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]

Re: How long should -O1 compiles take?



I said:

> Is the global register allocation that much better in gcc-2.96 that it
> should take 89 seconds now instead of 2.45 seconds as in egcs-1.1.2?
> Or is flow analysis/control so much better that gcc should be spending
> 353 seconds on it rather than 38 seconds?  (I tried -fno-thread-jumps in
> gcc-2.95.1 to try to reduce the compile time, but instead it increased
> to 478.56u.)

Sorry, I worded that very poorly.  What I should have said is: Do the
gcc developers want a global register allocator that takes 89 seconds
in gcc -O1, whether or not it is significantly better than the old one
that took 2.45 seconds?  And the same question for other passes in
the compiler.  Is there a specific design trade-off in compilation
speed/generated code speed for -O1 that decides things like this?

Brad


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