Experience report on org 9.6-dev perf improvement for tangling noweb-ref heavy files

Experience report on org 9.6-dev perf improvement for tangling noweb-ref heavy files

[pareto optimal]

[-- Attachment #1: Type: text/plain, Size: 750 bytes --]

Hi! I reported an issue with org-babel-tangle performance issues with lots of noweb references here:

https://lists.gnu.org/archive/html/emacs-orgmode/2022-01/msg00484.html

Using 2b49d6fd9c04020a8132bfcfffa449e3e05b47e1 on my personal (but not yet public) config that demonstrated that issue I see a great improvement!

It has 124 source blocks, 70 of which are noweb-ref, and I believe some of those noweb-refs also refer to eachother if that matters for performance considerations.

org 9.5.2:               1.28, 1.33, 1.32
org 9.6-dev:           0.46, 0.44, 0.43

Nearly 3x or 200% faster!

Thanks to Ihor Radchenko and others who worked on this!
-- Sent with https://mailfence.com  Secure and private email

[-- Attachment #2: Type: text/html, Size: 1072 bytes --]

Re: Experience report on org 9.6-dev perf improvement for tangling noweb-ref heavy files

[Ihor Radchenko]

pareto optimal <pareto.optimal@mailfence.com> writes:

> Hi! I reported an issue with org-babel-tangle performance issues with lots of noweb references here:
>
> https://lists.gnu.org/archive/html/emacs-orgmode/2022-01/msg00484.html
>
> Using 2b49d6fd9c04020a8132bfcfffa449e3e05b47e1 on my personal (but not yet public) config that demonstrated that issue I see a great improvement!
>
> It has 124 source blocks, 70 of which are noweb-ref, and I believe some of those noweb-refs also refer to eachother if that matters for performance considerations.
>
> org 9.5.2:               1.28, 1.33, 1.32
> org 9.6-dev:           0.46, 0.44, 0.43
>
> Nearly 3x or 200% faster!

Would you mind sharing the profiler report for org 9.6-dev with your
config? I am trying to gather more information on the remaining
bottlenecks.

Best,
Ihor