Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Performance audit, Spring 2017 #41469

Merged
merged 8 commits into from
Apr 22, 2017
Merged

Performance audit, Spring 2017 #41469

merged 8 commits into from
Apr 22, 2017

Commits on Apr 22, 2017

  1. remove cleanup branches to the resume block 

    This improves LLVM performance by 10% lost during the shimmir transition.
    @arielb1
    Ariel Ben-Yehuda authored and arielb1 committed Apr 22, 2017
    Configuration menu
    Copy the full SHA
    3bf0045 View commit details
    Browse the repository at this point in the history

  2. avoid calling mk_region unnecessarily 

    this improves typeck & trans performance by 1%. This looked hotter on
callgrind than it is on a CPU.
    @arielb1
    Ariel Ben-Yehuda authored and arielb1 committed Apr 22, 2017
    Configuration menu
    Copy the full SHA
    e1377a4 View commit details
    Browse the repository at this point in the history

  3. short-cut SharedCrateContext::layout_of 

    That method is *incredibly* hot, so this ends up saving 10% of trans
time.

BTW, we really should be doing dependency tracking there - and possibly be
taking the respective perf hit (got to find a way to make DTMs fast), but
`layout_cache` is a non-dep-tracking map.
    @arielb1
    Ariel Ben-Yehuda authored and arielb1 committed Apr 22, 2017
    Configuration menu
    Copy the full SHA
    acd0e40 View commit details
    Browse the repository at this point in the history

  4. cache attributes of items from foreign crates 

    this avoids parsing item attributes on each call to `item_attrs`, which takes
off 33% (!) of translation time and 50% (!) of trans-item collection time.
    @arielb1
    Ariel Ben-Yehuda authored and arielb1 committed Apr 22, 2017
    Configuration menu
    Copy the full SHA
    ece6c84 View commit details
    Browse the repository at this point in the history

  5. weak_lang_items: check for lang attribute before calling value_str 

    improves trans performance by *another* 10%.
    @arielb1
    Ariel Ben-Yehuda authored and arielb1 committed Apr 22, 2017
    Configuration menu
    Copy the full SHA
    266d36f View commit details
    Browse the repository at this point in the history

  6. allocate less strings in symbol_names 

    this improves trans performance by *another* 10%.
    @arielb1
    Ariel Ben-Yehuda authored and arielb1 committed Apr 22, 2017
    Configuration menu
    Copy the full SHA
    81af6fb View commit details
    Browse the repository at this point in the history

  7. add a cache to impl_polarity 

    this is another one of these things that looks *much* worse on valgrind.
    @arielb1
    Ariel Ben-Yehuda authored and arielb1 committed Apr 22, 2017
    Configuration menu
    Copy the full SHA
    a0f145b View commit details
    Browse the repository at this point in the history

  8. bail out of selection when there are multiple surviving candidates 

    In some cases (e.g. <[int-var] as Add<[int-var]>>), selection can turn up
a large number of candidates. Bailing out early avoids O(n^2) performance.

This improves item-type checking time by quite a bit, resulting in ~2% of total
time-to-typeck.
    @arielb1
    Ariel Ben-Yehuda authored and arielb1 committed Apr 22, 2017
    Configuration menu
    Copy the full SHA
    a660ad8 View commit details
    Browse the repository at this point in the history