From: Mikko Perttunen <cyndis@kapsi.fi>
To: Manikanta Maddireddy <mmaddireddy@nvidia.com>,
thierry.reding@gmail.com, jonathanh@nvidia.com,
robh+dt@kernel.org, frowand.list@gmail.com, bhelgaas@google.com,
rjw@rjwysocki.net, tglx@linutronix.de
Cc: vidyas@nvidia.com, kthota@nvidia.com,
linux-tegra@vger.kernel.org, devicetree@vger.kernel.org,
linux-pci@vger.kernel.org, linux-pm@vger.kernel.org
Subject: Re: [PATCH V2 8/9] PCI: tegra: Broadcast PME_turn_Off message before link goes to L2
Date: Wed, 29 Nov 2017 14:18:54 +0200 [thread overview]
Message-ID: <dd3cb6a5-76ec-98c2-4132-54678c2d74ce@kapsi.fi> (raw)
In-Reply-To: <1511638333-22951-9-git-send-email-mmaddireddy@nvidia.com>
On 25.11.2017 21:32, Manikanta Maddireddy wrote:
> Per PCIe r3.0, sec 5.3.3.2.1, PCIe root port shoould broadcast PME_turn_Off
> message before PCIe link goes to L2. PME_turn_Off broadcast mechanism is
> implemented in AFI module. Each Tegra PCIe root port has its own
> PME_turn_Off and PME_TO_Ack bitmap in AFI_PME register, program this
> register to broadcast PME_turn_Off message.
>
> Signed-off-by: Manikanta Maddireddy <mmaddireddy@nvidia.com>
> ---
> V2:
> * no change in this patch
>
> drivers/pci/host/pci-tegra.c | 76 ++++++++++++++++++++++++++++++++++++++++++++
> 1 file changed, 76 insertions(+)
>
> diff --git a/drivers/pci/host/pci-tegra.c b/drivers/pci/host/pci-tegra.c
> index bbc2807bcd4a..b380958a3deb 100644
> --- a/drivers/pci/host/pci-tegra.c
> +++ b/drivers/pci/host/pci-tegra.c
> @@ -155,6 +155,8 @@
> #define AFI_INTR_EN_FPCI_TIMEOUT (1 << 7)
> #define AFI_INTR_EN_PRSNT_SENSE (1 << 8)
>
> +#define AFI_PCIE_PME 0xf0
> +
> #define AFI_PCIE_CONFIG 0x0f8
> #define AFI_PCIE_CONFIG_PCIE_DISABLE(x) (1 << ((x) + 1))
> #define AFI_PCIE_CONFIG_PCIE_DISABLE_ALL 0xe
> @@ -315,6 +317,7 @@
> #define PADS_REFCLK_CFG_DRVI_SHIFT 12 /* 15:12 */
>
> #define LINK_RETRAIN_TIMEOUT 100000
> +#define PME_ACK_TIMEOUT 10000
>
> struct tegra_msi {
> struct msi_controller chip;
> @@ -1503,6 +1506,76 @@ static int tegra_pcie_put_resources(struct tegra_pcie *pcie)
> return 0;
> }
>
> +static inline u32 get_pme_turnoff_bitmap(struct tegra_pcie_port *port)
> +{
> + struct device *dev = port->pcie->dev;
> + struct device_node *np = dev->of_node;
> + int ret = 0;
> +
> + switch (port->index) {
> + case 0:
> + ret = 0;
> + case 1:
> + ret = 8;
> + case 2:
> + if (of_device_is_compatible(np, "nvidia,tegra30-pcie"))
> + ret = 16;
> + else
> + ret = 12;
> + }
> + return ret;
> +}
> +
> +static inline u32 get_pme_ack_bitmap(struct tegra_pcie_port *port)
> +{
> + struct device *dev = port->pcie->dev;
> + struct device_node *np = dev->of_node;
> + int ret = 0;
> +
> + switch (port->index) {
> + case 0:
> + ret = 5;
> + case 1:
> + ret = 10;
> + case 2:
> + if (of_device_is_compatible(np, "nvidia,tegra30-pcie"))
> + ret = 18;
> + else
> + ret = 14;
> + }
> + return ret;
> +}
From what I can tell, the port 2 bit is 12/14 on everything after
Tegra30 as well, so I don't think this actually works?
I think simpler would be to add a SoC data fields 'u8
pme_turnoff_bit[3]' and 'u8 pme_ack_bit[3]' and then set that
to '.pme_turnoff_bit = { 0, 8, 16 }' and so on.
> +
> +static void tegra_pcie_pme_turnoff(struct tegra_pcie_port *port)
> +{
> + struct tegra_pcie *pcie = port->pcie;
> + ktime_t deadline;
> + unsigned int data;
> +
> + data = afi_readl(pcie, AFI_PCIE_PME);
> + data |= (0x1 << get_pme_turnoff_bitmap(port));
> + afi_writel(pcie, data, AFI_PCIE_PME);
> +
> + deadline = ktime_add_us(ktime_get(), PME_ACK_TIMEOUT);
> + do {
> + data = afi_readl(pcie, AFI_PCIE_PME);
> + data &= (0x1 << get_pme_ack_bitmap(port));
> + udelay(1);
> + if (ktime_after(ktime_get(), deadline))
> + break;
> + } while (!data);
Since this is a normal MMIO read, we could replace the whole loop with a
call to readl_poll_timeout (or readl_relaxed_poll_timeout, or if we
really must delay and not sleep, readl_poll_timeout_atomic etc.) from
iopoll.h
int err;
u32 val;
err = readl_poll_timeout(pcie->afi + AFI_PCIE_PME, val,
val & (0x1 << pcie->soc->pme_ack_bit[port]),
1, PME_ACK_TIMEOUT);
if (err)
...
> +
> + if (data)
> + dev_err(pcie->dev, "PME Ack is not receieved on port: %d\n",
> + port->index);
> +
Typo here, s/receieved/received/
Cheers,
Mikko
> + usleep_range(10000, 11000);
> +
> + data = afi_readl(pcie, AFI_PCIE_PME);
> + data &= ~(0x1 << get_pme_turnoff_bitmap(port));
> + afi_writel(pcie, data, AFI_PCIE_PME);
> +}
> +
> static int tegra_msi_alloc(struct tegra_msi *chip)
> {
> int msi;
> @@ -2828,6 +2901,7 @@ static int tegra_pcie_remove(struct platform_device *pdev)
> {
> struct tegra_pcie *pcie = platform_get_drvdata(pdev);
> struct pci_host_bridge *host = pci_host_bridge_from_priv(pcie);
> + struct tegra_pcie_port *port, *tmp;
>
> if (IS_ENABLED(CONFIG_DEBUG_FS))
> tegra_pcie_debugfs_exit(pcie);
> @@ -2835,6 +2909,8 @@ static int tegra_pcie_remove(struct platform_device *pdev)
> pci_remove_root_bus(host->bus);
> if (IS_ENABLED(CONFIG_PCI_MSI))
> tegra_pcie_disable_msi(pcie);
> + list_for_each_entry_safe(port, tmp, &pcie->ports, list)
> + tegra_pcie_pme_turnoff(port);
> tegra_pcie_disable_ports(pcie);
> tegra_pcie_free_resources(pcie);
> tegra_pcie_disable_controller(pcie);
>
next prev parent reply other threads:[~2017-11-29 12:18 UTC|newest]
Thread overview: 24+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-11-25 19:32 [PATCH V2 0/9] Add loadable kernel module and power management support Manikanta Maddireddy
2017-11-25 19:32 ` [PATCH V2 3/9] ARM: tegra: Export tegra_cpuidle_pcie_irqs_in_use() Manikanta Maddireddy
[not found] ` <1511638333-22951-1-git-send-email-mmaddireddy-DDmLM1+adcrQT0dZR+AlfA@public.gmane.org>
2017-11-25 19:32 ` [PATCH V2 1/9] genirq: Export irq_set_msi_desc() Manikanta Maddireddy
2017-11-25 19:32 ` [PATCH V2 2/9] of: Export of_pci_range_to_resource() Manikanta Maddireddy
[not found] ` <1511638333-22951-3-git-send-email-mmaddireddy-DDmLM1+adcrQT0dZR+AlfA@public.gmane.org>
2017-11-26 22:31 ` Rob Herring
2017-11-25 19:32 ` [PATCH V2 4/9] PCI: Export pci_find_host_bridge() Manikanta Maddireddy
2017-11-29 17:35 ` Christoph Hellwig
2017-11-25 19:32 ` [PATCH V2 5/9] PCI: Export pci_flags Manikanta Maddireddy
2017-11-29 17:01 ` Bjorn Helgaas
[not found] ` <20171129170133.GC6469-1RhO1Y9PlrlHTL0Zs8A6p5iNqAH0jzoTYJqu5kTmcBRl57MIdRCFDg@public.gmane.org>
2017-11-30 10:24 ` Lorenzo Pieralisi
2017-11-30 18:42 ` Bjorn Helgaas
2017-11-30 19:38 ` Manikanta Maddireddy
2017-11-25 19:32 ` [PATCH V2 6/9] PCI: tegra: free resources on probe failure Manikanta Maddireddy
2017-11-29 11:59 ` Mikko Perttunen
2017-11-29 17:02 ` Bjorn Helgaas
2017-11-25 19:32 ` [PATCH V2 7/9] PCI: tegra: Add loadable kernel module support Manikanta Maddireddy
[not found] ` <1511638333-22951-8-git-send-email-mmaddireddy-DDmLM1+adcrQT0dZR+AlfA@public.gmane.org>
2017-11-29 12:01 ` Mikko Perttunen
2017-11-30 18:39 ` Manikanta Maddireddy
2017-11-25 19:32 ` [PATCH V2 8/9] PCI: tegra: Broadcast PME_turn_Off message before link goes to L2 Manikanta Maddireddy
2017-11-29 12:18 ` Mikko Perttunen [this message]
2017-12-01 8:51 ` Mikko Perttunen
[not found] ` <1511638333-22951-9-git-send-email-mmaddireddy-DDmLM1+adcrQT0dZR+AlfA@public.gmane.org>
2017-11-29 16:51 ` Bjorn Helgaas
2017-11-30 18:43 ` Manikanta Maddireddy
2017-11-25 19:32 ` [PATCH V2 9/9] PCI: tegra: Add power management support Manikanta Maddireddy
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=dd3cb6a5-76ec-98c2-4132-54678c2d74ce@kapsi.fi \
--to=cyndis@kapsi.fi \
--cc=bhelgaas@google.com \
--cc=devicetree@vger.kernel.org \
--cc=frowand.list@gmail.com \
--cc=jonathanh@nvidia.com \
--cc=kthota@nvidia.com \
--cc=linux-pci@vger.kernel.org \
--cc=linux-pm@vger.kernel.org \
--cc=linux-tegra@vger.kernel.org \
--cc=mmaddireddy@nvidia.com \
--cc=rjw@rjwysocki.net \
--cc=robh+dt@kernel.org \
--cc=tglx@linutronix.de \
--cc=thierry.reding@gmail.com \
--cc=vidyas@nvidia.com \
/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