[V3,12/22] vfio-pci: cpr part 1

Enable vfio-pci devices to be saved and restored across an exec restart
of qemu.

At vfio creation time, save the value of vfio container, group, and device
descriptors in the environment.

In cprsave, suspend the use of virtual addresses in DMA mappings with
VFIO_DMA_UNMAP_FLAG_VADDR, because guest ram will be remapped at a
different VA after exec.  DMA to already-mapped pages continues.  Save
the msi message area as part of vfio-pci vmstate, save the interrupt and
notifier eventfd's in the environment, and clear the close-on-exec flag
for the vfio descriptors.  The flag is not cleared earlier because the
descriptors should not persist across miscellaneous fork and exec calls
that may be performed during normal operation.

On qemu restart, vfio_realize() finds the descriptor env vars, uses
the descriptors, and notes that the device is being reused.  Device and
iommu state is already configured, so operations in vfio_realize that
would modify the configuration are skipped for a reused device, including
vfio ioctl's and writes to PCI configuration space.  The result is that
vfio_realize constructs qemu data structures that reflect the current
state of the device.  However, the reconstruction is not complete until
cprload is called. cprload loads the msi data and finds eventfds in the
environment.  It rebuilds vector data structures and attaches the
interrupts to the new KVM instance.  cprload then walks the flattened
ranges of the vfio_address_spaces and calls VFIO_DMA_MAP_FLAG_VADDR to
inform the kernel of the new VA's.  Lastly, it starts the VM and suppresses
vfio device reset.

This functionality is delivered by 2 patches for clarity.  Part 2 adds
eventfd and vector support.

Signed-off-by: Steve Sistare <steven.sistare@oracle.com>
---
 hw/pci/msi.c                  |   4 ++
 hw/pci/pci.c                  |   4 ++
 hw/vfio/common.c              |  59 ++++++++++++++++++-
 hw/vfio/cpr.c                 | 131 ++++++++++++++++++++++++++++++++++++++++++
 hw/vfio/meson.build           |   1 +
 hw/vfio/pci.c                 |  65 +++++++++++++++++++--
 hw/vfio/trace-events          |   1 +
 include/hw/pci/pci.h          |   1 +
 include/hw/vfio/vfio-common.h |   5 ++
 linux-headers/linux/vfio.h    |  27 +++++++++
 migration/cpr.c               |   7 +++
 11 files changed, 298 insertions(+), 7 deletions(-)
 create mode 100644 hw/vfio/cpr.c

On Fri,  7 May 2021 05:25:10 -0700
Steve Sistare <steven.sistare@oracle.com> wrote:

> Enable vfio-pci devices to be saved and restored across an exec restart
> of qemu.
> 
> At vfio creation time, save the value of vfio container, group, and device
> descriptors in the environment.
> 
> In cprsave, suspend the use of virtual addresses in DMA mappings with
> VFIO_DMA_UNMAP_FLAG_VADDR, because guest ram will be remapped at a
> different VA after exec.  DMA to already-mapped pages continues.  Save
> the msi message area as part of vfio-pci vmstate, save the interrupt and
> notifier eventfd's in the environment, and clear the close-on-exec flag
> for the vfio descriptors.  The flag is not cleared earlier because the
> descriptors should not persist across miscellaneous fork and exec calls
> that may be performed during normal operation.
> 
> On qemu restart, vfio_realize() finds the descriptor env vars, uses
> the descriptors, and notes that the device is being reused.  Device and
> iommu state is already configured, so operations in vfio_realize that
> would modify the configuration are skipped for a reused device, including
> vfio ioctl's and writes to PCI configuration space.  The result is that
> vfio_realize constructs qemu data structures that reflect the current
> state of the device.  However, the reconstruction is not complete until
> cprload is called. cprload loads the msi data and finds eventfds in the
> environment.  It rebuilds vector data structures and attaches the
> interrupts to the new KVM instance.  cprload then walks the flattened
> ranges of the vfio_address_spaces and calls VFIO_DMA_MAP_FLAG_VADDR to
> inform the kernel of the new VA's.  Lastly, it starts the VM and suppresses
> vfio device reset.
> 
> This functionality is delivered by 2 patches for clarity.  Part 2 adds
> eventfd and vector support.
> 
> Signed-off-by: Steve Sistare <steven.sistare@oracle.com>
> ---
>  hw/pci/msi.c                  |   4 ++
>  hw/pci/pci.c                  |   4 ++
>  hw/vfio/common.c              |  59 ++++++++++++++++++-
>  hw/vfio/cpr.c                 | 131 ++++++++++++++++++++++++++++++++++++++++++
>  hw/vfio/meson.build           |   1 +
>  hw/vfio/pci.c                 |  65 +++++++++++++++++++--
>  hw/vfio/trace-events          |   1 +
>  include/hw/pci/pci.h          |   1 +
>  include/hw/vfio/vfio-common.h |   5 ++
>  linux-headers/linux/vfio.h    |  27 +++++++++
>  migration/cpr.c               |   7 +++
>  11 files changed, 298 insertions(+), 7 deletions(-)
>  create mode 100644 hw/vfio/cpr.c
> 
> diff --git a/hw/pci/msi.c b/hw/pci/msi.c
> index 47d2b0f..39de6a7 100644
> --- a/hw/pci/msi.c
> +++ b/hw/pci/msi.c
> @@ -225,6 +225,10 @@ int msi_init(struct PCIDevice *dev, uint8_t offset,
>      dev->msi_cap = config_offset;
>      dev->cap_present |= QEMU_PCI_CAP_MSI;
>  
> +    if (dev->reused) {
> +        return 0;
> +    }
> +
>      pci_set_word(dev->config + msi_flags_off(dev), flags);
>      pci_set_word(dev->wmask + msi_flags_off(dev),
>                   PCI_MSI_FLAGS_QSIZE | PCI_MSI_FLAGS_ENABLE);
> diff --git a/hw/pci/pci.c b/hw/pci/pci.c
> index e08d981..27019ca 100644
> --- a/hw/pci/pci.c
> +++ b/hw/pci/pci.c
> @@ -308,6 +308,10 @@ static void pci_do_device_reset(PCIDevice *dev)
>  {
>      int r;
>  
> +    if (dev->reused) {
> +        return;
> +    }
> +
>      pci_device_deassert_intx(dev);
>      assert(dev->irq_state == 0);
>  
> diff --git a/hw/vfio/common.c b/hw/vfio/common.c
> index 9220e64..00d07b2 100644
> --- a/hw/vfio/common.c
> +++ b/hw/vfio/common.c
> @@ -31,6 +31,7 @@
>  #include "exec/memory.h"
>  #include "exec/ram_addr.h"
>  #include "hw/hw.h"
> +#include "qemu/env.h"
>  #include "qemu/error-report.h"
>  #include "qemu/main-loop.h"
>  #include "qemu/range.h"
> @@ -440,6 +441,10 @@ static int vfio_dma_unmap(VFIOContainer *container,
>          return vfio_dma_unmap_bitmap(container, iova, size, iotlb);
>      }
>  
> +    if (container->reused) {
> +        return 0;
> +    }
> +
>      while (ioctl(container->fd, VFIO_IOMMU_UNMAP_DMA, &unmap)) {
>          /*
>           * The type1 backend has an off-by-one bug in the kernel (71a7d3d78e3c
> @@ -463,6 +468,11 @@ static int vfio_dma_unmap(VFIOContainer *container,
>          return -errno;
>      }
>  
> +    if (unmap.size != size) {
> +        warn_report("VFIO_UNMAP_DMA(0x%lx, 0x%lx) only unmaps 0x%llx",
> +                     iova, size, unmap.size);
> +    }
> +
>      return 0;
>  }
>  
> @@ -477,6 +487,10 @@ static int vfio_dma_map(VFIOContainer *container, hwaddr iova,
>          .size = size,
>      };
>  
> +    if (container->reused) {
> +        return 0;
> +    }
> +
>      if (!readonly) {
>          map.flags |= VFIO_DMA_MAP_FLAG_WRITE;
>      }
> @@ -1603,6 +1617,10 @@ static int vfio_init_container(VFIOContainer *container, int group_fd,
>      if (iommu_type < 0) {
>          return iommu_type;
>      }
> +    if (container->reused) {
> +        container->iommu_type = iommu_type;
> +        return 0;
> +    }
>  
>      ret = ioctl(group_fd, VFIO_GROUP_SET_CONTAINER, &container->fd);
>      if (ret) {
> @@ -1703,6 +1721,8 @@ static int vfio_connect_container(VFIOGroup *group, AddressSpace *as,
>  {
>      VFIOContainer *container;
>      int ret, fd;
> +    bool reused;
> +    char name[40];
>      VFIOAddressSpace *space;
>  
>      space = vfio_get_address_space(as);
> @@ -1739,16 +1759,29 @@ static int vfio_connect_container(VFIOGroup *group, AddressSpace *as,
>          return ret;
>      }
>  
> +    snprintf(name, sizeof(name), "vfio_container_%d", group->groupid);

For more clarity, maybe "vfio_container_for_group_%d"?

> +    fd = getenv_fd(name);
> +    reused = (fd >= 0);
> +
>      QLIST_FOREACH(container, &space->containers, next) {
> +        if (fd >= 0 && container->fd == fd) {

Test @reused rather than @fd?  I'm not sure the first half of this test
is even needed though, <0 should never match container->fd, right?

> +            group->container = container;
> +            QLIST_INSERT_HEAD(&container->group_list, group, container_next);
> +            return 0;
> +        }

This looks unnecessarily sensitive to the order of containers in the
list, if the fd doesn't match above we try to set a new container below?
It seems like you only want to create a new container object if none of
the existing ones match.

There's also a lot of duplication that seems like it could be combined

if (container->fd == fd || (!reused && !ioctl(...)) {

>          if (!ioctl(group->fd, VFIO_GROUP_SET_CONTAINER, &container->fd)) {
>              group->container = container;
>              QLIST_INSERT_HEAD(&container->group_list, group, container_next);
>              vfio_kvm_device_add_group(group);

Why is this kvm device setup missing in the reuse case?


if (!reused) {
> +            setenv_fd(name, container->fd);
}

>              return 0;
>          }
>      }
>  
> -    fd = qemu_open_old("/dev/vfio/vfio", O_RDWR);
> +    if (fd < 0) {

if (!reused)?

> +        fd = qemu_open_old("/dev/vfio/vfio", O_RDWR);
> +    }
> +
>      if (fd < 0) {
>          error_setg_errno(errp, errno, "failed to open /dev/vfio/vfio");
>          ret = -errno;
> @@ -1766,6 +1799,7 @@ static int vfio_connect_container(VFIOGroup *group, AddressSpace *as,
>      container = g_malloc0(sizeof(*container));
>      container->space = space;
>      container->fd = fd;
> +    container->reused = reused;
>      container->error = NULL;
>      container->dirty_pages_supported = false;
>      QLIST_INIT(&container->giommu_list);
> @@ -1893,6 +1927,7 @@ static int vfio_connect_container(VFIOGroup *group, AddressSpace *as,
>      }
>  
>      container->initialized = true;
> +    setenv_fd(name, fd);

Maybe we don't need the test around the previous setenv_fd if we can
overwrite existing env values, which would seem to be the case for a
restart here.

>  
>      return 0;
>  listener_release_exit:
> @@ -1920,6 +1955,7 @@ static void vfio_disconnect_container(VFIOGroup *group)
>  
>      QLIST_REMOVE(group, container_next);
>      group->container = NULL;
> +    unsetenv_fdv("vfio_container_%d", group->groupid);
>  
>      /*
>       * Explicitly release the listener first before unset container,
> @@ -1978,7 +2014,12 @@ VFIOGroup *vfio_get_group(int groupid, AddressSpace *as, Error **errp)
>      group = g_malloc0(sizeof(*group));
>  
>      snprintf(path, sizeof(path), "/dev/vfio/%d", groupid);
> -    group->fd = qemu_open_old(path, O_RDWR);
> +
> +    group->fd = getenv_fd(path);
> +    if (group->fd < 0) {
> +        group->fd = qemu_open_old(path, O_RDWR);
> +    }
> +
>      if (group->fd < 0) {
>          error_setg_errno(errp, errno, "failed to open %s", path);
>          goto free_group_exit;
> @@ -2012,6 +2053,8 @@ VFIOGroup *vfio_get_group(int groupid, AddressSpace *as, Error **errp)
>  
>      QLIST_INSERT_HEAD(&vfio_group_list, group, next);
>  
> +    setenv_fd(path, group->fd);
> +
>      return group;
>  
>  close_fd_exit:
> @@ -2036,6 +2079,7 @@ void vfio_put_group(VFIOGroup *group)
>      vfio_disconnect_container(group);
>      QLIST_REMOVE(group, next);
>      trace_vfio_put_group(group->fd);
> +    unsetenv_fdv("/dev/vfio/%d", group->groupid);
>      close(group->fd);
>      g_free(group);
>  
> @@ -2049,8 +2093,14 @@ int vfio_get_device(VFIOGroup *group, const char *name,
>  {
>      struct vfio_device_info dev_info = { .argsz = sizeof(dev_info) };
>      int ret, fd;
> +    bool reused;
> +
> +    fd = getenv_fd(name);
> +    reused = (fd >= 0);
> +    if (fd < 0) {

if (!reused) ?

> +        fd = ioctl(group->fd, VFIO_GROUP_GET_DEVICE_FD, name);
> +    }
>  
> -    fd = ioctl(group->fd, VFIO_GROUP_GET_DEVICE_FD, name);
>      if (fd < 0) {
>          error_setg_errno(errp, errno, "error getting device from group %d",
>                           group->groupid);
> @@ -2095,6 +2145,8 @@ int vfio_get_device(VFIOGroup *group, const char *name,
>      vbasedev->num_irqs = dev_info.num_irqs;
>      vbasedev->num_regions = dev_info.num_regions;
>      vbasedev->flags = dev_info.flags;
> +    vbasedev->reused = reused;
> +    setenv_fd(name, fd);
>  
>      trace_vfio_get_device(name, dev_info.flags, dev_info.num_regions,
>                            dev_info.num_irqs);
> @@ -2111,6 +2163,7 @@ void vfio_put_base_device(VFIODevice *vbasedev)
>      QLIST_REMOVE(vbasedev, next);
>      vbasedev->group = NULL;
>      trace_vfio_put_base_device(vbasedev->fd);
> +    unsetenv_fd(vbasedev->name);
>      close(vbasedev->fd);
>  }
>  
> diff --git a/hw/vfio/cpr.c b/hw/vfio/cpr.c
> new file mode 100644
> index 0000000..c5ad9f2
> --- /dev/null
> +++ b/hw/vfio/cpr.c
> @@ -0,0 +1,131 @@
> +/*
> + * Copyright (c) 2021 Oracle and/or its affiliates.
> + *
> + * This work is licensed under the terms of the GNU GPL, version 2.
> + * See the COPYING file in the top-level directory.
> + */
> +
> +#include "qemu/osdep.h"
> +#include <sys/ioctl.h>
> +#include <linux/vfio.h>
> +#include "hw/vfio/vfio-common.h"
> +#include "sysemu/kvm.h"
> +#include "qapi/error.h"
> +#include "trace.h"
> +
> +static int
> +vfio_dma_unmap_vaddr_all(VFIOContainer *container, Error **errp)
> +{
> +    struct vfio_iommu_type1_dma_unmap unmap = {
> +        .argsz = sizeof(unmap),
> +        .flags = VFIO_DMA_UNMAP_FLAG_VADDR | VFIO_DMA_UNMAP_FLAG_ALL,
> +        .iova = 0,
> +        .size = 0,
> +    };
> +    if (ioctl(container->fd, VFIO_IOMMU_UNMAP_DMA, &unmap)) {
> +        error_setg_errno(errp, errno, "vfio_dma_unmap_vaddr_all");
> +        return -errno;
> +    }
> +    return 0;
> +}
> +
> +static int vfio_dma_map_vaddr(VFIOContainer *container, hwaddr iova,
> +                              ram_addr_t size, void *vaddr,
> +                              Error **errp)
> +{
> +    struct vfio_iommu_type1_dma_map map = {
> +        .argsz = sizeof(map),
> +        .flags = VFIO_DMA_MAP_FLAG_VADDR,
> +        .vaddr = (__u64)(uintptr_t)vaddr,
> +        .iova = iova,
> +        .size = size,
> +    };
> +    if (ioctl(container->fd, VFIO_IOMMU_MAP_DMA, &map)) {
> +        error_setg_errno(errp, errno,
> +                         "vfio_dma_map_vaddr(iova %lu, size %ld, va %p)",
> +                         iova, size, vaddr);
> +        return -errno;
> +    }
> +    return 0;
> +}
> +
> +static int
> +vfio_region_remap(MemoryRegionSection *section, void *handle, Error **errp)
> +{
> +    MemoryRegion *mr = section->mr;
> +    VFIOContainer *container = handle;
> +    const char *name = memory_region_name(mr);
> +    ram_addr_t size = int128_get64(section->size);
> +    hwaddr offset, iova, roundup;
> +    void *vaddr;
> +
> +    if (vfio_listener_skipped_section(section) || memory_region_is_iommu(mr)) {
> +        return 0;
> +    }
> +
> +    offset = section->offset_within_address_space;
> +    iova = TARGET_PAGE_ALIGN(offset);
> +    roundup = iova - offset;
> +    size = (size - roundup) & TARGET_PAGE_MASK;
> +    vaddr = memory_region_get_ram_ptr(mr) +
> +            section->offset_within_region + roundup;
> +
> +    trace_vfio_region_remap(name, container->fd, iova, iova + size - 1, vaddr);
> +    return vfio_dma_map_vaddr(container, iova, size, vaddr, errp);
> +}
> +
> +bool vfio_cpr_capable(VFIOContainer *container, Error **errp)
> +{
> +    if (!ioctl(container->fd, VFIO_CHECK_EXTENSION, VFIO_UPDATE_VADDR) ||
> +        !ioctl(container->fd, VFIO_CHECK_EXTENSION, VFIO_UNMAP_ALL)) {
> +        error_setg(errp, "VFIO container does not support VFIO_UPDATE_VADDR "
> +                         "or VFIO_UNMAP_ALL");
> +        return false;
> +    } else {
> +        return true;
> +    }
> +}
> +
> +int vfio_cprsave(Error **errp)
> +{
> +    VFIOAddressSpace *space;
> +    VFIOContainer *container;
> +
> +    QLIST_FOREACH(space, &vfio_address_spaces, list) {
> +        QLIST_FOREACH(container, &space->containers, next) {
> +            if (!vfio_cpr_capable(container, errp)) {
> +                return 1;
> +            }
> +            if (vfio_dma_unmap_vaddr_all(container, errp)) {
> +                return 1;
> +            }
> +        }
> +    }


Seems like you'd want to test that all containers are capable before
unmapping any vaddrs.  I also hope we'll find an unwind somewhere that
remaps vaddrs should any fail.

> +    return 0;
> +}
> +
> +int vfio_cprload(Error **errp)
> +{
> +    VFIOAddressSpace *space;
> +    VFIOContainer *container;
> +    VFIOGroup *group;
> +    VFIODevice *vbasedev;
> +
> +    QLIST_FOREACH(space, &vfio_address_spaces, list) {
> +        QLIST_FOREACH(container, &space->containers, next) {
> +            if (!vfio_cpr_capable(container, errp)) {
> +                return 1;
> +            }
> +            container->reused = false;
> +            if (as_flat_walk(space->as, vfio_region_remap, container, errp)) {
> +                return 1;
> +            }
> +        }
> +    }

What state are we in if any of these fail?

> +    QLIST_FOREACH(group, &vfio_group_list, next) {
> +        QLIST_FOREACH(vbasedev, &group->device_list, next) {
> +            vbasedev->reused = false;
> +        }
> +    }
> +    return 0;
> +}
> diff --git a/hw/vfio/meson.build b/hw/vfio/meson.build
> index da9af29..e247b2b 100644
> --- a/hw/vfio/meson.build
> +++ b/hw/vfio/meson.build
> @@ -5,6 +5,7 @@ vfio_ss.add(files(
>    'migration.c',
>  ))
>  vfio_ss.add(when: 'CONFIG_VFIO_PCI', if_true: files(
> +  'cpr.c',
>    'display.c',
>    'pci-quirks.c',
>    'pci.c',
> diff --git a/hw/vfio/pci.c b/hw/vfio/pci.c
> index 7a4fb6c..f7ac9f03 100644
> --- a/hw/vfio/pci.c
> +++ b/hw/vfio/pci.c
> @@ -29,6 +29,8 @@
>  #include "hw/qdev-properties.h"
>  #include "hw/qdev-properties-system.h"
>  #include "migration/vmstate.h"
> +#include "migration/cpr.h"
> +#include "qemu/env.h"
>  #include "qemu/error-report.h"
>  #include "qemu/main-loop.h"
>  #include "qemu/module.h"
> @@ -1612,6 +1614,14 @@ static void vfio_mmap_set_enabled(VFIOPCIDevice *vdev, bool enabled)
>      }
>  }
>  
> +static void vfio_config_sync(VFIOPCIDevice *vdev, uint32_t offset, size_t len)
> +{
> +    if (pread(vdev->vbasedev.fd, vdev->pdev.config + offset, len,
> +          vdev->config_offset + offset) != len) {
> +        error_report("vfio_config_sync pread failed");
> +    }
> +}
> +
>  static void vfio_bar_prepare(VFIOPCIDevice *vdev, int nr)
>  {
>      VFIOBAR *bar = &vdev->bars[nr];
> @@ -1652,6 +1662,7 @@ static void vfio_bars_prepare(VFIOPCIDevice *vdev)
>  static void vfio_bar_register(VFIOPCIDevice *vdev, int nr)
>  {
>      VFIOBAR *bar = &vdev->bars[nr];
> +    PCIDevice *pdev = &vdev->pdev;
>      char *name;
>  
>      if (!bar->size) {
> @@ -1672,7 +1683,10 @@ static void vfio_bar_register(VFIOPCIDevice *vdev, int nr)
>          }
>      }
>  
> -    pci_register_bar(&vdev->pdev, nr, bar->type, bar->mr);
> +    pci_register_bar(pdev, nr, bar->type, bar->mr);
> +    if (pdev->reused) {
> +        vfio_config_sync(vdev, pci_bar(pdev, nr), 8);

Assuming 64-bit BARs?  This might be the first case where we actually
rely on the kernel BAR values, IIRC we usually use QEMU's emulation.

> +    }
>  }
>  
>  static void vfio_bars_register(VFIOPCIDevice *vdev)
> @@ -2884,6 +2898,7 @@ static void vfio_realize(PCIDevice *pdev, Error **errp)
>          vfio_put_group(group);
>          goto error;
>      }
> +    pdev->reused = vdev->vbasedev.reused;
>  
>      vfio_populate_device(vdev, &err);
>      if (err) {
> @@ -3046,9 +3061,11 @@ static void vfio_realize(PCIDevice *pdev, Error **errp)
>                                               vfio_intx_routing_notifier);
>          vdev->irqchip_change_notifier.notify = vfio_irqchip_change;
>          kvm_irqchip_add_change_notifier(&vdev->irqchip_change_notifier);
> -        ret = vfio_intx_enable(vdev, errp);
> -        if (ret) {
> -            goto out_deregister;
> +        if (!pdev->reused) {
> +            ret = vfio_intx_enable(vdev, errp);
> +            if (ret) {
> +                goto out_deregister;
> +            }
>          }
>      }
>  
> @@ -3098,6 +3115,11 @@ static void vfio_realize(PCIDevice *pdev, Error **errp)
>      vfio_register_req_notifier(vdev);
>      vfio_setup_resetfn_quirk(vdev);
>  
> +    vfio_config_sync(vdev, pdev->msix_cap + PCI_MSIX_FLAGS, 2);
> +    if (pdev->reused) {
> +        pci_update_mappings(pdev);
> +    }
> +

Are the msix flag sync and mapping update related?  They seem
independent to me.  A blank line and comment would be helpful.  I
expect we'd need to call msix_enabled() somewhere for the msix flag
sync to be effective.

Is there an assumption here of msi-x only support or is it not needed
for msi or intx?

>      return;
>  
>  out_deregister:
> @@ -3153,6 +3175,10 @@ static void vfio_pci_reset(DeviceState *dev)
>  {
>      VFIOPCIDevice *vdev = VFIO_PCI(dev);
>  
> +    if (vdev->pdev.reused) {
> +        return;
> +    }
> +
>      trace_vfio_pci_reset(vdev->vbasedev.name);
>  
>      vfio_pci_pre_reset(vdev);
> @@ -3260,6 +3286,36 @@ static Property vfio_pci_dev_properties[] = {
>      DEFINE_PROP_END_OF_LIST(),
>  };
>  
> +static int vfio_pci_post_load(void *opaque, int version_id)
> +{
> +    VFIOPCIDevice *vdev = opaque;
> +    PCIDevice *pdev = &vdev->pdev;
> +    bool enabled;
> +
> +    pdev->reused = false;
> +    enabled = pci_get_word(pdev->config + PCI_COMMAND) & PCI_COMMAND_MASTER;
> +    memory_region_set_enabled(&pdev->bus_master_enable_region, enabled);
> +
> +    return 0;
> +}
> +
> +static bool vfio_pci_needed(void *opaque)
> +{
> +    return cpr_active();
> +}
> +
> +static const VMStateDescription vfio_pci_vmstate = {
> +    .name = "vfio-pci",
> +    .unmigratable = 1,
> +    .version_id = 0,
> +    .minimum_version_id = 0,
> +    .post_load = vfio_pci_post_load,
> +    .needed = vfio_pci_needed,
> +    .fields = (VMStateField[]) {
> +        VMSTATE_END_OF_LIST()
> +    }
> +};
> +
>  static void vfio_pci_dev_class_init(ObjectClass *klass, void *data)
>  {
>      DeviceClass *dc = DEVICE_CLASS(klass);
> @@ -3267,6 +3323,7 @@ static void vfio_pci_dev_class_init(ObjectClass *klass, void *data)
>  
>      dc->reset = vfio_pci_reset;
>      device_class_set_props(dc, vfio_pci_dev_properties);
> +    dc->vmsd = &vfio_pci_vmstate;
>      dc->desc = "VFIO-based PCI device assignment";
>      set_bit(DEVICE_CATEGORY_MISC, dc->categories);
>      pdc->realize = vfio_realize;
> diff --git a/hw/vfio/trace-events b/hw/vfio/trace-events
> index 079f53a..0f8b166 100644
> --- a/hw/vfio/trace-events
> +++ b/hw/vfio/trace-events
> @@ -118,6 +118,7 @@ vfio_region_sparse_mmap_header(const char *name, int index, int nr_areas) "Devic
>  vfio_region_sparse_mmap_entry(int i, unsigned long start, unsigned long end) "sparse entry %d [0x%lx - 0x%lx]"
>  vfio_get_dev_region(const char *name, int index, uint32_t type, uint32_t subtype) "%s index %d, %08x/%0x8"
>  vfio_dma_unmap_overflow_workaround(void) ""
> +vfio_region_remap(const char *name, int fd, uint64_t iova_start, uint64_t iova_end, void *vaddr) "%s fd %d 0x%"PRIx64" - 0x%"PRIx64" [%p]"
>  
>  # platform.c
>  vfio_platform_base_device_init(char *name, int groupid) "%s belongs to group #%d"
> diff --git a/include/hw/pci/pci.h b/include/hw/pci/pci.h
> index bef3e49..add7f46 100644
> --- a/include/hw/pci/pci.h
> +++ b/include/hw/pci/pci.h
> @@ -360,6 +360,7 @@ struct PCIDevice {
>      /* ID of standby device in net_failover pair */
>      char *failover_pair_id;
>      uint32_t acpi_index;
> +    bool reused;
>  };
>  
>  void pci_register_bar(PCIDevice *pci_dev, int region_num,
> diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h
> index 00acb85..b46d850 100644
> --- a/include/hw/vfio/vfio-common.h
> +++ b/include/hw/vfio/vfio-common.h
> @@ -85,6 +85,7 @@ typedef struct VFIOContainer {
>      Error *error;
>      bool initialized;
>      bool dirty_pages_supported;
> +    bool reused;
>      uint64_t dirty_pgsizes;
>      uint64_t max_dirty_bitmap_size;
>      unsigned long pgsizes;
> @@ -124,6 +125,7 @@ typedef struct VFIODevice {
>      bool no_mmap;
>      bool ram_block_discard_allowed;
>      bool enable_migration;
> +    bool reused;
>      VFIODeviceOps *ops;
>      unsigned int num_irqs;
>      unsigned int num_regions;
> @@ -200,6 +202,9 @@ VFIOGroup *vfio_get_group(int groupid, AddressSpace *as, Error **errp);
>  void vfio_put_group(VFIOGroup *group);
>  int vfio_get_device(VFIOGroup *group, const char *name,
>                      VFIODevice *vbasedev, Error **errp);
> +int vfio_cprsave(Error **errp);
> +int vfio_cprload(Error **errp);
> +bool vfio_cpr_capable(VFIOContainer *container, Error **errp);
>  
>  extern const MemoryRegionOps vfio_region_ops;
>  typedef QLIST_HEAD(VFIOGroupList, VFIOGroup) VFIOGroupList;
> diff --git a/linux-headers/linux/vfio.h b/linux-headers/linux/vfio.h
> index 609099e..bc3a66e 100644
> --- a/linux-headers/linux/vfio.h
> +++ b/linux-headers/linux/vfio.h
> @@ -46,6 +46,12 @@
>   */
>  #define VFIO_NOIOMMU_IOMMU		8
>  
> +/* Supports VFIO_DMA_UNMAP_FLAG_ALL */
> +#define VFIO_UNMAP_ALL                        9
> +
> +/* Supports VFIO DMA map and unmap with the VADDR flag */
> +#define VFIO_UPDATE_VADDR              10
> +
>  /*
>   * The IOCTL interface is designed for extensibility by embedding the
>   * structure length (argsz) and flags into structures passed between
> @@ -1074,12 +1080,22 @@ struct vfio_iommu_type1_info_dma_avail {
>   *
>   * Map process virtual addresses to IO virtual addresses using the
>   * provided struct vfio_dma_map. Caller sets argsz. READ &/ WRITE required.
> + *
> + * If flags & VFIO_DMA_MAP_FLAG_VADDR, record the new base vaddr for iova, and
> + * unblock translation of host virtual addresses in the iova range.  The vaddr
> + * must have previously been invalidated with VFIO_DMA_UNMAP_FLAG_VADDR.  To
> + * maintain memory consistency within the user application, the updated vaddr
> + * must address the same memory object as originally mapped.  Failure to do so
> + * will result in user memory corruption and/or device misbehavior.  iova and
> + * size must match those in the original MAP_DMA call.  Protection is not
> + * changed, and the READ & WRITE flags must be 0.
>   */
>  struct vfio_iommu_type1_dma_map {
>  	__u32	argsz;
>  	__u32	flags;
>  #define VFIO_DMA_MAP_FLAG_READ (1 << 0)		/* readable from device */
>  #define VFIO_DMA_MAP_FLAG_WRITE (1 << 1)	/* writable from device */
> +#define VFIO_DMA_MAP_FLAG_VADDR (1 << 2)
>  	__u64	vaddr;				/* Process virtual address */
>  	__u64	iova;				/* IO virtual address */
>  	__u64	size;				/* Size of mapping (bytes) */
> @@ -1102,6 +1118,7 @@ struct vfio_bitmap {
>   * field.  No guarantee is made to the user that arbitrary unmaps of iova
>   * or size different from those used in the original mapping call will
>   * succeed.
> + *
>   * VFIO_DMA_UNMAP_FLAG_GET_DIRTY_BITMAP should be set to get the dirty bitmap
>   * before unmapping IO virtual addresses. When this flag is set, the user must
>   * provide a struct vfio_bitmap in data[]. User must provide zero-allocated
> @@ -1111,11 +1128,21 @@ struct vfio_bitmap {
>   * indicates that the page at that offset from iova is dirty. A Bitmap of the
>   * pages in the range of unmapped size is returned in the user-provided
>   * vfio_bitmap.data.
> + *
> + * If flags & VFIO_DMA_UNMAP_FLAG_ALL, unmap all addresses.  iova and size
> + * must be 0.  This cannot be combined with the get-dirty-bitmap flag.
> + *
> + * If flags & VFIO_DMA_UNMAP_FLAG_VADDR, do not unmap, but invalidate host
> + * virtual addresses in the iova range.  Tasks that attempt to translate an
> + * iova's vaddr will block.  DMA to already-mapped pages continues.  This
> + * cannot be combined with the get-dirty-bitmap flag.
>   */
>  struct vfio_iommu_type1_dma_unmap {
>  	__u32	argsz;
>  	__u32	flags;
>  #define VFIO_DMA_UNMAP_FLAG_GET_DIRTY_BITMAP (1 << 0)
> +#define VFIO_DMA_UNMAP_FLAG_ALL              (1 << 1)
> +#define VFIO_DMA_UNMAP_FLAG_VADDR            (1 << 2)
>  	__u64	iova;				/* IO virtual address */
>  	__u64	size;				/* Size of mapping (bytes) */
>  	__u8    data[];
> diff --git a/migration/cpr.c b/migration/cpr.c
> index e0da1cf..e9a189b 100644
> --- a/migration/cpr.c
> +++ b/migration/cpr.c
> @@ -132,6 +132,9 @@ void cprsave(const char *file, CprMode mode, Error **errp)
>          shutdown_action = SHUTDOWN_ACTION_POWEROFF;
>          qemu_system_shutdown_request(SHUTDOWN_CAUSE_GUEST_SHUTDOWN);
>      } else if (restart) {
> +        if (vfio_cprsave(errp)) {
> +            goto err;
> +        }
>          walkenv(FD_PREFIX, preserve_fd, 0);
>          setenv("QEMU_START_FREEZE", "", 1);
>          qemu_system_exec_request();
> @@ -176,6 +179,10 @@ void cprload(const char *file, Error **errp)
>          return;
>      }
>  
> +    if (vfio_cprload(errp)) {
> +        return;
> +    }
> +
>      state = global_state_get_runstate();
>      if (state == RUN_STATE_RUNNING) {
>          vm_start();

I didn't find that unwind I was hoping for or anywhere that the msix
flags come into play.  Thanks,

Alex

On 5/21/2021 6:24 PM, Alex Williamson wrote:> On Fri,  7 May 2021 05:25:10 -0700
> Steve Sistare <steven.sistare@oracle.com> wrote:
> 
>> Enable vfio-pci devices to be saved and restored across an exec restart
>> of qemu.
>>
>> At vfio creation time, save the value of vfio container, group, and device
>> descriptors in the environment.
>>
>> In cprsave, suspend the use of virtual addresses in DMA mappings with
>> VFIO_DMA_UNMAP_FLAG_VADDR, because guest ram will be remapped at a
>> different VA after exec.  DMA to already-mapped pages continues.  Save
>> the msi message area as part of vfio-pci vmstate, save the interrupt and
>> notifier eventfd's in the environment, and clear the close-on-exec flag
>> for the vfio descriptors.  The flag is not cleared earlier because the
>> descriptors should not persist across miscellaneous fork and exec calls
>> that may be performed during normal operation.
>>
>> On qemu restart, vfio_realize() finds the descriptor env vars, uses
>> the descriptors, and notes that the device is being reused.  Device and
>> iommu state is already configured, so operations in vfio_realize that
>> would modify the configuration are skipped for a reused device, including
>> vfio ioctl's and writes to PCI configuration space.  The result is that
>> vfio_realize constructs qemu data structures that reflect the current
>> state of the device.  However, the reconstruction is not complete until
>> cprload is called. cprload loads the msi data and finds eventfds in the
>> environment.  It rebuilds vector data structures and attaches the
>> interrupts to the new KVM instance.  cprload then walks the flattened
>> ranges of the vfio_address_spaces and calls VFIO_DMA_MAP_FLAG_VADDR to
>> inform the kernel of the new VA's.  Lastly, it starts the VM and suppresses
>> vfio device reset.
>>
>> This functionality is delivered by 2 patches for clarity.  Part 2 adds
>> eventfd and vector support.
>>
>> Signed-off-by: Steve Sistare <steven.sistare@oracle.com>
>> ---
>>  hw/pci/msi.c                  |   4 ++
>>  hw/pci/pci.c                  |   4 ++
>>  hw/vfio/common.c              |  59 ++++++++++++++++++-
>>  hw/vfio/cpr.c                 | 131 ++++++++++++++++++++++++++++++++++++++++++
>>  hw/vfio/meson.build           |   1 +
>>  hw/vfio/pci.c                 |  65 +++++++++++++++++++--
>>  hw/vfio/trace-events          |   1 +
>>  include/hw/pci/pci.h          |   1 +
>>  include/hw/vfio/vfio-common.h |   5 ++
>>  linux-headers/linux/vfio.h    |  27 +++++++++
>>  migration/cpr.c               |   7 +++
>>  11 files changed, 298 insertions(+), 7 deletions(-)
>>  create mode 100644 hw/vfio/cpr.c
>>
>> diff --git a/hw/pci/msi.c b/hw/pci/msi.c
>> index 47d2b0f..39de6a7 100644
>> --- a/hw/pci/msi.c
>> +++ b/hw/pci/msi.c
>> @@ -225,6 +225,10 @@ int msi_init(struct PCIDevice *dev, uint8_t offset,
>>      dev->msi_cap = config_offset;
>>      dev->cap_present |= QEMU_PCI_CAP_MSI;
>>  
>> +    if (dev->reused) {
>> +        return 0;
>> +    }
>> +
>>      pci_set_word(dev->config + msi_flags_off(dev), flags);
>>      pci_set_word(dev->wmask + msi_flags_off(dev),
>>                   PCI_MSI_FLAGS_QSIZE | PCI_MSI_FLAGS_ENABLE);
>> diff --git a/hw/pci/pci.c b/hw/pci/pci.c
>> index e08d981..27019ca 100644
>> --- a/hw/pci/pci.c
>> +++ b/hw/pci/pci.c
>> @@ -308,6 +308,10 @@ static void pci_do_device_reset(PCIDevice *dev)
>>  {
>>      int r;
>>  
>> +    if (dev->reused) {
>> +        return;
>> +    }
>> +
>>      pci_device_deassert_intx(dev);
>>      assert(dev->irq_state == 0);
>>  
>> diff --git a/hw/vfio/common.c b/hw/vfio/common.c
>> index 9220e64..00d07b2 100644
>> --- a/hw/vfio/common.c
>> +++ b/hw/vfio/common.c
>> @@ -31,6 +31,7 @@
>>  #include "exec/memory.h"
>>  #include "exec/ram_addr.h"
>>  #include "hw/hw.h"
>> +#include "qemu/env.h"
>>  #include "qemu/error-report.h"
>>  #include "qemu/main-loop.h"
>>  #include "qemu/range.h"
>> @@ -440,6 +441,10 @@ static int vfio_dma_unmap(VFIOContainer *container,
>>          return vfio_dma_unmap_bitmap(container, iova, size, iotlb);
>>      }
>>  
>> +    if (container->reused) {
>> +        return 0;
>> +    }
>> +
>>      while (ioctl(container->fd, VFIO_IOMMU_UNMAP_DMA, &unmap)) {
>>          /*
>>           * The type1 backend has an off-by-one bug in the kernel (71a7d3d78e3c
>> @@ -463,6 +468,11 @@ static int vfio_dma_unmap(VFIOContainer *container,
>>          return -errno;
>>      }
>>  
>> +    if (unmap.size != size) {
>> +        warn_report("VFIO_UNMAP_DMA(0x%lx, 0x%lx) only unmaps 0x%llx",
>> +                     iova, size, unmap.size);
>> +    }
>> +
>>      return 0;
>>  }
>>  
>> @@ -477,6 +487,10 @@ static int vfio_dma_map(VFIOContainer *container, hwaddr iova,
>>          .size = size,
>>      };
>>  
>> +    if (container->reused) {
>> +        return 0;
>> +    }
>> +
>>      if (!readonly) {
>>          map.flags |= VFIO_DMA_MAP_FLAG_WRITE;
>>      }
>> @@ -1603,6 +1617,10 @@ static int vfio_init_container(VFIOContainer *container, int group_fd,
>>      if (iommu_type < 0) {
>>          return iommu_type;
>>      }
>> +    if (container->reused) {
>> +        container->iommu_type = iommu_type;
>> +        return 0;
>> +    }
>>  
>>      ret = ioctl(group_fd, VFIO_GROUP_SET_CONTAINER, &container->fd);
>>      if (ret) {
>> @@ -1703,6 +1721,8 @@ static int vfio_connect_container(VFIOGroup *group, AddressSpace *as,
>>  {
>>      VFIOContainer *container;
>>      int ret, fd;
>> +    bool reused;
>> +    char name[40];
>>      VFIOAddressSpace *space;
>>  
>>      space = vfio_get_address_space(as);
>> @@ -1739,16 +1759,29 @@ static int vfio_connect_container(VFIOGroup *group, AddressSpace *as,
>>          return ret;
>>      }
>>  
>> +    snprintf(name, sizeof(name), "vfio_container_%d", group->groupid);
> 
> For more clarity, maybe "vfio_container_for_group_%d"?

OK, that is clearer.

>> +    fd = getenv_fd(name);
>> +    reused = (fd >= 0);
>> +
>>      QLIST_FOREACH(container, &space->containers, next) {
>> +        if (fd >= 0 && container->fd == fd) {
> 
> Test @reused rather than @fd?  > I'm not sure the first half of this test
> is even needed though, <0 should never match container->fd, right?

OK, I will drop the first test.

>> +            group->container = container;
>> +            QLIST_INSERT_HEAD(&container->group_list, group, container_next);
>> +            return 0;
>> +        }
> 
> This looks unnecessarily sensitive to the order of containers in the
> list, if the fd doesn't match above we try to set a new container below?
> It seems like you only want to create a new container object if none of
> the existing ones match.
> 
> There's also a lot of duplication that seems like it could be combined
> 
> if (container->fd == fd || (!reused && !ioctl(...)) {

OK, I will rewrite to avoid depending on creation order, and de-dup:

    QLIST_FOREACH(container, &space->containers, next) {
        if (container->fd == fd ||
            !ioctl(group->fd, VFIO_GROUP_SET_CONTAINER, &container->fd)) {
            break;
        }
    }

    if (container) {
        group->container = container;
        QLIST_INSERT_HEAD(&container->group_list, group, container_next);
        if (!reused) {
            vfio_kvm_device_add_group(group);
            setenv_fd(name, container->fd);
        }
        return 0;
    }


>>          if (!ioctl(group->fd, VFIO_GROUP_SET_CONTAINER, &container->fd)) {
>>              group->container = container;
>>              QLIST_INSERT_HEAD(&container->group_list, group, container_next);
>>              vfio_kvm_device_add_group(group);
> 
> Why is this kvm device setup missing in the reuse case?

vfio_kvm_device_add_group only calls ioctls, and they were already called when the device 
was initially created.

> if (!reused) {
>> +            setenv_fd(name, container->fd);
> }
> 
>>              return 0;
>>          }
>>      }
>>  
>> -    fd = qemu_open_old("/dev/vfio/vfio", O_RDWR);
>> +    if (fd < 0) {
> 
> if (!reused)?

OK.

>> +        fd = qemu_open_old("/dev/vfio/vfio", O_RDWR);
>> +    }
>> +
>>      if (fd < 0) {
>>          error_setg_errno(errp, errno, "failed to open /dev/vfio/vfio");
>>          ret = -errno;
>> @@ -1766,6 +1799,7 @@ static int vfio_connect_container(VFIOGroup *group, AddressSpace *as,
>>      container = g_malloc0(sizeof(*container));
>>      container->space = space;
>>      container->fd = fd;
>> +    container->reused = reused;
>>      container->error = NULL;
>>      container->dirty_pages_supported = false;
>>      QLIST_INIT(&container->giommu_list);
>> @@ -1893,6 +1927,7 @@ static int vfio_connect_container(VFIOGroup *group, AddressSpace *as,
>>      }
>>  
>>      container->initialized = true;
>> +    setenv_fd(name, fd);
> 
> Maybe we don't need the test around the previous setenv_fd if we can
> overwrite existing env values, which would seem to be the case for a
> restart here.

Yes, setenv_fd overwrites, so I omitted the test for clarity, at a cost of a few cycles.

>>      return 0;
>>  listener_release_exit:
>> @@ -1920,6 +1955,7 @@ static void vfio_disconnect_container(VFIOGroup *group)
>>  
>>      QLIST_REMOVE(group, container_next);
>>      group->container = NULL;
>> +    unsetenv_fdv("vfio_container_%d", group->groupid);
>>  
>>      /*
>>       * Explicitly release the listener first before unset container,
>> @@ -1978,7 +2014,12 @@ VFIOGroup *vfio_get_group(int groupid, AddressSpace *as, Error **errp)
>>      group = g_malloc0(sizeof(*group));
>>  
>>      snprintf(path, sizeof(path), "/dev/vfio/%d", groupid);
>> -    group->fd = qemu_open_old(path, O_RDWR);
>> +
>> +    group->fd = getenv_fd(path);
>> +    if (group->fd < 0) {
>> +        group->fd = qemu_open_old(path, O_RDWR);
>> +    }
>> +
>>      if (group->fd < 0) {
>>          error_setg_errno(errp, errno, "failed to open %s", path);
>>          goto free_group_exit;
>> @@ -2012,6 +2053,8 @@ VFIOGroup *vfio_get_group(int groupid, AddressSpace *as, Error **errp)
>>  
>>      QLIST_INSERT_HEAD(&vfio_group_list, group, next);
>>  
>> +    setenv_fd(path, group->fd);
>> +
>>      return group;
>>  
>>  close_fd_exit:
>> @@ -2036,6 +2079,7 @@ void vfio_put_group(VFIOGroup *group)
>>      vfio_disconnect_container(group);
>>      QLIST_REMOVE(group, next);
>>      trace_vfio_put_group(group->fd);
>> +    unsetenv_fdv("/dev/vfio/%d", group->groupid);
>>      close(group->fd);
>>      g_free(group);
>>  
>> @@ -2049,8 +2093,14 @@ int vfio_get_device(VFIOGroup *group, const char *name,
>>  {
>>      struct vfio_device_info dev_info = { .argsz = sizeof(dev_info) };
>>      int ret, fd;
>> +    bool reused;
>> +
>> +    fd = getenv_fd(name);
>> +    reused = (fd >= 0);
>> +    if (fd < 0) {
> 
> if (!reused) ?

OK.

>> +        fd = ioctl(group->fd, VFIO_GROUP_GET_DEVICE_FD, name);
>> +    }
>>  
>> -    fd = ioctl(group->fd, VFIO_GROUP_GET_DEVICE_FD, name);
>>      if (fd < 0) {
>>          error_setg_errno(errp, errno, "error getting device from group %d",
>>                           group->groupid);
>> @@ -2095,6 +2145,8 @@ int vfio_get_device(VFIOGroup *group, const char *name,
>>      vbasedev->num_irqs = dev_info.num_irqs;
>>      vbasedev->num_regions = dev_info.num_regions;
>>      vbasedev->flags = dev_info.flags;
>> +    vbasedev->reused = reused;
>> +    setenv_fd(name, fd);
>>  
>>      trace_vfio_get_device(name, dev_info.flags, dev_info.num_regions,
>>                            dev_info.num_irqs);
>> @@ -2111,6 +2163,7 @@ void vfio_put_base_device(VFIODevice *vbasedev)
>>      QLIST_REMOVE(vbasedev, next);
>>      vbasedev->group = NULL;
>>      trace_vfio_put_base_device(vbasedev->fd);
>> +    unsetenv_fd(vbasedev->name);
>>      close(vbasedev->fd);
>>  }
>>  
>> diff --git a/hw/vfio/cpr.c b/hw/vfio/cpr.c
>> new file mode 100644
>> index 0000000..c5ad9f2
>> --- /dev/null
>> +++ b/hw/vfio/cpr.c
>> @@ -0,0 +1,131 @@
>> +/*
>> + * Copyright (c) 2021 Oracle and/or its affiliates.
>> + *
>> + * This work is licensed under the terms of the GNU GPL, version 2.
>> + * See the COPYING file in the top-level directory.
>> + */
>> +
>> +#include "qemu/osdep.h"
>> +#include <sys/ioctl.h>
>> +#include <linux/vfio.h>
>> +#include "hw/vfio/vfio-common.h"
>> +#include "sysemu/kvm.h"
>> +#include "qapi/error.h"
>> +#include "trace.h"
>> +
>> +static int
>> +vfio_dma_unmap_vaddr_all(VFIOContainer *container, Error **errp)
>> +{
>> +    struct vfio_iommu_type1_dma_unmap unmap = {
>> +        .argsz = sizeof(unmap),
>> +        .flags = VFIO_DMA_UNMAP_FLAG_VADDR | VFIO_DMA_UNMAP_FLAG_ALL,
>> +        .iova = 0,
>> +        .size = 0,
>> +    };
>> +    if (ioctl(container->fd, VFIO_IOMMU_UNMAP_DMA, &unmap)) {
>> +        error_setg_errno(errp, errno, "vfio_dma_unmap_vaddr_all");
>> +        return -errno;
>> +    }
>> +    return 0;
>> +}
>> +
>> +static int vfio_dma_map_vaddr(VFIOContainer *container, hwaddr iova,
>> +                              ram_addr_t size, void *vaddr,
>> +                              Error **errp)
>> +{
>> +    struct vfio_iommu_type1_dma_map map = {
>> +        .argsz = sizeof(map),
>> +        .flags = VFIO_DMA_MAP_FLAG_VADDR,
>> +        .vaddr = (__u64)(uintptr_t)vaddr,
>> +        .iova = iova,
>> +        .size = size,
>> +    };
>> +    if (ioctl(container->fd, VFIO_IOMMU_MAP_DMA, &map)) {
>> +        error_setg_errno(errp, errno,
>> +                         "vfio_dma_map_vaddr(iova %lu, size %ld, va %p)",
>> +                         iova, size, vaddr);
>> +        return -errno;
>> +    }
>> +    return 0;
>> +}
>> +
>> +static int
>> +vfio_region_remap(MemoryRegionSection *section, void *handle, Error **errp)
>> +{
>> +    MemoryRegion *mr = section->mr;
>> +    VFIOContainer *container = handle;
>> +    const char *name = memory_region_name(mr);
>> +    ram_addr_t size = int128_get64(section->size);
>> +    hwaddr offset, iova, roundup;
>> +    void *vaddr;
>> +
>> +    if (vfio_listener_skipped_section(section) || memory_region_is_iommu(mr)) {
>> +        return 0;
>> +    }
>> +
>> +    offset = section->offset_within_address_space;
>> +    iova = TARGET_PAGE_ALIGN(offset);
>> +    roundup = iova - offset;
>> +    size = (size - roundup) & TARGET_PAGE_MASK;
>> +    vaddr = memory_region_get_ram_ptr(mr) +
>> +            section->offset_within_region + roundup;
>> +
>> +    trace_vfio_region_remap(name, container->fd, iova, iova + size - 1, vaddr);
>> +    return vfio_dma_map_vaddr(container, iova, size, vaddr, errp);
>> +}
>> +
>> +bool vfio_cpr_capable(VFIOContainer *container, Error **errp)
>> +{
>> +    if (!ioctl(container->fd, VFIO_CHECK_EXTENSION, VFIO_UPDATE_VADDR) ||
>> +        !ioctl(container->fd, VFIO_CHECK_EXTENSION, VFIO_UNMAP_ALL)) {
>> +        error_setg(errp, "VFIO container does not support VFIO_UPDATE_VADDR "
>> +                         "or VFIO_UNMAP_ALL");
>> +        return false;
>> +    } else {
>> +        return true;
>> +    }
>> +}
>> +
>> +int vfio_cprsave(Error **errp)
>> +{
>> +    VFIOAddressSpace *space;
>> +    VFIOContainer *container;
>> +
>> +    QLIST_FOREACH(space, &vfio_address_spaces, list) {
>> +        QLIST_FOREACH(container, &space->containers, next) {
>> +            if (!vfio_cpr_capable(container, errp)) {
>> +                return 1;
>> +            }
>> +            if (vfio_dma_unmap_vaddr_all(container, errp)) {
>> +                return 1;
>> +            }
>> +        }
>> +    }
> 
> Seems like you'd want to test that all containers are capable before
> unmapping any vaddrs.  I also hope we'll find an unwind somewhere that
> remaps vaddrs should any fail.

That is verified earlier if one runs qemu with the -only-cpr-capable option:

  vfio_get_iommu_type()
      for (i = 0; i < ARRAY_SIZE(iommu_types); i++) {
          if (ioctl(container->fd, VFIO_CHECK_EXTENSION, iommu_types[i])) {
              if (only_cpr_capable && !vfio_cpr_capable(container, errp)) {
                  error_prepend(errp, "only-cpr-capable is specified: ");
                  return -EINVAL;

But I will check here as well in case only-cpr-capable was not specified.

I will add code to unwind and remap.

>> +    return 0;
>> +}
>> +
>> +int vfio_cprload(Error **errp)
>> +{
>> +    VFIOAddressSpace *space;
>> +    VFIOContainer *container;
>> +    VFIOGroup *group;
>> +    VFIODevice *vbasedev;
>> +
>> +    QLIST_FOREACH(space, &vfio_address_spaces, list) {
>> +        QLIST_FOREACH(container, &space->containers, next) {
>> +            if (!vfio_cpr_capable(container, errp)) {
>> +                return 1;
>> +            }
>> +            container->reused = false;
>> +            if (as_flat_walk(space->as, vfio_region_remap, container, errp)) {
>> +                return 1;
>> +            }
>> +        }
>> +    }
> 
> What state are we in if any of these fail?

The guest has not resumed, but we cannot recover.  Since we verified vfio_cpr_capable in 
cprsave, vfio_cprload should never fail, sans bugs.

>> +    QLIST_FOREACH(group, &vfio_group_list, next) {
>> +        QLIST_FOREACH(vbasedev, &group->device_list, next) {
>> +            vbasedev->reused = false;
>> +        }
>> +    }
>> +    return 0;
>> +}
>> diff --git a/hw/vfio/meson.build b/hw/vfio/meson.build
>> index da9af29..e247b2b 100644
>> --- a/hw/vfio/meson.build
>> +++ b/hw/vfio/meson.build
>> @@ -5,6 +5,7 @@ vfio_ss.add(files(
>>    'migration.c',
>>  ))
>>  vfio_ss.add(when: 'CONFIG_VFIO_PCI', if_true: files(
>> +  'cpr.c',
>>    'display.c',
>>    'pci-quirks.c',
>>    'pci.c',
>> diff --git a/hw/vfio/pci.c b/hw/vfio/pci.c
>> index 7a4fb6c..f7ac9f03 100644
>> --- a/hw/vfio/pci.c
>> +++ b/hw/vfio/pci.c
>> @@ -29,6 +29,8 @@
>>  #include "hw/qdev-properties.h"
>>  #include "hw/qdev-properties-system.h"
>>  #include "migration/vmstate.h"
>> +#include "migration/cpr.h"
>> +#include "qemu/env.h"
>>  #include "qemu/error-report.h"
>>  #include "qemu/main-loop.h"
>>  #include "qemu/module.h"
>> @@ -1612,6 +1614,14 @@ static void vfio_mmap_set_enabled(VFIOPCIDevice *vdev, bool enabled)
>>      }
>>  }
>>  
>> +static void vfio_config_sync(VFIOPCIDevice *vdev, uint32_t offset, size_t len)
>> +{
>> +    if (pread(vdev->vbasedev.fd, vdev->pdev.config + offset, len,
>> +          vdev->config_offset + offset) != len) {
>> +        error_report("vfio_config_sync pread failed");
>> +    }
>> +}
>> +
>>  static void vfio_bar_prepare(VFIOPCIDevice *vdev, int nr)
>>  {
>>      VFIOBAR *bar = &vdev->bars[nr];
>> @@ -1652,6 +1662,7 @@ static void vfio_bars_prepare(VFIOPCIDevice *vdev)
>>  static void vfio_bar_register(VFIOPCIDevice *vdev, int nr)
>>  {
>>      VFIOBAR *bar = &vdev->bars[nr];
>> +    PCIDevice *pdev = &vdev->pdev;
>>      char *name;
>>  
>>      if (!bar->size) {
>> @@ -1672,7 +1683,10 @@ static void vfio_bar_register(VFIOPCIDevice *vdev, int nr)
>>          }
>>      }
>>  
>> -    pci_register_bar(&vdev->pdev, nr, bar->type, bar->mr);
>> +    pci_register_bar(pdev, nr, bar->type, bar->mr);
>> +    if (pdev->reused) {
>> +        vfio_config_sync(vdev, pci_bar(pdev, nr), 8);
> 
> Assuming 64-bit BARs?  This might be the first case where we actually
> rely on the kernel BAR values, IIRC we usually use QEMU's emulation.

No asssumptions.  vfio_config_sync() preads a piece of config space using a single 
system call, copying directly to the qemu buffer, not looking at words or calling any
action functions.

>> +    }
>>  }
>>  
>>  static void vfio_bars_register(VFIOPCIDevice *vdev)
>> @@ -2884,6 +2898,7 @@ static void vfio_realize(PCIDevice *pdev, Error **errp)
>>          vfio_put_group(group);
>>          goto error;
>>      }
>> +    pdev->reused = vdev->vbasedev.reused;
>>  
>>      vfio_populate_device(vdev, &err);
>>      if (err) {
>> @@ -3046,9 +3061,11 @@ static void vfio_realize(PCIDevice *pdev, Error **errp)
>>                                               vfio_intx_routing_notifier);
>>          vdev->irqchip_change_notifier.notify = vfio_irqchip_change;
>>          kvm_irqchip_add_change_notifier(&vdev->irqchip_change_notifier);
>> -        ret = vfio_intx_enable(vdev, errp);
>> -        if (ret) {
>> -            goto out_deregister;
>> +        if (!pdev->reused) {
>> +            ret = vfio_intx_enable(vdev, errp);
>> +            if (ret) {
>> +                goto out_deregister;
>> +            }
>>          }
>>      }
>>  
>> @@ -3098,6 +3115,11 @@ static void vfio_realize(PCIDevice *pdev, Error **errp)
>>      vfio_register_req_notifier(vdev);
>>      vfio_setup_resetfn_quirk(vdev);
>>  
>> +    vfio_config_sync(vdev, pdev->msix_cap + PCI_MSIX_FLAGS, 2);
>> +    if (pdev->reused) {
>> +        pci_update_mappings(pdev);
>> +    }
>> +
> 
> Are the msix flag sync and mapping update related?  They seem
> independent to me.  A blank line and comment would be helpful.

OK.

> I expect we'd need to call msix_enabled() somewhere for the msix flag
> sync to be effective.

Yes, vfio_pci_post_load in cpr part 2 calls msix_enabled.

> Is there an assumption here of msi-x only support or is it not needed
> for msi or intx?

The code supports msi-x and msi.  However, I should only be sync'ing PCI_MSIX_FLAGS
if pdev->cap_present & QEMU_PCI_CAP_MSIX.  And, I am missing a sync for PCI_MSI_FLAGS.
I'll fix that.

> [...]
> 
> I didn't find that unwind I was hoping for or anywhere that the msix
> flags come into play.  Thanks,

Let me know f I have not adequately answered those questions.

- Steve

On 5/24/2021 2:29 PM, Steven Sistare wrote:
> On 5/21/2021 6:24 PM, Alex Williamson wrote:> On Fri,  7 May 2021 05:25:10 -0700
>> Steve Sistare <steven.sistare@oracle.com> wrote:
>>
>>>[...]
>>> diff --git a/hw/vfio/pci.c b/hw/vfio/pci.c
>>> index 7a4fb6c..f7ac9f03 100644
>>> --- a/hw/vfio/pci.c
>>> +++ b/hw/vfio/pci.c
>>> @@ -29,6 +29,8 @@
>>>  #include "hw/qdev-properties.h"
>>>  #include "hw/qdev-properties-system.h"
>>>  #include "migration/vmstate.h"
>>> +#include "migration/cpr.h"
>>> +#include "qemu/env.h"
>>>  #include "qemu/error-report.h"
>>>  #include "qemu/main-loop.h"
>>>  #include "qemu/module.h"
>>> @@ -1612,6 +1614,14 @@ static void vfio_mmap_set_enabled(VFIOPCIDevice *vdev, bool enabled)
>>>      }
>>>  }
>>>  
>>> +static void vfio_config_sync(VFIOPCIDevice *vdev, uint32_t offset, size_t len)
>>> +{
>>> +    if (pread(vdev->vbasedev.fd, vdev->pdev.config + offset, len,
>>> +          vdev->config_offset + offset) != len) {
>>> +        error_report("vfio_config_sync pread failed");
>>> +    }
>>> +}
>>> +
>>>  static void vfio_bar_prepare(VFIOPCIDevice *vdev, int nr)
>>>  {
>>>      VFIOBAR *bar = &vdev->bars[nr];
>>> @@ -1652,6 +1662,7 @@ static void vfio_bars_prepare(VFIOPCIDevice *vdev)
>>>  static void vfio_bar_register(VFIOPCIDevice *vdev, int nr)
>>>  {
>>>      VFIOBAR *bar = &vdev->bars[nr];
>>> +    PCIDevice *pdev = &vdev->pdev;
>>>      char *name;
>>>  
>>>      if (!bar->size) {
>>> @@ -1672,7 +1683,10 @@ static void vfio_bar_register(VFIOPCIDevice *vdev, int nr)
>>>          }
>>>      }
>>>  
>>> -    pci_register_bar(&vdev->pdev, nr, bar->type, bar->mr);
>>> +    pci_register_bar(pdev, nr, bar->type, bar->mr);
>>> +    if (pdev->reused) {
>>> +        vfio_config_sync(vdev, pci_bar(pdev, nr), 8);
>>
>> Assuming 64-bit BARs?  This might be the first case where we actually
>> rely on the kernel BAR values, IIRC we usually use QEMU's emulation.
> 
> No asssumptions.  vfio_config_sync() preads a piece of config space using a single 
> system call, copying directly to the qemu buffer, not looking at words or calling any
> action functions.
> 
>[...] 
>>> @@ -3098,6 +3115,11 @@ static void vfio_realize(PCIDevice *pdev, Error **errp)
>>>      vfio_register_req_notifier(vdev);
>>>      vfio_setup_resetfn_quirk(vdev);
>>>  
>>> +    vfio_config_sync(vdev, pdev->msix_cap + PCI_MSIX_FLAGS, 2);
>>> +    if (pdev->reused) {
>>> +        pci_update_mappings(pdev);
>>> +    }
>>> +
>>
>> Are the msix flag sync and mapping update related?  They seem
>> independent to me.  A blank line and comment would be helpful.
> 
> OK.
> 
>> I expect we'd need to call msix_enabled() somewhere for the msix flag
>> sync to be effective.
> 
> Yes, vfio_pci_post_load in cpr part 2 calls msix_enabled.
> 
>> Is there an assumption here of msi-x only support or is it not needed
>> for msi or intx?
> 
> The code supports msi-x and msi.  However, I should only be sync'ing PCI_MSIX_FLAGS
> if pdev->cap_present & QEMU_PCI_CAP_MSIX.  And, I am missing a sync for PCI_MSI_FLAGS.
> I'll fix that.

Hi Alex, FYI, I am making more changes here.  The calls to vfio_config_sync fix pdev->config[]
words that are initialized during vfio_realize(), by pread'ing from the live kernel config.
However, it makes more sense to suppress the undesired re-initialization, rather than undo
the damage later.  Thus I will add a few more 'if (!pdev->reused)' guards in msix and pci bar
init functions, and delete vfio_config_sync.

Most of the config is preserved in the kernel across restart.  However, the bits that are
purely emulated (indicated by the emulated_config_bits mask) may be rejected when they 
are written through to the kernel, and thus are currently lost on restart.  I need to save 
pdev->config[] in the vmstate file, and in vfio_pci_post_load, merge it with the kernel 
config using emulated_config_bits.

Sound sane?

- Steve

On 6/11/2021 2:15 PM, Steven Sistare wrote:
> On 5/24/2021 2:29 PM, Steven Sistare wrote:
>> On 5/21/2021 6:24 PM, Alex Williamson wrote:> On Fri,  7 May 2021 05:25:10 -0700
>>> Steve Sistare <steven.sistare@oracle.com> wrote:
>>>
>>>> [...]
>>>> diff --git a/hw/vfio/pci.c b/hw/vfio/pci.c
>>>> index 7a4fb6c..f7ac9f03 100644
>>>> --- a/hw/vfio/pci.c
>>>> +++ b/hw/vfio/pci.c
>>>> @@ -29,6 +29,8 @@
>>>>  #include "hw/qdev-properties.h"
>>>>  #include "hw/qdev-properties-system.h"
>>>>  #include "migration/vmstate.h"
>>>> +#include "migration/cpr.h"
>>>> +#include "qemu/env.h"
>>>>  #include "qemu/error-report.h"
>>>>  #include "qemu/main-loop.h"
>>>>  #include "qemu/module.h"
>>>> @@ -1612,6 +1614,14 @@ static void vfio_mmap_set_enabled(VFIOPCIDevice *vdev, bool enabled)
>>>>      }
>>>>  }
>>>>  
>>>> +static void vfio_config_sync(VFIOPCIDevice *vdev, uint32_t offset, size_t len)
>>>> +{
>>>> +    if (pread(vdev->vbasedev.fd, vdev->pdev.config + offset, len,
>>>> +          vdev->config_offset + offset) != len) {
>>>> +        error_report("vfio_config_sync pread failed");
>>>> +    }
>>>> +}
>>>> +
>>>>  static void vfio_bar_prepare(VFIOPCIDevice *vdev, int nr)
>>>>  {
>>>>      VFIOBAR *bar = &vdev->bars[nr];
>>>> @@ -1652,6 +1662,7 @@ static void vfio_bars_prepare(VFIOPCIDevice *vdev)
>>>>  static void vfio_bar_register(VFIOPCIDevice *vdev, int nr)
>>>>  {
>>>>      VFIOBAR *bar = &vdev->bars[nr];
>>>> +    PCIDevice *pdev = &vdev->pdev;
>>>>      char *name;
>>>>  
>>>>      if (!bar->size) {
>>>> @@ -1672,7 +1683,10 @@ static void vfio_bar_register(VFIOPCIDevice *vdev, int nr)
>>>>          }
>>>>      }
>>>>  
>>>> -    pci_register_bar(&vdev->pdev, nr, bar->type, bar->mr);
>>>> +    pci_register_bar(pdev, nr, bar->type, bar->mr);
>>>> +    if (pdev->reused) {
>>>> +        vfio_config_sync(vdev, pci_bar(pdev, nr), 8);
>>>
>>> Assuming 64-bit BARs?  This might be the first case where we actually
>>> rely on the kernel BAR values, IIRC we usually use QEMU's emulation.
>>
>> No asssumptions.  vfio_config_sync() preads a piece of config space using a single 
>> system call, copying directly to the qemu buffer, not looking at words or calling any
>> action functions.
>>
>> [...] 
>>>> @@ -3098,6 +3115,11 @@ static void vfio_realize(PCIDevice *pdev, Error **errp)
>>>>      vfio_register_req_notifier(vdev);
>>>>      vfio_setup_resetfn_quirk(vdev);
>>>>  
>>>> +    vfio_config_sync(vdev, pdev->msix_cap + PCI_MSIX_FLAGS, 2);
>>>> +    if (pdev->reused) {
>>>> +        pci_update_mappings(pdev);
>>>> +    }
>>>> +
>>>
>>> Are the msix flag sync and mapping update related?  They seem
>>> independent to me.  A blank line and comment would be helpful.
>>
>> OK.
>>
>>> I expect we'd need to call msix_enabled() somewhere for the msix flag
>>> sync to be effective.
>>
>> Yes, vfio_pci_post_load in cpr part 2 calls msix_enabled.
>>
>>> Is there an assumption here of msi-x only support or is it not needed
>>> for msi or intx?
>>
>> The code supports msi-x and msi.  However, I should only be sync'ing PCI_MSIX_FLAGS
>> if pdev->cap_present & QEMU_PCI_CAP_MSIX.  And, I am missing a sync for PCI_MSI_FLAGS.
>> I'll fix that.
> 
> Hi Alex, FYI, I am making more changes here.  The calls to vfio_config_sync fix pdev->config[]
> words that are initialized during vfio_realize(), by pread'ing from the live kernel config.
> However, it makes more sense to suppress the undesired re-initialization, rather than undo
> the damage later.  Thus I will add a few more 'if (!pdev->reused)' guards in msix and pci bar
> init functions, and delete vfio_config_sync.
> 
> Most of the config is preserved in the kernel across restart.  However, the bits that are
> purely emulated (indicated by the emulated_config_bits mask) may be rejected when they 
> are written through to the kernel, and thus are currently lost on restart.  I need to save 
> pdev->config[] in the vmstate file, and in vfio_pci_post_load, merge it with the kernel 
> config using emulated_config_bits.
> 
> Sound sane?

Furthermore, there is no need to check reused and suppress initialization of msix and pci bar, 
as the vmstate loader fixes them up.

- Steve