rtl8187 wireless issues

rtl8187 wireless issues

[P.G. Richardson]

Hello,

I recently bought a new laptop from Novatech.co.uk
(http://www.novatech.co.uk/novatech/range.html?t=3Dnb&c=3Dall&r=3DX60).

I have installed Mandriva 2009 (2.6.27-7) on the laptop and despite bes=
t
efforts cannot make the internal wifi card connect to my router. The ca=
rd
is reported by lsusb as

id: 0x8189 chipset: rtl8187b

I use WPA2 with a pre-shared key on my router. Would you mind giving me
guidance as to whether this is relibaly working with the rtl8187 driver=
?

Basically, I am unable to associate with the router. Looking at
wpa_supplicant debug output, it indicates the card simply times out whe=
n
trying to associate. However, on one brief occasion a couple of days ag=
o,
I was able to associate according to iwconfig, yet despite best efforts=
 I
was unable to ping anything.

It seems as well that when the driver is installed (currently blacklist=
ed
in modprobe.conf), it stops other network interfaces operating correctl=
y.
=46or example, I am unable to ping from a wired ethernet and unable to
associate with the AP with an externel wireless USB dongle.

I have since tried the 2.6.29 kernel from its git repository but that
version of the rtl8187 driver did not associate either.

At this point I am constrained by lack of debugging knowledge. The only
tools I know of is iwconfig and friends, wpa_supplicant -ddd and ifconf=
ig.
If someone wouldnt mind giving me an idea of how to approach debugging
this problem then I could try and understand what is going on myself.

Any help very much appreciated.

Thanks

P. G. Richardson
(a.k.a phantomjinx)

--=20
Laws are partly formed for the sake of good men, in order to instruct t=
hem
how they may live on friendly terms with one another, and partly for th=
e
sake of those who refuse to be instructed, whose spirit cannot be subdu=
ed,
or softened, or hindered from plunging into evil. [The Laws, Plato]

You fiend! Never have I encountered such corrupt and foul-minded
perversity=85 Have you ever considered a career in the church?
Bishop of Bath and Wells (Blackadder II)

--
To unsubscribe from this list: send the line "unsubscribe linux-wireles=
s" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: rtl8187 wireless issues

[Hin-Tak Leung]

On Thu, Jan 29, 2009 at 12:46 PM, P.G. Richardson
<p.g.richardson@phantomjinx.co.uk> wrote:
> id: 0x8189 chipset: rtl8187b

That's an older rtl8187b - there is no point trying mainline kernels,
as the interesting development are happening
in wireless-testing and takes a few months to get to mainline. i.e.
trying 2.6.29 won't help. Can you download the current compat-wireless
tar ball (http://linuxwireless.org/en/users/Download) and give it a
try?

Also can you show your "dmesg | grep rtl" result ?

The "can't ping" , etc behavior with multiple network interfaces is
normal and quite simple to work around - you need to read up on
"default route" and the "route" command.

Are you using NetworkManager or configuring by hand? If you are not
using NetworkManager, you need to set up dhcp manually. On 2nd
thought, I am suprised you have problem with ping, because if you are
using NetworkManager it should switch the default route, etc if you
choose to switch to the wired interface.

(You have come to the right place - Larry and Herton and I are the
current maintainers of the rtl8187 driver).

Re: rtl8187 wireless issues

[P.G. Richardson]

Hi,

Have compiled and install the latest driver and success!

Just to answer your questions ...

dmesg didnt offer much information other than "link is not ready"

Should have explained the detail of "cannot ping". Essentially I take t=
he
network down doing /etc/init.d/network stop. Change around which
ifcfg-interfaces start on boot then do /etc/init.d/network start. Thus,
the default route was always the active interface.

Do not use NetworkManager and use static ip addresses.

Anyway, hopefully everything is working.

Thanks for the response. Let me know if you require any more info or wo=
uld
like some testing done.

Regards

P.G. Richardson

PS. A minor thing which is probably my fault. I compiled the driver usi=
ng
make then ran make install. The latter installed the new mac82011, ifcf=
g
and b44 drivers. None of the other specific wireless drivers were copie=
d
to the kernel update directory. I had to copy the rtl8187 there manuall=
y.


> On Thu, Jan 29, 2009 at 12:46 PM, P.G. Richardson
> <p.g.richardson@phantomjinx.co.uk> wrote:
>> id: 0x8189 chipset: rtl8187b
>
> That's an older rtl8187b - there is no point trying mainline kernels,
> as the interesting development are happening
> in wireless-testing and takes a few months to get to mainline. i.e.
> trying 2.6.29 won't help. Can you download the current compat-wireles=
s
> tar ball (http://linuxwireless.org/en/users/Download) and give it a
> try?
>
> Also can you show your "dmesg | grep rtl" result ?
>
> The "can't ping" , etc behavior with multiple network interfaces is
> normal and quite simple to work around - you need to read up on
> "default route" and the "route" command.
>
> Are you using NetworkManager or configuring by hand? If you are not
> using NetworkManager, you need to set up dhcp manually. On 2nd
> thought, I am suprised you have problem with ping, because if you are
> using NetworkManager it should switch the default route, etc if you
> choose to switch to the wired interface.
>
> (You have come to the right place - Larry and Herton and I are the
> current maintainers of the rtl8187 driver).
> --
> To unsubscribe from this list: send the line "unsubscribe linux-wirel=
ess"
> in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>


--=20
Laws are partly formed for the sake of good men, in order to instruct t=
hem
how they may live on friendly terms with one another, and partly for th=
e
sake of those who refuse to be instructed, whose spirit cannot be subdu=
ed,
or softened, or hindered from plunging into evil. [The Laws, Plato]

You fiend! Never have I encountered such corrupt and foul-minded
perversity=85 Have you ever considered a career in the church?
Bishop of Bath and Wells (Blackadder II)

--
To unsubscribe from this list: send the line "unsubscribe linux-wireles=
s" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: rtl8187 wireless issues

[Hin-Tak Leung]

On Fri, Jan 30, 2009 at 1:59 PM, P.G. Richardson
<p.g.richardson@phantomjinx.co.uk> wrote:
> Hi,
>
> Have compiled and install the latest driver and success!

Glad to hear.

> Just to answer your questions ...
>
> dmesg didnt offer much information other than "link is not ready"

No, my asking for "dmesg | grep rtl " wasn't for disagnostic purposes.
There are 4-5 variants of the rtl8187b hardware, and they shows
slightly different parts. e.g. mine says:

phy0: hwaddr <address>, RTL8187BvE V0 + rtl8225z2

and yours probably will say 8187BvB with a B at the end, and some
other for the 2nd part. I just like the info for curiosity, since you
have a different rtl8187b from mine.

> Should have explained the detail of "cannot ping". Essentially I take the
> network down doing /etc/init.d/network stop. Change around which
> ifcfg-interfaces start on boot then do /etc/init.d/network start. Thus,
> the default route was always the active interface.
>
> Do not use NetworkManager and use static ip addresses.

NetworkManager gives you the ability to switch interfaces quite easily
and automatically and somewhat painlessly (when it works). If you want
to have multiple interfaces active and working simultaneously, you
need to learn a bit about routing, and do some homework. The issue you
experienced is not specific to wireless but you would have the same
issue with multiple wired ethernet interfaces; and somewhat
irrelevant.

> Anyway, hopefully everything is working.
>
> Thanks for the response. Let me know if you require any more info or would
> like some testing done.

If it breaks or not working well, you can come back :-). What's in
compat-wireless now will eventually end up in mainline, but meanwhile
if you upgrade your kernel and find that you need compat-wireless, it
would be nice to know just for the record that "distribution X is not
quite up-to-date". (some distros like Fedora do selective back-ports
of some drivers or some fixes). Support of rtl8187b in 2.6.27 wasn't
very good, but the changes in wireless-testing/compat-wireless is
gradually being pushed into mainline; what is in compat-wireless today
is mostly 2.6.30 material.

> PS. A minor thing which is probably my fault. I compiled the driver using
> make then ran make install. The latter installed the new mac82011, ifcfg
> and b44 drivers. None of the other specific wireless drivers were copied
> to the kernel update directory. I had to copy the rtl8187 there manually.

That's wierd - are you sure? Compat-wireless is supposed to put the
whole lot (about 50MB of kernel modules, nearly 50 of them) into
updates, although only about 4-5 kernel modules are used for rtl8187b.

Re: rtl8187 wireless issues

[P.G. Richardson]

ok,

Actually dmesg reports the same as you

phy0: hwaddr f729a040, RTL8187BvE V0 + rtl8225z2
usbcore: registered new interface driver rtl8187

NetworkManager advice taken with thanks.

>
> That's wierd - are you sure? Compat-wireless is supposed to put the
> whole lot (about 50MB of kernel modules, nearly 50 of them) into
> updates, although only about 4-5 kernel modules are used for rtl8187b=
=2E
>

After running make uninstall and make install a few times, this is the
structure of files I end with in /lib/modules/`uname -r`.

updates
updates/net
updates/net/mac80211
updates/net/mac80211/mac80211.ko
updates/net/wireless
updates/net/wireless/ifcfg80211.ko
updates/net/wireless/lib80211_crypt_tkip.ko
updates/net/wireless/lib80211.ko
updates/net/wireless/lib80211_crypt_ccmp.ko
updates/net/wireless/lib80211_crypt_wep.ko
updates/drivers
updates/drivers/net
updates/drivers/net/b44.ko
updates/drivers/net/wireless
updates/drivers/net/usb
updates/drivers/net/usb/rndis_host.ko
updates/drivers/net/usb/usbnet.ko
updates/drivers/net/usb/cdc_ether.ko
updates/drivers/misc
updates/drivers/misc/eeprom
updates/drivers/misc/eeprom/eeprom_93cx6.ko
updates/drivers/ssb
updates/drivers/ssb/ssb.ko

I manually added in to complete it:
updates/drivers/net/wireless/rtl818x
updates/drivers/net/wireless/rtl818x/rtl8187.ko

I am not that familiar with kernel makefiles but it would seem that the
drivers/net/wireless line below is not recursive:

obj-y :=3D net/wireless/ net/mac80211/
ifeq ($(ONLY_CORE),)
obj-$(CONFIG_B44) +=3D drivers/net/b44.o
obj-y +=3D drivers/ssb/ \
        drivers/misc/eeprom/ \
        drivers/net/usb/ \
        drivers/net/wireless/
endif

Anyway, its only a minor thing.

Cheers

PGR



--=20
Laws are partly formed for the sake of good men, in order to instruct t=
hem
how they may live on friendly terms with one another, and partly for th=
e
sake of those who refuse to be instructed, whose spirit cannot be subdu=
ed,
or softened, or hindered from plunging into evil. [The Laws, Plato]

You fiend! Never have I encountered such corrupt and foul-minded
perversity=85 Have you ever considered a career in the church?
Bishop of Bath and Wells (Blackadder II)

--
To unsubscribe from this list: send the line "unsubscribe linux-wireles=
s" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: rtl8187 wireless issues

[Hin-Tak Leung]

On Fri, Jan 30, 2009 at 10:04 PM, P.G. Richardson
<p.g.richardson@phantomjinx.co.uk> wrote:

> I am not that familiar with kernel makefiles but it would seem that the
> drivers/net/wireless line below is not recursive:
>
> obj-y := net/wireless/ net/mac80211/
> ifeq ($(ONLY_CORE),)
> obj-$(CONFIG_B44) += drivers/net/b44.o
> obj-y += drivers/ssb/ \
>        drivers/misc/eeprom/ \
>        drivers/net/usb/ \
>        drivers/net/wireless/
> endif

It is recursive - it runs Makefile in drivers/net/wireless/ . Your
list is very small - it should build and install 40+ kernel modules.
curious. some kernel config means the modules aren't built, but in
your case they are built but not installed... I wonder if it is
affected by blacklist.

Re: rtl8187 wireless issues

[P.G. Richardson]

That is a fair possibility. I upgraded the kernel on Saturday hence I h=
ad
to rebuild against the new kernel. Upon install, I noticed that rtl8187
was correctly installed. Of course, this second time of building and
installing, I had removed the blacklist entries for rtl8187 as I had
successfully used it for 24 hours.

One minor point, I notice in iwconfig is that it doesnt report a bit ra=
te
of greater than 1Mb/sec. Is this a known issue?

Regards

PGR

> On Fri, Jan 30, 2009 at 10:04 PM, P.G. Richardson
> <p.g.richardson@phantomjinx.co.uk> wrote:
>
>> I am not that familiar with kernel makefiles but it would seem that =
the
>> drivers/net/wireless line below is not recursive:
>>
>> obj-y :=3D net/wireless/ net/mac80211/
>> ifeq ($(ONLY_CORE),)
>> obj-$(CONFIG_B44) +=3D drivers/net/b44.o
>> obj-y +=3D drivers/ssb/ \
>>        drivers/misc/eeprom/ \
>>        drivers/net/usb/ \
>>        drivers/net/wireless/
>> endif
>
> It is recursive - it runs Makefile in drivers/net/wireless/ . Your
> list is very small - it should build and install 40+ kernel modules.
> curious. some kernel config means the modules aren't built, but in
> your case they are built but not installed... I wonder if it is
> affected by blacklist.
>


--=20
Laws are partly formed for the sake of good men, in order to instruct t=
hem
how they may live on friendly terms with one another, and partly for th=
e
sake of those who refuse to be instructed, whose spirit cannot be subdu=
ed,
or softened, or hindered from plunging into evil. [The Laws, Plato]

You fiend! Never have I encountered such corrupt and foul-minded
perversity=85 Have you ever considered a career in the church?
Bishop of Bath and Wells (Blackadder II)

--
To unsubscribe from this list: send the line "unsubscribe linux-wireles=
s" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: rtl8187 wireless issues

[Larry Finger]

P.G. Richardson wrote:
> That is a fair possibility. I upgraded the kernel on Saturday hence I had
> to rebuild against the new kernel. Upon install, I noticed that rtl8187
> was correctly installed. Of course, this second time of building and
> installing, I had removed the blacklist entries for rtl8187 as I had
> successfully used it for 24 hours.
> 
> One minor point, I notice in iwconfig is that it doesnt report a bit rate
> of greater than 1Mb/sec. Is this a known issue?

Please excuse my jumping into this thread, but the rate-setting
problem should be fixed in the latest compat-wireless. What version
are you using?

I just looked back at the patch log for wireless-testing. To get the
best code for the RTL8187, you should have compat-wireless from no
earlier than Jan. 28.

Larry

Re: rtl8187 wireless issues

[P.G. Richardson]

> P.G. Richardson wrote:
>> That is a fair possibility. I upgraded the kernel on Saturday hence =
I
>> had
>> to rebuild against the new kernel. Upon install, I noticed that rtl8=
187
>> was correctly installed. Of course, this second time of building and
>> installing, I had removed the blacklist entries for rtl8187 as I had
>> successfully used it for 24 hours.
>>
>> One minor point, I notice in iwconfig is that it doesnt report a bit
>> rate
>> of greater than 1Mb/sec. Is this a known issue?
>
> Please excuse my jumping into this thread, but the rate-setting
> problem should be fixed in the latest compat-wireless. What version
> are you using?
>
> I just looked back at the patch log for wireless-testing. To get the
> best code for the RTL8187, you should have compat-wireless from no
> earlier than Jan. 28.
>
> Larry

Thanks for letting me know. Compat-wireless dated 30/01/2009.

Regards

PGR

--=20
Laws are partly formed for the sake of good men, in order to instruct t=
hem
how they may live on friendly terms with one another, and partly for th=
e
sake of those who refuse to be instructed, whose spirit cannot be subdu=
ed,
or softened, or hindered from plunging into evil. [The Laws, Plato]

You fiend! Never have I encountered such corrupt and foul-minded
perversity=85 Have you ever considered a career in the church?
Bishop of Bath and Wells (Blackadder II)

--
To unsubscribe from this list: send the line "unsubscribe linux-wireles=
s" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: rtl8187 wireless issues

[Hin-Tak Leung]

On Mon, Feb 2, 2009 at 3:56 PM, P.G. Richardson
<p.g.richardson@phantomjinx.co.uk> wrote:

>> I just looked back at the patch log for wireless-testing. To get the
>> best code for the RTL8187, you should have compat-wireless from no
>> earlier than Jan. 28.
>>
>> Larry
>
> Thanks for letting me know. Compat-wireless dated 30/01/2009.

The "stuck-at-1Mb" problem happens a few times recently, for different
reasons - I suggest doing an "iwconfig rate <something>" and/or
"iwconfig rate <something> auto" to see if it works better. (the
latter means if it can take auto, so it is different from the former).
I say use the latest compat-wireless tarball whenever you need to
re-compile. If a problem persists, then ask. Note that if you let it
idle for a long time the rate change machanism will down-regulate.
(i.e. you should check *while* you are trying to push a lot of traffic
through).

About dates in wireless-testing log - I had an impression (I could be
wrong) that the dates are sometimes git-format-patch/posting dates and
not dates at which John commits to wireless-testing? (i.e. the dates
shown could be a lot earlier than time of commit, or time of commit is
not shown by the dates associated with a patch - I thought this makes
sense since the date of a patch should be the first time it appears in
the "system", not the first time it gets in-coporated in a particular
tree, and patches can go between trees and should *not* pick up date
change on the way).

Re: rtl8187 wireless issues

[John W. Linville]

On Mon, Feb 02, 2009 at 11:48:55PM +0000, Hin-Tak Leung wrote:

> About dates in wireless-testing log - I had an impression (I could be
> wrong) that the dates are sometimes git-format-patch/posting dates and
> not dates at which John commits to wireless-testing? (i.e. the dates
> shown could be a lot earlier than time of commit, or time of commit is
> not shown by the dates associated with a patch - I thought this makes
> sense since the date of a patch should be the first time it appears in
> the "system", not the first time it gets in-coporated in a particular
> tree, and patches can go between trees and should *not* pick up date
> change on the way).

Patch dates in 'git log' are based on their reported creation time or
(for patches not from git) their posting time.  That's just the way
git works (at least by default).

I generally tag wireless-testing updates on the master branch with
a date-based tag.  You can see them like this:

	git fetch --tags # just in case...
	git tag | grep master

Hth!

John
-- 
John W. Linville		Someday the world will need a hero, and you
linville@tuxdriver.com			might be all we have.  Be ready.

Re: rtl8187 wireless issues

[P.G. Richardson]

> On Mon, Feb 02, 2009 at 11:48:55PM +0000, Hin-Tak Leung wrote:
>
>> About dates in wireless-testing log - I had an impression (I could b=
e
>> wrong) that the dates are sometimes git-format-patch/posting dates a=
nd
>> not dates at which John commits to wireless-testing? (i.e. the dates
>> shown could be a lot earlier than time of commit, or time of commit =
is
>> not shown by the dates associated with a patch - I thought this make=
s
>> sense since the date of a patch should be the first time it appears =
in
>> the "system", not the first time it gets in-coporated in a particula=
r
>> tree, and patches can go between trees and should *not* pick up date
>> change on the way).
>
> Patch dates in 'git log' are based on their reported creation time or
> (for patches not from git) their posting time.  That's just the way
> git works (at least by default).
>
> I generally tag wireless-testing updates on the master branch with
> a date-based tag.  You can see them like this:
>
> 	git fetch --tags # just in case...
> 	git tag | grep master
>
> Hth!

Apologies. there might be some confusion. When I mentioned the date of =
the
30/01/2009, I was referring to the snapshot date of the tarball
downloaded, ie. compat-wireless-$(date -I).tar.bz2.

I haven't used wireless-testing's git repo thus far. If it will help wi=
th
the rate issue I have then will definitely install that version instead=
=2E

Cheers

PGR



--=20
Laws are partly formed for the sake of good men, in order to instruct t=
hem
how they may live on friendly terms with one another, and partly for th=
e
sake of those who refuse to be instructed, whose spirit cannot be subdu=
ed,
or softened, or hindered from plunging into evil. [The Laws, Plato]

You fiend! Never have I encountered such corrupt and foul-minded
perversity=85 Have you ever considered a career in the church?
Bishop of Bath and Wells (Blackadder II)

--
To unsubscribe from this list: send the line "unsubscribe linux-wireles=
s" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: rtl8187 wireless issues

[Larry Finger]

P.G. Richardson wrote:
> 
> Apologies. there might be some confusion. When I mentioned the date of the
> 30/01/2009, I was referring to the snapshot date of the tarball
> downloaded, ie. compat-wireless-$(date -I).tar.bz2.
> 
> I haven't used wireless-testing's git repo thus far. If it will help with
> the rate issue I have then will definitely install that version instead.

Using the wireless-testing git tree would always keep your wireless
with the latest (and likely the greatest), that would probably be
overkill for you. I would just grab the latest snapshot from
compat-wireless. Something from 02/02 would have the latest rtl8187 code.

Larry

rtl8187 sleeping/hanging

[P.G. Richardson]

Hi folks,

Installed current snapshot -> compat-wireless-2009-02-05

Still trying to get to grips with the rtl8187 driver. Not sure at the
moment whether my current problems are the fault of the laptop hardware=
 or
the driver. Maybe someone who has a similar wireless chip would care to
compare notes.

My router is situated upstairs. I have, this week, changed it to a new =
one
with a 5db gain level antenna rather than the original that had a 2db.

I have no problems with my old laptop using the ipw2200 driver connecti=
ng
to either of my routers. I can walk anywhere in the house with doors sh=
ut
and still get a solid ping.

Sadly, it seems my new laptop is not so reliable...
- Moving around when constantly pinging causes the pinging to halt,
missing packets.
- If I dont use the network connection for a few minutes (think it tend=
s
to be when on battery but not definitely tested), it seems the wlan0 di=
es
/ goes to sleep / hangs. Not sure which. Sometimes it comes back with a
quick ping to the router. Other occasions I have to re-modprobe the
driver.

I thought it might be the hardware is not as powerful as my old laptop =
but
iwconfig reports both have a 20db tx value so figured they are connecti=
ng
with the same transmit power.

Wondered if it might be to do with cpu frequency scaling as the default
kde4 powersave settings have the following settings:
- AC adapter: cpu freq scaling =3D performance
- Battery: cpu freq scaling =3D dynamic (less aggressive)
(Changed Battery to performance to see if this makes any difference)

Either way if people have any ideas they could throw at me then I would=
 be
much appreciated as doing anything networky such as a quick google is
turning into a full blown network marathon!

Distro: Mandriva 2009
Kernel: 2.6.27.10-desktop586-1mnb
IP Address: static
Network: handled by ifcfg script
Wireless: WPA2-PSK using TKIP
Channel: Tried both 11 and 8 and has no affect

Happy to provide any more info as required.

Regards

PGR

--=20
Laws are partly formed for the sake of good men, in order to instruct t=
hem
how they may live on friendly terms with one another, and partly for th=
e
sake of those who refuse to be instructed, whose spirit cannot be subdu=
ed,
or softened, or hindered from plunging into evil. [The Laws, Plato]

You fiend! Never have I encountered such corrupt and foul-minded
perversity=85 Have you ever considered a career in the church?
Bishop of Bath and Wells (Blackadder II)

--
To unsubscribe from this list: send the line "unsubscribe linux-wireles=
s" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: rtl8187 sleeping/hanging

[Larry Finger]

P.G. Richardson wrote:
> Sadly, it seems my new laptop is not so reliable...
> - Moving around when constantly pinging causes the pinging to halt,
> missing packets.
> - If I dont use the network connection for a few minutes (think it tends
> to be when on battery but not definitely tested), it seems the wlan0 dies
> / goes to sleep / hangs. Not sure which. Sometimes it comes back with a
> quick ping to the router. Other occasions I have to re-modprobe the
> driver.

On my RTL8187, this does not happen.

> I thought it might be the hardware is not as powerful as my old laptop but
> iwconfig reports both have a 20db tx value so figured they are connecting
> with the same transmit power.

That number comes from mac80211 and is set by the regulatory data
base. It has no relation to the actual power being transmitted!!!

> Wondered if it might be to do with cpu frequency scaling as the default
> kde4 powersave settings have the following settings:
> - AC adapter: cpu freq scaling = performance
> - Battery: cpu freq scaling = dynamic (less aggressive)
> (Changed Battery to performance to see if this makes any difference)

Shouldn't affect the results.

One thing you might do is get the vendor driver for the RTL8187L from
the Realtek site, compile it (good luck), and try it. There may be
differences between their version and the in-kernel driver. Some
8187's work great (mine is one of them), and others do not.
Unfortunately, there is nothing to fix for my stick.

Larry

Re: rtl8187 sleeping/hanging

[P.G. Richardson]

Larry,

Thanks for the response. I am coming to the conclusion that it is the
transmission strength of the wireless. However, I am no expert in this.

Have put the old router downstairs and maintained a connection for the
rest of the evening. Clearly it would appear that the laptop is not
picking up the router's signal as effectively as my old laptop. Only
conclusion since both laptops in the same location and 2 different rout=
ers
tried.

The 20db reported in iwconfig by mac80211 is not the actual data, is th=
e
signal strength please? I can see the signal strength change upon repea=
ted
calls to iwconfig and downstairs even when seemingly "out of range" and
failing to ping anything, it still reports a signal strength of above 5=
0%.

Since there is nothing to fix per se with the driver, is there anything
further to interrogate to eek out any info on what is going on when the
network seemingly drops?

Cheers for your help.

PGR

> P.G. Richardson wrote:
>> Sadly, it seems my new laptop is not so reliable...
>> - Moving around when constantly pinging causes the pinging to halt,
>> missing packets.
>> - If I dont use the network connection for a few minutes (think it t=
ends
>> to be when on battery but not definitely tested), it seems the wlan0
>> dies
>> / goes to sleep / hangs. Not sure which. Sometimes it comes back wit=
h a
>> quick ping to the router. Other occasions I have to re-modprobe the
>> driver.
>
> On my RTL8187, this does not happen.
>
>> I thought it might be the hardware is not as powerful as my old lapt=
op
>> but
>> iwconfig reports both have a 20db tx value so figured they are
>> connecting
>> with the same transmit power.
>
> That number comes from mac80211 and is set by the regulatory data
> base. It has no relation to the actual power being transmitted!!!
>
>> Wondered if it might be to do with cpu frequency scaling as the defa=
ult
>> kde4 powersave settings have the following settings:
>> - AC adapter: cpu freq scaling =3D performance
>> - Battery: cpu freq scaling =3D dynamic (less aggressive)
>> (Changed Battery to performance to see if this makes any difference)
>
> Shouldn't affect the results.
>
> One thing you might do is get the vendor driver for the RTL8187L from
> the Realtek site, compile it (good luck), and try it. There may be
> differences between their version and the in-kernel driver. Some
> 8187's work great (mine is one of them), and others do not.
> Unfortunately, there is nothing to fix for my stick.
>
> Larry
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-wirel=
ess"
> in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>


--=20
Laws are partly formed for the sake of good men, in order to instruct t=
hem
how they may live on friendly terms with one another, and partly for th=
e
sake of those who refuse to be instructed, whose spirit cannot be subdu=
ed,
or softened, or hindered from plunging into evil. [The Laws, Plato]

You fiend! Never have I encountered such corrupt and foul-minded
perversity=85 Have you ever considered a career in the church?
Bishop of Bath and Wells (Blackadder II)

--
To unsubscribe from this list: send the line "unsubscribe linux-wireles=
s" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: rtl8187 sleeping/hanging

[Larry Finger]

P.G. Richardson wrote:

> The 20db reported in iwconfig by mac80211 is not the actual data, is the
> signal strength please? I can see the signal strength change upon repeated
> calls to iwconfig and downstairs even when seemingly "out of range" and
> failing to ping anything, it still reports a signal strength of above 50%.

Signal strength may have even less physical meaning that the power.
Some devices, such as those by Broadcom, report the received signal
strength with some accuracy. The RTL8187B is supposed to have such an
output; however, my testing showed it to be essentially random. The
only reported quantity that could be used was the automatic gain
control setting of the receiver. This was calibrated against the
BCM4311 at two points, assumed to be linear, and the signal strength
and signal quality numbers were derived. For the RTL8187, a similar
scheme was used. As you can tell, only relative values for these
quantities are meaningful. The fact that your signal strength is above
50% means very little.

Larry

Re: rtl8187 sleeping/hanging

[P.G. Richardson]

> P.G. Richardson wrote:
>
>> The 20db reported in iwconfig by mac80211 is not the actual data, is=
 the
>> signal strength please? I can see the signal strength change upon
>> repeated
>> calls to iwconfig and downstairs even when seemingly "out of range" =
and
>> failing to ping anything, it still reports a signal strength of abov=
e
>> 50%.
>
> Signal strength may have even less physical meaning that the power.
> Some devices, such as those by Broadcom, report the received signal
> strength with some accuracy. The RTL8187B is supposed to have such an
> output; however, my testing showed it to be essentially random. The
> only reported quantity that could be used was the automatic gain
> control setting of the receiver. This was calibrated against the
> BCM4311 at two points, assumed to be linear, and the signal strength
> and signal quality numbers were derived. For the RTL8187, a similar
> scheme was used. As you can tell, only relative values for these
> quantities are meaningful. The fact that your signal strength is abov=
e
> 50% means very little.

Good to know.

With a repeater in the living room now, I have 100/100 signal strength =
and
fewer problems with connectivity. Not perfect I think due to the positi=
on
of the router. Still network connectivity improving.

Thanks for the info.

PGR

--=20
Laws are partly formed for the sake of good men, in order to instruct t=
hem
how they may live on friendly terms with one another, and partly for th=
e
sake of those who refuse to be instructed, whose spirit cannot be subdu=
ed,
or softened, or hindered from plunging into evil. [The Laws, Plato]

You fiend! Never have I encountered such corrupt and foul-minded
perversity=85 Have you ever considered a career in the church?
Bishop of Bath and Wells (Blackadder II)

--
To unsubscribe from this list: send the line "unsubscribe linux-wireles=
s" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html