Re: New hashtable power 2 rehash policy

2016-05-25 Thread Jonathan Wakely
On 25/05/16 10:15 +0200, Andreas Schwab wrote: François Dumont writes: * include/bits/c++config (_GLIBCXX14_USE_CONSTEXPR): New. FAIL: ext/profile/mutex_extensions_neg.cc (test for errors, line 324) FAIL: ext/profile/mutex_extensions_neg.cc (test for excess errors) Excess errors: /usr/l

Re: New hashtable power 2 rehash policy

2016-05-25 Thread Andreas Schwab
François Dumont writes: > * include/bits/c++config (_GLIBCXX14_USE_CONSTEXPR): New. FAIL: ext/profile/mutex_extensions_neg.cc (test for errors, line 324) FAIL: ext/profile/mutex_extensions_neg.cc (test for excess errors) Excess errors: /usr/local/gcc/gcc-20160525/Build/ia64-suse-linux/libst

Re: New hashtable power 2 rehash policy

2016-05-24 Thread François Dumont
Attached patch applied then. I had to regorganize things a little now that some pieces have been integrated in 71181 patch. 2016-05-24 François Dumont * include/bits/c++config (_GLIBCXX14_USE_CONSTEXPR): New. * include/bits/hashtable_policy.h (_Prime_rehash_policy::__has_load_f

Re: New hashtable power 2 rehash policy

2016-05-23 Thread Jonathan Wakely
On 17/05/16 22:28 +0200, François Dumont wrote: On 14/05/2016 19:06, Daniel Krügler wrote: 1) The function __clp2 is declared using _GLIBCXX14_CONSTEXPR, which means that it is an inline function if and *only* if _GLIBCXX14_CONSTEXPR really expands to constexpr, otherwise it is *not* inline, whi

Re: New hashtable power 2 rehash policy

2016-05-17 Thread François Dumont
On 14/05/2016 19:06, Daniel Krügler wrote: 2016-05-14 18:13 GMT+02:00 François Dumont : New patch attached, tested under linux x86_64. François 1) The function __clp2 is declared using _GLIBCXX14_CONSTEXPR, which means that it is an inline function if and *only* if _GLIBCXX14_CONSTEXPR really

Re: New hashtable power 2 rehash policy

2016-05-14 Thread Jonathan Wakely
On 14/05/16 18:13 +0200, François Dumont wrote: sizeof() returns number of char elements in the given type ? I thought it was the number of bytes and that a byte was always 8 bits. But I know that you know your stuff so fixed in the new patch. A char is a byte, by definition, so sizeof returns

Re: New hashtable power 2 rehash policy

2016-05-14 Thread Daniel Krügler
2016-05-14 18:13 GMT+02:00 François Dumont : > New patch attached, tested under linux x86_64. > > François 1) The function __clp2 is declared using _GLIBCXX14_CONSTEXPR, which means that it is an inline function if and *only* if _GLIBCXX14_CONSTEXPR really expands to constexpr, otherwise it is *n

Re: New hashtable power 2 rehash policy

2016-05-14 Thread François Dumont
Le 28/04/2016 12:22, Jonathan Wakely a écrit : On 23/04/16 10:27 +0200, François Dumont wrote: Hi Here is a patch to introduce a new power of 2 based rehash policy. It enhances performance as it avoids modulo float operations. I have updated performance benches and here is the result: T

Re: New hashtable power 2 rehash policy

2016-04-28 Thread Jonathan Wakely
I'm making this small change to some comments in hashtable_policy.h Tested x86_64-linux, committing to trunk. commit c387345f7c68df8812c7909f9187445a79bd5dcb Author: Jonathan Wakely Date: Thu Apr 28 11:25:29 2016 +0100 * include/bits/hashtable_policy.h (__detail::_Insert_base, __d

Re: New hashtable power 2 rehash policy

2016-04-28 Thread Jonathan Wakely
On 23/04/16 10:27 +0200, François Dumont wrote: Hi Here is a patch to introduce a new power of 2 based rehash policy. It enhances performance as it avoids modulo float operations. I have updated performance benches and here is the result: 54075.cctr1 benches 455r

New hashtable power 2 rehash policy

2016-04-23 Thread François Dumont
Hi Here is a patch to introduce a new power of 2 based rehash policy. It enhances performance as it avoids modulo float operations. I have updated performance benches and here is the result: 54075.cctr1 benches 455r 446u8s 0mem0pf 54075.cc