[LTP] [PATCH] nextafter01: Check exact results and convert to new API

[LTP] [PATCH] nextafter01: Check exact results and convert to new API

[Jan Stancek via ltp]

The test can hardly fail. Block 1 accepts nextafter(1.0, 1.1) whenever
(answer + 1.0) / 2 rounds to 1.0 as a float, which is also true when
nextafter() returns 1.0 unchanged. Block 2 never recomputes 'check'
and compares the stale value from block 1, which is exactly 1.0, so its
'check != 1.0' condition is always false and any return value passes.
Only the x == y case in block 3 is really verified. The file header also
names the test "scalb".

Compare the results with the exact next doubles instead: 1.0 + 2^-52
above and 1.0 - 2^-53 below. Both are exactly representable, so a plain
equality test is sufficient. Rewrite the test with the new API while at
it and fix the description.

Signed-off-by: Jan Stancek <jstancek@redhat.com>
---
 testcases/misc/math/nextafter/nextafter01.c | 145 +++++---------------
 1 file changed, 33 insertions(+), 112 deletions(-)

diff --git a/testcases/misc/math/nextafter/nextafter01.c b/testcases/misc/math/nextafter/nextafter01.c
index f4b212d4daa2..a74b7dc8e974 100644
--- a/testcases/misc/math/nextafter/nextafter01.c
+++ b/testcases/misc/math/nextafter/nextafter01.c
@@ -1,127 +1,48 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
 /*
- *
- *   Copyright (c) International Business Machines  Corp., 2002
- *
- *   This program is free software;  you can redistribute it and/or modify
- *   it under the terms of the GNU General Public License as published by
- *   the Free Software Foundation; either version 2 of the License, or
- *   (at your option) any later version.
- *
- *   This program is distributed in the hope that it will be useful,
- *   but WITHOUT ANY WARRANTY;  without even the implied warranty of
- *   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See
- *   the GNU General Public License for more details.
- *
- *   You should have received a copy of the GNU General Public License
- *   along with this program;  if not, write to the Free Software
- *   Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
+ * Copyright (c) International Business Machines Corp., 2002
+ * Copyright (c) 2026 Jan Stancek <jstancek@redhat.com>
  */
 
-/* 01/02/2003	Port to LTP	avenkat@us.ibm.com */
-/* 06/30/2001	Port to Linux	nsharoff@us.ibm.com */
-
-/*
- * NAME
- *      scalb
+/*\
+ * Verify that nextafter() returns the next representable double in the
+ * direction of its second argument:
  *
- * CALLS
- *      nextafter(3C)
- *
- * ALGORITHM
- *	Check results from the above functions against expected values.
- *
- * RESTRICTIONS
- * 	Checks for basic functionality, nothing fancy
+ * - nextafter(1.0, 1.1) is 1.0 + 2^-52 (DBL_EPSILON)
+ * - nextafter(1.0, 0.9) is 1.0 - 2^-53
+ * - nextafter(1.0, 1.0) is 1.0
  */
 
-#include <errno.h>
 #include <math.h>
-#include <stdio.h>
-#include <stdlib.h>
-#include "test.h"
+#include "tst_test.h"
 
-#define	FAILED 0
-#define	PASSED 1
+/* DBL_EPSILON spelled out, as sparse does not know __DBL_EPSILON__ */
+#define EPSILON (1.0 / (1ULL << 52))
 
-char *TCID = "nextafter01";
+static struct tcase {
+	double x;
+	double y;
+	double exp;
+} tcases[] = {
+	{1.0, 1.1, 1.0 + EPSILON},
+	{1.0, 0.9, 1.0 - EPSILON / 2},
+	{1.0, 1.0, 1.0},
+};
 
-int local_flag = PASSED;
-int block_number;
-FILE *temp;
-int TST_TOTAL = 1;
-
-void setup();
-void blenter();
-void blexit();
-
-/*--------------------------------------------------------------*/
-int main()
+static void run(unsigned int n)
 {
-	double answer;
-	double check;		/* tmp variable */
+	struct tcase *tc = &tcases[n];
+	double ret = nextafter(tc->x, tc->y);
 
-	setup();		/* temp file is now open */
-/*--------------------------------------------------------------*/
-	blenter();
-
-	answer = nextafter(1.0, 1.1);
-	check = (answer + 1.0) / 2;
-	if ((check != answer) && ((float)check != 1.0)) {
-		fprintf(temp, "nextafter returned %e, expected answer or 1.0\n",
-			answer);
-		local_flag = FAILED;
+	if (ret == tc->exp) {
+		tst_res(TPASS, "nextafter(%g, %g) == %.17g", tc->x, tc->y, ret);
+	} else {
+		tst_res(TFAIL, "nextafter(%g, %g) == %.17g, expected %.17g",
+			tc->x, tc->y, ret, tc->exp);
 	}
-
-	blexit();
-/*--------------------------------------------------------------*/
-	blenter();
-
-	answer = nextafter(1.0, 0.9);
-	if ((check != answer) && (check != 1.0)) {
-		fprintf(temp, "nextafter returned %e, expected answer or 1.0\n",
-			answer);
-		local_flag = FAILED;
-	}
-
-	blexit();
-/*--------------------------------------------------------------*/
-	blenter();
-
-	answer = nextafter(1.0, 1.0);
-	if (answer != 1.0) {
-		fprintf(temp, "nextafter 3 returned %e, expected 1.0\n",
-			answer);
-		local_flag = FAILED;
-	}
-
-	blexit();
-/*--------------------------------------------------------------*/
-
-	tst_exit();
 }
 
-/*--------------------------------------------------------------*/
-
-/*****	*****	LTP Port	*****/
-
-/* FUNCTIONS */
-
-void setup()
-{
-	temp = stderr;
-}
-
-void blenter()
-{
-	local_flag = PASSED;
-}
-
-void blexit()
-{
-	if (local_flag == PASSED)
-		tst_resm(TPASS, "Test passed");
-	else
-		tst_resm(TFAIL, "Test failed");
-}
-
-/*****	*****		*****/
+static struct tst_test test = {
+	.test = run,
+	.tcnt = ARRAY_SIZE(tcases),
+};
-- 
2.56.0


-- 
Mailing list info: https://lists.linux.it/listinfo/ltp

Re: [LTP] [PATCH] nextafter01: Check exact results and convert to new API

[Petr Vorel]

Hi Jan,

> The test can hardly fail. Block 1 accepts nextafter(1.0, 1.1) whenever
> (answer + 1.0) / 2 rounds to 1.0 as a float, which is also true when
> nextafter() returns 1.0 unchanged. Block 2 never recomputes 'check'
> and compares the stale value from block 1, which is exactly 1.0, so its
> 'check != 1.0' condition is always false and any return value passes.
> Only the x == y case in block 3 is really verified. The file header also
> names the test "scalb".

> Compare the results with the exact next doubles instead: 1.0 + 2^-52
> above and 1.0 - 2^-53 below. Both are exactly representable, so a plain
> equality test is sufficient. Rewrite the test with the new API while at
> it and fix the description.

Thanks!
Reviewed-by: Petr Vorel <pvorel@suse.cz>

...
> -char *TCID = "nextafter01";
> +static struct tcase {
> +	double x;
> +	double y;
> +	double exp;
> +} tcases[] = {
> +	{1.0, 1.1, 1.0 + EPSILON},
> +	{1.0, 0.9, 1.0 - EPSILON / 2},
> +	{1.0, 1.0, 1.0},
> +};

...
> +	if (ret == tc->exp) {
Hm, we still does not have TST_EXP_EQ_* variant for double.
Likely not worth of having it (not many tests comparing double).

Kind regards,
Petr

> +		tst_res(TPASS, "nextafter(%g, %g) == %.17g", tc->x, tc->y, ret);
> +	} else {
> +		tst_res(TFAIL, "nextafter(%g, %g) == %.17g, expected %.17g",
> +			tc->x, tc->y, ret, tc->exp);

-- 
Mailing list info: https://lists.linux.it/listinfo/ltp

Re: [LTP] [PATCH] nextafter01: Check exact results and convert to new API

[Andrea Cervesato via ltp]

Hi Petr,

> ...
> > +	if (ret == tc->exp) {
> Hm, we still does not have TST_EXP_EQ_* variant for double.
> Likely not worth of having it (not many tests comparing double).
> 
> Kind regards,
> Petr

actually I think it's a good idea, even if we don't have many examples
like this one. Maybe we can skip it for this time, keeping an eye on
patches to check for similar patterns.

--
Andrea Cervesato
SUSE QE Automation Engineer Linux
andrea.cervesato@suse.com

-- 
Mailing list info: https://lists.linux.it/listinfo/ltp

Re: [LTP] [PATCH] nextafter01: Check exact results and convert to new API

[Petr Vorel]

> Hi Petr,

> > ...
> > > +	if (ret == tc->exp) {
> > Hm, we still does not have TST_EXP_EQ_* variant for double.
> > Likely not worth of having it (not many tests comparing double).

> > Kind regards,
> > Petr

> actually I think it's a good idea, even if we don't have many examples
> like this one. Maybe we can skip it for this time, keeping an eye on
> patches to check for similar patterns.

+1, once we have it, people can start using it.

I suggest: Jan, feel free to merge as is and either send a separate patch to add
it or I send a patch to add it later (let's don't block a test rewrite unless
you have time to add TST_EXP_EQ_*).

Kind regards,
Petr

-- 
Mailing list info: https://lists.linux.it/listinfo/ltp

Re: [LTP] nextafter01: Check exact results and convert to new API

[linuxtestproject.agent]

Hi Jan,

On Mon, 19 Oct 2026 at 14:35:31 +0200, Jan Stancek via ltp wrote:
> nextafter01: Check exact results and convert to new API

Verdict - Reviewed

---
Note:

The agent can sometimes produce false positives although often its
findings are genuine. If you find issues with the review, please
comment this email or ignore the suggestions.

Regards,
LTP AI Reviewer

-- 
Mailing list info: https://lists.linux.it/listinfo/ltp

Re: [LTP] [PATCH] nextafter01: Check exact results and convert to new API

[Avinesh Kumar via ltp]

Hi,

> +/*\
> + * Verify that nextafter() returns the next representable double in the
nit:
:manpage:`nextafter(3)`

Reviewed-by: Avinesh Kumar <avinesh.kumar@suse.com>

Regards,
Avinesh


-- 
Mailing list info: https://lists.linux.it/listinfo/ltp