[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <CAOMZO5CfokeopsZuf2j905hzsU7cJ4x5iEyzujnFXr+EjD7nEg@mail.gmail.com>
Date: Thu, 15 Sep 2022 11:10:06 -0300
From: Fabio Estevam <festevam@...il.com>
To: Han Xu <han.xu@....com>
Cc: Abel Vesa <abelvesa@...nel.org>,
Michael Turquette <mturquette@...libre.com>,
Stephen Boyd <sboyd@...nel.org>,
Shawn Guo <shawnguo@...nel.org>,
Sascha Hauer <s.hauer@...gutronix.de>,
Pengutronix Kernel Team <kernel@...gutronix.de>,
NXP Linux Team <linux-imx@....com>,
"open list:NXP i.MX CLOCK DRIVERS" <linux-clk@...r.kernel.org>,
"moderated list:ARM/FREESCALE IMX / MXC ARM ARCHITECTURE"
<linux-arm-kernel@...ts.infradead.org>,
open list <linux-kernel@...r.kernel.org>,
miquel.raynal@...tlin.com, linux-mtd@...ts.infradead.org
Subject: Re: [PATCH] clk: imx: imx6sx: remove the SET_RATE_PARENT flag for
QSPI clocks
Hi Han Xu,
On Wed, Sep 14, 2022 at 5:51 PM Fabio Estevam <festevam@...il.com> wrote:
>
> Hi Han Xu,
>
> On Wed, Sep 14, 2022 at 5:11 PM Han Xu <han.xu@....com> wrote:
> >
> > There is no dedicate parent clock for QSPI so SET_RATE_PARENT flag
> > should not be used. For instance, the default parent clock for QSPI is
> > pll2_bus, which is also the parent clock for quite a few modules, such
> > as MMDC, once GPMI NAND set clock rate for EDO5 mode can cause system
> > hang due to pll2_bus rate changed.
>
> Thanks a lot for your patch.
>
> This fixes the kernel hang issue on a custom imx6sx board with NAND without the
> need of using the workaround found in the NXP kernel:
>
> https://source.codeaurora.org/external/imx/linux-imx/commit/drivers/mtd/nand/raw/gpmi-nand/gpmi-nand.c?h=lf-5.15.y&id=d03eb99c393f8732f70a1d7d29a3b9c42cccbe48
>
> Tested-by: Fabio Estevam <festevam@...x.de>
Just noticed the build error. I manually applied it to 5.15, so that's
why I didn't notice the problem
with your original patch. It is missing the closing parenthesis:
--- a/drivers/clk/imx/clk-imx6sx.c
+++ b/drivers/clk/imx/clk-imx6sx.c
@@ -280,13 +280,13 @@ static void __init imx6sx_clocks_init(struct
device_node *ccm_node)
hws[IMX6SX_CLK_SSI3_SEL] =
imx_clk_hw_mux("ssi3_sel", base + 0x1c, 14, 2,
ssi_sels, ARRAY_SIZE(ssi_sels));
hws[IMX6SX_CLK_SSI2_SEL] =
imx_clk_hw_mux("ssi2_sel", base + 0x1c, 12, 2,
ssi_sels, ARRAY_SIZE(ssi_sels));
hws[IMX6SX_CLK_SSI1_SEL] =
imx_clk_hw_mux("ssi1_sel", base + 0x1c, 10, 2,
ssi_sels, ARRAY_SIZE(ssi_sels));
- hws[IMX6SX_CLK_QSPI1_SEL] =
imx_clk_hw_mux_flags("qspi1_sel", base + 0x1c, 7, 3, qspi1_sels,
ARRAY_SIZE(qspi1_sels), CLK_SET_RATE_PARENT);
+ hws[IMX6SX_CLK_QSPI1_SEL] =
imx_clk_hw_mux("qspi1_sel", base + 0x1c, 7, 3, qspi1_sels,
ARRAY_SIZE(qspi1_sels));
hws[IMX6SX_CLK_PERCLK_SEL] =
imx_clk_hw_mux("perclk_sel", base + 0x1c, 6, 1,
perclk_sels, ARRAY_SIZE(perclk_sels));
hws[IMX6SX_CLK_VID_SEL] = imx_clk_hw_mux("vid_sel",
base + 0x20, 21, 3, vid_sels,
ARRAY_SIZE(vid_sels));
hws[IMX6SX_CLK_ESAI_SEL] =
imx_clk_hw_mux("esai_sel", base + 0x20, 19, 2,
audio_sels, ARRAY_SIZE(audio_sels));
hws[IMX6SX_CLK_CAN_SEL] = imx_clk_hw_mux("can_sel",
base + 0x20, 8, 2, can_sels,
ARRAY_SIZE(can_sels));
hws[IMX6SX_CLK_UART_SEL] =
imx_clk_hw_mux("uart_sel", base + 0x24, 6, 1,
uart_sels, ARRAY_SIZE(uart_sels));
- hws[IMX6SX_CLK_QSPI2_SEL] =
imx_clk_hw_mux_flags("qspi2_sel", base + 0x2c, 15, 3, qspi2_sels,
ARRAY_SIZE(qspi2_sels), CLK_SET_RATE_PARENT);
+ hws[IMX6SX_CLK_QSPI2_SEL] =
imx_clk_hw_mux("qspi2_sel", base + 0x2c, 15, 3, qspi2_sels,
ARRAY_SIZE(qspi2_sels));
hws[IMX6SX_CLK_SPDIF_SEL] =
imx_clk_hw_mux("spdif_sel", base + 0x30, 20, 2,
audio_sels, ARRAY_SIZE(audio_sels));
hws[IMX6SX_CLK_AUDIO_SEL] =
imx_clk_hw_mux("audio_sel", base + 0x30, 7, 2,
audio_sels, ARRAY_SIZE(audio_sels));
hws[IMX6SX_CLK_ENET_PRE_SEL] =
imx_clk_hw_mux("enet_pre_sel", base + 0x34, 15, 3,
enet_pre_sels, ARRAY_SIZE(enet_pre_sels));
Powered by blists - more mailing lists