Re:Re:Re: Puzzled at storeSwapOutAble. :-(

Thanks, Henrik pal!

I am sorry if I send more than one copies of the message, my
MTA is really slow and in trouble. :-)

Best regards,
George Ma

----- Original Message -----
From: Henrik Nordstrom
To: maer727@sohu.com
Subject: Re:Re: Puzzled at storeSwapOutAble. :-(
Sent: Tue Apr 23 18:01:22 CST 2002

> Why did you send me 4 of these???
>
>
>
> You wrote:
> > Thanks, Henrik pal!
> >
> > What is the variable e->mem_obj->swapout.sio stands for? I think
> > it must be start of something. But the meaning and function of the
> > variable really puzzled me. Can you give me a simple explanation?
> >
> > Another question, I think e->mem_obj->inmem_lo indicates the start
> > buffer address of the object in memory. How can you infer when
> > e->mem_obj->inmem_lo>0 that the start of the object is always missing?
> > What you mean "missing" here? Cache missing?
> > Can you give me a simple explabation?
> >
> > Best regards,
> > George Ma
> >
> > ----- Original Message -----
> > From: Henrik Nordstrom
> > To: maer727@sohu.com
> > Cc: squid-dev@squid-cache.org
> > Subject: Re: Puzzled at storeSwapOutAble. :-(
> > Sent: Tue Apr 23 16:10:00 CST 2002
> >
> > > maer727@sohu.com wrote:
> > > > if (e->mem_obj->swapout.sio != NULL)
> > > > return 1;
> > > > if (e->mem_obj->inmem_lo > 0)
> > > > return 0;
> > > >
> > > > I do not understand why should we check swapout.sio and
> > > > inmem_lo?
> > >
> > > If it is already being swapped out, then we can continue swapping out
> > > out on disk..
> > >
> > > If it is not being swapped out, and the start of the object is missing,
> > > then we cannot swap it out on disk as we don't have all required data.
> > >
> > > Regards
> > > Henrik
Received on Tue Apr 23 2002 - 21:21:12 MDT