similar to: [PATCHv4 0/6] qemu-kvm: vhost net support

Add raw network backend option which uses a packet socket to provide raw networking access. Once the socket is opened it's bound to a provided host interface, such that packets received on the interface are delivered to the VM and packets sent by the VM are sent to the interface. This is functionally similar to the existing pcap network backend, with the same advantages and problems.

Add raw network backend option which uses a packet socket to provide raw networking access. Once the socket is opened it's bound to a provided host interface, such that packets received on the interface are delivered to the VM and packets sent by the VM are sent to the interface. This is functionally similar to the existing pcap network backend, with the same advantages and problems.

This adds support for vhost-net virtio kernel backend. This patch is not intended to being merged yet. I'm posting it for the benefit of people testing the backend. Usage instructions: vhost currently requires MSI-X support in guest virtio. This means guests kernel version should be >= 2.6.31. To enable vhost, simply add ",vhost" flag to nic options. Example with tap backend:

This adds support for vhost-net virtio kernel backend. This patch is not intended to being merged yet. I'm posting it for the benefit of people testing the backend. Usage instructions: vhost currently requires MSI-X support in guest virtio. This means guests kernel version should be >= 2.6.31. To enable vhost, simply add ",vhost" flag to nic options. Example with tap backend:

This adds support for vhost-net virtio kernel backend. To enable (assuming device eth2): 1. enable promisc mode or program guest mac in device eth2 2. disable tso, gso, lro, jumbo frames on the card (disabling lro + jumbo frames should be sufficient, haven't tested this) 3. add vhost=eth2 to -net flag 4. run with CAP_NET_ADMIN priviledge (e.g. root) This patch is RFC, but works without

This adds support for vhost-net virtio kernel backend. To enable (assuming device eth2): 1. enable promisc mode or program guest mac in device eth2 2. disable tso, gso, lro, jumbo frames on the card (disabling lro + jumbo frames should be sufficient, haven't tested this) 3. add vhost=eth2 to -net flag 4. run with CAP_NET_ADMIN priviledge (e.g. root) This patch is RFC, but works without

Hello all: This seires is an update of last version of multiqueue support to add multiqueue capability to both tap and virtio-net. Some kinds of tap backends has (macvatp in linux) or would (tap) support multiqueue. In such kind of tap backend, each file descriptor of a tap is a qeueu and ioctls were prodived to attach an exist tap file descriptor to the tun/tap device. So the patch let qemu to

Hello all: This seires is an update of last version of multiqueue support to add multiqueue capability to both tap and virtio-net. Some kinds of tap backends has (macvatp in linux) or would (tap) support multiqueue. In such kind of tap backend, each file descriptor of a tap is a qeueu and ioctls were prodived to attach an exist tap file descriptor to the tun/tap device. So the patch let qemu to

Hello all: This seires is an update of last version of multiqueue support to add multiqueue capability to both tap and virtio-net. Some kinds of tap backends has (macvatp in linux) or would (tap) support multiqueue. In such kind of tap backend, each file descriptor of a tap is a qeueu and ioctls were prodived to attach an exist tap file descriptor to the tun/tap device. So the patch let qemu to

Hello all: This seires is an update of last version of multiqueue support to add multiqueue capability to both tap and virtio-net. Some kinds of tap backends has (macvatp in linux) or would (tap) support multiqueue. In such kind of tap backend, each file descriptor of a tap is a qeueu and ioctls were prodived to attach an exist tap file descriptor to the tun/tap device. So the patch let qemu to

Here is the port of MSI-X support patches to upstream qemu. Please comment or commit. This patchset adds generic support for MSI-X, adds implementation in APIC, and uses MSI-X in virtio-net. Changelog: - since v3 call to resize_region on load split patches a bit differently to address style comments by Glauber update commit message to clarify what msix_support flag does - since v2 rename

Here is the port of MSI-X support patches to upstream qemu. Please comment or commit. This patchset adds generic support for MSI-X, adds implementation in APIC, and uses MSI-X in virtio-net. Changelog: - since v3 call to resize_region on load split patches a bit differently to address style comments by Glauber update commit message to clarify what msix_support flag does - since v2 rename

Here's a patch series to fix known regressions in virtio_pci, by refactoring code along the lines suggested by Rusty. Changes since v3: whitespace fixed in PATCH 1/2 This is on top of patch virtio: fix memory leak on device removal that has been applied by Rusty. This supercedes patches: [PATCHv3] virtio: recover from vector assignment failure [PATCHv2] virtio: fix double free_irq on

Here's a patch series to fix known regressions in virtio_pci, by refactoring code along the lines suggested by Rusty. Changes since v3: whitespace fixed in PATCH 1/2 This is on top of patch virtio: fix memory leak on device removal that has been applied by Rusty. This supercedes patches: [PATCHv3] virtio: recover from vector assignment failure [PATCHv2] virtio: fix double free_irq on

On Tuesday 20 October 2015 18:45:55 Maxim Perevedentsev wrote: > +int64_t > +do_vfs_minimum_size (const mountable_t *mountable) > +{ > + int64_t r; > + > + /* How we set the label depends on the filesystem type. */ Just noticed now: this comment should be really fitting more the new function... :) -- Pino Toscano

On Tue, Oct 20, 2015 at 06:45:54PM +0300, Maxim Perevedentsev wrote: > Tried to make it in accordance with your comments. > > Difference to v1: > Added reply_with_error where necessary. > Changed name get_min_size -> vfs_min_size. > > Difference to v2: > Changed name to vfs_minimum_size. > Changed parsing to xstrtol + STR* macros where possible. > >

On 10/20/2015 07:49 PM, Pino Toscano wrote: > On Tuesday 20 October 2015 18:45:55 Maxim Perevedentsev wrote: >> +int64_t >> +do_vfs_minimum_size (const mountable_t *mountable) >> +{ >> + int64_t r; >> + >> + /* How we set the label depends on the filesystem type. */ > Just noticed now: this comment should be really fitting more the new > function...

I have an issue with btrfs: There's an API > btrfs inspect-internal min-dev-size /sysroot/ But this API is available only in btrfs-progs >= 4.2. Before this, there's no reliable way to get minimum size (I could parse "btrfs filesystem show", but it is veeery inaccurate). Can we require btrfs-progs v4.2? Or how should I work-around this issue? -- Your sincerely, Maxim

On 10/20/2015 08:58 PM, Richard W.M. Jones wrote: > On Tue, Oct 20, 2015 at 08:56:25PM +0300, Maxim Perevedentsev wrote: >> I have an issue with btrfs: >> >> There's an API >>> btrfs inspect-internal min-dev-size /sysroot/ >> But this API is available only in btrfs-progs >= 4.2. >> Before this, there's no reliable way to get minimum size (I

Rusty, could you review and comment on the patches please? Since most of the code deals with virtio from host side, I think it will make sense to merge them through your tree. What do you think? One comment on placement: I put files under a separate vhost directory to avoid confusion with virtio-net which runs in guest. Does this sound sane? If not let me know and I'll move them. Thanks!