linux-kernel - [PATCH 1/2] speedstep-smi: enable interrupts when waiting

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]

Message-ID: <alpine.LRH.2.02.1405141012040.11670@file01.intranet.prod.int.rdu2.redhat.com>
Date:	Wed, 14 May 2014 10:24:44 -0400 (EDT)
From:	Mikulas Patocka <mpatocka@...hat.com>
To:	Viresh Kumar <viresh.kumar@...aro.org>, cpufreq@...r.kernel.org
cc:	linux-kernel@...r.kernel.org,
	"Rafael J. Wysocki" <rjw@...ysocki.net>, linux-pm@...r.kernel.org
Subject: [PATCH 1/2] speedstep-smi: enable interrupts when waiting

Hi Viresh

Some times ago, I sent these two cpufreq patches to fix frequency 
transition problem on speedstep-smi, but you didn't put them to the 
upstream kernel. I am resending the patches. If you would like a different 
solution to this problem, please explain it.

Mikulas


On Dell Latitude C600 laptop with Pentium 3 850MHz processor, the
speedstep-smi driver sometimes loads and sometimes doesn't load with
"change to state X failed" message.

The hardware sometimes refuses to change frequency and in this case, we 
need to retry later. I found out that we need to enable interrupts while 
waiting for retry. When we enable interrupts, the hardware blockage that 
prevents frequency transition resolves and the transition is possible. 
With disabled interrupts, the blockage doesn't resolve (no matter how long 
do we wait). The exact reasons for this hardware behavior are unknown.

This patch enables interrupts in the function speedstep_set_state that can
be called with disabled interrupts. However, this function is called with
disabled interrupts only from speedstep_get_freqs, so it shouldn't cause
any problem.

Signed-off-by: Mikulas Patocka <mpatocka@...hat.com
Cc: stable@...r.kernel.org

---
 drivers/cpufreq/speedstep-lib.c |    3 +++
 drivers/cpufreq/speedstep-smi.c |   12 ++++++++++++
 2 files changed, 15 insertions(+)

Index: linux-3.14-rc1/drivers/cpufreq/speedstep-smi.c
===================================================================
--- linux-3.14-rc1.orig/drivers/cpufreq/speedstep-smi.c	2014-02-03 18:56:44.000000000 +0100
+++ linux-3.14-rc1/drivers/cpufreq/speedstep-smi.c	2014-02-03 19:06:21.000000000 +0100
@@ -156,6 +156,7 @@ static void speedstep_set_state(unsigned
 		return;
 
 	/* Disable IRQs */
+	preempt_disable();
 	local_irq_save(flags);
 
 	command = (smi_sig & 0xffffff00) | (smi_cmd & 0xff);
@@ -166,9 +167,19 @@ static void speedstep_set_state(unsigned
 
 	do {
 		if (retry) {
+			/*
+			 * We need to enable interrupts, otherwise the blockage
+			 * won't resolve.
+			 *
+			 * We disable preemption so that other processes don't
+			 * run. If other processes were running, they could
+			 * submit more DMA requests, making the blockage worse.
+			 */
 			pr_debug("retry %u, previous result %u, waiting...\n",
 					retry, result);
+			local_irq_enable();
 			mdelay(retry * 50);
+			local_irq_disable();
 		}
 		retry++;
 		__asm__ __volatile__(
@@ -185,6 +196,7 @@ static void speedstep_set_state(unsigned
 
 	/* enable IRQs */
 	local_irq_restore(flags);
+	preempt_enable();
 
 	if (new_state == state)
 		pr_debug("change to %u MHz succeeded after %u tries "
Index: linux-3.14-rc1/drivers/cpufreq/speedstep-lib.c
===================================================================
--- linux-3.14-rc1.orig/drivers/cpufreq/speedstep-lib.c	2014-02-03 18:54:39.000000000 +0100
+++ linux-3.14-rc1/drivers/cpufreq/speedstep-lib.c	2014-02-03 19:06:21.000000000 +0100
@@ -400,6 +400,7 @@ unsigned int speedstep_get_freqs(enum sp
 
 	pr_debug("previous speed is %u\n", prev_speed);
 
+	preempt_disable();
 	local_irq_save(flags);
 
 	/* switch to low state */
@@ -464,6 +465,8 @@ unsigned int speedstep_get_freqs(enum sp
 
 out:
 	local_irq_restore(flags);
+	preempt_enable();
+
 	return ret;
 }
 EXPORT_SYMBOL_GPL(speedstep_get_freqs);
--
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/