All of lore.kernel.org
 help / color / mirror / Atom feed
From: Ard Biesheuvel <ardb+git@google.com>
To: linux-efi@vger.kernel.org
Cc: Ard Biesheuvel <ardb@kernel.org>, Kees Cook <kees@kernel.org>,
	 "Guilherme G. Piccoli" <gpiccoli@igalia.com>,
	stable@vger.kernel.org
Subject: [PATCH 1/2] efi: pstore: Make pstore_disable parameter read-only again
Date: Fri, 20 Nov 2026 14:18:29 +0100	[thread overview]
Message-ID: <20261120131827.480055-5-ardb+git@google.com> (raw)
In-Reply-To: <20261120131827.480055-4-ardb+git@google.com>

From: Ard Biesheuvel <ardb@kernel.org>

Commit a28655c330ab ("efi: pstore: Allow dynamic initialization based on
module parameter") wired up a kernel_param_ops::set callback for
pstore_disable that invokes efivars_pstore_init() directly.

Because parse_args() runs before module initialization (and before
mm_core_init() during early boot when built-in), calling
efivars_pstore_init() from the parameter setter is fundamentally broken:

 - Loading efi-pstore as a module with pstore_disable=N initializes and
   registers efi_pstore_info during parse_args(), and then runs
   efivars_pstore_init() a second time from do_init_module(), leaking
   the first buffer, failing pstore_register() with -EBUSY, and zeroing
   efi_pstore_info.buf/bufsize while leaving efi_pstore_info registered
   with pstore.
 - If load_module() fails after parse_args() (e.g., due to a subsequent
   unknown parameter), the module is freed without calling module_exit,
   leaving pstore with a dangling backend pointer.
 - Passing pstore_disable=N before record_size=<N> initializes the
   buffer using the default record_size before record_size is parsed.

Module parameters are meant for boot/load-time configuration rather than
runtime driver state control via sysfs. Revert the dynamic parameter
callback changes and tighten the sysfs permissions of pstore_disable from
0644 to 0444 (matching record_size) so it is no longer misleadingly
writable at runtime.

Fixes: a28655c330ab ("efi: pstore: Allow dynamic initialization based on module parameter")
Cc: stable@vger.kernel.org
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
---
 drivers/firmware/efi/efi-pstore.c | 39 +++-----------------
 1 file changed, 6 insertions(+), 33 deletions(-)

diff --git a/drivers/firmware/efi/efi-pstore.c b/drivers/firmware/efi/efi-pstore.c
index a5db3534f0a6..19a7e4fe5b52 100644
--- a/drivers/firmware/efi/efi-pstore.c
+++ b/drivers/firmware/efi/efi-pstore.c
@@ -19,37 +19,10 @@ MODULE_PARM_DESC(record_size, "size of each pstore UEFI var (in bytes, min/defau
 	 EFI_VARIABLE_BOOTSERVICE_ACCESS | \
 	 EFI_VARIABLE_RUNTIME_ACCESS)
 
-static bool pstore_disable = IS_ENABLED(CONFIG_EFI_VARS_PSTORE_DEFAULT_DISABLE);
+static bool efivars_pstore_disable =
+	IS_ENABLED(CONFIG_EFI_VARS_PSTORE_DEFAULT_DISABLE);
 
-static int efivars_pstore_init(void);
-static void efivars_pstore_exit(void);
-
-static int efi_pstore_disable_set(const char *val, const struct kernel_param *kp)
-{
-	int err;
-	bool old_pstore_disable = pstore_disable;
-
-	err = param_set_bool(val, kp);
-	if (err)
-		return err;
-
-	if (old_pstore_disable != pstore_disable) {
-		if (pstore_disable)
-			efivars_pstore_exit();
-		else
-			efivars_pstore_init();
-	}
-
-	return 0;
-}
-
-static const struct kernel_param_ops pstore_disable_ops = {
-	.set	= efi_pstore_disable_set,
-	.get	= param_get_bool,
-};
-
-module_param_cb(pstore_disable, &pstore_disable_ops, &pstore_disable, 0644);
-__MODULE_PARM_TYPE(pstore_disable, "bool");
+module_param_named(pstore_disable, efivars_pstore_disable, bool, 0444);
 
 static int efi_pstore_open(struct pstore_info *psi)
 {
@@ -255,12 +228,12 @@ static struct pstore_info efi_pstore_info = {
 	.erase		= efi_pstore_erase,
 };
 
-static int efivars_pstore_init(void)
+static __init int efivars_pstore_init(void)
 {
 	if (!efivar_supports_writes())
 		return 0;
 
-	if (pstore_disable)
+	if (efivars_pstore_disable)
 		return 0;
 
 	/*
@@ -287,7 +260,7 @@ static int efivars_pstore_init(void)
 	return 0;
 }
 
-static void efivars_pstore_exit(void)
+static __exit void efivars_pstore_exit(void)
 {
 	if (!efi_pstore_info.bufsize)
 		return;
-- 
2.56.1.750.gf70dc682cf-goog


  reply	other threads:[~2026-11-20 13:18 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-11-20 13:18 [PATCH 0/2] efi: Fix pstore runtime en/disable Ard Biesheuvel
2026-11-20 13:18 ` Ard Biesheuvel [this message]
2026-11-20 13:18 ` [PATCH 2/2] efi: pstore: Convert into a platform driver for 'efivars' Ard Biesheuvel
2026-11-20 13:39   ` sashiko-bot

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=20261120131827.480055-5-ardb+git@google.com \
    --to=ardb+git@google.com \
    --cc=ardb@kernel.org \
    --cc=gpiccoli@igalia.com \
    --cc=kees@kernel.org \
    --cc=linux-efi@vger.kernel.org \
    --cc=stable@vger.kernel.org \
    /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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.