linux-kernel - [PATCH v4] misc: alcor_pci: fix null-ptr-deref when there is no PCI bridge

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-Id: <20210518192820.181500-1-ztong0001@gmail.com>
Date:   Tue, 18 May 2021 15:28:18 -0400
From:   Tong Zhang <ztong0001@...il.com>
To:     Arnd Bergmann <arnd@...db.de>,
        Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
        Tong Zhang <ztong0001@...il.com>, linux-kernel@...r.kernel.org
Cc:     Colin Ian King <colin.king@...onical.com>
Subject: [PATCH v4] misc: alcor_pci: fix null-ptr-deref when there is no PCI bridge

There is an issue with the ASPM(optional) capability checking function.
A device might be attached to root complex directly, in this case,
bus->self(bridge) will be NULL, thus priv->parent_pdev is NULL.
Since alcor_pci_init_check_aspm(priv->parent_pdev) checks the PCI link's
ASPM capability and populate parent_cap_off, which will be used later by
alcor_pci_aspm_ctrl() to dynamically turn on/off device, what we can do
here is to avoid checking the capability if we are on the root complex.
This will make pdev_cap_off 0 and alcor_pci_aspm_ctrl() will simply
return when bring called, effectively disable ASPM for the device.

[    1.246492] BUG: kernel NULL pointer dereference, address: 00000000000000c0
[    1.248731] RIP: 0010:pci_read_config_byte+0x5/0x40
[    1.253998] Call Trace:
[    1.254131]  ? alcor_pci_find_cap_offset.isra.0+0x3a/0x100 [alcor_pci]
[    1.254476]  alcor_pci_probe+0x169/0x2d5 [alcor_pci]

Signed-off-by: Tong Zhang <ztong0001@...il.com>
Co-developed-by: Greg Kroah-Hartman <gregkh@...uxfoundation.org>
Co-developed-by: Colin Ian King <colin.king@...onical.com>
---
v2: check before calling alcor_pci_find_cap_offset()
v3: Add comment. Enable the dev_dbg() output when priv->parent_pdev is NULL.
v4: fix inverted if statement, thanks to Colin <colin.king@...onical.com>

 drivers/misc/cardreader/alcor_pci.c | 8 +++++++-
 1 file changed, 7 insertions(+), 1 deletion(-)

diff --git a/drivers/misc/cardreader/alcor_pci.c b/drivers/misc/cardreader/alcor_pci.c
index cd402c89189e..de6d44a158bb 100644
--- a/drivers/misc/cardreader/alcor_pci.c
+++ b/drivers/misc/cardreader/alcor_pci.c
@@ -139,7 +139,13 @@ static void alcor_pci_init_check_aspm(struct alcor_pci_priv *priv)
 	u32 val32;
 
 	priv->pdev_cap_off    = alcor_pci_find_cap_offset(priv, priv->pdev);
-	priv->parent_cap_off = alcor_pci_find_cap_offset(priv,
+	/*
+	 * A device might be attached to root complex directly and
+	 * priv->parent_pdev will be NULL. In this case we don't check its
+	 * capability and disable ASPM completely.
+	 */
+	if (priv->parent_pdev)
+		priv->parent_cap_off = alcor_pci_find_cap_offset(priv,
 							 priv->parent_pdev);
 
 	if ((priv->pdev_cap_off == 0) || (priv->parent_cap_off == 0)) {
-- 
2.25.1