Ekdohibs
/
ocaml
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Modernize the description of target-specific ocamlopt aspects 

No more "Intel Pentium" and "AMD Athlon".
master
Xavier Leroy 2020-10-16 10:12:30 +02:00
parent 0fac4b030b
commit e7d2c394c0
1 changed files with 9 additions and 8 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

17
manual/manual/cmds/native.etex
View File

@ -113,12 +113,13 @@ exclusive.
% compilers and toplevel
\input{unified-options.tex}
\paragraph{Options for the IA32 architecture}
The IA32 code generator (Intel Pentium, AMD Athlon) supports the
\paragraph{Options for the 32-bit x86 architecture}
The 32-bit code generator for Intel/AMD x86 processors ("i386"
architecture) supports the
following additional option:
\begin{options}
\item["-ffast-math"] Use the IA32 instructions to compute
\item["-ffast-math"] Use the processor instructions to compute
trigonometric and exponential functions, instead of calling the
corresponding library routines. The functions affected are:
"atan", "atan2", "cos", "log", "log10", "sin", "sqrt" and "tan".
@ -128,9 +129,9 @@ trigonometric operations "cos", "sin", "tan" have their range reduced to
$[-2^{64}, 2^{64}]$.
\end{options}
\paragraph{Options for the AMD64 architecture}
The AMD64 code generator (64-bit versions of Intel Pentium and AMD
Athlon) supports the following additional options:
\paragraph{Options for the 64-bit x84 architecture}
The 64-bit code generator for Intel/AMD x86 processors ("amd64"
architecture) supports the following additional options:
\begin{options}
\item["-fPIC"] Generate position-independent machine code. This is
@ -219,8 +220,8 @@ until the next heap allocation.
beneficial, but produces floating-point results that differ slightly
from those produced by the bytecode interpreter.
\item On IA32 processors only (Intel and AMD x86 processors in 32-bit
mode), some intermediate results in floating-point computations are
\item On Intel/AMD x86 processors in 32-bit mode,
some intermediate results in floating-point computations are
kept in extended precision rather than being rounded to double
precision like the bytecode compiler always does. Floating-point
results can therefore differ slightly between bytecode and native code.