📄 http:^^www.cs.wisc.edu^~cs354-2^cs354^solutions^q3.j.html
字号:
Date: Tue, 05 Nov 1996 20:51:28 GMTServer: NCSA/1.5Content-type: text/htmlLast-modified: Wed, 23 Oct 1996 20:23:48 GMTContent-length: 4641<HTML><HEAD> <META NAME="GENERATOR" CONTENT="Adobe PageMill 2.0 Mac"> <TITLE>Q3.ans.html</TITLE></HEAD><BODY><H3 ALIGN=CENTER>CS354, Fall 1996</H3><H3 ALIGN=CENTER>Quiz 3, Sections 1 & 3, 10/9/96 & 10/11/96</H3><P><TABLE WIDTH="100%" HEIGHT="36" BORDER="1" CELLSPACING="2" CELLPADDING="0"><TR><TD WIDTH="50%"> Name (printed): </TD><TD WIDTH="50%"> Name (signed):</TD></TR></TABLE></P><P>(1, ver 1) (10 points) Write a SAL code segment that prints a 1 if variablewd contains the bit sequence "101101" and prints a 0 if it doesnot. The code segment should leave variable wd unchanged.</P><PRE><B>ANSWER</B> move a, 0 # assume "101101" is not in wd move pattern, 0x0000002d # "101101" as a 6 bit pattern move mask, 0x0000003f # used to get 6 bits of wd move count, 0 # do at most 32-6+1=27 testsloop: and x, wd, mask # get 6 bits of wd bne pattern, x, cont move a, 1 # "101101" is in wd b end # might as well quitcont: add count, count, 1 sll mask, mask, 1 # "ror wd, wd, 1" is no good sll pattern, pattern, 1 bne count, 27, loopend: put a </PRE><P>(1, ver 2) (10 points) Write a SAL code segment that prints a 1 if variablewd does not contain the bit sequence "1001100" and prints a 0if it does. The code seqment should leave variable wd unchanged.</P><PRE><B>ANSWER</B> move a, 1 # assume "1001100" is not in wd move pattern, 0x0000004c # "10001100" as a 7 bit pattern move mask, 0x0000007f # used to get 7 bits of wd move count, 0 # do at most 32-7+1=26 testsloop: and x, wd, mask bne pattern, x, cont move a, 0 # "1001100" is in wd b end # might as well quitcont: add count, count, 1 sll mask, mask, 1 # "ror wd, wd, 1" is no good sll pattern, pattern, 1 bne count, 26, loopend: put a</PRE><P>(2, ver 1) (5 points) In the simp simulator, the "not" instructionwas left out. Explain why:</P><BLOCKQUOTE> <BLOCKQUOTE> <BLOCKQUOTE> <P>"xor a, a, -1" is equivalent to "not a, a"</P> </BLOCKQUOTE> </BLOCKQUOTE></BLOCKQUOTE><PRE><B>ANSWER</B>1 xor 1 is 00 xor 1 is 1so xor'ing with 1 acts like not'ing and-1 is all 1's in a 32 bit, 2's complement representation</PRE><P>(2, ver 2) (5 points) In the simp simulator, the "not" instructionwas left out. Explain why:</P><BLOCKQUOTE> <BLOCKQUOTE> <BLOCKQUOTE> <P>"xnor a, a, 0" is equivalent to "not a, a"</P> </BLOCKQUOTE> </BLOCKQUOTE></BLOCKQUOTE><PRE><B>ANSWER</B>1 xnor 0 is 00 xnor 0 is 1so xnor'ing with 0 acts like not'ing and0 is all 0's in a 32 bit, 2's complement representation</PRE><P>(3, ver 1) (10 points) Add the following two IEEE floating point numbers(hidden bit is shown). Show the IEEE form of the result, with hidden bitshowing (8 points). What is the value of the result in decimal (2 points)?</P><PRE> S E F 0 1000 1010 (1)100 0110 0000 0000 0000 0000 0 1000 0111 (1)001 1000 0000 0000 0000 0000<B>ANSWER</B>S E F0 1000 1010 (1)100 0110 0000 0000 0000 00000 1000 0111 (1)001 1000 0000 0000 0000 0000 <- smaller E, shift rt 3 placesS E F (3 = 10 - 7)0 1000 1010 (1)100 0110 0000 0000 0000 00000 1000 1010 (0)001 0011 0000 0000 0000 00000 1000 1010 (1)101 1001 0000 0000 0000 0000 <- ans = 3472.0</PRE><P>(3, ver 2) (10 points) Add the following two IEEE floating point numbers(hidden bit is shown). Show the IEEE form of the result, with hidden bitshowing (8 points). What is the value of the result in decimal (2 points)?</P><PRE> S E F 1 1000 0010 (1)100 1100 0000 0000 0000 0000 1 1000 0111 (1)001 1000 0000 0000 0000 0000<B>ANSWER</B>S E F1 1000 0010 (1)100 1100 0000 0000 0000 0000 <- smaller E,1 1000 0111 (1)001 1000 0000 0000 0000 0000 shift rt 5 places (5 = 7 - 2)S E F1 1000 0111 (0)000 0110 0110 0000 0000 00001 1000 0111 (1)001 1000 0000 0000 0000 00001 1000 0111 (1)001 1110 0110 0000 0000 0000 <- ans = -316.75</PRE></BODY></HTML>⌨️ 快捷键说明
复制代码
Ctrl + C
搜索代码
Ctrl + F
全屏模式
F11
切换主题
Ctrl + Shift + D
显示快捷键
?
增大字号
Ctrl + =
减小字号
Ctrl + -