Re: [RFC 3/4] Documentation/scheduler/sched-deadline.txt: Some notes on EDF schedulability

2015-04-09 Thread Luca Abeni
On 04/09/2015 12:10 PM, Henrik Austad wrote: [...] @@ -43,7 +43,13 @@ CONTENTS "deadline", to schedule tasks. A SCHED_DEADLINE task should receive "runtime" microseconds of execution time every "period" microseconds, and these "runtime" microseconds are available within "deadline"

Re: [RFC 3/4] Documentation/scheduler/sched-deadline.txt: Some notes on EDF schedulability

2015-04-09 Thread Henrik Austad
On Thu, Apr 09, 2015 at 11:34:37AM +0200, Luca Abeni wrote: > Hi Henrik, > > On 04/09/2015 11:06 AM, Henrik Austad wrote: > >On Wed, Apr 08, 2015 at 01:59:39PM +0200, Luca Abeni wrote: > [...] > >Also, density is equivalent > >to 'utilization', right? (which is referred to in sec 4.1 > No; the

Re: [RFC 3/4] Documentation/scheduler/sched-deadline.txt: Some notes on EDF schedulability

2015-04-09 Thread Luca Abeni
Hi Henrik, On 04/09/2015 11:06 AM, Henrik Austad wrote: On Wed, Apr 08, 2015 at 01:59:39PM +0200, Luca Abeni wrote: Add a short discussion about sufficient and necessary schedulability tests, and a simple example showing that if D_i != P_i then density based tests are only sufficient. Also add

Re: [RFC 3/4] Documentation/scheduler/sched-deadline.txt: Some notes on EDF schedulability

2015-04-09 Thread Henrik Austad
On Wed, Apr 08, 2015 at 01:59:39PM +0200, Luca Abeni wrote: > Add a short discussion about sufficient and necessary schedulability tests, > and a simple example showing that if D_i != P_i then density based tests > are only sufficient. > Also add some references to scientific papers on

Re: [RFC 3/4] Documentation/scheduler/sched-deadline.txt: Some notes on EDF schedulability

2015-04-09 Thread Henrik Austad
On Thu, Apr 09, 2015 at 11:34:37AM +0200, Luca Abeni wrote: Hi Henrik, On 04/09/2015 11:06 AM, Henrik Austad wrote: On Wed, Apr 08, 2015 at 01:59:39PM +0200, Luca Abeni wrote: [...] Also, density is equivalent to 'utilization', right? (which is referred to in sec 4.1 No; the utilisation

Re: [RFC 3/4] Documentation/scheduler/sched-deadline.txt: Some notes on EDF schedulability

2015-04-09 Thread Luca Abeni
Hi Henrik, On 04/09/2015 11:06 AM, Henrik Austad wrote: On Wed, Apr 08, 2015 at 01:59:39PM +0200, Luca Abeni wrote: Add a short discussion about sufficient and necessary schedulability tests, and a simple example showing that if D_i != P_i then density based tests are only sufficient. Also add

Re: [RFC 3/4] Documentation/scheduler/sched-deadline.txt: Some notes on EDF schedulability

2015-04-09 Thread Henrik Austad
On Wed, Apr 08, 2015 at 01:59:39PM +0200, Luca Abeni wrote: Add a short discussion about sufficient and necessary schedulability tests, and a simple example showing that if D_i != P_i then density based tests are only sufficient. Also add some references to scientific papers on schedulability

Re: [RFC 3/4] Documentation/scheduler/sched-deadline.txt: Some notes on EDF schedulability

2015-04-09 Thread Luca Abeni
On 04/09/2015 12:10 PM, Henrik Austad wrote: [...] @@ -43,7 +43,13 @@ CONTENTS deadline, to schedule tasks. A SCHED_DEADLINE task should receive runtime microseconds of execution time every period microseconds, and these runtime microseconds are available within deadline microseconds -

[RFC 3/4] Documentation/scheduler/sched-deadline.txt: Some notes on EDF schedulability

2015-04-08 Thread Luca Abeni
Add a short discussion about sufficient and necessary schedulability tests, and a simple example showing that if D_i != P_i then density based tests are only sufficient. Also add some references to scientific papers on schedulability tests for EDF that are both necessary and sufficient, and on

[RFC 3/4] Documentation/scheduler/sched-deadline.txt: Some notes on EDF schedulability

2015-04-08 Thread Luca Abeni
Add a short discussion about sufficient and necessary schedulability tests, and a simple example showing that if D_i != P_i then density based tests are only sufficient. Also add some references to scientific papers on schedulability tests for EDF that are both necessary and sufficient, and on