Pull request

v2:
 * Removed clang-format call from scripts/block-coroutine-wrapper.py. This
   avoids the issue with clang version incompatibility. It could be added back
   in the future but the code is readable without reformatting and it also
   makes the build less dependent on the environment.

# gpg: Signature made Mon 05 Oct 2020 16:42:28 BST
# gpg:                using RSA key 8695A8BFD3F97CDAAC35775A9CA4ABB381AB73C8
# gpg: Good signature from "Stefan Hajnoczi <stefanha@redhat.com>" [full]
# gpg:                 aka "Stefan Hajnoczi <stefanha@gmail.com>" [full]
# Primary key fingerprint: 8695 A8BF D3F9 7CDA AC35  775A 9CA4 ABB3 81AB 73C8

* remotes/stefanha-gitlab/tags/block-pull-request:
  util/vfio-helpers: Rework the IOVA allocator to avoid IOVA reserved regions
  util/vfio-helpers: Collect IOVA reserved regions
  docs: add 'io_uring' option to 'aio' param in qemu-options.hx
  include/block/block.h: drop non-ascii quotation mark
  block/io: refactor save/load vmstate
  block: drop bdrv_prwv
  block: generate coroutine-wrapper code
  scripts: add block-coroutine-wrapper.py
  block: declare some coroutine functions in block/coroutines.h
  block/io: refactor coroutine wrappers
  block: return error-code from bdrv_invalidate_cache
  block/nvme: Replace magic value by SCALE_MS definition
  block/nvme: Use register definitions from 'block/nvme.h'
  block/nvme: Drop NVMeRegs structure, directly use NvmeBar
  block/nvme: Reduce I/O registers scope
  block/nvme: Map doorbells pages write-only
  util/vfio-helpers: Pass page protections to qemu_vfio_pci_map_bar()

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

The readthedocs build is failing because we do not support Python 3.5 anymore.
Bump the Python version in the configuration.

Fixes: 1b11f28d ("configure: Bump the minimum required Python version to 3.6", 2020-10-02)
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-id: 20201005150122.446472-1-pbonzini@redhat.com
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

* move target configuration to default-configs/targets (myself)
* Memory failure event (Zhenwei)

# gpg: Signature made Mon 05 Oct 2020 08:14:29 BST
# gpg:                using RSA key F13338574B662389866C7682BFFBD25F78C7AE83
# gpg:                issuer "pbonzini@redhat.com"
# gpg: Good signature from "Paolo Bonzini <bonzini@gnu.org>" [full]
# gpg:                 aka "Paolo Bonzini <pbonzini@redhat.com>" [full]
# Primary key fingerprint: 46F5 9FBD 57D6 12E7 BFD4  E2F7 7E15 100C CD36 69B1
#      Subkey fingerprint: F133 3857 4B66 2389 866C  7682 BFFB D25F 78C7 AE83

* remotes/bonzini-gitlab/tags/for-upstream:
  dockerfiles: add diffutils to Fedora
  tests: tcg: do not use implicit rules
  target-i386: post memory failure event to QMP
  qapi/run-state.json: introduce memory failure event
  target-i386: seperate MCIP & MCE_MASK error reason
  meson: move sparse detection to Meson and rewrite check_sparse.py
  default-configs: remove redundant keys
  default-configs: use TARGET_ARCH key
  configure: move OpenBSD W^X test to meson
  default-configs: remove default-configs/devices for user-mode targets
  configure: remove target configuration
  configure: remove useless config-target.mak symbols
  configure: compute derivatives of target name in meson
  configure: remove dead variable
  configure: move accelerator logic to meson
  configure: rewrite accelerator defaults as tests
  configure: convert accelerator variables to meson options
  default-configs: move files to default-configs/devices/
  travis: remove TCI test

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

# Conflicts:
#	configure

Update capstone submodule from v3.0.5 to v5 ("next").
Convert submodule build to meson.
Enable capstone disassembly for s390x.
Code cleanups in disas.c

# gpg: Signature made Sat 03 Oct 2020 10:33:44 BST
# gpg:                using RSA key 7A481E78868B4DB6A85A05C064DF38E8AF7E215F
# gpg:                issuer "richard.henderson@linaro.org"
# gpg: Good signature from "Richard Henderson <richard.henderson@linaro.org>" [full]
# Primary key fingerprint: 7A48 1E78 868B 4DB6 A85A  05C0 64DF 38E8 AF7E 215F

* remotes/rth-gitlab/tags/pull-cap-20201003:
  disas/capstone: Add skipdata hook for s390x
  disas: Enable capstone disassembly for s390x
  disas: Split out capstone code to disas/capstone.c
  disas: Configure capstone for aarch64 host without libvixl
  disas: Cleanup plugin_disas
  disas: Use qemu/bswap.h for bfd endian loads
  disas: Clean up CPUDebug initialization
  disas: Move host asm annotations to tb_gen_code
  capstone: Require version 4.0 from a system library
  capstone: Update to upstream "next" branch
  capstone: Convert Makefile bits to meson bits

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

Introduce the qemu_vfio_find_fixed/temp_iova helpers which
respectively allocate IOVAs from the bottom/top parts of the
usable IOVA range, without picking within host IOVA reserved
windows. The allocation remains basic: if the size is too big
for the remaining of the current usable IOVA range, we jump
to the next one, leaving a hole in the address map.

Signed-off-by: Eric Auger <eric.auger@redhat.com>
Message-id: 20200929085550.30926-3-eric.auger@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>

The IOVA allocator currently ignores host reserved regions.
As a result some chosen IOVAs may collide with some of them,
resulting in VFIO MAP_DMA errors later on. This happens on ARM
where the MSI reserved window quickly is encountered:
[0x8000000, 0x8100000]. since 5.4 kernel, VFIO returns the usable
IOVA regions. So let's enumerate them in the prospect to avoid
them, later on.

Signed-off-by: Eric Auger <eric.auger@redhat.com>
Message-id: 20200929085550.30926-2-eric.auger@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>

When we added io_uring AIO engine, we forgot to update qemu-options.hx,
so qemu(1) man page and qemu help were outdated.

Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Julia Suvorova <jusual@redhat.com>
Reviewed-by: Pankaj Gupta <pankaj.gupta.linux@gmail.com>
Message-Id: <20200924151511.131471-1-sgarzare@redhat.com>

This is the only non-ascii character in the file and it doesn't really
needed here. Let's use normal "'" symbol for consistency with the rest
11 occurrences of "'" in the file.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>

Like for read/write in a previous commit, drop extra indirection layer,
generate directly bdrv_readv_vmstate() and bdrv_writev_vmstate().

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20200924185414.28642-8-vsementsov@virtuozzo.com>

Now that we are not maintaining boilerplate code for coroutine
wrappers, there is no more sense in keeping the extra indirection layer
of bdrv_prwv().  Let's drop it and instead generate pure bdrv_preadv()
and bdrv_pwritev().

Currently, bdrv_pwritev() and bdrv_preadv() are returning bytes on
success, auto generated functions will instead return zero, as their
_co_ prototype. Still, it's simple to make the conversion safe: the
only external user of bdrv_pwritev() is test-bdrv-drain, and it is
comfortable enough with bdrv_co_pwritev() instead. So prototypes are
moved to local block/coroutines.h. Next, the only internal use is
bdrv_pread() and bdrv_pwrite(), which are modified to return bytes on
success.

Of course, it would be great to convert bdrv_pread() and bdrv_pwrite()
to return 0 on success. But this requires audit (and probably
conversion) of all their users, let's leave it for another day
refactoring.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Signed-off-by: Stefan Hajnoczi <stefanha@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: <20200924185414.28642-7-vsementsov@virtuozzo.com>

Use code generation implemented in previous commit to generated
coroutine wrappers in block.c and block/io.c

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20200924185414.28642-6-vsementsov@virtuozzo.com>

We have a very frequent pattern of creating a coroutine from a function
with several arguments:

  - create a structure to pack parameters
  - create _entry function to call original function taking parameters
    from struct
  - do different magic to handle completion: set ret to NOT_DONE or
    EINPROGRESS or use separate bool field
  - fill the struct and create coroutine from _entry function with this
    struct as a parameter
  - do coroutine enter and BDRV_POLL_WHILE loop

Let's reduce code duplication by generating coroutine wrappers.

This patch adds scripts/block-coroutine-wrapper.py together with some
friends, which will generate functions with declared prototypes marked
by the 'generated_co_wrapper' specifier.

The usage of new code generation is as follows:

    1. define the coroutine function somewhere

        int coroutine_fn bdrv_co_NAME(...) {...}

    2. declare in some header file

        int generated_co_wrapper bdrv_NAME(...);

       with same list of parameters (generated_co_wrapper is
       defined in "include/block/block.h").

    3. Make sure the block_gen_c declaration in block/meson.build
       mentions the file with your marker function.

Still, no function is now marked, this work is for the following
commit.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Message-Id: <20200924185414.28642-5-vsementsov@virtuozzo.com>
[Added encoding='utf-8' to open() calls as requested by Vladimir. Fixed
typo and grammar issues pointed out by Eric Blake. Removed clang-format
dependency that caused build test issues.
--Stefan]
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>

We are going to keep coroutine-wrappers code (structure-packing
parameters, BDRV_POLL wrapper functions) in separate auto-generated
files. So, we'll need a header with declaration of original _co_
functions, for those which are static now. As well, we'll need
declarations for wrapper functions. Do these declarations now, as a
preparation step.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Signed-off-by: Stefan Hajnoczi <stefanha@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: <20200924185414.28642-4-vsementsov@virtuozzo.com>

Most of our coroutine wrappers already follow this convention:

We have 'coroutine_fn bdrv_co_<something>(<normal argument list>)' as
the core function, and a wrapper 'bdrv_<something>(<same argument
list>)' which does parameter packing and calls bdrv_run_co().

The only outsiders are the bdrv_prwv_co and
bdrv_common_block_status_above wrappers. Let's refactor them to behave
as the others, it simplifies further conversion of coroutine wrappers.

This patch adds an indirection layer, but it will be compensated by
a further commit, which will drop bdrv_co_prwv together with the
is_write logic, to keep the read and write paths separate.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Signed-off-by: Stefan Hajnoczi <stefanha@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: <20200924185414.28642-3-vsementsov@virtuozzo.com>

This is the only coroutine wrapper from block.c and block/io.c which
doesn't return a value, so let's convert it to the common behavior, to
simplify moving to generated coroutine wrappers in a further commit.

Also, bdrv_invalidate_cache is a void function, returning error only
through **errp parameter, which is considered to be bad practice, as
it forces callers to define and propagate local_err variable, so
conversion is good anyway.

This patch leaves the conversion of .bdrv_co_invalidate_cache() driver
callbacks and bdrv_invalidate_cache_all() for another day.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Signed-off-by: Stefan Hajnoczi <stefanha@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: <20200924185414.28642-2-vsementsov@virtuozzo.com>

Use self-explicit SCALE_MS definition instead of magic value
(missed in similar commit e4f310fe).

Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20200922083821.578519-7-philmd@redhat.com>

Use the NVMe register definitions from "block/nvme.h" which
ease a bit reviewing the code while matching the datasheet.

Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20200922083821.578519-6-philmd@redhat.com>

NVMeRegs only contains NvmeBar. Simplify the code by using NvmeBar
directly.

This triggers a checkpatch.pl error:

  ERROR: Use of volatile is usually wrong, please add a comment
  #30: FILE: block/nvme.c:691:
  +    volatile NvmeBar *regs;

This is a false positive as in our case we are using I/O registers,
so the 'volatile' use is justified.

Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20200922083821.578519-5-philmd@redhat.com>

We only access the I/O register in nvme_init().
Remove the reference in BDRVNVMeState and reduce its scope.

Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20200922083821.578519-4-philmd@redhat.com>

Per the datasheet sections 3.1.13/3.1.14:
  "The host should not read the doorbell registers."

As we don't need read access, map the doorbells with write-only
permission. We keep a reference to this mapped address in the
BDRVNVMeState structure.

Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20200922083821.578519-3-philmd@redhat.com>

Pages are currently mapped READ/WRITE. To be able to use different
protections, add a new argument to qemu_vfio_pci_map_bar().

Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20200922083821.578519-2-philmd@redhat.com>

For some reason diffutils is not included in the Fedora containers anymore,
causing the build to fail.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Neal Gompa <ngompa13@gmail.com>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-Id: <20201003085054.332992-2-pbonzini@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

Use pattern rules to clarify which targets are going to match the
rule and to provide clearer error messages.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

Post memory failure event through QMP to handle hardware memory corrupted
event. Rather than simply printing to the log, QEMU could report more
effective message to the client. For example, if a guest receives an MCE,
evacuating the host could be a good idea.

Signed-off-by: zhenwei pi <pizhenwei@bytedance.com>
Message-Id: <20200930100440.1060708-4-pizhenwei@bytedance.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

Introduce memory failure events for hypervisor and guest. This lets
mft: Need exactly one file argument.  Try `mft --help' for more
information.

Suggested by Peter Maydell, rename events name&description to make
them architecture-neutral; and suggested by Paolo, add more info to
distinguish a mce is AR/AO, and if a previous MCE was still being
processed in the guest.

Signed-off-by: zhenwei pi <pizhenwei@bytedance.com>
Message-Id: <20200930100440.1060708-3-pizhenwei@bytedance.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

Previously we would only get a simple string "Triple fault" in qemu
log. Add detailed message for the two reasons to describe why qemu
has to reset the guest.

Signed-off-by: zhenwei pi <pizhenwei@bytedance.com>
Message-Id: <20200930100440.1060708-2-pizhenwei@bytedance.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

Pass the path to the program to scripts/check_sparse.py, which
previously was not included in config-host.mak.  Change
scripts/check_sparse.py to work with cgcc, which seems to
work better with sparse 0.6.x.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

TARGET_BASE_ARCH and TARGET_ABI_DIR are 99% of the time the same
as TARGET_ARCH, remove them if so.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

Replace the individual TARGET_*=y lines with TARGET_ARCH,
similar to how TARGET_BASE_ARCH is handled already.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

The TCG option is now passed through to Meson.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

We no longer need dummy files to detect targets, since
default-configs/targets/ exists.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

The config-target.mak files are small constant, we can therefore just
write them down explicitly.

This removes a pretty large part of the configure script, including the
whole logic to detect which accelerators are supported by each target.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

Omit symbols that are not needed by softmmu or bsd-user targets,
in preparation for moving the generated config-target.mak files
into the source tree.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

Several CONFIG_* symbols in config-target.mak are easily computed from just
the target name.  We do not need them in config-target.mak, and can instead
place them in the config_target dictionary only.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

Move to meson the code to detect the presence of accelerators, and
to define accelerator-specific config-target.h symbols.

The logic for now is duplicated in configure because it is still
in use to build the list of targets (which is in turn used to
create the config-target.mak files).  The next patches remove it.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

Prepare to process "auto" in meson rather than configure: standardize the
shape of the code that changes "auto" to enabled/disabled, to ease the review
when it will be moved to meson.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

Prepare for moving the tests to meson.  For now they only have
enabled/disabled as the possible values when meson is invoked,
but "auto" will be a possibility later, when configure will only
parse the command line options.

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

Make room for target files in default-configs/targets/

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

TCI is already covered on gitlab CI, so we can remove it.

Cc: Thomas Huth <thuth@redhat.com>
Cc: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>