/jtag/glyn_tonga2/

ow any performance impact. lcs_upper_bound's second sum is required all the time and try to handle endians 2010-12-27T07:14:42Z Eric Anderson ejona86@gmail.com 2010-12-27T07:14:42Z urn:sha1:34f3a25cf608320e8b95ada8cbe412ac7ec46be5 Only do second sum if it can make a difference 2010-12-26T16:18:22Z Eric Anderson ejona86@gmail.com 2010-12-26T16:18:22Z urn:sha1:e4c49643c47f4d69282a8d1ad2f05619bd4d86b9 Auto-choose at runtime to do second sum in lcs_upper_bound 2010-12-26T16:09:03Z Eric Anderson ejona86@gmail.com 2010-12-26T16:09:03Z urn:sha1:ff2d93e1b4ea34ddaa42be5e9c987861849e4b2b This has no impact for non-DO_SECOND_SUM, but puts DO_SECOND_SUM less than a percent behind. Simple movement of return calculation in lcs_upper_bound 2010-12-26T16:01:42Z Eric Anderson ejona86@gmail.com 2010-12-26T16:01:42Z urn:sha1:4c91b8ea5950834693130343899e0fa379fd3be1 This increased performance of the non-DO_SECOND_SUM code a _lot_ for something so small. Only add one to lcs_upper_bound or don't do the second sum. 2010-12-26T15:40:15Z Eric Anderson ejona86@gmail.com 2010-12-26T15:40:15Z urn:sha1:3824e95334e06d9c1309591e7a040dc7b4049ff5 The DO_SECOND_SUM define has reduced false positives, but without it has a reasonable amount less work. Thus, you must empirically test to find which has the best boost. Right now, DO_SECOND_SUM is faster. Add lcs_upper_bound check before lcs_32_rough for added speed 2010-12-26T14:47:02Z Eric Anderson ejona86@gmail.com 2010-12-26T14:47:02Z urn:sha1:fbe623251cc4c8ebcab2cd7c6a4e8e2c9a5f6ed7