public inbox for libc-alpha@sourceware.org
 help / color / mirror / Atom feed
From: Florian Weimer <fweimer@redhat.com>
To: Joseph Myers <joseph@codesourcery.com>
Cc: Letu Ren <fantasquex@gmail.com>,  <libc-alpha@sourceware.org>
Subject: Re: [PATCH v2] stdlib/strfrom: Change -NAN test to multiple possible results (bug 29501)
Date: Sun, 18 Sep 2022 10:31:14 +0200	[thread overview]
Message-ID: <87tu5584od.fsf@oldenburg.str.redhat.com> (raw)
In-Reply-To: <alpine.DEB.2.22.394.2209180529320.3209615@digraph.polyomino.org.uk> (Joseph Myers's message of "Sun, 18 Sep 2022 05:32:19 +0000")

* Joseph Myers:

> On Sat, 17 Sep 2022, Letu Ren via Libc-alpha wrote:
>
>> strfromf will convert the arguement with type float to double first.
>
> The *current implementation* will.  That's not part of the specification, 
> it's a quality-of-implementation bug (IEEE 754 leaves the presence of a 
> sign for a NaN in the results of convertToDecimalCharacter and 
> convertToHexCharacter optional, but glibc ought to be consistent here 
> between types and architectures).  That bug should be fixed; this patch 
> should not be applied.

I think the conversion to double is required because according to the
quote from ISO/IEC TS 18661-1:2014 in the bug, strfromf is defined in
terms of snprintf, and calling snprintf promotes float arguments that do
not correspond to a declared parameter to double as part of the default
argument promotions.

(Admittedly, this could be a bug in the TS.)

Thanks,
Florian


  reply	other threads:[~2022-09-18  8:31 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-09-17  6:13 Letu Ren
2022-09-18  5:32 ` Joseph Myers
2022-09-18  8:31   ` Florian Weimer [this message]
2022-09-18  8:54     ` Letu Ren
2022-09-18 16:25     ` Joseph Myers
2022-09-19  7:38       ` Florian Weimer
2022-09-21 10:36         ` Letu Ren
2022-09-21 16:34         ` Joseph Myers

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=87tu5584od.fsf@oldenburg.str.redhat.com \
    --to=fweimer@redhat.com \
    --cc=fantasquex@gmail.com \
    --cc=joseph@codesourcery.com \
    --cc=libc-alpha@sourceware.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).