Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1209388 > unrolled thread

[PATCH] staging: wilc1000: Set all options in region debugfs file

Back to article view | Back to linux.kernel

Contents

  [PATCH] staging: wilc1000: Set all options in region debugfs file Chandra S Gorentla <csgorentla@gmail.com> - 2015-08-18 19:10 +0200
    Re: [PATCH] staging: wilc1000: Set all options in region debugfs file Greg KH <gregkh@linuxfoundation.org> - 2015-08-19 05:10 +0200
      Re: [PATCH] staging: wilc1000: Set all options in region debugfs file Chandra Gorentla <csgorentla@gmail.com> - 2015-08-19 14:40 +0200
        Re: [PATCH] staging: wilc1000: Set all options in region debugfs file Greg KH <gregkh@linuxfoundation.org> - 2015-08-19 17:10 +0200

#1209388 — [PATCH] staging: wilc1000: Set all options in region debugfs file

This patch allows setting all options in the module's debug region
options file 'wilc_debug_region'.  This functionality allows the user
to enable logging from all regions (initialization, locks, firmware
etc.) of the driver.  Logging from the following regions is enabled
during the driver initialization:

INIT_DBG, GENERIC_DBG, CFG80211_DBG, FIRM_DBG and HOSTAPD_DBG

Before this change, the numerical value set is equal first byte of 
input minus 0x30 (ASCII value of '0').  Because of this, after a write 
to this debugfs file, it is difficult to predict the regions on which
logging is enabled.

The DBG_REGION_ALL now includes 3 additional regions TCP_ENH, SPIN_DEBUG
and FIRM_DBG.

Before this change, the region flag FIRM_DBG is enabled during
initialization but cleared after a write to the debugfs file.


Signed-off-by: Chandra S Gorentla <csgorentla@gmail.com>
---

It is verified by code walk that while generating logs for the flags
TCP_ENH and SPIN_DEBUG code does not try to print strings that are not
terminated by NULL and variables do not try to access invalid
memory locations.

 drivers/staging/wilc1000/wilc_debugfs.c | 24 ++++++++++++------------
 1 file changed, 12 insertions(+), 12 deletions(-)

diff --git a/drivers/staging/wilc1000/wilc_debugfs.c b/drivers/staging/wilc1000/wilc_debugfs.c
index ae11186..70d98ee 100644
--- a/drivers/staging/wilc1000/wilc_debugfs.c
+++ b/drivers/staging/wilc1000/wilc_debugfs.c
@@ -24,7 +24,10 @@ static struct dentry *wilc_dir;
  * --------------------------------------------------------------------------------
  */
 
-#define DBG_REGION_ALL	(GENERIC_DBG | HOSTAPD_DBG | HOSTINF_DBG | CORECONFIG_DBG | CFG80211_DBG | INT_DBG | TX_DBG | RX_DBG | LOCK_DBG | INIT_DBG | BUS_DBG | MEM_DBG)
+#define DBG_REGION_ALL	(GENERIC_DBG | HOSTAPD_DBG | HOSTINF_DBG |	\
+			CORECONFIG_DBG | CFG80211_DBG | INT_DBG |	\
+			TX_DBG | RX_DBG | LOCK_DBG | INIT_DBG |		\
+			BUS_DBG | MEM_DBG | TCP_ENH | SPIN_DEBUG | FIRM_DBG)
 #define DBG_LEVEL_ALL	(DEBUG | INFO | WRN | ERR)
 atomic_t REGION = ATOMIC_INIT(INIT_DBG | GENERIC_DBG | CFG80211_DBG | FIRM_DBG | HOSTAPD_DBG);
 atomic_t DEBUG_LEVEL = ATOMIC_INIT(ERR);
@@ -87,23 +90,20 @@ static ssize_t wilc_debug_region_read(struct file *file, char __user *userbuf, s
 	return simple_read_from_buffer(userbuf, count, ppos, buf, res);
 }
 
-static ssize_t wilc_debug_region_write(struct file *filp, const char *buf, size_t count, loff_t *ppos)
+static ssize_t wilc_debug_region_write(struct file *filp,
+					const char __user *buf,
+					size_t count, loff_t *ppos)
 {
-	char buffer[128] = {};
 	int flag;
+	int ret;
 
-	if (count > sizeof(buffer))
-		return -EINVAL;
-
-	if (copy_from_user(buffer, buf, count)) {
-		return -EFAULT;
-	}
-
-	flag = buffer[0] - '0';
+	ret = kstrtouint_from_user(buf, count, 16, &flag);
+	if (ret)
+		return ret;
 
 	if (flag > DBG_REGION_ALL) {
 		printk("%s, value (0x%08x) is out of range, stay previous flag (0x%08x)\n", __func__, flag, atomic_read(&REGION));
-		return -EFAULT;
+		return -EINVAL;
 	}
 
 	atomic_set(&REGION, (int)flag);
-- 
2.5.0

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

[toc] | [next] | [standalone]

#1209582

On Tue, Aug 18, 2015 at 10:32:17PM +0530, Chandra S Gorentla wrote:
> This patch allows setting all options in the module's debug region
> options file 'wilc_debug_region'.  This functionality allows the user
> to enable logging from all regions (initialization, locks, firmware
> etc.) of the driver.  Logging from the following regions is enabled
> during the driver initialization:
> 
> INIT_DBG, GENERIC_DBG, CFG80211_DBG, FIRM_DBG and HOSTAPD_DBG
> 
> Before this change, the numerical value set is equal first byte of 
> input minus 0x30 (ASCII value of '0').  Because of this, after a write 
> to this debugfs file, it is difficult to predict the regions on which
> logging is enabled.
> 
> The DBG_REGION_ALL now includes 3 additional regions TCP_ENH, SPIN_DEBUG
> and FIRM_DBG.

Why did you add these extra ones?

All of this should eventually just be deleted, as network drivers need
to use the networking driver debug interfaces, not their own crazy ones.

thanks,

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

[toc] | [prev] | [next] | [standalone]

#1209816

On Tue, Aug 18, 2015 at 08:01:00PM -0700, Greg KH wrote:
> On Tue, Aug 18, 2015 at 10:32:17PM +0530, Chandra S Gorentla wrote:
> > This patch allows setting all options in the module's debug region
> > options file 'wilc_debug_region'.  This functionality allows the user
> > to enable logging from all regions (initialization, locks, firmware
> > etc.) of the driver.  Logging from the following regions is enabled
> > during the driver initialization:
> > 
> > INIT_DBG, GENERIC_DBG, CFG80211_DBG, FIRM_DBG and HOSTAPD_DBG
> > 
> > Before this change, the numerical value set is equal first byte of 
> > input minus 0x30 (ASCII value of '0').  Because of this, after a write 
> > to this debugfs file, it is difficult to predict the regions on which
> > logging is enabled.
> > 
> > The DBG_REGION_ALL now includes 3 additional regions TCP_ENH, SPIN_DEBUG
> > and FIRM_DBG.
> 
> Why did you add these extra ones?
I added them because there is code support them and to avoid a holes in
the range of the options.
> 
> All of this should eventually just be deleted, as network drivers need
> to use the networking driver debug interfaces, not their own crazy ones.
In that case, can I assume that we are not going forward with this change?
> 
> thanks,
> 
> greg k-h

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

[toc] | [prev] | [next] | [standalone]

#1209884

On Wed, Aug 19, 2015 at 06:00:12PM +0530, Chandra Gorentla wrote:
> On Tue, Aug 18, 2015 at 08:01:00PM -0700, Greg KH wrote:
> > On Tue, Aug 18, 2015 at 10:32:17PM +0530, Chandra S Gorentla wrote:
> > > This patch allows setting all options in the module's debug region
> > > options file 'wilc_debug_region'.  This functionality allows the user
> > > to enable logging from all regions (initialization, locks, firmware
> > > etc.) of the driver.  Logging from the following regions is enabled
> > > during the driver initialization:
> > > 
> > > INIT_DBG, GENERIC_DBG, CFG80211_DBG, FIRM_DBG and HOSTAPD_DBG
> > > 
> > > Before this change, the numerical value set is equal first byte of 
> > > input minus 0x30 (ASCII value of '0').  Because of this, after a write 
> > > to this debugfs file, it is difficult to predict the regions on which
> > > logging is enabled.
> > > 
> > > The DBG_REGION_ALL now includes 3 additional regions TCP_ENH, SPIN_DEBUG
> > > and FIRM_DBG.
> > 
> > Why did you add these extra ones?
> I added them because there is code support them and to avoid a holes in
> the range of the options.

But why do you need to debug such things?

> > All of this should eventually just be deleted, as network drivers need
> > to use the networking driver debug interfaces, not their own crazy ones.
> In that case, can I assume that we are not going forward with this change?

I hope not, please work on fixing the driver up to work properly (i.e.
not with this type of stuff...)

thanks,

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

[toc] | [prev] | [standalone]

Back to top | Article view | linux.kernel

csiph-web