Commit Graph

89 Commits

Author SHA1 Message Date
Gerd Hoffmann 4423184376 spice/gl: render DisplaySurface via opengl
This switches over spice (in opengl mode) to render DisplaySurface
updates into a opengl texture, using the helper functions in
ui/console-gl.c.  With this patch applied spice (with gl=on) will
stop using qxl rendering ops, it will use dma-buf passing all the
time, i.e. for bios/bootloader (before virtio-gpu driver is loaded)
too.

This should improve performance even using spice (with gl=on) with
non-accelerated stdvga because we stop squeezing all display updates
through a unix/tcp socket and basically using a shared memory transport
instead.

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Message-id: 1474617028-3979-3-git-send-email-kraxel@redhat.com
2016-09-28 12:49:36 +02:00
Marc-André Lureau 9d8256ebc0 virgl: pass whole GL scanout dimensions
Spice client needs the whole GL texture dimension to be able to show a
scanout with a monitor offset (different than +0+0).

Furthermore, this fixes a crash when calling surface_{width,height}()
after dpy_gfx_replace_surface(con, NULL) was called in
virgl_cmd_set_scanout()

Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-id: 1465911849-30423-4-git-send-email-marcandre.lureau@redhat.com
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
2016-07-06 10:32:14 +02:00
Marc-André Lureau c61d8126fc spice: use the right head for multi-monitor
Look up the associated head monitor config.

Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-id: 1465911849-30423-3-git-send-email-marcandre.lureau@redhat.com
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
2016-07-06 10:32:14 +02:00
John Snow 015e02f880 spice: avoid .set_mm_time on >= 0.12.6
Spice deprecated this callback in 0.12.6.
It's not a problem yet, but it will cause Clang to fail in a -Werror
build due to the deprecated tag.

Signed-off-by: John Snow <jsnow@redhat.com>
Message-id: 1467240095-12507-2-git-send-email-jsnow@redhat.com
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
2016-07-06 10:31:57 +02:00
Laurent Vivier 5d61cafd0b SPICE: Use DIV_ROUND_UP
Replace (((n) + (d) - 1) /(d)) by DIV_ROUND_UP(n,d).

This patch is the result of coccinelle script
scripts/coccinelle/round.cocci

CC: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Laurent Vivier <lvivier@redhat.com>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
2016-06-07 18:19:25 +03:00
Gonglei 28f4a7083d spice: fix coverity complains
Remove the unnecessary NULL check.

Signed-off-by: Gonglei <arei.gonglei@huawei.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Message-id: 1463047028-123868-3-git-send-email-arei.gonglei@huawei.com
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
2016-05-12 16:41:46 +02:00
Gerd Hoffmann 39414ef4e9 spice/gl: add & use qemu_spice_gl_monitor_config
Cc: qemu-stable@nongnu.org
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
2016-05-12 16:41:46 +02:00
Gerd Hoffmann 22672a3798 spice/gl: tweak debug messages.
Adjust message levels, make messages more verbose.

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
2016-02-23 12:04:40 +01:00
Gerd Hoffmann 8e388e907b spice/gl: add unblock timer
Pure debug aid, print a warning in case unblocking
doesn't happen within one second.

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
2016-02-23 12:04:40 +01:00
Gerd Hoffmann 474114b730 spice: add opengl/virgl/dmabuf support
This adds support for dma-buf passing to spice.  This makes virtio-gpu
with 3d acceleration work with spice.

Workflow:
 * virglrenderer renders the guest command stream into a texture.
 * qemu exports the texture as dma-buf and passes on that dma-buf
   to spice-server.
 * spice-server passes the dma-buf to spice-client, using unix
   socket file descriptor passing.
 * spice-client asks the window systems composer to render the
   dma-buf to the screen.

Requires cutting edge spice (server) and spice-gtk (client) builds,
from git master branch.

Also requires libvirt managing your qemu instance, and using
"virt-viewer --attach $guest".  libvirt will connect spice-server and
spice-client using unix sockets instead of tcp sockets then, which
is required for file descriptor passing.

Works for the local case (spice server and client on the same machine)
only.  Supporting remote too is planned (by feeding the dma-bufs into
gpu-assisted video encoder), but not there yet.

gl mode is turned off by default, use "-spice gl=on,$otherargs" to
enable it.

Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
2016-02-23 12:04:39 +01:00
Marc-André Lureau 58c7b618f3 spice: reset cursor on resize
Spice server will clear the cursor on resize. QXL driver reset it after
resize, however, virtio and other devices do not. Teach qemu to set it
back.

Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
2016-02-23 12:04:39 +01:00
Gerd Hoffmann b5e751b51f spice: init dcl before registering qxl interface
Without this spice might callback into qemu before ssd->dcl.con is
initialized, resulting in a segfault due to NULL pointer dereference.

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
2016-02-23 12:04:39 +01:00
Peter Maydell e16f4c8770 ui: Clean up includes
Clean up includes so that osdep.h is included first and headers
which it implies are not included manually.

This commit was created with scripts/clean-includes.

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Message-id: 1454089805-5470-2-git-send-email-peter.maydell@linaro.org
2016-02-04 17:01:04 +00:00
Gerd Hoffmann b2af43cc37 spice: surface switch fast path requires same format too.
Commit "555e72f spice: rework mirror allocation, add no-resize fast path"
adds a fast path for surface switches which does't go through the full
primary surface destroy and re-recreation in case the new surface is
identical to the old one (page-flip).  It checks the size only though,
but the format must be identical too.  This patch adds the format check.

Commit "0002a51 ui/spice: Support shared surface for most pixman
formats" increases the chance to actually trigger this.

https://bugzilla.redhat.com/show_bug.cgi?id=1247479

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
2015-09-21 09:52:07 +02:00
Daniel P. Berrange ef1e1e0782 maint: avoid useless "if (foo) free(foo)" pattern
The free() and g_free() functions both happily accept
NULL on any platform QEMU builds on. As such putting a
conditional 'if (foo)' check before calls to 'free(foo)'
merely serves to bloat the lines of code.

Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
2015-09-11 10:21:38 +03:00
Markus Armbruster a0b1a66ea3 Include monitor/monitor.h exactly where needed
In particular, don't include it into headers.

Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Luiz Capitulino <lcapitulino@redhat.com>
2015-06-22 18:20:41 +02:00
Gerd Hoffmann 5a9259a0b5 spice: ui_info tweaks
Use the new dpy_ui_info_supported function.
Clarifies the control flow.

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
2015-06-11 09:06:14 +02:00
Gerd Hoffmann c6e484707f spice-display: fix segfault in qemu_spice_create_update
Although it is pretty unusual the stride for the guest image and the
mirror image maintained by spice-display can be different.  So use
separate variables for them.

https://bugzilla.redhat.com/show_bug.cgi?id=1163047

Cc: qemu-stable@nongnu.org
Reported-by: perrier vincent <clownix@clownix.net>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
2015-06-11 09:06:14 +02:00
Chih-Min Chao c9f88ce330 hw/display : remove 'struct' from 'typedef QXL struct'
Signed-off-by: Chih-Min Chao <cmchao@gmail.com>
Reviewed-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
2015-04-30 16:05:48 +03:00
Chih-Min Chao 9425c004fe ui/console : remove 'struct' from 'typedef struct' type
Signed-off-by: Chih-Min Chao <cmchao@gmail.com>
Reviewed-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
2015-04-30 16:05:48 +03:00
Marc-André Lureau 700cd855de spice: learn to hide cursor
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
2015-04-27 12:47:04 +02:00
Marc-André Lureau dc8dceee64 spice: set pointer position on hotspot
The Spice protocol uses cursor position on hotspot: the client is
applying hotspot offset when drawing the cursor.

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
2015-04-27 12:47:04 +02:00
Marc-André Lureau d0df04a156 spice: fix mouse cursor position
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
2015-04-27 12:47:04 +02:00
Gerd Hoffmann c1d37cd353 spice: fix simple display on bigendian hosts
Denis Kirjanov is busy getting spice run on ppc64 and trapped into this
one.  Spice wire format is little endian, so we have to explicitly say
we want little endian when letting pixman convert the data for us.

Reported-by: Denis Kirjanov <kirjanov@gmail.com>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
2015-04-27 12:47:03 +02:00
Gerd Hoffmann 51a0909914 spice: fix coverity reported defect in display code
Report:

1. Condition surface, taking false branch
406    if (surface && ssd->surface &&
407        surface_width(surface) == pixman_image_get_width(ssd->surface) &&
408        surface_height(surface) == pixman_image_get_height(ssd->surface)) {
409        /* no-resize fast path: just swap backing store */
...

10. alias_transfer: Assigning: ssd->ds = surface.
440    ssd->ds = surface;

11. var_deref_op: Dereferencing null pointer ssd->ds.
CID 1264334 (#1 of 1): Dereference after null check (FORWARD_NULL)
441    ssd->surface = pixman_image_ref(ssd->ds->image);

Fix:

Move code block dereferencing ssd->ds into the already existing
if (ssd->ds) { ... } block.

Cc: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
2015-01-22 11:18:41 +01:00
Gerd Hoffmann 0002a51889 ui/spice: Support shared surface for most pixman formats
Just hook up qemu_pixman_check_format.

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
2015-01-19 13:33:26 +01:00
Gerd Hoffmann 555e72f2d0 spice: rework mirror allocation, add no-resize fast path
Add fast path to qemu_spice_display_switch in case old and new
displaysurface have identical size (happens with display panning
and page flipping).  We just swap the backing store then and don't
go through the whole process of deleting and creating the primary
surface.

To simplify the code a bit move mirror surface allocation to
qemu_spice_display_switch().

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
2014-12-16 14:15:29 +01:00
Gerd Hoffmann 0b2824e5e4 spice: use bottom half instead of refresh timer for cursor updates
Calling directly doesn't work due to the qxl-render code running in
spice server thread context.  Meanwhile bottom half scheduling is
thread-safe though, so we can use that to kick a cursor update in
main i/o thread context.

Cc: Marc-André Lureau <marcandre.lureau@gmail.com>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
2014-12-16 14:15:29 +01:00
Gerd Hoffmann ab9509ccea spice: make sure we don't overflow ssd->buf
Related spice-only bug.  We have a fixed 16 MB buffer here, being
presented to the spice-server as qxl video memory in case spice is
used with a non-qxl card.  It's also used with qxl in vga mode.

When using display resolutions requiring more than 16 MB of memory we
are going to overflow that buffer.  In theory the guest can write,
indirectly via spice-server.  The spice-server clears the memory after
setting a new video mode though, triggering a segfault in the overflow
case, so qemu crashes before the guest has a chance to do something
evil.

Fix that by switching to dynamic allocation for the buffer.

CVE-2014-3615

Cc: qemu-stable@nongnu.org
Cc: secalert@redhat.com
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Reviewed-by: Laszlo Ersek <lersek@redhat.com>
2014-09-05 12:19:50 +02:00
Gerd Hoffmann e8e23b7dcf spice: fix 32bit build
Tested-by: Luiz Capitulino <lcapitulino@redhat.com>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Message-id: 1403244764-8622-1-git-send-email-kraxel@redhat.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
2014-06-20 16:22:07 +01:00
Gerd Hoffmann 5643fc012c spice: add mouse cursor support
So you'll have a mouse pointer when running non-qxl gfx cards with
mouse pointer support (virtio-gpu, IIRC vmware too).

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
2014-06-13 12:34:57 +02:00
Gerd Hoffmann dc491cfc14 spice: monitors_config: check pointer before dereferencing
Reported-by: Fabio Fantoni <fabio.fantoni@m2r.biz>
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
2014-04-07 12:18:43 +02:00
Gerd Hoffmann 9b74d0d598 spice: QemuUIInfo windup
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
2014-03-10 13:55:25 +01:00
Gerd Hoffmann 4b87dc4c97 spice: fix simple display surface handling
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
2014-03-10 13:55:20 +01:00
Gerd Hoffmann 7cc6a25fe9 spice: move qemu_spice_display_*() from spice-graphics to spice-core
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
2013-12-16 10:12:20 +01:00
Gerd Hoffmann 9fa032866d spice: fix multihead support
This patch fixes spice display initialization to handle
multihead properly.

spice-core now keeps track of which QemuConsole has a spice
display channel attached to it and which has not.  It also
manages display channel ids.

spice-display looks at all QemuConsoles and will pick up any
graphic console not yet bound to a spice channel (which in practice
are all non-qxl graphic devices).

Result is that
 (a) you'll get a spice client window for each graphical device
     now (first only without this patch), and
 (b) mixing qxl and non-qxl vga cards works properly.

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
2013-10-17 12:42:54 +02:00
Gerd Hoffmann 35b2122db4 spice-display: add display channel id to the debug messages.
And s/__FUNCTION__/__func__/ while being at it.

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
2013-10-17 12:41:03 +02:00
Marc-André Lureau 26defe81f6 spice: replace use of deprecated API
hose API are deprecated since 0.11, and qemu depends on 0.12 already.

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
2013-10-17 12:25:25 +02:00
Gerd Hoffmann 5209089fcd console: zap ds arg from register_displaychangelistener
We don't have multiple DisplayStates any more,
so passing it in as argument is not needed.

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
2013-04-25 14:45:46 -05:00
Gerd Hoffmann 284d1c6b3b console: allow pinning displaychangelisteners to consoles
DisplayChangeListener gets a new QemuConsole field, which can be set to
non-NULL before registering.  This will pin the QemuConsole, so that
particular DisplayChangeListener will not follow console switches.

spice+gtk (which don't support text console input anyway) are switched
over to be pinned to console 0, which usually is the graphical display.

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
2013-04-16 09:26:20 +02:00
Gerd Hoffmann 1dbfa00503 console: rename vga_hw_*, add QemuConsole param
Add QemuConsole parameter to vga_hw_*, so the interface allows to update
non-active consoles (the actual code can't handle this yet, see next
patch).  Passing NULL is allowed and updates the active console, like
the functions do today.

While touching all vga_hw_* calls anyway rename that to the functions to
hardware-neutral graphics_hw_*

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
2013-04-16 09:03:48 +02:00
Gerd Hoffmann c78f71378a console: stop using DisplayState in gfx hardware emulation
Use QemuConsole instead.  Updates interfaces in console.[ch] and adapts
gfx hardware emulation code.

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
2013-03-18 10:21:59 +01:00
Gerd Hoffmann bc2ed9704f console: zap displaystate from dcl callbacks
Now that nobody depends on DisplayState in DisplayChangeListener
callbacks any more we can remove the parameter from all callbacks.

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
2013-03-18 10:21:59 +01:00
Gerd Hoffmann 71874c1751 spice: stop using DisplayState
Rework DisplayStateListener callbacks to not use the DisplayState
any more.

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
2013-03-18 10:21:59 +01:00
Gerd Hoffmann c12aeb860c console: rework DisplaySurface handling [dcl/ui side]
Replace the dpy_gfx_resize and dpy_gfx_setdata DisplayChangeListener
callbacks with a dpy_gfx_switch callback which notifies the ui code
when the framebuffer backing storage changes.

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
2013-03-18 10:21:58 +01:00
Gerd Hoffmann 9c80a315b4 spice: zap sdpy global
DisplayChangeListener is passed now to all DisplayChangeListenerOps
callbacks, so we can use that to access the spice display state and
kill the sdpy global variable.

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
2013-03-18 10:21:58 +01:00
Gerd Hoffmann 7c20b4a374 console: fix displaychangelisteners interface
Split callbacks into separate Ops struct.  Pass DisplayChangeListener
pointer as first argument to all callbacks.  Uninline a bunch of
display functions and move them from console.h to console.c

Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
2013-03-18 10:21:58 +01:00
Paolo Bonzini 9c17d615a6 softmmu: move include files to include/sysemu/
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2012-12-19 08:32:45 +01:00
Paolo Bonzini 1de7afc984 misc: move include files to include/qemu/
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2012-12-19 08:32:39 +01:00
Paolo Bonzini 83c9089e73 monitor: move include files to include/monitor/
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2012-12-19 08:31:32 +01:00