Linux PCI subsystem development
 help / color / mirror / Atom feed
From: Niklas Cassel <cassel@kernel.org>
To: James Le Cuirot <chewi@gentoo.org>
Cc: Wilfred Mallawa <wilfred.mallawa@wdc.com>,
	Manivannan Sadhasivam <manivannan.sadhasivam@oss.qualcomm.com>,
	Bjorn Helgaas <bhelgaas@google.com>,
	linux-pci@vger.kernel.org, linux-rockchip@lists.infradead.org,
	Shawn Lin <shawn.lin@rock-chips.com>
Subject: Re: PCI: dw-rockchip: Unable to boot Quartz64 Model B from NVMe since reset changes
Date: Thu, 19 Nov 2026 11:55:59 +0100	[thread overview]
Message-ID: <av7Vvw5CpQo0gt1d@ryzen> (raw)
In-Reply-To: <08cc96c5171fc969b074371692e6d95d46eba60b.camel@gentoo.org>

On Wed, Nov 18, 2026 at 04:03:11PM +0000, James Le Cuirot wrote:
> Hello Wilfred,
> 
> Unfortunately, since your commit "PCI: dw-rockchip: Implement
> .reset_root_port() and use for link down" (b376b3ff9cb0527), I have been
> basically unable to boot my Quartz64 Model B from NVMe. Sometimes if I wait
> long enough, it will eventually work, albeit very slowly. If I revert this
> (and 4e9f69d08253) against 7.3.5, then it boots normally. Here is the relevant
> output.
> 
> rockchip-dw-pcie 3c0000000.pcie: host bridge /pcie@fe260000 ranges:
> rockchip-dw-pcie 3c0000000.pcie:       IO 0x00f4100000..0x00f41fffff -> 0x00f4100000
> rockchip-dw-pcie 3c0000000.pcie:      MEM 0x00f4200000..0x00f5ffffff -> 0x00f4200000
> rockchip-dw-pcie 3c0000000.pcie:      MEM 0x0300000000..0x033fffffff -> 0x0300000000
> rockchip-dw-pcie 3c0000000.pcie: iATU: unroll T, 8 ob, 8 ib, align 64K, limit 8G
> rockchip-dw-pcie 3c0000000.pcie: PCIe Gen.2 x1 link up

Here the link is reported as being up, at Gen2 speed.


> rockchip-dw-pcie 3c0000000.pcie: PCI host bridge to bus 0000:00
> pci_bus 0000:00: root bus resource [bus 00-0f]
> pci_bus 0000:00: root bus resource [io  0x0000-0xfffff] (bus address [0xf4100000-0xf41fffff])
> pci_bus 0000:00: root bus resource [mem 0xf4200000-0xf5ffffff]
> pci_bus 0000:00: root bus resource [mem 0x300000000-0x33fffffff]
> pci 0000:00:00.0: [1d87:3566] type 01 class 0x060400 PCIe Root Port
> pci 0000:00:00.0: ROM [mem 0x00000000-0x0000ffff pref]
> pci 0000:00:00.0: PCI bridge to [bus 01-ff]
> pci 0000:00:00.0:   bridge window [io  0x0000-0x0fff]
> pci 0000:00:00.0:   bridge window [mem 0x00000000-0x000fffff]
> pci 0000:00:00.0:   bridge window [mem 0x00000000-0x000fffff 64bit pref]
> pci 0000:00:00.0: supports D1 D2
> pci 0000:00:00.0: PME# supported from D0 D1 D3hot
> pci_bus 0000:01: busn_res: can not insert [bus 01-ff] under [bus 00-0f] (conflicts with (null) [bus 00-0f])
> pci 0000:01:00.0: [1e0f:0009] type 00 class 0x010802 PCIe Endpoint
> pci 0000:01:00.0: BAR 0 [mem 0x00000000-0x00003fff 64bit]
> pci 0000:01:00.0: 4.000 Gb/s available PCIe bandwidth, limited by 5.0 GT/s PCIe x1 link at 0000:00:00.0 (capable of 31.504 Gb/s with 8.0 GT/s PCIe x4 link)
> pci 0000:01:00.0: ASPM: default states L1
> pci 0000:00:00.0: bridge window [mem 0xf4200000-0xf42fffff]: assigned
> pci 0000:00:00.0: ROM [mem 0xf4300000-0xf430ffff pref]: assigned
> pci 0000:01:00.0: BAR 0 [mem 0xf4200000-0xf4203fff 64bit]: assigned
> pci 0000:00:00.0: PCI bridge to [bus 01-ff]
> pci 0000:00:00.0:   bridge window [mem 0xf4200000-0xf42fffff]
> pci_bus 0000:00: resource 4 [io  0x0000-0xfffff]
> pci_bus 0000:00: resource 5 [mem 0xf4200000-0xf5ffffff]
> pci_bus 0000:00: resource 6 [mem 0x300000000-0x33fffffff]
> pci_bus 0000:01: resource 1 [mem 0xf4200000-0xf42fffff]
> pcieport 0000:00:00.0: PME: Signaling with IRQ 86
> pcieport 0000:00:00.0: AER: enabled with IRQ 95
> pcieport 0000:00:00.0: Recovering Root Port due to Link Down

Here you get a link down IRQ.

Why is the link going down?

You seem to have already trained in Gen2, and your board seems to only
support Gen2 (so even if it was related to speed change - normally a
speed should not trigger a link down - a speed change does not seem to
happen here).


Perhaps you can enable CONFIG_PCIE_DW_DEBUGFS and then look at
/sys/kernel/debug/dwc_pcie_<dev>/
to try to see if that gives any more information about why the link goes down.

Like Wilfred said, you could also try to disable ASPM.

Also make sure that your device is attached to the slot properly (and
hopefully you are not using any extender cables or similar).


I guess we could also try to clear the interrupt just before unmasking it,
to see if this is a spurious interrupt, something like:

--- a/drivers/pci/controller/dwc/pcie-dw-rockchip.c
+++ b/drivers/pci/controller/dwc/pcie-dw-rockchip.c
@@ -692,7 +692,7 @@ static int rockchip_pcie_configure_rc(struct platform_device *pdev,
        struct device *dev = &pdev->dev;
        struct dw_pcie_rp *pp;
        int irq, ret;
-       u32 val;
+       u32 reg, val;

        if (!IS_ENABLED(CONFIG_PCIE_ROCKCHIP_DW_HOST))
                return -ENODEV;
@@ -739,6 +739,9 @@ static int rockchip_pcie_configure_rc(struct platform_device *pdev,
                return ret;
        }

+       reg = rockchip_pcie_readl_apb(rockchip, PCIE_CLIENT_INTR_STATUS_MISC);
+       rockchip_pcie_writel_apb(rockchip, reg, PCIE_CLIENT_INTR_STATUS_MISC);
+
        /* unmask hot reset/link-down reset */
        val = FIELD_PREP_WM16(PCIE_LINK_REQ_RST_NOT_INT, 0);
        rockchip_pcie_writel_apb(rockchip, val, PCIE_CLIENT_INTR_MASK_MISC);



Kind regards,
Niklas

      parent reply	other threads:[~2026-11-19 10:56 UTC|newest]

Thread overview: 5+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-11-18 16:03 PCI: dw-rockchip: Unable to boot Quartz64 Model B from NVMe since reset changes James Le Cuirot
2026-11-18 16:46 ` Niklas Cassel
2026-11-18 22:30   ` James Le Cuirot
2026-11-18 22:47     ` Wilfred Mallawa
2026-11-19 10:55 ` Niklas Cassel [this message]

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=av7Vvw5CpQo0gt1d@ryzen \
    --to=cassel@kernel.org \
    --cc=bhelgaas@google.com \
    --cc=chewi@gentoo.org \
    --cc=linux-pci@vger.kernel.org \
    --cc=linux-rockchip@lists.infradead.org \
    --cc=manivannan.sadhasivam@oss.qualcomm.com \
    --cc=shawn.lin@rock-chips.com \
    --cc=wilfred.mallawa@wdc.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