All of lore.kernel.org
 help / color / mirror / Atom feed
From: "Masami Hiramatsu (Google)" <mhiramat@kernel.org>
To: Masami Hiramatsu <mhiramat@kernel.org>,
	Andrew Morton <akpm@linux-foundation.org>,
	Christian Brauner <brauner@kernel.org>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>, Jan Kara <jack@suse.cz>,
	Breno Leitao <leitao@debian.org>,
	Ryan Roberts <ryan.roberts@arm.com>,
	Steven Rostedt <rostedt@goodmis.org>,
	linux-kernel@vger.kernel.org, linux-fsdevel@vger.kernel.org,
	linux-trace-kernel@vger.kernel.org
Subject: [PATCH 2/2] bootconfig: Free the initrd memory used by bootconfig
Date: Wed, 18 Nov 2026 08:47:40 +0900	[thread overview]
Message-ID: <179495926058.237695.12800083196053168375.stgit@devnote2> (raw)
In-Reply-To: <179495924003.237695.10973676765152260584.stgit@devnote2>

From: Masami Hiramatsu (Google) <mhiramat@kernel.org>

get_boot_config_from_initrd() removes the bootconfig from the initrd by
decrementing initrd_end in place. However, the whole initrd including
the bootconfig is reserved, e.g. reserve_initrd_mem() reserves
[initrd_start, PAGE_ALIGN(original initrd_end)), and do_populate_rootfs()
frees it with free_initrd_mem(initrd_start, initrd_end). Thus the full
pages in [PAGE_ALIGN(new initrd_end), PAGE_ALIGN(original initrd_end))
are never freed. This leaks up to 8 pages for a 32KB bootconfig.

Save the original initrd_end in initrd_mem_end when trimming the
bootconfig, and restore initrd_end before freeing the initrd memory.
The bootconfig data has already been copied by xbc_init() at that point.
The trimmed initrd_end is kept when the initrd is retained, so that
/sys/firmware/initrd does not expose the bootconfig.

Fixes: de462e5f1071 ("bootconfig: Fix to remove bootconfig data from initrd while boot")
Cc: stable@vger.kernel.org
Signed-off-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
---
 include/linux/initrd.h |    2 ++
 init/initramfs.c       |   10 ++++++++++
 init/main.c            |    8 +++++++-
 3 files changed, 19 insertions(+), 1 deletion(-)

diff --git a/include/linux/initrd.h b/include/linux/initrd.h
index 7e5d26c8136f..3086a462da76 100644
--- a/include/linux/initrd.h
+++ b/include/linux/initrd.h
@@ -17,6 +17,8 @@ extern unsigned long initrd_start, initrd_end;
 extern void free_initrd_mem(unsigned long, unsigned long);
 
 #ifdef CONFIG_BLK_DEV_INITRD
+/* initrd_end before trailing data (e.g. bootconfig) was trimmed off */
+extern unsigned long initrd_mem_end;
 extern void __init reserve_initrd_mem(void);
 extern void wait_for_initramfs(void);
 #else
diff --git a/init/initramfs.c b/init/initramfs.c
index ebddde9c8f0d..11030cc1f35d 100644
--- a/init/initramfs.c
+++ b/init/initramfs.c
@@ -582,6 +582,7 @@ char * __init unpack_to_rootfs(char *buf, unsigned long len)
 }
 
 static int __initdata do_retain_initrd;
+unsigned long initrd_mem_end __initdata;
 
 static int __init retain_initrd_param(char *str)
 {
@@ -750,6 +751,15 @@ static void __init do_populate_rootfs(void *unused, async_cookie_t cookie)
 		security_initramfs_populated();
 	}
 
+	/*
+	 * The bootconfig loader trims the bootconfig off initrd_end. Restore
+	 * it to free the memory used by the bootconfig too, unless the initrd
+	 * is retained (which must not expose the bootconfig).
+	 */
+	if (!do_retain_initrd && initrd_start && initrd_end < initrd_mem_end)
+		initrd_end = initrd_mem_end;
+	initrd_mem_end = 0;
+
 	/*
 	 * If the initrd region is overlapped with crashkernel reserved region,
 	 * free only memory that is not part of crashkernel region.
diff --git a/init/main.c b/init/main.c
index 9471ece3d7d5..f14780345051 100644
--- a/init/main.c
+++ b/init/main.c
@@ -323,7 +323,13 @@ static void * __init get_boot_config_from_initrd(size_t *_size)
 		return NULL;
 	}
 
-	/* Remove bootconfig from initramfs/initrd */
+	/*
+	 * Remove bootconfig from initramfs/initrd. Since the initrd memory
+	 * (including the bootconfig) is freed with initrd_end, save the
+	 * original end to free the whole region later.
+	 */
+	if (!initrd_mem_end)
+		initrd_mem_end = initrd_end;
 	initrd_end = (unsigned long)data;
 	if (_size)
 		*_size = size;


      parent reply	other threads:[~2026-11-17 23:47 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-11-17 23:47 [PATCH 0/2] bootconfig: Fix max size padding and initrd memory leak Masami Hiramatsu (Google)
2026-11-17 23:47 ` [PATCH 1/2] bootconfig: Exclude alignment padding from bootconfig max size Masami Hiramatsu (Google)
2026-11-17 23:47 ` Masami Hiramatsu (Google) [this message]

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=179495926058.237695.12800083196053168375.stgit@devnote2 \
    --to=mhiramat@kernel.org \
    --cc=akpm@linux-foundation.org \
    --cc=brauner@kernel.org \
    --cc=jack@suse.cz \
    --cc=leitao@debian.org \
    --cc=linux-fsdevel@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-trace-kernel@vger.kernel.org \
    --cc=rostedt@goodmis.org \
    --cc=ryan.roberts@arm.com \
    --cc=viro@zeniv.linux.org.uk \
    /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.