* [PATCH RT 1/8] genirq: Unmask oneshot irqs when thread was not woken
2012-02-09 1:59 [PATCH RT 0/8] [ANNOUNCE] 3.0.20-rt36-rc1 Steven Rostedt
@ 2012-02-09 1:59 ` Steven Rostedt
2012-02-09 1:59 ` [PATCH RT 2/8] genirq: Handle pending irqs in irq_startup() Steven Rostedt
` (6 subsequent siblings)
7 siblings, 0 replies; 9+ messages in thread
From: Steven Rostedt @ 2012-02-09 1:59 UTC (permalink / raw)
To: linux-kernel, linux-rt-users
Cc: Thomas Gleixner, Carsten Emde, John Kacur, stable-rt, Lothar Wassmann
[-- Attachment #1: 0001-genirq-Unmask-oneshot-irqs-when-thread-was-not-woken.patch --]
[-- Type: text/plain, Size: 1960 bytes --]
From: Thomas Gleixner <tglx@linutronix.de>
When the primary handler of an interrupt which is marked IRQ_ONESHOT
does not wake the thread (e.g. due to a spurious interrupt) then
nothing ever unmasks the interrupt line again. Handle it gracefully.
Cc: stable-rt@vger.kernel.org
Reported-by: Lothar Wassmann <lw@karo-electronics.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
---
kernel/irq/chip.c | 25 +++++++++++++++++++++++--
1 files changed, 23 insertions(+), 2 deletions(-)
diff --git a/kernel/irq/chip.c b/kernel/irq/chip.c
index dc5114b..3322a34 100644
--- a/kernel/irq/chip.c
+++ b/kernel/irq/chip.c
@@ -312,6 +312,24 @@ out_unlock:
}
EXPORT_SYMBOL_GPL(handle_simple_irq);
+/*
+ * Called unconditionally from handle_level_irq() and only for oneshot
+ * interrupts from handle_fasteoi_irq()
+ */
+static void cond_unmask_irq(struct irq_desc *desc)
+{
+ /*
+ * We need to unmask in the following cases:
+ * - Standard level irq (IRQF_ONESHOT is not set)
+ * - Oneshot irq which did not wake the thread (caused by a
+ * spurious interrupt or a primary handler handling it
+ * completely).
+ */
+ if (!irqd_irq_disabled(&desc->irq_data) &&
+ irqd_irq_masked(&desc->irq_data) && !desc->threads_oneshot)
+ unmask_irq(desc);
+}
+
/**
* handle_level_irq - Level type irq handler
* @irq: the interrupt number
@@ -344,8 +362,8 @@ handle_level_irq(unsigned int irq, struct irq_desc *desc)
handle_irq_event(desc);
- if (!irqd_irq_disabled(&desc->irq_data) && !(desc->istate & IRQS_ONESHOT))
- unmask_irq(desc);
+ cond_unmask_irq(desc);
+
out_unlock:
raw_spin_unlock(&desc->lock);
}
@@ -399,6 +417,9 @@ handle_fasteoi_irq(unsigned int irq, struct irq_desc *desc)
preflow_handler(desc);
handle_irq_event(desc);
+ if (desc->istate & IRQS_ONESHOT)
+ cond_unmask_irq(desc);
+
out_eoi:
desc->irq_data.chip->irq_eoi(&desc->irq_data);
out_unlock:
--
1.7.8.3
^ permalink raw reply related [flat|nested] 9+ messages in thread* [PATCH RT 2/8] genirq: Handle pending irqs in irq_startup()
2012-02-09 1:59 [PATCH RT 0/8] [ANNOUNCE] 3.0.20-rt36-rc1 Steven Rostedt
2012-02-09 1:59 ` [PATCH RT 1/8] genirq: Unmask oneshot irqs when thread was not woken Steven Rostedt
@ 2012-02-09 1:59 ` Steven Rostedt
2012-02-09 1:59 ` [PATCH RT 3/8] signal/x86: Delay calling signals in atomic Steven Rostedt
` (5 subsequent siblings)
7 siblings, 0 replies; 9+ messages in thread
From: Steven Rostedt @ 2012-02-09 1:59 UTC (permalink / raw)
To: linux-kernel, linux-rt-users
Cc: Thomas Gleixner, Carsten Emde, John Kacur, stable-rt
[-- Attachment #1: 0002-genirq-Handle-pending-irqs-in-irq_startup.patch --]
[-- Type: text/plain, Size: 3585 bytes --]
From: Thomas Gleixner <tglx@linutronix.de>
Irqs might be pending when irq_startup() is called. Call the resend
function in irq_startup() otherwise the irq line might be blocked
forever.
Cc: stable-rt@vger.kernel.org
Reported-and-tested-by: Russell King <rmk+kernel@arm.linux.org.uk>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
---
kernel/irq/autoprobe.c | 4 ++--
kernel/irq/chip.c | 17 ++++++++++-------
kernel/irq/internals.h | 2 +-
kernel/irq/manage.c | 2 +-
4 files changed, 14 insertions(+), 11 deletions(-)
diff --git a/kernel/irq/autoprobe.c b/kernel/irq/autoprobe.c
index 342d8f4..0119b9d 100644
--- a/kernel/irq/autoprobe.c
+++ b/kernel/irq/autoprobe.c
@@ -53,7 +53,7 @@ unsigned long probe_irq_on(void)
if (desc->irq_data.chip->irq_set_type)
desc->irq_data.chip->irq_set_type(&desc->irq_data,
IRQ_TYPE_PROBE);
- irq_startup(desc);
+ irq_startup(desc, false);
}
raw_spin_unlock_irq(&desc->lock);
}
@@ -70,7 +70,7 @@ unsigned long probe_irq_on(void)
raw_spin_lock_irq(&desc->lock);
if (!desc->action && irq_settings_can_probe(desc)) {
desc->istate |= IRQS_AUTODETECT | IRQS_WAITING;
- if (irq_startup(desc))
+ if (irq_startup(desc, false))
desc->istate |= IRQS_PENDING;
}
raw_spin_unlock_irq(&desc->lock);
diff --git a/kernel/irq/chip.c b/kernel/irq/chip.c
index 3322a34..ca14f5d 100644
--- a/kernel/irq/chip.c
+++ b/kernel/irq/chip.c
@@ -157,19 +157,22 @@ static void irq_state_set_masked(struct irq_desc *desc)
irqd_set(&desc->irq_data, IRQD_IRQ_MASKED);
}
-int irq_startup(struct irq_desc *desc)
+int irq_startup(struct irq_desc *desc, bool resend)
{
+ int ret = 0;
+
irq_state_clr_disabled(desc);
desc->depth = 0;
if (desc->irq_data.chip->irq_startup) {
- int ret = desc->irq_data.chip->irq_startup(&desc->irq_data);
+ ret = desc->irq_data.chip->irq_startup(&desc->irq_data);
irq_state_clr_masked(desc);
- return ret;
+ } else {
+ irq_enable(desc);
}
-
- irq_enable(desc);
- return 0;
+ if (resend)
+ check_irq_resend(desc, desc->irq_data.irq);
+ return ret;
}
void irq_shutdown(struct irq_desc *desc)
@@ -596,7 +599,7 @@ __irq_set_handler(unsigned int irq, irq_flow_handler_t handle, int is_chained,
irq_settings_set_noprobe(desc);
irq_settings_set_norequest(desc);
irq_settings_set_nothread(desc);
- irq_startup(desc);
+ irq_startup(desc, true);
}
out:
irq_put_desc_busunlock(desc, flags);
diff --git a/kernel/irq/internals.h b/kernel/irq/internals.h
index 6546431..62efdc4 100644
--- a/kernel/irq/internals.h
+++ b/kernel/irq/internals.h
@@ -67,7 +67,7 @@ extern int __irq_set_trigger(struct irq_desc *desc, unsigned int irq,
extern void __disable_irq(struct irq_desc *desc, unsigned int irq, bool susp);
extern void __enable_irq(struct irq_desc *desc, unsigned int irq, bool resume);
-extern int irq_startup(struct irq_desc *desc);
+extern int irq_startup(struct irq_desc *desc, bool resend);
extern void irq_shutdown(struct irq_desc *desc);
extern void irq_enable(struct irq_desc *desc);
extern void irq_disable(struct irq_desc *desc);
diff --git a/kernel/irq/manage.c b/kernel/irq/manage.c
index e1bd49d..3eb233e 100644
--- a/kernel/irq/manage.c
+++ b/kernel/irq/manage.c
@@ -1020,7 +1020,7 @@ __setup_irq(unsigned int irq, struct irq_desc *desc, struct irqaction *new)
desc->istate |= IRQS_ONESHOT;
if (irq_settings_can_autoenable(desc))
- irq_startup(desc);
+ irq_startup(desc, true);
else
/* Undo nested disables: */
desc->depth = 1;
--
1.7.8.3
^ permalink raw reply related [flat|nested] 9+ messages in thread* [PATCH RT 3/8] signal/x86: Delay calling signals in atomic
2012-02-09 1:59 [PATCH RT 0/8] [ANNOUNCE] 3.0.20-rt36-rc1 Steven Rostedt
2012-02-09 1:59 ` [PATCH RT 1/8] genirq: Unmask oneshot irqs when thread was not woken Steven Rostedt
2012-02-09 1:59 ` [PATCH RT 2/8] genirq: Handle pending irqs in irq_startup() Steven Rostedt
@ 2012-02-09 1:59 ` Steven Rostedt
2012-02-09 1:59 ` [PATCH RT 4/8] x86: Do not disable preemption in int3 on 32bit Steven Rostedt
` (4 subsequent siblings)
7 siblings, 0 replies; 9+ messages in thread
From: Steven Rostedt @ 2012-02-09 1:59 UTC (permalink / raw)
To: linux-kernel, linux-rt-users
Cc: Thomas Gleixner, Carsten Emde, John Kacur, stable-rt, Oleg Nesterov
[-- Attachment #1: 0003-signal-x86-Delay-calling-signals-in-atomic.patch --]
[-- Type: text/plain, Size: 5221 bytes --]
From: Oleg Nesterov <oleg@redhat.com>
On x86_64 we must disable preemption before we enable interrupts
for stack faults, int3 and debugging, because the current task is using
a per CPU debug stack defined by the IST. If we schedule out, another task
can come in and use the same stack and cause the stack to be corrupted
and crash the kernel on return.
When CONFIG_PREEMPT_RT_FULL is enabled, spin_locks become mutexes, and
one of these is the spin lock used in signal handling.
Some of the debug code (int3) causes do_trap() to send a signal.
This function calls a spin lock that has been converted to a mutex
and has the possibility to sleep. If this happens, the above issues with
the corrupted stack is possible.
Instead of calling the signal right away, for PREEMPT_RT and x86_64,
the signal information is stored on the stacks task_struct and
TIF_NOTIFY_RESUME is set. Then on exit of the trap, the signal resume
code will send the signal when preemption is enabled.
[ rostedt: Switched from #ifdef CONFIG_PREEMPT_RT_FULL to
ARCH_RT_DELAYS_SIGNAL_SEND and added comments to the code. ]
Cc: stable-rt@vger.kernel.org
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
---
arch/x86/include/asm/signal.h | 13 +++++++++++++
arch/x86/kernel/signal.c | 9 +++++++++
include/linux/sched.h | 4 ++++
kernel/signal.c | 37 +++++++++++++++++++++++++++++++++++--
4 files changed, 61 insertions(+), 2 deletions(-)
diff --git a/arch/x86/include/asm/signal.h b/arch/x86/include/asm/signal.h
index 598457c..1213ebd 100644
--- a/arch/x86/include/asm/signal.h
+++ b/arch/x86/include/asm/signal.h
@@ -31,6 +31,19 @@ typedef struct {
unsigned long sig[_NSIG_WORDS];
} sigset_t;
+/*
+ * Because some traps use the IST stack, we must keep
+ * preemption disabled while calling do_trap(), but do_trap()
+ * may call force_sig_info() which will grab the signal spin_locks
+ * for the task, which in PREEMPT_RT_FULL are mutexes.
+ * By defining ARCH_RT_DELAYS_SIGNAL_SEND the force_sig_info() will
+ * set TIF_NOTIFY_RESUME and set up the signal to be sent on exit
+ * of the trap.
+ */
+#if defined(CONFIG_PREEMPT_RT_FULL) && defined(CONFIG_X86_64)
+#define ARCH_RT_DELAYS_SIGNAL_SEND
+#endif
+
#else
/* Here we must cater to libcs that poke about in kernel headers. */
diff --git a/arch/x86/kernel/signal.c b/arch/x86/kernel/signal.c
index 40a2493..7d88255 100644
--- a/arch/x86/kernel/signal.c
+++ b/arch/x86/kernel/signal.c
@@ -840,6 +840,15 @@ do_notify_resume(struct pt_regs *regs, void *unused, __u32 thread_info_flags)
mce_notify_process();
#endif /* CONFIG_X86_64 && CONFIG_X86_MCE */
+#ifdef ARCH_RT_DELAYS_SIGNAL_SEND
+ if (unlikely(current->forced_info.si_signo)) {
+ struct task_struct *t = current;
+ force_sig_info(t->forced_info.si_signo,
+ &t->forced_info, t);
+ t->forced_info.si_signo = 0;
+ }
+#endif
+
/* deal with pending signal delivery */
if (thread_info_flags & _TIF_SIGPENDING)
do_signal(regs);
diff --git a/include/linux/sched.h b/include/linux/sched.h
index 4cd3278..77e132f 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -1405,6 +1405,10 @@ struct task_struct {
sigset_t blocked, real_blocked;
sigset_t saved_sigmask; /* restored if set_restore_sigmask() was used */
struct sigpending pending;
+#ifdef CONFIG_PREEMPT_RT_FULL
+ /* TODO: move me into ->restart_block ? */
+ struct siginfo forced_info;
+#endif
unsigned long sas_ss_sp;
size_t sas_ss_size;
diff --git a/kernel/signal.c b/kernel/signal.c
index 3f15914..6650b2c 100644
--- a/kernel/signal.c
+++ b/kernel/signal.c
@@ -1200,8 +1200,8 @@ int do_send_sig_info(int sig, struct siginfo *info, struct task_struct *p,
* We don't want to have recursive SIGSEGV's etc, for example,
* that is why we also clear SIGNAL_UNKILLABLE.
*/
-int
-force_sig_info(int sig, struct siginfo *info, struct task_struct *t)
+static int
+do_force_sig_info(int sig, struct siginfo *info, struct task_struct *t)
{
unsigned long int flags;
int ret, blocked, ignored;
@@ -1226,6 +1226,39 @@ force_sig_info(int sig, struct siginfo *info, struct task_struct *t)
return ret;
}
+int force_sig_info(int sig, struct siginfo *info, struct task_struct *t)
+{
+/*
+ * On some archs, PREEMPT_RT has to delay sending a signal from a trap
+ * since it can not enable preemption, and the signal code's spin_locks
+ * turn into mutexes. Instead, it must set TIF_NOTIFY_RESUME which will
+ * send the signal on exit of the trap.
+ */
+#ifdef ARCH_RT_DELAYS_SIGNAL_SEND
+ if (in_atomic()) {
+ if (WARN_ON_ONCE(t != current))
+ return 0;
+ if (WARN_ON_ONCE(t->forced_info.si_signo))
+ return 0;
+
+ if (is_si_special(info)) {
+ WARN_ON_ONCE(info != SEND_SIG_PRIV);
+ t->forced_info.si_signo = sig;
+ t->forced_info.si_errno = 0;
+ t->forced_info.si_code = SI_KERNEL;
+ t->forced_info.si_pid = 0;
+ t->forced_info.si_uid = 0;
+ } else {
+ t->forced_info = *info;
+ }
+
+ set_tsk_thread_flag(t, TIF_NOTIFY_RESUME);
+ return 0;
+ }
+#endif
+ return do_force_sig_info(sig, info, t);
+}
+
/*
* Nuke all other threads in the group.
*/
--
1.7.8.3
^ permalink raw reply related [flat|nested] 9+ messages in thread* [PATCH RT 4/8] x86: Do not disable preemption in int3 on 32bit
2012-02-09 1:59 [PATCH RT 0/8] [ANNOUNCE] 3.0.20-rt36-rc1 Steven Rostedt
` (2 preceding siblings ...)
2012-02-09 1:59 ` [PATCH RT 3/8] signal/x86: Delay calling signals in atomic Steven Rostedt
@ 2012-02-09 1:59 ` Steven Rostedt
2012-02-09 1:59 ` [PATCH RT 5/8] timer-fd: Prevent live lock Steven Rostedt
` (3 subsequent siblings)
7 siblings, 0 replies; 9+ messages in thread
From: Steven Rostedt @ 2012-02-09 1:59 UTC (permalink / raw)
To: linux-kernel, linux-rt-users
Cc: Thomas Gleixner, Carsten Emde, John Kacur, stable-rt
[-- Attachment #1: 0004-x86-Do-not-disable-preemption-in-int3-on-32bit.patch --]
[-- Type: text/plain, Size: 3691 bytes --]
From: Steven Rostedt <rostedt@goodmis.org>
Preemption must be disabled before enabling interrupts in do_trap
on x86_64 because the stack in use for int3 and debug is a per CPU
stack set by th IST. But 32bit does not have an IST and the stack
still belongs to the current task and there is no problem in scheduling
out the task.
Keep preemption enabled on X86_32 when enabling interrupts for
do_trap().
The name of the function is changed from preempt_conditional_sti/cli()
to conditional_sti/cli_ist(), to annotate that this function is used
when the stack is on the IST.
Cc: stable-rt@vger.kernel.org
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
---
arch/x86/kernel/traps.c | 32 +++++++++++++++++++++++---------
1 files changed, 23 insertions(+), 9 deletions(-)
diff --git a/arch/x86/kernel/traps.c b/arch/x86/kernel/traps.c
index b9b6716..5d47242 100644
--- a/arch/x86/kernel/traps.c
+++ b/arch/x86/kernel/traps.c
@@ -96,9 +96,21 @@ static inline void conditional_sti(struct pt_regs *regs)
local_irq_enable();
}
-static inline void preempt_conditional_sti(struct pt_regs *regs)
+static inline void conditional_sti_ist(struct pt_regs *regs)
{
+#ifdef CONFIG_X86_64
+ /*
+ * X86_64 uses a per CPU stack on the IST for certain traps
+ * like int3. The task can not be preempted when using one
+ * of these stacks, thus preemption must be disabled, otherwise
+ * the stack can be corrupted if the task is scheduled out,
+ * and another task comes in and uses this stack.
+ *
+ * On x86_32 the task keeps its own stack and it is OK if the
+ * task schedules out.
+ */
inc_preempt_count();
+#endif
if (regs->flags & X86_EFLAGS_IF)
local_irq_enable();
}
@@ -109,11 +121,13 @@ static inline void conditional_cli(struct pt_regs *regs)
local_irq_disable();
}
-static inline void preempt_conditional_cli(struct pt_regs *regs)
+static inline void conditional_cli_ist(struct pt_regs *regs)
{
if (regs->flags & X86_EFLAGS_IF)
local_irq_disable();
+#ifdef CONFIG_X86_64
dec_preempt_count();
+#endif
}
static void __kprobes
@@ -231,9 +245,9 @@ dotraplinkage void do_stack_segment(struct pt_regs *regs, long error_code)
if (notify_die(DIE_TRAP, "stack segment", regs, error_code,
12, SIGBUS) == NOTIFY_STOP)
return;
- preempt_conditional_sti(regs);
+ conditional_sti_ist(regs);
do_trap(12, SIGBUS, "stack segment", regs, error_code, NULL);
- preempt_conditional_cli(regs);
+ conditional_cli_ist(regs);
}
dotraplinkage void do_double_fault(struct pt_regs *regs, long error_code)
@@ -471,9 +485,9 @@ dotraplinkage void __kprobes do_int3(struct pt_regs *regs, long error_code)
return;
#endif
- preempt_conditional_sti(regs);
+ conditional_sti_ist(regs);
do_trap(3, SIGTRAP, "int3", regs, error_code, NULL);
- preempt_conditional_cli(regs);
+ conditional_cli_ist(regs);
}
#ifdef CONFIG_X86_64
@@ -567,12 +581,12 @@ dotraplinkage void __kprobes do_debug(struct pt_regs *regs, long error_code)
return;
/* It's safe to allow irq's after DR6 has been saved */
- preempt_conditional_sti(regs);
+ conditional_sti_ist(regs);
if (regs->flags & X86_VM_MASK) {
handle_vm86_trap((struct kernel_vm86_regs *) regs,
error_code, 1);
- preempt_conditional_cli(regs);
+ conditional_cli_ist(regs);
return;
}
@@ -591,7 +605,7 @@ dotraplinkage void __kprobes do_debug(struct pt_regs *regs, long error_code)
si_code = get_si_code(tsk->thread.debugreg6);
if (tsk->thread.debugreg6 & (DR_STEP | DR_TRAP_BITS) || user_icebp)
send_sigtrap(tsk, regs, error_code, si_code);
- preempt_conditional_cli(regs);
+ conditional_cli_ist(regs);
return;
}
--
1.7.8.3
^ permalink raw reply related [flat|nested] 9+ messages in thread* [PATCH RT 5/8] timer-fd: Prevent live lock
2012-02-09 1:59 [PATCH RT 0/8] [ANNOUNCE] 3.0.20-rt36-rc1 Steven Rostedt
` (3 preceding siblings ...)
2012-02-09 1:59 ` [PATCH RT 4/8] x86: Do not disable preemption in int3 on 32bit Steven Rostedt
@ 2012-02-09 1:59 ` Steven Rostedt
2012-02-09 1:59 ` [PATCH RT 6/8] genirq: Allow disabling of softirq processing in irq thread context Steven Rostedt
` (2 subsequent siblings)
7 siblings, 0 replies; 9+ messages in thread
From: Steven Rostedt @ 2012-02-09 1:59 UTC (permalink / raw)
To: linux-kernel, linux-rt-users
Cc: Thomas Gleixner, Carsten Emde, John Kacur, Sankara Muthukrishnan,
stable-rt
[-- Attachment #1: 0005-timer-fd-Prevent-live-lock.patch --]
[-- Type: text/plain, Size: 893 bytes --]
From: Thomas Gleixner <tglx@linutronix.de>
If hrtimer_try_to_cancel() requires a retry, then depending on the
priority setting te retry loop might prevent timer callback completion
on RT. Prevent that by waiting for completion on RT, no change for a
non RT kernel.
Reported-by: Sankara Muthukrishnan <sankara.m@gmail.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: stable-rt@vger.kernel.org
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
---
fs/timerfd.c | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/fs/timerfd.c b/fs/timerfd.c
index dffeb37..57f0e4e 100644
--- a/fs/timerfd.c
+++ b/fs/timerfd.c
@@ -313,7 +313,7 @@ SYSCALL_DEFINE4(timerfd_settime, int, ufd, int, flags,
if (hrtimer_try_to_cancel(&ctx->tmr) >= 0)
break;
spin_unlock_irq(&ctx->wqh.lock);
- cpu_relax();
+ hrtimer_wait_for_timer(&ctx->tmr);
}
/*
--
1.7.8.3
^ permalink raw reply related [flat|nested] 9+ messages in thread* [PATCH RT 6/8] genirq: Allow disabling of softirq processing in irq thread context
2012-02-09 1:59 [PATCH RT 0/8] [ANNOUNCE] 3.0.20-rt36-rc1 Steven Rostedt
` (4 preceding siblings ...)
2012-02-09 1:59 ` [PATCH RT 5/8] timer-fd: Prevent live lock Steven Rostedt
@ 2012-02-09 1:59 ` Steven Rostedt
2012-02-09 1:59 ` [PATCH RT 7/8] futex: Fix bug on when a requeued RT task times out Steven Rostedt
2012-02-09 1:59 ` [PATCH RT 8/8] Linux 3.0.20-rt36-rc1 Steven Rostedt
7 siblings, 0 replies; 9+ messages in thread
From: Steven Rostedt @ 2012-02-09 1:59 UTC (permalink / raw)
To: linux-kernel, linux-rt-users
Cc: Thomas Gleixner, Carsten Emde, John Kacur, stable-rt
[-- Attachment #1: 0006-genirq-Allow-disabling-of-softirq-processing-in-irq-.patch --]
[-- Type: text/plain, Size: 5119 bytes --]
From: Thomas Gleixner <tglx@linutronix.de>
The processing of softirqs in irq thread context is a performance gain
for the non-rt workloads of a system, but it's counterproductive for
interrupts which are explicitely related to the realtime
workload. Allow such interrupts to prevent softirq processing in their
thread context.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: stable-rt@vger.kernel.org
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
---
include/linux/interrupt.h | 2 ++
include/linux/irq.h | 4 +++-
kernel/irq/manage.c | 13 ++++++++++++-
kernel/irq/settings.h | 12 ++++++++++++
kernel/softirq.c | 7 +++++++
5 files changed, 36 insertions(+), 2 deletions(-)
diff --git a/include/linux/interrupt.h b/include/linux/interrupt.h
index 3142442..71c2c0b 100644
--- a/include/linux/interrupt.h
+++ b/include/linux/interrupt.h
@@ -61,6 +61,7 @@
* IRQF_NO_THREAD - Interrupt cannot be threaded
* IRQF_EARLY_RESUME - Resume IRQ early during syscore instead of at device
* resume time.
+ * IRQF_NO_SOFTIRQ_CALL - Do not process softirqs in the irq thread context (RT)
*/
#define IRQF_DISABLED 0x00000020
#define IRQF_SAMPLE_RANDOM 0x00000040
@@ -75,6 +76,7 @@
#define IRQF_FORCE_RESUME 0x00008000
#define IRQF_NO_THREAD 0x00010000
#define IRQF_EARLY_RESUME 0x00020000
+#define IRQF_NO_SOFTIRQ_CALL 0x00040000
#define IRQF_TIMER (__IRQF_TIMER | IRQF_NO_SUSPEND | IRQF_NO_THREAD)
diff --git a/include/linux/irq.h b/include/linux/irq.h
index baa397e..2df0acc 100644
--- a/include/linux/irq.h
+++ b/include/linux/irq.h
@@ -65,6 +65,7 @@ typedef void (*irq_preflow_handler_t)(struct irq_data *data);
* IRQ_NO_BALANCING - Interrupt cannot be balanced (affinity set)
* IRQ_MOVE_PCNTXT - Interrupt can be migrated from process context
* IRQ_NESTED_TRHEAD - Interrupt nests into another thread
+ * IRQ_NO_SOFTIRQ_CALL - No softirq processing in the irq thread context (RT)
*/
enum {
IRQ_TYPE_NONE = 0x00000000,
@@ -87,12 +88,13 @@ enum {
IRQ_MOVE_PCNTXT = (1 << 14),
IRQ_NESTED_THREAD = (1 << 15),
IRQ_NOTHREAD = (1 << 16),
+ IRQ_NO_SOFTIRQ_CALL = (1 << 18),
};
#define IRQF_MODIFY_MASK \
(IRQ_TYPE_SENSE_MASK | IRQ_NOPROBE | IRQ_NOREQUEST | \
IRQ_NOAUTOEN | IRQ_MOVE_PCNTXT | IRQ_LEVEL | IRQ_NO_BALANCING | \
- IRQ_PER_CPU | IRQ_NESTED_THREAD)
+ IRQ_PER_CPU | IRQ_NESTED_THREAD | IRQ_NO_SOFTIRQ_CALL)
#define IRQ_NO_BALANCING_MASK (IRQ_PER_CPU | IRQ_NO_BALANCING)
diff --git a/kernel/irq/manage.c b/kernel/irq/manage.c
index 3eb233e..cd98592 100644
--- a/kernel/irq/manage.c
+++ b/kernel/irq/manage.c
@@ -739,7 +739,15 @@ irq_forced_thread_fn(struct irq_desc *desc, struct irqaction *action)
local_bh_disable();
ret = action->thread_fn(action->irq, action->dev_id);
irq_finalize_oneshot(desc, action, false);
- local_bh_enable();
+ /*
+ * Interrupts which have real time requirements can be set up
+ * to avoid softirq processing in the thread handler. This is
+ * safe as these interrupts do not raise soft interrupts.
+ */
+ if (irq_settings_no_softirq_call(desc))
+ _local_bh_enable();
+ else
+ local_bh_enable();
return ret;
}
@@ -1031,6 +1039,9 @@ __setup_irq(unsigned int irq, struct irq_desc *desc, struct irqaction *new)
irqd_set(&desc->irq_data, IRQD_NO_BALANCING);
}
+ if (new->flags & IRQF_NO_SOFTIRQ_CALL)
+ irq_settings_set_no_softirq_call(desc);
+
/* Set default affinity mask once everything is setup */
setup_affinity(irq, desc, mask);
diff --git a/kernel/irq/settings.h b/kernel/irq/settings.h
index f166783..e1f617f 100644
--- a/kernel/irq/settings.h
+++ b/kernel/irq/settings.h
@@ -13,6 +13,7 @@ enum {
_IRQ_MOVE_PCNTXT = IRQ_MOVE_PCNTXT,
_IRQ_NO_BALANCING = IRQ_NO_BALANCING,
_IRQ_NESTED_THREAD = IRQ_NESTED_THREAD,
+ _IRQ_NO_SOFTIRQ_CALL = IRQ_NO_SOFTIRQ_CALL,
_IRQF_MODIFY_MASK = IRQF_MODIFY_MASK,
};
@@ -24,6 +25,7 @@ enum {
#define IRQ_NOTHREAD GOT_YOU_MORON
#define IRQ_NOAUTOEN GOT_YOU_MORON
#define IRQ_NESTED_THREAD GOT_YOU_MORON
+#define IRQ_NO_SOFTIRQ_CALL GOT_YOU_MORON
#undef IRQF_MODIFY_MASK
#define IRQF_MODIFY_MASK GOT_YOU_MORON
@@ -34,6 +36,16 @@ irq_settings_clr_and_set(struct irq_desc *desc, u32 clr, u32 set)
desc->status_use_accessors |= (set & _IRQF_MODIFY_MASK);
}
+static inline bool irq_settings_no_softirq_call(struct irq_desc *desc)
+{
+ return desc->status_use_accessors & _IRQ_NO_SOFTIRQ_CALL;
+}
+
+static inline void irq_settings_set_no_softirq_call(struct irq_desc *desc)
+{
+ desc->status_use_accessors |= _IRQ_NO_SOFTIRQ_CALL;
+}
+
static inline bool irq_settings_is_per_cpu(struct irq_desc *desc)
{
return desc->status_use_accessors & _IRQ_PER_CPU;
diff --git a/kernel/softirq.c b/kernel/softirq.c
index 3489d06..729cd0c 100644
--- a/kernel/softirq.c
+++ b/kernel/softirq.c
@@ -431,6 +431,13 @@ void local_bh_enable_ip(unsigned long ip)
}
EXPORT_SYMBOL(local_bh_enable_ip);
+void _local_bh_enable(void)
+{
+ current->softirq_nestcnt--;
+ migrate_enable();
+}
+EXPORT_SYMBOL(_local_bh_enable);
+
/* For tracing */
int notrace __in_softirq(void)
{
--
1.7.8.3
^ permalink raw reply related [flat|nested] 9+ messages in thread* [PATCH RT 7/8] futex: Fix bug on when a requeued RT task times out
2012-02-09 1:59 [PATCH RT 0/8] [ANNOUNCE] 3.0.20-rt36-rc1 Steven Rostedt
` (5 preceding siblings ...)
2012-02-09 1:59 ` [PATCH RT 6/8] genirq: Allow disabling of softirq processing in irq thread context Steven Rostedt
@ 2012-02-09 1:59 ` Steven Rostedt
2012-02-09 1:59 ` [PATCH RT 8/8] Linux 3.0.20-rt36-rc1 Steven Rostedt
7 siblings, 0 replies; 9+ messages in thread
From: Steven Rostedt @ 2012-02-09 1:59 UTC (permalink / raw)
To: linux-kernel, linux-rt-users
Cc: Thomas Gleixner, Carsten Emde, John Kacur, stable-rt
[-- Attachment #1: 0007-futex-Fix-bug-on-when-a-requeued-RT-task-times-out.patch --]
[-- Type: text/plain, Size: 3978 bytes --]
From: Steven Rostedt <rostedt@goodmis.org>
Requeue with timeout causes a bug with PREEMPT_RT_FULL.
The bug comes from a timed out condition.
TASK 1 TASK 2
------ ------
futex_wait_requeue_pi()
futex_wait_queue_me()
<timed out>
double_lock_hb();
raw_spin_lock(pi_lock);
if (current->pi_blocked_on) {
} else {
current->pi_blocked_on = PI_WAKE_INPROGRESS;
run_spin_unlock(pi_lock);
spin_lock(hb->lock); <-- blocked!
plist_for_each_entry_safe(this) {
rt_mutex_start_proxy_lock();
task_blocks_on_rt_mutex();
BUG_ON(task->pi_blocked_on)!!!!
The BUG_ON() actually has a check for PI_WAKE_INPROGRESS, but the
problem is that, after TASK 1 sets PI_WAKE_INPROGRESS, it then tries to
grab the hb->lock, which it fails to do so. As the hb->lock is a mutex,
it will block and set the "pi_blocked_on" to the hb->lock.
When TASK 2 goes to requeue it, the check for PI_WAKE_INPROGESS fails
because the task1's pi_blocked_on is no longer set to that, but instead,
set to the hb->lock.
The fix:
When calling rt_mutex_start_proxy_lock() a check is made to see
if the proxy tasks pi_blocked_on is set. If so, exit out early.
Otherwise set it to a new flag PI_REQUEUE_INPROGRESS, which notifies
the proxy task that it is being requeued, and will handle things
appropriately.
Cc: stable-rt@vger.kernel.org
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
---
kernel/rtmutex.c | 32 +++++++++++++++++++++++++++++++-
kernel/rtmutex_common.h | 1 +
2 files changed, 32 insertions(+), 1 deletions(-)
diff --git a/kernel/rtmutex.c b/kernel/rtmutex.c
index 15eaf30..789744a 100644
--- a/kernel/rtmutex.c
+++ b/kernel/rtmutex.c
@@ -75,7 +75,8 @@ static void fixup_rt_mutex_waiters(struct rt_mutex *lock)
static int rt_mutex_real_waiter(struct rt_mutex_waiter *waiter)
{
- return waiter && waiter != PI_WAKEUP_INPROGRESS;
+ return waiter && waiter != PI_WAKEUP_INPROGRESS &&
+ waiter != PI_REQUEUE_INPROGRESS;
}
/*
@@ -1345,6 +1346,35 @@ int rt_mutex_start_proxy_lock(struct rt_mutex *lock,
return 1;
}
+#ifdef CONFIG_PREEMPT_RT_FULL
+ /*
+ * In PREEMPT_RT there's an added race.
+ * If the task, that we are about to requeue, times out,
+ * it can set the PI_WAKEUP_INPROGRESS. This tells the requeue
+ * to skip this task. But right after the task sets
+ * its pi_blocked_on to PI_WAKEUP_INPROGRESS it can then
+ * block on the spin_lock(&hb->lock), which in RT is an rtmutex.
+ * This will replace the PI_WAKEUP_INPROGRESS with the actual
+ * lock that it blocks on. We *must not* place this task
+ * on this proxy lock in that case.
+ *
+ * To prevent this race, we first take the task's pi_lock
+ * and check if it has updated its pi_blocked_on. If it has,
+ * we assume that it woke up and we return -EAGAIN.
+ * Otherwise, we set the task's pi_blocked_on to
+ * PI_REQUEUE_INPROGRESS, so that if the task is waking up
+ * it will know that we are in the process of requeuing it.
+ */
+ raw_spin_lock(&task->pi_lock);
+ if (task->pi_blocked_on) {
+ raw_spin_unlock(&task->pi_lock);
+ raw_spin_unlock(&lock->wait_lock);
+ return -EAGAIN;
+ }
+ task->pi_blocked_on = PI_REQUEUE_INPROGRESS;
+ raw_spin_unlock(&task->pi_lock);
+#endif
+
ret = task_blocks_on_rt_mutex(lock, waiter, task, detect_deadlock);
if (ret && !rt_mutex_owner(lock)) {
diff --git a/kernel/rtmutex_common.h b/kernel/rtmutex_common.h
index a688a29..6ec3dc1 100644
--- a/kernel/rtmutex_common.h
+++ b/kernel/rtmutex_common.h
@@ -105,6 +105,7 @@ static inline struct task_struct *rt_mutex_owner(struct rt_mutex *lock)
* PI-futex support (proxy locking functions, etc.):
*/
#define PI_WAKEUP_INPROGRESS ((struct rt_mutex_waiter *) 1)
+#define PI_REQUEUE_INPROGRESS ((struct rt_mutex_waiter *) 2)
extern struct task_struct *rt_mutex_next_owner(struct rt_mutex *lock);
extern void rt_mutex_init_proxy_locked(struct rt_mutex *lock,
--
1.7.8.3
^ permalink raw reply related [flat|nested] 9+ messages in thread* [PATCH RT 8/8] Linux 3.0.20-rt36-rc1
2012-02-09 1:59 [PATCH RT 0/8] [ANNOUNCE] 3.0.20-rt36-rc1 Steven Rostedt
` (6 preceding siblings ...)
2012-02-09 1:59 ` [PATCH RT 7/8] futex: Fix bug on when a requeued RT task times out Steven Rostedt
@ 2012-02-09 1:59 ` Steven Rostedt
7 siblings, 0 replies; 9+ messages in thread
From: Steven Rostedt @ 2012-02-09 1:59 UTC (permalink / raw)
To: linux-kernel, linux-rt-users; +Cc: Thomas Gleixner, Carsten Emde, John Kacur
[-- Attachment #1: 0008-Linux-3.0.20-rt36-rc1.patch --]
[-- Type: text/plain, Size: 291 bytes --]
From: Steven Rostedt <srostedt@redhat.com>
---
localversion-rt | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/localversion-rt b/localversion-rt
index 366440d..a7827dc 100644
--- a/localversion-rt
+++ b/localversion-rt
@@ -1 +1 @@
--rt35
+-rt36-rc1
--
1.7.8.3
^ permalink raw reply related [flat|nested] 9+ messages in thread