[PATCH v2 0/1] MR2892: wined3d: Fix uninitialized variable warning.

[Davide Beatrici (＠davidebeatrici)]

```c
dlls/wined3d/context_vk.c:2377:42: warning: ‘null_binding’ may be used uninitialized in this function [-Wmaybe-uninitialized]
 2377 |                     *null_buffer_binding = b->binding = null_binding;
      |                     ~~~~~~~~~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~~
```

--
  v2: wined3d: Fix uninitialized variable warning.

https://gitlab.winehq.org/wine/wine/-/merge_requests/2892

[PATCH v2 1/1] wined3d: Fix uninitialized variable warning.

[Davide Beatrici]

From: Davide Beatrici

dlls/wined3d/context_vk.c:2377:42: warning: ‘null_binding’ may be used uninitialized in this function [-Wmaybe-uninitialized]
 2377 |                     *null_buffer_binding = b->binding = null_binding;
      |                     ~~~~~~~~~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~~
---
 dlls/wined3d/context_vk.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/dlls/wined3d/context_vk.c b/dlls/wined3d/context_vk.c
index 10713985e7e..8076ac54166 100644
--- a/dlls/wined3d/context_vk.c
+++ b/dlls/wined3d/context_vk.c
@@ -2337,7 +2337,7 @@ static bool wined3d_context_vk_update_graphics_pipeline_key(struct wined3d_conte
         {
             struct wined3d_shader_signature_element *element;
             struct wined3d_shader_signature *signature;
-            uint32_t null_binding, location;
+            uint32_t null_binding = UINT32_MAX, location;

for (i = 0; i < ARRAY_SIZE(state->streams); ++i)
             {

[Zebediah Figura (＠zfigura)]

With what gcc version and compiler flags does this appear?

[Davide Beatrici (＠davidebeatrici)]

On Wed May 24 17:50:18 2023 +0000, Zebediah Figura wrote:
> With what gcc version and compiler flags does this appear?

```
gcc-9 (Ubuntu 9.4.0-1ubuntu1~18.04) 9.4.0
x86_64-w64-mingw32-gcc (GCC) 9.3-win32 20200422
```

```
export CFLAGS_X32="-march=i686 -msse2 -mfpmath=sse -Og -ftree-vectorize"
export CFLAGS_X64="-march=x86-64 -msse3 -mfpmath=sse -Og -ftree-vectorize"
export LDFLAGS="-Wl,-O1,--sort-common,--as-needed"
```

[Zebediah Figura (＠zfigura)]

On Wed May 24 17:52:33 2023 +0000, Davide Beatrici wrote:
> gcc-9 (Ubuntu 9.4.0-1ubuntu1~18.04) 9.4.0
> x86_64-w64-mingw32-gcc (GCC) 9.3-win32 20200422
> export CFLAGS_X32="-march=i686 -msse2 -mfpmath=sse -Og -ftree-vectorize"
> export CFLAGS_X64="-march=x86-64 -msse3 -mfpmath=sse -Og -ftree-vectorize"
> export LDFLAGS="-Wl,-O1,--sort-common,--as-needed"

That's both old and nonstandard, and while I'm not going to argue too hard against fixing these warnings, it doesn't seem the best use of time. I have to wonder how many other such warnings exist across the codebase with those options...

Anyway, suppressing -Wmaybe-uninitialized by initializing the variable to a dummy value is bad for clarity, and in my experience almost never the right solution. In this case I'd rather replace the for + if pattern (which I pretty much universally dislike) with a helper function.

[Davide Beatrici (＠davidebeatrici)]

> That's both old and nonstandard, and while I'm not going to argue too hard against fixing these warnings, it doesn't seem the best use of time. I have to wonder how many other such warnings exist across the codebase with those options...

Quite a few, not too many though.
> Anyway, suppressing -Wmaybe-uninitialized by initializing the variable to a dummy value is bad for clarity, and in my experience almost never the right solution. In this case I'd rather replace the for + if pattern (which I pretty much universally dislike) with a helper function.

I strongly agree, I honestly didn't like this "solution" anyway. !2863 solved the warning the right way.