migration: per-mode notifiers

Keep a separate list of migration notifiers for each migration mode.

Suggested-by: Peter Xu <peterx@redhat.com>
Signed-off-by: Steve Sistare <steven.sistare@oracle.com>
Reviewed-by: Peter Xu <peterx@redhat.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Link: https://lore.kernel.org/r/1708622920-68779-8-git-send-email-steven.sistare@oracle.com
Signed-off-by: Peter Xu <peterx@redhat.com>

migration: MigrationNotifyFunc

Define MigrationNotifyFunc to improve type safety and simplify migration
notifiers.

Signed-off-by: Steve Sistare <steven.sistare@oracle.com>
Reviewed-by: Peter Xu <peterx@redhat.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Link: https://lore.kernel.org/r/1708622920-68779-7-git-send-email-steven.sistare@oracle.com
Signed-off-by: Peter Xu <peterx@redhat.com>

migration: remove postcopy_after_devices

postcopy_after_devices and migration_in_postcopy_after_devices are no
longer used, so delete them.

Signed-off-by: Steve Sistare <steven.sistare@oracle.com>
Reviewed-by: Peter Xu <peterx@redhat.com>
Link: https://lore.kernel.org/r/1708622920-68779-6-git-send-email-steven.sistare@oracle.com
Signed-off-by: Peter Xu <peterx@redhat.com>

migration: MigrationEvent for notifiers

Passing MigrationState to notifiers is unsound because they could access
unstable migration state internals or even modify the state.  Instead, pass
the minimal info needed in a new MigrationEvent struct, which could be
extended in the future if needed.

Suggested-by: Peter Xu <peterx@redhat.com>
Signed-off-by: Steve Sistare <steven.sistare@oracle.com>
Reviewed-by: Peter Xu <peterx@redhat.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Link: https://lore.kernel.org/r/1708622920-68779-5-git-send-email-steven.sistare@oracle.com
Signed-off-by: Peter Xu <peterx@redhat.com>

migration: convert to NotifierWithReturn

Change all migration notifiers to type NotifierWithReturn, so notifiers
can return an error status in a future patch.  For now, pass NULL for the
notifier error parameter, and do not check the return value.

Signed-off-by: Steve Sistare <steven.sistare@oracle.com>
Reviewed-by: Peter Xu <peterx@redhat.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Link: https://lore.kernel.org/r/1708622920-68779-4-git-send-email-steven.sistare@oracle.com
[peterx: dropped unexpected update to roms/seabios-hppa]
Signed-off-by: Peter Xu <peterx@redhat.com>

migration: remove error from notifier data

Remove the error object from opaque data passed to notifiers.
Use the new error parameter passed to the notifier instead.

Signed-off-by: Steve Sistare <steven.sistare@oracle.com>
Reviewed-by: Peter Xu <peterx@redhat.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Link: https://lore.kernel.org/r/1708622920-68779-3-git-send-email-steven.sistare@oracle.com
Signed-off-by: Peter Xu <peterx@redhat.com>

notify: pass error to notifier with return

Pass an error object as the third parameter to "notifier with return"
notifiers, so clients no longer need to bundle an error object in the
opaque data.  The new parameter is used in a later patch.

Signed-off-by: Steve Sistare <steven.sistare@oracle.com>
Reviewed-by: Peter Xu <peterx@redhat.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Link: https://lore.kernel.org/r/1708622920-68779-2-git-send-email-steven.sistare@oracle.com
Signed-off-by: Peter Xu <peterx@redhat.com>

migration/multifd: Drop unnecessary helper to destroy IOC

Both socket_send_channel_destroy() and multifd_send_channel_destroy() are
unnecessary wrappers to destroy an IOC, as the only thing to do is to
release the final IOC reference.  We have plenty of code that destroys an
IOC using direct unref() already; keep that style.

Reviewed-by: Fabiano Rosas <farosas@suse.de>
Link: https://lore.kernel.org/r/20240222095301.171137-6-peterx@redhat.com
Signed-off-by: Peter Xu <peterx@redhat.com>

migration/multifd: Cleanup outgoing_args in state destroy

outgoing_args is a global cache of socket address to be reused in multifd.
Freeing the cache in per-channel destructor is more or less a hack.  Move
it to multifd_send_cleanup_state() so it only get checked once.  Use a
small helper to do so because it's internal of socket.c.

Reviewed-by: Fabiano Rosas <farosas@suse.de>
Link: https://lore.kernel.org/r/20240222095301.171137-5-peterx@redhat.com
Signed-off-by: Peter Xu <peterx@redhat.com>

migration/multifd: Make multifd_channel_connect() return void

It never fails, drop the retval and also the Error**.

Suggested-by: Avihai Horon <avihaih@nvidia.com>
Reviewed-by: Fabiano Rosas <farosas@suse.de>
Link: https://lore.kernel.org/r/20240222095301.171137-4-peterx@redhat.com
Signed-off-by: Peter Xu <peterx@redhat.com>

migration/multifd: Drop registered_yank

With a clear definition of p->c protocol, where we only set it up if the
channel is fully established (TLS or non-TLS), registered_yank boolean will
have equal meaning of "p->c != NULL".

Drop registered_yank by checking p->c instead.

Reviewed-by: Fabiano Rosas <farosas@suse.de>
Link: https://lore.kernel.org/r/20240222095301.171137-3-peterx@redhat.com
Signed-off-by: Peter Xu <peterx@redhat.com>

migration/multifd: Cleanup TLS iochannel referencing

Commit a1af605bd5 ("migration/multifd: fix hangup with TLS-Multifd due to
blocking handshake") introduced a thread for TLS channels, which will
resolve the issue on blocking the main thread.  However in the same commit
p->c is slightly abused just to be able to pass over the pointer "p" into
the thread.

That's the major reason we'll need to conditionally free the io channel in
the fault paths.

To clean it up, using a separate structure to pass over both "p" and "tioc"
in the tls handshake thread.  Then we can make it a rule that p->c will
never be set until the channel is completely setup.  With that, we can drop
the tricky conditional unref of the io channel in the error path.

Reviewed-by: Fabiano Rosas <farosas@suse.de>
Link: https://lore.kernel.org/r/20240222095301.171137-2-peterx@redhat.com
Signed-off-by: Peter Xu <peterx@redhat.com>

migration/multifd: Release recv sem_sync earlier

Now that multifd_recv_terminate_threads() is called only once, release
the recv side sem_sync earlier like we do for the send side.

Signed-off-by: Fabiano Rosas <farosas@suse.de>
Reviewed-by: Peter Xu <peterx@redhat.com>
Link: https://lore.kernel.org/r/20240220224138.24759-6-farosas@suse.de
Signed-off-by: Peter Xu <peterx@redhat.com>

migration/multifd: Remove p->quit from recv side

Like we did on the sending side, replace the p->quit per-channel flag
with a global atomic 'exiting' flag.

Signed-off-by: Fabiano Rosas <farosas@suse.de>
Reviewed-by: Peter Xu <peterx@redhat.com>
Link: https://lore.kernel.org/r/20240220224138.24759-5-farosas@suse.de
Signed-off-by: Peter Xu <peterx@redhat.com>

tests/qtest/migration: Add a fd + file test

The fd URI supports an fd that is backed by a file. The code should
select between QIOChannelFile and QIOChannelSocket, depending on the
type of the fd. Add a test for that.

Signed-off-by: Fabiano Rosas <farosas@suse.de>
Reviewed-by: Peter Xu <peterx@redhat.com>
Link: https://lore.kernel.org/r/20240220224138.24759-4-farosas@suse.de
Signed-off-by: Peter Xu <peterx@redhat.com>

tests/qtest/migration: Rename fd_proto test

Next patch adds another fd test. Rename the existing one closer to
what's used on other tests, with the 'precopy' prefix.

Signed-off-by: Fabiano Rosas <farosas@suse.de>
Reviewed-by: Peter Xu <peterx@redhat.com>
Link: https://lore.kernel.org/r/20240220224138.24759-3-farosas@suse.de
Signed-off-by: Peter Xu <peterx@redhat.com>

docs/devel/migration.rst: Document the file transport

When adding the support for file migration with the file: transport,
we missed adding documentation for it.

Signed-off-by: Fabiano Rosas <farosas@suse.de>
Reviewed-by: Peter Xu <peterx@redhat.com>
Link: https://lore.kernel.org/r/20240220224138.24759-2-farosas@suse.de
Signed-off-by: Peter Xu <peterx@redhat.com>

Merge tag 'pull-target-arm-20240227-1' of https://git.linaro.org/people/pmaydell/qemu-arm into staging

target-arm queue:
 * Handle atomic updates of page tables entries in MMIO during PTW
 * Advertise Cortex-A53 erratum #843419 fix via REVIDR
 * MAINTAINERS: Cover hw/ide/ahci-allwinner.c with AllWinner A10 machine
 * misc: m48t59: replace qemu_system_reset_request() call with watchdog_perform_action()
 * misc: pxa2xx_timer: replace qemu_system_reset_request() call with watchdog_perform_action()
 * xlnx-versal-ospi: disable reentrancy detection for iomem_dac
 * sbsa-ref: Simplify init since PCIe is always enabled
 * stm32l4x5: Use TYPE_OR_IRQ when connecting STM32L4x5 EXTI fan-in IRQs
 * pl031: Update last RTCLR value on write in case it's read back
 * block: m25p80: Add support of mt35xu02gbba
 * xlnx-versal-virt: Add machine property ospi-flash
 * reset: refactor system reset to be three-phase aware
 * new board model raspi4b

# -----BEGIN PGP SIGNATURE-----
#
# iQJNBAABCAA3FiEE4aXFk81BneKOgxXPPCUl7RQ2DN4FAmXeAMEZHHBldGVyLm1h
# eWRlbGxAbGluYXJvLm9yZwAKCRA8JSXtFDYM3syyD/4lJzzstbDIAsu94Z4Hi0So
# CFLAMJFsPy3fMsU2IqVP+TDTyhUeMPebwfj7sQHUtQcXVh5i1/HlYgdUgXsnjGWQ
# pc6BxycpW6uJWYb7Ma3CdSGS+hxEpQ+U8Qeijwqg0kAqhjNtrSIkTRQ4u8p8T+kN
# dWtQzp7D15BpEVhWl/2dLWWJwV3H6TThmr1FbK5wl/c7hJzy2uaXqmmCvercU0Zo
# 6ab3SnGyhaujdd/FsDvhnVEYqcmcO2p9UtSnGAbdfw0zsf4p8cS2Q6M9q4DHBFYn
# 6Bt51DFP5D+114VpqRSXF2Lv9K8swjTgqhDld9vCoios6pS3LMwcTAcONUxE8JU+
# uD7kXTN/lv3atNEy4MTFkTeNtKgbYJJuPwWrDRNdbVXPwrEHGWN3+ZYISmuYb+p+
# XL2/7HeP7/qEVMW2d18+7OCriZwKiBRZRKUrtG7mQSBZEMetbhpA+mLcxAZT0FAl
# 18O/mcvEJrrE7x6Bqyv96b8PE0/er5cVg/b/wrkKS8DL4NWU9bJSjJNRrvt9bvvl
# jSzPGo4ngHlfO0OpurLoFOZCVxKWVXgaKkQ3pOz301nsDyhEndNLeCxrITac8G2Q
# C/WQuMaeOoV1x7N2MzaCQmyRzy8yGkG9av0aI/8feobfV/Yg4wPsfhcEn/XQWXKv
# NUJ4/z78FbJlI2JeDP2QSA==
# =xaMv
# -----END PGP SIGNATURE-----
# gpg: Signature made Tue 27 Feb 2024 15:33:21 GMT
# gpg:                using RSA key E1A5C593CD419DE28E8315CF3C2525ED14360CDE
# gpg:                issuer "peter.maydell@linaro.org"
# gpg: Good signature from "Peter Maydell <peter.maydell@linaro.org>" [ultimate]
# gpg:                 aka "Peter Maydell <pmaydell@gmail.com>" [ultimate]
# gpg:                 aka "Peter Maydell <pmaydell@chiark.greenend.org.uk>" [ultimate]
# gpg:                 aka "Peter Maydell <peter@archaic.org.uk>" [ultimate]
# Primary key fingerprint: E1A5 C593 CD41 9DE2 8E83  15CF 3C25 25ED 1436 0CDE

* tag 'pull-target-arm-20240227-1' of https://git.linaro.org/people/pmaydell/qemu-arm: (36 commits)
  docs/system/arm: Add RPi4B to raspi.rst
  hw/misc/bcm2835_property: Add missed BCM2835 properties
  tests/avocado/boot_linux_console.py: Add Rpi4b boot tests
  hw/arm/bcm2838_peripherals: Add clock_isp stub
  hw/arm: Add memory region for BCM2837 RPiVid ASB
  hw/arm/raspi4b: Temporarily disable unimplemented rpi4b devices
  hw/arm: Introduce Raspberry PI 4 machine
  hw/arm: Add GPIO and SD to BCM2838 periph
  hw/gpio: Connect SD controller to BCM2838 GPIO
  hw/gpio: Implement BCM2838 GPIO functionality
  hw/gpio: Add BCM2838 GPIO stub
  hw/arm/bcm2838: Add GIC-400 to BCM2838 SoC
  hw/arm: Introduce BCM2838 SoC
  hw/arm/raspi: Split out raspi machine common part
  hw/arm/bcm2853_peripherals: Split out common part of peripherals
  hw/arm/bcm2836: Split out common part of BCM283X classes
  docs/devel/reset: Update to discuss system reset
  hw/core/machine: Use qemu_register_resettable for sysbus reset
  hw/core/reset: Implement qemu_register_reset via qemu_register_resettable
  hw/core/reset: Add qemu_{register, unregister}_resettable()
  ...

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

Merge tag 'pull-aspeed-20240227' of https://github.com/legoater/qemu into staging

aspeed queue:

* Add support for UART0, in preparation of AST2700 models

# -----BEGIN PGP SIGNATURE-----
#
# iQIzBAABCAAdFiEEoPZlSPBIlev+awtgUaNDx8/77KEFAmXd2nMACgkQUaNDx8/7
# 7KErPBAAjKRmJQF9aMEgf7uqsPnJojAVumFe63NE9Gqnvy4MzgoZWfdSnLl2Ddba
# im5IfR7MYv0tzJtqCVtz7o4JwXhhDwesWALQZBM/ms48aacPSNP+7Gn141yLuCCS
# Vr8NBSIz156lSsnFGnRUArcQTDKjDp/1TLRiGcS8SDm/S4Nn++nur+T054EZgbKR
# CMWDeavgzZRb9HPepvWDwqb9qs11hq5/onCqC886dVNznxEKAVYcd0FVbSn3OfDF
# 2EPvKh+fxHlW37wcctlGPnbJK5rRvFi78yZf5utSt+mlVhyiEXjQJ6p8zBIh2w5A
# NlsmUo/UYv1F41yC/vCFRR8KJ2wO5VW7zL6UCGMV6I9hxhu/Qw+FYqWdBbAZWsOO
# GFOkFbe8zbJFXTr/W7P5upBlA7U1/B9VbRj71eu01dqT+n8OGsk8yfnWVs1SjpoD
# 89ZIhpb7lSolQmjPPxrVyfUe3/8ncTx64+CZuAZjxPh/9HA8wDXwVRPtAbIvvGaZ
# YPQ4Qmd4m6nAANAvTg2ufj19WT64XKwrQ6O3IkmGcn0BzHl08GFjru8IUp6rbduG
# m6WqulL1Ej1PrYaiw5ktpJ4Fkoy6iEFXJOWfl3oTLp2KWE5VAohyRKI00AFnHiAC
# frK+cxT4bqDtJR8QbNyJy5d3ZGZV1R6ZA0XjQ1jtb8ty2qISysw=
# =gFeX
# -----END PGP SIGNATURE-----
# gpg: Signature made Tue 27 Feb 2024 12:49:55 GMT
# gpg:                using RSA key A0F66548F04895EBFE6B0B6051A343C7CFFBECA1
# gpg: Good signature from "Cédric Le Goater <clg@kaod.org>" [undefined]
# gpg: WARNING: This key is not certified with a trusted signature!
# gpg:          There is no indication that the signature belongs to the owner.
# Primary key fingerprint: A0F6 6548 F048 95EB FE6B  0B60 51A3 43C7 CFFB ECA1

* tag 'pull-aspeed-20240227' of https://github.com/legoater/qemu:
  aspeed: fix hardcode boot address 0
  aspeed: introduce a new UART0 device name

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

docs/system/arm: Add RPi4B to raspi.rst

Signed-off-by: Sergey Kambalin <sergey.kambalin@auriga.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Message-id: 20240226000259.2752893-42-sergey.kambalin@auriga.com
[PMM: list PCIE and GENET as 'missing' for now, until we land
 the patches which add those devices]
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

hw/misc/bcm2835_property: Add missed BCM2835 properties

Our model of the bcm2835 mailbox is missing a few properties
that we need for the raspi4 kernel:
 * RPI_FWREQ_GET_CLOCKS
 * RPI_FWREQ_GET_THROTTLED
 * RPI_FWREQ_VCHIQ_INIT

Add minimal implementations of them.

Signed-off-by: Sergey Kambalin <sergey.kambalin@auriga.com>
Message-id: 20240226000259.2752893-40-sergey.kambalin@auriga.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
[PMM: improved commit message]
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>

tests/avocado/boot_linux_console.py: Add Rpi4b boot tests

Signed-off-by: Sergey Kambalin <sergey.kambalin@auriga.com>
Message-id: 20240226000259.2752893-31-sergey.kambalin@auriga.com
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
[PMM: Comment out use of USB, which depends on PCI]
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

hw/arm/bcm2838_peripherals: Add clock_isp stub

Signed-off-by: Sergey Kambalin <sergey.kambalin@auriga.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

hw/arm: Add memory region for BCM2837 RPiVid ASB

Signed-off-by: Sergey Kambalin <sergey.kambalin@auriga.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Message-id: 20240226000259.2752893-13-sergey.kambalin@auriga.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

hw/arm/raspi4b: Temporarily disable unimplemented rpi4b devices

This commit adds RPi4B device tree modifications:
- disable pcie, rng200, thermal sensor and genet devices
  (they're going to be re-enabled in the following commits)
- create additional memory region in device tree
  if RAM amount exceeds VC base address.

Signed-off-by: Sergey Kambalin <sergey.kambalin@auriga.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Message-id: 20240226000259.2752893-12-sergey.kambalin@auriga.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

hw/arm: Introduce Raspberry PI 4 machine

Signed-off-by: Sergey Kambalin <sergey.kambalin@auriga.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Message-id: 20240226000259.2752893-11-sergey.kambalin@auriga.com
[PMM: Change name to 'raspi4b', not 'raspi4b-2g']
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

hw/arm: Add GPIO and SD to BCM2838 periph

Signed-off-by: Sergey Kambalin <sergey.kambalin@auriga.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Message-id: 20240226000259.2752893-10-sergey.kambalin@auriga.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

hw/gpio: Connect SD controller to BCM2838 GPIO

Signed-off-by: Sergey Kambalin <sergey.kambalin@auriga.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Message-id: 20240226000259.2752893-9-sergey.kambalin@auriga.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

hw/gpio: Implement BCM2838 GPIO functionality

Signed-off-by: Sergey Kambalin <sergey.kambalin@auriga.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Message-id: 20240226000259.2752893-8-sergey.kambalin@auriga.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

hw/gpio: Add BCM2838 GPIO stub

Signed-off-by: Sergey Kambalin <sergey.kambalin@auriga.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Message-id: 20240226000259.2752893-7-sergey.kambalin@auriga.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

hw/arm/bcm2838: Add GIC-400 to BCM2838 SoC

Signed-off-by: Sergey Kambalin <sergey.kambalin@auriga.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Message-id: 20240226000259.2752893-6-sergey.kambalin@auriga.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

hw/arm: Introduce BCM2838 SoC

Signed-off-by: Sergey Kambalin <sergey.kambalin@auriga.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Message-id: 20240226000259.2752893-5-sergey.kambalin@auriga.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

hw/arm/raspi: Split out raspi machine common part

Pre-setup for raspberry pi 4 introduction

Signed-off-by: Sergey Kambalin <sergey.kambalin@auriga.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Message-id: 20240226000259.2752893-4-sergey.kambalin@auriga.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

hw/arm/bcm2853_peripherals: Split out common part of peripherals

Pre-setup for BCM2838 introduction

Signed-off-by: Sergey Kambalin <sergey.kambalin@auriga.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Message-id: 20240226000259.2752893-3-sergey.kambalin@auriga.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

hw/arm/bcm2836: Split out common part of BCM283X classes

Pre setup for BCM2838 introduction

Signed-off-by: Sergey Kambalin <sergey.kambalin@auriga.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Message-id: 20240226000259.2752893-2-sergey.kambalin@auriga.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

docs/devel/reset: Update to discuss system reset

Now that system reset uses a three-phase-reset, update the reset
documentation to include a section describing how this works.
Include documentation of the current major beartrap in reset, which
is that only devices on the qbus tree will get automatically reset.

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Message-id: 20240220160622.114437-11-peter.maydell@linaro.org
Reviewed-by: Zhao Liu <zhao1.liu@intel.com>

hw/core/machine: Use qemu_register_resettable for sysbus reset

Move the reset of the sysbus (and thus all devices and buses anywhere
on the qbus tree) from qemu_register_reset() to qemu_register_resettable().

This is a behaviour change: because qemu_register_resettable() is
aware of three-phase reset, this now means that:
 * 'enter' phase reset methods of devices and buses are called
   before any legacy reset callbacks registered with qemu_register_reset()
 * 'exit' phase reset methods of devices and buses are called
   after any legacy qemu_register_reset() callbacks

Put another way, a qemu_register_reset() callback is now correctly
ordered in the 'hold' phase along with any other 'hold' phase methods.

The motivation for doing this is that we will now be able to resolve
some reset-ordering issues using the three-phase mechanism, because
the 'exit' phase is always after the 'hold' phase, even when the
'hold' phase function was registered with qemu_register_reset().

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Message-id: 20240220160622.114437-10-peter.maydell@linaro.org
Reviewed-by: Zhao Liu <zhao1.liu@intel.com>

hw/core/reset: Implement qemu_register_reset via qemu_register_resettable

Reimplement qemu_register_reset() via qemu_register_resettable().

We define a new LegacyReset object which implements Resettable and
defines its reset hold phase method to call a QEMUResetHandler
function.  When qemu_register_reset() is called, we create a new
LegacyReset object and add it to the simulation_reset
ResettableContainer.  When qemu_unregister_reset() is called, we find
the LegacyReset object in the container and remove it.

This implementation of qemu_unregister_reset() means we'll end up
scanning the ResetContainer's list of child objects twice, once
to find the LegacyReset object, and once in g_ptr_array_remove().
In theory we could avoid this by having the ResettableContainer
interface include a resettable_container_remove_with_equal_func()
that took a callback method so that we could use
g_ptr_array_find_with_equal_func() and g_ptr_array_remove_index().
But we don't expect qemu_unregister_reset() to be called frequently
or in hot paths, and we expect the simulation_reset container to
usually not have many children.

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Message-id: 20240220160622.114437-9-peter.maydell@linaro.org
Reviewed-by: Zhao Liu <zhao1.liu@intel.com>

hw/core/reset: Add qemu_{register, unregister}_resettable()

Implement new functions qemu_register_resettable() and
qemu_unregister_resettable().  These are intended to be
three-phase-reset aware equivalents of the old qemu_register_reset()
and qemu_unregister_reset().  Instead of passing in a function
pointer and opaque, you register any QOM object that implements the
Resettable interface.

The implementation is simple: we have a single global instance of a
ResettableContainer, which we reset in qemu_devices_reset(), and
the Resettable objects passed to qemu_register_resettable() are
added to it.

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Message-id: 20240220160622.114437-8-peter.maydell@linaro.org
Reviewed-by: Zhao Liu <zhao1.liu@intel.com>

hw/core: Add ResetContainer which holds objects implementing Resettable

Implement a ResetContainer.  This is a subclass of Object, and it
implements the Resettable interface.  The container holds a list of
arbitrary other objects which implement Resettable, and when the
container is reset, all the objects it contains are also reset.

This will allow us to have a 3-phase-reset equivalent of the old
qemu_register_reset() API: we will have a single "simulation reset"
top level ResetContainer, and objects in it are the equivalent of the
old QEMUResetHandler functions.

The qemu_register_reset() API manages its list of callbacks using a
QTAILQ, but here we use a GPtrArray for our list of Resettable
children: we expect the "remove" operation (which will need to do an
iteration through the list) to be fairly uncommon, and we get simpler
code with fewer memory allocations.

Since there is currently no listed owner in MAINTAINERS for the
existing reset-related source files, create a new section for
them, and add these new files there also.

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Message-id: 20240220160622.114437-7-peter.maydell@linaro.org
Reviewed-by: Zhao Liu <zhao1.liu@intel.com>

hw/core: Add documentation and license comments to reset.h

Add the usual boilerplate license/copyright comment to reset.h (using
the text from reset.c), and document the existing functions.

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Message-id: 20240220160622.114437-6-peter.maydell@linaro.org

include/qom/object.h: New OBJECT_DEFINE_SIMPLE_TYPE{, _WITH_INTERFACES} macros

We have an OBJECT_DEFINE_TYPE_EXTENDED macro, plus several variations
on it, which emits the boilerplate for the TypeInfo and ensures it is
registered with the type system.  However, all the existing macros
insist that the type being defined has its own FooClass struct, so
they aren't useful for the common case of a simple leaf class which
doesn't have any new methods or any other need for its own class
struct (that is, for the kind of type that OBJECT_DECLARE_SIMPLE_TYPE
declares).

Pull the actual implementation of OBJECT_DEFINE_TYPE_EXTENDED out
into a new DO_OBJECT_DEFINE_TYPE_EXTENDED which parameterizes the
value we use for the class_size field.  This lets us add a new
OBJECT_DEFINE_SIMPLE_TYPE which does the same job as the various
existing OBJECT_DEFINE_*_TYPE_* family macros for this kind of simple
type, and the variant OBJECT_DEFINE_SIMPLE_TYPE_WITH_INTERFACES for
when the type will implement some interfaces.

Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Message-id: 20240220160622.114437-5-peter.maydell@linaro.org
Reviewed-by: Zhao Liu <zhao1.liu@intel.com>

system/bootdevice: Don't unregister reset handler in restore_boot_order()

Currently the qemu_register_reset() API permits the reset handler functions
registered with it to remove themselves from within the callback function.
This is fine with our current implementation, but is a bit odd, because
generally reset is supposed to be idempotent, and doesn't fit well in a
three-phase-reset world where a resettable object will get multiple
callbacks as the system is reset.

We now have only one user of qemu_register_reset() which makes use of
the ability to unregister itself within the callback:
restore_boot_order().  We want to change our implementation of
qemu_register_reset() to something where it would be awkward to
maintain the "can self-unregister" feature.  Rather than making that
reimplementation complicated, change restore_boot_order() so that it
doesn't unregister itself but instead returns doing nothing for any
calls after it has done the "restore the boot order" work.

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Acked-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20240220160622.114437-4-peter.maydell@linaro.org
Reviewed-by: Zhao Liu <zhao1.liu@intel.com>

arm/ptw: Handle atomic updates of page tables entries in MMIO during PTW.

I'm far from confident this handling here is correct. Hence
RFC.  In particular not sure on what locks I should hold for this
to be even moderately safe.

The function already appears to be inconsistent in what it returns
as the CONFIG_ATOMIC64 block returns the endian converted 'eventual'
value of the cmpxchg whereas the TCG_OVERSIZED_GUEST case returns
the previous value.

Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Message-id: 20240219161229.11776-1-Jonathan.Cameron@huawei.com
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

arm: xlnx-versal-virt: Add machine property ospi-flash

This property allows users to change flash model on command line as
below.

   ex: "-M xlnx-versal-virt,ospi-flash=mt35xu02gbba"

Signed-off-by: Sai Pavan Boddu <sai.pavan.boddu@amd.com>
Message-id: 20240220091721.82954-3-sai.pavan.boddu@amd.com
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

block: m25p80: Add support of mt35xu02gbba

Add Micro 2Gb OSPI flash part with sfdp data.

Signed-off-by: Sai Pavan Boddu <sai.pavan.boddu@amd.com>
Reviewed-by: Francisco Iglesias <frasse.iglesias@gmail.com>
Message-id: 20240220091721.82954-2-sai.pavan.boddu@amd.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

pl031: Update last RTCLR value on write in case it's read back

The PL031 allows you to read RTCLR, which is meant to give you the last
value written. PL031State has an lr field which is used when reading
from RTCLR, and is present in the VM migration state, but we never
actually update it, so it always reads as its initial 0 value.

Cc: qemu-stable@nongnu.org
Signed-off-by: Jessica Clarke <jrtc27@jrtc27.com>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Message-id: 20240222000341.1562443-1-jrtc27@jrtc27.com
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

tests/qtest: Check that EXTI fan-in irqs are correctly connected

This commit adds a QTest that verifies each input line of a specific
EXTI OR gate can influence the output line.

Signed-off-by: Inès Varhol <ines.varhol@telecom-paris.fr>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Message-id: 20240220184145.106107-3-ines.varhol@telecom-paris.fr
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

hw/arm: Use TYPE_OR_IRQ when connecting STM32L4x5 EXTI fan-in IRQs

Fixes: 52671f69f7a4 ("[PATCH v8 0/3] Add device STM32L4x5 EXTI")
Signed-off-by: Inès Varhol <ines.varhol@telecom-paris.fr>
Message-id: 20240220184145.106107-2-ines.varhol@telecom-paris.fr
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

xlnx-versal-ospi: disable reentrancy detection for iomem_dac

The OSPI DMA reads flash data through the OSPI linear address space (the
iomem_dac region), because of this the reentrancy guard introduced in
commit a2e1753b ("memory: prevent dma-reentracy issues") is disabled for
the memory region.

Signed-off-by: Sai Pavan Boddu <sai.pavan.boddu@amd.com>
Message-id: 20240219105637.65052-1-sai.pavan.boddu@amd.com
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

misc: pxa2xx_timer: replace qemu_system_reset_request() call with watchdog_perform_action()

A few watchdog devices use qemu_system_reset_request(). This is not ideal since
behaviour of watchdog-expiry can't be changed by QMP using `watchdog_action`.
As stated in BiteSizedTasks wiki page, instead of using qemu_system_reset_request()
to reset when a watchdog timer expires, let watchdog_perform_action() decide
what to do.

Resolves: https://gitlab.com/qemu-project/qemu/-/issues/2124
Signed-off-by: Abhiram Tilak <atp.exp@gmail.com>
Message-id: 20240216192612.30838-5-atp.exp@gmail.com
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

misc: m48t59: replace qemu_system_reset_request() call with watchdog_perform_action()

A few watchdog devices use qemu_system_reset_request(). This is not ideal since
behaviour of watchdog-expiry can't be changed by QMP using `watchdog_action`.
As stated in BiteSizedTasks wiki page, instead of using qemu_system_reset_request()
to reset when a watchdog timer expires, let watchdog_perform_action() decide
what to do.

Resolves: https://gitlab.com/qemu-project/qemu/-/issues/2124
Signed-off-by: Abhiram Tilak <atp.exp@gmail.com>
Message-id: 20240216192612.30838-4-atp.exp@gmail.com
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

MAINTAINERS: Cover hw/ide/ahci-allwinner.c with AllWinner A10 machine

This code -- which was moved many times around -- was added in
commit 377e214539 ("ahci: Add allwinner AHCI") and belong to the
AllWinner machines. See also commit dca625768a ("arm: allwinner-a10:
Add SATA").

Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Message-id: 20240215160713.80409-1-philmd@linaro.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

target/arm: Advertise Cortex-A53 erratum #843419 fix via REVIDR

The Cortex-A53 r0p4 revision that QEMU emulates is affected by a CatA
erratum #843419 (i.e., the most severe), which requires workarounds in
the toolchain as well as the OS.

Since the emulation is obviously not affected in the same way, we can
indicate this via REVIDR bit #8, which on r0p4 has the meaning that no
workarounds for erratum #843419 are needed.

Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20240215160202.2803452-1-ardb+git@google.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

hw/arm/sbsa-ref: Simplify init since PCIe is always enabled

There is no point in checking do we have PCIe if first thing after check
is adding PCIe card without checking.

Signed-off-by: Marcin Juszkiewicz <marcin.juszkiewicz@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Message-id: 20240215153311.186772-1-marcin.juszkiewicz@linaro.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

aspeed: fix hardcode boot address 0

In the previous design of ASPEED SOCs QEMU model, it set the boot
address at "0" which was the hardcode setting for ast10x0, ast2600,
ast2500 and ast2400.

According to the design of ast2700, it has a bootmcu(riscv-32) which
is used for executing SPL and initialize DRAM and copy u-boot image
from SPI/Flash to DRAM at address 0x400000000 at SPL boot stage.
Then, CPUs(cortex-a35) execute u-boot, kernel and rofs.

Currently, qemu not support emulate two CPU architectures
at the same machine. Therefore, qemu will only support
to emulate CPU(cortex-a35) side for ast2700 and the boot
address is "0x4 00000000".

Fixed hardcode boot address "0" for future models using
a different mapping address.

Signed-off-by: Troy Lee <troy_lee@aspeedtech.com>
Signed-off-by: Jamin Lin <jamin_lin@aspeedtech.com>
Reviewed-by: Cédric Le Goater <clg@kaod.org>
Signed-off-by: Cédric Le Goater <clg@kaod.org>

aspeed: introduce a new UART0 device name

The Aspeed datasheet refers to the UART controllers
as UART1 - UART13 for the ast10x0, ast2600, ast2500
and ast2400 SoCs and the Aspeed ast2700 introduces an UART0
and the UART controllers as UART0 - UART12.

To keep the naming in the QEMU models
in sync with the datasheet, let's introduce a new  UART0 device name
and do the required adjustements.

Signed-off-by: Troy Lee <troy_lee@aspeedtech.com>
Signed-off-by: Jamin Lin <jamin_lin@aspeedtech.com>
Reviewed-by: Cédric Le Goater <clg@kaod.org>
[ clg: - Kept original assert() in aspeed_soc_uart_set_chr()
       - Fixed 'i' range in connect_serial_hds_to_uarts() loop ]
Signed-off-by: Cédric Le Goater <clg@kaod.org>

Merge tag 'hw-misc-20240227' of https://github.com/philmd/qemu into staging

Misc HW patch queue

hw: Remove sysbus_address_space() (Phil)
hw/nubus: Add nubus-virtio-mmio device (Mark)
hw/usb: Remove usb_bus_find() (Paolo)
hw/usb: Extract sysbus-ohci from ohci (Paolo)
hw/nvme: Fix invalid endian conversion
hw/i386: More PC machine housekeeping (Bernhard, Phil)
hw/ide: Restrict "ide-internal.h" (Zoltan, Phil)
qom: Expose object_resolve_type_unambiguous() (Paolo)

# -----BEGIN PGP SIGNATURE-----
#
# iQIzBAABCAAdFiEE+qvnXhKRciHc/Wuy4+MsLN6twN4FAmXdn3MACgkQ4+MsLN6t
# wN6JhxAAsiLQlbCH+qQw7DvbNBrbYbmDZxx4vlJ2VGFO9SrChFR/2Jnd4z93B1Ry
# IfMyAUm1iu1tgGn9jotGFYsx6tEaL+E8xLdFGNFs85xXNv7q0q3EGlPA0QqNT3Lw
# PJvFk4aBHEDoHzyZk1VwUD3ZyAFcTl6RwcUfCD3/VsqqBBlKJ6mf9tkaGrOqG/cy
# Jjg0jIZOOrWQ77FAlIaPeqZMHfzU556SlKXRyU2TpvC/Q17YsRmwSpJyUr0dNy6A
# VNnBchJpaHxxEFVXrGxu5dkZ6am+pnNWloCILFqqbshpwrpWVsnarB3sbo78L+tV
# tEW3mKGanVpFAzeOT5CKIxHi5V+DSlriKD36XUeOr+A2CBjYkpVAWTluRaAX4BaK
# 9qMkt8AhQdDZd5+Vqvp6UwYfZT1gnPhAbhFNapINGJhtTz5xY2qCLFBBsFYkcf4x
# XvTJr/JgvxyTReH12Sl1HbOooC4Vuw/JpK/YPIA/UA2UfGGCOR+i5z6Lp6dNTswo
# rYv/aUdw2g2zdOHHJqMC9diCu2movfe8+W8o8pK0Z40zbkuLk+zW+j0lcqbEe6w4
# qa32kD75OisvPhcphOEzXOpihdIugqyKkLOCB/+Rg5t+K5hAEBMcGv2ZXvUMq41W
# mI5h9DEcmGxix4lCxmR4Cjxdet/Z3yp85GAoZC3piqURrNLCbM8=
# =2MiE
# -----END PGP SIGNATURE-----
# gpg: Signature made Tue 27 Feb 2024 08:38:11 GMT
# gpg:                using RSA key FAABE75E12917221DCFD6BB2E3E32C2CDEADC0DE
# gpg: Good signature from "Philippe Mathieu-Daudé (F4BUG) <f4bug@amsat.org>" [full]
# Primary key fingerprint: FAAB E75E 1291 7221 DCFD  6BB2 E3E3 2C2C DEAD C0DE

* tag 'hw-misc-20240227' of https://github.com/philmd/qemu: (30 commits)
  hw/ide: Include 'ide-internal.h' from current path
  hw/ide: Remove last two uses of ide/internal.h outside of hw/ide/
  hw/arm/sbsa-ref: Do not open-code ahci_ide_create_devs()
  hw/i386/pc: Populate RTC attribute directly
  hw/i386/pc: Remove unneeded class attribute "kvmclock_enabled"
  hw/i386/pc_{piix, q35}: Eliminate local pci_bus/pci_host variables
  hw/i386/pc: Rename "bus" attribute to "pcibus"
  hw/i386/q35: Include missing 'hw/acpi/acpi.h' header
  hw/i386/sgx: Use QDev API
  hw/i386/x86: Let ioapic_init_gsi() take parent as pointer
  hw/nvme: fix invalid endian conversion
  hw/usb: remove duplicate file in system_ss
  hw/usb: extract sysbus-ohci to a separate file
  hw/usb: remove usb_bus_find
  hw/ppc/pseries: do not require CONFIG_USB
  hw/ppc/mac_newworld: do not require CONFIG_USB
  hw/hppa: do not require CONFIG_USB
  hw/mips/loongson3_virt: do not require CONFIG_USB
  hw/sh4/r2d: do not use usb_bus_find()
  hw/ppc/sam460ex: do not use usb_bus_find()
  ...

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

hw/ide: Include 'ide-internal.h' from current path

Rename "internal.h" as "ide-internal.h", and include
it via its relative local path, instead of absolute
to the project root path.

Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Message-Id: <20240226080632.9596-4-philmd@linaro.org>

hw/ide: Remove last two uses of ide/internal.h outside of hw/ide/

Remove last two includes of hw/ide/intarnal.h outside of hw/ide and
replace them with newly added public header to allow moving internal.h
into hw/ide to really stop exposing it.

Fixes: a11f439a0e (hw/ide: Stop exposing internal.h to non-IDE files)
Signed-off-by: BALATON Zoltan <balaton@eik.bme.hu>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Message-ID: <20240223142633.933694E6004@zero.eik.bme.hu>
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>

hw/arm/sbsa-ref: Do not open-code ahci_ide_create_devs()

Use ahci_ide_create_devs() instead of open-coding it.
Not accessing AHCIDevice internals anymore allows to
remove "hw/ide/ahci_internal.h" (which isn't really a
public header).

Reviewed-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Message-Id: <20240226080632.9596-2-philmd@linaro.org>

hw/i386/pc: Populate RTC attribute directly

Both the piix and the q35 machines introduce an rtc_state variable and defer the
initialization of the X86MachineState::rtc attribute to pc_cmos_init(). Resolve
this complication which makes pc_cmos_init() do what it says on the tin.

Signed-off-by: Bernhard Beschow <shentey@gmail.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Zhao Liu <zhao1.liu@intel.com>
Message-ID: <20240224135851.100361-6-shentey@gmail.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>

hw/i386/pc: Remove unneeded class attribute "kvmclock_enabled"

PCMachineClass introduces the attribute into the class hierarchy and sets it to
true. There is no sub class overriding the attribute. Commit 30d2a17b46e9
"hw/i386: Remove the deprecated machines 0.12 up to 0.15" removed the last
overrides of this attribute. The attribute is now unneeded and can be removed.

Fixes: 30d2a17b46e9 "hw/i386: Remove the deprecated machines 0.12 up to 0.15"
Cc: Thomas Huth <thuth@redhat.com>
Signed-off-by: Bernhard Beschow <shentey@gmail.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Zhao Liu <zhao1.liu@intel.com>
Message-ID: <20240224135851.100361-5-shentey@gmail.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>

hw/i386/pc_{piix, q35}: Eliminate local pci_bus/pci_host variables

There is no advantage in having these local variables which 1/ needlessly have
different identifiers in both machines and 2/ which are redundant to pcms->bus
which is almost as short.

Signed-off-by: Bernhard Beschow <shentey@gmail.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Zhao Liu <zhao1.liu@intel.com>
Message-ID: <20240224135851.100361-4-shentey@gmail.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>

hw/i386/pc: Rename "bus" attribute to "pcibus"

The attribute is of type PCIBus; reflect that in the name. It will also make the
next change more intuitive.

Suggested-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Bernhard Beschow <shentey@gmail.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Reviewed-by: Zhao Liu <zhao1.liu@intel.com>
Message-ID: <20240224135851.100361-3-shentey@gmail.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>

hw/i386/q35: Include missing 'hw/acpi/acpi.h' header

"hw/acpi/acpi.h" is implicitly included. Include it
explicitly to avoid the following error when refactoring
headers:

hw/i386/pc_q35.c:209:43: error: use of undeclared identifier 'ACPI_PM_PROP_ACPI_PCIHP_BRIDGE'
                              ACPI_PM_PROP_ACPI_PCIHP_BRIDGE,
                              ^

Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Reviewed-by: Zhao Liu <zhao1.liu@intel.com>
Message-Id: <20240226090600.31952-3-philmd@linaro.org>

hw/i386/sgx: Use QDev API

Prefer the QDev API over the low level QOM one.
No logical change intended.

Reviewed-by: Michael Tokarev <mjt@tls.msk.ru>
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Message-Id: <20240215142035.73331-4-philmd@linaro.org>

hw/i386/x86: Let ioapic_init_gsi() take parent as pointer

Rather than taking a QOM name which has to be resolved, let's pass the parent
directly as pointer. This simplifies the code.

Signed-off-by: Bernhard Beschow <shentey@gmail.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Reviewed-by: Zhao Liu <zhao1.liu@intel.com>
Message-ID: <20240224135851.100361-2-shentey@gmail.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>

hw/nvme: fix invalid endian conversion

numcntl is one byte and so is max_vfs. Using cpu_to_le16 on big endian
hosts results in numcntl being set to 0.

Fix by dropping the endian conversion.

Fixes: 99f48ae7ae ("hw/nvme: Add support for Secondary Controller List")
Reported-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Klaus Jensen <k.jensen@samsung.com>
Reviewed-by: Minwoo Im <minwoo.im@samsung.com>
Message-ID: <20240222-fix-sriov-numcntl-v1-1-d60bea5e72d0@samsung.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>

hw/usb: remove duplicate file in system_ss

Because USB_EHCI_SYSBUS selects USB_EHCI, there is no need to include
hcd-ehci.c explicitly.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Message-ID: <20240223124406.234509-11-pbonzini@redhat.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>

hw/usb: extract sysbus-ohci to a separate file

Split the sysbus version to a separate file so that it is not
included in PCI-only machines, and adjust Kconfig for machines
that do need sysbus-ohci.  The copyrights are based on the
time and employer of balrog and Paul Brook's contributions.

While adjusting the SM501 dependency, move it to the right place
instead of keeping it in the R4D machine.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Message-ID: <20240223124406.234509-10-pbonzini@redhat.com>
[PMD: Rename some functions using 'ohci_sysbus_' prefix]
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>

hw/usb: remove usb_bus_find

Inline the sole remaining use, which is for the -usbdevice command line.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Message-ID: <20240223124406.234509-9-pbonzini@redhat.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>

hw/ppc/pseries: do not require CONFIG_USB

With --without-default-devices it is possible to build a binary that
does not include any USB host controller and therefore that does not
include the code guarded by CONFIG_USB.  While the simpler creation
functions such as usb_create_simple can be inlined, this is not true
of usb_bus_find().  Remove it, replacing it with a search of the single
USB bus on the machine.

Suggested-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Message-ID: <20240223124406.234509-8-pbonzini@redhat.com>
[PMD: Fixed style]
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>

hw/ppc/mac_newworld: do not require CONFIG_USB

With --without-default-devices it should not be required to have
devices in the binary that are removed by -nodefaults.  It should be
therefore possible to build a binary that does not include any USB
host controller or any of the code guarded by CONFIG_USB.  While the
simpler creation functions such as usb_create_simple can be inlined,
this is not true of usb_bus_find().  Remove it, replacing it with a
search of the single USB bus on the machine.

With this change, it is possible to change "select USB_OHCI_PCI" into
an "imply" directive.

Suggested-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Message-ID: <20240223124406.234509-7-pbonzini@redhat.com>
[PMD: Fixed style]
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>

hw/hppa: do not require CONFIG_USB

With --without-default-devices it is possible to build a binary that
does not include any USB host controller and therefore that does not
include the code guarded by CONFIG_USB.  While the simpler creation
functions such as usb_create_simple can be inlined, this is not true
of usb_bus_find().  Remove it, replacing it with a search of the single
USB bus on the machine.

Suggested-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Message-ID: <20240223124406.234509-6-pbonzini@redhat.com>
[PMD: Fixed style]
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>

hw/mips/loongson3_virt: do not require CONFIG_USB

Once the Kconfig for hw/mips is cleaned up, it will be possible to build a
binary that does not include any USB host controller and therefore that
does not include the code guarded by CONFIG_USB.  While the simpler
creation functions such as usb_create_simple can be inlined, this is not
true of usb_bus_find().  Remove it, replacing it with a search of the
single USB bus created by loongson3_virt_devices_init().

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-ID: <20240223124406.234509-5-pbonzini@redhat.com>
[PMD: Fixed style]
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>

hw/sh4/r2d: do not use usb_bus_find()

usb_bus_find() is always used with argument -1; it can be replaced with
a search of the single USB bus on the machine.

Suggested-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Message-ID: <20240223124406.234509-4-pbonzini@redhat.com>
[PMD: Fixed style]
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>

hw/ppc/sam460ex: do not use usb_bus_find()

usb_bus_find() is always used with argument -1; it can be replaced with
a search of the single USB bus on the machine.

Suggested-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Message-ID: <20240223124406.234509-3-pbonzini@redhat.com>
[PMD: Fixed style]
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>

hw/acpi: move object_resolve_type_unambiguous to core QOM

object_resolve_type_unambiguous provides a useful functionality, that
is currently emulated for example by usb_bus_find().  Move it to core
code and add error reporting for increased generality.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Message-ID: <20240223124406.234509-2-pbonzini@redhat.com>
[PMD: Fixed style]
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>

hw/nubus: add nubus-virtio-mmio device

The nubus-virtio-mmio device is a Nubus card that contains a set of 32 virtio-mmio
devices and a goldfish PIC similar to the m68k virt machine that can be plugged
into the m68k q800 machine.

There are currently a number of drivers under development that can be used in
conjunction with this device to provide accelerated and/or additional hypervisor
services to 68k Classic MacOS.

Signed-off-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>
Reviewed-by: Laurent Vivier <laurent@vivier.eu>
Message-ID: <20240111102954.449462-4-mark.cave-ayland@ilande.co.uk>
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>

hw/nubus: increase maximum Declaration ROM size from 128k to 1Mb

Whilst 128k is more than enough for a typical Declaration ROM, a C compiler
configured to produce an unstripped debug binary can generate a ROM image that
exceeds this limit. Increase the maximum size to 1Mb to help make life easier
for developers.

Signed-off-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>
Reviewed-by: Laurent Vivier <laurent@vivier.eu>
Message-ID: <20240111102954.449462-3-mark.cave-ayland@ilande.co.uk>
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>

hw/nubus-device: round Declaration ROM memory region address to qemu_target_page_size()

Declaration ROM binary images can be any arbitrary size, however if a host ROM
memory region is not aligned to qemu_target_page_size() then we fail the
"assert(!(iotlb & ~TARGET_PAGE_MASK))" check in tlb_set_page_full().

Ensure that the host ROM memory region is aligned to qemu_target_page_size()
and adjust the offset at which the Declaration ROM image is loaded, since Nubus
ROM images are unusual in that they are aligned to the end of the slot address
space.

Signed-off-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Reviewed-by: Laurent Vivier <laurent@vivier.eu>
Message-ID: <20240111102954.449462-2-mark.cave-ayland@ilande.co.uk>
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>

Merge tag 'edk2-stable202402-20240226-pull-request' of https://gitlab.com/kraxel/qemu into staging

firmware: update to edk2-stable202402

Update edk2 to the latest release tagged end of last week.

Cc stable this time because we should move away from the
git snapshot update done in January.

# -----BEGIN PGP SIGNATURE-----
#
# iQIzBAABCgAdFiEEoDKM/7k6F6eZAf59TLbY7tPocTgFAmXctQ0ACgkQTLbY7tPo
# cTjOkA/+OIMaKRl+ZQAQQYlfLoJK/XK51QHyZugzehWs1o5LBHh4aWrtcxTLhPQb
# Zvofzm59XNb5drBuZIZbdso6hFt7jZbzw7VsIJW/CFlqIJTGDffdmmJzPoGVoHlj
# SX9aVFlLT2Gu5qlRG5yaZKQF3gnJDKmOSfM8DJxu9yvxcY+Y812k1LV1M0zoedBR
# gx2283UISUQcT/3iazCWRpJVxV60i7iN4BsHUUgiJNOUIlMoNeFPZxL2OZl4m72+
# bA0EklTgMcCesctsDQvF+Ql30mY/KTtK7CtEhVvB13n87sDj73YuQng9lsL4x2oY
# NiokFlTKbvoDek4hjuyfHuf8MwwFPUHMHTFa5A61Er4UaiejgNE1I6rnyoIDAKz/
# bcZm3jYlu4La32mVsXyVxW+5812cR0PE6hZqBPvyMl2VtXhWTXLMjseTM8TzHo5a
# u/CYvKF7gDppoaPsy/sASC4Q4ko49U83BYC/ftJQHBmkIslI1jeP9wfqXOQk40Cn
# H/GyfTQZvam9Al6nxh++cU3cUoux3YQpNW5RE5Au4qizH/kyO8vFHPs4e4jbaOlc
# pI5g2bQESMM14tR5xEN5EKuxs0+GOlnzziCsWaVjBCU7s6YsFHclLSK/0zf2RUas
# uNiI6jjsgojUfpm896Xp6zaXUDWiQ6oyMdFBa72S8XJmzOufPV0=
# =4SrN
# -----END PGP SIGNATURE-----
# gpg: Signature made Mon 26 Feb 2024 15:58:05 GMT
# gpg:                using RSA key A0328CFFB93A17A79901FE7D4CB6D8EED3E87138
# gpg: Good signature from "Gerd Hoffmann (work) <kraxel@redhat.com>" [full]
# gpg:                 aka "Gerd Hoffmann <gerd@kraxel.org>" [full]
# gpg:                 aka "Gerd Hoffmann (private) <kraxel@gmail.com>" [full]
# Primary key fingerprint: A032 8CFF B93A 17A7 9901  FE7D 4CB6 D8EE D3E8 7138

* tag 'edk2-stable202402-20240226-pull-request' of https://gitlab.com/kraxel/qemu:
  update edk2 binaries to edk2-stable202402
  update edk2 submodule to edk2-stable202402

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

hw/sysbus: Remove now unused sysbus_address_space()

sysbus_address_space() is not more used, remove it.

Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20240216153517.49422-7-philmd@linaro.org>

hw/i386/kvmvapic: Inline sysbus_address_space()

sysbus_address_space(...) is a simple wrapper to
get_system_memory(). Use it in place, since KVM
VAPIC doesn't distinct address spaces.

Rename the 'as' variable as 'mr' since it is a
MemoryRegion type, not an AddressSpace one.

Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20240216153517.49422-6-philmd@linaro.org>

hw/display/exynos4210_fimd: Pass frame buffer memory region as link

Add the Exynos4210fimdState::'framebuffer-memory' property. Have
the board set it. We don't need to call sysbus_address_space()
anymore.

Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20240226173805.289-5-philmd@linaro.org>

hw/arm/exynos4210: Inline sysbus_create_varargs(EXYNOS4210_FIMD)

We want to set another qdev property (a link) for the FIMD
device, we can not use sysbus_create_varargs() which only
passes sysbus base address and IRQs as arguments. Inline
it so we can set the link property in the next commit.

Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20240216153517.49422-4-philmd@linaro.org>

hw/display/pl110: Pass frame buffer memory region as link property

Add the PL110::'framebuffer-memory' property. Have the different
ARM boards set it. We don't need to call sysbus_address_space()
anymore.

Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20240226173805.289-3-philmd@linaro.org>

hw/arm: Inline sysbus_create_simple(PL110 / PL111)

We want to set another qdev property (a link) for the pl110
and pl111 devices, we can not use sysbus_create_simple() which
only passes sysbus base address and IRQs as arguments. Inline
it so we can set the link property in the next commit.

Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20240226173805.289-2-philmd@linaro.org>

Merge tag 'pull-qapi-2024-02-26' of https://repo.or.cz/qemu/armbru into staging

QAPI patches patches for 2024-02-26

# -----BEGIN PGP SIGNATURE-----
#
# iQJGBAABCAAwFiEENUvIs9frKmtoZ05fOHC0AOuRhlMFAmXcXWUSHGFybWJydUBy
# ZWRoYXQuY29tAAoJEDhwtADrkYZTMqQP/3aZ8Jm48Bh4sG5JxJ01aM3iX//cuA8X
# 0xoKBfDWkwKvUDE+sv+2qH1WxcNGS4wb/lLaT6Jf0S8EzzelDx5zYGwXXsmBljoZ
# Kh8BI7Oe1SQwI5Z36q/efG9isQ6nqGyxR9xu1BTbJCSOZXzxpOBljbSyXOtKc5C/
# HM4TK82geLLgm04gMkoPmFdu3BXDfPYnGJ3UNLE9hTPoW7XL4EGFAOOWuDcrSF5n
# OdHjfK9TK9PcxsJGqVUqLHwfRQYLMBni6OkurdFOVdLM1v4C707NuryjaGQc1WEI
# xKwsJDKR+zG7vGu4y594HN/Ivoaqci8MMTbDgVmHZ3LaI3RUfSplGTDSYLjCp8jz
# XDx82+hhmb/2ZMmE0tarUQyv0dimrZSEh6cWWHMvp63edKTywoB/eIDR9lBteTZe
# wRvkSKmN6oKJI8cNiiXZqw5y2JPvhNag4Xdr8kHKwHgxVWP+SneInLCC+T2SMgio
# EeC+S4CVTdjPvEC96dOGrsqKn+gl/h74PK5ZdTaD1B6XCuIalsRn6REujqW6Ew6n
# rj7Iec/noejeOsflzBWRKT91t2Zck/MRLhX9nYqybBxyxUFvFS7M6ok/iq4oEtZR
# lJooF6iiq8xtEzoLselfGFAZTUxhwLdUfXPVDx7p5HDpJci88xv6zmav9eE84JbH
# mBD55GEH17ka
# =81Zq
# -----END PGP SIGNATURE-----
# gpg: Signature made Mon 26 Feb 2024 09:44:05 GMT
# gpg:                using RSA key 354BC8B3D7EB2A6B68674E5F3870B400EB918653
# gpg:                issuer "armbru@redhat.com"
# gpg: Good signature from "Markus Armbruster <armbru@redhat.com>" [full]
# gpg:                 aka "Markus Armbruster <armbru@pond.sub.org>" [full]
# Primary key fingerprint: 354B C8B3 D7EB 2A6B 6867  4E5F 3870 B400 EB91 8653

* tag 'pull-qapi-2024-02-26' of https://repo.or.cz/qemu/armbru:
  qapi: Divorce QAPIDoc from QAPIParseError
  qapi: Reject multiple and empty feature descriptions
  qapi: Rewrite doc comment parser
  qapi: Merge adjacent untagged sections
  qapi: Call QAPIDoc.check() always
  qapi: Recognize section tags and 'Features:' only after blank line
  qapi: Require descriptions and tagged sections to be indented
  qapi: Reject section heading in the middle of a doc comment
  qapi: Rename QAPIDoc.Section.name to .tag
  qapi: Improve error message for empty doc sections
  qapi: Improve error position for bogus invalid "Returns" section
  qapi: Improve error position for bogus argument descriptions
  sphinx/qapidoc: Drop code to generate doc for simple union branch
  tests/qapi-schema: Cover 'Features:' not followed by descriptions
  tests/qapi-schema: Cover duplicate 'Features:' line
  tests/qapi-schema: Fix test 'QAPI rST doc'
  qapi: Misc cleanups to migrate QAPIs

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

qapi: Divorce QAPIDoc from QAPIParseError

QAPIDoc stores a reference to QAPIParser just to pass it to
QAPIParseError.  The resulting error position depends on the state of
the parser.  It happens to be the current comment line.  Servicable,
but action at a distance.

The commit before previous moved most uses of QAPIParseError from
QAPIDoc to QAPIParser.  There are just three left.  Convert them to
QAPISemError.  This involves passing info to a few methods.  Then drop
the reference to QAPIParser.

The three errors lose the column number.  Not really interesting here:
it's the comment line's indentation.

Signed-off-by: Markus Armbruster <armbru@redhat.com>
Message-ID: <20240216145841.2099240-17-armbru@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>

qapi: Reject multiple and empty feature descriptions

The parser recognizes only the first "Features:" line.  Any subsequent
ones are treated as ordinary text, as visible in test case
doc-duplicate-features.  Recognize "Features:" lines anywhere.  A
second one is an error.

A 'Features:' line without any features is useless, but not an error.
Make it an error.  This makes detecting a second "Features:" line
easier.

qapi/run-state.json actually has an instance of this since commit
fe17522d854 (qapi: Remove deprecated 'singlestep' member of
StatusInfo).  Clean it up.

Signed-off-by: Markus Armbruster <armbru@redhat.com>
Message-ID: <20240216145841.2099240-16-armbru@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>

qapi: Rewrite doc comment parser

QAPISchemaParser is a conventional recursive descent parser.  Except
QAPISchemaParser.get_doc() delegates most of the doc comment parsing
work to a state machine in QAPIDoc.  The state machine doesn't get
tokens like a recursive descent parser, it is fed tokens.

I find this state machine rather opaque and hard to maintain.

Replace it by a conventional parser, all in QAPISchemaParser.  Less
code, and (at least in my opinion) easier to understand.

Signed-off-by: Markus Armbruster <armbru@redhat.com>
Message-ID: <20240216145841.2099240-15-armbru@redhat.com>
Tested-by: Daniel P. Berrangé <berrange@redhat.com>

qapi: Merge adjacent untagged sections

The parser mostly doesn't create adjacent untagged sections, and
merging the ones it does create is hardly worth the bother.  I'm doing
it to avoid behavioral change in the next commit.

Signed-off-by: Markus Armbruster <armbru@redhat.com>
Message-ID: <20240216145841.2099240-14-armbru@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>

qapi: Call QAPIDoc.check() always

We currently call QAPIDoc.check() only for definition documentation.
Calling it for free-form documentation as well is simpler.  No change,
because it doesn't actually do anything there.

Signed-off-by: Markus Armbruster <armbru@redhat.com>
Message-ID: <20240216145841.2099240-13-armbru@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>

qapi: Recognize section tags and 'Features:' only after blank line

Putting a blank line before section tags and 'Features:' is good,
existing practice.  Enforce it.

Signed-off-by: Markus Armbruster <armbru@redhat.com>
Message-ID: <20240216145841.2099240-12-armbru@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>

qapi: Require descriptions and tagged sections to be indented

By convention, we indent the second and subsequent lines of
descriptions and tagged sections, except for examples.

Turn this into a hard rule, and apply it to examples, too.

Signed-off-by: Markus Armbruster <armbru@redhat.com>
Message-ID: <20240216145841.2099240-11-armbru@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
[Straightforward conflicts in qapi/migration.json resolved]

qapi: Reject section heading in the middle of a doc comment

docs/devel/qapi-code-gen.txt claims "A heading line must be the first
line of the documentation comment block" since commit
55ec69f8b16 (docs/devel/qapi-code-gen.txt: Update to new rST backend
conventions).  Not true, we have code to make it work anywhere in a
free-form doc comment: commit dcdc07a97cb (qapi: Make section headings
start a new doc comment block).

Make it true, for simplicity's sake.

Signed-off-by: Markus Armbruster <armbru@redhat.com>
Message-ID: <20240216145841.2099240-10-armbru@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>

qapi: Rename QAPIDoc.Section.name to .tag

Since the previous commit, QAPIDoc.Section.name is either
None (untagged section) or the section's tag string ('Returns',
'@name', ...).  Rename it to .tag.

Signed-off-by: Markus Armbruster <armbru@redhat.com>
Message-ID: <20240216145841.2099240-9-armbru@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>

qapi: Improve error message for empty doc sections

Improve the message for an empty tagged section from

    empty doc section 'Note'

to

    text required after 'Note:'

and the message for an empty argument or feature description from

    empty doc section 'foo'

to

    text required after '@foo:'

Improve the error position to refer to the beginning of the empty
section instead of its end.

Signed-off-by: Markus Armbruster <armbru@redhat.com>
Message-ID: <20240216145841.2099240-8-armbru@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>