Re: strftime test on Solaris based OS

[Paul Eggert]

Bruno Haible wrote:
> On Dyson OS, a Solaris based OS with Debian packaging system [1],
> the gnulib 'strftime' test fails:
>
>
> FAIL: test-strftime
> ===================
>
> NZST-12NZDT,M9.5.0,M4.1.0/3: expected "1970-01-01 13:00:00 +1300 (NZDT)", got 
> "1970-01-01 13:00:00 +1300 (NZST)"
> NZST-12NZDT,M9.5.0,M4.1.0/3: expected "1985-11-05 13:53:21 +1300 (NZDT)", got 
> "1985-11-05 13:53:21 +1300 (NZST)"
> PST8PDT,M3.2.0,M11.1.0: expected "2001-09-08 18:46:42 -0700 (PDT)", got "2001-09-08 
> 18:46:42 -0700 (PST)"
> CET-1CEST,M3.5.0,M10.5.0/3: expected "2001-09-09 03:46:42 +0200 (CEST)", got 
> "2001-09-09 03:46:42 +0200 (CET)"
> FAIL test-strftime (exit status: 1)
>
>
> Any ideas? Paul?

Apparently strftime %Z is not working even though %z works. I assume 
HAVE_TM_ZONE is not defined and that HAVE_TZNAME is defined (that's how things 
should be on Solaris; you can check this in config.h). On the troublesome cases, 
__strftime_internal should set zone=NULL, then tzname_vec = tz->tzname_copy, 
then zone = tzname_vec[tp->tm_isdst != 0]; since tp->tm_isdst should be 1, this 
should set zone = tz->tzname_copy[1]. Since 'zone' is wrong, I suggest running 
test-strftime under GDB and investigating whether 'zone' is wrong because 
tz->tzname_copy[1] is wrong, or because tp->tm_isdst is wrong.

For what it's worth I can't reproduce the bug on either Solaris 10 sparc, or on 
Solaris 11 x86-64. I tried 32- and 64-bit builds on both platforms. I couldn't 
download Dyson OS and to be honest would rather not fight with installing it.