aboutsummaryrefslogtreecommitdiff
AgeCommit message (Collapse)AuthorFilesLines
2018-12-14hw/sd/sdhci: Don't leak memory region in sdhci_sysbus_realize()Peter Maydell1-4/+1
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(). Spotted by clang's leak sanitizer. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com> Message-id: 20181204132952.2601-4-peter.maydell@linaro.org
2018-12-14hw/arm/mps2-tz.c: Free mscname string in make_dma()Peter Maydell1-0/+1
The clang leak sanitizer spots a (one-off, trivial) memory leak in make_dma() due to a missing free. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com> Message-id: 20181204132952.2601-3-peter.maydell@linaro.org
2018-12-14target/arm: Free name string in ARMCPRegInfo hashtable entriesPeter Maydell1-1/+15
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. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com> Message-id: 20181204132952.2601-2-peter.maydell@linaro.org
2018-12-14include/hw/loader.h: Document load_image_size()Peter Maydell1-0/+16
Add a documentation comment for load_image_size(). Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Michael S. Tsirkin <mst@redhat.com> Reviewed-by: Eric Blake <eblake@redhat.com> Message-id: 20181130151712.2312-11-peter.maydell@linaro.org
2018-12-14hw/core/loader.c: Remove load_image()Peter Maydell2-26/+0
The load_image() function is now no longer used anywhere, so we can remove it completely. (Use load_image_size() or g_file_get_contents() instead.) Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Michael S. Tsirkin <mst@redhat.com> Reviewed-by: Eric Blake <eblake@redhat.com> Message-id: 20181130151712.2312-10-peter.maydell@linaro.org
2018-12-14device_tree.c: Don't use load_image()Peter Maydell1-1/+1
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(). Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Michael S. Tsirkin <mst@redhat.com> Reviewed-by: Eric Blake <eblake@redhat.com> Message-id: 20181130151712.2312-9-peter.maydell@linaro.org
2018-12-14hw/block/tc58128.c: Don't use load_image()Peter Maydell1-1/+2
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(). Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Michael S. Tsirkin <mst@redhat.com> Reviewed-by: Eric Blake <eblake@redhat.com> Message-id: 20181130151712.2312-8-peter.maydell@linaro.org
2018-12-14hw/i386/multiboot.c: Don't use load_image()Peter Maydell1-1/+5
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. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Michael S. Tsirkin <mst@redhat.com> Reviewed-by: Eric Blake <eblake@redhat.com> Message-id: 20181130151712.2312-7-peter.maydell@linaro.org
2018-12-14hw/i386/pc.c: Don't use load_image()Peter Maydell1-10/+12
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. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Michael S. Tsirkin <mst@redhat.com> Reviewed-by: Eric Blake <eblake@redhat.com> Message-id: 20181130151712.2312-6-peter.maydell@linaro.org
2018-12-14hw/pci/pci.c: Don't use load_image()Peter Maydell1-1/+5
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. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Michael S. Tsirkin <mst@redhat.com> Reviewed-by: Eric Blake <eblake@redhat.com> Message-id: 20181130151712.2312-5-peter.maydell@linaro.org
2018-12-14hw/smbios/smbios.c: Don't use load_image()Peter Maydell1-1/+1
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(). Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Michael S. Tsirkin <mst@redhat.com> Reviewed-by: Eric Blake <eblake@redhat.com> Message-id: 20181130151712.2312-4-peter.maydell@linaro.org
2018-12-14hw/ppc/ppc405_boards: Don't use load_image()Peter Maydell1-4/+8
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(). Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Michael S. Tsirkin <mst@redhat.com> Reviewed-by: Eric Blake <eblake@redhat.com> Acked-by: David Gibson <david@gibson.dropbear.id.au> Message-id: 20181130151712.2312-3-peter.maydell@linaro.org
2018-12-14hw/ppc/mac_newworld, mac_oldworld: Don't use load_image()Peter Maydell2-12/+8
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. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Michael S. Tsirkin <mst@redhat.com> Reviewed-by: Eric Blake <eblake@redhat.com> Acked-by: David Gibson <david@gibson.dropbear.id.au> Message-id: 20181130151712.2312-2-peter.maydell@linaro.org
2018-12-14elf_ops.h: Use address_space_write() to write memoryPeter Maydell1-1/+3
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. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com> Message-id: 20181122172653.3413-4-peter.maydell@linaro.org
2018-12-14monitor: Use address_space_read() to read memoryPeter Maydell1-1/+7
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. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com> Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com> Message-id: 20181122172653.3413-3-peter.maydell@linaro.org
2018-12-14disas.c: Use address_space_read() to read memoryPeter Maydell1-1/+4
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. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com> Message-id: 20181122172653.3413-2-peter.maydell@linaro.org
2018-12-14Rename cpu_physical_memory_write_rom() to address_space_write_rom()Peter Maydell8-35/+60
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(). Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com> Acked-by: Michael S. Tsirkin <mst@redhat.com> Message-id: 20181122133507.30950-3-peter.maydell@linaro.org
2018-12-14exec.c: Rename cpu_physical_memory_write_rom_internal()Peter Maydell1-7/+13
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(). Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com> Acked-by: Michael S. Tsirkin <mst@redhat.com> Message-id: 20181122133507.30950-2-peter.maydell@linaro.org
2018-12-14scripts/checkpatch.pl: Enforce multiline comment syntaxPeter Maydell1-0/+48
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 Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Acked-by: Thomas Huth <thuth@redhat.com> Reviewed-by: Markus Armbruster <armbru@redhat.com>
2018-12-14block/mirror: add missing coroutine_fn annotationsStefan Hajnoczi1-8/+11
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. Cc: Max Reitz <mreitz@redhat.com> Cc: John Snow <jsnow@redhat.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2018-12-14iotests: make 235 work on s390 (and others)Christian Borntraeger1-1/+3
"-machine pc" will not work all architectures. Lets fall back to the default machine by not specifying it. In addition we also need to specify -no-shutdown on s390 as qemu will exit otherwise. Cc: qemu-stable@nongnu.org Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com> Reviewed-by: Eric Blake <eblake@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2018-12-14block: Assert that flags are up-to-date in bdrv_reopen_prepare()Alberto Garcia1-0/+6
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. Signed-off-by: Alberto Garcia <berto@igalia.com> Reviewed-by: Max Reitz <mreitz@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2018-12-14block: Remove assertions from update_flags_from_options()Alberto Garcia3-4/+16
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: $ qemu-io -c 'reopen -o read-only=foo' hd.qcow2 block.c:1126: update_flags_from_options: Assertion `qemu_opt_find(opts, BDRV_OPT_CACHE_DIRECT)' failed. Aborted 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: $ qemu-io -c 'reopen -o read-only=foo' hd.qcow2 Parameter 'read-only' expects 'on' or 'off' Signed-off-by: Alberto Garcia <berto@igalia.com> Reviewed-by: Max Reitz <mreitz@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2018-12-14block: Stop passing flags to bdrv_reopen_queue_child()Alberto Garcia1-29/+19
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(). Signed-off-by: Alberto Garcia <berto@igalia.com> Reviewed-by: Max Reitz <mreitz@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2018-12-14block: Remove flags parameter from bdrv_reopen_queue()Alberto Garcia4-9/+7
Now that all callers are passing all flag changes as QDict options, the flags parameter is no longer necessary, so we can get rid of it. Signed-off-by: Alberto Garcia <berto@igalia.com> Reviewed-by: Max Reitz <mreitz@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2018-12-14block: Clean up reopen_backing_file() in block/replication.cAlberto Garcia1-24/+21
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. Signed-off-by: Alberto Garcia <berto@igalia.com> Reviewed-by: Max Reitz <mreitz@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2018-12-14qemu-io: Put flag changes in the options QDict in reopen_f()Alberto Garcia3-1/+43
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. Signed-off-by: Alberto Garcia <berto@igalia.com> Reviewed-by: Max Reitz <mreitz@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2018-12-14block: Drop bdrv_reopen()Alberto Garcia2-22/+0
No one is using this function anymore, so we can safely remove it. Signed-off-by: Alberto Garcia <berto@igalia.com> Reviewed-by: Max Reitz <mreitz@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2018-12-14block: Use bdrv_reopen_set_read_only() in the mirror driverAlberto Garcia1-7/+12
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. Signed-off-by: Alberto Garcia <berto@igalia.com> Reviewed-by: Max Reitz <mreitz@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2018-12-14block: Use bdrv_reopen_set_read_only() in external_snapshot_commit()Alberto Garcia1-2/+1
This patch replaces the bdrv_reopen() call that set and remove the BDRV_O_RDWR flag with the new bdrv_reopen_set_read_only() function. Signed-off-by: Alberto Garcia <berto@igalia.com> Reviewed-by: Max Reitz <mreitz@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2018-12-14block: Use bdrv_reopen_set_read_only() in qmp_change_backing_file()Alberto Garcia1-6/+2
This patch replaces the bdrv_reopen() calls that set and remove the BDRV_O_RDWR flag with the new bdrv_reopen_set_read_only() function. Signed-off-by: Alberto Garcia <berto@igalia.com> Reviewed-by: Max Reitz <mreitz@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2018-12-14block: Use bdrv_reopen_set_read_only() in stream_start/complete()Alberto Garcia1-10/+10
This patch replaces the bdrv_reopen() calls that set and remove the BDRV_O_RDWR flag with the new bdrv_reopen_set_read_only() function. Signed-off-by: Alberto Garcia <berto@igalia.com> Reviewed-by: Max Reitz <mreitz@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2018-12-14block: Use bdrv_reopen_set_read_only() in bdrv_commit()Alberto Garcia1-4/+3
This patch replaces the bdrv_reopen() calls that set and remove the BDRV_O_RDWR flag with the new bdrv_reopen_set_read_only() function. Signed-off-by: Alberto Garcia <berto@igalia.com> Reviewed-by: Max Reitz <mreitz@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2018-12-14block: Use bdrv_reopen_set_read_only() in commit_start/complete()Alberto Garcia1-10/+6
This patch replaces the bdrv_reopen() calls that set and remove the BDRV_O_RDWR flag with the new bdrv_reopen_set_read_only() function. Signed-off-by: Alberto Garcia <berto@igalia.com> Reviewed-by: Max Reitz <mreitz@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2018-12-14block: Use bdrv_reopen_set_read_only() in bdrv_backing_update_filename()Alberto Garcia1-5/+5
This patch replaces the bdrv_reopen() calls that set and remove the BDRV_O_RDWR flag with the new bdrv_reopen_set_read_only() function. Signed-off-by: Alberto Garcia <berto@igalia.com> Reviewed-by: Max Reitz <mreitz@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2018-12-14block: Add bdrv_reopen_set_read_only()Alberto Garcia2-0/+19
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()). Signed-off-by: Alberto Garcia <berto@igalia.com> Reviewed-by: Max Reitz <mreitz@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2018-12-14file-posix: Avoid aio_worker() for QEMU_AIO_IOCTLKevin Wolf1-40/+17
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. Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2018-12-14file-posix: Switch to .bdrv_co_ioctlKevin Wolf4-28/+24
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. Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2018-12-14file-posix: Remove paio_submit_co()Kevin Wolf1-34/+0
The function is not used any more, remove it. Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2018-12-14file-posix: Avoid aio_worker() for QEMU_AIO_READ/WRITEKevin Wolf1-8/+19
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. Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2018-12-14file-posix: Move read/write operation logic out of aio_worker()Kevin Wolf1-20/+20
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. Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2018-12-14file-posix: Avoid aio_worker() for QEMU_AIO_FLUSHKevin Wolf1-5/+11
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. Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2018-12-14file-posix: Avoid aio_worker() for QEMU_AIO_DISCARDKevin Wolf1-8/+24
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. Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2018-12-14file-posix: Avoid aio_worker() for QEMU_AIO_WRITE_ZEROESKevin Wolf1-19/+34
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. Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2018-12-14file-posix: Avoid aio_worker() for QEMU_AIO_COPY_RANGEKevin Wolf1-5/+17
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. Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2018-12-14file-posix: Avoid aio_worker() for QEMU_AIO_TRUNCATEKevin Wolf1-6/+6
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. Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2018-12-14file-posix: Factor out raw_thread_pool_submit()Kevin Wolf1-7/+10
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. Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2018-12-14file-posix: Reorganise RawPosixAIODataKevin Wolf1-40/+49
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. Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2018-12-14qcow2: do decompression in threadsVladimir Sementsov-Ogievskiy1-7/+27
Do decompression in threads, like it is already done for compression. This improves asynchronous compressed reads performance. Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Alberto Garcia <berto@igalia.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2018-12-14qcow2: aio support for compressed cluster readVladimir Sementsov-Ogievskiy2-51/+49
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. Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Alberto Garcia <berto@igalia.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
generated by cgit v1.1 at 2025-01-01 01:53:25 +0000