On May 16 2013, at 10:31 , David Chase wrote: > > On 2013-05-16, at 12:17 PM, Mike Duigou <mike.dui...@oracle.com> wrote: >> I haven't looked at the details for the PCLMULQDQ instruction but a caryless >> multiply could be of use to some of the crypto primitives as well (GHASH, >> GMAC and probably others). Perhaps the property could be >> "sun.hotspot.x64.clmulSupported" or something less specific to the usage. > > Note that the instruction also works on 32-bit, and once the builds all use > sufficiently modern compilers, the same source code works for both.
Understood, but it's only going to be available on processors that support also support EMT64, correct? I am not sure what the best way to characterize the required architecture should be. >> What's our actual experience with needing switches like -XX:-UseCLMUL >> or-XX:-UseAVX for other features? Faulty implementations? Feature >> misreporting? Performance regressions? Virtualization interactions? > > I haven't seen any virtualization problems -- I've also been testing on VMs, > Ubuntu and Solaris on VMWare. > Should I also give it a go in VirtualBox? What should I be looking for? I don't know of any specific problems but was more curious whether we really need the ability to disable the feature. Have we seen problems in the past? > David >
