Skip to contentSkip to conversation

parport: add device-model to parport subsystem

PATCH Superseded Sudip Mukherjee 4 participants 11 messages Apr 21, 2015 – Apr 25, 2015

Subsystems (2)

v2 of 5 Superseded Latest is v5 →

lore
Share & tools
lore

Sudip Mukherjee

author maintainer
#0 lore raw

PATCH v2 1/2 parport: add device-model to parport subsystem

This is again another WIP for your review.
almost all the points raised by Greg and Dan has been covered in this patch.

apart from those points, I found another problem with my previous code, that whenever any driver is trying to register, it is getting bound to all the available parallel ports. To solve that probe was required and it has been used in this version. Now the driver is binding only to the device it has registered. And that device will appear as a subdevice of the particular parallel port it wants to use.

In the previous version Greg mentioned that we do not need to maintain our own list now. That has been partially done in this version, we are no longet maintaining the list of drivers. But we still need to maintain the list of ports and devices as that will be used by drivers which are not yet converted to device model. When all drivers are converted to use the device-model parallel port all these lists can be removed.

Just a little off-topic: I am using the staging/panel to test this code, and last few days i had a very tough time finding a stackdump. I was always thinking that it is happening because of the changes I have done. But it turned out to be a bug in the panel code. And incidentally, with the changes done in this version that bug is also resolved.

---
 drivers/parport/procfs.c |  15 ++-
 drivers/parport/share.c  | 243 ++++++++++++++++++++++++++++++++++++++++++++---
 include/linux/parport.h  |  41 +++++++-
 3 files changed, 283 insertions(+), 16 deletions(-)
drivers/parport/procfs.c +14 -1
index 3b47080..c776333 100644
@@ -21,6 +21,7 @@
 #include <linux/parport.h>
 #include <linux/ctype.h>
 #include <linux/sysctl.h>
+#include <linux/device.h>
 
 #include <asm/uaccess.h>
 
@@ -558,8 +559,18 @@ int parport_device_proc_unregister(struct pardevice *device)
 
 static int __init parport_default_proc_register(void)
 {
+	int ret;
+
 	parport_default_sysctl_table.sysctl_header =
 		register_sysctl_table(parport_default_sysctl_table.dev_dir);
+	if (!parport_default_sysctl_table.sysctl_header)
+		return -ENOMEM;
+	ret = parport_bus_init();
+	if (ret) {
+		unregister_sysctl_table(parport_default_sysctl_table.
+					sysctl_header);
+		return ret;
+	}
 	return 0;
 }
 
@@ -570,6 +581,7 @@ static void __exit parport_default_proc_unregister(void)
 					sysctl_header);
 		parport_default_sysctl_table.sysctl_header = NULL;
 	}
+	parport_bus_exit();
 }
 
 #else /* no sysctl or no procfs*/
@@ -596,11 +608,12 @@ int parport_device_proc_unregister(struct pardevice *device)
 
 static int __init parport_default_proc_register (void)
 {
-	return 0;
+	return parport_bus_init();
 }
 
 static void __exit parport_default_proc_unregister (void)
 {
+	parport_bus_exit();
 }
 #endif
 
drivers/parport/share.c +229 -14

332 lines not on this page · raw message

include/linux/parport.h +40 -1

100 lines not on this page · raw message

Signature · 1 line
1.8.1.2
author maintainer
#1 lore raw

PATCH v2 2/2 staging: panel: modify driver to use new parport device model

modify panel driver to use the new parallel port device model.

---
 drivers/staging/panel/panel.c | 53 +++++++++++++++----------------------------
 1 file changed, 18 insertions(+), 35 deletions(-)
drivers/staging/panel/panel.c +18 -35
index ea54fb4..e9511ee 100644
@@ -2188,6 +2188,19 @@ static struct notifier_block panel_notifier = {
 	0
 };
 
+static int panel_probe(struct device *dev)
+{
+	if (strcmp(dev_name(dev), "panel"))
+		return -ENODEV;
+
+	return 0;
+}
+
+struct pardev_cb panel_cb = {
+	.flags = 0,	/*PARPORT_DEV_EXCL */
+	.private = &pprt,
+};
+
 static void panel_attach(struct parport *port)
 {
 	if (port->number != parport)
@@ -2199,10 +2212,7 @@ static void panel_attach(struct parport *port)
 		return;
 	}
 
-	pprt = parport_register_device(port, "panel", NULL, NULL,  /* pf, kf */
-				       NULL,
-				       /*PARPORT_DEV_EXCL */
-				       0, (void *)&pprt);
+	pprt = parport_register_dev(port, "panel", &panel_cb);
 	if (pprt == NULL) {
 		pr_err("%s: port->number=%d parport=%d, parport_register_device() failed\n",
 		       __func__, port->number, parport);
@@ -2240,38 +2250,10 @@ err_unreg_device:
 	pprt = NULL;
 }
 
-static void panel_detach(struct parport *port)
-{
-	if (port->number != parport)
-		return;
-
-	if (!pprt) {
-		pr_err("%s: port->number=%d parport=%d, nothing to unregister.\n",
-		       __func__, port->number, parport);
-		return;
-	}
-
-	unregister_reboot_notifier(&panel_notifier);
-
-	if (keypad.enabled && keypad_initialized) {
-		misc_deregister(&keypad_dev);
-		keypad_initialized = 0;
-	}
-
-	if (lcd.enabled && lcd.initialized) {
-		misc_deregister(&lcd_dev);
-		lcd.initialized = false;
-	}
-
-	parport_release(pprt);
-	parport_unregister_device(pprt);
-	pprt = NULL;
-}
-
 static struct parport_driver panel_driver = {
 	.name = "panel",
-	.attach = panel_attach,
-	.detach = panel_detach,
+	.check = panel_attach,
+	.probe = panel_probe,
 };
 
 /* init function */
@@ -2380,7 +2362,7 @@ static int __init panel_init_module(void)
 		return -ENODEV;
 	}
 
-	err = parport_register_driver(&panel_driver);
+	err = parport_register_drv(&panel_driver);
 	if (err) {
 		pr_err("could not register with parport. Aborting.\n");
 		return err;
@@ -2398,6 +2380,7 @@ static int __init panel_init_module(void)
 
 static void __exit panel_cleanup_module(void)
 {
+	unregister_reboot_notifier(&panel_notifier);
 
 	if (scan_timer.function != NULL)
 		del_timer_sync(&scan_timer);
Signature · 1 line
1.8.1.2
#2 lore raw
In reply to Sudip Mukherjee
Quoted 97 lines · Sudip Mukherjee

This is again another WIP for your review.
almost all the points raised by Greg and Dan has been covered in this patch.

apart from those points, I found another problem with my previous code, that whenever any driver is trying to register, it is getting bound to all the available parallel ports. To solve that probe was required and it has been used in this version. Now the driver is binding only to the device it has registered. And that device will appear as a subdevice of the particular parallel port it wants to use.

In the previous version Greg mentioned that we do not need to maintain our own list now. That has been partially done in this version, we are no longet maintaining the list of drivers. But we still need to maintain the list of ports and devices as that will be used by drivers which are not yet converted to device model. When all drivers are converted to use the device-model parallel port all these lists can be removed.

Just a little off-topic: I am using the staging/panel to test this code, and last few days i had a very tough time finding a stackdump. I was always thinking that it is happening because of the changes I have done. But it turned out to be a bug in the panel code. And incidentally, with the changes done in this version that bug is also resolved.

---
 drivers/parport/procfs.c |  15 ++-
 drivers/parport/share.c  | 243 ++++++++++++++++++++++++++++++++++++++++++++---
 include/linux/parport.h  |  41 +++++++-
 3 files changed, 283 insertions(+), 16 deletions(-)
drivers/parport/procfs.c +14 -1
index 3b47080..c776333 100644
@@ -21,6 +21,7 @@
 #include <linux/parport.h>
 #include <linux/ctype.h>
 #include <linux/sysctl.h>
+#include <linux/device.h>
 
 #include <asm/uaccess.h>
 
@@ -558,8 +559,18 @@ int parport_device_proc_unregister(struct pardevice *device)
 
 static int __init parport_default_proc_register(void)
 {
+	int ret;
+
 	parport_default_sysctl_table.sysctl_header =
 		register_sysctl_table(parport_default_sysctl_table.dev_dir);
+	if (!parport_default_sysctl_table.sysctl_header)
+		return -ENOMEM;
+	ret = parport_bus_init();
+	if (ret) {
+		unregister_sysctl_table(parport_default_sysctl_table.
+					sysctl_header);
+		return ret;
+	}
 	return 0;
 }
 
@@ -570,6 +581,7 @@ static void __exit parport_default_proc_unregister(void)
 					sysctl_header);
 		parport_default_sysctl_table.sysctl_header = NULL;
 	}
+	parport_bus_exit();
 }
 
 #else /* no sysctl or no procfs*/
@@ -596,11 +608,12 @@ int parport_device_proc_unregister(struct pardevice *device)
 
 static int __init parport_default_proc_register (void)
 {
-	return 0;
+	return parport_bus_init();
 }
 
 static void __exit parport_default_proc_unregister (void)
 {
+	parport_bus_exit();
 }
 #endif
 
drivers/parport/share.c +1 -0
index 3fa6624..8a28c54 100644
@@ -29,6 +29,7 @@
 #include <linux/slab.h>
 #include <linux/sched.h>
 #include <linux/kmod.h>
+#include <linux/device.h>
 
 #include <linux/spinlock.h>
 #include <linux/mutex.h>

drivers/parport/share.c static struct parport_operations dead_ops = { · line 106

Quoted 6 lines
drivers/parport/share.c +3 -0
@@ -100,6 +101,20 @@ static struct parport_operations dead_ops = {
 	.owner		= NULL,
 };
 
+struct bus_type parport_bus_type = {
+	.name           = "parport",
+};

Can you make this static? The indenting is a bit off.

Quoted 27 lines
drivers/parport/share.c +13 -0
+
+int parport_bus_init(void)
+{
+	return bus_register(&parport_bus_type);
+}
+
+void parport_bus_exit(void)
+{
+	bus_unregister(&parport_bus_type);
+}
+
 /* Call attach(port) for each registered driver. */
 static void attach_driver_chain(struct parport *port)
 {
@@ -157,6 +172,7 @@ int parport_register_driver (struct parport_driver *drv)
 
 	if (list_empty(&portlist))
 		get_lowlevel_driver ();
+	drv->devmodel = false;
 
 	mutex_lock(&registration_lock);
 	list_for_each_entry(port, &portlist, list)
@@ -167,6 +183,61 @@ int parport_register_driver (struct parport_driver *drv)
 	return 0;
 }
 
+/*

drivers/parport/share.c int parport_register_driver (struct parport_driver *drv) · line 196

Quoted 10 lines
drivers/parport/share.c +10 -0
+ * iterates through all the devices connected to the bus and sends the device
+ * details to the check callback of the driver, so that the driver can know
+ * what are all the ports that are connected to the bus and choose the port to
+ * which it wants to register its device.
+ */
+static int port_check(struct device *dev, void *drv)
+{
+	((struct parport_driver *)drv)->check(to_parport_dev(dev));
+	return 0;
+}

The cast isn't beautiful. Do this:

static int port_check(struct device *dev, void *_drv)

{
	struct parport_driver *drv = _drv;

	drv->check(to_parport_dev(dev));
	return 0;
}

What is the point of the check function really? The name isn't clear.

Since it always returns zero that means we loop through all the devices and then returns NULL. It feels like a function called bus_find_device() should find something. We have bus_for_each_dev() if we just want to iterate.

Quoted 141 lines
drivers/parport/share.c +69 -11
+
+/*
+ * __parport_register_drv - register a new parport driver
+ * @drv: the driver structure to register
+ * @owner: owner module of drv
+ * @mod_name: module name string
+ *
+ * Adds the driver structure to the list of registered drivers.
+ * Returns a negative value on error, otherwise 0.
+ * If no error occurred, the driver remains registered even if
+ * no device was claimed during registration.
+ */
+int __parport_register_drv(struct parport_driver *drv,
+			   struct module *owner, const char *mod_name)
+{
+	int ret;
+
+	if (!drv->probe) {
+		pr_err("please use probe in %s\n", drv->name);
+		return -ENODEV;
+	}
+
+	if (list_empty(&portlist))
+		get_lowlevel_driver();
+
+	/* initialize common driver fields */
+	drv->driver.name = drv->name;
+	drv->driver.bus = &parport_bus_type;
+	drv->driver.owner = owner;
+	drv->driver.mod_name = mod_name;
+	drv->driver.probe = drv->probe;
+	drv->devmodel = true;
+	ret = driver_register(&drv->driver);
+	if (ret)
+		return ret;
+
+	mutex_lock(&registration_lock);
+	bus_find_device(&parport_bus_type, NULL, drv, port_check);
+	mutex_unlock(&registration_lock);
+
+	return 0;
+}
+EXPORT_SYMBOL(__parport_register_drv);
+
 /**
  *	parport_unregister_driver - deregister a parallel port device driver
  *	@drv: structure describing the driver that was given to
@@ -189,15 +260,21 @@ void parport_unregister_driver (struct parport_driver *drv)
 	struct parport *port;
 
 	mutex_lock(&registration_lock);
-	list_del_init(&drv->list);
-	list_for_each_entry(port, &portlist, list)
-		drv->detach(port);
+	if (drv->devmodel) {
+		driver_unregister(&drv->driver);
+	} else {
+		list_del_init(&drv->list);
+		list_for_each_entry(port, &portlist, list)
+			drv->detach(port);
+	}
 	mutex_unlock(&registration_lock);
 }
 
-static void free_port (struct parport *port)
+static void free_port(struct device *dev)
 {
 	int d;
+	struct parport *port = to_parport_dev(dev);
+
 	spin_lock(&full_list_lock);
 	list_del(&port->full_list);
 	spin_unlock(&full_list_lock);
@@ -223,8 +300,9 @@ static void free_port (struct parport *port)
 
 struct parport *parport_get_port (struct parport *port)
 {
-	atomic_inc (&port->ref_count);
-	return port;
+	struct device *dev = get_device(&port->ddev);
+
+	return to_parport_dev(dev);
 }
 
 /**
@@ -237,11 +315,7 @@ struct parport *parport_get_port (struct parport *port)
 
 void parport_put_port (struct parport *port)
 {
-	if (atomic_dec_and_test (&port->ref_count))
-		/* Can destroy it now. */
-		free_port (port);
-
-	return;
+	put_device(&port->ddev);
 }
 
 /**
@@ -281,6 +355,7 @@ struct parport *parport_register_port(unsigned long base, int irq, int dma,
 	int num;
 	int device;
 	char *name;
+	int ret;
 
 	tmp = kzalloc(sizeof(struct parport), GFP_KERNEL);
 	if (!tmp) {
@@ -333,6 +408,9 @@ struct parport *parport_register_port(unsigned long base, int irq, int dma,
 	 */
 	sprintf(name, "parport%d", tmp->portnum = tmp->number);
 	tmp->name = name;
+	tmp->ddev.bus = &parport_bus_type;
+	tmp->ddev.release = free_port;
+	dev_set_name(&tmp->ddev, name);
 
 	for (device = 0; device < 5; device++)
 		/* assume the worst */
@@ -340,6 +418,12 @@ struct parport *parport_register_port(unsigned long base, int irq, int dma,
 
 	tmp->waithead = tmp->waittail = NULL;
 
+	ret = device_register(&tmp->ddev);
+	if (ret) {
+		put_device(&tmp->ddev);
+		return NULL;
+	}
+
 	return tmp;
 }
 
@@ -575,6 +659,7 @@ parport_register_device(struct parport *port, const char *name,
 	tmp->irq_func = irq_func;
 	tmp->waiting = 0;
 	tmp->timeout = 5 * HZ;
+	tmp->devmodel = false;
 
 	/* Chain this onto the list */
 	tmp->prev = NULL;
@@ -630,6 +715,133 @@ parport_register_device(struct parport *port, const char *name,
 	return NULL;
 }
 

drivers/parport/share.c parport_register_device(struct parport *port, const char *name, · line 727

Quoted 10 lines
drivers/parport/share.c +10 -0
+static void free_pardevice(struct device *dev)
+{
+	struct pardevice *par_dev = to_pardevice(dev);
+
+	kfree(par_dev->name);
+	kfree(par_dev);
+}
+
+struct pardevice *
+parport_register_dev(struct parport *port, const char *name,

Please choose a better name like parport_register_dev_model() or something.

drivers/parport/share.c parport_register_device(struct parport *port, const char *name, · line 737

Quoted 10 lines
drivers/parport/share.c +10 -0
+		     struct pardev_cb *par_dev_cb)
+{
+	struct pardevice *par_dev;
+	int ret;
+	char *devname;
+
+	if (port->physport->flags & PARPORT_FLAG_EXCL) {
+		/* An exclusive device is registered. */
+		pr_debug("%s: no more devices allowed\n",
+			 port->name);

This should probably be a pr_err().

Quoted 19 lines
drivers/parport/share.c +19 -0
+		return NULL;
+	}
+
+	if (par_dev_cb->flags & PARPORT_DEV_LURK) {
+		if (!par_dev_cb->preempt || !par_dev_cb->wakeup) {
+			pr_info("%s: refused to register lurking device (%s) without callbacks\n",
+				port->name, name);
+			return NULL;
+		}
+	}
+
+	if (!try_module_get(port->ops->owner))
+		return NULL;
+
+	parport_get_port(port);
+
+	par_dev = kzalloc(sizeof(*par_dev), GFP_KERNEL);
+	if (!par_dev)
+		goto err_par_dev;

drivers/parport/share.c parport_register_device(struct parport *port, const char *name, · line 766

Quoted 10 lines
drivers/parport/share.c +10 -0
+
+	par_dev->state = kzalloc(sizeof(*par_dev->state), GFP_KERNEL);
+	if (!par_dev->state)
+		goto err_put_par_dev;
+
+	devname = kstrdup(name, GFP_KERNEL);
+	if (!devname)
+		goto err_free_par_dev;
+
+	par_dev->name = devname;

The existing code is buggy here as we discussed previously. Could you just fix that before we do anything else? It's freaking me out.

drivers/parport/share.c parport_register_device(struct parport *port, const char *name, · line 779

Quoted 13 lines
drivers/parport/share.c +13 -0
+	par_dev->port = port;
+	par_dev->daisy = -1;
+	par_dev->preempt = par_dev_cb->preempt;
+	par_dev->wakeup = par_dev_cb->wakeup;
+	par_dev->private = par_dev_cb->private;
+	par_dev->flags = par_dev_cb->flags;
+	par_dev->irq_func = par_dev_cb->irq_func;
+	par_dev->waiting = 0;
+	par_dev->timeout = 5 * HZ;
+
+	par_dev->dev.parent = &port->ddev;
+	par_dev->dev.bus = &parport_bus_type;
+	dev_set_name(&par_dev->dev, "%s", devname);

dev_set_name() allocates a copy of "devname". It can fail with -ENOMEM. It's not likely but we check all the other allocations so we may as well check here.

drivers/parport/share.c parport_register_device(struct parport *port, const char *name, · line 784

Quoted 5 lines
drivers/parport/share.c +5 -0
+	par_dev->dev.release = free_pardevice;
+	par_dev->devmodel = true;
+	ret = device_register(&par_dev->dev);
+	if (ret)
+		goto err_free_all;

This is a bad name because as soon as you add a new allocation you will have to change the name. Use something like err_put_dev.

Quoted 117 lines
drivers/parport/share.c +66 -3
+
+	/* Chain this onto the list */
+	par_dev->prev = NULL;
+	/*
+	 * This function must not run from an irq handler so we don' t need
+	 * to clear irq on the local CPU. -arca
+	 */
+	spin_lock(&port->physport->pardevice_lock);
+
+	if (par_dev_cb->flags & PARPORT_DEV_EXCL) {
+		if (port->physport->devices) {
+			spin_unlock(&port->physport->pardevice_lock);
+			pr_debug("%s: cannot grant exclusive access for device %s\n",
+				 port->name, name);
+			goto err_free_all;
+		}
+		port->flags |= PARPORT_FLAG_EXCL;
+	}
+
+	par_dev->next = port->physport->devices;
+	wmb();	/*
+		 * Make sure that tmp->next is written before it's
+		 * added to the list; see comments marked 'no locking
+		 * required'
+		 */
+	if (port->physport->devices)
+		port->physport->devices->prev = par_dev;
+	port->physport->devices = par_dev;
+	spin_unlock(&port->physport->pardevice_lock);
+
+	init_waitqueue_head(&par_dev->wait_q);
+	par_dev->timeslice = parport_default_timeslice;
+	par_dev->waitnext = NULL;
+	par_dev->waitprev = NULL;
+
+	/*
+	 * This has to be run as last thing since init_state may need other
+	 * pardevice fields. -arca
+	 */
+	port->ops->init_state(par_dev, par_dev->state);
+	port->proc_device = par_dev;
+	parport_device_proc_register(par_dev);
+
+	return par_dev;
+
+err_free_all:
+	put_device(&par_dev->dev);
+err_free_par_dev:
+	kfree(par_dev->state);
+err_put_par_dev:
+	if (!par_dev->devmodel)
+		kfree(par_dev);
+err_par_dev:
+	parport_put_port(port);
+	module_put(port->ops->owner);
+
+	return NULL;
+}
+EXPORT_SYMBOL(parport_register_dev);
+
 /**
  *	parport_unregister_device - deregister a device on a parallel port
  *	@dev: pointer to structure representing device
@@ -691,7 +903,10 @@ void parport_unregister_device(struct pardevice *dev)
 	spin_unlock_irq(&port->waitlist_lock);
 
 	kfree(dev->state);
-	kfree(dev);
+	if (dev->devmodel)
+		device_unregister(&dev->dev);
+	else
+		kfree(dev);
 
 	module_put(port->ops->owner);
 	parport_put_port (port);
@@ -926,8 +1141,8 @@ int parport_claim_or_block(struct pardevice *dev)
 			       dev->port->physport->cad ?
 			       dev->port->physport->cad->name:"nobody");
 #endif
-	}
-	dev->waiting = 0;
+	} else if (r == 0)
+		dev->waiting = 0;
 	return r;
 }
 
include/linux/parport.h +5 -0
index c22f125..83e1a6e 100644
@@ -13,6 +13,7 @@
 #include <linux/wait.h>
 #include <linux/irqreturn.h>
 #include <linux/semaphore.h>
+#include <linux/device.h>
 #include <asm/ptrace.h>
 #include <uapi/linux/parport.h>
 
@@ -145,6 +146,8 @@ struct pardevice {
 	unsigned int flags;
 	struct pardevice *next;
 	struct pardevice *prev;
+	struct device dev;
+	bool devmodel;
 	struct parport_state *state;     /* saved status over preemption */
 	wait_queue_head_t wait_q;
 	unsigned long int time;
@@ -156,6 +159,8 @@ struct pardevice {
 	void * sysctl_table;
 };
 
+#define to_pardevice(n) container_of(n, struct pardevice, dev)
+
 /* IEEE1284 information */
 
 /* IEEE1284 phases. These are exposed to userland through ppdev IOCTL

include/linux/parport.h struct parport { · line 203

Quoted 5 lines
include/linux/parport.h +1 -1
@@ -195,7 +200,7 @@ struct parport {
 				 * This may unfortulately be null if the
 				 * port has a legacy driver.
 				 */
-
+	struct device ddev;	/* to link with the bus */

What does ddev stand for? Anyway, it's probably better to call it bus_dev?

regards,
dan carpenter

author maintainer
#3 lore raw
Quoted 1 line · Dan Carpenter

On Tue, Apr 21, 2015 at 07:22:34PM +0530, Sudip Mukherjee wrote:

<snip>

Quoted 8 lines

drivers/parport/share.c static struct parport_operations dead_ops = { · line 106

Quoted 6 lines
drivers/parport/share.c +3 -0
 	.owner		= NULL,
 };
 
+struct bus_type parport_bus_type = {
+	.name           = "parport",
+};

Can you make this static? The indenting is a bit off.

done in v3.

Quoted 1 line

<snip>

Quoted 11 lines

drivers/parport/share.c int parport_register_driver (struct parport_driver *drv) · line 196

Quoted 7 lines

+ * which it wants to register its device.
+ */

drivers/parport/share.c +5 -0
+static int port_check(struct device *dev, void *drv)
+{
+	((struct parport_driver *)drv)->check(to_parport_dev(dev));
+	return 0;
+}

The cast isn't beautiful. Do this:

static int port_check(struct device *dev, void *_drv)

Quoted 8 lines
{
	struct parport_driver *drv = _drv;

	drv->check(to_parport_dev(dev));
	return 0;
}

What is the point of the check function really? The name isn't clear.

yes, i am a bit blunt in thinking of new names, i hope you have noticed that in my naming of the labels .. :)

as the name was not sufficient i mentioned it in the comments. This check function will receive the device details and will decide if it wants to connect to that device. If it wants to connect then it registers its device and mark the port as claimed.
Infact, on second thought, i will return the success or error from check, then if the driver has found the device to connect then we can stop the iteration there.

maybe a better name can be check_port() ?

Quoted 6 lines

Since it always returns zero that means we loop through all the devices and then returns NULL. It feels like a function called bus_find_device() should find something. We have bus_for_each_dev() if we just want to iterate.

yes, bus_for_each_dev() will be better here. thanks.

Quoted 2 lines
Quoted 2 lines
(quoted diff) +2 -0
+
+/*

<snip>

Quoted 5 lines

Review

Quoted 2 lines
(quoted diff) +2 -0
+
+	par_dev->name = devname;

The existing code is buggy here as we discussed previously. Could you just fix that before we do anything else? It's freaking me out.

quoting from your previous mail:

Quoted 1 line

My concern is that it gets freed before we are done using it or something

here, i have modified that and we are no longer using the string passed as an argument. we have duplicated it using kstrdup and using that and it gets freed in free_pardevice().
or am i missing something here?

Quoted 10 lines
Quoted 10 lines
drivers/parport/share.c +9 -0
+	par_dev->port = port;
+	par_dev->daisy = -1;
+	par_dev->preempt = par_dev_cb->preempt;
+	par_dev->wakeup = par_dev_cb->wakeup;
+	par_dev->private = par_dev_cb->private;
+	par_dev->flags = par_dev_cb->flags;
+	par_dev->irq_func = par_dev_cb->irq_func;
+	par_dev->waiting = 0;
+	par_dev->timeout = 5 * HZ;
Quoted 8 lines

drivers/parport/share.c parport_register_device(struct parport *port, const char *name, · line 779

Quoted 4 lines
drivers/parport/share.c +4 -0
+
+	par_dev->dev.parent = &port->ddev;
+	par_dev->dev.bus = &parport_bus_type;
+	dev_set_name(&par_dev->dev, "%s", devname);

dev_set_name() allocates a copy of "devname". It can fail with -ENOMEM. It's not likely but we check all the other allocations so we may as well check here.

ok.

Quoted 9 lines

drivers/parport/share.c parport_register_device(struct parport *port, const char *name, · line 784

Quoted 6 lines
drivers/parport/share.c +5 -0
+	par_dev->dev.release = free_pardevice;
+	par_dev->devmodel = true;
+	ret = device_register(&par_dev->dev);
+	if (ret)
+		goto err_free_all;

This is a bad name because as soon as you add a new allocation you will have to change the name. Use something like err_put_dev.

you have said this in your last mail also. i missed it .. sorry .. :(

Quoted 7 lines

include/linux/parport.h struct parport { · line 203

Quoted 4 lines
include/linux/parport.h +2 -1
+
-
+	struct device ddev;	/* to link with the bus */

What does ddev stand for? Anyway, it's probably better to call it bus_dev?

ok.
I think I will wait today for some comments from Greg and then send in the v3 tomorrow.

thanks for your time and the review.

regards
sudip

Quoted 4 lines

regards,
dan carpenter

#4 lore raw
Quoted 12 lines · Sudip Mukherjee
Quoted 1 line

What is the point of the check function really? The name isn't clear.

yes, i am a bit blunt in thinking of new names, i hope you have noticed that in my naming of the labels .. :)

as the name was not sufficient i mentioned it in the comments. This check function will receive the device details and will decide if it wants to connect to that device. If it wants to connect then it registers its device and mark the port as claimed.
Infact, on second thought, i will return the success or error from check, then if the driver has found the device to connect then we can stop the iteration there.

Quoted 1 line

maybe a better name can be check_port() ?

match() or match_port() something.

Quoted 7 lines
Quoted 6 lines

Since it always returns zero that means we loop through all the devices and then returns NULL. It feels like a function called bus_find_device() should find something. We have bus_for_each_dev() if we just want to iterate.

yes, bus_for_each_dev() will be better here. thanks.

If we're match then bus_find_device() is correct. It's just that's not what v2 did.

Quoted 10 lines
Quoted 3 lines
Quoted 3 lines
(quoted diff) +2 -0
+
+/*

<snip>

Quoted 5 lines

Review

Quoted 2 lines
(quoted diff) +2 -0
+
+	par_dev->name = devname;

The existing code is buggy here as we discussed previously. Could you just fix that before we do anything else? It's freaking me out.

Quoted 7 lines

quoting from your previous mail:

Quoted 1 line

My concern is that it gets freed before we are done using it or something

here, i have modified that and we are no longer using the string passed as an argument. we have duplicated it using kstrdup and using that and it gets freed in free_pardevice().
or am i missing something here?

Ah. Ok. Thanks. I missed that and I don't think the patch has hit linux-next yet.

regards,
dan carpenter

author maintainer
#5 lore raw
Quoted 23 lines · Dan Carpenter
Quoted 12 lines · Sudip Mukherjee
Quoted 1 line

What is the point of the check function really? The name isn't clear.

yes, i am a bit blunt in thinking of new names, i hope you have noticed that in my naming of the labels .. :)

as the name was not sufficient i mentioned it in the comments. This check function will receive the device details and will decide if it wants to connect to that device. If it wants to connect then it registers its device and mark the port as claimed.
Infact, on second thought, i will return the success or error from check, then if the driver has found the device to connect then we can stop the iteration there.

Quoted 1 line

maybe a better name can be check_port() ?

match() or match_port() something.

Quoted 6 lines
Quoted 6 lines

Since it always returns zero that means we loop through all the devices and then returns NULL. It feels like a function called bus_find_device() should find something. We have bus_for_each_dev() if we just want to iterate.

Quoted 4 lines
Quoted 1 line

yes, bus_for_each_dev() will be better here. thanks.

If we're match then bus_find_device() is correct. It's just that's not what v2 did.

that was the main intention but but bus_find_device() will also do a get_device() once a match is found, then in that case I will have to do a put_device() immediately after bus_find_device() completes.

Quoted 14 lines
Quoted 14 lines
Quoted 4 lines
Quoted 4 lines
(quoted diff) +2 -0
+
+/*

<snip>

Quoted 5 lines

Review

Quoted 2 lines
(quoted diff) +2 -0
+
+	par_dev->name = devname;

The existing code is buggy here as we discussed previously. Could you just fix that before we do anything else? It's freaking me out.

quoting from your previous mail:

Quoted 1 line

My concern is that it gets freed before we are done using it or something

Quoted 7 lines
Quoted 4 lines

here, i have modified that and we are no longer using the string passed as an argument. we have duplicated it using kstrdup and using that and it gets freed in free_pardevice().
or am i missing something here?

Ah. Ok. Thanks. I missed that and I don't think the patch has hit linux-next yet.

hehe , no. Greg has to apply the patch, and in the last patch he found some points in his review. I will send in a v3 as soon as the confusion about the bus_find_device() or bus_for_each_dev() clears.

regards
sudip

Quoted 4 lines

regards,
dan carpenter

maintainer
#6 lore raw
In reply to Sudip Mukherjee
Quoted 5 lines · Sudip Mukherjee
Quoted 2 lines

What does ddev stand for? Anyway, it's probably better to call it bus_dev?

ok.
I think I will wait today for some comments from Greg and then send in the v3 tomorrow.

Don't wait for me, it's going to be a week or so at the earliest before I have a chance to look at this, sorry.

greg "burried in email" k-h

author maintainer
#7 lore raw
Quoted 9 lines · Greg KH
Quoted 5 lines · Sudip Mukherjee
Quoted 2 lines

What does ddev stand for? Anyway, it's probably better to call it bus_dev?

ok.
I think I will wait today for some comments from Greg and then send in the v3 tomorrow.

Don't wait for me, it's going to be a week or so at the earliest before I have a chance to look at this, sorry.

No problem. I will change what Dan has suggested and will request Alan and Dean to check on their hardware as well (ofcourse if they have time).

but finally it has to be you as you have to add it to your tree and I know you can not modify your tree during the merge window.

regards
sudip

Quoted 2 lines

greg "burried in email" k-h

#8 lore raw
Quoted 8 lines · On Fri, 24 Apr 2015 13:08:25 +0530 Sudip Mukherjee…
Quoted 7 lines · Greg KH
Quoted 5 lines · Sudip Mukherjee
Quoted 2 lines

What does ddev stand for? Anyway, it's probably better to call it bus_dev?

ok.
I think I will wait today for some comments from Greg and then send in the v3 tomorrow.

Quoted 6 lines
Quoted 2 lines

Don't wait for me, it's going to be a week or so at the earliest before I have a chance to look at this, sorry.

No problem. I will change what Dan has suggested and will request Alan and Dean to check on their hardware as well (ofcourse if they have time).

Let me know when you have a set of patches you feel are good for testing and I'll give them a go. I just picked up a backpack CD-ROM drive on ebay for 99p as well 8)

Alan

author maintainer
#9 lore raw

On Fri, Apr 24, 2015 at 07:37:02PM +0100, One Thousand Gnomes wrote:

Hi Alan,

Quoted 14 lines
Quoted 11 lines · On Fri, 24 Apr 2015 13:08:25 +0530 Sudip Mukherjee…
Quoted 9 lines · Greg KH
Quoted 5 lines · Sudip Mukherjee
Quoted 2 lines

What does ddev stand for? Anyway, it's probably better to call it bus_dev?

ok.
I think I will wait today for some comments from Greg and then send in the v3 tomorrow.

Don't wait for me, it's going to be a week or so at the earliest before I have a chance to look at this, sorry.

Quoted 6 lines
Quoted 3 lines

No problem. I will change what Dan has suggested and will request Alan and Dean to check on their hardware as well (ofcourse if they have time).

Let me know when you have a set of patches you feel are good for testing and I'll give them a go.

This series of WIP is also good for testing. But for next version of WIP I need to know something from Jean about his i2c-parport.
For all other drivers using the parallel port, they are choosing which port to use. But i2c-parport is not specifying that. So as a result, in my test system which has multiple parallel ports, when I load the module, i2c-parport is claiming all the ports with exclusive access. So if i want to use i2c-parport on one parallel port and any other device on another port then that is not possible. If that is the intended behaviour then I need to change some parts in the way I was making the parallel port device-model. Anyways, I have mailed Jean yesterday, and waiting for his comments now. Adding cc to Jean also to this mail.

Quoted 2 lines

I just picked up a backpack CD-ROM drive on ebay
for 99p as well 8)

wow, share me the link please. I will also get one.
I am able to manage only one parallel printer and staging/panel for testing. I am looking for a parallel port zip drive or cdrom or scanner to manage and test the code better. but :(

regards
sudip

Quoted 2 lines

Alan

author maintainer
#10 lore raw
Quoted 2 lines · Sudip Mukherjee

On Fri, Apr 24, 2015 at 07:37:02PM +0100, One Thousand Gnomes wrote:

<snip>

Quoted 3 lines
Quoted 2 lines

Let me know when you have a set of patches you feel are good for testing and I'll give them a go.

This series of WIP is also good for testing.

well, this series will create some ghost parport in /sys/bus/parport, which were being created while probing.
It has been fixed in my local tree and next WIP patch should be a fully working one. But if you are testig this patchset then please use only parport0.

regards
sudip