[so] [tema5] rezolvarea dilemei

Silviu-Ionut Ganceanu silviug at gmail.com
Sun Jan 6 21:58:06 EET 2008


Nicio parere oficiala apropo de chestia asta?

On Jan 6, 2008 7:00 PM, Silviu-Ionut Ganceanu <silviug at gmail.com> wrote:

> Plusand, cred ca ce vream check_ram0 de fapt este:
>
> static void check_ram0 (size_t ppages)
> {
> ....
>     for (i = 0; i < ppages * p_sz / 4; i++) {
>         read (ram_fd, &read_val, 4);
>         if (read_val != SWAP_POISON) {
>             check = 0;
>             break;
>         }
>     }
>
>     test ("ram_clean", check);
>
> }
>
>
> On Jan 6, 2008 6:49 PM, Andrei Deftu < andreideftu at yahoo.com> wrote:
>
> > Nu, are dreptate Silviu ca se bat cap in cap testele.
> > Daca ar fi sa scrii toate paginile in RAM pe care le
> > aduci din swap, asta ar insemna sa pice ram_clean. De
> > ce? Se acceseaza niste pagini care nu sunt in RAM. Ce
> > se intampla cand se acceseaza pagini care nu sunt in
> > RAM: swapin from swap. Deci in RAM file ar trebui sa
> > apara poisonul din swap file.
> >
> > --- Catalina Caloian <catalina.caloian at gmail.com>
> > wrote:
> >
> > > Buna.
> > >
> > > La 'ram_sync' ar trebui scrise in fisierul care
> > > simuleaza RAM-ul si paginile
> > > care au fost aduse din SWAP, indiferent de modul in
> > > care au fost accesate
> > > ulterior.
> > >
> > > Catalina
> > >
> > >
> > >
> > > _______________________________________________
> > > so mailing list
> > > so at cursuri.cs.pub.ro
> > > http://cursuri.cs.pub.ro/cgi-bin/mailman/listinfo/so
> > >
> >
> >
> >
> >
> >  ____________________________________________________________________________________
> >
> > Never miss a thing.  Make Yahoo your home page.
> > http://www.yahoo.com/r/hs
> > _______________________________________________
> > so mailing list
> > so at cursuri.cs.pub.ro
> > http://cursuri.cs.pub.ro/cgi-bin/mailman/listinfo/so
> >
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://cursuri.cs.pub.ro/pipermail/so/attachments/20080106/1f3b48d6/attachment-0001.htm


More information about the so mailing list