Hi
I assume this needs to be analyzed from System Device Tree perspective:
https://trustedfirmware-a.readthedocs.io/en/latest/components/psa-ffa-manif…
And this is to be included in the DT Technical Report.
Cheers
FF
--
François-Frédéric Ozog | *Director Linaro Edge & Fog Computing Group*
T: +33.67221.6485
francois.ozog(a)linaro.org | Skype: ffozog
Hi everyone.
I have to do this, but I have another unexpected conflict for the EBBR biweekly on the 14th.
Rather than cancelling outright, does anyone else want to chair the meeting? The major planned orientatio item on the agenda was to talk about EBBR testing, with Heinrich sharing what he is currently doing.
If I don't hear anything by about 1pm GMT tomorrow then I'll just cancel. Our next meeting will be in January as I believe most of us will already be on Christmas holiday on the 21st
g.
Get Outlook for Android<https://aka.ms/ghei36>
IMPORTANT NOTICE: The contents of this email and any attachments are confidential and may also be privileged. If you are not the intended recipient, please notify the sender immediately and do not disclose the contents to any other person, use it for any purpose, or store or copy the information in any medium. Thank you.
Hi,
I thought perhaps it might be worth starting a thread on this, as
despite Grant and Heinrich kinding spending a bit of time talking
about this, I am still very much in the dark about how 'embedded' and
distro/other boot flows are going to come together with EBBR. Of
course this would be easier f2f.
Case 1:
Firmware loads the kernel to a particular address, selects DT and
boots it. The kernel may require EFI boot services, or may not, but in
the general case the firmware provides them.
Case 2:
Firmware loads EFI app and provides EFI boot services to it. How the
system actually boots is under control of the app.
I feel that a lot of the confusion about verified boot, DT selections,
boot menus, etc. is coming from the introduction of an EFI app which
has no specification (it can be grub, shim or something else, as I
understand it). Certainly this is very flexible and future-proof, but
it is also arbitrarily complex, unpredictable and hard to secure.
I am wondering if we can come up with a way to deterministically
specify how a system will boot and how to make it boot a different way
(i.e. with a different kernel, initrd, DT).
Heinrich mentioned EFI variables as a way of selecting
kernel/initrd/DT. Then the problem becomes just a case of being able
to change those variables from Linux userspace. Is that right?
We are talking about having a 'secure' part of EBBR, which allows for
secure boot. Should we have a 'defined boot' part of EBBR, that
defines how the kernel/DT/initrd are selected, based on EFI variables?
Unfortunately I just don't know enough about all the different boot
flows used by the different distros. It seems like crazy town. Does
anyone have some pointers so I can do some study?
Regards,
SImon
Hi,
As I am thinking about conformance testing for SystemReady and Trusted
Substrate, I'd like to get your feedback on the following.
There are 7 values in the reg entry of interrupt-controller@210000 from the
below DT. This corresponds to 3 valid {address,size} plus a single
{address}.
The spec does not state anything on incomplete {address,size} pairs... I
understand that #size-cell can be zero, indicating that the reg will
contain only {address} "tuples" and not {address,size} tuples. But that
should be for all reg tuples, not just one.
In this case, I assume the driver will get what it wants, but from a
certification perspective:
- I would reject this DT.
- I would document proper tuple forming in the spec (no incomplete pairs)
Last, I would also add some "notes" in the spec about where to get the
"#*-cells" for the reg property of a device. If you think "hardware" it is
obvious that the information must be retrieved from the immediate parent
and "inheritance" does not make sense. But as I Googled the topic, I have
seen a number of discussions and wrong patches around that. So I would add
a non normative text (properly identified as such) to describe that in the
spec.
Thank you for your help
Cheers
FF
config-space@f0000000 {
#address-cells = <0x01>;
#size-cells = <0x01>;
compatible = "simple-bus";
ranges = <0x00 0x00 0xf0000000 0x1000000>;
interrupt-controller@210000 {
compatible = "arm,gic-400";
#interrupt-cells = <0x03>;
#address-cells = <0x01>;
#size-cells = <0x01>;
ranges;
interrupt-controller;
interrupts = <0x01 0x09 0xf04>;
reg = <0x210000 0x10000 0x220000 0x20000 0x240000 0 0x20000>;
phandle = <0x01>;
v2m@280000 {
compatible = "arm,gic-v2m-frame";
msi-controller;
reg = <0x280000 0x1000>;
arm,msi-base-spi = <0xa0>;
arm,msi-num-spis = <0x20>;
phandle = <0x03>;
};
--
François-Frédéric Ozog | *Director Linaro Edge & Fog Computing Group*
T: +33.67221.6485
francois.ozog(a)linaro.org | Skype: ffozog
All who participate or wish to participate in the DTE call,
Please fill in this pool before you go out for the holidays:
https://doodle.com/poll/fup2swsb9yi2u5u4?utm_source=poll&utm_medium=link
The current meeting time does not work well for people on the west coast
of North America as it is 6am Pacific time.
I have selected the following times:
7 am PT, 15:00 UTC
8 am PT, 16:00 UTC
9 am PT, 17:00 UTC
I have found options for Monday, Wednesday, Thursday, and Friday.
Tuesday does not work at all for me at any reasonable time for the EU.
All meeting times will repeat every two weeks except for the Wednesday
7am & 8am slots.
These will normally be the 2nd and 4th Wednesdays of the month but in
January will be the 6th and 27th. (This is to avoid the Linaro TSC
calls that a good number us participate in.)
If you want me to insert new times into the poll please respond to this
message.
Thanks,
Bill
--
Bill Mills
Principal Technical Consultant, Linaro
+1-240-643-0836
TZ: US Eastern
Work Schedule: Tues/Wed/Thur
Dear all,
we have discussed that a protocol is needed to fix-up device trees
loaded by GRUB or other boot managers.
In the U-Boot code we have actually the following actions for
device-trees:
* load the device-tree to memory
* copy the device-tree to an allocated memory region which
has 12 KiB free space for device-tree fix-ups
* do the actual fix-ups, i.e. add new nodes or change properties
* reserve memory according to the /reserved-memory node and the
memory reservation block as EfiBootServicesData or as
EfiReservedMemoryType for no-map regions
* install the device-tree as configuration table
So I think the usage of a DT fix-up protocol could take the following steps:
* GRUB loads the device-tree allocating sufficient memory for fix-ups
* GRUB calls the protocol a fist time to add extra nodes and properties
* GRUB applies its own device-tree fix-ups
* GRUB calls the protocol a second time which
- reserves memory according to the /reserved-memory node
- installs the device-tree as configuration table
One could join both service calls if GRUB applies its own fix-ups first.
But maybe GRUB wants to analyze U-Boot's fix-ups before committing its
own ones.
So lets define a bit-field to pass to the fix-up protocol:
/* Add nodes and update properties */
#define EFI_DT_APPLY_FIXUPS 0x00000001
/*
* Reserve memory according to the /reserved-memory node
* and the memory reservation block
*/
#define EFI_DT_RESERVE_MEMORY 0x00000002
/* Install the device-tree as configuration table */
#define EFI_DT_INSTALL_TABLE 0x00000004
Here is the rest of the proposed protocol definition:
#define EFI_DT_FIXUP_PROTOCOL_GUID \
{ 0xe617d64c, 0xfe08, 0x46da, \
{ 0xf4, 0xdc, 0xbb, 0xd5, 0x87, 0x0c, 0x73, 0x00 } }
typedef struct _EFI_DT_FIXUP_PROTOCOL {
EFI_DT_FIXUP fixup;
} EFI_DT_FIXUP_PROTOCOL;
typedef EFI_STATUS
(EFIAPI *EFI_DT_FIXUP) (
IN EFI_DT_FIXUP_PROTOCOL *This,
IN VOID *Fdt,
IN OUT UINTN *BufferSize,
IN UINT32 Flags,
);
This: Pointer to the protocol
Fdt: Buffer with the device-tree. This shall be memory
of type EfiACPIReclaimMemory if Flags contains
EFI_DT_INSTALL_TABLE.
BufferSize: Pointer to the size of the buffer including
trailing unused bytes for fix-ups.
If the buffer size is too small,
the required buffer size is returned.
Flags: Bitmap containing at least one of the values
EFI_DT_APPLY_FIXUPS, EFI_DT_RESERVE_MEMORY,
EFI_DT_INSTALL_TABLE. Indicates the actions
to be applied to the device-tree.
The selected actions indicated in Flags are applied in the
sequence:
* Add nodes and update properties.
* Reserve memory according to the /reserved-memory node
and the memory reservation block
* Install the device-tree as configuration table
Memory is reserved as EfiBootServicesData if the reservation does not
carry the no-map property and as EfiReservedMemoryType if it is marked
as no-map.
If *BufferSize exceeds the value of the totalsize field header of
device-tree header upon entry to the service, the totalsize field is set
to *BufferSize.
Return values:
EFI_INVALID_PARAMETER - This is NULL or does not point to a valid
EFI_DT_FIXUP_PROTOCOL implementation.
EFI_INVALID_PARAMETER - Fdt or BufferSize is NULL
EFI_INVALID_PARAMETER - *Fdt is not a valid device-tree
(e.g. incorrect value of magic)
EFI_INVALID_PARAMETER - Invalid value of Flags (zero or unknown bit)
EFI_BUFFER_TOO_SMALL - The buffer is too small to apply the fix-ups.
EFI_SUCCESS - All steps succeeded
If EFI_BUFFER_TOO_SMALL is returned, the device-tree is unmodified
and *BufferSize is updated with the required buffer size for the
provided device-tree.
The required buffer size when called with EFI_DT_APPLY_FIXUPS should
enforce at least 4 KiB unused space for additional fix-ups by the
operating system or the caller. The available space in the device-tree
shall be determined using the device-tree header fields:
Available = header->totalsize
- header->off_dt_strings
- header->size_dt_strings;
(The strings block is always last in the flattened device-tree. There
might be more space between blocks but not all device-tree libraries can
use it.)
The required buffer size when called without EFI_DT_APPLY_FIXUPS shall
be the value of the totalsize field of the flattened device tree header.
If any other error code is returned, the state of the device-tree is
undefined. The caller should discard the buffer content.
The extent to which the validity of the device-tree is checked is
implementation dependent. But a buffer without the correct value of the
magic field of the flattened device tree header should always be rejected.
The protocol implementation is not required to check if the device-tree
is in memory of type EfiACPIReclaimMemory.
Looking forward to your feedback.
Best regards
Heinrich
All,
Sorry for the late notice.
We will have our normal DTE call at 2PM GMT today.
My suggested topic for today is the open items in the DTB ABI slides here:
https://docs.google.com/presentation/d/1Hq7-42EfM4xC_1N1HMO20vXVKSpkA_S-Los…
This will be the last call for 2020.
My suggestion for the DTE calls after this is to start in January with
either:
Option 1:
2nd & 4th Wednesday @ 4 PM GMT / 11 AM US Eastern
Option 2:
Alternate weeks with the EBBR call, Mondays at 4 PM GMT
Thanks,
Bill
--
Bill Mills
Principal Technical Consultant, Linaro
+1-240-643-0836
TZ: US Eastern
Work Schedule: Tues/Wed/Thur
Hello Atish,
the UEFI spec has this sentence:
"When UEFI firmware handoff control to OS, the RISC-V is operated in
machine-mode privilege." (M-mode is the equivalent to EL3 in ARM).
This does not make any sense to me when using a secure execution
environement (SEE) like OpenSBI.
The hand-off should occur in S-Mode if the CPU supports it and only in
M-Mode when the CPU only supports M-mode.
We should prescribe this in the EBBR and somehow get the UEFI spec fixed
afterwards.
An other issue is the calling convention. Chapter "2.3.7.3 Detailed
Calling Convention" does not describe which registers are saved by the
UEFI payload's entry point and restored by the payload before calling
the UEFI API or returning to the UEFI payload. This concerns especially
registers gp (x3) and tp (x4).
Into the EBBR or UEFI spec we should put a link to the "RISC-V ELF psABI
specification"
https://github.com/riscv/riscv-elf-psabi-doc/blob/master/riscv-elf.md
which is referenced by "The RISC-V Instruction Set Manual".
>From the "RISC-V ELF psABI specification" one might conclude that the
UEFI payload should not be allowed to change gp and tp before calling
ExitBootServices() or SetVirtualAddressMap() whichever occurs last.
Due to this missing clarification U-Boot is currently saving gp before
calling the entry point of the payload and restores it on reentry or on
entry of an API call. Nothing is done for tp.
Best regards
Heinrich
I have a conflict this week and need to cancel. I propose pushing out to next week (Dec 14th), and cancelling the meeting on the 21st when many people will be on holiday anyway. Let me know if you want anything added to the meeting agenda before next week.
Draft agenda:
* Action item review
* EBBR Testing Efforts (SCT, FWTS, etc)
* UEFI Exception text changes
* Next release schedule
* Other business
Time: This is a recurring meeting Meet anytime
Join Zoom Meeting
https://armltd.zoom.us/j/92081365511?pwd=SFZpRitXUEp3Zy9GM0h3UUZ1b1pnUT09
Meeting ID: 920 8136 5511
Password: 490324
One tap mobile
+14086380968,,92081365511#,,#,490324# US (San Jose)
+16465189805,,92081365511#,,#,490324# US (New York)
Dial by your location
+1 408 638 0968 US (San Jose)
+1 646 518 9805 US (New York)
+1 346 248 7799 US (Houston)
Meeting ID: 920 8136 5511
Password: 490324
Find your local number: https://armltd.zoom.us/u/adYiWaDyys
IMPORTANT NOTICE: The contents of this email and any attachments are confidential and may also be privileged. If you are not the intended recipient, please notify the sender immediately and do not disclose the contents to any other person, use it for any purpose, or store or copy the information in any medium. Thank you.
