linux-kernel - [PATCH 4.19 02/46] i2c: fix bus recovery stop mode timing

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: <20200114094340.919318891@linuxfoundation.org>
Date:   Tue, 14 Jan 2020 11:01:19 +0100
From:   Greg Kroah-Hartman <gregkh@...uxfoundation.org>
To:     linux-kernel@...r.kernel.org
Cc:     Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
        stable@...r.kernel.org, Russell King <rmk+kernel@...linux.org.uk>,
        Wolfram Sang <wsa@...-dreams.de>
Subject: [PATCH 4.19 02/46] i2c: fix bus recovery stop mode timing

From: Russell King <rmk+kernel@...linux.org.uk>

commit cf8ce8b80f8bf9669f6ec4e71e16668430febdac upstream.

The I2C specification states that tsu:sto for standard mode timing must
be at minimum 4us. Pictographically, this is:

SCL: ____/~~~~~~~~~
SDA: _________/~~~~
       ->|    |<- 4us minimum

We are currently waiting 2.5us between asserting SCL and SDA, which is
in violation of the standard. Adjust the timings to ensure that we meet
what is stipulated as the minimum timings to ensure that all devices
correctly interpret the STOP bus transition.

This is more important than trying to generate a square wave with even
duty cycle.

Signed-off-by: Russell King <rmk+kernel@...linux.org.uk>
Signed-off-by: Wolfram Sang <wsa@...-dreams.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@...uxfoundation.org>

---
 drivers/i2c/i2c-core-base.c |   13 ++++++++++---
 1 file changed, 10 insertions(+), 3 deletions(-)

--- a/drivers/i2c/i2c-core-base.c
+++ b/drivers/i2c/i2c-core-base.c
@@ -194,10 +194,11 @@ int i2c_generic_scl_recovery(struct i2c_
 	 * If we can set SDA, we will always create a STOP to ensure additional
 	 * pulses will do no harm. This is achieved by letting SDA follow SCL
 	 * half a cycle later. Check the 'incomplete_write_byte' fault injector
-	 * for details.
+	 * for details. Note that we must honour tsu:sto, 4us, but lets use 5us
+	 * here for simplicity.
 	 */
 	bri->set_scl(adap, scl);
-	ndelay(RECOVERY_NDELAY / 2);
+	ndelay(RECOVERY_NDELAY);
 	if (bri->set_sda)
 		bri->set_sda(adap, scl);
 	ndelay(RECOVERY_NDELAY / 2);
@@ -219,7 +220,13 @@ int i2c_generic_scl_recovery(struct i2c_
 		scl = !scl;
 		bri->set_scl(adap, scl);
 		/* Creating STOP again, see above */
-		ndelay(RECOVERY_NDELAY / 2);
+		if (scl)  {
+			/* Honour minimum tsu:sto */
+			ndelay(RECOVERY_NDELAY);
+		} else {
+			/* Honour minimum tf and thd:dat */
+			ndelay(RECOVERY_NDELAY / 2);
+		}
 		if (bri->set_sda)
 			bri->set_sda(adap, scl);
 		ndelay(RECOVERY_NDELAY / 2);