Michael Haggerty [off-list ref] writes:
On 06/03/2016 11:33 PM, Junio C Hamano wrote:
quoted
Michael Haggerty [off-list ref] writes:
quoted
This patch series applies on top of mh/split-under-lock. It can also
be obtained from my GitHub repo [2] as branch "ref-iterators".
Ah, that reminds me. What's the doneness of the dependent topic?
If you mean the vtable implementation, check your inbox :-) I already
submitted it (38 patches, but most of them are straightforward). I've
been iterating on that patch series for quite a while so I'm pretty
confident that it's solid.
What I meant was the doneness of mh/split-under-lock actually.
quoted
The patches in this series looked all good to me.
Thanks for the review (and your patience), both David and Junio!
Will take a look at the updated vtable series. Thanks.
On 06/04/2016 01:40 AM, Junio C Hamano wrote:
Michael Haggerty [off-list ref] writes:
quoted
On 06/03/2016 11:33 PM, Junio C Hamano wrote:
quoted
[...]
Ah, that reminds me. What's the doneness of the dependent topic?
[...]
What I meant was the doneness of mh/split-under-lock actually.
I don't know of any open issues or unaddressed feedback in
split-under-lock. I've gone over the code many times myself and am
pretty confident about it.
I just reviewed the series again, and one thing that I wasn't so happy
about was that it makes many changes to the ref transaction
error-handling code, but that code is not very well tested (and the
content of the error messages is not really tested at all). So I just
submitted another patch series to add more tests in this area [1]. Those
tests turned up a couple of cosmetic problems but nothing serious.
So as far as I know, this series is ready to go. But it is a big and
rather intricate change to an important area of the code, so I hope it
can get merged early in the next release cycle to get wider testing.
Michael
[1] http://thread.gmane.org/gmane.comp.version-control.git/296626