[question] Can the execution of the atomtic operation instruction pair lwarx/stwcx be interrrupted by local HW interruptions?

[question] Can the execution of the atomtic operation instruction pair lwarx/stwcx be interrrupted by local HW interruptions?

[Gavin Hu]

[-- Attachment #1: Type: text/plain, Size: 725 bytes --]

Hi

I notice that there is a pair ppc instructions lwarx and stwcx used to
atomtic operation for instance, atomic_inc/atomic_dec.

In some ppc manuals, they more emphasize its mechanism is that lwarx can
reseve the target memory address preventing other CORE from modifying it.

I assume that there is atomtic operation executing on the CORE0 in a
multicore system. In this situation, does the CORE0 disable the local HW
interrupt?
Can the executing process from the beginning of lwarx and end of stwcx be
interrupted by HW interruptions/exceptions?  Anyway, they are two assembly
instructions.

 Thanks a lot!

"1:    lwarx    %0,0,%2        # atomic_inc\n\
    addic    %0,%0,1\n"
"    stwcx.    %0,0,%2 \n\


BR
Gavin. Hu

[-- Attachment #2: Type: text/html, Size: 899 bytes --]

Re: [question] Can the execution of the atomtic operation instruction pair lwarx/stwcx be interrrupted by local HW interruptions?

[wyang]

[-- Attachment #1: Type: text/plain, Size: 1150 bytes --]

On 12/28/2013 01:41 PM, Gavin Hu wrote:
> Hi
>
> I notice that there is a pair ppc instructions lwarx and stwcx used to 
> atomtic operation for instance, atomic_inc/atomic_dec.
>
> In some ppc manuals, they more emphasize its mechanism is that lwarx 
> can reseve the target memory address preventing other CORE from 
> modifying it.
>
> I assume that there is atomtic operation executing on the CORE0 in a 
> multicore system. In this situation, does the CORE0 disable the local 
> HW interrupt?
> Can the executing process from the beginning of lwarx and end of stwcx 
> be interrupted by HW interruptions/exceptions? Anyway, they are two 
> assembly instructions.

It should just like other arch, the processor should response any 
interrupt after the execution of a instruction, so the local HW 
interrupt is not disabled.

Thanks
Wei
>
>  Thanks a lot!
>
> "1:    lwarx    %0,0,%2        # atomic_inc\n\
>     addic    %0,%0,1\n"
> "    stwcx.    %0,0,%2 \n\
>
>
> BR
> Gavin. Hu
>
>
> _______________________________________________
> Linuxppc-dev mailing list
> Linuxppc-dev@lists.ozlabs.org
> https://lists.ozlabs.org/listinfo/linuxppc-dev


[-- Attachment #2: Type: text/html, Size: 2655 bytes --]

Re: [question] Can the execution of the atomtic operation instruction pair lwarx/stwcx be interrrupted by local HW interruptions?

[Gavin Hu]

[-- Attachment #1: Type: text/plain, Size: 1898 bytes --]

Thanks your response.  :)
But that means that these optimitive operations like atomic_add() aren't
optimitive actully in PPC architecture, right? Becuase they can be
interrupted by loacl HW interrupts. Theoretically, the ISR also can access
the atomic gloable variable.


The following codes are complete atomic_inc() copied from arch/
static __inline__ void atomic_add(int a, atomic_t *v)
{
    int t;

    __asm__ __volatile__(
"1:    lwarx    %0,0,%3        # atomic_add\n\
    add    %0,%2,%0\n"
    PPC405_ERR77(0,%3)
"    stwcx.    %0,0,%3 \n\
    bne-    1b"
    : "=&r" (t), "+m" (v->counter)
    : "r" (a), "r" (&v->counter)
    : "cc");
}


BR
Gavin. Hu


On Mon, Dec 30, 2013 at 9:54 AM, wyang <w90p710@gmail.com> wrote:

>  On 12/28/2013 01:41 PM, Gavin Hu wrote:
>
> Hi
>
> I notice that there is a pair ppc instructions lwarx and stwcx used to
> atomtic operation for instance, atomic_inc/atomic_dec.
>
>  In some ppc manuals, they more emphasize its mechanism is that lwarx can
> reseve the target memory address preventing other CORE from modifying it.
>
>  I assume that there is atomtic operation executing on the CORE0 in a
> multicore system. In this situation, does the CORE0 disable the local HW
> interrupt?
>  Can the executing process from the beginning of lwarx and end of stwcx
> be interrupted by HW interruptions/exceptions?  Anyway, they are two
> assembly instructions.
>
>
> It should just like other arch, the processor should response any
> interrupt after the execution of a instruction, so the local HW interrupt
> is not disabled.
>
> Thanks
> Wei
>
>
>  Thanks a lot!
>
> "1:    lwarx    %0,0,%2        # atomic_inc\n\
>     addic    %0,%0,1\n"
> "    stwcx.    %0,0,%2 \n\
>
>
>  BR
>  Gavin. Hu
>
>
> _______________________________________________
> Linuxppc-dev mailing listLinuxppc-dev@lists.ozlabs.orghttps://lists.ozlabs.org/listinfo/linuxppc-dev
>
>
>

[-- Attachment #2: Type: text/html, Size: 3427 bytes --]

Re: [question] Can the execution of the atomtic operation instruction pair lwarx/stwcx be interrrupted by local HW interruptions?

[wyang]

[-- Attachment #1: Type: text/plain, Size: 2400 bytes --]


On 01/06/2014 11:41 AM, Gavin Hu wrote:
> Thanks your response.  :)
> But that means that these optimitive operations like atomic_add() 
> aren't optimitive actully in PPC architecture, right? Becuase they can 
> be interrupted by loacl HW interrupts. Theoretically, the ISR also can 
> access the atomic gloable variable.

Nope, my understand is that if you wanna sync kernel primitive code with 
ISR, you have responsibility to disable local interrupts. atomic_add 
does not guarantee to handle such case.

Thanks
Wei

>
>
> The following codes are complete atomic_inc() copied from arch/
> static __inline__ void atomic_add(int a, atomic_t *v)
> {
>     int t;
>
>     __asm__ __volatile__(
> "1:    lwarx    %0,0,%3        # atomic_add\n\
>     add    %0,%2,%0\n"
>     PPC405_ERR77(0,%3)
> "    stwcx.    %0,0,%3 \n\
>     bne-    1b"
>     : "=&r" (t), "+m" (v->counter)
>     : "r" (a), "r" (&v->counter)
>     : "cc");
> }
>
>
> BR
> Gavin. Hu
>
>
> On Mon, Dec 30, 2013 at 9:54 AM, wyang <w90p710@gmail.com 
> <mailto:w90p710@gmail.com>> wrote:
>
>     On 12/28/2013 01:41 PM, Gavin Hu wrote:
>>     Hi
>>
>>     I notice that there is a pair ppc instructions lwarx and stwcx
>>     used to atomtic operation for instance, atomic_inc/atomic_dec.
>>
>>     In some ppc manuals, they more emphasize its mechanism is that
>>     lwarx can reseve the target memory address preventing other CORE
>>     from modifying it.
>>
>>     I assume that there is atomtic operation executing on the CORE0
>>     in a multicore system. In this situation, does the CORE0 disable
>>     the local HW interrupt?
>>     Can the executing process from the beginning of lwarx and end of
>>     stwcx be interrupted by HW interruptions/exceptions?  Anyway,
>>     they are two assembly instructions.
>
>     It should just like other arch, the processor should response any
>     interrupt after the execution of a instruction, so the local HW
>     interrupt is not disabled.
>
>     Thanks
>     Wei
>>
>>      Thanks a lot!
>>
>>     "1:    lwarx    %0,0,%2        # atomic_inc\n\
>>         addic    %0,%0,1\n"
>>     "    stwcx.    %0,0,%2 \n\
>>
>>
>>     BR
>>     Gavin. Hu
>>
>>
>>     _______________________________________________
>>     Linuxppc-dev mailing list
>>     Linuxppc-dev@lists.ozlabs.org  <mailto:Linuxppc-dev@lists.ozlabs.org>
>>     https://lists.ozlabs.org/listinfo/linuxppc-dev
>
>


[-- Attachment #2: Type: text/html, Size: 5868 bytes --]

Re: [question] Can the execution of the atomtic operation instruction pair lwarx/stwcx be interrrupted by local HW interruptions?

[Gavin Hu]

[-- Attachment #1: Type: text/plain, Size: 2366 bytes --]

Get it. Thanks!  :)


BR
Gavin. Hu


On Mon, Jan 6, 2014 at 1:27 PM, wyang <w90p710@gmail.com> wrote:

>
> On 01/06/2014 11:41 AM, Gavin Hu wrote:
>
>  Thanks your response.  :)
> But that means that these optimitive operations like atomic_add() aren't
> optimitive actully in PPC architecture, right? Becuase they can be
> interrupted by loacl HW interrupts. Theoretically, the ISR also can access
> the atomic gloable variable.
>
>
> Nope, my understand is that if you wanna sync kernel primitive code with
> ISR, you have responsibility to disable local interrupts. atomic_add does
> not guarantee to handle such case.
>
> Thanks
> Wei
>
>
>
>
> The following codes are complete atomic_inc() copied from arch/
> static __inline__ void atomic_add(int a, atomic_t *v)
> {
>     int t;
>
>     __asm__ __volatile__(
> "1:    lwarx    %0,0,%3        # atomic_add\n\
>     add    %0,%2,%0\n"
>     PPC405_ERR77(0,%3)
> "    stwcx.    %0,0,%3 \n\
>     bne-    1b"
>     : "=&r" (t), "+m" (v->counter)
>     : "r" (a), "r" (&v->counter)
>     : "cc");
> }
>
>
>  BR
>  Gavin. Hu
>
>
> On Mon, Dec 30, 2013 at 9:54 AM, wyang <w90p710@gmail.com> wrote:
>
>>  On 12/28/2013 01:41 PM, Gavin Hu wrote:
>>
>> Hi
>>
>> I notice that there is a pair ppc instructions lwarx and stwcx used to
>> atomtic operation for instance, atomic_inc/atomic_dec.
>>
>>  In some ppc manuals, they more emphasize its mechanism is that lwarx
>> can reseve the target memory address preventing other CORE from modifying
>> it.
>>
>>  I assume that there is atomtic operation executing on the CORE0 in a
>> multicore system. In this situation, does the CORE0 disable the local HW
>> interrupt?
>>  Can the executing process from the beginning of lwarx and end of stwcx
>> be interrupted by HW interruptions/exceptions?  Anyway, they are two
>> assembly instructions.
>>
>>
>>  It should just like other arch, the processor should response any
>> interrupt after the execution of a instruction, so the local HW interrupt
>> is not disabled.
>>
>> Thanks
>> Wei
>>
>>
>>  Thanks a lot!
>>
>> "1:    lwarx    %0,0,%2        # atomic_inc\n\
>>     addic    %0,%0,1\n"
>> "    stwcx.    %0,0,%2 \n\
>>
>>
>>  BR
>>  Gavin. Hu
>>
>>
>>  _______________________________________________
>> Linuxppc-dev mailing listLinuxppc-dev@lists.ozlabs.orghttps://lists.ozlabs.org/listinfo/linuxppc-dev
>>
>>
>>
>
>

[-- Attachment #2: Type: text/html, Size: 5799 bytes --]

Re: [question] Can the execution of the atomtic operation instruction pair lwarx/stwcx be interrrupted by local HW interruptions?

[Gavin Hu]

[-- Attachment #1: Type: text/plain, Size: 2484 bytes --]

So, these primitive funcitons like atomic_add() and so on also can't
prevent processes schedule switch on local CPU core? right?

Thanks!


BR
Gvain. Hu


On Mon, Jan 6, 2014 at 1:27 PM, wyang <w90p710@gmail.com> wrote:

>
> On 01/06/2014 11:41 AM, Gavin Hu wrote:
>
>  Thanks your response.  :)
> But that means that these optimitive operations like atomic_add() aren't
> optimitive actully in PPC architecture, right? Becuase they can be
> interrupted by loacl HW interrupts. Theoretically, the ISR also can access
> the atomic gloable variable.
>
>
> Nope, my understand is that if you wanna sync kernel primitive code with
> ISR, you have responsibility to disable local interrupts. atomic_add does
> not guarantee to handle such case.
>
> Thanks
> Wei
>
>
>
>
> The following codes are complete atomic_inc() copied from arch/
> static __inline__ void atomic_add(int a, atomic_t *v)
> {
>     int t;
>
>     __asm__ __volatile__(
> "1:    lwarx    %0,0,%3        # atomic_add\n\
>     add    %0,%2,%0\n"
>     PPC405_ERR77(0,%3)
> "    stwcx.    %0,0,%3 \n\
>     bne-    1b"
>     : "=&r" (t), "+m" (v->counter)
>     : "r" (a), "r" (&v->counter)
>     : "cc");
> }
>
>
>  BR
>  Gavin. Hu
>
>
> On Mon, Dec 30, 2013 at 9:54 AM, wyang <w90p710@gmail.com> wrote:
>
>>  On 12/28/2013 01:41 PM, Gavin Hu wrote:
>>
>> Hi
>>
>> I notice that there is a pair ppc instructions lwarx and stwcx used to
>> atomtic operation for instance, atomic_inc/atomic_dec.
>>
>>  In some ppc manuals, they more emphasize its mechanism is that lwarx
>> can reseve the target memory address preventing other CORE from modifying
>> it.
>>
>>  I assume that there is atomtic operation executing on the CORE0 in a
>> multicore system. In this situation, does the CORE0 disable the local HW
>> interrupt?
>>  Can the executing process from the beginning of lwarx and end of stwcx
>> be interrupted by HW interruptions/exceptions?  Anyway, they are two
>> assembly instructions.
>>
>>
>>  It should just like other arch, the processor should response any
>> interrupt after the execution of a instruction, so the local HW interrupt
>> is not disabled.
>>
>> Thanks
>> Wei
>>
>>
>>  Thanks a lot!
>>
>> "1:    lwarx    %0,0,%2        # atomic_inc\n\
>>     addic    %0,%0,1\n"
>> "    stwcx.    %0,0,%2 \n\
>>
>>
>>  BR
>>  Gavin. Hu
>>
>>
>>  _______________________________________________
>> Linuxppc-dev mailing listLinuxppc-dev@lists.ozlabs.orghttps://lists.ozlabs.org/listinfo/linuxppc-dev
>>
>>
>>
>
>

[-- Attachment #2: Type: text/html, Size: 5930 bytes --]

Re: [question] Can the execution of the atomtic operation instruction pair lwarx/stwcx be interrrupted by local HW interruptions?

[wyang]

[-- Attachment #1: Type: text/plain, Size: 3063 bytes --]


On 01/06/2014 02:24 PM, Gavin Hu wrote:
> So, these primitive funcitons like atomic_add() and so on also can't 
> prevent processes schedule switch on local CPU core? right?

You are right!

BR
Wei
>
> Thanks!
>
>
> BR
> Gvain. Hu
>
>
> On Mon, Jan 6, 2014 at 1:27 PM, wyang <w90p710@gmail.com 
> <mailto:w90p710@gmail.com>> wrote:
>
>
>     On 01/06/2014 11:41 AM, Gavin Hu wrote:
>>     Thanks your response.  :)
>>     But that means that these optimitive operations like atomic_add()
>>     aren't optimitive actully in PPC architecture, right? Becuase
>>     they can be interrupted by loacl HW interrupts. Theoretically,
>>     the ISR also can access the atomic gloable variable.
>
>     Nope, my understand is that if you wanna sync kernel primitive
>     code with ISR, you have responsibility to disable local
>     interrupts. atomic_add does not guarantee to handle such case.
>
>     Thanks
>     Wei
>
>
>>
>>
>>     The following codes are complete atomic_inc() copied from arch/
>>     static __inline__ void atomic_add(int a, atomic_t *v)
>>     {
>>         int t;
>>
>>         __asm__ __volatile__(
>>     "1:    lwarx    %0,0,%3        # atomic_add\n\
>>         add    %0,%2,%0\n"
>>         PPC405_ERR77(0,%3)
>>     "    stwcx.    %0,0,%3 \n\
>>         bne-    1b"
>>         : "=&r" (t), "+m" (v->counter)
>>         : "r" (a), "r" (&v->counter)
>>         : "cc");
>>     }
>>
>>
>>     BR
>>     Gavin. Hu
>>
>>
>>     On Mon, Dec 30, 2013 at 9:54 AM, wyang <w90p710@gmail.com
>>     <mailto:w90p710@gmail.com>> wrote:
>>
>>         On 12/28/2013 01:41 PM, Gavin Hu wrote:
>>>         Hi
>>>
>>>         I notice that there is a pair ppc instructions lwarx and
>>>         stwcx used to atomtic operation for instance,
>>>         atomic_inc/atomic_dec.
>>>
>>>         In some ppc manuals, they more emphasize its mechanism is
>>>         that lwarx can reseve the target memory address preventing
>>>         other CORE from modifying it.
>>>
>>>         I assume that there is atomtic operation executing on the
>>>         CORE0 in a multicore system. In this situation, does the
>>>         CORE0 disable the local HW interrupt?
>>>         Can the executing process from the beginning of lwarx and
>>>         end of stwcx be interrupted by HW interruptions/exceptions? 
>>>         Anyway, they are two assembly instructions.
>>
>>         It should just like other arch, the processor should response
>>         any interrupt after the execution of a instruction, so the
>>         local HW interrupt is not disabled.
>>
>>         Thanks
>>         Wei
>>>
>>>          Thanks a lot!
>>>
>>>         "1:    lwarx    %0,0,%2        # atomic_inc\n\
>>>             addic    %0,%0,1\n"
>>>         "    stwcx.    %0,0,%2 \n\
>>>
>>>
>>>         BR
>>>         Gavin. Hu
>>>
>>>
>>>         _______________________________________________
>>>         Linuxppc-dev mailing list
>>>         Linuxppc-dev@lists.ozlabs.org  <mailto:Linuxppc-dev@lists.ozlabs.org>
>>>         https://lists.ozlabs.org/listinfo/linuxppc-dev
>>
>>
>
>


[-- Attachment #2: Type: text/html, Size: 9361 bytes --]

Re: [question] Can the execution of the atomtic operation instruction pair lwarx/stwcx be interrrupted by local HW interruptions?

[Scott Wood]

On Mon, 2014-01-06 at 13:27 +0800, wyang wrote:
> 
> On 01/06/2014 11:41 AM, Gavin Hu wrote:
> 
> > Thanks your response.  :) 
> > But that means that these optimitive operations like atomic_add()
> > aren't optimitive actully in PPC architecture, right? Becuase they
> > can be interrupted by loacl HW interrupts. Theoretically, the ISR
> > also can access the atomic gloable variable.
> > 
> 
> Nope, my understand is that if you wanna sync kernel primitive code
> with ISR, you have responsibility to disable local interrupts.
> atomic_add does not guarantee to handle such case.

atomic_add() and other atomics do handle that case.  Interrupts are not
disabled, but there's a stwcx. in the interrupt return code to make sure
the reservation gets cleared.

-Scott

Re: [question] Can the execution of the atomtic operation instruction pair lwarx/stwcx be interrrupted by local HW interruptions?

[wyang]

On 01/07/2014 06:05 AM, Scott Wood wrote:
> On Mon, 2014-01-06 at 13:27 +0800, wyang wrote:
>> On 01/06/2014 11:41 AM, Gavin Hu wrote:
>>
>>> Thanks your response.  :)
>>> But that means that these optimitive operations like atomic_add()
>>> aren't optimitive actully in PPC architecture, right? Becuase they
>>> can be interrupted by loacl HW interrupts. Theoretically, the ISR
>>> also can access the atomic gloable variable.
>>>
>> Nope, my understand is that if you wanna sync kernel primitive code
>> with ISR, you have responsibility to disable local interrupts.
>> atomic_add does not guarantee to handle such case.
> atomic_add() and other atomics do handle that case.  Interrupts are not
> disabled, but there's a stwcx. in the interrupt return code to make sure
> the reservation gets cleared.

Yeah, Can you provide more detail info about why they can handle that 
case? The following is my understand:

Let us assume that there is a atomic global variable(var_a) and its 
initial value is 0.

The kernel attempts to execute atomic_add(1, var_a), after lwarx a async 
interrupt happens, and the ISR also accesses "var_a" variable and 
executes atomic_add.

static __inline__ void atomic_add(int a, atomic_t *v)
{
     int t;

     __asm__ __volatile__(
"1:    lwarx    %0,0,%3        # atomic_add\n\
----------------------------------  <----------- interrupt 
happens------->        ISR also operates this global variable "var_a" 
such as also executing atomic_add(1, var_a). so the
               var_a would is 1.
     add    %0,%2,%0\n"
     PPC405_ERR77(0,%3)
"    stwcx.    %0,0,%3 \n\ <----- After interrupt code returns, the 
reservation is cleared. so CR0 is not equal to 0, and then jump the 1 
label. the var_a will be 2.
     bne-    1b"
     : "=&r" (t), "+m" (v->counter)
     : "r" (a), "r" (&v->counter)
     : "cc");
}

So the value of var_a is 2 rather than 1. Thats why i said that 
atomic_add does not handle such case. If I miss something, please 
correct me.:-)

Wei
>
> -Scott
>
>
>

Re: [question] Can the execution of the atomtic operation instruction pair lwarx/stwcx be interrrupted by local HW interruptions?

[Scott Wood]

On Tue, 2014-01-07 at 09:00 +0800, wyang wrote:
> Yeah, Can you provide more detail info about why they can handle that 
> case? The following is my understand:
> 
> Let us assume that there is a atomic global variable(var_a) and its 
> initial value is 0.
> 
> The kernel attempts to execute atomic_add(1, var_a), after lwarx a async 
> interrupt happens, and the ISR also accesses "var_a" variable and 
> executes atomic_add.
> 
> static __inline__ void atomic_add(int a, atomic_t *v)
> {
>      int t;
> 
>      __asm__ __volatile__(
> "1:    lwarx    %0,0,%3        # atomic_add\n\
> ----------------------------------  <----------- interrupt 
> happens------->        ISR also operates this global variable "var_a" 
> such as also executing atomic_add(1, var_a). so the
>                var_a would is 1.
>      add    %0,%2,%0\n"
>      PPC405_ERR77(0,%3)
> "    stwcx.    %0,0,%3 \n\ <----- After interrupt code returns, the 
> reservation is cleared. so CR0 is not equal to 0, and then jump the 1 
> label. the var_a will be 2.
>      bne-    1b"
>      : "=&r" (t), "+m" (v->counter)
>      : "r" (a), "r" (&v->counter)
>      : "cc");
> }
> 
> So the value of var_a is 2 rather than 1. Thats why i said that 
> atomic_add does not handle such case. If I miss something, please 
> correct me.:-)

2 is the correct result, since atomic_add(1, var_a) was called twice
(once in the ISR, once in the interrupted context).

-Scott

Re: [question] Can the execution of the atomtic operation instruction pair lwarx/stwcx be interrrupted by local HW interruptions?

[wyang]

On 01/07/2014 02:35 PM, Scott Wood wrote:
> On Tue, 2014-01-07 at 09:00 +0800, wyang wrote:
>> Yeah, Can you provide more detail info about why they can handle that
>> case? The following is my understand:
>>
>> Let us assume that there is a atomic global variable(var_a) and its
>> initial value is 0.
>>
>> The kernel attempts to execute atomic_add(1, var_a), after lwarx a async
>> interrupt happens, and the ISR also accesses "var_a" variable and
>> executes atomic_add.
>>
>> static __inline__ void atomic_add(int a, atomic_t *v)
>> {
>>       int t;
>>
>>       __asm__ __volatile__(
>> "1:    lwarx    %0,0,%3        # atomic_add\n\
>> ----------------------------------  <----------- interrupt
>> happens------->        ISR also operates this global variable "var_a"
>> such as also executing atomic_add(1, var_a). so the
>>                 var_a would is 1.
>>       add    %0,%2,%0\n"
>>       PPC405_ERR77(0,%3)
>> "    stwcx.    %0,0,%3 \n\ <----- After interrupt code returns, the
>> reservation is cleared. so CR0 is not equal to 0, and then jump the 1
>> label. the var_a will be 2.
>>       bne-    1b"
>>       : "=&r" (t), "+m" (v->counter)
>>       : "r" (a), "r" (&v->counter)
>>       : "cc");
>> }
>>
>> So the value of var_a is 2 rather than 1. Thats why i said that
>> atomic_add does not handle such case. If I miss something, please
>> correct me.:-)
> 2 is the correct result, since atomic_add(1, var_a) was called twice
> (once in the ISR, once in the interrupted context).
Scott, thanks for your confirmation. I guess that Gavin thought that 1 
is a correct result. So thats why I said that if he wanna get 1,
he should have responsibility to disable local interrupts. I mean that 
atomic_add is not able to guarantee that 1 is a correct result.:-)

Wei
>
> -Scott
>
>
>

Re: [question] Can the execution of the atomtic operation instruction pair lwarx/stwcx be interrrupted by local HW interruptions?

[Scott Wood]

On Tue, 2014-01-07 at 15:22 +0800, wyang wrote:
> On 01/07/2014 02:35 PM, Scott Wood wrote:
> > On Tue, 2014-01-07 at 09:00 +0800, wyang wrote:
> >> Yeah, Can you provide more detail info about why they can handle that
> >> case? The following is my understand:
> >>
> >> Let us assume that there is a atomic global variable(var_a) and its
> >> initial value is 0.
> >>
> >> The kernel attempts to execute atomic_add(1, var_a), after lwarx a async
> >> interrupt happens, and the ISR also accesses "var_a" variable and
> >> executes atomic_add.
> >>
> >> static __inline__ void atomic_add(int a, atomic_t *v)
> >> {
> >>       int t;
> >>
> >>       __asm__ __volatile__(
> >> "1:    lwarx    %0,0,%3        # atomic_add\n\
> >> ----------------------------------  <----------- interrupt
> >> happens------->        ISR also operates this global variable "var_a"
> >> such as also executing atomic_add(1, var_a). so the
> >>                 var_a would is 1.
> >>       add    %0,%2,%0\n"
> >>       PPC405_ERR77(0,%3)
> >> "    stwcx.    %0,0,%3 \n\ <----- After interrupt code returns, the
> >> reservation is cleared. so CR0 is not equal to 0, and then jump the 1
> >> label. the var_a will be 2.
> >>       bne-    1b"
> >>       : "=&r" (t), "+m" (v->counter)
> >>       : "r" (a), "r" (&v->counter)
> >>       : "cc");
> >> }
> >>
> >> So the value of var_a is 2 rather than 1. Thats why i said that
> >> atomic_add does not handle such case. If I miss something, please
> >> correct me.:-)
> > 2 is the correct result, since atomic_add(1, var_a) was called twice
> > (once in the ISR, once in the interrupted context).
> Scott, thanks for your confirmation. I guess that Gavin thought that 1 
> is a correct result. So thats why I said that if he wanna get 1,
> he should have responsibility to disable local interrupts.

If you disable interrupts, that will just delay the interrupt until
afterward, at which point the interrupt will increment var_a to 2.

> I mean that 
> atomic_add is not able to guarantee that 1 is a correct result.:-)

Well, no.  It's atomic_add(), not set_var_to_one(). :-)

-Scott