[PATCH 00/14] qemu: arm64: Add support for uefi capsule update on qemu arm64 platform

The following series adds support for the uefi capsule update feature on the qemu arm64 platform, along with adding support for the capsule authentication feature.
The capsule update feature is supported on a platform configuration booting in a non-secure mode, i.e with -machine virt,secure=off option set. This results in the platform booting u-boot directly without the presence of trusted firmware(tf-a). Steps that need to be followed for using this feature have been provided as part of the documentation.
Support has also been added for enabling the capsule authentication feature. Capsule authentication, as defined by the uefi specification is very much on similar lines to the logic used for variable authentication. As a result, most of the signature verification code already in use for variable authentication has been used for capsule authentication.
Storage of the public key certificate, needed for the signature verification process is in form of the efi signature list(esl) structure. This public key is stored on the platform's device tree blob. The public key esl file can be embedded into the dtb using the mkeficapsule utility that has been added as part of the capsule update support series[1]. Steps needed for enabling capsule authentication have been provided as part of the documentation.
This patch series needs to be applied on top of the capsule update support patch series from Takahiro Akashi[1]
[1] - https://patchwork.ozlabs.org/project/uboot/cover/20201117002805.13902-1-taka...
Sughosh Ganu (14): qemu: arm: Use the generated DTB only when CONGIG_OF_BOARD is defined mkeficapsule: Add support for embedding public key in a dtb qemu: arm: Scan the pci bus in board_init crypto: Fix the logic to calculate hash with authattributes set qemu: arm64: Add support for dynamic mtdparts for the platform qemu: arm64: Set dfu_alt_info variable for the platform efi_loader: Add config option to indicate fmp header presence dfu_mtd: Add provision to unlock mtd device efi_loader: Make the pkcs7 header parsing function an extern efi_loader: Re-factor code to build the signature store from efi signature list efi: capsule: Add support for uefi capsule authentication efi_loader: Enable uefi capsule authentication efidebug: capsule: Add a command to update capsule on disk qemu: arm64: Add documentation for capsule update
board/emulation/qemu-arm/qemu-arm.c | 170 ++++++++++++++++++++++++ cmd/efidebug.c | 14 ++ doc/board/emulation/qemu-arm.rst | 157 ++++++++++++++++++++++ drivers/dfu/dfu_mtd.c | 20 ++- include/configs/qemu-arm.h | 8 ++ include/efi_api.h | 18 +++ include/efi_loader.h | 12 ++ lib/crypto/pkcs7_verify.c | 37 ++++-- lib/efi_loader/Kconfig | 24 ++++ lib/efi_loader/efi_capsule.c | 122 +++++++++++++++++ lib/efi_loader/efi_firmware.c | 49 ++++++- lib/efi_loader/efi_signature.c | 192 ++++++++++++++++++++------- lib/efi_loader/efi_variable.c | 93 +------------ tools/Makefile | 1 + tools/mkeficapsule.c | 198 ++++++++++++++++++++++++++-- 15 files changed, 954 insertions(+), 161 deletions(-)

The Qemu platform emulator generates a device tree blob and places it at the start of the dram, which is then used by u-boot. Use this dtb only if CONFIG_OF_BOARD is defined. This allows using a different device tree, using the CONFIG_OF_SEPARATE option. This dtb is attached to the u-boot binary as a u-boot-fdt.bin file
Signed-off-by: Sughosh Ganu sughosh.ganu@linaro.org --- board/emulation/qemu-arm/qemu-arm.c | 2 ++ 1 file changed, 2 insertions(+)
diff --git a/board/emulation/qemu-arm/qemu-arm.c b/board/emulation/qemu-arm/qemu-arm.c index f18f2ed7da..e146d1cc50 100644 --- a/board/emulation/qemu-arm/qemu-arm.c +++ b/board/emulation/qemu-arm/qemu-arm.c @@ -89,11 +89,13 @@ int dram_init_banksize(void) return 0; }
+#if defined(CONFIG_OF_BOARD) void *board_fdt_blob_setup(void) { /* QEMU loads a generated DTB for us at the start of RAM. */ return (void *)CONFIG_SYS_SDRAM_BASE; } +#endif
void enable_caches(void) {

On 11/26/20 7:40 PM, Sughosh Ganu wrote:
The Qemu platform emulator generates a device tree blob and places it at the start of the dram, which is then used by u-boot. Use this dtb only if CONFIG_OF_BOARD is defined. This allows using a different device tree, using the CONFIG_OF_SEPARATE option. This dtb is attached to the u-boot binary as a u-boot-fdt.bin file
Dear Sughosh,
thank your for this series which will allow us to better demonstrate and test capsule updates.
I am not sure if the approach that you take at device-trees here is the right one.
On QEMU the device-tree is generated on the fly by the QEMU binary depending on which devices the user has specified.
Your idea is to replace this device-tree completely to be able to add extra elements (the EFI signature list, see patch 2/14). Thus a device-tree might be loaded that does not match the user selected devices.
An alternative approach would be to apply all additions to the device-tree as an FDT overlay (or fixup). This would allow the dynamic parts of the QEMU device-tree still to be passed through.
Could you, please, assess the pros and cons of the two approaches with respect to:
* usability for capsule updates * applicability for non-QEMU systems * integration of DTB overlays with FIT images for other use cases
Best regards
Heinrich
Signed-off-by: Sughosh Ganu sughosh.ganu@linaro.org
board/emulation/qemu-arm/qemu-arm.c | 2 ++ 1 file changed, 2 insertions(+)
diff --git a/board/emulation/qemu-arm/qemu-arm.c b/board/emulation/qemu-arm/qemu-arm.c index f18f2ed7da..e146d1cc50 100644 --- a/board/emulation/qemu-arm/qemu-arm.c +++ b/board/emulation/qemu-arm/qemu-arm.c @@ -89,11 +89,13 @@ int dram_init_banksize(void) return 0; }
+#if defined(CONFIG_OF_BOARD) void *board_fdt_blob_setup(void) { /* QEMU loads a generated DTB for us at the start of RAM. */ return (void *)CONFIG_SYS_SDRAM_BASE; } +#endif
void enable_caches(void) {

hello Heinrich,
On Sat, 5 Dec 2020 at 15:01, Heinrich Schuchardt xypron.glpk@gmx.de wrote:
On 11/26/20 7:40 PM, Sughosh Ganu wrote:
The Qemu platform emulator generates a device tree blob and places it at the start of the dram, which is then used by u-boot. Use this dtb only if CONFIG_OF_BOARD is defined. This allows using a different device tree, using the CONFIG_OF_SEPARATE option. This dtb is attached to the u-boot binary as a u-boot-fdt.bin file
Dear Sughosh,
thank your for this series which will allow us to better demonstrate and test capsule updates.
I am not sure if the approach that you take at device-trees here is the right one.
On QEMU the device-tree is generated on the fly by the QEMU binary depending on which devices the user has specified.
Your idea is to replace this device-tree completely to be able to add extra elements (the EFI signature list, see patch 2/14). Thus a device-tree might be loaded that does not match the user selected devices.
An alternative approach would be to apply all additions to the device-tree as an FDT overlay (or fixup). This would allow the dynamic parts of the QEMU device-tree still to be passed through.
I will take a look at storing the public key as part of the fdt overlay, with a runtime fixup. Although, I think the issue that you are pointing to would be specific to Qemu and not other platforms. But I do see the merit in having the public-key certificate stored as part of an overlay. If I hit any issues while implementing this, I will get back to you. Thanks.
-sughosh
Could you, please, assess the pros and cons of the two approaches with respect to:
- usability for capsule updates
- applicability for non-QEMU systems
- integration of DTB overlays with FIT images for other use cases
Best regards
Heinrich
Signed-off-by: Sughosh Ganu sughosh.ganu@linaro.org
board/emulation/qemu-arm/qemu-arm.c | 2 ++ 1 file changed, 2 insertions(+)
diff --git a/board/emulation/qemu-arm/qemu-arm.c
b/board/emulation/qemu-arm/qemu-arm.c
index f18f2ed7da..e146d1cc50 100644 --- a/board/emulation/qemu-arm/qemu-arm.c +++ b/board/emulation/qemu-arm/qemu-arm.c @@ -89,11 +89,13 @@ int dram_init_banksize(void) return 0; }
+#if defined(CONFIG_OF_BOARD) void *board_fdt_blob_setup(void) { /* QEMU loads a generated DTB for us at the start of RAM. */ return (void *)CONFIG_SYS_SDRAM_BASE; } +#endif
void enable_caches(void) {

On 07.12.20 06:15, Sughosh Ganu wrote:
hello Heinrich,
On Sat, 5 Dec 2020 at 15:01, Heinrich Schuchardt <xypron.glpk@gmx.de mailto:xypron.glpk@gmx.de> wrote:
On 11/26/20 7:40 PM, Sughosh Ganu wrote: > The Qemu platform emulator generates a device tree blob and places it > at the start of the dram, which is then used by u-boot. Use this dtb > only if CONFIG_OF_BOARD is defined. This allows using a different > device tree, using the CONFIG_OF_SEPARATE option. This dtb is attached > to the u-boot binary as a u-boot-fdt.bin file Dear Sughosh, thank your for this series which will allow us to better demonstrate and test capsule updates. I am not sure if the approach that you take at device-trees here is the right one. On QEMU the device-tree is generated on the fly by the QEMU binary depending on which devices the user has specified. Your idea is to replace this device-tree completely to be able to add extra elements (the EFI signature list, see patch 2/14). Thus a device-tree might be loaded that does not match the user selected devices. An alternative approach would be to apply all additions to the device-tree as an FDT overlay (or fixup). This would allow the dynamic parts of the QEMU device-tree still to be passed through.
I will take a look at storing the public key as part of the fdt overlay, with a runtime fixup. Although, I think the issue that you are pointing to would be specific to Qemu and not other platforms. But I do see the merit in having the public-key certificate stored as part of an overlay. If I hit any issues while implementing this, I will get back to you. Thanks.
-sughosh
OpenSBI can supply a device-tree to U-Boot. So this would be an equivalent case.
Best regards
Heinrich
Could you, please, assess the pros and cons of the two approaches with respect to: * usability for capsule updates * applicability for non-QEMU systems * integration of DTB overlays with FIT images for other use cases Best regards Heinrich > > Signed-off-by: Sughosh Ganu <sughosh.ganu@linaro.org <mailto:sughosh.ganu@linaro.org>> > --- > board/emulation/qemu-arm/qemu-arm.c | 2 ++ > 1 file changed, 2 insertions(+) > > diff --git a/board/emulation/qemu-arm/qemu-arm.c b/board/emulation/qemu-arm/qemu-arm.c > index f18f2ed7da..e146d1cc50 100644 > --- a/board/emulation/qemu-arm/qemu-arm.c > +++ b/board/emulation/qemu-arm/qemu-arm.c > @@ -89,11 +89,13 @@ int dram_init_banksize(void) > return 0; > } > > +#if defined(CONFIG_OF_BOARD) > void *board_fdt_blob_setup(void) > { > /* QEMU loads a generated DTB for us at the start of RAM. */ > return (void *)CONFIG_SYS_SDRAM_BASE; > } > +#endif > > void enable_caches(void) > { >

On 07.12.20 13:50, Heinrich Schuchardt wrote:
On 07.12.20 06:15, Sughosh Ganu wrote:
hello Heinrich,
On Sat, 5 Dec 2020 at 15:01, Heinrich Schuchardt <xypron.glpk@gmx.de mailto:xypron.glpk@gmx.de> wrote:
On 11/26/20 7:40 PM, Sughosh Ganu wrote: > The Qemu platform emulator generates a device tree blob and places it > at the start of the dram, which is then used by u-boot. Use this dtb > only if CONFIG_OF_BOARD is defined. This allows using a different > device tree, using the CONFIG_OF_SEPARATE option. This dtb is attached > to the u-boot binary as a u-boot-fdt.bin file Dear Sughosh, thank your for this series which will allow us to better demonstrate and test capsule updates. I am not sure if the approach that you take at device-trees here is the right one. On QEMU the device-tree is generated on the fly by the QEMU binary depending on which devices the user has specified. Your idea is to replace this device-tree completely to be able to add extra elements (the EFI signature list, see patch 2/14). Thus a device-tree might be loaded that does not match the user selected devices. An alternative approach would be to apply all additions to the device-tree as an FDT overlay (or fixup). This would allow the dynamic parts of the QEMU device-tree still to be passed through.
I will take a look at storing the public key as part of the fdt overlay, with a runtime fixup. Although, I think the issue that you are pointing to would be specific to Qemu and not other platforms. But I do see the merit in having the public-key certificate stored as part of an overlay. If I hit any issues while implementing this, I will get back to you. Thanks.
-sughosh
OpenSBI can supply a device-tree to U-Boot. So this would be an equivalent case.
Best regards
Heinrich
<sng_>: "I am unable to think of a simple and elegant way to generate the overlay dtb, since this would require creation of a corresponding dts file, compiling it into a dtb and then using this on the platform."
Jean-Jacques' patch series introduced some of the necessary code:
[PATCH PATCH v6 00/13] Add support for applications of overlays in SPL https://lists.denx.de/pipermail/u-boot/2019-October/387653.html https://patchwork.ozlabs.org/project/uboot/list/?series=137810&state=%2A...
CONFIG_OF_OVERLAY_LIST is used to specify the overlays to be compiled. You will have to remove the CONFIG_SPL_LOAD_FIT_APPLY_OVERLAY and CONFIG_SPL_LOAD_FIT dependency.
Best regards
Heinrich
Could you, please, assess the pros and cons of the two approaches with respect to: * usability for capsule updates * applicability for non-QEMU systems * integration of DTB overlays with FIT images for other use cases Best regards Heinrich > > Signed-off-by: Sughosh Ganu <sughosh.ganu@linaro.org <mailto:sughosh.ganu@linaro.org>> > --- > board/emulation/qemu-arm/qemu-arm.c | 2 ++ > 1 file changed, 2 insertions(+) > > diff --git a/board/emulation/qemu-arm/qemu-arm.c b/board/emulation/qemu-arm/qemu-arm.c > index f18f2ed7da..e146d1cc50 100644 > --- a/board/emulation/qemu-arm/qemu-arm.c > +++ b/board/emulation/qemu-arm/qemu-arm.c > @@ -89,11 +89,13 @@ int dram_init_banksize(void) > return 0; > } > > +#if defined(CONFIG_OF_BOARD) > void *board_fdt_blob_setup(void) > { > /* QEMU loads a generated DTB for us at the start of RAM. */ > return (void *)CONFIG_SYS_SDRAM_BASE; > } > +#endif > > void enable_caches(void) > { >

On Mon, 7 Dec 2020 at 23:28, Heinrich Schuchardt xypron.glpk@gmx.de wrote:
On 07.12.20 13:50, Heinrich Schuchardt wrote:
On 07.12.20 06:15, Sughosh Ganu wrote:
hello Heinrich,
On Sat, 5 Dec 2020 at 15:01, Heinrich Schuchardt <xypron.glpk@gmx.de mailto:xypron.glpk@gmx.de> wrote:
On 11/26/20 7:40 PM, Sughosh Ganu wrote: > The Qemu platform emulator generates a device tree blob and
places it
> at the start of the dram, which is then used by u-boot. Use this
dtb
> only if CONFIG_OF_BOARD is defined. This allows using a different > device tree, using the CONFIG_OF_SEPARATE option. This dtb is
attached
> to the u-boot binary as a u-boot-fdt.bin file Dear Sughosh, thank your for this series which will allow us to better
demonstrate and
test capsule updates. I am not sure if the approach that you take at device-trees here is
the
right one. On QEMU the device-tree is generated on the fly by the QEMU binary depending on which devices the user has specified. Your idea is to replace this device-tree completely to be able to
add
extra elements (the EFI signature list, see patch 2/14). Thus a device-tree might be loaded that does not match the user selected devices. An alternative approach would be to apply all additions to the device-tree as an FDT overlay (or fixup). This would allow the
dynamic
parts of the QEMU device-tree still to be passed through.
I will take a look at storing the public key as part of the fdt overlay, with a runtime fixup. Although, I think the issue that you are pointing to would be specific to Qemu and not other platforms. But I do see the merit in having the public-key certificate stored as part of an overlay. If I hit any issues while implementing this, I will get back to you.
Thanks.
-sughosh
OpenSBI can supply a device-tree to U-Boot. So this would be an equivalent case.
Best regards
Heinrich
<sng_>: "I am unable to think of a simple and elegant way to generate the overlay dtb, since this would require creation of a corresponding dts file, compiling it into a dtb and then using this on the platform."
Jean-Jacques' patch series introduced some of the necessary code:
[PATCH PATCH v6 00/13] Add support for applications of overlays in SPL https://lists.denx.de/pipermail/u-boot/2019-October/387653.html
https://patchwork.ozlabs.org/project/uboot/list/?series=137810&state=%2A...
CONFIG_OF_OVERLAY_LIST is used to specify the overlays to be compiled. You will have to remove the CONFIG_SPL_LOAD_FIT_APPLY_OVERLAY and CONFIG_SPL_LOAD_FIT dependency.
What i meant was that the process to generate the fdt overlay on the host, embed it with the public-key certificate is more cumbersome than the current solution. So, for comparing the embedding the pub-key cert in the fdt overlay, as against the platform dtb
Embedding the certificate in the overlay 1) Generate a skeleton overlay dts file 2) Convert it to a dtb 3) Run it through the mkeficapsule utility to embed the pub-key certificate in the overlay(dtb) 4) Store the overlay dtb on some nv storage on the platform(ESP partition?) 5) Load the overlay and apply it to the platform's dtb 6) Retrieve the certificate from the dtb at the time of capsule authentication
Embedding the certificate in the platform's dtb 1) Run the platform's dtb through the mkeficapsule utility to embed the pub-key certificate 2) Boot the platform with the platform's dtb 3) Retrieve the certificate from the dtb at the time of capsule authentication
You had mentioned OpenSBI passing the dtb to u-boot. Does the OpenSBI generate the device tree for the platform on the fly even for non-qemu platforms. If it does not, the dtb that OpenSBI uses can be run through the mkeficapsule utility to embed the certificate.
-sughosh

On 08.12.20 06:28, Sughosh Ganu wrote:
On Mon, 7 Dec 2020 at 23:28, Heinrich Schuchardt <xypron.glpk@gmx.de mailto:xypron.glpk@gmx.de> wrote:
On 07.12.20 13:50, Heinrich Schuchardt wrote: > On 07.12.20 06:15, Sughosh Ganu wrote: >> hello Heinrich, >> >> On Sat, 5 Dec 2020 at 15:01, Heinrich Schuchardt <xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de> >> <mailto:xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de>>> wrote: >> >> On 11/26/20 7:40 PM, Sughosh Ganu wrote: >> > The Qemu platform emulator generates a device tree blob and places it >> > at the start of the dram, which is then used by u-boot. Use this dtb >> > only if CONFIG_OF_BOARD is defined. This allows using a different >> > device tree, using the CONFIG_OF_SEPARATE option. This dtb is attached >> > to the u-boot binary as a u-boot-fdt.bin file >> >> Dear Sughosh, >> >> thank your for this series which will allow us to better demonstrate and >> test capsule updates. >> >> I am not sure if the approach that you take at device-trees here is the >> right one. >> >> On QEMU the device-tree is generated on the fly by the QEMU binary >> depending on which devices the user has specified. >> >> Your idea is to replace this device-tree completely to be able to add >> extra elements (the EFI signature list, see patch 2/14). Thus a >> device-tree might be loaded that does not match the user selected >> devices. >> >> An alternative approach would be to apply all additions to the >> device-tree as an FDT overlay (or fixup). This would allow the dynamic >> parts of the QEMU device-tree still to be passed through. >> >> >> I will take a look at storing the public key as part of the fdt overlay, >> with a runtime fixup. Although, I think the issue that you are pointing >> to would be specific to Qemu and not other platforms. But I do see the >> merit in having the public-key certificate stored as part of an overlay. >> If I hit any issues while implementing this, I will get back to you. Thanks. >> >> -sughosh > > OpenSBI can supply a device-tree to U-Boot. So this would be an > equivalent case. > > Best regards > > Heinrich <sng_>: "I am unable to think of a simple and elegant way to generate the overlay dtb, since this would require creation of a corresponding dts file, compiling it into a dtb and then using this on the platform." Jean-Jacques' patch series introduced some of the necessary code: [PATCH PATCH v6 00/13] Add support for applications of overlays in SPL https://lists.denx.de/pipermail/u-boot/2019-October/387653.html <https://lists.denx.de/pipermail/u-boot/2019-October/387653.html> https://patchwork.ozlabs.org/project/uboot/list/?series=137810&state=%2A&archive=both <https://patchwork.ozlabs.org/project/uboot/list/?series=137810&state=%2A&archive=both> CONFIG_OF_OVERLAY_LIST is used to specify the overlays to be compiled. You will have to remove the CONFIG_SPL_LOAD_FIT_APPLY_OVERLAY and CONFIG_SPL_LOAD_FIT dependency.
What i meant was that the process to generate the fdt overlay on the host, embed it with the public-key certificate is more cumbersome than the current solution. So, for comparing the embedding the pub-key cert in the fdt overlay, as against the platform dtb
Embedding the certificate in the overlay
- Generate a skeleton overlay dts file
- Convert it to a dtb
- Run it through the mkeficapsule utility to embed the pub-key
certificate in the overlay(dtb) 4) Store the overlay dtb on some nv storage on the platform(ESP partition?) 5) Load the overlay and apply it to the platform's dtb 6) Retrieve the certificate from the dtb at the time of capsule authentication
Embedding the certificate in the platform's dtb
- Run the platform's dtb through the mkeficapsule utility to embed
the pub-key certificate 2) Boot the platform with the platform's dtb 3) Retrieve the certificate from the dtb at the time of capsule authentication
You had mentioned OpenSBI passing the dtb to u-boot. Does the OpenSBI generate the device tree for the platform on the fly even for non-qemu platforms. If it does not, the dtb that OpenSBI uses can be run through the mkeficapsule utility to embed the certificate.
OpenSBI applies fix-ups before passing the device-tree. The prototype device-tree can either be built into OpenSBI or can be passed from an earlier firmware stage.
Best regards
Heinrich
-sughosh

On Tue, 8 Dec 2020 at 14:32, Heinrich Schuchardt xypron.glpk@gmx.de wrote:
On 08.12.20 06:28, Sughosh Ganu wrote:
On Mon, 7 Dec 2020 at 23:28, Heinrich Schuchardt <xypron.glpk@gmx.de mailto:xypron.glpk@gmx.de> wrote:
On 07.12.20 13:50, Heinrich Schuchardt wrote: > On 07.12.20 06:15, Sughosh Ganu wrote: >> hello Heinrich, >> >> On Sat, 5 Dec 2020 at 15:01, Heinrich Schuchardt <xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de> >> <mailto:xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de>>> wrote: >> >> On 11/26/20 7:40 PM, Sughosh Ganu wrote: >> > The Qemu platform emulator generates a device tree blob and places it >> > at the start of the dram, which is then used by u-boot. Use this dtb >> > only if CONFIG_OF_BOARD is defined. This allows using a different >> > device tree, using the CONFIG_OF_SEPARATE option. This dtb is attached >> > to the u-boot binary as a u-boot-fdt.bin file >> >> Dear Sughosh, >> >> thank your for this series which will allow us to better demonstrate and >> test capsule updates. >> >> I am not sure if the approach that you take at device-trees here is the >> right one. >> >> On QEMU the device-tree is generated on the fly by the QEMU binary >> depending on which devices the user has specified. >> >> Your idea is to replace this device-tree completely to be able to add >> extra elements (the EFI signature list, see patch 2/14). Thus
a
>> device-tree might be loaded that does not match the user
selected
>> devices. >> >> An alternative approach would be to apply all additions to the >> device-tree as an FDT overlay (or fixup). This would allow the dynamic >> parts of the QEMU device-tree still to be passed through. >> >> >> I will take a look at storing the public key as part of the fdt overlay, >> with a runtime fixup. Although, I think the issue that you are pointing >> to would be specific to Qemu and not other platforms. But I do see the >> merit in having the public-key certificate stored as part of an overlay. >> If I hit any issues while implementing this, I will get back to you. Thanks. >> >> -sughosh > > OpenSBI can supply a device-tree to U-Boot. So this would be an > equivalent case. > > Best regards > > Heinrich <sng_>: "I am unable to think of a simple and elegant way to generate the overlay dtb, since this would require creation of a corresponding dts file, compiling it into a dtb and then using this on the
platform."
Jean-Jacques' patch series introduced some of the necessary code: [PATCH PATCH v6 00/13] Add support for applications of overlays in
SPL
https://lists.denx.de/pipermail/u-boot/2019-October/387653.html <https://lists.denx.de/pipermail/u-boot/2019-October/387653.html>
https://patchwork.ozlabs.org/project/uboot/list/?series=137810&state=%2A...
<
https://patchwork.ozlabs.org/project/uboot/list/?series=137810&state=%2A...
CONFIG_OF_OVERLAY_LIST is used to specify the overlays to be
compiled.
You will have to remove the CONFIG_SPL_LOAD_FIT_APPLY_OVERLAY and CONFIG_SPL_LOAD_FIT dependency.
What i meant was that the process to generate the fdt overlay on the host, embed it with the public-key certificate is more cumbersome than the current solution. So, for comparing the embedding the pub-key cert in the fdt overlay, as against the platform dtb
Embedding the certificate in the overlay
- Generate a skeleton overlay dts file
- Convert it to a dtb
- Run it through the mkeficapsule utility to embed the pub-key
certificate in the overlay(dtb) 4) Store the overlay dtb on some nv storage on the platform(ESP
partition?)
- Load the overlay and apply it to the platform's dtb
- Retrieve the certificate from the dtb at the time of capsule
authentication
Embedding the certificate in the platform's dtb
- Run the platform's dtb through the mkeficapsule utility to embed
the pub-key certificate 2) Boot the platform with the platform's dtb 3) Retrieve the certificate from the dtb at the time of capsule authentication
You had mentioned OpenSBI passing the dtb to u-boot. Does the OpenSBI generate the device tree for the platform on the fly even for non-qemu platforms. If it does not, the dtb that OpenSBI uses can be run through the mkeficapsule utility to embed the certificate.
OpenSBI applies fix-ups before passing the device-tree. The prototype device-tree can either be built into OpenSBI or can be passed from an earlier firmware stage.
So, if qemu if the only platform where the device tree is generated on the fly, and passed along, based on the comparison that i have stated above for the two scenarios(overlay vs platform dtb), I feel that using the platform's dtb and embedding the certificate is more easier to use than using the overlay.
Even on the qemu platform, I retrieved the dtb from the platform, for the given set of devices and interfaces used, and then ran the dtb through the mkeficapsule utility.
-sughosh

On 12/8/20 10:19 AM, Sughosh Ganu wrote:
On Tue, 8 Dec 2020 at 14:32, Heinrich Schuchardt <xypron.glpk@gmx.de mailto:xypron.glpk@gmx.de> wrote:
On 08.12.20 06:28, Sughosh Ganu wrote: > > On Mon, 7 Dec 2020 at 23:28, Heinrich Schuchardt <xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de> > <mailto:xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de>>> wrote: > > On 07.12.20 13:50, Heinrich Schuchardt wrote: > > On 07.12.20 06:15, Sughosh Ganu wrote: > >> hello Heinrich, > >> > >> On Sat, 5 Dec 2020 at 15:01, Heinrich Schuchardt > <xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de> <mailto:xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de>> > >> <mailto:xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de> <mailto:xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de>>>> wrote: > >> > >> On 11/26/20 7:40 PM, Sughosh Ganu wrote: > >> > The Qemu platform emulator generates a device tree blob and > places it > >> > at the start of the dram, which is then used by u-boot. Use > this dtb > >> > only if CONFIG_OF_BOARD is defined. This allows using a > different > >> > device tree, using the CONFIG_OF_SEPARATE option. This dtb > is attached > >> > to the u-boot binary as a u-boot-fdt.bin file > >> > >> Dear Sughosh, > >> > >> thank your for this series which will allow us to better > demonstrate and > >> test capsule updates. > >> > >> I am not sure if the approach that you take at device-trees > here is the > >> right one. > >> > >> On QEMU the device-tree is generated on the fly by the QEMU > binary > >> depending on which devices the user has specified. > >> > >> Your idea is to replace this device-tree completely to be > able to add > >> extra elements (the EFI signature list, see patch 2/14). Thus a > >> device-tree might be loaded that does not match the user selected > >> devices. > >> > >> An alternative approach would be to apply all additions to the > >> device-tree as an FDT overlay (or fixup). This would allow > the dynamic > >> parts of the QEMU device-tree still to be passed through. > >> > >> > >> I will take a look at storing the public key as part of the fdt > overlay, > >> with a runtime fixup. Although, I think the issue that you are > pointing > >> to would be specific to Qemu and not other platforms. But I do > see the > >> merit in having the public-key certificate stored as part of an > overlay. > >> If I hit any issues while implementing this, I will get back to > you. Thanks. > >> > >> -sughosh > > > > OpenSBI can supply a device-tree to U-Boot. So this would be an > > equivalent case. > > > > Best regards > > > > Heinrich > > <sng_>: "I am unable to think of a simple and elegant way to generate > the overlay dtb, since this would require creation of a corresponding > dts file, compiling it into a dtb and then using this on the platform." > > Jean-Jacques' patch series introduced some of the necessary code: > > [PATCH PATCH v6 00/13] Add support for applications of overlays in SPL > https://lists.denx.de/pipermail/u-boot/2019-October/387653.html <https://lists.denx.de/pipermail/u-boot/2019-October/387653.html> > <https://lists.denx.de/pipermail/u-boot/2019-October/387653.html <https://lists.denx.de/pipermail/u-boot/2019-October/387653.html>> > https://patchwork.ozlabs.org/project/uboot/list/?series=137810&state=%2A&archive=both <https://patchwork.ozlabs.org/project/uboot/list/?series=137810&state=%2A&archive=both> > <https://patchwork.ozlabs.org/project/uboot/list/?series=137810&state=%2A&archive=both <https://patchwork.ozlabs.org/project/uboot/list/?series=137810&state=%2A&archive=both>> > > CONFIG_OF_OVERLAY_LIST is used to specify the overlays to be compiled. > You will have to remove the CONFIG_SPL_LOAD_FIT_APPLY_OVERLAY and > CONFIG_SPL_LOAD_FIT dependency. > > > What i meant was that the process to generate the fdt overlay on the > host, embed it with the public-key certificate is more cumbersome than > the current solution. So, for comparing the embedding the pub-key cert > in the fdt overlay, as against the platform dtb > > Embedding the certificate in the overlay > 1) Generate a skeleton overlay dts file > 2) Convert it to a dtb > 3) Run it through the mkeficapsule utility to embed the pub-key > certificate in the overlay(dtb) > 4) Store the overlay dtb on some nv storage on the platform(ESP partition?) > 5) Load the overlay and apply it to the platform's dtb > 6) Retrieve the certificate from the dtb at the time of capsule > authentication > > Embedding the certificate in the platform's dtb > 1) Run the platform's dtb through the mkeficapsule utility to embed > the pub-key certificate > 2) Boot the platform with the platform's dtb > 3) Retrieve the certificate from the dtb at the time of capsule > authentication > > You had mentioned OpenSBI passing the dtb to u-boot. Does the OpenSBI > generate the device tree for the platform on the fly even for non-qemu > platforms. If it does not, the dtb that OpenSBI uses can be run through > the mkeficapsule utility to embed the certificate. OpenSBI applies fix-ups before passing the device-tree. The prototype device-tree can either be built into OpenSBI or can be passed from an earlier firmware stage.
So, if qemu if the only platform where the device tree is generated on the fly, and passed along, based on the comparison that i have stated above for the two scenarios(overlay vs platform dtb), I feel that using the platform's dtb and embedding the certificate is more easier to use than using the overlay.
Even on the qemu platform, I retrieved the dtb from the platform, for the given set of devices and interfaces used, and then ran the dtb through the mkeficapsule utility.
-sughosh
As I understand the whole series is not about productive firmware updates for QEMU but about a demonstration how it could be used and allowing testing.
Wouldn't it be the enough for testing to apply an overlay with the public key using the fdt command?
Best regards
Heinrich

On Wed, 9 Dec 2020 at 03:24, Heinrich Schuchardt xypron.glpk@gmx.de wrote:
On 12/8/20 10:19 AM, Sughosh Ganu wrote:
On Tue, 8 Dec 2020 at 14:32, Heinrich Schuchardt <xypron.glpk@gmx.de mailto:xypron.glpk@gmx.de> wrote:
On 08.12.20 06:28, Sughosh Ganu wrote: > > On Mon, 7 Dec 2020 at 23:28, Heinrich Schuchardt <xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de> > <mailto:xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de>>> wrote: > > On 07.12.20 13:50, Heinrich Schuchardt wrote: > > On 07.12.20 06:15, Sughosh Ganu wrote: > >> hello Heinrich, > >> > >> On Sat, 5 Dec 2020 at 15:01, Heinrich Schuchardt > <xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de> <mailto:xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de>> > >> <mailto:xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de> <mailto:xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de>>>> wrote: > >> > >> On 11/26/20 7:40 PM, Sughosh Ganu wrote: > >> > The Qemu platform emulator generates a device tree blob and > places it > >> > at the start of the dram, which is then used by u-boot. Use > this dtb > >> > only if CONFIG_OF_BOARD is defined. This allows
using a
> different > >> > device tree, using the CONFIG_OF_SEPARATE option. This dtb > is attached > >> > to the u-boot binary as a u-boot-fdt.bin file > >> > >> Dear Sughosh, > >> > >> thank your for this series which will allow us to
better
> demonstrate and > >> test capsule updates. > >> > >> I am not sure if the approach that you take at device-trees > here is the > >> right one. > >> > >> On QEMU the device-tree is generated on the fly by the QEMU > binary > >> depending on which devices the user has specified. > >> > >> Your idea is to replace this device-tree completely to
be
> able to add > >> extra elements (the EFI signature list, see patch 2/14). Thus a > >> device-tree might be loaded that does not match the user selected > >> devices. > >> > >> An alternative approach would be to apply all additions to the > >> device-tree as an FDT overlay (or fixup). This would
allow
> the dynamic > >> parts of the QEMU device-tree still to be passed
through.
> >> > >> > >> I will take a look at storing the public key as part of the fdt > overlay, > >> with a runtime fixup. Although, I think the issue that you
are
> pointing > >> to would be specific to Qemu and not other platforms. But
I do
> see the > >> merit in having the public-key certificate stored as part of an > overlay. > >> If I hit any issues while implementing this, I will get back to > you. Thanks. > >> > >> -sughosh > > > > OpenSBI can supply a device-tree to U-Boot. So this would
be an
> > equivalent case. > > > > Best regards > > > > Heinrich > > <sng_>: "I am unable to think of a simple and elegant way to generate > the overlay dtb, since this would require creation of a corresponding > dts file, compiling it into a dtb and then using this on the platform." > > Jean-Jacques' patch series introduced some of the necessary
code:
> > [PATCH PATCH v6 00/13] Add support for applications of overlays in SPL > https://lists.denx.de/pipermail/u-boot/2019-October/387653.html <https://lists.denx.de/pipermail/u-boot/2019-October/387653.html> > <https://lists.denx.de/pipermail/u-boot/2019-October/387653.html <https://lists.denx.de/pipermail/u-boot/2019-October/387653.html>> >
https://patchwork.ozlabs.org/project/uboot/list/?series=137810&state=%2A...
<
https://patchwork.ozlabs.org/project/uboot/list/?series=137810&state=%2A...
> <
https://patchwork.ozlabs.org/project/uboot/list/?series=137810&state=%2A... < https://patchwork.ozlabs.org/project/uboot/list/?series=137810&state=%2A...
> > CONFIG_OF_OVERLAY_LIST is used to specify the overlays to be compiled. > You will have to remove the CONFIG_SPL_LOAD_FIT_APPLY_OVERLAY
and
> CONFIG_SPL_LOAD_FIT dependency. > > > What i meant was that the process to generate the fdt overlay on
the
> host, embed it with the public-key certificate is more cumbersome than > the current solution. So, for comparing the embedding the pub-key cert > in the fdt overlay, as against the platform dtb > > Embedding the certificate in the overlay > 1) Generate a skeleton overlay dts file > 2) Convert it to a dtb > 3) Run it through the mkeficapsule utility to embed the pub-key > certificate in the overlay(dtb) > 4) Store the overlay dtb on some nv storage on the platform(ESP partition?) > 5) Load the overlay and apply it to the platform's dtb > 6) Retrieve the certificate from the dtb at the time of capsule > authentication > > Embedding the certificate in the platform's dtb > 1) Run the platform's dtb through the mkeficapsule utility to
embed
> the pub-key certificate > 2) Boot the platform with the platform's dtb > 3) Retrieve the certificate from the dtb at the time of capsule > authentication > > You had mentioned OpenSBI passing the dtb to u-boot. Does the
OpenSBI
> generate the device tree for the platform on the fly even for non-qemu > platforms. If it does not, the dtb that OpenSBI uses can be run through > the mkeficapsule utility to embed the certificate. OpenSBI applies fix-ups before passing the device-tree. The prototype device-tree can either be built into OpenSBI or can be passed from an earlier firmware stage.
So, if qemu if the only platform where the device tree is generated on the fly, and passed along, based on the comparison that i have stated above for the two scenarios(overlay vs platform dtb), I feel that using the platform's dtb and embedding the certificate is more easier to use than using the overlay.
Even on the qemu platform, I retrieved the dtb from the platform, for the given set of devices and interfaces used, and then ran the dtb through the mkeficapsule utility.
-sughosh
As I understand the whole series is not about productive firmware updates for QEMU but about a demonstration how it could be used and allowing testing.
True, but the process of embedding the public key certificate into the platform's dtb for use in capsule authentication can be common for all platforms. If you are fine, we can keep this patch to be used on other non-qemu platforms, while i work on the overlay method for qemu. Will you be fine with this approach.
-sughosh

On 12/9/20 6:25 AM, Sughosh Ganu wrote:
On Wed, 9 Dec 2020 at 03:24, Heinrich Schuchardt xypron.glpk@gmx.de wrote:
On 12/8/20 10:19 AM, Sughosh Ganu wrote:
On Tue, 8 Dec 2020 at 14:32, Heinrich Schuchardt <xypron.glpk@gmx.de mailto:xypron.glpk@gmx.de> wrote:
On 08.12.20 06:28, Sughosh Ganu wrote: > > On Mon, 7 Dec 2020 at 23:28, Heinrich Schuchardt <xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de> > <mailto:xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de>>> wrote: > > On 07.12.20 13:50, Heinrich Schuchardt wrote: > > On 07.12.20 06:15, Sughosh Ganu wrote: > >> hello Heinrich, > >> > >> On Sat, 5 Dec 2020 at 15:01, Heinrich Schuchardt > <xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de> <mailto:xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de>> > >> <mailto:xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de> <mailto:xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de>>>> wrote: > >> > >> On 11/26/20 7:40 PM, Sughosh Ganu wrote: > >> > The Qemu platform emulator generates a device tree blob and > places it > >> > at the start of the dram, which is then used by u-boot. Use > this dtb > >> > only if CONFIG_OF_BOARD is defined. This allows
using a
> different > >> > device tree, using the CONFIG_OF_SEPARATE option. This dtb > is attached > >> > to the u-boot binary as a u-boot-fdt.bin file > >> > >> Dear Sughosh, > >> > >> thank your for this series which will allow us to
better
> demonstrate and > >> test capsule updates. > >> > >> I am not sure if the approach that you take at device-trees > here is the > >> right one. > >> > >> On QEMU the device-tree is generated on the fly by the QEMU > binary > >> depending on which devices the user has specified. > >> > >> Your idea is to replace this device-tree completely to
be
> able to add > >> extra elements (the EFI signature list, see patch 2/14). Thus a > >> device-tree might be loaded that does not match the user selected > >> devices. > >> > >> An alternative approach would be to apply all additions to the > >> device-tree as an FDT overlay (or fixup). This would
allow
> the dynamic > >> parts of the QEMU device-tree still to be passed
through.
> >> > >> > >> I will take a look at storing the public key as part of the fdt > overlay, > >> with a runtime fixup. Although, I think the issue that you
are
> pointing > >> to would be specific to Qemu and not other platforms. But
I do
> see the > >> merit in having the public-key certificate stored as part of an > overlay. > >> If I hit any issues while implementing this, I will get back to > you. Thanks. > >> > >> -sughosh > > > > OpenSBI can supply a device-tree to U-Boot. So this would
be an
> > equivalent case. > > > > Best regards > > > > Heinrich > > <sng_>: "I am unable to think of a simple and elegant way to generate > the overlay dtb, since this would require creation of a corresponding > dts file, compiling it into a dtb and then using this on the platform." > > Jean-Jacques' patch series introduced some of the necessary
code:
> > [PATCH PATCH v6 00/13] Add support for applications of overlays in SPL > https://lists.denx.de/pipermail/u-boot/2019-October/387653.html <https://lists.denx.de/pipermail/u-boot/2019-October/387653.html> > <https://lists.denx.de/pipermail/u-boot/2019-October/387653.html <https://lists.denx.de/pipermail/u-boot/2019-October/387653.html>> >
https://patchwork.ozlabs.org/project/uboot/list/?series=137810&state=%2A...
<
https://patchwork.ozlabs.org/project/uboot/list/?series=137810&state=%2A...
> <
https://patchwork.ozlabs.org/project/uboot/list/?series=137810&state=%2A... < https://patchwork.ozlabs.org/project/uboot/list/?series=137810&state=%2A...
> > CONFIG_OF_OVERLAY_LIST is used to specify the overlays to be compiled. > You will have to remove the CONFIG_SPL_LOAD_FIT_APPLY_OVERLAY
and
> CONFIG_SPL_LOAD_FIT dependency. > > > What i meant was that the process to generate the fdt overlay on
the
> host, embed it with the public-key certificate is more cumbersome than > the current solution. So, for comparing the embedding the pub-key cert > in the fdt overlay, as against the platform dtb > > Embedding the certificate in the overlay > 1) Generate a skeleton overlay dts file > 2) Convert it to a dtb > 3) Run it through the mkeficapsule utility to embed the pub-key > certificate in the overlay(dtb) > 4) Store the overlay dtb on some nv storage on the platform(ESP partition?) > 5) Load the overlay and apply it to the platform's dtb > 6) Retrieve the certificate from the dtb at the time of capsule > authentication > > Embedding the certificate in the platform's dtb > 1) Run the platform's dtb through the mkeficapsule utility to
embed
> the pub-key certificate > 2) Boot the platform with the platform's dtb > 3) Retrieve the certificate from the dtb at the time of capsule > authentication > > You had mentioned OpenSBI passing the dtb to u-boot. Does the
OpenSBI
> generate the device tree for the platform on the fly even for non-qemu > platforms. If it does not, the dtb that OpenSBI uses can be run through > the mkeficapsule utility to embed the certificate. OpenSBI applies fix-ups before passing the device-tree. The prototype device-tree can either be built into OpenSBI or can be passed from an earlier firmware stage.
So, if qemu if the only platform where the device tree is generated on the fly, and passed along, based on the comparison that i have stated above for the two scenarios(overlay vs platform dtb), I feel that using the platform's dtb and embedding the certificate is more easier to use than using the overlay.
Even on the qemu platform, I retrieved the dtb from the platform, for the given set of devices and interfaces used, and then ran the dtb through the mkeficapsule utility.
-sughosh
As I understand the whole series is not about productive firmware updates for QEMU but about a demonstration how it could be used and allowing testing.
True, but the process of embedding the public key certificate into the platform's dtb for use in capsule authentication can be common for all platforms. If you are fine, we can keep this patch to be used on other non-qemu platforms, while i work on the overlay method for qemu. Will you be fine with this approach.
-sughosh
This patch only changes board/emulation/qemu-arm/qemu-arm.c. Hence "keep this patch to be used on other non-qemu platforms" is not applicable to this specific patch.
Best regards
Heinrich

On Wed, 9 Dec 2020 at 12:56, Heinrich Schuchardt xypron.glpk@gmx.de wrote:
On 12/9/20 6:25 AM, Sughosh Ganu wrote:
On Wed, 9 Dec 2020 at 03:24, Heinrich Schuchardt xypron.glpk@gmx.de
wrote:
On 12/8/20 10:19 AM, Sughosh Ganu wrote:
On Tue, 8 Dec 2020 at 14:32, Heinrich Schuchardt <xypron.glpk@gmx.de mailto:xypron.glpk@gmx.de> wrote:
On 08.12.20 06:28, Sughosh Ganu wrote: > > On Mon, 7 Dec 2020 at 23:28, Heinrich Schuchardt <xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de> > <mailto:xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de>>>
wrote:
> > On 07.12.20 13:50, Heinrich Schuchardt wrote: > > On 07.12.20 06:15, Sughosh Ganu wrote: > >> hello Heinrich, > >> > >> On Sat, 5 Dec 2020 at 15:01, Heinrich Schuchardt > <xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de> <mailto:xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de>> > >> <mailto:xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de> <mailto:xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de>>>> wrote: > >> > >> On 11/26/20 7:40 PM, Sughosh Ganu wrote: > >> > The Qemu platform emulator generates a device tree blob and > places it > >> > at the start of the dram, which is then used by u-boot. Use > this dtb > >> > only if CONFIG_OF_BOARD is defined. This allows
using a
> different > >> > device tree, using the CONFIG_OF_SEPARATE option. This dtb > is attached > >> > to the u-boot binary as a u-boot-fdt.bin file > >> > >> Dear Sughosh, > >> > >> thank your for this series which will allow us to
better
> demonstrate and > >> test capsule updates. > >> > >> I am not sure if the approach that you take at device-trees > here is the > >> right one. > >> > >> On QEMU the device-tree is generated on the fly by
the
QEMU > binary > >> depending on which devices the user has specified. > >> > >> Your idea is to replace this device-tree completely
to
be
> able to add > >> extra elements (the EFI signature list, see patch 2/14). Thus a > >> device-tree might be loaded that does not match the user selected > >> devices. > >> > >> An alternative approach would be to apply all additions to the > >> device-tree as an FDT overlay (or fixup). This would
allow
> the dynamic > >> parts of the QEMU device-tree still to be passed
through.
> >> > >> > >> I will take a look at storing the public key as part of the fdt > overlay, > >> with a runtime fixup. Although, I think the issue that
you
are
> pointing > >> to would be specific to Qemu and not other platforms.
But
I do
> see the > >> merit in having the public-key certificate stored as
part
of an > overlay. > >> If I hit any issues while implementing this, I will get back to > you. Thanks. > >> > >> -sughosh > > > > OpenSBI can supply a device-tree to U-Boot. So this would
be an
> > equivalent case. > > > > Best regards > > > > Heinrich > > <sng_>: "I am unable to think of a simple and elegant way
to
generate > the overlay dtb, since this would require creation of a corresponding > dts file, compiling it into a dtb and then using this on
the
platform." > > Jean-Jacques' patch series introduced some of the necessary
code:
> > [PATCH PATCH v6 00/13] Add support for applications of overlays in SPL >
https://lists.denx.de/pipermail/u-boot/2019-October/387653.html
<https://lists.denx.de/pipermail/u-boot/2019-October/387653.html> > <
https://lists.denx.de/pipermail/u-boot/2019-October/387653.html
<https://lists.denx.de/pipermail/u-boot/2019-October/387653.html
>
https://patchwork.ozlabs.org/project/uboot/list/?series=137810&state=%2A...
<
https://patchwork.ozlabs.org/project/uboot/list/?series=137810&state=%2A...
> <
https://patchwork.ozlabs.org/project/uboot/list/?series=137810&state=%2A...
<
https://patchwork.ozlabs.org/project/uboot/list/?series=137810&state=%2A...
> > CONFIG_OF_OVERLAY_LIST is used to specify the overlays to
be
compiled. > You will have to remove the
CONFIG_SPL_LOAD_FIT_APPLY_OVERLAY
and
> CONFIG_SPL_LOAD_FIT dependency. > > > What i meant was that the process to generate the fdt overlay
on
the
> host, embed it with the public-key certificate is more
cumbersome
than > the current solution. So, for comparing the embedding the
pub-key
cert > in the fdt overlay, as against the platform dtb > > Embedding the certificate in the overlay > 1) Generate a skeleton overlay dts file > 2) Convert it to a dtb > 3) Run it through the mkeficapsule utility to embed the pub-key > certificate in the overlay(dtb) > 4) Store the overlay dtb on some nv storage on the platform(ESP partition?) > 5) Load the overlay and apply it to the platform's dtb > 6) Retrieve the certificate from the dtb at the time of capsule > authentication > > Embedding the certificate in the platform's dtb > 1) Run the platform's dtb through the mkeficapsule utility to
embed
> the pub-key certificate > 2) Boot the platform with the platform's dtb > 3) Retrieve the certificate from the dtb at the time of capsule > authentication > > You had mentioned OpenSBI passing the dtb to u-boot. Does the
OpenSBI
> generate the device tree for the platform on the fly even for non-qemu > platforms. If it does not, the dtb that OpenSBI uses can be run through > the mkeficapsule utility to embed the certificate. OpenSBI applies fix-ups before passing the device-tree. The
prototype
device-tree can either be built into OpenSBI or can be passed
from an
earlier firmware stage.
So, if qemu if the only platform where the device tree is generated on the fly, and passed along, based on the comparison that i have stated above for the two scenarios(overlay vs platform dtb), I feel that using the platform's dtb and embedding the certificate is more easier to use than using the overlay.
Even on the qemu platform, I retrieved the dtb from the platform, for the given set of devices and interfaces used, and then ran the dtb through the mkeficapsule utility.
-sughosh
As I understand the whole series is not about productive firmware updates for QEMU but about a demonstration how it could be used and allowing testing.
True, but the process of embedding the public key certificate into the platform's dtb for use in capsule authentication can be common for all platforms. If you are fine, we can keep this patch to be used on other non-qemu platforms, while i work on the overlay method for qemu. Will you be fine with this approach.
-sughosh
This patch only changes board/emulation/qemu-arm/qemu-arm.c. Hence "keep this patch to be used on other non-qemu platforms" is not applicable to this specific patch.
Sorry, i did not mean this patch, but patch 02/14, "mkeficapsule: Add support for embedding public key in a dtb". I meant that that patch can be used for other non-qemu platforms.
-sughosh

On Wed, 9 Dec 2020 at 03:24, Heinrich Schuchardt xypron.glpk@gmx.de wrote:
On 12/8/20 10:19 AM, Sughosh Ganu wrote:
On Tue, 8 Dec 2020 at 14:32, Heinrich Schuchardt <xypron.glpk@gmx.de mailto:xypron.glpk@gmx.de> wrote:
On 08.12.20 06:28, Sughosh Ganu wrote: > > On Mon, 7 Dec 2020 at 23:28, Heinrich Schuchardt <xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de> > <mailto:xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de>>> wrote: > > On 07.12.20 13:50, Heinrich Schuchardt wrote: > > On 07.12.20 06:15, Sughosh Ganu wrote: > >> hello Heinrich, > >> > >> On Sat, 5 Dec 2020 at 15:01, Heinrich Schuchardt > <xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de> <mailto:xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de>> > >> <mailto:xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de> <mailto:xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de>>>> wrote: > >> > >> On 11/26/20 7:40 PM, Sughosh Ganu wrote: > >> > The Qemu platform emulator generates a device tree blob and > places it > >> > at the start of the dram, which is then used by u-boot. Use > this dtb > >> > only if CONFIG_OF_BOARD is defined. This allows
using a
> different > >> > device tree, using the CONFIG_OF_SEPARATE option. This dtb > is attached > >> > to the u-boot binary as a u-boot-fdt.bin file > >> > >> Dear Sughosh, > >> > >> thank your for this series which will allow us to
better
> demonstrate and > >> test capsule updates. > >> > >> I am not sure if the approach that you take at device-trees > here is the > >> right one. > >> > >> On QEMU the device-tree is generated on the fly by the QEMU > binary > >> depending on which devices the user has specified. > >> > >> Your idea is to replace this device-tree completely to
be
> able to add > >> extra elements (the EFI signature list, see patch 2/14). Thus a > >> device-tree might be loaded that does not match the user selected > >> devices. > >> > >> An alternative approach would be to apply all additions to the > >> device-tree as an FDT overlay (or fixup). This would
allow
> the dynamic > >> parts of the QEMU device-tree still to be passed
through.
> >> > >> > >> I will take a look at storing the public key as part of the fdt > overlay, > >> with a runtime fixup. Although, I think the issue that you
are
> pointing > >> to would be specific to Qemu and not other platforms. But
I do
> see the > >> merit in having the public-key certificate stored as part of an > overlay. > >> If I hit any issues while implementing this, I will get back to > you. Thanks. > >> > >> -sughosh > > > > OpenSBI can supply a device-tree to U-Boot. So this would
be an
> > equivalent case. > > > > Best regards > > > > Heinrich > > <sng_>: "I am unable to think of a simple and elegant way to generate > the overlay dtb, since this would require creation of a corresponding > dts file, compiling it into a dtb and then using this on the platform." > > Jean-Jacques' patch series introduced some of the necessary
code:
> > [PATCH PATCH v6 00/13] Add support for applications of overlays in SPL > https://lists.denx.de/pipermail/u-boot/2019-October/387653.html <https://lists.denx.de/pipermail/u-boot/2019-October/387653.html> > <https://lists.denx.de/pipermail/u-boot/2019-October/387653.html <https://lists.denx.de/pipermail/u-boot/2019-October/387653.html>> >
https://patchwork.ozlabs.org/project/uboot/list/?series=137810&state=%2A...
<
https://patchwork.ozlabs.org/project/uboot/list/?series=137810&state=%2A...
> <
https://patchwork.ozlabs.org/project/uboot/list/?series=137810&state=%2A... < https://patchwork.ozlabs.org/project/uboot/list/?series=137810&state=%2A...
> > CONFIG_OF_OVERLAY_LIST is used to specify the overlays to be compiled. > You will have to remove the CONFIG_SPL_LOAD_FIT_APPLY_OVERLAY
and
> CONFIG_SPL_LOAD_FIT dependency. > > > What i meant was that the process to generate the fdt overlay on
the
> host, embed it with the public-key certificate is more cumbersome than > the current solution. So, for comparing the embedding the pub-key cert > in the fdt overlay, as against the platform dtb > > Embedding the certificate in the overlay > 1) Generate a skeleton overlay dts file > 2) Convert it to a dtb > 3) Run it through the mkeficapsule utility to embed the pub-key > certificate in the overlay(dtb) > 4) Store the overlay dtb on some nv storage on the platform(ESP partition?) > 5) Load the overlay and apply it to the platform's dtb > 6) Retrieve the certificate from the dtb at the time of capsule > authentication > > Embedding the certificate in the platform's dtb > 1) Run the platform's dtb through the mkeficapsule utility to
embed
> the pub-key certificate > 2) Boot the platform with the platform's dtb > 3) Retrieve the certificate from the dtb at the time of capsule > authentication > > You had mentioned OpenSBI passing the dtb to u-boot. Does the
OpenSBI
> generate the device tree for the platform on the fly even for non-qemu > platforms. If it does not, the dtb that OpenSBI uses can be run through > the mkeficapsule utility to embed the certificate. OpenSBI applies fix-ups before passing the device-tree. The prototype device-tree can either be built into OpenSBI or can be passed from an earlier firmware stage.
So, if qemu if the only platform where the device tree is generated on the fly, and passed along, based on the comparison that i have stated above for the two scenarios(overlay vs platform dtb), I feel that using the platform's dtb and embedding the certificate is more easier to use than using the overlay.
Even on the qemu platform, I retrieved the dtb from the platform, for the given set of devices and interfaces used, and then ran the dtb through the mkeficapsule utility.
-sughosh
As I understand the whole series is not about productive firmware updates for QEMU but about a demonstration how it could be used and allowing testing.
Wouldn't it be the enough for testing to apply an overlay with the public key using the fdt command?
Is there a way I can build qemu such that the device-tree that is generated is with the __symbols__ node. Without this node, I am unable to apply the fdt overlay on the base platform devicetree.
-sughosh

On 15.12.20 12:10, Sughosh Ganu wrote:
On Wed, 9 Dec 2020 at 03:24, Heinrich Schuchardt <xypron.glpk@gmx.de mailto:xypron.glpk@gmx.de> wrote:
On 12/8/20 10:19 AM, Sughosh Ganu wrote: > > On Tue, 8 Dec 2020 at 14:32, Heinrich Schuchardt <xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de> > <mailto:xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de>>> wrote: > > On 08.12.20 06:28, Sughosh Ganu wrote: > > > > On Mon, 7 Dec 2020 at 23:28, Heinrich Schuchardt > <xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de> <mailto:xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de>> > > <mailto:xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de> <mailto:xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de>>>> wrote: > > > > On 07.12.20 13:50, Heinrich Schuchardt wrote: > > > On 07.12.20 06:15, Sughosh Ganu wrote: > > >> hello Heinrich, > > >> > > >> On Sat, 5 Dec 2020 at 15:01, Heinrich Schuchardt > > <xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de> <mailto:xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de>> > <mailto:xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de> <mailto:xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de>>> > > >> <mailto:xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de> <mailto:xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de>> > <mailto:xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de> <mailto:xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de>>>>> wrote: > > >> > > >> On 11/26/20 7:40 PM, Sughosh Ganu wrote: > > >> > The Qemu platform emulator generates a device tree > blob and > > places it > > >> > at the start of the dram, which is then used by > u-boot. Use > > this dtb > > >> > only if CONFIG_OF_BOARD is defined. This allows using a > > different > > >> > device tree, using the CONFIG_OF_SEPARATE option. > This dtb > > is attached > > >> > to the u-boot binary as a u-boot-fdt.bin file > > >> > > >> Dear Sughosh, > > >> > > >> thank your for this series which will allow us to better > > demonstrate and > > >> test capsule updates. > > >> > > >> I am not sure if the approach that you take at > device-trees > > here is the > > >> right one. > > >> > > >> On QEMU the device-tree is generated on the fly by the > QEMU > > binary > > >> depending on which devices the user has specified. > > >> > > >> Your idea is to replace this device-tree completely to be > > able to add > > >> extra elements (the EFI signature list, see patch > 2/14). Thus a > > >> device-tree might be loaded that does not match the > user selected > > >> devices. > > >> > > >> An alternative approach would be to apply all > additions to the > > >> device-tree as an FDT overlay (or fixup). This would allow > > the dynamic > > >> parts of the QEMU device-tree still to be passed through. > > >> > > >> > > >> I will take a look at storing the public key as part of > the fdt > > overlay, > > >> with a runtime fixup. Although, I think the issue that you are > > pointing > > >> to would be specific to Qemu and not other platforms. But I do > > see the > > >> merit in having the public-key certificate stored as part > of an > > overlay. > > >> If I hit any issues while implementing this, I will get > back to > > you. Thanks. > > >> > > >> -sughosh > > > > > > OpenSBI can supply a device-tree to U-Boot. So this would be an > > > equivalent case. > > > > > > Best regards > > > > > > Heinrich > > > > <sng_>: "I am unable to think of a simple and elegant way to > generate > > the overlay dtb, since this would require creation of a > corresponding > > dts file, compiling it into a dtb and then using this on the > platform." > > > > Jean-Jacques' patch series introduced some of the necessary code: > > > > [PATCH PATCH v6 00/13] Add support for applications of > overlays in SPL > > https://lists.denx.de/pipermail/u-boot/2019-October/387653.html <https://lists.denx.de/pipermail/u-boot/2019-October/387653.html> > <https://lists.denx.de/pipermail/u-boot/2019-October/387653.html <https://lists.denx.de/pipermail/u-boot/2019-October/387653.html>> > > > <https://lists.denx.de/pipermail/u-boot/2019-October/387653.html <https://lists.denx.de/pipermail/u-boot/2019-October/387653.html> > <https://lists.denx.de/pipermail/u-boot/2019-October/387653.html <https://lists.denx.de/pipermail/u-boot/2019-October/387653.html>>> > > > https://patchwork.ozlabs.org/project/uboot/list/?series=137810&state=%2A&archive=both <https://patchwork.ozlabs.org/project/uboot/list/?series=137810&state=%2A&archive=both> > <https://patchwork.ozlabs.org/project/uboot/list/?series=137810&state=%2A&archive=both <https://patchwork.ozlabs.org/project/uboot/list/?series=137810&state=%2A&archive=both>> > > > <https://patchwork.ozlabs.org/project/uboot/list/?series=137810&state=%2A&archive=both <https://patchwork.ozlabs.org/project/uboot/list/?series=137810&state=%2A&archive=both> <https://patchwork.ozlabs.org/project/uboot/list/?series=137810&state=%2A&archive=both <https://patchwork.ozlabs.org/project/uboot/list/?series=137810&state=%2A&archive=both>>> > > > > CONFIG_OF_OVERLAY_LIST is used to specify the overlays to be > compiled. > > You will have to remove the CONFIG_SPL_LOAD_FIT_APPLY_OVERLAY and > > CONFIG_SPL_LOAD_FIT dependency. > > > > > > What i meant was that the process to generate the fdt overlay on the > > host, embed it with the public-key certificate is more cumbersome > than > > the current solution. So, for comparing the embedding the pub-key > cert > > in the fdt overlay, as against the platform dtb > > > > Embedding the certificate in the overlay > > 1) Generate a skeleton overlay dts file > > 2) Convert it to a dtb > > 3) Run it through the mkeficapsule utility to embed the pub-key > > certificate in the overlay(dtb) > > 4) Store the overlay dtb on some nv storage on the platform(ESP > partition?) > > 5) Load the overlay and apply it to the platform's dtb > > 6) Retrieve the certificate from the dtb at the time of capsule > > authentication > > > > Embedding the certificate in the platform's dtb > > 1) Run the platform's dtb through the mkeficapsule utility to embed > > the pub-key certificate > > 2) Boot the platform with the platform's dtb > > 3) Retrieve the certificate from the dtb at the time of capsule > > authentication > > > > You had mentioned OpenSBI passing the dtb to u-boot. Does the OpenSBI > > generate the device tree for the platform on the fly even for > non-qemu > > platforms. If it does not, the dtb that OpenSBI uses can be run > through > > the mkeficapsule utility to embed the certificate. > > OpenSBI applies fix-ups before passing the device-tree. The prototype > device-tree can either be built into OpenSBI or can be passed from an > earlier firmware stage. > > > So, if qemu if the only platform where the device tree is generated on > the fly, and passed along, based on the comparison that i have stated > above for the two scenarios(overlay vs platform dtb), I feel that using > the platform's dtb and embedding the certificate is more easier to use > than using the overlay. > > Even on the qemu platform, I retrieved the dtb from the platform, for > the given set of devices and interfaces used, and then ran the dtb > through the mkeficapsule utility. > > -sughosh As I understand the whole series is not about productive firmware updates for QEMU but about a demonstration how it could be used and allowing testing. Wouldn't it be the enough for testing to apply an overlay with the public key using the fdt command?
Is there a way I can build qemu such that the device-tree that is generated is with the __symbols__ node. Without this node, I am unable to apply the fdt overlay on the base platform devicetree.
According to https://www.kernel.org/doc/Documentation/devicetree/overlay-notes.txt __symbols__ is only needed if you refer to phandles in the overlay. When using paths instead __symbols__ is not required.
I have not tested if this holds true for U-Boot too.
Best regards
Heinrich
-sughosh

On Tue, 15 Dec 2020 at 18:25, Heinrich Schuchardt xypron.glpk@gmx.de wrote:
On 15.12.20 12:10, Sughosh Ganu wrote:
On Wed, 9 Dec 2020 at 03:24, Heinrich Schuchardt <xypron.glpk@gmx.de mailto:xypron.glpk@gmx.de> wrote:
On 12/8/20 10:19 AM, Sughosh Ganu wrote: > > On Tue, 8 Dec 2020 at 14:32, Heinrich Schuchardt <xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de> > <mailto:xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de>>> wrote: > > On 08.12.20 06:28, Sughosh Ganu wrote: > > > > On Mon, 7 Dec 2020 at 23:28, Heinrich Schuchardt > <xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de> <mailto:xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de>> > > <mailto:xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de> <mailto:xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de>>>> wrote: > > > > On 07.12.20 13:50, Heinrich Schuchardt wrote: > > > On 07.12.20 06:15, Sughosh Ganu wrote: > > >> hello Heinrich, > > >> > > >> On Sat, 5 Dec 2020 at 15:01, Heinrich Schuchardt > > <xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de> <mailto:xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de>> > <mailto:xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de> <mailto:xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de>>> > > >> <mailto:xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de> <mailto:xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de>> > <mailto:xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de> <mailto:xypron.glpk@gmx.de <mailto:xypron.glpk@gmx.de>>>>> wrote: > > >> > > >> On 11/26/20 7:40 PM, Sughosh Ganu wrote: > > >> > The Qemu platform emulator generates a device
tree
> blob and > > places it > > >> > at the start of the dram, which is then used by > u-boot. Use > > this dtb > > >> > only if CONFIG_OF_BOARD is defined. This allows using a > > different > > >> > device tree, using the CONFIG_OF_SEPARATE
option.
> This dtb > > is attached > > >> > to the u-boot binary as a u-boot-fdt.bin file > > >> > > >> Dear Sughosh, > > >> > > >> thank your for this series which will allow us to better > > demonstrate and > > >> test capsule updates. > > >> > > >> I am not sure if the approach that you take at > device-trees > > here is the > > >> right one. > > >> > > >> On QEMU the device-tree is generated on the fly by the > QEMU > > binary > > >> depending on which devices the user has
specified.
> > >> > > >> Your idea is to replace this device-tree completely to be > > able to add > > >> extra elements (the EFI signature list, see patch > 2/14). Thus a > > >> device-tree might be loaded that does not match
the
> user selected > > >> devices. > > >> > > >> An alternative approach would be to apply all > additions to the > > >> device-tree as an FDT overlay (or fixup). This would allow > > the dynamic > > >> parts of the QEMU device-tree still to be passed through. > > >> > > >> > > >> I will take a look at storing the public key as part
of
> the fdt > > overlay, > > >> with a runtime fixup. Although, I think the issue that you are > > pointing > > >> to would be specific to Qemu and not other platforms. But I do > > see the > > >> merit in having the public-key certificate stored as part > of an > > overlay. > > >> If I hit any issues while implementing this, I will
get
> back to > > you. Thanks. > > >> > > >> -sughosh > > > > > > OpenSBI can supply a device-tree to U-Boot. So this would be an > > > equivalent case. > > > > > > Best regards > > > > > > Heinrich > > > > <sng_>: "I am unable to think of a simple and elegant way to > generate > > the overlay dtb, since this would require creation of a > corresponding > > dts file, compiling it into a dtb and then using this on the > platform." > > > > Jean-Jacques' patch series introduced some of the necessary code: > > > > [PATCH PATCH v6 00/13] Add support for applications of > overlays in SPL > > https://lists.denx.de/pipermail/u-boot/2019-October/387653.html <https://lists.denx.de/pipermail/u-boot/2019-October/387653.html> > <https://lists.denx.de/pipermail/u-boot/2019-October/387653.html <https://lists.denx.de/pipermail/u-boot/2019-October/387653.html>> > > > <https://lists.denx.de/pipermail/u-boot/2019-October/387653.html <https://lists.denx.de/pipermail/u-boot/2019-October/387653.html> > <https://lists.denx.de/pipermail/u-boot/2019-October/387653.html <https://lists.denx.de/pipermail/u-boot/2019-October/387653.html>>> > > >
https://patchwork.ozlabs.org/project/uboot/list/?series=137810&state=%2A...
<
https://patchwork.ozlabs.org/project/uboot/list/?series=137810&state=%2A...
> <
https://patchwork.ozlabs.org/project/uboot/list/?series=137810&state=%2A...
<
https://patchwork.ozlabs.org/project/uboot/list/?series=137810&state=%2A...
> > > <
https://patchwork.ozlabs.org/project/uboot/list/?series=137810&state=%2A...
<
https://patchwork.ozlabs.org/project/uboot/list/?series=137810&state=%2A...
<
https://patchwork.ozlabs.org/project/uboot/list/?series=137810&state=%2A...
<
https://patchwork.ozlabs.org/project/uboot/list/?series=137810&state=%2A...
> > > > CONFIG_OF_OVERLAY_LIST is used to specify the overlays to be > compiled. > > You will have to remove the CONFIG_SPL_LOAD_FIT_APPLY_OVERLAY and > > CONFIG_SPL_LOAD_FIT dependency. > > > > > > What i meant was that the process to generate the fdt overlay on the > > host, embed it with the public-key certificate is more cumbersome > than > > the current solution. So, for comparing the embedding the pub-key > cert > > in the fdt overlay, as against the platform dtb > > > > Embedding the certificate in the overlay > > 1) Generate a skeleton overlay dts file > > 2) Convert it to a dtb > > 3) Run it through the mkeficapsule utility to embed the
pub-key
> > certificate in the overlay(dtb) > > 4) Store the overlay dtb on some nv storage on the
platform(ESP
> partition?) > > 5) Load the overlay and apply it to the platform's dtb > > 6) Retrieve the certificate from the dtb at the time of
capsule
> > authentication > > > > Embedding the certificate in the platform's dtb > > 1) Run the platform's dtb through the mkeficapsule utility to embed > > the pub-key certificate > > 2) Boot the platform with the platform's dtb > > 3) Retrieve the certificate from the dtb at the time of
capsule
> > authentication > > > > You had mentioned OpenSBI passing the dtb to u-boot. Does the OpenSBI > > generate the device tree for the platform on the fly even
for
> non-qemu > > platforms. If it does not, the dtb that OpenSBI uses can be
run
> through > > the mkeficapsule utility to embed the certificate. > > OpenSBI applies fix-ups before passing the device-tree. The prototype > device-tree can either be built into OpenSBI or can be passed from an > earlier firmware stage. > > > So, if qemu if the only platform where the device tree is
generated on
> the fly, and passed along, based on the comparison that i have
stated
> above for the two scenarios(overlay vs platform dtb), I feel that using > the platform's dtb and embedding the certificate is more easier to
use
> than using the overlay. > > Even on the qemu platform, I retrieved the dtb from the platform,
for
> the given set of devices and interfaces used, and then ran the dtb > through the mkeficapsule utility. > > -sughosh As I understand the whole series is not about productive firmware updates for QEMU but about a demonstration how it could be used and allowing testing. Wouldn't it be the enough for testing to apply an overlay with the public key using the fdt command?
Is there a way I can build qemu such that the device-tree that is generated is with the __symbols__ node. Without this node, I am unable to apply the fdt overlay on the base platform devicetree.
According to https://www.kernel.org/doc/Documentation/devicetree/overlay-notes.txt __symbols__ is only needed if you refer to phandles in the overlay. When using paths instead __symbols__ is not required.
I have not tested if this holds true for U-Boot too.
This does work. Thanks. I was referring to the document mentioned in the README.fdt-overlays[1] and that does not have mention of target-path.
-sughosh
[1] - https://git.kernel.org/pub/scm/utils/dtc/dtc.git/tree/Documentation/dt-objec...

Add options for embedding the public key esl(efi signature list) file to the platform's dtb. The esl file is then retrieved and used for authenticating the capsule to be used for updating firmare components on the platform.
The esl file can now be embedded in the dtb by invoking the following command mkeficapsule -K <pub_key.esl> -D <dtb>
This will create a node named 'signature' in the dtb, and the esl file will be stored as 'capsule-key'
Signed-off-by: Sughosh Ganu sughosh.ganu@linaro.org --- tools/Makefile | 1 + tools/mkeficapsule.c | 198 ++++++++++++++++++++++++++++++++++++++++--- 2 files changed, 186 insertions(+), 13 deletions(-)
diff --git a/tools/Makefile b/tools/Makefile index 66d9376803..6d7b48fb57 100644 --- a/tools/Makefile +++ b/tools/Makefile @@ -218,6 +218,7 @@ hostprogs-$(CONFIG_MIPS) += mips-relocs hostprogs-$(CONFIG_ASN1_COMPILER) += asn1_compiler HOSTCFLAGS_asn1_compiler.o = -idirafter $(srctree)/include
+mkeficapsule-objs := mkeficapsule.o $(LIBFDT_OBJS) hostprogs-$(CONFIG_EFI_HAVE_CAPSULE_SUPPORT) += mkeficapsule
# We build some files with extra pedantic flags to try to minimize things diff --git a/tools/mkeficapsule.c b/tools/mkeficapsule.c index 45e27d74a5..ce05da12d6 100644 --- a/tools/mkeficapsule.c +++ b/tools/mkeficapsule.c @@ -4,16 +4,22 @@ * Author: AKASHI Takahiro */
+#include <errno.h> #include <getopt.h> #include <malloc.h> #include <stdbool.h> #include <stdio.h> #include <stdlib.h> #include <string.h> +#include <unistd.h> #include <linux/types.h> + +#include <sys/mman.h> #include <sys/stat.h> #include <sys/types.h>
+#include "fdt_host.h" + typedef __u8 u8; typedef __u16 u16; typedef __u32 u32; @@ -23,6 +29,8 @@ typedef __s32 s32;
#define aligned_u64 __aligned_u64
+#define SIGNATURE_NODENAME "signature" + #ifndef __packed #define __packed __attribute__((packed)) #endif @@ -43,6 +51,8 @@ static struct option options[] = { {"raw", required_argument, NULL, 'r'}, {"index", required_argument, NULL, 'i'}, {"instance", required_argument, NULL, 'I'}, + {"dtb", required_argument, NULL, 'D'}, + {"public key", required_argument, NULL, 'K'}, {"version", required_argument, NULL, 'v'}, {"help", no_argument, NULL, 'h'}, {NULL, 0, NULL, 0}, @@ -52,15 +62,154 @@ static void print_usage(void) { printf("Usage: %s [options] <output file>\n" "Options:\n" - "\t--fit <fit image> new FIT image file\n" - "\t--raw <raw image> new raw image file\n" - "\t--index <index> update image index\n" - "\t--instance <instance> update hardware instance\n" - "\t--version <version> firmware version\n" - "\t--help print a help message\n", + "\t--fit <fit image> new FIT image file\n" + "\t--raw <raw image> new raw image file\n" + "\t--index <index> update image index\n" + "\t--instance <instance> update hardware instance\n" + "\t--version <version> firmware version\n" + "\t--public-key <key file> public key esl file\n" + "\t--dtb <dtb file> dtb file\n" + "\t--help print a help message\n", tool_name); }
+static int fdt_add_pub_key_data(void *sptr, void *dptr, size_t key_size) +{ + int parent; + int ret = 0; + + parent = fdt_subnode_offset(dptr, 0, SIGNATURE_NODENAME); + if (parent == -FDT_ERR_NOTFOUND) { + parent = fdt_add_subnode(dptr, 0, SIGNATURE_NODENAME); + if (parent < 0) { + ret = parent; + if (ret != -FDT_ERR_NOSPACE) { + fprintf(stderr, + "Couldn't create signature node: %s\n", + fdt_strerror(parent)); + } + } + } + if (ret) + goto done; + + /* Write the key to the FDT node */ + ret = fdt_setprop(dptr, parent, "capsule-key", + sptr, key_size); + +done: + if (ret) + ret = ret == -FDT_ERR_NOSPACE ? -ENOSPC : -EIO; + + return ret; +} + +static int add_public_key(const char *pkey_file, const char *dtb_file) +{ + int ret; + int srcfd = 0; + int destfd = 0; + void *sptr = NULL; + void *dptr = NULL; + off_t src_size; + struct stat pub_key; + struct stat dtb; + + /* Find out the size of the public key */ + srcfd = open(pkey_file, O_RDONLY); + if (srcfd == -1) { + fprintf(stderr, "%s: Can't open %s: %s\n", + __func__, pkey_file, strerror(errno)); + goto err; + } + + ret = fstat(srcfd, &pub_key); + if (ret == -1) { + fprintf(stderr, "%s: Can't stat %s: %s\n", + __func__, pkey_file, strerror(errno)); + goto err; + } + + src_size = pub_key.st_size; + + /* mmap the public key esl file */ + sptr = mmap(0, src_size, PROT_READ, MAP_SHARED, srcfd, 0); + if ((sptr == MAP_FAILED) || (errno != 0)) { + fprintf(stderr, "%s: Failed to mmap %s:%s\n", + __func__, pkey_file, strerror(errno)); + goto err; + } + + /* Open the dest FDT */ + destfd = open(dtb_file, O_RDWR); + if (destfd == -1) { + fprintf(stderr, "%s: Can't open %s: %s\n", + __func__, dtb_file, strerror(errno)); + goto err; + } + + ret = fstat(destfd, &dtb); + if (ret == -1) { + fprintf(stderr, "%s: Can't stat %s: %s\n", + __func__, dtb_file, strerror(errno)); + goto err; + } + + dtb.st_size += src_size; + if (ftruncate(destfd, dtb.st_size)) { + fprintf(stderr, "%s: Can't expand %s: %s\n", + __func__, dtb_file, strerror(errno)); + goto err;; + } + + errno = 0; + /* mmap the dtb file */ + dptr = mmap(0, dtb.st_size, PROT_READ | PROT_WRITE, MAP_SHARED, + destfd, 0); + if ((dptr == MAP_FAILED) || (errno != 0)) { + fprintf(stderr, "%s: Failed to mmap %s:%s\n", + __func__, dtb_file, strerror(errno)); + goto err; + } + + if (fdt_check_header(dptr)) { + fprintf(stderr, "%s: Invalid FDT header\n", __func__); + goto err; + } + + ret = fdt_open_into(dptr, dptr, dtb.st_size); + if (ret) { + fprintf(stderr, "%s: Cannot expand FDT: %s\n", + __func__, fdt_strerror(ret)); + goto err; + } + + /* Copy the esl file to the expanded FDT */ + ret = fdt_add_pub_key_data(sptr, dptr, src_size); + if (ret < 0) { + fprintf(stderr, "%s: Unable to add public key to the FDT\n", + __func__); + goto err; + } + + return 0; + +err: + if (sptr) + munmap(sptr, src_size); + + if (dptr) + munmap(dptr, dtb.st_size); + + if (srcfd >= 0) + close(srcfd); + + if (destfd >= 0) + close(destfd); + + return -1; +} + static int create_fwbin(char *path, char *bin, efi_guid_t *guid, unsigned long version, unsigned long index, unsigned long instance) @@ -171,17 +320,22 @@ err_1: int main(int argc, char **argv) { char *file; + char *pkey_file; + char *dtb_file; efi_guid_t *guid; unsigned long index, instance, version; int c, idx; + int ret;
file = NULL; + pkey_file = NULL; + dtb_file = NULL; guid = NULL; index = 0; instance = 0; version = 0; for (;;) { - c = getopt_long(argc, argv, "f:r:i:I:v:h", options, &idx); + c = getopt_long(argc, argv, "f:r:i:I:v:D:K:h", options, &idx); if (c == -1) break;
@@ -211,22 +365,40 @@ int main(int argc, char **argv) case 'v': version = strtoul(optarg, NULL, 0); break; + case 'K': + if (pkey_file) { + printf("Public Key already specified\n"); + return -1; + } + pkey_file = optarg; + break; + case 'D': + if (dtb_file) { + printf("DTB file already specified\n"); + return -1; + } + dtb_file = optarg; + break; case 'h': print_usage(); return 0; } }
- /* need a output file */ - if (argc != optind + 1) { + /* need a fit image file or raw image file */ + if (!file && !pkey_file && !dtb_file) { print_usage(); return -1; }
- /* need a fit image file or raw image file */ - if (!file) { - print_usage(); - return -1; + if (pkey_file && dtb_file) { + ret = add_public_key(pkey_file, dtb_file); + if (ret == -1) { + printf("Adding public key to the dtb failed\n"); + return -1; + } else { + return 0; + } }
if (create_fwbin(argv[optind], file, guid, version, index, instance)

Scan the pci bus in board_init routine before scanning the virtio devices. This enumerates all the virtio devices, including devices found on the pci bus.
Signed-off-by: Sughosh Ganu sughosh.ganu@linaro.org --- board/emulation/qemu-arm/qemu-arm.c | 8 ++++++++ 1 file changed, 8 insertions(+)
diff --git a/board/emulation/qemu-arm/qemu-arm.c b/board/emulation/qemu-arm/qemu-arm.c index e146d1cc50..b3d5b3d5c2 100644 --- a/board/emulation/qemu-arm/qemu-arm.c +++ b/board/emulation/qemu-arm/qemu-arm.c @@ -65,6 +65,14 @@ struct mm_region *mem_map = qemu_arm64_mem_map;
int board_init(void) { + + /* + * Scan the pci bus before calling virtio_init. This + * enumerates all virtio devices, including devices + * on the pci bus. + */ + pci_init(); + /* * Make sure virtio bus is enumerated so that peripherals * on the virtio bus can be discovered by their drivers

On 11/26/20 7:40 PM, Sughosh Ganu wrote:
Scan the pci bus in board_init routine before scanning the virtio devices. This enumerates all the virtio devices, including devices found on the pci bus.
Signed-off-by: Sughosh Ganu sughosh.ganu@linaro.org
board/emulation/qemu-arm/qemu-arm.c | 8 ++++++++ 1 file changed, 8 insertions(+)
diff --git a/board/emulation/qemu-arm/qemu-arm.c b/board/emulation/qemu-arm/qemu-arm.c index e146d1cc50..b3d5b3d5c2 100644 --- a/board/emulation/qemu-arm/qemu-arm.c +++ b/board/emulation/qemu-arm/qemu-arm.c @@ -65,6 +65,14 @@ struct mm_region *mem_map = qemu_arm64_mem_map;
int board_init(void) {
- /*
* Scan the pci bus before calling virtio_init. This
* enumerates all virtio devices, including devices
* on the pci bus.
*/
- pci_init();
This does not compile if CONFIG_PCI=n.
aarch64-linux-gnu-ld.bfd: board/emulation/qemu-arm/built-in.o: in function `board_init': board/emulation/qemu-arm/qemu-arm.c:74: undefined reference to `pci_init'
Cf. arch/arm/mach-mvebu/arm64-common.c-106-#ifdef CONFIG_DM_PCI board/socrates/socrates.c-134-#if defined(CONFIG_DM_PCI)
I found these lines in common/board_r.c:250
#ifdef CONFIG_PCI static int initr_pci(void) { if (IS_ENABLED(CONFIG_PCI_INIT_R)) pci_init();
return 0; } #endif
Would selecting CONFIG_PCI_INIT_R be enough to solve your problem?
Best regards
Heinrich
- /*
- Make sure virtio bus is enumerated so that peripherals
- on the virtio bus can be discovered by their drivers

On Sat, 5 Dec 2020 at 15:15, Heinrich Schuchardt xypron.glpk@gmx.de wrote:
On 11/26/20 7:40 PM, Sughosh Ganu wrote:
Scan the pci bus in board_init routine before scanning the virtio devices. This enumerates all the virtio devices, including devices found on the pci bus.
Signed-off-by: Sughosh Ganu sughosh.ganu@linaro.org
board/emulation/qemu-arm/qemu-arm.c | 8 ++++++++ 1 file changed, 8 insertions(+)
diff --git a/board/emulation/qemu-arm/qemu-arm.c
b/board/emulation/qemu-arm/qemu-arm.c
index e146d1cc50..b3d5b3d5c2 100644 --- a/board/emulation/qemu-arm/qemu-arm.c +++ b/board/emulation/qemu-arm/qemu-arm.c @@ -65,6 +65,14 @@ struct mm_region *mem_map = qemu_arm64_mem_map;
int board_init(void) {
/*
* Scan the pci bus before calling virtio_init. This
* enumerates all virtio devices, including devices
* on the pci bus.
*/
pci_init();
This does not compile if CONFIG_PCI=n.
aarch64-linux-gnu-ld.bfd: board/emulation/qemu-arm/built-in.o: in function `board_init': board/emulation/qemu-arm/qemu-arm.c:74: undefined reference to `pci_init'
Cf. arch/arm/mach-mvebu/arm64-common.c-106-#ifdef CONFIG_DM_PCI board/socrates/socrates.c-134-#if defined(CONFIG_DM_PCI)
I found these lines in common/board_r.c:250
#ifdef CONFIG_PCI static int initr_pci(void) { if (IS_ENABLED(CONFIG_PCI_INIT_R)) pci_init();
return 0;
} #endif
Would selecting CONFIG_PCI_INIT_R be enough to solve your problem?
Will check this out. Thanks.
-sughosh

RFC 2315 Section 9.3 describes the message digesting process. The digest calculated depends on whether the authenticated attributes are present. In case of a scenario where the authenticated attributes are present, the message digest that gets signed and is part of the pkcs7 message is computed from the auth attributes rather than the contents field.
Check if the auth attributes are present, and if set, use the auth attributes to compute the hash that would be compared with the encrypted hash on the pkcs7 message.
Signed-off-by: Sughosh Ganu sughosh.ganu@linaro.org --- lib/crypto/pkcs7_verify.c | 37 ++++++++++++++++++++++++++----------- 1 file changed, 26 insertions(+), 11 deletions(-)
diff --git a/lib/crypto/pkcs7_verify.c b/lib/crypto/pkcs7_verify.c index 320ba49f79..58683ef614 100644 --- a/lib/crypto/pkcs7_verify.c +++ b/lib/crypto/pkcs7_verify.c @@ -50,8 +50,15 @@ static int pkcs7_digest(struct pkcs7_message *pkcs7, struct image_region regions[2]; int ret = 0;
- /* The digest was calculated already. */ - if (sig->digest) + /* + * [RFC2315 9.3] + * If the authenticated attributes are present, + * the message-digest is calculated on the + * attributes present in the + * authenticatedAttributes field and not just + * the contents field + */ + if (!sinfo->authattrs && sig->digest) return 0;
if (!sinfo->sig->hash_algo) @@ -63,17 +70,25 @@ static int pkcs7_digest(struct pkcs7_message *pkcs7, else return -ENOPKG;
- sig->digest = calloc(1, sig->digest_size); - if (!sig->digest) { - pr_warn("Sig %u: Out of memory\n", sinfo->index); - return -ENOMEM; - } + /* + * Calculate the hash only if the data is present. + * In case of authenticated variable and capsule, + * the hash has already been calculated on the + * efi_image_regions and populated + */ + if (pkcs7->data) { + sig->digest = calloc(1, sig->digest_size); + if (!sig->digest) { + pr_warn("Sig %u: Out of memory\n", sinfo->index); + return -ENOMEM; + }
- regions[0].data = pkcs7->data; - regions[0].size = pkcs7->data_len; + regions[0].data = pkcs7->data; + regions[0].size = pkcs7->data_len;
- /* Digest the message [RFC2315 9.3] */ - hash_calculate(sinfo->sig->hash_algo, regions, 1, sig->digest); + /* Digest the message [RFC2315 9.3] */ + hash_calculate(sinfo->sig->hash_algo, regions, 1, sig->digest); + }
/* However, if there are authenticated attributes, there must be a * message digest attribute amongst them which corresponds to the

On 11/26/20 7:41 PM, Sughosh Ganu wrote:
RFC 2315 Section 9.3 describes the message digesting process. The digest calculated depends on whether the authenticated attributes are present. In case of a scenario where the authenticated attributes are present, the message digest that gets signed and is part of the pkcs7 message is computed from the auth attributes rather than the contents field.
Check if the auth attributes are present, and if set, use the auth attributes to compute the hash that would be compared with the encrypted hash on the pkcs7 message.
Signed-off-by: Sughosh Ganu sughosh.ganu@linaro.org
Dear Takahiro,
Could you, please, review this patch.
Best regards
Heinrich
lib/crypto/pkcs7_verify.c | 37 ++++++++++++++++++++++++++----------- 1 file changed, 26 insertions(+), 11 deletions(-)
diff --git a/lib/crypto/pkcs7_verify.c b/lib/crypto/pkcs7_verify.c index 320ba49f79..58683ef614 100644 --- a/lib/crypto/pkcs7_verify.c +++ b/lib/crypto/pkcs7_verify.c @@ -50,8 +50,15 @@ static int pkcs7_digest(struct pkcs7_message *pkcs7, struct image_region regions[2]; int ret = 0;
- /* The digest was calculated already. */
- if (sig->digest)
/*
* [RFC2315 9.3]
* If the authenticated attributes are present,
* the message-digest is calculated on the
* attributes present in the
* authenticatedAttributes field and not just
* the contents field
*/
if (!sinfo->authattrs && sig->digest) return 0;
if (!sinfo->sig->hash_algo)
@@ -63,17 +70,25 @@ static int pkcs7_digest(struct pkcs7_message *pkcs7, else return -ENOPKG;
- sig->digest = calloc(1, sig->digest_size);
- if (!sig->digest) {
pr_warn("Sig %u: Out of memory\n", sinfo->index);
return -ENOMEM;
- }
- /*
* Calculate the hash only if the data is present.
* In case of authenticated variable and capsule,
* the hash has already been calculated on the
* efi_image_regions and populated
*/
- if (pkcs7->data) {
sig->digest = calloc(1, sig->digest_size);
if (!sig->digest) {
pr_warn("Sig %u: Out of memory\n", sinfo->index);
return -ENOMEM;
}
- regions[0].data = pkcs7->data;
- regions[0].size = pkcs7->data_len;
regions[0].data = pkcs7->data;
regions[0].size = pkcs7->data_len;
- /* Digest the message [RFC2315 9.3] */
- hash_calculate(sinfo->sig->hash_algo, regions, 1, sig->digest);
/* Digest the message [RFC2315 9.3] */
hash_calculate(sinfo->sig->hash_algo, regions, 1, sig->digest);
}
/* However, if there are authenticated attributes, there must be a
- message digest attribute amongst them which corresponds to the

Add support for setting the default values for mtd partitions on the platform for the nor flash. This would be used for updating the firmware image using uefi capsule update with the dfu mtd backend driver.
Signed-off-by: Sughosh Ganu sughosh.ganu@linaro.org --- board/emulation/qemu-arm/qemu-arm.c | 70 +++++++++++++++++++++++++++++ include/configs/qemu-arm.h | 7 +++ 2 files changed, 77 insertions(+)
diff --git a/board/emulation/qemu-arm/qemu-arm.c b/board/emulation/qemu-arm/qemu-arm.c index b3d5b3d5c2..d5ed3eebaf 100644 --- a/board/emulation/qemu-arm/qemu-arm.c +++ b/board/emulation/qemu-arm/qemu-arm.c @@ -197,3 +197,73 @@ void flash_write32(u32 value, void *addr) { asm("str %" __W "1, %0" : "=m"(*(u32 *)addr) : "r"(value)); } + +#if defined(CONFIG_EFI_CAPSULE_FIRMWARE_MANAGEMENT) + +#include <mtd.h> + +static void board_get_mtdparts(const char *dev, const char *partition, + char *mtdids, char *mtdparts) +{ + /* mtdids: "<dev>=<dev>, ...." */ + if (mtdids[0] != '\0') + strcat(mtdids, ","); + strcat(mtdids, dev); + strcat(mtdids, "="); + strcat(mtdids, dev); + + /* mtdparts: "mtdparts=<dev>:<mtdparts_<dev>>;..." */ + if (mtdparts[0] != '\0') + strncat(mtdparts, ";", MTDPARTS_LEN); + else + strcat(mtdparts, "mtdparts="); + + strncat(mtdparts, dev, MTDPARTS_LEN); + strncat(mtdparts, ":", MTDPARTS_LEN); + strncat(mtdparts, partition, MTDPARTS_LEN); +} + +void board_mtdparts_default(const char **mtdids, const char **mtdparts) +{ + struct mtd_info *mtd; + struct udevice *dev; + const char *mtd_partition; + static char parts[3 * MTDPARTS_LEN + 1]; + static char ids[MTDIDS_LEN + 1]; + static bool mtd_initialized; + + if (mtd_initialized) { + *mtdids = ids; + *mtdparts = parts; + return; + } + + memset(parts, 0, sizeof(parts)); + memset(ids, 0, sizeof(ids)); + + /* probe all MTD devices */ + for (uclass_first_device(UCLASS_MTD, &dev); dev; + uclass_next_device(&dev)) { + debug("mtd device = %s\n", dev->name); + } + + mtd = get_mtd_device_nm("nor0"); + if (!IS_ERR_OR_NULL(mtd)) { + mtd_partition = MTDPARTS_NOR0; + board_get_mtdparts("nor0", mtd_partition, ids, parts); + put_mtd_device(mtd); + } + + mtd = get_mtd_device_nm("nor1"); + if (!IS_ERR_OR_NULL(mtd)) { + mtd_partition = MTDPARTS_NOR1; + board_get_mtdparts("nor1", mtd_partition, ids, parts); + put_mtd_device(mtd); + } + + mtd_initialized = true; + *mtdids = ids; + *mtdparts = parts; + debug("%s:mtdids=%s & mtdparts=%s\n", __func__, ids, parts); +} +#endif /* CONFIG_EFI_CAPSULE_FIRMWARE_MANAGEMENT */ diff --git a/include/configs/qemu-arm.h b/include/configs/qemu-arm.h index 273fa1a7d7..69ff329434 100644 --- a/include/configs/qemu-arm.h +++ b/include/configs/qemu-arm.h @@ -32,6 +32,13 @@
#include <config_distro_bootcmd.h>
+#if defined(CONFIG_EFI_CAPSULE_FIRMWARE_MANAGEMENT) +#define CONFIG_SYS_MTDPARTS_RUNTIME +#endif + +#define MTDPARTS_NOR0 "64m(u-boot)\0" +#define MTDPARTS_NOR1 "64m(u-boot-env)\0" + #define CONFIG_EXTRA_ENV_SETTINGS \ "fdt_high=0xffffffff\0" \ "initrd_high=0xffffffff\0" \

On 11/26/20 7:41 PM, Sughosh Ganu wrote:
Add support for setting the default values for mtd partitions on the platform for the nor flash. This would be used for updating the firmware image using uefi capsule update with the dfu mtd backend driver.
Signed-off-by: Sughosh Ganu sughosh.ganu@linaro.org
board/emulation/qemu-arm/qemu-arm.c | 70 +++++++++++++++++++++++++++++ include/configs/qemu-arm.h | 7 +++ 2 files changed, 77 insertions(+)
diff --git a/board/emulation/qemu-arm/qemu-arm.c b/board/emulation/qemu-arm/qemu-arm.c index b3d5b3d5c2..d5ed3eebaf 100644 --- a/board/emulation/qemu-arm/qemu-arm.c +++ b/board/emulation/qemu-arm/qemu-arm.c
Is this development really QEMU specific or is it something we would could reuse for other systems too? If yes, we should put it into another directory (drivers/mtd/ or drivers/dfu/).
Best regards
Heinrich
@@ -197,3 +197,73 @@ void flash_write32(u32 value, void *addr) { asm("str %" __W "1, %0" : "=m"(*(u32 *)addr) : "r"(value)); }
+#if defined(CONFIG_EFI_CAPSULE_FIRMWARE_MANAGEMENT)
+#include <mtd.h>
+static void board_get_mtdparts(const char *dev, const char *partition,
char *mtdids, char *mtdparts)
+{
- /* mtdids: "<dev>=<dev>, ...." */
- if (mtdids[0] != '\0')
strcat(mtdids, ",");
- strcat(mtdids, dev);
- strcat(mtdids, "=");
- strcat(mtdids, dev);
- /* mtdparts: "mtdparts=<dev>:<mtdparts_<dev>>;..." */
- if (mtdparts[0] != '\0')
strncat(mtdparts, ";", MTDPARTS_LEN);
- else
strcat(mtdparts, "mtdparts=");
- strncat(mtdparts, dev, MTDPARTS_LEN);
- strncat(mtdparts, ":", MTDPARTS_LEN);
- strncat(mtdparts, partition, MTDPARTS_LEN);
+}
+void board_mtdparts_default(const char **mtdids, const char **mtdparts) +{
- struct mtd_info *mtd;
- struct udevice *dev;
- const char *mtd_partition;
- static char parts[3 * MTDPARTS_LEN + 1];
- static char ids[MTDIDS_LEN + 1];
- static bool mtd_initialized;
- if (mtd_initialized) {
*mtdids = ids;
*mtdparts = parts;
return;
- }
- memset(parts, 0, sizeof(parts));
- memset(ids, 0, sizeof(ids));
- /* probe all MTD devices */
- for (uclass_first_device(UCLASS_MTD, &dev); dev;
uclass_next_device(&dev)) {
debug("mtd device = %s\n", dev->name);
- }
- mtd = get_mtd_device_nm("nor0");
- if (!IS_ERR_OR_NULL(mtd)) {
mtd_partition = MTDPARTS_NOR0;
board_get_mtdparts("nor0", mtd_partition, ids, parts);
put_mtd_device(mtd);
- }
- mtd = get_mtd_device_nm("nor1");
- if (!IS_ERR_OR_NULL(mtd)) {
mtd_partition = MTDPARTS_NOR1;
board_get_mtdparts("nor1", mtd_partition, ids, parts);
put_mtd_device(mtd);
- }
- mtd_initialized = true;
- *mtdids = ids;
- *mtdparts = parts;
- debug("%s:mtdids=%s & mtdparts=%s\n", __func__, ids, parts);
+} +#endif /* CONFIG_EFI_CAPSULE_FIRMWARE_MANAGEMENT */ diff --git a/include/configs/qemu-arm.h b/include/configs/qemu-arm.h index 273fa1a7d7..69ff329434 100644 --- a/include/configs/qemu-arm.h +++ b/include/configs/qemu-arm.h @@ -32,6 +32,13 @@
#include <config_distro_bootcmd.h>
+#if defined(CONFIG_EFI_CAPSULE_FIRMWARE_MANAGEMENT) +#define CONFIG_SYS_MTDPARTS_RUNTIME +#endif
+#define MTDPARTS_NOR0 "64m(u-boot)\0" +#define MTDPARTS_NOR1 "64m(u-boot-env)\0"
- #define CONFIG_EXTRA_ENV_SETTINGS \ "fdt_high=0xffffffff\0" \ "initrd_high=0xffffffff\0" \

On Sat, 5 Dec 2020 at 16:00, Heinrich Schuchardt xypron.glpk@gmx.de wrote:
On 11/26/20 7:41 PM, Sughosh Ganu wrote:
Add support for setting the default values for mtd partitions on the platform for the nor flash. This would be used for updating the firmware image using uefi capsule update with the dfu mtd backend driver.
Signed-off-by: Sughosh Ganu sughosh.ganu@linaro.org
board/emulation/qemu-arm/qemu-arm.c | 70 +++++++++++++++++++++++++++++ include/configs/qemu-arm.h | 7 +++ 2 files changed, 77 insertions(+)
diff --git a/board/emulation/qemu-arm/qemu-arm.c
b/board/emulation/qemu-arm/qemu-arm.c
index b3d5b3d5c2..d5ed3eebaf 100644 --- a/board/emulation/qemu-arm/qemu-arm.c +++ b/board/emulation/qemu-arm/qemu-arm.c
Is this development really QEMU specific or is it something we would could reuse for other systems too? If yes, we should put it into another directory (drivers/mtd/ or drivers/dfu/).
I think the functionality added in this patch is specific for a particular board, given that the number and type of mtd devices would be specific to a board. Even if the mtd device detection can be done dynamically, it would be required to get the partition information from the board specific files. Merging all the functionality under a common directory like drivers/mtd/, if done, should be taken up as a separate effort. Thanks.
-sughosh
Best regards
Heinrich
@@ -197,3 +197,73 @@ void flash_write32(u32 value, void *addr) { asm("str %" __W "1, %0" : "=m"(*(u32 *)addr) : "r"(value)); }
+#if defined(CONFIG_EFI_CAPSULE_FIRMWARE_MANAGEMENT)
+#include <mtd.h>
+static void board_get_mtdparts(const char *dev, const char *partition,
char *mtdids, char *mtdparts)
+{
/* mtdids: "<dev>=<dev>, ...." */
if (mtdids[0] != '\0')
strcat(mtdids, ",");
strcat(mtdids, dev);
strcat(mtdids, "=");
strcat(mtdids, dev);
/* mtdparts: "mtdparts=<dev>:<mtdparts_<dev>>;..." */
if (mtdparts[0] != '\0')
strncat(mtdparts, ";", MTDPARTS_LEN);
else
strcat(mtdparts, "mtdparts=");
strncat(mtdparts, dev, MTDPARTS_LEN);
strncat(mtdparts, ":", MTDPARTS_LEN);
strncat(mtdparts, partition, MTDPARTS_LEN);
+}
+void board_mtdparts_default(const char **mtdids, const char **mtdparts) +{
struct mtd_info *mtd;
struct udevice *dev;
const char *mtd_partition;
static char parts[3 * MTDPARTS_LEN + 1];
static char ids[MTDIDS_LEN + 1];
static bool mtd_initialized;
if (mtd_initialized) {
*mtdids = ids;
*mtdparts = parts;
return;
}
memset(parts, 0, sizeof(parts));
memset(ids, 0, sizeof(ids));
/* probe all MTD devices */
for (uclass_first_device(UCLASS_MTD, &dev); dev;
uclass_next_device(&dev)) {
debug("mtd device = %s\n", dev->name);
}
mtd = get_mtd_device_nm("nor0");
if (!IS_ERR_OR_NULL(mtd)) {
mtd_partition = MTDPARTS_NOR0;
board_get_mtdparts("nor0", mtd_partition, ids, parts);
put_mtd_device(mtd);
}
mtd = get_mtd_device_nm("nor1");
if (!IS_ERR_OR_NULL(mtd)) {
mtd_partition = MTDPARTS_NOR1;
board_get_mtdparts("nor1", mtd_partition, ids, parts);
put_mtd_device(mtd);
}
mtd_initialized = true;
*mtdids = ids;
*mtdparts = parts;
debug("%s:mtdids=%s & mtdparts=%s\n", __func__, ids, parts);
+} +#endif /* CONFIG_EFI_CAPSULE_FIRMWARE_MANAGEMENT */ diff --git a/include/configs/qemu-arm.h b/include/configs/qemu-arm.h index 273fa1a7d7..69ff329434 100644 --- a/include/configs/qemu-arm.h +++ b/include/configs/qemu-arm.h @@ -32,6 +32,13 @@
#include <config_distro_bootcmd.h>
+#if defined(CONFIG_EFI_CAPSULE_FIRMWARE_MANAGEMENT) +#define CONFIG_SYS_MTDPARTS_RUNTIME +#endif
+#define MTDPARTS_NOR0 "64m(u-boot)\0" +#define MTDPARTS_NOR1 "64m(u-boot-env)\0"
- #define CONFIG_EXTRA_ENV_SETTINGS \ "fdt_high=0xffffffff\0" \ "initrd_high=0xffffffff\0" \

On Sat, Dec 05, 2020 at 11:29:58AM +0100, Heinrich Schuchardt wrote:
On 11/26/20 7:41 PM, Sughosh Ganu wrote:
Add support for setting the default values for mtd partitions on the platform for the nor flash. This would be used for updating the firmware image using uefi capsule update with the dfu mtd backend driver.
Signed-off-by: Sughosh Ganu sughosh.ganu@linaro.org
board/emulation/qemu-arm/qemu-arm.c | 70 +++++++++++++++++++++++++++++ include/configs/qemu-arm.h | 7 +++ 2 files changed, 77 insertions(+)
diff --git a/board/emulation/qemu-arm/qemu-arm.c b/board/emulation/qemu-arm/qemu-arm.c index b3d5b3d5c2..d5ed3eebaf 100644 --- a/board/emulation/qemu-arm/qemu-arm.c +++ b/board/emulation/qemu-arm/qemu-arm.c
[snip]
+#if defined(CONFIG_EFI_CAPSULE_FIRMWARE_MANAGEMENT) +#define CONFIG_SYS_MTDPARTS_RUNTIME +#endif
This symbol is in Kconfig and needs to be enabled that way.

On Tue, 8 Dec 2020 at 00:14, Tom Rini trini@konsulko.com wrote:
On Sat, Dec 05, 2020 at 11:29:58AM +0100, Heinrich Schuchardt wrote:
On 11/26/20 7:41 PM, Sughosh Ganu wrote:
Add support for setting the default values for mtd partitions on the platform for the nor flash. This would be used for updating the firmware image using uefi capsule update with the dfu mtd backend driver.
Signed-off-by: Sughosh Ganu sughosh.ganu@linaro.org
board/emulation/qemu-arm/qemu-arm.c | 70
+++++++++++++++++++++++++++++
include/configs/qemu-arm.h | 7 +++ 2 files changed, 77 insertions(+)
diff --git a/board/emulation/qemu-arm/qemu-arm.c
b/board/emulation/qemu-arm/qemu-arm.c
index b3d5b3d5c2..d5ed3eebaf 100644 --- a/board/emulation/qemu-arm/qemu-arm.c +++ b/board/emulation/qemu-arm/qemu-arm.c
[snip]
+#if defined(CONFIG_EFI_CAPSULE_FIRMWARE_MANAGEMENT) +#define CONFIG_SYS_MTDPARTS_RUNTIME +#endif
This symbol is in Kconfig and needs to be enabled that way.
Will make the necessary change in the next version. Thanks.
-sughosh

The dfu framework uses the dfu_alt_info environment variable to get information that is needed for performing the firmware update. Set the dfu_alt_info for the platform to reflect the two mtd partitions created for the u-boot env and the firmware image.
Signed-off-by: Sughosh Ganu sughosh.ganu@linaro.org --- board/emulation/qemu-arm/qemu-arm.c | 55 +++++++++++++++++++++++++++++ include/configs/qemu-arm.h | 1 + 2 files changed, 56 insertions(+)
diff --git a/board/emulation/qemu-arm/qemu-arm.c b/board/emulation/qemu-arm/qemu-arm.c index d5ed3eebaf..8cad54c76f 100644 --- a/board/emulation/qemu-arm/qemu-arm.c +++ b/board/emulation/qemu-arm/qemu-arm.c @@ -200,8 +200,63 @@ void flash_write32(u32 value, void *addr)
#if defined(CONFIG_EFI_CAPSULE_FIRMWARE_MANAGEMENT)
+#include <memalign.h> #include <mtd.h>
+#define MTDPARTS_LEN 256 +#define MTDIDS_LEN 128 + +#define DFU_ALT_BUF_LEN SZ_1K + +static void board_get_alt_info(struct mtd_info *mtd, char *buf) +{ + struct mtd_info *part; + bool first = true; + const char *name; + int len, partnum = 0; + + name = mtd->name; + len = strlen(buf); + + if (buf[0] != '\0') + len += snprintf(buf + len, DFU_ALT_BUF_LEN - len, "&"); + len += snprintf(buf + len, DFU_ALT_BUF_LEN - len, + "mtd %s=", name); + + list_for_each_entry(part, &mtd->partitions, node) { + partnum++; + if (!first) + len += snprintf(buf + len, DFU_ALT_BUF_LEN - len, ";"); + first = false; + + len += snprintf(buf + len, DFU_ALT_BUF_LEN - len, + "%s part %d", + part->name, partnum); + } +} + +void set_dfu_alt_info(char *interface, char *devstr) +{ + struct mtd_info *mtd; + + ALLOC_CACHE_ALIGN_BUFFER(char, buf, DFU_ALT_BUF_LEN); + + if (env_get("dfu_alt_info")) + return; + + memset(buf, 0, sizeof(buf)); + + /* probe all MTD devices */ + mtd_probe_devices(); + + mtd = get_mtd_device_nm("nor0"); + if (!IS_ERR_OR_NULL(mtd)) + board_get_alt_info(mtd, buf); + + env_set("dfu_alt_info", buf); + printf("dfu_alt_info set\n"); +} + static void board_get_mtdparts(const char *dev, const char *partition, char *mtdids, char *mtdparts) { diff --git a/include/configs/qemu-arm.h b/include/configs/qemu-arm.h index 69ff329434..726f985d35 100644 --- a/include/configs/qemu-arm.h +++ b/include/configs/qemu-arm.h @@ -33,6 +33,7 @@ #include <config_distro_bootcmd.h>
#if defined(CONFIG_EFI_CAPSULE_FIRMWARE_MANAGEMENT) +#define CONFIG_SET_DFU_ALT_INFO #define CONFIG_SYS_MTDPARTS_RUNTIME #endif

On 11/26/20 7:41 PM, Sughosh Ganu wrote:
The dfu framework uses the dfu_alt_info environment variable to get information that is needed for performing the firmware update. Set the dfu_alt_info for the platform to reflect the two mtd partitions created for the u-boot env and the firmware image.
Signed-off-by: Sughosh Ganu sughosh.ganu@linaro.org
I can't see anything QEMU specific in this patch. Why is the code under board/emulation/?
Best regards
Heinrich
board/emulation/qemu-arm/qemu-arm.c | 55 +++++++++++++++++++++++++++++ include/configs/qemu-arm.h | 1 + 2 files changed, 56 insertions(+)
diff --git a/board/emulation/qemu-arm/qemu-arm.c b/board/emulation/qemu-arm/qemu-arm.c index d5ed3eebaf..8cad54c76f 100644 --- a/board/emulation/qemu-arm/qemu-arm.c +++ b/board/emulation/qemu-arm/qemu-arm.c @@ -200,8 +200,63 @@ void flash_write32(u32 value, void *addr)
#if defined(CONFIG_EFI_CAPSULE_FIRMWARE_MANAGEMENT)
+#include <memalign.h> #include <mtd.h>
+#define MTDPARTS_LEN 256 +#define MTDIDS_LEN 128
+#define DFU_ALT_BUF_LEN SZ_1K
+static void board_get_alt_info(struct mtd_info *mtd, char *buf) +{
- struct mtd_info *part;
- bool first = true;
- const char *name;
- int len, partnum = 0;
- name = mtd->name;
- len = strlen(buf);
- if (buf[0] != '\0')
len += snprintf(buf + len, DFU_ALT_BUF_LEN - len, "&");
- len += snprintf(buf + len, DFU_ALT_BUF_LEN - len,
"mtd %s=", name);
- list_for_each_entry(part, &mtd->partitions, node) {
partnum++;
if (!first)
len += snprintf(buf + len, DFU_ALT_BUF_LEN - len, ";");
first = false;
len += snprintf(buf + len, DFU_ALT_BUF_LEN - len,
"%s part %d",
part->name, partnum);
- }
+}
+void set_dfu_alt_info(char *interface, char *devstr) +{
- struct mtd_info *mtd;
- ALLOC_CACHE_ALIGN_BUFFER(char, buf, DFU_ALT_BUF_LEN);
- if (env_get("dfu_alt_info"))
return;
- memset(buf, 0, sizeof(buf));
- /* probe all MTD devices */
- mtd_probe_devices();
- mtd = get_mtd_device_nm("nor0");
- if (!IS_ERR_OR_NULL(mtd))
board_get_alt_info(mtd, buf);
- env_set("dfu_alt_info", buf);
- printf("dfu_alt_info set\n");
+}
- static void board_get_mtdparts(const char *dev, const char *partition, char *mtdids, char *mtdparts) {
diff --git a/include/configs/qemu-arm.h b/include/configs/qemu-arm.h index 69ff329434..726f985d35 100644 --- a/include/configs/qemu-arm.h +++ b/include/configs/qemu-arm.h @@ -33,6 +33,7 @@ #include <config_distro_bootcmd.h>
#if defined(CONFIG_EFI_CAPSULE_FIRMWARE_MANAGEMENT) +#define CONFIG_SET_DFU_ALT_INFO #define CONFIG_SYS_MTDPARTS_RUNTIME #endif

On Sat, 5 Dec 2020 at 16:01, Heinrich Schuchardt xypron.glpk@gmx.de wrote:
On 11/26/20 7:41 PM, Sughosh Ganu wrote:
The dfu framework uses the dfu_alt_info environment variable to get information that is needed for performing the firmware update. Set the dfu_alt_info for the platform to reflect the two mtd partitions created for the u-boot env and the firmware image.
Signed-off-by: Sughosh Ganu sughosh.ganu@linaro.org
I can't see anything QEMU specific in this patch. Why is the code under board/emulation/?
The device that the board wants to use to set the dfu_alt_info to would be very much board specific. For example, the dfu_alt_info is being set for the qemu platform for an mtd as the interface, with nor0 as the device. Different platforms might have different requirements, like the setting of the variable done for st platforms(board/st/common/stm32mp_dfu.c). I think this information is very much platform specific.
-sughosh
Best regards
Heinrich
board/emulation/qemu-arm/qemu-arm.c | 55 +++++++++++++++++++++++++++++ include/configs/qemu-arm.h | 1 + 2 files changed, 56 insertions(+)
diff --git a/board/emulation/qemu-arm/qemu-arm.c
b/board/emulation/qemu-arm/qemu-arm.c
index d5ed3eebaf..8cad54c76f 100644 --- a/board/emulation/qemu-arm/qemu-arm.c +++ b/board/emulation/qemu-arm/qemu-arm.c @@ -200,8 +200,63 @@ void flash_write32(u32 value, void *addr)
#if defined(CONFIG_EFI_CAPSULE_FIRMWARE_MANAGEMENT)
+#include <memalign.h> #include <mtd.h>
+#define MTDPARTS_LEN 256 +#define MTDIDS_LEN 128
+#define DFU_ALT_BUF_LEN SZ_1K
+static void board_get_alt_info(struct mtd_info *mtd, char *buf) +{
struct mtd_info *part;
bool first = true;
const char *name;
int len, partnum = 0;
name = mtd->name;
len = strlen(buf);
if (buf[0] != '\0')
len += snprintf(buf + len, DFU_ALT_BUF_LEN - len, "&");
len += snprintf(buf + len, DFU_ALT_BUF_LEN - len,
"mtd %s=", name);
list_for_each_entry(part, &mtd->partitions, node) {
partnum++;
if (!first)
len += snprintf(buf + len, DFU_ALT_BUF_LEN - len,
";");
first = false;
len += snprintf(buf + len, DFU_ALT_BUF_LEN - len,
"%s part %d",
part->name, partnum);
}
+}
+void set_dfu_alt_info(char *interface, char *devstr) +{
struct mtd_info *mtd;
ALLOC_CACHE_ALIGN_BUFFER(char, buf, DFU_ALT_BUF_LEN);
if (env_get("dfu_alt_info"))
return;
memset(buf, 0, sizeof(buf));
/* probe all MTD devices */
mtd_probe_devices();
mtd = get_mtd_device_nm("nor0");
if (!IS_ERR_OR_NULL(mtd))
board_get_alt_info(mtd, buf);
env_set("dfu_alt_info", buf);
printf("dfu_alt_info set\n");
+}
- static void board_get_mtdparts(const char *dev, const char *partition, char *mtdids, char *mtdparts) {
diff --git a/include/configs/qemu-arm.h b/include/configs/qemu-arm.h index 69ff329434..726f985d35 100644 --- a/include/configs/qemu-arm.h +++ b/include/configs/qemu-arm.h @@ -33,6 +33,7 @@ #include <config_distro_bootcmd.h>
#if defined(CONFIG_EFI_CAPSULE_FIRMWARE_MANAGEMENT) +#define CONFIG_SET_DFU_ALT_INFO #define CONFIG_SYS_MTDPARTS_RUNTIME #endif

On 12/7/20 6:42 AM, Sughosh Ganu wrote:
On Sat, 5 Dec 2020 at 16:01, Heinrich Schuchardt <xypron.glpk@gmx.de mailto:xypron.glpk@gmx.de> wrote:
On 11/26/20 7:41 PM, Sughosh Ganu wrote: > The dfu framework uses the dfu_alt_info environment variable to get > information that is needed for performing the firmware update. Set the > dfu_alt_info for the platform to reflect the two mtd partitions > created for the u-boot env and the firmware image. > > Signed-off-by: Sughosh Ganu <sughosh.ganu@linaro.org <mailto:sughosh.ganu@linaro.org>> I can't see anything QEMU specific in this patch. Why is the code under board/emulation/?
The device that the board wants to use to set the dfu_alt_info to would be very much board specific. For example, the dfu_alt_info is being set for the qemu platform for an mtd as the interface, with nor0 as the device. Different platforms might have different requirements, like the setting of the variable done for st platforms(board/st/common/stm32mp_dfu.c). I think this information is very much platform specific.
In the function below you simply collect MTD partitions. Isn't this something that could be reused for other boards?
On other boards (e.g. board/samsung/common/misc.c) CONFIG_SET_DFU_ALT_INFO controls if set_dfu_alt_info() exists. Should the same be done for QEMU ARM?
Reading through patch 14/14 it is unclear to me how you control to which partition you write TF-A, SPL, U-Boot depending on which of these exists on the MTD device.
Wouldn't it make more sense to encode in the capsule to where it will be written?
Best regards
Heinrich
-sughosh
Best regards Heinrich > --- > board/emulation/qemu-arm/qemu-arm.c | 55 +++++++++++++++++++++++++++++ > include/configs/qemu-arm.h | 1 + > 2 files changed, 56 insertions(+) > > diff --git a/board/emulation/qemu-arm/qemu-arm.c b/board/emulation/qemu-arm/qemu-arm.c > index d5ed3eebaf..8cad54c76f 100644 > --- a/board/emulation/qemu-arm/qemu-arm.c > +++ b/board/emulation/qemu-arm/qemu-arm.c > @@ -200,8 +200,63 @@ void flash_write32(u32 value, void *addr) > > #if defined(CONFIG_EFI_CAPSULE_FIRMWARE_MANAGEMENT) > > +#include <memalign.h> > #include <mtd.h> > > +#define MTDPARTS_LEN 256 > +#define MTDIDS_LEN 128 > + > +#define DFU_ALT_BUF_LEN SZ_1K > + > +static void board_get_alt_info(struct mtd_info *mtd, char *buf) > +{ > + struct mtd_info *part; > + bool first = true; > + const char *name; > + int len, partnum = 0; > + > + name = mtd->name; > + len = strlen(buf); > + > + if (buf[0] != '\0') > + len += snprintf(buf + len, DFU_ALT_BUF_LEN - len, "&"); > + len += snprintf(buf + len, DFU_ALT_BUF_LEN - len, > + "mtd %s=", name); > + > + list_for_each_entry(part, &mtd->partitions, node) { > + partnum++; > + if (!first) > + len += snprintf(buf + len, DFU_ALT_BUF_LEN - len, ";"); > + first = false; > + > + len += snprintf(buf + len, DFU_ALT_BUF_LEN - len, > + "%s part %d", > + part->name, partnum); > + } > +} > + > +void set_dfu_alt_info(char *interface, char *devstr) > +{ > + struct mtd_info *mtd; > + > + ALLOC_CACHE_ALIGN_BUFFER(char, buf, DFU_ALT_BUF_LEN); > + > + if (env_get("dfu_alt_info")) > + return; > + > + memset(buf, 0, sizeof(buf)); > + > + /* probe all MTD devices */ > + mtd_probe_devices(); > + > + mtd = get_mtd_device_nm("nor0"); > + if (!IS_ERR_OR_NULL(mtd)) > + board_get_alt_info(mtd, buf); > + > + env_set("dfu_alt_info", buf); > + printf("dfu_alt_info set\n"); > +} > + > static void board_get_mtdparts(const char *dev, const char *partition, > char *mtdids, char *mtdparts) > { > diff --git a/include/configs/qemu-arm.h b/include/configs/qemu-arm.h > index 69ff329434..726f985d35 100644 > --- a/include/configs/qemu-arm.h > +++ b/include/configs/qemu-arm.h > @@ -33,6 +33,7 @@ > #include <config_distro_bootcmd.h> > > #if defined(CONFIG_EFI_CAPSULE_FIRMWARE_MANAGEMENT) > +#define CONFIG_SET_DFU_ALT_INFO > #define CONFIG_SYS_MTDPARTS_RUNTIME > #endif > >

On Mon, 7 Dec 2020 at 12:26, Heinrich Schuchardt xypron.glpk@gmx.de wrote:
On 12/7/20 6:42 AM, Sughosh Ganu wrote:
On Sat, 5 Dec 2020 at 16:01, Heinrich Schuchardt <xypron.glpk@gmx.de mailto:xypron.glpk@gmx.de> wrote:
On 11/26/20 7:41 PM, Sughosh Ganu wrote: > The dfu framework uses the dfu_alt_info environment variable to
get
> information that is needed for performing the firmware update. Set the > dfu_alt_info for the platform to reflect the two mtd partitions > created for the u-boot env and the firmware image. > > Signed-off-by: Sughosh Ganu <sughosh.ganu@linaro.org <mailto:sughosh.ganu@linaro.org>> I can't see anything QEMU specific in this patch. Why is the code
under
board/emulation/?
The device that the board wants to use to set the dfu_alt_info to would be very much board specific. For example, the dfu_alt_info is being set for the qemu platform for an mtd as the interface, with nor0 as the device. Different platforms might have different requirements, like the setting of the variable done for st platforms(board/st/common/stm32mp_dfu.c). I think this information is very much platform specific.
In the function below you simply collect MTD partitions. Isn't this something that could be reused for other boards?
The function probes all mtd devices, but uses only the nor0 device for setting up of the dfu_alt_info variable. The other nor1 device is not used, since, on this platform configuration(non-secure boot), it has the u-boot environment. So which of the mtd devices to use for setting of the dfu_alt_info is very much platform specific.
On other boards (e.g. board/samsung/common/misc.c) CONFIG_SET_DFU_ALT_INFO controls if set_dfu_alt_info() exists. Should the same be done for QEMU ARM?
Yes, I can add a check for this on the lines that you mention. Will do.
Reading through patch 14/14 it is unclear to me how you control to which partition you write TF-A, SPL, U-Boot depending on which of these exists on the MTD device.
This gets added in patch 05/14. This sets the mtd partitions that are to be set for the platform. On this particular configuration of the platform(non-secure boot), the partition 1 is used for the u-boot image, and the partition 2 is used for the u-boot environment.
Wouldn't it make more sense to encode in the capsule to where it will be written?
We have discussed this earlier[1] that there is no way to encode the target device and partition information as part of the capsule, the way the capsule structure is defined by the uefi specification. That is the reason that it was decided to use the dfu backend for the update using the dfu_alt_info variable to indicate the partitions and device to be used for the update.
-sughosh
[1] - https://lists.denx.de/pipermail/u-boot/2020-May/411236.html
Best regards
Heinrich
-sughosh
Best regards Heinrich > --- > board/emulation/qemu-arm/qemu-arm.c | 55 +++++++++++++++++++++++++++++ > include/configs/qemu-arm.h | 1 + > 2 files changed, 56 insertions(+) > > diff --git a/board/emulation/qemu-arm/qemu-arm.c b/board/emulation/qemu-arm/qemu-arm.c > index d5ed3eebaf..8cad54c76f 100644 > --- a/board/emulation/qemu-arm/qemu-arm.c > +++ b/board/emulation/qemu-arm/qemu-arm.c > @@ -200,8 +200,63 @@ void flash_write32(u32 value, void *addr) > > #if defined(CONFIG_EFI_CAPSULE_FIRMWARE_MANAGEMENT) > > +#include <memalign.h> > #include <mtd.h> > > +#define MTDPARTS_LEN 256 > +#define MTDIDS_LEN 128 > + > +#define DFU_ALT_BUF_LEN SZ_1K > + > +static void board_get_alt_info(struct mtd_info *mtd, char *buf) > +{ > + struct mtd_info *part; > + bool first = true; > + const char *name; > + int len, partnum = 0; > + > + name = mtd->name; > + len = strlen(buf); > + > + if (buf[0] != '\0') > + len += snprintf(buf + len, DFU_ALT_BUF_LEN - len,
"&");
> + len += snprintf(buf + len, DFU_ALT_BUF_LEN - len, > + "mtd %s=", name); > + > + list_for_each_entry(part, &mtd->partitions, node) { > + partnum++; > + if (!first) > + len += snprintf(buf + len, DFU_ALT_BUF_LEN - len, ";"); > + first = false; > + > + len += snprintf(buf + len, DFU_ALT_BUF_LEN - len, > + "%s part %d", > + part->name, partnum); > + } > +} > + > +void set_dfu_alt_info(char *interface, char *devstr) > +{ > + struct mtd_info *mtd; > + > + ALLOC_CACHE_ALIGN_BUFFER(char, buf, DFU_ALT_BUF_LEN); > + > + if (env_get("dfu_alt_info")) > + return; > + > + memset(buf, 0, sizeof(buf)); > + > + /* probe all MTD devices */ > + mtd_probe_devices(); > + > + mtd = get_mtd_device_nm("nor0"); > + if (!IS_ERR_OR_NULL(mtd)) > + board_get_alt_info(mtd, buf); > + > + env_set("dfu_alt_info", buf); > + printf("dfu_alt_info set\n"); > +} > + > static void board_get_mtdparts(const char *dev, const char *partition, > char *mtdids, char *mtdparts) > { > diff --git a/include/configs/qemu-arm.h
b/include/configs/qemu-arm.h
> index 69ff329434..726f985d35 100644 > --- a/include/configs/qemu-arm.h > +++ b/include/configs/qemu-arm.h > @@ -33,6 +33,7 @@ > #include <config_distro_bootcmd.h> > > #if defined(CONFIG_EFI_CAPSULE_FIRMWARE_MANAGEMENT) > +#define CONFIG_SET_DFU_ALT_INFO > #define CONFIG_SYS_MTDPARTS_RUNTIME > #endif > >

On Sat, Dec 05, 2020 at 11:31:49AM +0100, Heinrich Schuchardt wrote:
On 11/26/20 7:41 PM, Sughosh Ganu wrote:
The dfu framework uses the dfu_alt_info environment variable to get information that is needed for performing the firmware update. Set the dfu_alt_info for the platform to reflect the two mtd partitions created for the u-boot env and the firmware image.
Signed-off-by: Sughosh Ganu sughosh.ganu@linaro.org
I can't see anything QEMU specific in this patch. Why is the code under board/emulation/?
Best regards
Heinrich
board/emulation/qemu-arm/qemu-arm.c | 55 +++++++++++++++++++++++++++++ include/configs/qemu-arm.h | 1 + 2 files changed, 56 insertions(+)
diff --git a/board/emulation/qemu-arm/qemu-arm.c b/board/emulation/qemu-arm/qemu-arm.c index d5ed3eebaf..8cad54c76f 100644 --- a/board/emulation/qemu-arm/qemu-arm.c +++ b/board/emulation/qemu-arm/qemu-arm.c @@ -200,8 +200,63 @@ void flash_write32(u32 value, void *addr)
#if defined(CONFIG_EFI_CAPSULE_FIRMWARE_MANAGEMENT)
+#include <memalign.h> #include <mtd.h>
+#define MTDPARTS_LEN 256 +#define MTDIDS_LEN 128
+#define DFU_ALT_BUF_LEN SZ_1K
+static void board_get_alt_info(struct mtd_info *mtd, char *buf) +{
- struct mtd_info *part;
- bool first = true;
- const char *name;
- int len, partnum = 0;
- name = mtd->name;
- len = strlen(buf);
- if (buf[0] != '\0')
len += snprintf(buf + len, DFU_ALT_BUF_LEN - len, "&");
- len += snprintf(buf + len, DFU_ALT_BUF_LEN - len,
"mtd %s=", name);
- list_for_each_entry(part, &mtd->partitions, node) {
partnum++;
if (!first)
len += snprintf(buf + len, DFU_ALT_BUF_LEN - len, ";");
first = false;
len += snprintf(buf + len, DFU_ALT_BUF_LEN - len,
"%s part %d",
part->name, partnum);
- }
+}
+void set_dfu_alt_info(char *interface, char *devstr) +{
- struct mtd_info *mtd;
- ALLOC_CACHE_ALIGN_BUFFER(char, buf, DFU_ALT_BUF_LEN);
- if (env_get("dfu_alt_info"))
return;
- memset(buf, 0, sizeof(buf));
- /* probe all MTD devices */
- mtd_probe_devices();
- mtd = get_mtd_device_nm("nor0");
- if (!IS_ERR_OR_NULL(mtd))
board_get_alt_info(mtd, buf);
- env_set("dfu_alt_info", buf);
- printf("dfu_alt_info set\n");
+}
- static void board_get_mtdparts(const char *dev, const char *partition, char *mtdids, char *mtdparts) {
diff --git a/include/configs/qemu-arm.h b/include/configs/qemu-arm.h index 69ff329434..726f985d35 100644 --- a/include/configs/qemu-arm.h +++ b/include/configs/qemu-arm.h @@ -33,6 +33,7 @@ #include <config_distro_bootcmd.h>
#if defined(CONFIG_EFI_CAPSULE_FIRMWARE_MANAGEMENT) +#define CONFIG_SET_DFU_ALT_INFO #define CONFIG_SYS_MTDPARTS_RUNTIME
First, CONFIG_SET_DFU_ALT_INFO is in Kconfig and needs to be enabled that way. Second, typically we just set the information in the environment part of the header. This is especially true if in both this case and the previous patch to add mtdparts, we don't really have this being dynamic? Or are we really expecting / supporting arbitrary sized flash as this is qemu? Thanks.

On Tue, 8 Dec 2020 at 00:17, Tom Rini trini@konsulko.com wrote:
On Sat, Dec 05, 2020 at 11:31:49AM +0100, Heinrich Schuchardt wrote:
On 11/26/20 7:41 PM, Sughosh Ganu wrote:
The dfu framework uses the dfu_alt_info environment variable to get information that is needed for performing the firmware update. Set the dfu_alt_info for the platform to reflect the two mtd partitions created for the u-boot env and the firmware image.
Signed-off-by: Sughosh Ganu sughosh.ganu@linaro.org
I can't see anything QEMU specific in this patch. Why is the code under board/emulation/?
Best regards
Heinrich
board/emulation/qemu-arm/qemu-arm.c | 55
+++++++++++++++++++++++++++++
include/configs/qemu-arm.h | 1 + 2 files changed, 56 insertions(+)
diff --git a/board/emulation/qemu-arm/qemu-arm.c
b/board/emulation/qemu-arm/qemu-arm.c
index d5ed3eebaf..8cad54c76f 100644 --- a/board/emulation/qemu-arm/qemu-arm.c +++ b/board/emulation/qemu-arm/qemu-arm.c @@ -200,8 +200,63 @@ void flash_write32(u32 value, void *addr)
#if defined(CONFIG_EFI_CAPSULE_FIRMWARE_MANAGEMENT)
+#include <memalign.h> #include <mtd.h>
+#define MTDPARTS_LEN 256 +#define MTDIDS_LEN 128
+#define DFU_ALT_BUF_LEN SZ_1K
+static void board_get_alt_info(struct mtd_info *mtd, char *buf) +{
- struct mtd_info *part;
- bool first = true;
- const char *name;
- int len, partnum = 0;
- name = mtd->name;
- len = strlen(buf);
- if (buf[0] != '\0')
len += snprintf(buf + len, DFU_ALT_BUF_LEN - len, "&");
- len += snprintf(buf + len, DFU_ALT_BUF_LEN - len,
"mtd %s=", name);
- list_for_each_entry(part, &mtd->partitions, node) {
partnum++;
if (!first)
len += snprintf(buf + len, DFU_ALT_BUF_LEN - len,
";");
first = false;
len += snprintf(buf + len, DFU_ALT_BUF_LEN - len,
"%s part %d",
part->name, partnum);
- }
+}
+void set_dfu_alt_info(char *interface, char *devstr) +{
- struct mtd_info *mtd;
- ALLOC_CACHE_ALIGN_BUFFER(char, buf, DFU_ALT_BUF_LEN);
- if (env_get("dfu_alt_info"))
return;
- memset(buf, 0, sizeof(buf));
- /* probe all MTD devices */
- mtd_probe_devices();
- mtd = get_mtd_device_nm("nor0");
- if (!IS_ERR_OR_NULL(mtd))
board_get_alt_info(mtd, buf);
- env_set("dfu_alt_info", buf);
- printf("dfu_alt_info set\n");
+}
- static void board_get_mtdparts(const char *dev, const char
*partition,
char *mtdids, char *mtdparts)
{ diff --git a/include/configs/qemu-arm.h b/include/configs/qemu-arm.h index 69ff329434..726f985d35 100644 --- a/include/configs/qemu-arm.h +++ b/include/configs/qemu-arm.h @@ -33,6 +33,7 @@ #include <config_distro_bootcmd.h>
#if defined(CONFIG_EFI_CAPSULE_FIRMWARE_MANAGEMENT) +#define CONFIG_SET_DFU_ALT_INFO #define CONFIG_SYS_MTDPARTS_RUNTIME
First, CONFIG_SET_DFU_ALT_INFO is in Kconfig and needs to be enabled that way.
Will change in the next version.
Second, typically we just set the information in the environment part of the header. This is especially true if in both this case and the previous patch to add mtdparts, we don't really have this being dynamic? Or are we really expecting / supporting arbitrary sized flash as this is qemu? Thanks.
I am not sure I understand this. One thing that can be done is to move the setting of the mtd partitions to the Kconfig file, on similar lines to what is being done for the st platforms, e.g MTDPARTS_NOR0_TEE. I can move the mtd partition definitions to the Kconfig file if that is what you are asking for.
-sughosh

On Tue, Dec 08, 2020 at 10:48:57AM +0530, Sughosh Ganu wrote:
On Tue, 8 Dec 2020 at 00:17, Tom Rini trini@konsulko.com wrote:
On Sat, Dec 05, 2020 at 11:31:49AM +0100, Heinrich Schuchardt wrote:
On 11/26/20 7:41 PM, Sughosh Ganu wrote:
The dfu framework uses the dfu_alt_info environment variable to get information that is needed for performing the firmware update. Set the dfu_alt_info for the platform to reflect the two mtd partitions created for the u-boot env and the firmware image.
Signed-off-by: Sughosh Ganu sughosh.ganu@linaro.org
I can't see anything QEMU specific in this patch. Why is the code under board/emulation/?
Best regards
Heinrich
board/emulation/qemu-arm/qemu-arm.c | 55
+++++++++++++++++++++++++++++
include/configs/qemu-arm.h | 1 + 2 files changed, 56 insertions(+)
diff --git a/board/emulation/qemu-arm/qemu-arm.c
b/board/emulation/qemu-arm/qemu-arm.c
index d5ed3eebaf..8cad54c76f 100644 --- a/board/emulation/qemu-arm/qemu-arm.c +++ b/board/emulation/qemu-arm/qemu-arm.c @@ -200,8 +200,63 @@ void flash_write32(u32 value, void *addr)
#if defined(CONFIG_EFI_CAPSULE_FIRMWARE_MANAGEMENT)
+#include <memalign.h> #include <mtd.h>
+#define MTDPARTS_LEN 256 +#define MTDIDS_LEN 128
+#define DFU_ALT_BUF_LEN SZ_1K
+static void board_get_alt_info(struct mtd_info *mtd, char *buf) +{
- struct mtd_info *part;
- bool first = true;
- const char *name;
- int len, partnum = 0;
- name = mtd->name;
- len = strlen(buf);
- if (buf[0] != '\0')
len += snprintf(buf + len, DFU_ALT_BUF_LEN - len, "&");
- len += snprintf(buf + len, DFU_ALT_BUF_LEN - len,
"mtd %s=", name);
- list_for_each_entry(part, &mtd->partitions, node) {
partnum++;
if (!first)
len += snprintf(buf + len, DFU_ALT_BUF_LEN - len,
";");
first = false;
len += snprintf(buf + len, DFU_ALT_BUF_LEN - len,
"%s part %d",
part->name, partnum);
- }
+}
+void set_dfu_alt_info(char *interface, char *devstr) +{
- struct mtd_info *mtd;
- ALLOC_CACHE_ALIGN_BUFFER(char, buf, DFU_ALT_BUF_LEN);
- if (env_get("dfu_alt_info"))
return;
- memset(buf, 0, sizeof(buf));
- /* probe all MTD devices */
- mtd_probe_devices();
- mtd = get_mtd_device_nm("nor0");
- if (!IS_ERR_OR_NULL(mtd))
board_get_alt_info(mtd, buf);
- env_set("dfu_alt_info", buf);
- printf("dfu_alt_info set\n");
+}
- static void board_get_mtdparts(const char *dev, const char
*partition,
char *mtdids, char *mtdparts)
{ diff --git a/include/configs/qemu-arm.h b/include/configs/qemu-arm.h index 69ff329434..726f985d35 100644 --- a/include/configs/qemu-arm.h +++ b/include/configs/qemu-arm.h @@ -33,6 +33,7 @@ #include <config_distro_bootcmd.h>
#if defined(CONFIG_EFI_CAPSULE_FIRMWARE_MANAGEMENT) +#define CONFIG_SET_DFU_ALT_INFO #define CONFIG_SYS_MTDPARTS_RUNTIME
First, CONFIG_SET_DFU_ALT_INFO is in Kconfig and needs to be enabled that way.
Will change in the next version.
Second, typically we just set the information in the environment part of the header. This is especially true if in both this case and the previous patch to add mtdparts, we don't really have this being dynamic? Or are we really expecting / supporting arbitrary sized flash as this is qemu? Thanks.
I am not sure I understand this. One thing that can be done is to move the setting of the mtd partitions to the Kconfig file, on similar lines to what is being done for the st platforms, e.g MTDPARTS_NOR0_TEE. I can move the mtd partition definitions to the Kconfig file if that is what you are asking for.
Environment manipulation via C code is possible, but discouraged. What can be set via Kconfig should be set that way, and what can be put in the environment part of the header should be done that way. Does that help clarify?

On Tue, 8 Dec 2020 at 17:50, Tom Rini trini@konsulko.com wrote:
On Tue, Dec 08, 2020 at 10:48:57AM +0530, Sughosh Ganu wrote:
On Tue, 8 Dec 2020 at 00:17, Tom Rini trini@konsulko.com wrote:
On Sat, Dec 05, 2020 at 11:31:49AM +0100, Heinrich Schuchardt wrote:
On 11/26/20 7:41 PM, Sughosh Ganu wrote:
The dfu framework uses the dfu_alt_info environment variable to get information that is needed for performing the firmware update. Set
the
dfu_alt_info for the platform to reflect the two mtd partitions created for the u-boot env and the firmware image.
Signed-off-by: Sughosh Ganu sughosh.ganu@linaro.org
I can't see anything QEMU specific in this patch. Why is the code
under
board/emulation/?
Best regards
Heinrich
board/emulation/qemu-arm/qemu-arm.c | 55
+++++++++++++++++++++++++++++
include/configs/qemu-arm.h | 1 + 2 files changed, 56 insertions(+)
diff --git a/board/emulation/qemu-arm/qemu-arm.c
b/board/emulation/qemu-arm/qemu-arm.c
index d5ed3eebaf..8cad54c76f 100644 --- a/board/emulation/qemu-arm/qemu-arm.c +++ b/board/emulation/qemu-arm/qemu-arm.c @@ -200,8 +200,63 @@ void flash_write32(u32 value, void *addr)
#if defined(CONFIG_EFI_CAPSULE_FIRMWARE_MANAGEMENT)
+#include <memalign.h> #include <mtd.h>
+#define MTDPARTS_LEN 256 +#define MTDIDS_LEN 128
+#define DFU_ALT_BUF_LEN SZ_1K
+static void board_get_alt_info(struct mtd_info *mtd, char *buf) +{
- struct mtd_info *part;
- bool first = true;
- const char *name;
- int len, partnum = 0;
- name = mtd->name;
- len = strlen(buf);
- if (buf[0] != '\0')
len += snprintf(buf + len, DFU_ALT_BUF_LEN - len, "&");
- len += snprintf(buf + len, DFU_ALT_BUF_LEN - len,
"mtd %s=", name);
- list_for_each_entry(part, &mtd->partitions, node) {
partnum++;
if (!first)
len += snprintf(buf + len, DFU_ALT_BUF_LEN -
len,
";");
first = false;
len += snprintf(buf + len, DFU_ALT_BUF_LEN - len,
"%s part %d",
part->name, partnum);
- }
+}
+void set_dfu_alt_info(char *interface, char *devstr) +{
- struct mtd_info *mtd;
- ALLOC_CACHE_ALIGN_BUFFER(char, buf, DFU_ALT_BUF_LEN);
- if (env_get("dfu_alt_info"))
return;
- memset(buf, 0, sizeof(buf));
- /* probe all MTD devices */
- mtd_probe_devices();
- mtd = get_mtd_device_nm("nor0");
- if (!IS_ERR_OR_NULL(mtd))
board_get_alt_info(mtd, buf);
- env_set("dfu_alt_info", buf);
- printf("dfu_alt_info set\n");
+}
- static void board_get_mtdparts(const char *dev, const char
*partition,
char *mtdids, char *mtdparts)
{ diff --git a/include/configs/qemu-arm.h
b/include/configs/qemu-arm.h
index 69ff329434..726f985d35 100644 --- a/include/configs/qemu-arm.h +++ b/include/configs/qemu-arm.h @@ -33,6 +33,7 @@ #include <config_distro_bootcmd.h>
#if defined(CONFIG_EFI_CAPSULE_FIRMWARE_MANAGEMENT) +#define CONFIG_SET_DFU_ALT_INFO #define CONFIG_SYS_MTDPARTS_RUNTIME
First, CONFIG_SET_DFU_ALT_INFO is in Kconfig and needs to be enabled that way.
Will change in the next version.
Second, typically we just set the information in the environment part of the header. This is especially true if in both
this
case and the previous patch to add mtdparts, we don't really have this being dynamic? Or are we really expecting / supporting arbitrary sized flash as this is qemu? Thanks.
I am not sure I understand this. One thing that can be done is to move
the
setting of the mtd partitions to the Kconfig file, on similar lines to
what
is being done for the st platforms, e.g MTDPARTS_NOR0_TEE. I can move the mtd partition definitions to the Kconfig file if that is what you are asking for.
Environment manipulation via C code is possible, but discouraged. What can be set via Kconfig should be set that way, and what can be put in the environment part of the header should be done that way. Does that help clarify?
Yes, that does clarify. I will move the settings of the partitions used for mtdparts in the Kconfig, similar to the way it is being done for the st platforms. Thanks.
-sughosh

When building the capsule using scripts in edk2, an fmp header is added on top of the binary payload. Add a config option to indicate the presence of the header. When enabled, the pointer to the image needs to be adjusted as per the size of the header to point to the actual binary payload.
Signed-off-by: Sughosh Ganu sughosh.ganu@linaro.org --- lib/efi_loader/Kconfig | 7 +++++++ lib/efi_loader/efi_firmware.c | 12 ++++++++++++ 2 files changed, 19 insertions(+)
diff --git a/lib/efi_loader/Kconfig b/lib/efi_loader/Kconfig index 0d1b1b5356..55e4787e32 100644 --- a/lib/efi_loader/Kconfig +++ b/lib/efi_loader/Kconfig @@ -138,6 +138,13 @@ config EFI_CAPSULE_FIRMWARE_MANAGEMENT Select this option if you want to enable capsule-based firmware update using Firmware Management Protocol.
+config EFI_CAPSULE_FMP_HEADER + bool "Capsule uses FMP header" + depends on EFI_CAPSULE_FIRMWARE_MANAGEMENT + help + Select this option if the capsule is built using the + scripts in edk2. + config EFI_CAPSULE_FIRMWARE_FIT bool "FMP driver for FIT image" depends on EFI_CAPSULE_FIRMWARE_MANAGEMENT diff --git a/lib/efi_loader/efi_firmware.c b/lib/efi_loader/efi_firmware.c index 7e56077383..6c97604d8b 100644 --- a/lib/efi_loader/efi_firmware.c +++ b/lib/efi_loader/efi_firmware.c @@ -385,10 +385,22 @@ efi_status_t EFIAPI efi_firmware_raw_set_image( if (!image) return EFI_EXIT(EFI_INVALID_PARAMETER);
+ if (CONFIG_IS_ENABLED(EFI_CAPSULE_FMP_HEADER)) { + /* + * When building the capsule with the scripts in + * edk2, a FMP header is inserted above the capsule + * payload. Compensate for this header to get the + * actual payload that is to be updated. + */ + image += 0x10; + image_size -= 0x10; + } + if (dfu_write_by_alt(image_index - 1, (void *)image, image_size, NULL, NULL)) return EFI_EXIT(EFI_DEVICE_ERROR);
+ printf("%s: Capsule update complete!\n", __func__); return EFI_EXIT(EFI_SUCCESS); }

On 11/26/20 7:41 PM, Sughosh Ganu wrote:
When building the capsule using scripts in edk2, an fmp header is added on top of the binary payload. Add a config option to indicate the presence of the header. When enabled, the pointer to the image needs to be adjusted as per the size of the header to point to the actual binary payload.
Why do we need a config option? Can't we detect the header?
Can we harmonize the capsule format in EDK II and U-Boot?
Best regards
Heinrich
Signed-off-by: Sughosh Ganu sughosh.ganu@linaro.org
lib/efi_loader/Kconfig | 7 +++++++ lib/efi_loader/efi_firmware.c | 12 ++++++++++++ 2 files changed, 19 insertions(+)
diff --git a/lib/efi_loader/Kconfig b/lib/efi_loader/Kconfig index 0d1b1b5356..55e4787e32 100644 --- a/lib/efi_loader/Kconfig +++ b/lib/efi_loader/Kconfig @@ -138,6 +138,13 @@ config EFI_CAPSULE_FIRMWARE_MANAGEMENT Select this option if you want to enable capsule-based firmware update using Firmware Management Protocol.
+config EFI_CAPSULE_FMP_HEADER
- bool "Capsule uses FMP header"
- depends on EFI_CAPSULE_FIRMWARE_MANAGEMENT
- help
Select this option if the capsule is built using the
scripts in edk2.
- config EFI_CAPSULE_FIRMWARE_FIT bool "FMP driver for FIT image" depends on EFI_CAPSULE_FIRMWARE_MANAGEMENT
diff --git a/lib/efi_loader/efi_firmware.c b/lib/efi_loader/efi_firmware.c index 7e56077383..6c97604d8b 100644 --- a/lib/efi_loader/efi_firmware.c +++ b/lib/efi_loader/efi_firmware.c @@ -385,10 +385,22 @@ efi_status_t EFIAPI efi_firmware_raw_set_image( if (!image) return EFI_EXIT(EFI_INVALID_PARAMETER);
if (CONFIG_IS_ENABLED(EFI_CAPSULE_FMP_HEADER)) {
/*
* When building the capsule with the scripts in
* edk2, a FMP header is inserted above the capsule
* payload. Compensate for this header to get the
* actual payload that is to be updated.
*/
image += 0x10;
image_size -= 0x10;
}
if (dfu_write_by_alt(image_index - 1, (void *)image, image_size, NULL, NULL)) return EFI_EXIT(EFI_DEVICE_ERROR);
printf("%s: Capsule update complete!\n", __func__); return EFI_EXIT(EFI_SUCCESS); }

On Sat, 5 Dec 2020 at 16:04, Heinrich Schuchardt xypron.glpk@gmx.de wrote:
On 11/26/20 7:41 PM, Sughosh Ganu wrote:
When building the capsule using scripts in edk2, an fmp header is added on top of the binary payload. Add a config option to indicate the presence of the header. When enabled, the pointer to the image needs to be adjusted as per the size of the header to point to the actual binary payload.
Why do we need a config option? Can't we detect the header?
We do have a header signature that can be read, so yes this can be detected at runtime.
Can we harmonize the capsule format in EDK II and U-Boot?
I am not sure about this though. The FMP payload header that gets added by the edk2 scripts is not defined by the uefi specification but is a edk2 specific structure. I am not sure if defining it in u-boot would make sense.
-sughosh
Best regards
Heinrich
Signed-off-by: Sughosh Ganu sughosh.ganu@linaro.org
lib/efi_loader/Kconfig | 7 +++++++ lib/efi_loader/efi_firmware.c | 12 ++++++++++++ 2 files changed, 19 insertions(+)
diff --git a/lib/efi_loader/Kconfig b/lib/efi_loader/Kconfig index 0d1b1b5356..55e4787e32 100644 --- a/lib/efi_loader/Kconfig +++ b/lib/efi_loader/Kconfig @@ -138,6 +138,13 @@ config EFI_CAPSULE_FIRMWARE_MANAGEMENT Select this option if you want to enable capsule-based firmware update using Firmware Management Protocol.
+config EFI_CAPSULE_FMP_HEADER
bool "Capsule uses FMP header"
depends on EFI_CAPSULE_FIRMWARE_MANAGEMENT
help
Select this option if the capsule is built using the
scripts in edk2.
- config EFI_CAPSULE_FIRMWARE_FIT bool "FMP driver for FIT image" depends on EFI_CAPSULE_FIRMWARE_MANAGEMENT
diff --git a/lib/efi_loader/efi_firmware.c
b/lib/efi_loader/efi_firmware.c
index 7e56077383..6c97604d8b 100644 --- a/lib/efi_loader/efi_firmware.c +++ b/lib/efi_loader/efi_firmware.c @@ -385,10 +385,22 @@ efi_status_t EFIAPI efi_firmware_raw_set_image( if (!image) return EFI_EXIT(EFI_INVALID_PARAMETER);
if (CONFIG_IS_ENABLED(EFI_CAPSULE_FMP_HEADER)) {
/*
* When building the capsule with the scripts in
* edk2, a FMP header is inserted above the capsule
* payload. Compensate for this header to get the
* actual payload that is to be updated.
*/
image += 0x10;
image_size -= 0x10;
}
if (dfu_write_by_alt(image_index - 1, (void *)image, image_size, NULL, NULL)) return EFI_EXIT(EFI_DEVICE_ERROR);
printf("%s: Capsule update complete!\n", __func__); return EFI_EXIT(EFI_SUCCESS);
}

Prior to writing to an mtd device, mtd_erase is called. This call fails in case the sector being erased is locked. Call mtd_unlock to unlock the region which is to be erased and later written to. Lock the region once the write to the region has completed.
Signed-off-by: Sughosh Ganu sughosh.ganu@linaro.org --- drivers/dfu/dfu_mtd.c | 20 +++++++++++++++++--- 1 file changed, 17 insertions(+), 3 deletions(-)
diff --git a/drivers/dfu/dfu_mtd.c b/drivers/dfu/dfu_mtd.c index 36cd4e945b..b34975dbb0 100644 --- a/drivers/dfu/dfu_mtd.c +++ b/drivers/dfu/dfu_mtd.c @@ -21,7 +21,7 @@ static bool mtd_is_aligned_with_block_size(struct mtd_info *mtd, u64 size) static int mtd_block_op(enum dfu_op op, struct dfu_entity *dfu, u64 offset, void *buf, long *len) { - u64 off, lim, remaining; + u64 off, lim, remaining, lock_ofs, lock_len; struct mtd_info *mtd = dfu->data.mtd.info; struct mtd_oob_ops io_op = {}; int ret = 0; @@ -34,7 +34,7 @@ static int mtd_block_op(enum dfu_op op, struct dfu_entity *dfu, return 0; }
- off = dfu->data.mtd.start + offset + dfu->bad_skip; + off = lock_ofs = dfu->data.mtd.start + offset + dfu->bad_skip; lim = dfu->data.mtd.start + dfu->data.mtd.size;
if (off >= lim) { @@ -56,12 +56,19 @@ static int mtd_block_op(enum dfu_op op, struct dfu_entity *dfu, if (op == DFU_OP_WRITE) { struct erase_info erase_op = {};
- remaining = round_up(*len, mtd->erasesize); + remaining = lock_len = round_up(*len, mtd->erasesize); erase_op.mtd = mtd; erase_op.addr = off; erase_op.len = mtd->erasesize; erase_op.scrub = 0;
+ debug("Unlocking the mtd device\n"); + ret = mtd_unlock(mtd, lock_ofs, lock_len); + if (ret && ret != -EOPNOTSUPP) { + printf("MTD device unlock failed\n"); + return 0; + } + while (remaining) { if (erase_op.addr + remaining > lim) { printf("Limit reached 0x%llx while erasing at offset 0x%llx\n", @@ -139,6 +146,13 @@ static int mtd_block_op(enum dfu_op op, struct dfu_entity *dfu, io_op.len = mtd->writesize; }
+ if (op == DFU_OP_WRITE) { + /* Write done, lock again */ + debug("Locking the mtd device\n"); + ret = mtd_lock(mtd, lock_ofs, lock_len); + if (ret && ret != -EOPNOTSUPP) + printf("MTD device lock failed\n"); + } return ret; }

The pkcs7 header parsing functionality is pretty generic, and can be used by other features like capsule authentication. Make the function an extern, also changing it's name to efi_parse_pkcs7_header
Signed-off-by: Sughosh Ganu sughosh.ganu@linaro.org --- include/efi_loader.h | 4 ++ lib/efi_loader/efi_signature.c | 85 +++++++++++++++++++++++++++++++ lib/efi_loader/efi_variable.c | 93 ++-------------------------------- 3 files changed, 93 insertions(+), 89 deletions(-)
diff --git a/include/efi_loader.h b/include/efi_loader.h index 76cd2b36f2..b9226208f5 100644 --- a/include/efi_loader.h +++ b/include/efi_loader.h @@ -810,6 +810,10 @@ bool efi_secure_boot_enabled(void); bool efi_image_parse(void *efi, size_t len, struct efi_image_regions **regp, WIN_CERTIFICATE **auth, size_t *auth_len);
+struct pkcs7_message *efi_parse_pkcs7_header(const void *buf, + size_t buflen, + u8 **tmpbuf); + /* runtime implementation of memcpy() */ void efi_memcpy_runtime(void *dest, const void *src, size_t n);
diff --git a/lib/efi_loader/efi_signature.c b/lib/efi_loader/efi_signature.c index 79dee27421..9ab071b611 100644 --- a/lib/efi_loader/efi_signature.c +++ b/lib/efi_loader/efi_signature.c @@ -27,6 +27,91 @@ const efi_guid_t efi_guid_cert_x509_sha256 = EFI_CERT_X509_SHA256_GUID; const efi_guid_t efi_guid_cert_type_pkcs7 = EFI_CERT_TYPE_PKCS7_GUID;
#ifdef CONFIG_EFI_SECURE_BOOT +static u8 pkcs7_hdr[] = { + /* SEQUENCE */ + 0x30, 0x82, 0x05, 0xc7, + /* OID: pkcs7-signedData */ + 0x06, 0x09, 0x2a, 0x86, 0x48, 0x86, 0xf7, 0x0d, 0x01, 0x07, 0x02, + /* Context Structured? */ + 0xa0, 0x82, 0x05, 0xb8, +}; + +/** + * efi_parse_pkcs7_header - parse a signature in payload + * @buf: Pointer to payload's value + * @buflen: Length of @buf + * @tmpbuf: Pointer to temporary buffer + * + * Parse a signature embedded in payload's value and instantiate + * a pkcs7_message structure. Since pkcs7_parse_message() accepts only + * pkcs7's signedData, some header needed be prepended for correctly + * parsing authentication data + * A temporary buffer will be allocated if needed, and it should be + * kept valid during the authentication because some data in the buffer + * will be referenced by efi_signature_verify(). + * + * Return: Pointer to pkcs7_message structure on success, NULL on error + */ +struct pkcs7_message *efi_parse_pkcs7_header(const void *buf, + size_t buflen, + u8 **tmpbuf) +{ + u8 *ebuf; + size_t ebuflen, len; + struct pkcs7_message *msg; + + /* + * This is the best assumption to check if the binary is + * already in a form of pkcs7's signedData. + */ + if (buflen > sizeof(pkcs7_hdr) && + !memcmp(&((u8 *)buf)[4], &pkcs7_hdr[4], 11)) { + msg = pkcs7_parse_message(buf, buflen); + if (IS_ERR(msg)) + return NULL; + return msg; + } + + /* + * Otherwise, we should add a dummy prefix sequence for pkcs7 + * message parser to be able to process. + * NOTE: EDK2 also uses similar hack in WrapPkcs7Data() + * in CryptoPkg/Library/BaseCryptLib/Pk/CryptPkcs7VerifyCommon.c + * TODO: + * The header should be composed in a more refined manner. + */ + EFI_PRINT("Makeshift prefix added to authentication data\n"); + ebuflen = sizeof(pkcs7_hdr) + buflen; + if (ebuflen <= 0x7f) { + EFI_PRINT("Data is too short\n"); + return NULL; + } + + ebuf = malloc(ebuflen); + if (!ebuf) { + EFI_PRINT("Out of memory\n"); + return NULL; + } + + memcpy(ebuf, pkcs7_hdr, sizeof(pkcs7_hdr)); + memcpy(ebuf + sizeof(pkcs7_hdr), buf, buflen); + len = ebuflen - 4; + ebuf[2] = (len >> 8) & 0xff; + ebuf[3] = len & 0xff; + len = ebuflen - 0x13; + ebuf[0x11] = (len >> 8) & 0xff; + ebuf[0x12] = len & 0xff; + + msg = pkcs7_parse_message(ebuf, ebuflen); + + if (IS_ERR(msg)) { + free(ebuf); + return NULL; + } + + *tmpbuf = ebuf; + return msg; +}
/** * efi_hash_regions - calculate a hash value diff --git a/lib/efi_loader/efi_variable.c b/lib/efi_loader/efi_variable.c index 0c689cfb47..ba0874e9e7 100644 --- a/lib/efi_loader/efi_variable.c +++ b/lib/efi_loader/efi_variable.c @@ -24,91 +24,6 @@ #include <asm/sections.h>
#ifdef CONFIG_EFI_SECURE_BOOT -static u8 pkcs7_hdr[] = { - /* SEQUENCE */ - 0x30, 0x82, 0x05, 0xc7, - /* OID: pkcs7-signedData */ - 0x06, 0x09, 0x2a, 0x86, 0x48, 0x86, 0xf7, 0x0d, 0x01, 0x07, 0x02, - /* Context Structured? */ - 0xa0, 0x82, 0x05, 0xb8, -}; - -/** - * efi_variable_parse_signature - parse a signature in variable - * @buf: Pointer to variable's value - * @buflen: Length of @buf - * @tmpbuf: Pointer to temporary buffer - * - * Parse a signature embedded in variable's value and instantiate - * a pkcs7_message structure. Since pkcs7_parse_message() accepts only - * pkcs7's signedData, some header needed be prepended for correctly - * parsing authentication data, particularly for variable's. - * A temporary buffer will be allocated if needed, and it should be - * kept valid during the authentication because some data in the buffer - * will be referenced by efi_signature_verify(). - * - * Return: Pointer to pkcs7_message structure on success, NULL on error - */ -static struct pkcs7_message *efi_variable_parse_signature(const void *buf, - size_t buflen, - u8 **tmpbuf) -{ - u8 *ebuf; - size_t ebuflen, len; - struct pkcs7_message *msg; - - /* - * This is the best assumption to check if the binary is - * already in a form of pkcs7's signedData. - */ - if (buflen > sizeof(pkcs7_hdr) && - !memcmp(&((u8 *)buf)[4], &pkcs7_hdr[4], 11)) { - msg = pkcs7_parse_message(buf, buflen); - if (IS_ERR(msg)) - return NULL; - return msg; - } - - /* - * Otherwise, we should add a dummy prefix sequence for pkcs7 - * message parser to be able to process. - * NOTE: EDK2 also uses similar hack in WrapPkcs7Data() - * in CryptoPkg/Library/BaseCryptLib/Pk/CryptPkcs7VerifyCommon.c - * TODO: - * The header should be composed in a more refined manner. - */ - EFI_PRINT("Makeshift prefix added to authentication data\n"); - ebuflen = sizeof(pkcs7_hdr) + buflen; - if (ebuflen <= 0x7f) { - EFI_PRINT("Data is too short\n"); - return NULL; - } - - ebuf = malloc(ebuflen); - if (!ebuf) { - EFI_PRINT("Out of memory\n"); - return NULL; - } - - memcpy(ebuf, pkcs7_hdr, sizeof(pkcs7_hdr)); - memcpy(ebuf + sizeof(pkcs7_hdr), buf, buflen); - len = ebuflen - 4; - ebuf[2] = (len >> 8) & 0xff; - ebuf[3] = len & 0xff; - len = ebuflen - 0x13; - ebuf[0x11] = (len >> 8) & 0xff; - ebuf[0x12] = len & 0xff; - - msg = pkcs7_parse_message(ebuf, ebuflen); - - if (IS_ERR(msg)) { - free(ebuf); - return NULL; - } - - *tmpbuf = ebuf; - return msg; -}
/** * efi_variable_authenticate - authenticate a variable @@ -215,10 +130,10 @@ static efi_status_t efi_variable_authenticate(u16 *variable, goto err;
/* ebuf should be kept valid during the authentication */ - var_sig = efi_variable_parse_signature(auth->auth_info.cert_data, - auth->auth_info.hdr.dwLength - - sizeof(auth->auth_info), - &ebuf); + var_sig = efi_parse_pkcs7_header(auth->auth_info.cert_data, + auth->auth_info.hdr.dwLength + - sizeof(auth->auth_info), + &ebuf); if (!var_sig) { EFI_PRINT("Parsing variable's signature failed\n"); goto err;

On 11/26/20 7:41 PM, Sughosh Ganu wrote:
The pkcs7 header parsing functionality is pretty generic, and can be used by other features like capsule authentication. Make the function an extern, also changing it's name to efi_parse_pkcs7_header
Signed-off-by: Sughosh Ganu sughosh.ganu@linaro.org
include/efi_loader.h | 4 ++ lib/efi_loader/efi_signature.c | 85 +++++++++++++++++++++++++++++++ lib/efi_loader/efi_variable.c | 93 ++-------------------------------- 3 files changed, 93 insertions(+), 89 deletions(-)
diff --git a/include/efi_loader.h b/include/efi_loader.h index 76cd2b36f2..b9226208f5 100644 --- a/include/efi_loader.h +++ b/include/efi_loader.h @@ -810,6 +810,10 @@ bool efi_secure_boot_enabled(void); bool efi_image_parse(void *efi, size_t len, struct efi_image_regions **regp, WIN_CERTIFICATE **auth, size_t *auth_len);
+struct pkcs7_message *efi_parse_pkcs7_header(const void *buf,
size_t buflen,
u8 **tmpbuf);
- /* runtime implementation of memcpy() */ void efi_memcpy_runtime(void *dest, const void *src, size_t n);
diff --git a/lib/efi_loader/efi_signature.c b/lib/efi_loader/efi_signature.c index 79dee27421..9ab071b611 100644 --- a/lib/efi_loader/efi_signature.c +++ b/lib/efi_loader/efi_signature.c @@ -27,6 +27,91 @@ const efi_guid_t efi_guid_cert_x509_sha256 = EFI_CERT_X509_SHA256_GUID; const efi_guid_t efi_guid_cert_type_pkcs7 = EFI_CERT_TYPE_PKCS7_GUID;
#ifdef CONFIG_EFI_SECURE_BOOT +static u8 pkcs7_hdr[] = {
- /* SEQUENCE */
- 0x30, 0x82, 0x05, 0xc7,
- /* OID: pkcs7-signedData */
- 0x06, 0x09, 0x2a, 0x86, 0x48, 0x86, 0xf7, 0x0d, 0x01, 0x07, 0x02,
- /* Context Structured? */
- 0xa0, 0x82, 0x05, 0xb8,
Dear Takahiro,
From where are these magic numbers taken? Could you, please, provide a reference that we can add as a comment.
Best regards
Heinrich
+};
+/**
- efi_parse_pkcs7_header - parse a signature in payload
- @buf: Pointer to payload's value
- @buflen: Length of @buf
- @tmpbuf: Pointer to temporary buffer
- Parse a signature embedded in payload's value and instantiate
- a pkcs7_message structure. Since pkcs7_parse_message() accepts only
- pkcs7's signedData, some header needed be prepended for correctly
- parsing authentication data
- A temporary buffer will be allocated if needed, and it should be
- kept valid during the authentication because some data in the buffer
- will be referenced by efi_signature_verify().
- Return: Pointer to pkcs7_message structure on success, NULL on error
- */
+struct pkcs7_message *efi_parse_pkcs7_header(const void *buf,
size_t buflen,
u8 **tmpbuf)
+{
- u8 *ebuf;
- size_t ebuflen, len;
- struct pkcs7_message *msg;
- /*
* This is the best assumption to check if the binary is
* already in a form of pkcs7's signedData.
*/
- if (buflen > sizeof(pkcs7_hdr) &&
!memcmp(&((u8 *)buf)[4], &pkcs7_hdr[4], 11)) {
msg = pkcs7_parse_message(buf, buflen);
if (IS_ERR(msg))
return NULL;
return msg;
- }
- /*
* Otherwise, we should add a dummy prefix sequence for pkcs7
* message parser to be able to process.
* NOTE: EDK2 also uses similar hack in WrapPkcs7Data()
* in CryptoPkg/Library/BaseCryptLib/Pk/CryptPkcs7VerifyCommon.c
* TODO:
* The header should be composed in a more refined manner.
*/
- EFI_PRINT("Makeshift prefix added to authentication data\n");
- ebuflen = sizeof(pkcs7_hdr) + buflen;
- if (ebuflen <= 0x7f) {
EFI_PRINT("Data is too short\n");
return NULL;
- }
- ebuf = malloc(ebuflen);
- if (!ebuf) {
EFI_PRINT("Out of memory\n");
return NULL;
- }
- memcpy(ebuf, pkcs7_hdr, sizeof(pkcs7_hdr));
- memcpy(ebuf + sizeof(pkcs7_hdr), buf, buflen);
- len = ebuflen - 4;
- ebuf[2] = (len >> 8) & 0xff;
- ebuf[3] = len & 0xff;
- len = ebuflen - 0x13;
- ebuf[0x11] = (len >> 8) & 0xff;
- ebuf[0x12] = len & 0xff;
- msg = pkcs7_parse_message(ebuf, ebuflen);
- if (IS_ERR(msg)) {
free(ebuf);
return NULL;
- }
- *tmpbuf = ebuf;
- return msg;
+}
/**
- efi_hash_regions - calculate a hash value
diff --git a/lib/efi_loader/efi_variable.c b/lib/efi_loader/efi_variable.c index 0c689cfb47..ba0874e9e7 100644 --- a/lib/efi_loader/efi_variable.c +++ b/lib/efi_loader/efi_variable.c @@ -24,91 +24,6 @@ #include <asm/sections.h>
#ifdef CONFIG_EFI_SECURE_BOOT -static u8 pkcs7_hdr[] = {
- /* SEQUENCE */
- 0x30, 0x82, 0x05, 0xc7,
- /* OID: pkcs7-signedData */
- 0x06, 0x09, 0x2a, 0x86, 0x48, 0x86, 0xf7, 0x0d, 0x01, 0x07, 0x02,
- /* Context Structured? */
- 0xa0, 0x82, 0x05, 0xb8,
-};
-/**
- efi_variable_parse_signature - parse a signature in variable
- @buf: Pointer to variable's value
- @buflen: Length of @buf
- @tmpbuf: Pointer to temporary buffer
- Parse a signature embedded in variable's value and instantiate
- a pkcs7_message structure. Since pkcs7_parse_message() accepts only
- pkcs7's signedData, some header needed be prepended for correctly
- parsing authentication data, particularly for variable's.
- A temporary buffer will be allocated if needed, and it should be
- kept valid during the authentication because some data in the buffer
- will be referenced by efi_signature_verify().
- Return: Pointer to pkcs7_message structure on success, NULL on error
- */
-static struct pkcs7_message *efi_variable_parse_signature(const void *buf,
size_t buflen,
u8 **tmpbuf)
-{
- u8 *ebuf;
- size_t ebuflen, len;
- struct pkcs7_message *msg;
- /*
* This is the best assumption to check if the binary is
* already in a form of pkcs7's signedData.
*/
- if (buflen > sizeof(pkcs7_hdr) &&
!memcmp(&((u8 *)buf)[4], &pkcs7_hdr[4], 11)) {
msg = pkcs7_parse_message(buf, buflen);
if (IS_ERR(msg))
return NULL;
return msg;
- }
- /*
* Otherwise, we should add a dummy prefix sequence for pkcs7
* message parser to be able to process.
* NOTE: EDK2 also uses similar hack in WrapPkcs7Data()
* in CryptoPkg/Library/BaseCryptLib/Pk/CryptPkcs7VerifyCommon.c
* TODO:
* The header should be composed in a more refined manner.
*/
- EFI_PRINT("Makeshift prefix added to authentication data\n");
- ebuflen = sizeof(pkcs7_hdr) + buflen;
- if (ebuflen <= 0x7f) {
EFI_PRINT("Data is too short\n");
return NULL;
- }
- ebuf = malloc(ebuflen);
- if (!ebuf) {
EFI_PRINT("Out of memory\n");
return NULL;
- }
- memcpy(ebuf, pkcs7_hdr, sizeof(pkcs7_hdr));
- memcpy(ebuf + sizeof(pkcs7_hdr), buf, buflen);
- len = ebuflen - 4;
- ebuf[2] = (len >> 8) & 0xff;
- ebuf[3] = len & 0xff;
- len = ebuflen - 0x13;
- ebuf[0x11] = (len >> 8) & 0xff;
- ebuf[0x12] = len & 0xff;
- msg = pkcs7_parse_message(ebuf, ebuflen);
- if (IS_ERR(msg)) {
free(ebuf);
return NULL;
- }
- *tmpbuf = ebuf;
- return msg;
-}
/**
- efi_variable_authenticate - authenticate a variable
@@ -215,10 +130,10 @@ static efi_status_t efi_variable_authenticate(u16 *variable, goto err;
/* ebuf should be kept valid during the authentication */
- var_sig = efi_variable_parse_signature(auth->auth_info.cert_data,
auth->auth_info.hdr.dwLength
- sizeof(auth->auth_info),
&ebuf);
- var_sig = efi_parse_pkcs7_header(auth->auth_info.cert_data,
auth->auth_info.hdr.dwLength
- sizeof(auth->auth_info),
if (!var_sig) { EFI_PRINT("Parsing variable's signature failed\n"); goto err;&ebuf);

The efi_sigstore_parse_sigdb function reads the uefi authenticated variable, stored in the signature database format and builds the signature store structure. Factor out the code for building the signature store. This can then be used by the capsule authentication routine to build the signature store even when the signature database is not stored as an uefi authenticated variable
Signed-off-by: Sughosh Ganu sughosh.ganu@linaro.org --- include/efi_loader.h | 2 + lib/efi_loader/efi_signature.c | 103 +++++++++++++++++++-------------- 2 files changed, 63 insertions(+), 42 deletions(-)
diff --git a/include/efi_loader.h b/include/efi_loader.h index b9226208f5..8d8a6649b5 100644 --- a/include/efi_loader.h +++ b/include/efi_loader.h @@ -803,6 +803,8 @@ efi_status_t efi_image_region_add(struct efi_image_regions *regs, int nocheck);
void efi_sigstore_free(struct efi_signature_store *sigstore); +struct efi_signature_store *efi_build_signature_store(void *sig_list, + efi_uintn_t size); struct efi_signature_store *efi_sigstore_parse_sigdb(u16 *name);
bool efi_secure_boot_enabled(void); diff --git a/lib/efi_loader/efi_signature.c b/lib/efi_loader/efi_signature.c index 9ab071b611..87525bdc80 100644 --- a/lib/efi_loader/efi_signature.c +++ b/lib/efi_loader/efi_signature.c @@ -736,6 +736,63 @@ err: return NULL; }
+/** + * efi_sigstore_parse_sigdb - parse the signature list and populate + * the signature store + * + * @sig_list: Pointer to the signature list + * @size: Size of the signature list + * + * Parse the efi signature list and instantiate a signature store + * structure. + * + * Return: Pointer to signature store on success, NULL on error + */ +struct efi_signature_store *efi_build_signature_store(void *sig_list, + efi_uintn_t size) +{ + struct efi_signature_list *esl; + struct efi_signature_store *sigstore = NULL, *siglist; + + esl = sig_list; + while (size > 0) { + /* List must exist if there is remaining data. */ + if (size < sizeof(*esl)) { + EFI_PRINT("Signature list in wrong format\n"); + goto err; + } + + if (size < esl->signature_list_size) { + EFI_PRINT("Signature list in wrong format\n"); + goto err; + } + + /* Parse a single siglist. */ + siglist = efi_sigstore_parse_siglist(esl); + if (!siglist) { + EFI_PRINT("Parsing of signature list of failed\n"); + goto err; + } + + /* Append siglist */ + siglist->next = sigstore; + sigstore = siglist; + + /* Next */ + size -= esl->signature_list_size; + esl = (void *)esl + esl->signature_list_size; + } + free(sig_list); + + return sigstore; + +err: + efi_sigstore_free(sigstore); + free(sig_list); + + return NULL; +} + /** * efi_sigstore_parse_sigdb - parse a signature database variable * @name: Variable's name @@ -747,8 +804,7 @@ err: */ struct efi_signature_store *efi_sigstore_parse_sigdb(u16 *name) { - struct efi_signature_store *sigstore = NULL, *siglist; - struct efi_signature_list *esl; + struct efi_signature_store *sigstore = NULL; const efi_guid_t *vendor; void *db; efi_uintn_t db_size; @@ -784,47 +840,10 @@ struct efi_signature_store *efi_sigstore_parse_sigdb(u16 *name) ret = EFI_CALL(efi_get_variable(name, vendor, NULL, &db_size, db)); if (ret != EFI_SUCCESS) { EFI_PRINT("Getting variable, %ls, failed\n", name); - goto err; - } - - /* Parse siglist list */ - esl = db; - while (db_size > 0) { - /* List must exist if there is remaining data. */ - if (db_size < sizeof(*esl)) { - EFI_PRINT("variable, %ls, in wrong format\n", name); - goto err; - } - - if (db_size < esl->signature_list_size) { - EFI_PRINT("variable, %ls, in wrong format\n", name); - goto err; - } - - /* Parse a single siglist. */ - siglist = efi_sigstore_parse_siglist(esl); - if (!siglist) { - EFI_PRINT("Parsing signature list of %ls failed\n", - name); - goto err; - } - - /* Append siglist */ - siglist->next = sigstore; - sigstore = siglist; - - /* Next */ - db_size -= esl->signature_list_size; - esl = (void *)esl + esl->signature_list_size; + free(db); + return NULL; } - free(db); - - return sigstore;
-err: - efi_sigstore_free(sigstore); - free(db); - - return NULL; + return efi_build_signature_store(db, db_size); } #endif /* CONFIG_EFI_SECURE_BOOT */

Add support for authenticating uefi capsules. Most of the signature verification functionality is shared with the uefi secure boot feature.
The root certificate containing the public key used for the signature verification is stored as part of the device tree blob. The root certificate is stored as an efi signature list(esl) file -- this file contains the x509 certificate which is the root certificate.
Signed-off-by: Sughosh Ganu sughosh.ganu@linaro.org --- board/emulation/qemu-arm/qemu-arm.c | 35 ++++++++ include/efi_api.h | 18 ++++ include/efi_loader.h | 6 ++ lib/efi_loader/Kconfig | 17 ++++ lib/efi_loader/efi_capsule.c | 122 ++++++++++++++++++++++++++++ lib/efi_loader/efi_signature.c | 4 +- 6 files changed, 200 insertions(+), 2 deletions(-)
diff --git a/board/emulation/qemu-arm/qemu-arm.c b/board/emulation/qemu-arm/qemu-arm.c index 8cad54c76f..5794d4c669 100644 --- a/board/emulation/qemu-arm/qemu-arm.c +++ b/board/emulation/qemu-arm/qemu-arm.c @@ -208,6 +208,41 @@ void flash_write32(u32 value, void *addr)
#define DFU_ALT_BUF_LEN SZ_1K
+int efi_get_public_key_data(void **pkey, efi_uintn_t *pkey_len) +{ + const void *fdt_blob = gd->fdt_blob; + const void *blob; + const char *cnode_name = "capsule-key"; + const char *snode_name = "signature"; + int sig_node; + int len; + + sig_node = fdt_subnode_offset(fdt_blob, 0, snode_name); + if (sig_node < 0) { + EFI_PRINT("Unable to get signature node offset\n"); + return -FDT_ERR_NOTFOUND; + } + + blob = fdt_getprop(fdt_blob, sig_node, cnode_name, &len); + + if (!blob || len < 0) { + EFI_PRINT("Unable to get capsule-key value\n"); + *pkey = NULL; + *pkey_len = 0; + return -FDT_ERR_NOTFOUND; + } + + *pkey = (void *)blob; + *pkey_len = len; + + return 0; +} + +bool efi_capsule_auth_enabled(void) +{ + return env_get("capsule_authentication_enabled") != NULL ? true : false; +} + static void board_get_alt_info(struct mtd_info *mtd, char *buf) { struct mtd_info *part; diff --git a/include/efi_api.h b/include/efi_api.h index c7038f863a..bcf6de6629 100644 --- a/include/efi_api.h +++ b/include/efi_api.h @@ -1808,6 +1808,24 @@ struct efi_variable_authentication_2 { struct win_certificate_uefi_guid auth_info; } __attribute__((__packed__));
+/** + * efi_firmware_image_authentication - Capsule authentication method + * descriptor + * + * This structure describes an authentication information for + * a capsule with IMAGE_ATTRIBUTE_AUTHENTICATION_REQUIRED set + * and should be included as part of the capsule. + * Only EFI_CERT_TYPE_PKCS7_GUID is accepted. + * + * @monotonic_count: Count to prevent replay + * @auth_info: Authentication info + */ +struct efi_firmware_image_authentication { + uint64_t monotonic_count; + struct win_certificate_uefi_guid auth_info; +} __attribute__((__packed__)); + + /** * efi_signature_data - A format of signature * diff --git a/include/efi_loader.h b/include/efi_loader.h index 8d8a6649b5..d311317fc1 100644 --- a/include/efi_loader.h +++ b/include/efi_loader.h @@ -809,6 +809,8 @@ struct efi_signature_store *efi_sigstore_parse_sigdb(u16 *name);
bool efi_secure_boot_enabled(void);
+bool efi_capsule_auth_enabled(void); + bool efi_image_parse(void *efi, size_t len, struct efi_image_regions **regp, WIN_CERTIFICATE **auth, size_t *auth_len);
@@ -836,6 +838,10 @@ efi_status_t EFIAPI efi_query_capsule_caps( u64 *maximum_capsule_size, u32 *reset_type);
+efi_status_t efi_capsule_authenticate(const void *capsule, + efi_uintn_t capsule_size, + void **image, efi_uintn_t *image_size); + #define EFI_CAPSULE_DIR L"\EFI\UpdateCapsule\"
/* Hook at initialization */ diff --git a/lib/efi_loader/Kconfig b/lib/efi_loader/Kconfig index 55e4787e32..597ad6b2a7 100644 --- a/lib/efi_loader/Kconfig +++ b/lib/efi_loader/Kconfig @@ -145,6 +145,23 @@ config EFI_CAPSULE_FMP_HEADER Select this option if the capsule is built using the scripts in edk2.
+config EFI_CAPSULE_AUTHENTICATE + bool "Update Capsule authentication" + depends on EFI_CAPSULE_FIRMWARE + depends on EFI_CAPSULE_ON_DISK + depends on EFI_CAPSULE_FIRMWARE_MANAGEMENT + select SHA256 + select RSA + select RSA_VERIFY + select RSA_VERIFY_WITH_PKEY + select X509_CERTIFICATE_PARSER + select PKCS7_MESSAGE_PARSER + select PKCS7_VERIFY + default n + help + Select this option if you want to enable capsule + authentication + config EFI_CAPSULE_FIRMWARE_FIT bool "FMP driver for FIT image" depends on EFI_CAPSULE_FIRMWARE_MANAGEMENT diff --git a/lib/efi_loader/efi_capsule.c b/lib/efi_loader/efi_capsule.c index f385e58378..ad2036fe93 100644 --- a/lib/efi_loader/efi_capsule.c +++ b/lib/efi_loader/efi_capsule.c @@ -14,6 +14,10 @@ #include <mapmem.h> #include <sort.h>
+#include <crypto/pkcs7.h> +#include <crypto/pkcs7_parser.h> +#include <linux/err.h> + const efi_guid_t efi_guid_capsule_report = EFI_CAPSULE_REPORT_GUID; static const efi_guid_t efi_guid_firmware_management_capsule_id = EFI_FIRMWARE_MANAGEMENT_CAPSULE_ID_GUID; @@ -191,6 +195,124 @@ skip: return NULL; }
+#if defined(CONFIG_EFI_CAPSULE_AUTHENTICATE) + +const efi_guid_t efi_guid_capsule_root_cert_guid = + EFI_FIRMWARE_MANAGEMENT_CAPSULE_ID_GUID; + +__weak int efi_get_public_key_data(void **pkey, efi_uintn_t *pkey_len) +{ + /* The platform is supposed to provide + * a method for getting the public key + * stored in the form of efi signature + * list + */ + return 0; +} + +efi_status_t efi_capsule_authenticate(const void *capsule, efi_uintn_t capsule_size, + void **image, efi_uintn_t *image_size) +{ + u8 *buf; + int ret; + void *fdt_pkey, *pkey; + efi_uintn_t pkey_len; + uint64_t monotonic_count; + struct efi_signature_store *truststore; + struct pkcs7_message *capsule_sig; + struct efi_image_regions *regs; + struct efi_firmware_image_authentication *auth_hdr; + efi_status_t status; + + status = EFI_SECURITY_VIOLATION; + capsule_sig = NULL; + truststore = NULL; + regs = NULL; + + /* Sanity checks */ + if (capsule == NULL || capsule_size == 0) + goto out; + + auth_hdr = (struct efi_firmware_image_authentication *)capsule; + if (capsule_size < sizeof(*auth_hdr)) + goto out; + + if (auth_hdr->auth_info.hdr.dwLength <= + offsetof(struct win_certificate_uefi_guid, cert_data)) + goto out; + + if (guidcmp(&auth_hdr->auth_info.cert_type, &efi_guid_cert_type_pkcs7)) + goto out; + + *image = (uint8_t *)capsule + sizeof(auth_hdr->monotonic_count) + + auth_hdr->auth_info.hdr.dwLength; + *image_size = capsule_size - auth_hdr->auth_info.hdr.dwLength - + sizeof(auth_hdr->monotonic_count); + memcpy(&monotonic_count, &auth_hdr->monotonic_count, + sizeof(monotonic_count)); + + /* data to be digested */ + regs = calloc(sizeof(*regs) + sizeof(struct image_region) * 2, 1); + if (!regs) + goto out; + + regs->max = 2; + efi_image_region_add(regs, (uint8_t *)*image, + (uint8_t *)*image + *image_size, 1); + + efi_image_region_add(regs, (uint8_t *)&monotonic_count, + (uint8_t *)&monotonic_count + sizeof(monotonic_count), + 1); + + capsule_sig = efi_parse_pkcs7_header(auth_hdr->auth_info.cert_data, + auth_hdr->auth_info.hdr.dwLength + - sizeof(auth_hdr->auth_info), + &buf); + if (IS_ERR(capsule_sig)) { + debug("Parsing variable's pkcs7 header failed\n"); + capsule_sig = NULL; + goto out; + } + + ret = efi_get_public_key_data(&fdt_pkey, &pkey_len); + if (ret < 0) + goto out; + + pkey = malloc(pkey_len); + if (!pkey) + goto out; + + memcpy(pkey, fdt_pkey, pkey_len); + truststore = efi_build_signature_store(pkey, pkey_len); + if (!truststore) + goto out; + + /* verify signature */ + if (efi_signature_verify(regs, capsule_sig, truststore, NULL)) { + debug("Verified\n"); + } else { + debug("Verifying variable's signature failed\n"); + goto out; + } + + status = EFI_SUCCESS; + +out: + efi_sigstore_free(truststore); + pkcs7_free_message(capsule_sig); + free(regs); + + return status; +} +#else +efi_status_t efi_capsule_authenticate(const void *capsule, efi_uintn_t capsule_size, + void **image, efi_uintn_t *image_size) +{ + return EFI_UNSUPPORTED; +} +#endif /* CONFIG_EFI_CAPSULE_AUTHENTICATE */ + + /** * efi_capsule_update_firmware - update firmware from capsule * @capsule_data: Capsule diff --git a/lib/efi_loader/efi_signature.c b/lib/efi_loader/efi_signature.c index 87525bdc80..c7ec275414 100644 --- a/lib/efi_loader/efi_signature.c +++ b/lib/efi_loader/efi_signature.c @@ -26,7 +26,7 @@ const efi_guid_t efi_guid_cert_x509 = EFI_CERT_X509_GUID; const efi_guid_t efi_guid_cert_x509_sha256 = EFI_CERT_X509_SHA256_GUID; const efi_guid_t efi_guid_cert_type_pkcs7 = EFI_CERT_TYPE_PKCS7_GUID;
-#ifdef CONFIG_EFI_SECURE_BOOT +#if defined(CONFIG_EFI_SECURE_BOOT) || defined(CONFIG_EFI_CAPSULE_AUTHENTICATE) static u8 pkcs7_hdr[] = { /* SEQUENCE */ 0x30, 0x82, 0x05, 0xc7, @@ -846,4 +846,4 @@ struct efi_signature_store *efi_sigstore_parse_sigdb(u16 *name)
return efi_build_signature_store(db, db_size); } -#endif /* CONFIG_EFI_SECURE_BOOT */ +#endif /* CONFIG_EFI_SECURE_BOOT || CONFIG_EFI_CAPSULE_AUTHENTICATE */

Add support for enabling uefi capsule authentication. This feature is enabled by setting the environment variable "capsule_authentication_enabled".
The following configs are needed for enabling uefi capsule update and capsule authentication features on the platform.
CONFIG_EFI_HAVE_CAPSULE_SUPPORT=y CONFIG_EFI_CAPSULE_ON_DISK=y CONFIG_EFI_CAPSULE_FIRMWARE_MANAGEMENT=y CONFIG_EFI_CAPSULE_FIRMWARE=y CONFIG_EFI_CAPSULE_FIRMWARE_RAW=y CONFIG_EFI_CAPSULE_AUTHENTICATE=y
Signed-off-by: Sughosh Ganu sughosh.ganu@linaro.org --- lib/efi_loader/efi_firmware.c | 37 ++++++++++++++++++++++++++++++++++- 1 file changed, 36 insertions(+), 1 deletion(-)
diff --git a/lib/efi_loader/efi_firmware.c b/lib/efi_loader/efi_firmware.c index 6c97604d8b..5e17b2ab5a 100644 --- a/lib/efi_loader/efi_firmware.c +++ b/lib/efi_loader/efi_firmware.c @@ -162,9 +162,16 @@ static efi_status_t efi_get_dfu_info( image_info[i].version_name = NULL; /* not supported */ image_info[i].size = 0; image_info[i].attributes_supported = - IMAGE_ATTRIBUTE_IMAGE_UPDATABLE; + IMAGE_ATTRIBUTE_IMAGE_UPDATABLE | + IMAGE_ATTRIBUTE_AUTHENTICATION_REQUIRED; image_info[i].attributes_setting = IMAGE_ATTRIBUTE_IMAGE_UPDATABLE; + + /* Check if the capsule authentication is enabled */ + if (env_get("capsule_authentication_enabled")) + image_info[0].attributes_setting |= + IMAGE_ATTRIBUTE_AUTHENTICATION_REQUIRED; + image_info[i].lowest_supported_image_version = 0; image_info[i].last_attempt_version = 0; image_info[i].last_attempt_status = LAST_ATTEMPT_STATUS_SUCCESS; @@ -379,12 +386,40 @@ efi_status_t EFIAPI efi_firmware_raw_set_image( efi_status_t (*progress)(efi_uintn_t completion), u16 **abort_reason) { + void *capsule_payload; + efi_status_t status; + efi_uintn_t capsule_payload_size; + EFI_ENTRY("%p %d %p %ld %p %p %p\n", this, image_index, image, image_size, vendor_code, progress, abort_reason);
if (!image) return EFI_EXIT(EFI_INVALID_PARAMETER);
+ /* Authenticate the capsule if authentication enabled */ + if (IS_ENABLED(CONFIG_EFI_CAPSULE_AUTHENTICATE) && + env_get("capsule_authentication_enabled")) { + capsule_payload = NULL; + capsule_payload_size = 0; + status = efi_capsule_authenticate(image, image_size, + &capsule_payload, + &capsule_payload_size); + + if (status == EFI_SECURITY_VIOLATION) { + printf("Capsule authentication check failed. Aborting update\n"); + return EFI_EXIT(status); + } else if (status != EFI_SUCCESS) { + return EFI_EXIT(status); + } + + debug("Capsule authentication successfull\n"); + image = capsule_payload; + image_size = capsule_payload_size; + } else { + debug("Capsule authentication disabled. "); + debug("Updating capsule without authenticating.\n"); + } + if (CONFIG_IS_ENABLED(EFI_CAPSULE_FMP_HEADER)) { /* * When building the capsule with the scripts in

On 11/26/20 7:41 PM, Sughosh Ganu wrote:
Add support for enabling uefi capsule authentication. This feature is enabled by setting the environment variable "capsule_authentication_enabled".
The following configs are needed for enabling uefi capsule update and capsule authentication features on the platform.
CONFIG_EFI_HAVE_CAPSULE_SUPPORT=y CONFIG_EFI_CAPSULE_ON_DISK=y CONFIG_EFI_CAPSULE_FIRMWARE_MANAGEMENT=y CONFIG_EFI_CAPSULE_FIRMWARE=y CONFIG_EFI_CAPSULE_FIRMWARE_RAW=y CONFIG_EFI_CAPSULE_AUTHENTICATE=y
Dear Takahiro, dear Sughosh,
could you, please, provide a documentation for capsule updates in /doc/uefi.
Best regards
Heinrich
Signed-off-by: Sughosh Ganu sughosh.ganu@linaro.org
lib/efi_loader/efi_firmware.c | 37 ++++++++++++++++++++++++++++++++++- 1 file changed, 36 insertions(+), 1 deletion(-)
diff --git a/lib/efi_loader/efi_firmware.c b/lib/efi_loader/efi_firmware.c index 6c97604d8b..5e17b2ab5a 100644 --- a/lib/efi_loader/efi_firmware.c +++ b/lib/efi_loader/efi_firmware.c @@ -162,9 +162,16 @@ static efi_status_t efi_get_dfu_info( image_info[i].version_name = NULL; /* not supported */ image_info[i].size = 0; image_info[i].attributes_supported =
IMAGE_ATTRIBUTE_IMAGE_UPDATABLE;
IMAGE_ATTRIBUTE_IMAGE_UPDATABLE |
image_info[i].attributes_setting = IMAGE_ATTRIBUTE_IMAGE_UPDATABLE;IMAGE_ATTRIBUTE_AUTHENTICATION_REQUIRED;
/* Check if the capsule authentication is enabled */
if (env_get("capsule_authentication_enabled"))
image_info[0].attributes_setting |=
IMAGE_ATTRIBUTE_AUTHENTICATION_REQUIRED;
- image_info[i].lowest_supported_image_version = 0; image_info[i].last_attempt_version = 0; image_info[i].last_attempt_status = LAST_ATTEMPT_STATUS_SUCCESS;
@@ -379,12 +386,40 @@ efi_status_t EFIAPI efi_firmware_raw_set_image( efi_status_t (*progress)(efi_uintn_t completion), u16 **abort_reason) {
void *capsule_payload;
efi_status_t status;
efi_uintn_t capsule_payload_size;
EFI_ENTRY("%p %d %p %ld %p %p %p\n", this, image_index, image, image_size, vendor_code, progress, abort_reason);
if (!image) return EFI_EXIT(EFI_INVALID_PARAMETER);
/* Authenticate the capsule if authentication enabled */
if (IS_ENABLED(CONFIG_EFI_CAPSULE_AUTHENTICATE) &&
env_get("capsule_authentication_enabled")) {
capsule_payload = NULL;
capsule_payload_size = 0;
status = efi_capsule_authenticate(image, image_size,
&capsule_payload,
&capsule_payload_size);
if (status == EFI_SECURITY_VIOLATION) {
printf("Capsule authentication check failed. Aborting update\n");
return EFI_EXIT(status);
} else if (status != EFI_SUCCESS) {
return EFI_EXIT(status);
}
debug("Capsule authentication successfull\n");
image = capsule_payload;
image_size = capsule_payload_size;
} else {
debug("Capsule authentication disabled. ");
debug("Updating capsule without authenticating.\n");
}
if (CONFIG_IS_ENABLED(EFI_CAPSULE_FMP_HEADER)) { /*
- When building the capsule with the scripts in

Add a efidebug subcommand to initiate a firmware update using the efi firmware management protocol(fmp) set_image routine.
The firmware update can be initiated through
'efidebug capsule disk-update'
This would locate the efi capsule file on the efi system partition, and call the platform's set_image fmp routine to initiate the firmware update.
Signed-off-by: Sughosh Ganu sughosh.ganu@linaro.org --- cmd/efidebug.c | 14 ++++++++++++++ 1 file changed, 14 insertions(+)
diff --git a/cmd/efidebug.c b/cmd/efidebug.c index 7d327c8268..b8fe9ad6d5 100644 --- a/cmd/efidebug.c +++ b/cmd/efidebug.c @@ -79,6 +79,16 @@ static int do_efi_capsule_update(struct cmd_tbl *cmdtp, int flag, return CMD_RET_SUCCESS; }
+static int do_efi_capsule_on_disk_update(struct cmd_tbl *cmdtp, int flag, + int argc, char * const argv[]) +{ + efi_status_t ret; + + ret = efi_launch_capsules(); + + return ret == EFI_SUCCESS ? CMD_RET_SUCCESS : CMD_RET_FAILURE; +} + /** * do_efi_capsule_show() - show capsule information * @@ -207,6 +217,8 @@ static struct cmd_tbl cmd_efidebug_capsule_sub[] = { "", ""), U_BOOT_CMD_MKENT(show, CONFIG_SYS_MAXARGS, 1, do_efi_capsule_show, "", ""), + U_BOOT_CMD_MKENT(disk-update, 0, 0, do_efi_capsule_on_disk_update, + "", ""), U_BOOT_CMD_MKENT(result, CONFIG_SYS_MAXARGS, 1, do_efi_capsule_res, "", ""), }; @@ -1540,6 +1552,8 @@ static char efidebug_help_text[] = #ifdef CONFIG_EFI_HAVE_CAPSULE_SUPPORT "efidebug capsule update [-v] <capsule address>\n" " - process a capsule\n" + "efidebug capsule disk-update\n" + " - update a capsule from disk\n" "efidebug capsule show <capsule address>\n" " - show capsule information\n" "efidebug capsule result [<capsule result var>]\n"

Add documentation highlighting the steps for using the uefi capsule update feature for updating the u-boot firmware image.
Signed-off-by: Sughosh Ganu sughosh.ganu@linaro.org --- doc/board/emulation/qemu-arm.rst | 157 +++++++++++++++++++++++++++++++ 1 file changed, 157 insertions(+)
diff --git a/doc/board/emulation/qemu-arm.rst b/doc/board/emulation/qemu-arm.rst index 8d7fda10f1..3978c13269 100644 --- a/doc/board/emulation/qemu-arm.rst +++ b/doc/board/emulation/qemu-arm.rst @@ -90,3 +90,160 @@ The debug UART on the ARM virt board uses these settings:: CONFIG_DEBUG_UART_PL010=y CONFIG_DEBUG_UART_BASE=0x9000000 CONFIG_DEBUG_UART_CLOCK=0 + +Enabling Uefi Capsule Update feature +------------------------------------ + +Support has been added for the uefi capsule update feature which +enables updating the u-boot image using the uefi firmware management +protocol (fmp). The capsules are not passed to the firmware through +the UpdateCapsule runtime service. Instead, capsule-on-disk +functionality is used for fetching the capsule from the EFI System +Partition (ESP). + +Currently, support has been added for updating the u-boot binary as a +raw image when the platform is booted in non-secure mode, i.e with +CONFIG_TFABOOT disabled. For this configuration, the qemu platform +needs to be booted with 'secure=off'. The u-boot binary placed on the +first bank of the Nor Flash at offset 0x0. The u-boot environment is +placed on the second Nor Flash bank at offset 0x4000000. + +The capsule update feature is enabled with the following configs:: + + CONFIG_MTD=y + CONFIG_FLASH_CFI_MTD=y + CONFIG_CMD_MTDPARTS=y + CONFIG_CMD_DFU=y + CONFIG_DFU_MTD=y + CONFIG_EFI_CAPSULE_ON_DISK=y + CONFIG_EFI_CAPSULE_FIRMWARE_MANAGEMENT=y + CONFIG_EFI_CAPSULE_FIRMWARE=y + CONFIG_EFI_CAPSULE_FIRMWARE_RAW=y + CONFIG_EFI_CAPSULE_FMP_HEADER=y + +In addition, the following config needs to be disabled:: + CONFIG_TFABOOT + +The capsule file can be generated by using the GenerateCapsule.py +script in edk2:: + + $ ./BaseTools/BinWrappers/PosixLike/GenerateCapsule -e -o \ + <capsule_file_name> --fw-version <val> --lsv <val> --guid \ + e2bb9c06-70e9-4b14-97a3-5a7913176e3f --verbose --update-image-index \ + <val> --verbose <u-boot.bin> + +If the above edk2 script is being used for generating the capsule, the +following additional config needs to be enabled:: + CONFIG_EFI_CAPSULE_FMP_HEADER=y + +As per the uefi specification, the capsule file needs to be placed on +the EFI System Partition, under the EFI/UpdateCapsule/ directory. The +EFI System Partition can be a virtio-blk-device. + +Before initiating the firmware update, the efi variables BootNext, +BootXXXX and OsIndications need to be set. The BootXXXX variable needs +to be pointing to the EFI System Partition which contains the capsule +file. The BootNext, BootXXXX and OsIndications variables can be set +using the following commands:: + + => efidebug boot add 0 Boot0000 virtio 0:1 <capsule_file_name> + => efidebug boot next 0 + => setenv -e -nv -bs -rt -v OsIndications =0x04 + => saveenv + +Finally, the capsule update can be initiated with the following +command:: + + => efidebug capsule disk-update + +The updated u-boot image will be booted on subsequent boot. + +Enabling Capsule Authentication +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ + +The uefi specification defines a way of authenticating the capsule to +be updated by verifying the capsule signature. The capsule signature +is computed and prepended to the capsule payload at the time of +capsule generation. This signature is then verified by using the +public key stored as part of the X509 certificate. This certificate is +in the form of an efi signature list (esl) file, which is embedded as +part of the platform's device tree blob using the mkeficapsule +utility. + +The capsule authentication feature can be enabled through the +following config, in addition to the configs listed above for capsule +update:: + + CONFIG_EFI_CAPSULE_AUTHENTICATE=y + +The public key esl file can be embedded in the dtb with the following +command:: + ./tools/mkeficapsule -K <pub_key.esl> -D <dtb> + +Running the above command results in the creation of a 'signature' +node in the dtb, under which the public key is stored as a +'capsule-key' property. + +Once the esl file has been embedded as part of the dtb, the platform +needs to be be booted with this dtb. This can be done by disabling the +CONFIG_OF_BOARD option, and then, passing the above dtb file to the +u-boot build. + +The capsule update with authentication can be enabled on the platform +with the following steps + +1. Install utility commands on your host + * openssl + * efitools + +2. Create signing keys and certificate files on your host:: + + $ openssl req -x509 -sha256 -newkey rsa:2048 -subj /CN=CRT/ \ + -keyout CRT.key -out CRT.crt -nodes -days 365 + $ cert-to-efi-sig-list CRT.crt CRT.esl + + $ openssl x509 -in CRT.crt -out CRT.cer -outform DER + $ openssl x509 -inform DER -in CRT.cer -outform PEM -out CRT.pub.pem + + $ openssl pkcs12 -export -out CRT.pfx -inkey CRT.key -in CRT.crt + $ openssl pkcs12 -in CRT.pfx -nodes -out CRT.pem + +3. Store the esl file generated above as part of the dtb:: + + $ ./tools/mkeficapsule -K <pub_key.esl> -D <dtb> + +4. The capsule file can be generated by using the GenerateCapsule.py + script in edk2:: + + $ ./BaseTools/BinWrappers/PosixLike/GenerateCapsule -e -o \ + <capsule_file_name> --monotonic-count <val> --fw-version \ + <val> --lsv <val> --guid \ + e2bb9c06-70e9-4b14-97a3-5a7913176e3f --verbose \ + --update-image-index <val> --signer-private-cert \ + /path/to/CRT.pem --trusted-public-cert \ + /path/to/CRT.pub.pem --other-public-cert /path/to/CRT.pub.pem \ + <u-boot.bin> + +Once the capsule has been generated, use the same instructions as +mentioned above for placing the capsule on the EFI System Partition + +5. Building u-boot with the following steps:: + + $ make qemu_arm64_defconfig + $ make menuconfig + Disable CONFIG_OF_BOARD and CONFIG_TFABOOT + $ make EXT_DTB=<dtb> all + +6. Enable capsule authentication by setting the following env + variable:: + + => setenv capsule_authentication_enabled 1 + => saveenv + +Setting the environment variable capsule_authentication_enabled +enables the capsule authentication. + +Once the capsule has been placed on the EFI System Partition and the +above env variable has been set, along with the BootXXXX and the +BootNext variables, the capsule update can be initiated +using the same command as that shown above.

On 11/26/20 7:41 PM, Sughosh Ganu wrote:
Add documentation highlighting the steps for using the uefi capsule update feature for updating the u-boot firmware image.
Signed-off-by: Sughosh Ganu sughosh.ganu@linaro.org
doc/board/emulation/qemu-arm.rst | 157 +++++++++++++++++++++++++++++++
Thank you for carefully documenting your enhancement.
Unfortunately this does not build with 'make htmldocs'. (You will need python3-sphinx with a version < 3 for building due to incompatible changes in python3-sphinx.)
Warning, treated as error: doc/board/emulation/qemu-arm.rst:137:Unexpected indentation. make[1]: *** [doc/Makefile:69: htmldocs] Error 2 make: *** [Makefile:2167: htmldocs] Error 2
Please, run your series through Travis CI before resubmitting.
Takahiro's patches have been added to origin/next. So this is what your series should be based on until the next is merged into master in January.
I am missing a documentation for mkeficapsule. Could you, please, try to set one up together with Takahiro. I guess we should create a new directory doc/tools/.
Best regards
Heinrich
1 file changed, 157 insertions(+)
diff --git a/doc/board/emulation/qemu-arm.rst b/doc/board/emulation/qemu-arm.rst index 8d7fda10f1..3978c13269 100644 --- a/doc/board/emulation/qemu-arm.rst +++ b/doc/board/emulation/qemu-arm.rst @@ -90,3 +90,160 @@ The debug UART on the ARM virt board uses these settings:: CONFIG_DEBUG_UART_PL010=y CONFIG_DEBUG_UART_BASE=0x9000000 CONFIG_DEBUG_UART_CLOCK=0
+Enabling Uefi Capsule Update feature +------------------------------------
+Support has been added for the uefi capsule update feature which +enables updating the u-boot image using the uefi firmware management +protocol (fmp). The capsules are not passed to the firmware through +the UpdateCapsule runtime service. Instead, capsule-on-disk +functionality is used for fetching the capsule from the EFI System +Partition (ESP).
+Currently, support has been added for updating the u-boot binary as a +raw image when the platform is booted in non-secure mode, i.e with +CONFIG_TFABOOT disabled. For this configuration, the qemu platform +needs to be booted with 'secure=off'. The u-boot binary placed on the +first bank of the Nor Flash at offset 0x0. The u-boot environment is +placed on the second Nor Flash bank at offset 0x4000000.
+The capsule update feature is enabled with the following configs::
- CONFIG_MTD=y
- CONFIG_FLASH_CFI_MTD=y
- CONFIG_CMD_MTDPARTS=y
- CONFIG_CMD_DFU=y
- CONFIG_DFU_MTD=y
- CONFIG_EFI_CAPSULE_ON_DISK=y
- CONFIG_EFI_CAPSULE_FIRMWARE_MANAGEMENT=y
- CONFIG_EFI_CAPSULE_FIRMWARE=y
- CONFIG_EFI_CAPSULE_FIRMWARE_RAW=y
- CONFIG_EFI_CAPSULE_FMP_HEADER=y
+In addition, the following config needs to be disabled::
- CONFIG_TFABOOT
+The capsule file can be generated by using the GenerateCapsule.py +script in edk2::
- $ ./BaseTools/BinWrappers/PosixLike/GenerateCapsule -e -o \
- <capsule_file_name> --fw-version <val> --lsv <val> --guid \
- e2bb9c06-70e9-4b14-97a3-5a7913176e3f --verbose --update-image-index \
- <val> --verbose <u-boot.bin>
+If the above edk2 script is being used for generating the capsule, the +following additional config needs to be enabled::
- CONFIG_EFI_CAPSULE_FMP_HEADER=y
+As per the uefi specification, the capsule file needs to be placed on +the EFI System Partition, under the EFI/UpdateCapsule/ directory. The +EFI System Partition can be a virtio-blk-device.
+Before initiating the firmware update, the efi variables BootNext, +BootXXXX and OsIndications need to be set. The BootXXXX variable needs +to be pointing to the EFI System Partition which contains the capsule +file. The BootNext, BootXXXX and OsIndications variables can be set +using the following commands::
- => efidebug boot add 0 Boot0000 virtio 0:1 <capsule_file_name>
- => efidebug boot next 0
- => setenv -e -nv -bs -rt -v OsIndications =0x04
- => saveenv
+Finally, the capsule update can be initiated with the following +command::
- => efidebug capsule disk-update
+The updated u-boot image will be booted on subsequent boot.
+Enabling Capsule Authentication +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
+The uefi specification defines a way of authenticating the capsule to +be updated by verifying the capsule signature. The capsule signature +is computed and prepended to the capsule payload at the time of +capsule generation. This signature is then verified by using the +public key stored as part of the X509 certificate. This certificate is +in the form of an efi signature list (esl) file, which is embedded as +part of the platform's device tree blob using the mkeficapsule +utility.
+The capsule authentication feature can be enabled through the +following config, in addition to the configs listed above for capsule +update::
- CONFIG_EFI_CAPSULE_AUTHENTICATE=y
+The public key esl file can be embedded in the dtb with the following +command::
- ./tools/mkeficapsule -K <pub_key.esl> -D <dtb>
+Running the above command results in the creation of a 'signature' +node in the dtb, under which the public key is stored as a +'capsule-key' property.
+Once the esl file has been embedded as part of the dtb, the platform +needs to be be booted with this dtb. This can be done by disabling the +CONFIG_OF_BOARD option, and then, passing the above dtb file to the +u-boot build.
+The capsule update with authentication can be enabled on the platform +with the following steps
+1. Install utility commands on your host
- openssl
- efitools
+2. Create signing keys and certificate files on your host::
$ openssl req -x509 -sha256 -newkey rsa:2048 -subj /CN=CRT/ \
-keyout CRT.key -out CRT.crt -nodes -days 365
$ cert-to-efi-sig-list CRT.crt CRT.esl
$ openssl x509 -in CRT.crt -out CRT.cer -outform DER
$ openssl x509 -inform DER -in CRT.cer -outform PEM -out CRT.pub.pem
$ openssl pkcs12 -export -out CRT.pfx -inkey CRT.key -in CRT.crt
$ openssl pkcs12 -in CRT.pfx -nodes -out CRT.pem
+3. Store the esl file generated above as part of the dtb::
$ ./tools/mkeficapsule -K <pub_key.esl> -D <dtb>
+4. The capsule file can be generated by using the GenerateCapsule.py
- script in edk2::
$ ./BaseTools/BinWrappers/PosixLike/GenerateCapsule -e -o \
- <capsule_file_name> --monotonic-count <val> --fw-version \
- <val> --lsv <val> --guid \
- e2bb9c06-70e9-4b14-97a3-5a7913176e3f --verbose \
- --update-image-index <val> --signer-private-cert \
- /path/to/CRT.pem --trusted-public-cert \
- /path/to/CRT.pub.pem --other-public-cert /path/to/CRT.pub.pem \
- <u-boot.bin>
+Once the capsule has been generated, use the same instructions as +mentioned above for placing the capsule on the EFI System Partition
+5. Building u-boot with the following steps::
$ make qemu_arm64_defconfig
$ make menuconfig
Disable CONFIG_OF_BOARD and CONFIG_TFABOOT
$ make EXT_DTB=<dtb> all
+6. Enable capsule authentication by setting the following env
- variable::
=> setenv capsule_authentication_enabled 1
=> saveenv
+Setting the environment variable capsule_authentication_enabled +enables the capsule authentication.
+Once the capsule has been placed on the EFI System Partition and the +above env variable has been set, along with the BootXXXX and the +BootNext variables, the capsule update can be initiated +using the same command as that shown above.
participants (3)
-
Heinrich Schuchardt
-
Sughosh Ganu
-
Tom Rini