Skip to contentSkip to conversation

LTTng "TIF_KERNEL_TRACE"

DISCUSSION Mathieu Desnoyers 4 participants 11 messages Apr 28, 2009

lore
Share & tools
lore
#0 lore raw

Hi Ingo,

Looking at the current -tip tree, I notice that the TIF_SYSCALL_FTRACE flag is only implemented for x86.

I have TIF_KERNEL_TRACE in my lttng tree which applies to all architectures to do the exact same thing :

lttng-kernel-trace-thread-flag-alpha.patch
lttng-kernel-trace-thread-flag-arm.patch
lttng-kernel-trace-thread-flag-avr32.patch
lttng-kernel-trace-thread-flag-blackfin.patch
lttng-kernel-trace-thread-flag-cris.patch
lttng-kernel-trace-thread-flag-frv.patch
lttng-kernel-trace-thread-flag-h8300.patch
lttng-kernel-trace-thread-flag-ia64.patch
lttng-kernel-trace-thread-flag-m32r.patch
lttng-kernel-trace-thread-flag-m68k.patch
lttng-kernel-trace-thread-flag-mips.patch
lttng-kernel-trace-thread-flag-parisc.patch
lttng-kernel-trace-thread-flag-powerpc.patch
lttng-kernel-trace-thread-flag-s390.patch
lttng-kernel-trace-thread-flag-sh.patch
lttng-kernel-trace-thread-flag-sparc.patch
lttng-kernel-trace-thread-flag-um.patch
lttng-kernel-trace-thread-flag-x86.patch
lttng-kernel-trace-thread-flag-xtensa.patch
lttng-kernel-trace-thread-flag-api.patch

Is there any way we could get this merged ?

One thing I like about the name TIF_KERNEL_TRACE compared to TIF_SYSCALL_FTRACE is that it gives us a per-thread flag that could eventually be used for more kernel tracing purposes than just syscalls.

Mathieu

Signature · 2 lines
Mathieu Desnoyers
OpenPGP key fingerprint: 8CD5 52C3 8E3C 4140 715F  BA06 3F25 A8FE 3BAE 9A68
#1 lore raw
Quoted 29 lines · Mathieu Desnoyers

Hi Ingo,

Looking at the current -tip tree, I notice that the TIF_SYSCALL_FTRACE flag is only implemented for x86.

I have TIF_KERNEL_TRACE in my lttng tree which applies to all architectures to do the exact same thing :

lttng-kernel-trace-thread-flag-alpha.patch
lttng-kernel-trace-thread-flag-arm.patch
lttng-kernel-trace-thread-flag-avr32.patch
lttng-kernel-trace-thread-flag-blackfin.patch
lttng-kernel-trace-thread-flag-cris.patch
lttng-kernel-trace-thread-flag-frv.patch
lttng-kernel-trace-thread-flag-h8300.patch
lttng-kernel-trace-thread-flag-ia64.patch
lttng-kernel-trace-thread-flag-m32r.patch
lttng-kernel-trace-thread-flag-m68k.patch
lttng-kernel-trace-thread-flag-mips.patch
lttng-kernel-trace-thread-flag-parisc.patch
lttng-kernel-trace-thread-flag-powerpc.patch
lttng-kernel-trace-thread-flag-s390.patch
lttng-kernel-trace-thread-flag-sh.patch
lttng-kernel-trace-thread-flag-sparc.patch
lttng-kernel-trace-thread-flag-um.patch
lttng-kernel-trace-thread-flag-x86.patch
lttng-kernel-trace-thread-flag-xtensa.patch
lttng-kernel-trace-thread-flag-api.patch

Quoted 6 lines

Is there any way we could get this merged ?

One thing I like about the name TIF_KERNEL_TRACE compared to TIF_SYSCALL_FTRACE is that it gives us a per-thread flag that could eventually be used for more kernel tracing purposes than just syscalls.

Yeah - TIF_KERNEL_TRACE indeed sounds more descriptive and less restrictive. TIF_SYSCALL_FTRACE was a bit ad-hoc.

Ingo

#2 lore raw
Quoted 39 lines · * Ingo Molnar ([email protected]) wrote:
Quoted 29 lines · Mathieu Desnoyers

Hi Ingo,

Looking at the current -tip tree, I notice that the TIF_SYSCALL_FTRACE flag is only implemented for x86.

I have TIF_KERNEL_TRACE in my lttng tree which applies to all architectures to do the exact same thing :

lttng-kernel-trace-thread-flag-alpha.patch
lttng-kernel-trace-thread-flag-arm.patch
lttng-kernel-trace-thread-flag-avr32.patch
lttng-kernel-trace-thread-flag-blackfin.patch
lttng-kernel-trace-thread-flag-cris.patch
lttng-kernel-trace-thread-flag-frv.patch
lttng-kernel-trace-thread-flag-h8300.patch
lttng-kernel-trace-thread-flag-ia64.patch
lttng-kernel-trace-thread-flag-m32r.patch
lttng-kernel-trace-thread-flag-m68k.patch
lttng-kernel-trace-thread-flag-mips.patch
lttng-kernel-trace-thread-flag-parisc.patch
lttng-kernel-trace-thread-flag-powerpc.patch
lttng-kernel-trace-thread-flag-s390.patch
lttng-kernel-trace-thread-flag-sh.patch
lttng-kernel-trace-thread-flag-sparc.patch
lttng-kernel-trace-thread-flag-um.patch
lttng-kernel-trace-thread-flag-x86.patch
lttng-kernel-trace-thread-flag-xtensa.patch
lttng-kernel-trace-thread-flag-api.patch

Quoted 6 lines

Is there any way we could get this merged ?

One thing I like about the name TIF_KERNEL_TRACE compared to TIF_SYSCALL_FTRACE is that it gives us a per-thread flag that could eventually be used for more kernel tracing purposes than just syscalls.

Quoted 4 lines

Yeah - TIF_KERNEL_TRACE indeed sounds more descriptive and less restrictive. TIF_SYSCALL_FTRACE was a bit ad-hoc.

Ingo

lttng-kernel-trace-thread-flag-api.patch adds the enable/disable API, with refcount, in kernel/sched.c just because I did not know where to put it. Should we move this to its own file in kernel/trace/???.c ?

One option is kernel/trace/trace.c, which is compiled as soon as CONFIG_TRACING is enabled.

I added the prototypes to linux/sched.h, they should probably be moved to... linux/trace.h ?

Mathieu

Signature · 2 lines
Mathieu Desnoyers
OpenPGP key fingerprint: 8CD5 52C3 8E3C 4140 715F  BA06 3F25 A8FE 3BAE 9A68
#3 lore raw
In reply to Ingo Molnar
Quoted 34 lines · * Ingo Molnar ([email protected]) wrote:
Quoted 31 lines · Mathieu Desnoyers

Hi Ingo,

Looking at the current -tip tree, I notice that the TIF_SYSCALL_FTRACE flag is only implemented for x86.

I have TIF_KERNEL_TRACE in my lttng tree which applies to all architectures to do the exact same thing :

lttng-kernel-trace-thread-flag-alpha.patch
lttng-kernel-trace-thread-flag-arm.patch
lttng-kernel-trace-thread-flag-avr32.patch
lttng-kernel-trace-thread-flag-blackfin.patch
lttng-kernel-trace-thread-flag-cris.patch
lttng-kernel-trace-thread-flag-frv.patch
lttng-kernel-trace-thread-flag-h8300.patch
lttng-kernel-trace-thread-flag-ia64.patch
lttng-kernel-trace-thread-flag-m32r.patch
lttng-kernel-trace-thread-flag-m68k.patch
lttng-kernel-trace-thread-flag-mips.patch
lttng-kernel-trace-thread-flag-parisc.patch
lttng-kernel-trace-thread-flag-powerpc.patch
lttng-kernel-trace-thread-flag-s390.patch
lttng-kernel-trace-thread-flag-sh.patch
lttng-kernel-trace-thread-flag-sparc.patch
lttng-kernel-trace-thread-flag-um.patch
lttng-kernel-trace-thread-flag-x86.patch
lttng-kernel-trace-thread-flag-xtensa.patch
lttng-kernel-trace-thread-flag-api.patch

Is there any way we could get this merged ?

Quoted 8 lines
Quoted 4 lines

One thing I like about the name TIF_KERNEL_TRACE compared to TIF_SYSCALL_FTRACE is that it gives us a per-thread flag that could eventually be used for more kernel tracing purposes than just syscalls.

Yeah - TIF_KERNEL_TRACE indeed sounds more descriptive and less restrictive. TIF_SYSCALL_FTRACE was a bit ad-hoc.

Second question :

LTTng :

        read_lock(&tasklist_lock);
        do_each_thread(p, t) {
                set_tsk_thread_flag(t, TIF_KERNEL_TRACE);
        } while_each_thread(p, t);
        read_unlock(&tasklist_lock);

Ftrace:

        read_lock_irqsave(&tasklist_lock, flags);

        do_each_thread(g, t) {
                clear_tsk_thread_flag(t, TIF_SYSCALL_FTRACE);
        } while_each_thread(g, t);

        read_unlock_irqrestore(&tasklist_lock, flags);

With or without irqsave ?

Arguments against irqsave for this read lock :

- it's not used consistently for this read lock all over the kernel. Sometimes the read lock is taken without irqsave.
- it can be a long iteration, and therefore disables interrupts for a long time.

Arguments for irqsave for this read lock :

- Taking any kind of spin/rwlock with inconsistent irq disabling leads to races where interrupts can be disabled for an unbounded amount of time if a spinlock with irqoff waits on a spinlock with irqs on. This is a general problem with current kernel rwlock usage. See my "priority sifting reader-writer lock" patchset for a fix to this problem.

Mathieu

Quoted 1 line

Ingo

Signature · 2 lines
Mathieu Desnoyers
OpenPGP key fingerprint: 8CD5 52C3 8E3C 4140 715F  BA06 3F25 A8FE 3BAE 9A68
#4 lore raw
Quoted 73 lines · Mathieu Desnoyers
Quoted 34 lines · * Ingo Molnar ([email protected]) wrote:
Quoted 31 lines · Mathieu Desnoyers

Hi Ingo,

Looking at the current -tip tree, I notice that the TIF_SYSCALL_FTRACE flag is only implemented for x86.

I have TIF_KERNEL_TRACE in my lttng tree which applies to all architectures to do the exact same thing :

lttng-kernel-trace-thread-flag-alpha.patch
lttng-kernel-trace-thread-flag-arm.patch
lttng-kernel-trace-thread-flag-avr32.patch
lttng-kernel-trace-thread-flag-blackfin.patch
lttng-kernel-trace-thread-flag-cris.patch
lttng-kernel-trace-thread-flag-frv.patch
lttng-kernel-trace-thread-flag-h8300.patch
lttng-kernel-trace-thread-flag-ia64.patch
lttng-kernel-trace-thread-flag-m32r.patch
lttng-kernel-trace-thread-flag-m68k.patch
lttng-kernel-trace-thread-flag-mips.patch
lttng-kernel-trace-thread-flag-parisc.patch
lttng-kernel-trace-thread-flag-powerpc.patch
lttng-kernel-trace-thread-flag-s390.patch
lttng-kernel-trace-thread-flag-sh.patch
lttng-kernel-trace-thread-flag-sparc.patch
lttng-kernel-trace-thread-flag-um.patch
lttng-kernel-trace-thread-flag-x86.patch
lttng-kernel-trace-thread-flag-xtensa.patch
lttng-kernel-trace-thread-flag-api.patch

Is there any way we could get this merged ?

Quoted 8 lines
Quoted 4 lines

One thing I like about the name TIF_KERNEL_TRACE compared to TIF_SYSCALL_FTRACE is that it gives us a per-thread flag that could eventually be used for more kernel tracing purposes than just syscalls.

Yeah - TIF_KERNEL_TRACE indeed sounds more descriptive and less restrictive. TIF_SYSCALL_FTRACE was a bit ad-hoc.

Second question :

LTTng :

        read_lock(&tasklist_lock);
        do_each_thread(p, t) {
                set_tsk_thread_flag(t, TIF_KERNEL_TRACE);
        } while_each_thread(p, t);
        read_unlock(&tasklist_lock);

Ftrace:

        read_lock_irqsave(&tasklist_lock, flags);

        do_each_thread(g, t) {
                clear_tsk_thread_flag(t, TIF_SYSCALL_FTRACE);
        } while_each_thread(g, t);

        read_unlock_irqrestore(&tasklist_lock, flags);

With or without irqsave ?

Arguments against irqsave for this read lock :

- it's not used consistently for this read lock all over the kernel. Sometimes the read lock is taken without irqsave.
- it can be a long iteration, and therefore disables interrupts for a long time.

Arguments for irqsave for this read lock :

Quoted 9 lines

- Taking any kind of spin/rwlock with inconsistent irq disabling leads to races where interrupts can be disabled for an unbounded amount of time if a spinlock with irqoff waits on a spinlock with irqs on. This is a general problem with current kernel rwlock usage. See my "priority sifting reader-writer lock" patchset for a fix to this problem.

Mathieu

I don't know why I used irqsave here, I guess I was tired.

$ git-grep "read_lock_irqsave(&tasklist_lock)" | wc -l
0
$ git-grep "write_lock_irqsave(&tasklist_lock)" | wc -l
0

It is never used in an irq safe fashion, unless one of these sites has the irqs disabled.
Lockdep should even have complained about this, when you hold a lock class in an irq safe fashion and thereafter you try to hold it in an irq unsafe fashion, then the state of the kernel becomes unsafe and lockdep is supposed to complain about that.

Frederic.

#5 lore raw

2009/4/28 Frederic Weisbecker <[email protected]>:

Quoted 92 lines · Mathieu Desnoyers
Quoted 73 lines · * Ingo Molnar ([email protected]) wrote:
Quoted 34 lines · Mathieu Desnoyers
Quoted 32 lines

Hi Ingo,

Looking at the current -tip tree, I notice that the TIF_SYSCALL_FTRACE flag is only implemented for x86.

I have TIF_KERNEL_TRACE in my lttng tree which applies to all architectures to do the exact same thing :

lttng-kernel-trace-thread-flag-alpha.patch
lttng-kernel-trace-thread-flag-arm.patch
lttng-kernel-trace-thread-flag-avr32.patch
lttng-kernel-trace-thread-flag-blackfin.patch
lttng-kernel-trace-thread-flag-cris.patch
lttng-kernel-trace-thread-flag-frv.patch
lttng-kernel-trace-thread-flag-h8300.patch
lttng-kernel-trace-thread-flag-ia64.patch
lttng-kernel-trace-thread-flag-m32r.patch
lttng-kernel-trace-thread-flag-m68k.patch
lttng-kernel-trace-thread-flag-mips.patch
lttng-kernel-trace-thread-flag-parisc.patch
lttng-kernel-trace-thread-flag-powerpc.patch
lttng-kernel-trace-thread-flag-s390.patch
lttng-kernel-trace-thread-flag-sh.patch
lttng-kernel-trace-thread-flag-sparc.patch
lttng-kernel-trace-thread-flag-um.patch
lttng-kernel-trace-thread-flag-x86.patch
lttng-kernel-trace-thread-flag-xtensa.patch
lttng-kernel-trace-thread-flag-api.patch

Is there any way we could get this merged ?

Quoted 8 lines
Quoted 4 lines

One thing I like about the name TIF_KERNEL_TRACE compared to TIF_SYSCALL_FTRACE is that it gives us a per-thread flag that could eventually be used for more kernel tracing purposes than just syscalls.

Yeah - TIF_KERNEL_TRACE indeed sounds more descriptive and less restrictive. TIF_SYSCALL_FTRACE was a bit ad-hoc.

Second question :

LTTng :
        read_lock(&tasklist_lock); do_each_thread(p, t) { set_tsk_thread_flag(t, TIF_KERNEL_TRACE); } while_each_thread(p, t); read_unlock(&tasklist_lock);

Ftrace:
        read_lock_irqsave(&tasklist_lock, flags);

        do_each_thread(g, t) { clear_tsk_thread_flag(t, TIF_SYSCALL_FTRACE); } while_each_thread(g, t);

        read_unlock_irqrestore(&tasklist_lock, flags);

With or without irqsave ?

Arguments against irqsave for this read lock :

- it's not used consistently for this read lock all over the kernel. Sometimes the read lock is taken without irqsave.
- it can be a long iteration, and therefore disables interrupts for a long time.

Arguments for irqsave for this read lock :

Quoted 9 lines

- Taking any kind of spin/rwlock with inconsistent irq disabling leads to races where interrupts can be disabled for an unbounded amount of time if a spinlock with irqoff waits on a spinlock with irqs on. This is a general problem with current kernel rwlock usage. See my "priority sifting reader-writer lock" patchset for a fix to this problem.

Mathieu

I don't know why I used irqsave here, I guess I was tired.

$ git-grep "read_lock_irqsave(&tasklist_lock)" | wc -l
0
$ git-grep "write_lock_irqsave(&tasklist_lock)" | wc -l
0
Quoted 6 lines

It is never used in an irq safe fashion, unless one of these sites has the irqs disabled.
Lockdep should even have complained about this, when you hold a lock class in an irq safe fashion and thereafter you try to hold it in an irq unsafe fashion, then the state of the kernel becomes unsafe and lockdep is supposed to complain about that.

I mean, it's usually a proof of an unsafe state. Not here because we actually never hold
it from irq context.

#6 lore raw
In reply to Frederic Weisbecker
Quoted 2 lines · Frederic Weisbecker

I don't know why I used irqsave here, I guess I was tired.

It was the "when in doubt use irqsave", or was it the "I don't want to think about it, use irqsave" ;-)

-- Steve

#8 lore raw
In reply to Frederic Weisbecker
Quoted 92 lines · * Frederic Weisbecker ([email protected]) wrote:
Quoted 73 lines · Mathieu Desnoyers
Quoted 34 lines · * Ingo Molnar ([email protected]) wrote:
Quoted 31 lines · Mathieu Desnoyers

Hi Ingo,

Looking at the current -tip tree, I notice that the TIF_SYSCALL_FTRACE flag is only implemented for x86.

I have TIF_KERNEL_TRACE in my lttng tree which applies to all architectures to do the exact same thing :

lttng-kernel-trace-thread-flag-alpha.patch
lttng-kernel-trace-thread-flag-arm.patch
lttng-kernel-trace-thread-flag-avr32.patch
lttng-kernel-trace-thread-flag-blackfin.patch
lttng-kernel-trace-thread-flag-cris.patch
lttng-kernel-trace-thread-flag-frv.patch
lttng-kernel-trace-thread-flag-h8300.patch
lttng-kernel-trace-thread-flag-ia64.patch
lttng-kernel-trace-thread-flag-m32r.patch
lttng-kernel-trace-thread-flag-m68k.patch
lttng-kernel-trace-thread-flag-mips.patch
lttng-kernel-trace-thread-flag-parisc.patch
lttng-kernel-trace-thread-flag-powerpc.patch
lttng-kernel-trace-thread-flag-s390.patch
lttng-kernel-trace-thread-flag-sh.patch
lttng-kernel-trace-thread-flag-sparc.patch
lttng-kernel-trace-thread-flag-um.patch
lttng-kernel-trace-thread-flag-x86.patch
lttng-kernel-trace-thread-flag-xtensa.patch
lttng-kernel-trace-thread-flag-api.patch

Is there any way we could get this merged ?

Quoted 8 lines
Quoted 4 lines

One thing I like about the name TIF_KERNEL_TRACE compared to TIF_SYSCALL_FTRACE is that it gives us a per-thread flag that could eventually be used for more kernel tracing purposes than just syscalls.

Yeah - TIF_KERNEL_TRACE indeed sounds more descriptive and less restrictive. TIF_SYSCALL_FTRACE was a bit ad-hoc.

Second question :

LTTng :

        read_lock(&tasklist_lock);
        do_each_thread(p, t) {
                set_tsk_thread_flag(t, TIF_KERNEL_TRACE);
        } while_each_thread(p, t);
        read_unlock(&tasklist_lock);

Ftrace:

        read_lock_irqsave(&tasklist_lock, flags);

        do_each_thread(g, t) {
                clear_tsk_thread_flag(t, TIF_SYSCALL_FTRACE);
        } while_each_thread(g, t);

        read_unlock_irqrestore(&tasklist_lock, flags);

With or without irqsave ?

Arguments against irqsave for this read lock :

- it's not used consistently for this read lock all over the kernel. Sometimes the read lock is taken without irqsave.
- it can be a long iteration, and therefore disables interrupts for a long time.

Arguments for irqsave for this read lock :

Quoted 9 lines

- Taking any kind of spin/rwlock with inconsistent irq disabling leads to races where interrupts can be disabled for an unbounded amount of time if a spinlock with irqoff waits on a spinlock with irqs on. This is a general problem with current kernel rwlock usage. See my "priority sifting reader-writer lock" patchset for a fix to this problem.

Mathieu

I don't know why I used irqsave here, I guess I was tired.

$ git-grep "read_lock_irqsave(&tasklist_lock)" | wc -l
0
$ git-grep "write_lock_irqsave(&tasklist_lock)" | wc -l
0
Quoted 6 lines

It is never used in an irq safe fashion, unless one of these sites has the irqs disabled.
Lockdep should even have complained about this, when you hold a lock class in an irq safe fashion and thereafter you try to hold it in an irq unsafe fashion, then the state of the kernel becomes unsafe and lockdep is supposed to complain about that.

No, read-write lock is a "special case" where it does not deadlock if you have an interrupt handler taking the read lock over another read lock. It's just the write lock that _must absolutely_ disable interrupts.

However, the "latency race" scenario I explained above applies here, because the write lock disables interrupts and the read locks doesn't.

Mathieu

Quoted 4 lines

Frederic.

Signature · 2 lines
Mathieu Desnoyers
OpenPGP key fingerprint: 8CD5 52C3 8E3C 4140 715F  BA06 3F25 A8FE 3BAE 9A68
#9 lore raw
Quoted 4 lines · Mathieu Desnoyers

No, read-write lock is a "special case" where it does not deadlock if you have an interrupt handler taking the read lock over another read lock. It's just the write lock that _must absolutely_ disable interrupts.

Ah, you're right, I was thinking with spinlock rules in mind :)

Quoted 5 lines

However, the "latency race" scenario I explained above applies here, because the write lock disables interrupts and the read locks doesn't.

Mathieu

#7 lore raw
In reply to Steven Rostedt
Quoted 7 lines · Steven Rostedt
Quoted 2 lines · Frederic Weisbecker

I don't know why I used irqsave here, I guess I was tired.

It was the "when in doubt use irqsave", or was it the "I don't want to think about it, use irqsave" ;-)

A bit of the former and the latter...
It was "Argh, it's too late, I should have gone to my bed several hours ago, just use irqsave..." :-/

Quoted 2 lines

-- Steve

#10 lore raw
In reply to Frederic Weisbecker
Quoted 9 lines · Frederic Weisbecker
Quoted 4 lines · Mathieu Desnoyers

No, read-write lock is a "special case" where it does not deadlock if you have an interrupt handler taking the read lock over another read lock. It's just the write lock that _must absolutely_ disable interrupts.

Ah, you're right, I was thinking with spinlock rules in mind :)

The only time you would want to do a read_lock_irqsave is if the write_lock is taken in irq context. But I do not know of any lock where that is the case.

-- Steve