[Click] [RFC] Update radiotap parser
Roberto Riggio
roberto.riggio at create-net.org
Thu May 12 03:51:42 EDT 2011
Hi,
in this branch:
git://github.com/rriggio/click.git
I have put an updated version of the radiotap parser together with two
new elements
radiotaencapht and settxrateht. The latter can be used to specify the
MCS index on a
per-packet basis.
R.
Il 17/04/2011 16:34, 欧阳鑫 ha scritto:
>
>
> 在 2011年4月15日 下午4:55，Roberto Riggio
> <roberto.riggio at create-net.org <mailto:roberto.riggio at create-net.org>>
> 写 道：
>
> Il 15/04/2011 09:44, 欧阳鑫 ha scritto:
> >
> > Hi, I have patched my click source files according to above. I
> want to
> > implement click on 802.11n mode, but I find the above patch
> >
> > does not deal with MCS arguments such
> >
> IEEE80211_RADIOTAP_MCS_HAVE_MCS，IEEE80211_RADIOTAP_MCS_BW_40，IEEE80211_RADIOTAP_MCS_SGI
> >
> >
> well they are anyway ignored by an unpatched kernel
>
> I have updated wireless driver ath9k to compat-wireless-2011-03-16 and
> it has supported MCS arguments such as above.
>
> >
> > and so on. So I modify radiotapencap.cc, radiotapdecap.cc,
> > settxrate.cc, settxrate.hh, ettstat.cc and athdescdecap.cc. The
> patches
> >
> > are as follow:
> > diff -urN click-1.8.0_bg_patch/elements/roofnet/sr/ettstat.cc
> > click-1.8.0_patch_11n/elements/roofnet/sr/ettstat.cc
> > --- click-1.8.0_bg_patch/elements/roofnet/sr/ettstat.cc 2011-04-01
> > 15:09:40.000000000 +0800
> > +++ click-1.8.0_patch_11n/elements/roofnet/sr/ettstat.cc 2011-04-14
> > 15:59:10.252273517 +0800
> > @@ -420,7 +420,15 @@
> > struct click_wifi_extra *ceh = (struct click_wifi_extra *)
> p->user_anno();
> > ceh->magic = WIFI_EXTRA_MAGIC;
> > ceh->rate = rate;
> > -ceh->max_tries = 7;
> > +
> > + ceh->flags |= WIFI_EXTRA_RX_FLAG_HT;
> > + ceh->flags |= WIFI_EXTRA_RX_FLAG_40MHZ;
> > + ceh->flags |= WIFI_EXTRA_RX_FLAG_SHORT_GI;
> > + ceh->rateidx = _ads_rs_index;
> > + ceh->max_tries = 1;
> > +
> > checked_output_push(0, p);
> > }
>
> this should probably be part of the element configuration rather that
> hard coded.
>
> I do not understand completely. I have tried to get rid of this part
> in ettstat.cc and only add MCS arguments in settxrate.cc, but I find
> that the rate of patckets is only 0Mb/s, I do not know why.
>
>
> > diff -urN click-1.8.0_bg_patch/elements/wifi/athdescdecap.cc
> > click-1.8.0_patch_11n/elements/wifi/athdescdecap.cc
> > --- click-1.8.0_bg_patch/elements/wifi/athdescdecap.cc 2011-04-01
> > 15:09:40.000000000 +0800
> > +++ click-1.8.0_patch_11n/elements/wifi/athdescdecap.cc 2011-04-14
> > 10:36:26.468273490 +0800
> > @@ -30,7 +30,7 @@
> >
> > AthdescDecap::AthdescDecap()
> > {
> > - static_assert(WIFI_EXTRA_ANNO_SIZE >= sizeof(click_wifi_extra));
> > + //static_assert(WIFI_EXTRA_ANNO_SIZE >= sizeof(click_wifi_extra));
> > }
> >
>
> is there a particular reason to remove this? anyway the atheros
> descriptor is supported only by the madwifi
> driver which does not support 11n cards.
>
> Because I add one member uint8_t rateidx of click_wifi_extra, so the
> size of click_wifi_extra is 25, but WIFI_EXTRA_ANNO_SIZE is 24, the
> static_assert is not pass , so I remove this. I do not know if I
> remove this if it will have some effect on other functions.
>
> > AthdescDecap::~AthdescDecap()
> > diff -urN click-1.8.0_bg_patch/elements/wifi/radiotapdecap.cc
> > click-1.8.0_patch_11n/elements/wifi/radiotapdecap.cc
> > --- click-1.8.0_bg_patch/elements/wifi/radiotapdecap.cc 2011-04-12
> > 10:33:15.000000000 +0800
> > +++ click-1.8.0_patch_11n/elements/wifi/radiotapdecap.cc 2011-04-14
> > 16:35:28.604273504 +0800
> > @@ -26,6 +26,7 @@
> > #include <clicknet/radiotap.h>
> > #include <clicknet/llc.h>
> >
> > @@ -105,6 +106,26 @@
> >
> > CLICK_DECLS
> >
> > +#define IEEE80211_MCS_RATE_COUNT 16
> > +int IEEE80211_RADIOTAP_MCS_RATE[IEEE80211_MCS_RATE_COUNT][2] = {
> > + {0, 150},
> > + {1, 300},
> > + {2, 450},
> > + {3, 600},
> > + {4, 900},
> > + {5, 1200},
> > + {6, 1350},
> > + {7, 1500},
> > + {8, 300},
> > + {9, 600},
> > + {10, 900},
> > + {11, 1200},
> > + {12, 1800},
> > + {13, 2400},
> > + {14, 2700},
> > + {15, 3000}
> > +};
> > +
> I do not get this, the MCS index already embed the rate
> information. As
> a matter of fact
> the corresponding radiotap field shall containt the index NOT the
> corresponding PHY
> data rate:
>
> http://www.radiotap.org/defined-fields/MCS
>
> I add the data rate only in order to intuitively display data rate
> rather according to index converting to data rate.
>
> > RadiotapDecap::RadiotapDecap()
> > {
> > }
> > @@ -221,9 +242,36 @@
> > if (flags & IEEE80211_RADIOTAP_F_TX_FAIL)
> > ceh->flags |= WIFI_EXTRA_TX_FAIL;
> > break;
> > + case IEEE80211_RADIOTAP_MCS:
> > + uint8_t known, flags;
> > + known = *(iter.this_arg);
> > + flags = *(iter.this_arg + 1);
> > +
> > + // set Band width
> > + if (known & IEEE80211_RADIOTAP_MCS_HAVE_BW) {
> > + if (flags & IEEE80211_RADIOTAP_MCS_BW_40)
> > + ceh->flags |= WIFI_EXTRA_RX_FLAG_HT;
> > + }
> where is this WIFI_EXTRA_RX_FLAG_HT defined?
>
> > + if (known & IEEE80211_RADIOTAP_MCS_HAVE_MCS)
> > + ceh->rateidx = *(iter.this_arg + 2);
> > + else
> > + // may not reasonable......
> > + ceh->rateidx = 0;
> > +
> > + // set Guard interval
> > + if (known & IEEE80211_RADIOTAP_MCS_HAVE_GI) {
> > + if (flags & IEEE80211_RADIOTAP_MCS_SGI)
> > + ceh->flags |= WIFI_EXTRA_RX_FLAG_SHORT_GI;
> and this
> > + }
> > + // set FMT
> > + if (known & IEEE80211_RADIOTAP_MCS_HAVE_FMT) {
> > + if (flags & IEEE80211_RADIOTAP_MCS_FMT_GF)
> > + ceh->flags |= WIFI_EXTRA_RX_FLAG_LDPC;
> > + }
> > +
> same here
> >
> > + ceh->rate = IEEE80211_RADIOTAP_MCS_RATE[ceh->rateidx][1] / 5;
> > + break;
> > +
> ok, i get you want to set rate to the actual phy rate. why an
> actual use
> for it?
> >
> > - //if (flags & IEEE80211_RADIOTAP_F_FCS) {
> > - // p->take(4);
> so what happens if the frame actually include the FCS ?
> > }
> > }
> >
> > diff -urN click-1.8.0_bg_patch/elements/wifi/radiotapencap.cc
> > click-1.8.0_patch_11n/elements/wifi/radiotapencap.cc
> > --- click-1.8.0_bg_patch/elements/wifi/radiotapencap.cc 2011-04-12
> > 10:51:03.000000000 +0800
> > +++ click-1.8.0_patch_11n/elements/wifi/radiotapencap.cc 2011-04-14
> > 16:21:15.760273037 +0800
> > @@ -46,6 +46,14 @@
> > (1 << IEEE80211_RADIOTAP_EXT) | \
> > 0)
> >
> > +//566fix: mcs
> > +#define CLICK_RADIOTAP_MCS_PRESENT ( \
> > + (1 << IEEE80211_RADIOTAP_MCS) | \
> > + 0)
> > +
> > +#define IEEE80211_MCS_RATE_COUNT 16
> > +extern int
> IEEE80211_RADIOTAP_MCS_RATE[IEEE80211_MCS_RATE_COUNT][2];
> > +
> >
> > struct click_radiotap_header {
> > struct ieee80211_radiotap_header wt_ihdr;
> > @@ -66,6 +74,12 @@
> > u_int8_t wt_rts_retries;
> > u_int8_t wt_data_retries;
> >
> > + u_int8_t wt_known;
> > + u_int8_t wt_flags;
> > + u_int8_t wt_mcs;
> > +
> > u_int8_t wt_rate1;
> > u_int8_t wt_data_retries1;
> >
> > } __attribute__((__packed__));
> >
> >
> > @@ -206,6 +219,29 @@
> > }
> > }
> >
> > +
> > + if (ceh->flags & WIFI_EXTRA_RX_FLAG_HT) {
> > + crh->wt_ihdr.it_present |= CLICK_RADIOTAP_MCS_PRESENT;
> > + crh->wt_known = IEEE80211_RADIOTAP_MCS_HAVE_BW | \
> > + IEEE80211_RADIOTAP_MCS_HAVE_MCS | \
> > + IEEE80211_RADIOTAP_MCS_HAVE_GI | \
> > + IEEE80211_RADIOTAP_MCS_HAVE_FMT | \
> > + IEEE80211_RADIOTAP_MCS_HAVE_FEC;
> > + crh->wt_mcs = ceh->rateidx;
> > + if (ceh->flags & WIFI_EXTRA_RX_FLAG_40MHZ)
> > + crh->wt_flags |= IEEE80211_RADIOTAP_MCS_BW_40;
> > + if (ceh->flags & WIFI_EXTRA_RX_FLAG_SHORT_GI)
> > + crh->wt_flags |= IEEE80211_RADIOTAP_MCS_SGI;
> > + if (ceh->flags & WIFI_EXTRA_RX_FLAG_GF)
> > + crh->wt_flags |= IEEE80211_RADIOTAP_MCS_FMT_GF;
> > + crh->wt_rate = IEEE80211_RADIOTAP_MCS_RATE[ceh->rateidx][1] / 5;
> > +
> > + }
> > +
> >
> Here you add the three new MCS fields (known, flags, mcs), but the
> present bitmask is updated
> only if the WIFI_EXTRA_RX_FLAG_HT is set. As a result if
> WIFI_EXTRA_RX_FLAG_HT is not set
> the three fields are still there resulting in an invalid radiotap
> header.
>
> Finally, I would try to stabilize the radiotap parser and the
> decap/encap elements before
> adapting other elements such as ettstat, settxrate, etc.
>
> R.
> Click can run with 802.11b/g mode on high rates, such as 48Mb/s,
> 54Mb/s(the highest rate of 802.11g mode)and so on. But when I run
> click with 802.11n mode, and I add some sources dealing with MCS
> arguments as above, but the data rate is always 0Mbs/s, and the
> rateidx is always 0. I do not know if there is some important
> arguments but I do not deal with, so the rate is always 0Mb/s and
> rateidx is always 0. If it is, I should modify which sources of click?
> Thank you very much!
> Ouyang.
>
--
--------------------------------------------------------
Roberto Riggio, Ph.D.
CREATE-NET
Network & Security Solutions for Pervasive Computing Systems (iNSPIRE)
Senior Researcher
Via alla Cascata 56/D - 38123 Povo Trento (Italy)
e-mail: roberto.riggio at create-net.org
Tel: (+39) 0461 408400 - interno/extension 708
Fax: (+39) 0461 421157
www.create-net.org/~rriggio
--------------------------------------------------------
The information transmitted is intended only for the person or entity to
which it is addressed and may contain confidential and/or privileged
material. Any review, retransmission, dissemination or other use of, or
taking of any action in reliance upon, this information by persons or
entities other than the intended recipient is prohibited according to
the Italian Law 196/2003 of the Legislature. If you received this in
error, please contact the sender and delete the material from any
computer.
Le informazioni contenute in questo messaggio di posta elettronica e nei
file allegati sono da considerarsi strettamente riservate. Il loro
utilizzo e' consentito esclusivamente al destinatario del messaggio, per
le finalita' indicate nel messaggio stesso. Qualora riceveste questo
messaggio senza esserne il destinatario, Vi preghiamo cortesemente di
darcene notizia via e-mail e di procedere alla cancellazione del
messaggio stesso dal Vostro sistema. Trattenere il messaggio stesso,
divulgarlo anche in parte, distribuirlo ad altri soggetti, copiarlo,
od utilizzarlo per finalita' diverse, costituisce comportamento
contrario ai principi dettati dal D. Lgs. 196/2003.
More information about the click
mailing list