Re: [Xenomai-core] x86: Endless minor faults

2009-07-07 Thread Jan Kiszka
Gilles Chanteperdrix wrote: Jan Kiszka wrote: Gilles Chanteperdrix wrote: Jan Kiszka wrote: Gilles Chanteperdrix wrote: Jan Kiszka wrote: Gilles Chanteperdrix wrote: Jan Kiszka wrote: Hi again, this is now basically the patch which seems to stabilized x86 /wrt mmu switches again:

Re: [Xenomai-core] x86: Endless minor faults

2009-07-06 Thread Gilles Chanteperdrix
Jan Kiszka wrote: Gilles Chanteperdrix wrote: Jan Kiszka wrote: Gilles Chanteperdrix wrote: Jan Kiszka wrote: Gilles Chanteperdrix wrote: Jan Kiszka wrote: Hi again, this is now basically the patch which seems to stabilized x86 /wrt mmu switches again: There were 3 race windows between

Re: [Xenomai-core] x86: Endless minor faults

2009-07-05 Thread Jan Kiszka
Gilles Chanteperdrix wrote: Jan Kiszka wrote: Gilles Chanteperdrix wrote: Jan Kiszka wrote: Hi again, this is now basically the patch which seems to stabilized x86 /wrt mmu switches again: There were 3 race windows between setting active_mm of the current task and actually switching it

Re: [Xenomai-core] x86: Endless minor faults

2009-07-05 Thread Jan Kiszka
Gilles Chanteperdrix wrote: Jan Kiszka wrote: Gilles Chanteperdrix wrote: Jan Kiszka wrote: Gilles Chanteperdrix wrote: Jan Kiszka wrote: Hi again, this is now basically the patch which seems to stabilized x86 /wrt mmu switches again: There were 3 race windows between setting active_mm

Re: [Xenomai-core] x86: Endless minor faults

2009-07-04 Thread Jan Kiszka
Hi again, this is now basically the patch which seems to stabilized x86 /wrt mmu switches again: There were 3 race windows between setting active_mm of the current task and actually switching it (that's a noarch issue), there were several calls into switch_mm without proper hard interrupt

Re: [Xenomai-core] x86: Endless minor faults

2009-07-04 Thread Jan Kiszka
Gilles Chanteperdrix wrote: Jan Kiszka wrote: Hi again, this is now basically the patch which seems to stabilized x86 /wrt mmu switches again: There were 3 race windows between setting active_mm of the current task and actually switching it (that's a noarch issue), there were several

Re: [Xenomai-core] x86: Endless minor faults

2009-07-04 Thread Gilles Chanteperdrix
Jan Kiszka wrote: Hi again, this is now basically the patch which seems to stabilized x86 /wrt mmu switches again: There were 3 race windows between setting active_mm of the current task and actually switching it (that's a noarch issue), there were several calls into switch_mm without

Re: [Xenomai-core] x86: Endless minor faults

2009-07-02 Thread Jan Kiszka
Gilles Chanteperdrix wrote: Philippe Gerum wrote: On Wed, 2009-07-01 at 20:15 +0200, Philippe Gerum wrote: On Wed, 2009-07-01 at 19:56 +0200, Jan Kiszka wrote: Jan Kiszka wrote: Jan Kiszka wrote: Gilles Chanteperdrix wrote: Jan Kiszka wrote: Jan Kiszka wrote: Gilles Chanteperdrix wrote:

Re: [Xenomai-core] x86: Endless minor faults

2009-07-02 Thread Gilles Chanteperdrix
Jan Kiszka wrote: Gilles Chanteperdrix wrote: Philippe Gerum wrote: On Wed, 2009-07-01 at 20:15 +0200, Philippe Gerum wrote: On Wed, 2009-07-01 at 19:56 +0200, Jan Kiszka wrote: Jan Kiszka wrote: Jan Kiszka wrote: Gilles Chanteperdrix wrote: Jan Kiszka wrote: Jan Kiszka wrote: Gilles

Re: [Xenomai-core] x86: Endless minor faults

2009-07-02 Thread Philippe Gerum
On Wed, 2009-07-01 at 20:56 +0200, Jan Kiszka wrote: Philippe Gerum wrote: On Wed, 2009-07-01 at 19:56 +0200, Jan Kiszka wrote: Jan Kiszka wrote: Jan Kiszka wrote: Gilles Chanteperdrix wrote: Jan Kiszka wrote: Jan Kiszka wrote: Gilles Chanteperdrix wrote: Jan Kiszka wrote: Jan

Re: [Xenomai-core] x86: Endless minor faults

2009-07-02 Thread Jan Kiszka
Gilles Chanteperdrix wrote: Jan Kiszka wrote: On archs with non-atomic switch_mm(), use_mm() will require a different strategy. I'm thinking about something like use_mm(): set_some_flag(); barrier(); current-mm = new_mm; current-active_mm = new_mm;

Re: [Xenomai-core] x86: Endless minor faults

2009-07-02 Thread Gilles Chanteperdrix
Jan Kiszka wrote: Gilles Chanteperdrix wrote: Jan Kiszka wrote: On archs with non-atomic switch_mm(), use_mm() will require a different strategy. I'm thinking about something like use_mm(): set_some_flag(); barrier(); current-mm = new_mm; current-active_mm = new_mm;

Re: [Xenomai-core] x86: Endless minor faults

2009-07-01 Thread Jan Kiszka
Jan Kiszka wrote: It's still unclear what goes on precisely, we are still digging, but the test system that can produce this is highly contended. Short update: Further instrumentation revealed that cr3 differs from active_mm-pgd while we are looping over that fault, ie. the kernel tries to

Re: [Xenomai-core] x86: Endless minor faults

2009-07-01 Thread Gilles Chanteperdrix
Jan Kiszka wrote: Jan Kiszka wrote: It's still unclear what goes on precisely, we are still digging, but the test system that can produce this is highly contended. Short update: Further instrumentation revealed that cr3 differs from active_mm-pgd while we are looping over that fault, ie.

Re: [Xenomai-core] x86: Endless minor faults

2009-07-01 Thread Jan Kiszka
Gilles Chanteperdrix wrote: Jan Kiszka wrote: Jan Kiszka wrote: It's still unclear what goes on precisely, we are still digging, but the test system that can produce this is highly contended. Short update: Further instrumentation revealed that cr3 differs from active_mm-pgd while we are

Re: [Xenomai-core] x86: Endless minor faults

2009-07-01 Thread Gilles Chanteperdrix
Jan Kiszka wrote: Gilles Chanteperdrix wrote: Jan Kiszka wrote: Jan Kiszka wrote: It's still unclear what goes on precisely, we are still digging, but the test system that can produce this is highly contended. Short update: Further instrumentation revealed that cr3 differs from

Re: [Xenomai-core] x86: Endless minor faults

2009-07-01 Thread Jan Kiszka
Jan Kiszka wrote: Gilles Chanteperdrix wrote: Jan Kiszka wrote: Jan Kiszka wrote: It's still unclear what goes on precisely, we are still digging, but the test system that can produce this is highly contended. Short update: Further instrumentation revealed that cr3 differs from

Re: [Xenomai-core] x86: Endless minor faults

2009-07-01 Thread Gilles Chanteperdrix
Jan Kiszka wrote: Jan Kiszka wrote: Gilles Chanteperdrix wrote: Jan Kiszka wrote: Jan Kiszka wrote: It's still unclear what goes on precisely, we are still digging, but the test system that can produce this is highly contended. Short update: Further instrumentation revealed that cr3 differs

Re: [Xenomai-core] x86: Endless minor faults

2009-07-01 Thread Jan Kiszka
Gilles Chanteperdrix wrote: Jan Kiszka wrote: Jan Kiszka wrote: Gilles Chanteperdrix wrote: Jan Kiszka wrote: Jan Kiszka wrote: It's still unclear what goes on precisely, we are still digging, but the test system that can produce this is highly contended. Short update: Further

Re: [Xenomai-core] x86: Endless minor faults

2009-07-01 Thread Jan Kiszka
Jan Kiszka wrote: Gilles Chanteperdrix wrote: Jan Kiszka wrote: Jan Kiszka wrote: Gilles Chanteperdrix wrote: Jan Kiszka wrote: Jan Kiszka wrote: It's still unclear what goes on precisely, we are still digging, but the test system that can produce this is highly contended. Short update:

Re: [Xenomai-core] x86: Endless minor faults

2009-07-01 Thread Jan Kiszka
Jan Kiszka wrote: Jan Kiszka wrote: Gilles Chanteperdrix wrote: Jan Kiszka wrote: Jan Kiszka wrote: Gilles Chanteperdrix wrote: Jan Kiszka wrote: Jan Kiszka wrote: It's still unclear what goes on precisely, we are still digging, but the test system that can produce this is highly

Re: [Xenomai-core] x86: Endless minor faults

2009-07-01 Thread Philippe Gerum
On Wed, 2009-07-01 at 19:56 +0200, Jan Kiszka wrote: Jan Kiszka wrote: Jan Kiszka wrote: Gilles Chanteperdrix wrote: Jan Kiszka wrote: Jan Kiszka wrote: Gilles Chanteperdrix wrote: Jan Kiszka wrote: Jan Kiszka wrote: It's still unclear what goes on precisely, we are still digging,

Re: [Xenomai-core] x86: Endless minor faults

2009-07-01 Thread Philippe Gerum
On Wed, 2009-07-01 at 20:15 +0200, Philippe Gerum wrote: On Wed, 2009-07-01 at 19:56 +0200, Jan Kiszka wrote: Jan Kiszka wrote: Jan Kiszka wrote: Gilles Chanteperdrix wrote: Jan Kiszka wrote: Jan Kiszka wrote: Gilles Chanteperdrix wrote: Jan Kiszka wrote: Jan Kiszka wrote:

Re: [Xenomai-core] x86: Endless minor faults

2009-07-01 Thread Jan Kiszka
Philippe Gerum wrote: On Wed, 2009-07-01 at 20:15 +0200, Philippe Gerum wrote: On Wed, 2009-07-01 at 19:56 +0200, Jan Kiszka wrote: Jan Kiszka wrote: Jan Kiszka wrote: Gilles Chanteperdrix wrote: Jan Kiszka wrote: Jan Kiszka wrote: Gilles Chanteperdrix wrote: Jan Kiszka wrote: Jan Kiszka

Re: [Xenomai-core] x86: Endless minor faults

2009-06-30 Thread Jan Kiszka
Jan Kiszka wrote: Hi all, seen such loops before? This particular trace is from a 2.6.29.3 kernel with ipipe-2.3-01 (SMP/PREEMPT_VOLUNTARY), but the same happens with 2.6.29.5/2.3-03: :| +func-6530.084 __ipipe_handle_exception+0x11 (page_fault+0x26) :| +func

Re: [Xenomai-core] x86: Endless minor faults

2009-06-30 Thread Gilles Chanteperdrix
Jan Kiszka wrote: Jan Kiszka wrote: Hi all, seen such loops before? This particular trace is from a 2.6.29.3 kernel with ipipe-2.3-01 (SMP/PREEMPT_VOLUNTARY), but the same happens with 2.6.29.5/2.3-03: :| +func-6530.084 __ipipe_handle_exception+0x11

Re: [Xenomai-core] x86: Endless minor faults

2009-06-30 Thread Jan Kiszka
Gilles Chanteperdrix wrote: Jan Kiszka wrote: Jan Kiszka wrote: Hi all, seen such loops before? This particular trace is from a 2.6.29.3 kernel with ipipe-2.3-01 (SMP/PREEMPT_VOLUNTARY), but the same happens with 2.6.29.5/2.3-03: :| +func-6530.084

Re: [Xenomai-core] x86: Endless minor faults

2009-06-30 Thread Jan Kiszka
Gilles Chanteperdrix wrote: Jan Kiszka wrote: Jan Kiszka wrote: Hi all, seen such loops before? This particular trace is from a 2.6.29.3 kernel with ipipe-2.3-01 (SMP/PREEMPT_VOLUNTARY), but the same happens with 2.6.29.5/2.3-03: :| +func-6530.084

Re: [Xenomai-core] x86: Endless minor faults

2009-06-30 Thread Philippe Gerum
On Tue, 2009-06-30 at 10:42 +0200, Gilles Chanteperdrix wrote: Jan Kiszka wrote: Jan Kiszka wrote: Hi all, seen such loops before? This particular trace is from a 2.6.29.3 kernel with ipipe-2.3-01 (SMP/PREEMPT_VOLUNTARY), but the same happens with 2.6.29.5/2.3-03: :| +func

Re: [Xenomai-core] x86: Endless minor faults

2009-06-30 Thread Philippe Gerum
On Tue, 2009-06-30 at 11:21 +0200, Gilles Chanteperdrix wrote: Philippe Gerum wrote: On Tue, 2009-06-30 at 10:42 +0200, Gilles Chanteperdrix wrote: Jan Kiszka wrote: Jan Kiszka wrote: Hi all, seen such loops before? This particular trace is from a 2.6.29.3 kernel with ipipe-2.3-01

Re: [Xenomai-core] x86: Endless minor faults

2009-06-30 Thread Gilles Chanteperdrix
Philippe Gerum wrote: On Tue, 2009-06-30 at 11:21 +0200, Gilles Chanteperdrix wrote: Philippe Gerum wrote: On Tue, 2009-06-30 at 10:42 +0200, Gilles Chanteperdrix wrote: Jan Kiszka wrote: Jan Kiszka wrote: Hi all, seen such loops before? This particular trace is from a 2.6.29.3 kernel

Re: [Xenomai-core] x86: Endless minor faults

2009-06-30 Thread Philippe Gerum
On Tue, 2009-06-30 at 11:26 +0200, Gilles Chanteperdrix wrote: Philippe Gerum wrote: On Tue, 2009-06-30 at 11:21 +0200, Gilles Chanteperdrix wrote: Philippe Gerum wrote: On Tue, 2009-06-30 at 10:42 +0200, Gilles Chanteperdrix wrote: Jan Kiszka wrote: Jan Kiszka wrote: Hi all, seen

Re: [Xenomai-core] x86: Endless minor faults

2009-06-30 Thread Gilles Chanteperdrix
Philippe Gerum wrote: On Tue, 2009-06-30 at 11:26 +0200, Gilles Chanteperdrix wrote: Philippe Gerum wrote: On Tue, 2009-06-30 at 11:21 +0200, Gilles Chanteperdrix wrote: Philippe Gerum wrote: On Tue, 2009-06-30 at 10:42 +0200, Gilles Chanteperdrix wrote: Jan Kiszka wrote: Jan Kiszka wrote:

Re: [Xenomai-core] x86: Endless minor faults

2009-06-30 Thread Jan Kiszka
Gilles Chanteperdrix wrote: Philippe Gerum wrote: On Tue, 2009-06-30 at 11:26 +0200, Gilles Chanteperdrix wrote: Philippe Gerum wrote: On Tue, 2009-06-30 at 11:21 +0200, Gilles Chanteperdrix wrote: Philippe Gerum wrote: On Tue, 2009-06-30 at 10:42 +0200, Gilles Chanteperdrix wrote: Jan

Re: [Xenomai-core] x86: Endless minor faults

2009-06-29 Thread Philippe Gerum
On Mon, 2009-06-29 at 19:35 +0200, Jan Kiszka wrote: Hi all, seen such loops before? This particular trace is from a 2.6.29.3 kernel with ipipe-2.3-01 (SMP/PREEMPT_VOLUNTARY), but the same happens with 2.6.29.5/2.3-03: Do you mean 2.6.29.5/2.4-03? -- Philippe.

Re: [Xenomai-core] x86: Endless minor faults

2009-06-29 Thread Jan Kiszka
Philippe Gerum wrote: On Mon, 2009-06-29 at 19:35 +0200, Jan Kiszka wrote: Hi all, seen such loops before? This particular trace is from a 2.6.29.3 kernel with ipipe-2.3-01 (SMP/PREEMPT_VOLUNTARY), but the same happens with 2.6.29.5/2.3-03: Do you mean 2.6.29.5/2.4-03? No, in fact I