linux-kernel - [PATCH 5/8] ipmi: Reduce polling when interrupts are available

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>] [day] [month] [year] [list]

Date:	Mon, 03 May 2010 08:36:23 -0500
From:	Corey Minyard <minyard@....org>
To:	Linux Kernel <linux-kernel@...r.kernel.org>,
	Andrew Morton <akpm@...l.org>
Cc:	OpenIPMI Developers <openipmi-developer@...ts.sourceforge.net>,
	Matthew Garrett <mjg@...hat.com>
Subject: [PATCH 5/8] ipmi: Reduce polling when interrupts are available

From: Matthew Garrett <mjg@...hat.com>

If we're not currently in the middle of a transaction, and if we have
interrupts, there's no real reason to poll the controller more frequently
than the core IPMI code does. Set the interrupt_disabled flag appropriately
as the interrupt state changes, and make the timeout code reset itself
only if the transaction is incomplete or we have no interrupts.

Signed-off-by: Matthew Garrett <mjg@...hat.com>
Signed-off-by: Corey Minyard <cminyard@...sta.com>
---

Index: linux-2.6/drivers/char/ipmi/ipmi_si_intf.c
===================================================================
--- linux-2.6.orig/drivers/char/ipmi/ipmi_si_intf.c
+++ linux-2.6/drivers/char/ipmi/ipmi_si_intf.c
@@ -453,6 +453,9 @@ static inline void disable_si_irq(struct
 	if ((smi_info->irq) && (!smi_info->interrupt_disabled)) {
 		start_disable_irq(smi_info);
 		smi_info->interrupt_disabled = 1;
+		if (!atomic_read(&smi_info->stop_operation))
+			mod_timer(&smi_info->si_timer,
+				  jiffies + SI_TIMEOUT_JIFFIES);
 	}
 }
 
@@ -705,6 +708,8 @@ static void handle_transaction_done(stru
 			printk(KERN_WARNING
 			       "ipmi_si: Could not enable interrupts"
 			       ", failed set, using polled mode.\n");
+		} else {
+			smi_info->interrupt_disabled = 0;
 		}
 		smi_info->si_state = SI_NORMAL;
 		break;
@@ -885,6 +890,8 @@ static void sender(void                *
 	printk("**Enqueue: %d.%9.9d\n", t.tv_sec, t.tv_usec);
 #endif
 
+	mod_timer(&smi_info->si_timer, jiffies + SI_TIMEOUT_JIFFIES);
+
 	if (smi_info->run_to_completion) {
 		/*
 		 * If we are running to completion, then throw it in
@@ -1085,7 +1092,8 @@ static void smi_timeout(unsigned long da
 	}
 
  do_add_timer:
-	add_timer(&(smi_info->si_timer));
+	if ((smi_result != SI_SM_IDLE) || smi_info->interrupt_disabled)
+		add_timer(&(smi_info->si_timer));
 }
 
 static irqreturn_t si_irq_handler(int irq, void *data)
@@ -3115,7 +3123,7 @@ static int try_smi_init(struct smi_info 
 	for (i = 0; i < SI_NUM_STATS; i++)
 		atomic_set(&new_smi->stats[i], 0);
 
-	new_smi->interrupt_disabled = 0;
+	new_smi->interrupt_disabled = 1;
 	atomic_set(&new_smi->stop_operation, 0);
 	new_smi->intf_num = smi_num;
 	smi_num++;
--
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/