2013-11-22 21:17:14 +04:00
|
|
|
/*
|
|
|
|
* ARM mach-virt emulation
|
|
|
|
*
|
|
|
|
* Copyright (c) 2013 Linaro Limited
|
|
|
|
*
|
|
|
|
* This program is free software; you can redistribute it and/or modify it
|
|
|
|
* under the terms and conditions of the GNU General Public License,
|
|
|
|
* version 2 or later, as published by the Free Software Foundation.
|
|
|
|
*
|
|
|
|
* This program is distributed in the hope it will be useful, but WITHOUT
|
|
|
|
* ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
|
|
|
|
* FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for
|
|
|
|
* more details.
|
|
|
|
*
|
|
|
|
* You should have received a copy of the GNU General Public License along with
|
|
|
|
* this program. If not, see <http://www.gnu.org/licenses/>.
|
|
|
|
*
|
|
|
|
* Emulate a virtual board which works by passing Linux all the information
|
|
|
|
* it needs about what devices are present via the device tree.
|
|
|
|
* There are some restrictions about what we can do here:
|
|
|
|
* + we can only present devices whose Linux drivers will work based
|
|
|
|
* purely on the device tree with no platform data at all
|
|
|
|
* + we want to present a very stripped-down minimalist platform,
|
|
|
|
* both because this reduces the security attack surface from the guest
|
|
|
|
* and also because it reduces our exposure to being broken when
|
|
|
|
* the kernel updates its device tree bindings and requires further
|
|
|
|
* information in a device binding that we aren't providing.
|
|
|
|
* This is essentially the same approach kvmtool uses.
|
|
|
|
*/
|
|
|
|
|
2015-12-07 19:23:45 +03:00
|
|
|
#include "qemu/osdep.h"
|
2019-05-23 17:35:08 +03:00
|
|
|
#include "qemu-common.h"
|
2020-10-28 14:36:57 +03:00
|
|
|
#include "qemu/datadir.h"
|
2019-03-04 13:13:32 +03:00
|
|
|
#include "qemu/units.h"
|
hw/arm/virt: Support firmware configuration with -blockdev
The ARM virt machines put firmware in flash memory. To configure it,
you use -drive if=pflash,unit=0,... and optionally -drive
if=pflash,unit=1,...
Why two -drive? This permits setting up one part of the flash memory
read-only, and the other part read/write. It also makes upgrading
firmware on the host easier. Below the hood, we get two separate
flash devices, because we were too lazy to improve our flash device
models to support sector protection.
The problem at hand is to do the same with -blockdev somehow, as one
more step towards deprecating -drive.
We recently solved this problem for x86 PC machines, in commit
ebc29e1beab. See the commit message for design rationale.
This commit solves it for ARM virt basically the same way: new machine
properties pflash0, pflash1 forward to the onboard flash devices'
properties. Requires creating the onboard devices in the
.instance_init() method virt_instance_init(). The existing code to
pick up drives defined with -drive if=pflash is replaced by code to
desugar into the machine properties.
There are a few behavioral differences, though:
* The flash devices are always present (x86: only present if
configured)
* Flash base addresses and sizes are fixed (x86: sizes depend on
images, mapped back to back below a fixed address)
* -bios configures contents of first pflash (x86: -bios configures ROM
contents)
* -bios is rejected when first pflash is also configured with -machine
pflash0=... (x86: bios is silently ignored then)
* -machine pflash1=... does not require -machine pflash0=... (x86: it
does).
The actual code is a bit simpler than for x86 mostly due to the first
two differences.
Before the patch, all the action is in create_flash(), called from the
machine's .init() method machvirt_init():
main()
machine_run_board_init()
machvirt_init()
create_flash()
create_one_flash() for flash[0]
create
configure
includes obeying -drive if=pflash,unit=0
realize
map
fall back to -bios
create_one_flash() for flash[1]
create
configure
includes obeying -drive if=pflash,unit=1
realize
map
update FDT
To make the machine properties work, we need to move device creation
to its .instance_init() method virt_instance_init().
Another complication is machvirt_init()'s computation of
@firmware_loaded: it predicts what create_flash() will do. Instead of
predicting what create_flash()'s replacement virt_firmware_init() will
do, I decided to have virt_firmware_init() return what it did.
Requires calling it a bit earlier.
Resulting call tree:
main()
current_machine = object_new()
...
virt_instance_init()
virt_flash_create()
virt_flash_create1() for flash[0]
create
configure: set defaults
become child of machine [NEW]
add machine prop pflash0 as alias for drive [NEW]
virt_flash_create1() for flash[1]
create
configure: set defaults
become child of machine [NEW]
add machine prop pflash1 as alias for drive [NEW]
for all machine props from the command line: machine_set_property()
...
property_set_alias() for machine props pflash0, pflash1
...
set_drive() for cfi.pflash01 prop drive
this is how -machine pflash0=... etc set
machine_run_board_init(current_machine);
virt_firmware_init()
pflash_cfi01_legacy_drive()
legacy -drive if=pflash,unit=0 and =1 [NEW]
virt_flash_map()
virt_flash_map1() for flash[0]
configure: num-blocks
realize
map
virt_flash_map1() for flash[1]
configure: num-blocks
realize
map
fall back to -bios
virt_flash_fdt()
update FDT
You have László to thank for making me explain this in detail.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Acked-by: Laszlo Ersek <lersek@redhat.com>
Message-id: 20190416091348.26075-4-armbru@redhat.com
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
2019-05-07 14:55:02 +03:00
|
|
|
#include "qemu/option.h"
|
2020-02-14 16:27:44 +03:00
|
|
|
#include "monitor/qdev.h"
|
include/qemu/osdep.h: Don't include qapi/error.h
Commit 57cb38b included qapi/error.h into qemu/osdep.h to get the
Error typedef. Since then, we've moved to include qemu/osdep.h
everywhere. Its file comment explains: "To avoid getting into
possible circular include dependencies, this file should not include
any other QEMU headers, with the exceptions of config-host.h,
compiler.h, os-posix.h and os-win32.h, all of which are doing a
similar job to this file and are under similar constraints."
qapi/error.h doesn't do a similar job, and it doesn't adhere to
similar constraints: it includes qapi-types.h. That's in excess of
100KiB of crap most .c files don't actually need.
Add the typedef to qemu/typedefs.h, and include that instead of
qapi/error.h. Include qapi/error.h in .c files that need it and don't
get it now. Include qapi-types.h in qom/object.h for uint16List.
Update scripts/clean-includes accordingly. Update it further to match
reality: replace config.h by config-target.h, add sysemu/os-posix.h,
sysemu/os-win32.h. Update the list of includes in the qemu/osdep.h
comment quoted above similarly.
This reduces the number of objects depending on qapi/error.h from "all
of them" to less than a third. Unfortunately, the number depending on
qapi-types.h shrinks only a little. More work is needed for that one.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
[Fix compilation without the spice devel packages. - Paolo]
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2016-03-14 11:01:28 +03:00
|
|
|
#include "qapi/error.h"
|
2013-11-22 21:17:14 +04:00
|
|
|
#include "hw/sysbus.h"
|
2019-05-23 16:47:43 +03:00
|
|
|
#include "hw/arm/boot.h"
|
2013-11-22 21:17:14 +04:00
|
|
|
#include "hw/arm/primecell.h"
|
2015-05-29 13:28:54 +03:00
|
|
|
#include "hw/arm/virt.h"
|
2019-03-08 12:46:00 +03:00
|
|
|
#include "hw/block/flash.h"
|
2017-11-25 18:16:06 +03:00
|
|
|
#include "hw/vfio/vfio-calxeda-xgmac.h"
|
|
|
|
#include "hw/vfio/vfio-amd-xgbe.h"
|
2018-06-13 15:29:46 +03:00
|
|
|
#include "hw/display/ramfb.h"
|
2013-11-22 21:17:14 +04:00
|
|
|
#include "net/net.h"
|
|
|
|
#include "sysemu/device_tree.h"
|
2016-05-12 15:22:27 +03:00
|
|
|
#include "sysemu/numa.h"
|
2019-08-12 08:23:59 +03:00
|
|
|
#include "sysemu/runstate.h"
|
2020-03-05 19:51:45 +03:00
|
|
|
#include "sysemu/tpm.h"
|
2013-11-22 21:17:14 +04:00
|
|
|
#include "sysemu/kvm.h"
|
2014-09-12 17:06:48 +04:00
|
|
|
#include "hw/loader.h"
|
hw/smbios: support for type 41 (onboard devices extended information)
Type 41 defines the attributes of devices that are onboard. The
original intent was to imply the BIOS had some level of control over
the enablement of the associated devices.
If network devices are present in this table, by default, udev will
name the corresponding interfaces enoX, X being the instance number.
Without such information, udev will fallback to using the PCI ID and
this usually gives ens3 or ens4. This can be a bit annoying as the
name of the network card may depend on the order of options and may
change if a new PCI device is added earlier on the commande line.
Being able to provide SMBIOS type 41 entry ensure the name of the
interface won't change and helps the user guess the right name without
booting a first time.
This can be invoked with:
$QEMU -netdev user,id=internet
-device virtio-net-pci,mac=50:54:00:00:00:42,netdev=internet,id=internet-dev \
-smbios type=41,designation='Onboard LAN',instance=1,kind=ethernet,pcidev=internet-dev
The PCI segment is assumed to be 0. This should hold true for most
cases.
$ dmidecode -t 41
# dmidecode 3.3
Getting SMBIOS data from sysfs.
SMBIOS 2.8 present.
Handle 0x2900, DMI type 41, 11 bytes
Onboard Device
Reference Designation: Onboard LAN
Type: Ethernet
Status: Enabled
Type Instance: 1
Bus Address: 0000:00:09.0
$ ip -brief a
lo UNKNOWN 127.0.0.1/8 ::1/128
eno1 UP 10.0.2.14/24 fec0::5254:ff:fe00:42/64 fe80::5254:ff:fe00:42/64
Signed-off-by: Vincent Bernat <vincent@bernat.ch>
Message-Id: <20210401171138.62970-1-vincent@bernat.ch>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2021-04-01 20:11:38 +03:00
|
|
|
#include "qapi/error.h"
|
2013-11-22 21:17:14 +04:00
|
|
|
#include "qemu/bitops.h"
|
|
|
|
#include "qemu/error-report.h"
|
2019-05-23 17:35:07 +03:00
|
|
|
#include "qemu/module.h"
|
2015-02-13 08:46:08 +03:00
|
|
|
#include "hw/pci-host/gpex.h"
|
2020-02-14 16:27:44 +03:00
|
|
|
#include "hw/virtio/virtio-pci.h"
|
2015-06-02 14:29:13 +03:00
|
|
|
#include "hw/arm/sysbus-fdt.h"
|
|
|
|
#include "hw/platform-bus.h"
|
2019-08-12 08:23:51 +03:00
|
|
|
#include "hw/qdev-properties.h"
|
2015-06-19 16:17:44 +03:00
|
|
|
#include "hw/arm/fdt.h"
|
2016-07-14 18:51:37 +03:00
|
|
|
#include "hw/intc/arm_gic.h"
|
|
|
|
#include "hw/intc/arm_gicv3_common.h"
|
2019-08-12 08:23:42 +03:00
|
|
|
#include "hw/irq.h"
|
2015-08-13 13:26:21 +03:00
|
|
|
#include "kvm_arm.h"
|
2018-12-11 19:34:06 +03:00
|
|
|
#include "hw/firmware/smbios.h"
|
2015-09-24 03:29:37 +03:00
|
|
|
#include "qapi/visitor.h"
|
2020-03-20 13:01:36 +03:00
|
|
|
#include "qapi/qapi-visit-common.h"
|
2015-12-17 16:37:14 +03:00
|
|
|
#include "standard-headers/linux/input.h"
|
2018-05-04 20:05:52 +03:00
|
|
|
#include "hw/arm/smmuv3.h"
|
2019-03-04 13:13:36 +03:00
|
|
|
#include "hw/acpi/acpi.h"
|
2019-03-04 13:13:38 +03:00
|
|
|
#include "target/arm/internals.h"
|
2019-09-18 16:06:26 +03:00
|
|
|
#include "hw/mem/pc-dimm.h"
|
|
|
|
#include "hw/mem/nvdimm.h"
|
2019-09-18 16:06:27 +03:00
|
|
|
#include "hw/acpi/generic_event_device.h"
|
2020-02-14 16:27:44 +03:00
|
|
|
#include "hw/virtio/virtio-iommu.h"
|
2020-02-25 01:22:23 +03:00
|
|
|
#include "hw/char/pl011.h"
|
2020-04-20 15:18:07 +03:00
|
|
|
#include "qemu/guest-random.h"
|
2013-11-22 21:17:14 +04:00
|
|
|
|
2016-06-14 17:59:12 +03:00
|
|
|
#define DEFINE_VIRT_MACHINE_LATEST(major, minor, latest) \
|
2016-06-14 17:59:12 +03:00
|
|
|
static void virt_##major##_##minor##_class_init(ObjectClass *oc, \
|
|
|
|
void *data) \
|
|
|
|
{ \
|
|
|
|
MachineClass *mc = MACHINE_CLASS(oc); \
|
|
|
|
virt_machine_##major##_##minor##_options(mc); \
|
|
|
|
mc->desc = "QEMU " # major "." # minor " ARM Virtual Machine"; \
|
2016-06-14 17:59:12 +03:00
|
|
|
if (latest) { \
|
|
|
|
mc->alias = "virt"; \
|
|
|
|
} \
|
2016-06-14 17:59:12 +03:00
|
|
|
} \
|
|
|
|
static const TypeInfo machvirt_##major##_##minor##_info = { \
|
|
|
|
.name = MACHINE_TYPE_NAME("virt-" # major "." # minor), \
|
|
|
|
.parent = TYPE_VIRT_MACHINE, \
|
|
|
|
.class_init = virt_##major##_##minor##_class_init, \
|
|
|
|
}; \
|
|
|
|
static void machvirt_machine_##major##_##minor##_init(void) \
|
|
|
|
{ \
|
|
|
|
type_register_static(&machvirt_##major##_##minor##_info); \
|
|
|
|
} \
|
|
|
|
type_init(machvirt_machine_##major##_##minor##_init);
|
|
|
|
|
2016-06-14 17:59:12 +03:00
|
|
|
#define DEFINE_VIRT_MACHINE_AS_LATEST(major, minor) \
|
|
|
|
DEFINE_VIRT_MACHINE_LATEST(major, minor, true)
|
|
|
|
#define DEFINE_VIRT_MACHINE(major, minor) \
|
|
|
|
DEFINE_VIRT_MACHINE_LATEST(major, minor, false)
|
|
|
|
|
2016-06-14 17:59:12 +03:00
|
|
|
|
2017-01-09 14:40:22 +03:00
|
|
|
/* Number of external interrupt lines to configure the GIC with */
|
|
|
|
#define NUM_IRQS 256
|
|
|
|
|
|
|
|
#define PLATFORM_BUS_NUM_IRQS 64
|
|
|
|
|
2019-03-04 13:13:39 +03:00
|
|
|
/* Legacy RAM limit in GB (< version 4.0) */
|
2019-03-04 13:13:36 +03:00
|
|
|
#define LEGACY_RAMLIMIT_GB 255
|
|
|
|
#define LEGACY_RAMLIMIT_BYTES (LEGACY_RAMLIMIT_GB * GiB)
|
2016-03-04 14:30:16 +03:00
|
|
|
|
2013-11-22 21:17:14 +04:00
|
|
|
/* Addresses and sizes of our components.
|
|
|
|
* 0..128MB is space for a flash device so we can run bootrom code such as UEFI.
|
|
|
|
* 128MB..256MB is used for miscellaneous device I/O.
|
|
|
|
* 256MB..1GB is reserved for possible future PCI support (ie where the
|
|
|
|
* PCI memory window will go if we add a PCI host controller).
|
|
|
|
* 1GB and up is RAM (which may happily spill over into the
|
|
|
|
* high memory region beyond 4GB).
|
|
|
|
* This represents a compromise between how much RAM can be given to
|
|
|
|
* a 32 bit VM and leaving space for expansion and in particular for PCI.
|
2014-06-29 21:38:39 +04:00
|
|
|
* Note that devices should generally be placed at multiples of 0x10000,
|
|
|
|
* to accommodate guests using 64K pages.
|
2013-11-22 21:17:14 +04:00
|
|
|
*/
|
2019-03-04 13:13:32 +03:00
|
|
|
static const MemMapEntry base_memmap[] = {
|
2013-11-22 21:17:14 +04:00
|
|
|
/* Space up to 0x8000000 is reserved for a boot ROM */
|
2015-06-02 14:29:14 +03:00
|
|
|
[VIRT_FLASH] = { 0, 0x08000000 },
|
|
|
|
[VIRT_CPUPERIPHS] = { 0x08000000, 0x00020000 },
|
2013-11-22 21:17:14 +04:00
|
|
|
/* GIC distributor and CPU interfaces sit inside the CPU peripheral space */
|
2015-06-02 14:29:14 +03:00
|
|
|
[VIRT_GIC_DIST] = { 0x08000000, 0x00010000 },
|
|
|
|
[VIRT_GIC_CPU] = { 0x08010000, 0x00010000 },
|
|
|
|
[VIRT_GIC_V2M] = { 0x08020000, 0x00001000 },
|
2018-08-14 19:17:21 +03:00
|
|
|
[VIRT_GIC_HYP] = { 0x08030000, 0x00010000 },
|
|
|
|
[VIRT_GIC_VCPU] = { 0x08040000, 0x00010000 },
|
2015-09-24 03:29:37 +03:00
|
|
|
/* The space in between here is reserved for GICv3 CPU/vCPU/HYP */
|
|
|
|
[VIRT_GIC_ITS] = { 0x08080000, 0x00020000 },
|
|
|
|
/* This redistributor space allows up to 2*64kB*123 CPUs */
|
|
|
|
[VIRT_GIC_REDIST] = { 0x080A0000, 0x00F60000 },
|
2015-06-02 14:29:14 +03:00
|
|
|
[VIRT_UART] = { 0x09000000, 0x00001000 },
|
|
|
|
[VIRT_RTC] = { 0x09010000, 0x00001000 },
|
2015-10-08 18:02:56 +03:00
|
|
|
[VIRT_FW_CFG] = { 0x09020000, 0x00000018 },
|
2015-12-17 16:37:13 +03:00
|
|
|
[VIRT_GPIO] = { 0x09030000, 0x00001000 },
|
2016-01-21 17:15:07 +03:00
|
|
|
[VIRT_SECURE_UART] = { 0x09040000, 0x00001000 },
|
2018-05-04 20:05:52 +03:00
|
|
|
[VIRT_SMMU] = { 0x09050000, 0x00020000 },
|
2019-09-18 16:06:27 +03:00
|
|
|
[VIRT_PCDIMM_ACPI] = { 0x09070000, MEMORY_HOTPLUG_IO_LEN },
|
|
|
|
[VIRT_ACPI_GED] = { 0x09080000, ACPI_GED_EVT_SEL_LEN },
|
2020-04-21 15:59:30 +03:00
|
|
|
[VIRT_NVDIMM_ACPI] = { 0x09090000, NVDIMM_ACPI_IO_LEN},
|
2020-10-01 09:17:18 +03:00
|
|
|
[VIRT_PVTIME] = { 0x090a0000, 0x00010000 },
|
2021-01-28 15:00:11 +03:00
|
|
|
[VIRT_SECURE_GPIO] = { 0x090b0000, 0x00001000 },
|
2015-06-02 14:29:14 +03:00
|
|
|
[VIRT_MMIO] = { 0x0a000000, 0x00000200 },
|
2013-11-22 21:17:14 +04:00
|
|
|
/* ...repeating for a total of NUM_VIRTIO_TRANSPORTS, each of that size */
|
2015-06-02 14:29:14 +03:00
|
|
|
[VIRT_PLATFORM_BUS] = { 0x0c000000, 0x02000000 },
|
2016-03-04 14:30:17 +03:00
|
|
|
[VIRT_SECURE_MEM] = { 0x0e000000, 0x01000000 },
|
2015-06-02 14:29:14 +03:00
|
|
|
[VIRT_PCIE_MMIO] = { 0x10000000, 0x2eff0000 },
|
|
|
|
[VIRT_PCIE_PIO] = { 0x3eff0000, 0x00010000 },
|
|
|
|
[VIRT_PCIE_ECAM] = { 0x3f000000, 0x01000000 },
|
2019-03-04 13:13:36 +03:00
|
|
|
/* Actual RAM size depends on initial RAM and device memory settings */
|
|
|
|
[VIRT_MEM] = { GiB, LEGACY_RAMLIMIT_BYTES },
|
2019-03-04 13:13:32 +03:00
|
|
|
};
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Highmem IO Regions: This memory map is floating, located after the RAM.
|
|
|
|
* Each MemMapEntry base (GPA) will be dynamically computed, depending on the
|
|
|
|
* top of the RAM, so that its base get the same alignment as the size,
|
|
|
|
* ie. a 512GiB entry will be aligned on a 512GiB boundary. If there is
|
|
|
|
* less than 256GiB of RAM, the floating area starts at the 256GiB mark.
|
|
|
|
* Note the extended_memmap is sized so that it eventually also includes the
|
|
|
|
* base_memmap entries (VIRT_HIGH_GIC_REDIST2 index is greater than the last
|
|
|
|
* index of base_memmap).
|
|
|
|
*/
|
|
|
|
static MemMapEntry extended_memmap[] = {
|
2018-06-22 15:28:36 +03:00
|
|
|
/* Additional 64 MB redist region (can contain up to 512 redistributors) */
|
2019-03-04 13:13:32 +03:00
|
|
|
[VIRT_HIGH_GIC_REDIST2] = { 0x0, 64 * MiB },
|
|
|
|
[VIRT_HIGH_PCIE_ECAM] = { 0x0, 256 * MiB },
|
|
|
|
/* Second PCIe window */
|
|
|
|
[VIRT_HIGH_PCIE_MMIO] = { 0x0, 512 * GiB },
|
2013-11-22 21:17:14 +04:00
|
|
|
};
|
|
|
|
|
|
|
|
static const int a15irqmap[] = {
|
|
|
|
[VIRT_UART] = 1,
|
2014-06-29 21:38:39 +04:00
|
|
|
[VIRT_RTC] = 2,
|
2015-02-13 08:46:08 +03:00
|
|
|
[VIRT_PCIE] = 3, /* ... to 6 */
|
2015-12-17 16:37:13 +03:00
|
|
|
[VIRT_GPIO] = 7,
|
2016-01-21 17:15:07 +03:00
|
|
|
[VIRT_SECURE_UART] = 8,
|
2019-09-18 16:06:27 +03:00
|
|
|
[VIRT_ACPI_GED] = 9,
|
2013-11-22 21:17:14 +04:00
|
|
|
[VIRT_MMIO] = 16, /* ...to 16 + NUM_VIRTIO_TRANSPORTS - 1 */
|
2015-06-02 16:56:23 +03:00
|
|
|
[VIRT_GIC_V2M] = 48, /* ...to 48 + NUM_GICV2M_SPIS - 1 */
|
2018-05-04 20:05:52 +03:00
|
|
|
[VIRT_SMMU] = 74, /* ...to 74 + NUM_SMMU_IRQS - 1 */
|
2015-06-02 14:29:13 +03:00
|
|
|
[VIRT_PLATFORM_BUS] = 112, /* ...to 112 + PLATFORM_BUS_NUM_IRQS -1 */
|
2013-11-22 21:17:14 +04:00
|
|
|
};
|
|
|
|
|
2017-01-09 14:40:21 +03:00
|
|
|
static const char *valid_cpus[] = {
|
2019-07-01 19:26:14 +03:00
|
|
|
ARM_CPU_TYPE_NAME("cortex-a7"),
|
2017-09-13 19:04:57 +03:00
|
|
|
ARM_CPU_TYPE_NAME("cortex-a15"),
|
|
|
|
ARM_CPU_TYPE_NAME("cortex-a53"),
|
|
|
|
ARM_CPU_TYPE_NAME("cortex-a57"),
|
2018-11-27 14:02:45 +03:00
|
|
|
ARM_CPU_TYPE_NAME("cortex-a72"),
|
2021-08-31 11:29:39 +03:00
|
|
|
ARM_CPU_TYPE_NAME("a64fx"),
|
2017-09-13 19:04:57 +03:00
|
|
|
ARM_CPU_TYPE_NAME("host"),
|
2018-03-09 20:09:44 +03:00
|
|
|
ARM_CPU_TYPE_NAME("max"),
|
2013-11-22 21:17:14 +04:00
|
|
|
};
|
|
|
|
|
2017-09-13 19:04:57 +03:00
|
|
|
static bool cpu_type_valid(const char *cpu)
|
2013-11-22 21:17:14 +04:00
|
|
|
{
|
|
|
|
int i;
|
|
|
|
|
2017-01-09 14:40:21 +03:00
|
|
|
for (i = 0; i < ARRAY_SIZE(valid_cpus); i++) {
|
|
|
|
if (strcmp(cpu, valid_cpus[i]) == 0) {
|
|
|
|
return true;
|
2013-11-22 21:17:14 +04:00
|
|
|
}
|
|
|
|
}
|
2017-01-09 14:40:21 +03:00
|
|
|
return false;
|
2013-11-22 21:17:14 +04:00
|
|
|
}
|
|
|
|
|
2021-03-03 20:36:36 +03:00
|
|
|
static void create_kaslr_seed(MachineState *ms, const char *node)
|
2020-04-20 15:18:07 +03:00
|
|
|
{
|
|
|
|
uint64_t seed;
|
|
|
|
|
2020-06-30 12:03:28 +03:00
|
|
|
if (qemu_guest_getrandom(&seed, sizeof(seed), NULL)) {
|
2020-04-20 15:18:07 +03:00
|
|
|
return;
|
|
|
|
}
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_u64(ms->fdt, node, "kaslr-seed", seed);
|
2020-04-20 15:18:07 +03:00
|
|
|
}
|
|
|
|
|
2017-01-09 14:40:21 +03:00
|
|
|
static void create_fdt(VirtMachineState *vms)
|
2013-11-22 21:17:14 +04:00
|
|
|
{
|
2019-08-09 09:57:22 +03:00
|
|
|
MachineState *ms = MACHINE(vms);
|
|
|
|
int nb_numa_nodes = ms->numa_state->num_nodes;
|
2017-01-09 14:40:21 +03:00
|
|
|
void *fdt = create_device_tree(&vms->fdt_size);
|
2013-11-22 21:17:14 +04:00
|
|
|
|
|
|
|
if (!fdt) {
|
|
|
|
error_report("create_device_tree() failed");
|
|
|
|
exit(1);
|
|
|
|
}
|
|
|
|
|
2021-03-03 20:36:36 +03:00
|
|
|
ms->fdt = fdt;
|
2013-11-22 21:17:14 +04:00
|
|
|
|
|
|
|
/* Header */
|
2013-11-11 12:14:41 +04:00
|
|
|
qemu_fdt_setprop_string(fdt, "/", "compatible", "linux,dummy-virt");
|
|
|
|
qemu_fdt_setprop_cell(fdt, "/", "#address-cells", 0x2);
|
|
|
|
qemu_fdt_setprop_cell(fdt, "/", "#size-cells", 0x2);
|
2013-11-22 21:17:14 +04:00
|
|
|
|
2018-06-29 17:11:01 +03:00
|
|
|
/* /chosen must exist for load_dtb to fill in necessary properties later */
|
2013-11-11 12:14:41 +04:00
|
|
|
qemu_fdt_add_subnode(fdt, "/chosen");
|
2021-03-03 20:36:36 +03:00
|
|
|
create_kaslr_seed(ms, "/chosen");
|
2013-11-22 21:17:14 +04:00
|
|
|
|
2020-04-20 15:18:06 +03:00
|
|
|
if (vms->secure) {
|
|
|
|
qemu_fdt_add_subnode(fdt, "/secure-chosen");
|
2021-03-03 20:36:36 +03:00
|
|
|
create_kaslr_seed(ms, "/secure-chosen");
|
2020-04-20 15:18:06 +03:00
|
|
|
}
|
|
|
|
|
2013-11-22 21:17:14 +04:00
|
|
|
/* Clock node, for the benefit of the UART. The kernel device tree
|
|
|
|
* binding documentation claims the PL011 node clock properties are
|
|
|
|
* optional but in practice if you omit them the kernel refuses to
|
|
|
|
* probe for the device.
|
|
|
|
*/
|
2017-01-09 14:40:21 +03:00
|
|
|
vms->clock_phandle = qemu_fdt_alloc_phandle(fdt);
|
2013-11-11 12:14:41 +04:00
|
|
|
qemu_fdt_add_subnode(fdt, "/apb-pclk");
|
|
|
|
qemu_fdt_setprop_string(fdt, "/apb-pclk", "compatible", "fixed-clock");
|
|
|
|
qemu_fdt_setprop_cell(fdt, "/apb-pclk", "#clock-cells", 0x0);
|
|
|
|
qemu_fdt_setprop_cell(fdt, "/apb-pclk", "clock-frequency", 24000000);
|
|
|
|
qemu_fdt_setprop_string(fdt, "/apb-pclk", "clock-output-names",
|
2013-11-22 21:17:14 +04:00
|
|
|
"clk24mhz");
|
2017-01-09 14:40:21 +03:00
|
|
|
qemu_fdt_setprop_cell(fdt, "/apb-pclk", "phandle", vms->clock_phandle);
|
2013-11-22 21:17:14 +04:00
|
|
|
|
2019-08-09 09:57:23 +03:00
|
|
|
if (nb_numa_nodes > 0 && ms->numa_state->have_numa_distance) {
|
2017-06-02 13:51:49 +03:00
|
|
|
int size = nb_numa_nodes * nb_numa_nodes * 3 * sizeof(uint32_t);
|
|
|
|
uint32_t *matrix = g_malloc0(size);
|
|
|
|
int idx, i, j;
|
|
|
|
|
|
|
|
for (i = 0; i < nb_numa_nodes; i++) {
|
|
|
|
for (j = 0; j < nb_numa_nodes; j++) {
|
|
|
|
idx = (i * nb_numa_nodes + j) * 3;
|
|
|
|
matrix[idx + 0] = cpu_to_be32(i);
|
|
|
|
matrix[idx + 1] = cpu_to_be32(j);
|
2019-08-09 09:57:24 +03:00
|
|
|
matrix[idx + 2] =
|
|
|
|
cpu_to_be32(ms->numa_state->nodes[i].distance[j]);
|
2017-06-02 13:51:49 +03:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
qemu_fdt_add_subnode(fdt, "/distance-map");
|
|
|
|
qemu_fdt_setprop_string(fdt, "/distance-map", "compatible",
|
|
|
|
"numa-distance-map-v1");
|
|
|
|
qemu_fdt_setprop(fdt, "/distance-map", "distance-matrix",
|
|
|
|
matrix, size);
|
|
|
|
g_free(matrix);
|
|
|
|
}
|
2014-06-19 21:06:27 +04:00
|
|
|
}
|
|
|
|
|
2017-01-09 14:40:22 +03:00
|
|
|
static void fdt_add_timer_nodes(const VirtMachineState *vms)
|
2013-11-22 21:17:14 +04:00
|
|
|
{
|
2017-01-09 14:40:21 +03:00
|
|
|
/* On real hardware these interrupts are level-triggered.
|
|
|
|
* On KVM they were edge-triggered before host kernel version 4.4,
|
|
|
|
* and level-triggered afterwards.
|
|
|
|
* On emulated QEMU they are level-triggered.
|
|
|
|
*
|
|
|
|
* Getting the DTB info about them wrong is awkward for some
|
|
|
|
* guest kernels:
|
|
|
|
* pre-4.8 ignore the DT and leave the interrupt configured
|
|
|
|
* with whatever the GIC reset value (or the bootloader) left it at
|
|
|
|
* 4.8 before rc6 honour the incorrect data by programming it back
|
|
|
|
* into the GIC, causing problems
|
|
|
|
* 4.8rc6 and later ignore the DT and always write "level triggered"
|
|
|
|
* into the GIC
|
|
|
|
*
|
|
|
|
* For backwards-compatibility, virt-2.8 and earlier will continue
|
|
|
|
* to say these are edge-triggered, but later machines will report
|
|
|
|
* the correct information.
|
2013-11-22 21:17:14 +04:00
|
|
|
*/
|
2014-10-24 15:19:11 +04:00
|
|
|
ARMCPU *armcpu;
|
2017-01-09 14:40:21 +03:00
|
|
|
VirtMachineClass *vmc = VIRT_MACHINE_GET_CLASS(vms);
|
|
|
|
uint32_t irqflags = GIC_FDT_IRQ_FLAGS_LEVEL_HI;
|
2021-03-03 20:36:36 +03:00
|
|
|
MachineState *ms = MACHINE(vms);
|
2017-01-09 14:40:21 +03:00
|
|
|
|
|
|
|
if (vmc->claim_edge_triggered_timers) {
|
|
|
|
irqflags = GIC_FDT_IRQ_FLAGS_EDGE_LO_HI;
|
|
|
|
}
|
2013-11-22 21:17:14 +04:00
|
|
|
|
2020-03-11 16:16:14 +03:00
|
|
|
if (vms->gic_version == VIRT_GIC_VERSION_2) {
|
2015-09-24 03:29:37 +03:00
|
|
|
irqflags = deposit32(irqflags, GIC_FDT_IRQ_PPI_CPU_START,
|
|
|
|
GIC_FDT_IRQ_PPI_CPU_WIDTH,
|
2020-12-15 20:48:15 +03:00
|
|
|
(1 << MACHINE(vms)->smp.cpus) - 1);
|
2015-09-24 03:29:37 +03:00
|
|
|
}
|
2013-11-22 21:17:14 +04:00
|
|
|
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_add_subnode(ms->fdt, "/timer");
|
2014-10-24 15:19:11 +04:00
|
|
|
|
|
|
|
armcpu = ARM_CPU(qemu_get_cpu(0));
|
|
|
|
if (arm_feature(&armcpu->env, ARM_FEATURE_V8)) {
|
|
|
|
const char compat[] = "arm,armv8-timer\0arm,armv7-timer";
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop(ms->fdt, "/timer", "compatible",
|
2014-10-24 15:19:11 +04:00
|
|
|
compat, sizeof(compat));
|
|
|
|
} else {
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_string(ms->fdt, "/timer", "compatible",
|
2014-10-24 15:19:11 +04:00
|
|
|
"arm,armv7-timer");
|
|
|
|
}
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop(ms->fdt, "/timer", "always-on", NULL, 0);
|
|
|
|
qemu_fdt_setprop_cells(ms->fdt, "/timer", "interrupts",
|
2015-05-29 13:28:56 +03:00
|
|
|
GIC_FDT_IRQ_TYPE_PPI, ARCH_TIMER_S_EL1_IRQ, irqflags,
|
|
|
|
GIC_FDT_IRQ_TYPE_PPI, ARCH_TIMER_NS_EL1_IRQ, irqflags,
|
|
|
|
GIC_FDT_IRQ_TYPE_PPI, ARCH_TIMER_VIRT_IRQ, irqflags,
|
|
|
|
GIC_FDT_IRQ_TYPE_PPI, ARCH_TIMER_NS_EL2_IRQ, irqflags);
|
2013-11-22 21:17:14 +04:00
|
|
|
}
|
|
|
|
|
2017-01-09 14:40:21 +03:00
|
|
|
static void fdt_add_cpu_nodes(const VirtMachineState *vms)
|
2013-11-22 21:17:14 +04:00
|
|
|
{
|
|
|
|
int cpu;
|
2015-09-07 12:39:31 +03:00
|
|
|
int addr_cells = 1;
|
2017-05-10 14:29:54 +03:00
|
|
|
const MachineState *ms = MACHINE(vms);
|
2021-10-20 17:21:20 +03:00
|
|
|
const VirtMachineClass *vmc = VIRT_MACHINE_GET_CLASS(vms);
|
2020-12-15 20:48:15 +03:00
|
|
|
int smp_cpus = ms->smp.cpus;
|
2015-09-07 12:39:31 +03:00
|
|
|
|
|
|
|
/*
|
2021-10-20 17:21:20 +03:00
|
|
|
* See Linux Documentation/devicetree/bindings/arm/cpus.yaml
|
|
|
|
* On ARM v8 64-bit systems value should be set to 2,
|
|
|
|
* that corresponds to the MPIDR_EL1 register size.
|
|
|
|
* If MPIDR_EL1[63:32] value is equal to 0 on all CPUs
|
|
|
|
* in the system, #address-cells can be set to 1, since
|
|
|
|
* MPIDR_EL1[63:32] bits are not used for CPUs
|
|
|
|
* identification.
|
2015-09-07 12:39:31 +03:00
|
|
|
*
|
2021-10-20 17:21:20 +03:00
|
|
|
* Here we actually don't know whether our system is 32- or 64-bit one.
|
|
|
|
* The simplest way to go is to examine affinity IDs of all our CPUs. If
|
|
|
|
* at least one of them has Aff3 populated, we set #address-cells to 2.
|
2015-09-07 12:39:31 +03:00
|
|
|
*/
|
2020-12-15 20:48:15 +03:00
|
|
|
for (cpu = 0; cpu < smp_cpus; cpu++) {
|
2015-09-07 12:39:31 +03:00
|
|
|
ARMCPU *armcpu = ARM_CPU(qemu_get_cpu(cpu));
|
|
|
|
|
|
|
|
if (armcpu->mp_affinity & ARM_AFF3_MASK) {
|
|
|
|
addr_cells = 2;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
2013-11-22 21:17:14 +04:00
|
|
|
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_add_subnode(ms->fdt, "/cpus");
|
|
|
|
qemu_fdt_setprop_cell(ms->fdt, "/cpus", "#address-cells", addr_cells);
|
|
|
|
qemu_fdt_setprop_cell(ms->fdt, "/cpus", "#size-cells", 0x0);
|
2013-11-22 21:17:14 +04:00
|
|
|
|
2020-12-15 20:48:15 +03:00
|
|
|
for (cpu = smp_cpus - 1; cpu >= 0; cpu--) {
|
2013-11-22 21:17:14 +04:00
|
|
|
char *nodename = g_strdup_printf("/cpus/cpu@%d", cpu);
|
|
|
|
ARMCPU *armcpu = ARM_CPU(qemu_get_cpu(cpu));
|
2017-05-10 14:29:54 +03:00
|
|
|
CPUState *cs = CPU(armcpu);
|
2013-11-22 21:17:14 +04:00
|
|
|
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_add_subnode(ms->fdt, nodename);
|
|
|
|
qemu_fdt_setprop_string(ms->fdt, nodename, "device_type", "cpu");
|
|
|
|
qemu_fdt_setprop_string(ms->fdt, nodename, "compatible",
|
2013-11-22 21:17:14 +04:00
|
|
|
armcpu->dtb_compatible);
|
|
|
|
|
2020-12-15 20:48:15 +03:00
|
|
|
if (vms->psci_conduit != QEMU_PSCI_CONDUIT_DISABLED && smp_cpus > 1) {
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_string(ms->fdt, nodename,
|
2013-11-22 21:17:14 +04:00
|
|
|
"enable-method", "psci");
|
|
|
|
}
|
|
|
|
|
2015-09-07 12:39:31 +03:00
|
|
|
if (addr_cells == 2) {
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_u64(ms->fdt, nodename, "reg",
|
2015-09-07 12:39:31 +03:00
|
|
|
armcpu->mp_affinity);
|
|
|
|
} else {
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_cell(ms->fdt, nodename, "reg",
|
2015-09-07 12:39:31 +03:00
|
|
|
armcpu->mp_affinity);
|
|
|
|
}
|
|
|
|
|
2017-05-10 14:29:54 +03:00
|
|
|
if (ms->possible_cpus->cpus[cs->cpu_index].props.has_node_id) {
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_cell(ms->fdt, nodename, "numa-node-id",
|
2017-05-10 14:29:54 +03:00
|
|
|
ms->possible_cpus->cpus[cs->cpu_index].props.node_id);
|
2016-05-12 15:22:27 +03:00
|
|
|
}
|
|
|
|
|
2021-10-20 17:21:20 +03:00
|
|
|
if (!vmc->no_cpu_topology) {
|
|
|
|
qemu_fdt_setprop_cell(ms->fdt, nodename, "phandle",
|
|
|
|
qemu_fdt_alloc_phandle(ms->fdt));
|
|
|
|
}
|
|
|
|
|
2013-11-22 21:17:14 +04:00
|
|
|
g_free(nodename);
|
|
|
|
}
|
2021-10-20 17:21:20 +03:00
|
|
|
|
|
|
|
if (!vmc->no_cpu_topology) {
|
|
|
|
/*
|
|
|
|
* Add vCPU topology description through fdt node cpu-map.
|
|
|
|
*
|
|
|
|
* See Linux Documentation/devicetree/bindings/cpu/cpu-topology.txt
|
|
|
|
* In a SMP system, the hierarchy of CPUs can be defined through
|
|
|
|
* four entities that are used to describe the layout of CPUs in
|
|
|
|
* the system: socket/cluster/core/thread.
|
|
|
|
*
|
|
|
|
* A socket node represents the boundary of system physical package
|
|
|
|
* and its child nodes must be one or more cluster nodes. A system
|
|
|
|
* can contain several layers of clustering within a single physical
|
|
|
|
* package and cluster nodes can be contained in parent cluster nodes.
|
|
|
|
*
|
|
|
|
* Given that cluster is not yet supported in the vCPU topology,
|
|
|
|
* we currently generate one cluster node within each socket node
|
|
|
|
* by default.
|
|
|
|
*/
|
|
|
|
qemu_fdt_add_subnode(ms->fdt, "/cpus/cpu-map");
|
|
|
|
|
|
|
|
for (cpu = smp_cpus - 1; cpu >= 0; cpu--) {
|
|
|
|
char *cpu_path = g_strdup_printf("/cpus/cpu@%d", cpu);
|
|
|
|
char *map_path;
|
|
|
|
|
|
|
|
if (ms->smp.threads > 1) {
|
|
|
|
map_path = g_strdup_printf(
|
|
|
|
"/cpus/cpu-map/socket%d/cluster0/core%d/thread%d",
|
|
|
|
cpu / (ms->smp.cores * ms->smp.threads),
|
|
|
|
(cpu / ms->smp.threads) % ms->smp.cores,
|
|
|
|
cpu % ms->smp.threads);
|
|
|
|
} else {
|
|
|
|
map_path = g_strdup_printf(
|
|
|
|
"/cpus/cpu-map/socket%d/cluster0/core%d",
|
|
|
|
cpu / ms->smp.cores,
|
|
|
|
cpu % ms->smp.cores);
|
|
|
|
}
|
|
|
|
qemu_fdt_add_path(ms->fdt, map_path);
|
|
|
|
qemu_fdt_setprop_phandle(ms->fdt, map_path, "cpu", cpu_path);
|
|
|
|
|
|
|
|
g_free(map_path);
|
|
|
|
g_free(cpu_path);
|
|
|
|
}
|
|
|
|
}
|
2013-11-22 21:17:14 +04:00
|
|
|
}
|
|
|
|
|
2017-01-09 14:40:21 +03:00
|
|
|
static void fdt_add_its_gic_node(VirtMachineState *vms)
|
2016-10-04 15:28:09 +03:00
|
|
|
{
|
2018-06-29 17:11:01 +03:00
|
|
|
char *nodename;
|
2021-03-03 20:36:36 +03:00
|
|
|
MachineState *ms = MACHINE(vms);
|
2018-06-29 17:11:01 +03:00
|
|
|
|
2021-03-03 20:36:36 +03:00
|
|
|
vms->msi_phandle = qemu_fdt_alloc_phandle(ms->fdt);
|
2018-06-29 17:11:01 +03:00
|
|
|
nodename = g_strdup_printf("/intc/its@%" PRIx64,
|
|
|
|
vms->memmap[VIRT_GIC_ITS].base);
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_add_subnode(ms->fdt, nodename);
|
|
|
|
qemu_fdt_setprop_string(ms->fdt, nodename, "compatible",
|
2016-10-04 15:28:09 +03:00
|
|
|
"arm,gic-v3-its");
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop(ms->fdt, nodename, "msi-controller", NULL, 0);
|
|
|
|
qemu_fdt_setprop_sized_cells(ms->fdt, nodename, "reg",
|
2017-01-09 14:40:21 +03:00
|
|
|
2, vms->memmap[VIRT_GIC_ITS].base,
|
|
|
|
2, vms->memmap[VIRT_GIC_ITS].size);
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_cell(ms->fdt, nodename, "phandle", vms->msi_phandle);
|
2018-06-29 17:11:01 +03:00
|
|
|
g_free(nodename);
|
2016-10-04 15:28:09 +03:00
|
|
|
}
|
|
|
|
|
2017-01-09 14:40:21 +03:00
|
|
|
static void fdt_add_v2m_gic_node(VirtMachineState *vms)
|
2013-11-22 21:17:14 +04:00
|
|
|
{
|
2021-03-03 20:36:36 +03:00
|
|
|
MachineState *ms = MACHINE(vms);
|
2018-06-29 17:11:01 +03:00
|
|
|
char *nodename;
|
|
|
|
|
|
|
|
nodename = g_strdup_printf("/intc/v2m@%" PRIx64,
|
|
|
|
vms->memmap[VIRT_GIC_V2M].base);
|
2021-03-03 20:36:36 +03:00
|
|
|
vms->msi_phandle = qemu_fdt_alloc_phandle(ms->fdt);
|
|
|
|
qemu_fdt_add_subnode(ms->fdt, nodename);
|
|
|
|
qemu_fdt_setprop_string(ms->fdt, nodename, "compatible",
|
2015-06-02 16:56:23 +03:00
|
|
|
"arm,gic-v2m-frame");
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop(ms->fdt, nodename, "msi-controller", NULL, 0);
|
|
|
|
qemu_fdt_setprop_sized_cells(ms->fdt, nodename, "reg",
|
2017-01-09 14:40:21 +03:00
|
|
|
2, vms->memmap[VIRT_GIC_V2M].base,
|
|
|
|
2, vms->memmap[VIRT_GIC_V2M].size);
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_cell(ms->fdt, nodename, "phandle", vms->msi_phandle);
|
2018-06-29 17:11:01 +03:00
|
|
|
g_free(nodename);
|
2015-06-02 16:56:23 +03:00
|
|
|
}
|
2013-11-22 21:17:14 +04:00
|
|
|
|
2017-01-09 14:40:22 +03:00
|
|
|
static void fdt_add_gic_node(VirtMachineState *vms)
|
2015-06-02 16:56:23 +03:00
|
|
|
{
|
2021-03-03 20:36:36 +03:00
|
|
|
MachineState *ms = MACHINE(vms);
|
2018-06-29 17:11:01 +03:00
|
|
|
char *nodename;
|
|
|
|
|
2021-03-03 20:36:36 +03:00
|
|
|
vms->gic_phandle = qemu_fdt_alloc_phandle(ms->fdt);
|
|
|
|
qemu_fdt_setprop_cell(ms->fdt, "/", "interrupt-parent", vms->gic_phandle);
|
2017-01-09 14:40:21 +03:00
|
|
|
|
2018-06-29 17:11:01 +03:00
|
|
|
nodename = g_strdup_printf("/intc@%" PRIx64,
|
|
|
|
vms->memmap[VIRT_GIC_DIST].base);
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_add_subnode(ms->fdt, nodename);
|
|
|
|
qemu_fdt_setprop_cell(ms->fdt, nodename, "#interrupt-cells", 3);
|
|
|
|
qemu_fdt_setprop(ms->fdt, nodename, "interrupt-controller", NULL, 0);
|
|
|
|
qemu_fdt_setprop_cell(ms->fdt, nodename, "#address-cells", 0x2);
|
|
|
|
qemu_fdt_setprop_cell(ms->fdt, nodename, "#size-cells", 0x2);
|
|
|
|
qemu_fdt_setprop(ms->fdt, nodename, "ranges", NULL, 0);
|
2020-03-11 16:16:14 +03:00
|
|
|
if (vms->gic_version == VIRT_GIC_VERSION_3) {
|
2018-06-22 15:28:36 +03:00
|
|
|
int nb_redist_regions = virt_gicv3_redist_region_count(vms);
|
|
|
|
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_string(ms->fdt, nodename, "compatible",
|
2015-09-24 03:29:37 +03:00
|
|
|
"arm,gic-v3");
|
2018-06-22 15:28:36 +03:00
|
|
|
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_cell(ms->fdt, nodename,
|
2018-06-22 15:28:36 +03:00
|
|
|
"#redistributor-regions", nb_redist_regions);
|
|
|
|
|
|
|
|
if (nb_redist_regions == 1) {
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_sized_cells(ms->fdt, nodename, "reg",
|
2018-06-22 15:28:36 +03:00
|
|
|
2, vms->memmap[VIRT_GIC_DIST].base,
|
|
|
|
2, vms->memmap[VIRT_GIC_DIST].size,
|
|
|
|
2, vms->memmap[VIRT_GIC_REDIST].base,
|
|
|
|
2, vms->memmap[VIRT_GIC_REDIST].size);
|
|
|
|
} else {
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_sized_cells(ms->fdt, nodename, "reg",
|
2019-03-04 13:13:31 +03:00
|
|
|
2, vms->memmap[VIRT_GIC_DIST].base,
|
|
|
|
2, vms->memmap[VIRT_GIC_DIST].size,
|
|
|
|
2, vms->memmap[VIRT_GIC_REDIST].base,
|
|
|
|
2, vms->memmap[VIRT_GIC_REDIST].size,
|
|
|
|
2, vms->memmap[VIRT_HIGH_GIC_REDIST2].base,
|
|
|
|
2, vms->memmap[VIRT_HIGH_GIC_REDIST2].size);
|
2018-06-22 15:28:36 +03:00
|
|
|
}
|
|
|
|
|
2017-01-20 14:15:11 +03:00
|
|
|
if (vms->virt) {
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_cells(ms->fdt, nodename, "interrupts",
|
2018-08-14 19:17:21 +03:00
|
|
|
GIC_FDT_IRQ_TYPE_PPI, ARCH_GIC_MAINT_IRQ,
|
2017-01-20 14:15:11 +03:00
|
|
|
GIC_FDT_IRQ_FLAGS_LEVEL_HI);
|
|
|
|
}
|
2015-09-24 03:29:37 +03:00
|
|
|
} else {
|
|
|
|
/* 'cortex-a15-gic' means 'GIC v2' */
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_string(ms->fdt, nodename, "compatible",
|
2015-09-24 03:29:37 +03:00
|
|
|
"arm,cortex-a15-gic");
|
2018-08-14 19:17:21 +03:00
|
|
|
if (!vms->virt) {
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_sized_cells(ms->fdt, nodename, "reg",
|
2018-08-14 19:17:21 +03:00
|
|
|
2, vms->memmap[VIRT_GIC_DIST].base,
|
|
|
|
2, vms->memmap[VIRT_GIC_DIST].size,
|
|
|
|
2, vms->memmap[VIRT_GIC_CPU].base,
|
|
|
|
2, vms->memmap[VIRT_GIC_CPU].size);
|
|
|
|
} else {
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_sized_cells(ms->fdt, nodename, "reg",
|
2018-08-14 19:17:21 +03:00
|
|
|
2, vms->memmap[VIRT_GIC_DIST].base,
|
|
|
|
2, vms->memmap[VIRT_GIC_DIST].size,
|
|
|
|
2, vms->memmap[VIRT_GIC_CPU].base,
|
|
|
|
2, vms->memmap[VIRT_GIC_CPU].size,
|
|
|
|
2, vms->memmap[VIRT_GIC_HYP].base,
|
|
|
|
2, vms->memmap[VIRT_GIC_HYP].size,
|
|
|
|
2, vms->memmap[VIRT_GIC_VCPU].base,
|
|
|
|
2, vms->memmap[VIRT_GIC_VCPU].size);
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_cells(ms->fdt, nodename, "interrupts",
|
2018-08-14 19:17:21 +03:00
|
|
|
GIC_FDT_IRQ_TYPE_PPI, ARCH_GIC_MAINT_IRQ,
|
|
|
|
GIC_FDT_IRQ_FLAGS_LEVEL_HI);
|
|
|
|
}
|
2015-09-24 03:29:37 +03:00
|
|
|
}
|
|
|
|
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_cell(ms->fdt, nodename, "phandle", vms->gic_phandle);
|
2018-06-29 17:11:01 +03:00
|
|
|
g_free(nodename);
|
2013-11-22 21:17:14 +04:00
|
|
|
}
|
|
|
|
|
2017-01-09 14:40:22 +03:00
|
|
|
static void fdt_add_pmu_nodes(const VirtMachineState *vms)
|
2016-06-14 17:59:12 +03:00
|
|
|
{
|
2020-10-01 09:17:16 +03:00
|
|
|
ARMCPU *armcpu = ARM_CPU(first_cpu);
|
2016-06-14 17:59:12 +03:00
|
|
|
uint32_t irqflags = GIC_FDT_IRQ_FLAGS_LEVEL_HI;
|
2021-03-03 20:36:36 +03:00
|
|
|
MachineState *ms = MACHINE(vms);
|
2016-06-14 17:59:12 +03:00
|
|
|
|
2020-10-01 09:17:16 +03:00
|
|
|
if (!arm_feature(&armcpu->env, ARM_FEATURE_PMU)) {
|
|
|
|
assert(!object_property_get_bool(OBJECT(armcpu), "pmu", NULL));
|
|
|
|
return;
|
2016-06-14 17:59:12 +03:00
|
|
|
}
|
|
|
|
|
2020-03-11 16:16:14 +03:00
|
|
|
if (vms->gic_version == VIRT_GIC_VERSION_2) {
|
2016-06-14 17:59:12 +03:00
|
|
|
irqflags = deposit32(irqflags, GIC_FDT_IRQ_PPI_CPU_START,
|
|
|
|
GIC_FDT_IRQ_PPI_CPU_WIDTH,
|
2020-12-15 20:48:15 +03:00
|
|
|
(1 << MACHINE(vms)->smp.cpus) - 1);
|
2016-06-14 17:59:12 +03:00
|
|
|
}
|
|
|
|
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_add_subnode(ms->fdt, "/pmu");
|
2016-06-14 17:59:12 +03:00
|
|
|
if (arm_feature(&armcpu->env, ARM_FEATURE_V8)) {
|
|
|
|
const char compat[] = "arm,armv8-pmuv3";
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop(ms->fdt, "/pmu", "compatible",
|
2016-06-14 17:59:12 +03:00
|
|
|
compat, sizeof(compat));
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_cells(ms->fdt, "/pmu", "interrupts",
|
2016-06-14 17:59:12 +03:00
|
|
|
GIC_FDT_IRQ_TYPE_PPI, VIRTUAL_PMU_IRQ, irqflags);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2019-12-09 12:03:06 +03:00
|
|
|
static inline DeviceState *create_acpi_ged(VirtMachineState *vms)
|
2019-09-18 16:06:27 +03:00
|
|
|
{
|
|
|
|
DeviceState *dev;
|
|
|
|
MachineState *ms = MACHINE(vms);
|
|
|
|
int irq = vms->irqmap[VIRT_ACPI_GED];
|
2019-09-18 16:06:30 +03:00
|
|
|
uint32_t event = ACPI_GED_PWR_DOWN_EVT;
|
2019-09-18 16:06:27 +03:00
|
|
|
|
|
|
|
if (ms->ram_slots) {
|
2019-09-18 16:06:30 +03:00
|
|
|
event |= ACPI_GED_MEM_HOTPLUG_EVT;
|
2019-09-18 16:06:27 +03:00
|
|
|
}
|
|
|
|
|
2020-04-21 15:59:31 +03:00
|
|
|
if (ms->nvdimms_state->is_enabled) {
|
|
|
|
event |= ACPI_GED_NVDIMM_HOTPLUG_EVT;
|
|
|
|
}
|
|
|
|
|
qdev: Convert uses of qdev_create() with Coccinelle
This is the transformation explained in the commit before previous.
Takes care of just one pattern that needs conversion. More to come in
this series.
Coccinelle script:
@ depends on !(file in "hw/arm/highbank.c")@
expression bus, type_name, dev, expr;
@@
- dev = qdev_create(bus, type_name);
+ dev = qdev_new(type_name);
... when != dev = expr
- qdev_init_nofail(dev);
+ qdev_realize_and_unref(dev, bus, &error_fatal);
@@
expression bus, type_name, dev, expr;
identifier DOWN;
@@
- dev = DOWN(qdev_create(bus, type_name));
+ dev = DOWN(qdev_new(type_name));
... when != dev = expr
- qdev_init_nofail(DEVICE(dev));
+ qdev_realize_and_unref(DEVICE(dev), bus, &error_fatal);
@@
expression bus, type_name, expr;
identifier dev;
@@
- DeviceState *dev = qdev_create(bus, type_name);
+ DeviceState *dev = qdev_new(type_name);
... when != dev = expr
- qdev_init_nofail(dev);
+ qdev_realize_and_unref(dev, bus, &error_fatal);
@@
expression bus, type_name, dev, expr, errp;
symbol true;
@@
- dev = qdev_create(bus, type_name);
+ dev = qdev_new(type_name);
... when != dev = expr
- object_property_set_bool(OBJECT(dev), true, "realized", errp);
+ qdev_realize_and_unref(dev, bus, errp);
@@
expression bus, type_name, expr, errp;
identifier dev;
symbol true;
@@
- DeviceState *dev = qdev_create(bus, type_name);
+ DeviceState *dev = qdev_new(type_name);
... when != dev = expr
- object_property_set_bool(OBJECT(dev), true, "realized", errp);
+ qdev_realize_and_unref(dev, bus, errp);
The first rule exempts hw/arm/highbank.c, because it matches along two
control flow paths there, with different @type_name. Covered by the
next commit's manual conversions.
Missing #include "qapi/error.h" added manually.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20200610053247.1583243-10-armbru@redhat.com>
[Conflicts in hw/misc/empty_slot.c and hw/sparc/leon3.c resolved]
2020-06-10 08:31:58 +03:00
|
|
|
dev = qdev_new(TYPE_ACPI_GED);
|
2019-09-18 16:06:27 +03:00
|
|
|
qdev_prop_set_uint32(dev, "ged-event", event);
|
|
|
|
|
|
|
|
sysbus_mmio_map(SYS_BUS_DEVICE(dev), 0, vms->memmap[VIRT_ACPI_GED].base);
|
|
|
|
sysbus_mmio_map(SYS_BUS_DEVICE(dev), 1, vms->memmap[VIRT_PCDIMM_ACPI].base);
|
2019-12-09 12:03:06 +03:00
|
|
|
sysbus_connect_irq(SYS_BUS_DEVICE(dev), 0, qdev_get_gpio_in(vms->gic, irq));
|
2019-09-18 16:06:27 +03:00
|
|
|
|
sysbus: Convert to sysbus_realize() etc. with Coccinelle
Convert from qdev_realize(), qdev_realize_and_unref() with null @bus
argument to sysbus_realize(), sysbus_realize_and_unref().
Coccinelle script:
@@
expression dev, errp;
@@
- qdev_realize(DEVICE(dev), NULL, errp);
+ sysbus_realize(SYS_BUS_DEVICE(dev), errp);
@@
expression sysbus_dev, dev, errp;
@@
+ sysbus_dev = SYS_BUS_DEVICE(dev);
- qdev_realize_and_unref(dev, NULL, errp);
+ sysbus_realize_and_unref(sysbus_dev, errp);
- sysbus_dev = SYS_BUS_DEVICE(dev);
@@
expression sysbus_dev, dev, errp;
expression expr;
@@
sysbus_dev = SYS_BUS_DEVICE(dev);
... when != dev = expr;
- qdev_realize_and_unref(dev, NULL, errp);
+ sysbus_realize_and_unref(sysbus_dev, errp);
@@
expression dev, errp;
@@
- qdev_realize_and_unref(DEVICE(dev), NULL, errp);
+ sysbus_realize_and_unref(SYS_BUS_DEVICE(dev), errp);
@@
expression dev, errp;
@@
- qdev_realize_and_unref(dev, NULL, errp);
+ sysbus_realize_and_unref(SYS_BUS_DEVICE(dev), errp);
Whitespace changes minimized manually.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Acked-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20200610053247.1583243-46-armbru@redhat.com>
[Conflicts in hw/misc/empty_slot.c and hw/sparc/leon3.c resolved]
2020-06-10 08:32:34 +03:00
|
|
|
sysbus_realize_and_unref(SYS_BUS_DEVICE(dev), &error_fatal);
|
2019-09-18 16:06:27 +03:00
|
|
|
|
|
|
|
return dev;
|
|
|
|
}
|
|
|
|
|
2019-12-09 12:03:06 +03:00
|
|
|
static void create_its(VirtMachineState *vms)
|
2016-10-04 15:28:09 +03:00
|
|
|
{
|
|
|
|
const char *itsclass = its_class_name();
|
|
|
|
DeviceState *dev;
|
|
|
|
|
2021-09-13 18:07:24 +03:00
|
|
|
if (!strcmp(itsclass, "arm-gicv3-its")) {
|
|
|
|
if (!vms->tcg_its) {
|
|
|
|
itsclass = NULL;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2016-10-04 15:28:09 +03:00
|
|
|
if (!itsclass) {
|
|
|
|
/* Do nothing if not supported */
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
qdev: Convert uses of qdev_create() with Coccinelle
This is the transformation explained in the commit before previous.
Takes care of just one pattern that needs conversion. More to come in
this series.
Coccinelle script:
@ depends on !(file in "hw/arm/highbank.c")@
expression bus, type_name, dev, expr;
@@
- dev = qdev_create(bus, type_name);
+ dev = qdev_new(type_name);
... when != dev = expr
- qdev_init_nofail(dev);
+ qdev_realize_and_unref(dev, bus, &error_fatal);
@@
expression bus, type_name, dev, expr;
identifier DOWN;
@@
- dev = DOWN(qdev_create(bus, type_name));
+ dev = DOWN(qdev_new(type_name));
... when != dev = expr
- qdev_init_nofail(DEVICE(dev));
+ qdev_realize_and_unref(DEVICE(dev), bus, &error_fatal);
@@
expression bus, type_name, expr;
identifier dev;
@@
- DeviceState *dev = qdev_create(bus, type_name);
+ DeviceState *dev = qdev_new(type_name);
... when != dev = expr
- qdev_init_nofail(dev);
+ qdev_realize_and_unref(dev, bus, &error_fatal);
@@
expression bus, type_name, dev, expr, errp;
symbol true;
@@
- dev = qdev_create(bus, type_name);
+ dev = qdev_new(type_name);
... when != dev = expr
- object_property_set_bool(OBJECT(dev), true, "realized", errp);
+ qdev_realize_and_unref(dev, bus, errp);
@@
expression bus, type_name, expr, errp;
identifier dev;
symbol true;
@@
- DeviceState *dev = qdev_create(bus, type_name);
+ DeviceState *dev = qdev_new(type_name);
... when != dev = expr
- object_property_set_bool(OBJECT(dev), true, "realized", errp);
+ qdev_realize_and_unref(dev, bus, errp);
The first rule exempts hw/arm/highbank.c, because it matches along two
control flow paths there, with different @type_name. Covered by the
next commit's manual conversions.
Missing #include "qapi/error.h" added manually.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20200610053247.1583243-10-armbru@redhat.com>
[Conflicts in hw/misc/empty_slot.c and hw/sparc/leon3.c resolved]
2020-06-10 08:31:58 +03:00
|
|
|
dev = qdev_new(itsclass);
|
2016-10-04 15:28:09 +03:00
|
|
|
|
qom: Put name parameter before value / visitor parameter
The object_property_set_FOO() setters take property name and value in
an unusual order:
void object_property_set_FOO(Object *obj, FOO_TYPE value,
const char *name, Error **errp)
Having to pass value before name feels grating. Swap them.
Same for object_property_set(), object_property_get(), and
object_property_parse().
Convert callers with this Coccinelle script:
@@
identifier fun = {
object_property_get, object_property_parse, object_property_set_str,
object_property_set_link, object_property_set_bool,
object_property_set_int, object_property_set_uint, object_property_set,
object_property_set_qobject
};
expression obj, v, name, errp;
@@
- fun(obj, v, name, errp)
+ fun(obj, name, v, errp)
Chokes on hw/arm/musicpal.c's lcd_refresh() with the unhelpful error
message "no position information". Convert that one manually.
Fails to convert hw/arm/armsse.c, because Coccinelle gets confused by
ARMSSE being used both as typedef and function-like macro there.
Convert manually.
Fails to convert hw/rx/rx-gdbsim.c, because Coccinelle gets confused
by RXCPU being used both as typedef and function-like macro there.
Convert manually. The other files using RXCPU that way don't need
conversion.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20200707160613.848843-27-armbru@redhat.com>
[Straightforwad conflict with commit 2336172d9b "audio: set default
value for pcspk.iobase property" resolved]
2020-07-07 19:05:54 +03:00
|
|
|
object_property_set_link(OBJECT(dev), "parent-gicv3", OBJECT(vms->gic),
|
2016-10-04 15:28:09 +03:00
|
|
|
&error_abort);
|
sysbus: Convert to sysbus_realize() etc. with Coccinelle
Convert from qdev_realize(), qdev_realize_and_unref() with null @bus
argument to sysbus_realize(), sysbus_realize_and_unref().
Coccinelle script:
@@
expression dev, errp;
@@
- qdev_realize(DEVICE(dev), NULL, errp);
+ sysbus_realize(SYS_BUS_DEVICE(dev), errp);
@@
expression sysbus_dev, dev, errp;
@@
+ sysbus_dev = SYS_BUS_DEVICE(dev);
- qdev_realize_and_unref(dev, NULL, errp);
+ sysbus_realize_and_unref(sysbus_dev, errp);
- sysbus_dev = SYS_BUS_DEVICE(dev);
@@
expression sysbus_dev, dev, errp;
expression expr;
@@
sysbus_dev = SYS_BUS_DEVICE(dev);
... when != dev = expr;
- qdev_realize_and_unref(dev, NULL, errp);
+ sysbus_realize_and_unref(sysbus_dev, errp);
@@
expression dev, errp;
@@
- qdev_realize_and_unref(DEVICE(dev), NULL, errp);
+ sysbus_realize_and_unref(SYS_BUS_DEVICE(dev), errp);
@@
expression dev, errp;
@@
- qdev_realize_and_unref(dev, NULL, errp);
+ sysbus_realize_and_unref(SYS_BUS_DEVICE(dev), errp);
Whitespace changes minimized manually.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Acked-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20200610053247.1583243-46-armbru@redhat.com>
[Conflicts in hw/misc/empty_slot.c and hw/sparc/leon3.c resolved]
2020-06-10 08:32:34 +03:00
|
|
|
sysbus_realize_and_unref(SYS_BUS_DEVICE(dev), &error_fatal);
|
2017-01-09 14:40:21 +03:00
|
|
|
sysbus_mmio_map(SYS_BUS_DEVICE(dev), 0, vms->memmap[VIRT_GIC_ITS].base);
|
2016-10-04 15:28:09 +03:00
|
|
|
|
2017-01-09 14:40:21 +03:00
|
|
|
fdt_add_its_gic_node(vms);
|
2020-07-03 18:59:42 +03:00
|
|
|
vms->msi_controller = VIRT_MSI_CTRL_ITS;
|
2016-10-04 15:28:09 +03:00
|
|
|
}
|
|
|
|
|
2019-12-09 12:03:06 +03:00
|
|
|
static void create_v2m(VirtMachineState *vms)
|
2015-06-02 16:56:23 +03:00
|
|
|
{
|
|
|
|
int i;
|
2017-01-09 14:40:21 +03:00
|
|
|
int irq = vms->irqmap[VIRT_GIC_V2M];
|
2015-06-02 16:56:23 +03:00
|
|
|
DeviceState *dev;
|
|
|
|
|
qdev: Convert uses of qdev_create() with Coccinelle
This is the transformation explained in the commit before previous.
Takes care of just one pattern that needs conversion. More to come in
this series.
Coccinelle script:
@ depends on !(file in "hw/arm/highbank.c")@
expression bus, type_name, dev, expr;
@@
- dev = qdev_create(bus, type_name);
+ dev = qdev_new(type_name);
... when != dev = expr
- qdev_init_nofail(dev);
+ qdev_realize_and_unref(dev, bus, &error_fatal);
@@
expression bus, type_name, dev, expr;
identifier DOWN;
@@
- dev = DOWN(qdev_create(bus, type_name));
+ dev = DOWN(qdev_new(type_name));
... when != dev = expr
- qdev_init_nofail(DEVICE(dev));
+ qdev_realize_and_unref(DEVICE(dev), bus, &error_fatal);
@@
expression bus, type_name, expr;
identifier dev;
@@
- DeviceState *dev = qdev_create(bus, type_name);
+ DeviceState *dev = qdev_new(type_name);
... when != dev = expr
- qdev_init_nofail(dev);
+ qdev_realize_and_unref(dev, bus, &error_fatal);
@@
expression bus, type_name, dev, expr, errp;
symbol true;
@@
- dev = qdev_create(bus, type_name);
+ dev = qdev_new(type_name);
... when != dev = expr
- object_property_set_bool(OBJECT(dev), true, "realized", errp);
+ qdev_realize_and_unref(dev, bus, errp);
@@
expression bus, type_name, expr, errp;
identifier dev;
symbol true;
@@
- DeviceState *dev = qdev_create(bus, type_name);
+ DeviceState *dev = qdev_new(type_name);
... when != dev = expr
- object_property_set_bool(OBJECT(dev), true, "realized", errp);
+ qdev_realize_and_unref(dev, bus, errp);
The first rule exempts hw/arm/highbank.c, because it matches along two
control flow paths there, with different @type_name. Covered by the
next commit's manual conversions.
Missing #include "qapi/error.h" added manually.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20200610053247.1583243-10-armbru@redhat.com>
[Conflicts in hw/misc/empty_slot.c and hw/sparc/leon3.c resolved]
2020-06-10 08:31:58 +03:00
|
|
|
dev = qdev_new("arm-gicv2m");
|
2017-01-09 14:40:21 +03:00
|
|
|
sysbus_mmio_map(SYS_BUS_DEVICE(dev), 0, vms->memmap[VIRT_GIC_V2M].base);
|
2015-06-02 16:56:23 +03:00
|
|
|
qdev_prop_set_uint32(dev, "base-spi", irq);
|
|
|
|
qdev_prop_set_uint32(dev, "num-spi", NUM_GICV2M_SPIS);
|
sysbus: Convert to sysbus_realize() etc. with Coccinelle
Convert from qdev_realize(), qdev_realize_and_unref() with null @bus
argument to sysbus_realize(), sysbus_realize_and_unref().
Coccinelle script:
@@
expression dev, errp;
@@
- qdev_realize(DEVICE(dev), NULL, errp);
+ sysbus_realize(SYS_BUS_DEVICE(dev), errp);
@@
expression sysbus_dev, dev, errp;
@@
+ sysbus_dev = SYS_BUS_DEVICE(dev);
- qdev_realize_and_unref(dev, NULL, errp);
+ sysbus_realize_and_unref(sysbus_dev, errp);
- sysbus_dev = SYS_BUS_DEVICE(dev);
@@
expression sysbus_dev, dev, errp;
expression expr;
@@
sysbus_dev = SYS_BUS_DEVICE(dev);
... when != dev = expr;
- qdev_realize_and_unref(dev, NULL, errp);
+ sysbus_realize_and_unref(sysbus_dev, errp);
@@
expression dev, errp;
@@
- qdev_realize_and_unref(DEVICE(dev), NULL, errp);
+ sysbus_realize_and_unref(SYS_BUS_DEVICE(dev), errp);
@@
expression dev, errp;
@@
- qdev_realize_and_unref(dev, NULL, errp);
+ sysbus_realize_and_unref(SYS_BUS_DEVICE(dev), errp);
Whitespace changes minimized manually.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Acked-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20200610053247.1583243-46-armbru@redhat.com>
[Conflicts in hw/misc/empty_slot.c and hw/sparc/leon3.c resolved]
2020-06-10 08:32:34 +03:00
|
|
|
sysbus_realize_and_unref(SYS_BUS_DEVICE(dev), &error_fatal);
|
2015-06-02 16:56:23 +03:00
|
|
|
|
|
|
|
for (i = 0; i < NUM_GICV2M_SPIS; i++) {
|
2019-12-09 12:03:06 +03:00
|
|
|
sysbus_connect_irq(SYS_BUS_DEVICE(dev), i,
|
|
|
|
qdev_get_gpio_in(vms->gic, irq + i));
|
2015-06-02 16:56:23 +03:00
|
|
|
}
|
|
|
|
|
2017-01-09 14:40:21 +03:00
|
|
|
fdt_add_v2m_gic_node(vms);
|
2020-07-03 18:59:42 +03:00
|
|
|
vms->msi_controller = VIRT_MSI_CTRL_GICV2M;
|
2015-06-02 16:56:23 +03:00
|
|
|
}
|
|
|
|
|
2021-09-13 18:07:24 +03:00
|
|
|
static void create_gic(VirtMachineState *vms, MemoryRegion *mem)
|
2014-05-01 18:24:46 +04:00
|
|
|
{
|
2019-05-18 23:54:26 +03:00
|
|
|
MachineState *ms = MACHINE(vms);
|
2015-09-24 03:29:37 +03:00
|
|
|
/* We create a standalone GIC */
|
2014-05-01 18:24:46 +04:00
|
|
|
SysBusDevice *gicbusdev;
|
2015-08-13 13:26:21 +03:00
|
|
|
const char *gictype;
|
2017-01-09 14:40:22 +03:00
|
|
|
int type = vms->gic_version, i;
|
2019-05-18 23:54:26 +03:00
|
|
|
unsigned int smp_cpus = ms->smp.cpus;
|
2018-06-22 15:28:37 +03:00
|
|
|
uint32_t nb_redist_regions = 0;
|
2014-05-01 18:24:46 +04:00
|
|
|
|
2015-09-24 03:29:37 +03:00
|
|
|
gictype = (type == 3) ? gicv3_class_name() : gic_class_name();
|
2014-05-01 18:24:46 +04:00
|
|
|
|
qdev: Convert uses of qdev_create() with Coccinelle
This is the transformation explained in the commit before previous.
Takes care of just one pattern that needs conversion. More to come in
this series.
Coccinelle script:
@ depends on !(file in "hw/arm/highbank.c")@
expression bus, type_name, dev, expr;
@@
- dev = qdev_create(bus, type_name);
+ dev = qdev_new(type_name);
... when != dev = expr
- qdev_init_nofail(dev);
+ qdev_realize_and_unref(dev, bus, &error_fatal);
@@
expression bus, type_name, dev, expr;
identifier DOWN;
@@
- dev = DOWN(qdev_create(bus, type_name));
+ dev = DOWN(qdev_new(type_name));
... when != dev = expr
- qdev_init_nofail(DEVICE(dev));
+ qdev_realize_and_unref(DEVICE(dev), bus, &error_fatal);
@@
expression bus, type_name, expr;
identifier dev;
@@
- DeviceState *dev = qdev_create(bus, type_name);
+ DeviceState *dev = qdev_new(type_name);
... when != dev = expr
- qdev_init_nofail(dev);
+ qdev_realize_and_unref(dev, bus, &error_fatal);
@@
expression bus, type_name, dev, expr, errp;
symbol true;
@@
- dev = qdev_create(bus, type_name);
+ dev = qdev_new(type_name);
... when != dev = expr
- object_property_set_bool(OBJECT(dev), true, "realized", errp);
+ qdev_realize_and_unref(dev, bus, errp);
@@
expression bus, type_name, expr, errp;
identifier dev;
symbol true;
@@
- DeviceState *dev = qdev_create(bus, type_name);
+ DeviceState *dev = qdev_new(type_name);
... when != dev = expr
- object_property_set_bool(OBJECT(dev), true, "realized", errp);
+ qdev_realize_and_unref(dev, bus, errp);
The first rule exempts hw/arm/highbank.c, because it matches along two
control flow paths there, with different @type_name. Covered by the
next commit's manual conversions.
Missing #include "qapi/error.h" added manually.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20200610053247.1583243-10-armbru@redhat.com>
[Conflicts in hw/misc/empty_slot.c and hw/sparc/leon3.c resolved]
2020-06-10 08:31:58 +03:00
|
|
|
vms->gic = qdev_new(gictype);
|
2019-12-09 12:03:06 +03:00
|
|
|
qdev_prop_set_uint32(vms->gic, "revision", type);
|
|
|
|
qdev_prop_set_uint32(vms->gic, "num-cpu", smp_cpus);
|
2014-05-01 18:24:46 +04:00
|
|
|
/* Note that the num-irq property counts both internal and external
|
|
|
|
* interrupts; there are always 32 of the former (mandated by GIC spec).
|
|
|
|
*/
|
2019-12-09 12:03:06 +03:00
|
|
|
qdev_prop_set_uint32(vms->gic, "num-irq", NUM_IRQS + 32);
|
2015-09-08 19:38:44 +03:00
|
|
|
if (!kvm_irqchip_in_kernel()) {
|
2019-12-09 12:03:06 +03:00
|
|
|
qdev_prop_set_bit(vms->gic, "has-security-extensions", vms->secure);
|
2015-09-08 19:38:44 +03:00
|
|
|
}
|
2018-06-22 15:28:36 +03:00
|
|
|
|
|
|
|
if (type == 3) {
|
|
|
|
uint32_t redist0_capacity =
|
|
|
|
vms->memmap[VIRT_GIC_REDIST].size / GICV3_REDIST_SIZE;
|
|
|
|
uint32_t redist0_count = MIN(smp_cpus, redist0_capacity);
|
|
|
|
|
2018-06-22 15:28:37 +03:00
|
|
|
nb_redist_regions = virt_gicv3_redist_region_count(vms);
|
|
|
|
|
2019-12-09 12:03:06 +03:00
|
|
|
qdev_prop_set_uint32(vms->gic, "len-redist-region-count",
|
2018-06-22 15:28:37 +03:00
|
|
|
nb_redist_regions);
|
2019-12-09 12:03:06 +03:00
|
|
|
qdev_prop_set_uint32(vms->gic, "redist-region-count[0]", redist0_count);
|
2018-06-22 15:28:37 +03:00
|
|
|
|
2021-09-13 18:07:24 +03:00
|
|
|
if (!kvm_irqchip_in_kernel()) {
|
|
|
|
if (vms->tcg_its) {
|
|
|
|
object_property_set_link(OBJECT(vms->gic), "sysmem",
|
|
|
|
OBJECT(mem), &error_fatal);
|
|
|
|
qdev_prop_set_bit(vms->gic, "has-lpi", true);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2018-06-22 15:28:37 +03:00
|
|
|
if (nb_redist_regions == 2) {
|
|
|
|
uint32_t redist1_capacity =
|
2019-03-04 13:13:31 +03:00
|
|
|
vms->memmap[VIRT_HIGH_GIC_REDIST2].size / GICV3_REDIST_SIZE;
|
2018-06-22 15:28:37 +03:00
|
|
|
|
2019-12-09 12:03:06 +03:00
|
|
|
qdev_prop_set_uint32(vms->gic, "redist-region-count[1]",
|
2018-06-22 15:28:37 +03:00
|
|
|
MIN(smp_cpus - redist0_count, redist1_capacity));
|
|
|
|
}
|
2018-08-14 19:17:21 +03:00
|
|
|
} else {
|
|
|
|
if (!kvm_irqchip_in_kernel()) {
|
2019-12-09 12:03:06 +03:00
|
|
|
qdev_prop_set_bit(vms->gic, "has-virtualization-extensions",
|
2018-08-14 19:17:21 +03:00
|
|
|
vms->virt);
|
|
|
|
}
|
2018-06-22 15:28:36 +03:00
|
|
|
}
|
2019-12-09 12:03:06 +03:00
|
|
|
gicbusdev = SYS_BUS_DEVICE(vms->gic);
|
sysbus: Convert to sysbus_realize() etc. with Coccinelle
Convert from qdev_realize(), qdev_realize_and_unref() with null @bus
argument to sysbus_realize(), sysbus_realize_and_unref().
Coccinelle script:
@@
expression dev, errp;
@@
- qdev_realize(DEVICE(dev), NULL, errp);
+ sysbus_realize(SYS_BUS_DEVICE(dev), errp);
@@
expression sysbus_dev, dev, errp;
@@
+ sysbus_dev = SYS_BUS_DEVICE(dev);
- qdev_realize_and_unref(dev, NULL, errp);
+ sysbus_realize_and_unref(sysbus_dev, errp);
- sysbus_dev = SYS_BUS_DEVICE(dev);
@@
expression sysbus_dev, dev, errp;
expression expr;
@@
sysbus_dev = SYS_BUS_DEVICE(dev);
... when != dev = expr;
- qdev_realize_and_unref(dev, NULL, errp);
+ sysbus_realize_and_unref(sysbus_dev, errp);
@@
expression dev, errp;
@@
- qdev_realize_and_unref(DEVICE(dev), NULL, errp);
+ sysbus_realize_and_unref(SYS_BUS_DEVICE(dev), errp);
@@
expression dev, errp;
@@
- qdev_realize_and_unref(dev, NULL, errp);
+ sysbus_realize_and_unref(SYS_BUS_DEVICE(dev), errp);
Whitespace changes minimized manually.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Acked-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20200610053247.1583243-46-armbru@redhat.com>
[Conflicts in hw/misc/empty_slot.c and hw/sparc/leon3.c resolved]
2020-06-10 08:32:34 +03:00
|
|
|
sysbus_realize_and_unref(gicbusdev, &error_fatal);
|
2017-01-09 14:40:21 +03:00
|
|
|
sysbus_mmio_map(gicbusdev, 0, vms->memmap[VIRT_GIC_DIST].base);
|
2015-09-24 03:29:37 +03:00
|
|
|
if (type == 3) {
|
2017-01-09 14:40:21 +03:00
|
|
|
sysbus_mmio_map(gicbusdev, 1, vms->memmap[VIRT_GIC_REDIST].base);
|
2018-06-22 15:28:37 +03:00
|
|
|
if (nb_redist_regions == 2) {
|
2019-03-04 13:13:31 +03:00
|
|
|
sysbus_mmio_map(gicbusdev, 2,
|
|
|
|
vms->memmap[VIRT_HIGH_GIC_REDIST2].base);
|
2018-06-22 15:28:37 +03:00
|
|
|
}
|
2015-09-24 03:29:37 +03:00
|
|
|
} else {
|
2017-01-09 14:40:21 +03:00
|
|
|
sysbus_mmio_map(gicbusdev, 1, vms->memmap[VIRT_GIC_CPU].base);
|
2018-08-14 19:17:21 +03:00
|
|
|
if (vms->virt) {
|
|
|
|
sysbus_mmio_map(gicbusdev, 2, vms->memmap[VIRT_GIC_HYP].base);
|
|
|
|
sysbus_mmio_map(gicbusdev, 3, vms->memmap[VIRT_GIC_VCPU].base);
|
|
|
|
}
|
2015-09-24 03:29:37 +03:00
|
|
|
}
|
2014-05-01 18:24:46 +04:00
|
|
|
|
2017-01-20 14:15:09 +03:00
|
|
|
/* Wire the outputs from each CPU's generic timer and the GICv3
|
|
|
|
* maintenance interrupt signal to the appropriate GIC PPI inputs,
|
|
|
|
* and the GIC's IRQ/FIQ/VIRQ/VFIQ interrupt outputs to the CPU's inputs.
|
2014-05-01 18:24:46 +04:00
|
|
|
*/
|
|
|
|
for (i = 0; i < smp_cpus; i++) {
|
|
|
|
DeviceState *cpudev = DEVICE(qemu_get_cpu(i));
|
2015-08-13 13:26:18 +03:00
|
|
|
int ppibase = NUM_IRQS + i * GIC_INTERNAL + GIC_NR_SGIS;
|
2015-08-13 13:26:22 +03:00
|
|
|
int irq;
|
|
|
|
/* Mapping from the output timer irq lines from the CPU to the
|
|
|
|
* GIC PPI inputs we use for the virt board.
|
2014-05-01 18:24:46 +04:00
|
|
|
*/
|
2015-08-13 13:26:22 +03:00
|
|
|
const int timer_irq[] = {
|
|
|
|
[GTIMER_PHYS] = ARCH_TIMER_NS_EL1_IRQ,
|
|
|
|
[GTIMER_VIRT] = ARCH_TIMER_VIRT_IRQ,
|
|
|
|
[GTIMER_HYP] = ARCH_TIMER_NS_EL2_IRQ,
|
|
|
|
[GTIMER_SEC] = ARCH_TIMER_S_EL1_IRQ,
|
|
|
|
};
|
|
|
|
|
|
|
|
for (irq = 0; irq < ARRAY_SIZE(timer_irq); irq++) {
|
|
|
|
qdev_connect_gpio_out(cpudev, irq,
|
2019-12-09 12:03:06 +03:00
|
|
|
qdev_get_gpio_in(vms->gic,
|
2015-08-13 13:26:22 +03:00
|
|
|
ppibase + timer_irq[irq]));
|
|
|
|
}
|
2014-05-01 18:24:46 +04:00
|
|
|
|
2018-08-14 19:17:21 +03:00
|
|
|
if (type == 3) {
|
2019-12-09 12:03:06 +03:00
|
|
|
qemu_irq irq = qdev_get_gpio_in(vms->gic,
|
2018-08-14 19:17:21 +03:00
|
|
|
ppibase + ARCH_GIC_MAINT_IRQ);
|
|
|
|
qdev_connect_gpio_out_named(cpudev, "gicv3-maintenance-interrupt",
|
|
|
|
0, irq);
|
|
|
|
} else if (vms->virt) {
|
2019-12-09 12:03:06 +03:00
|
|
|
qemu_irq irq = qdev_get_gpio_in(vms->gic,
|
2018-08-14 19:17:21 +03:00
|
|
|
ppibase + ARCH_GIC_MAINT_IRQ);
|
|
|
|
sysbus_connect_irq(gicbusdev, i + 4 * smp_cpus, irq);
|
|
|
|
}
|
|
|
|
|
2017-09-04 17:21:53 +03:00
|
|
|
qdev_connect_gpio_out_named(cpudev, "pmu-interrupt", 0,
|
2019-12-09 12:03:06 +03:00
|
|
|
qdev_get_gpio_in(vms->gic, ppibase
|
2017-09-04 17:21:53 +03:00
|
|
|
+ VIRTUAL_PMU_IRQ));
|
2017-01-20 14:15:09 +03:00
|
|
|
|
2014-05-01 18:24:46 +04:00
|
|
|
sysbus_connect_irq(gicbusdev, i, qdev_get_gpio_in(cpudev, ARM_CPU_IRQ));
|
2015-05-12 13:57:18 +03:00
|
|
|
sysbus_connect_irq(gicbusdev, i + smp_cpus,
|
|
|
|
qdev_get_gpio_in(cpudev, ARM_CPU_FIQ));
|
2017-01-20 14:15:09 +03:00
|
|
|
sysbus_connect_irq(gicbusdev, i + 2 * smp_cpus,
|
|
|
|
qdev_get_gpio_in(cpudev, ARM_CPU_VIRQ));
|
|
|
|
sysbus_connect_irq(gicbusdev, i + 3 * smp_cpus,
|
|
|
|
qdev_get_gpio_in(cpudev, ARM_CPU_VFIQ));
|
2014-05-01 18:24:46 +04:00
|
|
|
}
|
|
|
|
|
2017-01-09 14:40:22 +03:00
|
|
|
fdt_add_gic_node(vms);
|
2015-06-02 16:56:23 +03:00
|
|
|
|
2017-02-28 15:08:16 +03:00
|
|
|
if (type == 3 && vms->its) {
|
2019-12-09 12:03:06 +03:00
|
|
|
create_its(vms);
|
2016-10-17 21:22:17 +03:00
|
|
|
} else if (type == 2) {
|
2019-12-09 12:03:06 +03:00
|
|
|
create_v2m(vms);
|
2015-09-24 03:29:37 +03:00
|
|
|
}
|
2014-05-01 18:24:46 +04:00
|
|
|
}
|
|
|
|
|
2019-12-09 12:03:06 +03:00
|
|
|
static void create_uart(const VirtMachineState *vms, int uart,
|
2016-12-07 16:20:22 +03:00
|
|
|
MemoryRegion *mem, Chardev *chr)
|
2013-11-22 21:17:14 +04:00
|
|
|
{
|
|
|
|
char *nodename;
|
2017-01-09 14:40:21 +03:00
|
|
|
hwaddr base = vms->memmap[uart].base;
|
|
|
|
hwaddr size = vms->memmap[uart].size;
|
|
|
|
int irq = vms->irqmap[uart];
|
2013-11-22 21:17:14 +04:00
|
|
|
const char compat[] = "arm,pl011\0arm,primecell";
|
|
|
|
const char clocknames[] = "uartclk\0apb_pclk";
|
qdev: Convert uses of qdev_create() with Coccinelle
This is the transformation explained in the commit before previous.
Takes care of just one pattern that needs conversion. More to come in
this series.
Coccinelle script:
@ depends on !(file in "hw/arm/highbank.c")@
expression bus, type_name, dev, expr;
@@
- dev = qdev_create(bus, type_name);
+ dev = qdev_new(type_name);
... when != dev = expr
- qdev_init_nofail(dev);
+ qdev_realize_and_unref(dev, bus, &error_fatal);
@@
expression bus, type_name, dev, expr;
identifier DOWN;
@@
- dev = DOWN(qdev_create(bus, type_name));
+ dev = DOWN(qdev_new(type_name));
... when != dev = expr
- qdev_init_nofail(DEVICE(dev));
+ qdev_realize_and_unref(DEVICE(dev), bus, &error_fatal);
@@
expression bus, type_name, expr;
identifier dev;
@@
- DeviceState *dev = qdev_create(bus, type_name);
+ DeviceState *dev = qdev_new(type_name);
... when != dev = expr
- qdev_init_nofail(dev);
+ qdev_realize_and_unref(dev, bus, &error_fatal);
@@
expression bus, type_name, dev, expr, errp;
symbol true;
@@
- dev = qdev_create(bus, type_name);
+ dev = qdev_new(type_name);
... when != dev = expr
- object_property_set_bool(OBJECT(dev), true, "realized", errp);
+ qdev_realize_and_unref(dev, bus, errp);
@@
expression bus, type_name, expr, errp;
identifier dev;
symbol true;
@@
- DeviceState *dev = qdev_create(bus, type_name);
+ DeviceState *dev = qdev_new(type_name);
... when != dev = expr
- object_property_set_bool(OBJECT(dev), true, "realized", errp);
+ qdev_realize_and_unref(dev, bus, errp);
The first rule exempts hw/arm/highbank.c, because it matches along two
control flow paths there, with different @type_name. Covered by the
next commit's manual conversions.
Missing #include "qapi/error.h" added manually.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20200610053247.1583243-10-armbru@redhat.com>
[Conflicts in hw/misc/empty_slot.c and hw/sparc/leon3.c resolved]
2020-06-10 08:31:58 +03:00
|
|
|
DeviceState *dev = qdev_new(TYPE_PL011);
|
2016-01-21 17:15:07 +03:00
|
|
|
SysBusDevice *s = SYS_BUS_DEVICE(dev);
|
2021-03-03 20:36:36 +03:00
|
|
|
MachineState *ms = MACHINE(vms);
|
2013-11-22 21:17:14 +04:00
|
|
|
|
2016-06-08 05:30:45 +03:00
|
|
|
qdev_prop_set_chr(dev, "chardev", chr);
|
sysbus: Convert to sysbus_realize() etc. with Coccinelle
Convert from qdev_realize(), qdev_realize_and_unref() with null @bus
argument to sysbus_realize(), sysbus_realize_and_unref().
Coccinelle script:
@@
expression dev, errp;
@@
- qdev_realize(DEVICE(dev), NULL, errp);
+ sysbus_realize(SYS_BUS_DEVICE(dev), errp);
@@
expression sysbus_dev, dev, errp;
@@
+ sysbus_dev = SYS_BUS_DEVICE(dev);
- qdev_realize_and_unref(dev, NULL, errp);
+ sysbus_realize_and_unref(sysbus_dev, errp);
- sysbus_dev = SYS_BUS_DEVICE(dev);
@@
expression sysbus_dev, dev, errp;
expression expr;
@@
sysbus_dev = SYS_BUS_DEVICE(dev);
... when != dev = expr;
- qdev_realize_and_unref(dev, NULL, errp);
+ sysbus_realize_and_unref(sysbus_dev, errp);
@@
expression dev, errp;
@@
- qdev_realize_and_unref(DEVICE(dev), NULL, errp);
+ sysbus_realize_and_unref(SYS_BUS_DEVICE(dev), errp);
@@
expression dev, errp;
@@
- qdev_realize_and_unref(dev, NULL, errp);
+ sysbus_realize_and_unref(SYS_BUS_DEVICE(dev), errp);
Whitespace changes minimized manually.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Acked-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20200610053247.1583243-46-armbru@redhat.com>
[Conflicts in hw/misc/empty_slot.c and hw/sparc/leon3.c resolved]
2020-06-10 08:32:34 +03:00
|
|
|
sysbus_realize_and_unref(SYS_BUS_DEVICE(dev), &error_fatal);
|
2016-01-21 17:15:07 +03:00
|
|
|
memory_region_add_subregion(mem, base,
|
|
|
|
sysbus_mmio_get_region(s, 0));
|
2019-12-09 12:03:06 +03:00
|
|
|
sysbus_connect_irq(s, 0, qdev_get_gpio_in(vms->gic, irq));
|
2013-11-22 21:17:14 +04:00
|
|
|
|
|
|
|
nodename = g_strdup_printf("/pl011@%" PRIx64, base);
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_add_subnode(ms->fdt, nodename);
|
2013-11-22 21:17:14 +04:00
|
|
|
/* Note that we can't use setprop_string because of the embedded NUL */
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop(ms->fdt, nodename, "compatible",
|
2013-11-22 21:17:14 +04:00
|
|
|
compat, sizeof(compat));
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_sized_cells(ms->fdt, nodename, "reg",
|
2013-11-22 21:17:14 +04:00
|
|
|
2, base, 2, size);
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_cells(ms->fdt, nodename, "interrupts",
|
2013-11-22 21:17:14 +04:00
|
|
|
GIC_FDT_IRQ_TYPE_SPI, irq,
|
2014-09-12 17:06:50 +04:00
|
|
|
GIC_FDT_IRQ_FLAGS_LEVEL_HI);
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_cells(ms->fdt, nodename, "clocks",
|
2017-01-09 14:40:21 +03:00
|
|
|
vms->clock_phandle, vms->clock_phandle);
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop(ms->fdt, nodename, "clock-names",
|
2013-11-22 21:17:14 +04:00
|
|
|
clocknames, sizeof(clocknames));
|
2014-09-12 17:06:47 +04:00
|
|
|
|
2016-01-21 17:15:07 +03:00
|
|
|
if (uart == VIRT_UART) {
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_string(ms->fdt, "/chosen", "stdout-path", nodename);
|
2016-01-21 17:15:07 +03:00
|
|
|
} else {
|
|
|
|
/* Mark as not usable by the normal world */
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_string(ms->fdt, nodename, "status", "disabled");
|
|
|
|
qemu_fdt_setprop_string(ms->fdt, nodename, "secure-status", "okay");
|
2018-10-05 11:07:29 +03:00
|
|
|
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_string(ms->fdt, "/secure-chosen", "stdout-path",
|
2018-10-05 11:07:29 +03:00
|
|
|
nodename);
|
2016-01-21 17:15:07 +03:00
|
|
|
}
|
|
|
|
|
2013-11-22 21:17:14 +04:00
|
|
|
g_free(nodename);
|
|
|
|
}
|
|
|
|
|
2019-12-09 12:03:06 +03:00
|
|
|
static void create_rtc(const VirtMachineState *vms)
|
2014-06-29 21:38:39 +04:00
|
|
|
{
|
|
|
|
char *nodename;
|
2017-01-09 14:40:21 +03:00
|
|
|
hwaddr base = vms->memmap[VIRT_RTC].base;
|
|
|
|
hwaddr size = vms->memmap[VIRT_RTC].size;
|
|
|
|
int irq = vms->irqmap[VIRT_RTC];
|
2014-06-29 21:38:39 +04:00
|
|
|
const char compat[] = "arm,pl031\0arm,primecell";
|
2021-03-03 20:36:36 +03:00
|
|
|
MachineState *ms = MACHINE(vms);
|
2014-06-29 21:38:39 +04:00
|
|
|
|
2019-12-09 12:03:06 +03:00
|
|
|
sysbus_create_simple("pl031", base, qdev_get_gpio_in(vms->gic, irq));
|
2014-06-29 21:38:39 +04:00
|
|
|
|
|
|
|
nodename = g_strdup_printf("/pl031@%" PRIx64, base);
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_add_subnode(ms->fdt, nodename);
|
|
|
|
qemu_fdt_setprop(ms->fdt, nodename, "compatible", compat, sizeof(compat));
|
|
|
|
qemu_fdt_setprop_sized_cells(ms->fdt, nodename, "reg",
|
2014-06-29 21:38:39 +04:00
|
|
|
2, base, 2, size);
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_cells(ms->fdt, nodename, "interrupts",
|
2014-06-29 21:38:39 +04:00
|
|
|
GIC_FDT_IRQ_TYPE_SPI, irq,
|
2014-09-12 17:06:50 +04:00
|
|
|
GIC_FDT_IRQ_FLAGS_LEVEL_HI);
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_cell(ms->fdt, nodename, "clocks", vms->clock_phandle);
|
|
|
|
qemu_fdt_setprop_string(ms->fdt, nodename, "clock-names", "apb_pclk");
|
2014-06-29 21:38:39 +04:00
|
|
|
g_free(nodename);
|
|
|
|
}
|
|
|
|
|
2016-03-30 19:27:22 +03:00
|
|
|
static DeviceState *gpio_key_dev;
|
2015-12-17 16:37:14 +03:00
|
|
|
static void virt_powerdown_req(Notifier *n, void *opaque)
|
|
|
|
{
|
2019-09-18 16:06:30 +03:00
|
|
|
VirtMachineState *s = container_of(n, VirtMachineState, powerdown_notifier);
|
|
|
|
|
|
|
|
if (s->acpi_dev) {
|
|
|
|
acpi_send_event(s->acpi_dev, ACPI_POWER_DOWN_STATUS);
|
|
|
|
} else {
|
|
|
|
/* use gpio Pin 3 for power button event */
|
|
|
|
qemu_set_irq(qdev_get_gpio_in(gpio_key_dev, 0), 1);
|
|
|
|
}
|
2015-12-17 16:37:14 +03:00
|
|
|
}
|
|
|
|
|
2021-03-03 20:36:36 +03:00
|
|
|
static void create_gpio_keys(char *fdt, DeviceState *pl061_dev,
|
2021-01-28 15:00:10 +03:00
|
|
|
uint32_t phandle)
|
|
|
|
{
|
|
|
|
gpio_key_dev = sysbus_create_simple("gpio-key", -1,
|
|
|
|
qdev_get_gpio_in(pl061_dev, 3));
|
|
|
|
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_add_subnode(fdt, "/gpio-keys");
|
|
|
|
qemu_fdt_setprop_string(fdt, "/gpio-keys", "compatible", "gpio-keys");
|
|
|
|
qemu_fdt_setprop_cell(fdt, "/gpio-keys", "#size-cells", 0);
|
|
|
|
qemu_fdt_setprop_cell(fdt, "/gpio-keys", "#address-cells", 1);
|
2021-01-28 15:00:10 +03:00
|
|
|
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_add_subnode(fdt, "/gpio-keys/poweroff");
|
|
|
|
qemu_fdt_setprop_string(fdt, "/gpio-keys/poweroff",
|
2021-01-28 15:00:10 +03:00
|
|
|
"label", "GPIO Key Poweroff");
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_cell(fdt, "/gpio-keys/poweroff", "linux,code",
|
2021-01-28 15:00:10 +03:00
|
|
|
KEY_POWER);
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_cells(fdt, "/gpio-keys/poweroff",
|
2021-01-28 15:00:10 +03:00
|
|
|
"gpios", phandle, 3, 0);
|
|
|
|
}
|
|
|
|
|
2021-01-28 15:00:11 +03:00
|
|
|
#define SECURE_GPIO_POWEROFF 0
|
|
|
|
#define SECURE_GPIO_RESET 1
|
|
|
|
|
2021-03-03 20:36:36 +03:00
|
|
|
static void create_secure_gpio_pwr(char *fdt, DeviceState *pl061_dev,
|
2021-01-28 15:00:11 +03:00
|
|
|
uint32_t phandle)
|
|
|
|
{
|
|
|
|
DeviceState *gpio_pwr_dev;
|
|
|
|
|
|
|
|
/* gpio-pwr */
|
|
|
|
gpio_pwr_dev = sysbus_create_simple("gpio-pwr", -1, NULL);
|
|
|
|
|
|
|
|
/* connect secure pl061 to gpio-pwr */
|
|
|
|
qdev_connect_gpio_out(pl061_dev, SECURE_GPIO_RESET,
|
|
|
|
qdev_get_gpio_in_named(gpio_pwr_dev, "reset", 0));
|
|
|
|
qdev_connect_gpio_out(pl061_dev, SECURE_GPIO_POWEROFF,
|
|
|
|
qdev_get_gpio_in_named(gpio_pwr_dev, "shutdown", 0));
|
|
|
|
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_add_subnode(fdt, "/gpio-poweroff");
|
|
|
|
qemu_fdt_setprop_string(fdt, "/gpio-poweroff", "compatible",
|
2021-01-28 15:00:11 +03:00
|
|
|
"gpio-poweroff");
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_cells(fdt, "/gpio-poweroff",
|
2021-01-28 15:00:11 +03:00
|
|
|
"gpios", phandle, SECURE_GPIO_POWEROFF, 0);
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_string(fdt, "/gpio-poweroff", "status", "disabled");
|
|
|
|
qemu_fdt_setprop_string(fdt, "/gpio-poweroff", "secure-status",
|
2021-01-28 15:00:11 +03:00
|
|
|
"okay");
|
|
|
|
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_add_subnode(fdt, "/gpio-restart");
|
|
|
|
qemu_fdt_setprop_string(fdt, "/gpio-restart", "compatible",
|
2021-01-28 15:00:11 +03:00
|
|
|
"gpio-restart");
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_cells(fdt, "/gpio-restart",
|
2021-01-28 15:00:11 +03:00
|
|
|
"gpios", phandle, SECURE_GPIO_RESET, 0);
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_string(fdt, "/gpio-restart", "status", "disabled");
|
|
|
|
qemu_fdt_setprop_string(fdt, "/gpio-restart", "secure-status",
|
2021-01-28 15:00:11 +03:00
|
|
|
"okay");
|
|
|
|
}
|
|
|
|
|
2021-01-28 15:00:10 +03:00
|
|
|
static void create_gpio_devices(const VirtMachineState *vms, int gpio,
|
|
|
|
MemoryRegion *mem)
|
2015-12-17 16:37:13 +03:00
|
|
|
{
|
|
|
|
char *nodename;
|
2016-03-30 19:27:22 +03:00
|
|
|
DeviceState *pl061_dev;
|
2021-01-28 15:00:10 +03:00
|
|
|
hwaddr base = vms->memmap[gpio].base;
|
|
|
|
hwaddr size = vms->memmap[gpio].size;
|
|
|
|
int irq = vms->irqmap[gpio];
|
2015-12-17 16:37:13 +03:00
|
|
|
const char compat[] = "arm,pl061\0arm,primecell";
|
2021-01-28 15:00:10 +03:00
|
|
|
SysBusDevice *s;
|
2021-03-03 20:36:36 +03:00
|
|
|
MachineState *ms = MACHINE(vms);
|
2015-12-17 16:37:13 +03:00
|
|
|
|
2021-01-28 15:00:10 +03:00
|
|
|
pl061_dev = qdev_new("pl061");
|
2021-07-02 13:40:15 +03:00
|
|
|
/* Pull lines down to 0 if not driven by the PL061 */
|
|
|
|
qdev_prop_set_uint32(pl061_dev, "pullups", 0);
|
|
|
|
qdev_prop_set_uint32(pl061_dev, "pulldowns", 0xff);
|
2021-01-28 15:00:10 +03:00
|
|
|
s = SYS_BUS_DEVICE(pl061_dev);
|
|
|
|
sysbus_realize_and_unref(s, &error_fatal);
|
|
|
|
memory_region_add_subregion(mem, base, sysbus_mmio_get_region(s, 0));
|
|
|
|
sysbus_connect_irq(s, 0, qdev_get_gpio_in(vms->gic, irq));
|
2015-12-17 16:37:13 +03:00
|
|
|
|
2021-03-03 20:36:36 +03:00
|
|
|
uint32_t phandle = qemu_fdt_alloc_phandle(ms->fdt);
|
2015-12-17 16:37:13 +03:00
|
|
|
nodename = g_strdup_printf("/pl061@%" PRIx64, base);
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_add_subnode(ms->fdt, nodename);
|
|
|
|
qemu_fdt_setprop_sized_cells(ms->fdt, nodename, "reg",
|
2015-12-17 16:37:13 +03:00
|
|
|
2, base, 2, size);
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop(ms->fdt, nodename, "compatible", compat, sizeof(compat));
|
|
|
|
qemu_fdt_setprop_cell(ms->fdt, nodename, "#gpio-cells", 2);
|
|
|
|
qemu_fdt_setprop(ms->fdt, nodename, "gpio-controller", NULL, 0);
|
|
|
|
qemu_fdt_setprop_cells(ms->fdt, nodename, "interrupts",
|
2015-12-17 16:37:13 +03:00
|
|
|
GIC_FDT_IRQ_TYPE_SPI, irq,
|
|
|
|
GIC_FDT_IRQ_FLAGS_LEVEL_HI);
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_cell(ms->fdt, nodename, "clocks", vms->clock_phandle);
|
|
|
|
qemu_fdt_setprop_string(ms->fdt, nodename, "clock-names", "apb_pclk");
|
|
|
|
qemu_fdt_setprop_cell(ms->fdt, nodename, "phandle", phandle);
|
2015-12-17 16:37:14 +03:00
|
|
|
|
2021-01-28 15:00:11 +03:00
|
|
|
if (gpio != VIRT_GPIO) {
|
|
|
|
/* Mark as not usable by the normal world */
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_string(ms->fdt, nodename, "status", "disabled");
|
|
|
|
qemu_fdt_setprop_string(ms->fdt, nodename, "secure-status", "okay");
|
2021-01-28 15:00:11 +03:00
|
|
|
}
|
2015-12-17 16:37:13 +03:00
|
|
|
g_free(nodename);
|
2021-01-28 15:00:10 +03:00
|
|
|
|
|
|
|
/* Child gpio devices */
|
2021-01-28 15:00:11 +03:00
|
|
|
if (gpio == VIRT_GPIO) {
|
2021-03-03 20:36:36 +03:00
|
|
|
create_gpio_keys(ms->fdt, pl061_dev, phandle);
|
2021-01-28 15:00:11 +03:00
|
|
|
} else {
|
2021-03-03 20:36:36 +03:00
|
|
|
create_secure_gpio_pwr(ms->fdt, pl061_dev, phandle);
|
2021-01-28 15:00:11 +03:00
|
|
|
}
|
2015-12-17 16:37:13 +03:00
|
|
|
}
|
|
|
|
|
2019-12-09 12:03:06 +03:00
|
|
|
static void create_virtio_devices(const VirtMachineState *vms)
|
2013-11-22 21:17:14 +04:00
|
|
|
{
|
|
|
|
int i;
|
2017-01-09 14:40:21 +03:00
|
|
|
hwaddr size = vms->memmap[VIRT_MMIO].size;
|
2021-03-03 20:36:36 +03:00
|
|
|
MachineState *ms = MACHINE(vms);
|
2013-11-22 21:17:14 +04:00
|
|
|
|
2015-02-05 16:37:23 +03:00
|
|
|
/* We create the transports in forwards order. Since qbus_realize()
|
|
|
|
* prepends (not appends) new child buses, the incrementing loop below will
|
|
|
|
* create a list of virtio-mmio buses with decreasing base addresses.
|
|
|
|
*
|
|
|
|
* When a -device option is processed from the command line,
|
|
|
|
* qbus_find_recursive() picks the next free virtio-mmio bus in forwards
|
|
|
|
* order. The upshot is that -device options in increasing command line
|
|
|
|
* order are mapped to virtio-mmio buses with decreasing base addresses.
|
|
|
|
*
|
|
|
|
* When this code was originally written, that arrangement ensured that the
|
|
|
|
* guest Linux kernel would give the lowest "name" (/dev/vda, eth0, etc) to
|
|
|
|
* the first -device on the command line. (The end-to-end order is a
|
|
|
|
* function of this loop, qbus_realize(), qbus_find_recursive(), and the
|
|
|
|
* guest kernel's name-to-address assignment strategy.)
|
|
|
|
*
|
|
|
|
* Meanwhile, the kernel's traversal seems to have been reversed; see eg.
|
|
|
|
* the message, if not necessarily the code, of commit 70161ff336.
|
|
|
|
* Therefore the loop now establishes the inverse of the original intent.
|
|
|
|
*
|
|
|
|
* Unfortunately, we can't counteract the kernel change by reversing the
|
|
|
|
* loop; it would break existing command lines.
|
|
|
|
*
|
|
|
|
* In any case, the kernel makes no guarantee about the stability of
|
|
|
|
* enumeration order of virtio devices (as demonstrated by it changing
|
|
|
|
* between kernel versions). For reliable and stable identification
|
|
|
|
* of disks users must use UUIDs or similar mechanisms.
|
2013-11-22 21:17:14 +04:00
|
|
|
*/
|
|
|
|
for (i = 0; i < NUM_VIRTIO_TRANSPORTS; i++) {
|
2017-01-09 14:40:21 +03:00
|
|
|
int irq = vms->irqmap[VIRT_MMIO] + i;
|
|
|
|
hwaddr base = vms->memmap[VIRT_MMIO].base + i * size;
|
2013-11-22 21:17:14 +04:00
|
|
|
|
2019-12-09 12:03:06 +03:00
|
|
|
sysbus_create_simple("virtio-mmio", base,
|
|
|
|
qdev_get_gpio_in(vms->gic, irq));
|
2013-11-22 21:17:14 +04:00
|
|
|
}
|
|
|
|
|
2015-02-05 16:37:23 +03:00
|
|
|
/* We add dtb nodes in reverse order so that they appear in the finished
|
|
|
|
* device tree lowest address first.
|
|
|
|
*
|
|
|
|
* Note that this mapping is independent of the loop above. The previous
|
|
|
|
* loop influences virtio device to virtio transport assignment, whereas
|
|
|
|
* this loop controls how virtio transports are laid out in the dtb.
|
|
|
|
*/
|
2013-11-22 21:17:14 +04:00
|
|
|
for (i = NUM_VIRTIO_TRANSPORTS - 1; i >= 0; i--) {
|
|
|
|
char *nodename;
|
2017-01-09 14:40:21 +03:00
|
|
|
int irq = vms->irqmap[VIRT_MMIO] + i;
|
|
|
|
hwaddr base = vms->memmap[VIRT_MMIO].base + i * size;
|
2013-11-22 21:17:14 +04:00
|
|
|
|
|
|
|
nodename = g_strdup_printf("/virtio_mmio@%" PRIx64, base);
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_add_subnode(ms->fdt, nodename);
|
|
|
|
qemu_fdt_setprop_string(ms->fdt, nodename,
|
2013-11-11 12:14:41 +04:00
|
|
|
"compatible", "virtio,mmio");
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_sized_cells(ms->fdt, nodename, "reg",
|
2013-11-11 12:14:41 +04:00
|
|
|
2, base, 2, size);
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_cells(ms->fdt, nodename, "interrupts",
|
2013-11-11 12:14:41 +04:00
|
|
|
GIC_FDT_IRQ_TYPE_SPI, irq,
|
|
|
|
GIC_FDT_IRQ_FLAGS_EDGE_LO_HI);
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop(ms->fdt, nodename, "dma-coherent", NULL, 0);
|
2013-11-22 21:17:14 +04:00
|
|
|
g_free(nodename);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
hw/arm/virt: Support firmware configuration with -blockdev
The ARM virt machines put firmware in flash memory. To configure it,
you use -drive if=pflash,unit=0,... and optionally -drive
if=pflash,unit=1,...
Why two -drive? This permits setting up one part of the flash memory
read-only, and the other part read/write. It also makes upgrading
firmware on the host easier. Below the hood, we get two separate
flash devices, because we were too lazy to improve our flash device
models to support sector protection.
The problem at hand is to do the same with -blockdev somehow, as one
more step towards deprecating -drive.
We recently solved this problem for x86 PC machines, in commit
ebc29e1beab. See the commit message for design rationale.
This commit solves it for ARM virt basically the same way: new machine
properties pflash0, pflash1 forward to the onboard flash devices'
properties. Requires creating the onboard devices in the
.instance_init() method virt_instance_init(). The existing code to
pick up drives defined with -drive if=pflash is replaced by code to
desugar into the machine properties.
There are a few behavioral differences, though:
* The flash devices are always present (x86: only present if
configured)
* Flash base addresses and sizes are fixed (x86: sizes depend on
images, mapped back to back below a fixed address)
* -bios configures contents of first pflash (x86: -bios configures ROM
contents)
* -bios is rejected when first pflash is also configured with -machine
pflash0=... (x86: bios is silently ignored then)
* -machine pflash1=... does not require -machine pflash0=... (x86: it
does).
The actual code is a bit simpler than for x86 mostly due to the first
two differences.
Before the patch, all the action is in create_flash(), called from the
machine's .init() method machvirt_init():
main()
machine_run_board_init()
machvirt_init()
create_flash()
create_one_flash() for flash[0]
create
configure
includes obeying -drive if=pflash,unit=0
realize
map
fall back to -bios
create_one_flash() for flash[1]
create
configure
includes obeying -drive if=pflash,unit=1
realize
map
update FDT
To make the machine properties work, we need to move device creation
to its .instance_init() method virt_instance_init().
Another complication is machvirt_init()'s computation of
@firmware_loaded: it predicts what create_flash() will do. Instead of
predicting what create_flash()'s replacement virt_firmware_init() will
do, I decided to have virt_firmware_init() return what it did.
Requires calling it a bit earlier.
Resulting call tree:
main()
current_machine = object_new()
...
virt_instance_init()
virt_flash_create()
virt_flash_create1() for flash[0]
create
configure: set defaults
become child of machine [NEW]
add machine prop pflash0 as alias for drive [NEW]
virt_flash_create1() for flash[1]
create
configure: set defaults
become child of machine [NEW]
add machine prop pflash1 as alias for drive [NEW]
for all machine props from the command line: machine_set_property()
...
property_set_alias() for machine props pflash0, pflash1
...
set_drive() for cfi.pflash01 prop drive
this is how -machine pflash0=... etc set
machine_run_board_init(current_machine);
virt_firmware_init()
pflash_cfi01_legacy_drive()
legacy -drive if=pflash,unit=0 and =1 [NEW]
virt_flash_map()
virt_flash_map1() for flash[0]
configure: num-blocks
realize
map
virt_flash_map1() for flash[1]
configure: num-blocks
realize
map
fall back to -bios
virt_flash_fdt()
update FDT
You have László to thank for making me explain this in detail.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Acked-by: Laszlo Ersek <lersek@redhat.com>
Message-id: 20190416091348.26075-4-armbru@redhat.com
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
2019-05-07 14:55:02 +03:00
|
|
|
#define VIRT_FLASH_SECTOR_SIZE (256 * KiB)
|
|
|
|
|
|
|
|
static PFlashCFI01 *virt_flash_create1(VirtMachineState *vms,
|
|
|
|
const char *name,
|
|
|
|
const char *alias_prop_name)
|
2014-09-12 17:06:48 +04:00
|
|
|
{
|
hw/arm/virt: Support firmware configuration with -blockdev
The ARM virt machines put firmware in flash memory. To configure it,
you use -drive if=pflash,unit=0,... and optionally -drive
if=pflash,unit=1,...
Why two -drive? This permits setting up one part of the flash memory
read-only, and the other part read/write. It also makes upgrading
firmware on the host easier. Below the hood, we get two separate
flash devices, because we were too lazy to improve our flash device
models to support sector protection.
The problem at hand is to do the same with -blockdev somehow, as one
more step towards deprecating -drive.
We recently solved this problem for x86 PC machines, in commit
ebc29e1beab. See the commit message for design rationale.
This commit solves it for ARM virt basically the same way: new machine
properties pflash0, pflash1 forward to the onboard flash devices'
properties. Requires creating the onboard devices in the
.instance_init() method virt_instance_init(). The existing code to
pick up drives defined with -drive if=pflash is replaced by code to
desugar into the machine properties.
There are a few behavioral differences, though:
* The flash devices are always present (x86: only present if
configured)
* Flash base addresses and sizes are fixed (x86: sizes depend on
images, mapped back to back below a fixed address)
* -bios configures contents of first pflash (x86: -bios configures ROM
contents)
* -bios is rejected when first pflash is also configured with -machine
pflash0=... (x86: bios is silently ignored then)
* -machine pflash1=... does not require -machine pflash0=... (x86: it
does).
The actual code is a bit simpler than for x86 mostly due to the first
two differences.
Before the patch, all the action is in create_flash(), called from the
machine's .init() method machvirt_init():
main()
machine_run_board_init()
machvirt_init()
create_flash()
create_one_flash() for flash[0]
create
configure
includes obeying -drive if=pflash,unit=0
realize
map
fall back to -bios
create_one_flash() for flash[1]
create
configure
includes obeying -drive if=pflash,unit=1
realize
map
update FDT
To make the machine properties work, we need to move device creation
to its .instance_init() method virt_instance_init().
Another complication is machvirt_init()'s computation of
@firmware_loaded: it predicts what create_flash() will do. Instead of
predicting what create_flash()'s replacement virt_firmware_init() will
do, I decided to have virt_firmware_init() return what it did.
Requires calling it a bit earlier.
Resulting call tree:
main()
current_machine = object_new()
...
virt_instance_init()
virt_flash_create()
virt_flash_create1() for flash[0]
create
configure: set defaults
become child of machine [NEW]
add machine prop pflash0 as alias for drive [NEW]
virt_flash_create1() for flash[1]
create
configure: set defaults
become child of machine [NEW]
add machine prop pflash1 as alias for drive [NEW]
for all machine props from the command line: machine_set_property()
...
property_set_alias() for machine props pflash0, pflash1
...
set_drive() for cfi.pflash01 prop drive
this is how -machine pflash0=... etc set
machine_run_board_init(current_machine);
virt_firmware_init()
pflash_cfi01_legacy_drive()
legacy -drive if=pflash,unit=0 and =1 [NEW]
virt_flash_map()
virt_flash_map1() for flash[0]
configure: num-blocks
realize
map
virt_flash_map1() for flash[1]
configure: num-blocks
realize
map
fall back to -bios
virt_flash_fdt()
update FDT
You have László to thank for making me explain this in detail.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Acked-by: Laszlo Ersek <lersek@redhat.com>
Message-id: 20190416091348.26075-4-armbru@redhat.com
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
2019-05-07 14:55:02 +03:00
|
|
|
/*
|
|
|
|
* Create a single flash device. We use the same parameters as
|
|
|
|
* the flash devices on the Versatile Express board.
|
2014-09-12 17:06:48 +04:00
|
|
|
*/
|
2020-06-10 08:31:59 +03:00
|
|
|
DeviceState *dev = qdev_new(TYPE_PFLASH_CFI01);
|
2014-09-12 17:06:48 +04:00
|
|
|
|
hw/arm/virt: Support firmware configuration with -blockdev
The ARM virt machines put firmware in flash memory. To configure it,
you use -drive if=pflash,unit=0,... and optionally -drive
if=pflash,unit=1,...
Why two -drive? This permits setting up one part of the flash memory
read-only, and the other part read/write. It also makes upgrading
firmware on the host easier. Below the hood, we get two separate
flash devices, because we were too lazy to improve our flash device
models to support sector protection.
The problem at hand is to do the same with -blockdev somehow, as one
more step towards deprecating -drive.
We recently solved this problem for x86 PC machines, in commit
ebc29e1beab. See the commit message for design rationale.
This commit solves it for ARM virt basically the same way: new machine
properties pflash0, pflash1 forward to the onboard flash devices'
properties. Requires creating the onboard devices in the
.instance_init() method virt_instance_init(). The existing code to
pick up drives defined with -drive if=pflash is replaced by code to
desugar into the machine properties.
There are a few behavioral differences, though:
* The flash devices are always present (x86: only present if
configured)
* Flash base addresses and sizes are fixed (x86: sizes depend on
images, mapped back to back below a fixed address)
* -bios configures contents of first pflash (x86: -bios configures ROM
contents)
* -bios is rejected when first pflash is also configured with -machine
pflash0=... (x86: bios is silently ignored then)
* -machine pflash1=... does not require -machine pflash0=... (x86: it
does).
The actual code is a bit simpler than for x86 mostly due to the first
two differences.
Before the patch, all the action is in create_flash(), called from the
machine's .init() method machvirt_init():
main()
machine_run_board_init()
machvirt_init()
create_flash()
create_one_flash() for flash[0]
create
configure
includes obeying -drive if=pflash,unit=0
realize
map
fall back to -bios
create_one_flash() for flash[1]
create
configure
includes obeying -drive if=pflash,unit=1
realize
map
update FDT
To make the machine properties work, we need to move device creation
to its .instance_init() method virt_instance_init().
Another complication is machvirt_init()'s computation of
@firmware_loaded: it predicts what create_flash() will do. Instead of
predicting what create_flash()'s replacement virt_firmware_init() will
do, I decided to have virt_firmware_init() return what it did.
Requires calling it a bit earlier.
Resulting call tree:
main()
current_machine = object_new()
...
virt_instance_init()
virt_flash_create()
virt_flash_create1() for flash[0]
create
configure: set defaults
become child of machine [NEW]
add machine prop pflash0 as alias for drive [NEW]
virt_flash_create1() for flash[1]
create
configure: set defaults
become child of machine [NEW]
add machine prop pflash1 as alias for drive [NEW]
for all machine props from the command line: machine_set_property()
...
property_set_alias() for machine props pflash0, pflash1
...
set_drive() for cfi.pflash01 prop drive
this is how -machine pflash0=... etc set
machine_run_board_init(current_machine);
virt_firmware_init()
pflash_cfi01_legacy_drive()
legacy -drive if=pflash,unit=0 and =1 [NEW]
virt_flash_map()
virt_flash_map1() for flash[0]
configure: num-blocks
realize
map
virt_flash_map1() for flash[1]
configure: num-blocks
realize
map
fall back to -bios
virt_flash_fdt()
update FDT
You have László to thank for making me explain this in detail.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Acked-by: Laszlo Ersek <lersek@redhat.com>
Message-id: 20190416091348.26075-4-armbru@redhat.com
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
2019-05-07 14:55:02 +03:00
|
|
|
qdev_prop_set_uint64(dev, "sector-length", VIRT_FLASH_SECTOR_SIZE);
|
2014-09-12 17:06:48 +04:00
|
|
|
qdev_prop_set_uint8(dev, "width", 4);
|
|
|
|
qdev_prop_set_uint8(dev, "device-width", 2);
|
2015-04-08 14:53:29 +03:00
|
|
|
qdev_prop_set_bit(dev, "big-endian", false);
|
2014-09-12 17:06:48 +04:00
|
|
|
qdev_prop_set_uint16(dev, "id0", 0x89);
|
|
|
|
qdev_prop_set_uint16(dev, "id1", 0x18);
|
|
|
|
qdev_prop_set_uint16(dev, "id2", 0x00);
|
|
|
|
qdev_prop_set_uint16(dev, "id3", 0x00);
|
|
|
|
qdev_prop_set_string(dev, "name", name);
|
qom: Drop parameter @errp of object_property_add() & friends
The only way object_property_add() can fail is when a property with
the same name already exists. Since our property names are all
hardcoded, failure is a programming error, and the appropriate way to
handle it is passing &error_abort.
Same for its variants, except for object_property_add_child(), which
additionally fails when the child already has a parent. Parentage is
also under program control, so this is a programming error, too.
We have a bit over 500 callers. Almost half of them pass
&error_abort, slightly fewer ignore errors, one test case handles
errors, and the remaining few callers pass them to their own callers.
The previous few commits demonstrated once again that ignoring
programming errors is a bad idea.
Of the few ones that pass on errors, several violate the Error API.
The Error ** argument must be NULL, &error_abort, &error_fatal, or a
pointer to a variable containing NULL. Passing an argument of the
latter kind twice without clearing it in between is wrong: if the
first call sets an error, it no longer points to NULL for the second
call. ich9_pm_add_properties(), sparc32_ledma_realize(),
sparc32_dma_realize(), xilinx_axidma_realize(), xilinx_enet_realize()
are wrong that way.
When the one appropriate choice of argument is &error_abort, letting
users pick the argument is a bad idea.
Drop parameter @errp and assert the preconditions instead.
There's one exception to "duplicate property name is a programming
error": the way object_property_add() implements the magic (and
undocumented) "automatic arrayification". Don't drop @errp there.
Instead, rename object_property_add() to object_property_try_add(),
and add the obvious wrapper object_property_add().
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20200505152926.18877-15-armbru@redhat.com>
[Two semantic rebase conflicts resolved]
2020-05-05 18:29:22 +03:00
|
|
|
object_property_add_child(OBJECT(vms), name, OBJECT(dev));
|
hw/arm/virt: Support firmware configuration with -blockdev
The ARM virt machines put firmware in flash memory. To configure it,
you use -drive if=pflash,unit=0,... and optionally -drive
if=pflash,unit=1,...
Why two -drive? This permits setting up one part of the flash memory
read-only, and the other part read/write. It also makes upgrading
firmware on the host easier. Below the hood, we get two separate
flash devices, because we were too lazy to improve our flash device
models to support sector protection.
The problem at hand is to do the same with -blockdev somehow, as one
more step towards deprecating -drive.
We recently solved this problem for x86 PC machines, in commit
ebc29e1beab. See the commit message for design rationale.
This commit solves it for ARM virt basically the same way: new machine
properties pflash0, pflash1 forward to the onboard flash devices'
properties. Requires creating the onboard devices in the
.instance_init() method virt_instance_init(). The existing code to
pick up drives defined with -drive if=pflash is replaced by code to
desugar into the machine properties.
There are a few behavioral differences, though:
* The flash devices are always present (x86: only present if
configured)
* Flash base addresses and sizes are fixed (x86: sizes depend on
images, mapped back to back below a fixed address)
* -bios configures contents of first pflash (x86: -bios configures ROM
contents)
* -bios is rejected when first pflash is also configured with -machine
pflash0=... (x86: bios is silently ignored then)
* -machine pflash1=... does not require -machine pflash0=... (x86: it
does).
The actual code is a bit simpler than for x86 mostly due to the first
two differences.
Before the patch, all the action is in create_flash(), called from the
machine's .init() method machvirt_init():
main()
machine_run_board_init()
machvirt_init()
create_flash()
create_one_flash() for flash[0]
create
configure
includes obeying -drive if=pflash,unit=0
realize
map
fall back to -bios
create_one_flash() for flash[1]
create
configure
includes obeying -drive if=pflash,unit=1
realize
map
update FDT
To make the machine properties work, we need to move device creation
to its .instance_init() method virt_instance_init().
Another complication is machvirt_init()'s computation of
@firmware_loaded: it predicts what create_flash() will do. Instead of
predicting what create_flash()'s replacement virt_firmware_init() will
do, I decided to have virt_firmware_init() return what it did.
Requires calling it a bit earlier.
Resulting call tree:
main()
current_machine = object_new()
...
virt_instance_init()
virt_flash_create()
virt_flash_create1() for flash[0]
create
configure: set defaults
become child of machine [NEW]
add machine prop pflash0 as alias for drive [NEW]
virt_flash_create1() for flash[1]
create
configure: set defaults
become child of machine [NEW]
add machine prop pflash1 as alias for drive [NEW]
for all machine props from the command line: machine_set_property()
...
property_set_alias() for machine props pflash0, pflash1
...
set_drive() for cfi.pflash01 prop drive
this is how -machine pflash0=... etc set
machine_run_board_init(current_machine);
virt_firmware_init()
pflash_cfi01_legacy_drive()
legacy -drive if=pflash,unit=0 and =1 [NEW]
virt_flash_map()
virt_flash_map1() for flash[0]
configure: num-blocks
realize
map
virt_flash_map1() for flash[1]
configure: num-blocks
realize
map
fall back to -bios
virt_flash_fdt()
update FDT
You have László to thank for making me explain this in detail.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Acked-by: Laszlo Ersek <lersek@redhat.com>
Message-id: 20190416091348.26075-4-armbru@redhat.com
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
2019-05-07 14:55:02 +03:00
|
|
|
object_property_add_alias(OBJECT(vms), alias_prop_name,
|
qom: Drop parameter @errp of object_property_add() & friends
The only way object_property_add() can fail is when a property with
the same name already exists. Since our property names are all
hardcoded, failure is a programming error, and the appropriate way to
handle it is passing &error_abort.
Same for its variants, except for object_property_add_child(), which
additionally fails when the child already has a parent. Parentage is
also under program control, so this is a programming error, too.
We have a bit over 500 callers. Almost half of them pass
&error_abort, slightly fewer ignore errors, one test case handles
errors, and the remaining few callers pass them to their own callers.
The previous few commits demonstrated once again that ignoring
programming errors is a bad idea.
Of the few ones that pass on errors, several violate the Error API.
The Error ** argument must be NULL, &error_abort, &error_fatal, or a
pointer to a variable containing NULL. Passing an argument of the
latter kind twice without clearing it in between is wrong: if the
first call sets an error, it no longer points to NULL for the second
call. ich9_pm_add_properties(), sparc32_ledma_realize(),
sparc32_dma_realize(), xilinx_axidma_realize(), xilinx_enet_realize()
are wrong that way.
When the one appropriate choice of argument is &error_abort, letting
users pick the argument is a bad idea.
Drop parameter @errp and assert the preconditions instead.
There's one exception to "duplicate property name is a programming
error": the way object_property_add() implements the magic (and
undocumented) "automatic arrayification". Don't drop @errp there.
Instead, rename object_property_add() to object_property_try_add(),
and add the obvious wrapper object_property_add().
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20200505152926.18877-15-armbru@redhat.com>
[Two semantic rebase conflicts resolved]
2020-05-05 18:29:22 +03:00
|
|
|
OBJECT(dev), "drive");
|
hw/arm/virt: Support firmware configuration with -blockdev
The ARM virt machines put firmware in flash memory. To configure it,
you use -drive if=pflash,unit=0,... and optionally -drive
if=pflash,unit=1,...
Why two -drive? This permits setting up one part of the flash memory
read-only, and the other part read/write. It also makes upgrading
firmware on the host easier. Below the hood, we get two separate
flash devices, because we were too lazy to improve our flash device
models to support sector protection.
The problem at hand is to do the same with -blockdev somehow, as one
more step towards deprecating -drive.
We recently solved this problem for x86 PC machines, in commit
ebc29e1beab. See the commit message for design rationale.
This commit solves it for ARM virt basically the same way: new machine
properties pflash0, pflash1 forward to the onboard flash devices'
properties. Requires creating the onboard devices in the
.instance_init() method virt_instance_init(). The existing code to
pick up drives defined with -drive if=pflash is replaced by code to
desugar into the machine properties.
There are a few behavioral differences, though:
* The flash devices are always present (x86: only present if
configured)
* Flash base addresses and sizes are fixed (x86: sizes depend on
images, mapped back to back below a fixed address)
* -bios configures contents of first pflash (x86: -bios configures ROM
contents)
* -bios is rejected when first pflash is also configured with -machine
pflash0=... (x86: bios is silently ignored then)
* -machine pflash1=... does not require -machine pflash0=... (x86: it
does).
The actual code is a bit simpler than for x86 mostly due to the first
two differences.
Before the patch, all the action is in create_flash(), called from the
machine's .init() method machvirt_init():
main()
machine_run_board_init()
machvirt_init()
create_flash()
create_one_flash() for flash[0]
create
configure
includes obeying -drive if=pflash,unit=0
realize
map
fall back to -bios
create_one_flash() for flash[1]
create
configure
includes obeying -drive if=pflash,unit=1
realize
map
update FDT
To make the machine properties work, we need to move device creation
to its .instance_init() method virt_instance_init().
Another complication is machvirt_init()'s computation of
@firmware_loaded: it predicts what create_flash() will do. Instead of
predicting what create_flash()'s replacement virt_firmware_init() will
do, I decided to have virt_firmware_init() return what it did.
Requires calling it a bit earlier.
Resulting call tree:
main()
current_machine = object_new()
...
virt_instance_init()
virt_flash_create()
virt_flash_create1() for flash[0]
create
configure: set defaults
become child of machine [NEW]
add machine prop pflash0 as alias for drive [NEW]
virt_flash_create1() for flash[1]
create
configure: set defaults
become child of machine [NEW]
add machine prop pflash1 as alias for drive [NEW]
for all machine props from the command line: machine_set_property()
...
property_set_alias() for machine props pflash0, pflash1
...
set_drive() for cfi.pflash01 prop drive
this is how -machine pflash0=... etc set
machine_run_board_init(current_machine);
virt_firmware_init()
pflash_cfi01_legacy_drive()
legacy -drive if=pflash,unit=0 and =1 [NEW]
virt_flash_map()
virt_flash_map1() for flash[0]
configure: num-blocks
realize
map
virt_flash_map1() for flash[1]
configure: num-blocks
realize
map
fall back to -bios
virt_flash_fdt()
update FDT
You have László to thank for making me explain this in detail.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Acked-by: Laszlo Ersek <lersek@redhat.com>
Message-id: 20190416091348.26075-4-armbru@redhat.com
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
2019-05-07 14:55:02 +03:00
|
|
|
return PFLASH_CFI01(dev);
|
|
|
|
}
|
2014-09-12 17:06:48 +04:00
|
|
|
|
hw/arm/virt: Support firmware configuration with -blockdev
The ARM virt machines put firmware in flash memory. To configure it,
you use -drive if=pflash,unit=0,... and optionally -drive
if=pflash,unit=1,...
Why two -drive? This permits setting up one part of the flash memory
read-only, and the other part read/write. It also makes upgrading
firmware on the host easier. Below the hood, we get two separate
flash devices, because we were too lazy to improve our flash device
models to support sector protection.
The problem at hand is to do the same with -blockdev somehow, as one
more step towards deprecating -drive.
We recently solved this problem for x86 PC machines, in commit
ebc29e1beab. See the commit message for design rationale.
This commit solves it for ARM virt basically the same way: new machine
properties pflash0, pflash1 forward to the onboard flash devices'
properties. Requires creating the onboard devices in the
.instance_init() method virt_instance_init(). The existing code to
pick up drives defined with -drive if=pflash is replaced by code to
desugar into the machine properties.
There are a few behavioral differences, though:
* The flash devices are always present (x86: only present if
configured)
* Flash base addresses and sizes are fixed (x86: sizes depend on
images, mapped back to back below a fixed address)
* -bios configures contents of first pflash (x86: -bios configures ROM
contents)
* -bios is rejected when first pflash is also configured with -machine
pflash0=... (x86: bios is silently ignored then)
* -machine pflash1=... does not require -machine pflash0=... (x86: it
does).
The actual code is a bit simpler than for x86 mostly due to the first
two differences.
Before the patch, all the action is in create_flash(), called from the
machine's .init() method machvirt_init():
main()
machine_run_board_init()
machvirt_init()
create_flash()
create_one_flash() for flash[0]
create
configure
includes obeying -drive if=pflash,unit=0
realize
map
fall back to -bios
create_one_flash() for flash[1]
create
configure
includes obeying -drive if=pflash,unit=1
realize
map
update FDT
To make the machine properties work, we need to move device creation
to its .instance_init() method virt_instance_init().
Another complication is machvirt_init()'s computation of
@firmware_loaded: it predicts what create_flash() will do. Instead of
predicting what create_flash()'s replacement virt_firmware_init() will
do, I decided to have virt_firmware_init() return what it did.
Requires calling it a bit earlier.
Resulting call tree:
main()
current_machine = object_new()
...
virt_instance_init()
virt_flash_create()
virt_flash_create1() for flash[0]
create
configure: set defaults
become child of machine [NEW]
add machine prop pflash0 as alias for drive [NEW]
virt_flash_create1() for flash[1]
create
configure: set defaults
become child of machine [NEW]
add machine prop pflash1 as alias for drive [NEW]
for all machine props from the command line: machine_set_property()
...
property_set_alias() for machine props pflash0, pflash1
...
set_drive() for cfi.pflash01 prop drive
this is how -machine pflash0=... etc set
machine_run_board_init(current_machine);
virt_firmware_init()
pflash_cfi01_legacy_drive()
legacy -drive if=pflash,unit=0 and =1 [NEW]
virt_flash_map()
virt_flash_map1() for flash[0]
configure: num-blocks
realize
map
virt_flash_map1() for flash[1]
configure: num-blocks
realize
map
fall back to -bios
virt_flash_fdt()
update FDT
You have László to thank for making me explain this in detail.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Acked-by: Laszlo Ersek <lersek@redhat.com>
Message-id: 20190416091348.26075-4-armbru@redhat.com
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
2019-05-07 14:55:02 +03:00
|
|
|
static void virt_flash_create(VirtMachineState *vms)
|
|
|
|
{
|
|
|
|
vms->flash[0] = virt_flash_create1(vms, "virt.flash0", "pflash0");
|
|
|
|
vms->flash[1] = virt_flash_create1(vms, "virt.flash1", "pflash1");
|
|
|
|
}
|
2014-09-12 17:06:48 +04:00
|
|
|
|
hw/arm/virt: Support firmware configuration with -blockdev
The ARM virt machines put firmware in flash memory. To configure it,
you use -drive if=pflash,unit=0,... and optionally -drive
if=pflash,unit=1,...
Why two -drive? This permits setting up one part of the flash memory
read-only, and the other part read/write. It also makes upgrading
firmware on the host easier. Below the hood, we get two separate
flash devices, because we were too lazy to improve our flash device
models to support sector protection.
The problem at hand is to do the same with -blockdev somehow, as one
more step towards deprecating -drive.
We recently solved this problem for x86 PC machines, in commit
ebc29e1beab. See the commit message for design rationale.
This commit solves it for ARM virt basically the same way: new machine
properties pflash0, pflash1 forward to the onboard flash devices'
properties. Requires creating the onboard devices in the
.instance_init() method virt_instance_init(). The existing code to
pick up drives defined with -drive if=pflash is replaced by code to
desugar into the machine properties.
There are a few behavioral differences, though:
* The flash devices are always present (x86: only present if
configured)
* Flash base addresses and sizes are fixed (x86: sizes depend on
images, mapped back to back below a fixed address)
* -bios configures contents of first pflash (x86: -bios configures ROM
contents)
* -bios is rejected when first pflash is also configured with -machine
pflash0=... (x86: bios is silently ignored then)
* -machine pflash1=... does not require -machine pflash0=... (x86: it
does).
The actual code is a bit simpler than for x86 mostly due to the first
two differences.
Before the patch, all the action is in create_flash(), called from the
machine's .init() method machvirt_init():
main()
machine_run_board_init()
machvirt_init()
create_flash()
create_one_flash() for flash[0]
create
configure
includes obeying -drive if=pflash,unit=0
realize
map
fall back to -bios
create_one_flash() for flash[1]
create
configure
includes obeying -drive if=pflash,unit=1
realize
map
update FDT
To make the machine properties work, we need to move device creation
to its .instance_init() method virt_instance_init().
Another complication is machvirt_init()'s computation of
@firmware_loaded: it predicts what create_flash() will do. Instead of
predicting what create_flash()'s replacement virt_firmware_init() will
do, I decided to have virt_firmware_init() return what it did.
Requires calling it a bit earlier.
Resulting call tree:
main()
current_machine = object_new()
...
virt_instance_init()
virt_flash_create()
virt_flash_create1() for flash[0]
create
configure: set defaults
become child of machine [NEW]
add machine prop pflash0 as alias for drive [NEW]
virt_flash_create1() for flash[1]
create
configure: set defaults
become child of machine [NEW]
add machine prop pflash1 as alias for drive [NEW]
for all machine props from the command line: machine_set_property()
...
property_set_alias() for machine props pflash0, pflash1
...
set_drive() for cfi.pflash01 prop drive
this is how -machine pflash0=... etc set
machine_run_board_init(current_machine);
virt_firmware_init()
pflash_cfi01_legacy_drive()
legacy -drive if=pflash,unit=0 and =1 [NEW]
virt_flash_map()
virt_flash_map1() for flash[0]
configure: num-blocks
realize
map
virt_flash_map1() for flash[1]
configure: num-blocks
realize
map
fall back to -bios
virt_flash_fdt()
update FDT
You have László to thank for making me explain this in detail.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Acked-by: Laszlo Ersek <lersek@redhat.com>
Message-id: 20190416091348.26075-4-armbru@redhat.com
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
2019-05-07 14:55:02 +03:00
|
|
|
static void virt_flash_map1(PFlashCFI01 *flash,
|
|
|
|
hwaddr base, hwaddr size,
|
|
|
|
MemoryRegion *sysmem)
|
|
|
|
{
|
|
|
|
DeviceState *dev = DEVICE(flash);
|
2014-09-12 17:06:48 +04:00
|
|
|
|
2020-05-11 23:52:46 +03:00
|
|
|
assert(QEMU_IS_ALIGNED(size, VIRT_FLASH_SECTOR_SIZE));
|
hw/arm/virt: Support firmware configuration with -blockdev
The ARM virt machines put firmware in flash memory. To configure it,
you use -drive if=pflash,unit=0,... and optionally -drive
if=pflash,unit=1,...
Why two -drive? This permits setting up one part of the flash memory
read-only, and the other part read/write. It also makes upgrading
firmware on the host easier. Below the hood, we get two separate
flash devices, because we were too lazy to improve our flash device
models to support sector protection.
The problem at hand is to do the same with -blockdev somehow, as one
more step towards deprecating -drive.
We recently solved this problem for x86 PC machines, in commit
ebc29e1beab. See the commit message for design rationale.
This commit solves it for ARM virt basically the same way: new machine
properties pflash0, pflash1 forward to the onboard flash devices'
properties. Requires creating the onboard devices in the
.instance_init() method virt_instance_init(). The existing code to
pick up drives defined with -drive if=pflash is replaced by code to
desugar into the machine properties.
There are a few behavioral differences, though:
* The flash devices are always present (x86: only present if
configured)
* Flash base addresses and sizes are fixed (x86: sizes depend on
images, mapped back to back below a fixed address)
* -bios configures contents of first pflash (x86: -bios configures ROM
contents)
* -bios is rejected when first pflash is also configured with -machine
pflash0=... (x86: bios is silently ignored then)
* -machine pflash1=... does not require -machine pflash0=... (x86: it
does).
The actual code is a bit simpler than for x86 mostly due to the first
two differences.
Before the patch, all the action is in create_flash(), called from the
machine's .init() method machvirt_init():
main()
machine_run_board_init()
machvirt_init()
create_flash()
create_one_flash() for flash[0]
create
configure
includes obeying -drive if=pflash,unit=0
realize
map
fall back to -bios
create_one_flash() for flash[1]
create
configure
includes obeying -drive if=pflash,unit=1
realize
map
update FDT
To make the machine properties work, we need to move device creation
to its .instance_init() method virt_instance_init().
Another complication is machvirt_init()'s computation of
@firmware_loaded: it predicts what create_flash() will do. Instead of
predicting what create_flash()'s replacement virt_firmware_init() will
do, I decided to have virt_firmware_init() return what it did.
Requires calling it a bit earlier.
Resulting call tree:
main()
current_machine = object_new()
...
virt_instance_init()
virt_flash_create()
virt_flash_create1() for flash[0]
create
configure: set defaults
become child of machine [NEW]
add machine prop pflash0 as alias for drive [NEW]
virt_flash_create1() for flash[1]
create
configure: set defaults
become child of machine [NEW]
add machine prop pflash1 as alias for drive [NEW]
for all machine props from the command line: machine_set_property()
...
property_set_alias() for machine props pflash0, pflash1
...
set_drive() for cfi.pflash01 prop drive
this is how -machine pflash0=... etc set
machine_run_board_init(current_machine);
virt_firmware_init()
pflash_cfi01_legacy_drive()
legacy -drive if=pflash,unit=0 and =1 [NEW]
virt_flash_map()
virt_flash_map1() for flash[0]
configure: num-blocks
realize
map
virt_flash_map1() for flash[1]
configure: num-blocks
realize
map
fall back to -bios
virt_flash_fdt()
update FDT
You have László to thank for making me explain this in detail.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Acked-by: Laszlo Ersek <lersek@redhat.com>
Message-id: 20190416091348.26075-4-armbru@redhat.com
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
2019-05-07 14:55:02 +03:00
|
|
|
assert(size / VIRT_FLASH_SECTOR_SIZE <= UINT32_MAX);
|
|
|
|
qdev_prop_set_uint32(dev, "num-blocks", size / VIRT_FLASH_SECTOR_SIZE);
|
sysbus: Convert to sysbus_realize() etc. with Coccinelle
Convert from qdev_realize(), qdev_realize_and_unref() with null @bus
argument to sysbus_realize(), sysbus_realize_and_unref().
Coccinelle script:
@@
expression dev, errp;
@@
- qdev_realize(DEVICE(dev), NULL, errp);
+ sysbus_realize(SYS_BUS_DEVICE(dev), errp);
@@
expression sysbus_dev, dev, errp;
@@
+ sysbus_dev = SYS_BUS_DEVICE(dev);
- qdev_realize_and_unref(dev, NULL, errp);
+ sysbus_realize_and_unref(sysbus_dev, errp);
- sysbus_dev = SYS_BUS_DEVICE(dev);
@@
expression sysbus_dev, dev, errp;
expression expr;
@@
sysbus_dev = SYS_BUS_DEVICE(dev);
... when != dev = expr;
- qdev_realize_and_unref(dev, NULL, errp);
+ sysbus_realize_and_unref(sysbus_dev, errp);
@@
expression dev, errp;
@@
- qdev_realize_and_unref(DEVICE(dev), NULL, errp);
+ sysbus_realize_and_unref(SYS_BUS_DEVICE(dev), errp);
@@
expression dev, errp;
@@
- qdev_realize_and_unref(dev, NULL, errp);
+ sysbus_realize_and_unref(SYS_BUS_DEVICE(dev), errp);
Whitespace changes minimized manually.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Acked-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20200610053247.1583243-46-armbru@redhat.com>
[Conflicts in hw/misc/empty_slot.c and hw/sparc/leon3.c resolved]
2020-06-10 08:32:34 +03:00
|
|
|
sysbus_realize_and_unref(SYS_BUS_DEVICE(dev), &error_fatal);
|
hw/arm/virt: Support firmware configuration with -blockdev
The ARM virt machines put firmware in flash memory. To configure it,
you use -drive if=pflash,unit=0,... and optionally -drive
if=pflash,unit=1,...
Why two -drive? This permits setting up one part of the flash memory
read-only, and the other part read/write. It also makes upgrading
firmware on the host easier. Below the hood, we get two separate
flash devices, because we were too lazy to improve our flash device
models to support sector protection.
The problem at hand is to do the same with -blockdev somehow, as one
more step towards deprecating -drive.
We recently solved this problem for x86 PC machines, in commit
ebc29e1beab. See the commit message for design rationale.
This commit solves it for ARM virt basically the same way: new machine
properties pflash0, pflash1 forward to the onboard flash devices'
properties. Requires creating the onboard devices in the
.instance_init() method virt_instance_init(). The existing code to
pick up drives defined with -drive if=pflash is replaced by code to
desugar into the machine properties.
There are a few behavioral differences, though:
* The flash devices are always present (x86: only present if
configured)
* Flash base addresses and sizes are fixed (x86: sizes depend on
images, mapped back to back below a fixed address)
* -bios configures contents of first pflash (x86: -bios configures ROM
contents)
* -bios is rejected when first pflash is also configured with -machine
pflash0=... (x86: bios is silently ignored then)
* -machine pflash1=... does not require -machine pflash0=... (x86: it
does).
The actual code is a bit simpler than for x86 mostly due to the first
two differences.
Before the patch, all the action is in create_flash(), called from the
machine's .init() method machvirt_init():
main()
machine_run_board_init()
machvirt_init()
create_flash()
create_one_flash() for flash[0]
create
configure
includes obeying -drive if=pflash,unit=0
realize
map
fall back to -bios
create_one_flash() for flash[1]
create
configure
includes obeying -drive if=pflash,unit=1
realize
map
update FDT
To make the machine properties work, we need to move device creation
to its .instance_init() method virt_instance_init().
Another complication is machvirt_init()'s computation of
@firmware_loaded: it predicts what create_flash() will do. Instead of
predicting what create_flash()'s replacement virt_firmware_init() will
do, I decided to have virt_firmware_init() return what it did.
Requires calling it a bit earlier.
Resulting call tree:
main()
current_machine = object_new()
...
virt_instance_init()
virt_flash_create()
virt_flash_create1() for flash[0]
create
configure: set defaults
become child of machine [NEW]
add machine prop pflash0 as alias for drive [NEW]
virt_flash_create1() for flash[1]
create
configure: set defaults
become child of machine [NEW]
add machine prop pflash1 as alias for drive [NEW]
for all machine props from the command line: machine_set_property()
...
property_set_alias() for machine props pflash0, pflash1
...
set_drive() for cfi.pflash01 prop drive
this is how -machine pflash0=... etc set
machine_run_board_init(current_machine);
virt_firmware_init()
pflash_cfi01_legacy_drive()
legacy -drive if=pflash,unit=0 and =1 [NEW]
virt_flash_map()
virt_flash_map1() for flash[0]
configure: num-blocks
realize
map
virt_flash_map1() for flash[1]
configure: num-blocks
realize
map
fall back to -bios
virt_flash_fdt()
update FDT
You have László to thank for making me explain this in detail.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Acked-by: Laszlo Ersek <lersek@redhat.com>
Message-id: 20190416091348.26075-4-armbru@redhat.com
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
2019-05-07 14:55:02 +03:00
|
|
|
|
|
|
|
memory_region_add_subregion(sysmem, base,
|
|
|
|
sysbus_mmio_get_region(SYS_BUS_DEVICE(dev),
|
|
|
|
0));
|
2016-03-04 14:30:17 +03:00
|
|
|
}
|
|
|
|
|
hw/arm/virt: Support firmware configuration with -blockdev
The ARM virt machines put firmware in flash memory. To configure it,
you use -drive if=pflash,unit=0,... and optionally -drive
if=pflash,unit=1,...
Why two -drive? This permits setting up one part of the flash memory
read-only, and the other part read/write. It also makes upgrading
firmware on the host easier. Below the hood, we get two separate
flash devices, because we were too lazy to improve our flash device
models to support sector protection.
The problem at hand is to do the same with -blockdev somehow, as one
more step towards deprecating -drive.
We recently solved this problem for x86 PC machines, in commit
ebc29e1beab. See the commit message for design rationale.
This commit solves it for ARM virt basically the same way: new machine
properties pflash0, pflash1 forward to the onboard flash devices'
properties. Requires creating the onboard devices in the
.instance_init() method virt_instance_init(). The existing code to
pick up drives defined with -drive if=pflash is replaced by code to
desugar into the machine properties.
There are a few behavioral differences, though:
* The flash devices are always present (x86: only present if
configured)
* Flash base addresses and sizes are fixed (x86: sizes depend on
images, mapped back to back below a fixed address)
* -bios configures contents of first pflash (x86: -bios configures ROM
contents)
* -bios is rejected when first pflash is also configured with -machine
pflash0=... (x86: bios is silently ignored then)
* -machine pflash1=... does not require -machine pflash0=... (x86: it
does).
The actual code is a bit simpler than for x86 mostly due to the first
two differences.
Before the patch, all the action is in create_flash(), called from the
machine's .init() method machvirt_init():
main()
machine_run_board_init()
machvirt_init()
create_flash()
create_one_flash() for flash[0]
create
configure
includes obeying -drive if=pflash,unit=0
realize
map
fall back to -bios
create_one_flash() for flash[1]
create
configure
includes obeying -drive if=pflash,unit=1
realize
map
update FDT
To make the machine properties work, we need to move device creation
to its .instance_init() method virt_instance_init().
Another complication is machvirt_init()'s computation of
@firmware_loaded: it predicts what create_flash() will do. Instead of
predicting what create_flash()'s replacement virt_firmware_init() will
do, I decided to have virt_firmware_init() return what it did.
Requires calling it a bit earlier.
Resulting call tree:
main()
current_machine = object_new()
...
virt_instance_init()
virt_flash_create()
virt_flash_create1() for flash[0]
create
configure: set defaults
become child of machine [NEW]
add machine prop pflash0 as alias for drive [NEW]
virt_flash_create1() for flash[1]
create
configure: set defaults
become child of machine [NEW]
add machine prop pflash1 as alias for drive [NEW]
for all machine props from the command line: machine_set_property()
...
property_set_alias() for machine props pflash0, pflash1
...
set_drive() for cfi.pflash01 prop drive
this is how -machine pflash0=... etc set
machine_run_board_init(current_machine);
virt_firmware_init()
pflash_cfi01_legacy_drive()
legacy -drive if=pflash,unit=0 and =1 [NEW]
virt_flash_map()
virt_flash_map1() for flash[0]
configure: num-blocks
realize
map
virt_flash_map1() for flash[1]
configure: num-blocks
realize
map
fall back to -bios
virt_flash_fdt()
update FDT
You have László to thank for making me explain this in detail.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Acked-by: Laszlo Ersek <lersek@redhat.com>
Message-id: 20190416091348.26075-4-armbru@redhat.com
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
2019-05-07 14:55:02 +03:00
|
|
|
static void virt_flash_map(VirtMachineState *vms,
|
|
|
|
MemoryRegion *sysmem,
|
|
|
|
MemoryRegion *secure_sysmem)
|
2016-03-04 14:30:17 +03:00
|
|
|
{
|
hw/arm/virt: Support firmware configuration with -blockdev
The ARM virt machines put firmware in flash memory. To configure it,
you use -drive if=pflash,unit=0,... and optionally -drive
if=pflash,unit=1,...
Why two -drive? This permits setting up one part of the flash memory
read-only, and the other part read/write. It also makes upgrading
firmware on the host easier. Below the hood, we get two separate
flash devices, because we were too lazy to improve our flash device
models to support sector protection.
The problem at hand is to do the same with -blockdev somehow, as one
more step towards deprecating -drive.
We recently solved this problem for x86 PC machines, in commit
ebc29e1beab. See the commit message for design rationale.
This commit solves it for ARM virt basically the same way: new machine
properties pflash0, pflash1 forward to the onboard flash devices'
properties. Requires creating the onboard devices in the
.instance_init() method virt_instance_init(). The existing code to
pick up drives defined with -drive if=pflash is replaced by code to
desugar into the machine properties.
There are a few behavioral differences, though:
* The flash devices are always present (x86: only present if
configured)
* Flash base addresses and sizes are fixed (x86: sizes depend on
images, mapped back to back below a fixed address)
* -bios configures contents of first pflash (x86: -bios configures ROM
contents)
* -bios is rejected when first pflash is also configured with -machine
pflash0=... (x86: bios is silently ignored then)
* -machine pflash1=... does not require -machine pflash0=... (x86: it
does).
The actual code is a bit simpler than for x86 mostly due to the first
two differences.
Before the patch, all the action is in create_flash(), called from the
machine's .init() method machvirt_init():
main()
machine_run_board_init()
machvirt_init()
create_flash()
create_one_flash() for flash[0]
create
configure
includes obeying -drive if=pflash,unit=0
realize
map
fall back to -bios
create_one_flash() for flash[1]
create
configure
includes obeying -drive if=pflash,unit=1
realize
map
update FDT
To make the machine properties work, we need to move device creation
to its .instance_init() method virt_instance_init().
Another complication is machvirt_init()'s computation of
@firmware_loaded: it predicts what create_flash() will do. Instead of
predicting what create_flash()'s replacement virt_firmware_init() will
do, I decided to have virt_firmware_init() return what it did.
Requires calling it a bit earlier.
Resulting call tree:
main()
current_machine = object_new()
...
virt_instance_init()
virt_flash_create()
virt_flash_create1() for flash[0]
create
configure: set defaults
become child of machine [NEW]
add machine prop pflash0 as alias for drive [NEW]
virt_flash_create1() for flash[1]
create
configure: set defaults
become child of machine [NEW]
add machine prop pflash1 as alias for drive [NEW]
for all machine props from the command line: machine_set_property()
...
property_set_alias() for machine props pflash0, pflash1
...
set_drive() for cfi.pflash01 prop drive
this is how -machine pflash0=... etc set
machine_run_board_init(current_machine);
virt_firmware_init()
pflash_cfi01_legacy_drive()
legacy -drive if=pflash,unit=0 and =1 [NEW]
virt_flash_map()
virt_flash_map1() for flash[0]
configure: num-blocks
realize
map
virt_flash_map1() for flash[1]
configure: num-blocks
realize
map
fall back to -bios
virt_flash_fdt()
update FDT
You have László to thank for making me explain this in detail.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Acked-by: Laszlo Ersek <lersek@redhat.com>
Message-id: 20190416091348.26075-4-armbru@redhat.com
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
2019-05-07 14:55:02 +03:00
|
|
|
/*
|
|
|
|
* Map two flash devices to fill the VIRT_FLASH space in the memmap.
|
2016-03-04 14:30:18 +03:00
|
|
|
* sysmem is the system memory space. secure_sysmem is the secure view
|
|
|
|
* of the system, and the first flash device should be made visible only
|
|
|
|
* there. The second flash device is visible to both secure and nonsecure.
|
|
|
|
* If sysmem == secure_sysmem this means there is no separate Secure
|
|
|
|
* address space and both flash devices are generally visible.
|
2016-03-04 14:30:17 +03:00
|
|
|
*/
|
2017-01-09 14:40:21 +03:00
|
|
|
hwaddr flashsize = vms->memmap[VIRT_FLASH].size / 2;
|
|
|
|
hwaddr flashbase = vms->memmap[VIRT_FLASH].base;
|
2014-09-12 17:06:48 +04:00
|
|
|
|
hw/arm/virt: Support firmware configuration with -blockdev
The ARM virt machines put firmware in flash memory. To configure it,
you use -drive if=pflash,unit=0,... and optionally -drive
if=pflash,unit=1,...
Why two -drive? This permits setting up one part of the flash memory
read-only, and the other part read/write. It also makes upgrading
firmware on the host easier. Below the hood, we get two separate
flash devices, because we were too lazy to improve our flash device
models to support sector protection.
The problem at hand is to do the same with -blockdev somehow, as one
more step towards deprecating -drive.
We recently solved this problem for x86 PC machines, in commit
ebc29e1beab. See the commit message for design rationale.
This commit solves it for ARM virt basically the same way: new machine
properties pflash0, pflash1 forward to the onboard flash devices'
properties. Requires creating the onboard devices in the
.instance_init() method virt_instance_init(). The existing code to
pick up drives defined with -drive if=pflash is replaced by code to
desugar into the machine properties.
There are a few behavioral differences, though:
* The flash devices are always present (x86: only present if
configured)
* Flash base addresses and sizes are fixed (x86: sizes depend on
images, mapped back to back below a fixed address)
* -bios configures contents of first pflash (x86: -bios configures ROM
contents)
* -bios is rejected when first pflash is also configured with -machine
pflash0=... (x86: bios is silently ignored then)
* -machine pflash1=... does not require -machine pflash0=... (x86: it
does).
The actual code is a bit simpler than for x86 mostly due to the first
two differences.
Before the patch, all the action is in create_flash(), called from the
machine's .init() method machvirt_init():
main()
machine_run_board_init()
machvirt_init()
create_flash()
create_one_flash() for flash[0]
create
configure
includes obeying -drive if=pflash,unit=0
realize
map
fall back to -bios
create_one_flash() for flash[1]
create
configure
includes obeying -drive if=pflash,unit=1
realize
map
update FDT
To make the machine properties work, we need to move device creation
to its .instance_init() method virt_instance_init().
Another complication is machvirt_init()'s computation of
@firmware_loaded: it predicts what create_flash() will do. Instead of
predicting what create_flash()'s replacement virt_firmware_init() will
do, I decided to have virt_firmware_init() return what it did.
Requires calling it a bit earlier.
Resulting call tree:
main()
current_machine = object_new()
...
virt_instance_init()
virt_flash_create()
virt_flash_create1() for flash[0]
create
configure: set defaults
become child of machine [NEW]
add machine prop pflash0 as alias for drive [NEW]
virt_flash_create1() for flash[1]
create
configure: set defaults
become child of machine [NEW]
add machine prop pflash1 as alias for drive [NEW]
for all machine props from the command line: machine_set_property()
...
property_set_alias() for machine props pflash0, pflash1
...
set_drive() for cfi.pflash01 prop drive
this is how -machine pflash0=... etc set
machine_run_board_init(current_machine);
virt_firmware_init()
pflash_cfi01_legacy_drive()
legacy -drive if=pflash,unit=0 and =1 [NEW]
virt_flash_map()
virt_flash_map1() for flash[0]
configure: num-blocks
realize
map
virt_flash_map1() for flash[1]
configure: num-blocks
realize
map
fall back to -bios
virt_flash_fdt()
update FDT
You have László to thank for making me explain this in detail.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Acked-by: Laszlo Ersek <lersek@redhat.com>
Message-id: 20190416091348.26075-4-armbru@redhat.com
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
2019-05-07 14:55:02 +03:00
|
|
|
virt_flash_map1(vms->flash[0], flashbase, flashsize,
|
|
|
|
secure_sysmem);
|
|
|
|
virt_flash_map1(vms->flash[1], flashbase + flashsize, flashsize,
|
|
|
|
sysmem);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void virt_flash_fdt(VirtMachineState *vms,
|
|
|
|
MemoryRegion *sysmem,
|
|
|
|
MemoryRegion *secure_sysmem)
|
|
|
|
{
|
|
|
|
hwaddr flashsize = vms->memmap[VIRT_FLASH].size / 2;
|
|
|
|
hwaddr flashbase = vms->memmap[VIRT_FLASH].base;
|
2021-03-03 20:36:36 +03:00
|
|
|
MachineState *ms = MACHINE(vms);
|
hw/arm/virt: Support firmware configuration with -blockdev
The ARM virt machines put firmware in flash memory. To configure it,
you use -drive if=pflash,unit=0,... and optionally -drive
if=pflash,unit=1,...
Why two -drive? This permits setting up one part of the flash memory
read-only, and the other part read/write. It also makes upgrading
firmware on the host easier. Below the hood, we get two separate
flash devices, because we were too lazy to improve our flash device
models to support sector protection.
The problem at hand is to do the same with -blockdev somehow, as one
more step towards deprecating -drive.
We recently solved this problem for x86 PC machines, in commit
ebc29e1beab. See the commit message for design rationale.
This commit solves it for ARM virt basically the same way: new machine
properties pflash0, pflash1 forward to the onboard flash devices'
properties. Requires creating the onboard devices in the
.instance_init() method virt_instance_init(). The existing code to
pick up drives defined with -drive if=pflash is replaced by code to
desugar into the machine properties.
There are a few behavioral differences, though:
* The flash devices are always present (x86: only present if
configured)
* Flash base addresses and sizes are fixed (x86: sizes depend on
images, mapped back to back below a fixed address)
* -bios configures contents of first pflash (x86: -bios configures ROM
contents)
* -bios is rejected when first pflash is also configured with -machine
pflash0=... (x86: bios is silently ignored then)
* -machine pflash1=... does not require -machine pflash0=... (x86: it
does).
The actual code is a bit simpler than for x86 mostly due to the first
two differences.
Before the patch, all the action is in create_flash(), called from the
machine's .init() method machvirt_init():
main()
machine_run_board_init()
machvirt_init()
create_flash()
create_one_flash() for flash[0]
create
configure
includes obeying -drive if=pflash,unit=0
realize
map
fall back to -bios
create_one_flash() for flash[1]
create
configure
includes obeying -drive if=pflash,unit=1
realize
map
update FDT
To make the machine properties work, we need to move device creation
to its .instance_init() method virt_instance_init().
Another complication is machvirt_init()'s computation of
@firmware_loaded: it predicts what create_flash() will do. Instead of
predicting what create_flash()'s replacement virt_firmware_init() will
do, I decided to have virt_firmware_init() return what it did.
Requires calling it a bit earlier.
Resulting call tree:
main()
current_machine = object_new()
...
virt_instance_init()
virt_flash_create()
virt_flash_create1() for flash[0]
create
configure: set defaults
become child of machine [NEW]
add machine prop pflash0 as alias for drive [NEW]
virt_flash_create1() for flash[1]
create
configure: set defaults
become child of machine [NEW]
add machine prop pflash1 as alias for drive [NEW]
for all machine props from the command line: machine_set_property()
...
property_set_alias() for machine props pflash0, pflash1
...
set_drive() for cfi.pflash01 prop drive
this is how -machine pflash0=... etc set
machine_run_board_init(current_machine);
virt_firmware_init()
pflash_cfi01_legacy_drive()
legacy -drive if=pflash,unit=0 and =1 [NEW]
virt_flash_map()
virt_flash_map1() for flash[0]
configure: num-blocks
realize
map
virt_flash_map1() for flash[1]
configure: num-blocks
realize
map
fall back to -bios
virt_flash_fdt()
update FDT
You have László to thank for making me explain this in detail.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Acked-by: Laszlo Ersek <lersek@redhat.com>
Message-id: 20190416091348.26075-4-armbru@redhat.com
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
2019-05-07 14:55:02 +03:00
|
|
|
char *nodename;
|
2014-09-12 17:06:48 +04:00
|
|
|
|
2016-03-04 14:30:18 +03:00
|
|
|
if (sysmem == secure_sysmem) {
|
|
|
|
/* Report both flash devices as a single node in the DT */
|
|
|
|
nodename = g_strdup_printf("/flash@%" PRIx64, flashbase);
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_add_subnode(ms->fdt, nodename);
|
|
|
|
qemu_fdt_setprop_string(ms->fdt, nodename, "compatible", "cfi-flash");
|
|
|
|
qemu_fdt_setprop_sized_cells(ms->fdt, nodename, "reg",
|
2016-03-04 14:30:18 +03:00
|
|
|
2, flashbase, 2, flashsize,
|
|
|
|
2, flashbase + flashsize, 2, flashsize);
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_cell(ms->fdt, nodename, "bank-width", 4);
|
2016-03-04 14:30:18 +03:00
|
|
|
g_free(nodename);
|
|
|
|
} else {
|
hw/arm/virt: Support firmware configuration with -blockdev
The ARM virt machines put firmware in flash memory. To configure it,
you use -drive if=pflash,unit=0,... and optionally -drive
if=pflash,unit=1,...
Why two -drive? This permits setting up one part of the flash memory
read-only, and the other part read/write. It also makes upgrading
firmware on the host easier. Below the hood, we get two separate
flash devices, because we were too lazy to improve our flash device
models to support sector protection.
The problem at hand is to do the same with -blockdev somehow, as one
more step towards deprecating -drive.
We recently solved this problem for x86 PC machines, in commit
ebc29e1beab. See the commit message for design rationale.
This commit solves it for ARM virt basically the same way: new machine
properties pflash0, pflash1 forward to the onboard flash devices'
properties. Requires creating the onboard devices in the
.instance_init() method virt_instance_init(). The existing code to
pick up drives defined with -drive if=pflash is replaced by code to
desugar into the machine properties.
There are a few behavioral differences, though:
* The flash devices are always present (x86: only present if
configured)
* Flash base addresses and sizes are fixed (x86: sizes depend on
images, mapped back to back below a fixed address)
* -bios configures contents of first pflash (x86: -bios configures ROM
contents)
* -bios is rejected when first pflash is also configured with -machine
pflash0=... (x86: bios is silently ignored then)
* -machine pflash1=... does not require -machine pflash0=... (x86: it
does).
The actual code is a bit simpler than for x86 mostly due to the first
two differences.
Before the patch, all the action is in create_flash(), called from the
machine's .init() method machvirt_init():
main()
machine_run_board_init()
machvirt_init()
create_flash()
create_one_flash() for flash[0]
create
configure
includes obeying -drive if=pflash,unit=0
realize
map
fall back to -bios
create_one_flash() for flash[1]
create
configure
includes obeying -drive if=pflash,unit=1
realize
map
update FDT
To make the machine properties work, we need to move device creation
to its .instance_init() method virt_instance_init().
Another complication is machvirt_init()'s computation of
@firmware_loaded: it predicts what create_flash() will do. Instead of
predicting what create_flash()'s replacement virt_firmware_init() will
do, I decided to have virt_firmware_init() return what it did.
Requires calling it a bit earlier.
Resulting call tree:
main()
current_machine = object_new()
...
virt_instance_init()
virt_flash_create()
virt_flash_create1() for flash[0]
create
configure: set defaults
become child of machine [NEW]
add machine prop pflash0 as alias for drive [NEW]
virt_flash_create1() for flash[1]
create
configure: set defaults
become child of machine [NEW]
add machine prop pflash1 as alias for drive [NEW]
for all machine props from the command line: machine_set_property()
...
property_set_alias() for machine props pflash0, pflash1
...
set_drive() for cfi.pflash01 prop drive
this is how -machine pflash0=... etc set
machine_run_board_init(current_machine);
virt_firmware_init()
pflash_cfi01_legacy_drive()
legacy -drive if=pflash,unit=0 and =1 [NEW]
virt_flash_map()
virt_flash_map1() for flash[0]
configure: num-blocks
realize
map
virt_flash_map1() for flash[1]
configure: num-blocks
realize
map
fall back to -bios
virt_flash_fdt()
update FDT
You have László to thank for making me explain this in detail.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Acked-by: Laszlo Ersek <lersek@redhat.com>
Message-id: 20190416091348.26075-4-armbru@redhat.com
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
2019-05-07 14:55:02 +03:00
|
|
|
/*
|
|
|
|
* Report the devices as separate nodes so we can mark one as
|
2016-03-04 14:30:18 +03:00
|
|
|
* only visible to the secure world.
|
|
|
|
*/
|
|
|
|
nodename = g_strdup_printf("/secflash@%" PRIx64, flashbase);
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_add_subnode(ms->fdt, nodename);
|
|
|
|
qemu_fdt_setprop_string(ms->fdt, nodename, "compatible", "cfi-flash");
|
|
|
|
qemu_fdt_setprop_sized_cells(ms->fdt, nodename, "reg",
|
2016-03-04 14:30:18 +03:00
|
|
|
2, flashbase, 2, flashsize);
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_cell(ms->fdt, nodename, "bank-width", 4);
|
|
|
|
qemu_fdt_setprop_string(ms->fdt, nodename, "status", "disabled");
|
|
|
|
qemu_fdt_setprop_string(ms->fdt, nodename, "secure-status", "okay");
|
2016-03-04 14:30:18 +03:00
|
|
|
g_free(nodename);
|
|
|
|
|
|
|
|
nodename = g_strdup_printf("/flash@%" PRIx64, flashbase);
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_add_subnode(ms->fdt, nodename);
|
|
|
|
qemu_fdt_setprop_string(ms->fdt, nodename, "compatible", "cfi-flash");
|
|
|
|
qemu_fdt_setprop_sized_cells(ms->fdt, nodename, "reg",
|
2016-03-04 14:30:18 +03:00
|
|
|
2, flashbase + flashsize, 2, flashsize);
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_cell(ms->fdt, nodename, "bank-width", 4);
|
2016-03-04 14:30:18 +03:00
|
|
|
g_free(nodename);
|
|
|
|
}
|
2014-09-12 17:06:48 +04:00
|
|
|
}
|
|
|
|
|
hw/arm/virt: Support firmware configuration with -blockdev
The ARM virt machines put firmware in flash memory. To configure it,
you use -drive if=pflash,unit=0,... and optionally -drive
if=pflash,unit=1,...
Why two -drive? This permits setting up one part of the flash memory
read-only, and the other part read/write. It also makes upgrading
firmware on the host easier. Below the hood, we get two separate
flash devices, because we were too lazy to improve our flash device
models to support sector protection.
The problem at hand is to do the same with -blockdev somehow, as one
more step towards deprecating -drive.
We recently solved this problem for x86 PC machines, in commit
ebc29e1beab. See the commit message for design rationale.
This commit solves it for ARM virt basically the same way: new machine
properties pflash0, pflash1 forward to the onboard flash devices'
properties. Requires creating the onboard devices in the
.instance_init() method virt_instance_init(). The existing code to
pick up drives defined with -drive if=pflash is replaced by code to
desugar into the machine properties.
There are a few behavioral differences, though:
* The flash devices are always present (x86: only present if
configured)
* Flash base addresses and sizes are fixed (x86: sizes depend on
images, mapped back to back below a fixed address)
* -bios configures contents of first pflash (x86: -bios configures ROM
contents)
* -bios is rejected when first pflash is also configured with -machine
pflash0=... (x86: bios is silently ignored then)
* -machine pflash1=... does not require -machine pflash0=... (x86: it
does).
The actual code is a bit simpler than for x86 mostly due to the first
two differences.
Before the patch, all the action is in create_flash(), called from the
machine's .init() method machvirt_init():
main()
machine_run_board_init()
machvirt_init()
create_flash()
create_one_flash() for flash[0]
create
configure
includes obeying -drive if=pflash,unit=0
realize
map
fall back to -bios
create_one_flash() for flash[1]
create
configure
includes obeying -drive if=pflash,unit=1
realize
map
update FDT
To make the machine properties work, we need to move device creation
to its .instance_init() method virt_instance_init().
Another complication is machvirt_init()'s computation of
@firmware_loaded: it predicts what create_flash() will do. Instead of
predicting what create_flash()'s replacement virt_firmware_init() will
do, I decided to have virt_firmware_init() return what it did.
Requires calling it a bit earlier.
Resulting call tree:
main()
current_machine = object_new()
...
virt_instance_init()
virt_flash_create()
virt_flash_create1() for flash[0]
create
configure: set defaults
become child of machine [NEW]
add machine prop pflash0 as alias for drive [NEW]
virt_flash_create1() for flash[1]
create
configure: set defaults
become child of machine [NEW]
add machine prop pflash1 as alias for drive [NEW]
for all machine props from the command line: machine_set_property()
...
property_set_alias() for machine props pflash0, pflash1
...
set_drive() for cfi.pflash01 prop drive
this is how -machine pflash0=... etc set
machine_run_board_init(current_machine);
virt_firmware_init()
pflash_cfi01_legacy_drive()
legacy -drive if=pflash,unit=0 and =1 [NEW]
virt_flash_map()
virt_flash_map1() for flash[0]
configure: num-blocks
realize
map
virt_flash_map1() for flash[1]
configure: num-blocks
realize
map
fall back to -bios
virt_flash_fdt()
update FDT
You have László to thank for making me explain this in detail.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Acked-by: Laszlo Ersek <lersek@redhat.com>
Message-id: 20190416091348.26075-4-armbru@redhat.com
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
2019-05-07 14:55:02 +03:00
|
|
|
static bool virt_firmware_init(VirtMachineState *vms,
|
|
|
|
MemoryRegion *sysmem,
|
|
|
|
MemoryRegion *secure_sysmem)
|
|
|
|
{
|
|
|
|
int i;
|
2020-10-26 17:30:16 +03:00
|
|
|
const char *bios_name;
|
hw/arm/virt: Support firmware configuration with -blockdev
The ARM virt machines put firmware in flash memory. To configure it,
you use -drive if=pflash,unit=0,... and optionally -drive
if=pflash,unit=1,...
Why two -drive? This permits setting up one part of the flash memory
read-only, and the other part read/write. It also makes upgrading
firmware on the host easier. Below the hood, we get two separate
flash devices, because we were too lazy to improve our flash device
models to support sector protection.
The problem at hand is to do the same with -blockdev somehow, as one
more step towards deprecating -drive.
We recently solved this problem for x86 PC machines, in commit
ebc29e1beab. See the commit message for design rationale.
This commit solves it for ARM virt basically the same way: new machine
properties pflash0, pflash1 forward to the onboard flash devices'
properties. Requires creating the onboard devices in the
.instance_init() method virt_instance_init(). The existing code to
pick up drives defined with -drive if=pflash is replaced by code to
desugar into the machine properties.
There are a few behavioral differences, though:
* The flash devices are always present (x86: only present if
configured)
* Flash base addresses and sizes are fixed (x86: sizes depend on
images, mapped back to back below a fixed address)
* -bios configures contents of first pflash (x86: -bios configures ROM
contents)
* -bios is rejected when first pflash is also configured with -machine
pflash0=... (x86: bios is silently ignored then)
* -machine pflash1=... does not require -machine pflash0=... (x86: it
does).
The actual code is a bit simpler than for x86 mostly due to the first
two differences.
Before the patch, all the action is in create_flash(), called from the
machine's .init() method machvirt_init():
main()
machine_run_board_init()
machvirt_init()
create_flash()
create_one_flash() for flash[0]
create
configure
includes obeying -drive if=pflash,unit=0
realize
map
fall back to -bios
create_one_flash() for flash[1]
create
configure
includes obeying -drive if=pflash,unit=1
realize
map
update FDT
To make the machine properties work, we need to move device creation
to its .instance_init() method virt_instance_init().
Another complication is machvirt_init()'s computation of
@firmware_loaded: it predicts what create_flash() will do. Instead of
predicting what create_flash()'s replacement virt_firmware_init() will
do, I decided to have virt_firmware_init() return what it did.
Requires calling it a bit earlier.
Resulting call tree:
main()
current_machine = object_new()
...
virt_instance_init()
virt_flash_create()
virt_flash_create1() for flash[0]
create
configure: set defaults
become child of machine [NEW]
add machine prop pflash0 as alias for drive [NEW]
virt_flash_create1() for flash[1]
create
configure: set defaults
become child of machine [NEW]
add machine prop pflash1 as alias for drive [NEW]
for all machine props from the command line: machine_set_property()
...
property_set_alias() for machine props pflash0, pflash1
...
set_drive() for cfi.pflash01 prop drive
this is how -machine pflash0=... etc set
machine_run_board_init(current_machine);
virt_firmware_init()
pflash_cfi01_legacy_drive()
legacy -drive if=pflash,unit=0 and =1 [NEW]
virt_flash_map()
virt_flash_map1() for flash[0]
configure: num-blocks
realize
map
virt_flash_map1() for flash[1]
configure: num-blocks
realize
map
fall back to -bios
virt_flash_fdt()
update FDT
You have László to thank for making me explain this in detail.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Acked-by: Laszlo Ersek <lersek@redhat.com>
Message-id: 20190416091348.26075-4-armbru@redhat.com
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
2019-05-07 14:55:02 +03:00
|
|
|
BlockBackend *pflash_blk0;
|
|
|
|
|
|
|
|
/* Map legacy -drive if=pflash to machine properties */
|
|
|
|
for (i = 0; i < ARRAY_SIZE(vms->flash); i++) {
|
|
|
|
pflash_cfi01_legacy_drive(vms->flash[i],
|
|
|
|
drive_get(IF_PFLASH, 0, i));
|
|
|
|
}
|
|
|
|
|
|
|
|
virt_flash_map(vms, sysmem, secure_sysmem);
|
|
|
|
|
|
|
|
pflash_blk0 = pflash_cfi01_get_blk(vms->flash[0]);
|
|
|
|
|
2020-10-26 17:30:16 +03:00
|
|
|
bios_name = MACHINE(vms)->firmware;
|
hw/arm/virt: Support firmware configuration with -blockdev
The ARM virt machines put firmware in flash memory. To configure it,
you use -drive if=pflash,unit=0,... and optionally -drive
if=pflash,unit=1,...
Why two -drive? This permits setting up one part of the flash memory
read-only, and the other part read/write. It also makes upgrading
firmware on the host easier. Below the hood, we get two separate
flash devices, because we were too lazy to improve our flash device
models to support sector protection.
The problem at hand is to do the same with -blockdev somehow, as one
more step towards deprecating -drive.
We recently solved this problem for x86 PC machines, in commit
ebc29e1beab. See the commit message for design rationale.
This commit solves it for ARM virt basically the same way: new machine
properties pflash0, pflash1 forward to the onboard flash devices'
properties. Requires creating the onboard devices in the
.instance_init() method virt_instance_init(). The existing code to
pick up drives defined with -drive if=pflash is replaced by code to
desugar into the machine properties.
There are a few behavioral differences, though:
* The flash devices are always present (x86: only present if
configured)
* Flash base addresses and sizes are fixed (x86: sizes depend on
images, mapped back to back below a fixed address)
* -bios configures contents of first pflash (x86: -bios configures ROM
contents)
* -bios is rejected when first pflash is also configured with -machine
pflash0=... (x86: bios is silently ignored then)
* -machine pflash1=... does not require -machine pflash0=... (x86: it
does).
The actual code is a bit simpler than for x86 mostly due to the first
two differences.
Before the patch, all the action is in create_flash(), called from the
machine's .init() method machvirt_init():
main()
machine_run_board_init()
machvirt_init()
create_flash()
create_one_flash() for flash[0]
create
configure
includes obeying -drive if=pflash,unit=0
realize
map
fall back to -bios
create_one_flash() for flash[1]
create
configure
includes obeying -drive if=pflash,unit=1
realize
map
update FDT
To make the machine properties work, we need to move device creation
to its .instance_init() method virt_instance_init().
Another complication is machvirt_init()'s computation of
@firmware_loaded: it predicts what create_flash() will do. Instead of
predicting what create_flash()'s replacement virt_firmware_init() will
do, I decided to have virt_firmware_init() return what it did.
Requires calling it a bit earlier.
Resulting call tree:
main()
current_machine = object_new()
...
virt_instance_init()
virt_flash_create()
virt_flash_create1() for flash[0]
create
configure: set defaults
become child of machine [NEW]
add machine prop pflash0 as alias for drive [NEW]
virt_flash_create1() for flash[1]
create
configure: set defaults
become child of machine [NEW]
add machine prop pflash1 as alias for drive [NEW]
for all machine props from the command line: machine_set_property()
...
property_set_alias() for machine props pflash0, pflash1
...
set_drive() for cfi.pflash01 prop drive
this is how -machine pflash0=... etc set
machine_run_board_init(current_machine);
virt_firmware_init()
pflash_cfi01_legacy_drive()
legacy -drive if=pflash,unit=0 and =1 [NEW]
virt_flash_map()
virt_flash_map1() for flash[0]
configure: num-blocks
realize
map
virt_flash_map1() for flash[1]
configure: num-blocks
realize
map
fall back to -bios
virt_flash_fdt()
update FDT
You have László to thank for making me explain this in detail.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Acked-by: Laszlo Ersek <lersek@redhat.com>
Message-id: 20190416091348.26075-4-armbru@redhat.com
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
2019-05-07 14:55:02 +03:00
|
|
|
if (bios_name) {
|
|
|
|
char *fname;
|
|
|
|
MemoryRegion *mr;
|
|
|
|
int image_size;
|
|
|
|
|
|
|
|
if (pflash_blk0) {
|
|
|
|
error_report("The contents of the first flash device may be "
|
|
|
|
"specified with -bios or with -drive if=pflash... "
|
|
|
|
"but you cannot use both options at once");
|
|
|
|
exit(1);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Fall back to -bios */
|
|
|
|
|
|
|
|
fname = qemu_find_file(QEMU_FILE_TYPE_BIOS, bios_name);
|
|
|
|
if (!fname) {
|
|
|
|
error_report("Could not find ROM image '%s'", bios_name);
|
|
|
|
exit(1);
|
|
|
|
}
|
|
|
|
mr = sysbus_mmio_get_region(SYS_BUS_DEVICE(vms->flash[0]), 0);
|
|
|
|
image_size = load_image_mr(fname, mr);
|
|
|
|
g_free(fname);
|
|
|
|
if (image_size < 0) {
|
|
|
|
error_report("Could not load ROM image '%s'", bios_name);
|
|
|
|
exit(1);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
return pflash_blk0 || bios_name;
|
|
|
|
}
|
|
|
|
|
2017-01-09 14:40:23 +03:00
|
|
|
static FWCfgState *create_fw_cfg(const VirtMachineState *vms, AddressSpace *as)
|
arm: add fw_cfg to "virt" board
fw_cfg already supports exposure over MMIO (used in ppc/mac_newworld.c,
ppc/mac_oldworld.c, sparc/sun4m.c); we can easily add it to the "virt"
board.
Because MMIO access is slow on ARM KVM, we enable the guest, with
fw_cfg_init_mem_wide(), to transfer up to 8 bytes with a single access.
This has been measured to speed up transfers up to 7.5-fold, relative to
single byte data access, on both ARM KVM and x86_64 TCG.
The MMIO register block of fw_cfg is advertized in the device tree. As
base address we pick 0x09020000, which conforms to the comment preceding
"a15memmap": it falls in the miscellaneous device I/O range 128MB..256MB,
and it is aligned at 64KB. The DTB properties follow the documentation in
the Linux source file "Documentation/devicetree/bindings/arm/fw-cfg.txt".
fw_cfg automatically exports a number of files to the guest; for example,
"bootorder" (see fw_cfg_machine_reset()).
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-id: 1419250305-31062-9-git-send-email-pbonzini@redhat.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
2014-12-22 15:11:42 +03:00
|
|
|
{
|
2019-05-18 23:54:26 +03:00
|
|
|
MachineState *ms = MACHINE(vms);
|
2017-01-09 14:40:21 +03:00
|
|
|
hwaddr base = vms->memmap[VIRT_FW_CFG].base;
|
|
|
|
hwaddr size = vms->memmap[VIRT_FW_CFG].size;
|
2016-11-15 15:17:15 +03:00
|
|
|
FWCfgState *fw_cfg;
|
arm: add fw_cfg to "virt" board
fw_cfg already supports exposure over MMIO (used in ppc/mac_newworld.c,
ppc/mac_oldworld.c, sparc/sun4m.c); we can easily add it to the "virt"
board.
Because MMIO access is slow on ARM KVM, we enable the guest, with
fw_cfg_init_mem_wide(), to transfer up to 8 bytes with a single access.
This has been measured to speed up transfers up to 7.5-fold, relative to
single byte data access, on both ARM KVM and x86_64 TCG.
The MMIO register block of fw_cfg is advertized in the device tree. As
base address we pick 0x09020000, which conforms to the comment preceding
"a15memmap": it falls in the miscellaneous device I/O range 128MB..256MB,
and it is aligned at 64KB. The DTB properties follow the documentation in
the Linux source file "Documentation/devicetree/bindings/arm/fw-cfg.txt".
fw_cfg automatically exports a number of files to the guest; for example,
"bootorder" (see fw_cfg_machine_reset()).
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-id: 1419250305-31062-9-git-send-email-pbonzini@redhat.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
2014-12-22 15:11:42 +03:00
|
|
|
char *nodename;
|
|
|
|
|
2016-11-15 15:17:15 +03:00
|
|
|
fw_cfg = fw_cfg_init_mem_wide(base + 8, base, 8, base + 16, as);
|
2019-05-18 23:54:26 +03:00
|
|
|
fw_cfg_add_i16(fw_cfg, FW_CFG_NB_CPUS, (uint16_t)ms->smp.cpus);
|
arm: add fw_cfg to "virt" board
fw_cfg already supports exposure over MMIO (used in ppc/mac_newworld.c,
ppc/mac_oldworld.c, sparc/sun4m.c); we can easily add it to the "virt"
board.
Because MMIO access is slow on ARM KVM, we enable the guest, with
fw_cfg_init_mem_wide(), to transfer up to 8 bytes with a single access.
This has been measured to speed up transfers up to 7.5-fold, relative to
single byte data access, on both ARM KVM and x86_64 TCG.
The MMIO register block of fw_cfg is advertized in the device tree. As
base address we pick 0x09020000, which conforms to the comment preceding
"a15memmap": it falls in the miscellaneous device I/O range 128MB..256MB,
and it is aligned at 64KB. The DTB properties follow the documentation in
the Linux source file "Documentation/devicetree/bindings/arm/fw-cfg.txt".
fw_cfg automatically exports a number of files to the guest; for example,
"bootorder" (see fw_cfg_machine_reset()).
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-id: 1419250305-31062-9-git-send-email-pbonzini@redhat.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
2014-12-22 15:11:42 +03:00
|
|
|
|
|
|
|
nodename = g_strdup_printf("/fw-cfg@%" PRIx64, base);
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_add_subnode(ms->fdt, nodename);
|
|
|
|
qemu_fdt_setprop_string(ms->fdt, nodename,
|
arm: add fw_cfg to "virt" board
fw_cfg already supports exposure over MMIO (used in ppc/mac_newworld.c,
ppc/mac_oldworld.c, sparc/sun4m.c); we can easily add it to the "virt"
board.
Because MMIO access is slow on ARM KVM, we enable the guest, with
fw_cfg_init_mem_wide(), to transfer up to 8 bytes with a single access.
This has been measured to speed up transfers up to 7.5-fold, relative to
single byte data access, on both ARM KVM and x86_64 TCG.
The MMIO register block of fw_cfg is advertized in the device tree. As
base address we pick 0x09020000, which conforms to the comment preceding
"a15memmap": it falls in the miscellaneous device I/O range 128MB..256MB,
and it is aligned at 64KB. The DTB properties follow the documentation in
the Linux source file "Documentation/devicetree/bindings/arm/fw-cfg.txt".
fw_cfg automatically exports a number of files to the guest; for example,
"bootorder" (see fw_cfg_machine_reset()).
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-id: 1419250305-31062-9-git-send-email-pbonzini@redhat.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
2014-12-22 15:11:42 +03:00
|
|
|
"compatible", "qemu,fw-cfg-mmio");
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_sized_cells(ms->fdt, nodename, "reg",
|
arm: add fw_cfg to "virt" board
fw_cfg already supports exposure over MMIO (used in ppc/mac_newworld.c,
ppc/mac_oldworld.c, sparc/sun4m.c); we can easily add it to the "virt"
board.
Because MMIO access is slow on ARM KVM, we enable the guest, with
fw_cfg_init_mem_wide(), to transfer up to 8 bytes with a single access.
This has been measured to speed up transfers up to 7.5-fold, relative to
single byte data access, on both ARM KVM and x86_64 TCG.
The MMIO register block of fw_cfg is advertized in the device tree. As
base address we pick 0x09020000, which conforms to the comment preceding
"a15memmap": it falls in the miscellaneous device I/O range 128MB..256MB,
and it is aligned at 64KB. The DTB properties follow the documentation in
the Linux source file "Documentation/devicetree/bindings/arm/fw-cfg.txt".
fw_cfg automatically exports a number of files to the guest; for example,
"bootorder" (see fw_cfg_machine_reset()).
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-id: 1419250305-31062-9-git-send-email-pbonzini@redhat.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
2014-12-22 15:11:42 +03:00
|
|
|
2, base, 2, size);
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop(ms->fdt, nodename, "dma-coherent", NULL, 0);
|
arm: add fw_cfg to "virt" board
fw_cfg already supports exposure over MMIO (used in ppc/mac_newworld.c,
ppc/mac_oldworld.c, sparc/sun4m.c); we can easily add it to the "virt"
board.
Because MMIO access is slow on ARM KVM, we enable the guest, with
fw_cfg_init_mem_wide(), to transfer up to 8 bytes with a single access.
This has been measured to speed up transfers up to 7.5-fold, relative to
single byte data access, on both ARM KVM and x86_64 TCG.
The MMIO register block of fw_cfg is advertized in the device tree. As
base address we pick 0x09020000, which conforms to the comment preceding
"a15memmap": it falls in the miscellaneous device I/O range 128MB..256MB,
and it is aligned at 64KB. The DTB properties follow the documentation in
the Linux source file "Documentation/devicetree/bindings/arm/fw-cfg.txt".
fw_cfg automatically exports a number of files to the guest; for example,
"bootorder" (see fw_cfg_machine_reset()).
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-id: 1419250305-31062-9-git-send-email-pbonzini@redhat.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
2014-12-22 15:11:42 +03:00
|
|
|
g_free(nodename);
|
2017-01-09 14:40:23 +03:00
|
|
|
return fw_cfg;
|
arm: add fw_cfg to "virt" board
fw_cfg already supports exposure over MMIO (used in ppc/mac_newworld.c,
ppc/mac_oldworld.c, sparc/sun4m.c); we can easily add it to the "virt"
board.
Because MMIO access is slow on ARM KVM, we enable the guest, with
fw_cfg_init_mem_wide(), to transfer up to 8 bytes with a single access.
This has been measured to speed up transfers up to 7.5-fold, relative to
single byte data access, on both ARM KVM and x86_64 TCG.
The MMIO register block of fw_cfg is advertized in the device tree. As
base address we pick 0x09020000, which conforms to the comment preceding
"a15memmap": it falls in the miscellaneous device I/O range 128MB..256MB,
and it is aligned at 64KB. The DTB properties follow the documentation in
the Linux source file "Documentation/devicetree/bindings/arm/fw-cfg.txt".
fw_cfg automatically exports a number of files to the guest; for example,
"bootorder" (see fw_cfg_machine_reset()).
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-id: 1419250305-31062-9-git-send-email-pbonzini@redhat.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
2014-12-22 15:11:42 +03:00
|
|
|
}
|
|
|
|
|
2021-03-03 20:36:36 +03:00
|
|
|
static void create_pcie_irq_map(const MachineState *ms,
|
2017-01-09 14:40:21 +03:00
|
|
|
uint32_t gic_phandle,
|
2015-02-13 08:46:08 +03:00
|
|
|
int first_irq, const char *nodename)
|
|
|
|
{
|
|
|
|
int devfn, pin;
|
2015-06-02 16:56:23 +03:00
|
|
|
uint32_t full_irq_map[4 * 4 * 10] = { 0 };
|
2015-02-13 08:46:08 +03:00
|
|
|
uint32_t *irq_map = full_irq_map;
|
|
|
|
|
|
|
|
for (devfn = 0; devfn <= 0x18; devfn += 0x8) {
|
|
|
|
for (pin = 0; pin < 4; pin++) {
|
|
|
|
int irq_type = GIC_FDT_IRQ_TYPE_SPI;
|
|
|
|
int irq_nr = first_irq + ((pin + PCI_SLOT(devfn)) % PCI_NUM_PINS);
|
|
|
|
int irq_level = GIC_FDT_IRQ_FLAGS_LEVEL_HI;
|
|
|
|
int i;
|
|
|
|
|
|
|
|
uint32_t map[] = {
|
|
|
|
devfn << 8, 0, 0, /* devfn */
|
|
|
|
pin + 1, /* PCI pin */
|
2015-06-02 16:56:23 +03:00
|
|
|
gic_phandle, 0, 0, irq_type, irq_nr, irq_level }; /* GIC irq */
|
2015-02-13 08:46:08 +03:00
|
|
|
|
|
|
|
/* Convert map to big endian */
|
2015-06-02 16:56:23 +03:00
|
|
|
for (i = 0; i < 10; i++) {
|
2015-02-13 08:46:08 +03:00
|
|
|
irq_map[i] = cpu_to_be32(map[i]);
|
|
|
|
}
|
2015-06-02 16:56:23 +03:00
|
|
|
irq_map += 10;
|
2015-02-13 08:46:08 +03:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop(ms->fdt, nodename, "interrupt-map",
|
2015-02-13 08:46:08 +03:00
|
|
|
full_irq_map, sizeof(full_irq_map));
|
|
|
|
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_cells(ms->fdt, nodename, "interrupt-map-mask",
|
2020-10-12 09:36:41 +03:00
|
|
|
cpu_to_be16(PCI_DEVFN(3, 0)), /* Slot 3 */
|
|
|
|
0, 0,
|
2015-02-13 08:46:08 +03:00
|
|
|
0x7 /* PCI irq */);
|
|
|
|
}
|
|
|
|
|
2019-12-09 12:03:06 +03:00
|
|
|
static void create_smmu(const VirtMachineState *vms,
|
2018-05-04 20:05:52 +03:00
|
|
|
PCIBus *bus)
|
|
|
|
{
|
|
|
|
char *node;
|
|
|
|
const char compat[] = "arm,smmu-v3";
|
|
|
|
int irq = vms->irqmap[VIRT_SMMU];
|
|
|
|
int i;
|
|
|
|
hwaddr base = vms->memmap[VIRT_SMMU].base;
|
|
|
|
hwaddr size = vms->memmap[VIRT_SMMU].size;
|
|
|
|
const char irq_names[] = "eventq\0priq\0cmdq-sync\0gerror";
|
|
|
|
DeviceState *dev;
|
2021-03-03 20:36:36 +03:00
|
|
|
MachineState *ms = MACHINE(vms);
|
2018-05-04 20:05:52 +03:00
|
|
|
|
|
|
|
if (vms->iommu != VIRT_IOMMU_SMMUV3 || !vms->iommu_phandle) {
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
qdev: Convert uses of qdev_create() with Coccinelle
This is the transformation explained in the commit before previous.
Takes care of just one pattern that needs conversion. More to come in
this series.
Coccinelle script:
@ depends on !(file in "hw/arm/highbank.c")@
expression bus, type_name, dev, expr;
@@
- dev = qdev_create(bus, type_name);
+ dev = qdev_new(type_name);
... when != dev = expr
- qdev_init_nofail(dev);
+ qdev_realize_and_unref(dev, bus, &error_fatal);
@@
expression bus, type_name, dev, expr;
identifier DOWN;
@@
- dev = DOWN(qdev_create(bus, type_name));
+ dev = DOWN(qdev_new(type_name));
... when != dev = expr
- qdev_init_nofail(DEVICE(dev));
+ qdev_realize_and_unref(DEVICE(dev), bus, &error_fatal);
@@
expression bus, type_name, expr;
identifier dev;
@@
- DeviceState *dev = qdev_create(bus, type_name);
+ DeviceState *dev = qdev_new(type_name);
... when != dev = expr
- qdev_init_nofail(dev);
+ qdev_realize_and_unref(dev, bus, &error_fatal);
@@
expression bus, type_name, dev, expr, errp;
symbol true;
@@
- dev = qdev_create(bus, type_name);
+ dev = qdev_new(type_name);
... when != dev = expr
- object_property_set_bool(OBJECT(dev), true, "realized", errp);
+ qdev_realize_and_unref(dev, bus, errp);
@@
expression bus, type_name, expr, errp;
identifier dev;
symbol true;
@@
- DeviceState *dev = qdev_create(bus, type_name);
+ DeviceState *dev = qdev_new(type_name);
... when != dev = expr
- object_property_set_bool(OBJECT(dev), true, "realized", errp);
+ qdev_realize_and_unref(dev, bus, errp);
The first rule exempts hw/arm/highbank.c, because it matches along two
control flow paths there, with different @type_name. Covered by the
next commit's manual conversions.
Missing #include "qapi/error.h" added manually.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20200610053247.1583243-10-armbru@redhat.com>
[Conflicts in hw/misc/empty_slot.c and hw/sparc/leon3.c resolved]
2020-06-10 08:31:58 +03:00
|
|
|
dev = qdev_new("arm-smmuv3");
|
2018-05-04 20:05:52 +03:00
|
|
|
|
qom: Put name parameter before value / visitor parameter
The object_property_set_FOO() setters take property name and value in
an unusual order:
void object_property_set_FOO(Object *obj, FOO_TYPE value,
const char *name, Error **errp)
Having to pass value before name feels grating. Swap them.
Same for object_property_set(), object_property_get(), and
object_property_parse().
Convert callers with this Coccinelle script:
@@
identifier fun = {
object_property_get, object_property_parse, object_property_set_str,
object_property_set_link, object_property_set_bool,
object_property_set_int, object_property_set_uint, object_property_set,
object_property_set_qobject
};
expression obj, v, name, errp;
@@
- fun(obj, v, name, errp)
+ fun(obj, name, v, errp)
Chokes on hw/arm/musicpal.c's lcd_refresh() with the unhelpful error
message "no position information". Convert that one manually.
Fails to convert hw/arm/armsse.c, because Coccinelle gets confused by
ARMSSE being used both as typedef and function-like macro there.
Convert manually.
Fails to convert hw/rx/rx-gdbsim.c, because Coccinelle gets confused
by RXCPU being used both as typedef and function-like macro there.
Convert manually. The other files using RXCPU that way don't need
conversion.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20200707160613.848843-27-armbru@redhat.com>
[Straightforwad conflict with commit 2336172d9b "audio: set default
value for pcspk.iobase property" resolved]
2020-07-07 19:05:54 +03:00
|
|
|
object_property_set_link(OBJECT(dev), "primary-bus", OBJECT(bus),
|
2018-05-04 20:05:52 +03:00
|
|
|
&error_abort);
|
sysbus: Convert to sysbus_realize() etc. with Coccinelle
Convert from qdev_realize(), qdev_realize_and_unref() with null @bus
argument to sysbus_realize(), sysbus_realize_and_unref().
Coccinelle script:
@@
expression dev, errp;
@@
- qdev_realize(DEVICE(dev), NULL, errp);
+ sysbus_realize(SYS_BUS_DEVICE(dev), errp);
@@
expression sysbus_dev, dev, errp;
@@
+ sysbus_dev = SYS_BUS_DEVICE(dev);
- qdev_realize_and_unref(dev, NULL, errp);
+ sysbus_realize_and_unref(sysbus_dev, errp);
- sysbus_dev = SYS_BUS_DEVICE(dev);
@@
expression sysbus_dev, dev, errp;
expression expr;
@@
sysbus_dev = SYS_BUS_DEVICE(dev);
... when != dev = expr;
- qdev_realize_and_unref(dev, NULL, errp);
+ sysbus_realize_and_unref(sysbus_dev, errp);
@@
expression dev, errp;
@@
- qdev_realize_and_unref(DEVICE(dev), NULL, errp);
+ sysbus_realize_and_unref(SYS_BUS_DEVICE(dev), errp);
@@
expression dev, errp;
@@
- qdev_realize_and_unref(dev, NULL, errp);
+ sysbus_realize_and_unref(SYS_BUS_DEVICE(dev), errp);
Whitespace changes minimized manually.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Acked-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20200610053247.1583243-46-armbru@redhat.com>
[Conflicts in hw/misc/empty_slot.c and hw/sparc/leon3.c resolved]
2020-06-10 08:32:34 +03:00
|
|
|
sysbus_realize_and_unref(SYS_BUS_DEVICE(dev), &error_fatal);
|
2018-05-04 20:05:52 +03:00
|
|
|
sysbus_mmio_map(SYS_BUS_DEVICE(dev), 0, base);
|
|
|
|
for (i = 0; i < NUM_SMMU_IRQS; i++) {
|
2019-12-09 12:03:06 +03:00
|
|
|
sysbus_connect_irq(SYS_BUS_DEVICE(dev), i,
|
|
|
|
qdev_get_gpio_in(vms->gic, irq + i));
|
2018-05-04 20:05:52 +03:00
|
|
|
}
|
|
|
|
|
|
|
|
node = g_strdup_printf("/smmuv3@%" PRIx64, base);
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_add_subnode(ms->fdt, node);
|
|
|
|
qemu_fdt_setprop(ms->fdt, node, "compatible", compat, sizeof(compat));
|
|
|
|
qemu_fdt_setprop_sized_cells(ms->fdt, node, "reg", 2, base, 2, size);
|
2018-05-04 20:05:52 +03:00
|
|
|
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_cells(ms->fdt, node, "interrupts",
|
2018-05-04 20:05:52 +03:00
|
|
|
GIC_FDT_IRQ_TYPE_SPI, irq , GIC_FDT_IRQ_FLAGS_EDGE_LO_HI,
|
|
|
|
GIC_FDT_IRQ_TYPE_SPI, irq + 1, GIC_FDT_IRQ_FLAGS_EDGE_LO_HI,
|
|
|
|
GIC_FDT_IRQ_TYPE_SPI, irq + 2, GIC_FDT_IRQ_FLAGS_EDGE_LO_HI,
|
|
|
|
GIC_FDT_IRQ_TYPE_SPI, irq + 3, GIC_FDT_IRQ_FLAGS_EDGE_LO_HI);
|
|
|
|
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop(ms->fdt, node, "interrupt-names", irq_names,
|
2018-05-04 20:05:52 +03:00
|
|
|
sizeof(irq_names));
|
|
|
|
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_cell(ms->fdt, node, "clocks", vms->clock_phandle);
|
|
|
|
qemu_fdt_setprop_string(ms->fdt, node, "clock-names", "apb_pclk");
|
|
|
|
qemu_fdt_setprop(ms->fdt, node, "dma-coherent", NULL, 0);
|
2018-05-04 20:05:52 +03:00
|
|
|
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_cell(ms->fdt, node, "#iommu-cells", 1);
|
2018-05-04 20:05:52 +03:00
|
|
|
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_cell(ms->fdt, node, "phandle", vms->iommu_phandle);
|
2018-05-04 20:05:52 +03:00
|
|
|
g_free(node);
|
|
|
|
}
|
|
|
|
|
2020-04-22 16:07:10 +03:00
|
|
|
static void create_virtio_iommu_dt_bindings(VirtMachineState *vms)
|
2020-02-14 16:27:44 +03:00
|
|
|
{
|
|
|
|
const char compat[] = "virtio,pci-iommu";
|
|
|
|
uint16_t bdf = vms->virtio_iommu_bdf;
|
2021-03-03 20:36:36 +03:00
|
|
|
MachineState *ms = MACHINE(vms);
|
2020-02-14 16:27:44 +03:00
|
|
|
char *node;
|
|
|
|
|
2021-03-03 20:36:36 +03:00
|
|
|
vms->iommu_phandle = qemu_fdt_alloc_phandle(ms->fdt);
|
2020-02-14 16:27:44 +03:00
|
|
|
|
|
|
|
node = g_strdup_printf("%s/virtio_iommu@%d", vms->pciehb_nodename, bdf);
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_add_subnode(ms->fdt, node);
|
|
|
|
qemu_fdt_setprop(ms->fdt, node, "compatible", compat, sizeof(compat));
|
|
|
|
qemu_fdt_setprop_sized_cells(ms->fdt, node, "reg",
|
2020-02-14 16:27:44 +03:00
|
|
|
1, bdf << 8, 1, 0, 1, 0,
|
|
|
|
1, 0, 1, 0);
|
|
|
|
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_cell(ms->fdt, node, "#iommu-cells", 1);
|
|
|
|
qemu_fdt_setprop_cell(ms->fdt, node, "phandle", vms->iommu_phandle);
|
2020-02-14 16:27:44 +03:00
|
|
|
g_free(node);
|
|
|
|
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_cells(ms->fdt, vms->pciehb_nodename, "iommu-map",
|
2020-02-14 16:27:44 +03:00
|
|
|
0x0, vms->iommu_phandle, 0x0, bdf,
|
|
|
|
bdf + 1, vms->iommu_phandle, bdf + 1, 0xffff - bdf);
|
|
|
|
}
|
|
|
|
|
2019-12-09 12:03:06 +03:00
|
|
|
static void create_pcie(VirtMachineState *vms)
|
2015-02-13 08:46:08 +03:00
|
|
|
{
|
2017-01-09 14:40:21 +03:00
|
|
|
hwaddr base_mmio = vms->memmap[VIRT_PCIE_MMIO].base;
|
|
|
|
hwaddr size_mmio = vms->memmap[VIRT_PCIE_MMIO].size;
|
2019-03-04 13:13:31 +03:00
|
|
|
hwaddr base_mmio_high = vms->memmap[VIRT_HIGH_PCIE_MMIO].base;
|
|
|
|
hwaddr size_mmio_high = vms->memmap[VIRT_HIGH_PCIE_MMIO].size;
|
2017-01-09 14:40:21 +03:00
|
|
|
hwaddr base_pio = vms->memmap[VIRT_PCIE_PIO].base;
|
|
|
|
hwaddr size_pio = vms->memmap[VIRT_PCIE_PIO].size;
|
2018-06-22 15:28:37 +03:00
|
|
|
hwaddr base_ecam, size_ecam;
|
2015-05-29 13:28:54 +03:00
|
|
|
hwaddr base = base_mmio;
|
2018-06-22 15:28:37 +03:00
|
|
|
int nr_pcie_buses;
|
2017-01-09 14:40:21 +03:00
|
|
|
int irq = vms->irqmap[VIRT_PCIE];
|
2015-02-13 08:46:08 +03:00
|
|
|
MemoryRegion *mmio_alias;
|
|
|
|
MemoryRegion *mmio_reg;
|
|
|
|
MemoryRegion *ecam_alias;
|
|
|
|
MemoryRegion *ecam_reg;
|
|
|
|
DeviceState *dev;
|
|
|
|
char *nodename;
|
2018-06-22 15:28:37 +03:00
|
|
|
int i, ecam_id;
|
2016-01-11 18:52:18 +03:00
|
|
|
PCIHostState *pci;
|
2021-03-03 20:36:36 +03:00
|
|
|
MachineState *ms = MACHINE(vms);
|
2015-02-13 08:46:08 +03:00
|
|
|
|
qdev: Convert uses of qdev_create() with Coccinelle
This is the transformation explained in the commit before previous.
Takes care of just one pattern that needs conversion. More to come in
this series.
Coccinelle script:
@ depends on !(file in "hw/arm/highbank.c")@
expression bus, type_name, dev, expr;
@@
- dev = qdev_create(bus, type_name);
+ dev = qdev_new(type_name);
... when != dev = expr
- qdev_init_nofail(dev);
+ qdev_realize_and_unref(dev, bus, &error_fatal);
@@
expression bus, type_name, dev, expr;
identifier DOWN;
@@
- dev = DOWN(qdev_create(bus, type_name));
+ dev = DOWN(qdev_new(type_name));
... when != dev = expr
- qdev_init_nofail(DEVICE(dev));
+ qdev_realize_and_unref(DEVICE(dev), bus, &error_fatal);
@@
expression bus, type_name, expr;
identifier dev;
@@
- DeviceState *dev = qdev_create(bus, type_name);
+ DeviceState *dev = qdev_new(type_name);
... when != dev = expr
- qdev_init_nofail(dev);
+ qdev_realize_and_unref(dev, bus, &error_fatal);
@@
expression bus, type_name, dev, expr, errp;
symbol true;
@@
- dev = qdev_create(bus, type_name);
+ dev = qdev_new(type_name);
... when != dev = expr
- object_property_set_bool(OBJECT(dev), true, "realized", errp);
+ qdev_realize_and_unref(dev, bus, errp);
@@
expression bus, type_name, expr, errp;
identifier dev;
symbol true;
@@
- DeviceState *dev = qdev_create(bus, type_name);
+ DeviceState *dev = qdev_new(type_name);
... when != dev = expr
- object_property_set_bool(OBJECT(dev), true, "realized", errp);
+ qdev_realize_and_unref(dev, bus, errp);
The first rule exempts hw/arm/highbank.c, because it matches along two
control flow paths there, with different @type_name. Covered by the
next commit's manual conversions.
Missing #include "qapi/error.h" added manually.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20200610053247.1583243-10-armbru@redhat.com>
[Conflicts in hw/misc/empty_slot.c and hw/sparc/leon3.c resolved]
2020-06-10 08:31:58 +03:00
|
|
|
dev = qdev_new(TYPE_GPEX_HOST);
|
sysbus: Convert to sysbus_realize() etc. with Coccinelle
Convert from qdev_realize(), qdev_realize_and_unref() with null @bus
argument to sysbus_realize(), sysbus_realize_and_unref().
Coccinelle script:
@@
expression dev, errp;
@@
- qdev_realize(DEVICE(dev), NULL, errp);
+ sysbus_realize(SYS_BUS_DEVICE(dev), errp);
@@
expression sysbus_dev, dev, errp;
@@
+ sysbus_dev = SYS_BUS_DEVICE(dev);
- qdev_realize_and_unref(dev, NULL, errp);
+ sysbus_realize_and_unref(sysbus_dev, errp);
- sysbus_dev = SYS_BUS_DEVICE(dev);
@@
expression sysbus_dev, dev, errp;
expression expr;
@@
sysbus_dev = SYS_BUS_DEVICE(dev);
... when != dev = expr;
- qdev_realize_and_unref(dev, NULL, errp);
+ sysbus_realize_and_unref(sysbus_dev, errp);
@@
expression dev, errp;
@@
- qdev_realize_and_unref(DEVICE(dev), NULL, errp);
+ sysbus_realize_and_unref(SYS_BUS_DEVICE(dev), errp);
@@
expression dev, errp;
@@
- qdev_realize_and_unref(dev, NULL, errp);
+ sysbus_realize_and_unref(SYS_BUS_DEVICE(dev), errp);
Whitespace changes minimized manually.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Acked-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20200610053247.1583243-46-armbru@redhat.com>
[Conflicts in hw/misc/empty_slot.c and hw/sparc/leon3.c resolved]
2020-06-10 08:32:34 +03:00
|
|
|
sysbus_realize_and_unref(SYS_BUS_DEVICE(dev), &error_fatal);
|
2015-02-13 08:46:08 +03:00
|
|
|
|
2018-06-22 15:28:37 +03:00
|
|
|
ecam_id = VIRT_ECAM_ID(vms->highmem_ecam);
|
|
|
|
base_ecam = vms->memmap[ecam_id].base;
|
|
|
|
size_ecam = vms->memmap[ecam_id].size;
|
|
|
|
nr_pcie_buses = size_ecam / PCIE_MMCFG_SIZE_MIN;
|
2015-02-13 08:46:08 +03:00
|
|
|
/* Map only the first size_ecam bytes of ECAM space */
|
|
|
|
ecam_alias = g_new0(MemoryRegion, 1);
|
|
|
|
ecam_reg = sysbus_mmio_get_region(SYS_BUS_DEVICE(dev), 0);
|
|
|
|
memory_region_init_alias(ecam_alias, OBJECT(dev), "pcie-ecam",
|
|
|
|
ecam_reg, 0, size_ecam);
|
|
|
|
memory_region_add_subregion(get_system_memory(), base_ecam, ecam_alias);
|
|
|
|
|
|
|
|
/* Map the MMIO window into system address space so as to expose
|
|
|
|
* the section of PCI MMIO space which starts at the same base address
|
|
|
|
* (ie 1:1 mapping for that part of PCI MMIO space visible through
|
|
|
|
* the window).
|
|
|
|
*/
|
|
|
|
mmio_alias = g_new0(MemoryRegion, 1);
|
|
|
|
mmio_reg = sysbus_mmio_get_region(SYS_BUS_DEVICE(dev), 1);
|
|
|
|
memory_region_init_alias(mmio_alias, OBJECT(dev), "pcie-mmio",
|
|
|
|
mmio_reg, base_mmio, size_mmio);
|
|
|
|
memory_region_add_subregion(get_system_memory(), base_mmio, mmio_alias);
|
|
|
|
|
2017-01-09 14:40:22 +03:00
|
|
|
if (vms->highmem) {
|
2015-09-07 12:39:29 +03:00
|
|
|
/* Map high MMIO space */
|
|
|
|
MemoryRegion *high_mmio_alias = g_new0(MemoryRegion, 1);
|
|
|
|
|
|
|
|
memory_region_init_alias(high_mmio_alias, OBJECT(dev), "pcie-mmio-high",
|
|
|
|
mmio_reg, base_mmio_high, size_mmio_high);
|
|
|
|
memory_region_add_subregion(get_system_memory(), base_mmio_high,
|
|
|
|
high_mmio_alias);
|
|
|
|
}
|
|
|
|
|
2015-02-13 08:46:08 +03:00
|
|
|
/* Map IO port space */
|
2015-05-29 13:28:54 +03:00
|
|
|
sysbus_mmio_map(SYS_BUS_DEVICE(dev), 2, base_pio);
|
2015-02-13 08:46:08 +03:00
|
|
|
|
|
|
|
for (i = 0; i < GPEX_NUM_IRQS; i++) {
|
2019-12-09 12:03:06 +03:00
|
|
|
sysbus_connect_irq(SYS_BUS_DEVICE(dev), i,
|
|
|
|
qdev_get_gpio_in(vms->gic, irq + i));
|
2017-09-14 20:43:19 +03:00
|
|
|
gpex_set_irq_num(GPEX_HOST(dev), i, irq + i);
|
2015-02-13 08:46:08 +03:00
|
|
|
}
|
|
|
|
|
2016-01-11 18:52:18 +03:00
|
|
|
pci = PCI_HOST_BRIDGE(dev);
|
2021-07-08 15:55:13 +03:00
|
|
|
pci->bypass_iommu = vms->default_bus_bypass_iommu;
|
2020-11-19 04:48:35 +03:00
|
|
|
vms->bus = pci->bus;
|
|
|
|
if (vms->bus) {
|
2016-01-11 18:52:18 +03:00
|
|
|
for (i = 0; i < nb_nics; i++) {
|
|
|
|
NICInfo *nd = &nd_table[i];
|
|
|
|
|
|
|
|
if (!nd->model) {
|
|
|
|
nd->model = g_strdup("virtio");
|
|
|
|
}
|
|
|
|
|
|
|
|
pci_nic_init_nofail(nd, pci->bus, nd->model, NULL);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2020-02-14 16:27:44 +03:00
|
|
|
nodename = vms->pciehb_nodename = g_strdup_printf("/pcie@%" PRIx64, base);
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_add_subnode(ms->fdt, nodename);
|
|
|
|
qemu_fdt_setprop_string(ms->fdt, nodename,
|
2015-02-13 08:46:08 +03:00
|
|
|
"compatible", "pci-host-ecam-generic");
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_string(ms->fdt, nodename, "device_type", "pci");
|
|
|
|
qemu_fdt_setprop_cell(ms->fdt, nodename, "#address-cells", 3);
|
|
|
|
qemu_fdt_setprop_cell(ms->fdt, nodename, "#size-cells", 2);
|
|
|
|
qemu_fdt_setprop_cell(ms->fdt, nodename, "linux,pci-domain", 0);
|
|
|
|
qemu_fdt_setprop_cells(ms->fdt, nodename, "bus-range", 0,
|
2015-02-13 08:46:08 +03:00
|
|
|
nr_pcie_buses - 1);
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop(ms->fdt, nodename, "dma-coherent", NULL, 0);
|
2015-02-13 08:46:08 +03:00
|
|
|
|
2017-01-09 14:40:21 +03:00
|
|
|
if (vms->msi_phandle) {
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_cells(ms->fdt, nodename, "msi-parent",
|
2017-01-09 14:40:21 +03:00
|
|
|
vms->msi_phandle);
|
2015-09-24 03:29:37 +03:00
|
|
|
}
|
2015-06-02 16:56:23 +03:00
|
|
|
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_sized_cells(ms->fdt, nodename, "reg",
|
2015-02-13 08:46:08 +03:00
|
|
|
2, base_ecam, 2, size_ecam);
|
2015-09-07 12:39:29 +03:00
|
|
|
|
2017-01-09 14:40:22 +03:00
|
|
|
if (vms->highmem) {
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_sized_cells(ms->fdt, nodename, "ranges",
|
2015-09-07 12:39:29 +03:00
|
|
|
1, FDT_PCI_RANGE_IOPORT, 2, 0,
|
|
|
|
2, base_pio, 2, size_pio,
|
|
|
|
1, FDT_PCI_RANGE_MMIO, 2, base_mmio,
|
|
|
|
2, base_mmio, 2, size_mmio,
|
|
|
|
1, FDT_PCI_RANGE_MMIO_64BIT,
|
|
|
|
2, base_mmio_high,
|
|
|
|
2, base_mmio_high, 2, size_mmio_high);
|
|
|
|
} else {
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_sized_cells(ms->fdt, nodename, "ranges",
|
2015-09-07 12:39:29 +03:00
|
|
|
1, FDT_PCI_RANGE_IOPORT, 2, 0,
|
|
|
|
2, base_pio, 2, size_pio,
|
|
|
|
1, FDT_PCI_RANGE_MMIO, 2, base_mmio,
|
|
|
|
2, base_mmio, 2, size_mmio);
|
|
|
|
}
|
2015-02-13 08:46:08 +03:00
|
|
|
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_cell(ms->fdt, nodename, "#interrupt-cells", 1);
|
|
|
|
create_pcie_irq_map(ms, vms->gic_phandle, irq, nodename);
|
2015-02-13 08:46:08 +03:00
|
|
|
|
2018-05-04 20:05:52 +03:00
|
|
|
if (vms->iommu) {
|
2021-03-03 20:36:36 +03:00
|
|
|
vms->iommu_phandle = qemu_fdt_alloc_phandle(ms->fdt);
|
2018-05-04 20:05:52 +03:00
|
|
|
|
2020-02-14 16:27:44 +03:00
|
|
|
switch (vms->iommu) {
|
|
|
|
case VIRT_IOMMU_SMMUV3:
|
2020-11-19 04:48:35 +03:00
|
|
|
create_smmu(vms, vms->bus);
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_setprop_cells(ms->fdt, nodename, "iommu-map",
|
2020-02-14 16:27:44 +03:00
|
|
|
0x0, vms->iommu_phandle, 0x0, 0x10000);
|
|
|
|
break;
|
|
|
|
default:
|
|
|
|
g_assert_not_reached();
|
|
|
|
}
|
2018-05-04 20:05:52 +03:00
|
|
|
}
|
2015-02-13 08:46:08 +03:00
|
|
|
}
|
|
|
|
|
2019-12-09 12:03:06 +03:00
|
|
|
static void create_platform_bus(VirtMachineState *vms)
|
2015-06-02 14:29:13 +03:00
|
|
|
{
|
|
|
|
DeviceState *dev;
|
|
|
|
SysBusDevice *s;
|
|
|
|
int i;
|
|
|
|
MemoryRegion *sysmem = get_system_memory();
|
|
|
|
|
qdev: Convert uses of qdev_create() with Coccinelle
This is the transformation explained in the commit before previous.
Takes care of just one pattern that needs conversion. More to come in
this series.
Coccinelle script:
@ depends on !(file in "hw/arm/highbank.c")@
expression bus, type_name, dev, expr;
@@
- dev = qdev_create(bus, type_name);
+ dev = qdev_new(type_name);
... when != dev = expr
- qdev_init_nofail(dev);
+ qdev_realize_and_unref(dev, bus, &error_fatal);
@@
expression bus, type_name, dev, expr;
identifier DOWN;
@@
- dev = DOWN(qdev_create(bus, type_name));
+ dev = DOWN(qdev_new(type_name));
... when != dev = expr
- qdev_init_nofail(DEVICE(dev));
+ qdev_realize_and_unref(DEVICE(dev), bus, &error_fatal);
@@
expression bus, type_name, expr;
identifier dev;
@@
- DeviceState *dev = qdev_create(bus, type_name);
+ DeviceState *dev = qdev_new(type_name);
... when != dev = expr
- qdev_init_nofail(dev);
+ qdev_realize_and_unref(dev, bus, &error_fatal);
@@
expression bus, type_name, dev, expr, errp;
symbol true;
@@
- dev = qdev_create(bus, type_name);
+ dev = qdev_new(type_name);
... when != dev = expr
- object_property_set_bool(OBJECT(dev), true, "realized", errp);
+ qdev_realize_and_unref(dev, bus, errp);
@@
expression bus, type_name, expr, errp;
identifier dev;
symbol true;
@@
- DeviceState *dev = qdev_create(bus, type_name);
+ DeviceState *dev = qdev_new(type_name);
... when != dev = expr
- object_property_set_bool(OBJECT(dev), true, "realized", errp);
+ qdev_realize_and_unref(dev, bus, errp);
The first rule exempts hw/arm/highbank.c, because it matches along two
control flow paths there, with different @type_name. Covered by the
next commit's manual conversions.
Missing #include "qapi/error.h" added manually.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20200610053247.1583243-10-armbru@redhat.com>
[Conflicts in hw/misc/empty_slot.c and hw/sparc/leon3.c resolved]
2020-06-10 08:31:58 +03:00
|
|
|
dev = qdev_new(TYPE_PLATFORM_BUS_DEVICE);
|
2021-10-08 16:34:35 +03:00
|
|
|
dev->id = g_strdup(TYPE_PLATFORM_BUS_DEVICE);
|
2018-05-10 20:10:56 +03:00
|
|
|
qdev_prop_set_uint32(dev, "num_irqs", PLATFORM_BUS_NUM_IRQS);
|
|
|
|
qdev_prop_set_uint32(dev, "mmio_size", vms->memmap[VIRT_PLATFORM_BUS].size);
|
sysbus: Convert to sysbus_realize() etc. with Coccinelle
Convert from qdev_realize(), qdev_realize_and_unref() with null @bus
argument to sysbus_realize(), sysbus_realize_and_unref().
Coccinelle script:
@@
expression dev, errp;
@@
- qdev_realize(DEVICE(dev), NULL, errp);
+ sysbus_realize(SYS_BUS_DEVICE(dev), errp);
@@
expression sysbus_dev, dev, errp;
@@
+ sysbus_dev = SYS_BUS_DEVICE(dev);
- qdev_realize_and_unref(dev, NULL, errp);
+ sysbus_realize_and_unref(sysbus_dev, errp);
- sysbus_dev = SYS_BUS_DEVICE(dev);
@@
expression sysbus_dev, dev, errp;
expression expr;
@@
sysbus_dev = SYS_BUS_DEVICE(dev);
... when != dev = expr;
- qdev_realize_and_unref(dev, NULL, errp);
+ sysbus_realize_and_unref(sysbus_dev, errp);
@@
expression dev, errp;
@@
- qdev_realize_and_unref(DEVICE(dev), NULL, errp);
+ sysbus_realize_and_unref(SYS_BUS_DEVICE(dev), errp);
@@
expression dev, errp;
@@
- qdev_realize_and_unref(dev, NULL, errp);
+ sysbus_realize_and_unref(SYS_BUS_DEVICE(dev), errp);
Whitespace changes minimized manually.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Acked-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20200610053247.1583243-46-armbru@redhat.com>
[Conflicts in hw/misc/empty_slot.c and hw/sparc/leon3.c resolved]
2020-06-10 08:32:34 +03:00
|
|
|
sysbus_realize_and_unref(SYS_BUS_DEVICE(dev), &error_fatal);
|
2018-05-10 20:10:56 +03:00
|
|
|
vms->platform_bus_dev = dev;
|
2015-06-02 14:29:13 +03:00
|
|
|
|
2018-05-10 20:10:56 +03:00
|
|
|
s = SYS_BUS_DEVICE(dev);
|
|
|
|
for (i = 0; i < PLATFORM_BUS_NUM_IRQS; i++) {
|
2019-12-09 12:03:06 +03:00
|
|
|
int irq = vms->irqmap[VIRT_PLATFORM_BUS] + i;
|
|
|
|
sysbus_connect_irq(s, i, qdev_get_gpio_in(vms->gic, irq));
|
2015-06-02 14:29:13 +03:00
|
|
|
}
|
|
|
|
|
|
|
|
memory_region_add_subregion(sysmem,
|
2018-05-10 20:10:56 +03:00
|
|
|
vms->memmap[VIRT_PLATFORM_BUS].base,
|
2015-06-02 14:29:13 +03:00
|
|
|
sysbus_mmio_get_region(s, 0));
|
|
|
|
}
|
|
|
|
|
2020-06-26 06:31:41 +03:00
|
|
|
static void create_tag_ram(MemoryRegion *tag_sysmem,
|
|
|
|
hwaddr base, hwaddr size,
|
|
|
|
const char *name)
|
|
|
|
{
|
|
|
|
MemoryRegion *tagram = g_new(MemoryRegion, 1);
|
|
|
|
|
|
|
|
memory_region_init_ram(tagram, NULL, name, size / 32, &error_fatal);
|
|
|
|
memory_region_add_subregion(tag_sysmem, base / 32, tagram);
|
|
|
|
}
|
|
|
|
|
2017-01-09 14:40:21 +03:00
|
|
|
static void create_secure_ram(VirtMachineState *vms,
|
2020-06-26 06:31:41 +03:00
|
|
|
MemoryRegion *secure_sysmem,
|
|
|
|
MemoryRegion *secure_tag_sysmem)
|
2016-03-04 14:30:17 +03:00
|
|
|
{
|
|
|
|
MemoryRegion *secram = g_new(MemoryRegion, 1);
|
|
|
|
char *nodename;
|
2017-01-09 14:40:21 +03:00
|
|
|
hwaddr base = vms->memmap[VIRT_SECURE_MEM].base;
|
|
|
|
hwaddr size = vms->memmap[VIRT_SECURE_MEM].size;
|
2021-03-03 20:36:36 +03:00
|
|
|
MachineState *ms = MACHINE(vms);
|
2016-03-04 14:30:17 +03:00
|
|
|
|
2017-07-07 17:42:53 +03:00
|
|
|
memory_region_init_ram(secram, NULL, "virt.secure-ram", size,
|
|
|
|
&error_fatal);
|
2016-03-04 14:30:17 +03:00
|
|
|
memory_region_add_subregion(secure_sysmem, base, secram);
|
|
|
|
|
|
|
|
nodename = g_strdup_printf("/secram@%" PRIx64, base);
|
2021-03-03 20:36:36 +03:00
|
|
|
qemu_fdt_add_subnode(ms->fdt, nodename);
|
|
|
|
qemu_fdt_setprop_string(ms->fdt, nodename, "device_type", "memory");
|
|
|
|
qemu_fdt_setprop_sized_cells(ms->fdt, nodename, "reg", 2, base, 2, size);
|
|
|
|
qemu_fdt_setprop_string(ms->fdt, nodename, "status", "disabled");
|
|
|
|
qemu_fdt_setprop_string(ms->fdt, nodename, "secure-status", "okay");
|
2016-03-04 14:30:17 +03:00
|
|
|
|
2020-06-26 06:31:41 +03:00
|
|
|
if (secure_tag_sysmem) {
|
|
|
|
create_tag_ram(secure_tag_sysmem, base, size, "mach-virt.secure-tag");
|
|
|
|
}
|
|
|
|
|
2016-03-04 14:30:17 +03:00
|
|
|
g_free(nodename);
|
|
|
|
}
|
|
|
|
|
2013-11-22 21:17:14 +04:00
|
|
|
static void *machvirt_dtb(const struct arm_boot_info *binfo, int *fdt_size)
|
|
|
|
{
|
2017-01-09 14:40:21 +03:00
|
|
|
const VirtMachineState *board = container_of(binfo, VirtMachineState,
|
|
|
|
bootinfo);
|
2021-03-03 20:36:36 +03:00
|
|
|
MachineState *ms = MACHINE(board);
|
|
|
|
|
2013-11-22 21:17:14 +04:00
|
|
|
|
|
|
|
*fdt_size = board->fdt_size;
|
2021-03-03 20:36:36 +03:00
|
|
|
return ms->fdt;
|
2013-11-22 21:17:14 +04:00
|
|
|
}
|
|
|
|
|
2017-01-09 14:40:22 +03:00
|
|
|
static void virt_build_smbios(VirtMachineState *vms)
|
2015-09-07 12:39:29 +03:00
|
|
|
{
|
2018-03-23 21:26:46 +03:00
|
|
|
MachineClass *mc = MACHINE_GET_CLASS(vms);
|
|
|
|
VirtMachineClass *vmc = VIRT_MACHINE_GET_CLASS(vms);
|
2015-09-07 12:39:29 +03:00
|
|
|
uint8_t *smbios_tables, *smbios_anchor;
|
|
|
|
size_t smbios_tables_len, smbios_anchor_len;
|
2015-10-16 13:14:53 +03:00
|
|
|
const char *product = "QEMU Virtual Machine";
|
2015-09-07 12:39:29 +03:00
|
|
|
|
2015-10-16 13:14:53 +03:00
|
|
|
if (kvm_enabled()) {
|
|
|
|
product = "KVM Virtual Machine";
|
|
|
|
}
|
|
|
|
|
|
|
|
smbios_set_defaults("QEMU", product,
|
2018-03-23 21:26:46 +03:00
|
|
|
vmc->smbios_old_sys_ver ? "1.0" : mc->name, false,
|
|
|
|
true, SMBIOS_ENTRY_POINT_30);
|
2015-09-07 12:39:29 +03:00
|
|
|
|
hw/smbios: support for type 41 (onboard devices extended information)
Type 41 defines the attributes of devices that are onboard. The
original intent was to imply the BIOS had some level of control over
the enablement of the associated devices.
If network devices are present in this table, by default, udev will
name the corresponding interfaces enoX, X being the instance number.
Without such information, udev will fallback to using the PCI ID and
this usually gives ens3 or ens4. This can be a bit annoying as the
name of the network card may depend on the order of options and may
change if a new PCI device is added earlier on the commande line.
Being able to provide SMBIOS type 41 entry ensure the name of the
interface won't change and helps the user guess the right name without
booting a first time.
This can be invoked with:
$QEMU -netdev user,id=internet
-device virtio-net-pci,mac=50:54:00:00:00:42,netdev=internet,id=internet-dev \
-smbios type=41,designation='Onboard LAN',instance=1,kind=ethernet,pcidev=internet-dev
The PCI segment is assumed to be 0. This should hold true for most
cases.
$ dmidecode -t 41
# dmidecode 3.3
Getting SMBIOS data from sysfs.
SMBIOS 2.8 present.
Handle 0x2900, DMI type 41, 11 bytes
Onboard Device
Reference Designation: Onboard LAN
Type: Ethernet
Status: Enabled
Type Instance: 1
Bus Address: 0000:00:09.0
$ ip -brief a
lo UNKNOWN 127.0.0.1/8 ::1/128
eno1 UP 10.0.2.14/24 fec0::5254:ff:fe00:42/64 fe80::5254:ff:fe00:42/64
Signed-off-by: Vincent Bernat <vincent@bernat.ch>
Message-Id: <20210401171138.62970-1-vincent@bernat.ch>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2021-04-01 20:11:38 +03:00
|
|
|
smbios_get_tables(MACHINE(vms), NULL, 0,
|
|
|
|
&smbios_tables, &smbios_tables_len,
|
|
|
|
&smbios_anchor, &smbios_anchor_len,
|
|
|
|
&error_fatal);
|
2015-09-07 12:39:29 +03:00
|
|
|
|
|
|
|
if (smbios_anchor) {
|
2017-01-09 14:40:23 +03:00
|
|
|
fw_cfg_add_file(vms->fw_cfg, "etc/smbios/smbios-tables",
|
2015-09-07 12:39:29 +03:00
|
|
|
smbios_tables, smbios_tables_len);
|
2017-01-09 14:40:23 +03:00
|
|
|
fw_cfg_add_file(vms->fw_cfg, "etc/smbios/smbios-anchor",
|
2015-09-07 12:39:29 +03:00
|
|
|
smbios_anchor, smbios_anchor_len);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2015-05-29 13:28:59 +03:00
|
|
|
static
|
2017-01-09 14:40:22 +03:00
|
|
|
void virt_machine_done(Notifier *notifier, void *data)
|
2015-05-29 13:28:59 +03:00
|
|
|
{
|
2017-01-09 14:40:22 +03:00
|
|
|
VirtMachineState *vms = container_of(notifier, VirtMachineState,
|
|
|
|
machine_done);
|
2019-08-09 09:57:21 +03:00
|
|
|
MachineState *ms = MACHINE(vms);
|
2018-05-10 20:10:56 +03:00
|
|
|
ARMCPU *cpu = ARM_CPU(first_cpu);
|
|
|
|
struct arm_boot_info *info = &vms->bootinfo;
|
|
|
|
AddressSpace *as = arm_boot_address_space(cpu, info);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* If the user provided a dtb, we assume the dynamic sysbus nodes
|
|
|
|
* already are integrated there. This corresponds to a use case where
|
|
|
|
* the dynamic sysbus nodes are complex and their generation is not yet
|
|
|
|
* supported. In that case the user can take charge of the guest dt
|
|
|
|
* while qemu takes charge of the qom stuff.
|
|
|
|
*/
|
|
|
|
if (info->dtb_filename == NULL) {
|
2021-03-03 20:36:36 +03:00
|
|
|
platform_bus_add_all_fdt_nodes(ms->fdt, "/intc",
|
2018-05-10 20:10:56 +03:00
|
|
|
vms->memmap[VIRT_PLATFORM_BUS].base,
|
|
|
|
vms->memmap[VIRT_PLATFORM_BUS].size,
|
|
|
|
vms->irqmap[VIRT_PLATFORM_BUS]);
|
|
|
|
}
|
2019-08-09 09:57:21 +03:00
|
|
|
if (arm_load_dtb(info->dtb_start, info, info->dtb_limit, as, ms) < 0) {
|
2018-05-10 20:10:56 +03:00
|
|
|
exit(1);
|
|
|
|
}
|
2017-01-09 14:40:22 +03:00
|
|
|
|
2020-11-19 04:48:35 +03:00
|
|
|
fw_cfg_add_extra_pci_roots(vms->bus, vms->fw_cfg);
|
|
|
|
|
2017-01-09 14:40:22 +03:00
|
|
|
virt_acpi_setup(vms);
|
|
|
|
virt_build_smbios(vms);
|
2015-05-29 13:28:59 +03:00
|
|
|
}
|
|
|
|
|
2017-05-03 15:56:56 +03:00
|
|
|
static uint64_t virt_cpu_mp_affinity(VirtMachineState *vms, int idx)
|
|
|
|
{
|
|
|
|
uint8_t clustersz = ARM_DEFAULT_CPUS_PER_CLUSTER;
|
|
|
|
VirtMachineClass *vmc = VIRT_MACHINE_GET_CLASS(vms);
|
|
|
|
|
|
|
|
if (!vmc->disallow_affinity_adjustment) {
|
|
|
|
/* Adjust MPIDR like 64-bit KVM hosts, which incorporate the
|
|
|
|
* GIC's target-list limitations. 32-bit KVM hosts currently
|
|
|
|
* always create clusters of 4 CPUs, but that is expected to
|
|
|
|
* change when they gain support for gicv3. When KVM is enabled
|
|
|
|
* it will override the changes we make here, therefore our
|
|
|
|
* purposes are to make TCG consistent (with 64-bit KVM hosts)
|
|
|
|
* and to improve SGI efficiency.
|
|
|
|
*/
|
2020-03-11 16:16:14 +03:00
|
|
|
if (vms->gic_version == VIRT_GIC_VERSION_3) {
|
2017-05-03 15:56:56 +03:00
|
|
|
clustersz = GICV3_TARGETLIST_BITS;
|
|
|
|
} else {
|
|
|
|
clustersz = GIC_TARGETLIST_BITS;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
return arm_cpu_mp_affinity(idx, clustersz);
|
|
|
|
}
|
|
|
|
|
2019-03-04 13:13:32 +03:00
|
|
|
static void virt_set_memmap(VirtMachineState *vms)
|
|
|
|
{
|
2019-03-04 13:13:36 +03:00
|
|
|
MachineState *ms = MACHINE(vms);
|
|
|
|
hwaddr base, device_memory_base, device_memory_size;
|
2019-03-04 13:13:32 +03:00
|
|
|
int i;
|
|
|
|
|
|
|
|
vms->memmap = extended_memmap;
|
|
|
|
|
|
|
|
for (i = 0; i < ARRAY_SIZE(base_memmap); i++) {
|
|
|
|
vms->memmap[i] = base_memmap[i];
|
|
|
|
}
|
|
|
|
|
2019-03-04 13:13:36 +03:00
|
|
|
if (ms->ram_slots > ACPI_MAX_RAM_SLOTS) {
|
|
|
|
error_report("unsupported number of memory slots: %"PRIu64,
|
|
|
|
ms->ram_slots);
|
|
|
|
exit(EXIT_FAILURE);
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* We compute the base of the high IO region depending on the
|
|
|
|
* amount of initial and device memory. The device memory start/size
|
|
|
|
* is aligned on 1GiB. We never put the high IO region below 256GiB
|
|
|
|
* so that if maxram_size is < 255GiB we keep the legacy memory map.
|
|
|
|
* The device region size assumes 1GiB page max alignment per slot.
|
|
|
|
*/
|
|
|
|
device_memory_base =
|
|
|
|
ROUND_UP(vms->memmap[VIRT_MEM].base + ms->ram_size, GiB);
|
|
|
|
device_memory_size = ms->maxram_size - ms->ram_size + ms->ram_slots * GiB;
|
|
|
|
|
|
|
|
/* Base address of the high IO region */
|
|
|
|
base = device_memory_base + ROUND_UP(device_memory_size, GiB);
|
|
|
|
if (base < device_memory_base) {
|
|
|
|
error_report("maxmem/slots too huge");
|
|
|
|
exit(EXIT_FAILURE);
|
|
|
|
}
|
|
|
|
if (base < vms->memmap[VIRT_MEM].base + LEGACY_RAMLIMIT_BYTES) {
|
|
|
|
base = vms->memmap[VIRT_MEM].base + LEGACY_RAMLIMIT_BYTES;
|
|
|
|
}
|
2019-03-04 13:13:32 +03:00
|
|
|
|
|
|
|
for (i = VIRT_LOWMEMMAP_LAST; i < ARRAY_SIZE(extended_memmap); i++) {
|
|
|
|
hwaddr size = extended_memmap[i].size;
|
|
|
|
|
|
|
|
base = ROUND_UP(base, size);
|
|
|
|
vms->memmap[i].base = base;
|
|
|
|
vms->memmap[i].size = size;
|
|
|
|
base += size;
|
|
|
|
}
|
2019-03-04 13:13:36 +03:00
|
|
|
vms->highest_gpa = base - 1;
|
|
|
|
if (device_memory_size > 0) {
|
|
|
|
ms->device_memory = g_malloc0(sizeof(*ms->device_memory));
|
|
|
|
ms->device_memory->base = device_memory_base;
|
|
|
|
memory_region_init(&ms->device_memory->mr, OBJECT(vms),
|
|
|
|
"device-memory", device_memory_size);
|
|
|
|
}
|
2019-03-04 13:13:32 +03:00
|
|
|
}
|
|
|
|
|
2020-03-11 16:16:15 +03:00
|
|
|
/*
|
|
|
|
* finalize_gic_version - Determines the final gic_version
|
|
|
|
* according to the gic-version property
|
|
|
|
*
|
|
|
|
* Default GIC type is v2
|
|
|
|
*/
|
|
|
|
static void finalize_gic_version(VirtMachineState *vms)
|
|
|
|
{
|
2020-03-11 16:16:18 +03:00
|
|
|
unsigned int max_cpus = MACHINE(vms)->smp.max_cpus;
|
|
|
|
|
2020-03-11 16:16:17 +03:00
|
|
|
if (kvm_enabled()) {
|
|
|
|
int probe_bitmap;
|
2020-03-11 16:16:16 +03:00
|
|
|
|
2020-03-11 16:16:17 +03:00
|
|
|
if (!kvm_irqchip_in_kernel()) {
|
|
|
|
switch (vms->gic_version) {
|
|
|
|
case VIRT_GIC_VERSION_HOST:
|
|
|
|
warn_report(
|
|
|
|
"gic-version=host not relevant with kernel-irqchip=off "
|
|
|
|
"as only userspace GICv2 is supported. Using v2 ...");
|
|
|
|
return;
|
|
|
|
case VIRT_GIC_VERSION_MAX:
|
|
|
|
case VIRT_GIC_VERSION_NOSEL:
|
|
|
|
vms->gic_version = VIRT_GIC_VERSION_2;
|
|
|
|
return;
|
|
|
|
case VIRT_GIC_VERSION_2:
|
|
|
|
return;
|
|
|
|
case VIRT_GIC_VERSION_3:
|
2020-03-11 16:16:15 +03:00
|
|
|
error_report(
|
2020-03-11 16:16:17 +03:00
|
|
|
"gic-version=3 is not supported with kernel-irqchip=off");
|
2020-03-11 16:16:15 +03:00
|
|
|
exit(1);
|
2020-03-11 16:16:17 +03:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
probe_bitmap = kvm_arm_vgic_probe();
|
|
|
|
if (!probe_bitmap) {
|
|
|
|
error_report("Unable to determine GIC version supported by host");
|
|
|
|
exit(1);
|
|
|
|
}
|
|
|
|
|
|
|
|
switch (vms->gic_version) {
|
|
|
|
case VIRT_GIC_VERSION_HOST:
|
|
|
|
case VIRT_GIC_VERSION_MAX:
|
|
|
|
if (probe_bitmap & KVM_ARM_VGIC_V3) {
|
|
|
|
vms->gic_version = VIRT_GIC_VERSION_3;
|
2020-03-11 16:16:16 +03:00
|
|
|
} else {
|
2020-03-11 16:16:17 +03:00
|
|
|
vms->gic_version = VIRT_GIC_VERSION_2;
|
2020-03-11 16:16:15 +03:00
|
|
|
}
|
2020-03-11 16:16:17 +03:00
|
|
|
return;
|
|
|
|
case VIRT_GIC_VERSION_NOSEL:
|
2020-03-11 16:16:18 +03:00
|
|
|
if ((probe_bitmap & KVM_ARM_VGIC_V2) && max_cpus <= GIC_NCPU) {
|
|
|
|
vms->gic_version = VIRT_GIC_VERSION_2;
|
|
|
|
} else if (probe_bitmap & KVM_ARM_VGIC_V3) {
|
|
|
|
/*
|
|
|
|
* in case the host does not support v2 in-kernel emulation or
|
|
|
|
* the end-user requested more than 8 VCPUs we now default
|
|
|
|
* to v3. In any case defaulting to v2 would be broken.
|
|
|
|
*/
|
|
|
|
vms->gic_version = VIRT_GIC_VERSION_3;
|
|
|
|
} else if (max_cpus > GIC_NCPU) {
|
|
|
|
error_report("host only supports in-kernel GICv2 emulation "
|
|
|
|
"but more than 8 vcpus are requested");
|
|
|
|
exit(1);
|
|
|
|
}
|
2020-03-11 16:16:17 +03:00
|
|
|
break;
|
|
|
|
case VIRT_GIC_VERSION_2:
|
|
|
|
case VIRT_GIC_VERSION_3:
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Check chosen version is effectively supported by the host */
|
|
|
|
if (vms->gic_version == VIRT_GIC_VERSION_2 &&
|
|
|
|
!(probe_bitmap & KVM_ARM_VGIC_V2)) {
|
|
|
|
error_report("host does not support in-kernel GICv2 emulation");
|
|
|
|
exit(1);
|
|
|
|
} else if (vms->gic_version == VIRT_GIC_VERSION_3 &&
|
|
|
|
!(probe_bitmap & KVM_ARM_VGIC_V3)) {
|
|
|
|
error_report("host does not support in-kernel GICv3 emulation");
|
|
|
|
exit(1);
|
2020-03-11 16:16:15 +03:00
|
|
|
}
|
2020-03-11 16:16:17 +03:00
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* TCG mode */
|
|
|
|
switch (vms->gic_version) {
|
|
|
|
case VIRT_GIC_VERSION_NOSEL:
|
2020-03-11 16:16:15 +03:00
|
|
|
vms->gic_version = VIRT_GIC_VERSION_2;
|
2020-03-11 16:16:17 +03:00
|
|
|
break;
|
|
|
|
case VIRT_GIC_VERSION_MAX:
|
|
|
|
vms->gic_version = VIRT_GIC_VERSION_3;
|
|
|
|
break;
|
|
|
|
case VIRT_GIC_VERSION_HOST:
|
|
|
|
error_report("gic-version=host requires KVM");
|
|
|
|
exit(1);
|
|
|
|
case VIRT_GIC_VERSION_2:
|
|
|
|
case VIRT_GIC_VERSION_3:
|
|
|
|
break;
|
2020-03-11 16:16:15 +03:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2020-10-01 09:17:15 +03:00
|
|
|
/*
|
|
|
|
* virt_cpu_post_init() must be called after the CPUs have
|
|
|
|
* been realized and the GIC has been created.
|
|
|
|
*/
|
2020-12-15 20:48:15 +03:00
|
|
|
static void virt_cpu_post_init(VirtMachineState *vms, MemoryRegion *sysmem)
|
2020-10-01 09:17:15 +03:00
|
|
|
{
|
2020-12-15 20:48:15 +03:00
|
|
|
int max_cpus = MACHINE(vms)->smp.max_cpus;
|
2020-10-01 09:17:18 +03:00
|
|
|
bool aarch64, pmu, steal_time;
|
2020-10-01 09:17:16 +03:00
|
|
|
CPUState *cpu;
|
2020-10-01 09:17:15 +03:00
|
|
|
|
|
|
|
aarch64 = object_property_get_bool(OBJECT(first_cpu), "aarch64", NULL);
|
2020-10-01 09:17:16 +03:00
|
|
|
pmu = object_property_get_bool(OBJECT(first_cpu), "pmu", NULL);
|
2020-10-01 09:17:18 +03:00
|
|
|
steal_time = object_property_get_bool(OBJECT(first_cpu),
|
|
|
|
"kvm-steal-time", NULL);
|
2020-10-01 09:17:15 +03:00
|
|
|
|
2020-10-01 09:17:16 +03:00
|
|
|
if (kvm_enabled()) {
|
2020-10-01 09:17:18 +03:00
|
|
|
hwaddr pvtime_reg_base = vms->memmap[VIRT_PVTIME].base;
|
|
|
|
hwaddr pvtime_reg_size = vms->memmap[VIRT_PVTIME].size;
|
|
|
|
|
|
|
|
if (steal_time) {
|
|
|
|
MemoryRegion *pvtime = g_new(MemoryRegion, 1);
|
|
|
|
hwaddr pvtime_size = max_cpus * PVTIME_SIZE_PER_CPU;
|
|
|
|
|
|
|
|
/* The memory region size must be a multiple of host page size. */
|
|
|
|
pvtime_size = REAL_HOST_PAGE_ALIGN(pvtime_size);
|
|
|
|
|
|
|
|
if (pvtime_size > pvtime_reg_size) {
|
|
|
|
error_report("pvtime requires a %" HWADDR_PRId
|
|
|
|
" byte memory region for %d CPUs,"
|
|
|
|
" but only %" HWADDR_PRId " has been reserved",
|
|
|
|
pvtime_size, max_cpus, pvtime_reg_size);
|
|
|
|
exit(1);
|
|
|
|
}
|
|
|
|
|
|
|
|
memory_region_init_ram(pvtime, NULL, "pvtime", pvtime_size, NULL);
|
|
|
|
memory_region_add_subregion(sysmem, pvtime_reg_base, pvtime);
|
|
|
|
}
|
|
|
|
|
2020-10-01 09:17:16 +03:00
|
|
|
CPU_FOREACH(cpu) {
|
|
|
|
if (pmu) {
|
|
|
|
assert(arm_feature(&ARM_CPU(cpu)->env, ARM_FEATURE_PMU));
|
|
|
|
if (kvm_irqchip_in_kernel()) {
|
|
|
|
kvm_arm_pmu_set_irq(cpu, PPI(VIRTUAL_PMU_IRQ));
|
|
|
|
}
|
|
|
|
kvm_arm_pmu_init(cpu);
|
|
|
|
}
|
2020-10-01 09:17:18 +03:00
|
|
|
if (steal_time) {
|
|
|
|
kvm_arm_pvtime_init(cpu, pvtime_reg_base +
|
|
|
|
cpu->cpu_index * PVTIME_SIZE_PER_CPU);
|
|
|
|
}
|
2020-10-01 09:17:16 +03:00
|
|
|
}
|
|
|
|
} else {
|
2020-10-01 09:17:15 +03:00
|
|
|
if (aarch64 && vms->highmem) {
|
|
|
|
int requested_pa_size = 64 - clz64(vms->highest_gpa);
|
|
|
|
int pamax = arm_pamax(ARM_CPU(first_cpu));
|
|
|
|
|
|
|
|
if (pamax < requested_pa_size) {
|
|
|
|
error_report("VCPU supports less PA bits (%d) than "
|
|
|
|
"requested by the memory map (%d)",
|
|
|
|
pamax, requested_pa_size);
|
|
|
|
exit(1);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2014-05-07 18:42:57 +04:00
|
|
|
static void machvirt_init(MachineState *machine)
|
2013-11-22 21:17:14 +04:00
|
|
|
{
|
2014-12-16 02:09:49 +03:00
|
|
|
VirtMachineState *vms = VIRT_MACHINE(machine);
|
2016-07-14 18:51:37 +03:00
|
|
|
VirtMachineClass *vmc = VIRT_MACHINE_GET_CLASS(machine);
|
2017-05-03 15:56:57 +03:00
|
|
|
MachineClass *mc = MACHINE_GET_CLASS(machine);
|
|
|
|
const CPUArchIdList *possible_cpus;
|
2013-11-22 21:17:14 +04:00
|
|
|
MemoryRegion *sysmem = get_system_memory();
|
2016-01-21 17:15:07 +03:00
|
|
|
MemoryRegion *secure_sysmem = NULL;
|
2020-06-26 06:31:41 +03:00
|
|
|
MemoryRegion *tag_sysmem = NULL;
|
|
|
|
MemoryRegion *secure_tag_sysmem = NULL;
|
2016-02-11 14:17:32 +03:00
|
|
|
int n, virt_max_cpus;
|
hw/arm/virt: Support firmware configuration with -blockdev
The ARM virt machines put firmware in flash memory. To configure it,
you use -drive if=pflash,unit=0,... and optionally -drive
if=pflash,unit=1,...
Why two -drive? This permits setting up one part of the flash memory
read-only, and the other part read/write. It also makes upgrading
firmware on the host easier. Below the hood, we get two separate
flash devices, because we were too lazy to improve our flash device
models to support sector protection.
The problem at hand is to do the same with -blockdev somehow, as one
more step towards deprecating -drive.
We recently solved this problem for x86 PC machines, in commit
ebc29e1beab. See the commit message for design rationale.
This commit solves it for ARM virt basically the same way: new machine
properties pflash0, pflash1 forward to the onboard flash devices'
properties. Requires creating the onboard devices in the
.instance_init() method virt_instance_init(). The existing code to
pick up drives defined with -drive if=pflash is replaced by code to
desugar into the machine properties.
There are a few behavioral differences, though:
* The flash devices are always present (x86: only present if
configured)
* Flash base addresses and sizes are fixed (x86: sizes depend on
images, mapped back to back below a fixed address)
* -bios configures contents of first pflash (x86: -bios configures ROM
contents)
* -bios is rejected when first pflash is also configured with -machine
pflash0=... (x86: bios is silently ignored then)
* -machine pflash1=... does not require -machine pflash0=... (x86: it
does).
The actual code is a bit simpler than for x86 mostly due to the first
two differences.
Before the patch, all the action is in create_flash(), called from the
machine's .init() method machvirt_init():
main()
machine_run_board_init()
machvirt_init()
create_flash()
create_one_flash() for flash[0]
create
configure
includes obeying -drive if=pflash,unit=0
realize
map
fall back to -bios
create_one_flash() for flash[1]
create
configure
includes obeying -drive if=pflash,unit=1
realize
map
update FDT
To make the machine properties work, we need to move device creation
to its .instance_init() method virt_instance_init().
Another complication is machvirt_init()'s computation of
@firmware_loaded: it predicts what create_flash() will do. Instead of
predicting what create_flash()'s replacement virt_firmware_init() will
do, I decided to have virt_firmware_init() return what it did.
Requires calling it a bit earlier.
Resulting call tree:
main()
current_machine = object_new()
...
virt_instance_init()
virt_flash_create()
virt_flash_create1() for flash[0]
create
configure: set defaults
become child of machine [NEW]
add machine prop pflash0 as alias for drive [NEW]
virt_flash_create1() for flash[1]
create
configure: set defaults
become child of machine [NEW]
add machine prop pflash1 as alias for drive [NEW]
for all machine props from the command line: machine_set_property()
...
property_set_alias() for machine props pflash0, pflash1
...
set_drive() for cfi.pflash01 prop drive
this is how -machine pflash0=... etc set
machine_run_board_init(current_machine);
virt_firmware_init()
pflash_cfi01_legacy_drive()
legacy -drive if=pflash,unit=0 and =1 [NEW]
virt_flash_map()
virt_flash_map1() for flash[0]
configure: num-blocks
realize
map
virt_flash_map1() for flash[1]
configure: num-blocks
realize
map
fall back to -bios
virt_flash_fdt()
update FDT
You have László to thank for making me explain this in detail.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Acked-by: Laszlo Ersek <lersek@redhat.com>
Message-id: 20190416091348.26075-4-armbru@redhat.com
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
2019-05-07 14:55:02 +03:00
|
|
|
bool firmware_loaded;
|
2018-06-22 15:28:37 +03:00
|
|
|
bool aarch64 = true;
|
2019-09-18 16:06:27 +03:00
|
|
|
bool has_ged = !vmc->no_ged;
|
2019-05-18 23:54:26 +03:00
|
|
|
unsigned int smp_cpus = machine->smp.cpus;
|
|
|
|
unsigned int max_cpus = machine->smp.max_cpus;
|
2013-11-22 21:17:14 +04:00
|
|
|
|
2019-03-04 13:13:37 +03:00
|
|
|
/*
|
|
|
|
* In accelerated mode, the memory map is computed earlier in kvm_type()
|
|
|
|
* to create a VM with the right number of IPA bits.
|
|
|
|
*/
|
|
|
|
if (!vms->memmap) {
|
|
|
|
virt_set_memmap(vms);
|
|
|
|
}
|
2019-03-04 13:13:32 +03:00
|
|
|
|
2015-09-24 03:29:37 +03:00
|
|
|
/* We can probe only here because during property set
|
|
|
|
* KVM is not available yet
|
|
|
|
*/
|
2020-03-11 16:16:15 +03:00
|
|
|
finalize_gic_version(vms);
|
2015-09-24 03:29:37 +03:00
|
|
|
|
2017-09-13 19:04:57 +03:00
|
|
|
if (!cpu_type_valid(machine->cpu_type)) {
|
|
|
|
error_report("mach-virt: CPU type %s not supported", machine->cpu_type);
|
2013-11-22 21:17:14 +04:00
|
|
|
exit(1);
|
|
|
|
}
|
|
|
|
|
hw/arm/virt: Support firmware configuration with -blockdev
The ARM virt machines put firmware in flash memory. To configure it,
you use -drive if=pflash,unit=0,... and optionally -drive
if=pflash,unit=1,...
Why two -drive? This permits setting up one part of the flash memory
read-only, and the other part read/write. It also makes upgrading
firmware on the host easier. Below the hood, we get two separate
flash devices, because we were too lazy to improve our flash device
models to support sector protection.
The problem at hand is to do the same with -blockdev somehow, as one
more step towards deprecating -drive.
We recently solved this problem for x86 PC machines, in commit
ebc29e1beab. See the commit message for design rationale.
This commit solves it for ARM virt basically the same way: new machine
properties pflash0, pflash1 forward to the onboard flash devices'
properties. Requires creating the onboard devices in the
.instance_init() method virt_instance_init(). The existing code to
pick up drives defined with -drive if=pflash is replaced by code to
desugar into the machine properties.
There are a few behavioral differences, though:
* The flash devices are always present (x86: only present if
configured)
* Flash base addresses and sizes are fixed (x86: sizes depend on
images, mapped back to back below a fixed address)
* -bios configures contents of first pflash (x86: -bios configures ROM
contents)
* -bios is rejected when first pflash is also configured with -machine
pflash0=... (x86: bios is silently ignored then)
* -machine pflash1=... does not require -machine pflash0=... (x86: it
does).
The actual code is a bit simpler than for x86 mostly due to the first
two differences.
Before the patch, all the action is in create_flash(), called from the
machine's .init() method machvirt_init():
main()
machine_run_board_init()
machvirt_init()
create_flash()
create_one_flash() for flash[0]
create
configure
includes obeying -drive if=pflash,unit=0
realize
map
fall back to -bios
create_one_flash() for flash[1]
create
configure
includes obeying -drive if=pflash,unit=1
realize
map
update FDT
To make the machine properties work, we need to move device creation
to its .instance_init() method virt_instance_init().
Another complication is machvirt_init()'s computation of
@firmware_loaded: it predicts what create_flash() will do. Instead of
predicting what create_flash()'s replacement virt_firmware_init() will
do, I decided to have virt_firmware_init() return what it did.
Requires calling it a bit earlier.
Resulting call tree:
main()
current_machine = object_new()
...
virt_instance_init()
virt_flash_create()
virt_flash_create1() for flash[0]
create
configure: set defaults
become child of machine [NEW]
add machine prop pflash0 as alias for drive [NEW]
virt_flash_create1() for flash[1]
create
configure: set defaults
become child of machine [NEW]
add machine prop pflash1 as alias for drive [NEW]
for all machine props from the command line: machine_set_property()
...
property_set_alias() for machine props pflash0, pflash1
...
set_drive() for cfi.pflash01 prop drive
this is how -machine pflash0=... etc set
machine_run_board_init(current_machine);
virt_firmware_init()
pflash_cfi01_legacy_drive()
legacy -drive if=pflash,unit=0 and =1 [NEW]
virt_flash_map()
virt_flash_map1() for flash[0]
configure: num-blocks
realize
map
virt_flash_map1() for flash[1]
configure: num-blocks
realize
map
fall back to -bios
virt_flash_fdt()
update FDT
You have László to thank for making me explain this in detail.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Acked-by: Laszlo Ersek <lersek@redhat.com>
Message-id: 20190416091348.26075-4-armbru@redhat.com
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
2019-05-07 14:55:02 +03:00
|
|
|
if (vms->secure) {
|
|
|
|
/*
|
|
|
|
* The Secure view of the world is the same as the NonSecure,
|
|
|
|
* but with a few extra devices. Create it as a container region
|
|
|
|
* containing the system memory at low priority; any secure-only
|
|
|
|
* devices go in at higher priority and take precedence.
|
|
|
|
*/
|
|
|
|
secure_sysmem = g_new(MemoryRegion, 1);
|
|
|
|
memory_region_init(secure_sysmem, OBJECT(machine), "secure-memory",
|
|
|
|
UINT64_MAX);
|
|
|
|
memory_region_add_subregion_overlap(secure_sysmem, 0, sysmem, -1);
|
|
|
|
}
|
|
|
|
|
|
|
|
firmware_loaded = virt_firmware_init(vms, sysmem,
|
|
|
|
secure_sysmem ?: sysmem);
|
|
|
|
|
2016-03-04 14:30:18 +03:00
|
|
|
/* If we have an EL3 boot ROM then the assumption is that it will
|
|
|
|
* implement PSCI itself, so disable QEMU's internal implementation
|
|
|
|
* so it doesn't get in the way. Instead of starting secondary
|
|
|
|
* CPUs in PSCI powerdown state we will start them all running and
|
|
|
|
* let the boot ROM sort them out.
|
2017-01-20 14:15:11 +03:00
|
|
|
* The usual case is that we do use QEMU's PSCI implementation;
|
|
|
|
* if the guest has EL2 then we will use SMC as the conduit,
|
|
|
|
* and otherwise we will use HVC (for backwards compatibility and
|
|
|
|
* because if we're using KVM then we must use HVC).
|
2016-03-04 14:30:18 +03:00
|
|
|
*/
|
2017-01-20 14:15:10 +03:00
|
|
|
if (vms->secure && firmware_loaded) {
|
|
|
|
vms->psci_conduit = QEMU_PSCI_CONDUIT_DISABLED;
|
2017-01-20 14:15:11 +03:00
|
|
|
} else if (vms->virt) {
|
|
|
|
vms->psci_conduit = QEMU_PSCI_CONDUIT_SMC;
|
2017-01-20 14:15:10 +03:00
|
|
|
} else {
|
|
|
|
vms->psci_conduit = QEMU_PSCI_CONDUIT_HVC;
|
|
|
|
}
|
2016-03-04 14:30:18 +03:00
|
|
|
|
2015-10-27 15:00:50 +03:00
|
|
|
/* The maximum number of CPUs depends on the GIC version, or on how
|
|
|
|
* many redistributors we can fit into the memory map.
|
|
|
|
*/
|
2020-03-11 16:16:14 +03:00
|
|
|
if (vms->gic_version == VIRT_GIC_VERSION_3) {
|
2019-03-04 13:13:31 +03:00
|
|
|
virt_max_cpus =
|
|
|
|
vms->memmap[VIRT_GIC_REDIST].size / GICV3_REDIST_SIZE;
|
|
|
|
virt_max_cpus +=
|
|
|
|
vms->memmap[VIRT_HIGH_GIC_REDIST2].size / GICV3_REDIST_SIZE;
|
2015-10-27 15:00:50 +03:00
|
|
|
} else {
|
2016-02-11 14:17:32 +03:00
|
|
|
virt_max_cpus = GIC_NCPU;
|
2015-10-27 15:00:50 +03:00
|
|
|
}
|
|
|
|
|
2016-02-11 14:17:32 +03:00
|
|
|
if (max_cpus > virt_max_cpus) {
|
2015-10-27 15:00:50 +03:00
|
|
|
error_report("Number of SMP CPUs requested (%d) exceeds max CPUs "
|
|
|
|
"supported by machine 'mach-virt' (%d)",
|
2016-02-11 14:17:32 +03:00
|
|
|
max_cpus, virt_max_cpus);
|
2015-10-27 15:00:50 +03:00
|
|
|
exit(1);
|
|
|
|
}
|
|
|
|
|
2017-01-20 14:15:11 +03:00
|
|
|
if (vms->virt && kvm_enabled()) {
|
|
|
|
error_report("mach-virt: KVM does not support providing "
|
|
|
|
"Virtualization extensions to the guest CPU");
|
|
|
|
exit(1);
|
|
|
|
}
|
|
|
|
|
2020-07-20 12:25:36 +03:00
|
|
|
if (vms->mte && kvm_enabled()) {
|
|
|
|
error_report("mach-virt: KVM does not support providing "
|
|
|
|
"MTE to the guest CPU");
|
|
|
|
exit(1);
|
|
|
|
}
|
|
|
|
|
2017-01-09 14:40:21 +03:00
|
|
|
create_fdt(vms);
|
2013-11-22 21:17:14 +04:00
|
|
|
|
2017-05-03 15:56:57 +03:00
|
|
|
possible_cpus = mc->possible_cpu_arch_ids(machine);
|
2020-12-15 20:48:15 +03:00
|
|
|
assert(possible_cpus->len == max_cpus);
|
2017-05-03 15:56:57 +03:00
|
|
|
for (n = 0; n < possible_cpus->len; n++) {
|
|
|
|
Object *cpuobj;
|
2017-05-03 15:56:58 +03:00
|
|
|
CPUState *cs;
|
2017-05-03 15:56:56 +03:00
|
|
|
|
2017-05-03 15:56:57 +03:00
|
|
|
if (n >= smp_cpus) {
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
2018-01-10 18:22:50 +03:00
|
|
|
cpuobj = object_new(possible_cpus->cpus[n].type);
|
qom: Put name parameter before value / visitor parameter
The object_property_set_FOO() setters take property name and value in
an unusual order:
void object_property_set_FOO(Object *obj, FOO_TYPE value,
const char *name, Error **errp)
Having to pass value before name feels grating. Swap them.
Same for object_property_set(), object_property_get(), and
object_property_parse().
Convert callers with this Coccinelle script:
@@
identifier fun = {
object_property_get, object_property_parse, object_property_set_str,
object_property_set_link, object_property_set_bool,
object_property_set_int, object_property_set_uint, object_property_set,
object_property_set_qobject
};
expression obj, v, name, errp;
@@
- fun(obj, v, name, errp)
+ fun(obj, name, v, errp)
Chokes on hw/arm/musicpal.c's lcd_refresh() with the unhelpful error
message "no position information". Convert that one manually.
Fails to convert hw/arm/armsse.c, because Coccinelle gets confused by
ARMSSE being used both as typedef and function-like macro there.
Convert manually.
Fails to convert hw/rx/rx-gdbsim.c, because Coccinelle gets confused
by RXCPU being used both as typedef and function-like macro there.
Convert manually. The other files using RXCPU that way don't need
conversion.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20200707160613.848843-27-armbru@redhat.com>
[Straightforwad conflict with commit 2336172d9b "audio: set default
value for pcspk.iobase property" resolved]
2020-07-07 19:05:54 +03:00
|
|
|
object_property_set_int(cpuobj, "mp-affinity",
|
|
|
|
possible_cpus->cpus[n].arch_id, NULL);
|
2015-02-13 08:46:08 +03:00
|
|
|
|
2017-05-03 15:56:58 +03:00
|
|
|
cs = CPU(cpuobj);
|
|
|
|
cs->cpu_index = n;
|
|
|
|
|
2017-05-30 19:23:56 +03:00
|
|
|
numa_cpu_pre_plug(&possible_cpus->cpus[cs->cpu_index], DEVICE(cpuobj),
|
|
|
|
&error_fatal);
|
2017-05-10 14:29:48 +03:00
|
|
|
|
2018-06-22 15:28:37 +03:00
|
|
|
aarch64 &= object_property_get_bool(cpuobj, "aarch64", NULL);
|
|
|
|
|
2014-12-16 02:09:49 +03:00
|
|
|
if (!vms->secure) {
|
qom: Put name parameter before value / visitor parameter
The object_property_set_FOO() setters take property name and value in
an unusual order:
void object_property_set_FOO(Object *obj, FOO_TYPE value,
const char *name, Error **errp)
Having to pass value before name feels grating. Swap them.
Same for object_property_set(), object_property_get(), and
object_property_parse().
Convert callers with this Coccinelle script:
@@
identifier fun = {
object_property_get, object_property_parse, object_property_set_str,
object_property_set_link, object_property_set_bool,
object_property_set_int, object_property_set_uint, object_property_set,
object_property_set_qobject
};
expression obj, v, name, errp;
@@
- fun(obj, v, name, errp)
+ fun(obj, name, v, errp)
Chokes on hw/arm/musicpal.c's lcd_refresh() with the unhelpful error
message "no position information". Convert that one manually.
Fails to convert hw/arm/armsse.c, because Coccinelle gets confused by
ARMSSE being used both as typedef and function-like macro there.
Convert manually.
Fails to convert hw/rx/rx-gdbsim.c, because Coccinelle gets confused
by RXCPU being used both as typedef and function-like macro there.
Convert manually. The other files using RXCPU that way don't need
conversion.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20200707160613.848843-27-armbru@redhat.com>
[Straightforwad conflict with commit 2336172d9b "audio: set default
value for pcspk.iobase property" resolved]
2020-07-07 19:05:54 +03:00
|
|
|
object_property_set_bool(cpuobj, "has_el3", false, NULL);
|
2014-12-16 02:09:49 +03:00
|
|
|
}
|
|
|
|
|
2020-09-14 16:56:17 +03:00
|
|
|
if (!vms->virt && object_property_find(cpuobj, "has_el2")) {
|
qom: Put name parameter before value / visitor parameter
The object_property_set_FOO() setters take property name and value in
an unusual order:
void object_property_set_FOO(Object *obj, FOO_TYPE value,
const char *name, Error **errp)
Having to pass value before name feels grating. Swap them.
Same for object_property_set(), object_property_get(), and
object_property_parse().
Convert callers with this Coccinelle script:
@@
identifier fun = {
object_property_get, object_property_parse, object_property_set_str,
object_property_set_link, object_property_set_bool,
object_property_set_int, object_property_set_uint, object_property_set,
object_property_set_qobject
};
expression obj, v, name, errp;
@@
- fun(obj, v, name, errp)
+ fun(obj, name, v, errp)
Chokes on hw/arm/musicpal.c's lcd_refresh() with the unhelpful error
message "no position information". Convert that one manually.
Fails to convert hw/arm/armsse.c, because Coccinelle gets confused by
ARMSSE being used both as typedef and function-like macro there.
Convert manually.
Fails to convert hw/rx/rx-gdbsim.c, because Coccinelle gets confused
by RXCPU being used both as typedef and function-like macro there.
Convert manually. The other files using RXCPU that way don't need
conversion.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20200707160613.848843-27-armbru@redhat.com>
[Straightforwad conflict with commit 2336172d9b "audio: set default
value for pcspk.iobase property" resolved]
2020-07-07 19:05:54 +03:00
|
|
|
object_property_set_bool(cpuobj, "has_el2", false, NULL);
|
2017-01-20 14:15:10 +03:00
|
|
|
}
|
|
|
|
|
2017-01-20 14:15:10 +03:00
|
|
|
if (vms->psci_conduit != QEMU_PSCI_CONDUIT_DISABLED) {
|
qom: Put name parameter before value / visitor parameter
The object_property_set_FOO() setters take property name and value in
an unusual order:
void object_property_set_FOO(Object *obj, FOO_TYPE value,
const char *name, Error **errp)
Having to pass value before name feels grating. Swap them.
Same for object_property_set(), object_property_get(), and
object_property_parse().
Convert callers with this Coccinelle script:
@@
identifier fun = {
object_property_get, object_property_parse, object_property_set_str,
object_property_set_link, object_property_set_bool,
object_property_set_int, object_property_set_uint, object_property_set,
object_property_set_qobject
};
expression obj, v, name, errp;
@@
- fun(obj, v, name, errp)
+ fun(obj, name, v, errp)
Chokes on hw/arm/musicpal.c's lcd_refresh() with the unhelpful error
message "no position information". Convert that one manually.
Fails to convert hw/arm/armsse.c, because Coccinelle gets confused by
ARMSSE being used both as typedef and function-like macro there.
Convert manually.
Fails to convert hw/rx/rx-gdbsim.c, because Coccinelle gets confused
by RXCPU being used both as typedef and function-like macro there.
Convert manually. The other files using RXCPU that way don't need
conversion.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20200707160613.848843-27-armbru@redhat.com>
[Straightforwad conflict with commit 2336172d9b "audio: set default
value for pcspk.iobase property" resolved]
2020-07-07 19:05:54 +03:00
|
|
|
object_property_set_int(cpuobj, "psci-conduit", vms->psci_conduit,
|
|
|
|
NULL);
|
2014-10-24 15:19:13 +04:00
|
|
|
|
2016-03-04 14:30:18 +03:00
|
|
|
/* Secondary CPUs start in PSCI powered-down state */
|
|
|
|
if (n > 0) {
|
qom: Put name parameter before value / visitor parameter
The object_property_set_FOO() setters take property name and value in
an unusual order:
void object_property_set_FOO(Object *obj, FOO_TYPE value,
const char *name, Error **errp)
Having to pass value before name feels grating. Swap them.
Same for object_property_set(), object_property_get(), and
object_property_parse().
Convert callers with this Coccinelle script:
@@
identifier fun = {
object_property_get, object_property_parse, object_property_set_str,
object_property_set_link, object_property_set_bool,
object_property_set_int, object_property_set_uint, object_property_set,
object_property_set_qobject
};
expression obj, v, name, errp;
@@
- fun(obj, v, name, errp)
+ fun(obj, name, v, errp)
Chokes on hw/arm/musicpal.c's lcd_refresh() with the unhelpful error
message "no position information". Convert that one manually.
Fails to convert hw/arm/armsse.c, because Coccinelle gets confused by
ARMSSE being used both as typedef and function-like macro there.
Convert manually.
Fails to convert hw/rx/rx-gdbsim.c, because Coccinelle gets confused
by RXCPU being used both as typedef and function-like macro there.
Convert manually. The other files using RXCPU that way don't need
conversion.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20200707160613.848843-27-armbru@redhat.com>
[Straightforwad conflict with commit 2336172d9b "audio: set default
value for pcspk.iobase property" resolved]
2020-07-07 19:05:54 +03:00
|
|
|
object_property_set_bool(cpuobj, "start-powered-off", true,
|
|
|
|
NULL);
|
2016-03-04 14:30:18 +03:00
|
|
|
}
|
2013-11-22 21:17:14 +04:00
|
|
|
}
|
2014-03-17 20:31:46 +04:00
|
|
|
|
2020-01-30 19:02:06 +03:00
|
|
|
if (vmc->kvm_no_adjvtime &&
|
2020-09-14 16:56:17 +03:00
|
|
|
object_property_find(cpuobj, "kvm-no-adjvtime")) {
|
qom: Put name parameter before value / visitor parameter
The object_property_set_FOO() setters take property name and value in
an unusual order:
void object_property_set_FOO(Object *obj, FOO_TYPE value,
const char *name, Error **errp)
Having to pass value before name feels grating. Swap them.
Same for object_property_set(), object_property_get(), and
object_property_parse().
Convert callers with this Coccinelle script:
@@
identifier fun = {
object_property_get, object_property_parse, object_property_set_str,
object_property_set_link, object_property_set_bool,
object_property_set_int, object_property_set_uint, object_property_set,
object_property_set_qobject
};
expression obj, v, name, errp;
@@
- fun(obj, v, name, errp)
+ fun(obj, name, v, errp)
Chokes on hw/arm/musicpal.c's lcd_refresh() with the unhelpful error
message "no position information". Convert that one manually.
Fails to convert hw/arm/armsse.c, because Coccinelle gets confused by
ARMSSE being used both as typedef and function-like macro there.
Convert manually.
Fails to convert hw/rx/rx-gdbsim.c, because Coccinelle gets confused
by RXCPU being used both as typedef and function-like macro there.
Convert manually. The other files using RXCPU that way don't need
conversion.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20200707160613.848843-27-armbru@redhat.com>
[Straightforwad conflict with commit 2336172d9b "audio: set default
value for pcspk.iobase property" resolved]
2020-07-07 19:05:54 +03:00
|
|
|
object_property_set_bool(cpuobj, "kvm-no-adjvtime", true, NULL);
|
2020-01-30 19:02:06 +03:00
|
|
|
}
|
|
|
|
|
2020-10-01 09:17:18 +03:00
|
|
|
if (vmc->no_kvm_steal_time &&
|
|
|
|
object_property_find(cpuobj, "kvm-steal-time")) {
|
|
|
|
object_property_set_bool(cpuobj, "kvm-steal-time", false, NULL);
|
|
|
|
}
|
|
|
|
|
2020-09-14 16:56:17 +03:00
|
|
|
if (vmc->no_pmu && object_property_find(cpuobj, "pmu")) {
|
qom: Put name parameter before value / visitor parameter
The object_property_set_FOO() setters take property name and value in
an unusual order:
void object_property_set_FOO(Object *obj, FOO_TYPE value,
const char *name, Error **errp)
Having to pass value before name feels grating. Swap them.
Same for object_property_set(), object_property_get(), and
object_property_parse().
Convert callers with this Coccinelle script:
@@
identifier fun = {
object_property_get, object_property_parse, object_property_set_str,
object_property_set_link, object_property_set_bool,
object_property_set_int, object_property_set_uint, object_property_set,
object_property_set_qobject
};
expression obj, v, name, errp;
@@
- fun(obj, v, name, errp)
+ fun(obj, name, v, errp)
Chokes on hw/arm/musicpal.c's lcd_refresh() with the unhelpful error
message "no position information". Convert that one manually.
Fails to convert hw/arm/armsse.c, because Coccinelle gets confused by
ARMSSE being used both as typedef and function-like macro there.
Convert manually.
Fails to convert hw/rx/rx-gdbsim.c, because Coccinelle gets confused
by RXCPU being used both as typedef and function-like macro there.
Convert manually. The other files using RXCPU that way don't need
conversion.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20200707160613.848843-27-armbru@redhat.com>
[Straightforwad conflict with commit 2336172d9b "audio: set default
value for pcspk.iobase property" resolved]
2020-07-07 19:05:54 +03:00
|
|
|
object_property_set_bool(cpuobj, "pmu", false, NULL);
|
2016-10-28 16:12:31 +03:00
|
|
|
}
|
|
|
|
|
2020-09-14 16:56:17 +03:00
|
|
|
if (object_property_find(cpuobj, "reset-cbar")) {
|
qom: Put name parameter before value / visitor parameter
The object_property_set_FOO() setters take property name and value in
an unusual order:
void object_property_set_FOO(Object *obj, FOO_TYPE value,
const char *name, Error **errp)
Having to pass value before name feels grating. Swap them.
Same for object_property_set(), object_property_get(), and
object_property_parse().
Convert callers with this Coccinelle script:
@@
identifier fun = {
object_property_get, object_property_parse, object_property_set_str,
object_property_set_link, object_property_set_bool,
object_property_set_int, object_property_set_uint, object_property_set,
object_property_set_qobject
};
expression obj, v, name, errp;
@@
- fun(obj, v, name, errp)
+ fun(obj, name, v, errp)
Chokes on hw/arm/musicpal.c's lcd_refresh() with the unhelpful error
message "no position information". Convert that one manually.
Fails to convert hw/arm/armsse.c, because Coccinelle gets confused by
ARMSSE being used both as typedef and function-like macro there.
Convert manually.
Fails to convert hw/rx/rx-gdbsim.c, because Coccinelle gets confused
by RXCPU being used both as typedef and function-like macro there.
Convert manually. The other files using RXCPU that way don't need
conversion.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20200707160613.848843-27-armbru@redhat.com>
[Straightforwad conflict with commit 2336172d9b "audio: set default
value for pcspk.iobase property" resolved]
2020-07-07 19:05:54 +03:00
|
|
|
object_property_set_int(cpuobj, "reset-cbar",
|
|
|
|
vms->memmap[VIRT_CPUPERIPHS].base,
|
|
|
|
&error_abort);
|
2014-03-17 20:31:46 +04:00
|
|
|
}
|
|
|
|
|
qom: Put name parameter before value / visitor parameter
The object_property_set_FOO() setters take property name and value in
an unusual order:
void object_property_set_FOO(Object *obj, FOO_TYPE value,
const char *name, Error **errp)
Having to pass value before name feels grating. Swap them.
Same for object_property_set(), object_property_get(), and
object_property_parse().
Convert callers with this Coccinelle script:
@@
identifier fun = {
object_property_get, object_property_parse, object_property_set_str,
object_property_set_link, object_property_set_bool,
object_property_set_int, object_property_set_uint, object_property_set,
object_property_set_qobject
};
expression obj, v, name, errp;
@@
- fun(obj, v, name, errp)
+ fun(obj, name, v, errp)
Chokes on hw/arm/musicpal.c's lcd_refresh() with the unhelpful error
message "no position information". Convert that one manually.
Fails to convert hw/arm/armsse.c, because Coccinelle gets confused by
ARMSSE being used both as typedef and function-like macro there.
Convert manually.
Fails to convert hw/rx/rx-gdbsim.c, because Coccinelle gets confused
by RXCPU being used both as typedef and function-like macro there.
Convert manually. The other files using RXCPU that way don't need
conversion.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20200707160613.848843-27-armbru@redhat.com>
[Straightforwad conflict with commit 2336172d9b "audio: set default
value for pcspk.iobase property" resolved]
2020-07-07 19:05:54 +03:00
|
|
|
object_property_set_link(cpuobj, "memory", OBJECT(sysmem),
|
2016-01-21 17:15:07 +03:00
|
|
|
&error_abort);
|
2016-01-21 17:15:07 +03:00
|
|
|
if (vms->secure) {
|
qom: Put name parameter before value / visitor parameter
The object_property_set_FOO() setters take property name and value in
an unusual order:
void object_property_set_FOO(Object *obj, FOO_TYPE value,
const char *name, Error **errp)
Having to pass value before name feels grating. Swap them.
Same for object_property_set(), object_property_get(), and
object_property_parse().
Convert callers with this Coccinelle script:
@@
identifier fun = {
object_property_get, object_property_parse, object_property_set_str,
object_property_set_link, object_property_set_bool,
object_property_set_int, object_property_set_uint, object_property_set,
object_property_set_qobject
};
expression obj, v, name, errp;
@@
- fun(obj, v, name, errp)
+ fun(obj, name, v, errp)
Chokes on hw/arm/musicpal.c's lcd_refresh() with the unhelpful error
message "no position information". Convert that one manually.
Fails to convert hw/arm/armsse.c, because Coccinelle gets confused by
ARMSSE being used both as typedef and function-like macro there.
Convert manually.
Fails to convert hw/rx/rx-gdbsim.c, because Coccinelle gets confused
by RXCPU being used both as typedef and function-like macro there.
Convert manually. The other files using RXCPU that way don't need
conversion.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20200707160613.848843-27-armbru@redhat.com>
[Straightforwad conflict with commit 2336172d9b "audio: set default
value for pcspk.iobase property" resolved]
2020-07-07 19:05:54 +03:00
|
|
|
object_property_set_link(cpuobj, "secure-memory",
|
|
|
|
OBJECT(secure_sysmem), &error_abort);
|
2016-01-21 17:15:07 +03:00
|
|
|
}
|
2016-01-21 17:15:07 +03:00
|
|
|
|
2020-07-20 12:25:36 +03:00
|
|
|
if (vms->mte) {
|
|
|
|
/* Create the memory region only once, but link to all cpus. */
|
2020-06-26 06:31:41 +03:00
|
|
|
if (!tag_sysmem) {
|
2020-07-20 12:25:36 +03:00
|
|
|
/*
|
|
|
|
* The property exists only if MemTag is supported.
|
|
|
|
* If it is, we must allocate the ram to back that up.
|
|
|
|
*/
|
2020-09-14 16:56:17 +03:00
|
|
|
if (!object_property_find(cpuobj, "tag-memory")) {
|
2020-07-20 12:25:36 +03:00
|
|
|
error_report("MTE requested, but not supported "
|
|
|
|
"by the guest CPU");
|
|
|
|
exit(1);
|
|
|
|
}
|
|
|
|
|
2020-06-26 06:31:41 +03:00
|
|
|
tag_sysmem = g_new(MemoryRegion, 1);
|
|
|
|
memory_region_init(tag_sysmem, OBJECT(machine),
|
|
|
|
"tag-memory", UINT64_MAX / 32);
|
|
|
|
|
|
|
|
if (vms->secure) {
|
|
|
|
secure_tag_sysmem = g_new(MemoryRegion, 1);
|
|
|
|
memory_region_init(secure_tag_sysmem, OBJECT(machine),
|
|
|
|
"secure-tag-memory", UINT64_MAX / 32);
|
|
|
|
|
|
|
|
/* As with ram, secure-tag takes precedence over tag. */
|
|
|
|
memory_region_add_subregion_overlap(secure_tag_sysmem, 0,
|
|
|
|
tag_sysmem, -1);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
qom: Put name parameter before value / visitor parameter
The object_property_set_FOO() setters take property name and value in
an unusual order:
void object_property_set_FOO(Object *obj, FOO_TYPE value,
const char *name, Error **errp)
Having to pass value before name feels grating. Swap them.
Same for object_property_set(), object_property_get(), and
object_property_parse().
Convert callers with this Coccinelle script:
@@
identifier fun = {
object_property_get, object_property_parse, object_property_set_str,
object_property_set_link, object_property_set_bool,
object_property_set_int, object_property_set_uint, object_property_set,
object_property_set_qobject
};
expression obj, v, name, errp;
@@
- fun(obj, v, name, errp)
+ fun(obj, name, v, errp)
Chokes on hw/arm/musicpal.c's lcd_refresh() with the unhelpful error
message "no position information". Convert that one manually.
Fails to convert hw/arm/armsse.c, because Coccinelle gets confused by
ARMSSE being used both as typedef and function-like macro there.
Convert manually.
Fails to convert hw/rx/rx-gdbsim.c, because Coccinelle gets confused
by RXCPU being used both as typedef and function-like macro there.
Convert manually. The other files using RXCPU that way don't need
conversion.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20200707160613.848843-27-armbru@redhat.com>
[Straightforwad conflict with commit 2336172d9b "audio: set default
value for pcspk.iobase property" resolved]
2020-07-07 19:05:54 +03:00
|
|
|
object_property_set_link(cpuobj, "tag-memory", OBJECT(tag_sysmem),
|
|
|
|
&error_abort);
|
2020-06-26 06:31:41 +03:00
|
|
|
if (vms->secure) {
|
qom: Put name parameter before value / visitor parameter
The object_property_set_FOO() setters take property name and value in
an unusual order:
void object_property_set_FOO(Object *obj, FOO_TYPE value,
const char *name, Error **errp)
Having to pass value before name feels grating. Swap them.
Same for object_property_set(), object_property_get(), and
object_property_parse().
Convert callers with this Coccinelle script:
@@
identifier fun = {
object_property_get, object_property_parse, object_property_set_str,
object_property_set_link, object_property_set_bool,
object_property_set_int, object_property_set_uint, object_property_set,
object_property_set_qobject
};
expression obj, v, name, errp;
@@
- fun(obj, v, name, errp)
+ fun(obj, name, v, errp)
Chokes on hw/arm/musicpal.c's lcd_refresh() with the unhelpful error
message "no position information". Convert that one manually.
Fails to convert hw/arm/armsse.c, because Coccinelle gets confused by
ARMSSE being used both as typedef and function-like macro there.
Convert manually.
Fails to convert hw/rx/rx-gdbsim.c, because Coccinelle gets confused
by RXCPU being used both as typedef and function-like macro there.
Convert manually. The other files using RXCPU that way don't need
conversion.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20200707160613.848843-27-armbru@redhat.com>
[Straightforwad conflict with commit 2336172d9b "audio: set default
value for pcspk.iobase property" resolved]
2020-07-07 19:05:54 +03:00
|
|
|
object_property_set_link(cpuobj, "secure-tag-memory",
|
|
|
|
OBJECT(secure_tag_sysmem),
|
|
|
|
&error_abort);
|
2020-06-26 06:31:41 +03:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
qdev: Convert bus-less devices to qdev_realize() with Coccinelle
All remaining conversions to qdev_realize() are for bus-less devices.
Coccinelle script:
// only correct for bus-less @dev!
@@
expression errp;
expression dev;
@@
- qdev_init_nofail(dev);
+ qdev_realize(dev, NULL, &error_fatal);
@ depends on !(file in "hw/core/qdev.c") && !(file in "hw/core/bus.c")@
expression errp;
expression dev;
symbol true;
@@
- object_property_set_bool(OBJECT(dev), true, "realized", errp);
+ qdev_realize(DEVICE(dev), NULL, errp);
@ depends on !(file in "hw/core/qdev.c") && !(file in "hw/core/bus.c")@
expression errp;
expression dev;
symbol true;
@@
- object_property_set_bool(dev, true, "realized", errp);
+ qdev_realize(DEVICE(dev), NULL, errp);
Note that Coccinelle chokes on ARMSSE typedef vs. macro in
hw/arm/armsse.c. Worked around by temporarily renaming the macro for
the spatch run.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Acked-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20200610053247.1583243-57-armbru@redhat.com>
2020-06-10 08:32:45 +03:00
|
|
|
qdev_realize(DEVICE(cpuobj), NULL, &error_fatal);
|
2017-02-28 15:08:15 +03:00
|
|
|
object_unref(cpuobj);
|
2013-11-22 21:17:14 +04:00
|
|
|
}
|
2017-01-09 14:40:22 +03:00
|
|
|
fdt_add_timer_nodes(vms);
|
2017-01-09 14:40:21 +03:00
|
|
|
fdt_add_cpu_nodes(vms);
|
2013-11-22 21:17:14 +04:00
|
|
|
|
2020-02-19 19:09:07 +03:00
|
|
|
memory_region_add_subregion(sysmem, vms->memmap[VIRT_MEM].base,
|
|
|
|
machine->ram);
|
2019-03-04 13:13:36 +03:00
|
|
|
if (machine->device_memory) {
|
|
|
|
memory_region_add_subregion(sysmem, machine->device_memory->base,
|
|
|
|
&machine->device_memory->mr);
|
|
|
|
}
|
2013-11-22 21:17:14 +04:00
|
|
|
|
2019-07-15 16:17:04 +03:00
|
|
|
virt_flash_fdt(vms, sysmem, secure_sysmem ?: sysmem);
|
2014-09-12 17:06:48 +04:00
|
|
|
|
2021-09-13 18:07:24 +03:00
|
|
|
create_gic(vms, sysmem);
|
2013-11-22 21:17:14 +04:00
|
|
|
|
2020-12-15 20:48:15 +03:00
|
|
|
virt_cpu_post_init(vms, sysmem);
|
2020-10-01 09:17:15 +03:00
|
|
|
|
2017-01-09 14:40:22 +03:00
|
|
|
fdt_add_pmu_nodes(vms);
|
2016-06-14 17:59:12 +03:00
|
|
|
|
2019-12-09 12:03:06 +03:00
|
|
|
create_uart(vms, VIRT_UART, sysmem, serial_hd(0));
|
2016-01-21 17:15:07 +03:00
|
|
|
|
|
|
|
if (vms->secure) {
|
2020-06-26 06:31:41 +03:00
|
|
|
create_secure_ram(vms, secure_sysmem, secure_tag_sysmem);
|
2019-12-09 12:03:06 +03:00
|
|
|
create_uart(vms, VIRT_SECURE_UART, secure_sysmem, serial_hd(1));
|
2016-01-21 17:15:07 +03:00
|
|
|
}
|
2013-11-22 21:17:14 +04:00
|
|
|
|
2020-06-26 06:31:41 +03:00
|
|
|
if (tag_sysmem) {
|
|
|
|
create_tag_ram(tag_sysmem, vms->memmap[VIRT_MEM].base,
|
|
|
|
machine->ram_size, "mach-virt.tag");
|
|
|
|
}
|
|
|
|
|
2018-06-22 15:28:37 +03:00
|
|
|
vms->highmem_ecam &= vms->highmem && (!firmware_loaded || aarch64);
|
|
|
|
|
2019-12-09 12:03:06 +03:00
|
|
|
create_rtc(vms);
|
2014-06-29 21:38:39 +04:00
|
|
|
|
2019-12-09 12:03:06 +03:00
|
|
|
create_pcie(vms);
|
2015-02-13 08:46:08 +03:00
|
|
|
|
2020-03-20 13:01:36 +03:00
|
|
|
if (has_ged && aarch64 && firmware_loaded && virt_is_acpi_enabled(vms)) {
|
2019-12-09 12:03:06 +03:00
|
|
|
vms->acpi_dev = create_acpi_ged(vms);
|
2019-09-18 16:06:30 +03:00
|
|
|
} else {
|
2021-01-28 15:00:10 +03:00
|
|
|
create_gpio_devices(vms, VIRT_GPIO, sysmem);
|
2019-09-18 16:06:27 +03:00
|
|
|
}
|
|
|
|
|
2021-01-28 15:00:11 +03:00
|
|
|
if (vms->secure && !vmc->no_secure_gpio) {
|
|
|
|
create_gpio_devices(vms, VIRT_SECURE_GPIO, secure_sysmem);
|
|
|
|
}
|
|
|
|
|
2019-09-18 16:06:29 +03:00
|
|
|
/* connect powerdown request */
|
|
|
|
vms->powerdown_notifier.notify = virt_powerdown_req;
|
|
|
|
qemu_register_powerdown_notifier(&vms->powerdown_notifier);
|
|
|
|
|
2013-11-22 21:17:14 +04:00
|
|
|
/* Create mmio transports, so the user can create virtio backends
|
|
|
|
* (which will be automatically plugged in to the transports). If
|
|
|
|
* no backend is created the transport will just sit harmlessly idle.
|
|
|
|
*/
|
2019-12-09 12:03:06 +03:00
|
|
|
create_virtio_devices(vms);
|
2013-11-22 21:17:14 +04:00
|
|
|
|
2017-01-09 14:40:23 +03:00
|
|
|
vms->fw_cfg = create_fw_cfg(vms, &address_space_memory);
|
|
|
|
rom_set_fw(vms->fw_cfg);
|
2015-05-29 13:28:59 +03:00
|
|
|
|
2019-12-09 12:03:06 +03:00
|
|
|
create_platform_bus(vms);
|
arm: add fw_cfg to "virt" board
fw_cfg already supports exposure over MMIO (used in ppc/mac_newworld.c,
ppc/mac_oldworld.c, sparc/sun4m.c); we can easily add it to the "virt"
board.
Because MMIO access is slow on ARM KVM, we enable the guest, with
fw_cfg_init_mem_wide(), to transfer up to 8 bytes with a single access.
This has been measured to speed up transfers up to 7.5-fold, relative to
single byte data access, on both ARM KVM and x86_64 TCG.
The MMIO register block of fw_cfg is advertized in the device tree. As
base address we pick 0x09020000, which conforms to the comment preceding
"a15memmap": it falls in the miscellaneous device I/O range 128MB..256MB,
and it is aligned at 64KB. The DTB properties follow the documentation in
the Linux source file "Documentation/devicetree/bindings/arm/fw-cfg.txt".
fw_cfg automatically exports a number of files to the guest; for example,
"bootorder" (see fw_cfg_machine_reset()).
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-id: 1419250305-31062-9-git-send-email-pbonzini@redhat.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
2014-12-22 15:11:42 +03:00
|
|
|
|
2020-04-21 15:59:30 +03:00
|
|
|
if (machine->nvdimms_state->is_enabled) {
|
|
|
|
const struct AcpiGenericAddress arm_virt_nvdimm_acpi_dsmio = {
|
|
|
|
.space_id = AML_AS_SYSTEM_MEMORY,
|
|
|
|
.address = vms->memmap[VIRT_NVDIMM_ACPI].base,
|
|
|
|
.bit_width = NVDIMM_ACPI_IO_LEN << 3
|
|
|
|
};
|
|
|
|
|
|
|
|
nvdimm_init_acpi_state(machine->nvdimms_state, sysmem,
|
|
|
|
arm_virt_nvdimm_acpi_dsmio,
|
|
|
|
vms->fw_cfg, OBJECT(vms));
|
|
|
|
}
|
|
|
|
|
2017-01-09 14:40:21 +03:00
|
|
|
vms->bootinfo.ram_size = machine->ram_size;
|
|
|
|
vms->bootinfo.nb_cpus = smp_cpus;
|
|
|
|
vms->bootinfo.board_id = -1;
|
|
|
|
vms->bootinfo.loader_start = vms->memmap[VIRT_MEM].base;
|
|
|
|
vms->bootinfo.get_dtb = machvirt_dtb;
|
2018-05-10 20:10:56 +03:00
|
|
|
vms->bootinfo.skip_dtb_autoload = true;
|
2017-01-09 14:40:21 +03:00
|
|
|
vms->bootinfo.firmware_loaded = firmware_loaded;
|
2019-08-09 09:57:21 +03:00
|
|
|
arm_load_kernel(ARM_CPU(first_cpu), machine, &vms->bootinfo);
|
2015-06-02 14:29:13 +03:00
|
|
|
|
2018-05-10 20:10:56 +03:00
|
|
|
vms->machine_done.notify = virt_machine_done;
|
|
|
|
qemu_add_machine_init_done_notifier(&vms->machine_done);
|
2013-11-22 21:17:14 +04:00
|
|
|
}
|
|
|
|
|
2014-12-16 02:09:44 +03:00
|
|
|
static bool virt_get_secure(Object *obj, Error **errp)
|
|
|
|
{
|
|
|
|
VirtMachineState *vms = VIRT_MACHINE(obj);
|
|
|
|
|
|
|
|
return vms->secure;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void virt_set_secure(Object *obj, bool value, Error **errp)
|
|
|
|
{
|
|
|
|
VirtMachineState *vms = VIRT_MACHINE(obj);
|
|
|
|
|
|
|
|
vms->secure = value;
|
|
|
|
}
|
|
|
|
|
2017-01-20 14:15:11 +03:00
|
|
|
static bool virt_get_virt(Object *obj, Error **errp)
|
|
|
|
{
|
|
|
|
VirtMachineState *vms = VIRT_MACHINE(obj);
|
|
|
|
|
|
|
|
return vms->virt;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void virt_set_virt(Object *obj, bool value, Error **errp)
|
|
|
|
{
|
|
|
|
VirtMachineState *vms = VIRT_MACHINE(obj);
|
|
|
|
|
|
|
|
vms->virt = value;
|
|
|
|
}
|
|
|
|
|
2015-09-07 12:39:29 +03:00
|
|
|
static bool virt_get_highmem(Object *obj, Error **errp)
|
|
|
|
{
|
|
|
|
VirtMachineState *vms = VIRT_MACHINE(obj);
|
|
|
|
|
|
|
|
return vms->highmem;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void virt_set_highmem(Object *obj, bool value, Error **errp)
|
|
|
|
{
|
|
|
|
VirtMachineState *vms = VIRT_MACHINE(obj);
|
|
|
|
|
|
|
|
vms->highmem = value;
|
|
|
|
}
|
|
|
|
|
2017-02-28 15:08:16 +03:00
|
|
|
static bool virt_get_its(Object *obj, Error **errp)
|
|
|
|
{
|
|
|
|
VirtMachineState *vms = VIRT_MACHINE(obj);
|
|
|
|
|
|
|
|
return vms->its;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void virt_set_its(Object *obj, bool value, Error **errp)
|
|
|
|
{
|
|
|
|
VirtMachineState *vms = VIRT_MACHINE(obj);
|
|
|
|
|
|
|
|
vms->its = value;
|
|
|
|
}
|
|
|
|
|
2021-01-19 03:32:13 +03:00
|
|
|
static char *virt_get_oem_id(Object *obj, Error **errp)
|
|
|
|
{
|
|
|
|
VirtMachineState *vms = VIRT_MACHINE(obj);
|
|
|
|
|
|
|
|
return g_strdup(vms->oem_id);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void virt_set_oem_id(Object *obj, const char *value, Error **errp)
|
|
|
|
{
|
|
|
|
VirtMachineState *vms = VIRT_MACHINE(obj);
|
|
|
|
size_t len = strlen(value);
|
|
|
|
|
|
|
|
if (len > 6) {
|
|
|
|
error_setg(errp,
|
|
|
|
"User specified oem-id value is bigger than 6 bytes in size");
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
2021-02-03 01:52:53 +03:00
|
|
|
strncpy(vms->oem_id, value, 6);
|
2021-01-19 03:32:13 +03:00
|
|
|
}
|
|
|
|
|
|
|
|
static char *virt_get_oem_table_id(Object *obj, Error **errp)
|
|
|
|
{
|
|
|
|
VirtMachineState *vms = VIRT_MACHINE(obj);
|
|
|
|
|
|
|
|
return g_strdup(vms->oem_table_id);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void virt_set_oem_table_id(Object *obj, const char *value,
|
|
|
|
Error **errp)
|
|
|
|
{
|
|
|
|
VirtMachineState *vms = VIRT_MACHINE(obj);
|
|
|
|
size_t len = strlen(value);
|
|
|
|
|
|
|
|
if (len > 8) {
|
|
|
|
error_setg(errp,
|
|
|
|
"User specified oem-table-id value is bigger than 8 bytes in size");
|
|
|
|
return;
|
|
|
|
}
|
2021-02-03 01:52:53 +03:00
|
|
|
strncpy(vms->oem_table_id, value, 8);
|
2021-01-19 03:32:13 +03:00
|
|
|
}
|
|
|
|
|
|
|
|
|
2020-03-20 13:01:36 +03:00
|
|
|
bool virt_is_acpi_enabled(VirtMachineState *vms)
|
|
|
|
{
|
|
|
|
if (vms->acpi == ON_OFF_AUTO_OFF) {
|
|
|
|
return false;
|
|
|
|
}
|
|
|
|
return true;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void virt_get_acpi(Object *obj, Visitor *v, const char *name,
|
|
|
|
void *opaque, Error **errp)
|
|
|
|
{
|
|
|
|
VirtMachineState *vms = VIRT_MACHINE(obj);
|
|
|
|
OnOffAuto acpi = vms->acpi;
|
|
|
|
|
|
|
|
visit_type_OnOffAuto(v, name, &acpi, errp);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void virt_set_acpi(Object *obj, Visitor *v, const char *name,
|
|
|
|
void *opaque, Error **errp)
|
|
|
|
{
|
|
|
|
VirtMachineState *vms = VIRT_MACHINE(obj);
|
|
|
|
|
|
|
|
visit_type_OnOffAuto(v, name, &vms->acpi, errp);
|
|
|
|
}
|
|
|
|
|
2020-05-12 06:06:01 +03:00
|
|
|
static bool virt_get_ras(Object *obj, Error **errp)
|
|
|
|
{
|
|
|
|
VirtMachineState *vms = VIRT_MACHINE(obj);
|
|
|
|
|
|
|
|
return vms->ras;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void virt_set_ras(Object *obj, bool value, Error **errp)
|
|
|
|
{
|
|
|
|
VirtMachineState *vms = VIRT_MACHINE(obj);
|
|
|
|
|
|
|
|
vms->ras = value;
|
|
|
|
}
|
|
|
|
|
2020-07-20 12:25:36 +03:00
|
|
|
static bool virt_get_mte(Object *obj, Error **errp)
|
|
|
|
{
|
|
|
|
VirtMachineState *vms = VIRT_MACHINE(obj);
|
|
|
|
|
|
|
|
return vms->mte;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void virt_set_mte(Object *obj, bool value, Error **errp)
|
|
|
|
{
|
|
|
|
VirtMachineState *vms = VIRT_MACHINE(obj);
|
|
|
|
|
|
|
|
vms->mte = value;
|
|
|
|
}
|
|
|
|
|
2015-09-24 03:29:37 +03:00
|
|
|
static char *virt_get_gic_version(Object *obj, Error **errp)
|
|
|
|
{
|
|
|
|
VirtMachineState *vms = VIRT_MACHINE(obj);
|
2020-03-11 16:16:14 +03:00
|
|
|
const char *val = vms->gic_version == VIRT_GIC_VERSION_3 ? "3" : "2";
|
2015-09-24 03:29:37 +03:00
|
|
|
|
|
|
|
return g_strdup(val);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void virt_set_gic_version(Object *obj, const char *value, Error **errp)
|
|
|
|
{
|
|
|
|
VirtMachineState *vms = VIRT_MACHINE(obj);
|
|
|
|
|
|
|
|
if (!strcmp(value, "3")) {
|
2020-03-11 16:16:14 +03:00
|
|
|
vms->gic_version = VIRT_GIC_VERSION_3;
|
2015-09-24 03:29:37 +03:00
|
|
|
} else if (!strcmp(value, "2")) {
|
2020-03-11 16:16:14 +03:00
|
|
|
vms->gic_version = VIRT_GIC_VERSION_2;
|
2015-09-24 03:29:37 +03:00
|
|
|
} else if (!strcmp(value, "host")) {
|
2020-03-11 16:16:14 +03:00
|
|
|
vms->gic_version = VIRT_GIC_VERSION_HOST; /* Will probe later */
|
2018-03-09 20:09:44 +03:00
|
|
|
} else if (!strcmp(value, "max")) {
|
2020-03-11 16:16:14 +03:00
|
|
|
vms->gic_version = VIRT_GIC_VERSION_MAX; /* Will probe later */
|
2015-09-24 03:29:37 +03:00
|
|
|
} else {
|
2015-12-17 19:35:15 +03:00
|
|
|
error_setg(errp, "Invalid gic-version value");
|
2018-03-09 20:09:44 +03:00
|
|
|
error_append_hint(errp, "Valid values are 3, 2, host, max.\n");
|
2015-09-24 03:29:37 +03:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2018-05-04 20:05:52 +03:00
|
|
|
static char *virt_get_iommu(Object *obj, Error **errp)
|
|
|
|
{
|
|
|
|
VirtMachineState *vms = VIRT_MACHINE(obj);
|
|
|
|
|
|
|
|
switch (vms->iommu) {
|
|
|
|
case VIRT_IOMMU_NONE:
|
|
|
|
return g_strdup("none");
|
|
|
|
case VIRT_IOMMU_SMMUV3:
|
|
|
|
return g_strdup("smmuv3");
|
|
|
|
default:
|
|
|
|
g_assert_not_reached();
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static void virt_set_iommu(Object *obj, const char *value, Error **errp)
|
|
|
|
{
|
|
|
|
VirtMachineState *vms = VIRT_MACHINE(obj);
|
|
|
|
|
|
|
|
if (!strcmp(value, "smmuv3")) {
|
|
|
|
vms->iommu = VIRT_IOMMU_SMMUV3;
|
|
|
|
} else if (!strcmp(value, "none")) {
|
|
|
|
vms->iommu = VIRT_IOMMU_NONE;
|
|
|
|
} else {
|
|
|
|
error_setg(errp, "Invalid iommu value");
|
|
|
|
error_append_hint(errp, "Valid values are none, smmuv3.\n");
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2021-07-08 15:55:13 +03:00
|
|
|
static bool virt_get_default_bus_bypass_iommu(Object *obj, Error **errp)
|
|
|
|
{
|
|
|
|
VirtMachineState *vms = VIRT_MACHINE(obj);
|
|
|
|
|
|
|
|
return vms->default_bus_bypass_iommu;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void virt_set_default_bus_bypass_iommu(Object *obj, bool value,
|
|
|
|
Error **errp)
|
|
|
|
{
|
|
|
|
VirtMachineState *vms = VIRT_MACHINE(obj);
|
|
|
|
|
|
|
|
vms->default_bus_bypass_iommu = value;
|
|
|
|
}
|
|
|
|
|
2017-05-10 14:29:45 +03:00
|
|
|
static CpuInstanceProperties
|
|
|
|
virt_cpu_index_to_props(MachineState *ms, unsigned cpu_index)
|
|
|
|
{
|
|
|
|
MachineClass *mc = MACHINE_GET_CLASS(ms);
|
|
|
|
const CPUArchIdList *possible_cpus = mc->possible_cpu_arch_ids(ms);
|
|
|
|
|
|
|
|
assert(cpu_index < possible_cpus->len);
|
|
|
|
return possible_cpus->cpus[cpu_index].props;
|
|
|
|
}
|
|
|
|
|
2017-06-01 13:53:28 +03:00
|
|
|
static int64_t virt_get_default_cpu_node_id(const MachineState *ms, int idx)
|
|
|
|
{
|
2019-08-09 09:57:22 +03:00
|
|
|
return idx % ms->numa_state->num_nodes;
|
2017-06-01 13:53:28 +03:00
|
|
|
}
|
|
|
|
|
2017-05-03 15:56:57 +03:00
|
|
|
static const CPUArchIdList *virt_possible_cpu_arch_ids(MachineState *ms)
|
|
|
|
{
|
|
|
|
int n;
|
2019-05-18 23:54:26 +03:00
|
|
|
unsigned int max_cpus = ms->smp.max_cpus;
|
2017-05-03 15:56:57 +03:00
|
|
|
VirtMachineState *vms = VIRT_MACHINE(ms);
|
|
|
|
|
|
|
|
if (ms->possible_cpus) {
|
|
|
|
assert(ms->possible_cpus->len == max_cpus);
|
|
|
|
return ms->possible_cpus;
|
|
|
|
}
|
|
|
|
|
|
|
|
ms->possible_cpus = g_malloc0(sizeof(CPUArchIdList) +
|
|
|
|
sizeof(CPUArchId) * max_cpus);
|
|
|
|
ms->possible_cpus->len = max_cpus;
|
|
|
|
for (n = 0; n < ms->possible_cpus->len; n++) {
|
2018-01-10 18:22:50 +03:00
|
|
|
ms->possible_cpus->cpus[n].type = ms->cpu_type;
|
2017-05-03 15:56:57 +03:00
|
|
|
ms->possible_cpus->cpus[n].arch_id =
|
|
|
|
virt_cpu_mp_affinity(vms, n);
|
|
|
|
ms->possible_cpus->cpus[n].props.has_thread_id = true;
|
|
|
|
ms->possible_cpus->cpus[n].props.thread_id = n;
|
|
|
|
}
|
|
|
|
return ms->possible_cpus;
|
|
|
|
}
|
|
|
|
|
2019-09-18 16:06:26 +03:00
|
|
|
static void virt_memory_pre_plug(HotplugHandler *hotplug_dev, DeviceState *dev,
|
|
|
|
Error **errp)
|
|
|
|
{
|
2019-09-18 16:06:27 +03:00
|
|
|
VirtMachineState *vms = VIRT_MACHINE(hotplug_dev);
|
2020-04-21 15:59:31 +03:00
|
|
|
const MachineState *ms = MACHINE(hotplug_dev);
|
2019-09-18 16:06:27 +03:00
|
|
|
const bool is_nvdimm = object_dynamic_cast(OBJECT(dev), TYPE_NVDIMM);
|
2019-09-18 16:06:26 +03:00
|
|
|
|
2019-09-18 16:06:27 +03:00
|
|
|
if (!vms->acpi_dev) {
|
|
|
|
error_setg(errp,
|
|
|
|
"memory hotplug is not enabled: missing acpi-ged device");
|
2019-09-18 16:06:26 +03:00
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
2020-07-20 12:25:36 +03:00
|
|
|
if (vms->mte) {
|
|
|
|
error_setg(errp, "memory hotplug is not enabled: MTE is enabled");
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
2020-04-21 15:59:31 +03:00
|
|
|
if (is_nvdimm && !ms->nvdimms_state->is_enabled) {
|
|
|
|
error_setg(errp, "nvdimm is not enabled: add 'nvdimm=on' to '-M'");
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
2019-09-18 16:06:26 +03:00
|
|
|
pc_dimm_pre_plug(PC_DIMM(dev), MACHINE(hotplug_dev), NULL, errp);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void virt_memory_plug(HotplugHandler *hotplug_dev,
|
|
|
|
DeviceState *dev, Error **errp)
|
|
|
|
{
|
|
|
|
VirtMachineState *vms = VIRT_MACHINE(hotplug_dev);
|
2020-04-21 15:59:30 +03:00
|
|
|
MachineState *ms = MACHINE(hotplug_dev);
|
|
|
|
bool is_nvdimm = object_dynamic_cast(OBJECT(dev), TYPE_NVDIMM);
|
2019-09-18 16:06:26 +03:00
|
|
|
|
2020-10-19 11:48:04 +03:00
|
|
|
pc_dimm_plug(PC_DIMM(dev), MACHINE(vms));
|
2019-09-18 16:06:26 +03:00
|
|
|
|
2020-04-21 15:59:30 +03:00
|
|
|
if (is_nvdimm) {
|
|
|
|
nvdimm_plug(ms->nvdimms_state);
|
|
|
|
}
|
|
|
|
|
2020-01-23 18:22:39 +03:00
|
|
|
hotplug_handler_plug(HOTPLUG_HANDLER(vms->acpi_dev),
|
|
|
|
dev, &error_abort);
|
2019-09-18 16:06:26 +03:00
|
|
|
}
|
|
|
|
|
|
|
|
static void virt_machine_device_pre_plug_cb(HotplugHandler *hotplug_dev,
|
|
|
|
DeviceState *dev, Error **errp)
|
|
|
|
{
|
2020-07-03 18:59:42 +03:00
|
|
|
VirtMachineState *vms = VIRT_MACHINE(hotplug_dev);
|
|
|
|
|
2019-09-18 16:06:26 +03:00
|
|
|
if (object_dynamic_cast(OBJECT(dev), TYPE_PC_DIMM)) {
|
|
|
|
virt_memory_pre_plug(hotplug_dev, dev, errp);
|
2020-07-03 18:59:42 +03:00
|
|
|
} else if (object_dynamic_cast(OBJECT(dev), TYPE_VIRTIO_IOMMU_PCI)) {
|
|
|
|
hwaddr db_start = 0, db_end = 0;
|
|
|
|
char *resv_prop_str;
|
|
|
|
|
|
|
|
switch (vms->msi_controller) {
|
|
|
|
case VIRT_MSI_CTRL_NONE:
|
|
|
|
return;
|
|
|
|
case VIRT_MSI_CTRL_ITS:
|
|
|
|
/* GITS_TRANSLATER page */
|
|
|
|
db_start = base_memmap[VIRT_GIC_ITS].base + 0x10000;
|
|
|
|
db_end = base_memmap[VIRT_GIC_ITS].base +
|
|
|
|
base_memmap[VIRT_GIC_ITS].size - 1;
|
|
|
|
break;
|
|
|
|
case VIRT_MSI_CTRL_GICV2M:
|
|
|
|
/* MSI_SETSPI_NS page */
|
|
|
|
db_start = base_memmap[VIRT_GIC_V2M].base;
|
|
|
|
db_end = db_start + base_memmap[VIRT_GIC_V2M].size - 1;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
resv_prop_str = g_strdup_printf("0x%"PRIx64":0x%"PRIx64":%u",
|
|
|
|
db_start, db_end,
|
|
|
|
VIRTIO_IOMMU_RESV_MEM_T_MSI);
|
|
|
|
|
|
|
|
qdev_prop_set_uint32(dev, "len-reserved-regions", 1);
|
|
|
|
qdev_prop_set_string(dev, "reserved-regions[0]", resv_prop_str);
|
|
|
|
g_free(resv_prop_str);
|
2019-09-18 16:06:26 +03:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2018-05-10 20:10:56 +03:00
|
|
|
static void virt_machine_device_plug_cb(HotplugHandler *hotplug_dev,
|
|
|
|
DeviceState *dev, Error **errp)
|
|
|
|
{
|
|
|
|
VirtMachineState *vms = VIRT_MACHINE(hotplug_dev);
|
|
|
|
|
|
|
|
if (vms->platform_bus_dev) {
|
2021-03-25 18:33:09 +03:00
|
|
|
MachineClass *mc = MACHINE_GET_CLASS(vms);
|
|
|
|
|
|
|
|
if (device_is_dynamic_sysbus(mc, dev)) {
|
2018-05-10 20:10:56 +03:00
|
|
|
platform_bus_link_device(PLATFORM_BUS_DEVICE(vms->platform_bus_dev),
|
|
|
|
SYS_BUS_DEVICE(dev));
|
|
|
|
}
|
|
|
|
}
|
2019-09-18 16:06:26 +03:00
|
|
|
if (object_dynamic_cast(OBJECT(dev), TYPE_PC_DIMM)) {
|
|
|
|
virt_memory_plug(hotplug_dev, dev, errp);
|
|
|
|
}
|
2020-02-14 16:27:44 +03:00
|
|
|
if (object_dynamic_cast(OBJECT(dev), TYPE_VIRTIO_IOMMU_PCI)) {
|
|
|
|
PCIDevice *pdev = PCI_DEVICE(dev);
|
|
|
|
|
|
|
|
vms->iommu = VIRT_IOMMU_VIRTIO;
|
|
|
|
vms->virtio_iommu_bdf = pci_get_bdf(pdev);
|
2020-04-22 16:07:10 +03:00
|
|
|
create_virtio_iommu_dt_bindings(vms);
|
2020-02-14 16:27:44 +03:00
|
|
|
}
|
2019-09-18 16:06:26 +03:00
|
|
|
}
|
|
|
|
|
2020-06-22 15:41:57 +03:00
|
|
|
static void virt_dimm_unplug_request(HotplugHandler *hotplug_dev,
|
|
|
|
DeviceState *dev, Error **errp)
|
|
|
|
{
|
|
|
|
VirtMachineState *vms = VIRT_MACHINE(hotplug_dev);
|
|
|
|
Error *local_err = NULL;
|
|
|
|
|
|
|
|
if (!vms->acpi_dev) {
|
|
|
|
error_setg(&local_err,
|
|
|
|
"memory hotplug is not enabled: missing acpi-ged device");
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (object_dynamic_cast(OBJECT(dev), TYPE_NVDIMM)) {
|
|
|
|
error_setg(&local_err,
|
|
|
|
"nvdimm device hot unplug is not supported yet.");
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
|
|
|
hotplug_handler_unplug_request(HOTPLUG_HANDLER(vms->acpi_dev), dev,
|
|
|
|
&local_err);
|
|
|
|
out:
|
|
|
|
error_propagate(errp, local_err);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void virt_dimm_unplug(HotplugHandler *hotplug_dev,
|
|
|
|
DeviceState *dev, Error **errp)
|
|
|
|
{
|
|
|
|
VirtMachineState *vms = VIRT_MACHINE(hotplug_dev);
|
|
|
|
Error *local_err = NULL;
|
|
|
|
|
|
|
|
hotplug_handler_unplug(HOTPLUG_HANDLER(vms->acpi_dev), dev, &local_err);
|
|
|
|
if (local_err) {
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
|
|
|
pc_dimm_unplug(PC_DIMM(dev), MACHINE(vms));
|
|
|
|
qdev_unrealize(dev);
|
|
|
|
|
|
|
|
out:
|
|
|
|
error_propagate(errp, local_err);
|
|
|
|
}
|
|
|
|
|
2019-09-18 16:06:26 +03:00
|
|
|
static void virt_machine_device_unplug_request_cb(HotplugHandler *hotplug_dev,
|
|
|
|
DeviceState *dev, Error **errp)
|
|
|
|
{
|
2020-06-22 15:41:57 +03:00
|
|
|
if (object_dynamic_cast(OBJECT(dev), TYPE_PC_DIMM)) {
|
|
|
|
virt_dimm_unplug_request(hotplug_dev, dev, errp);
|
|
|
|
} else {
|
|
|
|
error_setg(errp, "device unplug request for unsupported device"
|
|
|
|
" type: %s", object_get_typename(OBJECT(dev)));
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static void virt_machine_device_unplug_cb(HotplugHandler *hotplug_dev,
|
|
|
|
DeviceState *dev, Error **errp)
|
|
|
|
{
|
|
|
|
if (object_dynamic_cast(OBJECT(dev), TYPE_PC_DIMM)) {
|
|
|
|
virt_dimm_unplug(hotplug_dev, dev, errp);
|
|
|
|
} else {
|
|
|
|
error_setg(errp, "virt: device unplug for unsupported device"
|
|
|
|
" type: %s", object_get_typename(OBJECT(dev)));
|
|
|
|
}
|
2018-05-10 20:10:56 +03:00
|
|
|
}
|
|
|
|
|
|
|
|
static HotplugHandler *virt_machine_get_hotplug_handler(MachineState *machine,
|
|
|
|
DeviceState *dev)
|
|
|
|
{
|
2021-03-25 18:33:09 +03:00
|
|
|
MachineClass *mc = MACHINE_GET_CLASS(machine);
|
|
|
|
|
|
|
|
if (device_is_dynamic_sysbus(mc, dev) ||
|
2019-09-18 16:06:26 +03:00
|
|
|
(object_dynamic_cast(OBJECT(dev), TYPE_PC_DIMM))) {
|
2018-05-10 20:10:56 +03:00
|
|
|
return HOTPLUG_HANDLER(machine);
|
|
|
|
}
|
2020-02-14 16:27:44 +03:00
|
|
|
if (object_dynamic_cast(OBJECT(dev), TYPE_VIRTIO_IOMMU_PCI)) {
|
|
|
|
VirtMachineState *vms = VIRT_MACHINE(machine);
|
2018-05-10 20:10:56 +03:00
|
|
|
|
2020-03-20 13:01:36 +03:00
|
|
|
if (!vms->bootinfo.firmware_loaded || !virt_is_acpi_enabled(vms)) {
|
2020-02-14 16:27:44 +03:00
|
|
|
return HOTPLUG_HANDLER(machine);
|
|
|
|
}
|
|
|
|
}
|
2018-05-10 20:10:56 +03:00
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
2019-03-04 13:13:37 +03:00
|
|
|
/*
|
|
|
|
* for arm64 kvm_type [7-0] encodes the requested number of bits
|
|
|
|
* in the IPA address space
|
|
|
|
*/
|
|
|
|
static int virt_kvm_type(MachineState *ms, const char *type_str)
|
|
|
|
{
|
|
|
|
VirtMachineState *vms = VIRT_MACHINE(ms);
|
hw/arm/virt: KVM: The IPA lower bound is 32
The virt machine already checks KVM_CAP_ARM_VM_IPA_SIZE to get the
upper bound of the IPA size. If that bound is lower than the highest
possible GPA for the machine, then QEMU will error out. However, the
IPA is set to 40 when the highest GPA is less than or equal to 40,
even when KVM may support an IPA limit as low as 32. This means KVM
may fail the VM creation unnecessarily. Additionally, 40 is selected
with the value 0, which means use the default, and that gets around
a check in some versions of KVM, causing a difficult to debug fail.
Always use the IPA size that corresponds to the highest possible GPA,
unless it's lower than 32, in which case use 32. Also, we must still
use 0 when KVM only supports the legacy fixed 40 bit IPA.
Suggested-by: Marc Zyngier <maz@kernel.org>
Signed-off-by: Andrew Jones <drjones@redhat.com>
Reviewed-by: Eric Auger <eric.auger@redhat.com>
Reviewed-by: Marc Zyngier <maz@kernel.org>
Message-id: 20210310135218.255205-3-drjones@redhat.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
2021-03-10 16:52:18 +03:00
|
|
|
int max_vm_pa_size, requested_pa_size;
|
|
|
|
bool fixed_ipa;
|
|
|
|
|
|
|
|
max_vm_pa_size = kvm_arm_get_max_vm_ipa_size(ms, &fixed_ipa);
|
2019-03-04 13:13:37 +03:00
|
|
|
|
|
|
|
/* we freeze the memory map to compute the highest gpa */
|
|
|
|
virt_set_memmap(vms);
|
|
|
|
|
|
|
|
requested_pa_size = 64 - clz64(vms->highest_gpa);
|
|
|
|
|
hw/arm/virt: KVM: The IPA lower bound is 32
The virt machine already checks KVM_CAP_ARM_VM_IPA_SIZE to get the
upper bound of the IPA size. If that bound is lower than the highest
possible GPA for the machine, then QEMU will error out. However, the
IPA is set to 40 when the highest GPA is less than or equal to 40,
even when KVM may support an IPA limit as low as 32. This means KVM
may fail the VM creation unnecessarily. Additionally, 40 is selected
with the value 0, which means use the default, and that gets around
a check in some versions of KVM, causing a difficult to debug fail.
Always use the IPA size that corresponds to the highest possible GPA,
unless it's lower than 32, in which case use 32. Also, we must still
use 0 when KVM only supports the legacy fixed 40 bit IPA.
Suggested-by: Marc Zyngier <maz@kernel.org>
Signed-off-by: Andrew Jones <drjones@redhat.com>
Reviewed-by: Eric Auger <eric.auger@redhat.com>
Reviewed-by: Marc Zyngier <maz@kernel.org>
Message-id: 20210310135218.255205-3-drjones@redhat.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
2021-03-10 16:52:18 +03:00
|
|
|
/*
|
|
|
|
* KVM requires the IPA size to be at least 32 bits.
|
|
|
|
*/
|
|
|
|
if (requested_pa_size < 32) {
|
|
|
|
requested_pa_size = 32;
|
|
|
|
}
|
|
|
|
|
2019-03-04 13:13:37 +03:00
|
|
|
if (requested_pa_size > max_vm_pa_size) {
|
|
|
|
error_report("-m and ,maxmem option values "
|
|
|
|
"require an IPA range (%d bits) larger than "
|
|
|
|
"the one supported by the host (%d bits)",
|
|
|
|
requested_pa_size, max_vm_pa_size);
|
hw/arm/virt: KVM: The IPA lower bound is 32
The virt machine already checks KVM_CAP_ARM_VM_IPA_SIZE to get the
upper bound of the IPA size. If that bound is lower than the highest
possible GPA for the machine, then QEMU will error out. However, the
IPA is set to 40 when the highest GPA is less than or equal to 40,
even when KVM may support an IPA limit as low as 32. This means KVM
may fail the VM creation unnecessarily. Additionally, 40 is selected
with the value 0, which means use the default, and that gets around
a check in some versions of KVM, causing a difficult to debug fail.
Always use the IPA size that corresponds to the highest possible GPA,
unless it's lower than 32, in which case use 32. Also, we must still
use 0 when KVM only supports the legacy fixed 40 bit IPA.
Suggested-by: Marc Zyngier <maz@kernel.org>
Signed-off-by: Andrew Jones <drjones@redhat.com>
Reviewed-by: Eric Auger <eric.auger@redhat.com>
Reviewed-by: Marc Zyngier <maz@kernel.org>
Message-id: 20210310135218.255205-3-drjones@redhat.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
2021-03-10 16:52:18 +03:00
|
|
|
exit(1);
|
2019-03-04 13:13:37 +03:00
|
|
|
}
|
|
|
|
/*
|
hw/arm/virt: KVM: The IPA lower bound is 32
The virt machine already checks KVM_CAP_ARM_VM_IPA_SIZE to get the
upper bound of the IPA size. If that bound is lower than the highest
possible GPA for the machine, then QEMU will error out. However, the
IPA is set to 40 when the highest GPA is less than or equal to 40,
even when KVM may support an IPA limit as low as 32. This means KVM
may fail the VM creation unnecessarily. Additionally, 40 is selected
with the value 0, which means use the default, and that gets around
a check in some versions of KVM, causing a difficult to debug fail.
Always use the IPA size that corresponds to the highest possible GPA,
unless it's lower than 32, in which case use 32. Also, we must still
use 0 when KVM only supports the legacy fixed 40 bit IPA.
Suggested-by: Marc Zyngier <maz@kernel.org>
Signed-off-by: Andrew Jones <drjones@redhat.com>
Reviewed-by: Eric Auger <eric.auger@redhat.com>
Reviewed-by: Marc Zyngier <maz@kernel.org>
Message-id: 20210310135218.255205-3-drjones@redhat.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
2021-03-10 16:52:18 +03:00
|
|
|
* We return the requested PA log size, unless KVM only supports
|
|
|
|
* the implicit legacy 40b IPA setting, in which case the kvm_type
|
|
|
|
* must be 0.
|
2019-03-04 13:13:37 +03:00
|
|
|
*/
|
hw/arm/virt: KVM: The IPA lower bound is 32
The virt machine already checks KVM_CAP_ARM_VM_IPA_SIZE to get the
upper bound of the IPA size. If that bound is lower than the highest
possible GPA for the machine, then QEMU will error out. However, the
IPA is set to 40 when the highest GPA is less than or equal to 40,
even when KVM may support an IPA limit as low as 32. This means KVM
may fail the VM creation unnecessarily. Additionally, 40 is selected
with the value 0, which means use the default, and that gets around
a check in some versions of KVM, causing a difficult to debug fail.
Always use the IPA size that corresponds to the highest possible GPA,
unless it's lower than 32, in which case use 32. Also, we must still
use 0 when KVM only supports the legacy fixed 40 bit IPA.
Suggested-by: Marc Zyngier <maz@kernel.org>
Signed-off-by: Andrew Jones <drjones@redhat.com>
Reviewed-by: Eric Auger <eric.auger@redhat.com>
Reviewed-by: Marc Zyngier <maz@kernel.org>
Message-id: 20210310135218.255205-3-drjones@redhat.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
2021-03-10 16:52:18 +03:00
|
|
|
return fixed_ipa ? 0 : requested_pa_size;
|
2019-03-04 13:13:37 +03:00
|
|
|
}
|
|
|
|
|
2016-03-16 20:05:59 +03:00
|
|
|
static void virt_machine_class_init(ObjectClass *oc, void *data)
|
|
|
|
{
|
2016-03-16 20:05:59 +03:00
|
|
|
MachineClass *mc = MACHINE_CLASS(oc);
|
2018-05-10 20:10:56 +03:00
|
|
|
HotplugHandlerClass *hc = HOTPLUG_HANDLER_CLASS(oc);
|
2016-03-16 20:05:59 +03:00
|
|
|
|
|
|
|
mc->init = machvirt_init;
|
2018-06-22 15:28:38 +03:00
|
|
|
/* Start with max_cpus set to 512, which is the maximum supported by KVM.
|
|
|
|
* The value may be reduced later when we have more information about the
|
2016-03-16 20:05:59 +03:00
|
|
|
* configuration of the particular instance.
|
|
|
|
*/
|
2018-06-22 15:28:38 +03:00
|
|
|
mc->max_cpus = 512;
|
2017-11-25 18:16:06 +03:00
|
|
|
machine_class_allow_dynamic_sysbus_dev(mc, TYPE_VFIO_CALXEDA_XGMAC);
|
|
|
|
machine_class_allow_dynamic_sysbus_dev(mc, TYPE_VFIO_AMD_XGBE);
|
2018-06-13 15:29:46 +03:00
|
|
|
machine_class_allow_dynamic_sysbus_dev(mc, TYPE_RAMFB_DEVICE);
|
2018-10-15 19:52:10 +03:00
|
|
|
machine_class_allow_dynamic_sysbus_dev(mc, TYPE_VFIO_PLATFORM);
|
2021-06-15 17:21:18 +03:00
|
|
|
#ifdef CONFIG_TPM
|
2020-03-05 19:51:45 +03:00
|
|
|
machine_class_allow_dynamic_sysbus_dev(mc, TYPE_TPM_TIS_SYSBUS);
|
2021-06-15 17:21:18 +03:00
|
|
|
#endif
|
2016-03-16 20:05:59 +03:00
|
|
|
mc->block_default_type = IF_VIRTIO;
|
|
|
|
mc->no_cdrom = 1;
|
|
|
|
mc->pci_allow_0_address = true;
|
2016-10-24 18:26:50 +03:00
|
|
|
/* We know we will never create a pre-ARMv7 CPU which needs 1K pages */
|
|
|
|
mc->minimum_page_bits = 12;
|
2017-05-03 15:56:57 +03:00
|
|
|
mc->possible_cpu_arch_ids = virt_possible_cpu_arch_ids;
|
2017-05-10 14:29:45 +03:00
|
|
|
mc->cpu_index_to_instance_props = virt_cpu_index_to_props;
|
2017-09-13 19:04:57 +03:00
|
|
|
mc->default_cpu_type = ARM_CPU_TYPE_NAME("cortex-a15");
|
2017-06-01 13:53:28 +03:00
|
|
|
mc->get_default_cpu_node_id = virt_get_default_cpu_node_id;
|
2019-03-04 13:13:37 +03:00
|
|
|
mc->kvm_type = virt_kvm_type;
|
2018-05-10 20:10:56 +03:00
|
|
|
assert(!mc->get_hotplug_handler);
|
2018-05-10 20:10:56 +03:00
|
|
|
mc->get_hotplug_handler = virt_machine_get_hotplug_handler;
|
2019-09-18 16:06:26 +03:00
|
|
|
hc->pre_plug = virt_machine_device_pre_plug_cb;
|
2018-05-10 20:10:56 +03:00
|
|
|
hc->plug = virt_machine_device_plug_cb;
|
2019-09-18 16:06:26 +03:00
|
|
|
hc->unplug_request = virt_machine_device_unplug_request_cb;
|
2020-06-22 15:41:57 +03:00
|
|
|
hc->unplug = virt_machine_device_unplug_cb;
|
2020-04-21 15:59:31 +03:00
|
|
|
mc->nvdimm_supported = true;
|
2019-09-18 16:06:28 +03:00
|
|
|
mc->auto_enable_numa_with_memhp = true;
|
2020-06-26 10:22:48 +03:00
|
|
|
mc->auto_enable_numa_with_memdev = true;
|
2020-02-19 19:09:07 +03:00
|
|
|
mc->default_ram_id = "mach-virt.ram";
|
2020-03-20 13:01:36 +03:00
|
|
|
|
|
|
|
object_class_property_add(oc, "acpi", "OnOffAuto",
|
|
|
|
virt_get_acpi, virt_set_acpi,
|
qom: Drop parameter @errp of object_property_add() & friends
The only way object_property_add() can fail is when a property with
the same name already exists. Since our property names are all
hardcoded, failure is a programming error, and the appropriate way to
handle it is passing &error_abort.
Same for its variants, except for object_property_add_child(), which
additionally fails when the child already has a parent. Parentage is
also under program control, so this is a programming error, too.
We have a bit over 500 callers. Almost half of them pass
&error_abort, slightly fewer ignore errors, one test case handles
errors, and the remaining few callers pass them to their own callers.
The previous few commits demonstrated once again that ignoring
programming errors is a bad idea.
Of the few ones that pass on errors, several violate the Error API.
The Error ** argument must be NULL, &error_abort, &error_fatal, or a
pointer to a variable containing NULL. Passing an argument of the
latter kind twice without clearing it in between is wrong: if the
first call sets an error, it no longer points to NULL for the second
call. ich9_pm_add_properties(), sparc32_ledma_realize(),
sparc32_dma_realize(), xilinx_axidma_realize(), xilinx_enet_realize()
are wrong that way.
When the one appropriate choice of argument is &error_abort, letting
users pick the argument is a bad idea.
Drop parameter @errp and assert the preconditions instead.
There's one exception to "duplicate property name is a programming
error": the way object_property_add() implements the magic (and
undocumented) "automatic arrayification". Don't drop @errp there.
Instead, rename object_property_add() to object_property_try_add(),
and add the obvious wrapper object_property_add().
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20200505152926.18877-15-armbru@redhat.com>
[Two semantic rebase conflicts resolved]
2020-05-05 18:29:22 +03:00
|
|
|
NULL, NULL);
|
2020-03-20 13:01:36 +03:00
|
|
|
object_class_property_set_description(oc, "acpi",
|
2020-05-05 18:29:15 +03:00
|
|
|
"Enable ACPI");
|
2020-11-11 21:38:16 +03:00
|
|
|
object_class_property_add_bool(oc, "secure", virt_get_secure,
|
|
|
|
virt_set_secure);
|
|
|
|
object_class_property_set_description(oc, "secure",
|
|
|
|
"Set on/off to enable/disable the ARM "
|
|
|
|
"Security Extensions (TrustZone)");
|
|
|
|
|
|
|
|
object_class_property_add_bool(oc, "virtualization", virt_get_virt,
|
|
|
|
virt_set_virt);
|
|
|
|
object_class_property_set_description(oc, "virtualization",
|
|
|
|
"Set on/off to enable/disable emulating a "
|
|
|
|
"guest CPU which implements the ARM "
|
|
|
|
"Virtualization Extensions");
|
|
|
|
|
|
|
|
object_class_property_add_bool(oc, "highmem", virt_get_highmem,
|
|
|
|
virt_set_highmem);
|
|
|
|
object_class_property_set_description(oc, "highmem",
|
|
|
|
"Set on/off to enable/disable using "
|
|
|
|
"physical address space above 32 bits");
|
|
|
|
|
|
|
|
object_class_property_add_str(oc, "gic-version", virt_get_gic_version,
|
|
|
|
virt_set_gic_version);
|
|
|
|
object_class_property_set_description(oc, "gic-version",
|
|
|
|
"Set GIC version. "
|
|
|
|
"Valid values are 2, 3, host and max");
|
|
|
|
|
|
|
|
object_class_property_add_str(oc, "iommu", virt_get_iommu, virt_set_iommu);
|
|
|
|
object_class_property_set_description(oc, "iommu",
|
|
|
|
"Set the IOMMU type. "
|
|
|
|
"Valid values are none and smmuv3");
|
|
|
|
|
2021-07-08 15:55:13 +03:00
|
|
|
object_class_property_add_bool(oc, "default_bus_bypass_iommu",
|
|
|
|
virt_get_default_bus_bypass_iommu,
|
|
|
|
virt_set_default_bus_bypass_iommu);
|
|
|
|
object_class_property_set_description(oc, "default_bus_bypass_iommu",
|
|
|
|
"Set on/off to enable/disable "
|
|
|
|
"bypass_iommu for default root bus");
|
|
|
|
|
2020-11-11 21:38:16 +03:00
|
|
|
object_class_property_add_bool(oc, "ras", virt_get_ras,
|
|
|
|
virt_set_ras);
|
|
|
|
object_class_property_set_description(oc, "ras",
|
|
|
|
"Set on/off to enable/disable reporting host memory errors "
|
|
|
|
"to a KVM guest using ACPI and guest external abort exceptions");
|
|
|
|
|
|
|
|
object_class_property_add_bool(oc, "mte", virt_get_mte, virt_set_mte);
|
|
|
|
object_class_property_set_description(oc, "mte",
|
|
|
|
"Set on/off to enable/disable emulating a "
|
|
|
|
"guest CPU which implements the ARM "
|
|
|
|
"Memory Tagging Extension");
|
2020-11-11 21:38:17 +03:00
|
|
|
|
|
|
|
object_class_property_add_bool(oc, "its", virt_get_its,
|
|
|
|
virt_set_its);
|
|
|
|
object_class_property_set_description(oc, "its",
|
|
|
|
"Set on/off to enable/disable "
|
|
|
|
"ITS instantiation");
|
|
|
|
|
2021-04-02 11:21:28 +03:00
|
|
|
object_class_property_add_str(oc, "x-oem-id",
|
2021-01-19 03:32:13 +03:00
|
|
|
virt_get_oem_id,
|
|
|
|
virt_set_oem_id);
|
2021-04-02 11:21:28 +03:00
|
|
|
object_class_property_set_description(oc, "x-oem-id",
|
2021-01-19 03:32:13 +03:00
|
|
|
"Override the default value of field OEMID "
|
|
|
|
"in ACPI table header."
|
|
|
|
"The string may be up to 6 bytes in size");
|
|
|
|
|
|
|
|
|
2021-04-02 11:21:28 +03:00
|
|
|
object_class_property_add_str(oc, "x-oem-table-id",
|
2021-01-19 03:32:13 +03:00
|
|
|
virt_get_oem_table_id,
|
|
|
|
virt_set_oem_table_id);
|
2021-04-02 11:21:28 +03:00
|
|
|
object_class_property_set_description(oc, "x-oem-table-id",
|
2021-01-19 03:32:13 +03:00
|
|
|
"Override the default value of field OEM Table ID "
|
|
|
|
"in ACPI table header."
|
|
|
|
"The string may be up to 8 bytes in size");
|
|
|
|
|
2016-03-16 20:05:59 +03:00
|
|
|
}
|
|
|
|
|
2018-12-05 23:58:23 +03:00
|
|
|
static void virt_instance_init(Object *obj)
|
2014-12-16 02:09:44 +03:00
|
|
|
{
|
|
|
|
VirtMachineState *vms = VIRT_MACHINE(obj);
|
2017-02-28 15:08:16 +03:00
|
|
|
VirtMachineClass *vmc = VIRT_MACHINE_GET_CLASS(vms);
|
2014-12-16 02:09:44 +03:00
|
|
|
|
2015-09-08 19:38:44 +03:00
|
|
|
/* EL3 is disabled by default on virt: this makes us consistent
|
|
|
|
* between KVM and TCG for this board, and it also allows us to
|
|
|
|
* boot UEFI blobs which assume no TrustZone support.
|
|
|
|
*/
|
|
|
|
vms->secure = false;
|
2015-09-07 12:39:29 +03:00
|
|
|
|
2017-01-20 14:15:11 +03:00
|
|
|
/* EL2 is also disabled by default, for similar reasons */
|
|
|
|
vms->virt = false;
|
|
|
|
|
2015-09-07 12:39:29 +03:00
|
|
|
/* High memory is enabled by default */
|
|
|
|
vms->highmem = true;
|
2020-03-11 16:16:15 +03:00
|
|
|
vms->gic_version = VIRT_GIC_VERSION_NOSEL;
|
2017-01-09 14:40:21 +03:00
|
|
|
|
2018-06-22 15:28:37 +03:00
|
|
|
vms->highmem_ecam = !vmc->no_highmem_ecam;
|
|
|
|
|
2017-02-28 15:08:16 +03:00
|
|
|
if (vmc->no_its) {
|
|
|
|
vms->its = false;
|
|
|
|
} else {
|
|
|
|
/* Default allows ITS instantiation */
|
|
|
|
vms->its = true;
|
2021-09-13 18:07:24 +03:00
|
|
|
|
|
|
|
if (vmc->no_tcg_its) {
|
|
|
|
vms->tcg_its = false;
|
|
|
|
} else {
|
|
|
|
vms->tcg_its = true;
|
|
|
|
}
|
2017-02-28 15:08:16 +03:00
|
|
|
}
|
|
|
|
|
2018-05-04 20:05:52 +03:00
|
|
|
/* Default disallows iommu instantiation */
|
|
|
|
vms->iommu = VIRT_IOMMU_NONE;
|
|
|
|
|
2021-07-08 15:55:13 +03:00
|
|
|
/* The default root bus is attached to iommu by default */
|
|
|
|
vms->default_bus_bypass_iommu = false;
|
|
|
|
|
2020-05-12 06:06:01 +03:00
|
|
|
/* Default disallows RAS instantiation */
|
|
|
|
vms->ras = false;
|
|
|
|
|
2020-07-20 12:25:36 +03:00
|
|
|
/* MTE is disabled by default. */
|
|
|
|
vms->mte = false;
|
|
|
|
|
2017-01-09 14:40:21 +03:00
|
|
|
vms->irqmap = a15irqmap;
|
hw/arm/virt: Support firmware configuration with -blockdev
The ARM virt machines put firmware in flash memory. To configure it,
you use -drive if=pflash,unit=0,... and optionally -drive
if=pflash,unit=1,...
Why two -drive? This permits setting up one part of the flash memory
read-only, and the other part read/write. It also makes upgrading
firmware on the host easier. Below the hood, we get two separate
flash devices, because we were too lazy to improve our flash device
models to support sector protection.
The problem at hand is to do the same with -blockdev somehow, as one
more step towards deprecating -drive.
We recently solved this problem for x86 PC machines, in commit
ebc29e1beab. See the commit message for design rationale.
This commit solves it for ARM virt basically the same way: new machine
properties pflash0, pflash1 forward to the onboard flash devices'
properties. Requires creating the onboard devices in the
.instance_init() method virt_instance_init(). The existing code to
pick up drives defined with -drive if=pflash is replaced by code to
desugar into the machine properties.
There are a few behavioral differences, though:
* The flash devices are always present (x86: only present if
configured)
* Flash base addresses and sizes are fixed (x86: sizes depend on
images, mapped back to back below a fixed address)
* -bios configures contents of first pflash (x86: -bios configures ROM
contents)
* -bios is rejected when first pflash is also configured with -machine
pflash0=... (x86: bios is silently ignored then)
* -machine pflash1=... does not require -machine pflash0=... (x86: it
does).
The actual code is a bit simpler than for x86 mostly due to the first
two differences.
Before the patch, all the action is in create_flash(), called from the
machine's .init() method machvirt_init():
main()
machine_run_board_init()
machvirt_init()
create_flash()
create_one_flash() for flash[0]
create
configure
includes obeying -drive if=pflash,unit=0
realize
map
fall back to -bios
create_one_flash() for flash[1]
create
configure
includes obeying -drive if=pflash,unit=1
realize
map
update FDT
To make the machine properties work, we need to move device creation
to its .instance_init() method virt_instance_init().
Another complication is machvirt_init()'s computation of
@firmware_loaded: it predicts what create_flash() will do. Instead of
predicting what create_flash()'s replacement virt_firmware_init() will
do, I decided to have virt_firmware_init() return what it did.
Requires calling it a bit earlier.
Resulting call tree:
main()
current_machine = object_new()
...
virt_instance_init()
virt_flash_create()
virt_flash_create1() for flash[0]
create
configure: set defaults
become child of machine [NEW]
add machine prop pflash0 as alias for drive [NEW]
virt_flash_create1() for flash[1]
create
configure: set defaults
become child of machine [NEW]
add machine prop pflash1 as alias for drive [NEW]
for all machine props from the command line: machine_set_property()
...
property_set_alias() for machine props pflash0, pflash1
...
set_drive() for cfi.pflash01 prop drive
this is how -machine pflash0=... etc set
machine_run_board_init(current_machine);
virt_firmware_init()
pflash_cfi01_legacy_drive()
legacy -drive if=pflash,unit=0 and =1 [NEW]
virt_flash_map()
virt_flash_map1() for flash[0]
configure: num-blocks
realize
map
virt_flash_map1() for flash[1]
configure: num-blocks
realize
map
fall back to -bios
virt_flash_fdt()
update FDT
You have László to thank for making me explain this in detail.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Acked-by: Laszlo Ersek <lersek@redhat.com>
Message-id: 20190416091348.26075-4-armbru@redhat.com
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
2019-05-07 14:55:02 +03:00
|
|
|
|
|
|
|
virt_flash_create(vms);
|
2021-01-19 03:32:13 +03:00
|
|
|
|
|
|
|
vms->oem_id = g_strndup(ACPI_BUILD_APPNAME6, 6);
|
|
|
|
vms->oem_table_id = g_strndup(ACPI_BUILD_APPNAME8, 8);
|
2014-12-16 02:09:44 +03:00
|
|
|
}
|
|
|
|
|
2018-12-05 23:58:23 +03:00
|
|
|
static const TypeInfo virt_machine_info = {
|
|
|
|
.name = TYPE_VIRT_MACHINE,
|
|
|
|
.parent = TYPE_MACHINE,
|
|
|
|
.abstract = true,
|
|
|
|
.instance_size = sizeof(VirtMachineState),
|
|
|
|
.class_size = sizeof(VirtMachineClass),
|
|
|
|
.class_init = virt_machine_class_init,
|
2018-12-14 16:30:55 +03:00
|
|
|
.instance_init = virt_instance_init,
|
2018-12-05 23:58:23 +03:00
|
|
|
.interfaces = (InterfaceInfo[]) {
|
|
|
|
{ TYPE_HOTPLUG_HANDLER },
|
|
|
|
{ }
|
|
|
|
},
|
|
|
|
};
|
|
|
|
|
|
|
|
static void machvirt_machine_init(void)
|
|
|
|
{
|
|
|
|
type_register_static(&virt_machine_info);
|
|
|
|
}
|
|
|
|
type_init(machvirt_machine_init);
|
|
|
|
|
2021-08-31 04:54:26 +03:00
|
|
|
static void virt_machine_6_2_options(MachineClass *mc)
|
|
|
|
{
|
|
|
|
}
|
|
|
|
DEFINE_VIRT_MACHINE_AS_LATEST(6, 2)
|
|
|
|
|
2021-03-31 14:19:00 +03:00
|
|
|
static void virt_machine_6_1_options(MachineClass *mc)
|
|
|
|
{
|
2021-09-13 18:07:24 +03:00
|
|
|
VirtMachineClass *vmc = VIRT_MACHINE_CLASS(OBJECT_CLASS(mc));
|
|
|
|
|
2021-08-31 04:54:26 +03:00
|
|
|
virt_machine_6_2_options(mc);
|
|
|
|
compat_props_add(mc->compat_props, hw_compat_6_1, hw_compat_6_1_len);
|
2021-09-29 05:58:14 +03:00
|
|
|
mc->smp_props.prefer_sockets = true;
|
2021-10-20 17:21:18 +03:00
|
|
|
vmc->no_cpu_topology = true;
|
2021-09-13 18:07:24 +03:00
|
|
|
|
|
|
|
/* qemu ITS was introduced with 6.2 */
|
|
|
|
vmc->no_tcg_its = true;
|
2021-03-31 14:19:00 +03:00
|
|
|
}
|
2021-08-31 04:54:26 +03:00
|
|
|
DEFINE_VIRT_MACHINE(6, 1)
|
2021-03-31 14:19:00 +03:00
|
|
|
|
2020-11-09 20:39:28 +03:00
|
|
|
static void virt_machine_6_0_options(MachineClass *mc)
|
|
|
|
{
|
2021-06-10 21:34:59 +03:00
|
|
|
virt_machine_6_1_options(mc);
|
|
|
|
compat_props_add(mc->compat_props, hw_compat_6_0, hw_compat_6_0_len);
|
2020-11-09 20:39:28 +03:00
|
|
|
}
|
2021-03-31 14:19:00 +03:00
|
|
|
DEFINE_VIRT_MACHINE(6, 0)
|
2020-11-09 20:39:28 +03:00
|
|
|
|
2020-08-19 17:40:16 +03:00
|
|
|
static void virt_machine_5_2_options(MachineClass *mc)
|
|
|
|
{
|
2021-01-28 15:00:11 +03:00
|
|
|
VirtMachineClass *vmc = VIRT_MACHINE_CLASS(OBJECT_CLASS(mc));
|
|
|
|
|
2020-11-09 20:39:28 +03:00
|
|
|
virt_machine_6_0_options(mc);
|
|
|
|
compat_props_add(mc->compat_props, hw_compat_5_2, hw_compat_5_2_len);
|
2021-01-28 15:00:11 +03:00
|
|
|
vmc->no_secure_gpio = true;
|
2020-08-19 17:40:16 +03:00
|
|
|
}
|
2020-11-09 20:39:28 +03:00
|
|
|
DEFINE_VIRT_MACHINE(5, 2)
|
2020-08-19 17:40:16 +03:00
|
|
|
|
2020-04-29 17:46:05 +03:00
|
|
|
static void virt_machine_5_1_options(MachineClass *mc)
|
|
|
|
{
|
2020-10-01 09:17:18 +03:00
|
|
|
VirtMachineClass *vmc = VIRT_MACHINE_CLASS(OBJECT_CLASS(mc));
|
|
|
|
|
2020-08-19 17:40:16 +03:00
|
|
|
virt_machine_5_2_options(mc);
|
|
|
|
compat_props_add(mc->compat_props, hw_compat_5_1, hw_compat_5_1_len);
|
2020-10-01 09:17:18 +03:00
|
|
|
vmc->no_kvm_steal_time = true;
|
2020-04-29 17:46:05 +03:00
|
|
|
}
|
2020-08-19 17:40:16 +03:00
|
|
|
DEFINE_VIRT_MACHINE(5, 1)
|
2020-04-29 17:46:05 +03:00
|
|
|
|
2019-11-12 13:48:11 +03:00
|
|
|
static void virt_machine_5_0_options(MachineClass *mc)
|
|
|
|
{
|
2020-07-03 18:59:43 +03:00
|
|
|
VirtMachineClass *vmc = VIRT_MACHINE_CLASS(OBJECT_CLASS(mc));
|
|
|
|
|
2020-04-29 17:46:05 +03:00
|
|
|
virt_machine_5_1_options(mc);
|
2020-06-22 19:22:00 +03:00
|
|
|
compat_props_add(mc->compat_props, hw_compat_5_0, hw_compat_5_0_len);
|
2020-06-09 16:56:35 +03:00
|
|
|
mc->numa_mem_supported = true;
|
2020-07-03 18:59:43 +03:00
|
|
|
vmc->acpi_expose_flash = true;
|
2020-06-26 10:22:48 +03:00
|
|
|
mc->auto_enable_numa_with_memdev = false;
|
2019-11-12 13:48:11 +03:00
|
|
|
}
|
2020-04-29 17:46:05 +03:00
|
|
|
DEFINE_VIRT_MACHINE(5, 0)
|
2019-11-12 13:48:11 +03:00
|
|
|
|
2019-07-24 13:35:24 +03:00
|
|
|
static void virt_machine_4_2_options(MachineClass *mc)
|
|
|
|
{
|
2020-01-30 19:02:06 +03:00
|
|
|
VirtMachineClass *vmc = VIRT_MACHINE_CLASS(OBJECT_CLASS(mc));
|
|
|
|
|
2020-01-30 19:02:06 +03:00
|
|
|
virt_machine_5_0_options(mc);
|
2019-11-05 21:22:17 +03:00
|
|
|
compat_props_add(mc->compat_props, hw_compat_4_2, hw_compat_4_2_len);
|
2020-01-30 19:02:06 +03:00
|
|
|
vmc->kvm_no_adjvtime = true;
|
2019-07-24 13:35:24 +03:00
|
|
|
}
|
2019-11-12 13:48:11 +03:00
|
|
|
DEFINE_VIRT_MACHINE(4, 2)
|
2019-07-24 13:35:24 +03:00
|
|
|
|
2019-04-11 13:20:25 +03:00
|
|
|
static void virt_machine_4_1_options(MachineClass *mc)
|
|
|
|
{
|
2019-09-18 16:06:27 +03:00
|
|
|
VirtMachineClass *vmc = VIRT_MACHINE_CLASS(OBJECT_CLASS(mc));
|
|
|
|
|
2019-07-24 13:35:24 +03:00
|
|
|
virt_machine_4_2_options(mc);
|
|
|
|
compat_props_add(mc->compat_props, hw_compat_4_1, hw_compat_4_1_len);
|
2019-09-18 16:06:27 +03:00
|
|
|
vmc->no_ged = true;
|
2019-09-18 16:06:28 +03:00
|
|
|
mc->auto_enable_numa_with_memhp = false;
|
2019-04-11 13:20:25 +03:00
|
|
|
}
|
2019-07-24 13:35:24 +03:00
|
|
|
DEFINE_VIRT_MACHINE(4, 1)
|
2019-04-11 13:20:25 +03:00
|
|
|
|
2018-12-04 19:27:16 +03:00
|
|
|
static void virt_machine_4_0_options(MachineClass *mc)
|
|
|
|
{
|
2019-04-11 13:20:25 +03:00
|
|
|
virt_machine_4_1_options(mc);
|
|
|
|
compat_props_add(mc->compat_props, hw_compat_4_0, hw_compat_4_0_len);
|
2018-12-04 19:27:16 +03:00
|
|
|
}
|
2019-04-11 13:20:25 +03:00
|
|
|
DEFINE_VIRT_MACHINE(4, 0)
|
2018-12-04 19:27:16 +03:00
|
|
|
|
2018-08-20 13:24:32 +03:00
|
|
|
static void virt_machine_3_1_options(MachineClass *mc)
|
|
|
|
{
|
2018-12-04 19:27:16 +03:00
|
|
|
virt_machine_4_0_options(mc);
|
2018-12-12 18:36:30 +03:00
|
|
|
compat_props_add(mc->compat_props, hw_compat_3_1, hw_compat_3_1_len);
|
2018-08-20 13:24:32 +03:00
|
|
|
}
|
2018-12-04 19:27:16 +03:00
|
|
|
DEFINE_VIRT_MACHINE(3, 1)
|
2018-08-20 13:24:32 +03:00
|
|
|
|
2018-06-22 15:28:37 +03:00
|
|
|
static void virt_machine_3_0_options(MachineClass *mc)
|
|
|
|
{
|
2018-08-20 13:24:32 +03:00
|
|
|
virt_machine_3_1_options(mc);
|
2018-12-12 18:36:30 +03:00
|
|
|
compat_props_add(mc->compat_props, hw_compat_3_0, hw_compat_3_0_len);
|
2018-06-22 15:28:37 +03:00
|
|
|
}
|
2018-08-20 13:24:32 +03:00
|
|
|
DEFINE_VIRT_MACHINE(3, 0)
|
|
|
|
|
2018-01-16 16:28:10 +03:00
|
|
|
static void virt_machine_2_12_options(MachineClass *mc)
|
|
|
|
{
|
2018-06-22 15:28:37 +03:00
|
|
|
VirtMachineClass *vmc = VIRT_MACHINE_CLASS(OBJECT_CLASS(mc));
|
|
|
|
|
2018-06-22 15:28:37 +03:00
|
|
|
virt_machine_3_0_options(mc);
|
2018-12-12 18:36:30 +03:00
|
|
|
compat_props_add(mc->compat_props, hw_compat_2_12, hw_compat_2_12_len);
|
2018-06-22 15:28:37 +03:00
|
|
|
vmc->no_highmem_ecam = true;
|
2018-06-22 15:28:38 +03:00
|
|
|
mc->max_cpus = 255;
|
2018-01-16 16:28:10 +03:00
|
|
|
}
|
2018-06-22 15:28:37 +03:00
|
|
|
DEFINE_VIRT_MACHINE(2, 12)
|
2018-01-16 16:28:10 +03:00
|
|
|
|
2017-11-24 12:43:46 +03:00
|
|
|
static void virt_machine_2_11_options(MachineClass *mc)
|
|
|
|
{
|
2018-03-23 21:26:46 +03:00
|
|
|
VirtMachineClass *vmc = VIRT_MACHINE_CLASS(OBJECT_CLASS(mc));
|
|
|
|
|
2018-01-16 16:28:10 +03:00
|
|
|
virt_machine_2_12_options(mc);
|
2018-12-12 18:36:30 +03:00
|
|
|
compat_props_add(mc->compat_props, hw_compat_2_11, hw_compat_2_11_len);
|
2018-03-23 21:26:46 +03:00
|
|
|
vmc->smbios_old_sys_ver = true;
|
2017-11-24 12:43:46 +03:00
|
|
|
}
|
2018-01-16 16:28:10 +03:00
|
|
|
DEFINE_VIRT_MACHINE(2, 11)
|
2017-11-24 12:43:46 +03:00
|
|
|
|
2017-08-07 14:49:41 +03:00
|
|
|
static void virt_machine_2_10_options(MachineClass *mc)
|
|
|
|
{
|
2017-11-24 12:43:46 +03:00
|
|
|
virt_machine_2_11_options(mc);
|
2018-12-12 18:36:30 +03:00
|
|
|
compat_props_add(mc->compat_props, hw_compat_2_10, hw_compat_2_10_len);
|
2018-10-08 16:55:02 +03:00
|
|
|
/* before 2.11 we never faulted accesses to bad addresses */
|
|
|
|
mc->ignore_memory_transaction_failures = true;
|
2017-08-07 14:49:41 +03:00
|
|
|
}
|
2017-11-24 12:43:46 +03:00
|
|
|
DEFINE_VIRT_MACHINE(2, 10)
|
2017-08-07 14:49:41 +03:00
|
|
|
|
2016-12-27 17:59:26 +03:00
|
|
|
static void virt_machine_2_9_options(MachineClass *mc)
|
|
|
|
{
|
2017-08-07 14:49:41 +03:00
|
|
|
virt_machine_2_10_options(mc);
|
2018-12-12 18:36:30 +03:00
|
|
|
compat_props_add(mc->compat_props, hw_compat_2_9, hw_compat_2_9_len);
|
2016-12-27 17:59:26 +03:00
|
|
|
}
|
2017-08-07 14:49:41 +03:00
|
|
|
DEFINE_VIRT_MACHINE(2, 9)
|
2016-12-27 17:59:26 +03:00
|
|
|
|
2016-10-04 15:28:08 +03:00
|
|
|
static void virt_machine_2_8_options(MachineClass *mc)
|
|
|
|
{
|
2017-01-09 14:40:21 +03:00
|
|
|
VirtMachineClass *vmc = VIRT_MACHINE_CLASS(OBJECT_CLASS(mc));
|
|
|
|
|
2016-12-27 17:59:26 +03:00
|
|
|
virt_machine_2_9_options(mc);
|
2018-12-12 18:36:30 +03:00
|
|
|
compat_props_add(mc->compat_props, hw_compat_2_8, hw_compat_2_8_len);
|
2017-01-09 14:40:21 +03:00
|
|
|
/* For 2.8 and earlier we falsely claimed in the DT that
|
|
|
|
* our timers were edge-triggered, not level-triggered.
|
|
|
|
*/
|
|
|
|
vmc->claim_edge_triggered_timers = true;
|
2016-10-04 15:28:08 +03:00
|
|
|
}
|
2016-12-27 17:59:26 +03:00
|
|
|
DEFINE_VIRT_MACHINE(2, 8)
|
2016-10-04 15:28:08 +03:00
|
|
|
|
2016-06-14 17:59:13 +03:00
|
|
|
static void virt_machine_2_7_options(MachineClass *mc)
|
|
|
|
{
|
2016-10-17 21:22:17 +03:00
|
|
|
VirtMachineClass *vmc = VIRT_MACHINE_CLASS(OBJECT_CLASS(mc));
|
|
|
|
|
2016-10-04 15:28:08 +03:00
|
|
|
virt_machine_2_8_options(mc);
|
2018-12-12 18:36:30 +03:00
|
|
|
compat_props_add(mc->compat_props, hw_compat_2_7, hw_compat_2_7_len);
|
2016-10-17 21:22:17 +03:00
|
|
|
/* ITS was introduced with 2.8 */
|
|
|
|
vmc->no_its = true;
|
2016-10-24 18:26:50 +03:00
|
|
|
/* Stick with 1K pages for migration compatibility */
|
|
|
|
mc->minimum_page_bits = 0;
|
2016-06-14 17:59:13 +03:00
|
|
|
}
|
2016-10-04 15:28:08 +03:00
|
|
|
DEFINE_VIRT_MACHINE(2, 7)
|
2016-06-14 17:59:13 +03:00
|
|
|
|
2016-06-14 17:59:12 +03:00
|
|
|
static void virt_machine_2_6_options(MachineClass *mc)
|
2014-12-16 02:09:43 +03:00
|
|
|
{
|
2016-07-14 18:51:37 +03:00
|
|
|
VirtMachineClass *vmc = VIRT_MACHINE_CLASS(OBJECT_CLASS(mc));
|
|
|
|
|
2016-06-14 17:59:13 +03:00
|
|
|
virt_machine_2_7_options(mc);
|
2018-12-12 18:36:30 +03:00
|
|
|
compat_props_add(mc->compat_props, hw_compat_2_6, hw_compat_2_6_len);
|
2016-07-14 18:51:37 +03:00
|
|
|
vmc->disallow_affinity_adjustment = true;
|
2016-10-28 16:12:31 +03:00
|
|
|
/* Disable PMU for 2.6 as PMU support was first introduced in 2.7 */
|
|
|
|
vmc->no_pmu = true;
|
2014-12-16 02:09:43 +03:00
|
|
|
}
|
2016-06-14 17:59:13 +03:00
|
|
|
DEFINE_VIRT_MACHINE(2, 6)
|