This module should not write directly to stdout/stderr. Instead, it
should either raise exceptions or just log the messages and let the
callers handle them and decide what to do. For example, scripts could
choose to send the log messages stderr or/and write them to a file if
verbose or debugging mode is enabled.
This patch replaces the writes to stderr by an exception in the
send_fd_scm() when _socket_scm_helper is not set or not present. In the
same method, the subprocess Popen will now redirect the stdout/stderr to
logging.debug instead of writing to system stderr. As consequence, since
the Popen.communicate() is now used (in order to get the stdout), the
further call to wait() became redundant and was replaced by
Popen.returncode.
The shutdown() message on negative exit code will now be logged
to logging.warn instead of written to system stderr.
Lukáš Doktor [Fri, 18 Aug 2017 14:26:12 +0000 (16:26 +0200)]
qmp.py: Avoid overriding a builtin object
The "id" is a builtin method to get object's identity and should not be
overridden. This might bring some issues in case someone was directly
calling "cmd(..., id=id)" but I haven't found such usage on brief search
for "cmd\(.*id=".
Lukáš Doktor [Fri, 18 Aug 2017 14:26:09 +0000 (16:26 +0200)]
qmp.py: Couple of pylint/style fixes
No actual code changes, just initializing attributes earlier to avoid
AttributeError on early introspection, a few pylint/style fixes and
docstring clarifications.
Lukáš Doktor [Fri, 18 Aug 2017 14:26:08 +0000 (16:26 +0200)]
qemu.py: Use custom exceptions rather than Exception
The naked Exception should not be widely used. It makes sense to be a
bit more specific and use better-suited custom exceptions. As a benefit
we can store the full reply in the exception in case someone needs it
when catching the exception.
Lukáš Doktor [Fri, 18 Aug 2017 14:26:07 +0000 (16:26 +0200)]
qemu.py: Simplify QMP key-conversion
The QMP key conversion consist of '_'s to be replaced with '-'s, which
can easily be done by a single `str.replace` method which is faster and
does not require `string` module import.
In this case the `args` is actually copied so it would be safe to keep
it, but it's not a good practice to keep it. The same issue applies in
inherited qtest module.
* remotes/huth/tags/check-20170915:
qtest: Avoid passing raw strings through hmp()
libqtest: Remove dead qtest_instances variable
numa-test: Use hmp()
qtest: Don't perform side effects inside assertion
test-qga: Kill broken and dead QGA_TEST_SIDE_EFFECTING code
tests: Fix broken ivshmem-server-msi/-irq tests
tests/libqtest: Use a proper error message if QTEST_QEMU_BINARY is missing
tests/test-hmp: Remove puv3 and tricore_testboard from the blacklist
tests: Introduce generic device hot-plug/hot-unplug functions
* remotes/dgibson/tags/ppc-for-2.11-20170915:
ppc/kvm: use kvm_vm_check_extension() in kvmppc_is_pr()
spapr_events: use QTAILQ_FOREACH_SAFE() in spapr_clear_pending_events()
spapr_cpu_core: cleaning up qdev_get_machine() calls
spapr_pci: don't create 64-bit MMIO window if we don't need to
spapr_pci: convert sprintf() to g_strdup_printf()
spapr_cpu_core: fail gracefully with non-pseries machine types
xics: fix several error leaks
vfio, spapr: Fix levels calculation
spapr_pci: handle FDT creation errors with _FDT()
spapr_pci: use the common _FDT() helper
spapr: fix CAS-generated reset
ppc/xive: fix OV5_XIVE_EXPLOIT bits
spapr: only update SDR1 once per-cpu during CAS
spapr_pci: use g_strdup_printf()
spapr_pci: drop useless check in spapr_populate_pci_child_dt()
spapr_pci: drop useless check in spapr_phb_vfio_get_loc_code()
hw/ppc/spapr.c: cleaning up qdev_get_machine() calls
net: Add SunGEM device emulation as found on Apple UniNorth
Eric Blake [Mon, 11 Sep 2017 17:20:14 +0000 (12:20 -0500)]
qtest: Avoid passing raw strings through hmp()
hmp() passes its string argument through the sprintf() family;
with a proper attribute, gcc -Wformat warns us when we do something
dangerous like passing a non-constant format string. Fortunately,
all our strings were safe, but checking whether the string can
contain an unintended % is easy to avoid and therefore worth doing.
Eric Blake [Mon, 11 Sep 2017 17:19:49 +0000 (12:19 -0500)]
libqtest: Remove dead qtest_instances variable
Prior to commit 063c23d9, we were tracking a list of parallel
qtest objects, in order to safely clean up a SIGABRT handler
only after the last connection quits. But when we switched to
more of glib's infrastructure, the list became dead code that
is never assigned to.
Eric Blake [Mon, 11 Sep 2017 17:19:45 +0000 (12:19 -0500)]
test-qga: Kill broken and dead QGA_TEST_SIDE_EFFECTING code
Back when the test was introduced, in commit 62c39b307, the
test was set up to run qemu-ga directly on the host performing
the test, and defaults to limiting itself to safe commands. At
the time, it was envisioned that setting QGA_TEST_SIDE_EFFECTING
in the environment could cover a few more commands, while noting
the potential danger of those side effects running in the host.
But this has NEVER been tested: if you enable the environment
variable, the test WILL fail. One obvious reason: if you are not
running as root, you'll probably get a permission failure when
trying to freeze the file systems, or when changing system time.
Less obvious: if you run the test as root (wow, you're brave), you
could end up hanging if the test tries to log things to a
temporarily frozen filesystem. But the cutest reason of all: if
you get past the above hurdles, the test uses invalid JSON in
test_qga_fstrim() (missing '' around the dictionary key 'minimum'),
and will thus fail an assertion in qmp_fd().
Rather than leave this untested time-bomb in place, rip it out.
Hopefully, as originally envisioned, we can find an opportunity
to test an actual sandboxed guest where the guest-agent has
full permissions and will not unduly affect the host running
the test - if so, 'git revert' can be used if desired, for
salvaging any useful parts of this attempt.
Thomas Huth [Tue, 29 Aug 2017 17:03:51 +0000 (19:03 +0200)]
tests: Fix broken ivshmem-server-msi/-irq tests
Broken with commit b4ba67d9a7025 ("libqos: Change PCI accessors to take
opaque BAR handle") a while ago, but nobody noticed since the tests are
not run by default: The msix_pba_bar is not correctly initialized
anymore if bir_pba has the same value as bir_table. With this fix,
"make check SPEED=slow" should work fine again.
Thomas Huth [Mon, 28 Aug 2017 10:25:45 +0000 (12:25 +0200)]
tests/libqtest: Use a proper error message if QTEST_QEMU_BINARY is missing
The user can currently still cause an abort() if running certain tests
(like the prom-env-test) without setting the QTEST_QEMU_BINARY first.
A similar problem has been fixed with commit 7c933ad61b8f3f51337
already, but forgot to also take care of the qtest_get_arch() function,
so let's introduce a proper wrapper around getenv("QTEST_QEMU_BINARY")
that can be used in both locations now.
Thomas Huth [Thu, 24 Aug 2017 05:00:11 +0000 (07:00 +0200)]
tests/test-hmp: Remove puv3 and tricore_testboard from the blacklist
The problem with puv3 has been fixed with 0ac241bcf9f9d99a252a352a162f
('unicore32: abort when entering "x 0" on the monitor') and the problem
with tricore_testboard has been fixed with b190f477e29c7cd03a8fee49c96d
('qemu-system-tricore: segfault when entering "x 0" on the monitor').
A lot of tests provide code for adding and removing a device via the
device_add and device_del QMP commands. Maintaining this code in so many
places is cumbersome and error-prone (some of the code parts check the
responses for device deletion in an incorrect way, for example, we've got
to deal with both, error code and DEVICE_DEL event here). So let's provide
some proper generic functions for adding and removing a device instead.
The code for correctly unplugging a device has been taken from a patch
from Peter Xu.
Greg Kurz [Thu, 14 Sep 2017 10:48:04 +0000 (12:48 +0200)]
ppc/kvm: use kvm_vm_check_extension() in kvmppc_is_pr()
If the host has both KVM PR and KVM HV loaded and we pass:
-machine pseries,accel=kvm,kvm-type=PR
the kvmppc_is_pr() returns false instead of true. Since the helper
is mostly used as fallback, it doesn't have any real impact with
recent kernels. A notable exception is the workaround to allow
migration between compatible hosts with different PVRs (eg, POWER8
and POWER8E), since KVM still doesn't provide a way to check if a
specific PVR is supported (see commit c363a37a450f for details).
According to the official KVM API documentation [1], KVM_PPC_GET_PVINFO
is "vm ioctl", but we check it as a global ioctl. The following function
in KVM is hence called with kvm == NULL and considers we're in HV mode.
int kvm_vm_ioctl_check_extension(struct kvm *kvm, long ext)
{
int r;
/* Assume we're using HV mode when the HV module is loaded */
int hv_enabled = kvmppc_hv_ops ? 1 : 0;
if (kvm) {
/*
* Hooray - we know which VM type we're running on. Depend on
* that rather than the guess above.
*/
hv_enabled = is_kvmppc_hv_enabled(kvm);
}
Let's use kvm_vm_check_extension() to fix the issue.
Greg Kurz [Tue, 12 Sep 2017 06:51:21 +0000 (08:51 +0200)]
spapr_cpu_core: cleaning up qdev_get_machine() calls
This patch removes the qdev_get_machine() calls that are made
in spapr_cpu_core.c in situations where we can get an existing
pointer for the MachineState by either passing it as an argument
to the function or by using other already available pointers.
Credits to Daniel Henrique Barboza for the idea and the changelog
text.
This happens because we always create a 64-bit MMIO window, even if
we didn't explicitely requested it (ie, mem64_win_size == 0) and the
32-bit window is below 2GiB. It doesn't seem to have an impact on the
guest though because spapr_populate_pci_dt() doesn't advertise the
bogus windows when mem64_win_size == 0.
Since these memory regions don't induce any state, we can safely
choose to not create them when their address is equal to -1,
without breaking migration from existing setups.
Greg Kurz [Mon, 11 Sep 2017 20:52:06 +0000 (22:52 +0200)]
spapr_cpu_core: fail gracefully with non-pseries machine types
Since commit 7cca3e466eb0 ("ppc: spapr: Move VCPU ID calculation into
sPAPR"), QEMU aborts when started with a *-spapr-cpu-core device and
a non-pseries machine.
Let's rely on the already existing call to object_dynamic_cast() instead
of using the SPAPR_MACHINE() macro.
Greg Kurz [Mon, 11 Sep 2017 22:04:40 +0000 (00:04 +0200)]
xics: fix several error leaks
If object_property_get_link() fails then it allocates an error, which
must be freed before returning. The error_get_pretty() function is
merely an accessor to the error message and doesn't free anything.
The error.h header indicates how to do it right:
* Pass an existing error to the caller with the message modified:
* error_propagate(errp, err);
* error_prepend(errp, "Could not frobnicate '%s': ", name);
The existing tries to round up the number of pages but @pages is always
calculated as the rounded up value minus one which makes ctz64() always
return 0 and have create.levels always set 1.
This removes wrong "-1" and allows having more than 1 levels. This becomes
handy for >128GB guests with standard 64K pages as this requires blocks
with zone order 9 and the popular limit of CONFIG_FORCE_MAX_ZONEORDER=9
means that only blocks up to order 8 are allowed.
Greg Kurz [Sat, 9 Sep 2017 15:06:33 +0000 (17:06 +0200)]
spapr_pci: handle FDT creation errors with _FDT()
libfdt failures when creating the FDT should cause QEMU to terminate.
Let's use the _FDT() macro which does just that instead of propagating
the error to the caller. spapr_populate_pci_child_dt() no longer needs
to return a value in this case.
Note that, on the way, this get rids of the following nonsensical lines:
Greg Kurz [Sat, 9 Sep 2017 15:06:25 +0000 (17:06 +0200)]
spapr_pci: use the common _FDT() helper
All other users in hw/ppc already consider an error when building
the FDT to be fatal, even on hotplug paths. There's no valid reason
for spapr_pci to behave differently. So let's used the common _FDT()
helper which terminates QEMU when libfdt fails.
The OV5_MMU_RADIX_300 requires special handling in the CAS negotiation
process. It is cleared from the option vector of the guest before
evaluating the changes and re-added later. But, when testing for a
possible CAS reset :
On POWER9, the Client Architecture Support (CAS) negotiation process
determines whether the guest operates in XIVE Legacy compatibility or
in XIVE exploitation mode. Now that we have initial guest support for
the XIVE interrupt controller, let's fix the bits definition which have
evolved in the latest specs.
The platform advertises the XIVE Exploitation Mode support using the
property "ibm,arch-vec-5-platform-support-vec-5", byte 23 bits 0-1 :
- 0b00 XIVE legacy mode Only
- 0b01 XIVE exploitation mode Only
- 0b10 XIVE legacy or exploitation mode
The OS asks for XIVE Exploitation Mode support using the property
"ibm,architecture-vec-5", byte 23 bits 0-1:
- 0b00 XIVE legacy mode Only
- 0b01 XIVE exploitation mode Only
Greg Kurz [Mon, 4 Sep 2017 21:46:55 +0000 (23:46 +0200)]
spapr: only update SDR1 once per-cpu during CAS
Commit b55d295e3ec9 added the possibility to support HPT resizing with KVM.
In the case of PR, we need to pass the userspace address of the HPT to KVM
using the SDR1 slot.
This is handled by kvmppc_update_sdr1() which uses CPU_FOREACH() to update
all CPUs. It is hence not needed to call kvmppc_update_sdr1() for each CPU.
hw/ppc/spapr.c: cleaning up qdev_get_machine() calls
This patch removes the qdev_get_machine() calls that are made in
spapr.c in situations where we can get an existing pointer for
the MachineState by either passing it as an argument to the function
or by using other already available pointers.
The following changes were made:
- spapr_node0_size: static function that is called two times:
at spapr_setup_hpt_and_vrma and ppc_spapr_init. In both cases we can
pass an existing MachineState pointer to it.
- spapr_build_fdt: MachineState pointer can be retrieved from
the existing sPAPRMachineState pointer.
- spapr_boot_set: the opaque in the first arg is a sPAPRMachineState
pointer as we can see inside ppc_spapr_init:
qemu_register_boot_set(spapr_boot_set, spapr);
We can get a MachineState pointer from it.
- spapr_machine_device_plug and spapr_machine_device_unplug_request: the
MachineState, sPAPRMachineState, MachineClass and sPAPRMachineClass pointers
can all be retrieved from the HotplugHandler pointer.
net: Add SunGEM device emulation as found on Apple UniNorth
This adds a simplistic emulation of the Sun GEM ethernet controller
found in Apple ASICs.
Currently we only support the Apple UniNorth 1.x variant, but the
other Apple or Sun variants should mostly be a matter of adding
PCI IDs options.
We have a very primitive emulation of a single Broadcom 5201 PHY
which is supported by the MacOS driver.
This model brings out-of-the-box networking to MacOS 9, and all
versions of OS X I tried with the mac99 platform.
Further improvements from Mark:
- Remove sungem.h file, moving constants into sungem.c as required
- Switch to using tracepoints for debugging
- Split register blocks into separate memory regions
- Use arrays in SunGEMState to hold register values
- Add state-saving support
* remotes/pmaydell/tags/pull-target-arm-20170914:
mps2-an511: Fix wiring of UART overflow interrupt lines
hw/pci-host/gpex: Implement PCI INTx routing
hw/arm/virt: Set INTx/gsi mapping
hw/pci-host/gpex: Set INTx index/gsi mapping
target/arm: Avoid an extra temporary for store_exclusive
AArch64: Fix single stepping of ERET instruction
xlnx-zcu102: Mark the EP108 machine as deprecated
xlnx-zcu102: Add a machine level virtualization property
xlnx-zcu102: Add a machine level secure property
xlnx-zcu102: Manually create the machines
xlnx-ep108: Rename to ZCU102
target/arm: Rename 'type' to 'excret' in do_v7m_exception_exit()
target/arm: Add and use defines for EXCRET constants
target/arm: Remove unnecessary '| 0xf0000000' from do_v7m_exception_exit()
nvic: Don't apply group priority mask to negative priorities
target/arm: Get PRECISERR and IBUSERR the right way round
target/arm: Clear exclusive monitor on v7M reset, exception entry/exit
target/arm: Use M_REG_NUM_BANKS rather than hardcoding 2
Peter Maydell [Thu, 14 Sep 2017 17:43:19 +0000 (18:43 +0100)]
mps2-an511: Fix wiring of UART overflow interrupt lines
Fix an error that meant we were wiring every UART's overflow
interrupts into the same inputs 0 and 1 of the OR gate,
rather than giving each its own input.
Now we are able to retrieve the gsi from the INTx pin, let's
enable intx_to_irq routing. From that point on, irqfd becomes
usable along with INTx when assigning a PCIe device.
To implement INTx to gsi routing we need to pass the gpex host
bridge the gsi associated to each INTx index. Let's introduce
irq_num array and gpex_set_irq_num setter function.
Previously when single stepping through ERET instruction via GDB
would result in debugger entering the "next" PC after ERET instruction.
When debugging in kernel mode, this will also cause unintended behavior,
because debugger will try to access memory from EL0 point of view.
Alistair Francis [Thu, 14 Sep 2017 17:43:18 +0000 (18:43 +0100)]
xlnx-zcu102: Add a machine level virtualization property
Add a machine level virtualization property. This defaults to false and can be
set to true using this machine command line argument:
-machine xlnx-zcu102,virtualization=on
This follows what the ARM virt machine does.
This property only applies to the ZCU102 machine. The EP108 machine does
not have this property.
Alistair Francis [Thu, 14 Sep 2017 17:43:18 +0000 (18:43 +0100)]
xlnx-zcu102: Add a machine level secure property
Add a machine level secure property. This defaults to false and can be
set to true using this machine command line argument:
-machine xlnx-zcu102,secure=on
This follows what the ARM virt machine does.
This property only applies to the ZCU102 machine. The EP108 machine does
not have this property.
Alistair Francis [Thu, 14 Sep 2017 17:43:17 +0000 (18:43 +0100)]
xlnx-ep108: Rename to ZCU102
The EP108 is a early access development board. Now that silicon is in
production people have access to the ZCU102. Let's rename the internal
QEMU files and variables to use the ZCU102.
There is no functional change here as the EP108 is still a valid board
option.
Peter Maydell [Thu, 14 Sep 2017 17:43:17 +0000 (18:43 +0100)]
target/arm: Rename 'type' to 'excret' in do_v7m_exception_exit()
In the v7M and v8M ARM ARM, the magic exception return values are
referred to as EXC_RETURN values, and in QEMU we use V7M_EXCRET_*
constants to define bits within them. Rename the 'type' variable
which holds the exception return value in do_v7m_exception_exit()
to excret, making it clearer that it does hold an EXC_RETURN value.
Peter Maydell [Thu, 14 Sep 2017 17:43:17 +0000 (18:43 +0100)]
target/arm: Add and use defines for EXCRET constants
The exception-return magic values get some new bits in v8M, which
makes some bit definitions for them worthwhile.
We don't use the bit definitions for the switch on the low bits
which checks the return type for v7M, because this is defined
in the v7M ARM ARM as a set of valid values rather than via
per-bit checks.
Peter Maydell [Thu, 14 Sep 2017 17:43:17 +0000 (18:43 +0100)]
target/arm: Remove unnecessary '| 0xf0000000' from do_v7m_exception_exit()
In do_v7m_exception_exit(), there's no need to force the high 4
bits of 'type' to 1 when calling v7m_exception_taken(), because
we know that they're always 1 or we could not have got to this
"handle return to magic exception return address" code. Remove
the unnecessary ORs.
Peter Maydell [Thu, 14 Sep 2017 17:43:17 +0000 (18:43 +0100)]
nvic: Don't apply group priority mask to negative priorities
In several places we were unconditionally applying the
nvic_gprio_mask() to a priority value. This is incorrect
if the priority is one of the fixed negative priority
values (for NMI and HardFault), so don't do it.
This bug would have caused both NMI and HardFault to be
considered as the same priority and so NMI wouldn't
correctly preempt HardFault.
Peter Maydell [Thu, 14 Sep 2017 17:43:17 +0000 (18:43 +0100)]
target/arm: Get PRECISERR and IBUSERR the right way round
For a bus fault, the M profile BFSR bit PRECISERR means a bus
fault on a data access, and IBUSERR means a bus fault on an
instruction access. We had these the wrong way around; fix this.
Peter Maydell [Thu, 14 Sep 2017 17:43:16 +0000 (18:43 +0100)]
target/arm: Use M_REG_NUM_BANKS rather than hardcoding 2
Use a symbolic constant M_REG_NUM_BANKS for the array size for
registers which are banked by M profile security state, rather
than hardcoding lots of 2s.
Peter Maydell [Thu, 14 Sep 2017 15:33:02 +0000 (16:33 +0100)]
Merge remote-tracking branch 'remotes/dgilbert/tags/pull-hmp-20170914' into staging
HMP pull 2017-09-14
# gpg: Signature made Thu 14 Sep 2017 15:57:30 BST
# gpg: using RSA key 0x0516331EBC5BFDE7
# gpg: Good signature from "Dr. David Alan Gilbert (RH2) <[email protected]>"
# gpg: WARNING: This key is not certified with sufficiently trusted signatures!
# gpg: It is not certain that the signature belongs to the owner.
# Primary key fingerprint: 45F5 C71B 4A0C B7FB 977A 9FA9 0516 331E BC5B FDE7
* remotes/dgilbert/tags/pull-hmp-20170914:
hmp: introduce 'info memory_size_summary' command
qmp: introduce query-memory-size-summary command
hmp: extend "info numa" with hotplugged memory information
tests/hmp: test "none" machine with memory
dump: do not dump non-existent guest memory
hmp: fix "dump-quest-memory" segfault (arm)
hmp: fix "dump-quest-memory" segfault (ppc)
Vadim Galitsyn [Tue, 29 Aug 2017 15:30:22 +0000 (17:30 +0200)]
hmp: introduce 'info memory_size_summary' command
Add 'info memory_size_summary' command which is a sibling
of QMP command query-memory-size-summary. It provides the
following memory information in bytes:
* base-memory - size of "base" memory specified with command line option -m.
* plugged-memory - amount of memory that was hot-plugged.
If target does not have CONFIG_MEM_HOTPLUG enabled, no
value is reported.
* remotes/kraxel/tags/usb-20170913-pull-request:
usb: only build usb-host with CONFIG_USB=y
usb: drop HOST_USB
MAINTAINERS: add missing USB entry
xhci: Avoid DMA when ERSTBA is set to zero
* remotes/kraxel/tags/vga-20170913-pull-request:
virtio-gpu: don't clear QemuUIInfo information on reset
vga/migration: Update memory map in post_load
qxl: add support for chunked cursors.
qxl: drop mono cursor support
vga: stop passing pointers to vga_draw_line* functions
vga: fix display update region calculation (split screen)
Nowdays we use libusb for usb-host, so we don't have different code
for linux vs. bsd any more. So there is little reason to have the
HOST_USB variable, we can just write things directly into the Makefile
and avoid a pointless indirection.
The existing XHCI code reads the Event Ring Segment Table Base Address
Register (ERSTBA) every time when it is changed. However zero is its
default state so one would think that zero there means it is not in use.
This adds a check for ERSTBA in addition to the existing check for
the Event Ring Segment Table Size Register (ERSTSZ).
Alexander Graf [Tue, 29 Aug 2017 11:38:18 +0000 (13:38 +0200)]
console: add question-mark escape operator
Some termcaps (found using SLES11SP1) use [? sequences. According to man
console_codes (http://linux.die.net/man/4/console_codes) the question mark
is a nop and should simply be ignored.
This patch does exactly that, rendering screen output readable when
outputting guest serial consoles to the graphical console emulator.
virtio-gpu can trigger the assert added by commit "6905b93447 console:
add same surface replace pre-condition" in multihead setups (where
surface can be NULL for secondary displays). Allow surface being NULL.
Drop pixman submodule and support for the "internal" pixman build.
pixman should be reasonably well established meanwhile so we don't
need the fallback submodule any more. While being at it also drop
some #ifdefs for pixman versions older than what we require in
configure anyway.
virtio-gpu: don't clear QemuUIInfo information on reset
Don't reset window layout information (passed via virtio_gpu_ui_info) on
device reset, so the user interface window layout will be kept intact
over reboots. The head size and position was commented out already, so
this patch just drops the dead code. Additionally the enabled head mask
must be kept so multihead setups work properly too.
tcg/tci: do not use ldst label (never implemented)
changed in 659ef5cbb893, this fixes building with --enable-tcg-interpreter:
/home/travis/build/qemu/qemu/tcg/tcg.c:116:14: error: ‘tcg_out_ldst_finalize’ used but never defined [-Werror]
static bool tcg_out_ldst_finalize(TCGContext *s);
^
target/m68k: Switch fpu_rom from make_floatx80() to make_floatx80_init()
GCC 4.7.2 on SunOS reports that the values assigned to array members are not
real constants:
target/m68k/fpu_helper.c:32:5: error: initializer element is not constant
target/m68k/fpu_helper.c:32:5: error: (near initialization for 'fpu_rom[0]')
rules.mak:66: recipe for target 'target/m68k/fpu_helper.o' failed
Convert the array to make_floatx80_init() to fix it.
Replace floatx80_pi-like constants with make_floatx80_init() as they are
defined as make_floatx80().
Peter Maydell [Fri, 8 Sep 2017 13:44:44 +0000 (14:44 +0100)]
Merge remote-tracking branch 'remotes/dgibson/tags/ppc-for-2.11-20170908' into staging
ppc patch queue 2017-09-08
This is the first batch of ppc related patches for qemu-2.11, and it's
accumulated quite a few things. Includes:
* A cleanup to handling of ppc cpu models from Igor
* First parts of fixes to handling of guest vs. host SMT modes from
Sam Bobroff
* Preliminary patches towards supporting the Sam460 board from
Balaton Zoltan
* Several fixes for hotplug logic
* Assorted other fixes and cleanups
* remotes/dgibson/tags/ppc-for-2.11-20170908: (40 commits)
ppc: spapr: Move VCPU ID calculation into sPAPR
ppc: remove non implemented cpu models
ppc: drop caching ObjectClass from PowerPCCPUAlias
ppc: simplify cpu model lookup by PVR
ppc: replace inter-function cyclic dependency/recurssion with 2 simple lookups
ppc: make cpu alias point only to real cpu models
ppc: make cpu_model translation to type consistent
ppc: use macros to make cpu type name from string literal
target/ppc: Remove old STATUS file
PPC: KVM: Support machine option to set VSMT mode
spapr: fallback to raw mode if best compat mode cannot be set during CAS
hw/nvram/spapr_nvram: Device can not be created by the users
hw/ppc/spapr_cpu_core: Add a proper check for spapr machine
ppc4xx: Export ECB and PLB emulation
ppc4xx_i2c: Move to hw/i2c
ppc4xx_i2c: QOMify
ppc4xx: Split off 4xx I2C emulation from ppc405_uc to its own file
ppc4xx: Make MAL emulation more generic
ppc4xx: Move MAL from ppc405_uc to ppc4xx_devs
spapr_iommu: Realloc guest visible TCE table when hot(un)plugging vfio-pci
...
Mark Cave-Ayland [Sun, 16 Jul 2017 20:27:34 +0000 (21:27 +0100)]
pci: add reserved slot check to do_pci_register_device()
Add a new slot_reserved_mask bitmask to PCIBus indicating whether or not each
PCI slot on the bus is reserved. Ensure that it is initialised to zero to
maintain the existing behaviour that all slots are available by default, and
add the additional check with appropriate error reporting to
do_pci_register_device().
This compat property sole function is to prevent the device from being
instantiated. Instead of requiring an extra compat property, check if
fw_cfg has DMA enabled.
fw_cfg is a built-in device that is initialized very early by the
machine init code. We have at least one other device that also
assumes fw_cfg_find() can be safely used on realize: pvpanic.
This has the additional benefit of handling other cases properly, like:
$ qemu-system-x86_64 -device vmgenid -machine none
qemu-system-x86_64: -device vmgenid: vmgenid requires DMA write support in fw_cfg, which this machine type does not provide
$ qemu-system-x86_64 -device vmgenid -machine pc-i440fx-2.9 -global fw_cfg.dma_enabled=off
qemu-system-x86_64: -device vmgenid: vmgenid requires DMA write support in fw_cfg, which this machine type does not provide
$ qemu-system-x86_64 -device vmgenid -machine pc-i440fx-2.6 -global fw_cfg.dma_enabled=on
[boots normally]
Peter Xu [Thu, 17 Aug 2017 05:56:14 +0000 (13:56 +0800)]
intel_iommu: fix missing BQL in pt fast path
In vtd_switch_address_space() we did the memory region switch, however
it's possible that the caller of it has not taken the BQL at all. Make
sure we have it.
hw/pci: add QEMU-specific PCI capability to the Generic PCI Express Root Port
To enable hotplugging of a newly created pcie-pci-bridge,
we need to tell firmware (e.g. SeaBIOS) to reserve
additional buses or IO/MEM/PREF space for pcie-root-port.
Additional bus reservation allows us to hotplug pcie-pci-bridge into this root port.
The number of buses and IO/MEM/PREF space to reserve are provided to the device via
a corresponding property, and to the firmware via new PCI capability.
The properties' default values are -1 to keep default behavior unchanged.
hw/pci: introduce bridge-only vendor-specific capability to provide some hints to firmware
On PCI init PCI bridges may need some extra info about bus number,
IO, memory and prefetchable memory to reserve. QEMU can provide this
with a special vendor-specific PCI capability.
projects / qemu.git / log