mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
* [PATCH] local_t Documentation update 2
@ 2007-08-29 12:19 Mathieu Desnoyers
  2007-08-30  1:08 ` Grant Grundler
  0 siblings, 1 reply; 3+ messages in thread
From: Mathieu Desnoyers @ 2007-08-29 12:19 UTC (permalink / raw)
  To: akpm; +Cc: linux-kernel, Grant Grundler

local_t Documentation update 2

Grant Grundler was asking for more detail about correct usage of local atomic
operations and suggested adding the resulting summary to local_ops.txt.

"Please add a bit more detail. If DaveM is correct (he normally is), then
there must be limits on how the local_t can be used in the kernel process
and interrupt contexts. I'd like those rules spelled out very clearly
since it's easy to get wrong and tracking down such a bug is quite painful."

It applies on top of 2.6.23-rc3-mm1.

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
CC: Grant Grundler <grundler@parisc-linux.org>
---
 Documentation/local_ops.txt |   23 +++++++++++++++++++++++
 1 file changed, 23 insertions(+)

Index: linux-2.6-lttng/Documentation/local_ops.txt
===================================================================
--- linux-2.6-lttng.orig/Documentation/local_ops.txt	2007-08-29 08:09:34.000000000 -0400
+++ linux-2.6-lttng/Documentation/local_ops.txt	2007-08-29 08:15:37.000000000 -0400
@@ -45,6 +45,29 @@ long fails. The definition looks like :
 typedef struct { atomic_long_t a; } local_t;
 
 
+* Rules to follow when using local atomic operations
+
+- Variables touched by local ops must be per cpu variables.
+- _Only_ the CPU owner of these variables must write to them.
+- This CPU can use local ops from any context (process, irq, softirq, nmi, ...)
+  to update its local_t variables.
+- Preemption (or interrupts) must be disabled when using local ops in
+  process context to   make sure the process won't be migrated to a
+  different CPU between getting the per-cpu variable and doing the
+  actual local op.
+- When using local ops in interrupt context, no special care must be
+  taken on a mainline kernel, since they will run on the local CPU with
+  preemption already disabled. I suggest, however, to explicitly
+  disable preemption anyway to make sure it will still work correctly on
+  -rt kernels.
+- Reading the local cpu variable will provide the current copy of the
+  variable.
+- Reads of these variables can be done from any CPU, because updates to
+  "long", aligned, variables are always atomic. Since no memory
+  synchronization is done by the writer CPU, an outdated copy of the
+  variable can be read when reading some _other_ cpu's variables.
+
+
 * How to use local atomic operations
 
 #include <linux/percpu.h>

-- 
Mathieu Desnoyers
Computer Engineering Ph.D. Student, Ecole Polytechnique de Montreal
OpenPGP key fingerprint: 8CD5 52C3 8E3C 4140 715F  BA06 3F25 A8FE 3BAE 9A68

^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: [PATCH] local_t Documentation update 2
  2007-08-29 12:19 [PATCH] local_t Documentation update 2 Mathieu Desnoyers
@ 2007-08-30  1:08 ` Grant Grundler
  0 siblings, 0 replies; 3+ messages in thread
From: Grant Grundler @ 2007-08-30  1:08 UTC (permalink / raw)
  To: Mathieu Desnoyers; +Cc: akpm, linux-kernel, Grant Grundler

On Wed, Aug 29, 2007 at 08:19:53AM -0400, Mathieu Desnoyers wrote:
> local_t Documentation update 2
> 
> Grant Grundler was asking for more detail about correct usage of local atomic
> operations and suggested adding the resulting summary to local_ops.txt.
> 
> "Please add a bit more detail. If DaveM is correct (he normally is), then
> there must be limits on how the local_t can be used in the kernel process
> and interrupt contexts. I'd like those rules spelled out very clearly
> since it's easy to get wrong and tracking down such a bug is quite painful."
> 
> It applies on top of 2.6.23-rc3-mm1.
> 
> Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
> CC: Grant Grundler <grundler@parisc-linux.org>

Signed-off-by: Grant Grundler <grundler@parisc-linux.org>

thanks!
grant

> ---
>  Documentation/local_ops.txt |   23 +++++++++++++++++++++++
>  1 file changed, 23 insertions(+)
> 
> Index: linux-2.6-lttng/Documentation/local_ops.txt
> ===================================================================
> --- linux-2.6-lttng.orig/Documentation/local_ops.txt	2007-08-29 08:09:34.000000000 -0400
> +++ linux-2.6-lttng/Documentation/local_ops.txt	2007-08-29 08:15:37.000000000 -0400
> @@ -45,6 +45,29 @@ long fails. The definition looks like :
>  typedef struct { atomic_long_t a; } local_t;
>  
>  
> +* Rules to follow when using local atomic operations
> +
> +- Variables touched by local ops must be per cpu variables.
> +- _Only_ the CPU owner of these variables must write to them.
> +- This CPU can use local ops from any context (process, irq, softirq, nmi, ...)
> +  to update its local_t variables.
> +- Preemption (or interrupts) must be disabled when using local ops in
> +  process context to   make sure the process won't be migrated to a
> +  different CPU between getting the per-cpu variable and doing the
> +  actual local op.
> +- When using local ops in interrupt context, no special care must be
> +  taken on a mainline kernel, since they will run on the local CPU with
> +  preemption already disabled. I suggest, however, to explicitly
> +  disable preemption anyway to make sure it will still work correctly on
> +  -rt kernels.
> +- Reading the local cpu variable will provide the current copy of the
> +  variable.
> +- Reads of these variables can be done from any CPU, because updates to
> +  "long", aligned, variables are always atomic. Since no memory
> +  synchronization is done by the writer CPU, an outdated copy of the
> +  variable can be read when reading some _other_ cpu's variables.
> +
> +
>  * How to use local atomic operations
>  
>  #include <linux/percpu.h>
> 
> -- 
> Mathieu Desnoyers
> Computer Engineering Ph.D. Student, Ecole Polytechnique de Montreal
> OpenPGP key fingerprint: 8CD5 52C3 8E3C 4140 715F  BA06 3F25 A8FE 3BAE 9A68

^ permalink raw reply	[flat|nested] 3+ messages in thread

* [PATCH] local_t Documentation update 2
  2007-10-30 18:58     ` Christoph Lameter
@ 2007-10-31  1:52       ` Mathieu Desnoyers
  0 siblings, 0 replies; 3+ messages in thread
From: Mathieu Desnoyers @ 2007-10-31  1:52 UTC (permalink / raw)
  To: Andrew Morton
  Cc: matthew, linux-kernel, linux-mm, penberg, linux-arch, Christoph Lameter

local_t Documentation update 2

(this patch seems to have fallen off the grid, but is still providing
useful information. It applies to 2.6.23-mm1.)

Grant Grundler was asking for more detail about correct usage of local atomic
operations and suggested adding the resulting summary to local_ops.txt.

"Please add a bit more detail. If DaveM is correct (he normally is), then
there must be limits on how the local_t can be used in the kernel process
and interrupt contexts. I'd like those rules spelled out very clearly
since it's easy to get wrong and tracking down such a bug is quite painful."

Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Signed-off-by: Grant Grundler <grundler@parisc-linux.org>
---
 Documentation/local_ops.txt |   23 +++++++++++++++++++++++
 1 file changed, 23 insertions(+)

Index: linux-2.6-lttng/Documentation/local_ops.txt
===================================================================
--- linux-2.6-lttng.orig/Documentation/local_ops.txt	2007-09-04 11:53:23.000000000 -0400
+++ linux-2.6-lttng/Documentation/local_ops.txt	2007-09-04 12:19:31.000000000 -0400
@@ -68,6 +68,29 @@ typedef struct { atomic_long_t a; } loca
   variable can be read when reading some _other_ cpu's variables.
 
 
+* Rules to follow when using local atomic operations
+
+- Variables touched by local ops must be per cpu variables.
+- _Only_ the CPU owner of these variables must write to them.
+- This CPU can use local ops from any context (process, irq, softirq, nmi, ...)
+  to update its local_t variables.
+- Preemption (or interrupts) must be disabled when using local ops in
+  process context to   make sure the process won't be migrated to a
+  different CPU between getting the per-cpu variable and doing the
+  actual local op.
+- When using local ops in interrupt context, no special care must be
+  taken on a mainline kernel, since they will run on the local CPU with
+  preemption already disabled. I suggest, however, to explicitly
+  disable preemption anyway to make sure it will still work correctly on
+  -rt kernels.
+- Reading the local cpu variable will provide the current copy of the
+  variable.
+- Reads of these variables can be done from any CPU, because updates to
+  "long", aligned, variables are always atomic. Since no memory
+  synchronization is done by the writer CPU, an outdated copy of the
+  variable can be read when reading some _other_ cpu's variables.
+
+
 * How to use local atomic operations
 
 #include <linux/percpu.h>
-- 
Mathieu Desnoyers
Computer Engineering Ph.D. Student, Ecole Polytechnique de Montreal
OpenPGP key fingerprint: 8CD5 52C3 8E3C 4140 715F  BA06 3F25 A8FE 3BAE 9A68

^ permalink raw reply	[flat|nested] 3+ messages in thread

end of thread, other threads:[~2007-10-31  1:58 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2007-08-29 12:19 [PATCH] local_t Documentation update 2 Mathieu Desnoyers
2007-08-30  1:08 ` Grant Grundler
2007-10-28  3:31 [patch 00/10] SLUB: SMP regression tests on Dual Xeon E5345 (8p) and new performance patches Christoph Lameter
2007-10-28  3:32 ` [patch 08/10] SLUB: Optional fast path using cmpxchg_local Christoph Lameter
2007-10-30 18:49   ` Andrew Morton
2007-10-30 18:58     ` Christoph Lameter
2007-10-31  1:52       ` [PATCH] local_t Documentation update 2 Mathieu Desnoyers

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox

Powered by JetHome