On 2024-04-12 14:13, Stefan Berger wrote:
On 4/12/24 04:39, Gary Lin via Grub-devel wrote:
From: Patrick Colp <patrick.c...@oracle.com>
If a protector is specified, but it fails to unlock the disk, fall back
to asking for the passphrase. However, an error was set indicating that
the protector(s) failed. Later code (e.g., LUKS code) fails as
`grub_errno` is now set. Print the existing errors out first, before
Though you are resetting grub_errno now so that LUKS code then does
not fail unnecessarily? It's not quite clear what the above wants to
convey.
Yeah, that's it. If we do nothing with `grub_errno` here, then code will
eventually execute that checks if `grub_errno` is set. It'll see that it
is and fail, even though there was no actual failure, it was just set
before from when the protector failed. I could change it to read
something like:
If `grub_errno` was previously set, print out the error and clear
`grub_errno`.
proceeding with the passphrase.
Signed-off-by: Patrick Colp <patrick.c...@oracle.com>
Signed-off-by: Gary Lin <g...@suse.com>
---
grub-core/disk/cryptodisk.c | 7 ++++++-
1 file changed, 6 insertions(+), 1 deletion(-)
diff --git a/grub-core/disk/cryptodisk.c b/grub-core/disk/cryptodisk.c
index 0ca1a5c4d..0dbf601c4 100644
--- a/grub-core/disk/cryptodisk.c
+++ b/grub-core/disk/cryptodisk.c
@@ -1191,11 +1191,16 @@ grub_cryptodisk_scan_device_real (const char
*name,
source->name, source->partition != NULL ? "," : "",
part != NULL ? part : N_("UNKNOWN"), dev->uuid);
grub_free (part);
- goto error;
}
if (!cargs->key_len)
{
+ if (grub_errno)
+ {
+ grub_print_error ();
+ grub_errno = GRUB_ERR_NONE;
+ }
+
/* Get the passphrase from the user, if no key data. */
askpass = 1;
part = grub_partition_get_name (source->partition);
_______________________________________________
Grub-devel mailing list
Grub-devel@gnu.org
https://lists.gnu.org/mailman/listinfo/grub-devel
