mirror of
https://github.com/vortexgpgpu/vortex.git
synced 2025-04-24 05:47:35 -04:00
Floating point clarification.
git-svn-id: http://www.cdkersey.com/harp/harptool@127 0246edb2-e076-4747-b392-db732a341fa2
This commit is contained in:
parent
45f85cd29d
commit
379cb7ed69
1 changed files with 4 additions and 1 deletions
|
@ -148,7 +148,7 @@ After this, the operands of the instruction are ordered corresponding to their o
|
|||
\begin{itemize}
|
||||
\item Register operands are $log_2(\mathrm{\#GPRs})$ bits long, or just enough bits to uniquely identify a register.
|
||||
\item Predicate register operands are $log_2(\mathrm{\#pred\_regs})$ bits long, or just enough bits to uniquely identify a predicate register.
|
||||
\item Immediate fields are always the last field and occupy the remaining bits of the instruction. All immediate fields are assumed to be sign extended.
|
||||
\item Immediate fields are always the last field and occupy the remaining bits of the instruction. All immediate fields are sign extended to the length of a machine word.
|
||||
\end{itemize}
|
||||
|
||||
\begin{center}
|
||||
|
@ -306,6 +306,9 @@ The flags register used by \texttt{tlbadd} stores, in its least-significant four
|
|||
\end{center}
|
||||
|
||||
\subsection{Floating Point Arithmetic}
|
||||
These operations operate on real numbers in an implementation-determined
|
||||
format, which can be fixed point or floating point.
|
||||
|
||||
\begin{center}
|
||||
\begin{tabular}{cl}
|
||||
\textbf{Instruction}&\textbf{Description}\\
|
||||
|
|
Loading…
Add table
Add a link
Reference in a new issue