git@vger.kernel.org mailing list mirror (one of many)
 help / color / mirror / code / Atom feed
From: Karthik Nayak <karthik.188@gmail.com>
To: Patrick Steinhardt <ps@pks.im>, git@vger.kernel.org
Cc: Han-Wen Nienhuys <hanwenn@gmail.com>, Justin Tobler <jltobler@gmail.com>
Subject: Re: [PATCH v2 09/10] reftable/block: reuse `zstream` state on inflation
Date: Wed, 10 Apr 2024 03:15:56 -0700	[thread overview]
Message-ID: <CAOLa=ZTnnmBvwXMbHqJ+58kPGe4jwLhu_GaEfkAXyfVa5imeYA@mail.gmail.com> (raw)
In-Reply-To: <587b5601c0c51a8f041258a27e693a2590468eed.1712578376.git.ps@pks.im>

[-- Attachment #1: Type: text/plain, Size: 1351 bytes --]

Patrick Steinhardt <ps@pks.im> writes:

> When calling `inflateInit()` and `inflate()`, the zlib library will
> allocate several data structures for the underlying `zstream` to keep
> track of various information. Thus, when inflating repeatedly, it is
> possible to optimize memory allocation patterns by reusing the `zstream`
> and then calling `inflateReset()` on it to prepare it for the next chunk
> of data to inflate.
>
> This is exactly what the reftable code is doing: when iterating through
> reflogs we need to potentially inflate many log blocks, but we discard
> the `zstream` every single time. Instead, as we reuse the `block_reader`
> for each of the blocks anyway, we can initialize the `zstream` once and
> then reuse it for subsequent inflations.
>
> Refactor the code to do so, which leads to a significant reduction in
> the number of allocations. The following measurements were done when
> iterating through 1 million reflog entries. Before:
>
>   HEAP SUMMARY:
>       in use at exit: 13,473 bytes in 122 blocks
>     total heap usage: 23,028 allocs, 22,906 frees, 162,813,552 bytes allocated
>
> After:
>
>   HEAP SUMMARY:
>       in use at exit: 13,473 bytes in 122 blocks
>     total heap usage: 302 allocs, 180 frees, 88,352 bytes allocated
>

Really nice how just reusing the data structure has such a significant
impact.

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 690 bytes --]

  reply	other threads:[~2024-04-10 10:16 UTC|newest]

Thread overview: 31+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-03-27  6:36 [PATCH 0/9] reftable: optimize table and block iterators Patrick Steinhardt
2024-03-27  6:36 ` [PATCH 1/9] reftable/block: rename `block_reader_start()` Patrick Steinhardt
2024-03-27  6:37 ` [PATCH 2/9] reftable/block: merge `block_iter_seek()` and `block_reader_seek()` Patrick Steinhardt
2024-03-27  6:37 ` [PATCH 3/9] reftable/block: better grouping of functions Patrick Steinhardt
2024-03-27  6:37 ` [PATCH 4/9] reftable/block: introduce `block_reader_release()` Patrick Steinhardt
2024-04-03 13:16   ` Karthik Nayak
2024-04-08 12:10     ` Patrick Steinhardt
2024-03-27  6:37 ` [PATCH 5/9] reftable/block: move ownership of block reader into `struct table_iter` Patrick Steinhardt
2024-04-03  4:52   ` Justin Tobler
2024-04-03 13:10     ` Patrick Steinhardt
2024-03-27  6:37 ` [PATCH 6/9] reftable/reader: iterate to next block in place Patrick Steinhardt
2024-03-27  6:37 ` [PATCH 7/9] reftable/block: reuse uncompressed blocks Patrick Steinhardt
2024-03-27  6:37 ` [PATCH 8/9] reftable/block: open-code call to `uncompress2()` Patrick Steinhardt
2024-03-27  6:37 ` [PATCH 9/9] reftable/block: reuse `zstream` state on inflation Patrick Steinhardt
2024-04-03 13:33 ` [PATCH 0/9] reftable: optimize table and block iterators Karthik Nayak
2024-04-08 12:16 ` [PATCH v2 00/10] " Patrick Steinhardt
2024-04-08 12:16   ` [PATCH v2 01/10] reftable/block: rename `block_reader_start()` Patrick Steinhardt
2024-04-08 12:16   ` [PATCH v2 02/10] reftable/block: merge `block_iter_seek()` and `block_reader_seek()` Patrick Steinhardt
2024-04-08 12:16   ` [PATCH v2 03/10] reftable/block: better grouping of functions Patrick Steinhardt
2024-04-08 12:16   ` [PATCH v2 04/10] reftable/block: introduce `block_reader_release()` Patrick Steinhardt
2024-04-08 12:16   ` [PATCH v2 05/10] reftable/block: move ownership of block reader into `struct table_iter` Patrick Steinhardt
2024-04-08 12:16   ` [PATCH v2 06/10] reftable/reader: iterate to next block in place Patrick Steinhardt
2024-04-08 12:16   ` [PATCH v2 07/10] reftable/block: reuse uncompressed blocks Patrick Steinhardt
2024-04-08 12:16   ` [PATCH v2 08/10] reftable/block: open-code call to `uncompress2()` Patrick Steinhardt
2024-04-08 12:17   ` [PATCH v2 09/10] reftable/block: reuse `zstream` state on inflation Patrick Steinhardt
2024-04-10 10:15     ` Karthik Nayak [this message]
2024-04-08 12:17   ` [PATCH v2 10/10] reftable/block: avoid copying block iterators on seek Patrick Steinhardt
2024-04-09  1:29     ` Justin Tobler
2024-04-09  3:18       ` Patrick Steinhardt
2024-04-09  1:32   ` [PATCH v2 00/10] reftable: optimize table and block iterators Justin Tobler
2024-04-10 11:35   ` Karthik Nayak

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

  List information: http://vger.kernel.org/majordomo-info.html

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to='CAOLa=ZTnnmBvwXMbHqJ+58kPGe4jwLhu_GaEfkAXyfVa5imeYA@mail.gmail.com' \
    --to=karthik.188@gmail.com \
    --cc=git@vger.kernel.org \
    --cc=hanwenn@gmail.com \
    --cc=jltobler@gmail.com \
    --cc=ps@pks.im \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
Code repositories for project(s) associated with this public inbox

	https://80x24.org/mirrors/git.git

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).