Re: [Qemu-devel] [PULL 00/38] vhost, pc, virtio features, fixes, cleanups

[Peter Maydell]

On 21 October 2015 at 11:26, Michael S. Tsirkin <address@hidden> wrote:
> The following changes since commit 6d57410a79d51d92673c54f26624b44f27fa6214:
>
>   Merge remote-tracking branch 
> 'remotes/pmaydell/tags/pull-target-arm-20151016' into staging (2015-10-17 
> 12:31:33 +0100)
>
> are available in the git repository at:
>
>   git://git.kernel.org/pub/scm/virt/kvm/mst/qemu.git tags/for_upstream
>
> for you to fetch changes up to 4f53cd7788185f437a4275df6456eb60b90d4837:
>
>   hw/isa/lpc_ich9: inject the SMI on the VCPU that is writing to APM_CNT 
> (2015-10-21 13:24:49 +0300)
>
> ----------------------------------------------------------------
> vhost, pc, virtio features, fixes, cleanups
>
> New features:
>     VT-d support for devices behind a bridge
>     vhost-user migration support
>
> Signed-off-by: Michael S. Tsirkin <address@hidden>
>
> ----------------------------------------------------------------

Test failure on 64-bit ARM and 32-bit ARM:

TEST: tests/vhost-user-test... (pid=28701)
Warning: path not on HugeTLBFS: /tmp/vhost-test-pLQQzA
qemu-system-i386: -netdev
vhost-user,id=net0,chardev=chr-test,vhostforce: vhost-net support is
not compiled in
qemu-system-i386: -netdev
vhost-user,id=net0,chardev=chr-test,vhostforce: failed to init
vhost_net for queue 0

Broken pipe
FAIL: tests/vhost-user-test

My clang build gives this:

Warning: path not on HugeTLBFS: /tmp/vhost-test-EFF6ul
/home/petmay01/linaro/qemu-for-merges/target-i386/translate.c:2421:26:
runtime error: left shift of negative value -8
Warning: path not on HugeTLBFS: /tmp/vhost-test-EFF6ul
/home/petmay01/linaro/qemu-for-merges/target-i386/translate.c:2421:26:
runtime error: left shift of negative value -8
Warning: path not on HugeTLBFS: /tmp/vhost-test-EFF6ul
/home/petmay01/linaro/qemu-for-merges/target-i386/translate.c:2421:26:
runtime error: left shift of negative value -8

The tests pass, but I'd prefer it if our tests
didn't print out random warnings in the non-verbose
case if possible; it makes it hard to sift out
"this is a problem" from "this test has always
been grumpy". (Alternatively, if "path not on HugeTLBFS"
is a problem then the test should fail.)

The negative-shift stuff is clang's undefined
sanitizer; the code in question has been doing a
"-8 << s->dflag" since at least 2013, so this is just
your test code exercising something it hadn't before.
Sanitizer issues aren't currently something I'm
requiring pull requests to deal with, so we can fix
this one later.

thanks
-- PMM