wireless: Print wiphy name in sysfs.
PATCH 1/1 wireless: Print wiphy name in sysfs.
From: Ben Greear <[email protected]>
The index cannot be used to reliably reconstruct a phy name, so explicitly add the phy name to sysfs so that scripts can figure out the parent phy device for a particular wireless interface.
---
:100644 100644 74a9e3c... 4294fa2... M net/wireless/sysfs.c
net/wireless/sysfs.c | 9 +++++++++ 1 files changed, 9 insertions(+), 0 deletions(-)
net/wireless/sysfs.c +9 -0
@@ -35,6 +35,14 @@ SHOW_FMT(index, "%d", wiphy_idx);SHOW_FMT (macaddress,"%pM" , wiphy.perm_addr);SHOW_FMT (address_mask,"%pM" , wiphy.addr_mask); +static ssize_t name_show (struct device *dev, +struct device_attribute *attr, +char *buf) { +struct wiphy *wiphy = &dev_to_rdev (dev)->wiphy; +return sprintf (buf,"%s\n" ,dev_name (&wiphy->dev)); +} + +static ssize_t addresses_show (struct device *dev,struct device_attribute *attr,char *buf)
@@ -57,6 +65,7 @@ static struct device_attribute ieee80211_dev_attrs[] = {__ATTR_RO (macaddress),__ATTR_RO (address_mask),__ATTR_RO (addresses), + __ATTR_RO(name), {} };
Signature · 1 line
1.7.2.2
Johannes Berg
maintainerQuoted 4 lines · On Mon, 2010-10-11 at 10:28 -0700, [email protected] wrote:
The index cannot be used to reliably reconstruct a phy name, so explicitly add the phy name to sysfs so that scripts can figure out the parent phy device for a particular wireless interface.
But you still need to know the name to get at this to start with, so what's the point?
cat /sys/class/ieee80211/<phyname>/name
<phyname>
or
cat /sys/devices/pci0000:00/0000:00:15.0/0000:03:00.0/ssb0:0/ieee80211/phy0/name phy0
johannes
Quoted 10 lines · Johannes Berg
Quoted 4 lines · On Mon, 2010-10-11 at 10:28 -0700, [email protected] wrote:
The index cannot be used to reliably reconstruct a phy name, so explicitly add the phy name to sysfs so that scripts can figure out the parent phy device for a particular wireless interface.
But you still need to know the name to get at this to start with, so what's the point?
Quoted 7 lines
cat /sys/class/ieee80211/<phyname>/name
<phyname>or
cat /sys/devices/pci0000:00/0000:00:15.0/0000:03:00.0/ssb0:0/ieee80211/phy0/name phy0
I use this to determine the phy device for a particular network device:
[root@dx-2u-ath5k lanforge]# cat /sys/class/net/wlan0/phy80211/name wiphy0
Thanks,
Ben
Signature · 2 lines
Ben Greear <[email protected]> Candela Technologies Inc http://www.candelatech.com
Johannes Berg
maintainerQuoted 4 lines · Ben Greear
I use this to determine the phy device for a particular network device:
[root@dx-2u-ath5k lanforge]# cat
/sys/class/net/wlan0/phy80211/namewiphy0
What's wrong with
basename $(readlink /sys/class/net/wlan0/phy80211)
?
johannes
Quoted 11 lines · Johannes Berg
Quoted 4 lines · Ben Greear
I use this to determine the phy device for a particular network device:
[root@dx-2u-ath5k lanforge]# cat
/sys/class/net/wlan0/phy80211/namewiphy0What's wrong with
basename $(readlink
/sys/class/net/wlan0/phy80211)
It looks like that could work, but it seems a lot easier to probe network devices via normal manner, and then check if I can just open and read /sys/class/net/[dev-name]/phy80211/name
I can do that easily from within a C/C++ program, where as doing system calls and grabbing standard output is more of a pain (and more overhead).
But, if you don't want to add my patch, it's not a big deal.
Thanks,
Ben
Signature · 2 lines
Ben Greear <[email protected]> Candela Technologies Inc http://www.candelatech.com
Johannes Berg
maintainerQuoted 11 lines · Ben Greear
Quoted 1 line
basename $(readlink
/sys/class/net/wlan0/phy80211)It looks like that could work, but it seems a lot easier to probe network devices via normal manner, and then check if I can just open and read /sys/class/net/[dev-name]/phy80211/name
I can do that easily from within a C/C++ program, where as doing system calls and grabbing standard output is more of a pain (and more overhead).
But, if you don't want to add my patch, it's not a big deal.
I don't really mind, it just didn't seem necessary. FWIW, readlink(2) is also a system call that your C program could use, but if you want the patch instead I guess that's fine.
johannes
Quoted 12 lines · Johannes Berg
Quoted 10 lines · Ben Greear
Quoted 1 line
basename $(readlink
/sys/class/net/wlan0/phy80211)It looks like that could work, but it seems a lot easier to probe network devices via normal manner, and then check if I can just open and read /sys/class/net/[dev-name]/phy80211/name
I can do that easily from within a C/C++ program, where as doing system calls and grabbing standard output is more of a pain (and more overhead).
Quoted 5 lines
Quoted 1 line
But, if you don't want to add my patch, it's not a big deal.
I don't really mind, it just didn't seem necessary. FWIW, readlink(2) is also a system call that your C program could use, but if you want the patch instead I guess that's fine.
I think it offers a bit of use at a small cost, but certainly it's not a critical improvement.
If the maintainer picks it up, I'll happily use it..if not, I can carry it in my own tree or use readdir or something.
Thanks,
Ben
Signature · 2 lines
Ben Greear <[email protected]> Candela Technologies Inc http://www.candelatech.com