[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <218349c6.155f.1863fe35b5e.Coremail.00107082@163.com>
Date: Sat, 11 Feb 2023 17:51:50 +0800 (CST)
From: "David Wang" <00107082@....com>
To: musl@...ts.openwall.com
Subject: Re:Re: Re:Re: qsort
At 2023-02-11 17:22:49, "Markus Wichmann" <nullplan@....net> wrote:
>Also note that glibc further reduces the need for memcpy() by sorting
>indirectly if the size of a single element exceeds 32 bytes. That is of
>course something they can do, since they already are allocating memory
>and have a fall-back strategy in place. Not sure if this is worthwhile
>for musl. At least in the static linking case, it would suddenly pull
>malloc() and free() into executables that previously did not have those.
>
>That last optimization would also not have any bearing on the current
>batch of benchmarks, since in those, a size of four is set in stone.
>
My test data is just for demo, all data item is int32, definitely not meant to be benchmarking...
>Ciao,
>Markus
Powered by blists - more mailing lists
Confused about mailing lists and their use?
Read about mailing lists on Wikipedia
and check out these
guidelines on proper formatting of your messages.
