cd cd cs31 cd weeklylab mkdir week02 cd week02 pwd cp ~newhall/public/cs31/week02/* . ls Makefile printtypes.c testprog.c
Bit-wise Operations: AND (&) , OR (|), NOT (~), XOR(^) A B A & B A | B ~A A ^ B ------------------------------------------------------ 0 0 0 0 1 0 0 1 0 1 1 1 1 0 0 1 0 1 1 1 1 1 0 0 smallest unit is really a byte so do bitwise ops on some number of bytes 01010101 01101010 10101010 ~10101111 | 00100001 & 10111011 ^ 01101001 --------- ----------- ---------- --------- 01010000 01110101 00101010 11000011 Bit shift: val << amt val >> amt 10110111 << 2 is 11011100 000111 >>2 is 00000001 signed (arithmetic) right shift: keeps sign bit 100011111 >>2 is 111000111 in a C program would usually shift a variable value: int y, x = 10234; y = x >>2;
C has many different types for storing integer values. These types differ by the number of bytes they use to store values, and by weather or not they store both positive and negative values.
1 byte: | 2 bytes: | 4 bytes: | 4 or 8 bytes: (depends on arch) |
8 bytes: | char | short | int | long | long long |
unsigned char | unsigned short | unsigned int | unsigned long | unsigned long long |
When you allocate a variable of a specific type, you get a storage location of the appropriate number of bytes associated with that variable name:
int x; // 4 bytes of storage space to store a signed integer value unsigned char ch; // 1 byte of storage space to store an unsigned integer value
### Specifying the numberic representation: %d: print out value in decimal (base 10) %u: print out value in unsigned decimal (base 10) %x: print out value in hexidecimal (base 16) %o: print out value in octal (base 8) (there is no formatting option to display the value in binary) ### Specifying the type: %d: int (to print numeric values of int, short, and char args) %ld: long int %lld: long long int %u: unsigned %lu: long unsigned %llu: long long unsigned %p: an address value %f: float or double %lf: double %e: float or double in scientific notation %g: float in either %e or %f format %c: char (ex. 'x') %s: string (ex. "hello there") ### Specifying field width: %5d: print out the value in decimal in a field of with 5 %-5d: print out the value in decimal in a field of with 5, left justified %6.4f: print out a float in a field with of 6 with a precision of 4
gdb's print command can be used to print the value of a expression in different representations (binary, decimal, hex); you can use it as a simple calculator to verify answers to hex, binary, and decimal arithmetic. For this use of gdb, we don't have to have an executable to run gdb on. We can just run gdb, and then call its print command:
$ gdb # print an expression in different representations: # (/t in binary, /x in hexidecimal, default is decimal): (gdb) print/t 1234 # print/t: print decimal value 1234 in binary format # p is shorthand for print (gdb) p/x 1234 # p/x: print value in hexidecimal format (gdb) p/d 1234 # p/d: print value in decimal format (the default) # 0x is the prefix for a hexidecimal literals # 0b is the prefix for a binary literals # no prefix: for decimal literals (gdb) p 0xabf1 # print the hex value abf1 as decimal (gdb) p 0b0101 # print the binary value 0101 as decimal (gdb) p/t 0x1234 # print the hex value 0x1234 as binary # (note: leading 0's are not printed out) (gdb) p/d 0b1010 # print the binary value 01010 as decimal (gdb) p/x 0b10100000 # print a binary value in hex (gdb) p/t 0b101001001 + 0xa2 # add a binary and a hex value, print result in binary # you can re-cast a value as a specific C type: (gdb) p/t (char)(12) # tell gdb that the value 12 is a char (1 byte) and print as binary (gdb) p/t (char)(-12) # print -12 char value (1 byte) in binary
Today we will try out gdb on an executable file, learn some basic gdb commands. Over the course of the semester we will learn a lot more gdb. Let's open up testprog.c in vim, and take a look at it. Next, let's compile it and run it inside gdb (note: we want to compile with -g when using gdb):
$ make $ gdb ./testprog (gdb) break main # sets a breakpoint in main (gdb) run (gdb) list # list out program source code around breakpoint (gdb) break 31 # set a break point at line 31 (gdb) cont # continue execution until hit a breakpoint (gdb) next # execute the next instruction, then gdb gets control again (gdb) list (gdb) print x # print out the values of program variables and expressions (gdb) p (y - 2) (gdb) p/t x # some bit-wise operators (gdb) p/t !x (gdb) p/t x & 0b10101010 (gdb) p/t x | 0b10101010 (gdb) p/t x ^ 0b10101010 (gdb) p/t x << 2 # bit shift (gdb) p x << 2 (gdb) p uch # operations on different C types (gdb) p/c uch (gdb) p uch+257 # this converts uch to an int, then adds to 257, the result is an int (gdb) p (char)(uch+257) # re-cast the result as a char # set: to change the value of program variables in gdb: (gdb) set x = 10 (gdb) p x (gdb) p/x x - 30 # the examine (or x) command is similar to print but takes a memory address # (or a variable location) as its argument (gdb) x &x # examine memory at the memory address of the variable x (gdb) x &y (gdb) x/s s # the name of a string or array is a synonym for its address (gdb) x/d s (gdb) x/d array (gdb) x/d &array[3] # the memory address of the 3rd bucket in array (gdb) quit
Try to identify some of the following:
Before leaving lab, please clean-up all spare parts, screws, etc. that you have removed by puting them in the boxes. You do not need to put the cases in the boxes, just all loose parts. CPU thermal glue is toxic, so just to be extra safe I'd recommend washing your hands after lab today.