wxMSW: A better (IMHO) RTL fixes for wx{Auto}BufferedDC (PR #26398)

AliKet

unread,

Apr 19, 2026, 8:51:03 PM (4 days ago) Apr 19

to wx-...@googlegroups.com, Subscribed

Also fix custom drawing where wxDC::DeviceToLogical() and related functions need to be used.

You can view, comment on, or merge this pull request online at:

https://github.com/wxWidgets/wxWidgets/pull/26398

Commit Summary

2e0d08c wxDC::LogicalToDevice{Rel}() now always return unmirrored coordinates under wxMSW
5827b4c Remove unnecessary hack from wxMSWDCImpl::DoSetClippingRegion()
ada3d84 Really fix using wx{Auto}BufferedDC in RTL layout under wxMSW
af13c88 Fix rendering GridFrame (from griddemo) in RTL layout
dadf3ab Update OneDevRegionRTL test after recent changes

File Changes

(3 files)

Patch Links:

—
Reply to this email directly, view it on GitHub, or unsubscribe.
You are receiving this because you are subscribed to this thread.

AliKet

unread,

Apr 19, 2026, 9:08:20 PM (4 days ago) Apr 19

to wx-...@googlegroups.com, Push

@AliKet pushed 1 commit.

da9c92e Try to fix OneDevRegionRTL test failures

—
View it on GitHub or unsubscribe.
You are receiving this because you are subscribed to this thread.

AliKet

unread,

Apr 19, 2026, 9:31:50 PM (4 days ago) Apr 19

to wx-...@googlegroups.com, Subscribed

AliKet left a comment (wxWidgets/wxWidgets#26398)

I don't have a real explanation of why we need this commit da9c92e to fix tests failures in OneDevRegionRTL, Just some weired rounding errors? I'm guessing...

—
Reply to this email directly, view it on GitHub, or unsubscribe.

You are receiving this because you are subscribed to this thread.

AliKet

unread,

Apr 19, 2026, 9:49:20 PM (4 days ago) Apr 19

to wx-...@googlegroups.com, Subscribed

AliKet left a comment (wxWidgets/wxWidgets#26398)

Micro-optimisation:

diff --git a/src/msw/dc.cpp b/src/msw/dc.cpp
index dc56fa3e53..dda510a168 100644
--- a/src/msw/dc.cpp
+++ b/src/msw/dc.cpp
@@ -2379,10 +2379,12 @@ bool wxMSWDCImpl::DoStretchBlit(wxCoord xdest, wxCoord ydest,
                 // automatically as it doesn't even work with the source HDC.
                 // So do this manually to ensure that the coordinates are
                 // interpreted in the same way here as in all the other cases.
-                xsrc = source->LogicalToDeviceX(xsrcOrig);
-                ysrc = source->LogicalToDeviceY(ysrcOrig);
-                srcWidth = source->LogicalToDeviceXRel(srcWidth);
-                srcHeight = source->LogicalToDeviceYRel(srcHeight);
+                const auto devPos = implSrc->LogicalToDevice(xsrcOrig, ysrcOrig);
+                const auto devSize = implSrc->LogicalToDeviceRel(srcWidth, srcHeight);
+                xsrc = devPos.x;
+                ysrc = devPos.y;
+                srcWidth = devSize.x;
+                srcHeight = devSize.y;
 
                 // Figure out what co-ordinate system we're supposed to specify
                 // ysrc in.

LogicalToDevice{X,Y}[Rel]() now internally call SetLayout() to force LTR before calling LPtoDP() as a pre-condition and restore the old layout when the function returns as a post-condition. So there will be 8 calls to SetLayout() with the old code, but 4 with the patch above. I didn't try to profile this to see whether the patch makes any difference, but I still prefer to apply it, though.

—
Reply to this email directly, view it on GitHub, or unsubscribe.

You are receiving this because you are subscribed to this thread.

AliKet

unread,

Apr 19, 2026, 9:54:39 PM (4 days ago) Apr 19

to wx-...@googlegroups.com, Subscribed

AliKet left a comment (wxWidgets/wxWidgets#26398)

@vslavik I would be grateful if you could test this and see whether it breaks something I’m not aware of. Thanks in advance!