[Qemu-devel] Re: [PATCH-RFC 11/13] vhost net support

[Anthony Liguori <anthony@codemonkey.ws>]

On 01/11/2010 11:23 AM, Michael S. Tsirkin wrote:
> This adds vhost net support in qemu. Will be tied to tap device and
> virtio later. Raw backend is currently missing, will be worked
> on/submitted separately.
>
> Signed-off-by: Michael S. Tsirkin<mst@redhat.com>
>    

Hit send too quick.. here it is.

> ---
>   Makefile.target |    1 +
>   hw/vhost.c      |  349 +++++++++++++++++++++++++++++++++++++++++++++++++++++++
>   hw/vhost.h      |   33 +++++
>   hw/vhost_net.c  |  145 +++++++++++++++++++++++
>   hw/vhost_net.h  |   20 +++
>   5 files changed, 548 insertions(+), 0 deletions(-)
>   create mode 100644 hw/vhost.c
>   create mode 100644 hw/vhost.h
>   create mode 100644 hw/vhost_net.c
>   create mode 100644 hw/vhost_net.h
>
> diff --git a/Makefile.target b/Makefile.target
> index 7c1f30c..61b7148 100644
> --- a/Makefile.target
> +++ b/Makefile.target
> @@ -157,6 +157,7 @@ obj-y = vl.o async.o monitor.o pci.o pci_host.o pcie_host.o machine.o gdbstub.o
>   # virtio has to be here due to weird dependency between PCI and virtio-net.
>   # need to fix this properly
>   obj-y += virtio-blk.o virtio-balloon.o virtio-net.o virtio-console.o virtio-pci.o
> +obj-y += vhost_net.o vhost.o
>   obj-$(CONFIG_KVM) += kvm.o kvm-all.o
>   obj-$(CONFIG_ISA_MMIO) += isa_mmio.o
>   LIBS+=-lz
> diff --git a/hw/vhost.c b/hw/vhost.c
> new file mode 100644
> index 0000000..d23d94c
> --- /dev/null
> +++ b/hw/vhost.c
> @@ -0,0 +1,349 @@
> +#include "linux/vhost.h"
> +#include<sys/ioctl.h>
> +#include<sys/eventfd.h>
> +#include "vhost.h"
> +#include "hw/hw.h"
> +/* For range_get_last */
> +#include "pci.h"
> +
> +static void vhost_dev_unassign_memory(struct vhost_dev *dev,
> +				      struct vhost_memory *mem,
> +				      uint64_t start_addr,
> +				      uint64_t size)
> +{
> +	int from, to;
> +	for (from = 0, to = 0; from<  dev->mem->nregions; ++from, ++to) {
> +		struct vhost_memory_region *reg = mem->regions + to;
> +		uint64_t reglast;
> +		uint64_t memlast;
> +		uint64_t change;
> +
> +		/* clone old region */
> +		memcpy(reg, dev->mem->regions + from, sizeof *reg);
> +
> +		/* No overlap is simple */
> +		if (!ranges_overlap(reg->guest_phys_addr, reg->memory_size,
> +				    start_addr, size)) {
> +			continue;
> +		}
> +		reglast = range_get_last(reg->guest_phys_addr, reg->memory_size);
> +		memlast = range_get_last(start_addr, size);
> +
> +		/* Remove whole region */
> +		if (start_addr<= reg->guest_phys_addr&&  memlast>= reglast) {
> +			--to;
> +			continue;
> +		}
> +
> +		/* Shrink region */
> +		if (memlast>= reglast) {
> +			reg->memory_size = start_addr - reg->guest_phys_addr;
> +			continue;
> +		}
> +
> +		/* Shift region */
> +		if (start_addr<= reg->guest_phys_addr) {
> +			change = memlast + 1 - reg->guest_phys_addr;
> +			reg->memory_size -= change;
> +			reg->guest_phys_addr += change;
> +			reg->userspace_addr += change;
> +			continue;
> +		}
> +
> +		/* Split region: shrink first part, shift second part. */
> +		memcpy(reg + 1, reg, sizeof *reg);
> +		reg[0].memory_size = start_addr - reg->guest_phys_addr;
> +		change = memlast + 1 - reg->guest_phys_addr;
> +		reg[1].memory_size -= change;
> +		reg[1].guest_phys_addr += change;
> +		reg[1].userspace_addr += change;
> +		++to;
> +	}
> +	mem->nregions = to;
> +}
> +
> +/* Called after unassign, so no regions overlap the given range. */
> +static void vhost_dev_assign_memory(struct vhost_dev *dev,
> +				    struct vhost_memory *mem,
> +				    uint64_t start_addr,
> +				    uint64_t size,
> +				    uint64_t uaddr)
> +{
> +	int from, to;
> +	struct vhost_memory_region *merged = NULL;
> +	for (from = 0, to = 0; from<  dev->mem->nregions; ++from, ++to) {
> +		struct vhost_memory_region *reg = mem->regions + to;
> +		uint64_t prlast, urlast;
> +		uint64_t pmlast, umlast;
> +		uint64_t s, e, u;
> +
> +		/* clone old region */
> +		memcpy(reg, dev->mem->regions + from, sizeof *reg);
> +		prlast = range_get_last(reg->guest_phys_addr, reg->memory_size);
> +		pmlast = range_get_last(start_addr, size);
> +		urlast = range_get_last(reg->userspace_addr, reg->memory_size);
> +		umlast = range_get_last(uaddr, size);
> +
> +		/* Not an adjecent region - do not merge. */
> +		if ((prlast + 1 != start_addr || urlast + 1 != uaddr)&&
> +		    (pmlast + 1 != reg->guest_phys_addr ||
> +		     umlast + 1 != reg->userspace_addr)) {
> +			continue;
> +		}
> +
> +		if (!merged) {
> +			--to;
> +		} else {
> +			merged = reg;
> +		}
> +		u = MIN(uaddr, reg->userspace_addr);
> +		s = MIN(start_addr, reg->guest_phys_addr);
> +		e = MAX(pmlast, prlast);
> +		uaddr = merged->userspace_addr = u;
> +		start_addr = merged->guest_phys_addr = s;
> +		size = merged->memory_size = e - s + 1;
> +	}
> +
> +	if (!merged) {
> +		struct vhost_memory_region *reg = mem->regions + to;
> +		reg->memory_size = size;
> +		reg->guest_phys_addr = start_addr;
> +		reg->userspace_addr = uaddr;
> +		++to;
> +	}
> +	mem->nregions = to;
> +}
> +
> +static void vhost_client_set_memory(CPUPhysMemoryClient *client,
> +				    target_phys_addr_t start_addr,
> +				    ram_addr_t size,
> +				    ram_addr_t phys_offset)
> +{
> +	struct vhost_dev *dev = container_of(client, struct vhost_dev, client);
> +	ram_addr_t flags = phys_offset&  ~TARGET_PAGE_MASK;
> +	int s = offsetof(struct vhost_memory, regions) +
> +		(dev->mem->nregions + 1)* sizeof dev->mem->regions[0];
> +	struct vhost_memory *mem = qemu_malloc(s);
> +	memcpy(mem, dev->mem, s);
> +
> +	/* First, remove old mapping for this memory, if any. */
> +	vhost_dev_unassign_memory(dev, mem, start_addr, size);
> +	if (flags == IO_MEM_RAM) {
> +		/* Add given mapping, merging adjacent regions if any */
> +		vhost_dev_assign_memory(dev, mem, start_addr, size,
> +				(uintptr_t)qemu_get_ram_ptr(phys_offset));
> +	}
> +	qemu_free(dev->mem);
> +	dev->mem = mem;
> +}
>    

This is gnarly and basically duplicated with kvm-all.c.  Dunno if code 
sharing is practical though.diff --git a/hw/vhost.h b/hw/vhost.h

> new file mode 100644
> index 0000000..9f82b42
> --- /dev/null
> +++ b/hw/vhost.h
> @@ -0,0 +1,33 @@
> +#ifndef VHOST_H
> +#define VHOST_H
> +
> +#include "hw/hw.h"
> +#include "hw/virtio.h"
> +
> +/* Generic structures common for any vhost based device. */
> +struct vhost_virtqueue {
> +	int kick;
> +	int call;
> +	void *desc;
> +	void *avail;
> +	void *used;
> +};
>    

Please follow CodingStyle.

Regards,

Anthony Liguori