linux-kernel - Re: [PATCH] memory: omap-gpmc: fix multi-device handling on the same CS

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <2a2b2412-752a-1c64-41b7-eb256ae55f42@linaro.org>
Date:   Wed, 11 Jan 2023 15:57:50 +0100
From:   Krzysztof Kozlowski <krzysztof.kozlowski@...aro.org>
To:     INAGAKI Hiroshi <musashino.open@...il.com>, rogerq@...nel.org,
        tony@...mide.com
Cc:     linux-omap@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] memory: omap-gpmc: fix multi-device handling on the same
 CS

On 11/01/2023 15:13, INAGAKI Hiroshi wrote:
> This patch fixes the handling of multiple devices on the same CS by

Do not use "This commit/patch".
https://elixir.bootlin.com/linux/v5.17.1/source/Documentation/process/submitting-patches.rst#L95

> replacing CS name to "name" of node instead of "full_name".
> 
> In c2ade654dbf7d02f09ad491f5621fc321d4af96b
> ("memory: omap-gpmc: Use of_node_name_eq for node name comparisons"),

Use syntax: commit short SHA (".....") as pointed by checkpatch.

> the name for setting to CS was replaced but it doesn't fit for the
> comparison by of_node_name_eq.
> In of_node_name_eq, the length for strncmp will be obtained from the
> node that trying to register and it doesn't contain the length of
> "@<cs>,<offset>". 

Skip explanation what is inside of_node_name_eq() but focus on what the
driver is doing.

> But the base name for comparison passed from
> registered CS name contains the prefix,

What is "the prefix"?

>  then, that two lengths won't
> match and false will be returned, and registration on the same CS
> will be failed.

Unfortunately, based on this, I don't get what is compare with what. I
bet the issue is simple, but based on the description it does not look
like that.

> 
> example (Century Systems MA-E350/N, AM3352):
> 
> - Device Tree
> 
>   /* memory mapped gpio controllers on GPMC */
>   gpio@2,2 {
>       reg = <2 0x2 0x1>; /* CS2, offset 0x2, IO size 0x1 */
>       ...
>   };
> 
>   gpio@2,10 {
>       reg = <2 0x10 0x1>; /* CS2, offset 0x10, IO size 0x1 */
>       ...
>   };
> 
>   gpio@2,12 {
>       reg = <2 0x12 0x1>; /* CS2, offset 0x12, IO size 0x1 */
>       ...
>   };
> 
>   gpio@2,14 {
>       reg = <2 0x14 0x1>; /* CS2, offset 0x14, IO size 0x1 */
>       ...
>   };

Trim it, two entries might be enough to illustrate it.

> 
> - dmesg
> 
>   [    1.596402] omap-gpmc 50000000.gpmc: cannot request GPMC CS 2
>   [    1.596434] omap-gpmc 50000000.gpmc: failed to probe DT child 'gpio': -16
>   [    1.596489] omap-gpmc 50000000.gpmc: cannot request GPMC CS 2
>   [    1.596511] omap-gpmc 50000000.gpmc: failed to probe DT child 'gpio': -16
>   [    1.596564] omap-gpmc 50000000.gpmc: cannot request GPMC CS 2
>   [    1.596586] omap-gpmc 50000000.gpmc: failed to probe DT child 'gpio': -16
> 
>   ("gpio@2,2" is ok, "gpio@2,10", "gpio@2,12", "gpio@...4" are fail)
> 
> Fixes: c2ade654dbf7d02f09ad491f5621fc321d4af96b
> ("memory: omap-gpmc: Use of_node_name_eq for node name comparisons")

Also not correct tag. Run checkpatch.

> 

No blank lines.

> Signed-off-by: INAGAKI Hiroshi <musashino.open@...il.com>


> ---
>  drivers/memory/omap-gpmc.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/drivers/memory/omap-gpmc.c b/drivers/memory/omap-gpmc.c
> index d78f73db37c8..3e3e84e34795 100644
> --- a/drivers/memory/omap-gpmc.c
> +++ b/drivers/memory/omap-gpmc.c
> @@ -2202,7 +2202,7 @@ static int gpmc_probe_generic_child(struct platform_device *pdev,
>  		dev_err(&pdev->dev, "cannot request GPMC CS %d\n", cs);
>  		return ret;
>  	}
> -	gpmc_cs_set_name(cs, child->full_name);
> +	gpmc_cs_set_name(cs, child->name);
>  
>  	gpmc_read_settings_dt(child, &gpmc_s);
>  	gpmc_read_timings_dt(child, &gpmc_t);
> 
> base-commit: 13f35b3c72f4075e13a974f439b20b9e26f8f243

Best regards,
Krzysztof