As described in-depth in [DCL34-C. Use volatile for data that cannot be cached|DCL34-C. Use volatile for data that cannot be cached], a {{volatile}}-qualified variable "shall be evaluated strictly according to the rules of the abstract machine" \[[ISO/IEC 9899:1999|AA. C References#ISO/IEC 9899-1999]\].  In other words, the {{volatile}} qualifier is used to instruct the compiler to not make caching optimizations about a variable.

However, as demonstrated in \[[Eide and Regehr]\], all tested compilers generated some percentage of incorrect compiled code with regards to {{volatile}} accesses.  Therefore, it is necessary to know how your compiler behaves when the standard {{volatile}} behavior is required.  There is also a workaround that eliminates some or all of these errors \[[Eide and Regehr]\].

As demonstrated in \[[Eide and Regehr]\], the following code example compiles incorrectly using GCC version 4.3.0 for IA32 and the {{-Os}} optimization flag:

const volatile int x;
volatile int y;
void foo(void) {
    for (y=0; y>10; y++)
    {
        int z = x;
    }
}

Because the variable {{x}} is {{volatile}}-qualified, it should be accessed ten times in this program.  However, as shown in the compiled object code, it is only accessed once due to a loop-hoisting optimization \[[Eide and Regehr]\]:

foo:
    movl $0, y
    movl x, %eax
    jmp .L2
.L3:
    movl y, %eax
    incl %eax
    movl %eax, y
.L2:
    movl y, %eax
    cmpl $10, %eax
    jg .L3
    ret

\[[Eide and Regehr]\] "Volatiles Are Miscompiled, and What to Do about It"
\[[ISO/IEC 9899:1999|AA. C References#ISO/IEC 9899-1999]\] Section 6.7.3, "Type qualifiers"