| Message ID | 20171003051523.17650-3-aik@ozlabs.ru |
|---|---|
| State | Superseded |
| Headers | show |
| Series | fdt: Pass the resulting device tree to QEMU | expand |
On Tue, Oct 03, 2017 at 04:15:20PM +1100, Alexey Kardashevskiy wrote: > This reverts commit 089fc18a9b > "libhvcall: drop unused KVMPPC_H_REPORT_MC_ERR and KVMPPC_H_NMI_MCE defines" > as a bigger hammer is coming soon > which will pass the entire device tree to QEMU, not just phandles. > > QEMU has never used the hcall this patch is removing so there will > be no compatibility issues. > > Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> Yeah. These probably shouldn't have been committed to SLOF without the corresponding qemu code being at least acked in principle. After all, the qemu source is effectively the authoritative source for what the vendor specific hcalls are. > --- > lib/libhvcall/libhvcall.h | 2 ++ > 1 file changed, 2 insertions(+) > > diff --git a/lib/libhvcall/libhvcall.h b/lib/libhvcall/libhvcall.h > index b2ea3f6..193b738 100644 > --- a/lib/libhvcall/libhvcall.h > +++ b/lib/libhvcall/libhvcall.h > @@ -25,6 +25,8 @@ > /* Client Architecture support */ > #define KVMPPC_H_CAS (KVMPPC_HCALL_BASE + 0x2) > #define KVMPPC_H_RTAS_UPDATE (KVMPPC_HCALL_BASE + 0x3) > +#define KVMPPC_H_REPORT_MC_ERR (KVMPPC_HCALL_BASE + 0x4) > +#define KVMPPC_HCALL_MAX KVMPPC_H_NMI_MCE > > #ifndef __ASSEMBLY__ >
On Tue, 3 Oct 2017 16:15:20 +1100 Alexey Kardashevskiy <aik@ozlabs.ru> wrote: > This reverts commit 089fc18a9b > "libhvcall: drop unused KVMPPC_H_REPORT_MC_ERR and KVMPPC_H_NMI_MCE defines" > as a bigger hammer is coming soon > which will pass the entire device tree to QEMU, not just phandles. > Not sure you should do a plain revert of 089fc18a9b because... > QEMU has never used the hcall this patch is removing so there will > be no compatibility issues. > > Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> > --- > lib/libhvcall/libhvcall.h | 2 ++ > 1 file changed, 2 insertions(+) > > diff --git a/lib/libhvcall/libhvcall.h b/lib/libhvcall/libhvcall.h > index b2ea3f6..193b738 100644 > --- a/lib/libhvcall/libhvcall.h > +++ b/lib/libhvcall/libhvcall.h > @@ -25,6 +25,8 @@ > /* Client Architecture support */ > #define KVMPPC_H_CAS (KVMPPC_HCALL_BASE + 0x2) > #define KVMPPC_H_RTAS_UPDATE (KVMPPC_HCALL_BASE + 0x3) > +#define KVMPPC_H_REPORT_MC_ERR (KVMPPC_HCALL_BASE + 0x4) > +#define KVMPPC_HCALL_MAX KVMPPC_H_NMI_MCE > ... KVMPPC_H_NMI_MCE isn't defined anywhere. We should have: #define KVMPPC_HCALL_MAX KVMPPC_H_RTAS_UPDATE > #ifndef __ASSEMBLY__ >
diff --git a/lib/libhvcall/libhvcall.h b/lib/libhvcall/libhvcall.h index b2ea3f6..193b738 100644 --- a/lib/libhvcall/libhvcall.h +++ b/lib/libhvcall/libhvcall.h @@ -25,6 +25,8 @@ /* Client Architecture support */ #define KVMPPC_H_CAS (KVMPPC_HCALL_BASE + 0x2) #define KVMPPC_H_RTAS_UPDATE (KVMPPC_HCALL_BASE + 0x3) +#define KVMPPC_H_REPORT_MC_ERR (KVMPPC_HCALL_BASE + 0x4) +#define KVMPPC_HCALL_MAX KVMPPC_H_NMI_MCE #ifndef __ASSEMBLY__
This reverts commit 089fc18a9b "libhvcall: drop unused KVMPPC_H_REPORT_MC_ERR and KVMPPC_H_NMI_MCE defines" as a bigger hammer is coming soon which will pass the entire device tree to QEMU, not just phandles. QEMU has never used the hcall this patch is removing so there will be no compatibility issues. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> --- lib/libhvcall/libhvcall.h | 2 ++ 1 file changed, 2 insertions(+)