Hi, Russ,
On Sun, Sep 28, 2008 at 11:01 PM, Russ Anderson <[email protected]> wrote:
> On Sat, Sep 27, 2008 at 09:21:35AM +0800, Huang Ying wrote:
>> On Fri, 2008-09-26 at 15:02 -0500, Russ Anderson wrote:
>> > Index: linux/include/asm-x86/efi.h
>> > ===================================================================
>> > --- linux.orig/include/asm-x86/efi.h 2008-09-26 14:13:24.000000000 -0500
>> > +++ linux/include/asm-x86/efi.h 2008-09-26 14:13:36.000000000 -0500
>> > @@ -49,6 +49,20 @@ extern u64 efi_call5(void *fp, u64 arg1,
>> > extern u64 efi_call6(void *fp, u64 arg1, u64 arg2, u64 arg3,
>> > u64 arg4, u64 arg5, u64 arg6);
>> >
>> > +
>> > +#ifndef CONFIG_EFI
>> > +/*
>> > + * IF EFI is not configured, have the EFI calls return -ENOSYS.
>> > + */
>> > +#define efi_call0(_f) (-ENOSYS)
>> > +#define efi_call1(_f, _a1) (-ENOSYS)
>> > +#define efi_call2(_f, _a1, _a2) (-ENOSYS)
>> > +#define efi_call3(_f, _a1, _a2, _a3) (-ENOSYS)
>> > +#define efi_call4(_f, _a1, _a2, _a3, _a4) (-ENOSYS)
>> > +#define efi_call5(_f, _a1, _a2, _a3, _a4, _a5) (-ENOSYS)
>> > +#define efi_call6(_f, _a1, _a2, _a3, _a4, _a5, _a6) (-ENOSYS)
>> > +#endif /* CONFIG_EFI */
>> > +
>> > #define efi_call_phys0(f) \
>> > efi_call0((void *)(f))
>> > #define efi_call_phys1(f, a1) \
>>
>> efi_call_virt<n> and efi_call_phys<n> is the API instead of efi_call<n>.
>
> #define efi_call_virt6(f, a1, a2, a3, a4, a5, a6) \
> efi_call6((void *)(efi.systab->runtime->f), (u64)(a1), (u64)(a2), \
> (u64)(a3), (u64)(a4), (u64)(a5), (u64)(a6))
>
> efi_call_virt6() uses efi.systab->runtime->f.
> My call needs to use uv_systab, not efi.systab.
Yes. efi_call_virt6() is tied with standard EFI runtime services, so
it may be not suitable for your need. What you need is just to do
calling convention converting, like efi_call6 does. So I think it is
better to just use your current implementation.
Acked-by: Huang Ying <[email protected]>
Best Regards,
Huang Ying
On Thu, Oct 02, 2008 at 04:02:00PM +0800, huang ying wrote:
> On Sun, Sep 28, 2008 at 11:01 PM, Russ Anderson <[email protected]> wrote:
> >
> > efi_call_virt6() uses efi.systab->runtime->f.
> > My call needs to use uv_systab, not efi.systab.
>
> Yes. efi_call_virt6() is tied with standard EFI runtime services, so
> it may be not suitable for your need. What you need is just to do
> calling convention converting, like efi_call6 does. So I think it is
> better to just use your current implementation.
>
> Acked-by: Huang Ying <[email protected]>
Thanks. Ingo, that is the v2 version of the patchset.
> Best Regards,
> Huang Ying
--
Russ Anderson, OS RAS/Partitioning Project Lead
SGI - Silicon Graphics Inc [email protected]
* Russ Anderson <[email protected]> wrote:
> On Thu, Oct 02, 2008 at 04:02:00PM +0800, huang ying wrote:
> > On Sun, Sep 28, 2008 at 11:01 PM, Russ Anderson <[email protected]> wrote:
> > >
> > > efi_call_virt6() uses efi.systab->runtime->f.
> > > My call needs to use uv_systab, not efi.systab.
> >
> > Yes. efi_call_virt6() is tied with standard EFI runtime services, so
> > it may be not suitable for your need. What you need is just to do
> > calling convention converting, like efi_call6 does. So I think it is
> > better to just use your current implementation.
> >
> > Acked-by: Huang Ying <[email protected]>
>
> Thanks. Ingo, that is the v2 version of the patchset.
ok - could you please resend the lot with all the Acked-by's in place?
(and with all the bits that are being objected left out)
Ingo
On Fri, Oct 03, 2008 at 11:10:12AM +0200, Ingo Molnar wrote:
> * Russ Anderson <[email protected]> wrote:
> > On Thu, Oct 02, 2008 at 04:02:00PM +0800, huang ying wrote:
> > > On Sun, Sep 28, 2008 at 11:01 PM, Russ Anderson <[email protected]> wrote:
> > > >
> > > > efi_call_virt6() uses efi.systab->runtime->f.
> > > > My call needs to use uv_systab, not efi.systab.
> > >
> > > Yes. efi_call_virt6() is tied with standard EFI runtime services, so
> > > it may be not suitable for your need. What you need is just to do
> > > calling convention converting, like efi_call6 does. So I think it is
> > > better to just use your current implementation.
> > >
> > > Acked-by: Huang Ying <[email protected]>
> >
> > Thanks. Ingo, that is the v2 version of the patchset.
>
> ok - could you please resend the lot with all the Acked-by's in place?
> (and with all the bits that are being objected left out)
>
> Ingo
Resent on Friday. Here are the links.
http://marc.info/?l=linux-kernel&m=122305331422380&w=2
http://marc.info/?l=linux-kernel&m=122305331522386&w=2
http://marc.info/?l=linux-kernel&m=122305331522389&w=2
http://marc.info/?l=linux-kernel&m=122305331622393&w=2
http://marc.info/?l=linux-kernel&m=122305331722396&w=2
Thanks,
--
Russ Anderson, OS RAS/Partitioning Project Lead
SGI - Silicon Graphics Inc [email protected]