qga-win: Increase VSS freeze timeout to 60 secs instead of 10
Currently Requester freeze times out after 10 seconds, while
the default timeout for Writer Freeze is 60 seconds. according to
VSS Documentation [1].
[1]: https://docs.microsoft.com/en-us/windows/win32/vss/overview-of-processing-a-backup-under-vss
* remotes/bonzini-gitlab/tags/for-upstream:
coverity-model: write models fully for non-array allocation functions
coverity-model: constrain g_malloc/g_malloc0/g_realloc as never returning NULL
coverity-model: clean up the models for array allocation functions
coverity-model: remove model for more allocation functions
coverity-model: make g_free a synonym of free
coverity-model: update address_space_read/write models
vl: stop recording -smp in QemuOpts
vl: introduce machine_merge_property
Peter Maydell [Mon, 2 Aug 2021 11:59:00 +0000 (12:59 +0100)]
Merge remote-tracking branch 'remotes/pmaydell/tags/pull-target-arm-20210802' into staging
target-arm queue:
* Add documentation of Arm 'mainstone', 'kzm', 'imx25-pdk' boards
* MAINTAINERS: Don't list Andrzej Zaborowski for various components
* docs: Remove stale TODO comments about license and version
* docs: Move licence/copyright from HTML output to rST comments
* docs: Format literal text correctly
* hw/arm/boot: Report error if there is no fw_cfg device in the machine
* docs: rSTify barrier.txt and bootindex.txt
* remotes/pmaydell/tags/pull-target-arm-20210802: (21 commits)
docs: Move user-facing barrier docs into system manual
ui/input-barrier: Move TODOs from barrier.txt to a comment
docs: Move the protocol part of barrier.txt into interop
docs: Move bootindex.txt into system section and rstify
hw/arm/boot: Report error if there is no fw_cfg device in the machine
docs/tools/virtiofsd.rst: Delete stray backtick
docs/about/removed-features: Fix markup error
docs: Format literals correctly
docs/system/arm/cpu-features.rst: Format literals correctly
docs/system/s390x/protvirt.rst: Format literals correctly
docs/devel: Format literals correctly
docs/devel/migration.rst: Format literals correctly
docs/devel/ebpf_rss.rst: Format literals correctly
docs/devel/build-system.rst: Correct typo in example code
docs/devel/build-system.rst: Format literals correctly
docs: Move licence/copyright from HTML output to rST comments
docs: Remove stale TODO comments about license and version
MAINTAINERS: Don't list Andrzej Zaborowski for various components
docs: Add documentation of Arm 'imx25-pdk' board
docs: Add documentation of Arm 'kzm' board
...
Peter Maydell [Tue, 27 Jul 2021 20:41:12 +0000 (21:41 +0100)]
docs: Move user-facing barrier docs into system manual
The remaining text in docs/barrier.txt is user-facing description
of what the device is and how to use it. Move this into the
system manual and rstify it.
Peter Maydell [Tue, 27 Jul 2021 20:41:10 +0000 (21:41 +0100)]
docs: Move the protocol part of barrier.txt into interop
Most of docs/barrier.txt is describing the protocol implemented
by the input-barrier device. Move this into the interop
section of the manual, and rstify it.
Peter Maydell [Tue, 27 Jul 2021 19:49:55 +0000 (20:49 +0100)]
docs: Move bootindex.txt into system section and rstify
Move bootindex.txt into the system section of the manual and turn it
into rST format. To make the document make more sense in the context
of the system manual, expand the title and introductory paragraphs to
give more context.
Peter Maydell [Mon, 26 Jul 2021 16:33:51 +0000 (17:33 +0100)]
hw/arm/boot: Report error if there is no fw_cfg device in the machine
If the user provides both a BIOS/firmware image and also a guest
kernel filename, arm_setup_firmware_boot() will pass the
kernel image to the firmware via the fw_cfg device. However we
weren't checking whether there really was a fw_cfg device present,
and if there wasn't we would crash.
This crash can be provoked with a command line such as
qemu-system-aarch64 -M raspi3 -kernel /dev/null -bios /dev/null -display none
It is currently only possible on the raspi3 machine, because unless
the machine sets info->firmware_loaded we won't call
arm_setup_firmware_boot(), and the only machines which set that are:
* virt (has a fw-cfg device)
* sbsa-ref (checks itself for kernel_filename && firmware_loaded)
* raspi3 (crashes)
But this is an unfortunate beartrap to leave for future machine
model implementors, so we should handle this situation in boot.c.
Check in arm_setup_firmware_boot() whether the fw-cfg device exists
before trying to load files into it, and if it doesn't exist then
exit with a hopefully helpful error message.
Because we now handle this check in a machine-agnostic way, we
can remove the check from sbsa-ref.
Peter Maydell [Mon, 26 Jul 2021 14:23:37 +0000 (15:23 +0100)]
docs/about/removed-features: Fix markup error
The section describing the removed feature "-usbdevice ccid" had a
typo so the markup started with single backtick and ended with double
backtick; fix it.
Peter Maydell [Mon, 26 Jul 2021 14:23:36 +0000 (15:23 +0100)]
docs: Format literals correctly
In rST markup, single backticks `like this` represent "interpreted
text", which can be handled as a bunch of different things if tagged
with a specific "role":
https://docutils.sourceforge.io/docs/ref/rst/restructuredtext.html#interpreted-text
(the most common one for us is "reference to a URL, which gets
hyperlinked").
The default "role" if none is specified is "title_reference",
intended for references to book or article titles, and it renders
into the HTML as <cite>...</cite> (usually comes out as italics).
This commit fixes various places in the manual which were
using single backticks when double backticks (for literal text)
were intended, and covers those files where only one or two
instances of these errors were made.
Peter Maydell [Mon, 26 Jul 2021 14:23:35 +0000 (15:23 +0100)]
docs/system/arm/cpu-features.rst: Format literals correctly
In rST markup, single backticks `like this` represent "interpreted
text", which can be handled as a bunch of different things if tagged
with a specific "role":
https://docutils.sourceforge.io/docs/ref/rst/restructuredtext.html#interpreted-text
(the most common one for us is "reference to a URL, which gets
hyperlinked").
The default "role" if none is specified is "title_reference",
intended for references to book or article titles, and it renders
into the HTML as <cite>...</cite> (usually comes out as italics).
To format a literal (generally rendered as fixed-width font),
double-backticks are required.
cpu-features.rst consistently uses single backticks when double backticks
are required; correct it.
Peter Maydell [Mon, 26 Jul 2021 14:23:34 +0000 (15:23 +0100)]
docs/system/s390x/protvirt.rst: Format literals correctly
In rST markup, single backticks `like this` represent "interpreted
text", which can be handled as a bunch of different things if tagged
with a specific "role":
https://docutils.sourceforge.io/docs/ref/rst/restructuredtext.html#interpreted-text
(the most common one for us is "reference to a URL, which gets
hyperlinked").
The default "role" if none is specified is "title_reference",
intended for references to book or article titles, and it renders
into the HTML as <cite>...</cite> (usually comes out as italics).
To format a literal (generally rendered as fixed-width font),
double-backticks are required.
protvirt.rst consistently uses single backticks when double backticks
are required; correct it.
Peter Maydell [Mon, 26 Jul 2021 14:23:33 +0000 (15:23 +0100)]
docs/devel: Format literals correctly
In rST markup, single backticks `like this` represent "interpreted
text", which can be handled as a bunch of different things if tagged
with a specific "role":
https://docutils.sourceforge.io/docs/ref/rst/restructuredtext.html#interpreted-text
(the most common one for us is "reference to a URL, which gets
hyperlinked").
The default "role" if none is specified is "title_reference",
intended for references to book or article titles, and it renders
into the HTML as <cite>...</cite> (usually comes out as italics).
Fix various places in the devel section of the manual which were
using single backticks when double backticks (for literal text)
were intended.
Peter Maydell [Mon, 26 Jul 2021 14:23:32 +0000 (15:23 +0100)]
docs/devel/migration.rst: Format literals correctly
In rST markup, single backticks `like this` represent "interpreted
text", which can be handled as a bunch of different things if tagged
with a specific "role":
https://docutils.sourceforge.io/docs/ref/rst/restructuredtext.html#interpreted-text
(the most common one for us is "reference to a URL, which gets
hyperlinked").
The default "role" if none is specified is "title_reference",
intended for references to book or article titles, and it renders
into the HTML as <cite>...</cite> (usually comes out as italics).
To format a literal (generally rendered as fixed-width font),
double-backticks are required.
Mostly migration.rst gets this right, but some places incorrectly use
single backticks where double backticks were intended; correct them.
Peter Maydell [Mon, 26 Jul 2021 14:23:31 +0000 (15:23 +0100)]
docs/devel/ebpf_rss.rst: Format literals correctly
In rST markup, single backticks `like this` represent "interpreted
text", which can be handled as a bunch of different things if tagged
with a specific "role":
https://docutils.sourceforge.io/docs/ref/rst/restructuredtext.html#interpreted-text
(the most common one for us is "reference to a URL, which gets
hyperlinked").
The default "role" if none is specified is "title_reference",
intended for references to book or article titles, and it renders
into the HTML as <cite>...</cite> (usually comes out as italics).
To format a literal (generally rendered as fixed-width font),
double-backticks are required.
ebpf_rss.rst gets this wrong in a few places; correct them.
Peter Maydell [Mon, 26 Jul 2021 14:23:30 +0000 (15:23 +0100)]
docs/devel/build-system.rst: Correct typo in example code
One of the example meson.build fragments incorrectly quotes some
symbols as 'CONFIG_FOO`; the correct syntax here is 'CONFIG_FOO'.
(This isn't a rST formatting mistake because the example is displayed
literally; it's just the wrong kind of quote.)
Peter Maydell [Mon, 26 Jul 2021 14:23:29 +0000 (15:23 +0100)]
docs/devel/build-system.rst: Format literals correctly
In rST markup, single backticks `like this` represent "interpreted
text", which can be handled as a bunch of different things if tagged
with a specific "role":
https://docutils.sourceforge.io/docs/ref/rst/restructuredtext.html#interpreted-text
(the most common one for us is "reference to a URL, which gets
hyperlinked").
The default "role" if none is specified is "title_reference",
intended for references to book or article titles, and it renders
into the HTML as <cite>...</cite> (usually comes out as italics).
build-system.rst seems to have been written under the mistaken
assumption that single-backticks mark up literal text (function
names, etc) which should be rendered in a fixed-width font.
The rST markup for this is ``double backticks``.
Peter Maydell [Thu, 22 Jul 2021 19:20:16 +0000 (20:20 +0100)]
docs: Move licence/copyright from HTML output to rST comments
Our built HTML documentation now has a standard footer which
gives the license for QEMU (and its documentation as a whole).
In almost all pages, we either don't bother to state the
copyright/license for the individual rST sources, or we put
it in an rST comment. There are just three pages which render
copyright or license information into the user-visible HTML.
Quoting a specific (different) license for an individual HTML
page within the manual is confusing. Downgrade the license
and copyright info to a comment within the rST source, bringing
these pages in line with the rest of our documents.
Peter Maydell [Thu, 22 Jul 2021 19:20:15 +0000 (20:20 +0100)]
docs: Remove stale TODO comments about license and version
Since commits 13f934e79fa and 3a50c8f3067aaf, our HTML docs include a
footer to all pages stating the license and version. We can
therefore delete the TODO comments suggesting we should do that from
our .rst files.
Peter Maydell [Thu, 22 Jul 2021 18:09:51 +0000 (19:09 +0100)]
MAINTAINERS: Don't list Andrzej Zaborowski for various components
Andrzej Zaborowski is listed as an "Odd Fixes" maintainer for the
nSeries, Palm and PXA2XX boards, as well as the "Maintained" status
Arm 32-bit TCG backend.
Andrzej's last email to qemu-devel was back in 2017, and the email
before that was all the way back in 2013. We don't really need to
fill his email up with CCs on QEMU patches any more...
Remove Andrzej from the various boards sections (leaving them still
Odd Fixes with me as the backup patch reviewer). Add Richard
Henderson as the maintainer for the Arm TCG backend, since removing
Andrzej would otherwise leave that section with no M: line at all.
Peter Maydell [Mon, 2 Aug 2021 08:47:07 +0000 (09:47 +0100)]
Merge remote-tracking branch 'remotes/jasowang/tags/net-pull-request' into staging
# gpg: Signature made Mon 02 Aug 2021 05:23:19 BST
# gpg: using RSA key EF04965B398D6211
# gpg: Good signature from "Jason Wang (Jason Wang on RedHat) <[email protected]>" [marginal]
# gpg: WARNING: This key is not certified with sufficiently trusted signatures!
# gpg: It is not certain that the signature belongs to the owner.
# Primary key fingerprint: 215D 46F4 8246 689E C77F 3562 EF04 965B 398D 6211
* remotes/jasowang/tags/net-pull-request:
hw/net: e1000e: Don't zero out the VLAN tag in the legacy RX descriptor
hw/net: e1000e: Correct the initial value of VET register
hw/net: e1000: Correct the initial value of VET register
hw/net/can: sja1000 fix buff2frame_bas and buff2frame_pel when dlc is out of std CAN 8 bytes
hw/net/vmxnet3: Do not abort QEMU if guest specified bad queue numbers
Christina Wang [Fri, 23 Jul 2021 07:55:12 +0000 (15:55 +0800)]
hw/net: e1000e: Don't zero out the VLAN tag in the legacy RX descriptor
In the legacy RX descriptor mode, VLAN tag was saved to d->special
by e1000e_build_rx_metadata() in e1000e_write_lgcy_rx_descr(), but
it was then zeroed out again at the end of the call, which is wrong.
Christina Wang [Fri, 23 Jul 2021 07:55:11 +0000 (15:55 +0800)]
hw/net: e1000e: Correct the initial value of VET register
The initial value of VLAN Ether Type (VET) register is 0x8100, as per
the manual and real hardware.
While Linux e1000e driver always writes VET register to 0x8100, it is
not always the case for everyone. Drivers relying on the reset value
of VET won't be able to transmit and receive VLAN frames in QEMU.
Unlike e1000 in QEMU, e1000e uses a field 'vet' in "struct E1000Core"
to cache the value of VET register, but the cache only gets updated
when VET register is written. To always get a consistent VET value
no matter VET is written or remains its reset value, drop the 'vet'
field and use 'core->mac[VET]' directly.
Christina Wang [Fri, 23 Jul 2021 07:55:10 +0000 (15:55 +0800)]
hw/net: e1000: Correct the initial value of VET register
The initial value of VLAN Ether Type (VET) register is 0x8100, as per
the manual and real hardware.
While Linux e1000 driver always writes VET register to 0x8100, it is
not always the case for everyone. Drivers relying on the reset value
of VET won't be able to transmit and receive VLAN frames in QEMU.
Thomas Huth [Wed, 21 Jul 2021 14:15:59 +0000 (16:15 +0200)]
hw/net/vmxnet3: Do not abort QEMU if guest specified bad queue numbers
QEMU should never terminate unexpectedly just because the guest is
doing something wrong like specifying wrong queue numbers. Let's
simply refuse to set the device active in this case.
* remotes/rth-gitlab/tags/pull-tcg-20210730:
target/nios2: Mark raise_exception() as noreturn
accel/tcg: Remove double bswap for helper_atomic_sto_*_mmu
Paolo Bonzini [Tue, 27 Jul 2021 16:03:16 +0000 (18:03 +0200)]
coverity-model: write models fully for non-array allocation functions
Coverity seems to have issues figuring out the properties of g_malloc0
and other non *_n functions. While this was "fixed" by removing the
custom second argument to __coverity_mark_as_afm_allocated__, inline
the code from the array-based allocation functions to avoid future
issues.
Paolo Bonzini [Wed, 28 Jul 2021 17:12:22 +0000 (19:12 +0200)]
coverity-model: constrain g_malloc/g_malloc0/g_realloc as never returning NULL
g_malloc/g_malloc0/g_realloc only return NULL if the size is 0; we do not need
to cover that in the model, and so far have expected __coverity_alloc__
to model a non-NULL return value. But that apparently does not work
anymore, so add some extra conditionals that invoke __coverity_panic__
for NULL pointers.
Paolo Bonzini [Tue, 27 Jul 2021 15:56:04 +0000 (17:56 +0200)]
coverity-model: remove model for more allocation functions
These models are not needed anymore now that Coverity does not check
anymore that the result is used with "g_free". Coverity understands
GCC attributes and uses them to detect leaks.
Paolo Bonzini [Tue, 27 Jul 2021 15:54:15 +0000 (17:54 +0200)]
coverity-model: make g_free a synonym of free
Recently, Coverity has started complaining about using g_free() to free
memory areas allocated by GLib functions not included in model.c,
such as g_strfreev. This unfortunately goes against the GLib
documentation, which suggests that g_malloc() should be matched
with g_free() and plain malloc() with free(); since GLib 2.46 however
g_malloc() is hardcoded to always use the system malloc implementation,
and g_free is just "free" plus a tracepoint. Therefore, this
should not cause any problem in practice.
Use void * for consistency with the actual function; provide a model
for MemoryRegionCache functions and for address_space_rw. These
let Coverity understand the bounds of the data that various functions
read and write even at very high levels of inlining (e.g. pci_dma_read).
Paolo Bonzini [Thu, 29 Jul 2021 14:03:43 +0000 (16:03 +0200)]
vl: stop recording -smp in QemuOpts
-readconfig is still recording SMP options in QemuOpts instead of
using machine_opts_dict. This means that SMP options from -readconfig
are ignored.
Just stop using QemuOpts for -smp, making it return false for
is_qemuopts_group. Configuration files will merge the values in
machine_opts_dict using the new function machine_merge_property.
At the same time, fix -mem-prealloc which looked at QemuOpts to find the
number of guest CPUs, which it used as the default number of preallocation
threads.
Paolo Bonzini [Thu, 29 Jul 2021 13:53:27 +0000 (15:53 +0200)]
vl: introduce machine_merge_property
It will be used to parse smp-opts config groups from configuration
files. The point to note is that it does not steal a reference
from the caller. This is better because this function will be called
from qemu_config_foreach's callback; qemu_config_foreach does not cede
its reference to the qdict to the callback, and wants to free it. To
balance that extra reference, machine_parse_property_opt now needs
a qobject_unref.
* remotes/stefanha-gitlab/tags/block-pull-request:
MAINTAINERS: Added myself as a reviewer for the NVMe Block Driver
block/io_uring: resubmit when result is -EAGAIN
MAINTAINERS: add Stefano Garzarella as io_uring reviewer
* remotes/kraxel/tags/usb-20210729-pull-request:
docs: Fold usb2.txt passthrough information into usb.rst
docs: Fold usb2.txt physical port addressing info into usb.rst
docs: Fold usb2.txt USB controller information into usb.rst
docs: Incorporate information in usb-storage.txt into rST manual
usbredir: fix free call
ci: add libusb for windows builds
usb-host: wire up timer for windows
Linux SCSI can throw spurious -EAGAIN in some corner cases in its
completion path, which will end up being the result in the completed
io_uring request.
Resubmitting such requests should allow block jobs to complete, even
if such spurious errors are encountered.
Peter Maydell [Thu, 29 Jul 2021 12:17:20 +0000 (13:17 +0100)]
Merge remote-tracking branch 'remotes/thuth-gitlab/tags/pull-request-2021-07-29' into staging
* Document GitLab custom CI/CD variables
* Fix 'when:' condition in gitlab-CI jobs
* Disable tests in the gitlab-CI that fail due to out-of-memory conditions
* Allow pushing to "staging" again for maintainers without s390x access
* Fix migration-test build failure on SPARC
* Compile without "pie" on Haiku
* remotes/thuth-gitlab/tags/pull-request-2021-07-29:
configure script fix for Haiku
tests: Fix migration-test build failure for sparc
gitlab-ci.d/custom-runners: Improve rules for the staging branch
gitlab-ci.d/buildtest: Mark the aarch64 and ppc64-s390x CFI jobs as manual
gitlab-ci.d/buildtest: Disable iotests 197 and 215
gitlab-ci: Fix 'when:' condition in OpenSBI jobs
gitlab-ci: Fix 'when:' condition in EDK2 jobs
gitlab-ci: Fix 'when:' condition in acceptance_test_job_template
docs: Document GitLab custom CI/CD variables
* remotes/dg-gitlab/tags/ppc-for-6.1-20210729:
kvm: ppc: Print meaningful message on KVM_CREATE_VM failure
ppc/vof: Fix Coverity issues
target/ppc: Ease L=0 requirement on cmp/cmpi/cmpl/cmpli for ppc32
i2c/smbus_eeprom: Add feature bit to SPD data
ppc/pegasos2: Fix spurious warning with -bios
data might point into the middle of a larger buffer, there is a separate
free_on_destroy pointer passed into bufp_alloc() to handle that. It is
only used in the normal workflow though, not when dropping packets due
to the queue being full. Fix that.
libvhost-user: fix -Werror=format= warnings with __u64 fields
../subprojects/libvhost-user/libvhost-user.c:1070:12: error: format ‘%lx’ expects argument of type ‘long unsigned int’, but argument 3 has type ‘__u64’ {aka ‘long long unsigned int’} [-Werror=format=]
1070 | DPRINT(" desc_user_addr: 0x%016" PRIx64 "\n", vra->desc_user_addr);
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ~~~~~~~~~~~~~~~~~~~
| |
| __u64 {aka long long unsigned int}
Rather than using %llx, which may fail if __u64 is declared differently
elsewhere, let's just cast the values. Feel free to propose a better solution!
meson: fix meson 0.58 warning with libvhost-user subproject
Meson now checks that subprojects do not access files from parent
project. While we all agree this is best practice, libvhost-user also
want to share a few headers with QEMU, and libvhost-user isn't really a
standalone project at this point (although this is making the dependency
a bit more explicit).
target/i386: Added consistency checks for event injection
VMRUN exits with SVM_EXIT_ERR if either:
* The event injected has a reserved type.
* When the event injected is of type 3 (exception), and the vector that
has been specified does not correspond to an exception.
This does not fix the entire exc_inj test in kvm-unit-tests.
ERROR: configure test passed without -Werror but failed with -Werror.
This is probably a bug in the configure script. The failing command
will be at the bottom of config.log.
You can run configure with --disable-werror to bypass this check.
What we really want from these two tests is whether the
entire code sequence is supported, including pragmas.
Adding -Werror makes the test properly fail for clang.
When searching for options like -n in MAKEFLAGS, current code may result
in a false positive match when make is invoked with long options like
--no-print-directory. This has been observed with certain versions of
host make (e.g. 3.82) while building the Qemu package in buildroot.
Filter out such long options before searching for one-character options.
Coverity reports potential NULL pointer dereference in
get_supported_hv_cpuid_legacy() when 'cs->kvm_state' is NULL. While
'cs->kvm_state' can indeed be NULL in hv_cpuid_get_host(),
kvm_hyperv_expand_features() makes sure that it only happens when
KVM_CAP_SYS_HYPERV_CPUID is supported and KVM_CAP_SYS_HYPERV_CPUID
implies KVM_CAP_HYPERV_CPUID so get_supported_hv_cpuid_legacy() is
never really called. Add asserts to strengthen the protection against
broken KVM behavior.
Signed-off-by: Richard Zak <[email protected]>
Message-Id: <CAOakUfM8zMpYiAEn-_f9s1DHdVB-Bq9fGMM=Hfr8hJW9ra6aWw@mail.gmail.com>
[thuth: Tweaked title and patch description] Signed-off-by: Thomas Huth <[email protected]>
Peter Xu [Wed, 28 Jul 2021 21:41:28 +0000 (17:41 -0400)]
tests: Fix migration-test build failure for sparc
Even if <linux/kvm.h> seems to exist for all archs on linux, however including
it with __linux__ defined seems to be not working yet as it'll try to include
asm/kvm.h and that can be missing for archs that do not support kvm.
To fix this (instead of any attempt to fix linux headers..), we can mark the
header to be x86_64 only, because it's so far only service for adding the kvm
dirty ring test.
Thomas Huth [Wed, 28 Jul 2021 17:38:57 +0000 (19:38 +0200)]
gitlab-ci.d/custom-runners: Improve rules for the staging branch
If maintainers are currently pushing to a branch called "staging"
in their repository, they are ending up with some stuck jobs - unless
they have a s390x CI runner machine available. That's ugly, we should
make sure that the related jobs are really only started if such a
runner is available. So let's only run these jobs if it's the
"staging" branch of the main repository of the QEMU project (where
we can be sure that the s390x runner is available), or if the user
explicitly set a S390X_RUNNER_AVAILABLE variable in their CI configs
to declare that they have such a runner available, too.
Thomas Huth [Wed, 28 Jul 2021 07:09:14 +0000 (09:09 +0200)]
gitlab-ci.d/buildtest: Mark the aarch64 and ppc64-s390x CFI jobs as manual
These two jobs are currently failing very often - the linker seems to
get killed due to out-of-memory problems. Since apparently nobody has
currently an idea how to fix that nicely, let's mark the jobs as manual
for the time being until someone comes up with a proper fix.
Thomas Huth [Tue, 27 Jul 2021 16:25:42 +0000 (18:25 +0200)]
gitlab-ci.d/buildtest: Disable iotests 197 and 215
The iotests 197 and 215 are occasionally failing in the gitlab-CI now.
According to the log, the failure is "./common.rc: Killed" which might
be an indication that the process has been killed due to out-of-memory
reasons. Both tests are doing a big read with 2G that likely causes
this issue. It used to work fine in the gitlab-CI in the past, but
either the program is now requiring more free memory, or the the CI
containers have changed, so that the OOM condition now sometimes occurs.
Anyway, these two tests are not really suitable for CI containers if
they are doing things like huge reads (which is likely also the reason
why they haven't been added to the "auto" group in the past), so let's
simply disable them in the gitlab-CI now, too.
Jobs depending on another should not use the 'when: always'
condition, because if a dependency failed we should not keep
running jobs depending on it. The correct condition is
'when: on_success'.
Jobs depending on another should not use the 'when: always'
condition, because if a dependency failed we should not keep
running jobs depending on it. The correct condition is
'when: on_success'.
gitlab-ci: Fix 'when:' condition in acceptance_test_job_template
Jobs depending on another should not use the 'when: always'
condition, because if a dependency failed we should not keep
running jobs depending on it. The correct condition is
'when: on_success'.
We introduced the QEMU_CI_AVOCADO_TESTING variable in commit f56bf4caf
("gitlab: Run Avocado tests manually (except mainstream CI)"), but
forgot to document it properly. Do it now.
kvm: ppc: Print meaningful message on KVM_CREATE_VM failure
PowerPC has two KVM types (HV, PR) that translate into three kernel
modules:
kvm.ko - common kvm code
kvm_hv.ko - kvm running with MSR_HV=1 or MSR_HV|PR=0 in a nested guest.
kvm_pr.ko - kvm running in usermode MSR_PR=1.
Since the two KVM types can both be running at the same time, this
creates a situation in which it is possible for one or both of the
modules to fail to initialize, leaving the generic one behind. This
leads QEMU to think it can create a guest, but KVM will fail when
calling the type-specific code:
Ideally this would be solved kernel-side, but it might be a while
until we can get rid of one of the modules. So in the meantime this
patch tries to make this less confusing for the end user by adding a
more elucidative message:
ioctl(KVM_CREATE_VM) failed: 22 Invalid argument
PPC KVM module is not loaded. Try 'modprobe kvm_hv'.
Coverity reported issues which are caused by mixing of signed return codes
from DTC and unsigned return codes of the client interface.
This introduces PROM_ERROR and makes distinction between the error types.
This fixes NEGATIVE_RETURNS, OVERRUN issues reported by Coverity.
This adds a comment about the return parameters number in the VOF hcall.
The reason for such counting is to keep the numbers look the same in
vof_client_handle() and the Linux (an OF client).
vmc->client_architecture_support() returns target_ulong and we want to
propagate this to the client (for example H_MULTI_THREADS_ACTIVE).
The VOF path to do_client_architecture_support() needs chopping off
the top 32bit but SLOF's H_CAS does not; and either way the return values
are either 0 or 32bit negative error code. For now this chops
the top 32bits.
This makes "claim" fail if the allocated address is above 4GB as
the client interface is 32bit. This still allows claiming memory above
4GB as potentially initrd can be put there and the client can read
the address from the FDT's "available" property.
target/ppc: Ease L=0 requirement on cmp/cmpi/cmpl/cmpli for ppc32
In commit 8f0a4b6a9b, we started to require L=0 for ppc32 to match what
The Programming Environments Manual say:
"For 32-bit implementations, the L field must be cleared, otherwise
the instruction form is invalid."
The stricter behavior, however, broke AROS boot on sam460ex, which is a
regression from 6.0. This patch partially reverts the change, raising
the exception only for CPUs known to require L=0 (e500 and e500mc) and
logging a guest error for other cases.
Both behaviors are acceptable by the PowerISA, which allows "the system
illegal instruction error handler to be invoked or yield boundedly
undefined results."
Add the differential clock input feature bit to the generated SPD
data. Most guests don't seem to care but pegasos2 firmware version 1.2
checks for this bit and stops with unsupported module type error if
it's not present. Since this feature is likely present on real memory
modules add it in the general code rather than patching the generated
SPD data in pegasos2 board only.
The -append option is currently not compatible with -bios (as we don't
yet emulate nvram so we can only put it in the environment with VOF).
Therefore a warning is printed if -append is used with -bios but
because the default value of kernel_cmdline seems to be an empty
string instead of NULL this warning was printed even without -append
when -bios is used. Only print warning if -append is given.
Peter Maydell [Tue, 27 Jul 2021 14:32:00 +0000 (15:32 +0100)]
Merge remote-tracking branch 'remotes/pmaydell/tags/pull-target-arm-20210727' into staging
target-arm queue:
* hw/arm/smmuv3: Check 31st bit to see if CD is valid
* qemu-options.hx: Fix formatting of -machine memory-backend option
* hw: aspeed_gpio: Fix memory size
* hw/arm/nseries: Display hexadecimal value with '0x' prefix
* Add sve-default-vector-length cpu property
* docs: Update path that mentions deprecated.rst
* hw/intc/armv7m_nvic: for v8.1M VECTPENDING hides S exceptions from NS
* hw/intc/armv7m_nvic: Correct size of ICSR.VECTPENDING
* hw/intc/armv7m_nvic: ISCR.ISRPENDING is set for non-enabled pending interrupts
* target/arm: Report M-profile alignment faults correctly to the guest
* target/arm: Add missing 'return's after calling v7m_exception_taken()
* target/arm: Enforce that M-profile SP low 2 bits are always zero
* remotes/pmaydell/tags/pull-target-arm-20210727:
hw: aspeed_gpio: Fix memory size
hw/arm/nseries: Display hexadecimal value with '0x' prefix
target/arm: Add sve-default-vector-length cpu property
target/arm: Export aarch64_sve_zcr_get_valid_len
target/arm: Correctly bound length in sve_zcr_get_valid_len
docs: Update path that mentions deprecated.rst
hw/intc/armv7m_nvic: for v8.1M VECTPENDING hides S exceptions from NS
hw/intc/armv7m_nvic: Correct size of ICSR.VECTPENDING
hw/intc/armv7m_nvic: ISCR.ISRPENDING is set for non-enabled pending interrupts
target/arm: Report M-profile alignment faults correctly to the guest
target/arm: Add missing 'return's after calling v7m_exception_taken()
target/arm: Enforce that M-profile SP low 2 bits are always zero
qemu-options.hx: Fix formatting of -machine memory-backend option
hw/arm/smmuv3: Check 31st bit to see if CD is valid
# gpg: Signature made Mon 26 Jul 2021 20:18:12 BST
# gpg: using RSA key 522833AA75E2DCE6A24766C04DE1AF316D4F0DE9
# gpg: Good signature from "Klaus Jensen <[email protected]>" [unknown]
# gpg: aka "Klaus Jensen <[email protected]>" [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: DDCA 4D9C 9EF9 31CC 3468 4272 63D5 6FC5 E55D A838
# Subkey fingerprint: 5228 33AA 75E2 DCE6 A247 66C0 4DE1 AF31 6D4F 0DE9
* remotes/nvme/tags/nvme-next-pull-request:
tests/qtest/nvme-test: add mmio read test
hw/nvme: fix mmio read
hw/nvme: fix out-of-bounds reads
hw/nvme: use symbolic names for registers
hw/nvme: split pmrmsc register into upper and lower
hw/nvme: fix controller hot unplugging
tests/qtest/nvme-test: add persistent memory region test
hw/nvme: error handling for too many mappings
hw/nvme: unregister controller with subsystem at exit
hw/nvme: mark nvme-subsys non-hotpluggable
hw/nvme: remove NvmeCtrl parameter from ns setup/check functions
Joel Stanley [Tue, 13 Jul 2021 06:58:52 +0000 (16:28 +0930)]
hw: aspeed_gpio: Fix memory size
The macro used to calculate the maximum memory size of the MMIO region
had a mistake, causing all GPIO models to create a mapping of 0x9D8.
The intent was to have it be 0x9D8 - 0x800.
This extra size doesn't matter on ast2400 and ast2500, which have a 4KB
region set aside for the GPIO controller.
On the ast2600 the 3.3V and 1.8V GPIO controllers are 2KB apart, so the
regions would overlap. Worse was the 1.8V controller would map over the
top of the following peripheral, which happens to be the RTC.
The mmio region used by each device is a maximum of 2KB, so avoid the
calculations and hard code this as the maximum.
target/arm: Add sve-default-vector-length cpu property
Mirror the behavour of /proc/sys/abi/sve_default_vector_length
under the real linux kernel. We have no way of passing along
a real default across exec like the kernel can, but this is a
decent way of adjusting the startup vector length of a process.
Resolves: https://gitlab.com/qemu-project/qemu/-/issues/482 Signed-off-by: Richard Henderson <[email protected]> Reviewed-by: Peter Maydell <[email protected]>
Message-id: 20210723203344[email protected]
[PMM: tweaked docs formatting, document -1 special-case,
added fixup patch from RTH mentioning QEMU's maximum veclen.] Signed-off-by: Peter Maydell <[email protected]>
target/arm: Correctly bound length in sve_zcr_get_valid_len
Currently, our only caller is sve_zcr_len_for_el, which has
already masked the length extracted from ZCR_ELx, so the
masking done here is a nop. But we will shortly have uses
from other locations, where the length will be unmasked.
Saturate the length to ARM_MAX_VQ instead of truncating to
the low 4 bits.
Peter Maydell [Fri, 23 Jul 2021 16:21:46 +0000 (17:21 +0100)]
hw/intc/armv7m_nvic: for v8.1M VECTPENDING hides S exceptions from NS
In Arm v8.1M the VECTPENDING field in the ICSR has new behaviour: if
the register is accessed NonSecure and the highest priority pending
enabled exception (that would be returned in the VECTPENDING field)
targets Secure, then the VECTPENDING field must read 1 rather than
the exception number of the pending exception. Implement this.
Peter Maydell [Fri, 23 Jul 2021 16:21:45 +0000 (17:21 +0100)]
hw/intc/armv7m_nvic: Correct size of ICSR.VECTPENDING
The VECTPENDING field in the ICSR is 9 bits wide, in bits [20:12] of
the register. We were incorrectly masking it to 8 bits, so it would
report the wrong value if the pending exception was greater than 256.
Fix the bug.
Peter Maydell [Fri, 23 Jul 2021 16:21:44 +0000 (17:21 +0100)]
hw/intc/armv7m_nvic: ISCR.ISRPENDING is set for non-enabled pending interrupts
The ISCR.ISRPENDING bit is set when an external interrupt is pending.
This is true whether that external interrupt is enabled or not.
This means that we can't use 's->vectpending == 0' as a shortcut to
"ISRPENDING is zero", because s->vectpending indicates only the
highest priority pending enabled interrupt.
Remove the incorrect optimization so that if there is no pending
enabled interrupt we fall through to scanning through the whole
interrupt array.
Peter Maydell [Fri, 23 Jul 2021 16:21:43 +0000 (17:21 +0100)]
target/arm: Report M-profile alignment faults correctly to the guest
For M-profile, we weren't reporting alignment faults triggered by the
generic TCG code correctly to the guest. These get passed into
arm_v7m_cpu_do_interrupt() as an EXCP_DATA_ABORT with an A-profile
style exception.fsr value of 1. We didn't check for this, and so
they fell through into the default of "assume this is an MPU fault"
and were reported to the guest as a data access violation MPU fault.
Report these alignment faults as UsageFaults which set the UNALIGNED
bit in the UFSR.
Peter Maydell [Fri, 23 Jul 2021 16:21:42 +0000 (17:21 +0100)]
target/arm: Add missing 'return's after calling v7m_exception_taken()
In do_v7m_exception_exit(), we perform various checks as part of
performing the exception return. If one of these checks fails, the
architecture requires that we take an appropriate exception on the
existing stackframe. We implement this by calling
v7m_exception_taken() to set up to take the new exception, and then
immediately returning from do_v7m_exception_exit() without proceeding
any further with the unstack-and-exception-return process.
In a couple of checks that are new in v8.1M, we forgot the "return"
statement, with the effect that if bad code in the guest tripped over
these checks we would set up to take a UsageFault exception but then
blunder on trying to also unstack and return from the original
exception, with the probable result that the guest would crash.
Peter Maydell [Fri, 23 Jul 2021 16:21:41 +0000 (17:21 +0100)]
target/arm: Enforce that M-profile SP low 2 bits are always zero
For M-profile, unlike A-profile, the low 2 bits of SP are defined to be
RES0H, which is to say that they must be hardwired to zero so that
guest attempts to write non-zero values to them are ignored.
Implement this behaviour by masking out the low bits:
* for writes to r13 by the gdbstub
* for writes to any of the various flavours of SP via MSR
* for writes to r13 via store_reg() in generated code
Note that all the direct uses of cpu_R[] in translate.c are in places
where the register is definitely not r13 (usually because that has
been checked for as an UNDEFINED or UNPREDICTABLE case and handled as
UNDEF).
All the other writes to regs[13] in C code are either:
* A-profile only code
* writes of values we can guarantee to be aligned, such as
- writes of previous-SP-value plus or minus a 4-aligned constant
- writes of the value in an SP limit register (which we already
enforce to be aligned)
Peter Maydell [Mon, 19 Jul 2021 10:52:57 +0000 (11:52 +0100)]
qemu-options.hx: Fix formatting of -machine memory-backend option
The documentation of the -machine memory-backend has some minor
formatting errors:
* Misindentation of the initial line meant that the whole option
section is incorrectly indented in the HTML output compared to
the other -machine options
* The examples weren't indented, which meant that they were formatted
as plain run-on text including outputting the "::" as text.
* The a) b) list has no rst-format markup so it is rendered as
a single run-on paragraph
Peter Maydell [Tue, 27 Jul 2021 09:55:50 +0000 (10:55 +0100)]
Merge remote-tracking branch 'remotes/dgilbert-gitlab/tags/pull-migration-20210726a' into staging
Migration fixes 2021-07-26
Peter's fix for a bunch of races
-> Seem to fix the occasional crash seen by Peter
Wei's fix for migration with free page hinting
-> Bug has been around for a while, but makes a huge difference
My fix for OpenBSD test corner case
Signed-off-by: Dr. David Alan Gilbert <[email protected]>
# gpg: Signature made Mon 26 Jul 2021 13:42:16 BST
# gpg: using RSA key 45F5C71B4A0CB7FB977A9FA90516331EBC5BFDE7
# gpg: Good signature from "Dr. David Alan Gilbert (RH2) <[email protected]>" [full]
# Primary key fingerprint: 45F5 C71B 4A0C B7FB 977A 9FA9 0516 331E BC5B FDE7
* remotes/dgilbert-gitlab/tags/pull-migration-20210726a:
migration: clear the memory region dirty bitmap when skipping free pages
migration: Move the yank unregister of channel_close out
migration: Teach QEMUFile to be QIOChannel-aware
migration: Introduce migration_ioc_[un]register_yank()
migration: Make from_dst_file accesses thread-safe
migration: Fix missing join() of rp_thread
tests/qtest/migration-test.c: use 127.0.0.1 instead of 0
* remotes/rth-gitlab/tags/pull-tcg-20210726:
tests/unit: Remove unused variable from test_io
linux-user/syscall: Remove unused variable from execve
hw/pci-hist/pnv_phb4: Fix typo in pnv_phb4_ioda_write
hw/ppc/spapr_events: Remove unused variable from check_exception
hw/audio/adlib: Remove unused variable in adlib_callback
net/checksum: Remove unused variable in net_checksum_add_iov
util/selfmap: Discard mapping on error
accel/tcg: Remove unused variable in cpu_exec
nbd/server: Mark variable unused in nbd_negotiate_meta_queries
bitops.h: revert db1ffc32dd ("qemu/bitops.h: add bitrev8 implementation")
accel/tcg: Remove unnecessary check on icount_extra in cpu_loop_exec_tb()
accel/tcg: Don't use CF_COUNT_MASK as the max value of icount_decr.u16.low
projects / qemu.git / log