From mboxrd@z Thu Jan 1 00:00:00 1970 From: Simon Glass Date: Tue, 16 May 2017 19:38:00 -0600 Subject: [U-Boot] [PATCH] board_f: skip timer_init() on Coldfire archs In-Reply-To: <7dc9f309-9e8f-db79-c331-a3c18ba6b717@sysam.it> References: <20170510215806.31659-1-angelo@sysam.it> <7dc9f309-9e8f-db79-c331-a3c18ba6b717@sysam.it> Message-ID: List-Id: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: u-boot@lists.denx.de Hi Angelo, On 10 May 2017 at 16:36, Angelo Dureghello wrote: > Hi Simon, > > > On 11/05/2017 00:03, Simon Glass wrote: >> >> Hi Angelo, >> >> On 10 May 2017 at 15:58, Angelo Dureghello wrote: >>> >>> Coldfire arch is not happy with timer_init since interrupt handlers >>> are still not set at that stage, and the boot hangs silently. >>> >>> Signed-off-by: Angelo Dureghello >>> --- >>> common/board_f.c | 2 ++ >>> 1 file changed, 2 insertions(+) >>> >>> diff --git a/common/board_f.c b/common/board_f.c >>> index d9431ee79a..30e588e213 100644 >>> --- a/common/board_f.c >>> +++ b/common/board_f.c >>> @@ -740,7 +740,9 @@ static const init_fnc_t init_sequence_f[] = { >>> /* get CPU and bus clocks according to the environment variable >>> */ >>> get_clocks, /* get CPU and bus clocks (etc.) */ >>> #endif >>> +#if !defined(CONFIG_M68K) >>> timer_init, /* initialize timer */ >>> +#endif >>> #if defined(CONFIG_BOARD_POSTCLK_INIT) >>> board_postclk_init, >>> #endif >>> -- >>> 2.11.0 >>> >> >> I'm really hoping we can get rid of all arch-specific things from the >> init sequence. >> > > Yes, i have seen you unified that step for all archs, and unfortunately i > discovered the issue now only updating u-boot on my mcf5307 based board. > >> Is there no way that m68k can init its timer here? Or perhaps it could >> be a nop function? > > > I checked now all the cpu/xxx/start.S. At that early stage, all the > vector handlers are set to _fault thats is just and endless loop. > > In Coldfire arch, timer_init() is implemented in lib/time.c, as to > setup and start the system timer overflow interrupt. It is called later > from board_init_r(), line 863, after interrupt_init(). > > So, no :( unless you don't have some suggestion, i don't see any easy > way to keep that timer_init() call enabled in board_init_f(). My only ideas are: - user driver model for timer and then call the interrupt init from your driver's probe() method - make timer_init() a nop for you arch Regards, Simon