Driver now flushes AMPDU packets for a specified station on Mac80211 calling
wl_ops_ampdu_action(IEEE80211_AMPDU_TX_STOP). Not all AMPDU packets are flushed
yet: there can still be AMPDU packets pending in hardware (DMA). That is the
subject of the next patch in this series.
Signed-off-by: Roland Vossen <rvossen@broadcom.com> Reviewed-by: Arend van Spriel <arend@broadcom.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Roland Vossen [Thu, 10 Mar 2011 10:35:07 +0000 (11:35 +0100)]
staging: brcm80211: bugfix for control.sta NULL ptr dereference
Mac80211 can transmit packets where the control.sta field is NULL.
The driver dereferenced this. Bugfix was to only dereference a non NULL
ieee80211_sta pointer.
Signed-off-by: Roland Vossen <rvossen@broadcom.com> Reviewed-by: Arend van Spriel <arend@broadcom.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Roland Vossen [Thu, 10 Mar 2011 10:35:06 +0000 (11:35 +0100)]
staging: brcm80211: bugfix for NULL scb ptr dereference
The driver uses a struct called 'scb', this struct is primarily used for AMPDU
functionality and is embedded in struct ieee80211_sta. To increase driver
robustness, the case in which this scb pointer is NULL is now handled graceful.
This paves the way for the next patch in this series.
Signed-off-by: Roland Vossen <rvossen@broadcom.com> Reviewed-by: Arend van Spriel <arend@broadcom.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Randy Dunlap [Fri, 11 Mar 2011 00:48:36 +0000 (16:48 -0800)]
usb-storage: ene_ub6250 depends on USB_STORAGE
Fix ene_ub6250 build: it uses usb_storage driver interfaces, so it
should depend on USB_STORAGE.
ene_ub6250.c:(.text+0x14ff19): undefined reference to `usb_stor_reset_resume'
ene_ub6250.c:(.text+0x14ffb1): undefined reference to `usb_stor_bulk_transfer_buf'
ene_ub6250.c:(.text+0x14ffdd): undefined reference to `usb_stor_bulk_srb'
ene_ub6250.c:(.text+0x14fff1): undefined reference to `usb_stor_bulk_transfer_sg'
ene_ub6250.c:(.text+0x1503dd): undefined reference to `usb_stor_set_xfer_buf'
ene_ub6250.c:(.text+0x15048e): undefined reference to `usb_stor_access_xfer_buf'
ene_ub6250.c:(.text+0x150723): undefined reference to `usb_stor_probe1'
ene_ub6250.c:(.text+0x150795): undefined reference to `usb_stor_probe2'
ene_ub6250.c:(.text+0x1507af): undefined reference to `usb_stor_disconnect'
drivers/built-in.o:(.data+0x10224): undefined reference to `usb_stor_suspend'
drivers/built-in.o:(.data+0x10230): undefined reference to `usb_stor_pre_reset'
drivers/built-in.o:(.data+0x10234): undefined reference to `usb_stor_post_reset'
Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
The offscreen memory area currently conflicts with the video memory
exported to the framebuffer layer. The driver does not utilize offscreen
memory, so the functionality can be deleted.
The patch also eliminates the one last memory leak when the driver
is unloaded.
Signed-off-by: Aaro Koskinen <aaro.koskinen@iki.fi> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
The driver does not utilize HW command queue in any way, so the code
can be dropped. The support for the default mode (MMIO) and AGP have
been disabled already anyway.
Signed-off-by: Aaro Koskinen <aaro.koskinen@iki.fi> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
staging/hv: add missing include causing build error
This fixes following build error:
In file included from drivers/staging/hv/channel.h:28,
from drivers/staging/hv/vmbus_private.h:30,
from drivers/staging/hv/hv.c:28:
drivers/staging/hv/channel_mgmt.h:234: error: field ‘work’ has incomplete type
wwang [Tue, 8 Mar 2011 07:22:14 +0000 (15:22 +0800)]
staging: rts_pstor: MSXC card power class
1, Initialize chip->ms_power_class_en in rtsx_init_options;
2, In reset_ms_pro, set different initial value of change_power_class
according to chip->ms_power_class_en.
The previous implementation flawed, in case some channels are not enabled.
The sorted array would then include channel information of disabled channels,
And misses the enabled ones, when the count is reached.
More troublesome, the loop would not even terminate.
The fix is twofold:
First we skip channels that are not enabled.
Then we use a tested bubble sort algorithm to sort the array.
Since we already allocated exactly the number of bytes we need.
We can exercise bubble sort on the original memory.
In all cases I've seen, the array is already sorted, so this sort
terminates immediately.
Changes since V1:
Fix coding style issues.
Signed-off-by: Michael Hennerich <michael.hennerich@analog.com> Acked-by: Jonathan Cameron <jic23@cam.ac.uk> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
This driver did not conform with the IIO ABI for such devices.
Also the sysfs files that this driver adds were not complete and
partially un-documented.
Update and document ABI
Change License notice, stick to GPL-v2.
Fix indention style
Add option to specify external reference voltage via the regulator framework.
Add mandatory name attribute
Add mandatory out_scale attribute
Changes since V1:
Refine outY_powerdown_mode description
Remove bonus white line
Signed-off-by: Michael Hennerich <michael.hennerich@analog.com> Acked-by: Jonathan Cameron <jic23@cam.ac.uk> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Consolidate all device related state in struct hv_device by
moving the device field from struct vm_device to
struct hv_device. As part of this, also get rid of struct
vm_device since the consolidation is complete.
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com> Signed-off-by: Abhishek Kane <v-abkane@microsoft.com> Signed-off-by: Haiyang Zhang <haiyangz@microsoft.com> Signed-off-by: Hank Janssen <hjanssen@microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Both device abstractions: vm_device and hv_device maintain state
to reperesent the device instance (and they refer to them by different
names - device_id in vm_device and dev_instance in hv_device).
In preparation for consolidating all device state in
struct hv_device; eliminate device_id from struct vm_device.
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com> Signed-off-by: Abhishek Kane <v-abkane@microsoft.com> Signed-off-by: Haiyang Zhang <haiyangz@microsoft.com> Signed-off-by: Hank Janssen <hjanssen@microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Both device abstractions: vm_device and hv_device maintain state
to reperesent the device type (and they refer to them by different
names - class_id in vm_device and dev_type in hv_device).
In preparation for consolidating all device state in
struct hv_device; eliminate class_id from struct vm_device.
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com> Signed-off-by: Abhishek Kane <v-abkane@microsoft.com> Signed-off-by: Haiyang Zhang <haiyangz@microsoft.com> Signed-off-by: Hank Janssen <hjanssen@microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
We need to move the following elements from struct driver_context:
class_id and driver in one step. As part of this operation get rid of
the struct driver_context. With this patch we will have
consolidated all driver state into one data structure:
struct hv_driver.
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com> Signed-off-by: Abhishek Kane <v-abkane@microsoft.com> Signed-off-by: Haiyang Zhang <haiyangz@microsoft.com> Signed-off-by: Hank Janssen <hjanssen@microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Staging: hv: Change the signature for vmbus_child_driver_unregister
In preperation for moving the element driver from the
struct driver_context to struct hv_driver, change the
signature for the function vmbus_child_driver_unregister()
to take a pointer to struct device_driver.
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com> Signed-off-by: Abhishek Kane <v-abkane@microsoft.com> Signed-off-by: Haiyang Zhang <haiyangz@microsoft.com> Signed-off-by: Hank Janssen <hjanssen@microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Staging: hv: Change the signature for vmbus_child_driver_register
In preparation for moving the element driver from the
struct driver_context to struct hv_driver, change the
signature for the function vmbus_child_driver_register()
to take a pointer to struct device_driver.
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com> Signed-off-by: Abhishek Kane <v-abkane@microsoft.com> Signed-off-by: Haiyang Zhang <haiyangz@microsoft.com> Signed-off-by: Hank Janssen <hjanssen@microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Staging: hv: Use generic device_driver shutdown function
In preparation for moving all the state from struct driver_context
to struct hv_driver, eliminate the shutdown() function from
struct driver_context and use generic device_driver shutdown()
function.
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com> Signed-off-by: Abhishek Kane <v-abkane@microsoft.com> Signed-off-by: Haiyang Zhang <haiyangz@microsoft.com> Signed-off-by: Hank Janssen <hjanssen@microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Staging: hv: Use generic device_driver remove function
In preparation for moving all the state from struct driver_context
to struct hv_driver, eliminate the remove() function from
struct driver_context and use generic device_driver remove()
function.
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com> Signed-off-by: Abhishek Kane <v-abkane@microsoft.com> Signed-off-by: Haiyang Zhang <haiyangz@microsoft.com> Signed-off-by: Hank Janssen <hjanssen@microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Staging: hv: Use generic device_driver probe function
In preparation for moving all the state from struct driver_context
to struct hv_driver, eliminate the probe() function from
struct driver_context and use generic device_driver probe
function.
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com> Signed-off-by: Abhishek Kane <v-abkane@microsoft.com> Signed-off-by: Haiyang Zhang <haiyangz@microsoft.com> Signed-off-by: Hank Janssen <hjanssen@microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Tomas Winkler [Sun, 6 Mar 2011 08:59:03 +0000 (10:59 +0200)]
staging/easycap: reduce code nesting in easycap_sound.c
Reshuffle error handling to reduce indentation nesting
This reduce number of lines exceeding 80 characters
from 41 to 15
use:
if (error)
(return, goto, continue)
CODE
instead of:
if (good)
<CODE>
else
<EXCEPTION HANDLING>
Cc: Dan Carpenter <error27@gmail.com> Cc: Mike Thomas <rmthomas@sciolus.org> Signed-off-by: Tomas Winkler <tomas.winkler@intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Dowan Kim [Sat, 5 Mar 2011 01:47:43 +0000 (17:47 -0800)]
staging: brcm80211: FIX for bug that prevents system from entering suspend state
The attempt to enter to suspend mode can be hindered when the network
interface is disabled.
When system enters the suspend mode with the network interface
disabled, network layer calls ifdown() followed by cfg80211 layer
calling wl_cfg80211_suspend() which is registered as suspend handler
for cfg80211 layer.
ifdown() call ultimately funnels down to __wl_cfg80211_down() call
where WL_STATUS_READY bit is cleared via call to
"clear_bit(WL_STATUS_READY, &wl->status)"
But CHECK_SYS_UP()checks WL_STATUS_READY bit thinking it's not ready
and returns -EIO from suspend handler which intern prevents entering
into system suspend state
CHECK_SYS_UP() is mainly used in the code path where upper layer would
request certain wifi related activity to be performed by the firmware,
where this calls helps to make sure our firmware would be in ready
state to respond to those requests
But in the case of wl_cfg80211_suspend() code path there is no need to
check for firmware status for any reason
Signed-off-by: Dowan Kim <dowan@broadcom.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
wwang [Fri, 4 Mar 2011 02:56:36 +0000 (10:56 +0800)]
staging: rts_pstor: fix a bug that a greenhouse sd card can't be recognized
A greenhouse sd card can't be recognized using rts5209.
To fix this bug, these modifications are applied:
1, Move some codes which clear sd internal variables from sd_init_type to
sd_prepare_reset. So sd_init_type is useless any more and is removed
entirely;
2, If a sd card can't pass sd3.0 mode, the action of tunning phase should be
avoided when retrying sd2.0 mode.