From: ArunKumar.Prakash <ArunKumar.Prakash@in.bosch.com>
To: Pantelis Antoniou <pantelis.antoniou@konsulko.com>,
Frank Rowand <frowand.list@gmail.com>,
Rob Herring <robh+dt@kernel.org>,
devicetree@vger.kernel.org, linux-kernel@vger.kernel.org
Subject: [PATCH] of: overlay: Fix for possible kfree() on already freed memory
Date: Tue, 2 Apr 2019 06:53:25 +0200 [thread overview]
Message-ID: <0171a15f-a1fc-bcc6-f10d-1bc0e401eadd@in.bosch.com> (raw)
While applying an overlay, on an error return from
init_overlay_changeset(), of_overlay_apply()
frees memory pointed by the function arguments 'fdt' and 'tree';
free_overlay_changeset() is then called which tries to free memory
at the same location as 'fdt' aka 'ovcs->fdt' and
'tree' aka 'ovcs->overlay_tree'.
This can produce kernel Oops/BUG_ON() assertion.
This is fixed by making sure elements in ovcs are zero/NULL initialized
until _all_ checks for errors pass.
Below kernel OOPS occurred when a device tree overlay
with an invalid target/target-path is applied.
OF: overlay: find target, node: /fragment@2, path '/external-pcb' not found
OF: overlay: init_overlay_changeset() failed, ret = -22
------------[ cut here ]------------
kernel BUG at mm/slub.c:3892!
Internal error: Oops - BUG: 0 [#1] PREEMPT SMP
------------[manual cut]------------------
task: ffff8004c1529c00 task.stack: ffff00001eea8000
PC is at kfree+0xe4/0x238
LR is at free_overlay_changeset+0x8c/0xac
pc : [<ffff0000081d13e0>] lr : [<ffff0000085b29e8>] ps
x0 : ffff7e0012795d20
Process cat (pid: 2586, stack limit = 0xffff00001eea8000)
Call trace:
Exception stack(0xffff00001eeaba20 to 0xffff00001eeabb60)
[<ffff0000081d13e0>] kfree+0xe4/0x238
[<ffff0000085b29e8>] free_overlay_changeset+0x8c/0xac
[<ffff0000085b33a0>] of_overlay_fdt_apply+0x4d4/0x614
[<ffff0000085ac2e8>] create_overlay+0xac/0x104
[<ffff0000085ac3a4>] cfs_overlay_item_dtbo_write+0x64/0x98
[<ffff00000825dda8>] configfs_release_bin_file+0x58/0x98
[<ffff0000081e587c>] __fput+0xe4/0x1a8
[<ffff0000081e5998>] ____fput+0xc/0x14
[<ffff0000080b9da0>] task_work_run+0x88/0xa4
[<ffff0000080a0968>] do_exit+0x428/0x8f4
[<ffff0000080a0ea4>] do_group_exit+0x40/0x98
[<ffff0000080a0f0c>] __wake_up_parent+0x0/0x28
Signed-off-by: Arun Kumar Prakash <ArunKumar.Prakash@in.bosch.com>
---
drivers/of/overlay.c | 5 ++---
1 file changed, 2 insertions(+), 3 deletions(-)
diff --git a/drivers/of/overlay.c b/drivers/of/overlay.c
index 68c281d..a0908ce 100644
--- a/drivers/of/overlay.c
+++ b/drivers/of/overlay.c
@@ -569,9 +569,6 @@ static int init_overlay_changeset(struct overlay_changeset *ovcs,
if (!of_node_is_root(tree))
pr_debug("%s() tree is not root\n", __func__);
- ovcs->overlay_tree = tree;
- ovcs->fdt = fdt;
-
INIT_LIST_HEAD(&ovcs->ovcs_list);
of_changeset_init(&ovcs->cset);
@@ -650,6 +647,8 @@ static int init_overlay_changeset(struct overlay_changeset *ovcs,
ovcs->id = id;
ovcs->count = cnt;
ovcs->fragments = fragments;
+ ovcs->overlay_tree = tree;
+ ovcs->fdt = fdt;
return 0;
--
2.7.4
reply other threads:[~2019-04-02 4:53 UTC|newest]
Thread overview: [no followups] expand[flat|nested] mbox.gz Atom feed
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=0171a15f-a1fc-bcc6-f10d-1bc0e401eadd@in.bosch.com \
--to=arunkumar.prakash@in.bosch.com \
--cc=devicetree@vger.kernel.org \
--cc=frowand.list@gmail.com \
--cc=linux-kernel@vger.kernel.org \
--cc=pantelis.antoniou@konsulko.com \
--cc=robh+dt@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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).