Re: 1.5 and 1.6 svnsync errors over serf

C. Michael Pilato wrote:
> Lieven Govaerts wrote:
>> Does the same thing happen with ra_neon? I don't really see any serf
>> specific file-opening happening.
>
> A quick grep for "unique" in the ra-serf sources reveals that serf commits
> use tempfiles for transmitting file data, set to delete on pool closure.
>

Ah, right.

I'd think that creating a temporary pool in
svnsync/replay.c:start_replay, using it for the editor drive and then
cleaning it up in end_replay should solve the problem.

Don't have time to work on it right now, but I can have a look this weekend.

Lieven

------------------------------------------------------
http://subversion.tigris.org/ds/viewMessage.do?dsForumId=462&dsMessageId=1273269
Received on 2009-03-05 20:01:02 CET