From: Andy Shevchenko <andriy.shevchenko@linux•intel.com>
To: Shawn Lin <shawn.lin@rock-chips•com>
Cc: Andy Shevchenko <andy.shevchenko@gmail•com>,
Bjorn Helgaas <bhelgaas@google•com>,
"Vaibhaav Ram T . L" <vaibhaavram.tl@microchip•com>,
Kumaravel Thiagarajan <kumaravel.thiagarajan@microchip•com>,
Even Xu <even.xu@intel•com>, Xinpeng Sun <xinpeng.sun@intel•com>,
Srinivas Pandruvada <srinivas.pandruvada@linux•intel.com>,
Jiri Kosina <jikos@kernel•org>,
Alexandre Belloni <alexandre.belloni@bootlin•com>,
Zhou Wang <wangzhou1@hisilicon•com>,
Longfang Liu <liulongfang@huawei•com>,
Vinod Koul <vkoul@kernel•org>, Lee Jones <lee@kernel•org>,
Jijie Shao <shaojijie@huawei•com>,
Jian Shen <shenjian15@huawei•com>,
Sunil Goutham <sgoutham@marvell•com>,
Andrew Lunn <andrew+netdev@lunn•ch>,
Heiner Kallweit <hkallweit1@gmail•com>,
"David S . Miller" <davem@davemloft•net>,
Jeff Hugo <jeff.hugo@oss•qualcomm.com>,
Oded Gabbay <ogabbay@kernel•org>,
Maciej Falkowski <maciej.falkowski@linux•intel.com>,
Karol Wachowski <karol.wachowski@linux•intel.com>,
Min Ma <mamin506@gmail•com>, Lizhi Hou <lizhi.hou@amd•com>,
Andreas Noever <andreas.noever@gmail•com>,
Mika Westerberg <westeri@kernel•org>,
Tomasz Jeznach <tjeznach@rivosinc•com>,
Will Deacon <will@kernel•org>,
Xinliang Liu <xinliang.liu@linaro•org>,
Tian Tao <tiantao6@hisilicon•com>,
Davidlohr Bueso <dave@stgolabs•net>,
Jonathan Cameron <jonathan.cameron@huawei•com>,
Srujana Challa <schalla@marvell•com>,
Bharat Bhushan <bbhushan2@marvell•com>,
Antoine Tenart <atenart@kernel•org>,
Herbert Xu <herbert@gondor•apana.org.au>,
Raag Jadav <raag.jadav@intel•com>,
Hans de Goede <hansg@kernel•org>,
Greg Kroah-Hartman <gregkh@linuxfoundation•org>,
Jiri Slaby <jirislaby@kernel•org>,
Andy Shevchenko <andy@kernel•org>,
Manivannan Sadhasivam <mani@kernel•org>,
Mika Westerberg <mika.westerberg@linux•intel.com>,
Andi Shyti <andi.shyti@kernel•org>,
Robert Richter <rric@kernel•org>, Mark Brown <broonie@kernel•org>,
Nirmal Patel <nirmal.patel@linux•intel.com>,
Kurt Schwemmer <kurt.schwemmer@microsemi•com>,
Logan Gunthorpe <logang@deltatee•com>,
Linus Walleij <linusw@kernel•org>,
Bartosz Golaszewski <brgl@kernel•org>,
Sakari Ailus <sakari.ailus@linux•intel.com>,
Bingbu Cao <bingbu.cao@intel•com>,
Ulf Hansson <ulf.hansson@linaro•org>,
Arnd Bergmann <arnd@arndb•de>,
Benjamin Tissoires <bentiss@kernel•org>,
linux-input@vger•kernel.org, linux-i3c@lists•infradead.org,
dmaengine@vger•kernel.org, Philipp Stanner <phasta@kernel•org>,
netdev@vger•kernel.org, nic_swsd@realtek•com,
linux-arm-msm@vger•kernel.org, dri-devel@lists•freedesktop.org,
linux-usb@vger•kernel.org, iommu@lists•linux.dev,
linux-riscv@lists•infradead.org, David Airlie <airlied@gmail•com>,
Simona Vetter <simona@ffwll•ch>,
linux-cxl@vger•kernel.org, linux-crypto@vger•kernel.org,
platform-driver-x86@vger•kernel.org,
linux-serial@vger•kernel.org, mhi@lists•linux.dev,
Jan Dabros <jsd@semihalf•com>,
linux-i2c@vger•kernel.org, Daniel Mack <daniel@zonque•org>,
Haojian Zhuang <haojian.zhuang@gmail•com>,
linux-spi@vger•kernel.org,
Jonathan Derrick <jonathan.derrick@linux•dev>,
linux-pci@vger•kernel.org, linux-gpio@vger•kernel.org,
Mauro Carvalho Chehab <mchehab@kernel•org>,
linux-media@vger•kernel.org, linux-mmc@vger•kernel.org
Subject: Re: [PATCH 0/37] PCI/MSI: Enforce explicit IRQ vector management by removing devres auto-free
Date: Mon, 23 Feb 2026 19:38:33 +0200 [thread overview]
Message-ID: <aZyQmc7nOt87jitX@smile.fi.intel.com> (raw)
In-Reply-To: <cb878741-7b61-b72c-5a72-6ed6d5091b1f@rock-chips.com>
On Tue, Feb 24, 2026 at 12:09:37AM +0800, Shawn Lin wrote:
> 在 2026/02/23 星期一 23:50, Andy Shevchenko 写道:
> > On Mon, Feb 23, 2026 at 5:32 PM Shawn Lin <shawn.lin@rock-chips•com> wrote:
> > >
> > > This patch series addresses a long-standing design issue in the PCI/MSI
> > > subsystem where the implicit, automatic management of IRQ vectors by
> > > the devres framework conflicts with explicit driver cleanup, creating
> > > ambiguity and potential resource management bugs.
> > >
> > > ==== The Problem: Implicit vs. Explicit Management ====
> > > Historically, `pcim_enable_device()` not only manages standard PCI resources
> > > (BARs) via devres but also implicitly triggers automatic IRQ vector management
> > > by setting a flag that registers `pcim_msi_release()` as a cleanup action.
> > >
> > > This creates an ambiguous ownership model. Many drivers follow a pattern of:
> > > 1. Calling `pci_alloc_irq_vectors()` to allocate interrupts.
> > > 2. Also calling `pci_free_irq_vectors()` in their error paths or remove routines.
> > >
> > > When such a driver also uses `pcim_enable_device()`, the devres framework may
> > > attempt to free the IRQ vectors a second time upon device release, leading to
> > > a double-free. Analysis of the tree shows this hazardous pattern exists widely,
> > > while 35 other drivers correctly rely solely on the implicit cleanup.
> >
> > Is this confirmed? What I read from the cover letter, this series was
> > only compile-tested, so how can you prove the problem exists in the
> > first place?
>
> Yes, it's confirmed. My debug of a double free issue of a out-of-tree
> PCIe wifi driver which uses
> pcim_enable_device + pci_alloc_irq_vectors + pci_free_irq_vectors expose
> it. And we did have a TODO to cleanup this hybrid usage, targeted in
> this cycle[1] suggested by Philipp:
Okay, fair enough. I think this bit was missing in the cover letter.
> [1] https://git.kernel.org/pub/scm/linux/kernel/git/pci/pci.git/log/?h=msi
> > > ==== The Solution: Making Management Explicit ====
> > > This series enforces a clear, predictable model:
> > > 1. New Managed API (Patch 1/37): Introduces pcim_alloc_irq_vectors() and
> > > pcim_alloc_irq_vectors_affinity(). Drivers that desire devres-managed IRQ
> > > vectors should use these functions, which set the is_msi_managed flag and
> > > ensure automatic cleanup.
> > > 2. Patches 2 through 36 convert each driver that uses pcim_enable_device() alongside
> > > pci_alloc_irq_vectors() and relies on devres for IRQ vector cleanup to instead
> > > make an explicit call to pcim_alloc_irq_vectors().
> > > 3. Core Change (Patch 37/37): With the former cleanup, now modifies pcim_setup_msi_release()
> > > to check only the is_msi_managed flag. This decouples automatic IRQ cleanup from
> > > pcim_enable_device(). IRQ vectors allocated via pci_alloc_irq_vectors*()
> > > are now solely the driver's responsibility to free with pci_free_irq_vectors().
> > >
> > > With these changes, we clear ownership model: Explicit resource management eliminates
> > > ambiguity and follows the "principle of least surprise." New drivers choose one model and
> > > be consistent.
> > > - Use `pci_alloc_irq_vectors()` + `pci_free_irq_vectors()` for explicit control.
> > > - Use `pcim_alloc_irq_vectors()` for devres-managed, automatic cleanup.
> >
> > Have you checked previous attempts? Why is your series better than those?
>
> There seems not previous attempts.
Maybe we are looking to the different projects...
https://lore.kernel.org/all/?q=pcim_alloc_irq_vectors
> > > ==== Testing And Review ====
> > > 1. This series is only compiled test with allmodconfig.
> > > 2. Given the substantial size of this patch series, I have structured the mailing
> > > to facilitate efficient review. The cover letter, the first patch and the last one will be sent
> > > to all relevant mailing lists and key maintainers to ensure broad visibility and
> > > initial feedback on the overall approach. The remaining subsystem-specific patches
> > > will be sent only to the respective subsystem maintainers and their associated
> > > mailing lists, reducing noise.
--
With Best Regards,
Andy Shevchenko
next prev parent reply other threads:[~2026-02-23 17:38 UTC|newest]
Thread overview: 22+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-02-23 15:29 [PATCH 0/37] PCI/MSI: Enforce explicit IRQ vector management by removing devres auto-free Shawn Lin
2026-02-23 15:29 ` [PATCH 01/37] PCI/MSI: Add Devres managed IRQ vectors allocation Shawn Lin
2026-02-24 0:04 ` Jakub Kicinski
2026-02-24 2:08 ` Shawn Lin
2026-02-24 7:47 ` Philipp Stanner
2026-02-24 8:21 ` Shawn Lin
2026-02-24 8:32 ` Philipp Stanner
2026-02-24 16:20 ` Jonathan Cameron
2026-02-23 15:29 ` [PATCH 37/37] PCI/MSI: Only check is_msi_managed in pcim_setup_msi_release() Shawn Lin
2026-02-23 15:50 ` [PATCH 0/37] PCI/MSI: Enforce explicit IRQ vector management by removing devres auto-free Andy Shevchenko
2026-02-23 16:09 ` Shawn Lin
2026-02-23 17:38 ` Andy Shevchenko [this message]
2026-02-24 2:29 ` Shawn Lin
2026-02-23 15:56 ` [PATCH 26/37] net: stmmac: Replace pci_alloc_irq_vectors() with pcim_alloc_irq_vectors() Shawn Lin
2026-02-23 15:56 ` [PATCH 27/37] r8169: " Shawn Lin
2026-02-23 15:57 ` [PATCH 28/37] net: thunder_bgx: " Shawn Lin
2026-02-23 15:57 ` [PATCH 29/37] net: hibmcge: " Shawn Lin
2026-02-24 4:14 ` [PATCH 0/37] PCI/MSI: Enforce explicit IRQ vector management by removing devres auto-free Simon Richter
2026-02-24 7:39 ` Philipp Stanner
2026-02-24 9:12 ` Andy Shevchenko
2026-02-24 10:30 ` Philipp Stanner
2026-02-24 10:39 ` Andy Shevchenko
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=aZyQmc7nOt87jitX@smile.fi.intel.com \
--to=andriy.shevchenko@linux$(echo .)intel.com \
--cc=airlied@gmail$(echo .)com \
--cc=alexandre.belloni@bootlin$(echo .)com \
--cc=andi.shyti@kernel$(echo .)org \
--cc=andreas.noever@gmail$(echo .)com \
--cc=andrew+netdev@lunn$(echo .)ch \
--cc=andy.shevchenko@gmail$(echo .)com \
--cc=andy@kernel$(echo .)org \
--cc=arnd@arndb$(echo .)de \
--cc=atenart@kernel$(echo .)org \
--cc=bbhushan2@marvell$(echo .)com \
--cc=bentiss@kernel$(echo .)org \
--cc=bhelgaas@google$(echo .)com \
--cc=bingbu.cao@intel$(echo .)com \
--cc=brgl@kernel$(echo .)org \
--cc=broonie@kernel$(echo .)org \
--cc=daniel@zonque$(echo .)org \
--cc=dave@stgolabs$(echo .)net \
--cc=davem@davemloft$(echo .)net \
--cc=dmaengine@vger$(echo .)kernel.org \
--cc=dri-devel@lists$(echo .)freedesktop.org \
--cc=even.xu@intel$(echo .)com \
--cc=gregkh@linuxfoundation$(echo .)org \
--cc=hansg@kernel$(echo .)org \
--cc=haojian.zhuang@gmail$(echo .)com \
--cc=herbert@gondor$(echo .)apana.org.au \
--cc=hkallweit1@gmail$(echo .)com \
--cc=iommu@lists$(echo .)linux.dev \
--cc=jeff.hugo@oss$(echo .)qualcomm.com \
--cc=jikos@kernel$(echo .)org \
--cc=jirislaby@kernel$(echo .)org \
--cc=jonathan.cameron@huawei$(echo .)com \
--cc=jonathan.derrick@linux$(echo .)dev \
--cc=jsd@semihalf$(echo .)com \
--cc=karol.wachowski@linux$(echo .)intel.com \
--cc=kumaravel.thiagarajan@microchip$(echo .)com \
--cc=kurt.schwemmer@microsemi$(echo .)com \
--cc=lee@kernel$(echo .)org \
--cc=linusw@kernel$(echo .)org \
--cc=linux-arm-msm@vger$(echo .)kernel.org \
--cc=linux-crypto@vger$(echo .)kernel.org \
--cc=linux-cxl@vger$(echo .)kernel.org \
--cc=linux-gpio@vger$(echo .)kernel.org \
--cc=linux-i2c@vger$(echo .)kernel.org \
--cc=linux-i3c@lists$(echo .)infradead.org \
--cc=linux-input@vger$(echo .)kernel.org \
--cc=linux-media@vger$(echo .)kernel.org \
--cc=linux-mmc@vger$(echo .)kernel.org \
--cc=linux-pci@vger$(echo .)kernel.org \
--cc=linux-riscv@lists$(echo .)infradead.org \
--cc=linux-serial@vger$(echo .)kernel.org \
--cc=linux-spi@vger$(echo .)kernel.org \
--cc=linux-usb@vger$(echo .)kernel.org \
--cc=liulongfang@huawei$(echo .)com \
--cc=lizhi.hou@amd$(echo .)com \
--cc=logang@deltatee$(echo .)com \
--cc=maciej.falkowski@linux$(echo .)intel.com \
--cc=mamin506@gmail$(echo .)com \
--cc=mani@kernel$(echo .)org \
--cc=mchehab@kernel$(echo .)org \
--cc=mhi@lists$(echo .)linux.dev \
--cc=mika.westerberg@linux$(echo .)intel.com \
--cc=netdev@vger$(echo .)kernel.org \
--cc=nic_swsd@realtek$(echo .)com \
--cc=nirmal.patel@linux$(echo .)intel.com \
--cc=ogabbay@kernel$(echo .)org \
--cc=phasta@kernel$(echo .)org \
--cc=platform-driver-x86@vger$(echo .)kernel.org \
--cc=raag.jadav@intel$(echo .)com \
--cc=rric@kernel$(echo .)org \
--cc=sakari.ailus@linux$(echo .)intel.com \
--cc=schalla@marvell$(echo .)com \
--cc=sgoutham@marvell$(echo .)com \
--cc=shaojijie@huawei$(echo .)com \
--cc=shawn.lin@rock-chips$(echo .)com \
--cc=shenjian15@huawei$(echo .)com \
--cc=simona@ffwll$(echo .)ch \
--cc=srinivas.pandruvada@linux$(echo .)intel.com \
--cc=tiantao6@hisilicon$(echo .)com \
--cc=tjeznach@rivosinc$(echo .)com \
--cc=ulf.hansson@linaro$(echo .)org \
--cc=vaibhaavram.tl@microchip$(echo .)com \
--cc=vkoul@kernel$(echo .)org \
--cc=wangzhou1@hisilicon$(echo .)com \
--cc=westeri@kernel$(echo .)org \
--cc=will@kernel$(echo .)org \
--cc=xinliang.liu@linaro$(echo .)org \
--cc=xinpeng.sun@intel$(echo .)com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
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