Alex Bennée [Fri, 14 Dec 2018 15:17:18 +0000 (15:17 +0000)]
.shippable.yml: disable the win cross tests
The pkg.mxe.cc package repositories have been down for the last two
weeks causing the builds to fail when shippable re-builds the
containers.
This is really just a sticking plaster until we can get our own docker
hub images properly setup so we can avoid having dependencies on
external repos.
Peter Maydell [Sun, 16 Dec 2018 16:32:42 +0000 (16:32 +0000)]
Merge remote-tracking branch 'remotes/pmaydell/tags/pull-misc-20181214' into staging
miscellaneous patches:
* checkpatch.pl: Enforce multiline comment syntax
* Rename cpu_physical_memory_write_rom() to address_space_write_rom()
* disas, monitor, elf_ops: Use address_space_read() to read memory
* Remove load_image() in favour of load_image_size()
* Fix some minor memory leaks in arm boards/devices
* virt: fix broken indentation
* remotes/pmaydell/tags/pull-misc-20181214: (22 commits)
virt: Fix broken indentation
target/arm: Create timers in realize, not init
tests/test-arm-mptimer: Don't leak string memory
hw/sd/sdhci: Don't leak memory region in sdhci_sysbus_realize()
hw/arm/mps2-tz.c: Free mscname string in make_dma()
target/arm: Free name string in ARMCPRegInfo hashtable entries
include/hw/loader.h: Document load_image_size()
hw/core/loader.c: Remove load_image()
device_tree.c: Don't use load_image()
hw/block/tc58128.c: Don't use load_image()
hw/i386/multiboot.c: Don't use load_image()
hw/i386/pc.c: Don't use load_image()
hw/pci/pci.c: Don't use load_image()
hw/smbios/smbios.c: Don't use load_image()
hw/ppc/ppc405_boards: Don't use load_image()
hw/ppc/mac_newworld, mac_oldworld: Don't use load_image()
elf_ops.h: Use address_space_write() to write memory
monitor: Use address_space_read() to read memory
disas.c: Use address_space_read() to read memory
Rename cpu_physical_memory_write_rom() to address_space_write_rom()
...
Peter Maydell [Sun, 16 Dec 2018 12:49:06 +0000 (12:49 +0000)]
Merge remote-tracking branch 'remotes/kevin/tags/for-upstream' into staging
Block layer patches:
- qcow2: Decompression worker threads
- dmg: lzfse compression support
- file-posix: Simplify delegation to worker thread
- Don't pass flags to bdrv_reopen_queue()
- iotests: make 235 work on s390 (and others)
# gpg: Signature made Fri 14 Dec 2018 10:55:09 GMT
# gpg: using RSA key 7F09B272C88F2FD6
# gpg: Good signature from "Kevin Wolf <[email protected]>"
# Primary key fingerprint: DC3D EB15 9A9A F95D 3D74 56FE 7F09 B272 C88F 2FD6
* remotes/kevin/tags/for-upstream: (42 commits)
block/mirror: add missing coroutine_fn annotations
iotests: make 235 work on s390 (and others)
block: Assert that flags are up-to-date in bdrv_reopen_prepare()
block: Remove assertions from update_flags_from_options()
block: Stop passing flags to bdrv_reopen_queue_child()
block: Remove flags parameter from bdrv_reopen_queue()
block: Clean up reopen_backing_file() in block/replication.c
qemu-io: Put flag changes in the options QDict in reopen_f()
block: Drop bdrv_reopen()
block: Use bdrv_reopen_set_read_only() in the mirror driver
block: Use bdrv_reopen_set_read_only() in external_snapshot_commit()
block: Use bdrv_reopen_set_read_only() in qmp_change_backing_file()
block: Use bdrv_reopen_set_read_only() in stream_start/complete()
block: Use bdrv_reopen_set_read_only() in bdrv_commit()
block: Use bdrv_reopen_set_read_only() in commit_start/complete()
block: Use bdrv_reopen_set_read_only() in bdrv_backing_update_filename()
block: Add bdrv_reopen_set_read_only()
file-posix: Avoid aio_worker() for QEMU_AIO_IOCTL
file-posix: Switch to .bdrv_co_ioctl
file-posix: Remove paio_submit_co()
...
* remotes/kraxel/tags/usb-20181214-pull-request:
usb-mtp: Limit filename to object information size
usb-mtp: use O_NOFOLLOW and O_CLOEXEC.
ehci: fix fetch qtd race
usb-host: reset and close libusb_device_handle before qemu exit
pvusb: set max grants only in initialise
Peter Maydell [Sat, 15 Dec 2018 21:19:06 +0000 (21:19 +0000)]
Merge remote-tracking branch 'remotes/armbru/tags/pull-qapi-2018-12-13-v2' into staging
QAPI patches for 2018-12-13
# gpg: Signature made Fri 14 Dec 2018 05:53:51 GMT
# gpg: using RSA key 3870B400EB918653
# gpg: Good signature from "Markus Armbruster <[email protected]>"
# gpg: aka "Markus Armbruster <[email protected]>"
# Primary key fingerprint: 354B C8B3 D7EB 2A6B 6867 4E5F 3870 B400 EB91 8653
* remotes/armbru/tags/pull-qapi-2018-12-13-v2: (32 commits)
qapi: add conditions to REPLICATION type/commands on the schema
qapi: add more conditions to SPICE
qapi: add condition to variants documentation
qapi: add 'If:' condition to struct members documentation
qapi: add 'If:' condition to enum values documentation
qapi: Add #if conditions to generated code members
qapi: add 'if' to alternate members
qapi: add 'if' to union members
qapi: Add 'if' to implicit struct members
qapi: add a dictionary form for TYPE
qapi-events: add 'if' condition to implicit event enum
qapi: add 'if' to enum members
qapi: add a dictionary form with 'name' key for enum members
qapi: improve reporting of unknown or missing keys
qapi: factor out checking for keys
tests: print enum type members more like object type members
qapi: change enum visitor and gen_enum* to take QAPISchemaMember
qapi: Do not define enumeration value explicitly
qapi: break long lines at 'data' member
qapi: rename QAPISchemaEnumType.values to .members
...
Peter Maydell [Fri, 14 Dec 2018 17:30:06 +0000 (17:30 +0000)]
Merge remote-tracking branch 'remotes/gkurz/tags/for-upstream' into staging
Most notable change in this PR is the full removal of the "handle" fsdev
backend.
# gpg: Signature made Wed 12 Dec 2018 13:20:42 GMT
# gpg: using RSA key 71D4D5E5822F73D6
# gpg: Good signature from "Greg Kurz <[email protected]>"
# gpg: aka "Gregory Kurz <[email protected]>"
# gpg: aka "[jpeg image of size 3330]"
# Primary key fingerprint: B482 8BAF 9431 40CE F2A3 4910 71D4 D5E5 822F 73D6
* remotes/gkurz/tags/for-upstream:
9p: remove support for the "handle" backend
xen/9pfs: use g_new(T, n) instead of g_malloc(sizeof(T) * n)
9p: use g_new(T, n) instead of g_malloc(sizeof(T) * n)
Peter Maydell [Fri, 14 Dec 2018 16:03:33 +0000 (16:03 +0000)]
Merge remote-tracking branch 'remotes/pmaydell/tags/pull-target-arm-20181213' into staging
target-arm queue:
* Convert various devices from sysbus init to instance_init
* Remove the now unused sysbus init support entirely
* Allow AArch64 processors to boot from a kernel placed over 4GB
* hw: arm: musicpal: drop TYPE_WM8750 in object_property_set_link()
* versal: minor fixes to virtio-mmio instantation
* arm: Implement the ARMv8.1-HPD extension
* arm: Implement the ARMv8.2-AA32HPD extension
* arm: Implement the ARMv8.1-LOR extension (as the trivial
"no limited ordering regions provided" minimum)
* remotes/pmaydell/tags/pull-target-arm-20181213: (37 commits)
target/arm: Implement the ARMv8.1-LOR extension
target/arm: Use arm_hcr_el2_eff more places
target/arm: Introduce arm_hcr_el2_eff
target/arm: Implement the ARMv8.2-AA32HPD extension
target/arm: Implement the ARMv8.1-HPD extension
target/arm: Tidy scr_write
target/arm: Fix HCR_EL2.TGE check in arm_phys_excp_target_el
target/arm: Add SCR_EL3 bits up to ARMv8.5
target/arm: Add HCR_EL2 bits up to ARMv8.5
target/arm: Move id_aa64mmfr* to ARMISARegisters
hw/arm: versal: Correct the nr of IRQs to 192
hw/arm: versal: Use IRQs 111 - 118 for virtio-mmio
hw/arm: versal: Reduce number of virtio-mmio instances
hw/arm: versal: Remove bogus virtio-mmio creation
core/sysbus: remove the SysBusDeviceClass::init path
xen_backend: remove xen_sysdev_init() function
usb/tusb6010: Convert sysbus init function to realize function
timer/puv3_ost: Convert sysbus init function to realize function
timer/grlib_gptimer: Convert sysbus init function to realize function
timer/etraxfs_timer: Convert sysbus init function to realize function
...
Eduardo Habkost [Fri, 14 Dec 2018 13:30:55 +0000 (13:30 +0000)]
virt: Fix broken indentation
I introduced indentation using tabs instead of spaces in another
commit. Peter reported the problem, and I failed to fix that
before sending my pull request.
Peter Maydell [Fri, 14 Dec 2018 13:30:55 +0000 (13:30 +0000)]
target/arm: Create timers in realize, not init
The timer_new() function allocates memory; this means that
if we call it in the CPU's init method we would need
to provide an instance_finalize method to free it. Defer
the timer creation to the realize function instead.
This fixes a memory leak spotted by clang LeakSanitizer
when a CPU object is created for introspection.
Peter Maydell [Fri, 14 Dec 2018 13:30:54 +0000 (13:30 +0000)]
tests/test-arm-mptimer: Don't leak string memory
The test-arm-mptimer setup creates a lot of test names using
g_strdup_printf() and never frees them. This is entirely
harmless since it's one-shot test code, but it clutters
up the output from clang's LeakSanitizer. Refactor to
use a helper function so we can free the memory.
Peter Maydell [Fri, 14 Dec 2018 13:30:54 +0000 (13:30 +0000)]
hw/sd/sdhci: Don't leak memory region in sdhci_sysbus_realize()
In sdhci_sysbus_realize() we override the initialization of
s->iomem that sdhci_common_realize() performs. However we
don't destroy the old memory region before reinitializing
it, which means that the memory allocated for mr->name in
memory_region_do_init() is leaked.
Since sdhci_initfn() already initializes s->io_ops to
&sdhci_mmio_ops, always use that in sdhci_common_realize()
and remove the now-unnecessary reinitialization of the
MMIO region from sdhci_sysbus_realize().
Peter Maydell [Fri, 14 Dec 2018 13:30:53 +0000 (13:30 +0000)]
target/arm: Free name string in ARMCPRegInfo hashtable entries
When we add a new entry to the ARMCPRegInfo hash table in
add_cpreg_to_hashtable(), we allocate memory for tehe
ARMCPRegInfo struct itself, and we also g_strdup() the
name string. So the hashtable's value destructor function
must free the name string as well as the struct.
Spotted by clang's leak sanitizer. The leak here is a
small one-off leak at startup, because we don't support
CPU hotplug, and so the only time when we destroy
hash table entries is for the case where ARM_CP_OVERRIDE
means we register a wildcard entry and then override it later.
Peter Maydell [Fri, 14 Dec 2018 13:30:52 +0000 (13:30 +0000)]
device_tree.c: Don't use load_image()
The load_image() function is deprecated, as it does not let the
caller specify how large the buffer to read the file into is.
Instead use load_image_size().
Peter Maydell [Fri, 14 Dec 2018 13:30:52 +0000 (13:30 +0000)]
hw/block/tc58128.c: Don't use load_image()
The load_image() function is deprecated, as it does not let the
caller specify how large the buffer to read the file into is.
Instead use load_image_size().
Peter Maydell [Fri, 14 Dec 2018 13:30:51 +0000 (13:30 +0000)]
hw/i386/multiboot.c: Don't use load_image()
The load_image() function is deprecated, as it does not let the
caller specify how large the buffer to read the file into is.
Instead use load_image_size().
While we are converting the code, add the missing error check.
Peter Maydell [Fri, 14 Dec 2018 13:30:51 +0000 (13:30 +0000)]
hw/i386/pc.c: Don't use load_image()
The load_image() function is deprecated, as it does not let the
caller specify how large the buffer to read the file into is.
Use the glib g_file_get_contents() function instead, which does
the whole "allocate memory for the file and read it in" operation.
Peter Maydell [Fri, 14 Dec 2018 13:30:51 +0000 (13:30 +0000)]
hw/pci/pci.c: Don't use load_image()
The load_image() function is deprecated, as it does not let the
caller specify how large the buffer to read the file into is.
Instead use load_image_size().
While we are converting this code, add an error-check
for read failure.
Peter Maydell [Fri, 14 Dec 2018 13:30:50 +0000 (13:30 +0000)]
hw/smbios/smbios.c: Don't use load_image()
The load_image() function is deprecated, as it does not let the
caller specify how large the buffer to read the file into is.
Instead use load_image_size().
Peter Maydell [Fri, 14 Dec 2018 13:30:50 +0000 (13:30 +0000)]
hw/ppc/ppc405_boards: Don't use load_image()
The load_image() function is deprecated, as it does not let the
caller specify how large the buffer to read the file into is.
Instead use load_image_size().
Peter Maydell [Fri, 14 Dec 2018 13:30:50 +0000 (13:30 +0000)]
hw/ppc/mac_newworld, mac_oldworld: Don't use load_image()
The load_image() function is deprecated, as it does not let the
caller specify how large the buffer to read the file into is.
Use the glib g_file_get_contents() function instead, which does
the whole "allocate memory for the file and read it in" operation.
Peter Maydell [Fri, 14 Dec 2018 13:30:49 +0000 (13:30 +0000)]
elf_ops.h: Use address_space_write() to write memory
Currently the load_elf function in elf_ops.h uses
cpu_physical_memory_write() to write the ELF file to
memory if it is not handling it as a ROM blob. This
means we ignore the AddressSpace that the function
is passed to define where it should be loaded.
Use address_space_write() instead.
Peter Maydell [Fri, 14 Dec 2018 13:30:49 +0000 (13:30 +0000)]
monitor: Use address_space_read() to read memory
Currently monitor.c reads physical memory using
cpu_physical_memory_read(). This effectively hard-codes
assuming that all CPUs have the same view of physical
memory. Switch to address_space_read() instead, which
lets us use the AddressSpace for the CPU we're
reading memory for (falling back to address_space_memory
if there is no CPU, as happens with the "none" board).
As a bonus, this allows us to detect failures to read memory.
Peter Maydell [Fri, 14 Dec 2018 13:30:49 +0000 (13:30 +0000)]
disas.c: Use address_space_read() to read memory
Currently disas.c reads physical memory using
cpu_physical_memory_read(). This effectively hard-codes
assuming that all CPUs have the same view of physical
memory. Switch to address_space_read() instead, which
lets us use the AddressSpace for the CPU we're
disassembling for.
Peter Maydell [Fri, 14 Dec 2018 13:30:48 +0000 (13:30 +0000)]
Rename cpu_physical_memory_write_rom() to address_space_write_rom()
The API of cpu_physical_memory_write_rom() is odd, because it
takes an AddressSpace, unlike all the other cpu_physical_memory_*
access functions. Rename it to address_space_write_rom(), and
bring its API into line with address_space_write().
Rename cpu_physical_memory_write_rom_internal() to
address_space_write_rom_internal(), and make it take
MemTxAttrs and return a MemTxResult. This brings its
API into line with address_space_write().
This is an internal function to exec.c; fixing its API
will allow us to change the global function
cpu_physical_memory_write_rom().
We now require Linux-kernel-style multiline comments:
/*
* line one
* line two
*/
Enforce this in checkpatch.pl, by backporting the relevant
parts of the Linux kernel's checkpatch.pl. (The only changes
needed are that Linux's checkpatch.pl WARN() function takes
an extra argument that ours does not, and the kernel has a
special case for networking code we don't want.)"
The kernel's checkpatch does not enforce "leading /* on
a line of its own, so that part is unique to QEMU's checkpatch.
Sample warning output:
WARNING: Block comments use a leading /* on a separate line
#34: FILE: hw/intc/arm_gicv3_common.c:39:
+ /* Older versions of QEMU had a bug in the handling of state save/restore
Marking a function coroutine_fn currently has no effect on the compiler,
but it documents that this function must be called from coroutine
context and it may yield. This is important information for the
programmer.
Also, if we ever transition to a stackless coroutine implementation,
then it's likely that the annotation will become mandatory so the
compiler can use the correct calling convention for coroutine functions.
Alberto Garcia [Mon, 12 Nov 2018 14:00:47 +0000 (16:00 +0200)]
block: Assert that flags are up-to-date in bdrv_reopen_prepare()
Towards the end of bdrv_reopen_queue_child(), before starting to
process the children, the update_flags_from_options() function is
called in order to have BDRVReopenState.flags in sync with the options
from the QDict.
This is necessary because during the reopen process flags must be
updated for all nodes in the queue so bdrv_is_writable_after_reopen()
and the permission checks work correctly.
Because of that, calling update_flags_from_options() again in
bdrv_reopen_prepare() doesn't really change the flags (they are
already up-to-date). But we need to call it in order to remove those
options from QemuOpts and that way indicate that they have been
processed.
Alberto Garcia [Mon, 12 Nov 2018 14:00:46 +0000 (16:00 +0200)]
block: Remove assertions from update_flags_from_options()
This function takes four options (cache.direct, cache.no-flush,
read-only and auto-read-only) from a QemuOpts object and updates the
flags accordingly.
If any of those options is not set (because it was missing from the
original QDict or because it had an invalid value) then the function
aborts with a failed assertion:
This assertion is unnecessary, and it forces any caller of
bdrv_reopen() to pass all the aforementioned four options. This may
have made sense in order to remove ambiguity when bdrv_reopen() was
taking both flags and options, but that's not the case anymore.
It's also unnecessary if we want to validate the option values,
because bdrv_reopen_prepare() already takes care of that, as we can
see if we remove the assertions:
Alberto Garcia [Mon, 12 Nov 2018 14:00:45 +0000 (16:00 +0200)]
block: Stop passing flags to bdrv_reopen_queue_child()
Now that all callers are passing the new options using the QDict we no
longer need the 'flags' parameter.
This patch makes the following changes:
1) The update_options_from_flags() call is no longer necessary
so it can be removed.
2) The update_flags_from_options() call is now used in all cases,
and is moved down a few lines so it happens after the options
QDict contains the final set of values.
3) The flags parameter is removed. Now the flags are initialized
using the current value (for the top-level node) or the parent
flags (after inherit_options()). In both cases the initial
values are updated to reflect the new options in the QDict. This
happens in bdrv_reopen_queue_child() (as explained above) and in
bdrv_reopen_prepare().
Alberto Garcia [Mon, 12 Nov 2018 14:00:43 +0000 (16:00 +0200)]
block: Clean up reopen_backing_file() in block/replication.c
This function is used to put the hidden and secondary disks in
read-write mode before launching the backup job, and back in read-only
mode afterwards.
This patch does the following changes:
- Use an options QDict with the "read-only" option instead of
passing the changes as flags only.
- Simplify the code (it was unnecessarily complicated and verbose).
- Fix a bug due to which the secondary disk was not being put back
in read-only mode when writable=false (because in this case
orig_secondary_flags always had the BDRV_O_RDWR flag set).
- Stop clearing the BDRV_O_INACTIVE flag.
The flags parameter to bdrv_reopen_queue() becomes redundant and we'll
be able to get rid of it in a subsequent patch.
Alberto Garcia [Mon, 12 Nov 2018 14:00:42 +0000 (16:00 +0200)]
qemu-io: Put flag changes in the options QDict in reopen_f()
When reopen_f() puts a block device in the reopen queue, some of the
new options are passed using a QDict, but others ("read-only" and the
cache options) are passed as flags.
This patch puts those flags in the QDict. This way the flags parameter
becomes redundant and we'll be able to get rid of it in a subsequent
patch.
Alberto Garcia [Mon, 12 Nov 2018 14:00:40 +0000 (16:00 +0200)]
block: Use bdrv_reopen_set_read_only() in the mirror driver
The 'block-commit' QMP command is implemented internally using two
different drivers. If the source image is the active layer then the
mirror driver is used (commit_active_start()), otherwise the commit
driver is used (commit_start()).
In both cases the destination image must be put temporarily in
read-write mode. This is done correctly in the latter case, but what
commit_active_start() does is copy all flags instead.
This patch replaces the bdrv_reopen() calls in that function with
bdrv_reopen_set_read_only() so that only the read-only status is
changed.
A similar change is made in mirror_exit(), which is also used by the
'drive-mirror' and 'blockdev-mirror' commands.
Alberto Garcia [Mon, 12 Nov 2018 14:00:33 +0000 (16:00 +0200)]
block: Add bdrv_reopen_set_read_only()
Most callers of bdrv_reopen() only use it to switch a BlockDriverState
between read-only and read-write, so this patch adds a new function
that does just that.
We also want to get rid of the flags parameter in the bdrv_reopen()
API, so this function sets the "read-only" option and passes the
original flags (which will then be updated in bdrv_reopen_prepare()).
Kevin Wolf [Wed, 31 Oct 2018 10:30:42 +0000 (11:30 +0100)]
file-posix: Avoid aio_worker() for QEMU_AIO_IOCTL
aio_worker() doesn't add anything interesting, it's only a useless
indirection. Call the handler function directly instead.
As we know that this handler function is only called from coroutine
context and the coroutine stays around until the worker thread finishes,
we can keep RawPosixAIOData on the stack.
This was the last user of aio_worker(), so the function goes away now.
Kevin Wolf [Wed, 31 Oct 2018 10:25:18 +0000 (11:25 +0100)]
file-posix: Switch to .bdrv_co_ioctl
No real reason to keep using the callback based mechanism here when the
rest of the file-posix driver is coroutine based. Changing it brings
ioctls more in line with how other request types work.
Kevin Wolf [Thu, 25 Oct 2018 13:18:58 +0000 (14:18 +0100)]
file-posix: Avoid aio_worker() for QEMU_AIO_READ/WRITE
aio_worker() doesn't add anything interesting, it's only a useless
indirection. Call the handler function directly instead.
As we know that this handler function is only called from coroutine
context and the coroutine stays around until the worker thread finishes,
we can keep RawPosixAIOData on the stack.
Kevin Wolf [Wed, 31 Oct 2018 08:43:17 +0000 (09:43 +0100)]
file-posix: Move read/write operation logic out of aio_worker()
aio_worker() for reads and writes isn't boring enough yet. It still does
some postprocessing for handling short reads and turning the result into
the right return value.
However, there is no reason why handle_aiocb_rw() couldn't do the same,
and even without duplicating code between the read and write path. So
move the code there.
Kevin Wolf [Thu, 25 Oct 2018 13:18:58 +0000 (14:18 +0100)]
file-posix: Avoid aio_worker() for QEMU_AIO_FLUSH
aio_worker() doesn't add anything interesting, it's only a useless
indirection. Call the handler function directly instead.
As we know that this handler function is only called from coroutine
context and the coroutine stays around until the worker thread finishes,
we can keep RawPosixAIOData on the stack.
Kevin Wolf [Thu, 25 Oct 2018 13:18:58 +0000 (14:18 +0100)]
file-posix: Avoid aio_worker() for QEMU_AIO_DISCARD
aio_worker() doesn't add anything interesting, it's only a useless
indirection. Call the handler function directly instead.
As we know that this handler function is only called from coroutine
context and the coroutine stays around until the worker thread finishes,
we can keep RawPosixAIOData on the stack.
Kevin Wolf [Thu, 25 Oct 2018 13:18:58 +0000 (14:18 +0100)]
file-posix: Avoid aio_worker() for QEMU_AIO_WRITE_ZEROES
aio_worker() doesn't add anything interesting, it's only a useless
indirection. Call the handler function directly instead.
As we know that this handler function is only called from coroutine
context and the coroutine stays around until the worker thread finishes,
we can keep RawPosixAIOData on the stack.
Kevin Wolf [Thu, 25 Oct 2018 13:18:58 +0000 (14:18 +0100)]
file-posix: Avoid aio_worker() for QEMU_AIO_COPY_RANGE
aio_worker() doesn't add anything interesting, it's only a useless
indirection. Call the handler function directly instead.
As we know that this handler function is only called from coroutine
context and the coroutine stays around until the worker thread finishes,
we can keep RawPosixAIOData on the stack.
Kevin Wolf [Thu, 25 Oct 2018 13:18:58 +0000 (14:18 +0100)]
file-posix: Avoid aio_worker() for QEMU_AIO_TRUNCATE
aio_worker() doesn't add anything interesting, it's only a useless
indirection. Call the handler function directly instead.
As we know that this handler function is only called from coroutine
context and the coroutine stays around until the worker thread finishes,
we can keep RawPosixAIOData on the stack.
Kevin Wolf [Fri, 26 Oct 2018 15:53:47 +0000 (16:53 +0100)]
file-posix: Factor out raw_thread_pool_submit()
Getting the thread pool of the AioContext of a block node and scheduling
some work in it is an operation that is already done twice, and we'll
get more instances. Factor it out into a separate function.
Kevin Wolf [Thu, 25 Oct 2018 15:21:14 +0000 (16:21 +0100)]
file-posix: Reorganise RawPosixAIOData
RawPosixAIOData contains a lot of fields for several separate operations
that are to be processed in a worker thread and that need different
parameters. The struct is currently rather unorganised, with unions that
cover some, but not all operations, and even one #define for field names
instead of a union.
Clean this up to have some common fields and a single union. As a side
effect, on x86_64 the struct shrinks from 72 to 48 bytes.
Allocate buffers locally and release qcow2 lock. Than, reads inside
qcow2_co_preadv_compressed may be done in parallel, however all
decompression is still done synchronously. Let's improve it in the
following commit.
- make it look more like a pair of qcow2_compress - rename the function
and its parameters
- drop extra out_len variable, check filling of output buffer by strm
structure itself
- fix code style
- add some documentation
qcow2: make more generic interface for qcow2_compress
Give explicit size both for source and destination buffers, to make it
similar with decompression path and than cleanly reuse parameter
structure for decompression threads.
After commit f8d59dfb40
"block/backup: fix fleecing scheme: use serialized writes" fleecing
(specifically reading from backup target, when backup source is in
backing chain of backup target) is safe, because all backup-job writes
to target are serialized. Therefore we don't need additional
synchronization for these reads.
Julio Faracco [Mon, 5 Nov 2018 15:08:06 +0000 (13:08 -0200)]
dmg: exchanging hardcoded dmg UDIF block types to enum.
This change is better to understand what kind of block type is being
handled by the code. Using a syntax similar to the DMG documentation is
easier than tracking all hex values assigned to a block type.
Julio Faracco [Mon, 5 Nov 2018 15:08:04 +0000 (13:08 -0200)]
configure: adding support to lzfse library.
This commit includes the support to lzfse opensource library. With this
library dmg block driver can decompress images with this type of
compression inside.
Julio Faracco [Mon, 5 Nov 2018 15:08:03 +0000 (13:08 -0200)]
block: adding lzfse decompressing support as a module.
QEMU dmg support includes zlib and bzip2, but it does not contains lzfse
support. This commit adds the source file to extend compression support
for new DMGs.
Peter Maydell [Fri, 14 Dec 2018 10:19:47 +0000 (10:19 +0000)]
Merge remote-tracking branch 'remotes/huth-gitlab/tags/pull-request-2018-12-12' into staging
- Explicitly check for minimum compiler versions
- Remove obsolete code for old compilers that is now not required anymore
- Fix a duplicated typedef for Clang 3.4
* remotes/huth-gitlab/tags/pull-request-2018-12-12:
i2c: Move typedef of bitbang_i2c_interface to i2c.h
Remove QEMU_ARTIFICIAL macro
includes: Replace QEMU_GNUC_PREREQ with "__has_builtin || !defined(__clang__)"
audio/alsaaudio: Remove compiler check around pragma
tcg/tcg.h: Remove GCC check for tcg_debug_assert() macro
configure: Remove old -fno-gcse workaround for GCC 4.6.x and 4.7.[012]
configure: Remove obsolete check for Clang < 3.2
configure: Add a test for the minimum compiler version
usb-mtp: Limit filename to object information size
The filename length in MTP metadata is specified by the guest. By
trusting it directly it'd theoretically be possible to get the host to
write memory parts outside the filename buffer into a filename. In
practice though there are usually NUL bytes stopping the string
operations.
Also use the opportunity to not assign the filename member twice.
Gerd Hoffmann [Thu, 13 Dec 2018 12:25:11 +0000 (13:25 +0100)]
usb-mtp: use O_NOFOLLOW and O_CLOEXEC.
Open files and directories with O_NOFOLLOW to avoid symlinks attacks.
While being at it also add O_CLOEXEC.
usb-mtp only handles regular files and directories and ignores
everything else, so users should not see a difference.
Because qemu ignores symlinks, carrying out a successful symlink attack
requires swapping an existing file or directory below rootdir for a
symlink and winning the race against the inotify notification to qemu.
Before the patch, we first register the commands unconditionally in
generated code (requires a stub), then conditionally unregister in
qmp_unregister_commands_hack().
Afterwards, we register only when CONFIG_REPLICATION. The command
fails exactly the same, with CommandNotFound.
Improvement, because now query-qmp-schema is accurate, and we're one
step closer to killing qmp_unregister_commands_hack().
* enum BlockdevDriver value "replication" in command blockdev-add
qapi: add 'If:' condition to enum values documentation
Use a common function to generate the "If:..." line.
While at it, get rid of the existing \n\n (no idea why it was
there). Use a line-break in member description, this seems to look
slightly better in the plaintext version.
qapi: Add #if conditions to generated code members
Wrap generated enum and struct members and their supporting code with
#if/#endif, using the .ifcond members added in the previous patches.
We do enum and struct in a single patch because union tag enum and the
associated variants tie them together, and dealing with that to split
the patch doesn't seem worthwhile.
qapi-events: add 'if' condition to implicit event enum
Add condition to QAPIEvent enum members based on the event 'if'.
The generated code remains unconditional for now. Later patches
generate the conditionals (also there is no additional coverage of
this change in qapi-schema-test.out since the event_names enum is an
implicit type created by qapi/events.py).
QAPISchemaMember gains .ifcond for enum members: inherited classes,
such as QAPISchemaObjectTypeMember, will thus have an ifcond member
after this (those different types will also use the .ifcond to store
the condition and generate conditional code in the following patches).
The generated code remains unconditional for now. Later patches
generate the conditionals.
Introduce a new helper function to check if the given keys are known,
and if mandatory keys are present. The function will be reused in
other places in the following code changes.
tests: print enum type members more like object type members
Commit 93bda4dd461 changed the internal representation of enum type
members from str to QAPISchemaMember, but we still print only a
string. Has been good enough, as the name is the member's only
attribute of interest, but that's about to change. To prepare, print
them more like object type members.
qapi: change enum visitor and gen_enum* to take QAPISchemaMember
This will allow to add and access more properties associated with enum
values/members, like the associated 'if' condition. We may want to
have a specialized type QAPISchemaEnumMember, for now this will do.
Modify gen_enum() and gen_enum_lookup() for the same reason.
The generated C enumeration types explicitly set the enumeration
constants to 0, 1, 2, ... That's exactly what you get when you don't
supply values.
Drop the explicit values. No change now, but it will avoid gaps in
the values when we later add support for 'if' conditions. Avoiding
such gaps will save us the trouble of changing the ENUM_lookup[]
tables to work without a sentinel.
We'll have to take care to ensure the headers required by the 'if'
conditions get always included before the generated QAPI code.
Fortunately, our convention to include "qemu/osdep.h" first in any .c
ensures that's the case for our CONFIG_FOO macros.
Let's break the line before 'data'. While at it, improve a bit
indentation/spacing. (I removed some alignment which are not helping
much readability and become quickly inconsistent)
qapi: rename QAPISchemaEnumType.values to .members
Rename QAPISchemaEnumType.values and related variables to members.
Makes sense ever since commit 93bda4dd4 changed .values from list of
string to list of QAPISchemaMember. Obvious no-op.
projects / qemu.git / log