linux-kernel - Re: [PATCH v4 01/12] clk: mediatek: Disable tuner

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <1551239460.1047.30.camel@mtksdaap41>
Date:   Wed, 27 Feb 2019 11:51:00 +0800
From:   Weiyi Lu <weiyi.lu@...iatek.com>
To:     Matthias Brugger <matthias.bgg@...il.com>
CC:     Nicolas Boichat <drinkcat@...omium.org>,
        Stephen Boyd <sboyd@...eaurora.org>,
        Rob Herring <robh@...nel.org>,
        James Liao <jamesjj.liao@...iatek.com>,
        Fan Chen <fan.chen@...iatek.com>,
        <linux-arm-kernel@...ts.infradead.org>,
        <linux-kernel@...r.kernel.org>,
        <linux-mediatek@...ts.infradead.org>, <linux-clk@...r.kernel.org>,
        <srv_heupstream@...iatek.com>, <stable@...r.kernel.org>,
        Owen Chen <owen.chen@...iatek.com>
Subject: Re: [PATCH v4 01/12] clk: mediatek: Disable tuner_en before change
 PLL rate

On Tue, 2019-02-26 at 16:59 +0100, Matthias Brugger wrote:
> 
> On 01/02/2019 09:30, Weiyi Lu wrote:
> > From: Owen Chen <owen.chen@...iatek.com>
> > 
> > PLLs with tuner_en bit, such as APLL1, need to disable
> > tuner_en before apply new frequency settings, or the new frequency
> > settings (pcw) will not be applied.
> > The tuner_en bit will be disabled during changing PLL rate
> > and be restored after new settings applied.
> > Another minor change is to correct the macro name of pcw change bit
> > to CON1_PCW_CHG because PCW_CHG(BIT31) is on CON1.
> > 
> > Fixes: e2f744a82d725 (clk: mediatek: Add MT2712 clock support)
> > Cc: <stable@...r.kernel.org>
> > Signed-off-by: Owen Chen <owen.chen@...iatek.com>
> > ---
> >  drivers/clk/mediatek/clk-pll.c | 33 +++++++++++++++++++++++++++++++--
> >  1 file changed, 31 insertions(+), 2 deletions(-)
> > 
> > diff --git a/drivers/clk/mediatek/clk-pll.c b/drivers/clk/mediatek/clk-pll.c
> > index f54e4015b0b1..f0ff5f535c7e 100644
> > --- a/drivers/clk/mediatek/clk-pll.c
> > +++ b/drivers/clk/mediatek/clk-pll.c
> > @@ -27,7 +27,7 @@
> >  #define CON0_BASE_EN		BIT(0)
> >  #define CON0_PWR_ON		BIT(0)
> >  #define CON0_ISO_EN		BIT(1)
> > -#define CON0_PCW_CHG		BIT(31)
> > +#define CON1_PCW_CHG		BIT(31)
> >  
> >  #define AUDPLL_TUNER_EN		BIT(31)
> >  
> > @@ -93,9 +93,31 @@ static void mtk_pll_set_rate_regs(struct mtk_clk_pll *pll, u32 pcw,
> >  {
> >  	u32 con1, val;
> >  	int pll_en;
> > +	u32 tuner_en = 0;
> > +	u32 tuner_en_mask;
> > +	void __iomem *tuner_en_addr = NULL;
> >  
> >  	pll_en = readl(pll->base_addr + REG_CON0) & CON0_BASE_EN;
> >  
> > +	/* disable tuner */
> > +	if (pll->tuner_en_addr) {
> > +		tuner_en_addr = pll->tuner_en_addr;
> > +		tuner_en_mask = BIT(pll->data->tuner_en_bit);
> > +	} else if (pll->tuner_addr) {
> > +		tuner_en_addr = pll->tuner_addr;
> > +		tuner_en_mask = AUDPLL_TUNER_EN;
> > +	}
> > +
> > +	if (tuner_en_addr) {
> > +		val = readl(tuner_en_addr);
> > +		tuner_en = val & tuner_en_mask;
> > +
> > +		if (tuner_en) {
> > +			val &= ~tuner_en_mask;
> > +			writel(val, tuner_en_addr);
> > +		}
> > +	}
> > +
> 
> Why don't we use a flag here, it would make the code easier to understand.
> I think it would also help if you put this code in a separate function.
> 
> Regards,
> Matthias
> 

Hi Matthias,

I guess you suggest to add a flag (e.g. CLK_TUNER_SUPPORT) and declare
the PLL clock with this flag. And inside this mtk_pll_set_rate_regs()
function, we could modify it as below. Am I right?

if (pll->data->flags & CLK_TUNER_SUPPORT)
    disable_tuner();
[...]
if (pll->data->flags & CLK_TUNER_SUPPORT)
    enable_tuner();


But I thought this flag might be a customized flags and can only be put
in clk-mtk.h not clk-provider.h so far. And it might cause 
potential problem of clock flag re-definition. May I have more
suggestion?

> 
> >  	/* set postdiv */
> >  	val = readl(pll->pd_addr);
> >  	val &= ~(POSTDIV_MASK << pll->data->pd_shift);
> > @@ -116,12 +138,19 @@ static void mtk_pll_set_rate_regs(struct mtk_clk_pll *pll, u32 pcw,
> >  	con1 = readl(pll->base_addr + REG_CON1);
> >  
> >  	if (pll_en)
> > -		con1 |= CON0_PCW_CHG;
> > +		con1 |= CON1_PCW_CHG;
> >  
> >  	writel(con1, pll->base_addr + REG_CON1);
> >  	if (pll->tuner_addr)
> >  		writel(con1 + 1, pll->tuner_addr);
> >  
> > +	/* restore tuner_en */
> > +	if (tuner_en_addr && tuner_en) {
> > +		val = readl(tuner_en_addr);
> > +		val |= tuner_en_mask;
> > +		writel(val, tuner_en_addr);
> > +	}
> > +
> >  	if (pll_en)
> >  		udelay(20);
> >  }
> >