On Thu, Jan 27, 2011 at 2:03 AM, David Gilbert <david.gilb...@linaro.org> wrote: > On 26 January 2011 12:12, Dave Martin <dave.mar...@linaro.org> wrote: >> Hi Vijay, >> >> On Sat, Jan 22, 2011 at 9:59 AM, Vijay Kilari <vijay.kil...@gmail.com> wrote: >>> Hello Dave, >>> >>> Thanks for this info. >>> >>> I have few more queries after looking at the results of memset on A9 & A8. >>> I agree that externel bus speed matters in comparision across platforms. >>> >>> 1) Why memset is performance is good on A8 than A9?. any justification? >> >> I've CC'd the linaro-toolchain list who have been working on this >> topic and may be able to provide you with more information. > > Unfortunately we don't know why Neon was a bad idea for memset etc > on A9, it's just the tests show it being worse and the advice we get > says to avoid it - we've just not got an explanation.
To add some hearsay to the mix, I understand that the NEON unit on the A8 has a 128 bit connection straight into the L2 cache while the core has a 64 bit connection into the L1. This means that a NEON memcpy() (and probably memset()) on the A8 does twice as well for large or cold writes. -- Michael _______________________________________________ linaro-toolchain mailing list linaro-toolchain@lists.linaro.org http://lists.linaro.org/mailman/listinfo/linaro-toolchain
