block.h currently contains a mix of functions:
some of them run under the BQL and modify the block layer graph,
others are instead thread-safe and perform I/O in iothreads.
Some others can only be called by either the main loop or the
iothread running the AioContext (and not other iothreads),
and using them in another thread would cause deadlocks, and therefore
it is not ideal to define them as I/O.
It is not easy to understand which function is part of which
group (I/O vs GS vs "I/O or GS"), and this patch aims to clarify it.
The "GS" functions need the BQL, and often use
aio_context_acquire/release and/or drain to be sure they
can modify the graph safely.
The I/O function are instead thread safe, and can run in
any AioContext.
"I/O or GS" functions run instead in the main loop or in
a single iothread, and use BDRV_POLL_WHILE().
By splitting the header in two files, block-io.h
and block-global-state.h we have a clearer view on what
needs what kind of protection. block-common.h
contains common structures shared by both headers.
block.h is left there for legacy and to avoid changing
all includes in all c files that use the block APIs.
Assertions are added in the next patch.
Signed-off-by: Emanuele Giuseppe Esposito <eesposit@redhat.com>
Message-Id: <20220303151616.325444-4-eesposit@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Righ now, IO_CODE and IO_OR_GS_CODE are nop, as there isn't
really a way to check that a function is only called in I/O.
On the other side, we can use qemu_in_main_thread() to check if
we are in the main loop.
The usage of macros makes easy to extend them in the future without
making changes in all callers. They will also visually help understanding
in which category each function is, without looking at the header.
Signed-off-by: Emanuele Giuseppe Esposito <eesposit@redhat.com>
Message-Id: <20220303151616.325444-3-eesposit@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
When invoked from the main loop, this function is the same
as qemu_mutex_iothread_locked, and returns true if the BQL is held.
When invoked from iothreads or tests, it returns true only
if the current AioContext is the Main Loop.
This essentially just extends qemu_mutex_iothread_locked to work
also in unit tests or other users like storage-daemon, that run
in the Main Loop but end up using the implementation in
stubs/iothread-lock.c.
Using qemu_mutex_iothread_locked in unit tests defaults to false
because they use the implementation in stubs/iothread-lock,
making all assertions added in next patches fail despite the
AioContext is still the main loop.
See the comment in the function header for more information.
Signed-off-by: Emanuele Giuseppe Esposito <eesposit@redhat.com>
Message-Id: <20220303151616.325444-2-eesposit@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
185 tests quitting qemu while a block job is active. It does not
specifically test quitting qemu while a mirror or active commit job is
in its READY phase.
Add two test cases for this, where we respectively mirror or commit to
an external QSD instance, which provides a throttled block device. qemu
is supposed to cancel the job so that it can quit as soon as possible
instead of waiting for the job to complete (which it did before 6.2).
Signed-off-by: Hanna Reitz <hreitz@redhat.com>
Message-Id: <20220303164814.284974-5-hreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
To implement this, we reuse the existing daemonizing functions from the
system emulator, which mainly do the following:
- Fork off a child process, and set up a pipe between parent and child
- The parent process waits until the child sends a status byte over the
pipe (0 means that the child was set up successfully; anything else
(including errors or EOF) means that the child was not set up
successfully), and then exits with an appropriate exit status
- The child process enters a new session (forking off again), changes
the umask, and will ignore terminal signals from then on
- Once set-up is complete, the child will chdir to /, redirect all
standard I/O streams to /dev/null, and tell the parent that set-up has
been completed successfully
In contrast to qemu-nbd's --fork implementation, during the set up
phase, error messages are not piped through the parent process.
qemu-nbd mainly does this to detect errors, though (while os_daemonize()
has the child explicitly signal success after set up); because we do not
redirect stderr after forking, error messages continue to appear on
whatever the parent's stderr was (until set up is complete).
Signed-off-by: Hanna Reitz <hreitz@redhat.com>
Message-Id: <20220303164814.284974-4-hreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
In contrast to qemu-nbd (where it is called --fork) and the system
emulator, QSD does not have a --daemonize switch yet. Just like them,
QSD allows setting up block devices and exports on the command line.
When doing so, it is often necessary for whoever invoked the QSD to wait
until these exports are fully set up. A --daemonize switch allows
precisely this, by virtue of the parent process exiting once everything
is set up.
Note that there are alternative ways of waiting for all exports to be
set up, for example:
- Passing the --pidfile option and waiting until the respective file
exists (but I do not know if there is a way of implementing this
without a busy wait loop)
- Set up some network server (e.g. on a Unix socket) and have the QSD
connect to it after all arguments have been processed by appending
corresponding --chardev and --monitor options to the command line,
and then wait until the QSD connects
Having a --daemonize option would make this simpler, though, without
having to rely on additional tools (to set up a network server) or busy
waiting.
Implementing a --daemonize switch means having to fork the QSD process.
Ideally, we should do this as early as possible: All the parent process
has to do is to wait for the child process to signal completion of its
set-up phase, and therefore there is basically no initialization that
needs to be done before the fork. On the other hand, forking after
initialization steps means having to consider how those steps (like
setting up the block layer or QMP) interact with a later fork, which is
often not trivial.
In order to fork this early, we must scan the command line for
--daemonize long before our current process_options() call. Instead of
adding custom new code to do so, just reuse process_options() and give
it a @pre_init_pass argument to distinguish the two passes. I believe
there are some other switches but --daemonize that deserve parsing in
the first pass:
- --help and --version are supposed to only print some text and then
immediately exit (so any initialization we do would be for naught).
This changes behavior, because now "--blockdev inv-drv --help" will
print a help text instead of complaining about the --blockdev
argument.
Note that this is similar in behavior to other tools, though: "--help"
is generally immediately acted upon when finding it in the argument
list, potentially before other arguments (even ones before it) are
acted on. For example, "ls /does-not-exist --help" prints a help text
and does not complain about ENOENT.
- --pidfile does not need initialization, and is already exempted from
the sequential order that process_options() claims to strictly follow
(the PID file is only created after all arguments are processed, not
at the time the --pidfile argument appears), so it makes sense to
include it in the same category as --daemonize.
- Invalid arguments should always be reported as soon as possible. (The
same caveat with --help applies: That means that "--blockdev inv-drv
--inv-arg" will now complain about --inv-arg, not inv-drv.)
This patch does make some references to --daemonize without having
implemented it yet, but that will happen in the next patch.
Signed-off-by: Hanna Reitz <hreitz@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20220303164814.284974-3-hreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
The daemonizing functions in os-posix (os_daemonize() and
os_setup_post()) only daemonize the process if the static `daemonize`
variable is set. Right now, it can only be set by os_parse_cmd_args().
In order to use os_daemonize() and os_setup_post() from the storage
daemon to have it be daemonized, we need some other way to set this
`daemonize` variable, because I would rather not tap into the system
emulator's arg-parsing code. Therefore, this patch adds an
os_set_daemonize() function, which will return an error on os-win32
(because daemonizing is not supported there).
Signed-off-by: Hanna Reitz <hreitz@redhat.com>
Message-Id: <20220303164814.284974-2-hreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
qemu_mutex_iothread_locked() may be used from coroutines. Standard
__thread variables cannot be used by coroutines. Use the coroutine TLS
macros instead.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20220222140150.27240-5-stefanha@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
RCU may be used from coroutines. Standard __thread variables cannot be
used by coroutines. Use the coroutine TLS macros instead.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20220222140150.27240-4-stefanha@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
QEMU TLS macros must be used to make TLS variables safe with coroutines.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20220222140150.27240-3-stefanha@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Compiler optimizations can cache TLS values across coroutine yield
points, resulting in stale values from the previous thread when a
coroutine is re-entered by a new thread.
Serge Guelton developed an __attribute__((noinline)) wrapper and tested
it with clang and gcc. I formatted his idea according to QEMU's coding
style and wrote documentation.
The compiler can still optimize based on analyzing noinline code, so an
asm volatile barrier with an output constraint is required to prevent
unwanted optimizations.
Buglink: https://bugzilla.redhat.com/show_bug.cgi?id=1952483
Suggested-by: Serge Guelton <sguelton@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20220222140150.27240-2-stefanha@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Split bdrv_co_invalidate cache in two: the Global State (under BQL)
code that takes care of permissions and running GS callbacks,
and leave only the I/O code (->bdrv_co_invalidate_cache) running in
the I/O coroutine.
The only side effect is that bdrv_co_invalidate_cache is not
recursive anymore, and so is every direct call to
bdrv_invalidate_cache().
Signed-off-by: Emanuele Giuseppe Esposito <eesposit@redhat.com>
Message-Id: <20220209105452.1694545-6-eesposit@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Following the bdrv_activate renaming, change also the name
of the respective callers.
bdrv_invalidate_cache_all -> bdrv_activate_all
blk_invalidate_cache -> blk_activate
test_sync_op_invalidate_cache -> test_sync_op_activate
No functional change intended.
Signed-off-by: Emanuele Giuseppe Esposito <eesposit@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Hanna Reitz <hreitz@redhat.com>
Message-Id: <20220209105452.1694545-5-eesposit@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
This function is currently just a wrapper for bdrv_invalidate_cache(),
but in future will contain the code of bdrv_co_invalidate_cache() that
has to always be protected by BQL, and leave the rest in the I/O
coroutine.
Replace all bdrv_invalidate_cache() invokations with bdrv_activate().
Signed-off-by: Emanuele Giuseppe Esposito <eesposit@redhat.com>
Reviewed-by: Hanna Reitz <hreitz@redhat.com>
Message-Id: <20220209105452.1694545-4-eesposit@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
block_crypto_amend_options_generic_luks uses the block layer
permission API, therefore it should be called with the BQL held.
However, the same function is being called by two BlockDriver
callbacks: bdrv_amend_options (under BQL) and bdrv_co_amend (I/O).
The latter is I/O because it is invoked by block/amend.c's
blockdev_amend_run(), a .run callback of the amend JobDriver.
Therefore we want to change this function to still perform
the permission check, but making sure it is done under BQL regardless
of the caller context.
Remove the permission check in block_crypto_amend_options_generic_luks()
and:
- in block_crypto_amend_options_luks() (BQL case, called by
.bdrv_amend_options()), reuse helper functions
block_crypto_amend_{prepare/cleanup} that take care of checking
permissions.
- for block_crypto_co_amend_luks() (I/O case, called by
.bdrv_co_amend()), don't check for permissions but delegate
.bdrv_amend_pre_run() and .bdrv_amend_clean() to do it,
performing these checks before and after the job runs in its aiocontext.
Signed-off-by: Emanuele Giuseppe Esposito <eesposit@redhat.com>
Message-Id: <20220209105452.1694545-3-eesposit@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Move the permission API calls into driver-specific callbacks
that always run under BQL. In this case, bdrv_crypto_luks
needs to perform permission checks before and after
qcrypto_block_amend_options(). The problem is that the caller,
block_crypto_amend_options_generic_luks(), can also run in I/O
from .bdrv_co_amend(). This does not comply with Global State-I/O API split,
as permissions API must always run under BQL.
Firstly, introduce .bdrv_amend_pre_run() and .bdrv_amend_clean()
callbacks. These two callbacks are guaranteed to be invoked under
BQL, respectively before and after .bdrv_co_amend().
They take care of performing the permission checks
in the same way as they are currently done before and after
qcrypto_block_amend_options().
These callbacks are in preparation for next patch, where we
delete the original permission check. Right now they just add redundant
control.
Then, call .bdrv_amend_pre_run() before job_start in
qmp_x_blockdev_amend(), so that it will be run before the job coroutine
is created and stay in the main loop.
As a cleanup, use JobDriver's .clean() callback to call
.bdrv_amend_clean(), and run amend-specific cleanup callbacks under BQL.
After this patch, permission failures occur early in the blockdev-amend
job to update a LUKS volume's keys. iotest 296 must now expect them in
x-blockdev-amend's QMP reply instead of waiting for the actual job to
fail later.
Signed-off-by: Emanuele Giuseppe Esposito <eesposit@redhat.com>
Message-Id: <20220209105452.1694545-2-eesposit@redhat.com>
Signed-off-by: Hanna Reitz <hreitz@redhat.com>
Message-Id: <20220304153729.711387-6-hreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Debug output was always being sent to STDERR.
This has been replaced with trace events.
Signed-off-by: Carwyn Ellis <carwynellis@gmail.com>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-Id: <20220206183956.10694-2-carwynellis@gmail.com>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
The clock field is 16-bits in EDID Detailed Timing Descriptor, but
edid_desc_timing assumed it is 32-bit. Write the 16-bit value if it fits
in 16-bit. Write DisplayID otherwise.
Signed-off-by: Akihiko Odaki <akihiko.odaki@gmail.com>
Message-Id: <20220213021529.2248-1-akihiko.odaki@gmail.com>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Services menu functionality of Cocoa is described at:
https://developer.apple.com/design/human-interface-guidelines/macos/extensions/services/
Signed-off-by: Akihiko Odaki <akihiko.odaki@gmail.com>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Tested-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-Id: <20220214091320.51750-1-akihiko.odaki@gmail.com>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
The same info may be used to update the clipboard, and may be freed
before being ref'ed again.
Fixes: 70a54b0169 ("ui: avoid compiler warnings from unused clipboard info variable")
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Message-Id: <20220214115917.1679568-1-marcandre.lureau@redhat.com>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
qemu_console_resize() will create a blank surface and replace the
current scanout with it if called while the current scanout is
GL (texture or dmabuf).
This is not only very costly, but also can produce glitches on the
display/listener side.
Instead, compare the current console size with the fitting console
functions, which also works when the scanout is GL.
Note: there might be still an unnecessary surface creation on calling
qemu_console_resize() when the size is actually changing, but display
backends currently rely on DisplaySurface details during
dpy_gfx_switch() to handle various resize aspects. We would need more
refactoring to handle resize without DisplaySurface, this is left for a
future improvement.
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-Id: <20220214201337.1814787-4-marcandre.lureau@redhat.com>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Make surface_gl_create_texture() idempotent: if the surface is already
bound to a texture, do not create a new one.
This fixes texture leaks when there are multiple DBus listeners, for
example.
Reported-by: Akihiko Odaki <akihiko.odaki@gmail.com>
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-Id: <20220214201337.1814787-3-marcandre.lureau@redhat.com>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
The commit 7cc712e98 ("ui: dispatch GL events to all listener")
mechanically replaced the dpy_gl calls with a dispatch loop, using the
same pre-conditions. However, it didn't take into account that all
listeners do not have to implement the GL callbacks.
Add the missing pre-conditions before calling the callbacks.
Fix crash when running a GL-enabled VM with "-device virtio-gpu-gl-pci
-display egl-headless -vnc :0".
Fixes: 7cc712e98 ("ui: dispatch GL events to all listener")
Reported-by: Akihiko Odaki <akihiko.odaki@gmail.com>
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-Id: <20220214201337.1814787-2-marcandre.lureau@redhat.com>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Add docs/specs/sev-guest-firmware.rst which describes the GUIDed table
in the end of OVMF's image which is parsed by QEMU, and currently used
to describe some values for SEV and SEV-ES guests.
Signed-off-by: Dov Murik <dovmurik@linux.ibm.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Message-Id: <20220103091413.2869-1-dovmurik@linux.ibm.com>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Replce the literal magic number 48 with length calculation (32 bytes at
the end of the firmware after the table footer + 16 bytes of the OVMF
table footer GUID).
No functional change intended.
Signed-off-by: Dov Murik <dovmurik@linux.ibm.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Message-Id: <20220222071906.2632426-3-dovmurik@linux.ibm.com>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
When pc_system_parse_ovmf_flash() parses the optional GUIDed table in
the end of the OVMF flash memory area, the table length field is checked
for sizes that are too small, but doesn't error on sizes that are too
big (bigger than the flash content itself).
Add a check for maximal size of the OVMF table, and add an error report
in case the size is invalid. In such a case, an error like this will be
displayed during launch:
qemu-system-x86_64: OVMF table has invalid size 4047
and the table parsing is skipped.
Signed-off-by: Dov Murik <dovmurik@linux.ibm.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Message-Id: <20220222071906.2632426-2-dovmurik@linux.ibm.com>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Otherwise, the audio subsystem tries to use the voice and
eventually aborts due to the maximum number of samples in the
buffer is not set.
Signed-off-by: Akihiko Odaki <akihiko.odaki@gmail.com>
Reviewed-by: Christian Schoenebeck <qemu_oss@crudebyte.com>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-Id: <20220226115953.60335-1-akihiko.odaki@gmail.com>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
qemu_oom_check() is a function which essentially says "if you pass me
a NULL pointer then print a message then abort()". On POSIX systems
the message includes strerror(errno); on Windows it includes the
GetLastError() error value printed as an integer.
Other than in the implementation of qemu_memalign(), we use this
function only in hw/usb/redirect.c, for three checks:
* on a call to usbredirparser_create()
* on a call to usberedirparser_serialize()
* on a call to malloc()
The usbredir library API functions make no guarantees that they will
set errno on errors, let alone that they might set the
Windows-specific GetLastError string. malloc() is documented as
setting errno, not GetLastError -- and in any case the only thing it
might set errno to is ENOMEM. So qemu_oom_check() isn't the right
thing for any of these. Replace them with straightforward
error-checking code. This will allow us to get rid of
qemu_oom_check().
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20220226180723.1706285-2-peter.maydell@linaro.org>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Fix the same samples vs. frames mix-up that the previous commit
fixed for the PulseAudio backend.
Signed-off-by: Volker Rümelin <vr_qemu@t-online.de>
Message-Id: <20220301191311.26695-15-vr_qemu@t-online.de>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Now that the mixing buffer size no longer adds to playback
latency, fix the samples vs. frames mix-up in the mixing buffer
size calculation. This change will go largely unnoticed as long
as the user doesn't use a buffer-size smaller than timer-period.
Signed-off-by: Volker Rümelin <vr_qemu@t-online.de>
Message-Id: <20220301191311.26695-14-vr_qemu@t-online.de>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Return the free buffer size for the mmapped case in function
oss_buffer_get_free() to reduce the effective playback buffer
size. All intermediate audio playback buffers become temporary
buffers.
Signed-off-by: Volker Rümelin <vr_qemu@t-online.de>
Message-Id: <20220301191311.26695-13-vr_qemu@t-online.de>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Add the buffer_get_free pcm_ops function to reduce the effective
playback buffer size. All intermediate audio playback buffers
become temporary buffers.
Signed-off-by: Volker Rümelin <vr_qemu@t-online.de>
Message-Id: <20220301191311.26695-12-vr_qemu@t-online.de>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Add the buffer_get_free pcm_ops function to reduce the effective
playback buffer size. All intermediate audio playback buffers
become temporary buffers.
Signed-off-by: Volker Rümelin <vr_qemu@t-online.de>
Message-Id: <20220301191311.26695-11-vr_qemu@t-online.de>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Commit ff095e5231 "audio: api for mixeng code free backends"
introduced another FIFO for the audio subsystem with exactly the
same size as the mixing-engine FIFO. Most audio backends use
this generic FIFO. The generic FIFO used together with the
mixing-engine FIFO doubles the audio FIFO size, because that's
just two independent FIFOs connected together in series.
For audio playback this nearly doubles the playback latency.
This patch restores the effective mixing-engine playback buffer
size to a pre v4.2.0 size by only accepting the amount of
samples for the mixing-engine queue which the downstream queue
accepts.
Signed-off-by: Volker Rümelin <vr_qemu@t-online.de>
Reviewed-by: Akihiko Odaki <akihiko.odaki@gmail.com>
Message-Id: <20220301191311.26695-10-vr_qemu@t-online.de>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
This reverts commit cbaf25d1f5.
Since previous commit every audio backend has a pcm_ops function
table. It's no longer necessary to test if the table is available.
Signed-off-by: Volker Rümelin <vr_qemu@t-online.de>
Message-Id: <20220301191311.26695-9-vr_qemu@t-online.de>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Add a pcm_ops function table for the capture backend. This avoids
additional code in the next patches to test if the pcm_ops table
is available.
Signed-off-by: Volker Rümelin <vr_qemu@t-online.de>
Message-Id: <20220301191311.26695-8-vr_qemu@t-online.de>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Change the code to copy the playback stream in sequential order.
The advantage can be seen in the next patches where the stream
copy operation effectively becomes a write through operation.
The following diagram shows the average buffer fill level and
the stream copy sequence. ### represents a timer_period sized
chunk. The rest of the buffer sizes are not to scale.
With current code:
|--------| |#####111| |---#####|
sw->buf mix_buf backend buffer
1. clip
|--------| |---#####| |111##222|
sw->buf mix_buf backend buffer
2. write to audio device
333 -> |--------| |---#####| |---111##| -> 222
sw->buf mix_buf backend buffer
3a. sw device write
|-----333| |---#####| |---111##|
sw->buf mix_buf backend buffer
3b. resample and mix
|--------| |333#####| |---111##|
sw->buf mix_buf backend buffer
With this patch:
111 -> |--------| |---#####| |---#####|
sw->buf mix_buf backend buffer
1a: sw device write
|-----111| |---#####| |---#####|
sw->buf mix_buf backend buffer
1b. resample and mix
|--------| |111##222| |---#####|
sw->buf mix_buf backend buffer
2. clip
|--------| |---111##| |222##333|
sw->buf mix_buf backend buffer
3. write to audio device
|--------| |---111##| |---222##| -> 333
sw->buf mix_buf backend buffer
The effective total playback buffer size is reduced by
timer_period.
Signed-off-by: Volker Rümelin <vr_qemu@t-online.de>
Message-Id: <20220301191311.26695-7-vr_qemu@t-online.de>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
The next patch reduces the effective qemu playback buffer size
by timer-period. Increase the number of jack audio buffers by
one to preserve the total effective buffer size. The size of one
jack audio buffer is 512 samples. With audio defaults that's
512 samples / 44100 samples/s = 11.6 ms and only slightly larger
than the timer-period of 10 ms.
The larger jack audio buffer increases audio dropout safety,
because the high priority jack-audio worker threads can provide
audio data for a longer period of time as with a smaller buffer
and more audio data in the mixing engine buffer that they can't
access.
Signed-off-by: Volker Rümelin <vr_qemu@t-online.de>
Reviewed-by: Christian Schoenebeck <qemu_oss@crudebyte.com>
Message-Id: <20220301191311.26695-6-vr_qemu@t-online.de>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
This is a patch to improve the pulseaudio playback experience.
Asking pulseaudio for a playback latency of 15ms is quite
demanding. Increase this to 46ms. The total playback latency
now is 31ms larger. One of the next patches will reduce the
total playback latency again by more than 46ms.
Here is a quote from the PulseAudio Latency Control
documentation: 'For the sake of (...) drop-out safety always
make sure to pick the highest latency possible that fulfills
your needs.'
Signed-off-by: Volker Rümelin <vr_qemu@t-online.de>
Message-Id: <20220301191311.26695-5-vr_qemu@t-online.de>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Simplify code by inlining function audio_pcm_sw_get_rpos_in()
at the only call site and remove the duplicated audio_bug()
test.
Signed-off-by: Volker Rümelin <vr_qemu@t-online.de>
Message-Id: <20220301191311.26695-4-vr_qemu@t-online.de>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Add a function audio_pcm_hw_conv_in() similar to the existing
counterpart function audio_pcm_hw_clip_out(). This function reduces
the number of calls to the pcm_ops functions get_buffer_in() and
put_buffer_in(). That's one less call to get_buffer_in() and
put_buffer_in() every time the conv_buffer wraps around.
Signed-off-by: Volker Rümelin <vr_qemu@t-online.de>
Message-Id: <20220301191311.26695-3-vr_qemu@t-online.de>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Move the function audio_pcm_hw_clip_out() into the correct
section 'Hard voice (playback)'.
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Signed-off-by: Volker Rümelin <vr_qemu@t-online.de>
Message-Id: <20220301191311.26695-2-vr_qemu@t-online.de>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Replace open-coded buffer arithmetic with the new function
audio_ring_posb(). That's the position in backward direction
of a given point at a given distance.
Signed-off-by: Volker Rümelin <vr_qemu@t-online.de>
Reviewed-by: Akihiko Odaki <akihiko.odaki@gmail.com>
Message-Id: <20220301191311.26695-1-vr_qemu@t-online.de>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Since isochronous transfers cannot be handled async (the function
returns error in that case) we don't need to remember the packet.
Avoid using the usb_packet field in OHCIState (as that can be a
waiting async packet on another endpoint) and allocate and use a local
USBPacket for the iso transfer instead. After this we don't have to
care if we're called from a completion callback or not so we can drop
that parameter as well.
Signed-off-by: BALATON Zoltan <balaton@eik.bme.hu>
Message-Id: <bf523d40f8088a84383cb00ffd2e6e82fa47790d.1643117600.git.balaton@eik.bme.hu>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
These two do the same and only used once so no need to have two
functions, simplify by merging them.
Signed-off-by: BALATON Zoltan <balaton@eik.bme.hu>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-Id: <5fc8ba0bbf55703014d22dd06ab2f9eabaf370bf.1643117600.git.balaton@eik.bme.hu>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
This is always done before calling this function so remove duplicated
code and do it within the function at one place.
Signed-off-by: BALATON Zoltan <balaton@eik.bme.hu>
Message-Id: <ce766722506bfd7145cccbec750692ff57072280.1643117600.git.balaton@eik.bme.hu>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
