route(KVMState *s, int virq, MSIMessage msg) }#endif /* !KVM_CAP_IRQ_ROUTING */ -int kvm_irqchip_add_irqfd_notifier (KVMState *s,EventNotifier *n, -EventNotifier *rn,int virq) +int kvm_irqchip_add_irqfd_notifier_gsi (KVMState *s,EventNotifier *n, +EventNotifier *rn,int virq) {return kvm_irqchip_assign_irqfd (s,event_notifier_get_fd (n), rn ?event_notifier_get_fd (rn) :-1 , virq, true); } -int kvm_irqchip_remove_irqfd_notifier (KVMState *s,EventNotifier *n,int virq) +int kvm_irqchip_remove_irqfd_notifier_gsi (KVMState *s,EventNotifier *n, +int virq) {return kvm_irqchip_assign_irqfd (s,event_notifier_get_fd (n),-1 , virq, false);
kvm-stub.c +4 -3
@@ -137,13 +137,14 @@ int kvm_irqchip_add_adapter_route(KVMState *s, AdapterInfo *adapter)return -ENOSYS ; } -int kvm_irqchip_add_irqfd_notifier (KVMState *s,EventNotifier *n, -EventNotifier *rn,int virq) +int kvm_irqchip_add_irqfd_notifier_gsi (KVMState *s,EventNotifier *n, +EventNotifier *rn,int virq) {return -ENOSYS ; } -int kvm_irqchip_remove_irqfd_notifier (KVMState *s,EventNotifier *n,int virq) +int kvm_irqchip_remove_irqfd_notifier_gsi (KVMState *s,EventNotifier *n, +int virq) {return -ENOSYS ; }
Signature · 1 line
1.8.3.2
Re: [Qemu-devel] [RESEND PATCH v16 2/6] kvm: rename kvm_irqchip_[add, remove]_irqfd_notifier with gsi suffix
Quoted 5 lines · Eric Auger
Anticipating for the introduction of new add/remove functions taking a
qemu_irqparameter, let's rename existing ones with a gsi suffix.Signed-off-by Eric Auger <[email protected]>Tested-by Vikram Sethi <[email protected]>
Quoted 150 lines
---
v15 -> v16:
- added Vikram's T-b
- resolve rebase conflict inkvm.h--- hw/s390x/virtio-ccw.c | 8 ++++---- hw/vfio/pci.c | 6 +++--- hw/virtio/virtio-pci.c | 4 ++-- include/sysemu/kvm.h | 7 ++++--- kvm-all.c | 7 ++++--- kvm-stub.c | 7 ++++--- 6 files changed, 21 insertions(+), 18 deletions(-)hw/s390x/virtio-ccw.c +4
-4@@ -1186,8 +1186,8 @@ static int virtio_ccw_add_irqfd(VirtioCcwDevice *dev, int n)VirtQueue *vq =virtio_get_queue (vdev, n);EventNotifier *notifier =virtio_queue_get_guest_notifier (vq); -return kvm_irqchip_add_irqfd_notifier (kvm_state, notifier,NULL , - dev->routes.gsi[n]); +return kvm_irqchip_add_irqfd_notifier_gsi (kvm_state, notifier,NULL , + dev->routes.gsi[n]); }static void virtio_ccw_remove_irqfd (VirtioCcwDevice *dev,int n)@@ -1197,8 +1197,8 @@ static void virtio_ccw_remove_irqfd(VirtioCcwDevice *dev, int n)EventNotifier *notifier =virtio_queue_get_guest_notifier (vq);int ret; - ret =kvm_irqchip_remove_irqfd_notifier (kvm_state, notifier, - dev->routes.gsi[n]); + ret =kvm_irqchip_remove_irqfd_notifier_gsi (kvm_state, notifier, + dev->routes.gsi[n]);assert (ret ==0 ); }hw/vfio/pci.c +3
-3@@ -597,7 +597,7 @@ static void vfio_add_kvm_msi_virq(VFIOMSIVector *vector, MSIMessage *msg,return ; } -if (kvm_irqchip_add_irqfd_notifier (kvm_state, &vector->kvm_interrupt, +if (kvm_irqchip_add_irqfd_notifier_gsi (kvm_state, &vector->kvm_interrupt,NULL , virq) <0 ) {kvm_irqchip_release_virq (kvm_state, virq);event_notifier_cleanup (&vector->kvm_interrupt);@@ -609,8 +609,8 @@ static void vfio_add_kvm_msi_virq(VFIOMSIVector *vector, MSIMessage *msg,static void vfio_remove_kvm_msi_virq (VFIOMSIVector *vector) { -kvm_irqchip_remove_irqfd_notifier (kvm_state, &vector->kvm_interrupt, - vector->virq); +kvm_irqchip_remove_irqfd_notifier_gsi (kvm_state, &vector->kvm_interrupt, + vector->virq);kvm_irqchip_release_virq (kvm_state, vector->virq); vector->virq =-1 ;event_notifier_cleanup (&vector->kvm_interrupt);hw/virtio/virtio-pci.c +2
-2@@ -495,7 +495,7 @@ static int kvm_virtio_pci_irqfd_use(VirtIOPCIProxy *proxy,VirtQueue *vq =virtio_get_queue (vdev, queue_no);EventNotifier *n =virtio_queue_get_guest_notifier (vq);int ret; - ret =kvm_irqchip_add_irqfd_notifier (kvm_state, n,NULL , irqfd->virq); + ret =kvm_irqchip_add_irqfd_notifier_gsi (kvm_state, n,NULL , irqfd->virq);return ret; }@@ -509,7 +509,7 @@ static void kvm_virtio_pci_irqfd_release(VirtIOPCIProxy *proxy,VirtIOIRQFD *irqfd = &proxy->vector_irqfd[vector];int ret; - ret =kvm_irqchip_remove_irqfd_notifier (kvm_state, n, irqfd->virq); + ret =kvm_irqchip_remove_irqfd_notifier_gsi (kvm_state, n, irqfd->virq);assert (ret ==0 ); }include/sysemu/kvm.h +4
-3@@ -416,9 +416,10 @@ void kvm_irqchip_release_virq(KVMState *s, int virq);int kvm_irqchip_add_adapter_route (KVMState *s,AdapterInfo *adapter); -int kvm_irqchip_add_irqfd_notifier (KVMState *s,EventNotifier *n, -EventNotifier *rn,int virq); -int kvm_irqchip_remove_irqfd_notifier (KVMState *s,EventNotifier *n,int virq); +int kvm_irqchip_add_irqfd_notifier_gsi (KVMState *s,EventNotifier *n, +EventNotifier *rn,int virq); +int kvm_irqchip_remove_irqfd_notifier_gsi (KVMState *s,EventNotifier *n, +int virq);void kvm_pc_gsi_handler (void *opaque,int n,int level);void kvm_pc_setup_irq_routing (bool pci_enabled);void kvm_init_irq_routing (KVMState *s);kvm-all.c +4
-3@@ -1325,14 +1325,15 @@ int kvm_irqchip_update_msi_route(KVMState *s, int virq, MSIMessage msg) }#endif /* !KVM_CAP_IRQ_ROUTING */ -int kvm_irqchip_add_irqfd_notifier (KVMState *s,EventNotifier *n, -EventNotifier *rn,int virq) +int kvm_irqchip_add_irqfd_notifier_gsi (KVMState *s,EventNotifier *n, +EventNotifier *rn,int virq) {return kvm_irqchip_assign_irqfd (s,event_notifier_get_fd (n), rn ?event_notifier_get_fd (rn) :-1 , virq, true); } -int kvm_irqchip_remove_irqfd_notifier (KVMState *s,EventNotifier *n,int virq) +int kvm_irqchip_remove_irqfd_notifier_gsi (KVMState *s,EventNotifier *n, +int virq) {return kvm_irqchip_assign_irqfd (s,event_notifier_get_fd (n),-1 , virq, false);kvm-stub.c +4
-3@@ -137,13 +137,14 @@ int kvm_irqchip_add_adapter_route(KVMState *s, AdapterInfo *adapter)return -ENOSYS ; } -int kvm_irqchip_add_irqfd_notifier (KVMState *s,EventNotifier *n, -EventNotifier *rn,int virq) +int kvm_irqchip_add_irqfd_notifier_gsi (KVMState *s,EventNotifier *n, +EventNotifier *rn,int virq) {return -ENOSYS ; } -int kvm_irqchip_remove_irqfd_notifier (KVMState *s,EventNotifier *n,int virq) +int kvm_irqchip_remove_irqfd_notifier_gsi (KVMState *s,EventNotifier *n, +int virq) {return -ENOSYS ; }
PATCH RESEND v16 3/6 kvm-all.c: add qemu_irq/gsi hash table and utility routines
In reply to Eric AugerVFIO platform device needs to setup irqfd but it does not know the gsi corresponding to the device qemu_irq. This series proposes to store a hash table in kvm_state using the qemu_irq as key and the gsi as a value.
kvm_irqchip_set_qemuirq_gsi allows to insert such a pair. The interrupt controller is supposed to use it.
kvm_irqchip_[add, remove]_irqfd_notifier allows to setup/tear down irqfd directly from the qemu_irq.
---
v15 -> v16:
- Added Vikram's T-b
v13 -> v14:
- correct checkpatch warning
v2 -> v3 (integration into VFIO series v13):
- rename kvm_irqchip_[add, remove]_qemuirq_irqfd_notifier into kvm_irqchip_[add, remove]_irqfd_notifier. Possible since legacy functions were also renamed with _gsi suffix.
V1 -> v2:
- qemu_irq get_gsi callback replaced by hash table stored in kvm
--- include/sysemu/kvm.h | 6 ++++++ kvm-all.c | 37 +++++++++++++++++++++++++++++++++++++ 2 files changed, 43 insertions(+)
include/sysemu/kvm.h +6 -0
@@ -19,6 +19,7 @@#include "qemu/queue.h" #include "qom/cpu.h" #include "exec/memattrs.h" +#include "hw/irq.h" #ifdef CONFIG_KVM #include <linux/kvm.h>
@@ -420,6 +421,11 @@ int kvm_irqchip_add_irqfd_notifier_gsi(KVMState *s, EventNotifier *n, EventNotifier *rn, int virq);int kvm_irqchip_remove_irqfd_notifier_gsi (KVMState *s,EventNotifier *n,int virq); +int kvm_irqchip_add_irqfd_notifier (KVMState *s,EventNotifier *n, +EventNotifier *rn,qemu_irq irq); +int kvm_irqchip_remove_irqfd_notifier (KVMState *s,EventNotifier *n, +qemu_irq irq); +void kvm_irqchip_set_qemuirq_gsi (KVMState *s,qemu_irq irq,int gsi);void kvm_pc_gsi_handler (void *opaque,int n,int level);void kvm_pc_setup_irq_routing (bool pci_enabled);void kvm_init_irq_routing (KVMState *s);
kvm-all.c +37 -0
@@ -37,6 +37,7 @@#include "exec/address-spaces.h" #include "qemu/event_notifier.h" #include "trace.h" +#include "hw/irq.h" #include "hw/boards.h"
@@ -98,6 +99,7 @@ struct KVMState* unsigned, and treating them as signed here can break things */ unsigned irq_set_ioctl;unsigned int sigmask_len; +GHashTable *gsimap;#ifdef KVM_CAP_IRQ_ROUTING struct kvm_irq_routing *irq_routes;int nr_allocated_irq_routes;
@@ -1339,6 +1341,39 @@ int kvm_irqchip_remove_irqfd_notifier_gsi(KVMState *s, EventNotifier *n, false); } +int kvm_irqchip_add_irqfd_notifier (KVMState *s,EventNotifier *n, +EventNotifier *rn,qemu_irq irq) +{ +gpointer key, gsi; +gboolean found =g_hash_table_lookup_extended (s->gsimap, irq, &key, &gsi); + +if (!found) { +return -ENXIO ; + }else { +return kvm_irqchip_add_irqfd_notifier_gsi (s, n, rn, +GPOINTER_TO_UINT (gsi)); + } +} + +int kvm_irqchip_remove_irqfd_notifier (KVMState *s,EventNotifier *n, +qemu_irq irq) +{ +gpointer key, gsi; +gboolean found =g_hash_table_lookup_extended (s->gsimap, irq, &key, &gsi); + +if (!found) { +return -ENXIO ; + }else { +return kvm_irqchip_remove_irqfd_notifier_gsi (s, n, +GPOINTER_TO_INT (gsi)); + } +} + +void kvm_irqchip_set_qemuirq_gsi (KVMState *s,qemu_irq irq,int gsi) +{ +g_hash_table_insert (s->gsimap, irq,GINT_TO_POINTER (gsi)); +} +static int kvm_irqchip_create (MachineState *machine,KVMState *s) {int ret;
@@ -1371,6 +1406,8 @@ static int kvm_irqchip_create(MachineState *machine, KVMState *s)kvm_init_irq_routing (s); + s->gsimap =g_hash_table_new (g_direct_hash, g_direct_equal); +return 0 ; }
Signature · 1 line
1.8.3.2
Re: [Qemu-devel] [RESEND PATCH v16 3/6] kvm-all.c: add qemu_irq/gsi hash table and utility routines
Quoted 5 lines · Eric Auger
VFIO platform device needs to setup irqfd but it does not know the gsi corresponding to the device
qemu_irq. This series proposes to store a hash table inkvm_stateusing theqemu_irqas key and the gsi as a value.
Quoted 8 lines
kvm_irqchip_set_qemuirq_gsiallows to insert such a pair. The interrupt controller is supposed to use it.kvm_irqchip_[add, remove]_irqfd_notifier allows to setup/tear down irqfd directly from the
qemu_irq.Signed-off-by Eric Auger <[email protected]>Tested-by Vikram Sethi <[email protected]>
Quoted 114 lines
---
v15 -> v16:
- Added Vikram's T-bv13 -> v14:
- correct checkpatch warningv2 -> v3 (integration into VFIO series v13):
- rename kvm_irqchip_[add, remove]_qemuirq_irqfd_notifier into kvm_irqchip_[add, remove]_irqfd_notifier. Possible since legacy functions were also renamed with_gsisuffix.V1 -> v2:
-qemu_irqget_gsicallback replaced by hash table stored in kvm--- include/sysemu/kvm.h | 6 ++++++ kvm-all.c | 37 +++++++++++++++++++++++++++++++++++++ 2 files changed, 43 insertions(+)include/sysemu/kvm.h +6
-0@@ -19,6 +19,7 @@#include "qemu/queue.h" #include "qom/cpu.h" #include "exec/memattrs.h" +#include "hw/irq.h" #ifdef CONFIG_KVM #include <linux/kvm.h> @@ -420,6 +421,11 @@ int kvm_irqchip_add_irqfd_notifier_gsi(KVMState *s, EventNotifier *n, EventNotifier *rn, int virq);int kvm_irqchip_remove_irqfd_notifier_gsi (KVMState *s,EventNotifier *n,int virq); +int kvm_irqchip_add_irqfd_notifier (KVMState *s,EventNotifier *n, +EventNotifier *rn,qemu_irq irq); +int kvm_irqchip_remove_irqfd_notifier (KVMState *s,EventNotifier *n, +qemu_irq irq); +void kvm_irqchip_set_qemuirq_gsi (KVMState *s,qemu_irq irq,int gsi);void kvm_pc_gsi_handler (void *opaque,int n,int level);void kvm_pc_setup_irq_routing (bool pci_enabled);void kvm_init_irq_routing (KVMState *s);kvm-all.c +37
-0@@ -37,6 +37,7 @@#include "exec/address-spaces.h" #include "qemu/event_notifier.h" #include "trace.h" +#include "hw/irq.h" #include "hw/boards.h" @@ -98,6 +99,7 @@ struct KVMState* unsigned, and treating them as signed here can break things */ unsigned irq_set_ioctl;unsigned int sigmask_len; +GHashTable *gsimap;#ifdef KVM_CAP_IRQ_ROUTING struct kvm_irq_routing *irq_routes;int nr_allocated_irq_routes;@@ -1339,6 +1341,39 @@ int kvm_irqchip_remove_irqfd_notifier_gsi(KVMState *s, EventNotifier *n, false); } +int kvm_irqchip_add_irqfd_notifier (KVMState *s,EventNotifier *n, +EventNotifier *rn,qemu_irq irq) +{ +gpointer key, gsi; +gboolean found =g_hash_table_lookup_extended (s->gsimap, irq, &key, &gsi); + +if (!found) { +return -ENXIO ; + }else { +return kvm_irqchip_add_irqfd_notifier_gsi (s, n, rn, +GPOINTER_TO_UINT (gsi)); + } +} + +int kvm_irqchip_remove_irqfd_notifier (KVMState *s,EventNotifier *n, +qemu_irq irq) +{ +gpointer key, gsi; +gboolean found =g_hash_table_lookup_extended (s->gsimap, irq, &key, &gsi); + +if (!found) { +return -ENXIO ; + }else { +return kvm_irqchip_remove_irqfd_notifier_gsi (s, n, +GPOINTER_TO_INT (gsi)); + } +} + +void kvm_irqchip_set_qemuirq_gsi (KVMState *s,qemu_irq irq,int gsi) +{ +g_hash_table_insert (s->gsimap, irq,GINT_TO_POINTER (gsi)); +} +static int kvm_irqchip_create (MachineState *machine,KVMState *s) {int ret;@@ -1371,6 +1406,8 @@ static int kvm_irqchip_create(MachineState *machine, KVMState *s)kvm_init_irq_routing (s); + s->gsimap =g_hash_table_new (g_direct_hash, g_direct_equal); +return 0 ; }
Re: [Qemu-devel] [RESEND PATCH v16 3/6] kvm-all.c: add qemu_irq/gsi hash table and utility routines
In reply to Eric AugerQuoted 83 lines · Eric Auger
VFIO platform device needs to setup irqfd but it does not know the gsi corresponding to the device
qemu_irq. This series proposes to store a hash table inkvm_stateusing theqemu_irqas key and the gsi as a value.
kvm_irqchip_set_qemuirq_gsiallows to insert such a pair. The interrupt controller is supposed to use it.kvm_irqchip_[add, remove]_irqfd_notifier allows to setup/tear down irqfd directly from the
qemu_irq.Signed-off-by Eric Auger <[email protected]>Tested-by Vikram Sethi <[email protected]>---
v15 -> v16:
- Added Vikram's T-bv13 -> v14:
- correct checkpatch warningv2 -> v3 (integration into VFIO series v13):
- rename kvm_irqchip_[add, remove]_qemuirq_irqfd_notifier into kvm_irqchip_[add, remove]_irqfd_notifier. Possible since legacy functions were also renamed with_gsisuffix.V1 -> v2:
-qemu_irqget_gsicallback replaced by hash table stored in kvm--- include/sysemu/kvm.h | 6 ++++++ kvm-all.c | 37 +++++++++++++++++++++++++++++++++++++ 2 files changed, 43 insertions(+)include/sysemu/kvm.h +6
-0@@ -19,6 +19,7 @@#include "qemu/queue.h" #include "qom/cpu.h" #include "exec/memattrs.h" +#include "hw/irq.h" #ifdef CONFIG_KVM #include <linux/kvm.h> @@ -420,6 +421,11 @@ int kvm_irqchip_add_irqfd_notifier_gsi(KVMState *s, EventNotifier *n, EventNotifier *rn, int virq);int kvm_irqchip_remove_irqfd_notifier_gsi (KVMState *s,EventNotifier *n,int virq); +int kvm_irqchip_add_irqfd_notifier (KVMState *s,EventNotifier *n, +EventNotifier *rn,qemu_irq irq); +int kvm_irqchip_remove_irqfd_notifier (KVMState *s,EventNotifier *n, +qemu_irq irq); +void kvm_irqchip_set_qemuirq_gsi (KVMState *s,qemu_irq irq,int gsi);void kvm_pc_gsi_handler (void *opaque,int n,int level);void kvm_pc_setup_irq_routing (bool pci_enabled);void kvm_init_irq_routing (KVMState *s);kvm-all.c +3
-0@@ -37,6 +37,7 @@#include "exec/address-spaces.h" #include "qemu/event_notifier.h" #include "trace.h" +#include "hw/irq.h" #include "hw/boards.h" @@ -98,6 +99,7 @@ struct KVMState* unsigned, and treating them as signed here can break things */ unsigned irq_set_ioctl;unsigned int sigmask_len; +GHashTable *gsimap;#ifdef KVM_CAP_IRQ_ROUTING struct kvm_irq_routing *irq_routes;int nr_allocated_irq_routes;@@ -1339,6 +1341,39 @@ int kvm_irqchip_remove_irqfd_notifier_gsi(KVMState *s, EventNotifier *n, false); } +int kvm_irqchip_add_irqfd_notifier(KVMState *s, EventNotifier *n,
kvm-all.c int kvm_irqchip_remove_irqfd_notifier_gsi(KVMState *s, EventNotifier *n, · line 1354
Quoted 10 lines
kvm-all.c +10
-0+ EventNotifier *rn, qemu_irq irq) +{ +gpointer key, gsi; +gboolean found =g_hash_table_lookup_extended (s->gsimap, irq, &key, &gsi); + +if (!found) { +return -ENXIO ; + }else { +return kvm_irqchip_add_irqfd_notifier_gsi (s, n, rn, +GPOINTER_TO_UINT (gsi));
Why do we use GPOINTER_TO_UINT() here...
Quoted 4 lines
kvm-all.c +4
-0+ } +} + +int kvm_irqchip_remove_irqfd_notifier(KVMState *s, EventNotifier *n,
kvm-all.c int kvm_irqchip_remove_irqfd_notifier_gsi(KVMState *s, EventNotifier *n, · line 1368
Quoted 10 lines
kvm-all.c +10
-0+ qemu_irq irq) +{ +gpointer key, gsi; +gboolean found =g_hash_table_lookup_extended (s->gsimap, irq, &key, &gsi); + +if (!found) { +return -ENXIO ; + }else { +return kvm_irqchip_remove_irqfd_notifier_gsi (s, n, +GPOINTER_TO_INT (gsi));
...but GPOINTER_TO_INT() here?
(Aside: you don't need to put this code in an else {} clause, because the if() part is returning anyway.)
Quoted 12 lines
kvm-all.c +8
-0+ } +} + +void kvm_irqchip_set_qemuirq_gsi (KVMState *s,qemu_irq irq,int gsi) +{ +g_hash_table_insert (s->gsimap, irq,GINT_TO_POINTER (gsi)); +} +static int kvm_irqchip_create (MachineState *machine,KVMState *s) {int ret;
kvm-all.c static int kvm_irqchip_create(MachineState *machine, KVMState *s) · line 1412
Quoted 7 lines
kvm-all.c +2
-0@@ -1371,6 +1406,8 @@ static int kvm_irqchip_create(MachineState *machine, KVMState *s)kvm_init_irq_routing (s); + s->gsimap =g_hash_table_new (g_direct_hash, g_direct_equal); +return 0 ; }
thanks
-- PMM
Re: [Qemu-devel] [RESEND PATCH v16 3/6] kvm-all.c: add qemu_irq/gsi hash table and utility routines
Hi Peter,
Quoted 83 lines · Peter Maydell
Quoted 82 lines · Eric Auger
VFIO platform device needs to setup irqfd but it does not know the gsi corresponding to the device
qemu_irq. This series proposes to store a hash table inkvm_stateusing theqemu_irqas key and the gsi as a value.
kvm_irqchip_set_qemuirq_gsiallows to insert such a pair. The interrupt controller is supposed to use it.kvm_irqchip_[add, remove]_irqfd_notifier allows to setup/tear down irqfd directly from the
qemu_irq.Signed-off-by Eric Auger <[email protected]>Tested-by Vikram Sethi <[email protected]>---
v15 -> v16:
- Added Vikram's T-bv13 -> v14:
- correct checkpatch warningv2 -> v3 (integration into VFIO series v13):
- rename kvm_irqchip_[add, remove]_qemuirq_irqfd_notifier into kvm_irqchip_[add, remove]_irqfd_notifier. Possible since legacy functions were also renamed with_gsisuffix.V1 -> v2:
-qemu_irqget_gsicallback replaced by hash table stored in kvm--- include/sysemu/kvm.h | 6 ++++++ kvm-all.c | 37 +++++++++++++++++++++++++++++++++++++ 2 files changed, 43 insertions(+)include/sysemu/kvm.h +6
-0@@ -19,6 +19,7 @@#include "qemu/queue.h" #include "qom/cpu.h" #include "exec/memattrs.h" +#include "hw/irq.h" #ifdef CONFIG_KVM #include <linux/kvm.h> @@ -420,6 +421,11 @@ int kvm_irqchip_add_irqfd_notifier_gsi(KVMState *s, EventNotifier *n, EventNotifier *rn, int virq);int kvm_irqchip_remove_irqfd_notifier_gsi (KVMState *s,EventNotifier *n,int virq); +int kvm_irqchip_add_irqfd_notifier (KVMState *s,EventNotifier *n, +EventNotifier *rn,qemu_irq irq); +int kvm_irqchip_remove_irqfd_notifier (KVMState *s,EventNotifier *n, +qemu_irq irq); +void kvm_irqchip_set_qemuirq_gsi (KVMState *s,qemu_irq irq,int gsi);void kvm_pc_gsi_handler (void *opaque,int n,int level);void kvm_pc_setup_irq_routing (bool pci_enabled);void kvm_init_irq_routing (KVMState *s);kvm-all.c +2
-0@@ -37,6 +37,7 @@#include "exec/address-spaces.h" #include "qemu/event_notifier.h" #include "trace.h" +#include "hw/irq.h" #include "hw/boards.h" @@ -98,6 +99,7 @@ struct KVMState* unsigned, and treating them as signed here can break things */ unsigned irq_set_ioctl;unsigned int sigmask_len; +GHashTable *gsimap;#ifdef KVM_CAP_IRQ_ROUTING struct kvm_irq_routing *irq_routes;int nr_allocated_irq_routes;@@ -1339,6 +1341,39 @@ int kvm_irqchip_remove_irqfd_notifier_gsi(KVMState *s, EventNotifier *n, false); }
Quoted 13 lines
kvm-all.c int kvm_irqchip_remove_irqfd_notifier_gsi(KVMState *s, EventNotifier *n, · line 1354
Quoted 11 lines
kvm-all.c +11
-0+int kvm_irqchip_add_irqfd_notifier (KVMState *s,EventNotifier *n, +EventNotifier *rn,qemu_irq irq) +{ +gpointer key, gsi; +gboolean found =g_hash_table_lookup_extended (s->gsimap, irq, &key, &gsi); + +if (!found) { +return -ENXIO ; + }else { +return kvm_irqchip_add_irqfd_notifier_gsi (s, n, rn, +GPOINTER_TO_UINT (gsi));Why do we use
GPOINTER_TO_UINT()here...
Hum yes GPOINTER_TO_INT is in accordance withkvm_irqchip_add_irqfd_notifier_gsi signature
Quoted 20 lines
Quoted 6 lines
kvm-all.c +4
-0+ } +} + +int kvm_irqchip_remove_irqfd_notifier(KVMState *s, EventNotifier *n,kvm-all.c int kvm_irqchip_remove_irqfd_notifier_gsi(KVMState *s, EventNotifier *n, · line 1368
Quoted 10 lines
kvm-all.c +10
-0+ qemu_irq irq) +{ +gpointer key, gsi; +gboolean found =g_hash_table_lookup_extended (s->gsimap, irq, &key, &gsi); + +if (!found) { +return -ENXIO ; + }else { +return kvm_irqchip_remove_irqfd_notifier_gsi (s, n, +GPOINTER_TO_INT (gsi));...but
GPOINTER_TO_INT()here?
Quoted 2 lines
(Aside: you don't need to put this code in an else {} clause, because the
if()part is returning anyway.)
Indeed.
Do you wish I re-submit all the VFIO remaining patches as v17 or only that one?
Thanks
Eric
Quoted 24 lines
Quoted 13 lines
kvm-all.c +8
-0+ } +} + +void kvm_irqchip_set_qemuirq_gsi (KVMState *s,qemu_irq irq,int gsi) +{ +g_hash_table_insert (s->gsimap, irq,GINT_TO_POINTER (gsi)); +} +static int kvm_irqchip_create (MachineState *machine,KVMState *s) {int ret;kvm-all.c static int kvm_irqchip_create(MachineState *machine, KVMState *s) · line 1412
Quoted 7 lines
kvm-all.c +2
-0@@ -1371,6 +1406,8 @@ static int kvm_irqchip_create(MachineState *machine, KVMState *s)kvm_init_irq_routing (s); + s->gsimap =g_hash_table_new (g_direct_hash, g_direct_equal); +return 0 ; }thanks
-- PMM
PATCH RESEND v16 4/6 intc: arm_gic_kvm: set the qemu_irq/gsi mapping
In reply to Eric AugerThe arm_gic_kvm now calls kvm_irqchip_set_qemuirq_gsi to build the hash table storing qemu_irq/gsi mappings. From that point on irqfd can be setup directly from the qemu_irq using kvm_irqchip_add_irqfd_notifier.
---
v15 -> v16:
- added Vikram's T-b
- Resolve rebase conflict
v2 -> v3:
- kvm_irqchip_add_qemuirq_irqfd_notifier renamed into kvm_irqchip_add_irqfd_notifier
--- hw/intc/arm_gic_kvm.c | 6 ++++++ 1 file changed, 6 insertions(+)
hw/intc/arm_gic_kvm.c +6 -0
@@ -570,6 +570,12 @@ static void kvm_arm_gic_realize(DeviceState *dev, Error **errp)*/ i += (GIC_INTERNAL * s->num_cpu);qdev_init_gpio_in (dev, kvm_arm_gic_set_irq, i); + +for (i =0 ; i < s->num_irq -GIC_INTERNAL ; i++) { +qemu_irq irq =qdev_get_gpio_in (dev, i); +kvm_irqchip_set_qemuirq_gsi (kvm_state, irq, i); + } +/* We never use our outbound IRQ/FIQ lines but provide them so that * we maintain the same interface as the non-KVM GIC. */
Signature · 1 line
1.8.3.2
Re: [Qemu-devel] [RESEND PATCH v16 4/6] intc: arm_gic_kvm: set the qemu_irq/gsi mapping
Quoted 7 lines · Eric Auger
The
arm_gic_kvmnow callskvm_irqchip_set_qemuirq_gsito build the hash table storing qemu_irq/gsi mappings. From that point on irqfd can be setup directly from theqemu_irqusingkvm_irqchip_add_irqfd_notifier.Signed-off-by Eric Auger <[email protected]>Tested-by Vikram Sethi <[email protected]>
Quoted 30 lines
---
v15 -> v16:
- added Vikram's T-b
- Resolve rebase conflictv2 -> v3:
-kvm_irqchip_add_qemuirq_irqfd_notifierrenamed intokvm_irqchip_add_irqfd_notifier--- hw/intc/arm_gic_kvm.c | 6 ++++++ 1 file changed, 6 insertions(+)hw/intc/arm_gic_kvm.c +6
-0@@ -570,6 +570,12 @@ static void kvm_arm_gic_realize(DeviceState *dev, Error **errp)*/ i += (GIC_INTERNAL * s->num_cpu);qdev_init_gpio_in (dev, kvm_arm_gic_set_irq, i); + +for (i =0 ; i < s->num_irq -GIC_INTERNAL ; i++) { +qemu_irq irq =qdev_get_gpio_in (dev, i); +kvm_irqchip_set_qemuirq_gsi (kvm_state, irq, i); + } +/* We never use our outbound IRQ/FIQ lines but provide them so that * we maintain the same interface as the non-KVM GIC. */
Re: [Qemu-devel] [RESEND PATCH v16 4/6] intc: arm_gic_kvm: set the qemu_irq/gsi mapping
In reply to Eric AugerQuoted 25 lines · Eric Auger
The
arm_gic_kvmnow callskvm_irqchip_set_qemuirq_gsito build the hash table storing qemu_irq/gsi mappings. From that point on irqfd can be setup directly from theqemu_irqusingkvm_irqchip_add_irqfd_notifier.Signed-off-by Eric Auger <[email protected]>Tested-by Vikram Sethi <[email protected]>---
v15 -> v16:
- added Vikram's T-b
- Resolve rebase conflictv2 -> v3:
-kvm_irqchip_add_qemuirq_irqfd_notifierrenamed intokvm_irqchip_add_irqfd_notifier--- hw/intc/arm_gic_kvm.c | 6 ++++++ 1 file changed, 6 insertions(+)hw/intc/arm_gic_kvm.c +0
-0
hw/intc/arm_gic_kvm.c static void kvm_arm_gic_realize(DeviceState *dev, Error **errp) · line 578
Quoted 9 lines
hw/intc/arm_gic_kvm.c +6
-0@@ -570,6 +570,12 @@ static void kvm_arm_gic_realize(DeviceState *dev, Error **errp)*/ i += (GIC_INTERNAL * s->num_cpu);qdev_init_gpio_in (dev, kvm_arm_gic_set_irq, i); + +for (i =0 ; i < s->num_irq -GIC_INTERNAL ; i++) { +qemu_irq irq =qdev_get_gpio_in (dev, i); +kvm_irqchip_set_qemuirq_gsi (kvm_state, irq, i); + } +
Is there documentation anywhere about what a "GSI" number is for ARM? Is this part of the kernel ABI?
thanks
-- PMM
Re: [Qemu-devel] [RESEND PATCH v16 4/6] intc: arm_gic_kvm: set the qemu_irq/gsi mapping
Quoted 30 lines · Peter Maydell
Quoted 29 lines · Eric Auger
The
arm_gic_kvmnow callskvm_irqchip_set_qemuirq_gsito build the hash table storing qemu_irq/gsi mappings. From that point on irqfd can be setup directly from theqemu_irqusingkvm_irqchip_add_irqfd_notifier.Signed-off-by Eric Auger <[email protected]>Tested-by Vikram Sethi <[email protected]>---
v15 -> v16:
- added Vikram's T-b
- Resolve rebase conflictv2 -> v3:
-kvm_irqchip_add_qemuirq_irqfd_notifierrenamed intokvm_irqchip_add_irqfd_notifier--- hw/intc/arm_gic_kvm.c | 6 ++++++ 1 file changed, 6 insertions(+)hw/intc/arm_gic_kvm.c +1
-0@@ -570,6 +570,12 @@ static void kvm_arm_gic_realize(DeviceState *dev, Error **errp)*/ i += (GIC_INTERNAL * s->num_cpu);qdev_init_gpio_in (dev, kvm_arm_gic_set_irq, i); +
Quoted 8 lines
hw/intc/arm_gic_kvm.c static void kvm_arm_gic_realize(DeviceState *dev, Error **errp) · line 578
Quoted 5 lines
hw/intc/arm_gic_kvm.c +5
-0+for (i =0 ; i < s->num_irq -GIC_INTERNAL ; i++) { +qemu_irq irq =qdev_get_gpio_in (dev, i); +kvm_irqchip_set_qemuirq_gsi (kvm_state, irq, i); + } +Is there documentation anywhere about what a "GSI" number is for ARM? Is this part of the kernel ABI?
Currently this is documented in kernel Documentation/virtual/kvm/api.txt in the KVM_IRQFD part. With current irqfd/arm implementation this GSI number matches an SPI index. Now we are also working on GSI routing support which will extend that meaning. The related RFC also documents ths gsi in the api.txt
Thanks
Eric
Quoted 4 lines
thanks
-- PMM
PATCH RESEND v16 5/6 sysbus: add irq_routing_notifier
In reply to Eric AugerAdd a new connect_irq_notifier notifier in the SysBusDeviceClass. This notifier, if populated, is called after sysbus_connect_irq.
This mechanism is used to setup VFIO signaling once VFIO platform devices get attached to their platform bus, on a machine init done notifier.
---
v15 -> v16:
- add Vikram's T-b
v14 -> v15:
- add Peter R-b
v2 -> v3 (integrated into this series v14):
- rename irq_routing_notifier into connect_irq_notifier
v1 -> v2:
- duly put the notifier in the class and not in the device
--- hw/core/sysbus.c | 6 ++++++ include/hw/sysbus.h | 1 + 2 files changed, 7 insertions(+)
hw/core/sysbus.c +6 -0
@@ -109,7 +109,13 @@ qemu_irq sysbus_get_connected_irq(SysBusDevice *dev, int n)void sysbus_connect_irq (SysBusDevice *dev,int n,qemu_irq irq) { +SysBusDeviceClass *sbd =SYS_BUS_DEVICE_GET_CLASS (dev); +qdev_connect_gpio_out_named (DEVICE (dev),SYSBUS_DEVICE_GPIO_IRQ , n, irq); + +if (sbd->connect_irq_notifier) { + sbd->connect_irq_notifier (dev, irq); + } }/* Check whether an MMIO region exists */
include/hw/sysbus.h +1 -0
@@ -41,6 +41,7 @@ typedef struct SysBusDeviceClass {/*< public >*/ int (*init)(SysBusDevice *dev); +void (*connect_irq_notifier)(SysBusDevice *dev,qemu_irq irq); } SysBusDeviceClass;struct SysBusDevice {
Signature · 1 line
1.8.3.2
PATCH RESEND v16 6/6 hw/vfio/platform: add irqfd support
In reply to Eric AugerThis patch aims at optimizing IRQ handling using irqfd framework.
Instead of handling the eventfds on user-side they are handled on kernel side using
- the KVM irqfd framework,
- the VFIO driver virqfd framework.
the virtual IRQ completion is trapped at interrupt controller This removes the need for fast/slow path swap.
Overall this brings significant performance improvements.
---
v15 -> v16:
- add Vikram's T-b
v13 -> v14:
- use connect_irq_notifier
- remove trace_vfio_platform_start_eventfd
v12 -> v13:
- setup the new mechanism for starting irqfd, based on LinkPropertySetter override
- use kvm_irqchip_[add,remove]_irqfd_notifier new functions: no need to bother about gsi (hence virtualID could be removed with small change in trace-events)
v10 -> v11:
- Add Alex' Reviewed-by
- introduce kvm_accel in this patch and initialize it
v5 -> v6
- rely on kvm_irqfds_enabled() and kvm_resamplefds_enabled()
- guard KVM code with #ifdef CONFIG_KVM
v3 -> v4:
[Alvise Rigo]
Use of VFIO Platform driver v6 unmask/virqfd feature and removal of resamplefd handler. Physical IRQ unmasking is now done in VFIO driver.
v3:
[Eric Auger]
initial support with resamplefd handled on QEMU side since the unmask was not supported on VFIO platform driver v5.
--- hw/vfio/platform.c | 107 ++++++++++++++++++++++++++++++++++++++++ include/hw/vfio/vfio-platform.h | 2 + trace-events | 1 + 3 files changed, 110 insertions(+)
hw/vfio/platform.c +107 -0
@@ -26,6 +26,7 @@#include "hw/sysbus.h" #include "trace.h" #include "hw/platform-bus.h" +#include "sysemu/kvm.h" /* * Functions used whatever the injection method
@@ -51,6 +52,7 @@ static VFIOINTp *vfio_init_intp(VFIODevice *vbasedev, intp->pin = info.index; intp->flags = info.flags; intp->state =VFIO_IRQ_INACTIVE ; + intp->kvm_accel = false;sysbus_init_irq (sbdev, &intp->qemuirq);
@@ -61,6 +63,13 @@ static VFIOINTp *vfio_init_intp(VFIODevice *vbasedev,error_report ("vfio: Error: trigger event_notifier_init failed " );return NULL ; } +/* Get an eventfd for resample/unmask */ + ret =event_notifier_init (&intp->unmask,0 ); +if (ret) { +g_free (intp); +error_report ("vfio: Error: resample event_notifier_init failed eoi" ); +return NULL ; + }QLIST_INSERT_HEAD (&vdev->intp_list, intp, next);return intp;
@@ -315,6 +324,95 @@ static int vfio_start_eventfd_injection(VFIOINTp *intp)return ret; } +/* +* Functions used for irqfd +*/ + +#ifdef CONFIG_KVM + +/** +* vfio_set_resample_eventfd - sets the resamplefd for an IRQ +* @intp: the IRQ struct handle +* programs the VFIO driver to unmask this IRQ when the +* intp->unmask eventfd is triggered +*/ +static int vfio_set_resample_eventfd (VFIOINTp *intp) +{ +VFIODevice *vbasedev = &intp->vdev->vbasedev; +struct vfio_irq_set *irq_set; +int argsz, ret; +int32_t *pfd; + + argsz =sizeof (*irq_set) +sizeof (*pfd); + irq_set =g_malloc0 (argsz); + irq_set->argsz = argsz; + irq_set->flags =VFIO_IRQ_SET_DATA_EVENTFD |VFIO_IRQ_SET_ACTION_UNMASK ; + irq_set->index = intp->pin; + irq_set->start =0 ; + irq_set->count =1 ; + pfd = (int32_t *)&irq_set->data; + *pfd =event_notifier_get_fd (&intp->unmask); +qemu_set_fd_handler (*pfd,NULL ,NULL ,NULL ); + ret =ioctl (vbasedev->fd,VFIO_DEVICE_SET_IRQS , irq_set); +g_free (irq_set); +if (ret <0 ) { +error_report ("vfio: Failed to set resample eventfd: %m" ); + } +return ret; +} + +static void vfio_start_irqfd_injection (SysBusDevice *sbdev,qemu_irq irq) +{ +VFIOPlatformDevice *vdev =VFIO_PLATFORM_DEVICE (sbdev); +VFIOINTp *intp; +bool found = false; + + QLIST_FOREACH(intp, &vdev->intp_list, next) { +if (intp->qemuirq == irq) { + found = true; +break ; + } + } +assert (found); + +/* Get to a known interrupt state */ +qemu_set_fd_handler (event_notifier_get_fd (&intp->interrupt), +NULL ,NULL , vdev); + +vfio_mask_single_irqindex (&vdev->vbasedev, intp->pin); +qemu_set_irq (intp->qemuirq,0 ); + +if (kvm_irqchip_add_irqfd_notifier (kvm_state, &intp->interrupt, + &intp->unmask, irq) <0 ) { + goto fail_irqfd; + } + +if (vfio_set_trigger_eventfd (intp,NULL ) <0 ) { + goto fail_vfio; + } +if (vfio_set_resample_eventfd (intp) <0 ) { + goto fail_vfio; + } + +/* Let'em rip */ +vfio_unmask_single_irqindex (&vdev->vbasedev, intp->pin); + + intp->kvm_accel = true; + +trace_vfio_platform_start_irqfd_injection (intp->pin, +event_notifier_get_fd (&intp->interrupt), +event_notifier_get_fd (&intp->unmask)); +return ; +fail_vfio: +kvm_irqchip_remove_irqfd_notifier (kvm_state, &intp->interrupt, irq); +fail_irqfd: +vfio_start_eventfd_injection (intp); +vfio_unmask_single_irqindex (&vdev->vbasedev, intp->pin); +return ; +} + +#endif /* CONFIG_KVM */ +/* VFIO skeleton */ static void vfio_platform_compute_needs_reset (VFIODevice *vbasedev)
@@ -548,6 +646,7 @@ static void vfio_platform_realize(DeviceState *dev, Error **errp) {VFIOPlatformDevice *vdev =VFIO_PLATFORM_DEVICE (dev);SysBusDevice *sbdev =SYS_BUS_DEVICE (dev); +SysBusDeviceClass *sbc =SYS_BUS_DEVICE_GET_CLASS (dev);VFIODevice *vbasedev = &vdev->vbasedev;VFIOINTp *intp;int i, ret;
@@ -555,6 +654,13 @@ static void vfio_platform_realize(DeviceState *dev, Error **errp) vbasedev->type =VFIO_DEVICE_TYPE_PLATFORM ; vbasedev->ops = &vfio_platform_ops; +#ifdef CONFIG_KVM +if (kvm_irqfds_enabled () &&kvm_resamplefds_enabled () && + vdev->irqfd_allowed) { + sbc->connect_irq_notifier = vfio_start_irqfd_injection; + } +#endif +trace_vfio_platform_realize (vbasedev->name, vdev->compat); ret =vfio_base_device_init (vbasedev);
@@ -584,6 +690,7 @@ static Property vfio_platform_dev_properties[] = {DEFINE_PROP_BOOL ("x-mmap" , VFIOPlatformDevice, vbasedev.allow_mmap, true),DEFINE_PROP_UINT32 ("mmap-timeout-ms" , VFIOPlatformDevice, mmap_timeout,1100 ), +DEFINE_PROP_BOOL ("x-irqfd" , VFIOPlatformDevice, irqfd_allowed, true), DEFINE_PROP_END_OF_LIST(), };
include/hw/vfio/vfio-platform.h +2 -0
@@ -41,6 +41,7 @@ typedef struct VFIOINTp {int state;/* inactive, pending, active */ uint8_t pin;/* index */ uint32_t flags;/* IRQ info flags */ +bool kvm_accel;/* set when QEMU bypass through KVM enabled */ } VFIOINTp;/* function type for user side eventfd handler */
@@ -57,6 +58,7 @@ typedef struct VFIOPlatformDevice {uint32_t mmap_timeout;/* delay to re-enable mmaps after interrupt */ QEMUTimer *mmap_timer;/* allows fast-path resume after IRQ hit */ QemuMutex intp_mutex;/* protect the intp_list IRQ state */ +bool irqfd_allowed;/* debug option to force irqfd on/off */ } VFIOPlatformDevice; typedef struct VFIOPlatformDeviceClass {
trace-events +1 -0
@@ -1594,6 +1594,7 @@ vfio_platform_intp_interrupt(int pin, int fd) "Inject IRQ #%d (fd = %d)" vfio_platform_intp_inject_pending_lockheld(int pin, int fd) "Inject pending IRQ #%d (fd = %d)" vfio_platform_populate_interrupts(int pin, int count, int flags) "- IRQ index %d: count %d, flags=0x%x" vfio_intp_interrupt_set_pending(int index) "irq %d is set PENDING" +vfio_platform_start_irqfd_injection(int index, int fd, int resamplefd) "IRQ index=%d, fd = %d, resamplefd = %d" #hw/acpi/memory_hotplug.c mhp_acpi_invalid_slot_selected(uint32_t slot) "0x%"PRIx32
Signature · 1 line
1.8.3.2
Re: [Qemu-devel] [RESEND PATCH v16 6/6] hw/vfio/platform: add irqfd support
Quoted 77 lines · Eric Auger
This patch aims at optimizing IRQ handling using irqfd framework.
Instead of handling the eventfds on user-side they are handled on kernel side using
- the KVM irqfd framework,
- the VFIO driver virqfd framework.the virtual IRQ completion is trapped at interrupt controller This removes the need for fast/slow path swap.
Overall this brings significant performance improvements.
Signed-off-by Alvise Rigo <[email protected]>Signed-off-by Eric Auger <[email protected]>Reviewed-by Alex Bennée <[email protected]>Tested-by Vikram Sethi <[email protected]>---
v15 -> v16:
- add Vikram's T-bv13 -> v14:
- useconnect_irq_notifier
- removetrace_vfio_platform_start_eventfdv12 -> v13:
- setup the new mechanism for starting irqfd, based on LinkPropertySetter override
- use kvm_irqchip_[add,remove]_irqfd_notifier new functions: no need to bother about gsi (hence virtualID could be removed with small change in trace-events)v10 -> v11:
- Add Alex' Reviewed-by
- introducekvm_accelin this patch and initialize itv5 -> v6
- rely onkvm_irqfds_enabled()andkvm_resamplefds_enabled()
- guard KVM code with #ifdefCONFIG_KVMv3 -> v4:
[Alvise Rigo]
Use of VFIO Platform driver v6 unmask/virqfd feature and removal of resamplefd handler. Physical IRQ unmasking is now done in VFIO driver.v3:
[Eric Auger]
initial support with resamplefd handled on QEMU side since the unmask was not supported on VFIO platform driver v5.--- hw/vfio/platform.c | 107 ++++++++++++++++++++++++++++++++++++++++ include/hw/vfio/vfio-platform.h | 2 + trace-events | 1 + 3 files changed, 110 insertions(+)hw/vfio/platform.c +2
-0@@ -26,6 +26,7 @@#include "hw/sysbus.h" #include "trace.h" #include "hw/platform-bus.h" +#include "sysemu/kvm.h" /* * Functions used whatever the injection method @@ -51,6 +52,7 @@ static VFIOINTp *vfio_init_intp(VFIODevice *vbasedev, intp->pin = info.index; intp->flags = info.flags; intp->state =VFIO_IRQ_INACTIVE ; + intp->kvm_accel = false;sysbus_init_irq (sbdev, &intp->qemuirq);
hw/vfio/platform.c static VFIOINTp *vfio_init_intp(VFIODevice *vbasedev, · line 70
Quoted 8 lines
hw/vfio/platform.c +5
-0@@ -61,6 +63,13 @@ static VFIOINTp *vfio_init_intp(VFIODevice *vbasedev,error_report ("vfio: Error: trigger event_notifier_init failed " );return NULL ; } +/* Get an eventfd for resample/unmask */ + ret =event_notifier_init (&intp->unmask,0 ); +if (ret) { +g_free (intp); +error_report ("vfio: Error: resample event_notifier_init failed eoi" );
This error message seems rather cryptic; what is it trying to tell the user?
Quoted 45 lines
hw/vfio/platform.c +38
-0+return NULL ; + }QLIST_INSERT_HEAD (&vdev->intp_list, intp, next);return intp;@@ -315,6 +324,95 @@ static int vfio_start_eventfd_injection(VFIOINTp *intp)return ret; } +/* +* Functions used for irqfd +*/ + +#ifdef CONFIG_KVM + +/** +* vfio_set_resample_eventfd - sets the resamplefd for an IRQ +* @intp: the IRQ struct handle +* programs the VFIO driver to unmask this IRQ when the +* intp->unmask eventfd is triggered +*/ +static int vfio_set_resample_eventfd (VFIOINTp *intp) +{ +VFIODevice *vbasedev = &intp->vdev->vbasedev; +struct vfio_irq_set *irq_set; +int argsz, ret; +int32_t *pfd; + + argsz =sizeof (*irq_set) +sizeof (*pfd); + irq_set =g_malloc0 (argsz); + irq_set->argsz = argsz; + irq_set->flags =VFIO_IRQ_SET_DATA_EVENTFD |VFIO_IRQ_SET_ACTION_UNMASK ; + irq_set->index = intp->pin; + irq_set->start =0 ; + irq_set->count =1 ; + pfd = (int32_t *)&irq_set->data; + *pfd =event_notifier_get_fd (&intp->unmask); +qemu_set_fd_handler (*pfd,NULL ,NULL ,NULL ); + ret =ioctl (vbasedev->fd,VFIO_DEVICE_SET_IRQS , irq_set); +g_free (irq_set); +if (ret <0 ) { +error_report ("vfio: Failed to set resample eventfd: %m" ); + } +return ret; +}
hw/vfio/platform.c static int vfio_start_eventfd_injection(VFIOINTp *intp) · line 372
Quoted 10 lines
hw/vfio/platform.c +10
-0+ +static void vfio_start_irqfd_injection (SysBusDevice *sbdev,qemu_irq irq) +{ +VFIOPlatformDevice *vdev =VFIO_PLATFORM_DEVICE (sbdev); +VFIOINTp *intp; +bool found = false; + + QLIST_FOREACH(intp, &vdev->intp_list, next) { +if (intp->qemuirq == irq) { + found = true;
Stray double-space.
Quoted 15 lines
hw/vfio/platform.c +15
-0+break ; + } + } +assert (found); + +/* Get to a known interrupt state */ +qemu_set_fd_handler (event_notifier_get_fd (&intp->interrupt), +NULL ,NULL , vdev); + +vfio_mask_single_irqindex (&vdev->vbasedev, intp->pin); +qemu_set_irq (intp->qemuirq,0 ); + +if (kvm_irqchip_add_irqfd_notifier (kvm_state, &intp->interrupt, + &intp->unmask, irq) <0 ) { + goto fail_irqfd;
hw/vfio/platform.c static int vfio_start_eventfd_injection(VFIOINTp *intp) · line 397
Quoted 10 lines
hw/vfio/platform.c +10
-0+ } + +if (vfio_set_trigger_eventfd (intp,NULL ) <0 ) { + goto fail_vfio; + } +if (vfio_set_resample_eventfd (intp) <0 ) { + goto fail_vfio; + } + +/* Let'em rip */
This comment could probably be phrased more informatively :-)
Quoted 30 lines
hw/vfio/platform.c +19
-0+vfio_unmask_single_irqindex (&vdev->vbasedev, intp->pin); + + intp->kvm_accel = true; + +trace_vfio_platform_start_irqfd_injection (intp->pin, +event_notifier_get_fd (&intp->interrupt), +event_notifier_get_fd (&intp->unmask)); +return ; +fail_vfio: +kvm_irqchip_remove_irqfd_notifier (kvm_state, &intp->interrupt, irq); +fail_irqfd: +vfio_start_eventfd_injection (intp); +vfio_unmask_single_irqindex (&vdev->vbasedev, intp->pin); + return; +} + +#endif /* CONFIG_KVM */ +/* VFIO skeleton */ static void vfio_platform_compute_needs_reset (VFIODevice *vbasedev)@@ -548,6 +646,7 @@ static void vfio_platform_realize(DeviceState *dev, Error **errp) {VFIOPlatformDevice *vdev =VFIO_PLATFORM_DEVICE (dev);SysBusDevice *sbdev =SYS_BUS_DEVICE (dev); +SysBusDeviceClass *sbc =SYS_BUS_DEVICE_GET_CLASS (dev);VFIODevice *vbasedev = &vdev->vbasedev;VFIOINTp *intp;int i, ret;
hw/vfio/platform.c static void vfio_platform_realize(DeviceState *dev, Error **errp) · line 662
Quoted 9 lines
hw/vfio/platform.c +6
-0@@ -555,6 +654,13 @@ static void vfio_platform_realize(DeviceState *dev, Error **errp) vbasedev->type =VFIO_DEVICE_TYPE_PLATFORM ; vbasedev->ops = &vfio_platform_ops; +#ifdef CONFIG_KVM +if (kvm_irqfds_enabled () &&kvm_resamplefds_enabled () && + vdev->irqfd_allowed) { + sbc->connect_irq_notifier = vfio_start_irqfd_injection; + } +#endif
You shouldn't need this CONFIG_KVM ifdef -- we deliberately provide "always false" versions of functions like kvm_irqfds_enabled() in sysemu/kvm.h so that we can avoid littering code with ifdefs.
It looks like your commit f41389ae3c54b failed to provide the non-KVM version of kvm_resamplefds_enabled(), but the correct fix is to provide it...
This implies removing the ifdefs earlier around vfio_start_irqfd_injection, but that should be OK -- we provide stub versions of functions like kvm_irqchip_add_irqfd_notifier() so the non-KVM code can still compile.
Quoted 45 lines
hw/vfio/platform.c +2
-0+trace_vfio_platform_realize (vbasedev->name, vdev->compat); ret =vfio_base_device_init (vbasedev);@@ -584,6 +690,7 @@ static Property vfio_platform_dev_properties[] = {DEFINE_PROP_BOOL ("x-mmap" , VFIOPlatformDevice, vbasedev.allow_mmap, true),DEFINE_PROP_UINT32 ("mmap-timeout-ms" , VFIOPlatformDevice, mmap_timeout,1100 ), +DEFINE_PROP_BOOL ("x-irqfd" , VFIOPlatformDevice, irqfd_allowed, true), DEFINE_PROP_END_OF_LIST(), };include/hw/vfio/vfio-platform.h +2
-0@@ -41,6 +41,7 @@ typedef struct VFIOINTp {int state;/* inactive, pending, active */ uint8_t pin;/* index */ uint32_t flags;/* IRQ info flags */ +bool kvm_accel;/* set when QEMU bypass through KVM enabled */ } VFIOINTp;/* function type for user side eventfd handler */ @@ -57,6 +58,7 @@ typedef struct VFIOPlatformDevice {uint32_t mmap_timeout;/* delay to re-enable mmaps after interrupt */ QEMUTimer *mmap_timer;/* allows fast-path resume after IRQ hit */ QemuMutex intp_mutex;/* protect the intp_list IRQ state */ +bool irqfd_allowed;/* debug option to force irqfd on/off */ } VFIOPlatformDevice; typedef struct VFIOPlatformDeviceClass {trace-events +1
-1@@ -1594,6 +1594,7 @@ vfio_platform_intp_interrupt(int pin, int fd) "Inject IRQ #%d (fd = %d)" vfio_platform_intp_inject_pending_lockheld(int pin, int fd) "Inject pending IRQ #%d (fd = %d)" vfio_platform_populate_interrupts(int pin, int count, int flags) "- IRQ index %d: count %d, flags=0x%x" vfio_intp_interrupt_set_pending(int index) "irq %d is set PENDING" +vfio_platform_start_irqfd_injection(int index, int fd, int resamplefd) "IRQ index=%d, fd = %d, resamplefd = %d" #hw/acpi/memory_hotplug.c mhp_acpi_invalid_slot_selected(uint32_t slot) "0x%"PRIx32 --
Quoted 2 lines
1.8.3.2
thanks
-- PMM
Re: [Qemu-devel] [RESEND PATCH v16 6/6] hw/vfio/platform: add irqfd support
Quoted 77 lines · Peter Maydell
Quoted 76 lines · Eric Auger
This patch aims at optimizing IRQ handling using irqfd framework.
Instead of handling the eventfds on user-side they are handled on kernel side using
- the KVM irqfd framework,
- the VFIO driver virqfd framework.the virtual IRQ completion is trapped at interrupt controller This removes the need for fast/slow path swap.
Overall this brings significant performance improvements.
Signed-off-by Alvise Rigo <[email protected]>Signed-off-by Eric Auger <[email protected]>Reviewed-by Alex Bennée <[email protected]>Tested-by Vikram Sethi <[email protected]>---
v15 -> v16:
- add Vikram's T-bv13 -> v14:
- useconnect_irq_notifier
- removetrace_vfio_platform_start_eventfdv12 -> v13:
- setup the new mechanism for starting irqfd, based on LinkPropertySetter override
- use kvm_irqchip_[add,remove]_irqfd_notifier new functions: no need to bother about gsi (hence virtualID could be removed with small change in trace-events)v10 -> v11:
- Add Alex' Reviewed-by
- introducekvm_accelin this patch and initialize itv5 -> v6
- rely onkvm_irqfds_enabled()andkvm_resamplefds_enabled()
- guard KVM code with #ifdefCONFIG_KVMv3 -> v4:
[Alvise Rigo]
Use of VFIO Platform driver v6 unmask/virqfd feature and removal of resamplefd handler. Physical IRQ unmasking is now done in VFIO driver.v3:
[Eric Auger]
initial support with resamplefd handled on QEMU side since the unmask was not supported on VFIO platform driver v5.--- hw/vfio/platform.c | 107 ++++++++++++++++++++++++++++++++++++++++ include/hw/vfio/vfio-platform.h | 2 + trace-events | 1 + 3 files changed, 110 insertions(+)hw/vfio/platform.c +2
-0@@ -26,6 +26,7 @@#include "hw/sysbus.h" #include "trace.h" #include "hw/platform-bus.h" +#include "sysemu/kvm.h" /* * Functions used whatever the injection method @@ -51,6 +52,7 @@ static VFIOINTp *vfio_init_intp(VFIODevice *vbasedev, intp->pin = info.index; intp->flags = info.flags; intp->state =VFIO_IRQ_INACTIVE ; + intp->kvm_accel = false;sysbus_init_irq (sbdev, &intp->qemuirq);
Quoted 11 lines
hw/vfio/platform.c static VFIOINTp *vfio_init_intp(VFIODevice *vbasedev, · line 70
Quoted 9 lines
hw/vfio/platform.c +5
-0@@ -61,6 +63,13 @@ static VFIOINTp *vfio_init_intp(VFIODevice *vbasedev,error_report ("vfio: Error: trigger event_notifier_init failed " );return NULL ; } +/* Get an eventfd for resample/unmask */ + ret =event_notifier_init (&intp->unmask,0 ); +if (ret) { +g_free (intp); +error_report ("vfio: Error: resample event_notifier_init failed eoi" );This error message seems rather cryptic; what is it trying to tell the user?
I acknowledge ;-)
I will replace by "vfio: Unable to init event notifier for resampling"
Quoted 50 lines
Quoted 50 lines
hw/vfio/platform.c +42
-0+return NULL ; + }QLIST_INSERT_HEAD (&vdev->intp_list, intp, next);return intp;@@ -315,6 +324,95 @@ static int vfio_start_eventfd_injection(VFIOINTp *intp)return ret; } +/* +* Functions used for irqfd +*/ + +#ifdef CONFIG_KVM + +/** +* vfio_set_resample_eventfd - sets the resamplefd for an IRQ +* @intp: the IRQ struct handle +* programs the VFIO driver to unmask this IRQ when the +* intp->unmask eventfd is triggered +*/ +static int vfio_set_resample_eventfd (VFIOINTp *intp) +{ +VFIODevice *vbasedev = &intp->vdev->vbasedev; +struct vfio_irq_set *irq_set; +int argsz, ret; +int32_t *pfd; + + argsz =sizeof (*irq_set) +sizeof (*pfd); + irq_set =g_malloc0 (argsz); + irq_set->argsz = argsz; + irq_set->flags =VFIO_IRQ_SET_DATA_EVENTFD |VFIO_IRQ_SET_ACTION_UNMASK ; + irq_set->index = intp->pin; + irq_set->start =0 ; + irq_set->count =1 ; + pfd = (int32_t *)&irq_set->data; + *pfd =event_notifier_get_fd (&intp->unmask); +qemu_set_fd_handler (*pfd,NULL ,NULL ,NULL ); + ret =ioctl (vbasedev->fd,VFIO_DEVICE_SET_IRQS , irq_set); +g_free (irq_set); +if (ret <0 ) { +error_report ("vfio: Failed to set resample eventfd: %m" ); + } +return ret; +} + +static void vfio_start_irqfd_injection (SysBusDevice *sbdev,qemu_irq irq) +{ +VFIOPlatformDevice *vdev =VFIO_PLATFORM_DEVICE (sbdev);
Quoted 8 lines
hw/vfio/platform.c static int vfio_start_eventfd_injection(VFIOINTp *intp) · line 372
Quoted 6 lines
hw/vfio/platform.c +6
-0+VFIOINTp *intp; +bool found = false; + + QLIST_FOREACH(intp, &vdev->intp_list, next) { +if (intp->qemuirq == irq) { + found = true;Stray double-space.
ok
Quoted 20 lines
Quoted 20 lines
hw/vfio/platform.c +19
-0+break ; + } + } +assert (found); + +/* Get to a known interrupt state */ +qemu_set_fd_handler (event_notifier_get_fd (&intp->interrupt), +NULL ,NULL , vdev); + +vfio_mask_single_irqindex (&vdev->vbasedev, intp->pin); +qemu_set_irq (intp->qemuirq,0 ); + +if (kvm_irqchip_add_irqfd_notifier (kvm_state, &intp->interrupt, + &intp->unmask, irq) <0 ) { + goto fail_irqfd; + } + +if (vfio_set_trigger_eventfd (intp,NULL ) <0 ) { + goto fail_vfio;
Quoted 8 lines
hw/vfio/platform.c static int vfio_start_eventfd_injection(VFIOINTp *intp) · line 397
Quoted 6 lines
hw/vfio/platform.c +6
-0+ } +if (vfio_set_resample_eventfd (intp) <0 ) { + goto fail_vfio; + } + +/* Let'em rip */ This comment could probably be phrased more informatively :-)
sure
Quoted 45 lines
Quoted 31 lines
hw/vfio/platform.c +19
-0+vfio_unmask_single_irqindex (&vdev->vbasedev, intp->pin); + + intp->kvm_accel = true; + +trace_vfio_platform_start_irqfd_injection (intp->pin, +event_notifier_get_fd (&intp->interrupt), +event_notifier_get_fd (&intp->unmask)); +return ; +fail_vfio: +kvm_irqchip_remove_irqfd_notifier (kvm_state, &intp->interrupt, irq); +fail_irqfd: +vfio_start_eventfd_injection (intp); +vfio_unmask_single_irqindex (&vdev->vbasedev, intp->pin); + return; +} + +#endif /* CONFIG_KVM */ +/* VFIO skeleton */ static void vfio_platform_compute_needs_reset (VFIODevice *vbasedev)@@ -548,6 +646,7 @@ static void vfio_platform_realize(DeviceState *dev, Error **errp) {VFIOPlatformDevice *vdev =VFIO_PLATFORM_DEVICE (dev);SysBusDevice *sbdev =SYS_BUS_DEVICE (dev); +SysBusDeviceClass *sbc =SYS_BUS_DEVICE_GET_CLASS (dev);VFIODevice *vbasedev = &vdev->vbasedev;VFIOINTp *intp;int i, ret;hw/vfio/platform.c static void vfio_platform_realize(DeviceState *dev, Error **errp) · line 662
Quoted 9 lines
hw/vfio/platform.c +6
-0@@ -555,6 +654,13 @@ static void vfio_platform_realize(DeviceState *dev, Error **errp) vbasedev->type =VFIO_DEVICE_TYPE_PLATFORM ; vbasedev->ops = &vfio_platform_ops; +#ifdef CONFIG_KVM +if (kvm_irqfds_enabled () &&kvm_resamplefds_enabled () && + vdev->irqfd_allowed) { + sbc->connect_irq_notifier = vfio_start_irqfd_injection; + } +#endif You shouldn't need this
CONFIG_KVMifdef -- we deliberately provide "always false" versions of functions likekvm_irqfds_enabled()insysemu/kvm.hso that we can avoid littering code with ifdefs.
Quoted 3 lines
It looks like your commit
f41389ae3c54bfailed to provide the non-KVM version ofkvm_resamplefds_enabled(), but the correct fix is to provide it...
OK I ignored that. I will add kvm_resamplefds_allowed in kvm-stub.c then.
Quoted 4 lines
This implies removing the ifdefs earlier around
vfio_start_irqfd_injection, but that should be OK -- we provide stub versions of functions likekvm_irqchip_add_irqfd_notifier()so the non-KVM code can still compile.
OK will try this out.
So this answers my previous question. Will send a v17
Thanks
Eric
Quoted 52 lines
Quoted 46 lines
hw/vfio/platform.c +2
-0+trace_vfio_platform_realize (vbasedev->name, vdev->compat); ret =vfio_base_device_init (vbasedev);@@ -584,6 +690,7 @@ static Property vfio_platform_dev_properties[] = {DEFINE_PROP_BOOL ("x-mmap" , VFIOPlatformDevice, vbasedev.allow_mmap, true),DEFINE_PROP_UINT32 ("mmap-timeout-ms" , VFIOPlatformDevice, mmap_timeout,1100 ), +DEFINE_PROP_BOOL ("x-irqfd" , VFIOPlatformDevice, irqfd_allowed, true), DEFINE_PROP_END_OF_LIST(), };include/hw/vfio/vfio-platform.h +2
-0@@ -41,6 +41,7 @@ typedef struct VFIOINTp {int state;/* inactive, pending, active */ uint8_t pin;/* index */ uint32_t flags;/* IRQ info flags */ +bool kvm_accel;/* set when QEMU bypass through KVM enabled */ } VFIOINTp;/* function type for user side eventfd handler */ @@ -57,6 +58,7 @@ typedef struct VFIOPlatformDevice {uint32_t mmap_timeout;/* delay to re-enable mmaps after interrupt */ QEMUTimer *mmap_timer;/* allows fast-path resume after IRQ hit */ QemuMutex intp_mutex;/* protect the intp_list IRQ state */ +bool irqfd_allowed;/* debug option to force irqfd on/off */ } VFIOPlatformDevice; typedef struct VFIOPlatformDeviceClass {trace-events +1
-1@@ -1594,6 +1594,7 @@ vfio_platform_intp_interrupt(int pin, int fd) "Inject IRQ #%d (fd = %d)" vfio_platform_intp_inject_pending_lockheld(int pin, int fd) "Inject pending IRQ #%d (fd = %d)" vfio_platform_populate_interrupts(int pin, int count, int flags) "- IRQ index %d: count %d, flags=0x%x" vfio_intp_interrupt_set_pending(int index) "irq %d is set PENDING" +vfio_platform_start_irqfd_injection(int index, int fd, int resamplefd) "IRQ index=%d, fd = %d, resamplefd = %d" #hw/acpi/memory_hotplug.c mhp_acpi_invalid_slot_selected(uint32_t slot) "0x%"PRIx32 --Quoted 2 lines
1.8.3.2
thanks
-- PMM
Re: [Qemu-devel] [RESEND PATCH v16 6/6] hw/vfio/platform: add irqfd support
In reply to Eric AugerQuoted 134 lines · Eric Auger
This patch aims at optimizing IRQ handling using irqfd framework.
Instead of handling the eventfds on user-side they are handled on kernel side using
- the KVM irqfd framework,
- the VFIO driver virqfd framework.the virtual IRQ completion is trapped at interrupt controller This removes the need for fast/slow path swap.
Overall this brings significant performance improvements.
Signed-off-by Alvise Rigo <[email protected]>Signed-off-by Eric Auger <[email protected]>Reviewed-by Alex Bennée <[email protected]>Tested-by Vikram Sethi <[email protected]>---
v15 -> v16:
- add Vikram's T-bv13 -> v14:
- useconnect_irq_notifier
- removetrace_vfio_platform_start_eventfdv12 -> v13:
- setup the new mechanism for starting irqfd, based on LinkPropertySetter override
- use kvm_irqchip_[add,remove]_irqfd_notifier new functions: no need to bother about gsi (hence virtualID could be removed with small change in trace-events)v10 -> v11:
- Add Alex' Reviewed-by
- introducekvm_accelin this patch and initialize itv5 -> v6
- rely onkvm_irqfds_enabled()andkvm_resamplefds_enabled()
- guard KVM code with #ifdefCONFIG_KVMv3 -> v4:
[Alvise Rigo]
Use of VFIO Platform driver v6 unmask/virqfd feature and removal of resamplefd handler. Physical IRQ unmasking is now done in VFIO driver.v3:
[Eric Auger]
initial support with resamplefd handled on QEMU side since the unmask was not supported on VFIO platform driver v5.--- hw/vfio/platform.c | 107 ++++++++++++++++++++++++++++++++++++++++ include/hw/vfio/vfio-platform.h | 2 + trace-events | 1 + 3 files changed, 110 insertions(+)hw/vfio/platform.c +49
-0@@ -26,6 +26,7 @@#include "hw/sysbus.h" #include "trace.h" #include "hw/platform-bus.h" +#include "sysemu/kvm.h" /* * Functions used whatever the injection method @@ -51,6 +52,7 @@ static VFIOINTp *vfio_init_intp(VFIODevice *vbasedev, intp->pin = info.index; intp->flags = info.flags; intp->state =VFIO_IRQ_INACTIVE ; + intp->kvm_accel = false;sysbus_init_irq (sbdev, &intp->qemuirq);@@ -61,6 +63,13 @@ static VFIOINTp *vfio_init_intp(VFIODevice *vbasedev,error_report ("vfio: Error: trigger event_notifier_init failed " );return NULL ; } +/* Get an eventfd for resample/unmask */ + ret =event_notifier_init (&intp->unmask,0 ); +if (ret) { +g_free (intp); +error_report ("vfio: Error: resample event_notifier_init failed eoi" ); +return NULL ; + }QLIST_INSERT_HEAD (&vdev->intp_list, intp, next);return intp;@@ -315,6 +324,95 @@ static int vfio_start_eventfd_injection(VFIOINTp *intp)return ret; } +/* +* Functions used for irqfd +*/ + +#ifdef CONFIG_KVM + +/** +* vfio_set_resample_eventfd - sets the resamplefd for an IRQ +* @intp: the IRQ struct handle +* programs the VFIO driver to unmask this IRQ when the +* intp->unmask eventfd is triggered +*/ +static int vfio_set_resample_eventfd (VFIOINTp *intp) +{ +VFIODevice *vbasedev = &intp->vdev->vbasedev; +struct vfio_irq_set *irq_set; +int argsz, ret; +int32_t *pfd; + + argsz =sizeof (*irq_set) +sizeof (*pfd); + irq_set =g_malloc0 (argsz); + irq_set->argsz = argsz; + irq_set->flags =VFIO_IRQ_SET_DATA_EVENTFD |VFIO_IRQ_SET_ACTION_UNMASK ; + irq_set->index = intp->pin; + irq_set->start =0 ; + irq_set->count =1 ; + pfd = (int32_t *)&irq_set->data; + *pfd =event_notifier_get_fd (&intp->unmask); +qemu_set_fd_handler (*pfd,NULL ,NULL ,NULL ); + ret =ioctl (vbasedev->fd,VFIO_DEVICE_SET_IRQS , irq_set); +g_free (irq_set); +if (ret <0 ) { +error_report ("vfio: Failed to set resample eventfd: %m" ); + } +return ret; +} + +static void vfio_start_irqfd_injection (SysBusDevice *sbdev,qemu_irq irq) +{ +VFIOPlatformDevice *vdev =VFIO_PLATFORM_DEVICE (sbdev);
hw/vfio/platform.c static int vfio_start_eventfd_injection(VFIOINTp *intp) · line 376
Quoted 10 lines
hw/vfio/platform.c +10
-0+VFIOINTp *intp; +bool found = false; + + QLIST_FOREACH(intp, &vdev->intp_list, next) { +if (intp->qemuirq == irq) { + found = true; +break ; + } + } +assert (found);
You can simply assert(intp) and get rid of found.
Quoted 51 lines
hw/vfio/platform.c +40
-0+ +/* Get to a known interrupt state */ +qemu_set_fd_handler (event_notifier_get_fd (&intp->interrupt), +NULL ,NULL , vdev); + +vfio_mask_single_irqindex (&vdev->vbasedev, intp->pin); +qemu_set_irq (intp->qemuirq,0 ); + +if (kvm_irqchip_add_irqfd_notifier (kvm_state, &intp->interrupt, + &intp->unmask, irq) <0 ) { + goto fail_irqfd; + } + +if (vfio_set_trigger_eventfd (intp,NULL ) <0 ) { + goto fail_vfio; + } +if (vfio_set_resample_eventfd (intp) <0 ) { + goto fail_vfio; + } + +/* Let'em rip */ +vfio_unmask_single_irqindex (&vdev->vbasedev, intp->pin); + + intp->kvm_accel = true; + +trace_vfio_platform_start_irqfd_injection (intp->pin, +event_notifier_get_fd (&intp->interrupt), +event_notifier_get_fd (&intp->unmask)); +return ; +fail_vfio: +kvm_irqchip_remove_irqfd_notifier (kvm_state, &intp->interrupt, irq); +fail_irqfd: +vfio_start_eventfd_injection (intp); +vfio_unmask_single_irqindex (&vdev->vbasedev, intp->pin); + return; +} + +#endif /* CONFIG_KVM */ +/* VFIO skeleton */ static void vfio_platform_compute_needs_reset (VFIODevice *vbasedev)@@ -548,6 +646,7 @@ static void vfio_platform_realize(DeviceState *dev, Error **errp) {VFIOPlatformDevice *vdev =VFIO_PLATFORM_DEVICE (dev);SysBusDevice *sbdev =SYS_BUS_DEVICE (dev); +SysBusDeviceClass *sbc =SYS_BUS_DEVICE_GET_CLASS (dev);VFIODevice *vbasedev = &vdev->vbasedev;VFIOINTp *intp;int i, ret;
hw/vfio/platform.c static void vfio_platform_realize(DeviceState *dev, Error **errp) · line 660
Quoted 7 lines
hw/vfio/platform.c +4
-0@@ -555,6 +654,13 @@ static void vfio_platform_realize(DeviceState *dev, Error **errp) vbasedev->type =VFIO_DEVICE_TYPE_PLATFORM ; vbasedev->ops = &vfio_platform_ops; +#ifdef CONFIG_KVM +if (kvm_irqfds_enabled () &&kvm_resamplefds_enabled () && + vdev->irqfd_allowed) { + sbc->connect_irq_notifier = vfio_start_irqfd_injection;
Should we be abstracting this to a sysbus_register_connect_irq_notifier()? It seems a littler personal to be reaching in and setting it ourselves and would avoid us needing to reference the class.
Quoted 46 lines
hw/vfio/platform.c +4
-0+ } +#endif +trace_vfio_platform_realize (vbasedev->name, vdev->compat); ret =vfio_base_device_init (vbasedev);@@ -584,6 +690,7 @@ static Property vfio_platform_dev_properties[] = {DEFINE_PROP_BOOL ("x-mmap" , VFIOPlatformDevice, vbasedev.allow_mmap, true),DEFINE_PROP_UINT32 ("mmap-timeout-ms" , VFIOPlatformDevice, mmap_timeout,1100 ), +DEFINE_PROP_BOOL ("x-irqfd" , VFIOPlatformDevice, irqfd_allowed, true), DEFINE_PROP_END_OF_LIST(), };include/hw/vfio/vfio-platform.h +2
-0@@ -41,6 +41,7 @@ typedef struct VFIOINTp {int state;/* inactive, pending, active */ uint8_t pin;/* index */ uint32_t flags;/* IRQ info flags */ +bool kvm_accel;/* set when QEMU bypass through KVM enabled */ } VFIOINTp;/* function type for user side eventfd handler */ @@ -57,6 +58,7 @@ typedef struct VFIOPlatformDevice {uint32_t mmap_timeout;/* delay to re-enable mmaps after interrupt */ QEMUTimer *mmap_timer;/* allows fast-path resume after IRQ hit */ QemuMutex intp_mutex;/* protect the intp_list IRQ state */ +bool irqfd_allowed;/* debug option to force irqfd on/off */ } VFIOPlatformDevice; typedef struct VFIOPlatformDeviceClass {trace-events +1
-0@@ -1594,6 +1594,7 @@ vfio_platform_intp_interrupt(int pin, int fd) "Inject IRQ #%d (fd = %d)" vfio_platform_intp_inject_pending_lockheld(int pin, int fd) "Inject pending IRQ #%d (fd = %d)" vfio_platform_populate_interrupts(int pin, int count, int flags) "- IRQ index %d: count %d, flags=0x%x" vfio_intp_interrupt_set_pending(int index) "irq %d is set PENDING" +vfio_platform_start_irqfd_injection(int index, int fd, int resamplefd) "IRQ index=%d, fd = %d, resamplefd = %d" #hw/acpi/memory_hotplug.c mhp_acpi_invalid_slot_selected(uint32_t slot) "0x%"PRIx32
Re: [Qemu-devel] [RESEND PATCH v16 6/6] hw/vfio/platform: add irqfd support
Quoted 8 lines · Alex Williamson
Quoted 4 lines
hw/vfio/platform.c static void vfio_platform_realize(DeviceState *dev, Error **errp) · line 660
Quoted 4 lines
hw/vfio/platform.c +4
-0+#ifdef CONFIG_KVM +if (kvm_irqfds_enabled () &&kvm_resamplefds_enabled () && + vdev->irqfd_allowed) { + sbc->connect_irq_notifier = vfio_start_irqfd_injection;Should we be abstracting this to a
sysbus_register_connect_irq_notifier()? It seems a littler personal to be reaching in and setting it ourselves and would avoid us needing to reference the class.
It's your class, so it's not too bad to touch it from that point of view; on the other hand it's ugly to do it here nevertheless. :)
I think you should always set "sbc->connect_irq_notifier = vfio_start_irqfd_injection" in the class_init function. The vfio_start_irqfd_injection function can just exit if it finds "!kvm_enabled() || !kvm_irqfds_enabled() || !kvm_resamplefds_enabled() || !vdev->irqfd_allowed".
Paolo
Re: [Qemu-devel] [RESEND PATCH v16 6/6] hw/vfio/platform: add irqfd support
Quoted 12 lines · Paolo Bonzini
Quoted 8 lines · Alex Williamson
Quoted 4 lines
hw/vfio/platform.c static void vfio_platform_realize(DeviceState *dev, Error **errp) · line 660
Quoted 4 lines
hw/vfio/platform.c +4
-0+#ifdef CONFIG_KVM +if (kvm_irqfds_enabled () &&kvm_resamplefds_enabled () && + vdev->irqfd_allowed) { + sbc->connect_irq_notifier = vfio_start_irqfd_injection;Should we be abstracting this to a
sysbus_register_connect_irq_notifier()? It seems a littler personal to be reaching in and setting it ourselves and would avoid us needing to reference the class.
Quoted 8 lines
It's your class, so it's not too bad to touch it from that point of view; on the other hand it's ugly to do it here nevertheless. :)
I think you should always set "
sbc->connect_irq_notifier=vfio_start_irqfd_injection" in theclass_initfunction. Thevfio_start_irqfd_injectionfunction can just exit if it finds "!kvm_enabled() || !kvm_irqfds_enabled() || !kvm_resamplefds_enabled() || !vdev->irqfd_allowed".
OK thanks for the guidance. Alex, are you OK with that solution. It avoids touching the other patch
Thanks
Eric
Quoted 3 lines
Paolo
Re: [Qemu-devel] [RESEND PATCH v16 6/6] hw/vfio/platform: add irqfd support
Quoted 16 lines · Eric Auger
Quoted 15 lines · Paolo Bonzini
Quoted 8 lines · Alex Williamson
Quoted 4 lines
hw/vfio/platform.c static void vfio_platform_realize(DeviceState *dev, Error **errp) · line 660
Quoted 4 lines
hw/vfio/platform.c +4
-0+#ifdef CONFIG_KVM +if (kvm_irqfds_enabled () &&kvm_resamplefds_enabled () && + vdev->irqfd_allowed) { + sbc->connect_irq_notifier = vfio_start_irqfd_injection;Should we be abstracting this to a
sysbus_register_connect_irq_notifier()? It seems a littler personal to be reaching in and setting it ourselves and would avoid us needing to reference the class.It's your class, so it's not too bad to touch it from that point of view; on the other hand it's ugly to do it here nevertheless. :)
Quoted 7 lines
Quoted 5 lines
I think you should always set "
sbc->connect_irq_notifier=vfio_start_irqfd_injection" in theclass_initfunction. Thevfio_start_irqfd_injectionfunction can just exit if it finds "!kvm_enabled() || !kvm_irqfds_enabled() || !kvm_resamplefds_enabled() || !vdev->irqfd_allowed".OK thanks for the guidance. Alex, are you OK with that solution. It avoids touching the other patch
Yeah, I'm ok with that, it's less awkward from the class_init. Thanks,
Alex
ping!
Peter acked and pulled the first patch file. This now makes possible to instantiate the calxeda xgmac VFIO platform device.
Remain the irqfd setup helpers [2-4], the irq routing notifier [5] and eventually the VFIO code using those pieces [6].
Those patches were already discussed on the ML:
- irqfd setup:
https://lists.gnu.org/archive/html/qemu-devel/2015-04/msg02837.html
- irq routing notifier:
https://lists.nongnu.org/archive/html/qemu-devel/2015-04/msg03019.html.
Irqfd usage is a must have to provide an efficient (platform) passthrough and it would be a pity to introduce the functionality in 2.4 without the perf I think.
Thank you in advance
Best Regards
Eric
Quoted 243 lines · Eric Auger
Now that the platform-bus is instantiated in virt, the ARM sysbus-fdt infrastructure, the VFIO platform device and its first Calxeda xgmac specialization also are on the master, let's complete the work by
- enabling the instantiation of the VFIO Calxeda xgmac device
- enabling irqfd usage, which is vital for performance.So the resend only relates to patches 4-9 of the v16 series.
http://git.linaro.org/people/eric.auger/qemu.git
(branchvfio_integ_v16_partial_resend)Tested by Vikram Sethi on a SATA device featuring 2 MMIO regions. Code also exercised successfully by VOSYS for AMBA derivative series.
Thanks
Eric
v15 -> v16:
- no code change
- add Vikram's T-b
- rebase on 0915aed
- remove linux header update patch filev14 -> v15:
- add Peter R-b on sysbus: addirq_routing_notifier
- correctg_malloc0_nusage in skeleton
- correct return values of reset related functions
- include Cornelia's patch for header updatev13 -> v14:
- remove v13 9, 10, 11 patch files and replace them by a single patch file "sysbus: addirq_routing_notifier".
- in skeleton, fix ENAMETOOLONG sign
- remove VFIOINTp virtualID in "add irq assignment" patch file
- removedtrace_vfio_platform_start_eventfdv12 -> v13:
- header update but same update was already sent by Cornelia
- Rework VFIO signaling & irqfd setup: restored 2-step setup featuring eventfd setup on realize and then irqfd setup on irq binding.
- irqfd setup now useskvm_irqchip_add_irqfd_notifierand sysbusirq_set_hookoverride. This leads to the introduction of 6 patch files enabling those 2 features. Paolo advised to introducekvm_irqchip_add_irqfd_notifierseries in the VFIO one. I did the same forirq_set_hookseries but if it is better I can submit it aside.
- above changes made possible to remove
x hw/vfio/platform: add capability to start IRQ propagation" x hw/arm/virt: start VFIO IRQ propagation
- insysbus-fdt.c, useplatform_bus_get_mmio_addrinstead of deprecated mmio[0] property. Thanks to Bharat who pointed this issue out. alsocpu_to_be32was used for size and base (Vikram input) .
- in skeleton misc corrections following Alex
v11->v12:
- add x-mmap property definition, without which the default value of vbasedev.allow_mmap is false, hence preventing the reg space from being mmapped.
v10->v11:
- rebase onto v2.3.0-rc0 (mainly related to PCIe support in virt)
- add dma-coherent property for calxeda midway (fix revealed by removal of kernel-side "vfio: type1: support for ARM SMMUS withVFIO_IOMMU_TYPE1")
- virt modifications to start VFIO IRQ forwarding are now in a separate patch
- rearrange linux header exports (those are still partial exports waiting for definitive 4.1-rc0)
- take into account Alex Bennécomments:
- useg_malloc0_ninstead ofg_malloc0
- use block declarations when possible
- rework readlink returned value treatment
- useg_strlcatin place strncat
- re-arrange mutex locking for multiple IRQ support (user-side handled eventfds)
- useg_snprintfinstead of snprintf
- change the order of functions to avoid pre-declaration inplatform.c
- add flags in VFIOINTp struct to detect whether the IRQ is automasked
- some comment rewriting
v9->v10:
- rebase on "vfio: cleanupvfio_get_deviceerror path, removevfio_populate_device":vfio_populate_deviceno more called invfio_get_devicebut invfio_base_device_init
- update VFIO header according to vfio platform driver v13 (no AMBA)
v8->v9:
- rebase on 2.2.0 and machvirt dynamic sysbus instantiation v10
- v8 1-11 were pulled
- patch files related to forwarding are moved in a seperate series since it depends on kernel series still in RFC.
- introduction of basic VFIO platform device split into 3 patch files to ease the review (hope it will help).
- add an author inplatform.c
- add deallocation invfio_populate_deviceerror case
- add patch file doing the VFIO header sync
- useVFIO_DEVICE_FLAGS_PLATFORMinvfio_populate_device
- renamecalxeda_xgmac.cintocalxeda-xgmac.c
- sysbus-fdt:add_calxeda_midway_xgmac_fdt_nodeg_freein case of errors
- reword of linux-headers patch files
v7->v8:
- rebase on v2.2.0-rc3 and integrate
"Addskip_dumpflag to ignore memory region during dump"
- KVM header evolution with subindex addition inkvm_arch_forwarded_irq
- split [PATCH v7 03/16] hw/vfio/pci: introduce VFIODevice into 4 patches
-vfio_compute_needs_resetdoes not return bool anymore
- add some comments about exposed MMIO region and IRQ in calxeda xgmac device
- vfio_[un]mask_irqindex renamed into vfio_[un]mask_single_irqindex
- rework IRQ startup: former machine init done notifier is replaced by a reset notifier. machine file passes the interrupt controller DeviceState handle (not the platform bus first irq parameter).
- sysbus-fdt:
- move theadd_fdt_node_functionsarray declaration between the device specific code and the generic code to avoid forward declarations of decice specific functions
- renameadd_basic_vfio_fdt_nodeintoadd_calxeda_midway_xgmac_fdt_nodeemphasizing the fact it is xgmac specific
v6->v7:
- fake injection test modality removed
-VFIO_DEVICE_TYPE_PLATFORMonly introduced with VFIO platform
- new helper functions to start VFIO IRQ on machine init done notifier (introduced in hw/vfio/platform: add vfio-platform support and notifier registration invoked in hw/arm/virt: add support for VFIO devices).vfio_start_irq_injectionis replaced byvfio_register_irq_starter.
v5->v6:
- rebase on 2.1rc5 PCI code
- forwarded IRQ first integraton
-vfio_deviceproperty renamed into host property
- split IRQ setup in different functions that match the 3 supported injection techniques (user handled eventfd, irqfd, forwarded IRQ):
removes dynamic switch between injection methods
- introduce fake interrupts as a test modality:
x makes possible to test multiple IRQ user-side handling. x this is a test feature only: enable to trigger a fd as if the
real physical IRQ hit. No virtual IRQ is injected into the guest but handling is simulated so that the state machine can be tested
- user handled eventfd:
x add mutex to protect IRQ state & list manipulation,
x correct misleading comment invfio_intp_interrupt.
x Fix bugs using fake interrupt modality
- irqfd no more advertised in this patchset (handled in [3])
- VFIOPlatformDeviceClass becomes abstract and Calxeda xgmac device and class is re-introduced (as per v4)
- all DPRINTF removed in platform and replaced by trace-points
- corrects compilation with configure --disable-kvm
- simplifies the split forvfio_get_deviceand introduce a unique specialized function namedvfio_populate_device
-group_listrenamed intovfio_group_list
-hw/arm/dyn_sysbus_devtree.ccurrently only support vfio-calxeda-xgmac instantiation. Needs to be specialized for other VFIO devices
- fix 2 bugs indyn_sysbus_devtree(reg_attr index and compat)
v4->v5:
- rebase on v2.1.0 PCI code
- take into account Alex Williamson comments on PCI code rework
- trace updates in vfio_region_write/read
- remove fd from VFIORegion
- get/put ckeanup
- bug fix: bar region's vbasedev field duly initialization
- misc cleanups in platform device
- device tree node generation removed from device and handled inhw/arm/dyn_sysbus_devtree.c
- remove "hw/vfio: add an examplecalxeda_xgmac": with removal of device tree node generation we do not have so many things to implement in that derived device yet. May be re-introduced later on if needed typically for reset/migration.
- no GSI routing table anymore
v3->v4changes (Eric Auger, Alvise Rigo)
- rebase on last VFIO PCI code (v2.1.0-rc0)
- full git history rework to ease PCI code change review
- mv include files in hw/vfio
- DPRINTF reformatting temporarily moved out
- support of VFIO virq (removal of resamplefd handler on user-side)
- integration with sysbus dynamic instantiation framwork
- removal of unrealize and cleanup routines until it is better understood what is really needed
- Support of VFIO for Amba devices should be handled in an inherited device to specialize the device tree generation (clock handle currently missing in framework however)
- "Always use eventfd as notifying mechanism" temporarily moved out
- static instantiation is not mainstream (although it remains possible) note if static instantiation is used, irqfd must be setup in machine file when virtual IRQ is known
- create the GSI routing table on qemu side
v2->v3changes (Alvise Rigo, Eric Auger):
- Following Alex W recommandations, further efforts to factorize the code between PCI:introduction of VFIODevice and VFIORegion as base classes
- unique reset handler for platform and PCI
- cleanup following Kim's comments
- multiple IRQ support mechanics should be in place although not tested
- Better handling of MMIO multiple regions
- New features and fixes by Alvise (multiple compat string, exec flag, force eventfd usage, amba device tree support)
- irqfd support
v1->v2changes (Kim Phillips, Eric Auger):
- IRQ initial support (legacy mode where eventfds are handled on user side)
- hacked dynamic instantiationv1 (Kim Phillips):
- initial split between PCI and platform
- MMIO support only
- static instantiationBest Regards
Eric
Eric Auger (6):
hw/arm/sysbus-fdt: enable vfio-calxeda-xgmac dynamic instantiation
kvm: rename kvm_irqchip_[add,remove]_irqfd_notifier with gsi suffixkvm-all.c: add qemu_irq/gsi hash table and utility routines
intc:arm_gic_kvm: set the qemu_irq/gsi mapping
sysbus: addirq_routing_notifier
hw/vfio/platform: add irqfd supporthw/arm/sysbus-fdt.c | 72 +++++++++++++++++++++++++++ hw/core/sysbus.c | 6 +++ hw/intc/arm_gic_kvm.c | 6 +++ hw/s390x/virtio-ccw.c | 8 +-- hw/vfio/pci.c | 6 +-- hw/vfio/platform.c | 107 ++++++++++++++++++++++++++++++++++++++++ hw/virtio/virtio-pci.c | 4 +- include/hw/sysbus.h | 1 + include/hw/vfio/vfio-platform.h | 2 + include/sysemu/kvm.h | 11 ++++- kvm-all.c | 44 +++++++++++++++-- kvm-stub.c | 7 +-- trace-events | 1 + 13 files changed, 258 insertions(+), 17 deletions(-)
Quoted 7 lines · Eric Auger
Now that the platform-bus is instantiated in virt, the ARM sysbus-fdt infrastructure, the VFIO platform device and its first Calxeda xgmac specialization also are on the master, let's complete the work by
- enabling the instantiation of the VFIO Calxeda xgmac device
- enabling irqfd usage, which is vital for performance.
Quoted 7 lines
So the resend only relates to patches 4-9 of the v16 series.
http://git.linaro.org/people/eric.auger/qemu.git
(branchvfio_integ_v16_partial_resend)Tested by Vikram Sethi on a SATA device featuring 2 MMIO regions. Code also exercised successfully by VOSYS for AMBA derivative series.
I think I've now acked everything that's needed from me?
Paolo
Quoted 228 lines
Thanks
Eric
v15 -> v16:
- no code change
- add Vikram's T-b
- rebase on 0915aed
- remove linux header update patch filev14 -> v15:
- add Peter R-b on sysbus: addirq_routing_notifier
- correctg_malloc0_nusage in skeleton
- correct return values of reset related functions
- include Cornelia's patch for header updatev13 -> v14:
- remove v13 9, 10, 11 patch files and replace them by a single patch file "sysbus: addirq_routing_notifier".
- in skeleton, fix ENAMETOOLONG sign
- remove VFIOINTp virtualID in "add irq assignment" patch file
- removedtrace_vfio_platform_start_eventfdv12 -> v13:
- header update but same update was already sent by Cornelia
- Rework VFIO signaling & irqfd setup: restored 2-step setup featuring eventfd setup on realize and then irqfd setup on irq binding.
- irqfd setup now useskvm_irqchip_add_irqfd_notifierand sysbusirq_set_hookoverride. This leads to the introduction of 6 patch files enabling those 2 features. Paolo advised to introducekvm_irqchip_add_irqfd_notifierseries in the VFIO one. I did the same forirq_set_hookseries but if it is better I can submit it aside.
- above changes made possible to remove
x hw/vfio/platform: add capability to start IRQ propagation" x hw/arm/virt: start VFIO IRQ propagation
- insysbus-fdt.c, useplatform_bus_get_mmio_addrinstead of deprecated mmio[0] property. Thanks to Bharat who pointed this issue out. alsocpu_to_be32was used for size and base (Vikram input) .
- in skeleton misc corrections following Alex
v11->v12:
- add x-mmap property definition, without which the default value of vbasedev.allow_mmap is false, hence preventing the reg space from being mmapped.
v10->v11:
- rebase onto v2.3.0-rc0 (mainly related to PCIe support in virt)
- add dma-coherent property for calxeda midway (fix revealed by removal of kernel-side "vfio: type1: support for ARM SMMUS withVFIO_IOMMU_TYPE1")
- virt modifications to start VFIO IRQ forwarding are now in a separate patch
- rearrange linux header exports (those are still partial exports waiting for definitive 4.1-rc0)
- take into account Alex Bennécomments:
- useg_malloc0_ninstead ofg_malloc0
- use block declarations when possible
- rework readlink returned value treatment
- useg_strlcatin place strncat
- re-arrange mutex locking for multiple IRQ support (user-side handled eventfds)
- useg_snprintfinstead of snprintf
- change the order of functions to avoid pre-declaration inplatform.c
- add flags in VFIOINTp struct to detect whether the IRQ is automasked
- some comment rewriting
v9->v10:
- rebase on "vfio: cleanupvfio_get_deviceerror path, removevfio_populate_device":vfio_populate_deviceno more called invfio_get_devicebut invfio_base_device_init
- update VFIO header according to vfio platform driver v13 (no AMBA)
v8->v9:
- rebase on 2.2.0 and machvirt dynamic sysbus instantiation v10
- v8 1-11 were pulled
- patch files related to forwarding are moved in a seperate series since it depends on kernel series still in RFC.
- introduction of basic VFIO platform device split into 3 patch files to ease the review (hope it will help).
- add an author inplatform.c
- add deallocation invfio_populate_deviceerror case
- add patch file doing the VFIO header sync
- useVFIO_DEVICE_FLAGS_PLATFORMinvfio_populate_device
- renamecalxeda_xgmac.cintocalxeda-xgmac.c
- sysbus-fdt:add_calxeda_midway_xgmac_fdt_nodeg_freein case of errors
- reword of linux-headers patch files
v7->v8:
- rebase on v2.2.0-rc3 and integrate
"Addskip_dumpflag to ignore memory region during dump"
- KVM header evolution with subindex addition inkvm_arch_forwarded_irq
- split [PATCH v7 03/16] hw/vfio/pci: introduce VFIODevice into 4 patches
-vfio_compute_needs_resetdoes not return bool anymore
- add some comments about exposed MMIO region and IRQ in calxeda xgmac device
- vfio_[un]mask_irqindex renamed into vfio_[un]mask_single_irqindex
- rework IRQ startup: former machine init done notifier is replaced by a reset notifier. machine file passes the interrupt controller DeviceState handle (not the platform bus first irq parameter).
- sysbus-fdt:
- move theadd_fdt_node_functionsarray declaration between the device specific code and the generic code to avoid forward declarations of decice specific functions
- renameadd_basic_vfio_fdt_nodeintoadd_calxeda_midway_xgmac_fdt_nodeemphasizing the fact it is xgmac specific
v6->v7:
- fake injection test modality removed
-VFIO_DEVICE_TYPE_PLATFORMonly introduced with VFIO platform
- new helper functions to start VFIO IRQ on machine init done notifier (introduced in hw/vfio/platform: add vfio-platform support and notifier registration invoked in hw/arm/virt: add support for VFIO devices).vfio_start_irq_injectionis replaced byvfio_register_irq_starter.
v5->v6:
- rebase on 2.1rc5 PCI code
- forwarded IRQ first integraton
-vfio_deviceproperty renamed into host property
- split IRQ setup in different functions that match the 3 supported injection techniques (user handled eventfd, irqfd, forwarded IRQ):
removes dynamic switch between injection methods
- introduce fake interrupts as a test modality:
x makes possible to test multiple IRQ user-side handling. x this is a test feature only: enable to trigger a fd as if the
real physical IRQ hit. No virtual IRQ is injected into the guest but handling is simulated so that the state machine can be tested
- user handled eventfd:
x add mutex to protect IRQ state & list manipulation,
x correct misleading comment invfio_intp_interrupt.
x Fix bugs using fake interrupt modality
- irqfd no more advertised in this patchset (handled in [3])
- VFIOPlatformDeviceClass becomes abstract and Calxeda xgmac device and class is re-introduced (as per v4)
- all DPRINTF removed in platform and replaced by trace-points
- corrects compilation with configure --disable-kvm
- simplifies the split forvfio_get_deviceand introduce a unique specialized function namedvfio_populate_device
-group_listrenamed intovfio_group_list
-hw/arm/dyn_sysbus_devtree.ccurrently only support vfio-calxeda-xgmac instantiation. Needs to be specialized for other VFIO devices
- fix 2 bugs indyn_sysbus_devtree(reg_attr index and compat)
v4->v5:
- rebase on v2.1.0 PCI code
- take into account Alex Williamson comments on PCI code rework
- trace updates in vfio_region_write/read
- remove fd from VFIORegion
- get/put ckeanup
- bug fix: bar region's vbasedev field duly initialization
- misc cleanups in platform device
- device tree node generation removed from device and handled inhw/arm/dyn_sysbus_devtree.c
- remove "hw/vfio: add an examplecalxeda_xgmac": with removal of device tree node generation we do not have so many things to implement in that derived device yet. May be re-introduced later on if needed typically for reset/migration.
- no GSI routing table anymore
v3->v4changes (Eric Auger, Alvise Rigo)
- rebase on last VFIO PCI code (v2.1.0-rc0)
- full git history rework to ease PCI code change review
- mv include files in hw/vfio
- DPRINTF reformatting temporarily moved out
- support of VFIO virq (removal of resamplefd handler on user-side)
- integration with sysbus dynamic instantiation framwork
- removal of unrealize and cleanup routines until it is better understood what is really needed
- Support of VFIO for Amba devices should be handled in an inherited device to specialize the device tree generation (clock handle currently missing in framework however)
- "Always use eventfd as notifying mechanism" temporarily moved out
- static instantiation is not mainstream (although it remains possible) note if static instantiation is used, irqfd must be setup in machine file when virtual IRQ is known
- create the GSI routing table on qemu side
v2->v3changes (Alvise Rigo, Eric Auger):
- Following Alex W recommandations, further efforts to factorize the code between PCI:introduction of VFIODevice and VFIORegion as base classes
- unique reset handler for platform and PCI
- cleanup following Kim's comments
- multiple IRQ support mechanics should be in place although not tested
- Better handling of MMIO multiple regions
- New features and fixes by Alvise (multiple compat string, exec flag, force eventfd usage, amba device tree support)
- irqfd support
v1->v2changes (Kim Phillips, Eric Auger):
- IRQ initial support (legacy mode where eventfds are handled on user side)
- hacked dynamic instantiationv1 (Kim Phillips):
- initial split between PCI and platform
- MMIO support only
- static instantiationBest Regards
Eric
Eric Auger (6):
hw/arm/sysbus-fdt: enable vfio-calxeda-xgmac dynamic instantiation
kvm: rename kvm_irqchip_[add,remove]_irqfd_notifier with gsi suffixkvm-all.c: add qemu_irq/gsi hash table and utility routines
intc:arm_gic_kvm: set the qemu_irq/gsi mapping
sysbus: addirq_routing_notifier
hw/vfio/platform: add irqfd supporthw/arm/sysbus-fdt.c | 72 +++++++++++++++++++++++++++ hw/core/sysbus.c | 6 +++ hw/intc/arm_gic_kvm.c | 6 +++ hw/s390x/virtio-ccw.c | 8 +-- hw/vfio/pci.c | 6 +-- hw/vfio/platform.c | 107 ++++++++++++++++++++++++++++++++++++++++ hw/virtio/virtio-pci.c | 4 +- include/hw/sysbus.h | 1 + include/hw/vfio/vfio-platform.h | 2 + include/sysemu/kvm.h | 11 ++++- kvm-all.c | 44 +++++++++++++++-- kvm-stub.c | 7 +-- trace-events | 1 + 13 files changed, 258 insertions(+), 17 deletions(-)
Hi Paolo
Quoted 12 lines · Paolo Bonzini
Quoted 9 lines · Eric Auger
Now that the platform-bus is instantiated in virt, the ARM sysbus-fdt infrastructure, the VFIO platform device and its first Calxeda xgmac specialization also are on the master, let's complete the work by
- enabling the instantiation of the VFIO Calxeda xgmac device
- enabling irqfd usage, which is vital for performance.So the resend only relates to patches 4-9 of the v16 series.
Quoted 7 lines
Quoted 5 lines
http://git.linaro.org/people/eric.auger/qemu.git
(branchvfio_integ_v16_partial_resend)Tested by Vikram Sethi on a SATA device featuring 2 MMIO regions. Code also exercised successfully by VOSYS for AMBA derivative series.
I think I've now acked everything that's needed from me?
Thanks. what about the
[RESEND PATCH v16 5/6] sysbus: add irq_routing_notifier?
Is that one for Peter C. or anyone else?
Thans
Eric
Quoted 231 lines
Paolo
Quoted 228 lines
Thanks
Eric
v15 -> v16:
- no code change
- add Vikram's T-b
- rebase on 0915aed
- remove linux header update patch filev14 -> v15:
- add Peter R-b on sysbus: addirq_routing_notifier
- correctg_malloc0_nusage in skeleton
- correct return values of reset related functions
- include Cornelia's patch for header updatev13 -> v14:
- remove v13 9, 10, 11 patch files and replace them by a single patch file "sysbus: addirq_routing_notifier".
- in skeleton, fix ENAMETOOLONG sign
- remove VFIOINTp virtualID in "add irq assignment" patch file
- removedtrace_vfio_platform_start_eventfdv12 -> v13:
- header update but same update was already sent by Cornelia
- Rework VFIO signaling & irqfd setup: restored 2-step setup featuring eventfd setup on realize and then irqfd setup on irq binding.
- irqfd setup now useskvm_irqchip_add_irqfd_notifierand sysbusirq_set_hookoverride. This leads to the introduction of 6 patch files enabling those 2 features. Paolo advised to introducekvm_irqchip_add_irqfd_notifierseries in the VFIO one. I did the same forirq_set_hookseries but if it is better I can submit it aside.
- above changes made possible to remove
x hw/vfio/platform: add capability to start IRQ propagation" x hw/arm/virt: start VFIO IRQ propagation
- insysbus-fdt.c, useplatform_bus_get_mmio_addrinstead of deprecated mmio[0] property. Thanks to Bharat who pointed this issue out. alsocpu_to_be32was used for size and base (Vikram input) .
- in skeleton misc corrections following Alex
v11->v12:
- add x-mmap property definition, without which the default value of vbasedev.allow_mmap is false, hence preventing the reg space from being mmapped.
v10->v11:
- rebase onto v2.3.0-rc0 (mainly related to PCIe support in virt)
- add dma-coherent property for calxeda midway (fix revealed by removal of kernel-side "vfio: type1: support for ARM SMMUS withVFIO_IOMMU_TYPE1")
- virt modifications to start VFIO IRQ forwarding are now in a separate patch
- rearrange linux header exports (those are still partial exports waiting for definitive 4.1-rc0)
- take into account Alex Bennécomments:
- useg_malloc0_ninstead ofg_malloc0
- use block declarations when possible
- rework readlink returned value treatment
- useg_strlcatin place strncat
- re-arrange mutex locking for multiple IRQ support (user-side handled eventfds)
- useg_snprintfinstead of snprintf
- change the order of functions to avoid pre-declaration inplatform.c
- add flags in VFIOINTp struct to detect whether the IRQ is automasked
- some comment rewriting
v9->v10:
- rebase on "vfio: cleanupvfio_get_deviceerror path, removevfio_populate_device":vfio_populate_deviceno more called invfio_get_devicebut invfio_base_device_init
- update VFIO header according to vfio platform driver v13 (no AMBA)
v8->v9:
- rebase on 2.2.0 and machvirt dynamic sysbus instantiation v10
- v8 1-11 were pulled
- patch files related to forwarding are moved in a seperate series since it depends on kernel series still in RFC.
- introduction of basic VFIO platform device split into 3 patch files to ease the review (hope it will help).
- add an author inplatform.c
- add deallocation invfio_populate_deviceerror case
- add patch file doing the VFIO header sync
- useVFIO_DEVICE_FLAGS_PLATFORMinvfio_populate_device
- renamecalxeda_xgmac.cintocalxeda-xgmac.c
- sysbus-fdt:add_calxeda_midway_xgmac_fdt_nodeg_freein case of errors
- reword of linux-headers patch files
v7->v8:
- rebase on v2.2.0-rc3 and integrate
"Addskip_dumpflag to ignore memory region during dump"
- KVM header evolution with subindex addition inkvm_arch_forwarded_irq
- split [PATCH v7 03/16] hw/vfio/pci: introduce VFIODevice into 4 patches
-vfio_compute_needs_resetdoes not return bool anymore
- add some comments about exposed MMIO region and IRQ in calxeda xgmac device
- vfio_[un]mask_irqindex renamed into vfio_[un]mask_single_irqindex
- rework IRQ startup: former machine init done notifier is replaced by a reset notifier. machine file passes the interrupt controller DeviceState handle (not the platform bus first irq parameter).
- sysbus-fdt:
- move theadd_fdt_node_functionsarray declaration between the device specific code and the generic code to avoid forward declarations of decice specific functions
- renameadd_basic_vfio_fdt_nodeintoadd_calxeda_midway_xgmac_fdt_nodeemphasizing the fact it is xgmac specific
v6->v7:
- fake injection test modality removed
-VFIO_DEVICE_TYPE_PLATFORMonly introduced with VFIO platform
- new helper functions to start VFIO IRQ on machine init done notifier (introduced in hw/vfio/platform: add vfio-platform support and notifier registration invoked in hw/arm/virt: add support for VFIO devices).vfio_start_irq_injectionis replaced byvfio_register_irq_starter.
v5->v6:
- rebase on 2.1rc5 PCI code
- forwarded IRQ first integraton
-vfio_deviceproperty renamed into host property
- split IRQ setup in different functions that match the 3 supported injection techniques (user handled eventfd, irqfd, forwarded IRQ):
removes dynamic switch between injection methods
- introduce fake interrupts as a test modality:
x makes possible to test multiple IRQ user-side handling. x this is a test feature only: enable to trigger a fd as if the
real physical IRQ hit. No virtual IRQ is injected into the guest but handling is simulated so that the state machine can be tested
- user handled eventfd:
x add mutex to protect IRQ state & list manipulation,
x correct misleading comment invfio_intp_interrupt.
x Fix bugs using fake interrupt modality
- irqfd no more advertised in this patchset (handled in [3])
- VFIOPlatformDeviceClass becomes abstract and Calxeda xgmac device and class is re-introduced (as per v4)
- all DPRINTF removed in platform and replaced by trace-points
- corrects compilation with configure --disable-kvm
- simplifies the split forvfio_get_deviceand introduce a unique specialized function namedvfio_populate_device
-group_listrenamed intovfio_group_list
-hw/arm/dyn_sysbus_devtree.ccurrently only support vfio-calxeda-xgmac instantiation. Needs to be specialized for other VFIO devices
- fix 2 bugs indyn_sysbus_devtree(reg_attr index and compat)
v4->v5:
- rebase on v2.1.0 PCI code
- take into account Alex Williamson comments on PCI code rework
- trace updates in vfio_region_write/read
- remove fd from VFIORegion
- get/put ckeanup
- bug fix: bar region's vbasedev field duly initialization
- misc cleanups in platform device
- device tree node generation removed from device and handled inhw/arm/dyn_sysbus_devtree.c
- remove "hw/vfio: add an examplecalxeda_xgmac": with removal of device tree node generation we do not have so many things to implement in that derived device yet. May be re-introduced later on if needed typically for reset/migration.
- no GSI routing table anymore
v3->v4changes (Eric Auger, Alvise Rigo)
- rebase on last VFIO PCI code (v2.1.0-rc0)
- full git history rework to ease PCI code change review
- mv include files in hw/vfio
- DPRINTF reformatting temporarily moved out
- support of VFIO virq (removal of resamplefd handler on user-side)
- integration with sysbus dynamic instantiation framwork
- removal of unrealize and cleanup routines until it is better understood what is really needed
- Support of VFIO for Amba devices should be handled in an inherited device to specialize the device tree generation (clock handle currently missing in framework however)
- "Always use eventfd as notifying mechanism" temporarily moved out
- static instantiation is not mainstream (although it remains possible) note if static instantiation is used, irqfd must be setup in machine file when virtual IRQ is known
- create the GSI routing table on qemu side
v2->v3changes (Alvise Rigo, Eric Auger):
- Following Alex W recommandations, further efforts to factorize the code between PCI:introduction of VFIODevice and VFIORegion as base classes
- unique reset handler for platform and PCI
- cleanup following Kim's comments
- multiple IRQ support mechanics should be in place although not tested
- Better handling of MMIO multiple regions
- New features and fixes by Alvise (multiple compat string, exec flag, force eventfd usage, amba device tree support)
- irqfd support
v1->v2changes (Kim Phillips, Eric Auger):
- IRQ initial support (legacy mode where eventfds are handled on user side)
- hacked dynamic instantiationv1 (Kim Phillips):
- initial split between PCI and platform
- MMIO support only
- static instantiationBest Regards
Eric
Eric Auger (6):
hw/arm/sysbus-fdt: enable vfio-calxeda-xgmac dynamic instantiation
kvm: rename kvm_irqchip_[add,remove]_irqfd_notifier with gsi suffixkvm-all.c: add qemu_irq/gsi hash table and utility routines
intc:arm_gic_kvm: set the qemu_irq/gsi mapping
sysbus: addirq_routing_notifier
hw/vfio/platform: add irqfd supporthw/arm/sysbus-fdt.c | 72 +++++++++++++++++++++++++++ hw/core/sysbus.c | 6 +++ hw/intc/arm_gic_kvm.c | 6 +++ hw/s390x/virtio-ccw.c | 8 +-- hw/vfio/pci.c | 6 +-- hw/vfio/platform.c | 107 ++++++++++++++++++++++++++++++++++++++++ hw/virtio/virtio-pci.c | 4 +- include/hw/sysbus.h | 1 + include/hw/vfio/vfio-platform.h | 2 + include/sysemu/kvm.h | 11 ++++- kvm-all.c | 44 +++++++++++++++-- kvm-stub.c | 7 +-- trace-events | 1 + 13 files changed, 258 insertions(+), 17 deletions(-)
Quoted 6 lines · Eric Auger
Quoted 1 line
I think I've now acked everything that's needed from me?
Thanks. what about the
[RESEND PATCH v16 5/6] sysbus: addirq_routing_notifier?Is that one for Peter C. or anyone else?
There is no real maintainer, so I think Peter's Reviewed-by is enough.
Paolo