Check that the input sp is 16 byte aligned, not 4.
Do that before the lock_user_struct check.
Validate the saved sp is 8 byte aligned.
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20210426025334.1168495-22-richard.henderson@linaro.org>
Signed-off-by: Laurent Vivier <laurent@vivier.eu>
Clean up a goto label with a single use. Remove #if 0.
Remove useless parentheses. Fold constants into __put_user.
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20210426025334.1168495-21-richard.henderson@linaro.org>
Signed-off-by: Laurent Vivier <laurent@vivier.eu>
Stub it out to zero, but at least include it.
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20210426025334.1168495-20-richard.henderson@linaro.org>
Signed-off-by: Laurent Vivier <laurent@vivier.eu>
Replace __siginfo_t with target_pt_regs, and move si_mask
into target_signal_frame directly.
Extract save/restore functions for target_pt_regs. Adjust
for sparc64 tstate. Use proper get/put functions for psr.
Turns out we were already writing to si_mask twice, so no
need to handle that in the new functions.
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20210426025334.1168495-16-richard.henderson@linaro.org>
Signed-off-by: Laurent Vivier <laurent@vivier.eu>
Move target_reg_window up and use it. Fold structptr and xxargs
into xargs -- the use of a host pointer was incorrect anyway.
Rename the structure to target_stackf for consistency.
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20210426025334.1168495-15-richard.henderson@linaro.org>
Signed-off-by: Laurent Vivier <laurent@vivier.eu>
All of the source and header files already defer to sparc
via #include. The syscall.tbl and syscallhdr.sh files
could not do the same, but are identical.
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20210426025334.1168495-12-richard.henderson@linaro.org>
Signed-off-by: Laurent Vivier <laurent@vivier.eu>
There are only a few differences in sparc32 vs sparc64.
This fixes target_shmlba for sparc32plus, which is v9.
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20210426025334.1168495-8-richard.henderson@linaro.org>
Signed-off-by: Laurent Vivier <laurent@vivier.eu>
Share code between sparc32 and sparc64, removing a bit of pointless
difference wrt psr/tstate. Use sizeof(abi_ulong) for allocating
initial register window. Use TARGET_STACK_BIAS.
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20210426025334.1168495-7-richard.henderson@linaro.org>
Signed-off-by: Laurent Vivier <laurent@vivier.eu>
Move TARGET_STACK_BIAS from signal.c. Generic code cares about the
logical stack pointer, not the physical one that has a bias applied
for sparc64.
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20210426025334.1168495-6-richard.henderson@linaro.org>
Signed-off-by: Laurent Vivier <laurent@vivier.eu>
In most cases we were already passing get_sp_from_cpustate
directly to the function. In other cases, we were passing
a local variable which already contained the same value.
In the rest of the cases, we were passing the stack pointer
out of env directly.
Reviewed by: Warner Losh <imp@bsdimp.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20210426025334.1168495-5-richard.henderson@linaro.org>
Signed-off-by: Laurent Vivier <laurent@vivier.eu>
Now that we have exactly one call, it's easy to pass
in env instead of passing in the sp value.
Use target_save_altstack, which required env.
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20210426025334.1168495-4-richard.henderson@linaro.org>
Signed-off-by: Laurent Vivier <laurent@vivier.eu>
Note that target_restore_altstack uses the host memory
pointer that we have already verified, so TARGET_EFAULT
is not a possible return value.
Note that using -EFAULT was a bug.
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20210426025334.1168495-3-richard.henderson@linaro.org>
Signed-off-by: Laurent Vivier <laurent@vivier.eu>
Create a function to match target_save_altstack.
Fix some style and unlock issues in do_sigaltstack.
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20210426025334.1168495-2-richard.henderson@linaro.org>
Signed-off-by: Laurent Vivier <laurent@vivier.eu>
Some target are using llseek instead of _llseek like riscv,
nios2, hexagon, and openrisc.
Signed-off-by: Kito Cheng <kito.cheng@sifive.com>
Reviewed-by: Laurent Vivier <laurent@vivier.eu>
Message-Id: <20210512101358.122781-1-kito.cheng@sifive.com>
Signed-off-by: Laurent Vivier <laurent@vivier.eu>
Use bit masking instead of an if tree.
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Message-Id: <20210423165413.338259-5-richard.henderson@linaro.org>
Signed-off-by: Laurent Vivier <laurent@vivier.eu>
There is no such decoding in linux/arch/arm/nwfpe/fpmodule.c.
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Message-Id: <20210423165413.338259-4-richard.henderson@linaro.org>
Signed-off-by: Laurent Vivier <laurent@vivier.eu>
Pull out the fpa11 emulation to a helper function.
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Message-Id: <20210423165413.338259-2-richard.henderson@linaro.org>
Signed-off-by: Laurent Vivier <laurent@vivier.eu>
Syscall unshare did not have custom print function for strace, but it's argument is same as flags in clone syscall, so it can be easily implemented.
Also updated missing flags from clone_flags.
Signed-off-by: Matus Kysel <mkysel@tachyum.com>
Reviewed-by: Laurent Vivier <laurent@vivier.eu>
Message-Id: <20210406144203.1020598-1-mkysel@tachyum.com>
Signed-off-by: Laurent Vivier <laurent@vivier.eu>
We can not use watchpoints in user-mode emulation because we
need the softmmu slow path to detect accesses to watchpointed
memory. This code is expanded as empty stub in "hw/core/cpu.h"
anyway, so we can drop it.
Reviewed-by: Laurent Vivier <laurent@vivier.eu>
Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-Id: <20210303214708.1727801-18-f4bug@amsat.org>
Signed-off-by: Laurent Vivier <laurent@vivier.eu>
This patch set adds a bus multiplexer and the necessary infrastructure
in the I2C code to allow it to work.
These are common on systems with lots of I2C devices, like an IPMI BMC.
-----BEGIN PGP SIGNATURE-----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=IamZ
-----END PGP SIGNATURE-----
Merge remote-tracking branch 'remotes/cminyard/tags/for-qemu-6.1-v1' into staging
Add a bus multiplexer device
This patch set adds a bus multiplexer and the necessary infrastructure
in the I2C code to allow it to work.
These are common on systems with lots of I2C devices, like an IPMI BMC.
# gpg: Signature made Thu 13 May 2021 22:48:07 BST
# gpg: using RSA key FD0D5CE67CE0F59A6688268661F38C90919BFF81
# gpg: Good signature from "Corey Minyard <cminyard@mvista.com>" [unknown]
# gpg: aka "Corey Minyard <minyard@acm.org>" [unknown]
# gpg: aka "Corey Minyard <corey@minyard.net>" [unknown]
# gpg: aka "Corey Minyard <minyard@mvista.com>" [unknown]
# gpg: WARNING: This key is not certified with a trusted signature!
# gpg: There is no indication that the signature belongs to the owner.
# Primary key fingerprint: FD0D 5CE6 7CE0 F59A 6688 2686 61F3 8C90 919B FF81
* remotes/cminyard/tags/for-qemu-6.1-v1:
hw/i2c: add pca954x i2c-mux switch
hw/i2c: move search to i2c_scan_bus method
hw/i2c: add match method for device search
hw/i2c: name I2CNode list in I2CBus
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
The default "alabaster" sphinx theme has a couple shortcomings:
- the navbar moves along the page
- the search bar is not always at the same place
- it lacks some contrast and colours
The "rtd" theme from readthedocs.org is a popular third party theme used
notably by the kernel, with a custom style sheet. I like it better,
perhaps others do too. It also simplifies the "Edit on Gitlab" links.
Tweak a bit the custom theme to match qemu.org style, use the
QEMU logo, and favicon etc.
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Tested-by: Bin Meng <bmeng.cn@gmail.com>
Message-Id: <20210323115328.4146052-1-marcandre.lureau@redhat.com>
Reviewed-by: John Snow <jsnow@redhat.com>
Fix of the 2021-05-11 version, with a fix to build on the armhf
cross.
The largest change in this set is David's changes for ram block size
changing; then there's a pile of other cleanups and fixes.
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
-----BEGIN PGP SIGNATURE-----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=RJvy
-----END PGP SIGNATURE-----
Merge remote-tracking branch 'remotes/dgilbert/tags/pull-migration-20210513a' into staging
Migration pull 2021-05-13
Fix of the 2021-05-11 version, with a fix to build on the armhf
cross.
The largest change in this set is David's changes for ram block size
changing; then there's a pile of other cleanups and fixes.
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
# gpg: Signature made Thu 13 May 2021 18:36:06 BST
# gpg: using RSA key 45F5C71B4A0CB7FB977A9FA90516331EBC5BFDE7
# gpg: Good signature from "Dr. David Alan Gilbert (RH2) <dgilbert@redhat.com>" [full]
# Primary key fingerprint: 45F5 C71B 4A0C B7FB 977A 9FA9 0516 331E BC5B FDE7
* remotes/dgilbert/tags/pull-migration-20210513a:
tests/migration: introduce multifd into guestperf
tests/qtest/migration-test: Use g_autofree to avoid leaks on error paths
tests/migration-test: Fix "true" vs true
migration/ram: Use offset_in_ramblock() in range checks
migration/multifd: Print used_length of memory block
migration/ram: Handle RAM block resizes during postcopy
migration/ram: Simplify host page handling in ram_load_postcopy()
migration/ram: Discard RAM when growing RAM blocks after ram_postcopy_incoming_init()
exec: Relax range check in ram_block_discard_range()
migration/ram: Handle RAM block resizes during precopy
numa: Make all callbacks of ram block notifiers optional
numa: Teach ram block notifiers about resizeable ram blocks
util: vfio-helpers: Factor out and fix processing of existing ram blocks
migration: Drop redundant query-migrate result @blocked
migration/ram: Optimize ram_save_host_page()
migration/ram: Reduce unnecessary rate limiting
migrate/ram: remove "ram_bulk_stage" and "fpo_enabled"
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Guestperf tool does not cover the multifd-enabled migration
currently, it is worth supporting so that developers can
analysis the migration performance with all kinds of
migration.
To request that multifd is enabled, with 4 channels:
$ ./tests/migration/guestperf.py \
--multifd --multifd-channels 4 --output output.json
To run the entire standardized set of multifd-enabled
comparisons, with unix migration:
$ ./tests/migration/guestperf-batch.py \
--dst-host localhost --transport unix \
--filter compr-multifd* --output outputdir
Signed-off-by: Hyman Huang(黄勇) <huangy81@chinatelecom.cn>
Message-Id: <cfeeb04d17ad932c42a9871294058b77429ad1b7.1616171924.git.huangy81@chinatelecom.cn>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Coverity notices that several places in the migration-test code fail
to free memory in error-exit paths. This is pretty unimportant in
test case code, but we can avoid having to manually free the memory
entirely by using g_autofree.
The places where Coverity spotted a leak were relating to early exits
not freeing 'uri' in test_precopy_unix(), do_test_validate_uuid(),
migrate_postcopy_prepare() and test_migrate_auto_converge(). This
patch converts all the string-allocation in the test code to
g_autofree for consistency.
Fixes: Coverity CID 1432313, 1432315, 1432352, 1432364
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Message-Id: <20210506185819.9010-1-peter.maydell@linaro.org>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Accidental use of "true" as a boolean; spotted by coverity
and Peter.
Fixes: b99784ef6c
Fixes: d795f47466
Reported-by: Peter Maydell <peter.maydell@linaro.org>
Reported-by: Coverity (CID 1432373, 1432292, 1432288)
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Message-Id: <20210504100545.112213-1-dgilbert@redhat.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
We never read or write beyond the used_length of memory blocks when
migrating. Make this clearer by using offset_in_ramblock() consistently.
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20210429112708.12291-11-david@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
We actually want to print the used_length, against which we check.
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20210429112708.12291-10-david@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Resizing while migrating is dangerous and does not work as expected.
The whole migration code works with the usable_length of a ram block and
does not expect this value to change at random points in time.
In the case of postcopy, relying on used_length is racy as soon as the
guest is running. Also, when used_length changes we might leave the
uffd handler registered for some memory regions, reject valid pages
when migrating and fail when sending the recv bitmap to the source.
Resizing can be trigger *after* (but not during) a reset in
ACPI code by the guest
- hw/arm/virt-acpi-build.c:acpi_ram_update()
- hw/i386/acpi-build.c:acpi_ram_update()
Let's remember the original used_length in a separate variable and
use it in relevant postcopy code. Make sure to update it when we resize
during precopy, when synchronizing the RAM block sizes with the source.
Reviewed-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20210429112708.12291-9-david@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Add two new helper functions. This will come in come handy once we want to
handle ram block resizes while postcopy is active.
Note that ram_block_from_stream() will already print proper errors.
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20210429112708.12291-8-david@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
dgilbert: Added brackets in host_page_from_ram_block_offset
to cause uintptr_t to cast the sum, to fix armhf-cross build
In case we grow our RAM after ram_postcopy_incoming_init() (e.g., when
synchronizing the RAM block state with the migration source), the resized
part would not get discarded. Let's perform that when being notified
about a resize while postcopy has been advised, but is not listening
yet. With precopy, the process is as following:
1. VM created
- RAM blocks are created
2. Incomming migration started
- Postcopy is advised
- All pages in RAM blocks are discarded
3. Precopy starts
- RAM blocks are resized to match the size on the migration source.
- RAM pages from precopy stream are loaded
- Uffd handler is registered, postcopy starts listening
4. Guest started, postcopy running
- Pagefaults get resolved, pages get placed
Reviewed-by: Peter Xu <peterx@redhat.com>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20210429112708.12291-7-david@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
We want to make use of ram_block_discard_range() in the RAM block resize
callback when growing a RAM block, *before* used_length is changed.
Let's relax the check. As RAM blocks always mmap the whole max_length area,
we cannot corrupt unrelated data.
Reviewed-by: Peter Xu <peterx@redhat.com>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20210429112708.12291-6-david@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Resizing while migrating is dangerous and does not work as expected.
The whole migration code works on the usable_length of ram blocks and does
not expect this to change at random points in time.
In the case of precopy, the ram block size must not change on the source,
after syncing the RAM block list in ram_save_setup(), so as long as the
guest is still running on the source.
Resizing can be trigger *after* (but not during) a reset in
ACPI code by the guest
- hw/arm/virt-acpi-build.c:acpi_ram_update()
- hw/i386/acpi-build.c:acpi_ram_update()
Use the ram block notifier to get notified about resizes. Let's simply
cancel migration and indicate the reason. We'll continue running on the
source. No harm done.
Update the documentation. Postcopy will be handled separately.
Reviewed-by: Peter Xu <peterx@redhat.com>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20210429112708.12291-5-david@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Manual merge
Let's make add/remove optional. We want to introduce a RAM block
notifier for RAM migration that is only interested in resize events.
Reviewed-by: Peter Xu <peterx@redhat.com>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20210429112708.12291-4-david@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Ram block notifiers are currently not aware of resizes. To properly
handle resizes during migration, we want to teach ram block notifiers about
resizeable ram.
Introduce the basic infrastructure but keep using max_size in the
existing notifiers. Supply the max_size when adding and removing ram
blocks. Also, notify on resizes.
Acked-by: Paul Durrant <paul@xen.org>
Reviewed-by: Peter Xu <peterx@redhat.com>
Cc: xen-devel@lists.xenproject.org
Cc: haxm-team@intel.com
Cc: Paul Durrant <paul@xen.org>
Cc: Stefano Stabellini <sstabellini@kernel.org>
Cc: Anthony Perard <anthony.perard@citrix.com>
Cc: Wenchao Wang <wenchao.wang@intel.com>
Cc: Colin Xu <colin.xu@intel.com>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20210429112708.12291-3-david@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Factor it out into common code when a new notifier is registered, just
as done with the memory region notifier. This keeps logic about how to
process existing ram blocks at a central place.
Just like when adding a new ram block, we have to register the max_length.
Ram blocks are only "fake resized". All memory (max_length) is mapped.
Print the warning from inside qemu_vfio_ram_block_added().
Reviewed-by: Peter Xu <peterx@redhat.com>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20210429112708.12291-2-david@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Result @blocked is redundant. Unfortunately, we realized this too
close to the release to risk dropping it, so we deprecated it
instead, in commit e11ce6c06.
Since it was deprecated from the start, we can delete it without
the customary grace period. Do so.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Message-Id: <20210429140424.2802929-1-armbru@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Starting from pss->page, ram_save_host_page() will check every page
and send the dirty pages up to the end of the current host page or
the boundary of used_length of the block. If the host page size is
a huge page, the step "check" will take a lot of time.
It will improve performance to use migration_bitmap_find_dirty().
Tested on Kunpeng 920; VM parameters: 1U 4G (page size 1G)
The time of ram_save_host_page() in the last round of ram saving:
before optimize: 9250us after optimize: 34us
Signed-off-by: Keqian Zhu <zhukeqian1@huawei.com>
Signed-off-by: Kunkun Jiang <jiangkunkun@huawei.com>
Reviewed-by: Peter Xu <peterx@redhat.com>
Message-Id: <20210316125716.1243-3-jiangkunkun@huawei.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
When the host page is a huge page and something is sent in the
current iteration, migration_rate_limit() should be executed.
If not, it can be omitted.
Signed-off-by: Keqian Zhu <zhukeqian1@huawei.com>
Signed-off-by: Kunkun Jiang <jiangkunkun@huawei.com>
Reviewed-by: David Edmondson <david.edmondson@oracle.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Message-Id: <20210316125716.1243-2-jiangkunkun@huawei.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
The bulk stage is kind of weird: migration_bitmap_find_dirty() will
indicate a dirty page, however, ram_save_host_page() will never save it, as
migration_bitmap_clear_dirty() detects that it is not dirty.
We already fill the bitmap in ram_list_init_bitmaps() with ones, marking
everything dirty - it didn't used to be that way, which is why we needed
an explicit first bulk stage.
Let's simplify: make the bitmap the single source of thuth. Explicitly
handle the "xbzrle_enabled after first round" case.
Regarding XBZRLE (implicitly handled via "ram_bulk_stage = false" right
now), there is now a slight change in behavior:
- Colo: When starting, it will be disabled (was implicitly enabled)
until the first round actually finishes.
- Free page hinting: When starting, XBZRLE will be disabled (was implicitly
enabled) until the first round actually finished.
- Snapshots: When starting, XBZRLE will be disabled. We essentially only
do a single run, so I guess it will never actually get disabled.
Postcopy seems to indirectly disable it in ram_save_page(), so there
shouldn't be really any change.
Cc: "Michael S. Tsirkin" <mst@redhat.com>
Cc: Juan Quintela <quintela@redhat.com>
Cc: "Dr. David Alan Gilbert" <dgilbert@redhat.com>
Cc: Andrey Gruzdev <andrey.gruzdev@virtuozzo.com>
Cc: Peter Xu <peterx@redhat.com>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20210216105039.40680-1-david@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
