migration/ram: Yield periodically to the main loop

Usually, incoming migration coroutine yields to the main loop
while its IO-channel is waiting for data to receive. But there is a case
when RAM migration and data receive have the same speed: VM with huge
zeroed RAM. In this case, IO-channel won't read and thus the main loop
is stuck and for instance, it doesn't respond to QMP commands.

For this case, yield periodically, but not too often, so as not to
affect the speed of migration.

Signed-off-by: Yury Kotov <yury-kotov@yandex-team.ru>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>

migration: savevm_state_handler_insert: constant-time element insertion

savevm_state's SaveStateEntry TAILQ is a priority queue.  Priority
sorting is maintained by searching from head to tail for a suitable
insertion spot.  Insertion is thus an O(n) operation.

If we instead keep track of the head of each priority's subqueue
within that larger queue we can reduce this operation to O(1) time.

savevm_state_handler_remove() becomes slightly more complex to
accomodate these gains: we need to replace the head of a priority's
subqueue when removing it.

With O(1) insertion, booting VMs with many SaveStateEntry objects is
more plausible.  For example, a ppc64 VM with maxmem=8T has 40000 such
objects to insert.

Signed-off-by: Scott Cheloha <cheloha@linux.vnet.ibm.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>

migration: add savevm_state_handler_remove()

Create a function to abstract common logic needed when removing a
SaveStateEntry element from the savevm_state.handlers queue.

For now we just remove the element.  Soon it will involve additional
cleanup.

Signed-off-by: Scott Cheloha <cheloha@linux.vnet.ibm.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>

misc: use QEMU_IS_ALIGNED

Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Stefan Berger <stefanb@linux.ibm.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>

migration: Fix the re-run check of the migrate-incoming command

The current check sets an error but doesn't fail the command.
This may cause a problem if new connection attempt by the same URI
affects the first connection.

Signed-off-by: Yury Kotov <yury-kotov@yandex-team.ru>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Darren Kenny <darren.kenny@oracle.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>

migration: Fix incorrect integer->float conversion caught by clang

Clang does not like qmp_migrate_set_downtime()'s code to clamp double
@value to 0..INT64_MAX:

    qemu/migration/migration.c:2038:24: error: implicit conversion from 'long' to 'double' changes value from 9223372036854775807 to 9223372036854775808 [-Werror,-Wimplicit-int-float-conversion]

The warning will be enabled by default in clang 10. It is not
available for clang <= 9.

The clamp is actually useless; @value is checked to be within
0..MAX_MIGRATE_DOWNTIME_SECONDS immediately before.  Delete it.

While there, make the conversion from double to int64_t explicit.

Signed-off-by: Fangrui Song <i@maskray.me>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
[Patch split, commit message improved]
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>

migration: Rate limit inside host pages

When using hugepages, rate limiting is necessary within each huge
page, since a 1G huge page can take a significant time to send, so
you end up with bursty behaviour.

Fixes: 4c011c37ecb3 ("postcopy: Send whole huge pages")
Reported-by: Lin Ma <LMa@suse.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Peter Xu <peterx@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>

ram.c: remove unneeded labels

ram_save_queue_pages() has an 'err' label that can be replaced by
'return -1' instead.

Same thing with ram_discard_range(), and in this case we can also
get rid of the 'ret' variable and return either '-1' on error
or the result of ram_block_discard_range().

CC: Juan Quintela <quintela@redhat.com>
CC: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>

runstate: ignore finishmigrate -> prelaunch transition

Commit 1bd71dce4bf2 tries to prevent a finishmigrate -> prelaunch
transition by exiting at the beginning of the main_loop_should_exit()
function if the state is already finishmigrate.

As the finishmigrate state is set in the migration thread it can
happen concurrently to the function. The migration thread and the
function are normally protected by the iothread mutex and thus the
state should no evolve between the start of the function and its end.

Unfortunately during the function life the lock is released by
pause_all_vcpus() just before the point we need to be sure we are
not in finishmigrate state and if the migration thread is waiting
for the lock it will take the opportunity to change the state
to finishmigrate.

The only way to be sure we are not in the finishmigrate state when
we need is to check the state after the pause_all_vcpus() function.

Fixes: 1bd71dce4bf2 ("runstate: ignore exit request in finish migrate state")
Signed-off-by: Laurent Vivier <lvivier@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>

migration-test: introduce functions to handle string parameters

Signed-off-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>

migration: Make sure that we don't call write() in case of error

If we are exiting due to an error/finish/.... Just don't try to even
touch the channel with one IO operation.

Signed-off-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>

migration-test: Add migration multifd test

We set multifd-channels.

Signed-off-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Tested-by: Wei Yang <richardw.yang@linux.intel.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>

multifd: Initialize local variable

Fill everything with zero, so the padding fields are also initialized.

Signed-off-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>

Merge remote-tracking branch 'remotes/pmaydell/tags/pull-target-arm-20200117-1' into staging

Add model of the Netduino Plus 2 board
Some allwinner-a10 code cleanup
New test cases for cubieboard
target/arm/arm-semi: fix SYS_OPEN to return nonzero filehandle
i.MX: add an emulation for RNGC device
target/arm: adjust program counter for wfi exception in AArch32
arm/gicv3: update virtual irq state after IAR register read
Set IL bit correctly for syndrome information for data aborts

# gpg: Signature made Fri 17 Jan 2020 14:27:40 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]
# Primary key fingerprint: E1A5 C593 CD41 9DE2 8E83  15CF 3C25 25ED 1436 0CDE

* remotes/pmaydell/tags/pull-target-arm-20200117-1:
  target/arm: Set ISSIs16Bit in make_issinfo
  target/arm: Return correct IL bit in merge_syn_data_abort
  arm/gicv3: update virtual irq state after IAR register read
  target/arm: adjust program counter for wfi exception in AArch32
  i.MX: add an emulation for RNGC
  target/arm/arm-semi: fix SYS_OPEN to return nonzero filehandle
  hw/arm/allwinner-a10: Remove local qemu_irq variables
  hw/arm/allwinner-a10: Simplify by passing IRQs with qdev_pass_gpios()
  hw/arm/allwinner-a10: Move SoC definitions out of header
  tests/boot_linux_console: Add a SD card test for the CubieBoard
  tests/boot_linux_console: Add initrd test for the CubieBoard
  hw/arm: Add the Netduino Plus 2
  hw/arm: Add the STM32F4xx SoC
  hw/misc: Add the STM32F4xx EXTI device
  hw/misc: Add the STM32F4xx Sysconfig device

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

Merge remote-tracking branch 'remotes/huth-gitlab/tags/pull-request-2020-01-17' into staging

* Various fixes for qtests
* Enable TCG tests with TCI in the gitlab CI

# gpg: Signature made Fri 17 Jan 2020 10:35:17 GMT
# gpg:                using RSA key 27B88847EEE0250118F3EAB92ED9D774FE702DB5
# gpg:                issuer "thuth@redhat.com"
# gpg: Good signature from "Thomas Huth <th.huth@gmx.de>" [full]
# gpg:                 aka "Thomas Huth <thuth@redhat.com>" [full]
# gpg:                 aka "Thomas Huth <huth@tuxfamily.org>" [full]
# gpg:                 aka "Thomas Huth <th.huth@posteo.de>" [unknown]
# Primary key fingerprint: 27B8 8847 EEE0 2501 18F3  EAB9 2ED9 D774 FE70 2DB5

* remotes/huth-gitlab/tags/pull-request-2020-01-17:
  gitlab-ci.yml: Run tcg test with tci
  tests/qtest/vhost-user-test: Fix memory leaks
  migration-test: ppc64: fix FORTH test program
  tests: acpi: update path in rebuild-expected-aml

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

target/arm: Set ISSIs16Bit in make_issinfo

During the conversion to decodetree, the setting of
ISSIs16Bit got lost.  This causes the guest os to
incorrectly adjust trapping memory operations.

Cc: qemu-stable@nongnu.org
Fixes: 46beb58efbb8a2a32 ("target/arm: Convert T16, load (literal)")
Reported-by: Jeff Kubascik <jeff.kubascik@dornerworks.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20200117004618.2742-3-richard.henderson@linaro.org
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

target/arm: Return correct IL bit in merge_syn_data_abort

The IL bit is set for 32-bit instructions, thus passing false
with the is_16bit parameter to syn_data_abort_with_iss() makes
a syn mask that always has the IL bit set.

Pass is_16bit as true to make the initial syn mask have IL=0,
so that the final IL value comes from or'ing template_syn.

Cc: qemu-stable@nongnu.org
Fixes: aaa1f954d4ca ("target-arm: A64: Create Instruction Syndromes for Data Aborts")
Signed-off-by: Jeff Kubascik <jeff.kubascik@dornerworks.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20200117004618.2742-2-richard.henderson@linaro.org
[rth: Extracted this as a self-contained bug fix from a larger patch]
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

arm/gicv3: update virtual irq state after IAR register read

The IAR0/IAR1 register is used to acknowledge an interrupt - a read of the
register activates the highest priority pending interrupt and provides its
interrupt ID. Activating an interrupt can change the CPU's virtual interrupt
state - this change makes sure the virtual irq state is updated.

Signed-off-by: Jeff Kubascik <jeff.kubascik@dornerworks.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-id: 20200113154607.97032-1-jeff.kubascik@dornerworks.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

target/arm: adjust program counter for wfi exception in AArch32

The wfi instruction can be configured to be trapped by a higher exception
level, such as the EL2 hypervisor. When the instruction is trapped, the
program counter should contain the address of the wfi instruction that
caused the exception. The program counter is adjusted for this in the wfi op
helper function.

However, this correction is done to env->pc, which only applies to AArch64
mode. For AArch32, the program counter is stored in env->regs[15]. This
adds an if-else statement to modify the correct program counter location
based on the the current CPU mode.

Signed-off-by: Jeff Kubascik <jeff.kubascik@dornerworks.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

i.MX: add an emulation for RNGC

Add an emulation for the RNGC random number generator and the compatible
RNGB variant. These peripherals are included (at least) in imx25 and
imx35 chipsets.

The emulation supports the initial self test, reseeding the prng and
reading random numbers.

Signed-off-by: Martin Kaiser <martin@kaiser.cx>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

target/arm/arm-semi: fix SYS_OPEN to return nonzero filehandle

According to the specification "Semihosting for AArch32 and Aarch64",
the SYS_OPEN operation should return:

 - A nonzero handle if the call is successful
 - -1 if the call is not successful

So, it should never return 0.

Prior to commit 35e9a0a8ce4b ("target/arm/arm-semi: Make semihosting
code hand out its own file descriptors"), the guest fd matched to the
host fd. It returned a nonzero handle on success since the fd 0 is
already used for stdin.

Now that the guest fd is the index of guestfd_array, it starts from 0.

I noticed this issue particularly because Trusted Firmware-A built with
PLAT=qemu is no longer working. Its io_semihosting driver only handles
a positive return value as a valid filehandle.

Basically, there are two ways to fix this:

  - Use (guestfd - 1) as the index of guestfs_arrary. We need to insert
    increment/decrement to convert the guestfd and the array index back
    and forth.

  - Keep using guestfd as the index of guestfs_array. The first entry
    of guestfs_array is left unused.

I thought the latter is simpler. We end up with wasting a small piece
of memory for the unused first entry of guestfd_array, but this is
probably not a big deal.

Fixes: 35e9a0a8ce4b ("target/arm/arm-semi: Make semihosting code hand out its own file descriptors")
Cc: qemu-stable@nongnu.org
Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20200109041228.10131-1-masahiroy@kernel.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

hw/arm/allwinner-a10: Remove local qemu_irq variables

We won't reuse the CPU IRQ/FIQ variables. Simplify by calling
qdev_get_gpio_in() in place.

Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-id: 20191230110953.25496-6-f4bug@amsat.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>

hw/arm/allwinner-a10: Simplify by passing IRQs with qdev_pass_gpios()

By calling qdev_pass_gpios() we don't need to hold a copy of the
IRQs from the INTC into the SoC state.
Instead of filling an array of qemu_irq and passing it around, we
can now directly call qdev_get_gpio_in() on the SoC.

Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-id: 20191230110953.25496-5-f4bug@amsat.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>

hw/arm/allwinner-a10: Move SoC definitions out of header

These definitions are specific to the A10 SoC and don't need
to be exported to the different Allwinner peripherals.

Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-id: 20191230110953.25496-4-f4bug@amsat.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>

tests/boot_linux_console: Add a SD card test for the CubieBoard

The kernel image and DeviceTree blob are built by the Armbian
project (based on Debian):
https://docs.armbian.com/Developer-Guide_Build-Preparation/

The cpio image used comes from the linux-build-test project:
https://github.com/groeck/linux-build-test

If ARM is a target being built, "make check-acceptance" will
automatically include this test by the use of the "arch:arm" tags.

Alternatively, this test can be run using:

  $ avocado --show=console run -t machine:cubieboard tests/acceptance/boot_linux_console.py
  console: Uncompressing Linux... done, booting the kernel.
  console: Booting Linux on physical CPU 0x0
  console: Linux version 4.20.7-sunxi (root@armbian.com) (gcc version 7.2.1 20171011 (Linaro GCC 7.2-2017.11)) #5.75 SMP Fri Feb 8 09:02:10 CET 2019
  [...]
  console: ahci-sunxi 1c18000.sata: Linked as a consumer to regulator.4
  console: ahci-sunxi 1c18000.sata: controller can't do 64bit DMA, forcing 32bit
  console: ahci-sunxi 1c18000.sata: AHCI 0001.0000 32 slots 1 ports 1.5 Gbps 0x1 impl platform mode
  console: ahci-sunxi 1c18000.sata: flags: ncq only
  console: scsi host0: ahci-sunxi
  console: ata1: SATA max UDMA/133 mmio [mem 0x01c18000-0x01c18fff] port 0x100 irq 27
  console: of_cfs_init
  console: of_cfs_init: OK
  console: vcc3v0: disabling
  console: vcc5v0: disabling
  console: usb1-vbus: disabling
  console: usb2-vbus: disabling
  console: ata1: SATA link up 1.5 Gbps (SStatus 113 SControl 300)
  console: ata1.00: ATA-7: QEMU HARDDISK, 2.5+, max UDMA/100
  console: ata1.00: 40960 sectors, multi 16: LBA48 NCQ (depth 32)
  console: ata1.00: applying bridge limits
  console: ata1.00: configured for UDMA/100
  console: scsi 0:0:0:0: Direct-Access     ATA      QEMU HARDDISK    2.5+ PQ: 0 ANSI: 5
  console: sd 0:0:0:0: Attached scsi generic sg0 type 0
  console: sd 0:0:0:0: [sda] 40960 512-byte logical blocks: (21.0 MB/20.0 MiB)
  console: sd 0:0:0:0: [sda] Write Protect is off
  console: sd 0:0:0:0: [sda] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA
  console: sd 0:0:0:0: [sda] Attached SCSI disk
  console: EXT4-fs (sda): mounting ext2 file system using the ext4 subsystem
  console: EXT4-fs (sda): mounted filesystem without journal. Opts: (null)
  console: VFS: Mounted root (ext2 filesystem) readonly on device 8:0.
  [...]
  console: cat /proc/partitions
  console: / # cat /proc/partitions
  console: major minor  #blocks  name
  console: 1        0       4096 ram0
  console: 1        1       4096 ram1
  console: 1        2       4096 ram2
  console: 1        3       4096 ram3
  console: 8        0      20480 sda
  console: reboot
  console: / # reboot
  [...]
  console: sd 0:0:0:0: [sda] Synchronizing SCSI cache
  console: reboot: Restarting system
  PASS (48.39 s)

Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-id: 20191230110953.25496-3-f4bug@amsat.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

tests/boot_linux_console: Add initrd test for the CubieBoard

This test boots a Linux kernel on a CubieBoard and verify
the serial output is working.

The kernel image and DeviceTree blob are built by the Armbian
project (based on Debian):
https://docs.armbian.com/Developer-Guide_Build-Preparation/

The cpio image used comes from the linux-build-test project:
https://github.com/groeck/linux-build-test

If ARM is a target being built, "make check-acceptance" will
automatically include this test by the use of the "arch:arm" tags.

Alternatively, this test can be run using:

  $ avocado --show=console run -t machine:cubieboard tests/acceptance/boot_linux_console.py
  console: Uncompressing Linux... done, booting the kernel.
  console: Booting Linux on physical CPU 0x0
  console: Linux version 4.20.7-sunxi (root@armbian.com) (gcc version 7.2.1 20171011 (Linaro GCC 7.2-2017.11)) #5.75 SMP Fri Feb 8 09:02:10 CET 2019
  console: CPU: ARMv7 Processor [410fc080] revision 0 (ARMv7), cr=50c5387d
  console: CPU: PIPT / VIPT nonaliasing data cache, VIPT nonaliasing instruction cache
  console: OF: fdt: Machine model: Cubietech Cubieboard
  [...]
  console: Boot successful.
  console: cat /proc/cpuinfo
  console: / # cat /proc/cpuinfo
  console: processor      : 0
  console: model name     : ARMv7 Processor rev 0 (v7l)
  console: BogoMIPS       : 832.51
  [...]
  console: Hardware       : Allwinner sun4i/sun5i Families
  console: Revision       : 0000
  console: Serial         : 0000000000000000
  console: cat /proc/iomem
  console: / # cat /proc/iomem
  console: 01c00000-01c0002f : system-control@1c00000
  console: 01c02000-01c02fff : dma-controller@1c02000
  console: 01c05000-01c05fff : spi@1c05000
  console: 01c0b080-01c0b093 : mdio@1c0b080
  console: 01c0c000-01c0cfff : lcd-controller@1c0c000
  console: 01c0d000-01c0dfff : lcd-controller@1c0d000
  console: 01c0f000-01c0ffff : mmc@1c0f000
  [...]
  PASS (54.35 s)

Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Wainer dos Santos Moschetta <wainersm@redhat.com>
Tested-by: Wainer dos Santos Moschetta <wainersm@redhat.com>
Message-id: 20191230110953.25496-2-f4bug@amsat.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

hw/arm: Add the Netduino Plus 2

Signed-off-by: Alistair Francis <alistair@alistair23.me>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Message-id: dad8d8d47f7625913e35e27a1c00f603a6b08f9a.1576658572.git.alistair@alistair23.me
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

hw/arm: Add the STM32F4xx SoC

Signed-off-by: Alistair Francis <alistair@alistair23.me>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Message-id: 1d145c4c13e5fa140caf131232a6f524c88fcd72.1576658572.git.alistair@alistair23.me
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

hw/misc: Add the STM32F4xx EXTI device

Signed-off-by: Alistair Francis <alistair@alistair23.me>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Message-id: ef941d59fd8658589d34ed432e1d6dfdcf7fb1d0.1576658572.git.alistair@alistair23.me
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

hw/misc: Add the STM32F4xx Sysconfig device

Signed-off-by: Alistair Francis <alistair@alistair23.me>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Message-id: 49b01423a09cef2ca832ff73a84a996568f1a8fc.1576658572.git.alistair@alistair23.me
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

Merge remote-tracking branch 'remotes/rth/tags/pull-or1k-20200116' into staging

Fix FPSCR masking

# gpg: Signature made Fri 17 Jan 2020 00:51:41 GMT
# gpg:                using RSA key 7A481E78868B4DB6A85A05C064DF38E8AF7E215F
# gpg:                issuer "richard.henderson@linaro.org"
# gpg: Good signature from "Richard Henderson <richard.henderson@linaro.org>" [full]
# Primary key fingerprint: 7A48 1E78 868B 4DB6 A85A  05C0 64DF 38E8 AF7E 215F

* remotes/rth/tags/pull-or1k-20200116:
  target/openrisc: Fix FPCSR mask to allow setting DZF

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

Merge remote-tracking branch 'remotes/rth/tags/pull-tcg-20200115' into staging

Add cpu_{ld,st}*_mmuidx_ra
Remove MMU_MODE*_SUFFIX
Move tcg headers under include/

# gpg: Signature made Thu 16 Jan 2020 01:36:41 GMT
# gpg:                using RSA key 7A481E78868B4DB6A85A05C064DF38E8AF7E215F
# gpg:                issuer "richard.henderson@linaro.org"
# gpg: Good signature from "Richard Henderson <richard.henderson@linaro.org>" [full]
# Primary key fingerprint: 7A48 1E78 868B 4DB6 A85A  05C0 64DF 38E8 AF7E 215F

* remotes/rth/tags/pull-tcg-20200115: (34 commits)
  MAINTAINERS: Replace Claudio Fontana for tcg/aarch64
  configure: Remove tcg/ from the preprocessor include search list
  tcg: Move TCG headers to include/tcg/
  tcg: Search includes in the parent source directory
  tcg: Search includes from the project root source directory
  cputlb: Expand cpu_ldst_template.h in cputlb.c
  cputlb: Remove support for MMU_MODE*_SUFFIX
  target/ppc: Use cpu_*_mmuidx_ra instead of MMU_MODE*_SUFFIX
  target/s390x: Use cpu_*_mmuidx_ra instead of MMU_MODE*_SUFFIX
  target/mips: Use cpu_*_mmuidx_ra instead of MMU_MODE*_SUFFIX
  target/m68k: Use cpu_*_mmuidx_ra instead of MMU_MODE{0,1}_SUFFIX
  target/xtensa: Remove MMU_MODE{0,1,2,3}_SUFFIX
  target/unicore32: Remove MMU_MODE{0,1}_SUFFIX
  target/sh4: Remove MMU_MODE{0,1}_SUFFIX
  target/microblaze: Remove MMU_MODE{0,1,2}_SUFFIX
  target/i386: Remove MMU_MODE{0,1,2}_SUFFIX
  target/cris: Remove MMU_MODE{0,1}_SUFFIX
  target/alpha: Remove MMU_MODE{0,1}_SUFFIX
  target/nios2: Remove MMU_MODE{0,1}_SUFFIX
  cputlb: Expand cpu_ldst_useronly_template.h in user-exec.c
  ...

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

target/openrisc: Fix FPCSR mask to allow setting DZF

The mask used when setting FPCSR allows setting bits 10 to 1.  However,
OpenRISC has flags and config bits in 11 to 1, 11 being Divide by Zero
Flag (DZF).  This seems like an off-by-one bug.

This was found when testing the GLIBC test suite which has test cases to
set and clear all bits.

Signed-off-by: Stafford Horne <shorne@gmail.com>
Message-Id: <20200110212843.27335-1-shorne@gmail.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

gitlab-ci.yml: Run tcg test with tci

Since commit 2f160e0f9797c7522bfd0d09218d0c9340a5137c ("tci: Add
implementation for INDEX_op_ld16u_i64") has been included now, we
can also run the TCG tests with tci, so let's enable them in our
Gitlab CI now.

Message-Id: <20191127155105.3784-1-thuth@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>

tests/qtest/vhost-user-test: Fix memory leaks

Do not allocate resources in case we return early, and make sure
to free dest_cmdline at the end.

Reported-by: Euler Robot <euler.robot@huawei.com>
Reported-by: Pan Nengyuan <pannengyuan@huawei.com>
Message-Id: <20200116140736.9498-1-thuth@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>

migration-test: ppc64: fix FORTH test program

Commit e51e711b1bef has moved the initialization of start_address and
end_address after the definition of the command line argument,
where the nvramrc is initialized, and thus the loop is between 0 and 0
rather than 1 MiB and 100 MiB.

It doesn't affect the result of the test if all the tests are run in
sequence because the two first tests don't run the loop, so the
values are correctly initialized when we actually need them.

But it hangs when we ask to run only one test, for instance:

    QTEST_QEMU_BINARY=ppc64-softmmu/qemu-system-ppc64 \
    tests/migration-test -m=quick -p /ppc64/migration/validate_uuid_error

Fixes: e51e711b1bef ("tests/migration: Add migration-test header file")
Cc: wei@redhat.com
Signed-off-by: Laurent Vivier <lvivier@redhat.com>
Message-Id: <20200107163437.52139-1-lvivier@redhat.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Acked-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Thomas Huth <thuth@redhat.com>

tests: acpi: update path in rebuild-expected-aml

Since commit 1e8a1fae7464("test: Move qtests to a separate
directory") qtests are now placed in a separate folder and
this breaks the script used to rebuild the expected ACPI
tables for bios-tables-test. Update the script with correct
path.

Fixes: 1e8a1fae7464("test: Move qtests to a separate directory")
Signed-off-by: Shameer Kolothum <shameerali.kolothum.thodi@huawei.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Message-Id: <20200114165138.15716-1-shameerali.kolothum.thodi@huawei.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>

ui/gtk: Get display refresh rate with GDK version 3.22 or later

Commit c4c00922cc introduced the use of the GdkMonitor API, which
was introduced in GTK+ 3.22:
https://developer.gnome.org/gdk3/stable/api-index-3-22.html#api-index-3.22

Unfortunately this break building with older versions, as on Ubuntu
Xenial which provides GTK+ 3.18:

  $ lsb_release -cd
  Description:    Ubuntu 16.04.5 LTS
  Codename:       xenial

  $ ./configure && make
  GTK support       yes (3.18.9)
  GTK GL support    no
  [...]
    CC      ui/gtk.o
  qemu/ui/gtk.c: In function ‘gd_vc_gfx_init’:
  qemu/ui/gtk.c:1973:5: error: unknown type name ‘GdkMonitor’
       GdkMonitor *monitor = gdk_display_get_monitor_at_window(dpy, win);
       ^
  qemu/ui/gtk.c:1973:27: error: implicit declaration of function ‘gdk_display_get_monitor_at_window’ [-Werror=implicit-function-declaration]
       GdkMonitor *monitor = gdk_display_get_monitor_at_window(dpy, win);
                             ^
  qemu/ui/gtk.c:1973:5: error: nested extern declaration of ‘gdk_display_get_monitor_at_window’ [-Werror=nested-externs]
       GdkMonitor *monitor = gdk_display_get_monitor_at_window(dpy, win);
       ^
  qemu/ui/gtk.c:1973:27: error: initialization makes pointer from integer without a cast [-Werror=int-conversion]
       GdkMonitor *monitor = gdk_display_get_monitor_at_window(dpy, win);
                             ^
  qemu/ui/gtk.c:2035:28: error: implicit declaration of function ‘gdk_monitor_get_refresh_rate’ [-Werror=implicit-function-declaration]
       refresh_rate_millihz = gdk_monitor_get_refresh_rate(monitor);
                              ^
  qemu/ui/gtk.c:2035:5: error: nested extern declaration of ‘gdk_monitor_get_refresh_rate’ [-Werror=nested-externs]
       refresh_rate_millihz = gdk_monitor_get_refresh_rate(monitor);
       ^
  cc1: all warnings being treated as errors
  qemu/rules.mak:69: recipe for target 'ui/gtk.o' failed
  make: *** [ui/gtk.o] Error 1

GTK+ provides convenient definition in <gdk/gdkversionmacros.h>
(already include by <gdk/gdk.h>) to check which API are available.

We only use the GdkMonitor API to get the monitor refresh rate.

Extract this code as a new gd_refresh_rate_millihz() function,
and check GDK_VERSION_3_22 is defined before calling its API.
If it is not defined, return 0. This is safe and fixes our build
failure (see https://travis-ci.org/qemu/qemu/builds/636992508).

Reported-by: Travis-CI
Reviewed-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-id: 20200116115413.31650-1-philmd@redhat.com
Fixes: c4c00922cc (display/gtk: get proper refreshrate)
Reviewed-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

MAINTAINERS: Replace Claudio Fontana for tcg/aarch64

Claudio's Huawei address has been defunct for quite a while.  In

  https://lists.gnu.org/archive/html/qemu-devel/2019-05/msg06872.html

he asked for his personal address to be removed as well.

I will take over officially.

Cc: Claudio Fontana <claudio.fontana@gmail.com>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

configure: Remove tcg/ from the preprocessor include search list

All tcg includes are relative to the repository root directory,
we can safely remove the tcg/ directory from the include search
path list.

Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Stefan Weil <sw@weilnetz.de>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-Id: <20200101112303.20724-5-philmd@redhat.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

tcg: Move TCG headers to include/tcg/

Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Stefan Weil <sw@weilnetz.de>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-Id: <20200101112303.20724-4-philmd@redhat.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

tcg: Search includes in the parent source directory

All the *.inc.c files included by tcg/$TARGET/tcg-target.inc.c
are in tcg/, their parent directory. To simplify the preprocessor
search path, include the relative parent path: '..'.

Patch created mechanically by running:

  $ for x in tcg-pool.inc.c tcg-ldst.inc.c; do \
    sed -i "s,#include \"$x\",#include \"../$x\"," \
      $(git grep -l "#include \"$x\""); \
    done

Acked-by: David Gibson <david@gibson.dropbear.id.au> (ppc parts)
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Stefan Weil <sw@weilnetz.de>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-Id: <20200101112303.20724-3-philmd@redhat.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

tcg: Search includes from the project root source directory

We currently search both the root and the tcg/ directories for tcg
files:

  $ git grep '#include "tcg/' | wc -l
  28

  $ git grep '#include "tcg[^/]' | wc -l
  94

To simplify the preprocessor search path, unify by expliciting the
tcg/ directory.

Patch created mechanically by running:

  $ for x in \
      tcg.h tcg-mo.h tcg-op.h tcg-opc.h \
      tcg-op-gvec.h tcg-gvec-desc.h; do \
    sed -i "s,#include \"$x\",#include \"tcg/$x\"," \
      $(git grep -l "#include \"$x\""); \
    done

Acked-by: David Gibson <david@gibson.dropbear.id.au> (ppc parts)
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Stefan Weil <sw@weilnetz.de>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-Id: <20200101112303.20724-2-philmd@redhat.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

cputlb: Expand cpu_ldst_template.h in cputlb.c

Reduce the amount of preprocessor obfuscation by expanding
the text of each of the functions generated.  The result is
only slightly smaller than the original.

Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Aleksandar Markovic <amarkovic@wavecomp.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

cputlb: Remove support for MMU_MODE*_SUFFIX

All users have now been converted to cpu_*_mmuidx_ra.

Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/ppc: Use cpu_*_mmuidx_ra instead of MMU_MODE*_SUFFIX

There are only two uses.  Within dcbz_common, the local variable
mmu_idx already contains the epid computation, and we can avoid
repeating it for the store.  Within helper_icbiep, the usage is
trivially expanded using PPC_TLB_EPID_LOAD.

Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Acked-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/s390x: Use cpu_*_mmuidx_ra instead of MMU_MODE*_SUFFIX

The generated functions aside from *_real are unused.
The *_real functions have a couple of users in mem_helper.c;
use *_mmuidx_ra instead, with MMU_REAL_IDX.

Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
---
v2: Use *_mmuidx_ra directly, without intermediate macros.

target/mips: Use cpu_*_mmuidx_ra instead of MMU_MODE*_SUFFIX

The separate suffixed functions were used to construct
some do_##insn function switched on mmu_idx.  The interface
is exactly identical to the *_mmuidx_ra functions.  Replace
them directly and remove the constructions.

Cc: Aurelien Jarno <aurelien@aurel32.net>
Cc: Aleksandar Rikalo <aleksandar.rikalo@rt-rk.com>
Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Aleksandar Markovic <amarkovic@wavecomp.com>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/m68k: Use cpu_*_mmuidx_ra instead of MMU_MODE{0,1}_SUFFIX

The generated *_user functions are unused.  The *_kernel functions
have a couple of users in op_helper.c; use *_mmuidx_ra instead,
with MMU_KERNEL_IDX.

Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Laurent Vivier <laurent@vivier.eu>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
---
v2: Use *_mmuidx_ra directly, without intermediate macros.

target/xtensa: Remove MMU_MODE{0,1,2,3}_SUFFIX

The functions generated by these macros are unused.

Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Acked-by: Max Filippov <jcmvbkbc@gmail.com>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/unicore32: Remove MMU_MODE{0,1}_SUFFIX

The functions generated by these macros are unused.

Cc: Guan Xuetao <gxt@mprc.pku.edu.cn>
Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/sh4: Remove MMU_MODE{0,1}_SUFFIX

The functions generated by these macros are unused.

Cc: Aurelien Jarno <aurelien@aurel32.net>
Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/microblaze: Remove MMU_MODE{0,1,2}_SUFFIX

The functions generated by these macros are unused.

Cc: Edgar E. Iglesias <edgar.iglesias@gmail.com>
Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/i386: Remove MMU_MODE{0,1,2}_SUFFIX

The functions generated by these macros are unused.

Cc: Eduardo Habkost <ehabkost@redhat.com>
Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Acked-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/cris: Remove MMU_MODE{0,1}_SUFFIX

The functions generated by these macros are unused.

Cc: Edgar E. Iglesias <edgar.iglesias@gmail.com>
Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/alpha: Remove MMU_MODE{0,1}_SUFFIX

The functions generated by these macros are unused.

Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/nios2: Remove MMU_MODE{0,1}_SUFFIX

The functions generated by these macros are unused.

Cc: Chris Wulff <crwulff@gmail.com>
Cc: Marek Vasut <marex@denx.de>
Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

cputlb: Expand cpu_ldst_useronly_template.h in user-exec.c

With the tracing hooks, the inline functions are no longer
so simple.  Reduce the amount of preprocessor obfuscation
by expanding the text of each of the functions generated.

Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/i386: Use cpu_*_mmuidx_ra instead of templates

Do not use exec/cpu_ldst_{,useronly_}template.h directly,
but instead use the functional interface.

Cc: Eduardo Habkost <ehabkost@redhat.com>
Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Acked-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

cputlb: Provide cpu_(ld,st}*_mmuidx_ra for user-only

This finishes the new interface began with the previous patch.
Document the interface and deprecate MMU_MODE<N>_SUFFIX.

Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Aleksandar Markovic <amarkovic@wavecomp.com>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

cputlb: Rename helper_ret_ld*_cmmu to cpu_ld*_code

There are no uses of the *_cmmu names other than the bare wrapping
within the *_code inlines.  Therefore rename the functions so we
can drop the inlines.

Use abi_ptr instead of target_ulong in preparation for user-only;
the two types are identical for softmmu.

Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Aleksandar Markovic <amarkovic@wavecomp.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

translator: Use cpu_ld*_code instead of open-coding

The DO_LOAD macros replicate the distinction already performed
by the cpu_ldst.h functions.  Use them.

Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

cputlb: Move body of cpu_ldst_template.h out of line

With the tracing hooks, the inline functions are no longer
so simple.  Once out-of-line, the current tlb_entry lookup
is redundant with the one in the main load/store_helper.

This also begins the introduction of a new target facing
interface, with suffix *_mmuidx_ra.  This is not yet
official because the interface is not done for user-only.

Use abi_ptr instead of target_ulong in preparation for
user-only; the two types are identical for softmmu.

What remains in cpu_ldst_template.h are the expansions
for _code, _data, and MMU_MODE<N>_SUFFIX.

Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

plugins: Include trace/mem.h in api.c

Code movement in an upcoming patch will show that this file
was implicitly depending on trace/mem.h being included beforehand.

Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reported-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

linux-user: Include trace-root.h in syscall-trace.h

Code movement in an upcoming patch will show that this file
was implicitly depending on trace-root.h being included beforehand.

Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

linux-user: Include tcg.h in syscall.c

Code movement in an upcoming patch will show that this file
was implicitly depending on tcg.h being included indirectly.

Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

accel/tcg: Include tcg.h in tcg-runtime.c

Code movement in an upcoming patch will show that this file
was implicitly depending on tcg.h being included indirectly.

Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/arm: Include tcg.h in sve_helper.c

Code movement in an upcoming patch will show that this file
was implicitly depending on tcg.h being included indirectly.

Cc: Peter Maydell <peter.maydell@linaro.org>
Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/s390x: Include tcg.h in mem_helper.c

Code movement in an upcoming patch will show that this file
was implicitly depending on tcg.h being included indirectly.

Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

trace: Remove trace_mem_build_info_no_se_[bl]e

It is easy for the atomic helpers to use trace_mem_build_info
directly, without resorting to symbol pasting.  For this usage,
we cannot use trace_mem_get_info, because the MemOp does not
support 16-byte accesses.

Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

cputlb: Use trace_mem_get_info instead of trace_mem_build_info

In the cpu_ldst templates, we already require a MemOp, and it
is cleaner and clearer to pass that instead of 3 separate
arguments describing the memory operation.

Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/xtensa: Use probe_access for itlb_hit_test

We don't actually need the result of the read, only to probe that the
memory mapping exists.  This is exactly what probe_access does.

This is also the only user of any cpu_ld*_code_ra function.
Removing this allows the interface to be removed shortly.

Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Acked-by: Max Filippov <jcmvbkbc@gmail.com>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

Merge remote-tracking branch 'remotes/armbru/tags/pull-qapi-2020-01-14' into staging

QAPI patches for 2020-01-14

# gpg: Signature made Tue 14 Jan 2020 10:15:22 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

* remotes/armbru/tags/pull-qapi-2020-01-14:
  qapi: Simplify QAPISchemaModularCVisitor
  qapi: Fix code generation for empty modules
  qapi: Proper intermediate representation for modules
  qapi: Generate command registration stuff into separate files
  tests/Makefile.include: Fix missing test-qapi-emit-events.[ch]
  qapi: Tweak "command returns a nice type" check for clarity

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

Merge remote-tracking branch 'remotes/kraxel/tags/ui-20200114-pull-request' into staging

ui: add "-display help", gtk refresh rate.

# gpg: Signature made Tue 14 Jan 2020 10:01:14 GMT
# gpg:                using RSA key 4CB6D8EED3E87138
# 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

* remotes/kraxel/tags/ui-20200114-pull-request:
  display/gtk: get proper refreshrate
  ui: Print available display backends with '-display help'

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

Merge remote-tracking branch 'remotes/stefanha/tags/block-pull-request' into staging

Pull request

# gpg: Signature made Tue 14 Jan 2020 09:30:45 GMT
# gpg:                using RSA key 8695A8BFD3F97CDAAC35775A9CA4ABB381AB73C8
# gpg: Good signature from "Stefan Hajnoczi <stefanha@redhat.com>" [full]
# gpg:                 aka "Stefan Hajnoczi <stefanha@gmail.com>" [full]
# Primary key fingerprint: 8695 A8BF D3F9 7CDA AC35  775A 9CA4 ABB3 81AB 73C8

* remotes/stefanha/tags/block-pull-request:
  linux-aio: increasing MAX_EVENTS to a larger hardcoded value

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

Merge remote-tracking branch 'remotes/stefanha/tags/tracing-pull-request' into staging

Pull request

# gpg: Signature made Tue 14 Jan 2020 09:29:11 GMT
# gpg:                using RSA key 8695A8BFD3F97CDAAC35775A9CA4ABB381AB73C8
# gpg: Good signature from "Stefan Hajnoczi <stefanha@redhat.com>" [full]
# gpg:                 aka "Stefan Hajnoczi <stefanha@gmail.com>" [full]
# Primary key fingerprint: 8695 A8BF D3F9 7CDA AC35  775A 9CA4 ABB3 81AB 73C8

* remotes/stefanha/tags/tracing-pull-request:
  trace: update qemu-trace-stap to Python 3

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

qapi: Simplify QAPISchemaModularCVisitor

Since the previous commit, QAPISchemaVisitor.visit_module() is called
just once.  Simplify QAPISchemaModularCVisitor accordingly.

Signed-off-by: Markus Armbruster <armbru@redhat.com>
Message-Id: <20191120182551.23795-7-armbru@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>

qapi: Fix code generation for empty modules

When a sub-module doesn't contain any definitions, we don't generate
code for it, but we do generate the #include.

We generate code only for modules that get visited.
QAPISchema.visit() visits only modules that have definitions.  It can
visit modules multiple times.

Clean this up as follows.  Collect entities in their QAPISchemaModule.
Have QAPISchema.visit() call QAPISchemaModule.visit() for each module.
Have QAPISchemaModule.visit() call .visit_module() for itself, and
QAPISchemaEntity.visit() for each of its entities.  This way, we visit
each module exactly once.

Signed-off-by: Markus Armbruster <armbru@redhat.com>
Message-Id: <20191120182551.23795-6-armbru@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>

qapi: Proper intermediate representation for modules

Modules are represented only by their names so far.  Introduce class
QAPISchemaModule.  So far, it merely wraps the name.  The next patch
will put it to more interesting use.

Once again, arrays spice up the patch a bit.  For any other type,
@info points to the definition, which lets us map from @info to
module.  For arrays, there is no definition, and @info points to the
first use instead.  We have to use the element type's module instead,
which is only available after .check().

Signed-off-by: Markus Armbruster <armbru@redhat.com>
Message-Id: <20191120182551.23795-5-armbru@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>

qapi: Generate command registration stuff into separate files

Having to include qapi-commands.h just for qmp_init_marshal() is
suboptimal.  Generate it into separate files.  This lets
monitor/misc.c, qga/main.c, and the generated qapi-commands-FOO.h
include less.

Signed-off-by: Markus Armbruster <armbru@redhat.com>
Message-Id: <20191120182551.23795-4-armbru@redhat.com>
[Typos in docs/devel/qapi-code-gen.txt fixed]
Reviewed-by: Eric Blake <eblake@redhat.com>

tests/Makefile.include: Fix missing test-qapi-emit-events.[ch]

Commit 5d75648b56 "qapi: Generate QAPIEvent stuff into separate files"
added tests/test-qapi-emit-events.[ch] to the set of generated files,
but neglected to update tests/.gitignore and tests/Makefile.include.
Commit a0af8cee3c "tests/.gitignore: ignore test-qapi-emit-events.[ch]
for in-tree builds" fixed the former.  Now fix the latter.

Signed-off-by: Markus Armbruster <armbru@redhat.com>
Message-Id: <20191120182551.23795-3-armbru@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>

qapi: Tweak "command returns a nice type" check for clarity

Signed-off-by: Markus Armbruster <armbru@redhat.com>
Message-Id: <20191120182551.23795-2-armbru@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>

display/gtk: get proper refreshrate

Because some VMs in QEMU can get GPU virtualization (using technologies
such as iGVT-g, as mentioned previously), they could produce a video
output that had a higher display refresh rate than of what the GTK
display was displaying. (fxp. Playing a video game inside of a Windows
VM at 60 Hz, while the output stood locked at 33 Hz because of defaults
set in include/ui/console.h)

Since QEMU does indeed have internal systems for determining frame
times as defined in ui/console.c.
The code checks for a variable called update_interval that it later
uses for time calculation. This variable, however, isn't defined
anywhere in ui/gtk.c and instead ui/console.c just sets it to
GUI_REFRESH_INTERVAL_DEFAULT which is 30

update_interval represents the number of milliseconds per display
refresh, and by doing some math we get that 1000/30 = 33.33... Hz

This creates the mentioned problem and what this patch does is that it
checks for the display refresh rate reported by GTK itself (we can take
this as a safe value) and just converts it back to a number of
milliseconds per display refresh.

Signed-off-by: Nikola Pavlica <pavlica.nikola@gmail.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-id: 20200108121342.29597-1-pavlica.nikola@gmail.com

[ kraxel: style tweak: add blank line between vars and code ]

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>

ui: Print available display backends with '-display help'

We already print availabled devices with "-device help", or available
backends with "-netdev help" or "-chardev help". Let's provide a way
for the users to query the available display backends, too.

Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
Message-id: 20200108144702.29969-1-thuth@redhat.com
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>

trace: update qemu-trace-stap to Python 3

qemu-trace-stap does not support Python 3 yet:

  $ scripts/qemu-trace-stap list path/to/qemu-system-x86_64
  Traceback (most recent call last):
    File "scripts/qemu-trace-stap", line 175, in <module>
      main()
    File "scripts/qemu-trace-stap", line 171, in main
      args.func(args)
    File "scripts/qemu-trace-stap", line 118, in cmd_list
      print_probes(args.verbose, "*")
    File "scripts/qemu-trace-stap", line 114, in print_probes
      if line.startswith(prefix):
  TypeError: startswith first arg must be bytes or a tuple of bytes, not str

Now that QEMU requires Python 3.5 or later we can switch to pure Python
3.  Use Popen()'s universal_newlines=True argument to treat stdout as
text instead of binary.

Fixes: 62dd1048c0bd ("trace: add ability to do simple printf logging via systemtap")
Buglink: https://bugzilla.redhat.com/show_bug.cgi?id=1787395
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Message-id: 20200107112438.383958-1-stefanha@redhat.com
Message-Id: <20200107112438.383958-1-stefanha@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>

linux-aio: increasing MAX_EVENTS to a larger hardcoded value

Since commit 6040aedddb5f474a9c2304b6a432a652d82b3d3c "virtio-blk:
make queue size configurable",if the user set the queue size to
more than 128 ,it will not take effect. That's because linux aio's
maximum outstanding requests at a time is always less than or equal
to 128.

This patch simply increase MAX_EVENTS to a larger hardcoded value of
1024 as a shortterm fix.

Signed-off-by: wangyong <wang.yongD@h3c.com>
Message-id: faa5781afd354a96a0be152b288f636f@h3c.com
Message-Id: <faa5781afd354a96a0be152b288f636f@h3c.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>

Merge remote-tracking branch 'remotes/kraxel/tags/usb-20200113-pull-request' into staging

usb: bugfixes for xhci, usb pass-through and usb redirection.

# gpg: Signature made Mon 13 Jan 2020 13:06:35 GMT
# gpg:                using RSA key 4CB6D8EED3E87138
# 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

* remotes/kraxel/tags/usb-20200113-pull-request:
  xhci: recheck slot status
  xhci: Fix memory leak in xhci_kick_epctx when poweroff GuestOS
  usbredir: Prevent recursion in usbredir_write
  usb-redir: remove 'remote wakeup' flag from configuration descriptor
  usb-host: remove 'remote wakeup' flag from configuration descriptor

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

Merge remote-tracking branch 'remotes/huth-gitlab/tags/pull-request-2020-01-12' into staging

* Move qtests into a separate directory
* Build index.html for docs

# gpg: Signature made Sun 12 Jan 2020 11:21:41 GMT
# gpg:                using RSA key 27B88847EEE0250118F3EAB92ED9D774FE702DB5
# gpg:                issuer "thuth@redhat.com"
# gpg: Good signature from "Thomas Huth <th.huth@gmx.de>" [full]
# gpg:                 aka "Thomas Huth <thuth@redhat.com>" [full]
# gpg:                 aka "Thomas Huth <huth@tuxfamily.org>" [full]
# gpg:                 aka "Thomas Huth <th.huth@posteo.de>" [unknown]
# Primary key fingerprint: 27B8 8847 EEE0 2501 18F3  EAB9 2ED9 D774 FE70 2DB5

* remotes/huth-gitlab/tags/pull-request-2020-01-12:
  docs: build an index page for the HTML docs
  tests/libqos: Move the libqos files under tests/qtest/
  tests/Makefile: Move qtest-related settings to a separate Makefile.include
  test: Move qtests to a separate directory
  tests/Makefile: Separate unit test dependencies from qtest dependencies
  tests/Makefile: Remove 'tests/' and '$(EXESUF)' from the check-qtest variables
  tests/ptimer: Remove unnecessary inclusion of libqtest.h
  tests/Makefile: test-char does not need libqtest

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

xhci: recheck slot status

Factor out slot status check into a helper function.  Add an additional
check after completing transfers.  This is needed in case a guest
queues multiple transfers in a row and a device unplug happens while
qemu processes them.

Buglink: https://bugzilla.redhat.com/show_bug.cgi?id=1786413
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-id: 20200107083606.12393-1-kraxel@redhat.com

xhci: Fix memory leak in xhci_kick_epctx when poweroff GuestOS

start vm with libvirt, when GuestOS running, enter poweroff command using
the xhci keyboard, then ASAN shows memory leak stack：

Direct leak of 80 byte(s) in 5 object(s) allocated from:
    #0 0xfffd1e6431cb in __interceptor_malloc (/lib64/libasan.so.4+0xd31cb)
    #1 0xfffd1e107163 in g_malloc (/lib64/libglib-2.0.so.0+0x57163)
    #2 0xaaad39051367 in qemu_sglist_init /qemu/dma-helpers.c:43
    #3 0xaaad3947c407 in pci_dma_sglist_init /qemu/include/hw/pci/pci.h:842
    #4 0xaaad3947c407 in xhci_xfer_create_sgl /qemu/hw/usb/hcd-xhci.c:1446
    #5 0xaaad3947c407 in xhci_setup_packet /qemu/hw/usb/hcd-xhci.c:1618
    #6 0xaaad3948625f in xhci_submit /qemu/hw/usb/hcd-xhci.c:1827
    #7 0xaaad3948625f in xhci_fire_transfer /qemu/hw/usb/hcd-xhci.c:1839
    #8 0xaaad3948625f in xhci_kick_epctx /qemu/hw/usb/hcd-xhci.c:1991
    #9 0xaaad3948f537 in xhci_doorbell_write /qemu/hw/usb/hcd-xhci.c:3158
    #10 0xaaad38bcbfc7 in memory_region_write_accessor /qemu/memory.c:483
    #11 0xaaad38bc654f in access_with_adjusted_size /qemu/memory.c:544
    #12 0xaaad38bd1877 in memory_region_dispatch_write /qemu/memory.c:1482
    #13 0xaaad38b1c77f in flatview_write_continue /qemu/exec.c:3167
    #14 0xaaad38b1ca83 in flatview_write /qemu/exec.c:3207
    #15 0xaaad38b268db in address_space_write /qemu/exec.c:3297
    #16 0xaaad38bf909b in kvm_cpu_exec /qemu/accel/kvm/kvm-all.c:2383
    #17 0xaaad38bb063f in qemu_kvm_cpu_thread_fn /qemu/cpus.c:1246
    #18 0xaaad39821c93 in qemu_thread_start /qemu/util/qemu-thread-posix.c:519
    #19 0xfffd1c8378bb  (/lib64/libpthread.so.0+0x78bb)
    #20 0xfffd1c77616b  (/lib64/libc.so.6+0xd616b)

Reported-by: Euler Robot <euler.robot@huawei.com>
Signed-off-by: Chen Qun <kuhn.chenqun@huawei.com>
Message-id: 20200110105855.81144-1-kuhn.chenqun@huawei.com
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>

usbredir: Prevent recursion in usbredir_write

I've got a case where usbredir_write manages to call back into itself
via spice; this patch causes the recursion to fail (0 bytes) the write;
this seems to avoid the deadlock I was previously seeing.

I can't say I fully understand the interaction of usbredir and spice;
but there are a few similar guards in spice and usbredir
to catch other cases especially onces also related to spice_server_char_device_wakeup

This case seems to be triggered by repeated migration+repeated
reconnection of the viewer; but my debugging suggests the migration
finished before this hits.

The backtrace of the hang looks like:
  reds_handle_ticket
  reds_handle_other_links
  reds_channel_do_link
  red_channel_connect
  spicevmc_connect
  usbredir_create_parser
  usbredirparser_do_write
  usbredir_write
  qemu_chr_fe_write
  qemu_chr_write
  qemu_chr_write_buffer
  spice_chr_write
  spice_server_char_device_wakeup
  red_char_device_wakeup
  red_char_device_write_to_device
  vmc_write
  usbredirparser_do_write
  usbredir_write
  qemu_chr_fe_write
  qemu_chr_write
  qemu_chr_write_buffer
  qemu_mutex_lock_impl

and we fail as we lang through qemu_chr_write_buffer's lock
twice.

Bug: https://bugzilla.redhat.com/show_bug.cgi?id=1752320

Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Message-Id: <20191218113012.13331-1-dgilbert@redhat.com>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>

Merge remote-tracking branch 'remotes/vivier2/tags/trivial-branch-pull-request' into staging

Fix some uninitialized variable warnings,
some memory leak warnings and update MAINTAINERS file.

# gpg: Signature made Wed 08 Jan 2020 16:02:11 GMT
# gpg:                using RSA key CD2F75DDC8E3A4DC2E4F5173F30C38BD3F2FBE3C
# gpg:                issuer "laurent@vivier.eu"
# gpg: Good signature from "Laurent Vivier <lvivier@redhat.com>" [full]
# gpg:                 aka "Laurent Vivier <laurent@vivier.eu>" [full]
# gpg:                 aka "Laurent Vivier (Red Hat) <lvivier@redhat.com>" [full]
# Primary key fingerprint: CD2F 75DD C8E3 A4DC 2E4F  5173 F30C 38BD 3F2F BE3C

* remotes/vivier2/tags/trivial-branch-pull-request:
  vl: fix memory leak in configure_accelerators
  arm/translate-a64: fix uninitialized variable warning
  nbd: fix uninitialized variable warning
  util/module: fix a memory leak
  MAINTAINERS: Update Yuval Shaia's email address

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

usb-redir: remove 'remote wakeup' flag from configuration descriptor

If the redirected device has this capability, Windows guest may
place the device into D2 and expect it to wake when the device
becomes active, but this will never happen. For example, when
internal Bluetooth adapter is redirected, keyboards and mice
connected to it do not work. Current commit removes this
capability (starting from machine 5.0)
Set 'usb-redir.suppress-remote-wake' property to 'off' to keep
'remote wake' as is or to 'on' to remove 'remote wake' on
4.2 or earlier.

Signed-off-by: Yuri Benditovich <yuri.benditovich@daynix.com>
Message-id: 20200108091044.18055-3-yuri.benditovich@daynix.com
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>

usb-host: remove 'remote wakeup' flag from configuration descriptor

If the redirected device has this capability, Windows guest may
place the device into D2 and expect it to wake when the device
becomes active, but this will never happen. For example, when
internal Bluetooth adapter is redirected, keyboards and mice
connected to it do not work. Current commit removes this
capability (starting from machine 5.0)
Set 'usb-host.suppress-remote-wake' property to 'off' to keep
'remote wake' as is or to 'on' to remove 'remote wake' on
4.2 or earlier.

Signed-off-by: Yuri Benditovich <yuri.benditovich@daynix.com>
Message-id: 20200108091044.18055-2-yuri.benditovich@daynix.com
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>

docs: build an index page for the HTML docs

There is no index.html start page for the QEMU HTML documentation.  An
index page is needed so that documentation can be browsed easily on the
web.

This patch adds an index.html.in template file where the QEMU version
number is expanded.  It is written in HTML instead of using the existing
sphinx (rST) and texi documentation generators because they are
heavyweight and would make this harder.

Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20191219135620.1626608-1-stefanha@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>

tests/libqos: Move the libqos files under tests/qtest/

The qos stuff belongs to qtest, so move it into that directory, too.

Message-Id: <20191218103059.11729-8-thuth@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>

tests/Makefile: Move qtest-related settings to a separate Makefile.include

tests/Makefile.include is pretty much overcrowded. Now that we have a
dedicated folder for the qtests, let's move the related settings
to a Makefile.include file in that directory instead.

Message-Id: <20191218103059.11729-7-thuth@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>

test: Move qtests to a separate directory

The tests directory itself is pretty overcrowded, and it's hard to
see which test belongs to which test subsystem (unit, qtest, ...).
Let's move the qtests to a separate folder for more clarity.

Message-Id: <20191218103059.11729-6-thuth@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>

tests/Makefile: Separate unit test dependencies from qtest dependencies

Some of the unit test dependencies are declared right in the block of
the qtest dependencies. Let's move them to the other unit tests instead.

Message-Id: <20191218103059.11729-5-thuth@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>

tests/Makefile: Remove 'tests/' and '$(EXESUF)' from the check-qtest variables

It's much easier if we simply add the folder prefix and the exe suffix
later via a substitution instead.

Message-Id: <20191218103059.11729-4-thuth@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>