Merge remote-tracking branch 'remotes/rth/tags/pull-tcg-20200212' into staging

Fix breakpoint invalidation.
Add support for tcg helpers with 7 arguments.
Add support for gvec helpers with 5 arguments.

# gpg: Signature made Thu 13 Feb 2020 00:21:34 GMT
# 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/tags/pull-tcg-20200212:
  tcg: Add tcg_gen_gvec_5_ptr
  tcg: Add support for a helper with 7 arguments
  exec: flush CPU TB cache in breakpoint_invalidate

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

Merge remote-tracking branch 'remotes/kraxel/tags/usb-20200212-pull-request' into staging

usb: bugfixes

# gpg: Signature made Wed 12 Feb 2020 20:57:44 GMT
# gpg:                using RSA key 4CB6D8EED3E87138
# 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/usb-20200212-pull-request:
  uas: fix super speed bMaxPacketSize0
  usb-host: wait for cancel complete

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

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

* various small fixes and cleanups
* fixes for the ucode revision patch from the previous pull request

# gpg: Signature made Wed 12 Feb 2020 15:30:24 GMT
# gpg:                using RSA key BFFBD25F78C7AE83
# 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/tags/for-upstream:
  target/i386: enable monitor and ucode revision with -cpu max
  target/i386: check for availability of MSR_IA32_UCODE_REV as an emulated MSR
  target/i386: fix TCG UCODE_REV access
  build: move TARGET_GPROF to config-host.mak
  exec: do not define use_icount for user-mode emulation
  minikconf: accept alnum identifiers
  Remove support for CLOCK_MONOTONIC not being defined
  seqlock: fix seqlock_write_unlock_impl function
  vl: Don't mismatch g_strsplit()/g_free()

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

Merge remote-tracking branch 'remotes/kraxel/tags/ui-20200212-pull-request' into staging

gtk: refresh rate fix.
cocoa: drop pre-10.12 support.
ui: rework show-cursor option.

# gpg: Signature made Wed 12 Feb 2020 16:18:17 GMT
# gpg:                using RSA key 4CB6D8EED3E87138
# 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/ui-20200212-pull-request:
  ui/cocoa: Drop workarounds for pre-10.12 OSX
  ui: deprecate legacy -show-cursor option
  ui: drop curor_hide global variable.
  ui/gtk: implement show-cursor option
  ui/cocoa: switch to new show-cursor option
  ui/sdl: switch to new show-cursor option
  ui: wire up legacy -show-cursor option
  ui: add show-cursor option
  ui/gtk: Fix gd_refresh_rate_millihz() when widget window is not realized
  ui/gtk: Update gd_refresh_rate_millihz() to handle VirtualConsole

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

tcg: Add tcg_gen_gvec_5_ptr

Extend the vector generator infrastructure to handle
5 vector arguments.

Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Taylor Simpson <tsimpson@quicinc.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

uas: fix super speed bMaxPacketSize0

For usb2 bMaxPacketSize0 is "n", for usb3 it is "1 << n",
so it must be 9 not 64 ...

rom "Universal Serial Bus 3.1 Specification":

   If the device is operating at Gen X speed, the bMaxPacketSize0
   field shall be set to 09H indicating a 512-byte maximum packet.
   An Enhanced SuperSpeed device shall not support any other maximum
   packet sizes for the default control pipe (endpoint 0) control
   endpoint.

We now announce a 512-byte maximum packet.

Fixes: 89a453d4a5c ("uas-uas: usb3 streams")
Reported-by: Benjamin David Lunt <fys@fysnet.net>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-id: 20200117073716.31335-1-kraxel@redhat.com

usb-host: wait for cancel complete

After canceling transfers call into libvirt so it can process the
request, and wait for it to complete.  Also cancel all pending
transfers before exiting qemu.

Buglink: https://bugzilla.redhat.com//show_bug.cgi?id=1749745
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Message-id: 20200203114108.23952-1-kraxel@redhat.com

target/i386: enable monitor and ucode revision with -cpu max

These two features were incorrectly tied to host_cpuid_required rather than
cpu->max_features.  As a result, -cpu max was not enabling either MONITOR
features or ucode revision.

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

target/i386: check for availability of MSR_IA32_UCODE_REV as an emulated MSR

Even though MSR_IA32_UCODE_REV has been available long before Linux 5.6,
which added it to the emulated MSR list, a bug caused the microcode
version to revert to 0x100000000 on INIT.  As a result, processors other
than the bootstrap processor would not see the host microcode revision;
some Windows version complain loudly about this and crash with a
fairly explicit MICROCODE REVISION MISMATCH error.

[If running 5.6 prereleases, the kernel fix "KVM: x86: do not reset
 microcode version on INIT or RESET" should also be applied.]

Reported-by: Alex Williamson <alex.williamson@redhat.com>
Message-id: <20200211175516.10716-1-pbonzini@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

target/i386: fix TCG UCODE_REV access

This was a very interesting semantic conflict that caused git to move
the MSR_IA32_UCODE_REV read to helper_wrmsr.  Not a big deal, but
still should be fixed...

Fixes: 4e45aff398 ("target/i386: add a ucode-rev property", 2020-01-24)
Message-id: <20200206171022.9289-1-pbonzini@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

build: move TARGET_GPROF to config-host.mak

TARGET_GPROF is the same for all targets, write it to
config-host.mak instead.

Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-id: <20200204161104.21077-1-pbonzini@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

exec: do not define use_icount for user-mode emulation

use_icount is also defined by stubs/cpu-get-icount.c, we do not need
to have a useless definition in exec.c.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-id: <20200204161036.20889-1-pbonzini@redhat.com>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>

minikconf: accept alnum identifiers

Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

Remove support for CLOCK_MONOTONIC not being defined

Some older parts of QEMU's codebase assume that CLOCK_MONOTONIC
might not be defined by the host OS, and have workarounds to
deal with this. However, more recently (notably in commit
50290c002c045280f8d for qemu-img in mid-2019, but also much
earlier in 2011 in commit 22795174a37e0 for ui/spice-display.c)
we've written code that assumes CLOCK_MONOTONIC is always
defined. The only host OS anybody's ever noticed this on
is OSX 10.11 and earlier, which we don't support.

So we can assume that all our host OSes have the #define,
and we can remove some now-unnecessary ifdefs.

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Message-Id: <20200201172252.6605-1-peter.maydell@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

seqlock: fix seqlock_write_unlock_impl function

The seqlock write unlock function was incorrectly calling
seqlock_write_begin() instead of seqlock_write_end(), and was releasing
the lock before incrementing the sequence. This could lead to a race
condition and a corrupted sequence number becoming odd even though the
lock is not held.

Signed-off-by: Luc Michel <luc.michel@greensocs.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-Id: <20200129144948.2161551-1-luc.michel@greensocs.com>
Fixes: 988fcafc73 ("seqlock: add QemuLockable support", 2018-08-23)
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

vl: Don't mismatch g_strsplit()/g_free()

It's a mismatch between g_strsplit and g_free, it will cause a memory leak as follow:

[root@localhost]# ./aarch64-softmmu/qemu-system-aarch64 -accel help
Accelerators supported in QEMU binary:
tcg
kvm
=================================================================
==1207900==ERROR: LeakSanitizer: detected memory leaks

Direct leak of 8 byte(s) in 2 object(s) allocated from:
    #0 0xfffd700231cb in __interceptor_malloc (/lib64/libasan.so.4+0xd31cb)
    #1 0xfffd6ec57163 in g_malloc (/lib64/libglib-2.0.so.0+0x57163)
    #2 0xfffd6ec724d7 in g_strndup (/lib64/libglib-2.0.so.0+0x724d7)
    #3 0xfffd6ec73d3f in g_strsplit (/lib64/libglib-2.0.so.0+0x73d3f)
    #4 0xaaab66be5077 in main /mnt/sdc/qemu-master/qemu-4.2.0-rc0/vl.c:3517
    #5 0xfffd6e140b9f in __libc_start_main (/lib64/libc.so.6+0x20b9f)
    #6 0xaaab66bf0f53  (./build/aarch64-softmmu/qemu-system-aarch64+0x8a0f53)

Direct leak of 2 byte(s) in 2 object(s) allocated from:
    #0 0xfffd700231cb in __interceptor_malloc (/lib64/libasan.so.4+0xd31cb)
    #1 0xfffd6ec57163 in g_malloc (/lib64/libglib-2.0.so.0+0x57163)
    #2 0xfffd6ec7243b in g_strdup (/lib64/libglib-2.0.so.0+0x7243b)
    #3 0xfffd6ec73e6f in g_strsplit (/lib64/libglib-2.0.so.0+0x73e6f)
    #4 0xaaab66be5077 in main /mnt/sdc/qemu-master/qemu-4.2.0-rc0/vl.c:3517
    #5 0xfffd6e140b9f in __libc_start_main (/lib64/libc.so.6+0x20b9f)
    #6 0xaaab66bf0f53  (./build/aarch64-softmmu/qemu-system-aarch64+0x8a0f53)

Reported-by: Euler Robot <euler.robot@huawei.com>
Signed-off-by: Pan Nengyuan <pannengyuan@huawei.com>
Message-Id: <20200110091710.53424-2-pannengyuan@huawei.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

ui/cocoa: Drop workarounds for pre-10.12 OSX

Our official OSX support policy covers the last two released versions.
Currently that is 10.14 and 10.15.  We also may work on older versions, but
don't guarantee it.

In commit 50290c002c045280f8d in mid-2019 we introduced some uses of
CLOCK_MONOTONIC which incidentally broke compilation for pre-10.12 OSX
versions (see LP:1861551). We don't intend to fix that, so we might
as well drop the code in ui/cocoa.m which caters for pre-10.12
versions as well. (For reference, 10.11 fell out of Apple extended
security support in September 2018.)

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Message-Id: <20200201170534.22123-1-peter.maydell@linaro.org>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>

ui: deprecate legacy -show-cursor option

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>

ui: drop curor_hide global variable.

No users left.

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>

ui/gtk: implement show-cursor option

When specified just set null_cursor to NULL so we get the default
pointer instead of a blank pointer.

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>

ui/cocoa: switch to new show-cursor option

Use DisplayOpts settings to set the new file-global cursor_hide
variable, stop using the qemu-global cursor_hide variable.

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>

ui/sdl: switch to new show-cursor option

Use DisplayOpts settings instead of cursor_hide global variable.
Also make "-display sdl,show-cursor=on" work.

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>

ui: wire up legacy -show-cursor option

Set new show-cursor display option when legacy -show-cursor
is specified on the command line.

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>

ui: add show-cursor option

When enabled, this forces showing the mouse cursor,
i.e. do not hide the pointer on mouse grabs.
Defaults to off.

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>

ui/gtk: Fix gd_refresh_rate_millihz() when widget window is not realized

gtk_widget_get_window() returns NULL if the widget's window is not
realized, and QEMU crashes. Example under gtk 3.22.30 (mate 1.20.1):

  qemu-system-x86_64: Gdk: gdk_window_get_origin: assertion 'GDK_IS_WINDOW (window)' failed
  (gdb) bt
  #0  0x00007ffff496cf70 in gdk_window_get_origin () from /usr/lib64/libgdk-3.so.0
  #1  0x00007ffff49582a0 in gdk_display_get_monitor_at_window () from /usr/lib64/libgdk-3.so.0
  #2  0x0000555555bb73e2 in gd_refresh_rate_millihz (window=0x5555579d6280) at ui/gtk.c:1973
  #3  gd_vc_gfx_init (view_menu=0x5555579f0590, group=0x0, idx=0, con=<optimized out>, vc=0x5555579d4a90, s=0x5555579d49f0) at ui/gtk.c:2048
  #4  gd_create_menu_view (s=0x5555579d49f0) at ui/gtk.c:2149
  #5  gd_create_menus (s=0x5555579d49f0) at ui/gtk.c:2188
  #6  gtk_display_init (ds=<optimized out>, opts=0x55555661ed80 <dpy>) at ui/gtk.c:2256
  #7  0x000055555583d5a0 in main (argc=<optimized out>, argv=<optimized out>, envp=<optimized out>) at vl.c:4358

Fixes: c4c00922cc and 28b58f19d2 (display/gtk: get proper refreshrate)
Reported-by: Jan Kiszka <jan.kiszka@web.de>
Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Tested-by: Jan Kiszka <jan.kiszka@web.de>
Message-id: 20200208161048.11311-3-f4bug@amsat.org
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>

ui/gtk: Update gd_refresh_rate_millihz() to handle VirtualConsole

Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Tested-by: Jan Kiszka <jan.kiszka@web.de>
Message-id: 20200208161048.11311-2-f4bug@amsat.org
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>

tcg: Add support for a helper with 7 arguments

Currently, helpers can only take up to 6 arguments.  This patch adds the
capability for up to 7 arguments.  I have tested it with the Hexagon port
that I am preparing for submission.

Signed-off-by: Taylor Simpson <tsimpson@quicinc.com>
Message-Id: <1580942510-2820-1-git-send-email-tsimpson@quicinc.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

exec: flush CPU TB cache in breakpoint_invalidate

When a breakpoint is inserted at location for which there's currently no
virtual to physical translation no action is taken on CPU TB cache. If a
TB for that virtual address already exists but is not visible ATM the
breakpoint won't be hit next time an instruction at that address will be
executed.

Flush entire CPU TB cache in breakpoint_invalidate to force
re-translation of all TBs for the breakpoint address.

This change fixes the following scenario:
- linux user application is running
- a breakpoint is inserted from QEMU gdbstub for a user address that is
  not currently present in the target CPU TLB
- an instruction at that address is executed, but the external debugger
  doesn't get control.

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
Message-Id: <20191127220602.10827-2-jcmvbkbc@gmail.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

Merge remote-tracking branch 'remotes/dgilbert-gitlab/tags/pull-virtiofs-20200210' into staging

virtiofsd pull 2020-02-10

Coverity fixes and a reworked man page.

# gpg: Signature made Mon 10 Feb 2020 17:26:29 GMT
# gpg:                using RSA key 45F5C71B4A0CB7FB977A9FA90516331EBC5BFDE7
# gpg: Good signature from "Dr. David Alan Gilbert (RH2) <dgilbert@redhat.com>" [full]
# Primary key fingerprint: 45F5 C71B 4A0C B7FB 977A  9FA9 0516 331E BC5B FDE7

* remotes/dgilbert-gitlab/tags/pull-virtiofs-20200210:
  docs: add virtiofsd(1) man page
  virtiofsd: do_read missing NULL check
  virtiofsd: load_capng missing unlock
  virtiofsd: fv_create_listen_socket error path socket leak
  virtiofsd: Remove fuse_req_getgroups

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

docs: add virtiofsd(1) man page

Document the virtiofsd(1) program and its command-line options.  This
man page is a rST conversion of the original texi documentation that I
wrote.

Reviewed-by: Liam Merwick <liam.merwick@oracle.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>

virtiofsd: do_read missing NULL check

Missing a NULL check if the argument fetch fails.

Fixes: Coverity CID 1413119
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>

virtiofsd: load_capng missing unlock

Missing unlock in error path.

Fixes: Covertiy CID 1413123
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>

virtiofsd: fv_create_listen_socket error path socket leak

If we fail when bringing up the socket we can leak the listen_fd;
in practice the daemon will exit so it's not really a problem.

Fixes: Coverity CID 1413121
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>

virtiofsd: Remove fuse_req_getgroups

Remove fuse_req_getgroups that's unused in virtiofsd; it came in
from libfuse but we don't actually use it.  It was called from
fuse_getgroups which we previously removed (but had left it's header
in).

Coverity had complained about null termination in it, but removing
it is the easiest answer.

Fixes: Coverity CID: 1413117 (String not null terminated)
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>

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

Pull request

# gpg: Signature made Mon 10 Feb 2020 09:23:42 GMT
# 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/tags/block-pull-request:
  hw/core: Allow setting 'virtio-blk-device.scsi' property on OSX host
  block: fix crash on zero-length unaligned write and read

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

Merge remote-tracking branch 'remotes/gkurz/tags/9p-next-2020-02-08' into staging

9p patches:
- some more protocol sanity checks
- qtest for readdir
- Christian Schoenebeck now official reviewer

# gpg: Signature made Sat 08 Feb 2020 08:38:10 GMT
# gpg:                using RSA key B4828BAF943140CEF2A3491071D4D5E5822F73D6
# gpg: Good signature from "Greg Kurz <groug@kaod.org>" [full]
# gpg:                 aka "Gregory Kurz <gregory.kurz@free.fr>" [full]
# gpg:                 aka "[jpeg image of size 3330]" [full]
# Primary key fingerprint: B482 8BAF 9431 40CE F2A3  4910 71D4 D5E5 822F 73D6

* remotes/gkurz/tags/9p-next-2020-02-08:
  MAINTAINERS: 9pfs: Add myself as reviewer
  tests/virtio-9p: added readdir test
  hw/9pfs/9p-synth: added directory for readdir test
  9pfs: validate count sent by client with T_readdir
  9pfs: require msize >= 4096
  tests/virtio-9p: add terminating null in v9fs_string_read()

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

Merge remote-tracking branch 'remotes/philmd-gitlab/tags/python-next-20200207' into staging

- Python 3 cleanups:
  . Remove text about Python 2 in qemu-deprecated (Thomas)
  . Remove shebang header (Paolo, Philippe)
  . scripts/checkpatch.pl now allows Python 3 interpreter (Philippe)
  . Explicit usage of Python 3 interpreter in scripts (Philippe)
  . Fix Python scripts permissions (Paolo, Philippe)
  . Drop 'from __future__ import print_function' (Paolo)
  . Specify minimum python requirements in ReadTheDocs configuration (Alex)
- Test UNIX/EXEC transports with migration (Oksana)
- Added extract_from_rpm helper, improved extract_from_deb (Liam)
- Allow to use other serial consoles than default one (Philippe)
- Various improvements in QEMUMonitorProtocol (Wainer)
- Fix kvm_available() on ppc64le (Wainer)

# gpg: Signature made Fri 07 Feb 2020 15:01:56 GMT
# 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-gitlab/tags/python-next-20200207: (46 commits)
  .readthedocs.yml: specify some minimum python requirements
  drop "from __future__ import print_function"
  make all Python scripts executable
  scripts/signrom: remove Python 2 support, add shebang
  tests/qemu-iotests/check: Only check for Python 3 interpreter
  scripts: Explicit usage of Python 3 (scripts without __main__)
  tests/qemu-iotests: Explicit usage of Python3 (scripts without __main__)
  tests/vm: Remove shebang header
  tests/acceptance: Remove shebang header
  scripts/tracetool: Remove shebang header
  scripts/minikconf: Explicit usage of Python 3
  scripts: Explicit usage of Python 3 (scripts with __main__)
  tests: Explicit usage of Python 3
  tests/qemu-iotests: Explicit usage of Python 3 (scripts with __main__)
  tests/qemu-iotests/check: Allow use of python3 interpreter
  scripts/checkpatch.pl: Only allow Python 3 interpreter
  tests/acceptance/migration: Default to -nodefaults
  tests/acceptance/migration: Add the 'migration' tag
  tests/acceptance/migration: Test EXEC transport when migrating
  tests/acceptance/migration: Test UNIX transport when migrating
  ...

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

MAINTAINERS: 9pfs: Add myself as reviewer

Currently 9pfs is only taken care of by Greg. Since I am actively working
on 9pfs and already became quite used to the code base, it makes sense to
volunteer as reviewer for 9pfs related patches.

Signed-off-by: Christian Schoenebeck <qemu_oss@crudebyte.com>
Cc: Greg Kurz <groug@kaod.org>
Message-Id: <E1j04TG-0001xn-JY@lizzy.crudebyte.com>
Signed-off-by: Greg Kurz <groug@kaod.org>

tests/virtio-9p: added readdir test

The first readdir test simply checks the amount of directory
entries returned by 9pfs server, according to the created amount
of virtual files on 9pfs synth driver side. Then the subsequent
readdir test also checks whether all directory entries have the
expected file names (as created on 9pfs synth driver side),
ignoring their precise order in result list though.

Signed-off-by: Christian Schoenebeck <qemu_oss@crudebyte.com>
Message-Id: <e0b4402722a877178f8fb6a8ad7b64bb20150613.1579567020.git.qemu_oss@crudebyte.com>
Reviewed-by: Greg Kurz <groug@kaod.org>
Signed-off-by: Greg Kurz <groug@kaod.org>

hw/9pfs/9p-synth: added directory for readdir test

This will provide the following virtual files by the 9pfs
synth driver:

  - /ReadDirDir/ReadDirFile99
  - /ReadDirDir/ReadDirFile98
  ...
  - /ReadDirDir/ReadDirFile1
  - /ReadDirDir/ReadDirFile0

This virtual directory and its virtual 100 files will be
used by the upcoming 9pfs readdir tests.

Signed-off-by: Christian Schoenebeck <qemu_oss@crudebyte.com>
Reviewed-by: Greg Kurz <groug@kaod.org>
Message-Id: <5408c28c8de25dd575b745cef63bf785305ccef2.1579567020.git.qemu_oss@crudebyte.com>
Signed-off-by: Greg Kurz <groug@kaod.org>

9pfs: validate count sent by client with T_readdir

A good 9p client sends T_readdir with "count" parameter that's sufficiently
smaller than client's initially negotiated msize (maximum message size).
We perform a check for that though to avoid the server to be interrupted
with a "Failed to encode VirtFS reply type 41" transport error message by
bad clients. This count value constraint uses msize - 11, because 11 is the
header size of R_readdir.

Signed-off-by: Christian Schoenebeck <qemu_oss@crudebyte.com>
Reviewed-by: Greg Kurz <groug@kaod.org>
Message-Id: <3990d3891e8ae2074709b56449e96ab4b4b93b7d.1579567020.git.qemu_oss@crudebyte.com>
[groug: added comment ]
Signed-off-by: Greg Kurz <groug@kaod.org>

9pfs: require msize >= 4096

A client establishes a session by sending a Tversion request along with a
'msize' parameter which client uses to suggest server a maximum message
size ever to be used for communication (for both requests and replies)
between client and server during that session. If client suggests a 'msize'
smaller than 4096 then deny session by server immediately with an error
response (Rlerror for "9P2000.L" clients or Rerror for "9P2000.u" clients)
instead of replying with Rversion.

So far any msize submitted by client with Tversion was simply accepted by
server without any check. Introduction of some minimum msize makes sense,
because e.g. a msize < 7 would not allow any subsequent 9p operation at
all, because 7 is the size of the header section common by all 9p message
types.

A substantial higher value of 4096 was chosen though to prevent potential
issues with some message types. E.g. Rreadlink may yield up to a size of
PATH_MAX which is usually 4096, and like almost all 9p message types,
Rreadlink is not allowed to be truncated by the 9p protocol. This chosen
size also prevents a similar issue with Rreaddir responses (provided client
always sends adequate 'count' parameter with Treaddir), because even though
directory entries retrieval may be split up over several T/Rreaddir
messages; a Rreaddir response must not truncate individual directory entries
though. So msize should be large enough to return at least one directory
entry with the longest possible file name supported by host. Most file
systems support a max. file name length of 255. Largest known file name
lenght limit would be currently ReiserFS with max. 4032 bytes, which is
also covered by this min. msize value because 4032 + 35 < 4096.

Furthermore 4096 is already the minimum msize of the Linux kernel's 9pfs
client.

Signed-off-by: Christian Schoenebeck <qemu_oss@crudebyte.com>
Reviewed-by: Greg Kurz <groug@kaod.org>
Message-Id: <8ceecb7fb9fdbeabbe55c04339349a36929fb8e3.1579567019.git.qemu_oss@crudebyte.com>
Signed-off-by: Greg Kurz <groug@kaod.org>

tests/virtio-9p: add terminating null in v9fs_string_read()

The 9p protocol sends strings in general without null termination
over the wire. However for future use of this functions it is
beneficial for the delivered string to be null terminated though
for being able to use the string with standard C functions which
often rely on strings being null terminated.

Signed-off-by: Christian Schoenebeck <qemu_oss@crudebyte.com>
Reviewed-by: Greg Kurz <groug@kaod.org>
Message-Id: <52c84e2ce3bcafc2a38eed13b8c8e23bc1a8ecb9.1579567019.git.qemu_oss@crudebyte.com>
Signed-off-by: Greg Kurz <groug@kaod.org>

Merge remote-tracking branch 'remotes/pmaydell/tags/pull-target-arm-20200207' into staging

target-arm queue:
 * monitor: fix query-cpu-model-expansion crash when using machine type none
 * Support emulation of the ARMv8.1-VHE architecture feature
 * bcm2835_dma: fix bugs in TD mode handling
 * docs/arm-cpu-features: Make kvm-no-adjvtime comment clearer
 * stellaris, stm32f2xx_timer, armv7m_systick: fix minor memory leaks

# gpg: Signature made Fri 07 Feb 2020 14:32:28 GMT
# gpg:                using RSA key E1A5C593CD419DE28E8315CF3C2525ED14360CDE
# gpg:                issuer "peter.maydell@linaro.org"
# gpg: Good signature from "Peter Maydell <peter.maydell@linaro.org>" [ultimate]
# gpg:                 aka "Peter Maydell <pmaydell@gmail.com>" [ultimate]
# gpg:                 aka "Peter Maydell <pmaydell@chiark.greenend.org.uk>" [ultimate]
# Primary key fingerprint: E1A5 C593 CD41 9DE2 8E83  15CF 3C25 25ED 1436 0CDE

* remotes/pmaydell/tags/pull-target-arm-20200207: (48 commits)
  stellaris: delay timer_new to avoid memleaks
  stm32f2xx_timer: delay timer_new to avoid memleaks
  armv7m_systick: delay timer_new to avoid memleaks
  docs/arm-cpu-features: Make kvm-no-adjvtime comment clearer
  bcm2835_dma: Re-initialize xlen in TD mode
  bcm2835_dma: Fix the ylen loop in TD mode
  target/arm: Raise only one interrupt in arm_cpu_exec_interrupt
  target/arm: Use bool for unmasked in arm_excp_unmasked
  target/arm: Pass more cpu state to arm_excp_unmasked
  target/arm: Move arm_excp_unmasked to cpu.c
  target/arm: Enable ARMv8.1-VHE in -cpu max
  target/arm: Update arm_cpu_do_interrupt_aarch64 for VHE
  target/arm: Update get_a64_user_mem_index for VHE
  target/arm: check TGE and E2H flags for EL0 pauth traps
  target/arm: Update {fp,sve}_exception_el for VHE
  target/arm: Update arm_phys_excp_target_el for TGE
  target/arm: Flush tlbs for E2&0 translation regime
  target/arm: Flush tlb for ASID changes in EL2&0 translation regime
  target/arm: Add VHE timer register redirection and aliasing
  target/arm: Add VHE system register redirection and aliasing
  ...

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

hw/core: Allow setting 'virtio-blk-device.scsi' property on OSX host

Commit ed65fd1a2750 ("virtio-blk: switch off scsi-passthrough by
default") changed the default value of the 'scsi' property of
virtio-blk, which is only available on Linux hosts. It also added
an unconditional compat entry for 2.4 or earlier machines.

Trying to set this property on a pre-2.5 machine on OSX, we get:

   Unexpected error in object_property_find() at qom/object.c:1201:
   qemu-system-x86_64: -device virtio-blk-pci,id=scsi0,drive=drive0: can't apply global virtio-blk-device.scsi=true: Property '.scsi' not found

Fix this error by marking the property optional.

Fixes: ed65fd1a27 ("virtio-blk: switch off scsi-passthrough by default")
Suggested-by: Cornelia Huck <cohuck@redhat.com>
Reviewed-by: Cornelia Huck <cohuck@redhat.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Message-id: 20200207001404.1739-1-philmd@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>

block: fix crash on zero-length unaligned write and read

Commit 7a3f542fbd "block/io: refactor padding" occasionally dropped
aligning for zero-length request: bdrv_init_padding() blindly return
false if bytes == 0, like there is nothing to align.

This leads the following command to crash:

./qemu-io --image-opts -c 'write 1 0' \
  driver=blkdebug,align=512,image.driver=null-co,image.size=512

>> qemu-io: block/io.c:1955: bdrv_aligned_pwritev: Assertion
    `(offset & (align - 1)) == 0' failed.
>> Aborted (core dumped)

Prior to 7a3f542fbd we does aligning of such zero requests. Instead of
recovering this behavior let's just do nothing on such requests as it
is useless.

Note that driver may have special meaning of zero-length reqeusts, like
qcow2_co_pwritev_compressed_part, so we can't skip any zero-length
operation. But for unaligned ones, we can't pass it to driver anyway.

This commit also fixes crash in iotest 80 running with -nocache:

./check -nocache -qcow2 80

which crashes on same assertion due to trying to read empty extra data
in qcow2_do_read_snapshots().

Cc: qemu-stable@nongnu.org # v4.2
Fixes: 7a3f542fbd
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-id: 20200206164245.17781-1-vsementsov@virtuozzo.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>

Merge remote-tracking branch 'remotes/berrange/tags/misc-fixes-pull-request' into staging

Misc fixes

* Use correct constants for 'bool' type
* Fix compatibility with noVNC in websocket impl
* Remove inaccurate docs aout default NIC model

# gpg: Signature made Fri 07 Feb 2020 12:23:53 GMT
# gpg:                using RSA key DAF3A6FDB26B62912D0E8E3FBE86EBB415104FDF
# gpg: Good signature from "Daniel P. Berrange <dan@berrange.com>" [full]
# gpg:                 aka "Daniel P. Berrange <berrange@redhat.com>" [full]
# Primary key fingerprint: DAF3 A6FD B26B 6291 2D0E  8E3F BE86 EBB4 1510 4FDF

* remotes/berrange/tags/misc-fixes-pull-request:
  docs: stop documenting the e1000 NIC model as the default
  authz: fix usage of bool in listfile.c
  io/channel-websock: treat 'binary' and no sub-protocol as the same

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

.readthedocs.yml: specify some minimum python requirements

QEMU is all about the Python 3 now so lets also hint that to
ReadTheDocs in its config file.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Message-Id: <20200206163120.31899-1-alex.bennee@linaro.org>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>

drop "from __future__ import print_function"

This is only needed for Python 2, which we do not support anymore.

Cc: Philippe Mathieu-Daudé <philmd@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Acked-by: Alex Bennée <alex.bennee@linaro.org>
Acked-by: Markus Armbruster <armbru@redhat.com>
Message-Id: <20200204160604.19883-1-pbonzini@redhat.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>

make all Python scripts executable

Scripts that have a Python shebang are meant to be executed directly from the
shell; give them 755 permissions.

Cc: Philippe Mathieu-Daudé <philmd@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-Id: <20200204160237.16889-1-pbonzini@redhat.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>

scripts/signrom: remove Python 2 support, add shebang

Cc: Philippe Mathieu-Daudé <philmd@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-Id: <20200204160028.16211-1-pbonzini@redhat.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>

tests/qemu-iotests/check: Only check for Python 3 interpreter

All the iotests Python scripts have been converted to search for
the Python 3 interpreter. Update the ./check script accordingly.

Acked-by: Stefan Hajnoczi <stefanha@redhat.com>
Acked-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20200130163232.10446-13-philmd@redhat.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>

scripts: Explicit usage of Python 3 (scripts without __main__)

Use the program search path to find the Python 3 interpreter.

Patch created mechanically by running:

  $ sed -i "s,^#\!/usr/bin/\(env\ \)\?python$,#\!/usr/bin/env python3," \
      $(git grep -lF '#!/usr/bin/env python' \
      | xargs grep -L 'if __name__.*__main__')

Reported-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Suggested-by: Daniel P. Berrangé <berrange@redhat.com>
Suggested-by: Stefan Hajnoczi <stefanha@redhat.com>
Acked-by: Stefan Hajnoczi <stefanha@redhat.com>
Acked-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20200130163232.10446-12-philmd@redhat.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>

tests/qemu-iotests: Explicit usage of Python3 (scripts without __main__)

Use the program search path to find the Python 3 interpreter.

Patch created mechanically by running:

  $ sed -i "s,^#\!/usr/bin/\(env\ \)\?python$,#\!/usr/bin/env python3," \
      $(git grep -lF '#!/usr/bin/env python' \
      | xargs grep -L 'if __name__.*__main__')

Reported-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Suggested-by: Daniel P. Berrangé <berrange@redhat.com>
Suggested-by: Stefan Hajnoczi <stefanha@redhat.com>
Acked-by: Stefan Hajnoczi <stefanha@redhat.com>
Acked-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20200130163232.10446-11-philmd@redhat.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>

tests/vm: Remove shebang header

Patch created mechanically by running:

  $ chmod 644 $(git grep -lF '#!/usr/bin/env python' \
      | xargs grep -L 'if __name__.*__main__')
  $ sed -i "/^#\!\/usr\/bin\/\(env\ \)\?python.\?$/d" \
      $(git grep -lF '#!/usr/bin/env python' \
      | xargs grep -L 'if __name__.*__main__')

Reported-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Suggested-by: Stefan Hajnoczi <stefanha@redhat.com>
Acked-by: Stefan Hajnoczi <stefanha@redhat.com>
Acked-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20200130163232.10446-10-philmd@redhat.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>

tests/acceptance: Remove shebang header

Patch created mechanically by running:

  $ chmod 644 $(git grep -lF '#!/usr/bin/env python' \
      | xargs grep -L 'if __name__.*__main__')
  $ sed -i "/^#\!\/usr\/bin\/\(env\ \)\?python.\?$/d" \
      $(git grep -lF '#!/usr/bin/env python' \
      | xargs grep -L 'if __name__.*__main__')

Reported-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Suggested-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Wainer dos Santos Moschetta <wainersm@redhat.com>
Acked-by: Stefan Hajnoczi <stefanha@redhat.com>
Acked-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20200130163232.10446-8-philmd@redhat.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>

scripts/tracetool: Remove shebang header

Patch created mechanically by running:

  $ chmod 644 $(git grep -lF '#!/usr/bin/env python' \
      | xargs grep -L 'if __name__.*__main__')
  $ sed -i "/^#\!\/usr\/bin\/\(env\ \)\?python.\?$/d" \
      $(git grep -lF '#!/usr/bin/env python' \
      | xargs grep -L 'if __name__.*__main__')

Reported-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Suggested-by: Stefan Hajnoczi <stefanha@redhat.com>
Acked-by: Stefan Hajnoczi <stefanha@redhat.com>
Acked-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20200130163232.10446-9-philmd@redhat.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>

scripts/minikconf: Explicit usage of Python 3

Use the program search path to find the Python 3 interpreter.

Patch created manually after running:

  $ git grep -l 'if __name__.*__main__' \
      | xargs grep -LF '#!/usr/bin/env python3'

Reported-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Suggested-by: Daniel P. Berrangé <berrange@redhat.com>
Suggested-by: Stefan Hajnoczi <stefanha@redhat.com>
Acked-by: Stefan Hajnoczi <stefanha@redhat.com>
Acked-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20200130163232.10446-7-philmd@redhat.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>

scripts: Explicit usage of Python 3 (scripts with __main__)

Use the program search path to find the Python 3 interpreter.

Patch created mechanically by running:

  $ sed -i "s,^#\!/usr/bin/\(env\ \)\?python$,#\!/usr/bin/env python3," \
       $(git grep -l 'if __name__.*__main__')

Reported-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Suggested-by: Daniel P. Berrangé <berrange@redhat.com>
Suggested-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Acked-by: Stefan Hajnoczi <stefanha@redhat.com>
Acked-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20200130163232.10446-6-philmd@redhat.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>

tests: Explicit usage of Python 3

Use the program search path to find the Python 3 interpreter.

Patch created mechanically by running:

  $ sed -i "s,^#\!/usr/bin/\(env\ \)\?python$,#\!/usr/bin/env python3," \
       $(git grep -l 'if __name__.*__main__')

Reported-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Suggested-by: Daniel P. Berrangé <berrange@redhat.com>
Suggested-by: Stefan Hajnoczi <stefanha@redhat.com>
Acked-by: Stefan Hajnoczi <stefanha@redhat.com>
Acked-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20200130163232.10446-5-philmd@redhat.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>

tests/qemu-iotests: Explicit usage of Python 3 (scripts with __main__)

Use the program search path to find the Python 3 interpreter.

Patch created mechanically by running:

  $ sed -i "s,^#\!/usr/bin/\(env\ \)\?python$,#\!/usr/bin/env python3," \
       $(git grep -l 'if __name__.*__main__')

Reported-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Suggested-by: Daniel P. Berrangé <berrange@redhat.com>
Suggested-by: Stefan Hajnoczi <stefanha@redhat.com>
Acked-by: Stefan Hajnoczi <stefanha@redhat.com>
Acked-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20200130163232.10446-4-philmd@redhat.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>

tests/qemu-iotests/check: Allow use of python3 interpreter

As we want to enforce a unique and explicit Python 3 interpreter,
we need let this script handle 'python3' too.

Suggested-by: Kevin Wolf <kwolf@redhat.com>
Acked-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20200130163232.10446-3-philmd@redhat.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>

scripts/checkpatch.pl: Only allow Python 3 interpreter

Since commit ddf9069963 QEMU requires Python >= 3.5.

PEP 0394 [*] states that 'python3' should be available and
that 'python' is optional.

To avoid problem with unsupported versions, enforce the
shebang interpreter to Python 3.

[*] https://www.python.org/dev/peps/pep-0394/

Reported-by: John Snow <jsnow@redhat.com>
Acked-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <20200130163232.10446-2-philmd@redhat.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>

tests/acceptance/migration: Default to -nodefaults

We don't need the default options to run this test.

This fixes errors when running a binary built with
--without-default-devices such:

  ERROR: qemu-system-arm: Unsupported NIC model: virtio-net-pci

Reviewed-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Wainer dos Santos Moschetta <wainersm@redhat.com>
Message-Id: <20200129212345.20547-28-philmd@redhat.com>
[PMD: Rebased]
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>

tests/acceptance/migration: Add the 'migration' tag

By using an Avocado tag, we can run all tests described by that
tag as once:

  $ avocado --show=app run -t migration tests/acceptance/
  JOB ID     : 165477737509503fcfa6d7108057a0a18f2a6559
  JOB LOG    : avocado/job-results/job-2020-02-04T17.29-1654777/job.log
   (1/3) tests/acceptance/migration.py:Migration.test_migration_with_tcp_localhost: PASS (0.38 s)
   (2/3) tests/acceptance/migration.py:Migration.test_migration_with_unix: PASS (0.33 s)
   (3/3) tests/acceptance/migration.py:Migration.test_migration_with_exec: PASS (0.07 s)
  RESULTS    : PASS 3 | ERROR 0 | FAIL 0 | SKIP 0 | WARN 0 | INTERRUPT 0 | CANCEL 0

Reviewed-by: Wainer dos Santos Moschetta <wainersm@redhat.com>
Message-Id: <20200204163304.14616-1-philmd@redhat.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>

tests/acceptance/migration: Test EXEC transport when migrating

Along with VM migration via TCP, we can use migration through
the EXEC transport protocol

Signed-off-by: Oksana Vohchana <ovoshcha@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-Id: <20200203111631.18796-3-ovoshcha@redhat.com>
[PMD: Split patch in 2, reworded subject and description]
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>

tests/acceptance/migration: Test UNIX transport when migrating

Along with VM migration via TCP, we can use migration through
the UNIX transport protocol

Signed-off-by: Oksana Vohchana <ovoshcha@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-Id: <20200203111631.18796-3-ovoshcha@redhat.com>
[PMD: Split patch in 2, reworded subject and description]
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>

tests/acceptance/migration: Factor out do_migrate()

We are going to reuse this code when testing different transport
methods, so factor it out first.

Signed-off-by: Oksana Vohchana <ovoshcha@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-Id: <20200203111631.18796-2-ovoshcha@redhat.com>
[PMD: Split patch in 2, reworded subject and description]
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>

tests/acceptance/migration: Factor out assert_migration()

We are going to reuse this code when testing different transport
methods, so factor it out first

Signed-off-by: Oksana Vohchana <ovoshcha@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-Id: <20200203111631.18796-2-ovoshcha@redhat.com>
[PMD: Split patch in 2, reworded subject and description]
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>

tests/acceptance/version: Default to -nodefaults

We don't need the default options to run this test.

This fixes errors when running a binary built with
--without-default-devices such:

  ERROR: qemu-system-arm: Unsupported NIC model: virtio-net-pci

Reviewed-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Wainer dos Santos Moschetta <wainersm@redhat.com>
Message-Id: <20200129212345.20547-29-philmd@redhat.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>

tests/acceptance/boot_linux_console: Do not use VGA on Clipper machine

As we only read the serial console, we don't need to force a
VGA display. This fixes when running a binary built with
--without-default-devices:

  ERROR: qemu-system-alpha: standard VGA not available

We also need the '-nodefaults' argument to avoid:

  ERROR: qemu-system-alpha: Unsupported NIC model: e1000

Reviewed-by: Thomas Huth <thuth@redhat.com>
Message-Id: <20200129212345.20547-27-philmd@redhat.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>

tests/acceptance/virtio_check_params: Disable the test

This test fails on various CI:

- Using QEMU 4.0:
  tests/acceptance/x86_cpu_model_versions.py:X86CPUModelAliases.test_none_alias:  ERROR: 'alias-of' (0.45 s)

- On OSX
  Unexpected error in object_property_find() at qom/object.c:1201:
  qemu-system-x86_64: -device virtio-blk-pci,id=scsi0,drive=drive0: can't apply global virtio-blk-device.scsi=true: Property '.scsi' not found

- When removing unavailable machine:
  VirtioMaxSegSettingsCheck.test_machine_types: ERROR: list.remove(x): x not in list (0.12 s)

- Using Xen:
  xencall: error: Could not obtain handle on privileged command interface: No such file or directory
  xen be core: xen be core: can't open xen interface

- On PPC:
  TestFail: machine type pseries-2.8: No Transactional Memory support in TCG, try appending -machine cap-htm=off

- On S390X configured with --without-default-devices:
  ERROR: qemu-system-s390x: -device virtio-scsi-pci,id=scsi0: 'virtio-scsi-pci' is not a valid device model name

Disable it for now.

Cc: Denis Plotnikov <dplotnikov@virtuozzo.com>
Acked-by: Cornelia Huck <cohuck@redhat.com>
Message-Id: <20200206171715.25041-1-philmd@redhat.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>

tests/acceptance/virtio_check_params: Default to -nodefaults

We don't need the default options to run this test.

This fixes errors when running a binary built with
--without-default-devices such:

  ERROR: qemu-system-s390x: Unknown device 'virtio-net-ccw' for bus 'virtual-css-bus'

Reviewed-by: Thomas Huth <thuth@redhat.com>
Message-Id: <20200129212345.20547-25-philmd@redhat.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>

tests/acceptance/virtio_check_params: List machine being tested

Add logging for easier debugging of failures:

  $ avocado --show=machine run tests/acceptance/virtio_check_params.py
   (1/1) tests/acceptance/virtio_check_params.py:VirtioMaxSegSettingsCheck.test_machine_types:
  machine: {'name': 'pc-i440fx-2.12', 'seg_max_adjust': 'false', 'device': 'virtio-scsi-pci'}
  machine: {'name': 'pc-i440fx-2.0', 'seg_max_adjust': 'false', 'device': 'virtio-scsi-pci'}
  machine: {'name': 'pc-q35-4.2', 'seg_max_adjust': 'false', 'device': 'virtio-scsi-pci'}
  machine: {'name': 'pc-i440fx-2.5', 'seg_max_adjust': 'false', 'device': 'virtio-scsi-pci'}
  machine: {'name': 'pc-i440fx-4.2', 'seg_max_adjust': 'false', 'device': 'virtio-scsi-pci'}
  ...

Reviewed-by: Cornelia Huck <cohuck@redhat.com>
Message-Id: <20200129212345.20547-19-philmd@redhat.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>

tests/acceptance/virtio_check_params: Improve exception logging

Message-Id: <20200129212345.20547-18-philmd@redhat.com>
Reviewed-by: Eduardo Habkost <ehabkost@redhat.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>

tests: rename virtio_seg_max_adjust to virtio_check_params

Since, virtio_seg_max_adjust checks not only seg_max, but also
virtqueue_size parameter, let's make the test more general and
add new parameters to be checked there in the future.

Signed-off-by: Denis Plotnikov <dplotnikov@virtuozzo.com>
Message-Id: <20200129140702.5411-5-dplotnikov@virtuozzo.com>
Acked-by: Cornelia Huck <cohuck@redhat.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>

tests/acceptance: Add boot tests for some of the QEMU advent calendar images

The 2018 edition of the QEMU advent calendar 2018 featured Linux images
for various non-x86 machines. We can use them for a boot tests in our
acceptance test suite.

Let's also make sure that we build the corresponding machines in Travis.

Signed-off-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Liam Merwick <liam.merwick@oracle.com>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Tested-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Wainer dos Santos Moschetta <wainersm@redhat.com>
Message-Id: <20200124170325.30072-1-thuth@redhat.com>
[PMD: Rebased, --python=python3 parameter dropped in commit 5311cb12e]
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>

tests/boot_linux_console: Tag Emcraft Smartfusion2 as running 'u-boot'

Avocado tags are handy to automatically select tests matching
the tags. Since this test also runs U-Boot, tag it.

We can run all the tests using U-Boot as once with:

  $ avocado --show=app run -t u-boot tests/acceptance/
  JOB LOG    : avocado/job-results/job-2020-01-21T00.16-ee9344e/job.log
   (1/3) tests/acceptance/boot_linux_console.py:BootLinuxConsole.test_arm_emcraft_sf2: PASS (16.59 s)
   (2/3) tests/acceptance/boot_linux_console.py:BootLinuxConsole.test_arm_raspi2_uboot: PASS (0.47 s)
   (3/3) tests/acceptance/boot_linux_console.py:BootLinuxConsole.test_aarch64_raspi3_uboot: PASS (2.43 s)
  RESULTS    : PASS 3 | ERROR 0 | FAIL 0 | SKIP 0 | WARN 0 | INTERRUPT 0 | CANCEL 0
  JOB TIME   : 19.78 s

Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Wainer dos Santos Moschetta <wainersm@redhat.com>
Message-Id: <20200120235159.18510-8-f4bug@amsat.org>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>

tests/boot_linux_console: fix extract_from_deb() comment

The second param in extract_from_deb() is 'path' not 'file'

Signed-off-by: Liam Merwick <liam.merwick@oracle.com>
Reviewed-by: Stefano Garzarella <sgarzare@redhat.com>
Reviewed-by: Wainer dos Santos Moschetta <wainersm@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-Id: <1580142994-1836-4-git-send-email-liam.merwick@oracle.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>

tests/boot_linux_console: use os.path for filesystem paths

Change extract_from_deb() to use os.path routines to manipulate the
filesystem path returned when extracting a file.

Suggested-by: Wainer dos Santos Moschetta <wainersm@redhat.com>
Signed-off-by: Liam Merwick <liam.merwick@oracle.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-Id: <1580914565-19675-7-git-send-email-liam.merwick@oracle.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>

tests/boot_linux_console: add extract_from_rpm method

Add a method to extract a specified file from an RPM to the test's
working directory and return the path to the extracted file.

Signed-off-by: Liam Merwick <liam.merwick@oracle.com>
Reviewed-by: Stefano Garzarella <sgarzare@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-Id: <1580914565-19675-5-git-send-email-liam.merwick@oracle.com>
[PMD: Rebased]
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>

travis.yml: install rpm2cpio for acceptance tests

The extract_from_rpm() method added for the PVH acceptance tests needs
rpm2cpio to extract a vmlinux binary from an RPM.

Signed-off-by: Liam Merwick <liam.merwick@oracle.com>
Reviewed-by: Stefano Garzarella <sgarzare@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-Id: <1580914565-19675-4-git-send-email-liam.merwick@oracle.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>

Acceptance tests: Add interrupt_interactive_console_until_pattern()

We need a function to interrupt interactive consoles.

Example: Interrupt U-Boot to set different environment values.

Tested-by: Niek Linnenbank <nieklinnenbank@gmail.com>
Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Liam Merwick <liam.merwick@oracle.com>
Tested-by: Liam Merwick <liam.merwick@oracle.com>
Reviewed-by: Wainer dos Santos Moschetta <wainersm@redhat.com>
Message-Id: <20200120235159.18510-4-f4bug@amsat.org>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>

Acceptance tests: Extract _console_interaction()

Since we are going to re-use the code shared between
wait_for_console_pattern() and exec_command_and_wait_for_pattern(),
extract the common part into a local function.

Tested-by: Niek Linnenbank <nieklinnenbank@gmail.com>
Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Liam Merwick <liam.merwick@oracle.com>
Tested-by: Liam Merwick <liam.merwick@oracle.com>
Reviewed-by: Wainer dos Santos Moschetta <wainersm@redhat.com>
Message-Id: <20200120235159.18510-3-f4bug@amsat.org>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>

python/qemu/machine: Allow to use other serial consoles than default

Currently the QEMU Python module limits the QEMUMachine class to
use the first serial console.

Some machines/guest might use another console than the first one as
the 'boot console'. For example the Raspberry Pi uses the second
(AUX) console.

To be able to use the Nth console as default, we simply need to
connect all the N - 1 consoles to the null chardev.

Add an index argument, so we can use a specific serial console as
default.

Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Liam Merwick <liam.merwick@oracle.com>
Tested-by: Liam Merwick <liam.merwick@oracle.com>
Reviewed-by: Wainer dos Santos Moschetta <wainersm@redhat.com>
Message-Id: <20200120235159.18510-5-f4bug@amsat.org>
[PMD: zero-initialize _console_index in __init__()]
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>

python: Treat None-return of greeting cmd

In case qemu process dies the "monitor.cmd" returns None which gets
passed to the "__negotiate_capabilities" and leads to unhandled
exception. Let's only check the resp in case it has a value.

Signed-off-by: Lukáš Doktor <ldoktor@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Wainer dos Santos Moschetta <wainersm@redhat.com>
Message-Id: <20200120071202.30646-1-ldoktor@redhat.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>

qemu-deprecated: Remove text about Python 2

Python 2 support has been removed, so we should now also remove
the announcement text for the deprecation.

Signed-off-by: Thomas Huth <thuth@redhat.com>
Reviewed by: Aleksandar Markovic <amarkovic@wavecomp.com>
Reviewed-by: John Snow <jsnow@redhat.com>
Message-Id: <20200109095116.18201-1-thuth@redhat.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>

python/qemu: accel: Fix kvm_available() on ppc64le

On ppc64le, the accel.kvm_available() check may wrongly
return False because the host arch (as returned by os.uname[4])
and the target arch (ppc64) mismatch. In order to solve this
it is added an ppc64le -> ppc64 mapping which is used as an
fallback verification.

Fixes: 53a049d7d78e5ccf6d4c0d7
Signed-off-by: Wainer dos Santos Moschetta <wainersm@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-Id: <20200205203250.30526-5-wainersm@redhat.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>

python/qemu: qmp: Remove unnused attributes

The `error` and `timeout` attributes in QEMUMonitorProtocol are
not used, so this delete them.

Signed-off-by: Wainer dos Santos Moschetta <wainersm@redhat.com>
Reviewed-by: John Snow <jsnow@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-Id: <20191227134101.244496-6-wainersm@redhat.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>

python/qemu: qmp: Make QEMUMonitorProtocol a context manager

This implement the __enter__ and __exit__ functions on
QEMUMonitorProtocol class so that it can be used on 'with'
statement and the resources will be free up on block end:

with QEMUMonitorProtocol(socket_path) as qmp:
    qmp.connect()
    qmp.command('query-status')

Signed-off-by: Wainer dos Santos Moschetta <wainersm@redhat.com>
Reviewed-by: John Snow <jsnow@redhat.com>
Message-Id: <20200204141111.3207-5-wainersm@redhat.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>

python/qemu: qmp: Make accept()'s timeout configurable

Currently the timeout of QEMUMonitorProtocol.accept() is
hard-coded to 15.0 seconds. This added the parameter `timeout`
so the value can be configured by the user.

Signed-off-by: Wainer dos Santos Moschetta <wainersm@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: John Snow <jsnow@redhat.com>
Message-Id: <20200204141111.3207-4-wainersm@redhat.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>

python/qemu: Delint the qmp module

This clean up the pylint-3 report on qmp:

************* Module qemu.qmp
python/qemu/qmp.py:1:0: C0111: Missing module docstring (missing-docstring)
python/qemu/qmp.py:17:0: C0111: Missing class docstring (missing-docstring)
python/qemu/qmp.py:21:0: C0111: Missing class docstring (missing-docstring)
python/qemu/qmp.py:25:0: C0111: Missing class docstring (missing-docstring)
python/qemu/qmp.py:29:0: C0111: Missing class docstring (missing-docstring)
python/qemu/qmp.py:33:0: C0111: Missing class docstring (missing-docstring)
python/qemu/qmp.py:33:0: R0205: Class 'QEMUMonitorProtocol' inherits from object, can be safely removed from bases in python3 (useless-object-inheritance)
python/qemu/qmp.py:80:4: R1710: Either all return statements in a function should return an expression, or none of them should. (inconsistent-return-statements)
python/qemu/qmp.py:131:4: R1710: Either all return statements in a function should return an expression, or none of them should. (inconsistent-return-statements)
python/qemu/qmp.py:159:4: R1710: Either all return statements in a function should return an expression, or none of them should. (inconsistent-return-statements)
python/qemu/qmp.py:245:4: C0111: Missing method docstring (missing-docstring)
python/qemu/qmp.py:249:4: C0111: Missing method docstring (missing-docstring)
python/qemu/qmp.py:252:4: C0111: Missing method docstring (missing-docstring)
python/qemu/qmp.py:255:4: C0111: Missing method docstring (missing-docstring)

Signed-off-by: Wainer dos Santos Moschetta <wainersm@redhat.com>
Reviewed-by: John Snow <jsnow@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-Id: <20191227134101.244496-3-wainersm@redhat.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>

python/qemu: qmp: Replace socket.error with OSError

The socket.error is deprecated from Python 3.3, instead it is
made a link to OSError. This change replaces the occurences
of socket.error with OSError.

Signed-off-by: Wainer dos Santos Moschetta <wainersm@redhat.com>
Reviewed-by: John Snow <jsnow@redhat.com>
Message-Id: <20191227134101.244496-2-wainersm@redhat.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>

stellaris: delay timer_new to avoid memleaks

There is a memory leak when we call 'device_list_properties' with typename = stellaris-gptm. It's easy to reproduce as follow:

  virsh qemu-monitor-command vm1 --pretty '{"execute": "device-list-properties", "arguments": {"typename": "stellaris-gptm"}}'

This patch delay timer_new in realize to fix it.

Reported-by: Euler Robot <euler.robot@huawei.com>
Signed-off-by: Pan Nengyuan <pannengyuan@huawei.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-id: 20200205070659.22488-4-pannengyuan@huawei.com
Cc: qemu-arm@nongnu.org
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

stm32f2xx_timer: delay timer_new to avoid memleaks

There is a memory leak when we call 'device_list_properties' with typename = stm32f2xx_timer. It's easy to reproduce as follow:

    virsh qemu-monitor-command vm1 --pretty '{"execute": "device-list-properties", "arguments": {"typename": "stm32f2xx_timer"}}'

This patch delay timer_new to fix this memleaks.

Reported-by: Euler Robot <euler.robot@huawei.com>
Signed-off-by: Pan Nengyuan <pannengyuan@huawei.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Message-id: 20200205070659.22488-3-pannengyuan@huawei.com
Cc: Alistair Francis <alistair@alistair23.me>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

armv7m_systick: delay timer_new to avoid memleaks

There is a memory leak when we call 'device_list_properties' with typename = armv7m_systick. It's easy to reproduce as follow:

  virsh qemu-monitor-command vm1 --pretty '{"execute": "device-list-properties", "arguments": {"typename": "armv7m_systick"}}'

This patch delay timer_new to fix this memleaks.

Reported-by: Euler Robot <euler.robot@huawei.com>
Signed-off-by: Pan Nengyuan <pannengyuan@huawei.com>
Message-id: 20200205070659.22488-2-pannengyuan@huawei.com
Cc: qemu-arm@nongnu.org
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

docs/arm-cpu-features: Make kvm-no-adjvtime comment clearer

The bold text sounds like 'knock knock'. Only bolding the
second 'not' makes it easier to read.

Fixes: dea101a1ae
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Andrew Jones <drjones@redhat.com>
Message-id: 20200206225148.23923-1-philmd@redhat.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

bcm2835_dma: Re-initialize xlen in TD mode

TD (two dimensions) DMA mode did not work, because the xlen variable
has not been re-initialized before each additional ylen run through
in bcm2835_dma_update(). Fix it.

Signed-off-by: Rene Stange <rsta2@o2online.de>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

bcm2835_dma: Fix the ylen loop in TD mode

In TD (two dimensions) DMA mode ylen has to be increased by one after
reading it from the TXFR_LEN register, because a value of zero has to
result in one run through of the ylen loop. This has been tested on a
real Raspberry Pi 3 Model B+. In the previous implementation the ylen
loop was not passed at all for a value of zero.

Signed-off-by: Rene Stange <rsta2@o2online.de>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

target/arm: Raise only one interrupt in arm_cpu_exec_interrupt

The fall through organization of this function meant that we
would raise an interrupt, then might overwrite that with another.
Since interrupt prioritization is IMPLEMENTATION DEFINED, we
can recognize these in any order we choose.

Unify the code to raise the interrupt in a block at the end.

Tested-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20200206105448.4726-42-richard.henderson@linaro.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>