Instead of current abort level default which is critical level.
The spice_return_if_fail and spice_return_val_if_fail macros
log with critical level upon failure. This currently aborts
the program.
Instead it's better to log an error message and let the application
continue running.
diff --git a/
diff --git a/common/gdi_canvas.c b/common/gdi_canvas.c
index deb7649..05edf68 100644
--- a/common/gdi_canvas.c
+++ b/common/gdi_canvas.c
@@ -354,6 +354,7 @@ static void set_path(GdiCanvas *canvas, SpicePath *s)
if (!LineTo(canvas->dc, (int)fix_to_double(point->x),
Instead of critical-level.
This default can be overwritten.
Note that I left all spice_critical code as is.
This was the intention of the original commit using spice_critical.
I could have replaced all spice_critical with spice_error, and later
change per-case to spice_critical.
How do people feel
which vdagent version are you using? which windows version?
please send %windir%\temp\vdagemt.log & vdservice.log.
I guess it's something to do with the different sessions used by spice &
rdp.
flooding Controlled wrote:
I do not know how you can use spice and RDP together, can you explain you
On 12/13/2012 03:54 PM, Uri Lublin wrote:
replaced with spice_error where abort is wanted.
Note, that I did not change all spice_critical to spice_error.
For example for lz/quic/glz _usr_error functions I kept it as
spice_critical.
(I'm not sure about that move as there is a comment in
glz_u
replaced with spice_error where abort is wanted.
---
server/red_worker.c | 26 +-
1 files changed, 13 insertions(+), 13 deletions(-)
diff --git a/server/red_worker.c b/server/red_worker.c
index 530562b..1ac439b 100644
--- a/server/red_worker.c
+++ b/server/red_worker.c
@
On 12/13/2012 06:42 AM, Alon Levy wrote:
All looks good except the assert. We should be removing them whenever they are
guest trigerable - maybe I'm not following the code, but if in red_get_image
sees no pallete in the qxl struct then palette will be NULL, which means it's
guest trigerable.
All looks good except the assert. We should be removing them whenever they are
guest trigerable - maybe I'm not following the code, but if in red_get_image
sees no pallete in the qxl struct then palette will be NULL, which means it's
guest trigerable.
> Non rgb bitmaps are allowed to not have a
