Hi Isaku,
On 2/16/21 3:12 AM, Isaku Yamahata wrote:
> Switch to making a VM ioctl() call for KVM_CAP_READONLY_MEM, which may
> be conditional on VM type in recent versions of KVM, e.g. when TDX is
> supported.
>
> Signed-off-by: Isaku Yamahata
> ---
> accel/kvm/kvm-all.c | 2 +-
> 1 file
On 2/16/21 3:12 AM, Isaku Yamahata wrote:
> From: Sean Christopherson
>
> Expose x86_cpu_get_supported_feature_word() outside of cpu.c so that it
> can be used by TDX to setup the VM-wide CPUID configuration.
>
> Signed-off-by: Sean Christopherson
> ---
> target/i386/cpu.c | 4 ++--
>
Hi Yuri,
On 2/16/21 6:29 AM, Yuri Benditovich wrote:
> https://bugzilla.redhat.com/show_bug.cgi?id=1743098
Maybe add backtrace in patch description?
(gdb) bt
#0 0xc5bbdf0d in virtio_pci_notify_write (opaque=0x55b6c6dff170,
addr=0, val=<...>, size=<...>)
at hw/virtio/virtio-pci.c:1360
#1
On 2/9/21 8:29 PM, Mark Cave-Ayland wrote:
> The QOM object representing ESPState is currently embedded within both the
> SYSBUS_ESP and PCI_ESP devices with migration state handled by embedding
> vmstate_esp within each device using VMSTATE_STRUCT.
>
> Since the vmstate_esp fields are embedded
On 2/9/21 8:29 PM, Mark Cave-Ayland wrote:
> The transfer direction is currently determined by checking the sign of ti_size
> but as this series progresses ti_size can be zero at the end of the transfer.
>
> Use the SCSI phase to determine the transfer direction as used in other SCSI
> controller
On 2/9/21 8:29 PM, Mark Cave-Ayland wrote:
> Perform the length adjustment whereby a value of 0 in the STC represents
> a transfer length of 0x1 at the point where the TC is loaded at the
0x1 -> 64 KiB?
> start of a DMA command rather than just when a TI (Transfer Information)
> command
On 2/9/21 8:29 PM, Mark Cave-Ayland wrote:
> Now that all SCSI commands are accumulated in cmdbuf, remove the buf and
> buflen
> parameters from get_cmd() since these always reference cmdbuf and
> ESP_CMDBUF_SZ
> respectively.
>
> Signed-off-by: Mark Cave-Ayland
> ---
> hw/scsi/esp.c | 11
Hi Mark,
On 2/15/21 11:35 PM, Mark Cave-Ayland wrote:
> On 12/02/2021 18:56, Philippe Mathieu-Daudé wrote:
>
>> On 2/9/21 8:30 PM, Mark Cave-Ayland wrote:
>>> The MacOS toolbox ROM performs 4 byte reads/writes when transferring
>>> data to
>>> and from the target. Since the SCSI bus is 16-bits
On 2/16/21 8:15 AM, Thomas Huth wrote:
> There were recently some patches on the list which had their "From:"
> line mangled like this:
>
> From: qemu_oss--- via
>
> Since our test in the checkpatch.pl script did not trigger here, the
> patches finally also ended up in a pull request, with the
There were recently some patches on the list which had their "From:"
line mangled like this:
From: qemu_oss--- via
Since our test in the checkpatch.pl script did not trigger here, the
patches finally also ended up in a pull request, with the wrong author
set. So let's improve the regular
On Mon, Feb 15, 2021 at 01:01:52PM -0800, Richard Henderson wrote:
> On 2/14/21 9:58 AM, Philippe Mathieu-Daudé wrote:
> > +/*
> > + * The TX79-specific instruction Store Quadword
> > + *
> > + * ++---+---++
> > + * | 01 | base |
On 2/16/21 7:55 AM, Thomas Huth wrote:
> On 15/02/2021 20.28, Philippe Mathieu-Daudé wrote:
>> While we are interested in building docker images in different
>> branches, it only makes sense to push 'latest' to the registry
>> when this is the project default branch (usually 'master').
>>
>> Else
On 15/02/2021 20.28, Philippe Mathieu-Daudé wrote:
While we are interested in building docker images in different
branches, it only makes sense to push 'latest' to the registry
when this is the project default branch (usually 'master').
Else when pushing different branches concurrently we might
https://bugzilla.redhat.com/show_bug.cgi?id=1743098
There is missing check for vdev in this procedure.
QEMU crash happens in it in hot unplug flow.
Signed-off-by: Yuri Benditovich
---
hw/virtio/virtio-pci.c | 8 +++-
1 file changed, 7 insertions(+), 1 deletion(-)
diff --git
15.02.2021 23:04, Eric Blake wrote:
On 2/15/21 3:22 AM, Kevin Wolf wrote:
With this patch applied, 'check unit-test' fails with:
Running test test-replication
Unexpected error in bdrv_open_driver() at ../block.c:1481:
Could not open '/tmp/p_local_disk.z1Ugyc': Invalid argument
ERROR
Some tests explicitly require a QEMU accelerator to be available.
Given that this depends on some runtime aspects not known before
the test is started, such as the currently set QEMU binary, it's
left to be checked also at runtime.
Signed-off-by: Cleber Rosa
Message-Id:
It's questionable whether it's necessary to create one brand new pair
for each test. It's not questionable that it takes less time and
resources to just use the keys available at "tests/keys" that exist
for that exact reason.
If a location for the public key is not given explicitly, the
From: Eric Blake
Git's default hunk pattern recognizer favors the C language, but it
also includes several built-in diff styles that give saner results in
other languages. In particular, telling git to treat all .py files as
python changes the beginning of diff hunks as follows:
| ---
Currently the path of the ssh public key is being set, but its
content is obviously what's needed.
Signed-off-by: Cleber Rosa
Message-Id: <20210203172357.1422425-18-cr...@redhat.com>
Reviewed-by: Marc-André Lureau
Reviewed-by: Wainer dos Santos Moschetta
Reviewed-by: Willian Rampazzo
This is basically the infrastructure around "boot_linux.py" tests, but
now made into a base class for general use.
Signed-off-by: Cleber Rosa
Message-Id: <20210203172357.1422425-15-cr...@redhat.com>
Reviewed-by: Wainer dos Santos Moschetta
Reviewed-by: Willian Rampazzo
Signed-off-by: Cleber
This version (and 84.0) contain improvements that address specific
QEMU use cases, including:
* Being able to download and use Fedora 31 images and thus
re-activate the "boot_linux.py" tests
* Being able to register local assets via "avocado assets register"
and use them in tests
Preserve log at location already prepared for keeping the test's log
files.
While at it, log info about its location (in the main test log
file), instead of printing it out.
Reference:
https://avocado-framework.readthedocs.io/en/85.0/api/test/avocado.html#avocado.Test.logdir
Signed-off-by:
Closing a file that is open for writing, and then reading from it
sounds like a better idea than the opposite, given that the content
will be flushed.
Reference: https://docs.python.org/3/library/io.html#io.IOBase.close
Signed-off-by: Cleber Rosa
Message-Id:
From: Max Reitz
>From the cancel message, it is not entirely clear why this parameter is
mandatory now, or that it will be optional in the future. Add such a
more detailed explanation as a comment in the test source file.
Suggested-by: Alex Bennée
Signed-off-by: Max Reitz
Message-Id:
The following changes since commit 8ba4bca570ace1e60614a0808631a517cf5df67a:
Merge remote-tracking branch 'remotes/kevin/tags/for-upstream' into staging=
(2021-02-15 17:13:57 +)
are available in the Git repository at:
https://gitlab.com/cleber.gnu/qemu.git/
If the block size is programmed to a different value from the
previous one, reset the data pointer of s->fifo_buffer[] so that
s->fifo_buffer[] can be filled in using the new block size in
the next transfer.
With this fix, the following reproducer:
outl 0xcf8 0x80001010
outl 0xcfc 0xe000
The codes to limit the maximum block size is only necessary when
SDHC_BLKSIZE register is writable.
Signed-off-by: Bin Meng
---
Changes in v2:
- new patch: sdhci: Limit block size only when SDHC_BLKSIZE register is writable
hw/sd/sdhci.c | 14 +++---
1 file changed, 7 insertions(+),
When an ADMA transfer is started, the codes forget to set the
controller status to indicate a transfer is in progress.
With this fix, the following 2 reproducers:
https://paste.debian.net/plain/1185136
https://paste.debian.net/plain/1185141
cannot be reproduced with the following QEMU command
s->prnsts is updated in both branches of the if () else () statement.
Move the common bits outside so that it is cleaner.
Signed-off-by: Bin Meng
---
(no changes since v1)
hw/sd/sdhci.c | 7 +++
1 file changed, 3 insertions(+), 4 deletions(-)
diff --git a/hw/sd/sdhci.c b/hw/sd/sdhci.c
Per "SD Host Controller Standard Specification Version 7.00"
chapter 2.2.1 SDMA System Address Register:
This register can be accessed only if no transaction is executing
(i.e., after a transaction has stopped).
With this fix, the following reproducer:
https://paste.debian.net/plain/1185137
At the end of sdhci_send_command(), it starts a data transfer if the
command register indicates data is associated. But the data transfer
should only be initiated when the command execution has succeeded.
With this fix, the following reproducer:
outl 0xcf8 0x80001810
outl 0xcfc 0xe1068000
outl
This series includes several fixes to CVE-2020-17380, CVE-2020-25085
and CVE-2021-3409 that are heap-based buffer overflow issues existing
in the sdhci model.
These CVEs are pretty much similar, and were filed using different
reproducers. With this series, current known reproducers I have
cannot
On Mon, Feb 15, 2021 at 06:03:33PM +0100, Philippe Mathieu-Daudé wrote:
> On 2/15/21 4:49 PM, Wainer dos Santos Moschetta wrote:
> > Hi,
> >
> > On 2/8/21 8:35 AM, Philippe Mathieu-Daudé wrote:
> >> On 2/3/21 6:23 PM, Cleber Rosa wrote:
> >>> This introduces a base class for tests that need to
On Thu, Feb 11, 2021 at 02:15:32PM +0400, Marc-André Lureau wrote:
> Hi
>
> On Wed, Feb 3, 2021 at 10:07 PM Cleber Rosa wrote:
> >
> > It's questionable wether it's necessary to create one brand new pair
>
> whether
>
Yep, thanks!
> > for each test. It's not questionable that it takes less
On Mon, Feb 15, 2021 at 04:06:45PM -0300, Willian Rampazzo wrote:
> On Wed, Feb 3, 2021 at 2:24 PM Cleber Rosa wrote:
> >
> > This is basically the infrastructure around "boot_linux.py" tests, but
> > now made into a base class for general use.
> >
> > Signed-off-by: Cleber Rosa
> > ---
> >
On Mon, Feb 15, 2021 at 04:25:24PM -0600, Eric Blake wrote:
> Git's default hunk pattern recognizer favors the C language, but it
> also includes several built-in diff styles that give saner results in
> other languages. In particular, telling git to treat all .py files as
> python changes the
Follows the qapi/introspect.py definition of the same; this adds a more
precise typing to _gen_tree's mtype parameter.
NB: print(SchemaMetaType.BUILTIN) would produce the string
"SchemaMetaType.BUILTIN", but when using format strings (.format or f-strings),
it relies on the __format__ method
On Mon, Feb 15, 2021 at 03:30:16PM -0300, Wainer dos Santos Moschetta wrote:
> Hi,
>
> On 2/11/21 7:01 PM, Cleber Rosa wrote:
> > Closing a file that is open for writing, and then reading from it
> > sounds like a better idea than the opposite, given that the content
> > will be flushed.
> >
> >
To reflect the work that went into strictly typing introspect.py,
punish myself by claiming credit.
Signed-off-by: John Snow
---
scripts/qapi/introspect.py | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/scripts/qapi/introspect.py b/scripts/qapi/introspect.py
index
Presently, we use a tuple to attach a dict containing annotations
(comments and compile-time conditionals) to a tree node. This is
undesirable because dicts are difficult to strongly type; promoting it
to a real class allows us to name the values and types of the
annotations we are expecting.
In
It is easier to give a name to all of the dictly-typed objects we pass
around in introspect.py by removing this helper, as it does not return
an object that has any knowable type by itself.
Inline it into its only caller instead.
Signed-off-by: John Snow
---
scripts/qapi/introspect.py | 9
NB: The type aliases (SchemaInfo et al) declare intent for some of the
"dictly-typed" objects we pass around in introspect.py. They do not
enforce the shape of those objects, and cannot, until Python 3.7 or
later. (And even then, it may not be "worth it".)
Signed-off-by: John Snow
---
Trivial; make the error message just a pinch more explicit in case we
trip this by accident in the future.
Signed-off-by: John Snow
---
scripts/qapi/introspect.py | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/scripts/qapi/introspect.py b/scripts/qapi/introspect.py
index
At present, we open-code this in _make_tree itself; but if the structure
of the tree changes, this is brittle. Use an explicit recursive call to
_make_tree when appropriate to help keep the interior node typing
consistent.
A consequence of doing this is that the 'ifcond' key of the features
dict
We don't need to create an empty, mutable list to pass to _gen_tree;
since it is now typed as a Sequence, we can use the empty tuple as a
default and omit the argument.
Signed-off-by: John Snow
---
scripts/qapi/introspect.py | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git
The types will be used in forthcoming patches to add typing. These types
describe the layout and structure of the objects passed to
_tree_to_qlit, but lack the power to describe annotations until the next
commit.
Signed-off-by: John Snow
---
scripts/qapi/introspect.py | 31
On Mon, Feb 15, 2021 at 05:04:24PM -0500, John Snow wrote:
> On 2/11/21 5:01 PM, Cleber Rosa wrote:
> > Closing a file that is open for writing, and then reading from it
> > sounds like a better idea than the opposite, given that the content
> > will be flushed.
> >
> > Reference:
The introspect visitor is stateful, but expects that it will have a
schema to refer to. Add assertions that state this.
Signed-off-by: John Snow
---
scripts/qapi/introspect.py | 5 +
1 file changed, 5 insertions(+)
diff --git a/scripts/qapi/introspect.py b/scripts/qapi/introspect.py
index
This is only used to pass in a dictionary with a comment already set, so
skip the runaround and just accept the (optional) comment.
Signed-off-by: John Snow
---
scripts/qapi/introspect.py | 15 +--
1 file changed, 9 insertions(+), 6 deletions(-)
diff --git
This mimics how a typed object works, where 'if' and 'comment' are
always set, regardless of if they have a value set or not.
It is safe to do this because of the way that _tree_to_qlit processes
these values (using dict.get with a default of None), resulting in no
change of output from
Optional[List] is clunky; an empty sequence can more elegantly convey
"no variants". By downgrading "List" to "Sequence", we can also accept
tuples; this is useful for the empty tuple specifically, which we may
use as a default parameter because it is immutable.
Signed-off-by: John Snow
---
Specify the initial value for RCX/R8 to be the address of the HOB.
Don't propagate the value to Qemu's cache of the registers so as to
avoid implying that the register state is valid, e.g. Qemu doesn't model
TDX-SEAM behavior for initializing other GPRs.
Signed-off-by: Isaku Yamahata
---
Signed-off-by: John Snow
---
scripts/qapi/introspect.py | 18 ++
1 file changed, 18 insertions(+)
diff --git a/scripts/qapi/introspect.py b/scripts/qapi/introspect.py
index 45284af1330..5d4f5e23f7e 100644
--- a/scripts/qapi/introspect.py
+++ b/scripts/qapi/introspect.py
@@ -99,6
_tree_to_qlit is called recursively on dict values (isolated from their
keys); at such a point in generating output it is too late to apply an
ifcond. Similarly, comments do not necessarily have a "tidy" place they
can be printed in such a circumstance.
Forbid this usage by renaming
Add, and optionally measure, TDVF memory via KVM_TDX_INIT_MEM_REGION as
part of finalizing the TD.
Signed-off-by: Isaku Yamahata
Co-developed-by: Sean Christopherson
Signed-off-by: Sean Christopherson
---
target/i386/kvm/tdx.c | 19 ---
1 file changed, 16 insertions(+), 3
Subjective, but I find getting rid of the comprehensions helps. Also,
divide the sections into scalar and non-scalar sections, and remove
old-style string formatting.
Signed-off-by: John Snow
---
scripts/qapi/introspect.py | 33 ++---
1 file changed, 18
Add definitions for literals, enums, structs, GUIDs, etc... that will be
used by TDX to build the UEFI Hand-Off Block (HOB) that is passed to the
Trusted Domain Virtual Firmware (TDVF). All values come from the UEFI
specification.
note: EFI_RESOURCE_ATTRIBUTE_{ENCRYPTED, UNACCEPTED}, will be
It does happen to be a list (as of now), but we can describe it in more
general terms with no loss in accuracy to allow tuples and other
constructs.
In the future, we can write "ifcond: Sequence[str] = ()" as a default
parameter, which we could not do with a Mutable type like a List.
Returning two different types conditionally can be complicated to
type. Return one type for consistency.
Signed-off-by: John Snow
---
scripts/qapi/introspect.py | 4 +---
1 file changed, 1 insertion(+), 3 deletions(-)
diff --git a/scripts/qapi/introspect.py b/scripts/qapi/introspect.py
index
_make_tree might receive a dict (a SchemaInfo object) or some other type
(usually, a string) for its obj parameter. Adding features information
should arguably be performed by the caller at such a time when we know
the type of the object and don't have to re-interrogate it.
Signed-off-by: John
Hi, this series adds static type hints to the QAPI module.
This is part two, and covers introspect.py.
Part 2: https://gitlab.com/jsnow/qemu/-/tree/python-qapi-cleanup-pt2
Everything: https://gitlab.com/jsnow/qemu/-/tree/python-qapi-cleanup-pt6
- Requires Python 3.6+
- Requires mypy 0.770 or
From: Sean Christopherson
Add a hook for TDX to denote that the TD Virtual Firmware must be
provided via the "generic" device loader. Error out if pflash is used
in conjuction with TDX.
Suggested-by: Isaku Yamahata
Signed-off-by: Sean Christopherson
---
hw/i386/pc_sysfw.c | 6
On 2/15/21 5:19 PM, Richard Henderson wrote:
On 2/15/21 1:58 PM, Rebecca Cran wrote:
@@ -960,6 +960,12 @@ static void cpsr_write_from_spsr_elx(CPUARMState *env,
val |= CPSR_DIT;
}
+/* Move SSBS to the correct location for CPSR */
+if (val & PSTATE_SSBS) {
+
From: Sean Christopherson
Add a machine option to disable the legacy PIC (8259), which cannot be
supported for TDX guests as TDX-SEAM doesn't allow directly interrupt
injection. Using posted interrupts for the PIC is not a viable option
as the guest BIOS/kernel will not do EOI for PIC IRQs,
Add support for loading TDX's Trusted Domain Virtual Firmware (TDVF) via
the generic loader. Prioritize the TDVF above plain hex to avoid false
positives with hex (TDVF has explicit metadata to confirm it's a TDVF).
Enumerate TempMem as added, private memory, i.e. E820_RESERVED,
otherwise TDVF
From: Sean Christopherson
TDX requires x2apic and "resets" vCPUs to have x2apic enabled. Model
this in QEMU and unconditionally enable x2apic interrupt routing.
This fixes issues where interrupts from IRQFD would not get forwarded to
the guest due to KVM silently dropping the invalid routing
From: Sean Christopherson
Add support for grabbing KVM_TDX_CAPABILITIES and use the new
kvm_get_supported_cpuid() hook to adjust the supported XCR0 bits.
Add TODOs for the remaining work.
Signed-off-by: Sean Christopherson
---
target/i386/kvm/kvm.c | 2 ++
target/i386/kvm/tdx.c | 84
Signed-off-by: Sean Christopherson
---
include/sysemu/tdx.h | 1 +
target/i386/kvm/kvm.c | 8
target/i386/kvm/tdx-stub.c | 4
target/i386/kvm/tdx.c | 20
4 files changed, 29 insertions(+), 4 deletions(-)
diff --git a/include/sysemu/tdx.h
From: Xiaoyao Li
Introduce tdx_ioctl() to invoke TDX specific sub-ioctls of
KVM_MEMORY_ENCRYPT_OP. Use tdx_ioctl() to invoke KVM_TDX_INIT, by way
of tdx_init(), during kvm_arch_init(). KVM_TDX_INIT configures global
TD state, e.g. the canonical CPUID config, and must be executed prior to
From: Sean Christopherson
Expose x86_cpu_get_supported_feature_word() outside of cpu.c so that it
can be used by TDX to setup the VM-wide CPUID configuration.
Signed-off-by: Sean Christopherson
---
target/i386/cpu.c | 4 ++--
target/i386/cpu.h | 3 +++
2 files changed, 5 insertions(+), 2
Build the TD HOB during machine late initialization, i.e. once guest
memory is fully defined.
Note, the attribute absolutely for MMIO HOB entries must include
UNCACHEABLE, else TDVF will effectively consider it a bad HOB entry
and ignore it.
Signed-off-by: Isaku Yamahata
Co-developed-by: Sean
From: Xiaoyao Li
Pull in recent TDX updates, which are not backwards compatible.
Signed-off-by: Xiaoyao Li
Co-developed-by: Sean Christopherson
Signed-off-by: Sean Christopherson
---
linux-headers/asm-x86/kvm.h | 55 +
linux-headers/linux/kvm.h | 2 ++
From: Sean Christopherson
Move the architectural (for lack of a better term) CPUID leaf generation
to a separate helper so that the generation code can be reused by TDX,
which needs to generate a canonical VM-scoped configuration.
Signed-off-by: Sean Christopherson
---
target/i386/kvm/kvm.c
Add constants and structs for the TD Virtual Firmware metadata, which
describes how the TDVF must be built to ensure correct functionality and
measurement.
Signed-off-by: Isaku Yamahata
Co-developed-by: Sean Christopherson
Signed-off-by: Sean Christopherson
---
include/hw/i386/tdvf.h | 55
Make kvm_synchronize_all_tsc() nop for TD-guest.
TDX module specification, 9.11.1 TSC Virtualization
"Virtual TSC values are consistent among all the TD;s VCPUs at the
level suppored by the CPU".
There is no need for qemu to synchronize tsc and VMM can't access
to guest TSC. Actually
Despite advertising MCE support to the guest, TDX-SEAM doesn't support
injecting #MCs into the guest. All of the associated setup is thus
rejected by KVM.
Signed-off-by: Isaku Yamahata
---
target/i386/kvm/kvm.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git
From: Sean Christopherson
Ignore get/put state of TDX VMs as accessing/mutating guest state of
producation TDs is not supported.
Allow kvm_arch_get_registers() to run as normal, except for MSRs, for
debug TDs, and silently ignores attempts to read guest state for
non-debug TDs.
Signed-off-by:
From: Xiaoyao Li
Invoke KVM_TDX_FINALIZEMR to finalize the TD's measurement and make the
TD vCPUs runnable once machine initialization is complete.
Signed-off-by: Xiaoyao Li
---
target/i386/kvm/kvm.c | 7 +++
target/i386/kvm/tdx.c | 20
target/i386/kvm/tdx.h | 3 +++
From: Isaku Yamahata
KVM_CAP_X86_SMM is VM-specific capability for TDX.
So use VM ioctl for KVM_CAP_X86_SMM.
Signed-off-by: Isaku Yamahata
---
target/i386/kvm/kvm.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/target/i386/kvm/kvm.c b/target/i386/kvm/kvm.c
index
Switch to making a VM ioctl() call for KVM_CAP_READONLY_MEM, which may
be conditional on VM type in recent versions of KVM, e.g. when TDX is
supported.
Signed-off-by: Isaku Yamahata
---
accel/kvm/kvm-all.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/accel/kvm/kvm-all.c
From: Xiaoyao Li
Introduce a machine property, kvm-type, to allow the user to create a
Trusted Domain eXtensions (TDX) VM, a.k.a. a Trusted Domain (TD), e.g.:
# $QEMU \
-machine ...,kvm-type=tdx \
...
Only two types are supported: "legacy" and "tdx", with "legacy" being
the
Introduce a new notifier, machine_init_done_late, that is notified after
machine_init_done. This will be used by TDX to generate the HOB for its
virtual firmware, which needs to be done after all guest memory has been
added, i.e. after machine_init_done notifiers have run.
some devices
This patch series is to enable TDX support.
This needs corresponding KVM patch for TDX[] and more patches are needed
that addresses generic corner cases, e.g. ACPI related stuff, are needed.
So This patch series is RFC.
More emulated devices and their behavior needs to be adjusted as some
Not disabling the UART leads to QEMU overwriting the UART receive buffer with
the newest received byte. The rx_level variable is added to allow the use of
the existing OpenTitan driver libraries.
Signed-off-by: Alexander Wagner
---
hw/char/ibex_uart.c | 20 +++-
On 1/26/21 10:15 AM, Vladimir Sementsov-Ogievskiy wrote:
OK, thanks for handling it!
When will we move to python 3.7?
"I don't know, but it seems like a very long time."
The nominal EOL for Python 3.6 is this December; see
https://www.python.org/dev/peps/pep-0494/
Debian 10 ships 3.7.
On 210216 0855, Bin Meng wrote:
> Hi Alex,
>
> On Tue, Feb 16, 2021 at 12:48 AM Alexander Bulekov wrote:
> >
> > Hi Bin,
> > Thank you for this. I ran through the OSS-Fuzz tests again, and it found
> > one thing:
>
> Thanks for testing. Are there instructions to run OSS-Fuzz tests myself?
Yes
From: Isaku Yamahata
The following patch will introduce incompatible behavior of SMM.
Introduce a property to keep the old behavior for compatibility.
To enable smm compat, use "-global ICH9-LPC.smm-compat=on" or
"-global PIIX4.smm-compat=on"
Suggested-by: Igor Mammedov
Signed-off-by: Isaku
From: Isaku Yamahata
If SMM is not supported, ACPI fixed hardware doesn't support
legacy-mode. ACPI-only platform. Where SCI_EN in PM1_CNT register is
always set.
The bit tells OS legacy mode(SCI_EN cleared) or ACPI mode(SCI_EN set).
With the next patch (setting fadt.smi_cmd = 0 when smm isn't
From: Isaku Yamahata
update golden master acpi tables and empty
bios-tables-test-allowed-diff.h.
Signed-off-by: Isaku Yamahata
---
tests/data/acpi/pc/DSDT.nohpet | Bin 0 -> 4923 bytes
tests/data/acpi/pc/FACP.nosmm | Bin 0 -> 116 bytes
tests/data/acpi/q35/DSDT
From: Isaku Yamahata
Reviewed-by: Igor Mammedov
Signed-off-by: Isaku Yamahata
---
tests/qtest/bios-tables-test.c | 76 ++
1 file changed, 76 insertions(+)
diff --git a/tests/qtest/bios-tables-test.c b/tests/qtest/bios-tables-test.c
index 669202fc95..592c074ec7
From: Isaku Yamahata
Reviewed-by: Igor Mammedov
Signed-off-by: Isaku Yamahata
---
tests/qtest/bios-tables-test.c | 24
1 file changed, 24 insertions(+)
diff --git a/tests/qtest/bios-tables-test.c b/tests/qtest/bios-tables-test.c
index 592c074ec7..3fb7ed0c46 100644
From: Isaku Yamahata
The following tests will modify acpi tables.
prepare qtests to allow acpi table change.
add new tables for new tests.
- tests/data/acpi/pc/DSDT.nohpet
- tests/data/acpi/pc/FACP.nosmm
- tests/data/acpi/q35/DSDT.nohpet
- tests/data/acpi/q35/FACP.nosmm
Acked-by: Igor Mammedov
From: Isaku Yamahata
Miscellaneous bug fixes related to ACPI to play nice with guest BIOSes/OSes
by conforming to ACPI spec better.
Changes from v3:
- rename compat property name to smm-compat and add it to ICH9-LPC and PIIX4_PM
- MMCFG memory region to use dword memory region when possible
-
From: Sean Christopherson
Omit HPET AML if the HPET is disabled, QEMU is not emulating it and the
guest may get confused by seeing HPET in the ACPI tables without a
"physical" device present.
The change of DSDT when -no-hpet is as follows.
@@ -141,47 +141,6 @@ DefinitionBlock ("", "DSDT", 1,
From: Isaku Yamahata
Declare PNP0C01 device to reserve MMCONFIG region to conform to the
spec better and play nice with guest BIOSes/OSes.
According to PCI Firmware Specification[0], MMCONFIG region must be
reserved by declaring a motherboard resource. It's optional to reserve
the region in
From: Isaku Yamahata
>From table 5.9 SMI_CMD of ACPI spec
> This field is reserved and must be zero on system
> that does not support System Management mode.
When smm is not enabled, set it to zero to comform to the spec.
When -machine smm=off is passed, the change to FACP is as follows.
@@
From: Isaku Yamahata
Newly created acpi data files(tests/data/acpi/) cause false positive
warning.
If file names are acpi expected file, don't emit warning.
Fixes: e625ba2a41 ("checkpatch: fix acpi check with multiple file name")
Signed-off-by: Isaku Yamahata
---
scripts/checkpatch.pl | 4
Hi Alex,
On Tue, Feb 16, 2021 at 12:48 AM Alexander Bulekov wrote:
>
> Hi Bin,
> Thank you for this. I ran through the OSS-Fuzz tests again, and it found
> one thing:
Thanks for testing. Are there instructions to run OSS-Fuzz tests myself?
> Maybe this is already much better than the current
On 2/15/21 2:21 AM, Peter Maydell wrote:
> For a long time now the UI layer has guaranteed that the console
> surface is always 32 bits per pixel, RGB. The TCX code already
> assumes 32bpp, but it still has some checks of is_surface_bgr()
> in an attempt to support 32bpp BGR. is_surface_bgr() will
On 2/15/21 2:32 AM, Peter Maydell wrote:
> Peter Maydell (9):
> hw/arm/musicpal: Remove dead code for non-32-bit-RGB surfaces
> hw/display/tc6393xb: Remove dead code for handling non-32bpp surfaces
> hw/display/tc6393xb: Expand out macros in template header
> hw/display/tc6393xb: Inline
1 - 100 of 427 matches
Mail list logo
