* [PATCH v2] tst: Provide Y2038 tests for mktime (tst-mktime4.c)
@ 2021-01-27 12:47 Lukasz Majewski
2021-01-27 18:34 ` Paul Eggert
0 siblings, 1 reply; 2+ messages in thread
From: Lukasz Majewski @ 2021-01-27 12:47 UTC (permalink / raw)
To: Joseph Myers, Adhemerval Zanella, Florian Weimer
Cc: Paul Eggert, Alistair Francis, Arnd Bergmann, Alistair Francis,
GNU C Library, Carlos O'Donell, Florian Weimer,
Zack Weinberg, Lukasz Majewski
This change adds new test to assess mktime's functionality.
To be more specific - following use cases are checked:
- Pass struct tm as epoch time
- Pass struct tm as value just before Y2038 threshold (returned
value shall be 0x7FFFFFFF)
- Pass struct tm as the first value after Y2038 threshold
(expected value - 0x80000000)
---
Changes for v2:
- Remove FAIL_UNSUPPORTED() when sizeof (time_t) <= 4
- Remove tzset () as it is already done in mktime ()
- Use TEST_COMPARE to check if correct result is received
- Use "TZ=:" glibc extension to guarantee UTC time zone
- Add two more checks - prepare struct tm in a way that mktime will
return 0xFFFFFFFF and 0x100000000
---
time/Makefile | 4 +-
time/tst-mktime4.c | 100 +++++++++++++++++++++++++++++++++++++++++++++
2 files changed, 103 insertions(+), 1 deletion(-)
create mode 100644 time/tst-mktime4.c
diff --git a/time/Makefile b/time/Makefile
index 7de2ce0196..1c8678e46b 100644
--- a/time/Makefile
+++ b/time/Makefile
@@ -51,7 +51,7 @@ tests := test_time clocktest tst-posixtz tst-strptime tst_wcsftime \
tst-clock tst-clock2 tst-clock_nanosleep tst-cpuclock1 \
tst-adjtime tst-clock-y2038 tst-clock2-y2038 \
tst-cpuclock1-y2038 tst-clock_nanosleep-y2038 tst-clock_settime \
- tst-clock_adjtime tst-ctime tst-difftime
+ tst-clock_adjtime tst-ctime tst-difftime tst-mktime4
include ../Rules
@@ -88,4 +88,6 @@ test_time-ARGS= EST5EDT CST
tst-tzname-ENV = TZDIR=${common-objpfx}timezone/testdata
CPPFLAGS-tst-tzname.c += -DTZDEFRULES='"$(posixrules-file)"'
+CFLAGS-tst-mktime4.c += -Wno-overflow
+
bug-getdate1-ARGS = ${objpfx}bug-getdate1-fmt
diff --git a/time/tst-mktime4.c b/time/tst-mktime4.c
new file mode 100644
index 0000000000..811120721e
--- /dev/null
+++ b/time/tst-mktime4.c
@@ -0,0 +1,100 @@
+/* Test for mktime (4)
+ Copyright (C) 2021 Free Software Foundation, Inc.
+ This file is part of the GNU C Library.
+
+ The GNU C Library is free software; you can redistribute it and/or
+ modify it under the terms of the GNU Lesser General Public
+ License as published by the Free Software Foundation; either
+ version 2.1 of the License, or (at your option) any later version.
+
+ The GNU C Library 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
+ Lesser General Public License for more details.
+
+ You should have received a copy of the GNU Lesser General Public
+ License along with the GNU C Library; if not, see
+ <https://www.gnu.org/licenses/>. */
+
+#include <time.h>
+#include <stdlib.h>
+#include <support/check.h>
+
+struct tm tm0 = {
+ .tm_year = 70,
+ .tm_mon = 0,
+ .tm_mday = 1,
+ .tm_hour = 0,
+ .tm_min = 0,
+ .tm_sec = 0,
+ .tm_wday = 4,
+ .tm_yday = 0
+};
+
+struct tm tmY2038 = {
+ .tm_year = 138,
+ .tm_mon = 0,
+ .tm_mday = 19,
+ .tm_hour = 3,
+ .tm_min = 14,
+ .tm_sec = 7,
+ .tm_wday = 2,
+ .tm_yday = 18
+};
+
+struct tm tm1 = {
+ .tm_year = 206,
+ .tm_mon = 1,
+ .tm_mday = 7,
+ .tm_hour = 6,
+ .tm_min = 28,
+ .tm_sec = 15,
+ .tm_wday = 0,
+ .tm_yday = 0
+};
+
+static
+void test_mktime_helper (struct tm *tm, time_t exp_val)
+{
+ time_t result = mktime (tm);
+ if (result == -1)
+ FAIL_EXIT1 ("*** mktime failed: %m");
+
+ TEST_COMPARE (result, exp_val);
+}
+
+static int
+do_test (void)
+{
+ /* Use glibc time zone extension "TZ=:" to to guarantee that UTC
+ without leap seconds is used for the test. */
+ TEST_VERIFY_EXIT (setenv ("TZ", ":", 1) == 0);
+
+ /* Check that mktime(1970-01-01 00:00:00) returns 0. */
+ test_mktime_helper (&tm0, 0);
+
+ /* Check that mktime(2038-01-19 03:14:07) returns 0x7FFFFFFF. */
+ test_mktime_helper (&tmY2038, 0x7fffffff);
+
+ if (sizeof (time_t) > 4)
+ {
+ /* Check that mktime(2038-01-19 03:14:08) returns 0x80000000
+ (time_t overflow). */
+ tmY2038.tm_sec++;
+ test_mktime_helper (&tmY2038, 0x80000000);
+
+ /* Check that mktime(2106-02-07 06:28:15) returns 0xFFFFFFFF. */
+ test_mktime_helper (&tm1, 0xFFFFFFFF);
+
+ /* Check that mktime(2106-02-07 06:28:16) returns 0x100000000.
+
+ This test is executed with -Wno-overflow to prevent build
+ breaks when it is compiled on ports with 32 bit time_t. */
+ tm1.tm_sec++;
+ test_mktime_helper (&tm1, 0x100000000);
+ }
+
+ return 0;
+}
+
+#include <support/test-driver.c>
--
2.20.1
^ permalink raw reply [flat|nested] 2+ messages in thread
* Re: [PATCH v2] tst: Provide Y2038 tests for mktime (tst-mktime4.c)
2021-01-27 12:47 [PATCH v2] tst: Provide Y2038 tests for mktime (tst-mktime4.c) Lukasz Majewski
@ 2021-01-27 18:34 ` Paul Eggert
0 siblings, 0 replies; 2+ messages in thread
From: Paul Eggert @ 2021-01-27 18:34 UTC (permalink / raw)
To: Lukasz Majewski, Joseph Myers, Adhemerval Zanella, Florian Weimer
Cc: Alistair Francis, Arnd Bergmann, Alistair Francis, GNU C Library,
Carlos O'Donell, Florian Weimer, Zack Weinberg
On 1/27/21 4:47 AM, Lukasz Majewski wrote:
> +struct tm tm0 = {
> + .tm_year = 70,
> + .tm_mon = 0,
> + .tm_mday = 1,
> + .tm_hour = 0,
> + .tm_min = 0,
> + .tm_sec = 0,
> + .tm_wday = 4,
> + .tm_yday = 0
> +};
The tm_wday and tm_yday components here and elsewhere are irrelevant,
and should be omitted.
> + time_t result = mktime (tm);
> + if (result == -1)
'-1' should be '(time_t) -1'; for portability to hypothetical platforms
where time_t is unsigned and narrower than int.
> + /* Check that mktime(2038-01-19 03:14:08) returns 0x80000000
> + (time_t overflow). */
> + tmY2038.tm_sec++;
> + test_mktime_helper (&tmY2038, 0x80000000);
This sort of thing uses the output of the previous test as input to the
next test, which is not good practice. I suggest making tm0, tmY2038
const, and having test_mktime_helper copying the struct tm before
testing it.
> + /* Check that mktime(2106-02-07 06:28:16) returns 0x100000000.
> +
> + This test is executed with -Wno-overflow to prevent build
> + breaks when it is compiled on ports with 32 bit time_t. */
"compiled", not "executed".
Can't you avoid the problem by having the 2nd argument to
test_mktime_helper be long long instead of time_t?
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2021-01-27 18:34 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2021-01-27 12:47 [PATCH v2] tst: Provide Y2038 tests for mktime (tst-mktime4.c) Lukasz Majewski
2021-01-27 18:34 ` Paul Eggert
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).