On Thu, May 07, 2009 at 02:18:02PM +0200, Kristian Nielsen wrote:
> I thought a bit more about the issues of merging PBXT (and later XtraDB)
> changes into MariaDB. I would like to understand better how to do it.
>
> The main bzr repositories of both PBXT and XtraDB contain just the storage
> engi
One of the dangers of microbenchmarks is that they do not consider cache
misses well. Unrolling one loop is no harm, but what happens when you
have hundreds of unrolled loops in your program? I would imagine that
the probability of cache misses increase greatly, perhaps so much that
it regresse
Have we tried running the benchmarks with and without -funroll-loops?
-j
Stewart Smith wrote:
So i broke out mtaylor's patch that we were bumming around with at the UC
to replace ptr_compare with a simple memcmp call.
At the UC I benched that this patch actually caused a measurable performance
Hi!
> "Kristian" == Kristian Nielsen writes:
Kristian> I noticed a difference between Bitkeeper and Bzr merges.
Thanks for the analyse and thinking about this!
First, in Sun when we switched to bzr, we didn't care about the
difference and continued as if nothing had changed. We did get
Hi!
> "Toby" == Toby Thain writes:
Toby> On 8-May-09, at 1:45 AM, Kristian Nielsen wrote:
>> Kristian Nielsen writes:
>>
>>> I suggest you re-commit all of the changes as a single commit
>>> (this is
>>> generally preferred in Maria development), or at least add a good
>>> commit
>>>
5 matches
Mail list logo
