public inbox for glibc-bugs@sourceware.org
help / color / mirror / Atom feed
* [Bug math/29193] New: sincos produces a different output than sin/cos
@ 2022-05-27 16:53 dangelog at gmail dot com
2022-05-27 19:33 ` [Bug math/29193] " schwab@linux-m68k.org
` (7 more replies)
0 siblings, 8 replies; 9+ messages in thread
From: dangelog at gmail dot com @ 2022-05-27 16:53 UTC (permalink / raw)
To: glibc-bugs
https://sourceware.org/bugzilla/show_bug.cgi?id=29193
Bug ID: 29193
Summary: sincos produces a different output than sin/cos
Product: glibc
Version: 2.31
Status: UNCONFIRMED
Severity: normal
Priority: P2
Component: math
Assignee: unassigned at sourceware dot org
Reporter: dangelog at gmail dot com
Target Milestone: ---
Hello,
Please consider this reduced testcase:
#define _GNU_SOURCE
#include <math.h>
#include <stdio.h>
double degreesToRadians(double degrees)
{
return degrees * (M_PI / 180);
}
double __attribute__ ((noinline)) a(double rad)
{
return sin(rad);
}
double __attribute__ ((noinline)) b(double rad)
{
double s;
double c;
sincos(rad, &s, &c);
if (c > 100)
return 0.0;
return s;
}
int main()
{
double d = 297;
double rad = degreesToRadians(d);
printf("%a\n%a\n", a(rad), b(rad));
}
Godbolt: https://c.godbolt.org/z/z45GGWEEK
Under GCC 12.1 -O2 -fno-builtin, the output is:
-0x1.c83201d3d2c6ep-1
-0x1.c83201d3d2c6dp-1
In other words, sincos seems to be introducing ~1ULP of error when compared to
calling sin/cos separately.
This is annoying, because GCC has a specific optimization where it replaces
nearby sin/cos calls with a call to sincos (slightly modified example here:
https://c.godbolt.org/z/rYEhrcPcx ). However, by doing so, it actually changes
the results. As sincos is not documented to be any different from sin+cos, the
optimization is correct, in principle. (If, on the other hand, sincos could
generate different results than sin/cos, then GCC must stop employing such
optimization without -ffast-math or similar flags).
--
You are receiving this mail because:
You are on the CC list for the bug.
^ permalink raw reply [flat|nested] 9+ messages in thread
* [Bug math/29193] sincos produces a different output than sin/cos
2022-05-27 16:53 [Bug math/29193] New: sincos produces a different output than sin/cos dangelog at gmail dot com
@ 2022-05-27 19:33 ` schwab@linux-m68k.org
2022-05-27 20:05 ` schwab@linux-m68k.org
` (6 subsequent siblings)
7 siblings, 0 replies; 9+ messages in thread
From: schwab@linux-m68k.org @ 2022-05-27 19:33 UTC (permalink / raw)
To: glibc-bugs
https://sourceware.org/bugzilla/show_bug.cgi?id=29193
--- Comment #1 from Andreas Schwab <schwab@linux-m68k.org> ---
Which specific optimized sin function are you using? With __sin_avx I cannot
reproduce that. Note that sincos always uses a non-optimized sin/cos
computation.
--
You are receiving this mail because:
You are on the CC list for the bug.
^ permalink raw reply [flat|nested] 9+ messages in thread
* [Bug math/29193] sincos produces a different output than sin/cos
2022-05-27 16:53 [Bug math/29193] New: sincos produces a different output than sin/cos dangelog at gmail dot com
2022-05-27 19:33 ` [Bug math/29193] " schwab@linux-m68k.org
@ 2022-05-27 20:05 ` schwab@linux-m68k.org
2022-05-27 22:50 ` dangelog at gmail dot com
` (5 subsequent siblings)
7 siblings, 0 replies; 9+ messages in thread
From: schwab@linux-m68k.org @ 2022-05-27 20:05 UTC (permalink / raw)
To: glibc-bugs
https://sourceware.org/bugzilla/show_bug.cgi?id=29193
--- Comment #2 from Andreas Schwab <schwab@linux-m68k.org> ---
I cannot reproduce that with __sin_sse2 either.
--
You are receiving this mail because:
You are on the CC list for the bug.
^ permalink raw reply [flat|nested] 9+ messages in thread
* [Bug math/29193] sincos produces a different output than sin/cos
2022-05-27 16:53 [Bug math/29193] New: sincos produces a different output than sin/cos dangelog at gmail dot com
2022-05-27 19:33 ` [Bug math/29193] " schwab@linux-m68k.org
2022-05-27 20:05 ` schwab@linux-m68k.org
@ 2022-05-27 22:50 ` dangelog at gmail dot com
2022-05-28 9:03 ` dangelog at gmail dot com
` (4 subsequent siblings)
7 siblings, 0 replies; 9+ messages in thread
From: dangelog at gmail dot com @ 2022-05-27 22:50 UTC (permalink / raw)
To: glibc-bugs
https://sourceware.org/bugzilla/show_bug.cgi?id=29193
--- Comment #3 from Giuseppe D'Angelo <dangelog at gmail dot com> ---
Hello,
I'm actually not sure; running gdb on the binary and stepping through
instructions leads me to __sin_fma:
(gdb) bt
#0 __sin_fma (x=5.1836278784231586) at ../sysdeps/ieee754/dbl-64/s_sin.c:202
#1 0x00005555555550d9 in main () at test.c:30
This is on Ubuntu 20.04 on a Broadwell CPU.
--
You are receiving this mail because:
You are on the CC list for the bug.
^ permalink raw reply [flat|nested] 9+ messages in thread
* [Bug math/29193] sincos produces a different output than sin/cos
2022-05-27 16:53 [Bug math/29193] New: sincos produces a different output than sin/cos dangelog at gmail dot com
` (2 preceding siblings ...)
2022-05-27 22:50 ` dangelog at gmail dot com
@ 2022-05-28 9:03 ` dangelog at gmail dot com
2022-05-31 11:37 ` dangelog at gmail dot com
` (3 subsequent siblings)
7 siblings, 0 replies; 9+ messages in thread
From: dangelog at gmail dot com @ 2022-05-28 9:03 UTC (permalink / raw)
To: glibc-bugs
https://sourceware.org/bugzilla/show_bug.cgi?id=29193
--- Comment #4 from Giuseppe D'Angelo <dangelog at gmail dot com> ---
As a data point, if I run the testcase under
GLIBC_TUNABLES=glibc.cpu.hwcaps=-AVX2_Usable
then I get identical results:
-0x1.c83201d3d2c6dp-1
-0x1.c83201d3d2c6dp-1
With that environment variable, I end up into __sin_avx.
--
You are receiving this mail because:
You are on the CC list for the bug.
^ permalink raw reply [flat|nested] 9+ messages in thread
* [Bug math/29193] sincos produces a different output than sin/cos
2022-05-27 16:53 [Bug math/29193] New: sincos produces a different output than sin/cos dangelog at gmail dot com
` (3 preceding siblings ...)
2022-05-28 9:03 ` dangelog at gmail dot com
@ 2022-05-31 11:37 ` dangelog at gmail dot com
2022-05-31 11:47 ` fweimer at redhat dot com
` (2 subsequent siblings)
7 siblings, 0 replies; 9+ messages in thread
From: dangelog at gmail dot com @ 2022-05-31 11:37 UTC (permalink / raw)
To: glibc-bugs
https://sourceware.org/bugzilla/show_bug.cgi?id=29193
--- Comment #5 from Giuseppe D'Angelo <dangelog at gmail dot com> ---
Hello,
In a nutshell, there are a few things crossing each other:
1) `__sin_fma` returns slightly different results than `__sin_avx` or
`__sin_sse2`. I don't think that this is a problem at all, but I might be
wrong, I just lack the necessary knowledge about what glibc specifically
promises here.
2) `sincos` seems to be consistent with `__sin_sse2`, and therefore
inconsistent with `__sin_fma`. Is this inconsistency OK? Should it be
documented?
3) In general, this opens the question: within the very same "environment"
(same process, CPU, libm, fpenv, etc.), is there any guarantee that `sin(x)`
always yields the very same result for identical values of `x`? Is `sin(x)`
meant to be perfectly reproducible? If yes, should GCC stop turning pairs of
`sin(x)` and `cos(x)` calls into one call to `sincos`?
--
You are receiving this mail because:
You are on the CC list for the bug.
^ permalink raw reply [flat|nested] 9+ messages in thread
* [Bug math/29193] sincos produces a different output than sin/cos
2022-05-27 16:53 [Bug math/29193] New: sincos produces a different output than sin/cos dangelog at gmail dot com
` (4 preceding siblings ...)
2022-05-31 11:37 ` dangelog at gmail dot com
@ 2022-05-31 11:47 ` fweimer at redhat dot com
2022-06-01 8:32 ` schwab@linux-m68k.org
2022-06-02 9:29 ` dangelog at gmail dot com
7 siblings, 0 replies; 9+ messages in thread
From: fweimer at redhat dot com @ 2022-05-31 11:47 UTC (permalink / raw)
To: glibc-bugs
https://sourceware.org/bugzilla/show_bug.cgi?id=29193
Florian Weimer <fweimer at redhat dot com> changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |fweimer at redhat dot com
--
You are receiving this mail because:
You are on the CC list for the bug.
^ permalink raw reply [flat|nested] 9+ messages in thread
* [Bug math/29193] sincos produces a different output than sin/cos
2022-05-27 16:53 [Bug math/29193] New: sincos produces a different output than sin/cos dangelog at gmail dot com
` (5 preceding siblings ...)
2022-05-31 11:47 ` fweimer at redhat dot com
@ 2022-06-01 8:32 ` schwab@linux-m68k.org
2022-06-02 9:29 ` dangelog at gmail dot com
7 siblings, 0 replies; 9+ messages in thread
From: schwab@linux-m68k.org @ 2022-06-01 8:32 UTC (permalink / raw)
To: glibc-bugs
https://sourceware.org/bugzilla/show_bug.cgi?id=29193
Andreas Schwab <schwab@linux-m68k.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
Resolution|--- |FIXED
Status|UNCONFIRMED |RESOLVED
Target Milestone|--- |2.36
--- Comment #6 from Andreas Schwab <schwab@linux-m68k.org> ---
Fixed in 2.36.
--
You are receiving this mail because:
You are on the CC list for the bug.
^ permalink raw reply [flat|nested] 9+ messages in thread
* [Bug math/29193] sincos produces a different output than sin/cos
2022-05-27 16:53 [Bug math/29193] New: sincos produces a different output than sin/cos dangelog at gmail dot com
` (6 preceding siblings ...)
2022-06-01 8:32 ` schwab@linux-m68k.org
@ 2022-06-02 9:29 ` dangelog at gmail dot com
7 siblings, 0 replies; 9+ messages in thread
From: dangelog at gmail dot com @ 2022-06-02 9:29 UTC (permalink / raw)
To: glibc-bugs
https://sourceware.org/bugzilla/show_bug.cgi?id=29193
--- Comment #7 from Giuseppe D'Angelo <dangelog at gmail dot com> ---
Thank you very much!
--
You are receiving this mail because:
You are on the CC list for the bug.
^ permalink raw reply [flat|nested] 9+ messages in thread
end of thread, other threads:[~2022-06-02 9:29 UTC | newest]
Thread overview: 9+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2022-05-27 16:53 [Bug math/29193] New: sincos produces a different output than sin/cos dangelog at gmail dot com
2022-05-27 19:33 ` [Bug math/29193] " schwab@linux-m68k.org
2022-05-27 20:05 ` schwab@linux-m68k.org
2022-05-27 22:50 ` dangelog at gmail dot com
2022-05-28 9:03 ` dangelog at gmail dot com
2022-05-31 11:37 ` dangelog at gmail dot com
2022-05-31 11:47 ` fweimer at redhat dot com
2022-06-01 8:32 ` schwab@linux-m68k.org
2022-06-02 9:29 ` dangelog at gmail dot com
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).