[PATCH 7 of 7] ancestors: return all ancestors - not just those who are furthest from a root

Mads Kiilerich mads at kiilerich.com
Tue Feb 25 17:54:24 CST 2014


On 02/24/2014 11:31 PM, Matt Mackall wrote:
> On Mon, 2014-02-24 at 23:19 +0100, Mads Kiilerich wrote:
>> # HG changeset patch
>> # User Mads Kiilerich <madski at unity3d.com>
>> # Date 1393278134 -3600
>> #      Mon Feb 24 22:42:14 2014 +0100
>> # Node ID 2b44103c2e866a3a06adc3b7f91cdb609754d4db
>> # Parent  c5b597a62f2231919df87fe2915a4edf1023166b
>> ancestors: return all ancestors - not just those who are furthest from a root
>>
>> The function seemed to solve a harder problem than necessary. It gave some
>> common ancestors priority over others. This is a low level function and it
>> should consider all equal and leave it to the higher levels to decide how to
>> handle the multiple heads. Now they are all equal.
> I'll need to think on this. But you should probably share some
> benchmarks.

What benchmarks do you have in mind?

My concern here is only correctness, not performance. The patch just 
removes a post processing step in the gca calculation and can thus only 
make it faster. The algorithm did two linear iterations over the full 
repo with the option of early exit - now it do one.

/Mads


More information about the Mercurial-devel mailing list