Yup. Sure looks that way. Dang.
I can't fix it right now tho... my svn builds are horked, and my
backup client isn't working either. Gotta fix a bug before I can do
anything :-P
(shouldn't take long tho...)
(or if you want to do it yourself, just copy that function over from
old-and-busted.c into entries.c; or expose it from old-and-busted?)
Cheers,
-g
On Tue, Feb 17, 2009 at 14:36, Hyrum K. Wright
<hyrum_wright_at_mail.utexas.edu> wrote:
> Excellent! I suspect that there are only a handful of problems which are
> causing the remainder of the tests to fail.
>
> In other news, I think you busted something on the branch:
>
> Undefined symbols:
> "_resolve_to_defaults", referenced from:
> _read_entries in libsvn_wc-1.a(entries.o)
> ld: symbol(s) not found
> collect2: ld returned 1 exit status
> make: *** [subversion/svnversion/svnversion] Error 1
> make: *** Waiting for unfinished jobs....
>
> -Hyrum
>
> On Feb 17, 2009, at 1:51 AM, Greg Stein wrote:
>
>> And with this change:
>>
>> 336 tests FAILED
>> 1 tests XPASSED
>>
>>
>> Quite a drop from the 407 failures!!
>>
>> Cheers,
>> -g
>>
>> On Tue, Feb 17, 2009 at 08:13, Greg Stein <gstein_at_gmail.com> wrote:
>>>
>>> Author: gstein
>>> Date: Mon Feb 16 23:13:58 2009
>>> New Revision: 35916
>>>
>>> Log:
>>> On the explore-wc branch:
>>>
>>> Fix some problems related to the propcaching removal.
>>>
>>> Note that r35796 was partially-reverted. We want to keep the old
>>> propcaching introduction stuff in lock.c.
>>>
>>> * subversion/libsvn_wc/props.c:
>>> (load_props): if we're loading WORKING props, then indicate when the
>>> props are not present by returning NULL for the hash.
>>> (svn_wc__load_props): load the base props if somebody asks for the
>>> working props (if there are no mods, then WORKING is copied from BASE
>>> props).
>>> (svn_wc_prop_list): restore previous call to svn_wc__load_props(),
>>> ignoring the overlay stuff. that was an attempt to compensate for the
>>> original bug in svn_wc__load_props.
>>> (svn_wc__has_props): also examine the BASE props. really old working
>>> copies always had WORKING, but newer ones may not.
>>> (svn_wc_props_modified_p): if there are no WORKING props, then there are
>>> definitely no mods. if there are some, then load 'em up and do a full
>>> examination.
>>>
>>> * subversion/libsvn_wc/lock.c:
>>> (introduce_propcaching): restore from r35796. add note that we need this
>>> dance in order to toss out WORKING props if they're unchanged.
>>> [ strictly speaking, we may not need to; safer for now tho ]
>>> (maybe_upgrade_format): call introduce_propcaching when needed.
>>>
>>> * subversion/libsvn_wc/old-and-busted.c:
>>> (read_entry): ignore all the propcaching fields that may be saved.
>>> (write_entry): write bupkis for the propcaching fields
>>> (write_entry_xml): don't write attributes for propcaching fields
>>>
>>> * subversion/libsvn_wc/entries.c:
>>> (svn_wc__atts_to_entry): add comment about the deprecated fields
>>> (svn_wc_entry_dup): fill in empty strings for the old propcach fields
>>>
>>> * subversion/libsvn_wc/log.c:
>>> (svn_wc__loggy_entry_modify): add comment about the deprecated fields
>>>
>>> Modified:
>>> branches/explore-wc/subversion/libsvn_wc/entries.c
>>> branches/explore-wc/subversion/libsvn_wc/lock.c
>>> branches/explore-wc/subversion/libsvn_wc/log.c
>>> branches/explore-wc/subversion/libsvn_wc/old-and-busted.c
>>> branches/explore-wc/subversion/libsvn_wc/props.c
>>>
>>> ...
>>
>> ------------------------------------------------------
>>
>> http://subversion.tigris.org/ds/viewMessage.do?dsForumId=462&dsMessageId=1177092
>
>
------------------------------------------------------
http://subversion.tigris.org/ds/viewMessage.do?dsForumId=462&dsMessageId=1179152
Received on 2009-02-17 15:33:25 CET