linux.git/net/wireless/reg.c, branch v5.12

cfg80211: initialize reg_rule in __freq_reg_info()

2021-02-12T07:56:19+00:00

Sparse started warning on this function because we can potentially
return an uninitialized value.  The reason is that if the caller
passes a min_bw value that is higher then the last value in bws[], we
will not go into the loop and reg_rule will remain initialized.  This
cannot happen because the only caller of this function uses either 1
or 20 in min_bw, but the function will be more robust if we
pre-initialize the value.

Signed-off-by: Luca Coelho 
Link: https://lore.kernel.org/r/iwlwifi.20210204154439.6c884ea7281c.I257278d03b0c1ae0aa6631672cfa48f1a95d5996@changeid
Signed-off-by: Johannes Berg

cfg80211: avoid holding the RTNL when calling the driver

2021-01-26T10:55:50+00:00

Currently, _everything_ in cfg80211 holds the RTNL, and if you
have a slow USB device (or a few) you can get some bad lock
contention on that.

Fix that by re-adding a mutex to each wiphy/rdev as we had at
some point, so we have locking for the wireless_dev lists and
all the other things in there, and also so that drivers still
don't have to worry too much about it (they still won't get
parallel calls for a single device).

Then, we can restrict the RTNL to a few cases where we add or
remove interfaces and really need the added protection. Some
of the global list management still also uses the RTNL, since
we need to have it anyway for netdev management, but we only
hold the RTNL for very short periods of time here.

Link: https://lore.kernel.org/r/20210122161942.81df9f5e047a.I4a8e1a60b18863ea8c5e6d3a0faeafb2d45b2f40@changeid
Tested-by: Marek Szyprowski  [marvell driver issues]
Signed-off-by: Johannes Berg

cfg80211: Save the regulatory domain with a lock

2021-01-08T20:03:27+00:00

Saving the regulatory domain while setting custom regulatory domain
was done while accessing a RCU protected pointer but without any
protection.

Fix this by using RTNL while accessing the pointer.

Signed-off-by: Ilan Peer 
Reported-by: syzbot+27771d4abcd9b7a1f5d3@syzkaller.appspotmail.com
Reported-by: syzbot+db4035751c56c0079282@syzkaller.appspotmail.com
Reported-by: Hans de Goede 
Fixes: beee24695157 ("cfg80211: Save the regulatory domain when setting custom regulatory")
Signed-off-by: Luca Coelho 
Link: https://lore.kernel.org/r/iwlwifi.20210105165657.613e9a876829.Ia38d27dbebea28bf9c56d70691d243186ede70e7@changeid
Signed-off-by: Johannes Berg

cfg80211: Save the regulatory domain when setting custom regulatory

2020-12-11T11:57:24+00:00

When custom regulatory was set, only the channels setting was updated, but
the regulatory domain was not saved. Fix it by saving it.

Signed-off-by: Ilan Peer 
Signed-off-by: Luca Coelho 
Link: https://lore.kernel.org/r/iwlwifi.20201129172929.290fa5c5568a.Ic5732aa64de6ee97ae3578bd5779fc723ba489d1@changeid
Signed-off-by: Johannes Berg

net: wireless: make a const array static, makes object smaller

2020-12-11T11:50:02+00:00

Don't populate the const array bws on the stack but instead it
static. Makes the object code smaller by 80 bytes:

Before:
   text	   data	    bss	    dec	    hex	filename
  85694	  16865	   1216	 103775	  1955f	./net/wireless/reg.o

After:
   text	   data	    bss	    dec	    hex	filename
  85518	  16961	   1216	 103695	  1950f	./net/wireless/reg.o

(gcc version 10.2.0)

Signed-off-by: Colin Ian King 
Link: https://lore.kernel.org/r/20201116181636.362729-1-colin.king@canonical.com
Signed-off-by: Johannes Berg

cfg80211: regulatory: Fix inconsistent format argument

2020-10-30T09:06:56+00:00

Fix follow warning:
[net/wireless/reg.c:3619]: (warning) %d in format string (no. 2)
requires 'int' but the argument type is 'unsigned int'.

Reported-by: Hulk Robot 
Signed-off-by: Ye Bin 
Link: https://lore.kernel.org/r/20201009070215.63695-1-yebin10@huawei.com
Signed-off-by: Johannes Berg

cfg80211: regulatory: remove a bogus initialization

2020-09-28T12:20:58+00:00

The the __freq_reg_info() never returns NULL and the callers don't check
for NULL.  This initialization to set "reg_rule = NULL;" is just there
to make GCC happy but it's not required in current GCCs.

The problem is that Smatch sees the initialization and concludes that
this function can return NULL so it complains that the callers are not
checking for it.

Smatch used to be able to parse this correctly but we recently changed
the code from:

-       for (bw = MHZ_TO_KHZ(20); bw >= min_bw; bw = bw / 2) {
+       for (bw = MHZ_TO_KHZ(bws[i]); bw >= min_bw; bw = MHZ_TO_KHZ(bws[i--])) {

Originally Smatch used to understand that this code always iterates
through the loop once, but the change from "MHZ_TO_KHZ(20)" to
"MHZ_TO_KHZ(bws[i])" is too complicated for Smatch.

Signed-off-by: Dan Carpenter 
Link: https://lore.kernel.org/r/20200923084203.GC1454948@mwanda
Signed-off-by: Johannes Berg

cfg80211: regulatory: handle S1G channels

2020-09-18T10:29:23+00:00

S1G channels have a minimum bandwidth of 1Mhz, and there
is a 1:1 mapping of allowed bandwidth to channel number.

Signed-off-by: Thomas Pedersen 
Link: https://lore.kernel.org/r/20200908190323.15814-4-thomas@adapt-ip.com
Signed-off-by: Johannes Berg

Merge git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net

2020-09-05T04:28:59+00:00

We got slightly different patches removing a double word
in a comment in net/ipv4/raw.c - picked the version from net.

Simple conflict in drivers/net/ethernet/ibm/ibmvnic.c. Use cached
values instead of VNIC login response buffer (following what
commit 507ebe6444a4 ("ibmvnic: Fix use-after-free of VNIC login
response buffer") did).

Signed-off-by: Jakub Kicinski

cfg80211: add helper fn for adjacent rule channels

2020-08-27T09:27:17+00:00

Some usable channels are located in the union of adjacent
regulatory rules, for example channel 144 in Germany.

Enable them, by also checking if a channel spans two adjacent
regulatory rules/frequency ranges.

All flags involved are disabling things, therefore we can build
the maximum by or-ing them together. Furthermore, take the maximum
of DFS CAC time values and the minimum of allowed power of both
adjacent channels in order to comply with both regulatory rules at
the same time.

Signed-off-by: Markus Theil 
Link: https://lore.kernel.org/r/20200803144353.305538-2-markus.theil@tu-ilmenau.de
[remove unrelated comment changes]
Signed-off-by: Johannes Berg