From: shawn.guo@linaro.org (Shawn Guo)
To: linux-arm-kernel@lists.infradead.org
Subject: [PATCH V2 3/3] ARM: imx: source gpt per clk from OSC for system timer
Date: Wed, 10 Sep 2014 19:49:38 +0800 [thread overview]
Message-ID: <20140910114936.GE22579@dragon> (raw)
In-Reply-To: <7fd82b8c08c54f4dadb6b7402231a086@BN1PR0301MB0628.namprd03.prod.outlook.com>
On Wed, Sep 10, 2014 at 07:43:41AM +0000, Anson.Huang at freescale.com wrote:
> diff --git a/arch/arm/mach-imx/clk-imx6q.c b/arch/arm/mach-imx/clk-imx6q.c index 4ee6e77a0fdf..3f0401e27b38 100644
> --- a/arch/arm/mach-imx/clk-imx6q.c
> +++ b/arch/arm/mach-imx/clk-imx6q.c
> @@ -245,6 +245,7 @@ static void __init imx6q_clocks_init(struct device_node *ccm_node)
> clk[IMX6QDL_CLK_PLL3_80M] = imx_clk_fixed_factor("pll3_80m", "pll3_usb_otg", 1, 6);
> clk[IMX6QDL_CLK_PLL3_60M] = imx_clk_fixed_factor("pll3_60m", "pll3_usb_otg", 1, 8);
> clk[IMX6QDL_CLK_TWD] = imx_clk_fixed_factor("twd", "arm", 1, 2);
> + clk[IMX6QDL_CLK_GPT_3M] = imx_clk_fixed_factor("gpt_3m", "osc", 1, 8);
> if (cpu_is_imx6dl()) {
> clk[IMX6QDL_CLK_GPU2D_AXI] = imx_clk_fixed_factor("gpu2d_axi", "mmdc_ch0_axi_podf", 1, 1);
> clk[IMX6QDL_CLK_GPU3D_AXI] = imx_clk_fixed_factor("gpu3d_axi", "mmdc_ch0_axi_podf", 1, 1); @@ -469,6 +470,13 @@ static void __init imx6q_clocks_init(struct device_node *ccm_node)
>
> clk_register_clkdev(clk[IMX6QDL_CLK_ENET_REF], "enet_ref", NULL);
>
> + /*
> + * The gpt_3m clock is not available on i.MX6Q TO1.0. Let's point it
> + * to clock gpt_ipg_per to ease the gpt driver code.
> + */
> + if (cpu_is_imx6q() && imx_get_soc_revision() == IMX_CHIP_REVISION_1_0)
> + clk[IMX6QDL_CLK_GPT_3M] = clk[IMX6QDL_CLK_GPT_IPG_PER];
> +
> if ((imx_get_soc_revision() != IMX_CHIP_REVISION_1_0) ||
> cpu_is_imx6dl()) {
> clk_set_parent(clk[IMX6QDL_CLK_LDB_DI0_SEL], clk[IMX6QDL_CLK_PLL5_VIDEO_DIV]);
>
> diff --git a/arch/arm/mach-imx/time.c b/arch/arm/mach-imx/time.c index bf92e5a351c0..c0ad839516b0 100644
> --- a/arch/arm/mach-imx/time.c
> +++ b/arch/arm/mach-imx/time.c
> @@ -312,10 +317,21 @@ static void __init _mxc_timer_init(int irq,
> __raw_writel(0, timer_base + MXC_TCTL);
> __raw_writel(0, timer_base + MXC_TPRER); /* see datasheet note */
>
> - if (timer_is_v2())
> - tctl_val = V2_TCTL_CLK_PER | V2_TCTL_FRR | V2_TCTL_WAITEN | MXC_TCTL_TEN;
> - else
> + if (timer_is_v2()) {
> + tctl_val = V2_TCTL_FRR | V2_TCTL_WAITEN | MXC_TCTL_TEN;
> + if (clk_get_rate(clk_per) == V2_TIMER_RATE_OSC_DIV8) {
> /* Is the assumption corrrect ??? */
> [Anson] I am afraid that i.MX6Q TO1.0 can NOT meet it, as there is no gpt_3m available for i.MX6Q TO1.0, but we have it in dtb and clk driver.
>
> + tctl_val |= V2_TCTL_CLK_OSC_DIV8;
> + if (cpu_is_imx6dl() || cpu_is_imx6sx()) {
> + /* 24 / 8 = 3 MHz */
> + __raw_writel(7 << V2_TPRER_PRE24M, timer_base + MXC_TPRER);
> + tctl_val |= V2_TCTL_24MEN;
> + }
> + } else {
> + tctl_val |= V2_TCTL_CLK_PER;
> + }
> + } else {
> tctl_val = MX1_2_TCTL_FRR | MX1_2_TCTL_CLK_PCLK1 | MXC_TCTL_TEN;
> + }
>
> __raw_writel(tctl_val, timer_base + MXC_TCTL);
>
> @@ -349,9 +365,13 @@ static void __init mxc_timer_init_dt(struct device_node *np)
> WARN_ON(!timer_base);
> irq = irq_of_parse_and_map(np, 0);
>
> - clk_per = of_clk_get_by_name(np, "per");
> clk_ipg = of_clk_get_by_name(np, "ipg");
>
> + /* Try osc_per clock first, and fall back to per clock otherwise */
> + clk_per = of_clk_get_by_name(np, "osc_per");
> + if (!IS_ERR(clk_per))
> + clk_per = of_clk_get_by_name(np, "per");
> +
> [Anson] For i.MX6Q TO1.0, there is no GPT_3M clock for GPT, but in dtb, there is osc_per, so clk_per will be initialized by osc_per, and the clk rate read in _mxc_timer_init function will be 3000000 and go to the GPT_3M clk source path, which will NOT work for i.MX6Q TO1.0.
>
I had the following change for clk-imx6q.c as above. Shouldn't it
handle the i.MX6Q TO1.0 case well?
if (cpu_is_imx6q() && imx_get_soc_revision() == IMX_CHIP_REVISION_1_0)
clk[IMX6QDL_CLK_GPT_3M] = clk[IMX6QDL_CLK_GPT_IPG_PER];
Shawn
> _mxc_timer_init(irq, clk_per, clk_ipg); } CLOCKSOURCE_OF_DECLARE(mx1_timer, "fsl,imx1-gpt", mxc_timer_init_dt);
>
next prev parent reply other threads:[~2014-09-10 11:49 UTC|newest]
Thread overview: 12+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-09-05 3:26 [PATCH V2 0/3] move gpt per clk parent from ipg_per to OSC Anson Huang
2014-09-05 3:26 ` [PATCH V2 1/3] ARM: imx: add gpt_3m clk for i.mx6qdl Anson Huang
2014-09-05 3:26 ` [PATCH V2 2/3] ARM: dts: imx6: make gpt per clock can be from OSC Anson Huang
2014-09-05 3:26 ` [PATCH V2 3/3] ARM: imx: source gpt per clk from OSC for system timer Anson Huang
2014-09-05 12:09 ` Fabio Estevam
2014-09-05 12:58 ` Anson.Huang at freescale.com
2014-09-10 7:33 ` Shawn Guo
2014-09-10 7:43 ` Anson.Huang at freescale.com
2014-09-10 11:49 ` Shawn Guo [this message]
[not found] ` <8DDD46F6-AA96-4E14-ADD6-45A1AD12423C@freescale.com>
2014-09-10 13:01 ` Shawn Guo
2014-09-10 7:36 ` Shawn Guo
2014-09-10 7:45 ` Anson.Huang at freescale.com
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20140910114936.GE22579@dragon \
--to=shawn.guo@linaro.org \
--cc=linux-arm-kernel@lists.infradead.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).