Re: svn commit: r1508170 - in /subversion/trunk:

2013-08-01 Thread Daniel Shahaf
Stefan Fuhrmann wrote on Thu, Aug 01, 2013 at 12:31:27 +0200: > On Thu, Aug 1, 2013 at 6:36 AM, Daniel Shahaf wrote: > > #define svn_hash_sets(ht, key, val) \ > > apr_hash_set(ht, key, APR_HASH_KEY_STRING, val) > > > > I would keep the current code for that. I allows for the > compile

Re: svn commit: r1508170 - in /subversion/trunk:

2013-08-01 Thread Stefan Fuhrmann
On Thu, Aug 1, 2013 at 12:51 PM, Daniel Shahaf wrote: > Stefan Fuhrmann wrote on Thu, Aug 01, 2013 at 12:27:08 +0200: > > On Thu, Aug 1, 2013 at 4:47 AM, Daniel Shahaf wrote: > > > > > Stefan Fuhrmann wrote on Wed, Jul 31, 2013 at 23:21:45 +0200: > > > > * revert svn_hash_gets to simply use APR_

Re: svn commit: r1508170 - in /subversion/trunk:

2013-08-01 Thread Daniel Shahaf
Stefan Fuhrmann wrote on Wed, Jul 31, 2013 at 23:21:45 +0200: > * revert svn_hash_gets to simply use APR_HASH_KEY_STRING Went ahead and did this in r1509166 + r1509168.

Re: svn commit: r1508170 - in /subversion/trunk:

2013-08-01 Thread Daniel Shahaf
Stefan Fuhrmann wrote on Thu, Aug 01, 2013 at 12:27:08 +0200: > On Thu, Aug 1, 2013 at 4:47 AM, Daniel Shahaf wrote: > > > Stefan Fuhrmann wrote on Wed, Jul 31, 2013 at 23:21:45 +0200: > > > * revert svn_hash_gets to simply use APR_HASH_KEY_STRING > > > > ... > > > > > * where it is being used, m

Re: svn commit: r1508170 - in /subversion/trunk:

2013-08-01 Thread Stefan Fuhrmann
On Thu, Aug 1, 2013 at 6:36 AM, Daniel Shahaf wrote: > Daniel Shahaf wrote on Thu, Aug 01, 2013 at 05:47:58 +0300: > > Stefan Fuhrmann wrote on Wed, Jul 31, 2013 at 23:21:45 +0200: > > > * revert svn_hash_gets to simply use APR_HASH_KEY_STRING > > > > ... > > > > > * where it is being used, make

Re: svn commit: r1508170 - in /subversion/trunk:

2013-08-01 Thread Stefan Fuhrmann
On Thu, Aug 1, 2013 at 4:47 AM, Daniel Shahaf wrote: > Stefan Fuhrmann wrote on Wed, Jul 31, 2013 at 23:21:45 +0200: > > * revert svn_hash_gets to simply use APR_HASH_KEY_STRING > > ... > > > * where it is being used, make svn_private_config.h the first #include > > Why would the order of include

Re: svn commit: r1508170 - in /subversion/trunk:

2013-07-31 Thread Daniel Shahaf
Daniel Shahaf wrote on Thu, Aug 01, 2013 at 05:47:58 +0300: > Stefan Fuhrmann wrote on Wed, Jul 31, 2013 at 23:21:45 +0200: > > * revert svn_hash_gets to simply use APR_HASH_KEY_STRING > > ... > > > * where it is being used, make svn_private_config.h the first #include > > Why would the order of

Re: svn commit: r1508170 - in /subversion/trunk:

2013-07-31 Thread Greg Stein
On Jul 31, 2013 5:54 PM, "Branko Čibej" wrote: > > On 31.07.2013 23:21, Stefan Fuhrmann wrote: > > * svn_hash_gets / svn_hash_sets have been introduced to > > simplify coding. Being used in 600+ places, this surely is > > a useful API. We should have kept it private, though. > > Given that 1.8

Re: svn commit: r1508170 - in /subversion/trunk:

2013-07-31 Thread Daniel Shahaf
Stefan Fuhrmann wrote on Wed, Jul 31, 2013 at 23:21:45 +0200: > * revert svn_hash_gets to simply use APR_HASH_KEY_STRING ... > * where it is being used, make svn_private_config.h the first #include Why would the order of includes matter? Do you plan to still use SVN_HAS_DUNDER_BUILTINS in svn_h

Re: svn commit: r1508170 - in /subversion/trunk:

2013-07-31 Thread Branko Čibej
On 31.07.2013 23:21, Stefan Fuhrmann wrote: > * svn_hash_gets / svn_hash_sets have been introduced to > simplify coding. Being used in 600+ places, this surely is > a useful API. We should have kept it private, though. Given that 1.8 is a month old, can't we just declare that making this API p

Re: svn commit: r1508170 - in /subversion/trunk:

2013-07-31 Thread Branko Čibej
On 31.07.2013 23:21, Stefan Fuhrmann wrote: > On Tue, Jul 30, 2013 at 9:56 PM, Greg Stein > wrote: > > It really doesn't make that much of a difference. APR_HASH_KEY_STRING > simply tells the hash function to watch for NUL rather than to > traverse N characters

Re: svn commit: r1508170 - in /subversion/trunk:

2013-07-31 Thread Stefan Fuhrmann
On Tue, Jul 30, 2013 at 9:56 PM, Greg Stein wrote: > It really doesn't make that much of a difference. APR_HASH_KEY_STRING > simply tells the hash function to watch for NUL rather than to > traverse N characters. Because N doesn't have to be tracked, you might > even argue the NUL-test is faster.

Re: svn commit: r1508170 - in /subversion/trunk:

2013-07-30 Thread Greg Stein
On Tue, Jul 30, 2013 at 8:02 PM, Daniel Shahaf wrote: >... > svn_hash__make uses a hash function which does a strlen(). And I recall a lot of discussion when that function was introduced. The current docstring for it in svn_subr_private.h provides ZERO information on why/when it should be used.

Re: svn commit: r1508170 - in /subversion/trunk:

2013-07-30 Thread Daniel Shahaf
On Tue, Jul 30, 2013 at 03:56:02PM -0400, Greg Stein wrote: > It really doesn't make that much of a difference. APR_HASH_KEY_STRING > simply tells the hash function to watch for NUL rather than to > traverse N characters. Because N doesn't have to be tracked, you might > even argue the NUL-test is

Re: svn commit: r1508170 - in /subversion/trunk:

2013-07-30 Thread Ben Reser
On Tue, Jul 30, 2013 at 4:26 PM, Branko Čibej wrote: > My vote is to revert the whole shebang, possibly up to and including the > introduction of svn_hash_sets/gets in the first place. Though that's in > 1.8 now I believe... Yes it's in 1.8. We should simply revert back to what was in 1.8. Greg'

Re: svn commit: r1508170 - in /subversion/trunk:

2013-07-30 Thread Daniel Shahaf
On Wed, Jul 31, 2013 at 01:26:16AM +0200, Branko Čibej wrote: > On 30.07.2013 21:56, Greg Stein wrote: > > It really doesn't make that much of a difference. APR_HASH_KEY_STRING > > simply tells the hash function to watch for NUL rather than to > > traverse N characters. Because N doesn't have to be

Re: svn commit: r1508170 - in /subversion/trunk:

2013-07-30 Thread Branko Čibej
On 30.07.2013 21:56, Greg Stein wrote: > It really doesn't make that much of a difference. APR_HASH_KEY_STRING > simply tells the hash function to watch for NUL rather than to > traverse N characters. Because N doesn't have to be tracked, you might > even argue the NUL-test is faster. > > IOW, pass

Re: svn commit: r1508170 - in /subversion/trunk:

2013-07-30 Thread Greg Stein
s. > > Bert From: Daniel Shahaf > Sent: =E2=80=8E30/=E2=80=8E07/=E2=80=8E2013 16:47 > To: Stefan Fuhrmann > Cc: Subversion Development; comm...@subversion.apache.org > Subject: Re: svn commit: r1508170 - in /subversion/trunk: > build/ac-macros/svn-macros.m4 configure.ac subvers

RE: svn commit: r1508170 - in /subversion/trunk:

2013-07-30 Thread Bert Huijben
: Stefan Fuhrmann Cc: Subversion Development; comm...@subversion.apache.org Subject: Re: svn commit: r1508170 - in /subversion/trunk: build/ac-macros/svn-macros.m4 configure.ac subversion/include/svn_hash.h Stefan Fuhrmann wrote on Tue, Jul 30, 2013 at 16:30:00 +0200: > On Tue, Jul 30, 2013 at 3:57 PM, Dan

Re: svn commit: r1508170 - in /subversion/trunk: build/ac-macros/svn-macros.m4 configure.ac subversion/include/svn_hash.h

2013-07-30 Thread Daniel Shahaf
Daniel Shahaf wrote on Tue, Jul 30, 2013 at 17:46:44 +0300: > Stefan Fuhrmann wrote on Tue, Jul 30, 2013 at 16:30:00 +0200: > > Well, I got kind of fed up after manually patching > > ~180 files (about the same number of files left to > > check). Even the commit took longer than 4 minutes. > > > >

Re: svn commit: r1508170 - in /subversion/trunk: build/ac-macros/svn-macros.m4 configure.ac subversion/include/svn_hash.h

2013-07-30 Thread Daniel Shahaf
Stefan Fuhrmann wrote on Tue, Jul 30, 2013 at 16:30:00 +0200: > On Tue, Jul 30, 2013 at 3:57 PM, Daniel Shahaf wrote: > > > Stefan Fuhrmann wrote on Tue, Jul 30, 2013 at 01:04:43 +0200: > > > On Mon, Jul 29, 2013 at 9:06 PM, Daniel Shahaf > > wrote: > > > > > > > danie...@apache.org wrote on Mon

Re: svn commit: r1508170 - in /subversion/trunk: build/ac-macros/svn-macros.m4 configure.ac subversion/include/svn_hash.h

2013-07-30 Thread Stefan Fuhrmann
On Tue, Jul 30, 2013 at 3:57 PM, Daniel Shahaf wrote: > Stefan Fuhrmann wrote on Tue, Jul 30, 2013 at 01:04:43 +0200: > > On Mon, Jul 29, 2013 at 9:06 PM, Daniel Shahaf > wrote: > > > > > danie...@apache.org wrote on Mon, Jul 29, 2013 at 18:35:26 -: > > > > Author: danielsh > > > > Date: Mon

Re: svn commit: r1508170 - in /subversion/trunk: build/ac-macros/svn-macros.m4 configure.ac subversion/include/svn_hash.h

2013-07-30 Thread Daniel Shahaf
Stefan Fuhrmann wrote on Tue, Jul 30, 2013 at 01:04:43 +0200: > On Mon, Jul 29, 2013 at 9:06 PM, Daniel Shahaf wrote: > > > danie...@apache.org wrote on Mon, Jul 29, 2013 at 18:35:26 -: > > > Author: danielsh > > > Date: Mon Jul 29 18:35:25 2013 > > > New Revision: 1508170 > > > > > > URL: ht

Re: svn commit: r1508170 - in /subversion/trunk: build/ac-macros/svn-macros.m4 configure.ac subversion/include/svn_hash.h

2013-07-29 Thread Stefan Fuhrmann
On Mon, Jul 29, 2013 at 9:06 PM, Daniel Shahaf wrote: > danie...@apache.org wrote on Mon, Jul 29, 2013 at 18:35:26 -: > > Author: danielsh > > Date: Mon Jul 29 18:35:25 2013 > > New Revision: 1508170 > > > > URL: http://svn.apache.org/r1508170 > > Log: > > Follow-up to r1507366: svn_hash_gets

Re: svn commit: r1508170 - in /subversion/trunk: build/ac-macros/svn-macros.m4 configure.ac subversion/include/svn_hash.h

2013-07-29 Thread Daniel Shahaf
danie...@apache.org wrote on Mon, Jul 29, 2013 at 18:35:26 -: > Author: danielsh > Date: Mon Jul 29 18:35:25 2013 > New Revision: 1508170 > > URL: http://svn.apache.org/r1508170 > Log: > Follow-up to r1507366: svn_hash_gets: compute the length of string literal > keys (common case) at compile-