Re: CVS update: MODIFIED: libsvn_subr ...

Karl Fogel wrote:

>Branko �ibej <brane@xbc.nu> writes:
>
>>I've already solved the problem. APR now has apr_implode_gmt(), which
>>always returns a UTC apr_time_t, as long as the exploded time has the
>>correct tm_gmtoff value. We're using this API in whatever
>>svn_wc__string_to_time() has become. And I pounded APR into
>>calculating the correct tm_gmtoff on all platforms I can get my hands
>>on, with all TZ settings I could find defined -- see the comments in
>>apr/time/unix/time.c:get_offset().
>>
>
>By the way, my understanding of the above is that it doesn't matter
>that svn_time_to_string() uses apr_explode_localtime() while
>svn_time_from_string() uses apr_implode_gmt(). They don't have to
>match, because svn_time_from_string() is going to read the gmt offset
>from the exploded time and return a UTC apr_time_t no matter what.
>
Yes, that is correct. The whole purpose of apr_implode_gmt is to always
return a UTC apr_time_t.

I'm looking at svn_fs__dag_commit_txn right now, and the code to
calculate the timestamps looks O.K.

>Just a sanity check.
>
>Hmmm, then I wonder why the heck things weren't working as expected in
>fs-test.c...
>

Don't know.

Can you please do one thing: go to apr/test, `make testtimg &&
./testtime'. If it fails one of the apr_implode_* tests, make a note of
the difference it prints It may be something specific to your platform
-- FreeBSD, right? Although, all BSDs have tm_gmtoff; if that isn't set
correctly, all bets are off.

If it *doesn't* fail, fire up GDB and fix libsvn_fs, is all I can say. :-)

    Brane

-- 
Brane �ibej
    home:   <brane_at_xbc.nu>             http://www.xbc.nu/brane/
    work:   <branko.cibej_at_hermes.si>   http://www.hermes-softlab.com/
     ACM :   <brane_at_acm.org>            http://www.acm.org/
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org