mirror of
https://github.com/reactos/reactos.git
synced 2024-09-10 21:04:26 +00:00
Add notes, no code change.
svn path=/trunk/; revision=33956
This commit is contained in:
parent
86853e2cea
commit
f7a11147ec
|
@ -1432,7 +1432,8 @@ GDI_MapHandleTable(PSECTION_OBJECT SectionObject, PEPROCESS Process)
|
||||||
(PGDIBRUSHOBJ)->pBrushAttr = IntGdiAllocObjAttr(GDIObjType_BRUSH_TYPE);
|
(PGDIBRUSHOBJ)->pBrushAttr = IntGdiAllocObjAttr(GDIObjType_BRUSH_TYPE);
|
||||||
|
|
||||||
// Kernel Space to User Space Pointer
|
// Kernel Space to User Space Pointer
|
||||||
(PGDI_TABLE_ENTRY)->UserData = UserHeapAddressToUser(pBrushAttr);
|
(PGDI_TABLE_ENTRY)->UserData = pBrushAttr;
|
||||||
|
// Gdi will adjust for heap delta.
|
||||||
|
|
||||||
Example Freeing:
|
Example Freeing:
|
||||||
|
|
||||||
|
@ -1441,11 +1442,9 @@ GDI_MapHandleTable(PSECTION_OBJECT SectionObject, PEPROCESS Process)
|
||||||
(PGDIBRUSHOBJ)->pBrushAttr = NULL;
|
(PGDIBRUSHOBJ)->pBrushAttr = NULL;
|
||||||
|
|
||||||
Notes:
|
Notes:
|
||||||
Testing with DC_ATTR works but has drawing difficulties. This could be due
|
Testing with DC_ATTR works but has drawing difficulties.
|
||||||
to a bug in user share heap allocation and deallocation. The test did not
|
Base on observation, (Over looking the obvious) we need to supply heap delta
|
||||||
include additional deallocations until the DC was freed. So it was always
|
to user space gdi. Now, with testing, looks all normal.
|
||||||
allocated to the DC. Space between allocations was not even and at less
|
|
||||||
than sizeof(DC_ATTR).
|
|
||||||
|
|
||||||
*/
|
*/
|
||||||
PVOID
|
PVOID
|
||||||
|
|
Loading…
Reference in a new issue