Document fast math

doc/man/julia.1

@@ -120,6 +120,11 @@ Enable or disable color text
 --check-bounds={yes|no}
 Emit bounds checks always or never (ignoring declarations)
 
+.TP
+--math-mode={ieee|user}
+Always use IEEE semantics for math (ignoring declarations),
+or adhere to declarations in source code
+
 .TP
 --int-literals={32|64}
 Select integer literal size independent of platform

doc/manual/getting-started.rst

@@ -130,6 +130,8 @@ those available for the ``perl`` and ``ruby`` programs::
  --track-allocation={none|user|all}
  Count bytes allocated by each source line
  --check-bounds={yes|no} Emit bounds checks always or never (ignoring declarations)
+ --math-mode={ieee|user} Always use IEEE semantics for math (ignoring declarations),
+ or adhere to declarations in source code
  -O, --optimize Run time-intensive code optimizations
  --int-literals={32|64} Select integer literal size independent of platform
  --dump-bitcode={yes|no} Dump bitcode for the system image (used with --build)

doc/manual/performance-tips.rst

@@ -546,6 +546,10 @@ properties.
 - Use :obj:`@inbounds` to eliminate array bounds checking within expressions.
  Be certain before doing this. If the subscripts are ever out of bounds,
  you may suffer crashes or silent corruption.
+- Use :obj:`@fastmath` to allow floating point optimizations that are
+ correct for real numbers, but lead to differences for IEEE numbers.
+ Be careful when doing this, as this may change numerical results.
+ This corresponds to the ``-ffast-math`` option of clang.
 - Write :obj:`@simd` in front of ``for`` loops that are amenable to vectorization.
  **This feature is experimental** and could change or disappear in future
  versions of Julia.