Message ID | 4EC6A8E3.7050302@esd.eu |
---|---|
State | Changes Requested |
Headers | show |
Dear Reinhard Arlt, In message <4EC6A8E3.7050302@esd.eu> you wrote: > From: Reinhard Arlt <reinhard.arlt@esd.eu> > > Do not enable the data cache in start.S, the decrementer do not work. > > Signed-off-by: Reinhard Arlt <reinhard.arlt@esd.eu> > > -- > > diff --git a/arch/powerpc/cpu/74xx_7xx/start.S b/arch/powerpc/cpu/74xx_7xx/start.S > index 75fb773..131fc14 100644 > --- a/arch/powerpc/cpu/74xx_7xx/start.S > +++ b/arch/powerpc/cpu/74xx_7xx/start.S > @@ -247,11 +247,12 @@ in_flash: > /* enable address translation */ > bl enable_addr_trans > sync > - > +#if !defined(CONFIG_CPCI750) > /* enable and invalidate the data cache */ > bl l1dcache_enable > sync > #endif > +#endif > #ifdef CONFIG_SYS_INIT_RAM_LOCK > bl lock_ram_in_cache > sync Sorry, but I don't want to have board specific code in arch/powerpc/cpu/74xx_7xx/start.S Please turn this into a feature dependent define, if it is really needed. Also, can you provide a rational explanation how the decrementer depends on data cache being enabled? This sounds pretty strange to me. Best regards, Wolfgang Denk
diff --git a/arch/powerpc/cpu/74xx_7xx/start.S b/arch/powerpc/cpu/74xx_7xx/start.S index 75fb773..131fc14 100644 --- a/arch/powerpc/cpu/74xx_7xx/start.S +++ b/arch/powerpc/cpu/74xx_7xx/start.S @@ -247,11 +247,12 @@ in_flash: /* enable address translation */ bl enable_addr_trans sync - +#if !defined(CONFIG_CPCI750) /* enable and invalidate the data cache */ bl l1dcache_enable sync #endif +#endif #ifdef CONFIG_SYS_INIT_RAM_LOCK bl lock_ram_in_cache sync