Possibility to specify user-defined default cache sizes for GEBP kernel

Some architectures have no convinient way to determine cache sizes at
runtime. Eigen's GEBP kernel falls back to default cache values in this
case which might not be correct in all situations.

This patch introduces three preprocessor directives

  `EIGEN_DEFAULT_L1_CACHE_SIZE`
  `EIGEN_DEFAULT_L2_CACHE_SIZE`
  `EIGEN_DEFAULT_L3_CACHE_SIZE`

to give users the possibility to set these default values explicitly.

2 files changed, 36 insertions, 13 deletions

diff --git a/Eigen/src/Core/products/GeneralBlockPanelKernel.h b/Eigen/src/Core/products/GeneralBlockPanelKernel.h
index 4b6a3ee6f..64e7f79cf 100644
--- a/Eigen/src/Core/products/GeneralBlockPanelKernel.h
+++ b/Eigen/src/Core/products/GeneralBlockPanelKernel.h
@@ -31,20 +31,42 @@ inline std::ptrdiff_t manage_caching_sizes_helper(std::ptrdiff_t a, std::ptrdiff
   return a<=0 ? b : a;
 }
 
+#if defined(EIGEN_DEFAULT_L1_CACHE_SIZE)
+#define EIGEN_SET_DEFAULT_L1_CACHE_SIZE(val) EIGEN_DEFAULT_L1_CACHE_SIZE
+#else
+#define EIGEN_SET_DEFAULT_L1_CACHE_SIZE(val) val
+#endif // defined(EIGEN_DEFAULT_L1_CACHE_SIZE)
+
+#if defined(EIGEN_DEFAULT_L2_CACHE_SIZE)
+#define EIGEN_SET_DEFAULT_L2_CACHE_SIZE(val) EIGEN_DEFAULT_L2_CACHE_SIZE
+#else
+#define EIGEN_SET_DEFAULT_L2_CACHE_SIZE(val) val
+#endif // defined(EIGEN_DEFAULT_L2_CACHE_SIZE)
+
+#if defined(EIGEN_DEFAULT_L3_CACHE_SIZE)
+#define EIGEN_SET_DEFAULT_L3_CACHE_SIZE(val) EIGEN_SET_DEFAULT_L3_CACHE_SIZE
+#else
+#define EIGEN_SET_DEFAULT_L3_CACHE_SIZE(val) val
+#endif // defined(EIGEN_DEFAULT_L3_CACHE_SIZE)
+  
 #if EIGEN_ARCH_i386_OR_x86_64
-const std::ptrdiff_t defaultL1CacheSize = 32*1024;
-const std::ptrdiff_t defaultL2CacheSize = 256*1024;
-const std::ptrdiff_t defaultL3CacheSize = 2*1024*1024;
+const std::ptrdiff_t defaultL1CacheSize = EIGEN_SET_DEFAULT_L1_CACHE_SIZE(32*1024);
+const std::ptrdiff_t defaultL2CacheSize = EIGEN_SET_DEFAULT_L2_CACHE_SIZE(256*1024);
+const std::ptrdiff_t defaultL3CacheSize = EIGEN_SET_DEFAULT_L3_CACHE_SIZE(2*1024*1024);
 #elif EIGEN_ARCH_PPC
-const std::ptrdiff_t defaultL1CacheSize = 64*1024;
-const std::ptrdiff_t defaultL2CacheSize = 512*1024;
-const std::ptrdiff_t defaultL3CacheSize = 4*1024*1024;
+const std::ptrdiff_t defaultL1CacheSize = EIGEN_SET_DEFAULT_L1_CACHE_SIZE(64*1024);
+const std::ptrdiff_t defaultL2CacheSize = EIGEN_SET_DEFAULT_L2_CACHE_SIZE(512*1024);
+const std::ptrdiff_t defaultL3CacheSize = EIGEN_SET_DEFAULT_L3_CACHE_SIZE(4*1024*1024);
 #else
-const std::ptrdiff_t defaultL1CacheSize = 16*1024;
-const std::ptrdiff_t defaultL2CacheSize = 512*1024;
-const std::ptrdiff_t defaultL3CacheSize = 512*1024;
+const std::ptrdiff_t defaultL1CacheSize = EIGEN_SET_DEFAULT_L1_CACHE_SIZE(16*1024);
+const std::ptrdiff_t defaultL2CacheSize = EIGEN_SET_DEFAULT_L2_CACHE_SIZE(512*1024);
+const std::ptrdiff_t defaultL3CacheSize = EIGEN_SET_DEFAULT_L3_CACHE_SIZE(512*1024);
 #endif
 
+#undef EIGEN_SET_DEFAULT_L1_CACHE_SIZE
+#undef EIGEN_SET_DEFAULT_L2_CACHE_SIZE
+#undef EIGEN_SET_DEFAULT_L3_CACHE_SIZE
+
 /** \internal */
 struct CacheSizes {
   CacheSizes(): m_l1(-1),m_l2(-1),m_l3(-1) {
@@ -60,7 +82,6 @@ struct CacheSizes {
   std::ptrdiff_t m_l3;
 };
 
-
 /** \internal */
 inline void manage_caching_sizes(Action action, std::ptrdiff_t* l1, std::ptrdiff_t* l2, std::ptrdiff_t* l3)
 {
diff --git a/doc/PreprocessorDirectives.dox b/doc/PreprocessorDirectives.dox
index 8adb7ec63..0f545b086 100644
--- a/doc/PreprocessorDirectives.dox
+++ b/doc/PreprocessorDirectives.dox
@@ -107,7 +107,7 @@ run time. However, these assertions do cost time and can thus be turned off.
  Let us emphasize that \c EIGEN_MAX_*_ALIGN_BYTES define only a diserable upper bound. In practice data is aligned to largest power-of-two common divisor of \c EIGEN_MAX_STATIC_ALIGN_BYTES and the size of the data, such that memory is not wasted.
  - \b \c EIGEN_DONT_PARALLELIZE - if defined, this disables multi-threading. This is only relevant if you enabled OpenMP.
    See \ref TopicMultiThreading for details.
- - \b EIGEN_DONT_VECTORIZE - disables explicit vectorization when defined. Not defined by default, unless 
+ - \b \c EIGEN_DONT_VECTORIZE - disables explicit vectorization when defined. Not defined by default, unless 
    alignment is disabled by %Eigen's platform test or the user defining \c EIGEN_DONT_ALIGN.
  - \b \c EIGEN_UNALIGNED_VECTORIZE - disables/enables vectorization with unaligned stores. Default is 1 (enabled).
    If set to 0 (disabled), then expression for which the destination cannot be aligned are not vectorized (e.g., unaligned
@@ -127,9 +127,11 @@ run time. However, these assertions do cost time and can thus be turned off.
    By default it is defined to \c __forceinline for MSVC and ICC, and to \c inline for other compilers. A tipical usage is to
    define it to \c inline for MSVC users wanting faster compilation times, at the risk of performance degradations in some rare
    cases for which MSVC inliner fails to do a good job.
+ - \b \c EIGEN_DEFAULT_L1_CACHE_SIZE - Sets the default L1 cache size that is used in Eigen's GEBP kernel when the correct cache size cannot be determined at runtime.
+ - \b \c EIGEN_DEFAULT_L2_CACHE_SIZE - Sets the default L2 cache size that is used in Eigen's GEBP kernel when the correct cache size cannot be determined at runtime.
+ - \b \c EIGEN_DEFAULT_L3_CACHE_SIZE - Sets the default L3 cache size that is used in Eigen's GEBP kernel when the correct cache size cannot be determined at runtime.
 
-
- - \c EIGEN_DONT_ALIGN - Deprecated, it is a synonym for \c EIGEN_MAX_ALIGN_BYTES=0. It disables alignment completely. %Eigen will not try to align its objects and does not expect that any objects passed to it are aligned. This will turn off vectorization if \b EIGEN_UNALIGNED_VECTORIZE=1. Not defined by default.
+ - \c EIGEN_DONT_ALIGN - Deprecated, it is a synonym for \c EIGEN_MAX_ALIGN_BYTES=0. It disables alignment completely. %Eigen will not try to align its objects and does not expect that any objects passed to it are aligned. This will turn off vectorization if \b \c EIGEN_UNALIGNED_VECTORIZE=1. Not defined by default.
  - \c EIGEN_DONT_ALIGN_STATICALLY - Deprecated, it is a synonym for \c EIGEN_MAX_STATIC_ALIGN_BYTES=0. It disables alignment of arrays on the stack. Not defined by default, unless \c EIGEN_DONT_ALIGN is defined.