Claude review: video: fbdev: omapfb: Add missing error check for clk_get()

[Claude Code Review Bot <claude-review@example.com>]

Patch Review

**The IS_ERR check is correct and needed.** The `clk_get()` call at line 952 is indeed unchecked, and adding the check is the right thing to do:

```c
hwa742.sys_ck = clk_get(NULL, "hwa_sys_ck");
+	if (IS_ERR(hwa742.sys_ck))
+		return PTR_ERR(hwa742.sys_ck);
```

This is fine — at this point in `hwa742_init()`, no resources have been acquired yet (the `int_ctrl->init()` and `extif->init()` calls happen later), so returning directly is correct rather than going through the error unwind labels.

**The `clk_put()` in `err1` is problematic.** Looking at the error paths:

```c
err1:
+	clk_put(hwa742.sys_ck);
	return r;
```

The `err1` label is reached from line 956-957:
```c
if ((r = hwa742.int_ctrl->init(fbdev, 1, req_vram)) < 0)
    goto err1;
```

And also falls through from `err2`:
```c
err2:
	hwa742.int_ctrl->cleanup();
err1:
+	clk_put(hwa742.sys_ck);
	return r;
```

Since `clk_get()` succeeded if we reach any of these labels (the IS_ERR check returns early), calling `clk_put()` at `err1` is correct. Similarly, `err3` and `err4` flow through `err2` → `err1`, so all error paths will now properly release the clock. This is actually fine.

**The `clk_put()` in `hwa742_cleanup()` is correct:**

```c
	clk_disable_unprepare(hwa742.sys_ck);
+	clk_put(hwa742.sys_ck);
```

The clock needs to be released on driver cleanup. The ordering (disable/unprepare first, then put) is correct.

**Minor observation:** This driver uses the legacy `clk_get(NULL, ...)` / `clk_put()` API rather than `devm_clk_get()`. Converting to the managed API would be a further improvement but is outside the scope of this fix and would be a separate cleanup patch.

**Overall:** The patch is correct and addresses a real bug. The error handling and cleanup are properly structured. 

Reviewed-by looks appropriate for this patch.

---
Generated by Claude Code Patch Reviewer