|
|
|
| Why mutex_locks cannot be held across context-switches ? Can anyone
explain this .
| |
| David Schwartz 2004-12-21, 7:49 am |
|
"Mona" <indian.monalisa@gmail.com> wrote in message
news:1103619833.967517.293940@f14g2000cwb.googlegroups.com...
> Why mutex_locks cannot be held across context-switches ? Can anyone
> explain this .
They can. What are you talking about exactly?
DS
| |
| Måns Rullgård 2004-12-21, 7:49 am |
| "Mona" <indian.monalisa@gmail.com> writes:
> Why mutex_locks cannot be held across context-switches ? Can anyone
> explain this .
What sort of mutexes? I'm not aware of any with this limitation,
which, if I understand you correctly, would make them quite pointless.
--
Måns Rullgård
mru@inprovide.com
| |
|
| I was reading abot mutex_locks properities in SVR4 MP, where it was
documented saying mutex_locks are acquired and required on
context_switches. The reason for this we don't want each sleep
systemcall to store data related locking. I was looking for e.g
explaination.
| |
|
|
| Casper H.S. Dik 2004-12-21, 5:56 pm |
| "Mona" <indian.monalisa@gmail.com> writes:
>I was reading abot mutex_locks properities in SVR4 MP, where it was
>documented saying mutex_locks are acquired and required on
>context_switches. The reason for this we don't want each sleep
>systemcall to store data related locking. I was looking for e.g
>explaination.
That's a really odd implementation; you'd need to ask whoever did
SVR4 MP. (No, Solaris does not derive from SVR4 MP)
Casper
--
Expressed in this posting are my opinions. They are in no way related
to opinions held by my employer, Sun Microsystems.
Statements on Sun products included here are not gospel and may
be fiction rather than truth.
|
|
|
|