mail archive of the barebox mailing list
 help / color / mirror / Atom feed
From: Ahmad Fatoum <a.fatoum@pengutronix.de>
To: barebox@lists.infradead.org
Cc: has@pengutronix.de, Ahmad Fatoum <a.fatoum@pengutronix.de>
Subject: [PATCH 3/3] block: efi: allow disabling /dev/usbdiskX renaming
Date: Thu, 30 Jun 2022 14:40:35 +0200	[thread overview]
Message-ID: <20220630124035.4019644-3-a.fatoum@pengutronix.de> (raw)
In-Reply-To: <20220630124035.4019644-1-a.fatoum@pengutronix.de>

Some buggy UEFI implementations have been observed to not set
EFI_USB_IO protocol on handles that were instantiated from
USB mass storage. This leads to confusing behavior when barebox
uses /dev/usbdisk for some USB sticks and doesn't for some others.

The proper behavior is not relying on the UEFI and instead check
e.g. GUID of the boot disk image:

  if [ "$bootsource" = usb ]; then
    mydisk = /dev/usbdisk${bootsource_instance}
  else
    mydisk = /dev/disk${bootsource_instance}
  fi

  devlookup -v bootdiskguid $mydisk guid
  if [ "$bootdiskguid" = "my-guid" ]; then
    boot ${mydisk}.1
  else
    boot boochooser
  fi

When going that way, the renaming to usbdisk is just annoying,
so allow disabling it to simplify scripting.

Signed-off-by: Ahmad Fatoum <a.fatoum@pengutronix.de>
---
 drivers/block/Kconfig        | 21 +++++++++++++++++++++
 drivers/block/Makefile       |  2 +-
 drivers/block/efi-block-io.c |  3 ++-
 3 files changed, 24 insertions(+), 2 deletions(-)

diff --git a/drivers/block/Kconfig b/drivers/block/Kconfig
index 68a11438dcf3..625e81a4ec97 100644
--- a/drivers/block/Kconfig
+++ b/drivers/block/Kconfig
@@ -5,3 +5,24 @@ config VIRTIO_BLK
         help
           This is the virtual block driver for virtio.  It can be used with
           QEMU based VMMs (like KVM or Xen).
+
+config EFI_BLK
+	bool "EFI block I/O driver"
+	default y
+	depends on EFI_BOOTUP
+
+config EFI_BLK_SEPARATE_USBDISK
+	bool "rename USB devices to /dev/usbdiskX"
+	default y
+	depends on EFI_BLK
+	help
+	  EFI block devices will be normally called /dev/diskX. Setting this
+	  option will cause block devices instantiated from handles with a
+	  EFI_USB_IO protocol to be called /dev/usbdiskX instead. Note that
+	  some buggy UEFI implementations have been observed to not do this
+	  consistently for all USB mass storage. If you need to absolutely
+	  be sure your boot device is a USB mass storage device and you can't
+	  fix your UEFI, consider disabling this options and setting a GUID
+	  for your disk and checking against it with
+
+	    devlookup -v $bootguid /dev/disk$bootsource_instance guid
diff --git a/drivers/block/Makefile b/drivers/block/Makefile
index c50bdc1d02bd..a4e14a559cf7 100644
--- a/drivers/block/Makefile
+++ b/drivers/block/Makefile
@@ -1,3 +1,3 @@
 # SPDX-License-Identifier: GPL-2.0-only
-obj-$(CONFIG_EFI_BOOTUP) += efi-block-io.o
+obj-$(CONFIG_EFI_BLK) += efi-block-io.o
 obj-$(CONFIG_VIRTIO_BLK) += virtio_blk.o
diff --git a/drivers/block/efi-block-io.c b/drivers/block/efi-block-io.c
index 086afb378ab4..120e4d8f1ad2 100644
--- a/drivers/block/efi-block-io.c
+++ b/drivers/block/efi-block-io.c
@@ -140,7 +140,8 @@ static void efi_bio_print_info(struct device_d *dev)
 
 static bool is_bio_usbdev(struct efi_device *efidev)
 {
-	return efi_device_has_guid(efidev, EFI_USB_IO_PROTOCOL_GUID);
+	return IS_ENABLED(CONFIG_EFI_BLK_SEPARATE_USBDISK) &&
+		efi_device_has_guid(efidev, EFI_USB_IO_PROTOCOL_GUID);
 }
 
 static int efi_bio_probe(struct efi_device *efidev)
-- 
2.30.2




  parent reply	other threads:[~2022-06-30 12:42 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-06-30 12:40 [PATCH 1/3] partitions: efi: register guid device parameter for disk GUID Ahmad Fatoum
2022-06-30 12:40 ` [PATCH 2/3] commands: implement devlookup to find device behind device file Ahmad Fatoum
2022-06-30 15:26   ` Michael Olbrich
2022-07-01  5:24     ` Ahmad Fatoum
2022-07-05  7:45   ` Sascha Hauer
2022-07-05  7:48     ` Ahmad Fatoum
2022-06-30 12:40 ` Ahmad Fatoum [this message]
2022-06-30 13:26   ` [PATCH] fixup! commands: implement new devlookup command Ahmad Fatoum

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20220630124035.4019644-3-a.fatoum@pengutronix.de \
    --to=a.fatoum@pengutronix.de \
    --cc=barebox@lists.infradead.org \
    --cc=has@pengutronix.de \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox