From: Aldy Hernandez <aldyh@redhat.com>
To: Jakub Jelinek <jakub@redhat.com>
Cc: GCC patches <gcc-patches@gcc.gnu.org>,
Andrew MacLeod <amacleod@redhat.com>
Subject: Re: [PATCH] [PR24021] Implement PLUS_EXPR range-op entry for floats.
Date: Mon, 31 Oct 2022 09:42:38 +0100 [thread overview]
Message-ID: <CAGm3qMWJvn5zU0hvS5+XZNR3C39NnevktHsG4a38LBey7s-7XQ@mail.gmail.com> (raw)
In-Reply-To: <CAGm3qMXK5CiauPUhOKo0yW8CCUXnq=4xhb15LNMxvyzzSEwZYg@mail.gmail.com>
[-- Attachment #1: Type: text/plain, Size: 5733 bytes --]
Ping ping
On Mon, Oct 24, 2022, 08:04 Aldy Hernandez <aldyh@redhat.com> wrote:
> PING
>
> On Mon, Oct 17, 2022 at 8:21 AM Aldy Hernandez <aldyh@redhat.com> wrote:
> >
> > On Thu, Oct 13, 2022 at 7:57 PM Jakub Jelinek <jakub@redhat.com> wrote:
> > >
> > > On Thu, Oct 13, 2022 at 02:36:49PM +0200, Aldy Hernandez wrote:
> > > > +// Like real_arithmetic, but round the result to INF if the
> operation
> > > > +// produced inexact results.
> > > > +//
> > > > +// ?? There is still one problematic case, i387. With
> > > > +// -fexcess-precision=standard we perform most SF/DFmode arithmetic
> in
> > > > +// XFmode (long_double_type_node), so that case is OK. But without
> > > > +// -mfpmath=sse, all the SF/DFmode computations are in XFmode
> > > > +// precision (64-bit mantissa) and only occassionally rounded to
> > > > +// SF/DFmode (when storing into memory from the 387 stack). Maybe
> > > > +// this is ok as well though it is just occassionally more precise.
> ??
> > > > +
> > > > +static void
> > > > +frange_arithmetic (enum tree_code code, tree type,
> > > > + REAL_VALUE_TYPE &result,
> > > > + const REAL_VALUE_TYPE &op1,
> > > > + const REAL_VALUE_TYPE &op2,
> > > > + const REAL_VALUE_TYPE &inf)
> > > > +{
> > > > + REAL_VALUE_TYPE value;
> > > > + enum machine_mode mode = TYPE_MODE (type);
> > > > + bool mode_composite = MODE_COMPOSITE_P (mode);
> > > > +
> > > > + bool inexact = real_arithmetic (&value, code, &op1, &op2);
> > > > + real_convert (&result, mode, &value);
> > > > +
> > > > + // If real_convert above has rounded an inexact value to towards
> > > > + // inf, we can keep the result as is, otherwise we'll adjust by 1
> ulp
> > > > + // later (real_nextafter).
> > > > + bool rounding = (flag_rounding_math
> > > > + && (real_isneg (&inf)
> > > > + ? real_less (&result, &value)
> > > > + : !real_less (&value, &result)));
> > >
> > > I thought the agreement during Cauldron was that we'd do this always,
> > > regardless of flag_rounding_math.
> > > Because excess precision (the fast one like on ia32 or -mfpmath=387 on
> > > x86_64), or -frounding-math, or FMA contraction can all increase
> precision
> > > and worst case it all behaves like -frounding-math for the ranges.
> > >
> > > So, perhaps use:
> > > if ((mode_composite || (real_isneg (&inf) ? real_less (&result,
> &value)
> > > : !real_less (&value,
> &result))
> > > && (inexact || !real_identical (&result, &value))))
> >
> > Done.
> >
> > > ?
> > > No need to do the real_isneg/real_less stuff for mode_composite, then
> > > we do it always for inexacts, but otherwise we check if the rounding
> > > performed by real.cc has been in the conservative direction (for upper
> > > bound to +inf, for lower bound to -inf), if yes, we don't need to do
> > > anything, if yes, we frange_nextafter.
> > >
> > > As discussed, for mode_composite, I think we want to do the extra
> > > stuff for inexact denormals and otherwise do the nextafter
> unconditionally,
> > > because our internal mode_composite representation isn't precise
> enough.
> > >
> > > > + // Be extra careful if there may be discrepancies between the
> > > > + // compile and runtime results.
> > > > + if ((rounding || mode_composite)
> > > > + && (inexact || !real_identical (&result, &value)))
> > > > + {
> > > > + if (mode_composite)
> > > > + {
> > > > + bool denormal = (result.sig[SIGSZ-1] & SIG_MSB) == 0;
> > >
> > > Use real_isdenormal here?
> >
> > Done.
> >
> > > Though, real_iszero needs the same thing.
> >
> > So... real_isdenormal() || real_iszero() as in the attached patch?
> >
> > >
> > > > + if (denormal)
> > > > + {
> > > > + REAL_VALUE_TYPE tmp;
> > >
> > > And explain here why is this, that IBM extended denormals have just
> > > DFmode precision.
> >
> > Done.
> >
> > > Though, now that I think about it, while this is correct for denormals,
> > >
> > > > + real_convert (&tmp, DFmode, &value);
> > > > + frange_nextafter (DFmode, tmp, inf);
> > > > + real_convert (&result, mode, &tmp);
> > > > + }
> > >
> > > there are also the cases where the higher double exponent is in the
> > > [__DBL_MIN_EXP__, __LDBL_MIN_EXP__] aka [-1021, -968] or so.
> > > https://en.wikipedia.org/wiki/Double-precision_floating-point_format
> > > If the upper double is denormal in the DFmode sense, so smaller
> absolute
> > > value than __DBL_MIN__, then doing nextafter in DFmode is the right
> thing to
> > > do, the lower double must be always +/- zero.
> > > Now, if the result is __DBL_MIN__, the upper double is already
> normalized
> > > but we can add __DBL_DENORM_MIN__ to it, which will make the number
> have
> > > 54-bit precision.
> > > If the result is __DBL_MIN__ * 2, we can again add __DBL_DENORM_MIN__
> > > and make it 55-bit precision. Etc. until we reach __DBL_MIN__ * 2e53
> > > where it acts like fully normalized 106-bit precision number.
> > > I must say I'm not really sure what real_nextafter is doing in those
> cases,
> > > I'm afraid it doesn't handle it correctly but the only other use
> > > of real_nextafter is guarded with:
> > > /* Don't handle composite modes, nor decimal, nor modes without
> > > inf or denorm at least for now. */
> > > if (format->pnan < format->p
> > > || format->b == 10
> > > || !format->has_inf
> > > || !format->has_denorm)
> > > return false;
> >
> > Dunno. Is there a conservative thing we can do for mode_composites
> > that aren't denormal or zero?
> >
> > How does this look?
> > Aldy
>
next prev parent reply other threads:[~2022-10-31 8:42 UTC|newest]
Thread overview: 37+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-10-13 12:36 Aldy Hernandez
2022-10-13 13:02 ` Toon Moene
2022-10-13 13:44 ` Aldy Hernandez
2022-10-13 13:52 ` Toon Moene
2022-10-14 8:04 ` Aldy Hernandez
2022-10-13 17:57 ` Jakub Jelinek
2022-10-17 6:21 ` Aldy Hernandez
2022-10-24 6:04 ` Aldy Hernandez
2022-10-29 4:55 ` Jeff Law
2022-10-31 8:42 ` Aldy Hernandez [this message]
2022-11-04 13:16 ` Jakub Jelinek
2022-11-04 19:14 ` Jakub Jelinek
2022-11-04 19:53 ` Jakub Jelinek
2022-11-07 12:35 ` Aldy Hernandez
2022-11-07 12:43 ` Jakub Jelinek
2022-11-07 12:48 ` Aldy Hernandez
2022-11-07 12:56 ` Jakub Jelinek
2022-11-07 15:38 ` Aldy Hernandez
2022-11-08 11:07 ` Jakub Jelinek
2022-11-08 12:47 ` Aldy Hernandez
2022-11-08 13:15 ` Jakub Jelinek
2022-11-08 14:02 ` Aldy Hernandez
2022-11-08 14:03 ` Jakub Jelinek
2022-11-07 15:41 ` Aldy Hernandez
2022-11-08 11:20 ` Jakub Jelinek
2022-11-08 13:06 ` Aldy Hernandez
2022-11-08 13:24 ` Jakub Jelinek
2022-11-08 13:47 ` Aldy Hernandez
2022-11-08 13:50 ` Jakub Jelinek
2022-11-08 14:06 ` Aldy Hernandez
2022-11-08 14:11 ` Jakub Jelinek
2022-11-08 14:14 ` Aldy Hernandez
2022-11-08 23:05 ` Aldy Hernandez
2022-11-09 6:59 ` Aldy Hernandez
2022-11-08 17:44 ` Andrew Waterman
2022-11-08 18:11 ` Jakub Jelinek
2022-11-08 18:17 ` Andrew Waterman
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=CAGm3qMWJvn5zU0hvS5+XZNR3C39NnevktHsG4a38LBey7s-7XQ@mail.gmail.com \
--to=aldyh@redhat.com \
--cc=amacleod@redhat.com \
--cc=gcc-patches@gcc.gnu.org \
--cc=jakub@redhat.com \
/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).