Re: [RFC PATCH v2 7/7] util/vfio-helpers: Allow opening device requesting for multiple IRQs

[Alex Williamson <alex.williamson@redhat.com>]

On Thu, 13 Aug 2020 19:29:57 +0200
Philippe Mathieu-Daudé <philmd@redhat.com> wrote:

> Now that our helper is ready for handling multiple IRQs, let
> qemu_vfio_open_pci() take an 'irq_count' argument.
> 
> Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
> ---

As with patch 2/ tying IRQ setup with the opening of a device seems
wrong.  Get the device open, then create an interface to configure the
interrupt.  Thanks,

Alex


>  include/qemu/vfio-helpers.h | 2 +-
>  block/nvme.c                | 5 ++++-
>  util/vfio-helpers.c         | 4 ++--
>  3 files changed, 7 insertions(+), 4 deletions(-)
> 
> diff --git a/include/qemu/vfio-helpers.h b/include/qemu/vfio-helpers.h
> index 5c2d8ee5b3..4773b116df 100644
> --- a/include/qemu/vfio-helpers.h
> +++ b/include/qemu/vfio-helpers.h
> @@ -16,7 +16,7 @@
>  typedef struct QEMUVFIOState QEMUVFIOState;
>  
>  QEMUVFIOState *qemu_vfio_open_pci(const char *device, int irq_type,
> -                                  Error **errp);
> +                                  unsigned irq_count, Error **errp);
>  void qemu_vfio_close(QEMUVFIOState *s);
>  int qemu_vfio_dma_map(QEMUVFIOState *s, void *host, size_t size,
>                        bool temporary, uint64_t *iova_list);
> diff --git a/block/nvme.c b/block/nvme.c
> index a5ef571492..2d7aac3903 100644
> --- a/block/nvme.c
> +++ b/block/nvme.c
> @@ -106,6 +106,9 @@ QEMU_BUILD_BUG_ON(offsetof(NVMeRegs, doorbells) != 0x1000);
>  #define INDEX_ADMIN     0
>  #define INDEX_IO(n)     (1 + n)
>  
> +/* This driver shares a single MSIX IRQ for the admin and I/O queues */
> +#define MSIX_IRQ_COUNT  1
> +
>  struct BDRVNVMeState {
>      AioContext *aio_context;
>      QEMUVFIOState *vfio;
> @@ -712,7 +715,7 @@ static int nvme_init(BlockDriverState *bs, const char *device, int namespace,
>      }
>  
>      s->vfio = qemu_vfio_open_pci(device, VFIO_PCI_MSIX_IRQ_INDEX,
> -                                 errp);
> +                                 MSIX_IRQ_COUNT, errp);
>      if (!s->vfio) {
>          ret = -EINVAL;
>          goto out;
> diff --git a/util/vfio-helpers.c b/util/vfio-helpers.c
> index 7a934d1a1b..36fafef0d3 100644
> --- a/util/vfio-helpers.c
> +++ b/util/vfio-helpers.c
> @@ -450,12 +450,12 @@ static void qemu_vfio_open_common(QEMUVFIOState *s)
>   * Open a PCI device, e.g. "0000:00:01.0".
>   */
>  QEMUVFIOState *qemu_vfio_open_pci(const char *device, int irq_type,
> -                                  Error **errp)
> +                                  unsigned irq_count, Error **errp)
>  {
>      int r;
>      QEMUVFIOState *s = g_new0(QEMUVFIOState, 1);
>  
> -    r = qemu_vfio_init_pci(s, device, irq_type, 1, errp);
> +    r = qemu_vfio_init_pci(s, device, irq_type, irq_count, errp);
>      if (r) {
>          g_free(s);
>          return NULL;