By deleting three redundant functions. The spinlock was related to
event queue operation. Event queue data members were also renamed
to make their function more clear.
Reported-by: Johannes Berg <[email protected]>
Reviewed-by: Arend van Spriel <[email protected]>
Reviewed-by: Franky Lin <[email protected]>
Signed-off-by: Franky Lin <[email protected]>
Signed-off-by: Greg Kroah-Hartman <[email protected]>
Using a block statement to scope function variables is not
common in linux kernel development. Browsed through the brcmfmac
to remove those.
Reported-by: Johannes Berg <[email protected]>
Reviewed-by: Roland Vossen <[email protected]>
Reviewed-by: Pieter-Paul Giesberts <[email protected]>
Reviewed-by: Franky (Zhenhui) Lin <[email protected]>
Signed-off-by: Franky Lin <[email protected]>
Signed-off-by: Greg Kroah-Hartman <[email protected]>
Now that the transformation of the util driver is complete,
get rid of hv_cb_utils[] and other unneeded code.
Signed-off-by: K. Y. Srinivasan <[email protected]>
Signed-off-by: Haiyang Zhang <[email protected]>
Signed-off-by: Greg Kroah-Hartman <[email protected]>
Now, properly handle util services in the util driver and eliminate code
that will not be necessary. In the current code, util services were
all handled not as other vmbus devices (net, block) but rather through
special handling (channel setup etc.). In this patch we handle all
services using the standard Linux Driver Model.
Signed-off-by: K. Y. Srinivasan <[email protected]>
Signed-off-by: Haiyang Zhang <[email protected]>
Signed-off-by: Greg Kroah-Hartman <[email protected]>
In preparation for modifying the util driver to fully conform to the
Linux Driver Model, perform some service specific init and de-init
operations in util_probe()/util_remove() as opposed to in
init_hyperv_utils()/exit_hyperv_utils() as is currently done.
Signed-off-by: K. Y. Srinivasan <[email protected]>
Signed-off-by: Haiyang Zhang <[email protected]>
Signed-off-by: Greg Kroah-Hartman <[email protected]>
... instead of using IWEVCUSTOM. Use the defined events for
michael mic failure, association request info and association
response info.
Signed-off-by: David Kilroy <[email protected]>
Signed-off-by: Greg Kroah-Hartman <[email protected]>
Since the driver is now in-kernel remove support for building
against ancient WIRELESS_EXT versions.
Signed-off-by: David Kilroy <[email protected]>
Signed-off-by: Greg Kroah-Hartman <[email protected]>
If the usermode app does an ioctl over this serial device by
using TIOCMIWAIT, then the code will wait by setting the current
task state to TASK_INTERRUPTIBLE and then calling schedule().
This will be woken up by the qt2_process_modem_status on URB
completion when the port_extra->shadowMSR is set to the new
modem status.
However, this could result in a lost wakeup scenario due to a race
in the logic in the qt2_ioctl(TIOCMIWAIT) loop and the URB completion
for new modem status in qt2_process_modem_status.
Due to this, the usermode app's task will continue to sleep despite a
change in the modem status.
Signed-off-by: Kautuk Consul <[email protected]>
Cc: stable <[email protected]>
Signed-off-by: Greg Kroah-Hartman <[email protected]>
In the first alteration, the MAX_CNTL_PKT_SIZE is the
maximum size of the control packet in ->Adapter->txctlpacket[]
which is defined in InitAdapter(). This caps the size of
kmalloc memory allocation. In the second change, this max
cap fixes a potential memory corruption bug when subsequent
memset and memcpy calls are invoked.
Signed-off-by: Kevin McKinney <[email protected]>
Signed-off-by: Greg Kroah-Hartman <[email protected]>
If IoBuffer.InputLength is zero then this will cause an Oops when
we dereference the ZERO_SIZE_PTR. Or if it's smaller than
sizeof(struct link_request) then we would get memory corruption
when we set ->PLength in CopyBufferToControlPacket().
Signed-off-by: Kevin McKinney <[email protected]>
Signed-off-by: Greg Kroah-Hartman <[email protected]>
Revert changes in definitions that were submitted on May 11 2011 and
committed on June 7 2011. No reason for the change in these values was
given in the patch comment, it broke compatibility with older
versions, and was difficult to detect by the simultaneous move of the
lines of code.
Signed-off-by: Alexander Thomas <[email protected]>
Cc: stable <[email protected]>
Signed-off-by: Greg Kroah-Hartman <[email protected]>
The source file contained some static global variables that were
needed per device handled by the driver. They have moved to the
struct brcmf_sdio_dev for that reason.
Reviewed-by: Franky (Zhenhui) Lin <[email protected]>
Reviewed-by: Roland Vossen <[email protected]>
Reviewed-by: Pieter-Paul Giesberts <[email protected]>
Signed-off-by: Franky Lin <[email protected]>
Signed-off-by: Greg Kroah-Hartman <[email protected]>
The brcmfmac had a module parameter which could be used to override
the name of the wireless net device name and it defaults to 'wlan'.
This is considered rather useless, because what is in a name.
Reported-by: Johannes Berg <[email protected]>
Reviewed-by: Roland Vossen <[email protected]>
Reviewed-by: Pieter-Paul Giesberts <[email protected]>
Reviewed-by: Franky (Zhenhui) Lin <[email protected]>
Signed-off-by: Franky Lin <[email protected]>
Signed-off-by: Greg Kroah-Hartman <[email protected]>
_brcmf_sysioc_thread is handling deferred jobs from 4 callers. Use
work queue to handle set_mac_address/set_multicast_list requests.
brcmf_add_if/brcmf_del_if never run in atomic context and can call
brcmf_op_if directly.
Reported-by: Johannes Berg <[email protected]>
Reviewed-by: Roland Vossen <[email protected]>
Reviewed-by: Arend van Spriel <[email protected]>
Signed-off-by: Franky Lin <[email protected]>
Signed-off-by: Greg Kroah-Hartman <[email protected]>
Three structures that are sent to the dongle have been annotated for correct
endianness: struct brcmf_iscan_params, brcmf_scan_params,
brcmf_cfg80211_scan_req.
Reported-by: Johannes Berg <[email protected]>
Reviewed-by: Arend van Spriel <[email protected]>
Reviewed-by: Franky Lin <[email protected]>
Signed-off-by: Franky Lin <[email protected]>
Signed-off-by: Greg Kroah-Hartman <[email protected]>
brcmf_dev_ioctl() does not provide a user space buffer to
brcmf_netdev_ioctl_priv. Hence these calls are not necessary.
Reported-by: Johannes Berg <[email protected]>
Reviewed-by: Arend van Spriel <[email protected]>
Reviewed-by: Franky Lin <[email protected]>
Signed-off-by: Franky Lin <[email protected]>
Signed-off-by: Greg Kroah-Hartman <[email protected]>
Sizeof a pointer-typed expression returns the size of the pointer, not that
of the pointed data.
The semantic patch that fixes this problem is as follows:
(http://coccinelle.lip6.fr/)
// <smpl>
@@
expression *e;
type T;
identifier f;
@@
f(...,(T)e,...,
-sizeof(e)
+sizeof(*e)
,...)
// </smpl>
Signed-off-by: Julia Lawall <[email protected]>
Signed-off-by: Greg Kroah-Hartman <[email protected]>
Fix long function prototypes and line up arguments nicely.
Signed-off-by: Leonid V. Fedorenchik <[email protected]>
Signed-off-by: Greg Kroah-Hartman <[email protected]>
Replace :? operator by equivalent if-else statement where in improves
readability. Don't add else branch if it is not needed.
Signed-off-by: Leonid V. Fedorenchik <[email protected]>
Signed-off-by: Greg Kroah-Hartman <[email protected]>
