居然发现了一个VS2012 C编译器 float的一个BUG！！！

今天做一个实验的时候，莫名其妙地出错。。。

然后不断地DEBUG，最后居然发现下面问题：

float check1=33359568.000,check2=272322.000;

printf("%f\n",check1+check2);

在GCC下可以输出正确结果33631890.000，但是在VS2012下，结果却是33631888.000！！！

知道怎么回事的可以回复或者在http://social.msdn.microsoft.com/Forums/en-US/80ca3c69-c78b-44f2-bf71-80e01bb9f0e3/it-is-a-bug-in-vs2012-c-compiler-of-float-add-operation?forum=msbuild 这里回答我的问题，谢谢。

答案就是精度的问题：

引自 Anna Cc 的回答

I suggest you read this thread,

Difference between float and double

As the name implies, a double has 2x the precision of float^[1]. In general a double has 15 to 16 decimal digits of precision, while float only has 7.

This precision loss could lead to truncation errors much easier to float up.

"don't use += to sum lots of floating point numbers as the errors accumulate quickly."

The C and C++ standards do not specify the representation of float, double and long double. It is possible that all three implemented as IEEE double-precision. Nevertheless, for most architectures (gcc, MSVC; x86, x64, ARM) float is indeed a IEEE single-precision floating point number (binary32), and double is a IEEE double-precision floating point number (binary64).