Linux PCI subsystem development
 help / color / mirror / Atom feed
From: sashiko-bot@kernel.org
To: "Koichiro Den" <den@valinux.co.jp>
Cc: ntb@lists.linux.dev, linux-pci@vger.kernel.org
Subject: Re: [PATCH 2/3] PCI: dw-rockchip: Support fixed MSI-X table and PBA on RK3588
Date: Thu, 27 Aug 2026 18:41:07 +0000	[thread overview]
Message-ID: <20260827184108.00F6F1F000E9@smtp.kernel.org> (raw)
In-Reply-To: <20260827182012.1984960-3-den@valinux.co.jp>

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Medium] Unmapped epf_bar[bir] access in dw_pcie_ep_raise_msix_irq() causes kernel panic when msix_hw_owned is true
--

commit d51d0db24756ed54917db6dbe19e86f2adfdc77f
Author: Koichiro Den <den@valinux.co.jp>

PCI: dw-rockchip: Support fixed MSI-X table and PBA on RK3588

This commit adds support for fixed MSI-X tables and PBA on the RK3588
controller by marking BAR4 as a reserved region. It introduces the
msix_hw_owned flag to track when a hardware-owned layout is selected. The
RK3588 then overrides the MSI-X raise behavior to use the controller
doorbell when this flag is set.

> diff --git a/drivers/pci/controller/dwc/pcie-designware-ep.c b/drivers/pci/controller/dwc/pcie-designware-ep.c
> index 147b043589f02..bc2b63814dea1 100644
> --- a/drivers/pci/controller/dwc/pcie-designware-ep.c
> +++ b/drivers/pci/controller/dwc/pcie-designware-ep.c
[ ... ]
> @@ -764,6 +784,7 @@ static int dw_pcie_ep_set_msix(struct pci_epc *epc, u8 func_no, u8 vfunc_no,
>  	reg = ep_func->msix_cap + PCI_MSIX_PBA;
>  	val = layout->pba_offset | layout->pba_bar;
>  	dw_pcie_ep_writel_dbi(ep, func_no, reg, val);
> +	ep_func->msix_hw_owned = dw_pcie_ep_msix_layout_is_hw_owned(ep, layout);

[Severity: Medium]
Does introducing the msix_hw_owned state here create a potential NULL pointer
dereference in the generic dw_pcie_ep_raise_msix_irq() handler?

When msix_hw_owned is true, the MSI-X table is backed by a hardware-reserved
BAR, meaning the EPF driver does not map epf_bar[bir].

While the RK3588 controller correctly intercepts this state to avoid calling
the generic dw_pcie_ep_raise_msix_irq(), the generic function itself (which
is exported and serves as the default MSI-X handler for DesignWare) is not
updated.

If any other DW controller adds a hardware-owned layout and falls back to
this default handler, it will unconditionally dereference the NULL
ep_func->epf_bar[bir]:

drivers/pci/controller/dwc/pcie-designware-ep.c:dw_pcie_ep_raise_msix_irq() {
	...
	tbl_offset &= PCI_MSIX_TABLE_OFFSET;

	msix_tbl = ep_func->epf_bar[bir]->addr + tbl_offset;
	...
}

Should dw_pcie_ep_raise_msix_irq() be updated to check msix_hw_owned and
return an error, or otherwise avoid accessing the unmapped array?

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260827182012.1984960-1-den@valinux.co.jp?part=2

  reply	other threads:[~2026-08-27 18:41 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-08-27 18:20 [PATCH 0/3] PCI: endpoint: Support hardware-owned MSI-X table and PBA Koichiro Den
2026-08-27 18:20 ` [PATCH 1/3] " Koichiro Den
2026-08-27 18:27   ` sashiko-bot
2026-08-27 18:20 ` [PATCH 2/3] PCI: dw-rockchip: Support fixed MSI-X table and PBA on RK3588 Koichiro Den
2026-08-27 18:41   ` sashiko-bot [this message]
2026-08-27 18:20 ` [PATCH 3/3] PCI: endpoint: pci-epf-vntb: Honor MSI-X selection Koichiro Den
2026-08-27 19:04   ` sashiko-bot
2026-08-27 19:23   ` Frank Li

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=20260827184108.00F6F1F000E9@smtp.kernel.org \
    --to=sashiko-bot@kernel.org \
    --cc=den@valinux.co.jp \
    --cc=linux-pci@vger.kernel.org \
    --cc=ntb@lists.linux.dev \
    --cc=sashiko-reviews@lists.linux.dev \
    /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