amd/amdkfd/kfd_flat_memory.c

49  * mapped into this pointer space as apertures.
51  * based on what apertures the request falls in.
112  * A 64b pointer is compared to the apertures that are defined (Base/Limit), in
127  * memory object cannot straddle multiple apertures in the 64b address
151  * apertures to set the Mtype field for S_LOAD/FLAT_* ops which is input to the
155  * address space that doesn’t hit any defined base/limit apertures.
156  * The following diagram is a complete picture of the gfxip7.x SUA apertures.
165  * In all cases (for SUA and DUA apertures discussed later), aperture base
175  * In no case is a payload that straddles multiple apertures expected to work.
181  * or a payload straddles multiple apertures.  The S_LOAD instruction
184  * Apertures cannot overlap.
210  * Private memories are mapped as apertures into the address space,
212  * to the LDS or Scratch memory when it falls into one of these apertures.
213  * Like the SUA apertures, the Shared/Private apertures are 64KB aligned and
215  * the Shared/Private apertures are always placed in a limited selection of
217  * Shared/Private apertures can be placed anywhere in the 32b space
221  * HSA64 Apertures for FLAT_* vector ops
223  * For HSA64 SUA mode, the Shared and Private apertures are always placed
232  * HSA32 Apertures for FLAT_* vector ops
234  * In HSA32 mode, the Private and Shared apertures can be placed anywhere
236  * the apertures). If the base address of the apertures are non-zero
237  * (ie apertures exists), the size is always 64KB.
240  * GPUVM Apertures for FLAT_* vector ops
242  * In GPUVM mode, the Shared/Private apertures are specified identically
295 /* On GFXv9 the LDS and scratch apertures are programmed independently
388 		 * For 64 bit process apertures will be statically reserved in  in kfd_init_apertures()
390 		 * amdkfd doesn't currently support apertures for 32 bit process  in kfd_init_apertures()