On 6/2/07, Eduardo Morras <[EMAIL PROTECTED]> wrote:
At 17:18 31/05/2007, you wrote:
>On 5/31/07, Eduardo Morras <[EMAIL PROTECTED]> wrote:
>>At 23:25 30/05/2007, you wrote:
>>>Setting and reading individual bytes (u8 in sqlite-speak) are not
>>>threadsafe either. Only reading/setting entire enti
At 17:18 31/05/2007, you wrote:
>On 5/31/07, Eduardo Morras <[EMAIL PROTECTED]> wrote:
>>At 23:25 30/05/2007, you wrote:
>>>Setting and reading individual bytes (u8 in sqlite-speak) are not
>>>threadsafe either. Only reading/setting entire entire words
>>>are threadsafe on most architectures.
>>
>>
One needs to make use of CAS or LL/SC hardware instructions. These can be used
to implement lock-free synchronization.
Nuno Lucas <[EMAIL PROTECTED]> wrote:
On 5/31/07, Eduardo Morras wrote:
> At 23:25 30/05/2007, you wrote:
> >Setting and reading individual bytes (u8 in sqlite-speak) are no
On 5/31/07, Eduardo Morras <[EMAIL PROTECTED]> wrote:
At 23:25 30/05/2007, you wrote:
>Setting and reading individual bytes (u8 in sqlite-speak) are not
>threadsafe either. Only reading/setting entire entire words
>are threadsafe on most architectures.
Using a uint32 for store the flags is threa
At 23:25 30/05/2007, you wrote:
>--- [EMAIL PROTECTED] wrote:
>> - Original Message
>> > MemPage bitfield patch below.
>> >
>> > sizeof(MemPage) on Linux:
>> >
>> > original: 84
>> > patched: 76
>> > ...
>> > Break-even for memory is 904/8 = 113 MemPage structs allocated.
>>
>> I
> > If the MemPage are malloced individually (instead of being put in arrays),
> > then they are 16
> byte
> > aligned on most platforms, making the allocated block effectively the same
> > size (well, that
> > depends on how many bytes are used by malloc before the user block in
> > memory).
>
Joe Wilson wrote:
Generally structs are aligned on 8-byte boundaries, so making isInited
a bitfield wouldn't save any additional space in this particular case.
You can combine the other bitfields with isInited at the beginning of
the structure and save a byte later in the structure.
Den
--- Dennis Cote <[EMAIL PROTECTED]> wrote:
> Joe Wilson wrote:
> >
> > If an external interface changed, sure. But these internal structs
> > change constantly from (minor) release to release.
> >
> > The struct in question is used solely by btree.c, so the ordering
> > and layout for bit fields
--- [EMAIL PROTECTED] wrote:
> - Original Message
> > MemPage bitfield patch below.
> >
> > sizeof(MemPage) on Linux:
> >
> > original: 84
> > patched: 76
> > ...
> > Break-even for memory is 904/8 = 113 MemPage structs allocated.
>
> I didn't look at the code, so mind me :)
>
>
- Original Message
> From: Dennis Cote <[EMAIL PROTECTED]>
> To: sqlite-users@sqlite.org
> Sent: Wednesday, May 30, 2007 12:09:25 PM
> Subject: Re: [sqlite] sqlite internal structs don't make use of C bitfields?
> You may want to look at how the isInited fiel
Joe Wilson wrote:
If an external interface changed, sure. But these internal structs
change constantly from (minor) release to release.
The struct in question is used solely by btree.c, so the ordering
and layout for bit fields on different compilers or different platforms
do not matter.
- Original Message
> MemPage bitfield patch below.
>
> sizeof(MemPage) on Linux:
>
> original: 84
> patched: 76
> ...
> Break-even for memory is 904/8 = 113 MemPage structs allocated.
I didn't look at the code, so mind me :)
If the MemPage are malloced individually (instead of b
--- Dennis Cote <[EMAIL PROTECTED]> wrote:
> There are also some issues with regard to the ordering and layout
> of bitifleds in cross platform applications. I suspect that is the
> reason they aren't used.
If an external interface changed, sure. But these internal structs
change constantly fro
MemPage bitfield patch below.
sizeof(MemPage) on Linux:
original: 84
patched: 76
Patched "make test" runs without regressions on Linux and Windows.
Timings for "make test" (elapsed):
original: 1:20.74
patched: 1:20.22
Size of sqlite3.o when compiled from almalogmation with all
sql
Joe Wilson wrote:
You could save a few bytes in some sqlite internal structs if you'd
use C bitfields for boolean flags:
For example:
struct MemPage {
u8 isInit; /* True if previously initialized. MUST BE FIRST! */
u8 idxShift; /* True if Cell indices have changed */
u8
struct MemPage2 {
u8 nOverflow;/* Number of overflow cell bodies in aCell[] */
u8 childPtrSize; /* 0 if leaf==1. 4 if leaf==0 */
u8 hdrOffset;/* 100 for page 1. 0 otherwise */
u8 isInit:1; /* True if previously initialized. MUST BE FIRST! */
Okay, maybe not
You could save a few bytes in some sqlite internal structs if you'd
use C bitfields for boolean flags:
For example:
struct MemPage {
u8 isInit; /* True if previously initialized. MUST BE FIRST! */
u8 idxShift; /* True if Cell indices have changed */
u8 nOverflow;/*
17 matches
Mail list logo
