Daniel Shahaf <d...@daniel.shahaf.name> writes: > phi...@apache.org wrote on Mon, Oct 15, 2012 at 22:20:32 -0000: >> Author: philip >> Date: Mon Oct 15 22:20:32 2012 >> New Revision: 1398541 >> >> URL: http://svn.apache.org/viewvc?rev=1398541&view=rev >> Log: >> * STATUS: Clarify effect of r1237779. >> >> Modified: >> subversion/branches/1.7.x/STATUS >> >> Modified: subversion/branches/1.7.x/STATUS >> URL: >> http://svn.apache.org/viewvc/subversion/branches/1.7.x/STATUS?rev=1398541&r1=1398540&r2=1398541&view=diff >> ============================================================================== >> --- subversion/branches/1.7.x/STATUS (original) >> +++ subversion/branches/1.7.x/STATUS Mon Oct 15 22:20:32 2012 >> @@ -71,7 +71,7 @@ Candidate changes: >> +1: cmpilato >> >> * r1237779 >> - Fix sense of property validation on load. >> + Fix sense of property validation on load and make the revprop change >> atomic. > > The code uses either svn_repos_fs_change_rev_prop4() or > svn_fs_change_rev_prop2(). Both of them support atomic revprop changes, > but for both of them the code passes NULL to disable that functinoality. > > IOW, that revision doesn't change atomicity, AFAICT.
svn_fs_change_rev_prop4 retrieves the old value automatically if NULL is passed so the non-validating code is atomic. I did see some WANdisco tests fail but I also had to read the code to make sure. -- Join us this October at Subversion Live 2012 http://www.wandisco.com/svn-live-2012
