From: Olivier Marin <dkr@xxxxxxxxxxx> Commit af66366a9feb0194ed04b1f538998021ece268a8 introduced the keyword "never" to be used with approxidate() but defined it with a fixed date without taking care of timezone. As a result approxidate() will return a timestamp in the future with a negative timezone. With this patch, approxidate("never") always return 0 whatever your timezone is. Also, print seconds since the Epoch in test-date. Signed-off-by: Olivier Marin <dkr@xxxxxxxxxxx> --- Without this patch: $ git --version git version 1.5.6.rc3.156.g1a01 (next) $ mkdir /tmp/repo1 && cd /tmp/repo1 && git init $ echo A > A && git add A && git commit -m A $ git config gc.reflogexpire never $ TZ=UTC git reflog expire --dry-run --verbose HEAD keep commit (initial): A $ TZ=UTC-2 git reflog expire --dry-run --verbose HEAD would prune commit (initial): A This test does not trigger the problem with Junio's "Per-ref reflog expiry configuration" patch because it explicitly tests for "never" without relying on approxidate() but the bug still remains with --expire option and in date.c This make me thinking about TZ=UTC forced into t/test-lib.sh. Should not test cases to be independant of the timezone? date.c | 6 ++---- test-date.c | 2 +- 2 files changed, 3 insertions(+), 5 deletions(-) diff --git a/date.c b/date.c index a74ed86..1a4eb87 100644 --- a/date.c +++ b/date.c @@ -682,10 +682,8 @@ static void date_am(struct tm *tm, int *num) static void date_never(struct tm *tm, int *num) { - tm->tm_mon = tm->tm_wday = tm->tm_yday - = tm->tm_hour = tm->tm_min = tm->tm_sec = 0; - tm->tm_year = 70; - tm->tm_mday = 1; + time_t n = 0; + localtime_r(&n, tm); } static const struct special { diff --git a/test-date.c b/test-date.c index 62e8f23..b84e000 100644 --- a/test-date.c +++ b/test-date.c @@ -14,7 +14,7 @@ int main(int argc, char **argv) printf("%s -> %s -> %s", argv[i], result, ctime(&t)); t = approxidate(argv[i]); - printf("%s -> %s\n", argv[i], ctime(&t)); + printf("%s -> %lu -> %s\n", argv[i], t, ctime(&t)); } return 0; } -- 1.5.6.rc3.204.gf01a.dirty -- To unsubscribe from this list: send the line "unsubscribe git" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html