target/arm: Add VHE system register redirection and aliasing

Several of the EL1/0 registers are redirected to the EL2 version when in
EL2 and HCR_EL2.E2H is set.  Many of these registers have side effects.
Link together the two ARMCPRegInfo structures after they have been
properly instantiated.  Install common dispatch routines to all of the
relevant registers.

The same set of registers that are redirected also have additional
EL12/EL02 aliases created to access the original register that was
redirected.

Omit the generic timer registers from redirection here, because we'll
need multiple kinds of redirection from both EL0 and EL2.

Tested-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20200206105448.4726-29-richard.henderson@linaro.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

target/arm: Update define_one_arm_cp_reg_with_opaque for VHE

For ARMv8.1, op1 == 5 is reserved for EL2 aliases of
EL1 and EL0 registers.

Tested-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20200206105448.4726-28-richard.henderson@linaro.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

target/arm: Update timer access for VHE

Tested-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20200206105448.4726-27-richard.henderson@linaro.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

target/arm: Add the hypervisor virtual counter

Tested-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20200206105448.4726-26-richard.henderson@linaro.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

target/arm: Update ctr_el0_access for EL2

Update to include checks against HCR_EL2.TID2.

Tested-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20200206105448.4726-25-richard.henderson@linaro.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

target/arm: Update aa64_zva_access for EL2

The comment that we don't support EL2 is somewhat out of date.
Update to include checks against HCR_EL2.TDZ.

Tested-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20200206105448.4726-24-richard.henderson@linaro.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

target/arm: Update arm_sctlr for VHE

Use the correct sctlr for EL2&0 regime.  Due to header ordering,
and where arm_mmu_idx_el is declared, we need to move the function
out of line.  Use the function in many more places in order to
select the correct control.

Tested-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20200206105448.4726-23-richard.henderson@linaro.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

target/arm: Update arm_mmu_idx for VHE

Return the indexes for the EL2&0 regime when the appropriate bits
are set within HCR_EL2.

Tested-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20200206105448.4726-22-richard.henderson@linaro.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

target/arm: Add regime_has_2_ranges

Create a predicate to indicate whether the regime has
both positive and negative addresses.

Tested-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20200206105448.4726-21-richard.henderson@linaro.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

target/arm: Reorganize ARMMMUIdx

Prepare for, but do not yet implement, the EL2&0 regime.
This involves adding the new MMUIdx enumerators and adjusting
some of the MMUIdx related predicates to match.

Tested-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20200206105448.4726-20-richard.henderson@linaro.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

target/arm: Tidy ARMMMUIdx m-profile definitions

Replace the magic numbers with the relevant ARM_MMU_IDX_M_* constants.
Keep the definitions short by referencing previous symbols.

Tested-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20200206105448.4726-19-richard.henderson@linaro.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

target/arm: Rearrange ARMMMUIdxBit

Define via macro expansion, so that renumbering of the base ARMMMUIdx
symbols is automatically reflected in the bit definitions.

Tested-by: Alex Bennée <alex.bennee@linaro.org>
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>
Message-id: 20200206105448.4726-18-richard.henderson@linaro.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

target/arm: Expand TBFLAG_ANY.MMUIDX to 4 bits

We are about to expand the number of mmuidx to 10, and so need 4 bits.
For the benefit of reading the number out of -d exec, align it to the
penultimate nibble.

Tested-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20200206105448.4726-17-richard.henderson@linaro.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

target/arm: Recover 4 bits from TBFLAGs

We had completely run out of TBFLAG bits.
Split A- and M-profile bits into two overlapping buckets.
This results in 4 free bits.

We used to initialize all of the a32 and m32 fields in DisasContext
by assignment, in arm_tr_init_disas_context.  Now we only initialize
either the a32 or m32 by assignment, because the bits overlap in
tbflags.  So zero the entire structure in gen_intermediate_code.

Tested-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20200206105448.4726-16-richard.henderson@linaro.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

target/arm: Rename ARMMMUIdx_S1E2 to ARMMMUIdx_E2

This is part of a reorganization to the set of mmu_idx.
The non-secure EL2 regime only has a single stage translation;
there is no point in pointing out that the idx is for stage1.

Tested-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20200206105448.4726-15-richard.henderson@linaro.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

target/arm: Rename ARMMMUIdx*_S1E3 to ARMMMUIdx*_SE3

This is part of a reorganization to the set of mmu_idx.
The EL3 regime only has a single stage translation, and
is always secure.

Tested-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20200206105448.4726-14-richard.henderson@linaro.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

target/arm: Rename ARMMMUIdx_S1SE[01] to ARMMMUIdx_SE10_[01]

This is part of a reorganization to the set of mmu_idx.
This emphasizes that they apply to the Secure EL1&0 regime.

Tested-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20200206105448.4726-13-richard.henderson@linaro.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

target/arm: Rename ARMMMUIdx_S1NSE* to ARMMMUIdx_Stage1_E*

This is part of a reorganization to the set of mmu_idx.
The EL1&0 regime is the only one that uses 2-stage translation.
Spelling out Stage avoids confusion with Secure.

Tested-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20200206105448.4726-12-richard.henderson@linaro.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

target/arm: Rename ARMMMUIdx_S2NS to ARMMMUIdx_Stage2

The EL1&0 regime is the only one that uses 2-stage translation.

Tested-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20200206105448.4726-11-richard.henderson@linaro.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

target/arm: Rename ARMMMUIdx*_S12NSE* to ARMMMUIdx*_E10_*

This is part of a reorganization to the set of mmu_idx.
This emphasizes that they apply to the EL1&0 regime.

The ultimate goal is

 -- Non-secure regimes:
    ARMMMUIdx_E10_0,
    ARMMMUIdx_E20_0,
    ARMMMUIdx_E10_1,
    ARMMMUIdx_E2,
    ARMMMUIdx_E20_2,

 -- Secure regimes:
    ARMMMUIdx_SE10_0,
    ARMMMUIdx_SE10_1,
    ARMMMUIdx_SE3,

 -- Helper mmu_idx for non-secure EL1&0 stage1 and stage2
    ARMMMUIdx_Stage2,
    ARMMMUIdx_Stage1_E0,
    ARMMMUIdx_Stage1_E1,

The 'S' prefix is reserved for "Secure".  Unless otherwise specified,
each mmu_idx represents all stages of translation.

Tested-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20200206105448.4726-10-richard.henderson@linaro.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

target/arm: Simplify tlb_force_broadcast alternatives

Rather than call to a separate function and re-compute any
parameters for the flush, simply use the correct flush
function directly.

Tested-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20200206105448.4726-9-richard.henderson@linaro.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

target/arm: Split out alle1_tlbmask

No functional change, but unify code sequences.

Tested-by: Alex Bennée <alex.bennee@linaro.org>
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>
Message-id: 20200206105448.4726-8-richard.henderson@linaro.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

target/arm: Split out vae1_tlbmask

No functional change, but unify code sequences.

Tested-by: Alex Bennée <alex.bennee@linaro.org>
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>
Message-id: 20200206105448.4726-7-richard.henderson@linaro.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

target/arm: Update CNTVCT_EL0 for VHE

The virtual offset may be 0 depending on EL, E2H and TGE.

Tested-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20200206105448.4726-6-richard.henderson@linaro.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

target/arm: Add TTBR1_EL2

At the same time, add writefn to TTBR0_EL2 and TCR_EL2.
A later patch will update any ASID therein.

Tested-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20200206105448.4726-5-richard.henderson@linaro.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

target/arm: Add CONTEXTIDR_EL2

Not all of the breakpoint types are supported, but those that
only examine contextidr are extended to support the new register.

Tested-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20200206105448.4726-4-richard.henderson@linaro.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

target/arm: Enable HCR_E2H for VHE

Tested-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20200206105448.4726-3-richard.henderson@linaro.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

target/arm: Define isar_feature_aa64_vh

Tested-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20200206105448.4726-2-richard.henderson@linaro.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

target/arm/monitor: query-cpu-model-expansion crashed qemu when using machine type none

Commit e19afd566781 mentioned that target-arm only supports queryable
cpu models 'max', 'host', and the current type when KVM is in use.
The logic works well until using machine type none.

For machine type none, cpu_type will be null if cpu option is not
set by command line, strlen(cpu_type) will terminate process.
So We add a check above it.

This won't affect i386 and s390x since they do not use current_cpu.

Signed-off-by: Liang Yan <lyan@suse.com>
Message-id: 20200203134251.12986-1-lyan@suse.com
Reviewed-by: Andrew Jones <drjones@redhat.com>
Tested-by: Andrew Jones <drjones@redhat.com>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

Merge remote-tracking branch 'remotes/awilliam/tags/vfio-update-20200206.0' into staging

VFIO update 2020-02-06

 - Split IGD to separate file and config option (Thomas Huth)

# gpg: Signature made Thu 06 Feb 2020 23:50:49 GMT
# gpg:                using RSA key 239B9B6E3BB08B22
# gpg: Good signature from "Alex Williamson <alex.williamson@redhat.com>" [full]
# gpg:                 aka "Alex Williamson <alex@shazbot.org>" [full]
# gpg:                 aka "Alex Williamson <alwillia@redhat.com>" [full]
# gpg:                 aka "Alex Williamson <alex.l.williamson@gmail.com>" [full]
# Primary key fingerprint: 42F6 C04E 540B D1A9 9E7B  8A90 239B 9B6E 3BB0 8B22

* remotes/awilliam/tags/vfio-update-20200206.0:
  hw/vfio: Move the IGD quirk code to a separate file

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

Merge remote-tracking branch 'remotes/ericb/tags/pull-nbd-2020-02-06' into staging

nbd patches for 2020-02-06

- Allow setting NBD description from QMP for parity with qemu-nbd
- Remove deprecated 'qemu-nbd --partition'

# gpg: Signature made Thu 06 Feb 2020 12:50:46 GMT
# gpg:                using RSA key 71C2CC22B1C4602927D2F3AAA7A16B4A2527436A
# gpg: Good signature from "Eric Blake <eblake@redhat.com>" [full]
# gpg:                 aka "Eric Blake (Free Software Programmer) <ebb9@byu.net>" [full]
# gpg:                 aka "[jpeg image of size 6874]" [full]
# Primary key fingerprint: 71C2 CC22 B1C4 6029 27D2  F3AA A7A1 6B4A 2527 436A

* remotes/ericb/tags/pull-nbd-2020-02-06:
  qemu-nbd: Removed deprecated --partition option
  docs: Fix typo in qemu-nbd -P replacement
  nbd: Allow description when creating NBD blockdev

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

hw/vfio: Move the IGD quirk code to a separate file

The IGD quirk code defines a separate device, the so-called
"vfio-pci-igd-lpc-bridge" which shows up as a user-creatable
device in all QEMU binaries that include the vfio code. This
is a little bit unfortunate for two reasons: First, this device
is completely useless in binaries like qemu-system-s390x.
Second we also would like to disable it in downstream RHEL
which currently requires some extra patches there since the
device does not have a proper Kconfig-style switch yet.

So it would be good if the device could be disabled more easily,
thus let's move the code to a separate file instead and introduce
a proper Kconfig switch for it which gets only enabled by default
if we also have CONFIG_PC_PCI enabled.

Signed-off-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>

Merge remote-tracking branch 'remotes/maxreitz/tags/pull-block-2020-02-06' into staging

Block patches:
- Drop BDRV_SECTOR_SIZE from qcow2
- Allow Python iotests to be added to the auto group
  (and add some)
- Fix for the backup job
- Fix memleak in bdrv_refresh_filename()
- Use GStrings in two places for greater efficiency (than manually
  handling string allocation)

# gpg: Signature made Thu 06 Feb 2020 12:50:14 GMT
# gpg:                using RSA key 91BEB60A30DB3E8857D11829F407DB0061D5CF40
# gpg:                issuer "mreitz@redhat.com"
# gpg: Good signature from "Max Reitz <mreitz@redhat.com>" [full]
# Primary key fingerprint: 91BE B60A 30DB 3E88 57D1  1829 F407 DB00 61D5 CF40

* remotes/maxreitz/tags/pull-block-2020-02-06:
  iotests: add test for backup-top failure on permission activation
  block/backup-top: fix failure path
  qcow2: Use BDRV_SECTOR_SIZE instead of the hardcoded value
  qcow2: Don't require aligned offsets in qcow2_co_copy_range_from()
  qcow2: Use bs->bl.request_alignment when updating an L1 entry
  qcow2: Tighten cluster_offset alignment assertions
  qcow2: Don't round the L1 table allocation up to the sector size
  iotests: Enable more tests in the 'auto' group to improve test coverage
  iotests: Skip Python-based tests if QEMU does not support virtio-blk
  iotests: Check for the availability of the required devices in 267 and 127
  iotests: Test 183 does not work on macOS and OpenBSD
  iotests: Test 041 only works on certain systems
  iotests: remove 'linux' from default supported platforms
  qcow2: Use a GString in report_unsupported_feature()
  block: fix memleaks in bdrv_refresh_filename
  block: Use a GString in bdrv_perm_names()
  qcow2: Assert that host cluster offsets fit in L2 table entries

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

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

trivial patches 20200206

# gpg: Signature made Thu 06 Feb 2020 12:49:19 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:
  MAINTAINERS: Cc the qemu-arm@nongnu.org for the ARM machines
  aspeed/i2c: Prevent uninitialized warning
  hw/pci/pci_bridge: Fix typo in comment
  qemu-img: Place the '-i aio' option in alphabetical order
  qemu-options: replace constant 1 with HAS_ARG
  MAINTAINERS: Cover hppa-softmmu.mak in the HP-PARISC Machines section
  hw/i386/vmmouse: Fix crash when using the vmmouse on a machine without vmport
  hw/bt: Remove empty Kconfig file
  hw/timer/m48t59: Convert debug printf()s to trace events
  MAINTAINERS: update Leif Lindholm's address
  monitor: fix memory leak in monitor_fdset_dup_fd_find_remove
  hw/smbios/smbios: Remove unused include

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

iotests: add test for backup-top failure on permission activation

This test checks that bug is really fixed by previous commit.

Cc: qemu-stable@nongnu.org # v4.2.0
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-id: 20200121142802.21467-3-vsementsov@virtuozzo.com
Signed-off-by: Max Reitz <mreitz@redhat.com>

block/backup-top: fix failure path

We can't access top after call bdrv_backup_top_drop, as it is already
freed at this time.

Also, no needs to unref target child by hand, it will be unrefed on
bdrv_close() automatically.

So, just do bdrv_backup_top_drop if append succeed and one bdrv_unref
otherwise.

Note, that in !appended case bdrv_unref(top) moved into drained section
on source. It doesn't really matter, but just for code simplicity.

Fixes: 7df7868b96404
Cc: qemu-stable@nongnu.org # v4.2.0
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-id: 20200121142802.21467-2-vsementsov@virtuozzo.com
Signed-off-by: Max Reitz <mreitz@redhat.com>

qcow2: Use BDRV_SECTOR_SIZE instead of the hardcoded value

This replaces all remaining instances in the qcow2 code.

Signed-off-by: Alberto Garcia <berto@igalia.com>
Message-id: b5f74b606c2d9873b12d29acdb7fd498029c4025.1579374329.git.berto@igalia.com
Reviewed-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>

qcow2: Don't require aligned offsets in qcow2_co_copy_range_from()

qemu-img's convert_co_copy_range() operates at the sector level and
block_copy() operates at the cluster level so this condition is always
true, but it is not necessary to restrict this here, so let's leave it
to the driver implementation return an error if there is any.

Signed-off-by: Alberto Garcia <berto@igalia.com>
Message-id: a4264aaee656910c84161a2965f7a501437379ca.1579374329.git.berto@igalia.com
Reviewed-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>

qcow2: Use bs->bl.request_alignment when updating an L1 entry

When updating an L1 entry the qcow2 driver writes a (512-byte) sector
worth of data to avoid a read-modify-write cycle. Instead of always
writing 512 bytes we should follow the alignment requirements of the
storage backend.

(the only exception is when the alignment is larger than the cluster
size because then we could be overwriting data after the L1 table)

Signed-off-by: Alberto Garcia <berto@igalia.com>
Message-id: 71f34d4ae4b367b32fb36134acbf4f4f7ee681f4.1579374329.git.berto@igalia.com
Reviewed-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>

qcow2: Tighten cluster_offset alignment assertions

qcow2_alloc_cluster_offset() and qcow2_get_cluster_offset() always
return offsets that are cluster-aligned so don't just check that they
are sector-aligned.

The check in qcow2_co_preadv_task() is also replaced by an assertion
for the same reason.

Signed-off-by: Alberto Garcia <berto@igalia.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-id: 558ba339965f858bede4c73ce3f50f0c0493597d.1579374329.git.berto@igalia.com
Signed-off-by: Max Reitz <mreitz@redhat.com>

qcow2: Don't round the L1 table allocation up to the sector size

The L1 table is read from disk using the byte-based bdrv_pread() and
is never accessed beyond its last element, so there's no need to
allocate more memory than that.

Signed-off-by: Alberto Garcia <berto@igalia.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-id: b2e27214ec7b03a585931bcf383ee1ac3a641a10.1579374329.git.berto@igalia.com
Signed-off-by: Max Reitz <mreitz@redhat.com>

iotests: Enable more tests in the 'auto' group to improve test coverage

According to Kevin, tests 030, 040 and 041 are among the most valuable
tests that we have, so we should always run them if possible, even if
they take a little bit longer.

According to Max, it would be good to have a test for iothreads and
migration. 127 and 256 seem to be good candidates for iothreads. For
migration, let's enable 181 and 203 (which also tests iothreads).
(091 would be a good candidate for migration, too, but Alex Bennée
reported that this test fails on ZFS file systems, so it can't be
included yet)

Reviewed-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
Message-id: 20200121095205.26323-7-thuth@redhat.com
Signed-off-by: Max Reitz <mreitz@redhat.com>

iotests: Skip Python-based tests if QEMU does not support virtio-blk

We are going to enable some of the python-based tests in the "auto" group,
and these tests require virtio-blk to work properly. Running iotests
without virtio-blk likely does not make too much sense anyway, so instead
of adding a check for the availability of virtio-blk to each and every
test (which does not sound very appealing), let's rather add a check for
this a central spot in the "check" script instead (so that it is still
possible to run "make check" for qemu-system-tricore for example).

Signed-off-by: Thomas Huth <thuth@redhat.com>
Message-id: 20200121095205.26323-6-thuth@redhat.com
Signed-off-by: Max Reitz <mreitz@redhat.com>

iotests: Check for the availability of the required devices in 267 and 127

We are going to enable 127 in the "auto" group, but it only works if
virtio-scsi and scsi-hd are available - which is not the case with
QEMU binaries like qemu-system-tricore for example, so we need a
proper check for the availability of these devices here.

A very similar problem exists in iotest 267 - it has been added to
the "auto" group already, but requires virtio-blk and thus currently
fails with qemu-system-tricore for example. Let's also add aproper
check there.

Reviewed-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
Message-id: 20200121095205.26323-5-thuth@redhat.com
Signed-off-by: Max Reitz <mreitz@redhat.com>

iotests: Test 183 does not work on macOS and OpenBSD

In the long run, we might want to add test 183 to the "auto" group
(but it still fails occasionally, so we cannot do that yet). However,
when running 183 in Cirrus-CI on macOS, or with our vm-build-openbsd
target, it currently always fails with an "Timeout waiting for return
on handle 0" error.

Let's mark it as supported only on systems where the test is working
most of the time (i.e. Linux, FreeBSD and NetBSD).

Reviewed-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
Message-id: 20200121095205.26323-4-thuth@redhat.com
Signed-off-by: Max Reitz <mreitz@redhat.com>

iotests: Test 041 only works on certain systems

041 works fine on Linux, FreeBSD, NetBSD and OpenBSD, but fails on macOS.
Let's mark it as only supported on the systems where we know that it is
working fine.

Reviewed-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
Message-id: 20200121095205.26323-3-thuth@redhat.com
Signed-off-by: Max Reitz <mreitz@redhat.com>

iotests: remove 'linux' from default supported platforms

verify_platform will check an explicit whitelist and blacklist instead.
The default will now be assumed to be allowed to run anywhere.

For tests that do not specify their platforms explicitly, this has the effect of
enabling these tests on non-linux platforms. For tests that always specified
linux explicitly, there is no change.

For Python tests on FreeBSD at least; only seven python tests fail:
045 147 149 169 194 199 211

045 and 149 appear to be misconfigurations,
147 and 194 are the AF_UNIX path too long error,
169 and 199 are bitmap migration bugs, and
211 is a bug that shows up on Linux platforms, too.

This is at least good evidence that these tests are not Linux-only. If
they aren't suitable for other platforms, they should be disabled on a
per-platform basis as appropriate.

Therefore, let's switch these on and deal with the failures.

Reviewed-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: John Snow <jsnow@redhat.com>
Message-id: 20200121095205.26323-2-thuth@redhat.com
Signed-off-by: Max Reitz <mreitz@redhat.com>

qcow2: Use a GString in report_unsupported_feature()

This is a bit more efficient than having to allocate and free memory
for each item.

The default size (60) is enough for all the existing incompatible
features or the "Unknown incompatible feature" message.

Suggested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Signed-off-by: Alberto Garcia <berto@igalia.com>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Message-id: 20200115135626.19442-1-berto@igalia.com
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Stefano Garzarella <sgarzare@redhat.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>

block: fix memleaks in bdrv_refresh_filename

If we call the qmp 'query-block' while qemu is working on
'block-commit', it will cause memleaks, the memory leak stack is as
follow:

Indirect leak of 12360 byte(s) in 3 object(s) allocated from:
    #0 0x7f80f0b6d970 in __interceptor_calloc (/lib64/libasan.so.5+0xef970)
    #1 0x7f80ee86049d in g_malloc0 (/lib64/libglib-2.0.so.0+0x5249d)
    #2 0x55ea95b5bb67 in qdict_new /mnt/sdb/qemu-4.2.0-rc0/qobject/qdict.c:29
    #3 0x55ea956cd043 in bdrv_refresh_filename /mnt/sdb/qemu-4.2.0-rc0/block.c:6427
    #4 0x55ea956cc950 in bdrv_refresh_filename /mnt/sdb/qemu-4.2.0-rc0/block.c:6399
    #5 0x55ea956cc950 in bdrv_refresh_filename /mnt/sdb/qemu-4.2.0-rc0/block.c:6399
    #6 0x55ea956cc950 in bdrv_refresh_filename /mnt/sdb/qemu-4.2.0-rc0/block.c:6399
    #7 0x55ea958818ea in bdrv_block_device_info /mnt/sdb/qemu-4.2.0-rc0/block/qapi.c:56
    #8 0x55ea958879de in bdrv_query_info /mnt/sdb/qemu-4.2.0-rc0/block/qapi.c:392
    #9 0x55ea9588b58f in qmp_query_block /mnt/sdb/qemu-4.2.0-rc0/block/qapi.c:578
    #10 0x55ea95567392 in qmp_marshal_query_block qapi/qapi-commands-block-core.c:95

Indirect leak of 4120 byte(s) in 1 object(s) allocated from:
    #0 0x7f80f0b6d970 in __interceptor_calloc (/lib64/libasan.so.5+0xef970)
    #1 0x7f80ee86049d in g_malloc0 (/lib64/libglib-2.0.so.0+0x5249d)
    #2 0x55ea95b5bb67 in qdict_new /mnt/sdb/qemu-4.2.0-rc0/qobject/qdict.c:29
    #3 0x55ea956cd043 in bdrv_refresh_filename /mnt/sdb/qemu-4.2.0-rc0/block.c:6427
    #4 0x55ea956cc950 in bdrv_refresh_filename /mnt/sdb/qemu-4.2.0-rc0/block.c:6399
    #5 0x55ea956cc950 in bdrv_refresh_filename /mnt/sdb/qemu-4.2.0-rc0/block.c:6399
    #6 0x55ea9569f301 in bdrv_backing_attach /mnt/sdb/qemu-4.2.0-rc0/block.c:1064
    #7 0x55ea956a99dd in bdrv_replace_child_noperm /mnt/sdb/qemu-4.2.0-rc0/block.c:2283
    #8 0x55ea956b9b53 in bdrv_replace_node /mnt/sdb/qemu-4.2.0-rc0/block.c:4196
    #9 0x55ea956b9e49 in bdrv_append /mnt/sdb/qemu-4.2.0-rc0/block.c:4236
    #10 0x55ea958c3472 in commit_start /mnt/sdb/qemu-4.2.0-rc0/block/commit.c:306
    #11 0x55ea94b68ab0 in qmp_block_commit /mnt/sdb/qemu-4.2.0-rc0/blockdev.c:3459
    #12 0x55ea9556a7a7 in qmp_marshal_block_commit qapi/qapi-commands-block-core.c:407

Fixes: bb808d5f5c0978828a974d547e6032402c339555
Reported-by: Euler Robot <euler.robot@huawei.com>
Signed-off-by: Pan Nengyuan <pannengyuan@huawei.com>
Message-id: 20200116085600.24056-1-pannengyuan@huawei.com
Signed-off-by: Max Reitz <mreitz@redhat.com>

block: Use a GString in bdrv_perm_names()

This is a bit more efficient than having to allocate and free memory
for each new permission.

The default size (30) is enough for "consistent read, write, resize".

Signed-off-by: Alberto Garcia <berto@igalia.com>
Message-id: 20200110171518.22168-1-berto@igalia.com
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Max Reitz <mreitz@redhat.com>

qcow2: Assert that host cluster offsets fit in L2 table entries

The standard cluster descriptor in L2 table entries has a field to
store the host cluster offset. When we need to get that offset from an
entry we use L2E_OFFSET_MASK to ensure that we only use the bits that
belong to that field.

But while that mask is used every time we read from an L2 entry, it
is never used when we write to it. Due to the QCOW_MAX_CLUSTER_OFFSET
limit set in the cluster allocation code QEMU can never produce
offsets that don't fit in that field so any such offset would indicate
a bug in QEMU.

Compressed cluster descriptors contain two fields (host cluster offset
and size of the compressed data) and the situation with them is
similar. In this case the masks are not constant but are stored in the
csize_mask and cluster_offset_mask fields of BDRVQcow2State.

Signed-off-by: Alberto Garcia <berto@igalia.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Message-id: 20200113161146.20099-1-berto@igalia.com
Signed-off-by: Max Reitz <mreitz@redhat.com>

MAINTAINERS: Cc the qemu-arm@nongnu.org for the ARM machines

Not all ARM machines sections Cc the qemu-arm@nongnu.org list,
fix this.

Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-Id: <20200120185928.25115-2-philmd@redhat.com>
Signed-off-by: Laurent Vivier <laurent@vivier.eu>

aspeed/i2c: Prevent uninitialized warning

Compiler reports uninitialized warning for cmd_flags variable.

Adding NULL initialization to prevent this warning.

Signed-off-by: Miroslav Rezanina <mrezanin@redhat.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Message-Id: <07957dcab31f65de3dd30efa91e6b9152ac79879.1579598240.git.mrezanin@redhat.com>
Signed-off-by: Laurent Vivier <laurent@vivier.eu>

hw/pci/pci_bridge: Fix typo in comment

Signed-off-by: Julia Suvorova <jusual@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20200205185123.210209-1-jusual@redhat.com>
Signed-off-by: Laurent Vivier <laurent@vivier.eu>

qemu-img: Place the '-i aio' option in alphabetical order

The '-i AIO' option was accidentally placed after '-n' and '-t'. Move it
after '--flush-interval'.

Signed-off-by: Julia Suvorova <jusual@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20200205163008.204493-1-jusual@redhat.com>
Signed-off-by: Laurent Vivier <laurent@vivier.eu>

qemu-options: replace constant 1 with HAS_ARG

This is the only instance of a non-zero constant not using a symbolic
constant.
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: John Snow <jsnow@redhat.com>
Message-Id: <20200204165638.25051-1-jsnow@redhat.com>
Signed-off-by: Laurent Vivier <laurent@vivier.eu>

MAINTAINERS: Cover hppa-softmmu.mak in the HP-PARISC Machines section

Modifications to default-configs/hppa-softmmu.mak should be
reviewed by the hppa-softmmu users (currently a single machine).

Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Acked-by: Helge Deller <deller@gmx.de>
Message-Id: <20200129190316.16901-1-philmd@redhat.com>
Signed-off-by: Laurent Vivier <laurent@vivier.eu>

hw/i386/vmmouse: Fix crash when using the vmmouse on a machine without vmport

QEMU currently crashes when the user tries to use the "vmmouse" on a
machine without vmport, e.g.:

 $ x86_64-softmmu/qemu-system-x86_64 -machine microvm -device vmmouse
 Segmentation fault (core dumped)

or:

 $ x86_64-softmmu/qemu-system-x86_64 -device vmmouse -M pc,vmport=off
 Segmentation fault (core dumped)

Let's avoid the crash by checking for the vmport device first.

Signed-off-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Darren Kenny <darren.kenny@oracle.com>
Message-Id: <20200129112954.4282-1-thuth@redhat.com>
Signed-off-by: Laurent Vivier <laurent@vivier.eu>

hw/bt: Remove empty Kconfig file

While removing the bluetooth code some weeks ago, I had to leave the
hw/bt/Kconfig file around. Otherwise some of the builds would have been
broken since the generated dependency files tried to include it before
they were rebuilt. Meanwhile, all those dependency files should have
been updated, so we can remove the empty Kconfig file now, too.

Signed-off-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Laurent Vivier <laurent@vivier.eu>
Message-Id: <20200123064525.6935-1-thuth@redhat.com>
Signed-off-by: Laurent Vivier <laurent@vivier.eu>

hw/timer/m48t59: Convert debug printf()s to trace events

Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Hervé Poussineau <hpoussin@reactos.org>
Message-Id: <20200117165809.31067-3-philmd@redhat.com>
Signed-off-by: Laurent Vivier <laurent@vivier.eu>

MAINTAINERS: update Leif Lindholm's address

Update address to reflect new employer.

Signed-off-by: Leif Lindholm <leif@nuviainc.com>
Reviewed-by: Leif Lindholm <leif.lindholm@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Acked-by: Leif Lindholm <leif@nuviainc.com>
Message-Id: <20200116174226.4780-1-leif@nuviainc.com>
[lv: added .mailmap changes]
Signed-off-by: Laurent Vivier <laurent@vivier.eu>

monitor: fix memory leak in monitor_fdset_dup_fd_find_remove

When remove dup_fd in monitor_fdset_dup_fd_find_remove function,
we need to free mon_fdset_fd_dup. ASAN shows memory leak stack:

Direct leak of 96 byte(s) in 3 object(s) allocated from:
    #0 0xfffd37b033b3 in __interceptor_calloc (/lib64/libasan.so.4+0xd33b3)
    #1 0xfffd375c71cb in g_malloc0 (/lib64/libglib-2.0.so.0+0x571cb)
    #2 0xaaae25bf1c17 in monitor_fdset_dup_fd_add /qemu/monitor/misc.c:1724
    #3 0xaaae265cfd8f in qemu_open /qemu/util/osdep.c:315
    #4 0xaaae264e2b2b in qmp_chardev_open_file_source /qemu/chardev/char-fd.c:122
    #5 0xaaae264e47cf in qmp_chardev_open_file /qemu/chardev/char-file.c:81
    #6 0xaaae264e118b in qemu_char_open /qemu/chardev/char.c:237
    #7 0xaaae264e118b in qemu_chardev_new /qemu/chardev/char.c:964
    #8 0xaaae264e1543 in qemu_chr_new_from_opts /qemu/chardev/char.c:680
    #9 0xaaae25e12e0f in chardev_init_func /qemu/vl.c:2083
    #10 0xaaae26603823 in qemu_opts_foreach /qemu/util/qemu-option.c:1170
    #11 0xaaae258c9787 in main /qemu/vl.c:4089
    #12 0xfffd35b80b9f in __libc_start_main (/lib64/libc.so.6+0x20b9f)
    #13 0xaaae258d7b63  (/qemu/build/aarch64-softmmu/qemu-system-aarch64+0x8b7b63)

Reported-by: Euler Robot <euler.robot@huawei.com>
Signed-off-by: Chen Qun <kuhn.chenqun@huawei.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-Id: <20200115072016.167252-1-kuhn.chenqun@huawei.com>
Signed-off-by: Laurent Vivier <laurent@vivier.eu>

hw/smbios/smbios: Remove unused include

Nothing from "sysemu/cpus.h" is used by smbios.c, remove the include.

Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Message-Id: <20200109112504.32622-1-philmd@redhat.com>
Signed-off-by: Laurent Vivier <laurent@vivier.eu>

qemu-nbd: Removed deprecated --partition option

The option was deprecated in 4.0.0 (commit 0ae2d546); it's now been
long enough with no complaints to follow through with that process.

Signed-off-by: Eric Blake <eblake@redhat.com>
Message-Id: <20200123164650.1741798-3-eblake@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>

docs: Fix typo in qemu-nbd -P replacement

The suggested replacement for the deprecated 'qemu-nbd -P' refers to
'file.backing.opt' instead of 'file.file.opt'; using the example
verbatim results in:

qemu-nbd: Failed to blk_new_open 'driver=raw,offset=1m,size=100m,file.driver=qcow2,file.backing.driver=file,file.backing.filename=file4': A block device must be specified for "file"

Correct this text, prior to actually finishing the deprecation process.

Fixes: 0ae2d54645eb
Reported-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Eric Blake <eblake@redhat.com>
Message-Id: <20200123164650.1741798-2-eblake@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>

nbd: Allow description when creating NBD blockdev

Allow blockdevs to match the feature already present in qemu-nbd -D.
Enhance iotest 223 to cover it.

Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Maxim Levitsky <mlevitsk@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20191114024635.11363-5-eblake@redhat.com>

Merge remote-tracking branch 'remotes/stsquad/tags/pull-testing-040220-1' into staging

Testing updates and build fixes:

  - move more cross compilers to buster
  - fix build breakage (hppa Kconfig)
  - disable docs on shippable
  - build docs under bionic with python3
  - travis.yml re-factoring
  - check capabilities of non-docker compilers
  - smarter make -j parallelism

# gpg: Signature made Tue 04 Feb 2020 17:16:40 GMT
# gpg:                using RSA key 6685AE99E75167BCAFC8DF35FBD0DB095A9E2A44
# gpg: Good signature from "Alex Bennée (Master Work Key) <alex.bennee@linaro.org>" [full]
# Primary key fingerprint: 6685 AE99 E751 67BC AFC8  DF35 FBD0 DB09 5A9E 2A44

* remotes/stsquad/tags/pull-testing-040220-1:
  .travis.yml: ensure python3-sphinx installed for docs
  .travis.yml: single thread build-tcg
  .travis.yml: drop cris-linux-user from the plugins test
  .travis.yml: drop the travis_retry from tests
  .travis.yml: introduce TEST_BUILD_CMD and use it for check-tcg
  tests/tcg: gate pauth-% tests on having compiler support
  tests/tcg: add a configure compiler check for ARMv8.1 and SVE
  .travis.yml: probe for number of available processors
  .travis.yml: move cache flushing to early common phase
  .travis.yml: build documents under bionic
  .travis.yml: Add description to each job
  .travis.yml: Drop superfluous use of --python=python3 parameter
  .shippable: --disable-docs for cross-compile tests
  travis.yml: Install genisoimage package
  tests/docker: better handle symlinked libs
  tests/docker: move most cross compilers to buster base

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

.travis.yml: ensure python3-sphinx installed for docs

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-Id: <20200204105142.21845-1-alex.bennee@linaro.org>

.travis.yml: single thread build-tcg

I've theorised that a parallel build-tcg is somehow getting confused
when two fedora-30 based cross compilers attempt to build at the same
time. From one data-point so far this may fix the problem although the
plugins job runs quite close to timeout.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-Id: <20200203090932.19147-18-alex.bennee@linaro.org>

.travis.yml: drop cris-linux-user from the plugins test

While it shouldn't cause problems we will never get useful information
from cris as it has yet to be converted to the common translator loop.
It also causes the Travis CI to fail for weird reasons which I have so
far been unable to replicate on a normal Xenial system.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-Id: <20200203090932.19147-17-alex.bennee@linaro.org>

.travis.yml: drop the travis_retry from tests

This was a crutch when we introduced it - however it does have the
disadvantage of causing tests to timeout with large amounts of logs.
Lets drop it and see if the stability has improved since.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-Id: <20200203090932.19147-16-alex.bennee@linaro.org>

.travis.yml: introduce TEST_BUILD_CMD and use it for check-tcg

At least for check-tcg we can split the build phase from the test
phase and do the former in parallel. While we are at it drop the V=1
for the check-tcg part as it just generates a lot more noise in the
logs.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-Id: <20200203090932.19147-15-alex.bennee@linaro.org>

tests/tcg: gate pauth-% tests on having compiler support

Otherwise we end up failing to build our tests on CI which may have
older compilers that the user expects. We can get rid of this once we
can fallback to multiarch containers.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-Id: <20200203090932.19147-14-alex.bennee@linaro.org>

tests/tcg: add a configure compiler check for ARMv8.1 and SVE

We will need this for some tests later. The docker images already
support it by default.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20200203090932.19147-13-alex.bennee@linaro.org>

.travis.yml: probe for number of available processors

The arm64 hardware was especially hit by only building on 3 of the 32
available cores. Introduce a JOBS environment variable which we use
for all parallel builds. We still run the main checks single threaded
though so to make it easier to spot hangs.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-Id: <20200203090932.19147-12-alex.bennee@linaro.org>

.travis.yml: move cache flushing to early common phase

We shall be adding more common early setup in a future commit.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-Id: <20200203090932.19147-11-alex.bennee@linaro.org>

.travis.yml: build documents under bionic

It looks like the xenial tooling doesn't like something in our setup.
We should probably be moving to bionic for everything soon
anyway (libssh aside).

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-Id: <20200203090932.19147-10-alex.bennee@linaro.org>

.travis.yml: Add description to each job

The NAME variable can be used to describe nicely a job (see [*]).
As we currently have 32 jobs, use it. This helps for quickly
finding a particular job.

  before: https://travis-ci.org/qemu/qemu/builds/639887646
  after: https://travis-ci.org/philmd/qemu/builds/641795043

[*] https://docs.travis-ci.com/user/customizing-the-build/#naming-jobs-within-matrices

Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-Id: <20200125183135.28317-1-f4bug@amsat.org>
Message-Id: <20200203090932.19147-9-alex.bennee@linaro.org>

.travis.yml: Drop superfluous use of --python=python3 parameter

As we require Python3 since commit ddf9069963, we don't need to
explicit it with the --python=/usr/bin/python3 configure option.

Reported-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-Id: <20200125184217.30034-1-f4bug@amsat.org>
Message-Id: <20200203090932.19147-8-alex.bennee@linaro.org>

.shippable: --disable-docs for cross-compile tests

The sphinx support is fairly new and we don't seem to have all the
bugs worked out for cross development environments right now.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-Id: <20200203090932.19147-6-alex.bennee@linaro.org>

travis.yml: Install genisoimage package

The genisoimage program is required for tests/cdrom-test
tests, otherwise they are skipped. The current Travis
environments do not provide it by default, so let's
explicitly require the genisoimage package.

Signed-off-by: Wainer dos Santos Moschetta <wainersm@redhat.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-Id: <20200110191254.11303-2-wainersm@redhat.com>
Message-Id: <20200203090932.19147-5-alex.bennee@linaro.org>

tests/docker: better handle symlinked libs

When we are copying we want to ensure we grab the first
resolution (the found in path section). However even that binary might
be a symlink so lets make sure we chase the symlinks to copy the right
binary to where it can be found.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Robert Foley <robert.foley@linaro.org>
Message-Id: <20200203090932.19147-3-alex.bennee@linaro.org>

tests/docker: move most cross compilers to buster base

This includes fixing up the dependencies (Which were already wrong for
one of the mips variants).

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-Id: <20200203090932.19147-2-alex.bennee@linaro.org>

Merge remote-tracking branch 'remotes/huth-gitlab/tags/pull-request-2020-02-04' into staging

* Current qtests queue
* Some Kconfig updates
* Some trivial clean-ups here and there

# gpg: Signature made Tue 04 Feb 2020 08:43:28 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-2020-02-04:
  configure: Fix typo of the have_afalg variable
  hw/hppa/Kconfig: LASI chipset requires PARALLEL port
  hw/input: Do not enable CONFIG_PCKBD by default
  Makefile: Do not use wildcard hw/*/Kconfig as input for minikconf
  hw/*/Makefile.objs: Move many .o files to common-objs
  trivial: Remove xenfb_enabled from sysemu.h
  include/sysemu/sysemu.h: Remove usused variable no_quit
  gitlab-ci: Refresh the list of iotests
  tests/qtest: update comments about bios-tables-test-allowed-diff.h
  boot-order-test: fix memleaks in boot-order-test
  tests/Makefile: Fix inclusion of the qos dependency files
  docs/devel: Fix qtest paths and info about check-block in testing.rst
  tests/vhost-user-bridge: Fix build
  test-logging: Fix -Werror=maybe-uninitialized warning

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

Merge remote-tracking branch 'remotes/amarkovic/tags/mips-queue-feb-04-2020' into staging

MIPS queue for February 4th, 2020

# gpg: Signature made Tue 04 Feb 2020 07:55:13 GMT
# gpg:                using RSA key D4972A8967F75A65
# gpg: Good signature from "Aleksandar Markovic <amarkovic@wavecomp.com>" [full]
# Primary key fingerprint: 8526 FBF1 5DA3 811F 4A01  DD75 D497 2A89 67F7 5A65

* remotes/amarkovic/tags/mips-queue-feb-04-2020:
  target/mips: Separate FPU-related helpers into their own file
  target/mips: Separate CP0-related helpers into their own file
  target/mips: Fix handling of LL/SC instructions after 7dd547e5ab

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

configure: Fix typo of the have_afalg variable

The variable is called 'have_afalg' and not 'hava_afalg'.

Fixes: f0d92b56d88 ('introduce some common functions for af_alg backend')
Message-Id: <20200203160046.23194-1-thuth@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Longpeng(Mike) <longpeng2@huawei.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>

hw/hppa/Kconfig: LASI chipset requires PARALLEL port

The PARISC Lasi chipset emulation requires some of the common parallel
support and fails to build on a --without-default-devices:

    LINK    hppa-softmmu/qemu-system-hppa
  /usr/bin/ld: hw/hppa/lasi.o: in function `lasi_init':
  hw/hppa/lasi.c:324: undefined reference to `parallel_mm_init'
  collect2: error: ld returned 1 exit status
  make[1]: *** [Makefile:206: qemu-system-hppa] Error 1

Fixes: 376b851909d
Reported-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-Id: <20200129192350.27143-1-philmd@redhat.com>
Acked-by: Helge Deller <deller@gmx.de>
Signed-off-by: Thomas Huth <thuth@redhat.com>

hw/input: Do not enable CONFIG_PCKBD by default

The i8042 PS/2 Controller should not be enabled by default. It has
to be selected by machines or chipsets (e.g. SuperIO chipsets).

Message-Id: <20200115113748.24757-1-thuth@redhat.com>
Acked-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Laurent Vivier <laurent@vivier.eu>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>

Makefile: Do not use wildcard hw/*/Kconfig as input for minikconf

The hw/*/Kconfig files should be sourced from hw/Kconfig, so there
is no need to pass them along as input files to minikconfig. We should
use the hw/*/Kconfig wildcard only for build dependencies in the Makefile.

With this change, there are now no duplicate entries in the generated
*-softmmu/config-devices.mak.d files anymore, and there is finally a
chance to get rid of stale Kconfig files like hw/bt/Kconfig, too (once
they do not show up in the config-devices.mak.d files now anymore).

Message-Id: <20200203153905.20544-1-thuth@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>

hw/*/Makefile.objs: Move many .o files to common-objs

We have many files that apparently do not depend on the target CPU
configuration, i.e. which can be put into common-obj-y instead of
obj-y. This way, the code can be shared for example between
qemu-system-arm and qemu-system-aarch64, or the various big and
little endian variants like qemu-system-sh4 and qemu-system-sh4eb,
so that we do not have to compile the code multiple times anymore.

Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20200130133841.10779-1-thuth@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>

trivial: Remove xenfb_enabled from sysemu.h

The define is only used in one other place. Move the code there
instead of keeping this xen-specific define in sysemu.h.

Message-Id: <20200121161747.10569-1-thuth@redhat.com>
Acked-by: Paul Durrant <paul@xen.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>

include/sysemu/sysemu.h: Remove usused variable no_quit

The no_quit variable has been removed in commit 78782712a62d56 ("vl: drop
no_quit variable"), so let's remove the extern declaration in the header
now, too.

Fixes: 78782712a62d ("vl: drop no_quit variable")
Message-Id: <20200108192402.19672-1-thuth@redhat.com>
Reviewed-by: Laurent Vivier <laurent@vivier.eu>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Thomas Huth <thuth@redhat.com>

gitlab-ci: Refresh the list of iotests

iotest 147 and 205 have recently been marked as "NBD-only", so they
are currently simply skipped and thus can be removed.

iotest 129 occasionally fails in the gitlab-CI, and according to Max,
there are some known issues with this test (see for example this URL:
https://lists.nongnu.org/archive/html/qemu-block/2019-06/msg00499.html ),
so for the time being, let's disable it until the problems are fixed.

The iotests 040, 127, 203 and 256 are scheduled to become part of "make
check-block", so we also do not have to test them seperately here anymore.

On the other side, new iotests have been added to the QEMU repository
in the past months, so we can now add some new test > 256 instead.

Message-Id: <20200121131936.8214-1-thuth@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>

tests/qtest: update comments about bios-tables-test-allowed-diff.h

Update comments in tests/qtest/bios-tables-test.c to reflect the
current path of bios-tables-test-allowed-diff.h, which is now under
tests/qtest/ as well.

Signed-off-by: Heyi Guo <guoheyi@huawei.com>
Message-Id: <20200202110009.51479-1-guoheyi@huawei.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Igor Mammedov <imammedo@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>

boot-order-test: fix memleaks in boot-order-test

It's not a big deal, but 'check qtest-ppc/ppc64' runs fail if sanitizers is enabled.
The memory leak stack is as follow:

Direct leak of 128 byte(s) in 4 object(s) allocated from:
    #0 0x7f11756f5970 in __interceptor_calloc (/lib64/libasan.so.5+0xef970)
    #1 0x7f1174f2549d in g_malloc0 (/lib64/libglib-2.0.so.0+0x5249d)
    #2 0x556af05aa7da in mm_fw_cfg_init /mnt/sdb/qemu/tests/libqos/fw_cfg.c:119
    #3 0x556af059f4f5 in read_boot_order_pmac /mnt/sdb/qemu/tests/boot-order-test.c:137
    #4 0x556af059efe2 in test_a_boot_order /mnt/sdb/qemu/tests/boot-order-test.c:47
    #5 0x556af059f2c0 in test_boot_orders /mnt/sdb/qemu/tests/boot-order-test.c:59
    #6 0x556af059f52d in test_pmac_oldworld_boot_order /mnt/sdb/qemu/tests/boot-order-test.c:152
    #7 0x7f1174f46cb9  (/lib64/libglib-2.0.so.0+0x73cb9)
    #8 0x7f1174f46b73  (/lib64/libglib-2.0.so.0+0x73b73)
    #9 0x7f1174f46b73  (/lib64/libglib-2.0.so.0+0x73b73)
    #10 0x7f1174f46f71 in g_test_run_suite (/lib64/libglib-2.0.so.0+0x73f71)
    #11 0x7f1174f46f94 in g_test_run (/lib64/libglib-2.0.so.0+0x73f94)

Reported-by: Euler Robot <euler.robot@huawei.com>
Signed-off-by: Pan Nengyuan <pannengyuan@huawei.com>
Message-Id: <20200203025935.36228-1-pannengyuan@huawei.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>

tests/Makefile: Fix inclusion of the qos dependency files

The qos dependency files can be found under tests/qtest/libqos and
not under tests/qtest/qos.

Fixes: 1cf4323ecd0 ("Move the libqos files under tests/qtest/")
Message-Id: <20200127140245.20065-1-thuth@redhat.com>
Reviewed-by: Laurent Vivier <lvivier@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>

docs/devel: Fix qtest paths and info about check-block in testing.rst

The qtests have recently been moved to a separate subdirectory, so
the paths that are mentioned in the documentation have to be adjusted
accordingly. And some of the iotests are now always run as part of
"make check", so this information has to be adjusted here, too.

Message-Id: <20200122134511.23806-1-thuth@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>

tests/vhost-user-bridge: Fix build

vhost-user-bridge isn't actually a test, it's just a helper
(that should probably move somewhere else) - but the build was
broken in the qtest move.

Fixes: 833884f37adc9f125fa2
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Message-Id: <20200117122648.137862-1-dgilbert@redhat.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>

test-logging: Fix -Werror=maybe-uninitialized warning

Checking for uninitialized variables raises warning for file path
variables in test_logfile_write and test_logfile_lock functions.

To suppress this warning, initialize varibles to NULL. This is safe
change as result of g_build_filename is stored to them before any usage.

Signed-off-by: Miroslav Rezanina <mrezanin@redhat.com>
Message-Id: <63b0fcedf7dfe799c8210b113e5dccf32414a89d.1579598240.git.mrezanin@redhat.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Robert Foley <robert.foley@linaro.org>
Signed-off-by: Thomas Huth <thuth@redhat.com>

target/mips: Separate FPU-related helpers into their own file

For clarity and easier maintenence, create target/mips/fpu_helper.c, and
move all FPU-related content form target/mips/op_helper.c to that file.

Signed-off-by: Aleksandar Markovic <amarkovic@wavecomp.com>
Reviewed-by: Aleksandar Rikalo <aleksandar.rikalo@rt-rk.com>
Message-Id: <1580745443-24650-3-git-send-email-aleksandar.markovic@rt-rk.com>