This is the mail archive of the gcc@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: Quality of GCC 3.1


Wit Jakuczun wrote:
I have few comments and questions on quality of code generated by g++ from GCC 3.1.

1) Compiler seems to be much slower than GCC 2.95.2 (3 times on average).
The code heavily uses STL and exceptions.
There has been some earlier email pointing out that the compile-times have increased. However, this was in reference to C code and bootstrapping the compiler rather than C++ code.


2) Code generated is much slower (3 times on average) especially if streams
	are used
Do you have some concise examples that demonstrate this problem? It is hard to fix the problem without test cases. Having a test case that allows other people to replicate the problem helps immensely.


Questions:
a)	What is the reason of such big decrease in performance especially if
	exceptions and/or streams are used?
b)	Is it a problem of setting correct options for GCC 3.1?
c)	Where can I find a good comparison of C++ compilers?

Options I used to make GCC3.1 (using GCC 2.92.2):
--enable-threads --enable-nls --enable-cstdio=stdio
Did you "make bootstrap"? If you are just using the GCC 3.1 executable generated by GCC 2.95.2(?), that may be part of the problem. If you do not bootstrap the compiler, you are measuring the performance and efficiency of the code generated by gcc 2.95.2 when you measure compile-time.

The runtime libraries can also influence the performance. How much time is the code spending in the libraries versus in the GCC user code?


-Will



Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]