Re: [kvm-devel] [PATCH/RFC 6/9] virtual block device driver

2007-05-16 Thread Avi Kivity
Carsten Otte wrote: > >>> System call latency adds to the in-kernel approach here. >> I don't understand this. > What I meant to state was: If the host side of the block driver runs > in userspace, we have the extra latency to leave the kernel system > call context, compute on behalf of the user

Re: [kvm-devel] [PATCH/RFC 6/9] virtual block device driver

2007-05-15 Thread Carsten Otte
Avi Kivity wrote: > I don't see an immediate need to put the host-side driver in the kernel, > but I don't want to embed the host fd (which is an implementation > detail) into the host/guest ABI. There may not even be a host fd. Your point is taken, it also punches a hole in the security barrier

Re: [kvm-devel] [PATCH/RFC 6/9] virtual block device driver

2007-05-14 Thread Avi Kivity
Carsten Otte wrote: > > Avi Kivity wrote: >> Is this the host file descriptor? If so, we want to use something >> more abstract (if the host side is in kernel, there will be no fd, or >> if the device is implemented using >1 files (or <1 files)). > This is indeed the host file descriptor. Host u

Re: [kvm-devel] [PATCH/RFC 6/9] virtual block device driver

2007-05-14 Thread Carsten Otte
Avi Kivity wrote: > Oh. Why not use Xen's pending block driver? It probably has everything > needed. We're not too eager to have our own device drivers become the solution of choice. I have'nt looked at it so far, will do. so long, Carsten ---

Re: [kvm-devel] [PATCH/RFC 6/9] virtual block device driver

2007-05-14 Thread Carsten Otte
Avi Kivity wrote: > Is this the host file descriptor? If so, we want to use something more > abstract (if the host side is in kernel, there will be no fd, or if the > device is implemented using >1 files (or <1 files)). This is indeed the host file descriptor. Host userland uses sys_open to re

Re: [kvm-devel] [PATCH/RFC 6/9] virtual block device driver

2007-05-14 Thread Avi Kivity
Carsten Otte wrote: > From: Carsten Otte <[EMAIL PROTECTED]> > > This driver provides access to virtual block devices. It does use its own > make_request function which passes the bio to a workqueue thread. The > workqueue > thread does use the diagnose hypervisor call to call the hosting Linux. >

Re: [kvm-devel] [PATCH/RFC 6/9] virtual block device driver

2007-05-14 Thread Avi Kivity
Carsten Otte wrote: > From: Carsten Otte <[EMAIL PROTECTED]> > > This driver provides access to virtual block devices. It does use its own > make_request function which passes the bio to a workqueue thread. The > workqueue > thread does use the diagnose hypervisor call to call the hosting Linux. >

[kvm-devel] [PATCH/RFC 6/9] virtual block device driver

2007-05-11 Thread Carsten Otte
From: Carsten Otte <[EMAIL PROTECTED]> This driver provides access to virtual block devices. It does use its own make_request function which passes the bio to a workqueue thread. The workqueue thread does use the diagnose hypervisor call to call the hosting Linux. The hypervisor code in host users