From mboxrd@z Thu Jan 1 00:00:00 1970 From: Tomasz Figa Subject: Re: [PATCH 1/5] ARM: mm: cache-l2x0: Add base address argument to write_sec callback Date: Wed, 11 Jun 2014 18:07:24 +0200 Message-ID: <53987EBC.70307@samsung.com> References: <1402500612-4778-1-git-send-email-t.figa@samsung.com> <1402500612-4778-2-git-send-email-t.figa@samsung.com> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 7bit Return-path: In-reply-to: Sender: linux-kernel-owner@vger.kernel.org To: Jon Loeliger Cc: linux-samsung-soc@vger.kernel.org, Kukjin Kim , Laura Abbott , Tony Lindgren , Linus Walleij , linux-kernel@vger.kernel.org, Tomasz Figa , Santosh Shilimkar , Robin Holt , Russell King , linux-omap@vger.kernel.org, linux-arm-kernel@lists.infradead.org List-Id: linux-omap@vger.kernel.org On 11.06.2014 18:00, Jon Loeliger wrote: >> diff --git a/arch/arm/include/asm/mach/arch.h b/arch/arm/include/asm/mach/arch.h >> index 060a75e..ddaebcd 100644 >> --- a/arch/arm/include/asm/mach/arch.h >> +++ b/arch/arm/include/asm/mach/arch.h >> @@ -46,7 +46,8 @@ struct machine_desc { >> enum reboot_mode reboot_mode; /* default restart mode */ >> unsigned l2c_aux_val; /* L2 cache aux value */ >> unsigned l2c_aux_mask; /* L2 cache aux mask */ >> - void (*l2c_write_sec)(unsigned long, unsigned); >> + void (*l2c_write_sec)(void __iomem *, >> + unsigned long, unsigned); >> struct smp_operations *smp; /* SMP operations */ >> bool (*smp_init)(void); >> void (*fixup)(struct tag *, char **); > >> diff --git a/arch/arm/mm/cache-l2x0.c b/arch/arm/mm/cache-l2x0.c >> index efc5cab..1695eab 100644 >> --- a/arch/arm/mm/cache-l2x0.c >> +++ b/arch/arm/mm/cache-l2x0.c >> @@ -72,7 +72,7 @@ static void l2c_write_sec(unsigned long val, void __iomem *base, unsigned reg) >> if (val == readl_relaxed(base + reg)) >> return; >> if (outer_cache.write_sec) >> - outer_cache.write_sec(val, reg); >> + outer_cache.write_sec(base, val, reg); >> else >> writel_relaxed(val, base + reg); >> } > > The parameter order (base, val, reg) seems very non-intuitive. > Are you matching some existing prototype or adhering to some > backwards compatibility issue? If not wouldn't, say, (base, reg, val) > or (val, base, reg) be more intuitive? Hmm, I didn't think too much about this, so this order is just whatever first came to my mind, probably because I'm used to xxx_write(ctx, val, reg) accessors found in many drivers. Anyway, l2c_write_sec() in arm/mm/cache-l2x0.c, which calls outer_cache.write_sec(), already uses (val, base, reg) convention, so probably this one would be most suitable. I'll change in v2. Best regards, Tomasz