Summary:
display: use WAIT_FOR_EVENT everywhere
- we already had this macro, it's a straight replace.
display: fix broken debug in checked
- we had ASSERT's with undefined variables
miniport: add QXL_IO_LOG
- this is relatively big - adds an snprintf implementation to allow
using our QXL_IO_LOG hack to debug the driver also for miniport, not just
for the video driver bit. (see MSDN if this doesn't make sense).
display/qxldd.h: add DEBUG_PRINT_ONCE
- used for patch after next for more verbose prints on DrvCopyBits and
DrvBitBlt
display/res: turn two asserts in CursorCacheRemove to debug prints
- fixes checked build (would assert). The real cause is unknown.
display/surface.h: make GetSurfaceId NULL safe
- wasn't actually a problem, but seems safer.
display/rop.c: more verbose debugging (for guestdebug>=3)
- what it says
CreateDeviceBitmap: more verbose debug prints on error paths
- what it says
Please review.
Alon Levy (8):
display: use WAIT_FOR_EVENT everywhere
display: fix broken debug in checked
miniport: add QXL_IO_LOG
display/qxldd.h: add DEBUG_PRINT_ONCE
display/res: turn two asserts in CursorCacheRemove to debug prints
display/surface.h: make GetSurfaceId NULL safe
display/rop.c: more verbose debugging (for guestdebug>=3)
CreateDeviceBitmap: more verbose debug prints on error paths
display/driver.c | 6 ++-
display/quic.c | 4 +
display/qxldd.h | 9 +++
display/res.c | 51 ++++-----------
display/rop.c | 21 ++++++-
display/surface.c | 11 ++-
display/surface.h | 3 +
miniport/minimal_snprintf.c | 150 +++++++++++++++++++++++++++++++++++++++++++
miniport/qxl.c | 55 +++++++++++++++-
miniport/qxl.h | 3 +-
miniport/sources | 1 +
11 files changed, 268 insertions(+), 46 deletions(-)
create mode 100644 miniport/minimal_snprintf.c
--
1.7.4.2
_______________________________________________
Spice-devel mailing list
Spice-devel@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/spice-devel
