On Tue, Jun 04, 2019 at 11:38:27PM +0200, Heinrich Schuchardt wrote:
On 6/4/19 8:52 AM, AKASHI Takahiro wrote:
Data cache will be read in from persistent storage after (re)boot to restore UEFI non-volatile variables.
Signed-off-by: AKASHI Takahiro takahiro.akashi@linaro.org
lib/efi_loader/efi_setup.c | 6 ++++++ 1 file changed, 6 insertions(+)
diff --git a/lib/efi_loader/efi_setup.c b/lib/efi_loader/efi_setup.c index 8691d686d29d..45d6aca051f3 100644 --- a/lib/efi_loader/efi_setup.c +++ b/lib/efi_loader/efi_setup.c @@ -8,6 +8,7 @@ #include <common.h> #include <bootm.h> #include <efi_loader.h> +#include <environment.h>
#define OBJ_LIST_NOT_INITIALIZED 1
@@ -102,6 +103,11 @@ efi_status_t efi_init_obj_list(void) /* On ARM switch from EL3 or secure mode to EL2 or non-secure mode */ switch_to_non_secure_mode();
+#ifdef CONFIG_EFI_VARIABLE_USE_ENV
No clue what ENV refers to here as we are not talking about U-Boot environment variables anymore. How about CONFIG_EFI_PERSISTENT_VARIABLES.
It will be trivial once you take a look at "menuconfig."
- /* Load non-volatile variables */
- env_efi_load();
Can't we make env_efi_load() a __weak function which does nothing. If we have a backend, that backend replaces the weak function. That way we restrict the config variables to the Makefile.
This is a discussion. There can be different approaches here, so I would like to deter to a developer who will implement a next backing storage (other than U-Boot env), which is likely to be Standalone MM services for secure boot. Unfortunately I'm not responsible for that(StMM).
I hope that some Linaro engineers may have comments here.
Thanks, -Takahiro Akashi
Regards
Heinrich
+#endif
- /* Define supported languages */ ret = efi_init_platform_lang(); if (ret != EFI_SUCCESS)