From mboxrd@z Thu Jan 1 00:00:00 1970 Subject: Re: [Adeos-main] what should __adeos_set_timer(hz) do? From: Philippe Gerum In-Reply-To: <200406291242.32834.maillist@domain.hid> References: <200406291242.32834.maillist@domain.hid> Content-Type: text/plain Message-Id: <1088507709.636.14.camel@domain.hid> Mime-Version: 1.0 Date: Tue, 29 Jun 2004 13:15:09 +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: Thomas Kuhlmann Cc: adeos-main@gna.org On Tue, 2004-06-29 at 12:42, Thomas Kuhlmann wrote: > Hello, > > based on the adeos-linux-2.4.22-armnommu-r12.patch for uClinux 2.4.22-uc0 from > http://download.gna.org/adeos/patches/v2.4/armnommu/ I'm trying to extend > (port) this patch for armnommu/arch-netarm/ (NetSilicon NS7520 processor). > > Has anybody tried a patch for armnommu/arch-netarm (just to prevent me doing > it again)? > > The function __adeos_set_time(unsigned long hz) has to be ported to netarm. > However what should it do? > In the given patch for the armnommu/arch-atmel/ it looks like the kernel timer > is disabled, the frequency is changed, and the timer is enabled again (not > one shot, but periodic mode). Is this the correct semantics of > __adeos_set_timer(hz)? > Strictly speaking, setting the hw timer in periodic mode is not required by Adeos. __adeos_set_timer() is a helper for the adeos_tune_timer() service that is expected to set the Linux timer hw to a given frequency. This helper is totally ARM-dependent and does not even exist for other archs. > Sorry if my question is misplaced here and I'd better contact the ARM > responsible persons (let me know of this). > Your question indeed belongs to this list. Maybe the ARM guru (who I'm not) is listening? > Thanks, Thomas. -- Philippe.