From: Andrea della Porta <andrea.porta@suse.com>
To: Florian Fainelli <florian.fainelli@broadcom.com>
Cc: Andrea della Porta <andrea.porta@suse.com>,
Ulf Hansson <ulf.hansson@linaro.org>,
Rob Herring <robh@kernel.org>,
Krzysztof Kozlowski <krzysztof.kozlowski+dt@linaro.org>,
Conor Dooley <conor+dt@kernel.org>,
Broadcom internal kernel review list
<bcm-kernel-feedback-list@broadcom.com>,
Linus Walleij <linus.walleij@linaro.org>,
Adrian Hunter <adrian.hunter@intel.com>,
Kamal Dasu <kamal.dasu@broadcom.com>,
Al Cooper <alcooperx@gmail.com>,
linux-mmc@vger.kernel.org, devicetree@vger.kernel.org,
linux-arm-kernel@lists.infradead.org,
linux-kernel@vger.kernel.org, linux-gpio@vger.kernel.org,
Jonathan Bell <jonathan@raspberrypi.com>,
Phil Elwell <phil@raspberrypi.com>
Subject: Re: [PATCH 6/6] mmc: sdhci-brcmstb: Add BCM2712 SD Express support
Date: Sat, 27 Apr 2024 13:21:06 +0200 [thread overview]
Message-ID: <ZizfokNsEExVRYaF@apocalypse> (raw)
In-Reply-To: <6042c0c7-bb8a-4898-8bed-92155b8e9c4f@broadcom.com>
On 08:55 Sun 14 Apr , Florian Fainelli wrote:
>
>
> On 4/13/2024 3:14 PM, Andrea della Porta wrote:
> > Broadcom BCM2712 SDHCI controller is SD Express capable. Add support
> > for sde capability where the implementation is based on downstream driver,
> > diverging from it in the way that init_sd_express callback is invoked:
> > in downstream the sdhci_ops structure has been augmented with a new
> > function ptr 'init_sd_express' that just propagate the call to the
> > driver specific callback so that the callstack from a structure
> > standpoint is mmc_host_ops -> sdhci_ops. The drawback here is in the
> > added level of indirection (the newly added init_sd_express is
> > redundant) and the sdhci_ops structure declaration has to be changed.
> > To overcome this the presented approach consist in patching the mmc_host_ops
> > init_sd_express callback to point directly to the custom function defined in
> > this driver (see struct brcmstb_match_priv).
> >
> > Signed-off-by: Andrea della Porta <andrea.porta@suse.com>
> > ---
> > drivers/mmc/host/Kconfig | 1 +
> > drivers/mmc/host/sdhci-brcmstb.c | 147 ++++++++++++++++++++++++++++++-
> > 2 files changed, 147 insertions(+), 1 deletion(-)
> >
> > diff --git a/drivers/mmc/host/Kconfig b/drivers/mmc/host/Kconfig
> > index aebc587f77a7..343ccac1a4e4 100644
> > --- a/drivers/mmc/host/Kconfig
> > +++ b/drivers/mmc/host/Kconfig
> > @@ -1018,6 +1018,7 @@ config MMC_SDHCI_BRCMSTB
> > depends on ARCH_BRCMSTB || BMIPS_GENERIC || COMPILE_TEST
> > depends on MMC_SDHCI_PLTFM
> > select MMC_CQHCI
> > + select OF_DYNAMIC
> > default ARCH_BRCMSTB || BMIPS_GENERIC
> > help
> > This selects support for the SDIO/SD/MMC Host Controller on
> > diff --git a/drivers/mmc/host/sdhci-brcmstb.c b/drivers/mmc/host/sdhci-brcmstb.c
> > index 907a4947abe5..56fb34a75ec2 100644
> > --- a/drivers/mmc/host/sdhci-brcmstb.c
> > +++ b/drivers/mmc/host/sdhci-brcmstb.c
> > @@ -29,6 +29,7 @@
> > #define BRCMSTB_PRIV_FLAGS_HAS_CQE BIT(0)
> > #define BRCMSTB_PRIV_FLAGS_GATE_CLOCK BIT(1)
> > +#define BRCMSTB_PRIV_FLAGS_HAS_SD_EXPRESS BIT(2)
> > #define SDHCI_ARASAN_CQE_BASE_ADDR 0x200
> > @@ -50,13 +51,19 @@ struct sdhci_brcmstb_priv {
> > unsigned int flags;
> > struct clk *base_clk;
> > u32 base_freq_hz;
> > + struct regulator *sde_1v8;
> > + struct device_node *sde_pcie;
> > + void *__iomem sde_ioaddr;
> > + void *__iomem sde_ioaddr2;
> > struct pinctrl *pinctrl;
> > struct pinctrl_state *pins_default;
> > + struct pinctrl_state *pins_sdex;
> > };
> > struct brcmstb_match_priv {
> > void (*hs400es)(struct mmc_host *mmc, struct mmc_ios *ios);
> > void (*cfginit)(struct sdhci_host *host);
> > + int (*init_sd_express)(struct mmc_host *mmc, struct mmc_ios *ios);
> > struct sdhci_ops *ops;
> > const unsigned int flags;
> > };
> > @@ -263,6 +270,105 @@ static void sdhci_brcmstb_cfginit_2712(struct sdhci_host *host)
> > }
> > }
> > +static int bcm2712_init_sd_express(struct mmc_host *mmc, struct mmc_ios *ios)
> > +{
> > + struct sdhci_host *host = mmc_priv(mmc);
> > + struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host);
> > + struct sdhci_brcmstb_priv *brcmstb_priv = sdhci_pltfm_priv(pltfm_host);
> > + struct device *dev = host->mmc->parent;
> > + u32 ctrl_val;
> > + u32 present_state;
> > + int ret;
> > +
> > + if (!brcmstb_priv->sde_ioaddr || !brcmstb_priv->sde_ioaddr2)
> > + return -EINVAL;
> > +
> > + if (!brcmstb_priv->pinctrl)
> > + return -EINVAL;
> > +
> > + /* Turn off the SD clock first */
> > + sdhci_set_clock(host, 0);
> > +
> > + /* Disable SD DAT0-3 pulls */
> > + pinctrl_select_state(brcmstb_priv->pinctrl, brcmstb_priv->pins_sdex);
> > +
> > + ctrl_val = readl(brcmstb_priv->sde_ioaddr);
> > + dev_dbg(dev, "ctrl_val 1 %08x\n", ctrl_val);
> > +
> > + /* Tri-state the SD pins */
> > + ctrl_val |= 0x1ff8;
>
> No magic values please.
Ack.
>
> > + writel(ctrl_val, brcmstb_priv->sde_ioaddr);
> > + dev_dbg(dev, "ctrl_val 1->%08x (%08x)\n", ctrl_val, readl(brcmstb_priv->sde_ioaddr));
> > + /* Let voltages settle */
> > + udelay(100);
>
> Why not usleep_range()?
No real reason. I assume only the lower boundary is critical so I can use usleep_range instead.
Will be fixed in a future patch, the SD express support will be drpped in V2 since nto strictly
necessary.
>
> > +
> > + /* Enable the PCIe sideband pins */
> > + ctrl_val &= ~0x6000;
>
> No magic values please.
>
> > + writel(ctrl_val, brcmstb_priv->sde_ioaddr);
> > + dev_dbg(dev, "ctrl_val 1->%08x (%08x)\n", ctrl_val, readl(brcmstb_priv->sde_ioaddr));
> > + /* Let voltages settle */
> > + udelay(100);
>
> Likewise.
Ditto.
>
> > +
> > + /* Turn on the 1v8 VDD2 regulator */
> > + ret = regulator_enable(brcmstb_priv->sde_1v8);
> > + if (ret)
> > + return ret;
> > +
> > + /* Wait for Tpvcrl */
> > + msleep(1);
> > +
> > + /* Sample DAT2 (CLKREQ#) - if low, card is in PCIe mode */
> > + present_state = sdhci_readl(host, SDHCI_PRESENT_STATE);
> > + present_state = (present_state & SDHCI_DATA_LVL_MASK) >> SDHCI_DATA_LVL_SHIFT;
> > + dev_dbg(dev, "state = 0x%08x\n", present_state);
> > +
> > + if (present_state & BIT(2)) {
>
> Likewise, replace with constant.
Ack.
>
> > + dev_err(dev, "DAT2 still high, abandoning SDex switch\n");
> > + return -ENODEV;
> > + }
> > +
> > + /* Turn on the LCPLL PTEST mux */
> > + ctrl_val = readl(brcmstb_priv->sde_ioaddr2 + 20); // misc5
> > + ctrl_val &= ~(0x7 << 7);
> > + ctrl_val |= 3 << 7;
> > + writel(ctrl_val, brcmstb_priv->sde_ioaddr2 + 20);
> > + dev_dbg(dev, "misc 5->%08x (%08x)\n", ctrl_val, readl(brcmstb_priv->sde_ioaddr2 + 20));
> > +
> > + /* PTEST diff driver enable */
> > + ctrl_val = readl(brcmstb_priv->sde_ioaddr2);
> > + ctrl_val |= BIT(21);
> > + writel(ctrl_val, brcmstb_priv->sde_ioaddr2);
> > +
> > + dev_dbg(dev, "misc 0->%08x (%08x)\n", ctrl_val, readl(brcmstb_priv->sde_ioaddr2));
> > +
> > + /* Wait for more than the minimum Tpvpgl time */
> > + msleep(100);
> > +
> > + if (brcmstb_priv->sde_pcie) {
> > + struct of_changeset changeset;
> > + static struct property okay_property = {
> > + .name = "status",
> > + .value = "okay",
> > + .length = 5,
> > + };
> > +
> > + /* Enable the pcie controller */
> > + of_changeset_init(&changeset);
> > + ret = of_changeset_update_property(&changeset,
> > + brcmstb_priv->sde_pcie,
> > + &okay_property);
> > + if (ret) {
> > + dev_err(dev, "%s: failed to update property - %d\n", __func__,
> > + ret);
> > + return -ENODEV;
> > + }
> > + ret = of_changeset_apply(&changeset);
> > + }
>
> Why are you doing this? Cannot the firmware enable/disable the node
> according to the boot mode or something else?
>
> This is not going to fly for upstream, sorry.
> --
> Florian
next prev parent reply other threads:[~2024-04-27 11:21 UTC|newest]
Thread overview: 39+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-04-13 22:14 [PATCH 0/6] Add support for BCM2712 SD card controller Andrea della Porta
2024-04-13 22:14 ` [PATCH 1/6] dt-bindings: pinctrl: Add support for BCM2712 pin controller Andrea della Porta
2024-04-13 23:22 ` Rob Herring
2024-04-14 6:09 ` Krzysztof Kozlowski
2024-04-14 15:45 ` Florian Fainelli
2024-04-16 12:59 ` Linus Walleij
2024-04-27 10:55 ` Andrea della Porta
2024-04-13 22:14 ` [PATCH 2/6] dt-bindings: mmc: Add support for BCM2712 SD host controller Andrea della Porta
2024-04-13 23:22 ` Rob Herring
2024-04-14 6:11 ` Krzysztof Kozlowski
2024-04-14 15:55 ` Florian Fainelli
2024-04-13 22:14 ` [PATCH 3/6] arm64: dts: broadcom: Add support for BCM2712 Andrea della Porta
2024-04-14 6:22 ` Krzysztof Kozlowski
2024-04-14 16:01 ` Florian Fainelli
2024-04-27 11:02 ` Andrea della Porta
2024-04-15 8:20 ` Stefan Wahren
2024-04-15 8:52 ` Phil Elwell
2024-04-15 9:06 ` Stefan Wahren
2024-04-15 10:43 ` Phil Elwell
2024-04-13 22:14 ` [PATCH 4/6] pinctrl: bcm: Add pinconf/pinmux controller driver " Andrea della Porta
2024-04-14 7:19 ` Christophe JAILLET
2024-04-27 11:04 ` Andrea della Porta
2024-04-14 16:00 ` Florian Fainelli
2024-04-27 11:06 ` Andrea della Porta
2024-04-16 13:07 ` Linus Walleij
2024-04-27 11:13 ` Andrea della Porta
2024-04-13 22:14 ` [PATCH 5/6] mmc: sdhci-brcmstb: Add BCM2712 support Andrea della Porta
2024-04-14 6:25 ` Krzysztof Kozlowski
2024-04-14 7:28 ` Christophe JAILLET
2024-04-14 15:53 ` Florian Fainelli
2024-04-13 22:14 ` [PATCH 6/6] mmc: sdhci-brcmstb: Add BCM2712 SD Express support Andrea della Porta
2024-04-14 7:34 ` Christophe JAILLET
2024-04-27 11:16 ` Andrea della Porta
2024-04-14 15:55 ` Florian Fainelli
2024-04-27 11:21 ` Andrea della Porta [this message]
2024-04-14 10:07 ` [PATCH 0/6] Add support for BCM2712 SD card controller Stefan Wahren
2024-05-02 9:12 ` Andrea della Porta
2024-04-14 15:54 ` Florian Fainelli
2024-04-15 18:47 ` Rob Herring
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=ZizfokNsEExVRYaF@apocalypse \
--to=andrea.porta@suse.com \
--cc=adrian.hunter@intel.com \
--cc=alcooperx@gmail.com \
--cc=bcm-kernel-feedback-list@broadcom.com \
--cc=conor+dt@kernel.org \
--cc=devicetree@vger.kernel.org \
--cc=florian.fainelli@broadcom.com \
--cc=jonathan@raspberrypi.com \
--cc=kamal.dasu@broadcom.com \
--cc=krzysztof.kozlowski+dt@linaro.org \
--cc=linus.walleij@linaro.org \
--cc=linux-arm-kernel@lists.infradead.org \
--cc=linux-gpio@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mmc@vger.kernel.org \
--cc=phil@raspberrypi.com \
--cc=robh@kernel.org \
--cc=ulf.hansson@linaro.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 read-only IMAP folder(s) and NNTP newsgroup(s).