EngSaveFloatingPointState function (winddi.h)

The EngSaveFloatingPointState function saves the current Windows 2000 (and later) kernel floating-point state.

Syntax

ULONG EngSaveFloatingPointState(
            VOID  *pBuffer,
  [in, out] ULONG cjBufferSize
);

Parameters

pBuffer

Pointer to the buffer that receives the floating-point state. This buffer must be zero-initialized, and must be in nonpaged memory.

[in, out] cjBufferSize

Specifies the size, in bytes, of the buffer to which pBuffer points.

Return value

If pBuffer is non-NULL, EngSaveFloatingPointState returns TRUE if the state is successfully saved. It returns FALSE if the specified buffer is too small or the state cannot be saved.

If pBuffer is NULL or cjBufferSize is zero, EngSaveFloatingPointState returns the size of the buffer needed to save the floating-point state. If the return value is zero, the processor does not have hardware floating-point capability. In this case, the driver must not use any floating-point instructions.

Remarks

The driver must save the current kernel floating-point state before using floating-point hardware instructions. On Intel architecture systems, this permits the use of MMX instructions if they are supported by the processor. Drivers that do not properly use EngSaveFloatingPointState and EngRestoreFloatingPointState when using floating-point or MMX hardware will cause random floating-point or MMX corruption in the calling application.

On every call to the driver, the driver must call EngSaveFloatingPointState once to preserve kernel state before using floating-point or MMX operations. It must also call EngRestoreFloatingPointState once after all floating-point or MMX operations are complete to reset the kernel state.

GDI automatically saves the floating-point state for any calls to a driver's DrvEscape routine when the escape is OPENGL_CMD, OPENGL_GETINFO, or MCDFUNCS.

Requirements

See also

DrvEscape

EngRestoreFloatingPointState