[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <20261111195904.GE5716@brightrain.aerifal.cx>
Date: Wed, 11 Nov 2026 14:59:05 -0500
From: Rich Felker <dalias@...c.org>
To: Markus Wichmann <nullplan@....net>
Cc: musl@...ts.openwall.com, Khem Raj <khem.raj@....qualcomm.com>
Subject: Re: [PATCH v2] math: enable FENV_ACCESS in expm1, expm1f,
hypot and hypotf
On Wed, Nov 11, 2026 at 08:46:49PM +0100, Markus Wichmann wrote:
> Hi all,
>
> I wonder if adding those pragmas (the OP also submitted a patch with an
> FP_CONTRACT pragma) to libm.h wouldn't just be more expedient. Or is
> there a reason to compile any maths code with FENV_ACCESS set to off?
I think I raised this issue a while back in response to the topic of
clang CFLAGS, and I'm split on which way we should do it. It's in some
sense cleaner to put the pragmas in a centralized place, but it's also
nice to have the core correct even if it's pulled outside of musl
without libm.h. At least files that don't use libm.h should have the
pragmas inline.
Rich
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.
