[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20191206124508.GA1360047@kroah.com>
Date: Fri, 6 Dec 2019 13:45:08 +0100
From: Greg KH <gregkh@...uxfoundation.org>
To: Sourabh Jain <sourabhjain@...ux.ibm.com>
Cc: mpe@...erman.id.au, mahesh@...ux.vnet.ibm.com,
hbathini@...ux.ibm.com, linux-kernel@...r.kernel.org,
linuxppc-dev@...abs.org, corbet@....net, linux-doc@...r.kernel.org
Subject: Re: [PATCH v4 3/6] powerpc/fadump: reorganize /sys/kernel/fadump_*
sysfs files
On Fri, Dec 06, 2019 at 05:54:31PM +0530, Sourabh Jain wrote:
> +static struct kobj_attribute release_attr = __ATTR(release_mem,
> 0200, NULL,
> fadump_release_memory_store);
> -static struct kobj_attribute fadump_attr = __ATTR(fadump_enabled,
> +static struct kobj_attribute enable_attr = __ATTR(enabled,
> 0444, fadump_enabled_show,
> NULL);
__ATTR_RO()?
> -static struct kobj_attribute fadump_register_attr = __ATTR(fadump_registered,
> +static struct kobj_attribute register_attr = __ATTR(registered,
> 0644, fadump_register_show,
> fadump_register_store);
__ATTR_RW()?
And then use an ATTRIBUTE_GROUP() macro to create a group so that you
then can do:
> @@ -1452,11 +1450,47 @@ static void fadump_init_files(void)
> printk(KERN_ERR "fadump: unable to create debugfs file"
> " fadump_region\n");
>
> + rc = sysfs_create_file(fadump_kobj, &enable_attr.attr);
> + if (rc)
> + pr_err("unable to create enabled sysfs file (%d)\n",
> + rc);
> + rc = sysfs_create_file(fadump_kobj, ®ister_attr.attr);
> + if (rc)
> + pr_err("unable to create registered sysfs file (%d)\n",
> + rc);
> + if (fw_dump.dump_active) {
> + rc = sysfs_create_file(fadump_kobj, &release_attr.attr);
> + if (rc)
> + pr_err("unable to create release_mem sysfs file (%d)\n",
> + rc);
> + }
a single call to sysfs_create_groups() here instead of trying to unwind
the mess if something went wrong.
thanks,
greg k-h
Powered by blists - more mailing lists