Not stopping them leave the device in a bad state when virtio-net
fronted device is unplugged with device_del monitor command.
This is not triggable in regular poweroff or qemu forces shutdown
because cleanup is called right after vhost_vdpa_dev_start(false). But
devices hot unplug does not call vdpa device cleanups. This lead to all
the vhost_vdpa devices without stop the SVQ but the last.
Fix it and clean the code, making it symmetric with
vhost_vdpa_svqs_start.
Fixes: dff4426fa6 ("vhost: Add Shadow VirtQueue kick forwarding capabilities")
Reported-by: Lei Yang <leiyang@redhat.com>
Signed-off-by: Eugenio Pérez <eperezma@redhat.com>
Message-Id: <20230209170004.899472-1-eperezma@redhat.com>
Tested-by: Laurent Vivier <lvivier@redhat.com>
Acked-by: Jason Wang <jasowang@redhat.com>
The expiration time provided for timer_mod() can overflow if a
ridiculously large value is set to the comparator register. The
resulting value can represent a past time after rounded, forcing the
timer to fire immediately. If the timer is configured as periodic, it
will rearm the timer again, and form an endless loop.
Check if the expiration value will overflow, and if it will, stop the
timer instead of rearming the timer with the overflowed time.
This bug was found by Alexander Bulekov when fuzzing igb, a new
network device emulation:
https://patchew.org/QEMU/20230129053316.1071513-1-alxndr@bu.edu/
The fixed test case is:
fuzz/crash_2d7036941dcda1ad4380bb8a9174ed0c949bcefd
Fixes: 16b29ae180 ("Add HPET emulation to qemu (Beth Kon)")
Signed-off-by: Akihiko Odaki <akihiko.odaki@daynix.com>
Acked-by: Michael S. Tsirkin <mst@redhat.com>
Message-Id: <20230131030037.18856-1-akihiko.odaki@daynix.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
vhost_dev_cleanup(), called from vu_gpio_disconnect(), clears vhost_dev
so vhost-user-gpio must set the members of vhost_dev each time
connecting.
do_vhost_user_cleanup() should also acquire the pointer to vqs directly
from VHostUserGPIO instead of referring to vhost_dev as it can be called
after vhost_dev_cleanup().
Fixes: 27ba7b027f ("hw/virtio: add boilerplate for vhost-user-gpio device")
Signed-off-by: Akihiko Odaki <akihiko.odaki@daynix.com>
Message-Id: <20230130140320.77999-1-akihiko.odaki@daynix.com>
Reviewed-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
While ioeventfds are needed for good performance with KVM guests it
should not be a gating requirement. We can run vhost-user backends using
simulated ioeventfds or inband signalling.
With this change I can run:
$QEMU $OPTS \
-display gtk,gl=on \
-device vhost-user-gpu-pci,chardev=vhgpu \
-chardev socket,id=vhgpu,path=vhgpu.sock
with:
./contrib/vhost-user-gpu/vhost-user-gpu \
-s vhgpu.sock \
-v
and at least see things start-up - although the display gets rotated by
180 degrees. Once lightdm takes over we never make it to the login
prompt and just get a blank screen.
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Cc: Gerd Hoffmann <kraxel@redhat.com>
Message-Id: <20221202132231.1048669-1-alex.bennee@linaro.org>
Message-Id: <20230130124728.175610-1-alex.bennee@linaro.org>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Since GUEST_ANNOUNCE is emulated the feature bit could be set without
backend support. This happens in the vDPA case.
However, backend vDPA parent may not have CVQ support. This causes an
incoherent feature set, and the driver may refuse to start. This
happens in virtio-net Linux driver.
This may be solved differently in the future. Qemu is able to emulate a
CVQ just for guest_announce purposes, helping guest to notify the new
location with vDPA devices that does not support it. However, this is
left as a TODO as it is way more complex to backport.
Tested with vdpa_net_sim, toggling manually VIRTIO_NET_F_CTRL_VQ in the
driver and migrating it with x-svq=on.
Fixes: 980003debd ("vdpa: do not handle VIRTIO_NET_F_GUEST_ANNOUNCE in vhost-vdpa")
Reported-by: Dawar, Gautam <gautam.dawar@amd.com>
Signed-off-by: Eugenio Pérez <eperezma@redhat.com>
Message-Id: <20230124161159.2182117-1-eperezma@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: David Edmondson <david.edmondson@oracle.com>
Reviewed-by: Gautam Dawar <gautam.dawar@amd.com>
Tested-by: Gautam Dawar <gautam.dawar@amd.com>
Tested-by: Lei Yang <leiyang@redhat.com>
This reverts commit 67f7e426e5.
Additionally to the automatic revert, I went over the code
and dropped all mentions of legacy_no_rng_seed manually,
effectively reverting a combination of 2 additional commits:
commit ffe2d2382e
Author: Jason A. Donenfeld <Jason@zx2c4.com>
Date: Wed Sep 21 11:31:34 2022 +0200
x86: re-enable rng seeding via SetupData
commit 3824e25db1
Author: Gerd Hoffmann <kraxel@redhat.com>
Date: Wed Aug 17 10:39:40 2022 +0200
x86: disable rng seeding via setup_data
Fixes: 67f7e426e5 ("hw/i386: pass RNG seed via setup_data entry")
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Tested-by: Nathan Chancellor <nathan@kernel.org>
Tested-by: Dov Murik <dovmurik@linux.ibm.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Since commit a0e61807a3 ("qapi: Remove QMP events and commands from
user-mode builds") we don't generate the "qapi-commands-machine.h"
header in a user-emulation-only build.
Move the QMP functions from cpu_init.c (which is always compiled)
to monitor.c (which is only compiled when system-emulation
is selected). Rename monitor.c to arm-qmp-cmds.c.
Note ppc_cpu_class_by_name() is used by both file units, so we
expose its prototype in "cpu-qom.h".
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Cédric Le Goater <clg@kaod.org>
Message-Id: <20230223155540.30370-5-philmd@linaro.org>
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Since commit a0e61807a3 ("qapi: Remove QMP events and commands from
user-mode builds") we don't generate the "qapi-commands-machine.h"
header in a user-emulation-only build.
Extract the QMP functions from cpu.c (which is always compiled)
to the new 'loongarch-qmp-cmds.c' unit (which is only compiled
when system emulation is selected).
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Message-Id: <20230223155540.30370-4-philmd@linaro.org>
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Since commit a0e61807a3 ("qapi: Remove QMP events and commands from
user-mode builds") we don't generate the "qapi-commands-machine.h"
header in a user-emulation-only build.
Guard qmp_query_cpu_definitions() within CONFIG_USER_ONLY; move
x86_cpu_class_check_missing_features() closer since it is only used
by this QMP command handler.
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20230223155540.30370-3-philmd@linaro.org>
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Since commit a0e61807a3 ("qapi: Remove QMP events and commands from
user-mode builds") we don't generate the "qapi-commands-machine.h"
header in a user-emulation-only build.
Move the QMP functions from helper.c (which is always compiled)
to monitor.c (which is only compiled when system-emulation
is selected). Rename monitor.c to arm-qmp-cmds.c.
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20230223155540.30370-2-philmd@linaro.org>
Signed-off-by: Markus Armbruster <armbru@redhat.com>
[Straightforward conflict with commit 9def656e7a resolved]
Daniel Henrique Barboza <dbarboza@ventanamicro.com> says:
This is a re-send of patch 1, which is already reviewed, with a
follow-up that uses riscv_cpu_cfg() in the remaining of the file. This
was suggested by Weiwei Li in the "[PATCH 0/4] RISCVCPUConfig related
cleanups" review. Patch 1 makes the work of patch 2 easier since it
eliminated some uses of env_archcpu() we want to avoid.
* b4-shazam-merge:
target/riscv/vector_helper.c: avoid env_archcpu() when reading RISCVCPUConfig
target/riscv/vector_helper.c: create vext_set_tail_elems_1s()
Message-ID: <20230226170514.588071-1-dbarboza@ventanamicro.com>
Signed-off-by: Palmer Dabbelt <palmer@rivosinc.com>
Commit 752614cab8 ("target/riscv: rvv: Add tail agnostic for vector
load / store instructions") added code to set the tail elements to 1 in
the end of vext_ldst_stride(), vext_ldst_us(), vext_ldst_index() and
vext_ldff(). Aside from a env->vl versus an evl value being used in the
first loop, the code is being repeated 4 times.
Create a helper to avoid code repetition in all those functions.
Arguments that are used in the callers (nf, esz and max_elems) are
passed as arguments. All other values are being derived inside the
helper.
Reviewed-by: Weiwei Li <liweiwei@iscas.ac.cn>
Reviewed-by: Frank Chang <frank.chang@sifive.com>
Signed-off-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com>
Message-ID: <20230226170514.588071-2-dbarboza@ventanamicro.com>
Signed-off-by: Palmer Dabbelt <palmer@rivosinc.com>
Daniel Henrique Barboza <dbarboza@ventanamicro.com> says:
These cleanups were suggested by LIU Zhiwei during the review of
d3e6d5762b ("Merge patch series "make write_misa a no-op and FEATURE_*
cleanups"").
* b4-shazam-merge:
target/riscv/csr.c: avoid env_archcpu() usages when reading RISCVCPUConfig
target/riscv/csr.c: use riscv_cpu_cfg() to avoid env_cpu() pointers
target/riscv/csr.c: simplify mctr()
target/riscv/csr.c: use env_archcpu() in ctr()
Message-ID: <20230224174520.92490-1-dbarboza@ventanamicro.com>
Signed-off-by: Palmer Dabbelt <palmer@rivosinc.com>
Weiwei Li <liweiwei@iscas.ac.cn> says:
This patchset adds support svadu extension. It also fixes some
relationship between *envcfg fields and Svpbmt/Sstc extensions.
Specification for Svadu extension can be found in:
https://github.com/riscv/riscv-svadu
* b4-shazam-merge:
target/riscv: Export Svadu property
target/riscv: Add *envcfg.HADE related check in address translation
target/riscv: Add *envcfg.PBMTE related check in address translation
target/riscv: Add csr support for svadu
target/riscv: Fix the relationship of PBMTE/STCE fields between menvcfg and henvcfg
target/riscv: Fix the relationship between menvcfg.PBMTE/STCE and Svpbmt/Sstc extensions
Message-ID: <20230224040852.37109-1-liweiwei@iscas.ac.cn>
Signed-off-by: Palmer Dabbelt <palmer@rivosinc.com>
When menvcfg.HADE is 1, hardware updating of PTE A/D bits is enabled
during single-stage address translation. When the hypervisor extension is
implemented, if menvcfg.HADE is 1, hardware updating of PTE A/D bits is
enabled during G-stage address translation.
Set *envcfg.HADE default true for backward compatibility.
Signed-off-by: Weiwei Li <liweiwei@iscas.ac.cn>
Signed-off-by: Junqiang Wang <wangjunqiang@iscas.ac.cn>
Reviewed-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com>
Message-ID: <20230224040852.37109-6-liweiwei@iscas.ac.cn>
Signed-off-by: Palmer Dabbelt <palmer@rivosinc.com>
menvcfg.PBMTE bit controls whether the Svpbmt extension is available
for use in S-mode and G-stage address translation.
henvcfg.PBMTE bit controls whether the Svpbmt extension is available
for use in VS-stage address translation.
Set *envcfg.PBMTE default true for backward compatibility.
Signed-off-by: Weiwei Li <liweiwei@iscas.ac.cn>
Signed-off-by: Junqiang Wang <wangjunqiang@iscas.ac.cn>
Reviewed-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com>
Message-ID: <20230224040852.37109-5-liweiwei@iscas.ac.cn>
Signed-off-by: Palmer Dabbelt <palmer@rivosinc.com>
Launch qemu-system-riscv64 with a given dtb for 'sifive_u' and 'virt'
machines, QEMU complains:
qemu_fdt_add_subnode: Failed to create subnode /soc: FDT_ERR_EXISTS
The whole DT generation logic should be skipped when a given DTB is
present.
Fixes: b1f19f238c ("hw/riscv: write bootargs 'chosen' FDT after riscv_load_kernel()")
Signed-off-by: Bin Meng <bmeng@tinylab.org>
Reviewed-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com>
Message-ID: <20230228074522.1845007-1-bmeng@tinylab.org>
Signed-off-by: Palmer Dabbelt <palmer@rivosinc.com>
This bug has a noticeable behavior of falling back to the main loop and
respawning a redundant translation block including a single instruction
when the end address of the compressive instruction is exactly on a page
boundary, and slows down running system performance.
Signed-off-by: Shaobo Song <songshaobo@eswincomputing.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-ID: <20230220072732.568-1-songshaobo@eswincomputing.com>
Signed-off-by: Palmer Dabbelt <palmer@rivosinc.com>
Bin Meng <bmeng@tinylab.org> says:
At present gdbstub reports an incorrect / incomplete CSR list in the
target description XML, for example:
- menvcfg is reported in 'sifive_u' machine
- fcsr is missing in a F/D enabled processor
The issue is caused by:
- priv spec version check is missing when reporting CSRs
- CSR predicate() routine is called without turning on the debugger flag
* b4-shazam-merge:
target/riscv: Group all predicate() routines together
target/riscv: Drop priv level check in mseccfg predicate()
target/riscv: Allow debugger to access sstc CSRs
target/riscv: Allow debugger to access {h, s}stateen CSRs
target/riscv: Allow debugger to access seed CSR
target/riscv: Allow debugger to access user timer and counter CSRs
target/riscv: gdbstub: Drop the vector CSRs in riscv-vector.xml
target/riscv: gdbstub: Turn on debugger mode before calling CSR predicate()
target/riscv: Avoid reporting odd-numbered pmpcfgX in the CSR XML for RV64
target/riscv: Simplify getting RISCVCPU pointer from env
target/riscv: Simplify {read, write}_pmpcfg() a little bit
target/riscv: Use 'bool' type for read_only
target/riscv: Coding style fixes in csr.c
target/riscv: gdbstub: Do not generate CSR XML if Zicsr is disabled
target/riscv: gdbstub: Minor change for better readability
target/riscv: Use g_assert() for the predicate() NULL check
target/riscv: Add some comments to clarify the priority policy of riscv_csrrw_check()
target/riscv: gdbstub: Check priv spec version before reporting CSR
Message-ID: <20230228104035.1879882-1-bmeng@tinylab.org>
Signed-off-by: Palmer Dabbelt <palmer@rivosinc.com>
riscv_csrrw_check() already does the generic privilege level check
hence there is no need to do the specific M-mode access check in
the mseccfg predicate().
With this change debugger can access the mseccfg CSR anytime.
Signed-off-by: Bin Meng <bmeng@tinylab.org>
Reviewed-by: Weiwei Li <liweiwei@iscas.ac.cn>
Message-ID: <20230228104035.1879882-18-bmeng@tinylab.org>
Signed-off-by: Palmer Dabbelt <palmer@rivosinc.com>
At present with a debugger attached sstc CSRs can only be accssed
when CPU is in M-mode, or configured correctly.
Fix it by adjusting their predicate() routine logic so that the
static config check comes before the run-time check, as well as
adding a debugger check.
Signed-off-by: Bin Meng <bmeng@tinylab.org>
Reviewed-by: Weiwei Li <liweiwei@iscas.ac.cn>
Message-ID: <20230228104035.1879882-17-bmeng@tinylab.org>
Signed-off-by: Palmer Dabbelt <palmer@rivosinc.com>
