When guests send misaligned discard requests, the block layer breaks
them up into a misaligned head, an aligned main body, and a misaligned
tail.
The file-posix block driver on Linux returns -EINVAL on misaligned
discard requests. This causes bdrv_co_pdiscard() to fail and guests
configured with werror=stop will pause.
Add a special case for misaligned head/tail requests. Simply continue
when EINVAL is encountered so that the aligned main body of the request
can be completed and the guest is not paused. This is the best we can do
when guest discard limits do not match the host discard limits.
Fixes: https://issues.redhat.com/browse/RHEL-86032
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Hanna Czenczek <hreitz@redhat.com>
Message-ID: <20250417150528.76470-3-stefanha@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Convert the existing includes with
sed -i ,exec/memory.h,system/memory.h,g
Move the include within cpu-all.h into a !CONFIG_USER_ONLY block.
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Reviewed-by: Pierrick Bouvier <pierrick.bouvier@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
qcow2_refresh_limits() assumes that s->crypto is non-NULL whenever
bs->encrypted is true. This is actually not the case: qcow2_do_open()
allows to open an image with a missing crypto header for BDRV_O_NO_IO,
and then bs->encrypted is true, but s->crypto is still NULL.
It doesn't make sense to open an invalid image, so remove the exception
for BDRV_O_NO_IO. This catches the problem early and any code that makes
the same assumption is safe now.
At the same time, in the name of defensive programming, we shouldn't
make the assumption in the first place. Let qcow2_refresh_limits() check
s->crypto rather than bs->encrypted. If s->crypto is NULL, it also can't
make any requirement on request alignment.
Finally, start a qcow2-encryption test case that only serves as a
regression test for this crash for now.
Reported-by: Leonid Reviakin <L.reviakin@fobos-nt.ru>
Reported-by: Denis Rastyogin <gerben@altlinux.org>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Message-ID: <20250318201143.70657-1-kwolf@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
For block drivers that don't advertise FUA support, we already call
bdrv_co_flush(), which considers BDRV_O_NO_FLUSH. However, drivers that
do support FUA still see the FUA flag with BDRV_O_NO_FLUSH and get the
associated performance penalty that cache.no-flush=on was supposed to
avoid.
Clear FUA for write requests if BDRV_O_NO_FLUSH is set.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Message-ID: <20250307221634.71951-3-kwolf@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Until now, FUA was always emulated with a separate flush after the write
for file-posix. The overhead of processing a second request can reduce
performance significantly for a guest disk that has disabled the write
cache, especially if the host disk is already write through, too, and
the flush isn't actually doing anything.
Advertise support for REQ_FUA in write requests and implement it for
Linux AIO and io_uring using the RWF_DSYNC flag for write requests. The
thread pool still performs a separate fdatasync() call. This can be
improved later by using the pwritev2() syscall if available.
As an example, this is how fio numbers can be improved in some scenarios
with this patch (all using virtio-blk with cache=directsync on an nvme
block device for the VM, fio with ioengine=libaio,direct=1,sync=1):
| old | with FUA support
------------------------------+---------------+-------------------
bs=4k, iodepth=1, numjobs=1 | 45.6k iops | 56.1k iops
bs=4k, iodepth=1, numjobs=16 | 183.3k iops | 236.0k iops
bs=4k, iodepth=16, numjobs=1 | 258.4k iops | 311.1k iops
However, not all scenarios are clear wins. On another slower disk I saw
little to no improvment. In fact, in two corner case scenarios, I even
observed a regression, which I however consider acceptable:
1. On slow host disks in a write through cache mode, when the guest is
using virtio-blk in a separate iothread so that polling can be
enabled, and each completion is quickly followed up with a new
request (so that polling gets it), it can happen that enabling FUA
makes things slower - the additional very fast no-op flush we used to
have gave the adaptive polling algorithm a success so that it kept
polling. Without it, we only have the slow write request, which
disables polling. This is a problem in the polling algorithm that
will be fixed later in this series.
2. With a high queue depth, it can be beneficial to have flush requests
for another reason: The optimisation in bdrv_co_flush() that flushes
only once per write generation acts as a synchronisation mechanism
that lets all requests complete at the same time. This can result in
better batching and if the disk is very fast (I only saw this with a
null_blk backend), this can make up for the overhead of the flush and
improve throughput. In theory, we could optionally introduce a
similar artificial latency in the normal completion path to achieve
the same kind of completion batching. This is not implemented in this
series.
Compatibility is not a concern for the kernel side of io_uring, it has
supported RWF_DSYNC from the start. However, io_uring_prep_writev2() is
not available before liburing 2.2.
Linux AIO started supporting it in Linux 4.13 and libaio 0.3.111. The
kernel is not a problem for any supported build platform, so it's not
necessary to add runtime checks. However, openSUSE is still stuck with
an older libaio version that would break the build.
We must detect the presence of the writev2 functions in the user space
libraries at build time to avoid build failures.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Message-ID: <20250307221634.71951-2-kwolf@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Block drivers assume in their .bdrv_open() implementation that their
state in bs->opaque has been zeroed; it is initially allocated with
g_malloc0() in bdrv_open_driver().
bdrv_snapshot_goto() needs to make sure that it is zeroed again before
calling drv->bdrv_open() to avoid that block drivers use stale values.
One symptom of this bug is VMDK running into a double free when the user
tries to apply an internal snapshot like 'qemu-img snapshot -a test
test.vmdk'. This should be a graceful error because VMDK doesn't support
internal snapshots.
==25507== Invalid free() / delete / delete[] / realloc()
==25507== at 0x484B347: realloc (vg_replace_malloc.c:1801)
==25507== by 0x54B592A: g_realloc (gmem.c:171)
==25507== by 0x1B221D: vmdk_add_extent (../block/vmdk.c:570)
==25507== by 0x1B1084: vmdk_open_sparse (../block/vmdk.c:1059)
==25507== by 0x1AF3D8: vmdk_open (../block/vmdk.c:1371)
==25507== by 0x1A2AE0: bdrv_snapshot_goto (../block/snapshot.c:299)
==25507== by 0x205C77: img_snapshot (../qemu-img.c:3500)
==25507== by 0x58FA087: (below main) (libc_start_call_main.h:58)
==25507== Address 0x832f3e0 is 0 bytes inside a block of size 272 free'd
==25507== at 0x4846B83: free (vg_replace_malloc.c:989)
==25507== by 0x54AEAC4: g_free (gmem.c:208)
==25507== by 0x1AF629: vmdk_close (../block/vmdk.c:2889)
==25507== by 0x1A2A9C: bdrv_snapshot_goto (../block/snapshot.c:290)
==25507== by 0x205C77: img_snapshot (../qemu-img.c:3500)
==25507== by 0x58FA087: (below main) (libc_start_call_main.h:58)
This error was discovered by fuzzing qemu-img.
Cc: qemu-stable@nongnu.org
Closes: https://gitlab.com/qemu-project/qemu/-/issues/2853
Closes: https://gitlab.com/qemu-project/qemu/-/issues/2851
Reported-by: Denis Rastyogin <gerben@altlinux.org>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Message-ID: <20250310104858.28221-1-kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Generic CPUs / accelerators patch queue
- Merge "qemu/clang-tsa.h" within "qemu/compiler.h"
- Various cleanups around accelerators initialization code
(better user/system split)
- Various trivial cleanups in accel/tcg/,
Guard few TCG calls with tcg_enabled()
- Explicit disassemble_info endianness
- Improve dual-endianness support for MicroBlaze
# -----BEGIN PGP SIGNATURE-----
#
# iQIzBAABCAAdFiEE+qvnXhKRciHc/Wuy4+MsLN6twN4FAmfJw08ACgkQ4+MsLN6t
# wN70whAAtfcdWtqseFfb6fvDtjflgxN51Ui0iaOECXUA18USKriGy34eBcMYMiM2
# +eKgU7+jI6JGE4+burcgWUsPpFFF951/A8+lyIbFgO5yToTDmC+qNe4XfmMAIyXq
# uf9Obr2c0Xk9luh4odb+jPAQodw/7G1fKgcCVIJNDCl/xEcPhS9eNpTaHwcVnkWI
# K6KrxWXOsqG6+evJBPWYoXtOOyt0+JcwAsJoGhprwtGm3P9+jSVXsgeGsJVyZcna
# f32JtjWL754O8XeMkOn4x6rt58VrCIMKI9xT7keDyuhTCq0Zki9RO2nMU2dSw5mN
# AfL9hxqUy0Nijnyslg3ugujDfTePsNyLdwwH7n0mnoD72ELi6WnhDsmOThuEB3Rd
# 4/kdwTJfA/rlWk/GF1tbKW7AvQZokRARtzmL3V0HmGJu57lX+2JuszEdYBkqDEP7
# GH1I10B2yANUm+C9y3X8qWOU7Ws433ebJeJoZuyfnbZ9Me+UfRmql/oS+V8ata2i
# fArEItpldUFrWRyYLkTbXrh2dgyV9yJTEir/lzOzeAZZzyabTbjf2z9qnh976GGO
# 1QnDy5QA4f54kDBUZe7JK26TZsHPch7cgqXW6f8tRlJF7A9hxGK8d2TUV/lC3/vx
# LUOlWNu03PhiruYmZEcWOsY3Jt9jRCF6lIryrnaJsqnVOVmMUMM=
# =3TRh
# -----END PGP SIGNATURE-----
# gpg: Signature made Thu 06 Mar 2025 23:46:23 HKT
# gpg: using RSA key FAABE75E12917221DCFD6BB2E3E32C2CDEADC0DE
# gpg: Good signature from "Philippe Mathieu-Daudé (F4BUG) <f4bug@amsat.org>" [full]
# Primary key fingerprint: FAAB E75E 1291 7221 DCFD 6BB2 E3E3 2C2C DEAD C0DE
* tag 'accel-cpus-20250306' of https://github.com/philmd/qemu: (54 commits)
include: Poison TARGET_PHYS_ADDR_SPACE_BITS definition
system: Open-code qemu_init_arch_modules() using target_name()
target/i386: Mark WHPX APIC region as little-endian
target/alpha: Do not mix exception flags and FPCR bits
target/riscv: Convert misa_mxl_max using GLib macros
target/riscv: Declare RISCVCPUClass::misa_mxl_max as RISCVMXL
target/xtensa: Finalize config in xtensa_register_core()
target/sparc: Constify SPARCCPUClass::cpu_def
target/i386: Constify X86CPUModel uses
disas: Remove target_words_bigendian() call in initialize_debug_target()
target/xtensa: Set disassemble_info::endian value in disas_set_info()
target/sh4: Set disassemble_info::endian value in disas_set_info()
target/riscv: Set disassemble_info::endian value in disas_set_info()
target/ppc: Set disassemble_info::endian value in disas_set_info()
target/mips: Set disassemble_info::endian value in disas_set_info()
target/microblaze: Set disassemble_info::endian value in disas_set_info
target/arm: Set disassemble_info::endian value in disas_set_info()
target: Set disassemble_info::endian value for big-endian targets
target: Set disassemble_info::endian value for little-endian targets
target/mips: Fix possible MSA int overflow
...
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
This error was discovered by fuzzing qemu-img.
In the QED block driver, the need_check_timer timer is freed in
bdrv_qed_detach_aio_context, but the pointer to the timer is not
set to NULL. This can lead to a use-after-free scenario
in bdrv_qed_drain_begin().
The need_check_timer pointer is set to NULL after freeing the timer.
Which helps catch this condition when checking in bdrv_qed_drain_begin().
Closes: https://gitlab.com/qemu-project/qemu/-/issues/2852
Signed-off-by: Denis Rastyogin <gerben@altlinux.org>
Message-ID: <20250304083927.37681-1-gerben@altlinux.org>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Although defaulting the handshake limit to 10 seconds was a nice QoI
change to weed out intentionally slow clients, it can interfere with
integration testing done with manual NBD_OPT commands over 'nbdsh
--opt-mode'. Expose a QMP knob 'handshake-max-secs' to allow the user
to alter the timeout away from the default.
The parameter name here intentionally matches the spelling of the
constant added in commit fb1c2aaa98, and not the command-line spelling
added in the previous patch for qemu-nbd; that's because in QMP,
longer names serve as good self-documentation, and unlike the command
line, machines don't have problems generating longer spellings.
Signed-off-by: Eric Blake <eblake@redhat.com>
Message-ID: <20250203222722.650694-6-eblake@redhat.com>
[eblake: s/max-secs/max-seconds/ in QMP]
Acked-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
The general expectation is that header files should follow the same
file/path naming scheme as the corresponding source file. There are
various historical exceptions to this practice in QEMU, with one of
the most notable being the include/qapi/qmp/ directory. Most of the
headers there correspond to source files in qobject/.
This patch corrects most of that inconsistency by creating
include/qobject/ and moving the headers for qobject/ there.
This also fixes MAINTAINERS for include/qapi/qmp/dispatch.h:
scripts/get_maintainer.pl now reports "QAPI" instead of "No
maintainers found".
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Zhao Liu <zhao1.liu@intel.com>
Acked-by: Halil Pasic <pasic@linux.ibm.com> #s390x
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Message-ID: <20241118151235.2665921-2-armbru@redhat.com>
[Rebased]
BLOCK_OP_TYPE_DATAPLANE prevents BlockDriverState from being used by
virtio-blk/virtio-scsi with IOThread. Commit b112a65c52 ("block:
declare blockjobs and dataplane friends!") eliminated the main reason
for this blocker in 2014.
Nowadays the block layer supports I/O from multiple AioContexts, so
there is even less reason to block IOThread users. Any legitimate
reasons related to interference would probably also apply to
non-IOThread users.
The only remaining users are bdrv_op_unblock(BLOCK_OP_TYPE_DATAPLANE)
calls after bdrv_op_block_all(). If we remove BLOCK_OP_TYPE_DATAPLANE
their behavior doesn't change.
Existing bdrv_op_block_all() callers that don't explicitly unblock
BLOCK_OP_TYPE_DATAPLANE seem to do so simply because no one bothered to
rather than because it is necessary to keep BLOCK_OP_TYPE_DATAPLANE
blocked.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-ID: <20250203182529.269066-1-stefanha@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Add an option in BlockExportOptions to allow creating an export on an
inactive node without activating the node. This mode needs to be
explicitly supported by the export type (so that it doesn't perform any
operations that are forbidden for inactive nodes), so this patch alone
doesn't allow this option to be successfully used yet.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Acked-by: Fabiano Rosas <farosas@suse.de>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-ID: <20250204211407.381505-13-kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Device models have a relatively complex way to set up their block
backends, in which blk_attach_dev() sets blk->disable_perm = true.
We want to support inactive images in exports, too, so that
qemu-storage-daemon can be used with migration. Because they don't use
blk_attach_dev(), they need another way to set this flag. The most
convenient is to do this automatically when an inactive node is attached
to a BlockBackend that can be inactivated.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Acked-by: Fabiano Rosas <farosas@suse.de>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-ID: <20250204211407.381505-10-kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
In order for block_resize to fail gracefully on an inactive node instead
of crashing with an assertion failure in bdrv_co_write_req_prepare()
(called from bdrv_co_truncate()), we need to check for inactive nodes
also when they are attached as a root node and make sure that
BLK_PERM_RESIZE isn't among the permissions allowed for inactive nodes.
To this effect, don't enumerate the permissions that are incompatible
with inactive nodes any more, but allow only BLK_PERM_CONSISTENT_READ
for them.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Acked-by: Fabiano Rosas <farosas@suse.de>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-ID: <20250204211407.381505-7-kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Commit 7452162ade introduced 'qom-path' argument to BLOCK_IO_ERROR
event but when the event is instantiated in 'send_qmp_error_event()' the
arguments for 'device' and 'qom_path' in
qapi_event_send_block_io_error() were reversed :
Generated code for sending event:
void qapi_event_send_block_io_error(const char *qom_path,
const char *device,
const char *node_name,
IoOperationType operation,
[...]
Call inside send_qmp_error_event():
qapi_event_send_block_io_error(blk_name(blk),
blk_get_attached_dev_path(blk),
bs ? bdrv_get_node_name(bs) : NULL, optype,
[...]
This results into reporting the QOM path as the device alias and vice
versa which in turn breaks libvirt, which expects the device alias being
either a valid alias or empty (which would make libvirt do the lookup by
node-name instead).
Cc: qemu-stable@nongnu.org
Fixes: 7452162ade ("qapi: add qom-path to BLOCK_IO_ERROR event")
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Message-ID: <09728d784888b38d7a8f09ee5e9e9c542c875e1e.1737973614.git.pkrempa@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Philippe Mathieu-Daudé
Stefan HajnocziandKevin Wolf
Markus Armbruster
Richard Henderson
Denis RastyoginandStefan Hajnoczi
Eric Blake
Daniel P. BerrangéandMarkus Armbruster
Peter KrempaandKevin Wolf