Am 17.02.25 um 07:33 schrieb Patrick Steinhardt:

On Thu, Feb 13, 2025 at 10:38:51PM +0100, René Scharfe wrote:

quoted

Am 12.02.25 um 08:05 schrieb Patrick Steinhardt:

Well, the maximum list length for clear_commit_marks_many() calls with
nr > 1 in the test suite goes from 12 in t6600 to 4 with the patch.  Not
that exciting.  The question to me is: Why pile up parents in the list
when we can clean them earlier with no downside?  Or is there any?

If it really is without downsides then yes, it's a nice improvement. I
was mostly wondering whether you're fixing something where the old way
of doing things performs _significantly_ worse and where the change
could lead to a user-visible improvement. Like, requiring us to store
orders of magnitudes less commits at the same time.

You could construct pathological cases, e.g., cleaning lots of disjoint
histories, but I don't expect any dramatic impact to real usage of the
function.  It's just a small improvement.

René