-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 11/07/11 13:28, Kai Tietz wrote:
> Sure. A more general question, which was raised by Richi here. > For BC optimization it is of course interesting to know real > instruction-costs and not just guessings. The current code in > fold-const guess a bit too optimistic, the code in that patch is > in some cases too pesimistic. So is there a facility to get > prediction of instruction-costs on gimple-tree? I would say this > isn't possible, as here we would need final RTL form to tell this, > isn't it? We don't have much costing information in the gimple IL; that is largely by design. We generally wanted to avoid introducing cost information into the gimple/ssa optimizers so that their behavior would be more predictable. The fact that BRANCH_COST is utilized in fold-const.c is a historical blunder that we haven't fixed yet. jeff -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.11 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iQEcBAEBAgAGBQJOvWuwAAoJEBRtltQi2kC7ElgIAIOuttaZN4ff6I4WnkN1cmUk 5dr2B4ROVe49cLyq2NOXPp2zLxtxfDG6PkWDi+w1y+XXEXerRn/ncHIczYSZi+Be /DMhJDsog1GWUzblHTYvE4RluaB26+Khaj8yLgxZKhASAPw4h8Tjf7HwVedjiy1k PDJC/as8lbGHHgOfxrzOL3t1NXx8RthrtwR7pKBvq1oq0oomjsIEPEa+ubZJfnBJ TuKLJczcYCPKkgKBm5+wdOe5dO9NdpdfH5GVdEAPa+2lnXdrK7PXuYGz75uF9Hm1 j1zlO+zGP6qIS7FtEm7N70TZWcmPfmIKtamZTqT9JLsrvPTpzLkC4/PN486PiwQ= =U8VZ -----END PGP SIGNATURE-----
