From: Thomas Gleixner <tglx@linutronix.de>
To: "Rafael J. Wysocki" <rjw@sisk.pl>
Cc: "Linus Torvalds" <torvalds@linux-foundation.org>,
"Linux Kernel Mailing List" <linux-kernel@vger.kernel.org>,
"Frédéric RISS" <frederic.riss@gmail.com>,
"Thomas Meyer" <thomas@m3y3r.de>, "Milan Broz" <mbroz@redhat.com>,
"Andrew Morton" <akpm@linux-foundation.org>,
"Ingo Molnar" <mingo@elte.hu>, "Adrian Bunk" <bunk@stusta.de>
Subject: [PATCH] clockevents: Fix suspend/resume to disk hangs
Date: Sat, 17 Mar 2007 00:25:52 +0100 [thread overview]
Message-ID: <1174087552.13341.332.camel@localhost.localdomain> (raw)
In-Reply-To: <1174078066.13341.279.camel@localhost.localdomain>
I finally found a dual core box, which survives suspend/resume without
crashing in the middle of nowhere. Sigh, I never figured out from the
code and the bug reports what's going on.
The observed hangs are caused by a stale state transition of the clock
event devices, which keeps the RCU synchronization away from completion,
when the non boot CPU is brought back up.
The suspend/resume in oneshot mode needs the similar care as the
periodic mode during suspend to RAM. My assumption that the state
transitions during the different shutdown/bringups of s2disk would go
through the periodic boot phase and then switch over to highres resp.
nohz mode were simply wrong.
Add the appropriate suspend / resume handling for the non periodic
modes.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
diff --git a/kernel/time/tick-broadcast.c b/kernel/time/tick-broadcast.c
index 5567745..eadfce2 100644
--- a/kernel/time/tick-broadcast.c
+++ b/kernel/time/tick-broadcast.c
@@ -307,12 +307,19 @@ int tick_resume_broadcast(void)
spin_lock_irqsave(&tick_broadcast_lock, flags);
bc = tick_broadcast_device.evtdev;
- if (bc) {
- if (tick_broadcast_device.mode == TICKDEV_MODE_PERIODIC &&
- !cpus_empty(tick_broadcast_mask))
- tick_broadcast_start_periodic(bc);
- broadcast = cpu_isset(smp_processor_id(), tick_broadcast_mask);
+ if (bc) {
+ switch (tick_broadcast_device.mode) {
+ case TICKDEV_MODE_PERIODIC:
+ if(!cpus_empty(tick_broadcast_mask))
+ tick_broadcast_start_periodic(bc);
+ broadcast = cpu_isset(smp_processor_id(),
+ tick_broadcast_mask);
+ break;
+ case TICKDEV_MODE_ONESHOT:
+ broadcast = tick_resume_broadcast_oneshot(bc);
+ break;
+ }
}
spin_unlock_irqrestore(&tick_broadcast_lock, flags);
@@ -347,6 +354,16 @@ static int tick_broadcast_set_event(ktime_t expires, int force)
}
}
+int tick_resume_broadcast_oneshot(struct clock_event_device *bc)
+{
+ clockevents_set_mode(bc, CLOCK_EVT_MODE_ONESHOT);
+
+ if(!cpus_empty(tick_broadcast_oneshot_mask))
+ tick_broadcast_set_event(ktime_get(), 1);
+
+ return cpu_isset(smp_processor_id(), tick_broadcast_oneshot_mask);
+}
+
/*
* Reprogram the broadcast device:
*
diff --git a/kernel/time/tick-common.c b/kernel/time/tick-common.c
index 43ba1bd..bfda3f7 100644
--- a/kernel/time/tick-common.c
+++ b/kernel/time/tick-common.c
@@ -298,18 +298,17 @@ static void tick_shutdown(unsigned int *cpup)
spin_unlock_irqrestore(&tick_device_lock, flags);
}
-static void tick_suspend_periodic(void)
+static void tick_suspend(void)
{
struct tick_device *td = &__get_cpu_var(tick_cpu_device);
unsigned long flags;
spin_lock_irqsave(&tick_device_lock, flags);
- if (td->mode == TICKDEV_MODE_PERIODIC)
- clockevents_set_mode(td->evtdev, CLOCK_EVT_MODE_SHUTDOWN);
+ clockevents_set_mode(td->evtdev, CLOCK_EVT_MODE_SHUTDOWN);
spin_unlock_irqrestore(&tick_device_lock, flags);
}
-static void tick_resume_periodic(void)
+static void tick_resume(void)
{
struct tick_device *td = &__get_cpu_var(tick_cpu_device);
unsigned long flags;
@@ -317,6 +316,8 @@ static void tick_resume_periodic(void)
spin_lock_irqsave(&tick_device_lock, flags);
if (td->mode == TICKDEV_MODE_PERIODIC)
tick_setup_periodic(td->evtdev, 0);
+ else
+ tick_resume_oneshot();
spin_unlock_irqrestore(&tick_device_lock, flags);
}
@@ -348,13 +349,13 @@ static int tick_notify(struct notifier_block *nb, unsigned long reason,
break;
case CLOCK_EVT_NOTIFY_SUSPEND:
- tick_suspend_periodic();
+ tick_suspend();
tick_suspend_broadcast();
break;
case CLOCK_EVT_NOTIFY_RESUME:
if (!tick_resume_broadcast())
- tick_resume_periodic();
+ tick_resume();
break;
default:
diff --git a/kernel/time/tick-internal.h b/kernel/time/tick-internal.h
index 75890ef..c9d203b 100644
--- a/kernel/time/tick-internal.h
+++ b/kernel/time/tick-internal.h
@@ -19,12 +19,13 @@ extern void tick_setup_oneshot(struct clock_event_device *newdev,
extern int tick_program_event(ktime_t expires, int force);
extern void tick_oneshot_notify(void);
extern int tick_switch_to_oneshot(void (*handler)(struct clock_event_device *));
-
+extern void tick_resume_oneshot(void);
# ifdef CONFIG_GENERIC_CLOCKEVENTS_BROADCAST
extern void tick_broadcast_setup_oneshot(struct clock_event_device *bc);
extern void tick_broadcast_oneshot_control(unsigned long reason);
extern void tick_broadcast_switch_to_oneshot(void);
extern void tick_shutdown_broadcast_oneshot(unsigned int *cpup);
+extern int tick_resume_broadcast_oneshot(struct clock_event_device *bc);
# else /* BROADCAST */
static inline void tick_broadcast_setup_oneshot(struct clock_event_device *bc)
{
@@ -43,6 +44,10 @@ void tick_setup_oneshot(struct clock_event_device *newdev,
{
BUG();
}
+static inline void tick_resume_oneshot(void)
+{
+ BUG();
+}
static inline int tick_program_event(ktime_t expires, int force)
{
return 0;
@@ -54,6 +59,10 @@ static inline void tick_broadcast_setup_oneshot(struct clock_event_device *bc)
}
static inline void tick_broadcast_oneshot_control(unsigned long reason) { }
static inline void tick_shutdown_broadcast_oneshot(unsigned int *cpup) { }
+static inline int tick_resume_broadcast_oneshot(struct clock_event_device *bc)
+{
+ return 0;
+}
#endif /* !TICK_ONESHOT */
/*
diff --git a/kernel/time/tick-oneshot.c b/kernel/time/tick-oneshot.c
index 2e8b7ff..f6997ab 100644
--- a/kernel/time/tick-oneshot.c
+++ b/kernel/time/tick-oneshot.c
@@ -41,6 +41,18 @@ int tick_program_event(ktime_t expires, int force)
}
/**
+ * tick_resume_onshot - resume oneshot mode
+ */
+void tick_resume_oneshot(void)
+{
+ struct tick_device *td = &__get_cpu_var(tick_cpu_device);
+ struct clock_event_device *dev = td->evtdev;
+
+ clockevents_set_mode(dev, CLOCK_EVT_MODE_ONESHOT);
+ tick_program_event(ktime_get(), 1);
+}
+
+/**
* tick_setup_oneshot - setup the event device for oneshot mode (hres or nohz)
*/
void tick_setup_oneshot(struct clock_event_device *newdev,
next prev parent reply other threads:[~2007-03-16 23:19 UTC|newest]
Thread overview: 259+ messages / expand[flat|nested] mbox.gz Atom feed top
2007-03-16 16:33 Linux 2.6.21-rc4 Linus Torvalds
2007-03-16 17:01 ` Takashi Iwai
2007-03-16 17:44 ` Michal Piotrowski
2007-03-16 18:26 ` Andrew Morton
2007-03-16 18:55 ` Michal Piotrowski
2007-03-16 23:23 ` Jan Engelhardt
2007-03-16 23:31 ` Michal Piotrowski
2007-03-17 8:19 ` Mariusz Kozlowski
2007-03-16 18:54 ` Takashi Iwai
2007-03-16 19:03 ` Michal Piotrowski
2007-03-17 23:46 ` Adrian Bunk
2007-03-18 13:04 ` Michal Piotrowski
2007-03-16 20:34 ` Rafael J. Wysocki
2007-03-16 20:47 ` Thomas Gleixner
2007-03-16 23:25 ` Thomas Gleixner [this message]
2007-03-17 9:35 ` [PATCH] clockevents: Fix suspend/resume to disk hangs Milan Broz
2007-03-17 10:07 ` Thomas Meyer
2007-03-17 21:47 ` Rafael J. Wysocki
2007-03-18 17:58 ` Adrian Bunk
2007-03-18 0:42 ` appletouch quirk doesn't run at resume Adrian Bunk
2007-03-18 18:45 ` Jiri Kosina
2007-03-18 19:01 ` Thomas Meyer
2007-03-18 19:22 ` Jiri Kosina
2007-03-27 21:02 ` Thomas Meyer
2007-03-28 12:26 ` Jiri Kosina
2007-03-28 13:24 ` Dmitry Torokhov
2007-03-28 16:51 ` Thomas Meyer
2007-03-28 17:06 ` Jiri Kosina
2007-03-28 17:35 ` Dmitry Torokhov
2007-03-20 9:35 ` [PATCH] clockevents: Fix suspend/resume to disk hangs Marcus Better
2007-03-21 14:04 ` Thomas Gleixner
2007-03-22 10:34 ` Marcus Better
2007-03-23 9:14 ` Marcus Better
2007-03-23 10:05 ` Tino Keitel
2007-03-23 13:47 ` Rafael J. Wysocki
2007-03-23 14:36 ` Marcus Better
2007-03-16 21:11 ` Linux 2.6.21-rc4 Randy Dunlap
2007-03-16 22:39 ` Randy Dunlap
2007-03-16 23:13 ` Chris Friesen
2007-03-16 23:27 ` Jan Engelhardt
2007-03-17 6:43 ` Sam Ravnborg
2007-03-18 12:39 ` Sam Ravnborg
2007-03-19 4:16 ` Randy Dunlap
2007-03-18 18:49 ` [1/6] 2.6.21-rc4: known regressions Adrian Bunk
2007-03-20 10:24 ` Tobias Diedrich
2007-03-20 11:14 ` Adrian Bunk
2007-03-22 3:45 ` Linus Torvalds
2007-03-22 4:18 ` Nick Piggin
2007-03-22 15:21 ` Linus Torvalds
2007-03-23 1:08 ` Mingming Cao
2007-03-23 1:40 ` Linus Torvalds
2007-03-23 2:11 ` Nick Piggin
2007-03-23 7:51 ` Michal Piotrowski
2007-03-23 9:37 ` Nick Piggin
2007-03-23 17:19 ` Adrian Bunk
2007-03-23 12:01 ` [patch] hrtimers debug patch Ingo Molnar
[not found] ` <4607BDD9.1010002@googlemail.com>
[not found] ` <6bffcb0e0703260720i37bbb956o3d20019fe4ac9879@mail.gmail.com>
2007-03-26 14:33 ` Thomas Gleixner
2007-03-26 14:42 ` Michal Piotrowski
2007-03-26 15:07 ` Michal Piotrowski
2007-03-26 17:02 ` Ingo Molnar
2007-03-26 17:50 ` Michal Piotrowski
2007-04-06 15:27 ` Michal Piotrowski
2007-04-06 16:39 ` Ingo Molnar
2007-03-23 11:42 ` [1/6] 2.6.21-rc4: known regressions Ingo Molnar
2007-03-23 11:56 ` Thomas Gleixner
2007-03-23 15:08 ` [PATCH] i386: add command line option "local_apic_timer_c2_ok" Thomas Gleixner
2007-03-26 12:31 ` Pavel Machek
2007-03-26 13:52 ` Thomas Gleixner
2007-03-27 21:19 ` Len Brown
2007-03-27 21:34 ` Linus Torvalds
2007-03-27 22:16 ` Len Brown
2007-03-28 2:18 ` Len Brown
2007-03-29 14:15 ` Andi Kleen
2007-03-29 14:53 ` Langsdorf, Mark
2007-03-29 16:50 ` Andi Kleen
2007-03-29 20:02 ` Mark Langsdorf
2007-03-29 20:49 ` Andi Kleen
2007-03-29 21:16 ` Linus Torvalds
2007-03-29 21:45 ` Andreas Mohr
2007-03-29 21:56 ` Linus Torvalds
2007-03-29 22:06 ` Andi Kleen
2007-03-29 22:05 ` Andi Kleen
2007-03-30 21:06 ` Grzegorz Chwesewicz
2007-03-31 7:47 ` Grzegorz Chwesewicz
2007-03-29 21:43 ` Grzegorz Chwesewicz
2007-03-29 21:55 ` Grzegorz Chwesewicz
2007-03-29 14:19 ` Andi Kleen
2007-03-23 18:13 ` [1/6] 2.6.21-rc4: known regressions Linus Torvalds
2007-03-23 18:16 ` Linus Torvalds
2007-03-23 18:28 ` Linus Torvalds
2007-03-23 18:43 ` Thomas Gleixner
2007-03-23 12:27 ` Ingo Molnar
2007-03-22 18:24 ` Mariusz Kozłowski
2007-03-18 18:49 ` [2/6] " Adrian Bunk
2007-03-18 19:25 ` Andi Kleen
2007-03-19 16:06 ` Randy Dunlap
2007-03-19 16:15 ` Adrian Bunk
2007-03-19 17:07 ` Randy Dunlap
2007-03-20 15:32 ` Ray Lee
2007-03-18 18:49 ` [3/6] " Adrian Bunk
2007-03-26 1:25 ` Jeff Chua
2007-03-26 4:05 ` Adrian Bunk
2007-03-26 5:37 ` Jeff Chua
2007-03-26 16:26 ` Thomas Gleixner
2007-03-26 17:46 ` Jeff Chua
2007-03-28 7:04 ` Thomas Gleixner
2007-03-28 13:43 ` Maxim
2007-03-28 14:41 ` Ingo Molnar
2007-03-28 15:01 ` Maxim
2007-03-28 16:38 ` Linus Torvalds
2007-03-28 19:38 ` [linux-pm] " David Brownell
2007-03-28 20:19 ` Maxim
2007-03-28 20:59 ` David Brownell
2007-03-28 21:27 ` Maxim
2007-03-29 22:33 ` David Brownell
2007-03-29 23:29 ` Maxim Levitsky
2007-03-30 0:09 ` David Brownell
2007-03-30 0:48 ` Maxim Levitsky
2007-03-28 20:42 ` Linus Torvalds
2007-03-28 21:17 ` David Brownell
2007-03-28 22:26 ` Maxim
2007-03-29 4:41 ` [ PATCH] Add suspend/resume for HPET was: " Maxim
2007-03-29 5:08 ` Linus Torvalds
2007-03-29 5:47 ` Maxim
2007-03-29 13:20 ` Sergei Shtylyov
2007-03-29 13:31 ` Maxim
2007-03-29 13:46 ` [PATCH v2] Add suspend/resume for HPET Maxim Levitsky
2007-03-29 16:53 ` Linus Torvalds
2007-03-29 17:28 ` Maxim Levitsky
2007-03-29 17:51 ` Ingo Molnar
2007-03-29 20:46 ` Andi Kleen
2007-03-29 18:11 ` Jeff Chua
2007-03-31 15:51 ` Thomas Gleixner
2007-03-31 16:01 ` Jeff Chua
2007-03-31 16:09 ` Thomas Gleixner
2007-03-31 16:09 ` Linus Torvalds
2007-03-31 16:33 ` Thomas Gleixner
2007-03-31 16:41 ` Greg KH
2007-03-31 16:53 ` Linus Torvalds
2007-03-31 17:02 ` Ingo Molnar
2007-03-31 18:18 ` [linux-pm] " David Brownell
2007-03-31 19:32 ` David Brownell
2007-04-01 3:13 ` Jeff Chua
2007-04-01 4:13 ` David Brownell
2007-03-31 17:08 ` Greg KH
2007-03-31 17:55 ` [linux-pm] " David Brownell
2007-03-31 16:56 ` Maxim Levitsky
2007-03-31 17:09 ` Linus Torvalds
2007-03-31 17:17 ` Ingo Molnar
2007-03-31 17:58 ` Daniel Walker
2007-03-29 16:35 ` [ PATCH] Add suspend/resume for HPET was: Re: [3/6] 2.6.21-rc4: known regressions Linus Torvalds
2007-03-29 16:51 ` Maxim Levitsky
2007-03-29 17:22 ` Linus Torvalds
2007-03-29 17:47 ` [patch, v2] add suspend/resume for HPET Ingo Molnar
2007-03-28 18:04 ` [3/6] 2.6.21-rc4: known regressions Michael S. Tsirkin
2007-03-28 18:32 ` Ingo Molnar
2007-03-28 18:35 ` Randy Dunlap
2007-03-29 14:24 ` Jeff Chua
2007-03-18 18:49 ` [4/6] " Adrian Bunk
2007-03-18 18:49 ` [5/6] " Adrian Bunk
2007-03-18 19:07 ` Maxim
2007-03-18 19:22 ` Adrian Bunk
2007-03-18 19:59 ` Maxim
2007-03-18 20:03 ` Maxim
2007-03-18 18:49 ` [6/6] " Adrian Bunk
2007-03-20 2:38 ` David Miller
2007-03-24 19:50 ` David Miller
2007-03-19 20:39 ` 2.6.21-rc4: known regressions with patches available Adrian Bunk
2007-03-20 11:02 ` [Alsa-devel] " Takashi Iwai
2007-03-23 18:48 ` [1/5] 2.6.21-rc4: known regressions (v2) Adrian Bunk
2007-03-25 4:45 ` David Miller
2007-03-25 5:08 ` Paul Collins
2007-03-25 12:22 ` Adrian Bunk
2007-03-23 18:48 ` [2/5] " Adrian Bunk
2007-03-23 21:08 ` Thomas Gleixner
2007-03-24 0:14 ` Ray Lee
2007-03-24 6:40 ` Thomas Gleixner
2007-03-24 18:17 ` Ray Lee
2007-03-24 19:11 ` [PATCH] x86_64: avoid sending LOCAL_TIMER_VECTOR IPI to itself Ingo Molnar
2007-03-25 19:24 ` Ray Lee
2007-03-26 10:01 ` [2/5] 2.6.21-rc4: known regressions (v2) Tejun Heo
2007-03-23 18:50 ` [3/5] " Adrian Bunk
2007-03-23 19:07 ` Maxim
2007-03-23 20:53 ` Rafael J. Wysocki
2007-03-24 17:04 ` Thomas Meyer
2007-03-24 18:02 ` Eric W. Biederman
2007-03-24 18:20 ` Thomas Meyer
2007-03-24 18:47 ` Eric W. Biederman
2007-03-24 20:34 ` Thomas Meyer
2007-03-25 3:39 ` Eric W. Biederman
2007-03-25 11:41 ` Thomas Meyer
2007-03-25 12:03 ` Eric W. Biederman
2007-03-25 12:28 ` Rafael J. Wysocki
2007-03-25 12:56 ` Eric W. Biederman
2007-03-25 19:14 ` Rafael J. Wysocki
2007-03-25 20:37 ` Eric W. Biederman
2007-03-26 21:03 ` Rafael J. Wysocki
2007-03-25 14:17 ` Thomas Meyer
2007-03-25 18:56 ` Rafael J. Wysocki
2007-03-25 13:54 ` Thomas Meyer
2007-03-25 14:48 ` Adrian Bunk
2007-03-25 17:25 ` Thomas Meyer
2007-03-25 19:06 ` Rafael J. Wysocki
2007-03-25 19:31 ` Rafael J. Wysocki
2007-03-26 20:01 ` Luck, Tony
2007-03-27 3:29 ` Eric W. Biederman
2007-04-02 15:38 ` Bjorn Helgaas
2007-04-02 16:38 ` Bjorn Helgaas
2007-04-02 19:50 ` Eric W. Biederman
2007-03-25 21:34 ` Frédéric Riss
2007-03-26 6:45 ` Frédéric RISS
2007-03-26 9:14 ` Thomas Gleixner
2007-03-26 10:36 ` Frederic Riss
2007-03-26 18:53 ` Frédéric Riss
2007-03-26 19:02 ` Adrian Bunk
2007-03-26 19:39 ` Frederic Riss
2007-03-26 19:46 ` Adrian Bunk
2007-03-26 10:00 ` Marcus Better
2007-03-26 12:35 ` Pavel Machek
2007-03-26 14:11 ` Marcus Better
2007-03-26 14:34 ` Adrian Bunk
2007-03-26 17:42 ` Marcus Better
2007-03-26 18:48 ` Adrian Bunk
2007-03-27 9:42 ` Marcus Better
2007-03-23 18:50 ` [4/5] " Adrian Bunk
2007-03-23 19:15 ` Thomas Gleixner
2007-03-23 19:15 ` Adrian Bunk
2007-03-23 19:21 ` Thomas Gleixner
2007-03-23 22:23 ` Chuck Ebbert
2007-03-23 22:43 ` Thomas Gleixner
2007-03-23 23:35 ` Thomas Gleixner
2007-03-25 12:42 ` [PATCH] clocksource: Fix thinko in watchdog selection Thomas Gleixner
2007-03-23 23:00 ` [4/5] 2.6.21-rc4: known regressions (v2) Adrian Bunk
2007-03-23 23:05 ` Chuck Ebbert
2007-03-23 19:22 ` Thomas Gleixner
2007-03-24 13:47 ` Thomas Gleixner
2007-03-25 12:31 ` [PATCH] dynticks: fix hrtimer rounding error in next_timer_interrupt Thomas Gleixner
2007-03-23 19:49 ` [4/5] 2.6.21-rc4: known regressions (v2) Thomas Gleixner
[not found] ` <20070325071023.GL17532@mellanox.co.il>
2007-03-25 7:37 ` Thomas Gleixner
2007-03-25 8:57 ` Michael S. Tsirkin
2007-03-25 10:17 ` Thomas Gleixner
2007-03-25 10:15 ` Michael S. Tsirkin
2007-03-25 10:27 ` Thomas Gleixner
2007-03-25 10:25 ` Michael S. Tsirkin
2007-03-25 10:38 ` Thomas Gleixner
2007-03-25 11:16 ` Ingo Molnar
2007-03-25 12:09 ` Thomas Gleixner
2007-03-26 14:19 ` Michael S. Tsirkin
2007-03-23 20:00 ` Thomas Gleixner
2007-03-23 20:08 ` Thomas Gleixner
2007-03-24 13:59 ` Michal Piotrowski
2007-03-24 15:14 ` Thomas Gleixner
2007-03-24 16:13 ` Michal Piotrowski
2007-03-23 21:43 ` john stultz
2007-03-23 21:54 ` Linus Torvalds
2007-03-24 0:44 ` john stultz
2007-03-23 18:50 ` [5/5] " Adrian Bunk
2007-03-24 11:25 ` 2.6.21-rc4: known regressions with patches (v2) Adrian Bunk
2007-03-26 12:37 ` Bob Tracy
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1174087552.13341.332.camel@localhost.localdomain \
--to=tglx@linutronix.de \
--cc=akpm@linux-foundation.org \
--cc=bunk@stusta.de \
--cc=frederic.riss@gmail.com \
--cc=linux-kernel@vger.kernel.org \
--cc=mbroz@redhat.com \
--cc=mingo@elte.hu \
--cc=rjw@sisk.pl \
--cc=thomas@m3y3r.de \
--cc=torvalds@linux-foundation.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox