From mboxrd@z Thu Jan 1 00:00:00 1970 Subject: Re: [Adeos-main] adeos_tune_timer() From: Philippe Gerum In-Reply-To: <40C03A5E.7030006@domain.hid> References: <40C03A5E.7030006@domain.hid> Content-Type: text/plain Message-Id: <1086254126.680.15.camel@domain.hid> Mime-Version: 1.0 Date: Thu, 03 Jun 2004 11:15:26 +0200 Content-Transfer-Encoding: 7bit Sender: adeos-main-admin@domain.hid Errors-To: adeos-main-admin@domain.hid List-Help: List-Post: List-Subscribe: , List-Id: General discussion about Adeos List-Unsubscribe: , List-Archive: To: Wolfgang Grandegger Cc: adeos-main On Fri, 2004-06-04 at 11:01, Wolfgang Grandegger wrote: > Hello, > > the function adeos_tune_timer() allows to change the tick frequency of > the hardware timer. I wonder if Linux is aware of this change? Doesn't > it still believe that the timer interrupt comes at a rate of 100 Hz? > No, Linux is not aware of that. It's the burden of the domain tuning the timer to also intercept the timer interrupt and propagate one of them down the pipeline at the proper pace (e.g. 1 from 10 if the domain sets the timer for ticking at 10Khz, and Linux expects 1Khz), so that Linux is not affected by the change. Of course, this only works if the timer is tuned by a domain more prioritary than Linux, so such propagation can occur; less prioritary domains should not have to grab the timer anyway. IIRC, some example code in the tree illustrates this principle for the timer IRQ. > Wolfgang. > > _______________________________________________ > Adeos-main mailing list > Adeos-main@domain.hid > https://mail.gna.org/listinfo/adeos-main -- Philippe.