| 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
| ||
|
Message-ID: <CAP6Zq1jOSKpzFvto1LMs=JftLK0fxrrg+73Sh34GunuLfcAfEA@mail.gmail.com>
Date: Mon, 22 May 2023 15:56:25 +0300
From: Tomer Maimon <tmaimon77@...il.com>
To: Christophe JAILLET <christophe.jaillet@...adoo.fr>
Cc: mturquette@...libre.com, sboyd@...nel.org, avifishman70@...il.com,
tali.perry1@...il.com, joel@....id.au, venture@...gle.com,
yuenn@...gle.com, benjaminfair@...gle.com,
openbmc@...ts.ozlabs.org, linux-clk@...r.kernel.org,
linux-kernel@...r.kernel.org
Subject: Re: [PATCH v16 1/1] clk: npcm8xx: add clock controller
Hi Christophe,
Thanks for your comments
On Sun, 21 May 2023 at 20:21, Christophe JAILLET
<christophe.jaillet@...adoo.fr> wrote:
>
> Le 21/05/2023 à 16:32, Tomer Maimon a écrit :
> > Nuvoton Arbel BMC NPCM8XX contains an integrated clock controller which
> > generates and supplies clocks to all modules within the BMC.
> >
> > Signed-off-by: Tomer Maimon <tmaimon77@...il.com>
>
> Hi,
>
> should there be a v17, below a few nits and questions.
Let's wait for Stephen and see if he has other comments.
>
> Jusrt my 2c,
>
> > ---
> > drivers/clk/Kconfig | 8 +
> > drivers/clk/Makefile | 1 +
> > drivers/clk/clk-npcm8xx.c | 566 ++++++++++++++++++++++++++++++++++++++
> > 3 files changed, 575 insertions(+)
> > create mode 100644 drivers/clk/clk-npcm8xx.c
> >
> > diff --git a/drivers/clk/Kconfig b/drivers/clk/Kconfig
> > index 016814e15536..2249de28a46a 100644
> > --- a/drivers/clk/Kconfig
> > +++ b/drivers/clk/Kconfig
> > @@ -325,6 +325,14 @@ config COMMON_CLK_LOCHNAGAR
> > This driver supports the clocking features of the Cirrus Logic
> > Lochnagar audio development board.
> >
> > +config COMMON_CLK_NPCM8XX
> > + tristate "Clock driver for the NPCM8XX SoC Family"
> > + depends on ARCH_NPCM || COMPILE_TEST
> > + help
> > + This driver supports the clocks on the Nuvoton BMC NPCM8XX SoC Family,
> > + all the clocks are initialized by the bootloader, so this driver
> > + allows only reading of current settings directly from the hardware.
> > +
> > config COMMON_CLK_LOONGSON2
> > bool "Clock driver for Loongson-2 SoC"
> > depends on LOONGARCH || COMPILE_TEST
> > diff --git a/drivers/clk/Makefile b/drivers/clk/Makefile
> > index 0aebef17edc6..88713cbfbab5 100644
> > --- a/drivers/clk/Makefile
> > +++ b/drivers/clk/Makefile
> > @@ -51,6 +51,7 @@ obj-$(CONFIG_ARCH_MILBEAUT_M10V) += clk-milbeaut.o
> > obj-$(CONFIG_ARCH_MOXART) += clk-moxart.o
> > obj-$(CONFIG_ARCH_NOMADIK) += clk-nomadik.o
> > obj-$(CONFIG_ARCH_NPCM7XX) += clk-npcm7xx.o
> > +obj-$(CONFIG_COMMON_CLK_NPCM8XX) += clk-npcm8xx.o
> > obj-$(CONFIG_ARCH_NSPIRE) += clk-nspire.o
> > obj-$(CONFIG_COMMON_CLK_OXNAS) += clk-oxnas.o
> > obj-$(CONFIG_COMMON_CLK_PALMAS) += clk-palmas.o
> > diff --git a/drivers/clk/clk-npcm8xx.c b/drivers/clk/clk-npcm8xx.c
> > new file mode 100644
> > index 000000000000..9e6ed073d124
> > --- /dev/null
> > +++ b/drivers/clk/clk-npcm8xx.c
> > @@ -0,0 +1,566 @@
> > +// SPDX-License-Identifier: GPL-2.0
> > +/*
> > + * Nuvoton NPCM8xx Clock Generator
> > + * All the clocks are initialized by the bootloader, so this driver allow only
>
> nit: allows (as in Kconfig)
will modify in V17
>
> [...]
>
> > +/* npcm8xx clock registers*/
> > +#define NPCM8XX_CLKSEL (0x04)
> > +#define NPCM8XX_CLKDIV1 (0x08)
> > +#define NPCM8XX_CLKDIV2 (0x2C)
> > +#define NPCM8XX_CLKDIV3 (0x58)
> > +#define NPCM8XX_CLKDIV4 (0x7C)
> > +#define NPCM8XX_PLLCON0 (0x0C)
> > +#define NPCM8XX_PLLCON1 (0x10)
> > +#define NPCM8XX_PLLCON2 (0x54)
> > +#define NPCM8XX_PLLCONG (0x60)
> > +#define NPCM8XX_THRTL_CNT (0xC0)
>
> nit: () around this values looks superfluous.
I am not sure, it can work both ways.
>
> > +
> > +#define PLLCON_LOKI BIT(31)
> > +#define PLLCON_LOKS BIT(30)
>
> [...]
>
> > +static struct clk_hw *
> > +npcm8xx_clk_register_pll(struct device *dev, void __iomem *pllcon,
> > + const char *name, const struct clk_parent_data *parent,
> > + unsigned long flags)
> > +{
> > + struct npcm8xx_clk_pll *pll;
> > + struct clk_init_data init = {};
> > + int ret;
> > +
> > + pll = kzalloc(sizeof(*pll), GFP_KERNEL);
>
> Everything looks devm_()'ed in this driver, except this kzalloc.
> Except the one below, there is no kfree to free this memory, and no
> .remove() function.
Also clk_hw_register_divider_parent_data doesn't use devm_
about free the pll, we use it, return at the end of the function.
about adding remove, we had a dissection about it in V4, since the
clock is a service driver it shouldn't be removed.
https://patchwork.kernel.org/project/linux-watchdog/patch/20220621131424.162355-7-tmaimon77@gmail.com/
>
> Is it on purpose?
>
> > + if (!pll)
> > + return ERR_PTR(-ENOMEM);
> > +
> > + init.name = name;
> > + init.ops = &npcm8xx_clk_pll_ops;
> > + init.parent_data = parent;
> > + init.num_parents = 1;
> > + init.flags = flags;
> > +
> > + pll->pllcon = pllcon;
> > + pll->hw.init = &init;
> > +
> > + ret = devm_clk_hw_register(dev, &pll->hw);
> > + if (ret) {
> > + kfree(pll);
> > + return ERR_PTR(ret);
> > + }
> > +
> > + return &pll->hw;
> > +}
> > +
> > +static DEFINE_SPINLOCK(npcm8xx_clk_lock);
> > +
> > +static int npcm8xx_clk_probe(struct platform_device *pdev)
> > +{
> > + struct clk_hw_onecell_data *npcm8xx_clk_data;
> > + struct device *dev = &pdev->dev;
> > + void __iomem *clk_base;
> > + struct resource *res;
> > + struct clk_hw *hw;
> > + unsigned int i;
> > + int err;
> > +
> > + npcm8xx_clk_data = devm_kzalloc(dev, struct_size(npcm8xx_clk_data, hws,
> > + NPCM8XX_NUM_CLOCKS),
> > + GFP_KERNEL);
> > + if (!npcm8xx_clk_data)
> > + return -ENOMEM;
> > +
> > + res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
> > + clk_base = devm_ioremap(dev, res->start, resource_size(res));
> > + if (!clk_base) {
> > + dev_err(&pdev->dev, "Failed to remap I/O memory\n");
>
> Here and below: mostly a matter of taste, but usually return
> dev_err_probe() saves a few LoC and keep the error code in the error
> message.
Thanks, but I prefer the shorter way :-)
>
> > + return -ENOMEM;
> > + }
> > +
> > + npcm8xx_clk_data->num = NPCM8XX_NUM_CLOCKS;
> > +
> > + for (i = 0; i < NPCM8XX_NUM_CLOCKS; i++)
> > + npcm8xx_clk_data->hws[i] = ERR_PTR(-EPROBE_DEFER);
> > +
> > + /* Register plls */
> > + for (i = 0; i < ARRAY_SIZE(npcm8xx_pll_clks); i++) {
> > + struct npcm8xx_pll_data *pll_clk = &npcm8xx_pll_clks[i];
> > +
> > + hw = npcm8xx_clk_register_pll(dev, clk_base + pll_clk->reg,
> > + pll_clk->name, &pll_clk->parent,
> > + pll_clk->flags);
> > + if (IS_ERR(hw)) {
> > + dev_err(dev, "npcm8xx_clk: Can't register pll\n");
> > + return PTR_ERR(hw);
> > + }
> > + pll_clk->hw = *hw;
> > + }
> > +
> > + /* Register fixed dividers */
> > + hw = devm_clk_hw_register_fixed_factor(dev, NPCM8XX_CLK_S_PLL1_DIV2,
> > + NPCM8XX_CLK_S_PLL1, 0, 1, 2);
> > + if (IS_ERR(hw)) {
> > + dev_err(dev, "npcm8xx_clk: Can't register fixed div\n");
> > + return PTR_ERR(hw);
> > + }
> > + hw_pll1_div2 = *hw;
> > +
> > + hw = devm_clk_hw_register_fixed_factor(dev, NPCM8XX_CLK_S_PLL2_DIV2,
> > + NPCM8XX_CLK_S_PLL2, 0, 1, 2);
> > + if (IS_ERR(hw)) {
> > + dev_err(dev, "npcm8xx_clk: Can't register pll2 div2\n");
> > + return PTR_ERR(hw);
> > + }
> > + hw_pll2_div2 = *hw;
> > +
> > + hw = devm_clk_hw_register_fixed_factor(dev, NPCM8XX_CLK_S_PLL_GFX_DIV2,
> > + NPCM8XX_CLK_S_PLL_GFX, 0, 1, 2);
> > + if (IS_ERR(hw)) {
> > + dev_err(dev, "npcm8xx_clk: Can't register gfx div2\n");
> > + return PTR_ERR(hw);
> > + }
> > + hw_gfx_div2 = *hw;
> > +
> > + /* Register muxes */
> > + for (i = 0; i < ARRAY_SIZE(npcm8xx_muxes); i++) {
> > + struct npcm8xx_clk_mux_data *mux_data = &npcm8xx_muxes[i];
> > +
> > + hw = devm_clk_hw_register_mux_parent_data_table(dev,
> > + mux_data->name,
> > + mux_data->parent_data,
> > + mux_data->num_parents,
> > + mux_data->flags,
> > + clk_base + NPCM8XX_CLKSEL,
> > + mux_data->shift,
> > + mux_data->mask,
> > + 0,
> > + mux_data->table,
> > + &npcm8xx_clk_lock);
> > + if (IS_ERR(hw)) {
> > + dev_err(dev, "npcm8xx_clk: Can't register mux\n");
> > + return PTR_ERR(hw);
> > + }
> > + mux_data->hw = *hw;
> > + }
> > +
> > + hw = devm_clk_hw_register_fixed_factor(dev, NPCM8XX_CLK_S_PRE_CLK,
> > + NPCM8XX_CLK_S_CPU_MUX, 0, 1, 2);
> > + if (IS_ERR(hw)) {
> > + dev_err(dev, "npcm8xx_clk: Can't register ckclk div2\n");
>
> ckclk or just clk?
will modify in V17
>
> Based on naming only, this div2 (here and below) looks strange.
> Is it on purpose, on just a cut'n'paste from the previous message?
>
> > + return PTR_ERR(hw);
> > + }
> > + hw_pre_clk = *hw;
> > +
> > + hw = devm_clk_hw_register_fixed_factor(dev, NPCM8XX_CLK_S_AXI,
> > + NPCM8XX_CLK_S_TH, 0, 1, 2);
> > + if (IS_ERR(hw)) {
> > + dev_err(dev, "npcm8xx_clk: Can't register axi div2\n");
> > + return PTR_ERR(hw);
> > + }
> > + npcm8xx_clk_data->hws[NPCM8XX_CLK_AXI] = hw;
> > +
> > + hw = devm_clk_hw_register_fixed_factor(dev, NPCM8XX_CLK_S_ATB,
> > + NPCM8XX_CLK_S_AXI, 0, 1, 2);
> > + if (IS_ERR(hw)) {
> > + dev_err(dev, "npcm8xx_clk: Can't register atb div2\n");
> > + return PTR_ERR(hw);
> > + }
> > + npcm8xx_clk_data->hws[NPCM8XX_CLK_ATB] = hw;
> > +
> > + /* Register clock dividers specified in npcm8xx_divs */
> > + for (i = 0; i < ARRAY_SIZE(npcm8xx_divs); i++) {
> > + const struct npcm8xx_clk_div_data *div_data = &npcm8xx_divs[i];
> > +
> > + hw = clk_hw_register_divider_parent_data(dev, div_data->name,
> > + &div_data->parent_data,
> > + div_data->flags,
> > + clk_base + div_data->reg,
> > + div_data->shift,
> > + div_data->width,
> > + div_data->clk_divider_flags,
> > + &npcm8xx_clk_lock);
> > + if (IS_ERR(hw)) {
> > + dev_err(dev, "npcm8xx_clk: Can't register div table\n");
> > + goto err_div_clk;
> > + }
> > +
> > + if (div_data->onecell_idx >= 0)
> > + npcm8xx_clk_data->hws[div_data->onecell_idx] = hw;
> > + }
> > +
> > + err = devm_of_clk_add_hw_provider(dev, of_clk_hw_onecell_get,
> > + npcm8xx_clk_data);
> > + if (err) {
> > + dev_err(dev, "unable to add clk provider\n");
> > + hw = ERR_PTR(err);
> > + goto err_div_clk;
> > + }
> > +
> > + return err;
> > +
> > +err_div_clk:
> > + while (i--) {
> > + if (npcm8xx_divs[i].onecell_idx >= 0)
> > + clk_hw_unregister_divider(npcm8xx_clk_data->hws[npcm8xx_divs[i].onecell_idx]);
> > + }
>
> There is an error handling path in the probe, but no .remove() function.
> Should there be one?
The same answer as above (V4)
>
> CJ
>
> [...]
Best regards,
Tomer
Powered by blists - more mailing lists