OSDN Git Service

qmiga/qemu.git
6 years agohw/arm: Allow manually specified /psci node
Andrey Smirnov [Tue, 10 Apr 2018 12:02:24 +0000 (13:02 +0100)]
hw/arm: Allow manually specified /psci node

Change the code to avoid exiting QEMU if user provided DTB contains
manually specified /psci node and skip any /psci related fixups
instead.

Fixes: 4cbca7d9b4 ("hw/arm: Move virt's PSCI DT fixup code to
arm/boot.c")

Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com>
Reported-by: Marc Zyngier <marc.zyngier@arm.com>
Tested-by: Marc Zyngier <marc.zyngier@arm.com>
Message-id: 20180402205654.14572-1-andrew.smirnov@gmail.com
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
6 years agoMerge remote-tracking branch 'remotes/xtensa/tags/20180409-xtensa' into staging
Peter Maydell [Tue, 10 Apr 2018 09:22:45 +0000 (10:22 +0100)]
Merge remote-tracking branch 'remotes/xtensa/tags/20180409-xtensa' into staging

Fix file offset for preadv/pwritev linux-user syscalls.

# gpg: Signature made Tue 10 Apr 2018 03:04:24 BST
# gpg:                using RSA key 51F9CC91F83FA044
# gpg: Good signature from "Max Filippov <filippov@cadence.com>"
# gpg:                 aka "Max Filippov <max.filippov@cogentembedded.com>"
# gpg:                 aka "Max Filippov <jcmvbkbc@gmail.com>"
# Primary key fingerprint: 2B67 854B 98E5 327D CDEB  17D8 51F9 CC91 F83F A044

* remotes/xtensa/tags/20180409-xtensa:
  linux-user: fix preadv/pwritev offsets

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
6 years agolinux-user: fix preadv/pwritev offsets
Max Filippov [Thu, 5 Apr 2018 00:30:41 +0000 (17:30 -0700)]
linux-user: fix preadv/pwritev offsets

preadv/pwritev accept low and high parts of file offset in two separate
parameters. When host bitness doesn't match guest bitness these parts
must be appropriately recombined.
Introduce target_to_host_low_high that does this recombination and use
it in preadv/pwritev syscalls.

This fixes glibc testsuite test misc/tst-preadvwritev64.

Reviewed-by: Laurent Vivier <laurent@vivier.eu>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
6 years agoMerge remote-tracking branch 'remotes/cohuck/tags/s390x-20180409' into staging
Peter Maydell [Mon, 9 Apr 2018 17:21:22 +0000 (18:21 +0100)]
Merge remote-tracking branch 'remotes/cohuck/tags/s390x-20180409' into staging

Fixes for s390x: kvm, vfio-ccw, ipl code, bios. Includes a rebuild
of s390-ccw.img and s390-netboot.img.

# gpg: Signature made Mon 09 Apr 2018 16:08:19 BST
# gpg:                using RSA key DECF6B93C6F02FAF
# gpg: Good signature from "Cornelia Huck <conny@cornelia-huck.de>"
# gpg:                 aka "Cornelia Huck <huckc@linux.vnet.ibm.com>"
# gpg:                 aka "Cornelia Huck <cornelia.huck@de.ibm.com>"
# gpg:                 aka "Cornelia Huck <cohuck@kernel.org>"
# gpg:                 aka "Cornelia Huck <cohuck@redhat.com>"
# Primary key fingerprint: C3D0 D66D C362 4FF6 A8C0  18CE DECF 6B93 C6F0 2FAF

* remotes/cohuck/tags/s390x-20180409:
  s390x: load_psw() should only exchange the PSW for KVM
  s390x/mmu: don't overwrite pending exception in mmu translate
  vfio-ccw: fix memory leaks in vfio_ccw_realize()
  pc-bios/s390: update images
  s390: Do not pass inofficial IPL type to the guest
  s390: Ensure IPL from SCSI works as expected
  s390: Refactor IPL parameter block generation
  s390x/kvm: call cpu_synchronize_state() on every kvm_arch_handle_exit()

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
6 years agoMerge remote-tracking branch 'remotes/bonzini/tags/for-upstream' into staging
Peter Maydell [Mon, 9 Apr 2018 16:29:09 +0000 (17:29 +0100)]
Merge remote-tracking branch 'remotes/bonzini/tags/for-upstream' into staging

Miscellaneous bugfixes, including crash fixes from Alexey, Peter M. and
Thomas.

# gpg: Signature made Mon 09 Apr 2018 15:37:15 BST
# gpg:                using RSA key BFFBD25F78C7AE83
# gpg: Good signature from "Paolo Bonzini <bonzini@gnu.org>"
# gpg:                 aka "Paolo Bonzini <pbonzini@redhat.com>"
# Primary key fingerprint: 46F5 9FBD 57D6 12E7 BFD4  E2F7 7E15 100C CD36 69B1
#      Subkey fingerprint: F133 3857 4B66 2389 866C  7682 BFFB D25F 78C7 AE83

* remotes/bonzini/tags/for-upstream:
  Add missing bit for SSE instr in VEX decoding
  maint: Add .mailmap entries for patches claiming list authorship
  dump: Fix build with newer gcc
  device-crash-test: Remove fixed isa-fdc entry
  qemu-pr-helper: Write pidfile more often
  qemu-pr-helper: Daemonize before dropping privileges
  virtio-serial: fix heapover-flow
  kvmclock: fix clock_is_reliable on migration from QEMU < 2.9
  hw/dma/i82374: Avoid double creation of the 82374 controller
  hw/scsi: support SCSI-2 passthrough without PI
  scsi-disk: allow customizing the SCSI version
  scsi-disk: Don't enlarge min_io_size to max_io_size
  configure: Add missing configure options to help text
  i386/hyperv: error out if features requested but unsupported
  i386/hyperv: add hv-frequencies cpu property
  target/i386: WHPX: set CPUID_EXT_HYPERVISOR bit
  memfd: fix vhost-user-test on non-memfd capable host
  scripts/checkpatch.pl: Bug fix
  target/i386: Fix andn instruction
  sys_membarrier: fix up include directives

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
6 years agoMerge remote-tracking branch 'remotes/mst/tags/for_upstream' into staging
Peter Maydell [Mon, 9 Apr 2018 15:25:38 +0000 (16:25 +0100)]
Merge remote-tracking branch 'remotes/mst/tags/for_upstream' into staging

virtio,vhost: fixes

Add a feature flag for new protocol messages.
Misc fixes.

Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
# gpg: Signature made Mon 09 Apr 2018 15:37:29 BST
# gpg:                using RSA key 281F0DB8D28D5469
# gpg: Good signature from "Michael S. Tsirkin <mst@kernel.org>"
# gpg:                 aka "Michael S. Tsirkin <mst@redhat.com>"
# Primary key fingerprint: 0270 606B 6F3C DF3D 0B17  0970 C350 3912 AFBE 8E67
#      Subkey fingerprint: 5D09 FD08 71C8 F85B 94CA  8A0D 281F 0DB8 D28D 5469

* remotes/mst/tags/for_upstream:
  virtio-serial: fix heap-over-flow
  vhost: Allow adjoining regions
  contrib/libvhost-user: add the protocol feature used for SET/GET message
  vhost-user: back SET/GET_CONFIG requests with a protocol feature
  vhost-user-blk: set config ops before vhost-user init

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
6 years agoAdd missing bit for SSE instr in VEX decoding
Eugene Minibaev [Fri, 6 Apr 2018 13:41:52 +0000 (16:41 +0300)]
Add missing bit for SSE instr in VEX decoding

The 2-byte VEX prefix imples a leading 0Fh opcode byte.

Signed-off-by: Eugene Minibaev <mail@kitsu.me>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
6 years agomaint: Add .mailmap entries for patches claiming list authorship
Eric Blake [Mon, 26 Mar 2018 18:41:47 +0000 (13:41 -0500)]
maint: Add .mailmap entries for patches claiming list authorship

The list did not author any patches, but it does rewrite the
'From:' header of messages sent from any domain with restrictive
SPF policies that would otherwise prevent the message from reaching
all list recipients.  If a maintainer is not careful to undo the
list header rewrite, and the author did not include a manual
'From:' line in the body to fix the munged header, then 'git am'
happily attributes the patch to the list.  Add some mailmap
entries to correct the few that have escaped our attention; while
we also work on improving the tooling to catch the problem in
the future before a merge is even made.

Also improve the comments occurring in the file, including line
length improvements.

Signed-off-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
6 years agodump: Fix build with newer gcc
Eric Blake [Tue, 27 Mar 2018 20:21:51 +0000 (15:21 -0500)]
dump: Fix build with newer gcc

gcc 8 on rawhide is picky enough to complain:

/home/dummy/qemu/dump.c: In function 'create_header32':
/home/dummy/qemu/dump.c:817:5: error: 'strncpy' output truncated before terminating nul copying 8 bytes from a string of the same length [-Werror=stringop-truncation]
     strncpy(dh->signature, KDUMP_SIGNATURE, strlen(KDUMP_SIGNATURE));
     ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

But we already have SIG_LEN defined as the right length without needing
to do a strlen(), and memcpy() is better than strncpy() when we know
we do not want a trailing NUL byte.

Signed-off-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
6 years agodevice-crash-test: Remove fixed isa-fdc entry
Thomas Huth [Fri, 16 Mar 2018 09:51:30 +0000 (10:51 +0100)]
device-crash-test: Remove fixed isa-fdc entry

Fixed by commit b3da551 ("fdc: Exit if ISA controller does not support DMA", 2018-03-16).

Signed-off-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
6 years agoqemu-pr-helper: Write pidfile more often
Michal Privoznik [Tue, 3 Apr 2018 13:12:15 +0000 (15:12 +0200)]
qemu-pr-helper: Write pidfile more often

Let's write pidfile even if user did not request --daemon but
they requested just --pidfile. Libvirt will use exactly this.

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
6 years agoqemu-pr-helper: Daemonize before dropping privileges
Michal Privoznik [Tue, 3 Apr 2018 13:12:14 +0000 (15:12 +0200)]
qemu-pr-helper: Daemonize before dropping privileges

After we've dropped privileges it might be not possible to write
pidfile. For instance, if this binary is run as root (because
user wants it to write pidfile to some privileged location)
writing pidfile fails because privileges are dropped before we
even get to that.

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
6 years agovirtio-serial: fix heapover-flow
linzhecheng [Wed, 28 Mar 2018 13:34:35 +0000 (21:34 +0800)]
virtio-serial: fix heapover-flow

Check device having the feature of VIRTIO_CONSOLE_F_EMERG_WRITE before
get config->emerg_wr. It is neccessary because sizeof(virtio_console_config)
is 8 byte if VirtIOSerial doesn't have the feature of
VIRTIO_CONSOLE_F_EMERG_WRITE(see virtio_serial_device_realize),
read/write emerg_wr will lead to heap-over-flow.

Signed-off-by: linzhecheng <linzhecheng@huawei.com>
Message-Id: <20180328133435.20112-1-linzhecheng@huawei.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
6 years agokvmclock: fix clock_is_reliable on migration from QEMU < 2.9
Michael Chapman [Fri, 6 Apr 2018 05:34:06 +0000 (15:34 +1000)]
kvmclock: fix clock_is_reliable on migration from QEMU < 2.9

When migrating from a pre-2.9 QEMU, no clock_is_reliable flag is
transferred. We should assume that the source host has an unreliable
KVM_GET_CLOCK, rather than using whatever was determined locally, to
ensure that any drift from the TSC-based value calculated by the guest
is corrected.

Signed-off-by: Michael Chapman <mike@very.puzzling.org>
Message-Id: <20180406053406.774-1-mike@very.puzzling.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
6 years agohw/dma/i82374: Avoid double creation of the 82374 controller
Philippe Mathieu-Daudé [Mon, 26 Mar 2018 15:34:37 +0000 (12:34 -0300)]
hw/dma/i82374: Avoid double creation of the 82374 controller

QEMU fails when used with the following command line:

    ./ppc64-softmmu/qemu-system-ppc64 -S -machine 40p -device i82374
    qemu-system-ppc64: hw/isa/isa-bus.c:110: isa_bus_dma: Assertion `!bus->dma[0] && !bus->dma[1]' failed.

The 40p machine type already creates the device i82374. If specified in the
command line, it will try to create it again, hence generating the error. The
function isa_bus_dma() isn't supposed to be called twice for the same bus.
Check the bus doesn't already have a DMA controller registered before creating
the device.

Fixes: https://bugs.launchpad.net/qemu/+bug/1721224
Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-Id: <20180326153441.32641-2-f4bug@amsat.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
6 years agohw/scsi: support SCSI-2 passthrough without PI
Daniel Henrique Barboza [Tue, 27 Mar 2018 21:14:51 +0000 (18:14 -0300)]
hw/scsi: support SCSI-2 passthrough without PI

QEMU SCSI code makes assumptions about how the PROTECT and BYTCHK
works in the protocol, denying support for PI (Protection
Information) in case the guest OS requests it. However, in SCSI versions 2
and older, there is no PI concept in the protocol.

This means that when dealing with such devices:

- there is no PROTECT bit in byte 5 of the standard INQUIRY response. The
whole byte is marked as "Reserved";

- there is no RDPROTECT in byte 2 of READ. We have 'Logical Unit Number'
in this field instead;

- there is no VRPROTECT in byte 2 of VERIFY. We have 'Logical Unit Number'
in this field instead. This also means that the BYTCHK bit in this case
is not related to PI.

Since QEMU does not consider these changes, a SCSI passthrough using
a SCSI-2 device will not work. It will mistake these fields with
PI information and return Illegal Request SCSI SENSE thinking
that the driver is asking for PI support.

This patch fixes it by adding a new attribute called 'scsi_version'
that is read from the standard INQUIRY response of passthrough
devices. This allows for a version verification before applying
conditions related to PI that doesn't apply for older versions.

Reported-by: Dac Nguyen <dacng@us.ibm.com>
Signed-off-by: Daniel Henrique Barboza <danielhb@linux.vnet.ibm.com>
Message-Id: <20180327211451.14647-1-danielhb@linux.vnet.ibm.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
6 years agoscsi-disk: allow customizing the SCSI version
Paolo Bonzini [Thu, 5 Apr 2018 16:09:51 +0000 (18:09 +0200)]
scsi-disk: allow customizing the SCSI version

We would like to have different behavior for passthrough devices
depending on the SCSI version they expose.  To prepare for that,
allow the user of emulated devices to specify the desired SCSI
level, and adjust the emulation according to the property value.
The next patch will set the level for scsi-block and scsi-generic
devices.

Based on a patch by Daniel Henrique Barboza
<danielhb@linux.vnet.ibm.com>.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
6 years agoscsi-disk: Don't enlarge min_io_size to max_io_size
Fam Zheng [Tue, 27 Mar 2018 16:41:41 +0000 (00:41 +0800)]
scsi-disk: Don't enlarge min_io_size to max_io_size

Some backends report big max_io_sectors. Making min_io_size the same
value in this case will make it impossible for guest to align memory,
therefore the disk may not be usable at all.

Do not enlarge them when they are zero.

Reported-by: David Gibson <dgibson@redhat.com>
Signed-off-by: Fam Zheng <famz@redhat.com>
Message-Id: <20180327164141.19075-1-famz@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
6 years agoconfigure: Add missing configure options to help text
Thomas Huth [Tue, 27 Mar 2018 15:09:30 +0000 (17:09 +0200)]
configure: Add missing configure options to help text

We forgot to mention --with-git, --libexecdir and --with-pkgversion
so far.

Signed-off-by: Thomas Huth <thuth@redhat.com>
Message-Id: <1522163370-18544-1-git-send-email-thuth@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
6 years agoi386/hyperv: error out if features requested but unsupported
Roman Kagan [Fri, 30 Mar 2018 17:02:09 +0000 (20:02 +0300)]
i386/hyperv: error out if features requested but unsupported

In order to guarantee compatibility on migration, QEMU should have
complete control over the features it announces to the guest via CPUID.

However, for a number of Hyper-V-related cpu properties, if the
corresponding feature is not supported by the underlying KVM, the
propery is silently ignored and the feature is not announced to the
guest.

Refuse to start with an error instead.

Signed-off-by: Roman Kagan <rkagan@virtuozzo.com>
Message-Id: <20180330170209.20627-3-rkagan@virtuozzo.com>
Reviewed-by: Eduardo Habkost <ehabkost@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
6 years agoi386/hyperv: add hv-frequencies cpu property
Roman Kagan [Fri, 30 Mar 2018 17:02:08 +0000 (20:02 +0300)]
i386/hyperv: add hv-frequencies cpu property

In order to guarantee compatibility on migration, QEMU should have
complete control over the features it announces to the guest via CPUID.

However, the availability of Hyper-V frequency MSRs
(HV_X64_MSR_TSC_FREQUENCY and HV_X64_MSR_APIC_FREQUENCY) depends solely
on the support for them in the underlying KVM.

Introduce "hv-frequencies" cpu property (off by default) which gives
QEMU full control over whether these MSRs are announced.

While at this, drop the redundant check of the cpu tsc frequency, and
decouple this feature from hv-time.

Signed-off-by: Roman Kagan <rkagan@virtuozzo.com>
Reviewed-by: Eduardo Habkost <ehabkost@redhat.com>
Message-Id: <20180330170209.20627-2-rkagan@virtuozzo.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
6 years agotarget/i386: WHPX: set CPUID_EXT_HYPERVISOR bit
Justin Terry (VM) [Mon, 26 Mar 2018 17:06:58 +0000 (10:06 -0700)]
target/i386: WHPX: set CPUID_EXT_HYPERVISOR bit

Implements the CPUID trap for CPUID 1 to include the
CPUID_EXT_HYPERVISOR flag in the ECX results. This was preventing some
older linux kernels from booting when trying to access MSR's that dont
make sense when virtualized.

Signed-off-by: Justin Terry (VM) <juterry@microsoft.com>
Message-Id: <20180326170658.606-1-juterry@microsoft.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
6 years agovirtio-serial: fix heap-over-flow
linzhecheng [Wed, 28 Mar 2018 13:34:35 +0000 (21:34 +0800)]
virtio-serial: fix heap-over-flow

Check device having the feature of VIRTIO_CONSOLE_F_EMERG_WRITE before
get config->emerg_wr. It is neccessary because sizeof(virtio_console_config)
is 8 byte if VirtIOSerial doesn't have the feature of
VIRTIO_CONSOLE_F_EMERG_WRITE(see virtio_serial_device_realize),
read/write emerg_wr will lead to heap-over-flow.

Signed-off-by: linzhecheng <linzhecheng@huawei.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
6 years agovhost: Allow adjoining regions
Dr. David Alan Gilbert [Fri, 23 Mar 2018 15:39:39 +0000 (15:39 +0000)]
vhost: Allow adjoining regions

My rework of section adding combines overlapping or adjoining regions,
but checks they're actually the same underlying RAM block.
Fix the case where two blocks adjoin but don't overlap; that new region
should get added (but not combined), but my previous patch was disallowing it.

Fixes: c1ece84e7c9

Reported-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Tested-by: Alex Williamson <alex.williamson@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
6 years agocontrib/libvhost-user: add the protocol feature used for SET/GET message
Changpeng Liu [Fri, 30 Mar 2018 02:46:16 +0000 (10:46 +0800)]
contrib/libvhost-user: add the protocol feature used for SET/GET message

Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
6 years agovhost-user: back SET/GET_CONFIG requests with a protocol feature
Maxime Coquelin [Thu, 29 Mar 2018 07:52:33 +0000 (09:52 +0200)]
vhost-user: back SET/GET_CONFIG requests with a protocol feature

Without a dedicated protocol feature, QEMU cannot know whether
the backend can handle VHOST_USER_SET_CONFIG and
VHOST_USER_GET_CONFIG messages.

This patch adds a protocol feature that is only advertised by
QEMU if the device implements the config ops. Vhost user init
fails if the device support the feature but the backend doesn't.

The backend should only send VHOST_USER_SLAVE_CONFIG_CHANGE_MSG
requests if the protocol feature has been negotiated.

Signed-off-by: Maxime Coquelin <maxime.coquelin@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Acked-by: Changpeng Liu <changpeng.liu@intel.com>
6 years agovhost-user-blk: set config ops before vhost-user init
Maxime Coquelin [Thu, 29 Mar 2018 07:52:32 +0000 (09:52 +0200)]
vhost-user-blk: set config ops before vhost-user init

As soon as vhost-user init is done, the backend may send
VHOST_USER_SLAVE_CONFIG_CHANGE_MSG, so let's set the
notification callback before it.

Also, it will be used to know whether the device supports
the config feature to advertize it or not.

Signed-off-by: Maxime Coquelin <maxime.coquelin@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Acked-by: Changpeng Liu <changpeng.liu@intel.com>
6 years agogdbstub: fix off-by-one in gdb_handle_packet()
Philippe Mathieu-Daudé [Sun, 8 Apr 2018 14:59:33 +0000 (11:59 -0300)]
gdbstub: fix off-by-one in gdb_handle_packet()

memtohex() adds an extra trailing NUL character.

Reported-by: AddressSanitizer
Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 20180408145933.1149-1-f4bug@amsat.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
6 years agos390x: load_psw() should only exchange the PSW for KVM
David Hildenbrand [Mon, 9 Apr 2018 11:30:19 +0000 (13:30 +0200)]
s390x: load_psw() should only exchange the PSW for KVM

Let's simplify it a bit. On some weird circumstances we would have
tried to recompute watchpoints when running under KVM. load_psw() is
called from do_restart_interrupt() during a SIGP RESTART if the target
CPU is STOPPED. Let's touch watchpoints only in the TCG case - where
they are used for PER emulation.

Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20180409113019.14568-3-david@redhat.com>
Acked-by: Christian Borntraeger <borntraeger@de.ibm.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>
6 years agos390x/mmu: don't overwrite pending exception in mmu translate
David Hildenbrand [Mon, 9 Apr 2018 11:30:18 +0000 (13:30 +0200)]
s390x/mmu: don't overwrite pending exception in mmu translate

If we already triggered another exception, don't overwrite it with a
protection exception.

Only applies to old KVM instances without the virtual memory access
IOCTL in KVM.

Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20180409113019.14568-2-david@redhat.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Acked-by: Christian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>
6 years agovfio-ccw: fix memory leaks in vfio_ccw_realize()
Greg Kurz [Sat, 7 Apr 2018 14:43:46 +0000 (16:43 +0200)]
vfio-ccw: fix memory leaks in vfio_ccw_realize()

If the subchannel is already attached or if vfio_get_device() fails, the
code jumps to the 'out_device_err' label and doesn't free the string it
has just allocated.

The code should be reworked so that vcdev->vdev.name only gets set when
the device has been attached, and freed when it is about to be detached.
This could be achieved  with the addition of a vfio_ccw_get_device()
function that would be the counterpart of vfio_put_device(). But this is
a more elaborate cleanup that should be done in a follow-up. For now,
let's just add calls to g_free() on the buggy error paths.

Signed-off-by: Greg Kurz <groug@kaod.org>
Message-Id: <152311222681.203086.8874800175539040298.stgit@bahia>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>
6 years agopc-bios/s390: update images
Cornelia Huck [Fri, 6 Apr 2018 12:03:28 +0000 (08:03 -0400)]
pc-bios/s390: update images

Contains the following commits:
- s390: Do not pass inofficial IPL type to the guest

For s390-netboot.img, this also contains the following commits (update
was forgotten last time):
- pc-bios/s390-ccw: Move string arrays from bootmap header to .c file
- pc-bios/s390-ccw: Increase virtio timeout to 30 seconds

Signed-off-by: Cornelia Huck <cohuck@redhat.com>
6 years agos390: Do not pass inofficial IPL type to the guest
Viktor Mihajlovski [Thu, 5 Apr 2018 15:07:24 +0000 (17:07 +0200)]
s390: Do not pass inofficial IPL type to the guest

IPL over a virtio-scsi device requires special handling not
available in the real architecture. For this purpose the IPL
type 0xFF has been chosen as means of communication between
QEMU and the pc-bios. However, a guest OS could be confused
by seeing an unknown IPL type.

This change sets the IPL parameter type to 0x02 (CCW) to prevent
this. Pre-existing Linux has looked up the IPL parameters only in
the case of FCP IPL. This means that the behavior should stay
the same even if Linux checks for the IPL type unconditionally.

Signed-off-by: Viktor Mihajlovski <mihajlov@linux.vnet.ibm.com>
Message-Id: <1522940844-12336-4-git-send-email-mihajlov@linux.vnet.ibm.com>
Reviewed-by: Christian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>
6 years agos390: Ensure IPL from SCSI works as expected
Viktor Mihajlovski [Thu, 5 Apr 2018 15:07:23 +0000 (17:07 +0200)]
s390: Ensure IPL from SCSI works as expected

Operating systems may request an IPL from a virtio-scsi device
by specifying an IPL parameter type of CCW. In this case QEMU
won't set up the IPLB correctly. The BIOS will still detect
it's a SCSI device to boot from, but it will now have to search
for the first LUN and attempt to boot from there.
However this may not be the original boot LUN if there's more than
one SCSI disk attached to the HBA.

With this change QEMU will detect that the request is for a
SCSI device and will rebuild the initial IPL parameter info
if it's the SCSI device used for the first boot. In consequence
the BIOS can use the boot LUN from the IPL information block.

In case a different SCSI device has been set, the BIOS will find
and use the first available LUN.

Signed-off-by: Viktor Mihajlovski <mihajlov@linux.vnet.ibm.com>
Message-Id: <1522940844-12336-3-git-send-email-mihajlov@linux.vnet.ibm.com>
Reviewed-by: Farhan Ali <alifm@linux.vnet.ibm.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>
6 years agos390: Refactor IPL parameter block generation
Viktor Mihajlovski [Thu, 5 Apr 2018 15:07:22 +0000 (17:07 +0200)]
s390: Refactor IPL parameter block generation

Splitting out the the CCW device extraction allows reuse.

Signed-off-by: Viktor Mihajlovski <mihajlov@linux.vnet.ibm.com>
Message-Id: <1522940844-12336-2-git-send-email-mihajlov@linux.vnet.ibm.com>
Reviewed-by: Farhan Ali <alifm@linux.vnet.ibm.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>
6 years agos390x/kvm: call cpu_synchronize_state() on every kvm_arch_handle_exit()
David Hildenbrand [Fri, 6 Apr 2018 09:35:52 +0000 (11:35 +0200)]
s390x/kvm: call cpu_synchronize_state() on every kvm_arch_handle_exit()

Manually having to use cpu_synchronize_state() is error prone. And as
Christian Borntraeger discovered, e.g. handle_diag() is currently
missing a cpu_synchronize_state(), as decode_basedisp_s() uses a
general purpose register value internally.

So let's do an overall cpu_synchronize_state(), which fixes at least the
one mentioned BUG. We will clean up the superfluous cpu_synchronize_state()
calls later.

We now also call it (although maybe not neded) for
- KVM_EXIT_S390_RESET -> s390_reipl_request()
- KVM_EXIT_DEBUG -> kvm_arch_handle_debug_exit()
- unmanagable/unimplemented intercepts
- ICPT_CPU_STOP -> do_stop_interrupt() -> cpu gets halted
- Scenarios where we inject an operation exception
- handle_stsi()

I don't think any of these are performance critical. Especially as we
have all information directly contained in kvm_run, there are no
additional IOCTLs to issue on modern kernels.

Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20180406093552.13016-1-david@redhat.com>
Acked-by: Christian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>
6 years agoMerge remote-tracking branch 'remotes/kraxel/tags/ui-20180409-pull-request' into...
Peter Maydell [Mon, 9 Apr 2018 11:02:19 +0000 (12:02 +0100)]
Merge remote-tracking branch 'remotes/kraxel/tags/ui-20180409-pull-request' into staging

sdl2: fix kbd regression (compared to sdl1), cleanups.

# gpg: Signature made Mon 09 Apr 2018 10:40:21 BST
# gpg:                using RSA key 4CB6D8EED3E87138
# gpg: Good signature from "Gerd Hoffmann (work) <kraxel@redhat.com>"
# gpg:                 aka "Gerd Hoffmann <gerd@kraxel.org>"
# gpg:                 aka "Gerd Hoffmann (private) <kraxel@gmail.com>"
# Primary key fingerprint: A032 8CFF B93A 17A7 9901  FE7D 4CB6 D8EE D3E8 7138

* remotes/kraxel/tags/ui-20180409-pull-request:
  sdl2: drop dead code
  sdl2: drop QEMU_KEY_BACKSPACE special case
  sdl2: enable ctrl modifier keys for text consoles
  sdl2: track kbd modifier state unconditionally
  ui: add ctrl modifier support to kbd_put_qcode_console()
  sdl2: Remove unused epoxy include

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
6 years agomemfd: fix vhost-user-test on non-memfd capable host
Marc-André Lureau [Wed, 28 Mar 2018 12:18:04 +0000 (14:18 +0200)]
memfd: fix vhost-user-test on non-memfd capable host

On RHEL7, memfd is not supported, and vhost-user-test fails:
TEST: tests/vhost-user-test... (pid=10248)
  /x86_64/vhost-user/migrate:
  qemu-system-x86_64: -object memory-backend-memfd,id=mem,size=2M,: failed to create memfd
FAIL

There is a qemu_memfd_check() to prevent running memfd path, but it
also checks for fallback implementation. Let's specialize
qemu_memfd_check() to check memfd only, while qemu_memfd_alloc_check()
checks for the qemu_memfd_alloc() API.

Reported-by: Miroslav Rezanina <mrezanin@redhat.com>
Tested-by: Miroslav Rezanina <mrezanin@redhat.com>
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-Id: <20180328121804.16203-1-marcandre.lureau@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
6 years agoMerge remote-tracking branch 'remotes/famz/tags/testing-pull-request' into staging
Peter Maydell [Mon, 9 Apr 2018 09:21:14 +0000 (10:21 +0100)]
Merge remote-tracking branch 'remotes/famz/tags/testing-pull-request' into staging

Testing patches

# gpg: Signature made Mon 09 Apr 2018 08:18:02 BST
# gpg:                using RSA key CA35624C6A9171C6
# gpg: Good signature from "Fam Zheng <famz@redhat.com>"
# Primary key fingerprint: 5003 7CB7 9706 0F76 F021  AD56 CA35 624C 6A91 71C6

* remotes/famz/tags/testing-pull-request:
  docker: fedora: test more components
  docker: Inline "prep_fail" in run script
  tests: Fix ubuntu.i386 image initialization
  docker: dump 'config.log' if ./configure fails

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
6 years agosdl2: drop dead code
Gerd Hoffmann [Wed, 21 Mar 2018 13:50:40 +0000 (14:50 +0100)]
sdl2: drop dead code

Leftover from sdl1 -> sdl2 port.

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Message-id: 20180321135041.15768-6-kraxel@redhat.com
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
6 years agosdl2: drop QEMU_KEY_BACKSPACE special case
Gerd Hoffmann [Wed, 21 Mar 2018 13:50:39 +0000 (14:50 +0100)]
sdl2: drop QEMU_KEY_BACKSPACE special case

Not needed, kbd_put_qcode_console() will handle that for us.

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Message-id: 20180321135041.15768-5-kraxel@redhat.com
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
6 years agosdl2: enable ctrl modifier keys for text consoles
Gerd Hoffmann [Wed, 21 Mar 2018 13:50:38 +0000 (14:50 +0100)]
sdl2: enable ctrl modifier keys for text consoles

Unbreaks ctrl-pageup/pagedown scrollback.

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Message-id: 20180321135041.15768-4-kraxel@redhat.com
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
6 years agosdl2: track kbd modifier state unconditionally
Gerd Hoffmann [Wed, 21 Mar 2018 13:50:37 +0000 (14:50 +0100)]
sdl2: track kbd modifier state unconditionally

For both grapical and text consoles.

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Message-id: 20180321135041.15768-3-kraxel@redhat.com
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
6 years agoui: add ctrl modifier support to kbd_put_qcode_console()
Gerd Hoffmann [Wed, 21 Mar 2018 13:50:36 +0000 (14:50 +0100)]
ui: add ctrl modifier support to kbd_put_qcode_console()

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Message-id: 20180321135041.15768-2-kraxel@redhat.com
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
6 years agodocker: fedora: test more components
Paolo Bonzini [Tue, 13 Mar 2018 12:05:52 +0000 (13:05 +0100)]
docker: fedora: test more components

Install optional dependencies of QEMU to get better coverage.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <1520942752-19449-1-git-send-email-pbonzini@redhat.com>
Signed-off-by: Fam Zheng <famz@redhat.com>
6 years agodocker: Inline "prep_fail" in run script
Fam Zheng [Mon, 26 Mar 2018 09:03:50 +0000 (17:03 +0800)]
docker: Inline "prep_fail" in run script

We don't source common.rc where prep_fail is defined, so spell out the
commands and do what was intended.

Signed-off-by: Fam Zheng <famz@redhat.com>
Message-Id: <20180326090350.30014-1-famz@redhat.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Fam Zheng <famz@redhat.com>
6 years agotests: Fix ubuntu.i386 image initialization
Fam Zheng [Thu, 22 Mar 2018 03:47:53 +0000 (11:47 +0800)]
tests: Fix ubuntu.i386 image initialization

The apt-get commands we run through ssh expect certain features of the
tty, and refuses to work if /dev/null is used. It is ugly, but easy to
satisfy.

Actually, there is no reason to hide the output. It just makes things
harder to diagnose. We can always redirect in the Makefile, so don't do
it conditionally here.

Reported-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Fam Zheng <famz@redhat.com>
Message-Id: <20180322034753.6301-1-famz@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Tested-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Fam Zheng <famz@redhat.com>
6 years agodocker: dump 'config.log' if ./configure fails
Philippe Mathieu-Daudé [Thu, 15 Mar 2018 14:27:13 +0000 (15:27 +0100)]
docker: dump 'config.log' if ./configure fails

Suggested-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Message-Id: <20180315142713.30960-1-f4bug@amsat.org>
Signed-off-by: Fam Zheng <famz@redhat.com>
6 years agoMerge remote-tracking branch 'remotes/berrange/tags/qcrypto-next-pull-request' into...
Peter Maydell [Fri, 6 Apr 2018 16:51:20 +0000 (17:51 +0100)]
Merge remote-tracking branch 'remotes/berrange/tags/qcrypto-next-pull-request' into staging

# gpg: Signature made Fri 06 Apr 2018 14:29:10 BST
# gpg:                using RSA key BE86EBB415104FDF
# gpg: Good signature from "Daniel P. Berrange <dan@berrange.com>"
# gpg:                 aka "Daniel P. Berrange <berrange@redhat.com>"
# Primary key fingerprint: DAF3 A6FD B26B 6291 2D0E  8E3F BE86 EBB4 1510 4FDF

* remotes/berrange/tags/qcrypto-next-pull-request:
  crypto: ensure we use a predictable TLS priority setting
  docs: Document -object tls-creds-x509 priority=xxx
  docs: update information for TLS certificate management

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
6 years agoMerge remote-tracking branch 'remotes/rth/tags/pull-hppa-20180407' into staging
Peter Maydell [Fri, 6 Apr 2018 14:58:48 +0000 (15:58 +0100)]
Merge remote-tracking branch 'remotes/rth/tags/pull-hppa-20180407' into staging

Update hppa-softmmu bios

# gpg: Signature made Fri 06 Apr 2018 14:15:31 BST
# gpg:                using RSA key 64DF38E8AF7E215F
# gpg: Good signature from "Richard Henderson <richard.henderson@linaro.org>"
# Primary key fingerprint: 7A48 1E78 868B 4DB6 A85A  05C0 64DF 38E8 AF7E 215F

* remotes/rth/tags/pull-hppa-20180407:
  Update seabios-hppa

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
6 years agoMerge remote-tracking branch 'remotes/rth/tags/pull-tcg-20180407' into staging
Peter Maydell [Fri, 6 Apr 2018 13:26:45 +0000 (14:26 +0100)]
Merge remote-tracking branch 'remotes/rth/tags/pull-tcg-20180407' into staging

Fix generic host vector compares.

# gpg: Signature made Fri 06 Apr 2018 14:10:38 BST
# gpg:                using RSA key 64DF38E8AF7E215F
# gpg: Good signature from "Richard Henderson <richard.henderson@linaro.org>"
# Primary key fingerprint: 7A48 1E78 868B 4DB6 A85A  05C0 64DF 38E8 AF7E 215F

* remotes/rth/tags/pull-tcg-20180407:
  tcg: Fix out-of-line generic vector compares

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
6 years agoUpdate seabios-hppa
Richard Henderson [Mon, 2 Apr 2018 02:02:36 +0000 (12:02 +1000)]
Update seabios-hppa

A dozen or so fixes from Helge upstream.

6 years agotcg: Fix out-of-line generic vector compares
Richard Henderson [Fri, 6 Apr 2018 00:53:38 +0000 (00:53 +0000)]
tcg: Fix out-of-line generic vector compares

A mistake in the type passed to sizeof, that happens to work
when the out-of-line fallback itself is using host vectors,
but fails when using only the base types.

Tested-by: Emilio G. Cota <cota@braap.org>
Reported-by: Emilio G. Cota <cota@braap.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
6 years agocrypto: ensure we use a predictable TLS priority setting
Daniel P. Berrangé [Wed, 28 Feb 2018 14:04:38 +0000 (14:04 +0000)]
crypto: ensure we use a predictable TLS priority setting

The TLS test cert generation relies on a fixed set of algorithms that are
only usable under GNUTLS' default priority setting. When building QEMU
with a custom distro specific priority setting, this can cause the TLS
tests to fail. By forcing the tests to always use "NORMAL" priority we
can make them more robust.

Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
6 years agodocs: Document -object tls-creds-x509 priority=xxx
Christophe Fergeau [Fri, 8 Dec 2017 14:14:30 +0000 (15:14 +0100)]
docs: Document -object tls-creds-x509 priority=xxx

This was added in 13f1243, but is missing from qemu-options.hx

Signed-off-by: Christophe Fergeau <cfergeau@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
6 years agodocs: update information for TLS certificate management
Daniel P. Berrange [Fri, 8 Dec 2017 11:28:55 +0000 (11:28 +0000)]
docs: update information for TLS certificate management

The current docs for TLS assume only VNC is using TLS. Some of the information
is also outdated (ie lacking subject alt name info for certs). Rewrite it to
more accurately reflect the current situation.

Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Kashyap Chamarthy <kchamart@redhat.com>
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
6 years agoMerge remote-tracking branch 'remotes/awilliam/tags/vfio-fixes-20180405.0' into staging
Peter Maydell [Fri, 6 Apr 2018 09:05:32 +0000 (10:05 +0100)]
Merge remote-tracking branch 'remotes/awilliam/tags/vfio-fixes-20180405.0' into staging

VFIO fixes 2018-04-05

 - Convert over zealous error_report to trace point and make more
   meaningful (Eric Auger)

# gpg: Signature made Thu 05 Apr 2018 19:18:06 BST
# gpg:                using RSA key 239B9B6E3BB08B22
# gpg: Good signature from "Alex Williamson <alex.williamson@redhat.com>"
# gpg:                 aka "Alex Williamson <alex@shazbot.org>"
# gpg:                 aka "Alex Williamson <alwillia@redhat.com>"
# gpg:                 aka "Alex Williamson <alex.l.williamson@gmail.com>"
# Primary key fingerprint: 42F6 C04E 540B D1A9 9E7B  8A90 239B 9B6E 3BB0 8B22

* remotes/awilliam/tags/vfio-fixes-20180405.0:
  vfio: Use a trace point when a RAM section cannot be DMA mapped

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
6 years agovfio: Use a trace point when a RAM section cannot be DMA mapped
Eric Auger [Wed, 4 Apr 2018 20:30:50 +0000 (22:30 +0200)]
vfio: Use a trace point when a RAM section cannot be DMA mapped

Commit 567b5b309abe ("vfio/pci: Relax DMA map errors for MMIO regions")
added an error message if a passed memory section address or size
is not aligned to the page size and thus cannot be DMA mapped.

This patch fixes the trace by printing the region name and the
memory region section offset within the address space (instead of
offset_within_region).

We also turn the error_report into a trace event. Indeed, In some
cases, the traces can be confusing to non expert end-users and
let think the use case does not work (whereas it works as before).

This is the case where a BAR is successively mapped at different
GPAs and its sections are not compatible with dma map. The listener
is called several times and traces are issued for each intermediate
mapping.  The end-user cannot easily match those GPAs against the
final GPA output by lscpi. So let's keep those information to
informed users. In mid term, the plan is to advise the user about
BAR relocation relevance.

Fixes: 567b5b309abe ("vfio/pci: Relax DMA map errors for MMIO regions")
Signed-off-by: Eric Auger <eric.auger@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
6 years agoscripts/checkpatch.pl: Bug fix
Su Hang [Mon, 26 Mar 2018 02:06:22 +0000 (10:06 +0800)]
scripts/checkpatch.pl: Bug fix

Commit 2b9aef6fcd96ba7ed8c1ee723e391901852d344c introduced a regression:
checkpatch.pl started complaining about the following valid pattern:
do {
     /* something */
} while (condition);

Fix the script to once again permit this pattern.

Signed-off-by: Su Hang <suhang16@mails.ucas.ac.cn>
Message-Id: <1522029982-4650-1-git-send-email-suhang16@mails.ucas.ac.cn>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
6 years agotarget/i386: Fix andn instruction
Alexandro Sanchez Bach [Thu, 5 Apr 2018 12:40:58 +0000 (14:40 +0200)]
target/i386: Fix andn instruction

In commit 7073fbada733c8d10992f00772c9b9299d740e9b, the `andn` instruction
was implemented via `tcg_gen_andc` but passes the operands in the wrong
order:
- X86 defines `andn dest,src1,src2` as: dest = ~src1 & src2
- TCG defines `andc dest,src1,src2` as: dest = src1 & ~src2

The following simple test shows the issue:

    #include <stdio.h>
    #include <stdint.h>

    int main(void) {
        uint32_t ret = 0;
        __asm (
            "mov $0xFF00, %%ecx\n"
            "mov $0x0F0F, %%eax\n"
            "andn %%ecx, %%eax, %%ecx\n"
            "mov %%ecx, %0\n"
          : "=r" (ret));
        printf("%08X\n", ret);
        return 0;
    }

This patch fixes the problem by simply swapping the order of the two last
arguments in `tcg_gen_andc_tl`.

Reported-by: Alexandro Sanchez Bach <alexandro@phi.nz>
Signed-off-by: Alexandro Sanchez Bach <alexandro@phi.nz>
Cc: qemu-stable@nongnu.org
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
6 years agosys_membarrier: fix up include directives
Bruce Rogers [Thu, 29 Mar 2018 15:10:18 +0000 (09:10 -0600)]
sys_membarrier: fix up include directives

Our rule right now is to use <> for external headers only.
util/sys_membarrier.c violates that. Fix it up.

Signed-off-by: Bruce Rogers <brogers@suse.com>
Message-Id: <20180329151018.15319-1-brogers@suse.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
6 years agoUpdate version for v2.12.0-rc2 release
Peter Maydell [Wed, 4 Apr 2018 19:37:20 +0000 (20:37 +0100)]
Update version for v2.12.0-rc2 release

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
6 years agoMerge remote-tracking branch 'remotes/cody/tags/block-pull-request' into staging
Peter Maydell [Wed, 4 Apr 2018 16:48:18 +0000 (17:48 +0100)]
Merge remote-tracking branch 'remotes/cody/tags/block-pull-request' into staging

# gpg: Signature made Wed 04 Apr 2018 17:07:57 BST
# gpg:                using RSA key BDBE7B27C0DE3057
# gpg: Good signature from "Jeffrey Cody <jcody@redhat.com>"
# gpg:                 aka "Jeffrey Cody <jeff@codyprime.org>"
# gpg:                 aka "Jeffrey Cody <codyprime@gmail.com>"
# Primary key fingerprint: 9957 4B4D 3474 90E7 9D98  D624 BDBE 7B27 C0DE 3057

* remotes/cody/tags/block-pull-request:
  block/rbd: remove processed options from qdict

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
6 years agoblock/rbd: remove processed options from qdict
Jeff Cody [Wed, 4 Apr 2018 15:40:45 +0000 (11:40 -0400)]
block/rbd: remove processed options from qdict

Commit 4bfb274 added some QAPIfication of option parsing in
qemu_rbd_open().  We need to remove all the options we processed,
otherwise in bdrv_open_inherit() we will think the remaining options are
invalid.

(This needs to go in 2.12 to avoid a regression that prevents rbd
from being opened.)

Suggested-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Jeff Cody <jcody@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
6 years agotcg: fix 16-byte vector operations detection
Laurent Vivier [Wed, 28 Mar 2018 13:31:52 +0000 (15:31 +0200)]
tcg: fix 16-byte vector operations detection

configure tries to detect if the compiler
supports 16-byte vector operations.

As stated in the comment of the detection
program, there is a problem with the system
compiler on GCC on Centos 7.

This program doesn't actually detect the problem
with GCC on RHEL7 on PPC64LE (Red Hat 4.8.5-28).

This patch updates the test to look more like
it is in QEMU helpers, and now detects the problem.

The error reported is:

  CC      ppc64-softmmu/accel/tcg/tcg-runtime-gvec.o
  ..//accel/tcg/tcg-runtime-gvec.c: In function ‘helper_gvec_shl8i’:
  ../accel/tcg/tcg-runtime-gvec.c:558:26: internal compiler error: in emit_move_insn, at expr.c:3495
           *(vec8 *)(d + i) = *(vec8 *)(a + i) << shift;
                            ^
Fixes: db43267 "tcg: Add generic vector expanders"
Signed-off-by: Laurent Vivier <lvivier@redhat.com>
Reviewed-by: Miroslav Rezanina <mrezanin@redhat.com>
Message-id: 20180328133152.24623-1-lvivier@redhat.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
6 years agosdl2: Remove unused epoxy include
Elie Tournier [Wed, 4 Apr 2018 09:30:40 +0000 (10:30 +0100)]
sdl2: Remove unused epoxy include

Signed-off-by: Elie Tournier <elie.tournier@collabora.com>
Message-id: 20180404093040.26009-1-tournier.elie@gmail.com
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
6 years agoMerge remote-tracking branch 'remotes/kevin/tags/for-upstream' into staging
Peter Maydell [Wed, 4 Apr 2018 13:00:07 +0000 (14:00 +0100)]
Merge remote-tracking branch 'remotes/kevin/tags/for-upstream' into staging

Block layer patches

# gpg: Signature made Tue 03 Apr 2018 16:48:53 BST
# gpg:                using RSA key 7F09B272C88F2FD6
# gpg: Good signature from "Kevin Wolf <kwolf@redhat.com>"
# Primary key fingerprint: DC3D EB15 9A9A F95D 3D74  56FE 7F09 B272 C88F 2FD6

* remotes/kevin/tags/for-upstream:
  iotests: Test abnormally large size in compressed cluster descriptor
  qemu-iotests: Use ppc64 qemu_arch on ppc64le host
  iotests: Test preallocated truncate of 2G image
  block/file-posix: Fix fully preallocated truncate
  iotests: fix 208 for luks format
  iotests: Update 186 after commit ac64273c66ab136c44043259162
  iotests: Update 051 and 186 after commit 1454509726719e0933c
  block: handle invalid lseek returns gracefully
  gluster: Fix blockdev-add with server.N.type=unix

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
6 years agoMerge remote-tracking branch 'remotes/cody/tags/block-pull-request' into staging
Peter Maydell [Wed, 4 Apr 2018 11:33:23 +0000 (12:33 +0100)]
Merge remote-tracking branch 'remotes/cody/tags/block-pull-request' into staging

# gpg: Signature made Tue 03 Apr 2018 17:10:22 BST
# gpg:                using RSA key BDBE7B27C0DE3057
# gpg: Good signature from "Jeffrey Cody <jcody@redhat.com>"
# gpg:                 aka "Jeffrey Cody <jeff@codyprime.org>"
# gpg:                 aka "Jeffrey Cody <codyprime@gmail.com>"
# Primary key fingerprint: 9957 4B4D 3474 90E7 9D98  D624 BDBE 7B27 C0DE 3057

* remotes/cody/tags/block-pull-request:
  gluster: Fix blockdev-add with server.N.type=unix
  blockjob: use qapi enum helpers
  blockjob: leak fix, remove from txn when failing early

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
6 years agoMerge remote-tracking branch 'remotes/ehabkost/tags/x86-next-pull-request' into staging
Peter Maydell [Wed, 4 Apr 2018 10:13:52 +0000 (11:13 +0100)]
Merge remote-tracking branch 'remotes/ehabkost/tags/x86-next-pull-request' into staging

Fix memory leaks when using object_property_get_str()

# gpg: Signature made Tue 03 Apr 2018 15:00:10 BST
# gpg:                using RSA key 2807936F984DC5A6
# gpg: Good signature from "Eduardo Habkost <ehabkost@redhat.com>"
# Primary key fingerprint: 5A32 2FD5 ABC4 D3DB ACCF  D1AA 2807 936F 984D C5A6

* remotes/ehabkost/tags/x86-next-pull-request:
  sev/i386: fix memory leak in sev_guest_init()
  exec: fix memory leak in find_max_supported_pagesize()

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
6 years agoMerge remote-tracking branch 'remotes/vivier2/tags/linux-user-for-2.12-pull-request...
Peter Maydell [Wed, 4 Apr 2018 08:36:14 +0000 (09:36 +0100)]
Merge remote-tracking branch 'remotes/vivier2/tags/linux-user-for-2.12-pull-request' into staging

# gpg: Signature made Tue 03 Apr 2018 11:33:31 BST
# gpg:                using RSA key F30C38BD3F2FBE3C
# gpg: Good signature from "Laurent Vivier <lvivier@redhat.com>"
# gpg:                 aka "Laurent Vivier <laurent@vivier.eu>"
# gpg:                 aka "Laurent Vivier (Red Hat) <lvivier@redhat.com>"
# Primary key fingerprint: CD2F 75DD C8E3 A4DC 2E4F  5173 F30C 38BD 3F2F BE3C

* remotes/vivier2/tags/linux-user-for-2.12-pull-request:
  linux-user: fix TARGET___O_TMPFILE for sparc
  linux-user: define TARGET_ARCH_HAS_KA_RESTORER
  linux-user: fix alpha signal emulation

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
6 years agoMerge remote-tracking branch 'remotes/cohuck/tags/s390x-20180403' into staging
Peter Maydell [Tue, 3 Apr 2018 22:23:58 +0000 (23:23 +0100)]
Merge remote-tracking branch 'remotes/cohuck/tags/s390x-20180403' into staging

- fix a memory leak in the ipl code introduced with this release
- increase timeout in the bios to avoid hangs during migration (and
  rebuild bios to activate the change)

# gpg: Signature made Tue 03 Apr 2018 09:28:30 BST
# gpg:                using RSA key DECF6B93C6F02FAF
# gpg: Good signature from "Cornelia Huck <conny@cornelia-huck.de>"
# gpg:                 aka "Cornelia Huck <huckc@linux.vnet.ibm.com>"
# gpg:                 aka "Cornelia Huck <cornelia.huck@de.ibm.com>"
# gpg:                 aka "Cornelia Huck <cohuck@kernel.org>"
# gpg:                 aka "Cornelia Huck <cohuck@redhat.com>"
# Primary key fingerprint: C3D0 D66D C362 4FF6 A8C0  18CE DECF 6B93 C6F0 2FAF

* remotes/cohuck/tags/s390x-20180403:
  pc-bios/s390-ccw: update image
  pc-bios/s390-ccw: Increase virtio timeout to 30 seconds
  hw/s390x: fix memory leak in s390_init_ipl_dev()

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
6 years agoMerge remote-tracking branch 'remotes/xtensa/tags/20180402-xtensa' into staging
Peter Maydell [Tue, 3 Apr 2018 18:02:46 +0000 (19:02 +0100)]
Merge remote-tracking branch 'remotes/xtensa/tags/20180402-xtensa' into staging

xtensa-specific fixes for linux-user:

- fix flushing registers for signal processing in call8 and call12 frames;
- fix PC value for restarted syscalls;
- fix sysv IPC structures;
- fix fadvise64 syscall;

generic fixes for linux-user:

- fix QEMU assertion in multithreaded application by calling cpu_copy
  under clone_lock;
- fix mq_getsetattr implementation;
- fix error propagation in clock_gettime;
- implement clock_settime.

# gpg: Signature made Mon 02 Apr 2018 18:07:08 BST
# gpg:                using RSA key 51F9CC91F83FA044
# gpg: Good signature from "Max Filippov <filippov@cadence.com>"
# gpg:                 aka "Max Filippov <max.filippov@cogentembedded.com>"
# gpg:                 aka "Max Filippov <jcmvbkbc@gmail.com>"
# Primary key fingerprint: 2B67 854B 98E5 327D CDEB  17D8 51F9 CC91 F83F A044

* remotes/xtensa/tags/20180402-xtensa:
  target/xtensa: linux-user: fix fadvise64 call
  linux-user: implement clock_settime
  linux-user: fix error propagation in clock_gettime
  target/xtensa: linux-user: fix sysv IPC structures
  linux-user: fix mq_getsetattr implementation
  linux-user: call cpu_copy under clone_lock
  target/xtensa: linux-user: rewind pc for restarted syscall
  target/xtensa: fix flush_window_regs

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
6 years agoMerge remote-tracking branch 'mreitz/tags/pull-block-2018-04-03' into queue-block
Kevin Wolf [Tue, 3 Apr 2018 15:48:45 +0000 (17:48 +0200)]
Merge remote-tracking branch 'mreitz/tags/pull-block-2018-04-03' into queue-block

A fix for preallocated truncation, a new iotest, and a fix to make the iotests work more comfortably on ppc64

# gpg: Signature made Tue Apr  3 17:40:57 2018 CEST
# gpg:                using RSA key F407DB0061D5CF40
# gpg: Good signature from "Max Reitz <mreitz@redhat.com>"
# Primary key fingerprint: 91BE B60A 30DB 3E88 57D1  1829 F407 DB00 61D5 CF40

* mreitz/tags/pull-block-2018-04-03:
  iotests: Test abnormally large size in compressed cluster descriptor
  qemu-iotests: Use ppc64 qemu_arch on ppc64le host
  iotests: Test preallocated truncate of 2G image
  block/file-posix: Fix fully preallocated truncate

Signed-off-by: Kevin Wolf <kwolf@redhat.com>
6 years agoiotests: Test abnormally large size in compressed cluster descriptor
Alberto Garcia [Thu, 29 Mar 2018 12:07:45 +0000 (15:07 +0300)]
iotests: Test abnormally large size in compressed cluster descriptor

L2 entries for compressed clusters have a field that indicates the
number of sectors used to store the data in the image.

That's however not the size of the compressed data itself, just the
number of sectors where that data is located. The actual data size is
usually not a multiple of the sector size, and therefore cannot be
represented with this field.

The way it works is that QEMU reads all the specified sectors and
starts decompressing the data until there's enough to recover the
original uncompressed cluster. If there are any bytes left that
haven't been decompressed they are simply ignored.

One consequence of this is that even if the size field is larger than
it needs to be QEMU can handle it just fine: it will read more data
from disk but it will ignore the extra bytes.

This test creates an image with two compressed clusters that use 5
sectors (2.5 KB) each, increases the size field to the maximum (8192
sectors, or 4 MB) and verifies that the data can be read without
problems.

This test is important because while the decompressed data takes
exactly one cluster, the maximum value allowed in the compressed size
field is twice the cluster size. So although QEMU won't produce images
with such large values we need to make sure that it can handle them.

Another effect of increasing the size field is that it can make
it include data from the following host cluster(s). In this case
'qemu-img check' will detect that the refcounts are not correct, and
we'll need to rebuild them.

Additionally, this patch also tests that decreasing the size corrupts
the image since the original data can no longer be recovered. In this
case QEMU returns an error when trying to read the compressed data,
but 'qemu-img check' doesn't see anything wrong if the refcounts are
consistent.

One possible task for the future is to make 'qemu-img check' verify
the sizes of the compressed clusters, by trying to decompress the data
and checking that the size stored in the L2 entry is correct.

Signed-off-by: Alberto Garcia <berto@igalia.com>
Message-id: 20180329120745.11154-1-berto@igalia.com
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
6 years agoqemu-iotests: Use ppc64 qemu_arch on ppc64le host
Lukáš Doktor [Thu, 29 Mar 2018 11:20:53 +0000 (13:20 +0200)]
qemu-iotests: Use ppc64 qemu_arch on ppc64le host

The qemu target does not always correspond to the host machine type. For
example ppc64le machine target is ppc64. Let's introduce "qemu_arch"
variable to store the matching qemu architecture related to the current
architecture and use it when auto-detecting the default qemu binary.

Signed-off-by: Lukáš Doktor <ldoktor@redhat.com>
Message-id: 20180329112053.5399-2-ldoktor@redhat.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
6 years agoiotests: Test preallocated truncate of 2G image
Max Reitz [Wed, 28 Feb 2018 13:13:15 +0000 (14:13 +0100)]
iotests: Test preallocated truncate of 2G image

Signed-off-by: Max Reitz <mreitz@redhat.com>
Message-id: 20180228131315.30194-3-mreitz@redhat.com
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
6 years agoblock/file-posix: Fix fully preallocated truncate
Max Reitz [Wed, 28 Feb 2018 13:13:14 +0000 (14:13 +0100)]
block/file-posix: Fix fully preallocated truncate

Storing the lseek() result in an int results in it overflowing when the
file is at least 2 GB big.  Then, we have a 50 % chance of the result
being "negative" and thus thinking an error occurred when actually
everything went just fine.

So we should use the correct type for storing the result: off_t.

Reported-by: Daniel P. Berrange <berrange@redhat.com>
Buglink: https://bugzilla.redhat.com/show_bug.cgi?id=1549231
Cc: qemu-stable@nongnu.org
Signed-off-by: Max Reitz <mreitz@redhat.com>
Message-id: 20180228131315.30194-2-mreitz@redhat.com
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>
6 years agoiotests: fix 208 for luks format
Vladimir Sementsov-Ogievskiy [Fri, 30 Mar 2018 14:44:12 +0000 (17:44 +0300)]
iotests: fix 208 for luks format

Support luks images creatins like in 205

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
6 years agoMerge remote-tracking branch 'remotes/ericb/tags/pull-nbd-2018-04-02' into staging
Peter Maydell [Tue, 3 Apr 2018 14:59:50 +0000 (15:59 +0100)]
Merge remote-tracking branch 'remotes/ericb/tags/pull-nbd-2018-04-02' into staging

nbd patches for 2018-04-02

- Eric Blake: nbd: Fix 32-bit compilation on BLOCK_STATUS
- Eric Blake: nbd/client: Correctly handle bad server REP_META_CONTEXT
- Eric Blake: nbd: trace meta context negotiation

# gpg: Signature made Mon 02 Apr 2018 15:15:01 BST
# gpg:                using RSA key A7A16B4A2527436A
# gpg: Good signature from "Eric Blake <eblake@redhat.com>"
# gpg:                 aka "Eric Blake (Free Software Programmer) <ebb9@byu.net>"
# gpg:                 aka "[jpeg image of size 6874]"
# Primary key fingerprint: 71C2 CC22 B1C4 6029 27D2  F3AA A7A1 6B4A 2527 436A

* remotes/ericb/tags/pull-nbd-2018-04-02:
  nbd: trace meta context negotiation
  nbd/client: Correctly handle bad server REP_META_CONTEXT
  nbd: Fix 32-bit compilation on BLOCK_STATUS

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
6 years agoiotests: Update 186 after commit ac64273c66ab136c44043259162
Alberto Garcia [Thu, 22 Mar 2018 14:45:51 +0000 (16:45 +0200)]
iotests: Update 186 after commit ac64273c66ab136c44043259162

Commit ac64273c66ab136c44 modified the output of iotest 186, changing
the QOM path of floppy drives from /machine/unattached/device[17] to
/machine/unattached/device[13].

Instead of updating the test output to reflect this change, this patch
adds a new filter that hides all QOM paths from the 'Attached to:'
line of the 'info block' command.

Signed-off-by: Alberto Garcia <berto@igalia.com>
Cc: Philippe Mathieu-Daudé <f4bug@amsat.org>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
6 years agoiotests: Update 051 and 186 after commit 1454509726719e0933c
Alberto Garcia [Thu, 22 Mar 2018 14:45:50 +0000 (16:45 +0200)]
iotests: Update 051 and 186 after commit 1454509726719e0933c

SCSI controllers are no longer created automatically for
-drive if=scsi, so this patch updates the tests that relied
on that.

Signed-off-by: Alberto Garcia <berto@igalia.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Cc: Thomas Huth <thuth@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
6 years agogluster: Fix blockdev-add with server.N.type=unix
Kevin Wolf [Tue, 3 Apr 2018 11:08:10 +0000 (13:08 +0200)]
gluster: Fix blockdev-add with server.N.type=unix

The legacy command line interface gets the socket path from an option
called 'socket'. QAPI in contract uses SocketAddress, where the
corresponding option is called 'path'.

Fix the gluster block driver to accept both 'socket' and 'path', with
'path' being the preferred syntax.

https://bugzilla.redhat.com/show_bug.cgi?id=1545155

Cc: qemu-stable@nongnu.org
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Message-id: 20180403110810.25624-1-kwolf@redhat.com
Signed-off-by: Jeff Cody <jcody@redhat.com>
6 years agoblockjob: use qapi enum helpers
Marc-André Lureau [Tue, 27 Mar 2018 15:30:11 +0000 (17:30 +0200)]
blockjob: use qapi enum helpers

QAPI generator provide #define helpers for looking up enum string.

Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: John Snow <jsnow@redhat.com>
Message-id: 20180327153011.29569-1-marcandre.lureau@redhat.com
Signed-off-by: Jeff Cody <jcody@redhat.com>
6 years agoblockjob: leak fix, remove from txn when failing early
Marc-André Lureau [Wed, 28 Mar 2018 14:09:26 +0000 (10:09 -0400)]
blockjob: leak fix, remove from txn when failing early

This fixes leaks found by ASAN such as:
  GTESTER tests/test-blockjob
=================================================================
==31442==ERROR: LeakSanitizer: detected memory leaks

Direct leak of 24 byte(s) in 1 object(s) allocated from:
    #0 0x7f88483cba38 in __interceptor_calloc (/lib64/libasan.so.4+0xdea38)
    #1 0x7f8845e1bd77 in g_malloc0 ../glib/gmem.c:129
    #2 0x7f8845e1c04b in g_malloc0_n ../glib/gmem.c:360
    #3 0x5584d2732498 in block_job_txn_new /home/elmarco/src/qemu/blockjob.c:172
    #4 0x5584d2739b28 in block_job_create /home/elmarco/src/qemu/blockjob.c:973
    #5 0x5584d270ae31 in mk_job /home/elmarco/src/qemu/tests/test-blockjob.c:34
    #6 0x5584d270b1c1 in do_test_id /home/elmarco/src/qemu/tests/test-blockjob.c:57
    #7 0x5584d270b65c in test_job_ids /home/elmarco/src/qemu/tests/test-blockjob.c:118
    #8 0x7f8845e40b69 in test_case_run ../glib/gtestutils.c:2255
    #9 0x7f8845e40f29 in g_test_run_suite_internal ../glib/gtestutils.c:2339
    #10 0x7f8845e40fd2 in g_test_run_suite_internal ../glib/gtestutils.c:2351
    #11 0x7f8845e411e9 in g_test_run_suite ../glib/gtestutils.c:2426
    #12 0x7f8845e3fe72 in g_test_run ../glib/gtestutils.c:1692
    #13 0x5584d270d6e2 in main /home/elmarco/src/qemu/tests/test-blockjob.c:377
    #14 0x7f8843641f29 in __libc_start_main (/lib64/libc.so.6+0x20f29)

Add an assert to make sure that the job doesn't have associated txn before free().

[Jeff Cody: N.B., used updated patch provided by John Snow]

Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Signed-off-by: Jeff Cody <jcody@redhat.com>
6 years agoblock: handle invalid lseek returns gracefully
Jeff Cody [Tue, 3 Apr 2018 04:37:26 +0000 (00:37 -0400)]
block: handle invalid lseek returns gracefully

In commit 223a23c198787328ae75bc65d84edf5fde33c0b6, we implemented a
workaround in the gluster driver to handle invalid values returned for
SEEK_DATA or SEEK_HOLE.

In some instances, these same invalid values can be seen in the posix
file handler as well - for example, it has been reported on FUSE gluster
mounts.

Calling assert() for these invalid values is overly harsh; we can safely
return -EIO and allow this case to be treated as a "learned nothing"
case (e.g., D4 / H4, as commented in the code).

This patch does the same thing that 223a23c198787 did for gluster.c,
except in file-posix.c

Signed-off-by: Jeff Cody <jcody@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
6 years agogluster: Fix blockdev-add with server.N.type=unix
Kevin Wolf [Tue, 3 Apr 2018 10:58:32 +0000 (12:58 +0200)]
gluster: Fix blockdev-add with server.N.type=unix

The legacy command line interface gets the socket path from an option
called 'socket'. QAPI in contract uses SocketAddress, where the
corresponding option is called 'path'.

Fix the gluster block driver to accept both 'socket' and 'path', with
'path' being the preferred syntax.

https://bugzilla.redhat.com/show_bug.cgi?id=1545155

Cc: qemu-stable@nongnu.org
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
6 years agolinux-user: fix TARGET___O_TMPFILE for sparc
Laurent Vivier [Mon, 2 Apr 2018 10:24:53 +0000 (12:24 +0200)]
linux-user: fix TARGET___O_TMPFILE for sparc

Signed-off-by: Laurent Vivier <laurent@vivier.eu>
Message-Id: <20180402102453.9883-3-laurent@vivier.eu>

6 years agolinux-user: define TARGET_ARCH_HAS_KA_RESTORER
Laurent Vivier [Mon, 2 Apr 2018 10:24:52 +0000 (12:24 +0200)]
linux-user: define TARGET_ARCH_HAS_KA_RESTORER

Sparc as an extended sigaction structure containing
the field ka_restorer used in place of sa_restorer.

Define TARGET_ARCH_HAS_KA_RESTORER and use it
with sparc.

Signed-off-by: Laurent Vivier <laurent@vivier.eu>
Message-Id: <20180402102453.9883-2-laurent@vivier.eu>

6 years agolinux-user: fix alpha signal emulation
Laurent Vivier [Sun, 1 Apr 2018 20:46:53 +0000 (22:46 +0200)]
linux-user: fix alpha signal emulation

setup_frame() doesn't set correctly the address of the trampoline code.
The offset of retcode array must be added to the stack frame address.

Signed-off-by: Laurent Vivier <laurent@vivier.eu>
Message-Id: <20180401204653.14211-1-laurent@vivier.eu>

6 years agopc-bios/s390-ccw: update image
Cornelia Huck [Thu, 29 Mar 2018 14:13:37 +0000 (10:13 -0400)]
pc-bios/s390-ccw: update image

Contains the following commits:
- pc-bios/s390-ccw: Move string arrays from bootmap header to .c file
- pc-bios/s390-ccw: Increase virtio timeout to 30 seconds

Signed-off-by: Cornelia Huck <cohuck@redhat.com>
6 years agopc-bios/s390-ccw: Increase virtio timeout to 30 seconds
Thomas Huth [Thu, 29 Mar 2018 09:37:31 +0000 (11:37 +0200)]
pc-bios/s390-ccw: Increase virtio timeout to 30 seconds

The current timeout is set to only three seconds - and considering that
vring_wait_reply() or rather get_second() is not doing any rounding,
the real timeout is likely rather 2 seconds in most cases. When the
host is really badly loaded, it's possible that we hit this timeout by
mistake; it's even more likely if we run the guest in TCG mode instead
of KVM.

So let's increase the timeout to 30 seconds instead to ease this situation
(30 seconds is also the timeout that is used by the Linux SCSI subsystem
for example, so this seems to be a sane value for block IO timeout).

Buglink: https://bugzilla.redhat.com/show_bug.cgi?id=1549079
Signed-off-by: Thomas Huth <thuth@redhat.com>
Message-Id: <1522316251-16399-1-git-send-email-thuth@redhat.com>
Acked-by: Christian Borntraeger <borntraeger@de.ibm.com>
[CH: tweaked commit message]
Signed-off-by: Cornelia Huck <cohuck@redhat.com>
6 years agohw/s390x: fix memory leak in s390_init_ipl_dev()
Greg Kurz [Thu, 29 Mar 2018 09:10:06 +0000 (11:10 +0200)]
hw/s390x: fix memory leak in s390_init_ipl_dev()

The string returned by object_property_get_str() is dynamically allocated.

Fixes: 3c4e9baacf4d9
Signed-off-by: Greg Kurz <groug@kaod.org>
Message-Id: <152231460685.69730.14860451936216690693.stgit@bahia.lan>
Reviewed-by: Cornelia Huck <cohuck@redhat.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>
6 years agosev/i386: fix memory leak in sev_guest_init()
Greg Kurz [Thu, 29 Mar 2018 09:10:21 +0000 (11:10 +0200)]
sev/i386: fix memory leak in sev_guest_init()

The string returned by object_property_get_str() is dynamically allocated.

Fixes: d8575c6c0242b
Signed-off-by: Greg Kurz <groug@kaod.org>
Message-Id: <152231462116.69730.14119625999092384450.stgit@bahia.lan>
Reviewed-by: Cornelia Huck <cohuck@redhat.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
6 years agoexec: fix memory leak in find_max_supported_pagesize()
Greg Kurz [Thu, 29 Mar 2018 09:09:46 +0000 (11:09 +0200)]
exec: fix memory leak in find_max_supported_pagesize()

The string returned by object_property_get_str() is dynamically allocated.

Signed-off-by: Greg Kurz <groug@kaod.org>
Message-Id: <152231458624.69730.1752893648612848392.stgit@bahia.lan>
Reviewed-by: Cornelia Huck <cohuck@redhat.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
6 years agonbd: trace meta context negotiation
Eric Blake [Fri, 30 Mar 2018 13:09:50 +0000 (08:09 -0500)]
nbd: trace meta context negotiation

Having a more detailed log of the interaction between client and
server is invaluable in debugging how meta context negotiation
actually works.

Signed-off-by: Eric Blake <eblake@redhat.com>
Message-Id: <20180330130950.1931229-1-eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
6 years agonbd/client: Correctly handle bad server REP_META_CONTEXT
Eric Blake [Thu, 29 Mar 2018 23:18:37 +0000 (18:18 -0500)]
nbd/client: Correctly handle bad server REP_META_CONTEXT

It's never a good idea to blindly read for size bytes as
returned by the server without first validating that the size
is within bounds; a malicious or buggy server could cause us
to hang or get out of sync from reading further messages.

It may be smarter to try and teach the client to cope with
unexpected context ids by silently ignoring them instead of
hanging up on the server, but for now, if the server doesn't
reply with exactly the one context we expect, it's easier to
just give up - however, if we give up for any reason other
than an I/O failure, we might as well try to politely tell
the server we are quitting rather than continuing.

Fix some typos in the process.

Signed-off-by: Eric Blake <eblake@redhat.com>
Message-Id: <20180329231837.1914680-1-eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
6 years agonbd: Fix 32-bit compilation on BLOCK_STATUS
Eric Blake [Tue, 27 Mar 2018 21:05:17 +0000 (16:05 -0500)]
nbd: Fix 32-bit compilation on BLOCK_STATUS

iotests 123 and 209 fail on 32-bit platforms.  The culprit:
sizeof(extent) is wrong; we want sizeof(*extent).  But since
the struct is 8 bytes, it happened to work on 64-bit platforms
where the pointer is also 8 bytes (nasty).

Fixes: 78a33ab58
Reported-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Eric Blake <eblake@redhat.com>
Message-Id: <20180327210517.1804242-1-eblake@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
6 years agotarget/xtensa: linux-user: fix fadvise64 call
Max Filippov [Sun, 1 Apr 2018 22:02:34 +0000 (15:02 -0700)]
target/xtensa: linux-user: fix fadvise64 call

fadvise64_64 on xtensa passes advice as the second argument and so must
be handled similar to PPC.

This fixes glibc testsuite tests posix/tst-posix_fadvise and
posix/tst-posix_fadvise64.

Reviewed-by: Laurent Vivier <laurent@vivier.eu>
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
6 years agolinux-user: implement clock_settime
Max Filippov [Sun, 1 Apr 2018 20:13:49 +0000 (13:13 -0700)]
linux-user: implement clock_settime

This fixes glibc testsuite test rt/tst-clock2.

Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
6 years agolinux-user: fix error propagation in clock_gettime
Max Filippov [Sun, 1 Apr 2018 20:14:04 +0000 (13:14 -0700)]
linux-user: fix error propagation in clock_gettime

host_to_target_timespec may return error if target address could not be
locked, but it is ignored.
Propagate return value of host_to_target_timespec to the caller of
clock_gettime.

Reviewed-by: Laurent Vivier <laurent@vivier.eu>
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>