target/arm: Add is_secure parameter to regime_translation_disabled

Remove the use of regime_is_secure from regime_translation_disabled,
using the new parameter instead.

This fixes a bug in S1_ptw_translate and get_phys_addr where we had
passed ARMMMUIdx_Stage2 and not ARMMMUIdx_Stage2_S to determine if
Stage2 is disabled, affecting FEAT_SEL2.

Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20221001162318.153420-5-richard.henderson@linaro.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

target/arm: Fix S2 disabled check in S1_ptw_translate

Pass the correct stage2 mmu_idx to regime_translation_disabled,
which we computed afterward.

Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Message-id: 20221001162318.153420-4-richard.henderson@linaro.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

target/arm: Add is_secure parameter to get_phys_addr_lpae

Remove the use of regime_is_secure from get_phys_addr_lpae,
using the new parameter instead.

Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20221001162318.153420-3-richard.henderson@linaro.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

target/arm: Make the final stage1+2 write to secure be unconditional

While the stage2 call to get_phys_addr_lpae should never set
attrs.secure when given a non-secure input, it's just as easy
to make the final update to attrs.secure be unconditional and
false in the case of non-secure input.

Suggested-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20221007152159.1414065-1-richard.henderson@linaro.org
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

target/arm: Split s2walk_secure from ipa_secure in get_phys_addr

The starting security state comes with the translation regime,
not the current state of arm_is_secure_below_el3().

Create a new local variable, s2walk_secure, which does not need
to be written back to result->attrs.secure -- we compute that
value later, after the S2 walk is complete.

Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Message-id: 20221001162318.153420-2-richard.henderson@linaro.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

docs/nuvoton: Update URL for images

openpower.xyz was retired some time ago. The OpenBMC Jenkins is where
images can be found these days.

Signed-off-by: Joel Stanley <joel@jms.id.au>
Reviewed-by: Hao Wu <wuhaotsh@google.com>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Tested-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-id: 20221004050042.22681-1-joel@jms.id.au
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

target/arm: allow setting SCR_EL3.EnTP2 when FEAT_SME is implemented

Updates write_scr() to allow setting SCR_EL3.EnTP2 when FEAT_SME is
implemented. SCR_EL3 being a 64-bit register, valid_mask is changed
to uint64_t and the SCR_* constants in target/arm/cpu.h are extended
to 64-bit so that masking and bitwise not (~) behave as expected.

This enables booting Linux with Trusted Firmware-A at EL3 with
"-M virt,secure=on -cpu max".

Cc: qemu-stable@nongnu.org
Fixes: 78cb9776662a ("target/arm: Enable SME for -cpu max")
Signed-off-by: Jerome Forissier <jerome.forissier@linaro.org>
Reviewed-by: Andre Przywara <andre.przywara@arm.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20221004072354.27037-1-jerome.forissier@linaro.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

target/arm/kvm: Retry KVM_CREATE_VM call if it fails EINTR

Occasionally the KVM_CREATE_VM ioctl can return EINTR, even though
there is no pending signal to be taken. In commit 94ccff13382055
we added a retry-on-EINTR loop to the KVM_CREATE_VM call in the
generic KVM code. Adopt the same approach for the use of the
ioctl in the Arm-specific KVM code (where we use it to create a
scratch VM for probing for various things).

For more information, see the mailing list thread:
https://lore.kernel.org/qemu-devel/8735e0s1zw.wl-maz@kernel.org/

Reported-by: Vitaly Chikunov <vt@altlinux.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Vitaly Chikunov <vt@altlinux.org>
Reviewed-by: Eric Auger <eric.auger@redhat.com>
Acked-by: Marc Zyngier <maz@kernel.org>
Message-id: 20220930113824.1933293-1-peter.maydell@linaro.org

kvm: allow target-specific accelerator properties

Several hypervisor capabilities in KVM are target-specific.  When exposed
to QEMU users as accelerator properties (i.e. -accel kvm,prop=value), they
should not be available for all targets.

Add a hook for targets to add their own properties to -accel kvm, for
now no such property is defined.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20220929072014.20705-3-chenyi.qiang@intel.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

i386: kvm: extend kvm_{get, put}_vcpu_events to support pending triple fault

For the direct triple faults, i.e. hardware detected and KVM morphed
to VM-Exit, KVM will never lose them. But for triple faults sythesized
by KVM, e.g. the RSM path, if KVM exits to userspace before the request
is serviced, userspace could migrate the VM and lose the triple fault.

A new flag KVM_VCPUEVENT_VALID_TRIPLE_FAULT is defined to signal that
the event.triple_fault_pending field contains a valid state if the
KVM_CAP_X86_TRIPLE_FAULT_EVENT capability is enabled.

Acked-by: Peter Xu <peterx@redhat.com>
Signed-off-by: Chenyi Qiang <chenyi.qiang@intel.com>
Message-Id: <20220929072014.20705-2-chenyi.qiang@intel.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

scsi-disk: support setting CD-ROM block size via device options

SunOS expects CD-ROM devices to have a block size of 512, and will
fail to mount or install using QEMU's default block size of 2048.

When initializing the SCSI device, allow the `physical_block_size'
block device option to override the default block size.

Signed-off-by: John Millikin <john@john-millikin.com>
Message-Id: <20220804122950.1577012-1-john@john-millikin.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

x86: pci: acpi: consolidate PCI slots creation

No functional changes nor AML bytecode changes.
Consolidate code that generates empty and populated slot
descriptors. Besides eliminating duplication,
it helps consolidate conditions for generating
parts of Device{} desriptor in one place, which makes
code more compact and easier to read.

Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Message-Id: <20220701133515.137890-18-imammedo@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

tests: acpi: update expected blobs

    Expected change:
      +            Device (SE8)
      +            {
      +                Name (_ADR, 0x001D0000)  // _ADR: Address
      +                Name (ASUN, 0x1D)
                       Method (_DSM, 4, Serialized)  // _DSM: Device-Specific Method
                       {
                           Local0 = Package (0x02)
                               {
                                   BSEL,
                                   ASUN
                               }
                           Return (PDSM (Arg0, Arg1, Arg2, Arg3, Local0))
                       }
      -            }

      -            Device (SE8)
      -            {
      -                Name (_ADR, 0x001D0000)  // _ADR: Address
      -                Name (ASUN, 0x1D)
                       Name (_SUN, 0x1D)  // _SUN: Slot User Number
                       Method (_EJ0, 1, NotSerialized)  // _EJx: Eject Device
                       {
                           PCEJ (BSEL, _SUN)
                       }
      +            }

Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Message-Id: <20220701133515.137890-17-imammedo@redhat.com>

x86: pci: acpi: reorder Device's _DSM method

align _DSM method in empty slot descriptor with
a populated slot position.
Expected change:
  +            Device (SE8)
  +            {
  +                Name (_ADR, 0x001D0000)  // _ADR: Address
  +                Name (ASUN, 0x1D)
                   Method (_DSM, 4, Serialized)  // _DSM: Device-Specific Method
                   {
                       Local0 = Package (0x02)
                           {
                               BSEL,
                               ASUN
                           }
                       Return (PDSM (Arg0, Arg1, Arg2, Arg3, Local0))
                   }
  -            }

  -            Device (SE8)
  -            {
  -                Name (_ADR, 0x001D0000)  // _ADR: Address
  -                Name (ASUN, 0x1D)
                   Name (_SUN, 0x1D)  // _SUN: Slot User Number
                   Method (_EJ0, 1, NotSerialized)  // _EJx: Eject Device
                   {
                       PCEJ (BSEL, _SUN)
                   }
  +            }

i.e. put _DSM right after ASUN, with _SUN/_EJ0 following it.

that will eliminate contextual changes (causing test failures)
when follow up patches merge code generating populated and empty
slots descriptors.

Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Message-Id: <20220701133515.137890-16-imammedo@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

tests: acpi: whitelist pc/q35 DSDT before moving _ADR field

Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Message-Id: <20220701133515.137890-15-imammedo@redhat.com>

tests: acpi: update expected blobs

Expected change:
  -                Name (_SUN, 0x0X)  // _SUN: Slot User Number
                   Name (_ADR, 0xY)  // _ADR: Address
  ...
  +                Name (_SUN, 0xX)  // _SUN: Slot User Number

Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Message-Id: <20220701133515.137890-14-imammedo@redhat.com>

x86: pci: acpi: reorder Device's _ADR and _SUN fields

no functional change, align order of fields in empty slot
descriptor with a populated slot ordering.
Expected diff:
  -                Name (_SUN, 0x0X)  // _SUN: Slot User Number
                   Name (_ADR, 0xY)  // _ADR: Address
  ...
  +                Name (_SUN, 0xX)  // _SUN: Slot User Number

that will eliminate contextual changes (causing test failures)
when follow up patches merge code generating populated and empty
slots descriptors.

Put mandatory _ADR as the 1st field, then ASUN as it can be
present for both pupulated and empty slots and only then _SUN
which is present only when slot is hotpluggable.

Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Message-Id: <20220701133515.137890-13-imammedo@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

tests: acpi: whitelist pc/q35 DSDT before moving _ADR field

Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Message-Id: <20220701133515.137890-12-imammedo@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

tests: acpi: update expected blobs

It's expected that hotpluggable slots will, get ASUN variable
and use that instead of _SUN with its _DSM method.

For example:

  @@ -979,8 +979,9 @@ DefinitionBlock ("", "DSDT", 1, "BOCHS ", "BXPC    ", 0x00000001)

               Device (S18)
               {
  -                Name (_SUN, 0x03)  // _SUN: Slot User Number
  +                Name (ASUN, 0x03)
                   Name (_ADR, 0x00030000)  // _ADR: Address
  +                Name (_SUN, 0x03)  // _SUN: Slot User Number
                   Method (_EJ0, 1, NotSerialized)  // _EJx: Eject Device
                   {
                       PCEJ (BSEL, _SUN)
  @@ -991,7 +992,7 @@ DefinitionBlock ("", "DSDT", 1, "BOCHS ", "BXPC    ", 0x00000001)
                       Local0 = Package (0x02)
                           {
                               BSEL,
  -                            _SUN
  +                            ASUN
                           }
                       Return (PDSM (Arg0, Arg1, Arg2, Arg3, Local0))
                   }

Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Message-Id: <20220701133515.137890-11-imammedo@redhat.com>

x86: acpi: cleanup PCI device _DSM duplication

add ASUN variable to hotpluggable slots and use it
instead of _SUN which has the same value to reuse
_DMS code on both branches (hot- and non-hotpluggable).
No functional change.

Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Message-Id: <20220701133515.137890-10-imammedo@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

tests: acpi: whitelist pc/q35 DSDT before switching _DSM to use ASUN

Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Message-Id: <20220701133515.137890-9-imammedo@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

tests: acpi: update expected blobs

An intermediate blobs update to keep changes (last 2 patches)
reviewable.

Includes refactored PDSM that uses Package argument for custom
parameters.

 ===== PDSM taking package as arguments

             Return (Local0)
         }

-        Method (PDSM, 6, Serialized)
+        Method (PDSM, 5, Serialized)
         {
-            If ((Arg0 == ToUUID ("e5c937d0-3553-4d7a-9117-ea4d19c3434d") /* Device Labeling Interface */))
+            If ((Arg2 == Zero))
             {
-                Local0 = AIDX (Arg4, Arg5)
-                If ((Arg2 == Zero))
-                {
-                    If ((Arg1 == 0x02))
+                Local0 = Buffer (One)
                     {
-                        If (!((Local0 == Zero) | (Local0 == 0xFFFFFFFF)))
-                        {
-                            Return (Buffer (One)
-                            {
-                                 0x81                                             // .
-                            })
-                        }
+                         0x00                                             // .
                     }
+                Local1 = Zero
+                If ((Arg0 != ToUUID ("e5c937d0-3553-4d7a-9117-ea4d19c3434d") /* Device Labeling Interface */))
+                {
+                    Return (Local0)
+                }

-                    Return (Buffer (One)
-                    {
-                         0x00                                             // .
-                    })
+                If ((Arg1 < 0x02))
+                {
+                    Return (Local0)
                 }
-                ElseIf ((Arg2 == 0x07))
+
+                Local2 = AIDX (DerefOf (Arg4 [Zero]), DerefOf (Arg4 [One]
+                    ))
+                If (!((Local2 == Zero) | (Local2 == 0xFFFFFFFF)))
                 {
-                    Local1 = Package (0x02)
-                        {
-                            Zero,
-                            ""
-                        }
-                    Local1 [Zero] = Local0
-                    Return (Local1)
+                    Local1 |= One
+                    Local1 |= (One << 0x07)
                 }
+
+                Local0 [Zero] = Local1
+                Return (Local0)
+            }
+
+            If ((Arg2 == 0x07))
+            {
+                Local0 = Package (0x02)
+                    {
+                        Zero,
+                        ""
+                    }
+                Local2 = AIDX (DerefOf (Arg4 [Zero]), DerefOf (Arg4 [One]
+                    ))
+                Local0 [Zero] = Local2
+                Return (Local0)
             }
         }
     }

 =====  PCI slot using Package to pass arguments to _DSM

                 Name (ASUN, Zero)
                 Method (_DSM, 4, Serialized)  // _DSM: Device-Specific Method
                 {
-                    Return (PDSM (Arg0, Arg1, Arg2, Arg3, BSEL, ASUN))
+                    Local0 = Package (0x02)
+                        {
+                            BSEL,
+                            ASUN
+                        }
+                    Return (PDSM (Arg0, Arg1, Arg2, Arg3, Local0))
                 }
             }

 ===== hotpluggable PCI slot using Package to pass arguments to _DSM

                 Method (_DSM, 4, Serialized)  // _DSM: Device-Specific Method
                 {
-                    Return (PDSM (Arg0, Arg1, Arg2, Arg3, BSEL, _SUN))
+                    Local0 = Package (0x02)
+                        {
+                            BSEL,
+                            _SUN
+                        }
+                    Return (PDSM (Arg0, Arg1, Arg2, Arg3, Local0))
                 }
             }

Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Message-Id: <20220701133515.137890-8-imammedo@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

x86: acpi: _DSM: use Package to pass parameters

Numer of possible arguments to pass to a method is limited
in ACPI. The following patches will need to pass over more
parameters to PDSM method, will hit that limit.

Prepare for this by passing structure (Package) to method,
which let us workaround arguments limitation.
Pass to PDSM all standard arguments of _DSM as is, and
pack custom parameters into Package that is passed as
the last argument to PDSM.

Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Message-Id: <20220701133515.137890-7-imammedo@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

acpi: x86: refactor PDSM method to reduce nesting

.., it will help with code readability and make easier
to extend method in followup patches

Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Message-Id: <20220701133515.137890-6-imammedo@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

tests: acpi: whitelist pc/q35 DSDT due to HPET AML move

Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Message-Id: <20220701133515.137890-5-imammedo@redhat.com>

tests: acpi: update expected blobs after HPET move

HPET AML moved after PCI host bridge description (no functional change)

diff example for PC machine:

@@ -54,47 +54,6 @@ DefinitionBlock ("", "DSDT", 1, "BOCHS ", "BXPC    ", 0x00000001)
         }
     }

-    Scope (_SB)
-    {
-        Device (HPET)
-        {
-            Name (_HID, EisaId ("PNP0103") /* HPET System Timer */)  // _HID: Hardware ID
-            Name (_UID, Zero)  // _UID: Unique ID
-            OperationRegion (HPTM, SystemMemory, 0xFED00000, 0x0400)
-            Field (HPTM, DWordAcc, Lock, Preserve)
-            {
-                VEND,   32,
-                PRD,    32
-            }
-
-            Method (_STA, 0, NotSerialized)  // _STA: Status
-            {
-                Local0 = VEND /* \_SB_.HPET.VEND */
-                Local1 = PRD /* \_SB_.HPET.PRD_ */
-                Local0 >>= 0x10
-                If (((Local0 == Zero) || (Local0 == 0xFFFF)))
-                {
-                    Return (Zero)
-                }
-
-                If (((Local1 == Zero) || (Local1 > 0x05F5E100)))
-                {
-                    Return (Zero)
-                }
-
-                Return (0x0F)
-            }
-
-            Name (_CRS, ResourceTemplate ()  // _CRS: Current Resource Settings
-            {
-                Memory32Fixed (ReadOnly,
-                    0xFED00000,         // Address Base
-                    0x00000400,         // Address Length
-                    )
-            })
-        }
-    }
-
     Scope (_SB.PCI0)
     {
         Device (ISA)
@@ -529,6 +488,47 @@ DefinitionBlock ("", "DSDT", 1, "BOCHS ", "BXPC    ", 0x00000001)
         }
     }

+    Scope (_SB)
+    {
+        Device (HPET)
+        {
+            Name (_HID, EisaId ("PNP0103") /* HPET System Timer */)  // _HID: Hardware ID
+            Name (_UID, Zero)  // _UID: Unique ID
+            OperationRegion (HPTM, SystemMemory, 0xFED00000, 0x0400)
+            Field (HPTM, DWordAcc, Lock, Preserve)
+            {
+                VEND,   32,
+                PRD,    32
+            }
+
+            Method (_STA, 0, NotSerialized)  // _STA: Status
+            {
+                Local0 = VEND /* \_SB_.HPET.VEND */
+                Local1 = PRD /* \_SB_.HPET.PRD_ */
+                Local0 >>= 0x10
+                If (((Local0 == Zero) || (Local0 == 0xFFFF)))
+                {
+                    Return (Zero)
+                }
+
+                If (((Local1 == Zero) || (Local1 > 0x05F5E100)))
+                {
+                    Return (Zero)
+                }
+
+                Return (0x0F)
+            }
+
+            Name (_CRS, ResourceTemplate ()  // _CRS: Current Resource Settings
+            {
+                Memory32Fixed (ReadOnly,
+                    0xFED00000,         // Address Base
+                    0x00000400,         // Address Length
+                    )
+            })
+        }
+    }
+
     Scope (_SB)
     {
         Device (\_SB.PCI0.PRES)

Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Message-Id: <20220701133515.137890-4-imammedo@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

acpi: x86: deduplicate HPET AML building

HPET AML doesn't depend on piix4 nor q35, move code buiding it
to common scope to avoid duplication.

Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Message-Id: <20220701133515.137890-3-imammedo@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

tests: acpi: whitelist pc/q35 DSDT due to HPET AML move

Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Message-Id: <20220701133515.137890-2-imammedo@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

hw/smbios: support for type 8 (port connector)

PATCH v1: add support for SMBIOS type 8 to qemu
PATCH v2: incorporate patch v1 feedback and add smbios type=8 to qemu-options

internal_reference: internal reference designator
external_reference: external reference designator
connector_type: hex value for port connector type (see SMBIOS 7.9.2)
port_type: hex value for port type (see SMBIOS 7.9.3)

After studying various vendor implementationsi (Dell, Lenovo, MSI),
the value of internal connector type was hard-coded to 0x0 (None).

Example usage:
-smbios type=8,internal_reference=JUSB1,external_reference=USB1,connector_type=0x12,port_type=0x10 \
-smbios type=8,internal_reference=JAUD1,external_reference="Audio Jack",connector_type=0x1f,port_type=0x1d \
-smbios type=8,internal_reference=LAN,external_reference=Ethernet,connector_type=0x0b,port_type=0x1f \
-smbios type=8,internal_reference=PS2,external_reference=Mouse,connector_type=0x0f,port_type=0x0e \
-smbios type=8,internal_reference=PS2,external_reference=Keyboard,connector_type=0x0f,port_type=0x0d

Signed-off-by: Hal Martin <hal.martin@gmail.com>
Message-Id: <20220812135153.17859-1-hal.martin@gmail.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

pci: Sanity check mask argument to pci_set_*_by_mask()

Coverity complains that in functions like pci_set_word_by_mask()
we might end up shifting by more than 31 bits. This is true,
but only if the caller passes in a zero mask. Help Coverity out
by asserting that the mask argument is valid.

Fixes: CID 1487168
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Message-Id: <20220818135421.2515257-3-peter.maydell@linaro.org>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>

pci: Remove unused pci_get_*_by_mask() functions

The helper functions pci_get_{byte,word,long,quad}_by_mask()
were added in 2012 in commit c9f50cea70a1596. In the decade
since we have never added a single use of them.

The helpers clearly aren't that helpful, so drop them
rather than carrying around dead code.

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Message-Id: <20220818135421.2515257-2-peter.maydell@linaro.org>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>

hmp: add virtio commands

This patch implements the HMP versions of the virtio QMP commands.

[Jonah: Adjusted hmp monitor output format for features / statuses
with their descriptions.]

Signed-off-by: Laurent Vivier <lvivier@redhat.com>
Signed-off-by: Jonah Palmer <jonah.palmer@oracle.com>
Message-Id: <1660220684-24909-7-git-send-email-jonah.palmer@oracle.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

qmp: add QMP command x-query-virtio-queue-element

This new command shows the information of a VirtQueue element.

[Note: Up until v10 of this patch series, virtio.json had many (15+)
 enums defined (e.g. decoded device features, statuses, etc.). In v10
 most of these enums were removed and replaced with string literals.
 By doing this we get (1) simpler schema, (2) smaller generated code,
 and (3) less maintenance burden for when new things are added (e.g.
 devices, device features, etc.).]

Signed-off-by: Laurent Vivier <lvivier@redhat.com>
Signed-off-by: Jonah Palmer <jonah.palmer@oracle.com>
Message-Id: <1660220684-24909-6-git-send-email-jonah.palmer@oracle.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

qmp: add QMP commands for virtio/vhost queue-status

These new commands show the internal status of a VirtIODevice's
VirtQueue and a vhost device's vhost_virtqueue (if active).

Signed-off-by: Laurent Vivier <lvivier@redhat.com>
Signed-off-by: Jonah Palmer <jonah.palmer@oracle.com>
Message-Id: <1660220684-24909-5-git-send-email-jonah.palmer@oracle.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

qmp: decode feature & status bits in virtio-status

Display feature names instead of bitmaps for host, guest, and
backend for VirtIODevices.

Display status names instead of bitmaps for VirtIODevices.

Display feature names instead of bitmaps for backend, protocol,
acked, and features (hdev->features) for vhost devices.

Decode features according to device ID. Decode statuses
according to configuration status bitmap (config_status_map).
Decode vhost user protocol features according to vhost user
protocol bitmap (vhost_user_protocol_map).

Transport features are on the first line. Undecoded bits (if
any) are stored in a separate field.

[Jonah: Several changes made to this patch from prev. version (v14):
 - Moved all device features mappings to hw/virtio/virtio.c
 - Renamed device features mappings (less generic)
 - Generalized @FEATURE_ENTRY macro for all device mappings
 - Virtio device feature map definitions include descriptions of
   feature bits
 - Moved @VHOST_USER_F_PROTOCOL_FEATURES feature bit from transport
   feature map to vhost-user-supported device feature mappings
   (blk, fs, i2c, rng, net, gpu, input, scsi, vsock)
 - New feature bit added for virtio-vsock: @VIRTIO_VSOCK_F_SEQPACKET
 - New feature bit added for virtio-iommu: @VIRTIO_IOMMU_F_BYPASS_CONFIG
 - New feature bit added for virtio-mem: @VIRTIO_MEM_F_UNPLUGGED_INACCESSIBLE
 - New virtio transport feature bit added: @VIRTIO_F_IN_ORDER
 - Added device feature map definition for virtio-rng
]

Signed-off-by: Laurent Vivier <lvivier@redhat.com>
Signed-off-by: Jonah Palmer <jonah.palmer@oracle.com>
Message-Id: <1660220684-24909-4-git-send-email-jonah.palmer@oracle.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

qmp: add QMP command x-query-virtio-status

This new command shows the status of a VirtIODevice, including
its corresponding vhost device's status (if active).

Next patch will improve output by decoding feature bits, including
vhost device's feature bits (backend, protocol, acked, and features).
Also will decode status bits of a VirtIODevice.

[Jonah: From patch v12; added a check to @virtio_device_find to ensure
 synchronicity between @virtio_list and the devices in the QOM
 composition tree.]

Signed-off-by: Laurent Vivier <lvivier@redhat.com>
Signed-off-by: Jonah Palmer <jonah.palmer@oracle.com>
Message-Id: <1660220684-24909-3-git-send-email-jonah.palmer@oracle.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

qmp: add QMP command x-query-virtio

This new command lists all the instances of VirtIODevices with
their canonical QOM path and name.

[Jonah: @virtio_list duplicates information that already exists in
 the QOM composition tree. However, extracting necessary information
 from this tree seems to be a bit convoluted.

 Instead, we still create our own list of realized virtio devices
 but use @qmp_qom_get with the device's canonical QOM path to confirm
 that the device exists and is realized. If the device exists but
 is actually not realized, then we remove it from our list (for
 synchronicity to the QOM composition tree).

 Also, the QMP command @x-query-virtio is redundant as @qom-list
 and @qom-get are sufficient to search '/machine/' for realized
 virtio devices. However, @x-query-virtio is much more convenient
 in listing realized virtio devices.]

Signed-off-by: Laurent Vivier <lvivier@redhat.com>
Signed-off-by: Jonah Palmer <jonah.palmer@oracle.com>
Message-Id: <1660220684-24909-2-git-send-email-jonah.palmer@oracle.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

Revert "intel_iommu: Fix irqchip / X2APIC configuration checks"

It's true that when vcpus<=255 we don't require the length of 32bit APIC
IDs.  However here since we already have EIM=ON it means the hypervisor
will declare the VM as x2apic supported (e.g. VT-d ECAP register will have
EIM bit 4 set), so the guest should assume the APIC IDs are 32bits width
even if vcpus<=255.  In short, commit 77250171bdc breaks any simple cmdline
that wants to boot a VM with >=9 but <=255 vcpus with:

  -device intel-iommu,intremap=on

For anyone who does not want to enable x2apic, we can use eim=off in the
intel-iommu parameters to skip enabling KVM x2apic.

This partly reverts commit 77250171bdc02aee106083fd2a068147befa1a38, while
keeping the valid bit on checking split irqchip, but revert the other change.

One thing to mention is that this patch may break migration compatibility
of such VM, however that's probably the best thing we can do, because the
old behavior was simply wrong and not working for >8 vcpus.  For <=8 vcpus,
there could be a light guest ABI change (by enabling KVM x2apic after this
patch), but logically it shouldn't affect the migration from working.

Also, this is not the 1st commit to change x2apic behavior.  Igor provided
a full history of how this evolved for the past few years:

https://lore.kernel.org/qemu-devel/20220922154617.57d1a1fb@redhat.com/

Relevant commits for reference:

  fb506e701e ("intel_iommu: reject broken EIM", 2016-10-17)
  c1bb5418e3 ("target/i386: Support up to 32768 CPUs without IRQ remapping", 2020-12-10)
  77250171bd ("intel_iommu: Fix irqchip / X2APIC configuration checks", 2022-05-16)
  dc89f32d92 ("target/i386: Fix sanity check on max APIC ID / X2APIC enablement", 2022-05-16)

We may want to have this for stable too (mostly for 7.1.0 only).  Adding a
fixes tag.

Cc: David Woodhouse <dwmw2@infradead.org>
Cc: Claudio Fontana <cfontana@suse.de>
Cc: Igor Mammedov <imammedo@redhat.com>
Fixes: 77250171bd ("intel_iommu: Fix irqchip / X2APIC configuration checks")
Signed-off-by: Peter Xu <peterx@redhat.com>
Message-Id: <20220926153206.10881-1-peterx@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Igor Mammedov <imammedo@redhat.com>

mem/cxl-type3: Add sn option to provide serial number for PCI ecap

The Device Serial Number Extended Capability PCI r6.0 sec 7.9.3
provides a standard way to provide a device serial number as
an IEEE defined 64-bit extended unique identifier EUI-64.

CXL 2.0 section 8.1.12.2 Memory Device PCIe Capabilities and
Extended Capabilities requires this to be used to uniquely
identify CXL memory devices.

Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Message-Id: <20220923161835.9805-1-Jonathan.Cameron@huawei.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Ben Widawsky <bwidawsk@kernel.org>

tests/acpi: virt: update ACPI GTDT binaries

Step 6 & 7 of the bios-tables-test.c documented procedure.

Differences between disassembled ASL files for GTDT:

    @@ -13,14 +13,14 @@
     [000h 0000   4]                    Signature : "GTDT"    [Generic Timer Description Table]
     [004h 0004   4]                 Table Length : 00000060
     [008h 0008   1]                     Revision : 02
    -[009h 0009   1]                     Checksum : 8C
    +[009h 0009   1]                     Checksum : 9C
     [00Ah 0010   6]                       Oem ID : "BOCHS "
     [010h 0016   8]                 Oem Table ID : "BXPC    "
     [018h 0024   4]                 Oem Revision : 00000001
     [01Ch 0028   4]              Asl Compiler ID : "BXPC"
     [020h 0032   4]        Asl Compiler Revision : 00000001

    -[024h 0036   8]        Counter Block Address : 0000000000000000
    +[024h 0036   8]        Counter Block Address : FFFFFFFFFFFFFFFF
     [02Ch 0044   4]                     Reserved : 00000000

     [030h 0048   4]         Secure EL1 Interrupt : 0000001D
    @@ -46,16 +46,16 @@
                                     Trigger Mode : 0
                                         Polarity : 0
                                        Always On : 0
    -[050h 0080   8]   Counter Read Block Address : 0000000000000000
    +[050h 0080   8]   Counter Read Block Address : FFFFFFFFFFFFFFFF

     [058h 0088   4]         Platform Timer Count : 00000000
     [05Ch 0092   4]        Platform Timer Offset : 00000000

     Raw Table Data: Length 96 (0x60)

    -    0000: 47 54 44 54 60 00 00 00 02 8C 42 4F 43 48 53 20  // GTDT`.....BOCHS
    +    0000: 47 54 44 54 60 00 00 00 02 9C 42 4F 43 48 53 20  // GTDT`.....BOCHS
         0010: 42 58 50 43 20 20 20 20 01 00 00 00 42 58 50 43  // BXPC    ....BXPC
    -    0020: 01 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00  // ................
    +    0020: 01 00 00 00 FF FF FF FF FF FF FF FF 00 00 00 00  // ................
         0030: 1D 00 00 00 00 00 00 00 1E 00 00 00 04 00 00 00  // ................
         0040: 1B 00 00 00 00 00 00 00 1A 00 00 00 00 00 00 00  // ................
    -    0050: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00  // ................
    +    0050: FF FF FF FF FF FF FF FF 00 00 00 00 00 00 00 00  // ................

Signed-off-by: Miguel Luis <miguel.luis@oracle.com>
Message-Id: <20220920162137.75239-4-miguel.luis@oracle.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Acked-by: Ani Sinha <ani@anisinha.ca>

acpi: arm/virt: build_gtdt: fix invalid 64-bit physical addresses

Per the ACPI 6.5 specification, on the GTDT Table Structure, the Counter Control
Block Address and Counter Read Block Address fields of the GTDT table should be
set to 0xFFFFFFFFFFFFFFFF if not provided, rather than 0x0.

Fixes: 41041e57085 ("acpi: arm/virt: build_gtdt: use acpi_table_begin()/acpi_table_end() instead of build_header()")
Signed-off-by: Miguel Luis <miguel.luis@oracle.com>
Message-Id: <20220920162137.75239-3-miguel.luis@oracle.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Ani Sinha <ani@anisinha.ca>

tests/acpi: virt: allow acpi GTDT changes

Step 3 from bios-tables-test.c documented procedure.

Signed-off-by: Miguel Luis <miguel.luis@oracle.com>
Message-Id: <20220920162137.75239-2-miguel.luis@oracle.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Acked-by: Ani Sinha <ani@anisinha.ca>

vhost-user-blk: dynamically resize config space based on features

Make vhost-user-blk backwards compatible when migrating from older VMs
running with modern features turned off, the same way it was done for
virtio-blk in 20764be0421c ("virtio-blk: set config size depending on the features enabled")

It's currently impossible to migrate from an older VM with
vhost-user-blk (with disable-legacy=off) because of errors like this:

qemu-system-x86_64: get_pci_config_device: Bad config data: i=0x10 read: 41 device: 1 cmask: ff wmask: 80 w1cmask:0
qemu-system-x86_64: Failed to load PCIDevice:config
qemu-system-x86_64: Failed to load virtio-blk:virtio
qemu-system-x86_64: error while loading state for instance 0x0 of device '0000:00:05.0:00.0:02.0/virtio-blk'
qemu-system-x86_64: load of migration failed: Invalid argument

This is caused by the newer (destination) VM requiring a bigger BAR0
alignment because it has to cover a bigger configuration space, which
isn't actually needed since those additional config fields are not
active (write-zeroes/discard).

Signed-off-by: Daniil Tatianin <d-tatianin@yandex-team.ru>
Reviewed-by: Raphael Norwitz <raphael.norwitz@nutanix.com>
Message-Id: <20220906073111.353245-6-d-tatianin@yandex-team.ru>
Reviewed-by: Raphael Norwitz <raphael.norwitz@nutanix.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

vhost-user-blk: make 'config_wce' part of 'host_features'

No reason to have this be a separate field. This also makes it more akin
to what the virtio-blk device does.

Signed-off-by: Daniil Tatianin <d-tatianin@yandex-team.ru>
Reviewed-by: Raphael Norwitz <raphael.norwitz@nutanix.com>
Message-Id: <20220906073111.353245-5-d-tatianin@yandex-team.ru>
Reviewed-by: Raphael Norwitz <raphael.norwitz@nutanix.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

vhost-user-blk: make it possible to disable write-zeroes/discard

It is useful to have the ability to disable these features for
compatibility with older VMs that don't have these implemented.

Signed-off-by: Daniil Tatianin <d-tatianin@yandex-team.ru>
Reviewed-by: Raphael Norwitz <raphael.norwitz@nutanix.com>
Message-Id: <20220906073111.353245-4-d-tatianin@yandex-team.ru>
Reviewed-by: Raphael Norwitz <raphael.norwitz@nutanix.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

virtio-blk: move config size params to virtio-blk-common

This way we can reuse it for other virtio-blk devices, e.g
vhost-user-blk, which currently does not control its config space size
dynamically.

Signed-off-by: Daniil Tatianin <d-tatianin@yandex-team.ru>
Reviewed-by: Raphael Norwitz <raphael.norwitz@nutanix.com>
Message-Id: <20220906073111.353245-3-d-tatianin@yandex-team.ru>
Reviewed-by: Raphael Norwitz <raphael.norwitz@nutanix.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

virtio: introduce VirtIOConfigSizeParams & virtio_get_config_size

This is the first step towards moving all device config size calculation
logic into the virtio core code. In particular, this adds a struct that
contains all the necessary information for common virtio code to be able
to calculate the final config size for a device. This is expected to be
used with the new virtio_get_config_size helper, which calculates the
final length based on the provided host features.

This builds on top of already existing code like VirtIOFeature and
virtio_feature_get_config_size(), but adds additional fields, as well as
sanity checking so that device-specifc code doesn't have to duplicate it.

An example usage would be:

    static const VirtIOFeature dev_features[] = {
        {.flags = 1ULL << FEATURE_1_BIT,
         .end = endof(struct virtio_dev_config, feature_1)},
        {.flags = 1ULL << FEATURE_2_BIT,
         .end = endof(struct virtio_dev_config, feature_2)},
        {}
    };

    static const VirtIOConfigSizeParams dev_cfg_size_params = {
        .min_size = DEV_BASE_CONFIG_SIZE,
        .max_size = sizeof(struct virtio_dev_config),
        .feature_sizes = dev_features
    };

    // code inside my_dev_device_realize()
    size_t config_size = virtio_get_config_size(&dev_cfg_size_params,
                                                host_features);
    virtio_init(vdev, VIRTIO_ID_MYDEV, config_size);

Currently every device is expected to write its own boilerplate from the
example above in device_realize(), however, the next step of this
transition is moving VirtIOConfigSizeParams into VirtioDeviceClass,
so that it can be done automatically by the virtio initialization code.

All of the users of virtio_feature_get_config_size have been converted
to use virtio_get_config_size so it's no longer needed and is removed
with this commit.

Signed-off-by: Daniil Tatianin <d-tatianin@yandex-team.ru>
Message-Id: <20220906073111.353245-2-d-tatianin@yandex-team.ru>
Reviewed-by: Raphael Norwitz <raphael.norwitz@nutanix.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

tests/qtest: enable tests for virtio-gpio

We don't have a virtio-gpio implementation in QEMU and only
support a vhost-user backend. The QEMU side of the code is minimal so
it should be enough to instantiate the device and pass some vhost-user
messages over the control socket. To do this we hook into the existing
vhost-user-test code and just add the bits required for gpio.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Cc: Viresh Kumar <viresh.kumar@linaro.org>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Cc: Eric Auger <eric.auger@redhat.com>
Message-Id: <20220408155704.2777166-1-alex.bennee@linaro.org>

Message-Id: <20220802095010.3330793-23-alex.bennee@linaro.org>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

tests/qtest: add a get_features op to vhost-user-test

As we expand this test for more virtio devices we will need to support
different feature sets. Add a mandatory op field to fetch the list of
features needed for the test itself.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-Id: <20220802095010.3330793-22-alex.bennee@linaro.org>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

tests/qtest: implement stub for VHOST_USER_GET_CONFIG

We don't implement the full solution because frankly none of the tests
need to at the moment. We may end up re-implementing libvhostuser in
the end.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-Id: <20220802095010.3330793-21-alex.bennee@linaro.org>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

tests/qtest: add assert to catch bad features

No device driver (which is what the qvirtio_ access functions
represent) should be setting UNUSED(30) in the feature space. Although
existing libqos users mask it out lets ensure nothing sneaks through.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-Id: <20220802095010.3330793-20-alex.bennee@linaro.org>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

tests/qtest: plain g_assert for VHOST_USER_F_PROTOCOL_FEATURES

checkpatch.pl warns that non-plain asserts should be avoided so
convert the check to a plain g_assert.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-Id: <20220802095010.3330793-19-alex.bennee@linaro.org>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

tests/qtest: catch unhandled vhost-user messages

We don't need to action every message but lets document the ones we
are expecting to consume so future tests don't get confused about
unhandled bits.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-Id: <20220802095010.3330793-18-alex.bennee@linaro.org>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

tests/qtest: use qos_printf instead of g_test_message

The vhost-user tests respawn qos-test as a standalone process. As a
result the gtester framework squashes all messages coming out of it
which make it hard to debug. As the test does not care about asserting
certain messages just convert the tests to use the direct qos_printf.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-Id: <20220802095010.3330793-17-alex.bennee@linaro.org>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

tests/qtest: add a timeout for subprocess_run_one_test

Hangs have been observed in the tests and currently we don't timeout
if a subprocess hangs. Rectify that.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Message-Id: <20220802095010.3330793-16-alex.bennee@linaro.org>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>

tests/qtest: pass stdout/stderr down to subtests

When trying to work out what the virtio-net-tests where doing it was
hard because the g_test_trap_subprocess redirects all output to
/dev/null. Lift this restriction by using the appropriate flags so you
can see something similar to what the vhost-user-blk tests show when
running.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Acked-by: Thomas Huth <thuth@redhat.com>
Message-Id: <20220407150042.2338562-1-alex.bennee@linaro.org>

Message-Id: <20220802095010.3330793-15-alex.bennee@linaro.org>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

hw/virtio: add vhost-user-gpio-pci boilerplate

This allows is to instantiate a vhost-user-gpio device as part of a PCI
bus. It is mostly boilerplate which looks pretty similar to the
vhost-user-fs-pci device.

Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Message-Id: <5f560cab92d0d789b1c94295ec74b9952907d69d.1641987128.git.viresh.kumar@linaro.org>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-Id: <20220802095010.3330793-14-alex.bennee@linaro.org>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

hw/virtio: add boilerplate for vhost-user-gpio device

This creates the QEMU side of the vhost-user-gpio device which connects
to the remote daemon. It is based of vhost-user-i2c code.

Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Message-Id: <5390324a748194a21bc99b1538e19761a8c64092.1641987128.git.viresh.kumar@linaro.org>
[AJB: fixes for qtest, tweaks to feature bits]
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Cc: Vincent Whitchurch <vincent.whitchurch@axis.com>
Message-Id: <20220802095010.3330793-13-alex.bennee@linaro.org>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

hw/virtio: move vhd->started check into helper and add FIXME

The `started` field is manipulated internally within the vhost code
except for one place, vhost-user-blk via f5b22d06fb (vhost: recheck
dev state in the vhost_migration_log routine). Mark that as a FIXME
because it introduces a potential race. I think the referenced fix
should be tracking its state locally.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-Id: <20220802095010.3330793-12-alex.bennee@linaro.org>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Raphael Norwitz <raphael.norwittz@nutanix.com>

hw/virtio: move vm_running check to virtio_device_started

All the boilerplate virtio code does the same thing (or should at
least) of checking to see if the VM is running before attempting to
start VirtIO. Push the logic up to the common function to avoid
getting a copy and paste wrong.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-Id: <20220802095010.3330793-11-alex.bennee@linaro.org>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

hw/virtio: add some vhost-user trace events

These are useful for tracing the lifetime of vhost-user connections.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-Id: <20220802095010.3330793-10-alex.bennee@linaro.org>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>

hw/virtio: log potentially buggy guest drivers

If the guest driver attempts to use the UNUSED(30) bit it is
potentially buggy as 6.3 Legacy Interface: Reserved Feature Bits
states it "SHOULD NOT be negotiated". For now just log this guest
error.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Acked-by: Jason Wang <jasowang@redhat.com>
Message-Id: <20220802095010.3330793-9-alex.bennee@linaro.org>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

hw/virtio: fix some coding style issues

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Acked-by: Jason Wang <jasowang@redhat.com>
Message-Id: <20220802095010.3330793-8-alex.bennee@linaro.org>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>

include/hw: document vhost_dev feature life-cycle

Try and explicitly document the various state of feature bits as
related to the vhost_dev structure. Importantly the backend_features
can advertise things like VHOST_USER_F_PROTOCOL_FEATURES which is
never exposed to the driver and is only present in the vhost-user
feature negotiation.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Acked-by: Jason Wang <jasowang@redhat.com>
Message-Id: <20220802095010.3330793-7-alex.bennee@linaro.org>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>

include/hw/virtio: more comment for VIRTIO_F_BAD_FEATURE

When debugging a new vhost user you may be surprised to see
VHOST_USER_F_PROTOCOL getting squashed in the maze of
backend_features, acked_features and guest_features. Expand the
description here to help the next poor soul trying to work through
this.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-Id: <20220802095010.3330793-6-alex.bennee@linaro.org>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

hw/virtio: incorporate backend features in features

There are some extra bits used over a vhost-user connection which are
hidden from the device itself. We need to set them here to ensure we
enable things like the protocol extensions.

Currently net/vhost-user.c has it's own inscrutable way of persisting
this data but it really should live in the core vhost_user code.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-Id: <20220726192150.2435175-7-alex.bennee@linaro.org>
Message-Id: <20220802095010.3330793-2-alex.bennee@linaro.org>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>

file-posix: Remove unused s->discard_zeroes

The field is unused (only ever set, but never read) since commit
ac9185603. Additionally, the commit message of commit 34fa110e already
explained earlier why it's unreliable. Remove it.

Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20220923142838.91043-1-kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

job: remove unused functions

These public functions are not used anywhere, thus can be dropped.
Also, since this is the final job API that doesn't use AioContext
lock and replaces it with job_lock, adjust all remaining function
documentation to clearly specify if the job lock is taken or not.

Also document the locking requirements for a few functions
where the second version is not removed.

Signed-off-by: Emanuele Giuseppe Esposito <eesposit@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20220926093214.506243-22-eesposit@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

blockjob: remove unused functions

These public functions are not used anywhere, thus can be dropped.

Signed-off-by: Emanuele Giuseppe Esposito <eesposit@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
Message-Id: <20220926093214.506243-21-eesposit@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

block_job_query: remove atomic read

Not sure what the atomic here was supposed to do, since job.busy
is protected by the job lock. Since the whole function
is called under job_mutex, just remove the atomic.

Signed-off-by: Emanuele Giuseppe Esposito <eesposit@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20220926093214.506243-20-eesposit@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

job.c: enable job lock/unlock and remove Aiocontext locks

Change the job_{lock/unlock} and macros to use job_mutex.

Now that they are not nop anymore, remove the aiocontext
to avoid deadlocks.

Therefore:
- when possible, remove completely the aiocontext lock/unlock pair
- if it is used by some other function too, reduce the locking
  section as much as possible, leaving the job API outside.
- change AIO_WAIT_WHILE in AIO_WAIT_WHILE_UNLOCKED, since we
  are not using the aiocontext lock anymore

The only functions that still need the aiocontext lock are:
- the JobDriver callbacks, already documented in job.h
- job_cancel_sync() in replication.c is called with aio_context_lock
  taken, but now job is using AIO_WAIT_WHILE_UNLOCKED so we need to
  release the lock.

Reduce the locking section to only cover the callback invocation
and document the functions that take the AioContext lock,
to avoid taking it twice.

Also remove real_job_{lock/unlock}, as they are replaced by the
public functions.

Signed-off-by: Emanuele Giuseppe Esposito <eesposit@redhat.com>
Message-Id: <20220926093214.506243-19-eesposit@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

job.h: categorize JobDriver callbacks that need the AioContext lock

Some callbacks implementation use bdrv_* APIs that assume the
AioContext lock is held. Make sure this invariant is documented.

Signed-off-by: Emanuele Giuseppe Esposito <eesposit@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
Message-Id: <20220926093214.506243-18-eesposit@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

blockjob: protect iostatus field in BlockJob struct

iostatus is the only field (together with .job) that needs
protection using the job mutex.

It is set in the main loop (GLOBAL_STATE functions) but read
in I/O code (block_job_error_action).

In order to protect it, change block_job_iostatus_set_err
to block_job_iostatus_set_err_locked(), always called under
job lock.

Signed-off-by: Emanuele Giuseppe Esposito <eesposit@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
Message-Id: <20220926093214.506243-17-eesposit@redhat.com>
[kwolf: Fixed up type of iostatus]
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

blockjob: rename notifier callbacks as _locked

They all are called with job_lock held, in job_event_*_locked()

Signed-off-by: Emanuele Giuseppe Esposito <eesposit@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20220926093214.506243-16-eesposit@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

blockjob.h: categorize fields in struct BlockJob

The same job lock is being used also to protect some of blockjob fields.
Categorize them just as done in job.h.

Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
Signed-off-by: Emanuele Giuseppe Esposito <eesposit@redhat.com>
Message-Id: <20220926093214.506243-15-eesposit@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

jobs: protect job.aio_context with BQL and job_mutex

In order to make it thread safe, implement a "fake rwlock",
where we allow reads under BQL *or* job_mutex held, but
writes only under BQL *and* job_mutex.

The only write we have is in child_job_set_aio_ctx, which always
happens under drain (so the job is paused).
For this reason, introduce job_set_aio_context and make sure that
the context is set under BQL, job_mutex and drain.
Also make sure all other places where the aiocontext is read
are protected.

The reads in commit.c and mirror.c are actually safe, because always
done under BQL.

Note: at this stage, job_{lock/unlock} and job lock guard macros
are *nop*.

Suggested-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Emanuele Giuseppe Esposito <eesposit@redhat.com>
Message-Id: <20220926093214.506243-14-eesposit@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

job: detect change of aiocontext within job coroutine

We want to make sure access of job->aio_context is always done
under either BQL or job_mutex. The problem is that using
aio_co_enter(job->aiocontext, job->co) in job_start and job_enter_cond
makes the coroutine immediately resume, so we can't hold the job lock.
And caching it is not safe either, as it might change.

job_start is under BQL, so it can freely read job->aiocontext, but
job_enter_cond is not.
We want to avoid reading job->aio_context in job_enter_cond, therefore:
1) use aio_co_wake(), since it doesn't want an aiocontext as argument
   but uses job->co->ctx
2) detect possible discrepancy between job->co->ctx and job->aio_context
   by checking right after the coroutine resumes back from yielding if
   job->aio_context has changed. If so, reschedule the coroutine to the
   new context.

Calling bdrv_try_set_aio_context() will issue the following calls
(simplified):
* in terms of  bdrv callbacks:
  .drained_begin -> .set_aio_context -> .drained_end
* in terms of child_job functions:
  child_job_drained_begin -> child_job_set_aio_context -> child_job_drained_end
* in terms of job functions:
  job_pause_locked -> job_set_aio_context -> job_resume_locked

We can see that after setting the new aio_context, job_resume_locked
calls again job_enter_cond, which then invokes aio_co_wake(). But
while job->aiocontext has been set in job_set_aio_context,
job->co->ctx has not changed, so the coroutine would be entering in
the wrong aiocontext.

Using aio_co_schedule in job_resume_locked() might seem as a valid
alternative, but the problem is that the bh resuming the coroutine
is not scheduled immediately, and if in the meanwhile another
bdrv_try_set_aio_context() is run (see test_propagate_mirror() in
test-block-iothread.c), we would have the first schedule in the
wrong aiocontext, and the second set of drains won't even manage
to schedule the coroutine, as job->busy would still be true from
the previous job_resume_locked().

The solution is to stick with aio_co_wake() and detect every time
the coroutine resumes back from yielding if job->aio_context
has changed. If so, we can reschedule it to the new context.

Check for the aiocontext change in job_do_yield_locked because:
1) aio_co_reschedule_self requires to be in the running coroutine
2) since child_job_set_aio_context allows changing the aiocontext only
   while the job is paused, this is the exact place where the coroutine
   resumes, before running JobDriver's code.

Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20220926093214.506243-13-eesposit@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

jobs: group together API calls under the same job lock

Now that the API offers also _locked() functions, take advantage
of it and give also the caller control to take the lock and call
_locked functions.

This makes sense especially when we have for loops, because it
makes no sense to have:

for(job = job_next(); ...)

where each job_next() takes the lock internally.
Instead we want

JOB_LOCK_GUARD();
for(job = job_next_locked(); ...)

In addition, protect also direct field accesses, by either creating a
new critical section or widening the existing ones.

Note: at this stage, job_{lock/unlock} and job lock guard macros
are *nop*.

Signed-off-by: Emanuele Giuseppe Esposito <eesposit@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
Message-Id: <20220926093214.506243-12-eesposit@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

block/mirror.c: use of job helpers in drivers

Once job lock is used and aiocontext is removed, mirror has
to perform job operations under the same critical section,
Note: at this stage, job_{lock/unlock} and job lock guard macros
are *nop*.

Signed-off-by: Emanuele Giuseppe Esposito <eesposit@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
Message-Id: <20220926093214.506243-11-eesposit@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

jobs: use job locks also in the unit tests

Add missing job synchronization in the unit tests, with
explicit locks.

We are deliberately using _locked functions wrapped by a guard
instead of a normal call because the normal call will be removed
in future, as the only usage is limited to the tests.

In other words, if a function like job_pause() is/will be only used
in tests to avoid:

WITH_JOB_LOCK_GUARD(){
    job_pause_locked();
}

then it is not worth keeping job_pause(), and just use the guard.

Note: at this stage, job_{lock/unlock} and job lock guard macros
are *nop*.

Signed-off-by: Emanuele Giuseppe Esposito <eesposit@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20220926093214.506243-10-eesposit@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

jobs: add job lock in find_* functions

Both blockdev.c and job-qmp.c have TOC/TOU conditions, because
they first search for the job and then perform an action on it.
Therefore, we need to do the search + action under the same
job mutex critical section.

Note: at this stage, job_{lock/unlock} and job lock guard macros
are *nop*.

Signed-off-by: Emanuele Giuseppe Esposito <eesposit@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20220926093214.506243-9-eesposit@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

blockjob: introduce block_job _locked() APIs

Just as done with job.h, create _locked() functions in blockjob.h

These functions will be later useful when caller has already taken
the lock. All blockjob _locked functions call job _locked functions.

Note: at this stage, job_{lock/unlock} and job lock guard macros
are *nop*.

Signed-off-by: Emanuele Giuseppe Esposito <eesposit@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20220926093214.506243-8-eesposit@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

job: move and update comments from blockjob.c

This comment applies more on job, it was left in blockjob as in the past
the whole job logic was implemented there.

Note: at this stage, job_{lock/unlock} and job lock guard macros
are *nop*.

No functional change intended.

Signed-off-by: Emanuele Giuseppe Esposito <eesposit@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20220926093214.506243-7-eesposit@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

job.c: add job_lock/unlock while keeping job.h intact

With "intact" we mean that all job.h functions implicitly
take the lock. Therefore API callers are unmodified.

This means that:
- many static functions that will be always called with job lock held
  become _locked, and call _locked functions
- all public functions take the lock internally if needed, and call _locked
  functions
- all public functions called internally by other functions in job.c will have a
  _locked counterpart (sometimes public), to avoid deadlocks (job lock already taken).
  These functions are not used for now.
- some public functions called only from exernal files (not job.c) do not
  have _locked() counterpart and take the lock inside. Others won't need
  the lock at all because use fields only set at initialization and
  never modified.

job_{lock/unlock} is independent from real_job_{lock/unlock}.

Note: at this stage, job_{lock/unlock} and job lock guard macros
are *nop*

Signed-off-by: Emanuele Giuseppe Esposito <eesposit@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
Message-Id: <20220926093214.506243-6-eesposit@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

aio-wait.h: introduce AIO_WAIT_WHILE_UNLOCKED

Same as AIO_WAIT_WHILE macro, but if we are in the Main loop
do not release and then acquire ctx_ 's aiocontext.

Once all Aiocontext locks go away, this macro will replace
AIO_WAIT_WHILE.

Signed-off-by: Emanuele Giuseppe Esposito <eesposit@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
Message-Id: <20220926093214.506243-5-eesposit@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

job.c: API functions not used outside should be static

job_event_* functions can all be static, as they are not used
outside job.c.

Same applies for job_txn_add_job().

Signed-off-by: Emanuele Giuseppe Esposito <eesposit@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20220926093214.506243-4-eesposit@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

job.h: categorize fields in struct Job

Categorize the fields in struct Job to understand which ones
need to be protected by the job mutex and which don't.

Signed-off-by: Emanuele Giuseppe Esposito <eesposit@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20220926093214.506243-3-eesposit@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

job.c: make job_mutex and job_lock/unlock() public

job mutex will be used to protect the job struct elements and list,
replacing AioContext locks.

Right now use a shared lock for all jobs, in order to keep things
simple. Once the AioContext lock is gone, we can introduce per-job
locks.

To simplify the switch from aiocontext to job lock, introduce
*nop* lock/unlock functions and macros.
We want to always call job_lock/unlock outside the AioContext locks,
and not vice-versa, otherwise we might get a deadlock. This is not
straightforward to do, and that's why we start with nop functions.
Once everything is protected by job_lock/unlock, we can change the nop into
an actual mutex and remove the aiocontext lock.

Since job_mutex is already being used, add static
real_job_{lock/unlock} for the existing usage.

Signed-off-by: Emanuele Giuseppe Esposito <eesposit@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
Message-Id: <20220926093214.506243-2-eesposit@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

quorum: Remove unnecessary forward declaration

Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20221006122607.162769-1-kwolf@redhat.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

test-coroutine: add missing coroutine_fn annotations

Callers of coroutine_fn must be coroutine_fn themselves, or the call
must be within "if (qemu_in_coroutine())".  Apply coroutine_fn to
functions where this holds.

Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Alberto Faria <afaria@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20220922084924.201610-27-pbonzini@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

migration: add missing coroutine_fn annotations

Callers of coroutine_fn must be coroutine_fn themselves, or the call
must be within "if (qemu_in_coroutine())".  Apply coroutine_fn to
functions where this holds.

Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Alberto Faria <afaria@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20220922084924.201610-26-pbonzini@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

9p: add missing coroutine_fn annotations

Callers of coroutine_fn must be coroutine_fn themselves, or the call
must be within "if (qemu_in_coroutine())".  Apply coroutine_fn to
functions where this holds.

Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Acked-by: Greg Kurz <groug@kaod.org>
Reviewed-by: Alberto Faria <afaria@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20220922084924.201610-25-pbonzini@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

raw-format: add missing coroutine_fn annotations

Callers of coroutine_fn must be coroutine_fn themselves, or the call
must be within "if (qemu_in_coroutine())".  Apply coroutine_fn to
functions where this holds.

Reviewed-by: Alberto Faria <afaria@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20220922084924.201610-24-pbonzini@redhat.com>
[kwolf: Fixed up coding style]
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

coroutine-lock: add missing coroutine_fn annotations

Callers of coroutine_fn must be coroutine_fn themselves, or the call
must be within "if (qemu_in_coroutine())".  Apply coroutine_fn to
functions where this holds.

Reviewed-by: Alberto Faria <afaria@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20220922084924.201610-23-pbonzini@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

job: add missing coroutine_fn annotations

Callers of coroutine_fn must be coroutine_fn themselves, or the call
must be within "if (qemu_in_coroutine())".  Apply coroutine_fn to
functions where this holds.

Reviewed-by: Alberto Faria <afaria@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20220922084924.201610-22-pbonzini@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

vmdk: add missing coroutine_fn annotations

Callers of coroutine_fn must be coroutine_fn themselves, or the call
must be within "if (qemu_in_coroutine())".  Apply coroutine_fn to
functions where this holds.

Reviewed-by: Alberto Faria <afaria@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20220922084924.201610-21-pbonzini@redhat.com>
[kwolf: Fixed up coding style]
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

throttle: add missing coroutine_fn annotations

Callers of coroutine_fn must be coroutine_fn themselves, or the call
must be within "if (qemu_in_coroutine())".  Apply coroutine_fn to
functions where this holds.

Reviewed-by: Alberto Faria <afaria@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20220922084924.201610-20-pbonzini@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

quorum: add missing coroutine_fn annotations

Callers of coroutine_fn must be coroutine_fn themselves, or the call
must be within "if (qemu_in_coroutine())".  Apply coroutine_fn to
functions where this holds.

Reviewed-by: Alberto Faria <afaria@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20220922084924.201610-19-pbonzini@redhat.com>
[kwolf: Fixed up coding style]
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

qed: add missing coroutine_fn annotations

Callers of coroutine_fn must be coroutine_fn themselves, or the call
must be within "if (qemu_in_coroutine())".  Apply coroutine_fn to
functions where this holds.

Reviewed-by: Alberto Faria <afaria@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20220922084924.201610-18-pbonzini@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

curl: add missing coroutine_fn annotations

Callers of coroutine_fn must be coroutine_fn themselves, or the call
must be within "if (qemu_in_coroutine())".  Apply coroutine_fn to
functions where this holds.

Reviewed-by: Alberto Faria <afaria@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20220922084924.201610-17-pbonzini@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>