[PATCH v2] Saving a few cycles, part 1/3

Hi devs, I reworked the patch set according to the feedback I got here on this list. This is what changed in the first part: * split original part 1/2 into 1/3 (append char) and 2/3 (apply text delta). * introduced new svn_stringbuf_appendbyte function and undid changes to svn_stringbuf_append

[PATCH v2] Saving a few cycles, part 2/3

Hi devs, I reworked the patch set according to the feedback I got here on this list. This is what changed in the first part: * split original part 1/2 into 1/3 (append char) and 2/3 (apply text delta). * move memcpy optimization to a new fast_memcpy * add rationales to the commentary -- Stefan^

[PATCH v2] Saving a few cycles, part 3/3

Hi devs, I reworked the patch set according to the feedback I got here on this list. This is what changed in this last part: * add rationales to the commentary * bring svndiff.c changes closer to the original code * use svn_ctype_* functions in checksum parser -- Stefan^2. [[[ Various local op

what's up with r842347's log message?

The log on r842347 (aka r2273) looks wrong. What happened to it? With some sleuthing, I found the log message *should* look like: http://subversion.tigris.org/ds/viewMessage.do?dsForumId=495&dsMessageId=80079 So where/how did it get monkeyed? And even worse: what others may be clipped like that

Re: what's up with r842347's log message?

On Sat, May 01, 2010 at 01:45:53PM -0400, Greg Stein wrote: > The log on r842347 (aka r2273) looks wrong. What happened to it? What are you seeing? I'm getting: r842347 | cmpilato | 2002-06-18 22:14:56 +0200 (Tue, 18 Jun 20

Re: what's up with r842347's log message?

Greg Stein wrote on Sat, 1 May 2010 at 13:45 -0400: > With some sleuthing, I found the log message *should* look like: > > http://subversion.tigris.org/ds/viewMessage.do?dsForumId=495&dsMessageId=80079 > > So where/how did it get monkeyed? And even worse: what others may be > clipped like that?

RE: [PATCH v2] Saving a few cycles, part 3/3

> -Original Message- > From: Stefan Fuhrmann [mailto:stefanfuhrm...@alice-dsl.de] > Sent: zaterdag 1 mei 2010 19:03 > To: dev@subversion.apache.org > Subject: [PATCH v2] Saving a few cycles, part 3/3 > > Hi devs, > > I reworked the patch set according to the feedback I got > here on thi

Re: that harvest_committables() patch

Just to tell Philip that I'm pursuing this patch further now. ~Neels signature.asc Description: OpenPGP digital signature

Re: what's up with r842347's log message?

Eep! Local bug. It was trimming the log message :-( On Sat, May 1, 2010 at 13:49, Stefan Sperling wrote: > On Sat, May 01, 2010 at 01:45:53PM -0400, Greg Stein wrote: >> The log on r842347 (aka r2273) looks wrong. What happened to it? > > What are you seeing? > > I'm getting: > > --

Re: that harvest_committables() patch

Just to tell Philip r940102 and r940107. Still taking a look if I can move on a little. ~Neels Neels J Hofmeyr wrote: > Just to tell Philip that I'm pursuing this patch further now. > ~Neels > signature.asc Description: OpenPGP digital signature

Some failing tests

Hi devs, As I said in the mail thread about the misaligned blame output [1], I wanted to get the test suite running first. However, after spending some time with it, I still have 15 failing tests. I'd like that number to be 0 before I start hacking. Or is it known/expected that certain tests are c

more harvest_committables() -- was: that harvest_committables() patch

I have this patch in progress which tries to get rid of one {entry->deleted || (entry->schedule one of _delete, _replace)} use in harvest_committables(). It seems to pass now (complete make check still running, will verify tomorrow), but it definitely needs polish. Even if this patch passes make c