Revision be14eb619108fa8b7120eb2c42d66d5f623ae10e authored by Brian King on 10 September 2010, 07:03:21 UTC, committed by Jens Axboe on 10 September 2010, 07:03:21 UTC
While testing CPU DLPAR, the following problem was discovered. We were DLPAR removing the first CPU, which in this case was logical CPUs 0-3. CPUs 0-2 were already marked offline and we were in the process of offlining CPU 3. After marking the CPU inactive and offline in cpu_disable, but before the cpu was completely idle (cpu_die), we ended up in __make_request on CPU 3. There we looked at the topology map to see which CPU to complete the I/O on and found no CPUs in the cpu_sibling_map. This resulted in the block layer setting the completion cpu to be NR_CPUS, which then caused an oops when we tried to complete the I/O. Fix this by sanity checking the value we return from blk_cpu_to_group to be a valid cpu value. Signed-off-by: Brian King <brking@linux.vnet.ibm.com> Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
1 parent edce682
lapb_timer.c
/*
* LAPB release 002
*
* This code REQUIRES 2.1.15 or higher/ NET3.038
*
* This module:
* This module is free software; you can redistribute it and/or
* modify it under the terms of the GNU General Public License
* as published by the Free Software Foundation; either version
* 2 of the License, or (at your option) any later version.
*
* History
* LAPB 001 Jonathan Naylor Started Coding
* LAPB 002 Jonathan Naylor New timer architecture.
*/
#include <linux/errno.h>
#include <linux/types.h>
#include <linux/socket.h>
#include <linux/in.h>
#include <linux/kernel.h>
#include <linux/jiffies.h>
#include <linux/timer.h>
#include <linux/string.h>
#include <linux/sockios.h>
#include <linux/net.h>
#include <linux/inet.h>
#include <linux/skbuff.h>
#include <net/sock.h>
#include <asm/uaccess.h>
#include <asm/system.h>
#include <linux/fcntl.h>
#include <linux/mm.h>
#include <linux/interrupt.h>
#include <net/lapb.h>
static void lapb_t1timer_expiry(unsigned long);
static void lapb_t2timer_expiry(unsigned long);
void lapb_start_t1timer(struct lapb_cb *lapb)
{
del_timer(&lapb->t1timer);
lapb->t1timer.data = (unsigned long)lapb;
lapb->t1timer.function = &lapb_t1timer_expiry;
lapb->t1timer.expires = jiffies + lapb->t1;
add_timer(&lapb->t1timer);
}
void lapb_start_t2timer(struct lapb_cb *lapb)
{
del_timer(&lapb->t2timer);
lapb->t2timer.data = (unsigned long)lapb;
lapb->t2timer.function = &lapb_t2timer_expiry;
lapb->t2timer.expires = jiffies + lapb->t2;
add_timer(&lapb->t2timer);
}
void lapb_stop_t1timer(struct lapb_cb *lapb)
{
del_timer(&lapb->t1timer);
}
void lapb_stop_t2timer(struct lapb_cb *lapb)
{
del_timer(&lapb->t2timer);
}
int lapb_t1timer_running(struct lapb_cb *lapb)
{
return timer_pending(&lapb->t1timer);
}
static void lapb_t2timer_expiry(unsigned long param)
{
struct lapb_cb *lapb = (struct lapb_cb *)param;
if (lapb->condition & LAPB_ACK_PENDING_CONDITION) {
lapb->condition &= ~LAPB_ACK_PENDING_CONDITION;
lapb_timeout_response(lapb);
}
}
static void lapb_t1timer_expiry(unsigned long param)
{
struct lapb_cb *lapb = (struct lapb_cb *)param;
switch (lapb->state) {
/*
* If we are a DCE, keep going DM .. DM .. DM
*/
case LAPB_STATE_0:
if (lapb->mode & LAPB_DCE)
lapb_send_control(lapb, LAPB_DM, LAPB_POLLOFF, LAPB_RESPONSE);
break;
/*
* Awaiting connection state, send SABM(E), up to N2 times.
*/
case LAPB_STATE_1:
if (lapb->n2count == lapb->n2) {
lapb_clear_queues(lapb);
lapb->state = LAPB_STATE_0;
lapb_disconnect_indication(lapb, LAPB_TIMEDOUT);
#if LAPB_DEBUG > 0
printk(KERN_DEBUG "lapb: (%p) S1 -> S0\n", lapb->dev);
#endif
return;
} else {
lapb->n2count++;
if (lapb->mode & LAPB_EXTENDED) {
#if LAPB_DEBUG > 1
printk(KERN_DEBUG "lapb: (%p) S1 TX SABME(1)\n", lapb->dev);
#endif
lapb_send_control(lapb, LAPB_SABME, LAPB_POLLON, LAPB_COMMAND);
} else {
#if LAPB_DEBUG > 1
printk(KERN_DEBUG "lapb: (%p) S1 TX SABM(1)\n", lapb->dev);
#endif
lapb_send_control(lapb, LAPB_SABM, LAPB_POLLON, LAPB_COMMAND);
}
}
break;
/*
* Awaiting disconnection state, send DISC, up to N2 times.
*/
case LAPB_STATE_2:
if (lapb->n2count == lapb->n2) {
lapb_clear_queues(lapb);
lapb->state = LAPB_STATE_0;
lapb_disconnect_confirmation(lapb, LAPB_TIMEDOUT);
#if LAPB_DEBUG > 0
printk(KERN_DEBUG "lapb: (%p) S2 -> S0\n", lapb->dev);
#endif
return;
} else {
lapb->n2count++;
#if LAPB_DEBUG > 1
printk(KERN_DEBUG "lapb: (%p) S2 TX DISC(1)\n", lapb->dev);
#endif
lapb_send_control(lapb, LAPB_DISC, LAPB_POLLON, LAPB_COMMAND);
}
break;
/*
* Data transfer state, restransmit I frames, up to N2 times.
*/
case LAPB_STATE_3:
if (lapb->n2count == lapb->n2) {
lapb_clear_queues(lapb);
lapb->state = LAPB_STATE_0;
lapb_stop_t2timer(lapb);
lapb_disconnect_indication(lapb, LAPB_TIMEDOUT);
#if LAPB_DEBUG > 0
printk(KERN_DEBUG "lapb: (%p) S3 -> S0\n", lapb->dev);
#endif
return;
} else {
lapb->n2count++;
lapb_requeue_frames(lapb);
}
break;
/*
* Frame reject state, restransmit FRMR frames, up to N2 times.
*/
case LAPB_STATE_4:
if (lapb->n2count == lapb->n2) {
lapb_clear_queues(lapb);
lapb->state = LAPB_STATE_0;
lapb_disconnect_indication(lapb, LAPB_TIMEDOUT);
#if LAPB_DEBUG > 0
printk(KERN_DEBUG "lapb: (%p) S4 -> S0\n", lapb->dev);
#endif
return;
} else {
lapb->n2count++;
lapb_transmit_frmr(lapb);
}
break;
}
lapb_start_t1timer(lapb);
}
![swh spinner](/static/img/swh-spinner.gif)
Computing file changes ...