| Age | Commit message (Collapse) | Author |
|---|
|
This closes bug #2267.
Signed-off-by: Mike Westerhof <mwester@dls.net>
|
|
A network device is supposed to disable its internals in the "close"
function. Doing so affords us protection against various races,
including the ioctl vs. rfkill conflict reported by Michael.
Signed-off-by: Werner Almesberger <werner@openmoko.org>
Reported-by: Michael Trimarch <michael@panicking.kicks-ass.org>
|
|
This is loosely based on a patch by Ivan Petrov.
Signed-off-by: Werner Almesberger <werner@openmoko.org>
Reported-by: Ivan Petrov <ivan_p@hotbox.ru>
|
|
IDs 0 and 1 are AR6002, not AR6001. This is based on a patch by
Ivan Petrov, but it differs in leaving the IDs in
include/linux/mmc/sdio_ids.h
Signed-off-by: Werner Almesberger <werner@openmoko.org>
Reported-by: Ivan Petrov <ivan_p@hotbox.ru>
|
|
e2c0650efa751a6a2220618695fa41a2a5e7d23c introduced a platform-specific
hack to check if the MMC driver and hardware handle busy signaling from
the device properly.
Since this seems to be the case, we can revert this hack now.
Signed-off-by: Werner Almesberger <werner@openmoko.org>
|
|
This patch cleans up a little after Ivan's netif_queue_stop fix.
Signed-off-by: Werner Almesberger <werner@openmoko.org>
|
|
This patch resolves the following issue:
http://lists.openmoko.org/pipermail/openmoko-kernel/2009-March/009643.html
Changed: prevent rescheduling network queue at interface opened/connected.
Removed: wake network queue at transmit complete.
Added: wake network queue at packet queue limit not reached.
Signed-off-by: Ivan Petrov <ivan_p@hotbox.ru>
|
|
... and check the return value in ar6000_activate.
[ Werner: changed ___FUNCTION__ to __func__ ]
Signed-off-by: Michael Trimarchi <michael@panicking.kicks-ass.org>
Signed-off-by: Werner Almesberger <werner@openmoko.org>
|
|
Signed-off-by: Andy Green <andy@openmoko.com>
|
|
lt-configura-1232397413
pending-tracking-hist top was MERGE-via-stable-tracking-update-gta02-default-configura-1232397413 / fdaef96a49003c3e907629757f489585ea6708ab ... parent commitmessage:
From: merge <null@invalid>
MERGE-via-stable-tracking-hist-update-gta02-default-configura
stable-tracking-hist top was update-gta02-default-configura / ae4b4dc366b578f90cd85eda9348d3f7e585670c ... parent commitmessage:
From: Werner Almesberger <werner@openmoko.org>
update GTA02 default configurations
Add the settings introduced by the previous two patches to the default
configurations.
Signed-off-by: Werner Almesberger <werner@openmoko.org>
|
|
lter-include-1232325217
pending-tracking-hist top was MERGE-via-stable-tracking-fix-touchscreen-filter-include-1232325217 / d063e8c6d85c48de80b3d158bfa98d5a97149711 ... parent commitmessage:
From: merge <null@invalid>
MERGE-via-stable-tracking-hist-fix-touchscreen-filter-include
stable-tracking-hist top was fix-touchscreen-filter-include / bb151f28fc8e8923baad96e0f3e8f0ae57af95f5 ... parent commitmessage:
From: Nelson Castillo <arhuaco@freaks-unidos.net>
Fix touchscreen filter includes
Fix #includes to make the kernel compile again.
Signed-off-by: Nelson Castillo <arhuaco@freaks-unidos.net>
|
|
-the-ar6000-1232104308
pending-tracking-hist top was MERGE-via-stable-tracking-rfkill-support-for-the-ar6000-1232104308 / 09faeb511e9e6872c79b4d1ea2f3b5fefc16f3a1 ... parent commitmessage:
From: merge <null@invalid>
MERGE-via-stable-tracking-hist-rfkill-support-for-the-ar6000-
stable-tracking-hist top was rfkill-support-for-the-ar6000- / e82f8ffa04ebcb05506c8e22268371c3c3b03173 ... parent commitmessage:
From: Werner Almesberger <werner@openmoko.org>
rfkill support for the AR6000 driver
This patch adds rfkill support to the AR6000 driver. The driver does
not directly implement an rfkill device but uses the help of a special
platform device, such that the latter can retain rfkill state when the
AR6k driver is removed (e.g., because of suspend).
If an attempt is made to bring the driver up (module load, bind, or
resume) while rfkill is blocking, only the driver's data structures
are initialized, but function activation and most of the rest of the
setup is deferred until the rfkill block is removed.
Signed-off-by: Werner Almesberger <werner@openmoko.org>
|
|
As I've mentioned before, telling the WLAN module not to sleep
with "wmiconfig -i eth0 --power maxperf" reportedly improves
network stability in some situations where otherwise disconnects
were frequent. (Sorry if this sounds rather vague. I don't have
much real data on this phenomenon. More details would be welcome.)
Paul Lever from Atheros suggested that one possible source of
trouble when not using --power maxperf could be that the module
very often takes a nap and then needs to delay new commands while
waking up, and some SDIO hosts may fail to handle the delay
properly.
The S3C2442 manual leaves it unclear whether the controller handles
this condition or not. The Linux MMC/SDIO stack explicitly polls
for busy status in MMC mode but not in SDIO mode.
I've attached a patch against andy-tracking that rather crudely
checks if the module is signalling that it's still busy while we're
about to start a new transfer, and delays in this case.
In my tests, I could not make the busy condition occur (unless I
explicitly violated the stack's synchronization), but then I hardly
see any disconnects anyway.
So it would be good if those affected by frequent loss of association
could try this patch. More precisely,
- if association with the base station is often lost, and
- "wmiconfig -i eth0 --power maxperf" improves this, then
- please apply this patch, leave WLAN power at default settings
(i.e., wmiconfig -i eth0 --power rec), and report
- if the patch had any effect on the stability of associations, and
- whether "READ WHILE BUSY !" or "WRITE WHILE BUSY !" gets
logged in dmesg.
It may make sense to temporarily put this patch into andy-tracking.
While it may be perfectly useless, it shouldn't create any new
trouble either. In case it turns out to make a difference, I'll have
to find a better place in the stack for it, so it'll be reverted
either way.
Signed-off-by: Werner Almesberger <werner@openmoko.org>
|
|
SIOCGIWSCAN can return an empty node list if there are indeed no
nodes but also if we wait too long (more than about 20 seconds)
between initiating the scan and retrieving the list of nodes.
There's nothing suspicious per se about either condition, yet the
AR6k driver returns an error code and printk's a complaint.
This is a bit excessive and can mislead user space into believing
a real error has occurred. This patch make this sort of failure
silent, like in other WLAN drivers, e.g.,
ieee80211_wx.c:ieee80211_wx_get_scan
Signed-off-by: Werner Almesberger <werner@openmoko.org>
|
|
patch
pending-tracking-hist top was MERGE-via-stable-tracking-clean-hif-warning-patch / b8c3cb615a862e2a2167127195ffbd7ff18113dd ... parent commitmessage:
From: merge <null@invalid>
MERGE-via-stable-tracking-hist-clean-hif-warning-patch
stable-tracking-hist top was clean-hif-warning-patch / 37b07c166a3bcc91a2ded54daa14412fbfb7cfd0 ... parent commitmessage:
From: Andy Green <andy@openmoko.com>
clean-hif-warning.patch
Signed-off-by: Andy Green <andy@openmoko.com>
|
|
n-insane-allocat
balaji-tracking-hist top was MERGE-via-stable-tracking-hist-ar6k-break-down-insane-allocat / 6a9c6db399ca60d371ec6b42986608e8bc4a636f ... parent commitmessage:
From: merge <null@invalid>
MERGE-via-stable-tracking-hist-ar6k-break-down-insane-allocat
stable-tracking-hist top was ar6k-break-down-insane-allocat / ac6b4c5fae9b497401eeb3e47cc21932839c18a2 ... parent commitmessage:
From: Werner Almesberger <werner@openmoko.org>
ar6k-break-down-insane-allocation.patch
The Atheros WLAN stack kmallocs almost 64kB of contiguous kernel
memory for a structure containing almost entirely buffers.
As is commonly known , this kind of large allocation has a very
high risk of failing as kernel memory fragments during the life of
a system.
This patch allocates the buffers indiviudually, thus shrinking the
structure to a size below 4kB.
Note: this is untested. These buffers are only used with Atheros'
raw interface, which none of the code we have, including wmiconfig,
even seems to know about.
This may fix bug #2133. Code follows Atheros' style, so checkpatch
hates it.
Signed-off-by: Werner Almesberger <werner@openmoko.org>
|
|
patches-tracking-via-master-s3c-hsmmc-clean
balaji-tracking-hist top was efb2d57c0e0ed62324d79d6c5793fe797c157266
|
|
|
|
This is a replacement for Atheros' HIF layer that uses the Linux SDIO
stack.
Using GPLv2, like Atheros' code this is based on.
Work in progress.
Not-Yet-Signed-off-by: Werner Almesberger <werner@openmoko.org>
|
|
Make the AR6000 WLAN driver compile after moving it outside the
Atheros SDIO stack. Note that the config option's name changes
as well.
The choice of a non-standard location (drivers/ar6000/) is
intentional. The driver is still very far from being in shape for
mainline inclusion, and the odd location should serve as an
immediate warning.
Not-Yet-Signed-off-by: Werner Almesberger <werner@openmoko.org>
|
|
Signed-off-by: Andy Green <andy@openmoko.com>
|
