On Tue, Jul 20, 2021 at 01:32:43PM +0200, Kevin Wolf wrote:
> Am 14.07.2021 um 11:29 hat Stefan Hajnoczi geschrieben:
> > v3:
> > * Rebased
> > v2:
> > * Clarify that object-memory-memfd enables share=on by default [Marc-André]
> > * Rebased
> >
> > vhost-user requires -object memory-backend-*,share=on option so that QEMU uses
> > mmap(MAP_SHARED) on guest RAM that is shared with the vhost-user device backend
> > process. This is needed so the QEMU process sees changes made by the vhost-user
> > device backend process, and vice versa.
> >
> > Today QEMU and the vhost-user device process will start up and then fail with a
> > confusing error message if the user forgot to specify share=on.
> >
> > This patch series adds a warning letting the user know that share=on is
> > required.
>
> Hmm... This seems to work when I use shared=off for the memory backend,
> even though the error message is easy to miss because more error
> messages follow:
>
> $ LANG=C build/qemu-system-x86_64 -object memory-backend-memfd,id=mem,size=1G,share=off -m 1G -M memory-backend=mem -chardev socket,path=/tmp/vhost.sock,id=vhost1 -device vhost-user-blk-pci,chardev=vhost1
> qemu-system-x86_64: -device vhost-user-blk-pci,chardev=vhost1: warning: Found vhost-user memory region without MAP_SHARED (did you forget -object memory-*,share=on?)
> qemu-system-x86_64: Unexpected end-of-file before all data were read
> qemu-system-x86_64: Failed to set msg fds.
> qemu-system-x86_64: vhost VQ 0 ring restore failed: -1: Resource temporarily unavailable (11)
>
> However, if I don't create the memory backend at all (which is probably
> the most common error users will make), then the new message isn't
> printed:
>
> $ LANG=C build/qemu-system-x86_64 -m 1G -chardev socket,path=/tmp/vhost.sock,id=vhost1 -device vhost-user-blk-pci,chardev=vhost1
> qemu-system-x86_64: Unexpected end-of-file before all data were read
> qemu-system-x86_64: Failed to set msg fds.
> qemu-system-x86_64: vhost VQ 0 ring restore failed: -1: Resource temporarily unavailable (11)
>
> Wouldn't we want a warning for the second case, too?
Yes. I think so. I'll investigate.
Stefan