From mboxrd@z Thu Jan 1 00:00:00 1970 From: Gilles Chanteperdrix MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Message-ID: <17593.4142.482000.586615@domain.hid> Date: Sat, 15 Jul 2006 17:56:30 +0200 Subject: Re: [Xenomai-core] [PATCH] optimise syscall mux-code calculation In-Reply-To: <44B8D412.1080002@domain.hid> References: <44B62AE7.8080503@domain.hid> <1152930271.5027.69.camel@domain.hid> <44B8A1E9.7080004@domain.hid> <44B8D412.1080002@domain.hid> List-Id: "Xenomai life and development \(bug reports, patches, discussions\)" List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Jan Kiszka Cc: xenomai-core Jan Kiszka wrote: > Jan Kiszka wrote: > > ... > > Ok, I will rework this patch. > > > > And here is -v2 (for post 2.2 trunk). It keeps the ABI but changes the > API wrt __xeno_mux_code. As this change gets widely caught by > xeno_user_skin_init, the modifications are limited. > > At this chance I also converted the rtdm library to the generic > initialisation code. I tweaked nucleus/bind.h a bit so that one can > control via XENO_DISABLE_MLOCKALL_DETECTION if the related code gets > built into the lib. This appeared to be reasonable to me because librtdm > doesn't create its own RT threads, thus the check should be better left > to those skins who do so. Anyone any better idea, or is this acceptable? > If a separate patch is preferred, I will break out this change. Another reason why rtdm did not use the generic initialization code is that RTDM did not call exit upon failure to bind, and let a later call to open fail. I do not know if it was desirable, but it was a feature of the previous implementation. -- Gilles Chanteperdrix.