linux-kernel - Re: [PATCH V2 1/6] SLIMbus: Device management on SLIMbus

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 for Android: free password hash cracker in your pocket

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

Date:	Wed, 17 Jun 2015 11:36:06 -0600
From:	Sagar Dharia <sdharia@...eaurora.org>
To:	Mark Brown <broonie@...nel.org>
CC:	gregkh@...uxfoundation.org, bp@...e.de, poeschel@...onage.de,
	treding@...dia.com, gong.chen@...ux.intel.com,
	andreas.noever@...il.com, alan@...ux.intel.com,
	mathieu.poirier@...aro.org, daniel@...ll.ch, oded.gabbay@....com,
	jkosina@...e.cz, sharon.dvir1@...l.huji.ac.il, joe@...ches.com,
	davem@...emloft.net, james.hogan@...tec.com,
	michael.opdenacker@...e-electrons.com, daniel.thompson@...aro.org,
	linux-kernel@...r.kernel.org, nkaje@...eaurora.org,
	kheitke@...ience.com, mlocke@...eaurora.org, agross@...eaurora.org,
	linux-arm-msm@...r.kernel.org
Subject: Re: [PATCH V2 1/6] SLIMbus: Device management on SLIMbus

On 6/17/2015 7:16 AM, Mark Brown wrote:
> On Tue, Jun 16, 2015 at 07:45:59PM -0600, Sagar Dharia wrote:
>
>> +	if (status) {
>> +		slim_dev->driver = NULL;
>> +	} else if (driver->device_up) {
>> +		ctrl = slim_dev->ctrl;
>> +		queue_work(ctrl->wq, &slim_dev->wd);
>> +	}
> Nothing ever cleans this work up if it didn't manage to run or
> complete.
>
>> +static void slim_report(struct work_struct *work)
>> +{
>> +	struct slim_driver *sbdrv;
>> +	struct slim_device *sbdev = container_of(work, struct slim_device, wd);
>> +
>> +	if (!sbdev->dev.driver)
>> +		return;
> So we just forget about the device if we don't have a driver for it?
If device comes up before driver (e.g. dynamic module driver, or 
late-init driver), then to begin with: this doesn't do anything.
But once the driver-binding happens and probe succeeds, we queue 
device_up again.
>
>> +	/* check if device-up or down needs to be called */
>> +	if ((!sbdev->reported && !sbdev->notified) ||
>> +	    (sbdev->reported && sbdev->notified))
>> +		return;
> No locking here?
You are right, first I thought this was only touched in workqueue 
(singlethreaded), but this is also touched from logical-address 
assignment, driver removal etc.
I will add device-level lock for these.
Thanks
Sagar
>
>> +/**
>> + * slim_ctrl_add_boarddevs: Add devices registered by board-info
>> + * @ctrl: Controller to which these devices are to be added to.
>> + * This API is called by controller when it is up and running.
>> + * If devices on a controller were registered before controller,
>> + * this will make sure that they get probed when controller is up.
>> + */
>> +void slim_ctrl_add_boarddevs(struct slim_controller *ctrl)
> My concerns about the split here still remain.


-- 
Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum,
hosted by The Linux Foundation

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/