| Message ID | 1447753261-7552-51-git-send-email-shannon.zhao@linaro.org (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
>>> On 17.11.15 at 10:40, <shannon.zhao@linaro.org> wrote: > From: Shannon Zhao <shannon.zhao@linaro.org> > > When preparing EFI tables for Dom0, it needs the information of EFI > system table. Here store it in efi structure. Again I'm pretty sure I had already pointed out that this not being needed on x86 calls for either more than the brief description above or this being done just for ARM or (preferably) the table pointer - where _really_ needed - to be passed around as function argument(s). Jan
Hi Shannon, On 17/11/15 09:40, shannon.zhao@linaro.org wrote: > From: Shannon Zhao <shannon.zhao@linaro.org> > > When preparing EFI tables for Dom0, it needs the information of EFI > system table. Here store it in efi structure. On ARM64 the EFI stub is completely independent and therefore it's not possible to share variable between Xen and the stub. If it's working today is only by luck as we zero the BSS after the stub jumped to Xen entry point. Regards, > Signed-off-by: Parth Dixit <parth.dixit@linaro.org> > Signed-off-by: Shannon Zhao <shannon.zhao@linaro.org> > --- > xen/common/efi/boot.c | 2 ++ > xen/include/xen/efi.h | 1 + > 2 files changed, 3 insertions(+) > > diff --git a/xen/common/efi/boot.c b/xen/common/efi/boot.c > index 78d8ae9..6a48624 100644 > --- a/xen/common/efi/boot.c > +++ b/xen/common/efi/boot.c > @@ -1155,6 +1155,8 @@ efi_start(EFI_HANDLE ImageHandle, EFI_SYSTEM_TABLE *SystemTable) > > efi_tables(); > > + efi.est = (long)SystemTable; > + > /* Collect PCI ROM contents. */ > setup_efi_pci(); > > diff --git a/xen/include/xen/efi.h b/xen/include/xen/efi.h > index e74dad1..e112296 100644 > --- a/xen/include/xen/efi.h > +++ b/xen/include/xen/efi.h > @@ -16,6 +16,7 @@ struct efi { > unsigned long acpi20; /* ACPI table (ACPI 2.0) */ > unsigned long smbios; /* SM BIOS table */ > unsigned long smbios3; /* SMBIOS v3 table */ > + unsigned long est; /* EFI system table */ > }; > > extern struct efi efi; >
On 2015/11/17 19:57, Julien Grall wrote: > Hi Shannon, > > On 17/11/15 09:40, shannon.zhao@linaro.org wrote: >> From: Shannon Zhao <shannon.zhao@linaro.org> >> >> When preparing EFI tables for Dom0, it needs the information of EFI >> system table. Here store it in efi structure. > > On ARM64 the EFI stub is completely independent and therefore it's not > possible to share variable between Xen and the stub. > > If it's working today is only by luck as we zero the BSS after the stub > jumped to Xen entry point. > Yes, that's right. Sorry, I didn't investigate this deeply while just getting this patch of Parth version. I re-think about it. The reason why we need that variable is that when we create EFI System table for Dom0 it copies the table header from that. Since it supplies Dom0 a Xen specific UEFI and it could get the information from other way, this change is not necessary. I'll drop this at next version. Thanks,
diff --git a/xen/common/efi/boot.c b/xen/common/efi/boot.c index 78d8ae9..6a48624 100644 --- a/xen/common/efi/boot.c +++ b/xen/common/efi/boot.c @@ -1155,6 +1155,8 @@ efi_start(EFI_HANDLE ImageHandle, EFI_SYSTEM_TABLE *SystemTable) efi_tables(); + efi.est = (long)SystemTable; + /* Collect PCI ROM contents. */ setup_efi_pci(); diff --git a/xen/include/xen/efi.h b/xen/include/xen/efi.h index e74dad1..e112296 100644 --- a/xen/include/xen/efi.h +++ b/xen/include/xen/efi.h @@ -16,6 +16,7 @@ struct efi { unsigned long acpi20; /* ACPI table (ACPI 2.0) */ unsigned long smbios; /* SM BIOS table */ unsigned long smbios3; /* SMBIOS v3 table */ + unsigned long est; /* EFI system table */ }; extern struct efi efi;