On Wed, Jul 25, 2012 at 2:50 AM, Julian Foad wrote:
>
> Bert pointed out on IRC that changes like this one (and I have been making
> other similar changes for some time) could potentially have an adverse
> performance effect in some cases, because reparenting an RA session is
> not free. Our IRC
I (Julian Foad) wrote:
> I (Julian Foad) wrote:
> Using the attached 'reparenting-monitor.patch' and running
> merge_reintegrate_tests-10, I found that each of the merge
> commands executed in that test performs between 2 and 50 reparentings.
>
> DBG: ... sessions: 3, reparentings: 4 (+ no-ops
I (Julian Foad) wrote:
> Julian Foad wrote:
>> Daniel Shahaf wrote:
>>> So... should we revv ra_svn so 1.8 clients/servers can talk to each
>>> other exclusively in repos-root-relative paths?
>>
>> That sounds good to me, but I don't understand the authz impact.
>>
>> In much of the merge code, i
Julian Foad wrote:
> Daniel Shahaf wrote:
>> C. Michael Pilato wrote on Tue, Jul 24, 2012 at 21:31:49 -0400:
>>> For ra_svn: I was totally wrong. This thing always requires network
>>> activity: a "reparent" command/response at best; at worst, the complete
>>> teardown and re-opening of the s
Daniel Shahaf wrote:
>C. Michael Pilato wrote on Tue, Jul 24, 2012 at 21:31:49 -0400:
>> For ra_svn: I was totally wrong. This thing always requires network
>> activity: a "reparent" command/response at best; at worst, the complete
>> teardown and re-opening of the session. This is just a side-
C. Michael Pilato wrote on Tue, Jul 24, 2012 at 21:31:49 -0400:
> For ra_svn: I was totally wrong. This thing always requires network
> activity: a "reparent" command/response at best; at worst, the complete
> teardown and re-opening of the session. This is just a side-effect of the
> stateful p
Philip Martin wrote on Wed, Jul 25, 2012 at 12:06:48 +0100:
> Daniel Shahaf writes:
>
> > Philip Martin wrote on Wed, Jul 25, 2012 at 11:09:54 +0100:
> >> Daniel Shahaf writes:
> >>
> >> > Philip Martin wrote on Wed, Jul 25, 2012 at 10:49:36 +0100:
> >> >>
> >> >> I attached to the issue a rep
"Bert Huijben" writes:
>>SVN_ERR(svn_sqlite__step(&have_row, stmt2));
>>max = svn_sqlite__column_revnum(stmt2, 0);
>> - SVN_ERR(svn_fs_fs__revision_exists(max, fs, iterpool));
>> + if (SVN_IS_VALID_REVNUM(max)) /* The rep-cache could be empty. */
>> +SVN_ERR(svn
Daniel Shahaf writes:
> Philip Martin wrote on Wed, Jul 25, 2012 at 11:09:54 +0100:
>> Daniel Shahaf writes:
>>
>> > Philip Martin wrote on Wed, Jul 25, 2012 at 10:49:36 +0100:
>> >>
>> >> I attached to the issue a repository that demonstrates the corruption;
>> >> verify doesn't report a prob
> -Original Message-
> From: phi...@apache.org [mailto:phi...@apache.org]
> Sent: woensdag 25 juli 2012 13:01
> To: comm...@subversion.apache.org
> Subject: svn commit: r1365518 -
> /subversion/trunk/subversion/libsvn_fs_fs/rep-cache.c
>
> Author: philip
> Date: Wed Jul 25 11:01:26 2012
Philip Martin wrote on Wed, Jul 25, 2012 at 11:09:54 +0100:
> Daniel Shahaf writes:
>
> > Philip Martin wrote on Wed, Jul 25, 2012 at 10:49:36 +0100:
> >>
> >> I attached to the issue a repository that demonstrates the corruption;
> >> verify doesn't report a problem on that repository. What do
Philip Martin writes:
> This has nothing to do with the corrupt repository. I get the same
> error for:
>
> svnadmin create repo
> svn import -mm repo/format file://`pwd`/repo/f
> svnadmin verify repo
>
> I get a different error for:
>
> svnadmin create repo
> svn mkdir -mm file://`pwd`/repo/A
>
Philip Martin writes:
> $ valgrind -q subversion/svnadmin/.libs/lt-svnadmin verify repo
> * Verified revision 0.
> ==9089== Invalid read of size 1
> ==9089==at 0x4C25FF8: memcpy (mc_replace_strmem.c:497)
> ==9089==by 0x54E73F5: svn_stringbuf_appendbytes (string.c:558)
> ==9089==by 0x5
Daniel Shahaf writes:
> Philip Martin wrote on Wed, Jul 25, 2012 at 10:49:36 +0100:
>>
>> I attached to the issue a repository that demonstrates the corruption;
>> verify doesn't report a problem on that repository. What does verify
>> check?
>
> validate_root_noderev() catches an instance of t
Philip Martin wrote on Wed, Jul 25, 2012 at 10:49:36 +0100:
> Stefan Fuhrmann writes:
>
> > Could you please fix the UI experience first?
> > Currently, svnadmin verify will first verify the
> > root nodes of all revisions of the whole repository
> > and *then* start verifying all revisions showi
Stefan Fuhrmann writes:
> Could you please fix the UI experience first?
> Currently, svnadmin verify will first verify the
> root nodes of all revisions of the whole repository
> and *then* start verifying all revisions showing
> some progress info.
>
> Two issues with that:
> (1) The first part
Stefan Fuhrmann wrote on Wed, Jul 25, 2012 at 11:03:24 +0200:
> On Wed, Jul 25, 2012 at 9:53 AM, Daniel Shahaf wrote:
>
> > Should we mark the issue as FIXED ?
> >
>
> Could you please fix the UI experience first?
> Currently, svnadmin verify will first verify the
> root nodes of all revisions o
I think we definitely fixed one bug that could have explained this
corruption. (Specifically the valgrind bug which you isolated no the
ticket.)
Let's close the ticket then, and reopen it if someone can reproduce it
with ≥{1.6.18,1.7.5}.
Philip Martin wrote on Wed, Jul 25, 2012 at 10:10:03 +010
On Wed, Jul 25, 2012 at 11:14 AM, Philip Martin
wrote:
> Stefan Fuhrmann writes:
>
> > On Wed, Jul 25, 2012 at 9:53 AM, Daniel Shahaf
> wrote:
> >
> >> Should we mark the issue as FIXED ?
> >>
> >
> > Could you please fix the UI experience first?
> > Currently, svnadmin verify will first verify
Stefan Fuhrmann writes:
> On Wed, Jul 25, 2012 at 9:53 AM, Daniel Shahaf wrote:
>
>> Should we mark the issue as FIXED ?
>>
>
> Could you please fix the UI experience first?
> Currently, svnadmin verify will first verify the
> root nodes of all revisions of the whole repository
> and *then* star
I was going to ask you!
Daniel Shahaf writes:
> Should we mark the issue as FIXED ?
>
> phi...@tigris.org wrote on Tue, Jul 24, 2012 at 16:37:56 -0700:
>> http://subversion.tigris.org/issues/show_bug.cgi?id=4129
>>
>>
>>
>>
>>
>>
>> --- Additional comments from phi...@tigris.org Tue Ju
On Wed, Jul 25, 2012 at 9:53 AM, Daniel Shahaf wrote:
> Should we mark the issue as FIXED ?
>
Could you please fix the UI experience first?
Currently, svnadmin verify will first verify the
root nodes of all revisions of the whole repository
and *then* start verifying all revisions showing
some p
Should we mark the issue as FIXED ?
phi...@tigris.org wrote on Tue, Jul 24, 2012 at 16:37:56 -0700:
> http://subversion.tigris.org/issues/show_bug.cgi?id=4129
>
>
>
>
>
>
> --- Additional comments from phi...@tigris.org Tue Jul 24 16:37:55 -0700
> 2012 ---
> r1302613 was backported
23 matches
Mail list logo
