6 Commits
Author SHA1 Message Date
Will Deacon 27347f76ac kvm tools: ioeventfd: replace bool parameters to __add_event with flags
A recent fix to virtio MMIO (72a7541ce305 ["kvm tools: virtio-mmio:
init_ioeventfd should use MMIO for ioeventfd__add_event()"]) highlighted
the confusing parameters expected by ioeventfd__add_event.

As per Pekka's suggestion, replace the bool parameters to this function
with a single `flags' argument instead.

Cc: Ying-Shiuan Pan <yingshiuan.pan@gmail.com>
Signed-off-by: Will Deacon <will.deacon@arm.com>
Signed-off-by: Pekka Enberg <penberg@kernel.org>
2015-06-01 16:39:54 +01:00
Asias He 627d68743d kvm tools: Do not poll ioeventfd if vhost is enabled
If vhost is enabled for a virtio device, vhost will poll the ioeventfd
in kernel side and there is no need to poll it in userspace. Otherwise,
both vhost kernel and userspace will race to poll.

Signed-off-by: Asias He <asias.hejun@gmail.com>
Signed-off-by: Pekka Enberg <penberg@kernel.org>
2015-06-01 16:39:52 +01:00
Asias He 9ff913391e kvm tools: Add is_pio flag to ioeventfd__add_event
Add is_pio flag to set KVM_IOEVENTFD_FLAG_PIO. This is useful for
attaching an ioeventfd to MMIO address as well as PIO address.
virtio-mmio needs an ioeventfd to MMIO address.

Signed-off-by: Asias He <asias.hejun@gmail.com>
Signed-off-by: Pekka Enberg <penberg@kernel.org>
2015-06-01 16:39:51 +01:00
Sasha Levin ea6eeb1c10 kvm tools: Fixes for ioeventfd module
Fixes include:
 - Error handling
 - Cleanup
 - Standard init/uninit

Signed-off-by: Sasha Levin <levinsasha928@gmail.com>
2015-06-01 16:39:51 +01:00
Sasha Levin e13377810d kvm tools: Don't use ioeventfds if no KVM_CAP_IOEVENTFD
Check KVM_CAP_IOEVENTFD before using ioeventfds.

Signed-off-by: Sasha Levin <levinsasha928@gmail.com>
Signed-off-by: Pekka Enberg <penberg@kernel.org>
2015-06-01 16:39:49 +01:00
Sasha Levin 37f3d50e6a kvm tools: Add ioeventfd support
ioeventfd is way provided by KVM to receive notifications about
reads and writes to PIO and MMIO areas within the guest.

Such notifications are usefull if all we need to know is that
a specific area of the memory has been changed, and we don't need
a heavyweight exit to happen.

The implementation uses epoll to scale to large number of ioeventfds.

Benchmarks ran on a seperate (non boot) 1GB virtio-blk device, formatted
as ext4, using bonnie++.

cmd line:
# bonnie++ -d temp/ -c 2 -s 768 -u 0

Before:
Version  1.96       ------Sequential Output------ --Sequential Input- --Random-
Concurrency   2     -Per Chr- --Block-- -Rewrite- -Per Chr- --Block-- --Seeks--
Machine        Size K/sec %CP K/sec %CP K/sec %CP K/sec %CP K/sec %CP  /sec %CP
tux            768M   498  99 381127  74 269712  48  2871  99 717109  50 +++++ +++
Latency             18368us   31063us   21281us    3017us    6875us     251ms
Version  1.96       ------Sequential Create------ --------Random Create--------
tux                 -Create-- --Read--- -Delete-- -Create-- --Read--- -Delete--
              files  /sec %CP  /sec %CP  /sec %CP  /sec %CP  /sec %CP  /sec %CP
                 16 +++++ +++ +++++ +++ +++++ +++ +++++ +++ +++++ +++ +++++ +++
Latency               148us     588us    2792us    1547us    1543us     218us

After:
Version  1.96       ------Sequential Output------ --Sequential Input- --Random-
Concurrency   2     -Per Chr- --Block-- -Rewrite- -Per Chr- --Block-- --Seeks--
Machine        Size K/sec %CP K/sec %CP K/sec %CP K/sec %CP K/sec %CP  /sec %CP
tux            768M   499  99 459779  73 350689  54  2997  99 860395  58 +++++ +++
Latency             17194us   14619us   26358us    4055us    7890us   44122us
Version  1.96       ------Sequential Create------ --------Random Create--------
tux                 -Create-- --Read--- -Delete-- -Create-- --Read--- -Delete--
              files  /sec %CP  /sec %CP  /sec %CP  /sec %CP  /sec %CP  /sec %CP
                 16 +++++ +++ +++++ +++ +++++ +++ +++++ +++ +++++ +++ +++++ +++
Latency               135us     567us    2564us     134us    1500us     231us

Signed-off-by: Sasha Levin <levinsasha928@gmail.com>
Signed-off-by: Pekka Enberg <penberg@kernel.org>
2015-06-01 16:39:44 +01:00