Greg Kurz [Wed, 12 Mar 2025 15:29:32 +0000 (16:29 +0100)]
tests/9p: Test `Tsetattr` can truncate unlinked file
Enhance the `use-after-unlink` test with a new check for the
case where the client wants to alter the size of an unlinked
file for which it still has an active fid.
Suggested-by: Christian Schoenebeck <qemu_oss@crudebyte.com> Signed-off-by: Greg Kurz <groug@kaod.org> Reviewed-by: Christian Schoenebeck <qemu_oss@crudebyte.com>
Message-Id: <20250312152933.383967-7-groug@kaod.org> Signed-off-by: Christian Schoenebeck <qemu_oss@crudebyte.com>
Add and implement functions to 9pfs test client for sending a 9p2000.L
'Tsetattr' request and receiving its 'Rsetattr' response counterpart.
Signed-off-by: Christian Schoenebeck <qemu_oss@crudebyte.com> Signed-off-by: Greg Kurz <groug@kaod.org>
Message-Id: <20250312152933.383967-6-groug@kaod.org>
Greg Kurz [Wed, 12 Mar 2025 15:29:30 +0000 (16:29 +0100)]
9pfs: Introduce futimens file op
Add an futimens operation to the fs driver and use if when a fid has
a valid file descriptor. This is required to support more cases where
the client wants to do an action on an unlinked file which it still
has an open file decriptor for.
Only 9P2000.L was considered.
Signed-off-by: Greg Kurz <groug@kaod.org> Reviewed-by: Christian Schoenebeck <qemu_oss@crudebyte.com>
Message-Id: <20250312152933.383967-5-groug@kaod.org> Signed-off-by: Christian Schoenebeck <qemu_oss@crudebyte.com>
Greg Kurz [Wed, 12 Mar 2025 15:29:29 +0000 (16:29 +0100)]
9pfs: Introduce ftruncate file op
Add an ftruncate operation to the fs driver and use if when a fid has
a valid file descriptor. This is required to support more cases where
the client wants to do an action on an unlinked file which it still
has an open file decriptor for.
Only 9P2000.L was considered.
Signed-off-by: Greg Kurz <groug@kaod.org> Reviewed-by: Christian Schoenebeck <qemu_oss@crudebyte.com>
Message-Id: <20250312152933.383967-4-groug@kaod.org> Signed-off-by: Christian Schoenebeck <qemu_oss@crudebyte.com>
Greg Kurz [Wed, 12 Mar 2025 15:29:28 +0000 (16:29 +0100)]
9pfs: Don't use file descriptors in core code
v9fs_getattr() currently peeks into V9fsFidOpenState to know if a fid
has a valid file descriptor or directory stream. Even though the fields
are accessible, this is an implementation detail of the local backend
that should not be manipulated directly by the server code.
Abstract that with a new has_valid_file_handle() backend operation.
Signed-off-by: Greg Kurz <groug@kaod.org> Reviewed-by: Christian Schoenebeck <qemu_oss@crudebyte.com>
Message-Id: <20250312152933.383967-3-groug@kaod.org> Signed-off-by: Christian Schoenebeck <qemu_oss@crudebyte.com>
Greg Kurz [Wed, 12 Mar 2025 15:29:27 +0000 (16:29 +0100)]
9pfs: local : Introduce local_fid_fd() helper
Factor out duplicated code to a single helper. More users to come.
Signed-off-by: Greg Kurz <groug@kaod.org> Reviewed-by: Christian Schoenebeck <qemu_oss@crudebyte.com>
Message-Id: <20250312152933.383967-2-groug@kaod.org> Signed-off-by: Christian Schoenebeck <qemu_oss@crudebyte.com>
9pfs: fix FD leak and reduce latency of v9fs_reclaim_fd()
This patch fixes two different bugs in v9fs_reclaim_fd():
1. Reduce latency:
This function calls v9fs_co_close() and v9fs_co_closedir() in a loop. Each
one of the calls adds two thread hops (between main thread and a fs driver
background thread). Each thread hop adds latency, which sums up in
function's loop to a significant duration.
Reduce overall latency by open coding what v9fs_co_close() and
v9fs_co_closedir() do, executing those and the loop itself altogether in
only one background thread block, hence reducing the total amount of
thread hops to only two.
2. Fix file descriptor leak:
The existing code called v9fs_co_close() and v9fs_co_closedir() to close
file descriptors. Both functions check right at the beginning if the 9p
request was cancelled:
if (v9fs_request_cancelled(pdu)) {
return -EINTR;
}
So if client sent a 'Tflush' message, v9fs_co_close() / v9fs_co_closedir()
returned without having closed the file descriptor and v9fs_reclaim_fd()
subsequently freed the FID without its file descriptor being closed, hence
leaking those file descriptors.
This 2nd bug is fixed by this patch as well by open coding v9fs_co_close()
and v9fs_co_closedir() inside of v9fs_reclaim_fd() and not performing the
v9fs_request_cancelled(pdu) check there.
Even though this function is serialized to be always called from main
thread, v9fs_reclaim_fd() is dispatching the coroutine to a worker thread
in between via its v9fs_co_*() calls, hence leading to the situation where
v9fs_reclaim_fd() is effectively executed multiple times simultaniously,
which renders its LRU algorithm useless and causes high latency.
Fix this by adding a simple boolean variable to ensure this function is
only called once at a time. No synchronization needed for this boolean
variable as this function is only entered and returned on main thread.
Fixes: 7a46274529c ('hw/9pfs: Add file descriptor reclaim support') Signed-off-by: Christian Schoenebeck <qemu_oss@crudebyte.com> Reviewed-by: Greg Kurz <groug@kaod.org>
Message-Id: <5c622067efd66dd4ee5eca740dcf263f41db20b2.1741339452.git.qemu_oss@crudebyte.com>
Steven Lee [Fri, 2 May 2025 10:34:45 +0000 (18:34 +0800)]
docs: Add support for ast2700fc machine
- Updated Aspeed family boards list to include `ast2700fc`.
- Added boot instructions for the `ast2700fc` machine.
- Detailed the configuration and loading of firmware for the
Cortex-A35 and Cortex-M4 processors.
Steven Lee [Fri, 2 May 2025 10:34:43 +0000 (18:34 +0800)]
hw/arm: Introduce ASPEED AST2700 A1 full core machine
- Added new machine type `ast2700fc` with full core support.
- Defined `Ast2700FCState` structure for the new machine type.
- Implemented initialization functions for CA35, SSP, and TSP components.
- Updated `ast2700fc_types` to include the new machine type.
- Set machine class properties for `ast2700fc`.
Test Step:
- Download ast2700-default-obmc.tar.gz from AspeedTech-BMC OpenBmc
release page.
- Run the following QEMU command:
AST2700 TSP(Tertiary Service Processor) is a Cortex-M4 coprocessor
The patch adds support for TSP with following update:
- Introduce Aspeed27x0TSPSoCState structure in aspeed_soc.h
- Implement initialization and realization functions
- Add support for UART, INTC, and SCU devices
- Map unimplemented devices for IPC and SCUIO
- Defined memory map and IRQ maps for AST27x0 A1 TSP SoC
The IRQ mapping is similar to AST2700 CA35 SoC, featuring a two-level
interrupt controller.
Difference from AST2700:
- AST2700
- Support GICINT128 to GICINT136 in INTC
- The INTCIO GIC_192_201 has 10 output pins, mapped as follows:
Bit 0 -> GIC 192
Bit 1 -> GIC 193
Bit 2 -> GIC 194
Bit 3 -> GIC 195
Bit 4 -> GIC 196
- AST2700-tsp
- Support TSPINT128 to TSPINT136 in INTC
- The INTCIO TSPINT_160_169 has 10 output pins, mapped as follows:
Bit 0 -> TSPINT 160
Bit 1 -> TSPINT 161
Bit 2 -> TSPINT 162
Bit 3 -> TSPINT 163
Bit 4 -> TSPINT 164
Signed-off-by: Steven Lee <steven_lee@aspeedtech.com>
Change-Id: I69eec2b68b26ef04187b2922c5f2e584b9076c66 Reviewed-by: Cédric Le Goater <clg@redhat.com> Link: https://lore.kernel.org/qemu-devel/20250502103449.3091642-7-steven_lee@aspeedtech.com
[ clg: removed local 'Error* err' in aspeed_soc_ast27x0tsp_realize() ] Signed-off-by: Cédric Le Goater <clg@redhat.com>
The AST2700 SSP (Secondary Service Processor) is a Cortex-M4 coprocessor.
This patch adds support for A1 SSP with the following updates:
- Introduce Aspeed27x0SSPSoCState structure in aspeed_soc.h
- Define memory map and IRQ map for AST27x0 A1 SSP SoC
- Implement initialization and realization functions
- Add support for UART, INTC, and SCU devices
- Map unimplemented devices for IPC and SCUIO
The IRQ mapping is similar to AST2700 CA35 SoC, featuring a two-level
interrupt controller.
Difference from AST2700:
- AST2700
- Support GICINT128 to GICINT136 in INTC
- The INTCIO GIC_192_201 has 10 output pins, mapped as follows:
Bit 0 -> GIC 192
Bit 1 -> GIC 193
Bit 2 -> GIC 194
Bit 3 -> GIC 195
Bit 4 -> GIC 196
- AST2700-ssp
- Support SSPINT128 to SSPINT136 in INTC
- The INTCIO SSPINT_160_169 has 10 output pins, mapped as follows:
Bit 0 -> SSPINT 160
Bit 1 -> SSPINT 161
Bit 2 -> SSPINT 162
Bit 3 -> SSPINT 163
Bit 4 -> SSPINT 164
Signed-off-by: Steven Lee <steven_lee@aspeedtech.com>
Change-Id: I924bf1a657f1e83f9e16d6673713f4a06ecdb496 Reviewed-by: Cédric Le Goater <clg@redhat.com> Link: https://lore.kernel.org/qemu-devel/20250502103449.3091642-6-steven_lee@aspeedtech.com
[ clg: removed local 'Error* err' in aspeed_soc_ast27x0ssp_realize() ] Signed-off-by: Cédric Le Goater <clg@redhat.com>
Steven Lee [Fri, 2 May 2025 10:34:40 +0000 (18:34 +0800)]
hw/intc/aspeed: Add support for AST2700 TSP INTC
- Define new types for ast2700tsp INTC and INTCIO
- Add register definitions for TSP INTC and INTCIO
- Implement write handlers for TSP INTC and INTCIO
- Register new types in aspeed_intc_register_types
The design of the TSP INTC and INTCIO controllers is similar to
AST2700, with the following differences:
- AST2700
Support GICINT128 to GICINT136 in INTC
The INTCIO GIC_192_201 has 10 output pins, mapped as follows:
Bit 0 -> GIC 192
Bit 1 -> GIC 193
Bit 2 -> GIC 194
Bit 3 -> GIC 195
Bit 4 -> GIC 196
- AST2700-tsp
Support TSPINT128 to TSPINT136 in INTC
The INTCIO TSPINT_160_169 has 10 output pins, mapped as follows:
Bit 0 -> TSPINT 160
Bit 1 -> TSPINT 161
Bit 2 -> TSPINT 162
Bit 3 -> TSPINT 163
Bit 4 -> TSPINT 164
Steven Lee [Fri, 2 May 2025 10:34:39 +0000 (18:34 +0800)]
hw/intc/aspeed: Add support for AST2700 SSP INTC
- Define new types for ast2700ssp INTC and INTCIO
- Add register definitions for SSP INTC and INTCIO
- Implement write handlers for SSP INTC and INTCIO
- Register new types in aspeed_intc_register_types
The design of the SSP INTC and INTCIO controllers is similar to
AST2700, with the following differences:
- AST2700
Support GICINT128 to GICINT136 in INTC
The INTCIO GIC_192_201 has 10 output pins, mapped as follows:
Bit 0 -> GIC 192
Bit 1 -> GIC 193
Bit 2 -> GIC 194
Bit 3 -> GIC 195
Bit 4 -> GIC 196
- AST2700-ssp
Support SSPINT128 to SSPINT136 in INTC
The INTCIO SSPINT_160_169 has 10 output pins, mapped as follows:
Bit 0 -> SSPINT 160
Bit 1 -> SSPINT 161
Bit 2 -> SSPINT 162
Bit 3 -> SSPINT 163
Bit 4 -> SSPINT 164
Steven Lee [Fri, 2 May 2025 10:34:37 +0000 (18:34 +0800)]
aspeed: ast27x0: Map unimplemented devices in SoC memory
Maps following unimplemented devices in SoC memory
- dpmcu
- iomem
- iomem0
- iomem1
- ltpi
Iomem, Iomem0 and Iomem1 include unimplemented controllers in the memory ranges 0x0 - 0x1000000, 0x120000000 - 0x121000000 and
0x14000000 - 0x141000000.
For instance:
- USB hub at 0x12010000
- eSPI at 0x14C5000
- PWM at 0x140C0000
DPMCU stands for Display Port MCU controller. LTPI is used to connect to AST1700.
AST1700 is an I/O expander that supports the DC-SCM 2.1 LTPI protocol.
It provides AST2700 with additional GPIO, UART, I3C, and other interfaces.
Jamin Lin [Thu, 24 Apr 2025 07:51:33 +0000 (15:51 +0800)]
docs/system/arm/aspeed: move AST2700 content to new section
Moved AST2700-related content from the general Aspeed board list into a
dedicated section for Aspeed 2700 family boards. Improves clarity and
readability.
Jamin Lin [Thu, 24 Apr 2025 07:51:32 +0000 (15:51 +0800)]
tests/functional/aspeed: Add to test vbootrom for AST2700
Add the AST2700 functional test to boot using the vbootrom image
instead of manually loading boot components with -device loader.
The boot ROM binary is now passed via the
-bios option, using the image located in pc-bios/ast27x0_bootrom.bin.
Jamin Lin [Thu, 24 Apr 2025 07:51:29 +0000 (15:51 +0800)]
hw/arm/aspeed_ast27x0 Introduce vbootrom memory region
Introduce a new vbootrom memory region. The region is mapped at address
"0x00000000" and has a size of 128KB, identical to the SRAM region size.
This memory region is intended for loading a vbootrom image file as part of the
boot process.
The vbootrom registered in the SoC's address space using the ASPEED_DEV_VBOOTROM
index.
Signed-off-by: Jamin Lin <jamin_lin@aspeedtech.com> Reviewed-by: Nabih Estefan <nabihestefan@google.com> Tested-by: Nabih Estefan <nabihestefan@google.com> Reviewed-by: Cédric Le Goater <clg@redhat.com> Link: https://lore.kernel.org/qemu-devel/20250424075135.3715128-2-jamin_lin@aspeedtech.com Signed-off-by: Cédric Le Goater <clg@redhat.com>
Jamin Lin [Wed, 23 Apr 2025 07:23:43 +0000 (15:23 +0800)]
tests/functional/aspeed: Move I2C test into shared helper for AST2700 reuse
Move the I2C test case into a common helper function (do_ast2700_i2c_test) so it
can be reused across multiple AST2700-based test cases. This reduces duplication
and improves maintainability.
Jamin Lin [Wed, 23 Apr 2025 07:23:37 +0000 (15:23 +0800)]
hw/arm/aspeed_ast27x0: Rename variable sram_name to name in ast2700 realize
The variable "sram_name" was only used for naming the SRAM memory region.
Rename it to "name" for consistency with similar code and avoid unnecessary
new local variable declarations.
cde3247651dc998da5dc1005148302a90d72f21f fixed atomicity for LDRD, which
ends up making accesses 64-bits wide. However, the AST2600 bootloader
can sometimes compile with LDRD instructions, which causes the acceses
to fail when accessing the memory-mapped SPI flash.
To fix this, increase the MMIO region valid access size to allow for
64-bit accesses.
Troy Lee [Mon, 17 Mar 2025 06:59:38 +0000 (14:59 +0800)]
hw/arm: ast27x0: Wire up EHCI controllers
AST27x0 has 4 EHCI controllers, where each CPU and I/O die has 2
instances. This patch use existing TYPE_PLATFORM_EHCI. After wiring up
the EHCI controller, the ast2700a1-evb can find up to 4 USB EHCI
interfaces.
ehci-platform 12061000.usb: EHCI Host Controller
ehci-platform 12061000.usb: new USB bus registered, assigned bus number 2
ehci-platform 12063000.usb: EHCI Host Controller
ehci-platform 12063000.usb: new USB bus registered, assigned bus number 3
ehci-platform 12061000.usb: irq 88, io mem 0x12061000
ehci-platform 12063000.usb: irq 90, io mem 0x12063000
ehci-platform 14121000.usb: EHCI Host Controller
ehci-platform 14123000.usb: EHCI Host Controller
ehci-platform 12061000.usb: USB 2.0 started, EHCI 1.00
ehci-platform 14121000.usb: new USB bus registered, assigned bus number 5
ehci-platform 14123000.usb: new USB bus registered, assigned bus number 6
ehci-platform 14121000.usb: irq 91, io mem 0x14121000
ehci-platform 14123000.usb: irq 92, io mem 0x14123000
ehci-platform 12063000.usb: USB 2.0 started, EHCI 1.00
usb usb2: Manufacturer: Linux 6.6.78-dirty-bafd2830c17c-gbafd2830c17c-dirty ehci_hcd
usb usb3: Manufacturer: Linux 6.6.78-dirty-bafd2830c17c-gbafd2830c17c-dirty ehci_hcd
ehci-platform 14121000.usb: USB 2.0 started, EHCI 1.00
usb usb5: Manufacturer: Linux 6.6.78-dirty-bafd2830c17c-gbafd2830c17c-dirty ehci_hcd
ehci-platform 14123000.usb: USB 2.0 started, EHCI 1.00
usb usb6: Manufacturer: Linux 6.6.78-dirty-bafd2830c17c-gbafd2830c17c-dirty ehci_hcd
Note that, AST27x0A0 only has 2 EHCI controllers due to hw issue.
monitor: don't wake up qmp_dispatcher_co coroutine upon cleanup
Since the commit 3e6bed61 ("monitor: cleanup detection of qmp_dispatcher_co
shutting down"), coroutine pointer qmp_dispatcher_co is set to NULL upon
cleanup. If a QMP command is sent after monitor_cleanup() (e.g. after
shutdown), this may lead to SEGFAULT on aio_co_wake(NULL).
As mentioned in the comment inside monitor_cleanup(), the intention is to
allow incoming requests while shutting down, but simply leave them
without any response. Let's do exactly that, and if qmp_dispatcher_co
coroutine pointer has already been set to NULL, let's simply skip the
aio_co_wake() part.
This property was added to preserve previous value when this was fixed
in version 2.1 but the last machine using it was already removed when
adding diva-gsp leaving this property unused and unnecessary.
Signed-off-by: BALATON Zoltan <balaton@eik.bme.hu> Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Helge Deller <deller@gmx.de> Link: https://lore.kernel.org/r/20250502095524.DE1F355D264@zero.eik.bme.hu Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Paolo Bonzini [Mon, 15 Jul 2024 08:35:06 +0000 (10:35 +0200)]
target/i386: do not block singlestep for STI
STI will trigger a singlestep exception even if it has inhibit-IRQ
behavior. Do not suppress single-step for all IRQ-inhibiting
instructions, instead special case MOV SS and POP SS.
Cc: qemu-stable@nongnu.org Fixes: f0f0136abba ("target/i386: no single-step exception after MOV or POP SS", 2024-05-25) Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Magnus Kulke [Tue, 29 Apr 2025 09:33:19 +0000 (11:33 +0200)]
target/i386/emulate: remove rflags leftovers
Fixes: c901905ea670 ("target/i386/emulate: remove flags_mask")
In c901905ea670 rflags have been removed from `x86_decode`, but there
were some leftovers.
Zhao Liu [Mon, 14 Apr 2025 14:49:42 +0000 (22:49 +0800)]
rust/hpet: Support migration
Based on commit 1433e38cc8 ("hpet: do not overwrite properties on
post_load"), add the basic migration support to Rust HPET.
The current migration implementation introduces multiple unsafe
callbacks. Before the vmstate builder, one possible cleanup approach is
to wrap callbacks in the vmstate binding using a method similar to the
vmstate_exist_fn macro.
However, this approach would also create a lot of repetitive code (since
vmstate has so many callbacks: pre_load, post_load, pre_save, post_save,
needed and dev_unplug_pending). Although it would be cleaner, it would
somewhat deviate from the path of the vmstate builder.
Therefore, firstly focus on completing the functionality of HPET, and
those current unsafe callbacks can at least clearly indicate the needed
functionality of vmstate. The next step is to consider refactoring
vmstate to move towards the vmstate builder direction.
Additionally, update rust.rst about Rust HPET can support migration.
Paolo Bonzini [Fri, 2 May 2025 08:56:06 +0000 (10:56 +0200)]
rust: assertions: Support index field wrapped in BqlCell
Currently, if the `num` field of a varray is not a numeric type, such as
being placed in a wrapper, the array variant of assert_field_type will
fail the check.
HPET currently wraps num_timers in BqlCell<>. Although BqlCell<> is not
necessary from strictly speaking, it makes sense for vmstate to respect
BqlCell.
The failure of assert_field_type is because it cannot convert BqlCell<T>
into usize for use as the index. Use a constant 0 instead for the index,
by avoiding $(...)? and extracting the common parts of
assert_field_type! into an internal case.
Commit message based on a patch by Zhao Liu <zhao1.liu@intel.com>.
Zhao Liu [Mon, 14 Apr 2025 14:49:35 +0000 (22:49 +0800)]
rust/vmstate: Add support for field_exists checks
Unfortunately, at present it's not possible to have a const
"with_exist_check" method to append test_fn after vmstate_struct (due
to error on "constant functions cannot evaluate destructors" for `F`).
Before the vmstate builder, the only way to support "test_fn" is to
extend vmstate_struct macro to add the such new optional member (and
fortunately, Rust can still parse the current expansion!).
Abstract the previous callback implementation of vmstate_validate into
a separate macro, and moves it before vmstate_struct for vmstate_struct
to call.
Note that there's no need to add any extra flag for a new test_fn added
in the VMStateField.
Thomas Huth [Tue, 29 Apr 2025 15:21:39 +0000 (17:21 +0200)]
scripts/vmstate-static-checker.py: Allow new name for ghes_addr_le field
ghes_addr_le has been renamed to hw_error_le in commit 652f6d86cbb
("acpi/ghes: better name the offset of the hardware error firmware").
Adjust the checker script to allow that changed field name.
I hit following error which testing migration in pure RoCE env:
"-incoming rdma:[::]:8089: RDMA ERROR: You only have RoCE / iWARP devices in your
systems and your management software has specified '[::]', but IPv6 over RoCE /
iWARP is not supported in Linux.#012'."
In our setup, we use rdma bind on ipv6 on target host, while connect from source
with ipv4, remove the qemu_rdma_broken_ipv6_kernel, migration just work
fine.
Checking the git history, the function was added since introducing of
rdma migration, which is more than 10 years ago. linux-rdma has
improved support on RoCE/iWARP for ipv6 over past years. There are a few fixes
back in 2016 seems related to the issue, eg: aeb76df46d11 ("IB/core: Set routable RoCE gid type for ipv4/ipv6 networks")
other fixes back in 2018, eg: 052eac6eeb56 RDMA/cma: Update RoCE multicast routines to use net namespace 8d20a1f0ecd5 RDMA/cma: Fix rdma_cm raw IB path setting for RoCE 9327c7afdce3 RDMA/cma: Provide a function to set RoCE path record L2 parameters 5c181bda77f4 RDMA/cma: Set default GID type as RoCE when resolving RoCE route 3c7f67d1880d IB/cma: Fix default RoCE type setting be1d325a3358 IB/core: Set RoCEv2 MGID according to spec 63a5f483af0e IB/cma: Set default gid type to RoCEv2
So remove the outdated function and it's usage.
Cc: Peter Xu <peterx@redhat.com> Cc: Li Zhijian <lizhijian@fujitsu.com> Cc: Yu Zhang <yu.zhang@ionos.com> Cc: Fabiano Rosas <farosas@suse.de> Cc: qemu-devel@nongnu.org Cc: linux-rdma@vger.kernel.org Cc: michael@flatgalaxy.com Signed-off-by: Jack Wang <jinpu.wang@ionos.com> Tested-by: Li zhijian <lizhijian@fujitsu.com> Reviewed-by: Michael Galaxy <mrgalaxy@nvidia.com> Link: https://lore.kernel.org/r/20250402051306.6509-1-jinpu.wang@ionos.com
[peterx: some cosmetic changes] Signed-off-by: Peter Xu <peterx@redhat.com>
Peter Xu [Thu, 24 Apr 2025 22:07:05 +0000 (18:07 -0400)]
migration/postcopy: Spatial locality page hint for preempt mode
The preempt mode postcopy has been introduced for a while. From latency
POV, it should always win the vanilla postcopy.
However there's one thing missing when preempt mode is enabled right now,
which is the spatial locality hint when there're page requests from the
destination side.
In vanilla postcopy, as long as a page request was unqueued, it will update
the PSS of the precopy background stream, so that after a page request the
background thread will move the pages after whatever was requested. It's
pretty much a natural behavior when there's only one channel anyway, and
one scanner to send the pages.
Preempt mode didn't follow that, because preempt mode has its own channel
and its own PSS (which doesn't linearly scan the guest memory, but
dedicated to resolve page requested from destination). So the page request
process and the background migration process are completely separate.
This patch adds the hint explicitly for preempt mode. With that, whenever
the preempt mode receives a page request on the source, it will service the
remote page fault in the return path, then it'll provide a hint to the
background thread so that we'll start sending the pages right after the
requested ones in the background, assuming the follow up pages have a
higher chance to be accessed later.
NOTE: since the background migration thread and return path thread run
completely concurrently, it doesn't always mean the hint will be applied
every single time. For example, it's possible that the return path thread
receives multiple page requests in a row without the background thread
getting the chance to consume one. In such case, the preempt thread only
provide the hint if the previous hint has been consumed. After all,
there's no point queuing hints when we only have one linear scanner.
This could measureably improve the simple sequential memory access pattern
during postcopy (when preempt is on). For random accesses, I can measure a
slight increase of remote page fault latency from ~500us -> ~600us, that
could be a trade-off to have such hint mechanism, and after all that's
still greatly improved comparing to vanilla postcopy on random (~10ms).
The patch is verified by our QE team in a video streaming test case, to
reduce the pause of the video from ~1min to a few seconds when switching
over to postcopy with preempt mode.
tests/qtest/migration: consolidate set capabilities
Migration capabilities are set in multiple '.start_hook'
functions for various tests. Instead, consolidate setting
capabilities in 'migrate_start_set_capabilities()' function
which is called from the 'migrate_start()' function.
While simplifying the capabilities setting, it helps
to declutter the qtest sources.
Add a savevm handler for a module to opt-in sending extra sections right
before postcopy starts, and before VM is stopped.
RAM will start to use this new savevm handler in the next patch to do flush
and sync for multifd pages.
Note that we choose to do it before VM stopped because the current only
potential user is not sensitive to VM status, so doing it before VM is
stopped is preferred to enlarge any postcopy downtime.
It is still a bit unfortunate that we need to introduce such a new savevm
handler just for the only use case, however it's so far the cleanest.
The various logical migration channels don't have a
standardized way of advertising themselves and their
connections may be seen out of order by the migration
destination. When a new connection arrives, the incoming
migration currently make use of heuristics to determine
which channel it belongs to.
The next few patches will need to change how the multifd
and postcopy capabilities interact and that affects the
channel discovery heuristic.
Refactor the channel discovery heuristic to make it less
opaque and simplify the subsequent patches.
Li Zhijian [Tue, 11 Mar 2025 02:42:21 +0000 (10:42 +0800)]
migration: Add qtest for migration over RDMA
This qtest requires there is a RDMA(RoCE) link in the host.
In order to make the test work smoothly, introduce a
scripts/rdma-migration-helper.sh to detect existing RoCE link before
running the test.
Test will be skipped if there is no available RoCE link.
# Start of rdma tests
# Running /x86_64/migration/precopy/rdma/plain
ok 1 /x86_64/migration/precopy/rdma/plain # SKIP No rdma link available
# To enable the test:
# Run 'scripts/rdma-migration-helper.sh setup' with root to setup a new rdma/rxe link and rerun the test
# Optional: run 'scripts/rdma-migration-helper.sh clean' to revert the 'setup'
# End of rdma tests
Cc: Philippe Mathieu-Daudé <philmd@linaro.org> Cc: Stefan Hajnoczi <stefanha@gmail.com> Reviewed-by: Peter Xu <peterx@redhat.com> Signed-off-by: Li Zhijian <lizhijian@fujitsu.com>
Message-ID: <20250311024221.363421-1-lizhijian@fujitsu.com>
[add 'head -1' to script, reformat test message] Signed-off-by: Fabiano Rosas <farosas@suse.de>
Li Zhijian [Wed, 5 Mar 2025 06:28:24 +0000 (14:28 +0800)]
migration: Unfold control_save_page()
control_save_page() is for RDMA only, unfold it to make the code more
clear.
In addition:
- Similar to other branches style in ram_save_target_page(), involve RDMA
only if the condition 'migrate_rdma()' is true.
- Further simplify the code by removing the RAM_SAVE_CONTROL_NOT_SUPP.
Reviewed-by: Peter Xu <peterx@redhat.com> Signed-off-by: Li Zhijian <lizhijian@fujitsu.com>
Message-ID: <20250305062825.772629-6-lizhijian@fujitsu.com> Signed-off-by: Fabiano Rosas <farosas@suse.de>
Li Zhijian [Wed, 5 Mar 2025 06:28:21 +0000 (14:28 +0800)]
migration: check RDMA and capabilities are compatible on both sides
Depending on the order of starting RDMA and setting capability,
they can be categorized into the following scenarios:
Source:
S1: [set capabilities] -> [Start RDMA outgoing]
Destination:
D1: [set capabilities] -> [Start RDMA incoming]
D2: [Start RDMA incoming] -> [set capabilities]
Previously, compatibility between RDMA and capabilities was verified only
in scenario D1, potentially causing migration failures in other situations.
For scenarios S1 and D1, we can seamlessly incorporate
migration_transport_compatible() to address compatibility between
channels and capabilities vs transport.
For scenario D2, ensure compatibility within migrate_caps_check().
Reviewed-by: Peter Xu <peterx@redhat.com> Signed-off-by: Li Zhijian <lizhijian@fujitsu.com>
Message-ID: <20250305062825.772629-3-lizhijian@fujitsu.com> Signed-off-by: Fabiano Rosas <farosas@suse.de>
Jamin Lin [Tue, 29 Apr 2025 06:28:20 +0000 (14:28 +0800)]
pc-bios: Add AST27x0 vBootrom
The boot ROM is a minimal implementation designed to load an AST27x0 boot image.
Its source code is available at:
https://github.com/google/vbootrom
Commit id: d6e3386709b3e49322a94ffadc2aaab9944ab77b
Build Information:
```
Build Date : Apr 29 2025 01:23:18
FW Version : git-d6e3386
```
Signed-off-by: Jamin Lin <jamin_lin@aspeedtech.com> Reviewed-by: Nabih Estefan <nabihestefan@google.com> Reviewed-by: Cédric Le Goater <clg@redhat.com> Tested-by: Nabih Estefan <nabihestefan@google.com> Link: https://lore.kernel.org/qemu-devel/20250429062822.1184920-2-jamin_lin@aspeedtech.com Signed-off-by: Cédric Le Goater <clg@redhat.com>
For change, pause, resume, complete, dismiss and finalize actions
corresponding job- and block-job commands are almost equal. The
difference is in find_block_job_locked() vs find_job_locked()
functions. What's different?
1. find_block_job_locked() checks whether the found job is a block-job.
This is OK when moving to more generic API, no needs to document this
change.
2. find_block_job_locked() reports DeviceNotActive on failure, when
find_job_locked() reports GenericError. So, let's document this
difference in deprecated.txt. Still, for dismiss and finalize errors
are not documented at all, so be silent in deprecated.txt as well.
ACKed-by: Peter Krempa <pkrempa@redhat.com> Reviewed-by: Eric Blake <eblake@redhat.com> Reviewed-by: Markus Armbruster <armbru@redhat.com> Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
Message-ID: <20250409084232.28201-3-vsementsov@yandex-team.ru>
qapi: synchronize jobs and block-jobs documentation
Actualize documentation and synchronize it for commands which actually
call the same functions internally.
Reviewed-by: Markus Armbruster <armbru@redhat.com> Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
Message-ID: <20250409084232.28201-2-vsementsov@yandex-team.ru>
Vincent Vanlaer [Sat, 26 Oct 2024 16:30:09 +0000 (18:30 +0200)]
block: add test non-active commit with zeroed data
Signed-off-by: Vincent Vanlaer <libvirt-e6954efa@volkihar.be> Tested-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru> Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
Message-Id: <20241026163010.2865002-6-libvirt-e6954efa@volkihar.be> Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
Vincent Vanlaer [Sat, 26 Oct 2024 16:30:08 +0000 (18:30 +0200)]
block: allow commit to unmap zero blocks
Non-active block commits do not discard blocks only containing zeros,
causing images to lose sparseness after the commit. This commit fixes
that by writing zero blocks using blk_co_pwrite_zeroes rather than
writing them out as any other arbitrary data.
Signed-off-by: Vincent Vanlaer <libvirt-e6954efa@volkihar.be> Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
Message-Id: <20241026163010.2865002-5-libvirt-e6954efa@volkihar.be> Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
Vincent Vanlaer [Sat, 26 Oct 2024 16:30:07 +0000 (18:30 +0200)]
block: refactor error handling of commit_iteration
Signed-off-by: Vincent Vanlaer <libvirt-e6954efa@volkihar.be>
Message-Id: <20241026163010.2865002-4-libvirt-e6954efa@volkihar.be>
[vsementsov]: move action declaration to the top of the function Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
accel/tcg: Return TCGTBCPUState from cpu_get_tb_cpu_state
Combine 3 different pointer returns into one structure return.
Include a cflags field in TCGTBCPUState, not filled in by
cpu_get_tb_cpu_state, but used by all callers. This fills
a hole in the structure and is useful in some subroutines.
Reviewed-by: Pierrick Bouvier <pierrick.bouvier@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
accel/tcg: Unconditionally use CPU_DUMP_CCOP in log_cpu_exec
This flag is only tested by target/i386, so including this
makes no functional change. This is similar to other places
like cpu-target.c which use CPU_DUMP_CCOP unconditionally.
Reviewed-by: Pierrick Bouvier <pierrick.bouvier@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
"exec/exec-all.h" is now fully empty, let's remove it.
Mechanical change running:
$ sed -i '/exec\/exec-all.h/d' $(git grep -wl exec/exec-all.h)
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Mark Cave-Ayland <mark.caveayland@nutanix.com> Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-ID: <20250424202412.91612-14-philmd@linaro.org>
accel/tcg: Extract probe API out of 'exec/exec-all.h'
Declare probe methods in "accel/tcg/probe.h" to emphasize
they are specific to TCG accelerator.
Suggested-by: Richard Henderson <richard.henderson@linaro.org> Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Mark Cave-Ayland <mark.caveayland@nutanix.com> Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-ID: <20250424202412.91612-13-philmd@linaro.org>
Restrict iotlb_to_section(), address_space_translate_for_iotlb()
and memory_region_section_get_iotlb() to TCG. Declare them in
the new "accel/tcg/iommu.h" header. Declare iotlb_to_section()
using the MemoryRegionSection typedef.
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-ID: <20250424202412.91612-12-philmd@linaro.org>
The next commit will restrict TCG specific code in physmem.c
using some #ifdef'ry. In order to keep it simple, move
iotlb_to_section() and memory_region_section_get_iotlb()
around close together.
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Mark Cave-Ayland <mark.caveayland@nutanix.com> Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-ID: <20250424202412.91612-11-philmd@linaro.org>
accel/tcg: Include 'accel/tcg/getpc.h' in 'exec/helper-proto'
Most files including "exec/helper-proto.h" call GETPC().
Include it there (in the common part) instead of the
unspecific "exec/exec-all.h" header.
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-ID: <20250424202412.91612-10-philmd@linaro.org>
target/riscv: Include missing 'accel/tcg/getpc.h' in csr.c
"accel/tcg/getpc.h" is pulled in indirectly. Include it
explicitly to avoid when refactoring unrelated headers:
target/riscv/csr.c:2117:25: error: call to undeclared function 'GETPC' [-Wimplicit-function-declaration]
2117 | if ((val & RVC) && (GETPC() & ~3) != 0) {
| ^
Note the TODO comment around GETPC() added upon introduction in
commit f18637cd611 ("RISC-V: Add misa runtime write support"):
2099 static RISCVException write_misa(CPURISCVState *env, int csrno,
2100 target_ulong val)
2101 {
...
2113 /*
2114 * Suppress 'C' if next instruction is not aligned
2115 * TODO: this should check next_pc
2116 */
2117 if ((val & RVC) && (GETPC() & ~3) != 0) {
2118 val &= ~RVC;
2119 }
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Mark Cave-Ayland <mark.caveayland@nutanix.com> Acked-by: Alistair Francis <alistair.francis@wdc.com> Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-ID: <20250424202412.91612-8-philmd@linaro.org>
Reviewed-by: Anton Johansson <anjo@rev.ng> Reviewed-by: Pierrick Bouvier <pierrick.bouvier@linaro.org> Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
include/exec: Move tb_invalidate_phys_range to translation-block.h
Reviewed-by: Anton Johansson <anjo@rev.ng> Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Pierrick Bouvier <pierrick.bouvier@linaro.org> Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
vaddr is defined in "exec/vaddr.h", hwaddr in "exec/hwaddr.h"
and MemTxAttrs in "exec/memattrs.h". All these headers are
indirectly pulled in via "exec/translation-block.h". Since
we will remove "exec/translation-block.h" in the next commit,
include the missing ones, otherwise we'd get errors such:
include/exec/exec-all.h:51:1: error: unknown type name 'hwaddr'
51 | hwaddr memory_region_section_get_iotlb(CPUState *cpu,
| ^
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Mark Cave-Ayland <mark.caveayland@nutanix.com> Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-ID: <20250424202412.91612-5-philmd@linaro.org>
Reviewed-by: Anton Johansson <anjo@rev.ng> Reviewed-by: Pierrick Bouvier <pierrick.bouvier@linaro.org> Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
projects / thirdparty / qemu.git / log
