On Mon, Sep 26, 2011 at 4:59 PM, Guy, Wey-Yi <[email protected]> wrote:
> few things I will suggest and please give a try:
> 1. Please try to apply the following two patches I attach here.
> 2. load the iwlagn with debug=0x45800006, the log will be big, but it
> will contain more information on the tx/rx path.
Here's the log you asked for:
http://diff.cc/tmp/kernel_iwlagn_2.log.bz2
As I said, this problem does not occur with the Windows driver. You
said you are actively working on it. Does that mean, the problem is
known to you?
Cheers,
Henrik
On Mon, 2011-09-26 at 09:10 -0700, Henrik Friedrichsen wrote:
> I see. I hope the log is useful to you. I'm not sure, but this issue
> seems to be slightly different to your average performance issue. My
> Android device needs 1 second to completely load the Router webpanel,
> whereas my laptop with the Intel chip needs 30 seconds.
>
> I'm afraid the patches did not make any difference. Oh btw, I did
> apply them before making the log. I believe that was your intention.
>
> On Mon, Sep 26, 2011 at 5:19 PM, Guy, Wey-Yi <[email protected]> wrote:
> > On Mon, 2011-09-26 at 09:03 -0700, Henrik Friedrichsen wrote:
> >> On Mon, Sep 26, 2011 at 4:59 PM, Guy, Wey-Yi <[email protected]> wrote:
> >> > few things I will suggest and please give a try:
> >> > 1. Please try to apply the following two patches I attach here.
> >> > 2. load the iwlagn with debug=0x45800006, the log will be big, but it
> >> > will contain more information on the tx/rx path.
> >>
> >> Here's the log you asked for:
> >> http://diff.cc/tmp/kernel_iwlagn_2.log.bz2
> >>
> >> As I said, this problem does not occur with the Windows driver. You
> >> said you are actively working on it. Does that mean, the problem is
> >> known to you?
> >>
> >
> > we are seeing performance issue and still can not root cause the
> > problem. that is the reason I send you 2 patches for testing. We are
> > very active working on fix the performance problem. the symptoms might
> > be different, but we aware there were issues.
> >
btw, what if you using "Open/no security"?
Thanks
Wey
> >
> >
> >
On Mon, 2011-09-26 at 09:03 -0700, Henrik Friedrichsen wrote:
> On Mon, Sep 26, 2011 at 4:59 PM, Guy, Wey-Yi <[email protected]> wrote:
> > few things I will suggest and please give a try:
> > 1. Please try to apply the following two patches I attach here.
> > 2. load the iwlagn with debug=0x45800006, the log will be big, but it
> > will contain more information on the tx/rx path.
>
> Here's the log you asked for:
> http://diff.cc/tmp/kernel_iwlagn_2.log.bz2
>
> As I said, this problem does not occur with the Windows driver. You
> said you are actively working on it. Does that mean, the problem is
> known to you?
>
we are seeing performance issue and still can not root cause the
problem. that is the reason I send you 2 patches for testing. We are
very active working on fix the performance problem. the symptoms might
be different, but we aware there were issues.
Thx
Wey
Hey Meenakshi
On Mon, Sep 26, 2011 at 7:58 PM, Venkataraman, Meenakshi
<[email protected]> wrote:
> [MV] Just to verify that this is a 11n issue, you said you tried disabling the 11n mode?
Yes. I just did, once again, to verify it. No change, unfortunately :(
> Also, in your first email you mentioned that you were using 3.0.4 earlier? What were the results in that version?
It was the same. I was hoping a newer kernel might contain a fix of
some sort, that's why I updated it.
> I wonder if we're somehow using very low rates. Can you please send debug logs from your experiment with debug flags 0x143fff?
Here are the logs: http://diff.cc/tmp/kernel_iwlagn_3.log.bz2
They were made with "11n_disable=1" and without the patches posted by Wey-Yi.
Cheers,
Henrik
I see. I hope the log is useful to you. I'm not sure, but this issue
seems to be slightly different to your average performance issue. My
Android device needs 1 second to completely load the Router webpanel,
whereas my laptop with the Intel chip needs 30 seconds.
I'm afraid the patches did not make any difference. Oh btw, I did
apply them before making the log. I believe that was your intention.
On Mon, Sep 26, 2011 at 5:19 PM, Guy, Wey-Yi <[email protected]> wrote:
> On Mon, 2011-09-26 at 09:03 -0700, Henrik Friedrichsen wrote:
>> On Mon, Sep 26, 2011 at 4:59 PM, Guy, Wey-Yi <[email protected]> wrote:
>> > few things I will suggest and please give a try:
>> > 1. Please try to apply the following two patches I attach here.
>> > 2. load the iwlagn with debug=0x45800006, the log will be big, but it
>> > will contain more information on the tx/rx path.
>>
>> Here's the log you asked for:
>> http://diff.cc/tmp/kernel_iwlagn_2.log.bz2
>>
>> As I said, this problem does not occur with the Windows driver. You
>> said you are actively working on it. Does that mean, the problem is
>> known to you?
>>
>
> we are seeing performance issue and still can not root cause the
> problem. that is the reason I send you 2 patches for testing. We are
> very active working on fix the performance problem. the symptoms might
> be different, but we aware there were issues.
>
> Thx
> Wey
>
>
>
On Mon, Sep 26, 2011 at 6:05 PM, Guy, Wey-Yi <[email protected]> wrote:
> btw, what if you using "Open/no security"?
Same situation :\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Hi Henrik,
On Mon, 2011-09-26 at 11:26 -0700, Henrik Friedrichsen wrote:
> Hey Meenakshi
>
> On Mon, Sep 26, 2011 at 7:58 PM, Venkataraman, Meenakshi
> <[email protected]> wrote:
> > [MV] Just to verify that this is a 11n issue, you said you tried disabling the 11n mode?
> Yes. I just did, once again, to verify it. No change, unfortunately :(
>
> > Also, in your first email you mentioned that you were using 3.0.4 earlier? What were the results in that version?
> It was the same. I was hoping a newer kernel might contain a fix of
> some sort, that's why I updated it.
>
> > I wonder if we're somehow using very low rates. Can you please send debug logs from your experiment with debug flags 0x143fff?
> Here are the logs: http://diff.cc/tmp/kernel_iwlagn_3.log.bz2
> They were made with "11n_disable=1" and without the patches posted by Wey-Yi.
>
One experiment if you can help us to test.
please reset the repository to
commit#1f7b6172db86e9ab2b4cd794441bb2c40ab287fc
and give a try to see if you still seeing the issue.
Thanks
Wey
On Mon, Sep 26, 2011 at 8:33 PM, Guy, Wey-Yi <[email protected]> wrote:
> please reset the repository to
> commit#1f7b6172db86e9ab2b4cd794441bb2c40ab287fc
>
> and give a try to see if you still seeing the issue.
No noticeable difference :(
On Mon, 2011-09-26 at 09:10 -0700, Henrik Friedrichsen wrote:
> I see. I hope the log is useful to you. I'm not sure, but this issue
> seems to be slightly different to your average performance issue. My
> Android device needs 1 second to completely load the Router webpanel,
> whereas my laptop with the Intel chip needs 30 seconds.
>
> I'm afraid the patches did not make any difference. Oh btw, I did
> apply them before making the log. I believe that was your intention.
>
> On Mon, Sep 26, 2011 at 5:19 PM, Guy, Wey-Yi <[email protected]> wrote:
> > On Mon, 2011-09-26 at 09:03 -0700, Henrik Friedrichsen wrote:
> >> On Mon, Sep 26, 2011 at 4:59 PM, Guy, Wey-Yi <[email protected]> wrote:
> >> > few things I will suggest and please give a try:
> >> > 1. Please try to apply the following two patches I attach here.
> >> > 2. load the iwlagn with debug=0x45800006, the log will be big, but it
> >> > will contain more information on the tx/rx path.
> >>
> >> Here's the log you asked for:
> >> http://diff.cc/tmp/kernel_iwlagn_2.log.bz2
> >>
> >> As I said, this problem does not occur with the Windows driver. You
> >> said you are actively working on it. Does that mean, the problem is
> >> known to you?
> >>
> >
> > we are seeing performance issue and still can not root cause the
> > problem. that is the reason I send you 2 patches for testing. We are
> > very active working on fix the performance problem. the symptoms might
> > be different, but we aware there were issues.
> >
thanks a lot, 30 seconds is really bad :-(, working on it!!!
Wey
> >
> >
> >