linux-pci.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Bjorn Helgaas <helgaas@kernel.org>
To: Edward Chow <persmule@gmail.com>
Cc: lpieralisi@kernel.org, toke@toke.dk, kvalo@kernel.org,
	linux-pci@vger.kernel.org, robh@kernel.org,
	linux-wireless@vger.kernel.org, ath10k@lists.infradead.org,
	Edward Chow <equu@openmail.cc>
Subject: Re: [PATCH 2/3] wifi: ath9k: stop loading incompatible DT cal data
Date: Tue, 17 Jan 2023 13:46:14 -0600	[thread overview]
Message-ID: <20230117194614.GA135447@bhelgaas> (raw)
In-Reply-To: <20230117092746.1149155-1-equu@openmail.cc>

On Tue, Jan 17, 2023 at 05:27:46PM +0800, Edward Chow wrote:
> Loading calibration data from an OF device tree node not declared
> compatible with the device (e.g. a PCI device with calibration data
> from corresponding DT node gets replaced, so the newly installed
> device become incompatible with the node) or driver may lead to fatal
> result, e.g. kernel panic.

Please include a link to a bug report and include a few lines of the
oops or panic directly in the commit log so when users see this
problem, they can search for the text and possibly find this fix.

> The driver should check whether the DT node corresponding to the
> device compatible with it, and load calibration data only from
> compatible node.

If you read this commit log carefully, it doesn't actually say what
this patch *does*.  It has some background and this assertion about
what drivers *should* do, but it doesn't say what this patch does.

Suggest structure like this (flesh out with the relevant DT property
names, etc):

  For PCI ath9k devices, load calibration data only if there is a DT
  node corresponding to the device with XXX ...

More details: https://chris.beams.io/posts/git-commit/

> Signed-off-by: Edward Chow <equu@openmail.cc>
> ---
>  drivers/net/wireless/ath/ath9k/ath9k.h |  1 +
>  drivers/net/wireless/ath/ath9k/init.c  | 26 ++++++++++++++++++++++++++
>  drivers/net/wireless/ath/ath9k/pci.c   |  2 +-
>  3 files changed, 28 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/net/wireless/ath/ath9k/ath9k.h b/drivers/net/wireless/ath/ath9k/ath9k.h
> index 2cc23605c9fc..4f6f0383a5f8 100644
> --- a/drivers/net/wireless/ath/ath9k/ath9k.h
> +++ b/drivers/net/wireless/ath/ath9k/ath9k.h
> @@ -35,6 +35,7 @@ struct ath_node;
>  struct ath_vif;
>  
>  extern struct ieee80211_ops ath9k_ops;
> +extern struct pci_driver ath_pci_driver;
>  extern int ath9k_modparam_nohwcrypt;
>  extern int ath9k_led_blink;
>  extern bool is_ath9k_unloaded;
> diff --git a/drivers/net/wireless/ath/ath9k/init.c b/drivers/net/wireless/ath/ath9k/init.c
> index 4f00400c7ffb..f88a48e8456b 100644
> --- a/drivers/net/wireless/ath/ath9k/init.c
> +++ b/drivers/net/wireless/ath/ath9k/init.c
> @@ -22,6 +22,7 @@
>  #include <linux/module.h>
>  #include <linux/of.h>
>  #include <linux/of_net.h>
> +#include <linux/pci.h>
>  #include <linux/nvmem-consumer.h>
>  #include <linux/relay.h>
>  #include <linux/dmi.h>
> @@ -577,6 +578,31 @@ static int ath9k_nvmem_request_eeprom(struct ath_softc *sc)
>  	size_t len;
>  	int err;
>  
> +	/* devm_nvmem_cell_get() will get a cell first from the OF
> +	 * DT node representing the given device with nvmem-cell-name
> +	 * "calibration", and from the global lookup table as a fallback,
> +	 * and an ath9k device could be either a pci one or a platform one.
> +	 *
> +	 * If the OF DT node is not compatible with the real device, the
> +	 * calibration data got from the node should not be applied.
> +	 *
> +	 * dev_is_pci(sc->dev) && ( no OF node || caldata not from node
> +	 * || not compatible ) -> do not use caldata .
> +	 *
> +	 * !dev_is_pci(sc->dev) -> always use caldata .
> +	 */
> +	if (dev_is_pci(sc->dev) &&
> +	    (!sc->dev->of_node ||
> +	     !of_property_match_string(sc->dev->of_node,
> +				       "nvmem-cell-names",
> +				       "calibration") ||
> +	     !of_pci_node_match_driver(sc->dev->of_node,
> +				       &ath_pci_driver)))
> +		/* follow the "just return 0;" convention as
> +		 * noted below.
> +		 */
> +		return 0;
> +
>  	cell = devm_nvmem_cell_get(sc->dev, "calibration");
>  	if (IS_ERR(cell)) {
>  		err = PTR_ERR(cell);
> diff --git a/drivers/net/wireless/ath/ath9k/pci.c b/drivers/net/wireless/ath/ath9k/pci.c
> index a074e23013c5..fcb19761e60d 100644
> --- a/drivers/net/wireless/ath/ath9k/pci.c
> +++ b/drivers/net/wireless/ath/ath9k/pci.c
> @@ -1074,7 +1074,7 @@ static SIMPLE_DEV_PM_OPS(ath9k_pm_ops, ath_pci_suspend, ath_pci_resume);
>  
>  MODULE_DEVICE_TABLE(pci, ath_pci_id_table);
>  
> -static struct pci_driver ath_pci_driver = {
> +struct pci_driver ath_pci_driver = {
>  	.name       = "ath9k",
>  	.id_table   = ath_pci_id_table,
>  	.probe      = ath_pci_probe,
> -- 
> 2.39.0
> 

  reply	other threads:[~2023-01-17 21:28 UTC|newest]

Thread overview: 43+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-01-07 15:14 [PATCH] pci: Add functions to match pci dev or driver against OF DT node Mad Horse
2023-01-10 15:24 ` Bjorn Helgaas
2023-01-17  9:27 ` [PATCH 1/3] PCI: of: Match pci devices or drivers against OF DT nodes Edward Chow
2023-01-17  9:27 ` [PATCH 2/3] wifi: ath9k: stop loading incompatible DT cal data Edward Chow
2023-01-17 19:46   ` Bjorn Helgaas [this message]
2023-02-01  3:02     ` Mad Horse
2023-02-01 21:43       ` Bjorn Helgaas
2023-02-02  4:18       ` [PATCH v2 0/3] PCI: of: Load extra data only from compatible DT nodes equu
2023-02-02  4:18         ` [PATCH v2 1/3] PCI: of: Match pci devices or drivers against OF " equu
2023-02-02  4:18         ` [PATCH v2 2/3] wifi: ath9k: stop loading incompatible DT cal data equu
2023-02-02  4:18         ` [PATCH v2 3/3] wifi: ath10k: only load compatible " equu
2023-02-02  7:26           ` kernel test robot
2023-02-02  7:55       ` [PATCH v3 0/3] PCI: of: Load extra data only from compatible DT nodes equu
2023-02-02  7:55         ` [PATCH v3 1/3] PCI: of: Match pci devices or drivers against OF " equu
2023-02-03  8:23           ` kernel test robot
2023-02-02  7:55         ` [PATCH v3 2/3] wifi: ath9k: stop loading incompatible DT cal data equu
2023-02-03  9:56           ` kernel test robot
2023-02-02  7:55         ` [PATCH v3 3/3] wifi: ath10k: only load compatible " equu
2023-02-03 11:38           ` kernel test robot
2023-02-03  8:37       ` [PATCH v4 0/3] PCI: of: Load extra data only from compatible DT nodes equu
2023-02-03  8:37         ` [PATCH v4 1/3] PCI: of: Match pci devices or drivers against OF " equu
2023-02-03  8:37         ` [PATCH v4 2/3] wifi: ath9k: stop loading incompatible DT cal data equu
2023-02-03  8:37         ` [PATCH v4 3/3] wifi: ath10k: only load compatible " equu
2023-02-03 10:48       ` [PATCH v5 0/3] PCI: of: Load extra data only from compatible DT nodes equu
2023-02-03 10:48         ` [PATCH v5 1/3] PCI: of: Match pci devices or drivers against OF " equu
2023-02-03 10:48         ` [PATCH v5 2/3] wifi: ath9k: stop loading incompatible DT cal data equu
2023-02-03 10:48         ` [PATCH v5 3/3] wifi: ath10k: only load compatible " equu
2023-02-03 15:57           ` Rob Herring
2023-02-03 17:15             ` equu
2023-02-03 18:45               ` Rob Herring
2023-02-04  4:26                 ` equu
2023-02-09  4:50       ` [PATCH v6 0/3] PCI: of: Load extra data only from compatible DT nodes equu
2023-02-09  4:50       ` [PATCH v6 1/3] PCI: of: Match pci devices or drivers against OF " equu
2023-02-09  4:50       ` [PATCH v6 2/3] wifi: ath9k: stop loading incompatible DT cal data equu
2023-02-09  4:50       ` [PATCH v6 3/3] wifi: ath10k: only load compatible " equu
2023-02-09 16:09         ` Rob Herring
2023-01-17  9:28 ` [PATCH " Edward Chow
2023-01-17 10:01 ` [PATCH 1/3] PCI: of: Match pci devices or drivers against OF DT nodes Mad Horse
2023-01-17 10:02 ` [PATCH 2/3] wifi: ath9k: stop loading incompatible DT cal data Mad Horse
2023-01-17 10:02 ` [PATCH 3/3] wifi: ath10k: only load compatible " Mad Horse
2023-01-17 10:29 ` [PATCH 1/3] PCI: of: Match pci devices or drivers against OF DT nodes Mad Horse
2023-01-21 10:00 ` [PATCH 2/3] wifi: ath9k: stop loading incompatible DT cal data persmule
2023-01-21 10:06 ` [PATCH 3/3] wifi: ath10k: only load compatible " persmule

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=20230117194614.GA135447@bhelgaas \
    --to=helgaas@kernel.org \
    --cc=ath10k@lists.infradead.org \
    --cc=equu@openmail.cc \
    --cc=kvalo@kernel.org \
    --cc=linux-pci@vger.kernel.org \
    --cc=linux-wireless@vger.kernel.org \
    --cc=lpieralisi@kernel.org \
    --cc=persmule@gmail.com \
    --cc=robh@kernel.org \
    --cc=toke@toke.dk \
    /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;
as well as URLs for NNTP newsgroup(s).