It's IMO not the final users job to decide on such internal
details on the .prg level.
Of course not, but I´m my intention is have a way to easy test what is
the best value, without recompile Harbour each time.
Okay and sorry, in this case just put the C code inside
#pragma BEGINDUMP/ENDDUMP an
>Hi Toninho,
>
>IMO, these values should be preconfigured by Harbour (at
>compile time or runtime) for best values decided by the
>development team's (our) best judgement and at most offered
>as overrides for defaults in the form of build-time tuning
>or internal cmdline options (but I'd even avoid
Hi Toninho,
IMO, these values should be preconfigured by Harbour (at
compile time or runtime) for best values decided by the
development team's (our) best judgement and at most offered
as overrides for defaults in the form of build-time tuning
or internal cmdline options (but I'd even avoid that,
>Hi Toninho,
>
>You can make some experiments with DEFAULT_GRANULARITY, f.e.:
> set C_USR=-DDEFAULT_GRANULARITY=0x10
Hi Przemek thanks for your hint,
I see that the best value for windows vista is about 1. Thinking
in this I made 3 functions to change this dlmalloc values without need
On Mon, 10 Nov 2008, [EMAIL PROTECTED] wrote:
Hi Toninho,
> Here is the results for Windows Vista 6.0.6000. Strange that DL-Alloc
> wins pratically in all parts except in last part of memtst. Maybe is
> possible do an optmization in this process?
You can make some experiments with DEFAULT_GRANUL
>Thanks Toninho.
>
>You could further help by sending speedtst and
>memtst results for both DLALLOC and STDALLOC
>builds, run on Vista and XP.
Hi Viktor, thank you, but please, wait until tonight. In my home I
have windows XP. I´ll collect all results and post here.
Regards,
Toninho.
__
>As a last resort we may want to disable DLALLOC
>for BCC >= 6.0 if this seems to be best from the
>results.
Hi Viktor,
Here is the results for Windows Vista 6.0.6000. Strange that DL-Alloc
wins pratically in all parts except in last part of memtst. Maybe is
possible do an optmization in this pro
Thanks Toninho.
You could further help by sending speedtst and
memtst results for both DLALLOC and STDALLOC
builds, run on Vista and XP.
As a last resort we may want to disable DLALLOC
for BCC >= 6.0 if this seems to be best from the
results.
Brgds,
Viktor
On 2008.11.10., at 13:04, [EMAIL PROT
>Could you post the details of your Harbour build
>switches and C compiler type/version, Harbour revision?
>Did you do your tests in MT or ST mode?
---cut---
SET C_USR=-DHB_GUI -DHB_FM_STATISTICS_OFF -DHB_NO_PROFILER
-DADS_LIB_VERSION=700 -DHB_HASH_MSG_ITEMS -DHB_FM_STD_ALLOC
-DHB_NO_DEBUG
---cut-
Hi Toninho,
Could you post the details of your Harbour build
switches and C compiler type/version, Harbour revision?
Did you do your tests in MT or ST mode?
Maybe also Vista version/SP level, CPU type and
physical memory could also help.
Experience from Vista users would also be great to
hear.
>> In windows vista my tests display bad results if we use DL_ALLOC,
>
>Can you tell some more about this? What does
>"bad results" mean? It'd be important to know about
>such problems before we settle with our default.
Hi Viktor, sorry for delay. I don´t have windows vista in my home, so
I was wa
Hi Toninho,
I use -DHB_HASH_MSG_ITEMS and -DHB_FM_STD_ALLOC too,
Yes, there is more, I was just listing the performance
related ones.
In windows vista my tests display bad results if we use DL_ALLOC,
Can you tell some more about this? What does
"bad results" mean? It'd be important to know
>To further tune Harbour performance vs. core features, below
>flags can be used:
>-DHB_FM_STATISTICS_OFF (for final all apps)
>-DHB_NO_DEBUG (if you never intend to use the debugger)
>-DHB_GUI (to build GUI apps exclusively)
Hi,
I use -DHB_HASH_MSG_ITEMS and -DHB_FM_STD_ALLOC too,
In windows vi
On Sun, 09 Nov 2008, Enrico Maria Giordano wrote:
Hi Enrico,
> Let's see if I understood correctly. I want Windows native I/O. Used
> HB_WIN32_IO so far. Do I have to replace it with HB_IO_WIN?
No you do not have to use it. You haven't to HB_WIN32_IO so far
and many times in the past we were wr
-Messaggio Originale-
Da: "Szakáts Viktor" <[EMAIL PROTECTED]>
A:
Data invio: domenica 9 novembre 2008 20.59
Oggetto: [Harbour] 2008-11-09 20:58 UTC+0100 Viktor Szakats
(harbour.01syenar hu)
2008-11-09 20:58 UTC+0100 Viktor Szakats (harbour.01 syenar hu)
* include/hbdefs.h
* inc
Hi Enrico,
* include/hbdefs.h
* include/hbapigt.h
* source/common/hbfsapi.c
* source/rtl/console.c
* source/rtl/fserror.c
* source/rtl/fstemp.c
* source/rtl/gtstd/gtstd.c
* source/rtl/gtpca/gtpca.c
* source/rtl/filesys.c
* Macro rename:
- HB_WIN32_IO_OFF -> HB_IO_WIN_OFF
- HB_WIN32_IO
16 matches
Mail list logo
