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] [day] [month] [year] [list]
Date:   Fri, 15 Mar 2019 11:22:10 +0800
From:   Nicolas Boichat <drinkcat@...omium.org>
To:     Yongqiang Niu <yongqiang.niu@...iatek.com>
Cc:     CK Hu <ck.hu@...iatek.com>, Philipp Zabel <p.zabel@...gutronix.de>,
        David Airlie <airlied@...ux.ie>,
        Rob Herring <robh+dt@...nel.org>,
        Mark Rutland <mark.rutland@....com>,
        Matthias Brugger <matthias.bgg@...il.com>,
        devicetree@...r.kernel.org, lkml <linux-kernel@...r.kernel.org>,
        dri-devel@...ts.freedesktop.org,
        "moderated list:ARM/Mediatek SoC support" 
        <linux-mediatek@...ts.infradead.org>,
        linux-arm Mailing List <linux-arm-kernel@...ts.infradead.org>
Subject: Re: [PATCH 03/18] drm/mediatek: redefine mtk_ddp_sout_sel

On Fri, Mar 15, 2019 at 10:06 AM Yongqiang Niu
<yongqiang.niu@...iatek.com> wrote:
>
> On Tue, 2018-12-25 at 11:57 +0800, Nicolas Boichat wrote:
> > On Mon, Dec 24, 2018 at 6:52 PM Yongqiang Niu
> > <yongqiang.niu@...iatek.com> wrote:
> > >
> > > This patch redefine mtk_ddp_sout_sel
> >
> > Can you describe a bit more why you are making this change?
>
> the format of "mtk_ddp_sout_sel"was not flexible, after we add more
> mediatek SOC support, that will be redundant
>
> set this function format like mtk_ddp_mout_en and mtk_ddp_sel_in

This needs to be 2 patches:
 1. Make the change to "cur == DDP_COMPONENT_BLS && next ==
DDP_COMPONENT_DPI0", along with an explanation of why this is
reasonable.
 2. Change the format to look like mtk_ddp_mout_en and mtk_ddp_sel_in

> >
> > > Signed-off-by: Yongqiang Niu <yongqiang.niu@...iatek.com>
> > > ---
> > >  drivers/gpu/drm/mediatek/mtk_drm_ddp.c | 32 ++++++++++++++++++++------------
> > >  1 file changed, 20 insertions(+), 12 deletions(-)
> > >
> > > diff --git a/drivers/gpu/drm/mediatek/mtk_drm_ddp.c b/drivers/gpu/drm/mediatek/mtk_drm_ddp.c
> > > index adb37e4..592f852 100644
> > > --- a/drivers/gpu/drm/mediatek/mtk_drm_ddp.c
> > > +++ b/drivers/gpu/drm/mediatek/mtk_drm_ddp.c
> > > @@ -405,21 +405,27 @@ static unsigned int mtk_ddp_sel_in(enum mtk_ddp_comp_id cur,
> > >         return value;
> > >  }
> > >
> > > -static void mtk_ddp_sout_sel(void __iomem *config_regs,
> > > -                            enum mtk_ddp_comp_id cur,
> > > -                            enum mtk_ddp_comp_id next)
> > > +static unsigned int mtk_ddp_sout_sel(void __iomem *config_regs,
> >
> > You don't use config_regs anymore, drop it.
>
> ok, will drop it in next version
>
> >
> > > +                                    enum mtk_ddp_comp_id cur,
> > > +                                    enum mtk_ddp_comp_id next,
> > > +                                    unsigned int *addr)
> > >  {
> > > +       unsigned int value;
> > > +
> > >         if (cur == DDP_COMPONENT_BLS && next == DDP_COMPONENT_DSI0) {
> > > -               writel_relaxed(BLS_TO_DSI_RDMA1_TO_DPI1,
> > > -                              config_regs + DISP_REG_CONFIG_OUT_SEL);
> > > +               *addr = DISP_REG_CONFIG_OUT_SEL;
> > > +               value = BLS_TO_DSI_RDMA1_TO_DPI1;
> >
> > You can directly return BLS_TO_DSI_RDMA1_TO_DPI1.
>
> just format this like mtk_ddp_mout_en and mtk_ddp_sel_in

Since there is precedent, sounds ok.

> >
> > >         } else if (cur == DDP_COMPONENT_BLS && next == DDP_COMPONENT_DPI0) {
> > > -               writel_relaxed(BLS_TO_DPI_RDMA1_TO_DSI,
> > > -                              config_regs + DISP_REG_CONFIG_OUT_SEL);
> > > -               writel_relaxed(DSI_SEL_IN_RDMA,
> > > -                              config_regs + DISP_REG_CONFIG_DSI_SEL);
> > > -               writel_relaxed(DPI_SEL_IN_BLS,
> > > -                              config_regs + DISP_REG_CONFIG_DPI_SEL);
> > > +               *addr = DISP_REG_CONFIG_OUT_SEL;
> > > +               value = BLS_TO_DPI_RDMA1_TO_DSI;
> >
> > I (kind of) understand the change above, as you still end up writing
> > BLS_TO_DSI_RDMA1_TO_DPI1 in DISP_REG_CONFIG_OUT_SEL.
> >
> > This changes the behaviour, as now you only write
> > BLS_TO_DPI_RDMA1_TO_DSI to DISP_REG_CONFIG_OUT_SEL, but the previous
> > revision of the code would also write to DISP_REG_CONFIG_DSI_SEL and
> > DISP_REG_CONFIG_DPI_SEL. Why?
> >
>
> DISP_REG_CONFIG_DSI_SEL set in the next lines.

Ok, but where is mtk_ddp_sout_sel(DDP_COMPONENT_RDMA1,
DDP_COMPONENT_DSI0) called from?

Before this change, we just needed to call:
mtk_ddp_sout_sel(DDP_COMPONENT_BLS, DDP_COMPONENT_DSI0)
and the following 3 registers would be modified:
BLS_TO_DPI_RDMA1_TO_DSI, DSI_SEL_IN_RDMA, DPI_SEL_IN_BLS.

Now, to get similar behaviour, we need to call:
mtk_ddp_sout_sel(DDP_COMPONENT_BLS, DDP_COMPONENT_DSI0)
mtk_ddp_sout_sel(DDP_COMPONENT_RDMA1, DDP_COMPONENT_DSI0)

Maybe that's ok, but you really need to explain the reason for this
change in the commit message.

> DPI_SEL_IN_BLS is 0 for DISP_REG_CONFIG_DPI_SEL set, and hardware
> default setting is also 0, so this one is no need anymore

That's somewhat reasonable, but this needs to be at the very least
described in the commit message.

> > > +       } else if (cur == DDP_COMPONENT_RDMA1 && next == DDP_COMPONENT_DSI0) {
> > > +               *addr = DISP_REG_CONFIG_DSI_SEL;
> > > +               value = DSI_SEL_IN_RDMA;
> > > +       } else {
> > > +               value = 0;
> > >         }
> > > +
> > > +       return value;
> > >  }
> > >
> > >  void mtk_ddp_add_comp_to_path(void __iomem *config_regs,
> > > @@ -434,7 +440,9 @@ void mtk_ddp_add_comp_to_path(void __iomem *config_regs,
> > >                 writel_relaxed(reg, config_regs + addr);
> > >         }
> > >
> > > -       mtk_ddp_sout_sel(config_regs, cur, next);
> > > +       value = mtk_ddp_sout_sel(cur, next, &addr);
> > > +       if (value)
> > > +               writel_relaxed(value, config_regs + addr);
> >
> > Why this change? I don't see mtk_ddp_sout_sel being used later in the
> > series, so I'm not sure why we don't directly write the value into the
> > register.
> >
> in the patch "[PATCH 04/18] drm/mediatek: move rdma sout from
> mtk_ddp_mout_en into mtk_ddp_sout_sel", i moved all rdma out to here,
> rdma only have single out, no multi out.
> if keep this format, there will many writel_relaxed in mtk_ddp_sout_sel.
> and modify this format like mtk_ddp_mout_en and mtk_ddp_sel_in looks
> better.
>
> >
> >
> > >         value = mtk_ddp_sel_in(cur, next, &addr);
> > >         if (value) {
> > > --
> > > 1.8.1.1.dirty
> > > _______________________________________________
> > > dri-devel mailing list
> > > dri-devel@...ts.freedesktop.org
> > > https://lists.freedesktop.org/mailman/listinfo/dri-devel
>
>

Powered by blists - more mailing lists