Prepare for a more dynamic #pragma unroll use

This commit is contained in:
jsteube
2016-05-09 21:32:12 +02:00
parent 34c3557d50
commit c79bed3b7d
129 changed files with 776 additions and 140 deletions
+162 -7
View File
@@ -3,15 +3,30 @@
* License.....: MIT
*/
#ifdef cl_khr_byte_addressable_store
#pragma OPENCL EXTENSION cl_khr_int64_base_atomics : enable
#pragma OPENCL EXTENSION cl_khr_byte_addressable_store : enable
/**
* device type
*/
#define DEVICE_TYPE_CPU 2
#define DEVICE_TYPE_GPU 4
#define DEVICE_TYPE_ACCEL 8
#if DEVICE_TYPE == DEVICE_TYPE_CPU
#define IS_CPU
#elif DEVICE_TYPE == DEVICE_TYPE_GPU
#define IS_GPU
#elif DEVICE_TYPE == DEVICE_TYPE_ACCEL
#define IS_ACCEL
#endif
/**
* vendor specific
*/
#if VENDOR_ID == (1 << 0)
#if VENDOR_ID == (1 << 0)
#define IS_AMD
#elif VENDOR_ID == (1 << 6)
#define IS_NV
@@ -29,15 +44,155 @@
#endif
/**
* NV specific
* Unrolling is generally enabled, for all device types and hash modes
* There's a few exception when it's better not to unroll
*/
// Some algorithms run into too much register pressure due to loop unrolling
#ifdef IS_NV
#ifdef IS_GPU
#if KERN_TYPE == 1500
#undef _unroll
#endif
#if KERN_TYPE == 1800
#undef _unroll
#endif
#if KERN_TYPE == 3000
#undef _unroll
#endif
#if KERN_TYPE == 6221
#undef _unroll
#endif
#if KERN_TYPE == 6222
#undef _unroll
#endif
#if KERN_TYPE == 6223
#undef _unroll
#endif
#if KERN_TYPE == 6500
#undef _unroll
#endif
#if KERN_TYPE == 7100
#undef _unroll
#endif
#if KERN_TYPE == 7400
#undef _unroll
#endif
#if KERN_TYPE == 8200
#undef _unroll
#endif
#if KERN_TYPE == 10400
#undef _unroll
#endif
#if KERN_TYPE == 10500
#undef _unroll
#endif
#if KERN_TYPE == 10700
#undef _unroll
#endif
#if KERN_TYPE == 12300
#undef _unroll
#endif
#if KERN_TYPE == 12400
#undef _unroll
#endif
/**
* Generic
*/
#endif
#endif
#ifdef IS_GENERIC
#ifdef IS_AMD
#ifdef IS_GPU
#if KERN_TYPE == 3200
#undef _unroll
#endif
#if KERN_TYPE == 5200
#undef _unroll
#endif
#if KERN_TYPE == 6100
#undef _unroll
#endif
#if KERN_TYPE == 6221
#undef _unroll
#endif
#if KERN_TYPE == 6222
#undef _unroll
#endif
#if KERN_TYPE == 6223
#undef _unroll
#endif
#if KERN_TYPE == 6400
#undef _unroll
#endif
#if KERN_TYPE == 6500
#undef _unroll
#endif
#if KERN_TYPE == 6800
#undef _unroll
#endif
#if KERN_TYPE == 7100
#undef _unroll
#endif
#if KERN_TYPE == 7400
#undef _unroll
#endif
#if KERN_TYPE == 8000
#undef _unroll
#endif
#if KERN_TYPE == 8200
#undef _unroll
#endif
#if KERN_TYPE == 10900
#undef _unroll
#endif
#if KERN_TYPE == 11600
#undef _unroll
#endif
#if KERN_TYPE == 12300
#undef _unroll
#endif
#if KERN_TYPE == 12800
#undef _unroll
#endif
#if KERN_TYPE == 12900
#undef _unroll
#endif
#if KERN_TYPE == 13000
#undef _unroll
#endif
#endif
#endif
// Some algorithms break due to loop unrolling, it's unknown why, probably compiler bugs
// Can overlap with above cases
#ifdef IS_AMD
#ifdef IS_GPU
#if KERN_TYPE == 1750
#undef _unroll
#endif
#if KERN_TYPE == 1760
#undef _unroll
#endif
#if KERN_TYPE == 6500
#undef _unroll
#endif
#if KERN_TYPE == 7100
#undef _unroll
#endif
#if KERN_TYPE == 9600
#undef _unroll
#endif
#if KERN_TYPE == 12200
#undef _unroll
#endif
#if KERN_TYPE == 12300
#undef _unroll
#endif
#endif
#endif