Re: [Xen-devel] [PATCH] xen/hvc: If we use xen_raw_printk let italso work on HVM guests.

From: Ian Campbell
Date: Thu Aug 15 2013 - 10:40:51 EST


On Wed, 2013-08-14 at 08:46 -0400, Konrad Rzeszutek Wilk wrote:
> On Wed, Aug 14, 2013 at 08:15:56AM +0100, Ian Campbell wrote:
> > On Tue, 2013-08-13 at 20:10 -0400, Konrad Rzeszutek Wilk wrote:
> > > On Tue, Aug 13, 2013 at 11:20:05PM +0100, Ian Campbell wrote:
> > > > On Tue, 2013-08-13 at 21:59 +0100, Ian Campbell wrote:
> > > >
> > > > > > What it should be is:
> > > > > > > >
> > > > > > > > void xen_raw_console_write(const char *str)
> > > > > > > > {
> > > > > > > > - dom0_write_console(0, str, strlen(str));
> > > > > > > > + if (!xen_domain())
> > > > > > > > + return;
> > > > > > > > +
> > > > > > > > + if (xen_pv_domain())
> > > > > > xen_domain()
> > > > > >
> > > > > > > > + dom0_write_console(0, str, strlen(str));
> > > > > > > > + else if (xen_hvm_domain() || xen_cpuid_base()) {
> > > > > >
> > > > > > else if (xen_cpuid_base()) {
> > > > > >
> > > > > > > > + /* The hyperpage has not been setup yet. */
> > > > > > > > + int i, len = strlen(str);
> > > > > > > > + for (i = 0; i < len; i++)
> > > > > > > > + outb(str[i], 0xe9);
> > > > > > > > + }
> > > > > > > > }
> > > > > >
> > > > > > And then that should adhere to what I wrote up.
> > > > >
> > > > > I think it does too.
> > > >
> > > > Except as Daniel notes in <520A7145.5010306@xxxxxxxxxxxxx> for unrelated
> > > > reasons:
> > > >
> > > > > HVM guests can still use the PV output - they just need to use the console
> > > > > write hypercall instead of the HVM I/O port. I would think that PVH guests
> > > > > would default to using the hypercall as it is more efficient (it takes a
> > > > > string rather than one character per write).
> > > > >
> > > > > Actually, checking... the console_io hypercall would need to be added to
> > > > > the hvm_hypercall{32,64}_table for an HVM guest to be able to use it; they
> > > > > currently must use the I/O port. I didn't check the PVH patches.
> > > >
> > > > Or did you actually try this code and it worked?
> > >
> > > The one I typed up above - no. The one I had sent - yes.
> > >
> > > But with that above mentioned comment from Daniel I think it is still
> > > worth trying to do dom0_write_console and if the hypercall returns -ENOSYS
> > > then fall back on 0xe9.
> > >
> > > And lastly send an patch to make hypercall_io work under HVM.
> >
> > So you didn't try under PVHVM? That's what I was asking when I asked if
> > you tried it, since that is the point of this patch.
>
> Not the one I typed up above. Just the one I sent.

And did it work? ecause if the console_io hypercall isn't wired up in
the hypervisor I can't see how it can have done...

Oh the original patch would buggily never have tried to use console_io,
and you probably didn't notice the lack of output after the hypercall
page was setup.

Ian.

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/