PS:
> I verified that this builds on GNU/Linux on i686, on x86_64 (where SSE2 is
> part of the baseline and can thus be assumed safely) and on ARM (where
> SSE2 does not exist at all and so the patch should have no effect). I have
> not done runtime testing yet.
I have now done a basic functionali
On January 24, 2016 21:11:18 Hausmann Simon
wrote:
> Hi,
>
> Could you elaborate where you see copy on write causing writes to shared
> cache lines? Are you concerned about the shared cache line for the reference
> count?
>
> For reading MESI allows for shared cache lines and for hyper threads
Hi,
Could you elaborate where you see copy on write causing writes to shared cache
lines? Are you concerned about the shared cache line for the reference count?
For reading MESI allows for shared cache lines and for hyper threads the shared
l1 data cache mode favors sharing and thus CoW.
What
On January 24, 2016 17:45:36 Kevin Kofler wrote:
> Marc Mutz wrote:
>> (numThread == 2, same box)
>>
>> Copying is still not significantly slower than ref-counting, even for 4K
>> elements.
>
> But it is already slower with as little as 32 elements, and stops being
> significantly faster alread
Marc Mutz wrote:
> (numThread == 2, same box)
>
> Copying is still not significantly slower than ref-counting, even for 4K
> elements.
But it is already slower with as little as 32 elements, and stops being
significantly faster already at 16 elements.
And now try with numThread == 1 for some ex
Folly string is doing CoW only for sizes bigger than 255 and I believe Facebook
has measured it because for them it is money. ;-) I am not sure if they use
atomics. Maybe you could benchmark them too.
On January 24, 2016 03:21:31 Marc Mutz wrote:
> On Sunday 24 January 2016 03:01:57 Kevin Ko
