contrib/gitdm: add domain-map for Crudebyte

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Christian Schoenebeck <qemu_oss@crudebyte.com>
Message-Id: <20210720232703.10650-15-alex.bennee@linaro.org>

contrib/gitdm: un-ironically add a mapping for LWN

I think this mainly comes from kernel-doc stuff imported into the QEMU
tree.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Cc: Jonathan Corbet <corbet@lwn.net>
Message-Id: <20210720232703.10650-14-alex.bennee@linaro.org>

contrib/gitdm: add domain-map/group-map for Wind River

As per discussion at:
  http://patchwork.ozlabs.org/project/qemu-devel/patch/20201004180443.2035359-19-f4bug@amsat.org/

I've added Bin's personal email as an individual contributor.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Acked-by: Bin Meng <bin.meng@windriver.com>
Cc: Ruimei Yan <ruimei.yan@windriver.com>
Cc: Xuzhou Cheng <xuzhou.cheng@windriver.com>
Message-Id: <20210720232703.10650-13-alex.bennee@linaro.org>

contrib/gitdm: add domain-map for Eldorado

Luis acked on IRC:

  #qemu@znc-oftc_2021-07-13.txt:[15:00:02] <lffpires> stsquad: "eldorado.org.br Eldorado" is fine

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Acked-by: Luis Pires <luis.pires@eldorado.org.br>
Message-Id: <20210720232703.10650-12-alex.bennee@linaro.org>

contrib/gitdm: add domain-map/group-map mappings for Samsung

Minwoo's work from their personal address are treated as personal
contributions.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Acked-by: Klaus Jensen <k.jensen@samsung.com>
Cc: Gollu Appalanaidu <anaidu.gollu@samsung.com>
Cc: Minwoo Im <minwoo.im.dev@gmail.com>
Message-Id: <20210720232703.10650-11-alex.bennee@linaro.org>

gitdm.config: sort the corporate GroupMap entries

Lets try and keep them that way.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-Id: <20210720232703.10650-10-alex.bennee@linaro.org>

contrib/gitdm: add a group mapping for robot scanners

This mostly affects Reported-by: tags

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-Id: <20210720232703.10650-9-alex.bennee@linaro.org>

contrib/gitdm: add domain-map for MontaVista

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Acked-by: Corey Minyard <cminyard@mvista.com>
Message-Id: <20210720232703.10650-8-alex.bennee@linaro.org>

.mailmap: fix up some broken commit authors

Fixes: 49a6f3bffb ("target/arm: Correct the encoding of MDCCSR_EL0 and DBGDSCRint")
Fixes: 5a07192a04 ("target/i386: Fix handling of k_gs_base register in 32-bit mode in gdbstub")
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Cc: Nick Hudson <hnick@vmware.com>
Cc: Marek Dolata <mkdolata@us.ibm.com>
Message-Id: <20210720232703.10650-7-alex.bennee@linaro.org>

contrib/gitdm: add some new aliases to fix up commits

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Cc: Yuval Shaia <yuval.shaia.ml@gmail.com>
Message-Id: <20210720232703.10650-6-alex.bennee@linaro.org>

configure: remove needless if leg

It was pointed out in review of the previous patch that the if leg
isn't needed as the for loop will not enter on an empty $device_archs.

Fixes: d1d5e9eefd ("configure: allow the selection of alternate config in the build")
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20210720232703.10650-5-alex.bennee@linaro.org>

docs: add a section on the generalities of vhost-user

While we do mention some of this stuff in the various daemons and
manuals the subtleties of the socket and memory sharing are sometimes
missed. This document attempts to give some background on vhost-user
daemons in general terms.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20210720232703.10650-4-alex.bennee@linaro.org>

docs: collect the disparate device emulation docs into one section

While we are at it add a brief preamble that explains some of the
common concepts in QEMU's device emulation which will hopefully lead
to less confusing about our dizzying command line options.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Cc: Daniel P. Berrangé <berrange@redhat.com>
Cc: Eduardo Habkost <ehabkost@redhat.com>
Message-Id: <20210720232703.10650-3-alex.bennee@linaro.org>

gitignore: Update with some filetypes

Update .gitignore to ignore .swp and .patch files.

Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Message-Id: <79262dbe1f7888eb02e1911501eebafa6f2f6400.1616583806.git.viresh.kumar@linaro.org>
Message-Id: <20210720232703.10650-2-alex.bennee@linaro.org>

Merge remote-tracking branch 'remotes/kraxel/tags/vga-20210723-pull-request' into staging

vga: fixes for qxl and virtio-gpu

# gpg: Signature made Fri 23 Jul 2021 06:54:34 BST
# gpg:                using RSA key A0328CFFB93A17A79901FE7D4CB6D8EED3E87138
# gpg: Good signature from "Gerd Hoffmann (work) <kraxel@redhat.com>" [full]
# gpg:                 aka "Gerd Hoffmann <gerd@kraxel.org>" [full]
# gpg:                 aka "Gerd Hoffmann (private) <kraxel@gmail.com>" [full]
# Primary key fingerprint: A032 8CFF B93A 17A7 9901  FE7D 4CB6 D8EE D3E8 7138

* remotes/kraxel/tags/vga-20210723-pull-request:
  hw/display: fix virgl reset regression
  vl: add virtio-vga-gl to the default_list
  hw/display: fail early when multiple virgl devices are requested
  Revert "qxl: add migration blocker to avoid pre-save assert"
  qxl: remove assert in qxl_pre_save.
  hw/display/virtio-gpu: Fix memory leak (CID 1453811)

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

Merge remote-tracking branch 'remotes/bonzini-gitlab/tags/for-upstream' into staging

Bugfixes.

# gpg: Signature made Thu 22 Jul 2021 14:11:27 BST
# gpg:                using RSA key F13338574B662389866C7682BFFBD25F78C7AE83
# gpg:                issuer "pbonzini@redhat.com"
# gpg: Good signature from "Paolo Bonzini <bonzini@gnu.org>" [full]
# gpg:                 aka "Paolo Bonzini <pbonzini@redhat.com>" [full]
# Primary key fingerprint: 46F5 9FBD 57D6 12E7 BFD4  E2F7 7E15 100C CD36 69B1
#      Subkey fingerprint: F133 3857 4B66 2389 866C  7682 BFFB D25F 78C7 AE83

* remotes/bonzini-gitlab/tags/for-upstream:
  configure: Let --without-default-features disable vhost-kernel and vhost-vdpa
  configure: Fix the default setting of the "xen" feature
  configure: Allow vnc to get disabled with --without-default-features
  configure: Fix --without-default-features propagation to meson
  meson: fix dependencies for modinfo
  configure: Drop obsolete check for the alloc_size attribute
  target/i386: Added consistency checks for EFER
  target/i386: Added consistency checks for CR4
  target/i386: Added V_INTR_PRIO check to virtual interrupts
  qemu-config: restore "machine" in qmp_query_command_line_options()
  usb: fix usb-host dependency check
  chardev-spice: add missing module_obj directive
  vl: Parse legacy default_machine_opts
  qemu-config: fix memory leak on ferror()
  qemu-config: never call the callback after an error, fix leak

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

hw/display: fix virgl reset regression

Before commit 49afbca3b00e8e517d54964229a794b51768deaf ("virtio-gpu: drop
use_virgl_renderer"), use_virgl_renderer was preventing calling GL
functions from non-GL context threads. The innocuously looking

  g->parent_obj.use_virgl_renderer = false;

was set the first time virtio_gpu_gl_reset() was called, during
pc_machine_reset() in the main thread. Further virtio_gpu_gl_reset()
calls in IO threads, without associated GL context, were thus skipping
GL calls and avoided warnings or crashes (see also
https://gitlab.freedesktop.org/virgl/virglrenderer/-/issues/226).

Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-Id: <20210702123221.942432-1-marcandre.lureau@redhat.com>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>

vl: add virtio-vga-gl to the default_list

Do not instantiate an extra default VGA device if -device virtio-vga-gl
is provided.

Related to commit b36eb8860f8f4a9c6f131c3fd380116a3017e022 ("virtio-gpu:
add virtio-vga-gl")

Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-Id: <20210701062421.721414-1-marcandre.lureau@redhat.com>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>

hw/display: fail early when multiple virgl devices are requested

This avoids failing to initialize virgl and crashing later on, and clear
the user expectations.

Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>
Message-Id: <20210705104218.1161101-1-marcandre.lureau@redhat.com>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>

Revert "qxl: add migration blocker to avoid pre-save assert"

This reverts commit 86dbcdd9c7590d06db89ca256c5eaf0b4aba8858.

The pre-save assert is gone now, so the migration blocker
is not needed any more.

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-Id: <20210721093347.338536-3-kraxel@redhat.com>

qxl: remove assert in qxl_pre_save.

Since commit 551dbd0846d2 ("migration: check pre_save return in
vmstate_save_state") the pre_save hook can fail.  So lets finally
use that to drop the guest-triggerable assert in qxl_pre_save().

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-Id: <20210721093347.338536-2-kraxel@redhat.com>

Merge remote-tracking branch 'remotes/rth-gitlab/tags/pull-tcg-20210721' into staging

Atomic build fixes for clang-12
Breakpoint reorg

# gpg: Signature made Wed 21 Jul 2021 20:57:50 BST
# gpg:                using RSA key 7A481E78868B4DB6A85A05C064DF38E8AF7E215F
# gpg:                issuer "richard.henderson@linaro.org"
# gpg: Good signature from "Richard Henderson <richard.henderson@linaro.org>" [full]
# Primary key fingerprint: 7A48 1E78 868B 4DB6 A85A  05C0 64DF 38E8 AF7E 215F

* remotes/rth-gitlab/tags/pull-tcg-20210721: (27 commits)
  accel/tcg: Record singlestep_enabled in tb->cflags
  accel/tcg: Hoist tb_cflags to a local in translator_loop
  accel/tcg: Remove TranslatorOps.breakpoint_check
  accel/tcg: Move breakpoint recognition outside translation
  accel/tcg: Merge tb_find into its only caller
  target/avr: Implement gdb_adjust_breakpoint
  hw/core: Introduce CPUClass.gdb_adjust_breakpoint
  target/i386: Implement debug_check_breakpoint
  target/arm: Implement debug_check_breakpoint
  hw/core: Introduce TCGCPUOps.debug_check_breakpoint
  accel/tcg: Use CF_NO_GOTO_{TB, PTR} in cpu_exec_step_atomic
  accel/tcg: Handle -singlestep in curr_cflags
  accel/tcg: Drop CF_NO_GOTO_PTR from -d nochain
  accel/tcg: Add CF_NO_GOTO_TB and CF_NO_GOTO_PTR
  target/alpha: Drop goto_tb path in gen_call_pal
  accel/tcg: Move curr_cflags into cpu-exec.c
  accel/tcg: Reduce CF_COUNT_MASK to match TCG_MAX_INSNS
  accel/tcg: Push trace info building into atomic_common.c.inc
  trace: Fold mem-internal.h into mem.h
  accel/tcg: Expand ATOMIC_MMU_LOOKUP_*
  ...

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

configure: Let --without-default-features disable vhost-kernel and vhost-vdpa

The vhost_kernel and vhost_vdpa variables should be pre-initialized with
the $default_feature setting so that these features get disabled when
the user runs the configure scripts with --without-default-features.

Reported-by: Cole Robinson <crobinso@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
Message-Id: <20210713093155.677589-5-thuth@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

configure: Fix the default setting of the "xen" feature

The "xen" variable should either contain "enabled", "disabled" or
nothing (for auto detection). But when the user currently runs the
configure script with --without-default-features, it gets set to
"no" instead. This does not work as expected, the feature will still
be enabled if the Xen headers are present. Thus set the variable
to "disabled" instead if default_feature switch has been set.

Reported-by: Cole Robinson <crobinso@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
Message-Id: <20210713093155.677589-4-thuth@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

configure: Allow vnc to get disabled with --without-default-features

There's no reason why we should keep VNC enabled when the user
specified --without-default-features.

Reported-by: Cole Robinson <crobinso@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
Message-Id: <20210713093155.677589-3-thuth@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

configure: Fix --without-default-features propagation to meson

A typo prevents that many features get disabled when the user
runs "configure" with the --without-default-features switch.

Reported-by: Cole Robinson <crobinso@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
Message-Id: <20210713093155.677589-2-thuth@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

meson: fix dependencies for modinfo

modinfo runs the preprocessor and therefore needs all generated input files
to be there.  The "depends" clause does not work in Meson 0.55.3, so for
now use "input".

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

configure: Drop obsolete check for the alloc_size attribute

We recently bumped our requirement for Clang to at least version 6.0.
And according to:

 https://releases.llvm.org/6.0.0/tools/clang/docs/AttributeReference.html

Clang v6.0 supports the alloc_size attribute. Thus we can drop this
check in the configure script now.

Signed-off-by: Thomas Huth <thuth@redhat.com>
Message-Id: <20210714072855.785566-1-thuth@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

target/i386: Added consistency checks for EFER

EFER.SVME has to be set, and EFER reserved bits must
be zero.
In addition the combinations
 * EFER.LMA or EFER.LME is non-zero and the processor does not support LM
 * non-zero EFER.LME and CR0.PG and zero CR4.PAE
 * non-zero EFER.LME and CR0.PG and zero CR0.PE
 * non-zero EFER.LME, CR0.PG, CR4.PAE, CS.L and CS.D
are all invalid.
(AMD64 Architecture Programmer's Manual, V2, 15.5)

Signed-off-by: Lara Lazier <laramglazier@gmail.com>
Message-Id: <20210721152651.14683-3-laramglazier@gmail.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

target/i386: Added consistency checks for CR4

All MBZ bits in CR4 must be zero. (APM2 15.5)
Added reserved bitmask and added checks in both
helper_vmrun and helper_write_crN.

Signed-off-by: Lara Lazier <laramglazier@gmail.com>
Message-Id: <20210721152651.14683-2-laramglazier@gmail.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

target/i386: Added V_INTR_PRIO check to virtual interrupts

The APM2 states that The processor takes a virtual INTR interrupt
if V_IRQ and V_INTR_PRIO indicate that there is a virtual interrupt pending
whose priority is greater than the value in V_TPR.

Signed-off-by: Lara Lazier <laramglazier@gmail.com>
Message-Id: <20210721152651.14683-1-laramglazier@gmail.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

qemu-config: restore "machine" in qmp_query_command_line_options()

Commit d8fb7d0969d5c32b3d1b9e20b63ec6c0abe80be4 ("vl: switch -M parsing
to keyval") stopped adding the "machine" QemuOptsList. This causes
"machine" options to not show up in QMP query-command-line-options
output. For example, libvirt cannot detect that kernel_irqchip support
is available.

Adjust the "machine" opts enumeration in
qmp_query_command_line_options() so that options are properly reported.

Fixes: d8fb7d0969d5 ("vl: switch -M parsing to keyval")
Cc: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20210721151055.424580-1-stefanha@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

usb: fix usb-host dependency check

Fixes: 90540f3289 ("configure, meson: convert libusb detection to meson", 2021-06-25)
Reported-by: Programmingkid <programmingkidx@gmail.com>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Message-Id: <20210721081718.301343-1-kraxel@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

chardev-spice: add missing module_obj directive

The chardev-spicevmc class was not listed in chardev/spice.c, causing
"-chardev spicevmc" to fail when modules are enabled.

Reported-by: Frederic Bezies <fredbezies@gmail.com>
Fixes: 9f4a0f0978 ("modules: use modinfo for qom load", 2021-07-09)
Resolves: //gitlab.com/qemu-project/qemu/-/issues/488
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Message-Id: <20210719164435.1227794-1-pbonzini@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

hw/display/virtio-gpu: Fix memory leak (CID 1453811)

To avoid leaking memory on the error path, reorder the
code as:
- check the parameters first
- check resource already existing
- finally allocate memory

Reported-by: Coverity (CID 1453811: RESOURCE_LEAK)
Fixes: e0933d91b1c ("virtio-gpu: Add virtio_gpu_resource_create_blob")
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-Id: <20210531101928.1662732-1-philmd@redhat.com>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>

Merge remote-tracking branch 'remotes/ericb/tags/pull-bitmaps-2021-07-21' into staging

block bitmaps patches for 2021-07-21

- fix 'qemu-img convert --bitmaps' handling of qcow2 files with
  inconsistent bitmaps

# gpg: Signature made Wed 21 Jul 2021 20:16:09 BST
# gpg:                using RSA key 71C2CC22B1C4602927D2F3AAA7A16B4A2527436A
# gpg: Good signature from "Eric Blake <eblake@redhat.com>" [full]
# gpg:                 aka "Eric Blake (Free Software Programmer) <ebb9@byu.net>" [full]
# gpg:                 aka "[jpeg image of size 6874]" [full]
# Primary key fingerprint: 71C2 CC22 B1C4 6029 27D2  F3AA A7A1 6B4A 2527 436A

* remotes/ericb/tags/pull-bitmaps-2021-07-21:
  qemu-img: Add --skip-broken-bitmaps for 'convert --bitmaps'
  qemu-img: Fail fast on convert --bitmaps with inconsistent bitmap
  iotests: Improve and rename test 291 to qemu-img-bitmap

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

qemu-img: Add --skip-broken-bitmaps for 'convert --bitmaps'

The point of 'qemu-img convert --bitmaps' is to be a convenience for
actions that are already possible through a string of smaller
'qemu-img bitmap' sub-commands.  One situation not accounted for
already is that if a source image contains an inconsistent bitmap (for
example, because a qemu process died abruptly before flushing bitmap
state), the user MUST delete those inconsistent bitmaps before
anything else useful can be done with the image.

We don't want to delete inconsistent bitmaps by default: although a
corrupt bitmap is only a loss of optimization rather than a corruption
of user-visible data, it is still nice to require the user to opt in
to the fact that they are aware of the loss of the bitmap.  Still,
requiring the user to check 'qemu-img info' to see whether bitmaps are
consistent, then use 'qemu-img bitmap --remove' to remove offenders,
all before using 'qemu-img convert', is a lot more work than just
adding a knob 'qemu-img convert --bitmaps --skip-broken-bitmaps' which
opts in to skipping the broken bitmaps.

After testing the new option, also demonstrate the way to manually fix
things (either deleting bad bitmaps, or re-creating them as empty) so
that it is possible to convert without the option.

Fixes: https://bugzilla.redhat.com/show_bug.cgi?id=1946084
Signed-off-by: Eric Blake <eblake@redhat.com>
Message-Id: <20210709153951.2801666-4-eblake@redhat.com>
[eblake: warning message tweak, test enhancements]
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>

qemu-img: Fail fast on convert --bitmaps with inconsistent bitmap

Waiting until the end of the convert operation (a potentially
time-consuming task) to finally detect that we can't copy a bitmap is
bad, comparing to failing fast up front.  Furthermore, this prevents
us from leaving a file behind with a bitmap that is not marked as
inconsistent even though it does not have sane contents.

This fixes the problems exposed in the previous patch to the iotest:
it adds a fast failure up front, and even if we don't fail early, it
ensures that any bitmap we add but do not properly populate is removed
again rather than left behind incomplete.

Signed-off-by: Eric Blake <eblake@redhat.com>
Message-Id: <20210709153951.2801666-3-eblake@redhat.com>
[eblake: add a hint to the warning message, simplify name computation]
Reviewed-by: Nir Soffer <nsoffer@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>

iotests: Improve and rename test 291 to qemu-img-bitmap

Enhance the test to demonstrate existing less-than-stellar behavior of
qemu-img with a qcow2 image containing an inconsistent bitmap: we
don't diagnose the problem until after copying the entire image (a
potentially long time), and when we do diagnose the failure, we still
end up leaving an empty bitmap in the destination.  This mess will be
cleaned up in the next patch.

While at it, rename the test now that we support useful iotest names,
and fix a missing newline in the error message thus exposed.

Signed-off-by: Eric Blake <eblake@redhat.com>
Message-Id: <20210709153951.2801666-2-eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Nir Soffer <nsoffer@redhat.com>

accel/tcg: Record singlestep_enabled in tb->cflags

Set CF_SINGLE_STEP when single-stepping is enabled.
This avoids the need to flush all tb's when turning
single-stepping on or off.

Tested-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

accel/tcg: Hoist tb_cflags to a local in translator_loop

The access internal to tb_cflags() is atomic.
Avoid re-reading it as such for the multiple uses.

Tested-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

accel/tcg: Remove TranslatorOps.breakpoint_check

The hook is now unused, with breakpoints checked outside translation.

Tested-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

accel/tcg: Move breakpoint recognition outside translation

Trigger breakpoints before beginning translation of a TB
that would begin with a BP.  Thus we never generate code
for the BP at all.

Single-step instructions within a page containing a BP so
that we are sure to check each insn for the BP as above.

We no longer need to flush any TBs when changing BPs.

Resolves: https://gitlab.com/qemu-project/qemu/-/issues/286
Resolves: https://gitlab.com/qemu-project/qemu/-/issues/404
Resolves: https://gitlab.com/qemu-project/qemu/-/issues/489
Tested-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

accel/tcg: Merge tb_find into its only caller

We are going to want two things:
(1) check for breakpoints will want to break out of the loop here,
(2) cflags can only be calculated with pc in hand.

Tested-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/avr: Implement gdb_adjust_breakpoint

Ensure at registration that all breakpoints are in
code space, not data space.

Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

hw/core: Introduce CPUClass.gdb_adjust_breakpoint

This will allow a breakpoint hack to move out of AVR's translator.

Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/i386: Implement debug_check_breakpoint

Return false for RF set, as we do in i386_tr_breakpoint_check.

Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

target/arm: Implement debug_check_breakpoint

Reuse the code at the bottom of helper_check_breakpoints,
which is what we currently call from *_tr_breakpoint_check.

Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

hw/core: Introduce TCGCPUOps.debug_check_breakpoint

New hook to return true when an architectural breakpoint is
to be recognized and false when it should be suppressed.

First use must wait until other pieces are in place.

Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

accel/tcg: Use CF_NO_GOTO_{TB, PTR} in cpu_exec_step_atomic

Request that the one TB returns immediately, so that
we release the exclusive lock as soon as possible.

Tested-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Message-Id: <20210717221851.2124573-7-richard.henderson@linaro.org>

accel/tcg: Handle -singlestep in curr_cflags

Exchange the test in translator_use_goto_tb for CF_NO_GOTO_TB,
and the test in tb_gen_code for setting CF_COUNT_MASK to 1.

Tested-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Message-Id: <20210717221851.2124573-6-richard.henderson@linaro.org>

accel/tcg: Drop CF_NO_GOTO_PTR from -d nochain

The purpose of suppressing goto_ptr from -d nochain had been
to return to the main loop so that -d cpu would be recognized.
But we now include -d cpu logging in helper_lookup_tb_ptr so
there is no need to exclude goto_ptr.

Tested-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Message-Id: <20210717221851.2124573-5-richard.henderson@linaro.org>

accel/tcg: Add CF_NO_GOTO_TB and CF_NO_GOTO_PTR

Move the -d nochain check to bits on tb->cflags.
These will be used for more than -d nochain shortly.

Set bits during curr_cflags, test them in translator_use_goto_tb,
assert we're not doing anything odd in tcg_gen_goto_tb.  The test
in tcg_gen_exit_tb is redundant with the assert for goto_tb_issue_mask.

Tested-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Message-Id: <20210717221851.2124573-4-richard.henderson@linaro.org>

target/alpha: Drop goto_tb path in gen_call_pal

We are certain of a page crossing here, entering the
PALcode image, so the call to use_goto_tb that should
have been here will never succeed.

We are shortly going to add an assert to tcg_gen_goto_tb
that would trigger for this case.

Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

accel/tcg: Move curr_cflags into cpu-exec.c

We will shortly have more than a simple member read here,
with stuff not necessarily exposed to exec/exec-all.h.

Tested-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-Id: <20210717221851.2124573-3-richard.henderson@linaro.org>

accel/tcg: Reduce CF_COUNT_MASK to match TCG_MAX_INSNS

The space reserved for CF_COUNT_MASK was overly large.
Reduce to free up cflags bits and eliminate an extra test.

Tested-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Message-Id: <20210717221851.2124573-2-richard.henderson@linaro.org>

accel/tcg: Push trace info building into atomic_common.c.inc

Use trace_mem_get_info instead of trace_mem_build_info,
using the TCGMemOpIdx that we already have.  Do this in
the atomic_trace_*_pre function as common subroutines.

Tested-by: Cole Robinson <crobinso@redhat.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

trace: Fold mem-internal.h into mem.h

Since the last thing that mem.h does is include mem-internal.h,
the symbols are not actually private.

Tested-by: Cole Robinson <crobinso@redhat.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

accel/tcg: Expand ATOMIC_MMU_LOOKUP_*

Unify the parameters of atomic_mmu_lookup between cputlb.c and
user-exec.c.  Call the function directly, and remove the macros.

Tested-by: Cole Robinson <crobinso@redhat.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

accel/tcg: Remove ATOMIC_MMU_DECLS

All definitions are now empty.

Tested-by: Cole Robinson <crobinso@redhat.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

accel/tcg: Fold EXTRA_ARGS into atomic_template.h

All instances of EXTRA_ARGS are now identical.

Tested-by: Cole Robinson <crobinso@redhat.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

accel/tcg: Standardize atomic helpers on softmmu api

Reduce the amount of code duplication by always passing
the TCGMemOpIdx argument to helper_atomic_*.  This is not
currently used for user-only, but it's easy to ignore.

Tested-by: Cole Robinson <crobinso@redhat.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

tcg: Rename helper_atomic_*_mmu and provide for user-only

Always provide the atomic interface using TCGMemOpIdx oi
and uintptr_t retaddr.  Rename from helper_* to cpu_* so
as to (mostly) match the exec/cpu_ldst.h functions, and
to emphasize that they are not callable from TCG directly.

Tested-by: Cole Robinson <crobinso@redhat.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

qemu/atomic: Add aligned_{int64,uint64}_t types

Use it to avoid some clang-12 -Watomic-alignment errors,
forcing some structures to be aligned and as a pointer when
we have ensured that the address is aligned.

Tested-by: Cole Robinson <crobinso@redhat.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

qemu/atomic: Remove pre-C11 atomic fallbacks

We now require c11, so the fallbacks are now dead code

Tested-by: Cole Robinson <crobinso@redhat.com>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

qemu/atomic: Use macros for CONFIG_ATOMIC64

Clang warnings about questionable atomic usage get localized
to the inline function in atomic.h.  By using a macro, we get
the full traceback to the original use that caused the warning.

Tested-by: Cole Robinson <crobinso@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

Update version for v6.1.0-rc0 release

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

Merge remote-tracking branch 'remotes/stefanha-gitlab/tags/block-pull-request' into staging

Pull request

Stefano's performance regression fix for commit 2558cb8dd4 ("linux-aio:
increasing MAX_EVENTS to a larger hardcoded value").

# gpg: Signature made Wed 21 Jul 2021 14:12:47 BST
# gpg:                using RSA key 8695A8BFD3F97CDAAC35775A9CA4ABB381AB73C8
# gpg: Good signature from "Stefan Hajnoczi <stefanha@redhat.com>" [full]
# gpg:                 aka "Stefan Hajnoczi <stefanha@gmail.com>" [full]
# Primary key fingerprint: 8695 A8BF D3F9 7CDA AC35  775A 9CA4 ABB3 81AB 73C8

* remotes/stefanha-gitlab/tags/block-pull-request:
  linux-aio: limit the batch size using `aio-max-batch` parameter
  iothread: add aio-max-batch parameter
  iothread: generalize iothread_set_param/iothread_get_param

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

linux-aio: limit the batch size using `aio-max-batch` parameter

When there are multiple queues attached to the same AIO context,
some requests may experience high latency, since in the worst case
the AIO engine queue is only flushed when it is full (MAX_EVENTS) or
there are no more queues plugged.

Commit 2558cb8dd4 ("linux-aio: increasing MAX_EVENTS to a larger
hardcoded value") changed MAX_EVENTS from 128 to 1024, to increase
the number of in-flight requests. But this change also increased
the potential maximum batch to 1024 elements.

When there is a single queue attached to the AIO context, the issue
is mitigated from laio_io_unplug() that will flush the queue every
time is invoked since there can't be others queue plugged.

Let's use the new `aio-max-batch` IOThread parameter to mitigate
this issue, limiting the number of requests in a batch.

We also define a default value (32): this value is obtained running
some benchmarks and it represents a good tradeoff between the latency
increase while a request is queued and the cost of the io_submit(2)
system call.

Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Message-id: 20210721094211.69853-4-sgarzare@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>

iothread: add aio-max-batch parameter

The `aio-max-batch` parameter will be propagated to AIO engines
and it will be used to control the maximum number of queued requests.

When there are in queue a number of requests equal to `aio-max-batch`,
the engine invokes the system call to forward the requests to the kernel.

This parameter allows us to control the maximum batch size to reduce
the latency that requests might accumulate while queued in the AIO
engine queue.

If `aio-max-batch` is equal to 0 (default value), the AIO engine will
use its default maximum batch size value.

Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Message-id: 20210721094211.69853-3-sgarzare@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>

iothread: generalize iothread_set_param/iothread_get_param

Changes in preparation for next patches where we add a new
parameter not related to the poll mechanism.

Let's add two new generic functions (iothread_set_param and
iothread_get_param) that we use to set and get IOThread
parameters.

Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
Message-id: 20210721094211.69853-2-sgarzare@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>

Merge remote-tracking branch 'remotes/cleber-gitlab/tags/python-next-pull-request' into staging

Acceptance Tests

- Fix for tests/acceptance/virtio-gpu.py to match the change in device
  name
- Fix for failure caught by tests/acceptance/multiprocess.py

PS: While not a maintainer for the subsystem in PATCH 7, I'm including
it as a one-off to facilitate the landing of the fix as discussed in
the mailing list.

# gpg: Signature made Wed 21 Jul 2021 00:26:17 BST
# gpg:                using RSA key 7ABB96EB8B46B94D5E0FE9BB657E8D33A5F209F3
# gpg: Good signature from "Cleber Rosa <crosa@redhat.com>" [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: 7ABB 96EB 8B46 B94D 5E0F  E9BB 657E 8D33 A5F2 09F3

* remotes/cleber-gitlab/tags/python-next-pull-request:
  remote/memory: Replace share parameter with ram_flags
  tests/acceptance/virtio-gpu.py: provide kernel and initrd hashes
  tests/acceptance/virtio-gpu.py: use virtio-vga-gl
  tests/acceptance/virtio-gpu.py: combine kernel command line
  tests/acceptance/virtio-gpu.py: combine CPU tags
  tests/acceptance/virtio-gpu.py: combine x86_64 arch tags
  tests/acceptance/virtio-gpu.py: use require_accelerator()

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

remote/memory: Replace share parameter with ram_flags

Fixes: d5015b801340 ("softmmu/memory: Pass ram_flags to
qemu_ram_alloc_from_fd()")

Signed-off-by: Yang Zhong <yang.zhong@intel.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Pankaj Gupta <pankaj.gupta@ionos.com>
Reviewed-by: Peter Xu <peterx@redhat.com>
Message-Id: <20210709052800.63588-1-yang.zhong@intel.com>
Signed-off-by: Cleber Rosa <crosa@redhat.com>

tests/acceptance/virtio-gpu.py: provide kernel and initrd hashes

By providing kernel and initrd hashes, the test guarantees the
integrity of the images used and avoids the warnings set by
fetch_asset() when hashes are lacking.

Signed-off-by: Cleber Rosa <crosa@redhat.com>
Message-Id: <20210714174051.28164-7-crosa@redhat.com>
Reviewed-by: Willian Rampazzo <willianr@redhat.com>
Signed-off-by: Cleber Rosa <crosa@redhat.com>

tests/acceptance/virtio-gpu.py: use virtio-vga-gl

Since 49afbca3b, the use of an optional virgl renderer is not
available anymore, and since b36eb8860f, the way to choose a GL based
rendered is to use the "virtio-vga-gl" device.

Signed-off-by: Cleber Rosa <crosa@redhat.com>
Message-Id: <20210714174051.28164-6-crosa@redhat.com>
Reviewed-by: Willian Rampazzo <willianr@redhat.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Signed-off-by: Cleber Rosa <crosa@redhat.com>

tests/acceptance/virtio-gpu.py: combine kernel command line

Both tests use the same kernel command line arguments, so there's no
need to have a common and then an additional set of arguments.

Signed-off-by: Cleber Rosa <crosa@redhat.com>
Message-Id: <20210714174051.28164-5-crosa@redhat.com>
Reviewed-by: Willian Rampazzo <willianr@redhat.com>
Signed-off-by: Cleber Rosa <crosa@redhat.com>

tests/acceptance/virtio-gpu.py: combine CPU tags

Like previously done with the arch tags, all tests use the same CPU
value so it's possible to combine them at the class level.

Signed-off-by: Cleber Rosa <crosa@redhat.com>
Message-Id: <20210714174051.28164-4-crosa@redhat.com>
Reviewed-by: Willian Rampazzo <willianr@redhat.com>
Signed-off-by: Cleber Rosa <crosa@redhat.com>

tests/acceptance/virtio-gpu.py: combine x86_64 arch tags

The test class in question is x86_64 specific, so it's possible to set
the tags at the class level.

Signed-off-by: Cleber Rosa <crosa@redhat.com>
Message-Id: <20210714174051.28164-3-crosa@redhat.com>
Reviewed-by: Willian Rampazzo <willianr@redhat.com>
Signed-off-by: Cleber Rosa <crosa@redhat.com>

tests/acceptance/virtio-gpu.py: use require_accelerator()

Since efe30d501 there's a shorthand for requiring specific
accelerators, and canceling the test if it's not available.

Signed-off-by: Cleber Rosa <crosa@redhat.com>
Message-Id: <20210714174051.28164-2-crosa@redhat.com>
Reviewed-by: Willian Rampazzo <willianr@redhat.com>
Signed-off-by: Cleber Rosa <crosa@redhat.com>

Merge remote-tracking branch 'remotes/philmd/tags/kconfig-20210720' into staging

Kconfig patches queue

Various Kconfig patches collected during v6.1 development cycle.

# gpg: Signature made Tue 20 Jul 2021 19:11:32 BST
# gpg:                using RSA key FAABE75E12917221DCFD6BB2E3E32C2CDEADC0DE
# gpg: Good signature from "Philippe Mathieu-Daudé (F4BUG) <f4bug@amsat.org>" [full]
# Primary key fingerprint: FAAB E75E 1291 7221 DCFD  6BB2 E3E3 2C2C DEAD C0DE

* remotes/philmd/tags/kconfig-20210720:
  hw/tricore: fix inclusion of tricore_testboard
  hw/ppc/Kconfig: Add dependency PEGASOS2 -> ATI_VGA
  hw/isa/vt82c686: Add missing Kconfig dependency (runtime error)
  hw/isa/vt82c686: Add missing Kconfig dependencies (build error)
  hw/pci-host/Kconfig: Add missing dependency MV64361 -> I8259
  hw/ppc/Kconfig: Add missing dependency E500 -> DS1338 RTC
  hw/riscv/Kconfig: Restrict NUMA to Virt & Spike machines
  hw/riscv/Kconfig: Add missing dependency MICROCHIP_PFSOC -> SERIAL
  hw/arm/Kconfig: Add missing SDHCI symbol to FSL_IMX25
  hw/arm/Kconfig: Remove unused DS1338 symbol from i.MX25 PDK Board
  hw/arm/Kconfig: Add missing dependency NPCM7XX -> SMBUS
  hw/ide/Kconfig: Add missing dependency PCI -> IDE_QDEV
  hw/i386/Kconfig: Add missing Kconfig dependency (runtime error)
  hw/acpi/Kconfig: Add missing Kconfig dependencies (build error)
  hw/acpi: Do not restrict ACPI core routines to x86 architecture
  hw/mips: Express dependencies of the Boston machine with Kconfig
  hw/mips: Add dependency MIPS_CPS -> MIPS_ITU

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

hw/tricore: fix inclusion of tricore_testboard

Because commit f4063f9c319e39 ("meson: Introduce target-specific
Kconfig") ended being merged after commit 582079c9d27fc8cfff9f49
("hw/tricore: Add testdevice for tests in tests/tcg/"), we
inadvertently added a symbol clash causing the build not to include
the testboard needed for check-tcg.

Fixes: f4063f9c31 ("meson: Introduce target-specific Kconfig")
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-Id: <20210720114057.32053-2-alex.bennee@linaro.org>
[PMD: Updated description mentioning commits merged]
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>

hw/ppc/Kconfig: Add dependency PEGASOS2 -> ATI_VGA

The ATI VGA device isn't a requisite for the Pegasos2 machine
because Linux only uses the serial console; see commit ba7e5ac18e7
("hw/ppc: Add emulation of Genesi/bPlan Pegasos II") for rationale.

Using the default devices we don't have any problem:

  $ qemu-system-ppc -M pegasos2
  qemu-system-ppc: standard VGA not available

But when trying to explicitly use the ATI device we get an error:

  $ qemu-system-ppc -M pegasos2 -vga none -bios pegasos2.rom -device ati-vga,romfile=
  qemu-system-ppc: -device ati-vga,romfile=: 'ati-vga' is not a valid device model name

Add it as an implicit Kconfig dependency.

Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-Id: <20210515173716.358295-13-philmd@redhat.com>
Acked-by: BALATON Zoltan <balaton@eik.bme.hu>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Acked-by: David Gibson <david@gibson.dropbear.id.au>

hw/isa/vt82c686: Add missing Kconfig dependency (runtime error)

When building the Pegasos2 machine stand-alone we get:

  $ qemu-system-ppc -M pegasos2 -bios pegasos2.rom
  ERROR:qom/object.c:714:object_new_with_type: assertion failed: (type != NULL)
  Bail out! ERROR:qom/object.c:714:object_new_with_type: assertion failed: (type != NULL)

Looking at the backtraces:

  Thread 1 "qemu-system-ppc" received signal SIGABRT, Aborted.
  (gdb) bt
  #0  0x00007ffff53877d5 in raise () at /lib64/libc.so.6
  #1  0x00007ffff5370895 in abort () at /lib64/libc.so.6
  #2  0x00007ffff6dc4b6c in g_assertion_message_expr.cold () at /lib64/libglib-2.0.so.0
  #3  0x00007ffff6e229ff in g_assertion_message_expr () at /lib64/libglib-2.0.so.0
  #4  0x0000555555a0c8f4 in object_new_with_type (type=0x0) at qom/object.c:714
  #5  0x0000555555a0c9d5 in object_new (typename=0x555555c7afe4 "isa-pit") at qom/object.c:747
  #6  0x0000555555a053b8 in qdev_new (name=0x555555c7afe4 "isa-pit") at hw/core/qdev.c:153
  #7  0x00005555557cdd05 in isa_new (name=0x555555c7afe4 "isa-pit") at hw/isa/isa-bus.c:160
  #8  0x00005555557cf518 in i8254_pit_init (bus=0x55555603d140, base=64, isa_irq=0, alt_irq=0x0) at include/hw/timer/i8254.h:54
  #9  0x00005555557d12f9 in vt8231_realize (d=0x5555563d9770, errp=0x7fffffffcc28) at hw/isa/vt82c686.c:704

  (gdb) bt
  #0  0x00007ffff54bd7d5 in raise () at /lib64/libc.so.6
  #1  0x00007ffff54a6895 in abort () at /lib64/libc.so.6
  #2  0x00005555558f7796 in object_new (typename=0x555555ad4889 "isa-parallel") at qom/object.c:749
  #3  object_new (typename=type0x555555ad4889 "isa-parallel") at qom/object.c:743
  #4  0x00005555558f0d46 in qdev_new (name=0x555555ad4889 "isa-parallel") at hw/core/qdev.c:153
  #5  0x000055555576b669 in isa_new (name=0x555555ad4889 "isa-parallel") at hw/isa/isa-bus.c:160
  #6  0x000055555576bbe8 in isa_superio_realize (dev=0x555555f15910, errp=<optimized out>) at hw/isa/isa-superio.c:54
  #7  0x000055555576d5ed in via_superio_realize (d=0x555555f15910, errp=0x7fffffffcb30) at hw/isa/vt82c686.c:292
  #8  0x00005555558f12c1 in device_set_realized (obj=<optimized out>, ...) at hw/core/qdev.c:761
  #9  0x00005555558f5066 in property_set_bool (obj=0x555555f15910, ..., errp=0x7fffffffcbb0) at qom/object.c:2262
  #10 0x00005555558f7f38 in object_property_set (obj=0x555555f15910, name=0x555555b1b1e3 "realized", ...) at qom/object.c:1407
  #11 0x00005555558fb2d0 in object_property_set_qobject (obj=0x555555f15910, name=0x555555b1b1e3 "realized", ...) at qom/qom-qobject.c:28
  #12 0x00005555558f8525 in object_property_set_bool (obj=0x555555f15910, name=0x555555b1b1e3 "realized", ...) at qom/object.c:1477
  #13 0x00005555558f18ee in qdev_realize (dev=0x555555f15910, bus=0x55555602a610, errp=<optimized out>) at hw/core/qdev.c:389
  #14 0x00005555558f197f in qdev_realize_and_unref (dev=0x555555f15910, bus=0x55555602a610, errp=<optimized out>) at hw/core/qdev.c:396
  #15 0x000055555576b709 in isa_realize_and_unref (errp=<optimized out>, bus=0x55555602a610, dev=0x555555f15910) at hw/isa/isa-bus.c:179
  #16 isa_create_simple (bus=0x55555602a610, name=0x555555adc33b "vt8231-superio") at hw/isa/isa-bus.c:173
  #17 0x000055555576d9b7 in vt8231_realize (d=0x555556186a50, errp=<optimized out>) at hw/isa/vt82c686.c:706

The "isa-pit" type (TYPE_I8254) and "isa-parallel" are missing. Add them.

Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: BALATON Zoltan <balaton@eik.bme.hu>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Message-Id: <20210515173716.358295-12-philmd@redhat.com>
[PMD: Added "isa-parallel" later]

hw/isa/vt82c686: Add missing Kconfig dependencies (build error)

The VT82C686 device model misses various dependencies:

  /usr/bin/ld: libcommon.fa.p/hw_isa_vt82c686.c.o: in function `vt82c686b_realize':
  hw/isa/vt82c686.c:622: undefined reference to `i8259_init'
  /usr/bin/ld: hw/isa/vt82c686.c:624: undefined reference to `i8257_dma_init'
  /usr/bin/ld: hw/isa/vt82c686.c:627: undefined reference to `mc146818_rtc_init'

Add them.

Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: BALATON Zoltan <balaton@eik.bme.hu>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Message-Id: <20210515173716.358295-11-philmd@redhat.com>

hw/pci-host/Kconfig: Add missing dependency MV64361 -> I8259

Looking at the MV64341 model source, there is a dependency on the
8259 interrupt controller:

  523     case MV64340_PCI_1_INTERRUPT_ACKNOWLEDGE_VIRTUAL_REG:
  524         /* FIXME: Should this be sent via the PCI bus somehow? */
  525         if (s->gpp_int_level && (s->gpp_value & BIT(31))) {
  526             ret = pic_read_irq(isa_pic);
  527         }
  528         break;

Add it to Kconfig to avoid the following build failure:

  /usr/bin/ld: libcommon.fa.p/hw_pci-host_mv64361.c.o: in function `mv64361_read':
  hw/pci-host/mv64361.c:526: undefined reference to `isa_pic'
  /usr/bin/ld: hw/pci-host/mv64361.c:526: undefined reference to `pic_read_irq'

Fixes: dcdf98a9015 ("hw/pci-host: Add emulation of Marvell MV64361 PPC system controller")
Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: BALATON Zoltan <balaton@eik.bme.hu>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Acked-by: David Gibson <david@gibson.dropbear.id.au>
Message-Id: <20210515173716.358295-10-philmd@redhat.com>

Merge remote-tracking branch 'remotes/kevin/tags/for-upstream' into staging

Block layer patches

- mirror: Fix active mirror deadlock
- replication: Fix crashes due to operations on wrong BdrvChild
- configure: Add option to use driver whitelist even in tools
- vvfat: Fix crash when opening image read-write
- export: Fix crash in error path with fixed-iothread=false

# gpg: Signature made Tue 20 Jul 2021 16:09:23 BST
# gpg:                using RSA key DC3DEB159A9AF95D3D7456FE7F09B272C88F2FD6
# gpg:                issuer "kwolf@redhat.com"
# gpg: Good signature from "Kevin Wolf <kwolf@redhat.com>" [full]
# Primary key fingerprint: DC3D EB15 9A9A F95D 3D74  56FE 7F09 B272 C88F 2FD6

* remotes/kevin/tags/for-upstream:
  iotests/307: Test iothread conflict for exports
  block/export: Conditionally ignore set-context error
  block/vvfat: fix: drop backing
  replication: Remove workaround
  replication: Properly attach children
  replication: Reduce usage of s->hidden_disk and s->secondary_disk
  replication: Remove s->active_disk
  block: Add option to use driver whitelist even in tools
  block/mirror: fix active mirror dead-lock in mirror_wait_on_conflicts
  iotest 151: add test-case that shows active mirror dead-lock
  block/mirror: set .co for active-write MirrorOp objects

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

vl: Parse legacy default_machine_opts

qemu can't start a xen vm after commit d8fb7d0969d5
"vl: switch -M parsing to keyval" with:

$ ./qemu-system-i386 -M xenfv
Unexpected error in object_property_find_err() at ../qom/object.c:1298:
qemu-system-i386: Property 'xenfv-3.1-machine.accel' not found
Aborted (core dumped)

The default_machine_opts handling doesn't process the legacy machine
options like "accel".  Call qemu_apply_legacy_machine_options to provide
the legacy handling.

Signed-off-by: Jason Andryuk <jandryuk@gmail.com>
Message-Id: <20210713021552.19110-1-jandryuk@gmail.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

qemu-config: fix memory leak on ferror()

The leak is basically impossible to reach, since the only common way
to get ferror(fp) is by passing a directory to -readconfig.  In that
case, the error occurs before qdict is set to anything non-NULL.
However, it's theoretically possible to get there after an EIO.

Cc: armbru@redhat.com
Reported-by: Peter Maydell <peter.maydell@linaro.org>
Fixes: f7544edcd3 ("qemu-config: add error propagation to qemu_config_parse", 2021-03-06)
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

qemu-config: never call the callback after an error, fix leak

Ensure that the callback to qemu_config_foreach is never called upon
an error, by moving the invocation before the "out" label.

Cc: armbru@redhat.com
Fixes: 3770141139 ("qemu-config: parse configuration files to a QDict", 2021-06-04)
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

iotests/307: Test iothread conflict for exports

Passing fixed-iothread=true should make iothread conflicts fatal,
whereas fixed-iothread=false should not.

Combine the second case with an error condition that is checked after
the iothread is handled, to verify that qemu does not crash if there is
such an error after changing the iothread failed.

Signed-off-by: Max Reitz <mreitz@redhat.com>
Message-Id: <20210624083825.29224-3-mreitz@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Tested-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

block/export: Conditionally ignore set-context error

When invoking block-export-add with some iothread and
fixed-iothread=false, and changing the node's iothread fails, the error
is supposed to be ignored.

However, it is still stored in *errp, which is wrong.  If a second error
occurs, the "*errp must be NULL" assertion in error_setv() fails:

  qemu-system-x86_64: ../util/error.c:59: error_setv: Assertion
  `*errp == NULL' failed.

So if fixed-iothread=false, we should ignore the error by passing NULL
to bdrv_try_set_aio_context().

Fixes: f51d23c80af73c95e0ce703ad06a300f1b3d63ef
       ("block/export: add iothread and fixed-iothread options")
Signed-off-by: Max Reitz <mreitz@redhat.com>
Message-Id: <20210624083825.29224-2-mreitz@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

block/vvfat: fix: drop backing

Most probably this fake backing child doesn't work anyway (see notes
about it in a8a4d15c1c34d).

Still, since 25f78d9e2de528473d52 drivers are required to set
.supports_backing if they want to call bdrv_set_backing_hd, so now
vvfat just doesn't work because of this check.

Let's finally drop this fake backing file.

Fixes: 25f78d9e2de528473d52acfcf7acdfb64e3453d4
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20210715124853.13335-1-vsementsov@virtuozzo.com>
Tested-by: John Arbuckle <programmingkidx@gmail.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

replication: Remove workaround

Remove the workaround introduced in commit
6ecbc6c52672db5c13805735ca02784879ce8285
"replication: Avoid blk_make_empty() on read-only child".

It is not needed anymore since s->hidden_disk is guaranteed to be
writable when secondary_do_checkpoint() runs. Because replication_start(),
_do_checkpoint() and _stop() are only called by COLO migration code
and COLO-migration activates all disks via bdrv_invalidate_cache_all()
before it calls these functions.

Signed-off-by: Lukas Straub <lukasstraub2@web.de>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <d3acfad43879e9f376bffa7dd797ae74d0a7c81a.1626619393.git.lukasstraub2@web.de>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

replication: Properly attach children

The replication driver needs access to the children block-nodes of
it's child so it can issue bdrv_make_empty() and bdrv_co_pwritev()
to manage the replication. However, it does this by directly copying
the BdrvChilds, which is wrong.

Fix this by properly attaching the block-nodes with
bdrv_attach_child() and requesting the required permissions.

This ultimatively fixes a potential crash in replication_co_writev(),
because it may write to s->secondary_disk if it is in state
BLOCK_REPLICATION_FAILOVER_FAILED, without requesting write
permissions first. And now the workaround in
secondary_do_checkpoint() can be removed.

Signed-off-by: Lukas Straub <lukasstraub2@web.de>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <5d0539d729afb8072d0d7cde977c5066285591b4.1626619393.git.lukasstraub2@web.de>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

replication: Reduce usage of s->hidden_disk and s->secondary_disk

In preparation for the next patch, initialize s->hidden_disk and
s->secondary_disk later and replace access to them with local variables
in the places where they aren't initialized yet.

Signed-off-by: Lukas Straub <lukasstraub2@web.de>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <1eb9dc179267207d9c7eccaeb30761758e32e9ab.1626619393.git.lukasstraub2@web.de>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

replication: Remove s->active_disk

s->active_disk is bs->file. Remove it and use local variables instead.

Signed-off-by: Lukas Straub <lukasstraub2@web.de>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <2534f867ea9be5b666dfce19744b7d4e2b96c976.1626619393.git.lukasstraub2@web.de>
Reviewed-by: Zhang Chen <chen.zhang@intel.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

hw/ppc/Kconfig: Add missing dependency E500 -> DS1338 RTC

Commit 7abb479c7ab ("PPC: E500: Add FSL I2C controller and integrate
RTC with it") added a global dependency on the DS1338 model, instead
of a machine one (via Kconfig). This gives trouble when building
standalone machines not exposing I2C bus:

  The following clauses were found for DS1338

      CONFIG_DS1338=y
      config DS1338 depends on I2C

Fix by selecting the DS1338 symbol in the single machine requiring
it, the E500.

Fixes: 7abb479c7ab ("PPC: E500: Add FSL I2C controller and integrate RTC with it")
Acked-by: Paolo Bonzini <pbonzini@redhat.com>
Acked-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-Id: <20210513163858.3928976-9-philmd@redhat.com>

hw/riscv/Kconfig: Restrict NUMA to Virt & Spike machines

Only the Virt and Spike machines use NUMA. Add a RISCV_NUMA Kconfig
symbol and only have these machines select it. Adapt the Meson file
to only built it if required.

Acked-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-Id: <20210515173716.358295-8-philmd@redhat.com>
Reviewed-by: Alistair Francis <alistair.francis@wdc.com>

hw/riscv/Kconfig: Add missing dependency MICROCHIP_PFSOC -> SERIAL

Commit a8fb0a500a6 ("hw/char: Add Microchip PolarFire SoC MMUART
emulation") added a dependency on the SERIAL model, but forgot to
add the Kconfig selector.
Add the dependency to the MCHP_PFSOC_MMUART symbol to fix when
building the MICROCHIP_PFSOC machine stand-alone:

  /usr/bin/ld: libcommon.fa.p/hw_char_mchp_pfsoc_mmuart.c.o: in function `mchp_pfsoc_mmuart_create':
  hw/char/mchp_pfsoc_mmuart.c:79: undefined reference to `serial_mm_init'

Fixes: a8fb0a500a6 ("hw/char: Add Microchip PolarFire SoC MMUART emulation")
Acked-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-Id: <20210515173716.358295-7-philmd@redhat.com>
Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>

hw/arm/Kconfig: Add missing SDHCI symbol to FSL_IMX25

Commit bfae1772c43 ("hw/arm/fsl-imx25: Wire up eSDHC controllers")
added a dependency on the TYPE_IMX_USDHC model, but forgot to add
the Kconfig selector. Fix that to solve when built stand-alone:

  $ qemu-system-arm -M imx25-pdk
  qemu-system-arm: missing object type 'imx-usdhc'
  Aborted (core dumped)

Fixes: bfae1772c43 ("hw/arm/fsl-imx25: Wire up eSDHC controllers")
Acked-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-Id: <20210515173716.358295-6-philmd@redhat.com>