On Sun, Jun 18, 2023 at 08:39:27AM +0200, Heinrich Schuchardt wrote:
On 6/15/23 16:39, Ilias Apalodimas wrote:
efi_uninstall_protocol() calls efi_disconnect_all_drivers() but never checks the return value. Honor that and return an appropriate error if the associated controllers failed to disconnect
Signed-off-by: Ilias Apalodimas ilias.apalodimas@linaro.org
lib/efi_loader/efi_boottime.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-)
diff --git a/lib/efi_loader/efi_boottime.c b/lib/efi_loader/efi_boottime.c index 5006c0e1e4af..68198e6b5ff6 100644 --- a/lib/efi_loader/efi_boottime.c +++ b/lib/efi_loader/efi_boottime.c @@ -1353,7 +1353,11 @@ static efi_status_t efi_uninstall_protocol if (r != EFI_SUCCESS) goto out; /* Disconnect controllers */
- efi_disconnect_all_drivers(efiobj, protocol, NULL);
- r = efi_disconnect_all_drivers(efiobj, protocol, NULL);
- if (r != EFI_SUCCESS) {
r = EFI_DEVICE_ERROR;This return code is not foreseen in this case by the UEFI specification. I only found:
EFI_ACCESS_DENIED: The interface was not removed because the interface is still being used by a driver.
EDK II sets this code in CoreDisconnectControllersUsingProtocolInterface().
Right, I only looked at DisconnectController() returns values... You are right, I'll switch this and the check in the selftests.
Thanks /Ilias
Best regards
Heinrich
goto out;- } /* Close protocol */ list_for_each_entry_safe(item, pos, &handler->open_infos, link) { if (item->info.attributes ==