Backends: SDL_Renderer: Fix texture atlas format on big-endian hardware (#4927)

SDL_PIXELFORMAT_RGBA32 is intended for byte arrays where the channels
are specifically in the RGBA order. However, this is not what
GetTexDataAsRGBA32 does: rather, it constructs an array where each
pixel is an unsigned int in ABGR order. On little-endian platforms,
this does indeed result in an RGBA byte order, however, on big-endian
platforms, this results in an ABGR byte order, which does not match
the PIXELFORMAT enum.

What should be used is the SDL_PIXELFORMAT_ABGR8888 enum, which
specifies that the pixels are in native-endian ABGR, which they are.

backends/imgui_impl_sdlrenderer.cpp

@@ -210,7 +210,7 @@ bool ImGui_ImplSDLRenderer_CreateFontsTexture()
     io.Fonts->GetTexDataAsRGBA32(&pixels, &width, &height);   // Load as RGBA 32-bit (75% of the memory is wasted, but default font is so small) because it is more likely to be compatible with user's existing shaders. If your ImTextureId represent a higher-level concept than just a GL texture id, consider calling GetTexDataAsAlpha8() instead to save on GPU memory.
 
     // Upload texture to graphics system
-    bd->FontTexture = SDL_CreateTexture(bd->SDLRenderer, SDL_PIXELFORMAT_RGBA32, SDL_TEXTUREACCESS_STATIC, width, height);
+    bd->FontTexture = SDL_CreateTexture(bd->SDLRenderer, SDL_PIXELFORMAT_ABGR8888, SDL_TEXTUREACCESS_STATIC, width, height);
     if (bd->FontTexture == NULL)
     {
         SDL_Log("error creating texture");

docs/CHANGELOG.txt

@@ -160,6 +160,7 @@ Other Changes:
 - Backends: Metal: Ignore ImDrawCmd where ElemCount == 0, which are normally not emitted by the library but
   can theorically be created by user code manipulating a ImDrawList. (#4857)
 - Backends: Vulkan: Added support for ImTextureID as VkDescriptorSet, add ImGui_ImplVulkan_AddTexture(). (#914) [@martty]
+- Backends: SDL_Renderer: Fix texture atlas format on big-endian hardware (#4927) [@Clownacy]
 - Backends: WebGPU: Fixed incorrect size parameters in wgpuRenderPassEncoderSetIndexBuffer() and
   wgpuRenderPassEncoderSetVertexBuffer() calls. (#4891) [@FeepsDev]