Revision 5fa10196bdb5f190f595ebd048490ee52dddea0f authored by H. Peter Anvin on 07 March 2014, 23:05:20 UTC, committed by H. Peter Anvin on 07 March 2014, 23:08:14 UTC
Don Zickus reports: A customer generated an external NMI using their iLO to test kdump worked. Unfortunately, the machine hung. Disabling the nmi_watchdog made things work. I speculated the external NMI fired, caused the machine to panic (as expected) and the perf NMI from the watchdog came in and was latched. My guess was this somehow caused the hang. ---- It appears that the latched NMI stays latched until the early page table generation on 64 bits, which causes exceptions to happen which end in IRET, which re-enable NMI. Therefore, ignore NMIs that come in during early execution, until we have proper exception handling. Reported-and-tested-by: Don Zickus <dzickus@redhat.com> Link: http://lkml.kernel.org/r/1394221143-29713-1-git-send-email-dzickus@redhat.com Signed-off-by: H. Peter Anvin <hpa@linux.intel.com> Cc: <stable@vger.kernel.org> # v3.5+, older with some backport effort
1 parent d4078e2
ax25_ds_subr.c
/*
* This program 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.
*
* Copyright (C) Jonathan Naylor G4KLX (g4klx@g4klx.demon.co.uk)
* Copyright (C) Joerg Reuter DL1BKE (jreuter@yaina.de)
*/
#include <linux/errno.h>
#include <linux/types.h>
#include <linux/socket.h>
#include <linux/in.h>
#include <linux/kernel.h>
#include <linux/timer.h>
#include <linux/string.h>
#include <linux/sockios.h>
#include <linux/spinlock.h>
#include <linux/net.h>
#include <linux/gfp.h>
#include <net/ax25.h>
#include <linux/inet.h>
#include <linux/netdevice.h>
#include <linux/skbuff.h>
#include <net/sock.h>
#include <asm/uaccess.h>
#include <linux/fcntl.h>
#include <linux/mm.h>
#include <linux/interrupt.h>
void ax25_ds_nr_error_recovery(ax25_cb *ax25)
{
ax25_ds_establish_data_link(ax25);
}
/*
* dl1bke 960114: transmit I frames on DAMA poll
*/
void ax25_ds_enquiry_response(ax25_cb *ax25)
{
ax25_cb *ax25o;
/* Please note that neither DK4EG's nor DG2FEF's
* DAMA spec mention the following behaviour as seen
* with TheFirmware:
*
* DB0ACH->DL1BKE <RR C P R0> [DAMA]
* DL1BKE->DB0ACH <I NR=0 NS=0>
* DL1BKE-7->DB0PRA-6 DB0ACH <I C S3 R5>
* DL1BKE->DB0ACH <RR R F R0>
*
* The Flexnet DAMA Master implementation apparently
* insists on the "proper" AX.25 behaviour:
*
* DB0ACH->DL1BKE <RR C P R0> [DAMA]
* DL1BKE->DB0ACH <RR R F R0>
* DL1BKE->DB0ACH <I NR=0 NS=0>
* DL1BKE-7->DB0PRA-6 DB0ACH <I C S3 R5>
*
* Flexnet refuses to send us *any* I frame if we send
* a REJ in case AX25_COND_REJECT is set. It is superfluous in
* this mode anyway (a RR or RNR invokes the retransmission).
* Is this a Flexnet bug?
*/
ax25_std_enquiry_response(ax25);
if (!(ax25->condition & AX25_COND_PEER_RX_BUSY)) {
ax25_requeue_frames(ax25);
ax25_kick(ax25);
}
if (ax25->state == AX25_STATE_1 || ax25->state == AX25_STATE_2 || skb_peek(&ax25->ack_queue) != NULL)
ax25_ds_t1_timeout(ax25);
else
ax25->n2count = 0;
ax25_start_t3timer(ax25);
ax25_ds_set_timer(ax25->ax25_dev);
spin_lock(&ax25_list_lock);
ax25_for_each(ax25o, &ax25_list) {
if (ax25o == ax25)
continue;
if (ax25o->ax25_dev != ax25->ax25_dev)
continue;
if (ax25o->state == AX25_STATE_1 || ax25o->state == AX25_STATE_2) {
ax25_ds_t1_timeout(ax25o);
continue;
}
if (!(ax25o->condition & AX25_COND_PEER_RX_BUSY) && ax25o->state == AX25_STATE_3) {
ax25_requeue_frames(ax25o);
ax25_kick(ax25o);
}
if (ax25o->state == AX25_STATE_1 || ax25o->state == AX25_STATE_2 || skb_peek(&ax25o->ack_queue) != NULL)
ax25_ds_t1_timeout(ax25o);
/* do not start T3 for listening sockets (tnx DD8NE) */
if (ax25o->state != AX25_STATE_0)
ax25_start_t3timer(ax25o);
}
spin_unlock(&ax25_list_lock);
}
void ax25_ds_establish_data_link(ax25_cb *ax25)
{
ax25->condition &= AX25_COND_DAMA_MODE;
ax25->n2count = 0;
ax25_calculate_t1(ax25);
ax25_start_t1timer(ax25);
ax25_stop_t2timer(ax25);
ax25_start_t3timer(ax25);
}
/*
* :::FIXME:::
* This is a kludge. Not all drivers recognize kiss commands.
* We need a driver level request to switch duplex mode, that does
* either SCC changing, PI config or KISS as required. Currently
* this request isn't reliable.
*/
static void ax25_kiss_cmd(ax25_dev *ax25_dev, unsigned char cmd, unsigned char param)
{
struct sk_buff *skb;
unsigned char *p;
if (ax25_dev->dev == NULL)
return;
if ((skb = alloc_skb(2, GFP_ATOMIC)) == NULL)
return;
skb_reset_network_header(skb);
p = skb_put(skb, 2);
*p++ = cmd;
*p++ = param;
skb->protocol = ax25_type_trans(skb, ax25_dev->dev);
dev_queue_xmit(skb);
}
/*
* A nasty problem arises if we count the number of DAMA connections
* wrong, especially when connections on the device already existed
* and our network node (or the sysop) decides to turn on DAMA Master
* mode. We thus flag the 'real' slave connections with
* ax25->dama_slave=1 and look on every disconnect if still slave
* connections exist.
*/
static int ax25_check_dama_slave(ax25_dev *ax25_dev)
{
ax25_cb *ax25;
int res = 0;
spin_lock(&ax25_list_lock);
ax25_for_each(ax25, &ax25_list)
if (ax25->ax25_dev == ax25_dev && (ax25->condition & AX25_COND_DAMA_MODE) && ax25->state > AX25_STATE_1) {
res = 1;
break;
}
spin_unlock(&ax25_list_lock);
return res;
}
static void ax25_dev_dama_on(ax25_dev *ax25_dev)
{
if (ax25_dev == NULL)
return;
if (ax25_dev->dama.slave == 0)
ax25_kiss_cmd(ax25_dev, 5, 1);
ax25_dev->dama.slave = 1;
ax25_ds_set_timer(ax25_dev);
}
void ax25_dev_dama_off(ax25_dev *ax25_dev)
{
if (ax25_dev == NULL)
return;
if (ax25_dev->dama.slave && !ax25_check_dama_slave(ax25_dev)) {
ax25_kiss_cmd(ax25_dev, 5, 0);
ax25_dev->dama.slave = 0;
ax25_ds_del_timer(ax25_dev);
}
}
void ax25_dama_on(ax25_cb *ax25)
{
ax25_dev_dama_on(ax25->ax25_dev);
ax25->condition |= AX25_COND_DAMA_MODE;
}
void ax25_dama_off(ax25_cb *ax25)
{
ax25->condition &= ~AX25_COND_DAMA_MODE;
ax25_dev_dama_off(ax25->ax25_dev);
}
![swh spinner](/static/img/swh-spinner.gif)
Computing file changes ...