Skip to contentSkip to conversation

add basic support for parsing country information element

PATCH In review Marcel Holtmann 2 participants 14 messages May 4, 2009

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 add basic support for parsing country information element

---
 scan.c |   11 +++++++++++
 1 files changed, 11 insertions(+), 0 deletions(-)
scan.c +11 -0
index 2976115..c503487 100644
@@ -91,6 +91,16 @@ static void print_ign(unsigned char type, unsigned char len, unsigned char *data
 	/* ignore for now, not too useful */
 }
 
+static void print_country(unsigned char type, unsigned char len, unsigned char *data)
+{
+	int i;
+
+	printf("\tCountry: %.*s data:", 3, data);
+	for(i=0; i<len-3; i++)
+		printf(" %.02x", data[i + 3]);
+	printf("\n");
+}
+
 static void print_erp(unsigned char type, unsigned char len, unsigned char *data)
 {
 	if (data[0] == 0x00)
@@ -245,6 +255,7 @@ static const printfn ieprinters[] = {
 	[1] = print_supprates,
 	[3] = print_ds,
 	[5] = print_ign,
+	[7] = print_country,
 	[42] = print_erp,
 	[48] = print_rsn,
 	[50] = print_supprates,
Signature · 1 line
1.6.0.6
#1 lore raw
Quoted 9 lines · Marcel Holtmann
---
 scan.c |   11 +++++++++++
 1 files changed, 11 insertions(+), 0 deletions(-)
scan.c +0 -0
index 2976115..c503487 100644

scan.c static void print_ign(unsigned char type, unsigned char len, unsigned char *data · line 98

Quoted 8 lines
scan.c +5 -0
@@ -91,6 +91,16 @@ static void print_ign(unsigned char type, unsigned char len, unsigned char *data
 	/* ignore for now, not too useful */
 }
 
+static void print_country(unsigned char type, unsigned char len, unsigned char *data)
+{
+	int i;
+
+	printf("\tCountry: %.*s data:", 3, data);

IIRC the third characters there is a space if it's in- and outdoor, it might make sense to parse it instead of just printing it?

johannes

#3 lore raw

Hi Johannes,

Quoted 1 line

Everything else looks good.

I re-sent the patches now in a different order so that the WPA stuff is the latest in case you want extra fixes. I didn't include all defined ciphers since I would have to dig though specs again. However it should now be pretty easy to extend it since the basic parsing is available.

This bring iw scan at the same stage what we had with iwlist and so another piece to deprecate WEXT.

Regards

Marcel

#4 lore raw

Hi Marcel,

Quoted 4 lines

I re-sent the patches now in a different order so that the WPA stuff is the latest in case you want extra fixes. I didn't include all defined ciphers since I would have to dig though specs again. However it should now be pretty easy to extend it since the basic parsing is available.

I've committed everything, and am doing some minor cleanups. I also noticed we don't validate everything as good as we should so I'm adding some support for that.

johannes

#5 lore raw

Marcel,

Quoted 3 lines

I've committed everything, and am doing some minor cleanups. I also noticed we don't validate everything as good as we should so I'm adding some support for that.

I've done that now, and pushed it out, but I think the RSN IE printing you added (and I renamed to print_rsn_ie) is missing some input length validation, could you take a look at that please? In particular it takes the count from the IE, but doesn't verify it fits, I think?

johannes

#6 lore raw

Hi Johannes,

Quoted 8 lines
Quoted 3 lines

I've committed everything, and am doing some minor cleanups. I also noticed we don't validate everything as good as we should so I'm adding some support for that.

I've done that now, and pushed it out, but I think the RSN IE printing you added (and I renamed to print_rsn_ie) is missing some input length validation, could you take a look at that please? In particular it takes the count from the IE, but doesn't verify it fits, I think?

I can have a look at it since it is by no means perfect, but at least we know have something that allows us to not bother with iwlist.

Btw. I named it WPA2 on purpose since not everybody using iw will understand that it is actually named RSN. In the end it is your call, just to let you know why I did it. One of my first patches actually did use RSN in it :)

Regards

Marcel

#7 lore raw

Hi,

Quoted 7 lines
Quoted 4 lines

I've done that now, and pushed it out, but I think the RSN IE printing you added (and I renamed to print_rsn_ie) is missing some input length validation, could you take a look at that please? In particular it takes the count from the IE, but doesn't verify it fits, I think?

I can have a look at it since it is by no means perfect, but at least we know have something that allows us to not bother with iwlist.

Certainly. I just don't want to have it crash on an AP that advertises completely bogus information :)

Quoted 4 lines

Btw. I named it WPA2 on purpose since not everybody using iw will understand that it is actually named RSN. In the end it is your call, just to let you know why I did it. One of my first patches actually did use RSN in it :)

:)
I was going to call it RSN/WPA2 but that was too long for nice displaying. Now that I think about it though, I think RSN is more correct because WPA2 certification doesn't include any 11w things which this will include. I suspect something like WPA3 will be done for that. Who knows :)

johannes

#8 lore raw

Hi Johannes,

Quoted 10 lines
Quoted 7 lines
Quoted 4 lines

I've done that now, and pushed it out, but I think the RSN IE printing you added (and I renamed to print_rsn_ie) is missing some input length validation, could you take a look at that please? In particular it takes the count from the IE, but doesn't verify it fits, I think?

I can have a look at it since it is by no means perfect, but at least we know have something that allows us to not bother with iwlist.

Certainly. I just don't want to have it crash on an AP that advertises completely bogus information :)

if nobody beats me to it, then I will fix this. Just have to deal with some other stuff first.

Quoted 11 lines
Quoted 4 lines

Btw. I named it WPA2 on purpose since not everybody using iw will understand that it is actually named RSN. In the end it is your call, just to let you know why I did it. One of my first patches actually did use RSN in it :)

:)
I was going to call it RSN/WPA2 but that was too long for nice displaying. Now that I think about it though, I think RSN is more correct because WPA2 certification doesn't include any 11w things which this will include. I suspect something like WPA3 will be done for that. Who knows :)

tried the RSN/WPA2 too, but it is too long :)

On a different note, what is your preferred coding style for loops?

	for (i = 0; i < len; i++) {

or

	for(i=0; i<len-3; i++)

I was using the latter one since I assumed that is what you want (it is not my preferred one). Then I realized that you are using both or that something slipped in. Do you want me to send a patch to fix this?

Regards

Marcel

#9 lore raw
Quoted 5 lines · Marcel Holtmann
Quoted 2 lines

Certainly. I just don't want to have it crash on an AP that advertises completely bogus information :)

if nobody beats me to it, then I will fix this. Just have to deal with some other stuff first.

Alright, thanks. Not too pressing I guess. Just should be done before I cut the next release.

Quoted 6 lines

tried the RSN/WPA2 too, but it is too long :)

On a different note, what is your preferred coding style for loops?

	for (i = 0; i < len; i++) {
Quoted 7 lines

or

	for(i=0; i<len-3; i++)

I was using the latter one since I assumed that is what you want (it is not my preferred one). Then I realized that you are using both or that something slipped in. Do you want me to send a patch to fix this?

Oh, well, I prefer the former, but some other people put in the latter, and I didn't really care too much. No need to send a patch though.

johannes

#10 lore raw

Hi Johannes,

Quoted 6 lines
Quoted 6 lines

On a different note, what is your preferred coding style for loops?

	for (i = 0; i < len; i++) {

or

Quoted 8 lines
Quoted 5 lines
	for(i=0; i<len-3; i++)

I was using the latter one since I assumed that is what you want (it is not my preferred one). Then I realized that you are using both or that something slipped in. Do you want me to send a patch to fix this?

Oh, well, I prefer the former, but some other people put in the latter, and I didn't really care too much. No need to send a patch though.

I will sent a patch since I prefer the former too, but assumed you liked that latter one and that was my mistake in cluttering the code with more mixed coding style than needed.

Regards

Marcel

#11 lore raw
In reply to Johannes Berg

Hi Johannes,

Quoted 9 lines

scan.c static void print_ign(unsigned char type, unsigned char len, unsigned char *data · line 98

Quoted 6 lines
scan.c +5 -0
+static void print_country(unsigned char type, unsigned char len, unsigned char *data)
+{
+	int i;
+
+	printf("\tCountry: %.*s data:", 3, data);

IIRC the third characters there is a space if it's in- and outdoor, it might make sense to parse it instead of just printing it?

that explains why I have seen USI here and couldn't make sense of it and just declared that AP broken :)

Regards

Marcel

#12 lore raw
Quoted 6 lines
Quoted 3 lines

IIRC the third characters there is a space if it's in- and outdoor, it might make sense to parse it instead of just printing it?

that explains why I have seen USI here and couldn't make sense of it and just declared that AP broken :)

Heh, no, I think only "I", "O" and " " are valid there as the third character :)

johannes

#13 lore raw

Hi Johannes,

Quoted 8 lines
Quoted 5 lines
Quoted 2 lines

IIRC the third characters there is a space if it's in- and outdoor, it might make sense to parse it instead of just printing it?

that explains why I have seen USI here and couldn't make sense of it and just declared that AP broken :)

Heh, no, I think only "I", "O" and " " are valid there as the third character :)

added decoding for I and O to it :)

Regards

Marcel