public inbox for libc-alpha@sourceware.org
 help / color / mirror / Atom feed
From: Florian Weimer <fweimer@redhat.com>
To: "Zack Weinberg" <zack@owlfolio.org>
Cc: "Adhemerval Zanella" <adhemerval.zanella@linaro.org>,
	 "Kuan-Wei Chiu" <visitorckw@gmail.com>,
	 "GNU libc development" <libc-alpha@sourceware.org>,
	 "Noah Goldstein" <goldstein.w.n@gmail.com>
Subject: Re: [PATCH] stdlib: Optimize number of calls to comparison function
Date: Sun, 17 Dec 2023 16:55:03 +0100	[thread overview]
Message-ID: <87a5q86cyw.fsf@oldenburg.str.redhat.com> (raw)
In-Reply-To: <e170150a-1f58-47ad-8be5-88d6eb393899@app.fastmail.com> (Zack Weinberg's message of "Sun, 17 Dec 2023 10:42:17 -0500")

* Zack Weinberg:

> It occurred to me late last night that *any* in-place sorting algorithm
> that operates on an array, can be made stable with no additional storage
> cost, by using each element's address as a tiebreaker for the comparison
> function.  It *doesn't* need to be the element's original address -- all
> that matters is that for every pair of elements that compare equal, the
> sorting algorithm preserves the < relation for their addresses across
> all swaps, which should happen naturally if address-< is used as a
> tiebreaker for comparisons.

It's not as simple as that because we might have A < B, so we can move A
to a lower address than B, but at the same time time, we must make sure
that we do not move it past an element C with A == C, otherwise the
result of the tie-breaking address comparison changes.  I haven't really
thought about it, but I suspect it's possible to do for some of the
quadratic algorithms, but quicksort and heapsort—not so much.

Thanks,
Florian


  reply	other threads:[~2023-12-17 15:55 UTC|newest]

Thread overview: 19+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-12-02 21:48 Kuan-Wei Chiu
2023-12-04  8:20 ` Florian Weimer
2023-12-04 18:31   ` Kuan-Wei Chiu
2023-12-05 10:44     ` Florian Weimer
2023-12-05 20:00       ` Kuan-Wei Chiu
2023-12-05 20:35     ` Zack Weinberg
2023-12-06 10:21       ` Florian Weimer
2023-12-06 12:51         ` Adhemerval Zanella Netto
2023-12-17 15:42           ` Zack Weinberg
2023-12-17 15:55             ` Florian Weimer [this message]
2023-12-17 16:47             ` Kuan-Wei Chiu
2023-12-17 18:02               ` Florian Weimer
2023-12-05  3:22   ` [PATCH v2 0/2] stdlib: Optimize number of calls to comparison function in qsort Kuan-Wei Chiu
2023-12-05  3:22     ` [PATCH v2 1/2] " Kuan-Wei Chiu
2023-12-05  3:22     ` [PATCH v2 2/2] stdlib: Adjust the factor in tst-qsort5 Kuan-Wei Chiu
2024-02-16  7:08 ` [PATCH] stdlib: Optimize number of calls to comparison function Kuan-Wei Chiu
2024-03-27 19:45   ` Adhemerval Zanella Netto
2024-03-27 19:59     ` Kuan-Wei Chiu
2024-03-27 20:42       ` Adhemerval Zanella Netto

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

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

  git send-email \
    --in-reply-to=87a5q86cyw.fsf@oldenburg.str.redhat.com \
    --to=fweimer@redhat.com \
    --cc=adhemerval.zanella@linaro.org \
    --cc=goldstein.w.n@gmail.com \
    --cc=libc-alpha@sourceware.org \
    --cc=visitorckw@gmail.com \
    --cc=zack@owlfolio.org \
    /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.
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).