Thread: Latest tree

Latest tree

From
"Ansley, Michael"
Date:
Hi, all,

Whoever is doing the SSL stuff:
In postmaster.c, at lines 995, and 1841, there is code that is not wrapped
in the USE_SSL define.

Last night I checked out the latest source, and couldn't get it to compile.
It seems that the function heap_openr() has a new parameter, and there are
calls that have not been updated yet.  I was a little hesitant to go adding
stuff, as the new parameter is a LOCKTYPE, and I wouldn't know what locks
are required where, so I just left well alone.  Any comments on this?  There
is another function which is paired up with it, but I forget the name, which
also has a new parameter, it seems, and also has calls which have not yet
been updated.


MikeA


Re: [HACKERS] Latest tree

From
Vadim Mikheev
Date:
"Ansley, Michael" wrote:
> 
> Hi, all,
> 
> Whoever is doing the SSL stuff:
> In postmaster.c, at lines 995, and 1841, there is code that is not wrapped
> in the USE_SSL define.
> 
> Last night I checked out the latest source, and couldn't get it to compile.
> It seems that the function heap_openr() has a new parameter, and there are
> calls that have not been updated yet.  I was a little hesitant to go adding
> stuff, as the new parameter is a LOCKTYPE, and I wouldn't know what locks
> are required where, so I just left well alone.  Any comments on this?  There
> is another function which is paired up with it, but I forget the name, which
> also has a new parameter, it seems, and also has calls which have not yet
> been updated.

I fixed all this yesterday while committing my WAL changes...

Vadim


Re: [INTERFACES] Re: [HACKERS] Latest tree

From
Tatsuo Ishii
Date:
>I fixed all this yesterday while committing my WAL changes...
>
>Vadim

Does this mean that now we can enjoy the WAL?:-)

BTW, your WAL implementation will allow database recovery from log
files by using roll-forward or similar techniques?
--
Tatsuo Ishii


Re: [INTERFACES] Re: [HACKERS] Latest tree

From
Vadim Mikheev
Date:
Tatsuo Ishii wrote:
> 
> >I fixed all this yesterday while committing my WAL changes...
> >
> >Vadim
> 
> Does this mean that now we can enjoy the WAL?:-)

No yet :))

> 
> BTW, your WAL implementation will allow database recovery from log
> files by using roll-forward or similar techniques?

Yes.

Vadim