Re: [HACK] convert i_alloc_sem for direct_io.c craziness!

!MAILaRCHIVE_VOTE_RePLACE
Previous message: [thread] [date] [author]
Next message: [thread] [date] [author]
To: Zach Brown <zach.brown@...>
Cc: Steven Rostedt <rostedt@...>, LKML <linux-kernel@...>, linux-rt-users <linux-rt-users@...>, <mingo@...>, Thomas Gleixner <tglx@...>
Date: Monday, October 1, 2007 - 4:39 pm

On Mon, 2007-10-01 at 12:52 -0700, Zach Brown wrote:


Basically, anything that maps to a simple mutex. Anything more complex
gets real messy real quick.

Locks that have non-exclusive states become non-deterministic because an
unbounded number of contexts can be in this state. Hence acquisition of
the exclusive state has unbounded time. Even when limited to a bounded
number, the ramifications to the PI graph will get you a head-ache.

Also, non-owner locks, ie. semaphores (asymetric acquisition vs release
contexts) are unusable because the lack of ownership undermines PI - who
to boost?


-
Previous message: [thread] [date] [author]
Next message: [thread] [date] [author]

Messages in current thread:
[HACK] convert i_alloc_sem for direct_io.c craziness!, Steven Rostedt, (Mon Sep 24, 5:14 pm)
Re: [HACK] convert i_alloc_sem for direct_io.c craziness!, Peter Zijlstra, (Mon Oct 1, 4:39 pm)
Re: [HACK] convert i_alloc_sem for direct_io.c craziness!, Steven Rostedt, (Tue Oct 2, 2:12 pm)
Re: [HACK] convert i_alloc_sem for direct_io.c craziness!, Peter Zijlstra, (Tue Sep 25, 8:42 am)
[PATCH RT] Don't let -rt rw_semaphors do _non_owner locks, Steven Rostedt, (Tue Sep 25, 11:29 am)