>
> >
> > This patch fixes 2 issues:
> >
> > 1. Framebuffer should only be used in vga mode,
> > therefore when QxlDevice is active
> > FrameBufferIsActive flag shouldn't be checked;
> > 2. FrameBufferIsActive flag should be set true
> > on successfull frame buffer allocation
>
> The PAGED_CODE macro ensures that the calling thread is running at an
> IRQL that is low enough to permit paging. A call to this macro should
> be made at the beginning of every driver routine that either contains
> pageable code or accesses pageable code.
>
> Based on a patch by Sandy Stutsm
>
> This patch fixes 2 issues:
>
> 1. Framebuffer should only be used in vga mode,
> therefore when QxlDevice is active
> FrameBufferIsActive flag shouldn't be checked;
> 2. FrameBufferIsActive flag should be set true
> on successfull frame buffer allocation only.
>
> Signed-o
>
> Framebuffer should only be used in VGA mode, as WDDM DOD
> driver doesn't use the frame buffer (bar0), so no
> reason to map in into memory. However the mode is only known
> at runtime therefore framebuffer logic should be active when the
> driver is operating in vga mode only.
>
> There were
>
> Part of source image mapped by PresentDisplayOnly
> should be big enough to cover all rectangles being
> transferred.
>
> Signed-off-by: Sameeh Jubran
> Signed-off-by: Dmitry Fleytman
> ---
> qxldod/QxlDod.cpp | 28 +++-
> qxldod/QxlDod.h | 2 ++
> 2 files change
On Mon, Sep 26, 2016 at 4:33 PM, Frediano Ziglio wrote:
> >
> > Interrupts seem to arrive to the driver before the initialization phase
> > is over (m_pHWDevice = NULL), in that case we can't handle interrupts
> yet.
>
> I would remove the "seem"
>
> > Even
> > when m_pHWDevice isn't NULL, other
>
> Disable execution bit on mapping improving security.
>
> MmMapIoSpaceEx is available only in Windows 10 thus
> the macros are used.
>
> Based on a patch by Sandy Stutsman
>
> Signed-off-by: Sameeh Jubran
Acked-by: Frediano Ziglio
I'll wait couple if days, there were different comments
>
> Interrupts seem to arrive to the driver before the initialization phase
> is over (m_pHWDevice = NULL), in that case we can't handle interrupts yet.
I would remove the "seem"
> Even
> when m_pHWDevice isn't NULL, other fields aren't necessarly fully intialized
> till
> the StartDevice functi
The PAGED_CODE macro ensures that the calling thread is running at an
IRQL that is low enough to permit paging. A call to this macro should
be made at the beginning of every driver routine that either contains
pageable code or accesses pageable code.
Based on a patch by Sandy Stutsman
Signed-off
Framebuffer should only be used in VGA mode, as WDDM DOD
driver doesn't use the frame buffer (bar0), so no
reason to map in into memory. However the mode is only known
at runtime therefore framebuffer logic should be active when the
driver is operating in vga mode only.
There were rare BSOD failur
This patch fixes 2 issues:
1. Framebuffer should only be used in vga mode,
therefore when QxlDevice is active
FrameBufferIsActive flag shouldn't be checked;
2. FrameBufferIsActive flag should be set true
on successfull frame buffer allocation only.
Signed-off-by: Dmitry Fleytma
Disable execution bit on mapping improving security.
MmMapIoSpaceEx is available only in Windows 10 thus
the macros are used.
Based on a patch by Sandy Stutsman
Signed-off-by: Sameeh Jubran
---
qxldod/QxlDod.cpp | 13 -
1 file changed, 12 insertions(+), 1 deletion(-)
diff --git a
This call to BlackOutScreen is not needed since the actual
display surface is destroyed and created (redrawn) again.
Acked-by: Frediano Ziglio
Signed-off-by: Sameeh Jubran
---
qxldod/QxlDod.cpp | 2 --
1 file changed, 2 deletions(-)
diff --git a/qxldod/QxlDod.cpp b/qxldod/QxlDod.cpp
index b6c
Interrupts seem to arrive to the driver before the initialization phase
is over (m_pHWDevice = NULL), in that case we can't handle interrupts yet. Even
when m_pHWDevice isn't NULL, other fields aren't necessarly fully intialized
till
the StartDevice function has finished initialization, thus the f
Part of source image mapped by PresentDisplayOnly
should be big enough to cover all rectangles being
transferred.
Signed-off-by: Sameeh Jubran
Signed-off-by: Dmitry Fleytman
---
qxldod/QxlDod.cpp | 28 +++-
qxldod/QxlDod.h | 2 ++
2 files changed, 25 insertions(+), 5
This series contains the latest patches to support Windows 10.
The patches were applied on the up-to-date master branch of
https://gitlab.com/spice/qxl-wddm-dod
Diffrences from v4:
* Edited the patch "Use MmMapIoSpaceEx instead of MmMapIoSpace" to
support Windows 8 as discussed in here:
https://p
>
> ---
> enums.js| 11 +++-
> main.js | 5 +++-
> port.js | 85
> +
> spice.html | 50 +++--
> spice_auto.html | 12 +++-
> spicemsg.js | 24 ++--
> utils.js
On Mon, 2016-09-26 at 12:27 +0100, Frediano Ziglio wrote:
> Signed-off-by: Frediano Ziglio
Acked-by: Pavel Grunt
> ---
> server/red-channel.c | 10 --
> server/red-channel.h | 5 -
> 2 files changed, 4 insertions(+), 11 deletions(-)
>
> Changes since v1:
> - make test_capability fu
Signed-off-by: Frediano Ziglio
---
server/red-channel.c | 10 --
server/red-channel.h | 5 -
2 files changed, 4 insertions(+), 11 deletions(-)
Changes since v1:
- make test_capability function inline
diff --git a/server/red-channel.c b/server/red-channel.c
index 927e6e9..af49824 10
I finally put all in the same patch with the Textdecoder part removed. Just
added an utility function to do that in utils.js
On Mon, Sep 26, 2016 at 1:02 PM, Oliver Gutierrez
wrote:
> ---
> enums.js| 11 +++-
> main.js | 5 +++-
> port.js | 85 +
---
enums.js| 11 +++-
main.js | 5 +++-
port.js | 85 +
spice.html | 50 +++--
spice_auto.html | 12 +++-
spicemsg.js | 24 ++--
utils.js| 7 +
7
On Mon, 2016-09-26 at 09:12 +0100, Frediano Ziglio wrote:
> Move large buffer field at the end of structure.
> Due to the way machine address memory this usually can reduce code
> size
> and make program sligthly faster.
> Actually reduce size by 100 bytes.
>
> Signed-off-by: Frediano Ziglio
Acke
On Mon, 2016-09-26 at 05:18 -0400, Frediano Ziglio wrote:
> >
> > Hi,
> >
> > does it make a sense to keep the function (or is it to avoid the
> > header inclusion)? What about using the macro directly ?
> >
> > Pavel
> >
>
>
> The macro could be misleading as it refers to the agent.
> Also m
>
> Hi,
>
> does it make a sense to keep the function (or is it to avoid the
> header inclusion)? What about using the macro directly ?
>
> Pavel
>
The macro could be misleading as it refers to the agent.
Also macro could have side effects and usually are less type safe..
But a simple inline
On Mon, 2016-09-26 at 09:12 +0100, Frediano Ziglio wrote:
> There's no reason for so hard optimisations so avoid
> having to maintain multiple labels.
>
> Signed-off-by: Frediano Ziglio
Acked-by: Pavel Grunt
> ---
> tools/reds_stat.c | 20 ++--
> 1 file changed, 10 insertions(+)
Yes. Of course. Will take a look at this and change it. I will resend the
patch as soon as I fix that.
Thanks
On Thu, Sep 22, 2016 at 7:28 PM, Jeremy White
wrote:
> Hey Oliver,
>
> Thanks for this patch; it broadly looks good to me, with one question:
>
> On 09/20/2016 09:30 AM, Oliver Gutierre
On Mon, 2016-09-26 at 09:12 +0100, Frediano Ziglio wrote:
> If channel is not connected red_channel_pipes_new_add_push will
> do nothing.
red_channel_pipes_new_add_push() does it through "foreach" over its
clients. I would prefer to keep the explicit check, imho it is more
readable.
Pavel
>
> Si
Hi,
does it make a sense to keep the function (or is it to avoid the
header inclusion)? What about using the macro directly ?
Pavel
On Mon, 2016-09-26 at 09:12 +0100, Frediano Ziglio wrote:
> Signed-off-by: Frediano Ziglio
> ---
> server/red-channel.c | 7 +--
> 1 file changed, 1 insertion
Hi,
Maybe it is clear from the commit message, but I would prefer an
explicit comment saying that in case of invalid value the original
compression is unchanged.
On Mon, 2016-09-26 at 09:12 +0100, Frediano Ziglio wrote:
> Signed-off-by: Frediano Ziglio
Acked-by: Pavel Grunt
Thanks,
Pavel
> ---
Signed-off-by: Frediano Ziglio
---
tools/reds_stat.c | 13 -
1 file changed, 8 insertions(+), 5 deletions(-)
diff --git a/tools/reds_stat.c b/tools/reds_stat.c
index 5e9705c..07fd732 100644
--- a/tools/reds_stat.c
+++ b/tools/reds_stat.c
@@ -24,6 +24,7 @@
#include
#include
#incl
Signed-off-by: Frediano Ziglio
---
server/stream.c | 3 +++
1 file changed, 3 insertions(+)
diff --git a/server/stream.c b/server/stream.c
index 9d878c8..95ff015 100644
--- a/server/stream.c
+++ b/server/stream.c
@@ -320,6 +320,9 @@ static void attach_stream(DisplayChannel *display, Drawable
*d
Signed-off-by: Frediano Ziglio
---
server/red-worker.c | 6 --
1 file changed, 4 insertions(+), 2 deletions(-)
diff --git a/server/red-worker.c b/server/red-worker.c
index e55a939..7c7eafe 100644
--- a/server/red-worker.c
+++ b/server/red-worker.c
@@ -956,9 +956,9 @@ static void handle_dev_s
This could have happened if mremap failed.
Signed-off-by: Frediano Ziglio
---
tools/reds_stat.c | 8
1 file changed, 4 insertions(+), 4 deletions(-)
diff --git a/tools/reds_stat.c b/tools/reds_stat.c
index 945659e..d7d82ec 100644
--- a/tools/reds_stat.c
+++ b/tools/reds_stat.c
@@ -66,7
vdi_port_read_buf_release is registered passing data as
RedVDIReadBuf*, not RedPipeItem*. Cast opaque to proper
pointer type to avoid the assumption that first field of
RedVDIReadBuf is a RedPipeItem.
Signed-off-by: Frediano Ziglio
---
server/reds.c | 3 ++-
1 file changed, 2 insertions(+), 1 de
Signed-off-by: Frediano Ziglio
---
server/memslot.c | 14 ++
server/memslot.h | 3 +++
server/red-parse-qxl.c | 11 +++
server/red-parse-qxl.h | 1 +
server/red-worker.c| 3 +--
5 files changed, 30 insertions(+), 2 deletions(-)
diff --git a/server/memslot.c
If channel is not connected red_channel_pipes_new_add_push will
do nothing.
Signed-off-by: Frediano Ziglio
---
server/inputs-channel.c | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git a/server/inputs-channel.c b/server/inputs-channel.c
index 840d5e9..e3a940c 100644
--- a/serve
Signed-off-by: Frediano Ziglio
---
server/red-channel.c | 7 +--
1 file changed, 1 insertion(+), 6 deletions(-)
diff --git a/server/red-channel.c b/server/red-channel.c
index 927e6e9..27c8632 100644
--- a/server/red-channel.c
+++ b/server/red-channel.c
@@ -331,12 +331,7 @@ void red_channel_r
Signed-off-by: Frediano Ziglio
---
server/cursor-channel.c | 4
1 file changed, 4 deletions(-)
diff --git a/server/cursor-channel.c b/server/cursor-channel.c
index 05b8942..3e1e9f2 100644
--- a/server/cursor-channel.c
+++ b/server/cursor-channel.c
@@ -55,10 +55,6 @@ struct CursorChannelPriv
Move large buffer field at the end of structure.
Due to the way machine address memory this usually can reduce code size
and make program sligthly faster.
Actually reduce size by 100 bytes.
Signed-off-by: Frediano Ziglio
---
server/image-encoders.h | 3 ++-
1 file changed, 2 insertions(+), 1 del
Attempt to use consistent naming.
Usually we use surface name for RedSurface.
Signed-off-by: Frediano Ziglio
---
server/display-channel.c | 12 ++--
1 file changed, 6 insertions(+), 6 deletions(-)
diff --git a/server/display-channel.c b/server/display-channel.c
index ad9dde6..217f5eb 10
Signed-off-by: Frediano Ziglio
---
server/main-channel-client.c | 3 ---
server/reds-private.h| 1 -
server/reds.c| 13 -
server/stat.h| 1 -
4 files changed, 18 deletions(-)
Not to mention that the computation was wrong!
diff --git a/serve
Signed-off-by: Frediano Ziglio
---
Makefile.am | 2 +-
configure.ac | 1 +
tools/Makefile.am | 19 +++
3 files changed, 21 insertions(+), 1 deletion(-)
create mode 100644 tools/Makefile.am
diff --git a/Makefile.am b/Makefile.am
index 7eb0f28..5272d5a 100644
--- a/Ma
These functions were implementing the same stuff as empty
messages functions provided by RedChannel so reuse them.
The implementation seems a bit different but the result
is the same. Specifically:
- RedEmptyMsgPipeItem::msg is int while RedVerbItem::verb was
uint16_t however this data goes into
Attempt to use consistent naming.
Usually we use surface name for RedSurface so make sure
code reader do not get confused using a different name
for RedSurfaceCmd.
Signed-off-by: Frediano Ziglio
---
server/display-channel.c | 23 ---
server/display-channel.h | 6 +++---
2 fi
Trying to collect different stuff not much related.
The most related ones are the compilation and updates
of reds_stat utility.
Frediano Ziglio (18):
Remove red_pipe_add_verb family function
Prevent setting invalid image compression values
Add small comment for detach_stream
Reuse VD_AGENT
There's no reason for so hard optimisations so avoid
having to maintain multiple labels.
Signed-off-by: Frediano Ziglio
---
tools/reds_stat.c | 20 ++--
1 file changed, 10 insertions(+), 10 deletions(-)
diff --git a/tools/reds_stat.c b/tools/reds_stat.c
index 07fd732..945659e 10
Signed-off-by: Frediano Ziglio
---
server/spice-bitmap-utils.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/server/spice-bitmap-utils.c b/server/spice-bitmap-utils.c
index a423ad9..72a9285 100644
--- a/server/spice-bitmap-utils.c
+++ b/server/spice-bitmap-utils.c
@@ -132,7
Signed-off-by: Frediano Ziglio
---
server/red-channel-client-private.h | 1 +
1 file changed, 1 insertion(+)
diff --git a/server/red-channel-client-private.h
b/server/red-channel-client-private.h
index 46d4fca..804b39a 100644
--- a/server/red-channel-client-private.h
+++ b/server/red-channel-cl
48 matches
Mail list logo
