[ath9k-devel] Fully control ethernet frame

[ath9k-devel] Fully control ethernet frame

[김상혁]

Hi all,

I want to make corrupted packet by wrong CRC/FCS

I have two questions.

1) Can I handle CRC/FCS when ethernet frame be made at Tx ?

2-1) If can, how can I ? or where is related source?

2-2) If can?t, is there any way to control CRC/FCS at Tx ?

Thanks.

[ath9k-devel] Fully control ethernet frame

[Adrian Chadd]

hi,

There's a transmit descriptor bit, AR_CorruptFCS, which you can set
and it will negate the FCS so it always fails.

Look in the ar9002_mac.c / ar9003_mac.c where NoAck is set; i think
AR_CorruptFCS should also be set in that (ctl1 for ar9002 and below,
ctl12 for ar9003)


-a


On 8 July 2016 at 00:08, ??? <tkdgur7896@gmail.com> wrote:
> Hi all,
>
> I want to make corrupted packet by wrong CRC/FCS
>
> I have two questions.
>
> 1) Can I handle CRC/FCS when ethernet frame be made at Tx ?
>
> 2-1) If can, how can I ? or where is related source?
>
> 2-2) If can?t, is there any way to control CRC/FCS at Tx ?
>
> Thanks.
> _______________________________________________
> ath9k-devel mailing list
> ath9k-devel at lists.ath9k.org
> https://lists.ath9k.org/mailman/listinfo/ath9k-devel

[ath9k-devel] Fully control ethernet frame

[Ben Greear]

On 07/08/2016 10:46 AM, Adrian Chadd wrote:
> hi,
>
> There's a transmit descriptor bit, AR_CorruptFCS, which you can set
> and it will negate the FCS so it always fails.
>
> Look in the ar9002_mac.c / ar9003_mac.c where NoAck is set; i think
> AR_CorruptFCS should also be set in that (ctl1 for ar9002 and below,
> ctl12 for ar9003)

Neat.

e1000e (I think?) and some other drivers I mucked with long ago have
the ability to accept FCS from user-space.  If someone adds that
same support to ath9k then that would make applications supporting
this API able to work on ath9k devices as well...

Thanks,
Ben

>
>
> -a
>
>
> On 8 July 2016 at 00:08, ??? <tkdgur7896@gmail.com> wrote:
>> Hi all,
>>
>> I want to make corrupted packet by wrong CRC/FCS
>>
>> I have two questions.
>>
>> 1) Can I handle CRC/FCS when ethernet frame be made at Tx ?
>>
>> 2-1) If can, how can I ? or where is related source?
>>
>> 2-2) If can?t, is there any way to control CRC/FCS at Tx ?
>>
>> Thanks.
>> _______________________________________________
>> ath9k-devel mailing list
>> ath9k-devel at lists.ath9k.org
>> https://lists.ath9k.org/mailman/listinfo/ath9k-devel
> _______________________________________________
> ath9k-devel mailing list
> ath9k-devel at lists.ath9k.org
> https://lists.ath9k.org/mailman/listinfo/ath9k-devel
>


-- 
Ben Greear <greearb@candelatech.com>
Candela Technologies Inc  http://www.candelatech.com

[ath9k-devel] Fully control ethernet frame

[Adrian Chadd]

On 8 July 2016 at 10:55, Ben Greear <greearb@candelatech.com> wrote:
> On 07/08/2016 10:46 AM, Adrian Chadd wrote:
>>
>> hi,
>>
>> There's a transmit descriptor bit, AR_CorruptFCS, which you can set
>> and it will negate the FCS so it always fails.
>>
>> Look in the ar9002_mac.c / ar9003_mac.c where NoAck is set; i think
>> AR_CorruptFCS should also be set in that (ctl1 for ar9002 and below,
>> ctl12 for ar9003)
>
>
> Neat.
>
> e1000e (I think?) and some other drivers I mucked with long ago have
> the ability to accept FCS from user-space.  If someone adds that
> same support to ath9k then that would make applications supporting
> this API able to work on ath9k devices as well...

I don't know if we can disable hardware FCS calculation. :-) It looks
like we can only make it be corrupted, for testing.



-adrian

>
> Thanks,
> Ben
>
>
>>
>>
>> -a
>>
>>
>> On 8 July 2016 at 00:08, ??? <tkdgur7896@gmail.com> wrote:
>>>
>>> Hi all,
>>>
>>> I want to make corrupted packet by wrong CRC/FCS
>>>
>>> I have two questions.
>>>
>>> 1) Can I handle CRC/FCS when ethernet frame be made at Tx ?
>>>
>>> 2-1) If can, how can I ? or where is related source?
>>>
>>> 2-2) If can?t, is there any way to control CRC/FCS at Tx ?
>>>
>>> Thanks.
>>> _______________________________________________
>>> ath9k-devel mailing list
>>> ath9k-devel at lists.ath9k.org
>>> https://lists.ath9k.org/mailman/listinfo/ath9k-devel
>>
>> _______________________________________________
>> ath9k-devel mailing list
>> ath9k-devel at lists.ath9k.org
>> https://lists.ath9k.org/mailman/listinfo/ath9k-devel
>>
>
>
> --
> Ben Greear <greearb@candelatech.com>
> Candela Technologies Inc  http://www.candelatech.com
>

[ath9k-devel] Fully control ethernet frame

[SangHyuk Kim]

Dear Adrian.

Thanks for your reply.

I could check the point you mentioned, but when I tried to modify source
code(ar9002.c) and rebuild it didn't any effects.

And I realized that my wifi usb uses ath9k_htc things.

So I'm researching about htc_*.c things, but I can't find any crew clue.

My new questions :

1) How can I corrupt CRC/FCS using htc driver product
2) Also, what is the 'txok' variable in htc_drv_txrx.c :

[image: ?? ??? 1]

Any replies, thanks.

2016-07-09 8:56 GMT+09:00 Adrian Chadd <adrian@freebsd.org>:

> On 8 July 2016 at 10:55, Ben Greear <greearb@candelatech.com> wrote:
> > On 07/08/2016 10:46 AM, Adrian Chadd wrote:
> >>
> >> hi,
> >>
> >> There's a transmit descriptor bit, AR_CorruptFCS, which you can set
> >> and it will negate the FCS so it always fails.
> >>
> >> Look in the ar9002_mac.c / ar9003_mac.c where NoAck is set; i think
> >> AR_CorruptFCS should also be set in that (ctl1 for ar9002 and below,
> >> ctl12 for ar9003)
> >
> >
> > Neat.
> >
> > e1000e (I think?) and some other drivers I mucked with long ago have
> > the ability to accept FCS from user-space.  If someone adds that
> > same support to ath9k then that would make applications supporting
> > this API able to work on ath9k devices as well...
>
> I don't know if we can disable hardware FCS calculation. :-) It looks
> like we can only make it be corrupted, for testing.
>
>
>
> -adrian
>
> >
> > Thanks,
> > Ben
> >
> >
> >>
> >>
> >> -a
> >>
> >>
> >> On 8 July 2016 at 00:08, ??? <tkdgur7896@gmail.com> wrote:
> >>>
> >>> Hi all,
> >>>
> >>> I want to make corrupted packet by wrong CRC/FCS
> >>>
> >>> I have two questions.
> >>>
> >>> 1) Can I handle CRC/FCS when ethernet frame be made at Tx ?
> >>>
> >>> 2-1) If can, how can I ? or where is related source?
> >>>
> >>> 2-2) If can?t, is there any way to control CRC/FCS at Tx ?
> >>>
> >>> Thanks.
> >>> _______________________________________________
> >>> ath9k-devel mailing list
> >>> ath9k-devel at lists.ath9k.org
> >>> https://lists.ath9k.org/mailman/listinfo/ath9k-devel
> >>
> >> _______________________________________________
> >> ath9k-devel mailing list
> >> ath9k-devel at lists.ath9k.org
> >> https://lists.ath9k.org/mailman/listinfo/ath9k-devel
> >>
> >
> >
> > --
> > Ben Greear <greearb@candelatech.com>
> > Candela Technologies Inc  http://www.candelatech.com
> >
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.ath9k.org/pipermail/ath9k-devel/attachments/20160713/29e0cf54/attachment-0001.htm 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image.png
Type: image/png
Size: 108875 bytes
Desc: not available
Url : http://lists.ath9k.org/pipermail/ath9k-devel/attachments/20160713/29e0cf54/attachment-0001.png 

[ath9k-devel] Fully control ethernet frame

[Adrian Chadd]

hi,

for ath9k_htc devices you have to hack the TX descriptor setup in the
firmware, rather than in ath9k_htc.



-adrian


On 13 July 2016 at 01:52, SangHyuk Kim <tkdgur7896@gmail.com> wrote:

> Dear Adrian.
>
> Thanks for your reply.
>
> I could check the point you mentioned, but when I tried to modify source
> code(ar9002.c) and rebuild it didn't any effects.
>
> And I realized that my wifi usb uses ath9k_htc things.
>
> So I'm researching about htc_*.c things, but I can't find any crew clue.
>
> My new questions :
>
> 1) How can I corrupt CRC/FCS using htc driver product
> 2) Also, what is the 'txok' variable in htc_drv_txrx.c :
>
> [image: ?? ??? 1]
>
> Any replies, thanks.
>
> 2016-07-09 8:56 GMT+09:00 Adrian Chadd <adrian@freebsd.org>:
>
>> On 8 July 2016 at 10:55, Ben Greear <greearb@candelatech.com> wrote:
>> > On 07/08/2016 10:46 AM, Adrian Chadd wrote:
>> >>
>> >> hi,
>> >>
>> >> There's a transmit descriptor bit, AR_CorruptFCS, which you can set
>> >> and it will negate the FCS so it always fails.
>> >>
>> >> Look in the ar9002_mac.c / ar9003_mac.c where NoAck is set; i think
>> >> AR_CorruptFCS should also be set in that (ctl1 for ar9002 and below,
>> >> ctl12 for ar9003)
>> >
>> >
>> > Neat.
>> >
>> > e1000e (I think?) and some other drivers I mucked with long ago have
>> > the ability to accept FCS from user-space.  If someone adds that
>> > same support to ath9k then that would make applications supporting
>> > this API able to work on ath9k devices as well...
>>
>> I don't know if we can disable hardware FCS calculation. :-) It looks
>> like we can only make it be corrupted, for testing.
>>
>>
>>
>> -adrian
>>
>> >
>> > Thanks,
>> > Ben
>> >
>> >
>> >>
>> >>
>> >> -a
>> >>
>> >>
>> >> On 8 July 2016 at 00:08, ??? <tkdgur7896@gmail.com> wrote:
>> >>>
>> >>> Hi all,
>> >>>
>> >>> I want to make corrupted packet by wrong CRC/FCS
>> >>>
>> >>> I have two questions.
>> >>>
>> >>> 1) Can I handle CRC/FCS when ethernet frame be made at Tx ?
>> >>>
>> >>> 2-1) If can, how can I ? or where is related source?
>> >>>
>> >>> 2-2) If can?t, is there any way to control CRC/FCS at Tx ?
>> >>>
>> >>> Thanks.
>> >>> _______________________________________________
>> >>> ath9k-devel mailing list
>> >>> ath9k-devel at lists.ath9k.org
>> >>> https://lists.ath9k.org/mailman/listinfo/ath9k-devel
>> >>
>> >> _______________________________________________
>> >> ath9k-devel mailing list
>> >> ath9k-devel at lists.ath9k.org
>> >> https://lists.ath9k.org/mailman/listinfo/ath9k-devel
>> >>
>> >
>> >
>> > --
>> > Ben Greear <greearb@candelatech.com>
>> > Candela Technologies Inc  http://www.candelatech.com
>> >
>>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.ath9k.org/pipermail/ath9k-devel/attachments/20160713/a9d53913/attachment-0001.htm 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image.png
Type: image/png
Size: 108875 bytes
Desc: not available
Url : http://lists.ath9k.org/pipermail/ath9k-devel/attachments/20160713/a9d53913/attachment-0001.png