linux-kernel - [PATCH 1/2] Staging: zram: Turn lockdep off during zram_init

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-next>] [day] [month] [year] [list]

Date:	Wed, 30 Nov 2011 14:14:32 +0100
From:	Jerome Marchand <jmarchan@...hat.com>
To:	Greg Kroah-Hartman <gregkh@...e.de>
CC:	Nitin Gupta <ngupta@...are.org>,
	Robert Jennings <rcj@...ux.vnet.ibm.com>,
	Witold Baryluk <baryluk@....if.uj.edu.pl>,
	Peter Zijlstra <peterz@...radead.org>,
	Linux Kernel Mailing List <linux-kernel@...r.kernel.org>
Subject: [PATCH 1/2] Staging: zram: Turn lockdep off during zram_init_device()


zram->init_lock can be hold over an allocation that may try to reclaim
memory in zram_init_device(). The same lock can later be taken from a
reclaim context in zram_make_request(), thus triggering a lockdep
warning. However, memory can not be reclaimed to an uninitialized zram
device. Therefore, this warning is a false positive.
To prevent the warning to occur, we turn lockdep off during while the
device is initialized.

Signed-off-by: Jerome Marchand <jmarchan@...hat.com>
Reported-by: Witold Baryluk <baryluk@....if.uj.edu.pl>
---
 zram_drv.c |   14 ++++++++++++++
 1 file changed, 14 insertions(+)

diff --git a/drivers/staging/zram/zram_drv.c b/drivers/staging/zram/zram_drv.c
index 09de99f..34f4b96 100644
--- a/drivers/staging/zram/zram_drv.c
+++ b/drivers/staging/zram/zram_drv.c
@@ -637,10 +637,22 @@ int zram_init_device(struct zram *zram)
 	int ret;
 	size_t num_pages;
 
+	/*
+	 * Here we hold init_lock over allocations that may try to reclaim
+	 * memory. The same lock is also taken from a reclaim context in
+	 * zram_make_request(). That end up in a lockdep warning.
+	 * Moreover the table can be quite big an it wouldn't be reasonable to
+	 * use kmalloc(GFP_NOFS,...) to allocate it.
+	 * Fortunately, memory can not be reclaimed to an uninitialized zram
+	 * device, so this warning is a false positive. Therefore, to prevent
+	 * the warning, we turn lockdep off here.
+	 */
+	lockdep_off();
 	down_write(&zram->init_lock);
 
 	if (zram->init_done) {
 		up_write(&zram->init_lock);
+		lockdep_on();
 		return 0;
 	}
 
@@ -682,6 +694,7 @@ int zram_init_device(struct zram *zram)
 
 	zram->init_done = 1;
 	up_write(&zram->init_lock);
+	lockdep_on();
 
 	pr_debug("Initialization done!\n");
 	return 0;
@@ -692,6 +705,7 @@ fail_no_table:
 fail:
 	__zram_reset_device(zram);
 	up_write(&zram->init_lock);
+	lockdep_on();
 	pr_err("Initialization failed: err=%d\n", ret);
 	return ret;
 }
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/