It's good to see that the P1's were resolved rather than downgraded to P2's
It should make for a better release
I'm just hoping LTO will finally be usable across the whole system now
Phoronix: GCC 4.6 Release Candidate Comes w/o P1 Regressions
The GNU developers responsible for GCC have eliminated all of the P1 regressions (their most serious class of regressions in this open-source compiler) in the GNU Compiler Collection 4.6.0 code-base, so they have went ahead and tagged the first release candidate...
http://www.phoronix.com/vr.php?view=OTIxMA
It's good to see that the P1's were resolved rather than downgraded to P2's
It should make for a better release
I'm just hoping LTO will finally be usable across the whole system now
The new optimization level is called -Ofast not -0fast
"GCC 4.6 delivers Intel Sandy Bridge AVX support and other Core i7 / Core i5 / Sandy Bridge optimizations"
Since AMD's upcoming bulldozer can do/emulate AVX instructions, AFAIK, my question is whether GCC 4.6 will be able to make use of this Intel feature on future AMD CPU's?