linux-kernel - [PATCH] ath5k: prevent infinite loop

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 for Android: free password hash cracker in your pocket

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

Message-ID: <alpine.DEB.2.00.0905182011270.7956@gandalf.stny.rr.com>
Date:	Mon, 18 May 2009 20:28:30 -0400 (EDT)
From:	Steven Rostedt <rostedt@...dmis.org>
To:	LKML <linux-kernel@...r.kernel.org>
cc:	jirislaby@...il.com, mickflemm@...il.com, lrodriguez@...eros.com,
	me@...copeland.com, linux-wireless@...r.kernel.org,
	ath5k-devel@...ts.ath5k.org,
	Andrew Morton <akpm@...ux-foundation.org>
Subject: [PATCH] ath5k: prevent infinite loop


After updating my laptop with the latest kernel (had 2.6.26 before that),
my laptop load went to 100%. The daemon phy0 was at 99.9% of the CPU.
Luckily I compiled with a preempt kernel otherwise this would have
been a lock up.

Using ftrace to dig into the problem I found that that the ath5k driver
was in an infinite loop. The code in ath5k_get_linear_pcdac_min has:

	pwr_i = pwrR[0];
	do {
		pwr_i--;
		tmp = (s8) ath5k_get_interpolated_value(pwr_i,
						pwrR[0], pwrR[1],
						stepR[0], stepR[1]);
	} while (tmp > 1);


But ath5k_get_interpolated returns stepR[0] if pwrR[0] == pwrR[1] or 
stepR[0] == stepR[1]. The pwr_i is ignored and we enter an infinite loop
because tmp never changes between iterations. Using ftrace, I was able to 
determine that is exactly what happened in the case of my laptop.

This patch tries to keep the same result that would happen when this case 
occurs. That is, the pwr_i becomes a minimal number. I used the minimum 
number that a signed short may be to initialize the min pwrL and pwrR.
Then if the case that the code would cause an infinite loop, we bypass it.

Signed-off-by: Steven Rostedt <rostedt@...dmis.org>

diff --git a/drivers/net/wireless/ath5k/phy.c b/drivers/net/wireless/ath5k/phy.c
index 9e2faae..2c916fc 100644
--- a/drivers/net/wireless/ath5k/phy.c
+++ b/drivers/net/wireless/ath5k/phy.c
@@ -1473,6 +1473,8 @@ ath5k_get_interpolated_value(s16 target, s16 x_left, s16 x_right,
 	return result;
 }
 
+#define MIN_PWR (-32768)
+
 /*
  * Find vertical boundary (min pwr) for the linear PCDAC curve.
  *
@@ -1486,29 +1488,32 @@ ath5k_get_linear_pcdac_min(const u8 *stepL, const u8 *stepR,
 				const s16 *pwrL, const s16 *pwrR)
 {
 	s8 tmp;
-	s16 min_pwrL, min_pwrR;
+	s16 min_pwrL = MIN_PWR, min_pwrR = MIN_PWR;
 	s16 pwr_i = pwrL[0];
 
-	do {
-		pwr_i--;
-		tmp = (s8) ath5k_get_interpolated_value(pwr_i,
-						pwrL[0], pwrL[1],
-						stepL[0], stepL[1]);
-
-	} while (tmp > 1);
-
-	min_pwrL = pwr_i;
+	/* Avoid infinite loop */
+	if (pwrL[0] != pwrL[1] && stepL[0] != stepL[1]) {
+		do {
+			pwr_i--;
+			tmp = (s8) ath5k_get_interpolated_value(pwr_i,
+							pwrL[0], pwrL[1],
+							stepL[0], stepL[1]);
+		} while (tmp > 1);
+		min_pwrL = pwr_i;
+	}
 
 	pwr_i = pwrR[0];
-	do {
-		pwr_i--;
-		tmp = (s8) ath5k_get_interpolated_value(pwr_i,
-						pwrR[0], pwrR[1],
-						stepR[0], stepR[1]);
-
-	} while (tmp > 1);
 
-	min_pwrR = pwr_i;
+	/* Avoid infinite loop */
+	if (pwrR[0] != pwrR[1] && stepR[0] != stepR[1]) {
+		do {
+			pwr_i--;
+			tmp = (s8) ath5k_get_interpolated_value(pwr_i,
+							pwrR[0], pwrR[1],
+							stepR[0], stepR[1]);
+		} while (tmp > 1);
+		min_pwrR = pwr_i;
+	}
 
 	/* Keep the right boundary so that it works for both curves */
 	return max(min_pwrL, min_pwrR);
--
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/