Linux PCI subsystem development
 help / color / mirror / Atom feed
From: Bjorn Helgaas <helgaas@kernel.org>
To: Arnd Bergmann <arnd@kernel.org>
Cc: "Bjorn Helgaas" <bhelgaas@google.com>,
	"Hans Zhang" <18255117159@163.com>,
	"Arnd Bergmann" <arnd@arndb.de>,
	"Ilpo Järvinen" <ilpo.jarvinen@linux.intel.com>,
	"Jonathan Cameron" <Jonathan.Cameron@huawei.com>,
	"Krzysztof Wilczyński" <kwilczynski@kernel.org>,
	"Kuppuswamy Sathyanarayanan"
	<sathyanarayanan.kuppuswamy@linux.intel.com>,
	"Philipp Stanner" <phasta@kernel.org>,
	"Keith Busch" <kbusch@kernel.org>,
	"Michał Winiarski" <michal.winiarski@intel.com>,
	"Krishna Chaitanya Chundru" <krishna.chundru@oss.qualcomm.com>,
	"Jiwei Sun" <sunjw10@lenovo.com>,
	"Niklas Cassel" <cassel@kernel.org>,
	linux-pci@vger.kernel.org, linux-kernel@vger.kernel.org
Subject: Re: [PATCH] PCI: fix out-of-bounds stack access
Date: Fri, 25 Jul 2025 15:05:52 -0500	[thread overview]
Message-ID: <20250725200552.GA3111469@bhelgaas> (raw)
In-Reply-To: <20250725090633.2481650-1-arnd@kernel.org>

On Fri, Jul 25, 2025 at 11:06:28AM +0200, Arnd Bergmann wrote:
> From: Arnd Bergmann <arnd@arndb.de>
> 
> The newly added PCI_FIND_NEXT_EXT_CAP function calls a helper that
> stores a result using a 32-bit pointer, but passes a shorter local
> variable into it. gcc-15 warns about this undefined behavior:
> 
> In function 'cdns_pcie_read_cfg',
>     inlined from 'cdns_pcie_find_capability' at drivers/pci/controller/cadence/pcie-cadence.c:31:9:
> drivers/pci/controller/cadence/pcie-cadence.c:22:22: error: write of 32-bit data outside the bound of destination object, data truncated into 8-bit [-Werror=extra]
>    22 |                 *val = cdns_pcie_readb(pcie, where);
>       |                 ~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> In function 'dw_pcie_read_cfg',
>     inlined from 'dw_pcie_find_capability' at drivers/pci/controller/dwc/pcie-designware.c:234:9:
> drivers/pci/controller/dwc/pcie-designware.c:225:22: error: write of 32-bit data outside the bound of destination object, data truncated into 8-bit [-Werror=extra]
>   225 |                 *val = dw_pcie_readb_dbi(pci, where);
>       |                 ~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> 
> Change the macro to remove the invalid cast and extend the target
> variables as needed.
> 
> Fixes: 1b07388f32e1 ("PCI: Refactor extended capability search into PCI_FIND_NEXT_EXT_CAP()")
> Signed-off-by: Arnd Bergmann <arnd@arndb.de>

Squashed into that commit, thanks, Arnd!

  https://git.kernel.org/cgit/linux/kernel/git/pci/pci.git/commit/?id=2502a619108b

> ---
>  drivers/pci/pci.h | 8 ++++----
>  1 file changed, 4 insertions(+), 4 deletions(-)
> 
> diff --git a/drivers/pci/pci.h b/drivers/pci/pci.h
> index da5912c2017c..ac954584d991 100644
> --- a/drivers/pci/pci.h
> +++ b/drivers/pci/pci.h
> @@ -109,17 +109,17 @@ int pci_bus_read_config(void *priv, unsigned int devfn, int where, u32 size,
>  ({									\
>  	int __ttl = PCI_FIND_CAP_TTL;					\
>  	u8 __id, __found_pos = 0;					\
> -	u8 __pos = (start);						\
> -	u16 __ent;							\
> +	u32 __pos = (start);						\
> +	u32 __ent;							\
>  									\
> -	read_cfg(args, __pos, 1, (u32 *)&__pos);			\
> +	read_cfg(args, __pos, 1, &__pos);				\
>  									\
>  	while (__ttl--) {						\
>  		if (__pos < PCI_STD_HEADER_SIZEOF)			\
>  			break;						\
>  									\
>  		__pos = ALIGN_DOWN(__pos, 4);				\
> -		read_cfg(args, __pos, 2, (u32 *)&__ent);		\
> +		read_cfg(args, __pos, 2, &__ent);			\
>  									\
>  		__id = FIELD_GET(PCI_CAP_ID_MASK, __ent);		\
>  		if (__id == 0xff)					\
> -- 
> 2.39.5
> 

      reply	other threads:[~2025-07-25 20:05 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-07-25  9:06 [PATCH] PCI: fix out-of-bounds stack access Arnd Bergmann
2025-07-25 20:05 ` Bjorn Helgaas [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=20250725200552.GA3111469@bhelgaas \
    --to=helgaas@kernel.org \
    --cc=18255117159@163.com \
    --cc=Jonathan.Cameron@huawei.com \
    --cc=arnd@arndb.de \
    --cc=arnd@kernel.org \
    --cc=bhelgaas@google.com \
    --cc=cassel@kernel.org \
    --cc=ilpo.jarvinen@linux.intel.com \
    --cc=kbusch@kernel.org \
    --cc=krishna.chundru@oss.qualcomm.com \
    --cc=kwilczynski@kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-pci@vger.kernel.org \
    --cc=michal.winiarski@intel.com \
    --cc=phasta@kernel.org \
    --cc=sathyanarayanan.kuppuswamy@linux.intel.com \
    --cc=sunjw10@lenovo.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