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]
Date:	Wed, 27 Jan 2016 21:51:41 -0800
From:	Brian Norris <computersforpeace@...il.com>
To:	<linux-mtd@...ts.infradead.org>
Cc:	Brian Norris <computersforpeace@...il.com>,
	Rafał Miłecki <zajec5@...il.com>,
	Ezequiel Garcia <ezequiel@...guardiasur.com.ar>,
	Boris Brezillon <boris.brezillon@...e-electrons.com>,
	linux-kernel@...r.kernel.org, Bayi Cheng <bayi.cheng@...iatek.com>,
	Marek Vasut <marex@...x.de>, djkurtz@...omium.org
Subject: [PATCH 2/8] mtd: spi-nor: guard against underflows in stm_is_locked_sr

Users of stm_is_locked_sr() might do arithmetic that could result in a
negative offset. For example, when stm_unlock() tries to check the
status of the eraseblock below the range, it doesn't check for:

  ofs - mtd->erasesize < 0

Instead of forcing callers to be extra careful, let's just make
stm_is_locked_sr() do the right thing and report errors for invalid
ranges.

Also, fixup the calculations in stm_unlock(), so we:
(a) can handle non-eraseblock-aligned offsets and
(b) don't look for a negative offset when checking the first block

Signed-off-by: Brian Norris <computersforpeace@...il.com>
---
 drivers/mtd/spi-nor/spi-nor.c |   13 ++++++++++---
 1 file changed, 10 insertions(+), 3 deletions(-)

diff --git a/drivers/mtd/spi-nor/spi-nor.c b/drivers/mtd/spi-nor/spi-nor.c
index ef89bed1e5ea..c19674573eec 100644
--- a/drivers/mtd/spi-nor/spi-nor.c
+++ b/drivers/mtd/spi-nor/spi-nor.c
@@ -447,6 +447,9 @@ static int stm_is_locked_sr(struct spi_nor *nor, loff_t ofs, uint64_t len,
 	loff_t lock_offs;
 	uint64_t lock_len;
 
+	if (ofs < 0 || ofs + len > nor->mtd.size)
+		return -EINVAL;
+
 	stm_get_locked_range(nor, sr, &lock_offs, &lock_len);
 
 	return (ofs + len <= lock_offs + lock_len) && (ofs >= lock_offs);
@@ -543,9 +546,13 @@ static int stm_unlock(struct spi_nor *nor, loff_t ofs, uint64_t len)
 	if (status_old < 0)
 		return status_old;
 
-	/* Cannot unlock; would unlock larger region than requested */
-	if (stm_is_locked_sr(nor, ofs - mtd->erasesize, mtd->erasesize,
-			     status_old))
+	/*
+	 * Check the eraseblock next to us; if locked, then this would unlock
+	 * larger region than requested
+	 */
+	if (ofs > 0 && stm_is_locked_sr(nor, ALIGN(ofs - mtd->erasesize,
+					mtd->erasesize), mtd->erasesize,
+					status_old))
 		return -EINVAL;
 
 	/*
-- 
1.7.9.5

Powered by blists - more mailing lists